Calibration, validation and application of the SWAT model to determine the hydrological benefit of wetland rehabilitation in KwaZulu-Natal, South Africa

Abstract

In South Africa, with highly variable and intense land-use practices, coupled with limited soil fertility and water resources, there has been a long history of encroachment of arable lands (sugarcane and timber plantations) into surrounding wetlands. Although wetland delineation within the timber and sugar sectors is well-defined in policy, and existing and proposed legislation, there are significant areas of non-compliance. The spatially-explicit Soil Water Assessment Tool (SWAT) was adopted to investigate the interactions of climate, land-use and soil on the water-use of natural and encroached wetlands. This paper documents the calibration, validation and application of the SWAT model on Quaternary Catchment (QC) U20G, which is a 498 km2 catchment that forms part of the uMngeni River basin. The SWAT-CUP parameter sensitivity and optimization model was tested with daily observed streamflow data for this catchment. Parameters were modified using the sequential uncertainty fitting (SUFI-2) analysis routine to calibrate the model. The simulated flow had a close fit to the observed flow with a regression coefficient (r2) of 0.87 and a Nash-Sutcliffe (NS) coefficient of 0.8. Through the buffer scenario analysis, the model showed that if the wetland and a 20-m buffer were to be returned to a natural state, there could be a 16% increase in the annual streamflow contribution, with an upper limit of a 60% increase in some hydrologic response units (HRUs). Thus there would be a hydrological gain if wetlands and sensitive buffer areas were to be cleared of commercial timber species and sugarcane.