The seasonal impact of climate on wildfire in the American Southwest (1600 1900)

Wildfires cause billions of dollars in damages in the United States each year. Improving predictions of when and where fires are likely to occur, particularly due to varying climatic conditions, is important for planning and managing resources both now and in the future. Previous work analyzing tree ring-based fire history records and climate reconstructions in the American Southwest indicates that widespread fire is associated with anomalously dry Palmer drought severity index (PDSI) values during the fire year, and lags anomalously wet PDSI values by several years. Given that PDSI integrates over several seasons, the more specific seasonal signal of these moisture anomalies is unknown. Also, the significance of wet conditions preceding fire years is not fully tested, such that the observed association could result from natural climate variability rather than being causal. We used superposed epoch analysis (SEA) to evaluate the relationships between warm and cool-season standardized precipitation index (SPI) and regional and sub-regional fire years embedded in a newly compiled network of tree-ring fire history chronologies across Arizona and New Mexico. We found that cool-season precipitation was anomalously high prior to widespread fire years (most prominent two to three years prior) and both cool and warm-season precipitation were anomalously low during the fire year. To test that the prior wet years were specifically associated with wildfire and not merely reflecting natural climate variability, we analyzed simulated fire-climate relationships on the 15 driest years in the study period that were not regional fire years. We found these drought years had no association to prior wet conditions. This result supports the historical importance of antecedent moisture in contributing to the abundance and continuity of surface fuels that promote widespread fire activity. Further, the direct associations of seasonal moisture conditions and fire activity empowers greater precision in predicting wildfire risk and potential.