Emmanuel Sakala, Francois D Fourie, Modreck Gomo, and Henk Coetzee
Artificial neutral network, Watershed Policy and Planning, acid mine drainage, groundwater vulnerability, Geographic Information System
This study highlights the usage of artificial neural networks in assessment of groundwater vulnerability. The network uses the DRIST input parameters (Depth to water level, Recharge, Impact of the vadose zone, Soils and Topography) as inputs and hydrochemistry data (Sulphate and Total Dissolved Solids (TDS)) as training data. The results of training and classification using sulphate and TDS were combined using a fuzzy (AND) operator to generate the groundwater vulnerability model. This technique was applied to Witbank Coalfield where acid mine drainage emanating from coal mining operations is a huge concern for surrounding environment and groundwater resources. The generated groundwater vulnerability model of Witbank Coalfield was validated using pH data from 25 groundwater samples. The results show very high negative correlation (0.7182) between the groundwater vulnerability model and pH. This shows that areas with high sulphate and high TDS values correlate with low pH indicative of the presence of possible acid mine drainage pollution. The approach was able to differentiate areas in terms of vulnerability to acid mine drainage which can aid policy and decision makers to make scientifically informed decisions on land use planning. The approach developed in this research need to be applied to other coalfields in order to evaluate its robustness to different hydrogeological and geological conditions.