Generalizing β-VDR-based derivative computation for robust source edge detection and depth estimation from potential field data

Abstract

Abstract The derivatives of potential field data are essential for source edge detection and depth estimation. A recently introduced vertical derivative method, named β-VDR, has received remarkable recognition in the literature due to its robust vertical derivative approximation. In this study, we compacted the β-VDR formula and extended the method to horizontal derivative calculation to reduce its theoretical computational cost and improve its robustness for edge detection and depth estimation from profile and gridded potential field data. We demonstrated the superior performance of our generalized method through 2D and 3D synthetic tests and showed its applicability on real data.