Abstract: The basic laws of water-sediment asynchrony in the reservoir area at sediment incoming moment were statistically analyzed based on the water-sediment data from main monitoring stations of the Three Gorges Reservoir in 2003-2018. The duration of sand peak transport affected by various factors during typical floods under different water levels was discussed. The regression formula for the transport time of sand peak with a good degree of fitting was also obtained via regression analysis. Results revealed that among 40 typical flood processes with peak discharge greater than 30,000 m³/s, the floods with sand peak arriving earlier than flood peak, sand peak arriving synchronously with flood peak, and sand peak lagging behind flood peak accounted for 42.5%, 42.5%, and 15%, respectively. After the operation of Xiangjiaba and other cascade reservoirs on the lower reaches of the Jinsha River, the propagation of sand peaks in the reservoir area elongated apparently. In view of the different ranges of water level fluctuation in front of dam during flood season, the rating and grouping of water level was considered to analyze the correlation between sand peak transport time and different influencing factors; furthermore, stepwise regression was adopted to establish the formula of sand peak propagation time with good fitting effect. The research results offer technical support for the sand peak dispatching of reservoirs.

Key words: Three Gorges Reservoir, water-sediment asynchrony, sand peak transport time, regression analysis, water level