HR: 0830h
AN: S21F-0406    [PDF]
TI: Shallow Seismic Trapping Structure in the San Jacinto Fault Zone, California
AU: * Lewis, M A
EM: malewis@usc.edu
AF: University of Southern California, Department of Earth Sciences, 3651 University Avenue, SCI 117, Los Angeles, CA 90089-0740 United States
AU: Peng, Z
EM: zpeng@terra.usc.edu
AF: University of Southern California, Department of Earth Sciences, 3651 University Avenue, SCI 117, Los Angeles, CA 90089-0740 United States
AU: Ben-Zion, Y
EM: benzion@usc.edu
AF: University of Southern California, Department of Earth Sciences, 3651 University Avenue, SCI 117, Los Angeles, CA 90089-0740 United States
AU: Vernon, F
EM: vernon@epicenter.ucsd.edu
AF: University of California, San Diego, Institute of Geophysics and Planetary Physics, Scripps Institution of Oceanography, 9500 Gilman Drive, La Jolla, San Diego, CA 92093-0225 United States
AB: We analyze a waveform data set generated by 385 events and recorded by linear seismic arrays across the Clark Valley and Coyote Creek faults in the trifurcation area of the San Jacinto fault zone (FZ). The goal is to obtain structural information from a comprehensive analysis of FZ trapped waves in the data. A previous work based on selected waveforms suggested a low-velocity waveguide continuous to a depth of at least 18 km (Li and Vernon, JGR, 2001). If so, events located clearly outside the FZ proper should not generate any FZ trapped waves. On the other hand, a shallow FZ waveguide can produce (Fohrmann et al., PAGEOPH, 2003) trapped waves for events clearly off the fault. Our analysis of trapped waves in the larger considered data set is compatible with the existence of shallow non continuous waveguide layers along both the Clark Valley and Coyote Creek faults. Ben-Zion et al. (GJI, 2003) found in the context of the North Anatolian fault that many events off the fault produce FZ trapped waves and suggested that a better term for such data may be FZ related site effects. Within a distance of roughly 90 km, a subset of 159 events including many off the fault are suitably recorded by the arrays for analysis. A spectral ratio method is used to calculate the concentration of seismic energy within the FZ stations, producing a systematic measure of the quality of trapped waves or FZ related site effects. We find that FZ site effects are observed at FZ stations for the majority of the earthquakes, with the generating events located at various distances from the fault trace and various angles and distances from the receivers. The distribution of the events implies that the trapping structures are not continuous along the strike of either fault branch and do not extend bellow the depth of the shallowest events (e.g., 5 km). A travel time analysis of the difference between the direct S and trapped wave group arrivals shows no systematic increase with hypocentral distance or event depth, compatible with shallow discontinuous waveguide structures. Synthetic waveform fits to the data generated using the 2D analytical solution of Ben-Zion \& Aki (BSSA, 1990) indicate propagation distances within the low velocity FZ layer of around 5km, supporting the other analyses. We obtain good fits to the data produced for the Coyote Creek fault using a model consisting of a single low velocity layer in a half space. However, on the Clark Valley fault the waveguide appears to be weaker and the synthetic waveform fits are not as good. A feature of all the waveform modeling for both the arrays is that the waveguide center is located to the NE of the surface trace.
DE: 7209 Earthquake dynamics and mechanics
DE: 8010 Fractures and faults
DE: 8123 Dynamics, seismotectonics
SC: Seismology [S]
MN: 2003 Fall Meeting