Wildland Fire Management: Better Information and a Systematic
Process Could Improve Agencies' Approach to Allocating Fuel
Reduction Funds and Selecting Projects (28-SEP-07, GAO-07-1168).
Recognizing that millions of acres are at risk from wildland
fire, the federal government expends substantial resources on
thinning brush, trees, and other potentially hazardous fuels to
reduce the fire risk to communities and the environment. However,
questions have been raised about how the agencies responsible for
wildland fire management--the Department of Agriculture's Forest
Service and the Department of the Interior's (Interior) Bureau of
Indian Affairs (BIA), Bureau of Land Management (BLM), Fish and
Wildlife Service (FWS), and National Park Service (NPS)--allocate
their fuel reduction budgets and select projects. GAO was asked
to report on the agencies' processes for allocating funds and
selecting projects, and on how, if at all, these processes could
be improved to better ensure that they contribute to the
agencies' overall goal of reducing risk. To obtain this
information, GAO visited headquarters and field offices of all
five agencies; obtained data on fuel reduction funding and
accomplishments; and reviewed previous evaluations of the fuel
reduction program.
-------------------------Indexing Terms-------------------------
REPORTNUM: GAO-07-1168
ACCNO: A76890
TITLE: Wildland Fire Management: Better Information and a
Systematic Process Could Improve Agencies' Approach to Allocating
Fuel Reduction Funds and Selecting Projects
DATE: 09/28/2007
SUBJECT: Allocation (Government accounting)
Budgeting
Budgets
Environmental protection
Federal agencies
Federal funds
Forest conservation
Forest fires
Forest management
Funds management
Land management
Policy evaluation
Risk assessment
Risk management
Wildfires
Policies and procedures
******************************************************************
** This file contains an ASCII representation of the text of a **
** GAO Product. **
** **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced. Tables are included, but **
** may not resemble those in the printed version. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
******************************************************************
GAO-07-1168
* [1]WILDLAND FIRE MANAGEMENT
* [2]Better Information and a Systematic Process Could Improve Agencies'
Approach to Allocating Fuel Reduction Funds and Selecting Projects
* [3]Contents
* [4]Results in Brief
* [5]Background
* [6]Wildland Fire Is a Natural Process
* [7]Five Agencies Are Responsible for Wildland Fire
Management
* [8]Expansion into the Wildland-Urban Interface Has
Increased, as Has the Federal Focus on Wildland Fire
Management
* [9]GAO Has Reviewed Agencies' Fuel Treatment Programs
* [10]The Forest Service Uses a Mix of Quantitative and
Judgmental Processes and Considers a Range of Factors in
Allocating Funds and Selecting Projects
* [11]At the National Level, the Forest Service's
Allocation Process Increasingly Relies on a
Quantitative Approach
* [12]Some Forest Service Regions Use Quantitative
Allocation Processes, While Others Rely More on
Professional Judgment
* [13]National Forests Select Projects Using Quantitative
and Judgmental Processes
* [14]Interior and Its Agencies Use a Mix of Quantitative and
Judgmental Processes and Consider a Range of Factors in
Allocating Funds and Selecting Projects
* [15]Interior Allocates Funds to Its Agencies Primarily
on the Basis of Historical Funding Levels
* [16]BLM Increasingly Uses Quantitative Processes in
Allocating Funds and Selecting Projects
* [17]BLM Allocates Funds to Its State Offices Primarily
on the Basis of Historical Funding Levels but Plans to
Use a More Quantitative Approach in 2008
* [18]The Majority of BLM State Offices Incorporate
Quantitative Approaches in Their Allocation Processes
* [19]The Majority of BLM Field Units Incorporate
Quantitative Approaches into Their Project Selection
Processes
* [20]BIA Allocates Funds Largely on the Basis of Units'
Performance History, while FWS and NPS Use Quantitative
and Judgmental Processes
* [21]BIA and FWS Headquarters Allocate Funds Using
Quantitative Processes, While NPS Headquarters
Allocates Funds Primarily on the Basis of
Historical Funding Levels
* [22]BIA, FWS, and NPS Regional Offices Allocate
Funds to Field Units Using Quantitative and
Judgmental Processes
* [23]Local BIA, FWS, and NPS Units Select Projects
Using Quantitative and Judgmental Processes
* [24]Several Improvements Could Help Better Ensure That Fuel
Reduction Funds Are Allocated to Effectively Reduce Risk
* [25]The Agencies Do Not Consistently Assess All
Elements of Risk When Allocating Funds
* [26]The Agencies Do Not Consider Treatment
Effectiveness in Their Allocation Processes Because
They Have No Measure for Effectiveness
* [27]The Agencies Often Consider Costs, but Not
Cost-Effectiveness, When Allocating Funds
* [28]The Agencies Have Not Established Clear Guidance on
the Relative Importance of Factors Used in Setting
Priorities
* [29]Agencies' Allocation Processes Are Not Always
Systematic
* [30]Conclusions
* [31]Recommendations for Executive Action
* [32]Agency Comments and Our Evaluation
* [33]Appendix I: Objectives, Scope, and Methodology
* [34]Fuel Reduction Funding Allocation and Project Selection
Processes
* [35]Potential Improvements to Agency Processes to Better
Ensure They Contribute to Reducing Risk
* [36]Appendix II: Forest Service and Interior Fuel Reduction
Funding Allocations, Fiscal Years 2005, 2006, and 2007
* [37]Appendix III: Summary of Fuel Treatment Accomplishments for
the Forest Service and Interior, Fiscal Years 2005 and 2006
* [38]Appendix IV: Comments from the Department of the Interior and
the Forest Service
* [39]Appendix V: GAO Contact and Staff Acknowledgments
* [40]Related GAO Products
* [41]Order by Mail or Phone
United States Government Accountability Office
Report to Congressional Requesters
GAO
September 2007
WILDLAND FIRE
MANAGEMENT
Better Information and a Systematic Process Could Improve Agencies' Approach to
Allocating Fuel Reduction Funds and Selecting Projects
GAO-07-1168
Contents
Letter Appendix I Results in Brief Background The Forest [42]1 6
Service Uses a Mix of Quantitative and 10 19
Judgmental Processes and Considers a Range of 29 44
Factors in Allocating Funds and Selecting 64 66
Projects Interior and Its Agencies Use a Mix 67 69
of Quantitative and Judgmental Processes and
Consider a Range of Factors in Allocating
Funds and Selecting Projects Several
Improvements Could Help Better Ensure That
Fuel Reduction Funds Are Allocated to
Effectively Reduce Risk Conclusions
Recommendations for Executive Action Agency
Comments and Our Evaluation Objectives,
Scope, and Methodology
Appendix II Forest Service and Interior Fuel Reduction
Funding Allocations, Fiscal Years 2005, 2006, [43]77
and 2007
Appendix III Summary of Fuel Treatment Accomplishments for
the Forest Service and Interior, Fiscal Years [44]91
2005 and 2006
Appendix IV Comments from the Department of the Interior [45]100
and the Forest Service
Appendix V GAO Contact and Staff Acknowledgments [46]101
Related GAO Products [47]102
Tables
Table 1: Factors Considered in Forest Service Fuel Reduction
Funding Allocation Model [48]21
Table 2: Forest Service Regions' Fiscal Year 2007 Fuel
Reduction
Priority Scores and Funding Allocations [49]23
Table 3: BLM Allocations to State Offices, Fiscal Year 2007 [50]35
Table 4: Factors and Factor Categories BLM Considers in BLM
Fuel Reduction Funding Allocation Model [51]37
Table 5: Regional Offices GAO Visited [52]71
Table 6: Field Units GAO Visited [53]73
Table 7: Total Appropriations to Forest Service, and
Allocations to
Interior Agencies, Fiscal Years 2005, 2006, and 2007 [54]78
Table 8: Forest Service Allocations to Regions and
Headquarters,
Fiscal Years 2005, 2006, and 2007 [55]80
Table 9: Interior Allocations to BLM, BIA, FWS, and NPS,
Including
WUI and Non-WUI Allocations, Fiscal Years 2005, 2006,
and 2007 [56]82
Table 10: BLM Allocations to State Offices and Headquarters,
Fiscal Years 2005, 2006, and 2007 [57]83
Table 11: BIA Allocations to Regions and the National
Interagency
Fire Center, Fiscal Years 2005, 2006, and 2007 [58]85
Table 12: NPS Allocations to Regions and the Washington
Office,
Fiscal Years 2005, 2006, and 2007 [59]87
Table 13: FWS Allocations to Regions and Headquarters,
Fiscal
Years 2005, 2006, and 2007 [60]89
Table 14: Summary of Fiscal Years 2005 and 2006 Fuel
Reduction
Accomplishments for Interior and Forest Service [61]92
Table 15: Summary of Fiscal Years 2005 and 2006 Fuel
Reduction
Accomplishments for Forest Service Regions [62]93
Table 16: Summary of Fiscal Years 2005 and 2006 Fuel
Reduction
Accomplishments for BLM State Offices [63]94
Table 17: Summary of Fiscal Years 2005 and 2006 Fuel
Reduction
Accomplishments for BIA Regions [64]96
Table 18: Summary of Fiscal Years 2005 and 2006 Fuel
Reduction
Accomplishments for NPS Regions [65]98
Table 19: Summary of Fiscal Years 2005 and 2006 Fuel
Reduction
Accomplishments for FWS Regions [66]99
Figures
Figure 1: Annual Appropriation and Allocation Process for Fuel
Reduction Funds [67]4
Figure 2: Distribution of Total Land Managed by the Forest
Service,
BIA, BLM, FWS, and NPS [68]11
Figure 3: A Mechanical Thinning Project for Fuel Reduction
on
BLM Land in California [69]13
Figure 4: Prescribed Fire for Fuel Reduction on Forest
Service
Land in South Carolina [70]13
Figure 5: Forest Service Regions' Fuel Reduction Priority
Scores as
a Percentage of Total, Compared to Regions' Funding
Allocations as a Percentage of Total Allocations, Fiscal
Year 2007 [71]24
Figure 6: Percentage of Interior's Total Fuel Reduction
Funds
Allocated to the Interior Agencies, Fiscal Years 2001
through 2007 [72]30
Figure 7: Density of Wildland-Urban Interface Treatments
and
Population Density, by ZIP Code [73]50
Figure 8: Location of Federal Lands and Populated Areas in
the
Continental United States [74]52
Figure 9: Map of Los Angeles County Wildland-Urban
Interface
Fuel Reduction Treatments Completed in 2005 and 2006,
and Population Density [75]54
Figure 10: Map of Rio Blanco County Wildland-Urban
Interface
Fuel Reduction Treatments Completed in 2005 and 2006,
and Population Density [76]56
Figure 11: Agency Funding Levels as a Percentage of Total
Fuel
Reduction Funding, Fiscal Year 2007 [77]79
Abbreviations
BIA Bureau of Indian Affairs
BLM Bureau of Land Management
FPA Fire Program Analysis
FWS Fish and Wildlife Service
GAO Government Accountability Office
GIS Geographic information system
HFI Healthy Forests Initiative
HFRA Healthy Forests Restoration Act
NEPA National Environmental Policy Act
NPS National Park Service
USDA United States Department of Agriculture
WUI Wildland-urban interface
This is a work of the U.S. government and is not subject to copyright
protection in the United States. The published product may be reproduced
and distributed in its entirety without further permission from GAO.
However, because this work may contain copyrighted images or other
material, permission from the copyright holder may be necessary if you
wish to reproduce this material separately.
United States Government Accountability Office
Washington, DC 20548
September 28, 2007
The Honorable Norman D. Dicks
Chairman
The Honorable Todd Tiahrt
Ranking Member
Subcommittee on Interior, Environment, and Related Agencies
Committee on Appropriations
House of Representatives
The Honorable Raul M. Grijalva
Chairman
The Honorable Rob Bishop
Ranking Member
Subcommittee on National Parks, Forests, and Public Lands
Committee on Natural Resources
House of Representatives
The Honorable Greg Walden
House of Representatives
Decades of fire suppression in the nation's forests, together with such
practices as logging followed by dense tree planting, have resulted in the
accumulation of brush, small trees, and other vegetation that can fuel
wildland fires. Similarly, the nation's rangelands have suffered from
decades of fire suppression and livestock overgrazing, which have degraded
ecosystems and made the rangelands vulnerable to the invasion of
flammable, nonnative species, such as cheat grass. This accumulation and
alteration of vegetation, as well as drought and other stresses related to
climate change, have fueled wildland fires. Collectively, these fires have
cost billions of dollars to suppress, forced thousands from their homes,
and damaged cultural and natural resources. The impacts of these fires
have intensified as more and more communities develop in areas that are
adjacent to fire-prone lands--the wildland-urban interface.
In response to the increasing threat of wildland fires, the federal
agencies responsible for wildland fire management developed the National
Fire Plan. ^[78]1 These agencies are the Department of Agriculture's
Forest Service and the Department of the Interior's (Interior) Bureau of
Indian Affairs (BIA), Bureau of Land Management (BLM), Fish and Wildlife
Service (FWS), and National Park Service (NPS). Two components of the
National Fire Plan are a 10-year strategy and an implementation plan for
protecting communities and the environment that were developed in 2001 and
2002, respectively, by the Secretaries of Agriculture and of the Interior,
along with governors of western states and other interested parties, and
updated in December 2006. The 2002 plan emphasized reducing hazardous fuel
in forests and rangelands to mitigate the risk from wildland fire. In
2003, Congress passed the Healthy Forests Restoration Act (HFRA), ^[79]2
with the stated purpose of reducing wildland fire risk to communities,
municipal water supplies, and other at-risk federal land through a
collaborative process of planning, setting priorities, and implementing
fuel reduction projects. ^[80]3 HFRA also authorized grants to commercial
facilities that use biomass--that is, small-diameter trees and
branches--to offset the costs incurred to purchase biomass. Fuel reduction
projects can generate substantial amounts of biomass.
According to the updated 10-Year Strategy Implementation Plan, the goal of
the fuel reduction program is to reduce the risk of wildland fire to
communities and the environment. Fuel reduction projects--using prescribed
fire, mechanical thinning, herbicides, grazing, or combinations of these
methods--are intended to remove or modify wildland fuel to reduce the
potential for severe wildland fires, lessen the damage caused by fires,
limit the spread of flammable invasive species, and restore and maintain
healthy ecosystems. Local land management units, such as
national forests and parks, are typically responsible for selecting and
implementing fuel reduction projects.
^1The National Fire Plan comprises multiple documents, including (1) a
September 2000 report from the Secretaries of Agriculture and of the
Interior to the President in response to the wildland fires of 2000, (2)
congressional direction accompanying substantial new appropriations in
fiscal year 2001, and (3) several strategies to implement all or parts of
the plan. For a description of these documents and their contents, goals,
and relationships to one another, see Severe Wildland Fires: Leadership
and Accountability Needed to Reduce Risks to Communities and Resources,
GAO-02-259 (Washington, D.C.: Jan. 31, 2002).
^2Pub. L. No. 108-148 (2003).
^3HFRA defines "federal land" to include land administered by the Forest
Service and BLM. Consequently, HFRA fuel reduction project authorities are
available only to the Forest Service and BLM, and its fuel reduction
project requirements apply only to these agencies as well. In some cases,
BIA, FWS, and NPS have chosen to comply with some of the requirements.
Since 2001, Congress has appropriated more than $3.2 billion in fuel
reduction funds to the Forest Service and Interior. For 2007, the Forest
Service received about $300 million, and Interior received about $200
million. ^[81]4 After receiving its annual appropriation, the Forest
Service allocates funds to its nine regional offices which in turn
allocate funds to individual national forests and grasslands. Interior,
upon receiving its annual appropriation, allocates funds to BIA, BLM, FWS,
and NPS. BLM generally receives about 50 percent of Interior's funding,
BIA about 20 percent, and FWS and NPS about 15 percent each. These
agencies then allocate funds to their regional or state offices, which, in
turn, allocate funds to individual field units, such as national parks or
wildlife refuges. (BIA, FWS, and NPS have regional offices, while BLM has
state offices. For the purposes of this report, we refer to all of these
as regional offices when we discuss the Interior agencies collectively.)
Figure 1 shows the annual appropriation and allocation process.
^4Years cited in this report refer to fiscal years except where otherwise
specified.
Figure 1: Annual Appropriation and Allocation Process for Fuel Reduction Funds
Recognizing that treating all of the land in need of fuel reduction may
take decades, the agencies have acknowledged the importance of setting
priorities for which lands are to receive treatment so that they can
select those treatments that will be the most effective at reducing the
risks from wildland fire. However, we have found a long-standing pattern
of shortcomings in the processes the Forest Service and Interior agencies
use to identify and set priorities for lands needing fuel reduction.
Between 1999 and 2003, we reported that the Forest Service and Interior
had made it a priority to treat lands at the highest risk from wildland
fire, but they had not identified the amount or location of such lands and
had not issued guidance specific enough for field staff to set priorities
for individual projects. We concluded that the agencies needed a cohesive
strategy outlining long-term options and associated costs for reducing
fuel. ^[82]5 In subsequent reports, we noted, among other things, the
progress the agencies had made in improving their data, but reiterated
that they needed to complete ongoing efforts to identify lands at risk
from wildland fire--by collecting information on the hazards, the
likelihood of fire occurring, and the values at risk--so funds could be
targeted to such lands. We also reiterated the need to develop a cohesive
strategy that included long-term options and associated costs so that
Congress could make informed decisions about cost-effective approaches to
fuel reduction. ^[83]6
In this context, you asked us to report on the agencies' current processes
for identifying and setting priorities for fuel reduction. Specifically,
you asked us to (1) identify the processes the Forest Service, Interior,
and the four Interior agencies use to allocate fuel reduction funds and
select projects for implementation, including the factors that influence
these processes, and (2) determine how, if at all, these processes could
be improved to better ensure that they contribute to the agencies' goal of
effectively reducing the risk of wildland fire to communities and the
environment.
To address these objectives, we met with national, regional, state, and
local officials of the Forest Service, Interior, and Interior agencies. At
the national level, we met with agency officials at their Washington,
D.C., headquarters, as well as at the National Interagency Fire Center in
Boise, Idaho. At the regional and state levels, we used a structured
interview guide to speak, in person or by telephone, with officials in all
Forest Service regional and BLM state offices, as well as with officials
in selected BIA, FWS, and NPS regional offices that collectively received
a substantial portion of each agency's fuel reduction funds. At the local
level, we visited
20 local units, such as national forests and BLM field offices, in eight
states to gain a better understanding of their processes for selecting
fuel reduction projects for implementation. We selected local units that
are diverse in geographic location, predominant vegetation type, and
proximity to communities and development. We also obtained and reviewed
applicable laws, regulations, and agencywide and regional policies; agency
data on funding allocations; and electronic data on the extent of agency
fuel treatment activities. We tested these data and found that they were
sufficiently reliable for the purposes of this review. Finally, we
interviewed several nonfederal parties, including representatives from
environmental groups and the Western Governors' Association. ^[84]7 We
conducted our work from August 2006 to September 2007 in accordance with
generally accepted government auditing standards. See appendix I for a
detailed description of our methodology.
^5GAO, Western National Forests: A Cohesive Strategy Is Needed to Address
Catastrophic Wildfire Threats, GAO/RCED-99-65 (Washington, D.C.: Apr. 2,
1999); Reducing Wildfire Threats: Funds Should Be Targeted to the Highest
Risk Areas, GAO/T-RCED-00-296 (Washington, D.C.: Sept. 13, 2000);
GAO-02-259; Wildland Fire Management: Additional Actions Required to
Better Identify and Prioritize Lands Needing Fuels Reduction, GAO-03-805
(Washington, D.C.: Aug. 15, 2003).
^6GAO, Wildland Fires: Forest Service and BLM Need Better Information and
a Systematic Approach for Assessing the Risks of Environmental Effects,
GAO-04-705 (Washington, D.C.: June 24, 2004); Wildland Fire Management:
Important Progress Has Been Made, but Challenges Remain to Completing a
Cohesive Strategy, GAO-05-147 (Washington, D.C.: Jan. 14, 2005); Wildland
Fire Management: Update on Federal Agency Efforts to Develop a Cohesive
Strategy to Address Wildland Fire Threats, GAO-06-671R (Washington, D.C.:
May 1, 2006).
Results in Brief
In allocating fuel reduction funds and selecting projects, the Forest
Service--at the national, regional, and local levels--uses both
quantitative processes (such as computer models or scoring systems) and
professional judgment and, in doing so, considers multiple factors, such
as risk assessments, treatment cost per acre, and collaboration with
communities or other entities. Specifically, for 2007, we found the
following:
o At headquarters, the Forest Service began using a computer model to
influence funding allocations to regions. To set priorities for each
region's fuel reduction funding, the model considers multiple factors,
including some intended to assess risk, such as the potential for fires
occurring in each region and their expected severity, as well as other
factors, such as regional use of biomass removed in fuel treatments and
treatment cost per acre. However, the Forest Service's funding allocations
to its regions were not consistent in all cases with the priority scores
resulting from the model, with some high-scoring regions receiving less
funding than some lower-scoring regions. These disparities occurred for a
number of reasons, such as the higher costs of fuel reduction in some
areas. However, the model did not substantially influence the agency's
2007 allocations; instead, the Service relied largely on prior year
funding levels and used results from the model only to make minor
adjustments. Officials said they expect
the model's results to have more influence on future allocation decisions,
but curbed this influence initially because they were still refining the
model and wanted to maintain relatively stable regional funding levels.
^7The Western Governors' Association is an independent, nonpartisan
organization of governors representing 19 western states. The governors
use the association to develop and advocate policies that reflect regional
interests.
o In the regions, each region determined how to allocate funds to
national forests and what factors to consider, as long as they were
consistent with the factors used in the national allocation model.
Four of the Service's nine regions relied primarily on quantitative
data in their allocation processes, while five relied primarily on
professional judgment. For example, the Rocky Mountain region used a
computer model that evaluated data on multiple factors, such as
vegetative conditions and areas of insect-killed trees, while the
Southern region convened a group of officials who used professional
judgment and considered historical funding levels, the capabilities of
the forests, per-acre treatment cost, local priorities, and acreage
targets when allocating funds. Beginning with the 2008 allocations,
the Forest Service plans to require regions to use the headquarters
model to inform allocation decisions.
o Locally, national forests had discretion in determining how to
select projects. Some used quantitative, data-driven processes,
while others relied primarily on professional judgment or
collaborative processes involving other agencies and local
communities. Forests considered a range of factors--similar to
those used at the national and regional levels--when selecting
projects.
o Like the Forest Service, Interior and its agencies' national,
regional, and local offices used both quantitative and judgmental
processes for allocating fuel reduction funds and selecting
projects and considered multiple factors that are similar to
those the Forest Service uses. More specifically, for fiscal year
2007, we found the following:
o Interior allocated funds to its four agencies primarily on the basis
of historical funding levels; however, Interior is developing a
computer model similar to the Forest Service's, and it used the model
to allocate 5 percent of its funds in 2007. Interior agencies, in
turn, had the flexibility to determine how to allocate funds, within
the parameters of departmental guidance.
o BLM headquarters allocated funds primarily on the basis of historical
funding levels, but officials told us that, starting in 2008, they
plan to use a quantitative process incorporating multiple factors,
such as the potential for fires to occur, treatment cost, and local
risk ratings.
o BIA headquarters allocated funds using quantitative processes, but
these processes generally emphasized a single factor--BIA units' past
performance (measured in acres treated) in carrying out fuel reduction
activities.
o FWS headquarters allocated funds to regional offices using a
quantitative process--a model that considers a range of factors, such
as the history of fires, fuel conditions, and communities at risk.
o NPS headquarters allocated funds to regions largely on the basis of
historical funding levels. These funding levels were originally
determined using a model that assessed the risk from wildland fire,
among other factors.
At the regional and local levels, regional offices and field units in all
four agencies used a variety of processes to allocate funds and select
projects for implementation. Some processes emphasized quantitative data,
while others emphasized professional judgment. The regional and local
offices also considered a range of factors, consistent with departmental
direction.
Although the Forest Service and Interior have begun taking action to
enhance their funding allocation processes, there are additional steps
they could take to improve these processes to better ensure they advance
the agencies' goal of effectively reducing the risk of wildland fire to
communities and the environment. Specifically, the agencies could improve
their processes by taking the following five steps:
o Consistently using risk assessments. The agencies did not consistently
use risk assessments in their 2007 allocation processes at the national,
regional, and local levels, in some cases because national or regional
offices expected local units to do so. However, agency officials cannot be
sure that projects identified as high risk locally would likewise be the
highest risk from a regional or national perspective. Even when the
agencies did conduct risk assessments, they found it difficult to
meaningfully distinguish between higher- and lower-priority locations
because one key value at risk--the wildland-urban interface--is broadly
defined and many different areas are classified as interface. Although the
agencies' guidance sets a priority on projects in the interface, it does
not specify whether some of the areas classified as interface ought to be
higher priority than others. As a result, projects as diverse as those
protecting remote power lines, individual ranch houses, or large suburban
subdivisions can all fall within the wildland-urban interface category
and, thus, be designated high priority--complicating agency officials'
attempts to identify and
direct their limited resources toward the highest-priority areas.
o Developing information on treatment effectiveness. The agencies did
not consider treatment effectiveness--that is, how much risk reduction
can be achieved, and for how long--when making allocation decisions
because they currently have no measure for effectiveness, although
they are working to develop such a measure. Without information on
treatment effectiveness, the agencies could be funding treatments that
have little effect on reducing risk.
o Developing information on cost effectiveness. The agencies often
considered costs when allocating funds, but not cost effectiveness--
primarily because they lack information on treatment effectiveness.
Without such information, it is difficult to know whether a
treatment's cost is warranted or to compare the cost effectiveness of
different potential treatments to decide how to optimally allocate
funds.
o Clarifying the importance of factors unrelated to risk or
effectiveness. The agencies often considered factors other than risk,
treatment effectiveness, and cost effectiveness when allocating funds
and selecting projects. When these external factors--such as funding
stability and the use of biomass resulting from fuel reduction
treatments--have considerable influence, it is difficult for the
agencies to ensure that they are allocating funds so that treatments
will most effectively reduce risk.
o Applying more systematic processes. The agencies sometimes relied
exclusively on professional judgment when allocating funds or
selecting projects. Although judgmental processes might result in
allocations that maximize risk reduction, the agencies cannot be
assured that they routinely do because such processes are not
necessarily systematic--that is, methodical, based on criteria, and
applied consistently.
To improve the agencies' ability to ensure that fuel reduction funds are
directed to most effectively reduce the risk from wildland fire, we are
recommending that the Secretaries of Agriculture and of the Interior take
actions to implement a more systematic allocation process; develop
additional information on risk, treatment effectiveness, and
cost-effectiveness to support the process; and clarify the relative
importance of multiple criteria for setting priorities in allocation and
project selection decisions. We provided a draft of this report to the
Secretaries of Agriculture and of the Interior for review and comment. The
Forest Service and the Department of the Interior generally agreed with
our report; their joint comment letter is presented in appendix IV.
Background
Wildland Fire Is a Natural Process
Although its effect on communities can be devastating, wildland fire is a
natural and necessary process that provides many benefits to ecosystems,
such as maintaining habitat diversity, recycling soil nutrients, limiting
the spread of insects and disease, and promoting new growth by causing the
seeds of fire-dependent species to germinate. Wildland fire also
periodically removes brush, small trees, and other vegetation that can
otherwise accumulate and increase the size, intensity, and duration of
subsequent fires. However, human uses and land management practices--
including decades of wildland fire suppression--have excluded fire from
ecosystems, reducing the normal frequency of wildland fire and
subsequently causing an accumulation of vegetation. Federal researchers
have estimated that unnaturally dense fuel accumulations on 90 million to
200 million acres of federal lands in the contiguous United States place
these lands at an elevated risk of severe wildland fire and that these
conditions also hold true for many nonfederal lands.
Most lands in the United States evolved with fire, and each ecosystem has
a characteristic fire regime that describes the role fire plays in the
ecosystem, including typical fire frequency, scale, intensity, and
duration. These regimes are numbered I through V, with fire regime I
characterized by low-severity fires that historically occurred every 35
years or less, fire regime II characterized by high-severity fires that
historically occurred every 35 years or less, fire regime III
characterized by mixed-severity fires that historically occurred every 35
to 100 or more years, fire regime IV characterized by high-severity fires
that historically occurred every 35 to 100 or more years, and fire regime
V characterized by high-severity fires that historically occurred every
200 or more years. Many ecosystems-- particularly those in fire regimes I
and II--have now missed numerous fire cycles as a result of past
suppression policies and other land management practices. This departure
from the natural fire regime is categorized by a measure called condition
class, which the agencies have used as a generalized rating for the risk
of uncharacteristic wildland fires that may cause undesirable ecological
consequences. Ecosystems in condition class 1 are generally within their
historical fire return interval, so fires in these areas pose little risk
to natural processes--although fires in such ecosystems may still pose a
high risk to communities. Areas in condition classes 2 and 3 have moderate
to significant departures from historical fire experiences. In such areas,
fuel that would typically have burned periodically has instead
accumulated, posing a higher risk that uncharacteristically large amounts
of vegetation and other natural resources would be lost from wildland
fire; fires in these areas may also pose a high risk to communities.
Five Agencies Are Responsible for Wildland Fire Management
The Forest Service, BIA, BLM, FWS, and NPS are responsible for wildland
fire management, including fuel reduction. These five agencies manage
about 700 million acres of land in the United States, including national
forests, national grasslands, Indian reservations, national parks, and
national wildlife refuges. The Forest Service and BLM manage the majority
of these lands, with the Forest Service managing about 190 million acres
and BLM managing about 260 million acres; BIA, FWS, and NPS each manage
less than 100 million acres. Figure 2 shows the distribution of land among
the five agencies. Each agency has between 7 and 12 regional offices that
oversee field units.
Figure 2: Distribution of Total Land Managed by the Forest Service, BIA,
BLM, FWS, and NPS
BLM BIA
NPS
FWS
Forest Service
Forest Service Interior agencies
Source: GAO analysis of Forest Service and Interior data.
Each year, the Forest Service, Interior, and Interior agencies set
performance targets for region- and state-level fuel reduction by
establishing the number of acres the agencies expect to be treated--both
within and outside of the wildland-urban interface--using the funds
allocated. For example, for fiscal year 2007, BLM assigned a target of
almost 90,000 acres to its Oregon/Washington state office and specified
that two-thirds of the acres should be in the wildland-urban interface.
Between 2001 and August 2007, land managers treated more than 18 million
acres under the fuel reduction program, including about
8.5 million acres near communities. ^[85]8 These acres include federal,
state, and private land, because, in addition to conducting fuel
treatments on federal lands, the agencies work with and grant funds to
local communities to conduct fuel reduction treatments on state and
private lands. These acres also include those that have been treated more
than once.
The agencies generally reduce fuel using either mechanical treatments, in
which equipment--such as chainsaws, chippers, bulldozers, or mowers-- is
used to cut vegetation, or prescribed burning, in which fires are
deliberately set by land managers to restore or maintain desired
vegetation conditions. ^[86]9 Figure 3 depicts a mechanical thinning
project, and figure 4 depicts a prescribed burn.
^8The agencies treated an additional 1.8 million acres from 2004 to August
2007 through other land management activities.
^9The agencies also conduct some treatments using other methods, such as
applying herbicides and allowing animals to graze on the land.
Figure 3: A Mechanical Thinning Project for Fuel Reduction on BLM Land in
California
Source: BLM.
Figure 4: Prescribed Fire for Fuel Reduction on Forest Service Land in
South Carolina
Source: Forest Service.
Although prescribed burning can be risky, burning under specified fuel and
weather conditions enables fire to be controlled at a relatively low
intensity level within a confined area. Prescribed burning is very
effective in removing smaller vegetation, such as grasses, leaves, pine
needles, and twigs, but is not as effective in removing larger fuel, such
as trees, or in thinning stands to desired densities. In contrast,
mechanical treatment methods are effective in thinning stands and removing
larger vegetation but may increase the amount of smaller fuel on the
ground, including tree tops and limbs (referred to as slash) and other
debris from thinning. As a result, some fuel reduction projects use
multiple treatment methods and may span several years. For example, a
field unit may first treat an area mechanically to thin accumulated
vegetation and then follow with a prescribed burn to remove remaining
slash and litter on the ground.
In addition to reducing the risk of fire to communities and the
environment, one of the long-term goals of the fuel reduction program is
to allow fire to resume its natural role. By conducting treatments,
including creating fire breaks to help contain the spread of fire, the
agencies increase the amount of land where naturally ignited fires can
safely be allowed to burn. Under wildland fire use policies, land managers
may allow wildland fires that are naturally ignited to continue to burn,
as long as fuel and weather conditions are appropriate and the fire is
located within an area designated for wildland fire use. ^[87]10 Managers
are thus able to use natural fire to meet resource objectives, such as
removing excess vegetation.
Although the five agencies all reduce fuel in order to reduce risk to
communities and the environment, their fuel reduction programs reflect
differences in their missions, predominant vegetation types, and allowable
land uses. For example, FWS's mission is focused on the conservation of
wildlife habitat, and the agency generally conducts more prescribed burns
than mechanical treatments because such burns frequently improve habitat
as well as reduce risk; the agency has been conducting prescribed burns
since the 1930s. Similarly, prescribed burns, as well as wildland fire
use, are the preferred fuel treatment methods at NPS and have been used by
the agency for decades. NPS prefers these treatment methods over
mechanical treatments because its mission emphasizes preservation of
natural and cultural resources, and fire is a natural process that better
aligns with this mission. Regarding predominant vegetation types, BLM's
lands are largely rangelands, while lands managed by other agencies, such
as the Forest Service and FWS, include more forests. As a result of this
difference, BLM not only conducts mechanical treatments and prescribed
burns, as do the other agencies, but also uses herbicides to reduce fuel,
especially where rangelands have been invaded by exotic plants such as
cheat grass. Agency differences in allowable land use also affect their
fuel reduction programs. For example, the Forest Service, BIA, and BLM
have active commercial timber programs, and field units may therefore
conduct fuel treatments that benefit both the timber and fuel reduction
programs. BIA and NPS also manage lands with numerous archaeological
sites, which must be considered when conducting treatments. In contrast,
the majority of BLM's land is used for grazing, and, as a result, BLM
coordinates fuel treatments with potentially affected ranchers.
^10Interagency policy directs land managers to select firefighting
strategies in accordance with local federal units' land and fire
management plans. If a plan has not been developed and approved, the
policy directs land managers to suppress the fire. Thus, under the policy,
the areas where wildland fire use is allowed must be defined in a fire
management plan, along with prescribed weather and other conditions. The
fires are monitored, and if weather conditions change in a way that would
potentially allow the fires to escape from the designated areas, the fires
are suppressed.
Expansion into the Wildland-Urban Interface Has Increased, as Has the
Federal Focus on Wildland Fire Management
Urban and suburban expansion into the wildland-urban interface has
increased the number of communities and structures at risk of wildland
fire near federal lands that the five agencies manage. Experts estimate
that almost 60 percent of all new housing units built in the 1990s were
located in the wildland-urban interface and that this growth trend
continues. They also estimate that more than 30 percent of housing units
overall are located in the wildland-urban interface and that the interface
covers about 9 percent of the nation's land. In addition to housing units,
other types of infrastructure are located in the wildland-urban interface,
including power lines, campgrounds and other recreation facilities, oil
and gas wells, communications towers, and roads.
After the National Fire Plan was developed, the agencies began receiving
sharp increases in funding for fuel reduction and, since 2001, Congress
has appropriated between about $400 million and $500 million annually for
fuel reduction under the plan. (App. II shows agency fuel reduction
funding appropriations and allocations for 2005 through 2007; app. III
shows the agencies' fuel treatment accomplishments.) In 2002, the
President announced the Healthy Forests Initiative (HFI), directing the
departments of Agriculture and of the Interior and the Council on
Environmental Quality to provide regulations to ensure more timely
decisions, increase efficiency, and improve results in reducing the risk
of catastrophic wildland fires.
In 2003, Congress passed HFRA to reduce wildland fire risk to communities,
municipal water supplies, and other at-risk federal lands through a
collaborative process of planning, setting priorities for, and
implementing fuel reduction projects. In funding authorized fuel reduction
projects on federal land, HFRA requires the agencies to use at least 50
percent of these funds in the wildland-urban interface. ^[88]11 The act
also established separate environmental analysis and administrative review
procedures for fuel reduction projects authorized under HFRA. In providing
assistance for fuel reduction activities on nonfederal lands, HFRA
requires the agencies, to the maximum extent practicable, to give priority
to communities that have adopted a community wildfire protection plan
(community plan) or have taken proactive measures to encourage willing
property owners to reduce fire risk on private property. A community plan
identifies and sets priorities for fuel reduction treatments and
recommends the types and methods of treatment on federal and nonfederal
land that will protect at-risk communities and essential infrastructure;
community plans also recommend measures to reduce structural ignitability
throughout the at-risk community. These plans are to be agreed upon by the
applicable local government, local fire department, and state forest
management agency, in consultation with other interested parties and the
federal land management agencies. As of February 2007, there were at least
1,100 completed community plans covering almost 3,300 communities
throughout the United States, and approximately 450 additional plans in
progress, according to the National Association of State Foresters. A
community plan may cover one or more communities, and some cover entire
counties.
According to the 10-Year Strategy Implementation Plan, the goal of the
fuel reduction program is to reduce the risk of wildland fire to
communities and the environment. However, some fuel treatments provide
other benefits in addition to this overall program goal; for example,
agency staff sometimes conduct prescribed burns to both reduce fuel and
enhance wildlife habitat, or conduct mechanical thinning projects before a
commercial timber sale. Similarly, in addition to the approximately $400
to $500 million appropriated for fuel reduction each year, funds from
other agency programs, such as wildlife management or timber, often are
used to conduct vegetation treatment projects that reduce fuels as a
secondary
benefit. In addition, the agencies sometimes receive partnership funding
from outside organizations, such as the Rocky Mountain Elk Foundation or
The Nature Conservancy, to conduct collaborative treatments. ^[89]12
^11This requirement applies only to projects conducted using HFRA
authorities. Agency officials told us they do not have reliable data on
the portion of their fuel reduction projects that used HFRA authorities.
GAO Has Reviewed Agencies' Fuel Treatment Programs
While the federal agencies acknowledge the importance of setting
priorities for lands needing fuel treatments, we have identified a
long-standing pattern of shortcomings in the processes the Forest Service
and Interior use to identify and set priorities for these lands. Between
1999 and 2007, we conducted several reviews of the agencies' wildland fire
management efforts, including the fuel reduction program. We found that,
while the agencies aimed to target fuel reduction efforts to the highest
risk areas, they could not ensure that they were doing so. For example, in
1999, we found that the Forest Service intended to give priority to
treatments in the wildland-urban interface but was hampered in doing so
because it had not fully defined and mapped such areas. ^[90]13 We
concluded that the Forest Service needed a cohesive strategy outlining
options and associated costs for reducing fuel. We reiterated the
agencies' need for a cohesive strategy in several additional reports and
testimonies issued between 2002 and 2007. ^[91]14 In 2000 and 2002, we
reported that the Forest Service and Interior did not know how many
communities were at high risk of severe wildland fire or their locations
and the cost to treat them and, therefore, could not set treatment
priorities. ^[92]15 We further reported in 2002 and 2003 that the agencies
did not have quantifiable long-term and annual performance measures to
assess progress in reducing the risks of wildland fire and that they
measured the performance of the fuel reduction program by number of acres
treated, which does not necessarily correlate to risk reduction. ^[93]16
Similarly, in 2004, we reported that the agencies did not systematically
assess the risks to environmental resources and ecosystems and,
therefore, could not target fuel reduction efforts to the resources and
ecosystems at highest risk. To set priorities for fuel reduction
activities, the agencies must first identify areas at risk from wildland
fire by considering three elements: hazard, risk, and value. ^[94]17 A
hazard is a potential event, such as a wildland fire, and the conditions
that cause it; in the case of wildland fire, both the fuel conditions and
the fire itself are the hazard. Risk is the probability that an event such
as a wildland fire will occur. Values are the resources and property that
could be lost or damaged because of a hazard; in the case of wildland
fire, values might include social, economic, or environmental values.
^[95]18 Without considering all three elements, the agencies may not be
appropriately setting priorities for areas needing fuel reduction. For
example, an area with high vegetation hazard may not be in an area where
fires are likely to occur, making it a lower priority for treatment;
likewise, a high hazard area might not be near something of value that
could be lost or damaged in a fire, also making it a lower priority for
treatment.
^12Our review was limited to fuel reduction work activities using federal
funds appropriated specifically for this purpose. As a result, fuel
reduction work funded by other agency programs or outside organizations is
beyond the scope of this review.
^13GAO/RCED-99-65.
^14GAO, Wildland Fire Management: Reducing the Threat of Wildland Fires
Requires Sustained and Coordinated Effort, GAO-02-843T (Washington, D.C.:
June 13, 2002); GAO05-147; GAO-06-671R; Wildland Fire Management: Lack of
a Cohesive Strategy Hinders Agencies' Cost-Containment Efforts,
GAO-07-427T (Washington, D.C.: Jan. 30, 2007).
^15GAO/T-RCED-00-296; GAO-02-259.
^16GAO-02-259; GAO-03-805.
We also found, through multiple reviews, that the agencies could benefit
from coordinating their efforts to manage wildland fires because wildland
fire is a shared problem that transcends administrative boundaries. For
example, in 2001 we reported that federal policy for managing wildland
fire required coordination, consistency, and agreement among the Forest
Service, Interior, and Interior agencies, but we found that the agencies
planned and managed wildland fire management activities largely on
agency-by-agency and unit-by-unit bases, and could not ensure, among other
things, that they were allocating funds to the highest-risk communities
and ecosystems. ^[96]19 In a 2002 report, we noted that the Forest Service
and Interior had either developed or were in the process of developing
numerous strategies that had different goals and objectives and that were
not linked, primarily because the agencies had been managing their lands
on an agency-by-agency basis for decades. ^[97]20 In a subsequent
testimony, which emphasized the agencies' need for a cohesive strategy as
well as clearly defined and effective leadership, we concluded that
effectively addressing wildland fire would require a sustained and
coordinated effort between departments. ^[98]21 (See Related GAO
Products.)
^17This approach, outlined by the National Academy of Public
Administration, uses risk as a specific term referring to the probability
of an event, as well as an umbrella term that encompasses all three of
these elements. See National Academy of Public Administration,
Managing Wildland Fire: Enhancing Capacity to Implement the Federal
Interagency Policy (Washington, D.C.: December 2001).
^18For more information on the hazard-risk-value framework, see
GAO-04-705.
^19GAO, The National Fire Plan: Federal Agencies Are Not Organized to
Effectively and Efficiently Implement the Plan, GAO-01-1022T (Washington,
D. C.: July 31, 2001).
^20GAO-02-259.
The Forest Service Uses a Mix of Quantitative and Judgmental Processes and
Considers a Range of Factors in Allocating Funds and Selecting Projects
The Forest Service uses both quantitative and judgmental processes in
deciding how to allocate fuel reduction funds. At headquarters, the agency
increasingly relies on a quantitative process--reflected in a computer
model--to determine the relative need for fuel reduction funds in each
region. At the regional level, some offices primarily use quantitative
processes to allocate resources while others rely on professional
judgment. Similarly, the national forests use a mix of quantitative and
judgmental processes to select projects.
At the National Level, the Forest Service's Allocation Process
Increasingly Relies on a Quantitative Approach
At the headquarters level, the Forest Service has developed a computer
model that assesses regions on various factors and assigns a score to each
region reflecting its relative priority for fuel reduction funds. ^[99]22
According to Forest Service officials, they developed the model to address
shortcomings that were highlighted by Congress and that were previously
identified by GAO, the Department of Agriculture's Office of Inspector
General, and the Office of Management and Budget. These shortcomings
included inadequate assessment of the risk of wildland fires to
communities, failure to clearly identify fuel reduction priorities, and
little
assurance that funding is targeted to these priorities. ^[100]23 In
addition, agency officials said they developed the model to provide
transparency, so that agency officials at all levels, as well as Congress
and others, can understand the rationale behind allocation decisions.
^21GAO-02-843T.
^22The Forest Service's system for setting priorities for reducing
hazardous fuels and allocating resources excluded the Alaska region (and
did not give it a priority score) because the Alaska region's program
accounts for less than 1 percent of the agency's fuel reduction funds.
In developing the model, the Forest Service brought together an
interdisciplinary group of senior leaders to determine the final list of
factors, which was based on an initial list developed by regional fuel
program managers. To determine the factor weightings, the group followed a
multistep process in which they determined the relative importance of each
factor by comparing it separately to every other factor, and then
synthesized the results to determine overall weightings. ^[101]24 The
model includes 18 weighted factors, as shown in table 1.
^23See GAO-03-805; U.S. Department of Agriculture, Office of Inspector
General, Audit Report: Implementation of the Healthy Forests Initiative,
08601-6-AT (Washington, D.C.: September 2006); Office of Management and
Budget, Program Assessment Rating Tool: Review of U.S. Department of
Agriculture's Wildland Fire Management Program
(Washington, D.C.: 2006); U.S. House of Representatives, Committee on
Appropriations,
Department of the Interior, Environment, and Related Agencies
Appropriations Bill of 2007, House Report 109-465, 109th Cong., 2nd Sess.
(Washington, D.C.: May 15, 2006); U.S. Senate, Committee on
Appropriations, Department of the Interior, Environment, and Related
Agencies Appropriation Bill of 2007, Senate Report 109-275, 109th Cong.,
2nd Sess. (Washington, D.C.: June 29, 2006).
^24The process the officials used to weight the factors is called the
analytical hierarchy process, which is a systematic process often used in
private industry to make complex decisions involving multiple criteria,
such as investment decisions. We did not assess the appropriateness of the
factors selected or the weights assigned, nor did we evaluate the model's
accuracy in applying these factors to determine priority scores.
Table 1: Factors Considered in Forest Service Fuel Reduction Funding
Allocation Model
Factors Weight (percent)
Treatment effectiveness^a 16.7
Wildfire potential 12.5
Wildland-urban interface
Treatment method availability
Wildlife habitat objectives
Municipal water supply
Ecosystem vulnerability^b
Associated benefits^a,c
Vegetative maintenance
Biomass opportunity
Insects and disease
Invasive species
Vegetation departure^b
Watershed condition
Life cycle cost^a,d
Commercial timber
Smoke emissions^e
Use of legislative tools^f
Total 100
Source: GAO analysis of Forest Service data.
Notes: Totals do not add to 100 percent due to rounding.
^aNo data were available for the 2007 allocation process.
^b"Ecosystem vulnerability" and "vegetation departure" are measures of
fire regime condition class.
^c"Associated benefits" is a measure of acres treated with fuel reduction
funds that achieve benefits for other programs, such as wildlife or
watershed.
^d"Life cycle cost" is intended to measure the cost of treatments per
year.
^e"Smoke emissions" is a measure of the acres of vegetation that produces
high levels of smoke during a wildland fire.
^f"Use of legislative tools" is a measure of acres treated in projects
authorized in HFRA or HFI, identified in community wildfire protection
plans, or implemented using stewardship contracts. Stewardship contracting
involves the use of any of several contracting authorities that were first
authorized for use by the Forest Service on a pilot basis in 1998, and
were subsequently extended to BLM. In practice, stewardship contracts
generally involve the exchange of goods, such as timber, for contract
services, such as thinning of brush.
Several of the factors--such as fire potential, ecosystem vulnerability,
and wildland-urban interface--are designed to assess the potential for
severe wildland fires in each region and the likelihood of damage
resulting from such fires. For example, to determine the potential for
severe fires, the model analyzes data such as the number and size of large
wildland fires in each region. To determine the likelihood of damage
resulting from wildland fires, the model includes data on values at risk,
such as the locations of municipal water supplies and wildland-urban
interface.
Other factors are intended to encourage efficiency and effectiveness
within the fuel reduction program and across multiple Forest Service
programs, such as the forest products or wildlife management program, to
take advantage of opportunities to achieve objectives in other programs.
Regarding effectiveness, the model included a factor intended to assess
effectiveness in the regions--method availability. However, in practice,
this factor used data on the total number of acres treated in each
region-- in effect rewarding regions for treating a large number of acres
regardless of how well the treatments reduced risk or of the risk level of
the areas treated. In addition, the model was designed to include a factor
to assess how effective individual fuel reduction treatments are likely to
be in reducing risk. However, the Forest Service does not currently have
data to make such an assessment; consequently, for 2007, this factor did
not influence allocations to the regions.
In 2007, officials used the model's results to inform their decisions
about funding allocations to the regions, although they relied mainly on
the prior year's funding levels along with their professional judgment.
Officials used the model's results only to make minor adjustments to
allocations because the model was still being refined and because they
wanted to phase in funding changes gradually in order to minimize
budget-related disruptions. Headquarters officials said they expect the
model to have a stronger influence on future allocation decisions.
The model assigned a numerical score to each region that indicated the
region's relative priority for fuel reduction funds, with higher scores
indicating higher priority. However, as shown in table 2 and figure 5, the
Forest Service's funding allocations to its regions are often at odds with
the priority scores resulting from the model, with some high-scoring
regions--such as the Northern and Eastern regions--receiving less funding
than some lower scoring regions such as the Pacific Southwest and the
Southwest regions.
Table 2: Forest Service Regions' Fiscal Year 2007 Fuel Reduction Priority
Scores and Funding Allocations
Dollars in thousands
Forest Service region Priority score Funding allocation
Southern 574 $29,092
Northern 455 15,782
Eastern 416 9,718
Intermountain 408 16,165
Rocky Mountain 399 25,445
Pacific Northwest 389 25,794
Pacific Southwest 388 43,737
Southwestern 367 37,341
Alaska^a a 805
Total $203,879
Source: Forest Service.
Notes: The Forest Service also allocated $2.265 million to its research
stations and $ 95.109 million to its headquarters office and to cost
pools, which are used for expenses that cannot reasonably be charged to a
single program, including indirect, support, and common services charges.
^aThe Forest Service excluded the Alaska region from its model because the
region has a small fuel reduction program relative to the other regions
and receives less than 1 percent of the agency's fuel reduction funds.
Figure 5: Forest Service Regions' Fuel Reduction Priority Scores as a
Percentage of Total, Compared to Regions' Funding Allocations as a
Percentage of Total Allocations, Fiscal Year 2007
Percentage
25
Region
Regional priority score as a percent of total Regional portion of national
allocations
Source: GAO analysis of Forest Service data.
According to Forest Service officials, the allocation amounts were not
more closely correlated with the priority scores for the following
reasons:
o As noted, the officials wanted to temper changes to regions' budget
allocations until they completed revisions to the model and developed more
confidence in its output in order to minimize funding shifts that might
prove inappropriate once the model is refined. ^[102]25 Agency officials
told us that even when they become confident in the model's output,
they will likely implement changes incrementally in order to minimize
disruption to regions and national forests.
^25GAO has previously noted the importance of the integrity, credibility,
and quality of data that underlie budget decisions and, thus, the value in
implementing changes gradually when the quality of such data is in
question. See, for example, GAO, Performance Budgeting: Opportunities and
Challenges, GAO-02-1106T (Washington, D.C.: Sept. 19, 2002).
o Until the revisions have been completed, the model's results will be
tentative. An important focus of the revisions will be those 3 of the
model's 18 factors for which the Forest Service had no data sources in
2007. Because of the lack of data, these elements had no effect on the
regions' 2007 priority scores, ^[103]26 but agency officials hope to
have data to inform these elements for future allocations.
o The relatively high priority score assigned to the Eastern region was
not consistent with agency officials' knowledge of the area--that is,
they believed that, relative to the other regions, there were fewer
destructive wildland fires in the Eastern region and, consequently,
they expected the region's priority score to be lower than it was.
When the officials consulted data on the number of structures burned
in wildland fires, their belief was confirmed. Consequently, agency
officials are reexamining the measures they used to assess risk and
exploring options for refining them.
o Fuel reduction costs vary widely from region to region, and when
making final allocations, the officials made adjustments to
accommodate this variation. For example, the Pacific Southwest region
received the largest allocation of any region, despite its relatively
low priority score, in part because treatment costs in the region are
very high (averaging about $535 per acre in 2006). Therefore, a
relatively large allocation is needed to fund even a moderate amount
of work. At the other end of the spectrum, treatment costs in the
Southern region are low (averaging about $32 per acre in 2006),
meaning that needed work can be accomplished with a smaller
allocation. In addition, agency officials said that, although the
Southern region's priority score might point toward a larger
allocation for the region, nonmonetary constraints--such as the size
of the workforce--limit the amount of work the region can accomplish
and, therefore, the amount of funds that can prudently be invested
there. Further, in order to maintain overall funding stability to the
regions, officials coordinated regional fuel reduction funding
allocations with those of other Forest Service resource programs, such
as watershed management or forest products. This coordination
sometimes resulted in officials adjusting fuel reduction funding
allocations in order to compensate for adjustments in these other
programs' funding levels.
^26To ensure that the elements without data had no effect on priority
scores, all of the regions were assigned the same score for each of these
elements.
o The Forest Service allocated a portion of its fuel reduction funds
according to congressional direction. Specifically, congressional
committee reports accompanying relevant appropriations acts directed the
Forest Service to spend about $34 million (12 percent) of its 2006 fuel
reduction funds in certain areas or on certain projects. The Forest
Service accommodated this congressional direction, regardless of whether
doing so was consistent with priority scores.
Some Forest Service Regions Use Quantitative Allocation Processes, While
Others Rely More on Professional Judgment
For 2007, the Forest Service allowed each region to determine how to
allocate funds to its national forests and what factors to consider in the
process, as long as the factors were consistent with those considered in
the national allocation process. Four of the Service's nine regions relied
primarily on quantitative data in their allocation processes to national
forests, while five relied primarily on professional judgment. In applying
these processes, all nine regions considered a combination of factors,
many of which were similar to those used at the national level. Beginning
with the 2008 allocations, the Forest Service plans to require regions to
use the headquarters model to inform allocation decisions.
Of the four regions that relied primarily on quantitative processes in
2007, one--the Rocky Mountain region--used a computer model that analyzed
geospatial data on vegetative condition and areas of insect-killed trees
to help assess relative wildland fire risk among the national forests in
the region. Regional officials then used their judgment to consider other
factors, such as lands in the wildland-urban interface and acreage
targets, to refine allocation amounts. Through the risk assessment
process, the region identified 6 emphasis forests out of the 11 forests in
the region and allocated over 70 percent of the region's fuel reduction
funds to these 6 forests. The Pacific Northwest region also used a model,
but its model incorporated regional data on a number of factors, including
the number of acres in fire regimes I, II, and III; the number of acres
identified in community plans as being in the wildland-urban interface;
and per-acre treatment costs. Using these data, regional officials
identified five forests in the region where an extremely wet climate made
the risk of damaging wildland fires so low that they decided not to
allocate any fuel reduction funds to these forests and excluded them from
the model. Another region that relied on a quantitative process--the
Pacific Southwest region--used a scoring system that ranked forests
primarily on the basis of a risk assessment; the assessment incorporated
multiple factors, such as the number of acres in condition classes 2 and 3
and in the wildland-urban interface. The region also used other factors,
such as the forests' capacity to conduct fuel treatment work, to make
smaller adjustments. The Intermountain region allocated about 80 percent
of its fuel reduction funds in accordance with forests' historical funding
levels. For the remaining funds, the region delegated priority decisions
to collaborative interagency groups in each of the region's states. These
groups scored and ranked proposed projects against a set of standard
criteria and made funding recommendations to the regional office.
The remaining five Forest Service regions relied primarily on professional
judgment and negotiation among agency officials when determining funding
allocations to national forests. Although these regions did not use
quantitative processes to assign priorities among forests, they
incorporated some of the same information included in other regions'
quantitative processes. For example, the Northern region conducted a risk
assessment for the region, but instead of using the risk assessment to
guide its allocations to the forests, the region directed forests to use
it to identify potential treatments. The region then allocated funds to
forests primarily on the basis of the forests' proposed annual workloads.
In the Southern region, officials used their professional judgment to
decide on allocations largely on the basis of forests' reported
capabilities, per-acre treatment costs, and local priorities, and how they
fit with expected regional targets and budgets.
Factors outside of the formal process influenced allocations, according to
Forest Service officials, but they did not always formally incorporate
these factors into the allocation process. For example, in several
regions, fuel reduction officials said they coordinated with officials
from other resource programs, such as the wildlife management and
vegetation management programs, when deciding on final allocations. In
doing so, they sometimes adjusted fuel reduction allocations to, for
example, prevent multiple programs from reducing allocations to a given
forest in the same year or to take advantage of efficiencies when
different programs' priorities overlapped in a given forest. In addition,
nearly every region reported considering acreage targets when making
allocation decisions--even those that did not report it as an official
part of their allocation processes. Regional officials noted that pressure
to meet the acreage targets established by Forest Service headquarters
sometimes trumped all other factors in allocation decisions, especially in
2007 when targets increased at a faster rate than funding levels. Another
factor, according to agency officials in several locations, was direction
contained in congressional committee reports accompanying relevant
appropriations acts that a certain amount of funding be allocated to
specific forests or specific districts within forests. As with
headquarters, regional offices allocated funds according to this
direction, apart from any priority-setting process. For example, in 2006,
the Pacific Southwest region allocated nearly $21 million (about 52
percent of the region's budget) on the basis of congressional committee
report direction, in part to treat areas of insect-killed trees in
southern California. Finally, regions reported shifting funds among
forests, after initial allocation decisions had been made, to accommodate
unexpected circumstances during the year, such as large wildland fires
that prevented fuel reduction treatments from being implemented as
planned.
National Forests Select Projects Using Quantitative and Judgmental
Processes
Like regional offices, national forests are allowed to determine what
processes to use and which factors to consider in selecting fuel reduction
projects to fund and implement, within the parameters of national and
regional direction. In practice, some forests rely more on quantitative,
data-driven processes, while others rely more on professional judgment.
Both consider a mix of factors, as the following examples, based on our
site visits to national forests, illustrate:
o Quantitatively based selection. The Arapaho-Roosevelt National Forest
in Colorado collaborated with another national forest, a national
park, Forest Service research scientists, and the Colorado State
Forest Service to develop a risk assessment that used quantitative
data to map the highest priority locations for fuel reduction
treatments in the area. Forest officials then used the risk assessment
to prepare a 10-year strategy with proposed annual treatments. Each
year, forest officials first consult the strategy and the risk
assessment to identify a list of projects to fund, and then adjust the
list to meet acreage targets within budget constraints. Similarly,
officials of the Angeles National Forest in Southern California
convened a diverse group of stakeholders and followed a step-by-step
process to identify priorities for fuel reduction treatments. During
the process, Forest Service officials provided information, such as
the locations of historical wildland fires and developed areas, as
well as places where fuel reduction was not feasible because, for
example, the topography was too steep to operate needed equipment.
They then used fire behavior models to show where fires could
potentially burn and how various proposed fuel reduction treatments
might affect such fires. The end result was a multiyear list of
proposed projects that forest officials used to select projects each
year.
o Judgmental based selection. At the Medicine Bow-Routt National Forest
in Wyoming and the Chattahoochee-Oconee National Forest in Georgia,
officials relied largely on their knowledge and experience about the
area to select fuel reduction projects. Some of these officials had
worked at the same forest for decades. At the Ocala National Forest in
Florida, officials use their professional judgment to select projects,
which are almost all prescribed burns. However, because the forest's
fuel reduction program is so large and the vegetation grows so
quickly, the project selection process is founded on a rotational
schedule. Under this schedule, the forest aims to treat nearly all of
its approximately 130,000 burnable acres over a 4-year period.
Consequently, officials try to treat about a quarter of the
acreage--or slightly over 30,000 acres--each year. Forest officials
also said they consider other factors, such as wind direction,
humidity, and human activity (for example, popular areas for weekend
recreation), when determining the specific timing of a prescribed
burn.
In addition to factors that national forests considered, unanticipated
factors influenced project selection decisions at the local level,
sometimes preventing planned projects from being implemented. In such
cases, agency staff frequently carried out lower priority projects in
place of the originally planned projects. For example, wildland fires
sometimes burned in locations planned for fuel reduction treatments,
making the treatments unnecessary; in other cases, litigation prevented
planned treatments from being implemented as scheduled.
Interior and Its Agencies Use a Mix of Quantitative and Judgmental Processes and
Consider a Range of Factors in Allocating Funds and Selecting Projects
Interior and its agencies--BLM, BIA, FWS, and NPS--use both quantitative
and judgmental processes for allocating fuel reduction funds and selecting
projects, and consider multiple factors, many of which are similar to
those used by the Forest Service. In 2007, Interior allocated funds to its
four agencies primarily on the basis of historical funding levels, but it
is currently developing a computer model similar to the Forest Service's.
Like Interior, the BLM national office allocated funds to its state
offices primarily on the basis of historical funding levels in 2007 but is
expecting to implement a new funding allocation model in 2008. The
majority of BLM state and local offices allocated funds and selected
projects using quantitative processes, many of which use scoring systems.
The other three Interior agencies' national, state, and local offices used
both quantitative and judgmental processes to allocate funds and select
projects, considering a range of factors.
Interior Allocates Funds to Its Agencies Primarily on the Basis of
Historical Funding Levels
Interior's allocations to BLM, BIA, FWS, and NPS have remained fairly
constant from year to year, measured on a percentage basis, because the
department primarily allocates fuel reduction funds on the basis of past
funding levels--what one departmental official called "allocation by
tradition." This funding pattern dates back to 2001, when the Interior
agencies began receiving a sharply increased amount of fuel reduction
funds as a result of the National Fire Plan. Since then, the percentage of
Interior's fuel reduction funding that is allocated to each of the
agencies has remained consistent, with BLM receiving about 50 percent of
the funding, BIA receiving about 20 percent, and FWS and NPS each
receiving about 15 percent. Figure 6 shows the percentage of Interior's
total fuel reduction appropriation that was distributed to each agency
from 2001 through 2007.
Figure 6: Percentage of Interior's Total Fuel Reduction Funds Allocated to
the Interior Agencies, Fiscal Years 2001 through 2007
Percentage 100
80
60
40
20
0
2001 2002 2003 2004 2005 2006 2007 Fiscal year
BLM
BIA
NPS
FWS
Source: GAO analysis of Interior data.
In 2001, Interior established initial funding allocations on the basis of
estimates of each agency's infrastructure and capacity (i.e., the amount
of work each could accomplish), which it determined by compiling field
requests from the four agencies. However, at the time, most of the
agencies and their field units had little infrastructure related to the
fuel reduction program--including limited staff--so many units did not
have the resources to collect extensive information on fuel reduction
needs, according to agency officials. As a result, agency officials had to
make allocation decisions based on limited information. Some agency and
departmental officials have stated that the allocations need to be
revisited now that the fuel reduction program has been in place for
several years.
Each year, the department tells the four agencies how much funding the
department has requested for the fuel reduction program and what its
acreage targets are for treatments within and outside of the
wildland-urban interface. The agencies' fuel program leads--the
headquarters officials in charge of each agency's fuel reduction
program--then meet to determine how to divide the funds and set targets
for each agency. However, the fuel program leads do not have the authority
to significantly adjust the funding allocations from previous levels;
rather, such changes would have to be determined at the department level,
according to headquarters officials. The agencies' field units submit
proposed project lists to the regions, which review these lists before
forwarding them to headquarters; these lists provide the fuel program
leads with an idea of each agency's needs and capabilities when
determining funding allocations. After the fuel program leads decide upon
initial allocations, they may shuffle funds within or between their
agencies throughout the year to adapt to uncontrollable circumstances,
such as weather conditions. The majority of fuel treatments that the
Interior agencies conduct depend on the weather, and sometimes weather
conditions prevent work from being completed. For example, if a drought in
the Southeast makes vegetation too dry for safe prescribed burns, Interior
may shift funds to units in the western United States. In practice, these
considerations may result in Interior's shifting funds from FWS and NPS,
which conduct a large number of fuel treatments in the Southeast, to BLM
or BIA, which conduct most of their fuel treatments in the West.
In 2007, Interior allocated 5 percent of its funds to the agencies using a
model similar to the Forest Service model, and it plans to use the model
to influence a greater portion of allocations in future years. ^[104]27
The department developed the model after Interior and the Forest Service
received congressional committee direction in 2005 to develop a common
method for setting project priorities. Interior's 2007 model included a
range of factors, such as the amount of land each agency manages with
certain fuel conditions and the degree to which each agency used biomass,
but included fewer factors than the Forest Service's model because some
data were not yet available. Because Interior does not currently have a
good method for measuring efficiency or effectiveness, its 2007 model used
the legislative tools factor, which measures the extent of use of HFI and
HFRA planning authorities, to measure efficiency, and the number of acres
treated to measure effectiveness. The following provides the complete list
of factors used in Interior's 2007 model: ^[105]28
o number of fire starts,
o number of large fires (defined as 500 acres or more),
o fuel conditions,
o biomass utilization,
o number of threatened and endangered species,
o fire regime condition class improvement,
o use of legislative tools (HFI/HFRA),
o number of acres treated, and
o wildland-urban interface.
^27The funds Interior allocated using the model represented 5 percent of
project funds--that is, funds expected to be spent on individual
projects--rather than 5 percent of the total allocation, which would
include program management expenses such as salaries, facility costs, and
so forth. NPS did not receive any of the 5 percent of 2007 funds that were
allocated using the new model because Interior allocated the funds late in
the fiscal year and NPS had already met its 2007 acreage targets.
^28Weights are not included in the list because Interior had not yet
finalized them at the time of our review.
Results from the 2007 model were generally consistent with Interior's
allocations to the agencies in previous years. However, Interior is still
making changes to the model, including determining how to weight the
factors, so this may not be the case in future years. According to
departmental officials, Interior intends to be cautious in applying the
new model and making significant changes to current allocations because
Interior and the Forest Service are currently developing the Fire Program
Analysis (FPA) system--an interagency fire management planning and
budgeting model--and they expect information from that system to inform
future allocation decisions. ^[106]29 By proceeding slowly, the department
hopes to avoid potentially disruptive fluctuations in regional and field
unit allocations.
Once they have received their allocations from the department, Interior
agencies determine how to allocate fuel reduction funds to the regions
within the parameters of departmental and congressional direction.
Interior officials have stated that they would like the agencies to use
more rigorous allocation processes in the future, though one departmental
official noted that he does not want the agencies to invest substantial
funding or time and effort to develop new allocation processes pending the
expected completion of the FPA. Interior guidance lists the following
priorities for selecting projects:
o All projects must result from a collaborative process.
o Funding will be targeted to the wildland-urban interface.
o Within the wildland-urban interface, focus should be on projects near
wildland-urban interface communities at greatest risk of fire;
communities that have completed a community plan or its equivalent;
and communities where there is an active partnership with volunteer
efforts, in-kind services, or partners who contribute funding.
o Outside of the wildland-urban interface, focus should be on areas in
condition class 2 or 3 in fire regimes I, II, or III, or those in
condition class 1 where landscape conditions could quickly deteriorate
to condition 2 or 3.
^29While agency officials told us the new model will be coordinated with
FPA, they did not provide details on how this coordination will occur.
o Priority should also be given to projects using mechanical treatments,
with special emphasis on projects yielding biomass that can be sold or
traded to companies or the local community; and projects using
contractors, particularly those projects conducted under contracts
that support rural communities' stability.
o Prescribed burning is to be used when weather and resource conditions
permit, where mechanical treatments are not appropriate, and as
maintenance treatments following mechanical work.
o Managers must make maximum practical use of tools provided by HFRA and
HFI.
BLM Increasingly Uses Quantitative Processes in Allocating Funds and
Selecting Projects
In 2007, BLM headquarters allocated funds to its state offices primarily
on the basis of historical funding levels; however, agency officials told
us that, starting in 2008, BLM plans to use a quantitative process
incorporating factors similar to those used in Interior's new model, with
a greater emphasis on collaboration and local priorities. BLM headquarters
provides flexibility to state offices and local units when allocating
funds and selecting projects but directs these offices to consider
Interior and agency guidance. The majority of BLM state offices and local
units used quantitative processes to allocate funds and select projects in
2007, frequently scoring projects against a set of weighted factors.
BLM Allocates Funds to Its State Offices Primarily on the Basis of
Historical Funding Levels but Plans to Use a More Quantitative Approach in
2008
In 2007, BLM headquarters allocated funds to its state offices largely on
the basis of past funding levels--as in previous years--as a way to ensure
that funding levels remain relatively stable, but it also considered
proposed projects, national priorities, and the extent to which state
offices met past acreage targets established by BLM. While the project
lists do not largely influence allocations to state offices, state offices
use these lists to allocate funds to field units, and field units use them
to select projects for implementation. Table 3 shows the 2007 allocations
to the BLM state offices. (App. II also shows 2005 and 2006 allocations.)
Table 3: BLM Allocations to State Offices, Fiscal Year 2007
Percent of BLM total
State office Allocation state office allocation
Oregon/Washington $24,878,000 27.1
Idaho 14,598,000 15.9
Utah 10,078,000 11.0
California 7,322,000 8.0
Colorado 6,843,000 7.5
Nevada 6,414,000 7.0
New Mexico 6,412,000 7.0
Montana 5,461,000 6.0
Arizona 4,355,000 4.7
Wyoming 3,684,000 4.0
Alaska 1,556,000 1.7
Eastern States 126,000 0.1
Total 91,727,000a 100.0
Source: GAO analysis of BLM data.
Notes: Total allocation includes the allocation for the current year plus
carryover from the previous fiscal year.
^aBLM allocated an additional $8,473,000 for BLM headquarters, science
centers, training costs, and other support costs.
As shown in table 3, the Oregon/Washington, Idaho, and Utah state offices
got substantially more funding than the other states--more than half of
BLM's total funding. The Oregon/Washington state office alone received
more than $24 million--27 percent of BLM's state office funding; one BLM
field unit in Oregon, the Medford district office, received over $9
million in 2007--more than nine state offices each received in total
funding. According to some agency officials, the relatively high level of
fuel reduction funding directed toward the Oregon/Washington state and
Medford district offices is, in part, the result of BLM's emphasis on
providing stable levels of funding to states and field units. According to
these officials, when BLM (along with other federal agencies) received a
sharp increase in fuel reduction funding in 2001, agency officials sought
to identify units that could implement fuel reduction projects quickly.
Because the Oregon/Washington and Medford offices were identified as
having the capacity to undertake a large number of fuel reduction
projects, they received a substantial portion of the new funding. However,
another agency official told us these large amounts are justified because
there is substantial wildland fire risk in Oregon and, therefore, a great
need for fuel treatments because vegetation grows very quickly in the
western part of the state, there is considerable wildland-urban interface,
and wildland fire suppression costs are high.
Starting in 2008, BLM plans to use a model to influence funding
allocations to state offices for fuel reduction. Use of the model is
intended to ensure that the highest priority work is funded and that BLM's
fuel reduction treatments are integrated with other vegetation treatments,
such as range improvement projects, to effectively achieve fire and
resource management goals and objectives. According to a headquarters
official, the new model is intended to facilitate comparison of risk and
needed work at the national and state levels in order to set priorities
for funding among states and communities. Headquarters officials will use
the model results to make allocation decisions but will shift no more than
20 percent of the previous year's allocations to each state in 2008 and
2009.
The model has three components: (1) treatment characteristics; (2) a
measure of the degree of threat; and (3) an efficiency measure. For the
first component--treatment characteristics--the model will score every
proposed project on a set of weighted factors, such as local priority
ratings, the availability of joint funding, and condition class; there are
separate factors and weights for projects within and outside of the
wildland-urban interface. The second component--the measure of the degree
of threat--currently combines three elements: the number of fire starts,
the number of large fires (i.e., fires greater than 300 acres), and local
risk ratings. The third component--efficiency--is currently measured by
past performance on acreage targets, past performance on estimating
treatment costs, and treatment cost per acre. According to agency
officials, they intend to eventually include a measure of effectiveness in
the model, which would indicate how well a treatment reduces risk or
achieves other objectives. However, because BLM does not currently have a
good way to measure effectiveness, it is using measures of efficiency
until it develops a better approach. Table 4 shows the complete list of
factors used in the model and their weights.
Table 4: Factors and Factor Categories BLM Considers in BLM Fuel Reduction
Funding Allocation Model
Weights for
wildland-urban Weights for treatments
Funding allocation interface outside the Overall
wildland-
model components Factors evaluated treatments urban weight
interface
Community plan or 0.14 0.01
equivalent
High local 0.14 0.11
prioritya
Mechanical 0.12 0.04
treatment
Joint funding 0.10 0.08
available
HFRA/HFI NEPA 0.10 0.08
typeb
Stewardship 0.10 0.08
projectc
Multiple land 0.08 0.03
ownership
Treatment Moderate local 0.08 0.06 0.45
prioritya
characteristics Biomass utilizedd 0.05 0.04
Large-scale 0.05 0.10
treatmente
Low local 0.02 0.02
prioritya
Condition class 2 0.01 0.11
or 3
Impacted species 0.01 0.10
Fire regime I, f 0.06
II, or III
Fire or other f 0.08
treatment methodg
Number of large
fires (greater 0.50 0.50
than 300 acres)h
Degree of threat Number of fire 0.25 0.25 0.35
startsh
Local risk 0.25 0.25
ratingi
Past performance
on acreage 0.50 0.50
targetsj
Past performance
Efficiency on treatment cost 0.40 0.40 0.20
estimatesk
Cost per acre 0.10 0.10
Source: GAO analysis of BLM data.
^aThe local priority rating is assessed at the local level and is a way
for the field to communicate project priorities that may not be
well-represented by other factors.
^bThe "HFRA/HFI NEPA-type" factor weights projects that use National
Environmental Policy Act (NEPA) planning tools authorized by HFRA or HFI.
^cStewardship projects are accomplished through the use of stewardship
contracting, which involves the use of any of several contracting
authorities that were first authorized for use by the Forest Service on a
pilot basis in 1998, and were subsequently extended to BLM. In practice,
stewardship contracts generally involve the exchange of goods, such as
timber, for contract services, such as thinning of brush.
^dThe "biomass utilized" factor weights projects that make use of
biomass--small-diameter trees, branches, and other organic
material--removed through fuel reduction.
^eLarge-scale treatments are treatments that are at least 150 percent
larger than the average treatment.
^fThis factor was not used to determine the treatment scores for
wildland-urban interface treatments.
^gThis factor weights projects treated with prescribed fire or other
treatment methods, such as grazing or herbicides.
^hThe "number of large fires" and "number of fire starts" factors are
determined at the field office or district level and applied to all
treatments within that field office or district.
^iThe risk rating is assessed at the local level and is to be determined
from community plans or risk assessment programs.
^jThe "past performance on acreage targets" factor is calculated at the
state level and applied to all treatments within the state.
^kThe "past performance on treatment cost estimates" factor is calculated
at the state level and applied to all treatments within the state.
The BLM national office also directs state offices and local units to
consider Interior and BLM priorities when allocating funds and selecting
projects. BLM-specific guidance directs state offices and local units to
coordinate fuel treatments with other resource management activities, such
as timber and wildlife habitat; target funds to wildland-urban interface
areas identified through a collaborative process; target nonwildland-urban
interface funds to ecosystems that have the highest risk-reduction
potential; and use HFI and HFRA planning tools.
The Majority of BLM State Offices Incorporate Quantitative Approaches in
Their Allocation Processes
The BLM national office allows state offices to choose the approach they
use in allocating funding to field units, as long as they take into
account departmental and BLM priorities, and state offices will continue
to have this flexibility with the implementation of the new national
allocation process, according to headquarters officials. In 2007, 6 of the
11 BLM state offices primarily used quantitative approaches to inform
their allocation processes, and 5 primarily used a judgmental approach.
^[107]30 Nine of 11 state offices considered targets or past performance,
and 10 considered at least one factor related to collaboration, such as
community plans. Eight of 11 state offices considered at least one factor
to estimate wildland fire risk, such as local- or state-level risk
assessments or fire regime condition class.
The six state offices that allocated funds using quantitative processes in
2007 primarily used weighted scoring systems--similar to the state scoring
component of BLM's new model--to set priorities for projects. While the
specific factors and their weights varied by state, many factors were
commonly used and were similar to those used in BLM's headquarters system;
each of the states had separate lists of factors for projects within and
outside of the wildland-urban interface. For wildland-urban interface
projects, five of the six state offices emphasized factors such as local
risk ratings or community hazard assessments to estimate risk from
wildland fire, and all six offices considered a variety of other factors,
including community support and joint funding, to measure the extent of
collaboration. For projects outside of the wildland-urban interface, all
six offices gave priority to projects in condition classes 2 or 3, jointly
funded or collaborative projects, and projects that improved threatened
and endangered species habitat, as well as a variety of other factors.
Once the state offices had the field offices' project lists, state and
field offices generally negotiated to determine final funding allocations.
^30While there are 12 BLM state offices--11 in the West and 1 in the
East--the vast majority of BLM-managed land is in the West, and the
Eastern States office receives only about
0.1 percent of BLM's total fuel reduction funding. Further, this funding
is allocated to just one field unit. As a result, the Eastern States
office is not included in our description of BLM state office allocation
processes.
The remaining five state offices primarily used judgmental processes to
allocate funding to field units. For example, in 2007, the Oregon/
Washington state office allocated funding using professional judgment and
negotiation, which included numerous discussions with field units' fuel
program staff to assess the units' priorities and capabilities. The state
office primarily considered capability and past performance of field
offices and BLM's national priorities when making the final allocations.
Starting in 2008, the office plans to use a model to allocate base funding
for fuel reduction, which covers salaries and other fixed costs, but will
continue to allocate project funding using the current approach, which
relies primarily on professional judgment and negotiation.
BLM officials told us that factors outside of the formal process
influenced allocations. For example, in several states, agency officials
said they coordinated with officials from other resource programs, such as
the range or weeds programs, at the state or local level when deciding on
final allocations or selecting projects. As a result, they sometimes
selected projects that used funding from multiple resource areas, or
benefited these areas, over other projects in order to take advantage of
efficiencies. Many state offices also reported that they considered
acreage targets when making allocation decisions. According to one state
official, acreage targets were the most influential factor in allocation
decisions, and several agency officials said that lower priority projects
were sometimes funded to meet acreage targets. Finally, state offices
reported shuffling funds among or within field units after allocation
decisions had been made to accommodate uncontrollable circumstances
throughout the year, such as weather conditions that prevented prescribed
burns from being implemented as planned.
BLM state offices also devote substantial effort and funding to assist in
the development of community plans, and allocate a significant portion of
fuel reduction funding to projects on private land. For example, the
Montana state office funded the development of 49 out of 54 completed
community plans throughout the state, according to an agency official.
Also, when the Montana state office and its field units allocate funding
to, and select projects in, the wildland-urban interface, proposed
projects on private land--which are submitted to BLM by counties--are
ranked using the same system as BLM projects. Consequently, BLM projects
on federal land essentially compete for the same funding as projects on
private land. In California, the BLM state office allocates more than half
of its wildland-urban interface funding to a community assistance program,
through which fuel treatments on private, state, or tribal lands adjacent
to or in the vicinity of federal lands are funded through an interagency
grant process.
The Majority of BLM Field Units Incorporate Quantitative Approaches into
Their Project Selection Processes
As with the BLM state offices, in 2007, the majority of BLM field units
used quantitative approaches that incorporated a range of factors--many of
which were similar or identical to the ones used by state offices--to
select and rank projects. For example, the Twin Falls district office in
Idaho scored all projects using a weighted scoring system developed by the
BLM Idaho state office. It then ranked the projects, considering factors
such as project scores and areas identified in community plans. The
Billings field office in Montana also used a weighted scoring system to
rank projects. Field staff initially identified projects using community
plans or the field office's risk assessment--which analyzed fuel type,
fire regime condition class, and fire occurrence to identify high-risk
fire areas--and then scored the projects using the Montana state office's
weighted scoring system to identify high-, medium-, and low-priority
projects.
Also, like national forests, BLM field units were sometimes influenced by
unanticipated factors when selecting projects. For example, agency
officials sometimes deferred planned projects because newly proposed
projects suddenly became a high priority. They pointed to situations in
which nonprofit organizations donated funds to pay for projects and agency
officials gave those projects a higher priority. In Colorado, recent oil
and gas development, as well as construction of new subdivisions in
high-risk areas, have caused field units to shift priorities to conduct
treatments near these developments, according to a BLM official.
BIA Allocates Funds Largely on the Basis of Units' Performance History,
while FWS and NPS Use Quantitative and Judgmental Processes
In 2007, the three remaining Interior agencies--BIA, FWS, and NPS--
allocated fuel reduction funds using quantitative and judgmental processes
and considering a variety of factors. Like BLM, these agencies provide
flexibility to regional offices and local units in determining how to
allocate funds and select projects and direct them to consider
departmental priorities. (See app. II for these agencies' 2005 through
2007 allocations to their regional offices.)
BIA and FWS Headquarters Allocate Funds Using Quantitative Processes,
While NPS Headquarters Allocates Funds Primarily on the Basis of
Historical Funding Levels
In 2007, BIA headquarters allocated fuel reduction funds to its regions
using a formula that considered past performance and proposed work and
that essentially rewarded regions for their accomplishments. The formula
allocated to each region a percentage of the region's total budget
request, based on the percentage of the prior 3 years' acreage targets
that the region met. For example, if a region had met 95 percent of its
total acreage target since 2004, the region would receive about 95 percent
of its requested budget for 2007. BIA placed a cap on the amount of
funding that regions could request, based on their previous year's
accomplishments. ^[108]31 According to a headquarters official, BIA
rewards those regions and units that achieve acreage targets because, in
many instances, units do not meet targets.
FWS headquarters allocated 2007 funds to regional offices using a
quantitative model that considers multiple factors, including historical
fire occurrence, fuel conditions, community assessments of risk, and field
unit past performance. The model has separate modules for projects within
and outside of the wildland-urban interface, and produces a weighted score
for each FWS field unit. In the wildland-urban interface module, the most
influential factors are communities at risk, local hazard rankings, and
fire conditions. For the non-wildland-urban interface module, the most
influential factors are past performance and proposed work.
NPS headquarters allocated 2007 funding to regional offices primarily on
the basis of historical funding levels. These levels were originally set
by a model that determined funding allocations through a risk assessment,
which considered vegetation types, fuel types, fire return intervals, and
other data, and through an effectiveness measure that examined treatment
success for different vegetation types. According to an NPS official, the
agency maintained funding proportions at the model's 2005 level after
Interior directed it to work on the FPA; NPS decided that it would have
been too much work for field staff to maintain the model while also
preparing data for the FPA. Furthermore, they believed that the model,
initially developed more than 20 years ago, was outdated and did not merit
additional financial investment while the FPA was being developed.
^31Regions that accomplished 90 percent or more of the previous year's
acreage target could request up to 120 percent of the prior year's funding
amount, while regions that accomplished less than 90 percent could request
only up to 105 percent of the previous year's amount.
BIA, FWS, and NPS Regional Offices Allocate Funds to Field Units Using
Quantitative and Judgmental Processes
BIA, FWS, and NPS allow their regions the flexibility to determine how to
allocate funding to field units, provided the processes and factors are
consistent with departmental and agency guidance. The BIA national office
encourages regions to adopt allocation strategies similar to the one used
at headquarters--which rewards past performance--and some of BIA's
regional offices have done so, such as the Rocky Mountain and Northwest
regions. The Rocky Mountain region, for example, used a quantitative
process to allocate fuel reduction funds, using an allocation formula
similar to the one used by BIA headquarters but using only the previous
year's accomplishment rate, rather than the 3-year average headquarters
used. Likewise, one FWS regional office that we visited used a
quantitative process to allocate fuel reduction funds in 2007: FWS's
Mountain-Prairie region allocated funds to local units using FWS's
national model, but regional officials adjusted the model's allocations on
the basis of their knowledge about local factors, such as community
support for projects and field unit staffing levels.
Other BIA and FWS regional offices, and all of the NPS regional offices
that we visited, allocated fuel reduction funds in 2007 using judgmental
processes that incorporated a range of factors. For example, BIA's
Southwest region allocated funds primarily on the basis of project
rankings (as determined at the local level) and cost efficiency, according
to a regional official. Likewise, FWS's Southeast region allocated 2007
funds according to a regional official's assessment of a variety of
factors, such as field units' programs of work and wildland fire activity;
this official has many years of experience managing the region's fuel
reduction program. In NPS's Pacific West region, a group of local and
regional fire and fuel program staff determined funding allocations on the
basis of park priorities, past performance, and conformance with NPS
policy, balanced against regional funding levels and acreage targets.
As in other agencies, officials told us that factors outside of the formal
process also influenced allocations. Several BIA and NPS officials told us
that staffing constraints at field units may affect allocations. For
example, many park units have very small fuel treatment programs and no
staff dedicated solely to the program; therefore, the fuel reduction
programs at
such units may be eliminated if staff, who have numerous collateral
duties, no longer have the time to plan or implement treatments.
Furthermore, the location of some field units makes it difficult to
recruit and retain qualified staff; the field units are located either in
areas with high costs of living or in remote areas. Without dedicated
staff to manage fuel reduction programs at such field units, their
capacity to plan and implement projects and spend any funding allocation
is limited, so capacity becomes the determining factor regardless of other
factors considered in the allocation process, according to agency
officials. Some BIA officials told us that self-determination limits BIA's
influence over the tribes; self-determination provides tribes with the
authority to manage federal programs when they choose to do so, as well as
the authority to choose not to emphasize a given program. In some regions,
acreage targets also affected allocation and project selection processes,
and one agency official told us that projects were sometimes developed and
implemented specifically to meet targets. However, other BIA, FWS, and NPS
regional officials told us that they did not assign acreage targets to
field units or that there was little pressure to meet targets. Finally,
regions reported shifting funds among field units after allocation
decisions had been made to adapt to uncontrollable circumstances, such as
weather conditions that prevented planned projects from being implemented.
Local BIA, FWS, and NPS Units Select Projects Using Quantitative and
Judgmental Processes
Some BIA, FWS, and NPS local units selected projects in 2007 using
quantitative processes. For example, in NPS's Sequoia and Kings Canyon
National Parks in California, agency officials identified projects using a
model that determined high-risk areas on the basis of several factors,
such as the risk of a fire starting and the location of the wildland-urban
interface. Park officials used the model information, as well as
additional factors, such as values at risk, sequencing of treatments, and
project accessibility, to select projects. BIA's Zuni Agency in New Mexico
also used a quantitative process to select projects. The fuels specialist
analyzed geographic information--for example, on housing density and
existing vegetation--to identify and rank projects.
Other BIA, FWS, and NPS units primarily used judgmental processes when
selecting projects for 2007. For example, at FWS's Merritt Island National
Wildlife Refuge in central Florida, field staff selected projects
primarily on the basis of the rotational schedule for prescribed burns.
Refuge officials also considered other factors, such as wildlife habitat,
to select which projects to complete that year. According to agency
officials, the refuge has habitat for the scrub jay, a threatened species,
and while prescribed burns generally improve this habitat, too much
prescribed burning can be disruptive. NPS's Cape Canaveral National
Seashore, which neighbors Merritt Island National Wildlife Refuge, also
selected projects judgmentally, and in coordination with refuge staff. The
process was primarily influenced by the location of the wildland-urban
interface and threatened and endangered species habitat.
BIA, FWS, and NPS field units, like national forests and BLM field
offices, also adapted to unanticipated events when selecting projects. In
some cases, field units were forced to accommodate unique circumstances.
For example, the Merritt Island National Wildlife Refuge is adjacent to a
National Aeronautics and Space Administration facility, and, during the
days immediately before and during scheduled rocket or shuttle launches,
the refuge must put all prescribed burns on hold.
Several Improvements Could Help Better Ensure That Fuel Reduction Funds Are
Allocated to Effectively Reduce Risk
Although the Forest Service and Interior are taking steps to enhance their
funding allocation and project selection processes--for example, by
developing models to assist in making allocation decisions--there are
several improvements they could make to better ensure that they allocate
fuel reduction funds to effectively reduce risk. Specifically, when
allocating funds and selecting projects, the agencies could improve their
processes by (1) consistently assessing all elements of wildland fire
risk, including hazard, risk, and values; (2) developing and using
measures of the effectiveness of fuel reduction treatments; (3) using this
information on effectiveness, once developed, to assess the
cost-effectiveness of potential treatments; (4) clarifying the relative
importance of the numerous factors they use in allocating funds, including
factors unrelated to risk or effectiveness; and (5) following a more
systematic process in allocating funds. While the agencies have recognized
the importance of these elements--particularly risk, treatment
effectiveness, and cost effectiveness--in several strategy documents, they
have not effectively incorporated them into their allocation processes.
The Agencies Do Not Consistently Assess All Elements of Risk When
Allocating Funds
The agencies have repeatedly stressed the importance of identifying
high-risk areas in setting priorities and allocating funds for fuel
reduction; for example, in their 2006 document Protecting People and
Natural Resources: A Cohesive Fuels Treatment Strategy (Cohesive
Strategy), ^[109]32 the Forest Service and Interior declared that they
"expect to ensure that fuel project investments are cost-effectively
allocated to achieve risk reductions." Similarly, in its 2007 budget
justification, the Forest Service declared that the fuel reduction program
focuses on reducing the risk of wildland fire and long-term damage to
resources and property; likewise, Interior's 2007 budget justification
declared that the department intended to reduce fuels in order to "provide
better risk reduction to communities and resources."
At the national level, the Forest Service and FWS headquarters
incorporated nationwide risk assessments into their 2007 allocation
processes; Interior did so for only 5 percent of the funds it allocated to
the four Interior agencies; and BIA, BLM, and NPS did not include risk
assessments in their national allocation processes at all, although BLM
officials said they are taking steps to do so in the future. According to
Forest Service and Interior agency officials, it has been difficult to
develop national risk assessments because they require nationally
consistent data, which have not always been available. ^[110]33
Furthermore, some of the available national data on vegetation type and
condition were designed for forests and, consequently, are not as accurate
for shrublands and grasslands.
At the regional and local levels, some agency offices used risk
assessments when allocating funds and selecting projects, while others did
not. One of the Forest Service's 9 regions and 2 of BLM's 11 state offices
considered all three elements required for a risk assessment in their 2007
allocation processes, and several other Forest Service regions considered
two of the three elements--hazard and values--but did not consider risk.
Some, but
not all, of the other Interior agencies' regional offices we visited
considered elements of risk assessments in their allocation processes as
well. Regional officials offered several reasons for not always
systematically considering risk assessments when allocating funds, such as
not having the necessary data for a regionwide risk assessment or only
informally considering risk.
^32U.S. Department of the Interior and USDA Forest Service, "Protecting
People and Natural Resources: A Cohesive Fuels Treatment Strategy,"
February 2006. Note that, although the document is referred to as a
cohesive strategy, previous GAO reports concluded that it does not contain
all the elements GAO called for in its earlier recommendations for such a
strategy. See, for example, GAO-06-671R.
^33The agencies expect that nationally consistent data will be available
through LANDFIRE, a geospatial data and modeling system currently being
implemented. LANDFIRE data are complete for some of the country, with data
for the remainder of the country expected to be completed by 2009.
Several agency officials told us that they do not consider the lack of a
formal national or regional risk assessment to be a significant problem
because they rely on field units to assess risk when selecting projects.
However, as with regions, not all local units used risk assessments when
selecting projects; some local units used only partial assessments or did
not use risk assessments at all. Even when field units do use risk
assessments to help select projects in high-risk areas at the local level,
agency officials cannot be confident that areas designated as high risk
locally would still be designated as high risk at the regional or national
level. For example, one BLM field office in Colorado oversees a rural area
with only two communities, neither of which is at risk from wildland fire,
according to BLM officials. For officials at this office, the most
important values at risk are rural power lines and oil and gas
infrastructure; therefore, they give the highest priority to projects that
protect these features. From a regional or national perspective, however,
other projects may be a higher priority for funding because the values at
risk are more important, the area is at higher risk from fire, the level
of hazard is greater because of fuel conditions, or some combination of
these reasons. Without using national, regional, and local-level risk
assessments that systematically assess hazards, risks, and values, it is
difficult to ensure that allocation decisions are grounded in a clear
understanding of which areas are at the highest risk.
Even when the agencies conduct risk assessments that include hazards,
risks, and values, they may find it difficult to distinguish between high-
and low-priority locations because one key value at risk--the
wildland-urban interface--has multiple definitions that leave considerable
room for interpretation on the part of agency officials. As a result, many
different areas can be classified as wildland-urban interface, and the
term's usefulness in helping agency officials identify, and direct funds
toward, the highest-priority lands is diminished. In 2001, the
agencies--in cooperation with tribes and states--defined the interface as
including three categories:
(1) dense populations (250 or more people per square mile) abutting
wildlands; (2) scattered populations (28 to 250 people per square mile)
intermixed with wildlands, and (3) development surrounding an island of
wildland fuel, such as a park or open space. Agency officials told us that
they developed this definition very quickly, in response to legislative
direction, but later came to believe that it overemphasized population
density and was not flexible enough to accommodate differences in
landscape features such as vegetation, terrain, and prevailing weather
patterns, which can affect the size and shape of areas in the
wildland-urban interface.
In 2003, HFRA defined the wildland-urban interface to include an area
within or adjacent to an at-risk community, that is identified in project
recommendations to a federal agency in a community wildfire protection
plan. For areas not in community plans, HFRA specified that areas within
one-half mile of an at-risk community were to be considered wildland-urban
interface, ^[111]34 as were areas within 1-1/2-miles of an at-risk
community under certain conditions, and areas adjacent to evacuation
routes for at-risk communities. According to agency officials, this
definition offered more flexibility by moving away from the focus on
population density, but it applies only to projects conducted using HFRA
authorities.
Most recently, the 2006 10-Year Strategy Implementation Plan developed by
the agencies, western governors, and others, defined the wildland-urban
interface as the "the zone where structures and other human development
meet at-risk forest and rangelands." While this definition provided broad
flexibility, agency officials told us it did not replace the 2001
definition (which focused on population densities), and both the 2001 and
2006 definitions apply to projects other than those conducted using HFRA
authorities. The end result is multiple definitions that--individually and
collectively--allow many different areas to be classified as
wildland-urban interface without specifying whether some ought to be given
higher priority than others.
^34In 2001, a Federal Register notice was published with a list of
wildland-urban interface communities identified by states as being "in the
vicinity of federal lands" and "at high risk from wildfire." However, the
states and tribes used inconsistent approaches to identify these
communities at risk. To standardize these approaches, the National
Association of State Foresters was tasked, in the 10-Year Implementation
Plan, with developing a definition for community at risk, and a process
for states and tribes to follow to identify and prioritize the
communities. Accordingly, in 2003, the National Association of State
Foresters finalized its guidance, defining community as "a group of people
living in the same locality and under the same government," and specifying
that a community was to be considered at risk from wildland fire if it was
located within the wildland-urban interface as defined in the 2001 Federal
Register, which stated that, "the urban-wildland interface community
exists where humans and their development meet or intermix with wildland
fuel."
In part because of this lack of clarity, agency officials we spoke with
reported including several types of locations under the category of
wildland-urban interface. Some units interpreted the interface to mean
only the area surrounding houses, while others also included roads, power
lines, oil and gas development, communications infrastructure, campgrounds
and recreation areas, and other features. For example, the BLM Colorado
state office defined industrial interface as a subcategory of the
wildland-urban interface, including features such as power lines or oil
and gas development, which are common features in or near some of BLM's
rural field units. In contrast, officials for two national forests near
urban areas (Atlanta and Los Angeles) determined that most or all of their
forests were in the wildland-urban interface because, they estimated, a
wildland fire could move into nearby urban and suburban areas within a
single day. In yet another interpretation, a BIA agency in New Mexico
tailored its definition of wildland-urban interface to accommodate
cultural differences between tribes, as the differences were reflected in
the arrangement of their homes: one tribe built its homes in clusters
while another built its homes in a scattered pattern.
Although each of these interpretations of wildland-urban interface may
have merit given the situations the field units face, the lack of clear
definition effectively allows a wide range of areas to be defined as
wildland-urban interface. The fluid nature of the wildland-urban interface
definition is illustrated by guidance that one FWS region issued to its
local units in 2006, when it notified them that it was expanding the
relatively strict definition of wildland-urban interface the region had
previously used to reflect interagency guidance. According to this region,
"this expanded definition may enhance our ability to fund a project with
[wildland-urban interface] ... funding and will help us meet the
[wildland-urban interface] treatment targets mandated by the Department."
Given the range of definitions available for wildland-urban interface, it
is not surprising to find that in 2005 and 2006 many of the fuel reduction
treatments the agencies identified as being in the wildland-urban
interface were in ZIP code areas with fewer than 28 people per square
mile, on average. ^[112]35 (See fig. 7.) Specifically, about 2.2 million
acres, or 65 percent of all acres treated in areas identified as the
wildland-urban interface during
that period, were in ZIP code areas with fewer than 28 people per square
mile. While the agencies may have had legitimate reasons for some of these
treatments--for example, to protect a critical evacuation route for a
larger community--it is not clear why, as a whole, so many acres treated
are far from more densely populated areas. Expressing its concern about
this situation in 2006, the Office of Management and Budget noted, "As the
agencies increase their emphasis on [wildland-urban interface] treatments
over time, field staff and/or project proponents may simply be defining
more projects as [wildland-urban interface] projects in order to increase
the likelihood of having their projects funded."
^35We conducted our analysis using census data on the average population
per square mile across areas defined by ZIP codes. However, especially in
larger ZIP codes, there may be smaller pockets where the population
density is higher or lower than the average used in our analysis.
Figure 7: Density of Wildland-Urban Interface Treatments and Population
Density, by ZIP Code
Density of 2005 and 2006 Fuel Reduction Treatments Identified as Being in
the Wildland-Urban Interface, by ZIP Code
Less than 5,000 acres per zip code 5,000 or more acres per zip code
Population Density in 2000 in the Continental United States, by ZIP Code
Fewer than 28 people per square mile 28 to 249 people per square mile 250
or more people per square mile
Source: GAO analysis of Forest Service, Interior, and U.S. Census data.
Note: We conducted our analysis using census data on the average
population per square mile across areas defined by ZIP codes. However,
especially in larger ZIP codes, there may be smaller pockets where the
population density is higher or lower than the average used in our
analysis. When mapping the data, we included a 1.5-mile buffer around the
ZIP code areas with 28 or more people per square mile to account for the
1.5-mile buffer specified in the HFRA definition for wildland-urban
interface.
Conversely, population density alone may not be sufficient justification
for selecting locations for fuel reduction. One Forest Service official
cautioned against "prioritization by census," because more densely
populated areas are not necessarily at greater risk than less populated
areas. For example, although Chicago is a densely populated urban area,
the Forest Service has not conducted more treatments in the nearby
grassland because the risk of a fire threatening the urban area is very
low, according to agency officials. In addition, highly populated urban
areas are often not as close to federal lands as are communities with
smaller populations, and the agencies conduct the majority of their fuel
reduction work on federal lands. Figure 8 shows the location of federal
lands relative to more densely populated areas in the continental United
States. Even if a dense urban area is near federal lands, the entire area
is not typically at risk from a fire originating on federal lands; only
the portion of structures closest to federal lands is at risk, according
to Forest Service officials. ^[113]36 Finally, vegetation and other
conditions on some federal lands make it unlikely that a fire would burn
or that a fire would threaten a nearby population.
^36However, in a fire behavior assessment of the June 2007 Angora Fire in
California, Forest Service officials stated that a large number of houses
ignited because of embers from other burning houses, rather than from
wildland fuel--suggesting that even homes that are not immediately
adjacent to federal lands could be at risk from wildland fire.
Figure 8: Location of Federal Lands and Populated Areas in the Continental
United States
Federal Lands Managed by the Forest Service, BLM, BIA, FWS, and NPS in the
Continental United States
Population Density in 2000 in the Continental United States, by ZIP Code
Fewer than 28 people per square mile 28 to 249 people per square mile 250
or more people per square mile
Source: GAO analysis of U.S. Censusand U.S. Geological Survey's National
Atlas Website data.
Note: We conducted our analysis using census data on the average
population per square mile across areas defined by ZIP codes. However,
especially in larger ZIP codes, there may be smaller pockets where the
population density is higher or lower than the average used in our
analysis. When mapping the data, we included a 1.5-mile buffer around the
ZIP code areas with 28 or more people per square mile to account for the
1.5-mile buffer specified in the HFRA definition for wildland-urban
interface.
While many important contextual details are not visible on a national map,
some can be seen at the county level. For example, in Los Angeles
County--the most populous U.S. county--many of the fuel reduction
treatments completed in 2005 and 2006 were adjacent to densely populated
areas, as shown in figure 9, but some were miles away and in ZIP code
areas with relatively low population.
Figure 9: Map of Los Angeles County Wildland-Urban Interface Fuel
Reduction Treatments Completed in 2005 and 2006, and Population Density
Angeles
National Forest
Note: We conducted our analysis using census data on the average
population per square mile across areas defined by ZIP codes. However,
especially in larger ZIP codes, there may be smaller pockets where the
population density is higher or lower than the average used in our
analysis.
Treatments occurred in these low-density areas for several reasons. First,
many of the treatments conducted in the county during that period, while
not immediately adjacent to the city of Los Angeles, were on the federal
land closest to the city, the Angeles National Forest, which of course is
not highly populated. Also, while the average population density for the
general area is low, individual communities with populations ranging from
about 1,000 to 3,000 are located inside the boundaries of the forest, and
the Forest Service conducted some treatments to protect them. Second,
developed sites--such as campgrounds, roads, and recreation areas-- where
people temporarily congregate may not be reflected on a census map of
population density. According to officials at the Angeles National Forest,
human-caused wildland fires generally coincide with such areas, making it
important to conduct fuel treatments around these sites. Finally,
low-density areas within the forest were more feasible to treat than some
areas closer to population centers because steep terrain across much of
the forest--including along its southern boundary adjacent to heavily
populated Los Angeles--makes it difficult and expensive to conduct fuel
treatments, and, in some cases, would make treatments ineffective,
according to agency officials.
In contrast to Los Angeles County, Rio Blanco County, Colorado, is a rural
county with a total population of about 6,000 and an average population
density throughout the county of less than 28 people per square mile.
Nevertheless, BLM classified some of its fuel reduction treatments in this
county as wildland-urban interface treatments. As figure 10 shows, these
treatments in 2005 and 2006 were generally located far from the largest
towns in the county--Meeker and Rangely--which each has a population of
about 2,000.
Population Density
Source: GAO analysis of Forest Service, Interior, U.S. Census, and U.S.
Geological Survey's National Atlas Web site data.
Note: We conducted our analysis using census data on the average
population per square mile across areas defined by ZIP codes. However,
especially in larger ZIP codes, there may be smaller pockets where the
population density is higher or lower than the average used in our
analysis. In Rio Blanco County, there were no ZIP code areas with an
average population of 28 or more people per square mile.
According to BLM officials, they did not conduct wildland-urban interface
treatments closer to these towns because the towns are not at significant
risk from wildland fire; they are surrounded in large part by rocky
outcroppings and irrigated agricultural fields, where fires would not
likely start, and area roads serve as fire breaks. In addition, the county
had prepared a community plan identifying its highest priorities, and the
federal lands surrounding the towns were not among them. Instead, many of
the fuel reduction treatments the agencies did implement--including the
five located southwest of Meeker--were conducted to protect energy
development facilities, such as a coal mine and oil and gas wells, or
power lines that service such facilities, according to BLM officials.
According to the officials, they selected these projects because they were
higher priority than other potential projects in the county, and because
the county's community plan had identified the protection of energy
development and power lines as its priorities, defining the wildland-urban
interface to include the areas around such infrastructure. While these
decisions may be reasonable given local priorities, it is not clear from a
national perspective whether the values at risk in this case are of higher
priority than the values at risk in other locations--in part because the
definitions of the wildland-urban interface do not distinguish the
relative importance of different values at risk, such as homes, power
lines, or oil and gas wells, among others.
The Agencies Do Not Consider Treatment Effectiveness in Their Allocation
Processes Because They Have No Measure for Effectiveness
Although the agencies recognize the importance of measuring the
effectiveness of fuel reduction treatments--that is, how much risk
reduction is achieved through a given treatment and for how long--none of
the agencies considered effectiveness when allocating funds in 2007
because they have not yet developed a method for measuring it. Without
understanding the potential effectiveness of fuel reduction treatments,
the agencies cannot ensure that funds are allocated appropriately, because
not all areas that rank high in a risk assessment can be treated with the
same degree of success. For example, parts of southern California are
dominated by chaparral ecosystems, which feature plants with
fire-resistant roots, enabling the plants to re-sprout quickly. Some of
the plants also encourage fire because their leaves are coated with a
flammable resin. Although these areas of chaparral ecosystems would score
high on a risk assessment--because there is a high vegetation hazard near
populated areas with considerable values at risk--agency officials told us
that fuel reduction treatments in chaparral may be effective for only a
short time because the vegetation often grows back quickly. In addition,
many of the damaging fires in southern California chaparral have been
fanned by the warm, dry, and extremely powerful Santa Ana winds, making it
difficult for fuel treatments to affect fire severity, according to some
Forest Service officials. As a result, some of these areas, though at high
risk from fire, might not be designated as high priority for fuel
treatments. In general, understanding the expected effectiveness of fuel
reduction treatments under different conditions can help the agencies
target their funds toward treatments that will achieve the most risk
reduction for a given cost. The agencies have, on multiple occasions,
recognized the significance of treatment effectiveness; for example, in
the 2006 10-Year Strategy Implementation Plan, the agencies identified the
need to "explore the feasibility of developing measures that determine the
degree and longevity of fire hazard reduction achieved by hazardous fuels
treatments."
Although the agencies have not yet developed a measure of effectiveness,
they have designed their allocation models to accommodate data on
effectiveness in the expectation that such data will eventually become
available. The Forest Service's model includes two elements intended to
assess effectiveness, but, because the agency does not have data on
effectiveness, one of the elements serves as a placeholder--by assigning
each region an identical score--and thus does not influence priority
scores, while the other uses data on the total number of acres treated in
each region instead. Forest Service officials acknowledged that the number
of acres treated does not reveal how effective the treatments are in
reducing risk, but told us they used this information because they wanted
a measure that would reflect the variation in accomplishment levels from
one region to the next. Interior and BLM also plan to include a measure of
effectiveness in their allocation models, but Interior--like the Forest
Service--currently uses total acres treated, and BLM uses data on
efficiency, including total acres treated and average cost per acre,
because these are the only data available. According to agency officials,
it is difficult to develop a single measure of effectiveness for different
geographic locations and vegetation types, because, for example, a
treatment in grass might be effective for 1 year, while a treatment in
some forests might be effective for 30 years. Nevertheless, as long as the
agencies continue to allocate funds without knowing how effective
treatments are likely to be, they cannot be sure that funds are being
spent on projects that substantially reduce overall risk.
According to Forest Service research scientists, developing a measure of
treatment effectiveness would require that the agencies first determine
how to estimate the level of risk in a given location so they could track
any changes in risk resulting from fuel treatments. For example, they
could use data on fire intensity, severity, or occurrence, or some
combination of these and other factors, to evaluate risk. Once agency
officials determined how to estimate risk, they could use the information
to measure treatment effectiveness. However, there is no consensus on how
best to do so and any method would likely require considerable effort. For
example, under one approach described by the researchers, available
scientific studies about fuel reduction treatments in various vegetation
types would be analyzed to ascertain where fuel treatments are more or
less effective, and effectiveness ratings would be calculated for each
vegetation type on the basis of this information. After establishing the
ratings, they would collect field data to verify their initial conclusions
and ratings--a costly and time-consuming exercise, according to some
researchers. Such an approach would have drawbacks, however; the
researchers told us that it would be difficult to establish a single
rating that would apply to vegetation types under all circumstances
because fuel conditions within a given vegetation type vary widely,
depending, for example, on geographic location and previous fuel reduction
activity. In addition, factors other than vegetation--such as terrain,
weather, and soil--also influence treatment effectiveness. Consequently,
some researchers have proposed alternative approaches, such as one that
would consider many factors, in addition to vegetation type, to assign
effectiveness ratings to individual treatment areas rather than general
vegetation types. However, developing an effectiveness rating scheme using
this approach--or others that incorporate numerous factors--would require
significant research and analysis over a long time period, according to
one researcher.
A less expensive, quicker approach outlined by another Forest Service
researcher would rely on expert opinion rather than field data. Under this
simplified approach, a panel of experts with knowledge about and
experience in fuel reduction treatments and their effectiveness would use
their professional judgment to collectively estimate the extent to which
fuel treatments would be effective in each of several vegetation or fire
regime condition class categories. The experts' estimates could then be
used to inform decisions on allocating funds.
The Agencies Often Consider Costs, but Not Cost-Effectiveness, When
Allocating Funds
The agencies also do not consider the cost-effectiveness of treatments
when allocating funds, primarily because they do not have data on
treatment effectiveness. Treatment costs can vary widely in different
areas, from as little as $10 per acre to well over $1,000 per acre, even
ranging as high as $30,000 per acre under unusual circumstances, and
allocating funds wisely involves not simply targeting those acres that can
be treated most cheaply, but those acres where treatments yield the most
cost-effective result. While considering costs is an important step in
making allocation decisions, it is equally important to consider
effectiveness in conjunction with costs to avoid funding ineffective
projects simply because they are cheap. However, until the agencies have
data on treatment effectiveness, they will find it difficult to do so. In
support of these considerations, the 2006 Cohesive Strategy emphasized the
importance of reducing fuel in the most cost effective manner possible,
because federal funds can support only a finite number of fuels treatments
each year covering a fraction of the acres at high risk from unusually
severe fires.
In practice, the agencies frequently consider costs when allocating funds
and selecting projects. They sometimes give priority to projects with low
per-acre costs in order to leave more funds available for other projects
or to treat more acres within their budgets--an important factor for
agencies trying to meet increasing acreage targets. Also, agencies
sometimes give priority to low-cost treatments in areas that have
previously been treated and are consequently of relatively low risk, in
order to prevent them from becoming higher risk. According to agency
officials, these treatments are a priority because they are a
cost-effective way to maintain low-risk conditions once achieved; it is
generally much cheaper to reduce fuel in areas that have recently been
treated than to do so in areas that have never been treated or have not
been treated for a long time. However, without knowing the effectiveness
of treatments in reducing risk, agency officials may not be able to
compare the relative benefits of potential projects when deciding where to
invest fuel reduction funds--and, thus, may not know which projects are
likely to be the most cost-effective.
In some cases, the agencies also give lower priority to treatments with
very high per-acre costs--even in high-risk areas--because the expected
benefit does not justify the expense. For example, the Desert National
Wildlife Refuge in southern Nevada identified a mechanical thinning
treatment to remove palm trees as its highest-priority fuel reduction
project in 2006. The proposed project was in the wildland-urban interface
and would also improve the habitat of an endangered fish, according to
agency officials. However, it would have cost hundreds of thousands of
dollars--nearly the entire budget for the region--and, therefore, FWS
regional officials did not fund the project.
The Agencies Have Not Established Clear Guidance on the Relative
Importance of Factors Used in Setting Priorities
In addition to more consistently using information on risk and developing
measures of treatment- and cost-effectiveness, the agencies could improve
their allocation process by clarifying the relative importance of the
different factors they use in setting priorities. Without such
clarification, it is not clear how agency officials are to resolve
conflicts that arise between competing factors. In addition, when factors
other than risk, treatment effectiveness, and cost effectiveness have
considerable influence on allocation decisions, it is difficult for the
agencies to ensure that funds are allocated to areas where they will most
effectively reduce risk.
The agencies consider such factors in part because they are directed to do
so; many of the factors they consider are tied to federal laws or
congressional direction. For example, fuel reduction projects authorized
under HFRA include, among others, projects on federal land in the
wildland-urban interface and certain projects in areas where ecological
restoration is needed because vegetation has departed significantly from
its historical regime. The act requires the agencies to develop annual
programs of work for federal land that give priority to authorized
hazardous fuel reduction projects that provide for the protection of
at-risk communities or watersheds or that implement community wildfire
protection plans. Congressional committee direction has also called for
the agencies to put a priority on fuel reduction work completed through
mechanical treatments and projects that use biomass.
Some of the factors the agencies consider are also intended to encourage
efficiency in the fuel reduction program, as well as more broadly in their
land management missions. Specifically, the agencies give priority to
projects that achieve benefits not only for the fuel reduction program but
also for other programs such as wildlife management and watershed
improvement--an approach referred to as integration among programs. Agency
officials said implementing such projects is a way to leverage funds and
coordinate resources. The Forest Service also emphasizes these projects
because its interpretation of the President's HFI calls for a focus on
integrated management, according to agency officials.
In the face of multiple directives and competing agency priorities, agency
officials must balance numerous factors when allocating funds and
selecting projects, as the following examples illustrate:
o Priorities in community plans may not always align with
agency-identified priorities, forcing agencies to choose between them.
According to Montana BLM officials, one community proposed a fuel
reduction project in an area the officials believed was relatively low
risk because it had vegetation that does not burn easily. However, the
officials agreed to implement the project because they are directed to
give priority to locally identified projects and because they did not
want to damage their relationship with the community. Several agency
officials told us that community plans did not always include federal
lands or propose projects in locations where the agencies could
feasibly implement a treatment. In such cases, agency officials
sometimes worked with the communities to identify project locations
agreeable to all, while other times they implemented agency-identified
projects instead of those identified in the plans.
o Direction to give priority to high-risk areas may also conflict with
the agencies' commitment to meet acreage targets. Several agency
officials told us that they sometimes implemented lower-priority
projects with low unit costs because they felt pressure to meet
acreage targets. In some cases, these projects, although low priority
for fuel reduction purposes, were a high priority for other resource
programs or achieved other management objectives.
o Direction to give priority to areas in the wildland-urban interface
may conflict with other agency priorities. For example, NPS officials
told us that giving priority to fuel reduction treatments at the
interface conflicted with the agency's mission to preserve natural
ecosystems and processes, which would call for giving priority to
treatments in undeveloped areas.
o Desire for stable funding and staff levels may make officials
reluctant to shift funds on the basis of risk assessments. When
allocating funds, the agencies frequently emphasized the importance of
maintaining stable funding levels and minimizing disruptions to staff,
which can conflict with the direction to emphasize high-risk areas.
According to agency officials, stable allocations to regions and field
units are needed to ensure predictability and enable regional and
field staff to plan ahead. In addition, a minimum level of funding is
needed to maintain the workforce and infrastructure required to
support viable fuel reduction programs in regions and field units.
Several agency officials told us they were reluctant to shift funding
on the basis of risk assessments because doing so could require staff
to relocate-- potentially multiple times--and the officials wanted to
avoid uprooting staff.
Agency guidance offers little in the way of clarification for staff
confronted with numerous, conflicting priorities, as the multiplicity of
priorities in the Forest Service and Interior's 2006 Cohesive Strategy
illustrates. In this strategy, the Forest Service and Interior outline a
set of national fuel treatment priorities but do not establish a hierarchy
of their relative importance. Among the treatment priorities are areas in
the wildland-urban interface as well as some areas outside the interface
in condition classes 2 or 3. In addition, some areas in condition class
1--the only remaining condition class--are to be given equal priority,
according to the strategy. Similarly, the strategy calls for priority to
be given to mechanical treatments where appropriate, but also to
prescribed burns where appropriate. After providing a list of priority
criteria, the strategy declares that the more criteria a fuel reduction
project meets, the higher its priority should be for funding. However, it
also acknowledges that, in exercising management discretion, the agencies
may need to make exceptions to the process described for ranking and
selecting projects.
Agencies' Allocation Processes Are Not Always Systematic
Although the agencies are working to develop and implement models that
will allow them to allocate funds more systematically, such systematic
approaches are not used by all agencies or at all levels within the
agencies. By allocating funds using a systematic process--one that is
methodical, based on established criteria, and applied consistently--the
agencies can better ensure that they uniformly consider all relevant
criteria and appropriately apply these criteria in all decisions.
In particular, when agency officials rely primarily on professional
judgment and negotiation to allocate funds, they do not always follow a
step-by-step approach or consistently apply a predetermined set of
criteria. We recognize that agency decision makers--particularly those who
have served in the same location for many years--often have detailed
knowledge about on-the-ground conditions and a thorough understanding of
fuel reduction needs. Nevertheless, without using a systematic approach,
even knowledgeable and well-meaning decision makers may be more
susceptible to influences that are not intended to be part of the
decisions, as illustrated by the following examples:
o According to several agency officials, they face considerable pressure
to meet acreage targets. Under these circumstances, and with no
predetermined set of criteria in an allocation process, targets could
have more influence than intended. That is, agency officials might
fund lower priority projects in order to treat more acres.
o In NPS's Southeast region, agency officials told us that the location
of full-time fuel reduction staff has considerable influence on
allocations, even though it is not officially a factor in the
allocation process. Few parks in this region have full-time staff
devoted to fuel reduction, and parks without such staff request and
receive much less fuel reduction funding than do the parks with
dedicated staff--potentially because there are fewer staff to perform
the work necessary to identify fuel reduction needs and request funds.
Consequently, according to agency officials, it is difficult to ensure
that all of the highest-priority areas for fuel reduction across the
region are identified and targeted for funding
because some high-priority areas may not be identified if they are
located in parks with fewer staff. NPS officials in another region
expressed a similar concern, stating that the agency needs to shift fuel
reduction funds within the region to direct them to high-priority
locations and acknowledging that it cannot do so without also shifting
personnel to high-priority locations.
Moving toward more systematic allocation processes also enhances
transparency and accountability. In many of the locations we visited, the
agency offices that relied primarily on professional judgment to allocate
fuel reduction funds and select projects did not document the rationale
for their decisions. As a result, the processes were not transparent, and
neither agency officials nor others--including Congress and the public--
could understand the rationale behind the decisions or have confidence
that the resulting allocations were directed to the highest-priority areas
for reducing risk to communities and the environment. For example,
officials in BLM's Oregon/Washington state office used their professional
judgment to determine allocations to its 10 district offices. Under this
process in 2007, BLM's Medford district office received an allocation of
about $9 million--over 7 times the average allocation received by the
other nine district offices that year. While this disparity may be
appropriate, without a transparent process it is difficult to determine
the extent to which the allocation reflects agency priorities for reducing
risk to communities and the environment, rather than other factors. The
agencies themselves have emphasized the importance of transparency and
accountability; for example, the 10-Year Strategy Implementation Plan
states that the agencies should "strive for maximum transparency in the
decision-making process."
Conclusions
Our nation's wildland fire problem has been decades in the making and will
not be solved quickly. Nevertheless, with careful choices about where to
spend their limited fuel reduction dollars, federal agencies can
meaningfully, if incrementally, reduce the risks faced by communities and
the environment. Doing so will require the agencies to continue moving
away from allocation by tradition to allocation by priority. Toward this
end, the agencies could improve their current approaches in three key
areas.
First, the agencies would benefit from routinely using an allocation
process that is systematic, and that is common to all the agencies. A
systematic process can help ensure that the agencies apply their
allocation and project selection criteria consistently, and can help
interested parties outside of the process--Congress, local communities,
and other entities-- understand the rationale for the funding and project
selection decisions that are made. While the models that some of the
agencies are developing represent substantial steps forward in this regard
and will affect larger portions of funding allocations over time, not all
of the agencies have models, and none consistently uses models at the
national, regional, and local levels. Further, the models, even where
used, often exert only a small influence on allocation decisions, partly
because the agencies do not yet have full confidence in the models' data.
As a result, the agencies often base decisions mainly on historical
funding patterns and professional judgment. We recognize that professional
judgment will always have a role in the allocation process to account for
difficult-to-quantify factors, such as local priorities or political
considerations. However, the agencies and the public are best served if a
systematic process, such as a model, serves as the foundation for
allocation decisions, and professional judgment plays a supporting, rather
than a lead, role. Also, given that wildland fire is a nationwide problem
that does not respect administrative boundaries, the agencies would do
well to develop and use a common process for allocating fuel reduction
funds--as Congress has called for--that can be customized to accommodate
differences in scale, type of ecosystem, agency mission, and other
criteria.
Second, the agencies could improve the information they use to make
allocation decisions. Because the agencies do not always use risk
assessments and currently lack data on treatment effectiveness, they often
make allocation decisions without knowing, on a broad scale, where the
acres at highest risk are located, which treatments are most effective at
reducing risk, and which areas respond best to treatment. To improve their
allocation decisions, they should continue, over the long term, to develop
and use information on risk and treatment effectiveness. The agencies can
then use this information, in concert with cost information, to
effectively assess tradeoffs among potential treatments and identify the
most cost-effective investments.
Finally, the agencies could strengthen their allocation processes by
sorting through the numerous prioritization factors that have accumulated
over the years and establishing a hierarchy for considering them. Without
such a hierarchy, the exercise of setting priorities can be
frustrating--or even meaningless--because virtually any project can
qualify as high priority. While we recognize that, in some cases, the
agencies are bound by law or congressional direction to give priority to
certain factors, we believe there may remain enough room within those
constraints not only to establish a hierarchy of factors, but also to
clarify the relative importance of categories within some factors--in
particular, various categories of wildland-urban interface. We do not
advocate prioritization by census-- simply directing fuel reduction funds
to areas with the highest populations--but neither do we believe that the
agencies or the public are well-served by the broad definitions of the
interface currently used. However, if the agencies determine, through
further analysis, that laws or congressional direction create conflicts
prohibiting them from implementing a consistent, systematic approach that
distinguishes the relative importance of various priorities, they should
so inform Congress and seek additional clarification.
It will not be easy to carry out these tasks. As they work to improve
their processes, the agencies will need to devote considerable effort to
developing measures and collecting data on risk and effectiveness and
considerable thought to balancing this information against the many goals
of the fuel reduction program--all in a way that yields transparent
results. And once these steps are carried out, the agencies face perhaps
an even more difficult decision: how best to redirect fuel reduction funds
in a way that improves the agencies' effective use of their limited funds
despite the potentially disruptive consequences for individual field units
or nearby communities. Our findings suggest that the agencies are
increasingly mindful of the merits of such an approach and that their
recent actions have begun to lay the necessary groundwork. Nevertheless,
many challenges remain, and a difficult road lies ahead.
Recommendations for Executive Action
We are recommending that the Secretaries of Agriculture and of the
Interior take the following five actions to improve their ability to
allocate fuel reduction funds so that these funds contribute most
effectively to risk reduction.
First, we recommend that the Secretaries of Agriculture and of the
Interior direct the agencies to develop a common, systematic funding
allocation process in order to enhance the transparency and accountability
of their allocation decisions and to ensure a common federal approach to
allocating funds. Such a systematic process should serve as the foundation
of each agency's allocation process and should be applied at all levels
within the agencies. Existing models or those under development may serve
as useful prototypes; for example, while we have not assessed its accuracy
or technical soundness, the Forest Service's model for allocating funds
shows promise as the foundation of a systematic process.
In addition, we recommend that the Secretaries of Agriculture and of the
Interior direct their agencies to develop information to support this
systematic process. Development of the information should include the
following actions:
o Develop and implement a common approach to risk assessment, to provide
for a broad, national assessment of hazard, risk, and values, as in
the Forest Service's allocation model, as well as more refined
regional and local assessments.
o Devote resources to developing a measure of, and subsequently
collecting data on, fuel reduction effectiveness, so that the agencies
can usefully estimate the extent and duration of risk reduction from
potential fuel treatments. Because developing the measure and
collecting data are likely to be difficult and time-consuming
endeavors, the agencies might find it useful to proceed with convening
a panel of experts to devise a rudimentary framework for estimating
treatment effectiveness.
o Use information on risk and fuel treatment effectiveness, once
available, in concert with information on the cost of treatments, to
assess the cost-effectiveness of various potential fuel reduction
treatments.
Finally, the Secretaries of Agriculture and of the Interior should provide
guidance that clearly distinguishes the relative importance of the various
factors used in allocating funds and selecting projects, including the
importance of risk, effectiveness, and cost in comparison with other
factors. This guidance should also distinguish the relative priority of
different values at risk, especially different elements within the
wildland-urban interface, such as homes, power lines, and municipal
watersheds.
Agency Comments
and Our Evaluation
We provided the Secretaries of Agriculture and of the Interior with a
draft of this report for review and comment. The Forest Service and the
Department of the Interior generally agreed with the findings and
recommendations in the report, noting their ongoing efforts to develop and
implement a risk-informed allocation process, and reiterating the
importance of including state, tribal, and local concerns in the
prioritization process. Their joint comment letter is reproduced in
appendix IV.
We are sending copies of this report to interested congressional
committees, the Secretaries of Agriculture and the Interior, the Chief of
the Forest Service, and other interested parties. We will also make copies
available to others upon request. In addition, the report will be
available at no charge on the [114]GAO Web site at http://www.gao.gov.
If you or your staffs have any questions about this report, please contact
me at (202) 512-3841 or [email protected]. Contact points for our Offices
of Public Affairs and Congressional Relations may be found on the last
page of this report. GAO staff who made major contributions to this report
are listed in appendix V.
Robin M. Nazzaro
Director, Natural Resources and Environment
Appendix I: Objectives, Scope, and Methodology
We were asked to (1) identify the processes the Forest Service, the
Department of the Interior (Interior), and Interior's agencies--the Bureau
of Indian Affairs (BIA), Bureau of Land Management (BLM), Fish and
Wildlife Service (FWS), and National Park Service (NPS)--use to allocate
fuel reduction funds and select projects for implementation, including the
factors that influence these processes; and (2) determine how, if at all,
the agencies could improve these processes to better ensure they
contribute to their goal of effectively reducing the risk of wildland fire
to communities and the environment. We focused our review primarily on the
Forest Service and BLM because these two agencies accounted for about 80
percent of the fuel reduction funds appropriated by Congress for 2005,
2006, and 2007, although we collected information on the other three
agencies as well. ^[115]1 We focused our review on fuel reduction work
funded through congressional fuel reduction appropriations; therefore,
fuel reduction work funded by other agency programs or outside
organizations is outside the scope of this review. To gain an
understanding of outside perspectives on the agencies' fuel reduction
efforts, we met with several nonfederal parties, including representatives
from the National Association of State Foresters, The Nature Conservancy,
the Western Governors' Association, and the Wilderness Society.
Fuel Reduction Funding Allocation and Project Selection Processes
To learn how the agencies allocate fuel reduction funds and select
projects, and to identify the factors that influence these processes, we
first obtained and reviewed documents on policies and procedures governing
the fuel reduction program. These included applicable laws, administrative
initiatives, congressional committee reports, and interagency agreements,
as well as guidance for fuel reduction from the departments, agency
headquarters, and regional offices. ^[116]2 We also obtained and analyzed
agency data on funding allocations.
To learn about the processes used to allocate fuel reduction funds at the
national level, we met with agency officials from the Forest Service and
Interior at their Washington, D.C., headquarters, and with officials from
all five agencies at the National Interagency Fire Center in Boise, Idaho.
We also met with agency researchers and modeling experts to better
understand the data used in the national models currently under
development by the Forest Service, Interior, and BLM. We did not, however,
assess the accuracy or technical soundness of these models.
^1Years cited in this appendix refer to fiscal years except where
otherwise specified.
^2BIA, FWS, and NPS have regional offices, while BLM has state offices.
For the purposes of this appendix, we refer to all of these as regional
offices when we discuss the Interior agencies collectively.
At the regional and state levels, we used a structured interview guide to
speak, in person or by telephone, with officials in all Forest Service
regional and BLM state offices, as well as with officials in selected BIA,
FWS, and NPS regional offices. The structured interview guide included
questions about the processes used to allocate fuel reduction funds, the
factors that influence those processes, the extent and nature of regional
guidance provided to field units, and the amount of oversight on the part
of the regional offices. Because developing and administering a structured
interview guide may introduce errors--caused by the way a particular
question is interpreted, for example--we included steps in the development
and administration of the interview guide to minimize such errors. We
pretested the guide at several locations and modified it to reflect
questions and comments we received. We also visited a number of the
agencies' regional offices to obtain a greater understanding of the
funding allocation processes in those regions. We selected regional
offices that collectively received a substantial portion of their agency's
fuel reduction funds and represented diversity with respect to fuel
reduction funding levels, fuel reduction acreage accomplishments,
predominant vegetation type, and geographic location. These selection
criteria are shown in table 5.
: Objectives, Scope, and
Methodology
: Regional Offices GAO Visited
Northern x x x x
Pacific Northwest x x x x
Forest
Pacific Southwest x x x x
Service
Rocky Mountain x x x x
Southern x x x x
Northwest x x x x x
Pacific (by phone) x x x x
BIA
Rocky Mountain x x x x x
Southwest x x x x
California x x x x x
Colorado x x x x
BLM Idaho x x x x
Montana x x x x
Oregon/Washington x x x x x
California-Nevada^c x x x x
FWS Mountain-Prairie x x x x
Southeast x x x x
d d dIntermountain x x x
d d dNPS Pacific West x x x
Southeast x x x x
: Objectives, Scope, and
Methodology
: Field Units GAO Visited
x
National Grassland
Bitterroot National Forest^a Montana and Idaho x
Boise National Forest Idaho x
Chattahoochee-Oconee National Forests Georgia x
Medicine Bow-Routt National Forests and Thunder Basin Wyoming and Colorado
x
National Grassland
National Forests in Florida (Ocala National Forest) Florida x
San Bernardino National Forest California x x
BIA
Crow Agency Montana x
Zuni Agency New Mexico x
BLM
Albuquerque District Office New Mexico x
Billings Field Office Montana x
Little Snake Field Office Colorado x x
Twin Falls District Office Idaho x
White River Field Office Colorado x x
Rocky Mountain Arsenal National Wildlife Refuge Colorado x
NPS
Cape Canaveral National Seashore Florida x
Rocky Mountain National Park Colorado x
Sequoia and Kings Canyon National Parks^a California x
Fuel reduction Predominant funding level^a Acres treated^b vegetation type
Geographic location
Greater Less Greater Less Region/ than than than than Agency state office
average average average average Forest Grass Shrub West Central East
Source: GAO analysis of Forest Service and Interior data.
^a"Greater than average" refers to regions that received more than the
average funding amount received by that agency's regions in 2007, and
"less than average" refers to regions that received less than the average
amount in 2007.
^b"Greater than average" refers to regions that treated more than the
average acres treated by that agency's regions in 2006, and "less than
average" refers to regions that treated less than the average acres
treated in 2006.
^cFWS's California-Nevada Operations office is officially part of the
Pacific region, but manages its own fuel reduction program.
^dThese regions each cover several states and have a large variety of
vegetation; therefore, no one vegetation type is predominant, according to
agency officials.
To learn about the project selection processes used by local units, we
selected a nonprobability sample of 20 local units in eight states to
interview. ^[117]3 The sample included 8 national forests, 5 BLM district
or field offices, 2 BIA agencies, 2 national wildlife refuges, and 3
national parks. Table 6 lists the units we visited. The local units
selected for interviews represented diversity with respect to geographic
location and predominant vegetation type. In addition, we selected units
that represented diversity with respect to their proximity to communities
and development, including units that were located in counties that were
predominantly rural or urban.
^3Results from nonprobability samples cannot be used to make inferences
about a population, because in a nonprobability sample, some elements of
the population being studied have no chance or an unknown chance of being
selected as part of the sample.
Table 6: Field Units GAO Visited
Predominant vegetation type Agency and unit State Forest Grass Shrub
Forest Service
Angeles National Forest California
Arapaho-Roosevelt National Forests and Pawnee Colorado
FWS
Merritt Island National Wildlife Refuge Florida
Source: GAO analysis of Forest Service and Interior data.
^aWe met with officials from these field units at off-site locations, in
order to facilitate cost-effective travel logistics.
During all of these visits, we collected documents and interviewed staff;
during some of these visits, we also observed fuel reduction treatments.
Because we conducted in-depth analyses of only a few selected units, we
cannot generalize our findings beyond the local units and officials we
contacted.
Potential Improvements to Agency Processes to Better Ensure They
Contribute to Reducing Risk
To identify potential improvements to the agencies' processes for
allocating fuel reduction funds and selecting treatments, we analyzed the
information we collected through our site visits, structured interviews,
agency documentation, and discussions with other agency officials. To
identify the overall goals of the fuel reduction program, and the extent
to which earlier assessments of the program identified shortcomings in the
agencies' ability to meet these goals, we also evaluated (1) agency policy
documents, including strategy documents, program guidance, and related
documents discussing the program's objectives; (2) legislative direction
associated with the fuel reduction program, including laws, congressional
committee report language, and other direction; and (3) previous reviews
of the fuel reduction program by GAO, the Inspectors General, and others.
In our interviews with agency officials, we asked about the factors they
considered when allocating funds and selecting projects--including the
influence of specific factors, such as acreage targets and risk
assessments--as well as factors that prevented high-priority work from
being accomplished. We also asked about regional and local definitions of
the wildland-urban interface. We assigned the allocation processes used by
the agencies' headquarters, regional offices, and local units to one of
two categories: quantitative or judgmental. We also verified the factors
used in allocation processes with agency officials.
To determine the extent to which the locations of wildland-urban interface
treatments, population centers, and federal lands coincided, we analyzed
fuel reduction data from a Forest Service and Interior database--the
National Fire Plan Operations and Reporting System (NFPORS)--as well as
population data from the U.S. Census Bureau, and federal lands data from
the U.S. Geological Survey's National Atlas Web site (NationalAtlas.gov).
Using the agency data on fuel reduction treatments, we used geographic
information system (GIS) tools to map the location and size of
wildland-urban interface treatments completed in 2005 and 2006. We also
applied GIS tools to Census data to map population density in three
categories: (1) fewer than 28 people per square mile, on average;
(2) 28 to 249 people per square mile, on average; and (3) 250 or more
people per square mile, on average. ^[118]4 We used these categories
because they reflect the definition of wildland-urban interface published
in the
January 4, 2001 Federal Register and used by the agencies. ^[119]5 We also
mapped the location of federal lands using the data from the U.S.
Geological Survey. In addition, we created maps of two counties--one urban
and one rural--showing the locations of wildland-urban interface
treatments completed in 2005 and 2006, population density, and federal
lands. We also contacted officials in those two counties to discuss the
location of specific wildland-urban interface projects and their rationale
for selecting those projects.
^4We conducted our analysis using U.S. Census data on the average
population per square mile across areas defined by ZIP codes. However,
especially in larger ZIP codes, there may be small pockets where the
population density is higher or lower than the average used in our
analysis.
To determine the reliability of the agencies' fuel reduction data, we
reviewed related documentation, such as the NFPORS database users' manual;
interviewed knowledgeable agency officials, including database
administrators; discussed data input and verification procedures with
regional and field staff; and conducted electronic data testing. We found
that these fuel reduction data were sufficiently reliable for the purposes
of this review. We obtained the federal lands data prepared by
NationalAtlas.gov and reviewed the documentation provided on the
limitations of the file. From this review, we determined that the federal
lands data were sufficiently reliable for our purposes. To measure
population density, we used Census ZIP Code Tabulation Area data from the
2000 U.S. Census and the geographic boundary for those areas. We reviewed
documentation provided on the limitations of these files and compared
their consistency with other Census sources. From this review,
we determined that the population density data were sufficiently reliable
for our purposes.
^5In the 2001 Federal Register, the agencies provide three categories of
wildland-urban interface communities. The first is "interface community,"
which exists where structures directly abut wildland fuel; an alternative
definition of the interface community specifies a population density of
250 or more people per square mile. The second is "intermix community,"
which exists where structures are scattered throughout a wildland area; an
alternative definition of intermix community specifies a population
density of between 28-250 people per square mile. The third is "occluded
community," where structures, often within a city, abut an island of
wildland fuel (e.g., park or open space).
We conducted our work from August 2006 to September 2007 in accordance
with generally accepted government auditing standards.
Appendix II: Forest Service and Interior Fuel Reduction Funding Allocations,
Fiscal Years 2005, 2006, and 2007
This appendix provides information on fuel reduction funding
appropriations and allocations to the Forest Service, the Department of
the Interior (Interior) and its agencies, and their regions for 2005,
2006, and 2007. ^[120]1 Interior allocates separate fuel reduction funds
to its four agencies for treatments within and outside of the
wildland-urban interface (WUI), while the Forest Service allocates one
single source of fuel funding to its regions. Therefore, information on
allocation amounts to WUI and non-WUI areas are included for Interior but
not for the Forest Service. Table 7 provides total appropriations and
allocations to the Forest Service and Interior agencies--Bureau of Indian
Affairs (BIA), Bureau of Land Management (BLM), National Park Service
(NPS), and Fish and Wildlife Service (FWS)--for 2005, 2006, and 2007. As
shown in table 7 and figure 11, of the approximately $500 million
appropriated to the Forest Service and Interior for fuel reduction in
2007, the Forest Service received about 61 percent of the total; BLM
received about 19 percent of the total; and the remaining 20 percent was
allocated to BIA, NPS, and FWS.
^1Years cited in this appendix refer to fiscal years except where
otherwise specified. BIA, FWS, and NPS have regional offices, while BLM
has state offices. For the purposes of this appendix, we refer to all of
these as regional offices when we discuss the Interior agencies
collectively.
Table 7: Total Appropriations to Forest Service, and Allocations to
Interior Agencies, Fiscal Years 2005, 2006, and 2007
Agency BLM 2005 Total allocation $91,386,000 Percentage of total
allocation 18.8
2006 96,299,000 19.9
2007 93,389,000 18.8
BIA
2005 42,488,000 8.7
2006 43,237,000 8.9
2007 40,664,000 8.2
NPS
2005 33,040,000 6.8
2006 33,299,000 6.9
2007 31,396,000 6.3
FWS
2005 27,527,000 5.7
2006 32,162,000 6.6
2007 30,666,000 6.2
Subtotal--Interior
agencies
2005 194,441,000 39.9
2006 204,997,000 42.3
2007 196,115,000 39.4
Forest Servicea
2005 292,389,000 60.1
2006 280,119,000 57.7
2007 301,258,000 60.6
Total
2005 486,830,000 100
2006 485,116,000 100
2007 497,373,000 100
Source: GAO analysis of Forest Service and Interior data.
Notes: Interior allocated additional amounts of $6,968,000 in 2005;
$5,115,000 in 2006; and
$3,672,000 in 2007 to the Office of Wildland Fire Coordination, which is
responsible for the
coordination, integration, and oversight of wildland fire management
programs within Interior.
Total allocations do not include carryover from the previous fiscal year.
Numbers may not total due to rounding.
^aForest Service figures represent appropriations.
Figure 11: Agency Funding Levels as a Percentage of Total Fuel Reduction
Funding, Fiscal Year 2007
Forest Service FWS NPS BIA
BLM
Forest Service
Interior agencies
Source: GAO analysis of Forest Service and Interior data.
Notes: Interior allocated an additional $3,672,000 to the Office of
Wildland Fire Coordination. Total allocations do not include carryover
from the previous fiscal year.
Table 8 shows the Forest Service's total allocations to its nine regions
and its headquarters for 2005, 2006, and 2007. In 2007, of the total
funding allocated to the Forest Service for fuel reduction, about 68
percent was allocated to the regions. Approximately 32 percent was
allocated to the Forest Service's headquarters, research stations, and
general cost pools that are used for expenses not charged to a single
program, including indirect, support, and common services charges. The
Pacific Southwest region received the most funding for 2005, 2006, and
2007; the Southwestern region received the second-most funding during that
time period.
Table 8: Forest Service Allocations to Regions and Headquarters, Fiscal
Years 2005, 2006, and 2007
Percentage of
Forest Service's
Region Total allocation total allocation
Pacific Southwest
2005 $66,656,000 22.8
2006 41,944,000 15.0
2007 43,737,000 14.5
Southwestern
2005 30,638,000 10.5
2006 36,891,000 13.2
2007 37,341,000 12.4
Southern
2005 25,478,000 8.7
2006 26,368,000 9.4
2007 29,092,000 9.7
Pacific Northwest
2005 24,622,000 8.4
2006 23,179,000 8.3
2007 25,794,000 8.6
Rocky Mountain
2005 21,032,000 7.2
2006 23,706,000 8.5
2007 25,445,000 8.4
Intermountain
2005 13,673,000 4.7
2006 15,881,000 5.7
2007 16,165,000 5.4
Northern
2005 11,875,000 4.1
2006 12,006,000 4.3
2007 15,782,000 5.2
Eastern
2005 8,633,000 3.0
2006 8,631,000 3.1
2007 9,718,000 3.2
Percentage of
Forest Service's
Region Total allocation total allocation
Alaska
2005 1,834,000 0.6
2006 853,000 0.3
2007 805,000 0.3
Subtotal, regions
2005 204,441,000 69.9
2006 189,459,000 67.6
2007 203,879,000 67.7
Headquarters, Research stations, and cost pools
2005 87,948,000 30.1
2006 90,659,000 32.4
2007 97,379,000 32.3
Total
2005 292,389,000 100
2006 280,119,000 100
2007 301,258,000 100
Source: GAO analysis of Forest Service data.
Notes: Total allocations do not include carryover from the previous fiscal
year. Numbers may not total due to rounding.
Table 9 shows Interior's allocations to BLM, BIA, NPS, and FWS-- including
WUI and non-WUI allocations--for 2005, 2006, and 2007. In 2007, about 65
percent of Interior's total allocation was to WUI areas and 35 percent was
to non-WUI areas. In 2007, BLM received the largest percentage of
Interior's fuel reduction funding allocation--almost 48 percent.
Table 9: Interior Allocations to BLM, BIA, FWS, and NPS, Including WUI and
Non-WUI Allocations, Fiscal Years 2005, 2006, and 2007
Percentage of
Total Interior's total Total WUI Total non-WUI
allocation allocation to allocation allocation
Agency BLM the agencies
2005 $91,386,000 47.0 $64,437,000 $26,949,000
2006 96,299,000 47.0 66,245,000 30,054,000
2007 93,389,000 47.6 66,590,000 26,799,000
BIA
2005 42,488,000 21.9 27,299,000 15,189,000
2006 43,237,000 21.1 27,494,000 15,743,000
2007 40,664,000 20.7 26,681,000 13,983,000
NPS
2005 33,040,000 17.0 15,320,000 17,720,000
2006 33,299,000 16.2 14,948,000 18,351,000
2007 31,396,000 16.0 14,583,000 16,813,000
FWS
2005 27,527,000 14.2 15,583,000 11,944,000
2006 32,162,000 15.7 19,772,000 12,390,000
2007 30,666,000 15.6 18,922,000 11,744,000
Total--Interior
agencies
2005 $194,441,000 100 $122,639,000 $71,802,000
2006 $204,997,000 100 $128,459,000 $76,538,000
2007 $196,115,000 100 $126,776,000 $69,339,000
Source: GAO analysis of Interior data.
Notes: Interior allocated an additional $6,968,000 in 2005, $5,115,000 in
2006, and $3,672,000 in 2007 to the Office of Wildland Fire Coordination.
Total allocations do not include carryover from the previous fiscal year.
Table 10 shows BLM's allocations to its 12 state offices and headquarters
for 2005, 2006, and 2007. In 2007, about 71 percent of BLM's total fuel
reduction funding was allocated to WUI areas, while about 29 percent was
allocated to non-WUI areas. In 2005, 2006, and 2007, the Oregon/
Washington state office received the most funding, followed by the Idaho
and Utah state offices. These three offices accounted for about 50 percent
of BLM's total annual fuel reduction allocation.
Table 10: BLM Allocations to State Offices and Headquarters, Fiscal Years
2005, 2006, and 2007
State office Total Percentage of Total WUI Total
Oregon/Washington allocation BLM's total allocation non-WUI
2005 $26,177,000 allocation 27.6 $19,027,000 allocation
$7,150,000
2006 24,596,000 24.1 17,966,000 6,630,000
2007 24,878,000 24.8 18,542,000 6,336,000
Idaho
2005 14,536,000 15.3 10,130,000 4,406,000
2006 14,787,000 14.5 10,033,000 4,754,000
2007 14,598,000 14.6 10,338,000 4,260,000
Utah
2005 8,557,000 9.0 5,479,000 3,078,000
2006 7,968,000 7.8 5,225,000 2,743,000
2007 10,078,000 10.1 6,164,000 3,914,000
California
2005 7,257,000 7.6 6,096,000 1,161,000
2006 6,364,000 6.2 5,382,000 982,000
2007 7,322,000 7.3 6,294,000 1,028,000
Nevada
2005 6,663,000 7.0 4,572,000 2,091,000
2006 5,794,000 5.7 3,881,000 1,913,000
2007 6,414,000 6.4 4,317,000 2,097,000
Colorado
2005 6,480,000 6.8 4,891,000 1,589,000
2006 6,068,000 5.9 4,589,000 1,479,000
2007 6,843,000 6.8 5,285,000 1,558,000
New Mexico
2005 5,676,000 6.0 2,930,000 2,746,000
2006 6,167,000 6.0 3,347,000 2,820,000
2007 6,412,000 6.4 3,630,000 2,782,000
Montana
2005 5,338,000 5.6 4,248,000 1,090,000
2006 4,871,000 4.8 4,000,000 871,000
2007 5,461,000 5.5 4,366,000 1,095,000
Table 11 shows BIA's allocations for 2005, 2006, and 2007 to its 12
regions and the National Interagency Fire Center. In 2007, about 67
percent of BIA's total fuel reduction funding was allocated to WUI areas,
while about 33 percent was allocated to non-WUI areas. In 2005, 2006, and
2007, the
State office Total Percentage of BLM's Total WUI Total non-WUI
Arizona 2005 allocation total allocation allocation allocation
4,219,000 4.4 2,509,000 1,710,000
2006 3,787,000 3.7 2,396,000 1,391,000
2007 4,355,000 4.3 2,608,000 1,747,000
Wyoming
2005 3,143,000 3.3 1,830,000 1,313,000
2006 2,898,000 2.8 1,786,000 1,112,000
2007 3,684,000 3.7 2,185,000 1,499,000
Alaska
2005 785,000 0.8 365,000 420,000
2006 1,044,000 1.0 502,000 542,000
2007 1,556,000 1.6 786,000 770,000
Eastern states
2005 98,000 0.1 78,000 20,000
2006 83,000 0.1 58,000 25,000
2007 126,000 0.1 90,000 36,000
Subtotal, state
offices
2005 88,929,000 93.7 62,155,000 26,774,000
2006 84,427,000 82.6 59,165,000 25,262,000
2007 91,727,000 91.5 64,605,000 27,122,000
Headquarters
2005 5,979,000 6.3 4,527,000 1,452,000
2006 17,822,000 17.4 12,029,000 5,793,000
2007 8,473,000 8.5 6,315,000 2,158,000
Total
2005 94,908,000 100 66,682,000 28,226,000
2006 102,249,000 100 71,194,000 31,055,000
2007 100,200,000 100 70,920,000 29,280,000
Source: GAO analysis of BLM data.
Notes: Total allocations include the allocation for the current year plus
carryover from the previous fiscal year.
Numbers may not total due to rounding.
Northwest region received the most funding of the BIA regions, followed by
the Southwest region. These two regions accounted for about 50 percent of
BIA's total fuel reduction allocation in 2007.
Table 11: BIA Allocations to Regions and the National Interagency Fire
Center, Fiscal Years 2005, 2006, and 2007
Region Total Percentage of BIA's Total WUI Total non-WUI
Northwest allocation total allocation allocation allocation
2005 $12,204,855 28.1 $7,823,617 $4,381,238
2006 11,387,925 26.4 8,012,751 3,375,174
2007 11,835,643 29.6 8,465,745 3,369,898
Southwest
2005 7,651,861 17.6 3,518,105 4,133,756
2006 9,175,694 21.3 4,115,843 5,059,851
2007 8,366,522 21.0 5,128,157 3,238,365
Western
2005 4,497,336 10.4 2,912,773 1,584,563
2006 4,020,742 9.3 2,218,682 1,802,060
2007 3,366,120 8.4 2,287,498 1,078,622
Pacific
2005 3,846,125 8.9 2,863,206 982,919
2006 3,096,619 7.2 2,546,681 549,938
2007 2,401,976 6.0 1,453,604 948,372
Great Plains
2005 2,464,878 5.7 1,461,855 1,003,023
2006 3,196,997 7.4 1,742,168 1,454,829
2007 2,281,299 5.7 1,061,894 1,219,405
Alaska
2005 2,417,117 5.6 2,247,117 170,000
2006 2,256,079 5.2 2,163,458 92,621
2007 1,780,638 4.5 1,538,479 242,159
Midwest
2005 1,876,013 4.3 1,280,813 595,200
2006 2,736,830 6.3 1,988,630 748,200
2007 2,913,975 7.3 2,497,657 416,318
Rocky
Mountain
2005 1,699,576 3.9 1,004,728 694,848
2006 1,769,815 4.1 1,015,550 754,265
Percentage of Total
Region Total BIA's total Total WUI non-WUI
allocation allocation allocation allocation
2007 1,795,609 4.5 1,066,157 729,452
Navajo
2005 1,258,919 2.9 342,421 916,498
2006 1,335,559 3.1 443,588 891,971
2007 933,183 2.3 357,126 576,057
Southern Plains
2005 572,247 1.3 27,552 544,695
2006 568,471 1.3 30,174 538,297
2007 489,663 1.2 21,867 467,796
Eastern
2005 503,175 1.2 425,375 77,800
2006 215,825 0.5 115,732 100,093
2007 433,609 1.1 272,952 160,657
Eastern Oklahoma
2005 284,557 0.7 216,900 67,657
2006 224,465 0.5 189,709 34,756
2007 240,074 0.6 164,172 75,902
Subtotal, regions
2005 39,276,659 90.4 24,124,462 15,152,197
2006 39,985,021 92.7 24,582,966 15,402,055
2007 36,838,311 92.3 24,315,308 12,523,003
National Interagency Fire
Center
2005 4,172,185 9.6 3,621,793 550,392
2006 3,156,566 7.3 2,720,552 436,014
2007 3,092,337 7.7 2,608,684 483,653
Total
2005 43,448,844 100 27,746,255 15,702,589
2006 43,141,587 100 27,303,518 15,838,069
2007 39,930,648 100 26,923,992 13,006,656
Source: GAO analysis of BIA data.
Notes: Total allocations include the allocation for the current year plus
carryover from the previous fiscal year.
Numbers may not total due to rounding.
Table 12 shows NPS's allocations to its seven regions and the Washington
Office for 2005, 2006, and 2007. In 2007, about 46 percent of NPS's total
fuel reduction funding was allocated to WUI areas, while about 54 percent
was allocated to non-WUI areas. NPS was the only Interior agency that
allocated more funds to non-WUI areas than to WUI areas. In 2005, 2006,
and 2007, the Pacific West region received the most funding; followed by
the Intermountain region. These two regions accounted for about 60 percent
of NPS's total annual fuel reduction allocation in 2007.
Table 12: NPS Allocations to Regions and the Washington Office, Fiscal
Years 2005, 2006, and 2007
Region Pacific Total Percentage of Total WUI Total non-WUI
West allocation NPS's total allocation allocation
allocation
2005 $12,550,296 35.5 $7,317,240 $5,233,056
2006 10,478,930 32.4 6,231,822 4,247,108
2007 10,693,592 31.6 6,401,068 4,292,524
Intermountain
2005 10,070,518 28.5 4,268,238 5,802,280
2006 8,578,937 26.5 3,138,962 5,439,975
2007 9,398,600 27.8 4,006,559 5,392,041
Southeast
2005 4,279,340 12.1 1,973,600 2,305,740
2006 4,047,002 12.5 1,648,782 2,398,220
2007 4,604,308 13.6 1,843,531 2,760,777
Midwest
2005 3,104,724 8.8 627,690 2,477,034
2006 3,341,288 10.3 634,129 2,707,159
2007 3,469,731 10.3 644,967 2,824,764
Northeast
2005 898,084 2.5 521,020 377,064
2006 879,070 2.7 406,936 472,134
2007 1,201,497 3.6 757,338 444,159
Alaska
2005 560,582 1.6 0 560,582
2006 813,140 2.5 0 813,140
2007 739,037 2.2 0 739,037
National Capital
2005 142,043 0.4 122,320 19,723
2006 99,851 0.3 99,851 0
2007 103,631 0.3 103,631 0
Region Total Percentage of NPS's total Total WUI Total non-WUI
Subtotal, allocation allocation allocation allocation
regions
2005 31,605,587 89.5 14,830,108 16,775,479
2006 28,238,218 87.4 12,160,482 16,077,736
2007 30,210,396 89.3 13,757,094 16,453,302
Washington
Office
2005 3,705,107 10.5 1,572,455 2,132,652
2006 4,080,417 12.6 1,904,137 2,176,280
2007 3,605,604 10.7 1,804,906 1,800,698
Total
2005 35,310,694 100 $16,402,563 18,908,131
2006 32,318,635 100 14,064,619 18,254,016
2007 33,816,000 100 15,562,000 18,254,000
Source: GAO analysis of NPS data.
Notes: Total allocations include the allocation for the current year plus
carryover from the previous fiscal year.
Numbers may not total due to rounding.
Table 13 shows FWS's allocations in 2005, 2006, and 2007 to its seven
regions, the California-Nevada Operations office, and headquarters.
^[121]2 In 2007, about 61 percent of FWS's total fuel reduction funding
was allocated to WUI areas, while about 39 percent was allocated to
non-WUI areas. In 2005, 2006, and 2007, the Southeast region received the
most funding, followed by the Great Lakes-Big Rivers region.
^2While FWS has seven regions, it has an eighth office--the
California-Nevada Operations Office--that, although officially part of the
Pacific region, manages its own fuel reduction program.
Table 13: FWS Allocations to Regions and Headquarters, Fiscal Years 2005,
2006, and 2007
Region Southeast Total Percentage of Total WUI Total non-WUI
2005 allocation FWS's total allocation allocation
$7,005,484 allocation 24.7 $3,715,115 $3,290,369
2006 7,966,857 23.8 4,780,616 3,186,241
2007 7,543,624 24.2 4,481,330 3,062,294
Great Lakes-Big
Rivers
2005 4,867,717 17.1 2,111,460 2,756,257
2006 5,438,168 16.2 2,658,862 2,779,306
2007 5,336,376 17.1 2,604,636 2,731,740
Mountain-Prairie
2005 3,376,546 11.9 1,281,699 2,094,847
2006 3,776,901 11.3 1,658,254 2,118,647
2007 3,690,940 11.8 1,566,284 2,124,656
Southwest
2005 3,363,824 11.9 2,139,589 1,224,235
2006 3,903,921 11.7 2,499,122 1,404,799
2007 3,721,205 11.9 2,434,533 1,286,672
Pacific
2005 2,556,707 9.0 1,504,681 1,052,026
2006 2,853,522 8.5 1,864,739 988,783
2007 2,566,156 8.2 1,736,153 830,003
Northeast
2005 2,201,297 7.8 1,751,683 449,614
2006 2,597,811 7.8 2,106,088 491,723
2007 2,416,798 7.7 1,977,669 439,129
California-Nevadaa
2005 1,712,138 6.0 986,919 725,219
2006 2,542,027 7.6 1,960,241 581,786
2007 2,254,492 7.2 1,460,866 793,626
Alaska
2005 1,000,439 3.5 815,607 184,832
2006 1,271,882 3.8 1,050,646 221,236
2007 1,224,552 3.9 1,074,688 149,864
Region Total Percentage of FWS's Total WUI Total non-WUI
Subtotal, allocation total allocation 91.9 allocation allocation
regions 2005 26,084,152 14,306,753 11,777,399
2006 30,351,089 90.6 18,578,568 11,772,521
2007 28,754,143 92.1 17,336,159 11,417,984
Washington
Office
2005 2,302,509 8.1 1,571,629 730,880
2006 3,153,074 9.4 2,037,897 1,115,177
2007 2,450,015 7.9 1,723,281 726,734
Total
2005 28,386,661 100 15,878,382 12,508,279
2006 33,504,163 100 20,616,465 12,887,698
2007 31,204,158 100 19,059,440 12,144,718
Source: GAO analysis of FWS data.
Notes: Total allocations include the allocation for the current year plus
carryover from the previous fiscal year.
Numbers may not total due to rounding.
^aWhile FWS has seven regions, it has an eighth office--the
California-Nevada Operations office--that, although officially part of
the Pacific region, manages its own fuel reduction program.
Appendix III: Summary of Fuel Treatment Accomplishments for the Forest Service
and Interior, Fiscal Years 2005 and 2006
The tables in this appendix summarize the fuel reduction accomplishments
of the Forest Service and the four Interior agencies we reviewed--Bureau
of Land Management (BLM), Bureau of Indian Affairs (BIA), National Park
Service (NPS), and Fish and Wildlife Service (FWS)-- for 2005 and 2006,
^[122]1 the most recent years for which complete data were available.
National and regional office data are presented for each agency. ^[123]2
Table 14 provides nationwide information for each of the five agencies,
including total acres treated; acres treated in the wildland-urban
interface (WUI) and in non-WUI areas; and acres treated with prescribed
fire, mechanical methods, and other treatment methods such as herbicides
and grazing. The Forest Service treated more acres than the four Interior
agencies combined--almost 1.7 million acres in 2005 and more than
1.5 million acres in 2006. Within Interior, BLM and FWS treated the most
acres, with BLM treating more than 500,000 acres in 2005 and almost
430,000 acres in 2006, and FWS treating almost 420,000 acres in 2005 and
more than 370,000 acres in 2006. In each year, about 60 percent of the
total acres treated by the five agencies were in the WUI, and about 40
percent of total acres treated were outside of the WUI. The majority of
acres were treated with prescribed fire--almost 75 percent in 2005 and
almost 65 percent in 2006.
^1Years cited in this appendix refer to fiscal years except where
otherwise specified.
^2BIA, FWS, and NPS have regional offices, while BLM has state offices.
For the purposes of this appendix, we refer to all of these as regional
offices when we discuss the Interior agencies collectively.
Table 14: Summary of Fiscal Years 2005 and 2006 Fuel Reduction
Accomplishments for Interior and Forest Service
Acres
Acres treated Acres
using
treated using mechanical treated
using
Agency Treated WUI Non-WUI prescribed means other
acres acres acres fire meansa
BLM
2005 506,168 253,001 253,167 194,553 211,852 99,763
2006 427,912 230,932 196,980 107,443 206,123 114,346
BIA
2005 193,617 71,983 121,634 96,881 94,168 2,568
2006 187,653 89,961 97,692 78,304 106,204 3,145
NPS
2005 153,972 58,873 95,099 139,455 13,036 1,481
2006 116,635 38,558 78,077 102,765 11,532 2,338
FWS
2005 415,646 158,711 256,935 389,686 21,734 4,226
2006 373,933 173,113 200,820 333,038 32,118 8,777
Interior subtotal
2005 1,269,403 542,568 726,835 820,575 340,790 108,038
2006 1,106,133 532,564 573,569 621,550 355,977 128,606
Forest Service
2005 1,672,909 1,198,663 474,246 1,366,988 303,002 2,919
2006 1,503,475 1,090,721 412,754 1,061,277 433,077 9,121
Total Forest Service and
Interior
2005 2,942,312 1,741,231 1,201,081 2,187,563 643,792 110,957
2006 2,609,608 1,623,285 986,323 1,682,827 789,054 137,727
Source: GAO analysis of Interior and Forest Service data.
^a"Other" category includes treatments such as herbicides and grazing.
Table 15 summarizes 2005 and 2006 fuel treatment information for the
Forest Service regions. In both years, the Southern region treated
substantially more acres than the other regions, treating more than half
of the Forest Service's total treated acres. The Forest Service treated
more than twice as many acres in the WUI as in non-WUI areas, and treated
most acres with prescribed fire.
Table 15: Summary of Fiscal Years 2005 and 2006 Fuel Reduction
Accomplishments for Forest Service Regions
Acres Acres Acres
treated treated treated
using using using
Treated Non-WUI prescribed mechanical other
Region Pacific acres WUI acres acres fire means meansa
Southwest 2005 100,540 59,194 41,346 41,565 58,785 190
2006 95,729 62,229 33,500 36,779 50,423 8,527
Southwestern
2005 164,506 69,929 94,577 118,326 46,180 0
2006 180,616 84,973 95,643 134,289 46,327 0
Southern
2005 976,176 803,654 172,522 969,528 6,616 32
2006 776,145 674,189 101,956 625,605 150,540 0
Pacific
Northwest
2005 139,470 79,975 59,495 65,955 73,515 0
2006 133,528 60,904 72,624 73,068 60,432 28
Rocky Mountain
2005 93,969 65,862 28,107 51,353 39,919 2,697
2006 102,953 77,650 25,303 49,313 53,394 246
Intermountain
2005 74,676 34,163 40,513 47,077 27,599 0
2006 87,957 33,995 53,962 58,075 29,562 320
Northern
2005 70,594 43,824 26,770 39,726 30,868 0
2006 68,639 46,892 21,747 43,480 25,159 0
Eastern
2005 51,472 40,556 10,916 33,089 18,383 0
2006 57,221 49,202 8,019 40,242 16,979 0
Alaska
2005 1,506 1,506 0 369 1,137 0
2006 687 687 0 426 261 0
Total
2005 1,672,909 1,198,663 474,246 1,366,988 303,002 2,919
2006 1,503,475 1,090,721 412,754 1,061,277 433,077 9,121
Source: GAO analysis of Forest Service data.
^a"Other" category includes treatments such as herbicides and grazing.
Table 16 summarizes 2005 and 2006 fuel treatment information for the BLM
state offices. In both years, the Oregon/Washington and Idaho state
offices treated the most acres, followed by the New Mexico state office.
BLM treated about the same number of acres in the WUI and in non-WUI areas
in 2005, and treated about 34,000 more acres in the WUI than in non-WUI
areas in 2006. Unlike the Forest Service, BLM treated more acres
mechanically than with prescribed fire, and also treated a substantial
number of acres using other treatment methods, such as herbicides or
grazing.
Table 16: Summary of Fiscal Years 2005 and 2006 Fuel Reduction
Accomplishments for BLM State Offices
Acres
Acres treated Acres
using
treated mechanical treated
using using
State office Treated WUI Non-WUI prescribed means other
acres acres acres fire meansa
Oregon/ Washington
2005 108,909 71,218 37,691 47,273 61,636 0
2006 92,918 69,521 23,397 30,138 62,762 18
Idaho
2005 112,254 54,460 57,794 13,321 46,231 52,702
2006 113,778 69,648 44,130 12,307 47,614 53,857
Utah
2005 40,706 26,616 14,090 6,140 33,966 600
2006 40,535 28,906 11,629 3,900 36,601 34
California
2005 24,191 21,439 2,752 2,342 19,349 2,500
2006 19,389 16,231 3,158 4,187 11,422 3,780
Nevada
2005 28,427 15,190 13,237 10,391 16,272 1,764
2006 35,465 9,655 25,810 15,242 15,014 5,209
Colorado
2005 20,417 13,616 6,801 6,950 13,012 455
2006 17,870 10,132 7,738 4,906 12,774 190
New Mexico
2005 48,107 3,397 44,710 21,297 7,704 19,106
2006 53,329 4,390 48,939 11,682 5,643 36,004
Montana
2005 10,867 6,124 4,743 4,577 5,340 950
2006 12,446 7,530 4,916 5,910 6,426 110
Acres Acres treated Acres
Treated Non-WUI treated using mechanical treated
State acres WUI acres acres using means using
office prescribed other
Arizona fire meansa
2005 35,424 17,078 18,346 17,297 4,391 13,736
2006 19,557 7,424 12,133 5,845 3,625 10,087
Wyoming
2005 30,839 1,976 28,863 19,885 3,004 7,950
2006 18,662 4,507 14,155 9,816 3,789 5,057
Alaska
2005 45,707 21,847 23,860 45,080 627 0
2006 3,963 2,988 975 3,510 453 0
Eastern
states
2005 320 40 280 0 320 0
2006 0 0 0 0 0 0
Total
2005 506,168 253,001 253,167 194,553 211,852 99,763
2006 427,912 230,932 196,980 107,443 206,123 114,346
Source: GAO analysis of Interior data.
^a"Other" category includes treatments such as herbicides and grazing.
Table 17 summarizes 2005 and 2006 fuel treatment information for the BIA
regions. The Northwest and Western regions treated the most acres in 2005,
with each region treating more than 38,000 acres. In 2006, the Northwest
and Southwest regions treated the most acres, with each region treating
more than 45,000 acres. The agency treated more acres in non-WUI areas
than in WUI areas in 2005 and 2006.
Table 17: Summary of Fiscal Years 2005 and 2006 Fuel Reduction
Accomplishments for BIA Regions
Acres Acres treated using Acres
Treated WUI Non-WUI treated mechanical means treated
acres acres acres using using
Region prescribed other
Northwest fire meansa
2005 38,284 17,297 20,987 15,589 21,488 1,207
2006 48,733 25,354 23,379 14,687 32,590 1,456
Southwest
2005 28,212 11,839 16,373 2,678 24,433 1,101
2006 45,132 20,558 24,574 8,096 36,049 987
Western
2005 38,753 16,210 22,543 13,813 24,880 60
2006 22,167 10,360 11,807 7,892 14,275 0
Pacific
2005 2,584 1,817 767 180 2,404 0
2006 4,431 3,179 1,252 331 4,100 0
Great
Plains
2005 14,386 6,986 7,400 7,595 6,591 200
2006 13,234 4,828 8,406 6,217 6,508 509
Alaska
2005 1,253 1,253 0 167 1,086 0
2006 2,222 1,497 725 563 1,659 0
Midwest
2005 21,356 6,478 14,878 17,792 3,564 0
2006 18,559 16,401 2,158 15,585 2,974 0
Rocky
Mountain
2005 11,347 2,856 8,491 6,616 4,731 0
2006 7,400 3,634 3,766 4,177 3,223 0
Navajo
2005 14,274 956 13,318 13,318 956 0
2006 11,065 470 10,595 10,595 470 0
Southern
Plains
2005 12,322 434 11,888 8,401 3,921 0
2006 8,796 672 8,124 4,770 3,833 193
Eastern
2005 7,788 5,616 2,172 7,718 70 0
2006 4,607 2,547 2,060 4,099 508 0
Acres Acres treated using Acres
Treated WUI Non-WUI treated mechanical means treated
Region acres acres acres using using
Eastern prescribed other
Oklahoma fire meansa
2005 3,058 241 2,817 3,014 44 0
2006 1,307 461 846 1,292 15 0
Total
2005 193,617 71,983 121,634 96,881 94,168 2,568
2006 187,653 89,961 97,692 78,304 106,204 3,145
Source: GAO analysis of Interior data.
^a"Other" category includes treatments such as herbicides and grazing.
Table 18 summarizes 2005 and 2006 fuel treatment information for the NPS
regions. In 2005 and 2006, the Southeast region treated the most acres,
followed by the Intermountain region. NPS treated more acres in non-WUI
areas than the WUI, and treated the vast majority of acres (more than 90
percent in 2005 and about 88 percent in 2006) using prescribed fire.
Table 18: Summary of Fiscal Years 2005 and 2006 Fuel Reduction
Accomplishments for NPS Regions
Acres treated Acres Acres
using treated treated
prescribed fire using using
Region Treated WUI Non-WUI 18,922 mechanical other
Pacific West acres acres acres means 5,681 meansa
2005 25,949 9,432 16,517 1,346
2006 22,433 9,042 13,391 16,220 4,101 2,112
Intermountain
2005 43,823 28,585 15,238 38,874 4,844 105
2006 25,350 14,447 10,903 19,397 5,727 226
Southeast
2005 63,602 17,963 45,639 62,491 1,081 30
2006 45,471 9,413 36,058 44,641 830 0
Midwest
2005 20,082 2,433 17,649 18,971 1,111 0
2006 22,872 5,432 17,440 22,150 722 0
Northeast
2005 453 417 36 188 265 0
2006 486 224 262 348 138 0
Alaska
2005 29 9 20 0 29 0
2006 23 0 23 9 14 0
National
Capital
2005 34 34 0 9 25 0
2006 0 0 0 0 0 0
Total
2005 153,972 58,873 95,099 139,455 13,036 1,481
2006 116,635 38,558 78,077 102,765 11,532 2,338
Source: GAO analysis of Interior data.
^a"Other" category includes treatments such as herbicides and grazing.
Table 19 summarizes 2005 and 2006 fuel treatment information for the FWS
regions. In both years, the Southeast region treated substantially more
acres than the other regions--about 35 percent of total acres treated in
2005 and about 30 percent in 2006--followed by the Southwest and Great
Lakes-Big Rivers regions. Like NPS, FWS treated most acres outside of the
WUI, and treated the vast majority of acres (about 94 percent in 2005 and
about 89 percent in 2006) using prescribed fire.
Table 19: Summary of Fiscal Years 2005 and 2006 Fuel Reduction
Accomplishments for FWS Regions
Acres
Acres treated Acres
using
treated mechanical treated
using using
Region Treated WUI Non-WUI prescribed means other
acres acres acres fire meansa
Southeast
2005 144,902 83,218 61,684 141,616 3,202 84
2006 114,212 75,024 39,188 106,864 7,348 0
Great Lakes-Big Rivers
2005 73,550 27,719 45,831 70,880 2,204 466
2006 70,756 27,499 43,257 68,854 1,548 354
Mountain-Prairie
2005 42,252 7,994 34,258 42,032 220 0
2006 39,095 10,100 28,995 38,862 233 0
Southwest
2005 76,495 21,786 54,709 71,820 4,424 251
2006 56,607 17,036 39,571 54,792 1,800 15
Pacific
2005 13,865 6,039 7,826 7,703 6,063 99
2006 23,996 15,704 8,292 11,225 11,683 1,088
Northeast
2005 18,596 6,609 11,987 13,166 2,104 3,326
2006 16,515 8,791 7,724 13,007 1,794 1,714
California-Nevadab
2005 45,216 4,595 40,621 42,176 3,040 0
2006 43,023 18,864 24,159 29,771 7,646 5,606
Alaska
2005 770 751 19 293 477 0
2006 9,729 95 9,634 9,663 66 0
Total
2005 415,646 158,711 256,935 389,686 21,734 4,226
2006 373,933 173,113 200,820 333,038 32,118 8,777
Source: GAO analysis of Interior data.
^a"Other" category includes treatments such as herbicides and grazing.
^bWhile FWS has only seven regions, it has an eighth office--the
California-Nevada Operations office--that, although technically part of
the Pacific region, manages its own fuel reduction program.
Appendix V: GAO Contact and Staff Acknowledgments
GAO Contact
Robin Nazzaro, (202) 512-3841 or [124][email protected]
Staff Acknowledgments
In addition to the individual named above, Steve Gaty, Assistant Director;
Christy Feehan; Rich Johnson; Ches Joy; Amanda Miller; John Mingus;
Lesley Rinner; and Carol Shulman made key contributions to
this report. Elizabeth Curda, Mehrzad Nadji, Jackie Nowicki, and Jena
Sinkfield also made important contributions to the report.
Related GAO Products
Wildland Fire Management: A Cohesive Strategy and Clear Cost-
Containment Goals Are Needed for Federal Agencies to Manage Wildland
Fire Activities Effectively. [125]GAO-07-1017T. Washington, D.C.:
June 19, 2007.
Wildland Fire Management: Lack of a Cohesive Strategy Hinders
Agencies' Cost-Containment Efforts. [126]GAO-07-427T. Washin gton, D.C.:
January 30, 2007.
Wildland Fire Management: Update on Federal Agency Efforts to Develop
a Cohesive Strategy to Address Wildland Fire Threats. [127]GAO-06-671R .
Washington, D.C.: May 1, 2006.
Wildland Fire Management: Timely Identification of Long-Term Options
and Funding Needs Is Critical. [128]GAO-05-923T . Washington, D.C.:
July 14, 2005.
Wildland Fire Management: Forest Service and Interior Need to Specify
Steps and a Schedule for Identifying Long-Term Options and Their
Costs. [129]GAO-05-353T . Washington, D.C.: February 17, 2005.
Wildland Fire Management: Important Progress Has Been Made, but
Challenges Remain to Completing a Cohesive Strategy. [130]GAO-05-147 .
Washington, D.C.: January 14, 2005.
Wildland Fires: Forest Service and BLM Need Better Information and a
Systematic Approach for Assessing the Risks of Environmental Effects.
[131]GAO-04-705 . Washington, D.C.: June 24, 2004.
Wildland Fire Management: Additional Actions Required to Better Identify
and Prioritize Lands Needing Fuels Reduction. [132]GAO-03-805. Washington,
D.C.: August 15, 2003.
Wildland Fire Management: Reducing the Threat of Wildland Fires Requires
Sustained and Coordinated Effort. [133]GAO-02-843T. Washington, D.C.: June
13, 2002.
Severe Wildland Fires: Leadership and Accountability Needed to Reduce
Risks to Communities and Resources. [134]GAO-02-259. Washington, D.C.:
January 31, 2002.
Related GAO Products
The National Fire Plan: Federal Agencies Are Not Organized to Effectively
and Efficiently Implement the Plan. [135]GAO-01-1022T . Washington, D.C.:
July 31, 2001.
Reducing Wildfire Threats: Funds Should Be Targeted to the Highest Risk
Areas. [136]GAO/T-RCED-00-296. Washington, D.C.: September 13, 2000.
Western National Forests: A Cohesive Strategy Is Needed to Address
Catastrophic Wildfire Threats. [137]GAO/RCED-99-65 . Washington, D.C.:
April 2, 1999.
Western National Forests: Catastrophic Wildfires Threaten Resources and
Communities. [138]GAO/T-RCED-98-273 . Washington, D.C.: September 28,
1998.
(360745)
GAO's Mission
The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.
Obtaining Copies of Reports and Testimony
The fastest and easiest way to obtain copies of GAO documents at no cost
[139]is through GAO's Web site (www.gao.gov ). Each weekday, GAO posts GAO
newly released reports, testimony, and correspondence on its
Web site. To have GAO e-mail you a list of newly posted products every
afternoon, go to [140]www.gao.gov and select "Subscribe to Updates."
Order by Mail or Phone
The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:
U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548
To order by Phone: Voice: (202) 512-6000
TDD: (202) 512-2537
Fax: (202) 512-6061
Contact:
To Report Fraud, Waste, and Abuse in Federal Programs
Web site: [141]www.gao.gov/fraudnet/fraudnet.htm
E-mail: [142][email protected]
Automated answering system: (800) 424-5454 or (202) 512-7470
Congressional
Gloria Jarmon, Managing Director, [143][email protected] (202) 512-4400
U.S. Government Accountability Office, 441 G Street NW, Room 7125
Relations Washington, D.C. 20548
Public Affairs
Susan Becker, Acting Manager, [144][email protected] (202) 512-4800
U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548
References
Visible links
124. mailto:[email protected]
125. http://www.gao.gov/cgi-bin/getrpt?GAO-07-1017T
126. http://www.gao.gov/cgi-bin/getrpt?GAO-07-427T
127. http://www.gao.gov/cgi-bin/getrpt?GAO-06-671R
128. http://www.gao.gov/cgi-bin/getrpt?GAO-05-923T
129. http://www.gao.gov/cgi-bin/getrpt?GAO-05-353T
130. http://www.gao.gov/cgi-bin/getrpt?GAO-05-147
131. http://www.gao.gov/cgi-bin/getrpt?GAO-04-705
132. http://www.gao.gov/cgi-bin/getrpt?GAO-03-805
133. http://www.gao.gov/cgi-bin/getrpt?GAO-02-843T
134. http://www.gao.gov/cgi-bin/getrpt?GAO-02-259
135. http://www.gao.gov/cgi-bin/getrpt?GAO-01-1022T
136. http://www.gao.gov/cgi-bin/getrpt?GAO/T-RCED-00-296
137. http://www.gao.gov/cgi-bin/getrpt?GAO/RCED-99-65
138. http://www.gao.gov/cgi-bin/getrpt?GAO/T-RCED-98-273
139. http://www.gao.gov/
140. http://www.gao.gov/
141. http://www.gao.gov/fraudnet/fraudnet.htm
142. mailto:[email protected]
143. mailto:[email protected]
144. mailto:[email protected]
*** End of document. ***