An Emerging Role for Numerical Modelling in Wildfire Behavior Research: Explorations, Explanations, and Hypothesis Development

Abstract:

Advancements in computing power have created new opportunities for the use of numerical models in wildfire research. Models like the Wildland Fire Dynamics Simulator (WFDS) and FIRETEC attempt to represent interactions between the dominant processes that determine wildfire behavior such as convective and radiative heat transfer, aerodynamic drag and buoyant response of the atmosphere to heat released by the fire. Such models are not practical for operational faster-than-real-time fire prediction due to their computational and data requirements. However, their process-based model-development approach creates an opportunity to provide additional perspectives concerning aspects of fire behavior that have been observed in the field and in the laboratory, allow for sensitivity analysis that is impractical through observations and pose new hypotheses that can be tested experimentally. Specific examples of the use of FIRETEC in this fashion include: 1) investigation of the 3D fire/atmosphere interaction that dictates multiscale fireline dynamics; 2) the influence of vegetation heterogeneity and variability in wind fields on predictability of fire spread; 3) the interaction between ecosystem disturbances such as insect attacks and potential fire behavior; and 4) the effects of nonlocal topography. Numerical studies support new conceptual models for the dominant roles of multi-scale fluid dynamics in determining the nature and viability of fire spread. Results from these studies highlight the critical roles of upwind buoyancy-induced vorticity and crosswind fireline-intensity variations on heat transfer to unburned fuels, the influence of canopy structure on convective heating and cooling of canopy fuels and the impact of heterogeneous vegetation distribution/aggregation on wind penetration into canopies and crown fire behavior. There needs to be continued efforts to meet the challenges of validating the results from these numerical investigations with lab experiments and field observations, but, even so, they help progress wildfire science by suggesting relationships, interactions and phenomenology that should be considered in the context of the interpretation of observations, design of fire behavior experiments and development of new operational models.