This study investigates the use of Remote Sensing technology for geological mapping in an area south of the Orange River near Alexander Bay in South Africa. Landsat, ASTER and Hymap images were used to discriminate geological lithologies and structures in the study area. The results of the images were compared with each other using the 1: 50 000 geological map as a reference.
The Minimum Noise Fraction (MNF) was the remote sensing technique used to identify different lithological units in the study area. The MNF bands were displayed in an RGB band (colour) combination in order to enhance the lithologies identified in each image. The MNF 459, 321, and 431 band combinations for Hymap, ASTER and Landsat images were used respectively. Hymap images were the most informative in lithological discrimination followed by ASTER and with Landsat being the least effective. The rock types of the Orange River Group (basaltic andesite and rhyolite) were clearly identified from the Vioolsdrif Suite rock types (granitoids) occurring in the study area by the Hymap image.
Pixel size, spectral and spatial resolution differences of Hymap, ASTER and Landsat images were the result of discriminatory qualities of lithological information retrieved by the different sensors. The cost effectiveness of each image type in relation to the study area size is one of the major challenges in data selection and therefore should be considered in the initial planning stages of geological mapping.