HR: 0830h
AN: S11D-0318    [PDF]
TI: Kappa at Permanent Stations of the Southern Great Basin Digital Seismic Network
AU: * Biasi, G P
EM: glenn@seismo.unr.edu
AF: University of Nevada Reno, Seismological Laboratory MS174, Reno, NV 89557-0141 United States
AU: Anderson, J G
EM: jga@seismo.unr.edu
AF: University of Nevada Reno, Seismological Laboratory MS174, Reno, NV 89557-0141 United States
AU: Brune, J N
EM: brune@seismo.unr.edu
AF: University of Nevada Reno, Seismological Laboratory MS174, Reno, NV 89557-0141 United States
AU: Smith, K D
EM: ken@seismo.unr.edu
AF: University of Nevada Reno, Seismological Laboratory MS174, Reno, NV 89557-0141 United States
AU: Kilb, D
EM: dkilb@ucsd.edu
AF: U.C. San Diego, IGPP/SIO MS0225, La Jolla, CA 92093
AB: We have undertaken a systematic study of spectral attenuation parameter kappa at permanent stations of the Southern Great Basin Digital Seismic Network (SGBDSN). Objectives included estimating kappa itself, reconciling disparate small and moderate earthquake estimates, and exploring magnitude and distance dependence of kappa. In seismic engineering, kappa can be the controlling parameter for site-specific seismic design amplitudes. We estimated kappa from the slope of S-wave spectral roll-offs relative to an omega-squared spectral shape, and for reference, an end-member model that assumes corner frequency effects can be ignored. Data include accelerograms and on-scale velocity recordings of over 40 earthquakes ranging in magnitude up to Ml 4.4 in and near the SGBDSN. To extend the magnitude range, a group of small earthquakes was also analyzed. Individual spectral fits exhibit a strong trade-off of corner frequency and kappa. Best-fit corner-frequencies may be identified but chi-squared significance tests and direct examination show that a wide range of corner frequencies and kappas would fit about as well. Fixing the stress drop predictably reduces scatter in kappa estimates. A linear distance correction improves the overall fit to kappa versus distance. However, systematic residuals in this model lead us to prefer a two-slope model where kappa does not increase with distance to ~35 km, then increases at about 0.27 msec/km thereafter. Physically the increase beyond 35 km may indicate attenuation for ray paths in the lower crust. Comparing kappa at sites on alluvium, tuff, and Paleozoic limestone does not suggest an obvious correlation with surficial geology. Kappa estimates for moderate earthquakes are roughly consistent with a previous study in the area, and average 19 to 22 msec, depending on how the distance correction is implemented. Slope kappa is generally larger for smallest earthquakes. Site kappa estimates are scattered over a range of +/-10 msec, even for similar source-receiver paths. Disagreement between small and moderate earthquake kappa estimates might be resolved if low kappa events are caused partial stress drop behavior of the source.
DE: 7212 Earthquake ground motions and engineering
DE: 7215 Earthquake parameters
DE: 7223 Seismic hazard assessment and prediction
SC: Seismology [S]
MN: 2003 Fall Meeting