Detection and Characterization of Mining Activity at a National Level

Session: Explosion Seismology Advances
Type: Oral
Date: 4/29/2020
Time: 08:45 AM
Room: 240
Description: 

The USGS National Earthquake Information Center (NEIC) earthquake bulletin contains all detected earthquakes magnitude 2.5 and larger within the US and its territories and magnitude 4.5 worldwide. Within the US, the NEIC bulletin augments the more complete Advanced National Seismic System ComCat earthquake bulletin by providing earthquake locations and magnitudes calculated using consistent methodologies. For the NEIC bulletin to be useful for seismotectonic research and earthquake hazard studies, it must correctly differentiate between routine mining explosions and tectonic events nationally. Potential contamination of mislabeled events from mining explosions is particularly severe in the eastern US where some regions have more than 100 mining explosions for every tectonic event. Mining explosions also complicate earthquake monitoring by introducing numerous phase picks into the association process. These picks can generate spurious detections that increase manual processing time.

Here we present initial results of a large-scale study using cross-correlation template matching to detect and characterize mining activity primarily in the central and eastern US. These techniques have the advantage over event-based techniques, such as regional P-to-S ratios, of only requiring a single station to robustly differentiate between explosions and earthquakes. For this study, we generate templates using waveforms from earthquake clusters and/or mining explosions detected using amplitude-based STA/LTA phase picks and traditional association techniques. Waveform templates are generally constructed from the closest few stations for each of the earthquake clusters and/or mines. Currently, real-time detectors are running on 616 event clusters using 4,875 three-component waveforms, which includes monitoring of 34 active mines across the contiguous U.S. We will examine the efficiency of these detectors for finding subsequent mining events and study the temporal stability of the correlations. We will show how this technique can be integrated into the NEIC earthquake processing pipeline.

Presenting Author: Paul Earle

Authors