Gina M. Gerlich
Published: 2022
Total Pages: 0
Get eBook
Drought and wildfire occurrences are predicted to compound due to global climate change, especially in Mediterranean climates. Therefore, researching potential wildfire determinants is imperative in preparing for and managing future wildfires. The primary goal of this research was to determine if specific environmental, spatial, and human-based variables can explain large wildfire occurrences in Southern California during four designated wildfire regimes, which are drought and post-drought years within the two fire seasons (i.e., dry and Santa Ana (SA) wind fire seasons), between 2012 and 2019 utilizing binary logistic regression models. The secondary goal was to map the predictive patterns of large wildfire occurrences in Southern California. This research used remotely sensed land surface temperature (LST), normalized difference vegetation index (NDVI), and evapotranspiration (ET) datasets. This research also used other raster datasets, such as precipitation, wind, aspect, slope, and digital elevation model (DEM). Various vector derived raster datasets were also used, such as distance to roads, powerlines, cities, and campgrounds, ecoregions, and the wildland-urban interface (WUI). Wildfire occurrences are influenced by anthropogenic, environmental, and spatial factors; however, once ignition occurs and wildfires begin to spread, the environmental factors become more significant in fueling large wildfires. The results indicated that lower NDVI values were the strongest predictor when wildfires were smaller in terms of area burned and when less wildfires occurred. Higher wind speeds were the strongest predictor when wildfires were larger. However, higher LST values were the strongest predictor when wind was not a significant contributor to the model. These conclusions determine that large wildfires are mostly explained by wind, and when wind is not a significant contributor, then LST takes on that role, as these two variables have the ability to dry vegetation and to spread wildfires. This research further establishes the potential for early detections of large wildfires based on wildfire prediction patterns, provides useful information for resource issuance and wildfire management, and enhances general knowledge of the predicted extreme wildfire events in Southern California.