Long-Period Microtremor Observations in the Santa Clara Valley, California

David Dolenc and Doug Dreger

Introduction

The 3D velocity structure of the Santa Clara Valley (SCV) was previously investigated by modeling the teleseismic P-waves (Dolenc, 2001) recorded by the 41 seismic stations of the SCV Seismic Experiment (USGS/UCB/PASSCAL, 6/98-12/98). To complement these results, we now focused on the microseisms that were recorded during the same SCV seismic experiment.

Results

Microseisms are generated by the pressure variations on the sea floor due to the ocean waves and can be observed in the 0.1 to 5 Hz frequency range. We first compared the noise level of several earthquake-free periods to the ocean wave heights recorded at the Santa Cruz weather buoy and showed that the two are correlated (Dolenc, 2001).

Horizontal to vertical (H/V) spectral ratios of microtremor signals for a 5-day earthquake-free period were then calculated for each station. The H/V spectral ratios for the 5-minute segments at the beginning of each hour were first calculated and then averaged over the 5 days. The results showed that the dominant period of the H/V spectral ratios in the 0.1 to 1 Hz frequency range is stable with time and is location dependent (Figure 20.1). The longer periods can be observed for the stations above the two basins and the shorter periods for the stations in-between the basins. The period of the H/V peaks as a function of the basin depth from the USGS model (Jachens, 2000) is shown in Figure 20.2.

There was no correlation between the amplitudes of the H/V spectral ratios and the USGS model basin depths.

In addition we applied the H/V method to the two local earthquakes recorded during the SCV seismic experiment (San Juan Bautista, $M_{L}$=5.4, and Gilroy, $M_{L}=$4.0). The results showed that for most SCV seismic stations the peaks of the H/V spectral ratios in the 0.1 to 1 Hz frequency range coincided with microseism H/V spectral ratio peaks.

The observations of the microseisms on the SCV stations show that the presence of the basins can be observed even for the time periods without earthquakes. Future work will include modeling the response of the sedimentary layers using the 1D structure from the USGS model under each SCV seismic station.

Figure 20.1: The periods of the dominant H/V spectral peaks in the 0.1 to 1 Hz frequency range. The numbers indicate locations of the SCV seismic stations and the circles indicate the periods of the H/V spectral peaks. Stations that showed no peak in the 0.1 to 1 Hz frequency range are shown by squares. Contours of the basins from the USGS model at 1 km, 3 km, 5 km, and 6 km are shown in gray. Dashed lines are the active faults in the region.

Figure 20.2: The period of the dominant H/V peaks as a function of the basin depth from the USGS model. Stations located above shallower than 4 km basins were used in regression and the resulting line fit is shown (y=0.93x+2.73, R-squared=0.61).

Acknowledgements

This research was supported by the USGS grants 99HQGR0057 and 00HQGR0048. The USGS velocity model provided by Robert C. Jachens of the U.S. Geological Survey was used. Ocean wave data were obtained from the National Data Buoy Center.

References

Dolenc, D., Basin structure influences on the teleseismic wave propagation in the Santa Clara Valley, California, MS Thesis, University of California, Berkeley, 2001.

Jachens, R.C., Personal communications, 2000.