Application of Robust Phase Algorithms for Seismic Emission Detection in the Area of Blasting Operations in Mines

In this paper, we developed algorithms for multichannel processing of seismic array records that allow detecting signals of weak seismic events observed against the background of natural seismic noise. The algorithm for detecting signals of weak seismic events is based on the coherence assessment of longitudinal waves (P-waves) recorded by a small aperture array. The advantage of this algorithm over the known single-channel short/long term averaging (STA/LTA) detector is the ability to detect time intervals of P-waves at a signal-to-noise ratio (SNR) less than unity. Based on the time intervals of the array records with the registered coherence measure values that exceed a certain threshold, the apparent slowness vector is estimated by using the phase bimforming algorithm that is robust to noise (Zhang et al., 2008; Kushnir and Varypaev, 2017). The developed algorithms of multichannel processing were tested using real seismic records of a small aperture array installed in the Kursk Magnetic Anomaly area near the mine workings for iron ore extraction. Signals of both strong (underground industrial explosions) and weak seismic events were detected for 20 h of seismic recording. In some cases, seismic events were shown to have a complex source mechanism. Using the Capon filter (Capon, 1969), for weak seismic events, P-waves with a duration not exceeding 0.2 s and a temporal shape similar to the known function of a Ricker pulse (Bording, 1996) are found to have resemblance.