Spatial modelling of groundwater quality across a land use and land cover gradient in Limpopo Province, South Africa

摘要

The spatial variability of groundwater quality in rural Mokopane District is investigated. This was achieved by evaluating the variability in groundwater physio-chemical parameters in relation to Land Use and Land Cover (LULC)using Kriging geo-statistical technique and key groundwater physio-chemical parameters namely; ammonia, lead, Total Dissolved Solids (TDS), nitrates and chloride. Firstly, we derived the LULC from 30-m Landsat 8-OLI satellite data, with an overall accuracy of 76.67%. Water quality parameters were modelled in geographic information systems using the kriging interpolation technique. We then compared measured water quality and the drinking standards recommended by the South African Water Quality Standards (SAWS) and the World Health Organization (WHO) guideline standards for drinking water. Results indicated that there were significant differences (p < 0.05) in the levels of physio-chemical contents between borehole water and standard values according to the SAWS and WHO. For example, TDS in Mosesetjane exceeded both SAWS and WHO drinking water levels. Nitrate concentrations were also high on the first and second quarter with concentration levels of 5.56-29 mg/l and 30.9 mg/l in Mosesetjane, respectively. However, lead and ammonia were found to be evenly distributed in all the four quarters, but exhibited some slight deviations from the standard levels (+/- 4.7 mg/l) in selected quarters. Results also indicated significant differences in TDS, chloride and nitrate concentrations across different LULC types. Subsistence farming and built-up areas were the main LULC types that were found to have high concentrations of TDS, ammonia and nitrates, exceeding the recommended standards. However, lead and ammonia, did not differ significantly (p > 0.05) and these parameters were fairly distributed, with less spatial variability. The second quarter, extending between October and December, and the fourth quarter, extending from April to June, exhibited high concentrations of nitrates, chloride and TDS during the wet season. On the other hand, the first and third quarter of the dry period were associated with low concentrations. This study indicated that LULC types and seasonal variability have influence on specific groundwater quality parameters.