Triggering conditions, runout, and downstream impacts of debris flows following the 2021 Flag Fire, Arizona, USA

Abstract

Debris flows pose a serious threat to communities in mountainous areas, particularly in the years following wildfire. These events have been widely studied in regions where post-wildfire debris flows have been historically frequent, such as southern California. However, the threat of post-wildfire debris flows is increasing in many regions where detailed data on debris-flow physical properties, volume, and runout potential are sparse, such as the Southwest United States (US) (Arizona and New Mexico). As the Southwest becomes more vulnerable to these hazards, there is an increasing need to better characterize the properties of post-wildfire debris flows in this region and to identify similarities and differences with nearby areas, particularly southern California, where there is a greater abundance of data. In this paper, we study the characteristics and downstream impacts of two post-wildfire debris flows that initiated following the 2021 Flag Fire in northern Arizona, US. We gathered data regarding soil hydraulic properties, rainfall characteristics, watershed response, and debris-flow initiation, runout, volume, grain size, and downstream impacts during the first two monsoon seasons following the containment of the Flag Fire. We also applied established debris-flow runout and volume models that were developed in southern California to our study watershed and compared the output with observations. In the first monsoon season following the fire, there were two post-wildfire debris flows, one of which resulted in damage to downstream infrastructure, and one major flood event. We found that, while more intense rainfall was observed to generate debris flows at our study site compared to southern California, burned watersheds in northern Arizona are still susceptible to debris flows during storms with low recurrence intervals in the first year following fire. During the second monsoon season, there were no major runoff events, despite more intense storms. This indicates that the temporal window for heightened debris-flow susceptibility at our study area was less than one year, due to the recovery of soil hydraulic properties and vegetation regrowth. We also found that the debris-flow properties at our study site, such as volume, mobility, and grain size distribution, may differ from those in other regions in the western US, including southern California, potentially due to regional differences in rainfall characteristics and sediment supply. Differences in rainfall characteristics and sediment supply may have also influenced the performance of the debris-flow runout and volume models, which overpredicted the observed runout distance by 400 m and predicted a volume more than 17 times greater than what was observed.