Application of rainfall-runoff modeling for irrigation development in lake Rukwa basin: a case study of Momba and Songwe catchments

Rainfall-runoff modeling in Lake Rukwa basin specifically Momba and Songwe sub- catchments were done using five different rainfall-runoff models, which ranges from system type of models to conceptual ones in five stations within the catchments. Simple Linear Model (SLM), Linear Perturbation Models (LPM), Seasonal Varying Runoff Coefficient Model (SVRC), Soil moisture accounting for Routing Model (SMAR) and NAM model were applied in Momba and Songwe catchments, which cover an area of 10,098 and 9,078 km2 respectively. Rainfall, stream flow and evaporation data used in this study ranged from 1963 to 2012 while the common period for the data sets was 1975 to 1997. The time series in the catchment were checked for the seasonality and it was found that the high flow occurs in two periods from January to April and October to December, the low flow occur in May to September. The Nash-Sutcliffe (1970) model efficiency Index (R2) and Index of Volumetric Fit (IVF) were used as the measures of the model performance. Each model performs differently in Momba and Songwe catchments. SMAR models gave reasonable results and therefore they can beused in simulation and forecasting discharges in Momba and Songwe catchments. An attempt was also done on finding out if the available water in the catchments can support the potential irrigation area of Ifumbo, Naming’ong’o and Mshewe irrigation schemes. It was noted that the low flows during dry season is a limiting factor for paddy production in dry season, however other crops which do not require a lot of water can be grown during this period