Hydrological modelling for estimating water availability for irrigation supply in selected schemes in Nzhelele Area

Abstract

South Africa is one of the driest countries in the world, and its smallholder irrigation schemes - particularly in semi-arid regions like Nzhelele in Limpopo Province - are under growing pressure due to water scarcity and climate variability. Many of these schemes suffer from poor monitoring of water inflows and a lack of reliable data on crop irrigation needs. This study aimed to quantify irrigation water availability and determine whether current water resources can meet the crop water demands of three smallholder schemes - Mandiwana, Raliphaswa, and Mamuhohi. Objectives included simulating streamflow using hydrological modelling, estimating crop water requirements, and identifying which crops are sustainable under existing water supply conditions. The Soil and Water Assessment Tool (SWAT) was used to simulate streamflow from 1987 to 2014 using daily hydrometeorological data. Calibration and validation were performed over two distinct periods: 1991–2000 and 2001–2009. Reference evapotranspiration (ETo) was calculated using the Hargreaves equation due to limited data, and crop coefficients (Kc) were used to estimate daily water needs for maize, sweet potato, groundnuts, and green beans. A 60% conveyance efficiency was applied to assess actual deliverable water. The model showed that maize and sweet potato had manageable water demands 540 mm and 475 mm in the dry season, respectively, while groundnuts and green beans had higher peak requirements that exceeded available supply. The SWAT model performed well, the R2, NSE and PBIAS values for calibration and validation ranged from 0.78, 0.56, and 7.7%, and 0.60, 0.56, and -8.3%, respectively. These findings can guide crop selection and water allocation, offering practical insights for improving smallholder irrigation planning in water-limited environments.