Mapping barite vein-type deposits using Sentinel-2 data and fused spectral methods in a Paleozoic sedimentary cover (Eastern Anti-Atlas)

Abstract

In Morocco, the southern part of the Saghro massif (Eastern Anti-Atlas) includes various economic deposits, including those of Barite (Ba). Previous studies revealed the high potential of these mineralizations, which are structurally controlled by a fault network. This study aims to improve the detection accuracy of barite by using fused remote sensing techniques with Sentinel-2 satellite imagery. This task was completed by comparing different methods, including independent component analysis (ICA), mixture-tuned matched filtering (MTMF), and spectral angle mapper (SAM) methods. These latter include novel fused classifications, namely directed ICA-MTMF (DIM) and directed ICA-SAM (DIS). The obtained results were checked and validated by a previous geological study, field observation, radar Sentinel-1 data, and the calculation of accuracy metrics. The analysis showed that the fused methods (especially the DIM classification) gave better results than the others. The combination of the high spatial resolution of Sentinel-2 data (10 m) and DIM fused classification allowed mapping the targeted veins accurately. Therefore, the adopted methodology can be applied to mineralogical and geological exploration and other remote sensing applications.