Potential field methods are extensively utilized for identifying and modeling subsurface geological structures and mineral deposits. Edge detection constitutes a critical step in processing and interpreting potential field data, aimed at determining the precise lateral position of subsurface anomalies. The edges of an image represent its high-frequency components; thus, the derivative operator plays a significant role in delineating these edges due to its remarkable capacity to effectively amplify high-frequency components.
    However, several challenges confront edge detection in potential field data. Notably, the imbalance between the edges of deep and shallow sources, alongside low resolution, stands as one of the most prominent challenges encountered with conventional methods such as total horizontal gradient and analytic signal techniques. This limitation primarily arises from the reliance on the amplitude of the gradient vector in edge detection processes. The amplitude of the gradient vector is closely correlated with the amplitude of the anomaly, which, in turn, is inherently associated with the depth of the anomaly source beneath the surface.
    In recent years, innovative edge detection methodologies based on phase information derived from the gradient vector have emerged as promising alternatives. Ferreira et al. (2013) proposed that by calculating the tilt angle on the total horizontal gradient, it is feasible to not only balance the edges of deep and shallow anomalies but also significantly enhance resolution. Pham et al. (2022b) introduced an improved approach for calculating the tilt angle on the total horizontal gradient to augment the resolution of anomaly edges by imposing a threshold on the derived tilt angle; this method preserves high values of the tilt angle to further improve resolution.
    The efficacy of the proposed method was rigorously evaluated using synthetic data under two distinct conditions: one devoid of noise and another subjected to noise, simulating real-world scenarios that geophysicists often encounter. The results obtained from synthetic data indicated that the proposed method, while effectively identifying edges of anomalies at varying depths in a balanced manner, surpasses other comparative methods discussed within the paper in terms of resolution and clarity. Furthermore, findings derived from the noisy scenario demonstrated that the proposed method exhibits greater stability against noise than many of the aforementioned methods, thereby ensuring more reliable outcomes in practical applications.
    An assessment of the proposed method on actual gravimetry data related to the Dida iron deposit located in Jilin province in northeast China, juxtaposed with results from traditional methods, revealed that the proposed approach is capable of not only eliminating false edge generation but also producing edges with superior resolution compared to alternative methods. The comparative analysis highlights its effectiveness in accurately capturing subsurface features crucial for geological exploration. Consequently, this proposed method may be regarded as a viable alternative for edge detection within potential field data. By enhancing our ability to interpret subsurface anomalies effectively, this advancement holds significant implications for various applications in geoscience, including mineral exploration, environmental studies, and resource management.