The use of GIS and remote sensing to identify and prioritise anomalies for gold from a regional exploration dataset located within the weathered lateritic environment of Mali, West Africa.

The work in this research has used a wide variety of spatial data taken from a regional mineral exploration program. The region of study lies in an area of known artisan gold mining, encompassing the Birimian volcano-sedimentary lithologies, in which the Syama mine is currently operating.;These various data sources of cartographic, airborne, satellite and surface sampling have all been presented in an unprocessed 'raw' state with a complex range of scales, area extents and formats. One of the initial aims of the research has been to construct a fully integrated spatially georeferenced database for their use within a GIS. A second aim has been to find the most suitable analysis for assisting in the identification of data anomalies, pertaining to possible gold mineralisation. The individual analysis of these datasets have in their own right produced interesting 'stand-alone' results. A Third aim has been to combine the individual results, along with knowledge gained from the study of landscape geochemistry as an additional layer, in an effort to prioritise potential gold areas, on the basis of collective anomalies, thereby reducing the risk decision. Through taking advantage of the spatial processing powers of GIS two contrasting methods are investigated with regard to data aggregation and anomaly prioritisation. The first examines the 'weights-of-evidence' approach, which operates through assessing the spatial association of 'known' gold occurrences to individual associate layers, and the second uses an 'expert-system' approach to assess the combined importance of the individual GIS layers to the identification and prioritisation of anomalies.;The value of each approach in this environment is discussed and compared, with final conclusions drawn from the entire study.