3D magnetic amplitude inversion in the presence of self-demagnetization and remanent magnetization

Abstract

We have developed a general 3D amplitude inversion algorithm for magnetic data in the presence of self-demagnetization and remanent magnetization. The algorithm uses a nonlinear conjugate gradient (NLCG) scheme to invert the amplitude of the magnetic anomaly vector within a partial differential equation framework. Three quantities - the amplitude of the anomalous magnetic field, the analytic signal, and the normalized source strength, defined as the amplitudes of magnetic data that are weakly dependent on the magnetization direction - are inverted to recover the 3D distribution of the subsurface magnetic susceptibility. Numerical experiments indicate that our NLCG amplitude inversion algorithm has a rapid convergence rate that provides a reasonable inversion solution in the absence of knowing the total magnetization direction. High-resolution aeromagnetic data collected from the Pea Ridge iron oxide-apatite-rare earth element deposit, southeast Missouri, USA, are used to illustrate the efficacy of our amplitude inversion algorithm. This algorithm is generally applicable for tackling the large-scale inversion problem in the presence of self-demagnetization and remanent magnetization.