Application of Noise Suppression Methods to Potential Field Data from The Gulf of Aden: Tectonic and Structural Interpretation

Abstract

Numerous studies have utilized edge detection filters based on potential field derivatives, ratios, and statistical methods to assess geological structures and offer interpretations. However, the varied responses from these filters can lead to misinterpretations, making it challenging for users to select the most effective filter. To ensure accurate evaluations and prevent misinterpretations, it is crucial to simultaneously assess filters on the same theoretical models. This study introduces new edge detection filters based on enhanced algebraic and geometric functions applied to potential field data derivatives. The performance of 23 edge detection candidates was evaluated using 2.5D gravity models and magnetic compositional data under different scenarios. The results indicate that the modified fast sigmoid function (MFSED) outperforms the other 22 filters in terms of accuracy for both gravity and magnetic models. This comparison enables the inference of tectonic structural patterns by generating multiple results from different filters and examining artefacts. Additionally, edge enhancement filters were applied to free-air and magnetic gravity data of the Eden Fault, leading to the identification of boundaries related to seafloor spreading, divergent plates, and transform faults. These boundaries align well with the locations of earthquakes at shallow depths. The study suggests that the local fault system contributes to the dynamics of the Gulf of Aden, with seismic activities along normal fault zones and their transformation being associated with defined structural boundaries. This application of new edge detection filters offers a novel approach to studying faulting processes through multiple monitoring techniques.