Resumen
We develop and present a novel integrated water balance model that accounts for lake water?groundwater interactions, and apply it to the semi-closed freshwater Lake Babati system, Northern Tanzania, East Africa. The model was calibrated and used to evaluate the lake level sensitivity to changes in key hydro-climatic variables such as temperature, precipitation, humidity and cloudiness. The lake response to the Coupled Model Intercomparison Project, Phase 5 (CMIP5) output on possible future climate outcomes was evaluated, an essential basis in understanding future water security and flooding risk in the region. Results show high lake level sensitivity to cloudiness. Increased focus on cloud fraction measurement and interpretation could likely improve projections of lake levels and surface water availability. Modelled divergent results on the future (21st century) development of Lake Babati can be explained by the precipitation output variability of CMIP5 models being comparable to the precipitation change needed to drive the water balance model from lake dry-out to overflow; this condition is likely shared with many other East African lake systems. The developed methodology could be useful in investigations on change-driving processes in complex climate?drainage basin?lake systems, which are needed to support sustainable water resource planning in data scarce tropical Africa.