Advanced multi-technique aeromagnetic and remote sensing characterization of a small-scale gold prospect: a case...

Significant gold deposits occur within the ophiolitic rocks of the Central Eastern Desert of Egypt; yet, the structural controls governing ore distribution remain insufficiently constrained. This study applies an integrated remote sensing-aeromagnetic-structural workflow to the Um Salim area along the ENE-trending Barramiya-Um Salatit ophiolitic belt to delineate surface and subsurface architectures and assess their influence on gold localization. High-resolution lithological mapping was achieved through machine-learning classification of hyperspectral EnMap data using the Random Forest (RF) and Support Vector Machine (SVM) algorithms. SVM applied to enhanced Minimum Noise Fraction (MNF) components yielded superior performance, achieving an overall accuracy of 95.8%, a kappa coefficient of 94.6%, and an F1-score of 96.6%. Reduced-to-pole magnetic data reveal a prominent positive anomaly (~ 1400 nT) directly over the Um Salim gold mine, attributed to magnetite-bearing metavolcanics and ophiolitic serpentinite rocks. Upward continuation to 0.5-1 km confirms that this anomaly reflects a deeply rooted magnetic source. To refine structural interpretation, advanced edge-detection filters (TAHG and ILTHG) were applied, enhancing ENE-WSW, NE-SW, NW-SE, and N-S trends consistent with multi-phase deformation (D1-D3), and SAR-derived lineaments. Center for Exploration Targeting (CET)-based lineament density and orientation entropy maps delineate zones of intense structural complexity that closely correspond with mineralized shear zones and dyke-like porphyry intrusions. Depth-to-source analyses using Euler Deconvolution (EUD) and Tilt-Depth (TD) methods reveal that most magnetic bodies occur within 0-1000 m, whereas deeper (~ 1500 m) sources likely represent intrusive conduits that focused hydrothermal fluids. The novelty of this study lies in its integrated, multi-technique application of aeromagnetic analysis combined with hyperspectral machine learning to characterize a small, gold prospect and to map alteration-linked structural features at high resolution in the Barramiya-Um Salatit belt. The integrated results highlight new structurally complex zones in the eastern and southeastern Um Salim region as high-priority exploration targets, particularly along the NE-SW shear system that controls gold mineralization. This workflow provides a high-resolution, transferable strategy for targeting gold in structurally complex Precambrian terranes worldwide.