Application Research Of Joint Inversion Of Vertical Electric Sounding Data And Rayleigh-wave Dispersion Data In The Ground Substrate Layer Survey

The vertical spatial survey of the ground substrate layer survey is a branch of the specialized survey of the ground substrate layer survey.The main objective is to identify the types and spatiotemporal evolution of the ground substrate layer composition within the vertical range of the work area,to construct a three-dimensional structure of the vertical spatial distribution of the surface matrix layer,and to provide data support for the evaluation of the prospects for vertical space development and utilization.Geophysical surveys are a part of the vertical spatial survey and are mainly used for the detailed structural layering of underground spaces within the survey area.This article,based on the project of the China Geological Survey Bureau: the survey of the ground substrate layer of black soil in the Lishu area of the Liaohe Plain(No.DD20211590),proposes the joint inversion of direct current resistivity method and Rayleigh wave method using swarm intelligence optimization algorithms.It relies on parameters such as resistivity,transverse wave velocity,and layer thickness to provide a reasonable interpretation of the layered structure of underground space,thus achieving the goal of fine layering.The article begins by discussing the significance of joint inversion of electrical and seismic methods,and explains why direct current resistivity and Rayleigh wave methods were chosen for the joint inversion.The authors review the applications of joint inversion and swarm intelligence algorithms in geophysical inversion,as well as the research status of vertical spatial survey of the ground substrate layer.They derive the direct current resistivity measurement depth based on linear filtering algorithms and the Rayleigh wave dispersion curve forward modeling process based on the fast vector transfer algorithm.Regarding the selection of swarm intelligence nonlinear inversion algorithms,the authors propose using both single-objective and multi-objective algorithms to jointly invert the direct current resistivity measurement depth and Rayleigh wave dispersion curves,based on the different forms of the joint inversion objective function.Given the strong nonlinearity,multiplicity,and numerous local optima problems inherent in both direct current resistivity and Rayleigh wave methods,the authors propose the COBL-IDACPSO single-objective algorithm and the L-MODSQPSO multi-objective algorithm with multiple strategy improvements.The two algorithms introduce a variety of strategies to adjust the optimization algorithm and improve the diversity of each algorithm population,enabling the optimal solution to quickly escape local optima and enhance the efficiency of random search in uncertain environments.Furthermore,the model settings were carefully considered by taking into account the common structures of the surface substrate sedimentary strata,as well as various scenarios including noise-free and noisy data,and known or unknown model layer numbers.The inversion capabilities of the single-objective and multi-objective algorithms were tested on the different scenarios.Sensitivity analysis was conducted to investigate the influence of each inversion parameter on the inversion results.From the complementary perspective of the sensitivity analysis of model parameters,the reasons for the improved inversion performance through joint inversion were further explained.Regarding the key issue of model inversion parameter settings in current related research,two basic principles were proposed for the property search interval: "wide search interval" and "independent of the true model." For the layer thickness interval,a method of setting the inversion initial model to be greater than the known number of model layers was proposed to reduce the layer thickness interval of the single-layer inversion,which is also a more practical approach.Finally,by jointly inverting the model and actual data,the advantages,and disadvantages of two inversion methods were compared,demonstrating that both the COBL-IDACPSO single-objective algorithm and the L-MODSQPSO multi-objective algorithm can achieve good inversion results under different conditions.The joint inversion of DC resistivity and Rayleigh wave dispersion curves achieved satisfactory results without the constraint of drilling information,which to some extent solves the problem of vertical spatial refinement of the current ground substrate layer survey.Moreover,this method also has certain application prospects in solving other shallow surface layering problems.