The potential of Edge recognition method on the basis of the improved hyperbolic tangent filter applied to potential field data: A case study from Madagascar

Edge recognition methods based on gravity and magnetic field data have been widely used, particularly in identifying faults and rock mass boundaries, due to the unique advantages of these methods. The prevailing edge recognition methods often have difficulty in effectively balancing the anomaly amplitudes generated by geological bodies located at different burial depths, resulting in edge recognition results with diminished resolutions. To overcome these problems, we propose a novel edge recognition method (IHRHG) based on the hyperbolic tangent function with regulatory factor $\alpha$ and the regularization derivative of horizontal gradient (RHG) of potential field data. Through synthetic gravity and magnetic model testing, we find that the proposed filter has better performance compared with other commonly used methods, which can balance anomalous signals and avoid the generation of spurious boundaries. Finally, we successfully use the method to study the characteristics of major fault distributions on Madagascar Island and adjacent areas. The effectiveness and practicality of IHRHG filter are demonstrated by evaluating a synthetic example and a real example from Madagascar Island. The main advantage of IHRHG filter is that the method can outline boundaries located at different depths simultaneously, and its detection result is clear and with high resolution and accuracy, while avoiding the generation of false signals.