Asynchrony of flood peaks and sediment peaks in the Three Gorges Reservoir from 1990 to 2021

Sediment deposition is a critical issue for the long-term operation and maintenance of reservoirs.Implementing sediment peak discharge regulations based on the asynchronous characteristics of flood and sediment peaks is an effective measure for reducing sediment deposition and maintaining the reservoir capacity.However, existing methods for analyzing asynchronous characteristics, such as peak appearance time subtraction and SSC-Q loop analysis, are still limited in reflecting the overall asynchronous characteristics of the water-sediment process and accurately classifying asynchronous types.Here, we employ the dynamic time warping (DTW) algorithm to characterize flood event processes.We calculate the duration of asynchrony and determine the type of asynchrony by quantifying the graphical features of flood and sediment peaks.By examining flood events that occurred in the Three Gorges Reservoir (TGR) from 1990 to 2021, we analyze the asynchronous characteristics of flood and sediment peaks in the reservoir and obtain their spatiotemporal variations.In this study, we reveal that after the construction of the TGR and the Lower Jinsha River terrace reservoirs, the proportion of the sediment peak lagging at each station along the front range of the TGR before the dam increases, and the duration of asynchrony decreases with decreasing distance from the dam.Compared to traditional methods, the entire water-sediment change process during individual flood events can be considered in the DTW method, which can be used to effectively manage complex flood processes, demonstrating wide applicability.The DTW method exhibits notable potential and prospects in reservoir operation management and scheduling.