Effects of the Long-Term Climate Change and Selective Discharge Schemes on the Thermal Stratification of a Large Deep Reservoir, Xin’anjiang Reservoir, China

The effects of global warming and precipitation changes on water temperature and thermocline parameters, such as thermocline depth, thickness, and strength, were assessed. A catchment model, coupled with a reservoir thermal model with meteorological input calculated by a downscaled general circulation model (GCM) projection under three representative concentration pathways (RCPs), was applied to the Xin’anjiang Reservoir, located in southeast China. The results indicate that water temperature in each layer increased (decreased) with the rise (decline) in air temperature, especially the surface water temperature. There was a significant negative (positive) correlation between thermocline depth (strength) and air temperature during the period of stratification weakness. The most sensitive phenomenon of water temperature-to-precipitation changes occurred in the middle layer (depth = 30 m). Additionally, the thermocline depth and thickness increased with decreases in hydraulic residence time, which were caused by precipitation increases. According to the simulation experiments driven by RCP outputs, mean water temperature in each water layer in the future (2096–2100) has a strong response to increases in air temperature, which is projected to increase by 0.11–0.62 °C for RCP2.6, 0.76–1.19 °C for RCP4.5, and 1.50–2.35 °C for RCP8.5, compared to the baseline (2012–2016). However, mean water temperature in each water layer from 2096 to 2100 underwent a slight decrease caused by precipitation changes, with a 0.03–0.25 °C decrease for RCP2.6, 0.07–0.40 °C for RCP4.5, and 0.04–0.29 °C for RCP8.5, compared to 2012–2016. The mean thermocline depth in the future (2096–2100) will be significantly decreased, while the mean thermocline thickness will be slightly increased. Over a multiyear timescale, the impacts of air temperature changes are stronger than those induced by precipitation variations. However, the effects of hydraulic residence time changes caused by precipitation changes (especially rainstorm) should be considered in the management of deep reservoirs.