Spatial datasets of probabilistic wildfire risk components for the conterminous United States (270m) for circa 2011 climate and projected future climate circa 2047

Metadata:
Identification_Information:
Citation:
Citation_Information:
Originator: Riley, Karin L.
Originator: Zimmer, Scott N.
Originator: Kodra, Evan
Originator: Grenfell, Isaac C.
Originator: Dillon, Gregory K.
Originator: Scott, Joe H.
Originator: Jaffe, Melissa R.
Originator: Olszewski, Julia H.
Originator: Vogler, Kevin C.
Originator: Finney, Mark A.
Originator: Short, Karen C.
Originator: Jolly, W. Matthew
Originator: Brittain, Stuart E.
Originator: Callahan, Michael N.
Publication_Date: 2025
Title:
Spatial datasets of probabilistic wildfire risk components for the conterminous United States (270m) for circa 2011 climate and projected future climate circa 2047
Geospatial_Data_Presentation_Form: raster digital data
Publication_Information:
Publication_Place: Fort Collins, CO
Publisher: Forest Service Research Data Archive
Other_Citation_Details:
Updated 01 July 2025
Online_Linkage: https://doi.org/10.2737/RDS-2025-0006
Description:
Abstract:
The Large Fire Simulation System (FSim) simulates the growth and behavior of hundreds of thousands of fire events for risk analysis using geospatial data on historical fire occurrence, weather, terrain, and fuel conditions. It can be used to model the frequency and intensity of fires across large spatial and temporal scales. We simulated fire activity in FSim across the conterminous United States with a 2020 landscape (LANDFIRE) and under two sets of climate conditions: 1) using recent climate patterns from 2004-2018 and 2) with modeled future climate conditions for 2040-2054 to address how fire activity may change under future climate. The purpose of this research is to address how climate itself is expected to impact fire activity. Changes in climate will impact the number of days with conditions that promote burning and affect the intensity of burning conditions, which will impact ultimate fire activity and behavior.

The data presented here represent modeled burn probability (BP) and conditional flame length probabilities (FLPs) for the conterminous United States (CONUS) at a 270-meter grid spatial resolution. Flame-length probability is estimated for six standard Fire Intensity Levels (FIL). The six FILs correspond to flame-length classes as follows: FLP1 = < 2 feet (ft); FLP2 = 2 - < 4 ft; FLP3 = 4 - < 6 ft; FLP4 = 6 - < 8 ft; FLP5 = 8 - < 12 ft; FLP6 = 12+ ft. Since they indicate conditional probabilities (i.e., representing the likelihood of burning at a certain intensity level, given that a fire occurs), the FLP data must be used in conjunction with the BP data for risk assessment. All calibration settings and input data used in this analysis, such as vegetation and fuels, were the same as those used in the prior 2020-landscape 2011-climate vintage national FSim run, Dillon et al. (2023) (referred to in the remainder of this document as either the "2020 national run" or the "2011 climate run").

The 2020 national run published here is distinct from the preceding one in two ways: 1) in the previously published version, burnable pixels that did not burn during any simulations were backfilled with the low burn probability of 0.00008 and were assigned flame length probabilities; in this publication we do not alter outputs in this way; and 2) a complete set of flame length and arrival time data are available for all simulated fires in this version. Because of stochasticity in FSim, the two 2020 national runs have minor differences in some areas.

The weather streams used in the circa 2047 climate simulation were updated to a 15-year climate period centered on 2047 using projected shifts in monthly temperature, precipitation, and relative humidity from an ensemble of six General Circulation Models (GCMs) from the Coupled Model Intercomparison Project Phase 5 (CMIP5). Burn probabilities and flame length probabilities from this circa 2047 run were compared to those from the 2020 national FSim run to assess expected changes at the scale of counties and pyromes, or areas of homogeneous fire regime.
Purpose:
National-scale assessment of wildfire risk offers a consistent means of evaluating threats to valued resources and assets, thereby facilitating investments in management activities that can mitigate those risks. We used a simulation system to estimate the probabilistic components of wildfire risk across the nation. These outputs have been generated to support a number of national planning and risk assessment efforts.

Climate-conditioned runs (c2047) were generated to simulate the effect of expected climatic changes on fire activity. These data have direct importance for disaster preparedness planning at a national scale. These data may also address how drivers of fire impact simulated fire activity.
Supplemental_Information:
This data package was originally published on 01/30/2025. These data are a newer version of the Short et al. (2016, 2020) data publications. This modified version is based on circa 2020 landscape data, which were the most current LANDFIRE products available at the time of production. The methods used to generate these data generally followed the same process used in Short et al. (2016, 2020), with improvements made at specific steps. The process steps outlined in the Data Quality, Lineage section of this metadata document are expanded to more fully explain each step and provide additional details on methods for this version of the data. Beyond the newer input landscape data from LANDFIRE, we also used updated datasets for other inputs such as fire occurrence, observed gridded daily weather, and wind data from weather stations. To better capture recent climate conditions, we also shortened the time period of historical weather records used to inform the generation of simulated weather streams for simulation runs, using the most recent 15 years this time (2004-2018) rather than full record from 1992-2012 in the second edition (Short et al. 2020).

The raster files in this package were updated on 07/01/2025. The original rasters included values of 0 outside of the extent of CONUS. We have updated each raster to remove those values. Minor metadata updates were also made on 097/16/2025.
Time_Period_of_Content:
Time_Period_Information:
Range_of_Dates/Times:
Beginning_Date: 2004
Ending_Date: 2054
Currentness_Reference:
Ground condition
Status:
Progress: Complete
Maintenance_and_Update_Frequency: As needed
Spatial_Domain:
Description_of_Geographic_Extent:
The data presented here span the conterminous United States (CONUS).
Bounding_Coordinates:
West_Bounding_Coordinate: -125.00000
East_Bounding_Coordinate: -66.90000
North_Bounding_Coordinate: 49.50000
South_Bounding_Coordinate: 24.50000
Keywords:
Theme:
Theme_Keyword_Thesaurus: ISO 19115 Topic Category
Theme_Keyword: biota
Theme_Keyword: climatologyMeteorologyAtmosphere
Theme_Keyword: environment
Theme:
Theme_Keyword_Thesaurus: National Research & Development Taxonomy
Theme_Keyword: Climate change
Theme_Keyword: Climate change effects
Theme_Keyword: Carbon
Theme_Keyword: Ecology, Ecosystems, & Environment
Theme_Keyword: Fire
Theme_Keyword: Natural Resource Management & Use
Theme:
Theme_Keyword_Thesaurus: None
Theme_Keyword: fire
Theme_Keyword: climate change
Theme_Keyword: burn probability
Theme_Keyword: wildfire risk
Place:
Place_Keyword_Thesaurus: None
Place_Keyword: United States
Place_Keyword: CONUS
Access_Constraints: None
Use_Constraints:
These data were collected using funding from the U.S. Government and can be used without additional permissions or fees. If you use these data in a publication, presentation, or other research product please use the following citation:

Riley, Karin L.; Zimmer, Scott N.; Kodra, Evan; Grenfell, Isaac C.; Dillon, Gregory K.; Scott, Joe H.; Jaffe, Melissa R.; Olszewski, Julia H.; Vogler, Kevin C.; Finney, Mark A.; Short, Karen C.; Jolly, W. Matthew; Brittain, Stuart E.; Callahan, Michael N. 2025. Spatial datasets of probabilistic wildfire risk components for the conterminous United States (270m) for circa 2011 climate and projected future climate circa 2047. Updated 01 July 2025. Fort Collins, CO: Forest Service Research Data Archive. https://doi.org/10.2737/RDS-2025-0006

Users are strongly encouraged to read and fully comprehend the metadata prior to data use. Users should acknowledge the Originator when using this dataset as a source. Users should share data products developed using the source dataset with the Originator. No warranty is made by the Originator as to the accuracy, reliability, or completeness of these data for individual use or aggregate use with other data, or for purposes not intended by the Originator. This dataset is intended to estimate probabilistic wildfire risk components that can support national strategic planning. The applicability of the data to support fire and land management planning on smaller areas will vary by location and specific intended use. Further investigation by local and regional experts should be conducted to inform decisions regarding local applicability. It is the sole responsibility of the local user, using this metadata document and local knowledge, to determine if and/or how these data can be used for particular areas of interest. National FSim products are not intended to replace local products where they exist, but rather serve as a back-up by providing wall-to-wall cross-boundary data coverage. It is the responsibility of the user to be familiar with the value, assumptions, and limitations of these national data publications. Managers and planners must evaluate these data according to the scale and requirements specific to their needs. Spatial information may not meet National Map Accuracy Standards. This information may be updated without notification.
Point_of_Contact:
Contact_Information:
Contact_Organization_Primary:
Contact_Organization: USDA Forest Service, Rocky Mountain Research Station, Missoula Fire Sciences Laboratory
Contact_Person: Karin Riley
Contact_Position: Research Ecologist
Contact_Address:
Address_Type: mailing and physical
Address: 5775 W Broadway St
City: Missoula
State_or_Province: MT
Postal_Code: 59808
Country: USA
Contact_Voice_Telephone: 406-329-4806
Contact_Electronic_Mail_Address: karin.l.riley@usda.gov
Contact Instructions: This contact information was current as of original publication date. For current information see Contact Us page on: https://doi.org/10.2737/RDS.
Data_Set_Credit:
Funding for this project was provided by USDA Forest Service, Fire and Aviation Management and Rocky Mountain Research Station.


Author Information:

Karin L. Riley
USDA Forest Service, Rocky Mountain Research Station
https://orcid.org/0000-0001-6593-5657

Scott N. Zimmer
USDA Forest Service, Rocky Mountain Research Station
https://orcid.org/0009-0000-6389-6826

Evan Kodra
Northeastern University, Civil and Environmental Engineering, Sustainability and Data Sciences Lab

Isaac C. Grenfell
USDA Forest Service, Rocky Mountain Research Station
https://orcid.org/0000-0002-3779-1681

Gregory K. Dillon
USDA Forest Service, Rocky Mountain Research Station
https://orcid.org/0009-0006-6304-650X

Joe H. Scott
Pyrologix, LLC
https://orcid.org/0009-0008-3246-1190

Melissa R. Jaffe
Pyrologix, LLC
https://orcid.org/0009-0002-8623-407X

Julia H. Olszewski
USDA Forest Service, Rocky Mountain Research Station
https://orcid.org/0000-0003-3205-7100

Kevin C. Vogler
Filsinger Energy Partners
https://orcid.org/0000-0002-7080-2557

Mark A. Finney
USDA Forest Service, Rocky Mountain Research Station
https://orcid.org/0000-0002-6584-1754

Karen C. Short
USDA Forest Service, Rocky Mountain Research Station
https://orcid.org/0000-0002-3383-0460

W. Matthew Jolly
USDA Forest Service, Rocky Mountain Research Station
https://orcid.org/0000-0002-0457-6563

Stuart E. Brittain
Alturas Solutions, LLC

Michael N. Callahan
Pyrologix, LLC
https://orcid.org/0009-0009-4937-5405
Cross_Reference:
Citation_Information:
Originator: Short, Karen C.
Originator: Finney, Mark A.
Originator: Scott, Joe H.
Originator: Gilbertson-Day, Julie W.
Originator: Grenfell, Isaac C.
Publication_Date: 2016
Title:
Spatial dataset of probabilistic wildfire risk components for the conterminous United States
Edition: 1st
Geospatial_Data_Presentation_Form: raster digital data
Publication_Information:
Publication_Place: Fort Collins, CO
Publisher: Forest Service Research Data Archive
Online_Linkage: https://doi.org/10.2737/RDS-2016-0034
Cross_Reference:
Citation_Information:
Originator: Short, Karen C.
Originator: Finney, Mark A.
Originator: Vogler, Kevin C.
Originator: Scott, Joe H.
Originator: Gilbertson-Day, Julie W.
Originator: Grenfell, Isaac C.
Publication_Date: 2020
Title:
Spatial datasets of probabilistic wildfire risk components for the United States (270m)
Edition: 2nd
Geospatial_Data_Presentation_Form: raster digital data
Publication_Information:
Publication_Place: Fort Collins, CO
Publisher: Forest Service Research Data Archive
Online_Linkage: https://doi.org/10.2737/RDS-2016-0034-2
Cross_Reference:
Citation_Information:
Originator: Dillon, Gregory K.
Originator: Scott, Joe H.
Originator: Jaffe, Melissa R.
Originator: Olszewski, Julia H.
Originator: Vogler, Kevin C.
Originator: Finney, Mark A.
Originator: Short, Karen C.
Originator: Riley, Karin L.
Originator: Grenfell, Isaac C.
Originator: Jolly, W. Matthew
Originator: Brittain, Stuart E.
Publication_Date: 2023
Title:
Spatial datasets of probabilistic wildfire risk components for the United States (270m)
Edition: 3rd
Geospatial_Data_Presentation_Form: raster digital data
Publication_Information:
Publication_Place: Fort Collins, CO
Publisher: Forest Service Research Data Archive
Online_Linkage: https://doi.org/10.2737/RDS-2016-0034-3
Cross_Reference:
Citation_Information:
Originator: Columbia Climate School, National Center for Disaster Preparedness
Publication_Date: 2025
Title:
U.S. natural hazards climate change projections
Geospatial_Data_Presentation_Form: website
Publication_Information:
Publisher: Columbia Climate School, National Center for Disaster Preparedness
Online_Linkage: https://ncdp.columbia.edu/us-natural-hazards-and-climate-change/
Analytical_Tool:
Analytical_Tool_Description:
FSim is often referred to as a "large fire simulator" because it attempts to model the ignition and growth of only those wildfires with a propensity to spread. Relatively large and generally fast-moving fires are the focus of this system because they account for the majority (~80-97%) of total area burned per simulation unit, etc., and thus contribute the greatest to the probability of a wildland fire burning a given parcel of land therein (i.e., wildfire hazard). Fire occurrence in FSim is stochastically modeled based on historical relationships between large fires and Energy Release Component (ERC), a fire danger index that reflects dryness based on temperature and precipitation over approximately a 40-day period. The size threshold for defining a large fire was calculated for each spatial simulation unit (pyrome) and ranged from 73 acres up to 4,380 acres.

Because its objective is to simulate the behavior of large, spreading fires, FSim only models fire growth on days when the ERC reaches or exceeds the 80th percentile condition, signifying dry fuel conditions. On those days, the length of the active burning period is set at 1 hour, 3 hours, and 5 hours for the 80th, 90th, and 97th percentile ERC conditions, respectively. Fire growth and behavior is calculated using standard FlamMap routines and a minimum travel time (MTT) algorithm. Suppression influences on growth are accounted for by two mechanisms. The first is a statistical model that determines fire duration by indicating probability of containment (cessation) based on spread rates, fuel types, and the length of time a fire has been burning throughout each fire simulation. The second mechanism is a ‘perimeter trimming’ function that simulates the effect of suppression actions on fire progression and shape, resulting in improved fire size distributions from the simulation.

The fire growth simulations, when run over a multitude of individual fire seasons, each with different ignition locations and weather streams, generate burn probabilities and fire behavior distributions at each landscape location (i.e., cell or pixel). Results are objectively evaluated through comparison with historical fire patterns and statistics, including the mean fire size and number of large fires per million acres for each simulation unit. This evaluation is part of the FSim calibration process, whereby simulation inputs are adjusted until the mean fire size and number of large fires per million acres fall within an acceptable range of the historical reference value (i.e., the 70% confidence interval for the mean).

For a technical overview of the Fire Simulation (FSim) system developed by the USDA Forest Service, Missoula Fire Sciences Laboratory to estimate probabilistic components of wildfire risk, see Finney et al. (2011).
Tool_Access_Information:
Tool_Access_Instructions:
Please send requests to: Fire Modeling Institute, USFS Missoula Fire Sciences Laboratory, 5775 US Highway 10 West, Missoula, Montana, 59808; SM.FS.mso_fmi@usda.gov
Tool_Citation:
Citation_Information:
Originator: Finney, Mark A.
Originator: McHugh, Charles W.
Originator: Grenfell, Isaac C.
Originator: Riley, Karin L.
Originator: Short, Karen C.
Publication_Date: 2011
Title:
A simulation of probabilistic wildfire risk components for the continental United States
Geospatial_Data_Presentation_Form: journal article
Series_Information:
Series_Name: Stochastic Environmental Research and Risk Assessment
Issue_Identification: 25(7): 973-1000
Online_Linkage: https://doi.org/10.1007/s00477-011-0462-z
Online_Linkage: https://research.fs.usda.gov/treesearch/39312
Back to Top
Data_Quality_Information:
Attribute_Accuracy:
Attribute_Accuracy_Report:
Model results from the 2020 national FSim run were objectively evaluated through comparison with historical fire patterns and statistics within each pyrome. This evaluation is part of the FSim calibration process, whereby simulation inputs are adjusted until the validation statistics fall within an acceptable range of the historical reference value (±10%). Statistics used as calibration targets were: a) mean annual number of large fires per million burnable acres; and b) mean annual large-fire area burned per million burnable acres. The reference period for calibration targets was the most recent 15 years of records from the Fire Occurrence Database (FOD) (2006-2020) (Short et al. 2022); at the time the simulations were started, this was the years 2004-2018. In addition to the calibration targets, several variables were graphed in each pyrome as visual checks on the number and sizes of fires produced in the simulation. These variables included: historical vs. simulated 15-year cumulative fire size distribution (plotted as fire size against annual fire size exceedance probability), full FOD period (1992-2020) mean annual number of large fires and large-fire area burned, and first 15-years FOD (1992-2006) mean annual number of fires and large-fire area burned. For more information on the calibration process see Thompson et al. (2016, 2022).

Results of the climate-conditioned FSim run cannot be validated in such a way because they represent fire activity under potential future climate conditions. Therefore, we could not calibrate simulation inputs in the same manner. Instead, we used the same simulation inputs and settings as the 2020 national run, with the exception of the weather stream. Published data of observed numbers of large fires and area burned are shown to correlate strongly with shifts in Energy Release Component (ERC); ERC is the fire danger metric that governs the number of ignitions in FSim, as well as affecting fire growth and cessation, therefore we expect modeled increases in area burned and number of large fires in this simulation to be plausible (Riley et al. 2011).

Riley, Karin L.; Abatzoglou, John T.; Grenfell, Isaac C.; Klene, Anna E.; Heinsch, Faith Ann. 2013. The relationship of large fire occurrence with drought and fire danger indices in the western USA, 1984-2008: The role of temporal scale. International Journal of Wildland Fire. 22: 894-909. https://doi.org/10.1071/WF12149 and https://research.fs.usda.gov/treesearch/49353

Short, Karen C. 2022. Spatial wildfire occurrence data for the United States, 1992-2020 [FPA_FOD_20221014]. 6th Edition. Fort Collins, CO: Forest Service Research Data Archive. https://doi.org/10.2737/RDS-2013-0009.6

Thompson, Matthew P.; Bowden, Phil; Brough, April; Scott, Joe H.; Gilbertson-Day, Julie; Taylor, Alan; Anderson, Jennifer; Haas, Jessica R. 2016. Application of wildfire risk assessment results to wildfire response planning in the southern Sierra Nevada, California, USA. Forests. 7(3): 64. https://doi.org/10.3390/f7030064 and https://research.fs.usda.gov/treesearch/50797

Thompson, Matthew P.; Vogler, Kevin C.; Scott, Joe H.; Miller, Carol. 2022. Comparing risk-based fuel treatment prioritization with alternative strategies for enhancing protection and resource management objectives. Fire Ecology. 18: 26. https://doi.org/10.1186/s42408-022-00149-0 and https://research.fs.usda.gov/treesearch/66231
Logical_Consistency_Report:
Pixels with nonzero values for burn probability (BP) have nonzero sum-total values in the FLP layers, and the six FLP layers sum to 1. Pixels with values of zero ("0") for BP have corresponding sum-total zero ("0") values in the FLP layers. Burn probability from this climate-conditioned FSim simulation were compared to burn probability from the 2020 national FSim run to evaluate logical consistency between modeling runs.
Completeness_Report:
We are not aware of any missing data.
Lineage:
Source_Information:
Source_Citation:
Citation_Information:
Originator: Short, Karen C.
Publication_Date: 2022
Title:
Spatial wildfire occurrence data for the United States, 1992-2020 [FPA_FOD_20221014]
Edition: 6th
Geospatial_Data_Presentation_Form: vector digital data
Publication_Information:
Publication_Place: Fort Collins, CO
Publisher: Forest Service Research Data Archive
Online_Linkage: https://doi.org/10.2737/RDS-2013-0009.6
Type_of_Source_Media: Online
Source_Time_Period_of_Content:
Time_Period_Information:
Range_of_Dates/Times:
Beginning_Date: 1992
Ending_Date: 2020
Source_Currentness_Reference:
Ground Condition
Source_Citation_Abbreviation:
FPA_FOD_20221014 (Short et al. 2022)
Source_Contribution:
This historical wildfire dataset was used in development of the FSim products.

The development of the historical fire-occurrence data is described in a companion paper:

Short, Karen C. 2014. A spatial database of wildfires in the United States, 1992-2011. Earth System Science Data. 6: 1-27. https://doi.org/10.5194/essd-6-1-2014 and https://research.fs.usda.gov/treesearch/45689
Source_Information:
Source_Citation:
Citation_Information:
Originator: Short, Karen C.
Originator: Finney, Mark A.
Originator: Vogler, Kevin C.
Originator: Scott, Joe H.
Originator: Gilbertson-Day, Julie W.
Originator: Grenfell, Isaac C.
Publication_Date: 2020
Title:
Spatial datasets of probabilistic wildfire risk components for the United States (270m)
Edition: 2nd
Geospatial_Data_Presentation_Form: raster digital data
Publication_Information:
Publication_Place: Fort Collins, CO
Publisher: Forest Service Research Data Archive
Online_Linkage: https://doi.org/10.2737/RDS-2016-0034-2
Type_of_Source_Media: Online
Source_Time_Period_of_Content:
Time_Period_Information:
Single_Date/Time:
Calendar_Date: 2014
Source_Currentness_Reference:
Ground Condition
Source_Citation_Abbreviation:
Short et al. (2020)
Source_Contribution:
270-m rasters
Source_Information:
Source_Citation:
Citation_Information:
Originator: U.S. Department of Agriculture, Forest Service
Originator: U.S. Department of the Interior
Publication_Date: 2022
Title:
LANDFIRE 2020
Edition: 2.2.0
Geospatial_Data_Presentation_Form: database
Online_Linkage: https://www.landfire.gov/
Type_of_Source_Media: Online
Source_Time_Period_of_Content:
Time_Period_Information:
Multiple_Dates/Times:
Single_Date/Time:
Calendar_Date: 2017
Single_Date/Time:
Calendar_Date: 2018
Single_Date/Time:
Calendar_Date: 2019
Single_Date/Time:
Calendar_Date: 2020
Source_Currentness_Reference:
Ground Condition
Source_Citation_Abbreviation:
LANDFIRE 2020 (LF 2020 - LF_2.2.0)
Source_Contribution:
Multiple files were obtained from this source and are listed below.

1. LANDFIRE 2020 (LF 2020 - LF_2.2.0) fuels data (for 2020) were used to represent surface and canopy fuels in the landscape file created as input to FSim.

Fuels datasets: 40 Scott and Burgan Fire Behavior Fuel Models (FBFM40), Forest Canopy Cover (CC), Forest Canopy Height (CH), Forest Canopy Bulk Density (CBD), Forest Canopy Base Height (CBH)

Scott, Joe H.; Burgan, Robert E. 2005. Standard fire behavior fuel models: a comprehensive set for use with Rothermel's surface fire spread model. Gen. Tech. Rep. RMRS-GTR-153. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 72 p. https://doi.org/10.2737/rmrs-gtr-153


2. LANDFIRE 2020 (LF 2020 - LF_2.2.0) topographic data (for 2020) were included in the landscape file created as input to FSim.

Topographic datasets: Elevation, Aspect, Slope


3. LANDFIRE annual disturbance rasters were used to identify areas that experience wildfires in 2017, 2018, 2019, and 2020. See process step 1.
Source_Information:
Source_Citation:
Citation_Information:
Originator: U.S. Department of Agriculture, Forest Service
Originator: U.S. Department of the Interior
Publication_Date: 2020
Title:
LANDFIRE 2016
Edition: 2.0.0
Geospatial_Data_Presentation_Form: database
Online_Linkage: https://www.landfire.gov/version_download.php
Online_Linkage: https://www.landfire.gov/fuel.php
Type_of_Source_Media: Online
Source_Time_Period_of_Content:
Time_Period_Information:
Single_Date/Time:
Calendar_Date: 2016
Source_Currentness_Reference:
Ground Condition
Source_Citation_Abbreviation:
LANDFIRE 2016 (LF 2016 - LF_2.0.0)
Source_Contribution:
LANDFIRE 2016 (LF 2016 - LF_2.0.0) fuels data (for 2016) were used in areas with significant recent disturbance to represent surface and canopy fuels in the landscape file used during FSim calibration as described in process step 1.
Source_Information:
Source_Citation:
Citation_Information:
Originator: Abatzoglou, John
Publication_Date: 2013
Title:
gridMET
Geospatial_Data_Presentation_Form: NetCDF
Online_Linkage: https://www.climatologylab.org/gridmet.html
Type_of_Source_Media: Online
Source_Time_Period_of_Content:
Time_Period_Information:
Range_of_Dates/Times:
Beginning_Date: 1992
Ending_Date: 2020
Source_Currentness_Reference:
Ground Condition
Source_Citation_Abbreviation:
Abatzoglou (2013)
Source_Contribution:
A gridded daily historical climatology was used to generate values of ERC and dead fuel moisture content for the period of 1992-2020. This provided input data needed to generate synthetic weather streams for the FSim simulations.

Abatzoglou, John T. 2013. Development of gridded surface meteorological data for ecological applications and modeling. International Journal of Climatology. 33: 121-131. https://doi.org/10.1002/joc.3413
Source_Information:
Source_Citation:
Citation_Information:
Originator: Western Regional Climate Center
Publication_Date: Unknown
Title:
RAWS USA Climate Archive
Geospatial_Data_Presentation_Form: tabular digital data
Publication_Information:
Publication_Place: Reno, NV
Publisher: Western Regional Climate Center
Online_Linkage: https://raws.dri.edu/
Type_of_Source_Media: Online
Source_Time_Period_of_Content:
Time_Period_Information:
Range_of_Dates/Times:
Beginning_Date: 1992
Ending_Date: 2020
Source_Currentness_Reference:
Ground Condition
Source_Citation_Abbreviation:
RAWS
Source_Contribution:
Wind data from a selected RAWS station from each pyrome was used to determine distributions of wind speeds and directions for simulations.
Process_Step:
Process_Description:
METHODOLOGY SUMMARY

FSim simulation methods in this analysis were identical to those in the 2020 national FSim run (Dillon et al. 2023). The same landscape files, representing vegetation and topographic characteristics, were used in this analysis. These vegetation characteristics are from the 2020 LANDFIRE data release and represent conditions as of the end of 2020. By using these inputs in the climate-conditioned run, no changes in vegetation types, cover, or heights between 2020 conditions and the future are accounted for. Projecting such vegetation changes has great uncertainty, so we elected to keep these conditions constant to isolate changes in burn probability and intensity resulting directly from climatic changes. Since the same landscape files as the 2020 run were used in this analysis, we also used the same calibration settings to tune simulated fire activity in each pyrome.

The future climate streams in this analysis were produced through a climate conditioning process. Rather than using raw outputs of downscaled global climate models (GCMs) to infer daily weather conditions, we adjusted historical weather time series using projected monthly temperature, relative humidity, and precipitation change factors derived from GCMs. This hybrid approach preserved the original meteorological patterns in current weather conditions, while reflecting projected future climatic changes. Climate change factors were obtained by computing a bias model for each GCM based on a comparison of 1979-2010 NARR and GCM experiments. We utilized the following six GCMs from CMIP5 RCP8.5 scenarios: ACCESS1-3, CanESM2, GFDL-CM3, HadGEM2-CC, Inmcm4, MPI-ESM-LR.

After computing the bias model for each GCM, we performed a Monte Carlo simulation by sampling from each GCM’s bias model to create thousands of bias-corrected projections. From these we obtained statistics on projected climate trends. This approach was performed for every climate variable of interest and for each month, then the median monthly trend of all simulations was calculated for each variable of interest. This represents the ensemble change factor for each climate variable and for each month, and these change factors were then added to the historical input weather stream to create a climate-conditioned weather stream.

We required precipitation duration rather than amount in this analysis, but this was not directly available from GCMs. Instead we developed a linear relationship between historical precipitation amount and duration. Making the assumption that this relationship will remain unchanged in the future, we used this linear model to predict future precipitation duration based on the GCM estimate of future precipitation amount.

Climate change factors were calculated for multiple time steps. In this analysis we utilized a climate stream for the years 2025 to 2054, but we preserved only the final 15 years of this period. Therefore, fire activity in this analysis is representative of the years 2040 to 2054. We refer to this as a 2047-vintage product because this time period is centered on the year 2047.

FSim requires wind speed and direction inputs, but wind data for future timesteps has high uncertainty. Therefore, we used historical wind data in this analysis.

FSim does not utilize daily climatic variables themselves, but rather the Energy Release Component (ERC), an aggregate fire danger index and proxy for fuel moisture content. After the necessary conversion from climatic observations to ERC, we took one further analysis step to synchronize ERC values nationally. Without this step, fire simulation years in FSim would not be coherent nationally. The synchronization process synchronizes ERC values temporally and spatially so that a given year of simulation in FSim has coherent weather patterns nationally, enabling analyses at the national level rather than restricting them to the pyrome level.


Dillon, Gregory K.; Scott, Joe H.; Jaffe, Melissa R.; Olszewski, Julia H.; Vogler, Kevin C.; Finney, Mark A.; Short, Karen C.; Riley, Karin L.; Grenfell, Isaac C.; Jolly, W. Matthew; Brittain, Stuart. 2023. Spatial datasets of probabilistic wildfire risk components for the United States (270m). 3rd Edition. Fort Collins, CO: Forest Service Research Data Archive. https://doi.org/10.2737/RDS-2016-0034-3

Riley, Karin L.; Williams, A. Park; Urbanski, Shawn P.; Calkin, David E.; Short, Karen C.; O’Connor, Christopher D. 2019. Will landscape fire increase in the future? A systems approach to climate, fire, fuel, and human drivers. Current Pollution Reports. 5(2): 9-24. https://doi.org/10.1007/s40726-019-0103-6 and https://research.fs.usda.gov/treesearch/57586
Process_Date: 2024
Process_Step:
Process_Description:
PART A: 2020 NATIONAL FSIM RUN PROCESS STEPS

1. Prepare the landscape file (LCP; fuelscape).

LANDFIRE version 2.2.0 (2020) fuel and topography data were downloaded as full-extent mosaics from landfire.gov. Data were acquired, processed, and stored. The fuel and topography rasters were resampled to 270-meter (m) resolution using the nearest-neighbor resampling method. These resampled 270-m rasters were snapped to the 270-m rasters published in the previous edition of national FSim data products (Short et al. 2020).

In each pyrome, the area included in the landscape file extended 60 kilometers (km) beyond the pyrome boundary to allow for fires to grow unhindered by the edge of the fuelscape, which would otherwise truncate fire growth and affect the simulated fire-size distribution.

Ideally, FSim would be calibrated to contemporary fire occurrence statistics using a fuelscape representing conditions before recent large disturbances, with an updated post-disturbance fuelscape then substituted in only in final production runs. This was not an option on this project because of vegetation and fuel mapping method changes between the previous version (2.0.0; 2016) and most recent version (2.2.0; 2020). However, during FSim calibration runs, analysts noted that significant large wildfire events had occurred in the last 5-10 years in some pyromes, causing simulated burn probability in non-disturbed portions of the pyrome to be too high because the fuel conditions that produced observed fires are no longer present. Given this situation, a decision was made in pyromes with the greatest fraction of area disturbed in the last 5-10 years to substitute pre-disturbance fuels within the footprint of areas burned between 2017 and 2020. LANDFIRE version 2.0.0 (2016) fuels were used those pixels for calibration runs, with version 2.2.0 fuels used everywhere else. Burned areas were identified using LANDFIRE’s individual year disturbance rasters for 2017-2020.

An edit was made to the LANDFIRE 2.2.0 canopy base height (CBH) raster in one vegetation type (Fuel Vegetation Type 2301; Laurentian-Acadian Sub-boreal Aspen-Birch Forest) in Pyromes 96 and 98 (the Superior National Forest area of Minnesota). Within that area, CBH values were multiplied by 0.35 on all pixels mapped with fuel model Timber Understory 2 (TU2). The edit was needed to increase the occurrence of torching and crowning because crown fire behavior is not uncommon here, but only surface fire was simulated with the off-the-shelf CBH raster.
Process_Date: 2022
Process_Step:
Process_Description:
2. Determine large fire size thresholds for each pyrome.

For each pyrome, a large fire size threshold was determined from the FOD records from 1992-2020 by identifying the fire size at which the slope of a Lorenz curve equaled 1 (Lorenz curve represents the cumulative number of fires plotted against the cumulative proportion of area burned). FSim fire-size distributions frequently exhibit a discontinuity at a fire size of four pixels (regardless of pixel resolution), making the results unreliable at that number of pixels and smaller. To avoid that issue, the minimum allowable large-fire size was 73 acres (just larger than four pixels at 270-m resolution). These pyrome-specific large-fire sizes were used for parameterizing FSim (building the logistic regression coefficients and fire-day distribution (FDist) table for the FDist file), establishing calibration targets (mean annual number of large fires and mean annual large-fire area burned), and generating the Ignition Density Grids (rasters representing the historical spatial pattern of large fire ignitions across the pyrome).
Process_Date: 2022
Process_Step:
Process_Description:
3. Create ignition density rasters.

For each pyrome, a 270-m Ignition Density Grid (IDG) was produced using the large-fire size threshold calculated specific to that pyrome from fires occurring from 1992-2020. The IDG is used by FSim to represent the relative historical spatial pattern of large fire ignitions and was created through a multi-step process designed to account for spatially variable ignitable land cover within a moving-window search radius. The IDGs across all pyromes were normalized to have a minimum value of 50 and a maximum value of 1000. In lieu of using an ignition mask in FSim simulations, IDG values in the buffer area around each pyrome were set to zero so fires would not ignite in the buffer during simulations.
Process_Date: 2022
Process_Step:
Process_Description:
4. Process gridMET data and produce ERC at virtual station locations in each pyrome.

Daily values of Energy Release Component for fire danger fuel model G (ERC-G) and dead fuel moisture content for the period 1992-2020 were calculated for a representative location within each pyrome from a gridded historical climatology derived from gridMET data. Corrections for precipitation duration were made to the gridMET data following methods described in Jolly et al. (2019). A representative location (i.e., virtual station) in each pyrome was selected to be the location within the pyrome with the highest density of large-fire ignitions (but not closer than about 10 km from the pyrome boundary). We generated an FW13 file (fire weather file format) with temperature, humidity, and precipitation data for each of these virtual station locations, and then used these FW13 daily observations to generate daily values of ERC and dead fuel moisture content using a custom command-line utility called NFDRScli. Station catalog data for the virtual stations used as input for the NFDRScli tool were obtained from nearby RAWS.

Jolly, W. Matt; Freeborn, Patrick H.; Page, Wesley G.; Butler, Bret W. 2019. Severe Fire Danger Index: A forecastable metric to inform firefighter and community wildfire risk management. Fire. 2: 47. https://doi.org/10.3390/fire2030047 and https://www.fs.usda.gov/research/treesearch/58973
Process_Date: 2021
Process_Step:
Process_Description:
5. Create the FDist file.

The FDist file lists the logistic regression coefficients needed by FSim for estimating daily large-fire ignition probability in relation to ERC-G, and the historical empirical distribution of the number of large fires per large-fire day. We used the BuildFDist command-line utility to generate an FDist file for each pyrome based on each pyrome’s daily historical ERC values and large-fire occurrence during the full 1992-2020 FOD reference period.
Process_Date: 2022
Process_Step:
Process_Description:
6. Generate simulated ERCs and produce SeasonERC file.

Using historical ERC data for the contemporary 15-year period (2004-2018), we generated 20,000 years of simulated daily ERC values, synchronous across CONUS, for all 128 CONUS pyromes. These are not 20,000 years into the future; rather, they are 20,000 possible realizations of a contemporary year based on statistics present in the 2006-2020 observational data. The simulated ERCs were formatted into the required SeasonERC.csv (SERC) format required by FSim.
Process_Date: 2022
Process_Step:
Process_Description:
7. Process wind data from RAWS and integrate with ERC data to produce FRISK file.

The FRISK file contains information required by FSim about three aspects of fire weather. The first section is for Time Series Data and stores historical ERC data. We populated this section with ERC data from the full FOD reference period of 1992-2020 but we did not use these data in the simulations. Instead, we used ERC information contained in the SeasonERC file. The second section of the FRISK is the percentiles section. It lists the ERC and moisture content values for percentiles between 0 and 100%. We populated this section initially using the command-line BuildFRISK utility from the FW13 files created for the selected virtual station locations using the full historical 1992-2020 period. We subsequently used the ModFRISK utility to update the ERC values for each percentile to use simulated ERCs for 1992-2020. We calculated percentiles from the simulated ERC values rather than the historical because the simulated ERCs sometimes had very different exceedance probabilities causing unpredictable results from FSim. The third section of the FRISK consists of tabular distributions of wind speed and direction by month. The FRISK wind distributions were generated with the BuildFRISK utility from the pyrome’s selected RAWS. The following are true about wind records included in creating the FRISK:

• Use as much of the full 1992-2020 period as available (minimum 10 years);
• Use observations from noon to 11pm;
• Use sustained wind speeds;
• Use the Weibull option for wind speed distributions;
• Allow a maximum sustained wind speed of 40 miles/hour.

Although the ERC values used in the simulations (SeasonERC file) were generated from the contemporary 15-year period (2004-2018), the ERC percentiles in the FRISK file are for the full 29-year FOD period (1992-2020). This was done to capture and represent changes in ERC values over time that could cause extreme values of the simulated ERCs to occur more often than the nominal percentile would suggest. For example, under a stationary climate the 80th percentile ERC would be exceeded on 20 percent of the simulation days. Under a changing climate, the 80th percentile could be exceeded on more than 20 percent of the simulation days, which could result in both more fires per year and longer-duration fires.
Process_Date: 2022
Process_Step:
Process_Description:
8. Prepare additional FSim input files.

The ADJ file is an input to FSim that adjusts rates of spread by surface fuel model. We populated the ADJ files initially with baseline values arrived at through many years of running FSim on other projects across the United States. We used a baseline ADJ value of 0.30 in all grass and grass-shrub fuel models, and 0.60 in all shrub, timber litter, timber understory, and slash-blowdown fuel models. These values were then adjusted during calibration runs to help achieve calibration targets.

The FMS file is an input file that allows fuel moisture values from the FRISK file to be overridden. We used the FMS file to enter fixed values of live herbaceous and live woody fuel moisture at the 80th, 90th, and 97th percentile ERC bins. We specified live herbaceous moisture content at 60%, 45%, and 30% at those three ERC levels respectively. We specified live woody moisture content at 110%, 90%, and 70% at those same ERC levels. Dead fuel moisture values still came from the FRISK file.
Process_Date: 2022
Process_Step:
Process_Description:
9. FSim model calibration.

FSim model calibration involved a series of iterative model runs in each pyrome with adjustments to input parameters until there was reasonable agreement with target values derived from fire occurrence data. Calibration targets for this project were, in each pyrome, the mean annual number of large fires per million burnable acres (±10%) and the mean annual large-fire area burned per million burnable acres (±10%), calculated from the last 15 years of the FOD (2006-2020).

For the initial calibration run in each pyrome we set the AcreFract parameter in FSim to 1.0 and used ADJ values of 0.30 for grass and grass-shrub fuel models and 0.60 for all other fuel models. Successive calibration runs were performed with gradually increasing numbers of iterations, ending with about 25-50% of the final number of iterations, until simulated occurrence was “within range” of the historical occurrence. A simulation was considered "within range" if (1) the simulated annual number of large fires was within 10% of the 2006-2020 historical mean, and (2) the simulated annual large-fire area burned was within 10% of the 2006-2020 historical mean. The spatial pattern of simulated BP was then visually checked against known fire perimeters (from the Monitoring Trends in Burn Severity program (MTBS): https://www.mtbs.gov/); and issues were addressed if needed. After calibration of all pyromes, the historical and simulated fire-size distributions were compared to ensure a reasonable fit.

There are no FSim inputs that directly affect mean annual area burned; they instead directly affect the annual number and/or mean size of fires (and consequently the slope of the fire-size distribution), which then affects mean annual area burned. Calibration adjustments were therefore chosen to affect the number and/or sizes of fires. Calibration of an individual pyrome generally followed the following steps:

First run: The base ADJ values (0.30 for grass/grass-shrub and 0.60 for all other fuel models) were used for the initial run. The AcreFract parameter was set to 1.0 for the initial run. Sustained winds for the initially selected RAWS were used for the initial calibration run. The fire size list for this simulation was pasted into a Microsoft Excel calibration workbook to visualize the result and determine inputs for the next run.

Second and subsequent runs: If the simulated mean large-fire size varied from the historical by more than a factor of three, a new RAWS with higher or lower winds was considered. If a new RAWS was not indicated, adjustments to the ADJ values were implemented to bring the simulated mean large-fire size closer to the historical. Standard guidance on subsequent-run ADJ values was developed based on experience on past FSim calibration projects. Additionally, a new value for the AcreFract parameter was calculated to bring the simulated mean annual number of large fires to within 10% of the historical (2006-2020) mean annual number of large fires. Adjustments to the ADJ values and AcreFract continued on successive calibration runs until the simulated occurrence fell within 10% of the historical target. Other FSim parameters that were adjusted when needed included the SuppressionFactor (only in pyrome 91) and the FireDayDistribution in the FDist input file.
Process_Date: 2023
Process_Step:
Process_Description:
10. FSim production runs.

Once each pyrome was considered within range of the calibration targets and had passed other quality assurance and quality control (QA/QC) checks, a final full-iteration FSim run was performed. Final runs used input parameters arrived at through the calibration process and 20,000-100,000 iterations (i.e., potential annual weather scenarios) depending on the number of iterations needed to generate enough simulated fire perimeters across most burnable pixels to calculate probabilities.
Process_Date: 2023
Process_Step:
Process_Description:
11. FSim post-processing and QA/QC.

After final FSim runs were completed in each pyrome, the individual pyrome rasters were mosaicked into products for CONUS. Because each pyrome included a 60 km buffer for simulation and fires were allowed only to start inside the pyrome boundary and burn out (by setting IDG values to no-data in the buffer), mosaics were created by summing BP values from adjacent pyromes. FLPs were mosaicked by first multiplying the conditional FLP by burn probability to obtain the absolute FLP. Then, FLPs of overlapping pyromes were summed and divided by the mosaicked burn probability to calculate the mosaicked conditional FLP.
Process_Date: 2023
Process_Step:
Process_Description:
PART B: CIRCA 2047 CLIMATE RUN PROCESS STEPS

1. Prepare climate-conditioned weather streams

Future climate-conditioned weather streams were produced through a climate conditioning process. We adjusted historical weather time series using projected monthly temperature, relative humidity, and precipitation change factors derived from GCMs. Climate change factors were obtained by computing a bias model for each GCM based on a comparison of 1979-2010 NARR and GCM experiments. We utilized the following six GCMs from CMIP5 RCP8.5 scenarios: ACCESS1-3, CanESM2, GFDL-CM3, HadGEM2-CC, Inmcm4, MPI-ESM-LR.

After computing the bias model for each GCM, we performed a Monte Carlo simulation by sampling from each GCM’s bias model to create thousands of bias-corrected projections. From these we obtained statistics on projected climate trends. This approach was performed for every climate variable of interest and for each month, then the median monthly trend of all simulations was calculated for each variable of interest. This represents the ensemble change factor for each climate variable and for each month, and these change factors were then added to the historical input weather stream to create a climate-conditioned weather stream.

Precipitation duration was produced by developing a linear relationship between historical precipitation amount and duration. Then we used this linear model to predict future precipitation duration based on the GCM estimate of future precipitation amount.

We generated an FW13 file (fire weather file format) with temperature, humidity, and precipitation data, and then used these FW13 daily observations to generate daily values of ERC and dead fuel moisture content using a custom command-line utility called NFDRScli.
Process_Date: 2022
Process_Step:
Process_Description:
2. Generate simulated ERCs and produce SeasonERC file.

Using the climate-conditioned daily ERC values for each pyrome, we generated 20,000 years of simulated daily ERC values, synchronous across CONUS, for all 128 CONUS pyromes. These are not 20,000 years into the future; rather, they are 20,000 possible realizations of climate based on statistics present in the 2040-2054 climate-conditioned weather stream. The simulated ERCs were formatted into the required SeasonERC.csv format required by FSim.
Process_Date: 2024
Process_Step:
Process_Description:
3. FSim production runs.

Utilizing the same calibration settings determined in Part A, final full-iteration FSim runs were performed using the climate-conditioned weather streams. The same number of iterations as in the 2020 national FSim run (20,000-100,000) were performed per pyrome.
Process_Date: 2024
Process_Step:
Process_Description:
4. FSim post-processing and Quality Assurance/Quality Control.

After final FSim runs were completed in each pyrome, the individual pyrome rasters were mosaicked into products for CONUS. Because each pyrome included a 60-km buffer for simulation and fires were allowed only to start inside the pyrome boundary and burn out (by setting IDG values to no-data in the buffer), mosaics were created by summing BP values from adjacent pyromes. FLPs were mosaicked by first multiplying the conditional FLP by burn probability to obtain the absolute FLP. Then, FLPs of adjacent pyromes were summed and divided by the mosaicked burn probability to calculate the mosaicked conditional FLP.
Process_Date: 2024
Process_Step:
Process_Description:
PART C: CHANGE BETWEEN CIRCA 2011 AND CIRCA 2047 CLIMATE RUNS

1. Calculate mean burn probability and fire intensity change.

Pyrome-level mean burn probability change was calculated by masking the burn probability rasters from the circa-2011 climate national run and the circa-2047 climate-conditioned run to each pyrome. Then the mean value of all pixels within the pyrome was calculated for both burn probability rasters to obtain mean burn probability by pyrome. The burn probability percent change for the pyrome was then calculated by subtracting the mean value of the 2011 climate run from the mean value of the 2047 climate run, dividing by the mean value from the 2011 climate run, and multiplying by 100 to convert to percent. Therefore, a 100% change represents a doubling of mean burn probability between the 2011 climate run and 2047 climate-conditioned run.

County-level mean burn probability change was calculated through an identical process as pyrome change.

We also calculated the percent change in low-intensity and high-intensity fire between the 2011 climate run and 2047 climate run at the pyrome and county level. We considered FIL1 (<2 ft) and FIL2 (2-4 ft) as low-intensity fire, and all other flame lengths as high-intensity fire. This distinction was chosen because typically direct firefighting techniques can be used when flame lengths are below 4 ft, while longer flame lengths require indirect firefighting techniques.

Changes in low-intensity and high-intensity fire were assessed by summing the conditional flame length probabilities of the low-intensity FIL layers (FIL1 - FIL2) to calculate total low-intensity probability. Total high-intensity probability was calculated by summing the high-intensity FIL layers (FIL3 - FIL 6). Mean change in low-intensity and high-intensity probability were calculated by a similar process as mean burn probability change described above. Both low-intensity and high-intensity layers were masked to individual pyromes and counties, mean values were calculated, and then mean change was calculated by subtracting the 2011 climate mean values from the 2047 climate mean value.
Process_Date: 2024
Process_Step:
Process_Description:
PACKAGE UPDATE

In the data published on 01/31/2025, the rasters included values of 0 outside of the extent of CONUS. We have updated each raster to remove those values, and the new appropriately masked files are now provided in this package.
Process_Date: 20250701
Back to Top
Spatial_Data_Organization_Information:
Direct_Spatial_Reference_Method: Raster
Raster_Object_Information:
Raster_Object_Type: Pixel
Back to Top
Spatial_Reference_Information:
Horizontal_Coordinate_System_Definition:
Planar:
Map_Projection:
Map_Projection_Name: USA Contiguous Albers Equal Area Conic USGS version
Albers_Conical_Equal_Area:
Standard_Parallel: 29.5
Standard_Parallel: 45.5
Longitude_of_Central_Meridian: -96.0
Latitude_of_Projection_Origin: 23.0
False_Easting: 0
False_Northing: 0
Planar_Coordinate_Information:
Planar_Coordinate_Encoding_Method: Coordinate Pair
Coordinate_Representation:
Abscissa_Resolution: 0.0000000037527980722984474
Ordinate_Resolution: 0.0000000037527980722984474
Planar_Distance_Units: Meters
Geodetic_Model:
Horizontal_Datum_Name: North American Datum of 1983
Ellipsoid_Name: Geodetic Reference System 80
Semi-major_Axis: 6378137.0000
Denominator_of_Flattening_Ratio: 298.25722210
Back to Top
Entity_and_Attribute_Information:
Overview_Description:
Entity_and_Attribute_Overview:
Below you will find a list and description of the files included in this data publication.

VARIABLE DESCRIPTION FILE (1)

1. \Data\_variable_descriptions.csv: Comma-separated values (CSV) file containing a list and description of variables found in all data files. (A description of these variables is also provided in the metadata below.)

Columns include:

Filename = Name of data file
Variable = Name of variable
Units = Units (if applicable)
Precision = Precision (if applicable)
Description = Description of variable



DATA FILES - BURN PROBABILITY (2)

All burn probablity data files are Georeferenced raster TIFF files (GeoTIFFs) and include additional associated files (e.g., *.tif.aux.xml and *.tfw).

1. \Data\2011ClimateRun\BP.tif: 2011 climate annual burn probability (BP) for the conterminous United States.

2. \Data\2047ClimateRun\BP.tif: 2047 climate annual burn probability for the conterminous United States.

Values in the BP data layer indicate, for each pixel, the number of times that cell was burned by an FSim-modeled fire, divided by the total number of annual iterations simulated. The burn probability layer depicts only one component of wildfire risk, indicating the tendency of any given pixel to burn, given the static circa 2020 landscape conditions depicted by the LANDFIRE data, contemporary and projected weather and ignition patterns, as well as contemporary fire management policies (entailing considerable fire prevention and suppression efforts).

The BP data do not, and are not intended to, depict fire-return intervals of any vintage, nor do they indicate projected fire footprints or routes of travel. Nothing about the expected shape or size of any actual fire incident can be interpreted from the burn probabilities. Instead, the BP data, in conjunction with the FLP layers, are intended to support an actuarial approach to quantitative wildfire risk analysis (e.g., see Thompson et al. 2011).



DATA FILES - CONDITIONAL FLAME-LENGTH PROBABILITY (14)

All conditional flame-length probability (FLP) data files are Georeferenced raster files (GeoTIFF) files and include additional associated files (e.g., *.tif.aux.xml and *.tfw).

1-6. \Data\2011ClimateRun\FLP#.tif: 2011 climate conditional flame-length probability (FLP) data for the conterminous United States at a 270-meter grid spatial resolution for the specified Fire Intensity Level (FIL) # = 1-6, which are defined below.

7-14. \Data\2047ClimateRun\FLP#.tif: 2047 climate conditional flame-length probability (FLP) data for the conterminous United States at a 270-meter grid spatial resolution for the specified FIL # = 1-6, which are defined below.

Values in the FLP layers indicate, of all simulated fires that burned a given cell, the proportion in each Flame Length Probability (flame-length class). The six FLPs correspond to flame-length classes as follows:
FLP1 = < 2 feet (ft)
FLP2 = 2 < 4 ft
FLP3 = 4 < 6 ft
FLP4 = 6 < 8 ft
FLP5 = 8 < 12 ft
FLP6 = 12+ ft

The utility of the calibrated FSim BP and FLP data for quantitative geospatial wildfire risk assessment is described in Thompson et al. (2011) and Scott et al. (2013).


DATA FILES - COUNTIES (2)

County data are provided in two formats: CSV file and Esri file geodatabase (GDB).

1. \Data\CountyChange\county_change.csv: Table of U.S. counties with mean and median BP values of the 2011 climate and 2047 climate-conditioned run, mean values of low-intensity and high-intensity conditional FLP, and percent change between these runs. (This same content is available in counties.gdb.)

2. \Data\CountyChange\counties.gdb: GDB containing a feature class (counties_NatFSim2020_CC) of U.S. counties including mean BP and intensity of the 2011 climate (2020-landscape) and projected 2047 climate run. (This same content is available in counties_change.csv.)

Variables in both files include:

NAME = County name

STATE_NAME = State name

STATE_FIPS = State FIPS code

CNTY_FIPS = County FIPS code

FIPS = Full FIPS code

mean_NatFSim2020_BP = Mean annual burn probability in 2020-landscape FSim run

mean_CC_BP = Mean annual burn probability in Climate Conditioned c2047 FSim run

mean_BP_Percent_Change = Percent change in mean annual burn probability between 2020-landscape and Climate Conditioned c2047 FSim run

median_NatFSim2020_BP = Median annual burn probability in 2020-landscape FSim run

median_CC_BP = Median annual burn probability in Climate Conditioned c2047 FSim run

median_BP_Percent_Change = Percent change in median annual burn probability between 2020-landscape and Climate Conditioned c2047 FSim run

FSim2020_mean_low_intensity = Conditional probability of burning in low-intensity fire (flame lengths < 4 feet) in 2020-landscape FSim run

CC_mean_low_intensity = Conditional probability of burning in low-intensity fire (flame lengths < 4 feet) in Climate Conditioned c2047 FSim run

mean_low_intensity_raw_percent_change = Raw percent change in conditional probability of burning in low-intensity fire (flame lengths < 4 feet) between 2020-landscape and Climate Conditioned c2047 FSim run

FSim2020_mean_high_intensity = Conditional probability of burning in high-intensity fire (> 4 feet) in 2020-landscape FSim run

CC_mean_high_intensity = Conditional probability of burning in high-intensity fire (> 4 feet) in Climate Conditioned c2047 FSim run

mean_high_intensity_raw_percent_change = Raw percent change in conditional probability of burning in high-intensity fire (flame lengths > 4 feet) between 2020-landscape and Climate Conditioned c2047 FSim run

OBJECTID = Internal feature number (only available in *.gdb file)

Shape = Feature geometry (only available in *.gdb file)

Shape_Length = Length of feature in internal units (only available in *.gdb file)

Shape_Area = Area of feature in internal units squared (only available in *.gdb file)



DATA FILES - PYROMES (2)

Pyrome data are provided in two formats: CSV file and Esri file geodatabase (GDB).

1. \Data\CountyChange\pyrome_change.csv: Table of pyromes with mean and median BP values of the 2011 climate and 2047 climate-conditioned run, mean values of low-intensity and high-intensity conditional FLP, and percent change between these runs. (This same content is available in pyromes.gdb.)

2. \Data\CountyChange\pyromes.gdb: GDB containing a feature class (pyromes_NatFSim2020_CC) of pyromes, including mean BP and intensity of the 2011 climate (2020-landscape) and projected 2047 climate run. (This same content is available in pyrome_change.csv.)

Variables in both files include:

PYROME = Pyrome number

NAME = Pyrome name

acres = Pyrome size in acres

mean_NatFSim2020_BP = Mean annual burn probability in 2020-landscape FSim run

mean_CC_BP = Mean annual burn probability in Climate Conditioned c2047 FSim run

mean_BP_Percent_Change = Percent change in mean annual burn probability between 2020-landscape and Climate Conditioned c2047 FSim run

median_NatFSim2020_BP = Median annual burn probability in 2020-landscape FSim run

median_CC_BP = Median annual burn probability in Climate Conditioned c2047 FSim run

median_BP_Percent_Change = Percent change in median annual burn probability between 2020-landscape and Climate Conditioned c2047 FSim run

FSim2020_mean_low_intensity = Conditional probability of burning in low-intensity fire (flame lengths < 4 feet) in 2020-landscape FSim run

CC_mean_low_intensity = Conditional probability of burning in low-intensity fire (flame lengths < 4 feet) in Climate Conditioned c2047 FSim run

mean_low_intensity_raw_percent_change = Raw percent change in conditional probability of burning in low-intensity fire (flame lengths < 4 feet) between 2020-landscape and Climate Conditioned c2047 FSim run

FSim2020_mean_high_intensity = Conditional probability of burning in high-intensity fire (> 4 feet) in 2020-landscape FSim run

CC_mean_high_intensity = Conditional probability of burning in high-intensity fire (> 4 feet) in Climate Conditioned c2047 FSim run

mean_high_intensity_raw_percent_change = Raw percent change in conditional probability of burning in high-intensity fire (flame lengths > 4 feet) between 2020-landscape and Climate Conditioned c2047 FSim run

OBJECTID = Internal feature number (only available in *.gdb file)

Shape = Feature geometry (only available in *.gdb file)

Shape_Length = Length of feature in internal units (only available in *.gdb file)

Shape_Area = Area of feature in internal units squared (only available in *.gdb file)
Entity_and_Attribute_Detail_Citation:
Thompson, Matthew P.; Calkin, David E.; Finney, Mark A.; Ager, Alan A.; Gilbertson-Day, Julie W. 2011. Integrated national-scale assessment of wildfire risk to human and ecological values. Stochastic Environmental Research and Risk Assessment 25: 761-780. https://doi.org/10.1007/s00477-011-0461-0 and https://research.fs.usda.gov/treesearch/37465

Scott, Joe H.; Thompson, Matthew P.; Calkin, David E. 2013. A wildfire risk assessment framework for land and resource management. Gen. Tech. Rep. RMRS-GTR-315. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 83 p. https://doi.org/10.2737/rmrs-gtr-315
Back to Top
Distribution_Information:
Distributor:
Contact_Information:
Contact_Organization_Primary:
Contact_Organization: USDA Forest Service, Research and Development
Contact_Position: Research Data Archivist
Contact_Address:
Address_Type: mailing and physical
Address: 240 West Prospect Road
City: Fort Collins
State_or_Province: CO
Postal_Code: 80526
Country: USA
Contact_Voice_Telephone: see Contact Instructions
Contact Instructions: This contact information was current as of July 2025. For current information see Contact Us page on: https://doi.org/10.2737/RDS.
Resource_Description: RDS-2025-0006
Distribution_Liability:
Metadata documents have been reviewed for accuracy and completeness. Unless otherwise stated, all data and related materials are considered to satisfy the quality standards relative to the purpose for which the data were collected. However, neither the author, the Archive, nor any part of the federal government can assure the reliability or suitability of these data for a particular purpose. The act of distribution shall not constitute any such warranty, and no responsibility is assumed for a user's application of these data or related materials.

The metadata, data, or related materials may be updated without notification. If a user believes errors are present in the metadata, data or related materials, please use the information in (1) Identification Information: Point of Contact, (2) Metadata Reference: Metadata Contact, or (3) Distribution Information: Distributor to notify the author or the Archive of the issues.
Standard_Order_Process:
Digital_Form:
Digital_Transfer_Information:
Format_Name: CSV
Format_Version_Number: see Format Specification
Format_Specification:
Comma-separated values file
Digital_Transfer_Option:
Online_Option:
Computer_Contact_Information:
Network_Address:
Network_Resource_Name: https://doi.org/10.2737/RDS-2025-0006
Digital_Form:
Digital_Transfer_Information:
Format_Name: TIFF
Format_Version_Number: see Format Specification
Format_Specification:
Georeferenced raster TIFF file (GeoTIFF)
Digital_Transfer_Option:
Online_Option:
Computer_Contact_Information:
Network_Address:
Network_Resource_Name: https://doi.org/10.2737/RDS-2025-0006
Digital_Form:
Digital_Transfer_Information:
Format_Name: GDB
Format_Version_Number: see Format Specification
Format_Specification:
File geodatabase
Digital_Transfer_Option:
Online_Option:
Computer_Contact_Information:
Network_Address:
Network_Resource_Name: https://doi.org/10.2737/RDS-2025-0006
Fees: None
Back to Top
Metadata_Reference_Information:
Metadata_Date: 20250716
Metadata_Contact:
Contact_Information:
Contact_Organization_Primary:
Contact_Organization: USDA Forest Service, Rocky Mountain Research Station, Missoula Fire Sciences Laboratory
Contact_Person: Karin Riley
Contact_Position: Research Ecologist
Contact_Address:
Address_Type: mailing and physical
Address: 5775 W Broadway St
City: Missoula
State_or_Province: MT
Postal_Code: 59808
Country: USA
Contact_Voice_Telephone: 406-533-5820
Contact_Electronic_Mail_Address: karin.l.riley@usda.gov
Contact Instructions: This contact information was current as of original publication date. For current information see Contact Us page on: https://doi.org/10.2737/RDS.
Metadata_Standard_Name: FGDC Content Standard for Digital Geospatial Metadata
Metadata_Standard_Version: FGDC-STD-001-1998
Back to Top