Wildfire Vulnerability Modeling and Vegetation Cover Change in Butte: An Analysis Based on Boosted Regression Tree

In recent years, California has experienced increasingly severe wildfire events, leading to substantial socio-economic losses and ecological degradation. Against this backdrop, this study aims to identify key indicators influencing forest fire occurrence, assess forest fire vulnerability, and examine vegetation cover changes in Butte County, California. First, we analyze and visualize 14 wildfire-relevant environmental and anthropogenic factors, capturing climatic, topographic, and land-use characteristics of the study area. To address multicollinearity among variables, the Variance Inflation Factor (VIF) is employed, resulting in the selection of 11 non-collinear indicators. Based on these selected variables, a Boosted Regression Tree (BRT) model is applied to evaluate spatial patterns of wildfire vulnerability in Butte County. Finally, we employ Vegetation Fractional Cover (VFC) to quantify post-fire vegetation cover changes, enabling an assessment of wildfire impacts on vegetation dynamics. The results indicate that rainfall, land use, and topographic conditions exert significant influences on wildfire vulnerability in Butte County. Moreover, VFC analysis reveals a notable decline in vegetation cover surrounding fire locations between July 2024 and September 2024, highlighting the short-term ecological impacts of recent wildfire events.