The Study Of Restraining Environmental Noise And Its Application In Magnetotelluric Sounding

The purpose of Magnetotelluric Sounding (MT) is to measure the conductivity structuralof underground medium on different depth.The whole process includes data collection and processing and inversioning. Data collection is basic and crucial. The application of Magnetotelluric Sounding success or not depend on the quality of data greatly.How to suppress various disturbtion efficiently,raise signal-to-noise ratio, guarantee the data quality, that is the key problem in the collection and handling of MT data all along.Based on plenty of measured data, theoretical with practice, having studied the MT signal feature and the influence law of environmental noise to MT data, (electromagnetism disturb with static and topography influence) for MT information law after, put forward or studied various methods to restrain environmental noise, raise the data quality. Pass through application example finally, have confirmed the practicality and validity of these methods.Analyse more than ten thousands sites actual measured Mt data have gotten the characteristic of natural electromagnetic field used in MT method, and research compare with the global average amplitude of electromagnetic field. Brought forward many methods to study the noise in MT data. Studied the noise source, frequence characteristic influence law to MT data systemically from uniform angle. Lead theoretical foundation of CSAMT into exploration for MT in electricity disturb and magnetism disturb study. Confirm the signal-to-noise ratio of magnetic field will be better than electric field in theoretically. Point out the electromagnetism signal produced by human campaign not be noise completely, in which, part has formed useful signal. Analyse through error theory, have gotten the mathematics relation of phase error and apparent resistivity error. This conclusion has guidance meaning to make Magnetotelluric Sounding (MT) standard as well as to establish inversion goal function of MT data. Have studied the influence law of geological noise to MT data with the method of sense degree analysis.Paper have studied the influence of error in the information of actual measurement to inversion result. Know the error transmiting law clearly, as well as the effect of static migration to inversion result andthe MT observation frequency distribution interval reasonability problem.Theoretic analysis nearly all method of corrections of static migration that have carried out by now combined with example, have made objective comment and pointed out the application condition and exist problem of them. Study and analyse the effect law of static migration to apparent resistivity curve, impedance phase curve, electric principle axe, two-dimensional diverge degree and etc. Study from the mechanism of static migration, this paper put forward a static migeration correction method-tensor impedance correction method. It correct the tensor impedance, at the same time all MT parameters calculated from tensor (apparent resistivity curve, impedance phase curve, electric principle axe, two-dimensional diverge degree and ellipticity and etc) will be corrected. Thus,these parameters can correctly reflect geological information, and solve the problem that other corrected methods only corrected the apparent resistivity.Analyse from signal correlation angle, paper have studied the far reference method principle systemically, correct some wrong knowledge of far reference method, through specific experiment(different reference signal road test,different reference distance test) and corresponding theoretic anylysis, have gotten more systematic technical parameter. Put forward that reference distance should be 14 times depth of the designed exploration goal for the first time. Point out that electric filed single can also be used as far reference road and it have no relationship to the geological character of the reference site. Paper affirms far reference method for suppress noise have certain superiority, but at the same time, point out far reference method can not change the data quality i...