Morphological model for the River Meuse : Sambeek-Grave v0.8

Recent events, such as the flood in July 2021 in the Meuse River in which large changes in bed level posed problems to infrastructure, have shown that up-to-date morphodynamic models are essential to aid in evaluation of measures and policy choices. The Meuse River at present does not have a validated numerical model for morphodynamics. To uniformise the approaches on the Rhine and the Meuse, a plan was developed for the construction of a two-dimensional morphodynamic model of the Meuse which builds on the recently developed hydrodynamic D-HYDRO model of the Meuse. The current report shows the development of the version v0.8 model for the reach between Sambeek and Grave. The simulation has undergone a 1D calibration, using the sedimentation volume in the summerbed lowering at rkm 155. Therefore, this model should not yet be used for studies. The morphodynamic model uses the Simulation Management Tool (SMT) to speed up the computation. An original flow time of 4 years reduced to 22.5 days, enabling the computation to run in just over 33 hours. The model uses a locally refined mesh in the main channel. The hydrodynamics of the model compare reasonably to water-level-slope and velocity measurements. The division of the flow between the main channel and floodplain show expected behaviour, but are not verified as measurements do not exist. An exploratory data analysis reveals the width-averaged changes in the bed level from measurements during the period after the summerbed lowering. A discharge-dependent parameter investigation shows the behaviour of the model and reveals that the water level does not match the steering water level at Mook for low discharges, indicating that the Q-h boundary condition does work correctly. Furthermore, the calibration procedure of the hydrodynamic model yields friction coefficients that cannot be explained from the physical point of view and that cause unrealistic patterns of sediment transport. The model calibration consists of the offline computation of transport conditions for the period between July 2014 and June 2018, where the volume of sediment deposited in the summerbed lowering is used as a calibration goal. Finally after different morphological computations reasonable agreement was found, except for the downstream reach near the weir of Grave where the model predicted sedimentation, whereas the measurements show a decrease in bed level. Further investigation is required to understand what causes this difference. The model settings and underlying assumptions are discussed, and an outlook for further model investigation are provided.