Graduation of Anna Kosters

04 juni 2019 16:00 - Locatie: Deltares (Boussinesqweg 1, Delft), Paviljoen 1 - Door: Webredactie

"Human interventions in river-estuary systems - A case study of the Loire, France" | Professor of graduation: prof. dr. ir. Z.B. Wang. supervisors: dr. ir. E. Mosselman (TU Delft/Deltares), dr. D.S. van Maren (TU Delft/Deltares).

Like many rivers around the world, the Loire river in France has a history of human interventions in order to facilitate navigation and port development. Next to affecting bed levels directly, the heavy modification of the river-estuary has induced significant changes in the hydrodynamic and morphodynamic behaviour of the Loire. As a result, the Loire river is experiencing bed degradation, with lowering of bed levels of up to 4 m over the last century (Pérard, 2018). In the estuary, the interventions have led to an amplification of the tidal amplitude, a shift of the tidal and salt intrusion limit in upstream direction and a larger tidal asymmetry. According toWinterwerp andWang (2013), this tidal deformation leads to an increased import of fine sediment in the estuary and a reduced hydraulic drag. This in turn enhances tidal deformation further. Due to this positive feedback mechanism the Loire estuary has evolved into a so-called hyper-turbid state, associated with large suspended sediment concentrations and the formation of fluid mud.

To counteract bed degradation, large-scale groyne remodelling, restoration of a secondary channel and the construction of a fixed layer with a higher roughness than the river bed are planned for implementation in the coming years. However, the behaviour of the river-estuary on large temporal and spatial scales and its response to historical and possible future interventions is not yet sufficiently understood. Therefore, an idealized, computationally efficient, three-dimensional morphodynamic model of the river-estuary is developed with the FLOW module of the Delft3D software suite (Deltares, 2014), in which the main processes influencing the morphology of the Loire river-estuary are included and interventions can be simulated. The system geometry is represented with one flow-carrying channel, with elevated floodplains and intertidal areas on both sides. Over the first 50 km from the river mouth, the width of the estuary converges exponentially, after which the channel is modelled as prismatic up to the landward model boundary, at 115 km upstream of the mouth. The converging section is called the estuary in this research, while the prismatic channel is referred to as the river section.

The current geometry of the estuary, lacking the presence of large intertidal areas, induces a flood-dominant tidal signal. The river discharge introduces a mean seaward directed velocity, but also enhances the flooddominance of the tide by preferential damping of the ebb tide. In the mouth of the estuary, the baroclinic pressure gradient introduces a mean landward directed velocity near the bed and amplifies the mean seaward directed velocity near the surface. Only when this pressure gradient is included, net sand and mud transport in the estuary mouth are landward-directed for almost all considered discharges (200 · Q · 6000 m3/s). The gravitational circulation causes the formation of an Estuarine Turbidity Maximum (ETM) at the tip of the salt wedge, which is strengthened further by the flood-dominance of the tide. The ETM forms closer to the mouth and has a smaller extent for larger discharges. Model results confirm the feedback mechanism between tidal deformation, the import of fine sediment and the effective hydraulic drag as described above. However, next to tidal deformation, strengthening of the gravitational circulation also plays a large role in this mechanism for the Loire. Furthermore,model results indicate that historical deepening of the estuary has led to retrogressive erosion in the river section.

To mitigate bed degradation, several measures are simulated that decrease the transport capacity of the flow, increase the sediment supply or do both. Over the first 10 years, sediment nourishments are the most effective according to the model. However, after 100 years of simulation the removal of all groynes present in the river section leads to the most sedimentation relative to the reference scenario. On this century-timescale, it is important to account for the influence of climate change when assessing the effects ofmeasures. To decrease the import of mud into the estuary, two measures have been investigated. Increasing bed levels in the estuary leads to weakening of both the tidal currents and the gravitational circulation. Large-scale restoration of tidal flats turns the tide from flood- into ebb-dominant. The salt intrusion length decreases as
well, which is here mainly due to weakening of the gravitational circulation. Both measures result in a tidally averaged export of fine sediment and very low mud concentrations, in the water column as well as in the bed. Model results appear sensitive to the schematization of the tidal motion at the downstream boundary. It is recommended to investigate further which simplifications of the tide are justified in situations where the interaction between river discharge and tide is significant. The sensitivity of model results to changes in morphological parameters and the configuration of measures should be tested as well. Furthermore, model results would benefit from an improved understanding and model representation of sand-mud interaction processes and fluidmud dynamics.