Evaluating the sensitivity of debris flow inundation patterns to initiation and upstream flow characteristics

Abstract:

Predicting the extent of debris flow inundation is an important aspect of debris flow risk assessment. However, the uncertainty of initiation mechanism (distributed erosion and accumulation vs mass failure, etc.) and upstream flow and volume characteristics complicate the prediction. We used the model D-Claw, a shallow-water model for two-phase granular-fluid mixtures over varying terrain, to evaluate the sensitivity of debris flow runout characteristics to different initiation scenarios in the Van Tassel watershed, Los Angeles County, California. This 4 km² watershed was burned in the 2016 San Gabriel Complex Fire and experienced a series of rainfall-triggered debris flows and debris floods the following winter. We compared the simulated inundation patterns resulting from different initiation assumptions having the same total volume of entrained debris flow material. We modeled several initiation conditions including: (1) equally distributed erosion throughout the watershed, (2) material mobilized along the main flow channels within the watershed, and (3) a hydrograph release of the debris flow volume at the fan apex. The debris flow material volume and storm hydrograph are constrained by data recorded at a monitoring site just upstream of the fan apex. These comparisons assess the extent to which accurate predictions of potential debris flow inundation can be achieved by initializing D-Claw with computationally simple hydrographs versus more intensive calculations of headwater debris flow initiation, erosion, and entrainment. Variability among hydrographs at the fan apex and the resulting patterns of inundation highlight the sensitivity of runout characteristics to the upstream flow conditions. Inundation patterns were compared to the cumulative inundation observed following the first rainy season after the 2016 San Gabriel Complex Fire. Evaluating the sensitivity of debris flow inundation to upstream flow conditions allows for improved hazard risk assessment.