Modeling of sediment transport in rapidly-varying flow for coastal morphological changes caused by tsunamis

Abstract

Tsunamis can cause significant coastal erosion and harbor sedimentation that exacerbate the concomitant flood hazards and hamper recovery efforts. Coupling of the non-hydrostatic model NEOWAVE and the sediment transport model STM provides a tool to understand and predict these morphological changes. The non-hydrostatic model can describe flow fields associated with tsunami generation, wave dispersion as well as shock-related and separation-driven coastal processes. The sediment transport module includes non-equilibrium states under rapidly-varying flows with a variable exchange rate between bed and suspended loads. A previous flume experiment of solitary wave runup on a sandy beach provides measurements for a systematic evaluation of sediment transport driven by shock-related processes. The extensive impacts at Rikuzentakata, Iwate, Japan and Crescent City Harbor, California, USA from the 2011 Tohoku tsunami provide pertinent case studies for model benchmarking. We utilize a self-consistent fault-slip model to define the tsunami source mechanism and field survey data to determine the characteristic grain sizes and morphological changes. The near-field modeling at Rikuzentakata gives reasonable fits with observed large-scale erosion and sedimentation associated with transition of the incoming wave into a surge and formation of a hydraulic jump in the receding flow. The non-hydrostatic module becomes instrumental in resolving tsunami waves at the far-field shore of Crescent City. The results show good agreement with local tide-gauge records as well as observed scour around coastal structures and deposition in basins resulting from separation-driven processes. While the erosion patterns in the laboratory and field cases can be explained by suspended sediment transport in the receding flow, bed load transport can be a dominant mechanism in sediment laden flows and scour around coastal structures.