A dynamic view on key regional factors governing megathrust earthquakes, subduction mechanics and tsunami genesis

Abstract:

Determining characteristics that control subduction zone earthquake and tsunami dynamics is critical to mitigate megathrust hazards, but this is impeded by structural complexity, large spatio-temporal scales, and scarce instrumental coverage. Megathrust earthquake variability has been qualitatively explained using conceptual models. Depth-varying rupture behavior has been attributed to the presence of weaker materials in the near trench region. Along-arc variations in tectonic loading, megathrust interface topography, and seismic coupling have been invoked in space and time.

Here we show that modeled tsunamigenesis and earthquake dynamics are controlled by along-arc variability in regional tectonic stresses together with depth-dependent variations in rigidity and yield strength of near-fault sediments. We perform data-integrated, geometrically complex, multi-physics earthquake-tsunami modeling to identify dominant factors controlling megathrust hazards. To this end, we demonstrate how to unify and verify the required initial conditions for multi-physics earthquake-tsunami modeling from interdisciplinary geophysical observations. We present large-scale rupture dynamics and tsunami modeling of the 2004 Sumatra-Andaman earthquake and Indian Ocean tsunami that reconciles near- and far-field seismic, geodetic, geological, and tsunami observations. We specifically find that sediment strength affects tsunamigenic trade-offs between slip to the trench, splay faulting, and bulk yielding of the accretionary wedge: shallow and large slip to the trench remains invisible to teleseismic and near-field geodetic measurements, but is an important factor in dynamically sourced tsunami models.

Our computational capabilities render possible the incorporation of present and emerging high-resolution observations into dynamic-rupture-tsunami models. Our findings highlight the importance of regional-scale structural heterogeneity for earthquake dynamics and tsunami genesis. Incorporating smaller scale heterogeneities and observational uncertainties into our models may enhance scenario focused modeling and physics-based probabilistic hazard assessment of the future to decipher megathrust hazards based on the assessment of regional factors.