Application of empirical mode decomposition method in processing geophysical data

Abstract

The relevance of research is caused by the feasibility of using advanced mathematical methods in processing results of the geophysical surveys. The aim of the research is to improve the effectiveness of the methods of applied Geophysics through adaptive extraction of informative components of the physical fields using the Empirical Mode Decomposition method (EMD). The method of research: decomposition of the relevant geophysical data. The iterative computational process allows allocating different frequency orthogonal signal components, which are called empirical mode functions (IMFs). The method is designed to represent non-stationary signals in the form of a series of signals with different frequency. The authors propose the original algorithm in which the sourcewise approximation is used for constructing functions, enveloping the extrema of the analyzed signal, and the residual component of decomposition can identify the background component of the field. The results. The EMD algorithm was implemented and tested on practical materials for processing the data of gravity, magnetic and electrical prospecting. The authors defined the possibility of applying the method to separate the informative component of the results of electric methods of horizontal profiling (Perm Krai); to evaluate the quality of high-precision gravity survey (Western Sayan); to determine the component of the magnetic field caused by the impact of sharply dissected terrain of the earth surface in the development of the effusive traps (Norilsk region). The last case uses a series of regression dependencies between the various frequency components of the magnetic field and heights. The authors made the conclusion on appropriateness of using the EMD technique in applied Geophysics. The prospects of its application the authors see in analysis of geophysical monitoring material of mineral deposits development.

Keywords

• signal
• Empirical Mode Decomposition
• algorithm
• approximation
• field component
• gravity survey