Detection, localization and moment tensor solution for an offshore slow earthquake

Previous studies (Dreger et al., 1998 and Tajima et al., 2002) have investigated the capability of a sparse broadband network, the Berkeley Digital Seismic Network (BDSN), at monitoring a region located outside the network, and tested the feasibility of a new automated moment-tensor determination system that continuously monitors seismic waveforms from the same network.

Figure 2.32: Low-frequency study of an offshore slow earthquake. Association of azimuthal rotation and seismic moment tensor grid search in the localization of an unusual event. The lines indicated the rotation axes plus or minus 3 degrees for each station.
\begin{figure}\begin{center}
\epsfig{file=guilhem_2_fig2.eps, width=6cm}\end{center}\end{figure}

Before using the moment and scanning method for events in Mendocino Triple Junction region, we need to test its feasibility for a recorded slow-earthquake. We have decided to apply it to a slow/low-stress drop event in December 2000 extracted from Goran Ekström's catalog. Ekström analysed and located it onshore in Northern California (Figure 2.32). However a first triangulation analysis of seismic records between 0.02 and 0.05 Hz from a few stations of the BDSN network (YBH, MOD, WDC, ARC and ORV) has given an offshore location. Despite the difficulty of identifying the body waves, analyses of P and Rayleigh wave particle motion have been done and were used in order to determine the azimuth between the seismic stations and the event. Such an azimuthal analysis has confirmed an offshore location (lines from seismic stations on Figure 2.32).

However, the uncertainty on location that we obtained is relatively large and depends on the quality of the stations used. We have completed our research by a grid search analysis of the seismic moment tensors using ORV, MOD, WDC and YBH to quantitatively locate the earthquake using the variance reductions from the computation. By this frequency waveform scanning and moment tensor grid search we are able to define a smaller area for the location of the event around the ridge segment between the Pacific and Gorda plates. Finally, the ANSS catalog location is similar to our location using this method, suggesting the potential success of a low-frequency continuous waveform scanning in locating events in the offshore region, as well as computing the seismic moment tensor. We also have to mention here that the moment tensor solutions indicate a high CLVD component with a normal mechanism which could be consistent with an explosion on the ridge segment. The ANSS catalog gives a $M_{b}$ 4.7 and a $M_{s}$ 3.9. With our method we have found a $M_{w}$ between 4.4 and 4.5 for the best moment tensor solutions.

With the goal of better constraining the location and mechanism of the offshore events, it would be important to consider a better coverage of broadband stations in the future. In the case of the earthquake in 2000, very few broadband stations were continuously recording. Some part of the analysis has considered the records at a station in Washington state, LON station (Figure 2.32). Today, many more high quality stations are operating in many places along the coast and can be easily used.

Berkeley Seismological Laboratory
215 McCone Hall, UC Berkeley, Berkeley, CA 94720-4760
Questions or comments? Send e-mail: www@seismo.berkeley.edu
© 2007, The Regents of the University of California