Abstract
Seismic records produced by different seismic sources vary. In this study, we compared the waveform records and time-frequency characteristics of tectonic earthquakes, artificial explosions, and mine collapses in China’s Capital Region. The results show that tectonic earthquakes are characterized by stronger S-wave energy than P-wave energy, obvious high-frequency components, and wide frequency bands of P and S waves. Artificial explosions are characterized by greater P-wave amplitude than S-wave amplitude and near-station surface wave development. Mine collapses are characterized by lower overall frequency, more obvious surface waves, and longer duration. We extracted quantitative discriminants based on the analysis of different event records, with 31 feature values in 7 categories (P/S maximum amplitude ratio, high/low frequency energy ratio, P/S spectral ratio, corner frequency, duration, the second-order moment of spectrum, and energy strongest point). A comparison of the ability of these feature values to recognize distinct events showed that the 6–17 Hz P/S spectral ratio was able to completely distinguish artificial explosions from the other two types of events. The S-wave corner frequency performed relatively well in identifying all three types of events, with an accuracy of over 90%. Additionally, a support vector machine was used to comprehensively distinguish multiple features, with an accuracy for all three types of events reaching up to 100%.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have