HR: 0800h
AN: S51D-1047    [Abstracts]
TI: Finite Source Strong Ground Motion Synthesis With Pseudo Green's Functions
AU: Scognamiglio, L
EM: scognamiglio@ingv.it
AF: Istituto Nazionale di Geofisica e Vulcanologia, via di Vigna Murata 605, Rome, Ita 00143 Italy
AU: * Hutchings, L
EM: hutchings2@llnl.gov
AF: Lawrence Livermore National Laboratory, P.O. Box 808, Livermore, CA 94551 United States
AU: Akinci, A
EM: akinci@ingv.it
AF: Istituto Nazionale di Geofisica e Vulcanologia, via di Vigna Murata 605, Rome, Ita 00143 Italy
AU: Foxall, W
EM: foxall1@llnl.gov
AF: Lawrence Livermore National Laboratory, P.O. Box 808, Livermore, CA 94551 United States
AB: We analyze the usefulness of stochastic time series as Green's functions in finite fault rupture modeling by modeling observed accelerograms from the 1997 Colfiorito earthquake. We create synthetic high frequency Green's functions (> 1.0 Hz) with Gaussian white noise shaped by an envelope function, and modified by attenuation, geometrical spreading, and site-dependent kappa. The pseudo Green's functions are used as records from impulsive point shear dislocation sources in the representation relation to calculate synthetic seismograms. Using Green's functions as the propagation contribution from point sources allows us to avoid using scaling relations between small and large earthquakes. Also, the spectra of the finite sources are entirely determined by the finite rupture parameters, so we do not rely on models of source spectra such as the Brune model. The shaping functions to the stochastic time series are obtained by regression on the observed seismicity using the method developed by Malagnini et al. (2000) for propagation characteristics and NetMoment (Hutchings and Gok, 2002) analysis of site-specific kappa. So that, the pseudo Green's functions have the wave propagation characteristics of the region on average, and site specific information when possible.
DE: 0560 Numerical solutions (4255)
SC: Seismology [S]
MN: Fall Meeting 2005