Identifying quarry blast events in seismicity catalogs based on the ground motion parameters

English Abstract:

Discrimination of quarry and mine blasts from natural seismicity or induced seismicity is a preliminary and fundamental step in compiling a seismicity catalog. It is particularly challenging to discriminate blasts from potentially-induced events, as both induced events and blasts will be characterized by shallow depths. Such discrimination can be done by examining waveforms and seeking information on nearby quarry operations. However, such an approach is tedious and time-consuming, requiring manual review by a seismologist. Using purely statistical approaches is more efficient, but may not be as reliable, as some natural tectonic or induced events may be falsely discriminated as blasts, while some blasts may not be properly identified. In this study, we investigate whether we can identify quarry blasts by an automated analysis of index ground motion parameters. We hypothesize that ground motion parameters from blasts may have definitive characteristics that could be used in automated discrimination procedures. To test the hypothesis, we analyze the data from the TransAlta/Nanometrics seismic network in Alberta, Canada. We focus on the parameters of peak ground acceleration (PGA), peak ground velocity (PGV), pseudo-spectral acceleration (PSA) at different frequencies, and their vertical to horizontal-component ratios. These ground motion parameters are commonly calculated in routine, automated data analysis; therefore, they can be used in a process that automatically identifies quarry blasts. Having a reliable automated process that is based on the spectral characteristics is very useful for the compilation of reliable seismicity catalogs in near-real time.