Analyzing morphodynamics along a river widening: Applicability of different transport equations in 2D numerical modeling

Over the past two centuries, river regulation practices in Europe have significantly altered river systems through straightening, channelization and bedload retention. These modifications, coupled with the implementation of transverse structures such as hydropower facilities, have adversely affected riverine ecosystems, floodplains and sediment dynamics. River widening projects aim to address these challenges by creating more space for rivers, thereby improving the health of these natural systems. In Tyrol, Austria, between Stams and Rietz, a restoration project on the Inn River included the removal of most bank protection, the widening of the river up to 75 meters, and the creation of a dead branch and a side channel. On a length of 3 km, this measure re-established aquatic as well as terrestrial habitats. Shortly after completion, a 50-year flood event caused significant changes along the restoration zone, including the breaching of the dead branch, which subsequently connected to the main channel. These morphodynamic changes were documented using two airborne laser bathymetry (ALB) surveys and an echo-sounding survey for cross-sectional profiles.

Morphodynamic models are key tools for understanding sediment transport processes in rivers and providing insights into riverbed dynamics and sediment budgets over time. For this study, the Telemac2D hydrodynamic model, coupled with the Gaia sediment transport module, was employed to simulate the hydrograph of the HQ50 flood event. The model accounted for complex bedload behavior, including lateral slope effects and bank failure, which are essential processes in river restoration. A stable sediment budget with inflow rates equal to outflow rates was assumed due to the uncertainty in bedload inflow rates. A sensitivity analysis was conducted using various transport equations, including Meyer-Peter & Müller, Einstein & Brown, Hunziker, and Wilcock & Crowe. To assess model performance, metrics such as mean change in elevation (MCE), root mean square error (RMSE), and a newly developed erosion and deposition pattern index (EDPI) were analyzed. All transport equations replicated the general survey patterns, with the Meyer-Peter & Müller equation achieving the lowest MCE and RMSE errors and the highest EDPI values. Despite these promising results, unrealistic behavior was observed since none of the transport models accounted for the movement of the coarsest sediment fraction, leading to bed coarsening as fine material was preferentially transported out of the model. Furthermore, the Hunziker and Wilcock & Crowe equations yielded unrealistically low transport rates, resulting in reduced erosion and deposition compared to the Meyer-Peter & Müller and Einstein & Brown equations. These limitations highlight uncertainties in shear stress calculations for alpine rivers characterized by large particle sizes. Further research is recommended to address these issues and enhance the accuracy of sediment transport modeling in similar contexts.