Currently, no methods of predicting microscale, terrain influenced winds are available to fire managers. This study evaluated three methods of providing surface wind information to fire growth models. One was simply a uniform wind speed and direction, a method that has been traditionally used, and the other two types were based on numerical models used in the pollutant dispersion and wind power generation fields. The models were a commercial computational fluid dynamics (CFD) model and a mass-consistent model. The accuracy of the models in simulating flow fields over two terrains with available measured data was assessed. While both models performed well on the upwind and tops of hills, the mass-consistent model had significant error on the lee side. The CFD model did better here. The effect of the simulated flow fields on fire spread was examined using the FARSITE model. Comparison of using the simulated wind fields and traditional uniform wind fields was highlighted. Two historical fires with available weather and fire progression were simulated. An important finding was that both wind models produced more accurate fire spread than the uniform winds. The CFD results were the most accurate, followed by the mass-consistent and then the uniform wind fields. Another important finding was that fire spread was less sensitive to input wind direction with the simulated winds than with the uniform winds. This could benefit fire modelers trying to predict fire spread when wind direction is not known accurately. The study showed that both the CFD and mass-consistent wind model would be a useful addition to current fire behavior prediction systems. The CFD model appears to be more accurate, but the mass-consistent model has much shorter run times which could be an advantage, given short time constraints or limited computing resources.

[This publication is referenced in the "Synthesis of knowledge of extreme fire behavior: volume I for fire managers" (Werth et al 2011).]