| In MT surveys, most of actual structure of the Earth which we concerned is oftenthree-dimensional. However the current state-of-the-art for MT data or LMT datainterpretation is two-dimensional inversion and three-dimensional inversion whichmainly depends on the forward’s efficiency and the strategy of inverse on will becomeavailable in the near future. Therefore, the main method of data inversion is still basedon two-dimensional. When interpretation of3D data with2D techniques, someproblems are still exist, Such as how to analyze the feature of3D galvanic distortion,how to understand2D approximate interpretation of3D models, how to use physicalmethod and mathematical method to overcome the influence of3D local distortion asmuch as possible, how to use impedance tensor decomposition techniques to get morereliable geological results, how to make full use of apparent resistivity and phase ofTE mode and TM mode, and how to deal with the influence of terrain and the effectof static. These are the focus of concern on theoretical research and the core issueswhile the developing of magnetotelluric sounding method.3D/2D model, the popular models we mainly researched refer to thethree-dimensional local anomaly superimposed on the two-dimensional areaconstructed. It will produce a great impact on the inversion theory developing fromthe two-dimensional interpretation techniques to the three-dimensional throughtheoretical analysis and numerical calculation. Based on3D/2D model, calculating thegoverning equation for the EM field of local distortion, deducing the integralexpression of anomalous electric and magnetic field, discussing the model of thephysics hypothesis for local distortion and the expression of impedance distortiontensor. Galvanic distortion and induction distortion in classic classification of local distortion and their effect on apparent resistivity. Some features of local distortioncaused by small-scale surface inhomogeneous are briefly discussed. Through theanalysis of mathematical principles of local distortion, what can provide theoreticalsupport to various impedance tensor decomposition method, also can guide all of thismethods to determine the parameters and its scope of application.The invariants of impedance tensor occupy a pivotal position in the process ofmagnetotelluric data analysis, which mainly used in geo-electric dimensionalityanalysis. Basic invariants of impedance tensor, WAL rotational invariants, B Qrotational invariants with its dimensional determination methods are introduced, basedon those theory, developed a program to analyze the magnetotelluric dimensional;some issues needing attention using these invariants are indicated such as Swift,Bahr, GB, Chave Smithand phase tensor, the merits and shortcoming of thesemethods are point out; developed some programs including analyze the spindleazimuth, analyze GB signal site, signal frequency correction algorithm, calculatedistortion factor; The qualitative analysis of impedance tensor such as polar plots andMohr are reviewed in detail, also develop a program based on polar plots to performqualitative analysis; Static effect is a frequency encountered problems in themagnetotelluric data processing, how to reasonably remove the impact of static fromregional structure is an important part of the data processing. Analyze two differentways of removing the static effect, one based on the high frequency factor ofimpedance tensor, the other based on wavelet transform, both of this two ways wereprogrammed and packaged, use the measured data showed that the two ways can playa role with the visual display.The mathematical decomposition approach of impedance tensor derived from theproperties of the matrix, its decomposition results were applicable to any geologicalmodel because of completely divorced from the physical model, and the parameterswere correct in any case. After analyzing the mainstream mathematical approach suchas feature decomposition method, singular value decomposition method, modifiedsingular value decomposition method and canonical decomposition method, theapplication situation and used characteristics of each method were discussed from thelevel of mathematical theory. According to analysis of above approach, indicated thatthe last three ways developed from the singular value decomposition method, with theanalysis of magnetotelluric model, the formula was corrected constantly, so that it can be better adapted to the requirements of the geological structure. After derived fromthe singular value decomposition method, developed a program which used toanalysis the measured data, use the image showed that the canonical decompositionmethod can obtain the maximum apparent resistivity and minimum apparentresistivity without rotating the impedance tensor, and can capable get the dimensionalinformation from the side.Refer to data process of long period magnetotelluric, as well as analyze commonproblems existed in the processing of measured data, develop a set of data processingsoftware handle distortion decomposition and calibration of impedance tensor whichcompatible deal with magentotelluric data, long period magnetotelluric data, andbroadband magnetotelluric data. After import original sounding data, can performsome function include analyzing the magnitude of impedance tensor, mergingbroadband data, mapping apparent resistivity and apparent phase curve, analyzingqualitative of curve mode, analyzing constructed dimensional of word area, showingspindle azimuth, rotating the sounding curve, solving local distortion factor, achievingcanonical decomposition, removing static effect, all of this function were packaged ina set of program module, what formed a data process software. Through the analysisof the measured data and the results of demonstrated, indicated that this software cancomplete the processing of the measured data, after verified through numericalsimulation and inversion calculation, proved that the impedance tensor analysissoftware’s correctness and rationality on the aspect such as the analysis ofimpedance tensor, decomposition of local distortion correction algorithm,magnetotelluric data processing method. |
PDF Full Text Request