Statistical Features of Georadar Signals Obtained by Probing Horizontal Layered Media in Permafrost

Abstract

Relevance. To study the structure and properties of rock masses in the permafrost zone, such a method of geophysics as ground penetrating radar is successfully used. The existing methods of GPR measurements make it possible to obtain large volumes of reliable data on the layered structure of permafrost rocks. The methodological support for processing and interpreting data is developed much worse. Thus, manual and semi-automatic methods are used to identify the in-phase axes of georadar signals confined to the boundaries of rock layers, which significantly slows down the process of cameral work. Existing methods for automating the selection of in-phase axes of georadar signals do not always successfully cope with this task, since the statistical characteristics of georadar traces obtained by probing a horizontally layered media are very variable. The purpose of the research is to establish the statistical features of georadar signals that form the in-phase axes obtained by probing horizontally layered media in permafrost. Research methods – statistical analysis of the results of field georadar measurements in the Anabar region of the Republic of Sakha (Yakutia) at the site of the Mayat diamond deposit (AO Almazy Anabara) using the GPR OKO ABDL-Triton. Conclusions. The research of the amplitude characteristics of GPR signals that form the in-phase axes showed the presence of a random component, which significantly affects the recorded GPR wave field. Using the results of field measurements of horizontally layered media as an example, a change with increasing depth of the main statistical characteristics, as well as the distribution of amplitude values (normal–uniform–normal) was established. The results obtained will be very useful both in interpreting georadar measurement data and in improving the accuracy of modeling the propagation of ultra-wideband electromagnetic waves in horizontally layered media, and can also be used to refine the previously developed model of a georadar section of a frozen rock mass.