Coupled Weather-Fire Model Simulations of Extreme Winds and Fire Behavior during Recent Windstorm-Driven Wildfire Events

Several of the most destructive recent U.S. wildfire events were associated with strong downslope winds in the lee of mountain ridges, variously referred to as Diablo winds in northern California, Santa Anas in southern California, chinooks in the Front Range of the Rocky Mountains, and the more general term föehnwindsin the eastern U.S. Mesoscale model simulations of the weather leading up to and during these events capture broad spatial patterns of accelerated winds, yet in several cases, ignitions and rapid early fire growth appear linked to strong winds and local extrema in some locations that destruction and anecdotal evidence suggest may have reached 30-40 m s^-1 . While these features are small, sub-mesoscale circulations, they are not weak nor merely 'gusts', but resolvable dynamic phenomena. Understanding the magnitude, structure, and spatial distribution of these peaks has significant consequences for infrastructure design and operation in mitigating the damage from future events. We apply convective-scale (hundreds of meters horizontal grid spacing) simulations of the weather leading up to and during some of these events (including events during the 2016 October North Bay wildfire outbreak and the 2017 Camp Fire) and early fire growth using the CAWFE^âcoupled weather-wildland fire modeling system, using satellite active fire detection data from the Visible and Infrared Imaging Radiometer Suite (VIIRS) instrument to evaluate event simulations. We find common factors such as a shallow river of fast-moving stable air but several different types of dynamic microscale flow regimes, many of which produce dynamic microscale airflow regimes not present in the scientific literature and wind extrema that are co-located with the ignition area of highly destructive wildfires. We discuss the adequacy of the current and proposed surface station network - a foundation for utility de-energization - in detecting and characterizing such flows.

Several of the most destructive recent U.S. wildfire events were associated with strong downslope winds in the lee of mountain ridges, variously referred to as Diablo winds in northern California, Santa Anas in southern California, chinooks in the Front Range of the Rocky Mountains, and the more general term föehnwindsin the eastern U.S. Mesoscale model simulations of the weather leading up to and during these events capture broad spatial patterns of accelerated winds, yet in several cases, ignitions and rapid early fire growth appear linked to strong winds and local extrema in some locations that destruction and anecdotal evidence suggest may have reached 30-40 m s^-1 . While these features are small, sub-mesoscale circulations, they are not weak nor merely 'gusts', but resolvable dynamic phenomena. Understanding the magnitude, structure, and spatial distribution of these peaks has significant consequences for infrastructure design and operation in mitigating the damage from future events. We apply convective-scale (hundreds of meters horizontal grid spacing) simulations of the weather leading up to and during some of these events (including events during the 2016 October North Bay wildfire outbreak and the 2017 Camp Fire) and early fire growth using the CAWFE coupled weather-wildland fire modeling system, using satellite active fire detection data from the Visible and Infrared Imaging Radiometer Suite (VIIRS) instrument to evaluate event simulations. We find common factors such as a shallow river of fast-moving stable air but several different types of dynamic microscale flow regimes, many of which produce dynamic microscale airflow regimes not present in the scientific literature and wind extrema that are co-located with the ignition area of highly destructive wildfires. We discuss the adequacy of the current and proposed surface station network - a foundation for utility de-energization - in detecting and characterizing such flows.