A Joint Inversion Of The Two-mode Passive Source Electromagnetic Sounding Data Based On Background Subtraction Method

Passive source electromagnetic sounding method has increasingly been used in practical work for its unique advantages. However, the inversion method is lag behind which restricts the level of data interpretation results. With the support of National High Technology Development "863" Project-"Multi-scale Self-adapting Two-mode Joint Inversion of Passive Source Electromagnetic Sounding", This article researched a two-mode joint inversion method of the passive source electromagnetic sounding data based on background subtraction.This paper studied a rapid two-mode joint inversion of the passive source electromagnetic sounding data based on the idea of minimum structure. The main focus is the fast computing of the sensitivity matrix and the background subtraction inversion method based on the theory of wavelet multi-resolution decomposition. By theoretical analysis, numerical simulation, model test and practical application, the method of improving the inversion accuracy with making effective use of joint two-mode data is researched; meanwhile, speed is also discussed.Firstly, the objective function of the joint inversion of the two-mode passive source sounding with model constraints is built. Model parameters and data residual errors are restricted individually with different weights which can use various kinds of data effectively to get a smoother model. For the practical situation of the passive source sounding, a double grid of forward and inversion is designed which accelerate the inversion speed without loss of forward significance, moreover, it's easy for operation and management. On this basis, the two-mode joint inversion of passive source electromagnetic sounding data based on minimum structure is implemented by programming. The effects of individual and joint inversion with four kind of sounding data is analysed by inversion test to seven typical geoelectricity model, which lead to some significative conclusions. The inversion method is applied to real data of a mining area in Inner Mongolia in the following part. Secondly, on the basis of the analysis of one-dimensional magnetotelluric sensitivity function, by using numerical simulation method, high-frequency magnetotelluric sounding on two-dimensional model is taken for example to show the relationships between the sensitivity of various observational data and model resistivities, frequency and spatial coordinates, and the conclusion summarized is that the sensitivity will decay fast when spatial coordinates changed in the condition of a constant skin depth. With this conclusion, a method of fast computing the sensitivity matrix by limiting the rang with skin depth is presented and applied to the passive source electromagnetic sounding two-mode joint inversion and proved by model test that it can greatly accelerated the inversion speed.Thirdly, based on the theory of Mallat wavelet multi-resolution decomposition, a discussion and computing formula of obtaining background information from the surface electromagnetic response of the passive source electromagnetic sounding method is provided and verified by test. On this Basis a two-mode joint inversion method of passive source sounding based on the background subtraction is presented. The inversion of the background electromagnetic response is done individually while the result resistivity is imposed on the background outside the abnormal regions of the model for restricting the inversion, only the abnormal regions changes will be calculated. Model tests show that this method can effectively reduce the redundant structure and the elapsed time.Finally, the main work and innovations involved and some limitations in current research are pointed out and recommendations for future research are given in the conclusion part.