Application of a wavelet technique for the detection of earthquake signatures in the geomagnetic field

We developed an algorithm especially adapted to single-station wavelet detection of geomagnetic events, which precede or accompany the earthquakes. The detection problem in this situation is complicated by a great variability of earthquakes and accompanied phenomena, which aggravates finding characteristic features of the events. Therefore we chose to search for the characteristic features of both "disturbed" intervals (containing earthquakes) and "quiet" recordings. In this paper we propose an algorithm for solving the problem of detecting the presence of signals produced by an earthquake via analysis of its signature against the existing database of magnetic signals. To achieve this purpose, we construct the magnetic signature of certain earthquakes using the distribution of the energies among blocks, which consist of wavelet packet coefficients.