This project introduces a groundbreaking approach to wildfire risk management at the wildland–urban interface. Its core innovation is an integrated framework that combines high-resolution physical fire modeling, behavioral decision analysis, and vulnerability assessment, domains that have historically operated in isolation.
At the technical level, the team is developing a fire spread model capable of simulating fire progression across natural and built environments at 10-meter resolution and hourly intervals. This advancement reduces computation time from days to hours, enabling near-real-time forecasting of fire perimeters, spread rates, and directions, critical intelligence for emergency planners.
The project’s impact extends beyond predictive modeling by incorporating proactive behavioral research and targeted applications for vulnerable populations. Through surveys designed to capture evacuation decisions under diverse hypothetical fire and policy scenarios, the team will generate forward-looking insights into human responses, laying the groundwork for AI-driven decision-support tools.
A pilot application focused on nursing homes demonstrates the framework’s practical value, producing tailored evacuation strategies for facilities with limited mobility and high vulnerability. By bridging predictive science with actionable planning, this work promises to transform wildfire preparedness, enhance community resilience, and set a new standard for integrated climate-risk management.
