Determination of recharge and groundwater potential zones in Mhinga Area, South Africa

Abstract

This study was focused on determining groundwater recharge and groundwater potential for Mhinga area in South Africa, which is a rural area that depends mainly on groundwater for domestic water supply. Numerical modelling was used to simulate the groundwater behaviour in the aquifer and estimate groundwater recharge. MIKE SHE and MIKE 11 models were coupled and used to estimate groundwater recharge within calibration and validation periods of 2007/07/01 to 2009/12/31 and 2010/01/01 to 2013/05/21, respectively. Due to limited data availability for Mhinga, modelling was carried out at quaternary scale and then localised to Mhinga area. Remotely sensed data (satellite images, shapefiles and maps) was used to produce the groundwater potential map for Mhinga. The data were assigned with weights using Saaty’s Analytical Hierarchy Process and overlain on ArcGIS platform. Borehole drilling statistics of the boreholes in A91H quaternary catchment were used to validate the groundwater potential map. In streamflow modelling using MIKE 11, values of Nash-Sutcliffe efficiency (NSE), correlation coefficient (R), root mean square error (RMSE) and mean absolute error (MAE) were 0.51-0.89, 0.73-0.97, 3.61-7.96 and 1.13-2.75, respectively. In integrated groundwater modelling using MIKE SHE, values of NSE, R, RMSE and MAE were 0.72-0.84, 0.87-0.93, 0.18-0.32 and 0.13-0.26, respectively. These values showed that MIKE SHE and MIKE 11 models had satisfactory performances. Groundwater recharge estimates were generally very low ranging from 0 to 2.75 mm/year, which constituted 0 – 0.42 % of Mean Annual Precipitation (MAP) for the A91H quaternary catchment. This was associated with high evapotranspiration (mean of approximately 4 mm/day) compared to the low precipitation levels (MAP of 656 mm/year). Moreover, in the low lying areas, with gentle slopes, low recharge between 0.2 – 0.4 mm was observed. The groundwater potential (GWP) map produced revealed that Mhinga is predominantly covered by regions of very low and low groundwater potential, which was associated with the type of geology. Area coverages of 34.47 % had very low, 51.39 % had low, 7.66 % had moderate and 6.48 % had high groundwater potential. Moderate to high groundwater potential zones were located along the geologic fault zones. In A91H, 112 unsuccessful boreholes were drilled, 69 (61.6 %) fell in the very low GWP zones, 16 (14.3 %) fell in the low GWP zones, 17 (15.2 %) fell in moderate GWP zones and 10 (8.9 %) fell in the high GWP zones. In the Mhinga, 19 unsuccessful boreholes were drilled of which, 11 (57.9%) fell in the very low GWP zones, while 6 (31.6%) fell in the low GWP zones and 2 (10.5%) fell in the moderate GWP zone. Hence 89.5% of all the unsuccessful boreholes drilled occurred in the very low to low GWP zones. It is concluded that the study area is mainly dominated by of areas with low recharge and very low to low groundwater potential. It is recommended that the MIKE SHE – MIKE 11 model and the GIS models should be developed further and improved as more data is collected to refine the conceptualisation of the aquifer.