Analysis of Regional Seismic Data for Underground Explosions

In recent years, numerous regional seismic event discrimination studies have shown that the high-frequency (usually greater than 3–4 Hz), P/S spectral amplitude ratios of regional P (Pg or Pn) and S (Sg, Sn or Lg) waves observed at distance range from few kilometers up to about 2,000 kilometers, provide an efficient method to classify regional earthquakes, quarry blasts and underground nuclear explosions. The high-frequency P/S spectral ratios of verticalcomponent records, P/S spectral ratios of rotated, three-component regional records, as well as the P/S spectral ratio of the three-component regional records corrected for the free surface effect have been used depending upon the availability of data. The distance correction is also applied to P/S ratios which improves significantly the discrimination power of P/S ratios over uncorrected ratios, particularly if the events have data from wide distance ranges (e.g., 3 to 17 degrees). Distance corrected and network averaged P/S spectral ratios provide transportability of the spectral ratio method to various regions worldwide. However, the P/S spectral ratio method is region dependent and we must find the best phases, frequency bands, and distance corrections for each region by using training data set with wide distance range, azimuthal coverage, and event size. There are many regions worldwide where there are lack of training data set. In this study, we focus on data collection and analysis of such regions – Novaya Zemlya, Russia and Semipalatinsk Test Site, Kazakstan, for evaluating the P/S ratios for regional seismic discrimination for CTBT monitoring. We analyzed regional data from earthquakes, peaceful nuclear explosions, underground nuclear test around Novaya Zemlya, Russia to evaluate applicability of P/S ratio discriminant. In this region, data are available from the past underground nuclear explosions with magnitude mb(P) range from 4.5 to 6.5, while the useful earthquake data are limited to mb(P) 3.5 – 4.5. We report our analysis of the regional data to examine possible size (magnitude) dependence of regional high-frequency P/S spectral amplitude ratios as reported by Ringdal et al. (1999) for signals recorded at NORSAR from the Novaya Zemlya underground nuclear explosions. We also collected and analyzed regional data around Semipalatinsk Test Site, Kazakstan, where there are limited data available from few earthquakes. The regional data from the recent calibration explosions in Balapan and Degelen Mt. at various depth of burials are analyzed to shed light on whether the reported magnitude dependence of regional high-frequency P/S spectral amplitude ratios are due to yields of the explosions, or due to efficiency of S wave excitation from various burial depths of the sources.