United States
Wildfires in California have become increasingly frequent and severe, driven by climate change, vegetation dynamics, and human activity. The recent simultaneous eruption of multiple distinct wildfires concentrated in the Los Angeles area - such as the Palisades, Eaton, and Hughes fires - led to widespread evacuations and significant property damage, highlighting the urgent need for improved wildfire prevention strategies. To mitigate such risks, data-driven approaches are essential for identifying high-risk areas before ignition occurs. This study aims to develop a wildfire risk prediction and prevention framework by integrating spatiotemporal modeling and data-driven decision-making. Using past wildfire data from CALFIRE, meteorological variables from NOAA, and vegetation types from California Department of Fish and Wildlife, and human activity indicators from California Department of Conservation, we aim to first construct predictive models that estimate wildfire susceptibility at a geographic scale, then derive proactive wildfire prevention strategies, optimizing resource allocation and mitigating disaster impact. The findings will serve as a foundation for AI-enhanced risk assessment frameworks, improving resilience against future wildfires in California and beyond.