Calculation of slowness-azimuth station corrections for the IMS seismological networks

The International Monitoring System (IMS) of the Comprehensive Nuclear-Test-Ban Treaty Organization was designed so that, when fully deployed, it could detect with high confidence and accurately locate nuclear explosions with a yield of about 1 kt TNT equivalent in the absence of special efforts at evasion [1]. Also, an indicative metric for event location precision is that the confidence ellipse area around the estimated event location should not exceed 1000 km², the maximum area for an on-site inspection request in case of a suspected Treaty violation, specified in the Treaty. The IDC, therefore, needs to meet standards for both event detection and location precision. The event detection standard is most of the time easily met during automatic processing for energetic events thanks to the sufficient number of time-defining detected phases (i.e., phases whose arrival time isused for locating an event), however for events with few detecting stations, azimuth- and slowness-defining phases play a significant role. Furthermore, the more the defining features used in event location, the higher the location precision, i.e. the smaller the confidence ellipse, is. Therefore, accurate arrival directional information —slowness and azimuth— values can be very important, especially for small, weak events.  

About 25 years ago, after the first few years of operation, systematic biases of measured slowness and azimuth (deviations of measurements from predicted values) of the stations that comprised the IMS at the time were observed. They were attributed largely to lateral heterogeneity and slowness and azimuth station corrections were calculated to mitigate them [2]. Since then, some of these stations have been relocated, had their instruments re-calibrated or replaced and in some cases had their orientation modified. Also, new stations, for which no corrections have ever been calculated, have been installed. There is therefore a need to calculate or in some cases update these corrections. To do so we have considered good quality events (events with body wave magnitude ≥ 4, at least four detecting stations and azimuthal gap < 180°) reviewed by IDC analysts. We analyse time-defining phases with slowness and azimuth measurements and calculate corrections for prespecified slowness and azimuth bins. We also calculate station-specific trends to use as default corrections for bins for which the number of observations is insufficient to draw statistical conclusions. Finally, we calculate bin-specific and default uncertainties (modelling errors) as the spread of the residuals for bins. We compare the new corrections for some stations to the ones calculated previously and discuss differences and similarities. We also assess the effect these corrections would have on event definition, phase type identification and few-station event location.  

 

[1] National Academy of Sciences (2002), Technical Issues Related to the Comprehensive Nuclear Test Ban Treaty. The National Academies Press. https://doi.org/10.17226/10471 

[2] Bondar, I., North, R. G., Beal G. (1999), Teleseismic slowness-azimuth station corrections for the International Monitoring System Seismic Network, Bulletin of the Seismological Society of America, 89, pp.989-1003, https://doi.org/10.1785/BSSA0890040989