Mapping areas at elevated risk of large-scale structure loss using Monte Carlo simulation and wildland fire modeling

This work presents, and demonstrates through application to California, a data-driven methodology that can be used to identify areas at elevated risk of experiencing wildland fires capable of causing large-scale structure loss. A 2D Eulerian level set fire spread model is used as the computational engine for Monte Carlo simulation with ignition points placed randomly across the landscape. For each randomly-placed ignition point, wind and weather conditions are also selected randomly from a 10-year climatology that has been developed by others using the Weather Research and Forecasting (WRF) mesoscale weather model at a resolution of 2km. Fuel and topography inputs are obtained from LANDFIRE. Housing density is estimated from 2010 Census block data. For each randomly-selected combination of ignition location and wind/weather, fire progression is modeled so that fire area and number of impacted structures can be recorded. This is repeated for over 100 million discrete ignition points across California to generate “heat maps” of fire probability, fire consequence, and fire risk. In this work, fire volume (spatial integral of burned area and flame length) is used as a proxy for fire probability since quickly spreading fires with large flame lengths are most likely to escape initial attack and become extended attack fires. Fire consequence is taken as the number of impacted structures. Fire risk is then estimated as the product of probability and consequence. The methodology is assessed comparing the resultant fire risk raster with perimeters from California's 20 most damaging fires as tabulated by the California Department of Forestry and Fire Protection (CALFIRE). It is found that these historical perimeters from damaging fires correlate well with areas identified as high risk in the Monte Carlo simulation, suggesting that this methodology may be capable of identifying areas where similarly damaging fires may occur in the future.