Mapping and modelling human exposure to uraniferous mine waste using a GIS-supported virtual geographic environment

Abstract

The exposure of residents to large volumes of uraniferous mine waste deposits scattered across densely populated areas in the South African goldfields has long been identified as a potential health threat. To quantify this threat for selected locations numerous studies determined concentration levels of uranium and other radionuclides in different environmental media of the Witwatersrand basin including streams, soils, tailings deposits, mine water effluents, fluvial sediments, aquatic organisms as well as in the food chain (fish, cattle, crops, vegetables etc.). This paper describes how the secondary data were extracted from various sources in order to map the spatial extent and intensity of uranium pollution in selected goldfields and to quantify associated exposure of residents. Steps included geo-referencing, normalisation against selected benchmark values, and superimposition of complementary data sets in GIS. For easing access to the data base for non-GIS experts a virtual geographic environment (VGE) was used, offering updated satellite imagery and additional functionality that aided in modelling exposure of target populations. Exposure pathways considered here include ingestion of polluted water, food and soil (geophagia) as well as the inhalation of wind-blown tailings dust. In addition, anecdotal evidence and media-based reports on mining-related pollution and impacts were captured. Maps derived from the VGE were used to identify hot spots of radioactive pollution aiding the selection of suitable sampling sites for a screening survey on U-uptake by exposed residents using hair as a bio-indicator.