3D Magnetic Susceptibility Imaging Based On The Amplitude Of Magnetic Anomalies

Magnetic exploration is a common geophysical technique, which is used to study subsurface distribution and structure of magnetic mineral. Three-dimensional (3D) inversion of magnetic data is a quantitative interpretation method, which has been used successfully in mineral exploration. Aeromagnetic anomalies reflect variations of magnetic properties in the subsurface, including distribution of magnetic minerals and magnetization. However, the magnetization direction of subsurface magnetic mineral is unknown. Therefore, it's limited that invert and interpret aeromagnetic data directly in the presence of remanent magnetization.In this paper, I present a general approach for inverting magnetic data in the presence of strong remanent magnetization, called the amplitude of the anomalous magnetic field. For a 2D anomaly of a homogeneous causative body, the amplitude data does not depend on the direction of magnetization. In the 3D case, the amplitude data is weakly sensitive to the magnetization orientation. Therefore,â… invert amplitude data directly to interpret or to image subsurface distributions of magnetic susceptibility. The amplitude data is provided with this property, the mostly reason is to inosculate the information of three-component magnetic data. Hereby, this paper also introduces combining inversion of three-component magnetic data, which can directly adapt to image subsurface distributions of magnetic susceptibility.Based on Occam's inversion method, several theoretical aspects in 3D inversion of magnetic data are mentioned, such as the addition of the depth weighted function and roughness matrix into our objective function. Since the amplitude data depends nonlinearly upon magnetic susceptibility, this inversion is a nonlinear inverse problem. For a nonlinear inverse objective function, the resulting optimization was solved by truncated Gauss-Newton method. The combining inversion of three-component magnetic data is a linear inverse problem. No matter a nonlinear inverse problem or a linear inverse problem, inversion finally all solve multi-dimensional linear equation. In this paper, I apply the conjugate gradient (CG) method to solve multi-dimensional linear equation. Whereafter, I formulate the inversion using Tikhonov regularization. Further more, the experiments of some theoretical model indicated that the inversion method seemed to be reliable and practical.Concerning the boundary constraint of magnetic susceptibility, selected the nonlinear transformation method in this paper, it via the iterative process of inversion for the progressive realization of the lower and upper bounds to magnetic susceptibility. Constrained inversion results which are significantly better than the unconstrained inversion results are more reliable as that imaging anomaly body boundaries is clearer and "energy" is more focused. Synthetic model experiment shows that 3D inversion of the amplitude data can best suppress affects by the strong remanent magnetization, but has a lower vertical resolution of magnetic susceptibility imaging results. Inversion of three-component magnetic anomaly can suppress the impact of the weak remanence with a high vertical resolution of magnetic susceptibility imaging results.Finally, I test the inversion algorithm by using the actual magnetic data in the Chengchao Iron mine. Results of magnetic susceptibility imaging demonstrate that the location of the magnetic anomaly imaging is consistent with other interpretations.