Research Of2D CSAMT Forward And Inversion

The CSAMT is a frequency-domain electromagnetic sounding method, which uses electric dipole as a source, through the emission of electro-magnetic waves of different frequencies to achieve the purpose of exploration.In recent years, it is a commonly used geophysical method,and have a wide range of applications in the field of metal mining, energy and engineering.Only far field has features of the plane wave,in the near field, it makes data inversion difficulty, thus limiting the developments of data processing. Magnetotelluric data processing uses two-dimensional and three-dimensional inversion technique, but controlled source audio magnetotelluric inversion mostly uses one-dimensional inversion method,2D&3D inversion is under research, usually magnetotelluric inversion method applied to the data processing of the csamt, but when the data don’t meet the far-field conditions, the results will be wrong. The study of forward and inverse two-dimensional controll source can improve the accuracy of the interpretation of controlled source exploration.Numerical simulation of2.5d electromagnetic method, with three-dimensional source and2D geoelectric model, this paper will use the secondary field method to avoid the singularity of the source.Compared to the total field algorithm, the secondary field algorithm can significantly improve the accuracy and don’t significantly increase computation time. Sparse matrix formed by the finite element forward modeling, solving the equation to get the result of different values of the wave, and then use the inverse cosine transform gets results of the spatial domain.Two-dimensional controlled source inversion study focus on the frequency-wave domain Jacobian matrix calculation,which is accompanied by adjont eqution method, Jacobian matrix can be obtained through the inverse Fourier transform of frequency wave domain Jacobian matrix.In this paper,in order to solve the above problems, needs to tear down the frequency-wave domain finite element formulas, studys of sparse matrix storage, improves the speed of forward modeling, getsthe Jacobian matrix.