[S-SS05] Fault Rheology and Earthquake Physics

[SSS05-07] Connecting Probabilistic and Physical Perspectives on Megathrust Rupture Processes

*Junle Jiang¹, Jose Viteri Lopez¹, Segun Bodunde¹ (1.School of Geosciences, University of Oklahoma, Norman, OK, USA)

Keywords:Megathrust seismicity, Bayesian inference, Space geodesy, Tsunami, Postseismic deformation

Geophysical research of subduction zones fundamentally relies on often sparse surface observations to infer deep subsurface processes and reveal the system dynamics despite limited spatiotemporal resolutions. Integrating observations, inferences, and physics of megathrust processes is essential to a deeper understanding and potential forecast of subduction zone hazards. Here we present a modeling framework to connect probabilistic inference and physical modeling of megathrust earthquake ruptures, stress changes, and postseismic fault behavior. First, we use a Bayesian inversion approach to incorporate multiple datasets—including transoceanic tsunami records and on-land geodetic displacements from Interferometric Synthetic Aperture Radar (InSAR) and Global Positioning System (GPS)—to explore detailed megathrust rupture processes associated with the 2010 M_w8.8 Maule (Chile) earthquake. To improve source inference, we characterize and account for modeling errors in simulating tsunami waveforms and ground displacements, e.g., due to one- or three-dimensional subsurface elastic structure. The Bayesian approach enables the quantification of spatial resolutions and information content of each dataset and produces plausible source models with minimal a priori assumptions on source complexities. Second, the posterior model ensembles allow us to investigate the near-trench, downdip, and along-strike variations in fault slip and associated stress changes due to the mainshock. We compare the updated Maule results with two other large Chilean earthquakes (2014 M_w8.1 Iquique and 2015 M_w8.3 Illapel) to assess characteristic patterns of stress changes and their uncertainties from the trench to the coast. Lastly, we consider multi-rheology postseismic models of the Maule earthquake to estimate the longer-term stress changes in space and time and their comparison with the aftershock properties. Our data-driven, physics-based framework improves the characterization of the megathrust fault behavior and may enhance our ability to forecast subduction zone hazards such as tsunamis and aftershocks.