Modelling Density-Driven Secondary Flow with a 2-D Depth-Integrated Model: Insights from the Yangtze River-Poyang Lake Confluence

In river confluences, the two branches may have different water temperatures and sediment loads, which induce strong transverse density gradients. These gradients, in turn, drive the formation of secondary currents, which interact with those generated by channel curvature. Specifically, density gradients can either enhance or counteract curvature-induced secondary flows, and their impacts on flow and sediment transport require proper modelling approaches. Three-dimensional (3D) models naturally account for these dynamics and provide detailed predictions of flow and temperature fields, but cannot be applied to long-term morphodynamic simulations because of prohibitive computational demand. By contrast, traditional two-dimensional (2D) models are computationally more efficient, but do not account for 3D flow structures that are particularly relevant for river confluences. To fill the gap, a 2D depth-integrated hydro-morphodynamic model is enhanced, through appropriate parametrization, to account for the density-driven secondary flows and their effects on the flow field, mixing, sediment redistribution and, ultimately, on the morphodynamic evolution of the riverbed.

The enhanced 2D model is applied to the Yangtze River-Poyang Lake confluence, where field measurements have shown that temperature-induced density gradients play a critical role in shaping flow patterns, secondary currents, and the riverbed evolution. Interestingly, these effects vary throughout the year due to seasonal differences in temperature and discharge between the two branches of the confluence. Density-induced secondary currents, which superimpose or modify the curvature-induced helical flows, develop at the confluence apex where the two streams merge. Their inclusion in the 2D modelling framework improves the agreement of numerical results with ADCP field measurements, thus supporting the reliability of the model.

The efficiency of the 2D model, combined with its ability to represent key physical processes through the parametrization of density-driven effects, also allows to perform long-term simulations with mobile bed conditions. These simulations highlight the significant role of secondary flows, driven by both streamline curvature and spanwise density gradients, in sediment transport and bed morphology at the river confluence, confirming that the enhanced 2D model is a valuable tool for long-term morphodynamic studies in large river systems.