Modeling Wind-Driven Fire Spread in the Wildland-Urban Interface

The rise in both the frequency and intensity of wildfires has resulted in a dramatic increase in the destruction of structures within communities adjacent to or intermingled with wildland vegetation, also known as wildland-urban interface (WUI) communities. As WUI communities continue to expand, the likelihood of more catastrophic wildfires is ever increasing, driven by growing human activities and exacerbated by climate change. This underscores the urgent need for strategies that reduce the vulnerability of these communities and effectively equip first responders with the tools necessary to protect lives and property. To this end, numerical simulation tools are indispensable for deepening our understanding of past WUI fires and pinpointing areas within communities that may be at heightened risk in future events.

Two simulation tools, the Weather Research and Forecasting (WRF) model and the Streamlined Wildland Urban Interface Fire Tracing (SWUIFT) model, are at the forefront of WUI fire modeling research. Aside from the traditional application of using WRF to simulate atmospheric flows from synoptic- to micro-scales, WRF also includes a fire behavior model (WRF-Fire) capable of simulating wildland fires. However, WRF-Fire is not intended for simulating WUI fires, leading to a modeling conundrum when a wildfire transitions into the WUI. Meanwhile, SWUIFT focuses on simulating fire spread within communities by capturing the effects of thermal radiation and fire spotting.

To overcome the grand challenge of seamlessly tracking fire spread from wildland to urban areas, our research team has successfully coupled WRF-Fire and SWUIFT by passing the wind velocity components and fire perimeter from WRF-Fire to SWUIFT. We apply this integrated approach to analyze two significant wind-driven wildfire events: California's Camp Fire in 2018 and Colorado's Marshall Fire in 2021, both of which began in the wildland before transitioning into the WUI. The simulations not only closely mimic the actual progression of these fires but also highlight specific locations within the communities where damage was most severe. Notably, the Marshall Fire simulation reveals that suppression prevented the destruction of hundreds of structures. The case studies also offer valuable insights into how community characteristics, such as vegetation cover and building types and layout, influence fire behavior within communities. In addition, we present coupled WRF/SWUIFT results from a wind-driven fire event that began in the WUI: Hawaii’s Lahaina Fire in 2023. For this case, fire spread simulated by SWUIFT is strongly affected by WRF’s resolution resulting from highly variable micro- and meso-scale flow characteristics.