Method for Constructing a Model of a Geoelectric Section Taking into Account Seasonal Variations Based on Data from Long-Term Vertical Electric Sounding Monitoring in Search of Earthquake Precursors

Abstract

The study analyzes data from high-precision measurements of the apparent resistivity by a stationary multielectrode vertical electric sounding (VES) system including 12 current and 4 potential lines spaced 2–650 m apart. Observations had been being carried out at the Garm test area on a daily basis for 12 years in an earthquake prediction experiment. The use of special technical methods during measurements ensured an instrumental error of about 0.01%. The virtual error of each individual measurement of apparent resistivity (taking into account all possible noise) was 0.1–0.2%. The availability of more than 3000 VES curves measured in different seasons allows us to propose a new approach to constructing a geoelectric section model. To solve the inverse VES problem, a set of 36 averaged 10-day VES curves was analyzed, each of which was obtained by averaging approximately 100 individual VES curves accumulated in the same 10-day period of the annual (seasonal) cycle in different years. Comparative analysis of these curves made it possible to calculate and include corrections for stationary geological noise in the model. As a result, it was possible to substantially reduce (by an order of magnitude) the discrepancies in fitting the curves and dramatically narrow the equivalence domain. Based on the results of our analysis, we have constructed a model of a four-layer horizontally layered geoelectric section of the Khazor-Chashma depression to adequately describe not only the averaged section, but also its seasonal variations throughout the year. The stability in estimating the model parameters is studied. To further reduce the equivalence domain, we propose that the layer thicknesses be fixed. This model can be used not only to study the aforementioned characteristics of the section, but also to monitor time variations of resistivity in individual layers of the section. This will significantly improve the resolving power of systems for detecting time variations in geoelectric sections, including when searching for earthquake precursors.