Title:

California fire severity prediction maps by region

Author(s):

Drury, Stacy A.; Benoit, John W.; Fleming, Sean P.; Harris, Lucas B.; Taylor, Alan H.

Publication Year:

2026

How to Cite:

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:

Drury, Stacy A.; Benoit, John W.; Fleming, Sean P.; Harris, Lucas B.; Taylor, Alan H. 2026. California fire severity prediction maps by region. Fort Collins, CO: Forest Service Research Data Archive. https://doi.org/10.2737/RDS-2026-0034

Abstract:

This data publication contains a spatial database of potential fire severity raster datasets for California, using data from 1984 to 2022. The data collection and raster production was conducted as part of the California Fire Severity Prediction Mapping Project which models fire severity across California by region: Klamath, North Coast, Northeast, North Sierra, East Sierra, South Sierra, South Coast and Central Coast. The California Fire Severity Project uses the Random Forest modeling platform to create two empirically based fire severity prediction (forecast) models for every pixel in each of the eight regions (16 total models) that are based on spatial, temporal, and environmental information. The non-reburn model produces fire severity potential estimates for areas that have not burned since 1984 while the reburn model estimates potential fire severity for landscapes that have burned at least once since 1984. Potential Fire Severity model output is classified into one of three classes based on the liklihood that a pixel within the area of interest will burn at low, moderate, or high severity when burned by future wildfire. Initially two maps are produced for each region, an initial fire severity potential map covers non-reburn landscapes (no fire since 1984), and a reburn fire severity potential map (areas have been burned at least once since 1984). The initial and reburn potential fire severity maps are merged into a single raster based map layer referred to as combined potential fire severity. This data package includes a geodatabase for each region in California, and within each geodatabase there are separate potential fire severity raster layers that for the initial non-reburn areas, reburned areas, and both the non-reburn and reburned areas combined. This package also includes the R scripts and Google Earth Engine scripts used to produce the fire severity rasters.

Keywords:

biota; environment; Fire; Fire ecology; Fire effects on environment; Natural Resource Management & Use; Forest management; fire severity; fire ecology; historical wildfire; California

Related publications:

• Drury, Stacy A.; Benoit, John W.; Fleming, Sean P.; Harris, Lucas B.; Taylor, Alan H. Unknown. California fire severity prediction mapping project. Journal for Fire Ecology. [In review].
• Taylor, Alan H.; Harris, Lucas B.; Drury, Stacy A. 2021. Drivers of fire severity shift as landscapes transition to an active fire regime, Klamath Mountains, USA. Ecosphere. 12(9): e03734. https://doi.org/10.1002/ecs2.3734 https://research.fs.usda.gov/treesearch/63135

Metrics:

Visit count : 63 0
Download count: 5
More details

Data Access:

• View metadata (HTML)
• View file index (HTML), which lists all files in this data publication and short description of their contents
• Download all files below for the complete publication:
  • RDS-2026-0034_Metadata_Fileindex.zip (28.33 KB; Checksum)
  • RDS-2026-0034.zip (184.76 MB; Checksum)

Need information about Using our Formats?