Using Shear Wave Travel Time Correction Surfaces to Improve Regional Seismic Location in China


Ground Truth Comparisons

Figure 5 Events surrounding the 08 NOV 1997 Tibet Mw=7.5 (combined EDR, REB, ISC). Yellow events are those with S picks. Magenta events are those with LANL S picks.

As a means of testing the effects of shear wave PPCs on regional locations in China, we selected events including and surrounding the 1997 November 8 Tibet Mw=7.5 mainshock. This event produced a significant surface rupture [Peltzer et al., 1999] and has a vertical strike-slip mechanism [Velasco et al., 2000]. 56 events with S picks are located within the InSAR region shown in Figure 1. The combined EDR/REB/ISC catalog contains 154 S picks for 33 out of these 56 events. We were able to add/modify 102 S picks for 41 out of the 56 events giving a total of 256 S picks. We assigned a relative measurement error of 2 and 3 seconds for the P and S phases from global catalogs and 1 and 2 seconds for the LANL P and S picks, respectively, placing more emphasis on those picks in which we were the most confident.




Figure 6 - Relocations of combined EDR/REB/ISC and LANL databases using only stations with P phase PPCs. 90% confidence error ellipses are shown. When using only corrected P phases, the main shock relocates approximately to the surface rupture [Steck et al., 2000]. 43 of the 56 events converged to a solution. When using P and S phases but correcting only the P phases, the mainshock is positioned in approximately the original biased global catalog location [Begnaud et al., 2000]. 39 of the 56 events converged. Using and correcting P and S phases moves the mainshock at about the same location as the P defined/corrected map. 41 of the 56 events converged.



Figure 7 - Example of mispick of shear phase from the EDR global catalog. Phase misidentifications such as this bring into doubt the accuracy of the shear wave picks from global earthquake catalogs. Because a few incorrect S picks can significantly degrade the resulting relocation, it is extremely important that we obtain the most accurate shear arrivals possible.


Introduction | Generation of Propagation Path Correction (PPC) Surfaces | Ground Truth Comparisons
Ground Truth Comparisons Using LANL S Picks | Focal Mechanism Comparison | Conclusions and Future Work | References