Application of statistical methods to take into account the influence of the terrain when conducting aeromagnetic surveys

Abstract

Introduction: The study presents a method for applying a new statistical approach to take into account the effect of the daylight terrain, composed of magnetised magmatic rocks, on the results of magnetic field measurements. Methodology: A computer-based technology is proposed that is based on the synthesis of the method of empirical mode decomposition (EMD) and the group method of data handling (GMDH). This technology is implemented in the processing of the data from aeromagnetic surveys, performed in difficult physical and geological conditions of the western part of the Putorana plateau. Results and discussion: The discrete values of the anomalous magnetic field ΔT and the relief elevations H were decomposed into empirical mode functions (IMF). The GMDH identification modelling was conducted for ten experimental profiles. Based on this data, the methodology for the further analysis was optimised. For each profile, for certain components of the magnetic field, unique equations of their correlation with the terrain were obtained, model values of the components were calculated, and the discrepancy between the primary and model values of the components of the magnetic field was estimated. Conclusions: The research results made it possible to apply the practices for implementing the new approach proposed by the authors to take into account the influence of the terrain on the data of aeromagnetic survey. The generalised model characteristics, obtained for the whole data set, are in agreement with the results from the experimental profiles. It proved that the preliminary estimations were correct and the proposed sequence of analysis was valid. The revealed interrelations between the components of the magnetic field and the relief made it possible to suppress the influence of the trap cover on the anomalous magnetic field and to distinguish weakly manifested geological bodies and structures.