| Magnetotelluric method has been played an important role in the detection of the crust and upper mantle structure,make the long-period and broadband magnetotelluric method combines the land of ultra-wideband electromagnetic magnetotelluric method is a further increase detection depth of the new techniques,this method solve the deep problems of Earth Sciences have been got many successful examples of the application. Current domestic long-period MT data for the practical stage pretreatment system is still in the blank, in the actual production or processing software relying on foreign countries, so by the application of mature technologies and integration, anatomical, and imitation of foreign excellent software, thus form an independent software system for the style of long-period magnetotelluric method in the application and development of our great practical significance.This paper analyzed and summarized on the basis of research results at home and abroad, aiming at the long-period electromagnetic data processing, in which several issues still need further studying, that is long-period magnetotelluric method of time domain noise suppression and time-frequency analysis, impedance tensor’s estimation and impedance tensor decomposition, and put forward specific technical ideas and programming to try to establish a quantitative inversion of the original from the instrument prior to the time domain sequence to the apparent resistivity/apparent phase as practical information on data processing.Hilbert-Huang transform (Hilbert-Huang Transform, HHT) is the latest developed powerful tool to process nonlinear, non-stationary signals. This article will introduce HHT to denoising and time domain frequency analysis of the natural electromagnetic field original time series, based on the significance of departure from the practice, and analysis the process of theoretical principles and algorithms, we made a programming, which makes data processing convenience. In addition, HHT has a strong time-frequency signal resolution and positioning capabilities on non-stationary signals, it can detect the noise in the time domain, and is conducive to data selection. Make denoising and selection the measured data before and after for comparison, it shows it’s a significant practical applications.Estimation of the impedance tensor magnetotelluric is to obtain information on underground distribution of electrical structure, this is the first and important steps. In this paper, we study the recent foreign literature in terms of impedance tensor estimation of new ideas, reference to foreign software PRC_MTMV to achieve the long-period magnetotelluric method in time domain series to the power spectrum calculation, to estimate the impedance tensor, and finally have reliable elements of the impedance tensor, and we described the methods of the theory and implementation process in detail. Consider the traditional robust estimation have limit effort for the suppression of "input" related noise, we proposed the magnetic field control of the remote reference method (Remote Reference with Magnetic Control, RRMC) to suppress related noise; and then, we introduced the Repeated Median Estimator to estimate the long-period magnetotelluric impedance tensor, instead of the traditional algorithm based on Robust M estimation algorithm, which improved the time domain raw data noise tolerance, this algorithm make the theoretical collapse value from30%to50%. The calculation process in the Matlab programming platform, we use the procedure to process measured data collected in Longmen Mountain region with Lemi-417instrument, it obtained a good result.The geological structure is generally three-dimensional, before make two dimensional inversion, overcoming the maximum impact of three-dimensional of local distortion, which is the key to success inversion interpretation. Based on various physical impedance tensor decomposition method, not only helps to correct distortion, extract the regional tectonic features, but also can break down a variety of parameters derived for the interpretation of underground electrical axis direction, geological background-dimensional structure, various distortion characteristics have significant effect. In this paper distorted impedance tensor sophisticated analytical techniques, reducing the under two-dimensional three-dimensional inversion interpretation distortion effect of two-dimensional model.the analysis of three dimensional/two dimensional model based on the distortion characteristics, deduced Swift rotation, Barr decomposition, Groom-Baily decomposition of the mathematical process, and present a new optimized solve algorithm of Groom-Baily decomposition, the computation of the theoretical model, demonstrated the application scope of the three methods and their calculations are correct; for the impedance tensor elements can be applied to explain the qualitative characteristics of the geological structure, we make a program of the polarization chart analysis, the impedance tensor main axis’s azimuth and Mohr circle analysis, it was applied to the measured data to determine the regional geological background information. Finally, given the measured data is often subject to static effects, the impedance tensor of a static level of effect correction methods were studied and realized.Finally, we applied the thesis research results to long-period magnetotelluric data processing, which was collected in the Longmen Shan seismic zone, after decomposing distorted impedance tensor and analyzing qualitative data, we obtained electrical spindle orientation and characteristics of the regional dimension, and through the two-dimensional inversion, we got the region’s deep electrical structure, and then discusses the dynamic characteristics of that region. |
PDF Full Text Request