HR: 1330h
AN: S42C-0173    [PDF]
TI: Quantifying properties of seismic spectra: an examination of 100's of spectra from southern California earthquakes recorded by the ANZA seismic network
AU: * Kilb, D
EM: dkilb@ucsd.edu
AF: UCSD/IGPP/SIO, MS 0225, La Jolla, CA 92093
AU: Biasi, G
EM: glenn@seismo.unr.edu
AF: Univ. Nevada Reno, Seismology Laboratory 174, Reno, NV 89557
AU: Brune, J
EM: brune@seismo.unr.edu
AF: Univ. Nevada Reno, Seismology Laboratory 174, Reno, NV 89557
AU: Anderson, J
EM: jga@unr.edu
AF: Univ. Nevada Reno, Seismology Laboratory 174, Reno, NV 89557
AU: Vernon, F L
EM: vernon@epicenter.ucsd.edu
AF: UCSD/IGPP/SIO, MS 0225, La Jolla, CA 92093
AB: Unraveling earthquake source properties from seismic spectra requires an understanding of the inherent uncertainties that can be introduced from site and path effects. Here, we quantify properties of hundreds of earthquake spectra, and the associated uncertainties, from ~0$<$Ml$<$~3 earthquakes in southern California recorded by the ANZA seismic network. From the spectrum that meet our data-quality and data-fit requirements we: (1) Determine the influence of site effects on the spectral decay characteristics; (2) Estimate the distance that site effects can be reliably extrapolated using data from small aperture arrays; and (3) Investigate the hypothesis that earthquakes behave as partial stress drop events. For each seismogram with an analyst-identified S-wave arrival, we test a range of viable corner frequencies and identify the one that minimizes the squared error fit of the data spectrum to the synthetic spectrum, which is computed assuming a Brune (1970) omega-square fall off. From this we can then compute the associated stress drop, seismic moment and AH-kappa, where AH-kappa is the residual slope to an omega-squared spectrum (Anderson \& Humphries, 1991). Small values of AH-kappa indicate a greater percentage of high frequency energy. Using our highest quality data, we find that AH-Kappa measurements from sites on hard-rock can, in general, be ~20 ms smaller than those from sites on unconsolidated rock. The uncertainty in AH-Kappa at a single station is ~10 ms; assuming the site effect is constant for common source-receiver paths, we attribute this variability to deviations in the seismic source. Based on the theory of Brune (1970), we then used an iterative method that makes spectral corrections to account for potential partial stress drop events. Because the spectral shape of the partial stress drop model is less than omega-square model we expect the AH-kappa values to increase, and indeed we find this to be true. For an assumed full stress drop of 100 bars, our results typically show an increase of ~10 ms in AH-kappa measurements, a value that is consistent with the ~10 ms observed scatter. If variable stress drops are responsible for the 10 ms scatter in AH-kappa measurements, then, for engineering purposes, an additional ~10 ms should be added to the mean observed AH-kappa values. Using AH-kappa to summarize small-scale site effects shows a +/-10 ms variation in AH-kappa measurements from individual event recordings at stations that are only ~10s of meters apart. This implies that small-scale (~10s of meters) heterogeneous site properties play a key role in a seismogram's frequency content.
DE: 7200 SEISMOLOGY
DE: 7209 Earthquake dynamics and mechanics
DE: 7215 Earthquake parameters
DE: 7230 Seismicity and seismotectonics
SC: Seismology [S]
MN: 2003 Fall Meeting