Abstract:
Understanding wildfire propagation is essential for improving prediction capabilities and
informing effective fire-management strategies. This work synthesizes two
complementary parametric studies conducted with FireStar3D, a physics-based CFD
wildfire model, to investigate how the rate of spread (ROS) and other fire behavior
characteristics relate to Byram’s convective number (a non-dimensional number
characterizing fire behavior) across both wind-driven and plume-dominated shrubland
fire regimes. A total of 184 simulations were performed, covering a broad range of wind
speeds, fuel-bed heights, fuel volume-fractions, and fuel moisture contents, enabling a
systematic evaluation of key physical drivers of fire dynamics.
Results consistently show that the ROS scales with wind speed in the wind-driven
regime, in line with established empirical rules of thumb. Across both regimes, Byram’s
convective number was found to inherently capture the integrated effects of wind, fuel
moisture, and fuel height on fire behavior. However, fuel volume-fraction emerged as an
additional and significant parameter influencing ROS due to its strong impact on heat
transfer, demonstrating that Byram’s number alone is insufficient to fully describe spread
dynamics without explicit consideration of fuel type. The study also confirmed that
certain fire properties, such as the thermal-plume angle in the plume-dominated regime,
scale solely with Byram’s number. The fire-establishment phase was examined in detail,
revealing contrasting behavior between regimes: in wind-driven conditions, the fire
spreads slightly faster during establishment than at steady state, whereas in plumedominated
fires, spread is slightly slower. In both regimes, the duration of this transient is
strongly influenced by fuel volume-fraction in wind-driven fires and predominantly by
Byram’s number in plume-dominated ones.
Collectively, these findings improve the physical understanding of how ROS and firestructure
characteristics link to Byram’s convective number, while highlighting the
independent role of fuel volume-fraction. The combined insights contribute to more
robust fire-spread modeling frameworks and support the development of enhanced
predictive tools for wildfire management.
