Slowness–azimuth corrections of teleseismic events for IMS primary arrays in China

The seismic arrays at Hailar (HILR) and at Lanzhou (LZDM) in China are both primary stations of the International Monitoring System for verifying compliance with the Comprehensive Nuclear Test Ban Treaty. These two stations became operational in 2002 and have since then provided continuous data. In this study, the so-called slowness–azimuth station corrections (SASC) were derived and used to improve the location accuracy of the two arrays. The SASC are found by comparing the back-azimuth and slownesses obtained from array processing to the theoretical values calculated from the reported event locations and the corresponding seismic velocity model. Events reported by the National Earthquake International Center in the time period 2002 to 2006 were used as reference events, and the IASP91 was used as the theoretical velocity model. Small correction vectors with random orientation were found for HILR. Larger correction vectors with systematic vector biases were found for LZDM. The LZDM correction vectors seem to point to the same direction in a large part of the slowness space and may be attributed to local structure. After introducing the SASC for HILR, the standard deviations of back-azimuth and slowness residuals drop from 7.1° to 4.6° and from 1.0 to 0.6 s/°, respectively. For LZDM, these values drop from 22.3° to 10.2° and from 2.9 to 1.1 s/°, respectively. The variations of back-azimuth and slowness residuals were reduced by 32% and 30.2%, respectively, for HILR after SASC and the reductions were 21% and 40.2% for LZDM. The improvements were 77% in back-azimuth and 67% in slowness location for HILR and were 79% and 81% for LZDM after SASC.