[Federal Register Volume 88, Number 18 (Friday, January 27, 2023)]
[Proposed Rules]
[Pages 5558-5719]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2023-00269]



[[Page 5557]]

Vol. 88

Friday,

No. 18

January 27, 2023

Part III





Environmental Protection Agency





-----------------------------------------------------------------------





40 CFR Parts 50, 53, and 58





Reconsideration of the National Ambient Air Quality Standards for 
Particulate Matter; Proposed Rule

  Federal Register / Vol. 88 , No. 18 / Friday, January 27, 2023 / 
Proposed Rules  

[[Page 5558]]


-----------------------------------------------------------------------

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Parts 50, 53, and 58

[EPA-HQ-OAR-2015-0072; FRL-8635-01-OAR]
RIN 2060-AV52


Reconsideration of the National Ambient Air Quality Standards for 
Particulate Matter

AGENCY: Environmental Protection Agency (EPA).

ACTION: Proposed 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 
proposes to revise the primary annual PM2.5 standard by 
lowering the level. The Agency proposes to retain the current primary 
24-hour PM2.5 standard and the primary 24-hour 
PM10 standard. The Agency also proposes not to change the 
secondary 24-hour PM2.5 standard, secondary annual 
PM2.5 standard, and secondary 24-hour PM10 
standard at this time. The EPA also proposes 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: Comments must be received on or before March 28, 2023.
    Public Hearings: The EPA will hold a virtual public hearing on this 
proposed rule. This hearing will be announced in a separate Federal 
Register document that provides details, including specific dates, 
times, and contact information for these hearings.

ADDRESSES: You may submit comments, identified by Docket ID No. EPA-HQ-
OAR-2015-0072, by any of the following means:
     Federal eRulemaking Portal: https://www.regulations.gov/ 
(our preferred method). Follow the online instructions for submitting 
comments.
     Email: [email protected]. Include the Docket ID No. 
EPA-HQ-OAR-2015-0072 in the subject line of the message.
     Mail: U.S. Environmental Protection Agency, EPA Docket 
Center, Air and Radiation Docket, Mail Code 28221T, 1200 Pennsylvania 
Avenue NW, Washington, DC 20460.
     Hand Delivery or Courier (by scheduled appointment only): 
EPA Docket Center, WJC West Building, Room 3334, 1301 Constitution 
Avenue NW, Washington, DC 20004. The Docket Center's hours of 
operations are 8:30 a.m.-4:30 p.m., Monday-Friday (except Federal 
Holidays).
    Instructions: All submissions received must include the Docket ID 
No. for this document. Comments received may be posted without change 
to https://www.regulations.gov, including any personal information 
provided. For detailed instructions on sending comments and additional 
information on the rulemaking process, see the SUPPLEMENTARY 
INFORMATION section of this document.

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: 

General Information

Preparing Comments for the EPA

    Follow the online instructions for submitting comments. Once 
submitted to the Federal eRulemaking Portal, comments cannot be edited 
or withdrawn. The EPA may publish any comment received to its public 
docket. Do not submit electronically any information you consider to be 
Confidential Business Information (CBI) or other information whose 
disclosure is restricted by statute. Multimedia submissions (audio, 
video, etc.) must be accompanied by a written submission. The written 
comment is considered the official comment and should include 
discussion of all points you wish to make. The EPA will generally not 
consider comments or comment contents located outside of the primary 
submission (i.e., on the web, the cloud, or other file sharing system). 
For additional submission methods, the full EPA public comment policy, 
information about CBI or multimedia submissions, and general guidance 
on making effective comments, please visit https://www.epa.gov/dockets/commenting-epa-dockets.
    When submitting comments, remember to:
     Identify the action by docket number and other identifying 
information (subject heading, Federal Register date and page number).
     Explain why you agree or disagree, suggest alternatives, 
and substitute language for your requested changes.
     Describe any assumptions and provide any technical 
information and/or data that you used.
     Provide specific examples to illustrate your concerns and 
suggest alternatives.
     Explain your views as clearly as possible, avoiding the 
use of profanity or personal threats.
     Make sure to submit your comments by the comment period 
deadline identified.

Availability of Information Related to This Action

    All documents in the dockets pertaining to this action are listed 
on the www.regulations.gov website. This includes documents in the 
docket for the proposed decision (Docket ID No. EPA-HQ-OAR-2015-0072) 
and a separate docket, established for the Integrated Science 
Assessment (ISA) (Docket ID No. EPA-HQ-ORD-2014-0859) that has been 
adopted by reference into the docket for this proposed decision. 
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 may be viewed with prior 
arrangement with the EPA Docket Center. Additionally, a number of the 
documents that are relevant to this proposed decision are available 
through the EPA's website at https://www.epa.gov/naaqs/particulate-matter-pm-air-quality-standards. These documents include the Integrated 
Science Assessment for Particulate Matter (U.S. EPA, 2019a), available 
at https://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=347534, the 
Supplement to the 2019 Integrated Science Assessment for Particulate 
Matter (U.S. EPA, 2022a), available at https://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=354490, and the Policy Assessment for the 
Reconsideration of the National Ambient Air Quality Standards for 
Particulate Matter (U.S. EPA, 2022b), available at https://www.epa.gov/naaqs/particulate-matter-pm-standards-integrated-science-assessments-current-review.

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

[[Page 5559]]

    5. Review Completed in 2020
    6. Reconsideration of the 2020 PM NAAQS Final Action
    a. Decision To Initiate a Reconsideration
    b. Process for Reconsideration of the 2020 PM NAAQS Decision
    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 Proposed Decisions on the Primary PM2.5 
Standards
    A. General Approach
    1. Background on the Current Standards
    a. Considerations Regarding the Adequacy of the Existing 
Standards in the 2020 Review
    2. General Approach and Key Issues in This Reconsideration of 
the 2020 Final Decision
    B. Overview of the Health Effects Evidence
    1. Nature of Effects
    a. Mortality
    b. Cardiovascular Effects
    c. Respiratory Effects
    d. Cancer
    e. Nervous System Effects
    f. Other Effects
    2. Public Health Implications and At-Risk Populations
    3. PM2.5 Concentrations in Key Studies Reporting 
Health Effects
    a. PM2.5 Exposure Concentrations Evaluated in 
Experimental Studies
    b. Ambient PM2.5 Concentrations in Locations of 
Epidemiologic Studies
    4. Uncertainties in the Health Effects Evidence
    C. Summary of Exposure and Risk Estimates
    1. Key Design Aspects
    2. Key Limitations and Uncertainties
    3. Summary of Risk Estimates
    D. Proposed Conclusions on the Primary PM2.5 
Standards
    1. CASAC Advice in This Reconsideration
    2. Evidence- and Risk-Based Considerations in the Policy 
Assessment
    a. Evidence-Based Considerations
    b. Risk-Based Considerations
    3. Administrator's Proposed Conclusions on the Primary 
PM2.5 Standards
    a. Adequacy of the Current Primary PM2.5 Standards
    b. Consideration of Alternative Primary Annual PM2.5 
Standard Levels
    E. Proposed Decisions on the Primary PM2.5 Standards
III. Rationale for Proposed Decisions on the Primary PM10 
Standard
    A. General Approach
    1. Background on the Current Standard
    i. Considerations Regarding the Adequacy of the Existing 
Standard in the 2020 Review
    2. General Approach and Key Issues in This Reconsideration of 
the 2020 Final Decision
    B. Overview of Health Effects Evidence
    1. Nature of Effects
    a. Mortality
    i. Long-Term Exposures
    ii. Short-Term Exposures
    b. Cardiovascular Effects
    i. Long-Term Exposures
    ii. Short-Term Exposures
    c. Respiratory Effects--Short-Term Exposures
    d. Cancer--Long-Term Exposures
    e. Metabolic Effects--Long-Term Exposures
    f. Nervous System Effects--Long-Term Exposures
    C. Proposed Conclusions on the Primary PM10 Standard
    1. CASAC Advice in This Reconsideration
    2. Evidence-Based Considerations in the Policy Assessment
    3. Administrator's Proposed Decision on the Current Primary 
PM10 Standard
IV. Communication of Public Health
    A. Air Quality Index Overview
    B. Air Quality Index Category Breakpoints for PM2.5
    1. Air Quality Index Values of 50, 100 and 150
    2. Air Quality Index Values of 200 and 300
    3. Air Quality Index Value of 500
    C. Air Quality Index Category Breakpoints for PM10
    D. Air Quality Index Reporting
V. Rationale for Proposed Decisions on the Secondary PM Standards
    A. General Approach
    1. Background on the Current Standards
    a. Non-Visibility Effects
    i. Considerations Regarding Adequacy of the Existing Standards 
for Non-Visibility Effects in the 2020 Review
    b. Visibility Effects
    i. Considerations Regarding Adequacy of the Existing Standards 
for Visibility Effects in the 2020 Review
    2. General Approach and Key Issues in This Reconsideration of 
the 2020 Final Decision
    B. Overview of Welfare Effects Evidence
    1. Nature of Effects
    a. Visibility
    b. Climate
    c. Materials
    C. Summary of Air Quality and Quantitative Information
    1. Visibility Effects
    a. Target Level of Protection in Terms of a PM2.5 
Visibility Index
    b. Relationship Between the PM2.5 Visibility Index 
and the Current Secondary 24-Hour PM2.5 Standard
    2. Non-Visibility Effects
    D. Proposed Conclusions on the Secondary PM Standards
    1. CASAC Advice in This Reconsideration
    2. Evidence- and Quantitative Information-Based Considerations 
in the Policy Assessment
    3. Administrator's Proposed Decision on the Current Secondary PM 
Standards
VI. Interpretation of the NAAQS for PM
    A. Proposed Amendments to Appendix K: Interpretation of the 
NAAQS for Particulate Matter
    1. Updating Design Value Calculations To Be on a Site-Level 
Basis
    2. Codifying Site Combinations To Maintain a Continuous Data 
Record
    3. Clarifying Daily Validity Requirements for Continuous 
Monitors
    B. Proposed Amendments to Appendix N: Interpretation of the 
NAAQS for PM2.5
    1. Updating References to the Proposed Revision(s) of the 
Standards
    2. Codifying Site Combinations To Maintain a Continuous Data 
Record
VII. Proposed Amendments to Ambient Monitoring and Quality Assurance 
Requirements
    A. Proposed Amendment in 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)
    1. Update to Program Title and Delivery Address for FRM and FEM 
Application and Modification Requests
    2. Requests for Delivery of a Candidate FRM or FEM Instrument
    3. Amendments to Requirements for Submission of Materials in 
Sec.  53.4(b)(7) for Language and Format
    4. Amendment to Designation of Reference and Equivalent Methods
    5. Amendment to One Test Field Campaign Requirement for Class 
III PM2.5 FEMs
    6. Amendment to Use of Monodisperse Aerosol Generator
    7. Corrections to 40 CFR Part 53 (Reference and Equivalent 
Methods)
    C. Proposed 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
    a. Quality System Requirements
    b. Measurement Quality Check Requirements
    c. Calculations for Data Quality Assessments
    d. References
    2. Quality Assurance Requirements for Prevention of Significant 
Deterioration (PSD) Air Monitoring
    a. Quality System Requirements
    b. Measurement Quality Check Requirements
    c. Calculations for Data Quality Assessments
    d. References
    3. Proposed Amendments to PM Ambient Air Quality Methodology
    a. Proposal To Revoke Approved Regional Methods (ARMs)
    b. Proposal for Calibration of PM Federal Equivalent Methods 
(FEMs)
    4. Proposed Amendment to the PM2.5 Monitoring Network 
Design Criteria To Address At-Risk Communities
    5. Proposed Revisions To Probe and Monitoring Path Siting 
Criteria
    a. Providing Separate Section for Open Path Monitoring 
Requirements

[[Page 5560]]

    b. Amending Distance Precision for Spacing Offsets
    c. Clarifying Summary Table of Probe Siting Criteria
    d. Adding Flexibility for the Spacing From Minor Sources
    e. Amendments and Clarification for the Spacing From 
Obstructions and Trees
    f. Reinstating Minimum 270-Degree Arc and Clarifying 180-Degree 
Arc in Regulatory Text
    g. Clarification on Obstacles That Act as an Obstruction
    h. Amending and Clarifying the 10-Meter Tree Dripline 
Requirement
    i. Amending Spacing Requirement for Microscale Monitoring
    j. Amending Waiver Provisions
    k. Broadening of Acceptable Probe Materials
    D. Taking Comment on Incorporating Data From Next Generation 
Technologies
    1. Background on Use of FRM and FEM Monitors
    2. Next Generation Technologies: Data Considerations
    3. PM2.5 Continuous FEMs
    4. PM2.5 Satellite Products
    5. Use of Air Sensors
    6. Summary
VIII. Clean Air Act Implementation Requirements for the PM NAAQS
    A. Designation of Areas
    B. Section 110(a)(1) and (2) Infrastructure SIP Requirements
    C. Implementing Any Revised 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 Proposed 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 13563: Improving Regulation and 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
References

Executive Summary

    This document presents the Administrator's proposed 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 proposed decisions to revise the primary annual 
PM2.5 standard by lowering the level from 12.0 [micro]g/m\3\ 
to within the range of 9.0 to 10.0 [micro]g/m\3\ while taking comment 
on alternative annual standard levels down to 8.0 [micro]g/m\3\ and up 
to 11.0 [micro]g/m\3\; to retain the current primary 24-hour 
PM2.5 standard (at a level of 35 [micro]g/m\3\) while taking 
comment on revising the level as low as 25 [micro]g/m\3\; to retain the 
primary 24-hour PM10 standard, without revision; and, not to 
change the secondary PM standards at this time, while taking comment on 
revising the level of the secondary 24-hour PM2.5 standard 
as low as 25 [micro]g/m\3\. In reaching his proposed decisions, the 
Administrator has 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 Policy Assessment (PA), and advice from the 
Clean Air Scientific Advisory Committee (CASAC). The EPA solicits 
comment on the proposed decisions described here and on the array of 
issues associated with the reconsideration of these standards, 
including the judgments of public health, public welfare and science 
policy inherent in the proposed decisions, and requests commenters also 
provide the rationales upon which views articulated in submitted 
comments are based.
    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 last review 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 is reconsidering the December 
2020 decision 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. 
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, 2020a). 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, 2020a).
---------------------------------------------------------------------------

    \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 proposed 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 PA, advice from the CASAC \2\ and public 
comment received as part of this reconsideration.\3\ The health effects 
evidence 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

[[Page 5561]]

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 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 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 risk 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 PA, also inform the proposed decisions on the primary 
PM2.5 standards. The risk assessment in this consideration 
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. 
Further analyses comparing the reductions in average national 
PM2.5 concentrations and risk rates within each demographic 
population estimate that the average percent PM2.5 
concentrations and risk reductions are slightly greater in the Black 
population than in the White population when meeting a revised annual 
standard with a lower level. The analyses are summarized in this 
document and described in detail in the PA.
    In its advice to the Administrator, the CASAC concurred with the 
draft PA 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 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 primary 24-hour 
PM2.5 standard was adequate and should be retained, without 
revision.
    In considering how to revise the suite of 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. Based on the current evidence and quantitative 
information, as well as consideration of CASAC advice and public 
comment thus far in this reconsideration, the Administrator proposes to 
conclude that the current primary PM2.5 standards are not 
adequate to protect public health with an adequate margin of safety.
    The Administrator also notes that the CASAC was unanimous in its 
advice regarding the need to revise the annual standard. In considering 
the appropriate level for a revised annual standard, the Administrator 
provisionally concludes that a standard set within the range of 9.0 to 
10.0 [micro]g/m\3\ would reflect his placing the most weight on the 
strongest available evidence while appropriately weighing the 
uncertainties. In addition, the Administrator recognizes that some 
members of CASAC advised, and the PA concluded, that the available 
scientific information provides support for considering a range that 
extends up to 11.0 [micro]g/m\3\ and down to 8.0 [micro]g/m\3\.
    With regard to the primary 24-hour PM2.5 standard, the 
Administrator finds it is less clear whether the available scientific 
evidence and quantitative

[[Page 5562]]

information calls into question the adequacy of the public health 
protection afforded by the current 24-hour standard. He notes that a 
more stringent annual standard is expected to reduce both average 
(annual) concentrations and peak (daily) concentrations. Furthermore, 
he notes that the CASAC did not reach consensus on whether revisions to 
the primary 24-hour PM2.5 standard were warranted at this 
time. The majority of the CASAC recommended that the level of the 
current primary 24-hour PM2.5 should be revised to within 
the range of 25 to 30 [micro]g/m\3\, while the minority of the CASAC 
recommended retaining the current standard. The Administrator proposes 
to conclude that the 24-hour standard should be retained, particularly 
when considered in conjunction with the protection provided by the 
suite of standards and the proposed decision to revise the annual 
standard to a level of 9.0 to 10.0 [micro]g/m\3\.
    The EPA solicits comment on the Administrator's proposed 
conclusions, and on the proposed decision to revise the primary annual 
PM2.5 standard and retain the primary 24-hour 
PM2.5 standard, without revision. The Administrator is 
conscious of his obligation to set primary standards with an adequate 
margin of safety and preliminarily determines that the proposed 
decision balances the need to provide protection against uncertain 
risks with the obligation to not set standards that are more stringent 
than necessary. 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. 
Reaching decisions on what standards are appropriate necessarily 
requires judgments of the Administrator about how to consider the 
information available from the epidemiologic studies and other relevant 
evidence. In the Administrator's judgment, the proposed suite of 
primary PM2.5 standards reflects the appropriate 
consideration of the strength of the available evidence and other 
information and their associated uncertainties and the advice of the 
CASAC. The final rulemaking will reflect the Administrator's ultimate 
judgments as to the suite of primary PM2.5 standards that 
are requisite to protect the public health with an adequate margin of 
safety. Consistent with these principles, the EPA also solicits public 
comment on alternative annual standard levels down to 8.0 [micro]g/m\3\ 
and up to 11.0 [micro]g/m\3\, on an alternative 24-hour standard level 
as low as 25 [micro]g/m\3\ and on the combination of annual and 24-hour 
standards that commenters may believe is appropriate, along with the 
approaches and scientific rationales used to support such levels. For 
example, the EPA solicits comments on the uncertainties in the reported 
associations between daily or annual average PM2.5 exposures 
and mortality or morbidity in the epidemiologic studies, the 
significance of the 25th percentile of ambient concentrations reported 
in studies, the relevance and limitations of international studies, and 
other topics discussed in section II.D.3.b.
    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 proposed 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 PA, advice from the 
CASAC and public comment received as part of this reconsideration. 
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 proposed 
decision, the Administrator recognizes that, while the available health 
effects evidence has expanded, recent studies are subjected to the same 
types of uncertainties that were judged to be important in previous 
reviews. He also recognizes that the CASAC generally agreed with the 
draft PA that it was reasonable to retain the primary 24-hour 
PM10 standard given the available scientific evidence, 
including PM10 as an appropriate indicator. He proposes to 
conclude that the newly available evidence does not call into question 
the adequacy of the current primary PM10 standard, and he 
proposes to retain that standard, without revision.
    This reconsideration of the secondary PM standards focuses on 
visibility, climate, and materials effects.\4\ The Administrator's 
proposed 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 PA; advice from the 
CASAC; and public comment received as part of this reconsideration. 
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 climate and materials 
effects, while the evidence has expanded since previous reviews, 
uncertainties remain in the evidence and there are still significant 
limitations in quantifying potential adverse effects from PM on climate 
and materials for purposes of setting a standard. With regard to 
visibility effects, the results of quantitative analyses of visibility 
impairment are similar to those in previous reviews, and 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 upper end of the range 
for the target level of protection (i.e., 30 deciviews (dv)). The CASAC 
generally agreed with the draft PA that substantial uncertainties 
remain in the scientific evidence for climate and materials effects. In 
considering the available scientific evidence for climate and materials 
effects, along with CASAC advice, the Administrator proposes to 
conclude that it is appropriate to retain the existing secondary 
standards and that it is not appropriate to establish any distinct 
secondary PM standards to address non-visibility PM-related welfare 
effects. With regard to visibility effects, while the Administrator 
notes that the CASAC did not recommend revising either the target level 
of protection for the visibility index or the level of the current 
secondary 24-hour PM2.5 standard, the Administrator

[[Page 5563]]

recognizes that, should an alternative level be considered for the 
visibility index, that the CASAC recommends also considering revisions 
to the secondary 24-hour PM2.5 standard. In considering the 
available evidence and quantitative information, with its inherent 
uncertainties and limitations, the Administrator proposes not to change 
the secondary PM standards at this time, and solicits comment on this 
proposed decision. In addition, the Administrator additionally solicits 
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\.
---------------------------------------------------------------------------

    \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 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 those welfare effects.
---------------------------------------------------------------------------

    Any proposed revisions to the PM NAAQS, if finalized, would 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 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 use 
tools to advance environmental justice, in this case for overburdened 
communities in areas with high PM concentrations above the NAAQS, as 
provided in current PM NAAQS implementation guidance to meet 
requirements (80 FR 58010, 58136, August 25, 2016). The EPA is not 
proposing changes to any of the current PM NAAQS implementation 
programs in this proposed rulemaking, and therefore is not requesting 
comment on any specific proposals related to implementation or 
designations.
    On other topics, the EPA proposes to make two sets of changes to 
the PM2.5 sub-index of the AQI. First, the EPA proposes 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 consistent with the levels of the primary 
PM2.5 annual and 24-hour standards and proposes to revise 
the lower breakpoints to be consistent with any changes to the primary 
PM2.5 standards that are part of this reconsideration. In so 
doing, the EPA proposes to revise the AQI value of 50 within the range 
of 9.0 and 10.0 [micro]g/m\3\ and proposes to retain 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 proposes to revise the upper AQI breakpoints (200 and 
above) and to replace 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 also proposes to revise 
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. The EPA proposes 
to finalize these changes to the PM2.5 AQI in conjunction 
with the Agency's final decisions on the primary annual and 24-hour 
PM2.5 standards, if proposed revisions to such standards are 
promulgated. The EPA is soliciting comment on the proposed revisions to 
the AQI. In addition, the EPA also proposes to revise 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 proposes 
revisions to data calculations and ambient air monitoring requirements 
for PM to improve the usefulness of 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 proposed 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 proposed 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 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 proposed 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 13563 and 12866. The RIA estimates the costs and 
monetized human health benefits of attaining three alternative annual 
PM2.5 standard levels and one alternative 24-hour 
PM2.5 standard level. Specifically, the RIA examines the 
proposed annual and 24-hour alternative standard levels of 10/35 
[micro]g/m\3\ and 9/35 [micro]g/m\3\, as well as the following two more 
stringent alternative standard levels: (1) An alternative annual 
standard level of 8 [micro]g/m\3\ in combination with the current 24-
hour standard (i.e., 8/35 [micro]g/m\3\), and (2) an alternative 24-
hour standard level of 30 [micro]g/m\3\ in combination with the 
proposed annual standard level of 10 [micro]g/m\3\ (i.e., 10/30 
[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 (12/35 [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

[[Page 5564]]

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)(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 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.\7\
---------------------------------------------------------------------------

    \7\ Some aspects of the CASAC's advice may not be relevant to 
the EPA's process of setting primary and secondary standards that 
are requisite to protect public health and welfare. Indeed, were the 
EPA to consider costs of implementation when reviewing and revising 
the standards ``it would be grounds for vacating the NAAQS.'' 
Whitman, 531 U.S. at 471 n.4. At the same time, the CAA directs the 
CASAC to provide advice on ``any adverse public health, welfare, 
social, economic, or energy effects which may result from various 
strategies for attainment and maintenance'' of the NAAQS to the 
Administrator under section 109(d)(2)(C)(iv). In Whitman, the Court 
clarified that most of that advice would be relevant to 
implementation but not standard setting, as it ``enable[s] the 
Administrator to assist the States in carrying out their statutory 
role as primary implementers of the NAAQS.'' Id. at 470 (emphasis in 
original). However, the Court also noted that the CASAC's ``advice 
concerning certain aspects of `adverse public health . . . effects' 
from various attainment strategies is unquestionably pertinent'' to 
the NAAQS rulemaking record and relevant to the standard setting 
process. Id. at 470 n.2.
---------------------------------------------------------------------------

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 and Part D, Subparts 1, 4 and 6 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 such standards 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

[[Page 5565]]

Criteria Document (AQCD) (DHEW, 1969).\8\ 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.
---------------------------------------------------------------------------

    \8\ 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.\9\ 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 annual arithmetic mean, averaged over three years.
---------------------------------------------------------------------------

    \9\ 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;\10\ 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.
---------------------------------------------------------------------------

    \10\ 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, holding that ``the growing empirical evidence 
demonstrating a relationship between fine particle pollution and 
adverse health effects amply justifies establishment of new fine 
particle standards.'' American Trucking Associations, Inc. v. EPA, 175 
F. 3d 1027, 1055-56 (D.C. Cir. 1999). The D.C. Circuit also found 
``ample support'' for the EPA's decision 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 v. EPA, 175 F. 3d at 1054-55. 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.
    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

[[Page 5566]]

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).\11\ 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.\12\ 
The 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.
---------------------------------------------------------------------------

    \11\ 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.
    \12\ In the 2006 proposal, the EPA proposed to revise the 24-
hour PM10 standard in part by establishing a new 
PM10-2.5 indicator for thoracic coarse particles (i.e., 
particles generally between 2.5 and 10 [micro]m in diameter). The 
EPA proposed to include any ambient mix of PM10-2.5 that 
was dominated by resuspended dust from high density traffic on paved 
roads and by PM from industrial sources and construction sources. 
The EPA proposed to exclude any ambient mix of PM10-2.5 
that was dominated by rural windblown dust and soils and by PM 
generated from agricultural and mining sources. In the final 
decision, the existing PM10 standard was retained, in 
part due to an ``inability . . . to effectively and precisely 
identify which ambient mixes are included in the 
[PM10-2.5] indicator and which are not'' (71 FR 61197, 
October 17, 2006).
---------------------------------------------------------------------------

    Several parties filed petitions for review following promulgation 
of the revised PM NAAQS in 2006. These petitions addressed the 
following issues: (1) Selecting the level of the primary annual 
PM2.5 standard; (2) retaining PM10 as the 
indicator of a standard for thoracic coarse particles, retaining the 
level and form of the 24-hour PM10 standard, and revoking 
the PM10 annual standard; and (3) setting the secondary 
PM2.5 standards identical to the primary standards. 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,\13\ 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), ISA (U.S. EPA, 
2009a), REA planning documents for health and welfare (U.S. EPA, 2009a, 
U.S. EPA, 2009c), a quantitative health risk assessment (U.S. EPA, 
2009a, U.S. EPA, 2009c), a quantitative health risk assessment (U.S. 
EPA, 2010b) and an urban-focused visibility assessment (U.S. EPA, 
2010a), and 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).
---------------------------------------------------------------------------

    \13\ The history of the NAAQS review process, including 
revisions to the process, is discussed athttps://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 \14\ 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 and annual PM2.5 
standards \15\ and the 24-hour PM10 standard to address 
visibility and non-visibility welfare effects.
---------------------------------------------------------------------------

    \14\ The EPA also eliminated the option for spatial averaging.
    \15\ 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 Completed in 2020
    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/

[[Page 5567]]

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 Administrator issued a memorandum describing a 
``back-to-basics'' process for reviewing the NAAQS (Pruitt, 2018). This 
memo announced 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. Following this memo, on October 10, 2018, 
the 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).\16\
---------------------------------------------------------------------------

    \16\ 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 EPA Administrator dated April 11, 2019 
(Cox, 2019a). The EPA took steps to address 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 December 2019 (84 FR 58713, November 1, 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 
EPA 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. 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 (U.S. EPA, 2020a), 
as described in detail in section I.C.5 of the 2020 proposal document 
(85 FR 24100, April 30, 2020).
    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.
    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. In order to consider 
whether reconsideration of the 2020 final action was warranted, the EPA 
moved for two 90-day abeyances in these consolidated cases, which the 
Court granted. After the EPA announced that it is reconsidering the 
2020 final decision, the EPA filed a motion with the Court to hold the 
consolidated cases in abeyance until March 1, 2023, which the court 
granted on October 1, 2021.
6. Reconsideration of the 2020 PM NAAQS Final Action
    On January 20, 2021, President Biden issued an ``Executive Order on 
Protecting Public Health and the Environment and Restoring Science to 
Tackle the Climate Crisis'' (Executive Order 13990; 86 FR 7037, January 
25, 2021),\17\ which directed review of certain agency actions. 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.\18\
---------------------------------------------------------------------------

    \17\ 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/.
    \18\ See https://www.whitehouse.gov/briefing-room/statements-releases/2021/01/20/fact-sheet-list-of-agency-actions-for-review/.
---------------------------------------------------------------------------

a. Decision To Initiate a Reconsideration
    On June 10, 2021, the Agency announced its decision to reconsider 
the 2020 PM NAAQS final action.\19\ The EPA is reconsidering the 
December 2020 decision 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. The EPA noted that the 2020 PA concluded that the scientific 
evidence and information supported 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, 2020a). 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, 2020a).
---------------------------------------------------------------------------

    \19\ The press release for this announcement is available at: 
https://www.epa.gov/newsreleases/epa-reexamine-health-standards-harmful-soot-previous-administration-left-unchanged.
---------------------------------------------------------------------------

b. Process for Reconsideration of the 2020 PM NAAQS Decision
    In its announcement of the reconsideration of the PM NAAQS, the 
Agency explained 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

[[Page 5568]]

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/f?p=113:29:1706195567016:::RP,29:P29_COMMITTEEON:CASAC.
    \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). 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). The EPA 
took steps to address these comments 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).
    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 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, 
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, 2020a). 
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, 2020a).'' 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, 2020a). 
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, 2020a).''
    \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, 2020a).''
    \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

[[Page 5569]]

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 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). However, recognizing the limited 
scope of the draft ISA Supplement, the CASAC stated that ``[a]lthough 
this limitation is appropriate for the targeted purpose of the Draft 
ISA Supplement . . . this limiting of scope applies only to this 
document and is not intended to establish a precedent for future ISAs'' 
(Sheppard, 2022b, p. 2 of letter).
    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 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 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 information were not evaluated 
(i.e., PM10-2.5-related health effects and non-visibility 
effects), the conclusions presented in the 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 
PA throughout this document).

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 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 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 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 over 30% of 
emissions of primary PM2.5 emissions (U.S. EPA, 2021).
    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

[[Page 5570]]

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). 
Landowners, land managers and government public safety agencies are 
strongly motivated to reduce the frequency and severity of human caused 
wildfires. Additionally, land managers, landowners, air agencies and 
communities may be able to lessen the impacts of wildfires by working 
collaboratively to take steps to minimize fuel loading in areas 
vulnerable to fire. Fuel load minimization steps can consist of both 
prescribed fire and mechanical treatments, such as using mechanical 
equipment to reduce accumulated understory (81 FR 68249, October 3, 
2016). There are specific Federal plans of the Department of the 
Interior \27\ and United States Forest Service \28\ to increase fuel 
load minimization efforts in areas at high risk of wildfire. The 
recently passed Bipartisan Infrastructure Law \29\ and Inflation 
Reduction Act \30\ further direct agencies and provide funding for such 
efforts at the Federal level as well as at state, Tribal, local, and 
private landowner levels.\31\
---------------------------------------------------------------------------

    \27\ See U.S. Department of the Interior, ``Infrastructure 
Investment and Jobs Act Wildfire Risk Five-Year Monitoring, 
Maintenance, and Treatment Plan'' (April 2022), available at: 
https://www.doi.gov/sites/doi.gov/files/bil-5-year-wildfire-risk-mmt-plan.04.2022.owf_.final_.pdf.
    \28\ See U.S. Department of Agriculture, Forest Service, 
``Confronting the Wildfire Crisis: A Strategy for Protecting 
Communities and Improving Resilience in America's Forests'', FS-
1187d (April 2022) available at: https://www.fs.usda.gov/sites/default/files/Confronting-Wildfire-Crisis.pdf.
    \29\ Inflation Reduction Act, Public Law 117-169 available at 
https://www.congress.gov/117/plaws/publ169/PLAW-117publ169.pdf.
    \30\ Infrastructure Investment and Jobs Act, Public Law 117-58, 
available at https://www.congress.gov/117/plaws/publ58/PLAW-117publ58.pdf.
    \31\ Inflation Reduction Act, Public Law 117-169 available at 
https://www.congress.gov/117/plaws/publ169/PLAW-117publ169.pdf.
---------------------------------------------------------------------------

    Wildfire events produce high PM emissions that impact the PM 
concentrations in ambient air to the extent that such days with high PM 
concentrations from wildfire smoke events may affect the design values 
in a given area. The annual and daily design values affected by 
potential exceptional events associated with wildfire smoke may qualify 
to be excluded from design value calculations used for comparison to 
the NAAQS. The EPA's Exceptional Events Rule (81 FR 68216, October 3, 
2016) describes the process by which exceedances caused by fire events, 
including certain prescribed fires, can be excluded from the design 
values. It should be noted that potential exceptional events associated 
with prescribed fires on wildland may also qualify to be excluded from 
design value calculations used for comparison to the NAAQS under the 
Exceptional Events Rule (as described in more detail in section VIII 
below).
    While the EPA is not proposing changes to implementation as a part 
of this proposal (as described in more detail in section VIII below), 
the EPA acknowledges that increases in PM2.5 emissions due 
to increases in wildfire and prescribed fire on wildland present a 
number of challenges relevant to the implementation of the PM NAAQS, 
particularly if one or more standards are strengthened. Stakeholders 
have expressed concern about the growing health challenges associated 
with such emissions, the importance of prescribed fire for managing 
fire-dependent ecosystems and reducing fuel loads, and the potential 
for further increases in the frequency and magnitude of wildfires due 
to climate change. Though such issues are outside the scope of this 
proposal, the EPA acknowledges that these topics may arise in the 
context of implementation of any revised PM2.5 NAAQS and 
intends to work with stakeholders to address these issues.
3. Monitoring of Ambient PM
    To promote uniform enforcement 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 established PM 
Federal Reference Methods (FRMs) for both PM10 and 
PM2.5 (appendices J and L to 40 CFR part 50). Amended 
following the 2006 and 2012 PM NAAQS reviews, the current PM monitoring 
network relies on FRMs and automated continuous Federal Equivalent 
Methods (FEMs), in part to support changes necessary for implementation 
of the revised PM standards. The requirement for measuring ambient air 
quality and reporting ambient air quality data and related information 
are the basis for appendices A through E to 40 CFR part 58. More 
information on PM ambient monitoring networks is available in section 
2.2 of the 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. For most urban locations, PM2.5 monitors are 
sited at the neighborhood scale,\32\ 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 area-wide air quality 
is sited in an area of expected maximum concentration. 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.
---------------------------------------------------------------------------

    \32\ For PM2.5, neighborhood scale is defined 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.
---------------------------------------------------------------------------

    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 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

[[Page 5571]]

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 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 [mu]g/m\3\ higher than at the 
highest measuring non-near-road sites (range is 0.1 to 2.1 [mu]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 sub-daily 
patterns, and because some health studies examine PM exposure durations 
shorter than 24-hours, it is useful to understand the broader 
distribution of sub-daily PM2.5 concentrations across the 
U.S. The 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 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.

[[Page 5572]]

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 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 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 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). 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

[[Page 5573]]

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 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 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. Analyses in the 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 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. 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 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 Fields 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 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 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 PA evaluates the DI2019 
\40\ and HA2020 \41\ hybrid surfaces, surfaces that are used in several 
of the key epidemiologic studies in the 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 
influenced by the approach used to compute that mean and how that value 
might compare to monitor reported concentrations.
---------------------------------------------------------------------------

    \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 community multiscale air quality (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 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

[[Page 5574]]

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 [mu]g/m\3\ to 8.1 [mu]g/m\3\, compared to the CBSA scale, which 
ranges from 5.7 [mu]g/m\3\ to 8.7 [mu]g/m\3\. (U.S. EPA, 2022b, 
section 2.3.3.2.4, Table 2-6).
---------------------------------------------------------------------------

    The 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 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 [mu]g/
m\3\. When population weighting is applied, the average annual 
PM2.5 concentrations are slightly higher, ranging from 8.2 
to 10.2 [mu]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 [mu]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 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 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 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 [mu]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 
[mu]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 Proposed Decisions on the Primary PM2.5 Standards

    This section presents the rationale for the Administrator's 
proposed decision to revise the primary annual PM2.5 
standard 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 presented 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 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, 
2022b).\47\ The Administrator's rationale also takes into account: (1) 
the 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, as 
reflected in discussions of the drafts of the ISA Supplement and PA at 
public meetings and in the CASAC's letters to the Administrator; 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, 2022b).
---------------------------------------------------------------------------

    In presenting the rationale for the Administrator's proposed 
decisions and its foundations, section II.A provides background and 
introductory information for this reconsideration of the primary 
PM2.5 standards. It includes background on the 2020 final 
decision to retain the primary PM2.5 standards (section 
II.A.1) and also describes the general approach for this 
reconsideration (section II.A.2). Section II.B summarizes the key 
aspects of the currently available health effects evidence, focusing on 
consideration of

[[Page 5575]]

the key policy-relevant aspects. Section II.C summarizes the risk 
information for this reconsideration, drawing on the quantitative 
analyses for PM2.5, presented in the PA. Section II.D 
presents the Administrator's proposed conclusions on the current 
primary annual and 24-hour PM2.5 standards (section II.D.3), 
drawing on both the evidence-based and risk-based considerations 
(section II.D.2) and advice from the CASAC (section II.D.1).

A. General Approach

    This reconsideration of the 2020 final decision on the primary 
PM2.5 standards relies 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 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 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 PA. In 
evaluating the public health protection afforded by the current 
standards, 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 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 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]).
---------------------------------------------------------------------------

    The subsections below provide background and introductory 
information. Background on the 2020 decision to retain the current 
standards, including the rationale for that decision, is summarized in 
section II.A.1. This is followed, in section II.A.2, by an overview of 
the general approach for the reconsideration of the 2020 final 
decision. Following this introductory section and subsections, the 
subsequent sections summarize current information and analyses, 
including that newly available in this reconsideration. The 
Administrator's proposed conclusions on the primary PM2.5 
standards, based on the current information, are provided in section 
II.D.3.
1. Background on the Current Standards
    The current primary PM2.5 standards were retained in 
2020 based on the scientific evidence and quantitative risk analyses 
available at that time, as well as the Administrator's judgments 
regarding the available scientific information, the appropriate degree 
of public health protection for the standards, and the available risk 
information regarding the exposures and risk that may be allowed by the 
current standards (85 FR 82718, December 18, 2020). With the 2020 final 
decision, the EPA retained the primary 24-hour PM2.5 
standard, with its level of 35 [mu]g/m\3\, and the primary annual 
PM2.5 standard, with its level of 12.0 [mu]g/m\3\, this 
decision was informed by the scientific evidence evaluated in the 2019 
ISA, the evidence and quantitative risk information in the 2020 PA, the 
advice and recommendations of the CASAC, and public comments on the 
proposed decision (85 FR 24094, April 30, 2020).
    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, 2009a). 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 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. Based on this evidence, the 2019 ISA concludes 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 
exposures and respiratory effects, and between long-term 
PM2.5 exposures 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). Some of the evidence that supported 
conclusions on at-risk populations and lifestages also

[[Page 5576]]

contributed to the conclusions of causal and likely to be causal 
relationships within the 2019 ISA, including:
     PM2.5-related mortality and cardiovascular 
effects in older adults (U.S. EPA, 2019a, sections 11.1, 11.2, 6.1, and 
6.2);
     PM2.5-related cardiovascular effects in people 
with pre-existing cardiovascular disease (U.S. EPA, 2019a, section 
6.1);
     PM2.5-related respiratory effects in people 
with pre-existing respiratory disease, particularly asthma (U.S. EPA, 
2019a, section 5.1);
     PM2.5-related impairments in lung function 
growth and asthma development in children (U.S. EPA, 2019a, sections 
5.1, 5.2, and 12.5.1.1).
    The 2019 ISA also noted that stratified analyses (i.e., analyses 
that allow for the comparison of PM-related health effects across 
different populations) provided strong evidence for racial and ethnic 
differences in PM2.5 exposures and PM2.5-related 
health risk. Such analyses indicated that certain racial and ethnic 
groups, specifically Hispanic and non-Hispanic Black populations have 
higher PM2.5 exposures than non-Hispanic White populations, 
thus contributing to risk of adverse PM2.5-related health 
effects in minority populations (U.S. EPA, 2019a, section 12.5.4). 
Stratified analyses focusing on other groups also suggested that 
populations with pre-existing cardiovascular or respiratory disease, 
populations that are overweight or obese, populations that have 
particular genetic variants, and populations that are of low 
socioeconomic status (SES) could be at increased risk for 
PM2.5-related adverse health effects (U.S. EPA, 2019a, 
chapter 12).
    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, 2022b, 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. Key aspects of the 
consideration are summarized in section II.A.1.a below.
a. Considerations Regarding the Adequacy of the Existing Standards in 
the 2020 Review
    With the 2020 final decision, the EPA retained the primary 24-hour 
PM2.5 standard, with its level of 35 [micro]g/m\3\, and the 
primary annual PM2.5 standard, with its level of 12.0 
[micro]g/m\3\. The Administrator's conclusions regarding the adequacy 
of the primary PM2.5 standards at the time of the 2020 
review was based on consideration of the evidence, analyses and 
conclusions contained in the 2019 ISA; the quantitative risk assessment 
in the 2020 PA; advice from the CASAC; and public comments. Key 
considerations informing the Administrator's decision to retain the 
standards that were promulgated in the 2012 review are summarized 
below.
    As an initial matter, the 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 Administrator recognized that, 
consistent with the evidence available in prior reviews, the scientific 
evidence in the 2020 review 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 
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 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 sub-daily estimates are less consistent and, in some 
cases, smaller in magnitude (U.S. EPA, 2019a, section 1.5.2.1; U.S. 
EPA, 2020a, section 3.5.2.2). In addition, controlled human exposure 
and panel-based studies of sub-daily 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 
concludes that epidemiologic studies did not indicate that sub-daily 
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, 2020a, section 3.2.3.1). Thus, these 
studies also did not suggest the need for additional protection against 
sub-daily PM2.5 exposures (U.S. EPA, 2020a, section 
3.5.2.2). Therefore, the 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 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, 2020a, 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

[[Page 5577]]

supported a 98th percentile form for a 24-hour standard. The 
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, 2020a, section 3.5.2.3). As such, 
the 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 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, 2020a), 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).
    As an initial matter, the 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 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 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 
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 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 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 
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 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, 2020a, 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, 2020a, 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 available 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, 
2020a, section 3.2.3.1; 85 FR 82716, December 18, 2020).
    With respect to the epidemiologic evidence, the 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, 2020a, section 3.5.1 The Administrator noted 
that this was consistent with CASAC consensus support for retaining the 
current 24-hour standard. Thus, the Administrator concluded that the 
24-hour standard with its level of 35 [mu]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 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

[[Page 5578]]

effects observed in epidemiologic studies. However, the 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 Administrator's judgment. Thus, consistent 
with the approaches in previous NAAQS reviews, the 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 Administrator emphasized uncertainties and limitations that 
were present in epidemiologic studies in previous reviews and persisted 
in the 2020 review. These uncertainties 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 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, 2019a, 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, 2019a, p. 8 
consensus responses).
    Taking into consideration the views expressed by the CASAC members 
who supported retaining the annual standard, the 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, 2020a, 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 Administrator viewed that when assessing the 
mean concentrations of the key short-term and long-term epidemiologic 
studies in the U.S. that use 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 [mu]g/m\3\, a 
concentration level above the existing level of the primary annual 
standard of 12 [mu]g/m\3\. The 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, 2020a, 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 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 [mu]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 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 82716-82717, December 18, 
2020).
---------------------------------------------------------------------------

    \49\ The median of the study-reported mean (or median) 
PM2.5 concentrations is 13.3 [mu]g/m\3\, which was also 
above the level of the existing standard.
---------------------------------------------------------------------------

    In addition to the evidence, the 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 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, 2020a, section 3.3.2.4). Additionally, 
the 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 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 Administrator additionally considered an emerging body of 
evidence from accountability studies that examined past reductions in

[[Page 5579]]

ambient PM2.5 and the degree to which those reductions 
resulted in public health improvements. While the 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, 2020a, 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, 2020a, 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 [mu]g/m\3\ (i.e., U.S. EPA, 2020a, Table 3-3). Given the lack of 
available accountability studies 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 
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 
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 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, the Administrator judged that the 
existing standards provided an adequate margin of safety. With respect 
to the annual standard, the level of 12 [mu]g/m\3\ was below the lowest 
``starting'' concentration (i.e., 13.2 [mu]g/m\3\) in the available 
accountability studies that showed public health improvements 
attributable to reductions in ambient PM2.5. In addition, 
while the 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 [mu]g/
m\3\). In addition, the 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 Administrator's review of the science, 
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 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, the 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 Administrator's view that there were important 
limitations and uncertainties that remained in the evidence. The 
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 [mu]g/m\3\, 35 [mu]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. General Approach and Key Issues in This Reconsideration of the 2020 
Final Decision
    To evaluate whether it is appropriate to consider retaining the 
current primary PM2.5 standards, or whether consideration of 
revision is appropriate, the EPA has adopted an approach in this 
reconsideration that builds upon the general approach used in past 
reviews. This includes the substantial assessments and evaluations 
performed in those reviews, and also takes into account the more recent 
scientific evidence and risk information now available to inform 
understanding of the key policy-relevant issues in the reconsideration. 
As summarized above, the Administrator's decisions in the 2020 review 
were based on an integration of PM health effects information with the 
judgments on the adversity and public health significance of key health 
effects, policy judgments as to when the standard is requisite to 
protect public health with an adequate margin of safety, and 
consideration of CASAC advice and public comments.
    Similarly, in this reconsideration, we draw on the current evidence 
and quantitative assessments of exposure

[[Page 5580]]

pertaining to the public health risk of PM in ambient air. In 
considering the scientific and technical information here, we consider 
both the information available at the time of the 2020 review and 
information more recently available, including that which has been 
critically analyzed and characterized in the 2019 ISA and ISA 
Supplement. The quantitative risk analyses, including a newly conducted 
at-risk analysis, provide a context for interpreting the evidence of 
mortality and the potential public health significance of risks 
associated with air quality conditions that just meet the current and 
potential alternative standards. The overarching purpose of these 
analyses is to inform the Administrator's conclusions on the public 
health protection afforded by the current primary standards, with an 
important focus on evaluating the potential for exposures and risks 
beyond those indicated by the information available at the time the 
current standards were established.

B. Overview of the Health Effects Evidence

    The information summarized here 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 and ISA Supplement and its policy implications are further 
discussed in the 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 mortality and cardiovascular effects for both 
short- and long-term PM2.5 exposure that 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, the ISA 
Supplement evaluates newly available evidence related to short- and 
long-term PM2.5 exposure and mortality and cardiovascular 
effects given the strength of the evidence available in the 2019 ISA 
and past ISAs and AQCDs, as well as the clear adversity of these 
endpoints. 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. While the ISA Supplement does not include 
information for health effects other than mortality and cardiovascular 
effects, the 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, with a focus on those health effects for which 
the 2019 ISA concluded a ``causal'' or ``likely to be causal'' 
relationship.
1. 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, 2009b; 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 summarize the relationships between long- and short-term 
PM2.5 exposures and mortality (II.B.1.a), cardiovascular 
effects (II.B.1.b), respiratory effects (II.B.1.c), cancer (II.B.1.d), 
and nervous system effects (II.B.1.e). 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.

---------------------------------------------------------------------------

[[Page 5581]]

a. Mortality
i. 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 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 (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 [mu]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 (Correia et al., 
2013). In this follow-up study, 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 [mu]g/m\3\ 
decrease in long-term PM2.5 concentration was associated 
with a 0.35-year increase in life expectancy). 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, reported that PM2.5 
concentrations above the lowest observed concentration (2.8 [mu]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 [mu]g/m\3\ to participants living in areas with 
PM2.5 concentrations <12 [mu]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, have 
emerged

[[Page 5582]]

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 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.
    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. Studies that 
employed these alternative methods for confounder control 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. For example, 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 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. For example, 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 chronic obstructive pulmonary disease (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

[[Page 5583]]

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 evaluated in the 2019 ISA, 
including 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 
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).
ii. 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.) \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 (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 [mu]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

[[Page 5584]]

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 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 analyses 
reporting generally 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. 
This evidence 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%) (National Heart, 
Lung, and Blood Institute (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 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 sub-daily metric (i.e., 1-hour 
max). These initial studies 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. 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 [mu]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

[[Page 5585]]

(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 [mu]g/m\3\.
---------------------------------------------------------------------------

    \56\ 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\.
---------------------------------------------------------------------------

    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 supported by evidence from recent 
epidemiologic studies evaluated in the ISA Supplement (U.S. EPA, 2022a, 
section 3.2.1.4, p. 3-69).
b. Cardiovascular Effects
i. 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., 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, 2009b).
    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.B.1.a), 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 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. 
Additionally, an accountability study by Henneman et al. (2019a) 
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.
    The positive associations reported in epidemiologic studies are 
supported by toxicological evidence for increased

[[Page 5586]]

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.
    Moreover, 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 
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).
    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).
ii. 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). When considered as a whole; however, 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.
    The ISA Supplement also includes some 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 report 
positive associations across a number of statistical approaches, 
providing additional support for a relationship between short-term 
PM2.5 exposure and cardiovascular effects, while also

[[Page 5587]]

reducing uncertainties related to potential confounder bias.
    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).
    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. (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, hospital admissions and mortality in some people.
    Animal toxicological studies published since the 2009 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. Direct evidence for an independent effect of 
PM2.5 on cardiovascular effects can be found in a number of 
controlled human exposure and animal toxicological studies, which 
supports the results of epidemiologic studies reporting that 
associations remain relatively unchanged in copollutant models. These 
results concur with epidemiologic panel studies reporting that 
PM2.5 exposure is associated with some of the same 
cardiovascular endpoints reported in experimental studies. 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 concentrations 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).
c. Respiratory Effects
i. 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

[[Page 5588]]

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 in children'' and ``with asthma 
development in children, with increased bronchitis symptoms in children 
with asthma, 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 such associations reported in 
epidemiologic studies of respiratory health effects, animal 
toxicological studies continue to provide direct evidence that long-
term exposure to PM2.5 results in a variety of respiratory 
effects. Animal studies in the 2019 ISA show 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).
ii. 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).
    Building off the studies evaluated in the 2009 ISA, epidemiologic 
studies evaluated in the 2019 ISA expand the assessment of potential 
copollutant confounding. 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 (i.e., O3, 
NO2, SO2, with more limited evidence for CO) and 
other particle sizes (i.e., PM10-2.5) (U.S. EPA, 2019a, 
section 5.1.10.1).
    In the 2019 ISA, the uncertainty related to whether there is an 
independent effect of PM2.5 on respiratory health is also 
partially addressed by findings from animal toxicological studies. 
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. Animal toxicological evidence 
for other respiratory effects is inconsistent and a recent study by 
Wyatt et al. (2020) that was evaluated in the ISA Supplement, conducted 
at near ambient PM2.5 concentrations, adds to the limited 
evidence base of controlled human

[[Page 5589]]

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).
d. 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 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, 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 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).
    The 2019 ISA evaluates the degree to which epidemiologic studies 
have addressed the potential for confounding by copollutants and the 
shape of the C-R relationship. 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 
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 PM exposures. Experimental and epidemiologic studies that 
examine epigenetic effects indicate changes in DNA methylation, 
providing some support for PM2.5 exposure contributing 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, but 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).

[[Page 5590]]

e. 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.
    In addition to the findings described above, which are most 
relevant to older adults, several studies of neurodevelopmental effects 
in children have also been conducted. Positive associations between 
long-term exposure to PM2.5 during the prenatal period and 
autism spectrum disorder (ASD) are observed in multiple epidemiologic 
studies (U.S. EPA, 2019a, section 8.2.7.2), while studies of cognitive 
function provide little support for an association (U.S. EPA, 2019a, 
section 8.2.5.2). 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 regarding 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, and 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).
f. 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 ultrafine particle (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

[[Page 5591]]

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).
---------------------------------------------------------------------------

2. 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 here 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 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

[[Page 5592]]

for asthma exacerbation, and controlled human exposure and animal 
toxicological studies demonstrate 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 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., MI 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 are 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 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 
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 Natives have the 
highest prevalence of all heart disease and coronary heart disease, 
while Blacks 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 prevalence is generally similar across racial groups 
for chronic bronchitis and emphysema. Overall, the prevalence for 
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.
3. 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 PA's evaluation of the 
PM2.5 exposure concentrations that have been examined in 
controlled human exposure studies, animal toxicological studies, and 
epidemiologic studies. The 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 
PM2.5 exposures (U.S. EPA, 2022b, section 3.3.3). As 
described in greater detail in section II.B.1 above, this includes 
mortality, cardiovascular effects, and respiratory effects associated 
with short- or long-term PM2.5 exposures and cancer and 
nervous system effects associated with long-term PM2.5 
exposures. While the causality determinations in the 2019 ISA are 
informed by studies evaluating a wide range of PM2.5 
concentrations, the sections below summarize the considerations in the 
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.

[[Page 5593]]

a. 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, Preamble p. 20). Two types of 
experimental studies are of particular importance in understanding the 
effects 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, Preamble, 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,\62\ 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.
---------------------------------------------------------------------------

    \62\ 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.
---------------------------------------------------------------------------

    Table 3-4 in the PA summarizes information from the 2019 ISA on 
available controlled human exposure studies the 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 in the PA have evaluated average 
PM2.5 concentrations at or above about 100 [mu]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 [mu]g/m\3\, with less consistent evidence for effects following 
exposures to concentrations lower than 120 [mu]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 [mu]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 [mu]g/m\3\ in 
Hemmingsen et al. (2015b), but not for other outcomes following 5-hour 
exposures to 24 [mu]g/m\3\ in Hemmingsen et al. (2015a) and not 
following 24-hour exposures to 10.5 [mu]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 [mu]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 [mu]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).
    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, 
2019, 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).\63\ 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., 
ranging from 2- to 5-hours), including

[[Page 5594]]

those studies that are conducted at near-ambient PM2.5 
concentrations.
---------------------------------------------------------------------------

    \63\ 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 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. These studies, therefore, often do not 
include including children, or 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 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\).\64\ Given these 
results, the 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.
---------------------------------------------------------------------------

    \64\ 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 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 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.
b. Ambient PM2.5 Concentrations in Locations of 
Epidemiologic Studies
    As summarized in section II.B.1 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 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 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 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 describes the consideration of the key epidemiologic 
studies and observations from these studies, as evaluated in the PA 
(U.S. EPA, 2022b, section 3.3.3.2).
i. PM2.5 Air Quality Distributions Associated With Mortality 
or Morbidity in Key Epidemiologic Studies
    As an initial matter, in considering the PM2.5 air 
quality distributions associated with mortality or morbidity in the key 
epidemiologic studies, the PA recognizes that in previous reviews, the 
decision framework used to judge

[[Page 5595]]

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 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 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 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, 2020a, 
sections 3.1.2 and 3.2.3), the 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 
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 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 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).\65\
---------------------------------------------------------------------------

    \65\ 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 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 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 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

[[Page 5596]]

(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 that amount varying 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 annual primary 
PM2.5 standard, the 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.\66\
---------------------------------------------------------------------------

    \66\ 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.\67\ 
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, to further complicate the comparison, when 
looking across these studies, variations in how exposure is estimated 
are present between such studies, which 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 level tract. 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 report 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 PA also considers the methods used to estimate 
PM2.5 concentrations, which vary from traditional methods 
using monitoring data from ground-based monitors \68\ to those using 
more complex hybrid modeling approaches.\69\
---------------------------------------------------------------------------

    \67\ More detailed information about hybrid model methods and 
performance is described in section 2.3.3.2 of the PA (U.S. EPA, 
2022b).
    \68\ 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.
    \69\ 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 PA (U.S. EPA, 2022b).
---------------------------------------------------------------------------

    Given the emergence of the hybrid model-based epidemiologic studies 
since the 2012 review, the 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 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 
PA presents analyses that compare the area annual design values, 
composite monitor PM2.5

[[Page 5597]]

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 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 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 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 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).\70\ The analysis shows 
that the means vary 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 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 are 
included with lower estimated PM2.5 concentrations. 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 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 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.\71\ 
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 and are 15% to 18% higher when 
compared to average annual PM2.5 concentrations 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, with the 
approach that does not utilize population weighting producing a much 
lower value.
---------------------------------------------------------------------------

    \70\ More details on the evaluation of the two hybrid modeling 
approaches is provided in section 2.3.3.2.4 of the PA (U.S. EPA, 
2022b).
    \71\ 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 PA generally concludes that study-reported mean 
concentrations in the studies that employ hybrid modeling approaches 
and population-weight the 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 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 assessment of the information must be recognized (U.S. EPA, 
2022a, section 3.3.3.2.1). First, the comparisons used only two hybrid 
modeling approaches. Although the 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 over time, 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

[[Page 5598]]

concentration to their study population, and the assessment presented 
in the 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 PA focuses on mean 
PM2.5 concentrations and a similar comparison for these 
lower percentiles 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 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 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 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 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.\72\ In particular the 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.\73\ However, and as 
noted above, the 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 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 PA summarize the study details for the key U.S. and 
Canadian epidemiologic studies (U.S. EPA, 2022b, section 
3.3.3.2.1).\74\
---------------------------------------------------------------------------

    \72\ 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.''
    \73\ 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, 2020a, 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.
    \74\ The cohorts examined in the studies included in Figure 3-4 
to Figure 3-7 of the 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 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.\75\ Additionally, for studies that estimate 
PM2.5 exposure using hybrid modeling approaches, the PA also 
considers the approach used to estimate PM2.5 concentrations 
and the approach used to validate hybrid model predictions when 
determining those studies considered as key epidemiologic studies \76\ 
and focuses on those studies that use recent methods based on surfaces 
with fused

[[Page 5599]]

with monitored PM2.5 concentration data (U.S. EPA, 2022b, 
section 3.3.3.2.1).
---------------------------------------------------------------------------

    \75\ 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 
PA focuses on the distribution of ambient PM2.5 
concentrations that could have contributed to reported health 
outcomes. Therefore, the PA identifies studies as key epidemiologic 
studies when the years of air quality data and health data overlap 
in their entirety.
    \76\ 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.\77\ For the small subset of 
studies with available information on the broader distributions of 
underlying data, Figure 1 below also identifies the study-period mean 
PM2.5 concentrations corresponding to the 25th and 10th 
percentiles of health events\78\ (see Appendix B, Section B.2 of the 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\79\ when available 
(see Appendix B, section B.3 for additional information).\80\
---------------------------------------------------------------------------

    \77\ Canadian studies that use ground-based monitors estimate 
long- or short-term PM2.5 exposures are found in Figure 
3-9 of the PA, including concentrations corresponding to the 25th 
and 10th percentiles of estimated exposures or health events, when 
available (U.S. EPA, 2022b).
    \78\ 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 PA and 10% of the total health events occurred in 
study locations with mean PM2.5 concentrations below the 
10th percentiles identified.
    \79\ For most studies in Figure 2 below (Figure 3-14 in the 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.
    \80\ 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 PA (U.S. EPA, 2022b).
---------------------------------------------------------------------------

BILLING CODE 6560-50-P

[[Page 5600]]

[GRAPHIC] [TIFF OMITTED] TP27JA23.001


[[Page 5601]]


[GRAPHIC] [TIFF OMITTED] TP27JA23.002


[[Page 5602]]


BILLING CODE 6560-50-C
    Based on its evaluation of study-reported mean concentrations, the 
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 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 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\ \81\ to 16.5 [mu]g/m\3\ 
(Figure 1 and 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 [mu]g/m\3\ (Figure 2 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 [mu]g/m\3\ to 12.2 
[mu]g/m\3\. In the one study that population-weighted the grid cell 
prior to averaging up to the ZIP code or census tract level report 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.
---------------------------------------------------------------------------

    \81\ 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 from 
the grid cell spatial resolution 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. Additionally, the study mean 
reported in these studies may not reflect the exposure concentrations 
used in the epidemiologic study to assess the reported association. 
Because of this, 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 in calculating the mean, 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 quartiles 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 [mu]g/m\3\ and 10th percentiles of health events 
correspond to PM2.5 concentrations at or above 9.8 [mu]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 and U.S. EPA, 2022b, Figure 3-8). Of the key U.S. 
epidemiologic studies that use hybrid modeling approaches and 
population-weighting to estimate long-term PM2.5 exposures, 
the ambient PM2.5 concentrations corresponding to 25th 
percentiles of estimated exposures are 9.1 [mu]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 [mu]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 [mu]g/m\3\ (U.S. EPA, 2022b, Figure 3-14).\82\ 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 [mu]g/m\3\ to 7.3 
[mu]g/m\3\ (Figure 2 and U.S. EPA, 2022b, Figure 3-14).
---------------------------------------------------------------------------

    \82\ As noted above, in this study (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 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 [mu]g/m\3\ to 9.0 [mu]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 [mu]g/m\3\ and 10th 
percentiles of health events correspond to PM2.5 
concentrations at or above 6.4 [mu]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 [mu]g/m\3\ to just below 
10.0 [mu]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 [mu]g/m\3\ to 8.0 [mu]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 [mu]g/m\3\ and 10.0 [mu]g/m\3\ 
(U.S. EPA, 2022b, Figure 3-11). The four studies that estimate exposure 
at the regional scale

[[Page 5603]]

report mean PM2.5 concentrations that range from 7.8 [mu]g/
m\3\ to 9.8 [mu]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 [mu]g/m\3\ in two studies, and 4.3 [mu]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 
methods for confounder control),\83\ and accountability studies (U.S. 
EPA, 2019a, U.S. EPA, 2021a, U.S. EPA, 2022b).
---------------------------------------------------------------------------

    \83\ 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 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 [mu]g/m\3\) or short-term exposures (e.g., 
daily concentrations <35 [mu]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 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. Further, these studies 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 [mu]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 [mu]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 [mu]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 [mu]g/m\3\ and 8.8 [mu]g/m\3\, which presumably have lower mean 
concentrations than the mean of 7.8 [mu]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 [mu]g/m\3\, but less in analyses 
restricted to <8.8 [mu]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 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 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 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 analyses, 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.\84\ The 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

[[Page 5604]]

quality are associated with improvements in public health.
---------------------------------------------------------------------------

    \84\ 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 to assess whether there was evidence 
of changes in associations with mortality or cardiovascular effects due 
to changes in annual PM2.5 concentrations, Corrigan et al. 
(2018), Henneman et al. (2019b), and Sanders et al. (2020a) present 
analyses with starting concentrations (or concentrations prior to the 
policy or intervention) below 12.0 [mu]g/m\3\. Henneman et al. (2019b) 
explored the 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 [mu]g/m\3\ in 2005 to mean annual PM2.5 
concentrations of 7.2 [mu]g/m\3\ in 2012 from coal-fueled power plants 
resulted in corresponding reductions in the number of cardiovascular-
related 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 for 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 [mu]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.
4. Uncertainties in the Health Effects Evidence
    The 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 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 describe the air quality 
status of a given area relative to the NAAQS.\85\ As discussed above in

[[Page 5605]]

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 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 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.
---------------------------------------------------------------------------

    \85\ 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 (Table A-1) to the 2019 ISA (U.S. EPA, 2019a), 
copollutant models are not without their limitations, such as instances 
for which correlations are high between pollutants resulting in greater 
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, they do not identify particular 
PM2.5 exposures that cause 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 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.

C. 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. Conducting such quantitative analyses, if 
appropriate, could inform judgments about the potential for additional 
public health improvements associated with

[[Page 5606]]

PM2.5 exposure and related health effects and could help to 
place the evidence for specific effects into a broader public health 
context.
    In addition to consideration of the scientific evidence, the 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 interest 
(e.g., ``just meeting'' the current or potential alternative 
standards).
1. 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).\86\ 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).\87\
---------------------------------------------------------------------------

    \86\ Additional detail regarding the selection of epidemiologic 
studies and specification of C-R functions is provided in the PA 
(U.S. EPA, 2022b, Appendix C, section C.1.1).
    \87\ 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 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 [mu]g/m\3\ 
and the 24-hour standard with its level of 35 [mu]g/m\3\). Air quality 
modeling was then used to simulate air quality just meeting an 
alternative standard with a level of 10.0 [mu]g/m\3\ (annual) and 30 
[mu]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 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 [mu]g/m\3\, in conjunction with estimating the health risks 
associated with meeting a revised annual standard with a level of 10.0 
[mu]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 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,\88\ 11 study areas where just meeting the current 
standards is controlled by the daily standard,\89\ 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).\90\
---------------------------------------------------------------------------

    \88\ 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.
    \89\ 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.
    \90\ 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 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 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).
2. 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 are 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\ (a concentrations which is below or near most of the study 
reported

[[Page 5607]]

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). Similarly, the at-risk analysis is also subject to many of 
these same uncertainties. 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 the PA (U.S. EPA, 2022, section 3.4.1.7, section 
3.4.1.8, Appendix C, section C.3).
3. 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, 2022, section 3.4.2.1). 
Additionally, as described in more detail in the PA, the at-risk 
analysis indicates that Black populations may experience 
disproportionally higher exposures and risk under air quality 
conditions just meeting the current primary annual PM2.5 
standard in the study areas, as compared to White populations. Risk 
disparities include exposure disparities, as well as the relationship 
between exposure and health effect and baseline rates of the health 
effect. While risk disparities may be a more meaningful metric, they 
are also subject to additional uncertainties.
    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% a level of 11.0 [mu]g/m\3\, 15-19% for a level 
of 10.0 [mu]g/m\3\, 22-28% for a level of 9.0 [mu]g/m\3\, and 30-37% 
for a level of 8.0 [mu]g/m\3\) (U.S. EPA, 2022b, Table 3-17). Meeting a 
revised annual standard with a lower level may also reduce exposure and 
risk in Black populations slightly more so than in White 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 [mu]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 then 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 [mu]g/m\3\ is associated with reductions 
in estimated risk of 9-13% (U.S. EPA, 2022b, section 3.4.2.3).

D. Proposed Conclusions on the Primary PM2.5 Standards

    In reaching proposed conclusions on the current primary 
PM2.5 standards (presented in section II.D.3), the 
Administrator has taken into account the current evidence and 
associated conclusions in the 2019 ISA and ISA Supplement, in light of 
the policy-relevant evidence-based and risk-based considerations 
discussed in the PA (summarized in section II.D.2), as well as advice 
from the CASAC, and public comment received on the standards thus far 
in the reconsideration (section II.D.1). In general, the role of the PA 
is to help ``bridge the gap'' between the Agency's assessment of the 
current evidence and quantitative analyses (of air quality, exposure, 
and risk), and the judgments required of the Administrator in 
determining whether it is appropriate to retain or revise the NAAQS. 
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 section II.B 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 section II.C 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 level that reduces 
risk sufficiently so as to protect public health, including the health 
of sensitive groups, with an adequate margin of safety.
    The proposed 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. The Administrator's final decisions will 
additionally consider public comments received on these proposed 
decisions.
1. CASAC Advice in This Reconsideration
    The CASAC has provided advice on the adequacy of the current 
primary PM2.5 standards in the context of its review of the 
draft PA.\91\ The range of

[[Page 5608]]

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.
---------------------------------------------------------------------------

    \91\ A limited number of public comments have also been received 
in this reconsideration to date, including comments focused on the 
draft PA. Of the public comments that addressed adequacy of the 
current primary PM2.5 standards, some expressed agreement 
with staff conclusions in the draft PA, while others expressed the 
view that the standards should be more stringent.
---------------------------------------------------------------------------

    In its comments on the draft PA, the CASAC stated that: ``[o]verall 
the CASAC finds the Draft PA to be well-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). 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, 2022a, 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 (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 means 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 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 reaching conclusions on a recommended 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, the CASAC did not reach consensus regarding the adequacy of 
the public health protection provided by the current standard. 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 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).
    The majority of the CASAC favored revising the level of the primary 
24-hour PM2.5 standard and suggested that a range of 25-30 
[mu]g/m\3\ would be adequately protective. In so doing, 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 adequacy of the 24-hour standard, 
recognizing that controlled

[[Page 5609]]

human exposure studies preferentially recruit less susceptible 
individuals and have a typical exposure duration much 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 draft PA to retain the current primary 24-hour 
PM2.5 standard, without revision. In so doing, the minority 
of the CASAC placed greater weight on the risk assessment, noting that 
the risk assessment accounts for both the level and the form of the 
current standard and the way 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 that, because of this, 
``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 that would 
inform future reviews of the primary PM2.5 NAAQS (Sheppard, 
2022a, pp. 14-15 of consensus responses).
2. Evidence- and Risk-Based Considerations in the Policy Assessment
    The main focus of the policy-relevant considerations in the PA is 
consideration of the question: Does the currently available scientific 
evidence- and exposure/risk-based information support or call into 
question the adequacy of the protection afforded by the current primary 
PM2.5 standards? The PA response to this overarching 
question takes into account discussions that address the specific 
policy-relevant questions for this reconsideration, focusing first on 
consideration of the scientific evidence, as evaluated in the 2019 ISA 
and ISA Supplement, including that newly available in this 
reconsideration (section II.D.2.a). The PA also considers the 
quantitative risk estimates drawn from the risk assessment (presented 
in detail in section 3.4 and Appendix C of the PA; U.S. EPA, 2022b) 
including associated limitations and uncertainties, and the extent to 
which they may indicate different conclusions from those in previous 
reviews regarding the magnitude of risk, as well as the level of 
protection from adverse effects, associated with the current and 
alternative standards (section II.D.2.b). The PA additionally considers 
the key aspects of the evidence and exposure/risk estimates that were 
emphasized in previous reviews of the current standards, as well as the 
associated public health policy judgments and judgments about the 
uncertainties inherent in the scientific evidence and quantitative 
analyses that are integral to consideration of whether the currently 
available information supports or calls into question the adequacy of 
the current primary PM2.5 standards (U.S. EPA, 2022b, 
section 3.6).
a. Evidence-Based Considerations
    The currently available evidence on the health effects of 
PM2.5, including evidence newly available in this 
reconsideration, is largely consistent with the evidence that was 
available in previous reviews regarding health effects causally related 
to PM2.5 exposures. Specifically, as in the 2012 review, 
mortality and cardiovascular effects are concluded to be causally 
related to long- and short-term exposures to PM2.5, while 
respiratory effects are concluded to likely be causally related to 
long- and short-term PM2.5 exposures. Also, since the 2012 
review, recent evidence provides additional support that is sufficient 
to conclude that the relationship between long-term PM2.5 
exposures and nervous system effects and cancer are likely to be causal 
(U.S. EPA, 2019a, Table ES-1). These determinations are based on 
evidence from experimental and epidemiologic studies that is newly 
available since the completion of the 2009 ISA (U.S. EPA, 2019, Table 
ES-1). The current evidence base is concluded to be suggestive of, but 
not sufficient to infer, causal relationships between nervous system 
effects and short-term PM2.5 exposures; metabolic effects, 
reproduction and fertility, and pregnancy and birth outcomes and long- 
and short-term PM2.5 exposures (U.S. EPA, 2019a, Table ES-
1). Additionally, the current evidence base supports a suggestive of, 
but not sufficient to infer, a causal relationship for 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, 2019a, Table ES-1).
    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 SES. Other factors that have the potential to 
contribute to increased risk, but for which the evidence is less clear, 
include obesity, diabetes, genetic factors, smoking status, sex, diet, 
and residential location (U.S. EPA, 2019a, chapter 12). In addition to 
these population groups, the 2019 ISA and ISA Supplement conclude that 
there is strong evidence for racial and ethnic differences in 
PM2.5 exposures and PM2.5-related health risk. 
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). Further, there is consistent evidence across 
multiple studies that demonstrate increased risk of PM2.5-
related health effects for Black populations, with the strongest 
evidence for health risk disparities for mortality (U.S. EPA, 2019a, 
section 12.5.4). In addition, studies assessed in the 2019 ISA and ISA 
Supplement also provide evidence of exposure and health risk 
disparities based on SES. The evidence indicates that lower SES 
communities are exposed to higher concentrations of PM2.5 
compared to higher SES communities (U.S. EPA, 2019a, section 12.5.3; 
U.S. EPA, 2022b, section 3.3.3.1.1). Additionally, evidence supports 
the conclusions that lower SES

[[Page 5610]]

is associated with cause-specific mortality and certain health 
endpoints (i.e., MI and CHF), but less so for all-cause or total (non-
accidental) mortality (U.S. EPA, 2019a, section 12.5.3; U.S. EPA, 
2022b, section 3.3.3.1).
    Consistent with the evidence available in the 2009 ISA, controlled 
human exposure studies have demonstrated effects on cardiovascular 
function following 1- to 5-hour exposures to PM2.5, with the 
most consistent evidence for impaired vascular function. The PA notes 
that most of the controlled human exposure studies have evaluated 
average PM2.5 concentrations at or above about 100 [mu]g/
m\3\, with exposure durations up to two hours. These studies have 
often, though not always, reported statistically significant effects on 
one or more indicators of cardiovascular function following 2-hour 
exposures to average PM2.5 concentrations at and above about 
120 [mu]g/m\3\, with less consistent effects following exposures to 
concentrations lower than 120 [mu]g/m\3\.
    In considering the controlled human exposure studies in reaching 
conclusions on the primary PM2.5 standards, the PA notes 
that air quality analyses indicate that 2-hour PM2.5 
concentrations to which individuals were exposed in most of these 
studies, including those that report the most consistent results, are 
well-above the ambient PM2.5 concentrations typically 
measured in locations meeting the current primary standards. 
Additionally, the PA recognizes that the results are variable across 
controlled human exposure studies that evaluated near-ambient 
PM2.5 concentrations.
    Furthermore, the PA recognizes that controlled human exposure 
studies often include small numbers of individuals and do not include 
populations that are at increased risk of PM2.5-related 
health effects (e.g., children). While the PA recognizes that the 
controlled human exposure studies are important in establishing 
biological plausibility, it emphasizes that it is unclear how the 
results from these studies alone, particularly in studies conducted at 
near-ambient PM2.5 concentrations, and the importance of the 
effects observed in the studies should be interpreted with respect to 
adversity to public health.
    With regard to the animal toxicological studies, the PA recognizes 
that, unlike the controlled human exposure studies that provide insight 
on the exposure concentrations that directly elicit health effects in 
humans, there is uncertainty associated with translating the 
observations in the animal toxicological studies to potential adverse 
health effects in humans. The PA notes that the interpretation of these 
studies is complicated by the fact that PM2.5 concentrations 
in animal toxicological studies are much higher than those shown to 
elicit effects in human populations. Moreover, the PA recognizes that 
there are also significant anatomical and physiological difference 
between animal models and humans. In considering the information from 
the animal toxicological studies, the PA specifically notes two 
studies, one of which is newly available in the 2019 ISA, that report 
serious effects following long-term exposures to PM2.5 
concentrations close to the ambient concentrations reported in some 
epidemiologic studies, although still above the ambient concentrations 
likely to occur in areas meeting the current primary standards (U.S. 
EPA, 2022b, section 3.3.3.1).
    Since the 2012 review, a large number of epidemiologic studies have 
become available that report generally positive, and often 
statistically significant, associations between long- and short-term 
PM2.5 exposures and mortality and morbidity. Available 
studies additionally indicate that PM2.5 health effect 
associations are robust across various approaches to estimating 
PM2.5 exposures and across various exposure windows. Since 
the 2012 review, there are also a number of studies that employ 
alternative methods for confounder control that further inform the 
causal nature of the relationship between long- or short-term term 
PM2.5 exposure and mortality, and these studies provide 
support for the findings from the broad body of epidemiologic studies.
    In addition to broadening our understanding of the health effects 
that can result from exposures to PM2.5 and strengthening 
support for some key effects (e.g., nervous system effects, cancer, and 
metabolic effects), recent epidemiologic studies strengthen support for 
health effect associations at relatively low ambient PM2.5 
concentrations. Studies that examine the shapes of C-R functions over 
the full distribution of ambient PM2.5 concentrations have 
not identified a threshold concentration below which associations no 
longer exist (U.S. EPA, 2019a, section 1.5.3; U.S. EPA, 2022a, sections 
2.2.3.1 and 2.2.3.2). While such analyses are complicated by the 
relatively sparse data available at the lower end of the air quality 
distribution (U.S. EPA, 2019a, section 1.5.3), 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.
    Consistent with previous reviews, the PA notes that the use of 
information from epidemiologic studies to inform conclusions on the 
current 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 do not identify a population-level 
threshold below which it can be concluded with confidence that PM-
associated health effects do not occur (U.S. EPA, 2019a, section 
1.5.3). However, the study-reported ambient PM2.5 
concentrations reflecting estimated exposure in the middle portion of 
the PM2.5 air quality distribution, which corresponds to the 
bulk of the underlying data, provide the strongest support for reported 
health effect associations and can inform conclusions on the current 
and potential alternative standards. In considering this information, 
the PA recognizes that the mean PM2.5 concentrations 
reported by key epidemiologic studies differ in how mean concentrations 
were calculated, as well as their interpretation in what means 
represent in the context of the current standards.
    In identifying key epidemiologic studies for consideration, the PA 
places the greatest emphasis on studies conducted in the U.S. and 
Canada, although recognizes a number of limitations associated with 
interpreting the results of Canadian studies compared to studies 
conducted in the U.S. Generally, there are differences in the exposure 
environments and population characteristics between the U.S. and other 
countries, including Canada, that can affect the study-reported mean 
PM2.5 concentration and its comparability with the annual 
standard level. A number of other differences, including sources and 
pollutant mixtures, concentration gradients, and populations densities, 
can make it challenging to interpret the mean PM2.5 
concentrations in Canadian studies in the context of a U.S.-based 
standard. Specifically, it may be difficult to use such studies to 
directly and quantitatively inform questions regarding the adequacy of 
the current or potential alternative levels of the annual standard. 
Therefore, while the PA considers the mean PM2.5

[[Page 5611]]

concentrations from U.S. and Canadian studies in reaching conclusions, 
it notes that the U.S.-based epidemiologic studies are most informative 
for comparisons with the annual standard metric and for reaching 
conclusions on the current standards and for informing potential 
alternative levels of the standard.
    Consistent with previous reviews, in considering information that 
can be used from the available epidemiologic evidence to inform 
proposed decisions on the current standards, the PA focuses on 
PM2.5 concentrations near or somewhat below long-term mean 
concentrations reported in epidemiologic studies. In so doing, the PA 
notes that, in previous reviews, the epidemiologic studies used ground-
based monitors to estimate exposures, and that, in addition to newly 
available monitor-based studies, there are also newly available 
epidemiologic studies estimate exposures using hybrid modeling 
approaches. In considering how the study-reported mean PM2.5 
concentrations reported in studies using hybrid modeling approaches 
compare to studies using ground-based monitors, the PA notes that the 
hybrid modeling approaches provide a broader estimation of 
PM2.5 exposures compared to monitor-based studies (i.e., 
because hybrid modeling studies include PM2.5 concentrations 
estimated in areas without monitors). However, compared to monitor-
based studies, the PA recognizes that it is more difficult to relate 
these means to an annual standard metric which relies on maximum 
monitor design values to assess compliance. Further complicating the 
comparison is the variability in how PM2.5 concentrations 
are estimated between studies that use hybrid modeling approaches. 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).
    As described in more detail in section I.D.5 above, the PA included 
analyses that considered how the study-reported mean PM2.5 
concentrations were computed and how the means 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 so doing, the PA included a comparison of 
PM2.5 fields in estimating exposure relative to design 
values using two hybrid modeling surface with annual average 
PM2.5 concentrations estimated per year at a 1 km x 1 km 
spatial resolution. The PA notes that the means vary when 
PM2.5 concentrations are estimated in urban areas only 
(CBSAs) versus when the averages were calculated with all or most grid 
cells nationwide. This is likely indicative of the fact that areas 
included outside of CBSAs tend to be more rural and have lower 
estimated PM2.5 concentrations. The PA acknowledges that 
this is an important consideration 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 are 
included with lower estimated PM2.5 concentrations. 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 long-term mean concentration across the study area and 
period.
    Additionally, the PA analyses indicate that for the studies using 
the hybrid modeling approaches, the use of population weighting in 
calculating study-reported mean PM2.5 concentrations, and 
not a difference in estimates of exposures in the study itself, can 
produce substantially different study-reported mean PM2.5 
concentrations compared to an approach that does not utilize population 
weighting. In studies that do not apply population weighting in the 
calculation of the mean PM2.5 concentrations, study-reported 
means are lower, as a result of including areas with lower estimated 
PM2.5 concentrations that may not be as densely populated, 
as well as areas that may not include health events. To elaborate, in 
hybrid modeling approaches that present mean PM2.5 
concentrations based on an average PM2.5 concentration 
across all grid cells (i.e., do not apply aspects of population 
weighting), health events may not exist in each grid cell, and thus the 
mean reported PM2.5 concentration is not necessarily based 
on the mean PM2.5 concentrations assigned as the exposure in 
the health study. In other words, the mean PM2.5 
concentration that is reported and based on an average of all grid 
cells is not necessarily the same as the mean PM2.5 
concentration for each person assigned an exposure in the study. This 
is an important consideration, as the purpose of the epidemiologic 
study is to evaluate whether an association between PM2.5 
exposure and health outcomes exists. As such, it is unclear whether the 
mean concentration reported using each grid cell is associated with a 
health outcome (i.e., not all grid cells have health events). This 
leads to uncertainty in evaluating how the mean concentration can be 
used in the context of the approach above to evaluate the adequacy of 
the standard as well as potential alternative levels of the annual 
standard.
    In considering the variability in how exposure in estimated between 
studies that use hybrid modeling approaches, the PA focuses on the key 
epidemiologic studies that use hybrid modeling approaches and apply 
population weighting in calculating the study-reported mean, as well as 
those studies that use monitors to estimate exposure, as described in 
more detail in section II.B.3.b above. For key U.S. epidemiologic 
studies that use monitors to estimate PM2.5 exposures, 
overall mean PM2.5 concentrations range between 9.9 [mu]g/
m\3\ \92\ to 16.5 [mu]g/m\3\ (U.S. EPA, 2022b, Figure 3-8). For U.S. 
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 12.2 [mu]g/m\3\ (U.S. EPA, 2022b, Figure 3-14). In 
U.S. studies that average up from the grid cell level to the ZIP code 
or census tract level, mean PM2.5 concentrations range from 
9.8 [mu]g/m\3\ to 12.2 [mu]g/m\3\. In the one U.S. study that 
population-weighted the grid cells prior to averaging up to the ZIP 
code or census tract level, the reported mean PM2.5 
concentration is 9.3 [mu]g/m\3\. As described above, the PA also 
considers the study-reported means from the key Canadian epidemiologic 
studies, which are consistently much lower than those reported for key 
U.S. epidemiologic studies, while noting that for the reasons described 
above, there are uncertainties and limitations associated with 
comparisons between Canadian studies and the annual standard metric. 
For the key Canadian epidemiologic studies that use monitors to 
estimate PM2.5 exposures, overall mean PM2.5 
concentrations range from 6.9 [mu]g/m\3\ to 13.3 [mu]g/m\3\, while the 
range of mean PM2.5 concentrations in Canadian studies that 
use hybrid modeling (all of which average up to postal codes and thus 
include some aspects of population weighting) is 5.9 [mu]g/m\3\ to 9.8 
[mu]g/m\3\.
---------------------------------------------------------------------------

    \92\ 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).

---------------------------------------------------------------------------

[[Page 5612]]

    As described in more detail in section II.B.3.b above, in assessing 
the range of reported exposure concentrations for which the strongest 
support exists for adverse health effects occurring, the PA evaluates 
whether the available evidence supports or calls into question the 
adequacy of public health protection afforded by the current primary 
annual PM2.5 standard against these exposure concentrations. 
This means, as in past reviews, the 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. Based on the 
air quality analyses in presented in the PA and discussed above 
(section I.D.5.a and section I.D.5.b), design values associated with 
the study-reported means in these key U.S. based epidemiologic studies 
are only somewhat higher: 10-20% for monitor-based studies and 15-18% 
higher for the studies that include hybrid modeling approaches and 
utilize population weighting. Based on these results, it can generally 
be concluded that the study-reported mean concentrations in the studies 
are associated with air quality conditions that would be achieved by 
meeting annual standard levels that are 10-20% higher and 15-18% higher 
than study-reported means for monitor-based studies and hybrid 
modeling-based studies that use population weighting, respectively. 
Therefore, an annual standard level that is no more than 10-20% higher 
than the study-reported means in the monitor-based studies (i.e., 9.9-
16.5 [mu]g/m\3\), and no more than 15-18% higher than the study-
reported means in the studies that include hybrid modeling approaches 
and utilize population weighting (i.e., 9.3-12.2 [mu]g/m\3\), would 
generally maintain air quality exposures at or below those associated 
with the study-reported mean PM2.5 concentrations, exposures 
for which we have the strongest support for adverse health effects 
occurring. This relationship is indicative of the fact that 
PM2.5 exposures in an area are represented by a distribution 
of concentrations across that area, with the annual standard level at 
the design value monitor being associated with the highest annual 
average exposure concentration for that area.
    In addition to the study-reported mean concentrations, in 
considering the level of the annual standard, the PA uses an approach 
consistent with that used in previous reviews and also considers 
reported PM2.5 concentrations corresponding to the 25th and 
10th percentiles of health data or exposure estimates when available in 
the key epidemiologic studies. In using such an approach, the PA 
recognized that there is an interrelatedness of the distributional 
statistics in epidemiologic studies (e.g., 10th and 25th percentiles of 
PM2.5 concentrations) and a range of one standard deviation 
around the mean which contains approximately 68% of normally 
distributed data, in that one standard deviation below the mean falls 
between the 25th and 10th percentiles (U.S. EPA, 2022b, p. 2-71). 
Further, the PA notes that in past reviews, some weight was placed on 
studies that provided mean PM2.5 concentrations around the 
25th percentile of the distributions of deaths and cardiovascular-
related hospitalizations and the Administrator judged the region around 
the 25th percentile as a reasonable part of the distribution to guide 
the decision on the appropriate standard level (78 FR 3161, January 15, 
2013).
    As such, the PA concludes that focusing on concentrations somewhat 
below the means (e.g., 25th and 10th percentiles), when such 
information is available from epidemiologic studies, is a reasonable 
approach for considering lower portions of the air quality 
distribution. However, the PA recognizes that the health data are 
appreciably more sparse and an understanding of the magnitude and 
significance of the associations correspondingly become more uncertain 
in the lower part of the air quality distribution. While health effects 
can occur over the entire distribution of ambient PM2.5 
concentrations evaluated, and epidemiologic studies do not identify a 
population-level threshold below which it can be concluded with 
confidence that PM-associated health effects do not occur (U.S. EPA, 
2019a, section 1.5.3), using values below the 10th percentile would 
lead to even greater uncertainties and diminished confidence in the 
magnitude and significance of the associations.
    In considering the available key U.S. epidemiologic studies, the PA 
notes that a small number of studies report PM2.5 
concentrations corresponding to the 25th and 10th percentiles of health 
data or exposure estimates that can be considered to provide insight 
into the concentrations that comprise the lower quartiles of the air 
quality distributions is examined below. 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 [mu]g/m\3\ 
and 10th percentiles of health events correspond to PM2.5 
concentrations at or above 9.8 [mu]g/m\3\ (i.e., 25% and 10% of health 
events, respectively, occur in study locations with PM2.5 
concentrations below these values) (U.S. EPA, 2022b, Figure 3-8). Of 
the key U.S. epidemiologic studies that use hybrid modeling approaches 
to estimate long-term PM2.5 exposures, the ambient 
PM2.5 concentrations corresponding to 25th percentiles of 
estimated exposures are 9.1 [mu]g/m\3\ (U.S. EPA, 2022b, Figure 3-14). 
In key U.S. epidemiologic studies that use hybrid modeling approaches 
to estimate short-term PM2.5 exposures, the ambient 
concentrations corresponding to 25th percentiles of estimated 
exposures, or health events, are 6.7 [mu]g/m\3\ and the ambient 
PM2.5 concentration corresponding to that 10th percentile 
range from 4.7 [mu]g/m\3\ to 7.3 [mu]g/m\3\ (U.S. EPA, 2022b, Figure 3-
14).
    As with the mean PM2.5 concentrations, in considering 
these values relative to an area annual design value, the PA notes the 
25th and 10th percentiles provide information about the lower quartiles 
of the air quality distributions, while the study-reported mean 
provides information about the average or typical exposures, and the 
corresponding area annual design value provides the highest average 
annual PM2.5 concentration being measured. In this way, the 
PA recognizes that all of these metrics (i.e., lower percentiles, study 
mean, annual design value) have a relationship relative to the other, 
and each of these metrics can be used to inform the consideration of 
the level of the current annual standard. Further, the PA recognizes 
that the air quality analyses described above (section I.D.5) and in 
the PA (U.S. EPA, 2022b, section 2.3.3.1 and section 2.3.3.2.4) that 
evaluated the relationship between a mean PM2.5 
concentration in an area and the design value focuses on mean 
PM2.5 concentrations and similar analyses were not conducted 
for other PM2.5 concentrations in the lower portion of the 
air quality distribution. Therefore, given the lack of additional 
information regarding the relationship between percentiles of the air 
quality distribution other than the mean and the annual design value, 
the PA concludes that any direct comparison of study-reported 
PM2.5 concentrations corresponding to lower percentiles 
(e.g., 25th and/or 10th) and annual design values is more uncertain 
than such comparisons with the mean.
    Since the completion of the 2009 ISA, a number of epidemiologic 
studies have become available that can provide

[[Page 5613]]

additional consideration to inform conclusions regarding the adequacy 
of the current standards. Studies that examine health effect 
associations in analyses that exclude the highest exposures (i.e., 
studies that restrict analyses below certain PM2.5 
concentrations), and which report positive and statistically 
significant associations in analyses restricted to annual average 
PM2.5 exposures at or below 12 [mu]g/m\3\ and/or to daily 
exposures below 35 [mu]g/m\3\ (section II.B.3.b above and U.S. EPA, 
2022b, Table 3-10). The PA notes that these restricted analyses provide 
additional support for effects at lower concentrations, exhibiting 
associations for mean concentrations presumably below the mean 
concentrations for the main analyses. While mean PM2.5 
concentrations for these restricted analyses may not be reported in 
most studies, the PA asserts that it would not be unreasonable to 
presume that the mean PM2.5 concentrations in the restricted 
analyses are less than the study-reported mean PM2.5 
concentrations in the main analyses. The two studies (Di et al., 2017b, 
and Dominici et al., 2019) which 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 report mean PM2.5 concentrations 
of 9.6 [mu]g/m\3\. However, it is important to note that, even if the 
other studies had reported the mean PM2.5 concentrations for 
the restricted analysis, these means would not necessarily have been 
useful in the context of the decision framework as was used in past 
reviews (above in section II.B.3.b.), given uncertainties associated 
with identifying the relationship between a calculated mean 
concentration that excludes specific daily or annual average 
concentrations above a certain threshold and the design value used to 
determine compliance with a standard (either the annual or 24-hour 
standard). Moreover, the PA emphasizes there is uncertainty in how 
studies exclude concentrations (e.g., at what spatial resolution are 
concentrations being excluded), which would make any comparisons of 
mean concentrations in restricted analyses difficult to compare to 
design values.
    The PA also takes note of studies that restrict 24-hour average 
PM2.5 concentrations to values of less than 35 [mu]g/m\3\ 
and again recognizes that these studies do not report the mean 
PM2.5 concentration for the restricted analysis, as noted 
above, although the mean of the restricted analysis is presumably less 
than the mean PM2.5 concentration in the main analysis. 
However, in some studies, the majority of PM2.5 
concentrations from the main study are already less than the restricted 
concentration (e.g., in Di et al., 2017a, where of all case and control 
days, 93.6% had PM2.5 concentrations below 25 [mu]g/m\3\), 
which contributes to the uncertainty in how much lower a mean 
concentration in a restricted study is compared to the mean 
PM2.5 concentration in the main analysis. As a result, the 
PA recognizes that there are limitations in how this information can be 
used in evaluating the adequacy of the current or potential alternative 
levels of the 24-hour standard. Additionally, the PA further recognizes 
that it is difficult to use the means, when reported, from studies of 
restricted analyses to evaluate the level of protection afforded by the 
current or potential alternative levels of the primary 24-hour 
PM2.5 standard because the relationship between the study-
reported mean concentration and the 98th percentile form of the 24-hour 
standard is not well understood, in particular for a short-term 
standard designed to limit exposures to peak PM2.5 
concentrations.
    Finally, the PA notes the availability of accountability studies, 
which evaluate whether environmental policies or air quality 
interventions led to changes in air quality and are also associated 
with improvements in public health, including a number of recent 
studies evaluated in the ISA Supplement (summarized above in section 
II.B.3.b and U.S. EPA, 2022b, Table 3-12). These studies report 
positive and significant associations, including some studies with 
annual PM2.5 concentrations below 12.0 [micro]g/m\3\ at the 
start of the study period, indicating that public health improvements 
may occur following PM2.5 reductions in areas that already 
meet the current annual PM2.5 standard. For example, the PA 
notes that the studies by Corrigan et al. (2018) and Sanders et al. 
(2020a) and both found improvements in mortality rates due to 
improvements in air quality in both attainment and nonattainment areas 
following implementation of the 1997 primary annual PM2.5 
NAAQS. Additionally, the PA notes that an accountability study by 
Henneman et al. (2019a) evaluated the changes in modeled 
PM2.5 concentrations following the retirement of coal fired 
power plants in the U.S found that reductions in PM2.5 
concentrations resulted in reductions of cardiovascular-related 
hospital admissions.\93\ Other recent studies additionally report that 
declines in ambient PM2.5 concentrations over a period of 
years have been associated with decreases in mortality rates and 
increases in life expectancy, improvements in respiratory development, 
and decreased incidence of respiratory disease in children, further 
supporting the robustness of PM2.5 health effect 
associations reported in the epidemiologic evidence.
---------------------------------------------------------------------------

    \93\ We note that the studies by Corrigan et al. (2018) and 
Sanders et al. (2020a) report monitor-based average PM2.5 
concentrations, and the study by reports model-based average 
PM2.5 concentrations, and that these studies do not 
report design values.
---------------------------------------------------------------------------

    In considering the available scientific evidence, the PA recognizes 
that there are a number of uncertainties associated with the evidence 
that persist from previous reviews. The PA notes that, for controlled 
human exposures studies, there are uncertainties related to 
inconsistent results observed at concentrations near ambient 
PM2.5 levels. Additionally, the PA recognizes that it is 
unclear how the results of controlled human exposure studies alone and 
the importance of the effects observed in these studies, particularly 
in studies conducted at near-ambient PM2.5 concentrations, 
should be interpreted with respect to adversity to public health. With 
respect to animal toxicological studies, the PA notes that while these 
studies also help establish biological plausibility, uncertainty exists 
in extrapolating the effects observed in animal toxicological studies, 
and the PM2.5 concentrations that cause those effects, to 
human populations.
    Furthermore, the PA recognizes that uncertainties associated with 
the epidemiologic evidence (e.g., the potential for copollutant 
confounding and exposure measurement error) remain, although new 
studies evaluated in the ISA Supplement employ statistical methods such 
as alternative methods for confounder control, to more extensively 
account for confounders, which are more robust to model 
misspecification. With regard to controlling for potential confounders 
in particular, the PA notes that the key epidemiologic studies use a 
wide array of approaches to control for potential confounders. Time-
series studies control for potential confounders that vary over short 
time intervals (e.g., including temperature, humidity, dew point 
temperature, and day of the week), while cohort studies control for 
community- and/or individual-level confounders that vary spatially 
(e.g., including income, race, age, SES,

[[Page 5614]]

smoking, body mass index, and annual weather variables such as 
temperature and humidity) (U.S. EPA, 2022b, Table B-4). Sensitivity 
analyses indicate that adding covariates to control for potential 
confounders can either increase or decrease the magnitude of 
PM2.5 effect estimates, depending on the covariate, and that 
none of the covariates examined can fully explain the association with 
mortality (e.g., Di et al., 2017b, Figure S2 in Supplementary 
Materials). Thus, while no individual study adjusts for all potential 
confounders, a broad range of approaches have been adopted across 
studies to examine confounding, supporting the robustness of reported 
associations. Available studies additionally indicate that 
PM2.5 health effect associations are robust across various 
approaches to estimating PM2.5 exposures and across various 
exposure windows. This includes recent studies that estimate exposures 
using ground-based monitors alone and studies that estimate exposures 
using data from multiple sources (e.g., satellites, land use 
information, modeling), in addition to monitors. While none of these 
approaches eliminates the potential for exposure error in epidemiologic 
studies, the PA concludes that such error does not call into question 
the fundamental findings of the broad body of PM2.5 
epidemiologic evidence.
    Additionally, the PA notes the uncertainties associated with the 
studies that examine the shapes of C-R functions over the full 
distribution of ambient PM2.5 concentrations have not 
identified a threshold concentration, below which associations no 
longer exist (section II.B.4 above, U.S. EPA, 2019a, section 1.5.3; 
U.S. EPA, 2022a, sections 2.2.3.1 and 2.2.3.2). While such analyses are 
complicated by the relatively sparse data available at the lower end of 
the air quality distribution (U.S. EPA, 2019a, section 1.5.3), 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.
    While studies using hybrid modeling methods have demonstrated 
reduced exposure measurement error and reduced uncertainty in the 
health effect estimates, these methodologies have inherent limitations 
and uncertainties, as described in more detail above in section 
II.B.3.b and in sections 2.3.3.1.5 and 3.3.4 of the PA, and the 
performance of the modeling approaches depends on the availability of 
monitoring data which varies by location. Factors likely contributing 
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, the PA concludes that the uncertainty in 
hybrid model predictions becomes an increasingly important 
consideration as lower predicted concentrations are considered.
    In addition, the PA recognizes that there are uncertainties and 
limitations in the analysis evaluating the comparison of estimated 
PM2.5 concentrations using hybrid modeling surfaces and 
their relationship to design values that should be considered (section 
II.B.3.b above; U.S. EPA, 2022b, section 2.3.3.2.4). While design 
values in general are higher than estimated PM2.5 
concentrations using these two hybrid modeling approaches (DI2019 and 
HA2020), the PA recognizes that these are just two hybrid modeling 
approaches to estimating PM2.5 concentrations and other 
models/approaches/spatial scales may result in somewhat different 
PM2.5 concentrations and relationships with design values. 
The analysis evaluating the relationship between two different hybrid 
modeling surfaces and design values estimates PM2.5 
concentrations by CBSAs, but not every health study uses 
PM2.5 estimates at this spatial scale, and spatial scales 
for exposure estimates can vary by study (section I.D.5 above; U.S. 
EPA, 2022b, section 2.3.3.2.4). The analysis completed was a nationwide 
analysis and ratios between design values and mean concentrations are 
based on national estimates. However, not all health studies are 
national studies (i.e., some studies are completed in different regions 
of the country, like the southeast or northeast) and ratios in 
different parts of the country could be higher or lower, depending on 
factors like population, as well as the proportion of rural versus 
urban areas. This analysis used specific air quality years (2000-2016) 
and the use of other air quality years could result in higher or lower 
ratios.
    Regardless of whether an epidemiologic study uses monitoring data 
or a hybrid modeling approach when estimating PM2.5 
exposures, the PA recognizes that it is challenging to interpret the 
study-reported mean PM2.5 concentrations and how they 
compare to design values. This is particularly true given the 
variability that exists across the various approaches to estimate 
exposure and to calculate the study-reported mean. The PA also 
acknowledges that these types of challenges are also present in using 
information from Canadian studies to directly and quantitatively inform 
questions on the level of the annual standard given the difficulty of 
interpreting what the Canadian study means represent relative to U.S. 
design values.
b. Risk-Based Considerations
    As in previous reviews, consideration of the scientific evidence in 
this reconsideration is informed by results from a quantitative 
analysis of risk. The overarching PA consideration regarding these 
results is whether they alter the overall conclusions from previous 
reviews regarding health risk associated with exposure to 
PM2.5 in ambient air and associated judgments on the 
adequacy of public health protection provided by the current primary 
PM2.5 standards. The risk assessment conducted for this 
reconsideration develops exposure and risk estimates for populations in 
47 urban study areas, as well as subsets of those study areas depending 
on which of the primary PM2.5 standards is controlling in a 
given study area. 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. These study areas include nearly 60 million people ages 30 
years or older and illustrate the differences likely to occur across 
various locations with such air quality as a result of area-specific 
differences in emissions, meteorological, and population 
characteristics. While the same conceptual air quality scenarios are 
simulated in all study areas (i.e., conditions that just meet the 
existing or alternate standards), source, meteorological and population 
characteristics in the study areas contribute to variability in the 
estimated magnitude of risk across study areas (U.S. EPA, 2022b, 
section 3.6.2.1). In this way, the 47 areas provide a variety of 
examples of exposure patterns that can be informative to the 
Administrator's consideration of potential exposures and risks that may 
be associated with air quality conditions occurring under the current 
and potential alternative PM2.5 standards.
    In considering the risk assessment in this reconsideration, the PA 
notes a number of ways in which the current analyses update and improve 
upon

[[Page 5615]]

those available in previous reviews. As an initial matter, the PA notes 
that, consistent with the overall approach for this reconsideration, 
the risk assessment has a targeted scope that focuses on all-cause or 
nonaccidental mortality associated with long- and short-term 
PM2.5 exposures (U.S. EPA, 2022b, section 3.4.1.2). As noted 
in section II.B.1 above, the evidence assessed in the 2019 ISA and ISA 
Supplement support a causal relationship between long- and short-term 
PM2.5 exposures and mortality. 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 and 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 2.1).
    The risk assessment also includes updates and improvements to input 
data and modeling approaches, summarized in section II.C above and in 
section 3.4 of the PA (U.S. EPA, 2022b). As in previous reviews, 
exposure and risk are estimated from air quality scenarios defined by 
the highest design value in the study area, which is the monitor 
location with the highest 3-year average of the annual mean 
PM2.5 concentrations (e.g., equal to 12.0 [mu]g/m\3\ for the 
current standard scenario) for the annual PM2.5 standard and 
with the highest 3-year average of the 98th percentile 24-hour 
PM2.5 concentrations (e.g., equal to 35 [mu]g/m\3\ for the 
current standard scenario) for the 24-hour PM2.5 standard. 
As described in more detail in section II.C above and in section 3.4 of 
the PA (U.S. EPA, 2022b), air quality modeling was used to simulate 
just meeting the existing annual and 24-hour standards of 12.0 [mu]g/
m\3\ and 35 [mu]g/m\3\ and to just meeting potential alternative annual 
and 24-hour standards of 10.0 [mu]g/m\3\ and 30 [mu]g/m\3\. In addition 
to the air quality modeling approach, linear interpolation and 
extrapolation were used 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\) in the subset of study areas controlled by 
the annual standard.
    In addition to the risk assessment described above, the PA presents 
quantitative analyses that also assess long-term PM2.5-
attributable exposure and mortality risk, stratified by racial/ethnic 
demographics. As described in more detail in section II.B.2 above, the 
evidence suggests that different racial and ethnic groups, such as 
Black and Hispanic populations residing in the study areas, have higher 
PM2.5 exposures than White and non-Hispanic populations also 
residing in the study areas, respectively, thus contributing to 
increased risk of PM-related effects. Of the available studies, Di et 
al. (2017b) was identified as best characterizing populations 
potentially at increased risk of long-term exposure-attributable all-
cause mortality effects and provides race- and ethnicity-stratified C-R 
functions for ages 65 and over (U.S. EPA, 2022b, section 3.4.1.6 and 
Appendix C). Risk and exposure are quantitatively assessed within 
racial and ethnic minority populations of older adults in the full set 
of 47 areas and the subset of 30 areas controlled by the annual 
PM2.5 standard. This analysis, when considered alongside 
estimates of risk across all populations in the 47 study areas, can 
help to inform conclusions on the annual primary PM2.5 
standards that would be requisite to protect the public health of 
demographic populations potentially at increased risk of long-term 
PM2.5-related mortality effects.
    In considering the risk results, the PA focuses first on estimates 
for the full set of 47 urban study areas. The risk assessment estimates 
that the current primary PM2.5 standards could allow a 
substantial number of deaths in the U.S., with the large majority of 
those deaths associated with long-term PM2.5 exposures. For 
example, when air quality in the 47 study areas is adjusted to just 
meet the current standards, the risk assessment estimates about 41,000 
to 45,000 deaths from all-cause mortality in a single year (e.g., for 
long-term exposures; confidence intervals range from about 30,000 to 
59,000) (U.S. EPA, 2022b, section 3.4.2.1). For the 30 study areas \94\ 
where just meeting the current standards is controlled by the annual 
standard,\95\ long-term PM2.5 exposures are estimated to be 
associated with as many as 39,000 (confidence intervals range from 
about 26,000 to 51,000) deaths from all-cause mortality in a single 
year (U.S. EPA, 2022b, section 3.4.2.2). For the 11 study areas \96\ 
where just meeting the current standards is controlled by the daily 
standard,\97\ long-term PM2.5 exposures are estimated to be 
associated with as many as 2,600 (confidence intervals ranging from 
1,700 to 3,400) deaths in a single year (U.S. EPA, 2022b, section 
3.4.2.3). The risk assessment estimates far fewer deaths in a single 
year for short-term PM2.5 exposures as compared to long-term 
PM2.5 exposures, across all of the study area subsets (U.S. 
EPA, 2022b, section 3.6.2.2).
---------------------------------------------------------------------------

    \94\ These 30 areas controlled by the annual standard under all 
scenarios evaluated include a population of approximately 48 million 
adults aged 30-99, or about 75% of the population included in the 
full set of 47 areas.
    \95\ 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.
    \96\ These 11 areas controlled by the 24-hour standard under all 
scenarios evaluated include a population of approximately 10 million 
adults aged 30-99, or about 17% of the population included in the 
full set of 47 areas.
    \97\ 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.
---------------------------------------------------------------------------

    While the absolute numbers of estimated deaths vary across exposure 
durations, populations, and C-R functions, the general magnitude of 
risk estimates supports the potential for significant public health 
impacts in locations meeting the current primary PM2.5 
standards. This is particularly the case given that the large majority 
of PM2.5-associated deaths for air quality just meeting the 
current standards are estimated at annual average PM2.5 
concentrations from about 10 to 12 [mu]g/m\3\. These annual average 
PM2.5 concentrations fall within the range of long-term 
average concentrations over which key epidemiologic studies provide 
strong support for reported positive and statistically significant 
health effect associations (U.S. EPA, 2022b, section 3.6.2.2).
    In the 47 urban study areas, when air quality is simulated to just 
meet alternative standards, the PA notes that there are substantially 
larger risk reductions associated with lowering the annual standard 
than with lowering the 24-hour standard. Risks are estimated to 
decrease by 13-17% when air quality is adjusted to just meet an 
alternative annual standard with a level of 10.0 [mu]g/m\3\ or by 1-2% 
when adjusted to just meet an alternative 24-hour standard with a level 
of 30 [mu]g/m\3\ (U.S. EPA, 2022b, section 3.4.2.1). The percentage 
decrease when just meeting an alternative annual standard with a level 
of 10.0 [mu]g/m\3\ corresponds to approximately 7,400 fewer deaths per 
year (confidence intervals ranging from about 4,100 to 9,800) 
attributable to long-term PM2.5 exposures (U.S. EPA, 2022b, 
section 3.4.2.1).
    In the 30 study areas where just meeting the current and 
alternative standards is controlled by the annual standard, air quality 
adjusted to meet alternative annual standards with lower

[[Page 5616]]

levels is associated with reductions in estimated all-cause mortality 
risk. These reductions in risk for alternative annual levels are as 
follows: 7-9% reduction for an alternative annual level of 11.0 
[micro]g/m\3\, 15-19% reduction for a level of 10.0 [micro]g/m\3\, 22-
28% reduction for a level of 9.0 [micro]g/m\3\, and 30-37% reduction 
for a level of 8.0 [micro]g/m\3\ (U.S. EPA, 2022b, section 3.4.2.2). 
For each of these standards, most of the risk remaining is estimated at 
annual average PM2.5 concentrations that fall somewhat below 
the alternative standard levels (U.S. EPA, 2022b, section 3.4.2.2).
    In considering the at-risk analysis, the PA notes that across all 
simulated air quality for both the full set of 47 and the subset of 30 
study areas, Blacks experience the highest average PM2.5 
concentrations of the demographic groups analyzed. Native Americans 
experienced the lowest average PM2.5 concentrations, 
particularly in the full set of 47 study areas. White, Hispanic, and 
Asian populations were exposed to similar average PM2.5 
concentrations. Additionally, as the levels of potential alternative 
annual PM2.5 standards decrease, there is comparatively less 
disproportionate exposure between demographic populations (U.S. EPA, 
2022b, section 3.4.2.4).
    The PA recognizes that the risk estimates can provide additional 
information beyond the exposure information to inform our understanding 
of potentially disproportionate impacts, in this instance by including 
demographic-specific information on baseline incidence and the 
relationship between exposure and health effect. Across all air quality 
scenarios and demographic groups evaluated, Black populations in the 
study areas are associated with the largest PM2.5-
attributable mortality risk rate per 100,000 people, while White 
populations in the study areas are associated with the smallest 
PM2.5-attributative mortality risk rate (U.S. EPA, 2022b, 
section 3.4.2.4, Figure 3-20). Generally, as the levels of potential 
alternative annual PM2.5 standards decrease in the 30 areas 
controlled by the annual standard, the average reductions in 
PM2.5 concentration and mortality risk rates increase across 
all demographic populations (U.S. EPA, 2022b, section 3.4.2.4, Figure 
3-21).
    In comparing the reductions in average national PM2.5 
concentrations and risk rates within each demographic population, the 
average percent PM2.5 concentrations and risk reductions are 
slightly greater in the Black population than in the White population 
for each alternative standard evaluated (11.0 [micro]g/m\3\, 10.0 
[micro]g/m\3\, 9.0 [micro]g/m\3\, and 8.0 [micro]g/m\3\), when shifting 
from the current annual PM2.5 standard (12.0 [micro]g/m\3\) 
in the full set of 47 areas and the subset of 30 areas controlled by 
the annual standard. Furthermore, the difference in average percent 
risk reductions increases slightly more in Blacks than in Whites as the 
level of the potential alternative annual standard decreases (U.S. EPA, 
2022b, section 3.4.2.4, Table 3-19 and Table 3-20).
    The PA also recognizes that there are several particularly 
important uncertainties that affect the quantitative estimates of risk 
rates and exposure in the at-risk analysis and their interpretation in 
the context of considering the current primary PM2.5 
standards. These include uncertainties related to the modeling and 
adjustment methods for simulating air quality scenarios; the potential 
influence of confounders on the relationship between PM2.5 
exposure and mortality; and the interpretation of the shapes of C-R 
functions, particularly at lower concentrations. It is also important 
to recognize the limited availability of studies to inform the at-risk 
analysis. As noted in section II.C above and in section 3.4 of the PA, 
the at-risk analysis included C-R functions from 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. Of the studies available from the 2019 ISA, Di et al. 
(2017b) was identified as best characterizing potentially at-risk 
minority populations across the U.S.\98\ While the at-risk analyses 
provide additional insight on the estimated exposures and risks for 
certain demographic groups, it is not clear how the results would vary 
if: (1) analyses included populations that were younger than 65 years 
old, (2) the analyses were conducted areas that are demographically 
different than the 47 study areas included in this analysis, and (3) 
the air quality adjustments reflected source-specific emissions 
reduction strategies. Therefore, in light of the limitations and 
uncertainties associated with the at-risk analyses, the results should 
be considered within the context of the full risk assessment. The 
uncertainties associated with the quantitative risk assessment and at-
risk analyses are described in more detail in the PA (U.S. EPA, 2022b, 
section 3.4.2.5 and Appendix C) and are summarized in section II.C.2 
above.
---------------------------------------------------------------------------

    \98\ Additional details on concentration-response function 
identification can be found in Appendix C, section C.3.2 of the PA.
---------------------------------------------------------------------------

    In considering the public health implications of the risk 
assessment, the PA notes that the purpose for the study areas is to 
illustrate circumstances that may occur in areas that just meet the 
current or potential alternative standards, and not to estimate risk 
associated with conditions occurring in those specific locations 
currently. The PA notes that some areas across the U.S. have air 
quality for PM2.5 that is near or above the existing 
standards. Risks associated with air quality above the current 
standards are not informative to decisions about the adequacy of the 
current standards. This is because the risk assessment uses an approach 
to adjust air quality to just meet the current standards, which means 
that areas that have air quality that is above the current standards 
would be adjusted to just meet the current standards such that the 
evaluation of changes in risk and risk remaining would be associated 
with those areas meeting the current standards. The same is true for 
air quality adjusted to simulate just meeting alternative standard 
levels as well. Thus, the air quality and exposure circumstances 
assessed in the study areas in the risk assessment are specifically 
designed to inform whether the currently available information calls 
into question the adequacy of the public health protection afforded by 
the current standards, as well as to provide information regarding 
potential alternative standard levels.
    The risk estimates for the study areas assessed in this 
reconsideration reflect differences in exposure circumstances among 
those areas and illustrate the exposures and risks that might be 
expected to occur in other areas with such circumstances under air 
quality conditions that just meet the current standards or the 
alternative standards assessed. Thus, the exposure and risk estimates 
indicate the magnitude of exposure and risk that might be expected in 
many areas of the U.S. with PM2.5 concentrations at or near 
the current or alternative standards. Although the methodologies and 
data used to estimate risks in this reconsideration differ in several 
ways from what was used in the 2020 review, the findings and 
considerations summarized in the PA present a pattern of exposure and 
risk that is generally similar to that considered in the 2020 review, 
and indicate a level of protection generally consistent with that 
described in the 2020 PA.
    The PA notes that the considerations related to the potential 
public health implications of the risk assessment and at-risk analysis 
are important to informing the Administrator's proposed

[[Page 5617]]

decisions regarding the public health significance of the risk 
assessment results. Specifically, the PA notes that available evidence 
and information suggests that both long- and short-term 
PM2.5 exposures are associated with adverse health effects, 
including more severe effects such as mortality. In addition, the PA 
further notes that such effects impact large segments of the U.S. 
population, including those populations that may have other factors 
that influence risk (i.e., lifestage, pre-existing cardiovascular and 
respiratory diseases, race/ethnicity), as well as disparities in 
PM2.5 exposures and health risks based on race and ethnicity 
(U.S. EPA, 2022b, section 3.6.2.5). Therefore, the PA recognizes that 
the air quality allowed by the current primary PM2.5 
standards could be judged to be associated with significant public 
health risk. The PA also recognizes that such conclusions also depend 
in part on public health policy judgments that will weigh in the 
Administrator's decision in this reconsideration with regard to the 
adequacy of protection afforded by the current standards. Such 
judgments that are common to NAAQS decisions include those related to 
public health implications of effects of differing severity. Such 
judgments also include those concerning the public health significance 
of effects at exposures for which evidence is limited or lacking, such 
as effects at lower concentrations than those demonstrated in the key 
epidemiologic studies and in those population groups for which 
population-specific information, such as C-R functions, are not 
available from the epidemiologic literature.
3. Administrator's Proposed Conclusions on the Primary PM2.5 
Standards
    This section summarizes the Administrator's considerations and 
proposed conclusions related to the adequacy of the current primary 
PM2.5 standards and presents his proposed decision to revise 
the primary annual PM2.5 standard 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 will be 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 will be 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 will be informed by the Administrator's 
consideration of advice from the CASAC and public comments received on 
this proposal document.
a. Adequacy of the Current Primary PM2.5 Standards
    In considering 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, the Administrator's proposed 
conclusions regarding the adequacy of the current standards will depend 
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.
i. Proposed Conclusions on the Adequacy of the Current Primary 
PM2.5 Standards
    In reaching proposed conclusions on the adequacy of the current 
primary PM2.5 standards, the Administrator has considered 
the scientific evidence, including that assessed in the 2019 ISA and 
the ISA Supplement. The Administrator has also considered the 
quantitative estimates of risk developed in this reconsideration, 
including associated uncertainties and limitations, and the extent to 
which they indicate differing conclusions regarding the magnitude of 
risk, as well as level of protection from adverse effects, associated 
with the current standards. The Administrator has additionally 
considered the key aspects of the evidence and risk estimates 
emphasized in establishing the current standards, and the associated 
public health policy judgments and judgments about the uncertainties 
inherent in the scientific evidence and quantitative analyses that are 
integral to the proposed conclusions on the adequacy of the current 
primary PM2.5 standards.
    First, 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. 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. 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.
    In general, the Administrator recognizes that the annual standard 
is most effective at controlling exposures to ``typical'' daily 
PM2.5 concentrations that are experienced over the year, 
while the 24-hour standard, with its 98th percentile form, is most 
effective at limiting peak daily or 24-hour PM2.5 
concentrations. In considering the combined effects of these standards, 
the Administrator recognizes that changes in PM2.5 air 
quality designed 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. Additionally, changes 
designed to meet a lower 24-hour standard, with a 98th percentile form, 
would most effectively result in fewer and lower peak 24-hour 
PM2.5 concentrations, but also have an effect on lowering 
the annual average PM2.5 concentrations. Thus, the 
Administrator acknowledges the focus in evaluating

[[Page 5618]]

the current primary standards is on the protection provided by the 
combination of the annual and 24-hour standards against the 
distribution of both short- and long-term PM2.5 exposures.
    The Administrator recognizes the longstanding body of health 
evidence supporting relationships between PM2.5 exposures 
(short- and long-term) and mortality or serious morbidity effects. The 
evidence available in this reconsideration (i.e., that assessed in the 
2019 ISA (U.S. EPA, 2019a) and ISA Supplement (U.S. EPA, 2022a) and 
summarized above in section II.B.1 and section II.D.2.a reaffirms, and 
in some cases strengthens, the conclusions from the 2009 ISA regarding 
the health effects of PM2.5 exposures (U.S. EPA, 2009a). As 
noted above, epidemiologic studies 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). Recent 
experimental evidence, as well as evidence from 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. The Administrator also recognizes 
that the PA notes that while the full body of health effects evidence 
is considered in this reconsideration of the PM NAAQS, the greatest 
emphasis in the PA is placed 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.e., mortality, cardiovascular effects, respiratory effects, cancer, 
and nervous system effects). In considering the available scientific 
evidence, 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 the public health protection afforded by the current 
primary PM2.5 standards 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 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 (U.S. EPA, 2022b, section 3.6.3.2.1). In its advice 
on the adequacy of the current primary PM2.5 standards, the 
CASAC reached consensus that the PM2.5 mass-based indicator 
should be retained, without revision (Sheppard, 2022a, p. 2 of 
consensus letter). Thus, as in the 2020 review (85 FR 82715, December 
18, 2020) and consistent with the advice from the CASAC, the 
Administrator proposes to conclude that it is appropriate to consider 
retaining PM2.5 mass as the indicator for the primary 
standards for fine particles.
    With respect to averaging time and form, the Administrator notes 
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) (U.S. EPA, 2022b, section 3.6.3.2.2). In this 
reconsideration, the epidemiologic and controlled human exposure 
studies have examined a variety of PM2.5 exposure durations. 
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 the EPA notes that 
associations with sub-daily estimates are less consistent and, in some 
cases, smaller in magnitude (U.S. EPA, 2019a, section 1.5.2.1; U.S. 
EPA, 2022b, section 3.6.3.2.2). In addition, controlled human exposure 
and panel-based studies of sub-daily 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). 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). 
Additionally, 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, 2022b, section 3.3.3.1). Therefore, these 
studies do not provide support for additional protection against sub-
daily PM2.5 exposures, beyond that provided by the current 
primary standards. In its advice on the adequacy of the current primary 
PM2.5 standards, the CASAC reached consensus that averaging 
times for the standards should be retained, without revision (Sheppard, 
2022a, p. 2 of consensus letter). Thus, as in the 2020 review (85 FR 
82715, December 18, 2020), and consistent with the advice from the 
CASAC, the Administrator reaches the proposed conclusion 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.
    With regard to form, the Administrator proposes to conclude that it 
is appropriate to consider retaining the current form of both the 
annual and the 24-hour standards. In so doing, he first notes that, in 
the 1997 review, the EPA set both an annual standard, to provide 
protection from health effects associated with both long- and short-
term exposures to PM2.5, and a 24-hour standard to a 
supplement the protection afforded by the annual standard (62 FR 38667, 
July 18, 1997). 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 advice on the adequacy of the current standards, 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). In relation to 
the form of the 24-hour standard (98th percentile, averaged over three 
years), the Administrator notes that epidemiologic studies continue to 
provide strong support for health effect associations with short-term 
(e.g., mostly 24-hour) PM2.5 exposures (U.S.

[[Page 5619]]

EPA, 2022b, section 3.6.3.2.3) and that controlled human exposure 
studies provide evidence for health effects following single short-term 
``peak'' PM2.5 exposures. Thus, 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. The Administrator further notes that this 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). While the CASAC provided recommendations regarding the 
adequacy of the current 24-hour standard conditional on the current 
form (i.e., 98th percentile, averaged over three years), they 
recommended that in future reviews, the EPA also consider alternative 
forms for the primary 24-hour PM2.5 standard (Sheppard, 
2022a, p. 18 of consensus responses). 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. Thus, in considering 
the information summarized above, and consistent with the advice from 
the CASAC, the Administrator reaches the preliminary conclusion that it 
is appropriate to consider retaining the forms of the current annual 
and 24-hour PM2.5 standards. The Administrator solicits 
public comment on the proposed decision to retain the current form 
(98th percentile, averaged over three years) of the primary 24-hour 
PM2.5 standard. The Administrator acknowledges that the 
CASAC recommended retaining the current form at this time but also 
recommended that the EPA consider alternatives to the current form in 
future reviews. The EPA agrees that it would be appropriate to gather 
additional air quality and scientific information and further consider 
these issues in future reviews. This information will not be utilized 
for this reconsideration process.
    With regard to the level of the current standards, the 
Administrator first considers the scientific evidence evaluated in the 
2019 ISA and ISA Supplement, and considerations regarding the evidence 
as presented in the PA. The Administrator recognizes that the PA places 
greater weight on epidemiologic studies conducted in the U.S. and 
Canada, as these studies are more directly applicable for quantitative 
considerations compared to studies conducted in other countries. 
Studies conducted in other countries outside of the U.S. and Canada 
generally reflect different populations, exposure characteristics, air 
pollution mixtures, and higher PM2.5 concentrations in 
ambient air than are currently found in the U.S. Therefore, consistent 
with approaches in previous reviews, the Administrator judges that it 
is appropriate to place greater weight on the U.S. and Canadian 
epidemiologic studies in reaching conclusions regarding the adequacy of 
the current standards. In so doing, the Administrator notes that the 
epidemiologic studies in the U.S. and Canada report health effect 
associations with mortality and/or morbidity across multiple 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, 
populations with pre-existing cardiovascular and respiratory disease, 
minority populations, and low SES communities). Further, he notes the 
epidemiologic studies that use a variety of statistical designs and 
employ a variety of methods to examine exposure measurement error as 
well as to control for confounding effects, and he acknowledges that 
results of these analyses support the robustness of the reported 
associations. Additionally, the Administrator notes findings from an 
expanded body of studies that employ alternative methods for confounder 
control and accountability methods further inform the causal nature of 
the relationship between long or short-term term PM2.5 
exposure and mortality as described in the 2019 ISA and ISA Supplement 
(U.S. EPA, 2019, sections 11.1.2.1, 11.2.2.4; U.S. EPA, 2022a, sections 
3.1.1.3, 3.1.2.3, 3.2.1.3, and 3.2.2.3). These studies, summarized 
above in II.B.3 above and in Table 3-11 and Table 3-12 of the PA (U.S. 
EPA, 2022b) examine both short- and long-term PM2.5 exposure 
and cardiovascular effects and mortality, and, using a variety of 
statistical methods to control for confounding bias, consistently 
report positive associations, which further supports the broader body 
of epidemiologic evidence for both cardiovascular effects and 
mortality. Moreover, the Administrator 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. 
Lastly, the Administrator notes that studies that examine the shape of 
the C-R relationship over the full distribution of ambient 
PM2.5 concentrations have not identified a threshold 
concentration, below which associations no longer exist (U.S. EPA, 
2019a, section 1.5.3; U.S. EPA, 2022a, sections 2.1.1.5.1 and 
2.1.1.5.2). However, the Administrator also notes that 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 (section II.B.4 above; U.S. EPA, 2019a, 
section 11.2.4; U.S. EPA, 2022a, section 2.2.3.2).
    In considering the available scientific evidence to inform proposed 
decisions on the adequacy of the current level of the annual standard, 
the Administrator acknowledges that the evidence available in this 
reconsideration provides support for adverse health effect associations 
at lower ambient PM2.5 concentrations than in previous 
reviews. The Administrator notes that in previous reviews (including 
1997, 2006 and 2012 reviews), evidence-based approaches 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 are supported in 
this reconsideration by the current CASAC, who also referenced the 
potential for considering other lines of epidemiologic evidence.\99\ 
The Administrator 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 [mu]g/m\3\ (i.e., Figure 1 and 
Figure 2 above with concentrations ranging down as low as 9.9 [mu]g/
m\3\ in U.S.-based monitor-based studies and 9.3 [mu]g/m\3\ in U.S.-
based hybrid model-based studies). The Administrator also recognizes 
that, while Canadian studies can be more difficult to directly compare 
to the annual design value used to determine in compliance in the U.S., 
the overall mean PM2.5 concentrations from the key Canadian 
epidemiologic studies are

[[Page 5620]]

close to, though somewhat lower than, those from the U.S. studies. The 
range of monitor-based mean PM2.5 concentrations is from 6.9 
[mu]g/m\3\ to 13.3 [mu]g/m\3\ while the range of mean PM2.5 
concentrations in studies that use hybrid modeling is 5.9 [mu]g/m\3\ to 
9.8 [mu]g/m\3\.
---------------------------------------------------------------------------

    \99\ The Administrator notes that some members of the CASAC 
advised that ``for the purpose of informing the adequacy of the 
standards'' (Sheppard, 2022a, p. 8 of consensus responses) that the 
EPA in future reviews 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.
---------------------------------------------------------------------------

    In assessing the adequacy of the current annual standard, the 
Administrator also examines additional epidemiologic studies, 
consistent with CASAC advice, that provide supplementary information 
for consideration in reaching conclusions regarding the current annual 
standard. These studies include analyses that restrict annual average 
PM2.5 concentrations to values below level the annual 
standard (described above in section II.B.3.b and in Table 3-10 of the 
PA) and the CASAC advised that ``for the purpose of informing the 
adequacy of the standards'' that the EPA evaluate the means from these 
studies. In this reconsideration, there are two key studies available 
that restrict average annual PM2.5 concentrations to less 
than 12 [mu]g/m\3\ (Di et al., 2017a, and Dominici et al., 2019). These 
restricted analyses report positive and statistically significant 
associations with all-cause mortality and report mean PM2.5 
concentrations of 9.6 [mu]g/m\3\. Thus, these two epidemiologic studies 
provide support for positive and statistically significant associations 
at lower mean PM2.5 concentrations. The Administrator does 
note that uncertainties exist in these analyses (described in more 
detail in sections II.B.3.b and II.D.2.a above), including uncertainty 
in how studies exclude concentrations (e.g., at what spatial resolution 
are concentrations being excluded), which would make any comparisons of 
concentrations in restricted analyses difficult to compare directly to 
design values.
    In considering the available key U.S. epidemiologic studies, the 
Administrator also notes that CASAC recommended looking at the 
distribution of concentrations reported in epidemiologic studies for 
purposes of informing the adequacy of the standards and notes that a 
small number of studies report PM2.5 concentrations 
corresponding to the 25th and 10th percentiles of health data or 
exposure estimates. He observes that 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 [mu]g/m\3\ and 10th 
percentiles of health events correspond to PM2.5 
concentrations at or above 9.8 [mu]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). The Administrator further observes that of the key U.S. 
epidemiologic studies that use hybrid modeling approaches to estimate 
long-term PM2.5 exposures, the ambient PM2.5 
concentrations corresponding to 25th percentiles of estimated exposures 
are 9.1 [mu]g/m\3\ (Figure 2 above and U.S. EPA, 2022b, Figure 3-14). 
In key U.S. epidemiologic studies that use hybrid modeling approaches 
to estimate short-term PM2.5 exposures, the ambient 
concentrations corresponding to 25th percentiles of estimated 
exposures, or health events, are 6.7 [mu]g/m\3\ and the ambient 
PM2.5 concentration corresponding to that 10th percentile 
range from 4.7 [mu]g/m\3\ to 7.3 [mu]g/m\3\ (Figure 2 above and U.S. 
EPA, 2022b, Figure 3-14). While the Administrator places less weight on 
the limited number of studies that report these lower quartiles of the 
air quality distributions, he notes these concentrations are generally 
below the level of the annual standard of 12 [mu]g/m\3\.
    In further assessing the adequacy of the current annual standard, 
the Administrator also evaluates what the accountability studies may 
indicate with respect to potential for improvements in public health 
with improvements in air quality. In so doing, he takes note of three 
accountability studies (Sanders et al., 2020b; Corrigan et al., 2018; 
and Henneman et al., 2019a) newly available in this reconsideration 
with starting concentrations at or below 12.0 [mu]g/m\3\ that indicate 
positive and significant associations with mortality and morbidity and 
reductions in ambient PM2.5 (described above in section 
II.B.3.b and in Table 3-12 of the PA) and notes that these studies 
suggest public health improvements may occur at concentrations below 12 
[mu]g/m\3\.
    Thus, in considering the available scientific evidence to inform 
proposed decisions on the adequacy of the current primary annual 
PM2.5 standard, the Administrator recognizes that there is a 
long-standing body of epidemiologic evidence that provides support for 
associations between PM2.5 exposures and health effects 
across a distribution of air quality that includes concentrations near 
(i.e., at, above, and below) the current standards. As such, the 
Administrator recognizes that the available scientific evidence, as 
assessed in the 2019 ISA and ISA Supplement, including the newly 
available epidemiologic studies and the supplemental information from 
specific types of epidemiologic studies, provides a strong scientific 
foundation for consideration of the adequacy of the level of the 
current annual standard.
    In considering the available scientific evidence to inform proposed 
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. When looking to the experimental studies, he notes 
that controlled human exposure studies provide evidence for health 
effects following single, short-term exposures to PM2.5 
concentrations that are greater than those typically present in ambient 
air. In the controlled human exposure studies, the Administrator 
observes that results are inconsistent, particularly at lower 
PM2.5 concentrations, but that 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). 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. However, as noted in the PA, 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 which may provide support that 
more adverse effects may be experienced following longer exposure 
durations and/or exposure to higher concentrations but such 
intermediate effects typically would not, by themselves, be judged as 
adverse. Additionally, he acknowledges, as noted by the 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. So, noting these points and 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 examines the air quality 
analyses, described in more detail in section II.B.3.a above, to assess 
whether during recent air quality conditions, areas meeting the current

[[Page 5621]]

standards would experience the concentrations reported in these 
controlled human exposure studies. He observes that these 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, thus suggesting 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 and the annual standard), the 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 monitoring sites violating the 
current standards, but generally remain below 16 [mu]g/m\3\ and rarely 
exceed 80 [mu]g/m\3\. Based on this information, the Administrator 
finds that the current suite of standards maintains sub-daily 
concentrations far below the current 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.
    In addition, the Administrator also notes that the majority of the 
CASAC provide support for their advice to revise the current daily 
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). In 
considering this advice from the majority of the CASAC, the 
Administrator notes that the substantial epidemiologic evidence 
available in this reconsideration, including the studies that restrict 
short-term (24-hour average PM2.5 concentrations) 
PM2.5 exposures below 25 [mu]g/m\3\, provides support for 
positive and statistically significant associations between exposure to 
short-term PM2.5 concentrations and all-cause mortality (Di 
et al., 2017a) and CVD hospital admissions (deSouza et al., 2021, and 
Di et al., 2017a). In particular, for the available epidemiologic 
studies that employ restricted analyses of short-term exposure studies, 
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), below 30 [mu]g/m\3\ 
(Shi et al., 2016), and below 25 [mu]g/m\3\ (Di et al., 2017a). Thus, 
the Administrator agrees that these studies help to provide additional 
support for reaching conclusions on causality in the 2019 ISA. 
Additionally, when considering these studies, the restricted approach 
in these short-term studies most clearly indicates that risks 
associated with short-term PM2.5 exposures are not 
disproportionately driven by the peaks of the air quality distribution. 
While this is useful information, it does not help to inform questions 
on the adequacy of the current 24-hour standard given that the 24-hour 
standard focuses on reducing ``peak'' exposures (with its 98th 
percentile form). In further evaluating these studies, the 
Administrator notes that the fact that there are positive and 
significant associations in these analyses does not mean that one can 
conclude that there would be short-term effects occurring in areas that 
meet a 24-hour standard at these levels. This is true for multiple 
reasons. First, there are uncertainties with respect to the 
methodologies used in these studies to exclude concentrations and the 
specific methodology used (e.g., are individual days with 
concentrations above the concentration of interest in the restricted 
analyses excluded at the modeled grid cell level or the ZIP code level 
rather than removing entire areas with day(s) that exceed that 
concentration) has direct implications for the resulting air quality 
scenario(s). This in turn affects how the adjusted air quality 
scenarios in these studies can be related to air quality distributions 
and exposures to PM2.5 concentrations in ambient air and 
thus how the data can be interpreted with regard to the current 
standard level. Second, given that these studies are only evaluating 
daily or annual average PM2.5 concentrations that would 
correspond to the levels of the standards, they do not consider these 
levels along with the forms and averaging times of the standards. This 
is quite limiting for use in judging the adequacy of the 24-hour 
standard given that the study-reported mean concentration is not useful 
in informing the level of a standard with a 98th percentile form that 
is designed to limit exposures to peak PM2.5 concentrations. 
Further, as noted in the PA, the study-reported means from these 
studies, are not useful in identifying a level at which we can say with 
some confidence that effects are occurring due to impacts from ``peak'' 
exposures (i.e., those most closely aligned with the protection provide 
by the 24-hour standard, with its 98th percentile form) but are instead 
more useful in informing questions about impacts from ``typical'' or 
average 24-hour exposures (i.e., those most closely aligned with the 
protection provided by the annual standard). These uncertainties and 
lack of information available from these studies are quite limiting and 
as such, the Administrator concludes that it is unclear how to apply 
these studies to a decision framework that could inform whether the 
level of the current 24-hour standard is or is not adequate. However, 
the Administrator notes this uncertainty may not be quite as limiting 
for using restricted analyses studies to inform conclusions regarding 
the adequacy of the annual standard, given that the study-reported 
means could be evaluated in the context of the decision framework 
described above for informing proposed decisions on the level of the 
annual standard. However, in considering the available evidence with 
regard to the current 24-hour PM2.5 standard, while the 
Administrator agrees with the majority of the CASAC's comment that the 
controlled human exposure studies have significant limitations which 
must be considered when reaching conclusions on the adequacy of the 
current 24-hour standard, he finds that restricted analyses studies 
have significant limitations and do not provide a stronger line of 
evidence with which to inform his proposed decisions on the current 24-
hour standard.
    In addition to the evidence above, the Administrator also considers 
what the risk assessment indicates with regard to the adequacy of the 
current primary annual and 24-hour PM2.5 standards. These 
analyses provide estimates of PM2.5-attributable mortality 
which are estimated based on 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. The Administrator 
observes that the risk assessment estimates that the current primary 
annual PM2.5 standard could allow a substantial number of 
deaths in the U.S. For example, when air quality in 30 study areas is 
adjusted to simulate just meeting the current annual standard, the risk 
assessment estimates long-term PM2.5 exposures to be 
associated with as many as 39,000 total deaths, with confidence 
intervals ranging from 26,000-51,000. The Administrator notes that 
these estimates do not reflect uncertainties in associations of health 
effects at lower

[[Page 5622]]

concentrations and simulated air quality improvements will always lead 
to proportional decreases in risk (i.e., each additional [mu]g/m\3\ 
reduction produces additional benefits with no clear stopping point). 
Noting these limitations and noting that the absolute numbers of 
estimated deaths vary across exposure durations, populations, and C-R 
functions, he also observes that the general magnitude of risk 
estimates supports the potential for significant public health impacts 
in locations meeting the current primary annual PM2.5 
standard. He observes that this is particularly the case given that the 
large majority of PM2.5-associated deaths for air quality 
just meeting the current annual standard are estimated at annual 
average PM2.5 concentrations from about 10 to 12 [mu]g/m\3\, 
annual average PM2.5 concentrations that fall well within 
the range of long-term average concentrations over which key 
epidemiologic studies provide strong support for reported positive and 
statistically significant PM2.5 health effect associations. 
With respect to the CASAC's advice on the risk assessment, the 
Administrator notes that the majority of the CASAC agreed that ``[t]he 
results support the conclusion that the current primary annual 
PM2.5 standard does not adequately protect public health'' 
(Sheppard, 2022a, p. 2 of consensus letter) and that ``[t]he CASAC 
concurs with the EPA's assessment that meaningful risk reductions will 
result from lowering the annual PM2.5 standard'' (Sheppard, 
2022a, p. 3 of consensus letter). Additionally, the minority of CASAC 
also agreed that the risk assessment results support revision to the 
annual standard but commented that there were important uncertainties 
in the analyses and interpretation of the analyses for annual standard 
levels below 10 [mu]g/m\3\ (Sheppard, 2022a, p. 3 of consensus letter).
    The Administrator also recognizes that the risk assessment was able 
to include a new analysis based on the availability of a new study in 
this reconsideration that provided mortality risk coefficients for 
older adults (i.e., 65 years and older) based on PM2.5 
exposure and stratified by racial and ethnic demographics. This at-risk 
analysis provided estimates of potential long-term PM2.5-
attributable exposure and mortality risk in older adults, stratified by 
racial/ethnic demographics, when meeting a revised annual standard with 
a lower level. The Administrator recognizes that this analysis is 
subject to the same uncertainties as those associated with the main 
risk assessment estimates, including being limited to a subset of areas 
across the U.S. and influenced by air quality adjustment methodologies 
that may not produce estimates of PM2.5 concentration 
exposures that match those that can result from control strategies 
implemented to meet more stringent standards, and that the results are 
based on the risk coefficients of only one epidemiologic study. Taking 
into account these uncertainties and limitations, he does judge that 
the analysis supports that a lower annual standard level (i.e., below 
12 [mu]g/m\3\ and down as low as 8 [mu]g/m\3\) will help to reduce 
PM2.5 exposure and may also help to mitigate risk 
disparities. The Administrator notes that what urban areas are included 
in the risk assessment analysis will greatly influence the results but 
notes that based on the areas included in the analyses, the results 
show the largest impact is on reducing exposure and risk in Black 
populations, who were estimated in the risk assessment case study areas 
to have the highest levels of exposures and the greatest rates of 
premature mortality risk.
    With respect to the 24-hour standard, the risk assessment indicates 
that the annual standard is the controlling standard across most of the 
urban study areas evaluated. 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, 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 CASAC advice, 
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 commented that they 
placed greater weight on the evidence-based considerations 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). The majority of the CASAC members further state 
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 majority of the CASAC members further state 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).
    In considering the application of the risk assessment in a decision 
framework assessing the adequacy of the current 24-hour 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, and that this 
quantitative approach does not incorporate any elements of uncertainty 
in associations of health effects at lower concentrations and simulated 
air quality improvements will always lead to proportional decreases in 
risk (i.e., each additional [mu]g/m\3\ reduction produces additional 
benefits with no clear stopping point). Therefore, 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 also notes that 
in the U.S., current air quality shows that the 24-hour standard is 
controlling in very few areas and thus, it is understandable that there 
are very few areas that would be included in the study areas in the 
risk assessment. The Administrator also recognizes that the risk 
assessment did not provide quantitative information on risk impacts 
associated with an alternative 24-hour standard level of 25 [mu]g/m\3\.
    Based on the above considerations, the Administrator reaches the 
proposed conclusion that the available scientific evidence (summarized 
above in section II.B) and quantitative risk assessment

[[Page 5623]]

(summarized above in section II.C), can reasonably be viewed as calling 
into question the adequacy of the public health protection afforded by 
the current annual standard. In reaching this conclusion, the 
Administrator places weight on the extensive epidemiologic evidence 
available in this reconsideration, strengthened from previous reviews, 
showing associations between adverse health effects (particularly 
cardiovascular effects and mortality) and long-term mean 
PM2.5 concentrations, and notes the number and strength of 
studies available showing associations with mean PM2.5 
concentrations well below the current annual standard of 12.0 [mu]g/
m\3\. The Administrator also takes note of the evidence supporting the 
biological plausibility of these associations, including toxicological 
studies and controlled human exposure studies. When turning to 
additional information from the epidemiologic evidence base, he notes 
the advice from CASAC to also consider the 25th percentile of the data 
that is available and the study reported means from long-term studies 
that restrict concentrations to below 12 [mu]g/m\3\. When considering 
the 25th percentile of the data, the Administrator notes that it is 
available from a limited number of epidemiologic studies and that the 
current level of the annual standard is above most of the 25th 
percentile values reported in the key epidemiologic studies. When 
looking to the restricted analyses studies, he notes that there are two 
studies that report positive and statistically significant associations 
with all-cause mortality, and report a study mean PM2.5 
concentration of 9.6 [mu]g/m\3\. While noting the limited nature of 
these two lines of evidence and the associated uncertainties, the 
Administrator does judge that these data support the need to revise the 
annual standard level. Lastly, with respect to the epidemiologic 
evidence, the Administrator also takes into account accountability 
studies newly available in this reconsideration with starting 
concentrations at or below 12.0 [mu]g/m\3\ that indicate positive and 
significant associations with mortality and morbidity and reductions in 
ambient PM2.5 and notes that these studies suggest public 
health improvements may occur at concentrations below 12 [mu]g/m\3\.
    The Administrator also considers the results of the risk assessment 
in light of the information it provides on risks associated with the 
current and more stringent levels of the annual standard. While he 
recognizes a number of uncertainties and limitations associated with 
the quantitative estimates of the risk assessment, he judges that the 
estimated risks remaining under air quality adjusted to just meet the 
current suite of standards are too high to be considered requisite to 
protect public health with an adequate margin of safety, noting in 
particular the large number of premature deaths estimated to remain 
with air quality that just meets the current annual standard. The 
Administrator also recognizes that the risk assessment was able to 
include a new analysis (at-risk analysis) that provided estimates of 
potential long-term PM2.5-attributable exposure and 
mortality risk in older adults, stratified by racial/ethnic 
demographics, when meeting a revised annual standard with a lower 
level. While the Administrator recognizes that this analysis is subject 
to multiple uncertainties and limitations (as noted above in sections 
II.C.2 and II.D.2.b), he does judge that the analysis suggests that a 
lower annual standard level (i.e., below 12 [mu]g/m\3\ and down as low 
as 8 [mu]g/m\3\) will help to reduce PM2.5 exposure and may 
also help to mitigate exposure and risk disparities. Finally, the 
Administrator considers the advice from the CASAC, who unanimously 
recommended revising the annual standard.
    The Administrator finds it is less clear whether the available 
scientific evidence and quantitative information call into question the 
adequacy of the public health protection afforded by the current 24-
hour standard, particularly when considered in conjunction with the 
protection provided by the suite of standards and the proposed decision 
to revise the annual standard. In considering the scientific evidence, 
he notes that the controlled human exposure studies do not provide a 
threshold below which no effects occur and they do not include the most 
at-risk populations. However, the concentrations reported in these 
studies are for observed effects that signal a change in the body 
likely due to short-term exposure to PM2.5 and which may be 
the prelude to more adverse effects following longer duration and/or 
higher concentration exposures but typically would not, by themselves, 
be judged as adverse. Balancing this with the observation that the air 
quality concentrations in areas meeting the current standards are well 
below the PM2.5 concentrations shown to elicit effects in 
these studies, the Administrator does not judge that these studies call 
into question the adequacy of the current 24-hour standard. With 
respect to the epidemiologic evidence, the Administrator notes that the 
body of epidemiologic evidence provides limited support for judging 
adequacy of the level of the 24-hour standard. As discussed in detail 
above (section II.B.3.b), epidemiologic studies provide the strongest 
support for reported health effect associations for the part of the air 
quality distribution corresponding to the bulk of the underlying data 
(i.e., estimated exposures and/or health events), often around the 
overall mean concentrations evaluated rather than near the upper end of 
the distribution. While there are three studies available in this 
reconsideration that restricted 24-hour concentrations to 
concentrations below 25 [mu]g/m\3\ and while some members of CASAC 
pointed to these studies as the basis for their recommendation to 
revise the 24-hour standard, the Administrator preliminarily concludes 
that the results from these studies, particularly in light of the 
uncertainties associated with these studies (as discussed above), are 
an inadequate basis for revising the level of the 24-hour 
PM2.5 standard.
    When evaluating the risk assessment information, the Administrator 
notes that the risk assessment estimates a reduction of 9-13% 
PM2.5 attributable mortality in areas where the 24-hour 
standard is controlling when the 24-hour PM2.5 standard is 
reduced from a level of 35 [mu]g/m\3\ to 30 [mu]g/m\3\. The 
Administrator notes that this estimated reduction in PM2.5-
associated risks is across a more limited population and is largely 
confined to a small number of areas located in the western U.S. Other 
areas included in the risk assessment were shown to experience risk 
reductions that were driven primarily by meeting a lower annual 
standard level (though the associated change in air quality also 
resulted in lower 24-hour standard concentrations). With respect to 
CASAC advice, the Administrator notes that the majority of CASAC 
advised that less weight be placed, while the minority of CASAC advised 
that these risk assessment results support the conclusion that the 
current 24-hour standard is adequate (Sheppard, 2022a, p. 4 of 
consensus letter), the majority of CASAC advised that less weight be 
placed on the risk assessment results and noted 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).
    Based on the current evidence and quantitative information, as well 
as consideration of CASAC advice and

[[Page 5624]]

public comment thus far in this reconsideration, the Administrator 
proposes to conclude that the current primary PM2.5 
standards are not adequate to protect public health with an adequate 
margin of safety. While he notes that the scientific evidence and 
quantitative information clearly call into question the adequacy of the 
public health protection afforded by the current annual standard, the 
Administrator finds it is less clear whether the available scientific 
evidence and quantitative information calls into question the adequacy 
of the public health protection afforded by the current 24-hour 
standard. In considering how to revise the suite of standards to 
provide the requisite degree of protection, 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. He also 
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. 
As noted above, the annual standard is targeted at controlling the 
typical exposures for which the evidence of adverse health effects is 
strongest. The Administrator places the most weight on the large number 
and strength of epidemiologic studies that report positive, and often 
statistically significant, associations with long-term mean reported 
PM2.5 concentrations well below the current level of the 
annual standard of 12.0 [mu]g/m\3\, as well as corroborating evidence 
from U.S. accountability studies with starting concentrations below 12 
[mu]g/m\3\ and studies that found positive and statistically 
significant associations in analyses restricted to concentrations less 
than 12 [mu]g/m\3\. In considering the risk assessment information, he 
notes that, for most of the U.S., the annual standard is the 
controlling standard and that the risk assessment estimates reductions 
in PM2.5-associated risks across more of the population and 
in more areas with alternative annual standard levels compared to 
estimates for alternative 24-hour standard levels. Moreover, the 
Administrator notes that a more stringent annual standard has been 
shown to effectively reduce both average (annual) concentrations and 
peak (daily) concentrations, ensuring the broadest protection of public 
health. Finally, the Administrator notes that the CASAC was unanimous 
in its advice regarding the need to revise the annual standard, 
although they did not reach consensus on what range of alternative 
levels would be most appropriate to consider. Thus, in considering how 
to revise the suite of standards to provide the requisite degree of 
protection, the Administrator proposes to conclude it is appropriate to 
focus on revising the annual standard.
b. Consideration of Alternative Primary Annual PM2.5 
Standard Levels
    This section summarizes the Administrator's conclusions and 
proposed decisions related to the current primary annual 
PM2.5 standard and presents his proposed decision to revise 
the level of the current annual standard within the range of 9.0 to 
10.0 [mu]g/m\3\, in conjunction with retaining the current indicator, 
averaging time, and form of that standard. The EPA is also soliciting 
public comment on alternative annual standard levels down to 8.0 [mu]g/
m\3\ and up to 11.0 [mu]g/m\3\, on an alternative 24-hour standard 
level as low as 25 [mu]g/m\3\ and on the combination of annual and 24-
hour standards that commenters may believe is appropriate, along with 
the approaches and rationales used to support such levels.
    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.
    Having reached the conclusion that the current indicator, averaging 
time, and form of the standard are appropriate for the reasons outlined 
above, the Administrator next considers the range of potential 
alternative standard levels that could be reasonably supported by the 
available scientific evidence and risk-based information to increase 
public health protection against short-term and long-term 
PM2.5 exposures. The evidence available in this 
reconsideration regarding PM2.5 exposures associated with 
health effects affirms and strengthens the evidence available at the 
completion of the 2009 ISA, taking into account studies evaluated in 
the 2019 ISA and ISA Supplement. The Administrator recognizes that the 
weight of evidence is strongest for health effects for which the 2019 
ISA concludes that the evidence provides support for a causal 
relationship between PM2.5 exposures and health effects, 
including those between long- and short-term PM2.5 exposures 
and mortality and cardiovascular effects. He recognizes that the weight 
of evidence is also strong for health effects for which the 2019 ISA 
concludes that the evidence supports a likely to be causal 
relationship, which include long- and short-term PM2.5 
exposures and respiratory effects and long-term PM2.5 
exposures and cancer, and nervous system effects.
    In considering the available scientific evidence that could inform 
conclusions regarding potential alternative levels of the annual 
PM2.5 standard, the Administrator notes that in past 
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 (i.e., causal or likely to 
be causal determinations). In so doing, the Administrator recognizes 
that the number of epidemiologic studies has expanded since the 
completion of the 2009 ISA and the epidemiologic studies evaluated in 
the 2019 ISA and the ISA Supplement continue to report positive and 
statistically significant associations between long- and short-term 
exposure to PM2.5 and mortality and morbidity.
    Additionally, the Administrator recognizes that the available 
epidemiologic studies enable the examination of the entire population 
and include, and even focus on, those that may be at comparatively 
higher risk of experiencing a PM2.5-related health effects. 
The Administrator notes that the 2019 ISA found that 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 respiratory disease), and SES, and that the 
ISA Supplement noted new evidence that further supported racial and 
ethnic differences in PM2.5 exposures and PM2.5-
related health risks. The Administrator also observes that at-risk 
populations make up a substantial portion of the U.S. population 
(section II.B.2 above), including children (22%) and older adults 
(16%), as well as non-Hispanic Black (12%) and Hispanic populations 
(18%) and that the prevalence of pre-existing diseases varies by 
lifestage and

[[Page 5625]]

race/ethnicity. The Administrator notes that the cohorts examined in 
the epidemiologic studies available in this reconsideration include 
diverse populations that are broadly representative of the U.S. 
population as a whole, and include those populations identified as at-
risk (i.e., children and older adults), as well as individuals in the 
general population with pre-existing disease, such as cardiovascular 
disease and respiratory disease.
    Recent epidemiologic studies also strengthen support for health 
effect associations at lower ambient PM2.5 concentrations 
than previous reviews and studies that examine the shapes of C-R 
functions over the full distribution of ambient PM2.5 
concentrations have not identified a threshold concentration, below 
which associations no longer exist (U.S. EPA, 2019a, section 1.5.3; 
U.S. EPA, 2022a, sections 2.2.3.1 and 2.2.3.2). Though these analyses 
are complicated by the relatively sparse data available at the lower 
end of the air quality distribution (U.S. EPA, 2019a, section 1.5.3), 
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\, though uncertainties 
remain about the shape of the C-R curve at PM2.5 
concentrations < 8 [mu]g/m\3\.
    With respect to uncertainties in epidemiologic studies, a broad 
range of approaches have been adopted across studies to examine 
confounding and the results of those examinations support the 
robustness of reported associations. Additionally, there is a 
considerable amount of new epidemiologic evidence in this 
reconsideration, including a large number of new epidemiologic studies 
that use varying study designs that reduce uncertainties, including 
studies that employ alternative methods for confounder control and 
support associations between exposure and adverse health effects at 
lower PM2.5 concentrations. Consistent findings from the 
broad body of epidemiologic studies are supported by studies employing 
alternative methods for confounder control, which used a variety of 
statistical methods to control for confounding bias and consistently 
report positive associations. The results of these studies support the 
positive and significant effects seen in cohort studies associated with 
short- and long-term exposure to PM2.5 and mortality. 
Moreover, epidemiologic studies continue to evaluate the uncertainty 
related to exposure measurement error, and while none of these 
approaches eliminates the potential for exposure error in epidemiologic 
studies, the consistent reporting of PM2.5 health effect 
associations across exposure estimation approaches, even in the face of 
exposure error, together with the larger effect estimates reported in 
some studies that have attempted to reduce exposure error, provides 
further support for the robustness of associations between 
PM2.5 exposures and mortality and morbidity. Therefore, 
given the strength of the available epidemiologic evidence, including 
the ability of these studies to provide information about impacts on 
the most at-risk populations, the Administrator concludes that the 
strongest available evidence for evaluating alternative levels of the 
annual standard continues to be the epidemiologic studies.
    The evidence base available in this reconsideration also consists 
of experimental studies that include controlled human exposure studies 
and animal toxicological studies. These studies demonstrate health 
outcomes following long-term and short-term exposure to 
PM2.5 at exposures that are well-above those typically found 
in ambient air. This body of evidence provides support for the 
biological mechanisms and the plausibility of the serious health 
effects associated with ambient PM2.5 exposures in 
epidemiologic studies. Thus, the Administrator recognizes that while 
experimental studies may not be as useful in a decision-making 
framework alone, results from these studies lend further support to the 
use of the epidemiologic evidence base in informing the level of the 
annual standard.
    In considering the level of the annual standard, the Administrator 
recognizes that the annual standard, with its form based on the 
arithmetic mean concentration, is most appropriately meant to limit the 
``typical'' daily and annual exposures that are most strongly 
associated with the health effects observed in epidemiologic studies. 
However, the Administrator also recognizes that while epidemiologic 
studies examine associations between distributions of PM2.5 
air quality and health outcomes, they do not identify particular 
PM2.5 exposures that cause effects. Thus, any approach that 
uses epidemiologic information in reaching decisions on what standards 
are appropriate necessarily requires judgments of the Administrator 
about how to consider the information available from the epidemiologic 
studies as a basis for appropriate standards. This includes 
consideration of how to weigh 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 CAA.
    Thus, in recognizing the need to weigh these uncertainties in 
reaching decisions on alternative standard levels to propose, the 
Administrator judges that it is most appropriate to examine where the 
evidence of associations observed in the epidemiologic studies is 
strongest and, conversely, where he has appreciably less confidence in 
the associations observed in the epidemiologic studies. Based on 
information evaluated in the 2019 ISA and ISA Supplement, the 
Administrator recognizes that health effects may occur over the full 
range of concentrations observed in the long- and short-term 
epidemiologic studies and that no discernible threshold for any effects 
can be identified based on the currently available evidence (U.S. EPA, 
2019a, section 1.5.3, U.S. EPA, 2022a, section 2.2.3.1 and 2.2.3.2). He 
also recognizes, in taking note of CASAC advice and the distributional 
statistics analysis discussed in section II.B.3.b above and in the PA, 
that there is significantly greater confidence in observed associations 
over certain parts of the air quality distributions in the studies, and 
conversely, that there is significantly diminished confidence in 
ascribing effects to concentrations toward the lower part of the 
distributions.
    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. The current CASAC also supported assessing 
the mean (or median) concentrations, but also suggested additional 
approaches that could be explored.\100\ In utilizing this evidence-
based approach, the Administrator looks to study-reported

[[Page 5626]]

means from the key epidemiologic studies (as shown in Figure 1 and 
Figure 2) available in this reconsideration. He notes that there have 
been new approaches to estimating exposure concentrations since the 
2012 review, such that many of the available key epidemiologic studies 
include new approaches that apply hybrid modeling techniques to 
estimate exposures. In looking at the epidemiologic studies, he 
considers these studies in two groups: (1) monitor-based studies 
(epidemiologic studies that used ground-based monitors to estimate 
exposure, similar to approaches used in past reviews), and (2) hybrid 
modeling-based studies (epidemiologic studies that used hybrid modeling 
approaches to estimate exposures). As such, he recognizes that 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. Further, he recognizes that studies that use hybrid 
modeling approaches employ methods to estimate ambient PM2.5 
concentrations across large geographical areas, including those 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 to the maximum monitor 
design values used to assess compliance. For the key U.S. monitor-based 
epidemiologic studies, the study reported mean concentrations range 
from 9.9-16.5 [mu]g/m\3\ and for the U.S. hybrid modeling based key 
epidemiologic studies, the mean concentrations range from 9.3-12.2 
[mu]g/m\3\.
---------------------------------------------------------------------------

    \100\ The Administrator notes that some members of the CASAC 
advised that ``use of the mean to define where the data provide the 
most evidence is conservative. . .'' (Sheppard, 2022a, p. 3 of 
consensus letter) and advised that ``for the purpose of informing 
the adequacy of the standards'' (Sheppard, 2022a, p. 8 of consensus 
responses) that the EPA in future reviews 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.
---------------------------------------------------------------------------

    In thinking further about the relationship between mean 
PM2.5 concentrations in key epidemiologic studies and annual 
design values, the Administrator specifically notes that in a given 
area, the area design value is determined by the monitor in an area 
with the highest PM2.5 concentrations and is used to 
determine compliance with the standard. He observes, as detailed above 
in the air quality analyses in section I.D.5, 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 concentrations, monitored areas will 
generally have higher concentrations than areas without monitors. Thus, 
when a study reports a mean that reflects the average of annual average 
measured concentrations for an area, the area design value will 
generally be higher. Similarly, when a study reports a mean that 
reflects the average of annual average concentrations estimated at 
various points across an area using a hybrid modeling approach, the 
area design value will generally be higher. More specifically, the 
Administrator observes that the additional air quality analyses 
(described in section I.D.5) suggest that the area annual design value 
is greater than the study-reported mean values by 10-20% for monitor-
based studies and 15-18% for hybrid modeling with population weighting 
applied.\101\ As such, the Administrator observes that a policy 
approach for setting a standard level that requires the design value 
monitor to meet study-reported means will 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 
PM2.5 concentrations are generally highest) will be exposed 
to PM2.5 concentrations below the air quality conditions 
reported in the epidemiologic studies where there is the highest 
confidence of an association.\102\ In addition, he specifically notes 
that an annual standard level that is no more than 10-20% higher than 
the study-reported means in the U.S. monitor-based studies (i.e., for 
the lowest study reported mean value of 9.9 [mu]g/m\3\, this means an 
annual standard level of approximately 10.9-11.9 [mu]g/m\3\) and no 
more than 15-18% higher for the U.S. hybrid modeling with population 
weighting applied (i.e., for the lowest study reported mean value of 
9.3 [mu]g/m\3\, this means an annual standard level of approximately 
10.7-11.0 [mu]g/m\3\), would generally maintain air quality exposures 
at or below those associated with the study-reported mean 
PM2.5 concentrations, exposures for which we have the 
strongest support for adverse health effects occurring. Based on this, 
the Administrator concludes that a revised standard level of 9.0 to 
10.0 [mu]g/m\3\ would generally limit air quality exposures to levels 
well below those associated with the study-reported mean 
PM2.5 concentrations in the key epidemiologic studies. A 
revised standard level of 11.0 [mu]g/m\3\ would maintain air quality 
exposures to below those associated with most of these study-reported 
means, and a revised standard level of 8.0 [mu]g/m\3\ would maintain 
air quality exposures to far below all of these study-reported means. 
The Administrator notes that every member of the CASAC found that the 
information on study-reported means supported revising the annual 
standard level to 10.0 [mu]g/m\3\, with the minority of the CASAC 
advising that these data also supported a revised annual standard level 
of 10.0-11.0 [mu]g/m\3\ and the majority of the CASAC advising that 
these study-reported means, in conjunction with additional bodies of 
evidence, supported a revised annual standard level of 8.0-10.0 [mu]g/
m\3\.
---------------------------------------------------------------------------

    \101\ The Administrator also notes that there are a limited 
number of studies that report a study mean that does not reflect the 
exposure concentrations used in the epidemiologic study to assess 
the reported association. These studies do not report population-
weighted study means and are not considered here given the 
substantial difference in concentrations used to assess the 
association versus those used to calculate the study-reported means.
    \102\ Based on the available air quality information, it would 
be expected that an area with a study reported mean of 10 [mu]g/m\3\ 
would have a gradient of concentrations across the area, with higher 
concentrations near the design value monitor and lower 
concentrations away from it. If the level of the standard were 
revised to 10.0 [mu]g/m\3\, then it would be expected that there 
would still be a gradient of concentrations, but the 
PM2.5 concentrations across the area would be reduced in 
order to meet the revised standard at the design value monitor, and 
therefore areas away from the design value monitor would be expected 
to have a gradient of PM2.5 concentrations at or below 
10.0 [mu]g/m\3\ as well.
---------------------------------------------------------------------------

    The Administrator also considers additional information from 
epidemiologic studies, consistent with CASAC advice, to take into 
account the broader distribution of PM2.5 concentrations, 
including the 25th percentiles of the distributions, and the degree of 
confidence in the observed associations over the broader air quality 
distribution. In considering this additional information, he 
understands that the PA presented information on the distributions of 
PM2.5 concentrations, when available, from key epidemiologic 
studies to provide a general frame of reference as to the part of the 
distribution within which the data become appreciably more sparse and, 
thus, where his confidence in the associations observed in 
epidemiologic studies would become appreciably less. As discussed in 
section II.B.3.b above and presented in Figure 1 and Figure 2 above, he 
observes that most studies do not report such data and the conclusions 
that can be drawn from such information across the full body of 
evidence are quite limited. However, the Administrator takes note of 
additional population-level data that are available and in considering 
the long-term PM2.5 concentrations associated with the 25th

[[Page 5627]]

percentile values of the population-level data for the studies for 
which such data are available, he observes that for the three key U.S. 
epidemiologic studies that use hybrid modeling approaches that apply 
population weighting and report these data, the values reported were 
6.7 [mu]g/m\3\, 9.1 [mu]g/m\3\ and 9.1 [mu]g/m\3\. For the U.S.-based 
studies that use ground-based monitors, the 25th percentiles ranged 
from 11.5 [mu]g/m\3\ to just below 13.0 [mu]g/m\3\.
    The Administrator notes that there are substantial uncertainties 
associated with using 25th percentile data for purposes of setting this 
standard and these uncertainties are heightened by the relatively few 
studies which report such data and the fact that, by definition, this 
data is relatively less common even within a study for which it is 
reported. At the same time, the Administrator is conscious of his 
obligation to set primary standards with an adequate margin of safety 
and recognizes that some members of the CASAC advised that these data 
indicate that effects are occurring below the reported means of 
studies. Balancing these concerns about the need to provide some 
protection against uncertain risks with the obligation to not set 
standards that are more stringent than necessary, the Administrator 
preliminarily concludes that a revised standard should limit exposures 
to ambient concentrations near the 25th percentile of reported studies. 
Given this consideration, the Administrator recognizes that a standard 
level of 8.0-10.0 [mu]g/m\3\ is generally within the range of these 
values, while a standard level of 11.0 [mu]g/m\3\ is above the 25th 
percentile values reported in the hybrid model-based studies but below 
the 25th percentile values in studies that use ground-based monitors. 
Based on this, the Administrator recognizes that a standard within the 
range of 8.0-11.0 [mu]g/m\3\ would limit exposures to ambient 
concentrations near the 25th percentile reported in the available 
studies, with the lower end of this range further limiting those 
exposures.
    The Administrator also takes into consideration the long-term mean 
PM2.5 concentrations reported in Canadian epidemiologic 
studies that, in the context of the larger body of available evidence, 
provided support for causal or likely to be causal determinations 
between PM2.5 exposure and health effects, as summarized in 
the 2019 ISA and ISA Supplement. He notes that the study-reported means 
from these Canadian studies tend to be somewhat lower than those 
reported from the key epidemiologic studies in the U.S. ranging from 
6.9-13.3 [mu]g/m\3\ for the monitor-based studies and 5.9-9.8 [mu]g/
m\3\ for the hybrid model-based studies. However, the Administrator 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., 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, as detailed above. He additionally notes 
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 8.0-10.0 [mu]g/m\3\. Based on this, 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.
    The Administrator also notes that the CASAC recommended looking at 
the studies that included analyses that restrict annual average 
PM2.5 concentrations to concentrations below the level of 
the current annual standard in evaluating an appropriate range of 
levels for a revised annual standard. In this reconsideration, there 
are two key studies available (Di et al., 2017b and Dominici et al., 
2019) that restrict annual average PM2.5 concentrations to 
less than 12 [mu]g/m\3\. These restricted analyses report positive and 
statistically significant associations with all-cause mortality, and 
both report mean PM2.5 concentrations of 9.6 [mu]g/m\3\. The 
Administrator does note that uncertainties exist in these analyses 
(described in more detail in sections II.B.3.b and II.D.2.a above), 
including uncertainty in how the studies exclude concentrations (e.g., 
at what spatial resolution are concentrations being excluded), which 
would make it difficult to compare concentrations in restricted 
analyses directly to design values. However, he does note that an 
annual standard level of 9.0-10.0 [mu]g/m\3\ would be close to these 
reported mean values, while a standard level of 11.0 [mu]g/m\3\ would 
be above and a standard level of 8.0 [mu]g/m\3\ would be much further 
below.
    The Administrator additionally considers 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). The Administrator notes that there are three studies 
available in this reconsideration (Henneman et al. (2019b), Corrigan et 
al. (2018), and Sanders et al. (2020a)) that account for changes in 
PM2.5 concentrations due to implementation of policies and 
assess whether there was evidence of changes in associations with 
mortality or cardiovascular morbidity due to changes in annual 
PM2.5 concentrations. The Administrator notes that in each 
of these studies, prior to implementation of the policies, mean 
PM2.5 concentrations were below the level of the current 
annual standard level (12.0 [mu]g/m\3\) and ranged from 10.0 [mu]g/m\3\ 
to 11.1 [mu]g/m\3\. The Administrator notes that these 
studies report positive and significant associations between mortality 
and cardiovascular morbidity and reductions in ambient PM2.5 
(described above in section II.B.3.b and in Table 3-12 of the PA) and 
notes that these studies suggest 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 [mu]g/m\3\. The Administrator notes that a revised annual 
standard level of 9.0-10.0 [mu]g/m\3\ would be at or below the lowest 
starting concentration of these accountability studies (i.e., 10.0 
[mu]g/m\3\).
    In addition to the evidence, the Administrator also considers the 
results of the risk assessment. The PA includes a risk assessment that 
estimates PM2.5-attributable mortality risk associated with 
PM2.5 air quality that has been adjusted to simulate ``just 
meeting'' the current standards, as well as potential alternative 
standards. These 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. Such an approach does not incorporate any elements of 
uncertainty in associations of health effects at lower concentrations 
and simulated air quality improvements will always lead to proportional 
decreases in risk (i.e., each additional [mu]g/m\3\ reduction produces 
additional benefits with no clear stopping point). Therefore, 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 risk assessment estimates that the current primary 
PM2.5 standards could allow a substantial number of 
PM2.5-associated deaths in the U.S. Additionally, compared 
to the current

[[Page 5628]]

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 [mu]g/m\3\, 15-19% for a level of 10.0 [mu]g/m\3\, 22-28% 
for a level of 9.0 [mu]g/m\3\, and 30-37% for a level of 8.0 [mu]g/
m\3\) (U.S. EPA, 2022a, Table 3-17). The CASAC concurred with the PA's 
assessment that meaningful risk reductions will result from lowering 
the annual PM2.5 standard (Sheppard, 2022a, p. 16 of 
consensus responses).
    The PA also provides information on the distribution of 
concentrations associated with the estimated mortality risk at each 
alternative standard level assessed (U.S. EPA, 2022a, sections 3.4.2.2 
and 3.6.2.2, Figure 3-18 and 3-19). Further evaluating these results 
can help clarify the percentage of the exposure reductions that 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. When meeting a standard level of 11.0 
[mu]g/m\3\, the risk is estimated to be associated with exposure 
concentrations that are generally greater than 10.0 [mu]g/m\3\, while 
for a standard level of 10.0 [mu]g/m\3\, the majority of the days 
contributing to the risk estimates are estimated to be below 10.0 
[mu]g/m\3\. When meeting an annual standard or 9.0 [mu]g/m\3\, the 
majority of the exposure concentrations are estimated to be 8.0-9.0 
[mu]g/m\3\, while for a standard level of 8.0 [mu]g/m\3\, most of the 
days are below 8.0 [mu]g/m\3\. The Administrator notes that the 
evidence suggests that majority of the study-reported means are above 
10.0 [mu]g/m\3\ (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) and that at PM2.5 
concentrations less than 8.0 [mu]g/m\3\, the 2019 ISA notes that 
uncertainties remain in the shape of the C-R curve. He thus recognizes 
that there is increasing uncertainty in quantitative estimates of 
PM2.5-associated mortality risk for alternative standard 
levels at the lower end of the range of 8.0-11.0 [mu]g/m\3\.
    As discussed more above, the Administrator also recognizes that the 
risk assessment was able to include an at-risk analysis that estimated 
the potential long-term PM2.5-attributable exposure and 
mortality risk in older adults, stratified by racial/ethnic 
demographics, when meeting a revised annual standard with a lower 
level. While the Administrator recognizes that this analysis is subject 
to the multiple uncertainties and limitations (sections II.C.2 and 
II.D.2.b), he does note that the analysis suggests that a revised 
annual standard level within the range of 8.0 to 11.0 [mu]g/m\3\ is 
estimated to reduce PM2.5 exposure and may also help to 
mitigate risks. Based on the case study areas included in the analysis, 
The Administrator notes that what urban areas are included in the risk 
assessment analysis will greatly influence the results but notes that 
based on the areas included in the analyses, the results show the 
largest impact is on reducing exposure and risk in Black populations, 
who were estimated in the risk assessment case study areas to have the 
highest levels of exposures and the greatest rates of premature 
mortality risk. The Administrator also notes that, similar to the main 
risk estimates discussed above, there is increasing uncertainty in 
quantitative estimates of stratified risk estimates at the lower end of 
the range of standard levels assessed.
    The Administrator recognizes that judgments about the appropriate 
weight to place on any of the factors discussed above should reflect 
consideration not only of the relative strength of the evidence but 
also of the important uncertainties that remain in the evidence and the 
quantitative information being considered in this reconsideration. The 
Administrator also recognizes that the CAA requires him to set 
standards that in his judgment are neither more stringent nor less 
stringent than necessary to protect public health with an adequate 
margin of safety. Based on the above considerations, the Administrator 
concludes that it is appropriate to propose to set a level for the 
primary annual PM2.5 standard within the range of 9.0 to 
10.0 [mu]g/m\3\, while also taking comment on a level for the primary 
annual PM2.5 standard as low as 8.0 [mu]g/m\3\ and as high 
as 11.0 [mu]g/m\3\. The Administrator provisionally concludes that a 
standard level within the range of 9.0 to 10.0 [mu]g/m\3\ would reflect 
appropriate approaches to placing the most weight on the strongest 
available evidence, while placing less weight on much more limited 
evidence and on more uncertain analyses of information available from a 
relatively small number of studies. He notes that a standard set at 9.0 
to 10.0 [mu]g/m\3\ would be at or below the study-reported mean 
PM2.5 concentrations in the key U.S. epidemiologic studies, 
exposures for which we have the strongest support for adverse health 
effects occurring. Further, in considering margin of safety, he notes 
that an annual standard level that is no more than 10-20% higher than 
the study-reported means in the U.S. monitor-based studies (i.e., for 
the lowest study reported mean value of 9.9 [mu]g/m\3\, this means an 
annual standard level of approximately 10.9-11.9 [mu]g/m\3\) and no 
more than 15-18% higher for the U.S. hybrid modeling with population 
weighting (i.e., for the lowest study reported mean value of 9.3 [mu]g/
m\3\, this means an annual standard level of approximately 10.7-11.0 
[mu]g/m\3\), would generally maintain air quality exposures at or below 
those associated with the study-reported mean PM2.5 
concentrations. Additionally, the Administrator also notes that these 
key U.S. epidemiologic studies utilize cohorts that include populations 
identified as at-risk, including children and older adults, as well as 
individuals in the general population with pre-existing disease, like 
cardiovascular disease and respiratory disease. Based on this 
information, he concludes that a revised standard level of 9.0-10.0 
[mu]g/m\3\ would limit air quality exposures to concentrations well 
below those associated with the study reported mean, studies which 
include and assess impacts on the most at-risk populations. Thus, the 
Administrator provisionally concludes that a standard level within this 
range would appropriately provide an adequate margin of safety for the 
populations most at risk for adverse health effects associated with 
exposure to PM2.5.
    The Administrator also considers other lines of evidence, including 
the study reported means from epidemiologic studies that restrict 
concentrations to levels below 12 [mu]g/m\3\, the 25th percentiles 
values reported by a subset of epidemiologic studies, and the 
information from the accountability studies. He notes that a standard 
in the range of 9.0 to 10.0 [mu]g/m\3\ would limit exposures to ambient 
concentrations near the 25th percentile reported in the available 
studies, with a standard level of 9.0 [mu]g/m\3\ limiting those 
exposures somewhat more than a standard level of 10.0 [mu]g/m\3\. He 
also notes that a standard in the range of 9.0 to 10.0 [mu]g/m\3\ would 
be near the value of the study reported means from the two available 
long-term restricted analyses studies (i.e., 9.6 [mu]g/m\3\). The 
Administrator notes a standard level of 9.0-10.0 [mu]g/m\3\, would also 
be at or below the lowest starting concentration of the newest 
available accountability studies (i.e., 10.0-11.1 [mu]g/m\3\). The 
Administrator also considers the results from the risk assessment. He 
recognizes that the risk estimates should be considered along with the 
inherent uncertainties and

[[Page 5629]]

limitations of such analyses when informing judgments about the 
potential for additional public health protection associated with 
PM2.5 exposure and related health effects. When looking at 
the risk assessment results, he notes that an annual standard level of 
9.0-10.0 [mu]g/m\3\ is estimated to reduce exposure concentrations such 
that those remaining risks are associated with exposure concentrations 
that are below most of the study-reported means in the key U.S. 
epidemiologic studies, where we have the strongest support for adverse 
health effects occurring, and below PM2.5 concentrations 
(i.e., 8 [mu]g/m\3\) where the 2019 ISA notes that uncertainties remain 
in the shape of the C-R curve, particularly for a standard level as low 
as 9.0 [mu]g/m\3\. Lastly, the Administrator also notes that every 
member of the CASAC found that the available scientific evidence and 
information supported revising the annual standard level to a level of 
10.0 [mu]g/m\3\. Additionally, the majority of the CASAC also 
recommended that the available evidence and information supported 
revision to a level of 9.0 [mu]g/m\3\. Thus, recognizing the 
uncertainties in the evidence and the necessity of providing requisite 
protection, with an adequate margin of safety, the Administrator is 
proposing to set the level of the annual standard in the range of 9.0-
10.0 [mu]g/m\3\, and solicits comments on the appropriate standard 
level within that range.
    While the Administrator recognizes that some members of the CASAC 
advised, and the PA concluded, that the available scientific 
information provides support for considering a range that extends up to 
11.0 [mu]g/m\3\ and down to 8.0 [mu]g/m\3\, he provisionally concludes 
that proposing such an extended range would not be appropriate at this 
time. More specifically, the Administrator provisionally concludes that 
proposing to revise the annual standard level to above 10.0 [mu]g/m\3\ 
and as high as 11.0 [mu]g/m\3\ would reflect a public health policy 
approach that would place less weight on setting a standard level at or 
below the study-reported means from a number of key U.S. epidemiologic 
studies and less weight on the risk assessment results. Such an 
approach would also place little or no weight on the study reported 
means from epidemiologic studies that restrict concentrations to below 
12 [mu]g/m\3\ and the 25th percentile concentrations reported by a 
subset of epidemiologic studies. The Administrator notes that such an 
approach may fail to provide an adequate margin of safety in light of 
the evidence available in this reconsideration. In considering revision 
to the annual standard level to below 9.0 [mu]g/m\3\ and as low as 8.0 
[mu]g/m\3\, the Administrator notes that such a level would be 
substantially below the study-reported means and would not recognize 
the controlling nature of the design value monitor with respect to the 
concentration gradients consistently occurring across urban areas. The 
Administrator also recognizes that the evidence and uncertainties for 
public health benefits of lower standards exists on a continuum across 
the range of possible standard levels. He preliminarily judges that the 
evidence is sufficient to support standards in the range of 9.0-10.0 
[mu]g/m\3\, recognizing that the selection of a final standard level 
will depend on judgments about the relative weight to place on various 
aspects of the evidence and how to provide for an adequate margin of 
safety. However, the Administrator preliminarily judges that the 
available information and evidence are not sufficient to warrant 
revising the level of the annual standard below 9.0 [mu]g/m\3\. He 
finds the uncertainties as to the public health risks and benefits 
associated with such a standard to be too great at this time. 
Nonetheless, while the Administrator notes these considerations above, 
he solicits comment on revising the annual standard down to a level 
below 9.0 [mu]g/m\3\ and as low as 8.0 [mu]g/m\3\, as well as to above 
10.0 [mu]g/m\3\ and as high as 11.0 [mu]g/m\3\, and on approaches for 
interpreting the scientific evidence and rationales that would support 
such a level.

E. Proposed Decisions on the Primary PM2.5 Standards

    Taking the above considerations into account, upon reconsidering 
the current primary PM2.5 standards in light of the 
currently available scientific evidence and quantitative information, 
the Administrator proposes to revise the level of the primary annual 
PM2.5 standard from 12.0 [mu]g/m\3\ to within the range of 
9.0 to 10.0 [mu]g/m\3\ and to retain the 24-hour standard level at 35 
[mu]g/m\3\. In the Administrator's judgment, such a suite of primary 
PM2.5 standards and the rationale supporting such levels 
could reasonably be judged to reflect the appropriate consideration of 
the strength of the available evidence and other information and their 
associated uncertainties and the advice of the CASAC.
    The Administrator recognizes that the final suite of standards will 
reflect the Administrator's ultimate judgments in the final rulemaking 
as to the suite of primary PM2.5 standards that are 
requisite to protect the public health with an adequate margin of 
safety from effects associated with PM2.5 exposures. The 
final judgments to be made by the Administrator will appropriately 
consider the requirement for standards that are neither more nor less 
stringent than necessary and will recognize 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.
    Having reached his provisional judgment to propose revising the 
annual standard level from 12.0 to within a range of 9.0 to 10.0 [mu]g/
m\3\ and to propose retaining the 24-hour standard level at 35 [mu]g/
m\3\, the Administrator solicits public comment on this range of levels 
and on approaches to considering the available evidence and information 
that would support the choice of levels within this range. The 
Administrator also solicits public comment on alternative annual 
standard levels down to 8.0 [mu]g/m\3\ and up to 11.0 [mu]g/m\3\, on an 
alternative 24-hour standard level as low as 25 [mu]g/m\3\ and on the 
combination of annual and 24-hour standards that commenters may believe 
is appropriate, along with the approaches and rationales used to 
support such levels. For example, the EPA solicits comments on the 
uncertainties in the reported associations between daily or annual 
average PM2.5 exposures and mortality or morbidity in the 
epidemiologic studies, the significance of the 25th percentile of 
ambient concentrations reported in studies, the relevance and 
limitations of international studies, and other topics discussed in 
section II.D.3.b.

III. Rationale for Proposed Decisions on the Primary PM10 Standard

    This section presents the rationale for the Administrator's 
proposed 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,\103\ and 
evaluated in the 2019 ISA, on human health effects associated with 
PM10-2.5 in ambient air. As described in section 1.2 of the 
ISA Supplement, the

[[Page 5630]]

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, 2022a). 
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 rationale also takes into 
account: (1) the PA evaluation of the policy-relevant information in 
the 2019 ISA; (2) CASAC advice and recommendations, as reflected in 
discussions of the draft of the PA at public meetings and in the 
CASAC's letter dated March 18, 2022, to the Administrator; and (3) 
public comments received during the development of the PA.
---------------------------------------------------------------------------

    \103\ In addition to the review's opening ``call for 
information'' (79 FR 71764, December 3, 2014), the current 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.
---------------------------------------------------------------------------

    In presenting the rationale for the Administrator's proposed 
decision and its foundations, section III.A provides background and 
introductory information for this reconsideration of the primary 
PM10 standard. It includes background on the 2020 final 
decision to retain the primary PM10 standard (section 
III.A.1) and also describes the general approach for this 
reconsideration (section III.A.2) Section III.B summarizes the key 
aspects of the currently available scientific evidence for 
PM10-2.5-related health effects. Section III.C presents the 
Administrator's proposed conclusions regarding the adequacy of the 
primary PM10 standard (section III.C.3), drawing on 
evidence-based considerations (section III.C.2) and advice from the 
CASAC (section III.C.1).

A. General Approach

    The current primary PM10 standard was affirmed in 2020 
based on the scientific information available at that time, 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 (85 FR 82725, December 18, 2020). 
With the 2020 decision, the Administrator retained the existing 24-hour 
primary PM10 standard, with its level of 150 [mu]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). The 
subsection below focuses on the key considerations, and the prior 
Administrator's conclusions, for PM10-2.5-related health 
effects and the adequacy of the primary PM10 standard in the 
2020 review.
1. Background on the Current Standard
    In the 2019 ISA, the strongest evidence for PM10-2.5-
related health effects was for cardiovascular effects, respiratory 
effects, and premature mortality following short-term exposures. For 
each of these categories of effects, the 2019 ISA concludes that the 
evidence was ``suggestive of, but not sufficient to infer, a causal 
relationship''. Specifically, the health effects evidence evaluated in 
the 2019 ISA included an expanded body of scientific evidence that has 
become available since the completion of the 2009 ISA linking short-
term PM10-2.5 to health outcomes such as premature death and 
hospital visits (U.S. EPA, 2009a; U.S. EPA, 2019a). This evidence base 
evaluated the causal relationships between short-term exposure to 
PM10-2.5 and a broad range of health effects (U.S. EPA, 
2019a, section 1.4.2). These effects associated with short-term 
exposure ranged from hospital admissions and emergency department 
visits for cardiovascular effects (documented in epidemiologic studies 
that reported PM10-2.5 associations with cardiovascular 
hospital admissions and emergency department visits in study locations 
with mean 24-hour average PM10-2.5 concentrations ranging 
from 7.4 to 13 [mu]g/m\3\) and respiratory effects (documented in 
epidemiologic studies that reported PM10-2.5 associations 
with respiratory hospital admissions and emergency department visits in 
study locations with mean 24-hour average concentrations ranging from 
5.6 to 16.2 [mu]g/m\3\) to mortality (documented in epidemiologic 
studies that reported PM10-2.5 associations with mortality 
in study areas with mean 24-hour average concentrations ranging from 
6.1 to 16.4 [mu]g/m\3\). In addition to the epidemiologic studies, the 
evidence base included a small number of controlled human exposure 
studies and animal toxicological studies that provided insight into the 
biological plausibility of these effects. Collectively, the 
epidemiologic studies, controlled human exposure, and animal 
toxicological studies, with their inherent uncertainties, contributed 
to 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, 
cancer, and mortality (U.S. EPA, 2019a, section 1.4.2). The 2019 ISA 
includes expanded evidence for the relationships between long-term 
exposures and cardiovascular effects, metabolic effects, nervous system 
effects, cancer, and mortality. While the evidence available in the 
2019 ISA included additional health outcomes, including those 
associated with long-term PM10-2.5 exposure, key limitations 
in the evidence that were identified in the 2009 ISA persist in studies 
evaluated in the 2019 ISA.
    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 \104\ 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

[[Page 5631]]

PM10-2.5 exposures using non-inhalation (e.g., intratracheal 
instillation) route.\105\ 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:
---------------------------------------------------------------------------

    \104\ 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).
    \105\ 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 the apparent variability in associations between short-
term PM10-2.5 exposures and health effects across study 
locations (U.S. EPA, 2019a). Specifically, the relative lack of 
information on the chemical and biological composition of 
PM10-2.5 as well as potential spatial and temporal 
variability in PM10-2.5 exposures complicates the 
interpretation of results between study locations (U.S. EPA, 2009b; 
U.S. EPA, 2019a).
    Consistent with the general approach routinely employed in NAAQS 
reviews, the initial consideration in the 2020 review of the primary 
PM10 standard was with regard to the adequacy of protection 
provided by the then-existing standard. Key aspects of that 
consideration are summarized below.
i. Considerations Regarding the Adequacy of the Existing Standard in 
the 2020 Review
    In the 2020 final decision, the EPA retained the existing 24-hour 
primary PM10 standard with its level of 150 [mu]g/m\3\ and 
its one-expected-exceedance form on average over three years to 
continue to provide public health protection against exposures to 
PM10-2.5 (85 FR 82727, December 18, 2020). In reaching his 
decision, the 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 
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 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 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 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 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 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 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 proposed decision, the 
Administrator emphasized evidence supporting ``causal'' and ``likely to 
be causal'' relationships, and therefore, judged that the 
PM10-2.5-related health effects evidence provided

[[Page 5632]]

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 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 Administrator concluded that the primary 
PM10 standard (in all of its elements) 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. General Approach and Key Issues in This Reconsideration of the 2020 
Final Decision
    To evaluate whether it is appropriate to consider retaining the 
current primary PM10 standard, or whether consideration of 
revision is appropriate, the EPA has adopted an approach in this 
reconsideration that builds upon the general approach used in past 
reviews and reflects the body of evidence and information now 
available, as well as the assessments and evaluations performed in 
those reviews. As summarized above, the Administrator's decision in the 
2020 review was based on an integration of PM10-2.5-related 
health effects information with the judgments on the public health 
significance of key effects, policy judgments as to when the standard 
is requisite, consideration of CASAC advice, and consideration of 
public comments.
    Similarly, in this reconsideration, information is drawn from 
recent studies of PM10-2.5-related health effects. In so 
doing, the PA considers information critically analyzed and 
characterized in the 2019 ISA, as well as consideration of the 
associated uncertainties and limitations for the available evidence.

B. 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 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.
1. 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.
a. Mortality
i. Long-Term Exposures
    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) 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

[[Page 5633]]

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).
ii. Short-Term Exposures
    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 a variety of 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).
b. Cardiovascular Effects
i. Long-Term Exposures
    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).
ii. Short-Term Exposures
    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 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

[[Page 5634]]

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).
c. Respiratory Effects--Short-Term Exposures
    Based on a small number of epidemiologic studies observing 
associations with some respiratory effects and limited evidence from 
experimental studies to support biological plausibility, 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.'' 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 great sense 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).
d. Cancer--Long-Term Exposures
    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).
e. Metabolic Effects--Long-Term Exposures
    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).
f. Nervous System Effects--Long-Term Exposures
    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).

C. Proposed Conclusions on the Primary PM10 Standard

    In reaching proposed conclusions on the current primary 
PM10 standard (presented in section III.C.3), the 
Administrator has taken into account policy-relevant evidence-based 
considerations discussed in the PA (summarized in section III.C.2), as 
well as advice from the CASAC and public comments on the standard 
received thus far in the reconsideration (section III.C.1). In general, 
the role of the PA is

[[Page 5635]]

to help ``bridge the gap'' between the Agency's assessment of the 
available evidence, and the judgments required of the Administrator in 
determining whether it is appropriate to retain or revise the NAAQS. 
Evidence-based considerations draw upon the EPA's integrated evaluation 
of the scientific evidence of PM10-2.5-related health 
effects presented in the 2019 ISA (summarized in section III.B above) 
to address key policy-relevant questions in the reconsideration.
    The approach to reviewing the primary PM10 standard is 
consistent with requirements of the provisions of the CAA related to 
the review of the NAAQS and 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 concentrations for which scientists generally agree that health 
effects are likely to occur, through lower concentrations at which the 
likelihood and magnitude of response becomes 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 proposed decision on the adequacy of the primary 
PM10 standard described below is a public health policy 
judgment by the Administrator that draws on the scientific evidence for 
health effects and judgments about how to consider the uncertainties 
and limitations that are inherent in the scientific evidence. The four 
basic elements of the NAAQS (i.e., indicator, averaging time, form, and 
level) have been considered collectively in evaluating the health 
protection afforded by the current standard. The Administrator's final 
decision will additionally consider public comments received on this 
proposed decision.
1. CASAC Advice in This Reconsideration
    The CASAC has provided advice on the adequacy of the current 
primary PM10 standard in the context of its review of the 
draft PA (Sheppard, 2022a).\106\ In this context, the CASAC supported 
the preliminary conclusion in the 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 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).
---------------------------------------------------------------------------

    \106\ A limited number of public comments have also been 
received in this reconsideration to date, including comments focused 
on the draft PA. Of the public comments that addressed the adequacy 
of the current primary PM10 standard, most commenters 
supported the preliminary conclusion that it is appropriate to 
consider retaining the current primary PM10 standard, 
without revision. However, one nonprofit organization suggested that 
the primary PM10 standard should be strengthened to a 
level of 45 [mu]g/m\3\, consistent with the World Health 
Organization Global Air Quality Guideline (WHO, 2021).
---------------------------------------------------------------------------

    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. Evidence-Based Considerations in the Policy Assessment
    With regard to the current evidence on health effects associated 
with long and short--term PM10-2.5 exposure health effects, 
the PA notes that 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 (U.S. EPA, 2022b, section 
4.3.1). While significant uncertainties remain, as described below and 
summarized in the PA (U.S. EPA, 2022b, section 4.5), the PA recognizes 
that this expanded body of evidence has broadened the range of effects 
that have been linked with PM10-2.5 exposures. The 
uncertainties in the available epidemiologic studies contribute to the 
determinations in the 2019 ISA that the evidence for 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 is ``suggestive of, but 
not sufficient to infer'' causal relationships (U.S. EPA, 2019a; U.S. 
EPA, 2022b, section 4.3.1). Drawing from this information, the PA 
concludes that the evidence continues to provide support for 
maintaining a standard that provides some measure of protection against 
exposures to PM10-2.5, regardless of location, sources of 
origin, or particle composition (U.S. EPA, 2022b, section 4.5).
    With regard to uncertainties, the PA recognizes that the 2019 ISA 
notes that important uncertainties remain in the evidence base for 
PM10-2.5-related health effects. As summarized in section 
III.B above and in the PA (U.S. EPA, 2022b, sections 4.3.1 and 4.5). 
These uncertainties include those related to variability in 
PM10-2.5 exposure estimates used in epidemiologic studies, 
in the independence of PM10-2.5 health effect associations, 
and in the biological plausibility of the PM10-2.5-related 
health effects. These uncertainties contribute to the determinations in 
the 2019 ISA that the evidence for short- and long-term 
PM10-2.5 exposure in key health effect categories is 
``suggestive of, but not sufficient to infer'' causal relationships 
(U.S. EPA, 2019a). Taking this information into consideration, the PA 
concludes that, as in previous reviews, such uncertainties raised 
questions regarding the degree to which additional public health 
protection would be achieved by revising the existing PM10 
standard (U.S. EPA, 2022b, section 4.5).
    With regard to the indicator for the primary PM10 
standard, the PA notes that the evidence continues to support retaining 
the PM10 indicator to provide public health protection 
against PM10-2.5-related effects. Consistent with the 
approaches in previous reviews, a standard with a PM10 mass-
based

[[Page 5636]]

indicator, in conjunction with a PM2.5 mass-based standard, 
will result in controlling allowable concentrations of 
PM10-2.5. Given that the use of the PM10 
indicator does include consideration of both PM2.5 and 
PM10-2.5 concentrations, the 2019 ISA provides a comparison 
of the relative contribution of PM2.5 and 
PM10-2.5 to PM10 concentrations, finding 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. In light of 
this and consistent with the approaches in previous reviews, the PA 
concludes that a PM10 standard, set at a single unvarying 
level, will generally result in lower allowable concentrations of 
PM10-2.5 in urban areas than in nonurban areas. In this way, 
the PM10 indicator will target protection by allowing less 
PM10-2.5 in areas that experience high concentrations of 
potentially contaminating PM2.5. Thus, the evidence 
continues to support retaining the PM10 indicator.
    When the above information is taken together, the PA concludes that 
available evidence does not call into question the adequacy of the 
public health protection provided by the current primary 
PM10 standard in order to protect against 
PM10-2.5 exposures. Specifically, the PA notes that while 
the evidence supports maintaining a PM10 standard to provide 
some measure of protection against PM10-2.5 exposures, 
uncertainties in the evidence lead to questions regarding the potential 
public health implications of revising the existing PM10 
standard. Thus, the PA concludes that the evidence does not call into 
question the adequacy of the public health protection afforded by the 
current primary PM10 standard (U.S. EPA, 2022b, section 
4.5).
3. Administrator's Proposed Decision on the Current Primary 
PM10 Standard
    This section summarizes the Administrator's considerations and 
proposed conclusions related to the current primary PM10 
standard and presents his proposed decision to retain that standard, 
without revision. 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. He 
recognizes that 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, 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 recognizes that his final decision will 
be based on an interpretation of the scientific evidence that neither 
overstates nor understates its strengths and limitations, nor the 
appropriate inferences to be drawn.
    Consistent with previous reviews, the Administrator first considers 
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 recognizes 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.B 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 notes that the evidence 
for several PM10-2.5-related health effects has expanded 
since the completion of the 2009 ISA, but important uncertainties 
remain. Epidemiologic studies evaluated in the 2019 ISA continue to 
report positive associations between short-term exposure to 
PM10-2.5 and mortality and morbidity in cities across North 
America, Europe, and Asia, where PM10-2.5 sources and 
composition are expected to vary widely, but across studies 
inconsistency remains in the approaches used to estimate 
PM10-2.5 exposures. While the Administrator recognizes that 
important uncertainties remain, he also recognizes that the expansion 
in the number of studies evaluating PM10-2.5 exposures and 
health effects since the completion of the 2009 ISA has broadened the 
range of effects that may be linked with PM10-2.5 exposures. 
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). Although most of 
these studies examined PM10-2.5 health effect associations 
in urban areas, some studies have also linked mortality and morbidity 
with relatively high ambient concentrations of particles of non-urban 
crustal origin from dust storm events (U.S. EPA, 2019a).
    In considering the available evidence, the Administrator recognizes 
that the evidence continues to provide support for maintaining a 
standard that provides some measure of protection against exposures to 
PM10-2.5, regardless of location, source of origin, or 
particle composition, consistent with previous reviews (78 FR 3176, 
January 15, 2013; 85 FR 82726, December 18, 2020). Drawing from the 
evidence evaluated in the 2019 ISA and consideration of the scientific 
evidence in the PA, the Administrator notes that, consistent with 
previous reviews, the 2019 ISA and the PA highlight a number of 
uncertainties associated with the evidence, including those related to 
PM10-2.5 exposure estimates used in epidemiologic studies, 
in the independence of PM10-2.5 health effect associations, 
and in the 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 places the most weight on evidence supporting 
``causal'' and ``likely to be causal'' relationships. In so doing, he 
notes that the available evidence for short-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 recognizes that, because of the 
uncertainties and limitations in the evidence base, the PA does not 
include a quantitative assessment of

[[Page 5637]]

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, specifically for 
cardiovascular effects, respiratory effects, cancer, and mortality and 
short-term exposures to PM10-2.5, and the lack of available 
quantitative assessments, the Administrator judges 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. Furthermore, the 
Administrator recognizes that the 2019 ISA also concludes that the 
evidence supports ``suggestive of, but not sufficient to infer'' causal 
relationships for long-term PM10-2.5-exposures and 
cardiovascular effects, metabolic effects, nervous system effects, 
cancer, and mortality. However, in considering the available evidence 
for long-term PM10-2.5 exposures, he notes 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 
concludes 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 proposed conclusions on adequacy of the current primary 
24-hour PM10 standard, the Administrator also considers 
advice from the CASAC. As noted above, the CASAC recognizes 
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). Given these uncertainties, the CASAC agrees with the PA 
conclusion that the scientific evidence does not call into question the 
adequacy of the primary PM10 standard and supports 
consideration of retaining the current standard, noting that ``[t]he 
CASAC supports this decision'' (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 afford by the current PM10 standard'' (Sheppard, 
2022a, p. 4 of consensus letter).
    When the above information is taken together, the Administrator 
proposes to conclude that the available scientific evidence continues 
to support a PM10 standard to provide some measure of 
protection against PM10-2.5 exposures. This proposed 
conclusion reflects the available evidence for PM10-2.5-
related health effects, for both short and long-term exposure, as 
evaluated in the 2019 ISA. However, he also recognizes that important 
limitations in the evidence remain. Consistent with the decisions in 
previous reviews, the Administrator proposes 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, as described above, and his 
consideration of CASAC advice on the primary PM10 standard, 
the Administrator proposes to retain the current standard, without 
revision. The Administrator solicits comments on this proposed 
decision.
    Having reached the proposed decision described here based on the 
interpretation of the PM10-2.5-related health effects 
evidence, as evaluated in the 2019 ISA; the evaluation of policy-
relevant aspects of the evidence in the PA; the advice and 
recommendations from the CASAC; public comments received to date in 
their reconsideration; and the public health policy judgments described 
above, the Administrator recognizes that other interpretations, 
assessments and judgments might be possible. Therefore, the 
Administrator solicits comment on the array of issues associated with 
reconsideration of the primary 24-hour PM10 standard, 
including public health and science policy judgments inherent in his 
proposed decision, as described above, and the rationales upon which 
such views are based.

IV. Communication of Public Health

A. Air Quality Index Overview

    Information on the public health implications of ambient 
concentrations of criteria pollutants is made available primarily by 
Air Quality Index (AQI) reporting through the EPA's AirNow 
website.\107\ The current AQI has been in use since its inception in 
1999.\108\ 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. In 
fact, many countries have created similar indices based on the AQI.
---------------------------------------------------------------------------

    \107\ See http://www.airnow.gov/.
    \108\ 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 index 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 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 
evaluated in the ISA\109\ 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, all of which

[[Page 5638]]

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.
---------------------------------------------------------------------------

    \109\ In some NAAQS reviews, there may also be an ISA Supplement 
or a Provisional Assessment of scientific evidence that becomes 
available during a review after an ISA is finalized. To the extent 
that such evidence can inform decisions on the AQI, that information 
is also considered.
---------------------------------------------------------------------------

    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 states 
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 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 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). The EPA and state, local and 
Tribal agencies recognize that these programs are interrelated with AQI 
reporting and with the information related to the effects of air 
pollution on public health that is evaluated through the periodic 
review, and revision when appropriate, of the NAAQS.

B. Air Quality Index Category Breakpoints for PM2.5

    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 AQI at higher PM2.5 
concentrations, the EPA proposes to make two sets of changes to the 
PM2.5 sub-index of the AQI. First, the EPA proposes 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 consistent with the levels of the primary 
PM2.5 annual and 24-hour standards and proposes to revise 
the lower breakpoints to be consistent with any changes to the primary 
PM2.5 standards that are part of this reconsideration. 
Second, the EPA proposes 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. Thus, the EPA considers it appropriate to consider scientific 
evidence for these purposes beyond the scope of the ISA. More details 
on these proposed revisions to the AQI are provided below.
    Although revisions of the air quality criteria and NAAQS for PM 
generally prompt changes to the AQI, the AQI is not part of the NAAQS. 
The AQI is aimed at communicating risks of ambient concentrations which 
may far exceed the level of the NAAQS. While the AQI was not originally 
developed to be used as a regulatory tool or for other purposes and EPA 
does not provide guidance on the use of the AQI for such purposes, the 
EPA acknowledges that some organizations and entities have identified 
other uses for the AQI.\110\ As such, the EPA is requesting information 
about how other organizations and entities are applying the AQI. The 
EPA's goal is to update the PM2.5 AQI in conjunction with 
the Agency's final decisions on the primary annual and 24-hour 
PM2.5 standards, if proposed revisions to such standards are 
promulgated.
---------------------------------------------------------------------------

    \110\ For example, the Occupational Safety and Health divisions 
in California, Oregon, and Washington have linked outdoor worker 
regulations to the upper AQI breakpoints.
---------------------------------------------------------------------------

1. Air Quality Index Values of 50, 100 and 150
    With respect to the lower AQI breakpoints, the EPA concludes that 
it is still appropriate to continue to set 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 that (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 sees no basis for 
deviating from this approach in this reconsideration. Thus, the EPA 
proposes 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. In this document, the EPA is proposing to 
revise the primary annual PM2.5 standard level to 9 to 10 
[mu]g/m\3\ and soliciting comments on levels down to 8 [mu]g/m\3\ and 
up to 11 [mu]g/m\3\ (section II.D.3.a).
    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 which 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 sees no basis to deviate from this approach in this 
reconsideration. In this proposal, the EPA is proposing to retain the 
current primary 24-hour PM2.5 standard with its level of 35 
[mu]g/m\3\ but is taking comment on revising the level of that standard 
to 25 [mu]g/m\3\ (section II.D.3.b). Thus, the EPA proposes 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), but if the level of the 24-hour

[[Page 5639]]

standard is revised to a different concentration, the EPA is proposing 
to set the final AQI value of 100 equal to any revised level of the 
primary 24-hour PM2.5 standard.
    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 sees no basis to deviate from this approach in this 
reconsideration. Therefore, the EPA proposes to set an AQI value of 150 
proportionally, depending on the breakpoint concentration of the AQI 
value of 100. This means that if the EPA retains the current primary 
24-hour PM2.5 standard of 35 [mu]g/m\3\, we propose to also 
retain the current AQI value of 150 at a daily (i.e., 24-hour average) 
concentration of 55 [mu]g/m\3\. If, however, the EPA revises the level 
of the primary 24-hour PM2.5 standard, we propose to adjust 
the AQI value of 150 proportional to that revision (e.g., a 24-hour 
standard of 30 [mu]g/m\3\ might result in an AQI value for 150 of 45 
[mu]g/m\3\).
2. Air Quality Index Values of 200 and Above
    In 1999, the EPA established AQI breakpoints for the AQI values of 
200 and above (64 FR 42530, August 4, 1999). For this approach 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.\111\ It was found that this linear relationship was 
generally consistent with the health effect 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.\112\ Due to limited ambient PM2.5 monitoring data 
available at that time, the decision on the 500 breakpoint 
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). However, the assumption of approximate 
equivalence between the British Smoke method and the current 
PM2.5 monitoring method is not consistent with the view 
cited in the 1987 Federal Register document about the PM10 
AQI value of 500, in which the British Smoke method was noted to have a 
particle size cutpoint of 4.5 microns (52 FR 24688, July 1, 1987). 
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.
---------------------------------------------------------------------------

    \111\ 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).
    \112\ The current AQI value of 500 for PM10 was set 
in 1987 at the concentration of 600 [mu]g/m\3\ based on a 24-hour 
average, on the basis of increased mortality associated with 
historical wintertime pollution episodes in London (52 FR 24687 to 
24688, July 1, 1987). Particle concentrations during these episodes, 
measured by the British Smoke method, were in the range of 500 to 
1000 [mu]g/m\3\. In the 1987 rulemaking that established the upper 
bound index value for PM10, the EPA cited a generally 
held opinion that the British Smoke method measures PM with a 
cutpoint of approximately 4.5 microns (52 FR 24688, July 1, 1987). 
In establishing this value for PM10, the EPA assumed that 
concentrations of PM10, which includes both coarse 
(PM10-2.5) and fine particles (PM2.5), during 
episodes of concern, would be about 100 [mu]g/m\3\ higher than the 
PM concentration measured in terms of British Smoke (52 FR 24688, 
July 1, 1987). The PM10 upper bound index value of 600 
[mu]g/m\3\ was developed by selecting the lower end of the range of 
concentrations during the historical wintertime pollution episodes 
in London (500 [mu]g/m\3\) and adding a margin of 100 [mu]g/m\3\ to 
account for this measurement difference.
---------------------------------------------------------------------------

    As part of this reconsideration, the EPA recognizes that the health 
effects evidence associated with PM2.5 exposure has greatly 
expanded in recent years. 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 studies, specifically controlled human exposure studies, 
that can provide information about health effects at concentrations 
well above the standard levels. Additionally, there are also studies 
now available and evaluated in other Agency documents that can inform 
health effects at higher PM2.5 concentrations. Thus, the EPA 
concludes that it is appropriate to reevaluate the upper AQI 
breakpoints, taking into account the expanded body of scientific 
evidence. In particular, because these breakpoints were established in 
1999 (64 FR 42530, August 4, 1999), several new epidemiologic studies 
have become available that provide information about exposures during 
high pollution events, such as wildfires. Additionally, multiple 
controlled human exposure studies have become available that provide 
information about health effects across a range of concentrations. 
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. For example, while human exposure studies 
generally report only subclinical effects, the consistent observation 
of these effects in multiple studies can provide an indication of 
subclinical effects that are on the pathway to more serious health 
effects as PM2.5 concentrations increase above 55 [micro]g/
m\3\. 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 concludes 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. Thus, the discussion below focuses on the 
EPA's proposed revisions to the AQI breakpoints of 200 and above and 
the EPA's interpretation of the available health effects evidence that 
supports those proposed revisions.
    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

[[Page 5640]]

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 associated with 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. (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), giving the EPA increased confidence in the association between 
exposure to PM2.5 and respiratory-related outcomes at 
concentrations experienced during this time period. 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 proposes 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\. This concentration is at the lower 
end of the concentrations consistently shown to be associated with 
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.\113\ 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.
---------------------------------------------------------------------------

    \113\ Although participants in Lucking et al. (2011) were 
exposed to 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 above, the EPA proposes 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 and epidemiologic studies 
discussed above. Specifically, the Agency proposes to set an AQI value 
of 300 at a daily (i.e., 24-hour average) PM2.5 
concentration of

[[Page 5641]]

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 can 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).
    Lastly, the EPA is also proposing 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, 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 considers several recent controlled human exposure 
studies that observe health effects which are clearly associated with 
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. 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.\114\ Also as discussed above, Lucking et al. 
(2011) exposed healthy subjects to DE with a mean PM2.5 
concentration of 320 [micro]g/m\3\ for 1 hour.\115\ The types of 
cardiovascular effects observed in these controlled human exposure 
studies have been linked with the exacerbation of ischemic heart 
disease (IHD) and heart failure as well as myocardial infarction (MI) 
and stroke.
---------------------------------------------------------------------------

    \114\ These effects were attenuated when the DE was filtered, to 
reduce PM2.5 concentrations, indicating the effects were 
likely associated with PM2.5 exposure.
    \115\ 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 
above, recent epidemiologic studies examining the relationship between 
wildfire smoke and respiratory health can also inform proposed 
decisions on the concentration for the AQI value of 500. As noted 
earlier in this section, Hutchinson et al. (2018) reported 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 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. For example, Orr et al. (2020) conducted a 
longitudinal study that examined whether exposure to wildfire smoke 
from a multi-month fire resulted in respiratory effects in subsequent 
years. The authors conducted respiratory health assessments of adults 
living in Seeley Lake and Thompson Falls, MT, during the 3-month summer 
wildfire event that occurred in 2017 as well as follow-up visits in 
each of the two years following the wildfire (Orr et al., 2020). 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\. This study 
included full spirometry tests for all study participants during the 
initial 2017 visit and again in 2018 and 2019 to assess lung function 
and reported that the average FEV1/FVC (forced expiratory 
volume in 1 second/forced vital capacity) decreased significantly in 
2018 (71.6% observed; 77.35% predicted) and 2019 (73.4% observed; 
76.52% predicted) (Orr et al., 2020). This study suggests that exposure 
to high PM2.5 concentrations during a multi-week fire event 
may lead to long-term health consequences in the future, such as 
declines in lung function.
    The controlled human exposure studies provide biological 
plausibility for increases in respiratory-related health care events 
during the wildfires documented in epidemiologic studies. 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 to inform proposed 
decisions on the level of the AQI value of 500. Given the 
concentrations observed in these studies, the Agency proposes 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).
3. Summary
    Table 1 below summarizes the proposed breakpoints for the 
PM2.5 sub-index.

                                Table 1--Proposed Breakpoints for PM2.5 Sub-Index
----------------------------------------------------------------------------------------------------------------
                                                                                  Current
                                                                                breakpoints        Proposed
                        AQI category                           Index values    ([mu]g/m\3\,       breakpoints
                                                                                 24- hour      ([mu]g/m\3\, 24-
                                                                                 average)        hour average)
----------------------------------------------------------------------------------------------------------------
Good........................................................            0-50        0.0-12.0      0.0-(9.0-10.0)

[[Page 5642]]

 
Moderate....................................................          51-100       12.1-35.4     (9.1-10.1)-35.4
Unhealthy for Sensitive Groups..............................         101-150       35.5-55.4           35.5-55.4
Unhealthy...................................................         151-200      55.5-150.4          55.5-125.4
Very Unhealthy..............................................         201-300     150.5-250.4         125.5-225.4
Hazardous \1\...............................................            301+           250.5               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.

    As discussed above, the EPA recognizes that the health effects 
evidence associated with PM2.5 exposure has greatly expanded 
in recent years and concludes that the body of scientific evidence 
supports the need to revise many of the AQI breakpoints. This is 
particularly true of the AQI values of 200 and above, where the EPA 
concludes that the available controlled human exposure and 
epidemiologic studies support offering advisories to the public for 
reducing exposures at lower concentrations than recommended with the 
current AQI breakpoints. However, the EPA also recognizes that there 
are interpretations and judgments that must be applied in making the 
determinations of these breakpoints. Thus, the EPA is soliciting 
comment on the proposed revisions to the AQI described above. In 
particular, for the AQI values of 50, 100 and 150, the EPA is 
soliciting comment on the proposed decision to continue to use the 
approach used in AQI revisions in 2012 (77 FR 38890, June 29, 2012) of 
setting the lower breakpoints (50, 100, and 150) to be consistent with 
the levels of the primary annual and 24-hour PM2.5 standards 
and proposed decision to revise the lower breakpoints to be consistent 
with any changes to the primary PM2.5 standards that are 
part of this reconsideration. With respect to the AQI values of 200 and 
above, the EPA is soliciting comment on the proposed decision to revise 
those AQI values, as well as comment on the approach being applied, the 
health studies viewed as most relevant in these proposed decisions, and 
the proposed AQI breakpoint concentrations. The EPA also notes that 
while the newer studies do provide more refined information about the 
concentration range in which health effects might occur, the evidence 
continues to support a continuum of effects in concentration exposures 
in the range of those defined by the upper AQI values, with increasing 
PM2.5 concentrations being associated with increasingly 
larger numbers of people likely experiencing serious health effects. 
Given this, the EPA is also soliciting comment on maintaining the 
linear relationship approach used to set the upper AQI values in 1999 
but using a different linear relationship (64 FR 42530, August 4, 
1999). For example, the EPA could set the AQI value of 150 based on the 
primary NAAQS and the AQI value of 300 (which is the breakpoint that 
identifies the starting concentration for the highest AQI category) 
based on the considerations discussed above and using those values to 
develop a linear relationship for the AQI values for 200 and 500. Under 
this approach, if the AQI breakpoint for 150 is set at 55.4 [mu]g/m\3\ 
and the AQI breakpoint for 300 is set at 225.4 [mu]g/m\3\, the AQI 
breakpoint for 200 would be 112.4 [mu]g/m\3\ and the AQI breakpoint for 
500 would be 452.4 [mu]g/m\3\. The EPA solicits comments on whether to 
use a linear approach for higher breakpoints, the appropriate 
breakpoints to use for such an approach, and the appropriate values for 
breakpoints under other approaches, falling within the range of the 
current breakpoints and the breakpoints identified by these various 
approaches, as well as to retain and not change the existing 
breakpoints at this time.

C. Air Quality Index Category Breakpoints for PM10

    The EPA proposes 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 above.

D. Air Quality Index Reporting

    With respect to the reporting requirements for the AQI, 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.
    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 is 
proposing that state, local, and Tribal agencies that report the AQI be 
required to report it seven days a week, ensuring that the public 
continues 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. Sub-daily 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

[[Page 5643]]

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 proposes to recommend that state, local, 
and Tribal agencies report the AQI in near-real time. Like air quality 
forecasting, which also allows the public to make health-protective, 
near-real time AQI reporting is recommended but not required.
    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 uses this near-real-time data in the 
National, Interactive and Fire and Smoke maps on the AirNow website, 
and to create products for use by weather service providers and the 
media. Some state, local, and Tribal air quality agencies also use 
these products on their own websites and in their own applications 
(i.e., the California Air Resources Board uses the data in its 
California Smoke Spotter application). To continue to ensure the 
availability of the products that the public and many stakeholders rely 
upon, the EPA is proposing 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.
    The Agency is updating the reporting requirements and near-real-
time reporting and data submission recommendations for the AQI. The 
Agency is reformatting the question-and-answer format used in appendix 
G to align with the current standard formatting used in the Code of 
Federal Regulations. The EPA is not taking comment on or reopening the 
language that has merely been moved or rearranged as there are no 
substantive changes.
    Another change the EPA is proposing to make to appendix G is with 
regard to Table 2- Breakpoints for the AQI for purposes of clarity. We 
are proposing 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. This breakpoint of 
400, along with those for 200 and 300, were defined and are the 
historical basis for the Alert, Warning, and Emergency episode levels 
included in 40 CFR part 51, appendix L, as part of the Prevention of 
Air Pollution Emergency Episodes program (44 FR 92, May 10, 1979). 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 is proposing 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 plans 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 will be moved from appendix 
G to the TAD includes the definitions of the sensitive (at-risk) 
populations for each pollutant. 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 a 
NAAQS review. Data calculations for non-required mathematical 
equations, (i.e., the NowCast), are currently and will continue to be 
included in the TAD. The EPA works with state, local, and Tribal air 
agencies to modify these calculations as needed, which may not be 
associated with a NAAQS review. Also, recognizing that the ways that 
air quality and health information is supplied to the news media and 
public changes regularly, information about suggested approaches will 
be taken out of appendix G and discussed in the TAD.

V. Rationale for Proposed Decisions on the Secondary PM Standards

    This section presents the rationale for the Administrator's 
proposed 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).\116\ This rationale 
is based on a thorough review of the scientific evidence generally 
published through December 2017,\117\ 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 
rationale 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 \118\ (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

[[Page 5644]]

evidence in the 2020 PA.\119\ 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 rationale also takes into 
account: (1) the 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; (2) CASAC 
advice and recommendations, as reflected in discussions of the drafts 
of the ISA Supplement and PA at public meetings and in the CASAC's 
letters to the Administrator; and (3) public comments received during 
the development of these documents.
---------------------------------------------------------------------------

    \116\ 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 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).
    \117\ 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.
    \118\ 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).
    \119\ 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, 2020a). 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 proposed 
decision and its foundations, section V.A provides background and 
introductory information for this reconsideration of the secondary PM 
standards. It includes background on the 2020 final decision to retain 
the secondary PM standards (section V.A.1) and also describes the 
general approach for this reconsideration (section V.A.2). Section V.B 
summarizes the key aspects of the currently available evidence and 
quantitative information for PM-related visibility impairment and 
section V.C summarizes the available information for other PM-related 
welfare effects. Section V.D presents the Administrator's proposed 
conclusions on the current secondary PM standards (V.D.III), drawing on 
both evidence- and quantitative information-based considerations 
(section V.D.1) and advice from the CASAC (V.D.2).

A. General Approach

    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 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 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 
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 will 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. 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.
    The subsections below provide background and introductory 
information. Background on the 2020 decision to retain the current 
standards, including the rationale for that decision, for non-
visibility effects and visibility effects is summarized in sections 
V.A.1.a and V.A.1.b below, respectively. This is followed, in section 
V.A.2, by an overview of the general approach for the reconsideration 
of the 2020 final decision. Following this introductory section and 
subsections, the subsequent sections summarize current information and 
analyses, including that newly available in this reconsideration. The 
Administrator's proposed conclusions on the secondary PM standards, 
based on the current information, are provided in section V.D.3.
1. Background on the Current Standards
    The current secondary PM standards were affirmed in 2020 based on 
the scientific and technical information available at that time, as 
well as the 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 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 Administrator's conclusions, 
for climate and materials effects (section V.A.1.a) and visibility 
effects (section V.A.2.b) in the 2020 review.
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

[[Page 5645]]

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.
    Consistent with the general approach routinely employed in NAAQS 
reviews, the initial consideration in the 2020 review of the secondary 
standards was with regard to the adequacy of protection provided by the 
existing standards. Key aspects of the consideration are summarized in 
section V.A.1.a.i below.
i. Considerations Regarding Adequacy of the Existing Standards for Non-
Visibility Effects in the 2020 Review
    In considering non-visibility welfare effects in the 2020 review, 
as discussed above, the 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 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 Administrator 
noted that significant limitations continued to exist related to 
quantifying the contributions of direct and indirect effects of PM and 
PM components on climate forcing (U.S. EPA, 2020a, 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 Administrator to 
conclude that the available scientific information in the 2020 review 
remained insufficient to quantify climate impacts associated with 
particular concentrations of PM in ambient air (U.S. EPA, 2020a, 
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 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, 
2020a, section 13.4). Additionally, the information assessed in the 
2019 ISA did not support quantitative analyses of PM-related materials 
effects in the 2020 review (U.S. EPA, 2020a, 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 the 2020 review, the CASAC agreed with the 2020 PA 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 the 2020 review, 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 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 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.
    Consistent with the general approach routinely employed in NAAQS 
reviews, the initial consideration in the 2020 review of the secondary 
PM standards was with regard to the adequacy of the protection provided 
by the then-existing standards. Key aspects of that consideration are 
summarized in section V.A.1.b.i below.
i. Consideration Regarding the Adequacy of the Existing Standards for 
Visibility Effects in the 2020 Review
    In considering the visibility effects in the 2020 review, the 
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

[[Page 5646]]

impairment was from studies from four areas in North America.\120\ 
These studies provided information to inform our understanding of 
levels of visibility impairment that the public judged to be 
``acceptable'' (U.S. EPA, 2010a; 85 FR 24131, April 30, 2020). In 
considering these public preference studies, the Administrator noted 
that, as described in the 2019 ISA, no new visibility studies had been 
conducted in the U.S. 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.
---------------------------------------------------------------------------

    \120\ 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 
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 
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 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 and PM10-2.5, along with estimates of 
relative humidity. The Administrator recognized that, while there have 
been some revisions to the IMPROVE algorithm since the time of the 2012 
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 
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 
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 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 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 Administrator judged that it was appropriate to 
establish a target level of protection of 30 deciviews 
(dv),121 122 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 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, with 
their inherent uncertainties and limitations, the prior 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).
---------------------------------------------------------------------------

    \121\ 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.
    \122\ 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 available in the 
2020 review, the Administrator's judgments were based on the same 
studies, with the same range of levels, available in the 2012 review. 
As identified in the 2020 PA (U.S. EPA, 2020a, section 5.5), there were 
a number of limitations and uncertainties associated with these 
studies, including the following:
     Available studies may not represent the full range of 
preferences for visibility in the U.S. population, particularly given 
the potential variability in preferences based on the conditions 
commonly encountered and the scenes being viewed.
     Available preference studies were conducted 15 to 30 years 
ago and may

[[Page 5647]]

not accurately represent the current day preferences of people in the 
U.S.
     The variety of methods used in the preference studies may 
potentially influence the responses as to what level of impairment is 
deemed acceptable.
     Factors that are not captured in the methods of the 
preference studies, such as the time of day when light extinction is 
the greatest or the frequency of impairment episodes, may influence 
people's judgment on acceptable visibility (U.S. EPA, 2020a, section 
5.2.1.1).
    Therefore, 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 
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 Administrator considered the updated analyses 
of visibility impairment presented in the 2020 PA (U.S. EPA, 2020a, 
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, 2020a, 
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, 2020a, 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, 2020a, section 5.2.1.2). While 
the inclusion of the coarse fraction had a relatively modest impact on 
the estimates of light extinction, the 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, 2020a, section 5.2.1.2). He 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 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. General Approach and Key Issues in This Reconsideration of the 2020 
Final Decision
    To evaluate whether it is appropriate to consider retaining the 
current secondary PM standards, or whether consideration of revision is 
appropriate, the EPA has adopted an approach in this reconsideration 
that builds upon the general approach used in past reviews and reflects 
the body of evidence and information now available. Accordingly, the 
approach in this reconsideration takes into consideration the 
approaches used in past reviews, including the substantial assessments 
and evaluations performed in those reviews, and also takes into account 
the more recent scientific information and air quality data now 
available to inform understanding of the key policy-relevant issues in 
the reconsideration. As summarized above, the Administrator's decisions 
in the 2020 review were based on an integration of PM welfare effects 
information with the judgments on the public welfare significance of 
key effects, policy judgments as to when the standard is requisite, 
consideration of CASAC advice, and consideration of public comments.
    Similarly, in this reconsideration, we draw on the current 
information from studies of PM-related visibility effects, quantitative 
analyses of PM-related visibility impairment, and information from 
studies of non-visibility welfare effects. In so doing, we consider 
both the information available at the time of the 2012 and 2020 reviews 
and information more recently available, including that which has been 
critically analyzed and characterized in the 2019 ISA and ISA 
Supplement \123\ for visibility, climate, and materials effects. The 
evaluations in the PA, of the potential implications of various aspects 
of the scientific evidence in the 2019 ISA and ISA Supplement (building 
on prior such assessments), augmented by the quantitative air quality, 
exposure or risk-based information, are also considered along with the 
associated uncertainties and limitations.
---------------------------------------------------------------------------

    \123\ As noted above and described in detail in section 1.4.2 of 
the PA, 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. The selection of 
the 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. 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.
---------------------------------------------------------------------------

B. 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 PA. 
While the 2019 ISA provides the broad scientific foundation for this 
reconsideration, we recognize that 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 PA, the selection of welfare effects to 
evaluate

[[Page 5648]]

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.
1. 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).
a. Visibility
    Visibility refers to the visual quality of a human's view with 
respect to color rendition and contrast definition. It is the ability 
to perceive landscape form, colors, and textures. Visibility involves 
optical and psychophysical properties involving human perception, 
judgment, and interpretation. Light between the observer and the object 
can be scattered into or out of the sight path and absorbed by PM or 
gases in the sight path. Consistent with conclusions of causality in 
the 2012 and 2020 reviews, the 2019 ISA concludes 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). These 
conclusions are based on the strong and consistent evidence that 
ambient PM can impair visibility in both urban and remote areas (U.S. 
EPA, 2019a, section 13.1; U.S. EPA, 2009a, section 9.2.5).
    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.\124\ 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).
---------------------------------------------------------------------------

    \124\ 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., 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

[[Page 5649]]

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,\125\ 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).
---------------------------------------------------------------------------

    \125\ 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 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.\126\ The detailed methodology for these studies 
are described in the 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, 2010a, p. 4-24; U.S. EPA, 2020a, 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).
---------------------------------------------------------------------------

    \126\ 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.\127\ 
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.\128\ 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\129\ 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).
---------------------------------------------------------------------------

    \127\ 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 PA (U.S. EPA, 2022b, Appendix D).
    \128\ 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 PA (U.S. EPA, 2022b, Appendix D).
    \129\ 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 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

[[Page 5650]]

them as an alternative to the light extinction approach.
b. 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 \130\ 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 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.
---------------------------------------------------------------------------

    \130\ 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 PA (U.S. EPA, 2022b, section 5.3.2.1.1), 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 which 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). However, 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.
c. 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.\131\ 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).
---------------------------------------------------------------------------

    \131\ 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 PA (U.S. EPA, 2022b, section 5.3.2.1.2) 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 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.

C. Summary of Air Quality and Quantitative Information

    Beyond the consideration of the scientific evidence, as discussed 
in section V.B 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.
1. 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

[[Page 5651]]

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 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 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 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.
a. Target Level of Protection in Terms of a PM2.5 Visibility 
Index
    In evaluating the adequacy of the current secondary PM standards, 
the PA first evaluates the appropriateness of the elements (indicator, 
averaging time, form, and level) identified for a distinct secondary 
standard to protect against visibility effects. In previous reviews, 
the visibility index was 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 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 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 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 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 PA concludes that it is appropriate to continue to 
focus on daily visibility impairment. In so doing, the PA relies on 
analyses that were conducted in the 2012 review that showed relatively 
strong correlations between 24-hour and sub-daily (i.e., 4-hour 
average) PM2.5 light extinction that indicated that a 24-
hour averaging time is an appropriate surrogate for the sub-daily 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 PA also notes that the 24-hour 
averaging time may be less influenced by atypical conditions and/or 
atypical instrument performance than a sub-daily 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 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 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, 2010a, 
Chapter 4). In considering this evaluation of statistical forms from 
the 2010 UFVA, consistent with the 2011 PA, the 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 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 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 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) 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.
    Therefore, the PA continues to rely on the same studies \132\ 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 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 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

[[Page 5652]]

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.
---------------------------------------------------------------------------

    \132\ 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 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 24.3 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 
\133\ 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.
---------------------------------------------------------------------------

    \133\ 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 picturesque features 
of the Washington, DC, cityscape are buildings relatively nearby 
without prominent and/or values 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 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.
b. Relationship Between the PM2.5 Visibility Index and the 
Current Secondary 24-Hour PM2.5 Standard
    The 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

[[Page 5653]]

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 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.\134\ 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; 
\135\ (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).
---------------------------------------------------------------------------

    \134\ The analyses presented in the 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).
    \135\ 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 sub-daily visibility 
impairment estimates. As detailed in the 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 sub-daily 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 sub-daily 
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 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 \136\ 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).\137\
---------------------------------------------------------------------------

    \136\ 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.
    \137\ 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,\138\ 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,\139\ and those from previous 
reviews.
---------------------------------------------------------------------------

    \138\ As described in more detail in the 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).
    \139\ 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).\140\ 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.
---------------------------------------------------------------------------

    \140\ 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 [mu]g/m\3\ (see U.S. EPA, 2022b, Appendix D, 
Table D-3).
---------------------------------------------------------------------------

    Nevertheless, when light extinction is calculated using the 
Lowenthal and 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 [mu]g/m\3\ was attained.
    In considering visibility impairment under recent air quality 
conditions, the 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 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 
for converting OC to organic matter (OM) was more appropriate, for

[[Page 5654]]

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 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, 2010a) 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 review addressed light extinction at 20 of the 67 PM2.5 
sites where collocated PM10-2.5 monitoring data were 
available. Since the 2020 review, PM10-2.5 monitoring data 
are available at more locations and the analyses presented in the 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, 2020a, section 5.2.1.2). 
However, the 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.
2. Non-Visibility Effects
    Consistent with the evidence available at the time of the 2012 and 
2020 reviews, and as described in detail in the 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 PA (U.S. EPA, 
2022b, section 5.3.2.1.1), 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 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).
    For PM-related materials effects, as explained in more detail in 
the PA (U.S. EPA, 2022b, section 5.3.2.1.2), 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

[[Page 5655]]

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).

D. Proposed Conclusions on the Secondary PM Standards

    In reaching proposed conclusions on the current secondary PM 
standards (presented in section IV.D.3), the Administrator has taken 
into account policy-relevant evidence- and quantitative information-
based considerations discussed in the PA (summarized in section 
IV.D.2), as well as advice from the CASAC and public comment on the 
standards received thus far in the reconsideration (section IV.D.1). In 
general, the role of the PA is to help ``bridge the gap'' between the 
Agency's assessment of the current evidence and quantitative analyses, 
and the judgments required of the Administrator in determining whether 
it is appropriate to retain or revise the NAAQS. Evidence-based 
considerations draw upon the EPA's integrated assessment of the 
scientific evidence of PM-related welfare effects presented in the 2019 
ISA and ISA Supplement (summarized in section V.B above) to address key 
policy-relevant questions in the reconsideration. Similarly, the 
quantitative information-based considerations (summarized in section 
V.C above) focused on the potential for PM-related welfare effects 
under recent air quality conditions for the purposes of addressing the 
policy-relevant questions.
    This approach to reviewing the secondary standards is consistent 
with the 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 secondary 
standards that, in the Administrator's judgment, are requisite (i.e., 
neither more nor less stringent than necessary) to protect the public 
welfare from known or anticipated adverse effects associated with the 
presence of the pollutant in ambient air. 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 welfare 
effects evidence generally reflects a continuum that includes ambient 
air exposures for which scientists generally agree that 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 secondary standards 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 proposed decision on the adequacy of the current secondary 
standards described below is a public welfare policy judgment by the 
Administrator that draws upon the scientific evidence for the relevant 
welfare effects, quantitative analyses of air quality, as available, 
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 welfare protection afforded by the current standard against PM-
related visibility, climate and materials effects. The Administrator's 
final decision will additionally consider public comments received on 
this proposed decision.
1. CASAC Advice in This Reconsideration
    The CASAC provided its advice regarding the current secondary 
standards in the context of its review of the draft PA (Sheppard, 
2022a).\141\ In its comments on the draft PA, the CASAC first 
recognized the scientific evidence is sufficient to support a causal 
relationship between PM and visibility effects, climate effects and 
materials effects.
---------------------------------------------------------------------------

    \141\ A limited number of public comments have also been 
received in this reconsideration to date, including comments focused 
on the draft PA. Of those public comments that addressed the 
adequacy of the secondary PM standards, the majority of commenters 
support the preliminary conclusion that it is appropriate to 
consider retaining the current secondary PM standards, without 
revision. These commenters generally cite to a lack of newly 
available evidence and information that would inform consideration 
of alternative secondary PM standards to protect against PM-related 
effects on visibility, climate, and materials. One commenter, 
however, supported the revision of the secondary PM standards to 
provide additional protection against PM-related visibility 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).
    With regard to climate and materials effects, the CASAC noted that 
substantial uncertainties remain in the scientific evidence for these 
effects. The CASAC suggested a number of areas for future research to 
further inform our understanding of these effects, including more 
climate-related research and research that would allow for quantitative 
assessment of the relationship between materials effects and PM in 
ambient air.
2. Evidence- and Quantitative Information-Based Considerations in the 
Policy Assessment
    The secondary PM standards include the 24-hour PM2.5 
standard, with its level of 35 [mu]g/m\3\ as the 98th percentile, 
averaged over three years; the annual PM2.5 standard, with 
its level of 15.0 [mu]g/m\3\ as the annual mean, averaged over three 
years; and the 24-hour PM10 standard, with its level of 150 
[mu]g/m\3\, not to be exceeded more than once per year on average over 
three years. Together, these standards provide protection against both 
long-term average and short-term peak PM concentrations. For example, 
the 24-hour PM2.5 standard is most effective at limiting 
peak 24-hour PM2.5 concentrations, but in doing so, also has 
an effect on annual average PM2.5 concentrations. 
Additionally, the annual standard is most effective in controlling 
``typical'' or average PM2.5 concentrations, but also 
provides some

[[Page 5656]]

measure of protection against peak exposures.
    The PA considers the degree to which the available scientific 
evidence and quantitative information supports or calls into question 
the adequacy of the protection afforded by the current secondary PM 
standards. In doing so, the PA considers the evidence assessed in the 
2019 ISA and ISA Supplement, including the extent to which the evidence 
for PM-related visibility impairment, climate effects, or materials 
effects alters key conclusions from the 2020 review. The PA also 
considers quantitative analyses of visibility impairment and the extent 
to which they may indicate different conclusions from those in the 2020 
review regarding the degree of protection from adverse effects provided 
by the current secondary standards.
    Consistent with the approaches used in previous reviews, the 
quantitative analyses in the PA utilized a two-phase assessment for 
visibility impairment. First, the PA considered the appropriateness of 
the elements (indicator, averaging time, form, and level) of the 
visibility index for providing protection against PM-related visibility 
effects. Second, the PA evaluated the relationship between the current 
secondary 24-hour PM2.5 standard and the visibility index.
    With regard to the appropriateness of the visibility index and its 
target level of protection against PM-related visibility effects, the 
PA notes that there is limited information available in this 
reconsideration beyond that available in previous reviews to inform 
conclusions on the elements (indicator, averaging time, form, and 
level) of the visibility index (described in more detail in section 
V.C.1.a above). In considering the available information, the PA 
concludes that the 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.
    With regard to the relationship between the current secondary 24-
hour PM2.5 standard and the visibility index, the PA 
presents updated analyses based on recent air quality information, with 
a focus on locations meeting the current secondary 24-hour 
PM2.5 and PM10 standards. In the absence of 
advances in the monitoring methods for directly measuring light 
extinction, and given the lack of a robust monitoring network for the 
routine measurement of light extinction across the U.S. (section 
V.B.1.a), as in previous reviews, the PA analyses use calculated light 
extinction to estimate PM-related visibility impairment (U.S. EPA, 
2022b, section 5.3.1.2). Compared to the 2012 review, updated analyses 
incorporate several refinements. These include (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; 
\142\ (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 PA's updated analyses include 60 
monitoring sites that measure PM2.5 and PM10 that 
are geographically distributed across the U.S. in both urban and rural 
areas (U.S. EPA, 2022b, Appendix D, Figure D-1).\143\
---------------------------------------------------------------------------

    \142\ 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 sub-daily visibility 
impairment estimates. Therefore, the inputs to these light 
extinction calculations are based on 24-hour average measurements of 
PM2.5 mass and components, rather than sub-daily 
information.
    \143\ These sites are those that have a valid 24-hour 
PM2.5 design value for the 2015-2017 period and met 
strict criteria for PM species for this analysis, based on 24-hour 
average PM2.5 and PM10-2.5 mass and component 
data that were available from monitors in the IMPROVE network, CSN, 
and NCore Multipollutant Monitoring Network (U.S. EPA, 2022b, 
Appendix D).
---------------------------------------------------------------------------

    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 using the original and revised IMPROVE 
equations (U.S. EPA, 2022b, section 5.3.1.2). In addition, the four 
locations that exceeds the current 24-hour PM2.5 standard 
have light extinction estimates that range from 22 to 27 dv when using 
the original IMPROVE equation (U.S. EPA, 2022b, Figure 5-3) and from 22 
to 29 dv when using the revised IMPROVE equation (U.S. EPA, 2022b, 
Figure 5-4). The analyses presented in the 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 (the upper end of the range of target levels 
of protection) 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).\144\
---------------------------------------------------------------------------

    \144\ As noted above in section V.1.C.b, 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 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 is calculated using the updated IMPROVE 
equation from Lowenthal and Kumar (2016), the resulting 3-year 
visibility metrics are slightly higher at all sites compared to light 
extinction calculated using the original and revised IMPROVE equations 
(U.S. EPA, 2022b, Figure 5-5). The slightly higher estimates of light 
extinction are consistent with the higher OC multiplier included in the 
IMPROVE equation from Lowenthal and Kumar (2016), reflecting the use of 
data from remote areas with higher concentrations of organic PM when 
validating that equation. As such, it is important to note that the 
Lowenthal and Kumar (2016) version of the IMPROVE equation may 
overestimate light extinction in non-remote areas, including in the 
urban areas included in the analyses presented in the PA.
    Nevertheless, when light extinction is calculated using the 
Lowenthal and 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.\145\ For the sites that exceed the 
current 24-hour PM2.5 standard, three of the 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 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 [mu]g/m\3\ was attained.
---------------------------------------------------------------------------

    \145\ As noted above in section V.1.C.b, 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 [mu]g/m\3\ (see 
U.S. EPA, 2022b, Appendix D, Table D-3).
---------------------------------------------------------------------------

    In 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 PM10-2.5 is often a minor 
contributor to visibility impairment (U.S. EPA,

[[Page 5657]]

2010a), though it may make a larger contribution in some areas in the 
desert southwestern region of the U.S. However, 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 review addressed light extinction at 20 of the 
67 PM2.5 sites where collocated PM10-2.5 
monitoring data were available. Since the 2020 review, 
PM10-2.5 monitoring data are available at more locations and 
the analyses presented in the 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, as assessed and presented in the 2020 PA (U.S. EPA, 2020a, 
section 5.2.1.2). However, the 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).
    In summary, 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 when light 
extinction is calculated using the original and revised IMPROVE 
equations, and generally at or below 28 dv when using the Lowenthal and 
Kumar (2016) equation to estimate light extinction. Small differences 
in the 3-year visibility metric were observed between the variations of 
the IMPROVE equation. When light extinction is calculated using the 
revised IMPROVE equation, there is a generally 1-2 dv at 
the study locations compared to light extinction calculated using the 
original IMPROVE equation (U.S. EPA, 2022b, Appendix D, Table D-3). 
When light extinction is calculated using the Lowenthal and Kumar 
(2016) equation, the difference compared to using either the original 
or revised IMPROVE equation generally ranges from no difference to up 
to 4 dv greater in areas that meet the current secondary 24-hour 
PM2.5 standard (U.S. EPA, 2022b, Appendix D, Table D-3). 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 (U.S. EPA, 2004a; U.S. EPA, 
2010a). Given that there is more variability when estimating light 
extinction using the Lowenthal and Kumar (2016) IMPROVE equation 
compared to the original or revised IMPROVE equations, it is important 
to recognize that the PA notes that the Lowenthal and Kumar (2016) 
equation may not be appropriate for all locations and source types. For 
example, the larger multiplier used in the Lowenthal and Kumar (2016) 
may be more appropriate for estimating light extinction in more remote 
areas where there is more aged and oxygenated organic PM compared to in 
urban areas. As such, the PA recognizes that one version of the IMPROVE 
equation is not necessarily more accurate or precise in estimating 
light extinction, and that differences in locations may support the 
selection of inputs to the IMPROVE equation or of the appropriate 
IMPROVE equation to estimate light extinction on a regional basis 
rather than on a national basis. Overall, regardless of the IMPROVE 
equation that is used to estimate light extinction, in areas that meet 
the current 24-hour PM2.5 standards, the 3-year visibility 
metric is at or below 28 dv, which is in the upper range of levels for 
the target level of protection identified from the public preference 
studies (i.e., 20 to 30 dv). In fact, even in areas that exceed the 
secondary 24-hour PM2.5 standard, and regardless of the 
IMPROVE equation that is used to calculate light extinction, all study 
locations have 3-year visibility index values at or below 30 dv, which 
is the upper end of the range of target levels of protection.
    With regard to PM-related climate effects, the PA recognizes that 
while the evidence base has expanded since the completion of the 2009 
ISA, the recent evidence has not appreciably improved the 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). Despite continuing research, there are still significant 
limitations in quantifying the contributions of PM and PM components to 
the direct and indirect effects on climate forcing (e.g., changes to 
the pattern of rainfall, changes to wind patterns, effects on vertical 
mixing in the atmosphere) (U.S. EPA, 2022b, sections 5.3.2.1.1 and 
5.5). In addition, while a number of improvements and refinements have 
been made to climate models since the 2012 review, these 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 
recent research has added to the understanding of climate forcing on a 
global scale, there remain significant limitations to quantifying 
potential adverse effects from PM on climate in the U.S. and how they 
would vary in response to incremental changes in PM concentrations in 
the U.S. Overall, the PA recognizes that while new research is 
available on climate forcing on a global scale, the remaining 
uncertainties and limitations are significant, and the new global scale 
research does not translate directly to use at regional spatial scales. 
Thus, the evidence does not provide a clear understanding at the 
spatial scales needed for the NAAQS of a quantitative relationship 
between concentrations of PM mass in ambient air and the associated 
climate-related effects (U.S. EPA, 2022b, sections 5.3.2.2.1 and 5.5). 
The PA concludes that the evidence does not call into question the 
adequacy of the current secondary PM standards for climate effects.
    With regard to materials effects, the PA notes the availability of 
recent evidence in this reconsideration related to the soiling process 
and the types of materials that are affected. Such evidence provides 
some limited information to inform dose-response relationships and 
damage functions associated with PM, though most recent studies have 
been conducted outside the U.S. in areas where PM concentrations in 
ambient air are higher than those observed in the U.S. (U.S. EPA, 
2022b, section 5.3.2.1.2; U.S. EPA, 2019a, section 13.4). The recent 
evidence includes studies examining PM-related effects on the energy 
efficiency of solar panels and passive cooling building materials, 
though there remains insufficient evidence 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 recent research has 
expanded the body of evidence for PM-related materials effects, the PA 
recognizes the lack of information to inform quantitative analyses 
assessing

[[Page 5658]]

materials effects or the potential public welfare implications of such 
effects (U.S. EPA, 2022b, section 5.3.2.2.2). Thus, the PA concludes 
that the evidence does not call into question the adequacy of the 
current secondary PM standards for materials effects.
    Overall, the PA recognizes that the newly available welfare effects 
evidence, critically assessed in the 2019 ISA as part of the full body 
of evidence, and visibility effects evidence, assessed in the ISA 
Supplement, reaffirms the conclusions on the visibility, climate, and 
materials effects of PM as recognized in the 2012 and 2020 reviews 
(U.S. EPA, 2022b, sections 5.3.1.1, 5.3.2.1, and 5.5). Further, there 
is a general consistency of the currently available evidence with the 
evidence that was available in previous reviews, including with regard 
to key aspects of the decision to retain the standards in the 2012 and 
2020 reviews (U.S. EPA, 2022b, sections 5.3.1.1, 5.3.2.1, and 5.5). The 
quantitative analyses for visibility impairment for recent air quality 
conditions indicate that estimated light extinction in areas meeting 
the current secondary 24-hour PM2.5 standards have a 3-year 
visibility index at or below 30 dv (i.e., the upper end of the range of 
target levels of protection identified in the 2012 and 2020 reviews) 
and most areas have 3-year visibility index values at or below the 
midpoint of the range of target levels of protection (i.e., 25 dv) 
(U.S. EPA, 2022b, sections 5.3.1.2 and 5.5). Collectively, the PA finds 
that the evidence and quantitative information-based considerations 
support consideration of retaining the current secondary PM standards, 
without revision (U.S. EPA, 2022b, section 5.5).
3. Administrator's Proposed Decision on the Current Secondary PM 
Standards
    This section summarizes the Administrator's considerations and 
conclusions related to the current secondary PM2.5 and 
PM10 standards and presents his proposed 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 will be 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 will be based on an interpretation of the scientific 
evidence and technical analyses that neither overstates nor understates 
their strengths and limitations, nor the appropriate inferences to be 
drawn.
    As an initial matter in considering the secondary standards, the 
Administrator notes the longstanding body of evidence for PM-related 
visibility impairment. As in previous reviews, this evidence continues 
to demonstrate a causal relationship between ambient PM 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. In so doing, and consistent with previous 
reviews, the Administrator considers the protection provided by the 
current secondary standards 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 country. Programs implemented to meet the secondary PM NAAQS, 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.
    In addition, the Administrator notes 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). The Administrator finds that 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, recognizing that the 
Regional Haze Program will continue to contribute to reductions in 
visibility impairment in Class I areas, the Administrator proposes 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.
    In further considering what standards are requisite to protect 
against adverse public welfare effects from visibility impairment, the 
Administrator adopts an approach consistent with the approach used in 
previous reviews (section V.A.1.b). That is, he first identifies 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 
considers 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).
    To identify a 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 recognizes 
that there is a lack of availability of methods

[[Page 5659]]

and an established network for directly measuring light extinction, 
consistent with the conclusions reached in the 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 (Sheppard, 
2022a, p. 21 of consensus responses). He notes that in the 2012 and 
2020 reviews, given the lack of such monitoring data, the EPA used an 
index based on estimates of light extinction by PM2.5 
components calculated using an adjusted version of the original IMPROVE 
algorithm. As described above (sections V.B.1.a and V.D.2), this 
algorithm allows the estimation of light extinction using routinely 
monitored components of PM2.5 and PM10-2.5,\146\ 
along with estimates of relative humidity. While revisions have been 
made to the IMPROVE algorithm since the 2012 review (U.S. EPA, 2022b, 
section 5.3.1.1), the Administrator recognizes that our fundamental 
understanding of the relationship between ambient PM and light 
extinction has changed little since the 2012 review. He further 
recognizes that the results of the quantitative analyses in the 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 equation. He does, 
however, recognize that the PA suggests 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 (U.S. EPA, 2022b, section 5.3.1.2), and he further 
notes the CASAC's recognition that PM-visibility relationships are 
region specific (Sheppard, 2022a, p. 21 of consensus responses). In the 
absence of a robust monitoring network to directly measure light 
extinction (sections V.B.1.a and V.D.2), he preliminarily judges 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.
---------------------------------------------------------------------------

    \146\ In the 2012 review, the focus was on PM2.5 
components given their prominent role in PM-related visibility 
impairment in urban areas and the limited data available for 
PM10-2.5 (77 FR 38980, June 29, 2012; U.S. EPA, 2022b, 
section 5.3.1.2).
---------------------------------------------------------------------------

    In further defining the characteristics of a visibility index based 
on estimates of light extinction, the Administrator considers the 
appropriate averaging time, form, and level 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. The 
Administrator recognizes that the evidence available in this 
reconsideration and described in the PA continue to provide support for 
the short-term nature of PM-related visibility effects. In so doing, he 
relies on analyses of 24-hour and sub-daily PM2.5 light 
extinction to inform his conclusions on averaging time. The 
Administrator notes that there are strong correlations between 24-hour 
and sub-daily (i.e., 4-hour average) PM2.5 light 
extinction), indicating that a 24-hour averaging time is an appropriate 
surrogate for the sub-daily time periods relevant for visual perception 
(U.S. EPA, 2011, Appendix G, section G.4). He further recognizes that 
the longer averaging time may be less influenced by atypical conditions 
and/or atypical instrument performance. Considering this information, 
and noting that the CASAC did not provide advice or recommendations 
with regard to the averaging time of the visibility index, the 
Administrator preliminarily judges that it the 24-hour averaging time 
continues to be appropriate for the visibility index.
    With regard to the form of the visibility index, the Administrator 
notes 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 (U.S. EPA, 2011, p. 4-58). 
In considering the percentile that would be appropriate with the 3-year 
average, 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.\147\ Based on analyses examining 90th, 95th, 
and 98th percentile forms, the Administrator preliminarily judges 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 index, would be reasonably expected to 
lead to improvements in visual air quality for the 20% most impaired 
days (U.S. EPA, 2011, p. 4-59). 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 concludes 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 notes that the CASAC did not provide 
advice or recommendations related to the form of the visibility index. 
Therefore, the Administrator preliminarily judges 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.
---------------------------------------------------------------------------

    \147\ As noted above, the Administrator views the Regional Haze 
Program as a complement to the secondary PM NAAQS, and thus takes 
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 notes that the information that is available regarding the range 
of levels of visibility impairment judged to be acceptable by at least 
50% of study participants in the visibility preference studies is 
largely the same as was in previous reviews.\148\ As such, the 
Administrator notes that the PA identifies a range of 20 to 30 dv as 
appropriate for considering the level for the visibility index. 
Furthermore, the Administrator notes that a level at the upper end of 
the range (i.e., 30 dv) was selected for the 2012 and 2020 reviews, 
given the uncertainties and limitations associated with the public 
preference studies (U.S. EPA, 2022b, section 5.3.1.1). In considering 
the available public preference studies and the range of target levels 
of protection derived from the studies, the Administrator notes that, 
while methodologically similar, the studies have inherent differences 
that impact the responses from the study participants. He notes that 
the images used to evaluate public preferences differed significantly 
depending on geographical location, and that public preferences for 
visual air

[[Page 5660]]

quality can vary depending on the scenic elements depicted in the 
images. He also recognizes that the older studies (i.e., those in 
Denver, CO, and British Columbia, Canada) used photographs, paired with 
ambient measurements of light extinction, as opposed to the computer-
generated images in more recent studies (i.e., those in Phoenix, AZ, 
and Washington, DC), which introduces more variability in scene 
appearance that can influence preferences. Furthermore, the distances 
of objects depicted in the images can influence the perceived 
visibility changes, as objects at a greater distance have more 
sensitivity to changes in visibility impairment compared to those at 
shorter distances. The Administrator recognizes that these differences, 
and the uncertainties and limitations that result from them, are 
important to consider when identifying a target level of protection for 
the visibility index, particularly in identifying the appropriate level 
of protection that would be neither more nor less stringent than 
necessary for a national standard.
---------------------------------------------------------------------------

    \148\ For reasons stated above, the Administrator does not find 
it appropriate to use the most recent preference study (Malm et al., 
2019) for purposes of identifying a target level of protection for 
the visibility index.
---------------------------------------------------------------------------

    In addition to the methodological differences across the public 
preferences studies, the Administrator takes note of the uncertainties 
and limitations associated with the studies and discussed in the PA. In 
particular, the Administrator notes that available studies may not 
capture the full range of visibility preferences in the U.S. 
population, particularly given the potential for preferences to vary 
based on the visibility conditions commonly encountered and the types 
of scenes being viewed and factors that are not captured by the methods 
used in available preference studies may influence people's judgments 
on acceptable visibility, including the duration of visibility 
impairment, the time of day during which light extinction is greatest, 
and the frequency of episodes of visibility impairment (U.S. EPA, 
2022b, section 5.3.1.1).
    In considering the appropriate target level of protection for the 
visibility index, the Administrator also takes note of the CASAC's 
advice. Specifically, he notes that the CASAC recognizes that such a 
judgment 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. While the CASAC 
did not specifically recommend a level for the visibility index, they 
did state that a visibility index of 30 deciviews ``needs to be 
justified'' (Sheppard, 2022a, p. 21 of consensus responses). In 
considering the available information and the CASAC's advice, the 
Administrator notes that the public preference studies were conducted 
in several geographical areas across the U.S., and while they provide 
insight to regional preferences for visibility impairment, none of 
these studies identify a specific level of visibility impairment that 
would be perceived as ``acceptable'' or ``unacceptable'' across the 
whole U.S. population. The Administrator notes 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. 
Similarly, in the 1997 review, the Administrator at that time noted 
significant differences in visibility in the eastern U.S. compared to 
the western U.S. due to background conditions, found that a standard 
set to protect against visibility impairment nationwide would be 
significantly overprotective and not justified for some parts of the 
country, and concluded it was appropriate to rely on the Regional Haze 
Program in conjunction with the secondary PM NAAQS to achieve the 
requisite degree of protection from visibility impairment (62 FR 38652, 
July 18, 1997). For the reasons noted above, the Administrator is not 
seeking to set a standard that would eliminate visibility impairment in 
Class I areas, but significant uncertainties remain regarding how to 
judge visibility impairment across the entire range of daily outdoor 
activities for Americans across the country. Thus, the Administrator 
recognizes 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. The 
Administrator proposes to conclude that the uncertainties and 
variability inherent in the public preference studies warrant setting a 
higher target level of protection than if the underlying methods and 
results from the public preference studies were more consistent. In so 
doing, the Administrator first preliminarily judges 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 proposes 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.
    Having provisionally concluded 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 considers the degree of 
protection from visibility impairment afforded by the existing 
secondary standards. He considers the updated analyses of PM-related 
visibility impairment presented in the PA (U.S. EPA, 2022b, section 
5.3.1.2), which reflect several improvements over 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). This approach provides an improved 
understanding of how variation in equation inputs impacts calculated 
light extinction (U.S. EPA, 2022b, Appendix D). In addition, 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 notes that the results of these updated analyses 
are consistent with the results from the 2012 and 2020 reviews. 
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 
concludes 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

[[Page 5661]]

light extinction in the analyses presented in the PA, he nevertheless 
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 
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 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 [mu]g/m\3\ should 
be considered'' (Sheppard, 2022a, p. 21 of consensus responses). The 
Administrator recognizes that the CASAC recommended 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 has considered the CASAC's 
advice, together with the available scientific evidence and 
quantitative information in reaching his proposed conclusions. The 
Administrator recognizes 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 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 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 
proposes to conclude that the current secondary standards provide 
requisite protection against PM-related visibility effects. With 
respect to non-visibility welfare effects, the Administrator considers 
the evidence for PM-related impacts on climate and on materials and 
concludes that it is generally appropriate to retain the existing 
secondary standards and that it is not appropriate to establish any 
distinct secondary PM standards to address non-visibility PM-related 
welfare effects. With regard to climate, he 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). The resulting uncertainty leads 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 base a national ambient standard on climate 
impacts.
    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 
preliminarily judges 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 support a distinct national 
ambient standard based on materials impacts.
    Taken together, the Administrator proposes 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 reflects 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 reflects 
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 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) \149\ effects related to PM in ambient air. His 
proposed 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. 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 proposes not to 
change those standards (i.e., the current 24-hour and annual 
PM2.5 standards, 24-hour PM10 standard) at this 
time. The Administrator solicits comments on this proposed conclusion.
---------------------------------------------------------------------------

    \149\ 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.
---------------------------------------------------------------------------

    The Administrator additionally recognizes 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 notes a

[[Page 5662]]

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), which 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. While the Administrator notes 
that CASAC did not recommend revising the level of the current 24-hour 
PM2.5 standard, the Administrator recognizes that, should an 
alternative level be considered for the visibility index, that the 
CASAC recommends also considering revisions to the secondary 24-hour 
PM2.5 standard (Sheppard, 2022a, p. 21 of consensus 
responses). Thus, the Administrator solicits 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 
[mu]g/m\3\. Any comments on such revisions should include an 
explanation of the basis for the commenters' views.

E. Proposed Decisions on the Secondary PM Standards

    Taking the above considerations into account, upon reconsidering 
the public welfare protection provided by the current secondary PM 
standards for the known and anticipated adverse effects within the 
scope of this reconsideration, in light of the currently available 
scientific evidence and quantitative information, the Administrator 
proposes not to change the current secondary PM standards at this time. 
In the Administrator's preliminary judgment, such a suite of secondary 
PM standards and the rationale supporting not revising the current 
standards are reasonably judged to reflect the appropriate 
consideration of the strength of the available evidence and other 
information and their associated uncertainties and the advice of CASAC.
    The Administrator recognizes that the final suite of standards will 
reflect his ultimate judgment in the final rulemaking, and in the on-
going review of the secondary NAAQS for oxides of nitrogen, oxides of 
sulfur, and PM, as to the suite of secondary PM 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 final judgment to be made by the Administrator will 
appropriately consider the requirement for standards that are neither 
more nor less stringent than necessary and will recognize that the CAA 
does not require that secondary 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 Administrator also solicits comment on whether it would be 
appropriate to revise the current secondary 24-hour PM2.5 
standard, in conjunction with considering a lower target level of 
protection for the visibility index below 30 dv, and as low as 25 dv. 
The Administrator takes note that, while the CASAC did not recommend 
changes to the current level of 35 [mu]g/m\3\ for the secondary 24-hour 
PM2.5 standard, they indicated that alternative levels 
should be considered if a lower target level of protection (i.e., lower 
than 30 dv) for the visibility index was judged to be appropriate. 
Thus, the Administrator additionally solicits comment on the 
appropriateness of revising the level of the current secondary 24-hour 
PM2.5 standard to a level as low as 25 [mu]g/m\3\. Any 
comments on such revisions should include an explanation of the basis 
for the commenters' views.
    Having reached the proposed decision described here based on 
interpretation of the welfare effects evidence for this 
reconsideration, as assessed in the 2019 ISA and ISA Supplement, and 
the quantitative analyses of visibility impairment in the PA; the 
evaluation of policy-relevant aspects of the evidence and quantitative 
analyses in the PA; the advice and recommendations from the CASAC; 
public comments received to date in this reconsideration; and the 
public welfare policy judgments described above, the Administrator 
recognizes that other interpretations, assessments and judgments might 
be possible. Therefore, the Administrator solicits comment on the array 
of issues associated with reconsideration of the secondary PM 
standards, including public welfare and science policy judgments 
inherent in his proposed decision, as described above, and the 
rationales upon which such views are based.

VI. Interpretation of the NAAQS for PM

A. Proposed Amendments to Appendix K: Interpretation of the NAAQS for 
Particulate Matter

    The EPA proposes to revise appendix K to make the PM10 
data handling procedures for the 24-hour PM10 standards 
specified in 40 CFR 50.6 more consistent with those for other NAAQS 
pollutants and to codify existing practices. The proposed revisions, 
which describe site-level computations, site-to-site combinations, and 
daily validity requirements are discussed in more detail below.
1. Updating Design Value Calculations To Be on a Site-Level Basis
    First, the EPA proposes to require PM10 design values be 
calculated on a site-level basis. Past practice has been to calculate a 
monitor-level design value for each individual PM10 monitor 
when more than one monitor is located at a single site; however, this 
practice is inconsistent with the data handling for PM2.5 
and several other NAAQS pollutants. This inconsistency with 
PM2.5 has led to public confusion about the applicable 
PM10 design value and data completeness criteria at a site 
because operators are more accustomed to site-level monitoring 
requirements. To resolve this confusion, the EPA believes it would be 
appropriate to identify a single design value for each site; the EPA is 
proposing an analytic approach to combine data collected from multiple 
PM10 monitors collocated at a site to obtain a single set of 
daily PM10 concentration data for that site. This proposal 
to move from monitor-level to site-level PM10 design values 
is supported by the high level of consistency in the measurement data 
obtained across the various Federal reference and equivalent 
PM10 monitoring instruments currently in operation (U.S. 
EPA, 2009a, section 3.4.1.1).
    The proposed approach would provide for monitoring agencies to 
designate in their annual network plan one monitor as the primary 
monitor for each site.\150\ Once a primary monitor has been determined 
for a site, missing daily PM10 concentrations for the 
primary monitor would be substituted from any other monitors located at 
the site. In the event of two or more monitors operating at the same 
site, missing daily PM10 concentrations for the primary 
monitor would be substituted with daily values averaged across the 
other collocated monitors. The EPA notes that at the time of this 
proposal, there were more than 100 sites nationwide with two or more 
monitors operating simultaneously.
---------------------------------------------------------------------------

    \150\ In the absence of a primary monitor designation, the 
primary monitor would default to the monitor with the most complete 
daily dataset in each year.
---------------------------------------------------------------------------

    This proposed approach for combining data across collocated 
monitors at a site is consistent with the existing approach described 
in appendix N to part 50 for the current PM2.5 NAAQS. The 
EPA invites public comment on the scientific validity of

[[Page 5663]]

combining data across PM10 monitors and the merits of the 
proposed approach for combining data across multiple PM10 
monitors collocated at a site.
2. Codifying Site Combinations To Maintain a Continuous Data Record
    Second, and complementary to the first proposed revision described 
above, the EPA proposes to maintain the existing practice of combining 
data from nearby monitoring sites to determine a valid design value, 
known as a ``site combination.'' Site combinations typically involve 
situations where one site closes and another begins monitoring a short 
distance away within a few days, and the monitoring agency wishes to 
combine the data from the two sites to maintain a continuous data 
record. The EPA Regional offices have approved over ten site 
combinations for PM10 since the promulgation of the 1987 
PM10 NAAQS; these will be considered approved site 
combinations if these revisions are promulgated.
    Relatedly, the EPA proposes to maintain the existing practice of 
allowing monitoring agencies to submit site combination requests to the 
appropriate Regional Administrator through the EPA's Air Quality System 
(AQS) database. Site combinations may be approved by the Regional 
Administrator after they determine that the measured air quality 
concentrations do not differ substantially between the two sites. To 
make this determination for a requested site combination, the Regional 
Administrator may request additional information from the Agency 
including detailed information on the locations and distance between 
the two sites, levels of ambient concentrations measured at the two 
sites, and local emissions or meteorology data. To improve 
transparency, the EPA will make records of all approved site 
combinations available in the AQS database and will update design value 
calculations in AQS when approved site combinations are implemented. 
The EPA invites public comment on the merits of the proposed process 
for approving site combinations to obtain valid design values for the 
PM10 NAAQS.
3. Clarifying Daily Validity Requirements for Continuous Monitors
    Third, the EPA proposes to maintain the existing practice of 
considering daily averages to be valid if at least 75 percent of the 
hourly averages (i.e., 18 hourly values) for the 24-hour period are 
available unless a substitution test can show validity on days with 
seven or more missing hours.

B. Proposed Amendments to Appendix N: Interpretation of the NAAQS for 
PM2.5

    The EPA proposes to revise appendix N by updating references to the 
proposed revision(s) of the standards and changing data handling 
provisions related to combining data from nearby monitoring sites to 
codify existing practices that are currently being implemented as EPA 
standard operating procedures.
1. Updating References to the Proposed Revision(s) of the Standards
    The EPA proposes to maintain the existing practice of combining 
data from nearby monitoring sites to determine a valid design value, 
known as a ``site combination.'' Site combinations typically involve 
situations where one site closes and another begins monitoring a short 
distance away within a few days, and the monitoring agency wishes to 
combine the data from the two sites to maintain a continuous data 
record. The EPA Regional offices have approved over 40 site 
combinations for PM2.5 since the promulgation of the 1997 
PM2.5 NAAQS; these will be considered approved site 
combinations if these revisions are promulgated.
2. Codifying Site Combinations To Maintain a Continuous Data Record
    Relatedly, the EPA proposes to maintain the existing practice of 
allowing monitoring agencies to submit site combination requests to the 
appropriate Regional Administrator through the EPA's Air Quality System 
(AQS) database. Site combinations may be approved by the Regional 
Administrator after they determine that the measured air quality 
concentrations do not differ substantially between the two sites. To 
make this determination for a requested site combination, the Regional 
Administrator may request additional information from the Agency 
including detailed information on the locations and distance between 
the two sites, levels of ambient concentrations measured at the two 
sites, and local emissions or meteorology data. To improve 
transparency, the EPA will make records of all approved site 
combinations available in the AQS database and will update design value 
calculations in AQS when approved site combinations are implemented. 
The EPA invites public comment on the merits of the proposed process 
for approving site combinations to obtain valid design values for the 
PM2.5 NAAQS.

VII. Proposed Amendments to Ambient Monitoring and Quality Assurance 
Requirements

    The EPA is proposing revisions to ambient air monitoring 
requirements for PM to improve the usefulness of and appropriateness of 
data used in regulatory decision making. These proposed 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 
proposed 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 last completed revisions to PM ambient air monitoring 
regulations as a part of the PM NAAQS review completed in 2012 (78 FR 
3085, January 15, 2013). This final rulemaking included revisions to 
ensure the suite of standards for PM provide requisite protection of 
public health and welfare as well as corresponding revisions to the 
data handling conventions for PM and to the ambient air monitoring, 
reporting, and network design requirements. Other pollutant-specific 
monitoring updates have occurred in conjunction with revisions to the 
NAAQS. In such cases, the monitoring revisions were typically finalized 
as part of the final rulemaking for the NAAQS.\151\ Specific proposed 
changes are described below.
---------------------------------------------------------------------------

    \151\ Links to the NAAQS final rules are available at: https://www.epa.gov/criteria-air-pollutants.
---------------------------------------------------------------------------

A. Proposed Amendment in 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 is proposing a technical change to appendix L to include 
the addition of an alternative PM2.5 particle size separator 
to that of the WINS and the VSCC size separators. The new separator is 
the TE-PM2.5C cyclone manufactured by Tisch Environmental 
Inc., Cleves, Ohio, and 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 service interval 
comparable to the VSCC separator and is significantly longer than the 
service

[[Page 5664]]

interval for the WINS. Generally, the TE-PM2.5C is also 
physically interchangeable with the WINS and VSCC where both are 
manufactured for the same sampler. The proposal would allow 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-PM2.5C 
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 proposed change to 
appendix L, these instruments would be redesignated as PM2.5 
and PM10-2.5 FRMs, respectively. Owners of such samplers 
would contact the sampler manufacturer to receive a new reference 
method label for the samplers.

B. Proposed Amendments to Ambient Air Monitoring Reference and 
Equivalent Methods in 40 CFR Part 53

    The EPA is proposing clarifications to the regulations associated 
with submittal of candidate FRM and FEM applications for review by the 
EPA. Revisions are also proposed in instances where current regulatory 
specifications are no longer pertinent and require updating. In 
addition, the EPA has compiled a list of noted minor errors to correct 
in regulations associated with the testing requirements and acceptance 
criteria for Federal reference methods (FRMs) and Federal equivalent 
methods (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 
Application and Modification Requests
    The EPA is proposing to update the name of the program and delivery 
address for the EPA review of FRM and FEM Applications and Modification 
Requests (Sec.  53.4). 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.
2. Requests for Delivery of a Candidate FRM or FEM Instrument
    As part of the current applicant review process, Sec.  53.4(d) 
allows the EPA to request only candidate PM2.5 FRMs and 
Class II or Class III equivalent methods for test purposes. The EPA 
proposes to revise this section to allow the EPA to request any 
candidate FRM, FEM, or a designated FRM or FEM associated with a 
Modification Request, regardless of NAAQS pollutant type or metric.
3. Amendments to Requirements for Submission of Materials in Sec.  
53.4(b)(7) for Language and Format
    The EPA proposes to amend Sec.  53.4(b)(7), which specifies the 
format(s) in which all submissions must be received, to specify that 
all written 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.
4. Amendment to Designation of Reference and Equivalent Methods
    The EPA proposes to clarify the terms of new FRM and FEM methods 
(Sec.  53.8(a)) to ensure that candidate samplers and analyzers are not 
publicly announced, marketed, or sold as FRMs until the EPA's approval 
has been formally announced in the Federal Register.
5. Amendment to One Test Field Campaign Requirement for Class III 
PM2.5 FEMs
    Field comparability tests for candidate Class III PM2.5 
FEMs include the requirement that a total of five field campaigns must 
be conducted at four separate sites: A, B, C, and D. The site D 
specifications of Sec.  53.35(b)(1)(ii)(D) 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 proposes to revise the 
site D specifications to read ``. . . shall be in a large city east of 
the Mississippi River, having characteristically high humidity 
levels.''
6. Amendment to Use of Monodisperse Aerosol Generator
    Wind tunnel evaluation of candidate PM10 inlets and 
evaluation of candidate PM2.5 fractionators under static 
conditions requires the generation and use of monodisperse calibration 
aerosols of specified aerodynamic sizes. In the current regulations 
(Sec.  53.61(g)), the TSI Incorporated Vibrating Orifice Aerosol 
Generator (VOAG) is the only approved monodisperse generator for this 
purpose. However, TSI Incorporated no longer manufacturers nor supports 
the VOAG. Therefore, the EPA proposes to add a commercially available 
monodisperse aerosol generator--the Model 1520 Flow-Focusing 
Monodisperse Aerosol Generator, MSP Corporation, Shoreview, MN--to the 
list of approved generators for this purpose.
7. Corrections to 40 CFR Part 53 (Reference and Equivalent Methods)
    Certain provisions of Sec.  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; October 26, 2015), 
all EPA response deadlines for modifications of reference or equivalent 
methods are 90 days from day of receipt.
    The EPA proposes corrections to the following tables: Table A-1 to 
Subpart A of Part 53--Summary of Applicable 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 SO2 states the interference equivalent for 
each interferent is 0.005 ppm for both the standard- 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.
    The EPA proposes corrections to the following figures: 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. 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

[[Page 5665]]

calculation formula and Figure B-5 lists an incorrect calculation 
metric: there is a typesetting error in the calculation of the standard 
deviation. The EPA proposes to correct the typesetting errors.
    The EPA proposes correcting typesetting errors in several formulas 
provided throughout Sec.  53.43.

C. Proposed 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
    The EPA has evaluated the quality system as part of the PM NAAQS 
reconsideration and identified several areas that could be improved in 
light of lower average ambient PM2.5 concentrations across 
the country and the proposed more revised primary annual 
PM2.5 NAAQS described in section II above. Thus, we assessed 
PM2.5 concentration data across a range of values to 
determine if any changes were warranted to their use in the statistics 
used to evaluate the data qualify in the PM2.5 network. This 
section describes that work and any proposed changes as a result. Other 
changes proposed in this section include clarifications and other 
improvements that will better assist with the consistency and 
operations of quality assurance programs.
a. Quality System Requirements
    The EPA has reconsidered the appendix A, section 2.3.1.1, goal for 
acceptable measurement uncertainty for automated and manual 
PM2.5 methods currently stated as an upper 90 percent 
confidence limit for the coefficient of variation (CV) of 10 percent 
and 10 percent for total bias. The average PM2.5 
concentrations across the nation have steadily declined since the 
promulgation of the first PM2.5 standard (U.S. EPA, 2022, 
section 2.3). As ambient concentrations decrease, the bias is inflated 
using the current bias statistic in 4.2.5. The EPA has developed a new 
bias statistic to minimize the effect of low PM2.5 
concentrations on bias and is proposing to revise section 4.2.5 to 
implement this new bias statistic. The EPA has concluded that with this 
change to the bias statistic, the coefficient of variation (CV) of 10 
percent and 10 percent for total bias is still an 
acceptable goal for estimating total bias in the networks. The 
technical justification and background for this change is documented in 
a technical memorandum to the docket for this rulemaking titled ``Task 
16 on PEP/NPAP Task Order: Bias and Precision DQOs for the 
PM2.5 Ambient Air Monitoring Network.'' \152\
---------------------------------------------------------------------------

    \152\ Noah, G. (2022). 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 is proposing to update and clarify 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 Protocol Gas Verification Program. Appendix A would be revised to 
clarify that in order to participate in the Ambient Air Protocol Gas 
Verification Program, 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 Protocol Gas Verification Program. The EPA is proposing 
to include an allowable uncertainty of 2.0 percent for EPA 
Protocol Gas standards used in ambient air monitoring. This allowable 
uncertainty limit would match the existing limit set by the EPA's 
continuous emission monitoring program found in part 75, appendix A, 
section 5.1.4(b), and would make the EPA's regulations of quality 
assurance of ambient air monitors more uniform and consistent.
b. Measurement Quality Check Requirements
    The EPA is proposing to remove section 3.1.2.2 from appendix A. 
This provision in the quality assurance requirements for ambient air 
monitoring allows for NO2 compressed gas standards to be 
used to generate audit standards. However, NO2 compressed 
gas standards are not currently designated by the EPA's Office of 
Research and Development (ORD) as an EPA Protocol Gas Standard. As 
such, this provision conflicts with section 2.6.1 of appendix A that 
requires that any standard used for generating test atmospheres be an 
EPA Protocol Gas Standard. The EPA is aware that there is a need for 
NO2 compressed gas standards for direct read NO2 
monitoring methods. If these NO2 compressed gas standards 
can, in the future, be proven to be stable and approvable as EPA 
Protocol Gas Standards, the EPA will consider restoring this provision 
to appendix A.
    The EPA is proposing to revise the requirement in section 3.1.3.3 
pertaining to the validation of the gaseous cylinders used for the 
National Performance Audit Program (NPAP). The EPA proposes to change 
the requirement for annual verification to the ORD-recommended 
certification periods for standards identified in Table 2-3 of the EPA 
Traceability Protocol for Assay and Certification of Gaseous 
Calibration Standards (appendix A, section 6.0(4)). These ORD-
recommended periods are based on the periods for which similar gas 
mixtures over specific concentration ranges have been shown to be 
stable, as documented in the peer-reviewed literature or in 
concentration stability data submitted by the National Institute of 
Standards and Technology (NIST) and specialty gas producers and 
reviewed by the EPA. In effect, this would decrease the cost and burden 
on the Protocol Gas Verification Program (PGVP), which performs these 
verifications annually. The EPA anticipates this will also decrease the 
delay in returning tanks back to the auditors. This would provide 
auditors with longer periods with valid certifications to perform 
audits without annual interruptions for the verification process.
    The EPA is proposing 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\. As discussed above, ambient PM2.5 
concentrations have decreased, and many samples being collected now are 
below the 3 [mu]g/m\3\ threshold and deemed invalid for purposes of a 
valid audit sample. Therefore, decreasing this threshold from 3 [mu]g/
m\3\ to 2 [mu]g/m\3\ would increase the number of valid PEP sample 
pairs collected, which would reduce the number of re-audits that need 
to be performed to compensate for invalid sample pairs. Inclusion of 
values down to 2 [mu]g/m\3\ would represent the concentrations 
occurring in routine monitoring operations and are included in annual 
mean concentrations of the networks. Reducing the number of re-audits 
would reduce audit costs to monitoring organizations while better 
representing the data in the networks. The technical justification and 
background for this change is documented in a technical memorandum to 
the docket for this rulemaking titled ``Task 16 on PEP/NPAP Task Order: 
Bias and Precision DQOs for the PM2.5 Ambient Air Monitoring 
Network.'' \153\
---------------------------------------------------------------------------

    \153\ Noah, G. (2022). 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.

---------------------------------------------------------------------------

[[Page 5666]]

c. Calculations for Data Quality Assessments
    The EPA is proposing to update the appendix A, section 4.2.1, 
Equations 6 and 7, for calculating the Collocated Quality Control 
Sampler Precision Estimate for PM10, PM2.5 and Pb.
    The proposed changes are:




    [GRAPHIC] [TIFF OMITTED] TP27JA23.003
    
    These new statistics are designed to address the inflated precision 
values that result from using these calculations to compare low 
concentrations that are now observed in the networks. The current 
precision estimate uses a relative percent difference (RPD) when 
comparing two collocated samplers. As the two numbers used in the 
comparison get smaller, the statistic generally produces a result that 
is inflated. A precision statistic calculated for low-concentration 
data may show poor agreement even if the nominal values are relatively 
close to each other. By using the square root in the denominator in 
these statistics, the variability is more constant across all 
concentrations thereby reducing the inflated effect. The EPA believes 
this proposed change would provide the correct context for considering 
inflated RPDs when calculating the bias estimate. The technical 
justification and background for this change is documented in a 
technical memorandum to the docket for this rulemaking titled ``Task 16 
on PEP/NPAP Task Order: Bias and Precision DQOs for the 
PM2.5 Ambient Air Monitoring Network.'' \154\
---------------------------------------------------------------------------

    \154\ Noah, G. (2022). 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 is proposing to update the appendix A, section 4.2.5, 
Equation 8, calculation for the Performance Evaluation Programs Bias 
Estimate for PM2.5 from

[[Page 5667]]

[GRAPHIC] [TIFF OMITTED] TP27JA23.004

    Again, because the average ambient PM concentrations across the 
nation have steadily declined since the promulgation of the 
PM2.5 standard, the current method of calculation may not be 
appropriate for determining bias for these lower ambient concentrations 
and newer sampling methodologies. The current bias estimate uses a 
percent difference (PD), referenced in appendix A, section 4.1.1, when 
comparing an audit sampler against a routine sampler. As the two 
numbers used in the comparison get smaller, the statistic generally 
produces a result that is inflated. A bias statistic calculated for 
low-concentration data may show poor agreement even if the nominal 
values are relatively close to each other. This may be misleading when 
trying to assess bias and summarizing data to be used in decision 
making. The EPA believes this proposed change would provide the correct 
context for considering inflated RPDs when calculating the bias 
estimate. The technical justification and background for this change is 
documented in a technical memorandum to the docket for this rulemaking 
titled ``Task 16 on PEP/NPAP Task Order: Bias and Precision DQOs for 
the PM2.5 Ambient Air Monitoring Network.'' \155\
---------------------------------------------------------------------------

    \155\ Noah, G. (2022). 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.
---------------------------------------------------------------------------

d. References
    The EPA proposes to update the references and hyperlinks in 
appendix A, section 6. Several of the reference documents have been 
updated and the web locations have changed. This proposal provides 
accuracy in identifying and locating essential supporting documentation 
so that historical documents that do not represent current practices 
are not used. The EPA believes that it is important that interested 
parties--especially ambient air monitoring organizations and 
stakeholders--have the most current materials that provide 
clarifications and guidance on the interpretation of the regulations.
    The EPA is also proposing to add a footnote to Table A-1 of 
Appendix A to Part 58--Minimum Data Assessment Requirements for NAAQS 
Related Criteria Pollutant Monitors. The proposed footnote would 
clarify the allowable time (i.e., every two weeks, once a month, once a 
quarter, once every 6 months, or distributed over all 4 quarters 
depending on the check) between checks and encourage monitoring 
organizations to perform data assessments at regular intervals. The EPA 
believes this proposal is appropriate because the current stipulation 
is unclear regarding the specified interval for required verifications. 
For example, under the current flow rate verification for 
PM10 (low vol.), PM2.5, and Pb-PM10, a 
flow check could be performed on April 1 and not checked again until 
May 31, leaving approximately two months between checks. Following this 
practice would leave large intervals of time between verifications, and 
if a check fails using the described practice, an unacceptably large 
data loss could result. Also, a check could be performed on the last 
day of a quality control (QC) check interval and then on the first day 
of the following interval, with only a day or two between checks. This 
is not the intended practice for QC measures that are meant to ensure 
equipment is continually operating properly over an operational period. 
For this reason, the EPA is proposing to clarify the allowable time 
between checks.
2. Quality Assurance Requirements for Prevention of Significant 
Deterioration (PSD) Air Monitoring
    This section on Quality Assurance Requirements for Prevention of 
Significant Deterioration (PSD) Air Monitoring was developed in 
parallel to the proposed changes associated with appendix A. Thus, this 
section includes similar detail and proposed changes for evaluating 
quality system statistics for PM2.5, clarifications, and 
other improvements that will better assist with the consistency and 
operations of quality assurance programs for PSD.

[[Page 5668]]

a. Quality System Requirements
    The EPA has reconsidered the appendix A, section 2.3.1.1, goal for 
acceptable measurement uncertainty for automated and manual 
PM2.5 methods currently stated as an upper 90 percent 
confidence limit for the CV of 10 percent and 10 percent 
for total bias. The average PM concentrations across the nation have 
steadily declined since the promulgation of the first PM2.5 
standard (U.S. EPA, 2022, section 2.3). As ambient concentrations 
decrease, the bias is inflated using the current bias statistic in 
section 4.2.5. Using a new statistic to replace the existing statistic 
in section 4.2.5 developed to eliminate the effect of low 
concentrations on bias, the EPA has concluded that the coefficient of 
variation (CV) of 10 percent and 10 percent for total bias 
is still an acceptable goal for estimating total bias in the networks. 
The technical justification and background for this change is 
documented in a technical memorandum to the docket for this rulemaking 
titled ``Task 16 on PEP/NPAP Task Order: Bias and Precision DQOs for 
the PM2.5 Ambient Air Monitoring Network.'' \156\
---------------------------------------------------------------------------

    \156\ Noah, G. (2022). 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 is proposing to update and clarify 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 Protocol Gas Verification Program. Appendix A would be revised to 
clarify that in order to participate in the Ambient Air Protocol Gas 
Verification Program, 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 Protocol Gas Verification Program. The EPA is proposing 
to include an allowable uncertainty of 2.0 percent for EPA 
Protocol Gas standards used in ambient air monitoring. This allowable 
uncertainty limit would match the existing limit set by the EPA's 
continuous emission monitoring program found in part 75, appendix A, 
section 5.1.4(b), and would make the EPA's regulations more uniform and 
consistent.
b. Measurement Quality Check Requirements
    The EPA is proposing to remove section 3.1.2.2 from appendix A. 
This provision in the quality assurance requirements for ambient air 
monitoring allows for NO2 compressed gas standards to be 
used to generate audit standards. However, NO2 compressed 
gas standards are not currently designated by the EPA's ORD as an EPA 
Protocol Gas Standard. As such, this provision conflicts with section 
2.6.1 of appendix A that requires that any standard used for generating 
test atmospheres be an EPA Protocol Gas Standard. The EPA is aware that 
there is a need for NO2 compressed gas standards for direct 
read NO2 monitoring methods. If these NO2 
compressed gas standards can, in the future, be proven to be stable and 
approvable as EPA Protocol Gas Standards, the EPA will consider 
restoring this provision to appendix A.
    The EPA is proposing to revise the requirement in section 3.1.3.3 
pertaining to the validation of the gaseous cylinders used for the 
NPAP. The EPA proposes to change the requirement for annual 
verification to the ORD-recommended certification periods for standards 
identified in Table 2-3 of the EPA Traceability Protocol for Assay and 
Certification of Gaseous Calibration Standards (appendix A, section 
6.0(4)). These ORD-recommended periods are based on the periods for 
which similar gas mixtures over specific concentration ranges have been 
shown to be stable, as documented in the peer-reviewed literature or in 
concentration stability data submitted by NIST and specialty gas 
producers and reviewed by the EPA. In effect, this would decrease the 
cost and burden on the PGVP, which performs these verifications 
annually. The EPA anticipates this will also decrease the delay in 
returning tanks back to the auditors. This would provide auditors with 
longer periods with valid certifications to perform audits without 
annual interruptions for the verification process.
    The EPA is proposing 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\. As discussed above, ambient PM2.5 
concentrations have decreased, and many samples being collected now are 
below the 3 [mu]g/m\3\ threshold and deemed invalid for purposes of a 
valid audit sample. Therefore, decreasing this threshold from 3 [mu]g/
m\3\ to 2 [mu]g/m\3\ would increase the number of valid PEP sample 
pairs collected, which would reduce the number of re-audits that need 
to be performed to compensate for invalid sample pairs. Inclusion of 
values down to 2 [mu]g/m\3\ would represent the concentrations 
occurring in routine monitoring operations and are included in annual 
mean concentrations of the networks. Reducing the number of re-audits 
would reduce audit costs to monitoring organizations while better 
representing the data in the networks. The technical justification and 
background for this change is documented in a technical memorandum to 
the docket for this rulemaking titled ``Task 16 on PEP/NPAP Task Order: 
Bias and Precision DQOs for the PM2.5 Ambient Air Monitoring 
Network.'' \157\
---------------------------------------------------------------------------

    \157\ Noah, G. (2022). 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.
---------------------------------------------------------------------------

c. Calculations for Data Quality Assessments
    The EPA is proposing to update the appendix A, section 4.2.5, 
Equation 8, calculation for the Performance Evaluation Programs Bias 
Estimate for PM2.5 from

[[Page 5669]]

[GRAPHIC] [TIFF OMITTED] TP27JA23.005

    Again, because the average ambient PM concentrations across the 
nation have steadily declined since the promulgation of the 
PM2.5 standard, the current method of calculation may not be 
appropriate for determining bias for these lower ambient concentrations 
and newer sampling methodologies. The current bias estimate uses a PD, 
referenced in appendix A, section 4.1.1, when comparing an audit 
sampler against a routine sampler. As the two numbers used in the 
comparison get smaller, the statistic generally produces a result that 
is inflated. A bias statistic calculated for low-concentration data may 
show poor agreement even if the nominal values are relatively close to 
each other. This may be misleading when trying to assess bias and 
summarizing data to be used in making decisions. The EPA believes this 
proposed change would provide the correct context for considering 
inflated RPDs when calculating the bias estimate. The technical 
justification and background for this change is documented in a 
technical memorandum to the docket for this rulemaking titled ``Task 16 
on PEP/NPAP Task Order: Bias and Precision DQOs for the 
PM2.5 Ambient Air Monitoring Network.'' \158\
---------------------------------------------------------------------------

    \158\ Noah, G. (2022). 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.
---------------------------------------------------------------------------

d. References
    The EPA proposes to update the references and hyperlinks in 
appendix A, section 6. Several of the reference documents have been 
updated and the web locations have changed. This proposal provides 
accuracy in identifying and locating essential supporting documentation 
so that historical documents that do not represent current practices 
are not used. The EPA believes that it is important that interested 
parties--especially ambient air monitoring organizations and 
stakeholders--have the most current materials that provide 
clarifications and guidance on the interpretation of the regulations.
    The EPA is also proposing to add a footnote to Table A-1 of 
Appendix A to Part 58--Minimum Data Assessment Requirements for NAAQS 
Related Criteria Pollutant Monitors. The proposed footnote would 
clarify 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 encourage monitoring 
organizations to perform data assessments at regular intervals. The EPA 
believes this proposal is appropriate because the current stipulation 
is unclear regarding the specified interval for required verifications. 
For example, under the current flow rate verification for 
PM10 (low vol.), PM2.5, and Pb-PM10, a 
flow check could be performed on April 1 and not checked again until 
May 31, leaving approximately two months between checks. Following this 
practice would leave large intervals of time between verifications, and 
if a check fails using the described practice, an unacceptably large 
data loss could result. Also, a check could be performed on the last 
day of a QC check interval and then on the first day of the following 
interval, with only a day or two between checks. This is not the 
intended practice for quality control measures that are meant to ensure 
equipment is continually operating properly over an operational period. 
For this reason, the EPA is proposing to clarify the allowable time 
between checks.
3. Proposed Amendments to PM Ambient Air Quality Methodology
a. Proposal To Revoke Approved Regional Methods (ARMs)
    The EPA is proposing to remove provisions for approval and use of 
Approved Regional Methods (ARMs) throughout parts 50 and 58 of the CFR. 
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. However, at this time, there are no approved ARMs, nor does 
the EPA anticipate any will be requested. There are, however, more than 
a dozen approved FEMs for PM2.5. These approved FEMs are 
eligible for comparison to the NAAQS and to meet other monitoring 
objectives.
    The EPA first proposed a process to approve and use ARMs in January 
of 2006 (71 FR 2709, January 17, 2006). At that time, there were no 
approved continuous PM2.5 methods available to compare to 
the NAAQS. The hope was that approved ARMs would quickly start the use 
of PM2.5 continuous methods that worked well in monitoring 
agency networks, since the benefits of regulatory-grade automated 
methods were not available at that time to air agency programs. It was 
hoped that the benefits of automated PM2.5 methods--
including real-time data reporting of PM2.5 to support 
forecasting and reporting of the AQI while also providing a regulatory 
dataset eligible for comparison to the PM2.5 NAAQS--would 
encourage the development of ARMs. The idea to encourage ARMs was 
conceived following review of data across the country demonstrating 
that some agencies were achieving acceptable data comparability with 
their PM2.5 methods compared to collocated FRMs; however, 
those methods did not necessarily provide consistent data across the 
country. At that time, there were no approved PM2.5 
continuous FEMs and it was unclear how soon any might be approved. 
However, by March 2008, the EPA's Reference and Equivalent Methods 
program had approved the first PM2.5 continuous FEM (73 FR 
13224, March 12, 2008). Over the next eight years, an additional 12 
PM2.5 continuous FEMs were approved. With many commercially 
available PM2.5 continuous FEMs available to air agencies, 
almost all agencies soon began implementing one or more 
PM2.5 FEMs in their network. By 2020, monitoring agencies 
were

[[Page 5670]]

reporting PM2.5 continuous FEM data from 660 sites across 
the country (U.S. EPA, 2022, section 2.2.3.1). Therefore, with a large 
and growing network of PM2.5 continuous FEMs and no approved 
applications for ARMs in the 16 years that this provision has been 
available, the EPA is proposing to remove this provision, including any 
related language, and to instead rely on the existing network of 
approved PM2.5 FEMs and future approved FEMs. The EPA notes 
that although references to ARMs occur across part 50 and part 58, the 
EPA is not reopening the substance of the provisions where these 
references occur and is only proposing regulatory text for these 
provisions for the purpose of removing the reference to ARMs.
b. Proposal for Calibration of PM Federal Equivalent Methods (FEMs)
    The EPA is proposing to modify its specifications for PM FEMs 
described in appendix C to part 58. Specifically, the EPA is proposing 
that valid State, local, and Tribal air monitoring data 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 would apply to any PM FEM methods (i.e., 
PM10, PM2.5, and PM10-2.5). The EPA is 
proposing this modification because there are some approved PM FEMs 
that are not currently meeting measurement quality objectives (MQOs) 
when evaluating data nationally (U.S. EPA, 2022, section 2.2.3.1) 
meaning that an update to a factory calibration may be appropriate; 
however, there is not a clearly defined process to update the 
calibration of an FEM. 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 data), we are 
proposing to use routinely operated State, local, and Tribal 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 MQOs across the networks in which they 
are operated. The EPA has received input from CASAC (Sheppard, 2022, p. 
2 of consensus responses) and State, local, and Tribal agencies 
(National Association of Clean Air Agencies (NACAA) Monitoring 
Committee 01/20/22; Association of Air Pollution Control Agencies 
(AAPCA) Ambient Monitoring Committee 01/26/2022; Tribal air quality 
professionals call on 02/17/22), all of which expressed strong interest 
in improving FEM data comparability to collocated FRMs. While there are 
other approaches that could improve data comparability between PM FEMs 
and collocated FRMs, The EPA believes that this approach represents the 
most reliable approach to update FEM factory calibrations, since the 
existing FRM network data that meets MQOs would be used to set updated 
factory calibrations. While the Agency is proposing 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. This 
section 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 the 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 
manufactures are encouraged to seek improvements to their approved FEM 
methods as needed to continue to meet data quality needs as operated 
across the network. Instrument manufactures have an option to pursue 
that now and may have an additional option in the future should we 
finalize this proposal for calibration of PM FEMs.
    In the PA (U.S. EPA, 2022b, section 2.2.3.1), the EPA analyzed the 
quality of data from FRM samplers and continuous PM2.5 FEM 
monitors operating in routine networks to determine whether they meet 
the MQOs for PM2.5 FRMs and FEMs (40 CFR part 58, appendix 
A, section 2.3.1.1): ``Measurement Uncertainty for Automated and Manual 
PM2.5 Methods. The goal for acceptable measurement 
uncertainty is defined for precision as an upper 90 percent confidence 
limit for the coefficient of variation (CV) of 10 percent and 10 percent for total bias.'' When aggregating data across the 
country, all PM2.5 FRMs meet the MQOs for these methods. But 
of PM2.5 continuous FEMs aggregated across the country, some 
meet the MQOs, and others do not.
    One of the major challenges to ensuring uniform data from PM 
methods is that there are no accepted standards against which to 
calibrate PM methods. This was discussed in the 2004 Air Quality 
Criteria for Particulate Matter (U.S. EPA, 2004b). PM reference methods 
typically include the design and performance requirements set forth in 
the 40 CFR part 50. This is a contrast to FRMs and FEMs for gaseous 
NAAQS pollutants for which there are accepted calibration standards; in 
the case of ozone, there is even a standard reference photometer that 
can be used to calibrate approved methods in the field or laboratory. 
For PM monitoring methods, in the absence of accepted calibration 
standards, acceptable data quality is determined by comparing to other 
PM FRMs. One challenge to comparing to other PM FRMs during the initial 
field testing for purposes of FEM approval is that the dataset will in 
almost all cases be substantially more limited than what's available in 
routine networks once deployed. Thus, we seek to encourage instrument 
manufacturers of approved FEMs to evaluate data in routine networks and 
consider improvements to their FEM calibration, as needed.
    The EPA is proposing to use routine and collocated FRM data 
operated by State, local, and Tribal agencies as the basis to update 
factory calibrations. Routine State, local, and Tribal 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. While 
the EPA is proposing to use routine FRM data, there are other reference 
datasets that could be considered. For example, the agency has an FRM 
audit program\159\ operated by independent operators and laboratories. 
This program is highly valuable to the success of the PM2.5 
monitoring program by providing independent data to assess the quality 
of routinely operated FRMs and FEMs. If we used the audit program data 
as the basis for calibrating continuous monitors, we would lose the 
ability to collect independent data from audit monitors to assess the 
operation of routine monitors. Therefore, by using routinely operated 
FRMs to calibrate continuous FEMs, the Agency will continue to maintain 
the independence

[[Page 5671]]

of the FRM audit program to assess the quality of routinely operated 
FRM and continuous FEM data. The EPA also has chemical speciation data 
available at sites where the Chemical Speciation Network (CSN) or 
IMPROVE samplers are operated; however, these samplers use technologies 
that operate at different flow rates and with different-size selective 
devices than approved FRMs, and neither of these programs use FRMs as 
the basis to collect samples. Therefore, while CSN and IMPROVE data can 
be useful to help determine the aerosol chemistry of PM2.5 
and may provide additional validation of collocated FRM or FEM data, by 
themselves these data are not appropriate to update factory calibration 
of continuous FEMs.
---------------------------------------------------------------------------

    \159\ See: https://www.epa.gov/amtic/national-pm25-performance-evaluation-program.
---------------------------------------------------------------------------

    The EPA proposes to direct instrument companies and other 
interested stakeholders to the EPA's Air Quality System (AQS) database 
\160\ to access the valid routine network data that the Agency proposes 
to allow for use in updating factory calibration of continuous FEMs. 
There are several ways to obtain data from the AQS database, and many 
do not require registration. For example, daily processed datasets by 
year are publicly available at the website of ``Pre-Generated Data 
Fields.'' \161\ The data utilized would need to be valid PM FRM and FEM 
data that are collocated and aligned to the same date. For example, for 
PM2.5 mass concentrations, there are files by year for `` 
PM2.5 FRM/FEM Mass'' identified with a parameter code of 
88101. This information, already aggregated to daily data, represent 
the time-period of midnight-to-midnight local standard time. While any 
years of data may be considered, instrument companies should normally 
use at least two years of recent data where we are past the 
certification period for the previous-year data, which is May 1st of 
each year. Including at least two years of data is intended to address 
cases where one of the years may have high or low air quality 
concentrations. Data in the current year and previous year when we are 
not past the May 1st certification date can be considered to test data 
with a correction established from a previous year or more than one 
year. If multiple factors are included, any new statistical correction 
or corrections should be based on one or more calendar years, with 
independent testing of that data on another year or more that was not 
used to develop the equation(s).
---------------------------------------------------------------------------

    \160\ See: https://www.epa.gov/aqs.
    \161\ See: https://aqs.epa.gov/aqsweb/airdata/download_files.html.
---------------------------------------------------------------------------

    The EPA also encourages instrument companies to consider and 
implement all the ways to optimize PM2.5 FEMs. This may 
include, but is not limited to, whether a method's data can be improved 
by operating the FEM inside a heating, ventilation, and air-
conditioning (HVAC)-controlled shelter or outside with minimal or no 
HVAC control; optimizing heating of the airstream to avoid condensation 
while retaining semi-volatile PM captured on the FRM; and any 
specialized guidance or training that may help monitoring agencies 
optimize their data quality and comparability to collocated FRMs. Other 
options might include updates to unique coefficients used in the 
factory calibration such as the density of the aerosol, where 
applicable. Such changes would normally need to be approved by the EPA 
according to existing rules found at 40 CFR 53.14.
    Another challenge to consider is how to deal with 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. One of the reasons to use data from the AQS 
database is that there are tens of thousands of collocated data pairs 
available that include many of the approved continuous PM2.5 
FEMs. Having a large data set will diminish the effect of any one or 
more outliers. However, acknowledging that the goal of this proposed 
change is to update factory calibrations to increase the number of 
routinely operating FEMs meeting MQOs across the networks in which they 
are operated, we propose that instrument companies may, but are not 
required to, check for and exclude any potential outliers. 
Additionally, we propose that the range of data may be limited to those 
concentrations that are within the normal operating ranges of most 
sites, but this is not required. This approach, for example, could 
include 24-hour average PM2.5 concentrations up to the level 
of the primary 24-hour PM2.5 NAAQS or some percentile above 
that level (e.g., 125% of the 24-hour NAAQS). The rationale for this is 
that there are very few sites with routine concentrations above the 
level of the primary 24-hour PM2.5 NAAQS, and the 
establishment of any equation with this data would need to be 
constructed carefully to avoid having data below the primary annual 
PM2.5 NAAQS drive the coefficients used above the level of 
the primary 24-hour PM2.5 NAAQS.
    Ideally, the geographic coverage of the data used in establishing a 
new factory calibration would be national in scope; however, instrument 
companies can only use the data that is available. For widely used 
PM2.5 FEMs, this will not be an issue, but for less-operated 
PM2.5 FEMs, there may be limitations in the geographic scope 
of data produced. Another challenge may be a large grouping of sites in 
one part of the country that drives development of an equation used 
across all networks. Instrument companies may limit the use of sites 
with large groupings in one or more geographical area so that the data 
are more geographically representative across the network so long as 
there is a reasonable rationale as to why data from certain sites are 
not being included. With a new factory calibration available, 
instrument companies will need to test the performance of the updated 
calibration across a variety of sites. Testing of an updated factory 
calibration can be accomplished by utilizing a different year or years 
other than the time-period used to establish the revised factory 
calibration or a subset of data across all years. Testing should also 
include the range of sites in which the method is used.
    Building off the geographic location of the sites in which an 
updated factory calibration is tested with previously collected data, 
the EPA considered what performance level should be acceptable. 
Ideally, an updated factory calibration would work such that a 
significantly larger number of, or all, individual sites operating with 
the updated factory calibration would meet the MQOs. However, due to 
several complicating factors such as seasonal changes in temperature 
and humidity, elevation, differences in aerosol composition, and 
differences in concentration between more polluted urban sites and 
relatively cleaner rural sites (some of which read well below the 
proposed revisions to the level of the primary annual PM2.5 
NAAQS discussed in section II above), the EPA should not expect that 
every site will necessarily meet the MQOs. Therefore, the goal of this 
proposal is to increase the number of routinely operating FEMs meeting 
MQOs across the networks in which they are operated, especially for 
sites near the level of the NAAQS proposed elsewhere in this proposal. 
Since there are multiple MQOs to consider, the EPA proposes to place 
the most attention on improvements to the bias MQO goal because this 
statistic will likely have the most influence on improving the 
resultant data collected. In attempting to address this goal, 
instrument companies may be interested in testing their original data 
used in field studies of their candidate FEMs with an updated

[[Page 5672]]

factory calibration. While this could be a useful exercise to 
understand the sensitivity of the original and any updated factory 
calibration, the EPA proposes not to require meeting the performance 
criteria of the original field testing as a condition of approving an 
updated factory calibration.
    Regarding how frequently factory calibrations should be updated, 
the EPA believes it would be most appropriate to not define a specific 
time-period for updates. Rather, updates should be based on the 
available of quality data being produced across the network. Monitoring 
agencies routinely check their data comparability to collocated FRMs, 
including as part of annual data certification where an AMP-256 report 
describing data quality is included as part of the certification 
package (Sec.  58.15(c)). In addition, monitoring agencies typically 
provide a more thorough review of their networks and accompanying data 
quality as part of the five-year assessments due to the EPA pursuant to 
40 CFR 58.10(d).
    Another important aspect to implementing updated factory 
calibrations is the treatment of data already collected under the 
original factory calibration. There are two time periods to consider. 
First, there is the time-period before the EPA approves an updated to a 
factory calibration. We propose that data collected prior to an 
approved update to a factory calibration be allowed to remain as 
measured based on the factory calibration that was approved at the time 
the data was collected. Second, there is the time-period between when 
an updated factory calibration is approved by the EPA and when that 
updated calibration is implemented in the field. While ideally, this 
time-period would be short, there may be reasons why some agencies and 
the sites they run cannot easily update the firmware with the updated 
factory calibration. We solicit comment on how to handle these 
situations and whether there should be an allowance to correct such 
data.
    The EPA sought early input from State, local, and Tribal monitoring 
agencies (NACAA Monitoring Committee 01/20/22; AAPCA Ambient Monitoring 
Committee 01/26/2022; Tribal air quality professionals call on 02/17/
22) regarding how best to address the issue of some PM2.5 
FEMs having bias issues. Many monitoring agencies identified that they 
strongly favor a national solution that can be accomplished and 
implemented through a firmware upgrade or similar resolution that is 
consistent with the approach described above. One State suggested that 
the EPA should consider and allow site-by-site corrections between FRM 
and collocated FEMs with ongoing collocation at a 1:6 sample frequency 
for FRMs. The rationale for site-by-site corrections was that there are 
differences in the types of aerosol composition and concentration 
between urban and rural locations and having site-by-site corrections 
would ensure that each type of location is individually calibrated to a 
collocated FRM rather than to a consistent factory calibration that may 
average out any differences. In contrast, other monitoring agencies 
expressed concern about the challenges of implementing a site-by-site 
approach, especially for those agencies who stated that they would not 
be able to redeploy the FRMs that would be necessary to perform the 
site-by-site corrections in their networks for reasons including no 
longer having FRMs, not having staff available to support and operate 
the FRMs, and no longer have gravimetric laboratory capacity to support 
a larger inventory of FRMs operating in their networks.
    The CASAC also provided input on the FEM bias issue. As part of 
their review of the draft PA, the CASAC stated that ``the FEM bias 
needs to be addressed to make the FRMs and FEMs more comparable'' 
(Sheppard, 2022a, p. 2 of consensus responses). The CASAC offered two 
options for the EPA to consider. ``One option would be to allow states 
to develop correction factors for co-located FRMs and FEMs. These 
correction factors could be used to adjust FEM concentrations downward 
(or upward) to be comparable to FRMs. Another option would be for the 
EPA to revise the `equivalency box' (EB) criteria used to judge whether 
the bias of a new continuous PM2.5 monitor relative to an 
FRM is acceptable during field testing'' (Sheppard, 2022a, p. 2 of 
consensus responses). The CASAC's first option is consistent with the 
input received during early input described above. The EPA believes 
that the second option should be considered in future reviews of the PM 
NAAQS to help establish updated goals for data quality from 
PM2.5 FEMs. The existing network of commercially available 
PM2.5 FRMs and some of the continuous FEMs are already 
meeting the MQOs at the existing concentrations, which are at or below 
the proposed revisions to the level of the primary annual 
PM2.5 NAAQS discussed in section II above. However, the EPA 
recognizes that not all PM2.5 FEMs are meeting MQOs and, 
therefore, the EPA intends to address improvements to existing FEMs 
that are not meeting MQOs as described above.
    In attempting to address the comparability of PM2.5 FEMs 
to collocated FRMs through our proposal to allow updates to factory 
calibrations, the EPA recognizes that other potential solutions do not 
need to be mutually exclusive. That is, there can be multiple 
approaches to improve the comparability of PM2.5 FRMs to 
continuous FEMs. Therefore, the EPA solicits comment on additional ways 
to improve PM2.5 data comparability between PM2.5 
FRMs and collocated continuous FEMs.
    The EPA encourages early dialogue with instrument companies 
considering an update to any part (e.g., hardware, software, and/or 
firmware revision) of an approved FEM designation. Dialogue with the 
EPA as well as applications by instrument manufactures can be initiated 
by contacting the EPA ORD's Reference and Equivalent (R&E) Methods 
Designation program. The contact information for this can be found at 
40 CFR 53.4, ``Applications for reference or equivalent method 
determinations.''
    In summary, the EPA is proposing that valid State, local, and 
Tribal air monitoring data generated in routine networks and submitted 
to the EPA may be used to update factory calibrations included as part 
of approved FEMs. This approach, initiated by instrument manufacturers, 
subject to EPA approval, would be implemented as a national solution in 
factory calibrations of approved FEMs through a firmware update. This 
would apply to any PM FEM methods (i.e., PM10, 
PM2.5, and PM10-2.5). As part of this process, 
the EPA proposes 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 two 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 five 
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 solicits input on this 
approach and 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

[[Page 5673]]

proposed to be revised in section II above.
4. Proposed Amendment 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 PM2.5 annual standard, the EPA 
proposes to modify our PM2.5 monitoring network design 
criteria to include an environmental justice factor that accounts for 
proximity of populations at increased risk of adverse health effects 
from PM2.5 exposures to sources of concern. Specifically, 
the EPA proposes to modify our existing requirement (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) \162\ 
populations, minority populations (particularly Black populations), and 
people with certain preexisting diseases (particularly cardiovascular 
disease and asthma). The EPA is proposing 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.
---------------------------------------------------------------------------

    \162\ 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 has led to a robust 
national network of PM2.5 monitoring stations. These 
monitoring stations are largely in Core-Based Statistical Areas (CBSAs) 
\163\ across the country that include many PM2.5 monitor 
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 if the level of the annual NAAQS is lowered, 
characterizing localized air quality issues may become even more 
important around local emission sources. The EPA believes that adding a 
network design requirement to specifically locate monitors in at-risk 
communities will improve our characterization of exposures for at-risk 
communities where localized air quality issues may exist. Requiring the 
siting of PM2.5 monitoring stations in at-risk communities 
allows other methods to be operated alongside PM2.5 
measurements to support multiple monitoring objectives (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 these 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 regulatory air quality agencies to assist communities 
to reduce exposures and to help inform future implementation and 
reviews of the NAAQS.
---------------------------------------------------------------------------

    \163\ CBSAs--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.
---------------------------------------------------------------------------

    As described in section II.B.2 above and in more detail in the PA 
(U.S. EPA, 2022b, section 3.3.2), the public health implications of 
health effects associated with PM2.5 in ambient air are 
dependent on the type and severity of effects, as well as the size of 
the population affected and whether there are populations and/or 
lifestages at increased risk of a PM2.5-related health 
effect. 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, 
2019, p. 12-1). Factors that may contribute to increased risk of 
PM2.5-related health effects include lifestage, pre-existing 
diseases (cardiovascular disease and respiratory disease), race/
ethnicity, and socioeconomic status. The increased risk faced by these 
sub-populations raises environmental justice \164\ concerns. Section II 
of this preamble, section 12.5 of the 2019 ISA (U.S. EPA, 2019a) and 
section 3.3.3 of the ISA Supplement (U.S. EPA, 2022a) provide extensive 
discussion on the evidence for disparities in PM2.5 
exposures and PM2.5-related health risks of these sub-
populations.
---------------------------------------------------------------------------

    \164\ The EPA defines environmental justice as the fair 
treatment and meaningful involvement of all people regardless of 
race, color, national origin, or income with respect to the 
development, implementation, and enforcement of environmental laws, 
regulations, and policies. The EPA further defines the term fair 
treatment to mean that ``no group of people should bear a 
disproportionate burden of environmental harms and risks, including 
those resulting from the negative environmental consequences of 
industrial, governmental, and commercial operations or programs and 
policies.''
---------------------------------------------------------------------------

    Consistent with the requirement of the Clean Air Act to protect 
sensitive sub-populations, the EPA is particularly concerned with 
protecting sub-populations identified as being at higher risk of 
adverse health effects from PM2.5 exposure in the 2019 ISA, 
ISA Supplement and PA (U.S. EPA, 2019a; U.S. EPA, 2022a; U.S. EPA, 
2022b). The EPA finds it appropriate to better characterize the 
localized air quality in communities with relatively higher proportions 
of these sub-populations to ensure these sub-populations receive the 
intended level of protection of a revised NAAQS proposed earlier in 
section II. Thus, the EPA is proposing to modify the PM2.5 
ambient monitoring network design criteria to add a provision 
pertaining to sub-populations identified as at increased risk for 
PM2.5 exposures and health risks associated with 
PM2.5 (``at-risk communities'').
    An enhanced network should include representation of at-risk 
communities who live near emission sources of concern such as, but not 
limited to, major ports, rail yards, airports, industrial areas, or 
major transportation corridors. The EPA finds it appropriate, in light 
of the evidence of increased risk to these communities, to better 
characterize exposures given proximity to local sources of concern. For 
example, the EPA believes it is worthwhile to characterize localized 
ambient concentrations occurring when there are emission sources 
located in a part of a metropolitan area that are different than the 
design value \165\ site of the same metropolitan area. Thus, while 
there may be sites with higher overall maximum concentrations in 
another part of the same metropolitan area, those sites are covered by 
our long-standing existing requirement that monitors be placed ``. . . 
in the area of expected maximum concentration'' [Sec.  58.1 and 
appendix D, section 4.7.1(b)(1)].
---------------------------------------------------------------------------

    \165\ Design value is defined in Sec.  58.1 as the calculated 
concentration according to the applicable appendix of 40 CFR part 50 
for the highest site in an attainment or nonattainment area.
---------------------------------------------------------------------------

    PM2.5 concentrations have generally trended down when 
averaged across all monitoring sites over the last two decades since 
PM2.5 measurements

[[Page 5674]]

commenced nationally in 1999.\166\ This downward trend has resulted in 
lower background concentrations being measured upwind of urban areas; 
however, the impact of local emissions on PM2.5 may not be 
known if there is not a requirement to monitor ambient air in these 
areas. For example, the presence of new local sources of fine particle 
air pollution proximate to at-risk communities, such as significant 
increases in heavy duty truck traffic since monitors were originally 
sited, should be taken into consideration. As explained in the PA (U.S. 
EPA, 2022b), measured PM2.5 at near-road monitoring stations 
include an increment relative to other sites in the same CBSA. The 
near-road sites will complement any new or moved sites located to 
specifically address at-risk communities near sources of concern. We 
anticipate the significance of local emissions may increase if, as 
proposed, the level of the annual PM2.5 NAAQS is lowered. 
Thus, the EPA seeks to support communities with at-risk populations in 
proximity to local sources of concern so that they have access to 
PM2.5 NAAQS-comparable data to ensure compliance with the 
PM2.5 NAAQS and for other data uses.
---------------------------------------------------------------------------

    \166\ See: https://www.epa.gov/air-trends/particulate-matter-pm25-trends.
---------------------------------------------------------------------------

    To successfully select and deploy an ambient air monitoring 
station, monitoring agencies must comply with the requirements of the 
EPA network design criteria (40 CFR part 58, appendix D, section 4.7), 
consider input from the community and other interested stakeholders, 
and then overlay the requirements and input with logistically available 
options in the neighborhoods they intend to monitor. Often, monitoring 
agencies partner with schools and other government agencies that have 
access to property in a neighborhood so that the desired monitoring 
stations can be sited, deployed, and maintained. Locating monitoring 
stations in neighborhoods should be done in a way that provides a good 
representation of the particulate matter exposures of the communities 
in which they are located. Alternatively, monitoring stations can be 
located directly next to emission sources of concern. However, these 
locations, known as ``source-oriented'' sites, may not necessarily 
represent the exposures in community or the effect of a multitude of 
emissions that can impact a neighborhood.
    To ensure monitoring sites are appropriately representing exposure 
in at-risk communities, we propose that sites represent ``area-wide'' 
air quality near local sources of concern. 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).\167\ 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 [part 58, appendix D, 
section 4.7.1(c)(3)]. Additionally, as described in Sec.  58.30, sites 
representing ``area-wide'' air quality have a long-standing 
applicability to both the annual and 24-hour PM2.5 NAAQS. 
Siting in a community representing ``area-wide'' air quality as 
proposed is consistent with other network design objectives pursuant to 
which we locate monitors where people live, work, and play.
---------------------------------------------------------------------------

    \167\ 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 types of sites that are minimally required as part of the 
PM2.5 network design are associated with two geopolitical 
levels: MSAs and states. The minimum number and type of sites that are 
required within an MSA are a function of the population of the MSA, 
based on the latest available information from the Census Bureau, and 
the design value of the existing network of PM2.5 sites 
reported for that MSA. MSAs with design values at or above 85% of any 
PM2.5 NAAQS are required to operate one more site than those 
MSAs with values that are less than 85% of any PM2.5 NAAQS 
(40 CFR part 58, appendix D, Table D-5). Each MSA required to operate 
at least one monitoring station is to site the monitor at neighborhood 
or larger scale in an area of expected maximum concentration. MSAs with 
a population of 1 million or more are required to operate a 
PM2.5 monitor at a NO2 near-road station in the 
same MSA. Thus, according to Table D-5 of appendix D to part 58, only 
those MSAs with a population of greater than 1 million with the most 
recent 3-year design value greater than or equal to 85% of any 
PM2.5 NAAQS are required to operate at least three 
PM2.5 monitoring stations. Since one of these sites would be 
the site in the area of expected maximum concentration, which most 
often will be the design value site, and the other the near-road site, 
only the third location would not address either of those two 
requirements.
    The requirement for a third monitoring station in a MSA, where it 
exists, would take on the revised network design requirement to address 
at-risk communities near sources of concern. Many existing sites in the 
area of expected maximum concentration or near-road sites that are 
located in at-risk communities. Thus, having multiple sites located in 
at-risk communities may be appropriate so long as each siting criteria 
is achieved. Also, while we are proposing this modification to our 
network design criteria, we recognize that the number of monitors to 
support key monitoring objectives, including addressing at-risk 
communities, could go well beyond what is currently minimally required. 
Many monitoring agencies already operate more monitoring sites than are 
minimally required and we expect this to continue in considering siting 
monitors in at-risk communities. Thus, the existing and robust network 
of almost 1,000 PM2.5 sites nationally will continue to 
protect all populations at the level of the NAAQS discussed in section 
II of this proposal, by always having at least one site in the area of 
expected maximum concentration for each CBSA where monitoring is 
required. Many existing and a few new sites will form an important sub-
component of the PM2.5 network by characterizing air quality 
in at-risk communities, particularly with respect to sources of 
concern.
    Monitoring requirements applicable at the state level include 
measuring regional background and regional transport (40 CFR part 58, 
appendix D, section 4.7.3). These required sites at the state level are 
largely located in rural areas and may include use of IMPROVE samplers 
or continuous PM2.5 monitors. The sites required at the 
state level complement sites required at the MSA level. Together the 
sites already required at the state level combined with existing siting 
requirements at the MSA level as well as the proposed revisions 
described herein to address at-risk communities will achieve several 
monitoring objectives, including comparison to the NAAQS and AQI. The 
availability of data from regional background and regional transport 
sites compared to data from design value sites already allow for 
calculating incremental exposure in communities with the highest design 
value location. With the proposed addition of a siting requirement for 
at-risk communities and the use of data from these sites compared to 
select regional background and regional transport sites as well as 
other sites in the same MSA, we can

[[Page 5675]]

assess the incremental burden of exposure from local emissions to at-
risk communities.
    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 for trends and as an 
input to long term health and epidemiology studies that support reviews 
of the PM NAAQS. Therefore, while we are proposing to add a 
PM2.5 network design criteria 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 criteria we are proposing for at-risk communities. Thus, 
acknowledging the value of having long-term data from a consistent set 
of network sites, on balance the EPA believes that the movement of 
sites should be minimized, especially in MSA's with a small number of 
sites. However, a small number of new sites \168\ are expected to be 
required due to the existing minimum monitoring requirements (Table D-5 
of appendix D to part 58) and the revised primary annual 
PM2.5 NAAQS proposed in section II of this proposal. Also, 
sites do on occasion need to move due to loss of leases, no longer 
meeting siting criteria, or other reasons. For any of these cases, we 
believe it is appropriate to include prioritizing establishing sites in 
at-risk communities near sources of concern, should new sites be 
established, or existing locations be lost, and replacement sites need 
to be identified. Therefore, the EPA proposes that annual monitoring 
network plans [40 CFR 58.10(a)(1)] that include the few newly required 
sites and five-year assessments [40 CFR 58.10(d)] include a provision 
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.
---------------------------------------------------------------------------

    \168\ 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.
---------------------------------------------------------------------------

    Assessing and prioritizing at-risk communities for monitoring can 
be accomplished through several approaches. 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 source(s) of 
concern. While many CBSA's have one or more sources of concern 
described above, some CBSA's will not have the level of emissions from 
sources of concern that result in an elevated level of measured 
PM2.5 concentrations in surrounding communities. Since one 
of our other siting criteria to ``. . . be in the area of expected 
concentration'' [Sec.  58.1 and 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, on balance the EPA 
believes we should include being in an at-risk community for CBSAs with 
a third site requirement when there are no sources of concern 
identified in a CBSA or such sources do exist but are not expected to 
lead to elevated levels of measured PM2.5 concentrations.
    To identify at-risk communities to consider for the proposed 
monitoring requirement, tools such as the EPA's EJSCREEN \169\ are 
available. The EPA solicits comment on other tools and/or datasets that 
can could be utilized to identify the at-risk communities described 
above. With information on 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, 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. 
Potential datasets include non-regulatory data (CSN, IMPROVE, and AQI 
non-regulatory PM2.5 continuous monitors), modelling data--
which utilizes emission inventory and meteorological data, emerging 
sensor networks such as used in the EPA's AirNow fire and smoke 
map,\170\ and satellites--which measure radiance and with computational 
algorithms are then used to estimate PM2.5 from aerosol 
optical depth (AOD). The 2019 ISA and PA (U.S. EPA, 2019a; U.S. EPA, 
2022b) include details on each these, except for the AirNow fire and 
smoke map, which first became operational in 2020. Each of these 
datasets have advantages and disadvantages, especially when attempting 
to determine exposure concentrations for the averaging times of the 
PM2.5 NAAQS described in section II (i.e., annual NAAQS and 
24-hour NAAQS). The EPA solicits comment on datasets most useful to 
identify communities with high exposures for PM2.5 NAAQS 
(i.e., annual or 24-hour), including any discussion on limitations or 
advantages of the dataset of interest. The EPA is soliciting comment on 
the use of these datasets for the purpose of identifying communities 
where the proposed monitoring requirement would apply and not for the 
purpose of satisfying the proposed monitoring requirement.
---------------------------------------------------------------------------

    \169\ See: https://www.epa.gov/ejscreen.
    \170\ See: https://fire.airnow.gov/.
---------------------------------------------------------------------------

    The monitoring methods appropriate for use at these proposed sites 
are FRMs and automated continuous FEMs. These are the methods that are 
eligible to compare to the PM2.5 NAAQS, which will be the 
primary objective for collecting this data. There are several other 
monitoring objectives that would benefit from 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 indicating 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. The EPA and the CASAC worked collaboratively 
in 2010 (Russell and Samet, 2010) to define a list of measurements that 
would be useful to implement in the near-road environment, and a subset 
of these measurements may additionally be of value to characterize the 
exposure in at-risk communities. While either FRMs or automated FEMs 
may be used at a site for comparison to the PM2.5 NAAQS, the 
EPA encourages use of automated continuous FEMs at sites in at-risk 
communities.
    Although there are only a few new sites required,\171\ plus any 
potentially moved sites in cases where a site lease is lost, EPA 
believes we should build upon our existing regulatory process for 
selecting and approving these sites (40

[[Page 5676]]

CFR 58.10). For example, the timeline to implement the proposed 
PM2.5 sites in at-risk communities should allow monitoring 
agencies enough time for communities and other interested parties to 
provide their input regarding moving or adding new sites, while also 
minimally disrupting ongoing operations of monitoring agency programs. 
Another important factor is to ensure all existing PM2.5 
sites have data available for comparison to a revised PM2.5 
NAAQS, which is discussed in section II of this proposal. With a final 
rule from this proposal expected in 2023, we believe it would be 
appropriate to provide at least 12 months from the effective date of a 
final rule for monitoring agencies to initiate planning to implement 
these measures by seeking input from communities and other interested 
parties, and to consider revisions to their PM2.5 networks 
or explain how the existing network meets the objectives of this 
proposed modification. Thus, the EPA proposes that monitoring agencies 
identify their initial approach to the question of whether any new or 
moved sites are needed and to identify the potential communities in 
which the agencies are considering adding monitoring, if applicable, as 
well as identifying how they intend to meet the proposed revised 
criteria for PM2.5 network design to address at-risk 
communities. These aspects that will potentially affect the siting of 
new and moved sites should be addressed in the agencies' annual 
monitoring network plans due to each applicable EPA Regional office no 
later than July 1, 2024 (40 CFR 58.10). Specifics on the resulting 
proposed new or moved sites for PM2.5 network design to 
address at-risk communities would need 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). We are proposing 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 solicits comment 
on whether less time is needed (e.g., 12 months from plan approval and/
or January 1, 2026).
---------------------------------------------------------------------------

    \171\ 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 summary, the EPA is proposing to modify our PM2.5 
network design criteria to include an environmental justice factor to 
address at-risk communities with a focus on exposures from sources of 
concern. While this proposal would require that sites be located in at-
risk communities, particularly those whose air quality is potentially 
affected by local sources of concern, such sites should still meet the 
requirement for being considered ``area-wide'' air quality. Specific 
areas of interest we seek comment on include how to identify at-risk 
communities, the sources of concern important to consider, the datasets 
to identify communities with high exposures, and the most useful 
measurements to collocate with PM2.5 in at-risk communities. 
The EPA seeks comment on these areas of interest as well as the 
proposed modification of our PM2.5 network design objectives 
and implementation as described herein.
5. Proposed Revisions to Probe and Monitoring Path Siting Criteria
    The EPA is proposing changes to monitoring requirements in the 
Appendix E--Probe and Monitoring Path Siting Criteria for Ambient Air 
Quality Monitoring. Since 2006, multiple rule revisions were made 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 multiple revisions to the regulatory text, some 
requirements were inadvertently omitted, and, over time, the clarity of 
this appendix was reduced through these omissions that, in a few 
instances, led to unintended and conflicting regulatory requirements. 
The EPA proposes to reinstate portions of previous Probe and Monitoring 
Path Siting Criteria Requirements from previous rulemaking where 
appropriate to restore the original intent. The proposed changes that 
affect the overall appendix follow, while those specific to the various 
sections of the appendix will be addressed under a specific section 
heading. The EPA notes that appendix E is being reprinted in its 
entirety with this proposal because this section is being reorganized 
for clarity in addition to being selectively revised as described in 
detail below. The EPA is soliciting comment on the specific provisions 
of appendix E proposed for revision. However, there are a number of 
provisions that are being reprinted solely for clarity to assist the 
public in understanding the changes being proposed and reconciling 
requirements between different portions of the text; the EPA is not 
soliciting comment on those provisions and considers changes to those 
provisions to be beyond the scope of this proposed rulemaking.
a. Providing Separate Section for Open Path Monitoring Requirements
    The current appendix E regulation combines open path monitor siting 
requirements with requirements for siting samplers and monitors that 
utilize probe inlets. While this approach allowed the EPA to promulgate 
an abbreviated regulation for probe-siting requirements, the EPA now 
has determined that the clarity of the requirements for each monitoring 
method type has been diminished by this combination. As such, the EPA 
is proposing to relocate all open path monitor siting criteria 
requirements to a separate section in this appendix. Providing separate 
sections for these distinct monitoring method types will allow the EPA 
to more clearly articulate minimum technical siting requirements for 
each. Further rationale for creating these separate sections is that 
the regulatory monitoring community has not submitted to AQS 
measurement results from open path monitors since 2009. Because these 
open path monitoring methods are rarely used for monitoring to compare 
to the NAAQS, the EPA believes that moving the open path siting 
criteria to their own section will make clearer the probe siting 
criteria for the ambient air monitoring methods that are now most 
commonly utilized by monitoring organizations.
b. Amending Distance Precision for Spacing Offsets
    The EPA proposes 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. The EPA proposes to communicate this rounding requirement in 
the regulatory text using footnotes in Table E-1, Table E-2, and Table 
E-3 of the current regulation.
c. Clarifying Summary Table of Probe Siting Criteria
    To provide additional specificity and flexibility to the summary 
table for probe siting criteria (see current Table E-4 in appendix E), 
the EPA proposes to change the ``>'' (greater than) symbols to ``>='' 
(greater than or equal to) symbols. This minor revision will more 
clearly express the EPA's intent that the distance offsets provided in 
the current Table E-4 in appendix E are acceptable for NAAQS compliance 
monitoring.
d. Adding Flexibility for the Spacing From Minor Sources
    Current requirements for the spacing of probe inlets and monitoring 
paths from minor sources of SO2 and NO2

[[Page 5677]]

stipulate that the probe inlets and monitoring paths must be away from 
these minor sources (see current section 3(b) in appendix E). The EPA 
proposes to clarify and provide flexibility by changing this 
requirement to a goal. The EPA proposes to replace the ``must'' in this 
regulation with a ``should''. As stated in section 1(c) of the current 
rule, a ``must'' defines a requirement while a ``should'' specifies a 
goal. Since the current rule does not specify how far the probe must be 
spaced from such minor sources, the EPA proposes that a ``should'' in 
this regulation is more appropriate. Minor sources can have adverse 
impacts on the representativeness of the ambient pollutant 
concentrations sampled by the probe inlet. As such, the EPA recommends 
that sites with these minor sources be avoided whenever practicable and 
probe inlets spaced as far from these minor sources as possible when 
alternative monitoring stations are not suitable.
e. Amendments and Clarification for the Spacing From Obstructions and 
Trees
    The EPA proposes to clarify and redefine that the minimum arc 
required to be free of obstructions for a probe inlet or monitoring 
path is 270 degrees. Currently this portion of the regulation (see 
current section 4(b) of appendix E) specifies 180 degrees as this 
minimum arc. However, this requirement is inconsistent with the 
requirement found in footnote 5 of Table E-4 in appendix E that 
specifies the probe inlet or monitoring path must have unrestricted 
airflow of 270 degrees around the probe and 180 degrees for the arc is 
only allowed if the probe is on the side of a building or a wall. These 
inconsistent regulatory requirements were introduced in the 2006 
rulemaking when the 270-degree requirement was omitted from the text of 
section 4(b) (see 71 FR 61236, October 17, 2006).
    There are also inconsistent requirements in the current regulation 
regarding the spacing of probe inlets from the driplines of trees. 
Section 5(a) of appendix E requires the probe inlet must be no closer 
than 10 meters to the driplines of any trees, while footnote 3 of Table 
E-4 of the appendix E qualifies that this minimum 10-meter offset is 
only required when the tree also acts as an obstruction.
f. Reinstating Minimum 270-Degree Arc and Clarifying 180-Degree Arc in 
Regulatory Text
    The EPA proposes to correct identified inconsistencies in this 
regulation by reinstating the 270-degree requirement in section 4(b) of 
appendix E. Additionally, the EPA proposes to further clarify this 
regulation by stating that the continuous 180-degree minimum arc of 
unrestricted airflow provision is reserved for monitors sited on the 
side of a building or a wall to comply with network design criteria 
requirements specified in appendix D of part 58. Examples include CO 
monitoring in urbanized areas that relies on monitoring in street 
canyons and near-road monitoring where a continuous arc of 270 degrees 
of unrestricted airflow is not routinely possible given limited monitor 
siting options.
g. Clarification on Obstacles That Act as an Obstruction
    The EPA proposes to clarify the definitions of ``obstructions'' and 
``obstacles'' in the regulatory text (see section 4 of the current 
appendix E). While obstacles should be avoided as much as is 
practicable, logistical constraints may dictate that some obstacles are 
present within the vicinity of the monitoring probe inlet. Obstructions 
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 does not generally consider objects or obstacles such as flag 
poles or site towers for NOy convertors or towers for 
meteorological sensors, etc. to be obstructions.
h. Amending and Clarifying the 10-Meter Tree Dripline Requirement
    The EPA proposes to reconcile the conflicting requirements in 
section 5(a) and Table E-4, footnote 3 of the current regulation by 
deleting the qualification in footnote 3 of Table E-4 to require that 
the probe inlet must always be no closer than 10 meters to the tree 
dripline. The EPA also proposes to reinstate the goal that was omitted 
from section 5(a) during previous rule revisions, that monitor probe 
inlets should be at least 20 meters from the driplines of trees. 
Additionally, the EPA proposes to clarify section 5(a) of the current 
regulation by adding that when the tree or group of trees is considered 
an obstruction, then the regulatory requirements of section 4(a) apply.
i. Amending Spacing Requirement for Microscale Monitoring
    To obtain representative ambient air monitoring measurements for 
source-oriented and microscale air monitoring stations, it is important 
to have unobstructed airflow between the monitor's probe inlet and the 
source under investigation. This reasoning was used by the EPA when 
near-road NO2 monitoring stations were required to have an 
unobstructed airflow between the monitor probe and the outside nearest 
edge of the traffic lane (see current section 4(d) of this regulation). 
To assist in further clarifying the monitoring siting criteria for the 
spacing from obstructions and spacing from trees, the EPA proposes to 
change from a goal to a requirement 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. The EPA proposes to communicate this requirement by 
changing the ``should'' to a ``shall'' in the regulatory text of 
section 5(c). The EPA does not consider small obstacles such as shrubs 
that are below this fetch to adversely impact the representativeness of 
the air quality measurements results. This proposed revision of section 
4(d) will bring more consistency to appendix E.
j. Amending Waiver Provisions
    The EPA believes the effects of any requirements in this proposal 
that may be considered to be new are minor. While we are attempting to 
clarify probe and siting criteria as part of our monitoring 
regulations, the Agency fully intends to maintain waiver provisions 
that exist in the regulation for these siting criteria (see current 
section 10). 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 these siting criteria is appropriate.
    Even though the current regulation adequately and clearly 
identifies which monitoring situations are eligible for the EPA to 
consider waiving the requirements for probe-siting criteria (see 
current section 10), these waiver provisions are silent regarding how 
long an approved waiver remains in force and effect. Environmental 
conditions (e.g., airflow due to changes in growth of trees, shrubs, 
construction of buildings or other obstructions) around monitoring 
stations are prone to change over time. As such, the EPA has identified 
that previously approved waivers should be periodically reevaluated to 
ensure that the conditions upon which the original waiver was approved 
still exist and that the siting conditions have not degraded to an 
unacceptable level. The EPA proposes to modify section 10.3 of the

[[Page 5678]]

current regulation to state that waivers from the probe-siting criteria 
must be renewed minimally every 5 years. Ideally, sites needing a 
waiver renewal should be inspected by the EPA such as during a 
Technical Systems Audit (TSA) typically conducted at a subset of sites 
within each Primary Quality Assurance Organization (PQAO) every three 
years. However, virtual inspections may also be acceptable using 
documentation such as photos and traffic counts. Dates for the most 
recent approval of a waiver must then be included in the applicable 
network assessment and annual monitoring network plan. The EPA proposes 
to revise Sec.  58.10(b)(10) of the regulation to maintain consistency 
in the text for probe siting criteria requirements and annual 
monitoring network plans. This proposal leverages the existing annual 
assessment requirements found in Sec.  58.10(a)(1) and (d).
k. Broadening of Acceptable Probe Materials
    The current regulatory specifications for acceptable probe 
materials for sampling reactive gases are limited to borosilicate 
glass, fluorinated ethylene propylene (FEP) Teflon[supreg], or their 
equivalent (see section 9 of the regulation). The EPA's selection of 
``or its equivalent'' in the current regulatory text was intended to 
allow flexibility to monitoring organizations when selecting suitable 
sampling train materials. In practice, however, this text has resulted 
in potentially suitable materials not being used for sampling trains 
due to concerns that the material may not meet these regulatory 
requirements. The current requirements for acceptable probe materials 
were promulgated in 1979. Since 1979, several potential alternatives to 
borosilicate glass and FEP were developed and are commercially 
available.
    Because some of these alternative materials have advantages over 
the currently approved materials (e.g., cost and durability), the EPA 
has received numerous inquiries from monitoring organizations regarding 
the regulatory suitability of these materials. Monitoring organizations 
have expressed particular interest in the potential use of PVDF 
(polyvinylidene fluoride) which is marketed under the registered 
tradename of Kynar[supreg] by Arkema Inc. (Colombes, France). In 
response to these inquiries, the EPA's Office of Research and 
Development (ORD) recently designed and conducted a laboratory study to 
determine the transport efficiency of O3, SO2, 
NO2, and CO through several candidate tubing materials 
(Johnson, 2022). Based on these tests results, the EPA is proposing to 
revise Section 9 of the current regulation to add polyvinylidene 
fluoride (PVDF), polytetrafluoroethylene (PTFE), and perfluoroalkoxy 
(PFA) to the list of approved materials for efficiently transporting 
gaseous criteria pollutants. The EPA also proposes to clarify that the 
residence-time criteria for sampling reactive gas through these 
approved materials applies to all O3, SO2, and 
NO2 monitors. In conjunction with the previously approved 
borosilicate glass and FEP materials, including these three new 
materials would provide monitoring organizations with a wider variety 
of efficient sampling and transport materials needed for conducting 
NAAQS compliance monitoring.
    The EPA has also studied and approved the use of Nafion\TM\ 
upstream of ozone analyzers to minimize measurement bias associated 
with high ambient RH levels (U.S. EPA, 2020b). Minimal loss of ozone 
occurred in these systems as long as the Nafion\TM\ system was 
conditioned beforehand. Nafion\TM\ is composed primarily of PTFE and 
can be considered equivalent to PTFE. It has been shown in ORD's recent 
tests described above to exhibit virtually no loss of ozone at 20 
second residence times.

D. Taking Comment on Incorporating Data From Next Generation 
Technologies

1. Background on Use of FRM and FEM Monitors
    The EPA approves FRM and FEM monitors for criteria pollutant 
measurements in the Federal Register after careful review of 
applications describing extensive testing of the methods operation and 
performance. The siting of these monitors across State, local, and 
Tribal networks is subject to detailed requirements for network design 
detailed in appendix D to 40 CFR part 58 with probe and siting criteria 
described in appendix E for 40 CFR part 58. The operation of these 
monitors is subject to extensive quality assurance requirements 
detailed in appendix A to 40 CFR part 58, which ensures data quality 
statistics are produced to inform the quality of the data needed to 
ensure regulatory grade decisions are made with data of known quality. 
The EPA believes these requirements are important for ensuring the 
degree of accurate and precise data which is appropriate for regulatory 
decision-making, particularly decisions about attainment or 
nonattainment of the NAAQS. However, the EPA also recognizes that the 
capital and operating costs of these monitors is substantial, which 
requires the EPA and states to prioritize where monitors should be 
deployed. The EPA recognizes that making use of broader air quality 
data sets which are less expensive can provide important benefits, even 
if the EPA does not consider those datasets suitable for all regulatory 
purposes. In some circumstances in the past, for example, the EPA has 
used non-FRM monitoring to inform decisions about the boundaries of a 
nonattainment area, although the data was not sufficient to support a 
finding that an area was in nonattainment. Likewise, the EPA has 
incorporated sensor data into its fire and smoke map for the purpose of 
informing the public of potential imminent health risks, even though 
that data would not be comparable to the NAAQS for purposes of 
determining attainment. There are multiple uses of air quality data and 
the EPA believes there may be additional opportunities to develop 
broader air quality datasets which provide benefits to the EPA and the 
public even where the data is not from FRM/FEM monitors and is not 
suitable for comparison to the NAAQS.
2. Next Generation Technologies: Data Considerations
    The EPA and our State, local, and Tribal partners in cooperation 
with other Federal agencies have made great strides in integrating data 
from routine air monitoring methods with data from next generation 
technologies to address emerging air quality issues. For example, the 
EPA and U.S. Forest Service (USFS), in consultation with other 
partners, launched the publicly available AirNow Fire and Smoke 
Map,\172\ which has received over 26 million page views since its 
release in July 2020. This fire and smoke map has been an invaluable 
tool for the public, providing refined spatial information on current 
Air Quality Index (AQI) conditions, fire and smoke plumes locations, 
actions for communities to take based on local air quality, and links 
to Smoke Forecast Outlooks developed by specially trained air resource 
advisors. Data are brought together from multiple systems including 
permanent and temporary PM2.5 continuous monitoring sites, 
sensors, and satellite derived fire and smoke data. With the success of 
the fire and smoke map and a robust and growing network of 
PM2.5 continuous FEMs and sensor network data, as well as 
existing and future satellites products, the EPA is interested in 
considering further enhancements to

[[Page 5679]]

the evolution of data products to meet new and emerging non-regulatory 
air quality data needs. Below we describe each of the major data sets, 
their advantages, and any challenges to their use. We then solicit 
input on additional approaches and/or products to incorporating data 
from next generation technologies that can help address important non-
regulatory air quality data needs.
---------------------------------------------------------------------------

    \172\ See: https://fire.airnow.gov.
---------------------------------------------------------------------------

3. PM2.5 Continuous FEMs
    As described in the PA, State, local, and Tribal monitoring 
agencies are using an increasing number of PM2.5 continuous 
FEMs. These methods are primarily deployed to meet two monitoring 
objectives: first, to compare to the NAAQS, and second, to report and 
support forecasting of the AQI. PM2.5 continuous FEMs have 
some key advantages over FRMs, most notably that they provide automated 
hourly measurement of PM2.5 available in near real time. The 
continuous PM2.5 data are reported as soon as practicable 
after the end of each hour, usually within 5-10 minutes, and are used 
in multiple applications of real-time data such as such as by State, 
local, and Tribal websites,173 174 the EPA's AirNow website, 
and national media outlets. Recent improvements in the availability and 
exchange of near real-time data through a dedicated AirNow Application 
Programming Interface (API) allow for efficient exchange of data 
between the EPA, other Federal agencies, and commercial data providers 
such as low-cost sensor networks. The efficient exchange of data 
through the AirNow API was a key advancement in the successful 
implementation of the EPA AirNow's fire and smoke map. The 
PM2.5 continuous FEM data are critical to ``ground 
truthing'' other datasets such as sensors and satellites for two 
important reasons. First, PM2.5 continuous FEMs are subject 
to extensive regulatory-grade quality assurance and quality control as 
required by appendix A to 40 CFR part 58. Second, PM2.5 
continuous FEMs are located in accordance with strict siting criteria 
according to appendix E to 40 CFR part 58. The siting criteria assure 
that measured data represent ambient air at ground level where people 
are breathing and are thus exposed to particle pollution. The EPA and 
State, local, and Tribal agencies are working to upgrade many existing 
FRM-only sites with PM2.5 continuous FEMs through use of 
American Rescue Plan funds.\175\ Despite these investments, there are 
major challenges to monitoring agencies' ability to have enough trained 
and available staff to support their regulatory monitoring networks, 
especially in remote locations, and to have the capital resources to 
implement new monitoring stations. So, while there may be some 
improvements to the existing network of almost 1,000 PM2.5 
regulatory-grade monitoring stations, regulatory instruments will not 
produce data everywhere that it is desired. Thus, the integration of 
PM2.5 continuous FEMs with other datasets is an important 
opportunity to address existing and emerging air quality data needs for 
non-regulatory purposes.
---------------------------------------------------------------------------

    \173\ See: https://www.airnow.gov/partners/state-and-local-partners/.
    \174\ See: https://www.airnow.gov/partners/tribal-partners/.
    \175\ See: https://www.epa.gov/arp/enhanced-air-quality-monitoring-funding-under-arp.
---------------------------------------------------------------------------

4. PM2.5 Satellite Products
    Satellite-based instruments provide measurements of radiance that 
can be used to calculate the aerosol optical depth (AOD) of the 
atmosphere. For over a decade, satellite AOD values have been used in 
models that incorporate multiple datasets to predict surface level 
PM2.5 concentrations over the U.S. (hereafter, satellite-
PM2.5). Despite some heterogeneity in performance under 
varying conditions, the satellite-PM2.5 datasets have 
significantly advanced in terms of accuracy in recent years (Di et al., 
2019; van Donkelaar et al., 2019; Zhang and Kondragunta, 2021). The EPA 
is using satellite-PM2.5 datasets in a variety of contexts. 
Satellite-PM2.5 data was included in a comparative analysis 
of hybrid modeling methods in the PA (U.S. EPA, 2022b). The EPA is also 
working with the National Aeronautics and Space Administration (NASA) 
and National Oceanic and Atmospheric Administration (NOAA) to use 
satellite-PM2.5 in the AirNow system.\176\ The EPA also uses 
satellite AOD and many other satellite data products in the development 
of our photochemical modeling platforms that are used in regulatory and 
policy assessments both by the EPA and by our State and local partners.
---------------------------------------------------------------------------

    \176\ The EPA provided an update on the Health and Air Quality 
Applied Scientist Team (HAQAST) AirNow Project at the NASA HAQAST 
meeting in Texas in June 2022. For more information, see: https://haqast.org/haqast-houston-june-1-2/.
---------------------------------------------------------------------------

    Each satellite data product has its own strengths and limitations. 
One strength is the spatial coverage, which can be once-a-day globally 
for polar orbiting satellites or over a fixed field of view 
continuously for geostationary satellites. Satellite-PM2.5 
data has the limitation that it is not a direct measurement of 
PM2.5 concentrations, but rather is derived through a model 
that connects the total column AOD to surface PM2.5. In 
addition, the satellite products are only capable of making daytime 
measurements because they rely on sunlight. In fact, most satellite-
PM2.5 data products use the surface monitor network as an 
input. As such, the satellite-PM2.5 data does not substitute 
for a ground-based monitor; rather it complements the monitor network. 
The EPA continues to explore ways to use the wealth of data from 
satellites to address important air quality questions consistent with 
their strengths and limitations.
5. Use of Air Sensors
    The term ``air sensor'' is a simplified way of referring to a class 
of technology that has expanded on the market in recent years and has 
common traits of directly reading a pollutant in the air, being smaller 
in size, and often sold at lower prices that support a wider number of 
monitoring locations than possible in the past. As explained on the 
EPA's Air Sensor Toolbox website,\177\ air sensor monitors that are 
lower in cost, portable, and generally easier to operate than 
regulatory-grade monitors are widely used in the United States to 
understand air quality conditions. Many refer to this class of 
technology as ``low-cost air sensors,'' ``air sensor devices,'' or 
``air quality sensors.'' Potential uses for these non-regulatory air 
sensor technologies include, but are not limited to, science education, 
supplementing regulatory air quality measurements, conducting research, 
measuring local air quality to better understand sources of pollution, 
locating leaks at industrial facilities, and emergency response.
---------------------------------------------------------------------------

    \177\ See: https://www.epa.gov/air-sensor-toolbox.
---------------------------------------------------------------------------

    The growth in use of sensors included in the EPA's fire and smoke 
map provides a platform to build upon. There are thousands of PM 
sensors whose data are coordinated and overlaid with routine and 
temporary PM2.5 continuous monitors as well as satellite-
derived data on fires and smoke. Sensors offer an opportunity to 
supplement higher-cost regulatory monitoring to provide data for the 
non-regulatory uses as described above. However, there are several 
challenges to using sensors. Each commercially available PM sensor 
appears to have its own data quality challenges depending on season, 
aerosol encountered, and meteorological conditions (typically 
temperature and relative humidity). The EPA has gone to considerable 
length to ensure the PM2.5 sensor data on the fire and smoke 
map have a correction available with collocated FRMs and

[[Page 5680]]

FEMs.\178\ This was possible due to the large number of air sensors 
that are the same make and model located across the country. Thus, an 
important challenge for the use of sensors is the spatial richness in 
sensor networks needed to make integrating the dataset with other 
monitoring data viable. Even with corrected sensor data in hand, 
publicly shared sensor data lacks reliability and accountability for 
ensuring that basic siting criteria are met. Sensors are often 
installed by members of the public who share data to the sensor 
network, which is generally understood as implicitly representing that 
the sensor is located in ambient air although, in fact, the sensor may 
be located inside a home or next to a highly localized source of 
emissions such as the flue of a home heating system. In areas with many 
reporting sensors, these concerns about siting may be lessened through 
site-to site comparison of data; however, the absence of any confirmed 
information about siting presents challenges for use of sensor data.
---------------------------------------------------------------------------

    \178\ See: https://www.epa.gov/research-states/airnow-fire-and-smoke-map-extension-us-wide-correction-purpleair-pm25-sensors.
---------------------------------------------------------------------------

6. Summary
    The near real-time integration of data from PM2.5 
continuous monitors, sensors, and satellites has been proven through 
use of the EPA's fire and smoke map. This mapping product is possible 
though the use of API's where data sets are automatically shared on 
pre-specified computer servers. Given the success of the fire and smoke 
map, the EPA is interested in pursuing additional approaches and/or 
products that can help address important non-regulatory air quality 
data needs. Therefore, the EPA solicits comment on the most important 
data uses and data sets to consider in future products. Such approaches 
and/or products could utilize historical or near real-time data. For 
example, what are the advantages and disadvantages of using existing 
data and tools to identify PM hot spots across an area of interest? 
Could satellite data or a combined surface layer (PM2.5 FRM 
and FEM data, sensor data, and satellite data) be useful in siting 
regulatory monitors? Could combined surfaces layers be useful in 
determining the boundaries of nonattainment areas? Could combined 
surface layers be useful in exploring potential emission sources to 
consider in SIP planning? To what extent would requirements for data 
formats, units, or timescales of interest need to evolve to best 
address these needs? What other datasets should the EPA consider 
merging with the data sets listed above to help better inform air 
quality management, including prioritizing network investments for 
potential new sites such as in at risk communities described elsewhere 
in this proposal? The EPA seeks input and prioritization on each of 
these questions to help improve the utility of data to better support 
air quality management to improve public health and the environment.

VIII. Clean Air Act Implementation Requirements for the PM NAAQS

    The proposed revision to the primary annual PM2.5 NAAQS 
discussed in section II above, if finalized, would trigger a process 
under which states \179\ will make recommendations to the Administrator 
regarding area designations. States also will be required to review 
their existing section 110 infrastructure state implementation plans 
and modify them if necessary to implement a revised NAAQS. A revised 
primary annual PM2.5 NAAQS will need to be incorporated into 
the implementation of applicable air permitting requirements and the 
transportation conformity and general conformity processes, and states 
will need to review existing regulations for these programs that 
already cover PM2.5 to determine the extent to which any 
changes are needed. This section provides background information for 
understanding the possible implications of the proposed NAAQS changes 
and describes the EPA's plans for providing states guidance needed to 
assist their implementation efforts. This section also describes 
existing EPA interpretations of CAA requirements and other EPA guidance 
relevant to implementation of a revised PM2.5 NAAQS. Given 
the strong scientific evidence for disparities in PM2.5 
exposures and PM2.5-related health risk among certain 
populations (as discussed in section II of this document), the EPA 
included in its 2016 PM2.5 State Implementation Plan (SIP) 
Requirements Rule (81 FR 58010, August 24, 2016) (which was written to 
be applicable for any future NAAQS revisions) included a number of key 
recommendations for states to advance environmental justice through 
their attainment planning process. In addition, as discussed throughout 
this section, environmental justice considerations are evaluated with 
regard to the several specific program elements of the overall 
implementation process. State and local air agencies have a critically 
important role in implementing the NAAQS, including this proposed 
PM2.5 NAAQS, should it become finalized. Given the 
information provided in this proposed rulemaking, state and local air 
agencies are encouraged to begin to consider how they might develop 
implementation plans that encourage early emission reductions as well 
as emission reductions that facilitate or amplify reductions affecting 
overburdened communities. The public is encouraged to share information 
on this important topic and although this rulemaking is not requesting 
comment specifically on this topic, information on this topic may be 
submitted for informational purposes to the docket for this proposed 
rulemaking. The EPA may consider whether additional guidance on the 
topic of environmental justice and PM2.5 implementation is 
appropriate, beyond what is already included in the existing 
PM2.5 SIP Requirements Rule. The EPA encourages air agencies 
and other stakeholders to review the existing PM2.5 SIP 
Requirements Rule and the information provided therein regarding 
environmental justice considerations in PM2.5 air planning. 
To be clear, nothing in the above text should be interpreted as seeking 
comment in this proposal on any aspect of the 2016 PM2.5 SIP 
Requirements Rule.
---------------------------------------------------------------------------

    \179\ This and all subsequent references to ``state'' are meant 
to include State, local, and Tribal agencies responsible for the 
implementation of a PM2.5 control program.
---------------------------------------------------------------------------

    With respect to the topics covered in this section, the EPA 
welcomes the public to provide input to the Agency through comments. 
However, because these issues are not relevant to the establishment of 
a revised primary annual PM2.5 NAAQS, and because no 
specific revisions are proposed for the regulations implementing the 
PM2.5 NAAQS (i.e., 40 CFR part 51, subpart Z), the EPA does 
not expect to respond to these comments in the final action on this 
proposal (nor is it required to do so).

A. Designation of Areas

    After the EPA establishes or revises a NAAQS, the CAA requires the 
EPA and the states to take steps to ensure that the new or revised 
NAAQS is met. The first step, known as the initial area designations, 
involves identifying areas of the country that either meet or do not 
meet the new or revised NAAQS, along with the nearby areas contributing 
to the violations.
    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 109, the Governor of each state shall

[[Page 5681]]

. . . submit to the Administrator a list of all areas (or portions 
thereof) in the State'' and that making recommendations for whether the 
EPA should designate those areas as nonattainment, attainment, or 
unclassifiable.\180\ The CAA provides the EPA discretion to require 
states to submit their designations recommendations within a reasonable 
amount of time not exceeding 1 year. The CAA also stipulates 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.\181\ One way 
the EPA intends to account for environmental justice in the 
implementation process is to promptly issue designations in accordance 
with the statutory requirements to ensure expeditious public health 
protections for all populations, including those currently experiencing 
disparities in PM2.5 exposures and PM2.5-related 
health risk.
---------------------------------------------------------------------------

    \180\ 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.
    \181\ API v. Costle, 609 F.2d 20 (D.C. Cir. 1979)
---------------------------------------------------------------------------

    If the EPA agrees with the designation 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 designations of 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 tentative designation decision. If 
a state elects not to provide designation recommendations, then the EPA 
must timely 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 (Page, 2011a).
    Monitoring data are currently available from numerous existing 
PM2.5 Federal Equivalent Methods (FEM) and Federal Reference 
Methods (FRM) sites to determine compliance with the proposed revised 
PM2.5 primary annual NAAQS. As discussed in section II 
above, the EPA is proposing to: (1) revise the level of the primary 
annual PM2.5 standard and retain the current primary 24-hour 
PM2.5 standard (section II.D.3); and (2) not change the 
current secondary annual and 24-hour PM2.5 standards at this 
time (section V.D.3). 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 \182\ to support area boundaries decisions for the revised 
standard. 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. For the reason noted earlier, the EPA believes 
it is important to consider environment justice within the framework of 
this area-specific analysis. Consistent with past practice, the EPA 
expects to address issues relevant to area designations more fully in a 
separate designations-specific memorandum around the time of 
promulgation of any revised PM2.5 NAAQS.\183\ Examples of 
issues that may be included in the separate designations-specific 
memorandum may include, but are not limited to, exceptional events 
demonstrations for wildfire and/or prescribed fires on wildland, 
factors to consider in identifying appropriate designations for areas 
and boundaries, among other relevant topics. For informational 
purposes, the public can comment on the process and schedule for the 
initial area designations and nonattainment boundary setting effort 
associated with a new or revised PM2.5 NAAQS. As noted 
above, the EPA does not expect to respond to these comments in the 
final regulatory action establishing the NAAQS.
---------------------------------------------------------------------------

    \182\ 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 any new or revised PM2.5 NAAQS.
    \183\ https://www3.epa.gov/pmdesignations/2012standards/docs/april2013guidance.pdf.
---------------------------------------------------------------------------

    As in past iterations of the PM2.5 NAAQS, the EPA 
intends to make the designations for any revised NAAQS based on the 
most recent 3 years of complete and valid air quality data. 
Accordingly, the EPA recommends that states base their initial 
designation recommendations on the most current available 3 years of 
complete and valid air quality data. The EPA intends to use available 
air quality data from the current PM2.5 mass and speciation 
monitoring networks and other technical information. The EPA will then 
base the final designations on 3 consecutive years of certified air 
quality monitoring data, likely 2021-2023.\184\
---------------------------------------------------------------------------

    \184\ In certain circumstances in which the Administrator has 
insufficient information to promulgate area designations within 2 
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 1 year.
---------------------------------------------------------------------------

    In some areas, State or Tribal air agencies may have flagged air 
quality data for certain days in the Air Quality System due to 
potential impacts from exceptional events (i.e., such as wildfires or 
high wind dust storms). Air quality concentrations on such days may 
affect the calculation of design values for regulatory air monitoring 
sites in determining whether such sites may violate the revised 
PM2.5 NAAQS, and therefore could influence the initial area 
designations for this revised NAAQS. Under the 2016 Exceptional Events 
Rule (see ``Treatment of Data Influenced by Exceptional Events; Final 
Rule,'' 81 FR 68216, October 3, 2016), an air agency may submit to the 
EPA a demonstration with supporting information and analyses for each 
monitor and day the air agency claims should be excluded from design 
value calculations for regulatory purposes. The EPA has provided a 
number of tools to assist air agencies in preparing their 
demonstrations \185\ and will continue to work with air agencies as 
they identify, prepare and submit exceptional events demonstrations. 
The EPA recognizes that some areas and stakeholders may be

[[Page 5682]]

concerned about wildfire and prescribed fire related impacts to 
designations and/or other forthcoming actions of regulatory 
significance for which a state may want to submit an exceptional events 
demonstration. The EPA has already issued guidance addressing 
development of exceptional events demonstrations for both wildfire and 
prescribed fires on wildland. Existing guidance and other tools are 
available on the EPA's website identified above. The air agency is 
required to follow the exceptional events demonstration submission 
deadlines that are identified in 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.'' Further, the EPA has notified states of areas subject 
to mitigation plan provisions. Within 2 years of the notification, if 
the air agency has not submitted a required mitigation plan, the EPA 
will not concur with the air agency's request to exclude data until the 
required plan is submitted and verified.
---------------------------------------------------------------------------

    \185\ See EPA's Exceptional Events homepage at https://www.epa.gov/air-quality-analysis/treatment-air-quality-data-influenced-exceptional-events-homepage-exceptional.
---------------------------------------------------------------------------

    As noted earlier, the EPA intends to provide designation guidance 
to the states and tribes around the time of the promulgation of a 
revised NAAQS, to assist in formulating these recommendations. With 
regard to the area designations process, if, after evaluating the state 
recommendations in light of the technical factors, the Administrator 
intends to modify any state area recommendation, the EPA will notify 
the appropriate state Governor no later than 120 days prior to making 
final designations decisions. A state that believes the Administrator's 
intended modification is inappropriate will have the opportunity to 
demonstrate to the EPA why it believes its original recommendation (or 
a revised recommendation) is more appropriate before final designations 
are promulgated. The Administrator will take any additional input from 
the state into account in making final designation decisions. If the 
Administrator departs from the stated intentions in the initial 120-day 
notification letter in a way that does not match the most recently 
received recommendation from the Governor (or tribe) as of the date of 
the final designation, the Administrator will provide an additional 
120-day notification letter notifying the Governor of such 
modifications. The EPA invites preliminary comment on all aspects of 
the designation process at this time, which the Agency will consider in 
developing any updated guidance.

B. Section 110(a)(1) and (2) Infrastructure SIP Requirements

    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 
including, among other things, enforceable emissions limitations, an 
ambient monitoring program, an enforcement program, air quality 
modeling capabilities, and adequate personnel, resources, and legal 
authority. 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 sections 110(a)(1), all 
states are required to make these infrastructure SIP submissions within 
3 years after promulgation 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 does not currently intend to do so.
    Under CAA section 110(a)(1) and (2), states are required to make 
SIP submissions that address a number of requirements pertaining to 
implementation, maintenance, and enforcement of a new or revised NAAQS. 
The specific subsections in CAA section 110(a)(2) require states to 
address a number of requirements, as applicable: (A) Emissions limits 
and other control measures, (B) Ambient air quality monitoring/data 
system, (C) Programs for enforcement of control measures and for 
construction or modification of stationary sources, (D)(i) Interstate 
pollution transport; and (D)(ii) Interstate and international pollution 
abatement, (E) Adequate resources and authority, conflict of interest, 
and oversight of local governments and regional agencies, (F) 
Stationary source monitoring and reporting, (G) Emergency episodes, (H) 
SIP revisions, (I) Plan revisions for nonattainment areas, (J) 
Consultation with government officials, public notification, PSD and 
visibility protection, (K) Air quality modeling and submission of 
modeling data, (L) Permitting fees, and (M) Consultation and 
participation by affected local entities. These requirements apply to 
all SIP submissions in general, but the EPA has provided specific 
guidance to states concerning its interpretation of these requirements 
in the specific context of infrastructure SIP submissions for a new or 
revised NAAQS (Page, 2013).
    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 plan 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. 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 proposed revised PM2.5 NAAQS. States will be required to 
submit infrastructure SIP submissions for a revised PM2.5 
NAAQS before they are required to submit nonattainment plan SIP 
submissions to demonstrate attainment with the same NAAQS. States are 
required to submit nonattainment plans to provide for attainment and 
maintenance of a revised 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's 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.\186\ This element is often referred to as the ``good 
neighbor'' or ``interstate transport''

[[Page 5683]]

provision.\187\ 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).\188\ 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.\189\ 
Thus, the EPA anticipates that states will need to address interstate 
transport obligations associated with any revised PM NAAQS, if 
finalized, 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, if finalized, as expeditiously as practicable but no later 
than the end of the sixth calendar year following nonattainment area 
designations. See CAA section 188(c).
---------------------------------------------------------------------------

    \186\ CAA section 110(a)(2)(D)(i)(I).
    \187\ 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.''
    \188\ See North Carolina v. EPA, 531 F.3d 896, 909-11 (D.C. Cir. 
2008).
    \189\ 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).
---------------------------------------------------------------------------

    The EPA anticipates developing further information and coordinating 
with states with respect to the requirements of CAA section 
110(a)(2)(D)(i)(I) for implementation of any revised PM NAAQS. We note 
that states may elect to make SIP submissions that address certain 
infrastructure SIP elements separately from the others. In recent 
years, due in part to the complexity of addressing interstate transport 
obligations, some states have found it efficient to make SIP 
submissions to address the interstate transport provisions separately 
from other infrastructure SIP elements.
    It is the responsibility of each state to review its air quality 
management program's existing SIP provisions in light of each new or 
revised NAAQS to determine if any revisions are necessary to implement 
a new or revised NAAQS. Most states have revised and updated their SIPs 
in recent years to address requirements associated with other revised 
NAAQS. For some states, it may be the case that for a number of 
infrastructure elements, the state may believe it already has adequate 
state regulations already adopted and approved into the SIP to address 
a particular requirement with respect to any revised PM2.5 
NAAQS. For such portions of the state's infrastructure SIP submission, 
the state may provide an explanation of how its existing SIP provisions 
are adequate.
    If a state determines that existing SIP-approved provisions are 
adequate in light of the revised PM2.5 NAAQS with respect to 
a given infrastructure SIP element (or sub-element), then the state may 
make a SIP submission ``certifying'' that the existing SIP contains 
provisions that address those requirements of the specific section 
110(a)(2) infrastructure elements.\190\ In the case of such a 
certification submission, the state does not have to include a copy of 
the relevant provision (e.g., rule or statute) itself. Rather, the 
state in its infrastructure SIP submission may provide citations to the 
SIP-approved state statutes, regulations, or non-regulatory measures, 
as appropriate, which meet the relevant CAA requirement. Like any other 
SIP submission, that state can make such a certification 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 submittal 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.
---------------------------------------------------------------------------

    \190\ A ``certification'' approach would not be appropriate for 
the interstate pollution control requirements of CAA section 
110(a)(2)(D)(i).
---------------------------------------------------------------------------

    In consultation with its EPA 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. The EPA recommends 
that states electronically submit their infrastructure SIPs to the EPA 
through the State Plan Electronic Collaboration System (SPeCS),\191\ an 
online system available through the EPA's Central Data Exchange.
---------------------------------------------------------------------------

    \191\ https://cdx.epa.gov/.
---------------------------------------------------------------------------

C. Implementing Any Revised PM2.5 NAAQS in Nonattainment Areas

    Part D of the CAA describes the various program requirements that 
apply to nonattainment areas for different NAAQS. Section 172 (found in 
subpart 1 of part D) includes general SIP requirements, and sections 
188-190 (found in subpart 4 of part D) include SIP requirements that 
specifically govern implementation for the PM10 and 
PM2.5 NAAQS. All PM2.5 nonattainment areas are 
initially classified as Moderate per CAA section 188(a). Under section 
189(a)(2), states are required to submit attainment plan SIP 
submissions to the EPA within 18 months of the effective date of area 
designations. These plans need to show how the nonattainment area will 
attain the primary PM2.5 standards ``as expeditiously as 
practicable,'' but presumptively by no later than the end of the 6th 
calendar year after the effective date of designations. For example, if 
the EPA finalizes nonattainment designations for a revised 
PM2.5 NAAQS in 2024, then the outermost statutory Moderate 
area attainment date would be December 31, 2030. If the state fails to 
attain the standard by the end of the 6th calendar year after the 
effective date of designations, the EPA is required to reclassify the 
area to Serious, and the state then must attain the standard by the end 
of the 10th calendar year after the effective date of designations 
(e.g., December 31, 2034).
    On August 24, 2016, the EPA issued a detailed 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 EPA also 
intended this implementation rule to apply to nonattainment areas 
designated pursuant to any future revisions of the PM2.5 
NAAQS. The rule covers a number of SIP requirements for nonattainment 
areas, including a nonattainment area emissions inventory, policies 
regarding PM2.5 precursor pollutants (i.e., SO2, 
NOX,

[[Page 5684]]

VOC, and ammonia), control strategies (such as reasonably available 
control measures and reasonably available control technology), air 
quality modeling, attainment demonstrations, reasonable further 
progress requirements, quantitative milestones, and contingency 
measures. Guidance provided in the PM2.5 SIP Requirements 
Rule is supplemented by other EPA guidance 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),\192\ 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).
---------------------------------------------------------------------------

    \192\ 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.
---------------------------------------------------------------------------

    Under the basic approach outlined in the PM2.5 SIP 
Requirements Rule, a state would first develop an updated emissions 
inventory of sources and emissions activities in the nonattainment 
area. It would then use air quality modeling or other tools to estimate 
the air quality improvement that can be expected in the nonattainment 
area by the attainment year due to enforceable and existing ``on the 
books'' Federal, state, and local emissions reduction measures. The 
state also would work with the regulated community and other 
stakeholders to evaluate potential control measures for emissions 
sources and activities in the nonattainment area, and identify the 
additional reasonably available control measures (RACM) and reasonably 
available control technology (RACT) that can be implemented by these 
sources in order to attain the standard as expeditiously as 
practicable, but no later than by the end of the 6th calendar year 
after the effective date of designations.
    The evaluation of air quality improvement associated with potential 
future emissions reductions is commonly performed with sophisticated 
air quality modeling tools. Given that fine particle concentrations are 
affected both by regionally-transported pollutants (e.g., 
SO2 and NOX emissions from power plants) and 
emissions of direct PM2.5 and other pollutants from local 
sources in the nonattainment area (e.g., steel mills, rail yards, 
highway mobile sources), the EPA recommends the use of regional 
photochemical models (such as CMAQ and CAMx), in combination with 
source-oriented dispersion models (such as the American Meteorological 
Society/Environmental Protection Agency Regulatory Model (AERMOD)), as 
needed, to develop PM2.5 attainment strategies for any 
revised PM2.5 NAAQS. The EPA SIP modeling guidance provides 
details on the development of attainment demonstrations, and the EPA 
will continue to assist air agencies in modeling and technical analyses 
(80 FR 8787, February 19, 2015; U.S. EPA, 2017).
    The PM2.5 SIP Requirements Rule provides recommendations 
to states regarding when and how to consider environmental justice in 
the context of PM2.5 attainment 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 (80 FR 58010, 
58136, August 25, 2016). The EPA expects states to consider these and 
other factors as part of their SIP development process.
    The PM2.5 SIP Requirements Rule outlines some examples 
of how states can implement these recommendations.\193\ 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. The EPA recognizes that states have flexibility under the CAA to 
concentrate state resources on controlling sources of PM2.5 
emissions that directly and adversely affect certain populations 
currently experiencing disparities in PM2.5 exposures and 
PM2.5-related health risk, thereby maximizing health 
benefits for those populations. Moreover, states can establish 
opportunities to bolster meaningful involvement in a number of ways, 
such as communicating with communities with disparities in exposures 
and risks in appropriate languages and developing enhanced notice-and-
comment opportunities for those communities.
---------------------------------------------------------------------------

    \193\ For more information on the EPA's recommendations and 
examples, see 81 FR 58010, 58137, August 24, 2016.
---------------------------------------------------------------------------

    As previously mentioned, the 2016 PM2.5 SIP Requirements 
Rule is structured in such a way that it 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. Thus, the EPA is not proposing changes to the 
current PM2.5 SIP Requirements Rule in this proposed 
rulemaking, and therefore is not requesting comment on that rule.

D. Implementing the Primary and Secondary PM10 NAAQS

    As summarized in sections III.C.3 and III.D.3 above, the EPA is 
proposing to retain the current primary and secondary 24-hour 
PM10 standards 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 intends to retain 
the existing implementation strategy for meeting the CAA requirements 
for the PM10 NAAQS. States and emissions sources should 
continue to follow the existing guidance and regulations for 
implementing the current standards.

E. Prevention of Significant Deterioration and Nonattainment New Source 
Review Programs for the Proposed 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 major source construction. 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 proposed 
revised PM2.5

[[Page 5685]]

NAAQS, the NSR provisions applicable under an area's 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 new designations for the proposed revised PM2.5 
NAAQS, if finalized, become effective, those designations will further 
inform whether PSD or NNSR applies to PM2.5 in a particular 
area. 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.\194\
---------------------------------------------------------------------------

    \194\ 40 CFR 51.166(i)(2) and 52.21(i)(2)
---------------------------------------------------------------------------

    The EPA has assessed the proposed revision of the level of the 
primary annual PM2.5 NAAQS and is not proposing any changes 
to the NSR program regulations as part of this proposal to revise the 
PM2.5 NAAQS. Sources and reviewing authorities will be able 
to use existing NSR regulatory provisions. Under the PSD program, the 
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 41) to address primary and secondary 
PM2.5 impacts in making this demonstration and 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).\195\ The EPA will consider whether changes 
or updates to PSD program guidance or associated tools are warranted as 
a result of the proposed revision to the primary annual 
PM2.5 NAAQS, should it be finalized, and would communicate 
such changes through separate action(s) following promulgation of a 
revised standard.
---------------------------------------------------------------------------

    \195\ 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 for 
PM2.5 in areas that have been 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 in 
the EPA regulations 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 all attainment and unclassifiable areas. See 40 CFR 
51.166(b)(49)(i), 40 CFR 52.21(b)(50)(i). 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), as applicable.\196\ 
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) 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.\197\ 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, 40 CFR 
51.166(p); 40 CFR 52.21(p)).\198\ The EPA has developed the Guideline 
on Air Quality Models and other documents to, among other things, 
provide methods and guidance for demonstrating compliance with the 
PM2.5 NAAQS and PSD increments for PM2.5.\199\
---------------------------------------------------------------------------

    \196\ 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.
    \197\ 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.
    \198\ 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.
    \199\ See 40 CFR part 51, appendix W; 82 FR 5182, January 17, 
2017; See also U.S. EPA, 2021c. The EPA provided an initial version 
of the same 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.
---------------------------------------------------------------------------

    The EPA has historically interpreted the requirement for an air 
quality impact analysis under CAA section 165(a)(3) and the 
implementing regulations to include a requirement to demonstrate that 
emissions from the proposed facility will not cause or contribute to a 
violation of any NAAQS

[[Page 5686]]

that is in effect as of the date a PSD permit is issued, except to the 
extent that a pending permit application was subject to grandfathering 
provisions that the EPA had established through rulemaking. The EPA is 
not proposing such provisions for this action. 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 grandfathering provisions that exempted certain 
pending PSD permit actions (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. In August 2019, the U.S. Court of 
Appeals for the D.C. Circuit vacated the grandfathering provision in 
the PSD rules applicable to 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).\200\ Based on that court decision, the EPA is not 
proposing any grandfathering provision for this proposed 
PM2.5 NAAQS revision, if finalized. Accordingly, PSD permits 
issued on or after the effective date of any final revised 
PM2.5 NAAQS would require a demonstration that the proposed 
emissions increases would not cause or contribute to a violation of the 
revised PM2.5 NAAQS.
---------------------------------------------------------------------------

    \200\ 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 grandfathering for a 
new or revised NAAQS.
---------------------------------------------------------------------------

    The EPA anticipates that, if this rule is finalized as proposed, 
the existing PM2.5 air quality in some areas will not be in 
attainment of the new revised primary annual PM2.5 NAAQS, 
and that these areas will be designated 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 
PM2.5 will continue to be subject to the PSD program 
requirements for PM2.5.\201\ This raises the question as to 
how a source can be issued a PSD permit in light of known existing 
ambient violations of the revised 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).\202\ 
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, a PSD permit may not be issued unless the 
permit applicant compensates for the adverse impact that would 
otherwise cause or contribute to a violation of the NAAQS. While the 
PSD regulations do not explicitly specify remedial actions that a 
prospective source can take to address such a situation, the EPA has 
historically recognized in regulations, and through other actions, that 
sources applying for PSD permits may utilize offsets as part of the 
required PSD demonstration under CAA section 165(a)(3)(B).\203\
---------------------------------------------------------------------------

    \201\ 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.
    \202\ 40 CFR 51.166(k) requires that SIPs shall provide that the 
owner or operator of the proposed source or modification shall 
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.
    \203\ See, e.g., Page, 2010; 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, Jan. 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 a NAAQS has been established (i.e., a criteria 
pollutant). The relevant part D requirements are typically referred to 
as the NNSR program. The EPA's regulations for the NNSR programs are 
contained in 40 CFR 51.165 and 52.24 and part 51, appendix S. 
Specifically, the EPA has developed minimum program requirements for an 
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 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 
PM2.5, in addition to the general threshold level of 100 
tpy, 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 reviewed 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.\204\ 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). 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

[[Page 5687]]

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).
---------------------------------------------------------------------------

    \204\ All recognized precursors to PM2.5 are 
regulated as precursors for NNSR. 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 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, as well as performing other analyses 
as required under section 173 of the CAA.
    Following the promulgation of any revised NAAQS for 
PM2.5, some new nonattainment areas for PM2.5 may 
result. Where a state does not have an 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 undergo preconstruction review pursuant to the NNSR 
program. States are required to submit nonattainment plans to provide 
for attainment and maintenance of a revised PM2.5 NAAQS 
within 18 months from the effective date of nonattainment area 
designations as required under CAA section 189(a)(2)(B). Therefore, 
states whose existing NNSR program requirements, if any, cannot be 
interpreted to apply to the revised primary annual PM2.5 
NAAQS at that time will be allowed to issue the necessary permits in 
accordance with the applicable nonattainment permitting requirements 
contained in 40 CFR part 51, appendix S, which would apply to the 
revised PM2.5 NAAQS upon its effective date. See 73 FR 
28321, 28340, May 16, 2008.
    Finally, the EPA recommends that, where appropriate, PSD and NNSR 
permitting authorities assess impacts to communities with environmental 
justice concerns. For example, this may include conducting a 
demographic analysis to inform development of a plan for community 
outreach and engagement, conducting a cumulative emissions impact 
analysis,\205\ or considering the environmental and social costs 
imposed on the impacted community when conducting an alternative sites 
analysis.\206\ Another option could be improving the understanding of 
the potential impact of minor sources by generating an emissions 
inventory for such minor sources, including sources that are not 
currently required to report emissions, to generate options on how 
emissions can be reduced in the target area. See 81 FR 58010, 58137. 
The EPA anticipates developing further information and consulting with 
permitting authorities on how to best address environmental justice in 
the permitting process.
---------------------------------------------------------------------------

    \205\ The permitting authority may conduct a cumulative analysis 
of the projected PM2.5 emissions from all emission units 
at the proposed facility and PM2.5 emissions from nearby 
facilities, to provide a more complete assessment of the ambient air 
impacts of the proposed facility on affected communities. See 40 CFR 
part 51, appendix W, section 9.2.3.
    \206\ Section 173(a)(5) of the CAA requires for an NNSR permit 
``an analysis of alternative sites, sizes, production processes, and 
environmental control techniques for such proposed source [that] 
demonstrates that benefits of the proposed source significantly 
outweigh the environmental and social costs imposed as a result of 
its location, construction, or modification.'' This requirement is 
referred to as the ``alternative sites analysis.''
---------------------------------------------------------------------------

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 
any new or revised NAAQS for PM2.5 does not apply until one 
year after the effective date of the nonattainment designation for that 
NAAQS. See CAA section 176(c)(6) and 40 CFR 93.102(d)). The EPA's 
Transportation Conformity Rule \207\ establishes the criteria and 
procedures for determining whether transportation activities conform to 
the SIP. The EPA is not proposing changes to the transportation 
conformity rule in this proposed rulemaking. The EPA notes that the 
transportation conformity rule already addresses the PM2.5 
and PM10 NAAQS. However, in the future, the EPA will review 
the need to issue or revise guidance describing how the current 
conformity rule applies in nonattainment and maintenance areas for any 
new or revised primary or secondary PM NAAQS, as needed.
---------------------------------------------------------------------------

    \207\ 40 CFR part 93, subpart A
---------------------------------------------------------------------------

G. General Conformity Program

    The general conformity program implements CAA section 176(c) and 
requires that Federal agencies do not adopt, accept, approve, or fund 
activities that are not consistent with state air quality goals. 
General conformity applies to any Federal action (e.g., funding, 
licensing, permitting, or approving) if (1) the action takes place in a 
nonattainment or maintenance area for any of the criteria pollutants 
and (2) it is not a Federal Highway Administration (FHWA) or Federal 
Transit Administration (FTA) project as defined in 40 CFR 93.101 (these 
projects are covered under the transportation conformity program 
described above).
    The EPA's General Conformity Rule \208\ establishes the criteria 
and procedures for determining if a Federal action conforms to the 
applicable attainment plan. General conformity for any revised 
PM2.5 NAAQS does not apply until one year after the 
effective date of the nonattainment designation for that NAAQS. The EPA 
is not proposing changes to the General Conformity Rule in this 
proposed rulemaking. The EPA notes that the General Conformity Rule 
already addresses the PM2.5 and PM10 NAAQS.
---------------------------------------------------------------------------

    \208\ 40 CFR 93.150 through 93.165
---------------------------------------------------------------------------

IX. Statutory and Executive Order Reviews

    Additional information about these statutes and Executive orders 
can be found at https://www.epa.gov/laws-regulations/laws-and-executive-orders.

A. Executive Order 12866: Regulatory Planning and Review and Executive 
Order 13563: Improving Regulation and Regulatory Review

    This action is an economically significant regulatory action that 
was submitted to the Office of Management and Budget (OMB) for review. 
Any changes made in response to OMB recommendations have been 
documented in the docket. The EPA prepared an illustrative analysis of 
the potential costs and benefits associated with this action. This 
analysis is contained in the document ``Regulatory Impact Analysis for 
the Proposed Reconsideration of the National Ambient Air Quality 
Standards for Particulate Matter,'' which is available in the 
Regulatory Impact Analysis (RIA) docket (EPA-HQ-OAR-2019-0587) and 
briefly summarized below. The RIA estimates the costs and monetized 
human health benefits in 2032, after

[[Page 5688]]

implementing existing and expected regulations and assessing emissions 
reductions to meet the current annual and 24-hour particulate matter 
NAAQS (12/35 [mu]g/m\3\), associated with applying national control 
strategies for the proposed annual and 24-hour alternative standard 
levels of 10/35 [mu]g/m\3\ and 9/35 [mu]g/m\3\, as well as the 
following two more stringent alternative standard levels: (1) an 
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 (2) an 
alternative 24-hour standard level of 30 [mu]g/m\3\ in combination with 
the proposed annual standard level of 10 [mu]g/m\3\ (i.e., 10/30 [mu]g/
m\3\). Table 2 provides a summary of the estimated monetized benefits, 
costs, and net benefits associated with applying national control 
strategies toward reaching alternative standard levels. 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 proposed rule.

  Table 2--Estimated Monetized Benefits, Costs, and Net Benefits of the Illustrative Control Strategies Applied Toward the Primary 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...................  $9,600 and $20,000...................  $21,000 and $43,000..................  $46,000 and $95,000
Costs \b\...........................  $95..................................  $260.................................  $390.................................  $1,800
                                     -----------------------------------------------------------------------------------------------------------------------------------------------------------
    Net Benefits....................  $8,400 and $17,000...................  $9,300 and $19,000...................  $20,000 and $43,000..................  $44,000 and $93,000
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Notes: Rows may not appear to add correctly due to rounding. We focus results to 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 proposed 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 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 an information collection burden under 
the PRA. There are no information collection requirements directly 
associated with a proposed decision to revise or retain a NAAQS under 
section 109 of the CAA.

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 
proposed 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).

D. Unfunded Mandates Reform Act (UMRA)

    This action does not contain any unfunded mandate 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).
    The EPA acknowledges, however, that if corresponding revisions to 
associated SIP requirements and air quality surveillance requirements 
are proposed at a later time, those revisions might result in such 
effects. Any such effects would be addressed as appropriate if and when 
such revisions are proposed.

E. Executive Order 13132: Federalism

    This action does not have federalism implications. It 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 will offer 
government-to-government consultation with tribes as requested.

[[Page 5689]]

G. Executive Order 13045: Protection of Children From Environmental 
Health Risks and Safety Risks

    This action is subject to Executive Order 13045 because it is an 
economically significant regulatory action as defined by 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. The Policy on Children's Health also applies to this 
action. 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 proposed 
standards to be requisite to protect public health, including the 
health of children, with an adequate margin of safety, are contained in 
sections II.B and II.D of this preamble. Copies of all documents have 
been placed in the public docket for this action.

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 
propose to revise the primary annual PM2.5 NAAQS and to 
retain the primary 24-hour PM2.5 NAAQS, primary 
PM10 NAAQS, and secondary PM 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 proposal does not constitute a significant energy action as 
defined in Executive Order 13211.

I. National Technology Transfer and Advancement Act (NTTAA)

    This action involves technical standards. The EPA proposes to use 
the current 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. 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

    The EPA believes that this action does not have disproportionately 
high and adverse human health or environmental effects on minority 
populations, low-income populations and/or indigenous peoples, as 
specified in Executive Order 12898 (59 FR 7629, February 16, 1994). The 
documentation for this assessment is contained in sections II.B.2, 
II.C.1, II.C.3, II.D.2, and II.D. of this preamble and also in the 2019 
Integrated Science Assessment, Supplement to the 2019 Integrated 
Science Assessment, and Policy Assessment. The EPA has carefully 
evaluated the potential impacts on minority populations and low SES 
populations as discussed in sections II.B.2, II.C.1, II.C.3, II.D.2, 
and II.D.3 of this preamble. The Integrated Science Assessment, 
Supplement to the Integrated Science Assessment, and Policy Assessment 
contain the evaluation of the scientific evidence, quantitative risk 
analyses and policy considerations that pertain to these populations. 
These documents are available as described in this SUPPLEMENTARY 
INFORMATION section and copies of all documents have been placed in the 
public docket for this action.

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.
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. http://

[[Page 5690]]

www.azdeq.gov/environ/air/download/vis_021903f.pdf.
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 human exposures. 
Occupational and Environmental Medicine 70(11): 761-767.
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.
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.
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, E. (2019). Email and attachment from Elizabeth Chan, EPA, 
Question re: Request for additional PM2.5 information 
from Di et al 2017. Available in PM NAAQS Review Docket (EPA-HQ-OAR-
2015-0072). December 3, 2019. Office of Air Quality Planning and 
Standards Research Triangle Park, NC.
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.
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 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://www.epa.gov/sites/default/files/2019-09/documents/draft_policy_assessment_for_pm_naaqs_09-05-2019.pdf.
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 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://yosemite.epa.gov/sab/sabproduct.nsf/LookupWebReportsLastMonthCASAC/932D1DF8C2A9043F852581000048170D?OpenDocument&TableRow=2.3#2.
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.
deSouza, P, Braun, D, Parks, RM, Schwartz, J, Dominici, F and 
Kioumourtzoglou, MA (2021). Nationwide Study of Short-term Exposure 
to Fine Particulate Matter and Cardiovascular Hospitalizations Among 
Medicaid Enrollees. Epidemiology 32(1): 6-13.
DHEW (1969). Air Quality Criteria for Particulate Matter. National 
Air Pollution Control Administration. Washington, D.C. 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.
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.
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.

[[Page 5691]]

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.
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, P, PM and Loft, S (2015b). 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, Choirat, C and Zigler, CM (2019a). Accountability 
assessment of health improvements in the United States associated 
with reduced coal emissions between 2005 and 2012. Epidemiology 
(Cambridge, Mass.) 30(4): 477.
Henneman, LRF, Choirat, C and Zigler, ACM (2019b). 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.
Iglesias, V, Balch, JK and Travis, WR (2022). U.S. fires became 
larger, more frequent, and more widespread in the 2000s. Sci Adv 
8(11): eabc0020.
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.
Johnson, CW, A.; Long, R.; Vanderpool, R. (2022). Investigation of 
Gaseous Criteria Pollutant Transport Efficiency as a Function of 
Tubing Material. U.S. Environmental Protection Agency, Office of 
Research and Development, Center for Environmental Measurements and 
Modeling. Research Triangle Park, NC. U.S. EPA. EPA/600/R-22/166. 
August 2022. Available at: https://cfpub.epa.gov/si/si_public_record_Report.cfm?dirEntryId=355813&Lab=CEMM.
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.
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, 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://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.
Page, D. (2010). 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

[[Page 5692]]

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.
Page, D. (2011). Memorandum from Stephen D. Page, Director, Office 
of Air Quality Planning and Standards to Regional Administrators, 
Regions I-X. Guidance to Regions for Working with Tribes during the 
National Ambient Air Quality Standards (NAAQS) Designations Process. 
December 20, 2011. Office of Air Quality Planning and Standards U.S. 
EPA, Research Triangle Park. 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
.
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, I, 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.
Russell, A and Samet, J. (2010). Letter from Armistead Russel, Chair 
CASAC Ambient Air Monitoring & Methods Committee and Jonathan Samet, 
Chair, Clean Air Scientific Advisory Committee and to Administrator 
Lisa Jackson. Re: Review of the ``Near-road Guidance Document--
Outline'' and ``Near-road Monitoring Pilot Study Objectives and 
Approach'' (November 2010). November 24, 2010. EPA-CASAC-11-001. 
Office of the Administrator, Science Advisory Board U.S. EPA HQ, 
Washington DC. Available at: https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=9101XP24.PDF.
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 
31(2): 160-167.
Sanders, NJ, Barreca, AI and Neidell, MJ (2020b). Estimating causal 
effects of particulate matter regulation on mortality. Epidemiology 
(Cambridge, Mass.) 31(2): 160.
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, 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/f?p=105:18:10792850355838:::RP,18:P18_ID:2607#report.
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.
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. http://yosemite.epa.gov/sab/sabproduct.nsf/
B55911DF9796E5E385257592006FB737/$File/
CRA+VAQ+Pref+Robustness+Study+3+30+09+final.pdf.
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,

[[Page 5693]]

Case, MW, Berntsen, J, Bromberg, PA, Cascio, WE, Diaz-Sanchez, D, 
Devlin, RB and Samet, JM (2015). Dietary 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 
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 (2004b). 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 (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 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 (2009b). 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 (2009c). 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 (2010a). 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 (2010b). 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 (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.
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.
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.
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.
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.
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.
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.
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.
U.S. EPA (2020a). 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.
U.S. EPA (2020b). Recommendations for Nationwide Approval of 
NafionTM Dryers Upstream of UV-Absorption Ozone 
Analyzers. Office of Air Quality Planning and Standards, Health and 
Environmental Impacts Division. Research Triangle Park, NC. U.S. 
EPA. EPA/600/R-20/390. November 2020.
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.
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). 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). 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.
U.S. EPA (2022b). 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.
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

[[Page 5694]]

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.
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, Wang, Y, Di, Q, Choirat, C, Wang, Y, Koutrakis, P, 
Zanobetti, A, Dominici, F and Schwartz, JD (2019). Short term 
exposure to fine particulate matter and hospital admission risks and 
costs in the Medicare population: time stratified, case crossover 
study. BMJ (Clinical Research Edition) 367: l6258.
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.
WHO (2021). World Health Organization global air quality guidelines: 
particulate matter (PM2.5 and PM10), ozone, nitrogen dioxide, sulfur 
dioxide and carbon monoxide. World Health Organization. Geneva.
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.
Zhang, H and Kondragunta, S (2021). Daily and hourly surface 
PM2.5 estimation from satellite AOD. Earth and Space 
Science 8(3): e2020EA001599.
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 proposed to be 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 to 10.0 micrograms per cubic meter ([mu]g/
m\3\) annual arithmetic mean concentration and 35 [mu]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 to 10.0 [mu]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 [mu]g/m\3\.
0
3. Amend appendix K to part 50 as follows:
0
a. In section 1.0 by revising paragraph (b);
0
b. In section 2.3 by adding paragraph (d); and
0
c. In section 3.0 by 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 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 [mu]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

[[Page 5695]]

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 seven 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 [mu]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 PM2.5 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 as follows:
0
a. In section 1.0 by revising paragraph (a); and
0
b. In section 3.0 by adding paragraph (d)(3); and
0
c. In section 4.1 by revising paragraph (a); and
0
d. In section 4.2 by revising paragraph (a).
    The addition and revisions read as follows.

Appendix N to Part 50--Interpretation of the National Ambient Air 
Quality Standards for PM2.5

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 PM2.5

* * * * *
    (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 PM2.5 NAAQS

    (a) Levels of the primary and secondary annual PM2.5 
National Ambient Air Quality Standards are specified in Sec. Sec.  
50.7, 50.13, 50.18, and 50.20 as applicable.
* * * * *

4.2 Twenty-Four-Hour PM2.5 NAAQS

    (a) Levels of the primary and secondary 24-hour PM2.5 
National Ambient Air Quality Standards are specified in Sec. Sec.  
50.7, 50.13, 50.18, and 50.20 as applicable.
* * * * *

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. sec. 
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 as follows:
0
a. By revising paragraph (a);
0
b. By adding paragraph (b)(7); and
0
c. By 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: 
Director, Center for Environmental Measurement and Modeling, Reference 
and Equivalent Methods Designation Program (MD-205-03), U.S. 
Environmental Protection Agency, 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-

[[Page 5696]]

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 a notice 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 Federal Register notice 
has been published.
* * * * *
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 appendix of            Applicable subparts of this part
        Pollutant         Reference or equivalent    Manual or automated        part 50 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]  .......  .......  .......
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]  .......  [check]  .......  [check]  [check]
                                                                                                                         \2\                     \1\ \2\
                          Equivalent Class III...  Automated..............  L \1\.................  [check]  .......  [check]  .......  [check]  [check]
                                                                                                                                                    \1\
PM10-2.5................  Reference..............  Manual.................  L, O \2\..............  [check]  .......  .......  .......  [check]  .......

[[Page 5697]]

 
                          Equivalent Class I.....  Manual.................  L, O \2\..............  [check]  .......  [check]  .......  [check]  .......
                          Equivalent Class II....  Manual.................  L, O \2\..............  [check]  .......  [check]  .......  [check]  [check]
                                                                                                                         \2\                     \1\ \2\
                          Equivalent Class III...  Automated..............  L,\1\ O \1\ \2\.......  [check]  .......  [check]  .......  [check]  [check]
                                                                                                                                                    \1\
--------------------------------------------------------------------------------------------------------------------------------------------------------
\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 5698]]



                                                     Table B-3 to Subpart B of Part 53--Interferent Test Concentration,\1\ Parts per Million
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
                                         Hydro-
    Pollutant      Analyzer type \2\     chloric     Ammonia    Hydrogen   Sulfur    Nitrogen   Nitric   Carbon    Ethylene   Ozone   M-xylene    Water     Carbon   Methane   Ethane    Naphth-
                                          acid                  sulfide    dioxide   dioxide    oxide    dioxide                                  vapor    monoxide                       alene
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
SO2.............  Ultraviolet          ..........  ..........    \5\ 0.1  \4\ 0.14        0.5      0.5  ........  .........     0.5        0.2     20,000  ........  .......  ........  \6\ 0.05
                   fluorescence.
SO2.............  Flame photometric..  ..........  ..........       0.01  \4\ 0.14  .........  .......       750  .........  ......  .........        \3\        50  .......  ........  ........
                                                                                                                                                   20,000
SO2.............  Gas chromatography.  ..........  ..........        0.1  \4\ 0.14  .........  .......       750  .........  ......  .........        \3\        50  .......  ........  ........
                                                                                                                                                   20,000
SO2.............  Spectrophotometric-         0.2         0.1        0.1  \4\ 0.14        0.5  .......       750  .........     0.5  .........  .........  ........  .......  ........  ........
                   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.............  Spectrophotometric-  ..........  ..........  .........  \4\ 0.14        0.5      0.5  ........  .........     0.5        0.2  .........  ........  .......  ........  ........
                   gas phase,
                   including DOAS.
O3..............  Ethylene             ..........  ..........    \3\ 0.1  ........  .........  .......       750  .........     \4\  .........        \3\  ........  .......  ........  ........
                   Chemiluminescence.                                                                                          0.08                20,000
O3..............  NO-                  ..........  ..........    \3\ 0.1  ........        0.5  .......       750  .........     \4\  .........        \3\  ........  .......  ........  ........
                   chemiluminescence.                                                                                          0.08                20,000
O3..............  Electrochemical....  ..........     \3\ 0.1  .........       0.5        0.5  .......  ........  .........     \4\  .........        \3\  ........  .......  ........  ........
                                                                                                                               0.08                20,000
O3..............  Spectrophotometric-  ..........     \3\ 0.1  .........       0.5        0.5  \3\ 0.5  ........  .........     \4\  .........  .........  ........  .......  ........  ........
                   wet chemical                                                                                                0.08
                   (potassium iodide).
O3..............  Spectrophotometric-  ..........  ..........  .........       0.5        0.5  \3\ 0.5  ........  .........     \4\       0.02     20,000  ........  .......  ........  ........
                   gas phase,                                                                                                  0.08
                   including
                   ultraviolet
                   absorption and
                   DOAS.
CO..............  Non-dispersive       ..........  ..........  .........  ........  .........  .......       750  .........  ......  .........     20,000    \4\ 10  .......  ........  ........
                   Infrared.
CO..............  Gas chromatography   ..........  ..........  .........  ........  .........  .......  ........  .........  ......  .........     20,000    \4\ 10  .......       0.5  ........
                   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 replacement- ..........  ..........  .........  ........  .........  .......  ........        0.2  ......  .........  .........    \4\ 10  .......       0.5  ........
                   UV photometric.
NO2.............  Chemiluminescent...  ..........     \3\ 0.1  .........       0.5    \4\ 0.1      0.5  ........  .........  ......  .........     20,000  ........  .......  ........  ........
NO2.............  Spectrophotometric-  ..........  ..........  .........       0.5    \4\ 0.1      0.5       750  .........     0.5  .........  .........  ........  .......  ........  ........
                   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.............  Spectrophotometric-  ..........     \3\ 0.1  .........       0.5    \4\ 0.1      0.5  ........  .........     0.5  .........     20,000        50  .......  ........  ........
                   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 5699]]

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))

BILLING CODE 6560-50-P
[GRAPHIC] [TIFF OMITTED] TP27JA23.006

BILLING CODE 6560-50-C
* * * * *

[[Page 5700]]

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))
[GRAPHIC] [TIFF OMITTED] TP27JA23.007

* * * * *

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), [mu]g/m\3\.
Minimum number of test sites....  2................  1................  2..................  4..................  2.................  4.
Minimum number of candidate       3................  3................  \1\ 3..............  \1\ 3..............  \1\ 3.............  \1\ 3.
 method samplers or analyzers
 per site.
Number of reference method        3................  3................  \1\ 3..............  \1\ 3..............  \1\ 3.............  \1\ 3.
 samplers per site.
Minimum number of acceptable
 sample sets per site for PM10
 methods:
    Rj <20 [mu]g/m\3\...........  3................
    Rj >20 [mu]g/m\3\...........  3................
                                 -----------------------------------------------------------------------------------------------------------------------
        Total...................  10...............

[[Page 5701]]

 
Minimum number of acceptable
 sample sets per site for PM2.5
 and PM10-2.5 candidate
 equivalent methods:
    Rj <15 [mu]g/m\3\ for 24-hr   .................  3................  3..................  3..................  3.................  3.
     or Rj <8 [mu]g/m\3\ for 48-
     hr samples.
    Rj>15 [mu]g/m\3\ for 24-hr    .................  3................  3..................  3..................  3.................  3.
     or Rj >8 [mu]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 0.10.  1 0.10  1 0.10.        minus>0.05.                                                                      minus>0.12.
Intercept of regression           0 5..  0 1..  Between: 13.55-      Between: 15.05-      Between: 62.05-     Between: 70.50-
 relationship, [mu]g/m\3\.                                               (15.05 x slope),     (17.32 x slope),     (70.5 x slope),     (82.93 x slope),
                                                                         but not less than-   but not less than-   but not less than-  but not less than-
                                                                         1.5; and 16.56-      2.0; and 15.05-      3.5; and 78.95-     7.0; and 70.50-
                                                                         (15.05 x slope),     (13.20 x slope),     (70.5 x slope),     (61.16 x slope),
                                                                         but not more than    but not more than    but not more than   but not more than
                                                                         +1.5.                +2.0.                +3.5.               +7.0.
                                                                       ---------------------------------------------------------------------------------
Correlation of reference method   >=0.97...........  >=0.97...........  >=0.93--for CCV <=0.4; >=0.85 + 0.2 x CCV--for 0.4 <= CCV <=0.5; >=0.95--for CCV
 and candidate method                                                                                         >=0.5.
 measurements.
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Some missing daily measurement values may be permitted; see test procedure.
\2\ Calculated as the root mean square over all measurement sets.

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 paragraph (c)(2)(iv) to read as follows:


Sec.  53.43  Test procedures.

    (a) * * *
    (2) * * *
    (xvi) * * *
    [GRAPHIC] [TIFF OMITTED] TP27JA23.008
    
* * * * *
    (c) * * *
    (2) * * *
    (iv) * * *

[[Page 5702]]

[GRAPHIC] [TIFF OMITTED] TP27JA23.009

if Cj is above 80 [mu]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.  [thinsp]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 the heading of paragraph (g) and 
paragraphs (g)(1) introductory text, (g)(1)(i) introductory text, and 
(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. * * *
    (1) Particle aerodynamic diameter. The VOAG and FMAG produce near-
monodisperse droplets through the controlled breakup of a liquid jet. 
When the liquid solution consists of a non-volatile solute dissolved in 
a volatile solvent, the droplets dry to form particles of near-
monodisperse size.
    (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 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:
Equation 5 to Paragraph (g)(2)(iii)
[GRAPHIC] [TIFF OMITTED] TP27JA23.010

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 as follows:
0
a. By removing the definition for ``Approved regional method (ARM)''; 
and
0
b. By revising the definition for ``Traceable.''
    The revision reads as follows:


Sec.  [thinsp]58.1  Definitions.

* * * * *

[[Page 5703]]

    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 as follows:
0
a. By revising paragraphs (a)(1) and (b)(10) and (13);
0
b. By adding paragraph (b)(14); and
0
c. By 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 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 five 
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), [mu]g/m\3\ may include 
values down to 0 [mu]g/m\3\.

[[Page 5704]]

    (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 paragraphs (d)(1) and (3) to read as 
follows:


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 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

[[Page 5705]]

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 as follows:
0
a. By revising section 2.6.1 and adding sections 2.6.1.1 and 2.6.1.2;
0
b. By 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. By revising sections 3.1.3.3, 3.2.4, 4.2.1, and 4.2.5; and
0
d. In section 6 by 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 [mu]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:

Equation 6 to Appendix A to Part 58
[GRAPHIC] [TIFF OMITTED] TP27JA23.011

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:

Equation 7 to Appendix A to Part 58
[GRAPHIC] [TIFF OMITTED] TP27JA23.012

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.

Equation 8 to Appendix A to Part 58
[GRAPHIC] [TIFF OMITTED] TP27JA23.013

* * * * *

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

[[Page 5706]]

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 Appendix A to Part 58--Minimum Data Assessment Requirements for NAAQS Related Criteria Pollutant Monitors
--------------------------------------------------------------------------------------------------------------------------------------------------------
               Method                   Assessment method           Coverage           Minimum frequency     Parameters reported    AQS assessment type
--------------------------------------------------------------------------------------------------------------------------------------------------------
Gaseous Methods (CO, NO, SO, O):
    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    See section 3.1.2 of    Each analyzer.........  Once per year........  Audit concentration    Annual PE.
     for 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 concentration    NPAP.
                                                                                                             \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 as follows:
0
a. By revising section 2.6.1 and adding sections 2.6.1.1 and 2.6.1.2;

[[Page 5707]]

0
b. By removing and reserving section 3.1.2.2;
0
c. By revising sections 3.1.3.3 and 3.2.4;
0
d. By adding sections 3.2.4.1 through 3.2.4.3;
0
e. By revising sections 4.2.1, and 4.2.5; and
0
f. In section 6 by 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, FEM, or ARM 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:

Equation 6 to Appendix B to Part 58
[GRAPHIC] [TIFF OMITTED] TP27JA23.014

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:

Equation 7 to Appendix B to Part 58
[GRAPHIC] [TIFF OMITTED] TP27JA23.015

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.

Equation 8 to Appendix B to Part 58
[GRAPHIC] [TIFF OMITTED] TP27JA23.016

* * * * *

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,

[[Page 5708]]

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 Appendix B to Part 58--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 concentration    One-Point QC.
     CO.                              concentration 0.005-                                                   \1\ and measured
                                      0.08 ppm SO2, NO2,                                                     concentration \2\.
                                      O3, & 0.5 and 5 ppm
                                      CO.
    Quarterly performance            See section 3.1.2 of    Each analyzer.........  Once per quarter \5\.  Audit concentration    Annual PE.
     evaluation for SO2, NO2, O3,     this appendix.                                                         \1\ and measured
     CO.                                                                                                     concentration \2\
                                                                                                             for each level.
    NPAP for SO2, NO2, O3, CO\3\...  Independent Audit.....  Each primary monitor..  Once per year........  Audit concentration    NPAP.
                                                                                                             \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.
    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 ([mu]g
                                      strips/filters.                                                        Pb/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 quarters    Primary sampler        PEP.
     PM2.5 \3\.                                               PQAOs with <5 sites.    \5\.                   concentration and
                                                             (2) 8 valid audits for                          performance
                                                              PQAOs with >5 sites.                           evaluation sampler
                                                             (3) All samplers in 6                           concentration.
                                                              years.
    Performance Evaluation Program   Collocated samplers...  (1) 1 valid audit and   Over all 4 quarters    Primary sampler        PEP.
     Pb \3\.                                                  4 collocated samples    \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 as follows:
0
a. By adding sections 2.2 and 2.2.1 through 2.2.19; and
0
b. By removing and reserving sections 2.4, 2.4.1, and 2.4.1.1 through 
2.4.1.7.
    The additions 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

[[Page 5709]]

instrument manufacturer of record or the EPA Administrator.
    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.
    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 1st 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.
    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.
* * * * *
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. 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, particularly where 
there are anticipated effects from sources in the area (e.g., a 
major port, rail yard, airport, industrial area, or major 
transportation 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.

[[Page 5710]]

    (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 how and why the proposed siting deviates 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 a ``must'' 
are defined as requirements and exceptions must be approved through 
the waiver provisions. However, siting criteria that are phrased 
with a ``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

    The probe must be located greater than or equal to 2.0 and less 
than or equal to 15. meters above ground level for all O3 
and SO2 monitoring, and for neighborhood or larger 
spatial scale Pb, PM10, PM10-2.5, 
PM2.5, NO2, and CO sites. Middle scale CO and 
NO2 monitors must also have sampler inlets greater than 
or equal to 2.0 and less than or equal to 15 meters above ground 
level. Middle scale PM10-2.5 sites are required to have 
sampler inlets greater than or equal to 2.0 and less than or equal 
to 7.0 meters above ground level. Microscale Pb, PM10, 
PM10-2.5, and PM2.5 sites are required to have 
sampler inlets greater than or equal to 2.0 and less than or equal 
to 7.0 meters above ground level. Microscale near-road 
NO2 monitoring sites are required to have sampler inlets 
greater than or equal to 2.0 and less than or equal to 7.0 meters 
above ground level. The probe inlets for microscale carbon monoxide 
monitors that are being used to measure concentrations near roadways 
must be greater than or equal to 2.0 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 
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 location in order to interpret this particular 
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 is likely to 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. The plume from the 
local minor sources 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 the probe inlet should be away 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) Buildings and other obstacles may possibly 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 away 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.
    (b) A probe inlet located near or along a vertical wall is 
undesirable because air moving along the wall may be subject to 
possible 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 network 
design criteria regulations specified in appendix D to 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, a minimum of 2.0 meters of 
horizontal separation from walls, parapets, and structures is 
required for rooftop site placement.
    (c) A sampling station having a probe inlet located closer to an 
obstacle than this criterion allows 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 cases where they 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 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 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, such as a roadway or a stationary source.

2.5 Spacing From Roadways

 Table E-1 to Appendix E to Part 58--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      Minimum      distance \1\
                 per 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.
\2\ Applicable for ozone monitors whose placement has not already been
  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

[[Page 5711]]

O3. Table E-1 to 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 having 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 and 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.

 Table E-2 to Appendix E to Part 58--Minimum Separation Distance Between
  Roadways and Probes 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.

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 from 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 the microscale traffic corridor site, the 
location must be greater than or equal 5.0 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 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 to this appendix. Figure E-1 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, 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.

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 to 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 to 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 having a monitor 
probe located closer to a roadway than allowed by the table E-1 
requirements should be classified as microscale or middle scale 
rather than neighborhood or urban scale.

Figure E-1 to Appendix E to Part 58

[[Page 5712]]

[GRAPHIC] [TIFF OMITTED] TP27JA23.017

2.6 Probe Material and Pollutant Sampler Residence Time

    (a) For the reactive gases (SO2, NO2, and 
O3), special probe material must be used for monitors. 
Studies 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 materials in the preceding sentence, 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 has specified borosilicate glass or 
FEP Teflon[supreg] 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 equivalent 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\ is composed primarily of PTFE and can be considered 
equivalent to PTFE. It has been shown in tests to exhibit virtually 
no loss of ozone at 20 second residence times.
    (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 its equivalent 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 also 
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. Other studies indicate that a 10 
second or less residence time is easily achievable. Therefore, 
sampling probes for reactive gas monitors (i.e., SO2, 
NO2, and O3) must have a sample residence time 
less than 20 seconds.

2.7 Summary

    Table E-3 to this appendix presents a summary of the general 
requirements for probe siting criteria with respect to distances and 
heights. It is apparent from table E-3 that different elevation 
distances above the ground are shown 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 Appendix E to Part 58--Summary of Probe Siting Criteria
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                   Horizontal or
                                                                                 vertical distance
                                                          Height from ground      from supporting      Distance from drip line of       Distance from
            Pollutant                      Scale             to probe \8\       structures \2\ \8\    trees to probe \8\ (meters)     roadways to probe
                                                               (meters)           to probe inlet                                        \8\ (meters)
                                                                                     (meters)
--------------------------------------------------------------------------------------------------------------------------------------------------------
SO2 \2\ \3\ \4\ \5\..............  Middle (300 m)        2.0-15..............  >=1.0...............  >=10.........................  N/A.
                                    Neighborhood Urban,
                                    and Regional (1 km).
CO \3\ \4\ \6\...................  Micro [downtown or    2.5-3.5; 2.0-7.0;     >=1.0...............  >=10.........................  2.0-10 for downtown
                                    street canyon         2.0-15.                                                                    areas or street
                                    sites], micro [near-                                                                             canyon microscale;
                                    road sites], middle                                                                              <=50 for near-road
                                    (300 m) and                                                                                      microscale; see
                                    Neighborhood (1 km).                                                                             Table E-2 to this
                                                                                                                                     appendix for middle
                                                                                                                                     and neighborhood
                                                                                                                                     scales.
O3 \2\ \3\ \4\...................  Middle (300 m)        2.0-15..............  >=1.0...............  >=10.........................  See Table E-1 to
                                    Neighborhood,                                                                                    this appendix for
                                    Urban, and Regional                                                                              all scales.
                                    (1 km).
NO2 \2\ \3\ \4\..................  Micro (Near-road [50- 2.0-7.0 (micro).....  >=1.0...............  >=10.........................  <=50 for near-road
                                    300 m]).                                                                                         micro-scale.
                                   Middle (300 m)......  2.0-15..............  >=1.0...............  >=10.........................
                                   Neighborhood, Urban,  2.0-15 (all other     >=1.0...............  >=10.........................  See Table E-1 to
                                    and Regional (1 km).  scales).                                                                   this appendix for
                                                                                                                                     all other scales.
Ozone precursors (for PAMS) \2\    Neighborhood and      2.0-15..............  >=1.0...............  >=10.........................  See Table E-1 to
 \3\ \4\.                           Urban (1 km).                                                                                    this appendix for
                                                                                                                                     all scales.

[[Page 5713]]

 
PM, Pb \2\ \3\ \4\ \7\...........  Micro, Middle,        2.0-7.0 (micro); 2.0- >=2.0 (all scales,    >=10 (all scales)............  2.0-10 (micro); see
                                    Neighborhood, Urban   7.0 (middle PM10-     horizontal distance                                  Figure E-1 to this
                                    and Regional.         2.5); 2.0-7.0 for     only).                                               appendix for all
                                                          near-road; 2.0-15                                                          other scales. <=50
                                                          (all other scales).                                                        for near-road.
--------------------------------------------------------------------------------------------------------------------------------------------------------
N/A--Not applicable.
\1\ When probe is located on a rooftop, this separation distance is in reference to walls, parapets, or penthouses located on 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 text).
\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'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.
\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 is in place as
  approved by the Regional Administrator pursuant to section 3 of appendix A to this part. For PM2.5, collocated monitor inlet heights should be within
  1 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.

3. Open Path Analyzers

3.1 Horizontal and Vertical Placement

    At least 80 percent of the monitoring path, must be located 
greater than or equal 2.0 and less than or equal to 15 meters above 
ground level for all O3 and SO2 monitoring 
sites, and for neighborhood or larger spatial scale NO2, 
and CO sites. Middle scale CO and NO2 sites must also 
have monitoring paths greater than or equal 2.0 and less than or 
equal to 15 meters above ground level. Microscale near-road 
monitoring sites are required to have monitoring paths greater than 
or equal 2.0 and less than or equal to 7.0 meters above ground 
level. The monitoring path for microscale carbon monoxide monitors 
that are being used to measure concentrations near roadways must be 
greater than or equal 2.0 and less than or equal to 7.0 meters above 
ground level. Those monitoring paths for microscale carbon monoxide 
monitors measuring concentrations near roadways in downtown areas or 
urban street canyons must be greater than or equal 2.5 and less than 
or equal to 3.5 meters above ground level. At least 90 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 location in order to interpret this particular 
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 is likely to 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. The plume from the local minor sources 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, at least 90 percent of the monitoring path 
must be away 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) Buildings and other obstacles may possibly 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 away 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.
    (b) A monitoring path located near or along a vertical wall is 
undesirable because air moving along the wall may be subject to 
possible 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 network design criteria 
regulations specified in appendix D to 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 due to their inherent potential sensitivity to certain 
types of interferences, or 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 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 pollutant 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 having a monitoring path located closer 
to an obstacle than this criterion allows 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.

[[Page 5714]]

    (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 cases where they 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 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, such as a roadway or a stationary source.

3.5 Spacing From Roadways

    Table E-4 to Appendix E to Part 58--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 distance
               Roadway average daily traffic, vehicles per day                Minimum distance     \1\ \2\ \3\
                                                                              \1\ \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 has not already been 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 to 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 having 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 those 
situations 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 
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. 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 and 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.

 Table E-5 to Appendix E to Part 58--Minimum Separation Distance Between
 Roadways and Monitoring Paths for Monitoring Neighborhood Scale Carbon
                                Monoxide
------------------------------------------------------------------------
                                                   Minimum distance \1\
Roadway average daily traffic, vehicles per day        \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.

[[Page 5715]]

 
\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 (NO[bdi2]) Monitoring Paths

    (a) In siting near-road NO2 monitors as required in 
section 4.3.2 of appendix D to 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 to 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. An open 
path analyzer having a monitoring path located closer to a roadway 
than allowed by the requirements in table E-4 to this appendix 
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 those situations 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 form table E-5. 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 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 to this appendix presents a summary of the general 
requirements for monitoring path siting criteria with respect to 
distances and heights. It is apparent from table E-6 that different 
elevation distances above the ground are shown 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 to Appendix E to Part 58--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        path \1\ \8\      to 90% of     path \1\ \8\       \1\ \8\
                                                    (meters)       monitoring       (meters)         (meters)
                                                                  path \1\ \8\
                                                                    (meters)
----------------------------------------------------------------------------------------------------------------
SO2 \3\ \4\ \5\ \6\..........  Middle (300 m)   2.0-15.........           >=1.0            >=10  N/A.
                                Neighborhood
                                Urban, and
                                Regional (1
                                km).
CO \4\ \5\ \7\...............  Micro [downtown  2.5-3.5; 2.0-             >=1.0            >=10  2.0-10 for
                                or street        7.0; 2.0-15.                                     downtown areas
                                canyon sites],                                                    or street
                                micro [near-                                                      canyon
                                road sites],                                                      microscale;
                                middle (300.                                                      <=50. for near-
                                m) and                                                            road
                                Neighborhood                                                      microscale;
                                (1.0 km).                                                         see Table E-5
                                                                                                  to this
                                                                                                  appendix for
                                                                                                  middle and
                                                                                                  neighborhood
                                                                                                  scales.
O3 \3\ \4\ \5\...............  Middle (300. m)  2.0-15.........           >=1.0            >=10  See Table E-4
                                Neighborhood,                                                     to this
                                Urban, and                                                        appendix for
                                Regional (1.0                                                     all scales.
                                km).
NO2 \3\ \4\ \5\..............  Micro (Near-     2.0-7.0                   >=1.0            >=10  <=50. for near-
                                road [50-300     (micro);.                                        road micro-
                                m]).                                                              scale.
                               Middle (300 m).  2.0-15.........           >=1.0            >=10
                               Neighborhood,    2.0-15 (all               >=1.0            >=10  See Table E-4
                                Urban, and       other scales).                                   to this
                                Regional (1                                                       appendix for
                                km).                                                              all other
                                                                                                  scales.
Ozone precursors (for PAMS)    Neighborhood     2.0-15.........           >=1.0            >=10  See Table E-4
 \3\ \4\ \5\.                   and Urban (1                                                      to this
                                km).                                                              appendix for
                                                                                                  all scales.
----------------------------------------------------------------------------------------------------------------
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 when the tree(s).
\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.

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 
still produce useful data for some purposes. The EPA will consider a 
written request from

[[Page 5716]]

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 establishing a new site, a waiver may be granted only if 
both of the following criteria are met:
    4.1.1 The 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 However, for an existing site, a waiver may be granted if 
either of the criteria in sections 4.1.1 and 4.1.2 of this appendix 
are 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, they in themselves, will not be acceptable 
reasons for granting a waiver. Written requests for waivers must be 
submitted to the Regional Administrator. Approved waivers must be 
renewed minimally every 5 years and ideally as part of the annual 
monitoring network plan accompanying 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 (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, 
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.

[[Page 5717]]

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.
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 general public in a nationally uniform and easy 
to understand manner. The AQI converts concentrations of pollutants 
for which the EPA has established 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 Appendix G to Part 58--AQI Categories
------------------------------------------------------------------------
                                                         And this color
         For this AQI             Use this descriptor          \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/.

    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, 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.
    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.
    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.
    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.
    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.
    i. 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.
    ii. 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.
    iii. 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.

[[Page 5718]]

    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.
    i. The AQI is the highest value calculated for each pollutant as 
follows:
    a. Identify the highest concentration among all of the monitors 
within each reporting area and truncate as follows:
    (1) 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

    (2) [Reserved]
    b. Using table 2 to this appendix, find the two breakpoints that 
contain the concentration.
    c. Using equation 1 to this appendix, calculate the index.
    d. Round the index to the nearest integer.

                                                Table 2 to Appendix G to Part 58--Breakpoints for the AQI
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                        These breakpoints                                                            Equal these AQI's
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                      PM10
  O3 (ppm) 8-     O3 (ppm) 1-   PM2.5  ([mu]g/m3)  ([mu]g/m3)   CO (ppm)    SO2 (ppb)   NO2 (ppb)
     hour          hour \1\           24-hour        24-hour     8-hour      1-hour      1-hour       AQI                      Category
 
--------------------------------------------------------------------------------------------------------------------------------------------------------
   0.000-0.054  ..............     0.0-(9.0-10.0)        0-54     0.0-4.4        0-35        0-53      0-50  Good.
   0.055-0.070  ..............    (9.1-10.1)-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-     361-649   151-200  Unhealthy.
                                                                                  304
   0.106-0.200     0.205-0.404        125.5-225.4     355-424   15.5-30.4    \3\ 305-    650-1249   201-300  Very Unhealthy.
                                                                                  604
   0.201-(\2\)          0.405+             225.5+        425+       30.5+    \3\ 605+       1250+      301+  Hazardous.\4\
--------------------------------------------------------------------------------------------------------------------------------------------------------
\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 [mu]g/m\3\; PM10 24-hour--604 [mu]g/m\3\; CO ppm--50.4 ppm; SO2 1-hour--1004 ppb; and NO2 1-hour--2049 ppb.

    ii. 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.

Equation 1 to Appendix G to Part 58
[GRAPHIC] [TIFF OMITTED] TP27JA23.018

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.

    iii. 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

    iv. 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:
    a. 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;
    b. 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;
    c. 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;
    d. Apply equation 1 to this appendix for 210 [micro]g/m\3\, 
PM10:

[[Page 5719]]

[GRAPHIC] [TIFF OMITTED] TP27JA23.019

    e. Apply equation 1 to this appendix for 0.156 ppm, 1-hour 
O3:
[GRAPHIC] [TIFF OMITTED] TP27JA23.020

    f. Apply equation 1 to this appendix for 0.130 ppm, 8-hour 
O3:
[GRAPHIC] [TIFF OMITTED] TP27JA23.021

    g. 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. 2023-00269 Filed 1-26-23; 8:45 am]
 BILLING CODE 6560-50-P