HR: 09:30h
AN: S51F-06    [Abstracts]
TI: Macroscopic Source Parameters From Multi-cycle Earthquake Simulation With Heterogeneity in Rate-and-State Friction
AU: * Mai, P M
EM: mai@sed.ethz.ch
AF: Institute of Geophysics, ETH Zurich, ETH Hoenggerberg, Zurich, 8093 Switzerland
AU: Hillers, G
EM: hillers@sed.ethz.ch
AF: Institute of Geophysics, ETH Zurich, ETH Hoenggerberg, Zurich, 8093 Switzerland
AB: Recent advances in multi-cycle earthquake simulation include, among other improvements, 2-D heterogeneous distributions of rheological properties. Using rate-and-state dependent friction, we model geometrical heterogeneities of fault structures by allowing variability in the critical slip distance L along strike and with depth (while the a-b profile remains constant in the seismogenic zone). Our 3D quasi-static and quasi-dynamic simulations of slip on strike-slip faults deploy 2D anisotropic correlated distributions of L with variable correlation lengths along strike (a_x) and downdip (a_z). Be varying also the underlying distribution of L values (normal or uniform), we explore systematically the significance of the geometrical parameters ax, az on the resulting large-scale earthquake source properties. Our earthquake-cycle simulations return large sets of model quakes that display remarkable similarities to observed characteristics of finite-source rupture models. In particular, slip distributions for moderate-to-large events (M ≥q 6) are highly variable over the fault plane, ruptures tend to start at the edges of asperities (regions of large slip), and source-scaling properties are approximately consistent with real data. For that reason, we statistically analyze in this study a large set of model quakes with respect to their macroscopic source-scaling characteristics, and relate these to proposed scaling relations based on observations. However, temporal rupture evolution is much slower due to radiation damping in our quasi-dynamic simulations. Likewise, the maximum displacement is generally lower than expected from data. We therefore also investigate to what extent correction factors could be introduced to render these simulated ruptures appropriate for strong-motion calculations.
DE: 4430 Complex systems
DE: 7205 Continental crust (1219)
DE: 7212 Earthquake ground motions and engineering seismology
DE: 7290 Computational seismology
SC: Seismology [S]
MN: Fall Meeting 2005