Inversion Of2.5Dimensional Time Domain Helicopter Airborne Electromagnetic Method

As a quick exploration method developed in recent years, airborne time domain electromagnetic (ATEM) is of some advantages such as lower cost, wider detection, greater probing depth, etc., so it was widely used in mineral resources exploration, geological mapping, prospecting of water resources and environmental monitoring and other fields. Imaging and inversion interpretation of the ATEM data is one of the key researches in ATEM. Currently, the research of the inversion of ATEM data is mainly focused on one-dimensional inverse problem, but less studies about high-dimensional inverse problem. Compared with the1D inversion,2.5D inversion of ATEM data has higher precision, and compared with the3D inversion,2.5D inversion of ATEM data has a smaller amount of calculation. Therefore, the study of2.5D inversion method has become one focus of the ATEM data inversion research. This paper discusses mainly2.5D inversion methods of ATEM, and lays the foundation for the data interpretation of ATEM. Specifically, the main contents of this article are described as follows.(1)2.5D forward modeling of ATEM:Firstly, on the basis of electromagnetic basic theory, this paper presents the boundary value problem which is comprised of the dual second-order partial differential equations of electromagnetic field along the strike direction component in Laplace-Fourier domain and the boundary conditions. Then, we describe the solution of this boundary value problem with finite element method, and analyze the solving of background field and linear equations. Finally, we analyze the forward response of several typical2D geological models.(2)2.5D inversion of ATEM:This paper describes the calculation process of damped eigenparameter inversion method and damped least-square inversion method, and discusses several key parts of the process:the Jacobian matrix computation, the choice of damping factor, iterative convergence criteria and inversion stopping criteria. Both these two inversion methods are based on singular value decomposition theory, and are related to the decomposition of the Jacobian matrix to solve singular matrix. The Jacobian matrix computation is a very important step in inversion problem, which directly affects the correctness of the inversion results. Damping factor is used to control the correction direction and step length in iterative process, which plays an important role in ensuring the convergence of iteration. Iterative convergence criteria and inversion stopping criteria can ensure the iteration process cycle and end normally. To ensure the stability and effectiveness of the program, we often set multiple stopping criteria, as long as meet one of stopping conditions, the iterative inversion process may be terminated.(3) Analysis of the ATEM2.5D inversion results:First, we analyze the inversion results of forward response data of several geological models with the same initial model. Second, we discuss the inversion results of above two inversion methods with different initial model for one same forward data. The inversion results show that damped eigenparameter inversion method can obtain better results than damped least-square inversion method.