| dc.description.abstract | Sandy sea beds of coastal regions, like the shallow Dutch part of the North sea, are often characterised by distinct rhythmic morphological features, such as sand waves. Sand waves are often located in intensely used regions of critical depth, thus their dynamics may endanger human activities, such as shipping and offshore engineering. Parameters controlling sand wave dynamics, like growth and migration are often investigated by means of (non-)linear stability models. These studies however, are only able to investigate the behaviour of small scale bed perturbations. This study aims to identify the finite sand wave characteristics in the Dutch coastal zone and to relate these to local environmental conditions. Therefore, an empirical analysis is used to identify the sand wave characteristics, observed in the approach routes of the Dutch harbours of Rotterdam and IJmuiden, situated in the North sea, and to find empirical relations to environmental parameters. In addition, these parameters are tested with a numerical shallow water model (Delft3D), with which a realistic, finite sand wave field can be simulated. First of all, it was found that the Rotterdam sand waves are about 250 m long and about 3.5 m high. The IJmuiden sand waves are 400 m long and 2.5 m high. Furthermore, clear differences in migration rates are found between the sand wave crests and troughs for IJmuiden (~2.5 m/year and ~0 m/year, respectively), while the sand wave crests and troughs migrate with about the same rates at Rotterdam (~0.5 m/year). From a principal component regression analysis of field observation of environmental parameters, follows that the strength of the dominant tidal constituent, the relative phase shift between the M2 and M4 tide, the median grain size, the local water depth and the sand wave height are important in explaining the differences in sand wave migration rates. Model results, using a process based numerical model, reveal that sand wave migration behaviour is simulated in good qualitative agreement with the field observations. Moreover, in order to obtain a difference in crest and trough migration, a difference in suspended load and bed load sediment transport over the sand wave is required. Finally, the same qualitative relations between sand wave migration and the relative phase shift between the M2 and M4 tidal constituents, the median grain size and the water depth are obtained with the model as was obtained from the empirical analysis. | |
