HR: 14:35h
AN: S42F-04    [PDF]
TI: Physically Self-Consistent Earthquake Source Models for Strong Ground Motion Prediction
AU: Guatteri, M
EM: mariagiovanna_guatteri@swissre.com
AF: Swiss Reinsurance America, Armonk, New York, NY 10504 United States
AU: * Mai, P
EM: mai@seismo.ifg.ethz.ch
AF: Institute of Geophysics, ETH Hoenggerberg,, Zurich, ZH CH-8093 Switzerland
AU: Beroza, G C
EM: beroza@pangea.stanford.edu
AF: Department of Geophysics, Stanford University, Stanford, CA 94305-2215 United States
AB: Accurate predictions of the intensity and variability of near-source strong ground motions for future large earthquakes depend strongly on our ability to simulate realistic models of earthquake rupture. Commonly used kinematic earthquake source models are characterized by heterogeneity in the slip distribution and a temporal rupture evolution (rise time, rupture velocity, slip-velocity function) that is assumed \emph{a priori}. Such kinematic source models may not be physically realizable as a self-sustaining fault rupture. In order to make near-field ground motion simulations more accurate, we have developed a procedure to generate physically consistent earthquake rupture models. This so-called \emph{pseudo-dynamic} source characterization constitutes inherently kinematic rupture models that are designed to emulate important characteristics of fully dynamic rupture models through a series of empirical relationships derived from the analysis of a set of spontaneous rupture models. \\ We construct pseudo-dynamic models by first generating a slip distribution using a spatial random-field model that is consistent in its overall scaling and spatial variability with slip distributions observed in past earthquakes [\emph{Mai and Beroza}, 2000; 2002]. We then compute the static stress drop $\Delta\sigma$ associated with this slip distribution [\emph{Andrews}, 1980], which is then used to calculate the local rupture velocity through empirical relations between stress drop $\Delta\sigma$, fracture energy $G_c$ and crack length $L_c$ [\emph{Andrews}, 1976; \emph{Guatteri et al}, 2003]. An equivalent approach is used to simulate the distribution of rise time over the fault and the parameters describing the slip-velocity function. A simple energy-budget calculation is used to discard source models that would not be realizable as spontaneous ruptures.\\ Our pseudo-dynamic source characterization reproduces the important features of dynamic rupture without full dynamic simulations, and hence circumvents the limits imposed by the computational demand of dynamic rupture modelling for multiple scenario earthquakes. While the relationships between source parameters described in this paper are significant simplifications of the true complexity of the physics of the rupture process, they help identify important relationships between source properties that are relevant for strong ground motion prediction.
DE: 7212 Earthquake ground motions and engineering
SC: Seismology [S]
MN: 2003 Fall Meeting