Projected Fire Severity (V9a, Classes), 2017-2022, Sierra Nevada, CA

This layer represents modeled composite burn index (CBI, a Landsat-based burn severity metric) projected across the Sierra Nevada region using 2017-2022 ground conditions. CBI was modeled using a random forest regression model (randomForest R package), 8331 sample points from 558 fires spanning 2010-2020, and the following predictors: number of fires over the last 20 years, structure density, topographic wetness index, stand density index, canopy base height, dissection index, insolation index, canopy layers, NDVI, annual precipitation as snow, annual heat-moisture index, summer heat-moisture index, elevation, temperature continentality (30-year normal), annual mean temperature of the coldest month, ladder fuel density, wind exposition index, maximum vapor pressure deficit (30-year normal), time since last fire, soil thickness, soil porosity, slope, relative humidity (30-year normal), precipitation (30-year normal), distance to roads, distance to development, proportion drought-induced forest die-off and unattributed browning over last 5 years and 1 year across 500m radius circular moving window, climatic water deficit (30-year normal), annual climatic water deficit, and canopy bulk density.