Wildfire scenario projections were provided by Dr. LeRoy Westerling at the University of California Merced, using a statistical model based on historical data of climate, vegetation, population density, and fire history coupled with regionally downscaled LOCA climate projections


The fire modeling ran simulations on following five variables on a monthly time step:


The modeling uses the downscaled LOCA climate projections as inputs and therefore is considered as secondary scenarios in the Fourth Assessment. The GCMs and RCPs used are the “priority” scenarios recommended for California’s Fourth Climate Change Assessment. Details are described in Westerling et al., forthcoming [4th Assessment report or white paper].

Population and vegetation projections were developed by external collaborators of the Fourth Assessment. California Department of Finance (DOF) projections of county-level population at five-year increments through 2060 were extended by Ethan Sharygin of DOF to 2100 with three trajectories — Central, Low, and High scenario. These population projections were used to drive a land use change simulation model (LUCAS) by the USGS. The land use/land cover scenarios represent changes in a suite of classes of land use and land cover related to urbanization, agricultural expansion and contraction, forest harvest, wildfire, and other processes. Development was simulated based on the DOF population projections, whereas all other land use changes were based on historical data. The LUCAS scenarios generated a set of 10 Monte Carlo simulations at 1 km spatial resolution and one-year time steps for each DOF population projection, which was converted into proportion of the 1/16º grid cells that were vegetated (i.e., burnable wildland fuel). Details of the population and land use scenarios are described in Sleeter et al., forthcoming (journal article in review at Earth's Future).

List of Global Climate Model (GCM), Representative Concentration Pathway (RCP) and Population/Vegetation (Cond):


For each combination of GCM, RCP and Cond, 100 simulations were run with the stochastic modeling framework. Area burned was aggregated by year and grid cell. If the burned area is larger than the average amount of vegetation in a grid cell, burned area was allocated to the surrounding grid cells. Thus 24 wildfire scenarios were generated (4 GCMs * 2 RCPs * 3 Cond), with 1000 simulations each (10 land use simulation * 100 fire simulations). The mean area burned was also calculated for all 10 stochastic variations for each GCM/RCP/condition combination.

The fire severity modeling estimated the fraction of burned area in various severity classes, which is defined by the proportion of basal area removed by fire. Managers were also interested in the potential effect on fire severity of the massive tree mortality experienced by California’s conifer forests during the recent historic drought. A sensitivity analysis was conducted that varied the historical probability of high severity fire (greater than 90% of basal area of vegetation removed by fire) in cells with greater than 50% tree mortality based on mapping data from the US Forest Service. The sensitivity analysis was only performed in the Sierra Nevada region because of lack of data from conifer forests in other forested regions and because wildfire in non-forest vegetation is typically high severity (e.g., grassland and chaparral).


Spatial Resolution and Extent

The spatial resolution of this data is 1/16º (approximately 6 km). The spatial extent available on Cal-Adapt for area burned and fire severity (forthcoming) covers the entire state of California. The fire severity/tree mortality sensitivity analysis projections (forthcoming) are limited to the Sierra Nevada region.

Temporal Resolution

Download Data

For large areas or full spatial extent, it is more efficient to download data in NetCDF or GeoTIFF formats and clip to desired spatial extent. For smaller areas, additional options for downloading data in CSV or JSON formats are available through download tool below.


Other Formats

Spatial Extent
Climate Variable
Population Growth Projections

Cal-Adapt API