Exploring the Role of Antecedent Land Surface Conditions Associated with 2015 Wildfire Events in the Northwestern United States

Wildfires in the United States are responsible for numerous casualties and billions of dollars in property losses each year, particularly in the Western U.S. The 2017 California wildfires alone resulted in over $13B in losses and nearly 50 fatalities. During the past several years, higher temporal and spatial resolution land information databases have been developed by NASA’s Short-term Prediction Research and Transition Center (SPoRT) that could be applied to wildfire development and behavior. The objective of the project is to develop a proof-of-concept approach for correlating high-resolution land surface model outputs to observed wildfire behavior (i.e., rate of spread/flame lengths/spotting distances) and fire danger (i.e., energy release component and burning index) metrics in the Pacific Northwest region of the Continental U.S. (CONUS). The initial focus is on the prolific summer 2015 wildfire season over the Northwestern U.S., building upon results from lightning-ignition wildfire research at NASA Marshall Space Flight Center. The tools and datasets invoked include a 1-km configuration of the NASA Land Information System (LIS) land surface modeling and data assimilation framework, as well as Suomi-NPP Visible Infrared Imaging Radiometer Suite (VIIRS) 1-km resolution Green Vegetation Fraction (GVF) daily regional composites over the northern/western hemisphere.

Preliminary research has been conducted that combined SPoRT’s experimental-operational LIS with National Lightning Detection Network lightning information to assess wildfire potential. Initial findings showed that the 0-10 cm soil moisture content best distinguished firestarters from non-firestarters using rank sum hypothesis testing. Deep (0-200 cm) soil moisture content and GVF output showed some statistical significance but were flash-polarity dependent. This project seeks to develop statistically meaningful correlations between land surface properties and wildfire events during the active 2015 Pacific Northwest fire season. Results will serve as a catalyst for developing other regional correlations that could be useful for short-term fire behavior and danger prediction and assessment systems, including ones found within the Wildland Fire Decision Support System (WFDSS) and the National Fire Danger Rating System (NFDRS).