Research Of The 2D/3D Magnetotelluric Forward And Inversion With Induced Polarization Effect

Magnetotelluric sounding(MT)is a geophysical method for studying the internal structure of the Earth using natural field sources.It obtains information about the underground electrical structure by measuring orthogonal electromagnetic field components on the ground.MT is characterized by low detection cost,wide applicability,lightweight equipment,and high work efficiency.It is widely used in many areas,including geological structure research,underground water exploration,energy(petroleum,natural gas,geothermal,coalfields),metallic mineral resource exploration,engineering surveys,archaeological research,and environmental monitoring.The signals observed by MT are caused by both conductive currents and polarized currents in the geological body.However,existing MT inversion methods only consider the electromagnetic effects of underground media,neglecting the Induced Polarization(IP)effect.Using IP effects in rocks and ores can significantly improve the search for metal deposits(such as sulphide and oxidized metal deposits).Existing induced polarization methods need to provide artificial field sources to stimulate the IP effect of the geological body,such as dual-frequency induced polarization,DC induced polarization,and spectral induced polarization.These methods are more complex than MT and have a lower exploration depth,usually only a few hundred meters.If IP information could be extracted from MT observation data,additional polarization rate parameters could be obtained beyond the underground resistivity parameters,saving the exploration cost of traditional IP exploration methods and increasing their exploration depth.The challenge is that natural sources are weaker than artificial ones,which makes the IP signal weak in the observed data and coupled with the electromagnetic induction signal.Identifying and extracting IP information from MT observation data is key to this technology.In this study,we used an inversion method that introduced a complex resistivity model to extract the induced polarization effect.The forward modeling method replaced the real resistivity in Maxwell’s equations with the Cole-Cole complex resistivity model,considering both the IP effect and the electromagnetic induction effect of the geological body in the equations,and then used numerical methods to solve the equations.This study examined the influence of various parameters of the Cole-Cole model on forward modeling responses.Through forward modeling simulations,we found that the anomalies caused by zero frequency resistivity and polarization rate parameters are significantly larger,several orders of magnitude higher than those caused by frequency-related coefficients and time constants.To address the increased non-uniqueness of inversion due to increased parameters,we used an inversion method based on a cross-gradient structure constraint for multi-parameter inversion of magnetotelluric data considering the IP effect.For two-dimensional and three-dimensional problems,we conducted research on two and three-dimensional inversion methods considering the IP effect based on OCCAM inversion theory and Nonlinear Conjugate Gradient(NLCG)inversion theory,respectively.Given the high computational time and storage space requirements of three-dimensional forward and inversion,we researched and provided a parallel computing scheme that greatly reduced the consumption of forward and inversion time.Then,we used the inversion algorithm proposed in this study for model calculations.The results show that in the inversion of simple models and two-dimensional complex models,the inversion results almost restored the theoretical models;in the inversion of three-dimensional complex models,the inversion results can basically delineate the position and numerical range of the anomalies.Finally,this paper explores the practical application of 3D MT survey data from the Cascadia region on the west coast of the United States.Using the method proposed in this study,the polarization rate information was extracted from the 3D magnetotelluric data of this area for the first time,and the polarization rate profile was drawn,which showed the distribution characteristics of the underground polarization rate in this area.Based on the polarization profile inferred from magnetotelluric data,it is inferred that the lower crust of the active tectonic region in the western United States has a low-resistance layer on the west side containing graphite or sulfide metamorphic rocks,and the east side contains metallic magma and high-salinity fluid mixtures;it is inferred that the magma channels below the volcanic arc area contain abundant metal sulfides.