Lithological discrimination and lineaments extraction using Landsat 8 & ASTER data, a case study of Jbel Saghro (Moroccan Anti-Atlas)

Abstract

Abstract The present work reveals the potential of Landsat 8 and ASTER imagery in the lithological discrimination and lineaments extraction in the region of Tiwit (Jbel Saghro). Various remote sensing and image processing techniques were applied to the Landsat 8 and ASTER scenes: False-color composites (RGB 751 & 531), Principal Component Analysis (PCA 653 & 821), Minimum Noise Fraction (MNF 643 & 541), and Independent Component Analysis (ICA 137 & 235). These techniques discriminate the granitic formations (Isk-n-Alla, Mimasmarane, Ibantarn, and Ikniwn), the rhyolitic and ignimbrite formation (Amtattouch, Ouzarzamand Assaka), and other various rock types (aphanitic basalts, sandstones, conglomerates, etc.). The automatic and manual lineaments extraction methods highlight the major lineaments in the study area, trending NE-SW, E-W, and ENE-WSW. The obtained results are consistent with the geologic map of Tiwit. Maximum Likelihood, Spectral Angle Mapper, and Mahalanobis distance classifiers show an overall accuracy of 88%, 56%, and 82.6%, respectively, for Landsat 8. ASTER data show a better result in classification with an overall accuracy of 90.6%, 84%, and 88% for the same classifiers.