Comparison of GeoClaw and HEC-RAS for Numerically Modeling of Dam Failure and Additional Uses for GeoClaw Shallow Water Solver

This study investigates two shallow-water problems: (1) the classic shallow-water dam-break and (2) the landslide-triggered tsunami. Worldwide dams fail and endanger downstream populations, so improving our understanding of numerical models is critical. Further, climate change is expected to increase both dam failure events and landslide-triggered tsunami events in occurrence and magnitude. Therefore, studying both problems is timely and vital. Dam failure modeling studies using historical events are rare, leading to remaining questions for parameterization (e.g., breach parameterization and reservoir volume influence). We compare HEC-RAS (Hydraulic Engineer Center-River Analysis System) to the research code, GeoClaw, using the Teton Dam case study. By comparing flood wave arrival times, depths, and lateral extent, we find GeoClaw results are commensurate with historical gauge records, field observational data, and HEC-RAS results. The GeoClaw dam-break model is validated by its stability, computational cost, and conservation properties. (2) Another focus of this work uses the 2D debris flow model, D-Claw, to model landslide-triggered tsunamis. D-Claw couples granular-fluid flows equations and the shallow-water equations to resolve these complex geohazards. Outputs of this research include a comprehensive database outlining landslide-triggered tsunami events worldwide.