Tidal Meanders in the Ameland Basin

Abstract

The fairway connecting Holwerd and Nes forms the main transport route between Friesland and Ameland (the Dutch Wadden Sea). It consists of several tidal channels that funnel the majority of the tidal currents propagating through the tidal basin. Over the past decades, several channel bends have expanded leading to a fairway length-increase of about 1 km, and the channel in front of Holwerd has become subject to sedimentation. Since 1990 this channel was dredged and the volume of dredged material has increased exponentially. A relation between dredging activities, channel migration and frequent delays of the ferry has been suggested. The main goal of this thesis is therefore to explain the difference in evolution between the non-natural (intervened) and natural tidal meanders in the Ameland basin. To achieve this goal, the evolution of geographic channel dimensions are analyzed by assessing bathymetric data and the outcome is linked to the physics by means of hydrodynamic modeling. Channels have been classified based on their locations and dimensions. Three meandering stages have been identified that describe the basic channel behavior. The first meandering stage is the developed meander, characterized by a clear sinusoidal shape and with flood chutes. In this stage, the depth at a cross-over is approximately two times smaller than at the channel section with a maximum curvature (the channel top), where the width profile is funnel shaped. The developed, variable meander has similar characteristics, but differs by a lower correlation in width and depth, and closer to the tidal watershed the ebb channel location varies more in time. The third stage is the undeveloped meander, which is characterized by small and relatively straight channels, the absence of flood chutes, and a rather constant width- and depth-profile along the channel. In this study, the channels are evaluated that have similar meandering features and distance from the Ameland inlet as the dredged channels. In general, the examined channels showed a width reduction that dominates over the bed level decrease. An irregularity in the width and/or depth profile is an indicator for the development of a bifurcation or flood chute. Numerical modeling of the tidal currents revealed a correlation between the depth averaged velocity and the funnel shaped channel top. From the depth averaged velocities it is found that flow accelerates around the channel tops, where flood dominance prevails and flood chutes (are able to) develop. During ebb, a similar acceleration occurs towards the point of inflection, where ebb dominance prevails. Increasing meanders correspond to a shift towards ebb dominance in both natural and non-natural channels. By evaluating the secondary flow velocities along the channels, an eastward shift of the cross-over location is observed. Whether the shift corresponds to the channel migration, or that it strengthens the potential for channel splitting is not proven. However, an indication is found that a correlation exists between the distance between channel tops and the cross-over location shift. When the meanders develop, the distance between the tops increases, enabling a larger area over which the direction of the secondary flow can change. Furthermore, the orientation of the currents over the tidal flats and the angle under which they flow into the channel seems to be correlated. Also the importance of flood chutes to the meandering development is brought forward. The development of a chute increases the flow area and reduces the flow velocities in the ebb channel adjacent to the flood chute, which thereafter becomes shallower. After a flood chute is established, and it obtained a more west-east orientation, the flood currents over the tidal flat are deflected towards the ebb channel. As a consequence flow velocities increase in the ebb channel, resulting in a westward migration of the ebb channel. Comparing the natural and non-natural channels, the meandering extent (average sinuosity) is approximately 10\% larger for the non-natural channels. This sinuosity also increases over time in the natural channels, but to a lesser extent. By dredging, the irregularities in width and depth are flattened out and the flow through the ebb channel was maintained. This resulted in the non-natural development of the eastern Kikkertgat-bend and the stagnation of the flood chute development. Therefore it is concluded that a non-natural tidal meander in the Ameland basin is distinguished by developing meanders, where the flood chutes are not restricting the development of the meandering ebb channels.