Exciting news! We're transitioning to the Statewide California Earthquake Center. Our new website is under construction, but we'll continue using this website for SCEC business in the meantime. We're also archiving the Southern Center site to preserve its rich history. A new and improved platform is coming soon!

Towards realistic simulated broadband ground motion of multi-faults earthquake using physics-based dynamic earthquake rupture models.

Thomas Ulrich, & Alice-Agnes Gabriel

Submitted September 11, 2022, SCEC Contribution #12345, 2022 SCEC Annual Meeting Poster #236

Supercomputing-empowered 3D physics-based earthquake scenarios, assimilating available regional knowledge, such as fault geometry, community velocity and stress models, topography, past earthquakes, can enable non-ergodic Probabilistic Seismic Hazard Assessment (PSHA), complementing traditional Ground Motion Prediction Equations (GMPEs).

The first step to unleash this potential is identifying parameters leading to simulated broadband ground motions matching empirical predictions. The parameter space is explored in a collaborative effort (Withers et al., SCEC 2022), aiming at simulating realistic broadband ground motion using dynamic rupture modeling. In this group effort, fault geometry, faulting mechanism, expected moment magnitude, and velocity structure are constrained, and efforts focus at near-source distances. Simulated metrics are compared with empirical predictions at frequencies up to 3 Hz. Here, we use SeisSol (www.seissol.org) and consider band-limited fractal rough faults embedded in a laterally-uniform regional stress regime. In combination with a short slip weakening distance (d_c=0.15m) governing linear slip weakening friction, and with stochastic variations of the dynamic friction coefficient we yield peak slip rates in the range of 5-15 m/s and realistic broadband ground motion.

Our dynamic rupture model utilizes tetrahedral meshes, which allows accounting for complicated geometries, including complex fault networks, topography, and sedimentary basins. Here, we demonstrate workflows to incorporate fault roughness into geometrically complex fault networks. A fine spatial mesh discretization is required to resolve the smallest wavelength of the fractal rough fault (e.g., we show that 50 m edge-length elements are required for resolving fault roughness at 200 m scale). Yet, a statically adaptive mesh, constrained by the velocity model, and a high-order numerical scheme (polynomial order 4) allow resolving 4-5 Hz broadband ground motion with reasonable computing resources (~ 20k CPUh per simulation). Fine mesh discretization requirements could be relaxed by hybrid dynamic rupture models utilizing computationally cheap strength/traction heterogeneities in combination with more challenging geometric complexity enabling extremely efficient simulations. In this study, we propose guidelines for defining the minimum wavelengths to be geometrically resolved based on the process zone width, an inherent length scale of dynamic earthquake rupture.

Key Words
ground motion, dynamic rupture, SeisSol

Citation
Ulrich, T., & Gabriel, A. (2022, 09). Towards realistic simulated broadband ground motion of multi-faults earthquake using physics-based dynamic earthquake rupture models.. Poster Presentation at 2022 SCEC Annual Meeting.


Related Projects & Working Groups
Ground Motions