HR: 1340h
AN: S43A-1041    [Abstracts]
TI: Kinematic Source Model of the 2004 Parkfield Earthquake
AU: * Kim, A
EM: ahyi@seismo.berkeley.edu
AF: Department of Earth and Planetary Science, University of California at Berkeley, 289 McCone Hall, Berkeley, CA 94720
AU: Dreger, D
EM: dreger@seismo.berkeley.edu
AF: Department of Earth and Planetary Science, University of California at Berkeley, 289 McCone Hall, Berkeley, CA 94720
AU: Murray, M H
EM: mhmurray@seismo.berkeley.edu
AF: Department of Earth and Planetary Science, University of California at Berkeley, 289 McCone Hall, Berkeley, CA 94720
AB: The September 28, 2004 Mw6.0 Parkfield earthquake is probably the best recorded moderated sized earthquake to date. Approximately 41 near-fault strong motion stations lie within 19 km of the ruptured fault. Preliminary kinematic models for this event obtained by various researchers show a rupture process that is principally unilateral to the NW, initiating near Gold Hill and propagating to Middle Mtn. This process differs from what is known about previous Parkfield earthquakes (1934, 1966), which nucleated near Middle Mtn, and ruptured to the SE. Although the kinematic models for the 2004 event are largely consistent, the unilateral rupture to the NW is inconsistent with the distribution of near-fault strong ground shaking. Large amplitude ground motions are observed both north and south of the epicenter, however the kinematic models fail to explain the elevated motions to the SE. The San Andreas fault zone in this area is complex with multiple strands, and previous studies (e.g. Eberhart-Phillips and Michael 1993) have shown that there is a marked velocity contrast across the fault. It is possible that 3D velocity structure in the fault zone could have contributed to the strong ground motions observed close to the fault SE of the epicenter, but it is also possible that the regional data used in the preliminary studies failed to map significant slip extending SE of the epicenter. In this study we develop a kinematic model making use of the near-fault data from the NSMP and CSMIP strong motion networks. We utilize a linear multiple time window approach to invert the data for the spatial distribution of fault slip, the average rupture velocity and its variation, and the slip rise time. The model obtained using the near-fault data will be compared to the models obtained with the regional data set, and a model consistent with the near-fault, regional seismic data, and continuous and campaign-mode GPS data will be developed. A Jackknife station substitution method will be used to assess the resolution of the kinematic model. Using finite-differences we will examine the effect that 3D fault zone structure has on the ground motions at stations located close to or are in the fault zone, and also compute the slip weakening constitutive relationships for the rupture. Reference Eberhart-Phillips, D., and A. J. Michael, Three-dimensional velocity structure, seismicity, and fault structure in the Parkfield region, central CA, J. Geophys. Res., 98, 15,737-15,758, 1993.
DE: 7200 SEISMOLOGY
DE: 7203 Body waves
DE: 7209 Earthquake dynamics (1242)
DE: 7215 Earthquake source observations (1240)
SC: Seismology [S]
MN: Fall Meeting 2005