AMT Forwarding Modeling And Inversion Algorithms And Software Development

Abstract:Magnetotelluric is a method using the nature electromagnetic as source, utilizing the skin effect of the signals of different frequencies to achieve the purpose of geophysical exploration; Audio-frequency magnetotelluric(AMT) research on the response to the audio section of electromagnetic signals. Obviously, for any geophysical methods, forward modeling is the extremely foundations for method researching and inverse methods aim to processing and interpretation of field data on the basis of the forward modeling. This paper studies the AMT forward modeling and inverse methods to provide theoretical support for the AMT data processing and interpretation software.Firstly, starting from the AMT theory, using finite element method (FEM)with bilinear interpolation to compiled the two dimensional AMT forward modeling. Secondly, from the inversion theory, study the Bostick inversion, Occam inversion and Active Constrain Balancing algorithm regular inversion method, respectively, implement the inversion of AMT using bostick inversion, Occam inversion and ACB inversion numerical simulation. Finally, we test them in models, study the underground media with a simple and several Anomalies, study the static effect causing of the topography, and test the field data in our program.The results show that:the finite element method using a bilinear interpolation achieves a good forward modeling results; the Bostick inversion can approximated reflect the one-dimensional layered medium, Occam inversion can reflect the one-dimensional medium well and approximate reflect two-dimensional model; ACB inversion can inverse a good results in the case of topography, the ACB inversion method achieve the expected results testing the field data.