Marine and Coastal Morphology: medium term and long-term area modelling

Abstract

This thesis documents development and application of a modelling concept developed in collaboration between DTU and DHI. The modelling concept is used in morphological modelling in coastal areas where the governing sediment transport processes are due to wave action. The modelling concept is defined: Hybrid morphological modelling and it is based on coupling calculated sediment transport fields from a traditional process based coastal area model with a parametrised morphological evolution model. The focus of this study is to explore possible parametric formulations of the morphological evolution model and apply them to problems concerning coastal protection strategies (both hard and soft measures). The applied coastal protection strategies involve morphological impact of detached shore parallel segmented breakwaters and shore normal impermeable groynes in groyne fields, and morphological evolution around seawalls and response to beach and shoreface nourishment. The hybrid morphological modelling concept is introduced and is put into context with existing models used for medium term and long-term morphological simulation. The modelling concept is intended to improve the long-term predictive capabilities of process based area models thereby bridging the gap between short term models and long-term models. A number of different implementations within the hybrid morphological modelling concept have been developed and tested in order to ascertain the usefulness of the concept. The implementations are grouped into shoreline models and models used to develop local features on the coastal profile such as alongshore migration of a bar front or redistribution of a shoreface nourishment. The implementations are tested on a variety of problems (idealised cases and case studies at specific locations). Applications of the modelling concept have been collected into 5 papers (1 accepted journal paper, 3 published conference papers and 1 unpublished journal manuscript) which each contain an introduction to the implementation, the motivation for the implementation and the overall results from the application. The 5 papers are enclosed as separate chapters of this thesis. The overall conclusions to each of the model implementations are given as follows. 1D shoreline model Implementing the hybrid morphological modelling concept with a 1D morphological model is a strong engineering tool. The model is robust and computationally efficient and it may be adapted to real engineering problems. The results of the 1D shoreline model are however greatly affected by the imposed freedom of the model, and exaggerated alongshore smoothing of the calculated shoreline may occur if the 1D model is applied to problems in which the true solution has a two dimensional nature. 1.5D shoreline model A so-called “1.5D” implementation which introduces redistribution of sediment within a coastal profile in response to horizontal 2D currents makes it possible to simulate the morphological development in areas where 2D evolution occurs. The coastal profiles tend however to drift due to the fact that the response of an entire profile is coupled thereby in some cases leading to morphological activity in inactive areas. Diffusion of the coastal profile is therefore introduced in order to weakly force the profile towards an equilibrium form. The 1.5D model is seen to produce reasonable results when subject to cases with detached breakwaters and groynes. The computational efficiency of the model is however reduced compared to the 1D model, because the increased freedom of the model reduces the maximum stable morphological time step. Bar models The models for redistributing sediment on the shoreface have been applied with limited success. The models can only be applied to cases where cross-shore transport processes are of minor importance, due to inaccuracies in the calculated cross-shore transport in the applied 2D coastal model. Stabilisation of the bar parameters using behaviour-oriented methods may be possible but will require a strong forcing which according to the author reduces the significance of the modelling concept.