Soil Moisture Drives Wildfire Occurrence in the Southern Great Plains.

Innovative wildfire danger assessments that incorporate soil moisture information are becoming possible as soil moisture monitoring becomes widespread, but research describing soil moisture-wildfire relationships is lacking.  Our goal was to answer the simple but so far unanswered question, how are soil moisture and wildfires related?  We first aimed to identify relationships between soil moisture and the size of wildfires during the growing (May-October) and dormant seasons (November-April) in Oklahoma, USA.  Next, we developed seasonal models describing the probability of large (≥ 405 ha) wildfire occurrence based on soil moisture and weather.  Our initial analysis revealed a striking relationship (P < 0.001) between soil moisture and wildfires size during the growing season, with 91% of wildfires ≥ 121 ha occurring at FAW < 0.5 (a typical threshold for water stress in plants) and 77% occurring at FAW < 0.2 (extreme drought).  The largest dormant-season wildfires (≥ 405 ha) generally occurred at lower FAW than smaller fires (P < 0.001), but the range of FAW over which they occurred was greater (0.05-1.05) than for growing-season fires (0.05-0.46).  Our seasonal models revealed that the probability of a large wildfire occurring during the growing season increased from 0.18 to 0.74 as FAW fell from 0.5 to 0.2.  Dormant-season wildfire probability was related negatively to FAW and positively to lagged FAW, with wildfire probability increasing four-fold as FAW during the previous growing season increased  to 0.9 (soil moisture near ideal for plant growth).  Therefore, both wet and dry growing seasons can increase wildfire danger, but at different times and for different reasons.  We found that soil moisture strongly affected wildfire activity throughout the year, but the underlying mechanisms were seasonally dependent.  Our findings highlight the need to develop methods to use soil moisture data in wildfire danger assessments.