The Feature Of Three-Dimensional Distortion In MT And Research Of New Correction Methods

In MT surveys, the actual structure of the Earth is often three-dimensional (3D). However the current state-of-the-art for MT data interpretation is two-dimensional (2D) inversion and 3D inversion will become available in the near future. When interpretation of 3D data with 2-D techniques, some problems are still exist, such as how to analyze the features of 3D galvanic distortion, how to overcome the influences of 3D local distortion as much as possible and get more reliable geological results, how to understand 2D approximate interpretation of 3D models, and how to make best use of apparent resistivity and phase of TE mode and TM mode. These are the focus of concern scientific problems and the key factors which delay the developing of MT. This thesis focuses on the problems in the above paragraph. MT responses of local distortion are computed using a 3D finite-difference on a staggered grid code. The effect of local distortion has been studied. New correction algorithm and new parameters for determination of geoelectric dimensionality are brought out. The study about 2D approximate inversion versus 3D data is done. The research is divided into two parts: one is the difference between 3D and 2D responses; another is validity of 2D inversion results. At last the research results of this thesis are applied to the interpretation of filed data. The conclusions of this thesis are summarized below.1. Basic theory and summary of research situation of galvanic distortion in MTBased on the governing equation for the EM field of local distortion, the integral expression of anomalous electric and magnetic field are deduced. The physics hypothesis for local distortion model is discussed and the expression of impedance distortion tensor and vertical magnetic distortion vector are derived from integral equation in detail. Galvanic distortion and induction distortion in classic classification of local distortion and their effect on apparent resistivity and phase curves of TE mode and TM mode are illustrated. Some features of local distortion caused by small-scale surface inhomogeneous are briefly discussed. Basic invariants of impedance tensor and WAL rotational invariants used in geoelectric dimensionality analysis are introduced. Some methods and B-Q parameters for determination of geoelectric dimensionality based on these invariants are presented, and some issues needing attention using these invariants are indicated. Several approaches close with this thesis such as Swift, Bahr, GB methods are reviewed in detail. The merits and shortcoming of these methods are point out. These work provide a research beginning for developing novel distortion correction algorithm in this thesis. The bug in formula describing distortion of electric and magnetic fields is fixed.2. Three-dimensional modeling and analysis of local distortion in MTBased on theory in the first part, several 3D local distortion models are designed and model responses are computed using a improved finite-difference on a staggered grid code. Qualitative analysis for model responses is done and some new insights are obtained. When apparent resistivity curve of TM mode across small conductive surface body, the curve will upward shift at first and then downward shift. While across resistive body, the curve will downward shift at first and then upward shift.. This phenomenon is associated with secondary electric field caused by the accumulative surface electric charges at the boundaries of the body. When small body is resistive or conductive, the polarity of the charges results in a secondary field additive to the primary field or opposing the primary fields. Upon vectorial addition of the primary field and secondary field, one can estimate the sense of distortion in the resulting total electric fields. In the resistive body, the total field is increased directly over the body, it is decreased off the ends, and an increase occurs along the sides of the body. With a conductive body, the total field is reduced directly over the body, it is enhanced off the ends of the body, and it is diminished along the sides.When the strike direction of 2D regional structure is NS direction, in the plane map of apparent resistivity of YX mode, 3D conductive body is elongated in NS direction, two zones with the values increasing are located respectively at east and west boundary. While in the plane map of apparent resistivity of XY mode, 3D conductive body is elongated in EW direction, two zones with the values increasing are located respectively at north and south boundary. For phase of XY mode, 3D conductive body is elongated in EW direction; two zones with the values decreasing are located respectively at north and south boundary. For phase of YX mode, 3D conductive body is elongated in NS direction; two zones with the values decreasing are located respectively at east and west boundary. This phenomenon is consistent with the case that curves of 2D TM mode across the electric boundaries. When the profile runs along EW direction, YX mode response in 3D corresponds to TM mode response in 2D. If the profile runs along NS direction, XY mode response in 3D corresponds to TM mode response in 2D.The distortion on 2D regional response caused by small conductive inhomogeneous are more severe than that of resistive inhomogeneous. The distortion on TM mode response is slightly larger than that on TE mode caused by resistive body. While for conductive body, the distortion on TE mode response is larger than that of TM mode. In the same mode the distortion on phase is less than that on apparent resistivity.At the position located directly over the small inhomogeneous, the difference between 3D distortion response and 2D corresponding regional response is much large. But this 3D distortion response can be equivalent to other 2D response. It is more important to pay more attention to this special distortion in MT interpretation.3. Research of new correction method for local distortion in MTPhase tensor approach put forward by overseas geophysicist is introduced and it is the first time applied to the interpretation of synthetic data and field data at home. New single site, single frequency correction algorithm (SSSF) and single site, multi-frequency algorithm (SSMF) are brought out. The formula for them are present in detail. Based on these research, new parameters for determination of geoelectric dimensionality are defined and a program to perform MT distortion analysis is developed. Firstly, synthetic data is used to check the validity of new methods. Secondly, quantitative analysis on 3D local distortion response is done using Swift, Bahr, GB, phase tensor, WAL, SSSF and SSMF approach. Finally, new methods are applied to field data. The research results manifest that the application of new methods are comparable to or even superior to other decomposition methods. The parameters for determination of geoelectric dimensionality and distortion factors provided by new methods are important complement to the analysis of MT data. According to problem occurred in application for WAL method, WAL invariants criteria is corrected.4. The choice of polarization mode in 2D inversion versus 3D MT distortion dataIn order to understand the difference between 3D response and 2D response, answer the questions how to understand 2D approximate interpretation of 3D models, and how to make best use of apparent resistivity and phase of TE mode and TM mode. the study about 2D approximate inversion versus 3D data is done. The research is divided into two parts: one is the difference between 3D response and 2D response, another is validity of 2D inversion results.The results manifest that in the 3D scenario the demand for approximate 2D structure in TE mode is much stricter than in TM mode. When 3D effect is relatively severe, by applying 2D inversion to 3D data, the results show that the inversion model using only TM mode data is more validity, and the phantom structures are less than using TE mode or TE+TM mode. In TM mode, the distortion on phase is less than apparent resistivity, so the weight of phase in inversion can be increased. For 2D inversion of field data, TM mode is first choice, next is TE+TM, TE mode is not proposed. The physics analysis basis is provided in this chapter.5. Application of research results in filed dataThe research results of this thesis are applied to the interpretation of Megu-Suijiang field data in eastern margin of Tibetan plateau. The dimensionality analysis is carried out by phase tensor, SSSF method and WAL invariants criteria. The strike direction of regional structure is determined by SSSF approach. SSMF method is used in local distortion correction. TM mode data is used in 2D inversion. Some new approaches are adopted in inversion, one is Site-Center Mesh (SCM) which used in constructing initial model, another is the regularization parametersτwhich determined through segmental variation. The validity of inversion model is checked by using effective skin depth of complex structure method and invariants of phase tensor. A good inversion results is obtained and it has laid a solid foundation for the construction of electric structure and geological interpretation.