Data sets used in the analysis presented in the manuscript “Characterizing grassland fire activity in the Flint Hills region and air quality using satellite and routine surface monitor data”. The datasets used for the analysis include data from routine monitor networks located in the central U.S., satellite fire detection data, and burn area estimates. The data supporting each of the Figures in the manuscript are provided in a file specific for that Figure, so there is one file for each Figure. Each file is in comma-separated value (csv) format and contains observation data used to generate that Figure. Meta data on what is included in each file is provided in the Data Dictionary. This dataset is associated with the following publication: Baker, K., S. Koplitz, K. Foley, L. Avey, and A. Hawkins. Characterizing grassland fire activity in the Flint Hills region and air quality using satellite and routine surface monitor data. SCIENCE OF THE TOTAL ENVIRONMENT. Elsevier BV, AMSTERDAM, NETHERLANDS, 659: 1555-1566, (2019).

The Monitoring Trends in Burn Severity MTBS project assesses the frequency, extent, and magnitude (size and severity) of all large wildland fires (includes wildfire, wildland fire use, and prescribed fire) in the conterminous United States (CONUS), Alaska, Hawaii, and Puerto Rico from the beginning of the Landsat Thematic Mapper archive to the present. All fires reported as greater than 1,000 acres in the western U.S. and greater than 500 acres in the eastern U.S. are mapped across all ownerships. MTBS produces a series of geospatial and tabular data for analysis at a range of spatial, temporal, and thematic scales and are intended to meet a variety of information needs that require consistent data about fire effects through space and time. This map layer is a vector point of the location of all currently inventoried and mappable fires occurring between calendar year 1984 and the current MTBS release for CONUS, Alaska, Hawaii and Puerto Rico. Please visit https://mtbs.gov/announcements to determine the current release. Fires omitted from this mapped inventory are those where suitable satellite imagery was not available or fires were not discernable from available imagery. The point location represents the geographic centroid for the _BURN_AREA_BOUNDARY polygon(s) associated with each fire. Metadata

This product ("Prairie fires") presents burned area boundaries for The Flint Hills Ecoregion (KS and OK), one of the most fire prone ecosystems in the United States where hundreds of thousands of acres burn annually as prescribed fire and wildfire. The prairie fire products provide the extent of larger prairie fires in the Flint Hills to record the occurrence of fire and can be used to identify individual burned areas within the perimeters. This product is published to provide fire information of the most fire prone ecosystems to individuals and land management communities for assessing burn extent and impacts on a time sensitive basis. The methods used to produce the prairie fire products from 2019 to present are different than Monitoring Trends in Burn Severity Program (MTBS) methods. The product is developed by running a classification tree model on Landsat and Sentinel imagery for all available image dates with visible fires and without greater than 80 percent cloud cover in the spring of each year. The model takes each image, uses all Landsat bands 2-7 or Sentinel 2b bands 2-4, 8, 11, and 12, and finds thresholds between burnt and unburnt areas to create perimeters. Fire perimeters are created by the model and no manual editing is performed. Thus, these data are 100 percent (model based) auto-generated, however, analysts do review and remove small polygons less than 3 acres. The Prairie Fire dataset will include multi-part polygons and have one record for each source image date. These new methods are optimized to efficiently map and characterize the large number of fires that occur in this region on an annual basis. Prior to 2019, the standard MTBS fire mapping methods were used. Because of the unique frequency and extent of fire in this prairie biome, these fire products are now delivered through the Burn Severity Portal and are no longer included as part of the MTBS products unless a fire is identified in IRWIN, NFPORS or a legacy federal fire occurrence database. The provided data products will vary slightly based on the mapping methodology applied at the time of fire occurrence (pre-2019 or 2019 and later). This map layer is a vector point shapefile of fires occurring three acres and greater in size between calendar year 2009 and 2024 for the Flint Hills Ecoregion.

This product ("Prairie fires") presents burned area boundaries for The Flint Hills Ecoregion (KS and OK), one of the most fire prone ecosystems in the United States where hundreds of thousands of acres burn annually as prescribed fire and wildfire. The prairie fire products provide the extent of larger prairie fires in the Flint Hills to record the occurrence of fire and can be used to identify individual burned areas within the perimeters. This product is published to provide fire information of the most fire prone ecosystems to individuals and land management communities for assessing burn extent and impacts on a time sensitive basis. The methods used to produce the prairie fire products from 2019 to present are different than Monitoring Trends in Burn Severity Program (MTBS) methods. The product is developed by running a classification tree model on Landsat and Sentinel imagery for all available image dates with visible fires and without greater than 80 percent cloud cover in the spring of each year. The model takes each image, uses all Landsat bands 2-7 or Sentinel 2b bands 2-4, 8, 11, and 12, and finds thresholds between burnt and unburnt areas to create perimeters. Fire perimeters are created by the model and no manual editing is performed. Thus, these data are 100 percent (model based) auto-generated, however, analysts do review and remove small polygons less than 3 acres. The Prairie Fire dataset will include multi-part polygons and have one record for each source image date. These new methods are optimized to efficiently map and characterize the large number of fires that occur in this region on an annual basis. Prior to 2019, the standard MTBS fire mapping methods were used. Because of the unique frequency and extent of fire in this prairie biome, these fire products are now delivered through the Burn Severity Portal and are no longer included as part of the MTBS products unless a fire is identified in IRWIN, NFPORS or a legacy federal fire occurrence database. The provided data products will vary slightly based on the mapping methodology applied at the time of fire occurrence (pre-2019 or 2019 and later). This map layer is a vector point shapefile of fires occurring three acres and greater in size between calendar year 2009 and 2024 for the Flint Hills Ecoregion.

This product ("Prairie fires") presents burned area boundaries for The Flint Hills Ecoregion (KS and OK), one of the most fire prone ecosystems in the United States where hundreds of thousands of acres burn annually as prescribed fire and wildfire. The prairie fire products provide the extent of larger prairie fires in the Flint Hills to record the occurrence of fire and can be used to identify individual burned areas within the perimeters. This product is published to provide fire information of the most fire prone ecosystems to individuals and land management communities for assessing burn extent and impacts on a time sensitive basis. The methods used to produce the prairie fire products from 2019 to present are different than Monitoring Trends in Burn Severity Program (MTBS) methods. The product is developed by running a classification tree model on Landsat and Sentinel imagery for all available image dates with visible fires and without greater than 80 percent cloud cover in the spring of each year. The model takes each image, uses all Landsat bands 2-7 or Sentinel 2b bands 2-4, 8, 11, and 12, and finds thresholds between burnt and unburnt areas to create perimeters. Fire perimeters are created by the model and no manual editing is performed. Thus, these data are 100 percent (model based) auto-generated, however, analysts do review and remove small polygons less than 3 acres. The Prairie Fire dataset will include multi-part polygons and have one record for each source image date. These new methods are optimized to efficiently map and characterize the large number of fires that occur in this region on an annual basis. Prior to 2019, the standard MTBS fire mapping methods were used. Because of the unique frequency and extent of fire in this prairie biome, these fire products are now delivered through the Burn Severity Portal and are no longer included as part of the MTBS products unless a fire is identified in IRWIN, NFPORS or a legacy federal fire occurrence database. The provided data products will vary slightly based on the mapping methodology applied at the time of fire occurrence (pre-2019 or 2019 and later). This map layer is a vector point shapefile of fires occurring three acres and greater in size between calendar year 2009 and 2024 for the Flint Hills Ecoregion.

This product ("Prairie fires") presents burned area boundaries for The Flint Hills Ecoregion (KS and OK), one of the most fire prone ecosystems in the United States where hundreds of thousands of acres burn annually as prescribed fire and wildfire. The prairie fire products provide the extent of larger prairie fires in the Flint Hills to record the occurrence of fire and can be used to identify individual burned areas within the perimeters. This product is published to provide fire information of the most fire prone ecosystems to individuals and land management communities for assessing burn extent and impacts on a time sensitive basis. The methods used to produce the prairie fire products from 2009 to present are different than Monitoring Trends in Burn Severity Program (MTBS) methods. The product is developed by running a classification tree model on Landsat and Sentinel imagery for all available image dates with visible fires and without greater than 80 percent cloud cover in the spring of each year. The model takes each image, uses all Landsat bands 2-7 or Sentinel 2b bands 2-4, 8, 11, and 12, and finds thresholds between burnt and unburnt areas to create perimeters. Fire perimeters are created by the model and no manual editing is performed. Thus, these data are 100 percent (model based) auto-generated, however, analysts do review and remove small polygons less than 3 acres. The Prairie Fire dataset will include multi-part polygons and have one record for each source image date. These new methods are optimized to efficiently map and characterize the large number of fires that occur in this region on an annual basis. Prior to 2009, the standard MTBS fire mapping methods were used. Because of the unique frequency and extent of fire in this prairie biome, these fire products are now delivered through the Burn Severity Portal and are no longer included as part of the MTBS products unless a fire is identified in IRWIN, NFPORS or a legacy federal fire occurrence database. The provided data products will vary slightly based on the mapping methodology applied at the time of fire occurrence (pre-2019 or 2019 and later). This map layer is a vector polygon shapefile of fires occurring three acres and greater in size between calendar year 1986 and 2024 for the Flint Hills Ecoregion.

This product ("Prairie fires") presents burned area boundaries for The Flint Hills Ecoregion (KS and OK), one of the most fire prone ecosystems in the United States where hundreds of thousands of acres burn annually as prescribed fire and wildfire. The prairie fire products provide the extent of larger prairie fires in the Flint Hills to record the occurrence of fire and can be used to identify individual burned areas within the perimeters. This product is published to provide fire information of the most fire prone ecosystems to individuals and land management communities for assessing burn extent and impacts on a time sensitive basis. The methods used to produce the prairie fire products from 2009 to present are different than Monitoring Trends in Burn Severity Program (MTBS) methods. The product is developed by running a classification tree model on Landsat and Sentinel imagery for all available image dates with visible fires and without greater than 80 percent cloud cover in the spring of each year. The model takes each image, uses all Landsat bands 2-7 or Sentinel 2b bands 2-4, 8, 11, and 12, and finds thresholds between burnt and unburnt areas to create perimeters. Fire perimeters are created by the model and no manual editing is performed. Thus, these data are 100 percent (model based) auto-generated, however, analysts do review and remove small polygons less than 3 acres. The Prairie Fire dataset will include multi-part polygons and have one record for each source image date. These new methods are optimized to efficiently map and characterize the large number of fires that occur in this region on an annual basis. Prior to 2009, the standard MTBS fire mapping methods were used. Because of the unique frequency and extent of fire in this prairie biome, these fire products are now delivered through the Burn Severity Portal and are no longer included as part of the MTBS products unless a fire is identified in IRWIN, NFPORS or a legacy federal fire occurrence database. The provided data products will vary slightly based on the mapping methodology applied at the time of fire occurrence (pre-2019 or 2019 and later). This map layer is a vector polygon shapefile of fires occurring three acres and greater in size between calendar year 1986 and 2024 for the Flint Hills Ecoregion.

This product ("Prairie fires") presents burned area boundaries for The Flint Hills Ecoregion (KS and OK), one of the most fire prone ecosystems in the United States where hundreds of thousands of acres burn annually as prescribed fire and wildfire. The prairie fire products provide the extent of larger prairie fires in the Flint Hills to record the occurrence of fire and can be used to identify individual burned areas within the perimeters. This product is published to provide fire information of the most fire prone ecosystems to individuals and land management communities for assessing burn extent and impacts on a time sensitive basis. The methods used to produce the prairie fire products from 2009 to present are different than Monitoring Trends in Burn Severity Program (MTBS) methods. The product is developed by running a classification tree model on Landsat and Sentinel imagery for all available image dates with visible fires and without greater than 80 percent cloud cover in the spring of each year. The model takes each image, uses all Landsat bands 2-7 or Sentinel 2b bands 2-4, 8, 11, and 12, and finds thresholds between burnt and unburnt areas to create perimeters. Fire perimeters are created by the model and no manual editing is performed. Thus, these data are 100 percent (model based) auto-generated, however, analysts do review and remove small polygons less than 3 acres. The Prairie Fire dataset will include multi-part polygons and have one record for each source image date. These new methods are optimized to efficiently map and characterize the large number of fires that occur in this region on an annual basis. Prior to 2009, the standard MTBS fire mapping methods were used. Because of the unique frequency and extent of fire in this prairie biome, these fire products are now delivered through the Burn Severity Portal and are no longer included as part of the MTBS products unless a fire is identified in IRWIN, NFPORS or a legacy federal fire occurrence database. The provided data products will vary slightly based on the mapping methodology applied at the time of fire occurrence (pre-2019 or 2019 and later). This map layer is a vector polygon shapefile of fires occurring three acres and greater in size between calendar year 1986 and 2024 for the Flint Hills Ecoregion.

The Monitoring Trends in Burn Severity MTBS project assesses the frequency, extent, and magnitude (size and severity) of all large wildland fires (includes wildfire, wildland fire use, and prescribed fire) in the conterminous United States (CONUS), Alaska, Hawaii, and Puerto Rico for the period of 1984 through 2018. All fires reported as greater than 1,000 acres in the western U.S. and greater than 500 acres in the eastern U.S. are mapped across all ownerships. MTBS produces a series of geospatial and tabular data for analysis at a range of spatial, temporal, and thematic scales and are intended to meet a variety of information needs that require consistent data about fire effects through space and time. This map layer is a vector point of the location of all currently inventoried and mappable fires occurring between calendar year 1984 and 2018 for the continental United States, Alaska, Hawaii and Puerto Rico. The point location represents the geographic centroid for the _BURN_AREA_BOUNDARY polygon(s) associated with each fire. Map Service Feature Layer

This product ("Prairie fires") presents burned area boundaries for The Flint Hills Ecoregion (KS and OK), one of the most fire prone ecosystems in the United States where hundreds of thousands of acres burn annually as prescribed fire and wildfire. The prairie fire products provide the extent of larger prairie fires in the Flint Hills to record the occurrence of fire and can be used to identify individual burned areas within the perimeters. This product is published to provide fire information of the most fire prone ecosystems to individuals and land management communities for assessing burn extent and impacts on a time sensitive basis. The methods used to produce the prairie fire products from 2009 to present are different than Monitoring Trends in Burn Severity Program (MTBS) methods. The product is developed by running a classification tree model on Landsat and Sentinel imagery for all available image dates with visible fires and without greater than 80 percent cloud cover in the spring of each year. The model takes each image, uses all Landsat bands 2-7 or Sentinel 2b bands 2-4, 8, 11, and 12, and finds thresholds between burnt and unburnt areas to create perimeters. Fire perimeters are created by the model and no manual editing is performed. Thus, these data are 100 percent (model based) auto-generated, however, analysts do review and remove small polygons less than 3 acres. The Prairie Fire dataset will include multi-part polygons and have one record for each source image date. These new methods are optimized to efficiently map and characterize the large number of fires that occur in this region on an annual basis. Prior to 2009, the standard MTBS fire mapping methods were used. Because of the unique frequency and extent of fire in this prairie biome, these fire products are now delivered through the Burn Severity Portal and are no longer included as part of the MTBS products unless a fire is identified in IRWIN, NFPORS or a legacy federal fire occurrence database. The provided data products will vary slightly based on the mapping methodology applied at the time of fire occurrence (pre-2009 or 2009 and later). This map layer is a vector polygon shapefile of fires occurring three acres and greater in size between calendar year 1986 and 2025 for the Flint Hills Ecoregion.