A Constrained Approach by Curie Point Depth Estimation for the 3‐D Inversion of Regional Lithospheric Magnetic Anomaly Data in Spherical Coordinates and Its Application to the Northern Xinjiang, China

Abstract

AbstractThree‐dimensional (3‐D) inversion of regional lithospheric magnetic anomaly data not only needs to be performed in the spherical coordinate system but also requires valid constraint information to suppress the nonuniqueness. We present a constrained inversion approach of regional magnetic data at near‐surface altitudes to recover a 3‐D susceptibility model in spherical coordinates. First, we use a synthetically forward modeling experiment to illustrate that a 3‐D model with susceptibility variations along the depth direction should be considered when the altitude of magnetic data is 5 km above the WGS84 ellipsoid. We then choose the least‐squares algorithm to recover the lithospheric susceptibility model. Meanwhile, a model objective function including sensitivity‐based weighting function and bottom depth constraint of the source (i.e., Curie point depth, CPD) is introduced to stabilize the inversion and guide an optimal solution. Both the synthetic test and the application to the real data over the Northern Xinjiang indicate that the constrained inversion approach is effective to obtain a geologically reasonable result above the CPD. The recovered 3‐D susceptibility distribution under the Northern Xinjiang corresponds well to the extension of the main tectonic frame. All the rigid and cold blocks are imaged as high‐susceptibility anomalies, whereas the pronounced low‐susceptibility anomalous body is observed beneath the Tianshan. In addition, the earthquakes are mostly (>95%) located in areas where the susceptibility is smaller than 0.03 SI and vertically integrated susceptibility is smaller than 0.8 SIkm.