» Return to Regional

Qinya Liu has developed and implemented a Centroid Moment-Tensor (CMT) inversion procedure to determine source parameters for southern California earthquakes. The method is based upon spectral-element simulations of regional seismic wave propagation in a recently developed three-dimensional southern California model. Sensitivity to source parameters is determined by numerically calculating the Frechet derivatives required for the CMT inversion. We use a combination of waveform and waveform-envelope misfit criteria, and perform pure double-couple or zero-trace moment-tensor inversions. The technique is applied to five recent southern California earthquakes: the September 9, 2001, Mw=4.2 Hollywood event, the October 31, 2001, Mw=4.9 Anza Event, the September 3, 2002, Mw=4.2 Yorba Linda event, the February 22, 2003, Mw = 5.2 Big Bear event, and Mw = 4.5 Big Bear Aftershock.Using more than half of the available three-component data at periods of 6 seconds and longer, the focal mechanisms, locations, and magnitudes we obtain are in good agreement with estimates based upon classical body-wave, surface-wave, and first-motion inversions.

(a) The inversion results table for Yorba Linda event.

9/29/06" border="0" alt="">

(b) The waveform fits of different parts of the seismograms, including the Pnl parts of the vertical and radial components, surface wave parts of the vertical and radial components (Rayleigh wave) and the transverse components (Love wave), the two numbers below the station names are the station distance and azimuth, the three numbers to the right of each trace are maximum amplitude in units of 0.01 mm, station correction in seconds for the best correlation and the best correlation value (between 0 ~ 1)

Conclusions

1. We perform centroid moment tensor (CMT) inversions for recent Southern California earthquakes using the spectral-element simulations of ground motion. The mechanisms and locations obtained are in good agreement with independent body- wave and surface-wave inversion results.
2. Because we obtain god fits to the data at periods of 6 seconds and longer, we can determine the mechanisms of events with magnitudes Mw >= 3.5.
3. We are implementing an automated procedure to routinely determine CMT solutions for events in Southern California.
4 . The time shift information collected for each event will be used to further improve the Southern California velocity structure.