Magnetization Vector Inversion Based on L1-TV1 Regularization and Amplitude Constraint

Magnetization vector inversion (MVI) faces significant non-uniqueness challenges due to the influence of multiple parameters, which complicates the recovery of subsurface structures. To tackle this issue, a new magnetic vector inversion method that integrates L1-TV1 regularization with amplitude constraints is proposed. Specifically, within the L1-TV1 inversion framework, amplitude constraints on the three magnetization components are incorporated into the model’s L1-norm term. This integration enhances the consistency among the magnetization components and effectively mitigates issues related to positional and shape misalignment. Additionally, four weighting factors for the L1-TV1 regularization are introduced. A comprehensive analysis reveals that carefully selecting optimal weights can significantly improve depth resolution. Model tests confirm the effectiveness in recovering the position, geometry and magnetic direction of subsurface anomaly sources. Finally, when applied to measured data from the Galinge ore deposit in China, the proposed method produces a high-resolution 3-D distribution of the orebody that is consistent with geological records, demonstrating its practicality for data interpretation.