A Modeling System for Predicting the Behavior of Wildland Fires by Simulating Their Two-Way Interaction with the Atmosphere

Abstract:

To aid Colorado’s tactical management of active wildland fires and strategic planning for future fires, the state funded the National Center for Atmospheric Research (NCAR) to develop the Colorado Fire Prediction System (CO-FPS). The system is based on the Weather Research and Forecasting (WRF) Model coupled with a fire behavior model. Through a web interface, users configure and launch fire predictions, then view output using a geographical information system. Atmospheric turbulence is explicitly resolved with large-eddy simulations. An upgraded fire module includes a high-order advection scheme, the level-set method for tracking fires, re-initialization of the level-set to improve the stability of fire propagation, manual and automatic initialization of fires from observed fire perimeters, and prediction of wildfire characteristics such as flame length and smoke. This presentation will describe the design of CO-FPS, its functionality, how well it performs, and how it is used.