Soil Moisture Controls on Wildfire Occurrence and Extent in Oklahoma.

The increasing availability of soil moisture data presents an opportunity to use the relationship bewteen soil moisture and live fuel moisture for wildfire risk assessments.  Our objectives were to describe characteristic soil moisture and weather conditions for a range of wildfire sizes in Oklahoma and to identify seasonal differences in those characteristic conditions.  Of the 38,420 wildfires reported in Oklahoma from 2000-2012, 36% of wildfires and 30% of area burned occurred during the growing season (May-October).  Soil moisture, as fraction of plant-available water (FAW, scaled from 0 to 1), and other variables (maximum  air temperature, minimum relative humidity, precipitation, maximum wind speed, solar radiation, 100-hr dead fuel moisture, 1000-hr dead fuel moisture, and Keetch-Byram Drought Index) were determined for each fire.  Wildfires were then grouped by size, and the average of each environmental variable was calculated for six wildfire size classes (≤ 0.40 ha, > 0.40 and ≤ 4.05 ha, > 4.05 and ≤ 40.5 ha, > 40.5 and ≤ 405 ha, > 405 and ≤ 4047 ha, and > 4047 ha).  Regression of wildfire size class against environmental variables showed that during the growing season the relationship between FAW and wildfire size class (r² = 0.96, p < 0.001) was among the strongest of examined variables.  All growing-season wildfires > 405 ha occurred when FAW was < 0.5, and 87% occurred when FAW was < 0.2.  During the dormant season, the relationship between FAW and fire size class was weaker (r² = 0.73, p = 0.03), with most other variables being more strongly related to fire size.  The importance of FAW during the growing season is a consequence of the influence of soil moisture on live plants, whereas weather variables are more important during the dormant season when vegetation is not actively growing.