This paper summarizes the use of regional (Δ ≦ 30°) seismic data in a test ban context for detecting, locating, identifying, and determining the yield of underground nuclear explosions. In many areas of the world (Eastern North America, Africa, Eastern USSR), Lg is the largest amplitude wave recorded on standard seismograph systems and thus is the most appropriate phase for monitoring small magnitude events. Excellent location capability for near-regional events has been demonstrated at the Norwegian small aperture array (NORESS) using Lg and P waves. Lg and other regional phases may contain information on source depth, but such information has not been exploited to date. Fifteen classes of regional discriminants have been identified including
Excitation of Short-Period SH Waves
Lg/Rg Amplitude Ratios
Pn/Lg, Pg/Lg, and Pmax/Lg Amplitude Ratios
Lg Group Velocity and Energy Ratios in Lg
Excitation of Sn
Third Moment of Frequency
Generation of Higher Mode Surface Waves
Peak Amplitudes of Love and Rayleigh Waves and Long-Period Surface Wave Energy Density
Prevailing Period of Long-Period Love Waves
Spectral Ratio—Long-Period S Waves to Rayleigh Waves
Spectral Ratio—Long-Period Love Waves to Rayleigh Waves
Frequency of the Peak Spectral Amplitudes and Spectral Ratios in Pn, Pg, S, and Lg.
Each of these proposed discriminants has, with differing degrees of success, separated some explosions from some earthquakes. However, most have been tested only on limited data, usually from one geographic region and only one or two recording stations. No systematic analyses have been done to determine the best individual discriminant or combination of them.
Preliminary evaluation of the use of Lg for yield determination suggests that regional waves hold promise in this application. Theoretical studies have contributed significantly to the understanding of propagation characteristics of regional waves but further studies are required emphasizing modeling for realistic anisotropic sources.
The major conclusion of this study is that a systematic and comparative evaluation of all the proposed regional discriminants is now required, utilizing a common data base derived from all present-day test sites. This evaluation would suggest the optimal discrimination procedure using regional waves, and would also define areas of needed research. Without such an integrated evaluation, it is still possible to speculate, using existing results, on the most promising regional discriminants.