A Complex Method for Automatic Detection of Geoacoustic Emission Pulses Preceding Earthquakes

Abstract

Currently, the use of acoustic emission method is widely used in various fields. The method consists in detecting, recording and analysis of elastic vibrations that arise as a result of plastic deformation of solid media. The Laboratory of Acoustic Research of IKIR FEB RAS. The laboratory actively uses this method to monitor the geoacoustic emission occurring in the sound range during rock deformation. Perturbations of geoacoustic emission signals are of great interest for research and can often act as operational predictors of earthquakes. Geoacoustic pulses are characterized by a complex internal structure, a wide variety of time-domain waveforms, large amount of noise and short duration. Such features together with a huge amount of incoming data exclude manual processing and require an effective way to detect impulse anomalies in automatic mode. This paper proposes an algorithm for automatic detection of geoacoustic pulses. The algorithm is based on the use of an adaptive threshold automatically adjusted to the noise level in the signal. Mechanisms for detecting various pulse anomalies are considered. A method for defining the boundaries of selected events is also described. The received algorithm was tested on real signals with different degrees of noisiness.