Impact of Previous Fuel Treatments on Wildfire Severity Measures in California's Coniferous Forests

Keywords: wildfires,california,remote sensing, land cover, land cover change, disasters
Abstract Type: Virtual Paper Abstract

Abstract

Due to fire suppression and climate change, the annual area burned by wildfire in California has increased by over 500% in the last several decades. Additionally, there has been an increase wildfire severity in the state, likely owing to extreme pre-fire fuel conditions and fire weather. The 21st century, and especially the last decade, has been punctuated by extremely large fires of high severity in coniferous forests in the state. While these ecosystems have traditionally been resilient to fire, large contiguous patches of high severity and extended droughts have left recovery in doubt.

However, antecedent fuel treatments such as prescribed fires or thinning have been known to reduce the effects of subsequent fires that do occur. However, as of now, many studies of the impacts of these treatments on fire severity have focused on single fires. This study will use Google Earth Engine to examine how these previous fuel treatments impact the subsequent of severity of all large wildfires (>100 acres) in coniferous forests in the last several decades in the state.

These wildfire severity measures will be determined using spectral indices, such as RBR or RdNBR, derived from Landsat data. Additionally other possible predictors of subsequent fire severity, such as climate, fuel moisture, previous wildfires, among others, have been examined in order examine the most important predictors of subsequent fire severity in the state. Random forest modeling was used to predict burn severity based on these variables.