Estimating the relative yields of Novaya Zemlya tests by waveform intercorrelation

The relative size of 15 underground nuclear explosions at the Novaya Zemlya test site is determined by applying a relative waveform analysis called intercorrelation to 566 teleseismic short-period P-waves. The waveforms are corrected for the effects of pP interference and yield scaling of the explosion source functions so that signals from different events recorded at the same station can be analytically compared. For events within a single test site, this procedure accounts for common path and receiver effects. Relative explosion source strengths in the 0.5 -2.0 Hz frequency band determined by intercorrelation are consistent with the results of conventional mb analysis. Absolute yield estimates from both intercorrelation and mb analyses have much greater uncertainty due to unknown baselines. The yield estimates for the 15 events range from 36 to 3886 k t , assuming that tE=0.5 s and that yield scaling relations for the Amchitka test site are appropriate for Novaya Zemlya. The pP delay times obtained from intercorrelation of Novaya Zemlya events are similar to the delay times for Amchitka events with equivalent yields, consistent with both being hardrock sites. The analytical waveform comparisons also provide criteria for identifying anomalous events. The Novaya Zemlya test site has two subsites separated by about 300 km. One of the 11 events in the Northern subsite and one of the four events in the Southern sub-site have anomalous waveforms. A waveform modelling study indicates that the anomalous Southern event was a double explosion. Northern and Southern Novaya Zemlya have different waveform and amplitude patterns, apparently caused by different focusing and defocusing of seismic energy by laterally varying structure near the sources.