Impact Versus Frictional Earthquake Models for High‐Frequency Radiation in Complex Fault Zones

Earthquakes occur within complex fault zones containing numerous intersecting fault strands. This complexity poses a computational challenge for rupture models, which typically simplify fault structure to a small number of rough fault surfaces, with all other deformation assumed to be off‐fault viscoplastic deformation. In such models, high‐frequency ground motions originate solely from frictionally mediated, heterogeneous slip on a small number of potentially rough fault surfaces or from off‐fault viscoplastic deformation. Alternative explanations for high‐frequency ground motion generation that can account for a larger number of fault surfaces remain difficult to assess. Here, we evaluate the efficacy of a recently proposed stochastic impact model in which high‐frequency ground motion is caused by elastic impacts of structures within a complex fault zone. Impacts are envisioned to occur in response to fault motion in the presence of geometrical incompatibilities, which promotes transfer of slip onto different fault strands on timescales mediated by elasticity. We investigate the role of a complex fault zone for high‐frequency ground motion by comparing the underlying assumptions and resulting predictions of impact and rough fault frictional models. Relative to rough fault frictional models, impact models are characterized by deformation timescales and corner frequencies that are set by elasticity rather than viscoplasticity, relatively angular rather than smoothly varying fault roughness geometries, high‐frequency radiation patterns that are more isotropic, and higher P / S radiated energies. We outline ways to discriminate whether impact or rough fault frictional models are more likely to explain observations of high‐frequency ground motions.