Pseudodynamic Source Characterization for Strike-Slip Faulting Including Stress Heterogeneity and Super-Shear Ruptures

Abstract

Reliable ground-motion prediction for future earthquakes depends on the ability to simulate realistic earthquake source models. Though dynamic rupture calculations have recently become more popular, they are still computationally demanding. An alternative is to invoke the framework of pseudodynamic (PD) source characterizations that use simple relationships between kinematic and dynamic source parameters to build physically self-consistent kinematic models. Based on the PD approach of Guatteri et al. (2004), we propose new relationships for PD models for moderate-to-large strike-slip earthquakes that include local supershear rupture speed due to stress heterogeneities. We conduct dynamic rupture simulations using stochastic initial stress distributions to generate a suite of source models in the magnitude M w 6-8. This set of models shows that local supershear rupture speed prevails for all earthquake sizes, and that the local rise-time distribution is not controlled by the overall fault geometry, but rather by local stress changes on the faults. Based on these findings, we derive a new set of relations for the proposed PD source characterization that accounts for earthquake size, buried and surface ruptures, and includes local risetime variations and supershear rupture speed. By applying the proposed PD source characterization to several well-recorded past earthquakes, we verify that significant improvements in fitting synthetic ground motion to observed ones is achieved when comparing our new approach with the model of Guatteri et al. (2004). The proposed PD methodology can be implemented into ground-motion simulation tools for more physically reliable prediction of shaking in future earthquakes.

Rupture complexity and the supershear transition on rough faults

“…Consequently, the fraction of supershear rupture increases with both the initial background shear stress 0 and the roughness . The increasing likelihood of supershear transients at larger 0 supports the results of Mena et al [2012] that supershear ruptures prevail in dynamic simulations with longer rupture length, which occurs more commonly in our simulations at larger background shear stresses. Also worth noticing is that the peak in the supershear velocity distribution is at approximately 0.9c P .…”

A Probable Earthquake Scenario near Istanbul Determined from Dynamic Simulations

“…The spatial heterogeneity introduced in this study originates only from the fault geometry, which principally controls the macroscopic parameters such as the final magnitude and the rupture directivity. Further heterogeneities at smaller scales could be considered (1) deterministically, if we know the spatial distribution of heterogeneity such as fault asperity (Ide and Aochi, 2013), seismic coupling (Schmittbuhl et al, 2014), and other geochemical factors (Tryon et al, 2010), or (2) stochastically, as this is often developed in pseudodynamic modelings (e.g., Mena et al, 2012;Schmedes et al, 2012;Song and Dalguer, 2013;Trugman and Dunham, 2014). Such heterogeneities may make the triggering of multiple fault segments difficult (Kase, 2010).…”

A Laboratory Earthquake‐Based Stochastic Seismic Source Generation Algorithm for Strike‐Slip Faults and its Application to the Southern San Andreas Fault

“…Such an approach is used by pseudodynamic stochastic rupture model generators (e.g., Song and Somerville, 2010;Mena et al, 2012;Song et al, 2014;Trugman and Dunham, 2014). They typically represent final slip, rupture speed, and slip rate as spatial random fields that are statistically characterized (and/or correlated) by synthetic data from dynamic rupture simulations.…”