Remote Sensing Contributions to Lithostructural Mapping Using Sentinel 2A Data: A Case Study of the Errachidia Area (Central High Atlas, Morocco)

Abstract

Remote sensing techniques have become a vital resource for geological mapping and analysis. This research focuses on litho-structural mapping in the Northern Errachidia area in the southern Atlas range, Morocco, formed by Jurassic and Cretaceous formations shaped by the Alpine orogeny. This study combines remote sensing with field investigations. The main goals are to expand geological mapping south and east of the Ait Atmane geological map and demonstrate the efficacy of remote sensing in geological mapping. Here, we develop a methodological approach applicable to other areas with the same characteristics. For this purpose, several enhancements were applied to the preprocessed data, including contrast enhancement, minimum noise fraction, principal component analysis, and independent component analysis. These steps produce reliable results for visual mapping and extracting training samples for classification algorithms. Machine learning techniques like Mahalanobis Distance Classification and Support Vector Machine were evaluated to determine the best classification method for lithological mapping. Subsequently, structural lineaments were automatically extracted using directional filters applied to Sentinel 2A data. The proposed methodology provides a geological map validated in metrics, spatial frequency, and accuracy based on the field and available data.