Oral Presentation

Instead of single planes, faults are volumes of rocks containing fault cores surrounded by hundreds-meter-thick fractured rocks, i.e., the so-called fault damage zones. Major fault zones accommodate a significant portion of plate motion through seismic and aseismic deformation. As they can host large earthquakes that cause strong ground shaking, it is essential to understand how earthquakes grow in fault zones and whether their source properties and ground motions are strongly affected by fractured rocks. Our dynamic rupture simulations of large earthquakes demonstrate that dynamic waves in fault zones cause complex rupture styles such as short rise-time slip pulses and supershear rupture that propagates stably at a range of speeds usually considered as unstable. The modulation of rise time and rupture speed enhances the high-frequency content of earthquake ground motions. We also show through kinematic models and seismic observations that P-wave reflections in fault zones can cause a high-frequency peak in the velocity spectra of small earthquakes, which in practice may be inferred as part of the source spectra.

Fault damage zones are also highly variable along the fault and evolve over time due to coseismic damage and interseismic healing. Dynamic rupture simulations with spatially heterogeneous fault zones find that earthquakes tend to initiate and rupture in more damaged fault zone segments and arrest when reaching less damaged segments. We also simultaneously consider dynamic fault zone waves that occur within seconds and earthquake cycles that span hundreds of years in fully dynamic earthquake cycle simulations. The fault zone simulations generate earthquakes with a wide range of magnitudes (M3–7) and bimodal earthquake depth distributions as observed in nature. As fault zone material properties evolve over earthquake cycles, we find fault structural maturity has a primary influence on earthquake magnitudes and recurrence intervals. Immature fault zones promote aseismic slip that propagates into the seismogenic zone and limits earthquake size, whereas earthquakes in mature fault zones occur more regularly and are larger since they can break through to the surface. Finally, we will present new insights into the interplay between precursory seismic velocity changes in fault zones and earthquake nucleation process, which can be used to guide future real-time monitoring of major faults.