Chevrel, S, Croukamp, L, Bourguignon, A & Cottard, F 2008, 'A Remote-Sensing and GIS-Based Integrated Approach for Risk-Based Prioritization of Gold Tailings Facilities — Witwatersrand, South Africa', in AB Fourie, M Tibbett, I Weiersbye & P Dye (eds), Mine Closure 2008: Proceedings of the Third International Seminar on Mine Closure, Australian Centre for Geomechanics, Perth, pp. 639-650, https://doi.org/10.36487/ACG_repo/852_59
(https://papers.acg.uwa.edu.au/p/852_59_Chevrel/)
Abstract: The Witwatersrand gold field, located in the vicinity of Johannesburg, has been mined for more than a 
century. The Witwatersrand Mining Basin is covered with an excess of 400 km² of covered tailings heaps, 
thus affecting a considerable area and large watersheds. Tailings dams, waste rocks and derelict lands can 
lead to regional environmental problems and disseminate potentially toxic materials into the environment. 
This paper describes a collaborative project between the South African Council for Geoscience (CGS) and 
the French Bureau de Recherches Géologiques et Minières (BRGM) Geological Surveys, on an Earth 
Observation and GIS-based integrated methodology for rapid spatial screening and source characterization 
of hazards associated with gold slime dam features at the local to the regional scale in the Witwatersrand 
gold fields.  
The following was performed: (i) a visual interpretation of land-use over time from satellite images, which 
shows the growing urbanization in the West and East Rand, as well as tailings dams-reclamation, in 
particular in the East Rand; (ii) a land-cover digital classification based on validated image spectra from 
multi-spectral ASTER imagery, with a focus on potential surface water contamination; and (iii) a simplified 
and Preliminary Risk Assessment (PRA), combining a number of information layers in a GIS, to rank the 
tailing dams waste facilities and identify the sites requiring remedial actions or water/soil usage restrictions. 
The GIS analysis is based on several data layers, including satellite images, digital elevation models 
(DEMs), topographic maps, airborne radiometric images, geotechnical maps, groundwater depths, 
boreholes, sinkholes and land cover. It uses a score-based ranking approach for each of the three 
components of the pollutant source – pathway – receptor pollutant model. The final product is presented in 
the form of maps that break down the final risk level related to the site, or the source classification, into 
three categories: (i) sites at ‘high risk’, corresponding to sites requiring remedial works, land-use 
restrictions or further detailed investigations; (ii) sites only requiring monitoring and for which it might be 
necessary to draw up land-use restrictions as well as investigate and monitor sites with uncertain conditions; 
and (iii) so-called ‘low-risk’ sites not requiring particular actions.