Microevent Detection Based on Waveform Cross-correlation in the Dogye Mining Area, Korea

Abstract:

We have studied induced seismicity associated with Dogye coal mine located in the eastern part of Korea. From May 2009 to March 2014, 222 events that occurred at the mining area were reported in our catalog with local magnitudes ranging from 0.6 to 2.4. For 67 events we can observe that the epicenters relocated by the double difference technique with Lg waveform cross-correlation image location of the six clusters classified according to waveform similarity. On May 2014 a broadband seismometer is installed in the mine office to understand seismicity of the mining area. We cross-correlate continuous data of the installed station recorded from May 2014 to April 2015 with a comb-like waveform observed regularly. The comb-like waveform with length of 30 to 60 minutes is a signal train composed of a blast every 30 seconds. We consider the comb-like signal related directly to mining activity from the fact that the signal train appears averagely four times a day on weekdays with its monotonic amplitude. Besides the comb-like signal, events with an irregular occurrence time and amplitude is detected from the one-year continuous record of the installed station. We suggests that most of the undefined events are formed from fracturing in response to stress-perturbation on an active mining face or represent slip in existing shear zones such as a fault or dike.