On the present and future of physics-based earthquake source modeling

Nadia Lapusta

Submitted August 17, 2018, SCEC Contribution #8841, 2018 SCEC Annual Meeting Talk on Tue 11:00

Accelerating streams of field observations, lab studies, and numerical modeling have significantly improved our understanding of earthquakes and physical factors that affect them. The main suspects have been known for a while. Tectonic loading, static and wave-mediated stress transfers, aseismic slip, rate-and-state friction, fault geometry and roughness, visco-plastic deformation at depth, shear heating during rapid slip, variations in pore fluid pressure, and off-fault damage/healing have all been shown to significantly affect, and in some cases dominate, the stress/strength evolution on faults and hence the earthquake patterns that result. Through combined field, lab, and numerical studies, our research community is well on the way to systematically quantifying the physical factors and evaluating their relative importance for different faults/phenomena/scales. A number of earthquake source models are being developed, focusing on different combinations of the physical factors and scales, to aggregate the available knowledge and identify crucial gaps. Together, the modeling efforts are advancing towards interpreting field observations in terms of tractable models with physically meaningful fault and bulk properties that can be evaluated, at least in principle, through lab, field, and smaller-scale numerical studies.

Earthquake simulators have taken on a necessary and formidable task of investigating earthquake patterns on realistic fault networks, which is presently tractable only with simplifications and omissions of the other physical factors. They have been successful in finding sets of parameters, some physical and some ad-hoc, that allow the models to match known statistical properties of regional earthquake sequences. The simulators are valuable research tools for studying fault system dynamics. However, their simplifications necessitate careful considerations about which conclusions they allow us to draw. For example, why would we expect the simulators to provide realistic probabilities of jumps over step-overs if they ignore the potentially dominating physics - dynamic stress changes brought by seismic waves? Including the simulators in hazard assessment requires clear communication of their limitations, verification through comparisons with smaller-scale models that more accurately capture relevant physics, and community consensus of what criteria and observations the models need to satisfy. We are not there yet, but we are on the right track.

Citation
Lapusta, N. (2018, 08). On the present and future of physics-based earthquake source modeling. Oral Presentation at 2018 SCEC Annual Meeting.


Related Projects & Working Groups
Fault and Rupture Mechanics (FARM)