Use of globally available data for determining water level variation in lakes: the case study of Lake Turkana, Kenya

Abstract

This study identifies relevant globally available data to develop basin and water level variation models which can be used for scenario modelling. The case study for the current research is Lake Turkana in Kenya, which is the outlet of 3 main river basins: Omo, Turkwel and Keiro. Due to the lack of data, the present paper only considers the Omo River Basin, which contributes around 90% of the inflow into the lake. Lumped and semi-lumped hydrological models were developed using the USACE’s HEC-HMS tool. The calibration and validation were carried out using the EU computed discharges of the Global Flood Awareness System as observed flow. The validated model was used to assess the impact on the lake due to the operation of three reservoirs in the basin; Gibe I, II and III. The variation of the water level of Lake Turkana was determined by using the model builder tool in ArcGIS. The lake water level model was validated using the Normalized Difference Water Index of Landsat 5 and Landsat 7 satellite imagery. Finally, using the semi-lumped hydrological model and the lake water level variation model, two scenarios were simulated to assess the lake‘s condition due to three consecutive driest and wettest years. The semi-lumped model showed satisfactory goodness of fit compared to the lumped model, therefore it is concluded that it can be used to assess the impact of upstream developments in the Omo basin, on the shoreline and water level variations of Lake Turkana with reasonable accuracy.