Abstract: The analysis of hydrological regime change and attribution is essential for water resource management in a changing environment. This study focuses on the Xinjiang River Basin and examines the non-coherent changes in annual average flow and water level over a 70-year period, using hydrological data collected from the Meigang hydrological station between 1953 and 2020. A random forest model is employed to simulate flow and water level during the impact period, and the contribution of climate change and human activities to the annual average flow and water level is quantified using the residual analysis method. Considering the evolving hydrological conditions at the Meigang station, a method is proposed to calculate the design lowest navigable water level. This involves using the guaranteed rate-frequency method to calculate the design minimum navigable flow, followed by determining the design lowest navigable water level under changing environmental conditions using water level-flow relationship curves from different periods. The results reveal a weak upward trend in the annual average flow series at Meigang Station, with the largest difference observed between the earlier and later sub-series occurring around 1991, primarily driven by climate change. Conversely, the annual average water level series exhibits a significant downward trend, with the largest difference between the earlier and later sub-series observed around 2002, predominantly influenced by human activities. When considering an 80% design frequency, the design lowest navigable water level at Meigang station in 2020 (representing the current environment) is 1.78 m and 1.79 m lower than those in 1991, corresponding to annual guarantee rates of 95% and 98%, respectively. The findings of this study provide valuable insights for navigation, as well as the sustainable development and utilization of water resources in a changing environment.