DEBRIS FLOODS AND FLOWS ALONG THE CENTRAL CALIFORNIA COAST FOLLOWING RECORD-BREAKING WILDFIRE ACTIVITY ACROSS THE STATE

The steep, tectonically active terrain along the central California coast is well known to produce deadly and destructive debris flows. However, the extent to which fire affects debris-flow susceptibility in this region is an open question and of widespread concern. We documented the occurrence of post-fire debris floods and flows via ground reconnaissance following the landfall of an atmospheric river storm that delivered intense rainfall across multiple burn scars. These occurrence data comprise the first post-fire hydrologic response inventory for our study region. We leveraged this inventory to evaluate the predictive performance of the U.S. Geological Survey M1 likelihood model, a tool that presently underlies the emergency assessment of post-fire debris-flow hazards in the western United States. To quantitatively test model performance, we used the Threat Score skill statistic and found that the rainfall thresholds estimated by the M1 model for the central California coast performed similar to training (southern California) and testing (Intermountain West) data associated with the original model calibration. Model performance decreased when differentiating between “minor” and “major” post-fire hydrologic response types which qualitatively weigh impacts to human life and infrastructure. Our results underscore that the problem of false positives is among the biggest challenges for developing accurate rainfall thresholds for the occurrence of post-fire debris flows. As wildfire activity increases throughout the western United States, so too will the demand for the assessment of post-fire debris-flow hazards. We conclude that additional collection of field-verified inventories of post-fire hydrologic response will be critical to prioritize which model variables may be suitable candidates for regional calibration or replacement in favor of more universal terrain-, fire-, or sediment-based metrics.