Spatially Constrained Joint Inversion On Magnetic Resonance Sounding And Transient Electromagnetic Method

The shortage of water resources and the frequent occurrence of water-induced disasters have severely restricted economic development and national production and construction,which threaten the safety of life and property.In order to alleviate water stress and reduce the occurrence of water-induced disasters,the combined detection method of magnetic resonance sounding(MRS)and transient electromagnetic(TEM)has been widely applied as an important means to detect water.However,conventional1-D and 2-D joint inversion methods are faced with bottlenecks in joint exploration of groundwater with complex formation structure and 3-D shape because of the limitations of the information acquisition for topographic and parameters of adjacent measurement points.In addition,factors such as system configuration,geological parameters,and data quality seriously affect the accuracy of the joint inversion results,which directly limit the depth detection resolution of aquifers,and greatly increase the difficulty of groundwater exploration in complex geological environments.In order to address the above problems,this paper focuses on the spatially constrained joint inversion on MRS and TEM.The main results are as follows:1.To solve the problem of slow forward calculation for MRS in traditional layered conductive geological models,a fast calculation scheme for induced magnetic field based on Shanks transformation recursion is proposed.In order to reduce the difficulty of calculation,the continuous integral is discretized first.Then the horizontal coordinate is modified by the Gaussian quadrature,and the partial integrals irrelevant to the variable parameters are calculated and saved in advance.Finally,the Shanks transformation is used to recursively calculate the divergent integral.This method improves the calculation efficiency for the MRS forward modeling and lays a foundation for realizing multi-point spatially constrained joint inversion.2.To solve the problem that conventional 1-D and 2-D inversion methods cannot accurately obtain spatial information for the groundwater,a quasi-3-D spatially constrained joint inversion method on MRS and TEM is proposed.Based on the inversion for the data of the independent measurement points,the spatial constraint matrix is designed to impose smooth continuous constraints on the parameters according to the spatial correlation of the physical properties and electrical structures of the adjacent measurement points.In addition,a multi-objective weighted genetic algorithm is proposed.The self-adjustment of crossover probability and mutation probability is optimized.The objective functions of MRS-TEM data,the parameter spacial constraint and the weight coefficient are simultaneously introduced into the inverse evolution process,so as to find the optimal joint inversion parameters and solve the multi-solution problem in isolated measurement point inversion.3.In view of the problem that the detection system parameters directly affect the depth resolution of groundwater detection,a scheme of uncertainty analysis of inversion parameters is proposed.Based on the joint inversion model,the parameter covariance matrix is established,and the uncertainty level is divided according to the calculated standard deviation factor.In this paper,multiple simulation models are set up,and the influences of important system parameters and geological parameters on the uncertainty of the inversion results are analyzed.In this way,the ultimate effects of various parameters on the inversion results are comprehensively and deeply evaluated,and the optimization adjustment direction of each parameter is further clarified.This is of profound significance for perfecting the actual detection process and improving the accuracy of groundwater inversion results.4.To solve the problem of inversion deviation caused by conventional inversion methods relying on observation data to fit water content information,a spatially constrained joint inversion method on MRS and TEM based on the uncertainty factor correction is proposed.The uncertainty factor is introduced into the parameter optimization process of joint inversion,which makes the inversion evolution direction under the current device conditions and geological parameters is guided by the uncertainty of model parameters,rather than limited by observation data.The effectiveness of the method is verified by simulation cases analyses.5.The method studied in this paper is applied to the task of investigating hidden water hazards of Halagou Coal Mine.According to the actual detection target,environmental noise and geological parameters,the optimal settings of the instrument system are obtained.Finally,the imaging of quasi-3-D underground water-bearing structure in complex detection environment is realized.The results are consistent with the geological data and the drilling data,which prove the practicability of the proposed joint inversion method.The spatially constrained joint inversion method on MRS and TEM studied in this paper can perform inversions for the joint detection data in a large survey area.The 3-D distribution of underground water content and resistivity can be located and quantitatively displayed by means of spatial constraint and parameter uncertainty analysis.As an effective data interpretation method,it provides a theoretical basis for the search for groundwater resources in a large area and the early warning of water-induced disasters,and provides technical support for water resources prospecting in western China and national safe production and construction.