Can Targeted Source Information Be Extracted From Superimposed Magnetic Anomalies?

Magnetic anomalies commonly contain anomalies generated by crustal rocks that have variable mineral compositions and natural remanent magnetizations. Recovery of magnetic parameters from superimposed magnetic anomaly sources is therefore important for mineral exploration and geological studies. Superposition of magnetic anomalies leads to nonnegligible errors of inversion and interpretation. Therefore, extraction of magnetic anomalies is essential for magnetic data processing; however, existing methods cannot extract magnetic anomalies caused by sources without substantial depth differences. A novel low‐rank method is proposed for extracting a target magnetic anomaly from a superimposed anomaly. The low‐rank feature of the magnetic anomaly is used to obtain the target anomaly by reducing the Schatten‐p norm of the superimposed anomaly according to the possible target source distribution. The target magnetic anomaly can then be extracted from the data generated by the source interference at the same or different depths. The accuracy of the proposed method was verified using synthetic modeling. Magnetization vector information for individual igneous rocks was recovered from extracted magnetic anomalies via 3‐D inversion of field data from Yeshan (eastern China) that contain complex magnetic anomaly superpositions. The distributions and lithologies of five buried igneous rocks, including basalt, diabase, and granodiorite, were then identified. The proposed method expands the problem of separating magnetic information from rocks in different layers with large depth differences to rocks within the same or other layers and recovers geometric and physical information for each targeted magnetic source from the superimposed anomaly. Plain Language Summary: Magnetic rocks inherit different magnetic characteristics that depend on the type of rock and the features of the geomagnetic field present when the rock formed. Even for rocks at the same locality, the magnitude and direction of the magnetic vectors caused by the rocks are often different. Magnetic field observations contain mixed magnetic information from the geomagnetic core field and from all the rocks under and near the observation area, which causes serious problems for certain types of geological study. Therefore, we developed a new method to recover the magnetic anomaly caused by a target source from mixed magnetic information. We explain how the proposed method works and present a mathematical model. The robustness of the parameters of the proposed method is assessed using synthetic models. Synthetic experiments designed to test the accuracy of the proposed method indicated that the magnetic anomaly of the target rock could be recovered with high accuracy. Finally, we applied the proposed method to study field magnetic information in Yeshan, eastern China. Five magnetic anomalies caused by five isolated rocks were obtained using the proposed method, and the magnetic information, distributions, and lithologies of the five rocks were determined. Key Points: Magnetic data contain superimposed signals from multiple rocks that make extraction of magnetic anomalies for targeted sources challengingA novel low‐rank framework is proposed to extract target anomalies from superimposed magnetic anomaliesDetailed information derived from superimposed anomalies includes a magnetization vector model for each target source [ABSTRACT FROM AUTHOR]