Effects of reduced precipitation and water level variability on phytoplankton dynamics and thermal structure in a large tropical reservoir

Drought is one of the most severe and complex natural hazards causing changes in water balance and impeding crucial water-dependent ecosystem services. Over the last 25 years, multiple severe droughts linked to El Niño events have affected the Zambian and Zimbabwean regions. As a result, Lake Kariba, world's largest artificial reservoir by volume, saw historically low water levels and power shortages from reduced hydropower generation. This study employs a novel integrated approach to investigate the effects of varying water levels on the spatio-temporal dynamics of phytoplankton blooms across the lake, combining remote sensing, in situ measurements, and a three-dimensional hydrodynamic model (Delft3D). Satellite Earth Observations of lake water level, surface water temperature, and Chlorophyll-a concentration are used, and spatially resolved water transparency is estimated from lake water leaving reflectance. The Delft3D model is calibrated and validated against in situ and satellite data, including Rainfall Estimates from Rain Gauge and Satellite Observations (CHIRPS), and in situ discharge data, while assimilating maps of water transparency to simulate surface thermal stratification accurately. Results show pronounced spatial and seasonal heterogeneity in transparency, thermal structure and phytoplankton dynamics across the three basins. Phytoplankton phenology varies from short, pulse-like blooms in the most sensitive riverine upstream basin to longer, more coherent bloom seasons in downstream transitional and lacustrine basins. Bloom hotspots occur near shorelines and tributary mouths during windy and wet seasons, following basin-scale circulation patterns and possibly nutrient inflows. Our integrated approach captures Lake Kariba's heterogeneity and identifies vulnerable areas, highlighting its importance for effective water resource management.