Effects of initial and boundary conditions on the evolution of gravel-bed river

Understanding and predicting the morphological adaptation of rivers and in particular of their width, depth, bed slope, sinuosity and sediment composition is key to their sustainable management as these changes can impact the natural and the anthropogenic environments. The factors leading to morphological changes are alterations in boundary conditions in terms of discharge and sediment input regimes, level of receiving water bodies and vegetation characteristics caused by human interventions or natural events. The initial conditions of the river define its sediment transport capacity and channel characteristics at the start of these changes, which dictate the processes involved in the morphological evolution. Initial and boundary conditions might interact with one another and for this it is not clear whether the initial conditions affect also the new final river configuration, at the end of the morphological evolution. The role of different sediment and discharge regimes on river channel evolution has been studied already, but without considering the effects of the starting conditions. An extensive experimental study was carried out by Singh (2015) on a small flume, looking at the effect of initial width and different boundary conditions on the evolution of laboratory channels. Many questions still remain related to the applicability of the experimental results to a real rivers and to their long-term evolution, which was not always achievable in the laboratory. This work complements this experimental study by performing a similar investigation, but using a numerical model at a real river scale. In order to study the effects of initial and boundary conditions on river channel evolution, the experimental channels of Singh (2015) are upscaled and compared with existing rivers. An idealized virtual river is then reproduced in a two-dimensional depth-averaged model using the Delft 3D code based on the real river that is best represented by the laboratory channels. The model is calibrated to best simulate the trends and processes observed in the experiments and is validated on qualitative reproduction of the long-term width evolution in the base-case scenario (constant discharge and no sediment supplied at the upstream boundary). The model is then used to study the effects of initial and boundary conditions on the long-term morphological evolution at a real river scale, using different combinations of initial width, discharge and sediment input regime. The results show that the experimental channels of Singh (2015) represent steep mountainous gravel-bed rivers with large sediment size during high flow conditions. Most morphological processes and trends observed in the experiments were well replicated by the numerical model. The initial width seems to play a role on the long-term evolution of braid corridor width depending on the boundary conditions, but the active width of the river channel seems to always evolve towards a similar value, which is independent of initial conditions. The results also show that variable flow results in wider channels than constant flow, but the difference decreases for wider initial width channel. Finally, the channels obtained with sediment supply become wider than those with no sediment supply. The results increase the understanding in the role of initial and boundary conditions on the long-term evolution of gravel-bed rivers, focusing on how the trends and processes evolve towards a new equilibrium without considering the time scale of this evolution. However, the extrapolation of these results to actual gravel-bed rivers needs considering also the effects of vegetation on bank erosion and accretion affecting the channel width evolution