Integration of WorldView-3 Remote Sensing and Classification Techniques for Mapping Iron Oxide Mineralization in MVT Pb-Zn Deposits, Jbel Bou Dahar, Morocco

Remote sensing is a vital tool in mineral exploration, especially in geologically complex terrains where mapping iron oxide minerals can serve as a key proxy for identifying associated ore deposits like Pb-Zn. This study utilizes WorldView-3 satellite data and various image processing techniques, including band ratio, Principal Component Analysis, Support Vector Machine, Maximum Likelihood, and Spectral Angle Mapper, to map iron oxide mineralization associated with Pb-Zn mineralization in the Jbel Bou Dahar region of Morocco. Principal component analysis and band ratio methods successfully identified ferric (Fe³⁺) and ferrous (Fe²⁺) iron. Fe³⁺ was concentrated in oxidizing zones and the Cp geological unit, while Fe²⁺ mainly appeared in the Do unit and vegetated areas. Principal component analysis proved more effective than band ratio in separating Fe³⁺ from Fe²⁺ due to reduced spectral overlap. Among classification methods, the support vector machine achieved the highest accuracy (91.34%) and best performance in mapping Fe³⁺ near Pb-Zn veins, outperforming the maximum likelihood and spectral angle mapper. Mineral Prospectivity Maps generated using fuzzy logic with principal component analysis and classification outputs identified high-potential zones with substantial predictive value. These findings were validated using 3D imagery, macroscopic observations, and X-ray Diffraction, confirming a close link between Fe³⁺ enrichment and Pb-Zn mineralization. Overall, the results highlight the effectiveness of WorldView-3 data and integrated remote sensing techniques in exploring carbonate-hosted mineral deposits.