Burn Severity Mapping with L-band UAVSAR Observations Over Los Angeles’ Largest Wildfire

Recent increases in fire season length and frequency of high-severity fires necessitate predicting and monitoring fire impacts. This paper focuses on fire burn severity, a parameter that is critical for estimating post-fire carbon fluxes, air pollution, and debris flows. Traditionally, burn severity is evaluated through in-situ measurements and optical imagery, but this study presents a new method of mapping burn severity using Synthetic Aperture Radar (SAR) polarimetric images from the L-band airborne sensor, UAVSAR. This study used a log-ratio of the cross-polarization channel (HV) to measure changes in radar backscatter and defined three severity categories for the 2017 La Tuna Fire in the Verdugo Mountains in Los Angeles. The SAR log ratio was compared with the Landsat-derived dNBR product and results suggest that SAR and optical products can offer complementary views of burn severity.