Large Wildfire Potential, Santa Monica Mountains Region, CA

This dataset is a 30-m cell size raster representing predicted large wildfire potential based on historical (1980-2010) conditions. This model used 10 explanatory variables: topographic heterogeneity, Santa Ana wind speed, elevation, topographic position, distance to electric transmission lines, distance to roads, distance to development, minimum temperature, climatic water defecit, and southwest index. This model had a mean 10-fold cross-validated test AUC of 0.74, overall accuracy of 0.86, and TSS of 0.71.

All models are abstractions of reality that are created for particular purposes and should be used with caution. These outputs map the potential for large wildfire based on statistical modeling and available landscape-scale and longer-term data. They should be used solely to evaluate broad, general patterns of relative fire risk across the landscape and inform management decisions. They should not be used to predict the behavior of individual fires or to evaluate risk to structures at the parcel scale.