An Improved Impedance Tensor Decomposition Method And Its Applications

For the impedance tensor distortion, resulted by horizontal galvanic heterogeneity near surface, it is necessary to decompose the distorted impedance tensor, in order to describe true electrical structural feature. If the distortion of the impedance data is not eliminated, the calculated apparent resistivity and phase may be wrong, which effluence on the result of data processing and interpretation.In this thesis, based on the distorted impedance tensor feature of 3D/1D or 3D/2D model, we discuss several distortion decomposition methods. By analyzing the process of GB decomposition method, an improved method is introduced, using the generalized inverse matrix algorithm to calculate regional impedance from the data in which the distortions have been eliminated. A computer program to carry out this method is written with Matlab, succinct and clear. By forwardly calculating a theoretical model under the effluence of the distortion factors, we detect the error of the generalized inverse matrix algorithm and prove it to be effective. Finally, the improved method is applied into the calculation of regional impedance from some given magnetotelluric data, which proved to be practical and effective.The method introduced in this thesis can be used to calculate the distortion factors. Therefore, the quantitative relation between the distortion factors and regional impedance can be studied. Gain enlarges or reduces the amplitude of tensor impedance. Principal axis direction of Strike is compressed and perpendicular axis is stretched if anisotropic parameter is a positive value. Oppositely, perpendicular axis is compressed, and Principal axis direction of Strike is stretched. Tensor impedance rotates in clockwise direction under positive twist parameter. Negative shear parameter makes tensor impedance rotating in anti-clockwise direction.