An optimized entropy-based model for estimating river confluence hydrodynamics: Accounting for the effects of velocity dip

The entropy model has been successfully used to calculate the flow velocity profile of rivers according to the measured surface-flow velocity. However, it has limitations in the large river where the obvious secondary flow and flow dip (i.e., the location of maximum flow velocity moves down from the surface) occur. The confluence of rivers is usually characterized by strong secondary circulations, posing great challenges for the accurate flow velocity and discharge estimation. This paper aims to assess the effectiveness of the previous entropy-based approach in velocity estimation of a large confluence and then propose an optimized method. The flow velocity data of the large confluence between the Yangtze River and the Poyang Lake under different discharge ratios were used. The performance of the previous entropy model got worse for the stronger secondary circulation survey. To better estimate the confluence hydrodynamics, a moving average method was introduced with its advantage of smoothing data and integrated into the framework of the entropic method for the better assessment of dip parameter, and hence the better estimation of flow velocity profiles. Furthermore, the maximum velocity was found to locate at one third of the total depth below the water surface. The range of entropic function 0.48 ~ 0.68 could provide a basis for the determination under various flow conditions in rivers. The value of entropic function and the dip parameter got larger as the discharge ratio decreased. This research has expanded the application for estimating hydrodynamics in large rivers.