Robust Edge Detection for Structural Mapping Beneath the Aristarchus Plateau on the Moon Using Gravity Data

Abstract

Accurately detecting the edges of subsurface geological structures from potential field anomalies remains a fundamental challenge. We applied the HTHG (Hyperbolic tangent function with horizontal and vertical derivatives of Total Horizontal Derivative) method to enhance subtle details in lunar gravity anomalies, focusing on the Aristarchus region and its surroundings. Initial assessments were conducted on synthetic noise-free and noisy gravity data sets and compared against eleven representative edge detectors. In the noise-free data case, HTHG demonstrated superior performance over other detectors in terms of accuracy, resolution, sharpness, and amplitude balancing. However, similar to other approaches, its directional derivative calculations are highly susceptible to noise amplification. To address this challenge, we implemented various noise reduction techniques, including the β-VDR and MNLM methods. Notably, we also presented different methods for estimating the tuning parameters of the involved noise attenuation methods. HTHG, in conjunction with MNLM, demonstrated the most superior performance. We subsequently applied the HTHG operator to lunar gravity anomalies from the Aristarchus region. Our results were compared with the outputs of 2D inversion employing a mixed-weighted function, a correlation imaging algorithm, and 3D inversion enhanced by spectral analysis. Our findings indicate that the Aristarchus crater hosts a low-density subsurface mass. The outcomes of this study confirm the robust performance of the HTHG method in addressing edge detection challenges and underscore the necessity of integrating various methods, including edge detection, noise suppression, fast imaging, and inversion, to guarantee the interpretation reliability and advance our understanding about the internal architecture of the Moon.