Selective water release from the deeper pools of reservoirs for energy generation alters the temperature of downstream rivers. Thermal destabilization of downstream rivers can be detrimental to riverine ecosystem by potentially disturbing the growth stages of various aquatic species. To predict this impact of planned hydropower dams worldwide, we developed, tested and implemented a framework called ‘FUture Temperatures Using River hISTory’ (FUTURIST). The framework used historical records of in-situ river temperatures from 107 dams in the U.S. to train an artificial neural network (ANN) model to predict temperature change between upstream and downstream rivers. The model was then independently validated over multiple existing hydropower dams in Southeast Asia. Application of the model over 216 planned dam sites afforded the prediction of their likely thermal impacts. Results predicted a consistent shift toward lower temperatures during summers and higher temperatures during winters. During Jun-Aug, 80% of the selected planned sites are likely to cool downstream rivers out of which 15% are expected to reduce temperatures by more than 6˚C. Reservoirs that experience strong thermal stratification tend to cool severely during warm seasons. Over the months of Dec-Feb, a relatively consistent pattern of moderate warming was observed with a likely temperature change varying between 1.0 to 4.5˚C. Such impacts, homogenized over time, raise concerns for the ecological biodiversity and native species. The presented outlook to future thermal pollution will help design sustainable hydropower expansion plans so that the upcoming dams do not face and cause the same problems identified with the existing ones.

River temperature is projected to increase in the southeastern United States (SEUS) due to climate change, exacerbating the invasion of warm-water species and reducing suitable habitats for cold- and cool-water species. However, the response of river thermal regimes to climate change is also influenced by human activities, especially dam construction and operation. Large dams impound deep reservoirs, expand water surface area and prolong water residence time, modifying the interaction of surface meteorology with river systems. During warm seasons, surface energy fluxes can only heat the top layer (epilimnion) in deep reservoirs with bottom layer (hypolimnion) remaining cold. This vertical temperature gradient is called thermal stratification. Cold hypolimnetic releases from stratified reservoirs changes downstream thermal regimes that can expel indigenous warm-water species yet provide an ideal habitat for introduced cold-water species. For example, multiple species of trout (Family: Salmonidae) have been introduced to tailwaters downstream of multiple dams operated by the Tennessee Valley Authority, which has become a popular and lucrative recreational fishing location in the SEUS. Previous research has shown that reservoir thermal stratification will be retained under climate change, but stronger surface energy fluxes warm downstream river temperature, suggesting there will be a future decline in cold-water species habitat and a corresponding increase in local warm-water species habitat. In this study, we used a physically-based modeling method to simulate river temperatures, explicitly considering the impact of thermal stratification. The SEUS has a highly regulated river system and diverse freshwater fish species. We mapped the suitable habitats for selected cold-water and warm-water fish species by comparing the simulated river temperature against their physiological constraints. Model experiments were designed to quantify the impacts of dam operation by simulating river temperature for both regulated and unregulated scenarios. Potential ecological consequences under climate change were analyzed through projected changes in river thermal regimes, e.g., shrinking habitats for cold-water species and restoring local warm-water species.