TWO DECADES OF POST-FIRE DEBRIS-FLOW RESEARCH IN ARIZONA AND NEW MEXICO: LESSONS LEARNED PROVIDE A ROADMAP FOR FUTURE RESEARCH

In 2002, the Arizona Geological Survey (AZGS) began monitoring post-wildfire debris flows as part of its Geologic Hazards program. Early on (2002-2006), monitoring observations were made in response to requests from local, state, and federal agencies to evaluate if, in specific drainages, post-wildfire debris flows had occurred. In most cases, detailed rainfall data were not available. In 2010, monitoring efforts expanded to document flow responses in multiple basins across fire scars, to map flood and debris-flow deposits and measure debris-flow deposit volumes where possible, and to calculate peak rainfall intensities for different durations (5-60 minutes) for each storm where rainfall data were available (e.g., near ALERT networks). In 2014, we began using time-lapse cameras and a rain-triggered video camera to determine timing of flow and augmenting existing rain gage networks with additional tipping bucket rain gages to increase coverage of rainfall data across study sites. In 2017, we began installing non-vented pressure transducers at watershed outlets to more accurately constrain flow timing and identify flow type (i.e., flood flow versus debris flow).

We compiled a geodatabase to synthesize data from 20 years of monitoring post-fire flow responses across 25 wildfire scars in Arizona and New Mexico, USA. We used these data to evaluate 15-minute rainfall intensities (I₁₅) that trigger post-fire debris flows during the first two years following a fire. Based on data from post-fire years 1 and 2, we find that I₁₅ thresholds vary with dominant vegetation type, from 42 mmh^-1 in forested ecosystems, to 58 mmh^-1 in oak/chaparral ecosystems, and 83 mmh^-1 in desertscrub ecosystems. Further, during debris-flow-producing storms where timing was constrained, we found that in 65% of cases the triggering I₁₅ was lower than the peak I₁₅, and that it was lower by roughly 33%. Understanding the lower triggering intensity is crucial for effectively setting rainfall thresholds to warn the public. Similarly, 57% of the triggering I₁₅ cases had a return interval of ≤ 2yrs compared to only 31% of the peak I₁₅ cases. The observations and measurements initiated in 2002 are helping us to understand the drivers of post-fire debris flows and recovery rates in the American Southwest.