An investigation into the effectiveness of Sentinel-2 in lithological identification through remote sensing techniques

The global demand for critical raw materials (CRMs) has led to an increased use of satellite imagery in mineral exploration. Remote sensing has proven to be an essential tool for facilitating mineral exploration and enhancing the accuracy of targeting specific minerals compared to conventional and intrusive methods. This work aims to prioritize and focus on a greenfield exploration study area with a complex lithology in the south of Spain. Band ratio and supervised and classification techniques were applied to Sentinel-2 imagery, which was masked for water and vegetation. A comparative analysis of Band Ratio (BR) and Support Vector Machine (SVM) was conducted. SVM surpassed BR in mapping the lithologies. Based on the comparison between results obtained from BR and SVM, the consistent pixels were used to create prospectivity binary maps as the target points for extra infield investigations. The application of remote sensing data and methods for lithological mapping is crucial for mining companies, as it can enhance efficiency and sustainability in mineral exploration and extraction within the preliminary stages.