Controls on debris flow avulsions: White Mountains of California and Nevada

Abstract

Debris flows are high-velocity landslides generated by intense precipitation events that saturate a mountain slope, leading to the rapid movement of rock and soil entrained in a water-rich fluid within channels downstream toward valleys. The process by which debris flows shift from an active channel and branch out into new channels or areas is termed avulsion. Debris flow avulsion poses serious risks for structures and populations residing on debris-flow fans, yet avulsion mechanisms are relatively unknown and unaccounted for in hazard assessments, as compared to avulsions of rivers and streams, which are better understood. However, avulsion is a critical mechanism controlling the distribution of debris flow deposits. This study analyzes six debris-flow fans in the White Mountains of California and Nevada to identify relationships between channel and avulsion characteristics, constrain the controlling factors on avulsion, and assess the probability that avulsion will occur at specified locations. This study aims to develop a method to predict avulsion based on the factors that control avulsion on debris-flow fans, toward the goal of its incorporation into debris flow hazard assessment. The fans on the western flank of the White Mountains are an ideal study area for this work, as they have a long record of debris flow and avulsion events. A database of avulsion locations and their channel characteristics was compiled in the field. These were compared to the characteristics of other positions on the fan surface that show evidence of debris flows that did not avulse. The database (n=58) of avulsion and non-avulsion characteristics was analyzed through stepwise, binary logistic regression. Results indicate that two-thirds of avulsion likelihood can be attributed to the percentage of boulders at the site, slope angle, channel width, and the ratio between flow thickness and average slope at the avulsion location. The accuracy of this model can be improved with the consideration of the presence of a coarse channel plug, which increases the likelihood of avulsion. Application of this model is demonstrated by runout simulations with forced avulsions from modeled channel plugs. The results of this project improve our ability to predict and model debris flow avulsion so that it may be readily incorporated into geohazard assessments in the future.