Three-Dimensional Anomalous Characteristics Inversion Of Fixed Loop Source Transient Electromagnetic Method

During the last some decades the transient electromagnetic (TEM) method has proven to be efficient geophysical investigation method in mineral exploration. In interpretation of TEM data, the one-dimensional, horizontally layered earth is still one of the most useful models. 1-D models, however, are not appropriate if the subsurface conductivity changes laterally. Therefore, we should seek much more complicated earth structures, such as 2-D or 3-D models. In this paper, the study was focused on developing a 3-D anomalous characteristics inversion of fixed loop TEM method. The main results of the study are as below:Based on other inversions of TEM method, we have developed, for the first time, a 3-D anomalous characteristics inversion using a modified generalized inverse linear inversion, estimating parameters of the subsurface models. Furthermore, a 3-D anomalous characteristics linear inversion code has been programmed with MATLAB language and FORTRAN language.While inversing parameters of the subsurface models from the anomalous characteristics data, the Jacobian matrix was computed with backward difference replacing differential;a generalized inverse matrix of the Jacobian matrix was calculated using SVD method;considering the inversion is nonlinear, the inversion linear equation on the right was dealed with so that the error could be ignored. Finally, parameters of the subsurface models were estimated using a modified generalized inverse linear inversion.A new scheme that can be used to obtain the anomalous data from the observed profile data has been designed. A best layered-earth model was determined from the observed profile data using 1-D generalized inverse linear inversion. Then the transient responses of the layered-earth model were estimated. Finally, the anomalous data was determined, which was thedifference between the observed profile data and the layered-earth model transient responses. Some models have been computed and the anomalous data was obtained. The data was analyzed and some rules have been found.Some 3-D finite-difference modeling results have been compared with the analytical solutions and EM Vision 2. 1 computed results, verifying our modified code with MATLAB language and FORTRAN language over a homogeneous half-space. The comparison results demonstrate that the modified 3-D finite-difference modeling code is available. Furthermore, some 3-D models synthetic data has been inversed using the developing 3-D anomalous characteristics inversion. The results show the inversion method and programmed code are correct.