THE APRIL-1992 CAPE-MENDOCINO-EARTHQUAKE SEQUENCE - SEISMO-ACOUSTIC ANALYSIS UTILIZING FIXED HYDROPHONE ARRAYS

The oceanic T-waves of earthquakes associated with the 1992 Cape Mendo cino earthquake sequence were recorded and analyzed using fixed hydrop hone arrays located throughout the northeast Pacific Ocean. The T-wave s of these events were well recorded with high S/N ratios and strong a coustic energy present over a 0-64 Hz bandwidth. The smallest event re corded by the hydrophone arrays from the sequence had a local magnitud e of 2.4. The hydrophone records of the three largest shocks in the se quence (M-L 6.9, 6.2, 6.5) exhibited both T-waves and lithospheric pha ses from these events. Low-pass filtering (2 Hz) of the lithospheric p hases yielded a clear P-wave arrival for epicentral distances of <10 d egrees, but no apparent S-wave. A seafloor cable-break was detected im mediately after the second M > 6 aftershock, possibly the result of a submarine slide. The direct P-wave hydrophone records from the second large aftershock showed a relatively high-amplitude, high-frequency ar rival, consistent with seismic analyses which used this information to infer rupture direction. The rupture direction was toward the locatio n of the cable break, thus rupture directivity possibly played a role in initiating the slide event. Modelling of the T-wave propagation pat h, using the Parabolic Equation model, produced estimates of the acous tic transmission loss from epicenter to receiver. The transmission los s to the most distant phones is typically 10-20 dB, and can be as larg e as 50-70 dB for acoustic propagation paths that cross the continenta l margin. The amount of acoustic energy each earthquake released into the ocean at the seafloor-water interface was estimated applying the t ransmission loss and instrument response to the recorded T-wave signal s. This acoustic source power level was calculated for 41 events with magnitudes over a recorded range of 2.4 less than or equal to M-L less than or equal to 6.9, with 17 of these events having their seismic mo ment estimates available through the NEIC. Ground displacement spectra were estimated from the acoustic power spectra and showed no indicati on of a corner frequency. Thus empirical analyses relating source leve l to magnitude and seismic moment were necessary to quantitatively der ive an earthquake's size from hydrophone records. The results of indic ator variable regression analyses suggest that T-wave source level inc reases linearly with the event's local magnitude and seismic moment. F urthermore, the source power level versus magnitude relationships for oceanic and continental earthquakes are significantly different, proba bly illustrating differences in the seismic and acoustic propagation p aths from hypocenter to the hydrophone receivers. The results indicate that acoustic measurements provide a reasonable estimate of magnitude and seismic moment of an oceanic earthquake that was not detected by land-based seismic networks.