Presentation at the 2025 American Meteorological Society Conference
Increases in wildfire frequency and severity across several regions of the United States have prompted a call to action in wildfire science. Many wildfire models are utilized to understand active-fire behaviors (e.g., spread). However, advancing early warning timelines enhances the potential for extensive collaboration, leading to reduced costs, more efficient resource allocation, and mitigated damages.
Our investigation of the Complex Fire case study highlights how atmospheric and surface conditions can drive wildfire activity on longer timescales (e.g., two weeks to two months). These insights are critical to building long-term, predictive wildfire models. The results suggest that key environmental variables provide predictive signals across multiple time horizons, informing the development of AI/ML-based early warning systems. We see strong potential for continued collaboration across sectors to apply this science in support of proactive planning, risk reduction, and resource efficiency in wildfire management.