[Federal Register Volume 89, Number 45 (Wednesday, March 6, 2024)]
[Rules and Regulations]
[Pages 16202-16406]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2024-02637]
[[Page 16201]]
Vol. 89
Wednesday,
No. 45
March 6, 2024
Part III
Environmental Protection Agency
-----------------------------------------------------------------------
40 CFR Parts 50, 53, and 58
Reconsideration of the National Ambient Air Quality Standards for
Particulate Matter; Final Rule
Federal Register / Vol. 89 , No. 45 / Wednesday, March 6, 2024 /
Rules and Regulations
[[Page 16202]]
-----------------------------------------------------------------------
ENVIRONMENTAL PROTECTION AGENCY
40 CFR Parts 50, 53, and 58
[EPA-HQ-OAR-2015-0072; FRL-8635-02-OAR]
RIN 2060-AV52
Reconsideration of the National Ambient Air Quality Standards for
Particulate Matter
AGENCY: Environmental Protection Agency (EPA).
ACTION: Final rule.
-----------------------------------------------------------------------
SUMMARY: Based on the Environmental Protection Agency's (EPA's)
reconsideration of the air quality criteria and the national ambient
air quality standards (NAAQS) for particulate matter (PM), the EPA is
revising the primary annual PM2.5 standard by lowering the
level from 12.0 [micro]g/m\3\ to 9.0 [micro]g/m\3\. The Agency is
retaining the current primary 24-hour PM2.5 standard and the
primary 24-hour PM10 standard. The Agency also is not
changing the secondary 24-hour PM2.5 standard, secondary
annual PM2.5 standard, and secondary 24-hour PM10
standard at this time. The EPA is also finalizing revisions to other
key aspects related to the PM NAAQS, including revisions to the Air
Quality Index (AQI) and monitoring requirements for the PM NAAQS.
DATES: This final rule is effective May 6, 2024.
ADDRESSES: The EPA has established a docket for this action under
Docket ID No. EPA-HQ-OAR-2015-0072. All documents in the docket are
listed on the https://www.regulations.gov website. Although listed in
the index, some information is not publicly available, e.g., CBI or
other information whose disclosure is restricted by statute. Certain
other material, such as copyrighted material, is not placed on the
internet and will be publicly available only in hard copy form.
Publicly available docket materials are available electronically
through https://www.regulations.gov.
FOR FURTHER INFORMATION CONTACT: Dr. Lars Perlmutt, Health and
Environmental Impacts Division, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Mail Code C539-04,
Research Triangle Park, NC 27711; telephone: (919) 541-3037; fax: (919)
541-5315; email: [email protected].
SUPPLEMENTARY INFORMATION:
Table of Contents
The following topics are discussed in this preamble:
Executive Summary
I. Background
A. Legislative Requirements
B. Related PM Control Programs
C. Review of the Air Quality Criteria and Standards for
Particulate Matter
1. Reviews Completed in 1971 and 1987
2. Review Completed in 1997
3. Review Completed in 2006
4. Review Completed in 2012
5. Review Initiated in 2014
a. 2020 Proposed and Final Decisions
b. Reconsideration of the 2020 PM NAAQS Final Action
D. Air Quality Information
1. Distribution of Particle Size in Ambient Air
2. Sources and Emissions Contributing to PM in the Ambient Air
3. Monitoring of Ambient PM
4. Ambient Concentrations and Trends
a. PM2.5 Mass
b. PM2.5 Components
c. PM10
d. PM10-2.5
e. UFP
5. Characterizing Ambient PM2.5 Concentrations for
Exposure
a. Predicted Ambient PM2.5 and Exposure Based on
Monitored Data
b. Comparison of PM2.5 Fields in Estimating Exposure
and Relative to Design Values
6. Background PM
II. Rationale for Decisions on the Primary PM2.5
Standards
A. Introduction
1. Background on the Current Standards
2. Overview of the Health Effects Evidence
a. Nature of Effects
i. Mortality
ii. Cardiovascular Effects
iii. Respiratory Effects
iv. Cancer
v. Nervous System Effects
vi. Other Effects
b. Public Health Implications and At-Risk Populations
c. PM2.5 Concentrations in Key Studies Reporting
Health Effects
i. PM2.5 Exposure Concentrations Evaluated in
Experimental Studies
ii. Ambient PM2.5 Concentrations in Locations of
Epidemiologic Studies
d. Uncertainties in the Health Effects Evidence
3. Summary of Exposure and Risk Estimates
a. Key Design Aspects
b. Key Limitations and Uncertainties
c. Summary of Risk Estimates
B. Conclusions on the Primary PM2.5 Standards
1. CASAC Advice
2. Basis for the Proposed Decision
3. Comments on the Proposed Decision
4. Administrator's Conclusions
C. Decisions on the Primary PM2.5 Standards
III. Rationale for Decisions on the Primary PM10 Standard
A. Introduction
1. Background on the Current Standard
2. Overview of Health Effects Evidence
a. Nature of Effects
i. Mortality
ii. Cardiovascular Effects
iii. Respiratory Effects
iv. Cancer
v. Metabolic Effects
vi. Nervous System Effects
B. Conclusions on the Primary PM10 Standard
1. CASAC Advice
2. Basis for the Proposed Decision
3. Comments on the Proposed Decision
4. Administrator's Conclusions
C. Decisions on the Primary PM10 Standard
IV. Communication of Public Health
A. Air Quality Index Overview
B. Air Quality Index Category Breakpoints for PM2.5
1. Summary of Proposed Revisions
a. Air Quality Index Values of 50, 100, and 150
b. Air Quality Index Values of 200 and Above
2. Summary of Significant Comments on Proposed Revisions
a. Air Quality Index Values of 50, 100, and 150
b. Air Quality Index Values of 200 and Above
c. Other Comments
3. Summary of Final Revisions
C. Air Quality Index Category Breakpoints for PM10
D. Air Quality Index Reporting
1. Summary of Proposed Revisions
2. Summary of Significant Comments on Proposed Revisions
3. Summary of Final Revisions
V. Rationale for Decisions on the Secondary PM Standards
A. Introduction
1. Background on the Current Standards
a. Non-Visibility Effects
b. Visibility Effects
2. Overview of Welfare Effects Evidence
a. Nature of Effects
i. Visibility
ii. Climate
iii. Materials
3. Summary of Air Quality and Quantitative Information
a. Visibility Effects
i. Target Level of Protection in Terms of a PM2.5
Visibility Index
ii. Relationship Between the PM2.5 Visibility Index
and the Current Secondary 24-Hour PM2.5 Standard
b. Non-Visibility Effects
B. Conclusions on the Secondary PM Standards
1. CASAC Advice
2. Basis for the Proposed Decision
3. Comments on the Proposed Decision
4. Administrator's Conclusions
C. Decisions on the Secondary PM Standards
VI. Interpretation of the NAAQS for PM
A. Amendments to Appendix K: Interpretation of the NAAQS for
Particulate Matter
B. Amendments to Appendix N: Interpretation of the NAAQS for
PM2.5
VII. Amendments to Ambient Monitoring and Quality Assurance
Requirements
[[Page 16203]]
A. Amendment to 40 CFR Part 50 (Appendix L): Reference Method
for the Determination of Fine Particulate Matter as PM2.5
in the Atmosphere--Addition of the Tisch Cyclone as an Approved
Second Stage Separator
B. Issues Related to 40 CFR Part 53 (Reference and Equivalent
Methods)
C. Changes to 40 CFR Part 58 (Ambient Air Quality Surveillance)
D. Incorporating Data From Next-Generation Technologies
VIII. Clean Air Act Implementation Requirements for the Revised
Primary Annual PM2.5 NAAQS
A. Designation of Areas
B. Section 110(a)(1) and (2) Infrastructure SIP Requirements
C. Implementing the Revised Primary Annual PM2.5
NAAQS in Nonattainment Areas
D. Implementing the Primary and Secondary PM10 NAAQS
E. Prevention of Significant Deterioration and Nonattainment New
Source Review Programs for the Revised Primary Annual
PM2.5 NAAQS
F. Transportation Conformity Program
G. General Conformity Program
IX. Statutory and Executive Order Reviews
A. Executive Order 12866: Regulatory Planning and Review and
Executive Order 14094: Modernizing Regulatory Review
B. Paperwork Reduction Act (PRA)
C. Regulatory Flexibility Act (RFA)
D. Unfunded Mandates Reform Act (UMRA)
E. Executive Order 13132: Federalism
F. Executive Order 13175: Consultation and Coordination With
Indian Tribal Governments
G. Executive Order 13045: Protection of Children From
Environmental Health and Safety Risks
H. Executive Order 13211: Actions Concerning Regulations That
Significantly Affect Energy Supply, Distribution or Use
I. National Technology Transfer and Advancement Act (NTTAA)
J. Executive Order 12898: Federal Actions To Address
Environmental Justice in Minority Populations and Low-Income
Populations and Executive Order 14096: Revitalizing Our Nation's
Commitment to Environmental Justice for All
K. Congressional Review Act (CRA)
References
Executive Summary
This document presents the Administrator's final decisions for the
reconsideration of the 2020 final decision on the primary (health-
based) and secondary (welfare-based) National Ambient Air Quality
Standards (NAAQS) for Particulate Matter (PM). More specifically, this
document summarizes the background and rationale for the
Administrator's final decisions to revise the primary annual
PM2.5 standard by lowering the level from 12.0 [micro]g/m\3\
to 9.0 [micro]g/m\3\; to retain the current primary 24-hour
PM2.5 standard (at a level of 35 [micro]g/m\3\); to retain
the primary 24-hour PM10 standard; and, not to change the
secondary PM standards at this time. In reaching his final decisions,
the Administrator considered the currently available scientific
evidence in the 2019 Integrated Science Assessment (2019 ISA) and the
Supplement to the 2019 ISA (ISA Supplement), quantitative and policy
analyses presented in the 2022 Policy Assessment (2022 PA), advice from
the Clean Air Scientific Advisory Committee (CASAC), and public
comments on the proposal. The EPA has established primary and secondary
standards for PM2.5, which includes particles with diameters
generally less than or equal to 2.5 [micro]m, and PM10,
which includes particles with diameters generally less than or equal to
10 [micro]m. The standards include two primary PM2.5
standards: an annual average standard, averaged over three years, with
a level of 12.0 [micro]g/m\3\, and a 24-hour standard with a 98th
percentile form, averaged over three years, and a level of 35 [micro]g/
m\3\. It also includes a primary PM10 standard with a 24-
hour averaging time, and a level of 150 [micro]g/m\3\, not to be
exceeded more than once per year on average over three years. Secondary
PM standards are set equal to the primary standards, except that the
level of the secondary annual PM2.5 standard is 15.0
[micro]g/m\3\.
The most recent of the PM NAAQS was completed in December 2020. In
that review, the EPA retained the primary and secondary NAAQS, without
revision (85 FR 82684, December 18, 2020). Following publication of the
2020 final action, several parties filed petitions for review and
petitions for reconsideration of the EPA's final decision.
In June 2021, the Agency announced its decision to reconsider the
2020 PM NAAQS final action.\1\ The EPA decided to reconsider the
December 2020 decision because the available scientific evidence and
technical information indicated that the current standards may not be
adequate to protect public health and welfare, as required by the Clean
Air Act. The EPA noted that the 2020 PA concluded that the scientific
evidence and information called into question the adequacy of the
primary PM2.5 standards and supported consideration of
revising the level of the primary annual PM2.5 standard to
below the current level of 12.0 [micro]g/m\3\ while retaining the
primary 24-hour PM2.5 standard (U.S. EPA, 2020b). The EPA
also noted that the 2020 PA concluded that the available scientific
evidence and information did not call into question the adequacy of the
primary PM10 or secondary PM standards and supported
consideration of retaining the primary PM10 standard and
secondary PM standards without revision (U.S. EPA, 2020b).
---------------------------------------------------------------------------
\1\ The press release for this announcement is available at:
https://www.epa.gov/newsreleases/epa-reexamine-health-standards-harmful-soot-previous-administration-left-unchanged.
---------------------------------------------------------------------------
The final decisions presented in this document on the primary
PM2.5 standards have been informed by key aspects of the
available health effects evidence and conclusions contained in the 2019
ISA and ISA Supplement, quantitative exposure/risk analyses and policy
evaluations presented in the 2022 PA, advice from the CASAC \2\ and
public comment received as part of this reconsideration.\3\ The health
effects evidence newly available in this reconsideration, in
conjunction with the full body of evidence critically evaluated in the
2019 ISA, supports a causal relationship between long- and short-term
exposures and mortality and cardiovascular effects, and the evidence
supports a likely to be a causal relationship between long-term
exposures and respiratory effects, nervous system effects, and cancer.
The longstanding evidence base, including animal toxicological studies,
controlled human exposure studies, and epidemiologic studies,
reaffirms, and in some cases strengthens, the conclusions from past
reviews regarding the health effects of PM2.5 exposures.
Epidemiologic studies available in this reconsideration demonstrate
generally positive, and often statistically significant,
PM2.5 health effect associations. Such studies report
associations between estimated PM2.5 exposures and non-
accidental, cardiovascular, or respiratory mortality; cardiovascular or
respiratory hospitalizations or emergency room visits; and other
mortality/morbidity outcomes (e.g., lung cancer mortality or incidence,
asthma development). The scientific evidence available in this
reconsideration, as evaluated in the 2019 ISA and ISA Supplement,
includes
[[Page 16204]]
a number of epidemiologic studies that use various methods to
characterize exposure to PM2.5 (e.g., ground-based monitors
and hybrid modeling approaches) and to evaluate associations between
health effects and lower ambient PM2.5 concentrations. There
are a number of recent epidemiologic studies that use varying study
designs that reduce uncertainties related to confounding and exposure
measurement error. The results of these analyses provide further
support for the robustness of associations between PM2.5
exposures and mortality and morbidity. Moreover, the Administrator
notes that recent epidemiologic studies strengthen support for health
effect associations at lower PM2.5 concentrations, with
these new studies finding positive and significant associations when
assessing exposure in locations and time periods with lower annual mean
and 25th percentile concentrations than those evaluated in
epidemiologic studies available at the time of previous reviews.
Additionally, the experimental evidence (i.e., animal toxicological and
controlled human exposure studies) strengthens the coherence of effects
across scientific disciplines and provides additional support for
potential biological pathways through which PM2.5 exposures
could lead to the overt population-level outcomes reported in
epidemiologic studies for the health effect categories for which a
causal relationship (i.e., short- and long-term PM2.5
exposure and mortality and cardiovascular effects) or likely to be
causal relationship (i.e., short- and long-term PM2.5
exposure and respiratory effects; and long-term PM2.5
exposure and nervous system effects and cancer) was concluded.
---------------------------------------------------------------------------
\2\ In 2021, the Administrator announced his decision to
reestablish the membership of the CASAC. The Administrator selected
seven members to serve on the chartered CASAC, and appointed a PM
CASAC panel to support the chartered CASAC's review of the draft ISA
Supplement and the draft PA as a part of this reconsideration (see
section I.C.6.b below for more information).
\3\ More information regarding the CASAC review of the draft ISA
Supplement and the draft PA, including opportunities for public
comment, can be found in the following Federal Register notices: 86
FR 54186, September 30, 2021; 86 FR 52673, September 22, 2021; 86 FR
56263, October 8, 2021; 87 FR 958, January 7, 2022.
---------------------------------------------------------------------------
The available evidence in the 2019 ISA continues to provide support
for factors that may contribute to increased risk of PM2.5-
related health effects including lifestage (children and older adults),
pre-existing diseases (cardiovascular disease and respiratory disease),
race/ethnicity, and socioeconomic status. For example, the 2019 ISA and
ISA Supplement conclude that there is strong evidence that Black and
Hispanic populations, on average, experience higher PM2.5
exposures and PM2.5-related health risks than non-Hispanic
White populations. In addition, studies evaluated in the 2019 ISA and
ISA Supplement also provide evidence indicating that communities with
lower socioeconomic status (SES), as assessed in epidemiologic studies
using indicators of SES including income and educational attainment
are, on average, exposed to higher concentrations of PM2.5
compared to higher SES communities.
The quantitative risk assessment, as well as policy considerations
in the 2022 PA, also inform the final decisions on the primary
PM2.5 standards. The risk assessment in this reconsideration
focuses on all-cause or nonaccidental mortality associated with long-
and short-term PM2.5 exposures. The primary analyses focus
on exposure and risk associated with air quality that might occur in an
area under air quality conditions that just meet the current and
potential alternative standards. The risk assessment estimates that the
current primary PM2.5 standards could allow a substantial
number of PM2.5-associated premature deaths in the United
States, and that public health improvements would be associated with
just meeting all of the alternative (more stringent) annual and 24-hour
standard levels modeled. Additionally, the results of the risk
assessment suggest that for most of the U.S., the annual standard is
the controlling standard and that revision to that standard has the
most potential to reduce PM2.5 exposure-related risk. The
analyses are summarized in this document and in the proposal and are
described in detail in the 2022 PA.
In its advice to the Administrator, in its review of the 2021 draft
PA, the CASAC concurred that the currently available health effects
evidence calls into question the adequacy of the primary annual
PM2.5 standard. With regard to the primary annual
PM2.5 standard, the majority of the CASAC concluded that the
level of the standard should be revised within the range of 8.0 to 10.0
[micro]g/m\3\, while the minority of the CASAC concluded that the
primary annual PM2.5 standard should be revised to a level
of 10.0 to 11.0 [micro]g/m\3\. With regard to the primary 24-hour
PM2.5 standard, the CASAC did not reach consensus on the
adequacy of the current standard. The majority of the CASAC concluded
that the primary 24-hour PM2.5 was not adequate and that the
level of the standard should be revised to within the range of 25 to 30
[micro]g/m\3\, while the minority of the CASAC concluded that the
standard was adequate and should be retained, without revision.
Additionally, in their review of the 2019 draft PA, the CASAC did not
reach consensus on the adequacy of the primary annual PM2.5
standard, with the minority recommending revision and the majority
recommending the standard be retained. In their review of the 2019
draft PA, the CASAC reached consensus regarding the adequacy of the
primary 24-hour PM2.5 standard, concluding that the standard
should be retained.
In considering how to revise the suite of primary PM2.5
standards to provide the requisite degree of protection, the
Administrator recognizes that the current annual standard and 24-hour
standard, together, are intended to provide public health protection
against the full distribution of short- and long-term PM2.5
exposures. Further, he recognizes that changes in PM2.5 air
quality designed to meet either the annual or the 24-hour standard
would likely result in changes to both long-term average and short-term
peak PM2.5 concentrations.
As in 2012, the Administrator concludes that the most effective way
to reduce total population risk associated with both long- and short-
term PM2.5 exposures is to set a generally controlling
annual standard, and to provide supplemental protection against the
occurrence of peak 24-hour PM2.5 concentrations by means of
a 24-hour standard set at the appropriate level. Based on the current
evidence and quantitative information, as well as consideration of
CASAC advice and public comments, the Administrator concludes that the
current primary annual PM2.5 standard is not adequate to
protect public health with an adequate margin of safety. The
Administrator notes that the CASAC was unanimous in its advice on the
2021 draft PA regarding the need to revise the annual standard. In
considering the appropriate level for a revised annual standard, the
Administrator concludes that a standard set at a level of 9.0 [micro]g/
m\3\ reflects his judgment about placing the most weight on the
strongest available evidence while appropriately weighing the
uncertainties.
With regard to the primary 24-hour PM2.5 standard, the
Administrator finds the available scientific evidence and quantitative
information to be insufficient to call into question the adequacy of
the public health protection afforded by the current 24-hour standard.
He further notes that a more stringent annual standard set at a level
of 9.0 [micro]g/m\3\ is expected to reduce both average (annual)
concentrations and peak (daily) concentrations. The Administrator also
notes that, in their review of the 2021 draft PA, the CASAC did not
reach consensus on whether revisions to the primary 24-hour
PM2.5 standard are warranted at this time. He also notes
that, in their review of the 2019 draft PA, the CASAC did reach
consensus that the primary 24-hour PM2.5 standard should be
retained. The Administrator concludes that the 24-hour standard should
be retained to
[[Page 16205]]
continue to provide requisite protection against short-term peak
PM2.5 concentrations, particularly when considered in
conjunction with the protection provided by the suite of standards and
the decision to revise the annual standard to a level of 9.0 [micro]g/
m\3\.
The primary PM10 standard is intended to provide public
health protection against health effects related to exposures to
PM10-2.5, which are particles with a diameter between 10
[micro]m and 2.5 [micro]m. The final decision to retain the current 24-
hour PM10 standard has been informed by key aspects of the
available health effects evidence and conclusions contained in the 2019
ISA, the policy evaluations presented in the 2022 PA, advice from the
CASAC and public comments. Specifically, the health effects evidence
for PM10-2.5 exposures is somewhat strengthened since past
reviews, although the strongest evidence still only provides support
for a suggestive of, but not sufficient to infer, causal relationship
with long- and short-term exposures and mortality and cardiovascular
effects, short-term exposures and respiratory effects, and long-term
exposures and cancer, nervous system effects, and metabolic effects. In
reaching his final decision on the primary PM10 standard,
the Administrator recognizes that, while the available health effects
evidence has expanded, recent studies are subject to the same types of
uncertainties that were judged to be important in previous reviews. He
also recognizes that, in their review of the 2019 draft PA and the 2021
draft PA, the CASAC generally agreed that it was reasonable to retain
the primary 24-hour PM10 standard given the available
scientific evidence, including retaining PM10 as the
indicator. He concludes that the newly available evidence does not call
into question the adequacy of the current primary PM10
standard, and retains that standard, without revision.
With respect to the secondary PM standards, this reconsideration
focuses on visibility, climate, and materials effects.\4\ The
Administrator's final decision to not change the current secondary
standards at this time has been informed by key aspects of the
currently available welfare effects evidence as well as the conclusions
contained in the 2019 ISA and ISA Supplement; quantitative analyses of
visibility impairment; policy evaluations presented in the 2022 PA;
advice from the CASAC; and public comments. Specifically, the welfare
effects evidence available in this reconsideration is consistent with
the evidence available in previous reviews and supports a causal
relationship between PM and visibility, climate, and materials effects.
With regard to visibility effects, the Administrator notes that he
judges that the evidence supports a target level of protection of 27
dv. He further notes that the results of quantitative analyses of
visibility impairment suggest that in areas that meet the current
secondary 24-hour PM2.5 standard that estimated light
extinction in terms of a 3-year visibility metric would be at or well
below the target level of protection. With regard to climate and
materials effects, while the evidence has expanded since previous
reviews, significant limitations and uncertainties remain in the
evidence. While the evidence has expanded since previous reviews, the
available scientific evidence remains insufficient to allow the
Administrator to make a reasoned judgment about what specific
standard(s) would be requisite to protect against known or anticipated
adverse effects to public welfare from PM's effects on materials damage
or climate.-In their review of the 2019 draft PA and the 2021 draft PA,
the CASAC did not recommend revising the secondary PM standards. In
considering the available evidence and quantitative information, with
its inherent uncertainties and limitations, the Administrator judges
that it is appropriate not to change the secondary PM standards at this
time.
---------------------------------------------------------------------------
\4\ Consistent with the 2016 Integrated Review Plan (U.S. EPA,
2016), other welfare effects of PM, such as ecological effects, are
being considered in the separate, on-going review of the secondary
NAAQS for oxides of nitrogen, oxides of sulfur and PM. Accordingly,
the public welfare protection provided by the secondary PM standards
against ecological effects such as those related to deposition of
nitrogen- and sulfur-containing compounds in vulnerable ecosystems
is being considered in that separate review. Thus, the
Administrator's conclusion in this reconsideration of the 2020 final
decision is focused only and specifically on the adequacy of public
welfare protection provided by the secondary PM standards from
effects related to visibility, climate, and materials and hereafter
``welfare effects'' refers to those welfare effects.
---------------------------------------------------------------------------
The final revisions to the primary annual PM2.5 NAAQS
trigger a process under which States (and Tribes, if they choose) make
recommendations to the Administrator regarding designations,
identifying areas of the country that either meet or do not meet the
new or revised PM NAAQS. Those areas that do not meet the revised PM
NAAQS will need to develop plans that demonstrate how they will meet
the standards. As part of these plans, states have the opportunity to
advance environmental justice, in this case for overburdened
communities in areas with high PM concentrations above the NAAQS, by
using the tools described in the current PM NAAQS implementation
guidance (80 FR 58010, 58136, August 25, 2016). The EPA is not making
changes to any of the current PM NAAQS implementation programs in this
final rulemaking.
On other topics, the EPA is finalizing two sets of changes to the
PM2.5 sub-index of the Air Quality Index (AQI). First, the
EPA is continuing to use the approach used in the revisions to the AQI
in 2012 (77 FR 38890, June 29, 2012) of setting the lower breakpoints
(50, 100 and 150) based on the levels of the primary annual and 24-hour
PM2.5 standards. In so doing, the EPA is revising the AQI
value of 50 to 9.0 [micro]g/m\3\ and is retaining the AQI values of 100
and 150 at 35.4 [micro]g/m\3\ and 55.4 [micro]g/m\3\, respectively.
Second, the EPA is revising the upper AQI breakpoints (200 and above),
and replacing the linear-relationship approach used in 1999 (64 FR
42530, August 4, 1999) to set these breakpoints, with an approach that
more fully considers the PM2.5 health effects evidence from
controlled human exposure and epidemiologic studies that has become
available in the last 20 years. The EPA is also revising the AQI values
of 200, 300 and 500 to 125.4 [micro]g/m\3\, 225.4 [micro]g/m\3\, and
325.4 [micro]g/m\3\, respectively. In addition, this final rule revises
the daily reporting requirement from 5 days per week to 7 days per
week, while also reformatting appendix G and providing clarifications.
With regard to monitoring-related activities, the EPA finalizes
revisions to data calculations and ambient air monitoring requirements
for PM to improve the usefulness and appropriateness of data used in
regulatory decision making and to better characterize air quality in
communities that are at increased risk of PM2.5 exposure and
health risk. These changes are found in 40 CFR part 50 (appendices K,
L, and N), part 53, and part 58 with associated appendices (A, B, C, D,
and E). These changes include addressing updates in data calculations,
approval of reference and equivalent methods, updates in quality
assurance statistical calculations to account for lower concentration
measurements, updates to support improvements in PM methods, a revision
to the PM2.5 network design to account for at-risk
populations, and updates to the Probe and Monitoring Path Siting
Criteria for NAAQS pollutants.
In setting the NAAQS, the EPA may not consider the costs of
implementing the standards. This was confirmed by the Supreme Court in
Whitman v. American Trucking Associations, 531 U.S. 457, 465-472, 475-
76 (2001), as discussed in section II.A of this document. As has
traditionally been
[[Page 16206]]
done in NAAQS rulemaking, the EPA prepared a Regulatory Impact Analysis
(RIA) to provide the public with information on the potential costs and
benefits of attaining several alternative PM2.5 standard
levels. In NAAQS rulemaking, the RIA is done for informational purposes
only, and the final decisions on the NAAQS in this rulemaking are not
based on consideration of the information or analyses in the RIA. The
RIA fulfills the requirements of Executive Orders 14094, 13563, and
12866. The RIA estimates the costs and monetized human health benefits
of attaining the revised and two alternative annual PM2.5
standard levels and one alternative 24-hour PM2.5 standard
level. Specifically, the RIA examines the revised annual standard level
of 9.0 [micro]g/m\3\ in combination with the current 24-hour standard
of 35 [micro]g/m\3\ (i.e., 9.0/35 [micro]g/m\3\), as well as the
following less and more stringent alternative standard levels: (1) An
alternative annual standard level of 10.0 [micro]g/m\3\ in combination
with the current 24-hour standard (i.e., 10.0/35 [micro]g/m\3\), (2) an
alternative annual standard level of 8.0 [micro]g/m\3\ in combination
with the current 24-hour standard (i.e., 8.0/35 [micro]g/m\3\), and (3)
an alternative 24-hour standard level of 30 [micro]g/m\3\ in
combination with an alternative annual standard level of 10 [micro]g/
m\3\ (i.e., 10.0/30 [micro]g/m\3\). The RIA presents estimates of the
costs and benefits of applying illustrative national control strategies
in 2032 after implementing existing and expected regulations and
assessing emissions reductions to meet the current annual and 24-hour
particulate matter NAAQS (12.0/35 [micro]g/m\3\).
I. Background
A. Legislative Requirements
Two sections of the Clean Air Act (CAA) govern the establishment
and revision of the NAAQS. Section 108 (42 U.S.C. 7408) directs the
Administrator to identify and list certain air pollutants and then to
issue air quality criteria for those pollutants. The Administrator is
to list those pollutants ``emissions of which, in his judgment, cause
or contribute to air pollution which may reasonably be anticipated to
endanger public health or welfare''; ``the presence of which in the
ambient air results from numerous or diverse mobile or stationary
sources''; and for which he ``plans to issue air quality criteria. . .
.'' (42 U.S.C. 7408(a)(1)). Air quality criteria are intended to
``accurately reflect the latest scientific knowledge useful in
indicating the kind and extent of all identifiable effects on public
health or welfare which may be expected from the presence of [a]
pollutant in the ambient air. . . .'' (42 U.S.C. 7408(a)(2)).
Section 109 [42 U.S.C. 7409] directs the Administrator to propose
and promulgate ``primary'' and ``secondary'' NAAQS for pollutants for
which air quality criteria are issued [42 U.S.C. 7409(a)]. Section
109(b)(1) defines primary standards as ones ``the attainment and
maintenance of which in the judgment of the Administrator, based on
such criteria and allowing an adequate margin of safety, are requisite
to protect the public health.'' \5\ Under section 109(b)(2), a
secondary standard must ``specify a level of air quality the attainment
and maintenance of which, in the judgment of the Administrator, based
on such criteria, is requisite to protect the public welfare from any
known or anticipated adverse effects associated with the presence of
[the] pollutant in the ambient air.'' \6\
---------------------------------------------------------------------------
\5\ The legislative history of section 109 indicates that a
primary standard is to be set at ``the maximum permissible ambient
air level . . . which will protect the health of any [sensitive]
group of the population,'' and that for this purpose ``reference
should be made to a representative sample of persons comprising the
sensitive group rather than to a single person in such a group.'' S.
Rep. No. 91-1196, 91st Cong., 2d Sess. 10 (1970).
\6\ Under CAA section 302(h) (42 U.S.C. 7602(h)), effects on
welfare include, but are not limited to, ``effects on soils, water,
crops, vegetation, manmade materials, animals, wildlife, weather,
visibility, and climate, damage to and deterioration of property,
and hazards to transportation, as well as effects on economic values
and on personal comfort and well-being.''
---------------------------------------------------------------------------
In setting primary and secondary standards that are ``requisite''
to protect public health and welfare, respectively, as provided in
section 109(b), the EPA's task is to establish standards that are
neither more nor less stringent than necessary. In so doing, the EPA
may not consider the costs of implementing the standards. See generally
Whitman v. American Trucking Associations, 531 U.S. 457, 465-472, 475-
76 (2001). Likewise, ``[a]ttainability and technological feasibility
are not relevant considerations in the promulgation of national ambient
air quality standards.'' American Petroleum Institute v. Costle, 665
F.2d 1176, 1185 (D.C. Cir. 1981); accord Murray Energy Corporation v.
EPA, 936 F.3d 597, 623-24 (D.C. Cir. 2019).
The requirement that primary standards provide an adequate margin
of safety was intended to address uncertainties associated with
inconclusive scientific and technical information available at the time
of standard setting. It was also intended to provide a reasonable
degree of protection against hazards that research has not yet
identified. See Lead Industries Association v. EPA, 647 F.2d 1130, 1154
(D.C. Cir. 1980); American Petroleum Institute v. Costle, 665 F.2d at
1186; Coalition of Battery Recyclers Ass'n v. EPA, 604 F.3d 613, 617-18
(D.C. Cir. 2010); Mississippi v. EPA, 744 F.3d 1334, 1353 (D.C. Cir.
2013). Both kinds of uncertainties are components of the risk
associated with pollution at levels below those at which human health
effects can be said to occur with reasonable scientific certainty.
Thus, in selecting primary standards that include an adequate margin of
safety, the Administrator is seeking not only to prevent pollution
levels that have been demonstrated to be harmful but also to prevent
lower pollutant levels that may pose an unacceptable risk of harm, even
if the risk is not precisely identified as to nature or degree. The CAA
does not require the Administrator to establish a primary NAAQS at a
zero-risk level or at background concentration levels, see Lead
Industries Ass'n v. EPA, 647 F.2d at 1156 n.51, Mississippi v. EPA, 744
F.3d at 1351, but rather at a level that reduces risk sufficiently so
as to protect public health with an adequate margin of safety.
In addressing the requirement for an adequate margin of safety, the
EPA considers such factors as the nature and severity of the health
effects involved, the size of the sensitive population(s), and the kind
and degree of uncertainties. The selection of any particular approach
to providing an adequate margin of safety is a policy choice left
specifically to the Administrator's judgment. See Lead Industries Ass'n
v. EPA, 647 F.2d at 1161-62; Mississippi v. EPA, 744 F.3d at 1353.
Section 109(d)(1) of the Act requires the review every five years
of existing air quality criteria and, if appropriate, the revision of
those criteria to reflect advances in scientific knowledge on the
effects of the pollutant on public health and welfare. Under the same
provision, the EPA is also to review every five years and, if
appropriate, revise the NAAQS, based on the revised air quality
criteria. Section 109(d)(1) also provides that the Administrator may
review and revise criteria or promulgate new standards earlier or more
frequently.
Section 109(d)(2) addresses the appointment and advisory functions
of an independent scientific review committee. Section 109(d)(2)(A)
requires the Administrator to appoint this committee, which is to be
composed of ``seven members including at least one member of the
National Academy of Sciences, one physician, and one person
representing State air
[[Page 16207]]
pollution control agencies.'' Section 109(d)(2)(B) provides that the
independent scientific review committee ``shall complete a review of
the criteria . . . and the national primary and secondary ambient air
quality standards . . . and shall recommend to the Administrator any
new . . . standards and revisions of existing criteria and standards as
may be appropriate. . . .'' Since the early 1980s, this independent
review function has been performed by the Clean Air Scientific Advisory
Committee (CASAC) of the EPA's Science Advisory Board.
As previously noted, the Supreme Court has held that section 109(b)
``unambiguously bars cost considerations from the NAAQS-setting
process.'' Whitman v. Am. Trucking Associations, 531 U.S. 457, 471
(2001). Accordingly, while some of these issues regarding which
Congress has directed the CASAC to advise the Administrator are ones
that are relevant to the standard setting process, others are not.
Issues that are not relevant to standard setting may be relevant to
implementation of the NAAQS once they are established.
B. Related PM Control Programs
States are primarily responsible for ensuring attainment and
maintenance of ambient air quality standards once the EPA has
established them. Under section 110, Part C, and Part D, Subparts 1 and
4 of the CAA, and related provisions and regulations, States are to
submit, for the EPA's approval, State implementation plans (SIPs) that
provide for the attainment and maintenance of the NAAQS for PM through
control programs directed to sources of the pollutants involved. The
States, in conjunction with the EPA, also administer the prevention of
significant deterioration of air quality program that covers these
pollutants (see 42 U.S.C. 7470-7479). In addition, Federal programs
provide for or result in nationwide reductions in emissions of PM and
its precursors under Title II of the Act, 42 U.S.C. 7521-7574, which
involves controls for motor vehicles and nonroad engines and equipment;
the new source performance standards under section 111 of the Act, 42
U.S.C. 7411; and the national emissions standards for hazardous
pollutants under section 112 of the Act, 42 U.S.C. 7412.
C. Review of the Air Quality Criteria and Standards for Particulate
Matter
1. Reviews Completed in 1971 and 1987
The EPA first established NAAQS for PM in 1971 (36 FR 8186, April
30, 1971), based on the original Air Quality Criteria Document (AQCD)
(DHEW, 1969).\7\ The Federal reference method (FRM) specified for
determining attainment of the original standards was the high-volume
sampler, which collects PM up to a nominal size of 25 to 45 [micro]m
(referred to as total suspended particulates or TSP). The primary
standards were set at 260 [micro]g/m\3\, 24-hour average, not to be
exceeded more than once per year, and 75 [micro]g/m\3\, annual
geometric mean. The secondary standards were set at 150 [micro]g/m\3\,
24-hour average, not to be exceeded more than once per year, and 60
[micro]g/m\3\, annual geometric mean.
---------------------------------------------------------------------------
\7\ Prior to the review initiated in 2007 (see below), the AQCD
provided the scientific foundation (i.e., the air quality criteria)
for the NAAQS. Beginning in that review, the Integrated Science
Assessment (ISA) has replaced the AQCD.
---------------------------------------------------------------------------
In October 1979 (44 FR 56730, October 2, 1979), the EPA announced
the first periodic review of the air quality criteria and NAAQS for PM.
Revised primary and secondary standards were promulgated in 1987 (52 FR
24634, July 1, 1987). In the 1987 decision, the EPA changed the
indicator for particles from TSP to PM10, in order to focus
on the subset of inhalable particles small enough to penetrate to the
thoracic region of the respiratory tract (including the
tracheobronchial and alveolar regions), referred to as thoracic
particles.\8\ The level of the 24-hour standards (primary and
secondary) was set at 150 [micro]g/m\3\, and the form was one expected
exceedance per year, on average over three years. The level of the
annual standards (primary and secondary) was set at 50 [micro]g/m\3\,
and the form was the annual arithmetic mean, averaged over three years.
---------------------------------------------------------------------------
\8\ PM10 refers to particles with a nominal mean
aerodynamic diameter less than or equal to 10 [micro]m. More
specifically, 10 [micro]m is the aerodynamic diameter for which the
efficiency of particle collection is 50 percent.
---------------------------------------------------------------------------
2. Review Completed in 1997
In April 1994, the EPA announced its plans for the second periodic
review of the air quality criteria and NAAQS for PM, and in 1997 the
EPA promulgated revisions to the NAAQS (62 FR 38652, July 18, 1997). In
the 1997 decision, the EPA determined that the fine and coarse
fractions of PM10 should be considered separately. This
determination was based on evidence that serious health effects were
associated with short- and long-term exposures to fine particles in
areas that met the existing PM10 standards. The EPA added
new standards, using PM2.5 as the indicator for fine
particles (with PM2.5 referring to particles with a nominal
mean aerodynamic diameter less than or equal to 2.5 [micro]m). The new
primary standards were as follows: (1) An annual standard with a level
of 15.0 [micro]g/m\3\, based on the 3-year average of annual arithmetic
mean PM2.5 concentrations from single or multiple community-
oriented monitors; \9\ and (2) a 24-hour standard with a level of 65
[micro]g/m\3\, based on the 3-year average of the 98th percentile of
24-hour PM2.5 concentrations at each monitor within an area.
Also, the EPA established a new reference method for the measurement of
PM2.5 in the ambient air and adopted rules for determining
attainment of the new standards. To continue to address the health
effects of the coarse fraction of PM10 (referred to as
thoracic coarse particles or PM10-2.5, generally including
particles with a nominal mean aerodynamic diameter greater than 2.5
[micro]m and less than or equal to 10 [micro]m), the EPA retained the
primary annual PM10 standard and revised the form of the
primary 24-hour PM10 standard to be based on the 99th
percentile of 24-hour PM10 concentrations at each monitor in
an area. The EPA revised the secondary standards by setting them equal
in all respects to the primary standards.
---------------------------------------------------------------------------
\9\ The 1997 annual PM2.5 standard was compared with
measurements made at the community-oriented monitoring site
recording the highest concentration or, if specific constraints were
met, measurements from multiple community-oriented monitoring sites
could be averaged (i.e., ``spatial averaging''). In the last review
(completed in 2012) the EPA replaced the term ``community-oriented''
monitor with the term ``area-wide'' monitor. Area-wide monitors are
those sited at the neighborhood scale or larger, as well as those
monitors sited at micro- or middle-scales that are representative of
many such locations in the same core-based statistical area (CBSA)
(78 FR 3236, January 15, 2013).
---------------------------------------------------------------------------
Following promulgation of the 1997 PM NAAQS, petitions for review
were filed by several parties, addressing a broad range of issues. In
May 1999, the U.S. Court of Appeals for the District of Columbia
Circuit (D.C. Circuit) upheld the EPA's decision to establish fine
particle standards and to regulate coarse particle pollution, but
vacated the 1997 PM10 standards, concluding that the EPA had
not provided a reasonable explanation justifying use of PM10
as an indicator for coarse particles. American Trucking Associations,
Inc. v. EPA, 175 F. 3d 1027 (D.C. Cir. 1999). Pursuant to the D.C.
Circuit's decision, the EPA removed the vacated 1997 PM10
standards, and the pre-existing 1987 PM10 standards remained
in place (65 FR 80776, December 22, 2000). The D.C. Circuit also upheld
the EPA's determination not to establish more stringent secondary
standards for fine particles to address effects on visibility. American
Trucking Associations v. EPA, 175 F. 3d at 1027.
[[Page 16208]]
The D.C. Circuit also addressed more general issues related to the
NAAQS, including issues related to the consideration of costs in
setting NAAQS and the EPA's approach to establishing the levels of
NAAQS. Regarding the cost issue, the court reaffirmed prior rulings
holding that in setting NAAQS the EPA is ``not permitted to consider
the cost of implementing those standards.'' American Trucking
Associations v. EPA, 175 F. 3d at 1040-41. Regarding the levels of
NAAQS, the court held that the EPA's approach to establishing the level
of the standards in 1997 (i.e., both for PM and for the ozone NAAQS
promulgated on the same day) effected ``an unconstitutional delegation
of legislative authority.'' American Trucking Associations v. EPA, 175
F. 3d at 1034-40. Although the court stated that ``the factors EPA uses
in determining the degree of public health concern associated with
different levels of ozone and PM are reasonable,'' it remanded the rule
to the EPA, stating that when the EPA considers these factors for
potential non-threshold pollutants ``what EPA lacks is any determinate
criterion for drawing lines'' to determine where the standards should
be set.
The D.C. Circuit's holding on the cost and constitutional issues
were appealed to the United States Supreme Court. In February 2001, the
Supreme Court issued a unanimous decision upholding the EPA's position
on both the cost and constitutional issues. Whitman v. American
Trucking Associations, 531 U.S. 457, 464, 475-76. On the constitutional
issue, the Court held that the statutory requirement that NAAQS be
``requisite'' to protect public health with an adequate margin of
safety sufficiently guided the EPA's discretion, affirming the EPA's
approach of setting standards that are neither more nor less stringent
than necessary.
The Supreme Court remanded the case to the D.C. Circuit for
resolution of any remaining issues that had not been addressed in that
court's earlier rulings. Id. at 475-76. In a March 2002 decision, the
D.C. Circuit rejected all remaining challenges to the standards,
holding that the EPA's PM2.5 standards were reasonably
supported by the administrative record and were not ``arbitrary and
capricious.'' American Trucking Associations v. EPA, 283 F. 3d 355,
369-72 (D.C. Cir. 2002).
3. Review Completed in 2006
In October 1997, the EPA published its plans for the third periodic
review of the air quality criteria and NAAQS for PM (62 FR 55201,
October 23, 1997). After the CASAC and public review of several drafts,
the EPA's National Center for Environmental Assessment (NCEA) finalized
the AQCD in October 2004 (U.S. EPA, 2004a). The EPA's Office of Air
Quality Planning and Standards (OAQPS) finalized a Risk Assessment and
Staff Paper in December 2005 (Abt Associates, 2005; U.S. EPA,
2005).\10\ On December 20, 2005, the EPA announced its proposed
decision to revise the NAAQS for PM and solicited public comment on a
broad range of options (71 FR 2620, January 17, 2006). On September 21,
2006, the EPA announced its final decisions to revise the primary and
secondary NAAQS for PM to provide increased protection of public health
and welfare, respectively (71 FR 61144, October 17, 2006). With regard
to the primary and secondary standards for fine particles, the EPA
revised the level of the 24-hour PM2.5 standards to 35
[micro]g/m\3\, retained the level of the annual PM2.5
standards at 15.0 [micro]g/m\3\, and revised the form of the annual
PM2.5 standards by narrowing the constraints on the optional
use of spatial averaging. With regard to the primary and secondary
standards for PM10, the EPA retained the 24-hour standards,
with levels at 150 [micro]g/m\3\, and revoked the annual standards. The
then-Administrator judged that the available evidence generally did not
suggest a link between long-term exposure to existing ambient levels of
coarse particles and health or welfare effects. In addition, a new
reference method was added for the measurement of PM10-2.5
in the ambient air in order to provide a basis for approving Federal
Equivalent Methods (FEMs) and to promote the gathering of scientific
data to support future reviews of the PM NAAQS.
---------------------------------------------------------------------------
\10\ Prior to the review initiated in 2007, the Staff Paper
presented the EPA staff's considerations and conclusions regarding
the adequacy of existing NAAQS and, when appropriate, the potential
alternative standards that could be supported by the evidence and
information. More recent reviews present this information in the
Policy Assessment.
---------------------------------------------------------------------------
Several parties filed petitions for review following promulgation
of the revised PM NAAQS in 2006. On February 24, 2009, the D.C. Circuit
issued its opinion in the case American Farm Bureau Federation v. EPA,
559 F. 3d 512 (D.C. Cir. 2009). The court remanded the primary annual
PM2.5 NAAQS to the EPA because the Agency had failed to
adequately explain why the standards provided the requisite protection
from both short- and long-term exposures to fine particles, including
protection for at-risk populations. Id. at 520-27. With regard to the
standards for PM10, the court upheld the EPA's decisions to
retain the 24-hour PM10 standard to provide protection from
thoracic coarse particle exposures and to revoke the annual
PM10 standard. Id. at 533-38. With regard to the secondary
PM2.5 standards, the court remanded the standards to the EPA
because the Agency failed to adequately explain why setting the
secondary PM standards identical to the primary standards provided the
required protection for public welfare, including protection from
visibility impairment. Id. at 528-32. The EPA responded to the court's
remands as part of the next review of the PM NAAQS, which was initiated
in 2007 (discussed below).
4. Review Completed in 2012
In June 2007, the EPA initiated the fourth periodic review of the
air quality criteria and the PM NAAQS by issuing a call for information
(72 FR 35462, June 28, 2007). Based on the NAAQS review process, as
revised in 2008 and again in 2009,\11\ the EPA held science/policy
issue workshops on the primary and secondary PM NAAQS (72 FR 34003,
June 20, 2007; 72 FR 34005, June 20, 2007), and prepared and released
the planning and assessment documents that comprise the review process
(i.e., Integrated Review Plan, (IRP; U.S. EPA, 2008), Integrated
Science Assessment (ISA; U.S. EPA, 2009a), Risk and Exposure Assessment
(REA) planning documents for health and welfare (U.S. EPA, 2009b, U.S.
EPA, 2009c), a quantitative health risk assessment (U.S. EPA, 2010a)
and an urban-focused visibility assessment (U.S. EPA, 2010b), and a
Policy Assessment (PA; U.S. EPA, 2011). In June 2012, the EPA announced
its proposed decision to revise the NAAQS for PM (77 FR 38890, June 29,
2012).
---------------------------------------------------------------------------
\11\ The history of the NAAQS review process, including
revisions to the process, is discussed at https://www.epa.gov/naaqs/historical-information-naaqs-review-process.
---------------------------------------------------------------------------
In December 2012, the EPA announced its final decisions to revise
the primary NAAQS for PM to provide increased protection of public
health (78 FR 3086, January 15, 2013). With regard to primary standards
for PM2.5, the EPA revised the level of the annual
PM2.5 standard \12\ to 12.0 [micro]g/m\3\ and retained the
24-hour PM2.5 standard, with its level of 35 [micro]g/m\3\.
For the primary PM10 standard, the EPA retained the 24-hour
standard to continue to provide protection against effects associated
with short-term exposure to thoracic coarse particles (i.e.,
PM10-2.5). With regard to the secondary PM standards, the
EPA generally retained the 24-hour
[[Page 16209]]
and annual PM2.5 standards \13\ and the 24-hour
PM10 standard to address visibility and non-visibility
welfare effects.
---------------------------------------------------------------------------
\12\ The EPA also eliminated the option for spatial averaging.
\13\ Consistent with the primary standard, the EPA eliminated
the option for spatial averaging with the annual standard.
---------------------------------------------------------------------------
As with previous reviews, petitioners challenged the EPA's final
rule. Petitioners argued that the EPA acted unreasonably in revising
the level and form of the annual standard and in amending the
monitoring network provisions. On judicial review, the revised
standards and monitoring requirements were upheld in all respects. NAM
v. EPA, 750 F.3d 921 (D.C. Cir. 2014).
5. Review Initiated in 2014
In December 2014, the EPA announced the initiation of the current
periodic review of the air quality criteria for PM and of the
PM2.5 and PM10 NAAQS and issued a call for
information (79 FR 71764, December 3, 2014). On February 9 to 11, 2015,
the EPA's NCEA and OAQPS held a public workshop to inform the planning
for the review of the PM NAAQS (announced in 79 FR 71764, December 3,
2014). Workshop participants, including a wide range of external
experts as well as the EPA staff representing a variety of areas of
expertise (e.g., epidemiology, human and animal toxicology, risk/
exposure analysis, atmospheric science, visibility impairment, climate
effects), were asked to highlight significant new and emerging PM
research, and to make recommendations to the Agency regarding the
design and scope of the review. This workshop provided for a public
discussion of the key science and policy-relevant issues around which
the EPA structured the review of the PM NAAQS and of the most
meaningful new scientific information that would be available in the
review to inform understanding of these issues.
The input received at the workshop guided the EPA staff in
developing a draft IRP, which was reviewed by the CASAC Particulate
Matter Panel and discussed on public teleconferences held in May 2016
(81 FR 13362, March 14, 2016) and August 2016 (81 FR 39043, June 15,
2016). Advice from the CASAC, supplemented by the Particulate Matter
Panel, and input from the public were considered in developing the
final IRP (U.S. EPA, 2016). The final IRP discusses the approaches to
be taken in developing key scientific, technical, and policy documents
in the review and the key policy-relevant issues that frame the EPA's
consideration of whether the primary and/or secondary NAAQS for PM
should be retained or revised.
In May 2018, the then-Administrator issued a memorandum announcing
the Agency's intention to conduct the review of the PM NAAQS in such a
manner as to ensure that any necessary revisions were finalized by
December 2020 (Pruitt, 2018). Following this memo, on October 10, 2018,
the then-Administrator additionally announced that the role of
reviewing the key assessments developed as part of the ongoing review
of the PM NAAQS (i.e., drafts of the ISA and PA) would be performed by
the seven-member chartered CASAC (i.e., rather than the CASAC
Particulate Matter Panel that reviewed the draft IRP).\14\
---------------------------------------------------------------------------
\14\ Announcement available at: https://www.regulations.gov/document/EPA-HQ-OAR-2015-0072-0223.
---------------------------------------------------------------------------
The EPA released the draft ISA in October 2018 (83 FR 53471,
October 23, 2018). The draft ISA was reviewed by the chartered CASAC at
a public meeting held in Arlington, VA in December 2018 (83 FR 55529,
November 6, 2018) and was discussed on a public teleconference in March
2019 (84 FR 8523, March 8, 2019). The CASAC provided its advice on the
draft ISA in a letter to the then-Administrator dated April 11, 2019
(Cox, 2019a). The EPA addressed these comments in the final ISA, which
was released in December 2019 (U.S. EPA, 2019a).
The EPA released the draft PA in September 2019 (84 FR 47944,
September 11, 2019). The draft PA was reviewed by the chartered CASAC
and discussed in October 2019 at a public meeting held in Cary, NC.
Public comments were received via a separate public teleconference (84
FR 51555, September 30, 2019). A public meeting to discuss the
chartered CASAC letter and response to charge questions on the draft PA
was held in Cary, NC, in October 2019 (84 FR 51555, September 30,
2019), and the CASAC provided its advice on the draft PA, including its
advice on the current primary and secondary PM standards, in a letter
to the then-Administrator dated December 16, 2019 (Cox, 2019b). With
regard to the primary standards, the CASAC recommended retaining the
current 24-hour PM2.5 and PM10 standards but did
not reach consensus on the adequacy of the current annual
PM2.5 standard. Some CASAC members expressed support for
retaining the current primary annual PM2.5 standard while
other members expressed support for revising that standard in order to
increase public health protection (Cox, 2019b, p. 1 of letter). These
views are described in greater detail in the letter to the then-
Administrator (Cox, 2019b) and in the notice of final rulemaking (85 FR
82706-82707, December 18, 2020), as well as below. With regard to the
secondary standards, the CASAC recommended retaining the current
standards. In response to the CASAC's comments, the 2020 final PA
incorporated a number of changes (Cox, 2019b, U.S. EPA, 2020b), as
described in detail in section I.C.5 of the 2020 proposal document (85
FR 24100, April 30, 2020).
a. 2020 Proposed and Final Actions
On April 14, 2020, the EPA proposed to retain all of the primary
and secondary PM standards, without revision. These proposed decisions
were published in the Federal Register on April 30, 2020 (85 FR 24094,
April 30, 2020). The EPA's final decision on the PM NAAQS was published
in the Federal Register on December 18, 2020 (85 FR 82684, December 18,
2020). In the 2020 rulemaking, the EPA retained the primary and
secondary PM2.5 and PM10 standards, without
revision. The then-Administrator's rationale for his decisions is
described in more detail in section II, III, and V below, and is
briefly summarized here.
In reaching his final decision to retain the primary annual and 24-
hour PM2.5 standards, the then-Administrator considered the
available scientific evidence, quantitative information, CASAC advice,
and public comments in his supporting rationale in the 2020 final
action (85 FR 82714, December 18, 2020). In so doing, he concluded that
the available controlled human exposure studies did not provide support
for additional public health protection against exposures to peak
PM2.5 concentrations, beyond the protection provided by the
combination of the current primary annual and 24-hour PM2.5
standards. He also noted that the available epidemiologic studies did
not indicate that associations in those studies are strongly influenced
by exposures to peak concentrations in the air quality distribution and
thus did not indicate the need for additional protection against short-
term exposures to peak PM2.5 concentrations. Accordingly,
and taking into account consensus CASAC advice to retain the current
primary 24-hour PM2.5 standard, the then-Administrator
concluded the primary 24-hour PM2.5 standard should be
retained.
With respect to the annual PM2.5 standard, the then-
Administrator recognized that important uncertainties and limitations
that were present in epidemiologic studies in previous
[[Page 16210]]
reviews remained in the evidence assessed in the 2019 ISA. In
considering the epidemiologic evidence, the then-Administrator noted
that: (1) The reported mean concentration in the majority of the key
U.S. epidemiologic studies using ground-based monitoring data are above
the level of the current annual standard; (2) the mean of the reported
study means (or medians) (i.e., 13.5 [micro]g/m\3\) is above the level
of the current primary annual PM2.5 standard of 12 [micro]g/
m\3\; (3) air quality analyses show the study means to be lower than
their corresponding design by 10-20%; and (4) that these analyses must
be considered in light of uncertainties inherent in the epidemiologic
evidence. The then-Administrator further considered other available
information, including the risk assessment, accountability studies, and
controlled human exposure studies, and found that, in considering all
of the evidence together along with advice from the CASAC, the suite of
primary PM2.5 standards were requisite to protect public
health with an adequate margin of safety, and should be retained,
without revision.
With regard to the primary PM10 standard, the then-
Administrator noted that the expanded body of evidence has broadened
the range of effects that have been linked with PM10-2.5
exposures. In light of that information, as well as continued
uncertainties in the evidence and advice from the CASAC to retain the
standard, the then-Administrator judged it appropriate to retain the
primary PM10 standard to provide the requisite degree of
public health protection against PM10-2.5 exposures,
regardless of location, source of origin, or particle composition (85
FR 82725, December 18, 2020).
With regard to the secondary PM standards, the then-Administrator
concluded that there was insufficient information available to
establish any distinct secondary PM standards to address climate and
materials effects of PM. For visibility effects, he found that in the
absence of a monitoring network for direct measurement of light
extinction, a calculated light extinction indicator that utilizes the
IMPROVE algorithms continued to provide a reasonable basis for defining
a target level of protection against PM-related visibility impairment.
He further found that a visibility index with a 24-hour averaging time
was reasonable based on its stability and suitability for representing
subdaily periods, and a form based on the 3-year average of annual 90th
percentile values was reasonable based on its stability and that it
represents the median of the 20 percent worst visibility days which are
targeted under the Regional Haze program. With regard to the level of a
visibility index, the then-Administrator judged it appropriate to
establish a target level of protection of 30 dv, reflecting the upper
end of the range of visibility impairment judged to be acceptable by at
least 50% of study participants in the available public preference
studies, taking into consideration the variability, limitations and
uncertainties of the public preference studies. The then-Administrator
judged that the secondary 24-hour PM2.5 standard with its
level of 35 [micro]g/m\3\ would provide at least the target level of
protection for visual air quality of 30 dv which he judged appropriate.
Accordingly, taking into consideration the advice of the CASAC to
retain the current secondary PM standards, the then-Administrator found
the current secondary standards provide the requisite degree of
protection and that they should be retained (85 FR 82742, December 18,
2020).
Following publication of the 2020 final action, several parties
filed petitions for review and petitions for reconsideration of the
EPA's final decision. The petitions for review were filed in the D.C.
Circuit and the Court consolidated the cases.\15\ Following EPA's
decision to reconsider the 2020 final decision, the Court ordered the
consolidated cases to be held in abeyance.
---------------------------------------------------------------------------
\15\ See California v. EPA, (D.C. Cir., No. 21-2014 consolidated
with Nos. 21-1027, 21-1054).
---------------------------------------------------------------------------
b. Reconsideration of the 2020 PM NAAQS Final Action
Executive Order 13990 directed review of certain agency actions (86
FR 7037, January 25, 2021).\16\ An accompanying fact sheet provided a
non-exclusive list of agency actions that agency heads should review in
accordance with that order, including the 2020 Particulate Matter NAAQS
Decision.\17\
---------------------------------------------------------------------------
\16\ See https://www.whitehouse.gov/briefing-room/presidential-actions/2021/01/20/executive-order-protecting-public-health-and-environment-and-restoring-science-to-tackle-climate-crisis/.
\17\ See https://www.whitehouse.gov/briefing-room/statements-releases/2021/01/20/fact-sheet-list-of-agency-actions-for-review/.
---------------------------------------------------------------------------
On June 10, 2021, the Agency announced its decision to reconsider
the 2020 PM NAAQS final action because the available scientific
evidence and technical information indicate that the current standards
may not be adequate to protect public health and welfare, as required
by the Clean Air Act.\18\ The Administrator reached this decision in
part based on the fact that the EPA noted that the 2020 PA concluded
that the scientific evidence and information called into question the
adequacy of the primary annual PM2.5 standard and supported
revising the level to below the current level of 12.0 [micro]g/m\3\
while retaining the primary 24-hour PM2.5 standard (U.S.
EPA, 2020b). The EPA also noted that the 2020 PA concluded that the
available scientific evidence and information supported retaining the
primary PM10 standard and secondary PM standards without
revision (U.S. EPA, 2020b).
---------------------------------------------------------------------------
\18\ The press release for this announcement is available at:
https://www.epa.gov/newsreleases/epa-reexamine-health-standards-harmful-soot-previous-administration-left-unchanged.
---------------------------------------------------------------------------
The EPA staff conclusions detailed in the 2020 PA in combination
with the CASAC advice that informed the Administrator's decisions
regarding the 2020 final action, studies highlighted by public comments
on the 2020 proposal, and the numerous studies published since the
literature cutoff date of the 2019 ISA all informed the scope of the
reconsideration.
In its review of the 2019 draft PA, some members of the CASAC had
recommended that greater attention should be given to accountability
studies and epidemiologic studies that employ alternative methods for
confounder control (also referred to as causal inference or causal
modeling studies) in order to ``more fully account for effects of
confounding, measurement and estimation errors, model uncertainty, and
heterogeneity'' in epidemiologic studies (Cox, 2019b, p. 8 of consensus
responses). In addition, public commenters submitted a number of recent
studies published after the literature cutoff date for the 2019 ISA
that would have been considered within the scope of the 2019 ISA. While
the EPA provisionally considered these studies in responding to public
comments,\19\ it was determined that, at the time of the 2020 final
action, these studies were generally consistent with the evidence
assessed in the 2019 ISA (85 FR 82690, December 18, 2020; U.S. EPA,
2020a). As such, and consistent with previous NAAQS reviews, the EPA
concluded that the new studies did not materially change any of the
broad scientific conclusions regarding the health and welfare effects
of PM in ambient air made in the air quality criteria, and therefore,
reopening of the air quality criteria was not warranted (85 FR 82691,
December 18, 2020). However, at that time, the EPA
[[Page 16211]]
recognized that its ``provisional consideration of these studies did
not and could not provide the kind of in-depth critical review'' (85 FR
82690, December 18, 2020) that studies undergo in the development of an
ISA.
---------------------------------------------------------------------------
\19\ The list of provisionally considered studies is included in
Appendix A to the 2020 Response to Comments document (U.S. EPA,
2020a).
---------------------------------------------------------------------------
In preparing to reconsider the 2020 final decision for the PM
NAAQS, the Agency revisited the need to reopen the air quality
criteria, given the amount of time that had passed since the literature
cutoff date of the 2019 ISA (i.e., approximately January 2018) and the
volume of literature that had become available, including those studies
provisionally considered in responding to comments in 2020. In so
doing, the EPA preliminarily concluded that at least some of these
studies were likely to be relevant to its reconsideration of the air
quality criteria and the PM NAAQS and that, in considering public
comments on any proposed decisions for the reconsideration, these
studies were likely to be raised by public commenters and would
potentially warrant a reopening of the air quality criteria. For
example, on February 16, 2021, the EPA received two petitions to
reconsider the PM NAAQS. One petition objected to the EPA's provisional
consideration of studies submitted in public comments on the 2020
proposal and suggested that the provisional consideration was
inadequate because the studies could be important in determining
whether the existing standards are adequately protective. See, Petition
for Reconsideration of National Ambient Air Quality Standards for
Particulate Matter, submitted by American Lung Association, et al,
dated Feb. 16, 2020. The other petition identified a number of new
studies, including one epidemiologic study that was published after the
provisional consideration was completed that could further inform the
concern expressed by the CASAC that associations reported in
epidemiologic studies do not adequately account for ``uncontrolled
confounding and other potential sources of error and bias.'' See
Petition for Reconsideration of ``Review of the National Ambient Air
Quality Standards for Particulate Matter,'' submitted by the State of
California, dated Feb. 16, 2020. This was also an uncertainty noted by
the then-Administrator in the 2020 decision, who also recognized ``that
methodological study designs to address confounding, such as causal
inference methods, are an emerging field of study.'' Thus, the Agency
concluded it was appropriate to reconsider not only the standards but
also the air quality criteria, in light of public comments during the
2020 PM NAAQS proposal and recent studies published since the cutoff
date of the 2019 ISA, as reflected in petitions. In deciding to reopen
the air quality criteria, the Agency concluded it was reasonable to
focus on studies that were most likely to inform decisions on the
appropriate standard, but not to reassess areas which, based on the
assessment of available science published since the cutoff date of the
2019 ISA and through 2021, were judged unlikely to have new information
that would be useful for the Administrator's decision making. The
Agency accordingly announced that, in support of the reconsideration,
it would develop a supplement to the 2019 ISA and a revised PA.
The EPA also explained that the draft ISA Supplement and draft PA
would be reviewed at a public meeting by the CASAC, and the public
would have opportunities to comment on these documents during the CASAC
review process, as well as to provide input during the rulemaking
through the public comment process and public hearings on the proposed
rulemaking.
On March 31, 2021, the Administrator announced his decision to
reestablish the membership of the CASAC to ``ensure the agency received
the best possible scientific insight to support our work to protect
human health and the environment.'' \20\ Consistent with this
memorandum, a call for nominations of candidates to the EPA's chartered
CASAC was published in the Federal Register (86 FR 17146, April 1,
2021). On June 17, 2021, the Administrator announced his selection of
the seven members to serve on the chartered CASAC.21 22
Additionally, a call for nominations of candidates to a PM-specific
panel was published in the Federal Register (86 FR 33703, June 25,
2021). The members of the PM CASAC panel were announced on August 30,
2021.\23\
---------------------------------------------------------------------------
\20\ The press release for this announcement is available at:
https://www.epa.gov/newsreleases/administrator-regan-directs-epa-reset-critical-science-focused-federal-advisory.
\21\ The press release for this announcement is available at:
https://www.epa.gov/newsreleases/epa-announces-selections-charter-members-clean-air-scientific-advisory-committee.
\22\ The list of members of the chartered CASAC and their
biosketches are available at: https://casac.epa.gov/ords/sab/r/sab_apex/casac/mems?p14_committeeon=2021%20CASAC%20PM%20Panel&session=17433386035954
.
\23\ The list of members of the PM CASAC panel and their
biosketches are available at: https://casac.epa.gov/ords/sab/f?p=105:14:9979229564047:::14:P14_COMMITTEEON:2021%20CASAC%20PM%20Panel.
---------------------------------------------------------------------------
The draft ISA Supplement was released in September 2021 (U.S. EPA,
2021a; 86 FR 54186, September 30, 2021), and included a discussion of
the rationale and scope of the Supplement. As explained therein, the
ISA Supplement focuses on a thorough evaluation of some studies that
became available after the literature cutoff date of the 2019 ISA that
could either further inform the adequacy of the current PM NAAQS or
address key scientific topics that have evolved since the literature
cutoff date for the 2019 ISA. In selecting the health effects to
evaluate within the ISA Supplement, the EPA focused on health effects
for which the evidence supported a ``causal relationship'' because
those were the health effects that were most useful in informing
conclusions in the 2020 PA (U.S. EPA, 2022a, section 1.2.1).\24\
Consistent with the rationale for the focus on certain health effects,
in selecting the non-ecological welfare effects to evaluate within the
ISA Supplement, the EPA focused on the non-ecological welfare effects
for which the evidence supported a ``causal relationship'' and for
which quantitative analyses could be supported by the evidence because
those were the welfare effects that were most useful in informing
conclusions in the 2020 PA.\25\ Specifically, for non-ecological
welfare effects, the focus within the ISA Supplement is on visibility
effects. The ISA Supplement also considers recent health effects
evidence that addresses key scientific topics where the literature has
evolved since the 2020 review was completed,
[[Page 16212]]
specifically since the literature cutoff date for the 2019 ISA.\26\
---------------------------------------------------------------------------
\24\ As described in section 1.2.1 of the ISA Supplement: ``In
considering the public health protection provided by the current
primary PM2.5 standards, and the protection that could be
provided by alternatives, [the U.S. EPA, within the 2020 PM PA]
emphasized health outcomes for which the ISA determined that the
evidence supports either a `causal' or a `likely to be causal'
relationship with PM2.5 exposures'' (U.S. EPA, 2020b).
Although the 2020 PA initially focused on this broader set of
evidence, the basis of the discussion on potential alternative
standards primarily focused on health effect categories where the
2019 PM ISA concluded a `causal relationship' (i.e., short- and
long-term PM2.5 exposure and cardiovascular effects and
mortality) as reflected in Figures 3-7 and 3-8 of the 2020 PA (U.S.
EPA, 2020b).''
\25\ As described in section 1.2.1 of the ISA Supplement: ``The
2019 PM ISA concluded a `causal relationship' for each of the
welfare effects categories evaluated (i.e., visibility, climate
effects and materials effects). While the 2020 PA considered the
broader set of evidence for these effects, for climate effects and
material effects, it concluded that there remained `substantial
uncertainties with regard to the quantitative relationships with PM
concentrations and concentration patterns that limit[ed] [the]
ability to quantitatively assess the public welfare protection
provided by the standards from these effects' (U.S. EPA, 2020b).''
\26\ These key scientific topics include experimental studies
conducted at near-ambient concentrations, epidemiologic studies that
employed alternative methods for confounder control or conducted
accountability analyses, studies that assess the relationship
between PM2.5 exposure and severe acute respiratory
syndrome coronavirus 2 (SARS-CoV-2) infection and coronavirus
disease 2019 (COVID-19) death; and in accordance with recent EPA
goals on addressing environmental justice, studies that examine
disparities in PM2.5 exposure and the risk of health
effects by race/ethnicity or socioeconomic status (SES) (U.S. EPA,
2022a, section 1.2.1).
---------------------------------------------------------------------------
Building on the rationale presented in section 1.2.1, the ISA
Supplement considers peer-reviewed studies published from approximately
January 2018 through March 2021 that meet the following criteria:
Health Effects
[cir] U.S. and Canadian epidemiologic studies for health effect
categories where the 2019 ISA concluded a ``causal relationship''
(i.e., short- and long-term PM2.5 exposure and
cardiovascular effects and mortality).
[ssquf] U.S. and Canadian epidemiologic studies that employed
alternative methods for confounder control or conducted accountability
analyses (i.e., examined the effect of a policy on reducing
PM2.5 concentrations).
Welfare Effects
[cir] U.S. and Canadian studies that provide new information on
public preferences for visibility impairment and/or developed
methodologies or conducted quantitative analyses of light extinction.
Key Scientific Topics
[cir] Experimental studies (i.e., controlled human exposure and
animal toxicological) conducted at near-ambient PM2.5
concentrations experienced in the U.S.
[cir] U.S.- and Canadian-based epidemiologic studies that examined
the relationship between PM2.5 exposures and severe acute
respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and
coronavirus disease 2019 (COVID-19) death.
[cir] At-Risk Populations.
[ssquf] U.S.- and Canadian-based epidemiologic or exposure studies
examining potential disparities in either PM2.5 exposures or
the risk of health effects by race/ethnicity or socioeconomic status
(SES).
Given the narrow scope of the ISA Supplement, it is important to
recognize that the evaluation does not encompass the full
multidisciplinary evaluation presented within the 2019 ISA that would
result in weight-of-evidence conclusions on causality (i.e., causality
determinations). The ISA Supplement critically evaluates and provides
key study-specific information for those recent studies deemed to be of
greatest significance for informing preliminary conclusions on the PM
NAAQS in the context of the body of evidence and scientific conclusions
presented in the 2019 ISA.
In developing a revised PA to support the reconsideration, the EPA
considered the available scientific evidence, including the evidence
presented in the 2019 ISA and ISA Supplement. The 2022 PA considered
the quantitative and technical information presented in the 2020 PA, in
addition to new and updated analyses conducted since the 2020 final
decision. For those health and welfare effects for which the ISA
Supplement evaluated recently available studies (i.e.,
PM2.5-related health effects and visibility effects), new
updated quantitative analyses were conducted as a part of the
development of the 2022 PA. The newly available scientific and
technical information presented in the 2022 PA were considered in
reaching conclusions regarding the adequacy of the current standards
and any potential alternative standards. For those health and welfare
effects for which newly available scientific and technical information
were not evaluated (i.e., PM10-2.5-related health effects
and non-visibility welfare effects), the conclusions presented in the
2022 PA rely heavily on the information that supported the conclusions
in the 2020 PA.
The CASAC PM panel met at a virtual public meeting in November 2021
to review the draft ISA Supplement (86 FR 52673, September 22, 2021). A
virtual public meeting was then held in February 2022, and during this
meeting the chartered CASAC considered the CASAC PM panel's draft
letter to the Administrator on the draft ISA Supplement (87 FR 958,
January 7, 2022).
The chartered CASAC provided its advice on the draft ISA Supplement
in a letter to the EPA Administrator dated March 18, 2022 (Sheppard,
2022b). In its review of the draft ISA Supplement, the CASAC noted that
they found ``the Draft ISA Supplement to be a well-written,
comprehensive evaluation of the new scientific information published
since the 2019 PM ISA'' (Sheppard, 2022b, p. 2 of letter). Furthermore,
the CASAC stated that ``the final Integrated Science Assessment (ISA)
Supplement . . . deserve[s] the Administrator's full consideration and
[is] adequate for rulemaking'' (Sheppard, 2022b, p. 2 of letter). The
CASAC generally endorsed EPA's decisions regarding the limited scope of
the draft ISA Supplement, stating that ``this limitation [on scope] is
appropriate for the targeted purpose of the Draft ISA Supplement''
although the CASAC noted it would not be appropriate for ISAs
generally, and recommended that the EPA provide additional
acknowledgment and explanation for the limited scope (Sheppard, 2022b,
p. 2 of letter; see also pp. 2-3 of consensus responses). The EPA
specifically noted in the final ISA Supplement, which was released in
May 2022 (U.S. EPA, 2022a; hereafter referred to as the ISA Supplement
throughout this document) that the ``targeted approach to developing
the Supplement to the 2019 PM ISA for the purpose of reconsidering the
2020 PM NAAQS decision does not reflect a change to EPA's approach for
developing ISAs for NAAQS reviews.'' Thus, the evidence presented
within the 2019 ISA, along with the targeted identification and
evaluation of new scientific information in the ISA Supplement,
provides the scientific basis for the reconsideration of the 2020 PM
NAAQS final decision.
The draft PA was released in October 2021 (86 FR 56263, October 8,
2021). The CASAC PM panel met at a virtual public meeting in December
2021 to review the draft PA (86 FR 52673, September 22, 2021). A
virtual public meeting was then held in February 2022 and March 2022,
and during this meeting the chartered CASAC considered the CASAC PM
panel's draft letter to the Administrator on the draft PA (87 FR 958,
January 7, 2022). The chartered CASAC provided its advice on the draft
PA in a letter to the EPA Administrator dated March 18, 2022 (Sheppard,
2022a). The EPA took steps to address these comments in revising and
finalizing the PA. The 2022 PA considers the scientific evidence
presented in the 2019 ISA and ISA Supplement and considers the
quantitative and technical information presented in the 2020 PA, along
with updated and newly available analyses since the completion of the
2020 review. For those health and welfare effects for which the ISA
Supplement evaluated recently available evidence and for which updated
quantitative analyses were supported (i.e., PM2.5-related
health effects and visibility effects), the 2022 PA includes
consideration of this newly available scientific and technical
information in reaching preliminary conclusions. For those health and
welfare effects for which newly available scientific and technical
[[Page 16213]]
information were not evaluated (i.e., PM10-2.5-related
health effects and non-visibility effects), the conclusions presented
in the 2022 PA rely heavily on the information that supported the
conclusions in the 2020 PA. The final PA was released in May 2022 (U.S.
EPA, 2022b; hereafter referred to as the 2022 PA throughout this
document).
Drawing from his consideration of the scientific evidence assessed
in the 2019 ISA and ISA Supplement and the analyses in the 2022 PA,
including the uncertainties in the evidence and analyses, and from his
consideration of advice from the CASAC, on January 5, 2023, the
Administrator proposed to revise the level of the primary annual
PM2.5 standard and to retain the primary 24-hour
PM2.5 standard, the primary 24-hour PM10
standard, and the secondary PM standards. These proposed decisions were
published in the Federal Register on January 27, 2023 (88 FR 5558,
January 27, 2023). The EPA held a multi-day virtual public hearing on
February 21-23, 2023 (88 FR 6215, January 31, 2023). In total, the EPA
received nearly 700,000 comments on the proposal from members of the
public by the close of the public comment period on March 28, 2023.
Major issues raised in the public comments are discussed throughout the
preamble of this final action. A more detailed summary of all
significant comments, along with the EPA's responses (henceforth
``Response to Comments'' document), can be found in the docket for this
rulemaking (Docket No. EPA-HQ-OAR-2015-0072).
As in prior reviews, the EPA is basing its decision in this
reconsideration on studies and related information in the air quality
criteria, which have undergone CASAC and public review. These studies
assessed in the 2019 ISA \27\ and ISA Supplement \28\ and the 2022 PA,
and the integration of the scientific evidence presented in them, have
undergone extensive critical review by the EPA, the CASAC, and the
public. Decisions on the NAAQS should be based on studies that have
been rigorously assessed in an integrative manner not only by the EPA
but also by the statutorily mandated independent scientific advisory
committee, as well as the public review that accompanies this process.
It is for this reason that the EPA preliminarily concluded that the
scientific evidence available since the completion of the 2019 ISA,
including those raised in public comments on the proposal in 2020,
warranted a partial reopening of the air quality criteria and prepared
an ISA Supplement to enable the EPA, the CASAC, and the public to
consider them further. Some commenters have referred to and discussed
additional individual scientific studies on the health effects of PM
that were not included in the 2019 ISA or ISA Supplement (``new
studies'') and that have not gone through this comprehensive review
process. In considering and responding to comments for which such
``new'' studies were cited in support, the EPA has provisionally
considered the cited studies in the context of the findings of the 2019
ISA and ISA Supplement. The EPA's provisional consideration of these
studies did not and could not provide the kind of in-depth critical
review described above, but rather was focused on determining whether
they warranted further reopening the review of the air quality criteria
to enable the EPA, the CASAC, and the public to consider them further.
---------------------------------------------------------------------------
\27\ In addition to the 2020 review's opening ``call for
information'' (79 FR 71764, December 3, 2014), the 2019 ISA
identified and evaluated studies and reports that have undergone
scientific peer review and were published or accepted for
publication between January 1, 2009, through approximately January
2018 (U.S. EPA, 2019a, p. ES-2). References that are cited in the
2019 ISA, the references that were considered for inclusion but not
cited, and electronic links to bibliographic information and
abstracts can be found at: https://hero.epa.gov/hero/particulate-matter.
\28\ As described above, the ISA Supplement represents an
evaluation of recent studies that are of greatest policy relevance
and utility to the reconsideration of the 2020 final decision on the
PM NAAQS (U.S. EPA, 2022a).
---------------------------------------------------------------------------
This approach, and the decision to rely on the studies and related
information in the air quality criteria, which have undergone CASAC and
public review, is consistent with the EPA's practice in prior NAAQS
reviews and its interpretation of the requirements of the CAA. Since
the 1970 amendments, the EPA has taken the view that NAAQS decisions
are to be based on scientific studies and related information that have
been assessed as a part of the pertinent air quality criteria, and the
EPA has consistently followed this approach. This longstanding
interpretation was strengthened by new legislative requirements enacted
in 1977, which added section 109(d)(2) of the Act concerning CASAC
review of air quality criteria. See 71 FR 6114, 61148 (October 17,
2006, final decision on review of NAAQS for particulate matter) for a
detailed discussion of this issue and the EPA's past practice.
As discussed in the EPA's 1993 decision not to review the
O3 NAAQS, ``new'' studies may sometimes be of such
significance that it is appropriate to delay a decision in a NAAQS
review and to supplement the pertinent air quality criteria so the
studies can be taken into account (58 FR 13013-13014, March 9, 1993).
In the present case, the EPA decided to partially reopen the air
quality criteria and prepared an ISA Supplement as a part of the
reconsideration to facilitate evaluation of these studies by the EPA,
the CASAC, and the public. The narrow scope of the ISA Supplement is
supported by EPA's provisional consideration of ``new'' studies
submitted in response to public comments on the 2020 proposal which
concluded that, taken in context, the ``new'' information and findings
do not materially change any of the broad scientific conclusions
regarding the health and welfare effects of PM in ambient air made in
the air quality criteria. Therefore, a full reopening of the air
quality criteria was not warranted to assess the health and welfare
effects of PM for purposes of the review.
Accordingly, the EPA is basing the final decisions in this
reconsideration on the studies and related information included in the
PM air quality criteria (including the 2019 PM ISA and ISA Supplement)
that have undergone rigorous review by the EPA, the CASAC, and the
public. The EPA will consider these ``new'' studies for inclusion in
the air quality criteria for the next PM NAAQS review, which the EPA
expects to begin soon after the conclusion of this reconsideration and
which will provide the opportunity to fully assess these studies
through a more rigorous review process involving the EPA, the CASAC,
and the public.
D. Air Quality Information
This section provides a summary of basic information related to PM
ambient air quality. It summarizes information on the distribution of
particle size in ambient air (section I.D.1), sources and emissions
contributing to PM in the ambient air (section I.D.2), monitoring
ambient PM in the U.S. (section I.D.3), ambient PM concentrations and
trends in the U.S. (I.D.4), characterizing ambient PM2.5
concentrations for exposure (section I.D.5), and background PM (section
I.D.6). Additional detail on PM air quality can be found in Chapter 2
of the 2022 PA (U.S. EPA, 2022b).
1. Distribution of Particle Size in Ambient Air
In ambient air, PM is a mixture of substances suspended as small
liquid and/or solid particles (U.S. EPA, 2019a, section 2.2) and
distinct health and welfare effects have been linked with exposures to
particles of different sizes. Particles in the atmosphere range in size
from less than 0.01 to more than 10 [mu]m
[[Page 16214]]
in diameter (U.S. EPA, 2019a, section 2.2). The EPA defines
PM2.5, also referred to as fine particles, as particles with
aerodynamic diameters generally less than or equal to 2.5 [mu]m. The
size range for PM10-2.5, also called coarse or thoracic
coarse particles, includes those particles with aerodynamic diameters
generally greater than 2.5 [mu]m and less than or equal to 10 [mu]m.
PM10, which is comprised of both fine and coarse fractions,
includes those particles with aerodynamic diameters generally less than
or equal to 10 [mu]m. In addition, ultrafine particles (UFP) are often
defined as particles with a diameter of less than 0.1 [mu]m based on
physical size, thermal diffusivity or electrical mobility (U.S. EPA,
2019a, section 2.2). Atmospheric lifetimes are generally longest for
PM2.5, which often remains in the atmosphere for days to
weeks (U.S. EPA, 2019a, Table 2-1) before being removed by wet or dry
deposition, while atmospheric lifetimes for UFP and PM10-2.5
are shorter and are generally removed from the atmosphere within hours,
through wet or dry deposition (U.S. EPA, 2019a, Table 2-1; U.S. EPA,
2022b, section 2.1).
2. Sources and Emissions Contributing to PM in the Ambient Air
PM is composed of both primary (directly emitted particles) and
secondary particles. Primary PM is derived from direct particle
emissions from specific PM sources while secondary PM originates from
gas-phase precursor chemical compounds present in the atmosphere that
have participated in new particle formation or condensed onto existing
particles (U.S. EPA, 2019a, section 2.3). As discussed further in the
2019 ISA (U.S. EPA, 2019a, section 2.3.2.1), secondary PM is formed in
the atmosphere by photochemical oxidation reactions of both inorganic
and organic gas-phase precursors. Precursor gases include sulfur
dioxide (SO2), nitrogen oxides (NOX), and
volatile organic compounds (VOC) (U.S. EPA, 2019a, section 2.3.2.1).
Ammonia also plays an important role in the formation of nitrate PM by
neutralizing sulfuric acid and nitric acid. Sources and emissions of PM
are discussed in more detail the 2022 PA (U.S. EPA, 2022b, section
2.1.1). Briefly, anthropogenic sources of PM include both stationary
(e.g., fuel combustion for electricity production and other purposes,
industrial processes, agricultural activities) and mobile (e.g.,
diesel- and gasoline-powered highway vehicles and other engine-driven
sources) sources. Natural sources of PM include dust from the wind
erosion of natural surfaces, sea salt, wildfires, primary biological
aerosol particles (PBAP) such as bacteria and pollen, oxidation of
biogenic hydrocarbons, such as isoprene and terpenes to produce
secondary organic aerosol (SOA), and geogenic sources, such as sulfate
formed from volcanic production of SO2. Wildland fire, which
encompass both wildfire and prescribed fire, accounts for 44% of
emissions of primary PM2.5 emissions (U.S. EPA, 2021b).
Emissions from wildfire comprises 29% of primary PM2.5
emissions.
In recent years, the frequency and magnitude of wildfires have
increased (U.S. EPA, 2019a). The magnitude of the public health impact
of wildfires is substantial both because of the increase in
PM2.5 concentrations as well as the duration of the wildfire
smoke season, which is considered to range from May to November.
Wildfire can make a large contribution to air pollution (including
PM2.5), and wildfire events can threaten public safety and
life. The impacts of wildfire events can be mitigated through
management of wildland vegetation, including through prescribed fire.
Prescribed fire (and some wildfires) can mimic the natural processes
necessary to maintain fire-dependent ecosystems, minimizing
catastrophic wildfires and the risks they pose to safety, property and
air quality (see, e.g., 81 FR 58010, 58038, August 24, 2016). The EPA
views the strategic use of prescribed fire as an important tool for
reducing wildfire risk and the severity of wildfires and wildfire smoke
(88 FR, 54118, 54126, August 9, 2023).\29\ As noted in the PM NAAQS
proposal, agencies have efforts in place to reduce the frequency and
severity of human-caused wildfires (88 FR 5570, January 27, 2023).
---------------------------------------------------------------------------
\29\ See also: https://www.usda.gov/sites/default/files/documents/usda-epa-doi-cdc-mou.pdf.
---------------------------------------------------------------------------
Wildfire events produce high PM emissions that may impact the PM
concentrations in ambient air to the extent that the concentrations
result in an exceedance or violation which may affect the design value
in a given area. The EPA's Exceptional Events Rule (81 FR 68216,
October 3, 2016) describes the process by which air agencies may
request to exclude `event-influenced' data caused by exceptional
events, which can include wildfires and prescribed fires on wildland.
The EPA has issued guidance specifically addressing exceptional events
demonstrations for both wildfires and prescribed fires on wildland.
These documents are available on EPA's Exceptional Events Program
website.\30\ The EPA will develop fire-related exceptional events
implementation tools, including updates as needed to existing guidance
to facilitate more efficient processing of PM2.5-related
exceptional events demonstrations for both the 24-hour and annual
standards.
---------------------------------------------------------------------------
\30\ See: https://www.epa.gov/air-quality-analysis/final-2016-exceptional-events-rule-supporting-guidance-documents-updated-faqs.
---------------------------------------------------------------------------
3. Monitoring of Ambient PM
To promote uniform application of the air quality standards set
forth under the CAA and to achieve the degree of public health and
welfare protection intended for the NAAQS, the EPA establishes PM
Federal Reference Methods (FRMs) for both PM10 and
PM2.5 in appendices J and L to 40 CFR part 50, both of which
were amended following the 2006 and 2012 PM NAAQS reviews. The current
PM monitoring network relies on FRMs and automated continuous Federal
Equivalent Methods (FEMs) approved pursuant to 40 CFR part 53, in part
to support changes necessary for implementation of the revised PM
standards. Additionally, 40 CFR part 58, appendices A through E, detail
the requirements to measure ambient air quality and report ambient air
quality data and related information. More information on PM ambient
monitoring networks is available in section 2.2 of the 2022 PA (U.S.
EPA, 2022b).
The PM2.5 monitoring program is one of the major ambient
air monitoring programs with a robust, nationally consistent network of
ambient air monitoring sites providing mass and/or chemical speciation
measurements. 40 CFR part 58, appendix D, section 4.7 provides the
applicable PM2.5 network design criteria. For most urban
locations, PM2.5 monitors are sited at the neighborhood
scale,\31\ where PM2.5 concentrations are reasonably
homogeneous throughout an entire urban sub-region. In each CBSA with a
monitoring requirement, at least one PM2.5 monitoring
station representing
[[Page 16215]]
area-wide air quality is sited in an area of expected maximum
concentration.\32\ By ensuring the area of expected maximum
concentration in a CBSA has a site compared to both the annual and 24-
hour NAAQS, all other similar locations are thus protected. Sites that
represent relatively unique microscale, localized hot-spot, or unique
middle scale impact sites are only eligible for comparison to the 24-
hour PM2.5 NAAQS.
---------------------------------------------------------------------------
\31\ For PM2.5, neighborhood scale is defined at 40
CFR part 58, appendix D, 4.7.1(c)(3) as follows: Measurements in
this category would represent conditions throughout some reasonably
homogeneous urban sub-region with dimensions of a few kilometers and
of generally more regular shape than the middle scale. Homogeneity
refers to the particulate matter concentrations, as well as the land
use and land surface characteristics. Much of the PM2.5
exposures are expected to be associated with this scale of
measurement. In some cases, a location carefully chosen to provide
neighborhood scale data would represent the immediate neighborhood
as well as neighborhoods of the same type in other parts of the
city. PM2.5 sites of this kind provide good information
about trends and compliance with standards because they often
represent conditions in areas where people commonly live and work
for periods comparable to those specified in the NAAQS. In general,
most PM2.5 monitoring in urban areas should have this
scale.
\32\ 40 CFR part 58, app. D, 4.7.1(b)(2).
---------------------------------------------------------------------------
Under 40 CFR part 50, appendix L, and 40 CFR part 53, and 40 CFR
part 58 appendix D there are three main methods components of the
PM2.5 monitoring program: filter-based FRMs measuring
PM2.5 mass, FEMs measuring PM2.5 mass, and other
samplers used to collect the aerosol used in subsequent laboratory
analysis for measuring PM2.5 chemical speciation. The FRMs
are primarily used for comparison to the NAAQS, but also serve other
important purposes, such as developing trends and evaluating the
performance of FEMs. PM2.5 FEMs are typically continuous
methods used to support forecasting and reporting of the Air Quality
Index (AQI) but are also used for comparison to the NAAQS. Samplers
that are part of the Chemical Speciation Network (CSN) and Interagency
Monitoring of Protected Visual Environments (IMPROVE) network are used
to provide chemical composition of the aerosol and serve a variety of
objectives. More detail on of each of these components of the
PM2.5 monitoring program and of recent changes to
PM2.5 monitoring requirements are described in detail in the
2022 PA (U.S. EPA, 2022b, section 2.2.3).
4. Ambient Concentrations and Trends
This section summarizes available information on recent ambient PM
concentrations in the U.S. and on trends in PM air quality. Sections
I.D.4.a and I.D.4.b summarize information on PM2.5 mass and
components, respectively. Section I.D.4.c summarizes information on
PM10. Sections I.D.4.d and I.D.4.e summarize the more
limited information on PM10-2.5 and UFP, respectively.
Additional detail on PM air quality and trends can be found in the 2022
PA (U.S. EPA, 2022b, section 2.3).
a. PM2.5 mass
At monitoring sites in the U.S., annual PM2.5
concentrations from 2017 to 2019 averaged 8.0 [mu]g/m\3\ (with the 10th
and 90th percentiles at 5.9 and 10.0 [mu]g/m\3\,
respectively) and the 98th percentiles of 24-hour concentrations
averaged 21.3 [mu]g/m\3\ (with the 10th and 90th percentiles at 14.0
and 29.7 [mu]g/m\3\, respectively) (U.S. EPA, 2022b, section 2.3.2.1).
The highest ambient PM2.5 concentrations occur in the
western U.S., particularly in California and the Pacific Northwest
(U.S. EPA, 2022b, Figure 2-15). Much of the eastern U.S. has lower
ambient concentrations, with annual average concentrations generally at
or below 12.0 [mu]g/m\3\ and 98th percentiles of 24-hour concentrations
generally at or below 30 [mu]g/m\3\ (U.S. EPA, 2022b, section 2.3.2.1).
Recent ambient PM2.5 concentrations reflect the
substantial reductions that have occurred across much of the U.S. (U.S.
EPA, 2022b, section 2.3.2.1). From 2000 to 2019, national annual
average PM2.5 concentrations declined from 13.5 [mu]g/m\3\
to 7.6 [mu]g/m\3\, a 43% decrease (U.S. EPA, 2022b, section
2.3.2.1).\33\ These declines have occurred at urban and rural
monitoring sites, although urban PM2.5 concentrations remain
consistently higher than those in rural areas (Chan et al., 2018) due
to the impact of local sources in urban areas. Analyses at individual
monitoring sites indicate that declines in ambient PM2.5
concentrations have been most consistent across the eastern U.S. and in
parts of coastal California, where both annual average and 98th
percentiles of 24-hour concentrations declined significantly (U.S. EPA,
2022b, section 2.3.2.1). In contrast, trends in ambient
PM2.5 concentrations have been less consistent over much of
the western U.S., with no significant changes since 2000 observed at
some sites in the Pacific Northwest, the northern Rockies and plains,
and the Southwest, particularly for 98th percentiles of 24-hour
concentrations (U.S. EPA, 2022b, section 2.3.2.1). As noted below, some
sites in the northwestern U.S. and California, where wildfire have been
relatively common in recent years, have experienced high concentrations
over shorter periods (i.e., 2-hour averages).
---------------------------------------------------------------------------
\33\ See https://www.epa.gov/air-trends/particulate-matter-pm25-trends for up-to-date PM2.5 trends information.
---------------------------------------------------------------------------
The recent deployment of PM2.5 monitors near major roads
in large urban areas provides information on PM2.5
concentrations near an important emissions source. For 2016-2018, Gantt
et al. (2021) reported that 52% and 24% of the time near-road sites
reported the highest annual and 24-hour PM2.5 design value
\34\ in the CBSA, respectively. Of the CBSAs with the highest annual
design values at near-road sites reported by Gantt et al. (2021), those
design values were, on average, 0.8 [micro]g/m\3\ higher than at the
highest measuring non-near-road sites (range is 0.1 to 2.1 [micro]g/
m\3\ higher at near-road sites). Although most near-road monitoring
sites do not have sufficient data to evaluate long-term trends in near-
road PM2.5 concentrations, analyses of the data at one near-
road-like site in Elizabeth, NJ, \35\ show that the annual average
near-road increment has generally decreased between 1999 and 2017 from
about 2.0 [mu]g/m\3\ to about 1.3 [mu]g/m\3\ (U.S. EPA, 2022b, section
2.3.2.1).
---------------------------------------------------------------------------
\34\ A design value is considered valid if it meets the data
handling requirements given in appendix N to 40 CFR part 50.
\35\ The Elizabeth Lab site in Elizabeth, NJ, is situated
approximately 30 meters from travel lanes of the Interchange 13 toll
plaza of the New Jersey Turnpike and within 200 meters of travel
lanes for Interstate 278 and the New Jersey Turnpike.
---------------------------------------------------------------------------
Ambient PM2.5 concentrations can exhibit a diurnal cycle
that varies due to impacts from intermittent emission sources,
meteorology, and atmospheric chemistry. The PM2.5 monitoring
network in the U.S. has an increasing number of continuous FEM monitors
reporting hourly PM2.5 mass concentrations that reflect this
diurnal variation. The 2019 ISA describes a two-peaked diurnal pattern
in urban areas, with morning peaks attributed to rush-hour traffic and
afternoon peaks attributed to a combination of rush hour traffic,
decreasing atmospheric dilution, and nucleation (U.S. EPA, 2019a,
section 2.5.2.3, Figure 2-32). Because a focus on annual average and
24-hour average PM2.5 concentrations could mask subdaily
patterns, and because some health studies examine PM exposure durations
shorter than 24-hours, it is useful to understand the broader
distribution of subdaily PM2.5 concentrations across the
U.S. The 2022 PA presents information on the frequency distribution of
2-hour average PM2.5 mass concentrations from all FEM
PM2.5 monitors in the U.S. for 2017-2019. At sites meeting
the current primary PM2.5 standards, these 2-hour
concentrations generally remain below 10 [mu]g/m\3\, and rarely exceed
30 [mu]g/m\3\. Two-hour concentrations are higher at sites violating
the current standards, generally remaining below 16 [mu]g/m\3\ and
rarely exceeding 80 [mu]g/m\3\ (U.S. EPA, 2022b, section 2.3.2.2.3).
The extreme upper end of the distribution of 2-hour PM2.5
concentrations is shifted higher during the warmer months, generally
corresponding to the period of peak wildfire frequency (April to
September) in the U.S. At sites meeting the current primary standards,
the highest 2-hour concentrations measured rarely occur outside of the
period of peak wildfire frequency. Most of the sites measuring
[[Page 16216]]
these very high concentrations are in the northwestern U.S. and
California, where wildfires have been relatively common in recent years
(see U.S. EPA, 2022b, Appendix A, Figure A-1). When the period of peak
wildfire frequency is excluded from the analysis, the extreme upper end
of the distribution is reduced (U.S. EPA, 2022b, section 2.3.2.2.3).
b. PM2.5 Components
Based on recent air quality data, the major chemical components of
PM2.5 have distinct spatial distributions. Sulfate
concentrations tend to be highest in the eastern U.S., while in the
Ohio Valley, Salt Lake Valley, and California nitrate concentrations
are highest, and relatively high concentrations of organic carbon are
widespread across most of the continental U.S. (U.S. EPA, 2022b,
section 2.3.2.3). Elemental carbon, crustal material, and sea salt are
found to have the highest concentrations in the northeast U.S.,
southwest U.S., and coastal areas, respectively.
An examination of PM2.5 composition trends can provide
insight into the factors contributing to overall reductions in ambient
PM2.5 concentrations. The biggest change in PM2.5
composition that has occurred in recent years is the reduction in
sulfate concentrations due to reductions in SO2 emissions.
Between 2000 and 2015, the nationwide annual average sulfate
concentration decreased by 17% at urban sites and 20% at rural sites.
This change in sulfate concentrations is most evident in the eastern
U.S. and has resulted in organic matter or nitrate now being the
greatest contributor to PM2.5 mass in many locations (U.S.
EPA, 2019a, Figure 2-19). The overall reduction in sulfate
concentrations has contributed substantially to the decrease in
national average PM2.5 concentrations as well as the decline
in the fraction of PM10 mass accounted for by
PM2.5 (U.S. EPA, 2019a, section 2.5.1.1.6; U.S. EPA, 2022b,
section 2.3.1).
c. PM10
At long-term monitoring sites in the U.S., the 2017-2019 average of
2nd highest 24-hour PM10 concentration was 68 [mu]g/m\3\
(with 10th and 90th percentiles at 28 and 124 [mu]g/m\3\, respectively)
(U.S. EPA, 2022b, section 2.3.2.4).\36\ The highest PM10
concentrations tend to occur in the western U.S. Seasonal analyses
indicate that ambient PM10 concentrations are generally
higher in the summer months than at other times of year, though the
most extreme high concentration events are more likely in the spring
(U.S. EPA, 2019a, Table 2-5). This is due to fact that the major
PM10 emission sources, dust and agriculture, are more active
during the warmer and drier periods of the year.
---------------------------------------------------------------------------
\36\ The form of the current 24-hour PM10 standard is
one-expected-exceedance, averaged over three years.
---------------------------------------------------------------------------
Recent ambient PM10 concentrations reflect reductions
that have occurred across much of the U.S. (U.S. EPA, 2022b, section
2.3.2.4). From 2000 to 2019, 2nd highest 24-hour PM10
concentrations have declined by about 46% (U.S. EPA, 2022b, section
2.3.2.4).\37\ Analyses at individual monitoring sites indicate that
annual average PM10 concentrations have generally declined
at most sites across the U.S., with much of the decrease in the eastern
U.S. associated with reductions in PM2.5 concentrations
(U.S. EPA, 2022b, section 2.3.2.4). Annual 2nd highest 24-hour
PM10 concentrations have generally declined in the eastern
U.S., while concentrations in much of the midwest and western U.S. have
remained unchanged or increased since 2000 (U.S. EPA, 2022b, section
2.3.2.4).
---------------------------------------------------------------------------
\37\ For more information, see https://www.epa.gov/air-trends/particulate-matter-pm10-trends#pmnat.
---------------------------------------------------------------------------
Compared to previous reviews, data available from the NCore
monitoring network in the current reconsideration allows a more
comprehensive analysis of the relative contributions of
PM2.5 and PM10-2.5 to PM10 mass.
PM2.5 generally contributes more to annual average
PM10 mass in the eastern U.S. than the western U.S. (U.S.
EPA, 2022b, Figure 2-23). At most sites in the eastern U.S., the
majority of PM10 mass is comprised of PM2.5. As
ambient PM2.5 concentrations have declined in the eastern
U.S. (U.S. EPA, 2022b, section 2.3.2.2), the ratios of PM2.5
to PM10 have also declined. For sites with days having
concurrently very high PM2.5 and PM10
concentrations (U.S. EPA, 2022b, Figure 2-24), the PM2.5/
PM10 ratios are typically higher than the annual average
ratios. This is particularly true in the northwestern U.S. where the
high PM10 concentrations can occur during wildfires with
high PM2.5 (U.S. EPA, 2022b, section 2.3.2.4).
d. PM10-2.5
Since the 2012 review, the availability of PM10-2.5
ambient concentration data has greatly increased because of additions
to the PM10-2.5 monitoring capabilities to the national
monitoring network. As illustrated in the 2022 PA (U.S. EPA, 2022b,
section 2.3.2.5), annual average and 98th percentile
PM10-2.5 concentrations exhibit less distinct differences
between the eastern and western U.S. than for either PM2.5
or PM10.
Due to the short atmospheric lifetime of PM10-2.5
relative to PM2.5, many of the high concentration sites are
isolated and likely near emission sources associated with wind-blown
and fugitive dust. The spatial distributions of annual average and 98th
percentile concentrations of PM10-2.5 are more similar than
that of PM2.5, suggesting that the same dust-related
emission sources are affecting both long-term and episodic
concentrations (U.S. EPA, 2022b, Figure 2-25). The highest
concentrations of PM10-2.5 are in the southwest U.S. where
widespread dry and windy conditions contribute to wind-blown dust
emissions. Additionally, compared to PM2.5 and
PM10, changes in PM10-2.5 concentrations have
been small in magnitude and inconsistent in direction (U.S. EPA, 2022b,
Figure 2-25). The majority of PM10-2.5 sites in the U.S. do
not have a concentration trend from 2000-2019, reflecting the
relatively consistent level of dust emissions across the U.S. during
the same time period (U.S. EPA, 2022b, section 2.3.2.5).\38\
---------------------------------------------------------------------------
\38\ PM from dust emissions in the National Emissions Inventory
(NEI) remain fairly consistent from year-to-year, except when there
are severe weather incursions or there is a dust event that
transports or causes major local dust storms to occur (particularly
in the western U.S.). These dust events and weather incursions
needed to effect dust emissions on a national level are not common
and only seldomly occur. In the emissions trends analysis presented
in the 2022 PA (U.S. EPA, 2022b, section 2.1.1), dust is included in
the NEI sector labeled ``miscellaneous.''
---------------------------------------------------------------------------
e. UFP
Compared to PM2.5 mass, there is relatively little data
on U.S. particle number concentrations, which are dominated by UFP. In
the published literature, annual average particle number concentrations
reaching about 20,000 to 30,000 cm\3\ have been reported in U.S. cities
(U.S. EPA, 2019a). In addition, based on UFP measurements in two urban
areas (New York City, Buffalo) and at a background site (Steuben
County) in New York, there is a pronounced difference in particle
number concentration between different types of locations (U.S. EPA,
2022b, Figure 2-26; U.S. EPA, 2019a, Figure 2-18). Urban particle
number counts were several times higher than at the background site,
and the highest particle number counts in an urban area with multiple
sites (Buffalo) were observed at a near-road location (U.S. EPA, 2022b,
section 2.3.2.6).
Long-term trends in UFP are not routinely available at U.S.
monitoring
[[Page 16217]]
sites. At one background site in Illinois with long-term data
available, the annual average particle number concentration declined
between 2000 and 2019, closely matching the reductions in annual
PM2.5 mass over that same period (U.S. EPA, 2022b, section
2.3.2.6). In addition, a small number of published studies have
examined UFP trends over time. While limited, these studies also
suggest that UFP number concentrations have declined over time along
with decreases in PM2.5 (U.S. EPA, 2022b, section 2.3.2.6).
However, the relationship between changes in ambient PM2.5
and UFPs cannot be comprehensively characterized due to the high
variability and limited monitoring of UFPs (U.S. EPA, 2022b, section
2.3.2.6).
5. Characterizing Ambient PM2.5 Concentrations for Exposure
Epidemiologic studies use various methods to characterize exposure
to ambient PM2.5. The methods used to estimate
PM2.5 concentrations can vary from traditional methods using
monitoring data from ground-based monitors to newer methods using more
complex hybrid modeling approaches. Studies using hybrid modeling
approaches aim to broaden the spatial coverage, as well as estimate
more spatially-resolved ambient PM2.5 concentrations, by
expanding beyond just those areas with monitors and providing estimates
in areas that do not have ground-based monitors (i.e., areas that are
generally less densely populated and tend to have lower
PM2.5 concentrations) and at finer spatial resolutions
(e.g., 1 km x 1 km grid cells). Ground-based PM2.5 monitors
are generally sited in areas of expected maximum concentration. As
such, the hybrid modeling approaches tend to broaden the areas captured
in the exposure assessment, and in doing so, the studies that utilize
these methods tend to report lower mean PM2.5 concentrations
than monitor-based approaches. Further, other aspects of the approaches
applied in the various epidemiologic studies to estimate
PM2.5 exposure and/or to calculate the related study-
reported mean concentration (i.e., population weighting, trim mean
approaches) can affect those data values. More detail related to hybrid
modeling methods, performance of the methods, and how the reported mean
concentrations compare across approaches is provided in section 2.3.3.2
of the 2022 PA (U.S. EPA, 2022b). The subsections below discuss the
characterization of PM2.5 concentrations based on monitoring
data (I.D.5.a) and using hybrid modeling approaches (I.D.5.b).
a. Predicted Ambient PM2.5 and Exposure Based on Monitored
Data
Ambient concentrations of PM2.5 are often characterized
using measurements from national monitoring networks due to the
accuracy and precision of the measurements and the public availability
of data. For applications requiring PM2.5 characterizations
across large areas or provide complete coverage from the site
measurements, data interpolation and averaging techniques (such as
Average Nearest Neighbor tools, and area-wide or population-weighted
averaging of monitors) are sometimes used (U.S. EPA, 2019a, chapter 3).
For an area to meet the NAAQS, all valid design values \39\ in that
area, including the highest annual and 24-hour design values, must be
at or below the levels of the standards. Because the monitoring network
siting requirements are specified to capture the high PM2.5
concentrations (U.S. EPA, 2022b, section 2.2.3), areas meeting an
annual PM2.5 standard with a particular level would be
expected to have long-term average monitored PM2.5
concentrations (i.e., averaged across space and over time in the area)
somewhat below that standard level. This means that the
PM2.5 design value in an area is associated with a
distribution of PM2.5 concentrations in that area, and,
based on monitoring siting requirements, should represent the highest
concentration location applicable to be monitored under the
PM2.5 NAAQS. Analyses in the 2022 PA indicate that, based on
recent air quality in U.S. CBSAs, maximum annual PM2.5
design values are often 10% to 20% higher than annual average
concentrations (i.e., averaged across multiple monitors in the same
CBSA) (U.S. EPA, 2022b, section 2.3.3.1, Figures 2-28 and 2-29). This
difference between the maximum annual design value and the average
concentration in an area can vary, depending on factors such as the
number of monitors, monitor siting characteristics, and the
distribution of ambient PM2.5 concentrations. Given that
higher PM2.5 concentrations have been reported at some near-
road monitoring sites relative to the surrounding area (U.S. EPA,
2022b, section 2.3.2.2.2), recent requirements for PM2.5
monitoring at near-road locations in large urban areas (U.S. EPA,
2022b, section 2.2.3.3) may increase the ratios of maximum design
values to average annual design values in some areas. Such ratios may
also depend on how the averages are calculated (i.e., averaged across
monitors versus across modeled grid cells, as described below in
section I.5.b). Compared to annual design values, the analysis in the
2022 PA indicates a more variable relationship between maximum 24-hour
PM2.5 design values and annual average concentrations (U.S.
EPA, 2022b, section 2.3.3.1, Figure 2-29).
---------------------------------------------------------------------------
\39\ For the annual PM2.5 standard, design values are
calculated as the annual arithmetic mean PM2.5
concentration, averaged over 3 years. For the 24-hour standard,
design values are calculated as the 98th percentile of the annual
distribution of 24-hour PM2.5 concentrations, averaged
over three years (appendix N of 40 CFR part 50).
---------------------------------------------------------------------------
b. Comparison of PM2.5 Hybrid Modeling Approaches in
Estimating Exposure and Relative to Design Values
Two types of hybrid approaches that have been utilized in several
key PM2.5 epidemiologic studies in the 2019 ISA and ISA
Supplement include neural network approaches and a satellite-based
method with regression of residual PM2.5 with land-use and
other variables to improve estimates of PM2.5 concentration
in the U.S. As such, the 2022 PA further compares these two types of
approaches across various scales (e.g., CBSA versus nationwide), taking
into account population weighting approaches utilized in epidemiologic
studies when estimating PM2.5 exposure (U.S. EPA, 2022b,
section 2.3.3.2.4). Additionally, the 2022 PA assesses how average
PM2.5 concentrations computed in epidemiologic studies using
these hybrid surfaces compare to the maximum design values measured at
ground-based monitors. For this assessment, the 2022 PA evaluates the
DI2019 \40\ and HA2020 \41\ hybrid surfaces, surfaces that are used in
several of the key epidemiologic studies in the 2022 PA. This analysis
is intended to help inform how the magnitude of the overall study-
reported mean PM2.5 concentrations in epidemiologic studies
may be
[[Page 16218]]
influenced by the approach used to compute that mean and how that value
might compare to monitor reported concentrations. The PM2.5
standards are expected to achieve a pattern of air quality through the
attainment of a specific design value at each monitor in the monitoring
network. As a result, it is important to be able to assess the
relationship between monitor concentrations and patterns of air quality
evaluated in the epidemiologic studies.
---------------------------------------------------------------------------
\40\ This analysis includes an updated version of the surface
used in Di et al. (2016). Predictions in Di et al. (2016) were for
2000 to 2012 using a neural network model. The Di et al. (2019)
study improved on that effort in several ways. First, a generalized
additive model was used that accounted for geographic variations in
performance to combine predictions from three models (neural
network, random forest, and gradient boosting) to make the final
optimal PM2.5 predictions. Second, the datasets were
updated that were used in model training and included additional
variables such as 12-km CMAQ modeling as predictors. Finally, more
recent years were included in the Di et al. (2019) study.
\41\ The HA2020 field is based on the V4.NA.03 product available
at: https://sites.wustl.edu/acag/datasets/surface-pm2-5/. The name
``HA2020'' comes from the references for this product (Hammer et
al., 2020; van Donkelaar et al., 2019).
---------------------------------------------------------------------------
In estimating exposure, some studies focus on estimating
concentrations in urban areas, while others examine the entire U.S. or
large portions of the country. In general, the areas that are not
included in the CBSA-only analysis tend to be more rural or less
densely populated areas, tend to have lower PM2.5
concentrations, and likely correspond to those locations where
monitoring data availability is limited or nonexistent (U.S. EPA,
2022b, section 2.3.3.2.4, Figure 2-37). To evaluate the differences in
mean PM2.5 concentrations across different spatial scales,
the 2022 PA analysis compares the DI2019 and HA2020 surfaces. At the
national scale, the two surfaces generally produce similar average
annual PM2.5 concentrations, with the DI2019 surface being
slightly higher compared to the HA2020 surface. The average annual
PM2.5 concentrations are also slightly higher using the
DI2019 surface compared to the HA2020 surface when the analyses are
conducted for CBSAs. Also, regardless of which surface is used, the
average annual and 3-year average of the average annual
PM2.5 concentrations for the CBSA-only analyses are somewhat
higher than for the nationwide analyses (4-8% higher) (U.S. EPA, 2022b,
section 2.3.3.2.4, Table 2-5).\42\ Overall, these analyses suggest that
there are only slight differences in the average PM2.5
concentrations depending on the hybrid modeling method employed, though
including other hybrid modeling methods in this comparison could result
in larger differences.
---------------------------------------------------------------------------
\42\ For the national scale, 3-year averages of the average
annual PM2.5 concentrations generally range from about
5.3 [micro]g/m\3\ to 8.1 [micro]g/m\3\, compared to the CBSA scale,
which ranges from 5.7 [micro]g/m\3\ to 8.7 [micro]g/m\3\. (U.S. EPA,
2022b, section 2.3.3.2.4, Table 2-6).
---------------------------------------------------------------------------
The 2022 PA next evaluates how the averages of the hybrid model
surfaces compare to regulatory design values using both the DI2019 and
HA2020 surfaces and how population weighting influences the mean
PM2.5 concentration.\43\ As presented in the 2022 PA, the
results using the DI2019 and HA2020 surfaces are similar for the
average annual PM2.5 concentrations, for each 3-year period.
When population weighting is not applied, the average annual
PM2.5 concentrations generally range from 7.0 to 8.6
[micro]g/m\3\. When population weighting is applied, the average annual
PM2.5 concentrations are slightly higher, ranging from 8.2
to 10.2 [micro]g/m\3\. As with CBSAs versus the national comparison
above, population weighting results in a higher average
PM2.5 concentration than when population weighting is not
applied (U.S. EPA, 2022b, section 2.3.3.2.4, Table 2-7). For the CBSAs
included in the population weighted analyses, the average maximum
annual design values generally range from 9.5 to 11.7 [micro]g/m\3\.
The results are similar for both the DI2019 and HA2020 surfaces and the
maximum annual PM2.5 design values measured at the monitors
are often 40% to 50% higher than average annual PM2.5
concentrations predicted by hybrid modeling methods when population
weighting is not applied. However, when population weighting is
applied, the ratio of the maximum annual PM2.5 design values
to the predicted average annual PM2.5 concentrations are
lower than when population weighting is not applied, with monitored
design values generally 15% to 18% higher than population-weighted
hybrid modeling average annual PM2.5 concentrations (U.S.
EPA, 2022b, section 2.3.3.2.4, Table 2-7).
---------------------------------------------------------------------------
\43\ For this analysis, the 2022 PA includes CBSAs with three or
more valid design values for the 3-year period. The regulatory
design values for the CBSAs were calculated for each 3-year period
for the CBSAs with 3 or more design values in each of the 3-year
periods. Using the maximum design value for each CBSA and by each 3-
year period, the ratio of maximum design values to modeled average
annual PM2.5 concentrations were calculated, for each 3-
year period. More details about the analytical methods used for this
analysis are described in section A.6 of Appendix A in the 2022 PA
(U.S. EPA, 2022b).
---------------------------------------------------------------------------
6. Background PM
In this reconsideration, background PM is defined as all particles
that are formed by sources or processes that cannot be influenced by
actions within the jurisdiction of concern. U.S. background PM is
defined as any PM formed from emissions other than U.S. anthropogenic
(i.e., manmade) emissions. Potential sources of U.S. background PM
include both natural sources (i.e., PM that would exist in the absence
of any anthropogenic emissions of PM or PM precursors) and
transboundary sources originating outside U.S. borders. Background PM
is discussed in more detail in the 2022 PA (U.S. EPA, 2022b, section
2.4). At annual and national scales, estimated background PM
concentrations in the U.S. are small compared to contributions from
domestic anthropogenic sources.\44\ For example, based on zero-out
modeling in the last review of the PM NAAQS, annual background
PM2.5 concentrations were estimated to range from 0.5-3
[micro]g/m\3\ across the sites examined. In addition, speciated
monitoring data from IMPROVE sites can provide some insights into how
contributions from different sources, including sources of background
PM, may have changed over time. Such data suggests the estimates of
background concentrations using speciated monitoring data from IMPROVE
monitors are around 1-3 [micro]g/m\3\ and have not changed
significantly since the 2012 review. Contributions to background PM in
the U.S. result mainly from sources within North America. Contributions
from intercontinental events have also been documented (e.g., transport
from dust storms occurring in deserts in North Africa and Asia), but
these events are less frequent and represent a relatively small
fraction of background PM in most of the U.S. (U.S. EPA, 2022b, section
2.4).
---------------------------------------------------------------------------
\44\ Sources that contribute to natural background PM include
dust from the wind erosion of natural surfaces, sea salt, wildland
fires, primary biological aerosol particles such as bacteria and
pollen, oxidation of biogenic hydrocarbons such as isoprene and
terpenes to produce secondary organic aerosols (SOA), and geogenic
sources such as sulfate formed from volcanic production of
SO2 and oceanic production of dimethyl-sulfide (U.S. EPA,
2022b, section 2.4). While most of these sources release or
contribute predominantly to fine aerosol, some sources including
windblown dust, and sea salt also produce particles in the coarse
size range (U.S. EPA, 2019a, section 2.3.3).
---------------------------------------------------------------------------
II. Rationale for Decisions on the Primary PM2.5 Standards
This section presents the rationale for the Administrator's
decision to revise the primary annual PM2.5 standard down to
a level of 9 [micro]g/m\3\ and retain the primary 24-hour
PM2.5 standard. This rationale is based on a thorough review
of the scientific evidence generally published through January
2018,\45\ as evaluated in the 2019 ISA (U.S. EPA, 2019a), on the human
health effects of PM2.5 associated with long- and short-term
exposures \46\ to PM2.5 in
[[Page 16219]]
the ambient air. Additionally, this rationale is based on a thorough
evaluation of some studies that became available after the literature
cutoff date of the 2019 ISA, as evaluated in the ISA Supplement, that
could either further inform the adequacy of the current PM NAAQS or
address key scientific topics that have evolved since the literature
cutoff date for the 2019 ISA, generally through March 2021 (U.S. EPA,
2022a).\47\ The Administrator's rationale also takes into account: (1)
The 2022 PA evaluation of the policy-relevant information in the 2019
ISA and ISA Supplement and presentation of quantitative analyses of air
quality and health risks; (2) CASAC advice and recommendations; and (3)
public comments received during the development of these documents.
---------------------------------------------------------------------------
\45\ In addition to the 2020 review's opening ``call for
information'' (79 FR 71764, December 3, 2014), the 2019 ISA
identified and evaluated studies and reports that have undergone
scientific peer review and were published or accepted for
publication between January 1, 2009, through approximately January
2018 (U.S. EPA, 2019a, p. ES-2). References that are cited in the
2019 ISA, the references that were considered for inclusion but not
cited, and electronic links to bibliographic information and
abstracts can be found at: https://hero.epa.gov/hero/particulate-matter.
\46\ Short-term exposures are defined as those exposures
occurring over hours up to 1 month, whereas long-term exposures are
defined as those exposures occurring over 1 month to years (U.S.
EPA, 2019a, section P.3.1).
\47\ The ISA Supplement represents an evaluation of recent
studies that are of greatest policy relevance to the reconsideration
of the 2020 final decision on the PM NAAQS. Specifically, the ISA
Supplement focuses on studies of health effects for which the
evidence in the 2019 ISA supported a ``causal relationship'' (i.e.,
short- and long-term PM2.5 exposure and mortality and
cardiovascular effects) because those were the health effects that
were most useful in informing conclusions in the 2020 PA. The ISA
Supplement does not include an evaluation of studies for other
PM2.5-related health effects (U.S. EPA, 2022a).
---------------------------------------------------------------------------
In presenting the rationale for the Administrator's decisions and
its foundations, section II.A provides background on the general
approach for this reconsideration and the basis for the existing
standard, and also presents brief summaries of key aspects of the
currently available health effects and risk information. Section II.B
summarizes the CASAC advice and the basis for the proposed conclusions,
addresses public comments received on the proposal and presents the
Administrator's conclusions on the adequacy of the current standards,
drawing on consideration of the scientific evidence and quantitative
risk information, advice from the CASAC, and comments from the public.
Section II.C summarizes the Administrator's decision on the primary
PM2.5 standards.
A. Introduction
The general approach for this reconsideration of the 2020 final
decision on the primary PM2.5 standards is fundamentally
based on using the EPA's assessment of the current scientific evidence
and associated quantitative analyses to inform the Administrator's
judgment regarding primary PM2.5 standards that protect
public health with an adequate margin of safety. The EPA's assessments
are primarily documented in the 2019 ISA, ISA Supplement, and 2022 PA,
all of which have received CASAC review and public comment (83 FR
53471, October 23, 2018; 83 FR 55529, November 6, 2018; 85 FR 4655,
January 27, 2020; 86 FR 52673, September 22, 2021; 86 FR 54186,
September 30, 2021; 86 FR 56263, October 8, 2021; 87 FR 958, January 7,
2022; 87 FR 22207, April 14, 2022; 87 FR 31965, May 26, 2022). In
bridging the gap between the scientific assessments of the 2019 ISA and
ISA Supplement and the judgments required of the Administrator in
determining whether the current standards provide the requisite public
health protection, the 2022 PA evaluates policy implications of the
evaluation of the current evidence in the 2019 ISA and ISA Supplement,
and the risk information documented in the 2022 PA. In evaluating the
public health protection afforded by the current standards, the four
basic elements of the NAAQS (i.e., indicator, averaging time, level,
and form) are considered collectively.
The final decision on the adequacy of the current primary
PM2.5 standards is a public health policy judgment to be
made by the Administrator. In reaching conclusions with regard to the
standards, the decision will draw on the scientific information and
analyses about health effects and population risks, as well as
judgments about how to consider the range and magnitude of
uncertainties that are inherent in the scientific evidence and
analyses. This approach is based on the recognition that the available
health effects evidence generally reflects a continuum, consisting of
levels at which scientists generally agree that health effects are
likely to occur, through lower levels at which the likelihood and
magnitude of the response become increasingly uncertain. This approach
is consistent with the requirements of the NAAQS provisions of the
Clean Air Act and with how the EPA and the courts have historically
interpreted the Act (summarized in section I.A above). These provisions
require the Administrator to establish primary standards that, in the
judgment of the Administrator, are requisite to protect public health
with an adequate margin of safety. In so doing, the Administrator seeks
to establish standards that are neither more nor less stringent than
necessary for this purpose. The Act does not require that primary
standards be set at a zero-risk level, but rather at a level that
avoids unacceptable risks to public health, including the health of
sensitive (also referred to as ``at-risk'') groups.\48\
---------------------------------------------------------------------------
\48\ As noted in section I.A above, the legislative history
describes such protection for the sensitive group of individuals and
not for a single person in the sensitive group (see S. Rep. No. 91-
1196, 91st Cong, 2d Sess. 10 [1970]); see also Am. Lung Ass'n v.
EPA, 134 F.3d 388, 389 (D.C. Cir. 1998).
---------------------------------------------------------------------------
1. Background on the Current Standards
The current primary PM2.5 standards were retained in
2020 based on the scientific evidence and quantitative risk information
available at that time, as well as the then-Administrator's judgments
regarding the available health effects evidence and the appropriate
degree of public health protection afforded by the existing standards
(85 FR 82718, December 18, 2020). With the 2020 decision, the then-
Administrator retained the primary annual PM2.5 standard
with its level of 12.0 [mu]g/m\3\ and retained the primary 24-hour
PM2.5 standard with its level of 35 [mu]g/m\3\. The key
considerations and the then-Administrator's conclusions regarding the
primary PM2.5 standards in the 2020 review are summarized
below.
The health effects evidence base available in the 2020 review
included extensive evidence from previous reviews as well as the
evidence that had emerged since the prior review had been completed in
2012. This evidence base, spanning several decades, documents the
relationship between short- and long-term PM2.5 exposure and
mortality or serious morbidity effects. The evidence available in the
2019 ISA reaffirmed, and in some cases strengthened, the conclusions
from the 2009 ISA regarding the health effects of PM2.5
exposures (U.S. EPA, 2019a). Much of the evidence came from
epidemiologic studies conducted in North America, Europe, or Asia
examining short-term and long-term exposures that demonstrated
generally positive, and often statistically significant,
PM2.5 health effect associations with a range of outcomes
including non- accidental, cardiovascular, or respiratory mortality;
cardiovascular- or respiratory-related hospitalizations or emergency
department visits; and other mortality/morbidity outcomes (e.g., lung
cancer mortality or incidence, asthma development). Experimental
evidence, as well as evidence from panel studies, strengthened support
for potential biological pathways through which PM2.5
exposures could lead to health effects reported in many population-
based epidemiologic studies, including support for pathways that could
lead to cardiovascular, respiratory, nervous system, and cancer-related
effects.
[[Page 16220]]
Based on this evidence, the 2019 ISA concluded there to be a causal
relationship between long- and short-term PM2.5 exposure and
mortality and cardiovascular effects, as well as likely to be causal
relationships between long- and short-term PM2.5 exposure
and respiratory effects, and between long-term PM2.5
exposure and cancer and nervous system effects (U.S. EPA, 2019a,
section 1.7).
Epidemiologic studies reported PM2.5 health effect
associations with mortality and/or morbidity across multiple U.S.
cities and in diverse populations, including in studies examining
populations and lifestages that may be at increased risk of
experiencing a PM2.5-related health effect (e.g., older
adults, children). The 2019 ISA cited extensive evidence indicating
that ``both the general population as well as specific populations and
lifestages are at risk for PM2.5-related health effects''
(U.S. EPA, 2019a, p. 12-1), including children and older adults, people
with pre-existing respiratory or cardiovascular disease, minority
populations, and low socioeconomic status (SES) populations.
The risk information available in the 2020 review included risk
estimates for air quality conditions just meeting the existing primary
PM2.5 standards, and also for air quality conditions just
meeting potential alternative standards. The general approach to
estimating PM2.5-associated health risks combined
concentration-response (C-R) functions from epidemiologic studies with
model-based PM2.5 air quality surfaces, baseline health
incidence data, and population demographics for 47 urban areas (U.S.
EPA, 2020b, section 3.3, Figure 3-10, Appendix C). The risk assessment
estimated that the existing primary PM2.5 standards could
allow a substantial number of PM2.5-associated deaths in the
U.S. Uncertainty in risk estimates (e.g., in the size of risk
estimates) can result from a number of factors, including assumptions
about the shape of the C-R relationship with mortality at low ambient
PM2.5 concentrations, the potential for confounding and/or
exposure measurement error, and the methods used to adjust
PM2.5 air quality.
Consistent with the general approach routinely employed in NAAQS
reviews, the initial consideration in the 2020 review of the primary
PM2.5 standards was with regard to the adequacy of the
protection provided by the existing standards.
As an initial matter, the then-Administrator considered the range
of scientific evidence evaluating these effects, including studies of
at-risk populations, to inform his review of the primary
PM2.5 standards, placing the greatest weight on evidence of
effects for which the 2019 ISA determined there to be a causal or
likely to be causal relationship with long- and short-term
PM2.5 exposures (85 FR 82714-82715, December 18, 2020).
With regard to indicator, the then-Administrator recognized that,
consistent with the evidence available in prior reviews, the scientific
evidence continued to provide strong support for health effects
following short- and long-term PM2.5 exposures. He noted the
2020 PA conclusions that the information continued to support the
PM2.5 mass-based indicator and remained too limited to
support a distinct standard for any specific PM2.5 component
or group of components, and too limited to support a distinct standard
for the ultrafine fraction. Thus, the then-Administrator concluded that
it was appropriate to retain PM2.5 as the indicator for the
primary standards for fine particles (85 FR 82715, December 18, 2020).
With respect to averaging time and form, the then-Administrator
noted that the scientific evidence continued to provide strong support
for health effects associations with both long-term (e.g., annual or
multi-year) and short-term (e.g., mostly 24-hour) exposures to
PM2.5, consistent with the conclusions in the 2020 PA. In
the 2019 ISA, epidemiologic and controlled human exposure studies
examined a variety of PM2.5 exposure durations.
Epidemiologic studies continued to provide strong support for health
effects associated with short-term PM2.5 exposures based on
24-hour PM2.5 averaging periods, and the EPA noted that
associations with subdaily estimates are less consistent and, in some
cases, smaller in magnitude (U.S. EPA, 2019a, section 1.5.2.1; U.S.
EPA, 2020b, section 3.5.2.2). In addition, controlled human exposure
and panel-based studies of subdaily exposures typically examined
subclinical effects, rather than the more serious population-level
effects that have been reported to be associated with 24-hour exposures
(e.g., mortality, hospitalizations). Taken together, the 2019 ISA
concluded that epidemiologic studies did not indicate that subdaily
averaging periods were more closely associated with health effects than
the 24-hour average exposure metric (U.S. EPA, 2019a, section 1.5.2.1).
Additionally, while controlled human exposure studies provided
consistent evidence for cardiovascular effects following
PM2.5 exposures for less than 24 hours (i.e., <30 minutes to
5 hours), exposure concentrations in the studies were well-above the
ambient concentrations typically measured in locations meeting the
existing standards (U.S. EPA, 2020b, section 3.2.3.1). Thus, these
studies also did not suggest the need for additional protection against
subdaily PM2.5 exposures (U.S. EPA, 2020b, section 3.5.2.2).
Therefore, the then-Administrator judged that the 24-hour averaging
time remained appropriate (85 FR 82715, December 18, 2020).
With regard to the form of the 24-hour standard (98th percentile,
averaged over three years), the then-Administrator noted that
epidemiologic studies continued to provide strong support for health
effect associations with short-term (e.g., mostly 24-hour)
PM2.5 exposures (U.S. EPA, 2020b, section 3.5.2.3) and that
controlled human exposure studies provided evidence for health effects
following single short-term ``peak'' PM2.5 exposures. Thus,
the evidence supported retaining a standard focused on providing
supplemental protection against short-term peak exposures and supported
a 98th percentile form for a 24-hour standard. The then-Administrator
further noted that this form also provided an appropriate balance
between limiting the occurrence of peak 24-hour PM2.5
concentrations and identifying a stable target for risk management
programs (U.S. EPA, 2020b, section 3.5.2.3). As such, the then-
Administrator concluded that the available information supported
retaining the form and averaging time of the current 24-hour standard
(98th percentile, averaged over three years) and annual standard
(annual average, averaged over three years) (85 FR 82715, December 18,
2020).
With regard to the level of the standards, in reaching his final
decision, the then-Administrator considered the large body of evidence
presented and assessed in the 2019 ISA (U.S. EPA, 2019a), the policy-
relevant and risk-based conclusions and rationales as presented in the
2020 PA (U.S. EPA, 2020b), advice from the CASAC, and public comments.
In particular, in considering the 2019 ISA and 2020 PA, he considered
key epidemiologic studies that evaluated associations between
PM2.5 air quality distributions and mortality and morbidity,
including key accountability studies; the availability of experimental
studies to support biological plausibility; controlled human exposure
studies examining effects following short-term PM2.5
exposures; air quality analyses; and the important uncertainties and
limitations associated with the information (85 FR 82715, December 18,
2020).
[[Page 16221]]
As an initial matter, the then-Administrator considered the
protection afforded by both the annual and 24-hour standards together
against long- and short-term PM2.5 exposures and health
effects. The Administrator recognized that the annual standard was most
effective in controlling ``typical'' PM2.5 concentrations
near the middle of the air quality distribution (i.e., around the mean
of the distribution), but also provided some control over short-term
peak PM2.5 concentrations. On the other hand, the 24-hour
standard, with its 98th percentile form, was most effective at limiting
peak 24-hour PM2.5 concentrations, but in doing so also had
an effect on annual average PM2.5 concentrations. Thus,
while either standard could be viewed as providing some measure of
protection against both average exposures and peak exposures, the 24-
hour and annual standards were not expected to be equally effective at
limiting both types of exposures. Thus, consistent with previous
reviews, the then-Administrator's consideration of the public health
protection provided by the existing primary PM2.5 standards
was based on his consideration of the combination of the annual and 24-
hour standards. Specifically, he recognized that the annual standard
was more likely to appropriately limit the ``typical'' daily and annual
exposures that are most strongly associated with the health effects
observed in epidemiologic studies. The then-Administrator concluded
that an annual standard (as the arithmetic mean, averaged over three
years) remained appropriate for targeting protection against the annual
and daily PM2.5 exposures around the middle portion of the
PM2.5 air quality distribution. Further, recognizing that
the 24-hour standard (with its 98th percentile form) was more directly
tied to short-term peak PM2.5 concentrations, and more
likely to appropriately limit exposures to such concentrations, the
then-Administrator concluded that the current 24-hour standard (with
its 98th percentile form, averaged over three years) remained
appropriate to provide a balance between limiting the occurrence of
peak 24-hour PM2.5 concentrations and identifying a stable
target for risk management programs. However, the then-Administrator
recognized that changes in PM2.5 air quality to meet an
annual standard would likely result not only in lower short- and long-
term PM2.5 concentrations near the middle of the air quality
distribution, but also in fewer and lower short-term peak
PM2.5 concentrations. The then-Administrator further
recognized that changes in air quality to meet a 24-hour standard, with
a 98th percentile form, would result not only in fewer and lower peak
24-hour PM2.5 concentrations, but also in lower annual
average PM2.5 concentrations (85 FR 82715-82716, December
18, 2020).
Thus, in considering the adequacy of the 24-hour standard, the
then-Administrator noted the importance of considering whether
additional protection was needed against short-term exposures to peak
PM2.5 concentrations. In examining the scientific evidence,
he noted the limited utility of the animal toxicological studies in
directly informing conclusions on the appropriate level of the standard
given the uncertainty in extrapolating from effects in animals to those
in human populations. The then-Administrator noted that controlled
human exposure studies provided evidence for health effects following
single, short-term PM2.5 exposures that corresponded best to
exposures that might be experienced in the upper end of the
PM2.5 air quality distribution in the U.S. (i.e., ``peak''
concentrations). However, most of these studies examined exposure
concentrations considerably higher than are typically measured in areas
meeting the standards (U.S. EPA, 2020b, section 3.2.3.1). In
particular, controlled human exposure studies often reported
statistically significant effects on one or more indicators of
cardiovascular function following 2-hour exposures to PM2.5
concentrations at and above 120 [mu]g/m\3\ (at and above 149 [mu]g/m\3\
for vascular impairment, the effect shown to be most consistent across
studies). To provide insight into what these studies may indicate
regarding the primary PM2.5 standards, the 2020 PA (U.S.
EPA, 2020b, p. 3-49) noted that 2-hour ambient concentrations of
PM2.5 at monitoring sites meeting the current standards
almost never exceeded 32 [mu]g/m\3\. In fact, even the extreme upper
end of the distribution of 2-hour PM2.5 concentrations at
sites meeting the primary PM2.5 standards remained well-
below the PM2.5 exposure concentrations consistently shown
in controlled human exposure studies to elicit effects (i.e., 99.9th
percentile of 2-hour concentrations at these sites is 68 [mu]g/m\3\
during the warm season). Thus, the experimental evidence did not
indicate the need for additional protection against exposures to peak
PM2.5 concentrations, beyond the protection provided by the
combination of the 24-hour and the annual standards (U.S. EPA, 2020b,
section 3.2.3.1; 85 FR 82716, December 18, 2020).
With respect to the epidemiologic evidence, the then-Administrator
noted that the studies did not indicate that associations in those
studies were strongly influenced by exposures to peak concentrations in
the air quality distribution and thus did not indicate the need for
additional protection against short-term exposures to peak
PM2.5 concentrations (U.S. EPA, 2020b, section 3.5.1). The
then-Administrator noted that this was consistent with CASAC consensus
support for retaining the current 24-hour standard. Thus, the then-
Administrator concluded that the 24-hour standard with its level of 35
[micro]g/m\3\ was adequate to provide supplemental protection (i.e.,
beyond that provided by the annual standard alone) against short-term
exposures to peak PM2.5 concentrations (85 FR 82716,
December 18, 2020).
With regard to the level of the annual standard, the then-
Administrator recognized that the annual standard, with its form based
on the arithmetic mean concentration, was most appropriately meant to
limit the ``typical'' daily and annual exposures that were most
strongly associated with the health effects observed in epidemiologic
studies. However, the then-Administrator also noted that while
epidemiologic studies examined associations between distributions of
PM2.5 air quality and health outcomes, they did not identify
particular PM2.5 exposures that cause effects and thus, they
could not alone identify a specific level at which the standard should
be set, as such a determination necessarily required the then-
Administrator's judgment. Thus, consistent with the approaches in
previous NAAQS reviews, the then-Administrator recognized that any
approach that used epidemiologic information in reaching decisions on
what standards are appropriate necessarily required judgments about how
to translate the information from the epidemiologic studies into a
basis for appropriate standards. This approach included consideration
of the uncertainties in the reported associations between daily or
annual average PM2.5 exposures and mortality or morbidity in
the epidemiologic studies. Such an approach is consistent with setting
standards that are neither more nor less stringent than necessary,
recognizing that a zero-risk standard is not required by the Clean Air
Act (CAA) (85 FR 82716, December 18, 2020).
The then-Administrator emphasized uncertainties and limitations
that were present in epidemiologic studies in previous reviews and
persisted in the 2020 review. These uncertainties
[[Page 16222]]
included exposure measurement error, potential confounding by
copollutants, increasing uncertainty of associations at lower
PM2.5 concentrations, and heterogeneity of effects across
different cities or regions (85 FR 82716, December 18, 2020). The then-
Administrator also noted the advice given by the CASAC on this matter.
As described in section I.C.5 above, the CASAC did not reach consensus
on the adequacy of the primary annual PM2.5 standard. ``Some
CASAC members'' expressed support for retaining the primary annual
PM2.5 standard while ``other members'' expressed support for
revising that standard in order to increase public health protection
(Cox, 2019b, p. 1 of consensus letter). The CASAC members who supported
retaining the annual standard expressed their concerns with the
epidemiologic studies, asserting that these studies did not provide a
sufficient basis for revising the existing standards. They also
identified several key concerns regarding the associations reported in
epidemiologic studies and concluded that ``while the data on
associations should certainly be carefully considered, this data should
not be interpreted more strongly than warranted based on its
methodological limitations'' (Cox, 2019b, p. 8 consensus responses).
Taking into consideration the views expressed by the CASAC members
who supported retaining the annual standard, the then-Administrator
recognized that epidemiologic studies examined associations between
distributions of PM2.5 air quality and health outcomes, and
they did not identify particular PM2.5 exposures that cause
effects (U.S. EPA, 2020b, section 3.1.2). While the Administrator
remained concerned about placing too much weight on epidemiologic
studies to inform conclusions on the adequacy of the primary standards,
he noted the approach to considering such studies in the 2012 review.
In the 2012 review, it was noted that the evidence of an association in
any epidemiologic study was ``strongest at and around the long-term
average where the data in the study are most concentrated'' (78 FR
3140, January 15, 2013). In considering the characterization of
epidemiologic studies, the then-Administrator viewed that when
assessing the mean concentrations of the key short-term and long-term
epidemiologic studies in the U.S. that used ground-based monitoring
(i.e., those studies where the mean is most directly comparable to the
current annual standard), the majority of studies had mean
concentrations at or above the level of the existing annual standard,
with the mean of the study-reported means or medians equal to 13.5
[micro]g/m\3\, a concentration level above the existing level of the
primary annual standard of 12 [micro]g/m\3\. The then-Administrator
further noted his caution in directly comparing the reported study mean
values to the standard level given that study-reported mean
concentrations, by design, are generally lower than the design value of
the highest monitor in an area, which determines compliance. In the
2020 PA, analyses of recent air quality in U.S. CBSAs indicated that
maximum annual PM2.5 design values for a given three-year
period were often 10% to 20% higher than average monitored
concentrations (i.e., averaged across multiple monitors in the same
CBSA) (U.S. EPA, 2020b, Appendix B, section B.7). He further noted his
concern in placing too much weight on any one epidemiologic study but
instead judged that it was more appropriate to focus on the body of
studies together and therefore noted the calculation of the mean of
study-reported means (or medians). Thus, while the then-Administrator
was cautious in placing too much weight on the epidemiologic evidence
alone, he noted that: (1) The reported mean concentration in the
majority of the key U.S. epidemiologic studies using ground-based
monitoring data were above the level of the existing annual standard;
(2) the mean of the reported study means (or medians) (i.e., 13.5
[micro]g/m\3\) was above the level of the current standard; \49\ (3)
air quality analyses showed the study means to be lower than their
corresponding design values by 10-20%; and (4) these analyses must be
considered in light of uncertainties inherent in the epidemiologic
evidence. When taken together, the then-Administrator judged that, even
if it were appropriate to place more weight on the epidemiologic
evidence, this information did not call into question the adequacy of
the current standards (85 FR 82716-17, December 18, 2020).
---------------------------------------------------------------------------
\49\ The median of the study-reported mean (or median)
PM2.5 concentrations is 13.3 [micro]g/m\3\, which was
also above the level of the existing standard.
---------------------------------------------------------------------------
In addition to the evidence, the then-Administrator also considered
the potential implications of the risk assessment. He noted that all
risk assessments have limitations and that he remained concerned about
the uncertainties in the underlying epidemiologic data used in the risk
assessment. The then-Administrator also noted that in previous reviews,
these uncertainties and limitations have often resulted in less weight
being placed on quantitative estimates of risk than on the underlying
scientific evidence itself (e.g., 78 FR 3086, 3098-99, January 15,
2013). These uncertainties and limitations included uncertainty in the
shapes of C-R functions, particularly at low concentrations;
uncertainties in the methods used to adjust air quality; and
uncertainty in estimating risks for populations, locations and air
quality distributions different from those examined in the underlying
epidemiologic study (U.S. EPA, 2020b, section 3.3.2.4). Additionally,
the then-Administrator noted similar concern expressed by some members
of the CASAC who support retaining the existing standards; they
highlighted similar uncertainties and limitations in the risk
assessment (Cox, 2019b). In light of all of this, the then-
Administrator judged it appropriate to place little weight on
quantitative estimates of PM2.5-associated mortality risk in
reaching conclusions about the level of the primary PM2.5
standards (85 FR 82717, December 18, 2020).
The then-Administrator additionally considered an emerging body of
evidence from accountability studies that examined past reductions in
ambient PM2.5 and the degree to which those reductions
resulted in public health improvements. While the then-Administrator
agreed with public commenters that well-designed and conducted
accountability studies can be informative, he viewed the interpretation
of such studies in the context of the primary PM2.5
standards as complicated by the fact that some of the available studies
had not evaluated PM2.5 specifically (e.g., as opposed to
PM10 or total suspended particulates), did not show changes
in PM2.5 air quality, or had not been able to disentangle
health impacts of the interventions from background trends in health
(U.S. EPA, 2020b, section 3.5.1). He further recognized that the small
number of available studies that did report public health improvements
following past declines in ambient PM2.5 had not examined
air quality meeting the existing standards (U.S. EPA, 2020b, Table 3-
3). This included U.S. studies that reported increased life expectancy,
decreased mortality, and decreased respiratory effects following past
declines in ambient PM2.5 concentrations. Such studies
examined ``starting'' annual average PM2.5 concentrations
(i.e., prior to the reductions being evaluated) ranging from about 13.2
to >20[micro]g/m\3\ (i.e., U.S. EPA, 2020b, Table 3-3). Given the lack
of available accountability studies
[[Page 16223]]
reporting public health improvements attributable to reductions in
ambient PM2.5 in locations meeting the existing standards,
together with his broader concerns regarding the lack of experimental
studies examining PM2.5 exposures typical of areas meeting
the existing standards, the then-Administrator judged that there was
considerable uncertainty in the potential for increased public health
protection from further reductions in ambient PM2.5
concentrations beyond those achieved under the existing primary
PM2.5 standards (85 FR 82717, December 18, 2020).
When the above considerations were taken together, the then-
Administrator concluded that the scientific evidence assessed in the
2019 ISA, together with the analyses in the 2020 PA based on that
evidence and consideration of CASAC advice and public comments, did not
call into question the adequacy of the public health protection
provided by the existing annual and 24-hour PM2.5 standards.
In particular, the then-Administrator judged that there was
considerable uncertainty in the potential for additional public health
improvements from reducing ambient PM2.5 concentrations
below the concentrations achieved under the existing primary standards
and that, therefore, standards more stringent than the existing
standards (e.g., with lower levels) were not supported. That is, he
judged that more stringent standards would be more than requisite to
protect the public health with an adequate margin of safety. This
judgment reflected the Administrator's consideration of the
uncertainties in the potential implications of the lower end of the air
quality distributions from the epidemiologic studies due in part to the
lack of supporting evidence from experimental studies and retrospective
accountability studies conducted at PM2.5 concentrations
meeting the existing standards (85 FR 82717, December 18, 2020).
In reaching this conclusion in the 2020 review, the then-
Administrator judged that the existing standards provided an adequate
margin of safety. With respect to the annual standard, the level of 12
[micro]g/m\3\ was below the lowest ``starting'' concentration (i.e.,
13.2 [micro]g/m\3\) in the available accountability studies that showed
public health improvements attributable to reductions in ambient
PM2.5. In addition, while the then-Administrator placed less
weight on the epidemiologic evidence for selecting a standard, he noted
that the level of the annual standard was below the reported mean (and
median) concentrations in the majority of the key U.S. epidemiologic
studies using ground-based monitoring data (noting that these means
tend to be 10-20% lower than their corresponding area design values
which is the more relevant metric when considering the level of the
standard) and below the mean of the reported means (or medians) of
these studies (i.e., 13.5 [micro]g/m\3\). In addition, the then-
Administrator recognized that concentrations in areas meeting the
existing 24-hour and annual standards remained well-below the
PM2.5 exposure concentrations consistently shown to elicit
effects in human exposure studies (85 FR 82717-82718, December 18,
2020).
In addition, based on the then-Administrator's review of the
science in the 2020 review, including controlled human exposure studies
examining effects following short-term PM2.5 exposures, the
epidemiologic studies, and accountability studies conducted at levels
just above the existing annual standard, he judged that the degree of
public health protection provided by the existing annual standard is
not greater than warranted. This judgment, together with the fact that
no CASAC member expressed support for a less stringent standard, led
the then- Administrator to conclude that standards less stringent than
the existing standards (e.g., with higher levels) were also not
supported (85 FR 82718, December 18, 2020).
In reaching his final decision in the 2020 review, the then-
Administrator concluded that the scientific evidence and technical
information continued to support the existing annual and 24-hour
PM2.5 standards. This conclusion reflected the then-
Administrator's view that there were important limitations and
uncertainties that remained in the evidence. The then-Administrator
concluded that these limitations contributed to considerable
uncertainty regarding the potential public health implications of
revising the existing primary PM2.5 standards. Given this
uncertainty, and noting the advice from some CASAC members, he
concluded that the primary PM2.5 standards, including the
indicators (PM2.5), averaging times (annual and 24-hour),
forms (arithmetic mean and 98th percentile, averaged over three years)
and levels (12.0 [micro]g/m\3\, 35 [micro]g/m\3\), when taken together,
remained requisite to protect the public health. Therefore, in the 2020
review, the Administrator reached the conclusion that the primary 24-
hour and annual PM2.5 standards, together, were requisite to
protect public health from fine particles with an adequate margin of
safety, including the health of at-risk populations, and retained the
standards, without revision (85 FR 82718, December 18, 2020).
2. Overview of the Health Effects Evidence
The information summarized here and further detailed in section
II.B of the proposal (88 FR 5580, January 27, 2023), is an overview of
the policy-relevant aspects of the health effects evidence available in
this reconsideration; the assessment of this evidence is documented in
the 2019 ISA (U.S. EPA, 2019a) and ISA Supplement (U.S. EPA, 2022a) and
its policy implications are further discussed in the 2022 PA (U.S. EPA,
2022b). While the 2019 ISA provides the broad scientific foundation for
this reconsideration, additional literature has become available since
the cutoff date of the 2019 ISA that expands the body of evidence
related to mortality and cardiovascular effects for both short- and
long-term PM2.5 exposure, which can inform the
Administrator's judgment on the adequacy of the current primary
PM2.5 standards. As such, the ISA Supplement builds on the
information presented within the 2019 ISA with a targeted
identification and evaluation of new scientific information (U.S. EPA,
2022a, section 1.2). The ISA Supplement focuses on PM2.5
health effects evidence where the 2019 ISA concludes a ``causal
relationship,'' because such health effects are given the most weight
in an Administrator's decisions in a NAAQS review. As such, in
selecting the health effects to evaluate within the ISA Supplement
(i.e., newly available evidence related to short- and long-term
PM2.5 exposure and mortality and cardiovascular effects),
the primary rationale is based on the causality determinations for
health effect categories presented in the 2019 PM ISA, and the
subsequent use of the health effects evidence in the 2020 PM PA.
Specifically, U.S. and Canadian epidemiologic studies for mortality and
cardiovascular effects, along with controlled human exposure studies
associated with cardiovascular effects at near ambient concentrations,
were considered to be of greatest utility in informing the
Administrator's conclusions on the adequacy of the current primary
PM2.5 standards. Additionally, studies examining
associations outside the U.S. or Canada reflect air quality and
exposure patterns that may be less typical of the U.S., and thus less
likely to be informative for purposes of reviewing the NAAQS (U.S. EPA,
2022b, p.1-3). While the ISA Supplement does not include information
for health effects other than mortality and cardiovascular effects, the
[[Page 16224]]
scientific evidence for other health effect categories is evaluated in
the 2019 ISA, which in combination with the ISA Supplement represents
the complete scientific record for the reconsideration of the 2020
final decision.
The ISA Supplement also assessed accountability studies because
these types of epidemiologic studies were part of the body of evidence
that was a focus of the 2020 review. Accountability studies inform our
understanding of the potential for public health improvements as
ambient PM2.5 concentrations have declined over time.
Further, the ISA Supplement considered studies that employed
statistical approaches that attempt to more extensively account for
confounders and are more robust to model misspecification (i.e., used
alternative methods for confounder control),\50\ given that such
studies were highlighted by the CASAC and identified in public comments
in the 2020 review. Since the literature cutoff date for the 2019 ISA,
multiple accountability studies and studies that employ alternative
methods for confounder control have become available for consideration
in the ISA Supplement and, subsequently, in this reconsideration.
---------------------------------------------------------------------------
\50\ As noted in the ISA Supplement (U.S. EPA, 2022a, p. 1-3):
``In the peer-reviewed literature, these epidemiologic studies are
often referred to as causal inference studies or studies that used
causal modeling methods. For the purposes of this Supplement, this
terminology is not used to prevent confusion with the main
scientific conclusions (i.e., the causality determinations)
presented within an ISA. In addition, as is consistent with the
weight-of-evidence framework used within ISAs and discussed in the
Preamble to the Integrated Science Assessments, an individual study
on its own cannot inform causality, but instead represents a piece
of the overall body of evidence.''
---------------------------------------------------------------------------
The ISA Supplement also considered recent health effects evidence
that addresses key scientific issues where the literature has expanded
since the completion of the 2019 ISA.\51\ The 2019 ISA evaluated a
couple of controlled human exposure studies that investigated the
effect of exposure to near-ambient concentrations of PM2.5
(U.S. EPA, 2019a, section 6.1.10 and 6.1.13). The ISA Supplement adds
to this limited evidence, including a recent study conducted in young
healthy individuals exposed to near-ambient PM2.5
concentrations (U.S. EPA, 2022a, section 3.3.1). Given the importance
of identifying populations at increased risk of PM2.5-
related effects, the ISA Supplement also included epidemiologic or
exposure studies that examined whether there is evidence of exposure or
risk disparities by race/ethnicity or SES. These types of studies
provide additional information related to factors that may increase
risk of PM2.5-related health effects and provide additional
evidence for consideration by the Administrator in reaching conclusions
regarding the adequacy of the current standards. In addition, the ISA
Supplement evaluated studies that examined the relationship between
short- and long-term PM2.5 exposures and SARS-CoV-2
infection and/or COVID-19 death, as these studies are a new area of
research and were raised by a number of public commenters in the 2020
review.
---------------------------------------------------------------------------
\51\ As with the epidemiologic studies for long- and short-term
PM2.5 exposure and mortality and cardiovascular effects,
epidemiologic studies of exposure or risk disparities and SARS-CoV-2
infection and/or COVID-19 death were limited to those conducted in
the U.S. and Canada.
---------------------------------------------------------------------------
The evidence presented within the 2019 ISA, along with the targeted
identification and evaluation of new scientific information in the ISA
Supplement, provides the scientific basis for the reconsideration of
the 2020 final decision on the primary PM2.5 standards. The
subsections below briefly summarize the nature of PM2.5-
related health effects (II.A.2.a), with a focus on those health effects
for which the 2019 ISA concluded a ``causal'' or ``likely to be
causal'' relationship, the potential public health implications and
populations at risk (II.A.2.b), and PM2.5 concentrations in
key studies reporting health effects (II.A.2.c).
a. Nature of Effects
The evidence base available in the reconsideration includes decades
of research on PM2.5-related health effects (U.S. EPA,
2004b; U.S. EPA, 2009a; U.S. EPA, 2019a), including the full body of
evidence evaluated in the 2019 ISA (U.S. EPA, 2019a), along with the
targeted evaluation of recent evidence in the ISA Supplement (U.S. EPA,
2022a). In considering the available scientific evidence, the sections
below, and in more detail in section II.B.1 of the proposal (88 FR
5580, January 27, 2023), summarize the relationships between long-and
short-term PM2.5 exposures and mortality (II.A.2.a.i),
cardiovascular effects (II.A.2.a.ii), respiratory effects
(II.A.2.a.iii), cancer (II.A.2.a.iv), nervous system effects
(II.A.2.a.v) and other effects (II.A.2.a.vi). For these outcomes, the
2019 ISA concluded that the evidence supports either a ``causal'' or a
``likely to be causal'' relationship.\52\
---------------------------------------------------------------------------
\52\ In this reconsideration of the PM NAAQS, the EPA considers
the full body of health evidence, placing the greatest emphasis on
the health effects for which the evidence has been judged in the
2019 ISA to demonstrate a ``causal'' or ``likely to be causal''
relationship with PM2.5 exposures.
---------------------------------------------------------------------------
i. Mortality
Long-Term PM2.5 Exposures
In the 2012 review, the 2009 ISA reported that the evidence was
``sufficient to conclude that the relationship between long-term
PM2.5 exposures and mortality is causal'' (U.S. EPA, 2009a,
p. 7-96). The strongest evidence supporting this conclusion was
provided by epidemiologic studies, particularly those examining two
seminal cohorts, the American Cancer Society (ACS) cohort and the
Harvard Six Cities cohort. Analyses of the Harvard Six Cities cohort
included evidence indicating that reductions in ambient
PM2.5 concentrations are associated with reduced mortality
risk (Laden et al., 2006) and increases in life expectancy (Pope et
al., 2009). Further support was provided by other cohort studies
conducted in North America and Europe that reported positive
associations between long-term PM2.5 exposure and mortality
(U.S. EPA, 2019a).
Cohort studies, which have become available since the completion of
the 2009 ISA and evaluated in the 2019 ISA, continue to provide
consistent evidence of positive associations between long-term
PM2.5 exposures and mortality. These studies add support for
associations with all-cause and total (non-accidental) mortality,\53\
as well as with specific causes of mortality, including cardiovascular
disease and respiratory disease (U.S. EPA, 2019a, section 11.2.2).
Several of these studies conducted analyses over longer study durations
and periods of follow-up than examined in the original ACS and Harvard
Six Cities cohort studies and continue to report positive associations
between long-term exposure to PM2.5 and mortality (U.S. EPA,
2019a, section 11.2.2.1; Figures 11-18 and 11-19). In addition to
studies focusing on the ACS and Harvard Six Cities cohorts, additional
studies examining other cohorts also provide evidence of consistent,
positive associations between long-term PM2.5 exposure and
mortality across a wide range of demographic groups (e.g., age, sex,
occupation), spatial and temporal extents, exposure assessment metrics,
and statistical techniques (U.S. EPA, 2019a, sections 11.2.2.1, 11.2.5;
U.S. EPA, 2022a, Table 11-8). This includes some of the largest cohort
studies conducted to date, such as analyses of the U.S. Medicare cohort
that includes
[[Page 16225]]
nearly 61 million enrollees and studies that control for a range of
individual and ecological covariates, including race, age, SES, smoking
status, body mass index, and annual weather variables (e.g.,
temperature, humidity) (U.S. EPA, 2019a).
---------------------------------------------------------------------------
\53\ The majority of these studies examined non-accidental
mortality outcomes, though some Medicare studies lack cause-specific
death information and, therefore, examine total mortality.
---------------------------------------------------------------------------
In addition to those cohort studies evaluated in the 2019 ISA,
recent North American cohort studies evaluated in the ISA Supplement
continue to examine the relationship between long-term PM2.5
exposure and mortality and report consistent, positive, and
statistically significant associations. These recent studies also
utilize large and demographically diverse cohorts that are generally
representative of the national populations in both the U.S. and Canada.
These ``studies published since the 2019 ISA support and extend the
evidence base that contributed to the conclusion of a causal
relationship between long-term PM2.5 exposure and
mortality'' (U.S. EPA, 2022a, section 3.2.2.2.1, Figure 3-19, Figure 3-
20).
Furthermore, studies evaluated in the 2019 ISA and the ISA
Supplement that examined cause-specific mortality expand upon previous
research that found consistent, positive associations between
PM2.5 exposure and specific mortality outcomes, which
include cardiovascular and respiratory mortality, as well as other
mortality outcomes. For cardiovascular-related mortality, the evidence
evaluated in the ISA Supplement is consistent with the evidence
evaluated in the 2019 ISA with recent studies reporting positive
associations with long-term PM2.5 exposure. When evaluating
cause-specific cardiovascular mortality, recent studies reported
positive associations for a number of outcomes, such as ischemic heart
disease (IHD) and stroke mortality (U.S. EPA, 2022a, Figure 3-23).
Moreover, recent studies also provide some initial evidence that
individuals with pre-existing health conditions, such as heart failure
and diabetes, are at an increased risk of PM2.5-related
health effects (U.S. EPA, 2022a, section 3.2.2.4) and that these
individuals have a higher risk of mortality overall, which was
previously only examined in studies that used stratified analyses
rather than a cohort of people with an underlying health condition
(U.S. EPA, 2022a, section 3.2.2.4). With regard to respiratory
mortality, epidemiologic studies evaluated in the 2019 ISA and ISA
Supplement continue to provide support for associations between long-
term PM2.5 exposure and respiratory mortality (U.S. EPA,
2019a, section 5.2.10; U.S. EPA, 2022a, Table 3-2).
A series of epidemiologic studies evaluated in the 2019 ISA tested
the hypothesis that past reductions in ambient PM2.5
concentrations are associated with increased life expectancy or a
decreased mortality rate and report that reductions in ambient
PM2.5 are associated with improvements in longevity (U.S.
EPA, 2022a, section 11.2.2.5). Pope et al. (2009) conducted a cross-
sectional analysis using air quality data from 51 metropolitan areas
across the U.S., beginning in the 1970s through the early 2000s, and
found that a 10 [micro]g/m\3\ decrease in long-term PM2.5
concentration was associated with a 0.61-year increase in life
expectancy. In a subsequent analysis, the authors extended the period
of analysis to include 2000 to 2007, a time period with lower ambient
PM2.5 concentrations and found a decrease in long-term
PM2.5 concentration continued to be associated with an
increase in life expectancy, though the magnitude of the increase was
smaller than during the earlier time period (i.e., a 10 [micro]g/m\3\
decrease in long-term PM2.5 concentration was associated
with a 0.35-year increase in life expectancy) (Correia et al., 2013).
Additional studies conducted in the U.S. or Europe similarly report
that reductions in ambient PM2.5 are associated with
improvements in longevity (U.S. EPA, 2022a, section 11.2.2.5).
Since the literature cutoff date for the 2019 ISA, a few
epidemiologic studies were published that examined the relationship
between long-term PM2.5 exposure and life-expectancy (U.S.
EPA, 2022a, section 3.2.1.3) and report results that are consistent
with and expand upon the body of evidence from the 2019 ISA. For
example, Bennett et al. (2019) reported that PM2.5
concentrations above the lowest observed concentration (2.8 [micro]g/
m\3\) were associated with a 0.15 year decrease in national life
expectancy for women and 0.13 year decrease in national life expectancy
for men (U.S. EPA, 2022a, section 3.2.2.2.4, Figure 3-25). Another
study compared participants living in areas with PM2.5
concentrations >12 [micro]g/m\3\ to participants living in areas with
PM2.5 concentrations <12 [micro]g/m\3\ and reported that the
number of years of life lost due to living in areas with higher
PM2.5 concentrations was 0.84 years over a 5-year period
(Ward-Caviness et al., 2020; U.S. EPA, 2022a, section 3.2.2.2.4).
Additionally, a number of accountability studies, which are
epidemiologic studies that evaluate whether an environmental policy or
air quality intervention resulted in reductions in ambient air
pollution concentrations and subsequent reductions in mortality or
morbidity, have emerged and were evaluated in the ISA Supplement (U.S.
EPA, 2022a, section 3.2.2.3). For example, Sanders et al. (2020a)
examined whether policy actions (i.e., the first annual
PM2.5 NAAQS implementation rule in 2005 for the 1997 annual
PM2.5 standard with a 3-year annual average of 15.0 [mu]g/
m\3\) reduced PM2.5 concentrations and mortality rates in
Medicare beneficiaries between 2000-2013, and found that following
implementation of the annual PM2.5 NAAQS, annual
PM2.5 concentrations decreased by 1.59 [mu]g/m\3\ (95% CI:
1.39, 1.80) which corresponded to a 0.93% reduction in mortality rates
among individuals 65 years and older ([95% CI: 0.10%, 1.77%) in non-
attainment counties relative to attainment counties.
The 2019 ISA also evaluated a small number of studies that used
alternative methods for confounder control to further assess
relationship between long-term PM2.5 exposure and mortality
(U.S. EPA, 2019a, section 11.2.2.4). In addition, multiple
epidemiologic studies that implemented alternative methods for
confounder control and were published since the literature cutoff date
of the 2019 ISA were evaluated in the ISA Supplement (U.S. EPA, 2022a,
section 3.2.2.3). These studies used a variety of statistical methods
including generalized propensity score (GPS), inverse probability
weighting (IPW), and difference-in-difference (DID) to reduce
uncertainties related to confounding bias in the association between
long-term PM2.5 exposure and mortality. These studies
reported consistent positive associations between long-term
PM2.5 exposure and total mortality (U.S. EPA, 2022a, section
3.2.2.3), and provided further support for the associations reported in
the cohort studies referenced above.
The 2019 ISA and ISA Supplement also evaluated the degree to which
recent studies examining the relationship between long-term
PM2.5 exposure and mortality addressed key policy-relevant
issues and/or previously identified data gaps in the scientific
evidence, including methods to estimate exposure, methods to control
for confounding (e.g., co-pollutant confounding), the shape of the C-R
relationship, as well as examining whether a threshold exists below
which mortality effects do not occur. With respect to exposure
assessment, based on its evaluation of the evidence, the 2019 ISA
concludes that positive associations between long-term PM2.5
exposures and mortality are robust
[[Page 16226]]
across recent analyses using various approaches to estimate
PM2.5 exposures (e.g., based on monitors, models, satellite-
based methods, or hybrid methods that combine information from multiple
sources) (U.S. EPA, 2019a, section 11.2.5.1). Hart et al. (2015) report
that correction for bias due to exposure measurement error increases
the magnitude of the hazard ratios (confidence intervals widen but the
association remains statistically significant), suggesting that failure
to correct for exposure measurement error could result in attenuation
or underestimation of risk estimates.
The 2019 ISA additionally concludes that positive associations
between long-term PM2.5 exposures and mortality are robust
across statistical models that use different approaches to control for
confounders or different sets of confounders (U.S. EPA, 2019a, sections
11.2.3 and 11.2.5), across diverse geographic regions and populations,
and across a range of temporal periods including periods of declining
PM concentrations (U.S. EPA, 2019a, sections 11.2.2.5 and 11.2.5.3).
Additional evidence further demonstrates that associations with
mortality remain robust in copollutants analyses (U.S. EPA, 2019a,
section 11.2.3), and that associations persist in analyses restricted
to long-term exposures (annual average PM2.5 concentrations)
below 12 [mu]g/m\3\ (Di et al., 2017b) or 10 [mu]g/m\3\ (Shi et al.,
2016), indicating that risks are not disproportionately driven by the
upper portions of the air quality distribution. Recent studies
evaluated in the ISA Supplement further assess potential copollutant
confounding and indicate that while there is some evidence of potential
confounding of the PM2.5-mortality association by
copollutants in some of the studies (i.e., those studies of the
Mortality Air Pollution Associations in Low Exposure Environments
(MAPLE) cohort), this result is inconsistent with other recent studies
evaluated in the 2019 ISA that were conducted in the U.S. and Canada
that found associations in both single and copollutant models (U.S.
EPA, 2019a; U.S. EPA, 2022a, section 3.2.2.4)
Additionally, a few studies use statistical techniques to reduce
uncertainties related to potential confounding to further inform
conclusions on causality for long-term PM2.5 exposure and
mortality, as further detailed in section II.B.1.a.i of the proposal
(88 FR 5582, January 27, 2023), studies by Greven et al. (2011), Pun et
al. (2017), and Eum et al. (2018) completed sensitivity analyses as
part of their Medicare cohort study in which they decompose ambient
PM2.5 into ``spatial'' and ``spatiotemporal'' components in
order to evaluate the potential for bias due to unmeasured spatial
confounding. Pun et al. (2017) observed positive associations for the
``temporal'' variation model and approximately null associations for
the ``spatiotemporal'' variation model for all causes of death except
for COPD mortality. The difference in the results of these two models
for most causes of death suggests the presence of unmeasured
confounding, though the authors do not indicate anything about the
direction or magnitude of this bias. It is important to note that the
``temporal'' and ``spatiotemporal'' coefficients are not directly
comparable to the results of other epidemiologic studies when examined
individually and can only be used in comparison with one another to
evaluate the potential for unmeasured confounding bias. Eum et al.
(2018) and Wu et al. (2020) also attempted to address long-term trends
and meteorological variables as potential confounders and found that
not adjusting for temporal trends could overestimate the association,
while effect estimates in analyses that excluded meteorological
variables remained unchanged compared to the main analyses. While
results of these analyses suggest the presence of some unmeasured
confounding, they do not indicate the direction or magnitude of the
bias.\54\
---------------------------------------------------------------------------
\54\ In public comments on the 2019 draft PA, the authors of the
Pun et al. (2017) study further note that ``the presence of
unmeasured confounding. . .was expected given that we did not
control for several potential confounders that may impact
PM2.5-mortality associations, such as smoking, socio-
economic status (SES), gaseous pollutants, PM2.5
components, and long-term time trends in PM2.5'' and that
``spatial confounding may bias mortality risks both towards and away
from the null'' (Docket ID EPA-HQ-OAR-2015-0072-0065; accessible in
https://www.regulations.gov/).
---------------------------------------------------------------------------
An additional important consideration in characterizing the public
health impacts associated with PM2.5 exposure is whether C-R
relationships are linear across the range of concentrations or if
nonlinear relationships exist along any part of this range. Studies
evaluated in the 2019 ISA and the ISA Supplement examine this issue,
and continue to provide evidence of linear, no-threshold relationships
between long-term PM2.5 exposures and all-cause and cause-
specific mortality (U.S. EPA, 2019a, section 11.2.4; U.S. EPA, 2022a,
section 3.2.2.2.7, Table 3-6). Across the studies evaluated in the 2019
ISA and the ISA Supplement, a variety of statistical methods have been
used to assess whether there is evidence of deviations in linearity
(U.S. EPA, 2019a, Table 11-7; U.S. EPA, 2022a, section 2.2.3.2).
Studies have also conducted cut-point analyses that focus on examining
risk at specific ambient PM2.5 concentrations. Generally,
the evidence remains consistent in supporting a no-threshold
relationship, and in supporting a linear relationship for
PM2.5 concentrations >8 [mu]g/m\3\. However, uncertainties
remain about the shape of the C-R curve at PM2.5
concentrations <8 [mu]g/m\3\, with some recent studies providing
evidence for either a sublinear, linear, or supralinear relationship at
these lower concentrations (U.S. EPA, 2019a, section 11.2.4; U.S. EPA,
2022a, section 2.2.3.2). There was also some limited evidence
indicating that the slope of the C-R function may be steeper
(supralinear) at lower concentrations for cardiovascular mortality
(U.S. EPA, 2022a, section 3.1.1.2.6).
The biological plausibility of PM2.5-attributable
mortality is supported by the coherence of effects across scientific
disciplines (i.e., animal toxicological, controlled human exposure
studies, and epidemiologic) when evaluating respiratory and
cardiovascular morbidity effects, which are some of the largest
contributors to total (nonaccidental) mortality. The 2019 ISA outlines
the available evidence for biologically plausible pathways by which
inhalation exposure to PM2.5 could progress from initial
events (e.g., pulmonary inflammation, autonomic nervous system
activation) to endpoints relevant to population outcomes, particularly
those related to cardiovascular diseases such as ischemic heart
disease, stroke and atherosclerosis (U.S. EPA, 2019a, section 6.2.1),
and to metabolic effects, including diabetes (U.S. EPA, 2019a, section
7.3.1). The 2019 ISA notes ``more limited evidence from respiratory
morbidity'' (U.S. EPA, 2019a, p. 11-101) such as development of chronic
obstructive pulmonary disease (COPD) (U.S. EPA, 2019a, section 5.2.1)
to support the biological plausibility of mortality due to long-term
PM2.5 exposures (U.S. EPA, 2019a, section 11.2.1).
Taken together, epidemiologic studies, including those evaluated in
the 2019 ISA and more recent studies evaluated in the ISA Supplement,
consistently report positive associations between long-term
PM2.5 exposure and mortality across different geographic
locations, populations, and analytic approaches (U.S. EPA, 2019a; U.S.
EPA, 2022a, section 3.2.2.4). As such, these studies reduce key
uncertainties identified in previous reviews, including those related
to potential
[[Page 16227]]
copollutant confounding, and provide additional information on the
shape of the C-R curve. As evaluated in the 2019 ISA, experimental and
epidemiologic evidence for cardiovascular effects, and respiratory
effects to a more limited degree, supports the plausibility of
mortality due to long-term PM2.5 exposures. Overall, studies
evaluated in the 2019 ISA support the conclusion of a causal
relationship between long-term PM2.5 exposure and mortality,
which is supported and extended by evidence from recent epidemiologic
studies evaluated in the ISA Supplement (U.S. EPA, 2022a, section
3.2.2.4).
Short-Term PM2.5 Exposures
The 2009 ISA concluded that ``a causal relationship exists between
short-term exposure to PM2.5 and mortality'' (U.S. EPA,
2009a). This conclusion was based on the evaluation of both multi- and
single-city epidemiologic studies that consistently reported positive
associations between short-term PM2.5 exposure and non-
accidental mortality. These associations were strongest, in terms of
magnitude and precision, primarily at lags of 0 to 1 days. Examination
of the potential confounding effects of gaseous copollutants was
limited, though evidence from single-city studies indicated that
gaseous copollutants have minimal effect on the PM2.5-
mortality relationship (i.e., associations remain robust to inclusion
of other pollutants in copollutant models). The evaluation of cause-
specific mortality found that effect estimates were larger in
magnitude, but also had larger confidence intervals, for respiratory
mortality compared to cardiovascular mortality. Although the largest
mortality risk estimates were for respiratory mortality, the
interpretation of the results was complicated by the limited coherence
from studies of respiratory morbidity. However, the evidence from
studies of cardiovascular morbidity provided both coherence and
biological plausibility for the relationship between short-term
PM2.5 exposure and cardiovascular mortality.
Multicity studies evaluated in the 2019 ISA and the ISA Supplement
provide evidence of primarily positive associations between daily
PM2.5 exposures and mortality, with percent increases in
total mortality ranging from 0.19% (Lippmann et al., 2013) to 2.80%
(Kloog et al., 2013) \55\ at lags of 0 to 1 days in single-pollutant
models. Whereas many studies assign exposures using data from ambient
monitors, other studies employ hybrid modeling approaches, which
estimate PM2.5 concentrations using data from a variety of
sources (i.e., from satellites, land use information, and modeling, in
addition to monitors) and enable the inclusion of less urban and more
rural locations in analyses (e.g., Kloog et al., 2013, Lee et al.,
2015, Shi et al., 2016).
---------------------------------------------------------------------------
\55\ As detailed in the Preface to the ISA, risk estimates are
for a 10 [micro]g/m\3\ increase in 24-hour avg PM2.5
concentrations, unless otherwise noted (U.S. EPA, 2019a).
---------------------------------------------------------------------------
Some studies have expanded the examination of potential confounders
including long-term temporal trends, weather, and co-occurring
pollutants. Mortality associations were found to remain positive,
although in some cases were attenuated, when using different approaches
to account for temporal trends or weather covariates (e.g., U.S. EPA,
2019a, section 11.1.5.1). For example, Sacks et al. (2012) examined the
influence of model specification using the approaches for confounder
adjustment from models employed in several multicity studies within the
context of a common data set (U.S. EPA, 2019a, section 11.1.5.1). These
models use different approaches to control for long-term temporal
trends and the potential confounding effects of weather. The authors
report that associations between daily PM2.5 and
cardiovascular mortality were similar across models, with the percent
increase in mortality ranging from 1.5-2.0% (U.S. EPA, 2019a, Figure
11-4). Thus, alternative approaches to controlling for long-term
temporal trends and for the potential confounding effects of weather
may influence the magnitude of the association between PM2.5
exposures and mortality but have not been found to influence the
direction of the observed association (U.S. EPA, 2019a, section
11.1.5.1). Taken together, the 2019 ISA and the ISA Supplement conclude
that recent multicity studies conducted in the U.S., Canada, Europe,
and Asia continue to provide consistent evidence of positive
associations between short-term PM2.5 exposures and total
mortality across studies that use different approaches to control for
the potential confounding effects of weather (e.g., temperature) (U.S.
EPA, 2019a, section 1.4.1.5.1; U.S. EPA, 2022a, section 3.2.1.2).
With regard to copollutants, studies evaluated in the 2019 ISA
provide additional evidence that associations between short-term
PM2.5 exposures and mortality remain positive and relatively
unchanged in copollutant models with both gaseous pollutants and
PM10-2.5 (U.S. EPA, 2019a, section 11.1.4). Additionally,
the low (r < 0.4) to moderate correlations (r = 0.4-0.7) between
PM2.5 and gaseous pollutants and PM10-2.5
increase the confidence in PM2.5 having an independent
effect on mortality (U.S. EPA, 2019a, section 11.1.4). Consistent with
the studies evaluated in the 2019 ISA, studies evaluated in the ISA
Supplement that used data from more recent years also indicate that
associations between short-term PM2.5 exposure and mortality
remain unchanged in copollutant models. However, the evidence indicates
that the association could be larger in magnitude in the presence of
some copollutants such as oxidant gases (Lavigne et al., 2018; Shin et
al., 2021).
The generally positive associations reported with mortality are
supported by a small group of studies employing alternative methods for
confounder control or quasi-experimental statistical approaches (U.S.
EPA, 2019a, section 11.1.2.1). For example, two studies by Schwartz et
al. report associations between PM2.5 instrumental variables
and mortality (U.S. EPA, 2019a, Table 11-2), including in an analysis
limited to days with 24-hour average PM2.5 concentrations
<30 [mu]g/m\3\ (Schwartz et al., 2015; Schwartz et al., 2017). In
addition to the main analyses, these studies conducted Granger-like
causality tests as sensitivity analyses to examine whether there was
evidence of an association between mortality and PM2.5 after
the day of death, which would support the possibility that unmeasured
confounders were not accounted for in the statistical model. Neither
study reports evidence of an association with PM2.5 after
death (i.e., they do not indicate unmeasured confounding). Yorifuji et
al. (2016) conducted a quasi-experimental study to examine whether a
specific regulatory action in Tokyo, Japan (i.e., a diesel emission
control ordinance) resulted in a subsequent reduction in daily
mortality (Yorifuji et al., 2016). The authors reported a reduction in
mortality in Tokyo due to the ordinance, compared to Osaka, which did
not have a similar diesel emission control ordinance in place. In
another study, Schwartz et al. (2018) utilized three statistical
methods including instrumental variable analysis, a negative exposure
control, and marginal structural models to estimate the association
between PM2.5 and daily mortality (Schwartz et al., 2018).
Results from this study continue to support a relationship between
short-term PM2.5 exposure and mortality. Additional
epidemiologic studies evaluated in the ISA Supplement that employed
alternative methods for confounder control to examine the association
between short-term PM2.5 exposure and
[[Page 16228]]
mortality also report consistent positive associations in studies that
examine effects across multiple cities in the U.S. (U.S. EPA, 2022a).
The positive associations for total mortality reported across the
majority of studies evaluated are further supported by cause-specific
mortality analyses, which generally report consistent, positive
associations with both cardiovascular and respiratory mortality (U.S.
EPA, 2019a, section 11.1.3). Recent multicity studies evaluated in the
ISA Supplement add to the body of evidence indicating a relationship
between short-term PM2.5 exposure and cause-specific
mortality, with more variability in the magnitude and precision of
associations for respiratory mortality (U.S. EPA, 2022a; Figure 3-14).
For both cardiovascular and respiratory mortality, there has been a
limited assessment of potential copollutant confounding, though initial
evidence indicates that associations remain positive and relatively
unchanged in models with gaseous pollutants and PM10-2.5,
which further supports the copollutant analyses conducted for total
mortality. The strong evidence for ischemic events and heart failure,
as detailed in the assessment of cardiovascular morbidity (U.S. EPA,
2019a, Chapter 6), provides biological plausibility for
PM2.5-related cardiovascular mortality, which comprises the
largest percentage of total mortality (i.e., ~33%) (NHLBI, 2017).
Although there is evidence for exacerbations of COPD and asthma, the
collective body of respiratory morbidity evidence provides limited
biological plausibility for PM2.5-related respiratory
mortality (U.S. EPA, 2019a, Chapter 5).
In the 2009 ISA, one of the main uncertainties identified was the
regional and city-to-city heterogeneity in PM2.5-mortality
associations. Studies evaluated in the 2019 ISA examine both city-
specific as well as regional characteristics to identify the underlying
contextual factors that could contribute to this heterogeneity (U.S.
EPA, 2019a, section 11.1.6.3). Analyses focusing on effect modification
of the PM2.5 mortality relationship by PM2.5
components, regional patterns in PM2.5 components and city
specific differences in composition and sources indicate some
differences in the PM2.5 composition and sources across
cities and regions, but these differences do not fully explain the
observed heterogeneity. Additional studies find that factors related to
potential exposure differences, such as housing stock and commuting, as
well as city specific factors (e.g., land use, port volume, and traffic
information), may also explain some of the observed heterogeneity (U.S.
EPA, 2019a, section 11.1.6.3). Collectively, studies evaluated in the
2019 ISA and the ISA Supplement indicate that the heterogeneity in
PM2.5 mortality risk estimates cannot be attributed to one
factor, but instead a combination of factors including, but not limited
to, PM composition and sources as well as community characteristics
that could influence exposures (U.S. EPA, 2019a, section 11.1.12; U.S.
EPA, 2022a, section 3.2.1.2.1).
A number of studies evaluated in the 2019 ISA and ISA Supplement
conducted systematic evaluations of the lag structure of associations
for the PM2.5-mortality relationship by examining either a
series of single day or multiday lags and these studies continue to
support an immediate effect (i.e., lag 0 to 1 days) of short-term
PM2.5 exposures on mortality (U.S. EPA, 2019a, section
11.1.8.1; U.S. EPA, 2022a, section 3.2.1.1). Recent studies also
conducted analyses comparing the traditional 24-hour average exposure
metric with a subdaily metric (i.e., 1-hour max) and provide evidence
of a similar pattern of associations for both the 24-hour average and
1-hour max metric, with the association larger in magnitude for the 24-
hour average metric.
Multicity studies indicate that positive and statistically
significant associations with mortality persist in analyses restricted
to short-term (24-hour average PM2.5 concentrations)
PM2.5 exposures below 35 [mu]g/m\3\ (Lee et al., 2015),\56\
below 30 [mu]g/m\3\ (Shi et al., 2016), and below 25 [mu]g/m\3\ (Di et
al., 2017a), indicating that risks associated with short-term
PM2.5 exposures are not disproportionately driven by the
peaks of the air quality distribution. Additional studies examined the
shape of the C-R relationship for short-term PM2.5 exposure
and mortality and whether a threshold exists below which mortality
effects do not occur (U.S. EPA, 2019a, section 11.1.10). These studies
used various statistical approaches and consistently demonstrate linear
C-R relationships with no evidence of a threshold.
---------------------------------------------------------------------------
\56\ Lee et al. (2015) restrict exposures below 35 [mu]g/m\3\
only in areas with annual average concentrations <12 [mu]g/m\3\.
Additionally, Lee et al. (2015) also report that positive and
statistically significant associations between short-term
PM2.5 exposures and mortality persist in analyses
restricted to areas with long-term concentrations below 12 [mu]g/
m\3\.
---------------------------------------------------------------------------
Moreover, recent studies evaluated in the ISA Supplement provide
additional support for a linear, no-threshold C-R relationship between
short-term PM2.5 exposure and mortality, with confidence in
the shape decreasing at concentrations below 5 [micro]g/m\3\ (Shi et
al., 2016; Lavigne et al., 2018). Recent analyses provide initial
evidence indicating that PM2.5-mortality associations
persist and may be stronger (i.e., a steeper slope) at lower
concentrations (e.g., Di et al., 2017a; Figure 11-12 in U.S. EPA,
2019). However, given the limited data available at the lower end of
the distribution of ambient PM2.5 concentrations, the shape
of the C-R curve remains uncertain at these low concentrations.
Although difficulties remain in assessing the shape of the short-term
PM2.5-mortality C-R relationship, to date, studies have not
conducted systematic evaluations of alternatives to linearity and
recent studies evaluated in the ISA Supplement continue to provide
evidence of a no-threshold linear relationship, with less confidence at
concentrations lower than 5 [micro]g/m\3\.
Overall, epidemiologic studies evaluated in the 2019 ISA and the
ISA Supplement build upon and extend the conclusions of the 2009 ISA
for the relationship between short-term PM2.5 exposures and
total mortality. Supporting evidence for PM2.5-related
cardiovascular morbidity, and more limited evidence from respiratory
morbidity, provide biological plausibility for mortality due to short-
term PM2.5 exposures. The primarily positive associations
observed across studies conducted in diverse geographic locations is
further supported by the results from copollutant analyses indicating
robust associations, along with evidence from analyses examining the C-
R relationship. Overall, studies evaluated in the 2019 ISA support the
conclusion of a causal relationship between short-term PM2.5
exposure and mortality, which is further supported by evidence from
recent epidemiologic studies evaluated in the ISA Supplement (U.S. EPA,
2022a, section 3.2.1.4, p. 3-69).
ii. Cardiovascular Effects
Long-Term PM2.5 Exposures
The scientific evidence reviewed in the 2009 ISA was ``sufficient
to infer a causal relationship between long-term PM2.5
exposure and cardiovascular effects'' (U.S. EPA, 2009a). The strongest
line of evidence comprised findings from several large epidemiologic
studies of U.S. and Canadian cohorts that reported consistent positive
associations between long-term PM2.5 exposure and
cardiovascular mortality (Pope et al., 2004; Krewski et al., 2009;
Miller et al.,
[[Page 16229]]
2007; Laden et al., 2006). Studies of long-term PM2.5
exposure and cardiovascular morbidity were limited in number.
Biological plausibility and coherence with the epidemiologic findings
were provided by studies using genetic mouse models of atherosclerosis
demonstrating enhanced atherosclerotic plaque development and
inflammation, as well as changes in measures of impaired heart
function, following 4- to 6-month exposures to PM2.5
concentrated ambient particles (CAPs), and by a limited number of
studies reporting CAPs-induced effects on coagulation factors, vascular
reactivity, and worsening of experimentally induced hypertension in
mice (U.S. EPA, 2009a).
Consistent with the evidence assessed in the 2009 ISA, the 2019 ISA
concludes that recent studies, together with the evidence available in
previous reviews, support a causal relationship between long-term
exposure to PM2.5 and cardiovascular effects. Additionally,
recent epidemiologic studies published since the completion of the 2019
ISA and evaluated in the ISA Supplement expands the body of evidence
and further supports such a conclusion (U.S. EPA, 2022a). As discussed
above (section II.A.2.a.i), results from U.S. and Canadian cohort
studies evaluated in the 2019 ISA conducted at varying spatial and
temporal scales and employing a variety of exposure assessment and
statistical methods consistently report positive associations between
long-term PM2.5 exposure and cardiovascular mortality (U.S.
EPA, 2019, Figure 6-19, section 6.2.10). Positive associations between
long-term PM2.5 exposures and cardiovascular mortality are
generally robust in copollutant models adjusted for ozone,
NO2, PM10-2.5, or SO2. In addition,
most of the results from analyses examining the shape of the C-R
relationship between long-term PM2.5 exposures and
cardiovascular mortality support a linear relationship and do not
identify a threshold below which mortality effects do not occur (U.S.
EPA, 2019a, section 6.2.16, Table 6-52).
The body of literature examining the relationship between long-term
PM2.5 exposure and cardiovascular morbidity has greatly
expanded since the 2009 ISA, with positive associations reported in
several cohorts evaluated in the 2019 ISA (U.S. EPA, 2019a, section
6.2). Though results for cardiovascular morbidity are less consistent
than those for cardiovascular mortality (U.S. EPA, 2019a, section 6.2),
studies in the 2019 ISA and the ISA Supplement provide some evidence
for associations between long-term PM2.5 exposures and the
progression of cardiovascular disease. Positive associations with
cardiovascular morbidity (e.g., coronary heart disease, stroke,
arrhythmias, myocardial infarction (MI), atherosclerosis progression)
are observed in several epidemiologic studies (U.S. EPA, 2019a,
sections 6.2.2 to 6.2.9; U.S. EPA, 2022a, section 3.1.2.2).
Additionally, studies evaluated in the ISA Supplement report positive
associations among those with pre-existing conditions, among patients
followed after a cardiac event procedure, and among those with a first
hospital admission for heart attacks among older adults enrolled in
Medicare (U.S. EPA, 2022a, sections 3.1.1 and 3.1.2).
Recent studies published since the literature cutoff date of the
2019 ISA and evaluated in the ISA Supplement further assessed the
relationship between long-term PM2.5 exposure and
cardiovascular effects by conducting accountability analyses or by
using alternative methods for confounder control in evaluating the
association between long-term PM2.5 exposure and
cardiovascular hospital admissions (U.S. EPA, 2022a, section 3.1.2.3).
Studies that apply alternative methods for confounder control increase
confidence in the relationship between long-term PM2.5
exposure and cardiovascular effects by using methods that reduce
uncertainties related to potential confounding through statistical and/
or study design approaches. For example, to control for potential
confounding Wei et al. (2021) used a doubly robust additive model
(DRAM) and found an association between long-term exposure to
PM2.5 and cardiovascular effects, including MI, stoke, and
atrial fibrillation, among the Medicare population. For example, an
accountability study by Henneman et al. (2019) utilized a difference-
in-difference (DID) approach to determine the relationship between
coal-fueled power plant emissions and cardiovascular effects and found
that reductions in PM2.5 concentrations resulted in
reductions of cardiovascular-related hospital admissions. Furthermore,
several recent epidemiologic studies evaluated in the ISA Supplement
reported that the association between long-term PM2.5
exposure with stroke persisted after adjustment for NO2 but
was attenuated in the model with O3 and oxidant gases
represented by the redox weighted average of NO2 and
O3 (U.S. EPA, 2022a, section 3.1.2.2.8). Overall, these
studies report consistent findings that long-term PM2.5
exposure is related to increased hospital admissions for a variety of
cardiovascular disease outcomes among large nationally representative
cohorts and provide additional support for a relationship between long-
term PM2.5 exposure and cardiovascular effects.
Positive associations reported in epidemiologic studies are
supported by toxicological evidence evaluated in the 2019 ISA. The
positive associations reported in epidemiologic studies are supported
by toxicological evidence for increased plaque progression in mice
following long-term exposure to PM2.5 collected from
multiple locations across the U.S. (U.S. EPA, 2019a, section 6.2.4.2).
A small number of epidemiologic studies also report positive
associations between long-term PM2.5 exposure and heart
failure, changes in blood pressure, and hypertension (U.S. EPA, 2019a,
sections 6.2.5 and 6.2.7). Associations with heart failure are
supported by animal toxicological studies demonstrating decreased
cardiac contractility and function, and increased coronary artery wall
thickness following long-term PM2.5 exposure (U.S. EPA,
2019a, section 6.2.5.2). Similarly, a limited number of animal
toxicological studies demonstrating a relationship between long-term
PM2.5 exposure and consistent increases in blood pressure in
rats and mice are coherent with epidemiologic studies reporting
positive associations between long-term exposure to PM2.5
and hypertension.
Additionally, a number of studies evaluated in the ISA Supplement
focusing on morbidity outcomes, including those that focused on
incidence of MI, atrial fibrillation (AF), stroke, and congestive heart
failure (CHF), expand the evidence pertaining to the shape of the C-R
relationship between long-term PM2.5 exposure and
cardiovascular effects. These studies use statistical techniques that
allow for departures from linearity (U.S. EPA, 2022a, Table 3-3), and
generally support the evidence characterized in the 2019 ISA showing
linear, no-threshold C-R relationship for most cardiovascular disease
(CVD) outcomes. However, there is evidence for a sublinear or
supralinear C-R relationship for some outcomes (U.S. EPA, 2022a,
section 3.1.2.2.9).\57\
---------------------------------------------------------------------------
\57\ As noted above for mortality, uncertainty in the shape of
the C-R relationship increases near the upper and lower ends of the
distribution due to limited data.
---------------------------------------------------------------------------
Longitudinal epidemiologic analyses also report positive
associations with markers of systemic inflammation (U.S. EPA, 2019a,
section 6.2.11), coagulation (U.S. EPA, 2019a, section 6.2.12), and
[[Page 16230]]
endothelial dysfunction (U.S. EPA, 2019a, section 6.2.13). These
results are coherent with animal toxicological studies generally
reporting increased markers of systemic inflammation, oxidative stress,
and endothelial dysfunction (U.S. EPA, 2019a, section 6.2.12.2 and
6.2.14).
In summary, the 2019 ISA concludes that there is consistent
evidence from multiple epidemiologic studies illustrating that long-
term exposure to PM2.5 is associated with mortality from
cardiovascular causes. Epidemiologic studies evaluated in the ISA
Supplement provide additional evidence of positive associations between
long-term PM2.5 exposure and cardiovascular morbidity (U.S.
EPA, 2022a, section 3.1.2.2). Associations with coronary heart disease
(CHD), stroke and atherosclerosis progression were observed in several
additional epidemiologic studies, providing coherence with the
mortality findings. Results from copollutant models generally support
an independent effect of PM2.5 exposure on mortality.
Additional evidence of the independent effect of PM2.5 on
the cardiovascular system is provided by experimental studies in
animals, which support the biological plausibility of pathways by which
long-term exposure to PM2.5 could potentially result in
outcomes such as CHD, stroke, CHF, and cardiovascular mortality.
Overall, studies evaluated in the 2019 ISA support the conclusion of a
causal relationship between long-term PM2.5 exposure and
cardiovascular effects, which is supported and extended by evidence
from recent epidemiologic studies evaluated in the ISA Supplement (U.S.
EPA, 2022a, section 3.1.2.2).
Short-Term PM2.5 Exposures
The 2009 ISA concluded that ``a causal relationship exists between
short-term exposure to PM2.5 and cardiovascular effects''
(U.S. EPA, 2009a). The strongest evidence in the 2009 ISA was from
epidemiologic studies of emergency department (ED) visits and hospital
admissions for IHD and heart failure (HF), with supporting evidence
from epidemiologic studies of cardiovascular mortality (U.S. EPA,
2009a). Animal toxicological studies provided coherence and biological
plausibility for the positive associations reported with MI, ED visits,
and hospital admissions. These included studies reporting reduced
myocardial blood flow during ischemia and studies indicating altered
vascular reactivity. In addition, effects of PM2.5 exposure
on a potential indicator of ischemia (i.e., ST segment depression on an
electrocardiogram) were reported in both animal toxicological and
epidemiologic panel studies.\58\ Key uncertainties from the last review
resulted from inconsistent results across disciplines with respect to
the relationship between short-term exposure to PM2.5 and
changes in blood pressure, blood coagulation markers, and markers of
systemic inflammation. In addition, while the 2009 ISA identified a
growing body of evidence from controlled human exposure and animal
toxicological studies, uncertainties remained with respect to
biological plausibility.
---------------------------------------------------------------------------
\58\ Some animal studies included in the 2009 ISA examined
exposures to mixtures, such as motor vehicle exhaust or woodsmoke.
In these studies, it was unclear if the resulting cardiovascular
effects could be attributed specifically to the fine particle
component of the mixture.
---------------------------------------------------------------------------
Studies evaluated in the 2019 ISA provide additional support for a
causal relationship between short-term PM2.5 exposure and
cardiovascular effects. This includes generally positive associations
observed in multicity epidemiologic studies of emergency department
visits and hospital admissions for IHD, heart failure (HF), and
combined cardiovascular-related endpoints. In particular, nationwide
studies of older adults (65 years and older) using Medicare records
report positive associations between PM2.5 exposures and
hospital admissions for HF (U.S. EPA, 2019a, section 6.1.3.1).
Moreover, recent multicity studies, published after the literature
cutoff date of the 2019 ISA and evaluated in the ISA Supplement, are
consistent with studies evaluated in the 2019 ISA that report positive
association between short-term PM2.5 exposure and ED visits
and hospital admission for IHD, heart attacks, and HF (U.S. EPA, 2022a,
section 3.1). Epidemiologic studies conducted in single cities
contribute some support to the causality determination, though
associations reported in single-city studies are less consistently
positive than in multicity studies, and include a number of studies
reporting null associations (U.S. EPA, 2019a, sections 6.1.2 and
6.1.3). As a whole, though, the recent body of IHD and HF epidemiologic
evidence supports the evidence from previous ISAs reporting mainly
positive associations between short-term PM2.5
concentrations and emergency department visits and hospital admissions.
Consistent with the evidence assessed in the 2019 ISA, some studies
evaluated in the ISA Supplement report no evidence of an association
with stroke, regardless of stroke subtype. Additionally, as in the 2019
ISA, evidence evaluated in the ISA Supplement continues to indicate an
immediate effect of PM2.5 on cardiovascular-related outcomes
primarily within the first few days after exposure, and that
associations generally persisted in models adjusted for copollutants
(U.S. EPA, 2022a, section 3.1.1.2).
The ISA Supplement includes additional epidemiologic studies,
published since the literature cutoff date for the 2019 ISA, including
accountability analyses and epidemiologic studies that employ
alternative methods for confounder control to evaluate the association
between short-term PM2.5 exposure and cardiovascular-related
effects (U.S. EPA, 2022a, section 3.1.1.3). These studies employ a
number of statistical approaches and report positive associations,
providing additional support for a relationship between short-term
PM2.5 exposure and cardiovascular effects, while also
reducing uncertainties related to potential confounder bias.
A number of controlled human exposure, animal toxicological, and
epidemiologic panel studies provide evidence that PM2.5
exposure could plausibly result in IHD or HF through pathways that
include endothelial dysfunction, arterial thrombosis, and arrhythmia
(U.S. EPA, 2019a, section 6.1.1). The most consistent evidence from
recent controlled human exposure studies is for endothelial
dysfunction, as measured by changes in brachial artery diameter or flow
mediated dilation. Multiple controlled human exposure studies that
examined the potential for endothelial dysfunction report an effect of
PM2.5 exposure on measures of blood flow (U.S. EPA, 2019a,
section 6.1.13.2). However, these studies report variable results
regarding the timing of the effect and the mechanism by which reduced
blood flow occurs (i.e., availability vs sensitivity to nitric oxide).
In addition, some controlled human exposure studies using CAPs report
evidence for small increases in blood pressure (U.S. EPA, 2019a,
section 6.1.6.3). Although not entirely consistent, there is also some
evidence across controlled human exposure studies for conduction
abnormalities/arrhythmia (U.S. EPA, 2019a, section 6.1.4.3), changes in
heart rate variability (HRV) (U.S. EPA, 2019a, section 6.1.10.2),
changes in hemostasis that could promote clot formation (U.S. EPA,
2019a, section 6.1.12.2), and increases in inflammatory cells and
markers (U.S. EPA, 2019a, section 6.1.11.2). A recent study by Wyatt et
al.
[[Page 16231]]
(2020), evaluated in the ISA Supplement, adds to the limited evidence
base of controlled human exposure studies conducted at near ambient
PM2.5 concentrations. The study, completed in healthy young
adults subject to intermittent exercise, found some significant
cardiovascular effects (e.g., systematic inflammation markers,
including C-reactive protein (CRP), and cardiac repolarization). Thus,
when taken as a whole, controlled human exposure studies are coherent
with epidemiologic studies in that they demonstrate that short-term
exposures to PM2.5 may result in the types of cardiovascular
endpoints that could lead to emergency department visits and hospital
admissions for IHD or HF, as well as mortality in some people.
Animal toxicological studies published since the 2009 ISA and
evaluated in the 2019 ISA also support a relationship between short-
term PM2.5 exposure and cardiovascular effects. A study
demonstrating decreased cardiac contractility and left ventricular
pressure in mice is coherent with the results of epidemiologic studies
that report associations between short-term PM2.5 exposure
and heart failure (U.S. EPA, 2019a, section 6.1.3.3). In addition, and
as with controlled human exposure studies, there is generally
consistent evidence in animal toxicological studies for indicators of
endothelial dysfunction (U.S. EPA, 2019a, section 6.1.13.3). Some
studies in animals also provide evidence for changes in a number of
other cardiovascular endpoints following short-term PM2.5
exposure including conduction abnormalities and arrhythmia (U.S. EPA,
2019a, section 6.1.4.4), changes in HRV (U.S. EPA, 2019a, section
6.1.10.3), changes in blood pressure (U.S. EPA, 2019a, section
6.1.6.4), and evidence for systemic inflammation and oxidative stress
(U.S. EPA, 2019a, section 6.1.11.3).
In summary, evidence evaluated in the 2019 ISA extends the
consistency and coherence of the evidence base evaluated in the 2009
ISA and prior assessments. Epidemiologic studies reporting robust
associations in copollutant models are supported by direct evidence
from controlled human exposure and animal toxicologic studies reporting
independent effects of PM2.5 exposures on endothelial
dysfunction as well as endpoints indicating impaired cardiac function,
increased risk of arrhythmia, changes in HRV, increases in BP, and
increases in indicators of systemic inflammation, oxidative stress, and
coagulation (U.S. EPA, 2019, section 6.1.16). For some cardiovascular
effects, there are inconsistencies in results across some animal
toxicological, controlled human exposure, and epidemiologic panel
studies, though this may be due to substantial differences in study
design and/or study populations. Overall, the results from
epidemiologic panel, controlled human exposure, and animal
toxicological studies, in particular those related to endothelial
dysfunction, impaired cardiac function, ST segment depression,
thrombosis, conduction abnormalities, and changes in blood pressure
provide coherence and biological plausibility for the consistent
results from epidemiologic studies observing positive associations
between short-term PM2.5 exposures and IHD and HF, and
ultimately cardiovascular mortality. Overall, studies evaluated in the
2019 ISA support the conclusion of a causal relationship between short-
term PM2.5 exposure and cardiovascular effects, which is
supported and extended by evidence from recent epidemiologic studies
evaluated in the ISA Supplement (U.S. EPA, 2022a, section 3.1.1.4).
iii. Respiratory Effects
Long-Term PM2.5 Exposures
The 2009 ISA concluded that ``a causal relationship is likely to
exist between long-term PM2.5 exposure and respiratory
effects'' (U.S. EPA, 2009a). This conclusion was based mainly on
epidemiologic evidence demonstrating associations between long-term
PM2.5 exposure and changes in lung function or lung function
growth in children. Biological plausibility was provided by a single
animal toxicological study examining pre- and post-natal exposure to
PM2.5 CAPs, which found impaired lung development.
Epidemiologic evidence for associations between long-term
PM2.5 exposure and other respiratory outcomes, such as the
development of asthma, allergic disease, and COPD; respiratory
infection; and the severity of disease was limited, both in the number
of studies available and the consistency of the results. Experimental
evidence for other outcomes was also limited, with one animal
toxicological study reporting that long-term exposure to
PM2.5 CAPs results in morphological changes in nasal airways
of healthy animals. Other animal studies examined exposure to mixtures,
such as motor vehicle exhaust and woodsmoke, and effects were not
attributed specifically to the particulate components of the mixture.
Cohort studies evaluated in the 2019 ISA provided additional
support for the relationship between long-term PM2.5
exposure and decrements in lung function growth (as a measure of lung
development), indicating a robust and consistent association across
study locations, exposure assessment methods, and time periods (U.S.
EPA, 2019a, section 5.2.13). This relationship was further supported by
a retrospective study that reports an association between declining
PM2.5 concentrations and improvements in lung function
growth in children (U.S. EPA, 2019a, section 5.2.11). Epidemiologic
studies also examine asthma development in children (U.S. EPA, 2019a,
section 5.2.3), with prospective cohort studies reporting generally
positive associations, though several are imprecise (i.e., they report
wide confidence intervals). Supporting evidence is provided by studies
reporting associations with asthma prevalence in children, with
childhood wheeze, and with exhaled nitric oxide, a marker of pulmonary
inflammation (U.S. EPA, 2019a, section 5.2.13). Additionally, the 2019
ISA includes an animal toxicological study showing the development of
an allergic phenotype and an increase in a marker of airway
responsiveness supports the biological plausibility of the development
of allergic asthma (U.S. EPA, 2019a, section 5.2.13). Other
epidemiologic studies report a PM2.5-related acceleration of
lung function decline in adults, while improvement in lung function was
observed with declining PM2.5 concentrations (U.S. EPA,
2019a, section 5.2.11). A longitudinal study found declining
PM2.5 concentrations are also associated with an improvement
in chronic bronchitis symptoms in children, strengthening evidence
reported in the 2009 ISA for a relationship between increased chronic
bronchitis symptoms and long-term PM2.5 exposure (U.S. EPA,
2019a, section 5.2.11). A common uncertainty across the epidemiologic
evidence is the lack of examination of copollutants to assess the
potential for confounding. While there is some evidence that
associations remain robust in models with gaseous pollutants, a number
of these studies examining copollutant confounding were conducted in
Asia, and thus have limited generalizability due to high annual
pollutant concentrations.
When taken together, the 2019 ISA concludes that the epidemiologic
evidence strongly supports a relationship with decrements in lung
function growth asthma development in children, as well as increased
bronchitis symptoms in children with asthma. Additionally, the
epidemiologic
[[Page 16232]]
evidence strongly supports a relationship with an acceleration of lung
function decline in adults, and with respiratory mortality and cause-
specific respiratory mortality for COPD and respiratory infection (U.S.
EPA, 2019a, p. 1-34). In support of the biological plausibility of
associations reported in epidemiologic studies associated with
respiratory health effects, animal toxicological studies evaluated in
the 2019 ISA continue to provide direct evidence that long-term
exposure to PM2.5 results in a variety of respiratory
effects, including pulmonary oxidative stress, inflammation, and
morphologic changes in the upper (nasal) and lower airways. Other
results show that changes are consistent with the development of
allergy and asthma, and with impaired lung development. Overall, the
2019 ISA concludes that ``the collective evidence is sufficient to
conclude that a causal relationship is likely to exist between long-
term PM2.5 exposure and respiratory effects'' (U.S. EPA,
2019a, section 5.2.13).
Short-Term PM2.5 Exposures
The 2009 ISA (U.S. EPA, 2009a) concluded that a ``causal
relationship is likely to exist'' between short-term PM2.5
exposure and respiratory effects. This conclusion was based mainly on
the epidemiologic evidence demonstrating positive associations with
various respiratory effects. Specifically, the 2009 ISA described
epidemiologic evidence as consistently showing PM2.5-
associated increases in hospital admissions and ED visits for COPD and
respiratory infection among adults or people of all ages, as well as
increases in respiratory mortality. These results were supported by
studies reporting associations with increased respiratory symptoms and
decreases in lung function in children with asthma, though the
epidemiologic evidence was inconsistent for hospital admissions or
emergency department visits for asthma. Studies examining copollutant
models showed that PM2.5 associations with respiratory
effects were robust to inclusion of CO or SO2 in the model,
but often were attenuated (though still positive) with inclusion of
O3 or NO2. In addition to the copollutant models,
evidence supporting an independent effect of PM2.5 exposure
on the respiratory system was provided by animal toxicological studies
of PM2.5 CAPs demonstrating changes in some pulmonary
function parameters, as well as inflammation, oxidative stress, injury,
enhanced allergic responses, and reduced host defenses. Many of these
effects have been implicated in the pathophysiology for asthma
exacerbation, COPD exacerbation, or respiratory infection. In the few
controlled human exposure studies conducted in individuals with asthma
or COPD, PM2.5 exposure mostly had no effect on respiratory
symptoms, lung function, or pulmonary inflammation. Available studies
in healthy people also did not clearly demonstrate respiratory effects
following short-term PM2.5 exposures.
Epidemiologic studies evaluated in the 2019 ISA continue to provide
strong evidence for a relationship between short-term PM2.5
exposure and several respiratory-related endpoints, including asthma
exacerbation (U.S. EPA, 2019a, section 5.1.2.1), COPD exacerbation
(U.S. EPA, 2019a, section 5.1.4.1), and combined respiratory-related
diseases (U.S. EPA, 2019a, section 5.1.6), particularly from studies
examining ED visits and hospital admissions. The generally positive
associations between short-term PM2.5 exposure and asthma
and COPD as well as ED visits and hospital admissions are supported by
epidemiologic studies demonstrating associations with other
respiratory-related effects such as symptoms and medication use that
are indicative of asthma and COPD exacerbations (U.S. EPA, 2019a,
sections 5.1.2.2 and 5.4.1.2). The collective body of epidemiologic
evidence for asthma exacerbation is more consistent in children than in
adults. Additionally, epidemiologic studies examining the relationship
between short-term PM2.5 exposure and respiratory mortality
provide evidence of consistent positive associations, demonstrating a
continuum of effects (U.S. EPA, 2019a, section 5.1.9).
Epidemiologic studies evaluated in the 2019 ISA expand the
assessment of potential copollutant confounding evaluated in the 2009
ISA. There is some evidence that PM2.5 associations with
asthma exacerbation, combined respiratory-related diseases, and
respiratory mortality remain relatively unchanged in copollutant models
with gaseous pollutants including O3, NO2,
SO2, and with more limited evidence for CO, as well as other
particle sizes (i.e., PM10-2.5) (U.S. EPA, 2019a, section
5.1.10.1).
Insight into whether there is an independent effect of
PM2.5 on respiratory health is also partially addressed by
findings from animal toxicological studies evaluated in the 2019 ISA.
Specifically, short-term exposure to PM2.5 enhanced asthma-
related responses in an animal model of allergic airways disease and
enhanced lung injury and inflammation in an animal model of COPD (U.S.
EPA, 2019a, sections 5.1.2.4.4 and 5.1.4.4.3). The experimental
evidence provides biological plausibility for some respiratory-related
endpoints, including limited evidence of altered host defense and
greater susceptibility to bacterial infection as well as consistent
evidence of respiratory irritant effects. However, animal toxicological
evidence for other respiratory effects is inconsistent. A recent study
evaluated in the ISA supplement by Wyatt et al. (2020) and conducted at
near ambient PM2.5 concentrations, adds to the limited
evidence base of controlled human exposure studies. The study,
completed in healthy young adults subject to intermittent exercise,
found some significant respiratory effects (including decrease in lung
function), however these findings were inconsistent with the controlled
human exposure studies evaluated in the 2019 ISA (U.S. EPA, 2019a,
section 5.1.7.2, 5.1.2.3, and 6.1.11.2.1).
The 2019 ISA concludes that ``[t]he strongest evidence of an effect
of short-term PM2.5 exposure on respiratory effects is
provided by epidemiologic studies of asthma and COPD exacerbation.
While animal toxicological studies provide biological plausibility for
these findings, some uncertainty remains with respect to the
independence of PM2.5 effects'' (U.S. EPA, 2019a, p. 5-155).
When taken together, the 2019 ISA concludes that this evidence ``is
sufficient to conclude that a causal relationship is likely to exist
between short-term PM2.5 exposure and respiratory effects''
(U.S. EPA, 2019a, p. 5-155).
iv. Cancer
The 2009 ISA concluded that the overall body of evidence was
``suggestive of a causal relationship between relevant PM2.5
exposures and cancer'' (U.S. EPA, 2009a). This conclusion was based
primarily on positive associations observed in a limited number of
epidemiologic studies of lung cancer mortality. The few epidemiologic
studies that had evaluated PM2.5 exposure and lung cancer
incidence or cancers of other organs and systems generally did not show
evidence of an association. Toxicological studies did not focus on
exposures to specific PM size fractions, but rather investigated the
effects of exposures to total ambient PM, or other source-based PM such
as wood smoke. Collectively, results of in vitro studies were
consistent with the larger body of evidence demonstrating that ambient
PM and PM from specific combustion sources are mutagenic and genotoxic.
However, animal inhalation studies
[[Page 16233]]
found little evidence of tumor formation in response to chronic
exposures. A small number of studies provided preliminary evidence that
PM exposure can lead to changes in methylation of DNA, which may
contribute to biological events related to cancer.
Since the completion of the 2009 ISA, additional cohort studies
provide evidence that long-term PM2.5 exposure is positively
associated with lung cancer mortality and with lung cancer incidence,
and provide initial evidence for an association with reduced cancer
survival (U.S. EPA, 2019a, section 10.2.5). Re-analyses of the ACS
cohort using different years of PM2.5 data and follow up,
along with various exposure assignment approaches, provide consistent
evidence of positive associations between long-term PM2.5
exposure and lung cancer mortality (U.S. EPA, 2019a, Figure 10-3).
Additional support for positive associations with lung cancer mortality
is provided by recent epidemiologic studies using individual level data
to control for smoking status, as well as by studies of people who have
never smoked (though such studies generally report wide confidence
intervals due to the small number of lung cancer mortality cases within
this population), and in additional analyses of cohorts that relied
upon proxy measures to account for smoking status (U.S. EPA, 2019a,
section 10.2.5.1.1). Although studies that evaluate lung cancer
incidence, including studies of people who have never smoked, are
limited in number, studies in the 2019 ISA generally report positive
associations with long-term PM2.5 exposures (U.S. EPA,
2019a, section 10.2.5.1.2). A subset of the studies focusing on lung
cancer incidence also examined histological subtype, providing some
evidence of positive associations for adenocarcinomas, the predominate
subtype of lung cancer observed in people who have never smoked (U.S.
EPA, 2019a, section 10.2.5.1.2). Associations between long-term
PM2.5 exposure and lung cancer incidence were found to
remain relatively unchanged, though in some cases confidence intervals
widened, in analyses that attempted to reduce exposure measurement
error by accounting for length of time at residential address or by
examining different exposure assignment approaches (U.S. EPA, 2019a,
section 10.2.5.1.2).
To date, relatively few studies have evaluated the potential for
copollutant confounding of the relationship between long-term
PM2.5 exposure and lung cancer mortality or incidence. A
small number of such studies have generally focused on O3
and report that PM2.5 associations remain relatively
unchanged in copollutant models (U.S. EPA, 2019a, section 10.2.5.1.3).
However, available studies have not systematically evaluated the
potential for copollutant confounding by other gaseous pollutants or by
other particle size fractions (U.S. EPA, 2019a, section 10.2.5.1.3).
Compared to total (non-accidental) mortality (U.S. EPA, 2019a,
section 10.2.4.1.4), fewer studies have examined the shape of the C-R
curve for cause-specific mortality outcomes, including lung cancer.
Several studies of lung cancer mortality and incidence have reported no
evidence of deviations from linearity in the shape of the C-R
relationship (Lepeule et al., 2012; Raaschou-Nielsen et al., 2013;
Puett et al., 2014), though authors provided only limited discussions
of results (U.S. EPA, 2019a, section 10.2.5.1.4).
In support of the biological plausibility of an independent effect
of PM2.5 on lung cancer, the 2019 ISA notes evidence from
experimental and epidemiologic studies demonstrating that
PM2.5 exposure can lead to a range of effects indicative of
mutagenicity, genotoxicity, and carcinogenicity, as well as epigenetic
effects (U.S. EPA, 2019a, section 10.2.7). For example, both in vitro
and in vivo toxicological studies have shown that PM2.5
exposure can result in DNA damage (U.S. EPA, 2019a, section 10.2.2).
Although such effects do not necessarily equate to carcinogenicity, the
evidence that PM exposure can damage DNA, and elicit mutations,
provides support for the plausibility of epidemiologic associations
exhibited with lung cancer mortality and incidence. Additional
supporting studies indicate the occurrence of micronuclei formation and
chromosomal abnormalities (U.S. EPA, 2019a, section 10.2.2.3), and
differential expression of genes that may be relevant to cancer
pathogenesis, following PM2.5 exposures. Experimental and
epidemiologic studies that examine epigenetic effects indicate changes
in DNA methylation, providing some support that PM2.5
exposure contributes to genomic instability (U.S. EPA, 2019a, section
10.2.3). Overall, there is limited evidence that long-term
PM2.5 exposure is associated with cancers in other organ
systems, though there is some evidence that PM2.5 exposure
may reduce survival in individuals with cancer (U.S. EPA, 2019a,
section 10.2.7; U.S. EPA, 2022a, section 2.1.1.4.1).
Epidemiologic evidence for associations between PM2.5
and lung cancer mortality and incidence, together with evidence
supporting the biological plausibility of such associations,
contributes to the 2019 ISA's conclusion that the evidence ``is
sufficient to conclude that a causal relationship is likely to exist
between long-term PM2.5 exposure and cancer'' (U.S. EPA,
2019, section 10.2.7).
v. Nervous System Effects
Reflecting the very limited evidence available in the 2012 review,
the 2009 ISA did not make a causality determination for long-term
PM2.5 exposures and nervous system effects (U.S. EPA,
2009c). Since the 2012 review, this body of evidence has grown
substantially (U.S. EPA, 2019, section 8.2). Animal toxicological
studies assessed in in the 2019 ISA report that long-term
PM2.5 exposures can lead to morphologic changes in the
hippocampus and to impaired learning and memory. This evidence is
consistent with epidemiologic studies reporting that long-term
PM2.5 exposure is associated with reduced cognitive function
(U.S. EPA, 2019a, section 8.2.5). Further, while the evidence is
limited, the presence of early markers of Alzheimer's disease pathology
has been demonstrated in rodents following long-term exposure to
PM2.5 CAPs. These findings support reported associations
with neurodegenerative changes in the brain (i.e., decreased brain
volume), all-cause dementia, or hospitalization for Alzheimer's disease
in a small number of epidemiologic studies (U.S. EPA, 2019a, section
8.2.6). Additionally, loss of dopaminergic neurons in the substantia
nigra, a hallmark of Parkinson disease, has been reported in mice (U.S.
EPA, 2019a, section 8.2.4), though epidemiologic studies provide only
limited support for associations with Parkinson's disease (U.S. EPA,
2019a, section 8.2.6). Overall, the lack of consideration of
copollutant confounding introduces some uncertainty in the
interpretation of epidemiologic studies of nervous system effects, but
this uncertainty is partly addressed by the evidence for an independent
effect of PM2.5 exposures provided by experimental animal
studies.
While the findings described above are most relevant to older
adults, several studies of neurodevelopmental effects in children have
also been conducted. Epidemiologic studies provided limited evidence of
an association between PM2.5 exposure during pregnancy and
childhood on cognitive and motor development (U.S. EPA, 2019, section
8.2.5.2). While some studies report positive associations between long-
term
[[Page 16234]]
exposure to PM2.5 during the prenatal period and autism
spectrum disorder (ASD) (U.S. EPA, 2019, section 8.2.7.2), the
interpretation of these epidemiologic studies is limited due to the
small number of studies, their lack of control for potential
confounding by copollutants, and uncertainty related to the critical
exposure windows. Biological plausibility is provided for the ASD
findings by a study in mice that found inflammatory and morphologic
changes in the corpus collosum and hippocampus, as well as
ventriculomegaly (i.e., enlarged lateral ventricles) in young mice
following prenatal exposure to PM2.5 CAPs.
Taken together, the 2019 ISA concludes that studies indicate long-
term PM2.5 exposures can lead to effects on the brain
associated with neurodegeneration (i.e., neuroinflammation and
reductions in brain volume), as well as cognitive effects in older
adults (U.S. EPA, 2019a, Table 1-2). Animal toxicological studies
provide evidence for a range of nervous system effects in adult
animals, including neuroinflammation and oxidative stress,
neurodegeneration, cognitive effects, and effects on neurodevelopment
in young animals. The epidemiologic evidence is more limited, but
studies generally support associations between long-term
PM2.5 exposure and changes in brain morphology, cognitive
decrements and dementia. There is also initial, and limited, evidence
for neurodevelopmental effects, particularly ASD. The consistency and
coherence of the evidence supports the 2019 ISA's conclusion that ``the
collective evidence is sufficient to conclude that a causal
relationship is likely to exist between long-term PM2.5
exposure and nervous system effects'' (U.S. EPA, 2019a, section 8.2.9).
vi. Other Effects
For other health effect categories that were evaluated for their
relationship with PM2.5 exposures (i.e., short-term
PM2.5 exposure and nervous system effects and short- and
long-term PM2.5 exposure and metabolic effects, reproduction
and fertility, and pregnancy and birth outcomes (U.S. EPA, 2022a, Table
ES-1), the currently available evidence is ``suggestive of, but not
sufficient to infer, a causal relationship,'' mainly due to
inconsistent evidence across specific outcomes and uncertainties
regarding exposure measurement error, the potential for confounding,
and potential modes of action (U.S. EPA, 2019a, sections 7.14, 7.2.10,
8.1.6, and 9.1.5). The causality determination for short-term
PM2.5 exposure and nervous system effects in the 2019 ISA
reflects a revision to the causality determination in the 2009 ISA from
``inadequate to infer a causal relationship,'' while this is the first-
time assessments of causality were conducted for long-term
PM2.5 exposure and nervous system effects, as well as short-
and long-term PM2.5 exposure and metabolic effects reflect.
Recent studies evaluated in the 2019 ISA also further explored the
relationship between short-and long-term UFP exposure and health
effects. (i.e., cardiovascular effects and short-term UFP exposures;
respiratory effects and short-term UFP exposures; and nervous system
effects and long- and short-term exposures (U.S. EPA, 2022a, Table ES-
1). The currently available evidence is ``suggestive of, but not
sufficient to infer, a causal relationship'' for short-term UFP
exposure and cardiovascular and respiratory effects and for short- and
long-term UFP exposure and nervous system effects, primarily due to
uncertainties and limitations in the evidence, specifically,
variability across studies in the definition of UFPs and the exposure
metric used (U.S. EPA, 2019a, P.3.1; U.S. EPA, 2022a, section
3.3.1.6.3). The causality determinations for the other health effect
categories evaluated in the 2019 ISA are ``inadequate to infer a causal
relationship.'' Additionally, this is the first time assessments of
causality were conducted for short- and long-term UFP exposure and
metabolic effects and long-term UFP exposure and nervous system effects
(U.S. EPA, 2022a, Table ES-1).
With the advent of the global COVID-19 pandemic, a number of recent
studies evaluated in the ISA Supplement examined the relationship
between ambient air pollution, specifically PM2.5, and SARS-
CoV-2 infections and COVID-19 deaths, including a few studies within
the U.S. and Canada (U.S. EPA, 2022a, section 3.3.2).\59\ Some studies
examined whether daily changes in PM2.5 can influence SARS-
CoV-2 infection and COVID-19 death (U.S. EPA, 2022a, section 3.3.2.1).
Additionally, several studies evaluated whether long-term
PM2.5 exposure increases the risk of SARS-CoV-2 infection
and COVID-19 death in North America (U.S. EPA, 2022a, section 3.3.2.2).
While there is initial evidence of positive associations with SARS-CoV-
2 infection and COVID-19 death, uncertainties remain due to
methodological issues that may influence the results, including: (1)
The use of ecological study design; (2) studies were conducted during
the ongoing pandemic when the etiology of COVID-19 was still not well
understood (e.g., specifically, there are important differences in
COVID-19-related outcomes by a variety of factors such as race and
SES); and (3) studies did not account for crucial factors that could
influence results (e.g., stay-at-home orders, social distancing, use of
masks, and testing capacity) (U.S. EPA, 2022a, chapter 5). Taken
together, while there is initial evidence of positive associations with
SARS-CoV-2 infection and COVID-19 death, uncertainties remain due to
methodological issues.
---------------------------------------------------------------------------
\59\ While there is no exact corollary within the 2019 ISA for
these types of studies, the 2019 ISA presented evidence that
evaluates the potential relationship between short- and long-term
PM2.5 exposure and respiratory infection (U.S. EPA,
2022a, section 5.1.5 and 5.2.6). Studies assessed in the 2019 ISA
report some evidence of positive associations between short-term
PM2.5 and hospital admissions and ED visits for
respiratory infections, however the interpretation of these studies
is complicated by the variability in the type of respiratory
infection outcome examined (U.S. EPA, 2022a, Figure 5-7). In the
2019 ISA, studies of long-term PM2.5 exposure were
limited and while there were some positive associations reported,
there was minimal overlap in respiratory infection outcomes examined
across studies. Exposure to PM2.5 has been shown to
impair host defense, specifically altering macrophage function,
providing a biological pathway by which PM2.5 exposure
could lead to respiratory infection (U.S. EPA, 2022a, sections 5.1.1
and 5.1.5.) There is some additional evidence that PM2.5
exposure can lead to decreases in an individual's immune response,
which can subsequently facilitate replication of respiratory viruses
(Bourdrel et al., 2021).
---------------------------------------------------------------------------
b. Public Health Implications and At-Risk Populations
The public health implications of the evidence regarding
PM2.5-related health effects, as for other effects, are
dependent on the type and severity of the effects, as well as the size
of the population affected. Such factors are discussed below in the
context of our consideration of the health effects evidence related to
PM2.5 in ambient air. This section also summarizes the
current information on population groups at increased risk of the
effects of PM2.5 in ambient air.
The information available in this reconsideration has not altered
our understanding of human populations at risk of health effects from
PM2.5 exposures. As recognized in the 2020 review, the 2019
ISA cites extensive evidence indicating that ``both the general
population as well as specific populations and lifestages are at risk
for PM2.5-related health effects'' (U.S. EPA, 2019a, p. 12-
1). Factors that may contribute to increased risk of PM2.5-
related health effects include lifestage (children and older adults),
pre-existing diseases (cardiovascular disease and
[[Page 16235]]
respiratory disease), race/ethnicity, and SES.\60\
---------------------------------------------------------------------------
\60\ As described in the 2019 ISA, other factors that have the
potential to contribute to increased risk include obesity, diabetes,
genetic factors, smoking status, sex, diet, and residential location
(U.S. EPA, 2019, chapter 12).
---------------------------------------------------------------------------
Children make up a substantial fraction of the U.S. population, and
often have unique factors that contribute to their increased risk of
experiencing a health effect due to exposures to ambient air pollutants
because of their continuous growth and development.\61\ Children may be
particularly at risk for health effects related to ambient
PM2.5 exposures compared with adults because they have (1) a
developing respiratory system, (2) increased ventilation rates relative
to body mass compared with adults, and (3) an increased proportion of
oral breathing, particularly in boys, relative to adults (U.S. EPA,
2019a, section 12.5.1.1). There is strong evidence that demonstrates
PM2.5 associated health effects in children, particularly
from epidemiologic studies of long-term PM2.5 exposure and
impaired lung function growth, decrements in lung function, and asthma
development. However, there is limited evidence from stratified
analyses that children are at increased risk of PM2.5-
related health effects compared to adults. Additionally, there is some
evidence that indicates that children receive higher PM2.5
exposures than adults, and dosimetric differences in children compared
to adults can contribute to higher doses (U.S. EPA, 2019a, section
12.5.1.1).
---------------------------------------------------------------------------
\61\ Children, as used throughout this document, generally
refers to those younger than 18 years old.
---------------------------------------------------------------------------
In the U.S., older adults, often defined as adults 65 years of age
and older, represent an increasing portion of the population and often
have pre-existing diseases or conditions that may compromise biological
function. While there is limited evidence to indicate that older adults
have higher exposures than younger adults, older adults may receive
higher doses of PM2.5 due to dosimetric differences. There
is consistent evidence from studies of older adults demonstrating
generally consistent positive associations in studies examining health
effects from short- and long-term PM2.5 exposure and
cardiovascular or respiratory hospital admissions, emergency department
visits, or mortality (U.S. EPA, 2019a, sections 6.1, 6.2, 11.1, 11.2,
12.5.1.2). Additionally, several animal toxicological, controlled human
exposure, and epidemiologic studies did not stratify results by
lifestage, but instead focused the analyses on older individuals, and
can provide coherence and biological plausibility for the occurrence
among this lifestage (U.S. EPA, 2019a, section 12.5.1.2).
Individuals with pre-existing disease may be considered at greater
risk of an air pollution-related health effect than those without
disease because they are likely in a compromised biological state that
can vary depending on the disease and severity. With regard to
cardiovascular disease, we first note that cardiovascular disease is
the leading cause of death in the U.S., accounting for one in four
deaths, and approximately 12% of the adult population in the U.S. has a
cardiovascular disease (U.S. EPA, 2019a, section 12.3.1). Strong
evidence demonstrates that there is a causal relationship between
cardiovascular effects and long- and short-term exposures to
PM2.5. Some of the evidence supporting this conclusion is
from studies of panels or cohorts with pre-existing cardiovascular
disease, which provide supporting evidence but do not directly
demonstrate an increased risk (U.S. EPA, 2019a, section 12.3.1).
Epidemiologic evidence indicates that individuals with pre-existing
cardiovascular disease may be at increased risk for PM2.5-
associated health effects compared to those without pre-existing
cardiovascular disease. While the evidence does not consistently
support increased risk for all pre-existing cardiovascular diseases,
there is evidence that certain pre-existing cardiovascular diseases
(e.g., hypertension) may be a factor that increases PM2.5-
related risk. Furthermore, there is strong evidence supporting a causal
relationship for long- and short-term PM2.5 exposure and
cardiovascular effects, particularly for IHD (U.S. EPA, 2019a, chapter
6, section 12.3.1).
With regard to respiratory disease, we first note that the most
chronic respiratory diseases in the U.S. are asthma and COPD. Asthma
affects a substantial fraction of the U.S. population and is the
leading chronic disease among children. COPD primarily affects older
adults and contributes to compromised respiratory function and
underlying pulmonary inflammation. The body of evidence indicates that
individuals with pre-existing respiratory diseases, particularly asthma
and COPD, may be at increased risk for PM2.5-related health
effects compared to those without pre-existing respiratory diseases
(U.S. EPA, 2019a, section 12.3.5). There is strong evidence indicating
PM2.5-associated respiratory effects among those with
asthma, which forms the primary evidence base for the likely to be
causal relationship between short-term exposures to PM2.5
and respiratory health effects (U.S. EPA, 2019a, section 12.3.5). For
asthma, epidemiologic evidence demonstrates associations between short-
term PM2.5 exposures and respiratory effects, particularly
evidence for asthma exacerbation, and controlled human exposure and
animal toxicological studies demonstrate support for the biological
plausibility for asthma exacerbation with PM2.5 exposures
(U.S. EPA, 2019a, section 12.3.5.1). For COPD, epidemiologic studies
report positive associations between short-term PM2.5
exposures and hospital admissions and emergency department visits for
COPD, with supporting evidence from panel studies demonstration COPD
exacerbation. Epidemiologic evidence is supported by some experimental
evidence of COPD-related effects, which provides support for the
biological plausibility for COPD in response to PM2.5
exposures (U.S. EPA, 2019a, section 12.3.5.2).
There is strong evidence for racial and ethnic disparities in
PM2.5 exposures and PM2.5-related health risk, as
assessed in the 2019 ISA and with even more evidence available since
the literature cutoff date for the 2019 ISA and evaluated in the ISA
Supplement. There is strong evidence demonstrating that Black and
Hispanic populations, in particular, have higher PM2.5
exposures than non-Hispanic White populations (U.S. EPA, 2019a, Figure
12-2; U.S. EPA, 2022a, Figure 3-38). Black populations or individuals
that live in predominantly Black neighborhoods experience higher
PM2.5 exposures, in comparison to non-Hispanic White
populations. There is also consistent evidence across multiple studies
that demonstrate increased risk of PM2.5-related health
effects, with the strongest evidence for health risk disparities for
mortality (U.S. EPA, 2019a, section 12.5.4). There is also evidence of
health risk disparities for both Hispanic and non-Hispanic Black
populations compared to non-Hispanic White populations for cause-
specific mortality and incident hypertension (U.S. EPA, 2022a, section
3.3.3.2).
Socioeconomic status (SES) is a composite measure that includes
metrics such as income, occupation, or education, and can play a role
in access to healthy environments as well as access to healthcare. SES
may be a factor that contributes to differential risk from
PM2.5-related health effects. Studies assessed in the 2019
ISA and ISA Supplement provide evidence that lower SES communities are
exposed to higher concentrations of PM2.5
[[Page 16236]]
compared to higher SES communities (U.S. EPA, 2019a, section 12.5.3;
U.S. EPA, 2022a, section 3.3.3.1.1). Studies using composite measures
of neighborhood SES consistently demonstrated a disparity in both
PM2.5 exposure and the risk of PM2.5-related
health outcomes. There is some evidence that supports associations
larger in magnitude between mortality and long-term PM2.5
exposures for those with low income or living in lower income areas
compared to those with higher income or living in higher income
neighborhoods (U.S. EPA, 2019a, section 12.5.3; U.S. EPA, 2022a,
section 3.3.3.1.1). Additionally, evidence supports conclusions that
lower SES is associated with cause-specific mortality and certain
health endpoints (i.e., HI and CHF), but less so for all-cause or total
(non-accidental) mortality (U.S. EPA, 2022a, section 3.3.3.1).
The magnitude and characterization of a public health impact is
dependent upon the size and characteristics of the populations
affected, as well as the type or severity of the effects. As summarized
above, lifestage (children and older adults), race/ethnicity and SES
are factors that increase the risk of PM2.5-related health
effects. The American Community Survey (ACS) for 2019 estimates that
approximately 22% and 16% of the U.S. population are children (age<18)
and older adults (age 65+), respectively. For all ages, non-Hispanic
Black and Hispanic populations comprise approximately 12% and 18% of
the overall U.S. population in 2019. Currently available information
that helps to characterize key features of these population is included
in the 2022 PA (U.S. EPA, 2022b, Table 3-2).
As noted above, individuals with pre-existing cardiovascular
disease and pre-existing respiratory disease may also be at increased
risk of PM2.5-related health effects. Currently available
information that helps to characterize key features of populations with
cardiovascular or respiratory diseases or conditions is included in the
2022 PA (U.S. EPA, 2022b, Table 3-3). The National Center for Health
Statistics data for 2018 indicate that, for adult populations, older
adults (e.g., those 65 years and older) have a higher prevalence of
cardiovascular diseases compared to younger adults (e.g., those 64
years and younger). For respiratory diseases, older adults also have a
higher prevalence of emphysema than younger adults, and adults 44 years
or older have a higher prevalence of chronic bronchitis. However, the
prevalence for asthma is generally similar across all adult age groups.
With respect to race, American Indians or Alaskan Native
populations have the highest prevalence of all heart disease and
coronary heart disease, while Black populations have the highest
prevalence of hypertension and stroke. Hypertension has the highest
prevalence across all racial groups compared to other cardiovascular
diseases or conditions, ranging from approximately 22% to 32% of each
racial group. Overall, the prevalence of cardiovascular diseases or
conditions is lowest for Asians compared to Whites, Blacks, and
American Indians or Alaskan Natives. Asthma prevalence is highest among
Black and American Indian or Alaska Native populations, while the
prevalence of chronic bronchitis and emphysema is generally similar
across racial groups. Overall, the prevalence of respiratory diseases
is lowest for Asians compared to Whites, Blacks, and American Indians
or Alaskan Natives. With regard to ethnicity, cardiovascular and
respiratory disease prevalence across all diseases or conditions is
generally similar between Hispanic and non-Hispanic populations,
although non-Hispanics have a slightly higher prevalence compared to
Hispanics.
Taken together, this information indicates that the groups at
increased risk of PM2.5-related health effects represent a
substantial portion of the total U.S. population. In evaluating the
primary PM2.5 standards, an important consideration is the
potential PM2.5-related public health impacts in these
populations.
c. PM2.5 Concentrations in Key Studies Reporting Health
Effects
To inform conclusions on the adequacy of the public health
protection provided by the current primary PM2.5 standards,
the sections below summarize the 2022 PA's evaluation of the
PM2.5 exposures, specifically the concentrations that have
been examined in controlled human exposure studies, animal
toxicological studies, and epidemiologic studies. The 2022 PA places
the greatest emphasis on the health outcomes for which the 2019 ISA
concludes that the evidence supports a ``causal'' or a ``likely to be
causal'' relationship with short- or long-term PM2.5
exposures (U.S. EPA, 2022b, section 3.3.3). As described in greater
detail in section II.A.2 above, this includes short- or long-term
PM2.5 exposures and mortality, cardiovascular effects, and
respiratory effects and long-term PM2.5 exposures and cancer
and nervous system effects. While the causality determinations in the
2019 ISA are informed by studies evaluating a wide range of
PM2.5 concentrations,\62\ the sections below summarize the
considerations in the 2022 PA regarding the degree to which the
evidence assessed in the 2019 ISA and ISA Supplement supports the
occurrence of PM-related health effects at concentrations relevant to
informing conclusions on the primary PM2.5 standards. In so
doing, the 2022 PA focuses on the available studies that are most
directly informative to reaching conclusions regarding the adequacy of
the current primary PM2.5 standards (e.g., epidemiologic
studies with annual mean PM2.5 concentrations near or below
the level of the standard; and controlled human exposure studies at
PM2.5 exposures that elicit consistent effects, as well as
examining PM2.5 exposures at concentrations that are at or
near the level of the standard).
---------------------------------------------------------------------------
\62\ As described in more detail in section 5 of the Preamble to
the ISAs, judgments regarding causality take into consideration a
number of aspects when evaluating the available scientific evidence
(U.S. EPA, 2015, Table I). In reaching conclusions regarding
causality, ``evidence is evaluated for major outcome categories or
groups of related endpoints (e.g., respiratory effects, vegetation
growth), integrating evidence from across disciplines, and
evaluating the coherence of evidence across a spectrum of related
endpoints'' (U.S. EPA, 2015, p. 24). Furthermore, ``[i]n drawing
judgments regarding causality for the criteria air pollutants, the
ISA focuses on evidence of effects in the range of relevant
pollutant exposures or doses and not on determination of causality
at any dose. Emphasis is placed on evidence of effects at doses
(e.g., blood Pb concentration) or exposures (e.g., air
concentrations) that are relevant to, or somewhat above, those
currently experienced by the population. The extent to which studies
of higher concentrations are considered varies by pollutant and
major outcome category, but generally includes those with doses or
exposures in the range of one to two orders of magnitude above
current or ambient conditions to account for intra-species
variability and toxicokinetic or toxicodynamic differences between
experimental animals and humans. Studies that use higher doses or
exposures may also be considered to the extent that they provide
useful information to inform understanding of mode of action, inter-
species differences, or factors that may increase risk of effects
for a population and if biological mechanisms have not been
demonstrated to differ based on exposure concentration. Thus, a
causality determination is based on weight-of-evidence evaluation
for health or welfare effects, focusing on the evidence from
exposures or doses generally ranging from recent ambient
concentrations to one or two orders of magnitude above recent
ambient concentrations'' (U.S. EPA, 2015, p. 24).
---------------------------------------------------------------------------
i. PM2.5 Exposure Concentrations Evaluated in Experimental
Studies
Evidence for a particular PM2.5-related health outcome
is strengthened when results from experimental studies demonstrate
biologically plausible mechanisms through which adverse human health
outcomes could occur (U.S. EPA, 2015, p. 20). Two types of experimental
studies are of particular importance in understanding the effects
[[Page 16237]]
of PM exposures: controlled human exposure and animal toxicological
studies. In such studies, investigators expose human volunteers or
laboratory animals, respectively, to known concentrations of air
pollutants under carefully regulated environmental conditions and
activity levels. Thus, controlled human exposure and animal
toxicological studies can provide information on the health effects of
experimentally administered pollutant exposures under highly controlled
laboratory conditions (U.S. EPA, 2015, p. 11).
Controlled human exposure studies have reported that
PM2.5 exposures lasting from less than one hour up to five
hours can impact cardiovascular function,\63\ and the most consistent
evidence from these studies is for impaired vascular function (U.S.
EPA, 2019a, section 6.1.13.2). In addition, although less consistent,
the 2019 ISA notes that studies examining PM2.5 exposures
also provide evidence for increased blood pressure (U.S. EPA, 2019a,
section 6.1.6.3), conduction abnormalities/arrhythmia (U.S. EPA, 2019a,
section 6.1.4.3), changes in heart rate variability (U.S. EPA, 2019a,
section 6.1.10.2), changes in hemostasis that could promote clot
formation (U.S. EPA, 2019a, section 6.1.12.2), and increases in
inflammatory cells and markers (U.S. EPA, 2019a, section 6.1.11.2). The
2019 ISA concludes that, when taken as a whole, controlled human
exposure studies demonstrate that short-term exposure to
PM2.5 may impact cardiovascular function in ways that could
lead to more serious outcomes (U.S. EPA, 2019a, section 6.1.16). Thus,
such studies can provide insight into the potential for specific
PM2.5 exposures to result in physiological changes that
could increase the risk of more serious effects. Table 3-4 in the 2022
PA summarizes information from the 2019 ISA and 2022 ISA supplement on
available controlled human exposure studies that evaluate effects on
markers of cardiovascular function following exposure to
PM2.5 (U.S. EPA, 2022b). Most of the controlled human
exposure studies in Table 3-4 of the 2022 PA have evaluated average
PM2.5 concentrations at or above about 100 [micro]g/m\3\,
with exposure durations typically up to about two hours. Statistically
significant effects on one or more indicators of cardiovascular
function are often, though not always, reported following 2-hour
exposures to average PM2.5 concentrations at and above about
120 [micro]g/m\3\, with less consistent evidence for effects following
exposures to concentrations lower than 120 [micro]g/m\3\. Impaired
vascular function, the effect identified in the 2019 ISA as the most
consistent across studies (U.S. EPA, 2019a, section 6.1.13.2) is shown
following 2-hour exposures to PM2.5 concentrations at and
above 149 [micro]g/m\3\. Mixed results are reported in the studies that
evaluated longer exposure durations (i.e., longer than 2 hours) and
lower (i.e., near-ambient) PM2.5 concentrations (U.S. EPA,
2022b, section 3.3.3.1). For example, significant effects for some
outcomes were reported following 5-hour exposures to 24 [micro]g/m\3\
in Hemmingsen et al. (2015b), but not for other outcomes following 5-
hour exposures to 24 [micro]g/m\3\ in Hemmingsen et al. (2015a) and not
following 24-hour exposures to 10.5 [micro]g/m\3\ in Br[auml]uner et
al. (2008). Additionally, Wyatt et al. (2020) found significant effects
for some cardiovascular (e.g., systematic inflammation markers, cardiac
repolarization, and decreased pulmonary function) effects following 4-
hour exposures to 37.8 [micro]g/m\3\ in healthy young participants (18-
35 years, n=21) who were subject to intermittent moderate exercise. The
higher ventilation rate and longer exposure duration in this study
compared to most controlled human exposure studies is roughly
equivalent to a 2-hour exposure of 75-100 [micro]g/m\3\ of
PM2.5. Therefore, dosimetric considerations may explain the
observed changes in inflammation in young healthy individuals. Though
this study provides evidence of some effects at lower PM2.5
concentrations, overall, there is inconsistent evidence for
inflammation in other controlled human exposure studies evaluated in
the 2019 ISA (U.S. EPA, 2019a, sections 5.1.7., 5.1.2.3.3, and
6.1.11.2.1; U.S. EPA, 2022a, section 3.3.1).
---------------------------------------------------------------------------
\63\ In contrast, controlled human exposure studies provide
little evidence for respiratory effects following short-term
PM2.5 exposures (U.S. EPA, 2019a, section 5.1, Table 5-
18). Therefore, this section focuses on cardiovascular effects
evaluated in controlled human exposure studies of PM2.5
exposure.
---------------------------------------------------------------------------
While controlled human exposure studies are important in
establishing biological plausibility, it is unclear how the results
from these studies alone and the importance of the effects observed in
these studies, should be interpreted with respect to adversity to
public health. More specifically, impaired vascular function can signal
an intermediate effect along the potential biological pathways for
cardiovascular effects following short-term exposure to
PM2.5 and show a role for exposure to PM2.5
leading to potential worsening of IHD and heart failure followed
potentially by ED visits, hospital admissions, or mortality (U.S. EPA,
2019a, section 6.1 and Figure 6-1). However, just observing the
occurrence of impaired vascular function alone does not clearly suggest
an adverse health outcome. Additionally, associated judgments regarding
adversity or health significance of measurable physiological responses
to air pollutants have been informed by guidance, criteria or
interpretative statements developed within the public health community,
including the American Thoracic Society (ATS) and the European
Respiratory Society (ERS), which cooperatively updated the ATS 2000
statement What Constitutes an Adverse Health Effect of Air Pollution
(ATS, 2000) with new scientific findings, including the evidence
related to air pollution and the cardiovascular system (Thurston et
al., 2017).\64\ With regard to vascular function, the ATS/ERS statement
considers the adversity of both chronic and acute reductions in
endothelial function. While the ATS/ERS statement concluded that
chronic endothelial and vascular dysfunction can be judged to be a
biomarker of an adverse health effect from air pollution, they also
conclude that ``the health relevance of acute reductions in endothelial
function induced by air pollution is less certain'' (Thurston et al.,
2017). This is particularly informative to our consideration of the
controlled human exposure studies which are short-term in nature (i.e.,
generally ranging from 2- to 5-hours), including those studies that are
conducted at near-ambient PM2.5 concentrations.
---------------------------------------------------------------------------
\64\ The ATS/ERS described its 2017 statement as one ``intended
to provide guidance to policymakers, clinicians and public health
professionals, as well as others who interpret the scientific
evidence on the health effects of air pollution for risk management
purposes'' and further notes that ``considerations as to what
constitutes an adverse health effect, in order to provide guidance
to researchers and policymakers when new health effects markers or
health outcome associations might be reported in future.'' The most
recent policy statement by the ATS, which once again broadens its
discussion of effects, responses and biomarkers to reflect the
expansion of scientific research in these areas, reiterates that
concept, conveying that it does not offer ``strict rules or
numerical criteria, but rather proposes considerations to be weighed
in setting boundaries between adverse and nonadverse health
effects,'' providing a general framework for interpreting evidence
that proposes a ``set of considerations that can be applied in
forming judgments'' for this context (Thurston et al., 2017).
---------------------------------------------------------------------------
The 2022 PA also notes that it is important to recognize that
controlled human exposure studies include a small number of individuals
compared to epidemiologic studies. Additionally, these studies tend to
include generally healthy adult individuals, who are at a lower risk of
experiencing health effects.
[[Page 16238]]
These studies, therefore, often do not include children, older adults,
or individuals with pre-existing conditions. As such, these studies are
somewhat limited in their ability to inform at what concentrations
effects may be elicited in at-risk populations.
Nonetheless, to provide some insight into what these controlled
human exposure studies may indicate regarding short-term exposure to
peak PM2.5 concentrations and how concentrations relate to
ambient PM2.5 concentrations, analyses in the 2022 PA (U.S.
EPA, 2022b, Figure 2-19) examine monitored 2-hour PM2.5
concentrations (the exposure window most often utilized in the
controlled human exposure studies) at sites meeting the current primary
PM2.5 standards to evaluate the degree to which 2-hour
ambient PM2.5 concentrations at such locations are likely to
exceed the 2-hour exposure concentrations in the controlled human
exposure studies at which statistically significant effects are
reported in multiple studies for one or more indicators of
cardiovascular function. At sites meeting the current primary
PM2.5 standards, most 2-hour concentrations are below 10
[mu]g/m\3\, and almost never exceed 30 [mu]g/m\3\. The extreme upper
end of the distribution of 2-hour PM2.5 concentrations is
shifted higher during the warmer months (April to September), generally
corresponding to the period of peak wildfire frequency in the U.S. At
sites meeting the current primary PM2.5 standards, the
highest 2-hour concentrations measured tend to occur during the period
of peak wildfire frequency (i.e., 99.9th percentile of 2-hour
concentrations is 62 [mu]g/m\3\ during the warm season considered as a
whole). Most of the sites measuring these very high concentrations are
in the northwestern U.S. and California (U.S. EPA, 2022b, Appendix A,
Figure A-1), where wildfires have been relatively common in recent
years. When the typical fire season is excluded from the analysis, the
extreme upper end of the distribution is reduced (i.e., 99.9th
percentile of 2-hour concentrations is 55 [mu]g/m\3\).\65\ Given these
results, the 2022 PA concludes that PM2.5 exposure
concentrations evaluated in most of these controlled human exposure
studies are well-above the 2-hour ambient PM2.5
concentrations typically measured in locations meeting the current
primary standards.
---------------------------------------------------------------------------
\65\ Similar analyses of 4-hour and 5-hour PM2.5
concentrations are presented in Appendix A, Figure A-2 and Figure A-
3, respectively of the 2022 PA (U.S. EPA, 2022b).
---------------------------------------------------------------------------
With respect to animal toxicological studies, the 2019 ISA relies
on animal toxicological studies to support the plausibility of a wide
range of PM2.5-related health effects. While animal
toxicological studies often examine more severe health outcomes and
longer exposure durations than controlled human exposure studies, there
is uncertainty in extrapolating the effects seen in animals, and the
PM2.5 exposures and doses that cause those effects, to human
populations. The 2022 PA considers these uncertainties when evaluating
what the available animal toxicological studies may indicate with
regard to the current primary PM2.5 standards.
As with controlled human exposure studies, most animal
toxicological studies evaluated in the 2019 ISA have examined effects
following exposure to PM2.5 well above the concentrations
likely to be allowed by the current PM2.5 standards. Such
studies have generally examined short-term exposures to
PM2.5 concentrations ranging from 100 to >1,000 [mu]g/m\3\
and long-term exposures to concentrations from 66 to >400 [mu]g/m\3\
(e.g., see U.S. EPA, 2019a, Table 1-2). Two exceptions are animal
toxicological studies reporting impaired lung development following
long-term exposures (i.e., 24 hours per day for several months
prenatally and postnatally) to an average PM2.5
concentration of 16.8 [mu]g/m\3\ (Mauad et al., 2008) and increased
carcinogenic potential following long-term exposures (i.e., 2 months)
to an average PM2.5 concentration of 17.7 [mu]g/m\3\
(Cangerana Pereira et al., 2011). These two studies report serious
effects following long-term exposures to PM2.5
concentrations similar to the ambient concentrations reported in some
PM2.5 epidemiologic studies (U.S. EPA, 2019a, Table 1-2),
though still above the ambient concentrations likely to occur in areas
meeting the current primary PM2.5 standards. However, noting
uncertainty in extrapolating the effects seen in animals, and the
PM2.5 exposures and doses that cause those effects to human
populations, animal toxicological studies are of limited utility in
informing decisions on the public health protection provided by the
current or alternative primary PM2.5 standards. Therefore,
the animal toxicological studies are most useful in providing further
evidence to support the biological mechanisms and plausibility of
various adverse effects.
ii. Ambient PM2.5 Concentrations in Locations of
Epidemiologic Studies
As summarized in section II.A.2.a above, epidemiologic studies
examining associations between daily or annual average PM2.5
exposures and mortality or morbidity represent a large part of the
evidence base supporting several of the 2019 ISA's ``causal'' and
``likely to be causal'' determinations. The 2022 PA considers the
ambient PM2.5 concentrations present in areas where
epidemiologic studies have evaluated associations with mortality or
morbidity, and what such concentrations may indicate regarding the
adequacy of the primary PM2.5 standards. The use of
information from epidemiologic studies to inform conclusions on the
primary PM2.5 standards is complicated by the fact that such
studies evaluate associations between distributions of ambient
PM2.5 and health outcomes, and do not identify the specific
exposures that can lead to the reported effects. Rather, health effects
can occur over the entire distribution of ambient PM2.5
concentrations evaluated, and epidemiologic studies conducted to date
do not identify a population-level threshold below which it can be
concluded with confidence that PM2.5-associated health
effects do not occur. Therefore, the 2022 PA evaluates the
PM2.5 air quality distributions over which epidemiologic
studies support health effect associations (U.S. EPA, 2022b, section
3.3.3.2). In the absence of discernible thresholds, the 2022 PA
considers the study-reported ambient PM2.5 concentrations
reflecting estimated exposure with a focus around the middle portion of
the PM2.5 air quality distribution, where the bulk of the
observed data reside and which provides the strongest support for
reported health effect associations. The section below, as well as in
more detail in section II.B.3.b.i of the proposal (88 FR 5594, January
27, 2023), describes the consideration of the key epidemiologic studies
and observations from these studies, as evaluated in the 2022 PA (U.S.
EPA, 2022b, section 3.3.3.2).
As an initial matter, in considering the PM2.5 air
quality distributions associated with mortality or morbidity in the key
epidemiologic studies, the 2022 PA recognizes that in previous reviews,
the decision framework used to judge adequacy of the existing
PM2.5 standards, and what levels of any potential
alternative standards should be considered, placed significant weight
on epidemiologic studies that assessed associations between
PM2.5 exposure and health outcomes that were most strongly
supported by the body of scientific evidence. In doing so, the decision
framework recognized that while there is no specific point in the air
quality distribution of any
[[Page 16239]]
epidemiologic study that represents a ``bright line'' at and above
which effects have been observed and below which effects have not been
observed, there is significantly greater confidence in the magnitude
and significance of observed associations for the part of the air
quality distribution corresponding to where the bulk of the health
events in each study have been observed, generally at or around the
mean concentration. This is the case both for studies of daily
PM2.5 exposures and for studies of annual average
PM2.5 exposures (U.S. EPA, 2022b, section 3.3.3.2.1).
As discussed further in the 2022 PA, studies of daily
PM2.5 exposures examine associations between day-to-day
variation in PM2.5 concentrations and health outcomes, often
over several years (U.S. EPA, 2022b, section 3.3.3.2.1). While there
can be considerable variability in daily exposures over a multi-year
study period, most of the estimated exposures reflect days with ambient
PM2.5 concentrations around the middle of the air quality
distributions examined (i.e., ``typical'' days rather than days with
extremely high or extremely low concentrations). Similarly, for studies
of annual PM2.5 exposures, most of the health events occur
at estimated exposures that reflect annual average PM2.5
concentrations around the middle of the air quality distributions
examined. In both cases, epidemiologic studies provide the strongest
support for reported health effect associations for this middle portion
of the PM2.5 air quality distribution, which corresponds to
the bulk of the underlying data, rather than the extreme upper or lower
ends of the distribution. Consistent with this, as noted in the 2022 PA
(U.S. EPA, 2022b, section 3.3.1.1), several epidemiologic studies
report that associations persist in analyses that exclude the upper
portions of the distributions of estimated PM2.5 exposures,
indicating that ``peak'' PM2.5 exposures are not
disproportionately responsible for reported health effect associations.
Thus, in considering PM2.5 air quality data from
epidemiologic studies, consistent with approaches in the 2012 and 2020
reviews (78 FR 3161, January 15, 2013; U.S. EPA, 2011, sections 2.1.3
and 2.3.4.1; 85 FR 82716-82717, December 18, 2020; U.S. EPA, 2020b,
sections 3.1.2 and 3.2.3), the 2022 PA evaluates study-reported means
(or medians) of daily and annual average PM2.5
concentrations as indicators for the middle portions of the air quality
distributions, over which studies generally provide strong support for
reported associations and for which confidence in the magnitude and
significance of associations observed in the epidemiologic studies is
greatest (78 FR 3101, January 15, 2013). In addition to the overall
study means, the 2022 PA also focuses on concentrations somewhat below
the means (e.g., 25th and 10th percentiles), when such information is
available from the epidemiologic studies, which again is consistent
with approaches used in previous reviews. In so doing, the 2022 PA
notes, as in previous reviews, that a relatively small portion of the
health events are observed in the lower part of the air quality
distribution and confidence in the magnitude and significance of the
associations begins to decrease in the lower part of the air quality
distribution. Furthermore, consistent with past reviews, there is no
single percentile value within a given air quality distribution that is
most appropriate or ``correct'' to use to characterize where our
confidence in associations becomes appreciably lower. However, and as
detailed further in the 2022 PA, the range from the 25th to 10th
percentiles is a reasonable range to consider as a region where there
is appreciably less confidence in the associations observed in
epidemiologic studies compared to the means (U.S. EPA, 2022b, p. 3-
69).\66\
---------------------------------------------------------------------------
\66\ As detailed in the 2011 PA, we note the interrelatedness of
the distributional statistics and a range of one standard deviation
around the mean which represents approximately 68% of normally
distributed data, and in that one standard deviation below the mean
falls between the 25th and 10th percentiles (U.S. EPA, 2011, p. 2-
71; U.S. EPA, 2005, p. 5-22).
---------------------------------------------------------------------------
In evaluating the overall study-reported means, and concentrations
somewhat below the means from epidemiologic studies, the 2022 PA
focuses on the form, averaging time and level of the current primary
annual PM2.5 standard. Consistent with the approaches used
in the 2012 and 2020 reviews (78 FR 3161-3162, January 15, 2013; 85 FR
82716-82717, December 18, 2020), the annual standard has been utilized
as the primary means of providing public health protection against the
bulk of the distribution of short- and long-term PM2.5
exposures. Thus, the evaluation of the study-reported mean
concentrations from key epidemiologic studies lends itself best to
evaluating the adequacy of the annual PM2.5 standard (rather
than the 24-hour standard with its 98th percentile form). This is true
for the study-reported means from both long-term and short-term
exposure epidemiologic studies, recognizing that the overall mean
PM2.5 concentrations reported in studies of short-term (24-
hour) exposures reflect averages across the study population and over
the years of the study. Thus, mean concentrations from short-term
exposure studies reflect long-term averages of 24-hour PM2.5
exposure estimates. In this manner, the examination of study-reported
means in key epidemiologic studies in the 2022 PA aims to evaluate the
protection provided by the annual PM2.5 standard against the
exposures where confidence is greatest for associations with mortality
and morbidity. In addition, the protection provided by the annual
standard is evaluated in conjunction with that provided by the 24-hour
standard, with its 98th percentile form, which aims to provide
supplemental protection against the short-term exposures to peak
PM2.5 concentrations that can occur in areas with strong
contributions from local or seasonal sources, even when overall ambient
mean PM2.5 concentrations in an area remain relatively low.
In focusing on the annual standard, and in evaluating the range of
study-reported exposure concentrations for which the strongest support
for adverse health effects exists, the 2022 PA examines exposure
concentrations in key epidemiologic studies to determine whether the
current primary annual PM2.5 standard provides adequate
protection against these exposure concentrations. This means, as in
past reviews, application of a decision framework based on assessing
means reported in key epidemiologic studies must also consider how the
study means were computed and how these values compare to the annual
standard metric (including the level, averaging time and form) and the
use of the monitor with the highest PM2.5 design value in an
area for compliance. In the 2012 review, it was recognized that the key
epidemiologic studies computed the study mean using an average across
monitor-based PM2.5 concentrations. As such, the Agency
noted that this decision framework applied an approach of using maximum
monitor concentrations to determine compliance with the standard, while
selecting the standard level based on consideration of composite
monitor concentrations. Further, the Agency included analyses (Hassett-
Sipple et al., 2010; Frank, 2012) that examined the differences in
these two metrics (i.e., maximum monitor concentrations and composite
monitor concentrations) across the U.S. and in areas included in the
key epidemiologic studies and found that the maximum design value in an
area was generally higher than the monitor average across that area,
with the difference varying
[[Page 16240]]
based on location and concentration. This information was taken into
account in the Administrator's final decision in selecting a level for
the primary annual PM2.5 standard the 2012 review and
discussed more specifically in her considerations on adequate margin of
safety.
Consistent with the approach taken in 2012, in assessing how the
overall mean (or median) PM2.5 concentrations reported in
key epidemiologic studies can inform conclusions on the primary annual
PM2.5 standard, the 2022 PA notes that the relationship
between mean PM2.5 concentrations and the area design value
continues to be an important consideration in evaluating the adequacy
of the current or potential alternative annual PM2.5
standard levels in this reconsideration. In a given area, the area
design value is based on the monitor in an area with the highest
PM2.5 concentrations and is used to determine compliance
with the standard. The highest PM2.5 concentrations
spatially distributed in the area would generally occur at or near the
area design value monitor and the distribution of PM2.5
concentrations would generally be lower in other locations and at
monitors in that area. As such, when an area is meeting a specific
annual standard level, the annual average exposures in that area are
expected to be at concentrations lower than that level and the average
of the annual average exposures across that area are expected (i.e., a
metric similar to the study-reported mean values) to be lower than that
level.\67\
---------------------------------------------------------------------------
\67\ In setting a standard level that would require the design
value monitor to meet a level equal to the study-reported mean
PM2.5 concentrations would generally result in lower
concentrations of PM2.5 across the entire area, such that
even those people living near an area design value monitor (where PM
concentrations are generally highest) will be exposed to
PM2.5 concentrations below the air quality conditions
reported in the epidemiologic studies.
---------------------------------------------------------------------------
Another important consideration is that there are a substantial
number of different types of epidemiologic studies available since the
2012 review, included in both the 2019 ISA and the ISA Supplement, that
make understanding the relationship between the mean PM2.5
concentrations and the area design value even more important (U.S. EPA,
2019a; U.S. EPA, 2022a). While the key epidemiologic studies in the
2012 review were all monitor-based studies, the newer studies include
hybrid modeling approaches, which have emerged in the epidemiologic
literature as an alternative to approaches that only use ground-based
monitors to estimate exposure. As assessed in the 2019 ISA and ISA
Supplement, a substantial number of epidemiologic studies used hybrid
model-based methods in evaluating associations between PM2.5
exposure and health effects (U.S. EPA, 2019a; U.S. EPA, 2022a). Hybrid
model-based studies employ various fusion techniques that combine
ground-based monitored data with air quality modeled estimates and/or
information from satellites to estimate PM2.5 exposures.\68\
Additionally, hybrid modeling approaches tend to broaden the areas
captured in the exposure assessment, and in so doing, tend to report
lower mean PM2.5 concentrations than monitor-based
approaches because they include more suburban and rural areas where
concentrations are lower. While these studies provide a broader
estimation of PM2.5 exposures compared to monitor-based
studies (i.e., PM2.5 concentrations are estimated in areas
without monitors), the hybrid modeling approaches result in study-
reported means that are more difficult to relate to the annual standard
metric and to the use of maximum monitor design values to assess
compliance. In addition, and to further complicate the comparison, when
looking across these studies, variations exist in how exposure is
estimated between such studies, which in turn affects how the study
means are calculated. Two important variations across studies include:
(1) Variability in spatial scale used (i.e., averages computed across
the nation (or large portions of the country) versus a focus on only
CBSAs) and (2) variability in exposure assignment methods (i.e.,
averaging across all grid cells [non-population weighting], averaging
across a scaled-up area like a ZIP code [aspects of population
weighting applied], and/or applying population weighting). To elaborate
further on the variability in exposure assignment methods, studies that
use hybrid modeling approaches can estimate PM2.5
concentrations at different spatial resolutions, including at 1 km x 1
km grid cells, at 12 km x 12 km grid cells, or at the census tract
level. Mean reported PM2.5 concentrations can then be
estimated either by averaging up to a larger spatial resolution that
corresponds to the spatial resolution for which health data exists
(e.g., ZIP code level) and therefore apply aspects of population
weighting. These values are then averaged across all study locations at
the larger spatial resolution (e.g., averaged across all ZIP codes in
the study) over the study period, resulting in the study-reported mean
24-hour average or average annual PM2.5 concentration. Other
studies that use hybrid modeling methods to estimate PM2.5
concentrations may use each grid cell to calculate the study-reported
mean 24-hour average or average annual PM2.5 concentration.
As such, these types of studies do not apply population weighting in
their mean concentrations. In studies that use each grid cell to report
a mean PM2.5 concentration and do not apply aspects of
population weighting, the study mean may not reflect the exposure
concentrations used in the epidemiologic study to assess the reported
association. The impact of the differences in methods is an important
consideration when comparing mean concentrations across studies (U.S.
EPA, 2022b, section 3.3.3.2.1). Thus, the 2022 PA also considers the
methods used to estimate PM2.5 concentrations, which vary
from traditional methods using monitoring data from ground-based
monitors \69\ to those using more complex hybrid modeling approaches
and how these methods calculate the study-reported mean
PM2.5 concentration.\70\
---------------------------------------------------------------------------
\68\ More detailed information about hybrid model methods and
performance is described in section 2.3.3.2 of the 2022 PA (U.S.
EPA, 2022b).
\69\ In those studies that use ground-based monitors alone to
estimate long- or short-term PM2.5 concentrations,
approaches include: (1) PM2.5 concentrations from a
single monitor within a city/county; (2) average of PM2.5
concentrations across all monitors within a city/county or other
defined study area (e.g., CBSA); or (3) population-weighted averages
of exposures. Once the study location average PM2.5
concentration is calculated, the study-reported long-term average is
derived by averaging daily/annual PM2.5 concentrations
across all study locations over the entire study period.
\70\ Detailed information on the methods by which mean
PM2.5 concentrations are calculated in key monitor- and
hybrid model-based U.S. and Canadian epidemiologic studies are
presented in Tables 3-6 through 3-9 in the 2022 PA (U.S. EPA,
2022b).
---------------------------------------------------------------------------
Given the emergence of the hybrid model-based epidemiologic studies
since the 2012 review, the 2022 PA explores the relationship between
the approaches used in these studies to estimate PM2.5
concentrations and the impact that the different methods have on the
study-reported mean PM2.5 concentrations. The 2022 PA
further seeks to understand how the approaches and resulting mean
concentrations compare across studies, as well as what the resulting
mean values represent relative to the annual standard. In so doing, the
2022 PA presents analyses that compare the area annual design values,
composite monitor PM2.5 concentrations, and mean
concentrations from two hybrid modeling approaches, including
evaluation of the means when population weighting is applied and when
population weighting is not
[[Page 16241]]
applied (U.S. EPA, 2022b, section 2.3.3.1).
In the air quality analyses comparing composite monitored
PM2.5 concentrations with annual PM2.5 design
values in U.S. CBSAs, maximum annual PM2.5 design values
were approximately 10% to 20% higher than annual average composite
monitor concentrations (i.e., averaged across multiple monitors in the
same CBSA) (sections I.D.5.a above and U.S. EPA, 2022b, section
2.3.3.1, Figure 2-28 and Table 2-3). The difference between the maximum
annual design value and average concentration in an area can be smaller
or larger than this range (10-20%), depending on a variety of factors
such as the number of monitors, monitor siting characteristics, the
distribution of ambient PM2.5 concentrations, and how the
average concentrations are calculated (i.e., averaged across monitors
versus across modeled grid cells). Results of this analysis suggest
that there will be a distribution of concentrations across an area and
the maximum annual average monitored concentration in an area (at the
design value monitor, used for compliance with the standard), will
generally be 10-20% higher than the average PM2.5
concentration across the other monitors in the area. Thus, in
considering how the annual standard levels would relate to the study-
reported means from key monitor-based epidemiologic studies, the 2022
PA generally concludes that an annual standard level that is no more
than 10-20% higher than monitor-based study-reported mean
PM2.5 concentrations would generally maintain air quality
exposures to be below those associated with the study-reported mean
PM2.5 concentrations, exposures for which the strongest
support for adverse health effects occurring is available.
The 2022 PA also evaluates data from two hybrid modeling approaches
(DI2019 and HA2020) that have been used in several recent epidemiologic
studies (U.S. EPA, 2022b, section 2.3.3.2.4).\71\ The analysis shows
that the means differ when PM2.5 concentrations are
estimated in urban areas only (CBSAs) versus when the averages were
calculated with all or most grid cells nationwide, likely because areas
included outside of CBSAs tend to be more rural and have lower
estimated PM2.5 concentrations. The 2022 PA recognizes the
importance of this variability in the means since the study areas
included in the calculation of the mean, and more specifically whether
a study is focused on nationwide, regional, or urban areas, will affect
the calculation of the study mean based on how many rural areas, with
lower estimated PM2.5 concentrations, are included in the
study area. While the determination of what spatial scale to use to
estimate PM2.5 concentrations does not inherently affect the
quality of the epidemiologic study, the spatial scale can influence the
calculated reported long-term mean concentration across the study area
and period. The results of the analysis show that, regardless of the
hybrid modeling approach assessed, the annual average PM2.5
concentrations in CBSA-only analyses are 4-8% higher than for
nationwide analyses, likely as a result of higher PM2.5
concentrations in more densely populated areas, and exclusion of more
rural areas (U.S. EPA, 2022b, Table 2-4). When evaluating comparisons
between surfaces that estimate exposure using aspects of population
weighting versus surfaces that do not calculate means using population
weighting, surfaces that calculate long-term mean PM2.5
concentrations with population-weighted averages have higher average
annual PM2.5 concentrations, compared to annual
PM2.5 concentrations in analyses that do not apply
population weighting.\72\ Analyses show that average maximum annual
design values are 40 to 50% higher when compared to annual average
PM2.5 concentrations estimated without population weighting
versus 15% to 18% higher when compared to average annual
PM2.5 concentrations estimated with population weighting
applied (similar to the differences observed for the composite monitor
comparison values for the monitor-based epidemiologic studies) (U.S.
EPA, 2022b, section 2.3.3.2.4). Given these results, it is worth noting
that for the studies using the hybrid modeling approaches, the choice
of methodology employed in calculating the study-reported means (i.e.,
using population weighting or not), and not a difference in estimates
of exposure in the study itself, can produce substantially different
study-reported mean values, where approaches that do not apply
population weighting leading to much lower estimated mean
PM2.5 concentrations.
---------------------------------------------------------------------------
\71\ More details on the evaluation of the two hybrid modeling
approaches is provided in section 2.3.3.2.4 of the 2022 PA (U.S.
EPA, 2022b).
\72\ The annual PM2.5 concentrations for the
population-weighted averages ranged from 8.2-10.2 [mu]g/m\3\, while
those that do not apply population weighting ranged from 7.0-8.6
[mu]g/m\3\. Average maximum annual design values ranged from 9.5 to
11.7 [mu]g/m\3\.
---------------------------------------------------------------------------
Based on these results, and similar to conclusions for the monitor-
based studies, the 2022 PA generally concludes that study-reported mean
concentrations in the studies that employ hybrid modeling approaches
and calculate a population-weighted mean are associated with air
quality conditions that would be achieved by meeting annual standard
levels that are 15-18% higher than study-reported means. Therefore, an
annual standard level that is no more than 15-18% higher than the
study-reported means would generally maintain air quality exposures to
be below those associated with the study-reported mean PM2.5
concentrations, exposures for which we have the strongest support for
adverse health effects occurring. For the studies that utilize hybrid
modeling approaches but do not incorporate population weighting in
calculating the mean, the annual design values associated with these
air quality conditions are expected to be much higher (i.e., 40-50%
higher) and this larger difference makes it more difficult to consider
how these studies can be used to determine the adequacy of the
protection afforded by the current or potential alternative annual
standards. Additionally, as noted above in studies that utilize hybrid
modeling approaches and that do not incorporate population weighting in
calculating the mean (e.g., use each grid cell to calculate a mean
PM2.5 concentration), the study mean does not reflect the
exposure concentrations used in the epidemiologic study to assess the
reported association.
The 2022 PA notes that while these analyses can be useful to
informing the understanding of the relationship between study-reported
mean concentrations and the level of the annual standard, some
limitations of this analysis must be recognized (U.S. EPA, 2022a,
section 3.3.3.2.1). First, the comparisons used only two hybrid
modeling approaches. Although these two hybrid modeling surfaces have
been used in a number of recent epidemiologic studies, they represent
just two of the many hybrid modeling approaches that have been used in
epidemiologic studies to estimate PM2.5 concentrations.
These methods continue to evolve, with further development and
improvement to prediction models that estimate PM2.5
concentrations in epidemiologic studies. In addition to differences in
hybrid modeling approaches, epidemiologic studies also use different
methods to assign a population weighted average PM2.5
concentration to their study population, and the assessment presented
in the
[[Page 16242]]
2022 PA does not evaluate all of the potential methods that could be
used.
Additionally, while some of these epidemiologic studies also
provide information on the broader distributions of exposure estimates
and/or health events and the PM2.5 concentrations
corresponding to the lower percentiles of those data (e.g., 25th and/or
10th), the air quality analysis in the 2022 PA focuses on mean
PM2.5 concentrations and a similar comparison for lower
percentiles of data was not assessed. Therefore, any direct comparison
of study-reported PM2.5 concentrations corresponding to
lower percentiles and annual design values is more uncertain than such
comparisons with the mean. Finally, air quality analysis presented in
the 2022 PA and detailed above in section I.D.5 included two hybrid
modeling-based approaches that used U.S.-based air quality information
for estimating PM2.5 concentrations. As such, the analyses
are most relevant to interpreting the study-reported mean
concentrations from U.S. epidemiologic studies and do not provide
additional information about how the mean exposures concentrations
reported in epidemiologic studies in other countries would compare to
annual design values observed in the U.S. In addition, while
information from Canadian studies can be useful in assessing the
adequacy of the annual standard, differences in the exposure
environments and population characteristics between the U.S. and other
countries can affect the study-reported mean value and its relationship
with the annual standard level. Sources and pollutant mixtures, as well
as PM2.5 concentration gradients, may be different between
countries, and the exposure environments in other countries may differ
from those observed in the U.S. Furthermore, differences in population
characteristics and population densities can also make it challenging
to directly compare studies from countries outside of the U.S. to a
design value in the U.S.
As with the experimental studies discussed above, the 2022 PA
focuses on epidemiologic studies assessed in the 2019 ISA and ISA
Supplement that have the potential to be most informative in reaching
decisions on the adequacy of the primary PM2.5 standards.
The 2022 PA focuses on epidemiologic studies that provide strong
support for ``causal'' or ``likely to be causal'' relationships with
PM2.5 exposures in the 2019 ISA. Further, the 2022 PA also
focuses on the health effect associations that are determined in the
2019 ISA and ISA Supplement to be consistent across studies, coherent
with the broader body of evidence (e.g., including animal and
controlled human exposure studies), and robust to potential confounding
by co-occurring pollutants and other factors.\73\ In particular the
2022 PA considers the U.S. and Canadian epidemiologic studies to be
more useful for reaching conclusions on the current standards than
studies conducted in other countries, given that the results of the
U.S. and Canadian studies are more directly applicable for quantitative
considerations, whereas studies conducted in other countries reflect
different populations, exposure characteristics, and air pollution
mixtures. Additionally, epidemiologic studies outside of the U.S. and
Canada generally reflect higher PM2.5 concentrations in
ambient air than are currently found in the U.S., and are less relevant
to informing questions about adequacy of the current standards.\74\
However, and as noted above, the 2022 PA also recognizes that while
information from Canadian studies can be useful in assessing the
adequacy of the annual standard, there are still important differences
between the exposure environments in the U.S. and Canada and
interpreting the data (e.g., mean concentrations) from the Canadian
studies in the context of a U.S.-based standard may present challenges
in directly and quantitatively informing questions regarding the
adequacy of the current or potential alternative the levels of the
annual standard. Lastly, the 2022 PA emphasizes multicity/multistate
studies that examine health effect associations, as such studies are
more encompassing of the diverse atmospheric conditions and population
demographics in the U.S. than studies that focus on a single city or
State. Figures 3-4 through 3-7 in the 2022 PA summarize the study
details for the key U.S. and Canadian epidemiologic studies (U.S. EPA,
2022b, section 3.3.3.2.1).\75\
---------------------------------------------------------------------------
\73\ As described in the Preamble to the ISAs (U.S. EPA, 2015),
``the U.S. EPA emphasizes the importance of examining the pattern of
results across various studies and does not focus solely on
statistical significance or the magnitude of the direction of the
association as criteria of study reliability. Statistical
significance is influenced by a variety of factors including, but
not limited to, the size of the study, exposure and outcome
measurement error, and statistical model specifications. Statistical
significance may be informative; however, it is just one of the
means of evaluating confidence in the observed relationship and
assessing the probability of chance as an explanation. Other
indicators of reliability such as the consistency and coherence of a
body of studies as well as other confirming data may be used to
justify reliance on the results of a body of epidemiologic studies,
even if results in individual studies lack statistical significance.
Traditionally, statistical significance is used to a larger extent
to evaluate the findings of controlled human exposure and animal
toxicological studies. Understanding that statistical inferences may
result in both false positives and false negatives, consideration is
given to both trends in data and reproducibility of results. Thus,
in drawing judgments regarding causality, the U.S. EPA emphasizes
statistically significant findings from experimental studies, but
does not limit its focus or consideration to statistically
significant results in epidemiologic studies.''
\74\ This emphasis on studies conducted in the U.S. or Canada is
consistent with the approach in the 2012 and 2020 reviews of the PM
NAAQS (U.S. EPA, 2011, section 2.1.3; U.S. EPA, 2020b, section
3.2.3.2.1) and with approaches taken in other NAAQS reviews.
However, the importance of studies in the U.S., Canada, and other
countries in informing an ISA's considerations of the weight of the
evidence that informs causality determinations is recognized.
\75\ The cohorts examined in the studies included in Figure 3-4
to Figure 3-7 of the 2022 PA include large numbers of individuals in
the general population, and often also include those populations
identified as at-risk (i.e., children, older adults, minority
populations, and individuals with pre-existing cardiovascular and
respiratory disease).
---------------------------------------------------------------------------
The key epidemiologic studies identified in the 2022 PA indicate
generally positive and statistically significant associations between
estimated PM2.5 exposures (short- or long-term) and
mortality or morbidity across a range of ambient PM2.5
concentrations (U.S. EPA, 2022b, section 3.3.3.2.1), report overall
mean (or median) PM2.5 concentrations, and include those for
which the years of PM2.5 air quality data used to estimate
exposures overlap entirely with the years during which health events
are reported.\76\ Additionally, for studies that estimate
PM2.5 exposure using hybrid modeling approaches, the 2022 PA
also considers the approach used to estimate PM2.5
concentrations and the approach used to validate hybrid model
predictions when evaluating those studies as key epidemiologic studies
\77\ and focuses on those studies that use recent methods based on
surfaces that are with fused with monitored PM2.5
[[Page 16243]]
concentration data (U.S. EPA, 2022b, section 3.3.3.2.1).
---------------------------------------------------------------------------
\76\ For some studies of long-term PM2.5 exposures,
exposure is estimated from air quality data corresponding to only
part of the study period, often including only the later years of
the health data, and are not likely to reflect the full ranges of
ambient PM2.5 concentrations that contributed to reported
associations. While this approach can be reasonable in the context
of an epidemiologic study that is evaluating health effect
associations with long-term PM2.5 exposures, under the
assumption that spatial patterns in PM2.5 concentrations
are not appreciably different during time periods for which air
quality information is not available (e.g., Chen et al., 2016), the
2022 PA focuses on the distribution of ambient PM2.5
concentrations that could have contributed to reported health
outcomes. Therefore, the 2022 PA identifies studies as key
epidemiologic studies when the years of air quality data and health
data overlap in their entirety.
\77\ Such studies are identified as those that use hybrid
modeling approaches for which recent methods and models were used
(e.g., recent versions and configurations of the air quality
models); studies that are fused with PM2.5 data from
national monitoring networks (i.e., FRM/FEM data); and studies that
reported a thorough model performance evaluation for core years of
the study.
---------------------------------------------------------------------------
Figure 1 below (U.S. EPA, 2022b, Figure 3-8) highlights the overall
mean (or median) PM2.5 concentrations reported in key U.S.
studies that use ground-based monitors alone to estimate long- or
short-term PM2.5 exposure.\78\ For the small subset of
studies with available information on the broader distributions of
underlying data, Figure 1 below also identifies the study-period
PM2.5 concentrations corresponding to the 25th and 10th
percentiles of health events \79\ (see Appendix B, Section B.2 of the
2022 PA for more information). Figure 2 (U.S. EPA, 2022a, Figure 3-14)
presents overall means of predicted PM2.5 concentrations for
key U.S. model-based epidemiologic studies that apply aspects of
population-weighting, and the concentrations corresponding to the 25th
and 10th percentiles of estimated exposures or health events \80\ when
available (see Appendix B, section B.3 for additional information).\81\
---------------------------------------------------------------------------
\78\ Canadian studies that use ground-based monitors estimate
long- or short-term PM2.5 exposures are found in Figure
3-9 of the 2022 PA, including concentrations corresponding to the
25th and 10th percentiles of estimated exposures or health events,
when available (U.S. EPA, 2022b).
\79\ That is, 25% of the total health events occurred in study
locations with mean PM2.5 concentrations (i.e., averaged
over the study period) below the 25th percentiles identified in
Figure 3-8 of the 2022 PA and 10% of the total health events
occurred in study locations with mean PM2.5
concentrations below the 10th percentiles identified.
\80\ For most studies in Figure 2 below (Figure 3-14 in the 2022
PA), 25th percentiles of exposure estimates are presented. The
exception is Di et al. (2017b), for which Figure 2 (U.S. EPA, 2022b,
Figure 3-14) presents the short-term PM2.5 exposure
estimates corresponding to the 25th and 10th percentiles of deaths
in the study population (i.e., 25% and 10% of deaths occurred at
concentrations below these concentrations). In addition, the authors
of Di et al. (2017b) provided population-weighted exposure values.
The 10th and 25th percentiles of these population-weighted exposure
estimates are 7.9 and 9.5 [mu]g/m\3\, respectively.
\81\ Overall mean (or median) PM2.5 concentrations
reported in key Canadian studies that use model-based approaches to
estimate long- or short-term PM2.5 concentrations and the
concentrations corresponding to the 25th and 10th percentiles of
estimated exposures or health events, when available are found in
Figure 3-9 of the 2022 PA (U.S. EPA, 2022b).
---------------------------------------------------------------------------
[[Page 16244]]
BILLING CODE 6560-50-P
[GRAPHIC] [TIFF OMITTED] TR06MR24.029
[[Page 16245]]
[GRAPHIC] [TIFF OMITTED] TR06MR24.030
BILLING CODE 6560-50-C
Based on its evaluation of study-reported mean concentrations, the
2022 PA notes that key epidemiologic studies conducted in the U.S. or
Canada report generally positive and statistically significant
associations between estimated PM2.5 exposures (short- or
long-term) and mortality or morbidity across a wide range of ambient
PM2.5 concentrations (U.S. EPA, 2022b, section 3.3.3.2.1).
The 2022 PA makes a number of observations with regard to the study-
reported PM2.5 concentrations in the key U.S. and Canadian
epidemiologic studies.
The 2022 PA first considers the PM2.5 concentrations
from the key U.S. epidemiologic studies. For studies that use monitors
to estimate PM2.5 exposures, overall mean PM2.5
concentrations range between 9.9 [mu]g/m\3\ \82\ to 16.5 [mu]g/m\3\
(Figure 1 above and
[[Page 16246]]
U.S. EPA, 2022b, Figure 3-8). For key U.S. epidemiologic studies that
use hybrid model-predicted exposures and apply aspects of population-
weighting, mean PM2.5 concentrations range from 9.3 [mu]g/
m\3\ to just above 12.2 [micro]g/m\3\ (Figure 2 above and U.S. EPA,
2022b, Figure 3-14). In studies that average up from the grid cell
level to the ZIP code, postal code, or census tract level, mean
PM2.5 concentrations range from 9.8 [micro]g/m\3\ to 12.2
[micro]g/m\3\. The one study that population-weighted the grid cell
prior to averaging up to the ZIP code or census tract level reported
mean PM2.5 concentrations of 9.3 [mu]g/m\3\. Based on air
quality analyses noted above, these hybrid modelled epidemiologic
studies are expected to report means similar to those from monitor-
based studies.
---------------------------------------------------------------------------
\82\ This is generally consistent with, but slightly below, the
lowest study-reported mean PM2.5 concentration from
monitor-based studies available in the 2020 PA, which was 10.7
[mu]g/m\3\ (U.S. EPA, 2020a, Figure 3-7).
---------------------------------------------------------------------------
Other key U.S. epidemiologic studies that use hybrid modeling
approaches estimate mean PM2.5 exposure by averaging each
grid cell across the entire study area, whether that be the nation or a
region of the country. These studies do not weight the estimated
exposure concentrations based on population density or location of
health events. As such, the study mean reported in these studies may
not reflect the exposure concentrations used in the epidemiologic study
to assess the reported association. As a result, these reported mean
concentrations are the most different (and much lower) than the means
reported in monitor-based studies. Due to the methodology employed in
calculating the study-reported means and not necessarily a difference
in estimates of exposure, these epidemiologic studies are expected to
report some of the lowest mean values. For these studies, the reported
mean PM2.5 concentrations range from 8.1 [mu]g/m\3\ to 11.9
[mu]g/m\3\ (U.S. EPA, 2022b, Figure 3-14). As noted above, for studies
that utilize hybrid modeling approaches but do not incorporate
population weighting into the reported mean calculation, the associated
annual design values would be expected to be much higher (i.e., 40-50%
higher) than the study-reported means. This larger difference between
design values and study-reported mean concentrations makes it more
difficult to consider how these studies can be used to determine the
adequacy of the protection afforded by the current or potential
alternative annual standards (U.S. EPA, 2022b, section 3.3.3.2.1).
In addition to the mean PM2.5 concentrations, a subset
of the key U.S. epidemiologic studies report PM2.5
concentrations corresponding to the 25th and 10th percentiles of health
data or exposure estimates to provide insight into the concentrations
that comprise the lower quartile of the air quality distributions. In
studies that use monitors to estimate PM2.5 exposures, 25th
percentiles of health events correspond to PM2.5
concentrations (i.e., averaged over the study period for each study
city) at or above 11.5 [micro]g/m\3\ and 10th percentiles of health
events correspond to PM2.5 concentrations at or above 9.8
[micro]g/m\3\ (i.e., 25% and 10% of health events, respectively, occur
in study locations with PM2.5 concentrations below these
values) (Figure 1 above and U.S. EPA, 2022b, Figure 3-8). Of the key
U.S. epidemiologic studies that use hybrid modeling approaches and
apply population-weighting to estimate long-term PM2.5
exposures, the ambient PM2.5 concentrations corresponding to
25th percentiles of estimated exposures are 9.1 [micro]g/m\3\ (Figure 2
and U.S. EPA, 2022b, Figure 3-14). In key U.S. epidemiologic studies
that use hybrid modeling approaches and apply population-weighting to
estimate short-term PM2.5 exposures, the ambient
concentrations corresponding to 25th percentiles of estimated
exposures, or health events, are 6.7 [micro]g/m\3\ (Figure 2 and U.S.
EPA, 2022b, Figure 3-14). In key U.S. epidemiologic studies that use
hybrid modeling approaches and do not apply population-weighting to
estimate PM2.5 exposures, the ambient concentrations
corresponding to 25th percentiles of estimated exposures, or health
events, range from 4.6 to 9.2 [micro]g/m\3\ (U.S. EPA, 2022b, Figure 3-
14).\83\ In the key epidemiologic studies that apply hybrid modeling
approaches with population-weighting and with information available on
the 10th percentile of health events, the ambient PM2.5
concentration corresponding to that 10th percentile range from 4.7
[micro]g/m\3\ to 7.3 [micro]g/m\3\ (Figure 2 and U.S. EPA, 2022b,
Figure 3-14).
---------------------------------------------------------------------------
\83\ In the one study that reports 25th percentile exposure
estimates of 4.6 [micro]g/m\3\ (Shi et al., 2016), the authors
report that most deaths occurred at or above the 75th percentile of
annual exposure estimates (i.e., 10 [mu]g/m\3\). The short-term
exposure estimates accounting for most deaths are not presented in
the published study.
---------------------------------------------------------------------------
The 2022 PA next considers the PM2.5 concentrations from
the key Canadian epidemiologic studies. Generally, the study-reported
mean concentrations in Canadian studies are lower than those reported
in the U.S. studies for both monitor-based and hybrid model methods.
For the majority of key Canadian epidemiologic studies that use
monitor-based exposure, mean PM2.5 concentrations generally
ranged from 7.0 [micro]g/m\3\ to 9.0 [micro]g/m\3\ (U.S. EPA, 2022b,
Figure 3-9). For these studies, 25th percentiles of health events
correspond to PM2.5 concentrations at or above 6.5 [micro]g/
m\3\ and 10th percentiles of health events correspond to
PM2.5 concentrations at or above 6.4 [micro]g/m\3\ (U.S.
EPA, 2022b, Figure 3-9). For the key Canadian epidemiologic studies
that use hybrid model-predicted exposure, the mean PM2.5
concentrations are generally lower than in U.S. model-based studies
(U.S. EPA, 2022b, Figure 3-10), ranging from approximately 6.0
[micro]g/m\3\ to just below 10.0 [micro]g/m\3\ (U.S. EPA, 2022b, Figure
3-11). The majority of the key Canadian epidemiologic studies that used
hybrid modeling were completed at the nationwide scale, while four
studies were completed at the regional geographic spatial scale. In
addition, all the key Canadian epidemiologic studies apply aspects of
population weighting, where all grid cells within a postal code are
averaged, individuals are assigned exposure at the postal code
resolution, and study mean PM2.5 concentrations are based on
the average of individual exposures. The majority of studies estimating
exposure nationwide range between just below 6.0 [micro]g/m\3\ to 8.0
[micro]g/m\3\ (U.S. EPA, 2022b, Figure 3-11). One study by Erickson et
al. (2020) presents an analysis related immigrant status and length of
residence in Canada versus non-immigrant populations, which accounts
for the four highest mean PM2.5 concentrations which range
between 9.0 [micro]g/m\3\ and 10.0 [micro]g/m\3\ (U.S. EPA, 2022b,
Figure 3-11). The four studies that estimate exposure at the regional
scale report mean PM2.5 concentrations that range from 7.8
[micro]g/m\3\ to 9.8 [micro]g/m\3\ (U.S. EPA, 2022b, Figure 3-11).
Three key Canadian epidemiologic studies report information on the 25th
percentile of health events. In these studies, the ambient
PM2.5 concentration corresponding to the 25th percentile is
approximately 8.0 [micro]g/m\3\ in two studies, and 4.3 [micro]g/m\3\
in a third study (U.S. EPA, 2022b, Figure 3-11).
In addition to the expanded body of evidence from the key U.S.
epidemiologic studies discussed above, there are also a subset of
epidemiologic studies that have emerged that further inform an
understanding of the relationship between PM2.5 exposure and
health effects, including studies with the highest exposures excluded
(restricted analyses), epidemiologic studies that employed statistical
approaches that attempt to more extensively account for confounders and
are more robust to model misspecification (i.e., used alternative
[[Page 16247]]
methods for confounder control),\84\ and accountability studies (U.S.
EPA, 2019a, U.S. EPA, 2021a, U.S. EPA, 2022a).
---------------------------------------------------------------------------
\84\ As noted in the ISA Supplement (U.S. EPA, 2022a, p. 1-3):
``In the peer-reviewed literature, these epidemiologic studies are
often referred to as alternative methods for confounder control. For
the purposes of this Supplement, this terminology is not used to
prevent confusion with the main scientific conclusions (i.e., the
causality determinations) presented within an ISA. In addition, as
is consistent with the weight-of-evidence framework used within ISAs
and discussed in the Preamble to the Integrated Science Assessments,
an individual study on its own cannot inform causality, but instead
represents a piece of the overall body of evidence.''
---------------------------------------------------------------------------
Restricted analyses are studies that examine health effect
associations in analyses with the highest exposures excluded,
restricting analyses to daily exposures less than the 24-hour primary
PM2.5 standard and annual exposures less than the annual
PM2.5 standard. The 2022 PA presents a summary of restricted
analyses evaluated in the 2019 ISA and ISA Supplement (U.S. EPA, 2022b,
Table 3-10). The restricted analyses can be informative in assessing
the nature of the association between long-term exposures (e.g., annual
average concentrations <12.0 [micro]g/m\3\) or short-term exposures
(e.g., daily concentrations <35 [micro]g/m\3\) when looking only at
exposures to lower concentrations, including whether the association
persists in such restricted analyses compared to the same analyses for
all exposures, as well as whether the association is stronger, in terms
of magnitude and precision, than when completing the same analysis for
all exposures. While these studies are useful in supporting the
confidence and strength of associations at lower concentrations, these
studies also have inherent uncertainties and limitations, including
uncertainty in how studies exclude concentrations (e.g., are they
excluded at the modeled grid cell level, the ZIP code level) and in how
concentrations in studies that restrict air quality data relate to
design values for the annual and 24-hour standards. Further, these
studies often do not report descriptive statistics (e.g., mean
PM2.5 concentrations, or concentrations at other
percentiles) that allow for additional consideration of this
information. As such, while these studies can provide additional
supporting evidence for associations at lower concentrations, the 2022
PA notes that there are also limitations in how to interpret these
studies when evaluating the adequacy of the current or potential
alternative standards.
Restricted analyses provide additional information on the nature of
the association between long- or short-term exposures when analyses are
restricted to lower PM2.5 concentrations and indicate that
effect estimates are generally greater in magnitude in the restricted
analyses for long- and short-term PM2.5 exposure compared to
the main analyses. In two U.S. studies that report mean
PM2.5 concentrations in restricted analyses and that
estimate effects associated with long-term exposure to
PM2.5, the effect estimates are greater in the restricted
analyses than in the main analyses. Di et al. (2017a) and Dominici et
al. (2019) report positive and statistically significant associations
in analyses restricted to concentrations less than 12.0 [micro]g/m\3\
for all-cause mortality and effect estimates are greater in the
restricted analyses than effect estimates reported in main analyses. In
addition, both studies report mean PM2.5 concentrations of
9.6 [micro]g/m\3\. While none of the U.S. studies of short-term
exposure present mean PM2.5 concentrations for the
restricted analyses, these studies generally have mean 24-hour average
PM2.5 concentrations in the main analyses below 12.0
[micro]g/m\3\, and report increases in the effect estimates in the
restricted analyses compared to the main analyses. Additionally, in the
one Canadian study of long-term PM2.5 exposure, Zhang et al.
(2021) conducted analyses where annual PM2.5 concentrations
were restricted to concentrations below 10.0 [micro]g/m\3\ and 8.8
[micro]g/m\3\, which presumably have lower mean concentrations than the
mean of 7.8 [micro]g/m\3\ reported in the main analyses, though
restricted analysis mean PM2.5 concentrations are not
reported. Effect estimates for non-accidental mortality are greater in
analyses restricted to PM2.5 concentrations less than 10.0
[micro]g/m\3\, but less in analyses restricted to <8.8 [micro]g/m\3\.
The second type of studies that have recently emerged and further
inform the consideration of the relationship between PM2.5
exposure and health effects in the 2022 PA are those that employ
alternative methods for confounder control. Alternative methods for
confounder control seek to mimic randomized experiments through the use
of study design and statistical methods to more extensively account for
confounders and are more robust to model misspecification. The 2022 PA
presents a summary of the studies that employ alternative methods for
confounder control, and employ a variety of statistical methods, which
are evaluated in the 2019 ISA and ISA Supplement (U.S. EPA, 2022b,
Table 3-11). These studies reported consistent results among large
study populations across the U.S. and can further inform the
relationship between long- and short-term PM2.5 exposure and
total mortality. Studies that employ alternative methods for confounder
control to assess the association between long-term exposure to
PM2.5 and mortality reduce uncertainties related to
confounding and provide additional support for the associations
reported in the broader body of cohort studies that examined long-term
PM2.5 exposure and mortality.
Lastly, there is a subset of epidemiologic studies that assess
whether long-term reductions in ambient PM2.5 concentrations
result in corresponding reductions in health outcomes. These include
studies that evaluate the potential for improvements in public health,
including reductions in mortality rates, increases in life expectancy,
and reductions in respiratory disease as ambient PM2.5
concentrations have declined over time. Some of these studies,
accountability studies, provide insight on whether the implementation
of environmental policies or air quality interventions result in
changes/reductions in air pollution concentrations and the
corresponding effect on health outcomes.\85\ The 2022 PA presents a
summary of these studies, which are assessed in the 2019 ISA and ISA
Supplement (U.S. EPA, 2022b, Table 3-12). These studies lend support
for the conclusion that improvements in air quality are associated with
improvements in public health.
---------------------------------------------------------------------------
\85\ Given the nature of these studies, the majority tend to
focus on time periods in the past during which ambient
PM2.5 concentrations were substantially higher than those
measured more recently (e.g., see U.S. EPA, 2022b, Figure 2-16).
---------------------------------------------------------------------------
More specifically, of the accountability studies that account for
changes in PM2.5 concentrations due to a policy or the
implementation of an intervention and whether there was evidence of
changes in associations with mortality or cardiovascular effects as a
result of changes in annual PM2.5 concentrations, Corrigan
et al. (2018), Henneman et al. (2019) and Sanders et al. (2020a)
present analyses with starting PM2.5 concentrations (or
concentrations prior to the policy or intervention) below 12.0
[micro]g/m\3\. Henneman et al. (2019) explored changes in modeled
PM2.5 concentrations following the retirement of coal fired
power plants in the U.S., and found that reductions from mean annual
PM2.5 concentrations of 10.0 [micro]g/m\3\ in 2005 to mean
annual PM2.5 concentrations of 7.2 [micro]g/m\3\ in 2012
from coal-fueled power plants resulted in corresponding reductions in
the number of cardiovascular-related
[[Page 16248]]
hospital admissions, including for all cardiovascular disease, acute
MI, stroke, heart failure, and ischemic heart disease in those aged 65
and older. Corrigan et al. (2018) examined whether there was a change
in the cardiovascular mortality rate before (2000-2004) and after
(2005-2010) implementation of the first annual PM2.5 NAAQS
implementation based on mortality data from the National Center for
Health Statistics and reported 1.10 (95% confidence interval (CI):
0.37, 1.82) fewer cardiovascular deaths per year per 100,000 people for
each 1 [mu]g/m\3\ reduction in annual PM2.5 concentrations.
When comparing whether counties met the annual PM2.5
standard (attainment counties), there were 1.96 (95% CI: 0.77, 3.15)
fewer cardiovascular deaths for each 1 [mu]g/m\3\ reduction in annual
PM2.5 concentrations between the two periods for attainment
counties, whereas in non-attainment counties (e.g., counties that did
not meet the annual PM2.5 standard), there were 0.59 (95%
CI: - 0.54, 1.71) fewer cardiovascular deaths between the two periods.
And lastly, Sanders et al. (2020a) examined whether policy actions
(i.e., the first annual PM2.5 NAAQS implementation rule in
2005 for the 1997 annual PM2.5 standard with a 3-year annual
average of 15 [mu]g/m\3\) reduced PM2.5 concentrations and
mortality rates in Medicare beneficiaries between 2000-2013. They
report evidence of changes in associations with mortality (a decreased
mortality rate of ~0.5 per 1,000 in attainment and non-attainment
areas) due to changes in annual PM2.5 concentrations in both
attainment and non-attainment areas. Additionally, attainment areas had
starting concentrations below 12.0 [micro]g/m\3\ prior to
implementation of the annual PM2.5 NAAQS in 2005. In
addition, following implementation of the annual PM2.5
NAAQS, annual PM2.5 concentrations decreased by 1.59 [mu]g/
m\3\ (95% CI: 1.39, 1.80) which corresponded to a reduction in
mortality rates among individuals 65 years and older (0.93% [95% CI:
0.10%, 1.77%]) in non-attainment counties relative to attainment
counties. In a life expectancy study, Bennett et al. (2019) reports
increases in life expectancy in all but 14 counties (1325 of 1339
counties) that have exhibited reductions in PM2.5
concentrations from 1999 to 2015. These studies provide support for
improvements in public health following the implementation of policies,
including in areas with PM2.5 concentrations below the level
of the current annual standard, as well as increases in life expectancy
in areas with reductions in PM2.5 concentrations.
d. Uncertainties in the Health Effects Evidence
The 2022 PA recognizes that there are a number of uncertainties and
limitations associated with the available health effects evidence.
Although the epidemiologic studies clearly demonstrate associations
between long- and short-term PM2.5 exposures and health
outcomes, several uncertainties and limitations in the health effects
evidence remain. Epidemiologic studies evaluating short-term
PM2.5 exposure and health effects have reported
heterogeneity in associations between cities and geographic regions
within the U.S. Heterogeneity in the associations observed across
epidemiologic studies may be due in part to exposure error related to
measurement-related issues, the use of central fixed-site monitors to
represent population exposure to PM2.5, and a limited
understanding of factors including exposure error related to
measurement-related issues, variability in PM2.5 composition
regionally, and factors that result in differential exposures (e.g.,
topography, the built environment, housing characteristics, personal
activity patterns). Heterogeneity is expected when the methods or the
underlying distribution of covariates vary across studies (U.S. EPA,
2019a, p. 6-221). Studies assessed in the 2019 ISA and ISA Supplement
have advanced the state of exposure science by presenting innovative
methodologies to estimate PM exposure, detailing new and existing
measurement and modeling methods, and further informing our
understanding of the influence of exposure measurement error due to
exposure estimation methods on the associations between
PM2.5 and health effects reported in epidemiologic studies
(U.S. EPA, 2019a, section 1.2.2; U.S. EPA, 2022a). Data from
PM2.5 monitors continue to be commonly used in health
studies as a surrogate for PM2.5 exposure, and often provide
a reasonable representation of exposures throughout a study area (U.S.
EPA, 2019a, section 3.4.2.2; U.S. EPA, 2022a, section 3.2.2.2.2).
However, an increasing number of studies employ hybrid modeling methods
to estimate PM2.5 exposure using data from several sources,
often including satellites and models, in addition to ground-based
monitors. These hybrid models typically have good cross-validation,
especially for PM2.5, and have the potential to reduce
exposure measurement error and uncertainty in the health effect
estimates from epidemiologic models of long-term exposure (U.S. EPA,
2019a, section 3.5; U.S. EPA, 2022a, section 2.3.3).
While studies using hybrid modeling methods have reduced exposure
measurement error and uncertainty in the health effect estimates, these
studies use a variety of approaches to estimate PM2.5
concentrations and to assign exposure to assess the association between
health outcomes and PM2.5 exposure. This variability in
methodology has inherent limitations and uncertainties, as described in
more detail in section 2.3.3.1.5 of the 2022 PA, and the performance of
the modeling approaches depends on the availability of monitoring data
which varies by location. Factors that likely contribute to poorer
model performance often coincide with relatively low ambient
PM2.5 concentrations, in areas where predicted exposures are
at a greater distance to monitors, and under conditions where the
reliability and availability of key datasets (e.g., air quality
modeling) are limited. Thus, uncertainty in hybrid model predictions
becomes an increasingly important consideration as lower predicted
concentrations are considered.
Regardless of whether a study uses monitoring data or a hybrid
modeling approach when estimating PM2.5 exposures, one key
limitation that persists is associated with the interpretation of the
study-reported mean PM2.5 concentrations and how they
compare to design values, the metric that describes the air quality
status of a given area relative to the NAAQS.\86\ As discussed above in
section II.B.3.b, the overall mean PM2.5 concentrations
reported by key epidemiologic studies reflect averaging of short- or
long-term PM2.5 exposure estimates across location (i.e.,
across multiple monitors or across modeled grid cells) and over time
(i.e., over several years). For monitor-based studies, the comparison
is somewhat more straightforward than for studies that use hybrid
modeling methods, as the monitors used to estimate exposure in the
epidemiologic studies are generally the same monitors that are used to
calculate design values for a given area. It is expected that areas
meeting a PM2.5 standard with a particular level would be
expected to have average PM2.5 concentrations (i.e.,
averaged across space and over time in the area) somewhat below that
standard level., but the difference between the maximum annual design
value and
[[Page 16249]]
average concentration in an area can be smaller or larger than analyses
presented above in section I.D.5.a, likely depending on factors such as
the number of monitors, monitor siting characteristics, and the
distribution of ambient PM2.5 concentrations. For studies
that use hybrid modeling methods to estimate PM2.5
concentrations, the comparison between study-reported mean
PM2.5 concentrations and design values is more complicated
given the variability in the modeling methods, temporal scales (i.e.,
daily versus annual), and spatial scales (i.e., nationwide versus
urban) across studies. Analyses above in section I.D.5.b and detailed
more in the 2022 PA (U.S. EPA, 2022b, section 2.3.3.2.4) present a
comparison between two hybrid modeling surfaces, which explored the
impact of these factors on the resulting mean PM2.5
concentrations and provided additional information about the
relationship between mean concentrations from studies using hybrid
modeling methods and design values. However, the results of those
analyses only reflect two surfaces and two types of approaches, so
uncertainty remains in understanding the relationship between estimated
modeled PM2.5 concentrations and design values more broadly
across hybrid modeling studies. Moreover, this analysis was completed
using two hybrid modeling methods that estimate PM2.5
concentrations in the U.S., thus an additional uncertainty includes
understanding the relationship between modeled PM2.5
concentrations and design values reported in Canada.
---------------------------------------------------------------------------
\86\ For the annual PM2.5 standard, design values are
calculated as the annual arithmetic mean PM2.5
concentration, averaged over 3 years. For the 24-hour standard,
design values are calculated as the 98th percentile of the annual
distribution of 24-hour PM2.5 concentrations, averaged
over three years (Appendix N of 40 CFR part 50).
---------------------------------------------------------------------------
In addition, where PM2.5 and other pollutants (e.g.,
ozone, nitrogen dioxide, and carbon monoxide) are correlated, it can be
difficult to distinguish whether attenuation of effects in some studies
results from copollutant confounding or collinearity with other
pollutants in the ambient mixture (U.S. EPA, 2019a, section 1.5.1; U.S.
EPA, 2022a, section 2.2.1). Studies evaluated in the 2019 ISA and ISA
Supplement further examined the potential confounding effects of both
gaseous and particulate copollutants on the relationship between long-
and short-term PM2.5 exposure and health effects. As noted
in the Appendix to the 2019 ISA (U.S. EPA, 2019a, Table A-1),
copollutant models are not without their limitations, such as instances
for which correlations are high between pollutants resulting in greater
copollutant confounding bias in results. However, the studies continue
to provide evidence indicating that associations with PM2.5
are relatively unchanged in copollutants models (U.S. EPA, 2019a,
section 1.5.1; U.S. EPA, 2022a, section 2.2.1).
Another area of uncertainty is associated with other potential
confounders, beyond copollutants. Some studies have expanded the
examination of potential confounders to not only include copollutants,
but also systematic evaluations of the potential impact of inadequate
control from long-term temporal trends and weather (U.S. EPA, 2019a,
section 11.1.5.1). Analyses examining these covariates further confirm
that the relationship between PM2.5 exposure and mortality
is unlikely to be biased by these factors. Other studies have explored
the use of alternative methods for confounder control to more
extensively account for confounders and are more robust to model
misspecification that can further inform the causality determination
for long-term and short-term PM2.5 and mortality and
cardiovascular effects (U.S. EPA, 2019a, section 11.2.2.4; U.S. EPA,
2022a, sections 3.1.1.3, 3.1.2.3, 3.2.1.2, and 3.2.2.3). These studies
indicate that bias from unmeasured confounders can occur in either
direction, although controlling for these confounders did not result in
the elimination of the association, but instead provided additional
support for associations between long-term PM2.5 exposure
and mortality when accounting for additional confounders (U.S. EPA,
2022a, section 3.2.2.2.6).
Another important limitation associated with the evidence is that,
while epidemiologic studies indicate associations between
PM2.5 and health effects, the currently available evidence
does not identify particular PM2.5 concentrations that do
not elicit health effects. Rather, health effects can occur over the
entire distribution of ambient PM2.5 concentrations
evaluated, and epidemiologic studies conducted to date do not identify
a population-level threshold below which it can be concluded with
confidence that PM2.5-related effects do not occur.
Overall, evidence assessed in the 2019 ISA and ISA Supplement
continues to indicate a linear, no-threshold C-R relationship for
PM2.5 concentrations >8 [mu]g/m\3\. However, uncertainties
remain about the shape of the C-R curve at PM2.5
concentrations <8 [mu]g/m\3\, with some recent studies providing
evidence for either a sublinear, linear, or supralinear relationship at
these lower concentrations (U.S. EPA, 2019a, section 11.2.4; U.S. EPA,
2022a, section 2.2.3.2).
There are also a number of uncertainties and limitations associated
with the experimental evidence (i.e., controlled human exposure studies
and animal toxicological studies). With respect to controlled human
exposure studies, the PA recognizes that these studies include a small
number of individuals compared to epidemiologic studies. Additionally,
these studies tend to include generally healthy adult individuals, who
are at a lower risk of experiencing health effects. These studies,
therefore, often do not include populations that are at increased risk
of PM2.5-related health effects, including children, older
adults, or individuals with pre-existing conditions. As such, these
studies are somewhat limited in their ability to inform at what
concentrations effects may be elicited in at-risk populations. With
respect to animal toxicological studies, while these studies often
examine more severe health outcomes and longer exposure durations and
higher exposure concentrations than controlled human exposure studies,
there is uncertainty in extrapolating the effects seen in animals, and
the PM2.5 exposures and doses that cause those effects, to
human populations.
Consideration of health effects are informed by the epidemiologic,
controlled human exposure, and animal toxicological studies. The
evaluation and integration of the scientific evidence in the ISA
focuses on evaluating the findings from the body of evidence across
disciplines, including evaluating the strengths and weaknesses in the
overall collection of studies across disciplines. Integrating evidence
across disciplines can strengthen causal inference, such that a weak
inference from one line of evidence can be addressed by other lines of
evidence, and coherence of these lines of evidence can add support to a
cause-effect interpretation of the association. Evaluation and
integration of the evidence also includes consideration of
uncertainties that are inherent in the scientific findings (U.S. EPA,
2015, pp. 13-15), some of which are described above.
3. Summary of Exposure and Risk Estimates
Beyond the consideration of the scientific evidence, discussed
above in section II.B, the EPA also considers the extent to which new
or updated quantitative analyses of PM2.5 air quality,
exposure, or health risks could inform conclusions on the adequacy of
the public health protection provided by the current primary
PM2.5 standards. Additionally, the 2022 PA includes an at-
risk analysis that assesses PM2.5-attributable risk
associated with PM2.5 air quality that has been adjusted to
simulate air quality scenarios of policy
[[Page 16250]]
interest (e.g., ``just meeting'' the current or potential alternative
standards). Drawing on the summary in section II.C of the proposal, the
sections below provide a brief overview of key aspects of the
assessment design (II.A.3.a), key limitations and uncertainties
(II.A.3.b), and exposure/risk estimates (II.A.3.c).
a. Key Design Aspects
Risk assessments combine data from multiple sources and involve
various assumptions and uncertainties. Input data for these analyses
includes C-R functions from epidemiologic studies for each health
outcome and ambient annual or 24-hour PM2.5 concentrations
for the study areas utilized in the risk assessment (U.S. EPA, 2022b,
section 3.4.1). Additionally, quantitative and qualitative methods were
used to characterize variability and uncertainty in the risk estimates
(U.S. EPA, 2022b, section 3.4.1.7).
Concentration-response functions used in the risk assessment are
from large, multicity U.S. epidemiologic studies that evaluate the
relationship between PM2.5 exposures and mortality.
Epidemiologic studies and concentration-response studies that were used
in the risk assessment to estimate risk were identified using criteria
that take into account factors such as study design, geographic
coverage, demographic populations, and health endpoints (U.S. EPA,
2022b, section 3.4.1.1).\87\ The risk assessment focuses on all-cause
or nonaccidental mortality associated with long-term and short-term
PM2.5 exposures, for which the 2019 ISA concluded that the
evidence provides support for a ``causal relationship'' (U.S. EPA,
2022b, section 3.4.1.2).\88\
---------------------------------------------------------------------------
\87\ Additional detail regarding the selection of epidemiologic
studies and specification of C-R functions is provided in the 2022
PA (U.S. EPA, 2022b, Appendix C, section C.1.1).
\88\ While the 2019 ISA also found that evidence supports the
determination of a ``causal relationship'' between long- and short-
term PM2.5 exposures and cardiovascular effects,
cardiovascular mortality was not included as a health outcome as it
will be captured in the estimates of all-cause mortality.
---------------------------------------------------------------------------
As described in more detail in the 2022 PA, the risk assessment
first estimated health risks associated with air quality for 2015
adjusted to simulate ``just meeting'' the current primary
PM2.5 standards (i.e., the annual standard with its level of
12.0 [micro]g/m\3\ and the 24-hour standard with its level of 35
[micro]g/m\3\). Air quality modeling was then used to simulate air
quality just meeting an alternative standard with a level of 10.0
[micro]g/m\3\ (annual) and 30 [micro]g/m\3\ (24-hour). In addition to
the model-based approach, for the subset of 30 areas controlled by the
annual standard linear interpolation and extrapolation were employed to
simulate just meeting alternative annual standards with levels of 11.0
(interpolated between 12.0 and 10.0 [mu]g/m\3\), 9.0 [mu]g/m\3\, and
8.0 [mu]g/m\3\ (both extrapolated from 12.0 and 10.0 [mu]g/m\3\) (U.S.
EPA, 2022b, section 3.4.1.3). The 2022 PA notes that there is greater
uncertainty regarding whether a revised 24-hour standard (i.e., with a
lower level) is needed to further limit ``peak'' PM2.5
concentration exposure and whether a lower 24-hour standard level would
most effectively reduce PM2.5-associated health risks
associated with ``typical'' daily exposures. The risk assessment
estimates health risks associated with air quality adjusted to meet a
revised 24-hour standard with a level of 30 [micro]g/m\3\, in
conjunction with estimating the health risks associated with meeting a
revised annual standard with a level of 10.0 [micro]g/m\3\ (U.S. EPA,
2022b, section 3.4.1.3). More details on the air quality adjustment
approaches used in the risk assessment are described in section 3.4.1.4
and Appendix C of the 2022 PA (U.S. EPA, 2022b).
When selecting U.S. study areas for inclusion in the risk
assessment, the available ambient monitors, geographic diversity, and
ambient PM2.5 air quality concentrations were taken into
consideration (U.S. EPA, 2022b, section 3.4.1.4). When these factors
were applied, 47 urban study areas were identified, which include
nearly 60 million people aged 30-99, or approximately 30% of the U.S
population in this age range (U.S. EPA, 2022b, section 3.4.1.5,
Appendix C, section C.1.3). Of the 47 study areas, there were 30 study
areas where just meeting the current standards is controlled by the
annual standard,\89\ 11 study areas where just meeting the current
standards is controlled by the daily standard,\90\ and 6 study areas
where the controlling standard differed depending on the air quality
adjustment approach (U.S. EPA, 2022b, section 3.4.1.5).\91\
---------------------------------------------------------------------------
\89\ For these areas, the annual standard is the ``controlling
standard'' because when air quality is adjusted to simulate just
meeting the current or potential alternative annual standards, that
air quality also would meet the 24-hour standard being evaluated.
\90\ For these areas, the 24-hour standard is the controlling
standard because when air quality is adjusted to simulate just
meeting the current or potential alternative 24-hour standards, that
air quality also would meet the annual standard being evaluated.
Some areas classified as being controlled by the 24-hour standard
also violate the annual standard.
\91\ In these 6 areas, the controlling standard depended on the
air quality adjustment method used and/or the standard scenarios
evaluated.
---------------------------------------------------------------------------
In addition to the overall risk assessment, the 2022 PA also
includes an at-risk analysis and estimates exposures and health risks
of specific populations identified as at-risk that would be allowed
under the current and potential alternative standards to further inform
the Administrator's conclusions regarding the adequacy of the public
health protection provided by the current primary PM2.5
standards. In so doing, the 2022 PA evaluates exposure and
PM2.5 mortality risk for older adults (e.g., 65 years and
older), stratified for White, Black, Asian, Native American, Non-
Hispanic, and Hispanic individuals residing in the same study areas
included in the overall risk assessment. This analysis utilizes a
recent epidemiologic study that provides race- and ethnicity-specific
risk coefficients (Di et al., 2017b).
b. Key Limitations and Uncertainties
Uncertainty in risk estimates (e.g., in the size of risk estimates)
can result from a number of factors, including the assumptions about
the shape of the C-R function with mortality at low ambient PM
concentrations, the potential for confounding and/or exposure
measurement error in the underlying epidemiologic studies, and the
methods used to adjust PM2.5 air quality. More specifically,
the use of air quality modeling to adjust PM2.5
concentrations are limited as they rely on model predictions, are based
on emission changes scaled by fixed percentages, and use only two of
the full set of possible emission scenarios and linear interpolation/
extrapolation to adjust air quality that may not fully capture
potential non-linearities associated with real-world changes in air
quality. Additionally, the selection of case study areas is limited to
urban areas predominantly located CA and in the Eastern U.S. that are
controlled by the annual standard. While the risk assessment does not
report quantitative uncertainty in the risk estimates as exposure
concentrations are reduced, it does provide information on the
distribution of concentrations associated with the risk estimates when
evaluating progressively lower alternative annual standards. Based on
these data, as lower alternative annual standards are evaluated, larger
proportions of the distributions in risk occur at or below 10 [mu]g/
m\3\ (at concentrations below or near most of the study-reported means
from the key U.S. epidemiologic studies) and at or below 8 [mu]g/m\3\
(the concentration at which the ISA reports increasing uncertainty in
the shape of the C-R curve based on the body of epidemiologic
evidence).
[[Page 16251]]
Similarly, the at-risk analysis is also subject to many of these
same uncertainties noted above. Additionally, the at-risk analysis
included C-R functions from only one study (Di et al., 2017b), which
reported associations between long-term PM2.5 exposures and
mortality, stratified by race/ethnicity, in populations age 65 and
older, as opposed to the multiple studies used in the overall risk
assessment to convey risk estimate variability. These and other sources
of uncertainty in the overall risk assessment and the at-risk analyses
are characterized in more depth in the 2022 PA (U.S. EPA, 2022b,
section 3.4.1.7, section 3.4.1.8, Appendix C, section C.3).
c. Summary of Risk Estimates
Although limitations in the underlying data and approaches lead to
some uncertainty regarding estimates of PM2.5-associated
risk, the risk assessment estimates that the current primary
PM2.5 standards could allow a substantial number of
PM2.5-associated deaths in the U.S. For example, when air
quality in the 47 study areas is adjusted to simulate just meeting the
current standards, the risk assessment estimates up to 45,100 deaths in
2015 are attributable to long-term PM2.5 exposures
associated with just meeting the current annual and 24-hour
PM2.5 standards (U.S. EPA, 2022b, section 3.4.2.1).
Additionally, as described in more detail in the 2022 PA, the at-risk
analysis suggests that a lower annual standard level (i.e., below 12
[micro]g/m\3\ and down as low as 8 [micro]g/m\3\) will help to reduce
PM2.5 exposure and may also help to mitigate exposure and
risk disparities in populations identified as particularly at-risk for
adverse effects from PM exposures (i.e., minority populations).
Compared to the current annual standard, meeting a revised annual
standard with a lower level is estimated to reduce PM2.5-
associated health risks in the 30 study areas controlled by the annual
standard by about 7-9% for a level of 11.0 [micro]g/m\3\, 15-19% for a
level of 10.0 [micro]g/m\3\, 22-28% for a level of 9.0 [micro]g/m\3\,
and 30-37% for a level of 8.0 [micro]g/m\3\) (U.S. EPA, 2022b, Table 3-
17). Meeting a revised annual standard with a lower level may also help
to mitigate exposure and risk disparities in populations identified as
particularly at-risk for adverse effects from PM exposures (i.e.,
minority populations) in simulated scenarios just meeting alternative
annual standards. However, though reduced, disparities by race and
ethnicity persist even at an alternative annual standard level of 8
[micro]g/m\3\, the lowest alternative annual standard included in the
risk assessment (U.S. EPA, 2022b, section 3.4.2.4).
Revising the level of the 24-hour standard to 30 [mu]g/m\3\ is
estimated to lower PM2.5-associated risks across a more
limited population and number of areas than revising the annual
standard (U.S. EPA, 2022, section 3.4.2.4). Risk reduction predictions
are largely confined to areas located in the western U.S., several of
which are also likely to experience risk reductions upon meeting a
revised annual standard. In the 11 areas controlled by the 24-hour
standard, when air quality is simulated to just meet the current 24-
hour standard, PM2.5 exposures are estimated to be
associated with as many as 2,570 deaths annual. Compared to just
meeting the current standard, air quality just meeting an alternative
24-hour standard level of 30 [micro]g/m\3\ is associated with
reductions in estimated risk of 9-13% (U.S. EPA, 2022b, section
3.4.2.3).
B. Conclusions on the Primary PM2.5 Standards
In drawing conclusions on the adequacy of the current primary
PM2.5 standards, in view of the advances in scientific
knowledge and additional information now available, the Administrator
has considered the evidence base, information, and policy judgments
that were the foundation of the 2012 and 2020 reviews and reflects upon
the body of evidence and information newly available in this
reconsideration. In so doing, the Administrator has taken into account
both evidence-based and risk-based considerations, as well as advice
from the CASAC and public comments. Evidence-based considerations draw
upon the EPA's integrated assessment of the scientific evidence of
health effects related to PM2.5 exposure presented in the
2019 ISA and ISA Supplement (summarized in the proposal in sections
II.B (88 FR 5580, January 27, 2023) and II.D.2.a (88 FR 5609, January
27, 2023), and also in section II.A.2 above) to address key policy-
relevant questions in the reconsideration. Similarly, the risk-based
considerations draw upon the assessment of population exposure and risk
(summarized in the proposal in sections II.C (88 FR 5605, January 27,
2023) and II.D.2.b (88 FR 5614, January 27, 2023), and also in section
II.A.3 above) in addressing policy-relevant questions focused on the
potential for PM2.5 exposures associated with mortality
under air quality conditions just meeting the current and potential
alternative standards.
The approach to reviewing the primary standards is consistent with
requirements of the provisions of the CAA related to the review of the
NAAQS and with how the EPA and the courts have historically interpreted
the CAA. As discussed in section I.A above, these provisions require
the Administrator to establish primary standards that, in the
Administrator's judgment, are requisite (i.e., neither more nor less
stringent than necessary) to protect public health with an adequate
margin of safety. Consistent with the Agency's approach across all
NAAQS reviews, the EPA's approach to informing these judgments is based
on a recognition that the available health effects evidence generally
reflects a continuum that includes ambient air exposures for which
scientists generally agree that health effects are likely to occur
through lower levels at which the likelihood and magnitude of response
become increasingly uncertain. The CAA does not require the
Administrator to establish a primary standard at a zero-risk level or
at background concentration levels, but rather at a level that reduces
risk sufficiently so as to protect public health, including the health
of sensitive groups, with an adequate margin of safety.
The decisions on the adequacy of the current primary
PM2.5 standards described below is a public health policy
judgment by the Administrator that draws on the scientific evidence for
health effects, quantitative analyses of population exposures and/or
health risks, and judgments about how to consider the uncertainties and
limitations that are inherent in the scientific evidence and
quantitative analyses. The four basic elements of the NAAQS (i.e.,
indicator, averaging time, form, and level) have been considered
collectively in evaluating the public health protection afforded by the
current standards.
Section II.B.2 below briefly summarizes the basis for the
Administrator's proposed decision, drawing from section II.D.3 of the
proposal (88 FR 5617, January 27, 2023). The advice and recommendations
of the CASAC and public comments on the proposed decision are addressed
below in sections II.B.1 and II.B.3, respectively. The Administrator's
final conclusions in this reconsideration regarding the adequacy of the
current primary PM2.5 standards and whether any revisions
are appropriate are described in section II.B.4.
1. CASAC Advice
As part of its review of the 2019 draft PA, the CASAC provided
advice on the adequacy of the public health protection afforded by the
current primary PM2.5 standards. Its advice is documented in
[[Page 16252]]
a letter sent to the EPA Administrator (Cox, 2019b). In this letter,
the committee recommended retaining the current 24-hour
PM2.5 standard but did not reach consensus on whether the
scientific and technical information support retaining or revising the
current annual standard. In particular, though the CASAC agreed that
there is a long-standing body of health evidence supporting
relationships between PM2.5 exposures and various health
outcomes, including mortality and serious morbidity effects, individual
CASAC members ``differ[ed] in their assessments of the causal and
policy significance of these associations'' (Cox, 2019b, p. 8 of
consensus responses). Drawing from this evidence, ``some CASAC
members'' expressed support for retaining the current annual standard
while ``other members'' expressed support for revising that standard in
order to increase public health protection (Cox, 2019b, p.1 of letter).
These views are summarized below.
The CASAC members who supported retaining the current annual
standard expressed the view that substantial uncertainty remains in the
evidence for associations between PM2.5 exposures and
mortality or serious morbidity effects. These committee members
asserted that ``such associations can reasonably be explained in light
of uncontrolled confounding and other potential sources of error and
bias'' (Cox, 2019b, p. 8 of consensus responses). They noted that
associations do not necessarily reflect causal effects, and they
contended that recent epidemiologic studies assessed in the 2019 ISA
that report positive associations at lower estimated exposure
concentrations mainly confirm what was anticipated or already assumed
in setting the 2012 NAAQS. In particular, they concluded that such
studies have some of the same limitations as prior studies and do not
provide new information calling into question the existing standard.
They further asserted that ``accountability studies provide potentially
crucial information about whether and how much decreasing
PM2.5 causes decreases in future health effects'' (Cox,
2019b, p. 10 of consensus responses), and they cited recent reviews
(i.e., Henneman et al., 2017; Burns et al., 2019) to support their
position that in such studies, ``reductions of PM2.5
concentrations have not clearly reduced mortality risks'' (Cox, 2019b,
p. 8 of consensus responses). Thus, the committee members who supported
retaining the current annual standard advise that, ``while the data on
associations should certainly be carefully considered, this data should
not be interpreted more strongly than warranted based on its
methodological limitations'' (Cox, 2019b, p. 8 of consensus responses).
These members of the CASAC further concluded that the quantitative
risk assessment included in the 2019 draft PA does not provide a valid
basis for revising the current standards. This conclusion was based on
concerns that (1) ``the risk assessment treats regression coefficients
as causal coefficients with no justification or validation provided for
this decision;'' (2) the estimated regression concentration-response
functions ``have not been adequately adjusted to correct for
confounding, errors in exposure estimates and other covariates, model
uncertainty, and heterogeneity in individual biological (causal)
[concentration-response] functions;'' (3) the estimated concentration-
response functions ``do not contain quantitative uncertainty bands that
reflect model uncertainty or effects of exposure and covariate
estimation errors;'' and (4) ``no regression diagnostics are provided
justifying the use of proportional hazards . . . and other modeling
assumptions'' (Cox, 2019b, p. 9 of consensus responses). These
committee members also contended that details regarding the derivation
of concentration-response functions, including specification of the
beta values and functional forms, were not well-documented, hampering
the ability of readers to evaluate these design details. Thus, these
members ``think that the risk characterization does not provide useful
information about whether the current standard is protective'' (Cox,
2019b, p. 11 of consensus responses).
Drawing from their evaluation of the evidence and the risk
assessment in the 2019 draft PA, these committee members concluded that
``the Draft PM PA does not establish that new scientific evidence and
data reasonably call into question the public health protection
afforded by the . . . 2012 PM2.5 annual standard'' (Cox,
2019b, p.1 of letter).
In contrast, ``[o]ther members of CASAC conclude[d] that the weight
of the evidence, particularly reflecting recent epidemiology studies
showing positive associations between PM2.5 and health
effects at estimated annual average PM2.5 concentrations
below the current standard, does reasonably call into question the
adequacy of the 2012 annual PM2.5 [standard] to protect
public health with an adequate margin of safety'' (Cox, 2019b, p.1 of
letter). The committee members who supported this conclusion noted that
the body of health evidence for PM2.5 not only includes the
repeated demonstration of associations in epidemiologic studies, but
also includes support for biological plausibility established by
controlled human exposure and animal toxicology studies. They pointed
to recent studies demonstrating that the associations between
PM2.5 and health effects occur in a diversity of locations,
in different time periods, with different populations, and using
different exposure estimation and statistical methods. They concluded
that ``the entire body of evidence for PM health effects justifies the
causality determinations made in the Draft PM ISA'' (Cox, 2019b, p. 8
of consensus responses).
The members of the CASAC who supported revising the current annual
standard particularly emphasized recent findings of associations with
PM2.5 in areas with average long-term PM2.5
concentrations below the level of the annual standard and studies that
show positive associations even when estimated exposures above 12
[mu]g/m\3\ are excluded from analyses. They found it ``highly
unlikely'' that the extensive body of evidence indicating positive
associations at low estimated exposures could be fully explained by
confounding or by other non-causal explanations (Cox, 2019b, p. 8 of
consensus responses). They additionally concluded that ``the risk
characterization does provide a useful attempt to understand the
potential impacts of alternate standards on public health risks'' (Cox,
2019b, p. 11 of consensus responses). These CASAC members concluded
that the available evidence reasonably calls into question the
protection provided by the current primary PM2.5 standards
and supports revising the annual standard to increase that protection
(Cox, 2019b).
As a part of this reconsideration, the CASAC reviewed the 2021
draft PA (developed to support the reconsideration as described in
section I.C.5 above). As a part of their review of the 2021 draft PA,
the CASAC provided advice on the adequacy of the current primary
PM2.5 standards. The range of views summarized here
generally reflects differing judgments as to the relative weight to
place on various types of evidence, the risk-based information, and the
associated uncertainties, as well as differing judgments about the
importance of various PM2.5-related health effects from a
public health perspective.
In its comments on the 2021 draft PA, the CASAC stated that:
``[o]verall the CASAC finds the Draft PA to be well-
[[Page 16253]]
written and appropriate for helping to `bridge the gap' between the
agency's scientific assessments and quantitative technical analyses,
and the judgments required of the Administrator in determining whether
it is appropriate to retain or revise the National Ambient Air Quality
Standards (NAAQS)'' (Sheppard, 2022a, p. 1 of consensus letter). The
CASAC also stated that the ``[d]raft PA adequately captures and
appropriately characterizes the key aspects of the evidence assessed
and integrated in the 2019 ISA and Draft ISA Supplement of
PM2.5-related health effects'' (Sheppard, 2022b, p. 2 of
consensus letter). The CASAC also stated that ``[t]he interpretation of
the risk assessment for the purpose of evaluating the adequacy of the
current primary PM2.5 annual standard is appropriate given
the scientific findings presented'' (Sheppard, 2022a, p. 2 of consensus
letter).
With regard to the adequacy of the current primary annual
PM2.5 standard, ``all CASAC members agree that the current
level of the annual standard is not sufficiently protective of public
health and should be lowered'' (Sheppard, 2022a, p. 2 of consensus
letter). Additionally, ``the CASAC reached consensus that the
indicator, form, and averaging time should be retained, without
revision'' (Sheppard, 2022a, p. 2 of consensus letter). With regard to
the level of the primary annual PM2.5 standard, the CASAC
had differing recommendations for the appropriate range for an
alternative level. The majority of the CASAC ``judge[d] that an annual
average in the range of 8-10 [mu]g/m\3\'' was most appropriate, while
the minority of the CASAC members stated that ``the range of the
alternative standard of 10-11 [mu]g/m\3\ is more appropriate''
(Sheppard, 2022a, p. 16 of consensus responses). The CASAC did
highlight, however, that ``the alternative standard level of 10 [mu]g/
m\3\ is within the range of acceptable alternative standards
recommended by all CASAC members, and that an annual standard below 12
[mu]g/m\3\ is supported by a larger and coherent body of evidence''
(Sheppard, 2022a, p. 16 of consensus responses).
In reaching conclusions on a recommended range of 8-10 [mu]g/m\3\
for the primary annual PM2.5 standard, the majority of the
CASAC placed weight on various aspects of the available scientific
evidence and quantitative risk assessment information discussed in the
2021 draft PA (Sheppard, 2022a, p. 16 of consensus responses). In
particular, these members cited recent U.S.- and Canadian-based
epidemiologic studies that show positive associations between
PM2.5 exposure and mortality with study-reported mean
concentrations below 10 [mu]g/m\3\. Further, these members also noted
that the lower portions of the air quality distribution (i.e.,
concentrations below the mean) provide additional information to
support associations between health effects and PM2.5
concentrations lower than the reported long-term mean concentration. In
addition, the CASAC members recognized that the available evidence has
not identified a threshold concentration, below which an association no
longer remains, pointing to the conclusion in the draft ISA Supplement
that the ``evidence remains clear and consistent in supporting a no-
threshold relationship, and in supporting a linear relationship for
PM2.5 concentrations >8 [mu]g/m\3\'' (Sheppard, 2022a, p. 16
of consensus responses). Finally, these CASAC members placed weight on
the at-risk analysis as providing support for protection of at-risk
demographic groups, including minority populations.
In recommending a range of 10-11 [mu]g/m\3\ for the primary annual
PM2.5 standard, the minority of the CASAC emphasized that
there were few key epidemiologic studies that reported positive and
statistically significant health effects associations for
PM2.5 air quality distributions with overall mean
concentrations below 9.6 [mu]g/m\3\ (Sheppard, 2022a, p. 17 of
consensus responses). In so doing, the minority of the CASAC
specifically noted the variability in the relationship between study-
reported means and area annual design values based on the methods
utilized in the studies, noting that design values are generally higher
than area average exposure levels. Further, the minority of the CASAC
stated that ``uncertainties related to copollutants and confounders
make it difficult to justify a recommendation below 10-11 [mu]g/m\3\''
(Sheppard, 2022a, p. 17 of consensus responses). Finally, the minority
of the CASAC placed less weight on the risk assessment results, noting
large uncertainties, including the approaches used for adjusting air
quality to simulate just meeting the current and alternative standards.
With regard to the current primary 24-hour PM2.5
standard, in their review of the 2021 draft PA, the CASAC did not reach
consensus regarding the adequacy of the public health protection
provided by the current standard. As described further below, the
majority of the CASAC members concluded ``that the available evidence
calls into question the adequacy of the current 24-hour standard''
(Sheppard, 2022a, p. 3 of consensus letter), while the minority of the
CASAC members agreed with ``the EPA's preliminary conclusion [in the
draft PA] to retain the current 24-hour PM2.5 standard
without revision'' (Sheppard, 2022a, p. 4 of consensus letter). The
CASAC recommended that in future reviews, the EPA should also consider
alternative forms for the primary 24-hour PM2.5 standard.
Specifically, the CASAC ``suggests considering a rolling 24-hour
average and examining alternatives to the 98th percentile of the 3-year
average,'' pointing to concerns that computing 24-hour average
PM2.5 concentrations using the current midnight-to-midnight
timeframe could potentially underestimate the effects of high 24-hour
exposures, especially in areas with wood-burning stoves and wintertime
stagnation (Sheppard, 2022a, p. 18 of consensus responses).
As noted above, the majority of the CASAC favored revising the
level of the primary 24-hour PM2.5 standard, suggesting that
a range of 25-30 [mu]g/m\3\ would be adequately protective. In so
doing, the majority of the CASAC placed weight on the available
epidemiologic evidence, including epidemiologic studies that restricted
analyses to 24-hour PM2.5 concentrations below 25 [mu]g/
m\3\. These members also placed weight on results of controlled human
exposure studies with exposures close to the current standard, which
they note provide support for the epidemiologic evidence to lower the
standard. These members noted the limitations in using controlled human
exposure studies alone in considering the adequacy of the 24-hour
standard, recognizing that controlled human exposure studies
preferentially recruit less susceptible individuals and have a typical
exposure duration shorter than 24 hours. These members also placed
``greater weight on the scientific evidence than on the values
estimated by the risk assessment,'' citing their concerns that the risk
assessment ``may not adequately capture areas with wintertime
stagnation and residential wood-burning where the annual standard is
less likely to be protective'' (Sheppard, 2022a, p. 17 of consensus
responses). Furthermore, these CASAC members ``also are less confident
that the annual standard could adequately protect against health
effects of short-term exposures'' (Sheppard, 2022a, p. 17 of consensus
responses).
The minority of the CASAC agreed with the EPA's preliminary
conclusion in the 2021 draft PA to retain the current primary 24-hour
PM2.5 standard. In so doing, the minority of the CASAC
placed greater weight on the risk assessment, noting that the risk
[[Page 16254]]
assessment accounts for both the level and the form of the current
standard and the manner by which attainment with the standard is
determined. Further, the minority of the CASAC stated that the ``risk
assessment indicates that the annual standard is the controlling
standard across most of the urban study areas evaluated and revising
the level of the 24-hour standard is estimated to have minimal impact
on the PM2.5-associated risks'' and therefore, ``the annual
standard can be used to limit both long- and short-term
PM2.5 concentrations'' (Sheppard, 2022a, p. 18 of consensus
responses). Further, the minority of the CASAC placed more weight on
the controlled human exposure studies, which show ``effects at
PM2.5 concentrations well above those typically measured in
areas meeting the current standards'' and which suggest that ``the
current standards are providing adequate protection against these
exposures'' (Sheppard, 2022a, p. 18 of consensus responses).
While the CASAC members expressed differing opinions on the
appropriate revisions to the current standards, they did ``find that
both primary standards, 24-hour and annual, are critical to protect
public health given the evidence on detrimental health outcomes at both
short-term and long-term exposures including peak events'' (Sheppard,
2022a, p. 13 of consensus responses). The comments from the CASAC also
took note of uncertainties that remain in this reconsideration of the
primary PM2.5 standards and they identified a number of
additional areas for future research and data gathering and
dissemination that would inform future reviews of the primary
PM2.5 NAAQS (Sheppard, 2022a, pp. 14-15 of consensus
responses).
2. Basis for the Proposed Decision
In reaching his proposed decisions to revise the level of the
primary annual PM2.5 standard from its current level of 12.0
[micro]g/m\3\ to within the range of 9.0 to 10.0 [micro]g/m\3\, and to
retain the current primary 24-hour PM2.5 standard (88 FR
5558, January 27, 2023), the Administrator carefully considered the
assessment of the current evidence and conclusions reached in the 2019
ISA and ISA Supplement; the currently available exposure and risk
information, including associated limitations and uncertainties,
described in detail in the 2022 PA; the considerations and staff
conclusions and associated rationales presented in the 2022 PA; the
advice and recommendations from the CASAC; and public comments that had
been offered up to that point (88 FR 5558, January 27, 2023).
In reaching his proposed conclusions on whether the currently
available scientific evidence and quantitative risk-based information
support or call into question the adequacy of the public health
protection afforded by the current primary PM2.5 standards,
and as is the case with NAAQS reviews in general, the extent to which
the current primary PM2.5 standards are judged to be
adequate will depend on a variety of factors, including science policy
and public health policy judgments to be made by the Administrator on
the strength and uncertainties of the scientific evidence. The factors
relevant to judging the adequacy of the standards also include the
interpretation of, and decisions as to the weight to place on,
different aspects of the results of the risk assessment for the study
areas included and the associated uncertainties. Thus, in reaching
proposed conclusions of the current standards, the Administrator
recognized that such a determination depends in part on judgments
regarding aspects of the evidence and risk estimates, and judgments
about the degree of protection that is requisite to protect public
health with an adequate margin of safety.
The Administrator's full rationale for his proposed conclusions is
presented in section II.D.3 of proposal (88 FR 5658, January 27, 2023),
but is also briefly summarized here. In reaching the proposed decision
to revise the annual standard level to 9-10 [micro]g/m\3\, the
Administrator placed weight on the full body of scientific information.
He noted that the 2019 ISA finds that exposure to PM2.5
causes mortality and cardiovascular effects and is likely to cause
respiratory effects, cancer, and nervous system effects as detailed
further in section II.B.1 of the proposal. As detailed further in
section II.B.4 of the proposal, he additionally noted that the 2019 ISA
identifies at-risk populations at greater risk of health effects from
exposure to PM2.5, including children, older adults, people
with pre-existing respiratory or cardiovascular disease, minority
populations, and low socioeconomic status (SES) populations.
The Administrator also recognized that epidemiologic studies
provide the strongest scientific evidence when evaluating the adequacy
of the level of the annual standard. He noted that there is no specific
point in the air quality distribution of any epidemiologic study that
represents a `bright line' at and above which effects have been
observed and below which effects have not been observed. In his
proposed decision, he noted previous decision-making frameworks, which
placed weight on values at or near the study-reported mean
PM2.5 concentrations, which is where the most confidence in
the reported association of the epidemiologic study exists. He further
noted that there are a number of epidemiologic studies available in
this reconsideration that use new PM2.5 exposure estimation
techniques (e.g., hybrid modeling) that were not used in epidemiologic
studies that were available in previous reviews. These recent
epidemiologic studies that use new exposure estimation techniques
report long-term mean PM2.5 concentrations that are well
below corresponding design values, which is an important consideration
in reaching decisions on the level of the annual PM2.5
standard.
In reaching his proposed decision, the Administrator noted that a
level of 9-10 [micro]g/m\3\ would near or below the reported 25th
percentiles in key U.S. based epidemiologic studies, while also
recognizing that he has less confidence in the magnitude and
significance of the association at even lower percentiles (e.g., 10th
percentile), where even fewer health events are observed. The
Administrator also noted that a proposed level of 9-10 [micro]g/m\3\
would be near the mean PM2.5 reported in Canadian based
studies, though he also recognized that there are a number of factors
associated with the studies in Canada (e.g., exposure environments)
that make it more difficult to compare mean concnetrations from
Canadian studies to design values, which determine compliance with the
standard in the U.S.
The Administrator took note of additional pieces of scientific
evidence, which were not available in previous reviews, including
restricted analyses, which support that the association seen in
epidemiologic studies does not just occur from the peaks of the
exposure distribution. Additionally, he notes that a level of 9-10
[micro]g/m\3\ would be below the starting concentration in newly
available accountability studies, though he did note that it is more
difficult to interpret these studies in the context of selecting the
level of the annual PM2.5 standard.
Further, the Administrator took into consideration the advice of
the CASAC, noting that all members included 10 [micro]g/m\3\ in their
recommended range, and that the proposed range of 9-10 [micro]g/m\3\
for the level of the primary annual PM2.5 standard was
within the range recommended by the majority of the CASAC.
In reaching the proposed conclusion of a range between 9-10
[micro]g/m\3\, the Administrator noted that a level as high
[[Page 16255]]
as 11 [micro]g/m\3\ might not provide an adequate margin of safety,
given that 11 [micro]g/m\3\ was well above many of the epidemiologic
study-reported mean PM2.5 concentrations. Additionally, the
Administrator noted the uncertainties associated with the scientific
and quantitative information supporting a level as low as 8 [micro]g/
m\3\, which call into question the potential public health improvements
of a standard below 9 [micro]g/m\3\. The Administrator specifically
noted the lack of key U.S. studies with mean concentrations below 9.3
[micro]g/m\3\ and he further noted that the risk assessment suggests
that the risk remaining under a standard of 8 [micro]g/m\3\ would occur
at very low concentrations (e.g., mainly 7 [micro]g/m\3\ and below).
As such, the Administrator's proposed decision noted that the
current PM2.5 annual standard did not adequately provide
requisite protection against exposures to PM2.5 and that a
proposed range of 9-10 [micro]g/m\3\ would provide an adequate margin
of safety.
In his proposed decision to retain the current primary 24-hour
PM2.5 standard with a level of 35 [micro]g/m\3\, the
Administrator first considered the scientific information related to
short-term exposures to PM2.5 and health effects. He noted
that the controlled human exposure studies are the strongest line of
evidence for informing his conclusions regarding the adequacy of the
current 24-hour standard. In so doing, the Administrator recognized
that controlled human exposure studies are conducted with healthy adult
volunteers and that these studies do not include individuals who may be
at increased risk of PM2.5-related health effects (i.e.,
children, older adults, people with pre-existing diseases). He also
noted that the effects observed in the controlled human exposure
studies (e.g., changes in vascular function) are not effects that are
judged to be clearly adverse. He recognized the most consistent
evidence of effects in these studies occurs at higher concentrations
(e.g., >120 [micro]g/m\3\) following 1-5 hour exposures, and that one
study observed effects at concentrations as low as 38 [micro]g/m\3\
following 4-hour exposures. However, the Administrator reiterated that
these studies do not tell us at exactly what concentrations an adverse
effect might occur, especially for at-risk populations. As noted above
in section II.A.2.c, controlled human exposure studies tend to include
generally healthy adult individuals who are at a lower risk of
experiencing health effects, and often do not include at-risk
populations (e.g., children, older adults, or individuals with pre-
existing conditions). As such, the Administrator recognized that these
studies are somewhat limited in their ability to inform at what
concentrations effects may be elicited in in at-risk populations. The
Administrator also considered air quality analyses in the 2022 PA that
demonstrate that there will be very few, if any, days with
PM2.5 concentrations at levels evaluated in controlled human
exposure studies that are associated with effects in areas that meet
the current primary 24-hour PM2.5 standard.
The Administrator also noted that as, in previous PM NAAQS reviews,
the protection provided by the suite of standards (e.g., annual and 24-
hour standards) is evaluated together. He noted that the annual
standard is the controlling standard in most areas of the country. He
also considered air quality analyses in the 2022 PA that suggest that
revision of the annual standard to a level between 9-10 [micro]g/m\3\
would also control 24-hour PM2.5 concentrations in most
areas to, or below, 30 [micro]g/m\3\. Finally, the Administrator noted
the agreement with the advice from the minority of CASAC and
additionally noted the limited rationale and evidence provided by the
majority CASAC's recommendation to support revision of the 24-hour
standard. As such, the Administrator proposed to retain the current 24-
hour standard with its level of 35 [micro]g/m\3\.
Additionally, the Administrator proposed to conclude that it is
appropriate to retain all other elements (i.e., indicator, averaging
time, and form) of the annual and 24-hour standards.
3. Comments on the Proposed Decision
With respect to the adequacy of the primary annual PM2.5
standard, a number of commenters, primarily those from industry and
industry groups, non-governmental organizations, and some State and
local governments, disagree with the EPA's proposed decision to revise
the level of the primary annual PM2.5 standard. These
commenters generally expressed the view that the current standards
provide the requisite degree of public health protection and should be
retained, consistent with the 2020 final decision. In supporting their
view, these commenters assert that the scientific evidence available in
this reconsideration is essentially unchanged since the 2020 final
decision and that the additional scientific evidence and quantitative
risk information available for the reconsideration does not support
strengthening the primary annual PM2.5 standard. These
commenters also assert that uncertainties associated with the available
scientific evidence have not changed since the 2020 final decision, and
they note that these uncertainties were essential factors in the then-
Administrator's decision to retain the primary annual PM2.5
standard. These commenters argue that, while the current Administrator
acknowledges these uncertainties, he does not place enough weight on
them in reaching his conclusions regarding the current standard. The
commenters specifically highlight uncertainties related to exposure
misclassification, confounding, and other sources of potential bias,
which they claim supports retaining the current level of the annual
standard. These commenters also note that these uncertainties were
emphasized by the minority of the CASAC in their review of the 2021
draft PA, and the commenters further suggest that the lack of consensus
from the CASAC on the appropriate level for the primary annual
PM2.5 standard show that the research is unclear. The
commenters contend that there is not support in this reconsideration
for deviating from the then-Administrator's decision in 2020.
In contrast, other commenters, primarily from public health and
environmental organizations, some State and local elected
representatives, and some State and local government agencies agree
with the EPA's proposed decision that the primary annual
PM2.5 standard is not adequate. These commenters support
revising the level of the primary annual PM2.5 standard and
emphasize that the available scientific evidence, in particular
epidemiologic studies, along with the CASAC's advice in their review
for the 2021 draft PA, provide strong support for the proposed
decision. In particular, these commenters agree with the EPA's
conclusions about the strength of the scientific evidence, including
uncertainties, and they emphasize that the CASAC reached consensus in
their review of the 2021 draft PA that the current primary annual
PM2.5 standard is not adequate. Some of these commenters
also note that a revised primary annual PM2.5 standard would
result in significant public health benefits by reducing morbidity and
mortality associated with PM2.5 exposure, especially for at-
risk populations.
The EPA agrees with commenters that the primary annual
PM2.5 standard is not adequate. The EPA recognizes the
longstanding body of health evidence supporting relationships between
PM2.5 exposures (short- and long-term) and both mortality
and serious morbidity effects. The evidence available in this
reconsideration (i.e., the studies
[[Page 16256]]
assessed in the 2019 ISA and ISA Supplement summarized above in section
II.A.2.a) reaffirms, and in some cases strengthens, the conclusions
from the 2009 ISA regarding the health effects of PM2.5
exposures. As noted above, epidemiologic studies demonstrate generally
positive and often statistically significant associations between
PM2.5 exposures and health effects. Such studies report
associations between estimated PM2.5 exposures and non-
accidental, cardiovascular, or respiratory mortality; cardiovascular or
respiratory hospitalizations or emergency room visits; and other
mortality/morbidity outcomes (e.g., lung cancer mortality or incidence,
asthma development). Recent experimental evidence, as well as evidence
from epidemiologic panel studies, strengthens support for potential
biological pathways through which PM2.5 exposures could lead
to the serious effects reported in many population-level epidemiologic
studies, including support for pathways that could lead to
cardiovascular, respiratory, nervous system, and cancer-related
effects. Moreover, these recent epidemiologic studies strengthen
support for health effect associations at PM2.5
concentrations lower than in those evaluated in epidemiologic studies
available at the time of previous reviews.
Additionally, as discussed in more detail in section I.C.5.b above,
the ISA Supplement focused on studies that were most likely to inform
decisions on the appropriate standard, but not to reassess areas that,
based on the assessment of available science published since the cutoff
date of the 2019 ISA and through 2021, were judged unlikely to have new
information that would be useful for the Administrator's decision
making. The ISA Supplement included U.S. and Canadian epidemiologic
studies for health effect categories where the 2019 ISA concluded a
causal relationship (i.e., short- and long-term PM2.5
exposure and cardiovascular effects and mortality), as well as U.S. and
Canadian epidemiologic studies that employed alternative methods for
confounder control or conducted accountability analyses (i.e., studies
that examined the effect of a policy on reducing PM2.5
concentrations). These studies, summarized in section II.A.2.a above,
examine both short- and long-term PM2.5 exposure and
cardiovascular effects and mortality. Additionally, studies that employ
alternative methods for confounder control, as described in II.A.2.a
above and in Table 3-11 and of the 2022 PA (U.S. EPA, 2022b), use a
variety of statistical methods to control for confounding bias. These
studies consistently report positive associations, which further
supports the broader body of epidemiologic evidence for both
cardiovascular effects and mortality.
In addition, there are epidemiologic studies that provide
supplemental information for consideration in reaching conclusions that
the current suite of PM2.5 standards is not adequate. These
studies include analyses that restrict annual average PM2.5
concentrations to concentrations below 12 [micro]g/m\3\ and provide
support for positive and statistically significant associations with
mortality and cardiovascular morbidity at mean PM2.5
concentrations below the current level of the primary annual
PM2.5 standard (described above in section II.A.2.c.ii and
in Table 3-10 of the 2022 PA (U.S. EPA, 2022b)). Recent accountability
studies that have starting annual PM2.5 concentrations at or
below 12 [micro]g/m\3\ suggest public health improvements may occur at
concentrations below 12 [micro]g/m\3\. These studies indicate positive
and statistically significant associations with mortality and morbidity
(e.g., cardiovascular hospital admissions) and reductions in
PM2.5 concentrations in ambient air (described above in
section II.A.2.c.ii and in Table 3-12 of the 2022 PA (U.S. EPA,
2022b)).
Thus, in considering the available scientific evidence to inform
conclusions on the adequacy of the primary PM2.5 standards,
the Administrator recognizes that the 2019 ISA and the ISA Supplement
together provides a strong scientific foundation for concluding that
the current primary PM2.5 standards are not adequate.
In addition to the scientific evidence above, the risk assessment
estimates that the current primary annual PM2.5 standard
could allow a substantial number of deaths in the U.S. Although the
Administrator recognizes that while the risk estimates can help to
place the evidence for specific health effects into a broader public
health context, they should be considered along with the inherent
uncertainties and limitations of such analyses when informing judgments
about the potential for additional public health protection associated
with PM2.5 exposures and related health effects. The
Administrator takes into consideration these uncertainties, which are
described in more detail in section II.A.3.b above, but notes that the
general magnitude of risk estimates supports the potential for
significant public health impacts, particularly for lower alternative
annual standard levels.
In the CASAC's review of the 2019 draft PA, the CASAC did not reach
consensus on whether the current annual standard is adequate, with the
majority of the CASAC recommending that the annual standard be retained
and the minority of the CASAC recommending that the standard be
revised. In their review of the 2021 draft PA, the CASAC unanimously
recommended that the current annual standard is not sufficiently
protective of public health (Sheppard, 2022a, p. 2 of consensus
letter).
The EPA disagrees with the commenters who state that the available
scientific and quantitative information available in this
reconsideration does not provide support for the current Administrator
to reach a different decision than the then-Administrator reached in
the 2020 final action. The EPA agrees with these commenters that there
are uncertainties associated with the currently available scientific
evidence. The EPA has considered these uncertainties extensively both
in reaching conclusions in the 2022 PA (U.S. EPA, 2022b, sections
3.4.3, 3.6.1, and 4.6.3) and in the proposal (88 FR 5604, 5609, January
27, 2023), and the EPA addresses more detailed public comments about
these uncertainties, including those related to copollutant
confounding, unmeasured confounding, and temporal and spatiotemporal
confounding, in the Response to Comments document. However, we disagree
with the commenters that the evidence does not provide support for the
Administrator's conclusion that the current primary annual
PM2.5 standard is not adequate to protect public health with
an adequate margin of safety, and should be revised. As described
above, epidemiologic studies in the 2019 ISA and the ISA Supplement
support and extend the evidence evaluated in the 2009 ISA, through
studies conducted in diverse populations and geographic locations,
using various statistical models and approaches to control for
potential confounders, and using a variety of exposure assessment
methodologies. Therefore, the consistent, positive associations
reported across studies (U.S. EPA, 2019a, Figures 11-1 and 11-18; U.S.
EPA, 2022a) are unlikely to be to be the result of unmeasured
confounding and other biases are unlikely to account for the consistent
positive associations observed across epidemiologic studies.
Additionally, this reconsideration includes epidemiologic studies
that were not before the then-Administrator for consideration in
reaching his final decisions at the time of the 2020 decision and that
specifically evaluate
[[Page 16257]]
confounding using alternative methods for confounder control). These
recent epidemiologic studies provide support for the current
Administrator's conclusion that the suite of primary PM2.5
standards are not adequate. While confounding was an uncertainty noted
by the then-Administrator in the 2020 decision, he recognized ``that
methodological study designs to address confounding, such as causal
inference methods, are an emerging field of study'' (85 FR 82710,
December 18, 2020). The ISA Supplement considered studies that employed
statistical approaches that attempt to more extensively account for
confounders and are more robust to model misspecification (i.e., used
alternative methods for confounder control),\92\ given that such
studies were highlighted by the CASAC in their review of the 2019 draft
PA and identified in public comments on the 2020 proposal. Since the
literature cutoff date for the 2019 ISA, multiple studies that employ
alternative methods for confounder control have become available for
consideration in the ISA Supplement and, subsequently, in this
reconsideration. For example, one study before the Administrator in
this reconsideration that was not available in the 2019 ISA is Schwartz
et al. (2021), which used a causal modeling approach focused on
exposure changes and controls for measured confounders by design in
order to evaluate the association between long-term PM2.5
exposure and mortality in the Medicare population. The study authors
found significant associations of PM2.5 with increased
mortality rates using a causal modeling approach robust to omitted
confounding. The results of this study and other studies in the ISA
Supplement that employ alternative methods to control for confounders
lend support to the robustness of positive associations between
PM2.5 exposure and multiple morbidity and mortality
endpoints exhibited across epidemiologic studies, and also indicate
that unmeasured confounding and other biases are unlikely to account
for the consistent positive associations observed across epidemiologic
studies (U.S. EPA, 2022b, sections 3.1.1.3, 3.1.2.3, 3.2.1.3, and
3.2.2.3).
---------------------------------------------------------------------------
\92\ As noted in the ISA Supplement: ``In the peer-reviewed
literature, these epidemiologic studies are often referred to as
causal inference studies or studies that used causal modeling
methods. For the purposes of this Supplement, this terminology is
not used to prevent confusion with the main scientific conclusions
(i.e., the causality determinations) presented within an ISA. In
addition, as is consistent with the weight-of-evidence framework
used within ISAs and discussed in the Preamble to the Integrated
Science Assessments, an individual study on its own cannot inform
causality, but instead represents a piece of the overall body of
evidence'' (U.S. EPA, 2022a, p. 1-3).
---------------------------------------------------------------------------
Further, the EPA disagrees with the commenters who argue that the
Administrator did not appropriately consider the strengths and
limitations of the health evidence in reaching his decision to revise
the current primary annual PM2.5 standard in this
reconsideration. In reaching his proposed decision, the Administrator
considered the entire body of evidence and how to appropriately weigh
the uncertainties associated with the health evidence (88 FR 5617,
January 27, 2023). Such an approach is consistent with setting
standards that are neither more nor less stringent than necessary,
recognizing that ``Congress provided that the Administrator is to use
his judgment in setting air quality standards precisely to permit him
to act in the face of uncertainty,'' the Administrator must set
standards on ``the frontiers of scientific and medical knowledge'' and
``Congress directed the Administrator to err on the side of caution in
making the necessary decisions.'' Lead Indus. Ass'n, Inc. v. EPA, 647
F.2d 1130, 1155 & n.50 (D.C. Cir. 1980) (quoting H.R. Rep. No. 95-294,
at 50). As such, a determination of identifying a specific level at
which the standard should be set necessarily requires the
Administrator's judgement (e.g., weighing the uncertainties and margin
of safety).
Additionally, the EPA disagrees with the commenters that contend
that there is no basis in this reconsideration for deviating from the
previous Administrator's decision in 2020. It is well-established that
in CAA section 109 Congress specifically left the determination of the
requisite NAAQS to the judgment of the Administrator and, moreover,
that ``decisions about the appropriate NAAQS level must `necessarily .
. . rest largely on policy judgments.' '' Mississippi v. EPA, 744 F.3d
1344, 1357 (D.C. Cir. 2013) (quoting Lead Industries Ass'n v. EPA, 647
F.2d 1130, 1147 (D.C. Cir. 1980)). As the Court of Appeals for the D.C.
Circuit has noted, ``Every time EPA reviews a NAAQS, it (presumably)
does so against contemporary policy judgments and the existing corpus
of scientific knowledge.'' Id., at 1343.
In this reconsideration, both the existing corpus of scientific
knowledge as well as the Administrator's policy judgments about how to
interpret and weigh that evidence to protect public health with an
adequate margin of safety have changed. The expansion of the air
quality criteria to encompass additional studies, information and
analyses in the ISA Supplement and 2022 PA, as well as the additional
consideration of the scientific record by the CASAC and the public
provided the Administrator with significant additional information on
which to base his decision.\93\ In addition, in this reconsideration,
the Administrator is reaching different judgments about how to weigh
the epidemiologic evidence, including the uncertainties in the
scientific evidence, and how to ensure an adequate margin of safety to
protect against uncertain harms, compared to the approach in the 2020
final decision. For example, as discussed in greater detail above in
section II.A.1 and in the 2020 notice of final rulemaking (85 FR 82717,
December 18, 2020), in considering the epidemiologic evidence as part
of his decision to retain the current primary annual PM2.5
standard in the 2020 decision, the then-Administrator placed weight on
the mean of the study-reported means (or medians) (i.e., 13.5 [micro]g/
m\3\) from key U.S. epidemiologic studies that are monitor-based being
above the level of the current primary annual PM2.5 standard
of 12.0 [micro]g/m\3\. By contrast, in this reconsideration, the
current Administrator has taken an approach more similar to how the EPA
has considered study-reported mean PM2.5 concentrations
relative to the level of the primary annual PM2.5 standard
in other recent PM NAAQS reviews. In so doing, in reaching his decision
to revise the level of the primary annual PM2.5 standard to
9.0 [micro]g/m\3\, he is using an approach that places weight on
selecting a level for the standard that is below the study-reported
mean PM2.5 concentrations reported in key U.S. epidemiologic
studies, including recent epidemiologic studies that use hybrid model-
based methods, as well as being near or below the 25th percentile
PM2.5 concentrations in those key U.S. epidemiologic studies
that report these concentrations.
---------------------------------------------------------------------------
\93\ The EPA notes that, in considering the additional
scientific evidence available in this reconsideration, one member of
the CASAC who reviewed both the 2019 draft PA and the 2021 draft PA
found that the available scientific and quantitative information
available in this reconsideration supported revising the level of
the primary annual PM2.5 standard, whereas he recommended
retaining the standard during the review of the 2019 draft PA.
---------------------------------------------------------------------------
As such and further detailed in section II.B.4 below, in
considering the adequacy of the current primary PM standards in this
reconsideration, the Administrator has carefully considered the: (1)
Policy-relevant evidence and conclusions contained in the 2019 ISA and
2022 ISA Supplement; (2) the quantitative information presented and
[[Page 16258]]
assessed in the 2022 PA; (3) the evaluation of this evidence, the
quantitative information, and the rationale and conclusions presented
in the 2022 PA; (4) the advice and recommendations from the CASAC; and
(5) public comments. The Administrator concludes that the current suite
of primary PM2.5 standards are not adequate to protect
public health with an adequate margin of safety.
The four basic elements of the NAAQS (indicator, averaging time,
form, and level) are considered collectively in evaluating the health
protection afforded by a standard. The EPA received relatively few
comments on the averaging time and form for the primary
PM2.5 standards, but those who did provide comments on these
elements were primarily from public health and environmental
organizations, State and local elected representatives, and State and
local government agencies. Some commenters assert that the current 24-
hour averaging time for the primary 24-hour PM2.5 standard
does not adequately protect against short-term peaks. These commenters
further state that the 24-hour averaging time protects against chronic
exposures but does not adequately protect against serious acute risks
from certain sources such as prescribed burning. Also, a few commenters
explicitly recommend that a subdaily averaging time would be more
appropriate, although none of the commenters recommended a specific
averaging time for consideration. Additionally, some commenters cite to
the CASAC's advice in their review of the 2021 draft PA that future
reviews of the PM NAAQS should include evaluation of alternative forms
and averaging times of the current primary 24-hour PM2.5
standard.
The EPA disagrees with commenters that the current primary 24-hour
PM2.5 standard, with its 24-hour averaging time, does not
adequately protect against short-term peaks and disagrees that that
there is sufficient information to conclude that a subdaily averaging
time would be more appropriate than a 24-hour averaging time. The EPA
has reviewed the currently available scientific evidence and finds that
it does not indicate that alternative averaging times would be more
appropriate for the primary PM2.5 standards. Accordingly,
the EPA concludes that it is appropriate to retain both the annual and
24-hour averaging times for standards meant to protect against long-
and short-term PM2.5.
As noted in the proposal, the 2019 ISA and ISA Supplement found
that the scientific evidence continues to provide strong support for
health effect associations with both long-term (e.g., annual or multi-
year) and short-term (e.g., mostly 24-hour) exposures to
PM2.5. Epidemiologic studies continue to provide strong
support for health effects associated with short-term PM2.5
exposures based on 24-hour PM2.5 averaging periods, and we
note that subdaily effect estimates are less consistent and, in some
cases, smaller in magnitude (88 FR 5618, January 27, 2023). Controlled
human exposure and panel-based studies of subdaily exposures typically
examine subclinical effects rather than the more serious population-
level effects that have been reported to be associated with 24-hour
exposures (e.g., mortality, hospitalizations). Collectively, the 2019
ISA concludes that epidemiologic studies do not indicate that subdaily
averaging periods are more closely associated with health effects than
the 24-hour average exposure metric (U.S. EPA, 2019a, section 1.5.2.1).
Additionally, the EPA notes that while recent controlled human exposure
studies provide consistent evidence for cardiovascular effects
following PM2.5 exposures for less than 24 hours (i.e., <30
minutes to 5 hours), exposure concentrations in these studies are well-
above the ambient concentrations typically measured in locations
meeting the current standards (U.S. EPA, 2022a, section 3.3.3.1).
Therefore, this information does not indicate that a revision to the
averaging time is needed to provide additional protection against
subdaily PM2.5 exposures, beyond that provided by the
current primary standards. This conclusion is also supported by the
advice given to EPA by the CASAC in their review of the 2021 draft PA,
which reached consensus that averaging times for the standards should
be retained, without revision (Sheppard, 2022a, p. 2 of consensus
letter).\94\ For all of these reasons, the Administrator concludes that
the currently available evidence does not support considering
alternatives to the annual and 24-hour averaging times for standards
meant to protect against long- and short-term PM2.5
exposures.
---------------------------------------------------------------------------
\94\ In providing advice on the 2019 draft PA, the CASAC did not
weigh in specifically on the averaging time of the primary 24-hour
PM2.5 standard but did recommend that the standard be
retained because the available evidence does not call into question
its adequacy (Cox, 2019b, p. 3 of consensus letter).
---------------------------------------------------------------------------
Multiple commenters, primarily from public health and environmental
organizations, recommend revising the form of the primary 24-hour
PM2.5 standard to a 99th percentile to provide increased
public health protection against peak PM2.5 exposures,
particularly for at-risk populations. These commenters express concern
that the current 98th percentile form allows 7 exceedances per year and
contend that a 99th percentile form that would allow half that number
is more appropriate. Commenters also cite to the CASAC's advice in
their review of the 2021 draft PA, which recommended that the EPA
consider alternative percentiles for the form of the primary 24-hour
PM2.5 standard in the future.
The EPA disagrees that the current 98th percentile form does not
provide the requisite public health protection against peak
PM2.5 exposures and concludes that the 98th percentile,
averaged over three years, remains appropriate for the primary 24-hour
PM2.5 standard. As noted in previous reviews and in the
proposal, the EPA has set both an annual standard and a 24-hour
standard to provide protection from health effects associated with both
long- and short-term exposures to PM2.5 (62 FR 38667, July
18, 1997; 88 FR 5620, January 27, 2023). With respect to the form of
the 24-hour standard, as described just above, the epidemiologic
studies continue to provide strong support for health effect
associations with short-term (e.g., mostly 24-hour) PM2.5
exposures and controlled human exposure studies provide evidence for
health effects following single short-term ``peak'' PM2.5
exposures (88 FR 5619, January 27, 2023). Both the 98th and the 99th
percentile form provide a very high degree of control of peak
concentrations. As the commenters point out, a 99th percentile would
reduce the number of allowable exceedances to four days per year. The
EPA anticipates, however, that such a revision to the form would make
the attainment status of an area more subject to change from
unpredictable nonanthropogenic factors, such as meteorological events.
The EPA has often noted that frequent shifts in attainment status that
are unrelated to long-term air quality trends is inconsistent with
providing a stable target for air quality planning and risk management
programs, which in turn provides for the most effective public health
protection in the long run (78 FR 3127, January 15, 2013; 80 FR 65351,
October 26, 2015). Thus, the EPA's interest in an appropriate degree of
stability is to ensure that the State air quality programs are
effective in controlling pollution and that the public health
protections of the standard are achieved. As discussed above, while
recent controlled human exposure studies provide consistent evidence
for cardiovascular effects following PM2.5
[[Page 16259]]
exposures for less than 24 hours (i.e., < 30 minutes to 5 hours),
exposure concentrations in these studies are well-above the ambient
concentrations typically measured in locations meeting the current
standards (U.S. EPA, 2022a, section 3.3.3.1), and the 98th percentile
form is very effective at limiting occurrences of exposures of concern.
Taking into consideration the available scientific information and
quantitative information, the EPA therefore concludes that the 98th
percentile form provides an appropriate balance between limiting the
occurrence of peak 24-hour PM2.5 concentrations and
identifying a stable target for risk management programs. This
conclusion is also supported by the advice given to the EPA by the
CASAC in their review of the 2021 draft PA, where they reached
consensus that the form for the standards should be retained, without
revision (Sheppard, 2022a, p. 2 of consensus letter).\95\
---------------------------------------------------------------------------
\95\ The CASAC did not provide advice or recommendations
regarding the forms of the primary PM2.5 standards in
their review of the 2019 draft PA (Cox, 2019b).
---------------------------------------------------------------------------
Additionally, the EPA recognizes the CASAC's advice in their review
of the 2021 draft PA, where they recommended ``that in future reviews,
the EPA provide a more comprehensive assessment of the 24-hour standard
that includes the form as well as the level'' (Sheppard, 2022a, p. 4 of
consensus letter). This advice is reflected in the proposal by the EPA,
which noted ``that it would be appropriate to gather additional air
quality and scientific information and further consider these issues in
future reviews'' (88 FR 5619, January 27, 2023). The EPA will consider
the information provided by the commenters regarding the form of the
24-hour PM2.5 standard in the next review of the PM NAAQS.
A number of commenters who support revising the level of the
primary annual PM2.5 standard, particularly those who
support a revised level of 8 [micro]g/m\3\, disagree with how the EPA
has emphasized the mean PM2.5 concentrations reported in key
epidemiologic studies to inform conclusions on the level of the primary
PM2.5 standard. These commenters argue that, in this
reconsideration, the EPA is arbitrarily emphasizing uncertainties in
key epidemiologic studies in the focus on mean concentrations. Many of
these commenters recommend that the EPA consider the full distribution
of PM2.5 concentrations from the key epidemiologic studies
in reaching conclusions on the appropriate level for the primary annual
PM2.5 standards, in particular concentrations below the
mean, such as the 25th percentile. In supporting this view, commenters
point to the CASAC's advice in their review of the 2021 draft PA, where
the majority of the CASAC stated that the ``use of the mean to define
where the data provide the most evidence is conservative since robust
data clearly indicate effects below the mean in concentration-response
functions'' (Sheppard, 2022a, p. 16 of consensus responses), and that
``[e]pidemiologic studies require consideration of distribution around
the mean of exposure to identify effects and thus lower levels than the
mean must be considered as part of the range where the data provide
higher confidence'' (Sheppard, 2022a, p. 13 of consensus responses).
As an initial matter, consistent with some previous approaches and
as detailed by the Administrator in reaching conclusions on the level
of the primary annual PM2.5 standard in section II.B.4
below, the EPA considers the long-term study-reported mean
PM2.5 concentrations from key epidemiologic studies and sets
the level of the standard to somewhat below the lowest long-term mean
PM2.5 concentration. Additionally, as discussed further
below, the EPA also considers the available information from a subset
of epidemiologic studies that report exposure estimates or health
events at the 25th and 10th percentiles of PM2.5
concentrations. The Administrator gives some weight to the 25th
percentile data, although he recognizes that his confidence in the
magnitude and significance in the reported concentrations, and their
ability to inform decisions on the appropriate level of the annual
standard, decreases with reduced data (below the mean) and diminishes
further at percentiles that are even further below the mean and the
25th percentile. Therefore, the Administrator places weight on the
reported 25th percentiles concentrations, rather than the reported 10th
percentile concentrations, for the subset of studies that report lower
percentile PM2.5 concentrations in reaching his conclusions
regarding the appropriate level for the primary annual PM2.5
standard.
In considering the available scientific evidence to reach decisions
on the adequacy of the suite of primary PM2.5 standards, the
EPA notes that in previous PM NAAQS reviews (including the 1997, 2006
and 2012 reviews), evidence-based approaches were used that focused on
identifying standard levels near or somewhat below long-term mean
concentrations reported in key epidemiologic studies. These approaches
were supported by the CASAC in previous reviews and were supported in
this reconsideration by the CASAC in their review of the 2021 draft
PA.\96\
---------------------------------------------------------------------------
\96\ The Administrator notes that, in their review of the 2021
draft PA, a majority of members of the CASAC noted that there are
some limitations for this approach ``for the purpose of informing
the adequacy of the standards'' (Sheppard, 2022a, p. 8 of consensus
responses) and advised that future reviews should include evaluation
of other metrics, including the distribution of concentrations
reported in epidemiologic studies and in analyses restricting
concentrations to below the current standard level. The
Administrator also notes that, in their review of the 2019 draft PA,
the CASAC lacked consensus on the inferences to be drawn from the
epidemiologic evidence, with a majority of CASAC having concerns
about confounding, error and bias and concluding that newer studies
did not provide a basis for revising the current standards, while a
minority concluded that the evidence, including more recent studies
showing associations in areas with average long-term
PM2.5 concentrations below the current annual standard,
supported their conclusion that the current standards are inadequate
(Cox, 2019b, pp. 8-9 of consensus responses).
---------------------------------------------------------------------------
In considering the available scientific evidence, the EPA notes the
strength of the epidemiologic evidence which includes multiple studies
that consistently report positive associations for short- and long-term
PM2.5 exposures and mortality and cardiovascular effects.
Some available studies also use a variety of statistical methods to
control for confounding bias and report similar associations, which
further supports the broader body of epidemiologic evidence for both
mortality and cardiovascular effects. Additionally, the EPA notes that
recent epidemiologic studies strengthen support for health effect
associations at PM2.5 concentrations lower than in those
evaluated in epidemiologic studies available at the time of previous
reviews.
While these epidemiologic studies evaluate associations between
distributions of ambient PM2.5 concentrations and health
outcomes, they do not identify the specific exposures that led to the
reported effects. As such, there is no specific point in the air
quality distribution of any epidemiologic study that represents a
``bright line'' at and above which effects have been observed and below
which effects have not been observed.
Studies of daily PM2.5 exposures examine associations
between day-to-day variation in PM2.5 concentrations and
health outcomes, often over several years. While there can be
considerable variability in daily exposures over a multi-year study
period, most of the estimated exposures reflect days with
[[Page 16260]]
ambient PM2.5 concentrations around the middle of the air
quality distributions examined (i.e., ``typical'' days rather than days
with extremely high or extremely low concentrations). Similarly, for
studies of annual PM2.5 exposures, most of the health events
occur at estimated exposures that reflect annual average
PM2.5 concentrations around the middle of the air quality
distributions examined. In both cases, epidemiologic studies provide
the strongest support for reported health effect associations for this
middle portion of the PM2.5 air quality distribution, which
corresponds to the bulk of the underlying data, rather than the extreme
upper or lower ends of the distribution. Therefore, in the absence of
discernible thresholds, long-term study-reported means--that is, the
study-reported ambient PM2.5 concentrations in the
epidemiologic studies that reflect estimated exposures with a focus
around the middle portion of the PM2.5 air quality
distribution where the bulk of the observed data reside--provide the
strongest support for reported health effect associations in
epidemiologic studies.
Based on the air quality criteria for this reconsideration, as
described in the 2019 ISA, ISA Supplement, 2022 PA and the proposal,
the EPA believes it is appropriate to continue to use the mean
PM2.5 concentrations from the key epidemiologic studies to
inform conclusions regarding the appropriate level for the primary
annual PM2.5 standard.
There are a large number of key epidemiologic studies available in
this reconsideration to inform conclusions regarding the level of the
primary annual PM2.5 standard. For the key U.S.
epidemiologic studies, the study-reported mean PM2.5
concentrations range from 9.9-16.5 [mu]g/m\3\ for monitor-based studies
(Figure 1 above) and range from 9.3-12.2 [mu]g/m\3\ for hybrid
modeling-based studies (Figure 2 above).
In addition to the study-reported mean PM2.5
concentrations, the EPA agrees with the CASAC's advice in their review
of the 2021 draft PA and public comments that information on other
percentiles below the mean can also be informative, and the EPA notes
that the CASAC advised that for the purpose of informing the adequacy
of the standards, future reviews should include an evaluation of other
metrics, including the distribution of concentrations reported in
epidemiologic studies (Sheppard, 2022a, p. 9 of consensus responses).
As such, in reaching conclusions in this reconsideration, the EPA takes
note of the additional study-reported PM2.5 concentrations
below the means (e.g., 25th and 10th percentiles) that are available
from a limited subset of key U.S. epidemiologic studies. As shown in
Figures 1 and 2 above, six key U.S. epidemiologic studies report
information on other percentiles (e.g., 10th and 25th percentiles of
PM2.5 concentrations or 10th and 25th percentiles of
PM2.5 concentrations associated with health events) that are
below the mean.\97\ Three of the studies are monitor-based and three
are hybrid model-based.
---------------------------------------------------------------------------
\97\ The Wang et al. (2017) study only reports the 25th
percentile of the estimated PM2.5 concentrations, not the
10th percentile.
---------------------------------------------------------------------------
The key U.S. epidemiologic studies that report percentiles below
the mean that are monitor based are older studies. These studies
included smaller numbers of people than the newer hybrid model-based
studies. For the three older, monitor-based studies, because the
cohorts were smaller in size, a relatively smaller portion of the
health events were observed in the lower part of the air quality
distribution. As such, our confidence in the magnitude and significance
of the associations begins to decrease in the lower part of the air
quality distribution of those older, monitor-based studies.
The three newer, hybrid model-based studies have larger cohort
sizes than the older, monitor-based studies and, as noted by
commenters, have more health events in the lower part of the air
quality distribution. For these reasons, the EPA notes that we have
more confidence in the reported association at concentrations lower
than the reported mean in these more recent hybrid model-based studies,
particularly at the 25th percentile compared to the 10th percentile.
While the cohort sizes in the more recent, hybrid model-based studies
are larger than the older, monitor-based studies, the EPA notes that
the 10th percentiles are well below the middle portion of the air
quality distribution for which we have the greatest confidence, and as
noted above, our confidence in the magnitude and significance of
associations in the lower parts of the air quality distribution begins
to decrease. While we have more confidence in the lower percentiles
because of the larger cohort sizes in the more recent hybrid model-
based studies, we also have more confidence in the 25th percentiles
than in the 10th percentiles, which are further from the means and
closer to the lower end of the air quality distribution.
In considering how the six studies that report percentiles lower
than the mean can be used to inform conclusions regarding the level of
the primary annual PM2.5 standard, the EPA first notes that
the three monitor-based epidemiologic studies (Bell et al., 2008;
Franklin et al., 2007; Zanobetti and Schwartz, 2009) report 25th
percentile concentrations that are at or above 11.5 [micro]g/m\3\. For
two of the more recent hybrid model-based studies (Di et al., 2017b;
Wang et al., 2017), the 25th percentile of estimated PM2.5
concentrations are just above 9 [micro]g/m\3\, while one study (Di et
al., 2017a) reports a PM2.5 concentrations corresponding to
25th percentiles of health events of just below 7 [micro]g/m\3\. For
the Di et al. (2017a) study, the 25th percentile PM2.5
concentration (6.7 [micro]g/m\3\) is based on the PM2.5
concentration at which the 25th percentile of deaths occur in the
study, while the reported mean (11.6 [micro]g/m\3\) is based on
estimated PM2.5 exposure concentrations. Additionally, the
25th percentiles of the other two recently available hybrid model-based
studies (Di et al., 2017b; Wang et al., 2017) are based on estimated
PM2.5 concentrations. As such, the PM2.5
concentration at which the 25th percentile of health events occur may
be different from the estimated 25th percentile PM2.5
concentration in this study (Di et al., 2017a), creating an uncertain
basis for comparison with the studies by Di et al. (2017b) and Wang et
al. (2017). The 25th percentiles from these studies, in particular
those that are more recently available, help to inform the
Administrator's judgments regarding the appropriate level for the
primary annual PM2.5 standard.
Some commenters disagree with the EPA's consideration of the
relationship between mean PM2.5 concentrations reported in
the key epidemiologic studies and design values to inform conclusions
on the appropriate level for the primary annual PM2.5
standards. Commenters contend that setting the level of the primary
annual standard below the design values in the epidemiologic studies,
rather than below the study-reported mean concentrations, might keep
overall mean PM2.5 concentrations throughout an area below
the study-reported means but allow PM2.5 concentrations in
some parts of the area, including near the ``design value monitor'' to
remain above the study-reported mean PM2.5 concentrations,
which are the concentrations where the evidence of health effects is
strongest. Commenters contend that such a decision framework would not
result in a standard that would provide requisite protection with an
adequate margin of safety,
[[Page 16261]]
particularly for at-risk populations. These commenters further support
this view by citing the CASAC's advice in their review of the 2021
draft PA, where the majority of CASAC stated that ``even if a design
value is somewhat higher than the area average, it reflects actual
exposure levels and thus any portion of the population living near the
design value monitor does experience exposures at that level and
consequent health effects of exposure to that higher concentration''
(Sheppard, 2022a, p. 14 of consensus responses). Additionally, these
commenters suggest that the EPA should not deviate from the approach
taken in the 2012 review, which was to set the standard at a level
``somewhat below'' the lowest mean PM2.5 concentration in
the key epidemiologic studies.
To the extent that commenters are suggesting that the EPA is
setting the level of the primary annual PM2.5 standard below
the design values in the epidemiologic studies, rather than below the
study-reported mean PM2.5 concentrations, we disagree with
the commenters. In reaching conclusions on the level of the primary
annual PM2.5 standard, the EPA considers the long-term
study-reported mean PM2.5 concentrations from key
epidemiologic studies and sets the level of the standard to somewhat
below the lowest long-term mean PM2.5 concentration, not
below the design values in the epidemiologic studies. Additionally, the
EPA also considers the available information from a subset of
epidemiologic studies that report exposure estimates or health events
at the 25th and 10th percentiles of PM2.5 concentrations.
The EPA particularly considers the 25th percentile data, while
recognizing that our confidence in the magnitude and significance in
the reported concentrations, and the ability of the lower percentile
PM2.5 concentrations to inform decisions on the appropriate
level of the annual standard, decreases with reduced data (below the
mean) and diminishes further at percentiles that are even further below
the mean and the 25th percentile.
However, the EPA notes that it is important to understand, and to
not ignore, the relationship between the study-reported mean
PM2.5 concentrations reported in key epidemiologic studies
and the area design value. As an initial matter, the NAAQS consists of
all four elements of the standard (indicator, averaging time, form, and
level) and setting a standard that is requisite to protect public
health includes consideration of all four elements together. Following
implementation of the NAAQS, the design value is the metric used to
determine compliance with the standard and is the statistic that
describes the air quality status of a given location relative to the
level of the primary annual PM2.5 NAAQS. The design value is
different from the study-reported mean PM2.5 concentrations.
This is because the study-reported mean PM2.5 concentrations
are an annual average PM2.5 concentration, similar to the
level of the standard, but the epidemiologic studies do not report
statistics that take into account the other elements of the standard
(i.e., averaging time and form). Therefore, when considering the
appropriate revisions to the annual PM2.5 standard, the EPA
must consider the protection provided by a revised standard taking into
account all of the elements of the standard, not just the annual
average PM2.5 concentration alone.
In considering the annual standard, and in assessing the range of
study-reported exposure concentrations for which we have the strongest
support for adverse health effects observed in epidemiologic studies,
the EPA focuses on whether the current primary annual PM2.5
standard provides adequate protection against these exposure
concentrations or if the level of the standard should be revised to
provide the appropriate public health protection. This means that, as
in some previous reviews, it is important to consider how the study
means were computed and how these concentrations compare to the annual
standard metric (including the level, averaging time and form) which
must be met at the monitor with the highest PM2.5 design
value in an area for compliance with the NAAQS. This approach is based
on the application of a decision framework based on assessing means (as
well as the lower distribution of reported PM2.5
concentration, as noted above) reported in key epidemiologic studies.
In the 2012 review, the available key epidemiologic studies computed
the mean PM2.5 concentrations using an average across
monitor-based PM2.5 concentrations. As such, at that time,
the decision framework used an approach based on maximum monitor
concentrations to determine compliance with the standard, while
selecting the standard level based on consideration of composite
monitor concentrations (i.e., selecting the standard level of 12.0
[micro]g/m\3\ was just below the long-term study-reported mean
PM2.5 concentrations in key epidemiologic studies). Further,
the EPA conducted analyses that examined the differences in these two
metrics (i.e., maximum monitor concentrations, which is how compliance
with the standard is assessed and composite monitor concentrations,
which is how key epidemiologic studies report their mean
concentrations) across the U.S. and in areas included in the key
epidemiologic studies and found that the maximum design value in an
area was generally higher than the monitor average across that area,
with the amount of difference between the two metrics varying based on
location and concentration (Hassett-Sipple et al., 2010; Frank, 2012).
This information was taken into account by the then-Administrator's
final decision in selecting a level of 12.0 [micro]g/m\3\ for the
primary annual PM2.5 standard in the 2012 review and
discussed more specifically in her considerations on adequate margin of
safety.
The relationship between the mean PM2.5 concentrations
and the area design value continues to be an important consideration in
evaluating the adequacy of the current or potential alternative annual
standard levels in this reconsideration. Again, in a given area, the
area design value is based on the monitor in an area with the highest
PM2.5 concentrations and is used to determine compliance
with the standard, including the averaging time and form of the
standard (i.e., an annual average over 3-years must not exceed the
level of the of the annual PM2.5 standard). The highest
PM2.5 concentrations spatially distributed in the area would
generally occur at or near the area design value monitor and the
distribution of PM2.5 concentrations would generally be
lower in other locations and at monitors in that area. As such, when an
area is meeting a specific annual standard level (e.g., 9.0 [micro]g/
m\3\), we would expect the annual average exposures (i.e., a metric
similar to the study-reported mean values) in that area to be at
concentrations lower than that level (e.g., lower than 9.0 [micro]g/
m\3\).
However, as described in section II.A.2.c.ii, we note that there
are a substantial number of different types of epidemiologic studies
available since the 2012 review, as assessed in both the 2019 ISA and
the ISA Supplement, that make understanding the relationship between
the mean PM2.5 concentrations and the area design value an
even more important consideration in this reconsideration (U.S. EPA,
2019a; U.S. EPA, 2022a). While the key epidemiologic studies in the
2012 review were all monitor-based studies, the recent epidemiologic
studies in this reconsideration include hybrid modeling approaches that
have emerged in the epidemiologic literature as an
[[Page 16262]]
alternative to approaches that only use ground-based monitors to
estimate PM2.5 exposure. As assessed in the 2019 ISA and ISA
Supplement, a substantial number of epidemiologic studies used hybrid
model-based methods in evaluating associations between PM2.5
exposure and health effects. Hybrid model-based studies employ various
fusion techniques that combine ground-based monitored data with air
quality modeled estimates and/or information from satellites to
estimate PM2.5 exposures. While these studies provide a
broader estimation of PM2.5 exposures compared to monitor-
based studies (i.e., PM2.5 concentrations are estimated in
areas without monitors), the hybrid modeling approaches result in
study-reported means that are more difficult to relate to the annual
standard metric and to the maximum monitor design values used to assess
compliance. In addition, to further complicate the comparison, when
looking across these studies, we find variations in how exposure is
estimated between such studies, and thus, how the study means are
calculated. Two important variations across studies include: (1)
Variability in spatial scale used (i.e., averages computed across the
national (or large portions of the country) versus a focus on only
CBSAs); and (2) variability in exposure assignment methods (i.e.,
averaging across all grid cells, averaging across a scaled-up area like
a ZIP code, and population weighting). The differences in these
approaches can result in studies reporting different study means, even
though the association between PM2.5 exposure and health
effects outcomes are similar.
To emphasize the importance of the differences between the studies,
we revisit the simplified example in the State of Georgia from the 2022
PA that evaluates monitors and hybrid modeling approaches, noting that
this example is useful to exhibit how the differences in the methods
used to estimate exposure can lead to differences in the reported mean
concentrations (U.S. EPA, 2022b, p. 3-71). In this example, for all
monitors within the Atlanta-Sandy Springs-Roswell CBSA, the average
PM2.5 concentration is 9.3 [micro]g/m\3\, while the area
design value (based on the highest monitored PM2.5
concentration in the area) is 10.4 [micro]g/m\3\. This comparison helps
to illustrate the fact that composite monitor values tend to be
somewhat lower than the highest area monitor values, consistent with
the key points made in the 2012 review. This example also illustrates
how monitors are sited to represent the higher concentrations within
the area and that the area's annual design value, which is used for
compliance with the standard, is calculated based on the highest
monitor in the area. Next, in this example, mean PM2.5
concentrations were calculated using similar approaches to those used
in hybrid modeling-based epidemiologic studies to compute study-
reported means, including (1) the average concentration across the
entire State of Georgia; (2) the population-weighted average across the
entire State; (3) the average concentration across the Atlanta-Sandy
Springs-Roswell CBSA; and (4) the population-weighted average across
the Atlanta-Sandy Springs-Roswell CBSA. At the urban level (e.g.,
Atlanta-Sandy Springs-Roswell CBSA), the average PM2.5
concentration when taking the mean of all grid cells is 9.2 [micro]g/
m\3\, whereas the population-weighted mean is 9.6 [micro]g/m\3\. Across
Georgia, the average PM2.5 concentration using the hybrid
approach and averaged across each grid cell is 8.3 [micro]g/m\3\, which
is lower than the population-weighted statewide average of 9.1
[micro]g/m\3\. While this is a simple example completed in one State
and one CBSA, it suggests that the lowest mean values tend to result
from the approaches that use concentrations from all or most grid cells
(e.g., did not apply population weighting), both urban and rural,
across the study area to compute the mean. Higher mean values are
observed when the approach focuses on the urban areas alone or when the
approach incorporates population weighting. Overall, this example
suggests that the means from studies using hybrid modeling approaches
are generally lower than the means from monitor-based approaches, and
means from both approaches are lower than the annual design values for
the same area. Population weighting tends to increase the calculated
mean concentration, likely because more densely populated areas also
tend to have higher PM2.5 concentrations. In other words,
this simplified example exhibits how not all reported mean
PM2.5 concentrations from key epidemiologic studies are the
same; some reported means are from monitored studies and some reported
means are from hybrid modeling studies, while some reported means
include only urban areas, and other reported means include both urban
and rural areas, and some reported means include aspects of population
weighting while others do not.
As detailed above in section I.D.5, in the air quality analyses
comparing composite monitored PM2.5 concentrations with
annual PM2.5 design values in U.S. CBSAs, maximum annual
PM2.5 design values were approximately 10% to 20% higher
than annual average composite monitor concentrations (i.e., averaged
across multiple monitors in the same CBSA). Based on these results,
this analysis suggests that there will be a distribution of
concentrations and the maximum annual average monitored concentration
in an area (at the design value monitor, used for compliance with the
standard), will generally be 10-20% higher than the average across the
other monitors in the area. Thus, in considering how the annual
standard levels would relate to the study-reported means from monitor-
based studies, we can generally conclude that an annual standard level
that is no more than 10-20% higher than monitor-based study-reported
mean PM2.5 concentrations would generally maintain air
quality exposures to be below those associated with the study-reported
mean PM2.5 concentrations, exposures for which we have the
strongest support for adverse health effects occurring.
Air quality analyses described in section I.D.5 above also consider
information from the epidemiologic studies that utilized the hybrid
modeling approaches. Analyses show that average maximum annual design
values are 40-50% higher when compared to annual average
PM2.5 concentrations estimated without population weighting
and are 15-18% higher when compared to average annual PM2.5
concentrations with population weighting applied. Given these results,
it is worth noting that for the studies using the hybrid modeling
approaches, the choice of methodology employed in calculating the
study-reported means (i.e., using population weighting versus not
applying aspects of population weighting), and not a difference in
estimates of exposure in the study itself, can produce substantially
different study-reported mean values, with the approach that does not
employ population weighting producing a much lower reported mean
PM2.5 concentration. Therefore, the impact of the
differences in methods is an important consideration when comparing
mean concentrations across studies.
Because of the differences in the methods employed by the key
epidemiologic studies, and as demonstrated by the example and air
quality analyses above, the application of any decision framework that
considers the study-reported mean PM2.5 concentrations, and
evaluates whether the current annual standard provides adequate
protection against these reported exposure concentrations, is more
complicated than the
[[Page 16263]]
approaches used in past reviews. As such, the EPA disagrees with
commenters who argue that the EPA's consideration of the relationship
between mean PM2.5 concentrations reported in key
epidemiologic studies and design values is not appropriate and should
be ignored.
In considering the information from the epidemiologic studies,
while the EPA does not dispute the reported associations of
epidemiologic studies in hybrid modeling studies that report long-term
mean concentrations and do not apply aspects of population weighting,
using the reported long-term mean concentration from these studies in
informing an appropriate level of the annual PM2.5 standard
is more uncertain. Given this, hybrid modeling studies that do not
apply aspects of population weighting provide less information on
conclusions regarding the appropriate level of the primary annual
PM2.5 standard. In support of this, some commenters also
noted this consideration and suggested that the Administrator place
lower weight on U.S. studies that did not use population weighting.
In considering the relationship between study-reported mean
PM2.5 concentrations and the design values, the EPA agrees
with commenters that setting the level of the primary annual standard
below the design values, rather than below the study-reported mean
concentrations, might allow PM2.5 concentrations in some
part of the area near the design value monitor to remain above the
study-reported mean PM2.5concentration, where evidence of
health effects is strongest. As discussed in the proposal and in
section II.B.4 below, the Administrator specifically notes that that
the highest PM2.5 concentrations spatially distributed in
the area would generally occur at or near the area design value monitor
and that PM2.5 concentrations will be equal to or lower at
other monitors in the area. Furthermore, since monitoring strategies
aim to site monitors in areas with higher PM2.5
concentrations, monitored areas will generally have higher
concentrations compared to areas without monitors. Therefore, by
setting the level of the standard to 9.0 [micro]g/m\3\ and just below
the lowest study-reported mean PM2.5 concentration (e.g.,
9.3 [micro]g/m\3\), the highest possible design value in a given area
would be just below the study-reported mean PM2.5
concentration, the concentration where we have the most confidence in
the reported health effect association, and we anticipate that, based
on our assessment of air quality data, the distribution of
PM2.5 concentrations would decrease even further with
distance from the highest monitor (i.e., the ``design value monitor'')
(see, for example, U.S. EPA, 2022a, section 2.3.3.2.4 and pp. 3-71 to
3-77). The Administrator further notes that when an epidemiologic study
reports a mean PM2.5 concentration that reflects the average
of annual average monitor-based concentrations across an area, the area
design value will generally be higher than the study-reported mean.
Similarly, he observes that when a study reports a mean that reflects
the average of annual average concentrations estimated at across an
area using a hybrid modeling approach, the area design value will
generally be higher. As such, by evaluating the difference between the
study-reported mean PM2.5 concentrations and design values,
the Administrator seeks to set the level of the standard below the
lowest study-reported mean, while ensuring that the primary annual
PM2.5 standard, including its averaging time and form,
provides protection against the exposures associated with health
effects observed in the key epidemiologic studies.
Additionally, the EPA disagrees with commenters who contend that
the approach taken may allow PM2.5 near the design value
monitor to remain above the study-reported mean PM2.5
concentrations. In following this approach of setting the annual
standard level somewhat below the lowest reported mean PM2.5
concentration, setting a standard level that requires the design value
monitor (which is the highest monitor in an area) to be just below the
lowest study-reported mean across key studies will generally result in
distributions of even lower concentrations of PM2.5 across
the entire area, such that even those people living near an area design
value monitor (where PM2.5 concentrations are generally
highest) will be exposed to PM2.5 concentrations below the
PM2.5 concentrations reported in the epidemiologic studies
where there is the highest confidence of an association. In their
review of the 2021 draft PA, the majority of the CASAC had some
concerns about the approach for comparing study means and design
values, questioning whether such an approach would provide adequate
protection for people who live in areas with higher concentrations,
such as those living in areas with higher concentrations (e.g., near
the design value monitor) (Sheppard, 2022a, p. 8 of consensus
responses). The minority of the CASAC, in considering the relationship
between the study-reported mean PM2.5 concentration and
design values, stated that ``the form of the standard and the way
attainment with the standard is determined (i.e., highest design value
in the CBSA) are important factors when determining the appropriate
level for the standard'' and noted that that design values are
generally higher than area average exposure levels (Sheppard, 2022a, p.
17 of consensus responses). For all of the reasons discussed above, and
consistent with the minority of the CASAC's advice in their review of
the 2021 draft PA, we disagree with the commenters that areas near the
design value monitors would be expected to experience PM2.5
concentrations above the study-reported mean concentrations.
Several commenters assert that epidemiologic studies that restrict
PM2.5 concentration to below 12 [micro]g/m\3\ provide
additional support for revising the level of the primary annual
PM2.5 standard to 8 [micro]g/m\3\. Some commenters disagree
with the EPA's assertion that the studies that employ restricted
analyses do not provide enough information to understand how the
studies were restricted to certain PM2.5 concentrations,
with commenters providing additional information on the methods for
restricted analyses. The commenters state that for the long-term
studies at issue here, the study authors simply examined their database
that linked subjects to long-term PM2.5 concentrations above
12 [mu]g/m\3\, removed those data from the analysis, and reran the
analysis. Additionally, one commenter provided an explanation of how
the restricted analyses were conducted in studies for which he was an
author. The commenter notes that for each year a subject was in the
study, annual PM2.5 concentrations were assigned at the ZIP
code level. If they moved, they were assigned the ZIP code level
PM2.5 concentration for the new ZIP code. The commenter
notes that these restricted analyses only included subjects whose
annual PM2.5 exposure never exceeded that restricted
concentration for any year of follow-up in the study. The commenter
suggested that the EPA may be concerned as to how PM2.5
concentrations in restricted analyses related to a design value since
these are exposures for individuals who may have relocated during the
study but argue that that is not the point. The commenters assert that
while the analyses were restricted to people never exposed above
certain concentrations over longer periods of time, the actual
PM2.5 exposure was one year of exposure in most of these
studies. Commenters also suggest that, since the
[[Page 16264]]
EPA has deviated from its approach from the 2012 review for considering
study-reported mean PM2.5 concentrations, the EPA should
dismiss its concerns regarding being able to relate the mean
PM2.5 concentrations from these studies to design values.
First, the EPA agrees with commenters that studies that employ
restricted analyses can be used for informing conclusions regarding the
appropriate level of the primary annual PM2.5 standard.
However, the EPA disagrees that the information provided by the
commenters provides a sufficient basis for an annual standard level of
8 [mu]g/m\3\. Restricted analyses provide additional support for
effects at lower concentrations, exhibiting associations for mean
concentrations presumably below the mean concentrations for the main
analyses. However, even though commenters note that any individual with
exposures over the restricted analyses is excluded from restricted
analyses, uncertainties remain with regard to how the mean
PM2.5 concentrations in restricted analyses compare to
design values, particularly in light of the removal of entire ZIP codes
from analyses. Design values are calculated based on all measured
PM2.5 concentrations. When an analysis is restricted below a
certain level, some parts of the air quality distribution are removed,
but comparing the restricted mean to a design value is not possible
because these are two different metrics. For example, in a study that
restricts concentrations below 12 [micro]g/m\3\, that represents only
part of the air quality distribution, whereas a design value for that
study area would include all PM2.5 concentrations, not just
the ones below 12 [micro]g/m\3\. Therefore, in contrast to means from
the main (unrestricted) analysis, it is not possible to compare mean
concentrations from restricted analyses to design values. Further, it
is unclear how one could evaluate such a relationship between design
values and mean PM2.5 concentrations from studies that use
restricted analyses because the standard is set based on all of its
elements (indicator, averaging time, form, and level) and removing
PM2.5 concentrations from the calculation of the design
value for such a comparison would result in a metric that is no longer
a design value that would provide the intended protection of the
standard. This leads to greater uncertainty in how to use the mean
PM2.5 concentrations from these studies that use restricted
analyses in a similar decision framework as the epidemiologic studies
that report long-term mean PM2.5 concentrations for health
effect associations for the full distribution of PM2.5
concentrations.
As described in reaching his conclusions in the section below, the
Administrator judges that, despite these uncertainties and limitations,
studies that use restricted analyses can provide supplemental
information for consideration in reaching conclusions regarding both
the adequacy and level of the standard. He notes two studies (Di et
al., 2017b and Dominici et al., 2019) are available in this
reconsideration that report means in their restricted analyses
(restricting annual average PM2.5 exposure below 12 [mu]g/
m\3\) and used population-weighted approaches to estimate
PM2.5 exposures and these studies report mean
PM2.5 concentrations of 9.6 [micro]g/m\3\. He recognizes
that these studies are just one line of evidence for consideration and
that along with the broader evidence base, including the key
epidemiologic studies, these studies provide support that the level of
the primary annual PM2.5 standard should be set below 10
[micro]g/m\3\.
We disagree with the commenters that concerns about relating the
mean PM2.5 concentrations from restricted analyses to design
values are not valid. As an initial matter, restricted analyses were
not available and did not inform the 2012 decision to revise the annual
PM2.5 standard level to 12.0 [micro]g/m\3\. The approach in
2012 in revising the annual standard was to set the level to somewhat
below the mean of key epidemiologic studies. As noted above, while the
EPA believes that restricted analyses can help inform conclusions
regarding the adequacy and the level of the primary annual
PM2.5 standard, in the context of placing the studies in a
decision framework to inform the appropriate level of the annual
PM2.5 standard, the EPA has not deviated from its approach
from the 2012 review. Given that restricted analyses are new since the
2012 review, the EPA disagrees with commenters that uncertainties
associated with these studies should not be considered, and that these
studies should be used in a similar manner to their main analyses in
taking an approach to set a level of the standard somewhat below the
lowest long-term reported mean PM2.5 concentration.
Specifically, as detailed above there are uncertainties and limitations
associated with relating the mean PM2.5 concentrations from
these studies to design values for studies that use restricted
analyses, and many of these studies did not expressly report a mean
PM2.5 concentration for the restricted analysis which makes
it impossible to make such a comparison.
Several commenters contend that in considering the accountability
studies, the EPA inappropriately reached conclusions regarding the
level of the primary annual PM2.5 standard based on the
starting PM2.5 concentrations of these studies, rather than
the ending concentrations (i.e., concentrations after a policy was
implemented). The commenters assert that these studies provide support
for revising the level of the primary annual PM2.5 standard
to below the proposed range of 9-10 [micro]g/m\3\ to protect public
health with an adequate margin of safety.
Accountability studies examine the effect of a policy on reducing
PM2.5 concentrations in ambient air and evaluate whether
such reductions were observed to also lead to reductions in
PM2.5- associated health outcomes (e.g., mortality).
Additionally, accountability studies can reduce uncertainties related
to residual confounding of temporal and spatial factors (U.S. EPA,
2022a, p. 3-25). Prior to implementation of the policies, three
accountability studies newly available in this reconsideration and
assessed in the ISA Supplement, report mean PM2.5
concentrations below the level of the current annual standard level
(12.0 [micro]g/m\3\) and ranged from 10.0 [micro]g/m\3\ to 11.1
[micro]g/m\3\ (Sanders et al., 2020b; Corrigan et al., 2018; and
Henneman et al., 2019). These studies suggest that public health
improvements may occur following the implementation of a policy that
reduces annual average PM2.5 concentrations below the level
of the current standard of 12.0 [micro]g/m\3\, and potentially below
the lowest ``starting'' concentrations in these studies of 10.0
[micro]g/m\3\. However, while the small number of studies may provide
limited information related to informing the adequacy and level of the
annual PM2.5 standard, we note that accountability studies
are only one line of evidence, and that these studies provide
supplemental information for consideration in addition to the full body
of evidence. Further, the EPA does not believe it would be appropriate
to determine the level of the standard by reference to ending
concentrations in accountability studies. Accountability studies are
most informative in demonstrating that public health improvements may
occur following the implementation of a policy that reduces annual
average PM2.5 concentrations below the level of the current
standard of 12.0 [micro]g/m\3\, and potentially below the lowest
``starting'' concentrations in these studies of 10.0 [micro]g/m\3\.
However, the EPA finds the available information from accountability
studies is too limited to support a conclusion that the
[[Page 16265]]
appropriate level at which to set the primary annual PM2.5
standard would be equal to the ending concentrations of those studies,
as the commenters suggest. These studies demonstrate that there are
reductions in health outcomes when PM2.5 concentrations are
reduced in these studies from the starting concentration to the ending
concentration, but do not provide support for health effect
associations at or below the ending concentrations that would warrant a
more stringent standard.
Commenters disagree with the Administrator placing less weight on
the epidemiologic studies conducted in Canada when reaching conclusions
regarding the level of the primary annual PM2.5 standard.
These commenters argue that the Canadian epidemiologic studies provide
support for setting the level at the lowest end of the proposed range
(i.e., 8 [micro]g/m\3\) because they report mean PM2.5
concentrations, in some cases, below 8 [micro]g/m\3\. Commenters
disagree with the EPA's reasoning for placing less weight on the
Canadian epidemiologic studies, suggesting it conflicts with the
approaches in previous PM NAAQS reviews and arguing that the findings
of the Canadian epidemiologic studies can be directly translated into a
primary annual PM2.5 standard. Additionally, while the
commenters disagree with the EPA's approach for considering the study-
reported mean PM2.5 concentrations and design values in
general, they note that the CASAC, in their review of the 2021 PA,
noted that ``while there may be no design value in Canada, there are
data that indicate what a U.S. design value would be if an area average
like that found in the Canadian studies were to occur in the U.S.''
(Sheppard, 2022a, p. 13 of consensus responses). The commenters contend
that the EPA failed to acknowledge this advice from the CASAC,
specifically noting that the majority of the CASAC highlighted Canadian
epidemiologic studies as a part of their rationale for revising the
level of the primary annual PM2.5 standard to within the
range of 8-10 [micro]g/m\3\.
In considering the information from the epidemiologic studies in
reaching his conclusions, the Administrator considered the full body of
evidence, including studies conducted in the U.S. and Canada. However,
as described in the proposal and in section II.B.4 below, the
Administrator also recognizes that the exposure environments in the
U.S. are different from those in Canada. In particular, the U.S.
population density is approximately 43 people per square kilometer in
the contiguous U.S.\98\ compared to Canada, which has one of the lowest
population densities on the Earth with 4.2 people per square kilometer
(Statistics Canada, 2023). This difference in population density
between the U.S. and Canada was not as apparent, and did not need to be
highlighted, in the 2012 review given that the available Canadian
epidemiologic studies used population-weighting and focused on urban
areas where monitors were available and population densities were more
comparable with those in the U.S. Given this, the study-reported mean
concentrations from U.S. and Canadian studies in the 2012 review were
very similar. The recent epidemiologic evidence available in this
reconsideration, however, includes studies that utilize approaches that
highlight the importance of considering the differences between the two
exposure environments in the U.S. versus Canada. When focusing on the
recently available Canadian monitor-based epidemiologic studies in this
reconsideration, the information indicates that these studies, unlike
the studies available in the 2012 review, do not apply population
weighting (e.g., Lavigne et al., 2018; Liu et al., 2019). As noted in
responding to other public comments above, the absence of population
weighting is an important consideration that limits the utility of
these studies in informing the appropriate level of the primary annual
PM2.5 standard. In addition, there are recently available
studies in the 2019 ISA and ISA Supplement that expand the geographical
extent of the epidemiologic study areas by estimating exposure
concentrations in areas where there are no monitors. To do this, these
studies use either a statistical extrapolation of monitored values or
use air quality modeling and other forms of data (e.g., hybrid model-
based approaches). For these Canadian studies, the EPA notes two
important considerations in using the information to directly translate
to policy decisions regarding the level of the annual standard in the
U.S. The first is that in incorporating a larger portion of Canada into
these recent studies, more rural areas are included, and as such, the
population densities and exposure environment differences become more
important. The second is that in analyses that evaluate and validate
hybrid models, there is less certainty in PM2.5 exposure
estimates in more rural areas, which are further from air quality
monitors and where PM2.5 concentrations in the ambient air
tend to be lower (U.S. EPA, 2022b, pp. 2-51 and 2-63). Additionally, it
is unclear what portion of the PM2.5 concentrations from
rural areas are contributing to the study reported mean. Given this,
studies that incorporate more rural areas into the epidemiologic
studies highlight the importance of considering the differences between
the population exposures in the studies themselves and in the U.S.
versus Canadian study areas, as well as the influence these differences
have on the interpretation of the epidemiologic study results. For
these reasons, while the Canadian epidemiologic studies provide
additional support for associations between PM2.5
concentrations and health effects, the long-term means from Canadian
epidemiologic studies are a less certain basis for informing the EPA's
selection of the annual standard level, given that it is a U.S.-based
standard.
---------------------------------------------------------------------------
\98\ All of the key U.S. epidemiologic studies considered in
this reconsideration focus on all or subsections of the continental
U.S.
---------------------------------------------------------------------------
With respect to the CASAC's advice in their review of the 2021
draft PA, the EPA recognizes that the majority of the CASAC pointed to
the Canadian studies as supporting their recommendation to revise the
annual standard level to within the range of 8-10 [micro]g/m\3\.
However, the EPA also notes that the CASAC did not advise the EPA to
revise the annual standard to a level that was below the study-reported
means in the key Canadian epidemiologic studies. Indeed, the CASAC
noted that some of the Canadian studies showed associations below 8
[micro]g/m\3\, but did not recommend that the Administrator consider
levels below 8 [micro]g/m\3\ for the annual standard. Further, based on
the CASAC's advice, the Administrator is not excluding Canadian studies
from his consideration in this reconsideration, but he is considering
them in light of the limitations and challenges presented and in the
context of the full body of available scientific evidence.
Lastly, the EPA disagrees with commenters that the findings of the
Canadian epidemiologic studies can be directly translated into a
primary annual PM2.5 standard based on the evaluation of the
relationship between U.S. study-reported mean PM2.5
concentrations and U.S. design values. It is unclear whether the
relationship between U.S. study-reported mean PM2.5
concentrations and U.S. design values (which, in the case of U.S.
hybrid model-based studies, indicates that design values are 15-18%
greater than area mean PM2.5 concentrations) would apply to
the Canadian epidemiologic studies and their reported mean
PM2.5
[[Page 16266]]
concentrations, given that these studies generally report lower
PM2.5 concentrations than the U.S.-based studies. As such,
interpreting the study-reported mean concentrations from the Canadian
studies in the context of a U.S.-based standard may present challenges
in directly and quantitatively informing decisions regarding potential
alternative levels of the annual standard, particularly noting the
different in exposure relationships in the U.S. versus Canada given the
large difference in population densities between the two countries.
Further, as mentioned above, while the CASAC advised the EPA to
consider the Canadian studies as relevant evidence and found that
placing weight on the Canadian studies supported their recommendation
to revise the annual standard level to within the range of 8-10
[micro]g/m\3\, the lower end of their recommended range for the level
of the annual standard did not extend below the lower study-reported
means from those studies.
Commenters who supported retaining and revising the primary annual
PM2.5 standard both raised concerns regarding how the EPA
used the scientific evidence and quantitative risk assessment related
to disparities in PM2.5 exposure and risk in informing
conclusions on the standard. Commenters who supported retaining the
standard assert that the available scientific evidence that
demonstrates disparities for minority populations do not support
revising the standard, noting that these studies are in areas that tend
to have large minority populations and more sources of PM. These
commenters contend that because the studies conclude that minority
populations experience more effects than others living in the same area
that something other than PM2.5 concentrations in ambient
air is causing the disproportionate impact on minority populations,
providing proximity to a source as an example. The commenters note that
it is unclear how a national standard will reduce exposure disparities
for population groups living in the same area, and further assert that
studies of exposure disparities among minority populations were
considered in reaching the 2020 final decision to retain the standards.
Conversely, commenters who support revising the standard assert
that the at-risk analyses conducted in the 2022 PA provide support for
revising the primary annual PM2.5 standard to a level of 8
[mu]g/m\3\. In particular, these commenters state that the at-risk
analysis demonstrated that while disparities in mortality risk remain
at a standard level of 9.0 [mu]g/m\3\, disparities in exposure are
significantly reduced for an alternative standard level of 8.0 [mu]g/
m\3\ (U.S. EPA, 2022b, p. 3-162).
As discussed in section I above, the primary (health-based) NAAQS
are established at a level that is requisite to protect public health,
including the health of sensitive or at-risk groups, with an adequate
margin of safety.\99\ In so doing, decisions on the NAAQS are based on
an explicit and comprehensive assessment of the current scientific
evidence and associated risk analyses. More specifically, the EPA
expressly considers the available information regarding health effects
among at-risk populations in decisions on the primary NAAQS. Where
populations with disparities in exposure and risk are among the at-risk
populations, the decision on the standards is based on providing
requisite protection for these and other at-risk populations and
lifestages.
---------------------------------------------------------------------------
\99\ The legislative history of section 109 indicates that a
primary standard is to be set at ``the maximum permissible ambient
air level . . . which will protect the health of any [sensitive]
group of the population,'' and that for this purpose ``reference
should be made to a representative sample of persons comprising the
sensitive group rather than to a single person in such a group.'' S.
Rep. No. 91-1196, 91st Cong., 2d Sess. 10 (1970); see also, e.g.,
Am. Lung Ass'n v. EPA, 134 F.3d 388, 389 (D.C. Cir. 1998) (``If a
pollutant adversely affects the health of these sensitive
individuals, EPA must strengthen the entire national standard'').
---------------------------------------------------------------------------
The Administrator expressly considered the available information
regarding health effects among at-risk populations in reaching the
proposed decisions that the current primary annual PM2.5
standard is not requisite to protect public health with an adequate
margin of safety, and should be revised. The 2019 ISA and ISA
Supplement identified children, older adults, people with pre-existing
diseases (cardiovascular disease and respiratory disease), minority
populations, and low SES populations as at-risk populations. The
Administrator is thus, in his final decision, establishing primary
PM2.5 standards which, in his judgment, will provide
protection for these at-risk populations, including minority
populations, with an adequate margin of safety.
With respect to the risk assessment, while the EPA notes that the
analyses support the conclusion that the primary PM2.5
standards are not adequate, as detailed further in the proposal and
above in section II.A.3, the EPA also cautions against an over-
interpretation of the absolute results. The quantitative risk
assessment provides estimates of PM2.5-attributable
mortality based on input data that include C-R functions from
epidemiologic studies that do not quantitatively account for
uncertainties in associations between PM2.5 exposure and
health effects at lower concentrations and are based on an air quality
adjustment approach that incorporates proportional decreases in
PM2.5 concentrations to meet lower alternative standard
levels. As a result, simulated air quality improvements used in the
risk assessment will always lead to proportional decreases in risk
(i.e., each additional [micro]g/m\3\ reduction produces additional
benefits with no clear stopping point), without considering the
substantially greater uncertainties associated with the relationship
between PM2.5 exposures and health effects at lower
concentrations.
The same is true for the new at-risk analysis in the risk
assessment presented in the 2022 PA that is based on a recent
epidemiologic study that is available in this reconsideration that
provides mortality risk coefficients for older adults (i.e., 65 years
and older) based on PM2.5 exposure and stratified by racial
and ethnic demographics. Generally, the results of at-risk analyses can
vary greatly depending on the inputs to the analyses, including the
representativeness of the populations and demographics captured by the
study areas that are a part of the analyses, as well as the available
C-R functions from epidemiologic studies that stratify by race and
ethnicity and the air quality adjustment approaches that are used to
simulate air quality at different standard levels. In fact, for this
at-risk analysis, the results are even more uncertain than similar
estimates from the overall risk assessment due to additional sources of
uncertainty specific to the at-risk analysis, such as using C-R
functions derived from smaller epidemiologic sample sizes along with
the sources of uncertainty that apply to the overall risk assessment
(U.S. EPA, 2022b, section 3.4.1.8). Additionally, in characterizing at-
risk populations, the at-risk analysis only used one of the air quality
adjustment approaches used in the overall risk assessment, which
decreases the potential representativeness of the PM2.5
concentrations across the study areas (U.S. EPA, 2022b, section
3.4.1.8). Lastly, this at-risk analysis relies on the stratified risk
coefficients from only one epidemiologic study.\100\ For these reasons,
the Administrator places little
[[Page 16267]]
weight on the absolute results of the risk assessment, including the
at-risk analysis, for purposes of selecting the level of the annual
standard that is requisite.
---------------------------------------------------------------------------
\100\ Additional information on all available at-risk
epidemiologic studies in this reconsideration are available in
section 3.4 and Appendix C of the 2022 PA (U.S. EPA, 2022b, section
3.4, Figure 3-17, and Appendix C, section C.3.2).
---------------------------------------------------------------------------
While there are substantial uncertainties in the absolute results
of the quantitative risk assessment, the EPA also notes that recent
scientific evidence evaluated in the ISA Supplement, which built upon
the 2019 PM ISA conclusions, found that the evidence ``[c]ontinue[s] to
support disparities in PM2.5 exposure and health risks by
race and ethnicity'' while studies of SES ``provide additional support
indicating there may be disparities in PM2.5 exposure and
health risk by SES'' (U.S. EPA, 2022a, p. 5-4). Thus, in light of the
statutory requirement to provide protection for at-risk populations, it
is not surprising that the stratified population results of the risk
assessment suggest that meeting a revised standard would result in
higher risk reductions for minority and low SES populations.
In conclusion, the EPA recognizes that the at-risk analysis was
based on one epidemiologic study that stratified by race/ethnicity for
older adults (e.g., 65+ years old) and that there is increasing
uncertainty in quantitative estimates of stratified risk estimates at
the lower end of the range of standard levels assessed. Moreover, the
EPA finds that the goal of the NAAQS is to provide the requisite
protection to at-risk groups, and where minority populations are
included among the at-risk groups, providing requisite protection to
minority populations will also result in protecting the public health
of other populations. Thus, in setting the NAAQS to protect the health
of at-risk groups with an adequate margin of safety, the Administrator
is selecting the standard that will provide requisite protection,
including for minority populations and other at-risk populations, which
also generally results in protecting the public health of other
populations and reducing risk disparities.
A number of commenters, primarily from industries and industry
groups and some States, support the EPA's proposed decision to retain
the primary 24-hour PM2.5 standard. Many of these commenters
contend that the available scientific evidence and quantitative
information has not significantly changed since the 2020 final decision
and note that important uncertainties remain. The commenters agree with
the EPA's conclusions regarding the controlled human exposure studies
and their relationship to short-term peak PM2.5
concentrations in ambient air. These commenters also noted the primary
annual and 24-hour PM2.5 standards work together to provide
public health protection, with the 98th percentile form of the 24-hour
standard effectively limiting peak daily concentrations. The commenters
agree with the EPA that the current suite of standards maintain
subdaily concentrations below the higher concentrations in controlled
human exposure studies where more consistent health effects are
observed. Commenters also agree with the EPA's conclusions that the
epidemiologic studies are not useful for informing decisions on the
level of the primary 24-hour PM2.5 standard because the
standard focuses on reducing peak exposures with its 98th percentile
form, while the epidemiologic studies often focus on the mean or median
as the percentile for which associations with short-term exposures are
observed. These commenters also agree with the EPA's focus on U.S.-
based studies because of differences compared to Canadian studies. The
commenters also generally agree with the Administrator's judgment that
it was appropriate to place less weight on the risk assessment, noting
that the annual standard is controlling in most areas of the country
and revising the annual standard would have the most potential to
reduce risk related to PM2.5 exposures and would reduce both
average (annual) and peak (daily) PM2.5 concentrations.
Finally, these commenters note that the CASAC did not reach consensus
on whether the current primary 24-hour PM2.5 standard should
be revised, and they agree with the minority of the CASAC's
recommendation in their review of the 2021 draft PA that the primary
24-hour primary PM2.5 standard should be retained. These
commenters also note the CASAC's support in their review of the 2019
draft PA for retaining the primary 24-hour PM2.5 standard.
A number of commenters, primarily from public health and
environmental organizations and some States, oppose the EPA's proposed
decision to retain the primary 24-hour PM2.5 standard. These
commenters support revising the level of the primary 24-hour
PM2.5 standard, contending that a more stringent standard is
necessary to provide requisite public health protection with an
adequate margin of safety, particularly for at-risk groups. In so
doing, these commenters place weight on the same aspects of the
available scientific evidence as the majority of the CASAC in their
review of the 2021 draft PA, and generally advocate for revising the
level of the standard to within the range of 25-30 [micro]g/m\3\ as
recommended by the majority of the CASAC. Some of these commenters
support a level no higher than 25 [micro]g/m\3\ and others support a
level of 20 [micro]g/m\3\. These commenters generally cite to the
available scientific evidence, including evidence of disproportionate
exposures and risks for certain at-risk groups, and the CASAC's advice
in support for their recommendation. Some of these commenters also
suggest that decisions regarding the primary 24-hour PM2.5
standard should not be related to decisions on the primary annual
PM2.5 standard.
As an initial matter, the EPA disagrees with commenters who suggest
that decisions regarding the primary 24-hour PM2.5 standard
should not be related to decisions on the primary annual
PM2.5 standard. In reviewing the adequacy of the public
health protection afforded by the primary PM2.5 standards,
the Administrator's consistent past practice has been to evaluate the
combination of the annual and 24-hour standards together. In 2012, the
then-Administrator concluded that the most effective and efficient way
to reduce total population risk associated with both long- and short-
term PM2.5 exposures was to set a generally controlling
annual standard, and to provide supplemental protection by means of a
24-hour standard set at the appropriate level. In so doing, the then-
Administrator explicitly recognized that potential air quality changes
associated with meeting a revised annual standard (with a level of 12
[micro]g/m\3\) would result in lowering risks associated with both
long- and short-term PM2.5 exposures by lowering the overall
distribution of air quality concentrations, and that retaining a 24-
hour standard at the appropriate level would ensure an adequate margin
of safety against short-term effects in areas with high peak-to-mean
ratios (78 FR 3163, January 15, 2013). In this reconsideration, also,
the Administrator considers it appropriate to rely on the annual
standard (arithmetic mean, averaged over three years) for targeting
protection against both long- and short-term PM2.5
exposures, noting that the annual standard is typically controlling,
while the 24-hour standard (98th percentile, averaged over three years)
can provide supplemental protection against the occurrence of peak 24-
hour PM2.5 concentrations (U.S. EPA, 2022b, section 3.6.3).
Further, the Administrator notes that, as in the 2012 review, changes
in PM2.5 air quality to meet a revised annual standard would
[[Page 16268]]
affect the entire distribution of long- and short-term concentrations,
thus likely resulting not only in lower short- and long-term
PM2.5 concentrations near the middle of the air quality
distribution, but also in fewer and lower short-term peak
PM2.5 concentrations.\101\ Thus, the Administrator continues
to conclude it is appropriate to consider whether the annual and 24-
hour standards together provide requisite protection of public health,
rather than considering each standard in isolation.
---------------------------------------------------------------------------
\101\ Similarly, the Administrator recognizes that changes in
air quality to meet a 24-hour standard, would result not only in
fewer and lower peak 24-hour PM2.5 concentrations, but
also in lower annual average PM2.5 concentrations.
However, as noted in 2012, an approach that relied on setting the
level of the 24-hour standard such that the 24-hour standard was
generally controlling would be less effective and result in less
uniform protection across the U.S. than an approach that focuses on
setting a generally controlling annual standard (78 FR 3163, January
15, 2013).
---------------------------------------------------------------------------
Regarding the appropriate basis for determining the level of the
24-hour standard, a number of commenters who support revising the
primary 24-hour PM2.5 standard to a lower level contend that
the EPA should not rely on the controlled human exposure studies in
evaluating the adequacy of the public health protection afforded by the
primary 24-hour PM2.5 standard. These commenters support
this view by citing the CASAC comments in their review of the 2019
draft PA which advised that controlled human exposure studies have
limitations that may impact their ability to inform conclusions on the
adequacy of the public health protection afforded by the primary 24-
hour PM2.5 standard. Commenters noted that these studies do
not include the most vulnerable populations and often involve exposure
to only one pollutant to elicit a response, and therefore are not
representative of real-world exposures.
Other commenters support the EPA's use of the controlled human
exposure studies to inform the adequacy of the public health protection
and note that the 24-hour standard must at least provide protection
against the health effects observed in controlled human exposure
studies. Some of the commenters cite the Wyatt et al. (2020) study that
demonstrated cardiovascular effects following 2-hour exposures to 120
[micro]g/m\3\ and 4-hour exposures to 37.8 [micro]g/m\3\. Some of these
commenters contend that the current primary 24-hour PM2.5
standard allows PM2.5 exposures comparable to those observed
to elicit effects in the controlled human exposure studies, and
therefore, the EPA must revise the level of the current standard to
protect public health. To support this view, some commenters submitted
an analysis of monitoring data from 2017-2020, which compares the
number of days per year where maximum daily PM2.5
concentrations exceed 120 [micro]g/m\3\ and 37.8 [micro]g/m\3\.
Additionally, other commenters assert that the EPA should focus
less on peak PM2.5 concentrations ``typically measured'' in
areas meeting the current primary PM2.5 standards even if
they do not exceed the concentrations in the controlled human exposure
studies because, in their view, the standard needs to protect against
atypical exposures to atypical peak PM2.5 concentrations.
These commenters conclude that, when considered together, the
controlled human exposure studies and the epidemiologic studies warrant
strengthening the level of the primary 24-hour PM2.5
standard.
The EPA generally disagrees with commenters who contend that it is
inappropriate to rely on the controlled human exposures studies in
evaluating the adequacy of the public health protection afforded by the
primary 24-hour PM2.5 standard. The Agency considers these
studies informative both for establishing biological plausibility and
for determining an appropriate level for the 24-hour standard. When
looking to the experimental studies, the EPA finds that the 2019 ISA
and ISA Supplement included controlled human exposure studies that
report statistically significant effects on one or more indicators of
cardiovascular function following 2-hour exposures to PM2.5
concentrations at and above 120 [mu]g/m\3\ (and at and above 149 [mu]g/
m\3\ for vascular impairment, the effect shown to be most consistent
across studies). As noted in the 2019 ISA, these studies are important
in establishing biological plausibility for PM2.5 exposures
causing more serious health effects, such as those seen in short-term
exposure epidemiologic studies, and they provide support that more
adverse effects may be experienced following longer exposure durations
and/or exposure to higher concentrations. Additionally, one controlled
human exposure study assessed in the ISA Supplement reports evidence of
some effects for cardiovascular markers at lower PM2.5
concentrations, 4-hour exposures to 37.8 [micro]g/m\3\ (Wyatt et al.,
2020). However, there is inconsistent evidence for inflammation in
other controlled human exposure studies evaluated in the 2019 ISA. The
EPA notes that although the controlled human exposure studies do not
provide a threshold below which no effects occur, the observed effects
in these controlled human exposures studies are ones that signal an
intermediate effect in the body, likely due to short-term exposure to
PM2.5, and typically would not, by themselves, be judged as
adverse (88 FR 5620, January 27, 2023) 102 103
---------------------------------------------------------------------------
\102\ Judgments regarding adversity or health significance of
measurable physiological responses to air pollutants have been
informed by guidance, criteria or interpretative statements
developed within the public health community, including the American
Thoracic Society (ATS) and the European Respiratory Society (ERS),
which cooperatively updated the ATS 2000 statement What Constitutes
an Adverse Health Effect of Air Pollution (ATS, 2000) with new
scientific findings, including the evidence related to air pollution
and the cardiovascular system (Thurston et al., 2017).
\103\ The ATS/ERS described its 2017 statement as one ``intended
to provide guidance to policymakers, clinicians and public health
professionals, as well as others who interpret the scientific
evidence on the health effects of air pollution for risk management
purposes'' and further notes that ``considerations as to what
constitutes an adverse health effect, in order to provide guidance
to researchers and policymakers when new health effects markers or
health outcome associations might be reported in future.'' The most
recent policy statement by the ATS, which once again broadens its
discussion of effects, responses and biomarkers to reflect the
expansion of scientific research in these areas, reiterates that
concept, conveying that it does not offer ``strict rules or
numerical criteria, but rather proposes considerations to be weighed
in setting boundaries between adverse and nonadverse health
effects,'' providing a general framework for interpreting evidence
that proposes a ``set of considerations that can be applied in
forming judgments'' for this context (Thurston et al., 2017).
---------------------------------------------------------------------------
The EPA notes that the majority of the CASAC, in their review of
the 2021 draft PA, commented that these controlled human exposure
studies generally do not include populations with substantially
increased risk from exposure to PM2.5, such as children,
older adults, or those with more severe underlying illness, and often
involve exposure to only one pollutant to elicit a response. However,
both the majority and the minority of the CASAC explained that, even
taking into consideration their limitations, the controlled human
exposure studies provide some support for assessing the adequacy of the
24-hour standard.\104\
---------------------------------------------------------------------------
\104\ In their review of the 2021 draft PA, the majority of the
CASAC advised that ``evidence of effects from controlled human
exposure studies with exposures close to the current standard
support epidemiologic evidence for lowering the standard''
(Sheppard, 2022a, p. 4 of consensus letter). The minority of the
CASAC also advised that it was appropriate to place ``more emphasis
on the controlled human exposure studies, showing effects at
PM2.5 concentrations well above those typically measured
in areas meeting the current standards'' (Sheppard, 2022a, p. 4 of
consensus letter), in evaluating adequacy of the 24-hour standard.
---------------------------------------------------------------------------
The EPA agrees with the CASAC that the controlled human exposure
studies generally do not include populations with substantially
increased risk from exposure to PM2.5, like children, older
adults, or those with pre-existing severe illness, like cardiovascular
effects. As such, and as an initial note, these
[[Page 16269]]
studies are therefore somewhat limited in their ability to inform at
what concentrations effects may be elicited in at-risk populations. In
spite of this limitation, the EPA also agrees with the CASAC, that even
taking into consideration the limitations of the controlled human
exposure studies, these studies can provide some support for evaluating
the adequacy of the 24-hour standard. However, the EPA further notes
that while the controlled human exposure studies are important in
establishing biological plausibility, the health outcomes observed in
these controlled human exposure studies are often ``intermediate''
outcomes (i.e., not always clearly adverse) and therefore it is unclear
how the importance of the effects observed in the studies should be
interpreted with respect to adversity to public health. The EPA finds
that it is appropriate to consider these study limitations in assessing
the information provided by controlled human exposure studies in
evaluating the adequacy of the primary 24-hour PM2.5
standard.
The EPA agrees with commenters that the primary 24-hour
PM2.5 standard must at least provide protection against the
health effects consistently observed in controlled human exposure
studies. As discussed in the proposal, the EPA looks at whether the
exposures that elicit a response following exposure to PM2.5
in the controlled human exposure studies occur under recent air quality
conditions in areas meeting the current standards. Based on these air
quality analyses, the EPA concludes that these types of exposures very
rarely occur when the current standards are being met.
The EPA did receive multiple comments questioning these results and
the approach in the EPA's analyses. For example, some commenters
submitted an analysis of monitoring data from 2017-2020, which compares
the number of days per year where maximum daily PM2.5
concentrations exceed 120 [micro]g/m\3\ and 37.8 [micro]g/m\3\ and
evaluate the number of days subset by groups of monitors with 4-year
average PM2.5 concentrations close to the levels of
combinations of current and proposed annual (+/- 0.2 [micro]g/m\3\) and
24-hour (+/-2 [micro]g/m\3\) PM2.5 standards. To support
their view that the primary PM2.5 standards should be
revised, the commenters describe decreases in days per monitor per year
with 2-hour maximum concentrations greater than 120 [micro]g/m\3\ and
4-hour maximum concentrations greater than 37.8 [micro]g/m\3\ when
comparing monitors that achieve close to 10 and 30 [micro]g/m\3\ versus
monitors that meet close to 8 [micro]g/m\3\ and 25 [micro]g/m\3\. The
commenters noted decreases in the number of days per monitor per year
with 2-hour maximum concentrations over 120 [micro]g/m\3\ and 4-hour
max concentration over 37.8 [micro]g/m\3\ were also seen when comparing
monitors close to achieving 24-hour standards with levels of 35
[micro]g/m\3\ versus 25 [micro]g/m\3\.
First, the EPA notes that this analysis submitted by commenters was
limited to a very small number of monitors and did not include a
national perspective. Second, the EPA notes that this analysis focused
on number of days (rather than the number of times) where there was a
2-hour maximum concentration over 120 [micro]g/m\3\ or a 4-hour max
concentration over 37.8 [micro]g/m\3\. In order to evaluate the
protection provided by the current 24-hour standard against peak
exposures, including exposures with 2-hour concentrations greater than
120 [micro]g/m\3\ and 4-hour concentrations greater than 37.8 [micro]g/
m\3\, the EPA considers it more informative and appropriate from a
public health perspective to assess the number of times a subdaily
exposure of concern occurs in a year, rather than the number of days on
which they occur because the former identifies more potential exposures
of concern and provides more information about the scale and scope of
the occurrences of those exposures. Lastly, the analyses allowed
monitors somewhat above the standards to be included. Therefore, it is
unclear whether the exceedances of the 2-hour or 4-hour benchmarks
would still have occurred if the area had actually been meeting the
current primary PM2.5 standards. However, in considering the
analyses submitted by the commenters, the EPA conducted new analyses
\105\ that looked at all individual monitors across the U.S. and
evaluated the percentage of times the monitors experienced a 2-hour
maximum concentration over 120 [micro]g/m\3\ or a 4-hour max
concentration over 37.8 [micro]g/m\3\ when that monitor was meeting the
current standards. Further, given that the Administrator concludes that
the level of the current primary annual PM2.5 is not
adequate and that it should be revised to 9.0 [micro]g/m\3\, the new
analysis evaluates the percentage of times during a recent 3-year
period (i.e. 2019-2021) that individual monitors experienced a 2-hour
maximum concentration over 120 [micro]g/m\3\ or a 4-hour max
concentration over 37.8 [micro]g/m\3\ when that monitor was meeting the
current primary 24-hour PM2.5 standard with its level of 35
[micro]g/m\3\ and a revised primary annual PM2.5 standard of
9.0 [micro]g/m\3\.
---------------------------------------------------------------------------
\105\ Jones et al. (2023). Comparison of Occurrence of
Scientifically Relevant Air Quality Observations Between Design
Value Groups. Memorandum to the Rulemaking Docket for the Review of
the National Ambient Air Quality Standards for Particulate Matter
(EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
In evaluating the results from the new analyses, it is important to
keep in mind that the 2019 ISA and ISA Supplement concluded that the
most consistent evidence from the controlled human exposures studies is
for impaired vascular function following 2-hour exposures to average
PM2.5 concentrations at and above about 120 [micro]g/m\3\,
with less consistent evidence for effects following exposures to
concentrations lower than 120 [micro]g/m\3\. The new analyses show that
across all monitors, on average, only 0.029 percent of 2-hour
observations reach PM2.5 concentrations higher than 120
[micro]g/m\3\ in areas meeting the current 24-hour standard and a
revised annual standard of 9.0 [micro]g/m\3\. Further, recognizing that
one purpose of the 24-hour standard is to protect against exposure in
areas with high peak-to-mean ratios, when assessing the monitors
individually across the U.S. under these same conditions, the monitors
reporting the highest PM2.5 concentrations have only 0.47
percent of 2-hour observations reach PM2.5 concentrations
higher than 120 [micro]g/m\3\.
Additionally, the analyses also evaluated the frequency of
reporting a 4-hour maximum concentration over 37.8 [micro]g/m\3\ when
monitors were meeting the current 24-hour standard and a revised annual
standard of 9.0 [micro]g/m\3\. For this part of the analysis, the EPA
finds that across all monitors, on average, only 0.41 percent of 4-hour
observations reach PM2.5 concentrations higher than 37.8
[micro]g/m\3\ in areas meeting the current 24-hour standard and a
revised annual standard of 9.0 [micro]g/m\3\. Further, when assessing
the monitors individually across the U.S. under these same conditions,
the monitors reporting the highest PM2.5 concentrations have
only 2.6 percent of 4-hour observations reach PM2.5
concentrations higher than 37.8 [micro]g/m\3\. Thus, the EPA disagrees
with commenters that the current primary 24-hour PM2.5
standard typically allows PM2.5 exposures at or above those
observed to cause health effects in controlled human exposure studies.
Furthermore, the EPA notes that in light of the small number of
occurrences and the intermediate nature of the effects observed in
Wyatt et al. (2020) at concentrations of 37.8 [micro]g/m\3\ (i.e.,
effects that typically would not, by themselves, be judged as adverse),
there is substantial basis to doubt whether further improvements in
public health
[[Page 16270]]
would be achieved by further reducing these exposures. In drawing this
conclusion, the EPA notes the lack of evidence of effects from
controlled human exposure studies at levels below the current 24-hour
standard and the fact that the results of Wyatt et al. (2020) are
inconsistent with other currently available studies, and this study
only observes intermediate effects.
In response to commenters that cited the majority of the CASAC's
view that, in general, ``[t]here is . . . less confidence that the
annual standard could adequately protect against health effects of
short-term exposures'' (Sheppard, 2022a, p. 4 of consensus letter), the
EPA disagrees with the majority of CASAC, noting that the results of
the EPA's analysis suggest that high peak concentrations are extremely
infrequent in areas meeting an annual standard of 9.0 [micro]g/m\3\,
occurring less than 0.029-0.41 percent of the time (for 2-hour
concentrations >120 [micro]g/m\3\ and 4-hour concentrations >37.8
[micro]g/m\3\, respectively). This suggests that in most locations,
even the upper tail of the distribution would be controlled quite well
under a revised annual standard. With regard to the likelihood that the
current standards would allow peak concentrations that are clearly of
concern from a health perspective, therefore, the EPA concludes that
such occurrences are extremely infrequent--and will be even less
frequent under the improved air quality conditions associated with
meeting a revised annual PM2.5 standard of 9.0 [micro]g/
m\3\.
A number of commenters who support revising the primary 24-hour
PM2.5 standard to a lower level suggest that the available
epidemiologic evidence provides support for such a revision. To support
their view, the commenters note that the currently available evidence,
including a number of epidemiologic studies that demonstrate
associations between short-term PM2.5 exposures and health
effects, provides support for causal relationships for short-term
PM2.5 exposures and health effects as described in the 2019
ISA and ISA Supplement. The commenters further note that the available
epidemiologic studies include diverse populations that are broadly
representative of the U.S. population, including at-risk populations,
which they assert is an advantage over the controlled human exposure
studies and the risk assessment, which are not as broadly
representative.
These commenters highlight a number of specific epidemiologic
studies that they suggest provide support for revising the level of the
24-hour standard. Additionally, commenters contend that there are
epidemiologic studies using restricted analyses that show that positive
and statistically significant associations between short-term
PM2.5 exposure and mortality persist at daily mean
concentrations below 25 [micro]g/m\3\. The commenters also cite several
studies that provide no evidence of a threshold. These commenters also
point to the CASAC advice in their review of the 2021 draft PA, where
the majority of the CASAC cited epidemiologic studies using restricted
analyses as offering support for revision. The commenters argue that
the EPA cannot base discretion on uncertainties related to the methods
used in restricted analyses in the epidemiologic studies. In so doing,
these commenters disagree with the EPA that it is important to take
into consideration that these studies do not consider the form or
averaging time of the 24-hour standard. Finally, the commenters claim
that while the EPA stated that the study-reported means from
epidemiologic studies that use restricted analyses are more useful for
identifying impacts from typical 24-hour exposures than for peak 24-
hour exposures, the commenters assert that the studies also indicate
that there are health risks at relatively high concentrations below the
current level of the primary 24-hour PM2.5 standard that
must be addressed.
As noted by the commenters, epidemiologic studies that show
positive and statistically significant associations between short-term
PM2.5 exposure and mortality provide support for the causal
determination in the 2019 ISA. The EPA also agrees that the available
epidemiologic studies include diverse populations that are broadly
representative of the U.S. population, including at-risk populations.
Further, the EPA agrees that studies evaluated in the 2019 ISA and the
ISA Supplement continue to provide evidence of linear, no-threshold
concentration-response relationships, but with less certainty in the
shape of the curve at lower concentrations (i.e., below about 8
[micro]g/m\3\), with some recent studies providing evidence for either
a sublinear, linear, or supralinear relationship at these lower
concentrations (U.S. EPA, 2019a, section 11.2.4; U.S. EPA, 2022a,
section 2.2.3.2).
However, findings of positive, significant associations in short-
term epidemiologic studies do not directly indicate that short-term
effects would occur in areas meeting the 24-hour standard and
therefore, do not directly address the question of whether the current
24-hour standard is adequate. While short-term epidemiologic studies
evaluate associations between distributions of ambient PM2.5
and health outcomes, they do not identify the specific exposures (i.e.,
a specific 24-hour concentration) that can lead to the reported
effects. Short-term epidemiologic studies evaluate the association
between day-to-day variation in daily (24-hour) PM2.5
exposure and health endpoints (e.g., mortality) to understand how these
changes in air pollution concentrations are associated with changes in
health outcomes. But these studies do not report daily concentrations;
rather, they report the long-term mean concentration of the 24-hour
PM2.5 concentrations over the entire multi-year period of
the study, and typically report their results as a relative risk (e.g.,
for each 10 [micro]g/m\3\ increase in PM2.5, the risk of
mortality or cardiovascular hospital admissions increases by a certain
percentage, across the full range of the 24-hour PM2.5
concentrations in the study). This means that there is no specific
point in the air quality distribution of any epidemiologic study that
represents a ``bright line'' at and above which effects have been
observed and below which effects have not been observed. Nor, as noted
above, do these studies allow for any direct inferences about health
impacts associated with the short-term ``peak'' exposures that the
primary 24-hour standard is designed to protect against. While there
can be considerable variability in daily exposures over a multi-year
study period, most of the estimated exposures in these epidemiologic
studies reflect days with ambient PM2.5 concentrations
around the mean or middle of the air quality distributions examined
(i.e., ``typical'' days rather than days with extremely high or
extremely low concentrations). This is true of long-term epidemiologic
studies as well. The difference between epidemiologic studies examining
associations with long-term exposures and short-term exposures is
comparing different levels of exposure over different exposure
durations (i.e., long-term studies exposures are defined as those that
are annual or multi-year, while short-term exposures are defined as
those that are mostly 24-hour) (U.S. EPA, 2019a, section P.3.1). Thus,
in both cases, and in the absence of a discernible threshold,
epidemiologic studies of short-term and long-term exposures provide the
strongest support and confidence for reported health effect
associations around the middle portion of the PM2.5 air
quality distribution (e.g., the study-reported mean PM2.5
[[Page 16271]]
concentration), which corresponds to the bulk of the underlying data,
rather than at the extreme upper or lower ends of the distribution.
However, the difference between the annual standard and the 24-hour
standard, aside from averaging times, is that the form of the annual
standard is a mean PM2.5 concentration, which is based on
the bulk of the air quality data, while the form of the 24-hour
standard is a 98th percentile form, which is based on peak
concentrations. Both long-term and short-term epidemiologic studies are
informative for determining the appropriate level of the annual
PM2.5 standard, which is designed to control ``typical''
daily exposures and risks, because these studies most often report
long-term mean (or median) PM2.5 concentrations that are
representative of ``typical'' exposures that are associated with health
effects. In contrast, while the short-term epidemiologic studies
examine health effects associated with shorter exposure durations
(e.g., mostly 24-hour exposures), these studies are less informative
for determining the appropriate level of the 24-hour PM2.5
standard because these studies do not report the 98th percentile
PM2.5 concentrations,\106\ which is more directly comparable
to the form of the 24-hour standard. Additionally, if the 98th
percentile of data were reported, the EPA would consider the peak
concentrations observed in these studies (which by definition rarely
occur) in conjunction with other supporting evidence. However, as
already noted, there is an absence of new information in this
reconsideration (either from controlled human exposure studies or
epidemiologic studies) suggesting that peak concentrations just below
the level of the current 24-hour standard (with its level of 35
[micro]g/m\3\) are associated with adverse effects. Instead, the
evidence links risk to more typical daily exposures near the middle of
the air quality distribution--exposures most effectively controlled
through a strengthening of the annual standard. As noted in the 2012
final rule, ``reducing the annual standard is the most efficient way to
reduce the risks from short-term exposures . . . as the bulk of the
risk comes from the large number of days across the bulk of the air
quality distribution, not the relatively small number of days with peak
concentrations'' (78 FR 3156, January 15, 2013).
---------------------------------------------------------------------------
\106\ In the 2022 PA, the EPA has identified a number of key
areas for additional research and data collection for
PM2.5, based on the uncertainties and limitations that
remain in the scientific evidence and technical information. In
addition to research and data collection, the EPA specifically
highlights additional information that could be reported in the
epidemiologic studies that may help inform future reviews of the
primary PM2.5 standards, including additional descriptive
statistics in the upper percentiles of the air quality distribution
(i.e., from the 95th to the 99th percentile), as well as the number
of days of concentrations and/or health events within each of these
percentiles (U.S. EPA, 2022a, section 3.7).
---------------------------------------------------------------------------
As noted above, in evaluating the adequacy of the current
standards, the EPA has consistently considered the annual standard
(based on arithmetic mean concentrations) and 24-hour standard (based
on 98th percentile concentrations) together in evaluating the public
health protection provided by the standards against the full
distribution of short- and long-term PM2.5 exposures.
Moreover, the EPA has previously noted that the annual standard is
generally controlling in most parts of the country, providing an
effective and efficient way to reduce total population risk to both
long- and short-term PM2.5 exposures, while the 24-hour
standard, with its 98th percentile form, provides supplemental
protection, particularly for areas with high peak-to-mean ratios of 24-
hour PM2.5 concentrations (78 FR 3158, January 15, 2013). In
such areas, annual average PM2.5 concentrations could be
quite low, and the 24-hour standard provides a means of ensuring
control of episodic peaks possibly associated with strong local or
seasonal sources, or PM2.5-related effects that may be
associated with shorter-than daily exposure periods. The approach taken
in evaluating the adequacy and alternative levels of the annual
standard has been to evaluate the long-term mean PM2.5
concentrations of both long-term and short-term key epidemiologic
studies, where we have the most confidence in the reported health
effects association, while also giving some consideration to lower
percentiles of the air quality distribution (e.g., 25th percentiles).
However, using a similar approach to evaluate the adequacy of the
current and any potential alternative levels of the 24-hour standard
with short-term epidemiologic studies, as the majority of CASAC and
some commenters are suggesting, presents challenges.
Short-term epidemiologic studies, including those that use
restricted analyses, often report metrics that include mean
PM2.5 concentrations, with some studies also reporting lower
percentiles, such as the 25th percentile. As previously noted above,
for studies of daily PM2.5 exposure, which examine
associations between day-to-day variation in PM2.5
concentrations and health outcomes, often over several years, most of
the estimated exposures reflect days with ambient PM2.5
concentrations around the middle of the air quality distributions
examined (i.e., the mean or median). However, there is not a metric or
statistic reported in short-term epidemiologic studies that allows for
a direct comparison to the current 24-hour PM2.5 standard
and its 98th percentile form. While a 98th percentile of
PM2.5 concentrations is a metric that might be more closely
compared to the 24-hour standard level, 98th percentile
PM2.5 concentrations were not reported in key epidemiologic
studies. Consistent with the Administrator's final decision in 2012,
the EPA notes that even if 98th percentile values were reported, it
would be inappropriate to focus on these concentrations without also
considering the impact of a revised annual standard on short-term
concentrations, since many areas would be expected to experience
decreasing short- and long-term PM2.5 concentrations in
response to a revised annual standard (78 FR 3156, January 15, 2013).
Furthermore, in light of the scarcity of days at the very upper end of
the distribution, and to avoid placing undue reliance on the peak
concentrations observed in these studies (which by definition rarely
occur), the EPA finds that such values would need to be considered in
conjunction with other supporting evidence. In addition, as described
above, the other lines of evidence available for consideration by the
EPA do not indicate that the current primary 24-hour standard requires
revision to protect public health with an adequate margin of safety.
The EPA notes again the lack of corroborating evidence from controlled
human exposure studies. While the EPA agrees with the CASAC that the
controlled human exposure studies are limited in their ability to speak
to the concentrations at which effects may be elicited in at-risk
populations, as discussed above the lowest concentration associated
with effects is 37.8 [micro]g/m\3\ and the effects observed were
``intermediate'' outcomes that are not by themselves considered
adverse. We also note that, as detailed in section II.A.2.a above, the
study that observed intermediate effects at concentrations of 37.8
[micro]g/m\3\ was evaluated in the ISA Supplement and the results of
this study were inconsistent with the controlled human exposure studies
evaluated in the 2019 ISA. Additionally, as noted above, the EPA finds
that across all monitors, on average, only 0.41 percent of 4-hour
observations reach PM2.5 concentrations higher than 38
[micro]g/m\3\ in areas meeting the current 24-hour
[[Page 16272]]
standard and a revised annual standard of 9.0 [micro]g/m\3\. Given the
rarity of these occurrences and the fact that the effects associated
with exposures to this PM2.5 concentration have not been
found to be adverse in and of themselves, the EPA finds it reasonable
to conclude that this pattern of air quality will protect at-risk
populations, even though such populations were not in the study groups.
The EPA concludes that further evidence would be needed at specific
short-term (i.e., hourly or daily) concentrations below the level of
the current 24-hour standard to support any revision to the current 24-
hour standard.
With regard to the data that are available from the short-term
epidemiologic studies (which, as noted, do not include 98th percentile
values), the EPA considers it inappropriate to utilize the study-
reported means from the short-term epidemiologic evidence to assess the
adequacy of the 24-hour standard, with its 98th percentile form,
considering that the study-reported mean concentrations do not provide
meaningful insight regarding the frequency or health significance of
peak concentrations occurring during the study period. As indicated in
the 2022 PA, the study-reported means of short-term epidemiologic
studies do not serve a purpose in determining a level at which we can
confidently attribute effects to the impact of ``peak'' exposures. The
24-hour standard is intended to provide supplemental protection against
short-term peak exposures and while there is a general relationship
between mean concentrations and 98th percentile concentrations in
individual locations, such relationships vary by location and there is
not an established relationship that can be relied upon to predict 98th
percentile concentrations based on mean PM2.5 concentrations
reported in multi-city epidemiologic studies. Instead, mean
concentrations from short-term epidemiologic studies are more useful in
addressing questions regarding the effects of ``typical'' or average
24-hour exposures, which are addressed through the annual standard. For
this reason, the EPA does consider the mean concentrations of short-
term studies (as well as the means from the long-term studies) in
evaluating the level of the annual standard, which the EPA recognizes
as the generally controlling standard for both long- and short-term
exposures. However, the EPA does not agree with commenters that it is
appropriate to use means from short-term epidemiologic studies as the
basis for a decision-making framework to determine the adequacy of the
current 24-hour standard, with its 98th percentile form.
As described in the proposal (88 FR 5613, January 27, 2023), the
2022 PA also noted the epidemiologic studies that restrict 24-hour
average PM2.5 concentrations to values of less than 35
[micro]g/m\3\, and in some cases less than 25 [micro]g/m\3\, and annual
average PM2.5 concentrations less than 12 [micro]g/m\3\.
Restricted analyses use a subset of data from their main analyses and
conduct an epidemiologic study with health events that occur at
concentrations below a certain concentration (e.g., 25 [micro]g/m\3\).
While some of these studies do not report the mean PM2.5
concentration for the restricted analysis, the mean of the restricted
analysis is presumably less than the mean PM2.5
concentration in the main analysis. Restricted analyses from long-term
and short-term exposure epidemiologic studies are informative in
providing support that the health effects associations are not driven
by just the upper peaks of the PM2.5 air quality
distributions and provide support for revision to the level of the
annual PM2.5 standard. Short-term restricted analyses also
report positive associations between short-term PM2.5
exposure and morbidity and mortality. As an example, in a restricted
analysis evaluating the association between short-term exposures and
PM2.5 concentrations less than 25 [micro]g/m\3\, Di et al.
(2017a) removed 6.3 percent of the data from their main analyses,
(i.e., all PM2.5 concentrations greater than 25 [micro]g/
m\3\), and still found a positive and significant association between
short-term PM2.5 exposure and mortality. This study provides
additional support that the association between short-term exposure to
PM2.5 and mortality in the main epidemiologic analysis is
not driven by the upper peaks of the PM2.5 air quality
distribution, which in turn supports the conclusion that lowering the
entire distribution of air quality concentrations through a revised
annual standard is an appropriate means of protecting against adverse
effects from short-term exposure, as discussed further below.
In their review of the 2021 draft PA, the majority of the CASAC
highlighted three U.S.-based epidemiologic studies that restricted 24-
hour average PM2.5 concentrations below 25 [mu]g/m\3\ as a
part of their rationale for recommending that the EPA revise the level
of the primary 24-hour PM2.5 standard. Similarly, in
evaluating positive associations in restricted analyses, some
commenters also suggest that because an association exists at 24-hour
concentrations below 25 [micro]g/m\3\, the 24-hour standard level
should be set at the concentration at which the analysis was restricted
(e.g., 25 [micro]g/m\3\). However, the EPA notes that neither the CASAC
nor public commenters provided any detail regarding, how, in their
view, these studies demonstrate that the level of the current 24-hour
standard is not adequate, and/or how these studies demonstrate what
revised level of the 24-hour standard would provide requisite public
health protection with an adequate margin of safety. The EPA considers
that such an approach would have several important limitations. First,
the approach assumes that a specific point on the air quality
distribution (e.g., the point at which the analysis was restricted) is
where health effects are exhibited and where we have the most
confidence in the reported association. However, in addition to the
limitations associated with the short-term epidemiologic studies
outlined above, the EPA does not agree that it would be appropriate to
identify the requisite level of the primary 24-hour PM2.5
standard based on the specific concentration at which the analyses
restrict their studies. The choice to restrict the data at a particular
concentration is in effect arbitrary, and does not establish that any
particular effects are attributable to that concentration as opposed to
other concentrations within the restricted analysis.
Further, these restricted analyses do not report the
PM2.5 concentration at the 98th percentile of data or other
metrics relating to the upper end of the distribution that could
provide information about health risks associated with peak exposures.
For example, the CASAC does not provide a discussion of what the
comparable 98th percentile concentration is in the distribution of
remaining 24-hour PM2.5 concentrations of restricted
analyses (because such data is not reported by the study authors) and
what degree of confidence the Administrator should place on those upper
percentile values (e.g., 98th percentile values). In order to identify
a level of the 24-hour standard based on associations between the
``upper end'' of exposures, either in the unrestricted or the
restricted analyses, and adverse health effects, it would be necessary
to have both greater detail on the distribution of air quality in the
study and greater confidence in the reported association at the peak
concentrations such as the 98th percentile--in other words, a better
understanding of how specific 24-hour concentrations correspond to the
frequency and total number of observed health events in the study.
Further, the EPA notes that when resulting analyses based on the
[[Page 16273]]
restricted dataset continue to find positive associations between the
remaining air quality distribution and health effects, it suggests that
the relationship was in fact not driven primarily by the upper tail
(now removed from the dataset) but rather by lower portions of the
distribution of air quality. In other words, we have no confidence that
the remaining upper end of the air quality distribution is driving the
remaining associations reported in the restricted analyses, as opposed
to the vast array of health events at and around the mean
PM2.5 concentration. In fact, it is reasonable to conclude
that to effectively address the health effects observed in the study,
it is necessary to control not just the peak concentrations but to
reduce the bulk of the exposures (occurring near the mean), a task more
effectively achieved, as noted above through a tightening of the annual
standard, which has the effect of shifting the entire distribution of
PM2.5 concentrations downward (both peaks and means).
Therefore, while the EPA agrees that both short- and long-term
epidemiologic studies that completed restricted analyses and reported
the resulting study means could be used to inform conclusions regarding
the adequacy of the annual standard, given that the resulting study
means (when reported) could be evaluated in the context of the decision
framework described above for informing decisions on the level of the
annual standard, the EPA considers that current short-term
epidemiologic studies that restrict analyses are subject to the same
limitations outlined above for current short-term epidemiologic studies
in how they can be used in a decision-making framework to inform the
adequacy and alternative level of the primary 24-hour PM2.5
standard. As such, while the available short-term epidemiologic studies
that restrict their analyses are useful for informing conclusions
regarding the strength of the associations for health outcomes, they
are not, as currently designed, as useful for informing conclusions
regarding the adequacy of the current primary 24-hour PM2.5
standard. In reaching this conclusion, the EPA notes that the majority
of the CASAC did not address the limitations of these studies outlined
in the 2021 draft PA, particularly in the context of the 24-hour
standard with its 98th percentile form. Among the future research needs
identified by the EPA in the 2022 final PA, the Agency noted a number
of gaps in the currently available information reported in the
epidemiologic studies of short-term exposure, including ``descriptive
statistics of PM2.5 concentrations at individual percentiles
from the 95th percentile to the 99th percentile, as well as the number
of days of concentrations and/or health events within each of these
percentiles'' and other descriptive statistics and details regarding
analytical design in studies employing restricted analyses (U.S. EPA,
2022b, pp. 3-225 to 3-226). Such information could significantly
improve the EPA's ability to draw conclusions from these studies with
regard to the adequacy of the current primary 24-hour PM2.5
standard.
Due to the limitations and uncertainties outlined above, in
reaching his decision on the primary 24-hour PM2.5 standard,
the Administrator judges that the information from currently available
short-term epidemiologic studies, including those that use restricted
analyses, is inadequate to inform decisions regarding the adequacy of
the current 24-hour standard. Additionally, consistent with the final
decision in 2012, the EPA continues to view an approach that focuses on
setting a generally controlling annual standard as the most effective
and efficient way to reduce total population risk associated with both
long- and short-term PM2.5 exposures. Potential air quality
changes associated with meeting an annual standard level of 9.0
[micro]g/m\3\ will result in lowering risk associated with both long-
and short-term PM2.5 exposure by lowering the overall air
quality distribution. As discussed above, reducing the annual standard
is the most efficient way to reduce the risks from short-term exposures
identified in the epidemiologic studies, as the available evidence
suggests the bulk of the risk comes from the large number of days
across the bulk of the air quality distribution, not the relatively
small number of days with peak concentrations. However, as in the 2012
review, the Administrator recognizes that an annual standard alone
would not be expected to offer sufficient protection with an adequate
margin of safety against the effects of short-term PM2.5
exposures in all parts of the country, particularly in areas with high
peak-to-mean ratios, and concludes that it is appropriate to continue
to provide supplemental protection by means of a 24-hour standard. In
so doing, the Administrator concludes that retaining the level of the
primary 24-hour PM2.5 standard of 35 [micro]g/m\3\ will
provide requisite protection against short-term peak PM2.5
concentrations, in conjunction with a revised annual standard level of
9.0 [micro]g/m\3\.
4. Administrator's Conclusions
This section summarizes the Administrator's considerations and
conclusions related to the adequacy of the current primary
PM2.5 standards and presents his decision to revise the
primary annual PM2.5 standard to a level of 9.0 [micro]g/
m\3\ and retain the primary 24-hour PM2.5 standard. In
establishing primary standards under the Act that are ``requisite'' to
protect public health with an adequate margin of safety, the
Administrator is seeking to establish standards that are neither more
nor less stringent than necessary for this purpose. He recognizes that
the requirement to provide an adequate margin of safety was intended to
address uncertainties associated with inconclusive scientific and
technical information and to provide a reasonable degree of protection
against hazards that research has not yet identified. However, the Act
does not require that primary standards be set at a zero-risk level;
rather, the NAAQS must be sufficiently protective, but not more
stringent than necessary.
Given these requirements, the Administrator's final decision in
this reconsideration is a public health policy judgment drawing upon
scientific and technical information examining the health effects of
PM2.5 exposures, including how to consider the range and
magnitude of uncertainties inherent in that information. This public
health policy judgment is based on an interpretation of the scientific
and technical information that neither overstates nor understates its
strengths and limitations, nor the appropriate inferences to be drawn,
and is informed by the Administrator's consideration of advice from the
CASAC and public comments received on the proposal.
The initial issue to be addressed in the reconsideration of the
primary PM2.5 standards is whether, in view of the advances
in scientific knowledge and other information reflected in the 2019
ISA, ISA Supplement, and 2022 PA, the current standards are requisite
to protect public health with an adequate margin of safety. In
considering the adequacy of the current suite of primary
PM2.5 standards, the Administrator has considered the large
body of evidence presented and assessed in the 2019 ISA and ISA
Supplement, the conclusions presented in the 2022 PA, the views
expressed by the CASAC, and public comments. The Administrator has
taken into account both evidence- and risk-based considerations in
developing final conclusions on the adequacy of the current primary
PM2.5 standards. The Administrator has additionally
[[Page 16274]]
considered the associated public health policy judgments and judgments
about the uncertainties inherent in the scientific evidence and
quantitative analyses that are integral to the conclusions on the
adequacy of the current primary PM2.5 standards.
In evaluating the adequacy of the current standards, the
Administrator first recognizes the longstanding body of health evidence
supporting relationships between PM2.5 exposures (short- and
long-term) and mortality and serious morbidity effects. The evidence
available in this reconsideration (i.e., that assessed in the 2019 ISA
and ISA Supplement) and summarized above in section II.A.2.a reaffirms,
and in some cases strengthens, the conclusions from the 2009 ISA
regarding the health effects of PM2.5 exposures. Recent
epidemiologic studies demonstrate generally positive and often
statistically significant associations between PM2.5
exposures and a number of health effects, including non-accidental,
cardiovascular, or respiratory mortality; cardiovascular or respiratory
hospitalizations or emergency room visits; and other mortality/
morbidity outcomes (e.g., lung cancer mortality or incidence, asthma
development). Recent controlled human exposure and animal toxicological
studies, as well as evidence from epidemiologic panel studies,
strengthens support for potential biological pathways through which
PM2.5 exposures could lead to the serious effects reported
in many population-level epidemiologic studies, including support for
pathways that could lead to cardiovascular, respiratory, nervous
system, and cancer-related effects. In considering the available
scientific evidence, and consistent with approaches employed in past
NAAQS reviews, the Administrator places the most weight on evidence
supporting ``causal'' or ``likely to be causal'' relationship with long
or short-term PM2.5 exposures. In addition, the
Administrator also takes note of those populations identified to be at
greater risk of PM2.5-related health effects, as
characterized in the 2019 ISA and ISA Supplement, and the potential
public health implications.
In evaluating what existing or revised standards may be requisite
to protect public health, as described above in section II.A.2, the
Administrator's approach recognizes that the current annual standard
(based on arithmetic mean concentrations) and 24-hour standard (based
on 98th percentile concentrations), together, are intended to provide
public health protection against the full distribution of short- and
long-term PM2.5 exposures. This approach recognizes that
changes in PM2.5 air quality designed to meet either the
annual or the 24-hour standard would likely result in changes to both
long-term average and short-term peak PM2.5 concentrations.
Further, consistent with the approach adopted in 2012, the
Administrator concludes that the most effective and efficient way to
reduce total population risk associated with both long- and short-term
PM2.5 exposures is to set a generally controlling annual
standard, and to provide supplemental protection against the occurrence
of peak 24-hour PM2.5 concentrations by means of a 24-hour
standard set at the appropriate level. In reaching this conclusion, the
Administrator explicitly recognizes that air quality changes associated
with meeting a revised annual standard would result in lowering risks
associated with both long- and short-term PM2.5 exposures by
lowering the overall distribution of air quality concentrations,
leading to not only in lower short- and long-term PM2.5
concentrations near the middle of the air quality distribution, but
also in fewer and lower short-term peak PM2.5
concentrations. Similarly, the Administrator recognizes that changes in
air quality to meet a 24-hour standard, would result not only in fewer
and lower peak 24-hour PM2.5 concentrations, but also in
lower annual average PM2.5 concentrations. However, as noted
in 2012, he also recognizes that an approach that relies on setting the
level of the 24-hour standard such that the 24-hour standard is
generally controlling would be less effective and result in less
uniform protection across the U.S. than an approach that focuses on
setting a generally controlling annual standard. Thus, he concludes
that relying on a revised annual standard as the controlling standard
will reduce aggregate risks associated with both long- and short-term
exposures more consistently than a generally controlling 24-hour
standard. He further concludes that retaining a 24-hour standard at the
appropriate level will ensure an adequate margin of safety against
short-term effects in areas with high peak-to-mean ratios.
In light of his focus on the annual standard as the generally
controlling standard, in considering whether the primary
PM2.5 standards are adequate, the Administrator first
considers information available to inform his final conclusions
regarding the primary annual PM2.5 standard. In so doing, he
notes that in this reconsideration, a large number of key U.S.
epidemiologic studies report positive and statistically significant
associations for air quality distributions with overall mean
PM2.5 concentrations that are well below the current level
of the annual standard of 12.0 [mu]g/m\3\. He further recognizes that
there is additional scientific evidence assessed in the 2019 ISA and
newly assessed in this reconsideration in the ISA Supplement that can
provide supplemental information to inform his decisions. In addition
to the key U.S. epidemiologic studies, the Administrator also
recognizes that key Canadian epidemiologic studies also demonstrate
positive and statistically significant associations at concentrations
below 12 [mu]g/m\3\. He also recognizes that epidemiologic studies that
restrict annual average PM2.5 concentrations to below 12
[mu]g/m\3\ also provide support for positive and statistically
significant associations at lower mean PM2.5 concentrations,
as do accountability studies that also suggest public health
improvements may occur at concentrations below 12 [mu]g/m\3\.
With regard to the available scientific evidence to inform his
final decisions on the adequacy of the current 24-hour standard, the
Administrator finds that there is less information available to support
decisions on the 24-hour standard than that summarized above for the
annual standard. The Administrator first notes that controlled human
exposure studies, including those newly available in this
reconsideration, demonstrate effects following short-term
PM2.5 exposures at concentrations higher than the current
24-hour standard. The Administrator also considers air quality analyses
conducted in the 2022 PA and in responding to public comments, as
described above in section II.B.3, that evaluate PM2.5
concentrations in ambient air for similar durations to the controlled
human exposure studies. As noted above, these air quality analyses
indicate that the current 24-hour standard, particularly in conjunction
with the revised level of the annual standard, provides a high degree
of protection against subdaily PM2.5 concentrations that
have been shown to elicit effects in controlled human exposure studies.
The Administrator considers a limited number of available epidemiologic
studies that report associations with health effects when the analyses
are restricted to daily PM2.5 concentrations below 35 [mu]g/
m\3\. As described above, although these studies are useful in
demonstrating that health effects are associated with exposure to daily
PM2.5 concentrations in the lower part of the air quality
distribution, they do not provide information about health effects
associated with the short-term
[[Page 16275]]
``peak'' exposures that the 24-hour standard is designed to protect
against. Accordingly, these studies have limited relevance in informing
a decision about the appropriate level of the 24-hour standard.
In addition to the scientific evidence, the Administrator also
considers the information from the risk assessment. In so doing, he
notes that the risk assessment estimates that the current primary
annual PM2.5 standard could allow a substantial number of
deaths in the U.S. With respect to the 24-hour standard, the
Administrator recognizes that there are only a small number of study
areas where the 24-hour standard is controlling and changes in the 24-
hour standard level are estimated to have a much smaller impact on
public health. The Administrator recognizes that while the risk
estimates can help to place the evidence for specific health effects
into a broader public health context, they should be considered along
with the inherent uncertainties and limitations of such analyses when
informing judgments about the potential for additional public health
protection associated with PM2.5 exposure and related health
effects. While the Administrator recognizes that these uncertainties
are important, he also notes that the general magnitude of the risk
estimates provide support for significant public health impacts,
particularly for lower alternative annual standard levels.
In reaching his final conclusions regarding the adequacy of the
primary PM2.5 standards, the Administrator also considers
the CASAC's advice and recommendations, as well as public comments.
With respect to the CASAC's advice, the Administrator recognizes that,
in their review of the 2021 draft PA, the CASAC reached consensus that
the current primary annual PM2.5 standard is not adequate
and that it is not sufficiently protective of public health. The
Administrator also takes note of the CASAC's advice in their review of
the 2019 draft PA, where the CASAC did not reach consensus on the
adequacy of the primary annual PM2.5 standard, with the
minority recommending revision and the majority recommending the
standard be retained. Furthermore, he recognizes that in reviewing the
2019 draft PA, the CASAC reached consensus regarding the adequacy of
the primary 24-hour PM2.5 standard, concluding that the
standard should be retained. Conversely, in their review of the 2021
draft PA, the majority of the CASAC advised that the current primary
24-hour PM2.5 standard is not adequate and recommended
revising the level of the standard, while the minority of the CASAC
concluded that the standard was adequate and should be retained.
However, in considering the advice of the CASAC collectively in the
context of this reconsideration, the Administrator recognizes that the
2021 draft PA included scientific evidence and quantitative risk
information that was not available in the 2019 draft PA, and therefore,
the advice and recommendations of the CASAC in their review of the 2021
draft PA are based on consideration of the full body of scientific
evidence available in this reconsideration, including the evidence
evaluated in the 2019 ISA and the ISA Supplement.
The Administrator recognizes that much of the scientific evidence
available in this reconsideration was also available in the 2019 ISA
and was considered by the then-Administrator when he decided that the
current primary PM2.5 standards are requisite to protect
public health with an adequate margin of safety. However, as described
in section I.C.5.b above, in reaching his decision to reconsider the
2020 final decision, the Administrator also recognized that there were
a number of studies published since the literature cutoff date of the
2019 ISA that were raised by some members of the CASAC in their review
of the 2019 draft PA, in public comments on the 2020 proposal, and in
the petitions for reconsideration. As such, the expansion of the air
quality criteria in this reconsideration to encompass both the 2019 ISA
and the additional scientific evidence evaluated in the ISA Supplement,
along with evidence and updated quantitative analyses in the 2022 PA
also provided an expanded record for the CASAC's review and public
comments as a part of this reconsideration. Taken together, the 2019
ISA, ISA Supplement, and 2022 PA, along with the CASAC's advice and
recommendations and public comments, provide the Administrator with
additional information for consideration in reaching his final
conclusions in this reconsideration. As a result, the record before him
notably expands upon and strengthens the basis for the conclusions of
the 2019 ISA while reducing some uncertainties that were identified in
the 2020 final action.
In considering the available information in this reconsideration,
the current Administrator reached different conclusions regarding the
appropriate weight to place on certain aspects of the evidence than the
then-Administrator in the 2020 final decision. For example, in reaching
his conclusions on the primary annual PM2.5 standard in
2020, the then-Administrator concluded that it was appropriate to place
more weight on epidemiologic studies that used ground-based monitors
and to place less weight on the studies that used hybrid model-based
approaches, citing to increased uncertainties associated with this new
and emerging approach to estimating exposure. In placing more weight on
the key U.S. monitor-based studies, the then-Administrator noted that
the majority of these studies had mean concentrations at or above the
level of the annual standard (12.0 [micro]g/m\3\). However, unlike the
approach for considering such studies in the 2012 review, the then-
Administrator concluded that it was appropriate to consider the study-
reported means collectively, and in so doing, he placed weight on the
average of the study-reported means (or medians) across the U.S.
monitor-based studies of 13.5 [micro]g/m\3\, and noted that this
concentration was above the level of the standard (85 FR 82717,
December 18, 2020). In contrast, in this reconsideration, the current
Administrator judges that it is appropriate to consider the individual
study-reported mean PM2.5 concentrations from not only the
U.S. monitor-based epidemiologic studies, but also the U.S. hybrid
model-based epidemiologic studies, which are an advancement in the
available science since the completion of the 2009 ISA. The current
Administrator also adopts an approach similar to some previous
approaches for the PM NAAQS in which he judges it most appropriate to
set the level of the standard to somewhat below the lowest long-term
study-reported mean PM2.5 concentration reported in key U.S.
epidemiologic studies, which is 9.3 [micro]g/m\3\. The study that
reports the long-term mean PM2.5 concentration of 9.3
[micro]g/m\3\ is newly available in this reconsideration and is
evaluated in the ISA Supplement. In the 2019 ISA, the lowest long-term
study-reported mean PM2.5 concentrations for U.S.-based
studies that use ground-based monitors and hybrid model-based
approaches are 9.9 [micro]g/m\3\ and 10.7 [micro]g/m\3\, respectively.
In judging that it is appropriate to consider both monitor- and hybrid
model-based epidemiologic studies and that it is appropriate to adopt
an approach to set the level of the standard to somewhat below the
lowest long-term mean PM2.5 concentration, the current
Administrator judges that the available scientific evidence--evaluated
in both the 2019 ISA and in the ISA Supplement--provide support for his
conclusion that that current primary
[[Page 16276]]
PM2.5 standard is not adequate and should be revised.
In addition to adopting a different approach than the previous
Administrator for considering the long-term mean PM2.5
concentrations from key U.S. epidemiologic studies (one more consistent
with the approach of the EPA in other prior reviews), the current
Administrator both has information newly available in this
reconsideration before him and is reaching different conclusions about
how to weigh the evidence before him in reaching his final conclusions.
For example, in reaching his final decision in 2020, the then-
Administrator was concerned about placing too much weight on
epidemiologic studies to inform his conclusions on the adequacy of the
primary PM2.5 standards, noting that the epidemiologic
studies do not identify particular PM2.5 concentrations that
cause effects and cannot alone identify a specific level at which to
set the standard. In so doing, the then-Administrator placed greater
weight on the uncertainties and limitations associated with the
epidemiologic studies, including exposure measurement error, potential
confounding by copollutants, increased uncertainty of associations at
lower PM2.5 concentrations, and heterogeneity of effects
across different cities or regions (85 FR 82716, December 18, 2020).
The Administrator recognizes that in reaching these judgments, the
then-Administrator took into consideration the views of some members of
the CASAC, who, in their advice on the 2019 draft PA, expressed the
view that the current PM NAAQS should be retained because reported
associations between short- and long-term PM2.5 exposures
and adverse health outcomes ``can reasonably be explained in light of
uncontrolled confounding and other potential sources of error and
bias'' (Cox, 2019b, p. 8 of consensus responses).
In this reconsideration, the current Administrator notes that the
ISA Supplement evaluates additional studies that employed statistical
approaches that attempted to more extensively account for confounders
and are more robust to model misspecification (i.e., used alternative
methods for confounder control, which are sometimes referred to as
causal modeling or causal inference methods) that build upon those
studies available and evaluated in the 2019 ISA (U.S. EPA, 2019,
sections 11.1.2.1 and 11.2.2.4). These studies report consistent
positive associations between long-term and short-term PM2.5
exposure and total mortality and cardiovascular effects (U.S. EPA,
2022a, section 3.2.2.3). In considering the epidemiologic evidence
evaluated in the 2019 ISA, along with the newly available studies
evaluated in the ISA Supplement, the current Administrator also
recognizes that there are uncertainties and limitations associated with
the epidemiologic studies, but judges that it is appropriate to place
less weight on these uncertainties than the then-Administrator placed
on them in reaching his final decision in 2020, given the strength of
the longstanding large body of epidemiologic evidence, employing a
variety of study designs, that demonstrates associations between long-
and short-term PM2.5 exposures and health effects across
multiple U.S. cities and in diverse populations, including in studies
examining populations and lifestages that may be at comparatively
higher risk of experiencing a PM2.5-related health effect
(e.g., older adults, children).
In reaching this final decision, the Administrator recognizes he is
differing not only with the prior Administrator but also with the
advice some members of the CASAC provided during their review of the
2019 draft PA. Specifically, taking into consideration the strength of
the evidence providing support for causality determinations, the advice
of other members of the CASAC and the need to protect public health
with an adequate margin of safety, the current Administrator disagrees
with these members of CASAC regarding the weight to be given to
epidemiologic evidence ``based on its methodological limitations''
(Cox, 2019b, p. 8 of consensus responses), such as the possibility
``that such associations could reasonably be explained by uncontrolled
confounding and other potential sources of error and bias'' (Cox,
2019b, p. 8 of consensus responses).
As another example of information that was not available to the
CASAC in providing advice to the Administrator in reaching his final
decision in 2020, the then-Administrator noted in his final decision
that, while some members of the CASAC and public commenters highlighted
a number of accountability studies that examined past reductions in
ambient PM2.5 concentrations and the degree to which those
reductions have resulted in public health improvements, the small
number of available accountability studies did not examine air quality
with starting concentrations meeting the primary annual
PM2.5 standard of 12.0 [micro]g/m\3\. The then-
Administrator took into consideration the absence of such
accountability studies, as part of his consideration of the full body
of scientific evidence, in reaching his judgment that there was
considerable uncertainty in the potential for increased public health
protection from further reductions in ambient PM2.5
concentrations beyond those achieved under the existing primary
PM2.5 NAAQS (85 FR 82717, December 18, 2020). However, there
are several accountability studies available since the literature
cutoff date of the 2019 ISA and evaluated in the ISA Supplement in this
reconsideration that have starting concentrations (or concentrations
prior to the policy or intervention) below 12.0 [micro]g/m\3\ (Corrigan
et al, 2018; Henneman et al., 2019; Sanders et al., 2020a). The current
Administrator concludes that, while the number of available
accountability studies is limited, he recognizes that these studies
provide supplemental information for consideration for informing
decisions on the appropriate level of the primary annual
PM2.5 standard along with the full body of evidence.
As EPA has frequently noted throughout this document, the extent to
which the current primary PM2.5 standards are judged to be
adequate depends in part on science policy and public health policy
judgments to be made by the Administrator on the strength and
uncertainties of the scientific evidence, such as how to consider
epidemiologic evidence and the need for an adequate margin of safety in
setting the standards. Thus, it would be pure speculation to guess
whether the then-Administrator would have reached the same or different
conclusions in the 2020 final decision had the record before him
included the newly available information in this reconsideration.\107\
However, the current Administrator concludes that, for the reasons
explained herein that, in his judgment, based on the record before him
in this reconsideration, it is necessary and appropriate to revise the
primary annual PM2.5 NAAQS to provide requisite protection
of public health with an adequate margin of safety.
---------------------------------------------------------------------------
\107\ The EPA notes that, in considering the additional
scientific evidence available in this reconsideration, one member of
the CASAC who reviewed both the 2019 draft PA and the 2021 draft PA
found that the available scientific and quantitative information
available in this reconsideration supported revising the level of
the primary annual PM2.5 standard to within the range of
10-11 [micro]g/m\3\, whereas he recommended retaining the standard
during the review of the 2019 draft PA.
---------------------------------------------------------------------------
Based on the available scientific evidence and quantitative
information, as well as consideration of the CASAC's advice and public
comments, the Administrator concludes that the
[[Page 16277]]
current primary annual PM2.5 standard is not adequate to
protect public health with an adequate margin of safety. In addition,
he finds the available information insufficient to call into question
the adequacy of the public health protection afforded by the current
primary 24-hour PM2.5 standard.
In considering how to revise the current suite of primary
PM2.5 standards in order to achieve the requisite protection
for public health, with an adequate margin of safety, against long- and
short-term PM2.5 exposures the Administrator considers the
four basic elements of the NAAQS (indicator, averaging time, form, and
level) collectively. With respect to indicator, the Administrator
recognizes that the scientific evidence in this reconsideration, as in
previous reviews, continues to provide strong support for health
effects associated with PM2.5 mass. He notes the 2022 PA
conclusion that the available information continues to support the
PM2.5 mass-based indicator and remains too limited to
support a distinct standard for any specific PM2.5 component
or group of components, and too limited to support a distinct standard
for the ultrafine fraction of PM (U.S. EPA, 2022b, section 3.6.3.2.1).
In its advice on the adequacy of the current primary PM2.5
standards in their review of the 2021 draft PA, the CASAC reached
consensus that the PM2.5 mass-based indicator should be
retained, without revision (Sheppard, 2022a, p. 2 of consensus
letter).\108\ Additionally, there was no information in the public
comments that provided a rationale for an alternative indicator. For
all of these reasons, the Administrator concludes that it is
appropriate to retain PM2.5 mass as the indicator for the
primary standards for fine particles.
---------------------------------------------------------------------------
\108\ The CASAC did not provide advice or recommendations
regarding the indicator of the primary PM2.5 standards in
their review of the 2019 draft PA (Cox, 2019b).
---------------------------------------------------------------------------
Consistent with his proposed conclusions regarding averaging time,
the Administrator notes that the scientific evidence continues to
provide strong support for health effect associations with both long-
and short-term PM2.5 exposures (88 FR 5618, January 27,
2023). Epidemiologic studies continue to provide strong support for
health effects associated with short-term PM2.5 exposures
based on 24-hour averaging periods, and associations in epidemiologic
studies with subdaily estimates are less consistent and, in some cases,
smaller in magnitude (88 FR 5618, January 27, 2023). Taken together,
the 2019 ISA concludes that epidemiologic studies do not indicate that
subdaily averaging periods are more closely associated with health
effects than the 24-hour average exposure metric (U.S. EPA, 2019a,
section 1.5.2.1). In addition, controlled human exposure and panel-
based studies of subdaily exposures typically examine subclinical
effects rather than the more serious population-level effects that have
been reported to be associated with 24-hour exposures (e.g., mortality,
hospitalizations). While recent controlled human exposure studies
provide consistent evidence for cardiovascular effects following
PM2.5 exposures for less than 24 hours (i.e., <30 minutes to
5 hours), air quality analyses have shown that the current averaging
times can effectively protect against the exposure concentrations in
these studies. This information does not indicate that a revision to
the averaging time is necessary to provide additional protection
against subdaily PM2.5 exposures, beyond that provided by
the current primary annual and 24-hour PM2.5 standards. The
Administrator also notes that this conclusion is also support by the
CASAC's advice in their review of the 2021 draft PA where they reached
consensus that averaging times for the primary PM2.5
standards should be retained, without revision (Sheppard, 2022a, p. 2
of consensus letter).\109\ The Administrator also considers the
relatively few public comments received that support a subdaily
averaging time, but concludes that the currently available information
does not provide support for an alternate averaging time. Consistent
with his proposed decision, the Administrator concludes that it is
appropriate to retain the annual and 24-hour averaging times for the
primary PM2.5 standards to protect against long- and short-
term PM2.5 exposures.
---------------------------------------------------------------------------
\109\ The CASAC did not provide advice or recommendations
regarding the averaging times of the primary PM2.5
standards in their review of the 2019 draft PA (Cox, 2019b).
---------------------------------------------------------------------------
With regard to form, the Administrator first notes that the EPA has
set both an annual standard and a 24-hour standard to provide
protection from health effects associated with both long- and short-
term exposures to PM2.5 (62 FR 38667, July 18, 1997; 88 FR
5620, January 27, 2023). With regard to the form of the annual
standard, the Administrator recognizes that a large majority of the
recently available epidemiologic studies continue to report
associations between health effects and annual average PM2.5
concentrations. These studies of annual average PM2.5
concentrations provide support for retaining the current form of the
annual standard to provide protection against long- and short-term
PM2.5 exposures. In its review of the 2021 draft PA, the
CASAC reached consensus that the form of the annual standard (i.e.,
annual mean, averaged over 3 years) should be retained, without
revision (Sheppard, 2022a, p. 2 of consensus letter).\110\ The
Administrator also notes that there were no public comments that
recommended an alternative form for the primary annual PM2.5
standard.
---------------------------------------------------------------------------
\110\ The CASAC did not provide advice or recommendations
regarding the forms of the primary PM2.5 standards in
their review of the 2019 draft PA (Cox, 2019b).
---------------------------------------------------------------------------
With regard to the form of the 24-hour standard (98th percentile,
averaged over three years), epidemiologic studies continue to provide
strong support for health effect associations with short-term (e.g.,
mostly 24-hour) PM2.5 exposures, and controlled human
exposure studies provide evidence for health effects following single
short-term ``peak'' PM2.5 exposures (88 FR 5618, January 27,
2023). Therefore, the Administrator concludes that the evidence
supports retaining a standard focused on providing supplemental
protection against short-term peak exposures and supports a 98th
percentile form for a 24-hour standard, in combination with a primary
annual PM2.5 standard with its annual mean averaged over
three years form. As described in the proposal and in responding to
comments in section II.B.3 above, the Administrator further notes that
the 98th percentile, averaged over three years, form also provides an
appropriate balance between limiting the occurrence of peak 24-hour
PM2.5 concentrations and identifying a stable target for
risk management programs (U.S. EPA, 2022b, section 3.6.3.2.3).
Furthermore, the Administrator notes that the multi-year percentile
form (i.e., averaged over three years) offers greater stability to the
air quality management process by reducing the possibility that
statistically unusual indicator values will lead to transient
violations of the standard. This conclusion is also supported by the
CASAC's advice in their review of the 2021 draft PA, where they reached
consensus that the form for the primary PM2.5 standards
should be retained, without revision (Sheppard, 2022a, p. 2 of
consensus letter).\111\
---------------------------------------------------------------------------
\111\ The CASAC did not provide advice or recommendations
regarding the forms of the primary PM2.5 standards in
their review of the 2019 draft PA (Cox, 2019b).
---------------------------------------------------------------------------
The Administrator also recognizes that the CASAC recommended that
in future reviews, the EPA also consider alternative forms for the
primary 24-hour PM2.5 standard (Sheppard, 2022a,
[[Page 16278]]
p. 18 of consensus responses). Based on the CASAC's advice, the
proposal solicited comment on alternatives to the current form for
consideration in future reviews (88 FR 5619, January 27, 2023). The
Administrator recognizes that there were a limited number of public
comments related to the form of the primary PM2.5 standards
as discussed in section II.D.3 above and in the Response to Comments
document, and notes that, the EPA will consider the information
provided by the commenters regarding the form of the 24-hour
PM2.5 standard in the next review of the PM NAAQS.
Consistent with his proposed decision, in considering the information
summarized above, the Administrator concludes that it is appropriate to
retain the forms of the current annual and 24-hour PM2.5
standards.
In considering how to revise the current suite of PM2.5
standards to provide the requisite public health protection with an
adequate margin of safety, the Administrator next evaluates the
appropriate levels of the primary PM2.5 standards, beginning
with the annual PM2.5 standard. In having carefully
considered public comments related to the primary annual
PM2.5 standard, the Administrator believes that the
fundamental conclusions regarding the scientific evidence and
quantitative information that supported his proposed conclusions (as
described in the 2019 ISA, ISA Supplement, 2022 PA, and the proposal)
remain valid. In considering the level at which the primary annual
PM2.5 standard should be set, the Administrator considers
the entire body of evidence and information, giving appropriate weight
to each part of that body of evidence and information. He continues to
place the greatest weight in this reconsideration on the available
scientific evidence that provides support for associations between
health effects and long- and short-term PM2.5 exposures. In
conjunction with his decisions to retain the current indicator,
averaging time, and form as described above, the Administrator is
revising the level of the primary annual PM2.5 standard to
9.0 [micro]g/m\3\. In so doing, he is selecting a primary annual
PM2.5 standard that, together with the primary 24-hour
PM2.5 standard, provides requisite public health protection
with an adequate margin of safety, based on his judgments about and
interpretation of the scientific evidence and quantitative risk
information.
The Administrator's decision to revise the level of the primary
annual PM2.5 standard to 9.0 [micro]g/m\3\ builds upon his
conclusion that the overall body of scientific evidence and
quantitative risk information calls into question the adequacy of
public health protection afforded by the current standard, particularly
for at-risk populations. Consistent with his consideration of the
available information in reaching his proposed decisions, the
Administrator's final decision on the level of the primary annual
PM2.5 standard places the greatest emphasis on key U.S.
epidemiologic studies that report associations between long- and short-
term PM2.5 exposures and mortality and morbidity. As in the
proposal, and as discussed further below, he views additional
epidemiologic studies (i.e., studies that employ alternative methods
for confounding control, studies that employ restricted analyses, and
accountability studies), the controlled human exposure studies, and the
risk assessment as providing supplemental information in support of his
decision to revise the current annual standard, but recognizes that
some of these lines of evidence and information provide a more limited
basis for selecting a particular standard level among a range of
options. See Mississippi, 744 F. 3d at 1351-52 (studies can
legitimately support a decision to revise the standard, but not provide
sufficient information to justify their use in setting the level of a
revised standard).
Given his consideration of the scientific evidence, quantitative
risk information, advice from the CASAC, and public comments, the
Administrator judges that a primary annual PM2.5 standard
with a level of 9.0 [micro]g/m\3\ is requisite to protect public health
with an adequate margin of safety. He notes that the determination of
what constitutes an adequate margin of safety is expressly left to the
judgment of the EPA Administrator. See Lead Industries Association v.
EPA, 647 F.2d at 1161-62; Mississippi, 744 F.3d at 1353. He further
notes that in evaluating how particular standards address the
requirement to provide an adequate margin of safety, it is appropriate
to consider such factors as the nature and severity of the health
effects, the size of the at-risk populations, and the kind and degree
of the uncertainties present. In considering the need for an adequate
margin of safety, the Administrator notes that a primary annual
PM2.5 standard with a level of 9.0 [micro]g/m\3\ would be
expected to provide substantial improvements in public health compared
to the current annual standard, including for at-risk groups such as
children, older adults, people with preexisting conditions, minority
populations, and low SES populations.
Consistent with his conclusions on the need for revision of the
current annual standard, in reaching a decision on level, the
Administrator places the most weight on information from epidemiologic
studies. In so doing, the Administrator notes that these studies
provide consistent evidence of positive and statistically significant
associations between long- and short-term exposure to PM2.5
and mortality and morbidity (88 FR 5624, January 27, 2023). The
Administrator recognizes that placing weight on the information from
the epidemiologic studies allows for examination of the entire
population, including those that may be at comparatively higher risk of
experiencing a PM2.5-related health effects (e.g., children,
older adults, minority populations) (88 FR 5624, January 27, 2023). The
Administrator also recognizes that recent epidemiologic studies
continue to support a no-threshold relationship, meaning that there is
no ``bright line'' below which no effects have been found. These
studies also support a linear relationship between health effects and
PM2.5 exposures at PM2.5 concentrations greater
than 8 [mu]g/m\3\, though uncertainties remain about the shape of the
C-R curve at PM2.5 concentrations less than 8 [mu]g/m\3\,
with some recent studies providing evidence for either a sublinear,
linear, or supralinear relationship at these lower concentrations (U.S.
EPA, 2019a, section 11.2.4; U.S. EPA, 2022a, section 2.2.3.2; 88 FR
5625, January 27, 2023).
As at the time of proposal, the Administrator notes that some
recent epidemiologic studies have adopted a broad range of approaches
to examine confounding and the results of those examinations support
the robustness of reported associations seen in epidemiologic studies.
These include studies that employ alternative methods for confounder
control and studies that evaluate the uncertainty related to exposure
measurement error, both of which continue to support associations
between PM2.5 exposures and health effects while taking
approaches to address uncertainties.
In considering the epidemiologic evidence, the Administrator judges
that, in reaching his decision on an appropriate level for the annual
standard that will protect public health with an adequate margin of
safety, in the absence of any discernible population-level thresholds,
and in recognizing the need to weigh uncertainties associated with the
epidemiologic evidence, it is most appropriate to examine where the
evidence of associations observed in the
[[Page 16279]]
epidemiologic studies is strongest and, conversely, to place less
weight where he has less confidence in the associations observed in the
epidemiologic studies. As at the time of proposal, the Administrator
notes that in previous reviews, evidence-based approaches noted that
the evidence of an association in any epidemiologic study is
``strongest at and around the long-term average where the data in the
study are most concentrated'' (78 FR 3140, January 15, 2013). Given
this, these approaches focused on identifying standard levels near or
somewhat below long-term mean concentrations reported in key
epidemiologic studies. These approaches were supported by previous
CASAC advice as well as the CASAC's advice in their review of the 2021
draft PA as a part of this reconsideration.
Additionally, the Administrator acknowledges that in the 2020 final
action, the then-Administrator decided to retain the standard based in
part on concerns about placing reliance on the epidemiologic studies
and his judgment that even if he did rely on them, the majority of the
studies had means or medians, as well as the mean of all of the key
study-reported means or medians, above the level of the current annual
standard. However, after considering the evidence, the advice of CASAC,
and public comments the Administrator judges that this approach is
insufficient to protect public health with an adequate margin of
safety. The Administrator's decision to reach a different judgment
about the appropriate level of the annual standard reflects the updated
and expanded scientific record available to the Administrator in this
reconsideration, as well as the additional advice from the CASAC and
the public comments based on this newly available information. In
addition, the Administrator observes the decision in this action to
place weight on the epidemiologic studies, and to revise the annual
primary standard to a level below the lowest long-term mean in the
U.S.-based epidemiologic studies, is consistent with the EPA's past
practice in PM NAAQS reviews.
In this reconsideration, the Administrator is considering the
scientific record which has been expanded and updated since the 2020
final action, as well as the additional advice from the CASAC and the
public comments that are based on the newly available information that
expands upon the information previously available. In addition, the
Administrator is exercising his judgment about how to interpret and
weigh the expanded evidence in a way that is more consistent with the
approaches used in prior PM NAAQS reviews. As a result, the
Administrator has concluded on reconsideration that the level of the
primary annual standard is not adequate and should be revised to
protect public health with an adequate margin of safety.
Consistent with his proposed decisions, in reaching conclusions on
the level of the primary annual PM2.5 standard, the
Administrator considers the long-term \112\ study-reported mean
PM2.5 concentrations from key long- and short-term
epidemiologic studies and sets the level of the standard to somewhat
below the lowest long-term mean PM2.5 concentration.\113\ He
notes that in previous PM NAAQS reviews (including the 1997, 2006 and
2012 reviews), evidence-based approaches focused on identifying
standard levels near or somewhat below long-term mean concentrations
reported in key long- and short-term epidemiologic studies. These
approaches were supported by the CASAC in previous reviews and were
supported in this reconsideration by the CASAC in their review of the
2021 draft PA. In considering the available scientific evidence to
inform such an approach, the Administrator notes the strength of the
epidemiologic evidence which includes multiple studies that
consistently report positive associations for short- and long-term
PM2.5 exposure and mortality and cardiovascular effects.
Some available studies also use a variety of statistical methods to
control for confounding bias and report similar associations, which
further supports the broader body of epidemiologic evidence for both
mortality and cardiovascular effects. Additionally, he notes that
recent epidemiologic studies available for consideration in reaching
his final decision strengthen support for health effect associations at
PM2.5 concentrations lower than in those evaluated in
epidemiologic studies available at the time of previous reviews. The
Administrator does recognize, however, that while these epidemiologic
studies evaluate associations between distributions of ambient
PM2.5 concentrations and health outcomes, they do not
identify the specific exposures that led to the reported effects. As
such, he notes that there is no specific point in the air quality
distribution of any epidemiologic study that represents a ``bright
line'' at and above which effects have been observed and below which
effects have not been observed. The Administrator further notes that
the epidemiologic studies provide the strongest support for reported
health effect associations for this middle portion of the
PM2.5 air quality distribution, which corresponds to the
bulk of the underlying data, rather than the extreme upper or lower
ends of the distribution, and concludes that the long-term study-
reported means from both long- and short-term studies provide the
strongest support for reported health effect associations in
epidemiologic studies. For these reasons, as described in the proposal
and in responding to public comments in section II.B.3 above, the
Administrator concludes that it is appropriate to continue to employ an
approach that focuses on the mean PM2.5 concentrations from
the key epidemiologic studies to inform his conclusions regarding the
appropriate level for the primary annual PM2.5 standard.
---------------------------------------------------------------------------
\112\ ``Long-term'' represents PM2.5 exposures and
concentrations that are annual or multi-year.
\113\ As described in section II.A.2.c above, key epidemiologic
studies are those that report overall mean (or median)
PM2.5 concentrations and for which the years of
PM2.5 air quality data used to estimate exposures overlap
entirely with the years during which health events are reported.
---------------------------------------------------------------------------
In adopting such an approach, the Administrator considers the long-
term mean concentrations reported in two types of key epidemiologic
studies: (1) Monitor-based studies \114\ (epidemiologic studies that
used ground-based monitors to estimate exposure, similar to approaches
used in past reviews), and (2) hybrid modeling-based studies \115\
(epidemiologic studies that used hybrid modeling approaches and apply
aspects of population weighting to estimate exposures). In reaching
conclusions regarding the level of a standard that would provide
requisite protection with an adequate margin of safety, the
Administrator recognizes that he must use his judgment regarding the
appropriate weight to place on the available evidence and technical
information, including uncertainties. As shown in Figures 1 and 2
above, for the key U.S. monitor-based epidemiologic studies,
[[Page 16280]]
the study-reported mean concentrations range from 9.9-16.5 [mu]g/m\3\,
and for the key U.S. hybrid modeling-based epidemiologic studies, the
mean concentrations range from 9.3-12.2 [mu]g/m\3\. The Administrator
also recognizes that, in their review of the 2021 draft PA, both the
majority and minority of the CASAC emphasized the epidemiologic studies
in support of their recommendations for the level of the annual
standard, but they weighed the studies in different ways (Sheppard,
2022a, p. 16-17 of consensus responses).
---------------------------------------------------------------------------
\114\ Reported mean PM2.5 concentrations in monitor-
based studies are averaged across monitors in each study area with
multiple monitors, referred to as a composite monitor concentration,
in contrast to the highest concentration monitored in the study
area, referred to as a maximum monitor concentration (i.e., the
``design value'' concentration), which is used to determine whether
an area meets a given standard.
\115\ Studies that use hybrid modeling approaches employ methods
to estimate ambient PM2.5 concentrations across large
geographical areas, including areas without monitors, and thus, when
compared to monitor-based studies, require additional information to
inform the relationship between the estimated PM2.5
concentrations across an area and the maximum monitor design values
used to assess compliance.
---------------------------------------------------------------------------
Based on this information, and in considering the CASAC's advice in
their review of the 2021 draft PA, the Administrator judges that it is
appropriate to set the level of the primary PM2.5 standard
at least as low as the lowest mean PM2.5 concentration from
these key U.S.-based epidemiologic studies, which is 9.3 [micro]g/m\3\.
The Administrator additionally notes that setting the annual standard
level at 9.0 [micro]g/m\3\, which is below the lowest study-reported
mean PM2.5 concentration of 9.3 [micro]g/m\3\, would be
expected to shift the distribution of PM2.5 concentrations
in an area such that the area's highest monitor would generally be at
or below 9.0 [micro]g/m\3\ annually, when meeting the annual standard.
In this situation, the resulting average or mean PM2.5
concentration for the entire area (measured across a number of
monitors) would be even further below the study-reported means,\116\
and will provide adequate protection not only in areas where the
highest allowable concentrations would be expected (i.e., near design
value monitors) but also in other parts of the area where
PM2.5 concentrations would be expected to be maintained even
lower.
---------------------------------------------------------------------------
\116\ Analyses in the 2022 PA suggest that the highest monitored
value would be expected to be greater than the study-reported mean
values by 10-20% for monitor-based studies and 15-18% for hybrid
modeling studies that apply aspects of population weighting.
---------------------------------------------------------------------------
As noted above, however, the Administrator must exercise his
judgment regarding the appropriate weight to place on the available
scientific evidence and quantitative information, including
uncertainties, in determining what level of the annual standard is
sufficient to protect public health with an adequate margin of safety.
In so doing, he considers other information available in this
reconsideration to inform his judgments, including study-reported
PM2.5 concentrations at lower percentiles in key
epidemiologic studies, supplemental information from other types of
epidemiologic studies, study-reported PM2.5 concentrations
from key Canadian epidemiologic studies, and the results from the
quantitative risk assessment.
In weighing the evidence in considering the requisite level of the
annual standard, the Administrator also takes into account additional
information from the key long- and short-term U.S. epidemiologic
studies available that provide study-reported PM2.5
concentrations below the mean and, in particular, the subset of
epidemiologic studies that report 25th and 10th percentile
concentrations. Consistent with his proposed conclusions, as well as
the CASAC's advice in their review of the 2021 draft PA and public
comments, the Administrator judges that it is appropriate to place some
weight on these lower percentiles in reaching his conclusions on the
level of the primary annual standard. There are six key U.S.
epidemiologic studies that report information on other percentiles
(e.g., 10th and 25th percentiles of PM2.5 concentrations or
10th and 25th percentiles of PM2.5 concentrations associated
with health events) that are below the mean.\117\ In considering the
information from these studies, the Administrator first notes that the
three older, monitor-based studies that report lower percentiles of
PM2.5 concentrations have smaller cohort sizes than the
three hybrid model-based studies. Thus, the Administrator recognizes
that the older, monitor-based studies had a relatively smaller portion
of the health events that were observed in the lower part of the air
quality distribution because of the generally smaller size of the
cohorts. He further notes that the recent hybrid model-based studies
have larger cohort sizes than the older, monitor-based studies, and
therefore, have more health events in the lower part of the air quality
distribution. Because of the larger cohort sizes and having a larger
portion of health events that are observed across the air quality
distribution, the Administrator has more confidence in the magnitude
and significance of the associations in the lower parts of the air
quality distribution for the recent, hybrid model-based studies
compared to the older, monitor-based studies. Given this, the
Administrator judges that it is appropriate to place weight on the 25th
percentile concentrations reported in the recently available hybrid
model-based studies in reaching his conclusions regarding the
appropriate level for the primary annual PM2.5 standard.
However, the Administrator also recognizes that his confidence in the
magnitude and significance in the reported concentrations, and their
ability to inform decisions on the appropriate level of the annual
standard, starts to diminish at percentiles that are even further below
the mean and the 25th percentile. For these reasons, the Administrator
places weight on the reported 25th percentiles concentrations in the
recent hybrid model-based studies, rather than the reported 10th
percentile concentrations, in reaching his conclusions regarding the
appropriate level for the primary annual PM2.5 standard.
---------------------------------------------------------------------------
\117\ The Wang et al. (2017) study only reports the 25th
percentile of the estimated PM2.5 concentrations, not the
10th percentile.
---------------------------------------------------------------------------
In considering the information from these studies, as described in
section II.A.2.c and in responding to public comments in section II.B.3
above, the Administrator notes that there are two hybrid model-based
studies with large cohort sizes that apply population weighting and
report lower percentile values. These studies are Di et al. (2017b) and
Wang et al. (2017) and the reported 25th percentile concentration is
9.1 [micro]g/m\3\ for both studies.\118\ In considering these studies,
the Administrator concludes that it is appropriate to place weight on
the 25th percentile concentrations of these newer hybrid model-based
studies (of 9.1 [micro]g/m\3\) such that setting the level of the
standard near these 25th percentile concentrations would provide
requisite protection. The Administrator observes that an annual
standard level of 9.0 [micro]g/m\3\ would be near the reported 25th
percentile concentrations in these studies.
---------------------------------------------------------------------------
\118\ There is a third hybrid model-based study, as described in
the 2022 PA and in section II.B.3 above in responding to public
comments, but it is not referenced here because it reports a 25th
percentile PM2.5 concentration based on the 25th
percentile of health events that occur in the study (Di et al.,
2017a) rather than report the 25th percentile based on air quality
concentrations.
---------------------------------------------------------------------------
As at the time of proposal, the Administrator also takes note of
the study-reported long-term mean PM2.5 concentrations in
long- and short-term Canadian epidemiologic studies, which ranged from
6.9 to 13.3 [micro]g/m\3\ for monitor-based studies and 5.9 to 9.8
[micro]g/m\3\ for hybrid model-based studies. While the Administrator
notes that these studies provide additional support for associations
between PM2.5 concentrations and health effects, he is also
mindful that there are important differences between the exposure
environments in the U.S. and Canada and that interpreting the data
(e.g., study-reported mean concentrations)
[[Page 16281]]
from the Canadian studies in the context of a U.S.-based standard may
present challenges in directly and quantitatively informing decisions
regarding potential alternative levels of the annual standard. For
example, in terms of people per square kilometer, the U.S. population
density is nearly 10 times in the contiguous U.S. compared to Canada.
As described in more detail in responding to public comments in section
II.B.3 above, in this reconsideration, the Administrator recognizes
that this difference in population density between the U.S. and Canada
is more apparent than in previous reviews because the studies available
in this reconsideration use different approaches than those previously
available. In the 2012 review, the available Canadian epidemiologic
studies used population-weighting and focused on urban areas where
monitors were available and population densities were more comparable
with those in the U.S., and at that time, the U.S. and Canadian studies
reported similar mean PM2.5 concentrations. However, in this
reconsideration, the Administrator takes note that for the new Canadian
epidemiologic studies: (1) The Canadian monitor-based studies available
in this reconsideration do not apply population weighting as the
previously available studies did; and (2) some of the studies now use
hybrid modeling approaches for estimating exposure. The Administrator
recognizes that these differences are important to consider in reaching
conclusions on how these Canadian epidemiologic studies should be
interpreted regarding decisions on the requisite level of the primary
annual PM2.5 standard. Specifically, the Administrator notes
that the more recent Canadian studies that use hybrid modeling
incorporate larger portions of the country, and therefore include more
rural areas. The more rural areas that are included in the study using
the hybrid modeling approaches, the more important it is to consider
how the population densities and exposure environments differ between
the U.S. and Canada. Additionally, the Administrator notes that for
hybrid modeling-based studies there is less certainty in
PM2.5 exposure estimates in more rural areas, which are
further from air quality monitors and where PM2.5
concentrations in the ambient air tend to be lower. For these hybrid
model-based studies, the portion of the rural areas that are
contributing to the study-reported mean PM2.5 concentrations
in these studies is unclear. For these reasons, the Administrator
concludes that it is important to consider the differences between the
population exposures in the U.S. and Canadian study areas and how these
differences influence the interpretation of the epidemiologic study
results.
Thus, the Administrator considers the Canadian studies to inform
his judgments on what level for the annual standard is requisite in
light of the limitations and challenges presented. The Administrator
also recognizes that the majority of the CASAC in their review of the
2021 draft PA, as well as a number of public commenters, place weight
on the Canadian epidemiologic studies in recommending that the level of
the primary annual PM2.5 standard be revised to 8-10
[micro]g/m\3\. The Administrator further notes while the majority of
the CASAC advised the EPA to consider the Canadian studies in revising
the annual standard level to within the range of 8.0-10.0 [micro]g/
m\3\, they did not advise the EPA to set the annual standard level
below the study-reported means from those studies. Given these
considerations, the Administrator judges that it is appropriate to set
the level of annual standard within the range of 8-10 [micro]g/m\3\ to
be consistent with the majority of the CASAC's advice in their
consideration of these studies.
The Administrator also recognizes that information from
epidemiologic studies that included analyses that restrict annual
average PM2.5 concentrations to concentrations below the
level of the current annual standard can be useful for informing
conclusions regarding the appropriate level of the primary annual
PM2.5 standard. In so doing, he particularly notes the two
key U.S. epidemiologic studies (Di et al., 2017b and Dominici et al.,
2019) that restrict annual average PM2.5 concentrations to
less than 12 [micro]g/m\3\ and report positive and statistically
significant associations with all-cause mortality and mean
PM2.5 concentrations of 9.6 [micro]g/m\3\. He also considers
these results along with the uncertainties and limitations associated
with studies that restricted analyses below certain PM2.5
concentrations. As described in responding to comments in section
II.B.3 above, uncertainties associated with how the studies exclude
PM2.5 concentrations from the analyses (e.g., at what
spatial resolution are concentrations being excluded), make it
difficult to understand how to interpret the results of the restricted
analyses in the context of the approach employed in this
reconsideration, which takes into consideration the relationship
between mean PM2.5 concentrations and design values.
The Administrator also recognizes that, in their review of the 2021
draft PA, the CASAC noted that epidemiologic studies that restrict
analyses below certain PM2.5 concentrations represent one
area for which the evidence has expanded in this reconsideration,
stating that these studies provide support for mortality effects at
concentrations below the current PM NAAQS (Sheppard, 2022a, p. 5 of
consensus responses). In their recommendations on alternative levels
for the primary annual PM2.5 standard, the majority of the
CASAC cited to studies that restrict PM2.5 concentrations to
below 12 [micro]g/m\3\ as a part of their rationale for supporting a
level within the range of 8-10 [micro]g/m\3\ (Sheppard, 2022a p. 16 of
consensus responses). Additionally, the Administrator notes that some
members of the CASAC, in their review of the 2019 draft PA, concluded
that the epidemiologic studies that restrict analyses below 12
[micro]g/m\3\ and show positive associations with health effects, along
with other aspects of the scientific evidence, provide support for
their conclusion that the primary annual PM2.5 standard is
not adequate (Cox, 2019b, p. 9 of consensus responses). Furthermore,
the Administrator takes note of public commenters who also noted that
the epidemiologic studies that restrict PM2.5 concentrations
to below the current standard provide support, along with the other
available information, for lowering the level of the primary annual
PM2.5 standard. In considering the studies that include
restricted analyses, along with the CASAC's advice and public comments
on these types of studies, the Administrator concludes that, although
there are inherent uncertainties associated with this limited body of
evidence, these studies that apply restricted analyses provide support
for serious effects (e.g., mortality) at concentrations below 10.0
[micro]g/m\3\. Given this, the Administrator concludes that it is
appropriate to place some weight on these studies, and in doing so,
notes that a standard level of 9.0 [micro]g/m\3\ would be below the
reported mean PM2.5 concentrations of 9.6 [micro]g/m\3\ in
these studies and would, thus, be expected to provide protection
against exposures related to these reported mean concentrations.
The Administrator also takes into consideration recent U.S.
accountability studies, which assess the health effects associated with
actions that improve air quality (e.g., air quality policies or
implementation of an intervention). These types of studies can also
reduce uncertainties related to residual confounding of temporal and
spatial factors (U.S. EPA, 2022a, p. 3-25). The
[[Page 16282]]
Administrator notes that in the 2020 review, the available
accountability studies had ``starting'' annual average PM2.5
concentrations (i.e., mean concentration prior to reductions being
evaluated) from 13.2-31.5 [micro]g/m\3\, and the then-Administrator
cited the lack of accountability studies in areas where the
``starting'' concentration met the current primary PM2.5
standards as part of his rationale for retaining the standards. As at
the time of proposal, the current Administrator notes that in three
studies newly available in this reconsideration and assessed in the ISA
Supplement, prior to implementation of the policies, mean
PM2.5 concentrations in these studies were below the level
of the current annual standard level (12.0 [micro]g/m\3\) and ranged
from 10.0 [micro]g/m\3\ to 11.1 [micro]g/m\3\. These studies report
positive and significant associations between mortality and
cardiovascular morbidity and reductions in ambient PM2.5
following the implementation of a policy (Henneman et al., 2019;
Corrigan et al., 2018; Sanders et al., 2020a; 88 FR 5627, January 27,
2023). These studies suggest that public health improvements may occur
following the implementation of a policy that reduces annual average
PM2.5 concentrations below the level of the current standard
of 12.0 [micro]g/m\3\. The Administrator recognizes that in their
review of the 2021 draft PA, the CASAC noted that the availability of
recent accountability studies was one area where the evidence had been
strengthened and that the studies assessed in the ISA Supplement
provide evidence of mortality effects at annual average
PM2.5 concentrations below the current NAAQS (Sheppard,
2022a, p. 5 of consensus responses). The Administrator recognizes that
the CASAC also concluded that, along with other lines of evidence, the
accountability studies with starting concentrations below the levels of
the current standards are appropriate to consider for informing
conclusions on alternative standard levels (Sheppard, 2022a, p. 13 of
consensus responses). The Administrator also notes the advice of the
CASAC in their review of the 2019 draft ISA, where they suggested that
accountability studies be taken into account and such studies provide
potentially crucial information about whether and how much decreasing
PM2.5 causes decreases in future health effects, which
reflects the primary purpose of the NAAQS (Cox, 2019b, p. 8 and 10 of
consensus responses). The Administrator also notes that in their review
of the 2019 draft ISA, some members of the CASAC cautioned against
placing more weight on the data from accountability studies based on
the methodological limitations of the studies (Cox, 2019b, p. 8 of
consensus responses). The Administrator notes that the CASAC did not
explicitly cite to accountability studies in their reviews of the 2019
draft PA or 2021 draft PA as support for their recommendations on the
adequacy of the primary annual PM2.5 standard or potential
alternative standard levels. A number of public commenters who support
revising the level of the standard to 8 [micro]g/m\3\ cite these
accountability studies, along with the broader evidence base, as
support for a more protective standard. The Administrator, in
considering the evidence, the advice from the CASAC, and public
comment, first recognizes that accountability studies are just one line
of evidence to be considered in the broader evaluations of the
information available to inform conclusions on the level of the
standard. In so doing, he notes that public health improvements may
occur following the implementation of a policy that reduces annual
average PM2.5 concentrations below the level of the current
standard of 12.0 [micro]g/m\3\, and potentially below the lowest
``starting'' concentrations in these studies of 10.0 [micro]g/m\3\.
However, the Administrator concludes that the limited number of
accountability studies provide limited information for informing
decisions on the appropriate level of the primary annual
PM2.5 standard but recognizes that these studies provide
supplemental information for consideration along with the full body of
evidence. Taken together, the Administrator notes a revised annual
standard level of 9.0 [micro]g/m\3\ is at or below the lowest starting
concentration of these accountability studies (i.e., 10.0 [micro]g/
m\3\), and judges that it is appropriate to place some weight on these
studies, particularly for informing his public policy judgments
regarding an adequate margin of safety.
In addition to his consideration of and conclusions regarding the
available scientific evidence, the Administrator also considers the
results of the quantitative risk assessment to inform his conclusions
regarding the appropriate level for the primary annual PM2.5
standard. The Administrator recognizes that the risk estimates can help
to place the evidence for specific health effects into a broader public
health context, but should be considered along with the inherent
uncertainties and limitations of such analyses when informing judgments
about the potential for additional public health protection associated
with PM2.5 exposure and related health effects. The
Administrator recognizes that the overall risk assessment estimates
suggest that the current primary annual PM2.5 standard could
allow a substantial number of PM2.5-associated deaths in the
U.S. The Administrator also recognizes that the CASAC concurred with
the 2021 draft PA's assessment that meaningful risk reductions will
result from lowering the annual PM2.5 standard (Sheppard,
2022a, p. 16 of consensus responses).
Additionally, with respect to the results of the quantitative risk
assessment, the Administrator recognizes that the 2022 PA also provides
information on the distribution of concentrations associated with the
estimated mortality risk at each alternative standard level assessed
(U.S. EPA, 2022b, sections 3.4.2.2 and 3.6.2.2, Figure 3-18 and 3-19).
When meeting an annual standard of 9.0 [micro]g/m\3\ at the design
value monitor, the exposure concentrations within an area are estimated
to be below 9 [micro]g/m\3\, with the majority of those exposures being
at concentrations of below 8 [micro]g/m\3\. The Administrator notes
that this range of concentrations is below the lowest means in the key
long- and short-term epidemiologic studies (concentrations at which the
evidence is the strongest in supporting an association between exposure
to PM2.5 and adverse health effects observed in the key
epidemiologic studies available in this reconsideration). Thus, the
Administrator concludes that the results of the quantitative risk
assessment suggest that a revised annual standard level of 9.0
[micro]g/m\3\ is estimated to reduce PM2.5 exposures to fall
within the range of concentrations in which there is the most
confidence in the associations and thus, confidence that estimated risk
reductions will actually occur.
The Administrator also notes the information provided by the
quantitative risk assessment on the distribution of concentrations
associated with the estimated mortality risk for a higher annual
standard level of 10.0 [micro]g/m\3\ and a lower standard level of 8.0
[micro]g/m\3\ (U.S. EPA, 2022b, sections 3.4.2.2 and 3.6.2.2, Figure 3-
18 and 3-19). The Administrator finds that, for an annual standard
level of 10.0 [micro]g/m\3\, the quantitative risk assessment estimates
that the standard would allow multiple exposures at concentrations
above the lowest means in the key epidemiologic studies, and therefore,
calls into question whether a standard level of 10.0 [micro]g/m\3\
would provide enough public health protection. Additionally, the
Administrator also finds that, for a lower annual standard level of 8.0
[micro]g/
[[Page 16283]]
m\3\, the quantitative risk assessment estimates the exposure
concentrations to be below 8 [micro]g/m\3\, with the majority of those
exposures being at concentrations of below 7 [micro]g/m\3\. The
Administrator observes that the majority of exposure concentrations
under this air quality scenario are estimated to fall outside of the
range of concentrations in which he has the most confidence in the
associations and that the additional risk reductions will actually
occur.
Recognizing and building upon the above considerations and
judgments, and with consideration of advice from the CASAC and public
comment, the Administrator concludes that the current body of
scientific evidence and quantitative risk assessment support his
judgment that the level of the primary annual PM2.5 standard
should be revised to a level of 9.0 [micro]g/m\3\. Revising the level
of the primary annual PM2.5 standard will, in the
Administrator's judgment, provide requisite public health protection
with an adequate margin of safety.
The Administrator recognizes that placing weight on the information
from the epidemiologic studies allows for examination of the entire
population, including those that may be at comparatively higher risk of
experiencing a PM2.5-related health effects (e.g., children,
older adults, minority populations) (88 FR 5624, January 27, 2023). In
considering the epidemiologic evidence, the Administrator judges that,
in reaching his decision on an appropriate level for the annual
standard that will protect public health with an adequate margin of
safety, in the absence of any discernible population-level thresholds,
and in recognizing the need to weigh uncertainties associated with the
epidemiologic evidence, it is most appropriate to examine where the
evidence of associations observed in the epidemiologic studies is
strongest and, conversely, to place less weight where he has less
confidence in the associations observed in the epidemiologic studies.
The Administrator notes that in previous reviews, evidence-based
approaches noted that the evidence of an association in any
epidemiologic study is ``strongest at and around the long-term average
where the data in the study are most concentrated'' (78 FR 3140,
January 15, 2013). These approaches were supported by previous CASAC
advice as well as the CASAC's advice in their review of the 2021 draft
PA as a part of this reconsideration. Given this, the Administrator
notes that in revising the annual PM2.5 standard to a level
of 9.0 [micro]g/m\3\, he is setting the standard at a level below the
long-term mean PM2.5 concentrations in the key long- and
short-term epidemiologic studies, including the lowest study reported
mean of 9.3 [micro]g/m\3\, following an approach that is consistent
with previous PM NAAQS reviews. The Administrator additionally notes
that air quality analyses in the 2022 PA demonstrate that areas meeting
a revised annual standard of 9.0 [micro]g/m\3\ would be expected to
shift the distribution of PM2.5 exposure concentrations in
an area such that the area's highest monitor would generally be at or
below 9.0 [micro]g/m\3\ annually, and most of the resulting
PM2.5 concentrations across the area would be even further
below the study-reported means.119 120 Thus, a standard
level of 9.0 [micro]g/m\3\ is expected to provide sufficient protection
not only in areas where the highest allowable concentration would be
located (i.e., near design value monitors) but also in other parts of
the area where PM2.5 concentrations would be expected to be
maintained even lower.
---------------------------------------------------------------------------
\119\ Analyses in the 2022 PA suggest that the highest monitored
value would be expected to be greater than the study-reported mean
values by 10-20% for monitor-based studies and 15-18% for hybrid
modeling studies that apply aspects of population weighting (U.S.
EPA, 2022b, section 2.3.3.2.4).
\120\ The risk assessment in the 2022 PA used air quality
adjustments to simulate just meeting the current primary
PM2.5 standards, as well as alternative standard levels
(U.S. EPA, 2022b, section 3.4.1.4 and Appendix C, section C.1.4).
---------------------------------------------------------------------------
Furthermore, the Administrator recognizes the CASAC's advice in
their review of the 2021 draft PA, as well as public comments, that
weight should be placed on study-reported PM2.5
concentrations that are somewhat below the mean, particularly for some
of the newer epidemiologic studies with larger cohort sizes. In
weighing uncertainties associated with using these data to inform a
revised annual standard level, as well as noting the limited studies
for which this information is available, the Administrator judges that
some weight should be placed on these data, but they should not receive
the same weight as the study-reported mean concentrations. Thus, the
Administrator concludes that it would be appropriate to set the annual
standard level near the 25th percentile PM2.5 concentrations
in the two newer key epidemiologic studies for which these values were
reported. In doing so, the Administrator notes that a decision to
revise the annual standard to 9.0 [micro]g/m\3\ would set a level of
the standard near and somewhat below the reported 25th percentile
PM2.5 concentrations of 9.1 [micro]g/m\3\ in these two more
recent hybrid model-based studies.
The Administrator also takes note of the study-reported long-term
mean PM2.5 concentrations in the key Canadian epidemiologic
studies. While the Administrator notes that these studies provide
additional support for associations between PM2.5
concentrations and health effects, he is also mindful that there are
important differences between the exposure environments in the U.S. and
Canada that affect interpretation of the data in the context of
informing decisions regarding potential alternative levels of the
annual standard. The Administrator also recognizes that the majority of
the CASAC in their review of the 2021 draft PA, as well as a number of
public commenters, placed weight on the Canadian epidemiologic studies
in recommending that the level of the primary annual PM2.5
standard be revised to 8-10 [micro]g/m\3\. The Administrator notes that
a decision to revise the annual standard to 9.0 [micro]g/m\3\ would set
the level of the standard within the range of levels recommended by the
majority of CASAC in their consideration of these studies.
Additionally, the Administrator also considers the information
provided by epidemiologic studies that use restricted analyses, as well
as accountability studies. With respect to the restricted analyses, the
Administrator, in considering the CASAC's advice in their review of the
2021 draft PA and many public comments on these types of studies,
concludes that, although there are inherent uncertainties associated
with this limited body of evidence, the studies that apply restricted
analyses provide support for serious effects (e.g., mortality) at
concentrations below 10.0 [micro]g/m\3\. Additionally, in considering
accountability studies, the Administrator concludes that while the
small number of these studies provide limited information for informing
decisions on the appropriate level of the primary annual
PM2.5 standard, these studies provide supplemental
information for consideration along with the full body of evidence. The
Administrator further notes that these studies suggest that public
health improvements may occur following the implementation of a policy
that reduces annual average PM2.5 concentrations below the
level of the current standard of 12.0 [micro]g/m\3\, and potentially
below the lowest ``starting'' concentrations in these studies of 10.0
[micro]g/m\3\. Taken together, the Administrator judges that it is
appropriate to place some weight on these types of studies,
particularly for informing his public policy judgments regarding an
adequate margin
[[Page 16284]]
of safety, and notes that a revised annual standard level of 9.0
[micro]g/m\3\ is below the lowest starting concentration of the
accountability studies (i.e., 10.0 [micro]g/m\3\), and below the
concentration at which studies that apply restricted analyses provide
support for serious effects (i.e., 9.6 [micro]g/m\3\).
The Administrator also judges that the results of the quantitative
risk assessment provide support for a primary annual PM2.5
standard with a level of 9.0 [micro]g/m\3\. The results of the risk
assessment suggest that when meeting an annual standard of 9.0
[micro]g/m\3\, PM2.5 exposures are maintained below 9
[micro]g/m\3\ at the design value monitor, with the majority of those
exposures being at concentrations below 8 [micro]g/m\3\. Thus, the
Administrator notes that an annual standard level of 9.0 [micro]g/m\3\
would be expected to provide protection from exposures where he has the
greatest confidence in the associations between health effects and
PM2.5 exposures (i.e. the long-term mean PM2.5
concentrations in the key U.S. epidemiologic studies, of which the
lowest is 9.3 [micro]g/m\3\) and would provide an adequate margin of
safety by maintaining most PM2.5 exposures even further
below 9.0 [micro]g/m\3\.
When considering adequate margin of safety, the Administrator notes
that in his decision to revise the annual standard level to 9.0
[micro]g/m\3\, he is placing weight on the information from the
epidemiologic studies which allows for examination of the entire
population, including those that may be at comparatively higher risk of
experiencing a PM2.5-related health effects (e.g., children,
older adults, minority populations). Additionally, as discussed above,
the Administrator also recognizes that setting the annual standard
level at 9.0 [micro]g/m\3\, which is below concentrations at which the
evidence is the strongest in supporting an association between exposure
to PM2.5 and adverse health effects observed in the key
epidemiologic studies available in this reconsideration, would be
expected to shift the distribution of PM2.5 exposure
concentrations in an area such that the area's highest monitor would
generally be at or below 9.0 [micro]g/m\3\ annually, and most of the
resulting PM2.5 concentrations across the area would be even
lower. In considering these air quality relationships, the
Administrator judges that a revised annual standard level of 9.0
[micro]g/m\3\ would provide requisite protection with adequate margin
of safety, for all populations, including those most at-risk.
In reaching this conclusion, the Administrator recognizes that in
establishing primary standards under the Act that are requisite to
protect public health with an adequate margin of safety, he is seeking
to establish standards that are neither more nor less stringent than
necessary for this purpose. The Act does not require that primary
standards be set at a zero-risk level or to protect the most sensitive
individual, but rather at a level that avoids unacceptable risks to
public health. In this context, the Administrator's conclusion is that
revised primary annual standard, in conjunction with the 24-hour
standard, provides the appropriate degree of protection, and that more
or less stringent standards would not be requisite.
In considering the requirement for an adequate margin of safety,
the Administrator notes that the determination of what constitutes an
adequate margin of safety is expressly left to the judgment of the EPA
Administrator. See Lead Industries Association v. EPA, 647 F.2d at
1161-62; Mississippi, 744 F.3d at 1353. He further notes that in
evaluating how particular standards address the requirement to provide
an adequate margin of safety, it is appropriate to consider such
factors as the nature and severity of the health effects, the size of
sensitive population(s) at risk, and the kind and degree of the
uncertainties present. Consistent with past practice and long-standing
judicial precedent, and as described in this section, the Administrator
takes the need for an adequate margin of safety into account as an
integral part of his decision making on a standard. See, e.g., NRDC v.
EPA, 902 F. 2d 962, 973-74 (D.C. Cir. 1990).
Given all of the evidence and information discussed above, the
Administrator judges that a standard with a level of 9.0 [micro]g/m\3\
is requisite to protect public health with an adequate margin of
safety. In so doing, he first recognizes that a less stringent standard
would allow the occurrence of higher long- and short-term
PM2.5 concentrations at a level at or above the mean
PM2.5 concentrations in key U.S. epidemiologic studies. That
is, a less stringent standard would be expected to allow more
PM2.5 exposures at concentrations at or above which the key
U.S. epidemiologic studies have reported associations between mean
PM2.5 concentrations and serious health effects and would
deviate from some past approaches for selecting the appropriate level
of the annual standard. A less stringent standard would also not
provide requisite protection with an adequate margin of safety against
PM2.5 exposures in the lower percentiles of the air quality
distribution (i.e., 25th percentile) for which associations with health
effects have been observed in a limited number of epidemiologic
studies. Furthermore, the Administrator notes that the primary annual
and 24-hour PM2.5 standards, together, are intended to
provide public health protection against the full distribution of long-
and short-term PM2.5 exposures. As noted above, the
Administrator recognizes that the changes in PM2.5 air
quality designed to meet a less stringent annual standard would likely
result in higher exposures across the distribution of air quality,
including both higher average (or typical) concentrations as well as
higher short-term peak PM2.5 concentrations. Taking into
consideration both the full evidence base for associations of
PM2.5 with mortality and other adverse health effects,
including the reported mean PM2.5 concentrations from key
long- and short-term U.S. epidemiologic studies, information from
epidemiologic studies that report 25th percentile PM2.5
concentrations, supplemental information from other epidemiologic
studies (i.e., epidemiologic studies that use restricted analyses,
accountability studies, and Canadian epidemiologic studies), and the
results of the risk assessment, as well as the advice from the CASAC
and public comments, the Administrator concludes that a less stringent
standard would allow risks of mortality and other adverse health
effects that are too great, and thus would not provide sufficient
protection for public health as required by the CAA.
Additionally, in considering a less stringent standard, the
Administrator recognizes that through its control of long- and short-
term PM2.5 concentrations, the annual standard provides a
margin of safety for less well-studied exposure levels and population
groups for which the evidence is limited or lacking. In so doing, he
recognizes that our understanding of the relationships between the
presence of a pollutant in ambient air and associated health effects is
based on a broad body of information encompassing not only more
established aspects of the evidence, such as the conclusion that long-
and short-term exposures to PM2.5 are causally related to
mortality and cardiovascular effects and likely to be causally related
to respiratory effects, but also aspects with which there may be
substantial uncertainty. In particular, the Administrator notes that
there are other categories of effects with causality determinations
that are suggestive of, but not sufficient to infer, a causal
[[Page 16285]]
relationship between PM2.5 exposure and health outcomes.
These include, but are not limited t,o short-term exposure and nervous
system effects, as well as long- and short-term exposure and pregnancy
and birth outcomes, where the evidence is less certain but which
represent potentially substantial additional risk to public health from
exposure to PM2.5. He recognizes the CAA requirement that
requires primary standards to provide an adequate margin of safety was
intended to address uncertainties associated with inconclusive
scientific and technical information as well as to provide a reasonable
degree of protection against hazards that research has not yet
identified and in his judgment, the primary NAAQS must be set at a
level that is adequately protective against these and other effects
which research has not yet identified. Thus, even if the Administrator
had somewhat greater concerns about the possibility of confounding,
error and bias in the epidemiologic studies, which reduced his
confidence in finding that PM2.5 is causally related to
mortality and cardiovascular effects, he would still find it
appropriate to set the primary NAAQS below the means of key U.S.
epidemiologic studies given the strength of the evidence providing
support for the association, as well as additional evidence linking
PM2.5 to other endpoints of substantial public health
concern, and the need to protect public health with an adequate margin
of safety. In considering the uncertainties in both the epidemiologic
evidence and the controlled human exposures studies, the Administrator
recognizes that collectively, the health effects evidence generally
reflects a continuum, consisting of levels at which scientists
generally agree that health effects are likely to occur, through lower
levels at which the likelihood and magnitude of the response become
increasingly uncertain. In light of these uncertainties, the
Administrator recognizes that the CAA requirement that primary
standards provide an adequate margin of safety, as summarized in
section I.A above, is intended to address uncertainties associated with
inconclusive scientific and technical information, as well as to
provide a reasonable degree of protection against hazards that research
has not yet identified. The Administrator has taken the need to provide
for an adequate margin of safety into account as an integral part of
his decision-making on the appropriate standards in setting the
standard at a level below the level where available epidemiologic
studies, which include diverse populations that are broadly
representative of the U.S. population including at-risk populations,
have provided the strongest evidence supporting effects, and in other
ways as well. For example, consideration of a margin of safety is
reflected in the approach of setting the level of the annual standard
near and somewhat below the 25th percentile PM2.5
concentrations from key U.S. epidemiologic studies (i.e., 9.1 [micro]g/
m\3\), as well as recognition that attaining a design value will
generally result in significantly broader and greater improvements of
air quality across an area (including but certainly not limited to
areas near the design value monitor) (U.S. EPA, 2022a, sections
2.3.3.2.4 and 3.3.3.2.1, Table 3-5). Based on all of the considerations
noted here, and considering the current body of evidence, including the
associated limitations and uncertainties, in combination with the
exposure/risk information, the Administrator concludes that a less
stringent standard than the current standard would not provide the
requisite protection of public health, including an adequate margin of
safety.
Having concluded that a less stringent standard would not provide
the requisite protection of public health, the Administrator next
considers whether a more stringent standard would be appropriate. In so
doing, he notes that a decision to set the level of the annual standard
to below 9.0 [micro]g/m\3\ would place a large amount of the emphasis
on potential public health importance of further reducing the
occurrence of PM2.5 concentrations of concern, though the
exposures about which he is most concerned are well controlled with an
annual standard level of 9.0 [micro]g/m\3\, as demonstrated by the
quantitative risk assessment. Such a decision would also place greater
weight on (1) further reducing ambient PM2.5 concentrations
relative to those observed in long-and short-term epidemiologic
studies, including those that he had judged to have significant
uncertainties, including Canadian studies, studies using restricted
analyses, and accountability studies; (2) shifting the air quality
distribution in areas such that the highest exposure concentrations are
reduced to below PM2.5 concentrations observed in
epidemiologic studies to be in the 25th or lower percentile, for which
the evidence is limited; and (3) further shifting exposure
concentrations to those shown at the lower end of the distribution in
the quantitative risk assessment, despite the important uncertainties
in the overall risk assessment. As discussed in this section and in
responses to significant comments above and in the Response to Comments
document, the Administrator has concluded that placing a large emphasis
on these factors and revising the standard to a level below 9.0
[micro]g/m\3\ would result in a standard that is more stringent than
the evidence indicates to be sufficient to protect public health with
an adequate margin of safety. Compared to a primary annual
PM2.5 standard set at a level of 9.0 [micro]g/m\3\, the
Administrator concludes that the extent to which lower standard levels
could result in further public health improvements becomes notably less
certain.
Thus, having carefully considered the scientific evidence,
quantitative information, CASAC advice, and public comments relevant to
his decision on the level of the primary annual PM2.5
standard, as discussed above and in the Response to Comments document,
the Administrator is revising the level of the primary annual
PM2.5 standard to 9.0 [micro]g/m\3\. In the Administrator's
judgment, based on the currently available evidence and information, an
annual standard set at this level and using the specified indicator,
averaging time, and form, in conjunction with the other primary PM
standards, would be requisite to protect public health with an adequate
margin of safety. The Administrator judges that such a standard would
protect, with an adequate margin of safety, the health of at-risk
populations, including children, older adults, those with pre-existing
cardiovascular and respiratory diseases, minority populations, and low
SES populations. The Administrator believes that a standard set at 9.0
[micro]g/m\3\ would be sufficient to protect public health with a
margin of safety, and believes that a lower standard would be more than
what is necessary to provide this degree of protection. This judgment
by the Administrator appropriately considers the degree of protection
that is neither more nor less stringent than necessary for this purpose
and recognizes that the CAA does not require that primary standards be
set at a zero-risk level, but rather at a level that reduces risk
sufficiently so as to protect public health with an adequate margin of
safety.
In reaching his conclusions on adequacy of the current suite of
primary PM2.5 standards, based on consideration of the
available scientific evidence and quantitative information, the CASAC's
advice and public comments, the Administrator finds that the available
information is insufficient to call into question the adequacy of the
public
[[Page 16286]]
health protection afforded by the current primary 24-hour
PM2.5 standard. As described earlier in this section, the
Administrator concludes that it is appropriate to retain the current
indicator (PM2.5), averaging time (24-hour), and form (98th
percentile, averaged over three years) for the primary 24-hour
PM2.5 standard and below explains the basis for his final
decision that is also appropriate to retain the current level of the
primary 24-hour PM2.5 standard.
In reaching his conclusion to retain the current primary 24-hour
PM2.5 standard the Administrator does so in light of the
conclusion that the epidemiologic evidence supports associations
between short- and long-term PM2.5 exposures and adverse
health effects, but that the epidemiologic evidence does not identify
specific concentrations at which those effects occur and the
Administrator has greatest confidence in effects where the bulk of the
data is reported (i.e., the mean PM2.5 concentration, with
some consideration for the 25th percentile of the air quality
distribution). Thus, in considering the epidemiologic evidence, the
Administrator concludes it is appropriate to focus on setting a
generally controlling annual standard as the most effective and
efficient way to reduce total population risk associated with both
long- and short-term PM2.5 exposures, and that it is
appropriate to revise the level of the annual standard level to 9.0
[micro]g/m\3\. In addition to the epidemiologic evidence, the
Administrator also considers the available controlled human exposure
studies, which provide evidence for health effects following single,
short-term PM2.5 exposures to concentrations that typically
correspond to upper end of the PM2.5 air quality
distribution in the U.S. (i.e., ``peak'' concentrations). In so doing,
the Administrator notes that these studies report statistically
significant effects on one or more indicators of cardiovascular
function following 2-hour exposures to PM2.5 concentrations
at and above 120 [mu]g/m\3\ and at and above 149 [mu]g/m\3\ for
vascular impairment, the effect shown to be most consistent across
studies. In particular, the Administrator notes that a single study is
assessed in the ISA Supplement that reports effects following 4-hour
exposures at 37.8 [micro]g/m\3\, although the results of this study are
inconsistent with the results of the controlled human exposure studies
assessed in the 2019 ISA. Along with the inconsistent results from the
controlled human exposure studies, the Administrator also recognizes
that effects observed in these studies are intermediate effects which
are not typically considered adverse and that the study participants
were healthy individuals. Taking into consideration the available
scientific evidence, including the uncertainties and limitations, along
with the CASAC's advice, the Administrator concludes that it is
appropriate to maintain a primary 24-hour PM2.5 standard to
protect against peak exposures.
Thus, the Administrator considers what primary 24-hour
PM2.5 standard is requisite to provide supplemental
protection against peak exposures. While having confidence that the
revised annual standard will result in lowering risk associated with
both long- and short-term PM2.5 exposure by lowering the
overall air quality distribution, as in the 2012 review, the
Administrator recognizes that an annual standard alone would not be
expected to offer sufficient protection with an adequate margin of
safety against the effects of short-term PM2.5 exposures in
all parts of the country. Therefore, he continues to conclude that it
is appropriate to continue to provide supplemental protection by means
of a 24-hour standard, in conjunction with a revised annual standard
level of 9.0 [micro]g/m\3\.
In considering the available scientific evidence assessed in the
2019 ISA and ISA Supplement, the Administrator first considers the
controlled human exposure studies for informing his decisions on the
primary 24-hour PM2.5 standard. In so doing, he notes that
in their review of the 2021 draft PA, the majority of CASAC members
expressed the view that controlled human exposure studies are not the
best evidence to use for justifying retaining the 24-hour standard
without revision, in part because these studies preferentially recruit
less susceptible individuals and have a typical exposure duration much
shorter than 24 hours. Thus, in the view of the majority, ``the
evidence of effects from controlled human exposure studies with
exposures close to the current 24-hour standard supports
epidemiological evidence for lowering the standard'' (Sheppard, 2022a,
p. 3-4 of consensus letter). In reviewing the controlled human exposure
studies, the Administrator agrees with the majority of CASAC that these
controlled human exposure studies generally do not include populations
with substantially increased risk from exposure to PM2.5,
such as children, older adults, or those with more severe underlying
illness. However, he disagrees with any conclusion that they should not
be used to inform a decision about the adequacy of the current
standard. The Administrator finds the information available from these
studies to be useful, noting that the recently available controlled
human exposure studies provide evidence for health effects following
single, short-term exposures to PM2.5 concentrations that
are greater than those allowed under the current standard. The results
of the controlled human exposure studies are inconsistent, particularly
at lower PM2.5 concentrations, but some studies do report
statistically significant effects on one or more indicators of
cardiovascular function following 2-hour exposures to PM2.5
concentrations at and above 120 [mu]g/m\3\ (and at and above 149 [mu]g/
m\3\ for vascular impairment, the effect shown to be most consistent
across studies). Additionally, one controlled human exposure study
assessed in the ISA Supplement reports evidence of some effects for
cardiovascular markers following 4-hour exposures to 37.8 [micro]g/m\3\
(Wyatt et al., 2020). However, there is inconsistent evidence for
inflammation in other controlled human exposure studies evaluated in
the 2019 ISA. The Administrator finds these studies are important in
establishing biological plausibility for PM2.5 exposures
causing more serious health effects, such as those seen in short-term
exposure epidemiologic studies, and they provide support that more
adverse effects may be experienced following longer exposure durations
and/or exposure to higher concentrations. As described in more detail
in responding to public comments in section II.B.3 above, he notes that
although the controlled human exposure studies do not provide a
threshold below which no effects occur, the observed effects in these
controlled human exposures studies are ones that signal an intermediate
effect in the body, likely due to short-term exposure to
PM2.5, and typically would not, by themselves, be judged as
adverse. As noted in sections II.A.2 and II.B.3 above, associated
judgments regarding adversity or health significance of measurable
physiological responses to air pollutants in previous NAAQS reviews
have been informed by guidance, criteria or interpretative statements
developed within the public health community. This type of information
on adversity of effects is particularly informative to the
Administrator's judgments regarding the adversity of the effects
observed in the controlled human exposure studies which are short-term
in nature (i.e., generally ranging from 2- to 5-hours), including those
studies that are
[[Page 16287]]
conducted at near-ambient PM2.5 concentrations. Based on the
observation that the effects observed in Wyatt et al. (2020) are not by
themselves adverse, and the fact that the findings of this study are
inconsistent with other currently available evidence regarding the
level at which effects are observed, the Administrator disagrees with
the view expressed by the majority of CASAC that this study supports
epidemiologic evidence for lowering the 24-hour standard.
Consistent with his approach in reaching his proposed decision and
taking into consideration these points as well as balancing these
limitations (i.e., that the health outcomes observed in these
controlled human exposure studies are not clearly adverse and that the
studies generally do not include those at increased risk from
PM2.5 exposure), the Administrator still considers it
appropriate to ensure that the 24-hour PM2.5 standard
provides protection against health effects consistently observed in the
controlled human exposure studies. He next examines the air quality
analyses, described in more detail in section II.A.c.i above, to assess
whether during recent air quality conditions, areas meeting the current
standards would experience PM2.5 concentrations reported in
these controlled human exposure studies. He observes that air quality
analyses demonstrate that the PM2.5 exposures shown to cause
consistent effects in the controlled human exposure studies are well
above the ambient concentrations typically measured in locations
meeting the current primary standards, and therefore suggest that the
current primary PM2.5 standards provide protection against
these ``peak'' concentrations. In fact, at air quality monitoring sites
meeting the current primary PM2.5 standards (i.e., the 24-
hour standard of 35 [mu]g/m\3\ and the annual standard of 12 [mu]g/
m\3\), the 2-hour concentrations generally remain below 10 [mu]g/m\3\,
and rarely exceed 30 [mu]g/m\3\. Though two-hour concentrations are
higher at monitoring sites violating the current standards, they
generally remain below 16 [mu]g/m\3\ and rarely exceed 80 [mu]g/m\3\,
still below concentrations in CHE studies where consistent effects are
observed (e.g., greater than 120 [mu]g/m\3\) (U.S. EPA, 2022b, section
2.3.2.2.3, Figure 2-19, and section 3.3.3.1). Additionally, and in
response to public comments, the Administrator notes additional air
quality analyses conducted by the EPA,\121\ that provide a more refined
analysis of whether areas that meet the current standards experience
peak concentrations reported in controlled human exposure studies. He
notes that 2-hour observations greater than 120 [mu]g/m\3\ and 4-hour
observations greater than 38 [mu]g/m\3\ rarely occur (e.g., 0.025% of
rolling 2-hour observations are greater than 120 [mu]g/m\3\ and 0.78%
of rolling 4-hour observations greater than 38 [mu]g/m\3\). Based on
this information, the Administrator finds that the current suite of
standards maintains subdaily concentrations of PM2.5 in
ambient air far below the exposure concentrations in controlled human
exposure studies where consistent effects have been observed, and notes
that while these studies generally do not include the most at-risk
individuals, the exposure concentrations in these studies also do not
elicit adverse effects.
---------------------------------------------------------------------------
\121\ Jones et al. (2023). Comparison of Occurrence of
Scientifically Relevant Air Quality Observations Between Design
Value Groups. Memorandum to the Rulemaking Docket for the Review of
the National Ambient Air Quality Standards for Particulate Matter
(EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
Further, in light of the Administrator's emphasis on the annual
standard as the controlling standard, with the 24-hour standard
providing supplemental protection against peak concentrations, he next
considers the potential impact of a revised annual standard of 9.0
[micro]g/m\3\ on the occurrence of peak sub-daily PM2.5
concentrations. Specifically, the Administrator takes note of the new
air quality analyses \122\ where he observes that lower percentages of
concentrations greater than 120 [micro]g/m\3\ and 38 [micro]g/m\3\
occur in areas meeting an annual standard of 9.0 [micro]g/m\3\ and a
24-hour standard of 35 [micro]g/m\3\, versus an annual standard of 12.0
[micro]g/m\3\ and a 24-hour standard of 35 [micro]g/m\3\. Thus, he
concludes that an annual standard that is controlling across most areas
of the country will continue to effectively limit peak daily
concentrations in conjunction with the existing 24-hour standard, with
its level of 35 [micro]g/m\3\ and 98th percentile form, which continues
to provide supplemental protection against peak concentrations.
---------------------------------------------------------------------------
\122\ Jones et al. (2023). Comparison of Occurrence of
Scientifically Relevant Air Quality Observations Between Design
Value Groups. Memorandum to the Rulemaking Docket for the Review of
the National Ambient Air Quality Standards for Particulate Matter
(EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
In addition, the Administrator also notes that the majority of the
CASAC in their review of the 2021 draft PA, as well as a number of
public commenters, support their recommendation to revise the current
24-hour standard by pointing to ``substantial epidemiologic evidence
from both morbidity and mortality studies'' which ``includes three U.S.
air pollution studies with analyses restricted to 24-hour
concentrations below 25 [mu]g/m\3\'' (Sheppard, 2022a, p. 17 consensus
responses). The Administrator notes that the epidemiologic evidence
available in this reconsideration, including the studies that restrict
short-term PM2.5 exposures (i.e., 24-hour PM2.5
concentrations) to levels below 25 [mu]g/m\3\, provides support for
positive and statistically significant associations between short-term
exposure to PM2.5 and all-cause mortality (Di et al., 2017a)
and CVD hospital admissions (deSouza et al., 2021; Di et al., 2017a).
He agrees that these studies help to provide additional support for
reaching conclusions on causality in the 2019 ISA. He further agrees
that the available epidemiologic studies provide important information
that it is appropriate to consider in this reconsideration, including
information on associations between health effects and PM2.5
exposures in diverse populations that are broadly representative of the
U.S. population, and include populations identified as at-risk (e.g.,
older adults, minority populations), as well as evidence of linear, no-
threshold concentration-response relationships in those associations,
although with less certainty in the shape of the curve at long-term
average concentrations below about 8 [mu]g/m\3\.
However, the Administrator also notes significant limitations in
the currently available epidemiologic information that limit his
ability to draw conclusions from the key short-term studies, including
those that employ restricted analyses, to inform his decision regarding
the level of the 24-hour PM2.5 standard. As a result of
these limitations, the Administrator does not find that the short-term
epidemiologic studies, or the other evidence such as the controlled
human exposure studies or the risk assessment, provide a sufficient
justification for revising the 24-hour standard.
First, he notes that short-term epidemiologic studies examine
associations between day-to-day variations in PM2.5
concentrations and health outcomes, often over multi-year study
periods. As such, these studies report long-term mean 24-hour
PM2.5 concentrations (e.g., mean 24-hour PM2.5
concentrations over multi-year study periods), rather than at specific
points in the distribution (i.e., 90th or 98th percentile 24-hour
concentrations) at which effects occur. Further, he notes
[[Page 16288]]
that while there can be considerable variability in daily exposures
over a multi-year study period, the bulk of the observations reflect
days with ambient PM2.5 concentrations in the middle of the
air quality distribution (i.e., ``typical'' days rather than days with
extremely low or extremely high concentrations). As a result, the
results of these studies are more directly applicable to decisions
regarding the annual standard (which is based on the long-term mean of
both short- and long-term epidemiologic studies), and the fact that
they do not report other air quality statistics, such as the 98th
percentile concentrations which might be more directly compared to the
level of the 24-hour standard, makes them less useful for informing
decisions on the 24-hour standard. As discussed in responding to
comments above, the form of the annual standard is based on the annual
mean PM2.5 concentration averaged over three years,\123\
which makes it better suited as a basis for controlling air quality to
avoid effects observed in both long-term and short-term epidemiologic
studies. By contrast, the form of the 24-hour standard is the 98th
percentile averaged over three years, which makes it appropriate for
controlling short-term peak concentrations. However, based on the
available air quality information, including distribution statistics of
PM2.5 concentrations and health events reported in the
short-term epidemiologic studies, these studies are too limited in
their ability to identify health effects attributable to specific
short-term peak concentrations that are necessary to evaluate whether
the 24-hour standard with its 98th percentile form should be revised
(e.g., restricted epidemiologic studies do not report the number or the
percentile of health events or the percentile of PM2.5
concentrations across the highest part of the restricted air quality
distribution, including the 98th percentile). Thus, the Administrator
does not consider it appropriate to use the reported means from short-
term studies to determine the appropriate level for a 24-hour standard
with a 98th percentile form.
---------------------------------------------------------------------------
\123\ The annual mean is calculated by averaging daily values in
a calendar quarter and then averaging calendar quarters. See 40 CFR
part 50 Appendix N, section 4.4.
---------------------------------------------------------------------------
Similarly, the Administrator does not consider the results of the
restricted analyses to be well suited to informing the choice of level
for a 24-hour standard. Restricted analyses use a subset of data from
their main analyses to evaluate health events that occur at
concentrations below a certain concentration (e.g., 25 [mu]g/m\3\). The
Administrator notes that the associations between the health effects
(e.g., mortality and cardiovascular morbidity) and PM2.5
concentrations remain even after excluding higher concentrations in the
restricted analyses, and he also recognizes that the magnitude of the
effect is generally greater in the restricted analyses compared to the
associations reported in the main analysis. He considers such analyses
to be informative in indicating that the health effects association
reported in the main (unrestricted) analysis are not driven only by the
upper peaks of the PM2.5 air quality distribution, but
rather persist at lower portions of the distribution (consistent with
his emphasis on the annual standard, which is focused on exposures near
the mean concentration, where the bulk of the exposure distribution is
concentrated). Indeed, he notes that if peak concentrations were the
principal driver of health effects associated with PM2.5
exposure, one might expect the associations to become weaker as the
upper portion of the data is excluded in the restricted analyses, which
is not what is reported by the analyses (e.g., the restricted analyses
generally report associations that are greater in magnitude compared to
the main analyses). However, he disagrees with the assertion by the
CASAC in their review of the 2021 draft PA and some public commenters
that it would be appropriate to focus on the specific PM2.5
concentration (e.g., 25 or 30 [mu]g/m\3\) at which the analysis was
restricted as the basis for choosing a 24-hour standard level. The
Administrator recognizes that in restricted analyses, while an
association continues to persist across the full range of the air
quality distribution, and that the cutpoint concentration at which the
analysis was restricted (e.g., 25 or 30 [micro]g/m\3\) becomes the
maximum PM2.5 concentration in the distribution, he also
notes that these studies do not provide information related to the
distribution of health events and PM2.5 concentrations, and
as such, he is more uncertain where the bulk of the data are and where
he has confidence in the reported association.\124\ He notes that no
evidence exists to support a conclusion that the PM2.5
concentration chosen as the cutpoint in a restricted analysis has any
bearing on the concentration at which effects are likely to occur (or
not occur). He notes that, as with long-term studies, the evidence does
not suggest there is a specific point in the air quality distribution
of these short-term studies that represents a ``bright line'' at and
above which effects have been observed and below which effects have not
been observed. In order to identify a level of the 24-hour standard
based on associations between the ``upper end'' of exposures, either in
the unrestricted or the restricted analyses, and adverse health
effects, it would be necessary to have a better understanding of how
specific 24-hour concentrations correspond to the frequency and total
number of observed health events in the study. Currently, such
information, including 98th percentile statistics, are not reported in
the key short-term epidemiologic studies (and if they were reported,
the Administrator would have to carefully consider how to weigh the
data). As such, in reaching his decision on the primary 24-hour
PM2.5 standard, the Administrator judges that the currently
available information from short-term epidemiologic studies, including
those that employ restricted analyses, does not provide a sufficient
basis to revise the current 24-hour standard, given that the 24-hour
standard focuses on reducing ``peak'' exposures (with its 98th
percentile form), but rather that such information supports his
judgment that it is appropriate to focus on revising the annual
standard for purposes of reducing all exposures, across the entire
distribution of air quality, to increase public health protection.
---------------------------------------------------------------------------
\124\ These studies do not report information about the
distribution of the health events and PM2.5
concentrations (e.g., means, medians, other percentiles) in the
restricted analyses.
---------------------------------------------------------------------------
In reaching final decisions regarding the adequacy of the primary
24-hour PM2.5 standard, the Administrator continues to view
an approach that focuses on setting a generally controlling annual
standard as the most effective and efficient way to reduce total
population risk associated with both long- and short-term
PM2.5 exposures. Additionally, he emphasizes that
improvements in air quality associated with meeting an annual standard
level of 9.0 [micro]g/m\3\ will result in lowering risk associated with
both long- and short-term PM2.5 exposure by lowering the
overall air quality distribution. The Administrator concludes that
reducing the annual standard is the most efficient way to reduce the
risks from short-term exposures identified in the epidemiologic
studies, as the available evidence suggests the bulk of the risk comes
from the large number of days across the bulk of the air quality
distribution, not the relatively small number of days with peak
concentrations. However, as in the 2012
[[Page 16289]]
review, the Administrator recognizes that an annual standard alone
would not be expected to offer sufficient protection with an adequate
margin of safety against the effects of short-term PM2.5
exposures in all parts of the country and concludes that, in
conjunction with a revised annual standard level of 9.0 [micro]g/m\3\,
it is appropriate to continue to provide supplemental protection by
means of a 24-hour standard, particularly for areas with high peak-to-
mean ratios possibly associated with strong local or seasonal sources.
In selecting the level of a 24-hour standard designed to provide
supplemental protection against peak exposures (in conjunction with a
revised annual standard of 9.0 [micro]g/m\3\), the Administrator
considers the information from the controlled human exposure studies
and the EPA's analysis of peak concentrations observed in areas meeting
the current standard of 35 [micro]g/m\3\ in conjunction with a revised
standard of 9.0 [micro]g/m\3\ to be of particular relevance. He notes
the controlled human exposure evidence includes studies reporting
effects on one or more indicators of cardiovascular function following
2-hour exposures at and above 120 [micro]g/m\3\, including effects
reported at and above 149 [micro]g/m\3\ for vascular impairment, the
effect shown to be most consistent across studies, and less consistent
effects at lower concentrations, including a single study at near
ambient concentrations (Wyatt et al., 2020) reporting effects following
4-hour exposures at 37.8 [micro]g/m\3\. He recognizes that the effects
observed (in those studies that observed effects) are ones that signal
an intermediate effect in the body, likely due to short-term exposure
to PM2.5, and typically would not, by themselves, be judged
as adverse, and the study participants were healthy individuals.
He notes in particular that, in the EPA's analysis, in areas
meeting the current 24-hour standard and the revised annual standard
0.029 percent of 2-hour observations and 0.41 percent of 4-hour
observations reach PM2.5 concentrations higher than 120
[micro]g/m\3\ and 37.8 [micro]g/m\3\, respectively. He also notes the
lack of evidence of effects from controlled human exposure studies at
levels below the current 24-hour standard and the fact that the results
of Wyatt et al. (2020) are inconsistent with other available studies,
as well as the intermediate nature of effects observed in this study.
In his judgment, the small number of occurrences of peak exposures
indicate that, in conjunction with a revised annual standard of 9.0
[micro]g/m\3\, the current 24-hour standard of 35 [micro]g/m\3\ remains
requisite to protect public health with an adequate margin of safety,
and that there is substantial basis to doubt whether further
improvements in public health would be achieved by further reducing
these exposures. Furthermore, the Administrator concludes that due to
the limitations and uncertainties outlined above, the information from
recent short-term epidemiologic studies, including those that use
restricted analyses, is inadequate to inform decisions regarding the
adequacy of the current 24-hour standard. Thus, in reaching his
decision on the primary 24-hour PM2.5 standard, the
Administrator concludes that currently available evidence does not call
into question the adequacy of the current standard.
In addition to the scientific evidence, the Administrator also
considers the risk assessment in evaluating the appropriate level of
the 24-hour PM2.5 standard. The risk assessment indicates
that the annual standard is the controlling standard across most of the
urban study areas evaluated (i.e., when air quality related to the
annual average PM2.5 concentrations decrease, daily average
PM2.5 concentrations are also expected to decrease). When
air quality is adjusted to just meet an alternative 24-hour standard
level of 30 [mu]g/m\3\ in the areas where the 24-hour standard is
controlling, the risk assessment estimates reductions in
PM2.5-associated risks across a more limited population and
number of areas compared to when air quality is adjusted to simulate
alternative levels for the annual standard (i.e., where the annual
standard is controlling), and these predictions are largely confined to
areas located in the western U.S., several of which are also likely to
experience risk reductions upon meeting a revised annual standard. With
respect to the CASAC's advice in their review of the 2021 draft PA, the
Administrator notes that the minority of CASAC advised that these
results suggest that the annual standard can be used to limit both
long- and short-term PM2.5 concentrations and views these
risk assessment results as supporting the conclusion that the current
24-hour standard is adequate (Sheppard, 2022a, p. 4 of consensus
letter). In contrast, the majority of CASAC members in their review of
the 2021 draft PA, as well as a number of public commenters that
support revision of the 24-hour standard, placed greater weight on the
evidence-based considerations (e.g. scientific evidence, like the
restricted analyses) than on the values estimated by the risk
assessment, noting the potential for uncertainties in how the risk
assessment was able to ``capture areas with wintertime stagnation and
residential wood-burning where the annual standard is less likely to be
protective'' (Sheppard, 2022a, p. 4 of consensus letter).
In considering the application of the risk assessment to judgments
about the adequacy of the current primary 24-hour PM2.5
standard, the Administrator again notes that the risk assessment
analyses of PM2.5-attributable mortality use input data that
include C-R functions from epidemiologic studies that have no threshold
and a linear C-R relationship down to zero, as well an air quality
adjustment approach that incorporates proportional decreases in
PM2.5 concentrations to meet lower standard levels. As such,
the Administrator notes that this quantitative approach does not
incorporate any elements of uncertainty in associations of health
effects at lower concentrations and that simulated air quality
improvements will always lead to proportional decreases in risk (i.e.,
each additional [micro]g/m\3\ reduction produces additional benefits
with no clear stopping point at any PM2.5 concentration).
Therefore, the Administrator recognizes that while the risk estimates
can help to place the evidence for specific health effects into a
broader public health context, the results should be considered along
with the inherent uncertainties and limitations of such analyses when
informing judgments about the potential for additional public health
protection associated with PM2.5 exposure and related health
effects. Further, the Administrator notes additionally that air quality
analyses have also been considered in looking at the adequacy of the
24-hour standard in controlling peak PM2.5 concentrations of
potential concern,\125\ and that those analyses included monitoring
information from across the entire U.S., specifically highlighting
areas with higher peak concentrations and including areas impacted by
wintertime stagnation and residential wood-burning. Thus, while the
risk assessment may have focused on a subset of areas across the U.S.
based on the study area selection criteria, the Administrator is
considering a broader set of information in reaching his conclusions
regarding the appropriateness of the current 24-
[[Page 16290]]
hour standard to control peak concentrations.
---------------------------------------------------------------------------
\125\ Jones et al. (2023). Comparison of Occurrence of
Scientifically Relevant Air Quality Observations Between Design
Value Groups. Memorandum to the Rulemaking Docket for the Review of
the National Ambient Air Quality Standards for Particulate Matter
(EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
The Administrator also considers the advice from the CASAC in their
reviews of the 2019 draft PA and 2021 draft PA. In their review of the
2019 draft PA, the CASAC ``agrees with the EPA and finds that the
available evidence does not call into question the adequacy of public
health protection afforded by the current 24-hour PM2.5
standard and concurs that it be retained'' (Cox, 2019b, p. 3 of
letter). He also notes that in their review of the 2021 draft PA, the
CASAC did not reach consensus on whether the current 24-hour standard
is adequate, with the majority of the CASAC recommending that the 24-
hour standard be revised and the minority of the CASAC recommending
that the standard be retained. The majority of the CASAC members
further stated that ``[t]here is also less confidence that the annual
standard could adequately protect against health effects of short-term
exposures. A range of 25-30 [mu]g/m\3\ for the 24-hour PM2.5
standard would be adequately protective'' (Sheppard, 2022a, p. 4 of
consensus letter). The Administrator also acknowledges that some public
commenters agreed with the majority of the CASAC in supporting a
revision to the level of the 24-hour standard to a range between 25-30
[micro]g/m\3\. These commenters cite a number of reasons, including:
(1) Results from controlled human exposure studies at near ambient
concentrations; (2) aspects of the scientific evidence, including
restricted analyses that report positive and significant associations
below 35 [micro]g/m\3\; and (3) quantitative risk analyses that show
decreasing risk with decreasing PM2.5 concentrations. In
responding to these comments, the Administrator recognizes that some
commenters have different interpretations of the evidence, air quality
information, and quantitative results from the risk assessment in this
review and would make different judgments about the weight to place on
the relative strength and limitations of the currently available
scientific evidence and information and how such information could be
used in making public health policy decisions on the 24-hour standard.
However, as outlined above, the Administrator has carefully considered
the information available from controlled human exposure studies and
short-term epidemiologic studies, and weighed the strengths and
limitations of this evidence in formulating his decisions. Furthermore,
as discussed above the Administrator has noted significant
uncertainties and limitations inherent in the risk estimates, as well
as noting that very few areas were included. In addition, he has given
careful consideration to the majority of the CASAC's advice in their
review of the 2021 draft PA, but has drawn different conclusions with
respect to how currently available evidence and air quality information
inform the selection of level for the 24-hour primary PM2.5
standard.
In considering the advice of the majority of CASAC, the
Administrator notes that a decision to set the level of the 24-hour
standard to below 35 [micro]g/m\3\ would place a large amount of
emphasis on the potential public health importance of further reducing
the occurrence of peak PM2.5 concentrations. However, the
Administrator concludes that there is insufficient basis to conclude
that a more stringent standard to further reduce peak concentrations is
needed or would benefit public health. As discussed above, he judges
that the PM2.5 exposures in controlled human exposure
studies that correspond to peak concentrations will already be well
controlled via the combination of the revised annual standard, with a
level of 9.0 [micro]g/m\3\, and the 24-hour standard with its level 35
[micro]g/m\3\ and its 98th percentile form. Taking into consideration
the inconsistent results reported in controlled human exposure studies,
the intermediate nature of the health effects observed in the
controlled human exposure studies that are not typically considered
adverse, the health status of the study participants, and how
infrequently peak concentrations of potential concern are anticipated
to occur in areas meeting the revised primary annual PM2.5
standard, he judges that the current 24-hour standard is requisite to
protect against the effects reported in these studies with an adequate
margin of safety. Likewise, he judges that neither the epidemiologic
studies (including the studies that use restricted analyses) nor the
risk assessment provide a sufficient basis for revising the 24-hour
standard. As discussed above, the epidemiologic studies, including
short-term studies and those with restricted analyses, are not well-
suited for identifying a level for a 24-hour standard to address health
effects associated with peak concentrations. The restricted analyses
support the conclusion that the health effects associated with
PM2.5 is not associated primarily with exposure to higher
concentrations of the main analyses, but like other epidemiologic
studies they typically report only long-term mean 24-hour
concentrations (e.g., restricted epidemiologic studies do not report
the number or the percentile of health events or the percentile of
PM2.5 concentrations across the highest part of the
restricted air quality distribution, including the 98th percentile) and
do not identify any particular concentration within the air quality
distribution above which effects have been observed and below which
effects have not been observed. Similarly, the risk assessment
highlights that the annual standard is controlling across much of the
U.S. and is generally more effective at reducing risk than the 24-hour
standard and, taking into account the limitations and assumptions of
the risk assessment discussed above, does not provide a basis for
revising the 24-hour standard. For the reasons discussed herein, the
Administrator judges that the uncertainties as to whether there would
be public health benefits from a more stringent 24-hour standard are
too great to justify revising the standard.
Thus, having carefully considered the scientific evidence,
quantitative information, CASAC advice, and public comments, the
Administrator is retaining the current primary 24-hour PM2.5
standard, with its level of to 35 [micro]g/m\3\ and its 98th percentile
form. In the Administrator's judgment, based on the currently available
evidence and information, a 24-hour standard set at this level and
using the specified indicator, averaging time, and form would be
requisite to protect public health with an adequate margin of safety,
in conjunction with the annual standard. As noted, in evaluating the
adequacy of the current standards, the Administrator focuses on
evaluating the public health protection afforded by the annual and 24-
hour standards, taken together, against adverse health effects
associated with long- or short-term PM2.5 exposures. A 24-
hour standard set at a level of 35 [micro]g/m\3\, in conjunction with a
revised annual standard level of 9.0 [micro]g/m\3\, in the judgment of
the Administrator, provides an appropriate level of public health
protection, for both long- and short-term PM2.5 exposures.
The Administrator believes that a 24-hour standard set at 35 [micro]g/
m\3\ would continue to be sufficient to protect public health with a
margin of safety, and believes that a lower standard would be more than
what is necessary to provide this degree of protection when considered
in conjunction with a revised annual standard. The Administrator
concludes the current 24-hour standard at a level of 35 [micro]g/m\3\,
in conjunction with a revised annual standard level of 9.0 [micro]g/
m\3\, will provide appropriate protection
[[Page 16291]]
in areas in which the long-term mean concentrations are already
relatively low (i.e., below 9 [micro]g/m\3\) but where there may be
elevated short-term peak PM2.5 concentrations, often
associated with strong local or seasonal sources. This judgment by the
Administrator appropriately considers the degree of protection that is
neither more nor less stringent than necessary for this purpose and
recognizes that the CAA does not require that primary standards be set
at a zero-risk level, but rather at a level that reduces risk
sufficiently so as to protect public health with an adequate margin of
safety.
In making this decision to retain the current level of the primary
PM2.5 24-hour standard at 35 [micro]g/m\3\ in conjunction
with revising the annual standard level from 12.0 [micro]g/m\3\ to 9.0
[micro]g/m\3\, given all of the evidence and information discussed
above, the Administrator judges that the revised suite of primary
PM2.5 standards and the rationale supporting these levels
appropriately reflects consideration of the strength of the available
evidence and other information and its associated uncertainties as well
as the advice of CASAC and consideration of public comments. He
additionally judges that this suite of primary PM2.5
standards is requisite to protect public health, including at-risk
populations, with an adequate margin of safety from effects associated
with long and short-term exposures to fine particles. This judgment by
the Administrator appropriately considers the requirement for standards
that are requisite to protect public health but are neither more nor
less stringent than necessary.
C. Decisions on the Primary PM2.5 Standards
For the reasons discussed above, and taking into account the
information and assessments presented in the 2019 ISA and ISA
Supplement, the scientific and quantitative risk information in the
2022 PA, the advice and recommendations of the CASAC, and public
comments, the Administrator revises the current suite of primary
PM2.5 standards. Specifically, the Administrator revises the
level of the primary annual PM2.5 standard to 9.0 [micro]g/
m\3\ while retaining its form, indicator and averaging time. In
conjunction with revising the primary annual PM2.5 standard
level to provide protection from effects associated with long- and
short-term PM2.5 exposures, the Administrator retains the
level of 35 [micro]g/m\3\ and the 98th percentile form, indicator and
averaging time of the primary 24-hour PM2.5 standard to
continue to provide supplemental protection for areas with high peak
PM2.5 concentrations. The Administrator concludes that this
suite of standards is requisite to protect public health with an
adequate margin of safety against health effects potentially associated
with long- and short-term PM2.5 exposures.
III. Rationale for Decisions on the Primary PM10 Standard
This section presents the rationale for the Administrator's
decision to retain the existing primary PM10 standard. This
decision is based on a thorough review of the latest scientific
information, published through January 2018 \126\ and evaluated in the
2019 ISA, on human health effects associated with PM10-2.5
in ambient air. As described in section I above and in section 1.2 of
the ISA Supplement, the scope of the updated scientific evaluation of
the health effects evidence is based on those PM size fractions,
exposure durations, and health effects category combinations where the
2019 ISA concluded a causal relationship exists (U.S. EPA, 2019a, U.S.
EPA, 2022b).). Therefore, because the 2019 ISA did not conclude a
causal relationship for PM10-2.5 for any exposure durations
or health effect categories, the ISA Supplement does not include an
evaluation of additional studies for PM10-2.5. As a result,
the 2019 ISA continues to serve as the scientific foundation for
assessing the adequacy of the primary PM10 standard in this
reconsideration of the 2020 final decision (U.S. EPA, 2019a, section
1.7; U.S. EPA, 2022a). The Administrator's decision also takes into
account the 2022 PA evaluation of the policy-relevant information in
the 2019 ISA, CASAC advice and recommendations, and public comments.
---------------------------------------------------------------------------
\126\ In addition to the review's opening ``call for
information'' (79 FR 71764, December 3, 2014), the 2019 ISA
identified and evaluated studies and reports that have undergone
scientific peer review and were published or accepted for
publication between January 1, 2009, through approximately January
2018 (U.S. EPA, 2019a, p. ES-2). References cited in the 2019 ISA,
the references considered for inclusion but not cited, and
electronic links to bibliographic information and abstracts can be
found at: https://hero.epa.gov/hero/particulate-matter.
---------------------------------------------------------------------------
In presenting the rationale for the Administrator's final decision
and its foundations, Section III.A provides background on the 2020
final decision to retain the primary PM10 and a brief
summary of key aspects of the currently available health effects
information. Section III.B summarizes the CASAC advice and the
Administrator's proposed conclusions to retain the existing primary
PM10 standard, addresses public comments received on the
proposal, and presents the Administrator's conclusions on the adequacy
of the current standard, drawing on consideration of information in the
2019 ISA and the 2022 PA, advice from the CASAC, and comments from the
public. Section III.C summarizes the Administrator's decision on the
primary PM10 standard.
A. Introduction
The general approach for this reconsideration of the 2020 final
decision on the primary PM10 standard relies on the
scientific information available for this review, as well as the
Administrator's judgments regarding the available public health effects
evidence, and the appropriate degree of public health protection for
the existing standards. With the 2020 decision, the then-Administrator
retained the existing primary 24-hour PM10 standard, with
its level of 150 [micro]g/m\3\ and its one-expected-exceedance form on
average over three years, to continue to provide public health
protection against short-term exposures to PM10-2.5 (85 FR
82725, December 18, 2020).
1. Background on the Current Standard
Consistent with the 2009 ISA, the 2019 ISA concluded that the
available epidemiologic, controlled human exposure, and animal
toxicological studies, including uncertainties, provided support for
the causality determinations of ``suggestive of, but not sufficient to
infer, a causal relationship'' between short-term exposures to
PM10-2.5 and cardiovascular effects, respiratory effects,
and mortality (U.S. EPA, 2019a, section 1.4.2). The 2019 ISA also
reached the conclusion that the evidence supports a ``suggestive of,
but not sufficient to infer, a causal relationship'' between short-term
PM10-2.5 exposures and metabolic effects, an endpoint that
was not evaluated in the 2009 ISA (U.S. EPA, 2019a, section 1.4.2).
Compared to the 2009 ISA, the 2019 ISA includes expanded evidence
for the relationships between long-term exposures and cardiovascular
effects, metabolic effects, nervous system effects, cancer, and
mortality. The 2019 ISA concluded that the small number of
epidemiologic and experimental studies, including uncertainties,
contribute to the determination that, ``the evidence is suggestive of,
but not sufficient to infer, a causal relationship between long-term
PM10-2.5 exposure and cardiovascular effects, metabolic
effects, nervous system effects, cancer, and mortality and cancer (U.S.
EPA, 2019a, p. 10-87). For long-term exposures and cardiovascular
effects,
[[Page 16292]]
cardiovascular effects, and cancer, this is an upgrade from the
``inadequate to infer the presence or absence of a causal
relationship'' conclusions in the 2009 ISA (U.S. EPA, 2019a, section
1.4.2). This determination is also the first for long-term exposures
and metabolic effects, as the 2009 ISA did not include metabolic
effects as an endpoint (U.S. EPA, 2019a section 1.4.2).
In considering the available body of evidence, it was noted in the
2020 review there were considerable uncertainties and limitations
associated with the experimental evidence for PM2.5
exposures and health effects, and as such more weight was placed on the
available epidemiologic evidence. Therefore, the primary focus in the
2020 review was on multi-city and single-city epidemiologic studies
that evaluated associations between short-term PM10-2.5 and
mortality, cardiovascular effects (hospital admissions and emergency
department visits, as well as blood pressure and hypertension), and
respiratory effects. Despite differences in the approaches \127\ used
to estimate ambient PM10-2.5 concentrations, the majority of
the studies reported positive, though often not statistically
significant, associations with short-term PM10-2.5
exposures. Most PM10-2.5 effect estimates remained positive
in copollutant models that included either gaseous pollutants or other
particulate matter size fractions (e.g., PM2.5). In U.S.
study locations likely to have met the PM10 standard during
the study period, a few studies reported positive associations between
PM10-2.5 and mortality that were statistically significant
and remained so in copollutant models (U.S. EPA, 2019a). In addition to
the epidemiologic studies, there were a small number of controlled
human exposure studies evaluated in the 2019 ISA that reported
alterations in heart rate variability or increased pulmonary
inflammation following short-term exposure to PM10-2.5,
providing some support for the associations in the epidemiologic
studies. Animal toxicological studies examined the effect of short-term
PM10-2.5 exposures using non-inhalation (e.g., intratracheal
instillation) route.\128\ Therefore, these studies provided limited
evidence for the biological plausibility of PM10-2.5-induced
effects (U.S. EPA, 2019a). Although the scientific evidence available
in the 2019 ISA expanded the understanding of health effects associated
with PM10-2.5 exposures, a number of important uncertainties
remained. These uncertainties, and their implications for interpreting
the scientific evidence, include the following:
---------------------------------------------------------------------------
\127\ As discussed further below, methods employed by the
epidemiologic studies to estimate ambient PM10-2.5
concentrations include: (1) Calculating the difference between
PM10 and PM2.5 at co-located monitors, (2)
calculating the difference between county-wide averages of monitored
PM10 and PM2.5 based on monitors that are not
necessarily co-located, and (3) direct measurement of
PM10-2.5 using a dichotomous sampler (U.S. EPA, 2019a,
section 1.4.2).
\128\ Non-inhalation exposure experiments (i.e., intratracheal
[IT] instillation) are informative for size fractions (e.g.,
PM10-2.5) that cannot penetrate the airway of a study
animal and may provide information relevant to biological
plausibility and dosimetry (U.S. EPA, 2019a, section A-12).
---------------------------------------------------------------------------
The potential for confounding by copollutants, notably
PM2.5, was addressed with copollutant models in a relatively
small number of PM10-2.5 epidemiologic studies (U.S. EPA,
2019a). This was particularly important given the relatively small body
of experimental evidence (i.e., controlled human exposure and animal
toxicological studies) available to support the independent effect of
PM10-2.5 on human health. This increases the uncertainty
regarding the extent to which PM10-2.5 itself, rather than
one or more copollutants, is responsible for the mortality and
morbidity effects reported in epidemiologic studies.
There was greater spatial variability in
PM10-2.5 concentrations than PM2.5
concentrations, resulting in the potential for increased exposure error
for PM10-2.5 (U.S. EPA, 2019a). Available measurements did
not provide sufficient information to adequately characterize the
spatial distribution of PM10-2.5 concentrations (U.S. EPA,
2019a). The limitations in estimates of ambient PM10-2.5
concentrations ``would tend to increase uncertainty and make it more
difficult to detect effects of PM10-2.5 in epidemiologic
studies'' (U.S. EPA, 2019a).
Estimation of PM10-2.5 concentrations over
which reported health outcomes occur remain highly uncertain. When
compared with PM2.5, there is uncertainty spanning all
epidemiologic studies examining associations with PM10-2.5
including deficiencies in the existing monitoring networks, the lack of
a systematic evaluation of the various methods used to estimate
PM10-2.5 concentrations and the resulting uncertainty in the
spatial as well as the temporal variability in PM10-2.5
concentration (U.S. EPA, 2019a).). Given these limitations in routine
monitoring, epidemiologic studies employed a number of different
approaches for estimating PM10-2.5 concentrations, including
(1) calculating the difference between PM10 and
PM2.5 at co-located monitors, (2) calculating the difference
between county-wide averages of monitored PM10 and
PM2.5 based on monitors that are not necessarily co-located,
and (3) direct measurement of PM10-2.5 using a dichotomous
sampler (U.S. EPA, 2019a, section 1.4.2). Given the relatively small
number of PM10-2.5 monitoring sites, the relatively large
spatial variability in ambient PM10-2.5 concentrations, the
use of different approaches to estimating ambient PM10-2.5
concentrations across epidemiologic studies, and the limitations
inherent in such estimates, the distributions of PM10-2.5
concentrations over which reported health outcomes occur remain highly
uncertain (U.S. EPA, 2019a).
There was relatively little information available to characterize
potential exposure differences that may inform the apparent variability
in associations between short-term PM10-2.5 exposures and
health effects across study locations (U.S. EPA, 2019a). Specifically,
the potential spatial and temporal variability in PM10-2.5
exposures complicates the interpretation of results between study
locations as well as the relative lack of information on the chemical
and biological composition of PM10-2.5 (U.S. EPA, 2009a U.S.
EPA, 2019a).
In reaching his decision in 2020 to retain the existing 24-hour
primary PM10 standard, the then-Administrator specifically
noted that, while the health effects evidence was somewhat expanded
since the prior reviews, the overall conclusions in the 2019 ISA,
including uncertainties and limitations, were generally consistent with
what was considered in the 2012 review (85 FR 82725, December 18,
2020). In addition, the then-Administrator recognized that there were
still a number of uncertainties and limitations associated with the
available evidence.
With regard to the evidence on PM10-2.5-related health
effects, the then-Administrator noted that epidemiologic studies
continued to report positive associations with mortality and morbidity
in cities across North America, Europe, and Asia, where
PM10-2.5 sources and composition were expected to vary
widely. While significant uncertainties remained in the 2020 review,
the then-Administrator recognized that this expanded body of evidence
had broadened the range of effects that have been linked with
PM10-2.5 exposures. The studies evaluated in the 2019 ISA
expanded the scientific foundation presented in the 2009 ISA and led to
revised causality determinations (and new determinations) for long-term
PM10-2.5
[[Page 16293]]
exposures and mortality, cardiovascular effects, metabolic effects,
nervous system effects, and cancer (85 FR 82726, December 18, 2020).
Drawing from his consideration of this evidence, the then-Administrator
concluded that the scientific information available since the time of
the last review supported a decision to maintain a primary
PM10 standard to provide public health protection against
PM10-2.5 exposures, regardless of location, source of
origin, or particle composition (85 FR 82726, December 18, 2020). With
regard to uncertainties in the available evidence, the then-
Administrator first noted that a number of limitations were identified
in the 2012 review related to: (1) Estimates of ambient
PM10-2.5 concentrations used in epidemiologic studies; (2)
limited evaluation of copollutant models to address the potential for
confounding; and (3) limited experimental studies supporting biological
plausibility for PM10-2.5-related effects. Despite the
expanded body of evidence for PM10-2.5 exposures and health
effects, the then-Administrator recognized that uncertainties in the
2020 review continued to include those associated with the exposure
estimates used in epidemiologic studies, the independence of the
PM10-2.5 health effect associations, and the biologically
plausible pathways for PM10-2.5 health effects (85 FR 82726,
December 18, 2020). These uncertainties contributed to the 2019 ISA
determinations that the evidence is at most ``suggestive of, but not
sufficient to infer'' causal relationships (85 FR 82726, December 18,
2020). In considering the available evidence in his basis for the
decision, the then-Administrator emphasized evidence supporting
``causal'' and ``likely to be causal'' relationships, and therefore,
judged that the PM10-2.5-related health effects evidence
provided an uncertain scientific foundation for making standard-setting
decisions. He further judged limitations in the evidence raised
questions as to whether additional public health improvements would be
achieved by revising the existing PM10 standard (85 FR
24126, April 30, 2020). In the 2020 decision, for all of the reasons
discussed above and recognizing the CASAC conclusion that the evidence
provided support for retaining the current standard, the then-
Administrator concluded that it was appropriate to retain the existing
primary PM10 standard, without revision. His decision was
consistent with the CASAC advice related to the primary PM10
standard. Specifically, the CASAC agreed with the 2020 PA conclusions
that, while these effects are important, the ``evidence does not call
into question the adequacy of the public health protection afforded by
the current primary PM10 standard'' and ``supports
consideration of retaining the current standard in this review'' (Cox,
2019b, p. 3 of consensus letter). Thus, the then-Administrator
concluded that the primary PM10 standard (in all of its
elements (i.e., indicator, averaging time, form, and level)) was
requisite to protect public health with an adequate margin of safety
against effects that have been associated with PM10-2.5. In
light of this conclusion, the EPA retained the existing PM10
standard.
2. Overview of the Health Effects Evidence
The information summarized here is based on the scientific
assessment of the health effects evidence available in this
reconsideration; this evaluation is documented in the 2019 ISA and its
policy implications are discussed further in the 2022 PA. As noted
above, the ISA Supplement does not include an evaluation of studies for
PM10-2.5, and the 2019 ISA continues to serve as the
scientific foundation for this reconsideration.
a. Nature of Effects
For the health effect categories and exposure duration combinations
evaluated, the 2019 ISA concludes that the evidence supports causality
determinations for PM10-2.5 that are at most ``suggestive
of, but not sufficient to infer, a causal relationship''. While the
evidence supporting the causal nature of relationships between exposure
to PM10-2.5 has been strengthened for some health effect
categories since the completion of the 2009 ISA, the 2019 ISA concludes
that overall ``the uncertainties in the evidence identified in the 2009
ISA have, to date, still not been addressed'' (U.S. EPA, 2019a, section
1.4.2, p. 1-41; U.S. EPA, 2022b, section 4.3.1). Specifically,
epidemiologic studies available in the 2012 review relied on various
methods to estimate PM10-2.5 concentrations, and these
methods had not been systematically compared to evaluate spatial and
temporal correlations in PM10-2.5 concentrations. Methods
included: (1) Calculating the difference between PM10 and
PM2.5 concentrations at co-located monitors, (2) calculating
the difference between county-wide averages of monitored
PM10- and PM2.5-based on monitors that are not
necessarily co-located, and (3) direct measurement of
PM10-2.5 using a dichotomous sampler (U.S. EPA, 2019a,
section 1.4.2). As described in the 2019 ISA, there continues to be
variability across epidemiologic studies in the approaches used to
estimate PM10-2.5 concentrations. Additionally, some studies
estimate long-term PM10-2.5 exposures as the difference
between PM10 and PM2.5 concentrations based on
information from spatiotemporal or land use regression (LUR) models, in
addition to monitors. The various methods used to estimate
PM10-2.5 concentrations have not been systematically
evaluated (U.S. EPA, 2019a, section 3.3.1.1), contributing to
uncertainty regarding the spatial and temporal correlations in
PM10-2.5 concentrations across methods and in the
PM10-2.5 exposure estimates used in epidemiologic studies
(U.S. EPA, 2019a, section 2.5.1.2.3). Given the greater spatial and
temporal variability of PM10-2.5 and the lower number of
PM10-2.5 monitoring sites, compared to PM2.5,
this uncertainty is particularly important for the coarse size
fraction. Beyond the uncertainty associated with PM10-2.5
exposure estimates in epidemiologic studies, the limited information on
the potential for confounding by copollutants and the limited support
available for the biological plausibility of health effects following
PM10-2.5 exposures also continue to contribute to
uncertainty in the PM10-2.5 health evidence. Uncertainty
related to potential confounding stems from the relatively small number
of epidemiologic studies that have evaluated PM10-2.5 health
effect associations in copollutants models with both gaseous pollutants
and other PM size fractions. On the other hand, uncertainty related to
the biological plausibility of effects attributed to
PM10-2.5 exposures results from the small number of
controlled human exposure and animal toxicological studies that have
evaluated the health effects of experimental PM10-2.5
inhalation exposures. The evidence supporting the 2019 ISA's
``suggestive of, but not sufficient to infer, a causal relationship''
causality determinations for PM10-2.5, including
uncertainties in this evidence, is summarized below in sections
III.B.1.a through III.B.1.f.
i. Mortality
Due to the dearth of studies examining the association between
long-term PM10-2.5 exposure and mortality, the 2009 ISA
concluded that the evidence was ``inadequate to determine if a causal
relationship exists'' (U.S. EPA, 2009a). As reported in the 2019 ISA,
some cohort studies conducted in the U.S. and Europe report positive
associations between long-term PM10-2.5 exposure and total
(nonaccidental)
[[Page 16294]]
mortality, though results are inconsistent across studies (U.S. EPA,
2019a, Table 11-11). The examination of copollutant models in these
studies remains limited and, when included, PM10-2.5 effect
estimates are often attenuated after adjusting for PM2.5
(U.S. EPA, 2019a, Table 11-11). Across studies, PM10-2.5
exposure concentrations are estimated using a variety of approaches,
including direct measurements from dichotomous samplers, calculating
the difference between PM10 and PM2.5
concentrations measured at collocated monitors, and calculating
difference of area-wide concentrations of PM10 and
PM2.5. As discussed above, temporal and spatial correlations
between these approaches have not been evaluated, contributing to
uncertainty regarding the potential for exposure measurement error
(U.S. EPA, 2019a, section 3.3.1.1 and Table 11-11). The 2019 ISA
concludes that this uncertainty ``reduces the confidence in the
associations observed across studies'' (U.S. EPA, 2019a, p. 11-125).
The 2019 ISA additionally concludes that the evidence for long-term
PM10-2.5 exposures and cardiovascular effects, respiratory
morbidity, and metabolic disease provide limited biological
plausibility for PM10-2.5-related mortality (U.S. EPA,
2019a, sections 11.4.1 and 11.4). Taken together, the 2019 ISA
concludes that, ``this body of evidence is suggestive, but not
sufficient to infer, that a causal relationship exists between long-
term PM10-2.5 exposure and total mortality'' (U.S. EPA,
2019a, p. 11-125).
With regard to short-term PM10-2.5 exposures and
mortality, the 2009 ISA concluded that the evidence is ``suggestive of
a causal relationship between short-term exposure to
PM10-2.5 and mortality'' (U.S. EPA, 2009a). The 2019 ISA
included multicity epidemiologic studies conducted primarily in Europe
and Asia that continue to provide consistent evidence of positive
associations between short-term PM10-2.5 exposure and total
(nonaccidental) mortality (U.S. EPA, 2019a, Table 11-9). Although these
studies contribute to increasing confidence in the PM10-2.5-
mortality relationship, the use of various approaches to estimate
PM10-2.5 exposures continues to contribute uncertainty to
the associations observed. Recent studies expand the assessment of
potential copollutant confounding of the PM10-2.5-mortality
relationship and provide evidence that PM10-2.5 associations
generally remain positive in copollutant models, though associations
are attenuated in some instances (U.S. EPA, 2019a, section 11.3.4.1,
Figure 11-28, Table 11-10). The 2019 ISA concludes that, overall, the
assessment of potential copollutant confounding is limited due to the
lack of information on the correlation between PM10-2.5 and
gaseous pollutants and the small number of locations in which
copollutant analyses have been conducted. Associations with cause-
specific mortality (i.e., cardiovascular and respiratory mortality)
provide some support for associations with total (nonaccidental)
mortality, though associations with respiratory mortality are more
uncertain (i.e., wider confidence intervals) and less consistent (U.S.
EPA, 2019a, section 11.3.7). The 2019 ISA concludes that the evidence
for PM10-2.5-related cardiovascular effects provides only
limited support for the biological plausibility of a relationship
between short-term PM10-2.5 exposure and cardiovascular
mortality (U.S. EPA, 2019a, section 11.3.7). Based on the overall
evidence, the 2019 ISA concludes that, ``this body of evidence is
suggestive, but not sufficient to infer, that a causal relationship
exists between short-term PM10-2.5 exposure and total
mortality'' (U.S. EPA, 2019a, p. 11-120).
ii. Cardiovascular Effects
In the 2009 ISA, the evidence describing the relationship between
long-term exposure to PM10-2.5 and cardiovascular effects
was characterized as ``inadequate to infer the presence or absence of a
causal relationship.'' The limited number of epidemiologic studies
reported contradictory results and experimental evidence demonstrating
an effect of PM10-2.5 on the cardiovascular system was
lacking (U.S. EPA, 2019a, section 6.4).
The evidence relating long-term PM10-2.5 exposures to
cardiovascular mortality remains limited, with no consistent pattern of
associations across studies and, as discussed above, uncertainty
stemming from the use of various approaches to estimate
PM10-2.5 concentrations (U.S. EPA, 2019a, Table 6-70). The
evidence for associations with cardiovascular morbidity has grown and,
while results across studies are not entirely consistent, some
epidemiologic studies report positive associations with ischemic heart
disease (IHD) and MI (U.S. EPA, 2019a, Figure 6-34); stroke (U.S. EPA,
2019a, Figure 6-35); atherosclerosis (U.S. EPA, 2019a, section 6.4.5);
venous thromboembolism (VTE) (U.S. EPA, 2019a, section 6.4.7); and
blood pressure and hypertension (U.S. EPA, 2019a, Section 6.4.6).
PM10-2.5 cardiovascular mortality effect estimates are often
attenuated, but remain positive, in copollutants models that adjust for
PM2.5. For morbidity outcomes, associations are inconsistent
in copollutant models that adjust for PM2.5, NO2,
and chronic noise pollution (U.S. EPA, 2019a, p. 6-276). The lack of
toxicological evidence for long-term PM10-2.5 exposures
represents a data gap (U.S. EPA, 2019a, section 6.4.10), resulting in
the 2019 ISA conclusion that ``evidence from experimental animal
studies is of insufficient quantity to establish biological
plausibility'' (U.S. EPA, 2019a, p. 6-277). Based largely on the
observation of positive associations in some epidemiologic studies, the
2019 ISA concludes that ``evidence is suggestive of, but not sufficient
to infer, a causal relationship between long-term PM10-2.5
exposure and cardiovascular effects'' (U.S. EPA, 2019a, p. 6-277).
With regard to short-term PM10-2.5 exposures and
cardiovascular effects, the 2009 ISA found that the available evidence
for short-term PM10-2.5 exposure and cardiovascular effects
was ``suggestive of a causal relationship.'' This conclusion was based
on several epidemiologic studies reporting associations between short-
term PM10-2.5 exposure and cardiovascular effects, including
IHD hospitalizations, supraventricular ectopy, and changes in heart
rate variability (HRV). In addition, dust storm events resulting in
high concentrations of crustal material were linked to increases in
total cardiovascular disease emergency department visits and hospital
admissions. However, the 2009 ISA noted the potential for exposure
measurement error primarily due to the different methods used across
studies to estimate PM10-2.5 concentrations and copollutant
confounding in these epidemiologic studies. In addition, there was only
limited evidence of cardiovascular effects from a small number of
experimental studies (e.g. animal toxicological studies and controlled
human exposure studies) that examined short-term PM10-2.5
exposures (U.S. EPA, 2009a, section 6.2.12.2). In the 2019 ISA, key
uncertainties included the potential for exposure measurement error,
copollutant confounding, and limited evidence of biological
plausibility for cardiovascular effects following inhalation exposure
(U.S. EPA, 2019a, section 6.3.13).
The evidence for short-term PM10-2.5 exposure and
cardiovascular outcomes has expanded since the 2009 ISA, though
important uncertainties remain. The 2019 ISA notes that there are a
small number of epidemiologic studies reporting positive associations
between short-term exposure to PM10-2.5 and cardiovascular-
related morbidity
[[Page 16295]]
outcomes. However, the 2019 ISA notes that there is limited evidence to
support that these associations are biologically plausible, or
independent of copollutant confounding. The 2019 ISA also concludes
that it remains unclear how the approaches used to estimate
PM10-2.5 concentrations in epidemiologic studies compare
amongst one another and subsequently how exposure measurement error
varies between each method. Specifically, it is unclear how well-
correlated PM10-2.5 concentrations are both temporally and
spatially across these methods and therefore whether exposure
measurement error varies across these methods. Taken together, the 2019
ISA concludes that ``the evidence is suggestive of, but not sufficient
to infer, a causal relationship between short-term PM10-2.5
exposures and cardiovascular effects'' (U.S. EPA, 2019a, p. 6-254).
iii. Respiratory Effects
With regard to short-term PM10-2.5 exposures and
respiratory effects, the 2009 ISA (U.S. EPA, 2009a) concluded that the
relationship between short-term exposure to PM10-2.5 and
respiratory effects is ``suggestive of a causal relationship'' based on
a small number of epidemiologic studies observing associations with
some respiratory effects and limited evidence from experimental studies
to support biological plausibility. Epidemiologic findings were
consistent for respiratory infection and combined respiratory-related
diseases, but not for COPD. Studies were characterized by overall
uncertainty in the exposure assignment approach and limited information
regarding potential copollutant confounding. Controlled human exposure
studies of short-term PM10-2.5 exposures found no lung
function decrements and inconsistent evidence for pulmonary
inflammation. Animal toxicological studies were limited to those using
non-inhalation (e.g., intra-tracheal instillation) routes of
PM10-2.5 exposure.
Recent epidemiologic findings consistently link PM10-2.5
exposure to asthma exacerbation and respiratory mortality, with some
evidence that associations remain positive (though attenuated in some
studies of mortality) in copollutant models that include
PM2.5 or gaseous pollutants. Epidemiologic studies provide
limited evidence for positive associations with other respiratory
outcomes, including COPD exacerbation, respiratory infection, and
combined respiratory-related diseases (U.S. EPA, 2019a, Table 5-36). As
noted above for other endpoints, an uncertainty in these epidemiologic
studies is the lack of a systematic evaluation of the various methods
used to estimate PM10-2.5 concentrations and the resulting
uncertainty in the spatial and temporal variability in
PM10-2.5 concentrations compared to PM2.5 (U.S.
EPA, 2019a, sections 2.5.1.2.3 and 3.3.1.1). Specifically, the existing
monitoring networks do not provide a good characterization of how well
correlated concentrations are both spatially and temporally across the
PM10-2.5 estimation methods and overall spatial and temporal
patterns in PM10-2.5 concentrations. Taken together, the
2019 ISA concludes that ``the collective evidence is suggestive of, but
not sufficient to infer, a causal relationship between short-term
PM10-2.5 exposure and respiratory effects'' (U.S. EPA,
2019a, p. 5-270).
iv. Cancer
In the 2012 review, little information was available from studies
of cancer following inhalation exposures to PM10-2.5. Thus,
the 2009 ISA determined the evidence was ``inadequate to evaluate the
relationship between long-term PM10-2.5 exposures and
cancer'' (U.S. EPA, 2009a). The scientific information evaluated in the
2019 ISA of long-term PM10-2.5 exposure and cancer remains
limited, with a few recent epidemiologic studies reporting positive,
but imprecise, associations with lung cancer incidence (U.S. EPA,
2019a). Moreover, uncertainty remains in these studies with respect to
exposure measurement error due to the use of PM10-2.5
predictions that have not been validated by monitored
PM10-2.5 concentrations (U.S. EPA, 2019a, sections 3.3.2.3
and 10.3.4). Relatively few experimental studies of PM10-2.5
have been conducted, though available studies indicate that
PM10-2.5 exhibits two key characteristics of carcinogens:
genotoxicity and oxidative stress. While limited, such experimental
studies provide some evidence of biological plausibility for the
findings in a small number of epidemiologic studies (U.S. EPA, 2019a,
section 10.3.4).
Taken together, the small number of epidemiologic and experimental
studies, along with uncertainty with respect to exposure measurement
error, contribute to the determination in the 2019 ISA that, ``the
evidence is suggestive of, but not sufficient to infer, a causal
relationship between long-term PM10-2.5 exposure and
cancer'' (U.S. EPA, 2019a, p. 10-87).
v. Metabolic Effects
The 2009 ISA did not make a causality determination for
PM10-2.5-related metabolic effects. One epidemiologic study
in the 2019 ISA reports an association between long-term
PM10-2.5 exposure and incident diabetes, while additional
cross-sectional studies report associations with effects on glucose or
insulin homeostasis (U.S. EPA, 2019a, section 7.4). As discussed above
for other outcomes, uncertainties with the epidemiologic evidence
include the potential for copollutant confounding and exposure
measurement error due to the different methods used across studies to
estimate PM10-2.5 concentrations (U.S. EPA, 2019a, Tables 7-
14 and 7-15). The evidence base to support the biological plausibility
of metabolic effects following PM10-2.5 exposures is
limited, but a cross-sectional study that investigated biomarkers of
insulin resistance and systemic and peripheral inflammation may support
a pathway leading to type 2 diabetes (U.S. EPA, 2019a, sections 7.4.1
and 7.4.3). Based on the expanded, though still limited evidence base,
the 2019 ISA concludes that, ``[o]verall, the evidence is suggestive
of, but not sufficient to infer, a causal relationship between [long]-
term PM10-2.5 exposure and metabolic effects'' (U.S. EPA,
2019a, p. 7-56).
vi. Nervous System Effects
The 2009 ISA did not make a causality determination for
PM10-2.5-related nervous system effects. In the 2019 ISA,
available epidemiologic studies report associations between
PM10-2.5 and impaired cognition and anxiety in adults in
longitudinal analyses (U.S. EPA, 2019a, Table 8-25, section 8.4.5).
Associations of long-term exposure with neurodevelopmental effects are
not consistently reported in children (U.S. EPA, 2019a, sections 8.4.4
and 8.4.5). Uncertainties in these studies include the potential for
copollutant confounding, as no studies examined copollutants models
(U.S. EPA, 2019a, section 8.4.5), and for exposure measurement error,
given the use of various methods to estimate PM10-2.5
concentrations (U.S. EPA, 2019a, Table 8-25). In addition, there is
limited animal toxicological evidence supporting the biological
plausibility of nervous system effects (U.S. EPA, 2019a, sections 8.4.1
and 8.4.5). Overall, the 2019 ISA concludes that, ``the evidence is
suggestive of, but not sufficient to infer, a causal relationship''
between long-term PM10-2.5 exposure and nervous system
effects (U.S. EPA, 2019a, p. 8-75).
[[Page 16296]]
B. Conclusions on the Primary PM10 Standard
In drawing conclusions on the adequacy of the current primary
PM10 standard, in view of the advances in scientific
knowledge and additional information now available, the Administrator
has considered the evidence base, information, and policy judgments
that were the foundation of the 2020 review and reflects upon the body
of information and evidence available in this reconsideration. In so
doing, the Administrator has taken into account both evidence-based and
quantitative information-based considerations, as well as advice from
the CASAC and public comments. Evidence-based considerations draw upon
the EPA's integrated synthesis of the scientific evidence from animal
toxicologic, controlled human exposure, and epidemiologic studies
evaluating health effects related to exposures to PM10-2.5
as presented in the 2019 ISA and discussed in section III.A.2. In
addition to the evidence, the Administrator has weighed a range of
policy-relevant considerations as discussed in the 2022 PA and
summarized in sections III.B and III.C of the proposal and summarized
in section III.B.2 below. These considerations, along with the advice
from the CASAC (section III.B.1) and public comments (section III.B.3),
are discussed below. A more detailed summary of all significant
comments, along with the EPA's responses in the Response to Comments
document, can be found in the docket for this rulemaking (Docket No.
EPA-HQ-OAR-2015-00072). This document is available for review in the
docket for this rulemaking and through EPA's NAAQS website (link). The
Administrator's conclusions in this reconsideration regarding the
adequacy of the current primary PM10 standard and whether
any revisions are appropriate are described in section III.B.4.
1. CASAC Advice
As described in section I.X, the EPA decided to prepare a revised
PA for the reconsideration of the 2020 final decision. The CASAC's
advice on the 2019 draft PA and the 2021 draft PA was documented in
letters to the prior and current Administrators (Cox, 2019b; Sheppard,
2022a) and is summarized below. In reviewing both the 2019 draft PA and
the 2021 draft PA, the CASAC agreed with the EPA's preliminary
conclusion that the available scientific evidence, including its
uncertainties and limitations, does not call into question the adequacy
of the current primary PM10 standard and that the standard
should be retained, without revision.
In its review of the 2019 draft PA, the CASAC concurred with the
overall preliminary conclusion that it is appropriate to consider
retaining the current primary PM10 standard, without
revision. In their agreement with the conclusions in the 2019 draft PA,
the CASAC stated that ``that key uncertainties identified in the last
review remain'' (Cox, 2019b) and that ``none of the identified health
outcomes linked to PM10-2.5'' were judged to be causal or
likely to be causal (Cox, 2019b, p. 12 of consensus responses).
Moreover, to reduce these uncertainties in future reviews, the CASAC
recommended improvements to PM10-2.5 exposure assessment,
including a more extensive network for direct monitoring of the
PM10-2.5 fraction (Cox, 2019b, p. 13 of consensus
responses). The CASAC also recommended additional controlled human
exposure and animal toxicological studies of the PM10-2.5
fraction to improve the understanding of biological mechanisms and
pathways (Cox, 2019b, p. 13 of consensus responses). Overall, the CASAC
agreed with the EPA's preliminary conclusion in the 2019 draft PA that
``. . . the available evidence does not call into question the adequacy
of the public health protection afforded by the current primary
PM10 standard and that evidence supports consideration of
retaining the current standard in this review'' (Cox, 2019b, p. 3 of
letter).
In its review of the 2021 draft PA, the CASAC provided advice on
the adequacy of the current primary PM10 standard in the
context of its review of the revised PA for this reconsideration
(Sheppard, 2022a) \129\.) \130\. In this context, the CASAC supported
the preliminary conclusion in the 2021 draft PA that the evidence
reviewed in the 2019 ISA does not call into question the public health
protection provided by the current primary PM10 standard
against PM10-2.5 exposures and concurs with the 2021 draft
PA's overall preliminary conclusion that it is appropriate to consider
retaining the current primary PM10 standard (Sheppard,
2022a, p. 4 of consensus letter). Additionally, the CASAC concurred
that ``. . . at this time, PM10 is an appropriate choice as
the indicator for PM10-2.5'' and ``that it is important to
retain the level of protection afforded by the current PM10
standard'' (Sheppard, 2022a, p. 4 of consensus letter). The CASAC also
recognized uncertainties associated with the scientific evidence,
including ``compared to PM2.5 studies, the more limited
number of epidemiology studies with positive statistically significant
findings, and the difficulty in extracting the sole contribution of
coarse PM to observed adverse health effects'' (Sheppard, 2022a, p. 19
of consensus responses).
---------------------------------------------------------------------------
\129\ As described in section I.C.5.b above, the scope of the
ISA Supplement did not include consideration of studies of health
effects associated with exposure to PM10-2.5. Therefore,
the information and conclusions presented in the 2022 PA are very
similar to those in the 2020 PA.
\130\ As described in section I.C.5.b above, the scope of the
ISA Supplement did not include consideration of studies of health
effects associated with exposure to PM10-2.5. Therefore,
the information and conclusions presented in the 2022 PA are very
similar to those in the 2020 PA.
---------------------------------------------------------------------------
The CASAC recommended several areas for additional research to
reduce uncertainties in the PM10-2.5 exposure estimates used
in the epidemiologic studies, to evaluate the independence of
PM10-2.5 health effect associations, to evaluate the
biological plausibility of PM10-2.5-related effects, and to
increase the number of studies examining PM10-2.5-related
health effects in at-risk populations (Sheppard, 2022a, p. 20 of
consensus responses). Furthermore, the CASAC ``recognizes a need for,
and supports investment in research and deployment of measurement
systems to better characterize PM10-2.5'' and to ``provide
information that can improve public health'' (Sheppard, 2022a, p. 20 of
consensus responses).
2. Basis for the Proposed Decision
At the time of the proposal, the Administrator carefully considered
the assessment of the current evidence and conclusions reached in the
2019 ISA, considerations and staff conclusions and associated
rationales presented in the 2020 PA and 2022 PA, and advice and
recommendations of the CASAC (88 FR 5634, January 27, 2023). Consistent
with previous reviews, the Administrator first considered the available
scientific evidence for PM10-2.5-related exposures and
health effects, as evaluated in the 2019 ISA. As an initial matter, the
Administrator recognized that the scientific evidence for
PM10-2.5-related effects available in this reconsideration
is the same body of evidence that was available at the time of the 2020
review, as evaluated in the 2019 ISA and summarized in section III.A.2
above. The 2019 ISA concludes that the evidence supports ``suggestive
of, but not sufficient to infer'' causal relationships between short-
and long-term exposures to PM10-2.5 and cardiovascular
effects, cancer, and mortality and long-term PM10-2.5
exposures and metabolic effects and nervous system effects (U.S. EPA,
2019a). The Administrator noted that
[[Page 16297]]
the evidence for several PM10-2.5-related health effects has
expanded since the completion of the 2009 ISA, but important
uncertainties remain. The uncertainties in the epidemiologic studies
contribute to the determinations in the 2019 ISA that the evidence for
short and long-term PM10-2.5 exposures and mortality,
cardiovascular effects, metabolic effects, nervous system effects, and
cancer is ``suggestive of, but not sufficient to infer'' causal
relationships (U.S. EPA, 2019a; U.S. EPA, 2022b, section 4.3.1).
Drawing from the evidence evaluated in the 2019 ISA and consideration
of the scientific evidence in the 2022 PA, the Administrator noted
that, consistent with previous reviews, the 2019 ISA and the 2022 PA
highlight a number of uncertainties associated with the evidence,
including: (1) PM10-2.5 exposure estimates used in
epidemiologic studies, (2) independence of PM10-2.5 health
effect associations, and (3) biological plausibility of the
PM10-2.5-related effects. These uncertainties contribute to
the determinations in the 2019 ISA that the evidence for short-term
PM10-2.5 exposures and key health effects is ``suggestive
of, but not sufficient to infer'' causal relationships. In considering
the available scientific evidence, consistent with approaches employed
in past NAAQS reviews, the Administrator placed the most weight on
evidence supporting ``causal'' and ``likely to be causal''
relationships. In so doing, he noted that the available evidence for
short- and long-term PM10-2.5 exposures and health effects
does not support causality determinations of a ``causal relationship''
or ``likely to be causal relationship.'' Furthermore, the Administrator
recognized that, because of the uncertainties and limitations in the
evidence base, the 2022 PA does not include a quantitative assessment
of PM10-2.5 exposures and risk that might further inform
decisions regarding the adequacy of the current 24-hour primary
PM10 standard. Therefore, in light of the 2019 ISA
conclusions that the evidence supports ``suggestive of, but not
sufficient to infer'' causal relationships. The Administrator judged
that there are substantial uncertainties that raise questions regarding
the degree to which additional public health improvements would be
achieved by revising the existing PM10 standard. In
considering the available evidence for long-term PM10-2.5
exposures, the Administrator noted that there is limited evidence that
would support consideration of an annual standard to provide protection
against such effects, in conjunction with the current primary 24-hour
PM10 standard. He preliminarily concluded that the current
primary 24-hour PM2.5 standard that reduces 24-hour
exposures also likely reduces long-term average exposures, and
therefore provides some margin of safety against the health effects
associated with long-term PM10-2.5 exposures.
In reaching his proposed decision on the adequacy of the current
primary 24-hour PM10 standard, the Administrator also
considered advice from the CASAC. As noted above in section III.B.1,
the CASAC recognized uncertainties associated with the scientific
evidence and agreed with the 2019 draft PA and 2021 draft PA
conclusions that the scientific evidence does not call into question
the adequacy of the primary PM10 standard and supports
consideration of retaining the current standard.
When considering the above information together, the Administrator
proposed to conclude that the available scientific evidence continues
to support a PM10 standard to provide some measure of
protection against PM10-2.5 exposures. Additionally, he
recognized that there are important uncertainties and limitations
associated with the available evidence for PM10-2.5-related
health effects, for both short and long-term exposure, as evaluated in
the 2019 ISA. Consistent with the decisions in the previous reviews,
the Administrator proposed to conclude that these limitations lead to
considerable uncertainty regarding the potential public health
implications of revising the level of the current primary 24-hour
PM10 standard. Thus, based on his consideration of the
evidence and associated uncertainties and limitations for
PM10-2.5-related health effects and his consideration of
CASAC advice on the primary PM10 standard, the Administrator
proposed to retain the current primary PM10 standard,
without revision.
3. Comments on the Proposed Decision
Of the public comments received on the proposal, very few
commenters provided comments on the primary PM10 standard.
Of those commenters who did provide comments on the primary
PM10 standard, the majority agree with the EPA's proposed
decision to retain the primary PM10 standard. In so doing,
these commenters agree with the EPA's rationale regarding the available
scientific information, including uncertainties and limitations, for
informing decisions on the standard. These commenters state that no new
scientific evidence or quantitative information has emerged since the
2020 decision to retain the current standard. Furthermore, these
commenters note that the EPA did not evaluate any new scientific
evidence related to PM10-2.5 exposures and health effects as
a part of the 2022 ISA Supplement developed for this reconsideration,
nor did the revised 2022 PA consider any new or different information
from the 2020 PA, and therefore, the EPA reached the same conclusion as
is the 2020 PA that the current standard is adequate and should be
retained. This group includes industries and industry groups, as well
as some State and local governments. All of these commenters generally
note their agreements with the rationale provided in the proposal and
the CASAC concurrence with the 2021 draft PA conclusion that the
available information does not call into question the adequacy of the
current standard, and therefore, does not support revision and that the
current standard should be retained.
Some commenters, including those from environmental and public
health organizations and groups, some states, and individuals,
disagreed with the Administrator's proposed decision to retain the
current primary PM10 standard. These commenters recommend
that the EPA revise the primary PM10 standard to a lower
level to provide increased public health protection, citing to the
available scientific evidence, as well as the proposed revision to the
primary PM2.5 standard.
Commenters who disagreed with the proposal to retain the current
standard state that revision to the primary PM10 standard is
necessary to protect public health with an adequate margin of safety.
In their recommendations for revising the standard, some commenters
contend that the current standard, with its indicator of
PM10 to target exposures to PM10-2.5, has become
less protective as ambient concentrations of PM2.5 have been
reduced with revisions to that standard. These commenters assert that
the current primary PM10 standard allows increased exposure
to PM10-2.5 in ambient air because retaining the primary
PM10 would allow proportionately more PM10-2.5
mass as the PM2.5 standard has been revised downward.
Moreover, in support of their recommendations, the commenters note that
the available evidence of PM10-2.5-related health effects
has been expanded and strengthened since the time of the last review.
Taken together, the commenters contend that the primary PM10
standard should be revised and failure to do so would be arbitrary and
capricious. Some of these
[[Page 16298]]
commenters assert that the level of the primary PM10
standard should be revised to 140 or 145 [micro]g/m\3\, concurrent with
a strengthened primary 24-hour PM2.5 standard, while other
commenters recommend revising the level of the standard to within the
range of 65-75 [micro]g/m\3\, to provide increased public health
protection.
We disagree with the commenters that the primary PM10
standard should be revised because of reductions in ambient
concentrations of PM2.5. As an initial matter, we note that
overall, ambient concentrations of both PM10 and
PM2.5 have declined significantly over time. Ambient
concentrations of PM10 have declined by 46% across the U.S.
from 2000 to 2019,\131\ while PM2.5 concentrations in
ambient air have declined by 43% during this same time period.\132\ As
noted in the 2022 PA (p. 2-41), the majority of PM10-2.5
sites have generally remained steady and do not exhibit a trend of
increasing or decreasing concentrations during this time period,
reflecting the relatively consistent level of dust emission across the
U.S. from 2000 to 2019 (U.S. EPA, 2022b).
---------------------------------------------------------------------------
\131\ PM10 concentrations presented as the annual
second maximum 24-hour concentration (in [micro]g/m\3\) at 262 sites
in the U.S. For more information, see: https://www.epa.gov/air-
trends/particulate-matter-pm10-trends
\132\ PM2.5 concentrations presented as the
seasonally-weighted annual average concentration (in [micro]g/m\3\)
at 406 sites in the U.S. For more information, see: https://www.epa.gov/air-trends/particulate-matter-pm25-trends
---------------------------------------------------------------------------
The 2019 ISA provides a comparison of the relative contribution of
PM2.5 and PM10-2.5 to PM10
concentrations by region and season using the more comprehensive
monitoring data from the NCore network available in this
reconsideration (U.S. EPA, 2019, section 2.5.1.1.4). The data indicate
that, for urban areas, there are roughly equivalent amounts of
PM2.5 and PM10-2.5 contributing to
PM10 in ambient air, while rural locations have a slightly
higher contribution of PM10-2.5 contributing to
PM10 concentrations than PM2.5 (U.S. EPA, 2019,
section 2.5.1.1.4, Table 2-7). There is generally a greater
contribution from the PM2.5 fraction in the East and a
greater contribution from the PM10-2.5 fraction in the West
and Midwest.
The EPA recognizes that when the primary annual PM2.5
standard was revised from 15.0 [micro]g/m\3\ to 12.0 [micro]g/m\3\
while leaving the 24-hour PM2.5 standards unchanged at 35
[micro]g/m\3\ and the 24-hour PM10 standard unchanged at 150
[micro]g/m\3\, the PM10-2.5 fraction of PM10
could increase in some areas as the PM2.5 fraction decreases
(78 FR 3085, March 03, 2013). As described in the 2019 ISA,
PM10 has become considerably coarser across the U.S.
compared to similar observations in the 2009 ISA such that, in urban
areas, the mass of the coarse fraction of PM is similar to or greater
than the mass of the fine fraction of PM (U.S. EPA, 2019, section
2.5.1.1.4; U.S. EPA, 2009c). However, in considering recent air quality
data, the EPA notes that in most areas of the country PM2.5
and PM10 concentrations have declined and are well below
their respective 24-hour standards. While the contribution of fine and
coarse PM to PM10 mass concentrations may vary spatially and
temporally, based on the trends in recent air quality data, the
Administrator concludes that the current primary 24-hour
PM10 standard is maintaining air quality at level that
provides requisite protection against PM10-2.5. That is,
recent air quality data does not suggest that PM10-2.5
concentrations have been increasing as PM2.5 concentrations
have been decreasing. In considering the available PM10-2.5
health effects evidence in this reconsideration, there continue to be
significant uncertainties and limitations, specifically with respect to
the exposure assessment methods used to estimate PM10-2.5
concentrations, that make it difficult to fully assess the public
health implications of revising the primary PM10 standard
even considering the possibility for additional variability in the
relative ratio of PM2.5 to PM10-2.5 in current
PM10 air quality across the U.S. As described in detail
above in section III.A.2 and in the proposal (85 FR 5558, January 27,
2023), the uncertainties and limitations in the health effects evidence
for PM10-2.5 contributed to the determinations in the 2019
ISA that the evidence for key PM10-2.5 health effects is
``suggestive of, but not sufficient to infer, a causal relationship''
or ``inadequate to infer the presence, or absence of a causal
relationship'' (U.S. EPA, 2019a). While the evidence base for
PM10-2.5-related health effects has somewhat expanded since
the 2009 ISA, the Administrator concludes that the evidence remains too
limited to inform judgments regarding whether a more protective primary
PM10 standard is warranted at this time.
Beyond the uncertainties and limitations associated with the
available scientific evidence, the EPA also notes that, while the NCore
monitoring network has been expanded since the time of the last review,
epidemiologic studies available in this review do not use
PM10-2.5 NCore data in evaluating associations between
PM10-2.5 in ambient air and long- or short-term exposures.
In the absence of such evidence, the public health implications of
changes in ambient PM10-2.5 concentrations as
PM2.5 concentrations decrease remain unclear. Therefore, the
EPA continues to recognize this as an area for future research, to
address the existing uncertainties (U.S. EPA, 2022b, section 4.6), and
inform future reviews of the PM NAAQS. Taken together, as at the time
of proposal, the Administrator concludes that these and other
limitations in the PM10-2.5 evidence raised questions as to
whether additional public health improvements would be achieved by
revising the existing PM10 standard, particularly when
considering such judgments along with his decision to retain the
current primary 24-hour PM2.5 standard. Therefore, the EPA
does not agree with the commenters that the currently available air
quality information or scientific evidence support revisions to the
primary PM10 standard in this reconsideration.
Consistent with their comments on the 2020 proposal, some
commenters disagreed with the Administrator's proposed conclusion to
retain the current primary PM10 standard, primarily focusing
their comments on the need for revisions to the form of the standard or
the level of the standard. With regard to comments on the form of the
standard, some commenters assert that the EPA should revise the
standard by adopting a separate form (or a ``compliance threshold'' in
their words)--the 99th percentile, averaged over three years--for the
primary PM10 standard for continuous monitors, which provide
data every day, while maintaining the current form of the standard (one
exceedance, averaged over three years) for 1-in-6 samplers, given the
increased use of continuous monitoring and to ease the burden of
demonstrating exceptional events. These commenters, in support of their
comment, contend that the 99th percentile would effectively change the
form from the 2nd highest to the 4th highest and would allow no more
than three exceedances per year, averaged over three years. These
commenters additionally highlight the EPA's decision in the 1997 review
to adopt a 99th percentile form, averaged over three years, citing to
advantages of a percentile-based form in the Administrator's rationale
in that review. The comments further assert that a 99th percentile form
for the primary PM10 standard is still more conservative
than the form for other short-term NAAQS (e.g., PM2.5 and
NO2).
[[Page 16299]]
First, the EPA has long recognized that the form is an integral
part of the NAAQS and must be selected together with the other elements
(i.e., indicator, averaging time, level) of the NAAQS to ensure the
appropriate stringency and requisite degree of public health
protection. Thus, if the EPA were to change the form according to the
monitoring method it would be establishing two different NAAQS, varying
based on the monitoring method. The EPA has not done this to date, did
not propose such an approach, and declines to adopt it for the final
rule, as we believe such a decision in this final rule is beyond the
scope of the proposal, and that each PM standard should have a single
form, indicator, level and averaging time, chosen by the Administrator
as necessary and appropriate. While certain continuous monitors may be
established and approved as a Federal Equivalent Method (FEM) for
PM10, as an alternative to a Federal Reference Method (FRM),
the use of an FEM is intended as an alternative means of determining
compliance with the NAAQS, not as authorizing a different NAAQS.
Even if the commenters had asked that the change in form be made
without regard to monitoring method, the EPA does not believe such a
change would be warranted. The change in form for continuous monitors
suggested by the commenters, without also lowering the level of such a
standard, would allow more exceedances and thereby reduce the public
health protection provided against exposures to PM10-2.5 in
ambient air, resulting in a less stringent primary PM10
standard than the current standard. These commenters have not provided
new evidence or analyses to support their conclusion that an
appropriate degree of public health protection could be achieved by
allowing the use of an alternative form (i.e., 99th percentile), while
retaining the other elements of the standard.
With regard to the commenters' assertion that an alternate form of
the standard would ease the burden of demonstrating exceptional events,
the EPA recognizes, consistent with the CAA, that it may be appropriate
to exclude monitoring data influenced by ``exceptional'' events when
making certain regulatory determinations. However, the EPA notes that
the cost of implementation of the standards may not be considered by
the EPA in reviewing the standards. The EPA continues to update and
develop documentation and tools to facilitate the implementation of the
2016 Exceptional Events Rule, including new PM2.5
implementation focused products under development that are intended to
assist air agencies with the development of demonstrations for specific
types of exceptional events. With regard to the commenters' specific
concerns for wildfires or high winds, the EPA released updated guidance
documents on the preparation of exceptional event demonstrations
related to wildfires in September 2016, high wind dust events in April
2019, and prescribed fires in August 2019. These guidance documents
outline the regulatory requirements and provide examples for air
agencies preparing demonstrations for wildfires, high wind dust, and
prescribed fire events. For all of the reasons discussed above, the EPA
does not agree with the commenters that the form of the primary
PM10 standard should be revised to a 99th percentile for
continuous monitors.
4. Administrator's Conclusions
This section summarizes the Administrator's considerations and
conclusions related to the current primary PM10 standard. In
establishing primary standards under the Act that are ``requisite'' to
protect the public health with an adequate margin of safety, the
Administrator is seeking to establish standards that are neither more
nor less stringent than necessary for this purpose. In so doing, the
Administrator notes that his final decision in this reconsideration is
a public health policy judgment that draws upon scientific information,
as well as judgments about how to consider the range and magnitude of
uncertainties that are inherent in the information. Accordingly, he
recognizes that his decision requires judgments based on the
interpretation of the evidence that neither overstates nor understates
the strength or limitations of the evidence nor the appropriate
inferences to be drawn. He recognizes, as described in section I.A
above, that the Act does not require that primary standards be set at a
zero-risk level; rather, the NAAQS must be sufficient but not more
stringent than necessary to protect public health, including the health
of sensitive groups with an adequate margin of safety.
Given these requirements, and consistent with the primary
PM2.5 standards discussed above (section II.C.3), the
Administrator's final decision in this reconsideration of the current
primary PM10 standard will be a public health policy
judgment that draws upon the scientific information examining the
health effects of PM10-2.5 exposures, including how to
consider the range and magnitude of uncertainties inherent in that
information. The Administrator's final decision is based on an
interpretation of the scientific evidence that neither overstates nor
understates its strengths and limitations, nor the appropriate
inferences to be drawn.
Having carefully considered advice from the CASAC and public
comments, as discussed above, the Administrator notes that the
fundamental scientific conclusions on health effects of
PM10-2.5 in ambient air that were reached in the 2019 ISA
and summarized in the 2020 PA and 2022 PA remain valid. Additionally,
the Administrator believes the judgments he proposed (85 FR 5558,
January 27, 2023) with regard to the evidence remain appropriate.
Further, in considering the adequacy of the current primary
PM10 standard in this reconsideration, the Administrator has
carefully considered the policy-relevant evidence and conclusions
contained in the 2019 ISA; the rationale and conclusions presented in
the 2020 PA and 2022 PA; the advice and recommendations from the CASAC
in their reviews of the 2019 draft PA and 2021 draft PA; and public
comments, as addressed in section III.B.3 above and in the RTC
document. In the discussion below, the Administrator gives weight to
the conclusions in the 2020 PA and 2022 PA, with which the CASAC has
concurred, as summarized in section III.C of the proposal and takes
note of the key aspects of the rationale for those conclusions that
contribute to his decision in this review. In considering this
information, the Administrator concludes that the preliminary
conclusions and policy judgments supporting his proposed decision
remain valid, and that the current primary PM10 standard
provides requisite protection of public health with an adequate margin
of safety and should be retained. In considering the 2020 PA and 2022
PA evaluations and conclusions, the Administrator notes that, while the
health effects evidence is somewhat expanded since the 2009 ISA as
described in section III.A.2 above, the overall conclusions are
generally consistent with those reached in the 2009 ISA (U.S. EPA,
2020b, section 4.4). In so doing, he additionally notes that the CASAC
supported the preliminary conclusion in the 2019 draft PA and 2021
draft PA that the evidence reviewed in the 2019 ISA does not call into
question the public health protection provided by the current primary
PM10 standard against PM10-2.5 exposures and
concurs that it is appropriate to consider retaining the current
primary PM10 standard (Cox,
[[Page 16300]]
2019b, p. 13 of consensus responses; Sheppard, 2022a, p. 4 of consensus
letter).
As noted below, the scientific evidence for PM10-2.5-
related health effects has expanded somewhat since the 2012 review, in
particular for long-term exposures. The Administrator recognizes,
however, that there are a number of uncertainties and limitations
associated with the available information, as described in the proposal
(85 FR 5558, January 27, 2023) and below. With regard to the current
evidence on PM10-2.5-related health effects, the
Administrator takes note of recent epidemiologic studies that continue
to report positive associations with mortality and morbidity in cities
across North America, Europe, and Asia, where PM10-2.5
sources and composition are expected to vary widely. While significant
uncertainties remain, as described below, the Administrator recognizes
that this expanded body of evidence has broadened the range of effects
that have been linked with PM10-2.5 exposures. These studies
provide an important part of the scientific foundation supporting the
2019 ISA's revised causality determinations (and new determinations)
for long-term PM10-2.5 exposures and mortality,
cardiovascular effects, metabolic effects, nervous system effects, and
cancer (U.S. EPA, 2019a; U.S. EPA, 2022b, section 4.2). Drawing from
his consideration of this evidence, the Administrator concludes that
the available scientific information supports a decision to maintain a
primary PM10 standard to provide public health protection
against PM10-2.5 exposures, regardless of location, source
of origin, or particle composition. With regard to uncertainties in the
evidence, the Administrator first notes that a number of limitations
were identified in the 2012 review related to: (1) Estimates of ambient
PM10-2.5 concentrations used in epidemiologic studies; (2)
limited evaluation of copollutant models to address the potential for
confounding; and (3) limited experimental studies supporting biological
plausibility for PM10-2.5-related effects. Despite the
expanded body of evidence for PM10-2.5 exposures and health
effects assessed in the 2019 ISA, the Administrator recognizes that
uncertainties remain, similar to those in the 2012 review. As
summarized in section III.A.2 above and in responding to public
comments, uncertainties in the available scientific evidence continue
to include those associated with the exposure estimates used in
epidemiologic studies, the independence of the PM10-2.5
health effect associations, and the biologically plausible pathways for
PM10-2.5 health effects (U.S. EPA, 2022b, section 4.3).
These uncertainties contribute to the 2019 ISA determinations that the
evidence is ``suggestive of, but not sufficient to infer'' causal
relationships (U.S. EPA, 2019a). The Administrator recognizes that the
NAAQS must allow for a margin of safety but also places emphasis on
evidence supporting ``causal'' or ``likely to be causal'' relationships
(as described in sections II.A.2 and III.A.2 above). Finding that there
is too much uncertainty that a more stringent standard would improve
public health, the Administrator judges that the available evidence
provides support for his conclusion that the current standard provides
the requisite level of protection from the effects of
PM10-2.5. In making this judgment, the Administrator
considers whether this level of protection is more than what is
requisite and whether a less stringent standard would be appropriate to
consider. He notes that there continues to be uncertainty associated
with the evidence, as reflected by the ``suggestive of, but not
sufficient to infer'' causal determinations. The Administrator
recognizes that the CAA requirement that primary standards provide an
adequate margin of safety, as summarized in section I.A above, is
intended to address uncertainties associated with inconclusive
scientific evidence and technical information, as well as to provide a
reasonable degree of protection against hazards that research has not
yet identified. In light of these considerations and the current body
of evidence, including uncertainties and limitations, the Administrator
concludes that a less stringent standard would not provide the
requisite protection of public health, including an adequate margin of
safety. The Administrator also considers whether the level of
protection associated with the current standard is less than what is
requisite and whether a more stringent standard would be appropriate to
consider. In so doing, the Administrator considers, as discussed above,
the level of protection offered from exposures for which public health
implications are less clear. In so doing, he again notes the
significant uncertainties and limitations that persist in the
scientific evidence. In particular, he notes limitations in the
approaches used to estimate ambient PM10-2.5 concentrations
in epidemiologic studies, limited examination of the potential for
confounding by co-occurring pollutants, and limited support for the
biological plausibility of the serious effects reported in many
epidemiologic studies that are reflected by the ``suggestive of, but
not sufficient to infer'' causal determinations. Thus, in light of the
currently available information, including the uncertainties and
limitations of the evidence base available to inform his judgments
regarding protection against PM10-2.5-related effects, the
Administrator does not find it appropriate to increase the stringency
of the standard in order to provide the requisite public health
protection. Rather, he judges it appropriate to maintain the level of
protection provided by the current primary PM10 standard for
PM10-2.5 exposures and he does not judge that the available
information and the associated uncertainties indicate the need for a
greater level of public health protection.
In reaching his conclusions on the primary PM10
standard, the Administrator also considers advice from the CASAC. In
their comments, the CASAC noted that uncertainties that were identified
in the 2012 review persist in the evidence for PM10-2.5-
related health effects (Cox, 2019b, p. 13 of consensus responses;
Sheppard, 2022a, p. 4 of consensus letter) In considering these
comments, the Administrator takes note of the CASAC consideration of
the evidence, and associated uncertainties, and its conclusion that the
evidence reviewed in the 2019 ISA does not call into question the
adequacy of the public health protection afforded by the current
primary PM10 standard (Cox, 2019b, p. 3 of letter; Sheppard,
2022a, p. 4 of consensus letter). The Administrator further notes the
unanimous conclusions of the CASAC that evidence supports consideration
of retaining the current primary PM10 standard (Cox, 2019b,
p. 3 of consensus letter; Sheppard, 2022a, p. 4 of consensus letter).
In addition to the CASAC's advice, the Administrator also considers
public comments, the majority of which supported retaining the primary
PM10 standard, citing to and agreeing with the
Administrator's rationale for his proposed decision. The Administrator
also recognizes that a few public commenters supported revising the
primary PM10 standard in order to provide increased
protection against PM10-2.5-related health effects.
The Administrator also notes that the scientific record for his
decision on the primary PM10 standard is the same as the
record before the then-Administrator in 2020, as the scope of the ISA
Supplement focused on health effect categories where the 2019 ISA
concluded a causal relationship (i.e.,
[[Page 16301]]
short- and long-term PM2.5 exposure and cardiovascular
effects and mortality). Therefore, because no health outcome categories
for short- or long-term PM10-2.5 exposure in the 2019 ISA
were greater than ``suggestive of, but not sufficient to infer, a
causal relationship'', the ISA Supplement did not evaluate studies
published after the literature cutoff date of the 2019 ISA related to
PM10-2.5 exposures and health effects. The Administrator
further notes his decision is consistent with the decision of the prior
Administrator in 2020 to retain the primary PM10 standard.
With regard to the indicator for the primary PM10
standard, the Administrator recognizes that the 2022 PA notes that the
evidence continues to support retaining the PM10 indicator
to provide public health protection against PM10-2.5-related
effects. He notes that, consistent with the approaches in previous
reviews, a standard with a PM10 mass-based indicator, in
conjunction with a PM2.5 mass-based standard, will result in
controlling allowable concentrations of PM10-2.5. The
Administrator also takes note of the 2019 ISA comparison that showed
that the relative contribution of PM2.5 and
PM10-2.5 to PM10 concentrations can vary across
the U.S. by region and season, with urban locations having a somewhat
higher contribution of PM2.5 contributing to PM10
concentrations than PM10-2.5 (U.S. EPA, 2019a, section
2.5.1.1.4, Table 2-7). In these urban locations, where PM2.5
concentrations are somewhat higher than in rural locations, the
toxicity of the PM10 may be higher due to contaminating
PM2.5. Further, although uncertainties with the evidence
persist, the strongest health effects evidence associated with
PM10-2.5 comes from epidemiologic studies conducted in urban
areas. He also notes that the CASAC agreed with the EPA's conclusions
that a PM10 indicator remained appropriate (Cox, 2019b, p.
13 of consensus responses; Sheppard, 2022a, p. 4 of letter). In light
of this information, the Administrator concludes that the
PM10 indicator remains appropriate and provides protection
from exposure to all coarse PM, regardless of location, source of
origin, or particle composition.
Similarly, with regard to averaging time, form, and level of the
standard, the Administrator takes note of uncertainties in the
available evidence and information and continues to find that the
current standard, as defined by in all of its elements, is requisite.
As an initial matter, the Administrator notes that the current primary
PM10 standard, with its level of 150 [micro]g/m\3\, 24-hour
averaging time, not to be exceeded more than once per year on average
over three years, is intended to protect against short-term peak
PM10-2.5 exposures. In so doing, while the Administrator
notes that changes in PM2.5 concentrations in ambient air
can influence the contribution of the fine and coarse fractions to
PM10 mass, such that reductions in PM2.5
concentrations can lead to more allowable PM10-2.5 under the
current primary PM10 standard, he recognizes that there is
no new information available in this reconsideration to suggest that
the public health protection provided by the current standard is not
requisite or that a more stringent standard is warranted at this time.
The Administrator concludes that, particularly in light of his decision
to retain the primary 24-hour PM2.5 standard with its level
of 35 [micro]g/m\3\ as described in section II.B.4 above, the primary
PM10 standard would be expected to maintain
PM10-2.5 concentrations in ambient air below those that have
been considered to be associated with serious health effects in past
NAAQS reviews. The Administrator also notes that while the scientific
evidence available in the 2019 ISA has expanded since the completion of
the 2009 ISA, he concludes that this information does not provide
support for the causal or likely to be causal relationships upon which
he places the greatest weight in considering the adequacy of the
current standards. He further concludes that the uncertainties and
limitations of the scientific evidence, along with the absence of
information to inform a quantitative exposure or risk assessment, make
it difficult to reach decisions regarding whether a more protective
standard is warranted at this time. He has additionally considered the
public comments regarding revisions to these elements of the standard
and continues to judge that the existing level and the existing form,
in all its aspects, together with the other elements of the existing
standard provide an appropriate level of public health protection. For
all of the reasons discussed above and recognizing the CASAC's
conclusion that the current evidence provides support for retaining the
current standard, the Administrator concludes that the current primary
PM10 standard (in all of its elements) is requisite to
protect public health with an adequate margin of safety from effects of
PM10-2.5 in ambient air and should be retained without
revision.
C. Decision on the Primary PM10 Standard
For the reasons discussed above and considering information and
assessments presented in the 2019 ISA and the 2022 PA, the advice from
the CASAC, and public comments, the Administrator concludes that the
current primary PM10 standard is requisite to protect public
health with an adequate margin of safety, including the health of at-
risk populations, and is retaining the current standard without
revision.
IV. Communication of Public Health
A. Air Quality Index Overview
Information about the public health implications of ambient
concentrations of criteria pollutants is communicated to the public
using the Air Quality Index (AQI) reported on the EPA's AirNow
website.\133\ The current AQI has been in use since its inception in
1999.\134\ It provides useful, timely, and easily understandable
information about the daily degree of pollution. The goal of the AQI is
to establish a nationally uniform system of indexing pollution
concentrations for ozone, carbon monoxide, nitrogen dioxide, PM, and
sulfur dioxide. The AQI is recognized internationally as a proven tool
to effectively communicate air quality information to the public as
demonstrated by the fact that many countries have created similar
indices based on the AQI.
---------------------------------------------------------------------------
\133\ See http://www.airnow.gov/.
\134\ In 1976, the EPA established a nationally uniform air
quality index, then called the Pollutant Standard Index (PSI), for
use by State and local agencies on a voluntary basis (41 FR 37660,
September 7, 1976; 52 FR 24634, July 1,1987). In August 1999, the
EPA adopted revisions to this air quality index (64 FR 42530, August
4, 1999) and renamed the index the AQI.
---------------------------------------------------------------------------
The AQI converts an individual pollutant concentration in a
community's air to a number on a scale from 0 to 500. Reported AQI
values for specific pollutants enable the public to know whether air
pollution levels in a particular location are characterized as good (0-
50), moderate (51-100), unhealthy for sensitive groups (101-150),
unhealthy (151-200), very unhealthy (201-300), or hazardous (301+).
Across criteria pollutants, the AQI value of 100 typically corresponds
to the level of the short-term (e.g., 24-hour, 8-hour, or 1-hour
standard) NAAQS for each pollutant. Below an index value of 100, an
intermediate value of 50 is defined either as the level of the annual
standard if an annual standard has been established (e.g.,
PM2.5, nitrogen dioxide), a
[[Page 16302]]
concentration equal to one-half the value of the 24-hour standard used
to define an index value of 100 (e.g., carbon monoxide), or a
concentration based directly on health effects evidence (e.g., ozone).
An AQI value greater than 100 means that a pollutant is in one of the
unhealthy categories (i.e., unhealthy for sensitive groups, unhealthy,
very unhealthy, or hazardous). An AQI value at or below 100 means that
a pollutant concentration is in one of the satisfactory categories
(i.e., moderate or good). The scientific evidence on pollutant-related
health effects for each NAAQS review support decisions related to
pollutant concentrations at which to set the various AQI breakpoints,
which delineate the AQI categories for each individual pollutant (i.e.,
the pollutant concentrations corresponding to index values of 150, 200,
300, and 500). The AQI is reported three ways by the EPA and State,
local and Tribal agencies, all of which are useful and complementary.
The daily AQI is reported for the previous day and used to observe
trends in community air quality, the AQI forecast helps people plan
their outdoor activities for the next day, and the near-real-time AQI,
or NowCast AQI, tells people whether it is a good time for outdoor
activity.
Historically, State and local agencies have primarily used the AQI
to provide general information to the public about air quality and its
relationship to public health. For more than two decades, many State
and local agencies, as well as the EPA and other Federal agencies, have
been developing new and innovative programs and initiatives to provide
more information related to air quality and health messaging to the
public in a more timely way. These initiatives, including air quality
forecasting, near real-time data reporting through the AirNow website,
use of data from air quality sensors on the EPA and U.S. Forest
Service's (USFS) Fire and Smoke Map, and air quality action day
programs, provide useful, up-to-date, and timely information to the
public about air pollution and its health effects. Such information can
help the public learn when their well-being may be compromised, so they
can take actions to avoid or to reduce exposures to ambient air
pollution at concentrations of concern. This information can also
encourage the public to take actions that will reduce air pollution on
days when concentrations are projected to be of concern to local
communities (e.g., air quality action day programs can encourage
individuals to drive less or carpool).
B. Air Quality Index Category Breakpoints for PM2.5
Recognizing the scientific information available and current AQI
reporting practices, the EPA proposed several revisions to the AQI
PM2.5 breakpoints. EPA solicited and received comments on
these proposed revisions. Upon reviewing the information in the
proposal and considering the comments received EPA is making final
revisions to the AQI category breakpoints for PM2.5. This
section summarizes the proposed revisions, which can be read in full in
the proposal (88 FR 5638, January 27, 2023), significant comments, and
final revisions.
1. Summary of Proposed Revisions
One purpose of the AQI is to communicate to the public when air
quality is poor and thus when they should consider taking actions to
reduce their exposures. The higher the AQI value, the higher the level
of air pollution and the greater the health concern. In recognition of
the scientific information available that is informing the
reconsideration of the 2020 final decision on the primary
PM2.5 standards, including a number of new controlled human
exposure and epidemiologic studies published since the completion of
the 2009 ISA, as well as additional epidemiologic studies from other
peer reviewed documents that evaluate the health effects of wildfire
smoke exposure and that can inform the selection of AQI breakpoints at
higher PM2.5 concentrations,\135\ the EPA proposed to make
two sets of changes to the PM2.5 sub-index of the AQI.
First, the EPA proposed to continue to use the approach used in the
revisions to the AQI in 2012 (77 FR 38890, June 29, 2012) of setting
the lower breakpoints (50, 100 and 150) to be based on the levels of
the primary PM2.5 annual and 24-hour standards and proposed
to revise the lower breakpoints to be consistent with changes to the
primary PM2.5 standards that are part of this
reconsideration. Second, the EPA proposed to revise the upper AQI
breakpoints (200 and above) and to replace the linear-relationship
approach used in 1999 to set these breakpoints, with an approach that
more fully considers the PM2.5 health effects evidence from
controlled human exposure and epidemiologic studies that have become
available in the last 20 years (64 FR 42530, August 4, 1999).
---------------------------------------------------------------------------
\135\ In evaluating the scientific evidence available to inform
decisions regarding the AQI breakpoints, the EPA considered studies
that were included as a part of the 2019 ISA and ISA Supplement, but
also considered other studies that were not included as a part of
the review of the air quality criteria. The ISAs have specific
criteria for study inclusion and consideration in reaching
conclusions regarding causal relationships, and some studies that
may not have met those criteria (e.g., epidemiologic studies that
evaluate the health effects of wildfire smoke exposure that would
have higher PM2.5 concentrations, which are outside of
the scope of the ISA) were identified as studies that could be used
to inform decisions on the AQI, particularly for the upper
breakpoints.
---------------------------------------------------------------------------
a. Air Quality Index Values of 50, 100 and 150
With respect to the lower AQI breakpoints in the proposal (88 FR
5638, January 27, 2023), the EPA proposed to conclude that it is
appropriate to continue setting these breakpoints to be consistent with
the primary annual and 24-hour PM2.5 standard levels. The
lowest AQI value of 50 provides the breakpoint between the ``good'' and
``moderate'' categories. At and below this concentration, air quality
is considered ``good'' for everyone. Above this concentration, in the
``moderate'' category, the AQI contains advisories for unusually
sensitive individuals. The EPA has historically set this breakpoint at
the level of the primary annual PM2.5 standard. In doing so,
the EPA has recognized that: (1) The annual standard is set to provide
protection to the public, including at-risk populations, from
PM2.5 concentrations, which, when experienced on average for
a year, have the potential to result in adverse health effects; and (2)
the AQI exposure period represents a shorter exposure period (e.g., 24-
hour (or less)) while focusing on the most sensitive individuals. The
EPA saw no basis for deviating from this approach in this
reconsideration. Thus, the EPA proposed to set the AQI value of 50 at a
daily (i.e., 24-hour) average concentration equal to the level of the
primary annual PM2.5 standard that is promulgated.
The historical approach to setting an AQI value of 100, which is
the breakpoint between the ``moderate'' and ``unhealthy for sensitive
groups'' categories, and above which advisories are generated for
sensitive groups, is to set it at the same level as the primary 24-hour
PM2.5 standard. In so doing, the EPA has recognized that the
primary 24-hour PM2.5 standard is set to provide protection
to the public, including at-risk populations, from short-term exposures
to PM2.5 concentrations that have the potential to result in
adverse health effects. Given this, it is appropriate to generate
advisories for sensitive groups at concentrations above this level. In
the past, State, local, and Tribal air quality agencies have expressed
strong support for this approach (78 FR 3086, January 15, 2013). The
EPA saw no basis to deviate
[[Page 16303]]
from this approach in this reconsideration. In the proposal (88 FR
5638, January 27, 2023), the EPA proposed to retain the current primary
24-hour PM2.5 standard with its level of 35 [mu]g/m\3\ but
took comment on revising the level of that standard to 25 [mu]g/m\3\
(section II.D.3.b). Thus, the EPA proposed to retain the AQI value of
100 set at the level of the current primary 24-hour PM2.5
standard concentration of 35 [mu]g/m\3\ (i.e., 24-hour average).
With respect to an AQI value of 150, which is the breakpoint
between the ``unhealthy for sensitive groups'' and ``unhealthy
categories,'' this breakpoint concentration in this reconsideration is
based upon the considering the same health effects information, as
assessed in the 2019 ISA and ISA Supplement and described in section II
above, that informs the proposed decisions on the level of the 24-hour
standard and the AQI value of 100. Previously, the Agency has used a
proportional adjustment in which the AQI value of 150 was set
proportionally to the AQI value of 100. This proportional adjustment
inherently recognizes that the available epidemiologic studies provide
no evidence of discernible thresholds, below which effects do not occur
in either sensitive groups or in the general population, that could
inform conclusions regarding concentrations at which to set this
breakpoint. Given that the epidemiologic evidence continues to be the
most relevant health effects evidence for informing this range of AQI
values, the EPA saw no basis to deviate from this approach in this
reconsideration. Therefore, the EPA proposed to set an AQI value of 150
proportionally, depending on the breakpoint concentration of the AQI
value of 100 (i.e., 55.4 for a 24-hour standard of 35 [mu]g/m\3\).
b. Air Quality Index Values of 200 and Above
In the proposal (88 FR 5639, January 27, 2023), the EPA summarized
the history of setting the AQI values of 300 and above in the 1999 rule
(64 FR 42530, August 4, 1999) and established breakpoints for
PM2.5 in that range. In general, the AQI values between 100
and 500 were based on PM2.5 concentrations that generally
reflected a linear relationship between increasing index values and
increasing PM2.5 concentrations.\136\ It was found that this
linear relationship was generally consistent with the health effects
evidence, which suggested that as PM2.5 concentrations
increase, increasingly larger numbers of people are likely to
experience serious health effects in this range of PM2.5
concentrations (64 FR 42536, August 4, 1999). For the AQI breakpoint of
500, the concentration was based on the method used to establish a
previously existing PM10 breakpoint that was informed by
studies conducted in London using the British Smoke method, which uses
a different particle size cutpoint as noted in the proposal (88 FR
5639, January 27, 2023). Due to limited ambient PM2.5
monitoring data available at that time, the decision on the 500 value
concentration for PM2.5 was based on the stated assumption
that PM concentrations measured by the British Smoke method were
approximately equivalent to PM2.5 concentrations (64 FR
42530, August 4, 1999). Given that the British Smoke method has a
larger particle size cutpoint than the current PM2.5
monitoring method, which has a cutpoint of 2.5 microns, a concentration
of 500 [mu]g/m\3\ based on the British Smoke method would be equivalent
to a lower PM2.5 concentration. With respect to the upper
breakpoints of the AQI, the EPA has historically been concerned about
establishing these upper breakpoints using evidence based on larger
size fractions of PM, given that PM2.5 is the indicator for
the AQI. While monitoring data for higher PM2.5
concentrations in ambient air has been available for many years, the
health effects evidence has only recently become available for
consideration in informing decisions on the upper breakpoints of the
AQI.
---------------------------------------------------------------------------
\136\ The AQI breakpoint at 150 was originally set in 1999 to be
linearly related to the concentrations at the 100 and 500
breakpoints but then revised in 2012 to be proportional to the AQI
breakpoint concentration at 100 (78 FR 3181, January 15, 2013).
---------------------------------------------------------------------------
As part of this reconsideration, the EPA recognized that the health
effects evidence associated with PM2.5 exposure has greatly
expanded in recent years. Multiple controlled human exposure studies
have become available that provide information about health effects
across a range of concentrations. While many of the new studies
evaluated in the 2019 ISA focused on examining health effects
associated with exposure to lower PM2.5 concentrations,
there are also several new controlled human exposure studies that
provide information about the health effects observed in study
participants at concentrations well above the standard levels.
Additionally, there are also epidemiologic studies now available and
evaluated in other Agency peer-reviewed documents that can inform
health effects associated with higher PM2.5 concentrations
(U.S. EPA, 2021b).\137\ Thus, the EPA concluded that it is appropriate
to reevaluate the upper AQI breakpoints, taking into account the
expanded body of scientific evidence, particularly given several new
epidemiologic studies conducted during high pollution events like
wildfires and multiple controlled human exposure studies. While it
remains unclear the exact PM2.5 concentrations at which
specific health effects occur, the more recent studies do provide more
refined information about the concentration range in which these
effects might occur in some populations. These studies provide support
for coherence of effects across scientific disciplines and potentially
biologically plausible pathways for the overt population-level health
effects observed in epidemiologic studies. Therefore, taking into
account the short exposure time period in these studies (e.g., 1-6
hours) and that the studies generally do not include at-risk (or
sensitive) populations, but rather young, healthy adults, these
studies, in conjunction with information from epidemiologic studies,
the EPA preliminarily concluded it would be appropriate to be more
cautionary and offer advisories to the public for reducing exposures at
lower concentrations than recommended with the current AQI breakpoints.
---------------------------------------------------------------------------
\137\ In this reconsideration, the controlled human exposure
studies were evaluated in the 2019 ISA, whereas the epidemiologic
studies of wildfire smoke exposures were included in the EPA
Comparative Assessment of the Impacts of Prescribed Fire Versus
Wildfire (CAIF): A Case Study in the Western U.S. (U.S. EPA 2021b).
---------------------------------------------------------------------------
The AQI value of 200 is the breakpoint between the ``unhealthy''
and ``very unhealthy'' categories. At AQI values above 200, the AQI
would be providing a health warning that the risk of anyone
experiencing a health effect following short-term exposures to these
PM2.5 concentrations has increased. To inform proposed
decisions on this breakpoint, the EPA takes note of studies indicating
the potential for respiratory or cardiovascular effects that are on
their own representative of or are on the biologically plausible
pathway to more serious health outcomes (e.g., emergency department
visits, hospital admissions). The controlled human exposure studies
evaluated in the 2009 and 2019 ISAs provide evidence of inflammation as
well as cardiovascular effects in healthy subjects at and above 120
[micro]g/m\3\. For example, Ramanathan et al. (2016) observed a
transient reduction in antioxidant/anti-inflammatory function after
exposing healthy young subjects to a mean concentration of 150
[micro]g/m\3\ of PM2.5 for 2 hours. Urch et al.
[[Page 16304]]
(2010) also reported increased markers of inflammation when exposing
both asthmatic and non-asthmatic subjects to a mean concentration of
140 [micro]g/m\3\ of PM2.5 for 3 hours. In studies
specifically examining cardiovascular effects, Ghio et al. (2000) and
Ghio et al. (2003) exposed healthy subjects to a mean concentration of
120 [micro]g/m\3\ for 2 hours and reported significantly increased
levels of fibrinogen, a marker of coagulation that increases during
inflammation. Sivagangabalan et al. (2011) exposed healthy subjects to
a mean concentration of 150 [micro]g/m\3\ of PM2.5 for 2
hours and noted an increased QT interval (3.4 1.4)
indicating some evidence for conduction abnormalities, an indicator of
possible arrhythmias. Lastly, Brook et al. (2009) reported a transient
increase of 2.9 mm Hg in diastolic blood pressure in healthy subjects
during the 2-hour exposure to a mean concentration of 148 [micro]g/m\3\
of PM2.5.
In addition to epidemiologic studies evaluated in the 2019 ISA that
analyzed exposures at ambient PM2.5 concentrations, there
are a number of recent epidemiologic studies focusing on wildfire smoke
that have become available that were evaluated in the EPA's recently
released peer-reviewed assessment on wildland fire (U.S. EPA, 2021b).
One of these studies, Hutchinson et al. (2018), conducted a
bidirectional case-crossover analysis to examine associations between
wildfire-specific PM2.5 exposure and respiratory-related
healthcare encounters (i.e., ED visits, inpatient hospital admissions,
and outpatient visits) prior and during the 2007 San Diego wildfires.
This study found positive and significant associations to
PM2.5 exposures and respiratory-related healthcare
encounters. Further, during the initial 5-day period of the wildfire
event, the study observed that there was evidence of increases in a
number of respiratory-related outcomes particularly ED visits for
asthma, upper respiratory infection, respiratory symptoms, acute
bronchitis, and all respiratory-related visits (Hutchinson et al.,
2018). When examining the air quality during the wildfire event,
PM2.5 concentrations were highest during the initial five
days of the wildfire, with 24-hour average PM2.5
concentrations of 89.1 [micro]g/m\3\ across all zip codes and with the
highest 24-hour average of 160 [micro]g/m\3\ on the first day
(Hutchinson et al., 2018).
When considering this collective body of evidence from controlled
human exposure and epidemiologic studies, the Agency proposed to set an
AQI value of 200 at a daily (i.e., 24-hour average) concentration of
PM2.5 of 125 [mu]g/m\3\. As discussed above and in the
proposal (88 FR 5640, January 27, 2023), this concentration is at the
lower end of the concentrations consistently shown to be associated
with respiratory and cardiovascular effects in controlled human
exposure studies following short-term exposures (e.g., 2-3 hours) and
in young, healthy adults (Ghio et al., 2000; Ghio et al., 2003; Urch et
al., 2010; Ramanathan et al., 2016; Sivagangabalan et al., 2011; and
Brook et al., 2009) and also within the range of 5-day average and
maximum concentrations observed to be associated with respiratory-
related outcomes following exposure to wildfire smoke (Hutchinson et
al., 2018).
The AQI value of 300 denotes the breakpoint between the ``very
unhealthy'' and ``hazardous'' categories, and thus marks the beginning
of the ``hazardous'' AQI category. At AQI values above 300, the AQI
provides a health warning that everyone is likely to experience effects
following short-term exposures to these PM2.5
concentrations. To inform decisions on this AQI breakpoint, the EPA
takes note of controlled human exposure studies that consistently show
subclinical effects which are often associated with more severe
cardiovascular outcomes. As discussed above, Brook et al. (2009)
reported a transient increase of 2.9 mm Hg in diastolic blood pressure
in healthy subjects during the 2-hour exposure to a mean concentration
of 148 [micro]g/m\3\ of PM2.5. Bellavia et al. (2013)
exposed healthy subjects to an average PM2.5 concentration
of 242 [micro]g/m\3\ for 2 hours and reported increased systolic blood
pressure (2.53 mm Hg). Tong et al. (2015) exposed healthy subjects to
an average PM2.5 concentration of 253 [micro]g/m\3\ for 2
hours and observed a significant increase in diastolic blood pressure
(2.1 mm Hg) and a nonsignificant increase in systolic blood pressure
(2.5 mm Hg). Lucking et al. (2011) reported impaired vascular function
and increased potential for coagulation when exposing healthy subjects
to diesel exhaust (DE) with an average PM2.5 concentration
of 320 [micro]g/m\3\ for a duration of 1 hour.\138\ These studies all
provided evidence of impaired vascular function, including
vasodilatation impairment and increased thrombus formation, with Tong
et al. (2015), Bellavia et al. (2013), Brook et al. (2009) all
reporting increases in blood pressure. Additionally, Behbod et al.
(2013) reported increased inflammatory markers following a 2-hour
exposure to an average PM2.5 concentration of 250 [micro]g/
m\3\ in healthy subjects.
---------------------------------------------------------------------------
\138\ Although participants in Lucking et al. (2011) were
exposed to diesel exhaust (DE), the authors also conducted analyses
using a particle trap, and as noted in the 2019 ISA, this type of
study design allows for the assessment of the role of
PM2.5 on the health effects observed by removing PM from
the DE mixture.
---------------------------------------------------------------------------
In addition to the controlled human exposure studies discussed
above, the epidemiologic study conducted by DeFlorio-Barker et al.
(2019) examined the relationship between wildfire smoke and
cardiopulmonary hospitalizations among adults 65 years of age and older
from 2008-2010 in 692 U.S. counties. The authors reported a 2.22%
increase in all-cause respiratory hospitalizations on wildfire smoke
days for a 10 [micro]g/m\3\ increase in 24-hour average
PM2.5 concentrations (DeFlorio-Barker et al., 2019). The
maximum 24-hour average concentration in this study on wildfire smoke
days was 212.5 [micro]g/m\3\ (DeFlorio-Barker et al., 2019). In
considering this study, the EPA notes the increased probability that
even healthy adults experience effects at this maximum exposure
concentration, particularly given that this maximum concentration is
near the exposure concentrations in controlled human exposure studies
that consistently reported evidence of impaired vascular function and
several that reported increases in blood pressure in healthy adults
following 2-hour exposures.
Based on the information discussed above and in the proposal (88 FR
5640, January 27, 2023), the EPA proposed to revise the 300 level of
the AQI, which marks the beginning of the ``hazardous'' AQI category,
to a concentration that is consistent with the PM2.5
concentrations associated with health effects as reported in the
controlled human exposure (Brook et al., 2009; Bellavia et al., 2013;
Tong et al., 2015; Behbod et al., 2013) and epidemiologic studies
(DeFlorio-Barker et al. (2019). Specifically, the Agency proposed to
set an AQI value of 300 at a daily (i.e., 24-hour average)
PM2.5 concentration of 225 [mu]g/m\3\. This concentration
falls between the 2-hour average concentrations reported in controlled
human exposure studies found to be consistently associated, in healthy
adults, with impaired vascular function and/or increases in blood
pressure, which could both be a precursor to more severe cardiovascular
effects following short-term (1- to 2-hour) exposures, and the maximum
24-hour average PM2.5 concentrations on wildfire smoke days
reported in the epidemiologic study conducted by DeFlorio-Barker et al.
(2019).
[[Page 16305]]
c. Air Quality Index Value of 500
Lastly, the EPA also proposed revisions to the 500 value of the
AQI. The 500 value of the AQI is within the ``hazardous'' category but
is specified and used to calculate the slope of the AQI values in the
``hazardous category'' above and below AQI values of 500. In the past,
this breakpoint had a very prominent role in determining the current
upper AQI values given that it was used as part of the linear
relationship with the concentration at the AQI value of 100 to
determine the AQI values of 200 and 300 in 1999 (64 FR 42530, August 4,
1999).
As discussed above and in the proposal (88 FR 5641, January 27,
2023), the current breakpoint concentration for the 500 value of the
AQI was set in 1999 at a 24-hour average PM2.5 concentration
of 500 [mu]g/m\3\ and was based on studies conducted in London using
the British Smoke method, which used a different particle size cutpoint
and likely overestimated the PM2.5 concentration. In looking
to improve upon that approach, the EPA considered several recent
controlled human exposure studies that observe health effects that are
on the biologically plausible pathway to more severe cardiovascular
outcomes and note that these seem to follow exposures to high
PM2.5 concentrations that are well above those typically
observed in ambient air. More specifically, in controlled human
exposure studies, Vieira et al. (2016a) and Vieira et al. (2016b)
exposed healthy subjects and subjects with heart failure to diesel
exhaust (DE) with a mean PM2.5 concentration of 325
[micro]g/m\3\ for 21 minutes and reported decreased stroke volume, and
increased arterial stiffness (an indicator of endothelial dysfunction)
in both healthy and heart failure subjects.\139\ Also as summarized
above and discussed in the proposal (88 FR 5641, January 27, 2023),
Lucking et al. (2011) exposed healthy subjects to DE with a mean
PM2.5 concentration of 320 [micro]g/m\3\ for 1 hour.\140\
Epidemiologic studies have linked the types of cardiovascular effects
observed in these controlled human exposure studies with the
exacerbation of ischemic heart disease (IHD) and heart failure as well
as myocardial infarction (MI) and stroke.
---------------------------------------------------------------------------
\139\ These effects were attenuated when the DE was filtered, to
reduce PM2.5 concentrations, indicating the effects were
likely associated with PM2.5 exposure.
\140\ When applying a particle trap, PM2.5
concentrations were reduced, and effects associated with
cardiovascular function including impaired vascular function, as
measured by vasodilatation and thrombus formation were attenuated
indicating associations with PM2.5.
---------------------------------------------------------------------------
In addition to the controlled human exposure studies discussed in
the proposal (88 FR 5641, January 27, 2023) and summarized above,
recent epidemiologic studies examining the relationship between
concentrations of PM2.5 during wildfires and respiratory
health also informed the proposed decisions on the concentration for
the AQI value of 500. As discussed in the proposal (88 FR 5641, January
27, 2023) and summarized earlier in this section, Hutchinson et al.
(2018) reported increases in a number of respiratory-related ED visits
for asthma, upper respiratory infection, respiratory symptoms, acute
bronchitis, and all combined respiratory-related visits based on data
from Medi-Cal claims for emergency department presentations, inpatient
hospitalizations, and outpatient visits during the initial 5-day period
of the 2007 San Diego fire. During the initial 5-day window,
PM2.5 concentrations were found to be at their highest with
the 95th percentile of 24-hour average concentrations of 333 [micro]g/
m\3\.
Although studies of short-term (i.e., daily) exposures to wildfire
smoke are more informative in considering alternative level for the AQI
value of 500 since they mirror the 24-hour exposure timeframe,
additional information from epidemiologic studies of longer-term
exposures (i.e., over many weeks) during wildfire events can provide
supporting information. As discussed in the proposal (88 FR 5641,
January 27, 2023) and summarized here, Orr et al. (2020) conducted a
longitudinal study that reported exposure to wildfire smoke from a
multi-month fire resulted in reduced lung function in subsequent years
and concluded that exposure to high PM2.5 concentrations
during a multi-week fire event may lead to health consequences, such as
declines in lung function. During the 2017 wildfire event (August 1 to
September 19, 2017), Orr et al. (2020) reported that many days during
the multi-month fire had PM2.5 concentrations above 300
[micro]g/m\3\, resulting in a daily average PM2.5
concentration of 220.9 [mu]g/m\3\ with a maximum PM2.5
concentration of 638 [micro]g/m\3\.
The controlled human exposure studies provide biological
plausibility for results of epidemiologic studies that document
increases in respiratory-related health care events during the
wildfires. The collective evidence from controlled human exposure and
epidemiologic studies, which includes decreases in stroke volume,
increased arterial stiffness, impaired vascular function and
respiratory-related healthcare encounters provide health-based evidence
that informed the proposed decisions on the level of the AQI value of
500. Given the concentrations observed in these studies, the Agency
proposed to revise the AQI value of 500 to a level set at a daily
(i.e., 24-hour average) PM2.5 concentration of 325 [mu]g/
m\3\. This concentration is at or below the lowest concentrations
observed in the controlled human exposure studies associated with more
severe effects discussed above and also at the low end of the daily
concentrations observed in the epidemiologic studies conducted by
Hutchinson et al. (2018) and Orr et al. (2020).
Table 1 below summarizes the proposed breakpoints for the
PM2.5 sub-index.
Table 1--Proposed Breakpoints for PM2.5 Sub-Index
----------------------------------------------------------------------------------------------------------------
Proposed breakpoints ([mu]g/
AQI category Index values m\3\, 24-hour average)
----------------------------------------------------------------------------------------------------------------
Good............................................................... 0-50 0.0-(9.0-10.0)
Moderate........................................................... 51-100 (9.1-10.1)-35.4
Unhealthy for Sensitive Groups..................................... 101-150 35.5-55.4
Unhealthy.......................................................... 151-200 55.5-125.4
Very Unhealthy..................................................... 201-300 125.5-225.4
[[Page 16306]]
Hazardous \1\...................................................... 301+ 225.5
----------------------------------------------------------------------------------------------------------------
\1\ AQI values between breakpoints are calculated using equation 1 in appendix G. For AQI values in the
hazardous category, AQI values greater than 500 should be calculated using equation 1 and the PM2.5
concentration specified for the AQI value of 500.
2. Summary of Significant Comments on Proposed Revisions
The EPA received many comments on the proposed changes to the
PM2.5 AQI breakpoints. Many commenters generally supported
all the proposed revisions to the AQI breakpoints based on the
revisions to the primary annual and daily PM2.5 standards
and recent scientific evidence discussed in the proposal (88 FR 5558,
January 27, 2023). However, we received specific comments on proposed
revisions to the breakpoints in the lower end of the AQI, related to
their linkage to the annual and daily PM2.5 standards, and
proposed revisions to the breakpoints at the upper end of the AQI,
based on EPA's interpretation of available health effects evidence.
a. Air Quality Index Values of 50, 100, and 150
Some commenters agreed with using the historical approach of
setting the 50, 100 and 150 breakpoints of the AQI to be consistent
with the primary PM2.5 standards. Some cited the reason that
this approach creates consistent communication with respect to air
quality and the standards, and this is how the other AQI sub-indices
are set. A few commenters disagreed with the historical approach and
suggested instead that the 50 breakpoint of the AQI should not be
revised at all, or that the 50 and 100 breakpoints of the AQI should be
supported directly by health data similar to the basis for the proposed
200, 300 and 500 breakpoints.
The few commenters that disagreed with the historical approach of
the 50 breakpoint of the AQI noted that setting a short-term breakpoint
to annual standard was not logical since it is a long-term standard and
not meant to be interpreted for short-term messaging with the AQI, in
particular when reported hourly via the NowCast. These commenters also
noted that additional studies are needed to identify the health impacts
of short-term exposures at low concentrations. They also noted that
lowering the 50 breakpoint of the AQI in conjunction with the annual
standard may cause confusion with the public because some State
programs and policy decisions are connected to the AQI while others are
based on PM concentrations, which could lead to inconsistent messaging
reducing the public's trust. These comments were supported by noting
that revised breakpoints could lead to more moderate days than in the
past, but the monitor values would be the same as before when the
commenters considered it ``healthy,'' possibly eroding trust in air
agencies' messaging. Commenters also noted if the breakpoints are
revised, the public will not visually be able to detect the difference
between what was considered a good AQI day versus a now moderate AQI
day.
The EPA disagrees with these commenters. With respect to setting a
short-term breakpoint to the level of a much longer-term (annual)
standard, setting the lower AQI breakpoints at the level of the annual
and daily PM2.5 standards for communication purposes was
discussed in the proposed reconsideration (88 FR 5558, January 27,
2023) and previously supported by State organizations in the 2012 PM
Final Rule (77 FR 38890, June 29, 2012). Both the AQI and the Pollutant
Standards Index, which came before it, have historically been
normalized across pollutants by defining an index value of 50 and 100
as the numerical level of the annual (when defined) and short-term
(i.e., averaging time of 24-hours or less) primary NAAQS for each
pollutant. This approach clearly communicates the air quality to the
public. The EPA considers this approach to be appropriate given the
available evidence and structure of the standard. As discussed in
section II.B above and in the notice of final rulemaking for the 2012
review (77 FR 38890, June 29, 2012), the primary annual and 24-hour
PM2.5 standards work together in concert to provide public
health protection. The annual PM2.5 standard is generally
viewed as the principal means of providing public health protection
against ``typical'' daily and annual PM2.5 exposures, while
the 24-hour PM2.5 standard is generally viewed as a means of
providing protection against short-term exposures to ``peak''
PM2.5 concentrations, such as can occur in areas with strong
contributions from local or seasonal sources, even when annual average
PM2.5 concentrations remain relatively low. Because the
annual standard provides public health protection for typical daily
PM2.5 exposures, the EPA thinks it is appropriate to use
that level for the 50 breakpoint of the AQI and describe daily air
quality at and below the level of the annual standard ``Good.'' Since
an annual standard allows for days with air quality above that level,
it is appropriate to call days just above it ``Moderate.'' If the 50
breakpoint of the AQI was set at a level above the annual standard, it
would be possible for the majority of days to be called ``good'' in a
year when an area exceeds the annual standard. This could cause
confusion with the public about air quality if the general perception
is that local air quality is ``good,'' but the area fails to meet the
annual standard. In addition, the EPA continues to find it appropriate
to use the NowCast with the PM2.5 AQI index to provide more
real-time information to the public. As discussed in the AQI Technical
Assistance Document, while the NowCast algorithm is approximating a 24-
hour average exposure, it can reflect concentrations observed over
shorter averaging times when air quality is changing rapidly (U.S. EPA,
2018a). The EPA continues to consider the use of the primary annual
standard level suitable in the NowCast given the health evidence
supporting the standard and given that the reported concentrations are
an approximation of ``typical'' daily exposure. Additionally, the EPA
reflects the nature of the NowCast in the associated health messaging.
With regard to the commenter stating the public may not be able to
visually detect a difference in the air quality, the EPA notes that the
AQI is intended to be a communication tool for public awareness
precisely because it is generally difficult for the public to visually
judge air quality risks when air pollution is ``moderate.'' Moreover,
since the establishment of the AQI, the EPA and State and local air
agencies and organizations have developed experience in educating the
public about changes in the standards and,
[[Page 16307]]
concurrently, related changes to AQI breakpoints and advisories. When
the standards change, the EPA and State and local agencies have sought
to help the public understand that air quality is not getting worse,
it's that the health evidence underlying the standards and the AQI has
changed. The EPA's Air Quality System (AQS), the primary repository for
air quality monitoring data, is also adjusted to reflect the revised
breakpoints. Specifically, all historical AQI values in AQS are
recomputed with the revised breakpoints, so that all data queries and
reports downstream of AQS will show appropriate trends in AQI values
over time. If any State, local or Tribal air agency is concerned that
people are or will be confused on a moderate AQI day, then they could
use the communication information that has been developed with this
rulemaking.
Some commenters stated that the AQI should not necessarily be
linked to the primary PM2.5 standards. One example is the
comment that if the annual standard is not lowered to 8 [micro]g/m\3\,
the EPA should lower the 50 breakpoint of the AQI to that level to
better inform the public of the need for behavioral modifications to
reduce the harm to health from PM2.5 exposure. Similar to
the reasons discussed above, the EPA concludes that setting the 50
breakpoint of the AQI at the level of the annual PM2.5
standard is appropriate from a health perspective and for communication
purposes. The Administrator has judged the primary annual standard (in
conjunction with the other primary standards) as revised in this final
action to be requisite to protect public health with an adequate margin
of safety, based on the health evidence discussed in section II.A.2.
Setting the 50 breakpoint lower than the annual standard also has the
potential to cause confusion with the public since it does not reflect
the standards and the Administrator's judgments about the standards as
well.
With regard to the 100 breakpoint of the AQI, several commenters
expressed the view that the level of the 24-hour PM2.5
standard and an AQI value of 100 should be set at 25 [mu]g/m\3\ based
on the body of evidence and lower end of the range recommended by
CASAC. These commenters noted that if the current 24-hour standard and
AQI value of 100 is retained at 35 [mu]g/m\3\ then the public will not
be able to make informed decisions about actions to take to protect
their health. Many of these commenters further recommended that the AQI
value of 100 should be lowered to 25 [mu]g/m\3\ even if the standard is
retained. Commenters expressed the view that this would more adequately
allow the public to take health-protective actions.
The EPA disagrees with these commenters and notes that many State,
Tribal and local air agencies have expressed strong support for
aligning the 100 breakpoint of the AQI with the short-term 24-hour
primary PM2.5 standards as discussed in the proposal (88 FR
5558, January 27, 2023). The EPA agrees with the view, expressed by
State, local and Tribal entities, that aligning the lower breakpoints
with the standards enables clear communication of the standards. This
alignment approach is also utilized in the other AQI sub-indices lower
breakpoints and taking a different approach with the PM2.5
AQI could cause confusion. Additionally, the Administrator has judged
that it is appropriate to retain the 24-hour standard at a level of 35
[mu]g/m\3\ (in conjunction with the other primary standards) to protect
public health with an adequate margin of safety, based on the health
evidence discussed in section II.A.2. Thus, EPA disagrees that it is
necessary or appropriate to set the 100 breakpoint at a lower
concentration to provide further information to the public. The 50
breakpoint, which is set at a level below 25 [mu]g/m\3\, will continue
to provide information to members of the public particularly concerned
about exposures to PM2.5. As with the 50 breakpoint,
aligning the breakpoint with the standard both reflects the
Administrator's judgment about the health risks and eliminates the
potential to cause confusion in the public about those risks.
b. Air Quality Index Values of 200 and Above
Some commenters supported the proposed revisions to the 200, 300
and 500 breakpoints that recognize the expanded body of scientific
evidence, particularly several new epidemiologic studies conducted
during high pollution events such as wildfires and multiple controlled
human exposure studies. A few commenters agreed with incorporating the
expanded body of scientific evidence into the 200, 300 and 500
breakpoints, but suggested a modified linear approach between 200 (115
[mu]g/m\3\) and 500 (312 [mu]g/m\3\, setting the 300 breakpoint to 187
[mu]g/m\3\) based on recent epidemiologic wildfire smoke studies.
Other commenters disagreed with the proposed revisions and
suggested the EPA should continue using the previous breakpoints that
follow the 1999 linear approach (64 FR 42530, August 4, 1999), because
not changing the breakpoints would simplify communications. A few
commenters stated the proposed revisions to the AQI upper breakpoints
are not justified because the scientific evidence supporting the
revisions is inadequate. To support this view, the commenters suggest
that only three epidemiologic studies were used in determining the
upper breakpoints and none of them were representative of potential
effects in the general public; of the 13 studies cited only three were
near the proposed revised breakpoints; four of the studies involved
exposure to PM from diesel and traffic pollution, which is different
than PM from wildfire smoke; and the data supporting the revisions only
indicated ``mild'' health effects that were mostly in sensitive
populations.
The EPA agrees with the majority of commenters that supported
utilizing the expanded body of scientific evidence to revise the 200,
300 and 500 breakpoints of the AQI. The EPA appreciates the suggestion
of using a revised linear approach from 200 to 500. But rather than
using the available evidence to only set the breakpoint of 500, the EPA
finds it appropriate to set the breakpoints for 200, 300 and 500 using
an evidence-based approach, by relying on information presented in both
controlled human exposure studies and epidemiologic studies that
examine relationships between high PM2.5 exposure episodes
(i.e., periods of wildfire smoke) and various health outcomes. Setting
these breakpoints based directly on health effects evidence, which can
be communicated, is more useful and appropriate than using a linear
approach, because it can better describe the potential health effects
and symptoms which also helps the public better understand why more
health protective actions are needed. By its nature, a linear approach
does not evaluate and identify associated health effects and risk
factors.
The EPA disagrees with the commenters that expressed the view that
these upper breakpoints should not be revised based largely on the
numerous peer-reviewed studies published since the 200, 300 and 500
breakpoints were originally established in 1999 (64 FR 42530, August 4,
1999). As discussed in the proposal (88 FR 5641, January 27, 2023), the
rationale behind the proposed revisions is rooted in the fact the upper
AQI breakpoints are based on outdated scientific evidence.
Specifically, the traditional linear approach was predicated on the 500
value of the AQI, which was estimated using health studies that used
the British Smoke Method. The British Smoke Method is based on a
particle size fraction (4.5 microns) that is larger
[[Page 16308]]
than PM2.5. Given that the British Smoke method has a larger
particle size cutpoint than the current PM2.5 monitoring
method, which has a cutpoint of 2.5 microns, a concentration of 500
[mu]g/m\3\ based on the British Smoke method would be equivalent to a
lower PM2.5 concentration (88 FR 5641, January 27, 2023).
The combination of a larger particle size fraction informing previous
decisions around upper AQI breakpoints and more recent scientific
evidence than the London Fog Episode, on the potential health
consequences of what we currently consider to be high PM2.5
exposures, provides the underlying basis for revising the upper
breakpoints to better inform the public about air quality to allow the
public to take health protective actions as appropriate. Moreover, as
discussed above, until recently there was limited information upon
which to base the breakpoints between 150 and 500, so the linear
approach was a reasonable substitute. While not changing the
breakpoints may be easier because there is no change to communicate,
using a health-based approach is more appropriate, because it helps the
public better understand that more health protective actions are
needed.
The Agency disagrees that the scientific evidence discussed in the
proposal is inadequate to revise the 200, 300 and 500 breakpoints of
the AQI (88 FR 5640, January 27, 2023). The EPA disagrees that these
studies should not be considered because they ``indicated mild health
effects in sensitive populations.'' The EPA notes that many of the
subclinical effects discussed in the proposal (88 FR 5640, January 27,
2023) that informed the breakpoints are on the biologically plausible
pathway (see 2019 ISA, section 6.1.1 and Figure 6-1) to more severe
cardiovascular outcomes, such as ED visits, hospital admissions, and
death as depicted in the large number of epidemiologic studies
evaluated in the 2019 ISA and ISA Supplement. From a public health
perspective, the purpose of the AQI is to inform the public when air
quality could adversely affect their health. The scientific evidence
informed revisions to the breakpoints at the upper end of the AQI allow
it to better reflect the risk of experiencing health effects at higher
PM2.5 concentrations. In addition, the EPA disagrees with
the commenter that the effects reported at these higher concentrations
were observed only in sensitive populations as these effects were also
reported in healthy populations (Ghio et al., 2000; Ghio et al., 2003;
Urch et al., 2010; Ramanathan et al., 2016; Sivagangabalan et al.,
2011; Brook et al., 2009; Bellavia et al. (2013); Tong et al. (2015);
Behbod et al. (2013); Vieira et al. (2016a) Vieira et al. (2016b); and
Lucking et al. (2011)).
c. Other Comments
The EPA received a few additional comments on elements of the
PM2.5 AQI, including the averaging time. Some commenters
expressed the view that the 24-hour averaging time was not useful when
informing the public how to protect their health, particularly during
rapidly changing conditions such as wildfire smoke events. Instead,
they suggested a subdaily averaging time of 1-3 hours would be more
effective because it more closely aligns with how people breathe.
A few of these commenters suggested that instead of changing the
AQI averaging time, which aligns with the short-term standard, the EPA
could create a public health warning system for unhealthy
PM2.5 levels. The commenters noted that aligning the AQI
averaging time with the short-term standard could be useful for
consistent communication with the standards and attainment but
suggested that a subdaily warning system could better allow the public
to take health protective actions.
The EPA disagrees that a shorter averaging period for the
PM2.5 AQI sub-index would be better. The health effects
evidence supporting a subdaily metric is limited and inconsistent. As
part of its review of the health effects evidence, the 2019 ISA
evaluated whether a subdaily metric would be more closely related to
health effects. Most epidemiologic studies that examined the
relationship between short-term PM2.5 exposures and health
effects evaluated an exposure metric averaged over 24-hours. Some
recent studies, focusing on respiratory and cardiovascular effects and
mortality, have examined whether there is evidence that subdaily
exposure metrics are more closely related to health effects than a
traditional 24-hour average metric. After evaluating this limited newer
evidence, the 2019 ISA concluded that ``collectively, the available
evidence does not indicate that subdaily averaging periods for
PM2.5 are more closely associated with health effects than
the 24-hour avg exposure metric,'' (2019 ISA, chapter 1, section
1.5.2.1, pp. 146-147; U.S. EPA, 2022a).
In addition, there are communication benefits to aligning the
averaging time of the AQI with the daily standard, as some of these
commenters note, such as providing consistent messages about when it
may be beneficial for people to take actions to reduce PM2.5
exposures. Furthermore, with regard to an additional warning system,
the EPA is concerned that having two air quality communication systems
operating at the same time would likely be confusing to the public and
reduce the effectiveness of the systems.
At the same time, the EPA recognizes that when air quality is
rapidly changing, such as during wildfire smoke events, reporting
information based on a 24-hour metric may not be as useful for the
public as reporting more frequently would be. The EPA has balanced
concerns about being able to provide timely communication of air
quality hazards when conditions are changing quickly with the goal of
limiting the number of air quality communications systems and its
judgment that the evidence supports a 24-hour-based metric linked to
the daily standard by establishing the NowCast, which takes into
consideration subdaily PM2.5 concentrations and provides a
near real-time AQI value based on the AQI colors and scale.
Specifically, the NowCast shows air quality conditions for the most
current hour of PM2.5 data available by using a calculation
that involves multiple hours of past data. As noted in the AQI
Technical Assistance Document, the NowCast currently uses longer
averages during periods of stable air quality and shorter averages
(down to a 3-hour average) when air quality is changing rapidly, such
as during a wildfire (U.S. EPA, 2018a). As discussed further in section
IV.D.2 of this notice, the EPA uses the NowCast to approximate the
complete daily AQI (24-hour average) during any given hour. This means
the subdaily NowCast is approximating a 24-hour average exposure, which
aligns with the health evidence and the existing AQI communications
network, while also being capable of communicating rapidly changing
conditions to the public.
3. Summary of Final Revisions
Upon reviewing and considering the comments on the proposed
revisions (summarized above in Section IV.C) along with the scientific
evidence outlined in the proposal (88 FR 5639, January 27, 2023) and
summarized above in section IV.A, the EPA is finalizing the proposed
changes to the AQI.
Thus, as discussed in section IV of the preamble (88 FR 5639,
January 27, 2023) to the proposed rule, the EPA is taking final action
to revise the AQI value of 50 to 9.0 [mu]g/m\3\, 24-hour average,
consistent with the final decision on the primary annual
PM2.5 standard level as summarized in section II.C of the
[[Page 16309]]
preamble to the final rule; retain the AQI value of 100 at 35 [mu]g/
m\3\, 24-hour average, consistent with the final decision on the
primary 24-hour PM2.5 standard level as summarized in
section II.C of the preamble to the final rule; and retain the AQI
value of 150 at 55 [mu]g/m\3\, 24-hour average. The EPA is also taking
action to revise the AQI value of 200 to 125 [mu]g/m\3\, 24-hour
average; 300 to 225 [mu]g/m\3\, 24-hour average; and 500 to 325 [mu]g/
m\3\, 24-hour average, consistent with the rationale discussed above
and the health evidence discussed in section IV of the preamble (88 FR
5639, January 27, 2023) to the proposed rule. The EPA has prepared
communications materials to assist States with adjusting to the revised
AQI and looks forward to working with, and learning from the
experiences of, State, local, and Tribal governments in implementing
these changes.
C. Air Quality Index Category Breakpoints for PM10
The EPA proposed to retain the PM10 sub-index of the AQI
consistent with the proposed decision to retain the primary
PM10 standard, and consistent with the health effects
information that supports this proposed decision, as discussed in
section III.D of the proposal (88 FR 5632, January 27, 2023). EPA did
not receive comments on this and is taking final action to retain the
PM10 sub-index of the AQI for the reasons stated in the
preamble to the proposed rule (88 FR 5642, January 27, 2023).
D. Air Quality Index Reporting
With respect to the reporting requirements for the AQI and as noted
in the proposal (88 FR 5642, January 27, 2023) there have been many
technological advances in air quality monitoring and data reporting
since the appendix G to 40 CFR part 58 was last revised in 1999.
Federal, State, local, and Tribal agencies have used these changes to
make health information and air quality data more readily available and
easier to access. Given this, it is useful to update the reporting
requirements and recommendations to match current practices and ensure
the public has the most useful and timely information to take health-
protective behaviors.
1. Summary of Proposed Revisions
Currently, appendix G defines daily reporting as five days per
week. When this reporting requirement was originated in 1999 the
technology available at that time was not sufficient to calculate and
report the AQI more than five days per week without requiring
additional staffing on the weekends. Since that time, advances in
technology have allowed for reporting seven days per week automatically
without expending additional resources on weekends. As a result, most
State, local, and Tribal air agencies now report the AQI seven days per
a week. Given these technological advances and noting that reporting
agencies currently report the AQI seven days per week, the EPA proposed
that State, local, and Tribal agencies that report the AQI be required
to report it seven days a week, ensuring that the members of the public
continue to have access to daily air quality and health information
that they can use to take steps to protect their health.
Improvements in monitoring networks and modeling capabilities have
also enabled the ability to report the AQI in near real-time. This
allows State, local, and Tribal air agencies to provide timely air
quality information to the public for making health-protective
decisions and to help satisfy AQI reporting requirements. The
availability of near real-time AQI data also allows for more timely
responses by the public when air quality conditions are changing
rapidly, such as during wildfire smoke events. Subdaily reporting of
the AQI can be critical when there are rapidly change conditions and/or
high pollution events so that the public is able to make informed
decisions to protect their health. Many State, local, and Tribal air
agencies currently report the AQI hourly to ensure that the public has
access to accurate and timely information. In recognition of these
advances, and to continue to provide for near-real time AQI reporting
that the public has come to rely on, the EPA proposed to recommend that
State, local, and Tribal agencies report the AQI in near-real time.
In lieu of or along with reporting the near-real-time AQI directly
to the public, most State/local and Tribal agencies submit hourly air
quality data to the EPA. The EPA and some State, local and Tribal air
quality agencies use this near-real-time data to create products for
use by the public, weather service providers and the media as discussed
in the proposal (88 FR 5643, January 27, 2023). To continue to ensure
the availability of the products that the public and many stakeholders
rely upon, the EPA proposed to recommend that State, local, and Tribal
air quality agencies submit hourly data to the EPA's air quality
database. Submitting hourly data to the EPA for use on the AirNow
website and in other products also enables State, local, and Tribal air
quality agencies to meet the recommendation to report the AQI in near-
real-time.
In addition to the proposed updates to the reporting requirements
and recommendations for near-real-time reporting and data submission
recommendations, the Agency also proposed reformatting the question-
and-answer format used in appendix G to align with the current standard
formatting used in the Code of Federal Regulations. In proposing to
update the format, the EPA did not reopen the language that has merely
been moved or rearranged as there are no substantive changes.
Another change the EPA proposed to make to appendix G is with
regard to Table 2--Breakpoints for the AQI for purposes of clarity. As
discussed in the proposal (88 FR 5642, January 27, 2023) and summarized
here, the EPA proposed to collapse the two rows presented for the
Hazardous Category into one. The two rows in the current table specify
pollutant concentrations for two AQI ranges within the Hazardous
category (301-400 and 401-500), with an intermediate break at 400. The
400 breakpoint for all criteria pollutants in the current Table 2 is
set at the proportional pollutant concentration approximately halfway
between the Index values of 300 and 500. In proposing updated AQI
breakpoints for PM2.5, the EPA considered adjusting the 400
breakpoint similarly. However, the EPA concluded that collapsing the
two rows into a single range (301-500) would provide a more transparent
and easy-to-follow presentation of the pollutant concentrations
corresponding to the AQI range for the Hazardous category. Moreover,
collapsing the Hazardous category into a single row in Table 2 has no
substantive effect on the Emergency Episode program in 40 CFR part 51,
appendix L. Thus, the EPA proposed to remove the breakpoint of 400 from
the table in appendix G but this change would not substantively affect
the derivation of the AQI for any pollutant.
In addition, the EPA proposed to move some information currently in
appendix G into the Technical Assistance Document for the Reporting of
Daily Air Quality, or TAD (U.S. EPA, 2018a), so that it can be updated
in a more timely manner to reflect current scientific and health
effects evidence and current communication methods, thereby assisting
State, local, and Tribal agencies in providing accurate and timely
information to the public. Information that was proposed to be moved
from appendix G to the TAD included the definitions of the sensitive
(at-risk) populations for each pollutant.
[[Page 16310]]
This definition is typically evaluated and updated, as warranted, in
most NAAQS reviews, even if the standard is not revised. Generally, if
the standard is not revised in a review of the NAAQS, then appendix G
is also not revised. Moving the definitions of sensitive groups to the
TAD allows them to be updated even when a NAAQS is not revised to be
consistent with the definitions of the sensitive (at-risk) populations
identified in the ISA for that NAAQS review. Also, the proposal (88 FR
5642, January 27, 2023) recognized that the ways that air quality and
health information is supplied to the news media and public changes
regularly and thus proposed that information about suggested approaches
for public communication be taken out of appendix G and discussed in
the TAD.
2. Summary of Significant Comments on the Proposed Revisions
The EPA received many comments on the proposed changes to AQI
reporting, many of which supported the proposed revisions. EPA
discusses several of the topics that received the most attention from
commenters below. Discussion of other comments received on the proposed
changes to the AQI can be found in section IV of the Responses to
Significant Comments on the 2023 Proposed Reconsideration of the
National Ambient Air Quality Standards for Particulate Matter.
Most commenters expressed support for revising the definition of
``daily reporting'' from five days a week to seven days a week. A
commenter did not support this change and recommended the EPA maintain
the definition of daily as five days per week, noting that State and
local air agencies do not routinely work seven days per week and would
not be available to perform quality control of this data and report it
reliably on weekends.
The EPA appreciates the support for this proposed revision and
disagrees that the proposed change would require personnel to perform
quality control of AQI data on weekends. 40 CFR part 58 Appendix D
defines continuous monitoring requirements for agencies participating
in the State/Local Air Monitoring Stations (SLAMS) network, and
Appendix G states that agencies `` . . . must use concentration data
from State/Local Air Monitoring Stations (SLAMS) required by 40 CFR
58.10'' when reporting the AQI. Therefore, as noted in Appendix D and
G, Agencies are required to report the AQI using monitors within SLAMS,
which are not subject to daily quality control/validation.
A few commenters noted that the proposal preamble language
mentioned AQI is reported three ways (88 FR 5637, 5638, January 27,
2023): ``The AQI is reported three ways all of which are useful and
complementary. The daily AQI is reported for the previous day and used
to observe trends in community air quality, the AQI forecast helps
people plan their outdoor activities for the next day, and the near-
real-time AQI, or NowCast AQI, tells people whether it is a good time
for outdoor activity.'' These commenters suggested that the NowCast is
being codified in 40 CFR part 58 Appendix G as a method of calculating
the AQI, which they oppose, saying that codifying its use is
inappropriate given the shortest averaging period of the
PM2.5 NAAQS remains at 24-hours. Some stated that NowCast
values have no direct correlation to the AQI calculation methodology
codified in 40 CFR part 58 Appendix G. These commenters say that
codifying the NowCast would impose a significant burden on States'
forecasting staff.
However, some other commenters noted they appreciate the public-
friendly format and near real-time data the NowCast provides and use it
in their clinical encounters with patients. One air agency recognized
the importance of the NowCast near real-time AQI during high pollution
events and suggested the EPA should provide more ``concrete'' health
messaging for these short-term spikes.
The EPA disagrees that the preamble language proposed to codify the
NowCast or to impose a burden on reporting agencies. The preamble to
the proposed rule references the AQI being reported in three ways and
it does so because the EPA and many State, local and Tribal air quality
agencies already report it these three ways. However, text included in
the preamble is generally explanatory and does not alter regulatory
provisions. Comments that State that EPA is codifying the NowCast into
Appendix G are incorrect. Further, in proposed revisions to 40 CFR part
58 Appendix G, the EPA recommended, but did not propose to require, the
use of air quality forecasts and a subdaily AQI. Consistent with the
proposal, the EPA is therefore not finalizing any additional
requirement or burden on States' forecasting staff relative to
forecasts or a subdaily AQI.
The EPA disagrees with the comment that the NowCast values have no
direct correlation to the AQI calculation methodology codified in 40
CFR part 58 Appendix G. As noted in the AQI Technical Assistance
Document (Technical Assistance Document for the Reporting of Daily Air
Quality--the Air Quality Index (AQI)), the NowCast algorithm is based
on the AQI methodology but provides more real-time information to the
public (U.S. EPA, 2018a). While the NowCast algorithm is approximating
a 24-hour average exposure, it can reflect concentrations observed over
shorter averaging times when air quality is changing rapidly (U.S. EPA,
2018a). The EPA reflects the nature of the NowCast in the health
messaging provided there.
As noted in the above discussion of the AQI, air quality can change
quickly during the day. A central purpose of the AQI is to help the
public know when it is prudent to take action to reduce their exposure
to pollution. Accordingly, the EPA developed the NowCast to estimate
the 24-hour AQI for the current hour to give people information and
tools to reduce their exposures to protect their health, particularly
when air quality may be changing. The NowCast gives people the
knowledge and ability to take timely action. They can use this
information to reduce their exposure--reducing exposures if
PM2.5 is high only during a few hours a day will help reduce
a person's 24-hour exposure--or be active when air quality is better.
The first NowCast method was developed in 2003 and was designed so
``current conditions'' represent the 24-hour PM2.5 standard
as closely as possible. This method proved to be slow to respond during
rapid air quality changes. In 2013, the EPA developed an updated
NowCast method for PM2.5 \141\ that responds more quickly to
rapidly changing air quality conditions, such as those we see during
wildfires, to make air quality alerts more timely. We analyzed millions
of data points in developing this NowCast method and presented this
information to State, local and Tribal air agencies. The updated
NowCast, which is still in use, was launched August 1, 2013, on
AirNow.gov. It was designed to represent a shorter average (target 3-
hour) when air quality is changing rapidly, in part because 3-hour
averages from some continuous monitors are more stable than 1-hour
averages. The NowCast reflects a longer-term (12-hour) average when air
quality is stable.
---------------------------------------------------------------------------
\141\ U.S. EPA. (2013). Transitioning to a New NowCast Method.
Presentation available in the Rulemaking Docket for the Review of
the National Ambient Air Quality Standards for Particulate Matter
(EPA-HQ-OAR-2015-0072), at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
After evaluating the 2013 NowCast method, the EPA concluded that it
matched the desired characteristics. The NowCast method responds to
rapid changes in air quality yet still reflects a
[[Page 16311]]
longer-term average when air quality is stable; will work in any
location with adequate air quality data and for any air quality
situation; gives people the best possible estimate of a 24-hour
exposure; allows the EPA to caution people in time for them to take
protective action and reduce their 24-hour exposure; and ensures that
AQI maps on AirNow more closely match what people see.
The AQI is designed to allow people to reduce their exposure when
pollution levels are higher and be active outdoors when pollution
levels are lower. Since air quality almost always changes during the
day, that level of granularity is not possible with a 24-hour forecast.
If the public has only the 24-hour forecast, they may miss the times to
be active outdoors when air quality is better and may be active
outdoors when air quality is worse.
Also as noted above, many entities appreciate the near real-time
reporting of the AQI that the NowCast provides and suggested more
specific messaging is needed. The EPA appreciates this insight and will
continue to consider ways to communicate air quality information most
effectively to the public. For example, in light of recent wildfire
events, the EPA worked with the USFS to pilot the AirNow Fire and Smoke
Map.
3. Summary of Final Revisions
Upon reviewing and considering the comments on the proposed
revisions (summarized above in Section IV.C) along with the rationale
outlined in the proposal (88 FR 5638, January 27, 2023) and summarized
above in section IV.C, the EPA is finalizing the proposed changes to
the AQI reporting requirements. Thus, as discussed in section IV of the
preamble to the proposed rule, the EPA is taking final action to
require the AQI be reported seven days a week; recommend that State,
local, and Tribal agencies report the AQI in near-real time; recommend
that State, local, and Tribal air quality agencies submit hourly data
to the EPA's air quality database; reformat appendix G to align with
the current standard formatting used in the Code of Federal
Regulations; collapse the two rows in Table 2 presented for the
Hazardous Category into one by removing the 400 breakpoint; and move
some information currently in appendix G into the Technical Assistance
Document for the Reporting of Daily Air Quality, or TAD (U.S. EPA,
2018a) such as including the definitions of the sensitive (at-risk)
populations for each pollutant and suggested approaches for public
communication as stated in the revised Appendix G.
Table 2 below summarizes the breakpoints for the PM2.5
sub-index.
Table 2--Breakpoints for PM2.5 Sub-Index
----------------------------------------------------------------------------------------------------------------
Breakpoints ([mu]g/m\3\,
AQI category Index values 24-hour average)
----------------------------------------------------------------------------------------------------------------
Good............................................................... 0-50 0.0-9.0
Moderate........................................................... 51-100 9.1-35.4
Unhealthy for Sensitive Groups..................................... 101-150 35.5-55.4
Unhealthy.......................................................... 151-200 55.5-125.4
Very Unhealthy..................................................... 201-300 125.5-225.4
Hazardous \1\...................................................... 301+ 225.5
----------------------------------------------------------------------------------------------------------------
\1\ AQI values between breakpoints are calculated using equation 1 in appendix G. For AQI values in the
hazardous category, AQI values greater than 500 should be calculated using equation 1 and the PM2.5
concentration specified for the AQI value of 500.
V. Rationale for Decisions on the Secondary PM Standards
This section presents the rationale for the Administrator's
decision that no change to the current secondary PM standards is
required at this time to provide requisite protection against the
public welfare effects of PM within the scope of this reconsideration
(i.e., visibility, climate, and materials effects).\142\ This decision
is based on a thorough review of the scientific evidence generally
published through December 2017,\143\ as presented in the 2019 ISA
(U.S. EPA, 2019a), on the non-ecological public welfare effects of PM
pertaining to the presence of PM in ambient air, specifically
visibility, climate, and materials effects. Additionally, this decision
is based on a thorough evaluation of some studies that became available
after the literature cutoff date of the 2019 ISA that could either
further inform the adequacy of the current PM NAAQS or address key
scientific topics that have evolved since the literature cutoff date
for the 2019 ISA, generally through March 2021, as presented in the ISA
Supplement \144\ (U.S. EPA, 2022a). The selection of welfare effects
evaluated within the ISA Supplement was based on the causality
determinations reported in the 2019 ISA and the subsequent use of
scientific evidence in the 2020 PA.\145\
[[Page 16312]]
Specifically, for welfare effects, the focus within the ISA Supplement
is on visibility effects. The ISA Supplement does not include an
evaluation of studies on climate or materials effects. The
Administrator's decision also takes into account the 2022 PA evaluation
of the policy-relevant information in the 2019 ISA and ISA Supplement
and presentation of quantitative analysis of air quality related to
visibility impairment; CASAC advice and recommendations, as reflected
in discussions of the drafts of the ISA Supplement and 2022 PA at
public meetings and in the CASAC's letters to the Administrator; and
public comments received on the proposal.
---------------------------------------------------------------------------
\142\ Consistent with the 2016 Integrated Review Plan (U.S. EPA,
2016), other welfare effects of PM, including ecological effects,
are being considered in the separate, on-going review of the
secondary NAAQS for oxides of nitrogen, oxides of sulfur and PM.
Accordingly, the public welfare protection provided by the secondary
PM standards against ecological effects such as those related to
deposition of nitrogen- and sulfur-containing compounds in
vulnerable ecosystems is being considered in that separate review.
Thus, the Administrator's decision in this reconsideration will be
focused only and specifically on the adequacy of public welfare
protection provided by the secondary PM standards from effects
related to visibility, climate, and materials and hereafter
``welfare effects'' refers to non-ecological welfare effects (i.e.,
visibility, climate, and materials effects).
\143\ In addition to the 2020 review's opening ``call for
information'' (79 FR 71764, December 3, 2014), the 2019 ISA
identified and evaluated studies and reports that have undergone
scientific peer review and were published or accepted for
publication between January 1, 2009 through approximately January
2018 (U.S. EPA, 2019a, p. ES-2). References that are cited in the
2019 ISA, the references that were considered for inclusion but not
cited, and electronic links to bibliographic information and
abstracts can be found at: https://hero.epa.gov/hero/particulate-matter.
\144\ As described in more detail in the ISA Supplement, ``the
scope of this Supplement provides specific criteria for the types of
studies considered for inclusion within the Supplement.
Specifically, studies must be peer reviewed and published between
approximately January 2018 and March 2021'' (U.S. EPA, 2022a,
section 1.2.2).
\145\ As described in section 1.2.1 of the ISA Supplement, ``the
selection of welfare effects to evaluate within this Supplement is
based on the causality determinations reported in the 2019 PM ISA
and the subsequent use of scientific evidence in the 2020 PM PA. The
2019 PM ISA concluded a causal relationship for each of the welfare
effects categories evaluated (i.e., visibility, climate effects, and
materials effects). While the 2020 PM PA considered the broader set
of evidence for these effects, for climate effects and material
effects, it concluded that there remained `substantial uncertainties
with regard to the quantitative relationships with PM concentrations
and concentration patterns that limit[ed] [the] ability to
quantitatively assess the public welfare protection provided by the
standards from these effects (U.S. EPA, 2020b). Given these
uncertainties and limitations, the basis of the discussion on
conclusions regarding the secondary standards in the 2020 PM PA
primarily focused on visibility effects. Therefore, this Supplement
focuses only on visibility effects in evaluating newly available
scientific information and is limited to studies conducted in the
U.S. and Canada'' (U.S. EPA, 2022a, section 1.2.1).
---------------------------------------------------------------------------
In presenting the rationale for the Administrator's final decision
and its foundations, section V.A provides background on the 2020 final
decision to retain the secondary PM standards (section V.A.1), and also
provides brief summaries of key aspects of the currently available
welfare effects evidence (section V.A.2) and quantitative information
(section V.A.3). Section V.B summarizes the CASAC's advice (section
V.B.1) and the proposed conclusions (section V.B.2), addresses public
comments received on the proposal (section V.B.3), and presents the
Administrator's conclusions on the adequacy of the current standards
(section V.B.4), drawing on consideration of the available scientific
and quantitative information, advice from the CASAC, and comments from
the public. Section V.C summarizes the Administrator's decision on the
secondary PM standards.
A. Introduction
The general approach for this reconsideration of the 2020 final
decision on the secondary PM standards relies on the EPA's assessments
of the current scientific evidence and associated quantitative analyses
to inform the Administrator's judgments regarding secondary standards
that are requisite to protect the public welfare from known or
anticipated adverse effects associated with the pollutant's presence in
the ambient air. The EPA's assessments are primarily documented in the
2019 ISA, ISA Supplement, and 2022 PA, which builds on the 2020 PA, all
of which have received CASAC review and public comment (83 FR 53471,
October 23, 2018; 83 FR 55529, November 6, 2018; 85 FR 4655, January
27, 2020; 86 FR 52673, September 22, 2021; 86 FR 54186, September 30,
2021; 86 FR 56263, October 8, 2021; 87 FR 958, January 7, 2022; 87 FR
22207, April 14, 2022; 87 FR 31965, May 26, 2022). In bridging the gap
between the scientific assessments of the 2019 ISA and ISA Supplement
and the judgments required of the Administrator in determining whether
the current standards provide the requisite public welfare protection,
the 2022 PA evaluates policy implications of the evaluation of the
current evidence in the 2019 ISA and ISA Supplement, and the
quantitative information documented in the 2022 PA. In evaluating the
public welfare protection afforded by the current standards against PM-
related effects within the scope of this reconsideration, the four
basic elements of the NAAQS (indicator, averaging time, level, and
form) are considered collectively.
The final decision on the adequacy of the current secondary
standards is a public welfare policy judgment to be made by the
Administrator. In reaching conclusions with regard to the standard, the
decision draws on the scientific information and analyses about welfare
effects, and associated public welfare significance, as well as
judgments about how to consider the range and magnitude of
uncertainties that are inherent in the scientific evidence and
analyses. This approach is based on the recognition that the available
evidence generally reflects a continuum that includes ambient air
exposures at which scientists agree that effects are likely to occur
through lower levels at which the likelihood and magnitude of responses
become increasingly uncertain. This approach is consistent with the
requirements of the provisions of the Clean Air Act related to the
review of NAAQS and with how the EPA and the courts have historically
interpreted the Act. These provisions require the Administrator to
establish secondary standards that, in the judgment of the
Administrator, are requisite to protect public welfare from known or
anticipated adverse effects associated with the presence of the
pollutant in the ambient air. In so doing, the Administrator seeks to
establish standards that are neither more nor less stringent than
necessary for this purpose. The Act does not require that standards be
set at a zero-risk level, but rather at a level that reduces risk
sufficiently so as to protect the public welfare from known or
anticipated adverse effects.
1. Background on the Current Standards
The current secondary PM standards were retained in 2020 based on
the scientific and technical information available at that time, as
well as the then-Administrator's judgments regarding the available
welfare effects evidence, the appropriate degree of public welfare
protection for the existing standards, and available air quality
information on visibility impairment that may be allowed by such a
standard (85 FR 82684, December 18, 2020). With the 2020 decision, the
then-Administrator retained the secondary 24-hour PM2.5
standard, with its level of 35 [micro]g/m\3\, the annual
PM2.5 standard, with its level of 15.0 [micro]g/m\3\, and
the 24-hour PM10 standard, with its level of 150 [micro]g/
m\3\. The subsections below focus on the key considerations and the
then-Administrator's conclusions in the 2020 final decision for climate
and materials effects (section V.A.1.a) and visibility effects (section
V.A.2.b).
a. Non-Visibility Effects
In light of the robust evidence base, the 2019 ISA concluded there
to be causal relationships between PM and climate effects and materials
effects (U.S. EPA, 2019a, sections 13.3.9 and 13.4.2). The 2020 final
decision was based on a thorough review in the 2019 ISA of the
scientific information on PM-induced climate and materials effects. The
decision also took into account: (1) Assessments in the 2020 PA of the
most policy-relevant information in the 2019 ISA regarding evidence of
adverse effects of PM to climate and materials, (2) uncertainties in
the available evidence to inform a quantitative assessment of PM-
related climate and materials effects, (3) CASAC advice and
recommendations, and (4) public comments received during the
development of these documents and on the proposal document.
In considering non-visibility welfare effects in the 2020 decision,
the then-Administrator concluded that, while it is important to
maintain an appropriate degree of control of fine and coarse particles
to address non-visibility welfare effects, ``it is generally
appropriate to retain the existing standards and that there is
insufficient information to establish any distinct secondary PM
standards to address climate and materials effects of PM'' (85 FR
82744, December 18, 2020).
With regard to climate, the then-Administrator recognized that
there were a number of improvements and refinements to climate models
since the 2012 review. However, while the evidence continued to support
a causal relationship between PM and climate effects, the then-
Administrator noted that significant limitations continued to exist
related to quantifying the contributions of direct and indirect
[[Page 16313]]
effects of PM and PM components on climate forcing (U.S. EPA, 2020b,
sections 5.2.2.1.1 and 5.4). He also recognized that the models
continued to exhibit considerable variability in estimates of PM-
related climate impacts at regional scales (e.g., ~100 km) as compared
to simulations at global scales. Therefore, the resulting uncertainty
led the then-Administrator to conclude in the 2020 decision that the
available scientific information remained insufficient to quantify
climate impacts associated with particular concentrations of PM in
ambient air (U.S. EPA, 2020b, section 5.2.2.2.1) or to evaluate or
consider a level of PM air quality in the U.S. to protect against
climate effects and that there was insufficient information available
to base a national ambient standard on climate impacts (85 FR 82744,
December 18, 2020).
With regard to materials effects, the then-Administrator noted that
the evidence available in the 2019 ISA continued to support a causal
relationship between materials effects and PM deposition (U.S. EPA,
2019a, section 13.4). He recognized that the deposition of fine and
coarse particles to materials can lead to physical damage and/or
impaired aesthetic qualities. Particles can contribute to materials
damage by adding to the natural weathering processes and by promoting
the corrosion of metals, the degradation of building materials, and the
weakening of material components. While some new information was
available in the 2019 ISA, the information was from studies primarily
conducted outside of the U.S. in areas where PM concentrations in
ambient air are higher than those observed in the U.S. (U.S. EPA,
2020b, section 13.4). Additionally, the information assessed in the
2019 ISA did not support quantitative analyses of PM-related materials
effects in the 2020 PA (U.S. EPA, 2020b, section 5.2.2.2.2). Given the
limited amount of information available and its inherent uncertainties
and limitations, the Administrator concluded that he was unable to
relate soiling or damage to specific levels of PM in ambient air or to
evaluate or consider a level of air quality to protect against such
materials effects, and that there was insufficient information
available to support a distinct national ambient standard based on
materials effects (85 FR 82744, December 18, 2020).
In reviewing the 2019 draft PA, the CASAC agreed with staff
conclusions that, while these effects are important, ``the available
evidence does not call into question the protection afforded by the
current secondary PM standards'' and recommended that the secondary
standards ``should be retained'' (Cox, 2019b, p. 3 of letter). In
reaching a final decision in 2020, for all of the reasons discussed
above and recognizing the CASAC conclusion that the evidence provided
support for retaining the current secondary PM standards, the then-
Administrator concluded that it was appropriate to retain the existing
secondary PM standards, without revision. For climate and materials
effects, this conclusion reflected his judgment that, although it
remains important to maintain secondary PM2.5 and
PM10 standards to provide some degree of control over long-
and short-term concentrations of both fine and coarse particles, there
was insufficient information to establish distinct secondary PM
standards to address non-visibility PM-related welfare effects (85 FR
82744, December 18, 2020).
b. Visibility Effects
The 2019 ISA concluded that, ``the evidence is sufficient to
conclude that a causal relationship exists between PM and visibility
impairment'' (U.S. EPA, 2019a, section 13.2.6). The 2020 decision on
the adequacy of the secondary standards with regard to visibility
effects was a public welfare policy judgment made by the then-
Administrator, which drew upon the available scientific evidence for
PM-related visibility effects and on analyses of visibility impairment,
as well as judgments about the appropriate weight to place on the range
of uncertainties inherent in the evidence and analyses. The 2020 final
decision was based on a thorough review in the 2019 ISA of the
scientific information on PM-related visibility effects. The decision
also took into account: (1) Assessments in the 2020 PA of the most
policy-relevant information in the 2019 ISA regarding evidence of
adverse effects of PM on visibility; (2) air quality analyses of the
PM2.5 visibility index and design values based on the form
and averaging time of the existing secondary 24-hour PM2.5
standard; (3) CASAC advice and recommendations; and (4) public comments
received during the development of these documents and on the 2020
proposal document.
In considering the visibility effects in the 2020 review, the then-
Administrator noted the long-standing body of evidence for PM-related
visibility impairment. This evidence, which is based on the fundamental
relationship between light extinction and PM mass, demonstrated that
ambient PM can impair visibility in both urban and remote areas, and
had changed very little since the 2012 review (U.S. EPA, 2019a, section
13.1; U.S. EPA, 2009a, section 9.2.5). The evidence related to public
perception of visibility impairment was from studies from four areas in
North America.\146\ These studies provided information to inform our
understanding of levels of visibility impairment that the public judged
to be ``acceptable'' (U.S. EPA, 2010b; 85 FR 24131, April 30, 2020). In
considering these public preference studies, the then-Administrator
noted that no new visibility studies conducted in the U.S. were
discussed in the 2019 ISA, and there was little newly available
information with regard to acceptable levels of visibility impairment
in the U.S. The Administrator recognized that visibility impairment can
have implications for people's enjoyment of daily activities and their
overall well-being, and therefore, considered the degree to which the
current secondary standards protect against PM-related visibility
impairment.
---------------------------------------------------------------------------
\146\ Preference studies were available in four urban areas.
Three western preference studies were available, including one in
Denver, Colorado (Ely et al., 1991), one in the lower Fraser River
valley near Vancouver, British Columbia, Canada (Pryor, 1996), and
one in Phoenix, Arizona (BBC Research & Consulting, 2003). A pilot
focus group study was also conducted for Washington, DC (Abt
Associates, 2001), and a replicate study with 26 participants was
also conducted for Washington, DC (Smith and Howell, 2009). More
details about these studies are available in Appendix D of the 2022
PA (U.S. EPA, 2022b).
---------------------------------------------------------------------------
Consistent with the 2012 review, in the 2020 review, the then-
Administrator first concluded that a target level of protection for a
secondary PM standard is most appropriately defined in terms of a
visibility index that directly takes into account the factors (i.e.,
species composition and relative humidity) that influence the
relationship between PM2.5 in ambient air and PM-related
visibility impairment. In defining a target level of protection, the
then-Administrator considered the specific aspects of such an index,
including the appropriate indicator, averaging time, form and level (78
FR 82742-82744, December 18, 2020).
First, with regard to indicator, the then-Administrator noted that
in the 2012 review, the EPA used an index based on estimates of light
extinction by PM2.5 components calculated using an adjusted
version of the IMPROVE algorithm, which allows the estimation of the
light extinction using routinely monitored components of
PM2.5, PM10-2.5 mass, and estimates of relative
humidity. The then-Administrator recognized that, while there have been
some revisions to the IMPROVE algorithm since the time of the 2012
[[Page 16314]]
review, our fundamental understanding of the relationship between PM in
ambient air and light extinction had changed little and the various
IMPROVE algorithms appropriately reflected this relationship across the
U.S. In the absence of a monitoring network for direct measurement of
light extinction, he concluded that a calculated light extinction
indicator that utilizes the IMPROVE algorithms continued to provide a
reasonable basis for defining a target level of protection against PM-
related visibility impairment (78 FR 82742-82744, December 18, 2020).
In further defining the characteristics of a visibility index, the
then-Administrator next considered the appropriate averaging time,
form, and level of the index. Given the available scientific
information the review, and in considering the CASAC's advice and
public comments, the then-Administrator concluded that, consistent with
the decision in the 2012 review, a visibility index with a 24-hour
averaging time and a form based on the 3-year average of annual 90th
percentile values remained reasonable. With regard to the averaging
time and form of such an index, the Administrator noted analyses
conducted in the last review that demonstrated relatively strong
correlations between 24-hour and subdaily (i.e., 4-hour average)
PM2.5 light extinction (78 FR 3226, January 15, 2013),
indicating that a 24-hour averaging time is an appropriate surrogate
for the subdaily time periods of the perception of PM-related
visibility impairment and the relevant exposure periods for segments of
the viewing public. This decision in the 2020 review also recognized
that a 24-hour averaging time may be less influenced by atypical
conditions and/or atypical instrument performance (78 FR 3226, January
15, 2013). The then-Administrator recognized that there was no new
information to support updated analyses of this nature, and therefore,
he believed these analyses continued to provide support for
consideration of a 24-hour averaging time for a visibility index in
this review. With regard to the statistical form of the index, the
Administrator noted that, consistent with the 2012 review: (1) A multi-
year percentile form offers greater stability from the occasional
effect of interannual meteorological variability (78 FR 3198, January
15, 2013; U.S. EPA, 2011, p. 4-58); (2) a 90th percentile represents
the median of the distribution of the 20 percent worst visibility days,
which are targeted in Federal Class I areas by the Regional Haze
Program; and (3) public preference studies did not provide information
to identify a different target than that identified for Federal Class I
areas (U.S. EPA, 2011, p. 4-59). Therefore, the then-Administrator
judged that a visibility index based on estimates of light extinction,
with a 24-hour averaging time and a 90th percentile form, averaged over
three years, remained appropriate (78 FR 82742-82744, December 18,
2020).
With regard to the level of a visibility index, consistent with the
2012 review, the then-Administrator judged that it was appropriate to
establish a target level of protection of 30 deciviews
(dv),147 148 reflecting the upper end of the range of
visibility impairment judged to be acceptable by at least 50% of study
participants in the available public preference studies (78 FR 3226,
January 15, 2013). The 2011 PA identified a range of levels from 20 to
30 dv based on the responses in the public preference studies available
at that time (U.S. EPA, 2011, section 4.3.4). At the time of the 2012
review, the then-Administrator noted a number of uncertainties and
limitations in public preference studies, including the small number of
stated preference studies available, the relatively small number of
study participants, the extent to which the study participants may not
be representative of the broader study area population in some of the
studies, and the variations in the specific materials and methods used
in each study. In considering the available preference studies in 2012,
with their inherent uncertainties and limitations, the then-
Administrator concluded that the substantial degree of variability and
uncertainty in the public preference studies should be reflected in a
target level of protection based on the upper end of the range of
candidate protection levels (CPLs).
---------------------------------------------------------------------------
\147\ Deciview (dv) refers to a scale for characterizing
visibility that is defined directly in terms of light extinction.
The deciview scale is frequently used in the scientific and
regulatory literature on visibility.
\148\ For comparison, 20 dv, 25 dv, and 30 dv are equivalent to
64, 112, and 191 megameters (Mm-1), respectively.
---------------------------------------------------------------------------
Given that there were no new preference studies in the 2019 ISA,
the then-Administrator's judgments in 2020 were based on the same
studies, with the same range of levels, available in the 2012 review.
The 2020 PA (U.S. EPA, 2020b, section 5.5), discussed a number of
limitations and uncertainties associated with these studies. In
considering the scientific information, with its uncertainties and
limitations, as well as public comments on the level of the target
level of protection against visibility impairment, the then-
Administrator concluded that it was appropriate to again use a level of
30 dv for the visibility index (78 FR 82742-82744, December 18, 2020).
Having concluded that the protection provided by a standard defined
in terms of a PM2.5 visibility index, with a 24-hour
averaging time, and a 90th percentile form, averaged over 3 years, set
at a level of 30 dv, was requisite to protect public welfare with
regard to visual air quality, the Administrator next considered the
degree of protection from visibility impairment afforded by the
existing suite of secondary PM standards.
In this context, the then-Administrator considered the updated
analyses of visibility impairment presented in the 2020 PA (U.S. EPA,
2020b, section 5.2.1.2), which reflected a number of improvements since
the 2012 review. Specifically, the updated analyses examined multiple
versions of the IMPROVE equation, including the version incorporating
revisions since the time of the 2012 review. These updated analyses
provided a further understanding of how variation in the inputs to the
algorithms affect the estimates of light extinction (U.S. EPA, 2020b,
Appendix D). Additionally, for a subset of monitoring sites with
available PM10-2.5 data, the updated analyses better
characterized the influence of coarse PM on light extinction than in
the 2012 review (U.S. EPA, 2020b, section 5.2.1.2).
The results of the updated analyses in the 2020 PA were consistent
with those from the 2012 review. Regardless of which version of the
IMPROVE equation was used, the analyses demonstrated that, based on
2015-2017 data, the 3-year visibility metric was at or below about 30
dv in all areas meeting the current 24-hour PM2.5 standard,
and below 25 dv in most of those areas. In locations with available
PM10-2.5 monitoring, which met both the current 24-hour
secondary PM2.5 and PM10 standards, 3-year
visibility index metrics were at or below 30 dv regardless of whether
the coarse fraction was included as an input to the algorithm for
estimating light extinction (U.S. EPA, 2020b, section 5.2.1.2). While
the inclusion of the coarse fraction had a relatively modest impact on
the estimates of light extinction, the then-Administrator recognized
the continued importance of the PM10 standard given the
potential for larger impacts on light extinction in areas with higher
coarse particle concentrations, which were not included in the analyses
in the 2020 PA due to a lack of available data (U.S. EPA, 2019a,
section 13.2.4.1; U.S. EPA, 2020b, section 5.2.1.2). He
[[Page 16315]]
noted that the air quality analyses showed that all areas meeting the
existing 24-hour PM2.5 standard, with its level of 35
[micro]g/m\3\, had visual air quality at least as good as 30 dv, based
on the visibility index. Thus, the secondary 24-hour PM2.5
standard would likely be controlling relative to a 24-hour visibility
index set at a level of 30 dv. Additionally, areas would be unlikely to
exceed the target level of protection for visibility of 30 dv without
also exceeding the existing secondary 24-hour PM2.5
standard. Thus, the then-Administrator judged that the 24-hour
PM2.5 standard provided sufficient protection in all areas
against the effects of visibility impairment, i.e., that the existing
24-hour PM2.5 standard would provide at least the target
level of protection for visual air quality of 30 dv which he judged
appropriate (78 FR 82742-82744, December 18, 2020).
2. Overview of Welfare Effects Evidence
The information summarized here is based on the scientific
assessment of the welfare effects evidence available in this
reconsideration; this assessment is documented in the 2019 ISA and ISA
Supplement and its policy implications are further discussed in the
2022 PA. While the 2019 ISA provides the broad scientific foundation
for this reconsideration, additional literature has become available
since the cutoff date of the 2019 ISA that expands the body of evidence
related to visibility effects that can inform the Administrator's
judgment on the adequacy of the current secondary PM standards. As
such, the ISA Supplement builds on the information in the 2019 ISA with
a targeted identification and evaluation of new scientific information
regarding visibility effects. As described in the ISA Supplement and
the 2022 PA, the selection of welfare effects to evaluate within the
ISA Supplement were based on the causality determinations reported in
the 2019 ISA and the subsequent use of scientific evidence in the 2020
PA (U.S. EPA, 2019a, section 1.2; U.S. EPA, 2022a, section 1.4.2). The
ISA Supplement focuses on U.S. and Canadian studies that provide new
information on public preferences for visibility impairment and/or
developed new methodologies or conducted quantitative analyses of light
extinction (U.S. EPA, 2022a, section 1.2). Such studies of visibility
effects and quantitative relationships between visibility impairment
and PM in ambient air were considered to be of greatest utility in
informing the Administrator's conclusions on the adequacy of the
current secondary PM standards. The visibility effects evidence
presented within the 2019 ISA, along with the targeted identification
and evaluation of new scientific information in the ISA Supplement,
provides the scientific basis for the reconsideration of the 2020 final
decision on the secondary PM standards for visibility effects. For
climate and materials effects, the 2020 PA concluded that there were
substantial uncertainties associated with the quantitative
relationships with PM concentrations and the concentration patterns
that limited the ability to quantitatively assess the public welfare
protection provided by the standards from these effects. Therefore, the
evaluation of the information related to these effects draws heavily
from the 2019 ISA and 2020 PA. The subsections below briefly summarize
the nature of PM-related visibility (section V.B.1.a), climate (section
V.B.1.b), and materials (section V.B.1.c) effects.
a. Nature of Effects
Visibility impairment can have implications for people's enjoyment
of daily activities and for their overall sense of well-being (U.S.
EPA, 2009a, section 9.2). The strongest evidence for PM-related
visibility impairment comes from the fundamental relationship between
light extinction and PM mass (U.S. EPA, 2009a), which confirms a well-
established ``causal relationship exists between PM and visibility
impairment'' (U.S. EPA, 2009a, p. 2-28). Beyond its effects on
visibility, the 2009 ISA also identified a causal relationship
``between PM and climate effects, including both direct effects of
radiative forcing and indirect effects that involve cloud and feedbacks
that influence precipitation formation and cloud lifetimes'' (U.S. EPA,
2009a, p. 2-29). The evidence also supports a causal relationship
between PM and effects on materials, including soiling effects and
materials damage (U.S. EPA, 2009a, p. 2-31).
The evidence available in this reconsideration is consistent with
the evidence available at the time of the 2012 and 2020 reviews and
supports the conclusions of causal relationships between PM and
visibility, climate, and materials effects (U.S. EPA, 2019a, chapter
13). Evidence newly available in this reconsideration augments the
previously available evidence of the relationship between PM and
visibility impairment (U.S. EPA, 2019a, section 13.2; U.S. EPA, 2022a,
section 4), climate effects (U.S. EPA, 2019a, section 13.3), and
materials effects (U.S. EPA, 2019a, section 13.4).
i. Visibility
The fundamental relationship between light extinction and PM mass,
and the EPA's understanding of this relationship, has changed little
since the 2009 ISA (U.S. EPA, 2009a). The combined effect of light
scattering and absorption by particles and gases is characterized as
light extinction, i.e., the fraction of light that is scattered or
absorbed per unit of distance in the atmosphere.\149\ Light extinction
is measured in units of 1/distance, which is often expressed in the
technical literature as visibility per megameter (abbreviated
Mm-1). Higher values of light extinction (usually given in
units of Mm-1 or dv) correspond to lower visibility. When PM
is present in the air, its contribution to light extinction is
typically much greater than that of gases (U.S. EPA, 2019a, section
13.2.1). The impact of PM on light scattering depends on particle size
and composition, as well as relative humidity. All particles scatter
light, as described by the Mie theory, which relates light scattering
to particle size, shape, and index of refraction (U.S. EPA, 2019a,
section 13.2.3; Mie, 1908, Van de Hulst, 1981). Fine particles scatter
more light than coarse particles on a per unit mass basis and include
sulfates, nitrates, organics, light-absorbing carbon, and soil (Malm et
al., 1994). Hygroscopic particles like ammonium sulfate, ammonium
nitrate, and sea salt increase in size as relative humidity increases,
leading to increased light scattering (U.S. EPA, 2019a, section
13.2.3).
---------------------------------------------------------------------------
\149\ All particles scatter light and, although a larger
particle scatters more light than a similarly shaped smaller
particle of the same composition, the light scattered per unit of
mass is greatest for particles with diameters from ~0.3-1.0 [micro]m
(U.S. EPA, 2009a, section 2.5.1; U.S. EPA, 2019a, section 13.2.1).
Particles with hygroscopic components (e.g., particulate sulfate and
nitrate) contribute more to light extinction at higher relative
humidity than at lower relative humidity because they change size in
the atmosphere in response to relative humidity.
---------------------------------------------------------------------------
As at the time of the 2012 and 2020 reviews, direct measurements of
PM light extinction, scattering, and absorption continue to be
considered more accurate for quantifying visibility than PM mass-based
estimates because measurements do not depend on assumptions about
particle characteristics (e.g., size, shape, density, component
mixture, etc.) (U.S. EPA, 2019a, section 13.2.2.2). Measurements of
light extinction can be made with high time resolution, allowing for
characterization of subdaily temporal patterns of visibility
impairment. A number of measurement methods have been used for
visibility impairment (e.g.,
[[Page 16316]]
transmissometers, integrating nephelometers, teleradiometers,
telephotometers, and photography and photographic modeling), although
each of these methods has its own strengths and limitations (U.S. EPA,
2019a, Table 13-1). While some recent research confirms and adds to the
body of knowledge regarding direct measurements as is described in the
2019 ISA and ISA Supplement, no major new developments have been made
with these measurement methods since prior reviews (U.S. EPA, 2019a,
section 13.2.2.2; U.S. EPA, 2022a, section 4.2).
In the absence of a robust monitoring network for the routine
measurement of light extinction across the U.S., estimation of light
extinction based on existing PM monitoring can be used. The theoretical
relationship between light extinction and PM characteristics, as
derived from Mie theory (U.S. EPA, 2019a, Equation 13.5), can be used
to estimate light extinction by combining mass scattering efficiencies
of particles with particle concentrations (U.S. EPA, 2019a, section
13.2.3; U.S. EPA, 2009a, sections 9.2.2.2 and 9.2.3.1). This estimation
of light extinction is consistent with the method used in previous
reviews. The algorithm used to estimate light extinction, known as the
IMPROVE algorithm,\150\ provides for the estimation of light extinction
(bext), in units of Mm\-1\, using routinely monitored
components of fine (PM2.5) and coarse (PM10-2.5)
PM. Relative humidity data are also needed to estimate the contribution
by liquid water that is in solution with the hygroscopic components of
PM. To estimate each component's contribution to light extinction,
their concentrations are multiplied by extinction coefficients and are
additionally multiplied by a water growth factor that accounts for
their expansion with moisture. Both the extinction efficiency
coefficients and water growth factors of the IMPROVE algorithm have
been developed by a combination of empirical assessment and theoretical
calculation using particle size distributions associated with each of
the major aerosol components (U.S. EPA, 2019a, sections 13.2.3.1 and
13.2.3.3).
---------------------------------------------------------------------------
\150\ The algorithm is referred to as the IMPROVE algorithm as
it was developed specifically to use monitoring data generated at
IMPROVE network sites and with equipment specifically designed to
support the IMPROVE program and was evaluated using IMPROVE optical
measurements at the subset of monitoring sites that make those
measurements (Malm et al., 1994).
---------------------------------------------------------------------------
At the time of the 2012 review, two versions of the IMPROVE
algorithm were available in the literature--the original IMPROVE
algorithm (Lowenthal and Kumar, 2004, Malm and Hand, 2007, Ryan et al.,
2005) and the revised IMPROVE algorithm (Pitchford et al., 2007). As
described in detail in the 2022 PA (U.S. EPA, 2022b, section 5.3.1.1)
and the 2019 ISA (U.S. EPA, 2019a, section 13.2.3), the algorithm has
been further evaluated and refined since the time of the 2012 review
(Lowenthal and Kumar, 2016), particularly for PM characteristics and
relative humidity in remote areas. All three versions of the IMPROVE
algorithm were considered in evaluating visibility impairment in this
reconsideration.
Consistent with the evidence available at the time of the 2012 and
2020 reviews, our understanding of public perception of visibility
impairment comes from visibility preference studies conducted in four
areas in North America.\151\ The detailed methodology for these studies
are described in the 2022 PA (U.S. EPA, 2022b, section 5.3.1.1), the
2019 ISA (U.S. EPA, 2019a), and the 2009 ISA (U.S. EPA, 2019a). In
summary, the study participants were queried regarding multiple images
that were either photographs of the same location and scenery that had
been taken on different days on which measured extinction data were
available or digitized photographs onto which a uniform ``haze'' had
been superimposed. Results of the studies indicated a wide range of
judgments on what study participants considered to be acceptable
visibility across the different study areas, depending on the setting
depicted in each photograph. Based on the results of the four cities, a
range encompassing the PM2.5 visibility index values from
images that were judged to be acceptable by at least 50 percent of
study participants across all four of the urban preference studies was
identified (U.S. EPA, 2010b, p. 4-24; U.S. EPA, 2020b, Figure 5-2).
Much lower visibility (considerably more haze resulting in higher
values of light extinction) was considered acceptable in Washington,
DC, than was in Denver, and 30 dv reflected the level of impairment
that was determined to be ``acceptable'' by at least 50 percent of
study participants (78 FR 3226-3227, January 15, 2013).
---------------------------------------------------------------------------
\151\ Preference studies were available in four urban areas in
the last review: Denver, Colorado (Ely et al., 1991), Vancouver,
British Columbia, Canada (Pryor, 1996), Phoenix, Arizona (BBC
Research & Consulting, 2003), and Washington, DC (Abt Associates,
2001; Smith and Howell, 2009).
---------------------------------------------------------------------------
Since the completion of the 2009 and 2019 ISAs, there has been only
one public preference study that has become available in the U.S. This
study uses images of the Grand Canyon, AZ, described in the ISA
Supplement (U.S. EPA, 2022a). The Grand Canyon study, conducted by Malm
et al. (2019), has a similar study design to that used in the public
preference studies discussed above; however, there are several
important differences that make it difficult to directly compare the
results of the Malm et al. (2019) study with other public preference
studies. As an initial matter, the Grand Canyon study was conducted in
a Federal Class I area, as opposed to in an urban area, with a scene
depicted in the photographs that did not include urban features.\152\
We recognize that public preferences with respect to visibility in
Federal Class 1 areas may well differ from visibility preferences in
urban areas and other contexts, although there is currently a lack of
information to on such questions. Further, the Malm et al. (2019) study
also used a much lower range of superimposed ``haze'' than the
preference studies discussed above.\153\ It is unclear whether the
participant preferences are a function in part of the range of
potential values presented, such that the participant preferences for
the Grand Canyon were generally lower \154\ than the other preference
studies in part because of the lower range of superimposed ``haze'' for
the images in that study, or if their preferences would vary if
presented with images with a range of superimposed ``haze'' more
comparable to the levels used in the other studies (i.e., more ``haze''
superimposed on the images).
---------------------------------------------------------------------------
\152\ The Grand Canyon study used a single scene looking west
down the canyon with a small landscape feature of a 100-km-distant
mountain (Mount Trumbull), along with other closer landscape
features. The scenes presented in the previously available
visibility preference studies are presented in more detail in Table
D-9 in the 2022 PA (U.S. EPA, 2022b, Appendix D).
\153\ The Grand Canyon study superimposed light extinction
ranging from 3 dv to 20 dv on the image slides shown to participants
compared to the previously available preference studies. In those
studies, the visibility ranges presented were as low as 9 dv and as
high as 45 dv. The visibility ranges presented in the previously
available visibility preference studies are described in more detail
in Table D-9 in the 2022 PA (U.S. EPA, 2022b, Appendix D).
\154\ In the Grand Canyon study, the level of impairment that
was determined to be ``acceptable'' by at least 50 percent of study
participants was 7 dv (Malm et al., 2019).
---------------------------------------------------------------------------
The Malm et al. (2019) study also explored alternate methods for
evaluating ``acceptable'' levels of visual air quality from the
preference studies, including the use of scene-specific visibility
indices as potential indicators of visibility levels as perceived by
the observer (Malm et al., 2019). In addition to measures of
atmospheric haze, such
[[Page 16317]]
as atmospheric extinction, used in previously available preference
studies, other indices for visual air quality include color and
achromatic contrast of single landscape figures, average and equivalent
contrast of an entire scene, edge detection algorithms such as the
Sobel index, and just-noticeable difference or change indexes. The
results reported by Malm et al. (2019) suggest that scene-dependent
metrics, such as contrast, may be useful alternate predictors of
preference levels compared to universal metrics like light extinction
(U.S. EPA, 2022a, section 4.2.1). This is because extinction alone is
not a measure of ``haze,'' but of light attenuation per unit distance,
and visible ``haze'' is dependent on both light extinction and distance
to a landscape feature (U.S. EPA, 2022a, section 4.2.1). However, there
are very few studies available that use scene-dependent metrics (i.e.,
contrast) to evaluate public preference information, which makes it
difficult to evaluate them as an alternative to the light extinction
approach.
ii. Climate
The available evidence continues to support the conclusion of a
causal relationship between PM and climate effects (U.S. EPA, 2019a,
section 13.3.9). Since the 2012 review, climate impacts have been
extensively studied and recent research reinforces and strengthens the
evidence evaluated in the 2009 ISA. Recent evidence provides greater
specificity about the details of radiative forcing effects \155\ and
increases the understanding of additional climate impacts driven by PM
radiative effects. The Intergovernmental Panel on Climate Change (IPCC)
assesses the role of anthropogenic activity in past and future climate
change, and since the completion of the 2009 ISA, has issued the Fifth
IPCC Assessment Report (AR5; IPCC, 2013), which summarizes any key
scientific advances in understanding the climate effects of PM since
the previous report. As in the 2009 ISA, the 2019 ISA draws
substantially on the IPCC report to summarize climate effects. As
discussed in more detail in the 2022 PA (U.S. EPA, 2022b, section
5.3.2.1.1), the general conclusions are similar between the IPCC AR4
and AR5 reports with regard to effects of PM on global climate.
Consistent with the evidence available in the 2012 review, the key
components, including sulfate, nitrate, organic carbon (OC), black
carbon (BC), and dust, that contribute to climate processes vary in
their reflectivity, forcing efficiencies, and direction of forcing.
Since the completion of the 2009 ISA, the evidence base has expanded
with respect to the mechanisms of climate responses and feedbacks to PM
radiative forcing; however, the recently published literature assessed
in the 2019 ISA does not reduce the considerable uncertainties that
continue to exist related these mechanisms.
---------------------------------------------------------------------------
\155\ Radiative forcing (RF) for a given atmospheric constituent
is defined as the perturbation in net radiative flux, at the
tropopause (or the top of the atmosphere) caused by that
constituent, in watts per square meter (Wm\-2\), after allowing for
temperatures in the stratosphere to adjust to the perturbation but
holding all other climate responses constant, including surface and
tropospheric temperatures (Fiore et al., 2015; Myhre et al., 2013).
A positive forcing indicates net energy trapped in the Earth system
and suggests warming of the Earth's surface, whereas a negative
forcing indicates net loss of energy and suggests cooling (U.S. EPA,
2019a, section 13.3.2.2).
---------------------------------------------------------------------------
As described in the proposal (88 FR 5650, January 27, 2023), PM has
a very heterogeneous distribution globally and patterns of forcing tend
to correlate with PM loading, with the greatest forcings centralized
over continental regions. The climate response to this PM forcing,
however, is more complicated since the perturbation to one climate
variable (e.g., temperature, cloud cover, precipitation) can lead to a
cascade of effects on other variables. While the initial PM radiative
forcing may be concentrated regionally, the eventual climate response
can be much broader spatially or be concentrated in remote regions, and
may be quite complex, affecting multiple climate variables with
possible differences in the direction of the forcing in different
regions or for different variables (U.S. EPA, 2019a, section 13.3.6).
The complex climate system interactions lead to variation among climate
models, which have suggested a range of factors that can influence
large-scale meteorological processes and may affect temperature,
including local feedback effects involving soil moisture and cloud
cover, changes in the hygroscopicity of the PM, and interactions with
clouds (U.S. EPA, 2019a, section 13.3.7). As a result, there remains
insufficient evidence to related climate effects to specific PM levels
in ambient air or to establish a quantitative relationship between PM
and climate effects, particularly at a regional scale. Further research
is needed to better characterize the effects of PM on regional climate
in the U.S. before PM climate effects can be quantified.
iii. Materials
Consistent with the evidence assessed in the 2009 ISA, the
available evidence continues to support the conclusion that there is a
causal relationship between PM deposition and materials effects.
Effects of deposited PM, particularly sulfates and nitrates, to
materials include both physical damage and impaired aesthetic
qualities, generally involving soiling and/or corrosion (U.S. EPA,
2019a, section 13.4.2). Because of their electrolytic, hygroscopic, and
acidic properties and their ability to sorb corrosive gases, particles
contribute to materials damage by adding to the effects of natural
weathering processes, by potentially promoting or accelerating the
corrosion of metals, degradation of painted surfaces, deterioration of
building materials, and weakening of material components.\156\ There is
a limited amount of recently available data for consideration in this
review from studies primarily conducted outside of the U.S. on
buildings and other items of cultural heritage. However, these studies
involved concentrations of PM in ambient air greater than those
typically observed in the U.S. (U.S. EPA, 2019a, section 13.4).
---------------------------------------------------------------------------
\156\ As discussed in the 2019 ISA (U.S. EPA, 2019a, section
13.4.1), corrosion typically involves reactions of acidic PM (i.e.,
acidic sulfate or nitrate) with material surfaces, but gases like
SO2 and nitric acid (HNO3) also contribute.
Because ``the impacts of gaseous and particulate N and S wet
deposition cannot be clearly distinguished'' (U.S. EPA, 2019a, p.
13-1), the assessment of the evidence in the 2019 ISA considers the
combined impacts.
---------------------------------------------------------------------------
Building on the evidence available in the 2009 ISA, and as
described in detail in the proposal (88 FR 5650, January 27, 2023) and
in the 2019 ISA (U.S. EPA, 2019a, section 13.4), research has
progressed on (1) the theoretical understanding of soiling of items of
cultural heritage; (2) the quantification of degradation rates and
further characterization of factors that influence damage of stone
materials; (3) materials damage from PM components besides sulfate and
black carbon and atmospheric gases besides SO2; (4) methods
for evaluating soiling of materials by PM mixtures; (5) PM-attributable
damage to other materials, including glass and photovoltaic panels; (6)
development of dose-response relationships for soiling of building
materials; and (7) damage functions to quantify material decay as a
function of pollutant type and load. While the evidence of PM-related
materials effects has expanded somewhat since the completion of the
2009 ISA, there remains insufficient evidence to relate soiling or
damage to specific PM levels in ambient air or to establish a
quantitative relationship between PM and materials degradation. The
recent evidence assessed in the 2019 ISA is generally similar to the
evidence available in the 2009 ISA, including
[[Page 16318]]
associated limitations and uncertainties and a lack of evidence to
inform quantitative relationships between PM and materials effects,
therefore leading to similar conclusions about the PM-related effects
on materials.
3. Summary of Air Quality and Quantitative Information
Beyond the consideration of the scientific evidence, as discussed
in section V.A.2 above, quantitative analyses of PM air quality, when
available, can also inform conclusions on the adequacy of the public
welfare protection provided by the current secondary PM standards.
a. Visibility Effects
In the 2012 and 2020 reviews, quantitative analyses for PM-related
visibility effects focused on daily visibility impairment, given the
short-term nature of PM-related visibility effects. The evidence and
information available in this reconsideration continues to provide
support for the short-term (i.e., hourly or daily) nature of PM-related
visibility impairment. As such, the quantitative analyses presented in
the 2022 PA continue to focus on daily visibility impairment and
utilize a two-phase assessment approach for visibility impairment,
consistent with the approaches taken in past reviews. First, the 2022
PA considers the appropriateness of the elements (indicator, averaging
time, form, and level) of the visibility index for providing protection
against PM-related visibility effects. Second, recent air quality was
used to evaluate the relationship between the current secondary 24-hour
PM2.5 standard and the visibility index. The information
available since the 2012 review includes an updated equation for
estimating light extinction, summarized in the 2022 PA (U.S. EPA,
2022b, section 5.3.1.1) and described in the 2019 ISA (U.S. EPA, 2019a,
section 13.2.3.3), as well as more recent air monitoring data, that
together allow for development of an updated assessment of PM-related
visibility impairment in study locations in the U.S.
i. Target Level of Protection in Terms of a PM2.5
Visibility Index
In evaluating the adequacy of the current secondary PM standards,
the 2022 PA first evaluates the appropriateness of the elements
(indicator, averaging time, form, and level) identified for a
visibility index to protect against visibility effects. In previous
reviews, the visibility index as set at a level of 30 dv, with
estimated light extinction as the indicator, a 24-hour averaging time,
and a 90th percentile form, averaged over three years.
With regard to an indicator for the visibility index, the 2022 PA
recognizes the lack of availability of methods and an established
network for directly measuring light extinction (U.S. EPA, 2022b,
section 5.3.1.1). Therefore, consistent with previous reviews, the 2022
PA concludes that a visibility index based on estimates of light
extinction by PM2.5 components derived from an adjusted
version of the original IMPROVE algorithm to be the most appropriate
indicator for the visibility index in this reconsideration. As
described in section 5.3.1.1 of the 2022 PA, the IMPROVE algorithm
estimates light extinction using routinely monitored components of
PM2.5 and PM10-2.5, along with estimates of
relative humidity (U.S. EPA, 2022b, section 5.3.1.1).
With regard to averaging time, the 2022 PA notes that the evidence
continues to provide support for the short-term nature of PM-related
visibility effects. Given that there is no new information available
regarding the time periods during which visibility impairment occurs or
public preferences related to specific time periods for visibility
impairment, the 2022 PA concludes that it is appropriate to continue to
focus on daily visibility impairment. In so doing, the 2022 PA relies
on analyses that were conducted in the 2012 review that showed
relatively strong correlations between 24-hour and subdaily (i.e., 4-
hour average) PM2.5 light extinction that indicated that a
24-hour averaging time is an appropriate surrogate for the subdaily
time periods relevant for visual perception (U.S. EPA, 2011, Figures G-
4 and G-5; Frank, 2012). These analyses continue to provide support for
a 24-hour averaging time for the visibility index in this
reconsideration. Consistent with previous reviews, the 2022 PA also
notes that the 24-hour averaging time may be less influenced by
atypical conditions and/or atypical instrument performance than a
subdaily averaging time (85 FR 82740, December 18, 2020; 78 FR 3226,
January 15, 2013).
With regard to the form for the visibility index, the available
information continues to provide support for a 3-year average of annual
90th percentile values. Given that there is no new information to
inform selection of an alternate form, as in previous reviews, the 2022
PA notes that the 3-year average form provides stability from the
occasional effect of inter-annual meteorological variability that can
result in unusually high pollution levels for a particular year (85 FR
82741, December 18, 2020; 78 FR 3198, January 15, 2013; U.S. EPA, 2011,
p. 4-58). In so doing, the 2022 PA considers the evaluation in the 2010
Urban-Focused Visibility Assessment (UFVA) of three different
statistical forms: 90th, 95th, and 98th percentiles (U.S. EPA, 2010b,
Chapter 4).). In considering this evaluation of statistical forms from
the 2010 UFVA, consistent with the 2011 PA, the 2022 PA notes that the
Regional Haze Program targets the 20 percent most impaired days for
visibility improvements in visual air quality in Federal Class I areas
and that the median of the distribution of these 20 percent most
impaired days would be the 90th percentile. The 2011 PA also noted that
strategies that are implemented so that 90 percent of days would have
visual air quality that is at or below the level of the visibility
index would reasonably be expected to lead to improvements in visual
air quality for the 20 percent most impaired days. Additionally, as in
the 2011 PA, the 2022 PA recognizes that the available public
preference studies do not address frequency of occurrence of different
levels of visibility (U.S. EPA, 2022b, section 5.3.1.2). Therefore, the
analyses and consideration for the form of a visibility index from the
2011 PA continue to provide support for a 90th percentile form,
averaged across three years, in defining the characteristics of a
visibility index in this reconsideration.
With regard to the level for the visibility index, the 2022 PA
recognizes that there is an additional public preference study (Malm et
al., 2019) available in this reconsideration. As noted above, however,
this study differs from the previously available public preference
studies in several ways, which makes it difficult to integrate this
newly available study with the previously available studies. Most
significantly, this study was evaluated public preferences for
visibility in the Grand Canyon, perhaps the most notable Class I area
in the country for visibility purposes. Therefore, the 2022 PA
concludes that the Grand Canyon study is not directly comparable to the
other available preferences studies and public preferences of
visibility impairment in the Malm et al. (2019) study are not
appropriate to consider in identifying a range of levels for the target
level of protection against visibility impairment for this
reconsideration of the secondary PM NAAQS.
[[Page 16319]]
Therefore, the 2022 PA continues to rely on the same studies \157\
and the range of 20 to 30 dv identified from those studies in previous
reviews. With regard to selecting the appropriate target level of
protection for visibility impairment within this range, the 2022 PA
notes that in previous reviews, a level at the upper end of the range
(i.e., 30 dv) was selected given the uncertainties and limitations
associated with the public preference studies (U.S. EPA, 2022b, section
5.3.1.1). However, the 2022 PA also recognizes that (1) the degree of
protection provided by a secondary PM NAAQS is not determined solely by
any one element of the standard but by all elements (i.e., indicator,
averaging time, form, and level) being considered together, and (2)
decisions regarding the adequacy of the current secondary standards is
a public welfare policy judgment to be made by the Administrator. As
such, the Administrator may judge that a target level of protection
below the upper end of the range (i.e., less than 30 dv) is
appropriate, depending on his public welfare policy judgments, which
draw upon the available scientific evidence for PM-related visibility
effects and on analyses of visibility impairment, as well as judgments
about the appropriate weight to place on the range of uncertainties
inherent in the evidence and analyses.
---------------------------------------------------------------------------
\157\ As noted above, the available public preference studies
include those conducted in Denver, Colorado (Ely et al., 1991),
Vancouver, British Columbia, Canada (Pryor, 1996), Phoenix, Arizona
(BBC Research & Consulting, 2003), and Washington, DC (Abt
Associates, 2001; Smith and Howell, 2009).
---------------------------------------------------------------------------
In considering the available public preference studies, consistent
with past reviews, the 2022 PA concludes that it is reasonable to
consider a range of 20 to 30 dv for selecting a target level of
protection, including a high value of 30 dv, a midpoint value of 25 dv,
and a low value of 20 dv. A target level of protection at or in the
upper end of the range would focus on the Washington, DC, preference
study results (Abt Associates, 2001; Smith and Howell, 2009), which
identified 30 dv as the level of impairment that was determined to be
``acceptable'' by at least 50 percent of study participants. The public
preferences of visibility impairment in the Washington, DC, study are
likely to be generally representative of urban areas that do not have
valued scenic elements (e.g., mountains) in the distant background.
This would be more representative of areas in the middle of the country
and many areas in the eastern U.S., as well as possibly some areas in
the western U.S.
A target level of protection in the middle of the range would be
most closely associated with the level of impairment that was
determined to be ``acceptable'' by at least 50 percent of study
participants in the Phoenix, AZ, study (BBC Research & Consulting,
2003), which was 24 dv. This study, while methodologically similar to
the other public preference studies, included participants that were
selected as a representative sample of the Phoenix area population
\158\ and used computer-generated images to depict specific uniform
visibility impairment conditions. This study yielded the best results
of the four public preference studies in terms of the least noisy
preference results and the most representative selection of
participants. Therefore, based on this study, the use of 25 dv to
represent a midpoint within the range of target levels protection is
well supported.
---------------------------------------------------------------------------
\158\ The other preference studies did not include populations
that were necessarily representative of the population in the area
for which the images being judged. For example, in the Denver, CO,
study, participants were from intact groups (i.e., those who were
meeting for other reasons) and were asked to provide a period of
time during a regularly scheduled meeting to participate in the
study (Ely et al., 1991). As another example, in the British
Columbia, Canada, study, participants were recruited from
undergraduate and graduate students enrolled in classes at the
University of British Columbia's Department of Geography (Pryor,
1996).
---------------------------------------------------------------------------
A target level of protection at or just above the lower end of the
range would focus on the Denver, CO, study, but may not be as strongly
supported as higher levels within the range (Ely et al., 1991). Older
studies, such as those conducted in Denver, CO (Ely et al., 1991), and
British Columbia, Canada (Pryor, 1996), used photographs that were
taken at different times of the day and on different days to capture a
range of light extinction levels needed for the preference studies.
Compared to studies that used computer-generated images (i.e., those in
Phoenix, AZ, and Washington, DC) there was more variability in scene
appearance in these older studies that could affect preference rating
and includes uncertainties associated with using ambient measurements
to represent sight path-averaged light extinction values rather than
superimposing a computer-generated amount of haze onto the images. When
using photographs, the intrinsic appearance of the scene can change due
to meteorological conditions (i.e., shadow patterns and cloud
conditions) and spatial variations in ambient air quality that can
result in ambient light extinction measurement not being representative
of the sight-path-averaged light extinction. Computer-generated images,
such as those generated with WinHaze, do not introduce such
uncertainties, as the same base photograph is used (i.e., there is no
intrinsic change in scene appearance) and the modeled haze that is
superimposed on the photograph is determined based on uniform light
extinction throughout the scene.
In addition to differences in preferences that may arise from
photographs versus computer-generated images, urban visibility
preference may differ by location, and such differences may arise from
differences in the cityscape scene that is depicted in the images.
These differences are related to the perceived value of objects and
scenes that are included in the image, as objects at a greater distance
have a greater sensitivity to perceived visibility changes as light
extinction is changed compared to similar scenes with objects at
shorter distances. For example, a person (regardless of their location)
evaluating visibility in an image with more scenic elements such as
mountains or natural views may value better visibility conditions in
these images compared to the same level of visibility impairment in an
image that only depicts urban features such as buildings and roads.
That is, if a person was shown the same level of visibility impairment
in two images depicting different scenes--one with mountains in the
background and urban features in the foreground and one with no
mountains in the background and nearby buildings in the image without
mountains in the distance--may find the amount of haze to be
unacceptable in the image with the mountains in the distance because of
a greater perceived value of viewing the mountains, while finding the
amount of haze to be acceptable in the image with the buildings because
of a lesser value of viewing the cityscape or an expectation that such
urban areas may generally have higher levels of haze in general. This
is consistent when comparing the differences between the Denver, CO,
study results (which found the 50% acceptance criteria occurred at the
best visual air quality levels among the four cities) and the
Washington, DC, results (which found the 50% acceptability criteria
occurred at the worst visual air quality levels among the four cities).
These results may occur because the most prominent and picturesque
feature of the cityscape of Denver is the visible snow-covered
mountains in the distance, while the prominent and
[[Page 16320]]
picturesque features of the Washington, DC, cityscape are buildings
relatively nearby without prominent and/or valued scenic features that
are more distant. Given these variabilities in preferences it is
unclear to what extent, the available evidence provides strong support
for a target level of protection at the lower end of the range. Future
studies that reduce sources of noisiness and uncertainty in the results
could provide more information that would support selection of a target
level of protection at or just above the lower end of the range.
Taken together, the 2022 PA concludes that available information
continues to support a visibility index with estimated light extinction
as the indicator, a 24-hour averaging time, and a 90th percentile form,
averaged over three years, with a level within the range of 20 to 30
dv.
ii. Relationship Between the PM2.5 Visibility Index and the
Current Secondary 24-Hour PM2.5 Standard
The 2022 PA presents quantitative analyses based on recent air
quality that evaluate the relationship between recent air quality and
calculated light extinction. As in previous reviews, these analyses
explored this relationship as an estimate of visibility impairment in
terms of the 24-hour PM2.5 standard and the visibility
index. Generally, the results of the updated analyses are similar to
those based on the data available at the time of the 2012 and 2020
reviews (U.S. EPA, 2022b, section 5.3.1.2). As discussed in section
V.C.1.a above, the 2022 PA concludes that the available evidence
continues to support a visibility index with estimated light extinction
as the indicator, a 24-hour averaging time, and a 90th percentile form,
averaged over three years, with a level within the range of 20 to 30
dv. These analyses evaluate visibility impairment in the U.S. under
recent air quality conditions, particularly those conditions that meet
the current standards, and the relative influence of various factors on
light extinction. Given the relationship of visibility with short-term
PM, we focus particularly on the short-term PM standards.\159\ Compared
to the 2012 review, updated analyses incorporate several refinements,
including (1) the evaluation of three versions of the IMPROVE equation
to calculate light extinction (U.S. EPA, 2022b, Appendix D, Equations
D-1 through D-3) in order to better understand the influence of
variability in equation inputs; \160\ (2) the use of 24-hour relative
humidity data, rather than monthly average relative humidity as was
used in the 2012 review (U.S. EPA, 2022b, section 5.3.1.2, Appendix D);
and (3) the inclusion of the coarse fraction in the estimation of light
extinction (U.S. EPA, 2022b, section 5.3.1.2, Appendix D). The analyses
in the reconsideration are updated from the 2012 and 2020 reviews and
include 60 monitoring sites that measure PM2.5 and
PM10 and are geographically distributed across the U.S. in
both urban and rural areas (U.S. EPA, 2022b, Appendix D, Figure D-1).
---------------------------------------------------------------------------
\159\ The analyses presented in the 2022 PA focus on the
visibility index and the current secondary 24-hour PM2.5
standard with a level of 35 [micro]g/m\3\. However, we recognize
that all three secondary PM standards influence the PM
concentrations associated with the air quality distribution. As
noted in section V.A.1 above, the current secondary PM standards
include the 24-hour PM2.5 standard, with its level of 35
[micro]g/m\3\, the annual PM2.5 standard, with its level
of 15.0 [micro]g/m\3\, and the 24-hour PM10 standard,
with its level of 150 [micro]g/m\3\. With regard to the annual
PM2.5 standard, we note that all 60 areas included in the
analyses meet the current secondary annual PM standard (U.S. EPA,
2022b, Table D-7).
\160\ While the PM2.5 monitoring network has an
increasing number of continuous FEM monitors reporting hourly
PM2.5 mass concentrations, there continue to be data
quality uncertainties associated with providing hourly
PM2.5 mass and component measurements that could be input
into IMPROVE equation calculations for subdaily visibility
impairment estimates. As detailed in the 2022 PA, there are
uncertainties associated with the precision and bias of 24-hour
PM2.5 measurements (U.S. EPA, 2022b, p. 2-18), as well as
to the fractional uncertainty associated with 24-hour PM component
measurements (U.S. EPA, 2022b, p. 2-21). Given the uncertainties
present when evaluating data quality on a 24-hour basis, the
uncertainty associated with subdaily measurements may be even
greater. Therefore, the inputs to these light extinction
calculations are based on 24-hour average measurements of
PM2.5 mass and components, rather than subdaily
information.
---------------------------------------------------------------------------
When light extinction was calculated using the revised IMPROVE
equation, in areas that meet the current 24-hour PM2.5
standard for the 2017-2019 time period, all sites have light extinction
estimates at or below 26 dv (U.S. EPA, 2022b, Figure 5-3). For the four
locations that exceed the current 24-hour PM2.5 standard,
light extinction estimates range from 22 dv to 27 dv (U.S. EPA, 2022b,
Figure 5-3). These findings are consistent with the findings of the
analyses using the same IMPROVE equation in the 2012 review with data
from 102 sites with data from 2008-2010 and in the 2020 review with
data from 67 sites with data from 2015-2017. The analyses presented in
the 2022 PA indicate similar findings to those from the analyses in the
2012 and 2020 reviews, i.e., the updated quantitative analysis shows
that the 3-year visibility metric was no higher than 30 dv \161\ at
sites meeting the current secondary PM standards, and at most such
sites the 3-year visibility index values are much lower (e.g., an
average of 20 dv across the 60 sites).\162\
---------------------------------------------------------------------------
\161\ A 3-year visibility metric with a level of 30 dv would be
at the upper end of the range of levels identified from the public
preference studies.
\162\ When light extinction is calculated using the original
IMPROVE equation, all 60 sites have 3-year visibility metrics below
30 dv, 58 sites are at or below 25 dv, and 26 sites are at or below
20 dv (see U.S. EPA, 2022b, Appendix D, Table D-3).
---------------------------------------------------------------------------
When light extinction was calculated using the revised IMPROVE
equation,\163\ the resulting 3-year visibility metrics are nearly
identical to light extinction estimates calculated using the original
IMPROVE equation (U.S. EPA, 2022b, Figure 5-4), but some sites are just
slightly higher. Using the revised IMPROVE equation, for those sites
that meet the current 24-hour PM2.5 standard, the 3-year
visibility metric is at or below 26 dv. For the four locations that
exceed the current 24-hour PM2.5 standard, light extinction
estimates range from 22 dv to 29 dv (U.S. EPA, 2022b, Figure 5-4).
These results are similar to those for light extinction calculated
using the original IMPROVE equation,\164\ and those from previous
reviews.
---------------------------------------------------------------------------
\163\ As described in more detail in the 2022 PA, the revised
IMPROVE equation divides PM components into smaller and larger sizes
of particles in PM2.5, with separate mass scattering
efficiencies and hygroscopic growth functions for each size category
(U.S. EPA, 2022b, section 5.3.1.1).
\164\ When light extinction is calculated using the revised
IMPROVE equation, all 60 sites have 3-year visibility metrics below
30 dv, 56 sites are at or below 25 dv, and 26 sites are at or below
20 dv (see U.S. EPA, 2022b, Appendix D, Table D-3).
---------------------------------------------------------------------------
When light extinction was calculated using the refined equation
from Lowenthal and Kumar (2016), the resulting 3-year visibility
metrics are slightly higher at all sites compared to light extinction
estimates calculated using the original IMPROVE equation (U.S. EPA,
2022b, Figure 5-5).\165\ These higher estimates are to be expected,
given the higher OC multiplier included in the IMPROVE equation from
Lowenthal and Kumar (2016), which reflects the use of data from remote
areas with higher concentrations of organic PM when validating the
equation. As such, it is important to note that the Lowenthal and Kumar
(2016) version of the equation may overestimate light extinction in
non-remote areas, including the urban areas in the updated analyses in
this reconsideration.
---------------------------------------------------------------------------
\165\ When light extinction is calculated using the Lowenthal
and Kumar IMPROVE equation, 59 sites have 3-year visibility metrics
below 30 dv, 45 sites are at or below 25 dv, and 15 sites are at or
below 20 dv. The one site with a 3-year visibility metric of 32 dv
exceeds the secondary 24-hour PM2.5 standard, with a
design value of 56 [micro]g/m\3\ (see U.S. EPA, 2022b, Appendix D,
Table D-3).
---------------------------------------------------------------------------
Nevertheless, when light extinction is calculated using the
Lowenthal and
[[Page 16321]]
Kumar (2016) equation for those sites that meet the current 24-hour
PM2.5 standard, the 3-year visibility metric is generally at
or below 28 dv. For those sites that exceed the current 24-hour
PM2.5 standard, three of these sites have a 3-year
visibility metric ranging between 26 dv and 30 dv, while one site in
Fresno, California that exceeds the current 24-hour PM2.5
standard and has a 3-year visibility index value of 32 dv (compared to
29 dv when light extinction is calculated with the original IMPROVE
equation) (see U.S. EPA, 2022b, Appendix D, Table D-3). At this site,
it is likely that the 3-year visibility metric using the Lowenthal and
Kumar (2016) equation would be below 30 dv if PM2.5
concentrations were reduced such that the 24-hour PM2.5
level of 35 [micro]g/m\3\ was attained.
In considering visibility impairment under recent air quality
conditions, the 2022 PA recognizes that the differences in the inputs
to equations estimating light extinction can influence the resulting
values. For example, given the varying chemical composition of
emissions from different sources, the 2.1 multiplier for converting OC
to organic matter (OM) in the Lowenthal and Kumar (2016) equation may
not be appropriate for all source types. At the time of the 2012
review, the EPA judged that a 1.6 multiplier was more appropriate, for
the purposes of estimating visibility index at sites across the U.S.,
than the 1.4 or 1.8 multipliers used in the original and revised
IMPROVE equations, respectively. A multiplier of 1.8 or 2.1 would
account for the more aged and oxygenated organic PM that tends to be
found in more remote regions than in urban regions, whereas a
multiplier of 1.4 may underestimate the contribution of organic PM
found in remote regions when estimating light extinction (78 FR 3206,
January 15, 2013; U.S. EPA, 2012, p. IV-5). The available scientific
information and results of the air quality analyses indicate that it
may be appropriate to select inputs to the IMPROVE equation (e.g., the
multiplier for OC to OM) on a regional basis rather than a national
basis when calculating light extinction. This is especially true when
comparing sites with localized PM sources (such as sites in urban or
industrial areas) to sites with PM derived largely from biogenic
precursor emissions (that contribute to widespread secondary organic
aerosol formation), such as those in the southeastern U.S. The 2022 PA
notes, however, that conditions involving PM from such different
sources have not been well studied in the context of applying a
multiplier to estimate light extinction, contributing uncertainty to
estimates of light extinction for such conditions.
At the time of the 2012 review, the EPA noted that PM2.5
is the size fraction of PM responsible for most of the visibility
impairment in urban areas (77 FR 38980, June 29, 2012). Data available
at the time of the 2012 review suggested that, generally,
PM10-2.5 was a minor contributor to visibility impairment
most of the time (U.S. EPA, 2010b) although the coarse fraction may be
a major contributor in some areas in the desert southwestern region of
the U.S. Moreover, at the time of the 2012 review, there were few data
available from PM10-2.5 monitors to quantify the
contribution of coarse PM to calculated light extinction. Since that
time, an expansion in PM10-2.5 monitoring efforts has
increased the availability of data for use in estimating light
extinction with both PM2.5 and PM10-2.5
concentrations included as inputs in the equations. The analysis in the
2020 PA addressed light extinction at 20 of the 67 PM2.5
sites where collocated PM10-2.5 monitoring data were
available. Since that time, PM10-2.5 monitoring data are
available at more locations and the analyses presented in the 2022 PA
include those for light extinction estimated with coarse and fine PM at
all 60 sites. Generally, the contribution of the coarse fraction to
light extinction at these sites is minimal, contributing less than 1 dv
to the 3-year visibility metric (U.S. EPA, 2020b, section 5.2.1.2).
However, the 2022 PA notes that in the updated quantitative analyses,
only a few sites were in locations that would be expected to have high
concentrations of coarse PM, such as the Southwest. These results are
consistent with those in the analyses in the 2019 ISA, which found that
mass scattering from PM10-2.5 was relatively small (less
than 10%) in the eastern and northwestern U.S., whereas mass scattering
was much larger in the Southwest (more than 20%) particularly in
southern Arizona and New Mexico (U.S. EPA, 2019a, section 13.2.4.1, p.
13-36).
Overall, the findings of these updated quantitative analyses are
generally consistent with those in the 2012 and 2020 reviews. The 3-
year visibility metric was generally below 26 dv in most areas that
meet the current 24-hour PM2.5 standard. Small differences
in the 3-year visibility metric were observed between the variations of
the IMPROVE equation, which may suggest that it may be more appropriate
to use one version over another in different regions of the U.S. based
on PM characteristics such as particle size and composition to more
accurately estimate light extinction.
b. Non-Visibility Effects
Consistent with the evidence available at the time of the 2012 and
2020 reviews, and as described in detail in the 2022 PA (U.S. EPA,
2022b, section 5.3.2.2), the data remain insufficient to conduct
quantitative analyses for PM effects on climate and materials. For PM-
related climate effects, as explained in more detail in the proposal
(88 FR 5654, January 27, 2023), our understanding of PM-related climate
effects is still limited by significant key uncertainties. The recently
available evidence does not appreciably improve our understanding of
the spatial and temporal heterogeneity of PM components that contribute
to climate forcing (U.S. EPA, 2022b, sections 5.3.2.1.1 and 5.5).
Significant uncertainties also persist related to quantifying the
contributions of PM and PM components to the direct and indirect
effects on climate forcing, such as changes to the pattern of rainfall,
changes to wind patterns, and effects on vertical mixing in the
atmosphere (U.S. EPA, 2022b, sections 5.3.2.1.1 and 5.5). Additionally,
while improvements have been made to climate models since the
completion of the 2009 ISA, the models continue to exhibit variability
in estimates of the PM-related climate effects on regional scales
(e.g., ~100 km) compared to simulations at the global scale (U.S. EPA,
2022b, sections 5.3.2.1.1 and 5.5). While our understanding of climate
forcing on a global scale is somewhat expanded since the 2012 review,
significant limitations remain to quantifying potential adverse PM-
related climate effects in the U.S. and how they would vary in response
to incremental changes in PM concentrations across the U.S. As such,
while recent research is available on climate forcing on a global
scale, the remaining limitations and uncertainties are significant, and
the recent global scale research does not translate directly for use at
regional spatial scales. Therefore, the evidence does not provide a
clear understanding at the necessary spatial scales for quantifying the
relationship between PM mass in ambient air and the associated climate-
related effects in the U.S. that would be necessary to evaluate or
consider a level of air quality to protect against such effects and for
informing consideration of a national PM standard on climate in this
reconsideration (U.S. EPA, 2022b, section 5.3.2.2.1; U.S. EPA, 2019a,
section 13.3).
[[Page 16322]]
For PM-related materials effects, as explained in more detail in
the 2022 PA (U.S. EPA, 2022b, section 5.3.2.2), the available evidence
has been somewhat expanded to include additional information about the
soiling process and the types of materials impacted by PM. This
evidence provides some limited information to inform dose-response
relationships and damage functions associated with PM, although most of
these studies were conducted outside of the U.S. where PM
concentrations in ambient air are typically above those observed in the
U.S. (U.S. EPA, 2022b, section 5.3.2.1.2; U.S. EPA, 2019a, section
13.4). The evidence on materials effects characterized in the 2019 ISA
also includes studies examining effects of PM on the energy efficiency
of solar panels and passive cooling building materials, although the
evidence remains insufficient to establish quantitative relationships
between PM in ambient air and these or other materials effects (U.S.
EPA, 2022b, section 5.3.2.1.2). While the available evidence assessed
in the 2019 ISA is somewhat expanded since the time of the 2012 review,
quantitative relationships have not been established for PM-related
soiling and corrosion and frequency of cleaning or repair that further
the understanding of the public welfare implications of materials
effects (U.S. EPA, 2022b, section 5.3.2.2.2; U.S. EPA, 2019a, section
13.4). Therefore, there is insufficient information to inform
quantitative analyses assessing materials effects to inform
consideration of a national PM standard on materials in this
reconsideration (U.S. EPA, 2022b, section 5.3.2.2.2; U.S. EPA, 2019a,
section 13.4).
B. Conclusions on the Secondary PM Standards
In drawing conclusions on the adequacy of the current secondary PM
standards, in view of the advances in scientific knowledge and
additional information now available, the Administrator has considered
the evidence base, information, and policy judgments that were the
foundation of the 2020 decision and reflects upon the body of
information and evidence available in this reconsideration. In so
doing, the Administrator has taken into account both evidence-based and
quantitative information-based considerations, as well as advice from
the CASAC and public comments. Evidence-based considerations draw upon
the EPA's assessment and integrated synthesis of the scientific
evidence from studies evaluating welfare effects related to visibility,
climate, and materials associated with PM in ambient air as discussed
in the 2022 PA (summarized in sections V.B and V.D.2 of the proposal,
section V.A.2 above). The quantitative information-based considerations
draw from the results of the quantitative analyses of visibility
impairment presented in the 2022 PA (as summarized in section V.C of
the proposal and V.A.3 above) and consideration of these results in the
2022 PA.
Consideration of the scientific evidence and quantitative
information in the 2022 PA and by the Administrator is framed by
consideration of a series of policy-relevant questions. Section V.B.2
below summarizes the rationale for the Administrators proposed
decision, drawing from section V.D.3 of the proposal. The advice and
recommendations of the CASAC and public comments on the proposed
decision are addressed below in sections V.B.1 and V.B.3, respectively.
The Administrator's conclusions in this reconsideration regarding the
adequacy of the secondary PM standards and whether any revisions are
appropriate are described in section V.D.4.
1. CASAC Advice
In comments on the 2019 draft PA, the CASAC concurred with the
staff's overall preliminary conclusions that it is appropriate to
consider retaining the current secondary standards without revision
(Cox, 2019b). The CASAC ``finds much of the information . . . on
visibility and materials effects of PM2.5 to be useful,
while recognizing that uncertainties and controversies remain about the
best ways to evaluate these effects'' (Cox, 2019b, p. 13 of consensus
responses). Regarding climate, while the CASAC agreed that research on
PM-related effects has expanded since the 2012 review, it also
concluded that ``there are still significant uncertainties associated
with the accurate measurement of PM to the direct and indirect effects
of PM on climate'' (Cox, 2019b, pp. 13-14 of consensus responses). The
committee recommended that the EPA summarize the ``current scientific
knowledge and quantitative modeling results for effects of reducing
PM2.5'' on several climate-related outcomes (Cox, 2019b, p.
14 of consensus responses), while also recognizing that ``it is
appropriate to acknowledge uncertainties in climate change impacts and
resulting welfare impacts in the United States of reductions in
PM2.5 levels'' (Cox, 2019b, p. 14 of consensus responses).
When considering the overall body of scientific evidence and technical
information for PM-related effects on visibility, climate, and
materials, the CASAC agreed with the EPA's preliminary conclusions in
the 2019 draft PA, stating that ``the available evidence does not call
into question the protection afforded by the current secondary PM
standards and concurs that they should be retained'' (Cox, 2019b, p. 3
of letter).
In this reconsideration, the CASAC provided its advice regarding
the current secondary PM standards in the context of its review of the
2021 draft PA (Sheppard, 2022a). In its comments on the 2021 draft PA,
the CASAC first recognized that the scientific evidence is sufficient
to support a causal relationship between PM and visibility effects,
climate effects and materials effects.
With regard to visibility effects, the CASAC recognized that the
identification of a target level of protection for the visibility index
is based on a limited number of studies and suggested that ``additional
region- and view-specific visibility preference studies and data
analyses are needed to support a more refined visibility target''
(Sheppard, 2022a, p. 21 of consensus responses). While the CASAC did
not recommend revising either the target level of protection for the
visibility index or the level of the current 24-hour PM2.5
standard, they did state that a visibility index of 30 deciviews
``needs to be justified'' and ``[i]f a value of 20-25 deciviews is
deemed to be an appropriate visibility target level of protection, then
a secondary 24-hour PM2.5 standard in the range of 25-35
[micro]g/m\3\ should be considered'' (Sheppard, 2022a, p. 21 of
consensus responses).
The CASAC also recognized the limited availability of monitoring
methods and networks for directly measuring light extinction. As such,
they suggest that ``[a] more extensive technical evaluation of the
alternatives for visibility indicators and practical measurement
methods (including the necessity for a visibility FRM) is need for
future reviews'' (Sheppard, 2022a, p. 22 of consensus letter). The
majority of the CASAC ``recommend[ed] that an FRM for a directly
measured PM2.5 light extinction indicator be developed'' to
inform the consideration of the protection afforded by the secondary PM
standards against visibility impairment, the minority of the CASAC
``believe that a light extinction FRM is not necessary to set a
secondary standard protective of visibility'' (Sheppard, 2022a, p. 22
of consensus responses).
[[Page 16323]]
With regard to climate, the CASAC noted that ``there is a causal
relationship between PM and climate change, but large uncertainties
remain'' and recommended additional research (Sheppard, 2022a, p. 22 of
consensus responses). With respect to materials damage, the CASAC noted
that ``[q]uantitative information on the relationship between PM and
material damage is lacking'' and suggested some additional studies and
research approaches that could provide additional information on the
effects of PM on materials and the quantitative assessment of the
relationship between materials effects and PM in ambient air (Sheppard,
2022a, p. 23 of consensus responses).
2. Basis for the Proposed Decision
In reaching his proposed conclusions, the Administrator first
recognized that, consistent with the scope of this reconsideration, his
decision in this reconsideration will be focused only and specifically
on the adequacy of public welfare protection provided by the secondary
PM standards from effects related to visibility, climate, and
materials. He then considered the assessment of the current evidence
and conclusions reached in the 2019 ISA and ISA Supplement; the
currently available quantitative information, including associated
limitations and uncertainties, described in detail and characterized in
the 2022 PA; considerations and staff conclusions and associated
rationales presented in the 2022 PA; and the advice and recommendations
from the CASAC (88 FR 5655, January 27, 2023).
With respect to visibility, the Administrator noted the
longstanding body of evidence that demonstrates a causal relationship
between ambient PM and effects on visibility (U.S. EPA, 2019a, section
13.2). and that visibility impairment can have implications for
people's enjoyment of daily activities and for their overall sense of
well-being. Therefore, as in previous reviews, he considered the degree
to which the current secondary standards protect against PM-related
visibility impairment. In so doing, and consistent with previous
reviews, the Administrator considered the protection provided by the
current secondary standards against PM-related visibility impairment in
conjunction with the Regional Haze Program \166\ for protecting
visibility in Class I areas,\167\ which together would be expected to
achieve appropriate visual air quality across all areas (88 FR 5658,
January 27, 2023). The Administrator proposed to conclude that
addressing visibility impairment in Class I areas is beyond the scope
of the secondary PM NAAQS and that setting the secondary PM NAAQS at a
level that would remedy visibility impairment in Class I areas would
result in standards that are more stringent than is requisite.
---------------------------------------------------------------------------
\166\ The Regional Haze Program was established by Congress
specifically to achieve ``the prevention of any future, and the
remedying of existing, impairment of visibility in mandatory Class I
areas, which impairment results from man-made air pollution,'' and
that Congress established a long-term program to achieve that goal
(CAA section 169A).
\167\ In adopting section 169A, Congress set a goal of
eliminating anthropogenic visibility impairment at Class I areas, as
well as a framework for achieving that goal which extends well
beyond the planning process and timeframe for attaining secondary
NAAQS. Thus, the Regional Haze Program will continue to contribute
to reductions in visibility impairment in Class I areas.
---------------------------------------------------------------------------
In further considering what standards are requisite to protect
against adverse public welfare effects from visibility impairment, the
Administrator adopted an approach consistent with the approach used in
previous reviews (88 FR 5645, January 27, 2023). That is, he first
identified an appropriate target level of protection in terms of a PM
visibility index that accounts for the factors that influence the
relationship between particles in the ambient air and visibility (i.e.,
size fraction, species composition, and relative humidity). He then
considered air quality analyses examining the relationship between this
PM visibility index and the current secondary 24-hour PM2.5
standard in locations meeting the current 24-hour PM2.5 and
PM10 standards (U.S. EPA, 2022b, section 5.3.1.2; 88 FR
5650, January 27, 2023).
To identify a target level of protection, the Administrator first
considered the characteristics of the visibility index and defines its
elements (indicator, averaging time, form, and level). With regard to
the indicator for the visibility index, the Administrator recognized
that there is a lack of availability of methods and an established
network for directly measuring light extinction, consistent with the
conclusions reached in the 2022 PA (U.S. EPA, 2022b, section 5.3.1.1)
and with the CASAC's recommendation for additional research on direct
measurement methods for light extinction in their review of the 2021
draft PA (Sheppard, 2022a, p. 22 of consensus responses). Consistent
with the approaches used in reaching decisions in 2012 and 2020, given
the lack of such monitoring data, the Administrator preliminarily
judged that estimated light extinction, as calculated using one or more
versions of the IMPROVE algorithms, continues to be the most
appropriate indicator for the visibility index in this reconsideration
(88 FR 5659, January 27, 2023).
In further defining the characteristics of a visibility index based
on estimates of light extinction, the Administrator considered the
appropriate averaging time, form, and level of the index. With regard
to the averaging time and form, the Administrator noted that in
previous reviews, a 24-hour averaging time was selected and the form
was defined as the 3-year average of annual 90th percentile values. The
Administrator recognized that the evidence available in this
reconsideration and described in the 2022 PA continue to provide
support for the short-term nature of PM-related visibility effects.
Considering the available analyses of 24-hour and subdaily
PM2.5 light extinction, and noting that the CASAC did not
provide advice or recommendations with regard to the averaging time of
the visibility index, the Administrator preliminarily judged that the
24-hour averaging time continues to be appropriate for the visibility
index (88 FR 5659, January 27, 2023).
With regard to the form of the visibility index, the Administrator
noted that, consistent with the approach taken in other NAAQS,
including the current secondary 24-hour PM2.5 NAAQS, a
multi-year percentile form offers greater stability to the air quality
management process by reducing the possibility that statistically
unusual indicator values will lead to transient violations of the
standard. Using a 3-year average provides stability from the occasional
effects of inter-annual meteorological variability that can result in
unusually high pollution levels for a particular year (88 FR 5659,
January 27, 2023). In considering the percentile that would be
appropriate with the 3-year average, the Administrator first noted that
the Regional Haze Program targets the 20% most impaired days for
improvements in visual air quality in Class I areas.\168\ Based on
analyses examining 90th, 95th, and 98th percentile forms, the
Administrator preliminarily judged that a focus similar to the Regional
Haze Program focused on improving the 20% most impaired days suggest
that the 90th percentile, which represents the median of the 20% most
impaired days, such that 90% of days have visual air quality that is at
or below the target level of protection of the visibility
[[Page 16324]]
index, would be reasonably expected to lead to improvements in visual
air quality for the 20% most impaired days (88 FR 5659, January 27,
2023). In the analyses of percentiles, the results suggest that a
higher percentile value could have the effect of limiting the
occurrence of days with peak PM-related light extinction in areas
outside of Federal Class I areas to a greater degree. However, the
Administrator preliminarily concluded that it is appropriate to balance
concerns about focusing on the group of most impaired days with
concerns about focusing on the days with peak visibility impairment.
Additionally, the Administrator noted that the CASAC did not provide
advice or recommendations related to the form of the visibility index.
Therefore, the Administrator preliminarily judged that it remains
appropriate to define a visibility index in terms of a 24-hour
averaging time and a form based on the 3-year average of annual 90th
percentile values (88 FR 5659, January 27, 2023).
---------------------------------------------------------------------------
\168\ As noted above, the Administrator viewed the Regional Haze
Program as a complement to the secondary PM NAAQS, and thus took
into consideration its approach to improving visibility in
considering how to address visibility outside of Class I areas.
---------------------------------------------------------------------------
With regard to the level of the visibility index, the Administrator
first noted that the scientific evidence that is available to inform
the level of the visibility index is largely the same as in previous
reviews, and continues to provide support for a level within the range
of 20 to 30 dv (88 FR 5659-5660, January 27, 2023). The Administrator
recognized that significant uncertainties and limitations remained, in
particular those related to the public preference studies, including
methodological differences between the studies, and that the available
studies may not capture the full range of visibility preferences in the
U.S. population (88 FR 5659-5660, January 27, 2023). The Administrator
also noted that, in their review of the 2021 draft PA, the CASAC
recognized that a judgment regarding the appropriate target level of
protection for the visibility index is based on a limited number of
visibility preference studies, with studies conducted in the western
U.S. reporting public preferences for visibility impairment associated
with the lower end of the range of levels, while studies conducted in
the eastern U.S. reporting public preferences associated with the upper
end of the range (Sheppard, 2022a, p. 21 of consensus responses). The
Administrator noted that there have long been significant questions
about how to set a national standard for visibility that is not
overprotective for some areas of the U.S. In establishing the Regional
Haze Program to improve visibility in Class I areas, Congress noted
that ``as a matter of equity, the national ambient air quality
standards cannot be revised to adequately protect visibility in all
areas of the country.'' H.R. Rep. 95-294 at 205. Thus, in reaching his
proposed conclusion, the Administrator recognized that there are
substantial uncertainties and limitations in the public preference
studies that should be considered when selecting a target level of
protection for the visibility index and took the uncertainties and
variability inherent in the public preference studies into account. In
so doing, the Administrator first preliminarily judged that, consistent
with similar judgments in past reviews, it is appropriate to recognize
that the secondary 24-hour PM2.5 standard is intended to
address visibility impairment across a wide range of regions and
circumstances, and that the current standard works in conjunction with
the Regional Haze Program to improve visibility, and therefore, it is
appropriate to establish a target level of protection based on the
upper end of the range of levels. In considering the information
available in this reconsideration and the CASAC's advice, the
Administrator proposed to conclude that the protection provided by a
visibility index based on estimated light extinction, a 24-hour
averaging time, and a 90th percentile form, averaged over 3 years, set
at a level of 30 dv (the upper end of the range of levels) would be
requisite to protect public welfare with regard to visibility
impairment (88 FR 5660, January 27, 2023).
In preliminarily concluding that it remains appropriate in this
reconsideration to define the target level of protection in terms of a
visibility index based on estimated light extinction as described above
(i.e., with a 24-hour averaging time; a 3-year, 90th percentile form;
and a level of 30 dv), the Administrator next considered the degree of
protection from visibility impairment afforded by the existing
secondary standards. He considered the updated analyses of PM-related
visibility impairment presented in the 2022 PA (U.S. EPA, 2022b,
section 5.3.1.2), which reflect several improvements over the analyses
conducted in the 2012 review. Specifically, the updated analyses
examine multiple versions of the IMPROVE algorithm, including the
version incorporating revisions since the 2012 review (section
V.B.1.a), which provides an improved understanding of how variation in
equation inputs impacts calculated light extinction (U.S. EPA, 2022b,
Appendix D). In addition, unlike the analyses in the 2012 review and
the 2020 PA, all of the sites included in the analyses had
PM10-2.5 data available, which allows for better
characterization of the influence of the coarse fraction on light
extinction (U.S. EPA, 2022b, section 5.3.1.2).
The Administrator noted that the results of these updated analyses
are consistent with the results from the 2012 and 2020 reviews (88 FR
5660, January 27, 2023). Regardless of the IMPROVE equation used, these
analyses demonstrate that the 3-year visibility metric is at or below
28 dv in all areas meeting the current 24-hour PM2.5
standard (section V.C.1.b). Given the results of these analyses, the
Administrator preliminarily concluded that the updated scientific
evidence and technical information support the adequacy of the current
secondary PM2.5 and PM10 standards to protect
against PM-related visibility impairment. While the inclusion of the
coarse fraction had a relatively modest impact on calculated light
extinction in the analyses presented in the 2022 PA, he nevertheless
recognized the continued importance of the PM10 standard
given the potential for larger impacts in locations with higher coarse
particle concentrations, such as in the southwestern U.S., for which
only a few sites met the criteria for inclusion in the analyses in the
2022 PA (U.S. EPA, 2019a, section 13.2.4.1; U.S. EPA, 2022b, section
5.3.1.2).
With regard to the adequacy of the secondary 24-hour
PM2.5 standard, the Administrator noted that the CASAC
stated that ``[i]f a value of 20-25 deciviews is deemed to be an
appropriate visibility target level of protection, then a secondary 24-
hour PM2.5 standard in the range of 25-35 [micro]g/m\3\
should be considered'' (Sheppard, 2022a, p. 21 of consensus responses).
The Administrator recognized that the CASAC recommended that the
Administrator provide additional justification for a visibility index
target of 30 dv but did not specifically recommend that he choose an
alternative level for the visibility index. The Administrator
considered the CASAC's advice, together with the available scientific
evidence and quantitative information, in reaching his proposed
conclusions. He recognized conclusions regarding the appropriate weight
to place on the scientific and technical information examining PM-
related visibility impairment including how to consider the range and
magnitude of uncertainties inherent in that information is a public
welfare policy judgment left to the Administrator. As such, the
Administrator noted his conclusion on
[[Page 16325]]
the appropriate visibility index (i.e., with a 24-hour averaging time;
a 3-year, 90th percentile form; and a level of 30 dv) and his
conclusions regarding the quantitative analyses of the relationship
between the visibility index and the current secondary 24-hour
PM2.5 standard. In so doing, he proposed to conclude that
the current secondary standards provide requisite protection against
PM-related visibility effects (88 FR 5661, January 27, 2023).
In reaching his proposed conclusions, the Administrator also
recognized that the available evidence on visibility impairment
generally reflects a continuum and that the public preference studies
did not identify a specific level of visibility impairment that would
be perceived as ``acceptable'' or ``unacceptable'' across the whole
U.S. population. However, he noted that a judgment regarding the
appropriate target level of protection would take into consideration
the appropriate weight to place on the individual public preference
studies. In so doing, he noted that placing more weight on the public
preference study from Washington, DC, could provide support for a
target level of protection at or near 30 dv, whereas placing more
weight on the public preference study performed in the Phoenix, AZ,
study could provide support for a target level of protection below 30
dv and down to 25 dv. While the Administrator noted that, in their
review of the 2021 draft PA, the CASAC did not recommend revising the
level of the current 24-hour PM2.5 standard, the
Administrator recognized that they did recommend greater justification
for a target level of protection of 30 dv, and noted that if a target
level of protection of 20-25 dv was identified, then a secondary 24-
hour PM2.5 standard in the range of 25-35 [mu]g/m\3\ should
be considered (Sheppard, 2022a, p. 21 of consensus responses). For
these reasons, the Administrator solicited comment on his proposed
decision to retain the current secondary 24-hour PM2.5
standard, as well as the appropriateness of a target level of
protection for visibility below 30 dv and as low as 25 dv, and on
revising the level of the current secondary 24-hour PM2.5
standard to a level as low as 25 [micro]g/m\3\.
With respect to climate effects, the Administrator recognized that
a number of improvements and refinements have been made to climate
models since the time of the 2012 review. However, despite continuing
research and the strong evidence supporting a causal relationship with
climate effects (U.S. EPA, 2019a, section 13.3.9), the Administrator
noted that there are still significant limitations in quantifying the
contributions of the direct and indirect effects of PM and PM
components on climate forcing (U.S. EPA, 2022b, sections 5.3.2.1.1 and
5.5). He also recognized that models continue to exhibit considerable
variability in estimates of PM-related climate impacts at regional
scales (e.g., ~100 km), compared to simulations at the global scale
(U.S. EPA, 2022b, sections 5.3.2.1.1 and 5.5). As noted above, the
CASAC recognized a causal relationship between PM and climate effects
but also the large uncertainties associated with quantitatively
assessing such effects, particularly on a national level in the context
of a U.S.-based standard. These uncertainties led the Administrator to
preliminarily conclude that the scientific information available in
this reconsideration remains insufficient to quantify, with confidence,
the impacts of ambient PM on climate in the U.S. (U.S. EPA, 2022b,
section 5.3.2.2.1) and that there is insufficient information at this
time to revise the current secondary PM standards or to promulgate a
distinct secondary standard to address PM-related climate effects (88
FR 5661, January 27, 2023).
With respect to materials effects, the Administrator noted that the
available evidence continues to support the conclusion that there is a
causal relationship with PM deposition (U.S. EPA, 2019a, section 13.4).
He recognized that deposition of particles in the fine or coarse
fractions can result in physical damage and/or impaired aesthetic
qualities. Particles can contribute to materials damage by adding to
the effects of natural weathering processes and by promoting the
corrosion of metals, the degradation of painted surfaces, the
deterioration of building materials, and the weakening of material
components. While some recent evidence on materials effects of PM is
available in the 2019 ISA, the Administrator noted that this evidence
is primarily from studies conducted outside of the U.S. in areas where
PM concentrations in ambient air are higher than those observed in the
U.S. (U.S. EPA, 2019a, section 13.4). The CASAC also noted the lack of
quantitative information relating PM and material effects. Given the
limited amount of information on the quantitative relationships between
PM and materials effects in the U.S., and uncertainties in the degree
to which those effects could be adverse to the public welfare, the
Administrator preliminarily judged that the scientific information
available in this reconsideration remains insufficient to quantify,
with confidence, the public welfare impacts of ambient PM on materials
and that there is insufficient information at this time to revise the
current secondary PM standards or to promulgate a distinct secondary
standard to address PM-related materials effects (88 FR 5661, January
27, 2023).
Taken together, the Administrator proposed to conclude that the
scientific and technical information for PM-related visibility
impairment, climate impacts, and materials effects, with its attendant
uncertainties and limitations, supports the current level of protection
provided by the secondary PM standards as being requisite to protect
against known and anticipated adverse effects on public welfare. For
visibility impairment, this proposed conclusion reflected his
consideration of the evidence for PM-related light extinction, together
with his consideration of updated analyses of the protection provided
by the current secondary PM2.5 and PM10
standards. For climate and materials effects, this conclusion reflected
his preliminary judgment that, although it remains important to
maintain secondary PM2.5 and PM10 standards to
provide some degree of control over long- and short-term concentrations
of both fine and coarse particles, it is generally appropriate not to
change the existing secondary standards at this time and that it is not
appropriate to establish any distinct secondary PM standards to address
PM-related climate and materials effects at this time. As such, the
Administrator recognized that current suite of secondary standards
(i.e., the 24-hour PM2.5, 24-hour PM10, and
annual PM2.5 standards) together provide such control for
both fine and coarse particles and long- and short-term visibility and
non-visibility (e.g., climate and materials) \169\ effects related to
PM in ambient air. His proposed conclusions on the secondary standards
were consistent with advice from the CASAC, which noted substantial
uncertainties remain in the scientific evidence for climate and
materials effects. Thus, based on his consideration of the evidence and
analyses for PM-related welfare effects, as described above, and his
consideration of CASAC advice on the secondary standards, the
Administrator proposed not to change those standards (i.e., the current
24-hour and annual PM2.5 standards, 24-hour PM10
standard) at this time (88 FR 5662, January 27, 2023).
---------------------------------------------------------------------------
\169\ As noted earlier, other welfare effects of PM, such as
ecological effects, are being considered in the separate, on-going
review of the secondary NAAQS for oxides of nitrogen, oxides of
sulfur and PM.
---------------------------------------------------------------------------
[[Page 16326]]
3. Comments on the Proposed Decision
Of the public comments received on the proposal, very few were
specific to the secondary PM standards. Of those commenters who did
provide comments on the secondary PM standards, the majority support
the Administrator's proposed decision to retain the current standards.
Some commenters disagree with the Administrator's proposed conclusion
to retain the current secondary standards, primarily focusing their
comments on the need for a revised standard to protect against
visibility impairment. In addition to the comments addressed in this
notice, the EPA has prepared a Response to Comments document that
addresses other specific comments related to setting the secondary PM
standards. This document is available for review in the docket for this
rulemaking and through the EPA's NAAQS website (https://www.epa.gov/naaqs/particulate-matter-pm-air-quality-standards).
We first note that some commenters raise questions about the
protection provided by the secondary PM standards for ecological
effects (e.g., effects on ecosystems, ecosystem services, or species).
However, consistent with the 2016 IRP and as described in the proposal
(88 FR 5643, January 27, 2023), other welfare effects of PM, such as
the ecological effects identified by commenters, are being considered
as part of the separate, ongoing review of the secondary standards for
oxides of sulfur, oxides of nitrogen and PM, and thus, those comments
are beyond the scope of this action.
Of the comments addressing the proposed decision for the secondary
PM standards, many of the commenters support the Administrator's
proposed decision to retain the current secondary PM standards, without
revision. This group includes industries and industry groups and State
and local governments and organizations. All of these commenters
generally note their agreement with the rationale provided in the
proposal, with a focus on the strength of the available scientific
evidence for PM-related welfare effects. Most also recognize that the
scientific evidence and quantitative information available in this
reconsideration have not substantially altered our previous
understanding of PM-related effects on non-ecological welfare effects
(i.e., visibility, climate, and materials) and do not call into
question the adequacy of the current secondary standards. They find the
proposed decision not to change the standards at this time to be well
supported and a reasonable exercise of the Administrator's public
welfare policy judgment under the CAA. The EPA agrees with these
comments regarding the adequacy of the current secondary PM standards
and the lack of support for revision of these standards at this time.
The EPA received relatively few comments on the proposed decision
that it is not appropriate to establish any distinct secondary PM
standards to address PM-related climate effects. Several commenters
agree that the available scientific evidence provides support for the
2019 conclusion that there is a causal relationship between PM and
climate effects, and the commenters also agree with the EPA that the
currently available information is not sufficient for supporting
quantitative analyses for the climate effects of PM in ambient air.
These commenters support the Administrator's proposed decision not to
set a distinct standard for climate.
There were also very few commenters who commented on the proposed
decision that it is not appropriate to establish any distinct secondary
PM standards to address PM-related materials effects. As with comments
on climate effects, commenters generally agree with the EPA that the
evidence is not sufficient to support quantitative analyses for PM-
related materials effects. However, some commenters contend that EPA
failed to explain in the proposal how the current standard is
appropriate to protect materials from the effects of PM. These
commenters disagree with the EPA's conclusion that quantitative
relationships have not been established for PM-related soiling and
corrosion and frequency of cleaning or repair of materials, and cite to
several studies conducted outside the U.S. that they contend that the
EPA should consider since the same materials are present in the U.S.
They further contend that, in discussing the available scientific
evidence in the 2019 ISA for studies conducted outside of the U.S., the
EPA did not provide references to these studies and, therefore, the
public is unable to comment on these studies. They further State that
EPA failed to consider the following information: (1) Recent work
related to soiling of photovoltaic modules and other surfaces, and; (2)
damage and degradation resulting from oxidant concentrations and solar
radiation for a number of materials, including polymeric materials,
plastic, paint, and rubber. These commenters further assert that the
EPA failed to propose a standard that provides requisite protection
against materials effects attributable to PM.
As an initial matter, we note that the commenters submitted the
same comments related to materials effects during the 2020 review.
Consistent with our response in the 2020 notice of final rulemaking (85
FR 82737, December 18, 2020), we disagree with the commenters that the
EPA failed to consider the relevant scientific information about
materials effects available in this reconsideration. The 2019 ISA
considered and included studies related to materials effects of PM,
including studies conducted in and outside of the U.S., on newly
studied materials including photovoltaic modules that were published
prior to the cutoff date for the literature search.\170\ These include
the Besson et al. (2017) study referenced by the commenters (U.S. EPA,
2019a, section 13.4.2). The Gr[oslash]ntoft et al. (2019) study
referenced by the same commenters was published after the cutoff date
for the literature search for the 2019 ISA. However, the EPA
provisionally considered new studies in responding to comments in the
2020 review, including the new studies highlighted by the commenters in
their comments on the 2020 notice of proposed rulemaking, in the
context of the findings of the 2019 ISA (see Appendix in U.S. EPA,
2020a).\171\ Based on the provisional consideration, the EPA concluded
in the 2020 review that the new studies are not sufficient to alter the
conclusions reached in the 2019 ISA regarding PM and materials effects.
For example, the Gr[oslash]ntoft et al. (2019) study was based on
European air pollution which as the EPA has noted has higher
concentrations (as well as diversity in sources, such as light duty
diesel engines) compared to the U.S.. Thus, the EPA did not find it
necessary or appropriate to reopen the air quality criteria to consider
this study because it would not have been an adequate basis on which to
set a NAAQS. As discussed in section I, when the EPA decided to
reconsider the standards, it also decided to reopen the air quality
criteria to a limited degree, based on its judgment that certain new
studies were likely to be useful in reconsidering the standards.
[[Page 16327]]
Based on the provisional consideration in the 2020 review and the
significant data gaps that existed at that time, the EPA did not
include these studies within the scope of the 2022 ISA Supplement
because, although these studies provide additional support for PM-
related materials, the studies would not support quantitative analyses
or alternative conclusions regarding these effects. As described in
section I.C.5.b above, the ISA Supplement focuses on a thorough
evaluation of some studies that became available after the literature
cutoff date of the 2019 ISA that could either further inform the
adequacy of the current PM NAAQS or address key scientific topics that
have evolved since the literature cutoff date for the 2019 ISA. In
developing the ISA Supplement, the EPA focused on the non-ecological
welfare effects for which the evidence supported a ``causal
relationship'' and for which quantitative analyses could be supported
by the evidence because those were the welfare effects that were most
useful in informing conclusions in the 2020 PA. While the 2020 PA
considered the broader set of evidence for materials effects, it
concluded that there remained `substantial uncertainties with regard to
the quantitative relationships with PM concentrations and concentration
patterns that limit[ed] [the] ability to quantitatively assess the
public welfare protection provided by the standards from these effects'
(U.S. EPA, 2020b).'' Therefore, the ISA Supplement did not include an
evaluation of scientific evidence for PM-related materials effects.
However, the EPA has once again provisionally considered new studies in
this reconsideration, including the studies highlighted by the
commenters, in the context of the 2019 ISA and concludes that, as in
the 2020 review, these studies are not sufficient to alter the
conclusions reached in the 2019 ISA regarding PM and materials effects
or to provide sufficient information on which to base a secondary
NAAQS. The EPA agrees there is a causal relationship between the
presence of PM in the ambient air and materials effects, but to set a
standard, the EPA needs not only to understand at what point materials
effects become adverse to public welfare but to be able to relate
specific concentrations of ambient PM to those levels of materials
effects. Given the significant gaps in the evidence, particularly given
that the majority of the recent evidence has been conducted outside of
the U.S., establishing any quantitative relationships between particle
size, concentration, chemical components, and specific measures of
materials damage, such as frequency of painting or repair of materials,
the EPA finds the evidence is insufficient to support a secondary NAAQS
to protect against materials effects.
---------------------------------------------------------------------------
\170\ As noted earlier in section V, the 2019 ISA ``identified
and evaluated studies and reports that that have undergone
scientific peer review and were published or accepted for
publication between January 1, 2009, and March 31, 2017. A limited
literature update identified some additional studies that were
published before December 31, 2017'' (U.S. EPA, 2019a, Appendix, p.
A-3).
\171\ As discussed in section I.D, the EPA has provisionally
considered studies that were highlighted by commenters and that were
published after the 2019 ISA. These studies are generally consistent
with the evidence assessed in the 2019 ISA, and they do not
materially alter our understanding of the scientific evidence or the
Agency's conclusions based on that evidence.
---------------------------------------------------------------------------
With regard to studies conducted outside of the U.S., including
those referenced by the commenters, as described in the proposal, in
reaching his proposed conclusion, the Administrator recognized that
while there was some newly available information related to materials
effects of PM included in the 2019 ISA, ``this evidence is primarily
from studies conducted outside of the U.S. in areas where PM
concentrations in ambient air are higher than those observed in the
U.S. (U.S. EPA, 2019a, section 13.4)'' (88 FR 5661, January 27, 2023).
We disagree with the commenters that EPA did not provide references for
these studies, nor that the lack of references inhibited the public's
ability to provide comment on this proposed conclusion. First, the
reference to section 13.4 in the 2019 ISA is a direct citation to the
evaluation of newly available studies on PM-related materials effects,
which includes citations for all materials effects evidence considered
in the 2020 review and in this reconsideration. Second, section
5.3.2.1.2 of the 2022 PA considers the available scientific evidence
for PM-related materials effects--including citations to the studies
newly available in the 2019 ISA--and how that evidence informs
conclusions regarding the adequacy of the standard (U.S. EPA, 2022b,
section 5.3.2.1.2). Therefore, the EPA disagrees that the proposal
failed to provide the proper references to the studies conducted
outside of the U.S., and that the public was not provided the
opportunity to provide comment on these studies.
Moreover, we disagree with the commenters that the EPA failed to
consider quantitative information from studies available in this
reconsideration. As detailed in sections 5.3.2.1.2 and 5.3.2.2 of the
2022 PA, and consistent with the information available in the 2020
review, a number of new studies are available that apply new methods to
characterize PM-related effects on previously studied materials;
however, the evidence remains insufficient to relate soiling or damage
to specific levels of PM in ambient air or to establish quantitative
relationships between PM and materials degradation. The uncertainties
in the evidence identified in the 2012 review persist in the evidence
in the 2020 review and in this reconsideration, with significant
uncertainties and limitations to establishing quantitative
relationships between particle size, concentration, chemical
components, and frequency of painting or repair of materials. While
some new evidence is available in the 2019 ISA, overall, the data are
insufficient to conduct quantitative analyses for PM-related materials
effects. Quantitative relationships have not been established between
characteristics of PM and frequency of repainting or cleaning of
materials, including photovoltaic panels and other energy-efficient
materials, that would help inform our understanding of the public
welfare implications of soiling in the U.S. (U.S. EPA, 2022b, section
5.3.2.2.2; U.S. EPA, 2019a, section 13.4). Similarly, the information
does not support quantitative analyses between microbial deterioration
of surfaces and the contribution of carbonaceous PM to the formation of
black crusts that contribute to soiling (U.S. EPA, 2022b, section
5.3.2.2.2; U.S. EPA, 2019a, section 13.4). We also note that
quantitative relationships are difficult to assess, in particular those
characterized using damage functions as these approaches depend on
human perception of the level of soiling deemed to be acceptable and
evidence in this area remains limited in this reconsideration (U.S.
EPA, 2022b, section 5.3.2.1.2). Additionally, we note the CASAC's
concurrence with conclusions in the 2020 PA (Cox, 2019b, p. 13 of
consensus responses) and the 2022 PA (Sheppard, 2022a, p. 23 of
consensus responses) that uncertainties remain about the best way to
evaluate materials effects of PM in ambient air. Further, no new
studies are available in this reconsideration to link human perception
of reduced aesthetic appeal of buildings and other objects to materials
effects and PM in ambient air. Finally, uncertainties remain about
deposition rates of PM in ambient air to surfaces and the interaction
of PM with copollutants on these surfaces (U.S. EPA, 2022b, section
5.6).
With respect to the commenters' assertion that the EPA failed to
consider information related to materials damage and degradation from
oxidant concentrations and solar radiation for a variety of materials,
we first note that, even assuming these sources of materials damage are
within the scope of this review of the PM NAAQS, the commenter did not
provide any references to the scientific studies that they suggest that
the EPA did not consider. Despite the lack of a list of specific
references from the commenter, we note that the 2019 ISA considered a
number of studies that examined the
[[Page 16328]]
relationships between PM and several of the materials listed by the
commenters (e.g., paint, plastic, rubber). However, as described in the
2022 PA, these studies did not provide additional information regarding
quantitative relationships between PM and materials that could inform
quantitative analyses (U.S. EPA, 2022b, sections 5.3.2.1.2 and
5.3.2.2.2), nor did they alter conclusions regarding the adequacy of
the current standard (U.S. EPA, 2022b, section 5.5).
As summarized above and in the proposal, the evidence in the 2020
review and in this reconsideration for PM-related effects on materials
is not substantively changed from that in the 2012 review. There
continues to be a lack of evidence related to materials effects that
establishes quantitative relationships and supports quantitative
analyses of PM-related materials soiling or damage. While the
information available in the 2020 review and in this reconsideration
continues to support a causal relationship between PM in ambient air
and materials effects (U.S. EPA, 2019a, section 13.4), the EPA is
unable to relate soiling or damage to specific levels of PM in ambient
air and is unable to evaluate or consider a level of air quality to
protect against such materials effects. Although the EPA did not
propose a distinct level of air quality or a national standard based on
air quality impacts (88 FR 5662, January 27, 2023), we did identify
data gaps that prevented us from doing so. The EPA identified a number
of key uncertainties and areas of future research (U.S. EPA, 2022b,
section 5.6) that may inform consideration of the materials effects of
PM in ambient air in future reviews of the PM NAAQS. The EPA notes that
one commenter objected to the Administrator's proposed conclusion in
the proposal (88 FR 5661, January 27, 2023) that in light of the
available evidence for PM-related impacts on climate and on materials
that it is appropriate not to change the existing secondary standards
at this time. The EPA has explained, in both the proposal and this
final action, the basis for its conclusion that there is insufficient
evidence to identify any particular secondary standard or standards
that would provide requisite protection against climate effects or
materials damage. The EPA acknowledges that, as a result, the adoption
of any distinct secondary PM standards for those effects would be
inconsistent with the requirements of the CAA. The EPA is clarifying
that it is not basing its decisions on secondary standards in this
reconsideration to address these welfare effects because it has
concluded that the available scientific evidence is insufficient to
allow the Administrator to make a reasoned judgment about what specific
standard(s) would be requisite to protect against known or anticipated
adverse effects to public welfare from PM-related materials damage or
climate effects.
Some commenters agree with the Administrator's proposed conclusion
that a target level of protection for visibility of 30 dv and the level
of the secondary 24-hour PM2.5 standard of 35 [micro]g/m\3\
continues to be adequate to protect visibility, highlighting
improvements in visibility in the U.S. Other commenters who disagree
with the proposed decision indicated support for a more stringent
standard for visibility impairment, although some of these commenters
did not necessarily specify the alternative standard that would, in
their judgment, address their concerns related to various aspects of
the EPA's proposal, including the available public preference studies,
specific aspects of the visibility index, and the target level of
protection identified by the Administrator. Rather, most commenters
focused on particular aspects of the visibility metric underlying the
current secondary 24-hour PM2.5 standard, including the
form, averaging time, and target level of protection necessary to
protect against visibility impairment.
With regard to the commenters' assertion that the current secondary
standards are inadequate to protect the public welfare from PM-related
visibility impairment, the EPA disagrees that the currently available
information is sufficient to suggest that a more stringent standard is
warranted. The EPA identified and addressed in great detail the
limitations and uncertainties associated with the public preference
studies as a part of the 2012 review (78 FR 3210, January 15, 2013).
Given that the evidence related to public preferences has not
substantially changed since the 2012 review, the EPA reiterated the
limitations and uncertainties inherent in the evidence as a part of the
2020 PA (U.S. EPA, 2020b, section 5.5), as well as in the 2022 PA for
this reconsideration (U.S. EPA, 2022b, section 5.6). The 2022 PA
highlights key uncertainties associated with public perception of
visibility impairment and identifies areas for future research to
inform future PM NAAQS reviews, including those raised by the
commenters (U.S. EPA, 2022b, section 5.6). Specifically, the EPA agrees
with commenters that there are several areas where additional
information would reduce uncertainty in our interpretation of the
available information for purposes of characterizing visibility
impairment. As described in more detail in the 2020 PA (U.S. EPA,
2020b, p. 5-41) and the 2022 PA (U.S. EPA, 2022b, p. 5-53), briefly,
these areas include: (1) Expanding the number and geographic coverage
of preference studies in urban, rural, and Class I areas; (2)
evaluating visibility preferences of the U.S. population today, given
that the preference studies were conducted more than 15 years ago,
during which time air quality in the U.S. has improved; (3) accounting
for the influence of varying study methods may have on an individual's
response as to what level of visibility impairment is acceptable, and;
(4) information on people's judgments on acceptable visibility based on
factors that can influence their perception of visibility (e.g.,
duration of impairment experiences, time of day, frequency of
impairment).
However, the EPA disagrees with the commenters that the current
secondary PM standards are inadequate and should be made more stringent
because of the limitations and uncertainties associated with the
available public preference studies. The EPA does not view the
limitations of the preference studies and other available evidence as
so significant as to render the EPA unable to identify a secondary
standard to protect against the adverse effects of PM on visibility,
but the EPA also does not believe that the limitations themselves mean
that the standards are inadequate. In fact, there is a limited amount
of recently available scientific evidence to further inform our
understanding of public preferences and visibility impairment is
recognized by the Administrator in reaching his proposed decision not
to change the current secondary PM standards at this time, given that
the evidence base is largely the same as at the time of the 2012 and
2020 reviews.
These same commenters further contend that the EPA failed to use
the latest science to develop a visibility index, stating that the EPA
failed to consider the contrast of distance methodology employed in a
recent meta-analysis of available preference studies (Malm et al.,
2019). Commenters claim that the EPA draws conclusions from the Malm et
al. (2019) study about how to relate contrast to acceptable visibility
preferences in the 2022 ISA Supplement, yet ignores the findings of the
study and fails to consider the ``contrast of distance'' methodology in
the 2022 PA and the proposal, thereby, in their view, departing from
the CASAC's advice to consider this evidence in setting the secondary
[[Page 16329]]
standard. Finally, the commenters assert that the EPA did not explain
why the available public preference studies are adequate for analysis
using a light extinction approach but not using the contrast of
distance approach, and that such differential treatment is arbitrary.
We disagree with the commenters that the EPA did not use the latest
science in evaluating the visibility index, and that the EPA failed to
consider the contrast of distance methodology used in Malm et al.
(2019). As the commenters state, the Malm et al. (2019) study was
included in the ISA Supplement (U.S. EPA, 2022a, section 4.2.1).
However, the EPA disagrees with the assertion that the ISA Supplement
reached conclusions about how to relate contrast to acceptable
visibility preferences. The ISA Supplement provided an overview of the
Malm et al. (2019) study, stating that ``[t]he main conclusion of this
study was that the level of acceptable visual air quality is more
consistent across studies using metrics that evaluate the distinction
of an object from a background than using metrics that evaluate the
greatest distance at which an object can be observed.'' Furthermore,
the statements that the commenters are referencing in support of this
statement (i.e., U.S. EPA, 2022b, pp. 4-5-4-6) are in fact the
conclusions of the study itself, rather than conclusions of the EPA.
For example, the ISA Supplement notes that ``Malm et al. (2019)
suggested that scene-dependent metrics like contrast, which integrate
the effects of bext along the sight paths between observers
and landscape features, are better predictors of preference levels than
universal metrics like light extinction.'' The suggestion that the
contrast of distance methodology is a better predictor than light
extinction is one of the study authors, not the EPA. The EPA has not
reached a conclusion on whether contrast of distance methodology would
be a more appropriate indicator for a visibility index than estimated
light extinction because the EPA finds that there is insufficient
information in the record at this time to support that it is practical
to evaluate, much less adopt, the contrast of distance methodology on a
national basis. Specifically, the Malm et al. (2019) study does not
provide as a part of their publication the specific input values to the
equation to calculate the contrast of distance associated with the
available public preference studies (e.g., sight paths from the
images), nor do the preference studies present or make publicly
available these data in their publications. In the absence of
additional studies or publicly available data to further evaluate the
contrast of distance methodology, the EPA is unable to consider
contrast of distance as an alternative to estimated light extinction in
this reconsideration, although we note that it may be appropriate to
evaluate it more closely in future reviews.
In reaching conclusions regarding the appropriate indicator for the
visibility index, the 2022 PA specifically notes ``that limited new
research is available on methods of characterizing visibility or on how
visibility is valued by the public, such as visibility preference
studies. Thus, while limited new research has further informed our
understanding of the influence of atmospheric components of
PM2.5 on light extinction, the available evidence to inform
consideration of the public welfare implications of PM-related
visibility impairment remains relatively unchanged'' (U.S. EPA, 2022b,
p. 5-50). The EPA again notes in the proposal that ``there are very few
studies available that use scene-dependent metrics (i.e., contrast) to
evaluate public preference information, which makes it difficult to
evaluate them as an alternative to the light extinction approach'' (88
FR 5649-5650, January 27, 2023). To further expand on this statement,
the Malm et al. (2019) study does not provide enough information to
replicate the results of their contrast of distance approach to allow
for a comprehensive evaluation of the potential use of this methodology
in considering the results of the public preference studies for
determining the target level of protection for visibility.
Some commenters suggests that the methodology could be approximated
by simply ensuring that people could always see distant scenic
elements, and that characterizing typical average and/or maximal
viewing distances cross different geographical areas and regions would
be a straightforward Geographical Information Systems (GIS) exercise.
The EPA disagrees that this assessment would be straightforward, given
the lack of data establishing viewing distances in the available
scientific record and the diversity of distance to scenic elements
across different areas and regions of the U.S., and finds that this
approach is also not practical to adopt in this reconsideration.
Finally, while the Malm et al. (2019) study is using an alternative
approach for evaluating public preferences and acceptability, we note
that this study is evaluating the same public preference studies that
have been available for the past several decades. For these reasons,
the EPA disagrees with the commenters' allegation that the EPA ignored
the findings of the Malm et al. (2019) study and failed to consider the
contrast of distance methodology in the 2022 PA and the proposal, and
ignored the CASAC's advice to consider this study. The ISA Supplement
and the 2022 PA considered the Malm et al. (2019) study, along with the
full body of available scientific evidence, and took into account the
uncertainties and limitations associated with the evidence for
visibility preferences, in reaching conclusions regarding the adequacy
of the secondary 24-hour PM2.5 standard (U.S. EPA, 2022b,
pp. 5-24-5.25, 5-50).
Several comments in support of revising the secondary 24-hour
PM2.5 standard to protect against visibility generally
recommend revisions to the elements of the standard and visibility
index (indicator, averaging time, form, and level) consistent with
those supported by the CASAC and public comments in previous PM NAAQS
reviews. Some commenters assert that the EPA's approach in the 2022 PA
and in the proposal for this reconsideration did not evaluate options
for alternative secondary PM standards and thereby is flawed. We
address comments on the elements of a visibility index and a revised
standard for visibility effects below.
As an initial matter, the EPA disagrees to the extent commenters
are suggesting that the PA is legally required to analyze options for
alternative standards. The PA is a document developed by the EPA in
order to assist the Administrator and the CASAC in reaching conclusions
regarding the adequacy of the current standards, and its scope is
determined by the EPA. Moreover, the 2022 PA did assess a wide range of
information relevant to the Administrator's decision and considered a
range of potential standards.
First, in developing the 2022 PA and in responding to CASAC's
advice and recommendations during its review of the 2021 draft PA, the
EPA expanded upon its discussion of determining the target level of
protection for the visibility index and considered the extent to which
the available scientific information would alter regarding the
visibility index and the appropriate target level of protection against
PM-related visibility effects (U.S. EPA, 2022b, pp. 5-27-5-29). This
detailed discussion expands the consideration of the target level of
protection for the visibility index presented in the 2020 PA (U.S. EPA,
2020b) and the 2021 draft PA (U.S. EPA, 2021c), neither of which
specifically considered the elements of the visibility index in
determining the appropriate target level of protection. In
[[Page 16330]]
considering the available information in the 2022 PA, the EPA concluded
that the available information continued to provide support for a
visibility index with a level of 30 dv, with estimated light extinction
as the indicator, a 24-hour averaging time, and a 90th percentile form,
averaged over three years.
Additionally, in summarizing the air quality and quantitative
information in the proposal for this reconsideration, the EPA further
expands upon the discussion added to the 2022 PA related to the target
level of protection in terms of a PM2.5 visibility index. In
so doing, the EPA considers even more extensively the available public
preference studies and quantitative analyses (88 FR 5651-5652, January
27, 2023). In particular, there is a more detailed discussion of the
public preference studies, including the levels of impairment
determined to be ``acceptable'' by at least 50 percent of study
participants and the methodologies used in the studies, including
uncertainties and limitations associated with the methodologies (88 FR
5652, January 27, 2023). In reaching a proposed decision regarding the
adequacy of the secondary PM standards, as well as the appropriate
target level of protection for the visibility index, the Administrator
considered the available scientific evidence and quantitative analyses,
as well as judgments about how to consider the range and magnitude of
uncertainties that are inherent in the scientific evidence and
analyses. In so doing, the Administrator proposed to conclude that the
protection provided by a visibility index based on estimated light
extinction, a 24-hour averaging time, and a 90th percentile form,
averaged over 3 years, set at a level of 30 dv would be requisite to
protect public welfare with regard to visibility impairment (88 FR
5660, January 27, 2023).
Having provisionally concluded that it was appropriate to define
the target level of protection in terms of a visibility index based on
estimated light extinction as described above (i.e., with a 24-hour
averaging time; a 3-year, 90th percentile form; and a level of 30 dv),
the Administrator next considered the degree of protection from
visibility afforded by the current secondary PM standards. In so doing,
he considered the updated analyses of PM-related visibility impairment
presented in the 2022 PA (U.S. EPA, 2022b, section 5.3.1.2) and
described in more detail in the proposal (88 FR 5656, January 27,
2023), which included estimating light extinction using multiple
versions of the IMPROVE algorithm and inclusion of PM10-2.5
data at all sites to allow for better characterization of the influence
of the coarse fraction of PM on light extinction. The Administrator
noted that the results of the analyses in the 2022 PA were consistent
with those from the 2012 and 2020 reviews. He also recognized that,
regardless of the IMPROVE equation that was used, the analyses
demonstrated that the 3-year visibility metric is at or below 28 dv in
all areas meting the current 24-hour PM2.5 standard (88 FR
5657, January 27, 2023). The Administrator also noted that, in their
review of the 2021 draft PA, the CASAC stated that ``[i]f a value of
20-25 deciviews is deemed to be an appropriate visibility target level
of protection, then a secondary 24-hour standard in the range of 25-35
[micro]g/m\3\ should be considered (Sheppard, 2022a, p. 21 of consensus
responses). The Administrator recognized that while the CASAC
recommended that additional justification be provided for a visibility
index target level of protection of 30 dv, they did not specifically
recommend that he choose an alternative level for the visibility index.
Therefore, the Administrator considered the available scientific
evidence, quantitative information, and the CASAC's advice in reaching
his proposed conclusions. The Administrator recognized conclusions
regarding the appropriate weight to place on the scientific and
technical information, including how to consider the range and
magnitude of uncertainties inherent in that information, is a public
welfare policy judgment left to the Administrator. As such, the
Administrator noted his preliminary conclusion on the appropriate
visibility index (i.e., with a 24-hour averaging time; a 3-year, 90th
percentile form; and a level of 30 dv) and his preliminary conclusions
regarding the quantitative analyses of the relationship between the
visibility index and the current secondary 24-hour PM2.5
standard. In so doing, he proposed to conclude that the current
secondary standards provide requisite protection against PM-related
visibility effects (88 FR 5661, January 27, 2023).
However, the Administrator additionally recognized that the
available evidence on visibility impairment generally reflects a
continuum and that the public preference studies did not identify a
specific level of visibility impairment that would be perceived as
``acceptable'' or ``unacceptable'' across the whole U.S. population. He
noted a judgment of a target level of protection, below 30 dv and down
to 25 dv, could be supported if more weight was put on the public
preference study performed in the Phoenix, AZ, study (BBC Research &
Consulting, 2003). As described above, while the Administrator noted
that the CASAC did not recommend revising the level of the current 24-
hour PM2.5 standard in their review of the 2021 draft PA,
they did state that, should an alternative level be considered for the
visibility index, revisions to the secondary 24-hour PM2.5
standard should also be considered (Sheppard, 2022a, p. 21 of consensus
responses). Thus, the Administrator solicited comment on the
appropriateness of a target level of protection for visibility below 30
dv and down as low as 25 dv, and of revising the level of the current
secondary 24-hour PM2.5 standard to a level as low as 25
[micro]g/m\3\ (88 FR 5662, January 27, 2023), and the Administrator
considered these public comments in reaching his final decision on the
secondary standards. Thus, the EPA disagrees that the 2022 PA and the
proposal did not adequately consider options for revising the secondary
PM NAAQS.
With regard to the elements of the visibility index, in considering
the adequacy of the current secondary 24-hour PM2.5 standard
to protect against visibility impairment, as described in the proposal
(88 FR 5658-5660, January 27, 2023), the Administrator first defined an
appropriate target level of protection in terms of a PM visibility
index. In considering the information available in this reconsideration
and the CASAC's advice, the Administrator proposed to conclude that the
protection provided by a visibility index based on estimated light
extinction, a 24-hour averaging time, and 90th percentile form,
averaged over 3 years, set at a level of 30 dv, would be requisite to
protect public welfare with regard to visibility impairment (88 FR
5660, January 27, 2023).
In defining this target level of protection, the Administrator
first considered the indicator of such an index. He noted that, given
the lack of availability of methods and an established network for
directly measuring light extinctions, a visibility index based on
estimates of light extinction by PM2.5 components derived
from an adjusted version of the original IMPROVE algorithm would be
most appropriate, consistent with the 2012 and 2020 reviews. As
described in the proposal (88 FR 5649, January 27, 2023) and above
(section V.A.2), the IMPROVE algorithm estimates light extinction using
routinely monitored components of PM2.5 and
PM10-2.5, along with estimates of relative humidity. The
[[Page 16331]]
Administrator, while recognizing that some revisions to the IMRPOVE
algorithm were newly available in the 2020 review, noted that the
fundamental relationship between ambient PM and light extinction has
changed very little and the different versions of the IMPROVE
algorithms can appropriately reflect this relationship across the U.S.
(88 FR 5658-5659, January 27, 2023). As such, he judged that defining a
target level of protection in terms of estimated light extinction
continues to be a reasonable approach in this reconsideration.
Some commenters who criticized the EPA's interpretation and
application of the Malm et al. (2019) study also contend that an
indicator based on the contrast of distance would be a significant
improvement over the current indicator for the visibility index and
would more accurately evaluate public preferences. However, as
described in the 2022 PA (U.S. EPA, 2022b, section 5.3.1.1), while
scene-dependent metrics, such as contrast, may be useful alternative
predictors of preferences compared to universal metrics like light
extinction, there are a very limited number of studies that use such
metrics to evaluate public preferences of visibility impairment and
there is a lack of scientific evidence that supports one metric over
another. Moreover, the EPA finds that even if the Administrator agreed
that the contrast of distance methodology was an improvement over light
extinction, there is insufficient information available to evaluate and
adopt contrast of distance as an indicator for a national visibility
target at this time. While, in its review of the 2021 draft PA the
CASAC suggested that the EPA consider this method in developing the
secondary PM standards, the CASAC also noted that ``more extensive
technical evaluation of the alternatives for visibility indicators and
practical measurement methods'' is needed to inform future reviews of
the secondary PM standards (Sheppard, 2022a, p. 22 of consensus
responses). The CASAC did not recommend using a different indicator for
this reconsideration, with the majority of CASAC members reiterated
past advice recommending development of a visibility FRM for a directly
measured PM2.5 light extinction indicator (Sheppard, 2022a,
p. 22 of consensus responses), a recommendation that was supported by
other public commenters as well, and the minority of the CASAC
suggested that such an FRM is not necessary. For these reasons, the EPA
does not consider it feasible or appropriate to define the visibility
index in terms of a contrast of distance indicator at this time.
With regard to averaging time, some commenters suggested to the EPA
that a secondary standard with a different form than the primary
standard may be a more relevant for welfare effects. While they do not
recommend a specific alternative form, the commenters point to CASAC
advice in past reviews where the CASAC stated that a subdaily standard
based on daylight hours better reflects visibility impairment.
In defining the characteristics of a visibility index, the EPA
continues to believe that a 24-hour averaging time is reasonable. This
is in part based on analyses conducted in the 2012 review that showed
relatively strong correlations between 24-hour and subdaily (i.e., 4-
hour average) PM2.5 light extinction (88 FR 5659, January
27, 2023; 85 FR 82740, December 18, 2020; 78 FR 3226, January 15,
2013), indicating that a 24-hour averaging time is an appropriate
surrogate for the subdaily time periods relevant for visual perception.
The EPA believes that these analyses continue to provide support for a
24-hour averaging time for the visibility index in this
reconsideration. The EPA also recognizes that the longer averaging time
may be less influenced by atypical conditions and/or atypical
instrument performance (88 FR 5659, January 27, 2023; 85 FR 82740,
December 18, 2020; 78 FR 3226, January 15, 2013). When taken together,
the available scientific information and updated analyses of calculated
light extinction available in this reconsideration continue to support
that a 24-hour averaging time is appropriate when defining a target
level of protection against visibility impairment in terms of a
visibility index.
Moreover, the EPA disagrees with commenters that a secondary
PM2.5 standard with a 24-hour averaging time does not
provide requisite protection against the public welfare impacts of
visibility impairment. At the time of the 2012 review, the EPA
recognized that hourly or subdaily (i.e., 4- to 6-hour) averaging
times, within daylight hours and excluding hours with high relative
humidity, are more directly related to the short-term nature of
visibility impairment and the relevant viewing periods for segments of
the viewing public than a 24-hour averaging time. At the time of the
2012 review, the EPA agreed that a subdaily averaging time would
generally be preferable. However, the Agency noted significant data
quality uncertainties associated with the instruments that would
provide hourly PM2.5 mass concentrations necessary to inform
a subdaily averaging time. These uncertainties, as described in the
2012 review, included short-term variability in hourly data from
available continuous monitoring methods, which would prohibit
establishing a subdaily averaging time (78 FR 3209, January 15, 2013).
For all of these reasons, and consistent with the 2020 review, the EPA
continues to believe that a subdaily averaging time is not supported by
the information available in this reconsideration.
With regard to the form of the visibility index, some commenters
contend that the form used in evaluating visibility impairment is not
appropriate. First, commenters contend that the EPA incorrectly stated
that the CASAC did not provide advice on the 3-year, 90th percentile
form of the visibility index and that the CASAC specifically
recommended that the EPA further justify the metric and form, and by
not doing so, the proposal arbitrarily departs from the CASAC's
recommendations. The commenters also contend that the EPA fails to
explain how averaging the form over three years is protective given
that the public does not perceive visibility in three-year averages.
We disagree with the commenters that the EPA departed from the
CASAC's recommendations that ``[t]he final PA should provide a robust
justification for the daily light extinction percentile used in the
analysis'' (Sheppard, 2022a, p. 22 of consensus responses). In this
statement, the CASAC did not make explicit recommendations for
revisions to the form of the visibility index, as the commenters
assert, but rather requested additional justification for the
percentile selected for the visibility index in the 2022 PA. In
response to the CASAC's recommendation after reviewing the 2021 draft
PA, the EPA included a new section in the 2022 PA that explicitly
discusses the elements (i.e., indicator, averaging time, form, and
level) of the visibility index, including additional justification for
the conclusions regarding the appropriate elements for the index (U.S.
EPA, 2022b, pp. 5-27-5-29). In so doing, the 2022 PA recognizes that
there is no new information available in this reconsideration to inform
selection of an alternative form of the visibility index, and
therefore, relied on the analyses presented in the 2010 UFVA that
evaluated the different statistical forms of the visibility index. The
2022 PA also discusses the approach to improving visual air quality in
Federal Class I areas as a part of the Regional Haze Program (U.S. EPA,
2022b, p. 5-28). Furthermore, as reflected in responding to public
comments below, and in
[[Page 16332]]
reaching his final conclusions in section V.B.4 below, the
Administrator further considers the available scientific and
quantitative information, the CASAC's advice, and public comments in
informing his final conclusions regarding the appropriate target level
of protection for the visibility index. With regard to the commenters'
assertion that the EPA did not justify why averaging the form over
three years is protective, we agree with the commenters that people do
not perceive visibility impairment in three year averages. As described
in the 2022 PA, visibility-related effects and perceived impairment are
often associated with short-term PM concentrations, and therefore, the
focus of the visibility analyses is centered on the adequacy of the 24-
hour PM2.5 standard (U.S. EPA, 2022b, p. 5-29). However, as
described in the 2022 PA, the 3-year average form provides stability
from the occasional effect of inter-annual meteorological variability
that can result in unusually high pollution levels for a particular
year (U.S. EPA, 2022b, p. 5-28). Occasional meteorological variability
is of particular concern for the visibility index, which can be
impacted by not only PM concentrations in ambient air but also relative
humidity. The D.C. Circuit has previously recognized that it is
legitimate for the EPA to consider overall stability of the standard
and its resulting promotion of overall effectiveness of NAAQS control
programs in setting a standard. See American Trucking Ass'ns v.
Whitman, 283 F.3d 355, 375-76 (D.C. Cir. 2002). The 2022 PA concluded
that the available information continues to provide support for a 90th
percentile form, averaged over three years, and the inclusion of
additional justification for the elements of the visibility index
responds to the CASAC's recommendation (U.S. EPA, 2022b, section
5.3.1.2).
Some commenters suggest that the 90th percentile form is too low
and would result in 36 days being excluded annually, presuming that the
public only finds it objectionable when visibility is worse than the
standard on 37 or more days per year. The commenters also contend that
the EPA's approach of using a 90th percentile form for the visibility
index is inconsistent with the goals of the Regional Haze Program. In
so doing, the commenters note that the Regional Haze Rule focuses on
improving conditions on the worst days, while they argue that a 90th
percentile form for the visibility index would ignore the 36 worst
visibility days, rather than identifying them and reducing pollution on
those days.
In reaching conclusions regarding the appropriate form of the
visibility index, the EPA is following the same approach employed in
past reviews of the secondary PM NAAQS, including those in the 2012 and
2020 rulemakings. In reaching conclusions regarding the appropriate
form of the visibility index in the 2011 PA, the EPA considered the
percentile forms of the visibility index assessed in the 2010 PA (i.e.,
90th, 95th, 98th) along with the approach for improving visual air
quality under the Regional Haze Program. In so doing, the 2011 PA notes
that the Regional Haze Program targets the 20% most impaired days for
improvements in visual air quality in Federal Class I areas (i.e., the
days more impaired than the 80th percentile). The 2011 PA recognized
that to increase the likelihood of improving visual air quality on the
worst days, the form of the visibility index should be set well above
the 80th percentile. The 2011 PA further concluded that a 90th
percentile form would represent the median of the distribution of the
20% most impaired days, and meeting a visibility index with a 90th
percentile form would mean that 90% of the days have visual air quality
that is at or below the level of the visibility index and would
reasonably expected to lead to improvements in visual air quality for
the 20% most impaired days (U.S. EPA, 2011, p. 4-59). The 2022 PA noted
that there is no new information from public preference studies that
would inform the Administrator's consideration of the appropriate form
for the visibility target index, and reached conclusions consistent
with those of 2011 PA. However, as discussed below, the EPA disagrees
that a focus on the 90th percentile ``ignores'' any days with worse
visibility. It is possible to examine past patterns of air quality to
judge the relationship between the 90th percentile and higher
percentiles, and to assess whether achieving a 90th percentile
visibility target will also result in air quality improvements, where
necessary, at higher percentiles. Based on its assessment of past air
quality and potential alternative percentiles for the form, the EPA
judged that a 90th percentile would appropriately achieve improved air
quality both above and below that percentile.
Some commenters suggest that the analyses conducted in the 2010
UFVA are based on a different metric than the 24-hour average being
considered in the reconsideration, that the analyses are outdated and
irrelevant. Therefore, the commenters assert that relying on the
analyses in the 2010 UFVA is not a rational justification for the use
of a 90th percentile for the visibility index in this reconsideration.
Moreover, these commenters state that, in past reviews, both the EPA
and the CASAC have considered and recommended a 98th percentile form,
but the proposal does not consider the 98th percentile.
[[Page 16333]]
These commenters assert that the 2010 UFVA was not considering the
same metric under consideration here. However, the EPA was citing to
the 2010 UFVA for the conclusion that there are correlations between
different statistical forms of the visibility index. To confirm whether
these correlations occur under recent air quality, we conducted
additional air quality analyses evaluating the visibility index using
the current percentile form (i.e., 90th) and two alternative forms
(i.e., 95th and 98th).\172\ While a higher percentile form would
further limit the number of days with peak PM-related light extinction,
the analyses confirm that a 90th percentile form is effective in
limiting visibility impairment at higher percentiles. Based on these
analyses, depending on which version of the IMPROVE equation is used to
estimate light extinction, the differences in the 3-year averages of
estimated light extinction for the 90th, 95th, and 98th percentile
forms are small. For example, in areas that meet the current 24-hour
PM2.5 standard, for light extinction estimated using the
original IMPROVE equation, all sites have light extinction estimates
for a 90th percentile form at or below 26 dv, for a 95th or 98th
percentile form at or below 29 dv.\173\ In most locations, when
estimating light extinction based on the original IMPROVE equation, the
difference between a 95th or 98th percentile form and a 90th percentile
form is generally less than 3 dv.\174\ As noted in previous reviews, a
change of 1 to 2 dv in light extinction under many viewing conditions
will be perceived as a small, but noticeable, change in the appearance
of a scene, regardless of the initial amount of visibility impairment
(88 FR 5657, January 27, 2023; U.S. EPA, 2004b; U.S. EPA, 2010b). Thus,
differences between a 90th percentile form and a 95th or 98th
percentile form remain small, and for any of these forms of the
visibility index, the estimated light extinction based on the original
IMPROVE equation in areas meeting the current secondary 24-hour
PM2.5 standard is below the upper end of the range of the
levels considered for the visibility index (i.e., below 30 dv).
---------------------------------------------------------------------------
\172\ Gantt, B., and Hagan, N. (2023). Analysis of Percentile
Forms of the Visibility Index. Memorandum to the Rulemaking Docket
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
\173\ Gantt, B., and Hagan, N. (2023). Analysis of Percentile
Forms of the Visibility Index. Memorandum to the Rulemaking Docket
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
\174\ Gantt, B., and Hagan, N. (2023). Analysis of Percentile
Forms of the Visibility Index. Memorandum to the Rulemaking Docket
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
Some commenters disagree with the EPA's proposed conclusion that a
level of 30 dv is appropriate for the visibility index and support a
lower level in order to provide increased protection against visibility
impairment. Commenters who support a revised level for the visibility
index state that a target level of protection of 30 dv would mean that
less than 10% of participants in the public preference studies, other
than the Washington, DC, study, would accept visibility conditions
above 29 dv. These commenters further suggest that a 75% acceptability,
rather than 50% acceptability, is requisite to protect visibility
sources, which would be on average a level of 21 dv when using the
light extinction method or 18 dv when using the contrast of distance
method. These commenters argue that, based on the available
information, a target level of protection for the visibility index of
approximately 20 dv would be more appropriate, and therefore, the level
of the secondary 24-hour PM2.5 standard should be
strengthened to 25 [mu]g/m\3\. Other commenters who support a revised
level for the visibility index suggest that public preference studies
with longer sight paths to distant landscape features or with lower
target levels than those in the Washington, DC study, such as the
Phoenix study, would support a lower level. These commenters support
revising the target level of protection for the visibility index to a
25 dv, and revising the level of the secondary 24-hour PM2.5
standard to a level as low as 25 [mu]g/m\3\, suggesting that in low
relative humidity environments, 25 dv is consistent with
PM2.5 concentrations of less than 25 [mu]g/m\3\.
Some commenters state that EPA's justification for setting a target
level of protection at the upper end of the 20 to 30 dv range is
arbitrary. These commenters state that the EPA's reliance on the
standard operating in many regions and circumstances as support for the
upper end of the range is irrational and illegal. Moreover, these
commenters contend that EPA provided no rational connection between the
Regional Haze Program and the proposed decision to set the target level
of protection at the upper end of the range. They suggest that the EPA
proposed to rely exclusively on the Regional Haze Program to protect
visibility in Class I areas and to give visibility in these areas no
weight in considering the secondary PM standard and that it is not
rational to entirely ignore visibility in Class I areas when setting
the secondary standard. These commenters assert that the Regional Haze
Program provides no rational basis for a target level of protection at
the upper end of the range, nor does the EPA identify one.
Some commenters contend that the EPA failed to justify the adequacy
of the current secondary annual PM2.5 standard, noting that
the secondary 24-hour and annual PM2.5 standards work
together to provide protection against short- and long-term effects of
PM2.5. These commenters point to CASAC comments on the 2021
draft PA and the comments of an individual CASAC member's support for
strengthening the secondary annual PM2.5 standard to provide
increased protection against climate and materials effects over time.
They contend that EPA arbitrarily failed to discuss the secondary
annual PM2.5 standard not only in the proposal, but also in
the 2022 PA and in the 2020 final decision.
The EPA recognizes that the selection of the target level of
protection for the visibility index is fundamentally a public welfare
policy judgment for the Administrator. The Administrator is tasked by
the CAA to judge when visibility impairment becomes an adverse effect
on public welfare. It is clear that visibility impairment can become
adverse to public welfare, but the Administrator does not consider that
every deciview of impairment is adverse to public welfare. In
considering the point at which visibility impairment becomes adverse to
public welfare, such that the attainment of the secondary PM NAAQS
would prevent the adverse effect, the Administrator gives weight to the
public preference studies as to when visibility impairment is
unacceptable. At the same time, the Administrator recognizes the
limitations of these studies, which have been detailed in the proposal
and the 2022 PA. Similarly, the EPA discussed the Regional Haze program
in the proposal to highlight that there is a distinct program to
protect against visibility impairment in Class I areas, and the
existence of that program is relevant to the Administrator's judgment
about the level of visibility impairment that is adverse to public
welfare under CAA 109(d), because in determining what is requisite the
Administrator is primarily considering visibility impairment outside of
Class I areas.
[[Page 16334]]
In considering how to use the results of the public preference
studies, the Administrator concludes that a 50th acceptability
criterion is an appropriate tool. The Administrator's task is to set
standards that are neither more stringent nor less stringent than
necessary, and a 50% acceptability criterion seems most appropriate to
use in judging when visibility impairments become adverse, because it
should more closely represent when the median person would find the
impairment to be adverse. The Administrator notes this conclusion is
consistent with the approach adopted in the Denver study by Ely et al.
(1991) where the 50% acceptability criterion for urban visibility was
first presented. This study discussed the use of the 50% acceptability
criteria as a reasonable basis for setting a standard to protect
visibility in urban areas. In doing so, Ely et al. (1991) noted that
the 50% acceptability criterion divided the slides into two groups--
those judged acceptable and those judged unacceptable by a majority of
people in the study--and therefore, was reasonable since it defines the
point where the majority of the study participants began to judge
levels of visibility impairment as unacceptable (Ely et al., 1991).
In considering the appropriate target level of protection, we next
look to the available public preference studies, noting that the
selecting of the range of 20 to 30 dv for the target level of
protection for the visibility index is informed by the 50%
acceptability values from these studies. The Denver, CO, (Ely et al.,
1991) and British Columbia, Canada, (Pryor, 1996) studies met the 50%
acceptability criteria at 20 dv and 19-23 dv, respectively (U.S. EPA,
2022b, Table D-8). As described in the proposal, these studies used
photographs that were taken at different times of the day and on
different days to capture a range of light extinction levels needed for
the preference studies (88 FR 5652, January 27, 2023). Compared to
studies that used computer-generated images (i.e., those in Phoenix,
AZ, and Washington, DC) there was more variability in scene appearance
in these older studies that could affect preference rating and includes
uncertainties associated with using ambient measurements to represent
sight path-averaged light extinction values rather than superimposing a
computer-generated amount of haze onto the images. When using
photographs, the intrinsic appearance of the scene can change due to
meteorological conditions (i.e., shadow patterns and cloud conditions)
and spatial variations in ambient air quality that can result in
ambient light extinction measurement not being representative of the
sight-path-averaged light extinction. Computer-generated images, such
as those generated with WinHaze, do not introduce such uncertainties,
as the same base photograph is used (i.e., there is no intrinsic change
in scene appearance) and the modeled haze that is superimposed on the
photograph is determined based on uniform light extinction throughout
the scene. Because of the uncertainties and limitations associated with
the Denver, CO, and British Columbia, Canada, the EPA concludes that it
is appropriate to place less weight on these studies, and to instead
focus on the public preference studies that were designed to reduce
these uncertainties and limitations.
In so doing, we focus on the public preference studies that use
computer-generated images (i.e., those in the Phoenix, AZ, and
Washington, DC) studies. As described in the proposal, the use of
computer-generated images have less variability in scene appears than
in those studies that use photographs taken on different days and at
different times of the days (i.e., those in the Denver, CO, study) that
would be likely to influence preference rating and introduces
uncertainties associated with using ambient measurements to present
sight path-averaged light extinction values rather than superimposing a
computer-generated amount of haze onto the images (88 FR 5652, January
27, 2023).
The Phoenix, AZ, public preference study (BBC Research &
Consulting, 2003) had several strengths compared to some of the other
public preference studies. The Phoenix, AZ, study had the largest
number of participants (385 in 27 separate focus group sessions) of all
of the public preference studies, with a sample group designed to be
demographically representative of the Phoenix population at that time.
The age range in the Phoenix study was also more inclusive (18-65+),
with the distribution of the study participants corresponding
reasonably well to the overall age distribution in the 2000 U.S. Census
for the Phoenix area (BBC Research & Consulting, 2003). Furthermore,
the 21 images used in the Phoenix, AZ, study were developed using the
WinHaze software with visual air quality ranging from 15 to 35 dv, and
the view was toward the southwest, including downtown Phoenix, with the
Sierra Estrella Mountains in the background at a distance of 25 miles.
This study had the least noisy preference results, perhaps because a
larger, more representative group of participants combined with the use
of computer-generated images resulted in the smoother distribution of
responses of ``acceptable'' visual air quality. Based on the EPA's
evaluation of the public preference studies in the 2012 review, the 50%
``acceptable'' criteria was met at approximately 24 dv (U.S. EPA, 2010,
Table 2-3).
We also consider the public preferences for the Washington, DC,
studies (Abt Associates, 2001; Smith and Howell, 2009). The 2001
Washington, DC study included nine participants, and the 2009
Washington, DC, study replicated the 2001 study with 26 additional
participants. Similar to the Phoenix study, the Washington, DC, studies
also had the strength of having the 20 images included in the study
generated using WinHaze with visual air quality ranging from 9 to 45
dv. The study depicted a scene of a panoramic view of the Potomac
River, the National Mall, and downtown Washington, DC. All of the
distinct buildings in the scene were within four miles and the higher
elevations in the background were less than 10 miles from where the
image was taken from the Arlington National Cemetary in Virginia. The
50% ``acceptable'' criteria was met at approximately 29 dv (U.S. EPA,
2010, Table 2-3).
As described in more detail in the proposal, visibility preferences
can vary by location, and such differences may arise based on the
differences in the cityscape scene that is depicted in the images (88
FR 5652, January 27, 2023). In considering the geographical differences
between the public preference studies, we recognize that the
methodological differences between the studies may influence the
resulting ``acceptable'' level of visibility impairment. In the
Phoenix, AZ, study, the image depicted mountains in the background and
urban features in the foreground, whereas the Washington, DC, study
depicted nearby buildings in the image without mountains in the
distance. As an initial matter, we note that the object of interest to
the study participant could differ across the studies based on the
scenes included in the images being evaluated--with the mountains being
of greater interest in the images in the Phoenix, AZ, study, despite
also depicting buildings that are similar to those shown and presumed
to be of interest in the images in the Washington, DC, study (88 FR
5652, January 27, 2023). We also agree with the commenters that the
distance between the object of interest and the camera is an important
consideration in
[[Page 16335]]
evaluating the public preference studies. Objects at greater distances
from the camera location (such as those in the Phoenix, AZ, study which
had a maximum distance of 42 km (U.S. EPA, 2022b, Table D-8)) have a
greater sensitivity to light extinction, which alone could explain
differences in preferences but coupled with an object of greater
interest results in lower acceptable levels of visibility impairment.
Conversely, objects at closer distances from the camera location (such
as those in the Washington, DC, study which had a maximum distance of 8
km (U.S. EPA, 2022b, Table D-8)) have less sensitivity to light
extinction, which coupled with objects of interest (compared to the
mountainous views in the Phoenix, AZ, study) result in higher
acceptable levels of visibility impairment. These studies clearly
demonstrate that there are differences in the public preferences across
the studies depending on the images that are used, in particular the
object of interest to the study participant depicted in the image and
the distance of the sight path to the object, and that such differences
can influence preference results.
However, we note that these uncertainties and limitations have
persisted from past reviews, and there is very little new information
to inform conclusions regarding the interpretation of these results
with regard to the target level of protection. In selecting a target
level of protection, and in considering the CASAC's advice in their
review of the 2021 draft PA and public comments, we conclude that it is
appropriate to consider the information from the public preference
studies in Washington, DC, and Phoenix, AZ, and in so doing, that it is
appropriate to place weight on both of these studies in reaching
conclusions on the appropriate target level of protection. The EPA
recognizes that the scenes depicted in these two studies are different
and may influence public preferences of visibility impairment, but
notes these studies can be considered together as providing information
about different areas across the U.S. with variations in the scenes
that people are likely to most commonly encounter. The scene depicted
in the images used in the Washington, DC, study have a mix of
buildings, landmarks, and open space. On the other hand, the scene
depicted in the Phoenix, AZ, study included a mix of buildings in the
foreground and with more distant mountains in the background. The
Administrator considers it appropriate to consider these studies
together because in combination, they provide a greater diversity of
scenes, which is more likely to be representative of scenes people
typically experience around the country (e.g., not only in eastern
metropolitan statistical areas, but also in western areas with
different vistas). In considering these two studies together, the EPA
recognizes that, first, the ``object of interest'' is a subjective
judgment left to the participants of the public preference studies, and
second, the images in these two studies may differ in terms of
sensitivity to changes in light extinction because of the distance
between the object of interest in the scene and the camera. As noted by
the public commenters, the sight path for the images in the public
preference studies is an important consideration in reaching
conclusions regarding the appropriate target level of protection for
the visibility index. In addition, the Administrator judges that giving
weight to multiple studies is a more appropriate approach than focusing
on a single study, particularly where the study design (including the
representativeness of the participants and the scenes depicted in the
images) may be important for interpreting the results of the public
preference studies for informing conclusions regarding the visibility
index. Given these considerations and taking into consideration public
comments on the target level of protection for the visibility index,
the Administrator recognizes that it is more appropriate to consider a
broader range of public preferences, reflecting a broader range of
scenes, by putting significant weight on both the Washington, DC, and
Phoenix, AZ, studies. In so doing, he reaches the conclusion that it
would be appropriate to identify secondary PM standards that generally
limit visibility impairment to a level between the two studies.
The Administrator next considers what target level of protection
would be appropriate based on the available information from these
public preference studies. He first recognizes that, in the 2012 and
2020 final decisions, the then-Administrators selected a target level
of protection of 30 dv, based on the upper end of the range. In so
doing, the then-Administrators judged that it was appropriate to place
more weight on the uncertainties associated with the public preference
studies in reaching their conclusions. However, in this
reconsideration, the current Administrator, while continuing to
recognize that substantial uncertainties remain and that there is
relatively limited new information regarding public preferences of
visibility impairment, judges that it is important to balance the
weight placed on uncertainties with the strength of the scientific
evidence. As such, the Administrator concludes that it is appropriate
to consider a target level of protection within the range of 20 to 30
dv. He further concludes that in selecting a target level within that
range it is appropriate to place weight on both the mid-point of the
range, as supported by the study in Phoenix, AZ, as well as the upper
end, as supported by the Washington, DC, study. The Administrator notes
that these two studies both employ similar methodologies that are
subject to fewer uncertainties than older public preference studies
(including their use of WinHaze to reduce uncertainties in the
preference solicitations) although he notes that the Phoenix, AZ, study
yielded the best results of the four public preference studies in terms
of the least noisy preference results and the most representative
selection of participants. Furthermore, he notes the differences
between the scenes used for each study and finds that consideration of
these studies together is more appropriate in selecting a national
target for visibility protection than considering either study alone.
Thus, in considering this information, along with the uncertainties and
limitations of the public preference studies, the Administrator judges
that it would be appropriate to select a target level of protection
based on placing equal weight on the upper end of the range (i.e., 30
dv) and the middle of the range (i.e., 24 dv based on the Phoenix, AZ,
study) in order to identify a nationwide target for protection against
visibility impairment. In so doing, the Administrator concludes that a
visibility index with a target level of protection of 27, defined in
terms of estimated light extinction, with a 24-hour averaging time and
a 3-year, 90th percentile form, would provide adequate protection
against PM-related visibility effects on public welfare. Such a target
level of protection balances the information from two key studies
reflecting different participant preferences for different vistas in
different parts of the country, appropriately weighting both near-field
and more distant landscape features that may be of importance to public
perceptions of visibility.
The Administrator notes that the available evidence indicates that
the relationship between PM and light extinction is complex, depending
on factors such as PM composition, size fraction, and age of the
particles in ambient air, as well as relative
[[Page 16336]]
humidity. These factors can vary across the country based on
differences in regional influences, as well as meteorological
conditions that can vary spatially and temporally in different areas.
The Administrator also recognizes that this variability, coupled with
the age of the PM depending on the distance from the source to the
monitor location, also complicates the selection of which IMPROVE
equation is most appropriate in different areas, although he notes that
different IMPROVE equations will yield similar, but not identical,
results. In so doing, the Administrator takes note of the figures
presented in the 2022 PA, which depict the comparisons using the
original IMPROVE equation (Figure 5-3), the revised IMPROVE equation
(Figure 5-4), and the Lowenthal & Kumar equation (Figure 5-6), as well
as the estimated light extinction values for the three different
equations presented in Table D-7.
The Administrator notes that when light extinction is calculated
using the original IMPROVE equation, all 60 sites have 3-year
visibility metrics below 28 dv, 58 sites are at or below 25 dv, 26
sites are at or below 20 dv, and of the two sites above 25 dv one is at
26 dv and the other has a 24-hour PM2.5 design value of 56
[mu]g/m\3\ (i.e., well above the current 24-hour standard). Results are
similar for other IMPROVE equations.\175\ Based on these analyses, and
consistent with the results of similar analyses in the 2012 review and
the 2020 PA, the Administrator concludes that the current secondary 24-
hour PM2.5 standard, with its level of 35 [mu]g/m\3\,
maintains the visibility index below 27 dv, and in fact, the current
standard maintains air quality such that many areas have visibility
index values that range between 15 and 25 dv for all three IMPROVE
equations. In the areas that meet the secondary 24-hour
PM2.5 standard, all locations were below 27 dv when using
the original and revised IMPROVE equation and all but three locations
were at or below 27 dv when using the Lowenthal & Kumar IMPROVE
equation. Three locations (two in California and one in Utah) had air
quality that was at 28 dv when the Lowenthal & Kumar IMPROVE equation
was used. As described in more detail in section V.A.1.3, we recognize
that there are differences in the inputs for the three IMPROVE
equations that can influence the resulting estimated light extinction
values. The higher multiplier for converting OC to OM in the Lowenthal
& Kumar IMPROVE equation (i.e., a multiplier of 2.1) may be more
appropriate in more remote locations where there is more aged and
oxygenated organic PM than in urban locations. The three locations with
air quality at 28 dv are all in urban areas (downtown Los Angeles, CA;
Rubidoux, CA; Salt Lake City, UT) and tend to have higher levels of
nitrate and OC, especially during the wintertime when peak
PM2.5 concentrations typically occur. In these locations, it
may be more appropriate to use either the original or revised IMPROVE
equation, which have multipliers of 1.4 and 1.8, respectively, in order
to refine the inputs such that estimated light extinction in these
locations is more accurately characterized based on site-specific
characteristics.
---------------------------------------------------------------------------
\175\ When light extinction is calculated using the revised
IMPROVE equation, all 60 sites have 3-year visibility metrics below
28 dv, 56 sites are at or below 25 dv, and 26 sites are at or below
20 dv. When light extinction is calculated using the Lowenthal and
Kumar IMPROVE equation, 59 sites have 3-year visibility metrics
below 28 dv, 45 sites are at or below 25 dv, and 15 sites are at or
below 20 dv. The one site with a 3-year visibility metric of 32 dv
exceeds the secondary 24-hour PM2.5 standard, with a
design value of 56 [mu]g/m\3\ (see U.S. EPA, 2022b, Appendix D,
Table D-3).
---------------------------------------------------------------------------
We also note that the four areas that exceed the secondary 24-hour
PM2.5 standard also generally had air quality that was below
27 dv in terms of the visibility index, with only two locations
experiencing a visibility index above 27 dv. One location that exceeds
the secondary 24-hour PM2.5 standard had a visibility index
of 29 dv using the original IMPROVE equation, while two locations were
30 and 32 dv using the Lowenthal & Kumar IMPROVE equation. We believe
attainment and maintenance of the secondary 24-hour PM2.5
standard will result in improved air quality in these areas, such that
the visibility index values for these areas will decrease even further.
The Administrator recognizes that in concluding that it is
appropriate to identify secondary PM standards that generally limit
visibility impairment to as low as 27 dv in terms of the visibility
index, the current secondary PM standards continue to provide
protection against visibility impairment associated with a visibility
index as low as, or even lower than, 27 dv. In so doing, he notes that
when meeting the current 24-hour PM2.5 standard, all sites
have a visibility index at or below 27 dv with the original and revised
IMPROVE equations, and all but three sites at or below 27 dv with the
Lowenthal and Kumar IMPROVE equation. Furthermore, the Administrator
notes that this conclusion is consistent with the CASAC's advice who,
in their review the 2021 draft PA, stated that ``[i]f a value of 20-25
deciviews is deemed to be an appropriate visibility target level of
protection, then a secondary 24-hour PM2.5 standard in the
range of 25-35 [mu]g/m\3\ should be considered'' (Sheppard, 2022a, p.
21 of consensus responses).
Thus, the Administrator concludes that weight on both the upper end
of the range of target levels of protection for the visibility index
identified in previous reviews and the mid-point of the range, as
presented by the Phoenix, AZ, public preference study, and focusing on
a target level of protection of 27 dv, he still judges the current
secondary 24-hour PM2.5 standard requisite to achieve that
target because the standard generally maintains the visibility index at
or below 27 dv such that more stringent standards are not warranted.
The EPA agrees with the commenters that the secondary PM standards
work together to provide protection against short- and long-term
effects of both fine and coarse particles (U.S. EPA, 2022b, section
5.5; 88 FR 5661, January 27, 2023). However, the EPA disagrees with
commenters that we failed to discuss the secondary annual
PM2.5 standard in the proposal, 2022 PA, and the 2020 final
notice and that we failed to justify the adequacy of the secondary
annual PM2.5 standard. As described in the 2022 PA and the
proposal, we recognize that PM2.5 is the size fraction of PM
responsible for most of the visibility impairment in urban areas (U.S.
EPA, 2022b, section 5.3.1.2; 88 FR 5654, January 27, 2023). Analyses in
the 2019 ISA found that mass scattering from PM10-2.5 was
relatively small (less than 10%) in the eastern and northwestern U.S.,
whereas mass scattering was much larger in the Southwest (more than
20%), particularly in southern Arizona and New Mexico (U.S. EPA, 2019,
section 13.2.4.1, p. 13-36). Given the relationship between visibility
and PM2.5 along with the short-term nature of visibility
effects, we focus more on the adequacy of the secondary 24-hour
PM2.5 standard for providing protection against visibility
impairment (U.S. EPA, 2022b, section 5.3.1.2; 88 FR 5653, January 27,
2023). In reaching his proposed conclusions, the Administrator clearly
states that he ``recognizes that the current suite of secondary
standards (i.e., the 24-hour PM2.5, 24-hour PM10,
and annual PM2.5 standards) together provide . . . control
for both fine and coarse particulates and long- and short-term
visibility and non-visibility (e.g., climate and materials) effects
related to PM in ambient air'' (88 FR 5661, January 27, 2023). Thus, by
explaining how the secondary standards work together to provide
protection from adverse effects, why we focus on the secondary 24-hour
PM2.5 standard as
[[Page 16337]]
most relevant to visibility impairment, and how the Administrator
selected the target level of protection for the visibility index, we
have addressed the CASAC's request to support the proposed decision to
revise the secondary 24-hour PM2.5 standard while retaining
the secondary annual PM2.5 standard. The commenters also
cite to an individual CASAC member's comments for the review of the
2021 draft PA who stated ``[f]or the limited scope of this
reconsideration review, I see no reason to not simply set the Secondary
equal to the Primary PM Standards, whatever they may be'' (Sheppard,
2022a, p. A-3). This CASAC member did not provide a supporting
rationale for revising the secondary standards to levels equal to the
primary standards. Although areas across the country are required to
attain both the primary and secondary PM2.5 standards so air
quality is unaffected by the Administrator's decision not to revise the
secondary standards to be equal to the primary standards, as described
in responding to comments above, the CAA provisions require the
Administrator to establish secondary standards that, in the judgment of
the Administrator, are requisite to protect public welfare from known
or anticipated adverse effects associated with the presence of the
pollutant in ambient air. In so doing, the Administrator seeks to
establish standards that are neither more nor less stringent than
necessary for this purpose. The Act does not require that standards be
set at a zero-risk level, but rather at a level that reduces risk
sufficiently so as to protect the public welfare from known or
anticipated adverse effects. The final decision on the adequacy of the
current secondary standards is a public welfare policy judgment to be
made by the Administrator. In reaching his proposed and final decisions
regarding the adequacy of the current secondary PM standards, the
Administrator considered the available scientific information and
analyses about welfare effects, and associated public welfare
significance, as well as judgments about how to consider the range and
magnitude of uncertainties that are inherent in the scientific evidence
and analyses. In so doing, the Administrator concluded that the
currently available scientific evidence and quantitative analyses,
including uncertainties and limitations, do not call into question the
adequacy of the current secondary PM standards and that the current
secondary PM standards should be retained, without revision. The
Administrator's judgments and decisions on the primary and secondary
standards are independent and consider different aspects of the
available scientific evidence and information in reaching conclusions
regarding the adequacy of the standards in protecting against PM-
related health and welfare effects.
4. Administrator's Conclusions
This section summarizes the Administrator's considerations and
conclusions related to the current secondary PM2.5 and
PM10 standards and presents the rationale for his decision
that no change is required for those standards at this time. The CAA
provisions require the Administrator to establish secondary standards
that, in the judgment of the Administrator, are requisite to protect
public welfare from known or anticipated adverse effects associated
with the presence of the pollutant in the ambient air. In so doing, the
Administrator seeks to establish standards that are neither more nor
less stringent than necessary for this purpose. The Act does not
require that standards be set at a zero-risk level, but rather at a
level that reduces risk sufficiently so as to protect the public
welfare from known or anticipated adverse effects. The final decision
on the adequacy of the current secondary standards is a public welfare
policy judgment to be made by the Administrator. The decision should
draw on the scientific information and analyses about welfare effects,
and associated public welfare significance, as well as judgments about
how to consider the range and magnitude of uncertainties that are
inherent in the scientific evidence and analyses. This approach is
based on the recognition that the available evidence generally reflects
a continuum that includes ambient air exposures at which scientists
agree that effects are likely to occur through lower levels at which
the likelihood and magnitude of responses become increasingly
uncertain. This approach is consistent with the requirements of the
provisions of the Clean Air Act related to the review of NAAQS and with
how the EPA and the courts have historically interpreted the Act.
Given these requirements, the Administrator's final decision in
this reconsideration is a public welfare policy judgment that draws
upon the scientific and technical information examining PM-related
visibility impairment, climate effects and materials effects, including
how to consider the range and magnitude of uncertainties inherent in
that information. The Administrator recognizes that his final decision
is based on an interpretation of the scientific evidence and technical
analyses that neither overstates nor understates their strengths and
limitations, or the appropriate inferences to be drawn. In particular,
the Administrator notes that the assessment of when visibility
impairment is adverse to public welfare requires a public welfare
policy judgment informed by available scientific and quantitative
information.
In considering the adequacy of the current secondary PM standards
in this reconsideration, the Administrator has carefully considered
the: (1) Policy-relevant evidence and conclusions contained in the 2019
ISA and 2022 ISA Supplement; (2) the quantitative information presented
and assessed in the 2022 PA; (3) the evaluation of this evidence, the
quantitative information, and the rationale and conclusions presented
in the 2022 PA; (4) the advice and recommendations from the CASAC; and
(5) public comments. In the discussion below, the Administrator gives
weight to the 2022 PA conclusions, with which the CASAC generally
concurred during their review of the 2019 draft PA and 2021 draft PA,
as summarized in section IV.B.1 of the 2020 final notice and section
V.D.1 of the 2022 proposal, and takes note of key aspects of the
rationale for those conclusions that contribute to his decision in this
reconsideration. After giving careful consideration to all of this
information, the Administrator judges that no change is required for
the secondary PM standards at this time.
In considering the 2022 PA evaluations and conclusions, the
Administrator takes note of the overall conclusions that the non-
ecological welfare effects evidence and quantitative information are
generally consistent with what was considered in the 2020 final
decision and in the 2012 review (U.S. EPA, 2022b, section 5.5). The
scientific evidence for non-ecological welfare effects in this
reconsideration is largely the same as that available in the 2019 ISA
and 2020 PA. As described in section I.C.5.b above, the 2022 ISA
Supplement included a limited number of newly available studies on PM-
related visibility effects. This newly available evidence on visibility
effects, along with the full body of non-ecological welfare effects
evidence assessed in the 2019 ISA, reaffirms conclusions on the
visibility, climate, and materials effects recognized in the 2020 final
decision and in the 2012 review, including key conclusions on which the
standards are based. Further, as discussed in more
[[Page 16338]]
detail above, the updated quantitative analyses of visibility
impairment for areas meeting the current standards in the 2022 PA
support the adequacy of the current secondary PM standards to protect
against PM-related visibility impairment. The Administrator also
recognizes that uncertainties and limitations continue to be associated
with the available scientific evidence and quantitative information.
With regard to the current evidence on visibility effects, as
summarized in the 2022 PA and discussed in detail in the 2019 ISA and
ISA Supplement, the Administrator notes the long-standing body of
evidence for PM-related visibility impairment. As in previous reviews,
this evidence continues to demonstrate a causal relationship between PM
in ambient air and effects on visibility (U.S. EPA, 2019a, section
13.2). The Administrator recognizes that visibility impairment can have
implications for people's enjoyment of daily activities and for their
overall sense of well-being. Therefore, as in previous reviews, he
considers the degree to which the current secondary standards protect
against PM-related visibility impairment and the degree to which PM-
related visibility impairment is adverse to public welfare. In
particular, in recognizing the short-term nature of visibility
impairment along with the fact that PM2.5 is the size
fraction that contributes most to light extinction, the Administrator
especially focuses on the adequacy of the current secondary 24-hour
PM2.5 standard in providing protection against PM-related
visibility effects judged to be adverse. The Administrator also
considers the protection provided by the current secondary 24-hour
PM2.5 standard against PM-related visibility impairment in
conjunction with the Regional Haze Program as a means of achieving
appropriate levels of protection against PM-related visibility
impairment in urban, suburban, rural, and Federal Class I areas across
the U.S. Programs implemented to meet the secondary PM standards, along
with the requirements of the Regional Haze Program established for
protecting against visibility impairment in Class I areas, would be
expected to improve visual air quality across all areas of the country.
As described in the proposal (88 FR 5658, January 27, 2023), the
Administrator recognizes that the Regional Haze Program was established
by Congress specifically to achieve ``the prevention of any future, and
the remedying of existing, impairment of visibility in mandatory Class
I areas, which impairment results from man-made air pollution,'' and
that Congress established a long-term program to achieve that goal (CAA
section 169A). In adopting section 169, Congress set a goal of
eliminating anthropogenic visibility impairment at Class I areas, as
well as a framework for achieving that goal which extends well beyond
the planning process and timeframe for attaining the secondary PM
NAAQS. Recognizing that the Regional Haze Program will continue to
contribute to reductions in visibility impairment in Class I areas,
consistent with his proposed conclusions, the Administrator concludes
that addressing visibility impairment in Class I areas is largely
beyond the scope of the secondary PM standards and that setting the
secondary 24-hour PM2.5 standard at a level that would
remedy visibility impairment in Class I areas would result in standards
that are more stringent than is requisite.
In further considering what standards are requisite to protect
against adverse public welfare effects from visibility impairment, the
Administrator concludes that it is appropriate to use an approach
consistent with the approach used past reviews (88 FR 5650, January 27,
2023). He first identifies an appropriate target level of protection in
terms of a PM visibility index that takes into account the factors that
influence the relationship between PM in ambient air and visibility
(i.e., size fraction, species composition, and relative humidity). He
then considers the air quality analyses conducted in the 2022 PA that
examine the relationship between the PM visibility index and the
current secondary 24-hour PM2.5 standard in locations that
meet the current 24-hour PM2.5 and PM10 standards
(U.S. EPA, 2022b, section 5.3.1.2).
In reaching conclusions regarding the target level of protection,
the Administrator first considers the characteristics of the visibility
index and defines its elements (indicator, averaging time, form, and
level). With regard to the indicator for the visibility index, the
Administrator continues to recognize that, consistent with the
conclusions of the 2022 PA and the CASAC's advice in their review of
the 2021 draft PA, there is a lack of availability of methods and an
established network for directly measuring light extinction. Therefore,
the Administrator concludes that it continues to be appropriate to
using an index based on estimates of light extinction by
PM2.5 components based on the IMPROVE algorithm. In so
doing, the Administrator recognizes that the fundamental understanding
of the relationship between ambient PM and light extinction has
generally changed very little over time; however, several versions of
the IMPROVE equation have been developed and evaluated that could be
used to estimate light extinction. As at the time of the proposal, the
Administrator recognizes that the results of the quantitative analyses
in the 2022 PA that examined three versions of the IMPROVE equation
indicate that there are very small differences in estimates of light
extinction between the equations, and that it is not always clear that
one version of the IMPROVE equation is more appropriate for estimating
light extinction across the U.S. than other versions of the IMPROVE
algorithm (88 FR 5659, January 27, 2023). He also recognizes that the
selection of inputs to the IMPROVE equation (e.g., the multiplier for
OC to OM) may be more appropriate on a regional basis rather than a
national basis when calculating light extinction, and notes the CASAC's
advice that PM-visibility relationships are region specific (Sheppard,
2022a, p. 21 of consensus responses). The Administrator further notes
that neither the CASAC nor public commenters recommended a specific
IMPROVE equation or an approach for using different IMPROVE equations
across the U.S. Therefore, given the absence of a robust monitoring
network to directly measure light extinction, the Administrator
concludes that light estimated light extinction, as calculated using
one or more versions of the IMPROVE algorithms, continues to be the
most appropriate indicator for the visibility index.
Having reached the conclusion that estimated light extinction is
the appropriate indicator for the visibility index, the Administrator
next considers the appropriate averaging time and form of the index.
With regard to the averaging time and form, the Administrator notes
that in previous reviews, a 24-hour averaging time was selected and the
form was defined as the 3-year average of annual 90th percentile
values. As at the time of proposal, the Administrator recognizes that
the available information continues to provide support for the short-
term nature of visibility effects. He further recognizes that no new
information is available in this reconsideration to inform his
conclusions regarding averaging time, and therefore, he considers past
analyses of 24-hour and subdaily PM2.5 light extinction to
inform his conclusions on averaging time. As described in the proposal
(88 FR 5659, January 27, 2023) and in responding to comments in section
V.B.3 above, prior
[[Page 16339]]
analyses demonstrated that there are strong correlations between 24-
hour and subdaily (i.e., 4-hour average) PM2.5 light
extinction, indicating that a 24-hour averaging time is an appropriate
surrogate for the subdaily time periods associated with when
individuals experience visibility impairment and that a longer
averaging time may also be less influenced by atypical conditions and/
or atypical instrument performance. The Administrator also notes that
the CASAC did not provide advice or recommendations with regard to the
averaging time of the visibility index, although some public commenters
referenced CASAC advice in past reviews that a subdaily standard based
on daylight hours would better reflect the public welfare effects of
public perceptions of visibility impairment than a 24-hour standard.
However, in considering the available scientific and quantitative
information, as well as the CASAC's advice in their reviews of the 2019
draft PA and 2021 draft PA, the Administrator concludes that the 24-
hour averaging time continues to be appropriate for the visibility
index because it is an appropriate surrogate for subdaily time periods
and results in a more stable target.
With regard to the form of the visibility index, the Administrator
notes the approach in other NAAQS that a multi-year percentile form
offers greater stability to the air quality management process by
reducing the possibility that statistically unusual indicator values
will lead to transient violations of the standard. He recognizes that
using a 3-year average provides stability from the occasional effects
of inter-annual meteorological variability (including relative
humidity) that can result in unusually high pollution levels for a
particular year (88 FR 5659, January 27, 2023) and recognizes that a
stable standard contributes to the benefits of the NAAQS by ensuring
that attainment strategies are designed to address non-transient
problems and achieve durable air quality improvements. For these
reasons, he concludes that a 3-year average continues to be
appropriate.
In considering the percentile that would be appropriate with the 3-
year average, the Administrator recognizes that there is very little
new information available in this reconsideration to inform selection
of an alternative form of the visibility index and that the appropriate
form requires the exercise of public welfare policy judgment. In
selecting the appropriate target level of protection for the visibility
index, the Administrator is required to assess when visibility
impairment becomes adverse to public welfare, weighing both the degree
of visibility impairment (in dv) and the frequency of such impairment
(through the form). As with the mass-based PM air quality standard, the
target level of protection for the visibility index must be selected in
conjunction with the form to determine the appropriate stringency. In
so doing, consistent with approaches in past reviews, the Administrator
first notes that the Regional Haze Program targets the 20% most
impaired days for improvements in visual air quality in Class I areas,
which are the days above the 80th percentile form of the visibility
index. The Administrator concludes that a percentile form set at the
80th percentile would not be likely to sufficiently improve visual air
quality on the worst days based on the visibility index. In considering
the information available in past reviews regarding the form of the
visibility index, as well as the analysis of alternative forms based on
recent air quality discussed above, the Administrator notes that a 90th
percentile form would represent the median of the distribution of the
20% most impaired days, and meeting a visibility index with a 90th
percentile form would reasonably be expected to lead to improvements in
visual air quality for days both above and below the 90th percentile
(88 FR 5660, January 27, 2023). In reaching his conclusion that a 90th
percentile would appropriately achieve improved air quality both above
and below that percentile, the Administrator took into consideration
assessments of air quality data and potential alternative percentiles
for the form. The Administrator further notes that, consistent with the
conclusions in the 2011 PA and 2020 PA, the 2022 PA concluded that
there is no new information from public preference studies that would
suggest that a 90th percentile form is not appropriate. The
Administrator also considers air quality analyses described above in
responding to public comments regarding the percentile form of the
visibility index. In particular, the Administrator notes that while a
higher percentile form (i.e., 95th or 98th) would somewhat further
limit the number of days with peak PM-related light extinction, the
differences in the 3-year averages of estimated light extinction for
the 90th, 95th, and 98th percentile forms are small. For example, he
notes that for the original IMPROVE equation, in areas that meet the
current 24-hour PM2.5 standard, all sites have light
extinction estimates for a 90th percentile form at or below 26 dv, and
for a 95th or 98th percentile form light extinction estimates are at or
below 29 dv.\176\ He further notes that, in most locations when
estimating light extinction based on the original IMPROVE equation, the
difference between a 95th or 98th percentile form and a 90th percentile
form is generally less than 3 dv.\177\ Moreover, the Administrator
concludes that a 90th percentile form achieves a very high degree of
control but appropriately targets the group of worst days, rather than
the few very worst days. Based on the available information and these
analyses, the Administrator concludes that the information does not
indicate that it would be appropriate to consider limiting the
occurrence of days with peak PM-related light extinction to a greater
degree, nor did the CASAC provide advice or recommendations related to
the form of the visibility index. Therefore, the Administrator judges
that it remains appropriate to define a visibility index in terms of a
24-hour averaging time and form based on the 3-year average of annual
90th percentile values.
---------------------------------------------------------------------------
\176\ Gantt, B., and Hagan, N. (2023). Analysis of Percentile
Forms of the Visibility Index. Memorandum to the Rulemaking Docket
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
\177\ Gantt, B., and Hagan, N. (2023). Analysis of Percentile
Forms of the Visibility Index. Memorandum to the Rulemaking Docket
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
With regard to the level of the visibility index, as at the time of
proposal, the Administrator continues to recognize that there is very
little new information available to inform his judgment regarding the
range of levels of visibility impairment judged to be acceptable by at
least 50% of study participants in the visibility preference
studies,\178\ and therefore, the range of 20 to 30 dv identified in the
2022 PA remains appropriate for considering the level of the visibility
index. The Administrator also recognizes that the uncertainties and
limitations associated with the public preferences identified in the
2012 and 2020 reviews continue to persist, and that these limitations
and uncertainties contributed to the decisions in 2012 and 2020 that a
level at the upper end of the range (i.e., 30 dv) was selected. The
Administrator specifically notes that, while the studies
[[Page 16340]]
are methodologically similar, there are a number of factors that can
influence comparability across the studies and that the available
studies may not capture the full range of visibility preferences in the
U.S. population, as described in more detail in section V.D.3 of the
2022 proposal (88 FR 5659-5660, January 27, 2023). The Administrator
also notes the CASAC's advice in their review of the 2021 draft PA that
there are a limited number of visibility preference studies available
to inform the Administrator's judgment regarding the appropriate target
level of protection for the visibility index (Sheppard, 2022a, p. 21 of
consensus responses). In considering the available information,
including uncertainties and limitation, and the CASAC's advice, the
Administrator proposed to conclude that it is appropriate to consider a
target level of protection for the visibility index within the range of
20 to 30 dv, and that establishing a target level of protection at the
upper end of the range was appropriate. In so doing, the Administrator
proposed to conclude that the protection provided by a visibility index
based on estimated light extinction, a 24-hour averaging time, and a
90th percentile form, averaged over 3 years, set to a level of 30 dv
would be requisite to protect public welfare with regard to visibility
impairment.
---------------------------------------------------------------------------
\178\ For reasons stated above and described in the 2022 PA and
proposal, the Administrator does not find it appropriate to use the
most recent preference study based on the Grand Canyon study area
(Malm et al., 2019) for purposes of identifying a target level of
protection for the visibility index.
---------------------------------------------------------------------------
However, at the time of proposal, the Administrator recognized that
the available evidence on visibility impairment generally reflects a
continuum and that the public preference studies do not provide
information about the specific level for which visibility impairment
would be ``acceptable'' or ``unacceptable'' across the country, and
that alternative target levels of protection could be supported. At
that time, in soliciting public comments, the Administrator recognized
that other interpretations, assessments, and judgments based on the
available welfare effects evidence for this reconsideration could be
possible (88 FR 5662, January 27, 2023).
With regard to the appropriate target level of protection for the
visibility index, the Administrator first notes that while the public
preference studies were conducted in several geographical areas across
the U.S., and they provide insight into regional preferences for
visibility impairment, none of the studies identify a specific level of
visibility impairment that would be perceived as ``acceptable'' or
``unacceptable'' across the whole U.S. population. He also noted that
there have been significant questions about how to set a standard for
visibility that is neither overprotective nor underprotective for some
areas of the U.S. As described in the proposal (88 FR 5660, January 27,
2023), in establishing the Regional Haze Program to improve visibility
in Class I areas, Congress noted that ``as a matter of equity, the
national ambient air quality standards cannot be revised to adequately
protect visibility in all areas of the country.'' H.R. Rep. 95-294 at
205. For the reasons noted above, in reaching his proposed decision
regarding visibility impairment, the Administrator recognized that he
is not seeking to set a standard that would eliminate visibility
impairment in Class I areas, but significant uncertainties remain
regarding how to judge when visibility impairment becomes adverse to
public welfare across the range of daily outdoor activities for
Americans across the country.
In reaching final conclusions regarding the available information,
along with the CASAC's advice and public comments, the Administrator
again considers what constitutes an appropriate target level of
protection, and in particular considers whether a target level of
protection below 30 dv is warranted. In so doing, he first notes the
variability in public preferences of visibility impairment as
demonstrated by the available public preferences, which support a range
of potential target levels of protection for the visibility index from
20 to 30 dv. He also notes that this range informed the 2012 and 2020
then-Administrators final decisions that a target level of protection
at the upper end of the range (i.e., 30 dv) would be most appropriate,
given the uncertainties and limitations associated with the public
preference studies. As described in in section V.B.3 above in
responding to public comments, the Administrator recognizes that a
number of factors can influence public preferences across studies, in
particular due to the types of scenes depicted in the images as well as
the distances at which the objects of interest are located from the
camera. Furthermore, the Administrator recognizes the small number of
public preference studies currently available makes precise
interpretations of their results challenging for determining a
nationally appropriate target level of visibility protection. The
Administrator also recognizes that the CASAC, in their review of 2021
draft PA, reiterated that PM-visibility relationships are region-
specific based on aerosol composition, and that several public
commenters emphasized the importance of the sight path distance in the
images when considering how to interpret the public preference studies.
In this reconsideration, the Administrator judges that in
determining when visibility impairment becomes adverse to public
welfare for purposes of the secondary NAAQS, while continuing to
recognize that substantial uncertainties remain and that there is
relatively limited new information regarding public preferences of
visibility impairment, it is important to balance the weight placed on
uncertainties with the strength of the scientific evidence. In so
doing, the Administrator first concludes that, consistent with previous
reviews and his proposed decision, it remains appropriate to consider a
target level of protection within the range of 20 to 30 dv. However, in
further considering the available scientific and quantitative
information, CASAC advice, and public comments, he further concludes
that in selecting a target level within that range it is appropriate to
place weight on both the middle of the range, as supported by the study
in Phoenix, AZ, as well as the upper end, as supported by the
Washington, DC, study. In so doing, he notes that the Washington, DC,
and Phoenix, AZ, studies employ similar methodologies that are subject
to fewer uncertainties than older public preference studies (including
their use of WinHaze to reduce uncertainties in the preference
solicitations) although he does note that the Phoenix, AZ, study
yielded the best results of the four public preference studies in terms
of the least noisy preference results and the most representative
selection of participants. Further, the Administrator judges that this
approach would take into account scenes that are similar to both the
Washington, DC, study and Phoenix, AZ, study, which would be more
representative of the ``typical'' scenes encountered across more areas
of the U.S. than an approach that places weight on just one study or on
studies conducted in certain geographical areas of the country. In
considering this information, along with the uncertainties and
limitations of the public preference studies, the Administrator judges
that it would be appropriate to select a target level of protection
based on placing equal weight on the upper end of the range (i.e., 30
dv) and the middle of the range (i.e., 24 dv based on the Phoenix, AZ,
study) in order to provide protection against visibility impairment in
different geographical areas of the U.S. For these reasons, the
Administrator concludes that a visibility index with a target level of
protection of 27 dv,
[[Page 16341]]
defined in terms of estimated light extinction, with a 24-hour
averaging time and a 3-year, 90th percentile form, would provide
adequate protection against PM-related visibility effects. In reaching
this conclusion, the Administrator judges that such a target level of
protection balances the information from these two key public
preference studies in such a way appropriately weighs both near-field
and more distant landscape features that may be of importance to public
perceptions of visibility.
In further considering the appropriate target level of protection
for the visibility index, the Administrator again recognizes the
complexity of the relationship between PM and light extinction which is
dependent on a number of factors, including PM composition, size
fraction, and age of the particles in ambient air, as well as relative
humidity. As noted in responding to comments above, these factors can
vary geographically across the U.S. and local or regional
meteorological conditions can also vary spatially and temporally. These
factors are critical inputs to the IMPROVE equation and can influence
the resulting estimated light extinction such that it is not a
straightforward comparison between estimated light extinction in one
area of the country versus another. Moreover, the Administrator
recognizes that there is variability in estimated light extinction
depending on the version of the IMPROVE equation that is used. As
described in more detail in the 2022 PA and the proposal, and in
reaching his decisions on the indicator of the visibility index above,
the Administrator notes that the 2022 PA concluded that one version of
the IMPROVE equation is not more accurate or precise in estimating
light extinction, and that difference in locations may support the
selection of inputs into the IMPROVE equation or of the appropriate
IMPROVE equation to estimate light extinction on a regional basis
rather than on a national basis.
In considering the available information, including variations in
both public preferences of visibility impairment and estimates of light
extinction using one or more IMPROVE equation, as well as the CASAC's
advice in their review of the 2019 draft PA and 2021 draft PA and
public comments, the Administrator judges that a target level of
protection of 27 dv would be appropriate. In so doing, he concludes
that a target level of protection above 27 dv would not provide
adequate protection against PM-related visibility impairment based on
the 50% acceptability values when both the Washington, DC, and Phoenix,
AZ, studies are considered. However, he also notes that when
considering the 50% acceptability values from studies conducted in
different areas of the U.S. and with different scenes and images
depicted, the available public preference studies do not provide a
``bright line'' at and above which visibility impairment is considered
adverse to public welfare. He further recognizes that, as discussed
just above, there are a number of region-specific factors that can
influence light extinction, and thereby influence visibility
impairment, as well as variations in public preferences of visibility
impairment based on the available studies, that complicate selection of
a single target level of protection that would be appropriate for a
national visibility index. While the Administrator recognizes that the
uncertainties and limitations associated with public preferences of
visibility and estimating light extinction have persisted over the last
several PM NAAQS reviews, he also recognizes that in reaching
conclusions regarding the appropriate target level of protection for
the visibility index also involves public welfare policy judgments
regarding how to appropriately consider the particular uncertainties
around identifying when visibility impairment becomes adverse to public
welfare, and the limitations on relying on the public preference
studies.
The Administrator also places weight on the high degree of spatial
and temporal variability in PM composition and relative humidity across
the U.S. in considering a target level of protection. This approach of
establishing a target level of protection that takes into account 50%
acceptability values from both eastern and western sites is a more
appropriate basis for determining the requisite level of protection
against known or anticipated adverse effects on public welfare across
diverse locations, i.e., a standard that is neither more nor less
stringent than necessary nationwide. Specifically, the Administrator
judges that a target level of protection for the visibility index
focused on maintaining estimated light extinction between the upper end
of the range of the target levels of protection (i.e., 30 dv based on
the Washington, DC, study) and the middle of the range (i.e., 24 dv
based on the Phoenix, AZ, study) to be more appropriate for a
nationwide standard to protect against visibility impairment compared
to a value derived from one location or one type of scene alone. For
these reasons, in selecting a target level of protection, the
Administrator concludes that a target level of protection somewhere
between the upper end and middle of the range is appropriate because he
judges that this approach, in conjunction with the Regional Haze
program, is sufficient, but not more stringent than necessary, to
protect against adverse effects on public welfare. Thus, he concludes a
secondary 24-hour PM2.5 NAAQS should be evaluated based on
its ability to provide protection against visibility impairment
associated with estimated light extinction of 27 dv based on estimated
light extinction, a 24-hour averaging time, and a 90th percentile form,
averaged over 3 years.
Having concluded that it is appropriate to identify a target level
of protection in terms of a visibility index based on estimated light
extinction as described above, the Administrator next considers the
degree of protection from visibility impairment afforded by the current
secondary PM standards. He considers the updated analyses of PM-related
visibility impairment presented in the 2022 PA (U.S. EPA, 2022b,
section 5.3.1.2) and described in section V.B.1.a of the proposal, and
notes that the results of the analyses are consistent with the results
from the 2012 and 2020 reviews.
Taking into consideration the full body of scientific evidence and
technical information concerning the known and anticipated effects of
PM on visibility impairment, the Administrator concludes that the
current secondary PM2.5 and PM10 standards are
requisite to protect against PM-related visibility impairment. While
the inclusion of the coarse fraction had a relatively modest impact on
calculated light extinction in the analyses presented in the 2022 PA,
he recognizes the continued importance of the PM10 standard
given the potential for larger impacts in locations with higher coarse
particle concentrations, such as in the southwestern U.S., for which
only a few sites met the criteria for inclusion in the analyses in the
2022 PA (U.S. EPA, 2019a, section 13.2.4.1; U.S. EPA, 2022b, section
5.3.1.2).
With regard to the adequacy of the secondary 24-hour
PM2.5 standard, the Administrator notes that, in their
review of the 2021 draft PA, the CASAC stated that ``[i]f a value of
20-25 deciviews is deemed to be an appropriate visibility target level
of protection, then a secondary 24-hour PM2.5 standard in
the range of 25-35 [micro]g/m\3\ should be considered'' (Sheppard,
2022a, p. 21 of consensus responses). The Administrator recognizes that
the CASAC recommended that the Administrator provide additional
justification for a visibility index target
[[Page 16342]]
of 30 dv but did not specifically recommend that he choose an
alternative level for the visibility index. The Administrator carefully
considered the advice of CASAC and the public comments and concluded
that a lower target level of visibility was appropriate in order to
properly reflect both a broader set of studies and a broader range of
vistas that were the subject of those studies. However, in their review
of the 2021 draft PA, the CASAC recognized that even a visibility index
target in the range of 20-25 dv could still warrant retention of the
current secondary 24-hour PM2.5 standard. The Administrator
also considers the advice from the CASAC in their review of the 2019
draft PA, who ``recogniz[ed] that uncertainties. . .remain about the
best way to evaluate'' PM-related visibility effects (Cox, 2019b, p. 13
consensus responses). The Administrator considered the CASAC's advice,
together with the available scientific evidence and quantitative
information, in reaching his conclusions.
The Administrator recognizes that conclusions regarding the
appropriate weight to place on the scientific and technical information
examining PM-related visibility impairment, including how to consider
the range and magnitude of uncertainties inherent in that information,
is a public welfare policy judgment left to the Administrator. In
reaching his final decision in 2020, the then-Administrator noted that
the available evidence regarding visibility effects had changed very
little since the 2012 review, specifically recognizing that, as
evaluated in the 2019 ISA, there were no new visibility studies that
were conducted in the U.S. and there was little new information
available with regard to acceptable levels of visibility impairment in
the U.S. (85 FR 82742, December 18, 2020). As such, the then-
Administrator concluded that the protection provided by a standard
defined in terms of a PM2.5 visibility index, with a 24-hour
averaging time, a 90th percentiles form averaged over three years, set
at a level of 30 dv, was requisite to protect public welfare against
visibility impairment (85 FR 82743, December 18, 2020). He also
recognized that there was some new information to inform quantitative
analyses of light extinction, but that the results of the analyses
conducted in the 2020 PA were consistent with those from the 2012
review. The then-Administrator recognized that the analyses
demonstrated that the 3-year visibility metric was at or below about 30
dv in all areas that met the current secondary 24-hour PM2.5
standard, and was below 25 dv in most of those areas (85 FR 82743,
December 18, 2020). Therefore, the Administrator judged that the
secondary 24-hour PM2.5 standard provided sufficient
protection for visual air quality of 30 dv, which he judged appropriate
(88 FR 82744, December 18, 2020). In this reconsideration, the ISA
Supplement evaluated newly available studies on public preferences for
visibility impairment and/or development methodologies or conducted
quantitative analyses of light extinction. In considering the available
scientific and quantitative information, including that newly available
in this reconsideration, the current Administrator reached the same
preliminary conclusions in the notice of proposed rulemaking regarding
the 3-year visibility index and the current secondary PM standards as
the then-Administrator in the 2020 final decision. However, in light of
public comments on the proposal, the Administrator has further
considered the available scientific evidence and information, as well
as the CASAC's advice regarding visibility effects in their review of
the 2021 draft PA. In so doing, the Administrator judges that it is
appropriate to place more weight on certain aspects of the evidence
that he had placed less weight on in reaching his proposed conclusions
(i.e., he focused on the both the middle and the upper end of the range
of the 50% acceptability values from the available public preference
studies). As such, the Administrator notes his conclusion on the
appropriate visibility index (i.e., with a 24-hour averaging time; a 3-
year, 90th percentile form; and a level of 27 dv), which takes into
account the regional variations in public preferences and equations for
estimating light extinction, and his conclusions regarding the
quantitative analyses of the relationship between the visibility index
and the current secondary 24-hour PM2.5 standard. In so
doing, the Administrator concludes that the current secondary standards
provide requisite protection against PM-related visibility effects.
With respect to climate effects, as at the time of proposal, the
Administrator recognizes that a number of improvements and refinements
have been made to climate models since the time of the 2012 review.
However, despite continuing research and the strong evidence supporting
a causal relationship with climate effects (U.S. EPA, 2019a, section
13.3.9), the Administrator notes that there are still significant
limitations in quantifying the contributions of the direct and indirect
effects of PM and PM components on climate forcing (U.S. EPA, 2022b,
sections 5.3.2.1.1 and 5.5). He also recognizes that models continue to
exhibit considerable variability in estimates of PM-related climate
impacts at regional scales (e.g., ~100 km), compared to simulations at
the global scale (U.S. EPA, 2022b, sections 5.3.2.1.1 and 5.5).
Moreover, the effects of PM on climate are diverse as well as
uncertain. Depending on the circumstances, the radiative forcing
effects of PM in the atmosphere can vary, such that positive forcing
could result in warming of the Earth's surface, whereas a negative
forcing could result in cooling (U.S. EPA, 2019a, section 13.3.2.2).
The resulting uncertainty leads the Administrator to conclude that the
scientific information available in this reconsideration remains
insufficient to quantify, with confidence, the impacts of ambient PM on
climate in the U.S. (U.S. EPA, 2022b, section 5.3.2.2.1) and that there
is not an adequate scientific basis to link attainment of any
particular PM concentration in ambient air in the U.S. to specific
climate effects. Consequently, the Administrator judges that there is
insufficient information at this time to revise the current secondary
PM standards or to promulgate a distinct secondary standard to address
PM-related climate effects.
With respect to materials effects, the Administrator notes that the
available evidence continues to support the conclusion that there is a
causal relationship with PM deposition (U.S. EPA, 2019a, section 13.4).
He recognizes that deposition of particles in the fine or coarse
fractions can result in physical damage and/or impaired aesthetic
qualities. Particles can contribute to materials damage by adding to
the effects of natural weathering processes and by promoting the
corrosion of metals, the degradation of painted surfaces, the
deterioration of building materials, and the weakening of material
components. While some recent evidence on materials effects of PM is
available in the 2019 ISA, the Administrator notes that this evidence
is primarily from studies conducted outside of the U.S. in areas where
PM concentrations in ambient air are higher than those observed in the
U.S. (U.S. EPA, 2019a, section 13.4). Given the limited amount of
information on the quantitative relationships between PM and materials
effects in the U.S., and uncertainties in the degree to which those
effects could be adverse to the public welfare, the Administrator
judges that the available scientific information
[[Page 16343]]
remains insufficient to quantify, with confidence, the public welfare
impacts of ambient PM on materials and that there is insufficient
information at this time to revise the current secondary PM standards
or to promulgate a distinct secondary standard to address PM-related
materials effects.
Taken together, the Administrator concludes that the scientific and
quantitative information for PM-related non-ecological welfare effects
(i.e., visibility, climate, and materials),\179\ along with the
uncertainties and limitations, supports the current level of protection
provided by the secondary PM standards as being requisite to protect
against known and anticipated adverse effects on public welfare. For
visibility impairment, this conclusion reflects his consideration of
the evidence for PM-related light extinction, together with his
consideration of updated air quality analyses of the relationship
between the visibility index and the current secondary 24-hour
PM2.5 standard and the protection provided by the current
secondary PM2.5 and PM10 standards. For climate
and materials effects, this conclusion reflects his judgment that,
although it remains important to maintain secondary PM2.5
and PM10 standards to provide some degree of control over
long- and short-term concentrations of both fine and coarse particles,
it is appropriate not to change the existing secondary standards at
this time and that it is not appropriate to establish any distinct
secondary PM standards to address PM-related climate and materials
effects at this time. As such, the Administrator recognizes that
current suite of secondary standards (i.e., the 24-hour
PM2.5, 24-hour PM10, and annual PM2.5
standards) together provide such control for both fine and coarse
particles and long- and short-term visibility and non-visibility (e.g.,
climate and materials) effects related to PM in ambient air. His
conclusions on the secondary standards are consistent with advice from
the CASAC, which noted substantial uncertainties remain in the
scientific evidence for climate and materials effects, as well as the
majority of public comments on the secondary PM standards. Thus, based
on his consideration of the evidence and analyses for PM-related
welfare effects, as described above, and his consideration of CASAC
advice and public comments on the secondary standards, the
Administrator concludes that it is appropriate not to change those
standards (i.e., the current 24-hour and annual PM2.5
standards, 24-hour PM10 standard) at this time.
---------------------------------------------------------------------------
\179\ As noted earlier, other welfare effects of PM, such as
ecological effects, are being considered in the separate, on-going
review of the secondary NAAQS for oxides of nitrogen, oxides of
sulfur and PM.
---------------------------------------------------------------------------
C. Decision on the Secondary PM Standards
For the reasons discussed above and taking into account information
and assessments presented in the 2019 ISA, ISA Supplement, and 2022 PA,
advice from the CASAC, and consideration of public comments, the
Administrator concludes that the current secondary PM standards are
requisite to protect public welfare from known or anticipated adverse
effects and is not changing the standards at this time.
VI. Interpretation of the NAAQS for PM
The EPA is finalizing revisions on data calculations in appendix K
for PM10 and appendix N for PM2.5. Revisions to
appendix K make the PM10 data handling procedures for the
24-hour PM10 standards more consistent with those of other
NAAQS pollutants and codify existing practices. Revisions to appendix N
update references to the revision(s) of the standards and change data
handling provisions related to combining data from nearby monitoring
sites to codify existing practices that are currently being implemented
as the EPA standard operating procedures.
A. Amendments to Appendix K: Interpretation of the NAAQS for
Particulate Matter
The EPA proposed to modify its data handling procedures for the 24-
hour PM10 standard in appendix K to part 50 (88 FR 5662,
January 27, 2023). The proposed modifications include: (1) Revising
design value calculations to be on a site-level basis, (2) codifying
site combinations to maintain a continuous data record, and (3)
clarifying daily validity requirements for continuous monitors. The
purpose of these modifications is to make the data handling procedures
for the 24-hour PM10 standard more consistent with those of
other NAAQS pollutants and codify existing practices that are currently
being implemented as EPA standard operating procedures.
The EPA received few comments on these proposed appendix K
revisions, the majority of which were supportive.
One commenter was not supportive of the proposed appendix K
revision to site-level PM10 design values, asserting that it
would amount to an imposition of a more stringent PM10
standard due to the potential high bias of FEMs. The EPA disagrees with
this assertion because site-level design values would combine data from
any high biased FEM with other monitors at the site rather than
calculate a monitor-level design value with data solely from that high-
biased FEM. The EPA tested the impact of calculating site-level
PM10 design values for the 2019-2021 period by assigning the
lowest parameter occurrence code as the primary monitor and calculating
site-level design values. Most resulting site-level design values were
either identical to or in-between the multiple monitor-level design
values at the site. Combining data from two or more monitors also has
the benefit of increasing the number of valid sample days at many
sites. For the 2019-2021 test period, approximately 10% of the sites
with more than one monitor went from having multiple invalid design
values to a single valid design value.
One commenter was not supportive of a footnote in the preamble of
the NPRM stating that in the absence of a designated primary monitor at
a given site, the default primary monitor would be one with the most
complete data record (88 FR 5662, January 27, 2023). Because the
procedure for calculating PM10 design values on a site-level
basis being finalized here will require monitoring agencies to
designate a primary monitor for each site in their annual network plans
(88 FR 5694, January 27, 2023; App. K, 1.0(b)), the EPA agrees with the
commenter that this footnote was unnecessary.
Therefore, the EPA is finalizing these appendix K revisions as
proposed.
B. Amendments to Appendix N: Interpretation of the NAAQS for PM2.5
The EPA proposed to modify its data handling procedures for the
annual and 24-hour PM2.5 standards in appendix N to part 50
(88 FR 5663, January 27, 2023). These proposed revisions include: (1)
Updating references to the revisions of the standards rather than
stating the specific level, and (2) codifying site combinations to
maintain a continuous data record. The purpose of both modifications is
to codify existing practices that are currently being implemented as
the EPA standard operating procedures.
The EPA received few comments on these revisions in the proposed
rule, with most supportive of the appendix N revisions.
Although the EPA did not propose or request comment on this issue,
one commenter suggested that appendix N be revised to only allow data
from the primary monitor to be used in PM2.5 NAAQS
designations asserting that it would add flexibility. The EPA disagrees
with the commenter's assertion that this would add flexibility because
it could force agencies to run
[[Page 16344]]
their FRMs on a daily schedule or potentially lead to invalid design
values if manual sampling interruptions or laboratory issues impact FRM
data completeness. This change would also be undesirable because it
could reduce by two-thirds the number of days used in calculations for
the annual and 24-hour PM2.5 design values at many sites.
Therefore, the EPA is finalizing these appendix N revisions as
proposed.
VII. Amendments to Ambient Monitoring and Quality Assurance
Requirements
The EPA is finalizing revisions to ambient air monitoring
requirements for PM to improve the usefulness of and appropriateness of
data used in regulatory decision making. These changes focus on ambient
monitoring requirements found in 40 CFR parts 50 (appendix L), 53, and
58 with associated appendices (A, B, C, D, and E). These changes
include addressing updates in the approval of reference and equivalent
methods, updates in quality assurance statistical calculations to
account for lower concentration measurements, updates to support
improvements in PM methods, a revision to the PM2.5 network
design to account for at-risk populations, and updates to the Probe and
Monitoring Path Siting Criteria for NAAQS pollutants. The EPA also took
comment on how to incorporate data from next generation technologies
into Agency efforts. A summary of the comments received is included in
this section.
A. Amendment to 40 CFR Part 50 (Appendix L): Reference Method for the
Determination of Fine Particulate Matter as PM2.5 in the Atmosphere--
Addition of the Tisch Cyclone as an Approved Second Stage Separator
The EPA proposed a change to the FRM for PM2.5 (40 CFR
part 50, appendix L), the addition of an alternative PM2.5
particle size separator to that of the Well Impactor Ninety-Six (WINS)
and the Very Shape Cut Cyclone (VSCC) size separators (88 FR 5663,
January 27, 2023). The new separator is the TE-PM2.5C
cyclone manufactured by Tisch Environmental Inc.,\180\ Cleves Ohio,
which has been shown to have performance equivalent to that of the
originally specified WINS impactor with regards to aerodynamic cutpoint
and PM2.5 concentration measurement. In addition, the new
TE-PM2.5C has a significantly longer service interval than
the WINS and is comparable to that of the VSCC separator. Generally,
the TE-PM2.5C is also physically interchangeable with the
WINS and VSCC where both are manufactured for the same sampler. The
proposed change would allow either the WINS, VSCC, or TE-
PM2.5C to be used in a PM2.5 FRM sampler. As is
the case for the WINS and VSCC, the TE-2.5C is now also an approved
size separator for candidate PM2.5 FEMs. Currently, the EPA
has designated one PM2.5 sampler configured with TE-
PM2.5C separator as a Class II PM2.5 equivalent
method and one as a PM10-2.5 equivalent method. Upon
promulgation of this change to appendix L, these instruments would be
redesignated as PM2.5 and PM10-2.5 FRMs,
respectively. Owners of such samplers should contact the sampler
manufacturer to receive a new reference method label for the samplers.
---------------------------------------------------------------------------
\180\ Mention of commercial names does not constitute EPA
endorsement.
---------------------------------------------------------------------------
The EPA received only one comment regarding this proposed change,
which was supportive. Therefore, the EPA is finalizing this change to
Appendix L as proposed.
B. Issues Related to 40 CFR Part 53 (Reference and Equivalent Methods)
The EPA proposed to clarify the regulations associated with FRM and
FEM applications for review by the EPA (88 FR 5664, January 27, 2023).
Revisions were also proposed in instances where current regulatory
specifications are no longer pertinent and require updating. In
addition, the EPA proposed to correct a compiled a list of noted minor
errors in the regulations associated with the testing requirements and
acceptance criteria for FRMs and FEMs in part 53. These errors are
typically not associated with the content of Federal Register documents
but often relate to transcription errors and typographical errors in
the electronic CFR (eCFR) and printed versions of the CFR.
1. Update to Program Title and Delivery Address for FRM and FEM
Applications
The EPA proposed a change to 40 CFR 53.4(a) to update the delivery
address for FRM and FEM Applications and Modification Requests, as well
as update the name of the program responsible for their review (88 FR
5664, January 27, 2023). These revisions are due solely to
organizational changes and do not affect the structure or role of the
Reference and Equivalent Methods Designation Program in reviewing new
FRM and FEM application requests and requests to modify existing
designated instruments. The EPA received no comments on this revision
and, therefore, the EPA is finalizing this revision as proposed.
2. Requests for Delivery of a Candidate FRM or FEM Instrument
The EPA proposed a change to 40 CFR 53.4(d), which currently allows
the EPA to request only candidate PM2.5 FRMs and Class II or
Class III equivalent methods for testing purposes as part of the
applicant review process (88 FR 5664, January 27, 2023). The EPA
proposed to revise this section to enable requesting any candidate FRM,
FEM, or a designated FRM or FEM associated with a Modification Request,
regardless of NAAQS pollutant type or metric. The EPA received no
comments on these revisions; therefore, the EPA is finalizing this
revision as proposed.
3. Amendments to Requirements for Submission of Materials in 40 CFR
53.4(b)(7) for Language and Format
The EPA proposed a change to 40 CFR 53.4(b)(7) to specify that all
written FRM and FEM application materials must be submitted to the EPA
in English in MS Word format and that submitted data must be submitted
in MS Excel format (88 FR 5664, January 27, 2023). The EPA received no
comments on these revisions; therefore, the EPA is finalizing this
section as proposed.
4. Amendment to Designation of Reference and Equivalent Methods
The EPA proposed a change to 40 CFR 53.8(a) to clarify the terms of
new FRM and FEM methods to ensure that candidate samplers and analyzers
are not publicly announced, marketed, or sold until the EPA's approval
has been formally announced in the Federal Register (88 FR 5664,
January 27, 2023). The EPA received no comments on these revisions;
therefore, the EPA is finalizing this section as proposed.
5. Amendment to One Test Field Campaign Requirement for Class III
PM2.5 FEMs
The EPA proposed a change to 40 CFR 53.35(b)(1)(ii)(D) that
involves field comparability tests for candidate Class III
PM2.5 FEMs, including the requirement that a total of five
field campaigns must be conducted at four separate sites, A, B, C, and
D (88 FR 5664, January 27, 2023). The existing Site D specifications
require that the site ``shall be in a large city east of the
Mississippi River, having characteristically high sulfate
concentrations and high humidity levels.'' However, dramatic decreases
in ambient sulfate concentration make it difficult for applicants to
routinely meet the high sulfate concentration requirement. Therefore,
the EPA proposed to revise the Site D specifications to read ``shall be
in a large
[[Page 16345]]
city east of the Mississippi River, having characteristically high
humidity levels.'' Only one comment was received on this proposed
revision, which was supportive. Therefore, the EPA is finalizing the
revision to 40 CFR 53.35(b)(1)(ii)(D), as proposed.
6. Amendment to Use of Monodisperse Aerosol Generator
The EPA proposed a change to 40 CFR 53.61(g), 53.62(e), and Table
F-1 that involves the wind tunnel evaluation of candidate
PM10 inlets and candidate PM2.5 fractionators
under static conditions, which requires the generation and use of
monodisperse calibration aerosols of specified aerodynamic sizes (88 FR
5664, January 27, 2023). In the current regulations, the TSI
Incorporated Vibrating Orifice Aerosol Generator (VOAG) is the only
monodisperse generator that is approved for this purpose. However, TSI
Incorporated no longer manufacturers nor supports the VOAG. Therefore,
a commercially available monodisperse aerosol generator (Model 1520
Fluidized Monodisperse Aerosol Generator, MSP Corporation, Shoreview,
MN) has been added to list of approved generators for this purpose. No
comments were received on this revision; therefore, the EPA is
finalizing this revision as proposed.
7. Corrections to 40 CFR Part 53 (Reference and Equivalent Methods)
Certain provisions of 40 CFR 53.14, Modification of a reference or
equivalent method, incorrectly state an EPA response deadline of 30
days for receipt of modification materials in response to an EPA
notice. Per a 2015 amendment (80 FR 65460, 65416, Oct. 26, 2015), all
EPA response deadlines for modifications of reference or equivalent
methods are 90 days from day of receipt. Thus, the EPA proposed a
correction to specify the correct 90-day deadline (88 FR 5664, January
27, 2023).
Requirements for Reference and Equivalent Methods for Air
Monitoring of Criteria Pollutants identifies the applicable 40 CFR part
50 appendices and 40 CFR part 53 subparts for each criteria pollutant.
The four rows in the section for PM10-2.5 erroneously do not
include the footnote instruction that the aforementioned pollutant
alternative Class III requirements may be substituted in regard to
Appendix O to Part 50--Reference Method for the Determination of Coarse
Particulate Matter as PM10-2.5 in the Atmosphere.
Table B-1 specifies that the interference equivalent for each
interferent is 0.005 ppm for both the standard-range and
lower-range limits, with the exception of nitric oxide (NO) for the
lower-range limit per note 4. When testing the lower range of
SO2, the limit for NO is 0.003 ppm, therefore,
an incorrect lower limit (0.0003) is currently stated in
note 4 for this exception to the SO2 lower range limit.
Thus, the EPA proposed a correction to Table B-1 to specify the correct
limit in note 4 (88 FR 5664, January 27, 2023).
After the EPA received an inquiry regarding the interaction of NO
and O3, the EPA investigated the interferent testing
requirements stated by 40 CFR part 53, subpart B. The EPA has
determined that during the 2011 SO2 amendment and subsequent
2015 O3 amendment, several typographical errors were
introduced into Table B-3, the most significant of which is the
omission of note 3, which instructs the applicant to not mix the
pollutant with the interferent. Thus, the EPA proposed revisions to
Table B-3 to correct these errors (88 FR 5664, January 27, 2023).
Additionally, appendix A to subpart B of part 53 provides figures
depicting optional forms for reporting test results. Figure B-3 lists
an incorrect formula: the lower detectible limit section is missing the
proper operator in the LDL calculation formula and Figure B-5 lists an
incorrect calculation metric, and there is a typesetting error in the
calculation of the standard deviation. The EPA proposed to correct the
typesetting errors and noted other errors to be corrected in several
formulas provided throughout Sec. 53.43 (88 FR 5664, January 27,
2023).
The EPA proposed a revision to 40 CFR 53.43(a)(2)(xvi),
53.43(b)(2)(iv), and 53.43(b)(2)(iv) to correct typographical errors in
equations.
The EPA proposed a revision to Table C-4 of part 53 Subpart C (88
FR 5700). This change is related to field comparability tests of
candidate PM2.5, PM10-2.5, and PM10
FEMs, which requires testing at wide range of ambient concentrations.
For this reason, Table C-4 specifies a minimum number of valid sample
sets to be conducted at specified high concentrations. However, due to
the dramatic decrease in ambient PM concentrations in the past two
decades, these number of valid test days at high concentrations has
been difficult to achieve. Accordingly, the EPA proposed to revise the
testing specifications for high concentration events in Table C-4 to
reflect current levels of ambient PM for all three PM metrics. In
addition to the revision of the ambient PM concentration specifications
to Table C-4, there are also several entry errors that required
correction.
The EPA received no comments on these proposed revisions;
therefore, the EPA is finalizing the changes as proposed.
C. Changes to 40 CFR Part 58 (Ambient Air Quality Surveillance)
1. Quality Assurance Requirements for Monitors Used in Evaluations for
National Ambient Air Quality Standards
In the proposal, the EPA described how we evaluated the quality
system as part of the PM NAAQS reconsideration (88 FR 5665, January 27,
2023). In this section, the EPA identified several areas for
improvement in steadily declining average ambient PM2.5
concentrations across the country and the final decision to revise
primary annual PM2.5 NAAQS described in section II above. We
assessed PM2.5 concentration data across a range of values
to determine if any changes to the statistical calculations used to
evaluate the data quality in the PM2.5 network were
warranted. This section describes the EPA's assessment, comments
received, and the EPA's final decisions on the proposed changes. Other
changes in this section include clarifications and other improvements
that will facilitate consistency and the operation of quality assurance
programs by State, local, and Tribal (SLT) agencies nationwide.
a. Quality System Requirements
The EPA reconsidered the appendix A, section 2.3.1.1 goal for
acceptable measurement uncertainty (88 FR 5665, January 27, 2023) for
automated and manual PM2.5 methods for total bias. The
existing total bias goal is an upper 90 percent confidence limit for
the coefficient of variation (CV) of 10 percent and 10
percent for total bias. The intent of the proposal was to investigate
if this bias goal is still realistic given updated precision and bias
statistic. The EPA received one comment that bias reevaluation may be
premature, since the final NAAQS standard had not yet been determined
at the time of the proposal. The EPA acknowledges this comment but
clarifies that the proposed new bias statistic was evaluated at a range
of levels including the range of proposed PM2.5 standards in
the technical memorandum, ``Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network.'' \181\ Considering the
[[Page 16346]]
justification in the technical memorandum and the lack of adverse
comments regarding this part of the proposal, the EPA is retaining the
appendix A, section 2.3.1.1, goal for acceptable measurement
uncertainty for automated and manual PM2.5 methods for total
bias.
---------------------------------------------------------------------------
\181\ Noah, G. (2023). Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network. Memorandum to the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for Particulate Matter (EPA-
HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
The EPA also proposed to update and clarify ambient air monitoring
requirements found in 40 CFR part 58, appendix A, section 2.6.1
pertaining to EPA Protocol Gas standards used for ambient air
monitoring and the Ambient Air Protocol Gas Verification Program (PGVP)
(88 FR 5665, January 27, 2023). The EPA proposed to revise appendix A
to clarify that in order to participate in the Ambient Air PGVP,
producers of Protocol Gases must adhere to the requirements of 40 CFR
75.21(g), and only regulatory ambient air monitoring programs may
submit cylinders for assay verification to the EPA Ambient Air PGVP.
The EPA received mixed comments in support of and in opposition to this
proposed revision. The sole commenter opposing the proposed revision
indicated that the proposed PGVP requirements would be additional and
is concerned with an increased resource burden. But the EPA responds
that the PGVP requirements that were proposed to be added are
consistent with the existing PGVP requirements in 40 CFR 75.21(g), and
PGVP has been defined as a regulatory requirement since 2016 (81 FR
17263, March 28, 2016), so the proposed part 58 changes are not
``additional'' to existing regulations. After consideration of the
comments, the EPA is finalizing the update and clarification of ambient
air monitoring requirements found in appendix A, section 2.6.1
pertaining to EPA Protocol Gas standards used for ambient air
monitoring and the Ambient Air PGVP as proposed.
b. Measurement Quality Check Requirements
The EPA proposed to remove section 3.1.2.2 from appendix A, which
allows NO2 compressed gas standards to be used to generate
audit standards (88 FR 5665, January 27, 2023). The EPA received one
comment supporting this change. As a result of the comment received and
other general supportive comments regarding quality assurance, the EPA
is finalizing the removal of section 3.1.2.2 from appendix A as
proposed.
The EPA proposed to revise the requirement in Appendix A, section
3.1.3.3 changing the National Performance Audit Program (NPAP)
requirement for annual verification of gaseous standards to the ORD-
recommended certification periods identified in Table 2-3 of the EPA
Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards (appendix A, section 6.0(4)) (88 FR 5665). The
EPA received one comment supporting this change. As a result of the
comment received and other general supportive comments regarding
quality assurance, the EPA is finalizing the updated NPAP gaseous
certification requirement in section 3.1.3.3 as proposed.
The EPA proposed to adjust the minimum value required by appendix
A, section 3.2.4, to be considered valid sample pairs for the
PM2.5 Performance Evaluation Program (PEP) from 3 [mu]g/m\3\
to 2 [mu]g/m\3\ (88 FR 5665, January 27, 2023). The EPA received
comments in support and against the change. In the only opposing
comment, the commenter expressed concern that the method detection
limit (MDL) for PM2.5 is 2 [mu]g/m\3\. The commenter also
indicated that the MDL ``typically has minimal value per the definition
of the MDL.'' 40 CFR part 50, appendix L states, ``The lower detection
limit of the mass concentration measurement range is estimated to be
approximately 2 [mu]g/m\3\, based on noted mass changes in field blanks
in conjunction with the 24 m\3\ nominal total air sample volume
specified for the 24-hour sample.'' The EPA notes that field blanks
currently average less than 10 [mu]g nationally, and when divided by
the 24 m\3\ nominal total air sample volume specified for a 24-hour
sample, the result is 0.4 [mu]g/m\3\. The appendix L MDL referenced by
the commenter was part of the 1997 PM NAAQS rulemaking (62 FR 38652,
July 18, 1997); current data shows that the MDL is substantially lower
than the EPA's original estimate. After review of the comments, and in
consideration of the recently calculated detection limit for the
PM2.5 FRM that is substantially lower than our original
estimate,\182\ the EPA is finalizing the revised minimum value for
valid sample pairs for the PM2.5 Performance Evaluation
Program (PEP) from 3 [mu]g/m\3\ to 2 [mu]g/m\3\ in appendix A, section
3.2.4 as proposed.
---------------------------------------------------------------------------
\182\ See the EPA's PM2.5 Data Quality Dashboard
available at https://sti-r-shiny.shinyapps.io/QVA_Dashboard/.
---------------------------------------------------------------------------
c. Calculations for Data Quality Assessments
The EPA proposed to change Equations 6 and 7 of appendix A, section
4.2.1 that are used to calculate the Collocated Quality Control Sampler
Precision Estimate for PM10, PM2.5 and Pb (88 FR 5666, January 27,
2023). The proposed new statistics are designed to address the high
imprecision values that result from using these calculations to compare
low concentrations that are now more routinely observed in the
networks. The EPA received several comments in support of this change
in general, but some commenters indicated that they believed there was
an error in the new calculation that may result in high imprecision
from the calculation of the equation. The EPA reviewed the technical
memorandum and confirmed that a multiplier of 100 was unintentionally
left in the proposed relative difference equation, Equation 6. Also,
equation 6 was corrected from a normalized percent difference to a
normalized relative percent difference that is appropriate for
comparing collocated pairs at low concentrations. The technical
memorandum titled ``Task 16 on PEP/NPAP Task Order: Bias and Precision
DQOs for the PM2.5 Ambient Air Monitoring Network'' has been
amended to correct the error and is included in the docket for this
action.\183\
---------------------------------------------------------------------------
\183\ Noah, G. (2023). Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network. Memorandum to the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for Particulate Matter (EPA-
HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
Equation 6 as proposed at 88 FR 5666 (January 27, 2023) was:
[[Page 16347]]
[GRAPHIC] [TIFF OMITTED] TR06MR24.031
As a result of the positive comments received and the correction to
the equation made in response to some comments, the EPA is finalizing
the updated Equation 6 as described and is finalizing Equation 7 as
proposed for the calculation of the Collocated Quality Control Sampler
Precision Estimate for PM10, PM2.5, and Pb in section 4.2.1.
The EPA proposed to update the appendix A, section 4.2.5, Equation
8, calculation for the Performance Evaluation Program Bias Estimate for
PM2.5 (88 FR 5666-67, January 27, 2023). Because average
ambient PM concentrations across the nation have steadily declined
since the promulgation of the PM2.5 standard, the EPA
proposed to replace the current percent difference equation with a
relative difference equation. The EPA received several comments in
support of this change in general, but some commenters identified a
potential error in the new calculation that resulted in an artificially
high estimate, which they do not support. The EPA reviewed the
technical memorandum and discovered that a multiplier of 100 was left
in the new relative difference equation used in the bias equation. The
technical memorandum, ``Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network'' has been amended to correct the error and is included in the
docket.\184\ The proposed Equation 8 proposed at 88 FR 5667 (January
27, 2023) was:
---------------------------------------------------------------------------
\184\ Noah, G. (2023). Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network. Memorandum to the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for Particulate Matter (EPA-
HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
[GRAPHIC] [TIFF OMITTED] TR06MR24.032
As a result of the supportive comments received and the correction
to the equation in response to some comments, the EPA is updating and
finalizing Equation 8 as described for the calculation for the
Performance Evaluation Program Bias Estimate for PM2.5, in
section 4.2.5.
d. References
The EPA proposed to update the references and hyperlinks in
appendix A, section 6 (88 FR 5667, January 27, 2023) to provide
accuracy in identifying and locating essential supporting documentation
and delete references to historical documents that do not represent
current practices. The EPA received only favorable comments, and as a
result, the EPA is finalizing the updated the references and hyperlinks
in appendix A, section 6, as proposed.
The EPA also proposed to add a footnote to Table A-1 of part 58,
appendix A--Minimum Data Assessment Requirements for NAAQS Related
Criteria Pollutant Monitors (88 FR 5669, January 27, 2023). The
proposed footnote clarifies the allowable time (i.e., every two weeks,
once a month, once a quarter, once every six months, or distributed
over all four quarters depending on the check) between checks and
encourages monitoring organizations to perform data assessments at
regular intervals. The EPA received two comments regarding this
proposed footnote. One commenter indicated that this change is
inconsistent with the QA Handbook for Air Pollution Measurement
Systems: ``Volume II: Ambient Air Quality Monitoring Program QA
Handbook.'' The EPA agrees with the commenter; because the QA Handbook
is guidance,
[[Page 16348]]
the EPA will revise it after this action is finalized to be consistent
with the updated CFR provision. Another commenter does not support the
addition of the footnote due to concerns about limiting flexibility. In
response, the EPA reiterates that the proposed revision is intended to
clarify intent and does not make any changes to the required
frequencies or acceptance criteria for data assessment. A ``weight of
evidence'' narrative is still found in 40 CFR part 58, appendix A,
section 1.2.3. As a result of the comments received and the rationale
discussed above, the EPA is finalizing the addition of the new footnote
to Table A-1 of part 58, appendix A--Minimum Data Assessment
Requirements for NAAQS Related Criteria Pollutant Monitors as proposed.
2. Quality Assurance Requirements for Prevention of Significant
Deterioration (PSD) Air Monitoring
The EPA proposed to revise appendix B, Quality Assurance
Requirements for Prevention of Significant Deterioration (PSD) Air
Monitoring (88 FR 5667, January 27, 2023), in parallel to the proposal
to revise appendix A. Thus, this section of the proposal included
similar detail and proposed revisions related to evaluating quality
system statistical calculations for PM2.5, clarifications
and other improvements that would facilitate consistency and the
operation of quality assurance programs for PSD by SLT agencies
nationwide.
a. Quality System Requirements
The EPA reconsidered the goal in appendix B, section 2.3.1.1 for
acceptable measurement uncertainty for automated and manual
PM2.5 methods for total bias (88 FR 5668, January 27,
2023).\185\ The current total bias goal is an upper 90 percent
confidence limit for the coefficient of variation (CV) of 10 percent
and 10 percent for total bias. The EPA's intent was to
investigate if this goal is still realistic given updated precision and
bias statistics. The EPA received one comment that bias reevaluation
may be premature, since the final NAAQS standard had not yet been
determined at the time of the proposal. The EPA acknowledges this
comment but clarifies that the proposed new bias statistic was
evaluated at a range of levels including the proposed range of
PM2.5 standards in the technical memorandum, ``Task 16 on
PEP/NPAP Task Order: Bias and Precision DQOs for the PM2.5
Ambient Air Monitoring Network.'' \186\ Considering the justification
in the technical memorandum and the lack of adverse comments regarding
the substantive proposal, the EPA is retaining the appendix B, section
2.3.1.1, goal for acceptable measurement uncertainty for automated and
manual PM2.5 methods for total bias.
---------------------------------------------------------------------------
\185\ In the proposal, in section VII.C.2 Quality Assurance
Requirements for Prevention of Significant Deterioration (PSD) Air
Monitoring (88 FR 5667-69), the EPA inadvertently referred to
``appendix A'' in the section rather than the correct ``appendix
B.'' The EPA's intent to have proposed changes to appendix B on
these pages is made clear by the section header, the Table of
Contents on page 5559, and the proposed regulatory text for appendix
B on pages 5707-08. See, e.g., id. at p.5668 (preamble erroneously
states that the EPA proposed to change appendix A, section 2.6.1);
id. at p.5668 (preamble erroneously states that the EPA proposed to
adjust the minimum value required by appendix A, section 3.2.4).
\186\ Noah, G. (2023). Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network. Memorandum to the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for Particulate Matter (EPA-
HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
The EPA also proposed to update and clarify ambient air monitoring
requirements found in 40 CFR part 58, appendix B, section 2.6.1
pertaining to EPA Protocol Gas standards used for ambient air
monitoring and the Ambient Air PGVP (88 FR 5668, January 27, 2023). The
EPA proposed to revise appendix B to clarify that in order to
participate in the Ambient Air PGVP, producers of Protocol Gases must
adhere to the requirements of 40 CFR 75.21(g), and only regulatory
ambient air monitoring programs may submit cylinders for assay
verification to the EPA Ambient Air PGVP. The EPA received comments in
support of and in opposition to this proposed revision. The commenter
opposing the revision indicated that the proposed PGVP requirements
would be additional and is concerned with an increased resource burden.
However, the EPA disagrees with the commenter because that the proposed
PGVP requirements are consistent with the existing PGVP requirements in
40 CFR 75.21(g). PGVP has been defined as a regulatory requirement
since 2016 (81 FR 17263, March 28, 2016), so the proposed part 58
changes are not ``additional'' to existing regulations. After
consideration of the comments, the EPA is finalizing the update and
clarification of ambient air monitoring requirements found in appendix
B, section 2.6.1 pertaining to EPA Protocol Gas standards used for
ambient air monitoring and the Ambient Air PGVP as proposed.
b. Measurement Quality Check Requirements
The EPA proposed to remove section 3.1.2.2 from appendix B, which
allows NO2 compressed gas standards to be used to generate
audit standards (88 FR 5668, January 27, 2023). The EPA received one
comment supporting this change. As a result of the comment received and
other general supportive comments regarding quality assurance, the EPA
is finalizing the removal of section 3.1.2.2 from appendix B as
proposed.
The EPA proposed to revise the requirement in Appendix B, section
3.1.3.3 changing the National Performance Audit Program (NPAP)
requirement for annual verification of gaseous standards to the ORD-
recommended certification periods identified in Table 2-3 of the EPA
Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards (appendix B, section 6.0(4)) (88 FR 5668, January
27, 2023). The EPA received one comment supporting this change. As a
result of the comment received and other general supportive comments
regarding quality assurance, the EPA is finalizing the updated NPAP
gaseous certification requirement in section 3.1.3.3 as proposed.
The EPA proposed to adjust the minimum value required by appendix
B, section 3.2.4, to be considered valid sample pairs for the
PM2.5 Performance Evaluation Program (PEP) from 3 [mu]g/m\3\
to 2 [mu]g/m\3\ (88 FR 5668, January 27, 2023). The EPA received
comments in support and against the change. In the only opposing
comment, the commenter expressed concern that the method detection
limit (MDL) for PM2.5 is 2 [mu]g/m\3\. The commenter also
indicated that the MDL ``typically has minimal value per the definition
of the MDL.'' 40 CFR part 50, appendix L states, ``The lower detection
limit of the mass concentration measurement range is estimated to be
approximately 2 [mu]g/m\3\, based on noted mass changes in field blanks
in conjunction with the 24 m\3\ nominal total air sample volume
specified for the 24-hour sample''. The EPA notes that field blanks
currently average less than 10 [mu]g nationally, and when divided by
the 24 m\3\ nominal total air sample volume specified for a 24-hour
sample, the result is 0.4 [mu]g/m\3\. The appendix L MDL referenced by
the commenter was part of the 1997 PM NAAQS rulemaking more than 20
years ago (62 FR 38652, July 18, 1997); current data shows that the MDL
is substantially lower than EPA's original estimate. After review of
the comments, and in consideration of the recently calculated
[[Page 16349]]
detection limit for the PM2.5 FRM that is substantially
lower than our original estimate, the EPA is revising the minimum value
for valid sample pairs for the PM2.5 Performance Evaluation
Program (PEP) from 3 [mu]g/m\3\ to 2 [mu]g/m\3\ in appendix B, section
3.2.4 as proposed.
c. Calculations for Data Quality Assessments
The EPA proposed to change Equations 6 and 7 of appendix B, section
4.2.1 used for calculating the Collocated Quality Control Sampler
Precision Estimate for PM10, PM2.5 and Pb (88 FR 5707, January 27,
2023). These new statistics are designed to address the high
imprecision values that result from using these calculations to compare
low concentrations that are now more routinely observed in the
networks. The EPA received several comments in support of this change
in general, but a couple commenters indicated that there could be an
error in the new calculation that resulted in high imprecision from the
calculation of the equation. The EPA reviewed the technical memorandum
and discovered that a multiplier of 100 was unintentionally left in the
proposed relative difference equation, Equation 6. Also, equation 6 was
corrected from a normalized percent difference to a normalized relative
percent difference that is appropriate for comparing collocated pairs
at low concentrations. The technical memorandum titled ``Task 16 on
PEP/NPAP Task Order: Bias and Precision DQOs for the PM2.5
Ambient Air Monitoring Network'' was amended to correct the error and
is included in the docket.\187\
---------------------------------------------------------------------------
\187\ Noah, G. (2023). Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network. Memorandum to the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for Particulate Matter (EPA-
HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
[GRAPHIC] [TIFF OMITTED] TR06MR24.033
As a result of the positive comments received and the correction to
the equation made in response to those comments, the EPA is finalizing
the update to Equation 6 and retaining Equation 7 as proposed for the
calculation of the Collocated Quality Control Sampler Precision
Estimate for PM10, PM2.5 and Pb in section 4.2.1.
The EPA proposed to update the appendix B, section 4.2.5, Equation
8, calculation for the Performance Evaluation Program Bias Estimate for
PM2.5 (88 FR 5668-59, January 27, 2023). Because average
ambient PM concentrations across the nation have steadily declined
since the promulgation of the PM2.5 standard, the EPA
proposed to replace the current percent difference equation with a
relative difference equation. The EPA received several comments in
support of this change in general, but some commenters identified a
potential error in the new calculation that resulted in an artificially
high estimate, which they do not support. The EPA reviewed the
technical memorandum and discovered that a multiplier of 100 was left
in the new relative difference equation used in the bias equation. The
technical memorandum, ``Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network'' has been amended to correct the error and is included in the
docket. The proposed Equation 8 (88 FR 5669, January 27, 2023) was:
[[Page 16350]]
[GRAPHIC] [TIFF OMITTED] TR06MR24.034
As a result of the supportive comments received and the correction
to the equation in response to some comments, the EPA is updating and
finalizing Equation 8 as described for the calculation for the
Performance Evaluation Program Bias Estimate for PM2.5, in
section 4.2.5.
d. References
The EPA proposed to update the references and hyperlinks in
appendix B, section 6 (88 FR 5669, January 27, 2023) to provide
accuracy in identifying and locating essential supporting documentation
and delete references to historical documents that do not represent
current practices. The EPA received only favorable comments, and as a
result, the EPA is finalizing the updated the references and hyperlinks
in appendix B, section 6, as proposed.
The EPA also proposed to add a footnote to Table B-1 of part 58,
appendix B--Minimum Data Assessment Requirements for NAAQS Related
Criteria Pollutant PSD Monitors (88 FR 5669, January 27, 2023). The
proposed footnote clarifies the allowable time (i.e., every two weeks,
once a month, once a quarter, once every six months, or distributed
over all four quarters depending on the check) between checks and
encourages monitoring organizations to perform data assessments at
regular intervals. The EPA received two comments regarding this
proposal. One commenter indicated that this change is inconsistent with
the QA Handbook. The EPA agrees with the commenter; because the QA
Handbook is guidance, the EPA will revise it after this action is
finalized to be consistent with the updated CFR provision. Another
commenter does not support the addition of the footnote due to concerns
about limiting flexibility. In response, the EPA reiterates that the
proposed revision is intended to clarify intent and does not make any
changes to the required frequencies or acceptance criteria for data
assessment. A ``weight of evidence'' narrative is still found in 40 CFR
part 58, appendix B, section 1.2.3. As a result of the comments
received and the rationale discussed above, the EPA is adding the new
footnote to Table B-1 of part 58, appendix B--Minimum Data Assessment
Requirements for NAAQS Related Criteria Pollutant PSD Monitors as
proposed.
3. Amendments to PM Ambient Air Quality Methodology
a. Revoking Approved Regional Methods (ARMs)
The EPA proposed to remove provisions for approval and use of
Approved Regional Methods (ARMs) throughout parts 50 and 58 of the CFR
(88 FR 5669, January 27, 2023). ARMs are continuous PM2.5
methods that have been approved specifically within a State or local
air agency monitoring network for purposes of comparison to the NAAQS
and to meet other monitoring objectives. Currently, there are no
approved ARMs. There are, however, more than a dozen approved Federal
Equivalent Methods (FEMs) for PM2.5. These approved FEMs are
eligible for comparison to the NAAQS and to meet other monitoring
objectives.
The EPA received comments from multiple State air programs in
support of the proposal to remove provisions for approval and use of
ARMs. One commenter cites that there are multiple FEMs available for
monitoring agencies to work with and that the agency was never able to
get a candidate ARM to meet the requirements for approval. With the
availability of multiple FEMs that now work in the monitoring agency's
network, the commenting agency does not anticipate the need to ever
pursue an ARM in the future and, therefore, suggests that the ARM
provision is no longer needed. Another commenter strongly supported the
proposed changes to remove the ARM provisions. The EPA also received
comments from a few agencies that supported retaining the ARM
provisions instead. One commenter cited the need to consider the rapid
advancement of various new technologies and that, in some cases,
approved continuous FEMs may have shortcomings, meaning that losing the
ability to propose an ARM in the future may limit useful alternative
options to monitoring agencies. Another commenter suggested that the
removal of the ARM would take away the ability and right to use locally
derived correction factors.
After considering the comments for and against removing the
provisions for ARMs, the EPA believes it is most appropriate to remove
the ARM provisions. As described in the proposal, when the EPA first
proposed the process for approving and using ARMs, there were no
continuous FEMs approved. There are now over a dozen approved
PM2.5 continuous FEMs and no approved ARMs. Therefore, the
EPA is finalizing the removal of ARMs throughout 40 CFR parts 50 and 58
as proposed.
b. Calibration of PM Federal Equivalent Methods (FEMs)
The EPA proposed to modify its specifications for PM FEMs in
appendix C to Part 58 (88 FR 5670-73, January 27, 2023). Specifically,
the EPA proposed that valid State, local, and Tribal (SLT) air
monitoring data from Federal Reference Methods (FRMs) generated in
routine networks and submitted to the EPA may be used to improve the PM
concentration measurement performance of approved FEMs. This approach,
initiated by instrument manufacturers, would be implemented as a
national solution in factory calibrations of approved FEMs through a
firmware update. This could apply to any PM FEM methods (i.e.,
PM10, PM2.5, and PM10-2.5).
The EPA proposed this modification because there are some approved
PM FEMs that are not currently meeting bias measurement quality
objectives (MQOs) when evaluating data nationally as described in the
2022 PA (U.S. EPA, 2022b, section 2.2.3.1), meaning that an update to
factory calibrations may be appropriate; however, there is no clearly
defined process to update the calibration of FEMs. While there are
several types of data available to use as the reference for such
updates (e.g., routinely operated FRMs, audit program FRMs, and
chemical speciation sampler
[[Page 16351]]
data), we proposed to use routinely operated SLT FRMs as the basis of
comparison upon which to calibrate FEMs. The goal of updating factory
calibrations would be to increase the number of routinely operating
FEMs meeting bias MQOs across the networks in which they are operated.
While there are other approaches that could improve data comparability
between PM FEMs and collocated FRMs, the EPA believes that the proposed
modification to calibrate PM FEMs represents the most reliable approach
to update FEM factory calibrations, since the existing FRM network data
that meet MQOs would be used to set updated factory calibrations.
While the Agency proposed to add this language to more expressly
define a process to update factory calibrations of approved PM FEMs,
the EPA believes that the existing rules for updating approved FRMs and
FEMs found at 40 CFR 53.14 may also continue to be utilized for this
purpose, as appropriate. 40 CFR 53.14 allows instrument manufactures to
submit to the EPA a ``Modification of a reference or equivalent
method.'' Submitting a modification request may be appropriate to
ensure an approved FEM continues to meet 40 CFR 53.9, ``Conditions of
designation.'' Specifically, 40 CFR 53.9(c) requires that, ``Any
analyzer, PM10 sampler, PM2.5 sampler, or
PM10-2.5 sampler offered for sale as part of an FRM or FEM
shall function within the limits of the performance specifications
referred to in Sec. 53.20(a), Sec. 53.30(a), Sec. 53.35, Sec.
53.50, or Sec. 53.60, as applicable, for at least 1 year after
delivery and acceptance when maintained and operated in accordance with
the manual referred to in Sec. 53.4(b)(3).'' Thus, instrument
manufacturers are encouraged to seek improvements to their approved FEM
methods as needed to continue to meet data quality needs as operated
across the network.
There are several technical components to EPA's proposed
modification, including: the reference data to be used in the
calibrations; implementing as a national solution in factory
calibrations of approved FEMs through firmware updates; application to
any PM FEM methods (i.e., PM10, PM2.5, and
PM10-2.5); the appropriate range of data to be used to
develop and test new factory calibrations, from just the most
representative concentrations up to all available concentrations; the
representative set of geographic locations that can be used; whether
outliers may be included or not included; that new factory calibrations
should be developed using data from at least 2 years and tested on data
from a separate year or years; that updates to factory calibrations can
occur as often as needed; that calibrations should be evaluated by
monitoring agencies as part of routine data assessments, e.g., during
certification of data and 5-year assessments; the EPA's recognition
that only data from existing operating sites is available; and finally,
that an updated factory calibration does not have to work with the
original field study data submitted that led to the original FEM
designation.
With the proposed modification, the EPA solicited input on these
technical issues as well as the overall approach and any alternatives
that could lead to more sites meeting the bias MQO with automated FEMs,
especially for those sites that are near the level of the primary
annual PM2.5 NAAQS, as proposed to be revised in section II
above. In response, the EPA received comments from about two dozen
entities, most of which were SLT air programs or Multi-Jurisdictional
Organizations (MJOs) comprised of these entities.
Overall, there was broad and strong support from a majority of
commenters for the proposed requirement to use FRM data generated in
routine networks and submitted to the EPA to update factory
calibrations included as part of approved FEMs. There were a smaller
number of critical comments on the proposed process as well as some
commentors that supported the proposed requirement but also provided
additional suggestions for the EPA's consideration. Below, we address
each of the areas on which the EPA requested comment regarding the
calibration of PM FEMs, as well as a few additional areas where
multiple commenters offered input on other areas related to our
proposal.
A majority of the commenters on the proposed PM FEM calibration
process support the process to use valid State, local, and Tribal FRM
data generated in routine networks and submitted to the EPA to improve
the PM concentration measurement performance of approved FEMs. Some
commenters suggested that this action is needed to ensure that data
reported from FRMs and FEMs are comparable and correction methods
applied to data from FEM monitors are defensible across the national PM
monitoring network. Others stated that they agree with the EPA that
this is a critical step in the right direction to account for the
discrepancies between PM2.5 FRM data and PM2.5
FEM data. Some commented that applying corrections includes a
recognition that, while different measurement principles may produce
differences in the resulting data, having an approach that minimizes
bias is extremely important. Finally, some stated their belief that a
correction factor is necessary to preserve data integrity with the FRM.
The EPA also received comments suggesting ways that the PM FEM
correction could be performed, including through detailed analysis of
data; by having PM FEM instrument manufacturers evaluate nationally
available valid FRM data to update factory calibrations; and, by having
the instrument manufacturers implement calibration adjustments at the
factory.
The EPA also received supportive comments on the PM FEMs
calibration relating to comparability to the NAAQS. For example, a
commenter stated that it is important to ensure bias MQOs are met for
FEMs run at sites potentially affected by revised standards as well as
the need to accurately designate areas as attaining or not attaining
the NAAQS. There were comments supporting the correction of PM FEM data
as helping the EPA and SLT monitoring programs continue to evolve
toward more automated methods. For example, one commenter appreciates
the EPA's support for the ongoing move from filter-based
PM2.5 FRMs to use of continuous FEMs, stating that they
concur with the EPA's assessment that there is monitoring bias between
FRMs and FEMs, and commending EPA for recognizing ongoing data quality
issues for FEMs and for taking action to improve these issues in
collaboration with instrument manufacturers and SLT agencies.
A small number of commenters were critical of the proposed FEM
calibration approach. One commenter noted that EPA should further
examine the handling of FEM PM2.5 data when used for
comparison to the NAAQS. In response, we note that monitoring agencies
and the EPA will continue to examine the comparability and use of FEM
data used in comparison to the NAAQS. Another commenter suggested that
the calibration process for a designated PM monitor should not be
altered following Class III designation approval. The EPA disagrees as
we believe it is appropriate for FEMs to be calibrated with routinely
operated FRMs, because doing so is an efficient way to work towards FEM
data meeting the bias MQO across the networks in which the FEMs are
currently being operated. Also, having continuous PM FEMs meeting bias
MQOs allows the use of the data in a variety of other ways that
manually operated FRMs samplers cannot support. Another commenter
stated that, if a particular FEM designated make or model of
[[Page 16352]]
instruments fails to meet MQOs, then that make or model should be
removed from the designations altogether. The EPA agrees and clarifies
that the modification would not prevent removal of FEM designation from
a make or model of instrument under the existing 40 CFR 53.11--
Cancellation of reference or equivalent method designation. This may be
appropriate if there are no other solutions to improve the method such
that it achieves bias MQOs.
A few commenters provided specific recommendations for how the
regulatory language could be improved. These included comments that the
new regulatory language proposed for 40 CFR part 58, appendix C,
section 2.2 must ensure consistency and transparency when requesting
changes to the factory calibration; that the EPA should incorporate
binding regulatory language in 40 CFR part 58, appendix C, section 2.2
(i.e., it currently lacks ``shall'' or ``must'') to ensure the language
is not open to inconsistency and does not provide unique deference to
instrument manufacturers without a mechanism for transparent
communication of the changes being made and the supporting technical
analysis. A commenter also requested that the EPA define the core
requirements needed to ensure all requests for updating factory
calibrations are required to follow the same process, using data of the
same known quality, and evaluating the effectiveness of the resulting
correction factors consistently.
In response to these comments, while the EPA agrees that the
proposed regulatory language for 40 CFR part 58, appendix C, section
2.2 must ensure consistency and transparency when entities request
changes to factory calibrations, the EPA disagrees that the regulations
cannot also provide some flexibility. For example, we believe that a
degree of flexibility is appropriate regarding whether outliers in the
data to be used for factory calibration should or should not be
included, the range of data to be included, and in utilizing collocated
FRM and FEM data for updated calibrations from a representative set of
geographic areas in which it is produced. The EPA believes that the
proposal defined the core requirements needed to ensure all requests
for updating FEM factory calibrations will follow the same process,
using data of the same known quality and evaluating the effectiveness
of the resulting correction factors consistently.
In its proposal, the EPA identified that while there are several
types of data available to use as the reference for FEM calibration
updates, including data from routinely operated FRMs, audit program
FRMs, and PM2.5 chemical speciation samplers, the EPA
proposed to use routinely operated State, local, and Tribal FRMs as the
basis of comparison upon which to calibrate FEMs (88 FR 5670-71,
January 27, 2023). Importantly, routine SLT agency FRM data form the
largest portion of the monitored air quality data used in epidemiologic
studies that are being used to inform proposed decisions regarding the
adequacy of the public health protection afforded by the primary
PM2.5 NAAQS, as discussed in section II above.
Overall, there was broad and strong support for utilizing
collocated FRM data from routine SLT networks to provide calibrations
of the continuous FEMs. For example, several commenters agree that
valid SLT air monitoring data generated in routine networks and
submitted to the EPA will improve the PM concentration measurement
performance of approved FEMs. Another commenter provided support for PM
FEM instrument manufacturers to evaluate nationally available valid FRM
data as well as other data sets such as the performance evaluation
audit program to update factory calibrations. The EPA believes that the
routinely operated PM FRMs represent the best and largest source of
data to calibrate continuous PM FEMs, and that performance evaluation
audit program data should be kept independent of the calibration
process. This will mean that assessments of the routine monitoring
operations, including both the FRM and any future updated PM FEMs, will
appropriately remain independent in evaluating whether updated methods
are meeting bias MQOs. The EPA is, therefore, finalizing its approach
to use routinely operated SLT FRMs as the basis of comparison upon
which to calibrate continuous PM FEMs as proposed.
Regarding the EPA's proposed requirement to utilize factory
calibrations (88 FR 5670-71, January 27, 2023), several commenters
agreed that factory calibrations provide the best option to improve PM
FEMs. For example, one commenter stated that the correction factors are
necessary to preserve data integrity with the FRM, and they support the
proposal that the approach be initiated by instrument manufacturers and
implemented as a national solution through firmware updates.
Regarding the proposed requirement that calibrations be initiated
by instrument manufacturers (88 FR 5671, January 27, 2023), most
commenters were supportive of the proposed approach that recalibration
of FEM PM instruments be initiated by instrument manufacturers. For
example, one commenter stated they support allowing instrument
companies submit improvements to their existing FEMs, as vendors should
be encouraged to improve their methods. Another commenter noted that
having a methodology initiated by the manufacturer will have nationwide
consistency. A few of commenters recommended that SLT air agencies
should have the additional ability to petition the EPA Administrator to
initiate factory calibrations of FEMs to better meet MQOs when data
collected by their agencies indicate disparities, because the
monitoring agencies are responsible for the quality of the data from
the specific makes and models of instrumentation used in their
networks. While the EPA believes that, in most cases, the instrument
companies should be the ones to initiate the process for calibration of
FEMs to routinely operated FRMs, we agree with the commenters who
suggested that other options should be available, including allowing
monitoring agencies or MJOs to work independently or together to pursue
improvements to designated FEMs. However, the EPA believes that any
such improvements initiated by monitoring agencies or MJOs should still
be facilitated through the responsible instrument company. Also, any
such effort to improve data quality should be employed across all the
networks in which the methods are operated and not limited to the
networks operated by the agency(s) pursuing such improvements.
Regarding how frequently factory calibrations should be updated,
our proposal identified that it would be most appropriate to not define
a specific time period for updates; rather, updates should be based on
whether or not quality data is being produced across a given network
(88 FR 5672, January 27, 2023). Regarding this issue, one commenter
recommended that instrument manufacturers be required to evaluate and,
if necessary, adjust PM FEMs factory calibrations on an ongoing basis
at regular intervals. The EPA notes that while it does not have the
authority to require instrument companies to evaluate the quality of
data from operating FEMs under 40 CFR part 58, the EPA does routinely
participate in conferences and workshops and makes assessments of data
quality specific to instrument makes and models publicly available. The
EPA also regularly summarizes relevant FRM and FEM data
[[Page 16353]]
quality in documents such as the 2022 PA (U.S. EPA, 2022b). Therefore,
consistent with the proposal, we are not finalizing any specifics
regarding how frequently factory calibrations should be updated but
commit to continue to routinely provide information to SLT agencies
regarding FEM data quality.
The EPA proposed that the calibration of FEMs could apply to any of
the PM FEM method indicators (i.e., PM10, PM2.5,
and PM10-2.5) (88 FR 5670, January 27, 2023). The EPA
received only supportive comments. All comments that included a
discussion of three PM metrics support their inclusion for calibration
of PM FEMs. Therefore, the EPA is finalizing the inclusion of all three
PM indicators (i.e., PM10, PM2.5, and
PM10-2.5) as proposed.
The EPA proposed that either all data available or a range of data
up to 125% of the 24-hour NAAQS for the PM indicator of interest may be
used to establish new factory calibrations, (88 FR 5671-73, January 27,
2023). The EPA received many comments supportive of the proposal and
one comment offering a different approach on the range of data to use.
One commenter recommends that the EPA should consider using all
``validated'' data because how these instruments behave under normal
operating ranges may be just as important as how they behave when
monitoring conditions are low or elevated, and that the full range of
data should be used when determining the appropriate level of the
standard, just as the full range of data is used in determining if an
area is attaining the standard. In response to this comment, the EPA
believes that making allowances for some flexibilities will increase
the likelihood of instrument companies pursuing such improvements.
Also, even though there is flexibility, the EPA will still be able to
evaluate the appropriateness of a range of concentration data included
as part of each application submitted. Also, the EPA notes that in
certain circumstances, States do petition the EPA to set aside data
under the Exceptional Events Rule (Sec. 50.14, ``Treatment of air
quality monitoring data influenced by exceptional events''). Where
approved, exceptional event data are set aside from use in regulatory
decisions. Thus, there is a process to set aside certain high
concentration data for certain purposes. Therefore, the EPA is
finalizing the provision that factory calibrations may be based on a
range of valid data as proposed.
The EPA solicited comment on the representative set of geographic
locations to use in the calibration of FEMs compared to collocated FEMs
(88 FR 5671, January 27, 2023). Most commenters were supportive of the
approach of using representative sites in SLT networks from across the
country. For example, several commenters provided their support for PM
FEM instrument manufacturers to evaluate nationally available valid FRM
data to update factory calibrations. Commenters disagreeing with a
national geographic approach preferred to allow local solutions to
correct data. For example, one commenter suggested having a local or
regional option because PM instruments are impacted by, and respond
differently to, a variety of local factors, including relative
humidity, temperature, concentration levels, and particle composition.
The EPA agrees that there are challenges in the response of PM FEMs to
a variety of local factors; however, this can be true of many methods
and are not specific to PM FEMs and, therefore, does not provide a
reason to reject this approach in this instance. Another commenter
stated that the proposed national correction factor is a ``flawed
concept,'' suggesting that it is ``widely understood throughout the
monitoring community that monitors perform best with a local correction
factor.'' This commentor offered no record or citation supporting this
point. The EPA counters that while monitoring agencies may
statistically correct data from a PM continuous monitor for AQI
purposes (40 CFR part 58, appendix G), there are both examples of well
performing statistically corrected PM continuous monitors being used
for AQI purposes; however, without proper attention and updates, there
are also examples of poorly performing ones. Finally, another commenter
believes that a national correction factor cannot possibly incorporate
data to represent all the scenarios across the nation that have an
impact on monitor performance and data quality. Although the EPA agrees
that there are a variety of local scenarios that could affect monitor
performance, the overall benefits of having nationally consistent
measurement of PM concentrations and national calibration of data
outweigh the potential advantages of locally specific calibrations.
Several commenters also disagreed with using local and regional
calibrations of data, including some monitoring agencies that asserted
being unable to reinvest in the operation of FRMs that would be
required to locally calibrate their own PM FEMs. Further, every
approved PM FEM method designated today is effectively calibrated
through demonstration of field testing in the areas in which it was
required to be tested (40 CFR 53.35(b)(1)). Moreover, the EPA proposed
to require instrument manufacturers to demonstrate that they can
improve the number of sites meeting bias MQOs by initiating a
recalibration of an FEM. Thus, the use of a national set of sites where
the methods are operated is essentially a fine-tuning of the PM FEMs
performance across all sites where it is used.
After considering all the comments received, the EPA believes it is
appropriate to finalize as proposed with a representative set
geographic locations at SLT sites to calibrate PM FEMs. Identification
of such sites would be made by the applicant of the planned updated
calibration, subject to EPA approval, and submitted to the EPA in
accordance with the requirements and application instructions in 40 CFR
part 58, appendix C, sections 2.2 and 2.7. The EPA encourages early
communication between an applicant seeking a method update and the EPA
to facilitate the most appropriate sites are included in any updated
application of the methods calibration.
The EPA proposed that instrument companies may, but are not
required to, check for and exclude any potential outliers that may
exist in the validated State, local, and Tribal agency network data
available from AQS that would be used to establish new factory
calibrations. The EPA received two comments regarding potential outlier
approaches. One commenter disagreed with the proposed approach and
instead recommended the use of all ``validated'' data, because how
these instruments behave under normal operating ranges may be just as
important as how they behave when monitoring conditions are low or
elevated. The EPA acknowledges this point; however, the proposal on
outliers allows flexibility in using standard outlier tests if needed
to include or exclude such data as part of the calibration process.
Ultimately, the true test of success for an updated method calibration
will be that a higher number of sites are meeting bias MQOs in the
areas in which the method is used, which will include all routine valid
data including any potential outliers. Another commenter asserted
concerns with the ability of instrument manufacturers to analyze data
within individual monitoring agencies. The EPA disagrees with the
commenter because decisions whether to include or exclude outliers
should be flexible and made on a case-by-case basis. Moreover, the
expected substantially larger dataset from routinely operated
collocated FRMs and FEMs compared to what was
[[Page 16354]]
used in the original FEM designation testing (Sec. 53.35 Test
procedure for Class II and Class III methods for PM2.5 and
PM10-2.5) will minimize the effect of any potential
outliers.
In contrast to these two comments, the EPA received many comments
supportive of the proposed outlier approach overall. Therefore, the EPA
is finalizing this part of the proposal that instrument companies may,
but are not required to, check for and exclude any potential outliers
that may exist in the validated State, local, and Tribal agency network
data available from AQS that would be used to establish new factory
calibrations.
Several commenters offered input on statistical criteria and
initial testing requirements for approval of candidate PM FEMs and the
role of instrument manufacturers in this process. The EPA did not
propose any changes related to these issues; however, these comments
have been considered below.
One commenter suggested that data quality objectives, bias, and
precision estimators for different monitoring methods should be based
on averages at both national and regional levels for purposes of
comparison. Another commenter asked to strengthen the criteria for
Class 3 Equivalency standards for candidate PM instrumentation. On
testing requirements, one commenter recommended that the EPA consider
updating the 40 CFR part 53 process for approving FEMs so that the
testing process more closely reflects the regulatory deployment and
data handling that generates NAAQS-comparable data. Another commenter
asked that the results from ``summer'' and ``winter'' field evaluations
not be averaged together because it allows agencies to minimize the
error of biased instruments by averaging poor results with data often
biased in the other direction. The same commenter also recommended that
candidate instruments data sets should not be averaged together as is
done currently where data from triplicate instruments are averaged for
each day. Another commenter asked that the EPA require FEM field
comparability tests in the northwest (e.g., in EPA Region 10) in areas
where particulate derived from biomass predominates to ensure that
certified instruments will perform reliably in regions influenced by
these sources. Related to the different measurement principles and the
instrument companies' role in PM FEMs, one commenter noted that FEMs
may never align perfectly with the FRMs due to the use of different
measurement principles. Another commenter asked that manufacturers of
FEM instruments be held accountable for ensuring that they continue to
meet FEM criteria, whether through calibration updates and/or follow-up
evaluations. Another commenter suggested that instrument manufacturers
should be required to further evaluate the FEM monitoring data at
defined intervals including, but not limited to, the 2-year and 5-year
approval anniversaries.
The EPA did not propose to make modifications to the statistical
criteria or testing requirements; however, we did solicit comment on
any alternatives that would lead to more sites meeting the bias MQO
with automated FEMs, especially for those sites that are near the level
of the primary annual PM2.5 NAAQS as proposed (88 FR 5672-
73, January 27, 2023). While the comments requesting that the
statistical criteria be strengthened may have merit, doing so would not
address the large inventory of already deployed PM FEMs used throughout
the country. Also, without performing a detailed Data Quality Objective
(DQO) design process, it is unclear how changing one or more
statistical criteria would help improve the number of sites meeting the
bias MQO now or in the future. Similarly, while the comments asking for
changes to the locations of testing may also have merit, the EPA
believes this could be a deterrent for instrument manufactures to seek
additional improvements since more testing would be required, at least
for candidate methods. Regarding the comment on the different
measurement principles, the EPA concurs that different measurement
principles may never align perfectly. Also, the EPA notes that the
Agency has longstanding goals for acceptable measurement uncertainty of
automated and manual PM2.5 methods in 40 CFR part 58,
appendix A, section 2.3.1.1. Therefore, while having different
measurement principles align is useful, meeting the goal for acceptable
measurement uncertainty is the objective.
Regarding the comments related to the instrument companies' role in
PM FEMs, the EPA notes that FEMs are already required to meet 40 CFR
53.9, ``Conditions of designation.'' Specifically, 40 CFR 53.9(c)
requires that, ``Any analyzer, PM10 sampler,
PM2.5 sampler, or PM10-2.5 sampler offered for
sale as part of an FRM or FEM shall function within the limits of the
performance specifications referred to in Sec. 53.20(a), Sec.
53.30(a), Sec. 53.35, Sec. 53.50, or Sec. 53.60, as applicable, for
at least 1 year after delivery and acceptance when maintained and
operated in accordance with the manual referred to in Sec.
53.4(b)(3).'' The EPA does not have the authority to require instrument
manufacturers to further evaluate the FEM monitoring data at defined
intervals, including but not limited to the 2-year and 5-year approval
anniversaries, as one commenter suggested.
In addition to these few recommendations, the EPA received many
comments supportive of the proposal that valid State, local, and Tribal
air monitoring data from FRMs generated in routine networks and
submitted to the EPA may be used to improve the PM concentration
measurement performance of approved FEMs; therefore, consistent with
the proposal we are not finalizing any updates to the statistical
criteria, testing requirements, or requirements on instrument
manufactures as proposed.
The EPA proposed that any new factory calibration should be
developed using data from at least 2 years and tested on a separate
year(s) of data (88 FR 5672, January 27, 2023). Comments on this part
of the proposal were generally supportive. One commenter requested that
at least a 3-year dataset, rather than the proposed 2 years, be used
for a representative design value comparison of the FEM and FRM
datasets to be evaluated. Another commenter pointed out that as large a
data set as possible should be used, but EPA should not limit it to
only data collected by instruments that have operated for more than 2
years.
In response to these comments, the EPA notes the broad support for
the proposal as written. Also, the EPA notes that the 2-year period for
using data to develop a factory calibration is a minimum, and that more
years may be used as appropriate. Therefore, the EPA is finalizing its
approach that any new factory calibration should be developed using
data from at least 2 years and tested on a separate year(s) of data as
proposed.
The EPA proposed several aspects of the FEM calibration on which we
did not receive specific comments, including a provision that FEM
methods should be evaluated by monitoring agencies as part of routine
data assessments, such as during certification of data and 5-year
assessments; the fact that the EPA recognizes only data from existing
operating sites are available for use in factory calibrations; and
recognition that an updated factory calibration does not have to work
with the original field study data submitted that led to the
designation as an FEM. With the broad general support from commenters
summarized above, the EPA is finalizing each of these
[[Page 16355]]
individual aspects of the FEM calibration as proposed.
In the proposal, the EPA identified that we should expect a lag
between the date when an already designated method is approved with a
new factory calibration as an updated method by the EPA and when it can
be implemented in the field. The EPA solicited comment on how to
approach the data produced during this lag. Commenters provided input
not only on how to address data during the lag, but also regarding how
to address data already collected prior to a method update that has the
potential to be used in regulatory decision making, particularly where
such collected data do not meet the bias MQO. In response to this
solicitation of comment, there was a consistent recommendation that
calibrations of data associated with method updates should be applied
to all relevant PM data prior to the EPA using it for designations
under a final NAAQS.
While the EPA appreciates these comments and recognizes their
support for retroactive data correction, at this time and following
this final rule, monitoring agencies should continue to report PM FEM
data as measured. This component of this final rule is focused only on
revising 40 CFR part 53, appendix C to implement an updated calibration
for approved PM FEMs. The issue of how prior and future monitoring data
will be used in the implementation of this NAAQS, such as for
designations, and for air quality regulatory programs is outside the
scope of this rulemaking and will therefore be addressed by the EPA in
a subsequent relevant action or actions.
The EPA received comments on whether updates to PM FEM methods
should be required to be implemented or there would flexibility in when
and if a monitoring agency implemented them. The commenters asked that
EPA be flexible in allowing the use of updated method correction
factors intended to improve the data comparability between the FRMs and
FEMs.
In most cases, the EPA expects that updating the FEMs will result
in improved data quality and more sites meeting bias MQOs; however, the
EPA is not finalizing an update requirement in this action. Monitoring
agencies can assess their data and make decisions on an update based on
whether they are meeting the bias MQOs. Such decisions on whether or
not to update a method may efficiently be included in those agencies'
annual monitoring network plans under 40 CFR 58.10, ``Annual monitoring
network plan and periodic assessment,'' which are already subject to
EPA Regional office approval. In some circumstances, it is possible the
original PM FEM may be revised in a manner where only the updated
method has an active approved designation. In these cases, monitoring
agencies would need to address updating their PM FEM in a timely
manner.
The EPA solicited input on any alternative approaches that could
lead to more sites meeting the bias MQO with automated PM FEMs,
especially for those sites that are near the level of the primary
annual PM2.5 NAAQS as proposed to be revised in section II
above. A few commentors provided input on potential options for
alternative approaches and several others offered input on how a local
or regional calibration of an FEM could work. Among alternative
approaches, one commenter suggested that manufacturers of FEMs could
provide settings that would allow for adjustments to make FEM data more
``FRM-like.'' Another commenter suggested working with the
manufacturers of FEM equipment to diagnose the cause of the bias and
then to address it appropriately.
The EPA received several comments on how to implement a local or
regional calibration of FEMs. One commenter suggested that EPA could
allow for SLT agencies to adjust FEM data to be more ``FRM-like'' prior
to submitting data to AQS. Another commenter suggested using a rolling
3-month linear regression based on a comparison of FEM data to
PM2.5 levels measured by a 1-in-6-day FRM. Another commenter
recommended that the EPA allow the application of a correction factor
that is from an area with a similar climate and other conditions.
Another commenter suggested that, for metropolitan statistical areas
(MSAs) where the re-calibrated FEMs still do not meet equivalency
criteria, monitoring agencies should be able to use the rolling linear
regression technique to further calibrate the FEMs within an MSA.
Another commenter suggested that developing a simple linear regression
could establish the relationship between FEM data and FRM data and be
used to adjust the FEM data at each site where they are collocated.
Another commenter suggested that averaging the results within a MSA and
applying it on an MSA basis with the previous 2 years of data could
provide an adjustment method for sites without a collocated FRM.
Another commenter identified that a regional correction factor
potentially could improve instrument accuracy to biomass sources, which
are a large component of PM in many communities.
Among the alternative approaches suggested, having settings that
would allow for adjustments to make FEM data more ``FRM-like'' has
merit, but assuming this was within a PM FEM itself, it would need to
be separately incorporated into each make and model of FEM. If EPA were
to pursue this alternative approach, the suggestion could be
incorporated into a future regulatory action as a potential condition
of designation because, without having the opportunity to thoughtfully
consider how every step of such an approach would need to work,
including what such requirements would look like and how potential
settings adjustments would be made, it is not appropriate for the EPA
to require the availability of such settings now, nor would it address
the inventory of currently available PM FEMs already operating.
Regarding the suggestion that the EPA and SLTs should work with the
manufacturers of FEM equipment to diagnose the cause of any biases and
then to address them appropriately, the EPA supports this
recommendation, but does not believe a regulatory change is required to
allow the monitoring community (EPA and SLTs) to work with instrument
manufacturers in this way.
Regarding the several comments on how to implement a local or
regional calibration of FEMs, the EPA acknowledges the desire for this
flexibility but believes that any such provisions for local or regional
calibration of FEMs would need to be thoroughly thought out and
proposed for consideration across the monitoring community. While
several commenters support such an approach, the EPA also received
adverse comments on the potential for local and regional calibration of
PM FEMs instead of national. Most of the criticism of local and
regional calibration of PM FEMs centered on both the lack of existing
operating PM FRMs in commenters' networks and monitoring agencies'
inability to staff the higher number of operating FRMs that would have
to be collocated with PM FEMs to calibrate. Thus, the commenters that
oppose local and regional calibrations of data prefer to utilize the
national calibration of FEM data as proposed. Acknowledging all of
these viewpoints, the EPA believes that it would not be appropriate to
institute such an approach at this time. As discussed throughout this
section, this final rule, the EPA is embarking on a new national
approach to calibration of FEMs where valid State, local, and
[[Page 16356]]
Tribal air monitoring data from FRMs generated in routine networks and
submitted to the EPA may be used to improve the PM concentration
measurement performance of approved FEMs. The EPA and the community of
SLT monitoring agencies can further consider other solutions to
improving PM FEM methods, including local and regional scale
calibration of FEMs, in a future review of the PM NAAQS.
In summary, the EPA is finalizing its proposal to allow valid
State, local, and Tribal air monitoring data from PM FRMs and FEMs
generated in routine networks and submitted to the EPA to update
factory calibrations included as part of approved FEMs (40 CFR part 58,
appendix C, sections 2.2 and 2.7). This approach, which will typically
be initiated by instrument manufacturers but can also be spurred by
monitoring agencies, MJOs of monitoring agencies, and the EPA itself,
is to be implemented as a national solution in factory calibrations of
approved FEMs through a firmware update, subject to EPA approval. FEM
calibrations can apply to any PM FEM methods (i.e., PM10,
PM2.5, and PM10-2.5). As part of this process,
the EPA is finalizing that a range of data based on the most
representative concentrations up to all available concentrations may be
used in developing and testing a new factory calibration; that a
representative set of geographic locations can be used; that outliers
may be included or not included; that a new factory calibration should
be developed using data from at least 2 years and tested on a separate
year(s) of data; that updates to factory calibrations can occur as
often as needed and should be evaluated by monitoring agencies as part
of routine data assessments such as during certification of data and 5-
year assessments; that the EPA recognizes only data from existing
operating sites is available; and that an updated factory calibration
does not have to work with the original field study data submitted that
led to the designation as an FEM. The EPA is finalizing this approach
as proposed with the intention of having more sites meet the bias MQOs
with automated PM FEMs.
4. Revisions to the PM2.5 Monitoring Network Design Criteria
To Address At-Risk Communities
To enhance protection of air quality in communities subject to
disproportionate air pollution risk, particularly in light of the
proposed range for a revised primary annual PM2.5 standard,
the EPA proposed to modify the PM2.5 monitoring network
design criteria to include an environmental justice (EJ) factor that
accounts for proximity of at-risk populations (i.e., those identified
in the 2019 ISA and ISA Supplement as being at increased risk of
adverse health effects from PM2.5 exposures to sources of
concern), consistent with the statutory requirement that the NAAQS
protect the health of at-risk populations (88 FR 5673, January 27,
2023). Specifically, the EPA proposed to modify the existing
requirement at 40 CFR part 58, appendix D, section 4.7.1(b)(3)): ``For
areas with additional required SLAMS, a monitoring station is to be
sited in an area of poor air quality,'' to additionally address at-risk
communities with a focus on anticipated exposures from local sources of
emissions. The scientific evidence evaluated in the 2019 ISA and ISA
Supplement indicates that sub-populations at potentially greater risk
from PM2.5 exposures include children, lower socioeconomic
status (SES) \188\ populations, minority populations (particularly
Black populations), and people with certain preexisting diseases
(particularly cardiovascular disease and asthma). The EPA proposed that
communities with relatively higher proportions of sub-populations at
greater risk from PM2.5 exposure within the jurisdiction of
a State or local monitoring agency should be considered ``at-risk
communities'' for these purposes.
---------------------------------------------------------------------------
\188\ SES is a composite measure that includes metrics such as
income, occupation, and education, and can play a role in
populations' access to healthy environments and healthcare.
---------------------------------------------------------------------------
The PM2.5 network design criteria have led to a robust
national network of PM2.5 monitoring stations. These
monitoring stations are largely in Core-Based Statistical Areas (CBSAs)
\189\ across the country that include many PM2.5 monitoring
sites in at-risk communities. Many of the epidemiologic studies
evaluated in the 2019 ISA and ISA Supplement, including those that
provide evidence of disparities in PM2.5 exposure and health
risk in minority populations and low-SES populations, often use data
from these existing PM2.5 monitoring sites. However, we
anticipate that with the more protective annual NAAQS finalized in
section II above, characterizing localized air quality issues around
local emission sources may become even more important. The EPA believes
that adding a network design requirement to locate monitors in at-risk
communities will improve our characterization of exposures for at-risk
communities where localized air quality issues may contribute to air
pollution exposures. Requiring that PM2.5 monitoring
stations be sited in at-risk communities will allow other methods to be
operated alongside PM2.5 measurements to support multiple
monitoring objectives per 40 CFR part 58, appendix D, section 1.1. The
EPA believes that it is appropriate to formalize the monitoring
network's characterization of PM2.5 concentrations in
communities at increased risk to provide such areas with the level of
protection intended with the PM2.5 NAAQS. The addition of
this requirement will also lead to enhanced local data that will allow
air quality regulators help communities reduce exposures and inform
future implementation and reviews of the NAAQS.
---------------------------------------------------------------------------
\189\ Metropolitan and Micropolitan Statistical Areas are
collectively referred to as ``Core-Based Statistical Areas.''
Metropolitan statistical areas have at least one urbanized area of
50,000 or more population, plus adjacent territory that has a high
degree of social and economic integration with the core as measured
by commuting ties. Micropolitan statistical areas are a set of
statistical areas that have at least one urban cluster of at least
10,000 but less than 50,000 population, plus adjacent territory that
has a high degree of social and economic integration with the core
as measured by commuting ties.
---------------------------------------------------------------------------
The EPA received comments concerning the proposed requirement to
modify the PM2.5 monitoring network design criteria to
include an EJ factor that accounts for the proximity of populations at
increased risk of adverse health effects from PM2.5
exposures to sources of concern. Commenters included State, local, and
Tribal air agencies and multijurisdictional organizations (MJOs)
comprised of those agencies; industry and industry groups; other
Federal, State, and local government entities; public health, medical,
and environmental nongovernmental organizations (NGOs); and private
citizens. The EPA proposed to require that sites located in at-risk
communities (particularly those whose air quality is potentially
affected by local sources of concern) should nonetheless meet the
requirements to be considered representative of ``areawide'' air
quality as this is consistent with all other minimally required sites.
There were several other technical components of the proposed
requirement for which we asked for comment, including: how to identify
at-risk communities; the PM sources of concern important to consider;
the datasets that can be used to identify communities with high
exposures; the most useful measurement methods to collocate with
PM2.5 in at-risk communities; and the timeline to implement
any new or moved sites.
Overall, most commenters were very supportive of the EPA's proposed
modification to the PM2.5 monitoring
[[Page 16357]]
network design criteria to include an EJ factor that accounts for
proximity of populations at increased risk of adverse health effects
from PM2.5 exposures to sources of concern. A few commenters
offered detailed supporting comments. For example, one commenter
recommended targeting investment in regulatory monitors in EJ
communities, opining that there is presently a lack of equitable
distribution of these monitors in low-income and minority communities.
Another commenter supports the inclusion of an EJ factor in
PM2.5 monitoring network design criteria as a means to
assess whether disparities in exposure are reduced in the future. The
EPA appreciates the support for the proposed requirement and
acknowledges the desirability of a goal to assess if disparities in
exposure are reduced in the future as a result of these monitoring
efforts.
Some commenters were generally supportive of the proposed
requirement but suggested that the EPA should recast the approach in a
more specific way or offered additional examples of sources of concern.
For example, one commenter stated that PM2.5 emissions from
residential and commercial wood burning result in localized hotspots
that are often not revealed by community air monitoring. Another
commenter asked that the EPA adopt a strategy to monitor EJ communities
near both larger well-known point sources of PM2.5 and along
traffic corridors as well as smaller sources that, when taken together,
may create a large amount of emissions and health harms in the area.
Another commenter stated that the national network of monitors operated
by the EPA captures data used for generalized modeling, but overall
monitoring is not as granular as one would expect, especially in urban
areas. For instance, the commenter suggested that EPA could monitor
suspected ``hot spots'' (e.g., residential development adjacent to
highways and active construction sites) to better manage and mitigate
PM2.5 pollution at their sites of origin, and that more
extensive and granular monitoring data would also facilitate essential
research and inform future evaluations and adjustments of the NAAQS.
The EPA acknowledges these comments identifying other sources of
concern, and we address these and other potential sources of concern
below.
Among adverse comments, a few commenters stated that ``at-risk
communities'' is not well defined. The EPA disagrees and directs those
commenters to the numerous places where this definition is covered,
including in Section II.B.2 of the proposal where we explained the term
related to a variety of at-risk populations (88 FR 5591-92, January 27,
2023) as well as section 12.5 of the 2019 ISA (U.S. EPA, 2019a) and
section 3.3.3 of the ISA Supplement (U.S. EPA, 2022a). Other commenters
oppose the addition of the proposed monitoring because they feel it
would reduce flexibility for agencies in deciding where they should
site monitors, advocating that monitoring agencies should be afforded
maximum flexibility to identify where to site monitors for at-risk
areas. Because the EPA recognizes the challenges cited by these
commenters related to establishing new ambient air monitoring stations,
the EPA is finalizing the modified requirement on PM2.5
monitoring network design criteria intended to address at-risk
communities that allows flexibility regarding which EJ communities
should be monitored. Finally, one commenter asked that the EPA clarify
a specific metric to judge how to site monitors in at-risk communities.
Instead, the EPA believes it is appropriate for agencies to recommend
what they believe to be the most important things to consider for their
sites to meet the PM2.5 network design requirements and,
thus, applying a new metric could take away from local priorities for
at-risk communities.
A few commenters asked that the EPA require more monitoring than
proposed. One commenter stated that it would be more beneficial to
overburdened communities if air monitoring were required in all at-risk
communities. A few commenters asked that EPA require additional
monitoring for attainment of PM2.5 NAAQS in EJ communities.
In response to these comments, the EPA supports the SLT agencies'
initiatives to conduct additional monitoring beyond the minimum
monitoring requirements and network design criteria. In addition, the
EPA supports agencies' use of alternative datasets such as sensors and
sensors networks, satellites, and other non-regulatory monitoring where
appropriate for non-regulatory data uses. The EPA notes that many
monitoring agencies already operate more monitoring sites than are
minimally required, and we expect this to continue as agencies consider
siting monitors in at-risk communities.
However, the EPA also received substantial concerns from monitoring
agencies about their resource constraints, including staffing to
support any potential new monitoring. The EPA also notes that the
existing and robust network of almost 1,000 PM2.5 sites
nationally is designed to continue to protect all populations at the
level of the NAAQS discussed in section II of this final action by
always having at least one site in the area of expected maximum
concentration for each CBSA where monitoring is required. As a result
of the revisions to the annual PM2.5 NAAQS being finalized
in this action, a small number of new monitoring sites will also be
required under EPA's current minimum monitoring requirements. With the
monitoring network design changes finalized in this rule, many of these
existing and new sites will form an important sub-component of the
PM2.5 network by better characterizing air quality in at-
risk communities, particularly with respect to sources of concern.
The EPA concludes that the requirements in this final rule for
siting of monitoring in at-risk communities will meaningfully improve
the PM2.5 monitoring network and its characterization of air
quality in at-risk communities, without placing substantial new
resource burdens on States and their monitoring agencies that would be
associated with requirements for additional monitoring sites.
Therefore, the EPA is finalizing this part of the proposed action
without requiring additional monitoring sites beyond what would be
associated with the revised annual PM2.5 NAAQS described in
section II as they pertain to the minimum requirements associated with
Table D-5 of Appendix D to Part 58--PM2.5 Minimum Monitoring
Requirements.
A few commenters asked that the EPA enhance monitoring in smaller
cities and rural areas. One commenter asked for the EPA to extend the
proposed monitoring network to Micropolitan Statistical Areas with
populations of 10,000-50,000 and to rural areas. Another commenter
pointed out that current air quality monitoring networks focus on urban
and densely populated areas; therefore, rural areas are often not
captured in this existing monitoring infrastructure, despite well-
documented examples of high PM concentration in rural communities. The
commenter believes this results in inadequate assessment of air
pollution exposures for a substantial segment of the U.S. population.
The EPA disagrees that there needs to be additional requirements for
small CBSA's and rural areas. Regarding these comments, the EPA points
out that we have a long-standing requirement for each State to monitor
at background and transport sites (40 CFR part 58, appendix D, section
4.7.3--Requirement for PM2.5 Background and Transport
Sites). Also, if an agency deems it appropriate to do so, monitoring
coverage of rural areas can be accomplished with other tools
[[Page 16358]]
such as sensors and sensors networks, satellites, and other non-
regulatory monitoring. Although there may be short-term high exposures
in rural areas, there is no evidence that long-term averages are higher
in rural areas compared to urban areas with significantly higher
density of populations and emissions. For smaller cities or rural areas
that may have concentrations near the level of the PM2.5
NAAQS finalized in section II above, monitoring agencies are encouraged
to monitor and address emissions as appropriate.
Some commenters disagree that the proposed revision to the
PM2.5 monitoring network design criteria to address at-risk
communities is needed. One commenter stated that including an EJ factor
is not necessary because the current network is designed to protect all
citizens. Another commenter stated that EJ factors could be cumbersome
to implement. Another commenter asserted the proposal to add SLAMS in
at-risk communities with higher PM2.5 concentrations might
create more granular data and provide for a greater margin of safety
for those communities and monitors in such a way that data from those
areas could misrepresent the larger area represented by the network. In
response to the comment on the current network protecting all citizens,
the EPA agrees that by measuring in the community with the highest
concentration of PM2.5 we protect other citizens; however,
as stated in the proposal, the EPA believes that adding a requirement
for sites with an EJ factor near sources of concern will enhance the
overall network to the benefit of all citizens. Also, we anticipate
that with the more protective annual NAAQS finalized in section II
above, characterizing localized air quality issues will become even
more important around local emission sources. As for EJ factors being
cumbersome to implement, the EPA disagrees because there are many such
locations already operating successfully in the current network.
Regarding the comment that sites in at-risk communities may
misrepresent the larger area represented by a particular network, the
EPA notes that pursuant to 40 CFR part 58, minimally required sites in
a given network are to represent area-wide air quality; therefore,
sites in at-risk communities, by definition, would be representative of
the communities within the network in which they are sited for the
level of protection intended under the annual PM2.5 NAAQS.
In the proposal, the EPA identified that, in light of the evidence
of increased risk to at-risk communities, it would be appropriate to
better characterize exposures for communities in proximity to local
sources of concern (88 FR 5673-76, January 27, 2023). Thus, the EPA
proposed that enhanced networks should include representation of at-
risk communities living near emission sources of concern (e.g., major
ports, rail yards, airports, industrial areas, or major transportation
corridors). The EPA requested comment on the types of sources of
concern most important to consider. In addition to supporting the types
of sources the EPA identified in the proposal, commenters also
identified several additional localized sources such as railroads,
stationary sources, transportation facilities, and communities with
high numbers of wood stoves.
A few commenters suggested the inclusion of sources that are often
considered line and/or area sources, e.g., traffic corridors and
emissions from federally regulated facilities, military installations,
and national forests. Commenters also identified other sources usually
associated with long-range transport such as smoke from wildfire and
prescribed fires and long-distance transport of PM, for example from
Saharan dust and other international transport. As explained in the
proposal, the site with the highest expected PM2.5 is
already required to have a monitor by our long-standing requirement
that monitors be placed ``. . . in the area of expected maximum
concentration'' (Sec. 58.1 and appendix D, section 4.7.1(b)(1)). The
EPA expects that both sites with the expected maximum concentration and
sites specifically placed in at-risk communities would be impacted by
any long-range transport in the area. Therefore, the EPA believes any
emphasis on the sources of concern should prioritize localized sources,
including point, area, and line sources of concern impacting the at-
risk community of interest. Therefore, based upon the comments, the EPA
is finalizing a broader example list of sources of concern to include
localized sources such as point sources and transportation facilities,
since these are the most commonly expected additional sources of
concern. In response to the other sources of concern suggested by
commenters, the EPA notes that while it has provided examples, the
siting of monitors in EJ communities would not be limited to these
examples. Thus, the revised set of examples would include ``a major
industrial area, point source(s), port, rail yard, airport, or other
transportation facility or corridor.'' In finalizing this modified list
of examples, the EPA is not looking to prioritize one type of source
category over another; rather, we intend to further illustrate the
types of localized sources of pollution that might impact at-risk
communities such that the siting of monitors nearby may be appropriate.
One commenter noted that the proposal may have unintentionally
taken out the requirement related to specific design criteria for
PM2.5 in 40 CFR part 58, appendix D, 4.7.1(b)(3) that, for
an area with a requirement for an additional SLAMS monitor, it should
``be sited in an area of poor air quality.'' Thus, the language as
proposed neither requires that such monitors be sited in areas of poor
air quality, nor does it require that the monitor be sited in an area
that is anticipated to experience poor air quality from unspecified
(and thus potentially relatively insignificant) sources in the area.
The EPA agrees that this was not our intention; the EPA wants to
protect populations in at-risk communities by ensuring they are
protected by the NAAQS when there are sources of concern that may be
impacting them (i.e., not insignificant sources). Thus, the EPA is
reinstating this requirement in the network design language and
combining it with the examples of the types of localized sources of
concern: ``For areas with additional required SLAMS, a monitoring
station is to be sited in an at-risk community with poor air quality,
particularly where there are anticipated effects from sources in the
area (e.g., a major industrial area, point source(s), port, rail yard,
airport, or other transportation facility or corridor).''
To ensure minimally required monitoring sites appropriately
represent exposures in at-risk communities, the EPA proposed that sites
represent ``area-wide'' air quality near local sources of concern (88
FR 5674, January 27, 2023). Sites representing ``area-wide'' air
quality are those monitors sited at neighborhood, urban, and regional
scales, as well as those monitors sited at either micro- or middle-
scale that are identified as being representative of many such
locations in the same Metropolitan Statistical Area (MSA).\190\ Most
existing--as well as new or moved sites--are expected to be
neighborhood-scale, which means that the monitoring stations would
typically represent conditions throughout some reasonably homogeneous
urban sub-region with dimensions of a few kilometers per part 58,
appendix D, section 4.7.1(c)(3). Additionally, as described in Sec.
58.30,
[[Page 16359]]
sites representing ``area-wide'' air quality have a long-standing
applicability to both the annual and 24-hour PM2.5 NAAQS.
Our proposed requirement for siting monitors in communities
representing ``area-wide'' air quality is consistent with other network
design objectives pursuant to which we seek to have monitors located
where people live, work, and play.
---------------------------------------------------------------------------
\190\ MSA means a CBSA associated with at least one urbanized
area of 50,000 population or greater. The central-county, plus
adjacent counties with a high degree of integration, comprise the
area.
---------------------------------------------------------------------------
The EPA received a few comments on its proposed requirement that
minimally required sites represent ``area-wide'' air quality. One
commenter stated that the inclusion of a provision for EJ would narrow
the location of monitors to certain communities that may not best
represent ``areawide'' air quality. Another commenter asked the EPA to
consider removing requirements that sites be area-wide, since 24-hour
and annual averaging times would miss short, elevated pollution events.
A couple commenters had concerns with the difference in the scale of
representation between EJ monitors using small scale and other NAAQS
monitors using area-wide scale, in that area-wide scale would not
protect those most at risk. However, another commenter agreed with the
EPA that sites representing at-risk communities should represent area-
wide air quality. In addition to these comments, the EPA received many
comments with support for its proposed modifications to the network
design criteria as whole.
Regarding whether narrowing the location to certain communities may
not best represent ``area-wide'' air quality, the EPA notes that sites
are either identified as being area-wide or not; the EPA did not
suggest it was seeking a best ``area-wide'' location. In response to
the comment that area-wide site may miss short, elevated pollution
events, the EPA is aware that there can be local, short-term spikes in
PM2.5 concentrations. However, the network design criteria
associated with minimally required sites is applicable to both the
annual and 24-hour PM2.5 NAAQS, and the EPA believes it is
appropriate to continue to ensure all minimally required sites have the
most utility and remain applicable to both forms of the
PM2.5 NAAQS. The identification of unique micro- and middle-
scale sites was directed at discretionary efforts of any monitoring
agency, with the recognition that such sites, (i.e., relatively unique
micro-scale, or localized hot spot, or unique middle-scale impact
sites), are not applicable to the annual NAAQS as described in Sec.
58.30--Special consideration for data comparison to the NAAQS.
After considering all the comments on this topic, the EPA is
finalizing this part of the modification to the network design criteria
to maintain, consistent with our long-standing network design criteria,
that all minimally required sites are to represent area-wide air
quality.
In addition to using data from the robust network of almost 1,000
PM2.5 sites for NAAQS and AQI purposes, having a stable
network of long-term sites is especially valuable to examine trends and
to inform long-term health and epidemiology studies that support
reviews of the PM NAAQS. Therefore, while we proposed to add a
PM2.5 network design criterion to address at-risk
communities, many sites are likely already in valuable locations
meeting one of the existing network design criteria (i.e., being in an
area-wide area of expected maximum concentration or collocated with
near-road sites) and supporting multiple monitoring objectives. Also,
in many communities, there may already be sites meeting the network
design criterion we proposed for at-risk communities. Thus,
acknowledging the value of having long-term data from a consistent set
of network sites, the EPA believes that moving sites should be
minimized, especially in MSAs with a small number of sites. However,
because a small number of new sites are expected to be required due to
the existing minimum monitoring requirements (40 CFR part 58, appendix
D, Table D-5) \191\ and the revised primary annual PM2.5
NAAQS detailed in section II, and because sites occasionally have to be
moved--due to, for example, loss of access to a site or a site no
longer meeting siting criteria--the EPA believes it is appropriate to
prioritize establishing sites in at-risk communities near sources of
concern, whenever new sites are established, whether because it is a
new site or a replacement for a prior site that must be moved. The EPA
accordingly proposed that annual monitoring network plans (40 CFR
58.10(a)(1)) and 5-year assessments (40 CFR 58.10(d)) that include any
of the few new sites that will be required include a commitment to
examine the ability of existing and proposed sites to support air
quality characterization for areas with at-risk populations in the
community and the objective discussed herein.
---------------------------------------------------------------------------
\191\ Gantt, B. (2022). Analyses of Minimally Required
PM2.5 Sites Under Alternative NAAQS. Memorandum to the
Rulemaking Docket for the Review of the National Ambient Air Quality
Standards for Particulate Matter (EPA-HQ-OAR-2015-0072). Available
at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
In the proposal, the EPA identified that assessing and prioritizing
at-risk communities for monitoring can be accomplished through several
approaches (88 FR 5675). The most critical aspect of prioritizing which
communities to monitor is their representation of the at-risk
populations described earlier in this section. The other major
consideration is whether the community is near a source or sources of
concern. While many CBSAs have one or more sources of concern described
above, some CBSAs will not have a quantity of emissions from sources of
concern that result in an elevated level of measured PM2.5
concentrations in surrounding communities. The siting criteria to be
``in the area of expected maximum concentration,'' Sec. 58.1 &
appendix D, section 4.7.1(b)(1) ensures there is a monitoring site in
the community with the highest exposure in each CBSA with a monitoring
requirement. Some CBSAs may also have a requirement to collocate a
PM2.5 monitor at a near-road NO2 station.
Therefore, the EPA believes that for cases where an additional
PM2.5 site is required, we should include a criterion that
the site be in an at-risk community when there are no sources of
concern identified in that CBSA, or such sources do exist but are not
expected to lead to elevated levels of measured PM2.5
concentrations.
In its proposal, the EPA highlighted that tools such as the EPA's
EJSCREEN \192\ are available to identify the at-risk communities
intended for monitoring as part of the proposed revision to the
PM2.5 network design criteria (88 FR 5675-76, January 27,
2023). The EPA solicited comment on other tools and/or datasets that
can be utilized to identify at-risk communities. In addition to support
for using EJSCREEN, commenters identified several other options to
identify at-risk communities intended for monitoring as part of the
proposed revision to the PM2.5 network design criteria.
Among similar tools, one commentor suggesting using
CalEnviroScreen.\193\ Commenters also identified different options for
models including InMAP,\194\ satellite-derived models that can be
employed to help identify EJ communities, and hybrid models. A few
commenters also suggested using sensors and sensor networks such as the
BlueSky \195\ and PurpleAir \196\ sensors.
---------------------------------------------------------------------------
\192\ See: https://www.epa.gov/ejscreen.
\193\ See: https://oehha.ca.gov/calenviroscreen.
\194\ See: https://inmap.run/#home.
\195\ Mention of commercial names does not constitute EPA
endorsement.
\196\ Mention of commercial names does not constitute EPA
endorsement.
---------------------------------------------------------------------------
The EPA supports the use of other State and local tools designed to
help identify the at-risk communities that
[[Page 16360]]
should be monitored to meet the revised network design criteria. The
EPA additionally agrees with commenters that the use of models as well
as sensors and sensor networks may be appropriate and helpful in
identifying the most appropriate at-risk communities in which to locate
monitors.
For at-risk communities, monitoring agencies need data that can
best inform where there may be elevated levels of exposures from
sources of concern. While we use FRMs and FEMs to determine compliance
with the NAAQS, data from these methods will only be available at
existing sites. However, there are several additional datasets
available that may be useful in evaluating the potential for elevated
levels of exposure to communities near sources of concern. In the
proposal, EPA identified potential non-regulatory monitoring datasets
such as CSN, IMPROVE, and AQI non-regulatory PM2.5
continuous monitors; modeling data that utilizes emission inventory and
meteorological data; emerging sensor networks such as those that
comprise EPA and the USFS's Fire and Smoke Map; \197\ and satellites
that measure radiance and, with computational algorithms, can be used
to estimate PM2.5 from aerosol optical depth (AOD) (88 FR
5675-76, January 27, 2023). The EPA solicited comment on datasets most
useful to identify communities with high exposures for PM2.5
NAAQS (i.e., annual or 24-hour). In addition to providing information
about datasets that can inform the NAAQS comparison, commenters
additionally identified several types of datasets that may be useful to
identify where there may be elevated levels of exposures from sources
of concern. These datasets include satellite measurements, sensors, and
sensor network data, which may all be useful to find hot spots in
communities. Commenters also identified EJScreen and CalEnviroScreen,
which are screening and mapping tools that utilize several datasets.
Another commenter stated that to better understand exposure differences
in disadvantaged communities, shorter measurement intervals should be
measured and reported.
---------------------------------------------------------------------------
\197\ See: https://fire.airnow.gov/.
---------------------------------------------------------------------------
In considering the datasets identified in the proposal as well as
the ones commenters provided, the EPA believes all the datasets have
value to help inform where there may be elevated levels of exposures
from sources of concern. However, each of them may also have
limitations and, therefore, users should be careful not to rely solely
on one dataset versus another for all purposes. Fortunately, many of
the available datasets are becoming easier to work with and more
accessible, which will allow interested parties and monitoring agencies
the opportunity to efficiently review the datasets and determine best
applicability. For all of these reasons, the EPA is not finalizing a
requirement to use a specific dataset or tool to identify at risk
communities; however, whatever datasets a monitoring agency elects to
use, its plan to use such data for purposes of meeting the network
design requirements will be subject to EPA approval as part of the 40
CFR 58.10 annual monitoring network plan. Regarding the comment
recommending shorter measurement intervals in measuring and reporting
data to better understand exposure differences in disadvantaged
communities, the EPA agrees and generally supports use of continuous
methods. While we generally support use of continuous methods, approved
filter-based technologies and methods also provide valuable air quality
information. Therefore, the EPA is not requiring the use of automated
continuous methods beyond what is already required in 40 CFR part 58,
appendix D, section 4.7.2--Requirement for Continuous PM2.5
Monitoring.
The monitoring methods appropriate for use at required
PM2.5 sites in at-risk communities are FRMs and automated
continuous FEMs (88 FR 5675-76, January 27, 2023). These are the
methods eligible to compare to the PM2.5 NAAQS, which is the
primary objective for collecting this data. There are several other
monitoring objectives that would benefit from the use of automated
continuous FEMs. For example, having hourly data available from
automated continuous FEMs would allow sites to provide data in near-
real time to support forecasting and near real-time reporting of the
AQI. Automated continuous methods are also useful to support evaluation
of other methods such as low-cost sensors. When used in combination
with on-site wind speed and wind direction measurements, automated FEMs
can provide useful pollution roses, which help in identifying the
origin of emissions that affect a community. Additionally, when
collocated with continuous carbon methods such as an aethalometer,
automated FEMs can help identify potential local carbon sources
contributing to increased exposure in the community. While either FRMs
or automated FEMs may be used at a site for comparison to the
PM2.5 NAAQS, the EPA supports use of automated continuous
FEMs at sites in at-risk communities.
The EPA requested comment on the measurement methods most useful to
collocate with PM2.5 in at-risk communities (88 FR 5675-76,
January 27, 2023), and a few commenters provided input. One commenter
recommended that the EPA should employ supplemental technologies and
systems to increase coverage of the regulatory monitoring network and
obtain more complete data to further protect public health and address
environmental injustice in air pollution exposure. Another commenter
recommended that the EPA invest in community-led monitoring and mobile
air quality monitoring with a goal of recording block-level
variabilities in data. And another commenter cited the value of
community-deployed PM2.5 monitoring.
The EPA appreciates the comments provided on the measurement
methods most useful to collocate with PM2.5 monitoring sites
in at-risk communities. Because the use of methods beyond the required
PM2.5 FRMs or FEMs or other criteria pollutant measurements
meeting a NAAQS monitoring requirement is voluntary, the establishment
of PM2.5 NAAQS comparable sites in at-risk communities will
allow for collaboration at multiple levels. The EPA strongly encourages
such collaboration with impacted communities, and the measurement
methods discussed here should be considered for use as appropriate.
In the proposal, the EPA identified that, to meet the revised
network design criteria, there will be only a few new sites
required,\198\ plus any potentially moved sites in cases where an
existing site lease is lost or otherwise requires relocation (88 FR
5675-76, January 27, 2023). To handle these new or relocated sites, the
EPA proposed to build upon our existing regulatory process for
selecting and approving these sites under 40 CFR 58.10 (88 FR 5676,
January 27, 2023). In the proposal, we stated it would be appropriate
to provide at least 12 months from the effective date of the final rule
to allow monitoring agencies to initiate planning to implement these
measures by seeking input from communities and other interested parties
and considering whether to revise their PM2.5 networks
[[Page 16361]]
or explain how their existing networks meet the objectives of the
proposed modification to the network design criteria. Thus, the EPA
proposed that monitoring agencies should address their approach to the
question of whether any new or moved sites are needed and identify the
potential communities in which the agencies are considering adding
monitoring, if applicable, as well as identifying how they intend to
meet the revised criteria for PM2.5 network design to
address at-risk communities in the agencies' annual monitoring network
plans due to each applicable EPA Regional office no later than July 1,
2024 (see 40 CFR 58.10). Specifics on the resulting new or moved sites
for PM2.5 network design to address at-risk communities were
proposed to be detailed in the annual monitoring network plans due to
each applicable EPA Regional office no later than July 1, 2025 (40 CFR
58.10). The EPA proposed that any new or moved sites would be required
to be implemented and fully operational no later than 24 months from
the date of approval of a plan or January 1, 2027, whichever comes
first, but the EPA solicited comment on whether less time is needed
(e.g., 12 months from plan approval and/or January 1, 2026).
---------------------------------------------------------------------------
\198\ Gantt, B. (2022). Analyses of Minimally Required
PM2.5 Sites Under Alternative NAAQS. Memorandum to the
Rulemaking Docket for the Review of the National Ambient Air Quality
Standards for Particulate Matter (EPA-HQ-OAR-2015-0072). Available
at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
The EPA received a few comments on its proposed timeline for
monitoring agencies to identify, propose, and ultimately bring any new
or moved sites online. One commenter asked that the timeline give
states more time to start or move sites. A few commenters asked that
the EPA only require meeting a timeline for identifying whether any new
or moved sites are needed after the EPA has provided the monitoring
agencies with guidance on the priority of the potential at-risk
communities. One of those commenters further requests that the EPA
allow at least 24 months from the date of approval of a Sec. 58.10
monitoring plan identifying any relocation of monitoring sites or
establishment of new monitoring sites to implement any changes to the
network, citing the need for more time to work with local officials,
procure monitoring equipment, and contract for services, all of which
can cause significant delays in establishing a monitoring site. Another
commenter asked that the EPA remain attentive to the challenges that
States, and air agencies face regarding recruiting and retaining the
specialized staff needed to support their existing regulatory
monitoring networks and the capital resources needed to implement and
sustain new monitoring stations in areas that are clearly meeting the
existing PM NAAQS or any revised PM NAAQS. Another commenter stated
that the July 1, 2024, timeline for a network evaluation this complex
is insufficient, noting that they submit their draft annual monitoring
network plan for public review and comment in mid-April for 30 days.
Because the final plan is due July 1 and must include all comments and
responses and describe any changes based on those comments, the
timeline does not take these requirements into consideration by
allowing for the more extensive assessment of changes that may be
needed to meet the proposed new monitoring requirements. The commenter
stated that it would be appropriate to provide at least 12 months from
the effective date of this final rule for monitoring agencies to
initiate planning to implement these measures, seek input, consider
revisions to their PM2.5 networks, and explain how their
existing networks meets the objectives of the final rule. The commenter
notes that that SLT agencies should be provided a minimum of 18 months
after the final recommendation is published to add this information to
their Sec. 58.10 annual monitoring network plans. Another commenter
encourages the EPA to retain the proposed deadline for any newly
required monitoring stations in at-risk communities to be operational
(i.e., 24 months after the July 2025 network plan approval or January
1, 2027, whichever is earlier). While the need for this data is urgent,
the commenter stated that the process for procuring instrumentation,
securing leases, and building permits, and other logistics in
constructing new monitoring sites can take a significant amount of
time, some of which are outside of agencies' control.
As stated earlier, the EPA received strong support for our proposal
to modify the PM2.5 monitoring network design criteria to
include an EJ factor that accounts for proximity of populations at
increased risk of adverse health effects from PM2.5
exposures to sources of concern from a wide range of commenters. A few
commenters support the timeline proposed, a few others support starting
any new or moved sites sooner than proposed, while other commenters
asked for more time or offered conditions regarding how to establish an
appropriate timeline.
The EPA disagrees with the commenter that suggested the EPA should
only require agencies to meet a timeline to identify whether any new or
moved sites are needed after the EPA has provided the monitoring
agencies with guidance on the priority of the potential at-risk
communities, because the regulatory text provides all the guidance
required for agencies to begin this process. As we explained above, the
EPA does not anticipate that many new or moved sites will be required
based on the final rule because we think most sites are already in
suitable locations and long-term sites are highly valued. Also,
monitoring agencies have discretion to provide to the EPA their
recommendations regarding how they intend to meet the modifications to
the PM2.5 monitoring network design criteria to include an
EJ factor that accounts for proximity of populations at increased risk
of adverse health effects from PM2.5 exposures to sources of
concern. Overall, the EPA believes that having sites in the areas of
expected maximum concentrations will best ensure that all communities
are protected. Since there may be multiple choices for sites in EJ
areas near sources of concern, the EPA acknowledges that there may be
many locations that can meet the revised PM2.5 network
design criteria. While, as we explained earlier, we want such sites to
also be in areas of poor air quality, the sites in the area of maximum
concentration will ensure that all communities are protected, there can
be more flexibility afforded in the selection amongst at-risk
communities to meet the revised requirements, since any alternative at-
risk communities would already be protected.
The EPA considered both the concerns and support for the timeline
proposed and clarifies that the component of the proposed requirement
regarding the need to identify potential new sites or an intention to
move sites to be included in the annual monitoring network plan due to
EPA on July 1, 2024, would be satisfied with a statement of intent to
pursue a new site per the revised network design criteria and in
consideration of the minimum monitoring requirements. While monitoring
agencies may provide as much detail as they deem appropriate regarding
the revised PM2.5 network design criteria in their annual
monitoring network plans due on July 1, 2024, there is no expectation
that any details on site-specific information would be included at that
stage. We encourage agencies to provide their initial thinking on the
communities they are most interested in monitoring pursuant to the
revised network design criteria. Therefore, the EPA is finalizing the
timeline as proposed, including the provision that monitoring agencies
report their intention to add or move sites, where required, in their
annual monitoring network plans due to each applicable EPA Regional
office no later than July 1, 2024 (40 CFR 58.10). The
[[Page 16362]]
monitoring agencies will then provide specifics on any new or moved
sites for PM2.5 network design to address at-risk
communities in the annual monitoring network plans due to each
applicable EPA Regional office no later than July 1, 2025 (40 CFR
58.10). And any new or moved sites shall be implemented and fully
operational no later than 24 months from the date of approval of a
Sec. 58.10 plan, or January 1, 2027, whichever comes first.
In summary, the EPA is finalizing modifications to the
PM2.5 network design criteria to include an EJ factor to
address at-risk communities with a focus on exposures from sources of
concern in areas of poor air quality. While this modification to the
PM2.5 network design requires sites to be located in at-risk
communities, particularly those whose air quality is potentially
affected by local sources of concern, such sites must still meet the
requirement for being considered ``area-wide'' air quality. In
finalizing this modification to the PM2.5 network design
requirement, the EPA is making two changes in the final rule response
to the comments received. First, the EPA is broadening our examples of
``sources of concern'' to include localized sources such as point
sources and major transportation facilities or corridors. Second, the
EPA is reinstating ``poor air quality'' in our requirement for the
modified network design criteria, meaning the revised PM2.5
network design requirement now states: ``For areas with additional
required SLAMS, a monitoring station is to be sited in an at-risk
community with poor air quality, particularly where there are
anticipated effects from sources in the area (e.g., a major industrial
area, point source(s), port, rail yard, airport, or other
transportation facility or corridor).'' All other aspects of the
PM2.5 network design requirements are being finalized as
proposed.
5. Revisions to Probe and Monitoring Path Siting Criteria
The EPA proposed changes to monitoring requirements in the Appendix
E--Probe and Monitoring Path Siting Criteria for Ambient Air Quality
Monitoring (88 FR 5676-78, January 27, 2023). Since 2006, the EPA
finalized multiple rule revisions to establish siting requirements for
PM10-2.5 and O3 monitoring sites (71 FR 2748,
January 17, 2006), Near-Road NO2 monitoring sites (75 FR
6535, February 9, 2010), Near-Road CO monitoring sites (76 FR 54342,
August 31, 2011), and Near-Road PM2.5 monitoring sites (78
FR 3285, January 15, 2013). Through these previous revisions to the
regulatory text, some requirements were inadvertently omitted, and,
over time, the clarity of this appendix was reduced through those
omissions that, in a few instances, led to unintended and conflicting
regulatory requirements. The EPA proposed to reinstate portions of
previous Probe and Monitoring Path Siting Criteria Requirements from
previous rulemakings, where appropriate, to restore the original
intent.
The EPA only received a few comments on the proposed rulemaking
pertaining to the proposed changes regarding probe and monitoring path
siting criteria for ambient air quality monitoring, most of which were
supportive of the proposed revisions. One commenter noted that the
image for Figure E-1 in Appendix E to part 58 was distorted and of
extremely poor quality, rendering the text in places almost unreadable
(88 FR 5712, January 27, 2023). The EPA makes several references to
Figure E-1, which provides detailed information needed for assessing a
range of acceptable probe distances from roadways based on a monitor's
spatial scale. The commenter also stated that a higher quality image is
needed for the figure so that agencies can fully interpret the figure
to the extent that EPA requires. The EPA agrees with the commenter that
a higher quality image for Figure E-1 is important and needed. Based on
this comment, the EPA is finalizing the revision to Figure E-1 to
clearly communicate the requirements of appendix E.
The EPA is revising appendix E in its entirety as proposed (88 FR
5709-5717, January 27, 2023) for clarity and as described in detail
below.
a. Separate Section for Open Path Monitoring Requirements
The EPA proposed to relocate all open path monitor siting criteria
requirements to a separate section in appendix E from those
requirements for siting samplers and monitors that utilize probe inlets
(88 FR 5676, January 27, 2023). Separate sections for these distinct
monitoring method types allows the EPA to more clearly articulate
minimum technical siting requirements for each.
The EPA received one supportive comment to adopt this change and
received no adverse comments. Another commenter stated the regulatory
text of the proposal improves the clarity of the appendix but
encouraged the EPA to break the summary tables down further into more
manageable components (perhaps by pollutant). The commenter stated that
summary tables for the proposed appendix continue to be a ``jumbled
mess of regulatory requirements.'' The EPA agrees that the summary
tables E-3 and E-6 in the proposal could be improved further. Also, the
EPA found that footnote 3 of Table E-6 in the proposed rule was
incomplete and corrected this editorial error.
Therefore, the EPA is making editorial changes to both summary
tables E-3 and E-6 and finalizing the remainder of the language as
proposed with the open path monitor siting criteria requirements placed
into a separate section of the appendix.
b. Distance Precision for Spacing Offsets
The EPA proposed to require that when rounding is performed to
assess compliance with these siting requirements, the distance
measurements will be rounded such as to retain at least two significant
figures (88 FR 5676, January 27, 2023). The EPA proposed to communicate
this rounding requirement in the regulatory text using footnotes in the
tables of this appendix.
The EPA received two supportive comments and no adverse comments
regarding this proposed change. While supportive of the proposal, one
of the two supporting comments suggested it would be clearer if EPA
explicitly defined a decimal in the distance values and round to the
nearest tenths place for these assessments. The EPA disagrees with this
recommendation because in some cases it would be more restrictive and
burdensome than the proposed requirement that was intended to provide
both clarity and flexibility. Therefore, the EPA is finalizing the
language as proposed.
c. Summary Table of Probe Siting Criteria
The EPA proposed to provide additional specificity and flexibility
to the summary table for probe siting criteria by changing the ``>''
(greater than) symbols to ``>='' (greater than or equal to) symbols in
the summary table E-4 (88 FR 5676, January 27, 2023). Because one
commenter pointed out to the EPA that in the prior version of the rule
there was no table E-4, as a clerical matter, we have renumbered this
summary table to table E-3 in the final rule. This proposed minor
revision to the summary table more clearly expresses the EPA's intent
that the distance offsets provided in the summary tables in appendix E
are acceptable for NAAQS compliance monitoring.
The EPA received one comment supporting the proposal. The EPA
received no adverse comments. Because
[[Page 16363]]
one commenter pointed out to the EPA that in the prior version of the
rule there was no table E-4, as a clerical matter, we have renumbered
this summary table to table E-3 in the final rule. Therefore, the EPA
is updating the table numbering and otherwise finalizing the tables as
proposed.
d. Spacing From Minor Sources
The EPA proposed to clarify and provide flexibility regarding
siting monitors near minor sources by changing a requirement to a goal
(88 FR 5676-77, January 27, 2023). To accomplish this, the EPA proposed
to replace the ``must'' in the regulation with a ``should.'' While the
EPA proposed to change this requirement to a goal, the EPA reiterated
in the proposal that it recommends that sites with minor sources be
avoided whenever practicable and probe inlets should be spaced as far
from minor sources as possible when alternative monitoring stations are
not suitable.
The EPA received one comment supporting the proposed revision and
received no adverse comments. Therefore, the EPA is finalizing the
language as proposed.
e. Spacing From Obstructions and Trees
The EPA proposed to clarify and redefine that the minimum arc
required to be free of obstructions for a probe inlet or monitoring
path is 270-degrees and that probe inlets must be no closer than 10-
meters to the driplines of any trees (88 FR 5677, January 27, 2023).
These changes were proposed because of inconsistencies introduced into
the rule with the 2006 rulemaking. Both are discussed in more detail in
the following sections.
The majority of comments received were supportive of these proposed
siting amendments and clarifications. Two commenters were not
supportive of this proposal. One adverse comment focused on the
potential that site modifications would be required if the minimum arc
required to be free of obstructions for a probe inlet is 270-degrees.
The second adverse comment pertained to the proposal to clarify
distance requirements from tree driplines. The commenter stated they
would expect significant challenges in meeting the proposed 20-meter
tree dripline distance. This comment is not a substantive negative
comment because the 20-meter distance provided in the proposal is a
goal and not a requirement. As such, monitoring organizations should
not expect additional challenges in meeting the probe siting
requirements. One supportive commenter on the 270-degree minimum arc
proposal also requested that the EPA acknowledge that some cases exist
where monitoring is desired or necessary to protect the public health,
but siting criteria cannot be met.
Based on the only two negative comments received from monitoring
agencies or organizations, one of which was not substantive, the EPA
believes most sites already meet these proposed requirements related to
the arc and distance from dripline. However, the EPA also acknowledges
that there may be limited cases where this proposed revision may
require site modifications, and some sites may not be able to be
achieve the proposed siting requirements, even with modifications to
the site. For cases where long-term trend sites or monitors that
determine the design value for their area cannot reasonably meet these
regulatory siting requirements, the EPA encourages monitoring
organizations to work with their respective EPA Regional offices to
determine if a waiver from this siting criteria would be appropriate
under appendix E, section 10.
These siting requirements are discussed in more detail below in
sections VII.B.5.f and VII.B.5.h.
f. Reinstating Minimum 270-Degree Arc and Clarified 180-Degree Arc
The EPA proposed to correct identified inconsistencies in the 270-
degree requirement for unrestricted airflow to the probe inlet by
reinstating the requirement stated in appendix E, paragraph 4(b), and
to clarify that the continuous 180-degree minimum arc of unrestricted
airflow provision is reserved for monitors sited on the side of a
building or wall to comply with network design criteria requirements
specified in appendix D of part 58 (88 FR 5677, January 27, 2023).
The EPA received two comments regarding this proposal, with one
being supportive and one being negative. The adverse comment focused on
the potential that site modifications would be required if this
revision was made. The commenter supporting the proposal also requested
that the EPA acknowledge that some cases exist where monitoring is
desired or necessary to protect the public health, but siting criteria
cannot be met. The EPA agrees with both commenters and acknowledges
that there does exist limited cases where this proposal would require
site modifications and some sites may not be able to be achieve the
proposed siting requirement even with modifications to the site. For
these cases, and especially when long-term trend sites or monitors that
determine the design value for their area cannot reasonably meet these
regulatory siting requirements, the EPA encourages monitoring
organizations to work with their respective EPA Regional Offices to
determine if a waiver from this siting criteria is appropriate through
the provisions found in Section 10 of this appendix.
Based on the EPA only receiving a single negative comment regard
the 270-degree and 180-degree provisions the EPA thinks most sites
already meet these proposed requirements. Additionally, as stated
above, the EPA is also retaining waiver provisions from these siting
requirements for the remaining cases that can be exercised when
appropriate. Therefore, the EPA is finalizing the language as proposed.
g. Obstacles That Act as Obstructions
The EPA proposed to clarify the definitions of ``obstructions'' and
``obstacles'' in the regulatory text (88 FR 5677, January 27, 2023).
Stating that, ``[o]bstructions to the air flow of the probe inlet are
those obstacles that are horizontally closer than twice the vertical
distance the obstacle protrudes above the probe inlet and can be
reasonably thought to scavenge reactive gases or to restrict the
airflow for any pollutant,'' the EPA proposed to reiterate that the EPA
does not generally consider objects or obstacles such as flag poles or
site towers used for NOy convertors and meteorological
sensors, etc., to be deemed obstructions.
The EPA received one comment supporting the proposal and received
no adverse comments. Therefore, the EPA is finalizing the definitions
as proposed.
h. 10-Meter Tree Dripline Requirement
The EPA proposed to reconcile the conflicting requirements in 5(a)
and the prior table E-4 footnote 3 by clarifying that the probe inlet
must always be no closer than 10 meters to the tree dripline (88 FR
5677, January 27, 2023). The EPA also proposed to reinstate the goal
``that monitor probe inlets should be at least 20-meters from the
driplines of trees,'' a goal that was inadvertently omitted during
previous rule revisions. In addition, the EPA proposed to clarify that
if a tree or group of trees is considered an ``obstruction,'' section
4(a) will apply.
As described above, the majority of comments received were
supportive of the EPA proposed amendments and clarification, with two
commenters focused on the possibility that monitoring agencies may not
be able to meet the revised siting requirements. Specific to the
proposed dripline requirement, the EPA reiterates that the
[[Page 16364]]
20-meter tree dripline offset is not a requirement, but rather a goal.
Monitoring programs should as much as practicable attempt to meet this
20-meter tree dripline offset goal but are only required to be at least
10 meters removed from tree driplines. If these requirements cannot be
met, the EPA encourages monitoring organizations to contact their
respective EPA Regional offices to determine if a waiver from this
siting criteria would be appropriate under appendix E, section 10.
Another commenter recommended that the proposal should also include
an elevation specification. For instance, if a monitor is on the roof
of a shelter, a tree below that roof should not be considered an
obstruction no matter the distance to the dripline. The EPA considers
this scenario to occur in practice only rarely. The EPA agrees that
when the overall tree height is less than the height of the probe
inlet, the tree is not obstructing the airflow to the probe inlet.
However, a tree in such proximity to the probe inlet in many cases is
not likely to remain at a height lower than the probe inlet. The EPA
considers a scenario such as this to be best addressed in the waiver
provisions of this appendix due both to the rarity of this occurring as
well as the need for the EPA to periodically reassess whether tree
growth has adversely impacted the site conditions.
For these reasons, the EPA is finalizing the language as proposed.
i. Spacing Requirement for Microscale Monitoring
The EPA proposed to require that microscale sites for any pollutant
shall have no trees or shrubs blocking the line-of-sight fetch between
the monitor's probe inlet and the source under investigation (88 FR
5677, January 27, 2023). This proposed revision would bring consistency
between near-road monitoring stations and other microscale monitoring.
The EPA received one comment on this proposed requirement
expressing concerns regarding its practicality and legality. The
commenter stated agencies may at times want to site a monitor close to
a source, but the closest location will have trees in the line of sight
on private property. Additionally, in some cases, the trees may have
been planted for the purpose of reducing off-property emissions from a
source such as a Concentrated Animal Feeding Operation (CAFO). The
commenter further stated that the proposal mandates that State agencies
order the removal of trees from private property to collect valid data.
The EPA disagrees that the proposed requirement is impractical or
unlawful. The proposed requirement would not require, mandate, or
otherwise empower monitoring agencies to force the removal of trees on
private property. The EPA agrees with the commenter that trees may at
times be planted as part of control strategies to reduce offsite
emissions and thus protect the public, but the EPA disagrees with the
commenter that the trees must be removed to perform ambient air
monitoring in these locations. Rather, if trees or shrubs block the
line-of-sight fetch between the monitor's probe inlet and the source
under investigation, it is the EPA's position that, for most cases, a
microscale designation does not accurately reflect the monitoring scale
for this location, and instead the EPA would recommend that the
monitoring scale be designated to a more representative monitoring
scale such as middle scale or neighborhood scale.
Moreover, for cases where long-term trend sites or monitors that
determine the design value for an area cannot reasonably meet this
regulatory siting requirement, the EPA encourages monitoring
organizations to work with their respective EPA Regional offices to
determine if a waiver from this siting criteria may be appropriate
under appendix E, section 10.
For these reasons, the EPA is finalizing the language as proposed.
j. Waiver Provisions
The EPA proposed to maintain the appendix E, section 10 waiver
provisions in the current regulation for siting criteria, but to modify
section 10.3 to require that waivers from the probe-siting criteria
must be reevaluated and renewed minimally every 5 years (88 FR 5677-78,
January 27, 2023).
The EPA received one comment supporting the proposal and no adverse
comments. Therefore, the EPA is finalizing the language as proposed.
k. Acceptable Probe Materials
The EPA proposed to expand the list of acceptable probe materials
for sampling reactive gases in appendix E, section 9, from just
borosilicate glass and fluorinated ethylene propylene (FEP)
Teflon[supreg], or their equivalents. The EPA proposed to add
polyvinylidene fluoride (PVDF), also known as Kynar[supreg],
polytetrafluoroethylene (PTFE), and perfluoroalkoxy (PFA) to the list
of approved materials for efficiently transporting gaseous criteria
pollutants, and the use of Nafion\TM\ upstream of ozone analyzers (88
FR 5678, January 27, 2023). Mention of trade names or commercial
products does not constitute endorsement.
The EPA received two comments supporting the proposal and received
no adverse comments. Therefore, the EPA is finalizing the language as
proposed.
D. Incorporating Data From Next Generation Technologies
In the proposal, the EPA requested comment on how to incorporate
data from next generation technologies into Agency efforts (88 FR 5678-
80, January 27, 2023). The near real-time integration of data from
PM2.5 continuous monitors, sensors, and satellites has
allowed the EPA to use data in certain informational applications such
as EPA and USFS's Fire and Smoke Map.\199\ This mapping product uses
Application Program Interfaces (APIs) where data sets are automatically
shared on prespecified computer servers. Given the success of the Fire
and Smoke Map, the EPA indicated interest in exploring the use of next-
generation technologies to develop additional approaches, products, and
applications to help address important non-regulatory air quality data
needs. Therefore, the EPA solicited comment on the most important data
uses and data sets to consider in such future initiatives. Such
approaches and/or products could utilize historical or near real-time
data. The EPA sought this input and prioritization on use of next
generation technologies to help improve the utility of data to better
support air quality management to improve public health and the
environment.
---------------------------------------------------------------------------
\199\ Available at https://fire.airnow.gov/.
---------------------------------------------------------------------------
The EPA received comments from about two dozen entities on its
request for comments on how to incorporate data from next generation
technologies. The entities that provided comment included federal
agencies; representatives of industry and industry groups; public
health, medical, and environmental organizations; State, local and
related multi-state organizations involved in air program management;
Tribes and Tribal organizations involved in air program management; and
other State and local governments.
While there were some differences across commenters, a majority of
the commenters support use of next generation data for non-regulatory
purposes, but not for regulatory decision making due to their inherent
uncertainties and limitations. The EPA also received comments from some
environmental organizations support using alternative data for
regulatory decision making.
[[Page 16365]]
Many commenters pointed out that they are already successfully
using sensor data and networks in supplemental and informational
applications and support further expansion of these capabilities.
Across many commenters, there was support for using next generation
data as ``fit for purpose,'' filling in gaps, finding hot spots,
identifying and addressing EJ concerns, and evaluating and informing
network siting. The EPA acknowledges the successful examples of sensor
data and networks for non-regulatory purposes. A few commenters support
expanding the use of sensor data to provide real-time AQI; the EPA is
interested in this use of next generation data as well. A few
commenters pointed the need for the EPA to work closely with them and
their communities to understand and use next generation data, while
others expressed a desire for help developing best practices around
collecting and using next generation data, developing products with
data analysis/visualization, and developing appropriate QA/QC for
sensor data. The EPA acknowledges each of these requests and expects to
continue to work closely with SLTs and other stakeholders to understand
and develop information on the collection and use of next generation
data.
A few commenters offered more detailed comments. Some recommended
that the EPA repropose implementation provisions related to next
generation technologies with greater clarity to provide for meaningful
comment. For example, the use of low-cost sensor and satellite data
could be used in drawing nonattainment area boundaries or identifying
sources for emissions control, but doing so would be such a significant
change from prior EPA policy that it warrants a more specific proposal,
beyond the scope of this request for comment. In response to this
comment, the EPA notes it did not propose or change the use of non-
regulatory measurement data as part of this proposal, but instead
opened an opportunity to comment about the use of next generation
technologies.
Another commenter stated that while low-cost sensor data can be
invaluable for some purposes, the potentially overwhelming amount of
data produced by sensors may present additional challenges to
communities without the resources or expertise to analyze it. Cost is
another concern associated with some next generation technologies of
which some communities may not be aware, as the initial cost of the
sensor alone is not indicative of the total cost of operation, which
can include costs of internet access and servers. The EPA appreciates
the need to consider all the costs of implementing and maintaining
sensor data.
Another commenter stated that having a dense sensor network
collocated with FRMs and FEMs could help ensure timely maintenance of
the regulatory measurements in the event there appears to be a
divergence of data. The EPA appreciates the comment that emphasizes how
sensors could be used to complement the FRM and FEM data with regard to
ensuring timely maintenance.
Another commentor strongly opposes incorporating sensor data into
any EPA systems unless robust quality assurance (QA) practices are
widely established and managed by qualified personnel. The EPA agrees
that QA is necessary, and notes that the ``fit for purpose'' aspect of
using sensor data will inform the appropriate QA associated with the
intended use of such data.
In summary, the EPA invited comment on how we should consider
incorporating data from next generation technologies into our air
monitoring efforts. In seeking comment on this topic, the EPA did not
propose to add, edit, or delete any regulatory language associated with
the PM NAAQS. The EPA received comments from a variety of entities that
largely support using next generation data for a variety of purposes
that supplement, but cannot replace, the measurement data from
monitoring methods required (i.e., FRMs and FEMs) for regulatory
decision making. Across many commenters, there was support for using
next generation technologies and data as ``fit for purpose,'' filling
in gaps, finding hot spots, identifying, and addressing EJ concerns,
and evaluating and informing network siting. Quality assurance of the
data will be an important component in the use of next generation
technology data. The EPA will consider these comments as it continues
its work with the co-regulated community comprised of SLT agencies and
other stakeholders to understand and use next generation data and joint
efforts to manage the nation's ambient air.
VIII. Clean Air Act Implementation Requirements for the Revised Primary
Annual PM2.5 NAAQS
The EPA's revision to the primary annual PM2.5 NAAQS
discussed in section II above triggers a number of implementation
related activities that were described in the NPRM. The two most
immediate implementation impacts following a final new or revised NAAQS
are related to stationary source permitting and the initial area
designations process. Permitting implications are discussed below in
section VIII.E. With regard to initial area designations, the EPA is
separately issuing a memorandum regarding the Initial Area Designations
for the Revised Primary Annual Fine Particle National Ambient Air
Quality Standard Memorandum (the ``Annual PM2.5 NAAQS
Designations Memorandum'') that will provide information about the
statutory schedule for the designations process. For other
implementation related implications, please refer back to the NPRM
section VIII.
The NPRM also referred to the PM2.5 State Implementation
Plan (SIP) Requirements Rule (81 FR 58010, August 24, 2016), which
specifies planning requirements for areas designated as nonattainment
for purposes of the PM2.5 NAAQS and includes a number of key
recommendations for areas to consider implications of environmental
justice through the attainment planning process, consistent with the
identification of at-risk groups in the 2019 ISA and ISA Supplement and
the statutory requirement to protect the health of at-risk groups. As
stated in the NPRM, State and local air agencies are encouraged to
consider how they might develop implementation plans that encourage
early emission reductions.
A. Designation of Areas
As discussed in section II, with respect to the PM2.5
NAAQS, the EPA is finalizing: (1) Revisions to the level of the primary
annual PM2.5 NAAQS and retaining the current primary 24-hour
PM2.5 NAAQS (section II.B.4); and (2) no change to the
current secondary annual and 24-hour PM2.5 NAAQS at this
time (section V.B.4). Upon promulgation of a new or revised NAAQS,
States and the EPA must initiate the process for initial designations.
The timeline for initial area designations begins with promulgation
of the revised primary annual PM2.5 NAAQS, as stated in the
CAA section 107(d)(1)(B)(i). Through this process, which provides for
input from States and others at various stages, the EPA identifies
areas of the country that either meet or do not meet the revised
primary annual PM2.5 NAAQS, along with the nearby areas
contributing to NAAQS violations. The following includes additional
information regarding the designations process described in the CAA.
Section 107(d)(1) of the CAA states that, ``By such date as the
Administrator may reasonably require, but not later than 1 year after
promulgation of a new or revised national ambient air quality standard
for any pollutant under section
[[Page 16366]]
109, the Governor of each State shall . . . submit to the Administrator
a list of all areas (or portions thereof) in the State'' and make
recommendations for whether the EPA should designate those areas as
nonattainment, attainment, or unclassifiable.\200\ The CAA provides the
EPA with discretion to require States to submit their designations
recommendations within a reasonable amount of time not exceeding one
additional year.\201\ Section 107(d)(1)(A) of the CAA also states that
``the Administrator may not require the Governor to submit the required
list sooner than 120 days after promulgating a new or revised national
ambient air quality standard.'' Section 107(d)(1)(B)(i) further
provides, ``Upon promulgation or revision of a NAAQS, the Administrator
shall promulgate the designations of all areas (or portions thereof) .
. . as expeditiously as practicable, but in no case later than 2 years
from the date of promulgation. Such period may be extended for up to
one year in the event the Administrator has insufficient information to
promulgate the designations.'' With respect to the NAAQS setting
process, courts have interpreted the term ``promulgation'' to be
signature and widespread dissemination of a final rule.\202\
---------------------------------------------------------------------------
\200\ While the CAA says ``designating'' with respect to the
Governor's letter, in the full context of the CAA section it is
clear that the Governor actually makes a recommendation to which the
EPA must respond via a specified process if the EPA does not accept
it.
\201\ In certain circumstances in which the Administrator has
insufficient information to promulgate area designations within two
years from the promulgation of the NAAQS, CAA section
107(d)(1)(B)(i) provides that the EPA may extend the designations
schedule by up to one year.
\202\ API v. Costle, 609 F.2d 20 (D.C. Cir. 1979).
---------------------------------------------------------------------------
If the EPA agrees with the designations recommendation of the
State, then it may proceed to promulgate the designations for such
areas. If, however, the EPA disagrees with the State's recommendation,
then the EPA may elect to make modifications to the recommended
designations. By no later than 120 days prior to promulgating the final
designations, the EPA is required to notify States of any intended
modifications to the State designation recommendations for any areas or
portions thereof, including the boundaries of areas, as the EPA may
deem necessary. States then have an opportunity to comment on the EPA's
intended modification and tentative designation decision. If a State
elects not to provide designation recommendations for any area, then
the EPA must itself promulgate the designation that it deems
appropriate.
While section 107(d) of the CAA specifically addresses the
designations process for States, the EPA intends to follow the same
process for Tribes to the extent practicable, pursuant to section
301(d) of the CAA regarding Tribal authority, and the Tribal Authority
Rule (63 FR 7254, February 12, 1998). To provide clarity and
consistency in doing so, the EPA issued a guidance memorandum to our
Regional Offices on working with Tribes during the designations
process.\203\
---------------------------------------------------------------------------
\203\ ``Guidance to Regions for Working with Tribes during the
National Ambient Air Quality Standards (NAAQS) Designations
Process,'' December 20, 2011, Memorandum from Stephen D. Page to
Regional Air Directors, Regions 1-X available at https://www.epa.gov/sites/default/files/2017-02/documents/12-20-11_guidance_to_regions_for_working_with_tribes_naaqs_designations.pdf
.
---------------------------------------------------------------------------
Consistent with the process used in previous area designations
efforts, the EPA will evaluate each area on a case-by-case basis
considering the specific facts and circumstances unique to the area to
support area boundary decisions for the revised standard. The EPA
intends to issue a designations memorandum which will provide
information regarding the designations process. In broad overview, the
EPA has historically used area-specific analyses to support
nonattainment area boundary recommendations and final boundary
determinations by evaluating factors such as air quality data,
emissions and emissions-related data (e.g., population density and
degree of urbanization, traffic and commuting patterns), meteorology,
geography/topography, and jurisdictional boundaries. We expect to
follow a similar process when establishing area designations for this
revised PM2.5 NAAQS. CAA section 107(d) explicitly requires
that the EPA designate as nonattainment not only the area that is
violating the pertinent standard, but also those nearby areas that
contribute to the violation in the violating area. In the
PM2.5 NAAQS Designations Memorandum, the EPA intends to
include information regarding consideration of federal land boundaries
that may be fully or partially included within the bounds of a county
otherwise identified as nonattainment.
As with past revisions of the PM2.5 NAAQS, the EPA
intends to make the designations decisions for the revised primary
annual PM2.5 NAAQS based on the most recent three years of
quality-assured, certified air quality data in the EPA's Air Quality
System (AQS). Accordingly, the EPA recommends that States base their
initial area designation recommendations on the most current available
three years of complete and certified air quality data at the time of
the recommendations. The EPA will then base the final designations on
the most recent three consecutive years of complete, certified air
quality monitoring data available at the time of final
designations.\204\
---------------------------------------------------------------------------
\204\ In certain circumstances in which the Administrator has
insufficient information to promulgate area designations within two
years from the promulgation of a new or revised NAAQS, CAA section
107(d)(1)(B)(i) provides the EPA may extend the designations
schedule by up to one year.
---------------------------------------------------------------------------
Monitoring data are currently available from numerous existing
PM2.5 Federal Equivalent Methods (FEM) and Federal Reference
Methods (FRM) sites to determine violations of the revised primary
annual PM2.5 NAAQS. As described in section VII.C.3.b, the
EPA took comment on how to deal with cases where an FEM is approved by
the EPA with an update and when it can be implemented in the field. The
EPA took comment on how to approach the data produced during this lag
and received input from over a dozen commenters. The commenters asked
that the EPA be flexible in allowing the use of updated method
correction factors intended to improve the data comparability between
the FRMs and FEMs. The EPA will address any data correction issues
between the FRMs and FEMs through a future Notice of Data Availability
(NOA).
Consistent with past practice and as noted in the NPRM, the EPA
intends to provide additional information concerning the designations
process, including information about the schedule and recommendations
for determining area boundaries in the forthcoming Annual
PM2.5 NAAQS Designations Memorandum. Other topics addressed
in this memorandum include the schedule for preparing and submitting
exceptional events initial notification and exceptional events
demonstrations relevant to the designations process, and information
related to wildfire and prescribed fire on wildlands as it pertains to
initial area designations, as well as addressing back-correction of PM
FEM data when a method has an approved factory calibration as part of a
method update. The Annual PM2.5 NAAQS Designations
Memorandum is intended to assist States and Tribes in formulating their
area recommendations.\205\
---------------------------------------------------------------------------
\205\ See: https://www.epa.gov/particle-pollution-designations.
---------------------------------------------------------------------------
As discussed in the proposal, the ``Treatment of Data Influenced by
Exceptional Events; Final Rule,'' (81 FR 68216, October 3, 2016) and
codified at 40 CFR 50.1, 40 CFR 50.14, and 40 CFR 51.930, contains
instructions and requirements for air agencies that may
[[Page 16367]]
flag air quality data for certain days in the Air Quality System due to
potential impacts from exceptional events (i.e., such as prescribed
fires on wildland, wildfires, or high wind dust storms). Accordingly,
for purposes of initial area designations for a new or revised NAAQS,
an air agency may submit to the EPA an exceptional events demonstration
with supporting information and analyses for each monitoring site and
day the air agency claims the EPA should exclude from design value
calculations for designations purposes.
The EPA has provided tools to assist air agencies in preparing
adequate exceptional events demonstrations.\206\ Further, the EPA will
continue to work with air agencies as they identify exceptional events
that may influence decisions related to the initial area designations
process, and to prepare and submit exceptional events demonstrations if
appropriate. Importantly, air quality monitoring data may be influenced
by emissions from prescribed fires on wildland and wildfires. The EPA's
Exceptional Events Rule provides for both of these types of events to
be considered as exceptional events, provided the affected air agencies
submit exceptional events demonstrations that meet the procedural and
technical requirements of the EPA's Exceptional Events Rule. To that
end, the EPA has issued guidance addressing development of exceptional
events demonstrations for both wildfire and prescribed fires on
wildland.\207\ In light of the growing frequency and severity of
wildfire events, and expected increases in the application of
prescribed fire as a means to achieve long-term reductions in high
severity wildfire risk and associated smoke impacts, the EPA seeks to
ensure that the Agency's exceptional events process provides an
efficient and clear pathway for excluding data that may be affected by
such events in a manner that is consistent with the Clean Air Act and
the public health objectives of the NAAQS. Accordingly, the EPA is
continuing to explore opportunities to develop additional tools that
could assist air agencies in preparing exceptional events
demonstrations for wildfires and prescribed fires on wildland. In
addition, EPA intends to continue engaging with the U.S. Department of
Agriculture, U.S. Department of the Interior, air agencies, and other
stakeholders on these issues. For more information regarding the
exceptional events demonstration submission deadlines for the area
designations process, please see Table 2 to 40 CFR 50.14(c)(2)(vi)--
``Schedule for Initial Notification and Demonstration Submission for
Data Influenced by Exceptional Events for Use in Initial Area
Designations.''
---------------------------------------------------------------------------
\206\ See the EPA's Exceptional Events homepage at https://www.epa.gov/air-quality-analysis/treatment-air-quality-data-influenced-exceptional-events-homepage-exceptional.
\207\ See EPA's ``Final Guidance on the Preparation of
Exceptional Events Demonstrations for Wildfire Events that May
Influence Ozone Concentrations and EPA's Exceptional Events
Guidance: Prescribed Fire on Wildland that May Influence Ozone and
Particulate Matter Concentrations,'' found on EPA's Exceptional
Events homepage at https://www.epa.gov/air-quality-analysis/treatment-air-quality-data-influenced-exceptional-events-homepage-exceptional.
---------------------------------------------------------------------------
B. Section 110(a)(1) and (2) Infrastructure SIP Requirements
As discussed in the NPRM, the CAA directs States to address basic
SIP requirements to implement, maintain, and enforce the NAAQS. Under
CAA sections 110(a)(1) and (2), states are required to have State
implementation plans that provide the necessary air quality management
infrastructure that provides for the implementation, maintenance, and
enforcement of the NAAQS. After the EPA promulgates a new or revised
NAAQS, States are required to make a new SIP submission to establish
that they meet the necessary structural requirements for such new or
revised NAAQS or make changes to do so. The EPA refers to this type of
SIP submission as an ``infrastructure SIP submission.'' Under CAA
section 110(a)(1), all States are required to make these infrastructure
SIP submissions within three years after the effective date of a new or
revised primary standard. While the CAA authorizes the EPA to set a
shorter time for States to make these SIP submissions, the EPA is
requiring submission of infrastructure SIPs within three years of the
effective date of this revised primary annual PM2.5 NAAQS.
The EPA has provided general guidance to States concerning its
interpretation of these requirements of CAA section 110(a)(1) and (2)
in the context of infrastructure SIP submissions for a new or revised
NAAQS.\208\ The EPA encourages States to use this guidance when
developing their infrastructure SIPs for this revised primary annual
PM2.5 NAAQS.
---------------------------------------------------------------------------
\208\ See ``Guidance on Infrastructure State Implementation Plan
(SIP) Elements under Clean Air Act Sections 110(a)(1) and
110(a)(2)'' September 2013, Memorandum from Stephen D. Page to
Regional Air Directors, Regions 1-10.
---------------------------------------------------------------------------
As a reminder, the EPA notes that States are not required to
address nonattainment plan requirements for purposes of the revised
primary annual PM2.5 NAAQS on the same schedule as
infrastructure SIP requirements. The EPA interprets the CAA such that
two elements identified in section 110(a)(2) are not subject to the 3-
year submission deadline of section 110(a)(1) and thus States are not
required to address them in the context of an infrastructure SIP
submission. The elements pertain to part D, in title I of the CAA,
which addresses additional SIP requirements for nonattainment areas.
Therefore, for the reasons explained below, the following section
110(a)(2) elements are considered by the EPA to be outside the scope of
infrastructure SIP actions: (1) The portion of section 110(a)(2)(C),
programs for enforcement of control measures and for construction or
modification of stationary sources that applies to permit programs
applicable in designated nonattainment areas (known as ``nonattainment
new source review'') under part D; and (2) section 110(a)(2)(I), which
requires a SIP submission pursuant to part D, in its entirety.
Accordingly, the EPA does not expect States to address the
requirement for a new or revised NAAQS in the infrastructure SIP
submissions to include regulations or emissions limits developed
specifically for attaining the relevant standard in areas designated
nonattainment for the revised primary annual PM2.5 NAAQS.
States are required to submit infrastructure SIP submissions for the
revised primary annual PM2.5 NAAQS before they will be
required to submit nonattainment plan SIP submissions to demonstrate
attainment with the same NAAQS. States are required to submit
nonattainment plan SIP submissions to provide for attainment and
maintenance of a revised primary annual PM2.5 NAAQS within
18 months from the effective date of nonattainment area designations as
required under CAA section 189(a)(2)(B). The EPA reviews and acts upon
these later SIP submissions through a separate process. For this
reason, the EPA does not expect States to address new nonattainment
area emissions controls per section 110(a)(2)(I) in their
infrastructure SIP submissions.
One of the required infrastructure SIP elements is that each State
SIP must contain adequate provisions to prohibit, consistent with the
provisions of title I of the CAA, emissions from within the State that
will significantly contribute to nonattainment in, or interfere with
maintenance by, any other State of the primary or secondary NAAQS.\209\
This
[[Page 16368]]
element is often referred to as the ``good neighbor'' or ``interstate
transport'' provision.\210\ The provision has two prongs: significant
contribution to nonattainment (prong 1), and interference with
maintenance (prong 2). The EPA and States must give independent
significance to prong 1 and prong 2 when evaluating downwind air
quality problems under CAA section 110(a)(2)(D)(i)(I).\211\ Further,
case law has established that the EPA and States must implement
requirements to meet interstate transport obligations in alignment with
the applicable statutory attainment schedule of the downwind areas
impacted by upwind-state emissions.\212\ Thus, the EPA anticipates that
States will need to address interstate transport obligations associated
with this revised PM NAAQS, in alignment with the provisions of subpart
4 of part D of the CAA, as discussed in more detail in section VIII.C
below. Specifically, States must implement any measures required to
address interstate transport obligations as expeditiously as
practicable and no later than the next statutory attainment date, i.e.,
for this NAAQS revision as expeditiously as practicable, but no later
than the end of the sixth calendar year following nonattainment area
designations. See CAA section 188(c). States may find it efficient to
make SIP submissions to address the interstate transport provisions
separately from other infrastructure SIP elements.
---------------------------------------------------------------------------
\209\ CAA section 110(a)(2)(D)(i)(I).
\210\ CAA section 110(a)(2)(D)(i)(II) also addresses certain
interstate effects that states must address and thus is also
sometimes referred to as relating to ``interstate transport.''
\211\ See North Carolina v. EPA, 531 F.3d 896, 909-11 (D.C. Cir.
2008).
\212\ See id. 911-13. See also Wisconsin v. EPA, 938 F.3d 303,
313-20 (D.C. Cir. 2019); Maryland v. EPA, 958 F.3d 1185, 1203-04
(D.C. Cir. 2020).
---------------------------------------------------------------------------
Each State has the authority and responsibility to review its air
quality management program's existing SIP provisions in light of a new
or revised NAAQS to determine if any revisions are necessary to
implement the new or revised NAAQS. Most States have revised and
updated their SIPs in recent years to address requirements associated
with other revised NAAQS. For certain infrastructure elements, some
States may believe they already have adequate State regulations adopted
and approved into the SIP to address a particular requirement with
respect to the revised primary annual PM2.5 NAAQS.
If a State determines that existing SIP-approved provisions are
adequate in light of this revised primary annual PM2.5 NAAQS
with respect to a given infrastructure SIP element (or sub-element),
then the State may make an infrastructure SIP submission ``certifying''
that the existing State's existing EPA approved SIP already contains
provisions that address one or more specific section 110(a)(2)
infrastructure elements.\213\ In the case of such a submission, the
State does not have to include a copy of the relevant provision (e.g.,
rule or statute) itself. Rather, this certification submission should
provide citations to the SIP-approved State statutes, regulations, or
non-regulatory measures, as appropriate, in or referenced by the
already EPA-approved SIP that meet particular infrastructure SIP
element requirements. The State's infrastructure SIP submission should
also include an explanation as to how the State has determined that
those existing provisions meet the relevant requirements.
---------------------------------------------------------------------------
\213\ A ``certification'' approach would not be appropriate for
the interstate pollution control requirements of CAA section
110(a)(2)(D)(i).
---------------------------------------------------------------------------
Like any other SIP submission, that State can make such an
infrastructure SIP submission certifying that it has already met some
or all of the applicable requirements only after it has provided
reasonable notice and opportunity for public hearing. This ``reasonable
notice and opportunity for public hearing'' requirement for
infrastructure SIP submissions is to meet the requirements of CAA
sections 110(a) and 110(l). Under the EPA's regulations at 40 CFR part
51, if a public hearing is held, an infrastructure SIP submission must
include a certification by the State that the public hearing was held
in accordance with the EPA's procedural requirements for public
hearings. See 40 CFR part 51, appendix V, section 2.1(g), and see 40
CFR 51.102.
In consultation with the EPA's Regional office, a State should
follow all applicable EPA regulations governing infrastructure SIP
submissions in 40 CFR part 51--e.g., subpart I (Review of New Sources
and Modifications), subpart J (Ambient Air Quality Surveillance),
subpart K (Source Surveillance), subpart L (Legal Authority), subpart M
(Intergovernmental Consultation), subpart O (Miscellaneous Plan Content
Requirements), subpart P (Protection of Visibility), and subpart Q
(Reports). For the EPA's general criteria for infrastructure SIP
submissions, refer to 40 CFR part 51, appendix V, Criteria for
Determining the Completeness of Plan Submissions. For additional
information on infrastructure SIP submission requirements, refer to the
EPA's 2013 guidance entitled ``Guidance on Infrastructure State
Implementation Plan (SIP) Elements under Clean Air Act Sections
110(a)(1) and 110(a)(2).'' The EPA recommends that States
electronically submit their infrastructure SIPs to the EPA through the
State Plan Electronic Collaboration System (SPeCS),\214\ an online
system available through the EPA's Central Data Exchange.
---------------------------------------------------------------------------
\214\ https://cdx.epa.gov/.
---------------------------------------------------------------------------
C. Implementing Revised Primary Annual PM2.5 NAAQS in Nonattainment
Areas
As discussed in the NPRM, the EPA issued a SIP Requirements Rule
for implementing the PM2.5 NAAQS (81 FR 58010, August 24,
2016) (PM2.5 SIP Requirements Rule). It provides guidance
and establishes additional regulatory requirements for States regarding
development of attainment plans for nonattainment areas for the 1997,
2006, and 2012 revisions of the PM2.5 NAAQS. The guidance
and regulations in the SIP Requirements Rule also apply to any States
for which the EPA promulgates nonattainment area designations for the
new revised primary annual PM2.5 NAAQS.
The PM2.5 SIP Requirements Rule provides comprehensive
information regarding nonattainment plan requirements including, among
other things: nonattainment area emissions inventories; policies
regarding PM2.5 precursor pollutants (i.e., SO2,
NOX, VOC, and ammonia); control strategies (such as
reasonably available control measures and reasonably available control
technology for direct PM2.5 and relevant precursors); air
quality modeling; attainment demonstrations; reasonable further
progress requirements; quantitative milestones; and contingency
measures. Information provided in the PM2.5 SIP Requirements
Rule is supplemented by other EPA documents, including guidance on
emissions inventory development (80 FR 8787, February 19, 2015; U.S.
EPA, 2017), optional PM2.5 precursor demonstrations (U.S.
EPA, 2019b),\215\ and guidance on air quality modeling for meeting air
quality goals for the ozone and PM2.5 NAAQS and regional
haze program (U.S. EPA, 2018b).
---------------------------------------------------------------------------
\215\ Provides guidance on developing demonstrations under
section 189(e) intended to show that a certain PM2.5
precursor in a particular nonattainment area does not significantly
contribute to PM2.5 concentrations that exceed the
standard.
---------------------------------------------------------------------------
As stated in the NPRM, the PM2.5 SIP Requirements Rule
provides recommendations to States regarding consideration of
environmental justice in the context of PM2.5 attainment
[[Page 16369]]
planning. Some of the considerations for States include: (1)
Identifying areas with overburdened communities where more ambient
monitoring may be warranted; (2) targeting emissions reductions that
may be needed to attain the PM2.5 NAAQS; and (3) increasing
opportunities for meaningful involvement for overburdened populations
(see 88 FR 5558, 5684, January 27, 2023; 80 FR 58010, 58136, August 25,
2016). In light of the identification of at-risk populations for this
reconsideration, the EPA encourages States to consider these and other
factors as part of their attainment plan SIP development process.
The PM2.5 SIP Requirements Rule outlines some examples
of how States can elect to implement these recommendations.\216\ For
instance, States can use modeling and screening tools to better
understand where sources of PM2.5 or PM2.5
precursor emissions are located and identify areas that may be
candidates for additional ambient monitoring. Furthermore, once these
target areas are identified, States can prioritize direct
PM2.5 or PM2.5 precursor control measures and
enforcement strategies in these areas to reduce ambient
PM2.5 and achieve the NAAQS. As articulated in the NPRM and
the PM2.5 SIP Requirements Rule, the EPA recognizes that
States have flexibility under the CAA to concentrate State resources on
controlling sources of PM2.5 emissions in light of
environmental justice considerations (see 88 FR 5558, 5684, January 27,
2023; 81 FR 58010, 58137, August 24, 2016). Moreover, States can
establish opportunities to bolster meaningful involvement in a number
of ways, such as communicating in appropriate languages, ensuring
access to draft SIPs and other information, and developing enhanced
notice-and-comment opportunities, as appropriate (see 88 FR 5558, 5684,
January 27, 2023; 80 FR 58010, 58136, August 25, 2016).
---------------------------------------------------------------------------
\216\ For more information on the EPA's recommendations and
examples, see 81 FR 58010, 58137, August 24, 2016.
---------------------------------------------------------------------------
As previously mentioned, the PM2.5 SIP Requirements Rule
provides guidance and regulatory requirements for remaining
nonattainment areas for the 1997, 2006, and 2012 revisions of the
PM2.5 NAAQS, as well as for nonattainment areas designated
pursuant to any future revisions of the PM2.5 NAAQS,
including the revised annual PM2.5 NAAQS being finalized in
this action. The EPA is not making any changes to the current
PM2.5 SIP Requirements Rule.
D. Implementing the Primary and Secondary PM10 NAAQS
As summarized in sections III.B.4 and V.B.4 above, the EPA is
retaining the current primary and secondary 24-hour PM10
NAAQS to protect against the health effects associated with short-term
exposures to thoracic coarse particles and against the welfare effects
considered in this reconsideration (i.e., visibility, climate, and
materials effects). The EPA is retaining the existing implementation
strategy for meeting the CAA requirements for the PM10
NAAQS. States and emissions sources should continue to follow the
existing regulations and guidance for implementing the current
standards.\217\
---------------------------------------------------------------------------
\217\ CAA Sections 110(a) and 172 contain general nonattainment
planning provisions, regarding the public review, adoption,
submittal, and content of implementation plans. CAA Section 189
specifies additional plan provisions for particulate matter
nonattainment areas. General Preamble for the Implementation of
Title I of the Clean Air Act Amendments of 1990 provides a detailed
discussion of the EPA's interpretation of the Title I requirements
(57 FR 13498, April 16, 1992; 59 FR 41998, August 16, 1994).
---------------------------------------------------------------------------
E. Prevention of Significant Deterioration and Nonattainment New Source
Review Programs for the Revised Primary Annual PM2.5 NAAQS
The CAA, at parts C and D of title I, contains preconstruction
review and permitting programs applicable to new major stationary
sources and major modifications of existing major sources. The
preconstruction review of each new major stationary source and major
modification applies on a pollutant-specific basis, and the
requirements that apply for each pollutant depend on whether the area
in which the source is situated is designated as attainment (or
unclassifiable) or nonattainment for that pollutant. In areas
designated attainment or unclassifiable for a pollutant, the Prevention
of Significant Deterioration (PSD) requirements under part C apply to
construction at major sources. In areas designated nonattainment for a
pollutant, the Nonattainment New Source Review (NNSR) requirements
under part D apply to construction at major sources. Collectively,
those two sets of permit requirements are commonly referred to as the
``major New Source Review'' or ``major NSR'' programs.
Until the EPA designates an area with respect to the revised
primary annual PM2.5 NAAQS, the NSR provisions applicable
under an area's current designation for the 1997, 2006, and 2012
PM2.5 NAAQS would continue to apply. See 40 CFR 51.166(i)(2)
and 52.21(i)(2). That is, for areas designated as attainment/
unclassifiable for the 1997, 2006, and 2012 PM2.5 NAAQS, PSD
will apply to new major stationary sources and major modifications that
trigger major source permitting requirements for PM2.5. For
areas designated nonattainment for the 1997, 2006, or 2012
PM2.5 NAAQS, NNSR requirements will apply for new major
stationary sources and major modifications that trigger major source
permitting requirements for PM2.5. When the initial area
designations for this revised primary annual PM2.5 NAAQS
become effective, those designations will further determine whether PSD
or NNSR applies to PM2.5 in a particular area, depending on
the designation status. New major sources and major modifications will
be subject to the PSD program requirements for PM2.5 if they
are located in an area that does not have a current nonattainment
designation under CAA section 107 for PM2.5.\218\
---------------------------------------------------------------------------
\218\ 40 CFR 51.166(i)(2) and 52.21(i)(2).
---------------------------------------------------------------------------
Under the PSD program, the permit applicant must demonstrate that
the new or modified source emissions increase does not cause or
contribute to a NAAQS violation. In 2017, the EPA revised the Guideline
on Air Quality Models (published as appendix W to 40 CFR part 51) to
address primary and secondary PM2.5 impacts in making this
demonstration. The EPA has since provided associated technical
guidance, models and tools, such as the recent ``Final Guidance for
Ozone and Fine Particulate Matter Permit Modeling'' (July 29,
2022).\219\ Additionally, in light of this NAAQS revision, the EPA is
updating its guidance that provides recommended significant impact
levels (SILs) for PM2.5 and expects that an updated SIL for
the revised primary annual PM2.5 NAAQS will be available
[[Page 16370]]
on or before the effective date of the final NAAQS.
---------------------------------------------------------------------------
\219\ On July 29, 2022, the EPA issued ``Final Guidance for
Ozone and Fine Particulate Matter Permit Modeling,'' available at
https://www.epa.gov/system/files/documents/2022-07/Guidance_for_O3_PM25_Permit_Modeling.pdf. This guidance provides the
EPA's recommendations for how a stationary source seeking a PSD
permit may demonstrate that it will not cause or contribute to a
violation of the National Ambient Air Quality Standards for Ozone
and PM2.5 and PSD increments for PM2.5, as
required under section 165(a)(3) of the Clean Air Act and 40 CFR
51.166(k) and 52.21(k). The EPA has also previously issued two
technical guidance documents for use in conducting these
demonstrations: ``Guidance on the Development of Modeled Emission
Rates for Precursors (MERPs) as a Tier 1 Demonstration Tool for
Ozone and PM2.5 under the PSD Permitting Program,''
available at https://www.epa.gov/sites/default/files/2020-09/documents/epa-454_r-19-003.pdf, and ``Guidance on the Use of Models
for Assessing the Impacts of Emissions from Single Sources on the
Secondarily Formed Pollutants: Ozone and PM2.5,''
available at https://www.epa.gov/sites/default/files/2020-09/documents/epa-454_r-16-005.pdf.
---------------------------------------------------------------------------
The statutory requirements for a PSD permit program set forth under
part C of title I of the CAA (sections 160 through 169) are addressed
by the EPA's PSD regulations found at 40 CFR 51.166 (minimum
requirements for an approvable PSD SIP) and 40 CFR 52.21 (PSD
permitting program for permits issued under the EPA's Federal
permitting authority). These regulations already apply to
PM2.5 in areas that are designated attainment or
unclassifiable for PM2.5 whenever a proposed new major
source or major modification triggers PSD requirements for
PM2.5.
For PSD, a ``major stationary source'' is one with the potential to
emit 250 tons per year (tpy) or more of any regulated NSR pollutant,
unless the new or modified source is classified under a list of 28
source categories contained in the statutory definition of ``major
emitting facility'' in section 169(1) of the CAA. For those 28 source
categories, a ``major stationary source'' is one with the potential to
emit 100 tpy or more of any regulated NSR pollutant. A ``major
modification'' is a physical change or a change in the method of
operation of an existing major stationary source that results, first,
in a significant emissions increase of a regulated NSR pollutant and,
second, in a significant net emissions increase of that pollutant. See
40 CFR 51.166(b)(2)(i), 40 CFR 52.21(b)(2)(i). The EPA PSD regulations
define the term ``regulated NSR pollutant'' to include any pollutant
for which a NAAQS has been promulgated and any pollutant identified by
the EPA as a constituent or precursor to such pollutant. See 40 CFR
51.166(b)(49), 40 CFR 52.21(b)(50). These regulations identify
SO2 and NOX as precursors to PM2.5 in
attainment and unclassifiable areas. See 40 CFR 51.166(b)(49)(i)(b), 40
CFR 52.21(b)(50)(i)(b).\220\ Thus, for PM2.5, the PSD
program currently requires the review and control of emissions of
direct PM2.5 emissions and SO2 and NOX
(as precursors to PM2.5), absent a demonstration otherwise
for NOX. Among other things, for each regulated NSR
pollutant emitted or increased in a significant amount, the PSD program
requires a new major stationary source or a major modification to apply
the ``best available control technology'' (BACT) to limit emissions and
to conduct an air quality impact analysis to demonstrate that the
proposed major stationary source or major modification will not cause
or contribute to a violation of any NAAQS or PSD increment.\221\ See
CAA section 165(a)(3) and (4), 40 CFR 51.166(j) and (k), 40 CFR
52.21(j) and (k). The PSD requirements may also include, in appropriate
cases, an analysis of potential adverse impacts on Class I areas. See
CAA sections 162(a) and 165(d), 40 CFR 51.166(p); 40 CFR
52.21(p)).\222\ The EPA developed the Guideline on Air Quality Models
and other documents to, among other things, provide methods and
guidance for demonstrating that increased emissions from construction
will not cause or contribute to exceedances of the PM2.5
NAAQS and PSD increments for PM2.5.\223\
---------------------------------------------------------------------------
\220\ Sulfur dioxide is a precursor to PM2.5 in all
attainment and unclassifiable areas. NOX is presumed to
be a precursor to PM2.5 in all attainment and
unclassifiable areas, unless a state or the EPA demonstrates that
emissions of NOX from sources in a specific area are not
a significant contributor to that area's ambient PM2.5
concentrations. VOC is presumed not to be a precursor to
PM2.5 in any attainment or unclassifiable area, unless a
state or the EPA demonstrates that emissions of VOC from sources in
a specific area are a significant contributor to that area's ambient
PM2.5 concentrations.
\221\ By establishing the maximum allowable level of ambient
pollutant concentration increase in a particular area, an increment
defines ``significant deterioration'' of air quality in that area.
Increments are defined by the CAA as maximum allowable increases in
ambient air concentrations above a baseline concentration and are
specified in the PSD regulations by pollutant and area
classification (Class I, II and III). 40 CFR 51.166(c), 40 CFR
52.21(c); 75 FR 64864 (October 20, 2010).
\222\ Congress established certain Class I areas in section
162(a) of the CAA, including international parks, national
wilderness areas, and national parks that meet certain criteria.
Such Class I areas, known as mandatory Federal Class I areas, are
afforded special protection under the CAA. In addition, States and
Tribal governments may establish Class I areas within their own
political jurisdictions to provide similar special air quality
protection.
\223\ See 40 CFR part 51, appendix W; 82 FR 5182 (January 17,
2017); See also U.S. EPA, 2021d. The EPA provided an initial version
of the 2021 guidance for public comment on February 10, 2020. Upon
consideration of the comments received, and consistent with
Executive Order 13990, the EPA revised the initial draft guidance
and posted the revised version for additional public comment.
---------------------------------------------------------------------------
Upon the effective date of the revised primary annual
PM2.5 NAAQS, the demonstration required under CAA Section
165(a)(3), and the associated regulations, must include the revised
primary annual PM2.5 NAAQS. In past NAAQS revision rules,
including the 2012 PM2.5 NAAQS (78 FR 3086, January 15,
2013) and 2015 Ozone NAAQS (80 FR 65292, October 26, 2015), the EPA
included limited provision that exempted certain sources with pending
PSD permit applications (those that had reached a particular stage in
the permitting process at the time the revised NAAQS was promulgated or
became effective) from the requirement to demonstrate that the proposed
emissions increases would not cause or contribute to a violation of the
revised NAAQS.\224\ In August 2019, the U.S. Court of Appeals for the
D.C. Circuit vacated the exemption provision in the PSD rules for the
2015 Ozone NAAQS, finding that the provision contradicted ``Congress's
`express policy choice' not to allow construction which will `cause or
contribute to' nonattainment of `any' effective NAAQS, regardless of
when they are adopted or when a permit was completed.'' Murray Energy
Corp. v. EPA, 936 F.3d 597, 627 (D.C. Cir. 2019).\225\ Based on that
court decision, the EPA is not establishing any PSD permitting
exemption provision in this action. Some commenters requested that the
EPA provide the same kind of relief for pending PSD permit applications
by extending the effective date of this new revised NAAQS beyond the 60
days that the EPA has traditionally used for such rules. Such comments
are addressed in the Response to Comments portion of this action. The
EPA is making this revised primary annual PM2.5 NAAQS
effective in 60 days.
---------------------------------------------------------------------------
\224\ This exemption was referred to as ``grandfathering'' in
the 2015 Ozone NAAQS and the D.C. Circuit's Murray Energy Corp.
decision on that exemption. See 80 FR 65292, 65431 (October 26,
2015); Murray Energy Corp. v. EPA, 936 F.3d 597, 627 (D.C. Cir.
2019). The EPA refers to this ``grandfathering'' provision in this
action as an exemption provision.
\225\ While the specifics of this case involved the 2015 ozone
NAAQS, the case was based upon an interpretation of CAA section
165(a) and therefore applies equally to any PSD permitting exemption
provision for a new or revised NAAQS.
---------------------------------------------------------------------------
The EPA anticipates that the existing PM2.5 air quality
in some areas will not be in attainment with the revised primary annual
PM2.5 NAAQS, and the EPA will designate these areas as
nonattainment at a later date, consistent with the designation process
described in the preceding sections. However, until such nonattainment
designation occurs, proposed new major sources and major modifications
located in any area currently designated attainment or unclassifiable
for all preexisting PM2.5 NAAQS will continue to be subject
to the PSD program requirements for PM2.5. Any proposed
major stationary source or major modification triggering PSD
requirements for PM2.5 that does not receive its PSD permit
by the effective date of a new nonattainment designation for the area
where the source would locate would then be required to satisfy
applicable NNSR preconstruction permit requirements for
PM2.5.
In areas where air pollution exceeds the level of the revised
primary annual PM2.5 NAAQS, a PSD permit applicant must
demonstrate that the source or modification will not cause or
[[Page 16371]]
contribute to a violation of the NAAQS. Section 165(a)(3)(B) of the CAA
states that a proposed source may not construct unless it demonstrates
that it will not cause or contribute to a violation of any NAAQS. This
statutory requirement is implemented through a provision contained in
the PSD regulations at 40 CFR 51.166(k) and 52.21(k).\226\ If a source
cannot make this demonstration, or if its initial air quality impact
analysis shows that the source's impact would cause or contribute to a
violation, the reviewing authority may not issue a PSD permit to that
source. However, a PSD permit applicant may be able to make this
demonstration if it compensates for the adverse impact that would
otherwise cause or contribute to a violation of the NAAQS. In contrast
to the NSR requirements for nonattainment areas, the PSD regulations do
not explicitly specify remedial actions that a prospective source must
take to address such a situation, but the EPA has historically
recognized that sources applying for PSD permits may utilize offsetting
reductions in emissions as part of the required PSD demonstration under
CAA section 165(a)(3)(B).\227\
---------------------------------------------------------------------------
\226\ 40 CFR 51.166(k) states that SIPs must require that the
owner or operator of the proposed source or modification demonstrate
that allowable emission increases from the proposed source or
modification, in conjunction with all other applicable emissions
increases or reductions (including secondary emissions), would not
cause or contribute to air pollution in violation of: (i) Any
national ambient air quality standard in any air quality control
region; or (ii) any applicable maximum allowable increase over the
baseline concentration in any area.
\227\ See, e.g., Memorandum from Stephen D. Page, Director,
Office of Air Quality Planning and Standards to Regional Air
Division Directors, Guidance Concerning Implementation of the 1-hour
SO2 NAAQS for the Prevention of Significant Deterioration
Program. August 23, 2010. Office of Air Quality Planning and
Standards U.S. EPA, Research Triangle Park. Available at: https://www.epa.gov/sites/default/files/2015-07/documents/appwso2.pdf; 44 FR
3274, 3278, January 16, 1979; See also In re Interpower of New York,
Inc., 5 E.A.D. 130, 141 (EAB 1994) (describing an EPA Region 2 PSD
permit that relied in part on offsets to demonstrate the source
would not cause or contribute to a violation of the NAAQS). 52 FR
24634, 24684, July 1, 1987; 78 FR 3085, 3261-62, January 15, 2013.
The EPA has recognized the ability of sources to obtain offsets in
the context of PSD though the PSD provisions of the Act do not
expressly reference offsets as the NNSR provisions of the Act do.
See 80 FR 65292, 65441, October 26, 2015.
---------------------------------------------------------------------------
Part D of title I of the CAA includes preconstruction review and
permitting requirements applicable to new major stationary sources and
major modifications located in areas designated nonattainment for a
pollutant for which the EPA has established a NAAQS (i.e., a criteria
pollutant). The relevant part D requirements are typically referred to
as the nonattainment NSR (NNSR) program. The EPA's regulations for the
NNSR program are contained in 40 CFR 51.165 and 52.24 and part 51,
appendix S. Specifically, the EPA has developed minimum program
requirements for a NNSR program that is approvable in a SIP, and those
requirements, which include requirements for PM2.5, are
contained in 40 CFR 51.165. In addition, 40 CFR part 51, appendix S,
contains requirements constituting an interim NNSR program. This
interim program enables NNSR permitting in nonattainment areas by
States that lack a SIP-approved NNSR permitting program during the time
between the date of the relevant designation and the date that the EPA
approves into the SIP a NNSR program. See 40 CFR part 51, appendix S,
section I; 40 CFR 52.24(k).
For NNSR, ``major stationary source'' is generally defined as a
source with the potential to emit at least 100 tpy of the regulated NSR
pollutant for which the area is designated nonattainment. In some
cases, however, the CAA and the NNSR regulations define ``major
stationary source'' for NNSR in terms of a lower rate dependent on the
pollutant and degree of nonattainment in the area. For purposes of the
PM2.5NAAQS, in addition to the general threshold level of
100 tpy in Moderate PM2.5 nonattainment areas, a lower major
source threshold of 70 tpy applies in Serious PM2.5
nonattainment areas pursuant to subpart 4 of part D, title I of the
CAA. See 40 CFR 51.165(a)(1)(iv)(A)(1)(vii) and (viii); 40 CFR part 51,
appendix S, II.A.4(i)(a)(7) and (8).
Under the NNSR program, direct PM2.5 emissions and
emissions of each PM2.5 precursor are considered separately
in accordance with the applicable major source threshold. For example,
the threshold for Serious PM2.5 nonattainment areas is 70
tpy of direct PM2.5, as well as for the PM2.5
precursors SO2, NOX, VOC, and ammonia.\228\ See
40 CFR 51.165(a)(1)(iv)(A)(1)(vii) and (viii); 40 CFR part 51, appendix
S, II.A.4.(i)(a)(7) and (8). A source qualifies as major for
nonattainment NSR in a PM2.5 nonattainment area if it emits
or has the potential to emit direct PM2.5 or any
PM2.5 precursor in an amount equal to or greater than the
applicable threshold.
---------------------------------------------------------------------------
\228\ All of these pollutants are identified as precursors to
PM2.5 in NNSR regulations. See 40 CFR
51.165(a)(1)(xxxvii)(C)(2). No significant emission rate is
established by the EPA for ammonia, and states are required to
define ``significant'' for ammonia for their respective areas unless
the state pursues the optional precursor demonstration to exclude
ammonia from planning requirements. See 40 CFR 51.165(a)(1)(x)(F);
40 CFR 51.165(a)(13).
---------------------------------------------------------------------------
For modifications, NNSR applies to proposed physical changes or
changes in the method of operation of an existing stationary source
where (1) the source is major for the nonattainment pollutant (or a
precursor for that pollutant) and (2) the physical change or change in
the method of operation of a major stationary source results, first, in
a significant emissions increase of a regulated NSR pollutant and,
second, in a significant net emissions increase of that same
nonattainment pollutant (or same precursor for that pollutant). See 40
CFR 51.165(a)(1)(v)(A); 40 CFR part 51, appendix S, II.A.5.(i). For
example, to qualify as a major modification for SO2 (as a
PM2.5 precursor) in a Moderate PM2.5
nonattainment area, the existing source would have to have the
potential to emit 100 tpy or more of SO2, and the project
would have to result in an increase in SO2 emissions of 40
tpy or more. See 40 CFR 51.165(a)(1)(x)(A).
New major stationary sources and major modifications for
PM2.5 subject to NNSR must comply with the ``lowest
achievable emission rate'' (LAER), as defined in the CAA and NNSR
rules. Such sources must also perform other analyses and obtain
emission offsets, as required under section 173 of the CAA and
applicable regulations.
Following the promulgation of this revised primary annual
PM2.5 NAAQS, some new areas may be designated nonattainment
for PM2.5. Where a State does not have an existing NNSR
program or where the current NNSR program does not apply to
PM2.5, that State will be required to submit the necessary
SIP revisions to ensure that new major stationary sources and major
modifications for PM2.5 or a PM2.5 precursor
undergo preconstruction review pursuant to the NNSR program. States
with designated nonattainment areas for the revised primary annual
PM2.5 NAAQS are required to make SIP submissions to meet
nonattainment plan requirements within 18 months from the effective
date of designations, as required under CAA section 189(a)(2)(B).
States that have existing NNSR program requirements that cannot be
interpreted to apply at the time of designation to the revised primary
annual PM2.5 NAAQS may, in the interim, issue permits in
accordance with the applicable nonattainment permitting requirements
contained in 40 CFR part 51, appendix S, which would apply to the
revised primary annual PM2.5 NAAQS upon its effective date.
See 73 FR 28321, 28340, May 16, 2008.
Finally, the EPA has released several documents that discuss air
permitting
[[Page 16372]]
and environmental justice, including, for example, a memorandum \229\
and attached permitting principles.\230\ The EPA recommends that PSD
and NNSR permitting authorities review this memorandum and the
principles and consider applying them in their air permitting actions
as appropriate to help identify, analyze, and address environmental
justice concerns in those air permitting actions to help ensure that
the NAAQS achieve their intended health benefits for at-risk
populations.
---------------------------------------------------------------------------
\229\ Memorandum from Joseph Goffman, Principal Deputy Assistant
Administrator, Office of Air and Radiation, to Air and Radiation
Division Directors, ``Principles for Addressing Environmental
Justice in Air Permitting'' (December 22, 2022), available at
https://www.epa.gov/caa-permitting/ej-air-permitting-principles-addressing-environmental-justice-concerns-air.
\230\ Id., Attachment, ``EJ in Air Permitting: Principles for
Addressing Environmental Justice Concerns in Air Permitting''
(December 2022), available at https://www.epa.gov/caa-permitting/ej-air-permitting-principles-addressing-environmental-justice-concerns-air.
---------------------------------------------------------------------------
F. Transportation Conformity Program
Transportation conformity is required under CAA section 176(c) to
ensure that transportation plans, transportation improvement programs
(TIPs) and federally supported highway and transit projects will not
cause or contribute to any new air quality violation, increase the
frequency or severity of any existing violation, or delay timely
attainment or any required interim emissions reductions or other
milestones. Transportation conformity applies to areas that are
designated as nonattainment or nonattainment areas that have been
redesignated to attainment with an approved CAA section 175A
maintenance plan (i.e., maintenance areas) for transportation-related
criteria pollutants: carbon monoxide, ozone, NO2,
PM2.5, and PM10. Transportation conformity for
the revised primary annual PM2.5 NAAQS does not apply until
one year after the effective date of nonattainment designations for
that NAAQS. See CAA section 176(c)(6) and 40 CFR 93.102(d)). The EPA's
Transportation Conformity Rule \231\ establishes the criteria and
procedures for determining whether transportation activities conform to
the SIP. No changes are being made to the transportation conformity
rule in this final rulemaking. The EPA notes that the transportation
conformity rule already addresses the PM2.5 and
PM10 NAAQS. However, in the future, the EPA intends to
review the need to issue or revise guidance describing how the current
conformity rule applies in nonattainment and maintenance areas for the
revised primary annual PM2.5 NAAQS, as needed.
---------------------------------------------------------------------------
\231\ 40 CFR part 93, subpart A.
---------------------------------------------------------------------------
G. General Conformity Program
The conformity requirement under CAA section 176(c) ensures that
federal activities implemented by federal agencies will not interfere
with a State's ability to attain and maintain the NAAQS. Under CAA
176(c)(1), the requirement prohibits Federal agencies from approving,
permitting, licensing, or funding activities that do not conform to the
purpose of the applicable SIP for the control and prevention of air
pollution. See CAA 176(c)(1)(A). Under CAA 176(c)(1)(B), conformity to
an implementation plan means that federal activities will not cause or
contribute to any new violations of the NAAQS, increase the frequency
or severity of any existing NAAQS violation, or delay timely attainment
or any required interim emissions reductions or other milestones
contained in the applicable SIP.
The general conformity program \232\ implements CAA section
176(c)(4)(A), and the criteria and procedures for determining
conformity of federal activities to the applicable SIP are established
under 40 CFR part 93 subpart B, sections 93.150 through 93.165. General
Conformity applies to federal activities that (1) would cause emissions
of relevant criteria or precursor pollutants to originate within
nonattainment areas or areas that have been redesignated to attainment
with an approved CAA section 175A maintenance plan (i.e., maintenance
areas), as set forth under 40 CFR 93.153, and (2) are not Federal
Highway Administration (FHWA) or Federal Transit Administration (FTA)
transportation projects as defined in 40 CFR 93.101 under the
transportation conformity requirements. See 40 CFR 93.153. General
conformity for the revised primary annual PM2.5 NAAQS does
not apply until one year after the effective date of the nonattainment
designation for that NAAQS. See 40 CFR 93.153(k).
---------------------------------------------------------------------------
\232\ 40 CFR part 93 subpart B.
---------------------------------------------------------------------------
With regard to issues regarding prescribed fires, which were
addressed earlier in this action, here is some additional information
regarding prescribed fires and General Conformity regulations. Under
the General Conformity regulations at 40 CFR 93.153(c)(4), a conformity
evaluation is not required to support a decision by a federal agency to
conduct or carry out prescribed burning when the burn is consistent
with the terms of a land management plan or other plan that includes
the prescribed burn at issue, where the overall plan that includes the
burn was previously evaluated under 40 CFR part 93 subpart B by the
responsible federal agency, and the agency found the plan conforms
under CAA paragraphs 176(c)(1)(A) and (1)(B). This assumes the burn at
issue will be conducted by meeting any conditions specified as
necessary for meeting conformity in the agency's decision to approve
the plan. Alternatively, a presumption of conformity applies also under
40 CFR 93.153(i)(2) for prescribed fires conducted in accordance with a
Smoke Management Program that meets the requirements of the EPA's 1998
Interim Air Quality Policy on Wildland and Prescribed Fires or an
equivalent replacement EPA policy. The preamble to the Exceptional
Events Rule explains that the EPA adapted language associated with the
six basic components of a certifiable Smoke Management Program for
exceptional events purposes from the 1998 Interim Air Quality Policy on
Wildland and Prescribed Fires (see, e.g., 81 FR 68216, 68252 (including
footnote 75), 68256, October 2, 2016). The Exceptional Events Rule at
40 CFR 50.14(a)(3)(ii)(A) also indicates that certain requirements
within the Exceptional Events Rule can be satisfied if a prescribed
fire is conducted under a certified Smoke Management Program or using
appropriate basic smoke management practices such as those identified
in Table 1 to 40 CFR 50.14 (see e.g., 81 FR 68216, 68250-68257, 68277-
68278, October 3, 2016).
No changes are being made to the general conformity regulations in
this final rulemaking and the EPA notes that the courts recognize the
regulations constitute control for the established PM2.5 and
PM10 NAAQS. However, in the future, the EPA intends to
review the need to issue or revise guidance describing how the current
General Conformity regulations apply within nonattainment and
maintenance areas for the revised primary annual PM2.5
NAAQS, as needed.\233\
---------------------------------------------------------------------------
\233\ Further, the EPA's current Unified Agenda and Regulatory
Plan includes its intention to issue a proposed rule to amend the
General Conformity Regulations. The EPA intends to address in that
regulatory action topics regarding prescribed fire, including
consideration of smoke management approaches such as those discussed
in the Exceptional Events Rule, among other topics. See, e.g.,
https://www.reginfo.gov/public/do/eAgendaViewRule?pubId=202310&RIN=2060-AV28.
---------------------------------------------------------------------------
IX. Statutory and Executive Order Reviews
Additional information about these statutes and Executive orders
can be
[[Page 16373]]
found at https://www.epa.gov/laws-regulations/laws-and-executive-orders.
A. Executive Order 12866: Regulatory Planning and Review and Executive
Order 14094: Modernizing Regulatory Review
This action is ``significant regulatory action'' as defined under
section 3(f)(1) of Executive Order 12866, as amended by Executive Order
14094. Accordingly, the EPA submitted this action to the Office of
Management and Budget (OMB) for review. Documentation of any changes
made in response to the Executive Order 12866 review is available in
the docket. The EPA prepared an illustrative analysis of the potential
costs and benefits associated with this action. This analysis,
``Regulatory Impact Analysis for the Reconsideration of the National
Ambient Air Quality Standards for Particulate Matter,'' is available in
the Regulatory Impact Analysis (RIA) docket (EPA-HQ-OAR-2019-0587) and
briefly summarized below. However, the CAA and judicial decisions make
clear that the economic and technical feasibility of attaining ambient
standards are not to be considered in setting or revising NAAQS,
although such factors may be considered in the development of State
plans to implement the standards. Accordingly, although an RIA has been
prepared, the results of the RIA have not been considered in issuing
this final rule.
The RIA estimates the costs and monetized human health benefits in
2032, after implementing existing and expected regulations and
assessing emissions reductions to meet the current primary annual and
24-hour particulate matter NAAQS (12/35 [mu]g/m\3\), associated with
applying national control strategies for the revised annual and 24-hour
standard levels of 9/35 [mu]g/m\3\, as well as the following less and
more stringent alternative standard levels: (1) A less stringent
alternative annual standard level of 10 [mu]g/m\3\ in combination with
the current 24-hour standard (i.e., 10/35 [mu]g/m\3\), (2) a more
stringent alternative annual standard level of 8 [mu]g/m\3\ in
combination with the current 24-hour standard (i.e., 8/35 [mu]g/m\3\),
and (3) a more stringent alternative 24-hour standard level of 30
[mu]g/m\3\ in combination with an annual standard level of 10 [mu]g/
m\3\ (i.e., 10/30 [mu]g/m\3\). Table 3 provides a summary of the
estimated monetized benefits, costs, and net benefits associated with
applying national control strategies toward reaching the revised and
alternative standard levels.
Table 3--Estimated Monetized Benefits, Costs, and Net Benefits of the Illustrative Control Strategies Applied
Toward the Primary Revised and Alternative Annual and Daily Standard Levels of 10/35 [mu]g/m\3\, 10/30 [mu]g/
m\3\, 9/35 [mu]g/m\3\, and 8/35 [mu]g/m\3\ in 2032 for the U.S.
[Millions of 2017$]
----------------------------------------------------------------------------------------------------------------
10/35 10/30 9/35 8/35
----------------------------------------------------------------------------------------------------------------
Benefits \a\.................... $8,500 and $17,000 $10,000 and $22,000 and $48,000 and
$21,000. $46,000. $99,000.
Costs \b\....................... $200.............. $340.............. $590.............. $1,500.
-------------------------------------------------------------------------------
Net Benefits................ $8,300 and $17,000 $9,900 and $21,000 $22,000 and $46,000 and
$46,000. $97,000.
----------------------------------------------------------------------------------------------------------------
Notes: Rows may not appear to add correctly due to rounding. We provide a snapshot of costs and benefits in
2032, using the best available information to approximate social costs and social benefits recognizing
uncertainties and limitations in those estimates. The estimated costs and monetized human health benefits
associated with applying national control strategies do not fully account for all the emissions reductions
needed to reach the final and more stringent alternative standard levels for some standard levels analyzed.
\a\ We assume that there is a cessation lag between the change in PM exposures and the total realization of
changes in mortality effects. Specifically, we assume that some of the incidences of premature mortality
related to PM2.5 exposures occur in a distributed fashion over the 20 years following exposure, which affects
the valuation of mortality benefits at different discount rates. Similarly, we assume there is a cessation lag
between the change in PM exposures and both the development and diagnosis of lung cancer. The benefits are
associated with two point estimates from two different epidemiologic studies, and we present the benefits
calculated at a real discount rate of 3 percent. The monetized benefits exclude additional health and welfare
benefits that could not be quantified.
\b\ The costs are annualized using a 7 percent interest rate.
B. Paperwork Reduction Act (PRA)
This action does not impose any new information collection burden
under the PRA. OMB has previously approved the information collection
activities contained in the existing regulations and has assigned OMB
control number 2060-0084. The data collected through this information
collection consist of ambient air concentration measurements for the
seven air pollutants with national ambient air quality standards (i.e.,
ozone, sulfur dioxide, nitrogen dioxide, lead, carbon monoxide,
PM2.5 and PM10), ozone precursors, air toxics,
meteorological variables at a select number of sites, and other
supporting measurements. Accompanying the pollutant concentration data
are quality assurance/quality control data and air monitoring network
design information. The EPA and others (e.g., State and local air
quality management agencies, tribal entities, environmental
organizations, academic institutions, industrial groups) use the
ambient air quality data for many purposes including informing the
public and other interested parties of an area's air quality, judging
an area's air quality in comparison with the established health or
welfare standards, evaluating an air quality management agency's
progress in achieving or maintaining air pollutant levels below the
national and local standards, developing and revising State
Implementation Plans (SIPs), evaluating air pollutant control
strategies, developing or revising national control policies, providing
data for air quality model development and validation, supporting
enforcement actions, documenting episodes and initiating episode
controls, assessing air quality trends, and conducting air pollution
research.
C. Regulatory Flexibility Act (RFA)
I certify that this action will not have a significant economic
impact on a substantial number of small entities under the RFA. This
action will not impose any requirements on small entities. Rather, this
final rule establishes national standards for allowable concentrations
of PM in ambient air as required by section 109 of the CAA. See also
American Trucking Associations v. EPA, 175 F.3d 1027, 1044-45 (D.C.
Cir. 1999) (NAAQS do not have significant impacts upon small entities
because NAAQS themselves impose no regulations upon small entities),
rev'd in part on other grounds, Whitman v. American Trucking
Associations, 531 U.S. 457 (2001).
[[Page 16374]]
D. Unfunded Mandates Reform Act (UMRA)
This action does not contain an unfunded mandate of $100 million or
more as described in the Unfunded Mandates Reform Act (UMRA), 2 U.S.C.
1531-1538, and does not significantly or uniquely affect small
governments. Furthermore, as indicated previously, in setting a NAAQS
the EPA cannot consider the economic or technological feasibility of
attaining ambient air quality standards, although such factors may be
considered to a degree in the development of State plans to implement
the standards. See also American Trucking Associations v. EPA, 175 F.
3d at 1043 (noting that because the EPA is precluded from considering
costs of implementation in establishing NAAQS, preparation of the RIA
pursuant to the Unfunded Mandates Reform Act would not furnish any
information that the court could consider in reviewing the NAAQS).
E. Executive Order 13132: Federalism
This action will not have substantial direct effects on the States,
on the relationship between the National Government and the States, or
on the distribution of power and responsibilities among the various
levels of government. However, the EPA recognizes that States will have
a substantial interest in this action and any future revisions to
associated requirements.
F. Executive Order 13175: Consultation and Coordination With Indian
Tribal Governments
This action does not have Tribal implications, as specified in
Executive Order 13175. It does not have a substantial direct effect on
one or more Indian Tribes as Tribes are not obligated to adopt or
implement any NAAQS. In addition, Tribes are not obligated to conduct
ambient monitoring for PM or to adopt the ambient monitoring
requirements of 40 CFR part 58. Thus, Executive Order 13175 does not
apply to this action. However, consistent with the EPA Policy on
Consultation and Coordination with Indian Tribes, the EPA offered
consultation to all 574 Federally Recognized Tribes during the
development of this action. Although no Tribes requested consultation,
the EPA provided informational meetings including an informational
meeting with the Pueblo de San Ildefonso and provided information on
the monthly National Tribal Air Association calls.
G. Executive Order 13045: Protection of Children From Environmental
Health Risks and Safety Risks
Executive Order 13045 directs federal agencies to include an
evaluation of the health and safety effects of the planned regulation
on children in federal health and safety standards and explain why the
regulation is preferable to potentially effective and reasonably
feasible alternatives. This action is subject to Executive Order 13045
because it is a significant regulatory action under section 3(f)(1) of
Executive Order 12866, and the EPA believes that the environmental
health or safety risk addressed by this action may have a
disproportionate effect on children. Accordingly, we have evaluated the
environmental health or safety effects of PM exposures on children. The
protection offered by these standards may be especially important for
children because childhood represents a lifestage associated with
increased susceptibility to PM-related health effects. Because children
have been identified as a susceptible population, we have carefully
evaluated the environmental health effects of exposure to PM pollution
among children. Children make up a substantial fraction of the U.S.
population, and often have unique factors that contribute to their
increased risk of experiencing a health effect due to exposures to
ambient air pollutants because of their continuous growth and
development. As described in the 2019 Integrated Science Assessment,
children may be particularly at risk for health effects related to
ambient air PM2.5 exposures compared with adults because
they have (1) a developing respiratory system, (2) increased
ventilation rates relative to body mass compared with adults, and (3)
an increased proportion of oral breathing, particularly in boys,
relative to adults. More detailed information on the evaluation of the
scientific evidence and policy considerations pertaining to children,
including an explanation for why the Administrator judges the revised
standards to be requisite to protect public health, including the
health of children, with an adequate margin of safety, are contained in
section II.A.2. ``Overview of the Health Effects Evidence'', section
II.A.2.b ``Public Health Implications and At-Risk Populations'' and
II.B ``Conclusions on the Primary PM2.5 Standards'' of this
preamble. Copies of all documents have been placed in the public docket
for this action. The Administrator judges that revising the primary
annual PM2.5 standard to a level of 9.0 [micro]g/m\3\ and
retaining the primary 24-hour PM2.5 standard provides
requisite public health protection with an adequate margin of safety,
including for children. Furthermore, the Policy on Children's Health
also applies to this action. Information on how the Policy was applied
is described in section II.A.2 ``Overview of the Health Effects
Evidence'', section II.A.2.b ``Public Health Implications and At-Risk
Populations'' and II.B ``Conclusions on the Primary PM2.5
Standards'' of this preamble.
H. Executive Order 13211: Actions Concerning Regulations That
Significantly Affect Energy Supply, Distribution or Use
This action is not a ``significant energy action'' because it is
not likely to have a significant adverse effect on the supply,
distribution, or use of energy. The purpose of this action is to revise
level of the primary annual PM2.5 NAAQS. The action does not
prescribe specific pollution control strategies by which these ambient
standards and monitoring revisions will be met. Such strategies will be
developed by States on a case-by-case basis, and the EPA cannot predict
whether the control options selected by States will include regulations
on energy suppliers, distributors, or users. Thus, the EPA concludes
that this action does not constitute a significant energy action as
defined in Executive Order 13211.
I. National Technology Transfer and Advancement Act (NTTAA)
This rulemaking involved environmental monitoring or measurement.
The EPA has decided it will continue to use the existing indicators for
fine (PM2.5) and coarse (PM10) particles. The
indicator for fine particles is measured using the Reference Method for
the Determination of Fine Particulate Matter as PM2.5 in the
Atmosphere (appendix L to 40 CFR part 50), which is known as the
PM2.5 FRM, and the indicator for coarse particles is
measured using the Reference Method for the Determination of
Particulate Matter as PM10 in the Atmosphere (appendix J to
40 CFR part 50), which is known as the PM10 FRM.
To the extent feasible, the EPA employs a Performance-Based
Measurement System (PBMS), which does not require the use of specific,
prescribed analytic methods. The PBMS is defined as a set of processes
wherein the data quality needs, mandates or limitations of a program or
project are specified and serve as criteria for selecting appropriate
methods to meet those needs in a cost-effective manner.
[[Page 16375]]
It is intended to be more flexible and cost effective for the regulated
community; it is also intended to encourage innovation in analytical
technology and improved data quality. Though the FRM defines the
particular specifications for ambient monitors, there is some
variability with regard to how monitors measure PM, depending on the
type and size of PM and environmental conditions. Therefore, it is not
practically possible to fully define the FRM in performance terms to
account for this variability. Nevertheless, our approach in the past
has resulted in multiple brands of monitors being approved as FRM for
PM, and we expect this to continue. Also, the FRMs described in 40 CFR
part 50 and the equivalency criteria described in 40 CFR part 53,
constitute a performance-based measurement system for PM, since methods
that meet the field testing and performance criteria can be approved as
FEMs. Since finalized in 2006 (71 FR 61236, October 17, 2006) the new
field and performance criteria for approval of PM2.5
continuous FEMs has resulted in the approval of 13 approved FEMs. In
summary, for measurement of PM2.5 and PM10, the
EPA relies on both FRMs and FEMs, with FEMs relying on a PBMS approach
for their approval. The EPA is not precluding the use of any other
method, whether it constitutes a voluntary consensus standard or not,
as long as it meets the specified performance criteria.
J. Executive Order 12898: Federal Actions To Address Environmental
Justice in Minority Populations and Low-Income Populations and
Executive Order 14096: Revitalizing Our Nation's Commitment to
Environmental Justice for All
The EPA believes that the human health or environmental conditions
associated with the primary PM2.5 NAAQS that exist prior to
this action result in or have the potential to result in
disproportionate and adverse human health or environmental effects on
communities with environmental justice concerns. There is strong
evidence for racial and ethnic disparities in PM2.5
exposures and PM2.5-related health risk, as assessed in the
2019 Integrated Science Assessment and with even more evidence
available since the literature cutoff date for the 2019 Integrated
Science Assessment and evaluated in the Supplement to the 2019
Integrated Science Assessment. There is strong evidence demonstrating
that Black and Hispanic populations, in particular, have higher
PM2.5 exposures than non-Hispanic White populations. Black
populations or individuals that live in predominantly Black
neighborhoods experience higher PM2.5 exposures, in
comparison to non-Hispanic White populations. There is also consistent
evidence across multiple studies that demonstrate increased risk of
PM2.5-related health effects, with the strongest evidence
for health risk disparities for mortality. There is also evidence of
health risk disparities for both Hispanic and non-Hispanic Black
populations compared to non-Hispanic White populations for cause-
specific mortality and incident hypertension.
Socioeconomic status (SES) is a composite measure that includes
metrics such as income, occupation, or education, and can play a role
in access to healthy environments as well as access to healthcare. SES
may be a factor that contributes to differential risk from
PM2.5-related health effects. Studies assessed in the 2019
Integrated Science Assessment and Supplement to the 2019 Integrated
Science Assessment provide evidence that lower SES communities are
exposed to higher concentrations of PM2.5 compared to higher
SES communities. Studies using composite measures of neighborhood SES
consistently demonstrated a disparity in both PM2.5 exposure
and the risk of PM2.5-related health outcomes. There is some
evidence that supports associations larger in magnitude between
mortality and long-term PM2.5 exposures for those with low
income or living in lower income areas compared to those with higher
income or living in higher income neighborhoods. Additionally, evidence
supports conclusions that lower SES is associated with cause-specific
mortality and certain health endpoints (i.e., HI and CHF), but less so
for all-cause or total (non-accidental) mortality.
The EPA believes that this action is likely to reduce existing
disproportionate and adverse effects on communities with environmental
justice concerns.
The EPA additionally identified and addressed environmental justice
concerns by providing opportunities for public input on the proposed
decisions. The EPA held a multi-day virtual public hearing for the
public to provide oral testimony and there was a 60-day public comment
period for the proposed action. As described in section II.A.3 above,
the EPA conducted a risk assessment to support this action that
included an at-risk analysis that evaluates exposure and
PM2.5 mortality risk for older adults (e.g., 65 years and
older), stratified for White, Black, Asian, Native American, Non-
Hispanic, and Hispanic individuals. This at-risk analysis found that
compared to a primary annual PM2.5 standard with a level of
12.0 [micro]g/m\3\, meeting a revised annual standard with a level of
9.0 [micro]g/m\3\ is estimated to reduce PM2.5-associated
health risks in the 30 study areas controlled by the annual standard by
about 22-28% and is expected to reduce disparities in exposure and risk
among these populations.
The information supporting this Executive Order review is is
contained in sections II.A.2, II.B.3.a, II.B.3.c, II.B.2, and II.B.4.
of this preamble and also in the 2019 Integrated Science Assessment,
Supplement to the 2019 Integrated Science Assessment, and 2022 Policy
Assessment. The EPA has carefully evaluated the potential impacts on
minority populations and low SES populations as discussed in sections
II.A.2, II.A.3, II.B.2, and II.B.4 of this preamble. The 2019
Integrated Science Assessment, Supplement to the Integrated Science
Assessment, and 2022 Policy Assessment contain the evaluation of the
scientific evidence, quantitative risk analyses and policy
considerations that pertain to these populations. These documents are
available in the public docket for this action.
K. Congressional Review Act (CRA)
This action is subject to the CRA, and the EPA will submit a rule
report to each House of the Congress and to the Comptroller General of
the United States. This action meets the criteria set forth in 5 U.S.C.
804(2).
References
Abt Associates, Inc. (2001). Assessing public opinions on visibility
impairment due to air pollution: Summary report. U.S. Environmental
Protection Agency. Research Triangle Park, NC. Available at: https://www3.epa.gov/ttn/naaqs/standards/pm/data/vis_rpt_final.pdf.
Abt Associates, Inc. (2005). Particulate matter health risk
assessment for selected urban areas: Draft report. EPA Contract No.
68-D-03-002. U.S. Environmental Protection Agency. Research Triangle
Park, NC. Available at: http://www3.epa.gov/ttn/naaqs/standards/pm/data/PMrisk20051220.pdf.
ATS (2000). What Constitutes an Adverse Health Effect of Air
Pollution? American Journal of Respiratory and Critical Care
Medicine 161(2): 665-673.
BBC Research & Consulting (2003). Phoenix area visibility survey.
Denver, CO. Available at: https://www.regulations.gov/document/EPA-HQ-OAR-2015-0072-0089.
Behbod, B, Urch, B, Speck, M, Scott, JA, Liu, L, Poon, R, Coull, B,
Schwartz, J, Koutrakis, P, Silverman, F and Gold, DR (2013).
Endotoxin in concentrated coarse and fine ambient particles induces
acute systemic inflammation in controlled
[[Page 16376]]
human exposures. Occupational and Environmental Medicine 70(11):
761-767.
Bell, ML, Ebisu, K, Peng, RD, Walker, J, Samet, JM, Zeger, SL and
Dominic, F (2008). Seasonal and regional short-term effects of fine
particles on hospital admissions in 202 U.S. counties, 1999-2005.
American Journal of Epidemiology 168(11): 1301-1310.
Bellavia, A, Urch, B, Speck, M, Brook, RD, Scott, JA, Albetti, B,
Behbod, B, North, M, Valeri, L, Bertazzi, PA, Silverman, F, Gold, D
and Baccarelli, AA (2013). DNA hypomethylation, ambient particulate
matter, and increased blood pressure: Findings from controlled human
exposure experiments. Journal of the American Heart Association
2(3): e000212.
Bennett, JE, Tamura-Wicks, H, Parks, RM, Burnett, RT, Pope, CA,
Bechle, MJ, Marshall, JD, Danaei, G and Ezzati, M (2019).
Particulate matter air pollution and national and county life
expectancy loss in the USA: A spatiotemporal analysis. PLoS Medicine
16(7): e1002856.
Besson, P, Mu[ntilde]oz, C, Ram[iacute]rez-Sagner, G, Salgado, M,
Escobar, R and Platzer, W (2017). Long-Term Soiling Analysis for
Three Photovoltaic Technologies in Santiago Region. IEEE Journal of
Photovoltaics 7(6): 1755-1760.
Bourdrel, T, Annesi-Maesano, I, Alahmad, B, Maesano, CN and Bind, MA
(2021). The impact of outdoor air pollution on COVID-19: a review of
evidence from in vitro, animal, and human studies. 30(159).
Br[auml]uner, EV, M[oslash]ller, P, Barregard, L, Dragsted, LO,
Glasius, M, W[aring]hlin, P, Vinzents, P, Raaschou-Nielsen, O and
Loft, S (2008). Exposure to ambient concentrations of particulate
air pollution does not influence vascular function or inflammatory
pathways in young healthy individuals. Particle and Fibre Toxicology
5: 13.
Brook, RD, Urch, B, Dvonch, JT, Bard, RL, Speck, M, Keeler, G,
Morishita, M, Marsik, FJ, Kamal, AS, Kaciroti, N, Harkema, J, Corey,
P, Silverman, F, Gold, DR, Wellenius, G, Mittleman, MA, Rajagopalan,
S and Brook, JR (2009). Insights into the mechanisms and mediators
of the effects of air pollution exposure on blood pressure and
vascular function in healthy humans. Hypertension 54(3): 659-667.
Burns, J, Boogaard, H, Polus, S, Pfadenhauer, LM, Rohwer, AC, van
Erp, AM, Turley, R and Rehfuess, E (2019). Interventions to reduce
ambient particulate matter air pollution and their effect on health.
Cochrane Database of Systematic Reviews (5).
Cangerana Pereira, FA, Lemos, M, Mauad, T, de Assuncao, JV and
Nascimento Saldiva, PH (2011). Urban, traffic-related particles and
lung tumors in urethane treated mice. Clinics 66(6): 1051-1054.
Chan, EAW, Gantt, B and McDow, S (2018). The reduction of summer
sulfate and switch from summertime to wintertime PM2.5
concentration maxima in the United States. Atmospheric Environment
175: 25-32.
Chen, H, Burnett, RT, Copes, R, Kwong, JC, Villeneuve, PJ, Goldberg,
MS, Brook, RD, van Donkelaar, A, Jerrett, M, Martin, RV, Brook, JR,
Kopp, A and Tu, JV (2016). Ambient fine particulate matter and
mortality among survivors of myocardial infarction: population-based
cohort study. Environmental Health Perspectives 124(9): 1421-1428.
Correia, AW, Pope, CA, III, Dockery, DW, Wang, Y, un, Ezzati, M and
Dominici, F (2013). Effect of air pollution control on life
expectancy in the United States: an analysis of 545 U.S. counties
for the period from 2000 to 2007. Epidemiology 24(1): 23-31.
Corrigan, AE, Becker, MM, Neas, LM, Cascio, WE and Rappold, AG
(2018). Fine particulate matters: The impact of air quality
standards on cardiovascular mortality. Environmental research 161:
364-369.
Cox, LA. (2019a). Letter from Louis Anthony Cox, Jr., Chair, Clean
Air Scientific Advisory Committee, to Administrator Andrew R.
Wheeler. Re: CASAC Review of the EPA's Integrated Science Assessment
for Particulate Matter (External Review Draft--October 2018). April
11, 2019. EPA-CASAC-19-002. Office of the Administrator, Science
Advisory Board U.S. EPA HQ, Washington DC. Available at: https://casac.epa.gov/ords/sab/r/sab_apex/casac/0?report_id=1069&request=APPLICATION_PROCESS%3DREPORT_DOC&session=7184955370570.
Cox, LA. (2019b). Letter from Louis Anthony Cox, Jr., Chair, Clean
Air Scientific Advisory Committee, to Administrator Andrew R.
Wheeler. Re: CASAC Review of the EPA's Policy Assessment for the
Review of the National Ambient Air Quality Standards for Particulate
Matter (External Review Draft--September 2019). December 16, 2019.
EPA-CASAC-20-001. Office of the Administrator, Science Advisory
Board U.S. EPA HQ, Washington DC. Available at: https://casac.epa.gov/ords/sab/r/sab_apex/casac/0?report_id=1073&request=APPLICATION_PROCESS%3DREPORT_DOC&session=6224161457429.
DeFlorio-Barker, S, Crooks, J, Reyes, J and Rappold, AG (2019).
Cardiopulmonary effects of fine particulate matter exposure among
older adults, during wildfire and non-wildfire periods, in the
United States 2008-2010. Environmental health perspectives 127(3):
037006.
DHEW (1969). Air Quality Criteria for Particulate Matter. National
Air Pollution Control Administration. Washington, DC U.S. Department
of Health. January 1969.
Di, Q, Amini, H, Shi, L, Kloog, I, Silvern, R, Kelly, J, Sabath, MB,
Choirat, C, Koutrakis, P and Lyapustin, A (2019). An ensemble-based
model of PM2.5 concentration across the contiguous United
States with high spatiotemporal resolution. Environment
International 130: 104909.
Di, Q, Dai, L, Wang, Y, Zanobetti, A, Choirat, C, Schwartz, JD and
Dominici, F (2017a). Association of short-term exposure to air
pollution with mortality in older adults. JAMA: Journal of the
American Medical Association 318(24): 2446-2456.
Di, Q, Kloog, I, Koutrakis, P, Lyapustin, A, Wang, Y and Schwartz, J
(2016). Assessing PM2.5 exposures with high
spatiotemporal resolution across the Continental United States.
Environmental Science and Technology 50(9): 4712-4721.
Di, Q, Wang, Y, Zanobetti, A, Wang, Y, Koutrakis, P, Choirat, C,
Dominici, F and Schwartz, JD (2017b). Air pollution and mortality in
the Medicare population. New England Journal of Medicine 376(26):
2513-2522.
Dominici, F, Schwartz, J, Di, Q, Braun, D, Choirat, C and Zanobetti,
A (2019). Assessing adverse health effects of long-term exposure to
low levels of ambient air pollution: Phase 1. Health Effects
Institute. Boston, MA. Available at: https://www.healtheffects.org/system/files/dominici-rr-200-report.pdf.
Ely, DW, Leary, JT, Stewart, TR and Ross, DM (1991). The
establishment of the Denver Visibility Standard. Colorado Department
of Health. Denver, Colorado.
Erickson, AC, Christidis, T, Pappin, A, Brook, JR, Crouse, DL,
Hystad, P, Li, C, Martin, RV, Meng, J, Pinault, L, von Donkelaar, A,
Weichenthal, S, Tjepkema, M, Burnett, RT and Brauer, M (2020).
Disease assimilation: The mortality impacts of fine particulate
matter on immigrants to Canada. Health Reports 31(3): 14-26.
Eum, K, Suh, HH, Pun, V and Manjourides, J (2018). Impact of long-
term temporal trends in fine particulate matter (PM2.5)
on association of annual PM2.5 exposure and mortality: an
analysis of over 20 million Medicare beneficiaries. Environmental
Epidemiology 2(2): e009.
Fiore, AM, Naik, V and Leibensperger, EM (2015). Air quality and
climate connections. Journal of the Air and Waste Management
Association 65(6): 645-685.
Frank, N. (2012). Memorandum to PM NAAQS Review Docket (EPA-HQ-OAR-
2007-0492) regarding the Differences between maximum and composite
monitor annual PM2.5 design values by CBSA. Dec 14, 2012.
Docket ID No. EPA-HQ-OAR-2007-0492. Office of Air Quality Planning
and Standards Research Triangle Park, NC. Available at: https://www.regulations.gov/document/EPA-HQ-OAR-2007-0492-10099.
Franklin, M, Zeka, A and Schwartz, J (2007). Association between
PM2.5 and all-cause and specific-cause mortality in 27 US
communities. Journal of Exposure Science and Environmental
Epidemiology 17(3): 279-287.
Gantt, B, Owen, RC and Watkins, N (2021). Characterizing Nitrogen
Oxides and Fine Particulate Matter near Major Highways in the United
States Using the National Near-Road Monitoring Network.
Environmental science & technology 55(5): 2831-2838.
[[Page 16377]]
Ghio, AJ, Hall, A, Bassett, MA, Cascio, WE and Devlin, RB (2003).
Exposure to concentrated ambient air particles alters hematologic
indices in humans. Inhalation Toxicology 15(14): 1465-1478.
Ghio, AJ, Kim, C and Devlin, RB (2000). Concentrated ambient air
particles induce mild pulmonary inflammation in healthy human
volunteers. American Journal of Respiratory and Critical Care
Medicine 162(3): 981-988.
Greven, S, Dominici, F and Zeger, S (2011). An Approach to the
Estimation of Chronic Air Pollution Effects Using Spatio-Temporal
Information. Journal of the American Statistical Association
106(494): 396-406.
Gr[oslash]ntoft, T, Verney-Carron, A and Tidblad, J (2019). Cleaning
Costs for European Sheltered White Painted Steel and Modern Glass
Surfaces Due to Air Pollution Since the Year 2000. Atmosphere 10(4):
167.
Hammer, MS, van Donkelaar, A, Li, C, Lyapustin, A, Sayer, AM, Hsu,
NC, Levy, RC, Garay, MJ, Kalashnikova, OV and Kahn, RA (2020).
Global estimates and long-term trends of fine particulate matter
concentrations (1998-2018). Environmental Science & Technology
54(13): 7879-7890.
Hart, JE, Liao, X, Hong, B, Puett, RC, Yanosky, JD, Suh, H,
Kioumourtzoglou, MA, Spiegelman, D and Laden, F (2015). The
association of long-term exposure to PM2.5 on all-cause
mortality in the Nurses' Health Study and the impact of measurement-
error correction. Environmental Health: A Global Access Science
Source 14: 38.
Hassett-Sipple, B, Schmidt, M and Rajan, P. (2010). Memorandum to PM
NAAQS Review Docket (EPA-HQ-OAR-2007-0492). Analysis of
PM2.5 (Particulate Matter Smaller than 2.5 Micrometers in
Diameter). Mar 30, 2010. Docket ID No. EPA-HQ-OAR-2007-0492. Office
of Air Quality Planning and Standards Research Triangle Park, NC.
Available at: https://www.regulations.gov/document/EPA-HQ-OAR-2007-0492-0077.
Hemmingsen, JG, Jantzen, K, M[oslash]ller, P and Loft, S (2015a). No
oxidative stress or DNA damage in peripheral blood mononuclear cells
after exposure to particles from urban street air in overweight
elderly. Mutagenesis 30(5): 635-642.Hemmingsen, JG, Rissler, J,
Lykkesfeldt, J, Sallsten, G, Kristiansen, J, M[oslash]ller, P and
Loft, S (2015). Controlled exposure to particulate matter from urban
street air is associated with decreased vasodilation and heart rate
variability in overweight and older adults. Particle and Fibre
Toxicology 12(1): 6.
Henneman, LR, Liu, C, Mulholland, JA and Russell, AG (2017).
Evaluating the effectiveness of air quality regulations: A review of
accountability studies and frameworks. Journal of the Air Waste
Management Association 67(2): 144-172.
Henneman, LRF, Choirat, C and Zigler, ACM (2019). Accountability
assessment of health improvements in the United States associated
with reduced coal emissions between 2005 and 2012. Epidemiology
30(4): 477-485.
Hutchinson, JA, Vargo, J, Milet, M, French, NH, Billmire, M,
Johnson, J and Hoshiko, S (2018). The San Diego 2007 wildfires and
Medi-Cal emergency department presentations, inpatient
hospitalizations, and outpatient visits: An observational study of
smoke exposure periods and a bidirectional case-crossover analysis.
PLoS medicine 15(7): e1002601.
IPCC (2013). Climate change 2013: The physical science basis.
Contribution of working group I to the fifth assessment report of
the Intergovernmental Panel on Climate Change. T. F. Stocker, D.
Qin, G. K. Plattner et al. Cambridge University Press. Cambridge,
UK.
Kloog, I, Ridgway, B, Koutrakis, P, Coull, BA and Schwartz, JD
(2013). Long- and short-term exposure to PM2.5 and
mortality: Using novel exposure models. Epidemiology 24(4): 555-561.
Krewski, D, Jerrett, M, Burnett, RT, Ma, R, Hughes, E, Shi, Y,
Turner, MC, Pope, CA, III, Thurston, G, Calle, EE, Thun, MJ,
Beckerman, B, Deluca, P, Finkelstein, N, Ito, K, Moore, DK, Newbold,
KB, Ramsay, T, Ross, Z, Shin, H and Tempalski, B (2009). Extended
follow-up and spatial analysis of the American Cancer Society study
linking particulate air pollution and mortality. ISSN 1041-5505, HEI
Research Report 140. Health Effects Institute. Boston, MA. Available
at: https://www.healtheffects.org/system/files/Krewski140Statement.pdf.
Laden, F, Schwartz, J, Speizer, FE and Dockery, DW (2006). Reduction
in fine particulate air pollution and mortality: extended follow-up
of the Harvard Six Cities study. American Journal of Respiratory and
Critical Care Medicine 173(6): 667-672.
Lavigne, E, Burnett, RT and Weichenthal, S (2018). Association of
short-term exposure to fine particulate air pollution and mortality:
effect modification by oxidant gases. Scientific Reports 8(1):
16097.
Lee, M, Koutrakis, P, Coull, B, Kloog, I and Schwartz, J (2015).
Acute effect of fine particulate matter on mortality in three
Southeastern states from 2007-2011. Journal of Exposure Science and
Environmental Epidemiology 26(2): 173-179.
Lepeule, J, Laden, F, Dockery, D and Schwartz, J (2012). Chronic
exposure to fine particles and mortality: an extended follow-up of
the Harvard Six Cities study from 1974 to 2009. Environmental Health
Perspectives 120(7): 965-970.
Lippmann, M, Chen, LC, Gordon, T, Ito, K and Thurston, GD (2013).
National Particle Component Toxicity (NPACT) Initiative: Integrated
epidemiologic and toxicologic studies of the health effects of
particulate matter components: Investigators' Report. 177. Health
Effects Institute. Boston, MA.
Liu, C, Chen, R, Sera, F, Vicedo-Cabrera, AM, Guo, Y, Tong, S,
Coelho, M, Saldiva, PHN, Lavigne, E, Matus, P, Valdes Ortega, N,
Osorio Garcia, S, Pascal, M, Stafoggia, M, Scortichini, M,
Hashizume, M, Honda, Y, Hurtado-D[iacute]az, M, Cruz, J, Nunes, B,
Teixeira, JP, Kim, H, Tobias, A, [Iacute][ntilde]iguez, C, Forsberg,
B, [Aring]str[ouml]m, C, Ragettli, MS, Guo, YL, Chen, BY, Bell, ML,
Wright, CY, Scovronick, N, Garland, RM, Milojevic, A, Kysel[yacute],
J, Urban, A, Orru, H, Indermitte, E, Jaakkola, JJK, Ryti, NRI,
Katsouyanni, K, Analitis, A, Zanobetti, A, Schwartz, J, Chen, J, Wu,
T, Cohen, A, Gasparrini, A and Kan, H (2019). Ambient Particulate
Air Pollution and Daily Mortality in 652 Cities. New England Journal
of Medicine 381(8): 705-715.
Lowenthal, DH and Kumar, N (2004). Variation of mass scattering
efficiencies in IMPROVE. Journal of the Air and Waste Management
Association (1990-1992) 54(8): 926-934.
Lowenthal, DH and Kumar, N (2016). Evaluation of the IMPROVE
Equation for estimating aerosol light extinction. Journal of the Air
and Waste Management Association 66(7): 726-737.
Lucking, AJ, Lundb[auml]ck, M, Barath, SL, Mills, NL, Sidhu, MK,
Langrish, JP, Boon, NA, Pourazar, J, Badimon, JJ, Gerlofs-Nijland,
ME, Cassee, FR, Boman, C, Donaldson, K, Sandstrom, T, Newby, DE and
Blomberg, A (2011). Particle traps prevent adverse vascular and
prothrombotic effects of diesel engine exhaust inhalation in men.
Circulation 123(16): 1721-1728.
Malm, WC and Hand, JL (2007). An examination of the physical and
optical properties of aerosols collected in the IMPROVE program.
Atmospheric Environment 41(16): 3407-3427.
Malm, WC, Schichtel, B, Molenar, J, Prenni, A and Peters, M (2019).
Which visibility indicators best represent a population's preference
for a level of visual air quality? Journal of the Air & Waste
Management Association 69(2): 145-161.
Malm, WC, Sisler, JF, Huffman, D, Eldred, RA and Cahill, TA (1994).
Spatial and seasonal trends in particle concentration and optical
extinction in the United States. Journal of Geophysical Research
99(D1): 1347-1370.
Mauad, T, Rivero, DH, de Oliveira, RC, Lichtenfels, AJ, Guimaraes,
ET, de Andre, PA, Kasahara, DI, Bueno, HM and Saldiva, PH (2008).
Chronic exposure to ambient levels of urban particles affects mouse
lung development. American Journal of Respiratory and Critical Care
Medicine 178(7): 721-728.
Mie, G (1908). Beitrage zur Optik truber Medien, speziell
kolloidaler Metallosungen [Optics of cloudy media, especially
colloidal metal solutions]. Annalen der Physik 25(3): 377-445.
Miller, KA, Siscovick, DS, Sheppard, L, Shepherd, K, Sullivan, JH,
Anderson, GL and Kaufman, JD (2007). Long-term exposure to air
pollution and incidence of cardiovascular events in women. New
England Journal of Medicine 356(5): 447-458.
Myhre, G, Shindell, D, Br[eacute]on, FM, Collins, W, Fuglestvedt, J,
Huang, J, Koch, D, Lamarque, JF, Lee, D, Mendoza, B, Nakajima, T,
Robock, A, Stephens, G,
[[Page 16378]]
Takemura, T and Zhang, H, Eds. (2013). Anthropogenic and natural
radiative forcing. Cambridge University Press Cambridge, UK.
NHLBI (2017). ``NHLBI fact book, fiscal year 2012: Disease
statistics.'' Retrieved August 23, 2017, from https://web.archive.org/web/20170711012213/http://www.nhlbi.nih.gov/about/documents/factbook/2012/chapter4.
Orr, A, AL Migliaccio, C, Buford, M, Ballou, S and Migliaccio, CT
(2020). Sustained effects on lung function in community members
following exposure to hazardous pm2. 5 levels from wildfire smoke.
Toxics 8(3): 53.
Pitchford, M, Maim, W, Schichtel, B, Kumar, N, Lowenthal, D and
Hand, J (2007). Revised algorithm for estimating light extinction
from IMPROVE particle speciation data. Journal of the Air and Waste
Management Association 57(11): 1326-1336.
Pope, CA, III, Burnett, RT, Thurston, GD, Thun, MJ, Calle, EE,
Krewski, D and Godleski, JJ (2004). Cardiovascular mortality and
long-term exposure to particulate air pollution: epidemiological
evidence of general pathophysiological pathways of disease.
Circulation 109(1): 71-77.
Pope, CA, III, Ezzati, M and Dockery, DW (2009). Fine-particulate
air pollution and life expectancy in the United States. New England
Journal of Medicine 360(4): 376-386.
Pruitt, E. (2018). Memorandum from E. Scott Pruitt, Administrator,
U.S. EPA to Assistant Administrators. Back-to-Basics Process for
Reviewing National Ambient Air Quality Standards. May 9, 2018.
Office of the Administrator U.S. EPA HQ, Washington DC. Available
at: https://www.epa.gov/criteria-air-pollutants/back-basics-process-reviewing-national-ambient-air-quality-standards.
Pryor, SC (1996). Assessing public perception of visibility for
standard setting exercises. Atmospheric Environment 30(15): 2705-
2716.
Puett, RC, Hart, JE, Yanosky, JD, Spiegelman, D, Wang, M, Fisher,
JA, Hong, B and Laden, F (2014). Particulate matter air pollution
exposure, distance to road, and incident lung cancer in the Nurses'
Health Study cohort. Environmental Health Perspectives 122(9): 926-
932.
Pun, VC, Kazemiparkouhi, F, Manjourides, J and Suh, HH (2017). Long-
term PM2.5 exposures and respiratory, cancer and
cardiovascular mortality in American older adults. American Journal
of Epidemiology 186(8): 961-969.
Raaschou-Nielsen, O, Andersen, ZJ, Beelen, R, Samoli, E, Stafoggia,
M, Weinmayr, G, Hoffmann, B, Fischer, P, Nieuwenhuijsen, MJ,
Brunekreef, B, Xun, WW, Katsouyanni, K, Dimakopoulou, K, Sommar, J,
Forsberg, B, Modig, L, Oudin, A, Oftedal, B, Schwarze, PE, Nafstad,
P, De Faire, U, Pedersen, NL, [Ouml]stenson, CG, Fratiglioni, L,
Penell, J, Korek, M, Pershagen, G, Eriksen, KT, S[oslash]rensen, M,
Tj[oslash]nneland, A, Ellermann, T, Eeftens, M, Peeters, PH,
Meliefste, K, Wang, M, Bueno-De-mesquita, B, Key, TJ, De Hoogh, K,
Concin, H, Nagel, G, Vilier, A, Grioni, S, Krogh, V, Tsai, MY,
Ricceri, F, Sacerdote, C, Galassi, C, Migliore, E, Ranzi, A,
Cesaroni, G, Badaloni, C, Forastiere, F, Tamayo, I, Amiano, P,
Dorronsoro, M, Trichopoulou, A, Bamia, C, Vineis, P and Hoek, G
(2013). Air pollution and lung cancer incidence in 17 European
cohorts: Prospective analyses from the European Study of Cohorts for
Air Pollution Effects (ESCAPE). The Lancet Oncology 14(9): 813-822.
Ramanathan, G, Yin, F, Speck, M, Tseng, CH, Brook, JR, Silverman, F,
Urch, B, Brook, RD and Araujo, JA (2016). Effects of urban fine
particulate matter and ozone on HDL functionality. Particle and
Fibre Toxicology 13(1): 26.
Ryan, PA, Lowenthal, D and Kumar, N (2005). Improved light
extinction reconstruction in interagency monitoring of protected
visual environments. Journal of the Air and Waste Management
Association 55(11): 1751-1759.
Sacks, JD, Ito, K, Wilson, WE and Neas, LM (2012). Impact of
covariate models on the assessment of the air pollution-mortality
association in a single- and multipollutant context. American
Journal of Epidemiology 176(7): 622-634.
Sanders, NJ, Barreca, AI and Neidell, MJ (2020a). Estimating causal
effects of particulate matter regulation on mortality. Epidemiology
(Cambridge, Mass.) 31(2): 160.
Sanders, NJ, Barreca, AI and Neidell, MJ (2020b). Estimating Causal
Effects of Particulate Matter Regulation on Mortality. Epidemiology
31(2): 160-167.
Schwartz, J, Austin, E, Bind, MA, Zanobetti, A and Koutrakis, P
(2015). Estimating causal associations of fine particles with daily
deaths in Boston. American Journal of Epidemiology 182(7): 644-650.
Schwartz, J, Bind, MA and Koutrakis, P (2017). Estimating causal
effects of local air pollution on daily deaths: Effect of low
levels. Environmental Health Perspectives 125(1): 23-29.
Schwartz, J, Wei, Y, Di, Q, Dominici, F and Zanobetti, A (2021). A
national difference in differences analysis of the effect of PM2. 5
on annual death rates. Environmental Research 194: 110649.
Schwartz, JD, Wang, Y, Kloog, I, Yitshak-Sade, M, Dominici, F and
Zanobetti, A (2018). Estimating the Effects of PM2.5 on
Life Expectancy Using Causal Modeling Methods. Environmental Health
Perspectives 126(12): 127002.
Sheppard, EA. (2022a). Letter from Elizabeth A. (Lianne) Sheppard,
Chair, Clean Air Scientific Advisory Committee, to Administrator
Michael S. Regan. Re: CASAC Review of the EPA's Policy Assessment
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (External Review Draft--October 2021). March 18,
2022. EPA-CASAC-22-002. Office of the Administrator, Science
Advisory Board U.S. EPA HQ, Washington DC. Available at: https://casac.epa.gov/ords/sab/r/sab_apex/casac/0?report_id=1094&request=APPLICATION_PROCESS%3DREPORT_DOC&session=7184955370570.
Sheppard, EA. (2022b). Letter from Elizabeth A. (Lianne) Sheppard,
Chair, Clean Air Scientific Advisory Committee, to Administrator
Michael S. Regan. Re: CASAC Review of the EPA's Supplement to the
2019 Integrated Science Assessment for Particulate Matter (External
Review Draft--October 2021). EPA-CASAC-22-001. Office of the
Administrator, Science Advisory Board U.S. EPA HQ, Washington DC.
Available at: https://casac.epa.gov/ords/sab/r/sab_apex/casac/0?report_id=1093&request=APPLICATION_PROCESS%3DREPORT_DOC&session=10813926997922.
Shi, L, Zanobetti, A, Kloog, I, Coull, BA, Koutrakis, P, Melly, SJ
and Schwartz, JD (2016). Low-concentration PM2.5 and
mortality: estimating acute and chronic effects in a population-
based study. Environmental Health Perspectives 124(1): 46-52.
Shin, HH, Gogna, P, Maquiling, A, Parajuli, RP, Haque, L and Burr, B
(2021). Comparison of hospitalization and mortality associated with
short-term exposure to ambient ozone and PM2.5 in Canada.
Chemosphere 265: 128683.
Sivagangabalan, G, Spears, D, Masse, S, Urch, B, Brook, RD,
Silverman, F, Gold, DR, Lukic, KZ, Speck, M, Kusha, M, Farid, T,
Poku, K, Shi, E, Floras, J and Nanthakumar, K (2011). The effect of
air pollution on spatial dispersion of myocardial repolarization in
healthy human volunteers. Journal of the American College of
Cardiology 57(2): 198-206.
Smith, AE and Howell, S (2009). An assessment of the robustness of
visual air quality preference study results. CRA International.
Washington, DC. Available at: https://www.regulations.gov/document/EPA-HQ-ORD-2007-0517-0085.
Statistics Canada (2023). Census Profile. 2021 Census of Population.
Statistics Canada Catalogue no. 98-316-X2021001. Ottawa. Released
March 29, 2023. Available at: https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/index.cfm?Lang=E.
Thurston, GD, Kipen, H, Annesi-Maesano, I, Balmes, J, Brook, RD,
Cromar, K, De Matteis, S, Forastiere, F, Forsberg, B, Frampton, MW,
Grigg, J, Heederik, D, Kelly, FJ, Kuenzli, N, Laumbach, R, Peters,
A, Rajagopalan, ST, Rich, D, Ritz, B, Samet, JM, Sandstrom, T,
Sigsgaard, T, Sunyer, J and Brunekreef, B (2017). A joint ERS/ATS
policy statement: what constitutes an adverse health effect of air
pollution? An analytical framework. European Respiratory Journal
49(1): 1600419.
Tong, H, Rappold, AG, Caughey, M, Hinderliter, AL, Bassett, M,
Montilla, T, Case, MW, Berntsen, J, Bromberg, PA, Cascio, WE, Diaz-
Sanchez, D, Devlin, RB and Samet, JM (2015). Dietary
[[Page 16379]]
supplementation with olive oil or fish oil and vascular effects of
concentrated ambient particulate matter exposure in human
volunteers. Environmental Health Perspectives 123(11): 1173-1179.
U.S. EPA (2004a). Air Quality Criteria for Particulate Matter. (Vol
I of II). Office of Research and Development. Research Triangle
Park, NC. U.S. EPA. EPA-600/P-99-002aF. October 2004. Available at:
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100LFIQ.txt.
U.S. EPA (2004b). Air Quality Criteria for Particulate Matter. (Vol
II of II). Office of Research and Development. Research Triangle
Park, NC. U.S. EPA. EPA-600/P-99-002bF. October 2004. Available at:
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100LG7Q.txt.
U.S. EPA (2005). Review of the National Ambient Air Quality
Standards for Particulate Matter: Policy Assessment of Scientific
and Technical Information, OAQPS Staff Paper. Office of Air Quality
Planning and Standards. Research Triangle Park, NC. U.S. EPA. EPA-
452/R-05-005a. December 2005. Available at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P1009MZM.txt.
U.S. EPA (2008). Integrated Review Plan for the National Ambient Air
Quality Standards for Particulate Matter. Office of Research and
Development, National Center for Environmental Assessment; Office of
Air Quality Planning and Standards, Health and Environmental Impacts
Division. Research Triangle Park, NC. U.S. EPA. EPA 452/R-08-004.
March 2008. Available at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P1001FB9.txt.
U.S. EPA (2009a). Particulate Matter National Ambient Air Quality
Standards: Scope and Methods Plan for Health Risk and Exposure
Assessment. Office of Air Quality Planning and Standards, Health and
Environmental Impacts Division. Research Triangle Park, NC. U.S.
EPA. EPA-452/P-09-002. February 2009. Available at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100FLWP.txt.
U.S. EPA (2009b). Particulate Matter National Ambient Air Quality
Standards: Scope and Methods Plan for Urban Visibility Impact
Assessment. Office of Air Quality Planning and Standards, Health and
Environmental Impacts Division. Research Triangle Park, NC. U.S.
EPA. EPA-452/P-09-001. February 2009. Available at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100FLUX.txt.
U.S. EPA (2009c). Integrated Science Assessment for Particulate
Matter (Final Report). Office of Research and Development, National
Center for Environmental Assessment. Research Triangle Park, NC.
U.S. EPA. EPA-600/R-08-139F. December 2009. Available at: https://cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=216546.
U.S. EPA (2010a). Quantitative Health Risk Assessment for
Particulate Matter (Final Report). Office of Air Quality Planning
and Standards, Health and Environmental Impacts Division. Research
Triangle Park, NC. U.S. EPA. EPA-452/R-10-005. June 2010. Available
at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P1007RFC.txt.
U.S. EPA (2010b). Particulate Matter Urban-Focused Visibility
Assessment (Final Document). Office of Air Quality Planning and
Standards, Health and Environmental Impacts Division. Research
Triangle Park, NC. U.S. EPA. EPA-452/R-10-004. July 2010. Available
at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100FO5D.txt.
U.S. EPA (2011). Policy Assessment for the Review of the Particulate
Matter National Ambient Air Quality Standards. Office of Air Quality
Planning and Standards, Health and Environmental Impacts Division.
Research Triangle Park, NC. U.S. EPA. EPA-452/R-11-003. April 2011.
Available at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100AUMY.txt.
U.S. EPA (2012). Responses to Significant Comments on the 2012
Proposed Rule on the National Ambient Air Quality Standards for
Particulate Matter (June 29, 2012; 77 FR 38890). Research Triangle
Park, NC. U.S. EPA. Docket ID No. EPA-HQ-OAR-2007-0492. Available
at: https://www3.epa.gov/ttn/naaqs/standards/pm/data/20121214rtc.pdf.
U.S. EPA (2015). Preamble to the integrated science assessments.
U.S. Environmental Protection Agency, Office of Research and
Development, National Center for Environmental Assessment, RTP
Division. Research Triangle Park, NC. U.S. EPA. EPA/600/R-15/067.
November 2015. Available at: https://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=310244.
U.S. EPA (2016). Integrated review plan for the national ambient air
quality standards for particulate matter. Office of Air Quality
Planning and Standards. Research Triangle Park, NC. U.S. EPA. EPA-
452/R-16-005. December 2016. Available at: https://www3.epa.gov/ttn/naaqs/standards/pm/data/201612-final-integrated-review-plan.pdf.
U.S. EPA (2017). Emissions Inventory Guidance for Implementation of
Ozone and Particulate Matter National Ambient Air Quality Standards
(NAAQS) and Regional Haze Regulations. Office of Air Quality
Planning and Standards, Office of Air and Radiation. Research
Triangle Park, NC. U.S. EPA. U.S. EPA-454/B-17-002. Available at:
https://www.epa.gov/sites/default/files/2017-07/documents/ei_guidance_may_2017_final_rev.pdf.
U.S. EPA (2018a). Technical Assistance Document (TAD) for the
Reporting of Daily Air Quality--the Air Quality Index (AQI). U.S.
Environmental Protection Agency, Office of Air Quality Planning and
Standards. Research Triangle Park, NC. U.S. EPA. EPA 454/B-18-007.
September 2018. Available at: https://www.airnow.gov/sites/default/files/2020-05/aqi-technical-assistance-document-sept2018.pdf.
U.S. EPA (2018b). Modeling Guidance for Demonstrating Air Quality
Goals for Ozone, PM2.5, and Regional Haze. Office of Air
Quality Planning and Standards, Air Quality Policy Division.
Research Triangle Park, NC. U.S. EPA. EPA 454/R-18-009. November
2018. Available at: https://www.epa.gov/sites/default/files/2020-10/documents/o3-pm-rh-modeling_guidance-2018.pdf.
U.S. EPA (2019a). Integrated Science Assessment (ISA) for
Particulate Matter (Final Report). U.S. Environmental Protection
Agency, Office of Research and Development, National Center for
Environmental Assessment. Washington, DC. U.S. EPA. EPA/600/R-19/
188. December 2019. Available at: https://www.epa.gov/naaqs/particulate-matter-pm-standards-integrated-science-assessments-current-review.
U.S. EPA (2019b). PM2.5 Precursor Demonstration Guidance.
Office of Air Quality Planning and Standards, Air Quality Policy
Division. Research Triangle Park, NC. U.S. EPA. EPA-454/R-19-004.
May 2019. Available at: https://nepis.epa.gov/Exe/ZyPDF.cgi/P100YD1Q.PDF?Dockey=P100YD1Q.PDF.
U.S. EPA (2020a). Responses to significant comments on the 2020
proposed rule on the National Ambient Air Quality Standards for
particulate matter (April 30, 2020; 85 FR 24094). EPA-HQ-OAR-2015-
0072. Available at: https://www.epa.gov/sites/production/files/2020-12/documents/pm_naaqs_response_to_comments_final.pdf.
U.S. EPA (2020b). Policy Assessment for the Review of the National
Ambient Air Quality Standards for Particulate Matter. Office of Air
Quality Planning and Standards, Health and Environmental Impacts
Division. Research Triangle Park, NC. U.S. EPA. EPA-452/R-20-002.
January 2020. Available at: https://www.epa.gov/system/files/documents/2021-10/final-policy-assessment-for-the-review-of-the-pm-naaqs-01-2020.pdf.
U.S. EPA (2021a). Supplement to the 2019 Integrated Science
Assessment for Particulate Matter (External Review Draft). U.S.
Environmental Protection Agency, Office of Research and Development,
Center for Public Health and Environmental Assessment. Research
Triangle Park, NC. U.S. EPA. EPA/600/R-21/198. December 2019.
Available at: https://www.epa.gov/naaqs/particulate-matter-pm-standards-integrated-science-assessments-current-review.
U.S. EPA (2021b). Comparative Assessment of the Impacts of
Prescribed Fire Versus Wildfire (CAIF): A Case Study in the Western
U.S. U.S. Environmental Protection Agency. Washington, DC. U.S. EPA.
EPA/600/R-21/197.
U.S. EPA (2021c). Policy Assessment for the Review of the National
Ambient Air Quality Standards for Particulate Matter (External
Review Draft). Office of Air Quality Planning and Standards, Health
and Environmental Impacts Division. Research Triangle Park, NC. U.S.
EPA. EPA-452/P-21-001. October 2021. Available at: https://
www.epa.gov/system/files/documents/2022-05/
[[Page 16380]]
Final%20Policy%20Assessment%20for%20the%20Reconsideration%20of%20the%
20PM%20NAAQS_May2022_0.pdf.
U.S. EPA (2021d). Guidance for Ozone and Fine Particulate Matter
Permit Modeling. U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Air Quality Assessment Division.
Research Triangle Park, NC. U.S. EPA. EPA-454/P-21-001. September
2021.
U.S. EPA (2022a). Policy Assessment for the Reconsideration of the
National Ambient Air Quality Standards for Particulate Matter.
Office of Air Quality Planning and Standards, Health and
Environmental Impacts Division. Research Triangle Park, NC. U.S.
EPA. EPA-452/R-22-004. May 2022. Available at: https://www.epa.gov/system/files/documents/2022-05/Final%20Policy%20Assessment%20for%20the%20Reconsideration%20of%20the%20PM%20NAAQS_May2022_0.pdf.
U.S. EPA (2022b). Supplement to the 2019 Integrated Science
Assessment for Particulate Matter (Final Report). U.S. Environmental
Protection Agency, Office of Research and Development, Center for
Public Health and Environmental Assessment. Research Triangle Park,
NC. U.S. EPA. EPA/600/R 22/028. May 2022. Available at: https://www.epa.gov/naaqs/particulate-matter-pm-standards-integrated-science-assessments-current-review.
Urch, B, Speck, M, Corey, P, Wasserstein, D, Manno, M, Lukic, KZ,
Brook, JR, Liu, L, Coull, B, Schwartz, J, Gold, DR and Silverman, F
(2010). Concentrated ambient fine particles and not ozone induce a
systemic interleukin-6 response in humans. Inhalation Toxicology
22(3): 210-218.
Van de Hulst, H (1981). Light scattering by small particles. Dover
Publications, Inc. New York.
van Donkelaar, A, Martin, RV, Li, C and Burnett, RT (2019). Regional
estimates of chemical composition of fine particulate matter using a
combined geoscience-statistical method with information from
satellites, models, and monitors. Environmental Science & Technology
53(5).
Vieira, JL, Guimaraes, GV, de Andre, PA, Cruz, FD, Nascimento
Saldiva, PH and Bocchi, EA (2016a). Respiratory filter reduces the
cardiovascular effects associated with diesel exhaust exposure a
randomized, prospective, double-blind, controlled study of heart
failure: the FILTER-HF trial. JACC: Heart Failure 4(1): 55-64.
Vieira, JL, Guimaraes, GV, de Andre, PA, Nascimento Saldiva, PH and
Bocchi, EA (2016b). Effects of reducing exposure to air pollution on
submaximal cardiopulmonary test in patients with heart failure:
Analysis of the randomized, double-blind and controlled FILTER-HF
trial. International Journal of Cardiology 215: 92-97.
Wang, Y, Shi, L, Lee, M, Liu, P, Di, Q, Zanobetti, A and Schwartz,
JD (2017). Long-term exposure to PM2.5 and mortality
among older adults in the Southeastern US. Epidemiology 28(2): 207-
214.
Ward-Caviness, CK, Weaver, AM, Buranosky, M, Pfaff, ER, Neas, LM,
Devlin, RB, Schwartz, J, Di, Q, Cascio, WE and Diaz-Sanchez, D
(2020). Associations between long-term fine particulate matter
exposure and mortality in heart failure patients. Journal of the
American Heart Association 9(6): e012517.
Wei, Y, Yazdi, MD, Di, Q, Requia, WJ, Dominici, F, Zanobetti, A and
Schwartz, J (2021). Emulating causal dose-response relations between
air pollutants and mortality in the Medicare population.
Environmental Health: A Global Access Science Source 20(1): 53.
Wu, X, Braun, D, Schwartz, J, Kioumourtzoglou, MA and Dominici, F
(2020). Evaluating the impact of long-term exposure to fine
particulate matter on mortality among the elderly. Science Advances
6(29): eaba5692.
Wyatt, LH, Devlin, RB, Rappold, AG, Case, MW and Diaz-Sanchez, D
(2020). Low levels of fine particulate matter increase vascular
damage and reduce pulmonary function in young healthy adults.
Particle and fibre toxicology 17(1): 1-12.
Yorifuji, T, Kashima, S and Doi, H (2016). Fine-particulate air
pollution from diesel emission control and mortality rates in Tokyo:
a quasi-experimental study. Epidemiology 27(6): 769-778.
Zanobetti, A and Schwartz, J (2009). The effect of fine and coarse
particulate air pollution on mortality: A national analysis.
Environmental Health Perspectives 117(6): 1-40.
Zhang, Z, Wang, J, Kwong, JC, Burnett, RT, van Donkelaar, A, Hystad,
P, Martin, RV, Bai, L, McLaughlin, J and Chen, H (2021). Long-term
exposure to air pollution and mortality in a prospective cohort: The
Ontario Health Study. Environment International 154: 106570.
List of Subjects
40 CFR Part 50
Environmental protection, Air pollution control, Carbon monoxide,
Lead, Nitrogen dioxide, Ozone, Particulate matter, Sulfur oxides.
40 CFR Part 53
Environmental protection, Administrative practice and procedure,
Air pollution control, Reporting and recordkeeping requirements.
40 CFR Part 58
Environmental protection, Administrative practice and procedure,
Air pollution control, Intergovernmental relations, Reporting and
recordkeeping requirements.
Michael S. Regan,
Administrator.
For the reasons set forth in the preamble, chapter I of title 40 of
the Code of Federal Regulations is amended as follows:
PART 50--NATIONAL PRIMARY AND SECONDARY AMBIENT AIR QUALITY
STANDARDS
0
1. The authority citation for part 50 continues to read as follows:
Authority: 42 U.S.C. 7401, et seq.
0
2. Add Sec. 50.20 to read as follows:
Sec. 50.20 National primary ambient air quality standards for PM2.5.
(a) The national primary ambient air quality standards for
PM2.5 are 9.0 micrograms per cubic meter ([micro]g/m\3\)
annual arithmetic mean concentration and 35 [micro]g/m\3\ 24-hour
average concentration measured in the ambient air as PM2.5
(particles with an aerodynamic diameter less than or equal to a nominal
2.5 micrometers) by either:
(1) A reference method based on appendix L to this part and
designated in accordance with part 53 of this chapter; or
(2) An equivalent method designated in accordance with part 53 of
this chapter.
(b) The primary annual PM2.5 standard is met when the
annual arithmetic mean concentration, as determined in accordance with
appendix N to this part, is less than or equal to 9.0 [micro]g/m\3\.
(c) The primary 24-hour PM2.5 standard is met when the
98th percentile 24-hour concentration, as determined in accordance with
appendix N to this part, is less than or equal to 35 [micro]g/m\3\.
0
3. Amend appendix K to part 50 by:
0
a. In section 1.0 revising paragraph (b);
0
b. In section 2.3 adding paragraph (d); and
0
c. In section 3.0 adding paragraphs (a) and (b).
The revision and additions read as follows:
Appendix K to Part 50--Interpretation of the National Ambient Air
Quality Standards for Particulate Matter
1.0 General
* * * * *
(b) The terms used in this appendix are defined as follows:
Average refers to the arithmetic mean of the estimated number of
exceedances per year, as per section 3.1 of this appendix.
Collocated monitors refer to two or more air measurement
instruments for the same parameter (e.g., PM10 mass)
operated at the same site location, and whose placement is
consistent with part 53 of this chapter. For purposes of considering
a combined site record in this appendix, when two or more monitors
are operated at the same site, one
[[Page 16381]]
monitor is designated as the ``primary'' monitor with any additional
monitors designated as ``collocated.'' It is implicit in these
appendix procedures that the primary monitor and collocated
monitor(s) are all reference or equivalent methods; however, it is
not a requirement that the primary and collocated monitors utilize
the same specific sampling and analysis method.
Combined site data record is the data set used for performing
computations in this appendix and represents data for the primary
monitors augmented with data from collocated monitors according to
the procedure specified in section 3.0(a) of this appendix.
Daily value for PM10 refers to the 24-hour average
concentration of PM10 calculated or measured from
midnight to midnight (local time).
Exceedance means a daily value that is above the level of the
24-hour standard after rounding to the nearest 10 [micro]g/m\3\
(i.e., values ending in 5 or greater are to be rounded up).
Expected annual value is the number approached when the annual
values from an increasing number of years are averaged, in the
absence of long-term trends in emissions or meteorological
conditions.
Primary monitors are suitable monitors designated by a State or
local agency in their annual network plan as the default data source
for creating a combined site data record. If there is only one
suitable monitor at a particular site location, then it is presumed
to be a primary monitor.
Year refers to a calendar year.
* * * * *
2.3 Data Requirements
* * * * *
(d) 24-hour average concentrations will be computed from
submitted hourly PM10 concentration data for each
corresponding day of the year and the result will be stored in the
first, or start, hour (i.e., midnight, hour `0') of the 24-hour
period. A 24-hour average concentration shall be considered valid if
at least 75 percent of the hourly averages (i.e., 18 hourly values)
for the 24-hour period are available. In the event that fewer than
all 24 hourly average concentrations are available (i.e., fewer than
24 but at least 18), the 24-hour average concentration shall be
computed on the basis of the hours available using the number of
available hours within the 24-hour period as the divisor (e.g., the
divisor is 19 if 19 hourly values are available). 24-hour periods
with 7 or more missing hours shall also be considered for
computations in this appendix if, after substituting zero for all
missing hourly concentrations, the resulting 24-hour average daily
value exceeds the level of the 24-hour standard specified in Sec.
50.6 after rounding to the nearest 10 [micro]g/m\3\.
* * * * *
3.0 Computational Equations for the 24-Hour Standards
(a) All computations shown in this appendix shall be implemented
on a site-level basis. Site level concentration data shall be
processed as follows:
(1) The default dataset for PM10 mass concentrations
for a site shall consist of the measured concentrations recorded
from the designated primary monitor(s). All daily values produced by
the primary monitor are considered part of the site record.
(2) If a daily value is not produced by the primary monitor for
a particular day, but a value is available from a single collocated
monitor, then that collocated monitor value shall be considered part
of the combined site data record. If daily value data is available
from two or more collocated monitors, the average of those
collocated values shall be used as the daily value. The data record
resulting from this procedure is referred to as the ``combined site
data record.''
(b) In certain circumstances, including but not limited to site
closures or relocations, data from two nearby sites may be combined
into a single site data record for the purpose of calculating a
valid design value. The appropriate Regional Administrator may
approve such combinations if the Regional Administrator determines
that the measured concentrations do not differ substantially between
the two sites, taking into consideration factors such as distance
between sites, spatial and temporal patterns in air quality, local
emissions and meteorology, jurisdictional boundaries, and terrain
features.
* * * * *
0
4. Amend appendix L to part 50 by revising section 7.3.4 and adding
section 7.3.4.5 to read as follows:
Appendix L to Part 50--Reference Method for the Determination of Fine
Particulate Matter as PM[bdi2].[bdi5] in the Atmosphere
* * * * *
7.3.4 Particle size separator. The sampler shall be configured
with one of the three alternative particle size separators described
in this section. One separator is an impactor-type separator (WINS
impactor) described in sections 7.3.4.1, 7.3.4.2, and 7.3.4.3 of
this appendix. One alternative separator is a cyclone-type separator
(VSCC\TM\) described in section 7.3.4.4 of this appendix. The other
alternative separator is also a cyclone-type separator (TE-
PM2.5C) described in section 7.3.4.5 of this appendix.
* * * * *
7.3.4.5 A second cyclone-type separator is identified as a Tisch
TE-PM2.5C Cyclone particle size separator specified as
part of EPA-designated reference method RFPS-1014-219 and as
manufactured by Tisch Environmental Incorporated, 145 S. Miami
Avenue, Village of Cleves, Ohio 45002.
* * * * *
0
5. Amend appendix N to part 50 by:
0
a. In section 1.0 revising paragraph (a);
0
b. In section 3.0 adding paragraph (d)(3);
0
c. In section 4.1 revising paragraph (a); and
0
d. In section 4.2 revising paragraph (a).
The addition and revisions read as follows.
Appendix N to Part 50--Interpretation of the National Ambient Air
Quality Standards for PM[bdi2].[bdi5]
1.0 General
(a) This appendix explains the data handling conventions and
computations necessary for determining when the national ambient air
quality standards (NAAQS) for PM2.5 are met, specifically
the primary and secondary annual and 24-hour PM2.5 NAAQS
specified in Sec. Sec. 50.7, 50.13, 50.18, and 50.20.
PM2.5 is defined, in general terms, as particles with an
aerodynamic diameter less than or equal to a nominal 2.5
micrometers. PM2.5 mass concentrations are measured in
the ambient air by a Federal Reference Method (FRM) based on
appendix L to this part, as applicable, and designated in accordance
with part 53 of this chapter or by a Federal Equivalent Method (FEM)
designated in accordance with part 53 of this chapter. Only those
FRM and FEM measurements that are derived in accordance with part 58
of this chapter (i.e., that are deemed ``suitable'') shall be used
in comparisons with the PM2.5 NAAQS. The data handling
and computation procedures to be used to construct annual and 24-
hour NAAQS metrics from reported PM2.5 mass
concentrations, and the associated instructions for comparing these
calculated metrics to the levels of the PM2.5 NAAQS, are
specified in sections 2.0, 3.0, and 4.0 of this appendix.
* * * * *
3.0 Requirements for Data Use and Data Reporting for Comparisons With
the NAAQS for PM[bdi2].[bdi5]
* * * * *
(d) * * *
(3) In certain circumstances, including but not limited to site
closures or relocations, data from two nearby sites may be combined
into a single site data record for the purpose of calculating a
valid design value. The appropriate Regional Administrator may
approve such site combinations if the Regional Administrator
determines that the measured concentrations do not differ
substantially between the two sites, taking into consideration
factors such as distance between sites, spatial and temporal
patterns in air quality, local emissions and meteorology,
jurisdictional boundaries, and terrain features.
* * * * *
4.1 Annual PM[bdi2].[bdi5] NAAQS
(a) Levels of the primary and secondary annual PM2.5
NAAQS are specified in Sec. Sec. 50.7, 50.13, 50.18, and 50.20 as
applicable.
* * * * *
4.2 Twenty-Four-Hour PM[bdi2].[bdi5] NAAQS
(a) Levels of the primary and secondary 24-hour PM2.5
NAAQS are specified in Sec. Sec. 50.7, 50.13, 50.18, and 50.20 as
applicable.
* * * * *
[[Page 16382]]
PART 53--AMBIENT AIR MONITORING REFERENCE AND EQUIVALENT METHODS
0
6. The authority citation for part 53 continues to read as follows:
Authority: Sec. 301(a) of the Clean Air Act (42 U.S.C.
1857g(a)), as amended by sec. 15(c)(2) of Pub. L. 91-604, 84 Stat.
1713, unless otherwise noted.
Subpart A--General Provisions
0
7. Amend Sec. 53.4 by:
0
a. Revising paragraph (a);
0
b. Adding paragraph (b)(7); and
0
c. Revising paragraph (d).
The revisions and addition read as follows:
Sec. 53.4 Applications for reference or equivalent method
determinations.
(a) Applications for FRM or FEM determinations and modification
requests of existing designated instruments shall be submitted to: U.S.
Environmental Protection Agency, Director, Center for Environmental
Measurement and Modeling, Reference and Equivalent Methods Designation
Program (MD-D205-03), 109 T.W. Alexander Drive, P.O. Box 12055,
Research Triangle Park, North Carolina 27711 (commercial delivery
address: 4930 Old Page Road, Durham, North Carolina 27703).
* * * * *
(b) * * *
(7) All written materials for new FRM and FEM applications and
modification requests must be submitted in English in MS Word format.
For any calibration certificates originally written in a non-English
language, the original non-English version of the certificate must be
submitted to EPA along with a version of the certificate translated to
English. All laboratory and field data associated with new FRM and FEM
applications and modification requests must be submitted in MS Excel
format. All worksheets in MS Excel must be unprotected to enable full
inspection as part of the application review process.
* * * * *
(d) For candidate reference or equivalent methods or for designated
instruments that are the subject of a modification request, the
applicant, if requested by EPA, shall provide to EPA a representative
sampler or analyzer for test purposes. The sampler or analyzer shall be
shipped free on board (FOB) destination to Director, Center for
Environmental Measurements and Modeling, Reference and Equivalent
Methods Designation Program (MD D205-03), U.S. Environmental Protection
Agency, 4930 Old Page Road, Durham, North Carolina 27703, scheduled to
arrive concurrently with or within 30 days of the arrival of the other
application materials. This sampler or analyzer may be subjected to
various tests that EPA determines to be necessary or appropriate under
Sec. 53.5(f), and such tests may include special tests not described
in this part. If the instrument submitted under this paragraph (d)
malfunctions, becomes inoperative, or fails to perform as represented
in the application before the necessary EPA testing is completed, the
applicant shall be afforded the opportunity to repair or replace the
device at no cost to the EPA. Upon completion of EPA testing, the
sampler or analyzer submitted under this paragraph (d) shall be
repacked by EPA for return shipment to the applicant, using the same
packing materials used for shipping the instrument to EPA unless
alternative packing is provided by the applicant. Arrangements for, and
the cost of, return shipment shall be the responsibility of the
applicant. The EPA does not warrant or assume any liability for the
condition of the sampler or analyzer upon return to the applicant.
0
8. Amend Sec. 53.8 by revising paragraph (a) to read as follows:
Sec. 53.8 Designation of reference and equivalent methods.
(a) A candidate method determined by the Administrator to satisfy
the applicable requirements of this part shall be designated as an FRM
or FEM (as applicable) by and upon publication of the designation in
the Federal Register. Applicants shall not publicly announce, market,
or sell the candidate sampler and analyzer as an approved FRM or FEM
(as applicable) until the designation is published in the Federal
Register.
* * * * *
0
9. Amend Sec. 53.14 by revising paragraphs (c)(4), (5), and (6) to
read as follows:
Sec. 53.14 Modification of a reference or equivalent method.
* * * * *
(c) * * *
(4) Send notice to the applicant that additional information must
be submitted before a determination can be made and specify the
additional information that is needed (in such cases, the 90-day period
shall commence upon receipt of the additional information).
(5) Send notice to the applicant that additional tests are
necessary and specify which tests are necessary and how they shall be
interpreted (in such cases, the 90-day period shall commence upon
receipt of the additional test data).
(6) Send notice to the applicant that additional tests will be
conducted by the Administrator and specify the reasons for and the
nature of the additional tests (in such cases, the 90-day period shall
commence 1 calendar day after the additional tests are completed).
* * * * *
0
10. Revise table A-1 to subpart A of part 53 to read as follows:
Table A-1 to Subpart A of Part 53--Summary of Applicable Requirements for Reference and Equivalent Methods for Air Monitoring of Criteria Pollutants
--------------------------------------------------------------------------------------------------------------------------------------------------------
Applicable Applicable subparts of this part
Pollutant Reference or Manual or automated appendix of part 50 -----------------------------------------------------------------------
equivalent of this chapter A B C D E F
--------------------------------------------------------------------------------------------------------------------------------------------------------
SO2.............. Reference.......... Manual............. A-2
Automated.......... A-1 [check] [check]
Equivalent......... Manual............. A-1 [check] .......... [check]
Automated.......... A-1 [check] [check] [check]
CO............... Reference.......... Automated.......... C [check] [check]
Equivalent......... Manual............. C [check] .......... [check]
Automated.......... C [check] [check] [check]
O3............... Reference.......... Automated.......... D [check] [check]
Equivalent......... Manual............. D [check] .......... [check]
Automated.......... D [check] [check] [check]
NO2.............. Reference.......... Automated.......... F [check] [check]
Equivalent......... Manual............. F [check] .......... [check]
Automated.......... F [check] [check] [check]
[[Page 16383]]
Pb............... Reference.......... Manual............. G
Equivalent......... Manual............. G [check] .......... [check]
Automated.......... G [check] .......... [check]
PM10-Pb.......... Reference.......... Manual............. Q
Equivalent......... Manual............. Q [check] .......... [check]
Automated.......... Q [check] .......... [check]
PM10............. Reference.......... Manual............. J [check] .......... .......... [check]
Equivalent......... Manual............. J [check] .......... [check] [check]
Automated.......... J [check] .......... [check] [check]
PM2.5............ Reference.......... Manual............. L [check] .......... .......... .......... [check]
Equivalent Class I. Manual............. L [check] .......... [check] .......... [check]
Equivalent Class II Manual............. L \1\ [check] .......... \2\ .......... [check] 1 2
[check] [check]
Equivalent Class Automated.......... L \1\ [check] .......... [check] .......... [check] \1\
III. [check]
PM10-2.5......... Reference.......... Manual............. L,\2\ O [check] .......... .......... .......... [check]
Equivalent Class I. Manual............. L,\2\ O [check] .......... [check] .......... [check]
Equivalent Class II Manual............. L,\2\ O [check] .......... \2\ .......... [check] \1,2\
[check] [check]
Equivalent Class Automated.......... \1\ L, 1 2 O [check] .......... [check] .......... [check] \1\
III. [check]
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Some requirements may apply, based on the nature of each particular candidate method, as determined by the Administrator.
\2\ Alternative Class III requirements may be substituted.
Subpart B--Procedures for Testing Performance Characteristics of
Automated Methods for SO2, CO, O3, and NO2
0
11. Amend table B-1 to subpart B of part 53 by revising footnote 4 to
read as follows:
Table B-1 to Subpart B of Part 53--Performance Limit Specifications for
Automated Methods
* * * * *
\4\ For nitric oxide interference for the SO2
ultraviolet fluorescence (UVF) method, interference equivalent is
0.003 ppm for the lower range.
* * * * *
0
12. Revise table B-3 to subpart B of part 53 to read as follows:
[[Page 16384]]
Table B-3 to Subpart B of Part 53--Interferent Test Concentration 1
[Parts per million]
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Hydro-
Pollutant Analyzer type chloric Ammonia Hydrogen Sulfur Nitrogen Nitric Carbon Ethylene Ozone M-xylene Water Carbon Methane Ethane Naphthalene
\2\ acid sulfide dioxide dioxide oxide dioxide vapor monoxide
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
SO2........... Ultraviolet ........ ........ \5\ 0.1 \4\ 0.14 0.5 0.5 ......... ......... 0.5 0.2 20,000 ......... ........ ........ \6\ 0.05
fluorescence.
SO2........... Flame ........ ........ 0.01 \4\ 0.14 ......... ........ 750 ......... ........ ......... \3\ 50 ........ ........ ............
photometric. 20,000
SO2........... Gas ........ ........ 0.1 \4\ 0.14 ......... ........ 750 ......... ........ ......... \3\ 50 ........ ........ ............
chromatography. 20,000
SO2........... Spectrophotometr 0.2 0.1 0.1 \4\ 0.14 0.5 ........ 750 ......... 0.5 ......... ........ ......... ........ ........ ............
ic-wet chemical
(pararosanaline
).
SO2........... Electrochemical. 0.2 0.1 0.1 \4\ 0.14 0.5 0.5 ......... 0.2 0.5 ......... \3\ ......... ........ ........ ............
20,000
SO2........... Conductivity.... 0.2 0.1 ......... \4\ 0.14 0.5 ........ 750 ......... ........ ......... ........ ......... ........ ........ ............
SO2........... Spectrophotometr ........ ........ ......... \4\ 0.14 0.5 0.5 ......... ......... 0.5 0.2 ........ ......... ........ ........ ............
ic-gas phase,
including DOAS.
O3............ Ethylene ........ ........ \3\ 0.1 ........ ......... ........ 750 ......... \4\ 0.08 ......... \3\ ......... ........ ........ ............
Chemiluminescen 20,000
ce.
O3............ NO- ........ ........ \3\ 0.1 ........ 0.5 ........ 750 ......... \4\ 0.08 ......... \3\ ......... ........ ........ ............
chemiluminescen 20,000
ce.
O3............ Electrochemical. ........ \3\ 0.1 ......... 0.5 0.5 ........ ......... ......... \4\ 0.08 ......... \3\ ......... ........ ........ ............
20,000
O3............ Spectrophotometr ........ \3\ 0.1 ......... 0.5 0.5 \3\ 0.5 ......... ......... \4\ 0.08 ......... ........ ......... ........ ........ ............
ic-wet chemical
(potassium
iodide).
O3............ Spectrophotometr ........ ........ ......... 0.5 0.5 \3\ 0.5 ......... ......... \4\ 0.08 0.02 20,000 ......... ........ ........ ............
ic-gas phase,
including
ultraviolet
absorption and
DOAS.
CO............ Non-dispersive ........ ........ ......... ........ ......... ........ 750 ......... ........ ......... 20,000 \4\ 10 ........ ........ ............
Infrared.
CO............ Gas ........ ........ ......... ........ ......... ........ ......... ......... ........ ......... 20,000 \4\ 10 ........ 0.5 ............
chromatography
with flame
ionization
detector.
CO............ Electrochemical. ........ ........ ......... ........ ......... 0.5 ......... 0.2 ........ ......... 20,000 \4\ 10 ........ ........ ............
CO............ Catalytic ........ 0.1 ......... ........ ......... ........ 750 0.2 ........ ......... 20,000 \4\ 10 5.0 0.5 ............
combustion-
thermal
detection.
CO............ IR fluorescence. ........ ........ ......... ........ ......... ........ 750 ......... ........ ......... 20,000 \4\ 10 ........ 0.5 ............
CO............ Mercury ........ ........ ......... ........ ......... ........ ......... 0.2 ........ ......... ........ \4\ 10 ........ 0.5 ............
replacement-UV
photometric.
NO2........... Chemiluminescent ........ \3\ 0.1 ......... 0.5 \4\ 0.1 0.5 ......... ......... ........ ......... 20,000 ......... ........ ........ ............
NO2........... Spectrophotometr ........ ........ ......... 0.5 \4\ 0.1 0.5 750 ......... 0.5 ......... ........ ......... ........ ........ ............
ic-wet chemical
(azo-dye
reaction).
NO2........... Electrochemical. 0.2 \3\ 0.1 ......... 0.5 \4\ 0.1 0.5 750 ......... 0.5 ......... 20,000 50 ........ ........ ............
NO2........... Spectrophotometr ........ \3\ 0.1 ......... 0.5 \4\ 0.1 0.5 ......... ......... 0.5 ......... 20,000 50 ........ ........ ............
ic-gas phase.
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Concentrations of interferent listed must be prepared and controlled to 10 percent of the stated value.
\2\ Analyzer types not listed will be considered by the Administrator as special cases.
\3\ Do not mix interferent with the pollutant.
\4\ Concentration of pollutant used for test. These pollutant concentrations must be prepared to 10 percent of the stated value.
\5\ If candidate method utilizes an elevated-temperature scrubber for removal of aromatic hydrocarbons, perform this interference test.
\6\ If naphthalene test concentration cannot be accurately quantified, remove the scrubber, use a test concentration that causes a full-scale response, reattach the scrubber, and evaluate
response for interference.
[[Page 16385]]
0
13. Amend appendix A to subpart B of part 53 by revising figures B-3
and B-5 to read as follows:
Appendix A to Subpart B of Part 53--Optional Forms for Reporting Test
Results
* * * * *
Figure B-3 to Appendix A to Subpart B of Part 53--Form for Test Data
and Calculations for Lower Detectable Limit (LDL) and Interference
Equivalent (IE) (see Sec. 53.23(c) and (d))
LDL Interference Test Data
Applicant--------------------------------------------------------------
Analyzer---------------------------------------------------------------
Date-------------------------------------------------------------------
Pollutant--------------------------------------------------------------
[GRAPHIC] [TIFF OMITTED] TR06MR24.035
* * * * *
Figure B-5 to Appendix A to Subpart B of Part 53--Form for Calculating
Zero Drift, Span Drift and Precision (see Sec. 53.23(e))
Calculation of Zero Drift, Span Drift, and Precision
Applicant--------------------------------------------------------------
Analyzer---------------------------------------------------------------
Date-------------------------------------------------------------------
Pollutant--------------------------------------------------------------
[GRAPHIC] [TIFF OMITTED] TR06MR24.036
[[Page 16386]]
* * * * *
Subpart C--Procedures for Determining Comparability Between
Candidate Methods and Reference Methods
0
14. Amend Sec. 53.35 by revising paragraph (b)(1)(ii)(D) to read as
follows:
Sec. 53.35 Test procedure for Class II and Class III methods for
PM2.5 and PM10-2.5.
* * * * *
(b) * * *
(1) * * *
(ii) * * *
(D) Site D shall be in a large city east of the Mississippi River,
having characteristically high humidity levels.
* * * * *
0
15. Revise table C-4 to subpart C of part 53 to read as follows:
Table C-4 to Subpart C of Part 53--Test Specifications for PM10, PM2.5, and PM10-2.5 Candidate Equivalent Methods
--------------------------------------------------------------------------------------------------------------------------------------------------------
PM2.5 PM10-2.5
Specification PM10 ---------------------------------------------------------------------------------------------------
Class I Class II Class III Class II Class III
--------------------------------------------------------------------------------------------------------------------------------------------------------
Acceptable concentration range 5-300............. 3-200............. 3-200............. 3-200............. 3-200............. 3-200.
(Rj), [micro]g/m\3\.
Minimum number of test sites.... 2................. 1................. 2................. 4................. 2................. 4.
Minimum number of candidate 3................. 3................. 3\1\.............. 3\1\.............. 3\1\.............. 3.\1\
method samplers or analyzers
per site.
Number of reference method 3................. 3................. 3\1\.............. 3\1\.............. 3\1\.............. 3.\1\
samplers per site.
Minimum number of acceptable
sample sets per site for PM10
methods:
Rj < 20 [micro]g/m\3\....... 3................. .................. .................. .................. .................. ..................
Rj > 20 [micro]g/m\3\....... 3................. .................. .................. .................. .................. ..................
Total................... 10................ .................. .................. .................. .................. ..................
Minimum number of acceptable
sample sets per site for PM2.5
and PM10-2.5 candidate
equivalent methods:
Rj < 15 [micro]g/m\3\ for 24- .................. 3................. 3................. 3................. 3................. 3.
hr or Rj < 8 [micro]g/m\3\
for 48-hr samples..
Rj > 15 [micro]g/m\3\ for 24- .................. 3................. 3................. 3................. 3................. 3.
hr or Rj > 8 [micro]g/m\3\
for 48-hr samples.
Each season................. .................. 10................ 23................ 23................ 23................ 23.
Total, each site........ .................. 10................ 23................ 23 (46 for two- 23................ 23 (46 for two-
season sites). season sites).
Precision of replicate reference 5 [mu]g/m\3\ or 2 [mu]g/m\3\ or 10%\2\............ 10%\2\............ 10%\2\............ 10%.\2\
method measurements, PRj or 7%.. 5%..
RPRj, respectively; RP for
Class II or III PM2.5 or PM10-
2.5, maximum.
Precision of PM2.5 or PM10-2.5 .................. .................. 10%\2\............ 15%\2\............ 15%\2\............ 15%.\2\
candidate method, CP, each site.
Slope of regression relationship 1 0.10 1 0.05 1 0.10 1 0.10 1 0.10 1 0.12.
Intercept of regression 0 5... 0 1... Between: 13.55-- Between: 15.05-- Between: 62.05-- Between: 70.50--
relationship, [micro]g/m\3\. (15.05 x slope), (17.32 x slope), (70.5 x slope), (82.93 x slope),
but not less but not less but not less but not less
than--1.5; and than--2.0; and than--3.5; and than--7.0; and
16.56--(15.05 x 15.05--(13.20 x 78.95--(70.5 x 70.50--(61.16 x
slope), but not slope), but not slope), but not slope), but not
more than +1.5. more than +2.0. more than +3.5. more than +7.0.
-------------------------------------------------------------------------------
Correlation of reference method >= 0.97........... >= 0.97........... >= 0.93--for CCV <= 0.4;
and candidate method >= 0.85 + 0.2 x CCV--for 0.4 <= CCV <= 0.5;
measurements. >= 0.95--for CCV >= 0.5
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Some missing daily measurement values may be permitted; see test procedure.
\2\ Calculated as the root mean square over all measurement sets.
[[Page 16387]]
Subpart D--Procedures for Testing Performance Characteristics of
Methods for PM10
0
16. Amend Sec. 53.43 by revising the formula in paragraph (a)(2)(xvi)
and the formula in paragraph (c)(2)(iv) to read as follows:
Sec. 53.43 Test procedures.
(a) * * *
(2) * * *
(xvi) * * *
[GRAPHIC] [TIFF OMITTED] TR06MR24.037
* * * * *
(c) * * *
(2) * * *
(iv) * * *
[GRAPHIC] [TIFF OMITTED] TR06MR24.038
if Cj is below 80 [micro]g/m\3\, or
[GRAPHIC] [TIFF OMITTED] TR06MR24.039
if Cj is above 80 [micro]g/m\3\.
Subpart E--Procedures for Testing Physical (Design) and Performance
Characteristics of Reference Methods and Class I and Class II
Equivalent Methods for PM2.5 or PM10-2.5
0
17. Amend Sec. 53.51 by revising paragraph (d)(2) to read as follows:
Sec. 53.51 Demonstration of compliance with design specifications and
manufacturing and test requirements.
* * * * *
(d) * * *
(2) VSCC and TE-PM2.5C separators. For samplers and
monitors utilizing the BGI VSCC or Tisch TE-PM2.5C particle
size separators specified in sections 7.3.4.4 and 7.3.4.5 of appendix L
to part 50 of this chapter, respectively, the respective manufacturers
shall identify the critical dimensions and manufacturing tolerances for
the separator, devise appropriate test procedures to verify that the
critical dimensions and tolerances are maintained during the
manufacturing process, and carry out those procedures on each separator
manufactured to verify conformance of the manufactured products. The
manufacturer shall also maintain records of these tests and their test
results and submit evidence that this procedure is incorporated into
the manufacturing procedure, that the test is or will be routinely
implemented, and that an appropriate procedure is in place for the
disposition of units that fail this tolerance tests.
* * * * *
Subpart F--Procedures for Testing Performance Characteristics of
Class II Equivalent Methods for PM2.5
0
18. Amend Sec. 53.61 by revising paragraph (g) introductory text, the
first sentence of paragraph (g)(1), the first sentence of (g)(1)(i),
(g)(2)(i) and adding paragraph (g)(2)(iii) to read as follows:
Sec. 53.61 Test conditions.
* * * * *
(g) Vibrating Orifice Aerosol Generator (VOAG) and Flow-Focusing
Monodisperse Aerosol Generator (FMAG) conventions. This section
prescribes conventions regarding the use of the vibrating orifice
aerosol generator (VOAG) and the flow-focusing monodisperse aerosol
generator (FMAG) for the size-selective performance tests outlined in
Sec. Sec. 53.62, 53.63, 53.64, and 53.65.
(1) Particle aerodynamic diameter. The VOAG and FMAG produce near-
monodisperse droplets through the controlled breakup of a liquid jet. *
* *
(i) The physical diameter of a generated spherical particle can be
calculated from the operational parameters of the VOAG and FMAG as:
* * * * *
(2) * * *
(i) Solid particle tests performed in this subpart shall be
conducted using particles composed of ammonium fluorescein. For use in
the VOAG or FMAG, liquid solutions of known volumetric concentration
can be prepared by diluting fluorescein powder
(C2OH12O5, FW = 332.31, CAS 2321-07-5)
with aqueous ammonia. Guidelines for preparation of fluorescein
solutions of the desired volume concentration (Cvol) are
presented in Vanderpool and Rubow (1988) (Reference 2 in appendix A to
this subpart). For purposes of converting particle physical diameter to
aerodynamic diameter, an ammonium fluorescein particle density of 1.35
g/cm\3\ shall be used.
* * * * *
(iii) Calculation of the physical diameter of the particles
produced by the VOAG and FMAG requires
[[Page 16388]]
knowledge of the liquid solution's volume concentration
(Cvol). Because uranine is essentially insoluble in oleic
acid, the total particle volume is the sum of the oleic acid volume and
the uranine volume. The volume concentration of the liquid solution
shall be calculated as:
[GRAPHIC] [TIFF OMITTED] TR06MR24.040
Where:
Vu = uranine volume, ml;
Voleic = oleic acid volume, ml;
Vsol = total solution volume, ml;
Mu = uranine mass, g;
Pu = uranine density, g/cm\3\;
Moleic = oleic acid mass, g; and
Poleic = oleic acid density, g/cm\3\.
* * * * *
PART 58--AMBIENT AIR QUALITY SURVEILLANCE
0
19. The authority citation for part 58 continues to read as follows:
Authority: 42 U.S.C. 7403, 7405, 7410, 7414, 7601, 7611, 7614,
and 7619.
Subpart A--General Provisions
0
20. Amend Sec. 58.1 by:
0
a. Removing the definition for ``Approved regional method (ARM)''; and
0
b. Revising the definition for ``Traceable.''
The revision reads as follows:
Sec. 58.1 Definitions.
* * * * *
Traceable means a measurement result from a local standard whereby
the result can be related to the International System of Units (SI)
through a documented unbroken chain of calibrations, each contributing
to the measurement uncertainty. Traceable measurement results must be
compared and certified, either directly or via not more than one
intermediate standard, to a National Institute of Standards and
Technology (NIST)-certified reference standard. Examples include but
are not limited to NIST Standard Reference Material (SRM), NIST-
traceable Reference Material (NTRM), or a NIST-certified Research Gas
Mixture (RGM). Traceability to the SI through other National Metrology
Institutes (NMIs) in addition to NIST is allowed if a Declaration of
Equivalence (DoE) exists between NIST and that NMI.
* * * * *
Subpart B--Monitoring Network
0
21. Amend Sec. 58.10 by:
0
a. Revising paragraphs (a)(1) and (b)(10) and (13);
0
b. Adding paragraph (b)(14); and
0
c. Revising paragraph (d).
The revisions and addition read as follows:
Sec. 58.10 Annual monitoring network plan and periodic network
assessment.
(a)(1) Beginning July 1, 2007, the State, or where applicable
local, agency shall submit to the Regional Administrator an annual
monitoring network plan which shall provide for the documentation of
the establishment and maintenance of an air quality surveillance system
that consists of a network of SLAMS monitoring stations that can
include FRM and FEM monitors that are part of SLAMS, NCore, CSN, PAMS,
and SPM stations. The plan shall include a statement of whether the
operation of each monitor meets the requirements of appendices A, B, C,
D, and E to this part, where applicable. The Regional Administrator may
require additional information in support of this statement. The annual
monitoring network plan must be made available for public inspection
and comment for at least 30 days prior to submission to the EPA and the
submitted plan shall include and address, as appropriate, any received
comments.
* * * * *
(b) * * *
(10) Any monitors for which a waiver has been requested or granted
by the EPA Regional Administrator as allowed for under appendix D or
appendix E to this part. For those monitors where a waiver has been
approved, the annual monitoring network plan shall include the date the
waiver was approved.
* * * * *
(13) The identification of any PM2.5 FEMs used in the
monitoring agency's network where the data are not of sufficient
quality such that data are not to be compared to the national ambient
air quality standards (NAAQS). For required SLAMS where the agency
identifies that the PM2.5 Class III FEM does not produce
data of sufficient quality for comparison to the NAAQS, the monitoring
agency must ensure that an operating FRM or filter-based FEM meeting
the sample frequency requirements described in Sec. 58.12 or other
Class III PM2.5 FEM with data of sufficient quality is
operating and reporting data to meet the network design criteria
described in appendix D to this part.
(14) The identification of any site(s) intended to address being
sited in an at-risk community where there are anticipated effects from
sources in the area as required in section 4.7.1(b)(3) of appendix D to
this part. An initial approach to the question of whether any new or
moved sites are needed and to identify the communities in which they
intend to add monitoring for meeting the requirement in this paragraph
(b)(14), if applicable, shall be submitted in accordance with the
requirements of section 4.7.1(b)(3) of appendix D to this part, which
includes submission to the EPA Regional Administrator no later than
July 1, 2024. Specifics on the resulting proposed new or moved sites
for PM2.5 network design to address at-risk communities, if
applicable, would need to be detailed in annual monitoring network
plans due to each applicable EPA Regional office no later than July 1,
2025. The plan shall provide for any required sites to be operational
no later than 24 months from date of approval of a plan or January 1,
2027, whichever comes first.
* * * * *
(d) The State, or where applicable local, agency shall perform and
submit to the EPA Regional Administrator an assessment of the air
quality surveillance system every 5 years to determine, at a minimum,
if the network meets the monitoring objectives defined in appendix D to
this part, whether new sites are needed, whether existing sites are no
longer needed and can be terminated, and whether new technologies are
appropriate for incorporation into the ambient air monitoring network.
The network assessment must consider the ability of existing and
proposed sites to support air quality characterization for areas with
relatively high populations of
[[Page 16389]]
susceptible individuals (e.g., children with asthma) and other at-risk
populations, and, for any sites that are being proposed for
discontinuance, the effect on data users other than the agency itself,
such as nearby States and Tribes or health effects studies. The State,
or where applicable local, agency must submit a copy of this 5-year
assessment, along with a revised annual network plan, to the Regional
Administrator. The assessments are due every 5 years beginning July 1,
2010.
* * * * *
0
22. Amend Sec. 58.11 by revising paragraphs (a)(2) and (e) to read as
follows:
Sec. 58.11 Network technical requirements.
(a) * * *
(2) Beginning January 1, 2009, State and local governments shall
follow the quality assurance criteria contained in appendix A to this
part that apply to SPM sites when operating any SPM site which uses an
FRM or an FEM and meets the requirements of appendix E to this part,
unless the Regional Administrator approves an alternative to the
requirements of appendix A with respect to such SPM sites because
meeting those requirements would be physically and/or financially
impractical due to physical conditions at the monitoring site and the
requirements are not essential to achieving the intended data
objectives of the SPM site. Alternatives to the requirements of
appendix A may be approved for an SPM site as part of the approval of
the annual monitoring plan, or separately.
* * * * *
(e) State and local governments must assess data from Class III
PM2.5 FEM monitors operated within their network using the
performance criteria described in table C-4 to subpart C of part 53 of
this chapter, for cases where the data are identified as not of
sufficient comparability to a collocated FRM, and the monitoring agency
requests that the FEM data should not be used in comparison to the
NAAQS. These assessments are required in the monitoring agency's annual
monitoring network plan described in Sec. 58.10(b) for cases where the
FEM is identified as not of sufficient comparability to a collocated
FRM. For these collocated PM2.5 monitors, the performance
criteria apply with the following additional provisions:
(1) The acceptable concentration range (Rj), [micro]g/m\3\ may
include values down to 0 [micro]g/m\3\.
(2) The minimum number of test sites shall be at least one;
however, the number of test sites will generally include all locations
within an agency's network with collocated FRMs and FEMs.
(3) The minimum number of methods shall include at least one FRM
and at least one FEM.
(4) Since multiple FRMs and FEMs may not be present at each site,
the precision statistic requirement does not apply, even if precision
data are available.
(5) All seasons must be covered with no more than 36 consecutive
months of data in total aggregated together.
(6) The key statistical metric to include in an assessment is the
bias (both additive and multiplicative) of the PM2.5
continuous FEM(s) compared to a collocated FRM(s). Correlation is
required to be reported in the assessment, but failure to meet the
correlation criteria, by itself, is not cause to exclude data from a
continuous FEM monitor.
0
23. Amend Sec. 58.12 by revising paragraph (d)(1):
Sec. 58.12 Operating schedules.
* * * * *
(d) * * *
(1)(i) Manual PM2.5 samplers at required SLAMS stations
without a collocated continuously operating PM2.5 monitor
must operate on at least a 1-in-3 day schedule unless a waiver for an
alternative schedule has been approved per paragraph (d)(1)(ii) of this
section.
(ii) For SLAMS PM2.5 sites with both manual and
continuous PM2.5 monitors operating, the monitoring agency
may request approval for a reduction to 1-in-6 day PM2.5
sampling or for seasonal sampling from the EPA Regional Administrator.
Other requests for a reduction to 1-in-6 day PM2.5 sampling
or for seasonal sampling may be approved on a case-by-case basis. The
EPA Regional Administrator may grant sampling frequency reductions
after consideration of factors (including but not limited to the
historical PM2.5 data quality assessments, the location of
current PM2.5 design value sites, and their regulatory data
needs) if the Regional Administrator determines that the reduction in
sampling frequency will not compromise data needed for implementation
of the NAAQS. Required SLAMS stations whose measurements determine the
design value for their area and that are within plus or minus 10
percent of the annual NAAQS, and all required sites where one or more
24-hour values have exceeded the 24-hour NAAQS each year for a
consecutive period of at least 3 years are required to maintain at
least a 1-in-3 day sampling frequency until the design value no longer
meets the criteria in this paragraph (d)(1)(ii) for 3 consecutive
years. A continuously operating FEM PM2.5 monitor satisfies
the requirement in this paragraph (d)(1)(ii) unless it is identified in
the monitoring agency's annual monitoring network plan as not
appropriate for comparison to the NAAQS and the EPA Regional
Administrator has approved that the data from that monitor may be
excluded from comparison to the NAAQS.
(iii) Required SLAMS stations whose measurements determine the 24-
hour design value for their area and whose data are within plus or
minus 5 percent of the level of the 24-hour PM2.5 NAAQS must
have an FRM or FEM operate on a daily schedule if that area's design
value for the annual NAAQS is less than the level of the annual
PM2.5 standard. A continuously operating FEM or
PM2.5 monitor satisfies the requirement in this paragraph
(d)(1)(iii) unless it is identified in the monitoring agency's annual
monitoring network plan as not appropriate for comparison to the NAAQS
and the EPA Regional Administrator has approved that the data from that
monitor may be excluded from comparison to the NAAQS. The daily
schedule must be maintained until the referenced design values no
longer meets the criteria in this paragraph (d)(1)(iii) for 3
consecutive years.
(iv) Changes in sampling frequency attributable to changes in
design values shall be implemented no later than January 1 of the
calendar year following the certification of such data as described in
Sec. 58.15.
* * * * *
0
24. Revise Sec. 58.15 to read as follows:
Sec. 58.15 Annual air monitoring data certification.
(a) The State, or where appropriate local, agency shall submit to
the EPA Regional Administrator an annual air monitoring data
certification letter to certify data collected by FRM and FEM monitors
at SLAMS and SPM sites that meet criteria in appendix A to this part
from January 1 to December 31 of the previous year. The head official
in each monitoring agency, or his or her designee, shall certify that
the previous year of ambient concentration and quality assurance data
are completely submitted to AQS and that the ambient concentration data
are accurate to the best of her or his knowledge, taking into
consideration the quality assurance findings. The annual data
certification letter is due by May 1 of each year.
(b) Along with each certification letter, the State shall submit to
the Regional Administrator an annual
[[Page 16390]]
summary report of all the ambient air quality data collected by FRM and
FEM monitors at SLAMS and SPM sites. The annual report(s) shall be
submitted for data collected from January 1 to December 31 of the
previous year. The annual summary serves as the record of the specific
data that is the object of the certification letter.
(c) Along with each certification letter, the State shall submit to
the Regional Administrator a summary of the precision and accuracy data
for all ambient air quality data collected by FRM and FEM monitors at
SLAMS and SPM sites. The summary of precision and accuracy shall be
submitted for data collected from January 1 to December 31 of the
previous year.
Subpart C--Special Purpose Monitors
0
25. Amend Sec. 58.20 by revising paragraphs (b) through (e) to read as
follows:
Sec. 58.20 Special purpose monitors (SPM).
* * * * *
(b) Any SPM data collected by an air monitoring agency using a
Federal reference method (FRM) or Federal equivalent method (FEM) must
meet the requirements of Sec. Sec. 58.11 and 58.12 and appendix A to
this part or an approved alternative to appendix A. Compliance with
appendix E to this part is optional but encouraged except when the
monitoring agency's data objectives are inconsistent with the
requirements in appendix E. Data collected at an SPM using a FRM or FEM
meeting the requirements of appendix A must be submitted to AQS
according to the requirements of Sec. 58.16. Data collected by other
SPMs may be submitted. The monitoring agency must also submit to AQS an
indication of whether each SPM reporting data to AQS monitor meets the
requirements of appendices A and E.
(c) All data from an SPM using an FRM or FEM which has operated for
more than 24 months are eligible for comparison to the relevant NAAQS,
subject to the conditions of Sec. Sec. 58.11(e) and 58.30, unless the
air monitoring agency demonstrates that the data came from a particular
period during which the requirements of appendix A, appendix C, or
appendix E to this part were not met, subject to review and EPA
Regional Office approval as part of the annual monitoring network plan
described in Sec. 58.10.
(d) If an SPM using an FRM or FEM is discontinued within 24 months
of start-up, the Administrator will not base a NAAQS violation
determination for the PM2.5 or ozone NAAQS solely on data
from the SPM.
(e) If an SPM using an FRM or FEM is discontinued within 24 months
of start-up, the Administrator will not designate an area as
nonattainment for the CO, SO2, NO2, or 24-hour
PM10 NAAQS solely on the basis of data from the SPM. Such
data are eligible for use in determinations of whether a nonattainment
area has attained one of these NAAQS.
* * * * *
0
26. Amend appendix A to part 58 by:
0
a. Revising section 2.6.1 and adding sections 2.6.1.1 and 2.6.1.2;
0
b. Removing section 3.1.2.2 and redesignating sections 3.1.2.3,
3.1.2.4, 3.1.2.5, and 3.1.2.6 as sections 3.1.2.2, 3.1.2.3, 3.1.2.4,
and 3.1.2.5, respectively;
0
c. Revising sections 3.1.3.3, 3.2.4, 4.2.1, and 4.2.5; and
0
d. In section 6 revising References (1), (4), (6), (7), (9), (10), and
(11) and table A-1.
The revisions and additions read as follows:
Appendix A to Part 58--Quality Assurance Requirements for Monitors used
in Evaluations of National Ambient Air Quality Standards
* * * * *
2.6.1 Gaseous pollutant concentration standards (permeation
devices or cylinders of compressed gas) used to obtain test
concentrations for CO, SO2, NO, and NO2 must
be EPA Protocol Gases certified in accordance with one of the
procedures given in Reference 4 of this appendix.
2.6.1.1 The concentrations of EPA Protocol Gas standards used
for ambient air monitoring must be certified with a 95-percent
confidence interval to have an analytical uncertainty of no more
than 2.0 percent (inclusive) of the certified
concentration (tag value) of the gas mixture. The uncertainty must
be calculated in accordance with the statistical procedures defined
in Reference 4 of this appendix.
2.6.1.2 Specialty gas producers advertising certification with
the procedures provided in Reference 4 of this appendix and
distributing gases as ``EPA Protocol Gas'' for ambient air
monitoring purposes must adhere to the regulatory requirements
specified in 40 CFR 75.21(g) or not use ``EPA'' in any form of
advertising. Monitoring organizations must provide information to
the EPA on the specialty gas producers they use on an annual basis.
PQAOs, when requested by the EPA, must participate in the EPA
Ambient Air Protocol Gas Verification Program at least once every 5
years by sending a new unused standard to a designated verification
laboratory.
* * * * *
3.1.3.3 Using audit gases that are verified against the NIST
standard reference methods or special review procedures and
validated per the certification periods specified in Reference 4 of
this appendix (EPA Traceability Protocol for Assay and Certification
of Gaseous Calibration Standards) for CO, SO2, and
NO2 and using O3 analyzers that are verified
quarterly against a standard reference photometer.
* * * * *
3.2.4 PM2.5 Performance Evaluation Program (PEP)
Procedures. The PEP is an independent assessment used to estimate
total measurement system bias. These evaluations will be performed
under the national performance evaluation program (NPEP) as
described in section 2.4 of this appendix or a comparable program. A
prescribed number of Performance evaluation sampling events will be
performed annually within each PQAO. For PQAOs with less than or
equal to five monitoring sites, five valid performance evaluation
audits must be collected and reported each year. For PQAOs with
greater than five monitoring sites, eight valid performance
evaluation audits must be collected and reported each year. A valid
performance evaluation audit means that both the primary monitor and
PEP audit concentrations are valid and equal to or greater than 2
[micro]g/m3. Siting of the PEP monitor must be consistent with
section 3.2.3.4(c) of this appendix. However, any horizontal
distance greater than 4 meters and any vertical distance greater
than one meter must be reported to the EPA regional PEP coordinator.
Additionally for every monitor designated as a primary monitor, a
primary quality assurance organization must:
* * * * *
4.2.1 Collocated Quality Control Sampler Precision Estimate for
PM10, PM2.5, and Pb. Precision is estimated
via duplicate measurements from collocated samplers. It is
recommended that the precision be aggregated at the PQAO level
quarterly, annually, and at the 3-year level. The data pair would
only be considered valid if both concentrations are greater than or
equal to the minimum values specified in section 4(c) of this
appendix. For each collocated data pair, calculate ti,
using equation 6 to this appendix:
[[Page 16391]]
[GRAPHIC] [TIFF OMITTED] TR06MR24.041
Where Xi is the concentration from the primary
sampler and Yi is the concentration value from the audit
sampler. The coefficient of variation upper bound is calculated
using equation 7 to this appendix:
[GRAPHIC] [TIFF OMITTED] TR06MR24.042
Where k is the number of valid data pairs being aggregated, and
X\2\0.1,k-1 is the 10th percentile of a chi-squared
distribution with k-1 degrees of freedom. The factor of 2 in the
denominator adjusts for the fact that each ti is
calculated from two values with error.
* * * * *
4.2.5 Performance Evaluation Programs Bias Estimate for
PM2.5. The bias estimate is calculated using the PEP
audits described in section 3.2.4. of this appendix. The bias
estimator is based on, si, the absolute difference in
concentrations divided by the square root of the PEP concentration.
[GRAPHIC] [TIFF OMITTED] TR06MR24.043
* * * * *
6. References
(1) American National Standard Institute--Quality Management Systems
For Environmental Information And Technology Programs--Requirements
With Guidance For Use. ASQ/ANSI E4-2014. February 2014. Available
from ANSI Webstore https://webstore.ansi.org/.
* * * * *
(4) EPA Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards. EPA-600/R-12/531. May, 2012. Available from
U.S. Environmental Protection Agency, National Risk Management
Research Laboratory, Research Triangle Park NC 27711. https://www.epa.gov/nscep.
* * * * *
(6) List of Designated Reference and Equivalent Methods. Available
from U.S. Environmental Protection Agency, Center for Environmental
Measurements and Modeling, Air Methods and Characterization
Division, MD-D205-03, Research Triangle Park, NC 27711. https://www.epa.gov/amtic/air-monitoring-methods-criteria-pollutants.
(7) Transfer Standards for the Calibration of Ambient Air Monitoring
Analyzers for Ozone. EPA-454/B-13-004 U.S. Environmental Protection
Agency, Research Triangle Park, NC 27711, October, 2013. https://www.epa.gov/sites/default/files/2020-09/documents/ozonetransferstandardguidance.pdf.
* * * * *
(9) Quality Assurance Handbook for Air Pollution Measurement
Systems, Volume 1--A Field Guide to Environmental Quality Assurance.
EPA-600/R-94/038a. April 1994. Available from U.S. Environmental
Protection Agency, ORD Publications Office, Center for Environmental
Research Information (CERI), 26 W. Martin Luther King Drive,
Cincinnati, OH 45268. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#documents.
(10) Quality Assurance Handbook for Air Pollution Measurement
Systems, Volume II: Ambient Air Quality Monitoring Program Quality
System Development. EPA-454/B-13-003. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#documents.
(11) National Performance Evaluation Program Standard Operating
Procedures. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#npep.
Table A-1 to Section 6 of Appendix A--Minimum Data Assessment Requirements for NAAQS Related Criteria Pollutant Monitors
--------------------------------------------------------------------------------------------------------------------------------------------------------
Method Assessment method Coverage Minimum frequency Parameters reported AQS assessment type
--------------------------------------------------------------------------------------------------------------------------------------------------------
Gaseous Methods (CO, NO2, SO2, O3):
[[Page 16392]]
One-Point QC for SO2, NO2, O3, Response check at Each analyzer......... Once per 2 weeks \5\. Audit concentration One-Point QC.
CO. concentration 0.005- \1\ and measured
0.08 ppm SO2, NO2, concentration.\2\.
O3, and.
0.5 and 5 ppm CO......
Annual performance evaluation for See section 3.1.2 of Each analyzer......... Once per year........ Audit concentration Annual PE.
SO2, NO2, O3, CO. this appendix. \1\ and measured
concentration \2\
for each level.
NPAP for SO2, NO2, O3, CO.......... Independent Audit..... 20% of sites each year Once per year........ Audit NPAP.
concentration\1\ and
measured
concentration \2\
for each level.
Particulate Methods:
Continuous \4\ method-- Collocated samplers... 15%................... 1-in-12 days......... Primary sampler No Transaction
collocated quality control concentration and reported as raw
sampling PM2.5. duplicate sampler data.
concentration.\3\.
Manual method--collocated Collocated samplers... 15%................... 1-in-12 days......... Primary sampler No Transaction
quality control sampling PM10, concentration and reported as raw
PM2.5, Pb-TSP, Pb-PM10. duplicate sampler data.
concentration.\3\.
Flow rate verification PM10 Check of sampler flow Each sampler.......... Once every month \5\. Audit flow rate and Flow Rate
(low Vol) PM2.5, Pb-PM10. rate. measured flow rate Verification.
indicated by the
sampler.
Flow rate verification PM10 Check of sampler flow Each sampler.......... Once every quarter Audit flow rate and Flow Rate
(High-Vol), Pb-TSP. rate. \5\. measured flow rate Verification.
indicated by the
sampler.
Semi-annual flow rate audit Check of sampler flow Each sampler.......... Once every 6 months Audit flow rate and Semi Annual Flow Rate
PM10, TSP, PM10-2.5, PM2.5, Pb- rate using \5\. measured flow rate Audit.
TSP, Pb-PM10. independent standard. indicated by the
sampler.
Pb analysis audits Pb-TSP, Pb- Check of analytical Analytical............ Once each quarter \5\ Measured value and Pb Analysis Audits.
PM10. system with Pb audit audit value (ug Pb/
strips/filters. filter) using AQS
unit code 077.
Performance Evaluation Program Collocated samplers... (1) 5 valid audits for Distributed over all Primary sampler PEP.
PM2.5. primary QA orgs, with 4 quarters \5\. concentration and
<=5 sites. performance
(2) 8 valid audits for evaluation sampler
primary QA orgs, with concentration.
>5 sites.
(3) All samplers in 6
years.
Performance Evaluation Program Collocated samplers... (1) 1 valid audit and Distributed over all Primary sampler PEP.
Pb-TSP, Pb-PM10. 4 collocated samples 4 quarters \5\. concentration and
for primary QA orgs, performance
with <=5 sites. evaluation sampler
(2) 2 valid audits and concentration.
6 collocated samples Primary sampler
for primary QA orgs concentration and
with >5 sites. duplicate sampler
concentration.
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Effective concentration for open path analyzers.
\2\ Corrected concentration, if applicable for open path analyzers.
\3\ Both primary and collocated sampler values are reported as raw data.
\4\ PM2.5 is the only particulate criteria pollutant requiring collocation of continuous and manual primary monitors.
\5\ EPA's recommended maximum number of days that should exist between checks to ensure that the checks are routinely conducted over time and to limit
data impacts resulting from a failed check.
* * * * *
0
27. Amend appendix B to part 58 by:
0
a. Revising section 2.6.1 and adding sections 2.6.1.1 and 2.6.1.2;
0
b. Removing and reserving section 3.1.2.2;
[[Page 16393]]
0
c. Revising sections 3.1.3.3 and 3.2.4;
0
d. Adding sections 3.2.4.1 through 3.2.4.3;
0
e. Revising sections 4.2.1, and 4.2.5; and
0
f. In section 6 revising References (1), (4), (6), (7), (9), (10), and
(11) and table B-1.
The revisions and additions read as follows:
Appendix B to Part 58--Quality Assurance Requirements for Prevention of
Significant Deterioration (PSD) Air Monitoring
* * * * *
2.6.1 Gaseous pollutant concentration standards (permeation devices
or cylinders of compressed gas) used to obtain test concentrations for
CO, SO2, NO, and NO2 must be EPA Protocol Gases
certified in accordance with one of the procedures given in Reference 4
of this appendix.
2.6.1.1 The concentrations of EPA Protocol Gas standards used for
ambient air monitoring must be certified with a 95-percent confidence
interval to have an analytical uncertainty of no more than 2.0 percent (inclusive) of the certified concentration (tag
value) of the gas mixture. The uncertainty must be calculated in
accordance with the statistical procedures defined in Reference 4 of
this appendix.
2.6.1.2 Specialty gas producers advertising certification with the
procedures provided in Reference 4 of this appendix and distributing
gases as ``EPA Protocol Gas'' for ambient air monitoring purposes must
adhere to the regulatory requirements specified in 40 CFR 75.21(g) or
not use ``EPA'' in any form of advertising. The PSD PQAOs must provide
information to the PSD reviewing authority on the specialty gas
producers they use (or will use) for the duration of the PSD monitoring
project. This information can be provided in the QAPP or monitoring
plan but must be updated if there is a change in the specialty gas
producers used.
* * * * *
3.1.3.3 Using audit gases that are verified against the NIST
standard reference methods or special review procedures and validated
per the certification periods specified in Reference 4 of this appendix
(EPA Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards) for CO, SO2, and NO2 and
using O3 analyzers that are verified quarterly against a
standard reference photometer.
* * * * *
3.2.4 PM2.5 Performance Evaluation Program (PEP)
Procedures. The PEP is an independent assessment used to estimate total
measurement system bias. These evaluations will be performed under the
NPEP as described in section 2.4 of this appendix or a comparable
program. Performance evaluations will be performed annually within each
PQAO. For PQAOs with less than or equal to five monitoring sites, five
valid performance evaluation audits must be collected and reported each
year. For PQAOs with greater than five monitoring sites, eight valid
performance evaluation audits must be collected and reported each year.
A valid performance evaluation audit means that both the primary
monitor and PEP audit concentrations are valid and equal to or greater
than 2 [micro]g/m3. Siting of the PEP monitor must be consistent with
section 3.2.3.4(c) of this appendix. However, any horizontal distance
greater than 4 meters and any vertical distance greater than one meter
must be reported to the EPA regional PEP coordinator. Additionally for
every monitor designated as a primary monitor, a primary quality
assurance organization must:
3.2.4.1 Have each method designation evaluated each year; and,
3.2.4.2 Have all FRM and FEM samplers subject to a PEP audit at
least once every 6 years, which equates to approximately 15 percent of
the monitoring sites audited each year.
3.2.4.3 Additional information concerning the PEP is contained in
Reference 10 of this appendix. The calculations for evaluating bias
between the primary monitor and the performance evaluation monitor for
PM2.5 are described in section 4.2.5 of this appendix.
* * * * *
4.2.1 Collocated Quality Control Sampler Precision Estimate for
PM10, PM2.5, and Pb. Precision is estimated via
duplicate measurements from collocated samplers. It is recommended that
the precision be aggregated at the PQAO level quarterly, annually, and
at the 3-year level. The data pair would only be considered valid if
both concentrations are greater than or equal to the minimum values
specified in section 4(c) of this appendix. For each collocated data
pair, calculate ti, using equation 6 to this appendix:
[GRAPHIC] [TIFF OMITTED] TR06MR24.044
Where Xi is the concentration from the primary sampler
and Yi is the concentration value from the audit sampler.
The coefficient of variation upper bound is calculated using equation 7
to this appendix:
[GRAPHIC] [TIFF OMITTED] TR06MR24.045
Where k is the number of valid data pairs being aggregated, and
X\2\0.1,k-1 is the 10th percentile of a chi-squared
distribution with k-1 degrees of freedom. The factor of 2 in the
denominator adjusts for the fact that
[[Page 16394]]
each ti is calculated from two values with error.
* * * * *
4.2.5 Performance Evaluation Programs Bias Estimate for
PM2.5. The bias estimate is calculated using the PEP audits
described in section 3.2.4. of this appendix. The bias estimator is
based on, si, the absolute difference in concentrations
divided by the square root of the PEP concentration.
[GRAPHIC] [TIFF OMITTED] TR06MR24.046
* * * * *
6. References
(1) American National Standard Institute--Quality Management Systems
For Environmental Information And Technology Programs--Requirements
With Guidance For Use. ASQ/ANSI E4-2014. February 2014. Available
from ANSI Webstore https://webstore.ansi.org/.
* * * * *
(4) EPA Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards. EPA-600/R-12/531. May, 2012. Available from
U.S. Environmental Protection Agency, National Risk Management
Research Laboratory, Research Triangle Park NC 27711. https://www.epa.gov/nscep.
* * * * *
(6) List of Designated Reference and Equivalent Methods. Available
from U.S. Environmental Protection Agency, Center for Environmental
Measurements and Modeling, Air Methods and Characterization
Division, MD-D205-03, Research Triangle Park, NC 27711. https://www.epa.gov/amtic/air-monitoring-methods-criteria-pollutants.
(7) Transfer Standards for the Calibration of Ambient Air Monitoring
Analyzers for Ozone. EPA-454/B-13-004 U.S. Environmental Protection
Agency, Research Triangle Park, NC 27711, October, 2013. https://www.epa.gov/sites/default/files/2020-09/documents/ozonetransferstandardguidance.pdf.
* * * * *
(9) Quality Assurance Handbook for Air Pollution Measurement
Systems, Volume 1--A Field Guide to Environmental Quality Assurance.
EPA-600/R-94/038a. April 1994. Available from U.S. Environmental
Protection Agency, ORD Publications Office, Center for Environmental
Research Information (CERI), 26 W. Martin Luther King Drive,
Cincinnati, OH 45268. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#documents.
(10) Quality Assurance Handbook for Air Pollution Measurement
Systems, Volume II: Ambient Air Quality Monitoring Program Quality
System Development. EPA-454/B-13-003. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#documents.
(11) National Performance Evaluation Program Standard Operating
Procedures. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#npep.
Table B-1 to Section 6 of Appendix B- Minimum Data Assessment Requirements for NAAQS Related Criteria Pollutant PSD Monitors
--------------------------------------------------------------------------------------------------------------------------------------------------------
Method Assessment method Coverage Minimum frequency Parameters reported AQS Assessment type
--------------------------------------------------------------------------------------------------------------------------------------------------------
Gaseous Methods (CO, NO2, SO2, O3):
One-Point QC for SO2, NO2, O3, Response check at Each analyzer......... Once per 2 weeks\5\.. Audit One-Point QC.
CO. concentration 0.005- concentration\1\ and
0.08 ppm SO2, NO2, measured
O3, & 0.5 and 5 ppm concentration\2\.
CO.
Quarterly performance See section 3.1.2 of Each analyzer......... Once per quarter\5\.. Audit Annual PE.
evaluation for SO2, NO2, O3, this appendix. concentration\1\ and
CO. measured
concentration\2\ for
each level.
NPAP for SO2, NO2, O3, CO\3\... Independent Audit..... Each primary monitor.. Once per year........ Audit NPAP.
concentration\1\ and
measured
concentration\2\ for
each level.
Particulate Methods:
Collocated sampling PM10, Collocated samplers... 1 per PSD Network per Every 6 days or every Primary sampler No Transaction
PM2.5, Pb. pollutant. 3 days if daily concentration and reported as raw
monitoring required. duplicate sampler data.
concentration\4\.
Flow rate verification PM10, Check of sampler flow Each sampler.......... Once every month\5\.. Audit flow rate and Flow Rate
PM2.5, Pb. rate. measured flow rate Verification.
indicated by the
sampler.
Semi-annual flow rate audit Check of sampler flow Each sampler.......... Once every 6 months Audit flow rate and Semi Annual Flow Rate
PM10, PM2.5, Pb. rate using or beginning, middle measured flow rate Audit.
independent standard. and end of indicated by the
monitoring\5\. sampler.
[[Page 16395]]
Pb analysis audits Pb-TSP, Pb- Check of analytical Analytical............ Each quarter\5\...... Measured value and Pb Analysis Audits.
PM10. system with Pb audit audit value (ug Pb/
strips/filters. filter) using AQS
unit code 077 for
parameters:.
14129--Pb (TSP) LC
FRM/FEM.
85129--Pb (TSP) LC
Non-FRM/FEM..
Performance Evaluation Program Collocated samplers... (1) 5 valid audits for Over all 4 Primary sampler PEP.
PM2.5\3\. PQAOs with <= 5 quarters\5\. concentration and
sites.. performance
(2) 8 valid audits for evaluation sampler
PQAOs with > 5 sites.. concentration.
(3) All samplers in 6
years.
Performance Evaluation Program Collocated samplers... (1) 1 valid audit and Over all 4 Primary sampler PEP.
Pb \3\. 4 collocated samples quarters\5\. concentration and
for PQAOs, with <=5 performance
sites.. evaluation sampler
(2) 2 valid audits and concentration.
6 collocated samples Primary sampler
for PQAOs with >5 concentration and
sites.. duplicate sampler
concentration.
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Effective concentration for open path analyzers.
\2\ Corrected concentration, if applicable for open path analyzers.
\3\ NPAP, PM2.5, PEP, and Pb-PEP must be implemented if data is used for NAAQS decisions otherwise implementation is at PSD reviewing authority
discretion.
\4\ Both primary and collocated sampler values are reported as raw data
\5\ A maximum number of days should be between these checks to ensure the checks are routinely conducted over time and to limit data impacts resulting
from a failed check.
0
28. Amend appendix C to part 58 by:
0
a. Adding sections 2.2 and 2.2.1 through 2.2.19;
0
b. Removing and reserving sections 2.4, 2.4.1;
0
c. Removing sections 2.4.1.1 through 2.4.1.7; and
0
d. Revising section 2.7.1.
The additions and revision reads as follows:
Appendix C to Part 58--Ambient Air Quality Monitoring Methodology
* * * * *
2.2 PM10, PM2.5, or PM10-2.5
continuous FEMs with existing valid designations may be calibrated
using network data from collocated FRM and continuous FEM data under
the following provisions:
2.2.1 Data to demonstrate a calibration may include valid data
from State, local, or Tribal air agencies or data collected by
instrument manufacturers in accordance with 40 CFR 53.35 or other
data approved by the Administrator.
2.2.2 A request to update a designated methods calibration may
be initiated by the instrument manufacturer of record or the EPA
Administrator. State, local, Tribal, and multijusistincional
organizations of these entities may work with an instrument
manufacture to update a designated method calibration.
2.2.3 Requests for approval of an updated PM10,
PM2.5, or PM10-2.5 continuous FEM calibration
must meet the general submittal requirements of section 2.7 of this
appendix.
2.2.4 Data included in the request should represent a subset of
representative locations where the method is operational. For cases
with a small number of collocated FRMs and continuous FEMs sites, an
updated candidate calibration may be limited to the sites where both
methods are in use.
2.2.5 Data included in a candidate method updated calibration
may include a subset of sites where there is a large grouping of
sites in one part of the country such that the updated calibration
would be representative of the country as a whole.
2.2.6 Improvements should be national in scope and ideally
implemented through a firmware change.
2.2.7 The goal of a change to a methods calibration is to
increase the number of sites meeting measurements quality objectives
of the method as identified in section 2.3.1.1 of appendix A to this
part.
2.2.8 For meeting measurement quality objectives (MQOs), the
primary objective is to meet the bias goal as this statistic will
likely have the most influence on improving the resultant data
collected.
2.2.9 Precision data are to be included, but so long as
precision data are at least as good as existing network data or meet
the MQO referenced in section 2.2.8 of this appendix, no further
work is necessary with precision.
2.2.10 Data available to use may include routine primary and
collocated data.
2.2.11 Audit data may be useful to confirm the performance of a
candidate updated calibration but should not be used as the basis of
the calibration to keep the independence of the audit data.
2.2.12 Data utilized as the basis of the updated calibration may
be obtained by accessing EPA's AQS database or future analogous EPA
database.
2.2.13 Years of data to use in a candidate method calibration
should include two recent years where we are past the certification
period for the previous year's data, which is May 1 of each year.
2.2.14 Data from additional years is to be used to test an
updated calibration such that the calibration is independent of the
test years of interest. Data from these additional years need to
minimally demonstrate that a larger number of sites are expected to
meet bias MQO especially at sites near the level of the NAAQS for
the PM indicator of interest.
2.2.15 Outliers may be excluded using routine outlier tests.
2.2.16 The range of data used in a calibration may include all
data available or alternatively use data in the range from the
lowest measured data available up to 125% of the 24-hour NAAQS for
the PM indicator of interest.
[[Page 16396]]
2.2.17 Other improvements to a PM continuous method may be
included as part of a recommended update so long as appropriate
testing is conducted with input from EPA's Office of Research and
Development (ORD) Reference and Equivalent (R&E) Methods Designation
program.
2.2.18 EPA encourages early communication by instrument
manufacturers considering an update to a PM method. Instrument
companies should initiate such dialogue by contacting EPA's ORD R&E
Methods Designation program. The contact information for this can be
found at 40 CFR 53.4.
2.2.19 Manufacturers interested in improving instrument's
performance through an updated factory calibration must submit a
written modification request to EPA with supporting rationale.
Because the testing requirements and acceptance criteria of any
field and/or lab tests can depend upon the nature and extent of the
intended modification, applicants should contact EPA's R&E Methods
Designation program for guidance prior to development of the
modification request.
* * * * *
2.7.1 Requests for approval under sections 2.2, 2.4, 2.6.2, or
2.8 of this appendix must be submitted to: Director, Center for
Environmental Measurement and Modeling, Reference and Equivalent
Methods Designation Program (MD-D205-03), U.S. Environmental
Protection Agency, P.O. Box 12055, Research Triangle Park, North
Carolina 27711.
0
29. Amend appendix D to part 58 by revising sections 1 and 1.1(b), the
introductory text before the table in section 4.7.1(a), and sections
4.7.1(b)(3) and 4.7.2 to read as follows:
Appendix D to Part 58--Network Design Criteria for Ambient Air Quality
Monitoring
* * * * *
1. Monitoring Objectives and Spatial Scales
The purpose of this appendix is to describe monitoring
objectives and general criteria to be applied in establishing the
required SLAMS ambient air quality monitoring stations and for
choosing general locations for additional monitoring sites. This
appendix also describes specific requirements for the number and
location of FRM and FEM sites for specific pollutants, NCore
multipollutant sites, PM10 mass sites, PM2.5
mass sites, chemically-speciated PM2.5 sites, and
O3 precursor measurements sites (PAMS). These criteria
will be used by EPA in evaluating the adequacy of the air pollutant
monitoring networks.
1.1 * * *
(b) Support compliance with ambient air quality standards and
emissions strategy development. Data from FRM and FEM monitors for
NAAQS pollutants will be used for comparing an area's air pollution
levels against the NAAQS. Data from monitors of various types can be
used in the development of attainment and maintenance plans. SLAMS,
and especially NCore station data, will be used to evaluate the
regional air quality models used in developing emission strategies,
and to track trends in air pollution abatement control measures'
impact on improving air quality. In monitoring locations near major
air pollution sources, source-oriented monitoring data can provide
insight into how well industrial sources are controlling their
pollutant emissions.
* * * * *
4.7.1 * * *
(a) State and where applicable, local, agencies must operate the
minimum number of required PM2.5 SLAMS sites listed in
table D-5 to this appendix. The NCore sites are expected to
complement the PM2.5 data collection that takes place at
non-NCore SLAMS sites, and both types of sites can be used to meet
the minimum PM2.5 network requirements. For many State
and local networks, the total number of PM2.5 sites
needed to support the basic monitoring objectives of providing air
pollution data to the general public in a timely manner, support
compliance with ambient air quality standards and emission strategy
development, and support for air pollution research studies will
include more sites than the minimum numbers required in table D-5 to
this appendix. Deviations from these PM2.5 monitoring
requirements must be approved by the EPA Regional Administrator.
* * * * *
(b) * * *
(3) For areas with additional required SLAMS, a monitoring
station is to be sited in an at-risk community with poor air
quality, particularly where there are anticipated effects from
sources in the area (e.g., a major industrial area, point source(s),
port, rail yard, airport, or other transportation facility or
corridor).
* * * * *
4.7.2 Requirement for Continuous PM2.5 Monitoring.
The State, or where appropriate, local agencies must operate
continuous PM2.5 analyzers equal to at least one-half
(round up) the minimum required sites listed in table D-5 to this
appendix. At least one required continuous analyzer in each MSA must
be collocated with one of the required FRM/FEM monitors, unless at
least one of the required FRM/FEM monitors is itself a continuous
FEM monitor in which case no collocation requirement applies. State
and local air monitoring agencies must use methodologies and quality
assurance/quality control (QA/QC) procedures approved by the EPA
Regional Administrator for these required continuous analyzers.
* * * * *
0
30. Revise appendix E to part 58 to read as follows:
Appendix E to Part 58--Probe and Monitoring Path Siting Criteria for
Ambient Air Quality Monitoring
1. Introduction
2. Monitors and Samplers with Probe Inlets
3. Open Path Analyzers
4. Waiver Provisions
5. References
1. Introduction
1.1 Applicability
(a) This appendix contains specific location criteria applicable
to ambient air quality monitoring probes, inlets, and optical paths
of SLAMS, NCore, PAMS, and other monitor types whose data are
intended to be used to determine compliance with the NAAQS. These
specific location criteria are relevant after the general location
has been selected based on the monitoring objectives and spatial
scale of representation discussed in appendix D to this part.
Monitor probe material and sample residence time requirements are
also included in this appendix. Adherence to these siting criteria
is necessary to ensure the uniform collection of compatible and
comparable air quality data.
(b) The probe and monitoring path siting criteria discussed in
this appendix must be followed to the maximum extent possible. It is
recognized that there may be situations where some deviation from
the siting criteria may be necessary. In any such case, the reasons
must be thoroughly documented in a written request for a waiver that
describes whether the resulting monitoring data will be
representative of the monitoring area and how and why the proposed
or existing siting must deviate from the criteria. This
documentation should help to avoid later questions about the
validity of the resulting monitoring data. Conditions under which
the EPA would consider an application for waiver from these siting
criteria are discussed in section 4 of this appendix.
(c) The pollutant-specific probe and monitoring path siting
criteria generally apply to all spatial scales except where noted
otherwise. Specific siting criteria that are phrased with ``shall''
or ``must'' are defined as requirements and exceptions must be
granted through the waiver provisions. However, siting criteria that
are phrased with ``should'' are defined as goals to meet for
consistency but are not requirements.
2. Monitors and Samplers with Probe Inlets
2.1 Horizontal and Vertical Placement
(a) For O3 and SO2 monitoring, and for
neighborhood or larger spatial scale Pb, PM10,
PM10-2.5, PM2.5, NO2, and CO sites,
the probe must be located greater than or equal to 2.0 meters and
less than or equal to 15 meters above ground level.
(b) Middle scale CO and NO2 monitors must have
sampler inlets greater than or equal to 2.0 meters and less than or
equal to 15 meters above ground level.
(c) Middle scale PM10-2.5 sites are required to have
sampler inlets greater than or equal to 2.0 meters and less than or
equal to 7.0 meters above ground level.
(d) Microscale Pb, PM10, PM10-2.5, and
PM2.5 sites are required to have sampler inlets greater
than or equal to 2.0 meters and less than or equal to 7.0 meters
above ground level.
(e) Microscale near-road NO2 monitoring sites are
required to have sampler inlets greater than or equal to 2.0 meters
and less than or equal to 7.0 meters above ground level.
(f) The probe inlets for microscale carbon monoxide monitors
that are being used to
[[Page 16397]]
measure concentrations near roadways must be greater than or equal
to 2.0 meters and less than or equal to 7.0 meters above ground
level. Those probe inlets for microscale carbon monoxide monitors
measuring concentrations near roadways in downtown areas or urban
street canyons must be greater than or equal to 2.5 meters and less
than or equal to 3.5 meters above ground level. The probe must be at
least 1.0 meter vertically or horizontally away from any supporting
structure, walls, parapets, penthouses, etc., and away from dusty or
dirty areas. If the probe is located near the side of a building or
wall, then it should be located on the windward side of the building
relative to the prevailing wind direction during the season of
highest concentration potential for the pollutant being measured.
2.2 Spacing From Minor Sources
(a) It is important to understand the monitoring objective for a
particular site in order to interpret this requirement. Local minor
sources of a primary pollutant, such as SO2, lead, or
particles, can cause high concentrations of that particular
pollutant at a monitoring site. If the objective for that monitoring
site is to investigate these local primary pollutant emissions, then
the site will likely be properly located nearby. This type of
monitoring site would, in all likelihood, be a microscale-type of
monitoring site. If a monitoring site is to be used to determine air
quality over a much larger area, such as a neighborhood or city, a
monitoring agency should avoid placing a monitor probe inlet near
local, minor sources, because a plume from a local minor source
should not be allowed to inappropriately impact the air quality data
collected at a site. Particulate matter sites should not be located
in an unpaved area unless there is vegetative ground cover year-
round, so that the impact of windblown dusts will be kept to a
minimum.
(b) Similarly, local sources of nitric oxide (NO) and ozone-
reactive hydrocarbons can have a scavenging effect causing
unrepresentatively low concentrations of O3 in the
vicinity of probes for O3. To minimize these potential
interferences from nearby minor sources, the probe inlet should be
placed at a distance from furnace or incineration flues or other
minor sources of SO2 or NO. The separation distance
should take into account the heights of the flues, type of waste or
fuel burned, and the sulfur content of the fuel.
2.3 Spacing From Obstructions
(a) Obstacles may scavenge SO2, O3, or
NO2, and can act to restrict airflow for any pollutant.
To avoid this interference, the probe inlet must have unrestricted
airflow pursuant to paragraph (b) of this section and should be
located at a distance from obstacles. The horizontal distance from
the obstacle to the probe inlet must be at least twice the height
that the obstacle protrudes above the probe inlet. An obstacle that
does not meet the minimum distance requirement is considered an
obstruction that restricts airflow to the probe inlet. The EPA does
not generally consider objects or obstacles such as flag poles or
site towers used for NOy convertors and meteorological sensors, etc.
to be deemed obstructions.
(b) A probe inlet located near or along a vertical wall is
undesirable because air moving along the wall may be subject to
removal mechanisms. A probe inlet must have unrestricted airflow
with no obstructions (as defined in paragraph (a) of this section)
in a continuous arc of at least 270 degrees. An unobstructed
continuous arc of 180 degrees is allowable when the applicable
network design criteria specified in appendix D of this part require
monitoring in street canyons and the probe is located on the side of
a building. This arc must include the predominant wind direction for
the season of greatest pollutant concentration potential. For
particle sampling, there must be a minimum of 2.0 meters of
horizontal separation from walls, parapets, and structures for
rooftop site placement.
(c) A sampling station with a probe inlet located closer to an
obstacle than required by the criteria in this section should be
classified as middle scale or microscale, rather than neighborhood
or urban scale, since the measurements from such a station would
more closely represent these smaller scales.
(d) For near-road monitoring stations, the monitor probe shall
have an unobstructed air flow, where no obstacles exist at or above
the height of the monitor probe, between the monitor probe and the
outside nearest edge of the traffic lanes of the target road
segment.
2.4 Spacing From Trees
(a) Trees can provide surfaces for SO2,
O3, or NO2 adsorption or reactions and
surfaces for particle deposition. Trees can also act as obstructions
in locations where the trees are between the air pollutant sources
or source areas and the monitoring site and where the trees are of a
sufficient height and leaf canopy density to interfere with the
normal airflow around the probe inlet. To reduce this possible
interference/obstruction, the probe inlet should be 20 meters or
more from the drip line of trees and must be at least 10 meters from
the drip line of trees. If a tree or group of trees is an obstacle,
the probe inlet must meet the distance requirements of section 2.3
of this appendix.
(b) The scavenging effect of trees is greater for O3
than for other criteria pollutants. Monitoring agencies must take
steps to consider the impact of trees on ozone monitoring sites and
take steps to avoid this problem.
(c) Beginning January 1, 2024, microscale sites of any air
pollutant shall have no trees or shrubs located at or above the
line-of-sight fetch between the probe and the source under
investigation, e.g., a roadway or a stationary source.
2.5 Spacing From Roadways
Table E-1 to Section 2.5 of Appendix E--Minimum Separation Distance
Between Roadways and Probes for Monitoring Neighborhood and Urban Scale
Ozone (O3) and Oxides of Nitrogen (NO, NO2, NOX, NOy)
------------------------------------------------------------------------
Minimum
Roadway average daily traffic, vehicles per Minimum distance\1
day distance\1 2 3\
3\ (meters) (meters)
------------------------------------------------------------------------
<=1,000....................................... 10 10
10,000........................................ 10 20
15,000........................................ 20 30
20,000........................................ 30 40
40,000........................................ 50 60
70,000........................................ 100 100
>=110,000..................................... 250 250
------------------------------------------------------------------------
\1\ Distance from the edge of the nearest traffic lane. The distance for
intermediate traffic counts should be interpolated from the table
values based on the actual traffic count./TNOTE>
\2\ Applicable for ozone monitors whose placement was not approved as of
December 18, 2006.
\3\ All distances listed are expressed as having 2 significant figures.
When rounding is performed to assess compliance with these siting
requirements, the distance measurements will be rounded such as to
retain at least two significant figures.
2.5.1 Spacing for Ozone Probes
In siting an O3 monitor, it is important to minimize
destructive interferences from sources of NO, since NO readily
reacts with O3. Table E-1 of this appendix provides the
required minimum separation distances between a roadway and a probe
inlet for various ranges of daily roadway traffic. A sampling site
with a monitor probe located closer to a roadway than allowed by the
Table E-1 requirements should be classified as middle scale or
microscale, rather than neighborhood or urban scale, since the
measurements from such a site would more closely represent these
smaller scales.
2.5.2 Spacing for Carbon Monoxide Probes
(a) Near-road microscale CO monitoring sites, including those
located in downtown areas, urban street canyons, and other near-road
locations such as those adjacent to highly trafficked roads, are
intended to provide a measurement of the influence of the immediate
source on the pollution exposure on the adjacent area.
(b) Microscale CO monitor probe inlets in downtown areas or
urban street canyon locations shall be located a minimum distance of
2.0 meters and a maximum distance of 10 meters from the edge of the
nearest traffic lane.
(c) Microscale CO monitor probe inlets in downtown areas or
urban street canyon locations shall be located at least 10 meters
from an intersection, preferably at a midblock location. Midblock
locations are preferable to intersection locations because
intersections represent a much smaller portion of downtown space
than do the streets between
[[Page 16398]]
them. Pedestrian exposure is probably also greater in street canyon/
corridors than at intersections.
(d) Neighborhood scale CO monitor probe inlets in downtown areas
or urban street canyon locations shall be located according to the
requirements in Table E-2 of this appendix.
Table E-2 to Section 2.5.2 of Appendix E--Minimum Separation Distance
Between Roadways and Probes for Monitoring Neighborhood Scale Carbon
Monoxide
------------------------------------------------------------------------
Minimum distance 1 2
Roadway average daily traffic, vehicles per day (meters)
------------------------------------------------------------------------
<=10,000....................................... 10
15,000......................................... 25
20,000......................................... 45
30,000......................................... 80
40,000......................................... 115
50,000......................................... 135
>=60,000....................................... 150
------------------------------------------------------------------------
1 Distance from the edge of the nearest traffic lane. The distance for
intermediate traffic counts should be interpolated from the table
values based on the actual traffic count.
2 All distances listed are expressed as having 2 significant figures.
When rounding is performed to assess compliance with these siting
requirements, the distance measurements will be rounded such as to
retain at least two significant figures.
2.5.3 Spacing for Particulate Matter (PM2.5, PM2.5 10, PM10, Pb)
Inlets
(a) Since emissions associated with the operation of motor
vehicles contribute to urban area particulate matter ambient levels,
spacing from roadway criteria are necessary for ensuring national
consistency in PM sampler siting.
(b) The intent is to locate localized hot-spot sites in areas of
highest concentrations, whether it be caused by mobile or multiple
stationary sources. If the area is primarily affected by mobile
sources and the maximum concentration area(s) is judged to be a
traffic corridor or street canyon location, then the monitors should
be located near roadways with the highest traffic volume and at
separation distances most likely to produce the highest
concentrations. For microscale traffic corridor sites, the location
must be greater than or equal 5.0 meters and less than or equal to
15 meters from the major roadway. For the microscale street canyon
site, the location must be greater than or equal 2.0 meters and less
than or equal to 10 meters from the roadway. For the middle scale
site, a range of acceptable distances from the roadway is shown in
Figure E-1 of this appendix. This figure also includes separation
distances between a roadway and neighborhood or larger scale sites
by default. Any PM probe inlet at a site, 2.0 to 15 meters high, and
further back than the middle scale requirements will generally be
neighborhood, urban or regional scale. For example, according to
Figure E-1 of this appendix, if a PM sampler is primarily influenced
by roadway emissions and that sampler is set back 10 meters from a
30,000 ADT (average daily traffic) road, the site should be
classified as microscale, if the sampler's inlet height is between
2.0 and 7.0 meters. If the sampler's inlet height is between 7.0 and
15 meters, the site should be classified as middle scale. If the
sampler is 20 meters from the same road, it will be classified as
middle scale; if 40 meters, neighborhood scale; and if 110 meters,
an urban scale.
[GRAPHIC] [TIFF OMITTED] TR06MR24.047
2.5.4 Spacing for Nitrogen Dioxide (NO2) Probes
(a) In siting near-road NO2 monitors as required in
section 4.3.2 of appendix D of this part, the monitor probe shall be
as near as practicable to the outside nearest edge of the traffic
lanes of the target road segment but shall not be located at a
distance greater than 50 meters, in the horizontal, from the outside
nearest edge of the traffic lanes of the target road segment. Where
possible, the near-road NO2 monitor probe should be
within 20 meters of the target road segment.
(b) In siting NO2 monitors for neighborhood and
larger scale monitoring, it is important to minimize near-road
influences. Table E-1 of this appendix provides the required minimum
separation distances between a roadway and a probe inlet for various
ranges of daily roadway traffic. A site with a monitor probe located
closer to a roadway than allowed by the Table E-1 requirements
should be classified
[[Page 16399]]
as microscale or middle scale rather than neighborhood or urban
scale.
2.6 Probe Material and Pollutant Sampler Residence Time
(a) For the reactive gases (SO2, NO2, and
O3), approved probe materials must be used for monitors.
Studies25 34 have been conducted to determine the
suitability of materials such as polypropylene, polyethylene,
polyvinyl chloride, Tygon[supreg], aluminum, brass, stainless steel,
copper, borosilicate glass, polyvinylidene fluoride (PVDF),
polytetrafluoroethylene (PTFE), perfluoroalkoxy (PFA), and
fluorinated ethylene propylene (FEP) for use as intake sampling
lines. Of the above materials, only borosilicate glass, PVDF, PTFE,
PFA, and FEP have been found to be acceptable for use as intake
sampling lines for all the reactive gaseous pollutants. Furthermore,
the EPA \25\ has specified borosilicate glass, FEP Teflon[supreg],
or their equivalents as the only acceptable probe materials for
delivering test atmospheres in the determination of reference or
equivalent methods. Therefore, borosilicate glass, PVDF, PTFE, PFA,
FEP, or their equivalents must be the only material in the sampling
train (from probe inlet to the back of the monitor) that can be in
contact with the ambient air sample for reactive gas monitors.
Nafion\TM\, which is composed primarily of PTFE, can be considered
equivalent to PTFE; it has been shown in tests to exhibit virtually
no loss of ozone at 20-second residence times.\35\
(b) For volatile organic compound (VOC) monitoring at PAMS, FEP
Teflon[supreg] is unacceptable as the probe material because of VOC
adsorption and desorption reactions on the FEP Teflon[supreg].
Borosilicate glass, stainless steel, or their equivalents are the
acceptable probe materials for VOC and carbonyl sampling. Care must
be taken to ensure that the sample residence time is kept to 20
seconds or less.
(c) No matter how nonreactive the sampling probe material is
initially, after a period of use, reactive particulate matter is
deposited on the probe walls. Therefore, the time it takes the gas
to transfer from the probe inlet to the sampling device is critical.
Ozone in the presence of nitrogen oxide (NO) will show significant
losses, even in the most inert probe material, when the residence
time exceeds 20 seconds.\26\ Other studies 27 28indicate
that a 10-second or less residence time is easily achievable.
Therefore, sampling probes for all reactive gas monitors for
SO2, NO2, and O3 must have a sample
residence time less than 20 seconds.
2.7 Summary
Table E-3 of this appendix presents a summary of the general
requirements for probe siting criteria with respect to distances and
heights. Table E-3 requires different elevation distances above the
ground for the various pollutants. The discussion in this appendix
for each of the pollutants describes reasons for elevating the
monitor or probe inlet. The differences in the specified range of
heights are based on the vertical concentration gradients. For
source oriented and near-road monitors, the gradients in the
vertical direction are very large for the microscale, so a small
range of heights are used. The upper limit of 15 meters is specified
for the consistency between pollutants and to allow the use of a
single manifold for monitoring more than one pollutant.
Table E-3 to Section 2.7 of Appendix E--Summary of Probe Siting Criteria
----------------------------------------------------------------------------------------------------------------
Horizontal or Distance
Height vertical distance from drip
from from supporting line of Distance from
Pollutant Scale \9\ ground to structures \1\ \8\ trees to roadways to probe
probe \8\ to probe inlet probe \8\ \8\ (meters)
(meters) (meters) (meters)
----------------------------------------------------------------------------------------------------------------
SO2\2 3 4 5\................... Middle, 2.0-15 >=1.0 >=10 N/A.
Neighborhood,
Urban, and
Regional.
CO3 4 6....................... Micro [downtown or 2.5-3.5 2.0-10 for
street canyon downtown areas
sites]. or street canyon
microscale.
Micro [Near-Road 2.0-7.0 >=1.0 >=10 <=50 for near-
sites]. road microscale.
Middle and 2.0-15 See Table E-2 of
Neighborhood. this appendix
for middle and
neighborhood
scales.
O32 3 4........................ Middle, 2.0-15 >=1.0 >=10 See Table E-1.
Neighborhood,
Urban, and
Regional.
Micro............. 2.0-7.0 <=50 for near-
road micro-
scale.
NO22 3 4....................... Middle, 2.0-15 >=1.0 >=10 See Table E-1.
Neighborhood,
Urban, and
Regional.
PAMS2 3 4 Ozone precursors..... Neighborhood and 2.0-15 >=1.0 >=10 See Table E-1.
Urban.
PM, Pb 2 3 4 7................. Micro............. 2.0-7.0
Middle, 2.0-15 >=2.0 (horizontal >=10 See Figure E-1.
Neighborhood, distance only)
Urban and
Regional.
----------------------------------------------------------------------------------------------------------------
N/A--Not applicable.
\1\ When a probe is located on a rooftop, this separation distance is in reference to walls, parapets, or
penthouses located on the roof.
\2\ Should be greater than 20 meters from the dripline of tree(s) and must be 10 meters from the dripline.
\3\ Distance from sampler or probe inlet to obstacle, such as a building, must be at least twice the height the
obstacle protrudes above the sampler or probe inlet. Sites not meeting this criterion may be classified as
microscale or middle scale (see paragraphs 2.3(a) and 2.3(c)).
\4\ Must have unrestricted airflow in a continuous arc of at least 270 degrees around the probe or sampler; 180
degrees if the probe is on the side of a building or a wall for street canyon monitoring.
\5\ The probe or sampler should be away from minor sources, such as furnace or incineration flues. The
separation distance is dependent on the height of the minor source emission point(s), the type of fuel or
waste burned, and the quality of the fuel (sulfur, ash, or lead content). This criterion is designed to avoid
undue influences from minor sources.
\6\ For microscale CO monitoring sites, the probe must be >=10 meters from a street intersection and preferably
at a midblock location.
\7\ Collocated monitor inlets must be within 4.0 meters of each other and at least 2.0 meters apart for flow
rates greater than 200 liters/min or at least 1.0 meter apart for samplers having flow rates less than 200
liters/min to preclude airflow interference, unless a waiver has been granted by the Regional Administrator
pursuant to paragraph 3.3.4.2(c) of appendix A of part 58. For PM2.5, collocated monitor inlet heights should
be within 1.0 meter of each other vertically.
\8\ All distances listed are expressed as having 2 significant figures. When rounding is performed to assess
compliance with these siting requirements, the distance measurements will be rounded such as to retain at
least two significant figures.
\9\ See section 1.2 of appendix D for definitions of monitoring scales.
3. Open Path Analyzers
3.1 Horizontal and Vertical Placement
(a) For all O3 and SO2 monitoring sites
and for neighborhood or larger spatial scale NO2, and CO
sites, at least 80 percent of the monitoring path must be located
greater than or equal 2.0 meters and less than or equal to 15 meters
above ground level.
(b) Middle scale CO and NO2 sites must have
monitoring paths greater than or equal 2.0 meters and less than or
equal to 15 meters above ground level.
(c) Microscale near-road monitoring sites are required to have
monitoring paths greater than or equal 2.0 meters and less than or
equal to 7.0 meters above ground level.
(d) For microscale carbon monoxide monitors that are being used
to measure concentrations near roadways, the monitoring path must be
greater than or equal 2.0 meters and less than or equal to 7.0
meters above ground level. If the microscale carbon monoxide
monitors measuring concentrations near roadways are in downtown
areas or urban street canyons, the monitoring path must be greater
than or equal 2.5 meters and less than or equal to 3.5 meters above
ground level and at least 90
[[Page 16400]]
percent of the monitoring path must be at least 1.0 meter vertically
or horizontally away from any supporting structure, walls, parapets,
penthouses, etc., and away from dusty or dirty areas. If a
significant portion of the monitoring path is located near the side
of a building or wall, then it should be located on the windward
side of the building relative to the prevailing wind direction
during the season of highest concentration potential for the
pollutant being measured.
3.2 Spacing From Minor Sources
(a) It is important to understand the monitoring objective for a
particular site in order to interpret this requirement. Local minor
sources of a primary pollutant, such as SO2 can cause
high concentrations of that particular pollutant at a monitoring
site. If the objective for that monitoring site is to investigate
these local primary pollutant emissions, then the site will likely
be properly located nearby. This type of monitoring site would, in
all likelihood, be a microscale type of monitoring site. If a
monitoring site is to be used to determine air quality over a much
larger area, such as a neighborhood or city, a monitoring agency
should avoid placing a monitoring path near local, minor sources,
because a plume from a local minor source should not be allowed to
inappropriately impact the air quality data collected at a site.
(b) Similarly, local sources of nitric oxide (NO) and ozone-
reactive hydrocarbons can have a scavenging effect causing
unrepresentatively low concentrations of O3 in the
vicinity of monitoring paths for O3. To minimize these
potential interferences from nearby minor sources, at least 90
percent of the monitoring path should be at a distance from furnace
or incineration flues or other minor sources of SO2 or
NO. The separation distance should take into account the heights of
the flues, type of waste or fuel burned, and the sulfur content of
the fuel.
3.3 Spacing From Obstructions
(a) Obstacles may scavenge SO2, O3, or
NO2, and can act to restrict airflow for any pollutant.
To avoid this interference, at least 90 percent of the monitoring
path must have unrestricted airflow and should be located at a
distance from obstacles. The horizontal distance from the obstacle
to the monitoring path must be at least twice the height that the
obstacle protrudes above the monitoring path. An obstacle that does
not meet the minimum distance requirement is considered an
obstruction that restricts airflow to the monitoring path. The EPA
does not generally consider objects or obstacles such as flag poles
or site towers used for NOy convertors and meteorological sensors,
etc. to be deemed obstructions.
(b) A monitoring path located near or along a vertical wall is
undesirable because air moving along the wall may be subject to
removal mechanisms. At least 90 percent of the monitoring path for
open path analyzers must have unrestricted airflow with no
obstructions (as defined in paragraph (a) of this section) in a
continuous arc of at least 270 degrees. An unobstructed continuous
arc of 180 degrees is allowable when the applicable network design
criteria specified in appendix D of this part require monitoring in
street canyons and the monitoring path is located on the side of a
building. This arc must include the predominant wind direction for
the season of greatest pollutant concentration potential.
(c) Special consideration must be given to the use of open path
analyzers given their inherent potential sensitivity to certain
types of interferences and optical obstructions. A monitoring path
must be clear of all trees, brush, buildings, plumes, dust, or other
optical obstructions, including potential obstructions that may move
due to wind, human activity, growth of vegetation, etc. Temporary
optical obstructions, such as rain, particles, fog, or snow, should
be considered when siting an open path analyzer. Any of these
temporary obstructions that are of sufficient density to obscure the
light beam will negatively affect the ability of the open path
analyzer to continuously measure pollutant concentrations.
Transient, but significant obscuration of especially longer
measurement paths, could occur as a result of certain meteorological
conditions (e.g., heavy fog, rain, snow) and/or aerosol levels that
are of a sufficient density to prevent the open path analyzer's
light transmission. If certain compensating measures are not
otherwise implemented at the onset of monitoring (e.g., shorter path
lengths, higher light source intensity), data recovery during
periods of greatest primary pollutant potential could be
compromised. For instance, if heavy fog or high particulate levels
are coincident with periods of projected NAAQS-threatening pollutant
potential, the representativeness of the resulting data record in
reflecting maximum pollution concentrations may be substantially
impaired despite the fact that the site may otherwise exhibit an
acceptable, even exceedingly high, overall valid data capture rate.
(d) A sampling station with a monitoring path located closer to
an obstacle than required by the criteria in this section should be
classified as middle scale or microscale, rather than neighborhood
or urban scale, since the measurements from such a station would
more closely represent these smaller scales.
(e) For near-road monitoring stations, the monitoring path shall
have an unobstructed air flow, where no obstacles exist at or above
the height of the monitoring path, between the monitoring path and
the outside nearest edge of the traffic lanes of the target road
segment.
3.4 Spacing From Trees
(a) Trees can provide surfaces for SO2,
O3, or NO2 adsorption or reactions. Trees can
also act as obstructions in locations where the trees are located
between the air pollutant sources or source areas and the monitoring
site, and where the trees are of a sufficient height and leaf canopy
density to interfere with the normal airflow around the monitoring
path. To reduce this possible interference/obstruction, at least 90
percent of the monitoring path should be 20 meters or more from the
drip line of trees and must be at least 10 meters from the drip line
of trees. If a tree or group of trees could be considered an
obstacle, the monitoring path must meet the distance requirements of
section 3.3 of this appendix.
(b) The scavenging effect of trees is greater for O3
than for other criteria pollutants. Monitoring agencies must take
steps to consider the impact of trees on ozone monitoring sites and
take steps to avoid this problem.
(c) Beginning January 1, 2024, microscale sites of any air
pollutant shall have no trees or shrubs located at or above the
line-of-sight fetch between the monitoring path and the source under
investigation, e.g., a roadway or a stationary source.
3.5 Spacing from Roadways
Table E-4 of Section 3.5 of Appendix E--Minimum Separation Distance
Between Roadways and Monitoring Paths for Monitoring Neighborhood and
Urban Scale Ozone (O3) and Oxides of Nitrogen (NO, NO2, NOX, NOy)
------------------------------------------------------------------------
Minimum Minimum
distance distance
Roadway average daily traffic, vehicles per day \1 3\ \1 2 3\
(meters) (meters)
------------------------------------------------------------------------
<=1,000........................................... 10 10
10,000............................................ 10 20
15,000............................................ 20 30
20,000............................................ 30 40
40,000............................................ 50 60
70,000............................................ 100 100
>=110,000......................................... 250 250
------------------------------------------------------------------------
\1\ Distance from the edge of the nearest traffic lane. The distance for
intermediate traffic counts should be interpolated from the table
values based on the actual traffic count.
\2\ Applicable for ozone open path monitors whose placement was not
approved as of December 18, 2006.
\3\ All distances listed are expressed as having 2 significant figures.
When rounding is performed to assess compliance with these siting
requirements, the distance measurements will be rounded such as to
retain at least two significant figures.
3.5.1 Spacing for Ozone Monitoring Paths
In siting an O3 open path analyzer, it is important
to minimize destructive interferences form sources of NO, since NO
readily reacts with O3. Table E-4 of this appendix
provides the required minimum separation distances between a roadway
and at least 90 percent of a monitoring path for various ranges of
daily roadway traffic. A monitoring site with a monitoring path
located closer to a roadway than allowed by the Table E-4
requirements should be classified as microscale or middle scale,
rather than neighborhood or urban scale, since the measurements from
such a site would more closely represent these smaller scales. The
monitoring path(s) must not cross over a roadway with an average
daily traffic count of 10,000 vehicles per day or more. For
locations where a monitoring path crosses a roadway with fewer than
10,000 vehicles per day, monitoring agencies must consider the
entire segment of the monitoring path in the area of potential
atmospheric interference from automobile emissions. Therefore, this
[[Page 16401]]
calculation must include the length of the monitoring path over the
roadway plus any segments of the monitoring path that lie in the
area between the roadway and minimum separation distance, as
determined from Table E-4 of this appendix. The sum of these
distances must not be greater than 10 percent of the total
monitoring path length.
3.5.2 Spacing for Carbon Monoxide Monitoring Paths
(a) Near-road microscale CO monitoring sites, including those
located in downtown areas, urban street canyons, and other near-road
locations such as those adjacent to highly trafficked roads, are
intended to provide a measurement of the influence of the immediate
source on the pollution exposure on the adjacent area.
(b) Microscale CO monitoring paths in downtown areas or urban
street canyon locations shall be located a minimum distance of 2.0
meters and a maximum distance of 10 meters from the edge of the
nearest traffic lane.
(c) Microscale CO monitoring paths in downtown areas or urban
street canyon locations shall be located at least 10 meters from an
intersection, preferably at a midblock location. Midblock locations
are preferable to intersection locations because intersections
represent a much smaller portion of downtown space than do the
streets between them. Pedestrian exposure is probably also greater
in street canyon/corridors than at intersections.
(d) Neighborhood scale CO monitoring paths in downtown areas or
urban street canyon locations shall be located according to the
requirements in Table E-5 of this appendix.
Table E-5 Section 3.5.2 of Appendix E--Minimum Separation Distance
Between Roadways and Monitoring Paths for Monitoring Neighborhood Scale
Carbon Monoxide
------------------------------------------------------------------------
Minimum
Roadway average daily traffic, vehicles per day distance \1 2\
(meters)
------------------------------------------------------------------------
<=10,000................................................ 10
15,000.................................................. 25
20,000.................................................. 45
30,000.................................................. 80
40,000.................................................. 115
50,000.................................................. 135
>=60,000................................................ 150
------------------------------------------------------------------------
\1\ Distance from the edge of the nearest traffic lane. The distance for
intermediate traffic counts should be interpolated from the table
values based on the actual traffic count.
\2\ All distances listed are expressed as having 2 significant figures.
When rounding is performed to assess compliance with these siting
requirements, the distance measurements will be rounded such as to
retain at least two significant figures.
3.5.3 Spacing for Nitrogen Dioxide (NO2) Monitoring Paths
(a) In siting near-road NO2 monitors as required in
section 4.3.2 of appendix D of this part, the monitoring path shall
be as near as practicable to the outside nearest edge of the traffic
lanes of the target road segment but shall not be located at a
distance greater than 50 meters, in the horizontal, from the outside
nearest edge of the traffic lanes of the target road segment.
(b) In siting NO2 open path monitors for neighborhood
and larger scale monitoring, it is important to minimize near-road
influences. Table E-5 of this appendix provides the required minimum
separation distances between a roadway and at least 90 percent of a
monitoring path for various ranges of daily roadway traffic. A site
with a monitoring path located closer to a roadway than allowed by
the Table E-4 requirements should be classified as microscale or
middle scale rather than neighborhood or urban scale. The monitoring
path(s) must not cross over a roadway with an average daily traffic
count of 10,000 vehicles per day or more. For locations where a
monitoring path crosses a roadway with fewer than 10,000 vehicles
per day, monitoring agencies must consider the entire segment of the
monitoring path in the area of potential atmospheric interference
form automobile emissions. Therefore, this calculation must include
the length of the monitoring path over the roadway plus any segments
of the monitoring path that lie in the area between the roadway and
minimum separation distance, as determined from Table E-5 of this
appendix. The sum of these distances must not be greater than 10
percent of the total monitoring path length.
3.6 Cumulative Interferences on a Monitoring Path
The cumulative length or portion of a monitoring path that is
affected by minor sources, trees, or roadways must not exceed 10
percent of the total monitoring path length.
3.7 Maximum Monitoring Path Length
The monitoring path length must not exceed 1.0 kilometer for
open path analyzers in neighborhood, urban, or regional scale. For
middle scale monitoring sites, the monitoring path length must not
exceed 300 meters. In areas subject to frequent periods of dust,
fog, rain, or snow, consideration should be given to a shortened
monitoring path length to minimize loss of monitoring data due to
these temporary optical obstructions. For certain ambient air
monitoring scenarios using open path analyzers, shorter path lengths
may be needed in order to ensure that the monitoring site meets the
objectives and spatial scales defined in appendix D to this part.
The Regional Administrator may require shorter path lengths, as
needed on an individual basis, to ensure that the SLAMS sites meet
the appendix D requirements. Likewise, the Administrator may specify
the maximum path length used at NCore monitoring sites.
3.8 Summary
Table E-6 of this appendix presents a summary of the general
requirements for monitoring path siting criteria with respect to
distances and heights. Table E-6 requires different elevation
distances above the ground for the various pollutants. The
discussion in this appendix for each of the pollutants describes
reasons for elevating the monitoring path. The differences in the
specified range of heights are based on the vertical concentration
gradients. For source oriented and near-road monitors, the gradients
in the vertical direction are very large for the microscale, so a
small range of heights are used. The upper limit of 15 meters is
specified for the consistency between pollutants and to allow the
use of a monitoring path for monitoring more than one pollutant.
Table E-6 Section 3.8 of Appendix E--Summary of Monitoring Path Siting Criteria
----------------------------------------------------------------------------------------------------------------
Horizontal or
vertical
Height from distance from Distance from Distance from
Maximum ground to 80% supporting trees to 90% roadways to
Pollutant monitoring path of monitoring structures 2 of monitoring monitoring path
length 9 10 path 1 8 to 90% of path 1 8 1 8 (meters)
(meters) monitoring (meters)
path 1 8
(meters)
----------------------------------------------------------------------------------------------------------------
SO2 3 4 5 6.................. <= 300 m for 2.0-15 >=1.0 >=10 N/A.
Middle.
<= 1.0 km for
Neighborhood,
Urban, and
Regional.
CO4 5 7...................... <= 300 m for 2.5-3.5 >=1.0 >=10 2.0-10 for
Micro [downtown downtown areas
or street or street
canyon sites]. canyon
microscale.
<= 300 m for 2.0-7.0 <=50 for near-
Micro [Near- road
Road sites]. microscale.
[[Page 16402]]
<= 300 m for 2.0-15 See Table E-5
Middle. of this
appendix for
middle and
neighborhood
scales.
<= 1.0 km for
Neighborhood.
O33 4 5...................... <= 300 m for
Middle.
<= 1.0 km for 2.0-15 >=1.0 >=10 See Table E-4.
Neighborhood,
Urban, and
Regional.
NO23 4 5..................... Between 50 m-300 2.0-7.0 <=50 for near-
m for Micro road micro-
(Near-Road). scale.
<= 300 m for >=1.0 >=10
Middle.
<= 1.0 km for 2.0-15 See Table E-4.
Neighborhood,
Urban, and
Regional.
PAMS3 4 5 Ozone precursors... <= 1.0 km for 2.0-15 >=1.0 >=10 See Table E-4.
Neighborhood
and Urban.
----------------------------------------------------------------------------------------------------------------
N/A--Not applicable.
\1\ Monitoring path for open path analyzers is applicable only to middle or neighborhood scale CO monitoring,
middle, neighborhood, urban, and regional scale NO2 monitoring, and all applicable scales for monitoring SO2,
O3, and O3 precursors.
\2\ When the monitoring path is located on a rooftop, this separation distance is in reference to walls,
parapets, or penthouses located on roof.
\3\ At least 90 percent of the monitoring path should be greater than 20 meters from the dripline of tree(s) and
must be 10-meters from the dripline.
\4\ Distance from 90 percent of monitoring path to obstacle, such as a building, must be at least twice the
height the obstacle protrudes above the monitoring path. Sites not meeting this criterion may be classified as
microscale or middle scale (see text).
\5\ Must have unrestricted airflow 270 degrees around at least 90 percent of the monitoring path; 180 degrees if
the monitoring path is adjacent to the side of a building or a wall for street canyon monitoring.
\6\ The monitoring path should be away from minor sources, such as furnace or incineration flues. The separation
distance is dependent on the height of the minor source's emission point (such as a flue), the type of fuel or
waste burned, and the quality of the fuel (sulfur, ash, or lead content). This criterion is designed to avoid
undue influences from minor sources.
\7\ For microscale CO monitoring sites, the monitoring path must be >=10. meters from a street intersection and
preferably at a midblock location.
\8\ All distances listed are expressed as having 2 significant figures. When rounding is performed to assess
compliance with these siting requirements, the distance measurements will be rounded such as to retain at
least two significant figures.
\9\ See section 1.2 of appendix D for definitions of monitoring scales.
\10\ See section 3.7 of this appendix.
4. Waiver Provisions
Most sampling probes or monitors can be located so that they
meet the requirements of this appendix. New sites, with rare
exceptions, can be located within the limits of this appendix.
However, some existing sites may not meet these requirements and may
still produce useful data for some purposes. The EPA will consider a
written request from the State, or where applicable local, agency to
waive one or more siting criteria for some monitoring sites
providing that the State or their designee can adequately
demonstrate the need (purpose) for monitoring or establishing a
monitoring site at that location.
4.1 For a proposed new site, a waiver may be granted only if
both the following criteria are met:
4.1.1 The proposed new site can be demonstrated to be as
representative of the monitoring area as it would be if the siting
criteria were being met.
4.1.2 The monitor or probe cannot reasonably be located so as to
meet the siting criteria because of physical constraints (e.g.,
inability to locate the required type of site the necessary distance
from roadways or obstructions).
4.2 For an existing site, a waiver may be granted if either the
criterion in section 4.1.1 or the criterion in 4.1.2 of this
appendix is met.
4.3 Cost benefits, historical trends, and other factors may be
used to add support to the criteria in sections 4.1.1 and 4.1.2 of
this appendix; however, by themselves, they will not be acceptable
reasons for the EPA to grant a waiver. Written requests for waivers
must be submitted to the Regional Administrator. Granted waivers
must be renewed minimally every 5 years and ideally as part of the
network assessment as defined in Sec. 58.10(d). The approval date
of the waiver must be documented in the annual monitoring network
plan to support the requirements of Sec. 58.10(a)(1) and
58.10(b)(10).
5. References
1. Bryan, R.J., R.J. Gordon, and H. Menck. Comparison of High
Volume Air Filter Samples at Varying Distances from Los Angeles
Freeway. University of Southern California, School of Medicine, Los
Angeles, CA. (Presented at 66th Annual Meeting of Air Pollution
Control Association. Chicago, IL. June 24-28, 1973. APCA 73-158.)
2. Teer, E.H. Atmospheric Lead Concentration Above an Urban
Street. Master of Science Thesis, Washington University, St. Louis,
MO. January 1971.
3. Bradway, R.M., F.A. Record, and W.E. Belanger. Monitoring and
Modeling of Resuspended Roadway Dust Near Urban Arterials. GCA
Technology Division, Bedford, MA. (Presented at 1978 Annual Meeting
of Transportation Research Board, Washington, DC. January 1978.)
4. Pace, T.G., W.P. Freas, and E.M. Afify. Quantification of
Relationship Between Monitor Height and Measured Particulate Levels
in Seven U.S. Urban Areas. U.S. Environmental Protection Agency,
Research Triangle Park, NC. (Presented at 70th Annual Meeting of Air
Pollution Control Association, Toronto, Canada. June 20-24, 1977.
APCA 77-13.4.)
5. Harrison, P.R. Considerations for Siting Air Quality Monitors
in Urban Areas. City of Chicago, Department of Environmental
Control, Chicago, IL. (Presented at 66th Annual Meeting of Air
Pollution Control Association, Chicago, IL. June 24-28, 1973. APCA
73-161.)
6. Study of Suspended Particulate Measurements at Varying
Heights Above Ground. Texas State Department of Health, Air Control
Section, Austin, TX. 1970. p.7.
7. Rodes, C.E. and G.F. Evans. Summary of LACS Integrated
Pollutant Data. In: Los Angeles Catalyst Study Symposium. U.S.
Environmental Protection Agency, Research Triangle Park, NC. EPA
Publication No. EPA-600/4-77-034. June 1977.
8. Lynn, D.A. et al. National Assessment of the Urban
Particulate Problem: Volume 1, National Assessment. GCA Technology
Division, Bedford, MA. U.S. Environmental Protection Agency,
Research Triangle Park, NC. EPA Publication No. EPA-450/3-75-024.
June 1976.
9. Pace, T.G. Impact of Vehicle-Related Particulates on TSP
Concentrations and Rationale for Siting Hi-Vols in the Vicinity of
Roadways. OAQPS, U.S. Environmental Protection Agency, Research
Triangle Park, NC. April 1978.
10. Ludwig, F.L., J.H. Kealoha, and E. Shelar. Selecting Sites
for Monitoring Total Suspended Particulates. Stanford Research
Institute, Menlo Park, CA. Prepared for U.S. Environmental
Protection Agency, Research Triangle Park, NC. EPA Publication No.
EPA-450/3-77-018. June 1977, revised December 1977.
11. Ball, R.J. and G.E. Anderson. Optimum Site Exposure Criteria
for SO2 Monitoring. The Center for the Environment and
Man,
[[Page 16403]]
Inc., Hartford, CT. Prepared for U.S. Environmental Protection
Agency, Research Triangle Park, NC. EPA Publication No. EPA-450/3-
77-013. April 1977.
12. Ludwig, F.L. and J.H.S. Kealoha. Selecting Sites for Carbon
Monoxide Monitoring. Stanford Research Institute, Menlo Park, CA.
Prepared for U.S. Environmental Protection Agency, Research Triangle
Park, NC. EPA Publication No. EPA-450/3-75-077. September 1975.
13. Ludwig, F.L. and E. Shelar. Site Selection for the
Monitoring of Photochemical Air Pollutants. Stanford Research
Institute, Menlo Park, CA. Prepared for U.S. Environmental
Protection Agency, Research Triangle Park, NC. EPA Publication No.
EPA-450/3-78-013. April 1978.
14. Lead Analysis for Kansas City and Cincinnati, PEDCo
Environmental, Inc., Cincinnati, OH. Prepared for U.S. Environmental
Protection Agency, Research Triangle Park, NC. EPA Contract No. 66-
02-2515, June 1977.
15. Barltrap, D. and C.D. Strelow. Westway Nursery Testing
Project. Report to the Greater London Council. August 1976.
16. Daines, R. H., H. Moto, and D. M. Chilko. Atmospheric Lead:
Its Relationship to Traffic Volume and Proximity to Highways.
Environ. Sci. and Technol., 4:318, 1970.
17. Johnson, D. E., et al. Epidemiologic Study of the Effects of
Automobile Traffic on Blood Lead Levels, Southwest Research
Institute, Houston, TX. Prepared for U.S. Environmental Protection
Agency, Research Triangle Park, NC. EPA-600/1-78-055, August 1978.
18. Air Quality Criteria for Lead. Office of Research and
Development, U.S. Environmental Protection Agency, Washington, DC
EPA-600/8-83-028 aF-dF, 1986, and supplements EPA-600/8-89/049F,
August 1990. (NTIS document numbers PB87-142378 and PB91-138420.)
19. Lyman, D. R. The Atmospheric Diffusion of Carbon Monoxide
and Lead from an Expressway, Ph.D. Dissertation, University of
Cincinnati, Cincinnati, OH. 1972.
20. Wechter, S.G. Preparation of Stable Pollutant Gas Standards
Using Treated Aluminum Cylinders. ASTM STP. 598:40-54, 1976.
21. Wohlers, H.C., H. Newstein and D. Daunis. Carbon Monoxide
and Sulfur Dioxide Adsorption On and Description From Glass, Plastic
and Metal Tubings. J. Air Poll. Con. Assoc. 17:753, 1976.
22. Elfers, L.A. Field Operating Guide for Automated Air
Monitoring Equipment. U.S. NTIS. p. 202, 249, 1971.
23. Hughes, E.E. Development of Standard Reference Material for
Air Quality Measurement. ISA Transactions, 14:281-291, 1975.
24. Altshuller, A.D. and A.G. Wartburg. The Interaction of Ozone
with Plastic and Metallic Materials in a Dynamic Flow System.
Intern. Jour. Air and Water Poll., 4:70-78, 1961.
25. Code of Federal Regulations. 40 CFR 53.22, July 1976.
26. Butcher, S.S. and R.E. Ruff. Effect of Inlet Residence Time
on Analysis of Atmospheric Nitrogen Oxides and Ozone, Anal. Chem.,
43:1890, 1971.
27. Slowik, A.A. and E.B. Sansone. Diffusion Losses of Sulfur
Dioxide in Sampling Manifolds. J. Air. Poll. Con. Assoc., 24:245,
1974.
28. Yamada, V.M. and R.J. Charlson. Proper Sizing of the
Sampling Inlet Line for a Continuous Air Monitoring Station.
Environ. Sci. and Technol., 3:483, 1969.
29. Koch, R.C. and H.E. Rector. Optimum Network Design and Site
Exposure Criteria for Particulate Matter, GEOMET Technologies, Inc.,
Rockville, MD. Prepared for U.S. Environmental Protection Agency,
Research Triangle Park, NC. EPA Contract No. 68-02-3584. EPA 450/4-
87-009. May 1987.
30. Burton, R.M. and J.C. Suggs. Philadelphia Roadway Study.
Environmental Monitoring Systems Laboratory, U.S. Environmental
Protection Agency, Research Triangle Park, N.C. EPA-600/4-84-070
September 1984.
31. Technical Assistance Document for Sampling and Analysis of
Ozone Precursors. Atmospheric Research and Exposure Assessment
Laboratory, U.S. Environmental Protection Agency, Research Triangle
Park, NC 27711. EPA 600/8-91-215. October 1991.
32. Quality Assurance Handbook for Air Pollution Measurement
Systems: Volume IV. Meteorological Measurements. Atmospheric
Research and Exposure Assessment Laboratory, U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711. EPA 600/4-90-
0003. August 1989.
33. On-Site Meteorological Program Guidance for Regulatory
Modeling Applications. Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, Research Triangle Park, NC
27711. EPA 450/4-87-013. June 1987F.
34. Johnson, C., A. Whitehill, R. Long, and R. Vanderpool.
Investigation of Gaseous Criteria Pollutant Transport Efficiency as
a Function of Tubing Material. U.S. Environmental Protection Agency,
Research Triangle Park, NC 27711. EPA/600/R-22/212. August 2022.
35. Hannah Halliday, Cortina Johnson, Tad Kleindienst, Russell
Long, Robert Vanderpool, and Andrew Whitehill. Recommendations for
Nationwide Approval of Nafion\TM\ Dryers Upstream of UV-Absorption
Ozone Analyzers. U.S. Environmental Protection Agency, Research
Triangle Park, NC 27711. EPA/600/R-20/390. November 2020.
0
31. Revise appendix G to part 58 to read as follows:
Appendix G to Part 58--Uniform Air Quality Index (AQI) and Daily
Reporting
1. General Information
2. Reporting Requirements
3. Data Handling
1. General Information
1.1 AQI Overview. The AQI is a tool that simplifies reporting
air quality to the public in a nationally uniform and easy to
understand manner. The AQI converts concentrations of pollutants,
for which the EPA has established a national ambient air quality
standard (NAAQS), into a uniform scale from 0-500. These pollutants
are ozone (O3), particulate matter (PM2.5,
PM10), carbon monoxide (CO), sulfur dioxide
(SO2), and nitrogen dioxide (NO2). The scale
of the index is divided into general categories that are associated
with health messages.
2. Reporting Requirements
2.1 Applicability. The AQI must be reported daily for a
metropolitan statistical area (MSA) with a population over 350,000.
When it is useful and possible, it is recommended, but not required
for an area to report a sub-daily AQI as well.
2.2 Contents of AQI Report.
2.2.1 Content of AQI Report Requirements. An AQI report must
contain the following:
a. The reporting area(s) (the MSA or subdivision of the MSA).
b. The reporting period (the day for which the AQI is reported).
c. The main pollutant (the pollutant with the highest index
value).
d. The AQI (the highest index value).
e. The category descriptor and index value associated with the
AQI and, if choosing to report in a color format, the associated
color. Use only the following descriptors and colors for the six AQI
categories:
Table 1 to Section 2 of Appendix G--AQI Categories
------------------------------------------------------------------------
For this AQI Use this descriptor And this color \1\
------------------------------------------------------------------------
0 to 50................ ``Good''............... Green.
51 to 100.............. ``Moderate''........... Yellow.
101 to 150............. ``Unhealthy for Orange.
Sensitive Groups''.
151 to 200............. ``Unhealthy''.......... Red.
201 to 300............. ``Very Unhealthy''..... Purple.
301 and above.......... ``Hazardous''.......... Maroon\1\.
------------------------------------------------------------------------
\1\Specific color definitions can be found in the most recent reporting
guidance (Technical Assistance Document for the Reporting of Daily Air
Quality), which can be found at https://www.airnow.gov/publications/air-quality-index/technical-assistance-document-for-reporting-the-daily-aqi/.
[[Page 16404]]
f. The pollutant specific sensitive groups for any reported
index value greater than 100. The sensitive groups for each
pollutant are identified as part of the periodic review of the air
quality criteria and the NAAQS. For convenience, the EPA lists the
relevant groups for each pollutant in the most recent reporting
guidance (Technical Assistance Document for the Reporting of Daily
Air Quality), which can be found at https://www.airnow.gov/publications/air-quality-index/technical-assistance-document-for-reporting-the-daily-aqi/.
2.2.2 Contents of AQI Report When Applicable. When appropriate,
the AQI report may also contain the following, but such information
is not required:
a. Appropriate health and cautionary statements.
b. The name and index value for other pollutants, particularly
those with an index value greater than 100.
c. The index values for sub-areas of your MSA.
d. Causes for unusually high AQI values.
e. Pollutant concentrations.
f. Generally, the AQI report applies to an area's MSA only.
However, if a significant air quality problem exists (AQI greater
than 100) in areas significantly impacted by the MSA but not in it
(for example, O3 concentrations are often highest
downwind and outside an urban area), the report should identify
these areas and report the AQI for these areas as well.
2.3. Communication, Timing, and Frequency of AQI Report. The
daily AQI must be reported 7 days per week and made available via
website or other means of public access. The daily AQI report
represents the air quality for the previous day. Exceptions to this
requirement are in section 2.4 of this appendix.
a. Reporting the AQI sub-daily is recommended, but not required,
to provide more timely air quality information to the public for
making health-protective decisions.
b. Submitting hourly data in real-time to the EPA's AirNow (or
future analogous) system is recommended, but not required, and
assists the EPA in providing timely air quality information to the
public for making health-protective decisions.
c. Submitting hourly data for appropriate monitors (referenced
in section 3.2 of this appendix) satisfies the daily AQI reporting
requirement because the AirNow system makes daily and sub-daily AQI
reports widely available through its website and other communication
tools.
d. Forecasting the daily AQI provides timely air quality
information to the public and is recommended but not required. Sub-
daily forecasts are also recommended, especially when air quality is
expected to vary substantially throughout the day, like during
wildfires. Long-term (multi-day) forecasts can also be made
available when useful.
2.4. Exceptions to Reporting Requirements.
a. If the index value for a particular pollutant remains below
50 for a season or year, then it may be excluded from the
calculation of the AQI in section 3 of this appendix.
b. If all index values remain below 50 for a year, then the AQI
may be reported at the discretion of the reporting agency. In
subsequent years, if pollutant levels rise to where the AQI would be
above 50, then the AQI must be reported as required in section 2 of
this appendix.
c. As previously mentioned in section 2.3 of this appendix,
submitting hourly data in real-time from appropriate monitors
(referenced in section 3.2 of this appendix) to the EPA's AirNow (or
future analogous) system satisfies the daily AQI reporting
requirement.
3. Data Handling.
3.1 Relationship of AQI and pollutant concentrations. For each
pollutant, the AQI transforms ambient concentrations to a scale from
0 to 500. As appropriate, the AQI is associated with the NAAQS for
each pollutant. In most cases, the index value of 100 is associated
with the numerical level of the short-term standard (i.e., averaging
time of 24-hours or less) for each pollutant. The index value of 50
is associated with the numerical level of the annual standard for a
pollutant, if there is one, at one-half the level of the short-term
standard for the pollutant or at the level at which it is
appropriate to begin to provide guidance on cautionary language.
Higher categories of the index are based on the potential for
increasingly serious health effects to occur following exposure and
increasing proportions of the population that are likely to be
affected. The reported AQI corresponds to the pollutant with the
highest calculated AQI. For the purposes of reporting the AQI, the
sub-indexes for PM10 and PM2.5 are to be
considered separately. The pollutant responsible for the highest
index value (the reported AQI) is called the ``main'' pollutant for
that day.
3.2 Monitors Used for AQI Reporting. Concentration data from
State/Local Air Monitoring Station (SLAMS) or parts of the SLAMS
required by 40 CFR 58.10 must be used for each pollutant except PM.
For PM, calculate and report the AQI on days for which air quality
data has been measured (e.g., from continuous PM2.5
monitors required in appendix D to this part). PM measurements may
be used from monitors that are not reference or equivalent methods
(for example, continuous PM10 or PM2.5
monitors). Detailed guidance for relating non-approved measurements
to approved methods by statistical linear regression is referenced
here:
Reference for relating non-approved PM measurements to approved
methods (Eberly, S., T. Fitz-Simons, T. Hanley, L. Weinstock., T.
Tamanini, G. Denniston, B. Lambeth, E. Michel, S. Bortnick. Data
Quality Objectives (DQOs) For Relating Federal Reference Method
(FRM) and Continuous PM2.5 Measurements to Report an Air
Quality Index (AQI). U.S. Environmental Protection Agency, Research
Triangle Park, NC. EPA-454/B-02-002, November 2002).
3.3 AQI Forecast. The AQI can be forecasted at least 24-hours in
advance using the most accurate and reasonable procedures
considering meteorology, topography, availability of data, and
forecasting expertise. The guidance document, ``Guidelines for
Developing an Air Quality (Ozone and PM2.5) Forecasting
Program,'' can be found at https://www.airnow.gov/publications/weathercasters/guidelines-developing-air-quality-forecasting-program/.
3.4 Calculation and Equations.
a. The AQI is the highest value calculated for each pollutant as
follows:
i. Identify the highest concentration among all of the monitors
within each reporting area and truncate as follows:
(A) Ozone--truncate to 3 decimal places
PM2.5--truncate to 1 decimal place
PM10--truncate to integer
CO--truncate to 1 decimal place
SO2--truncate to integer
NO2--truncate to integer
(B) [Reserved]
ii. Using table 2 to this appendix, find the two breakpoints
that contain the concentration.
iii. Using equation 1 to this appendix, calculate the index.
iv. Round the index to the nearest integer.
Table 2 to Section 3.4 of Appendix G--Breakpoints for the AQI
--------------------------------------------------------------------------------------------------------------------------------------------------------
These breakpoints Equal these AQI's
--------------------------------------------------------------------------------------------------------------------------------------------------------
PM2.5 PM10 ([micro]g/
O3 (ppm) 8-hour O3 (ppm) 1- ([micro]g/ m\3\) 24-hour CO (ppm) 8- SO2 (ppb) 1- NO2 (ppb) 1- AQI Category
hour\1\ m\3\) 24-hour hour hour hour
--------------------------------------------------------------------------------------------------------------------------------------------------------
0.000-0.054......... .............. 0.0-9.0 0-54 0.0-4.4 0-35 0-53 0-50 Good.
0.055-0.070......... .............. 9.1-35.4 55-154 4.5-9.4 36-75 54-100 51-100 Moderate.
0.071-0.085......... 0.125-0.164 35.5-55.4 155-254 9.5-12.4 76-185 101-360 101-150 Unhealthy for
Sensitive Groups.
0.086-0.105......... 0.165-0.204 55.5-125.4 255-354 12.5-15.4 \3\ 186-304 361-649 151-200 Unhealthy.
0.106-0.200......... 0.205-0.404 125.5--225.4 355-424 15.5-30.4 \3\ 305-604 650-1249 201-300 Very Unhealthy.
[[Page 16405]]
0.201-(\2\)......... 0.405+ 225.5+ 425+ 30.5+ \3\ 605+ 1250+ 301+ \4\ Hazardous.
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Areas are generally required to report the AQI based on 8-hour ozone values. However, there are a small number of areas where an AQI based on 1-hour
ozone values would be more precautionary. In these cases, in addition to calculating the 8-hour ozone index value, the 1-hour ozone index value may be
calculated, and the maximum of the two values reported.
\2\ 8-hour O3 concentrations do not define higher AQI values (>301). AQI values > 301 are calculated with 1-hour O3 concentrations.
\3\ 1-hr SO2 concentrations do not define higher AQI values (>=200). AQI values of 200 or greater are calculated with 24-hour SO2 concentration.
\4\ AQI values between breakpoints are calculated using equation 1 to this appendix. For AQI values in the hazardous category, AQI values greater than
500 should be calculated using equation 1 and the concentration specified for the AQI value of 500. The AQI value of 500 are as follows: O3 1-hour--
0.604 ppm; PM2.5 24-hour--325.4 [micro]g/m\3\; PM10 24-hour--604 [micro]g/m\3\; CO ppm--50.4 ppm; SO2 1-hour--1004 ppb; and NO2 1-hour--2049 ppb.
b. If the concentration is equal to a breakpoint, then the index
is equal to the corresponding index value in table 2 to this
appendix. However, equation 1 to this appendix can still be used.
The results will be equal. If the concentration is between two
breakpoints, then calculate the index of that pollutant with
equation 1. It should also be noted that in some areas, the AQI
based on 1-hour O3 will be more precautionary than using
8-hour values (see footnote 1 to table 2). In these cases, the 1-
hour values as well as 8-hour values may be used to calculate index
values and then use the maximum index value as the AQI for
O3.
[GRAPHIC] [TIFF OMITTED] TR06MR24.048
Where:
Ip = the index value for pollutantp.
Cp = the truncated concentration of
pollutantp.
BPHi = the breakpoint that is greater than or equal to
Cp.
BPLo = the breakpoint that is less than or equal to
Cp.
IHi = the AQI value corresponding to BPHi.
Ilo = the AQI value corresponding to BPLo.
c. If the concentration is larger than the highest breakpoint in
table 2 to this appendix then the last two breakpoints in table 2
may be used when equation 1 to this appendix is applied.
Example:
d. Using table 2 and equation 1 to this appendix, calculate the
index value for each of the pollutants measured and select the one
that produces the highest index value for the AQI. For example, if a
PM10 value of 210 [micro]g/m\3\ is observed, a 1-hour
O3 value of 0.156 ppm, and an 8-hour O3 value
of 0.130 ppm, then do this:
i. Find the breakpoints for PM10 at 210 [micro]g/m\3\
as 155 [micro]g/m\3\ and 254 [micro]g/m\3\, corresponding to index
values 101 and 150;
ii. Find the breakpoints for 1-hour O3 at 0.156 ppm
as 0.125 ppm and 0.164 ppm, corresponding to index values 101 and
150;
iii. Find the breakpoints for 8-hour O3 at 0.130 ppm
as 0.116 ppm and 0.374 ppm, corresponding to index values 201 and
300;
iv. Apply equation 21 to this appendix for 210 [micro]g/m\3\,
PM10:
[GRAPHIC] [TIFF OMITTED] TR06MR24.049
v. Apply equation 3 to this appendix for 0.156 ppm, 1-hour
O3:
[GRAPHIC] [TIFF OMITTED] TR06MR24.050
vi. Apply equation 4 to this appendix for 0.130 ppm, 8-hour
O3:
[GRAPHIC] [TIFF OMITTED] TR06MR24.051
[[Page 16406]]
vii. Find the maximum, 206. This is the AQI. A minimal AQI
report could read: ``Today, the AQI for my city is 206, which is
Very Unhealthy, due to ozone.'' It would then reference the
associated sensitive groups.
[FR Doc. 2024-02637 Filed 3-5-24; 8:45 am]
BILLING CODE 6560-50-P