The massif of Saghro in Moroccan Anti-Atlas is known for several important economic deposits. Thanks to its high potential in terms of production, the deposit of Bouskour is considered among the most important and promising ones on the national scale. The objective of this work is to evaluate the potential of the new Multispectral Landsat-8 OLI and Sentinel-2A sensors in comparison with the Terra ASTER sensor in mineralogical mapping. This task was completed using the band ratio (BR), independent component analysis (ICA) and mixture tuned matched filtering (MTMF) methods. In addition, a field survey was carried out in order to control and validate the obtained mineralogical abundances. The approach followed gave satisfactory results with respect to the field data and the previous studies. Therefore, the sensors used showed great potential for mineralogical mapping and they can be exploited in other regions.

In Morocco, the oriental Anti Atlas encompasses many valuable deposits, including that of copper (Cu) which is located in the Sidi Flah-Bouskour inlier. This deposit is known for its high importance on the national scale. The present work aims to compare the Hyperion raw and fused data for mapping Cu mineralisation. After applying the pre-processing steps on the images used, the fusion (hyperspectral/multispectral) process was done using the Colour Normalised (CN) spectral sharpening method. Subsequently, the mineralogical mapping task was performed using the Mixture Tuned Matched Filtering (MTMF) and the Independent Component Analysis (ICA) methods. In the last step, a geophysical data, a field survey and spectroradiometric measurements were done for checking and validating the results obtained. The analysis showed a high performance of the fused images for mineralogical mapping. Consequently, the methodology proposed can be exploited in mineralogical mapping, and also in other remote sensing works.

Lineament mapping is a very important step in geological studies and mineral exploration, with the evolution of remote sensing, processing methods as well as the availability of several optical and radar satellite data, lineaments can be detected without using traditional methods. The objective of this work is to recommend the most suitable data for lineaments to be used in the field of geological science, by comparing the performance of three different RADAR data, namely ALOS PALSAR (The Advanced Land Observing Satellite Phased Array type L-band Synthetic Aperture Radar), Radarsat-1 and Sentinel-1 for automatic lineament extraction using a combination of edge detection and line linking algorithms. The methodology consists to link between the length, number, orientation and density of lineaments with surface characteristics such as the slope, faults, lithology, discontinuities and mineral veins. The results obtained show that the extracted lineaments from Sentinel-1 VH polarization have a better correlation with geological units, tectonic system direction, as well as shading and slope maps, which is due to the effectiveness of VH polarization, which is independent of soil properties compared to the other polarizations that have shown an overestimation of the lineaments.