Constraining the Volume and Runout of Post-Wildfire Debris Flows

Abstract

Debris flows pose a serious threat to downstream communities in mountainous regions around the world. This is especially true in the years following wildfire, as burned watersheds are more likely to generate debris flows and tend to produce larger debris flows than unburned watersheds. Our ability to forecast debris-flow hazards and mitigate the impacts of future events would benefit from an improved understanding of postfire debris-flow processes, from initiation to deposition. In this dissertation, I present a series of studies that improve our understanding of postfire debris-flow processes by exploring the controls on debris-flow initiation, volume, and runout and by introducing new methods for predicting postfire debris-flow volume and runout. I first explore how the properties of a postfire debris flow evolve over the course of a 7 km runout path in northern Arizona. I identify two distinct flow types that I classify based on differences in maximum clast size and flow-margin thickness, and I determine that changes in slope and channel confinement facilitate the transition between the two. This study improves our understanding of the controls on postfire debris-flow runout and provides a dataset that can be used to develop and test tools for predicting runout (Appendix A). I then introduce a new, reduced-complexity debris-flow runout model that is capable of accurately and efficiently reproducing the inundation extent and peak-flow depths of five postfire debris flows in southern California. The computational efficiency of the model indicates that it is a promising tool for rapid postfire hazard assessment (Appendix B). Next, I study the hydrologic response of a burned watershed in northern Arizona for two years following fire. I find that, despite producing multiple debris flows in the first year following fire, the watershed did not have a response in the second year. This study provides information on the triggering conditions and the temporal-persistence of debris-flow hazards in northern Arizona. I also assess the performance of postfire debris-flow runout and volume models developed in southern California, finding that improved methods are needed to accurately predict postfire debris-flow hazards in the Southwest US (Appendix C). Finally, I present a new dataset of 54 postfire debris-flow volumes that I use to develop three models for predicting postfire debris-flow volume in the Southwest. These models, which predict volume given variables related to rainfall, terrain, and/or fire severity, outperform three existing volume models that were developed elsewhere in the western US. This study improves our ability to predict postfire debris-flow volume in the Southwest and provides insight into regional differences in the factors that control debris-flow volume (Appendix D).