The morphological response of the Dutch Rhine branches to interventions dependssensitively on the morphological development of the bifurcations. Computations inthe 1990s revealed that proper modelling of this morphological development requiresthe inclusion of physical mechanisms for grain sorting. Therefore we apply a two-dimensional (2D) morphological Delft3D model with graded sediment to the Rhinebifurcation at Pannerden ("Pannerdensche Kop"). We find that such a computation,unlike computations with uniform sediment, hardly produces any changes in bedtopography at a discharge of 2400 m3/s. Only a less frequent flood discharge of6000 m~3/s is found to produce topography changes. Apparently, the highest floodsneed to be included when modelling the combined evolution of bed topography andbed sediment composition. We explain this from the time scales for bed topographydevelopment and the development of bed sediment composition. If the compositionpattern develops fast with respect to bed topography, the sediment is rearrangedimmediately in a way that eliminates the gradients in sediment transport capacity,after which the bed topography remains unchanged. We show theoretically that theratio of transport layer thickness to flow depth is the key parameter that determinesthe ratio of these time scales. The findings also suggest that existing theories tend tounderestimate the thickness of the transport layer, with the important implicationthat computational results may give the false idea of a stable bed topography in caseswhere the real topography is not stable at all.
展开▼
