Miguel L. Villarreal Raymond L. LeBeau Sandra Haire James Pierce Todd Hawbaker Emma Sherwood 20260113 raster digital data https://doi.org/10.5066/P14BHA49 The data release contains spatial data sets in raster and vector formats that are used to quantify, analyze and visualize recent fire regimes for the Mogollon Rim-Colorado Plateau (also called the Mogollon Highlands), the Sky Islands, or Madrean Archipelago and the Northern Sierra Madre Occidental regions of the United States and Mexico. The data were generated as part of the Southwest Climate Adaptation Science Center-funded project titled "Assessing Vulnerability of Vegetation and Wildlife Communities to Post-Fire Transformations to Guide Management of Southwestern Pine Forests and Woodlands." Specifically, the data release is comprised of 3 different compressed (zip) files containing the following data: Dataset 1 Fire Perimeters - A fire perimeter (burned area) dataset ranging in years from 1985-2022 that was created by merging fire information from existing wildfire databases available in the United States, and custom Landsat burned area mapping products generated for areas in Mexico. The zip file contains separate polygon files for each ecoregion. Dataset 2 Burn Severity – Fire severity rasters calculated as the differenced Normalized Burn Ratio (dNBR) from Landsat data for each fire perimeter. This dataset contains dNBR images of all individual fires in a zipped folder. Dataset 3 Fire Regimes and Land Use Variables – Fire regime metrics were generated from the above data sets including: 1) number of times burned raster @ 100m, 2) Fire deficit/surplus layers based on the expected number of fires for each vegetation class, 3) maximum fire severity, mean fire severity and fire order-weighted burn severity all @ 30 m, 4) fire order-weighted seasonality (ignition date) surface high resolution, 5) Anthropogenic Biomes (Anthromes) describing biotic community type and population density within a 1km focal window and 6) roadless volume rasters at 100 m resolution. The data were developed as a contribution to a larger effort to quantify, analyze and visualize recent fire regimes across the globe and specifically fill an information gap in the cross-border region that includes the southwestern US and northern Mexico. The datasets were intended for application in studies that link disturbance patterns to ecological processes including dynamics of wildlife habitat; improve understanding of changes through time; and identify connections with human influence and management. 19850101 20221231 publication date Complete None planned North American Cordillera of the southwestern United States and northern Mexico -113.4027 -106.2330 36.0541 26.2138 ISO 19115 Topic Category economy None wildfire burn severity remote sensing fire regime Landsat USGS Thesaurus fires remote sensing USGS Metadata Identifier USGS:6920ef6dd4be025bc609c70e None Arizona New Mexico Sonora, Mexico Chihuahua, Mexico No access constraints. Please see 'Distribution Information' for details. These data are marked with a Creative Common CC0 1.0 Universal License. These data are in the public domain and do not have any use constraints. Users are advised to read the dataset's metadata thoroughly to understand appropriate use and data limitations. Miguel Villarreal USGS - SOUTHWEST REGION Research Geographer mailing and physical

NASA AMES-Moffett Field, Moffett Field - Building 19

Moffett Field CA 94035 650-439-2761 mvillarreal@usgs.gov Python, QGIS, ArcPro and Google Earth Engine Commission errors from poor image quality or sensor artifacts were flagged as errors to be removed from the project. BA polygons that did not exhibit signs of burning were also flagged as errors, in addition to agricultural fires. No formal logical accuracy tests were conducted. Data set is considered complete for the information presented, as described in the abstract. Users are advised to read the rest of the metadata record carefully for additional details. A formal accuracy assessment of the horizontal positional information in the dataset has not been conducted. A formal accuracy assessment of the vertical positional information in the dataset has not been conducted. Boston University NASA MEaSUREs "GLANCE" project 20190627 1 vector digital data https://github.com/measures-glance/glance-grids Digital and/or Hardcopy 20190627 publication date GLANCE GLANCE tiles were used to constrain and process landsat burned area product INEGI 20170101 tabular digital data https://databasin.org/datasets/a5e789aa10fb4efbbd8b7b5a1cf8d423/ Digital and/or Hardcopy 19340101 20170101 publication date Ejidos Used to assign location to fire name Monitoring Trends in Burn Severity 20211231 vector digital data https://www.mtbs.gov/ Digital and/or Hardcopy 19840101 20211231 publication date MTBS Addition of fire perimeters to existing database U.S. Forest Service 20221231 vector digital data https://apps.fs.usda.gov/arcx/rest/services/EDW/EDW_FireOccurrenceAndPerimeter_01/MapServer Digital and/or Hardcopy 20220101 20221231 publication date USFS Addition of fire perimeters to existing database AI and Data for Good at Meta 20190101 raster digital data https://data.humdata.org/organization/meta?q=population%20density&sort=if(gt(last_modified%2Creview_date)%2Clast_modified%2Creview_date Digital and/or Hardcopy 20190101 publication date Population Classified and combined with biome data Open Street Map 2025 tabular digital data https://download.geofabrik.de/north-america/us.html Digital and/or Hardcopy 2025 publication date OSM Roadless volume Federal Highway Administration 2025 2025 vector digital data https://planningdivisiongis-nmdot.hub.arcgis.com/datasets/NMDOT::functional-classification-system-hpms-2025-submittal-of-2024-data/about Digital and/or Hardcopy 2025 publication date HPMS NM Roadless volume calculation National Institute of Statistics and Geography (INEGI) 2024 vector digital data https://www.inegi.org.mx/temas/viascomunicacion/#downloads Digital and/or Hardcopy 2012 2024 publication date INEGI Used for roadless volume Arizona Department of Transportation (ADOT) 2024 2024 vector digital data https://www.arcgis.com/home/item.html?id=c2a321f877984ceabe64dd5762f2087a Digital and/or Hardcopy 2024 publication date ADOT Roadless volume data Instituto Nacional de Estadística y Geografía (INEGI) 2016 2016 vector digital data https://data.humdata.org/dataset/cod-ab-mex Digital and/or Hardcopy 2016 publication date Municipo Used to attribute mexico fire data Dataset 1 Fire Perimeters- Mexico burned area products: We implemented the Landsat Burned Area (BA) algorithm (Hawbaker et al. 2020) for the Northern Sierra Madre Occidental (NSMO) ecoregion to map fires from 1990-2022. This process requires a time series of analysis ready data (ARD) Landsat imagery, which is currently only available for the United States. We therefore adjusted the python code to convert Landsat path/row data into ARD tiles using GLANCE (Global LANd Cover mapping and Estimation) grids (https://measures-glance.github.io/glance-grids/grid_layouts) then applied the CONUS Landsat BA algorithm to the 11GLANCE tiles that covered the study region in Mexico. Burned area rasters were generated for every Landsat scene (1990-2022) within the 11 GLANCE grid tiles. These were inspected for potential commission errors (from clouds or sensor noise) and then converted to a polygon file. The individual BA polygons were then examined on a tile-by-tile basis and edited in QGIS for accuracy. When examining the BA polygon geometries, if the BA algorithm did not capture the full burn perimeter in the latest Landsat image then the BA polygons were manually adjusted in QGIS. This involved manual polygon reshaping and/or merge processes to combine multiple sub-polygons into a single polygon feature. Polygon merging was important in cases where Landsat 7 images exhibited ‘striping’ errors that resulted in fragmented BA polygons, as well as cases where individual fires or fire complexes occurred across multiple image dates. Merge processes were also used to combine related BA polygons (i.e., identical fire data/source) that were separated between two or more GLANCE tiles. Commission errors from poor image quality or sensor artifacts were flagged as errors to be removed from the project. BA polygons that did not exhibit signs of burning were also flagged as errors, in addition to agricultural fires. After BA polygon edits were complete, the polygons were cleaned using the “Delete Holes” algorithm in Qgis that removed any polygon holes and maintained only the external polygon ring and original attributes. The “Merge Vector Layer” algorithm was used to combine the BA polygons for each GLANCE tile into one shapefile. Fires under 400 acres were removed from the data set to be consistent with MTBS fire products. Each BA polygon was assigned a unique “FIRE_ID” field, composed of a location designation and the recorded minimum fire date (YYYYMMDD). In the Sierra Madre Occidental region, the location designation was determined by overlaying the BA polygons onto the Ejidos of Mexico dataset. For example, a fire occurring on May 24, 1994, in Los Ocotes would be labeled as 'LOS_OCOTES_19940524’. If no Ejido designation was available, the FIRE_ID was based on the GLANCE tile name, such as 'SMO_h0044v0048_19960529'. A similar approach was used for the Sky Islands area, where the FIRE_ID was based on the Sky Island name and the minimum fire date. To ensure uniqueness, a sequential suffix was appended to FIRE_IDs that had more than one occurrence (e.g., ACONCHI-19890704-01, ACONCHI-19890704-02). Fire IDs from external sources, such as MTBS or USFS, were merged into the Burned Area dataset while preserving their original IDs. Other regional fire data: Fire perimeters in the Mogollon Plateau region were previously mapped by the Monitoring Trends in Burn Severity (MTBS) program (1984–2021), and this dataset was updated with new fire perimeters from the USFS for 2022. The final regional dataset included 853 fires. The Sky Islands dataset was created by combining the previous data release of 1985-2017 perimeters (Villarreal et al. 2020) with MTBS and 2022 fire perimeters from USFS database as well as any newly identified Landsat burned area (BA) Polygons generated for the Sky Islands in Mexico. The final regional dataset included 448 fires. Hawbaker, T. J., Vanderhoof, M. K., Schmidt, G. L., Beal, Y. J., Picotte, J. J., Takacs, J. D., ... & Dwyer, J. L. (2020). The Landsat Burned Area algorithm and products for the conterminous United States. Remote Sensing of Environment, 244, 111801. Villarreal, M. L., Poitras, T. B., & Conrad, C. R. (2020). Mapped fire perimeters from the Sky Island Mountains of US and Mexico: 1985-2017. US Geological Survey (USGS) Data Release, 714. GLANCE Ejidos MTBS USFS Municipo 2024 Raymond Lebeau USGS - SOUTHWEST REGION Geographer mailing and physical

NASA AMES-Moffett Field, Moffett Field - Building 19

Moffett Field CA 94035 na rlebeau@usgs.gov Dataset 2 Burn Severity – Burn severity was calculated using Google Earth Engine (GEE), employing mean composite fire severity metrics as developed by Parks et al. (2018 and 2021). The burn severity code was updated to accommodate the migration in GEE from Landsat Collection 1 to Landsat Collection 2 data. This update involved applying and mapping new surface reflectance and temperature band scaling factors across the image collections before processing. To quantify fire severity, the differenced Normalized Burn Ratio (dNBR) was calculated for each fire perimeter. The dNBR was computed using the equation: dNBR= (preNBR−postNBR) ×1000 where preNBR is the pre-fire Normalized Burn Ratio and postNBR is the post-fire Normalized Burn Ratio. We calculated the dNBR for all mapped fire perimeters in the 3 ecoregions. This approach ensured accurate and consistent burn severity measurements across the entire dataset, providing insight into fire impacts within the studied regions. Note - there are 6 fires in the perimeter data Parks, S. A., Holsinger, L. M., Voss, M. A., Loehman, R. A., & Robinson, N. P. (2018). Mean composite fire severity metrics computed with Google Earth Engine offer improved accuracy and expanded mapping potential. Remote Sensing, 10(6), 879. Parks, S. A., Holsinger, L. M., Voss, M. A., Loehman, R. A., & Robinson, N. P. (2021). Correction: Parks et al. Mean composite fire severity metrics computed with Google Earth Engine offer improved accuracy and expanded mapping potential. Remote Sensing. 13: 4580. https://doi. org/10.3390/rs13224580., 13, 4580. 20241201 Raymond Lebeau USGS - SOUTHWEST REGION Geographer mailing and physical

NASA AMES-Moffett Field, Moffett Field - Building 19

Moffett Field CA 94035 na rlebeau@usgs.gov Dataset 3 Fire Regimes and Land Use Variables – 1. Fire regime variable: Fire count. Created by summarizing overlapping fires. 2. Fire regime variable. Fire deficit/surplus layers. Fire count data were joined with a vegetation/biome map (Rehfeldt et al. 2017) to derive an attribute of times burned for 9 major vegetation community types reclassified from the original 19 Rehfeldt classes. These include: 3 - Thornscrub (Sinaloa Dry Deciduous Forest and Pacific Coast Thornscrub), 6- Chaparral (California Chaparral and Interior Chaparral), 8- Pinon Juniper Woodland (Great Basin Conifer Woodland), 10- Pine-oak Forests (10 Madrean Montane Conifer Forest), 33 -Madrean Evergreen Forests and Woodlands (Madrean-Transvolcanic Woodland and San Lucas Pine-Oak Woodland), 35- Desertscrub (Sonoran Desertscrub, Great Basin Desertscrub and Chihuahuan Desertscrub), 40- Grassland (Semidesert Grassland, Great Basin Shrub-Grassland and Plains Grassland), 41 Spruce Fir (Rocky Mountain Subalpine Conifer Forest ) and 43- Mixed Conifer Forest (Rocky Mountain Montane Conifer Forest). The Mean Fire Interval (MFI) was calculated as the expected number of fires in the period of record from Landsat satellite burn data (37 years for Plateau and Sky Islands and 32 years for the Sierra Madre Occidental). Expected number of fires was determined for each vegetation type based on data from the paleoecological (tree-ring) literature, please see Table 1 in Villarreal et al. (2020) for more information on vegetation-specific fire regime characteristics. Note that Pinon Juniper Woodland and Spruce Fir classes had fire regimes divided into 2 different MFI classes - Frequent Low Severity Fire and Infrequent High Severity Fire. These were differentiated by identifying pixels of these two classes that contained dNBR of >440 and assigned to the high severity fire regime classes. We then classified the vegetation/fire count rasters into five classes: 1) Fire deficit -unburned 2) Fire deficit - burned 3) Expected fire regime - unburned 4) Expected fire regime 5) Fire surplus 3. Fire regime variable: Burn severity surfaces. Fire order-weighted burn severity surfaces were calculated for overlapping dNRB rasters using the R code provided by Steel et al. 2021 (https://github.com/zacksteel/pyrodiversity/blob/main/code/season_surface.R). A decay rate of 0.5 is used for earlier fires. Between [0,1); a rate of .5 means each subsequent value receives 1/2 of the previous. We also calculated, mean, maximum (of any burn), and maximum recent burn composites of overlapping dNBRs. 4. Fire regime variable: Fire order-weighted seasonality. This was calculated using the R code provided by Steel et al. 2021 (https://github.com/zacksteel/pyrodiversity/blob/main/code/season_surface.R). Input data were the fire perimeter data which include fire year and julian date in the attribute table. A decay rate of 0.5 is used for earlier fires. Between [0,1); a rate of .5 means each subsequent value receives 1/2 of the previous 5. Land use variable: Anthromes. Anthromes, also known as anthropogenic biomes or human biomes, are global ecological patterns shaped by sustained direct human interactions with ecosystems (Ellis and Ramankutty 2008). They represent the terrestrial biosphere in its contemporary, human-altered form, acknowledging that humans have fundamentally transformed ecosystems. Anthromes encompass various land uses and land covers, including dense settlements, villages, croplands, rangelands, and semi-natural lands. Global anthromes data sets are ~4km pixel size (Ellis and Ramankutty 2008), which are not an appropriate resolution for regional-local scale investigations. For this proejct we aimed to generate a high resolution (1km) anthromes dataset that is specific to the biotic communities of the North American Cordillera of the southwestern United States and northern Mexico study areas. This was accomplished by combining high resolution 30m population data (Tiecke et al. 2017) with 1km resolution vegetation community maps (Rehfeldt et al. 2012). The processing steps are as follows: First combined 30 m tif of USA population data from 2 zones that covered the study area with 30 m Mexico data. Executed a moving window focal (sum) on the 30m population data with a 33x33 window to approximate average population in 1 km2 area, then classified population summaries into 3 classes based on the anthromes (Ellis and Ramankutty 2008) population density cutoffs: Remote, 0-1 person/km2 Populated, 1-10 person/km2 Residential, >10 person/km2 The classified population data were then combined with a 18 class vegetation/biome map (Rehfeldt et al. 2017) to create the 23-class anthromes raster. Note that several of the desert and grassland vegetation classes were combined into a single "Rangeland" class that includes Sonoran Desertscrub, Mohave Desertscrub, Western Alpine Tundra, Great Basin Desertscrub, Chihuahuan Desertscrub, Semidesert Grassland, Great Basin Shrub-Grassland and Plains Grassland. Likewise, Pacific Coast Thornscrub, California Chaparral and Interior Chaparral were combined into Chaparral & Thornscrub class. 6. Land use variable: Roadless Volume (Emma Sherwood (emma.sherwood@nrcan-rncan.gc.ca) and Marc-André Parisien authors) steps: Road data for AZ, NM and Mexico were accessed from Open Street Map, using files named gis_osm_roads_free_1.shp. All files were merged. Feature classes are then filtered for roads: NOT IN ('bridleway', 'cycleway', 'footway', 'living_street', 'path', 'pedestrian', 'steps', 'track', 'track_grade1', 'track_grade2', 'track_grade3', 'track_grade4', 'track_grade5') Other road sources include NM: (https://planningdivisiongis-nmdot.hub.arcgis.com/datasets/0e0d859a98b3495f9e1ff404a479d6cc_0/explore?location=32.398696%2C-108.743102%2C13.71), Mexico (https://en.www.inegi.org.mx/temas/viascomunicacion/#downloads) and AZ (https://azgeo-open-data-agic.hub.arcgis.com/datasets/c2a321f877984ceabe64dd5762f2087a_0/explore?location=33.988529%2C-112.338846%2C11.52) All road data are then merged into a single file and selected for only features within 100km of study region. Run Euclidean Distance tool with the roads layer as the source data, with output cell size as 100, and geodesic distance method. Multiply the raster by the cell area & convert units. In this case, the distance to the nearest road is in meters, and the cells are 100m x 100m, and divided by 100,000 to do the volume & conversion in the same step (using 'Divide' tool) resulting in km3. Focal statistics with a circular neighborhood with a 56-cell radius (which makes as close to 100km^2 moving window as possible with 100m cells). Mask to study area (no data values have been set as -999). References: Ellis, E. C. and N. Ramankutty. 2008. Putting people in the map: anthropogenic biomes of the world. Frontiers in Ecology and the Environment 6(8):439-447. Rehfeldt, G. E., Crookston, N. L., Sáenz-Romero, C., & Campbell, E. M. (2012). North American vegetation model for land‐use planning in a changing climate: A solution to large classification problems. Ecological Applications, 22(1), 119-141. Steel, Z. L. , B. M. Collins, D. B. Sapsis, and S. L. Stephens, “Quantifying pyrodiversity and its drivers,” Proceedings of the Royal Society B: Biological Sciences, vol. 288, no. 1948, Apr. 2021. https://github.com/zacksteel/pyrodiversity Tiecke, T. G., Liu, X., Zhang, A., Gros, A., Li, N., Yetman, G., ... & Dang, H. A. H. (2017). Mapping the world population one building at a time. arXiv preprint arXiv:1712.05839. Villarreal, M. L., Iniguez, J. M., Flesch, A. D., Sanderlin, J. S., Cortés Montaño, C., Conrad, C. R., & Haire, S. L. (2020). Contemporary fire regimes provide a critical perspective on restoration needs in the Mexico-United States borderlands. Air, Soil and Water Research, 13, 1178622120969191. Population OSM HPMS NM INEGI ADOT 20250701 Miguel Villarreal USGS - SOUTHWEST REGION Research Geographer mailing and physical

NASA AMES-Moffett Field, Moffett Field - Building 19

Moffett Field CA 94035 650-439-2761 mvillarreal@usgs.gov Raster Pixel various various 0.0197844184 0.0229902536 Decimal seconds WGS_1984 WGS 84 6378137.0 298.257223563 Anthromes_Rehfeldt_1km.tif Raster geospatial data file. Producer Defined OID Internal object identifier. Producer Defined Sequential unique whole numbers that are automatically generated. Value Anthrome class number Producer Defined 0 803 Class numbers Count Number of raster cells with this value. Producer Defined 2.0 348764.0 Pixels Name Anthrome name Producer Defined No data No data value Producer defined Remote Rangeland Rangeland with 0-1 person/km2 Producer defined Populated Rangeland Rangeland with 1-10 person/km2 Producer defined Residential Rangeland Rangeland with >10 person/km2 Producer defined Remote Chaparral & Thornscrub Chaparral & Thornscrub with 0-1 person/km2 Producer defined Populated Chaparral & Thornscrub Chaparral & Thornscrub with 1-10 person/km2 Producer defined Residential Chaparral & Thornscrub Chaparral & Thornscrub with >10 person/km2 Producer defined Remote Sinaloa Dry Deciduous Forest Sinaloa Dry Deciduous Forest with 0-1 person/km2 Producer defined Populated Sinaloa Dry Deciduous Forest Sinaloa Dry Deciduous Forest with 1-10 person/km2 Producer defined Residential Sinaloa Dry Deciduous Forest Sinaloa Dry Deciduous Forest with >10 person/km2 Producer defined Remote Great Basin Conifer Woodland Great Basin Conifer Woodland with 0-1 person/km2 Producer defined Populated Great Basin Conifer Woodland Great Basin Conifer Woodland with 1-10 person/km2 Producer defined Residential Great Basin Conifer Woodland Great Basin Conifer Woodland with >10 person/km2 Producer defined Remote Madrean Montane Conifer Forest Madrean Montane Conifer Forest with 0-1 person/km2 Producer defined Populated Madrean Montane Conifer Forest Madrean Montane Conifer Forest with 1-10 person/km2 Producer defined Residential Madrean Montane Conifer Forest Madrean Montane Conifer Forest with >10 person/km2 Producer defined Remote Madrean-Transvolcanic Woodland Madrean-Transvolcanic Woodland with 0-1 person/km2 Producer defined Populated Madrean-Transvolcanic Woodland Madrean-Transvolcanic Woodland with 1-10 person/km2 Producer defined Residential Madrean-Transvolcanic Woodland Madrean-Transvolcanic Woodland with >10 person/km2 Producer defined Remote Rocky Mountain Subalpine Conifer Forest Rocky Mountain Subalpine Conifer Forest with 0-1 person/km2 Producer defined Populated Rocky Mountain Subalpine Conifer Forest Rocky Mountain Subalpine Conifer Forest with 1-10 person/km2 Producer defined Remote Rocky Mountain Montane Conifer Forest Rocky Mountain Montane Conifer Forest with 0-1 person/km2 Producer defined Populated Rocky Mountain Montane Conifer Forest Rocky Mountain Montane Conifer Forest with 1-10 person/km2 Producer defined Residential Rocky Mountain Montane Conifer Forest Rocky Mountain Montane Conifer Forest with >10 person/km2 Producer defined RV_100.tif Roadless volume Producer Defined Value Volume of roads (km3) in surrounding area Producer Defined 22.058307647705078 751.1968383789062 km3 Plateau_LargeFires_WGS84_zone12_Final_v3.shp Attribute Table Table containing attribute information associated with the fire perimeter data set. Producer Defined FID Internal feature number. ESRI Sequential unique whole numbers that are automatically generated. Shape Feature geometry. ESRI Shape type. Event_ID Unique Fire ID assigned from source data Producer Defined Unique Fire ID assigned from source data Map_Prog Original source of fire perimeter information Producer Defined MTBS Data from Monitoring Trends in Burn Severity (MTBS) database Producer defined USFS Data from National USFS Final Fire Perimeter database Producer defined Fire_ID Unique Fire ID obtained from source data and modified to append fire date information (YYYYMMDD) Producer Defined Unique Fire ID obtained from source data and modified to append fire date information (YYYYMMDD) Year The year the fire occurred. Producer Defined 1984 2022 Year Country Country where fire occurred. Producer Defined USA USA Producer defined State State where fire occurred. Producer Defined Arizona Arizona Producer defined New Mexico New Mexico Producer defined Ecoregion Ecoregion where fire occurred. Producer Defined Plateau Mogollon Rim-Colorado Plateau Producer defined Hectares Total area in acres of each fire in hectares. Producer Defined 402.82 228106.841 Hectares Fire_Date Date of the start of each fire. YYYY/MM/DD Producer Defined 19840529 20220703 YYYY/MM/DD Area Total area in meters. Producer Defined 4028230.59 2281068406.0 Meters Perimeter Fire perimeter Producer Defined 8254.797 594276.56 Meters Julian_Day Julan day of fire detection Producer Defined 4 350 Julian day SkyIsland_LargeFires_1985_to_2022_Final_v3.shp Attribute Table Table containing attribute information associated with the fire perimeter data set. Producer Defined FID Internal feature number. ESRI Sequential unique whole numbers that are automatically generated. Shape Feature geometry. ESRI Shape type. Event_ID Unique Fire ID assigned from source data Producer Defined Unique Fire ID assigned from source data Map_Prog Original source of fire perimeter information Producer Defined << empty cell >> No info Producer defined USFS Data from National USFS Final Fire Perimeter database Producer defined MTBS Data from Monitoring Trends in Burn Severity (MTBS) database Producer defined USGS Perimeter developed in house by USGS Producer defined Fire_ID Unique Fire ID obtained from source data or if in Mexico, modified to append fire date information (YYYYMMDD) to sky island name Producer Defined Unique Fire ID obtained from source data or if in Mexico, modified to append fire date information (YYYYMMDD) to sky island name Year The year the fire occurred. Producer Defined 1985 2022 Year Country Country where fire occurred. Producer Defined USA USA Producer defined Mexico Mexico Producer defined State State where fire occurred. Producer Defined New Mexico New Mexico Producer defined Arizona Arizona Producer defined Chihuahua Chihuahua Producer defined Sonora Sonora Producer defined Ecoregion Ecoregion where fire occurred. Producer Defined Sky Island Sky Islands, aka Madrean Archipelago Producer defined Hectares Total area in acres of each fire in hectares. Producer Defined 400.909 91490.908 Hectares Fire_Date Date of the start of each fire. YYYY/MM/DD Producer Defined 19850704 20220614 YYYY/MM/DD Area Total area in meters Producer Defined 4009086.714 914909077.5 Meters Perimeter Fire perimeter size Producer Defined 8476.634 306571.232 Meters Complex_na Name of the Sky Island Complex Producer Defined Name of the Sky Island Complex Julian_Day Julan day of fire detection Producer Defined 13 325 Julain Day seasonal_surface_decay05_.tif Raster geospatial data file. fire order-weighted seasonality Producer Defined Value Data represent the weighted average julian date of fire(s), older fires received a decay rate of 0.5.Data represent the weighted average julian date of fire(s), older fires received a decay rate of 0.5. Producer Defined 4.0 350.0 seasonal_surface_recentfire_.tif Raster geospatial data file. Data represent the julian date of the most recent fire. There is one raster for each ecoregion. Producer Defined Value Data represent the julian date of the most recent fire. Producer Defined 4.0 350.0 Julian date dNBR.tif This is a generic attribute description for 2278 of the differenced Normalized Burn Ratio (dNBR) rasters that were calculated for each individual fire in the study region during the study period. Producer Defined Value dNBR value contained in each raster cell. Higher values generally correspond to higher severity fire. Producer Defined -706 1125 dNBR FDS_ecoregion_clip.tif Raster geospatial data file. Producer Defined OID Internal object identifier. Producer Defined Sequential unique whole numbers that are automatically generated. Value Combination of original vegetation class (first 1-2 digits) + high severity (value of 10 or 0) + fire count . Example is vegetation class 41 Spruce-Fir: 4102 is Spuce Fir, low severity with 2 burns. 4112 is Spuce Fir, high severity with 2 burns Producer Defined 600 4316 Count Number of raster cells with this value. Producer Defined 2.0 2661958.0 Pixels FDS_value 1) Fire deficit -unburned 2) Fire deficit - burned 3) Expected fire regime - unburned 4) Expected fire regime 5) Fire surplus Producer Defined 3 Expected fire regime - unburned Producer defined 4 Expected fire regime Producer defined 5 Fire surplus Producer defined 1 Fire deficit -unburned Producer defined 2 Fire deficit - burned Producer defined Plateau_LargeFires_WGS84_zone12_Final_v3.shp Attribute Table Table containing attribute information associated with the data set. Producer Defined FID Internal feature number. ESRI Sequential unique whole numbers that are automatically generated. Shape Feature geometry. ESRI Shape type. Event_ID Unique Fire ID assigned from source data Producer Defined Unique Fire ID assigned from source data Map_Prog Original source of fire perimeter information Producer Defined MTBS Data from Monitoring Trends in Burn Severity (MTBS) database Producer defined USFS Data from National USFS Final Fire Perimeter database Producer defined Fire_ID Unique Fire ID obtained from source data and modified to append fire date information (YYYYMMDD) Producer Defined Unique Fire ID obtained from source data and modified to append fire date information (YYYYMMDD) Year The year the fire occurred. Producer Defined 1984 2022 year Country The country of origin. Producer Defined USA USA Producer defined State State of origin. Producer Defined Arizona Arizona Producer defined New Mexico New Mexico Producer defined Ecoregion Ecoregion where fire occurred. Producer Defined Plateau Mogollon Rim-Colorado Plateau (also called the Mogollon Highlands) Producer defined Hectares Total area in acres of each fire in hectares. Producer Defined 402.82 228106.841 Hectare Fire_Date Date of the start of each fire. YYYY/MM/DD Producer Defined 19840529 20220703 YYYY/MM/DD Area Area of fire in meters Producer Defined 4028230.59 2281068406.0 meters Perimeter Fire perimeter in meters Producer Defined 8254.797 594276.56 meters Julian_Day Julian day Producer Defined 4 350 Julian day SMO_LargeFires_WGS84_zone12_Final_v3.shp Attribute Table Table containing attribute information associated with the data set. Producer Defined FID Internal feature number. ESRI Sequential unique whole numbers that are automatically generated. Shape Feature geometry. ESRI Shape type. Year Year of the fire. Producer Defined 1990 2022 Year Map_from Code that references the GLANCE ARD tile and Landsat Path Row of the primary image the fire was mapped from Producer Defined Code that references the GLANCE ARD tile and Landsat Path Row of the primary image the fire was mapped from Country Country were fire occurred. Producer Defined Mexico Mexico Producer defined Fire_Date Date of the fire, YYYYMMDD Producer Defined 19900404.0 20220606.0 YYYYMMDD State Mexican state where fire occurred Producer Defined Chihuahua Chihuahua Producer defined Sonora Sonora Producer defined Sinaloa Sinaloa Producer defined Durango Durango Producer defined Ecoregion Ecoregion where fire occured Producer Defined SMO Northern Sierra Madre Occidental Producer defined Fire_ID Unique Fire ID obtained from source data and modified to append fire date information (YYYYMMDD) Producer Defined Unique Fire ID obtained from source data and modified to append fire date information (YYYYMMDD) Perimeter Fire perimeter in meters Producer Defined 8557.9 450438.76 Meters Hectares Fire area in hectares Producer Defined 399.576 36916.933 Hectares Area Fire area in meters Producer Defined 3995762.165 369169328.562 Meters Municipali Municipality in Mexico, used to assign unique Fire_ID Producer Defined Municipality in Mexico, used to assign unique Fire_ID Julian_Day Julain Day of the fire start Producer Defined 11 255 Julain day dNBR_max_.tif Raster geospatial data file. Producer Defined Value Maximum dNBR value (if overlapping fires) contained in each raster cell. Higher values generally correspond to higher severity fire. Producer Defined -284.0 960.0 dNBR dNBR_mean.tif Raster geospatial data file. Producer Defined Value Mean dNBR value (if overlapping fires) contained in each raster cell. Higher values generally correspond to higher severity fire. Producer Defined -284.0 831.0 dNBR dNBR_order_weighted_.tif Raster geospatial data file. Producer Defined Value Fire order weighted dNBR value (if overlapping fires) contained in each raster cell. Higher values generally correspond to higher severity fire. Decay value of 0.5 was used for older fires. Producer Defined 0.0 831.0 dNBR List of data included: Dataset 1 Fire Perimeters- 1.Burned Area/Fire Perimeter, 3 shapefiles one for each region. Dataset 2 Burn Severity – 2. Burn severity - individual dNBRs calculated for each fire perimeter, in zipped folder. Dataset 3 Fire Regimes and Land Use Variables – 3. Fire count, 3 rasters one for each region. 4. Fire Deficit/Surplus layers, 3 rasters one for each region. 5. Burn severity surface - 1990-2022 composites (mean, max, recent max, fire ordered weighted) 6. Fire order-weighted seasonality (ignition date) surface 7. Anthropogenic Biomes (Anthromes) 8. Roadless Volume na U.S. Geological Survey - ScienceBase mailing address

Denver Federal Center

Building 810

Mail Stop 302

Denver CO 80225 1-888-275-8747 sciencebase@usgs.gov Unless otherwise stated, all data, metadata and related materials are considered to satisfy the quality standards relative to the purpose for which the data were collected. Although these data and associated metadata have been reviewed for accuracy and completeness and approved for release by the U.S. Geological Survey (USGS), no warranty expressed or implied is made regarding the display or utility of the data for other purposes, nor on all computer systems, nor shall the act of distribution constitute any such warranty. Digital Data https://doi.org/10.5066/P14BHA49 None 20260113 Miguel Villarreal USGS - SOUTHWEST REGION Research Geographer mailing and physical

NASA AMES-Moffett Field, Moffett Field - Building 19

Moffett Field CA 94035 650-439-2761 mvillarreal@usgs.gov FGDC Biological Data Profile of the Content Standard for Digital Geospatial Metadata FGDC-STD-001.1-1999