On Dislocation Modeling of Megathrust Tsunami Sources

Abstract

Modeling tsunamis due to subduction earthquakes for scientific research and hazard assessment requires accurate quantification of coseismic seafloor deformation. Although the widely used analytical model of shear dislocation in a uniform elastic half space can accommodate complex fault geometry and slip distribution, it fails to capture the sloping seafloor topography and heterogeneous rock rigidity (shear modulus) in real subduction zones. In practice, these real-world complexities are either ignored or addressed by adjusting fault geometry and/or applying corrections to the deformation results, with consequences poorly understood. This study investigates the validity or errors of these simplifications by comparing dislocation model results with those from finite element models that account for these complexities. Our analysis reveals that the absence of the seafloor slope can be accurately compensated by adjusting the shallow geometry of the megathrust such that the fault depth below the flat model surface approximates the actual fault depth below the seafloor. Effects of short-wavelength bathymetry can be effectively incorporated by adding a commonly used gradient-based correction. For slip-to-trench ruptures, it is crucial to adjust fault geometry to ensure that the fault reaches the model surface at the trench; otherwise, the abrupt slip termination at a small depth creates an uplift spike which is a commonly seen artifact in tsunami source models. Our findings highlight the secondary or minimal effects of heterogeneous rigidity on tsunamigenic deformation if fault slip is kinematically assigned. This research offers guidance for the development of more accurate tsunami source models using analytical dislocation solution.