Enhancing manual P-phase arrival detection and automatic onset time picking in a noisy microseismic data in underground mines

Abstract

Accurate detection and picking of the P-phase onset time in noisy microseismic data from underground mines remains a big challenge. Reliable P-phase onset time picking is necessary for accurate source location needed for planning and rescue operations in the event of failures. In this paper, a new technique based on the discrete stationary wavelet transform (DSWT) and higher order statistics is proposed for processing noisy data from underground mines. The objectives of this method are to (i) improve manual detection and picking of P-phase onset; and (ii) provide an automatic means of detecting and picking P-phase onset time accurately. The DSWT is first used to filter the signal over several scales. The manual P-phase onset detection and picking are then obtained by computing the signal energy across selected scales with frequency bands that capture the signal of interest. The automatic P-phase onset, on the other hand, is achieved by using skewness- and kurtosis-based criterion applied to selected scales in a time-frequency domain. The method was tested using synthetic and field data from an underground limestone mine. Results were compared with results obtained by using the short-term to long-term average (STA/LTA) ratio and that by Reference Ge et al. (2009). The results show that the method provides a more reliable estimate of the P-phase onset arrival than the STA/LTA method when the signal to noise ratio is very low. Also, the results obtained from the field data matched accurately with the results from Reference Ge et al. (2009). Keywords: Manual P-phase detection, Automatic onset picking, Noisy microseismic data, Kurtosis, Skewness