A Wildland Fire Spotting Parameterization for the Weather Research and Forecasting Model

Wildland fires are ubiquitous, inherently complex phenomena that occur in many regions throughout the world. A large number of fires – particularly those occurring in the wildland-urban interface – overwhelm firefighting activity and cause significant losses of life and property. Some of these highly destructive wildfires spread rapidly due to a process known as fire spotting. During the spotting process, the turbulent atmospheric flow generates and transports firebrands (flaming or glowing particles), some of which are capable of igniting new fires far away from the main fire front. Numerical representation of the spotting processes is challenging, and it requires special attention in the context of coupled fire-atmosphere modeling.

To this end, we have developed a wildland fire spotting parameterization for the Weather Research and Forecasting (WRF) model coupled to a fire behavior model based on the Coupled Atmosphere-Wildland Fire Environment. The coupled model, known as WRF-Fire, is utilized as an operational tool for the Colorado Fire Prediction System. We have adapted the particle transport physics from the Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) model to aid in the development of a new module within the WRF physics package. Resolved fire-atmosphere interactions are allowed to influence particle transport in the new parameterization, and individual firebrand properties (namely temperature and mass/diameter) are calculated and tracked. We conduct numerical simulations of idealized and real-world atmospheric conditions to test the utility of incorporating a wildland fire spotting parameterization into an operational model, as well as to quantify the sensitivity of the new parameterization to changes in environmental conditions and initial firebrand properties.