A New Vector for Mapping Gold Mineralization Potential and Proposed Pathways in Highly Weathered Basement Rocks using Multispectral, Radar, and Magnetic Data in Random Forest Algorithm

Various deformation events and weathering processes are among the obstacles to exploration of natural resources, especially mineral ones. This study addressed these obstacles by proposing solutions through the utilization of multiple multispectral, radar, and magnetic datasets. The study region, situated in a region between two blocks associated with the Arabian-Nubian Shield, is known for the presence of these obstacles. Also, the region is characterized by the conflicting nature of these blocks in relation to occurrences of gold mineralization. Various optical and multi-frequency radar datasets were used for detailed geological/mineralogical discrimination and surface lineaments mapping, respectively. Numerous composite enhancement, processing, and classification methods were applied to aeromagnetic and multispectral data to derive as much information as possible from these data. The results of the classification of minerals, the separation of rock units, and the identification of deep-seated geological structures were combined to produce a map for predicting gold prospectivity by generating a random forest model using 10 controlling factors and the previously defined zones of gold mineralization. A three-dimensional model was generated to show the locations of the expected gold mineralization potential areas with regard to depth in the investigated region. This model figured out the strong relation between gold potentialities related to their suggested pathways, particularly WNW- and N-trending shear zones. Our model indicates that the region is promising for mining activities and presents a new approach for mapping pathways of gold-mineralized-altered zones that can be applied worldwide.