HR: 1330h
AN: S52A-0122    [PDF]
TI: Determination of source, path, and site parameters from Yokohama High-concentration Seismograph Network
AU: * Tsuda, K
EM: kenichi@crustal.ucsb.edu
AF: Institute for Crustal Studies, University of California Santa Barbara, Institute for Crustal Studies, University of california SAnta Barbara, Santa Barbara, CA 93106 United States
AU: Archuleta, R
EM: ralph@crustal.ucsb.edu
AF: Institute for Crustal Studies, University of California Santa Barbara, Institute for Crustal Studies, University of california SAnta Barbara, Santa Barbara, CA 93106 United States
AB: The data from the Yokohama high-density accelerometer array have been used to try to separate path, source and site effects. The Yokohama array has 150 surface accelerometers and 9 borehole accelerometers located in a 20 km x 20 km region. The main objective of this study is to measure the variance of site effects for closely spaced sites on similar surficial geology. Since 1998, the array has recorded more than 30 earthquakes, 13 of these events were recorded at all borehole stations, and selected for our analysis (9 deep events and 4 shallow events). Using the data from these 13 events we estimate the seismic moment and corner frequency for each event, Qo, frequency dependence of Q, and the frequency dependent site effect. We use Boatwright's (1978) representation of the omega-squared spectrum for the source. The quality factor is assumed to be constant for frequencies less than 1.0 Hz while having a power law relation for frequencies greater than 1.0 Hz. Using a Heat Bath algorithm, we estimate the parameters for the source and path using only the borehole stations. Initially the borehole response is assumed to be unity and path attenuation fixed while Mo is determined by inverting the data for f$\dots$ 1Hz. The difference between the observed and predicted spectrum at each borehole is then used as the site response of the borehole records and Qo, Q(f) and fc are determined by inverting the borehole data from the 9 deep events, keeping Mo fixed to the previously determined values. We iterate until changes in the residuals between observed and predicted spectra are negligible. Once stable values of Mo, fc and Q(f) are obtained, we invert only the source parameters (Mo and fc) using the 9 borehole stations for each of the 9 events individually. The same procedure is applied for 4 shallow events. Knowing Mo, fc, and Q(f), we can then determine the site response for the 150 surface stations by averaging the ratio between predicted and observed amplitude spectra. The Q results show a very weak function of frequency for deep events: Q(f)=285f**0.055, however, events shallower than 20 km have a stronger dependence:Q(f)=80f**0.265.The results for site response correlate with the averaged S-wave velocity for upper 30m for the low frequency range.
DE: 7203 Body wave propagation
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