Abstract:In order to investigate the variation of water temperature at the head area of the Three Gorges Reservoir (TGR) from 2003 to 2018 and its impact on the ecological environment, the LSTM-Logistic model was used to simulate the multi-year water temperature process based on the field data at the Miaohe hydrological station. The influences of its variation on algal blooms in the tributary bay and the spawning water temperature of the downstream fish are discussed. The results show that the LSTM-Logistic model is applicable to the TGR, and it can accurately simulate the daily variation of water temperature. From 2006 to 2013, thermal stratification occurred at the head of the TGR from April to June. In 2014, the vertical differences of water temperature began to decrease and stratification disappeared. The average water temperature increased in spring, autumn, and winter, and the discharge water temperature further flattened. The long-term warming trend of the water temperature at the head area of the TGR will reduce the depth of the plunging point of the density current from the tributary bay, affect the pattern of the density current, and reduce the risk of algal blooms in the tributary bay. After the impoundment of the upstream cascade reservoirs, the impact caused by cold discharge water in spring is partly mitigated, while the increase of the water temperature in autumn and winter further threatened the spawning of Chinese sturgeon.