HYDROLOGICAL MODELLING USING SWAT MODEL FOR KIKULETWA AND SIGI WATERSHEDS, TANZANIA

Abstract

In recent years, rapid population growth, urbanization, industrialization, agriculture, change in economic and social activities have resulted in the increase and diversification of water demand. These changes have now capture social and political tension from the local, national and international levels. Tanzania is also facing similar problems, especially in Kikuletwa and Sigi watersheds. The streams and rivers in these watersheds are being abstracted for agricultural and domestic uses. This abstraction causes water stresses, which lead to social conflict among the water users and to the denial of water for the fragile ecosystem. To address the issue, this study was conducted in order to model the two watersheds and assess the hydrological responses by utilizing the Soil and Water Assessment Tool (SWAT). The Sequential Uncertainty Fitting (SUFI-2) optimization technique available in SWAT CUP. The irrigation and livestock demands were taken from secondary data available. The domestic, industrial and environmental demands were calculated from IS 1172-1993 based on population and the critical minimum requirement for environmental flow. The model performance was analyzed using a graphical comparison of observed and simulated flow as well as a statistical approach. The period from 2000 to 2007 and 2008 to 2012 were used for calibration and validation respectively. The comparative analysis results for Kikuletwa watershed on a daily basis showed very good values of Nash Sutcliffe Efficiency (0.74 and 0.82) and coefficient of determination (0.75 and 0.79) for calibration and validation respectively and also NSE (0.76 and 0.84) and R2 (0.72 and 0.80) for Sigi. The available water is estimated to be about 363.28 MCM/year and 779.9 MCM/year in Kikuletwa and Sigi watersheds, respectively. While the demand was estimated to be 523.12 MCM/year and 788.33MCM/year. The water balance showed that Kikuletwa watershed has a high deficit of 159.84 MCM and Sigi having 8.46 MCM per year. This study concludes that the SWAT model can be used confidently to simulate and assess the hydrological responses for Kikuletwa and Sigi watersheds. It is farther recommended, to adapt soil-water and best management practices and finding alternative ways to mitigate water demands in the dry season by focusing on groundwater sources and rainwater harvesting.