Geological mapping of the area of Lawra Gold Belt (Ghana) based on remotely sensed images

Abstract

Remote sensing techniques and spatial data analysis through Geographic Information Systems (GIS) have been jointly applied in lithological discrimination in the Lawra gold belt. Spectral signatures of the study area recorded by the Landsat 8 were studied to aid geological unit identification. As expected the savannah vegetation coupled with tropical geomorphic events (weathering, erosion and deposition), spectral inseparability and positional errors have the tendency to affect significantly the accuracy of the classification. Final detection of lithologies was based on correlation between classes generated in the thematic map and the reference geological map with a moderate overall accuracy of 20.192%. Results show that for unsupervised classification using Iterative Self-Organizing Data Analysis Technique (ISODATA) small number of classes produces more accurate results. The results once more show the limitations of spectral based approaches to geological mapping in vegetated terrains. The development of this data will provide a baseline on which to base future studies of the area. Integrated approach (geology, textural and spectral features) could be used to improve overall accuracy.