Multidimensional Wavelet Analysis of Time Series of Electrotelluric Observations in Kamchatka

Abstract

The paper presents results of joint multidimensional wavelet analysis of three series of variations in the electrotelluric potential observed at the Verkhnyaya Paratunka station in the Kamchatka Peninsula from October 1, 1996, to June 23, 2001. The analysis was made in order to identify common components in the signals analyzed and compare them with the seismic regime and variations in meteorological parameters. The analysis was based on the method of robust wavelet-aggregated signals, developed by one of the authors for monitoring problems. The average prognostic efficiency of the inferred anomalies is estimated. The analysis revealed the effect of frequency migration of the collectiveness measure peak in the behavior of the study series toward higher frequencies; this effect took place throughout the observation interval. Recently, specific features of this type in the behavior of geophysical characteristics have more often been regarded as a basically new class of strong earthquake precursors.