Geological mapping and spectral based classification of basement rocks using remote sensing data analysis: The Korbiai-Gerf nappe complex, South Eastern Desert, Egypt

The Pan-African Neoproterozoic Korbiai-Gerf nappe complex in the extreme South Eastern Desert of Egypt comprises dismembered ophiolite assemblages tectonically thrusted over pelite-psammopelite, quartzo-feldspathic gneiss and island-arc schistose metavolcanics. The whole sequence is intruded by syn-late to post tectonic mafic and felsic intrusions. The enhanced Landsat-8 band ratio (bands 6/2, 6/7 and 6/5 x 4/5) and Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) Principal Component (PC2, PC6, and PC5) successfully discriminated most of the exposed lithological units and produced a detailed geological map. Granitoids, psammopelite-pelite, gneiss and serpentinite-talc carbonate rocks have been discriminated using ASTER kaolinite, clay, sericite-muscovite and calcite-carbonate indices respectively. Three spectral based classification algorithms have been compared using Landsat-8 and the Advanced Space-borne Thermal Emission and Reflection Radiometer (ASTER) datasets to obtain the best litho logical classification for the exposed basement rock units. Results from the present study revealed that, Support Vector Machine (SVM) classifier algorithm provided the best lithological classification accuracy (97.72%) using the combination of 9 ASTER bands and 20 ASTER derivative images. The results of the present study concluded that, the integrated data of ASTER and Landsat-8 enhanced images are effective in the discrimination and classification of the basement rock units exposed at Korbiai-Gerf nappe complex and can be applied in similar areas in the Arabian-Nubian Shield. (C) 2017 Elsevier Ltd. All rights reserved.