Spatio-temporal clustering of seismicity enabled by off-fault plasticity

Abstract

Earthquakes are among nature’s deadliest and costliest hazards. Understanding mechanisms for earthquake nucleation, propagation, and arrest is key for developing reliable operational forecasts and next generation seismic hazard models. While significant progress has been made in understanding source processes in linear elastic domains, the response of the rocks near the fault is complex and likely to be inelastic due to the extreme stresses and deformations associated with fault slip. The effect of this more realistic fault zone response on seismic and aseismic fault slip is poorly understood. Here, we simulate sequence of earthquake and aseismic slip of a fault embedded in an elastic-viscoplastic bulk subject to slow tectonic loading. We show that off-fault plasticity significantly influences the source characteristics. Specifically, off-fault plasticity may lead to partial ruptures and emergence of spatial segmentation as well as hierarchical temporal seismic clustering. Furthermore, co-evolution of fault slip and off-fault bulk plasticity may lead to heterogeneous rupture propagation and results in pockets of slip deficit. While the energy dissipated through plastic deformation remains a small fraction of the total energy budget, its impact on the source characteristics is disproportionally large through the redistribution of stresses and viscous relaxation. Our results suggest a new mechanism of dynamic heterogeneity in earthquake physics that can be active for both small and large earthquakes and may have important implications on earthquake size distribution and energy budget. Furthermore, this plasticity-induced self-limiting crack dynamics may be relevant for other dynamic fracture applications and design of dynamically tough materials.

A discontinuous Galerkin method for sequences of earthquakes and aseismic slip on multiple faults using unstructured curvilinear grids

“…Inspecting the results, see Figure 14a, reveals that strain becomes very large at the intersection between a splay and the main fault, and that the change in traction changes sign, indicated by red lines which show the fault system geometry warped by the change in shear traction in the direction of the normal of the main fault. We might overcome this issue by future implementation of appropriate jump conditions for tensile stresses (Day et al 2005) or accounting for off-fault inelastic deformation (Templeton & Rice 2008;Gabriel et al 2013;Erickson et al 2017;Wollherr et al 2018;Mia et al 2021).…”

Modeling Sequences of Earthquakes and Aseismic Slip (SEAS) in Elasto‐Plastic Fault Zones With a Hybrid Finite Element Spectral Boundary Integral Scheme

“…Initially, FEBE was implemented to reduce the computational cost associated with simulating dynamic rupture propagation in complex fault zones (Albertini et al., 2021; Hajarolasvadi & Elbanna, 2017; X. Ma & Elbanna, 2019; X. Ma et al., 2018). (Abdelmeguid et al., 2019; Mia et al., 2022) later extend the framework to SEAS models in anti‐plane setting with fault zone heterogeneity. The main advantage of FEBE is its flexibility in incorporating local fault zone features (complexities) through finite element method (FEM), while truncating the majority of homogeneous elastic bulk through spectral boundary integral equation (SBIE) method.…”

“…Spatial discretization methods for SEAS generally fall under two main categories: domain‐based approaches and boundary integral approaches. The flexibility of domain‐based methods allows for handling small‐scale heterogeneities, material nonlinearities, as well as complexities of fault geometry (Aagaard et al., 2013; Allison & Dunham, 2018; Barbot, 2019; Erickson & Dunham, 2014; Erickson et al., 2017; Kaneko et al., 2011; Kuna, 2013; Mia et al., 2022; Taborda & Bielak, 2011; Thakur et al., 2020). However, modeling sequences of earthquakes and seismic slip with domain‐based methods requires substantial computational effort due to the different spatial, and temporal scales (Allison & Dunham, 2018; Biemiller & Lavier, 2017; Kaneko et al., 2008; Mckay et al., 2019; Tong & Lavier, 2018; Uphoff et al., 2022; Van Dinther et al., 2013).…”

“…The role of viscosity has been explored in an anti‐plane setting by Mia et al. (2022), for in‐plane deformations we relegate the exploration of the role of viscosity to a future study.…”

“…Since their emergence early 2000s, SEAS models have provided crucial insights on natural earthquake phenomena such as spontaneous nucleation of earthquakes (Lapusta & Rice, 2003; Lapusta et al., 2000), slow slip events (Barbot, 2019), small repeating earthquakes (Chen & Lapusta, 2009), and seismicity swarms (Zhu et al., 2020). Additionally, advancement in computational tools enabled investigations of the long term effects of thermal pressurization (Noda & Lapusta, 2013), poroelasticity (Torberntsson et al., 2018), quasidynamic slip evolution and fault roughness (Cattania & Segall, 2021; Heimisson, 2020), bi‐material effects (Abdelmeguid & Elbanna, 2022), low velocity fault zones (Abdelmeguid et al., 2019; Thakur et al., 2020), and inelastic deformations (Erickson et al., 2017; Mia et al., 2022; Tal & Faulkner, 2022) on the evolution of aseismic and coseismic slip. In most elastic models the overall pattern would converge to a statistically steady solution independent of the initial conditions after this transitional spin‐up period (Erickson & Jiang, 2018).…”