Abstract

Most ripple-fired explosions have a unique signature in the spectrum that can be used to quickly identify them when monitoring the Comprehensive Test Ban Treaty. We want to know the importance of high-frequency (>16 Hz) seismic data in identifying these explosions. The deployable seismic verification system (DSVS) in Wyoming records data up to 50 Hz with good signal to noise, so it is an excellent source for determining the benefits of recording data over 16 Hz. We developed a database of 118 known earthquakes and 176 known ripple-fired explosions. We calculated a binary spectrogram and an average binary spectrum for each event and developed criteria based on the number of peaks and nulls above certain thresholds to classify events as ripple-fired explosions. Using data at frequencies up to 50 Hz, 50% of the known ripple-fired explosions were correctly identified, and another 22% had banding in the spectrogram but did not meet the established criteria. The other 27% of the known ripple-fired explosions had no banding. To test the importance of high-frequency (>16 Hz) data, we recalculated spectrograms between 0 and 16 Hz for the 135 known ripple-fired explosions and earthquakes that had banding. Eleven of the 135 could not be identified correctly if data over 16 Hz was excluded; 5 had information at frequencies over 16 Hz that was critical to seeing any banding. But the vast majority of events were correctly identified in the pass band of 0 to 16 Hz, suggesting that higher-frequency data do not significantly help to identify most ripple-fired explosions in Wyoming.

Return to Table of Contents