Water dynamics of large controlled lakes in a global flood model: the case of Lake Victoria

A new satellite-based reservoir operation module is implemented in the Hydrological Modeling and Analysis Platform (HyMAP) global flood model and tested over Lake Victoria, in Eastern Africa, where water levels have been controlled at the Nalubaale hydropower dam since 1954. Lake Victoria is the world's second largest freshwater lake by surface area. HyMAP simulates river and floodplain dynamics through a one-dimensional local inertia formulation, allowing the representation of backwater effects, essential for a physically-based simulation of lakes and reservoirs. River network parameters at 0.1°were derived from the upscaling of the Multi-Error-Removed Improved-Terrain (MERIT) digital elevation model combined with a ground-based high-resolution lake bathymetry. Radar altimetry data, available since 1992, is used as a proxy to inform the model on dam operation rules. Daily runoff was generated by the Noah land surface model with multiparameterization options(Noah-MP) and used as input in HyMAP. Results show that lake water extent and storage reached extremes in 1998 and 2006. Lake extent simulations show high accuracy when compared to MODIS data and we find that lake storage variability is a major component of GRACE-based terrestrial water storage estimates. Comparisons against a simulation of the naturalized system (i.e., no anthropogenic impacts) show that dam operation rules significantly affect the outflow peak lag and amplitudes. These findings show the impact of lakes and reservoirs in global flood models and directly contribute to their representation in water monitoring systems.