Numerical models become attractive means of study, when considering the limited knowledge and guidance on functional design of coastal groins. In the present article, a process-based 3D numerical model (Delft3D) is verified against two different datasets. For application on design of straight open (normal) groins, Badiei et. al. 1994 laboratory measurements, and in case of T-Head groins Ozolcer et. al. 2004 field measurements are considered. In case of normal groins, the model showed good agreement with the observations, hence, a design exercise and method has been developed. In case of T-Head groins, the model showed sensitivity to the numerical representation of the structure, while still satisfactory resemblance was perceived. Comparing to the present design rules, application of a numerical study is of crucial importance.

Deterministic and probabilistic Profile models have been compared with hydrodynamic and morphodynamic data of laboratory and field experiments on the time scale of storms and seasons. The large-scale laboratory experiment is a pure 2D case and offers and ideal test case for cross-shore Profile models, as disturbing alongshore non-uniformities are absent. The field experiments are performed at the Eginond site (The Netherlands) during the EU-COAST3D project and represent storm time scale (Oct.-Nov. 1998) as well as seasonal time scale conditions (May 1998-Sep. 1999). The objective of the paper is to present information of coastal processes on these time scales and to assess the predictive capabilities Coastal Process-based Profile models with respect to hydrodynamics and morphodynamics at sandy beaches on the time scales of storms and seasons. Profile models can quite accurately (errors smaller than 10 represent the cross-shore significant wave height distribution in the surf zone, if the wave breaking model is properly calibrated. The wave breaking coefficient should be a function of local wave steepness and bottom slope for most accurate results. Profile models can reasonably represent the cross-shore and longshore currents (undertow) in a pure 2D case and in 3D field conditions. Profile models including cross-shore mixing effects and breaker delay effects do not produce better predictions of the longshore and cross-shore current velocities. Profile models using default settings can quite reasonably simulate the behaviour of the outer and inner bars on the storm time scale; the behaviour of the beach cannot be modelled with sufficient accuracy on the storm time scale. Profile models can reasonably simulate the post-storm onshore bar migration, provided that the near-bed orbital velocities and wave asymmetry-related sand transport are represented in a sufficiently accurate way (using non-linear wave theories). Profile models cannot simulate the beach recovery processes on the post-storm time scale, because these essentially 3D processes are not sufficiently known to be included in the models. Profile models using default settings cannot simulate the behaviour of the outer and inner bars and the beach on the seasonal time scale; the behaviour of the outer bar on the seasonal time scale can only be represented properly after tuning using measured bed profiles, The simulation of the inner bar and beach morphology on the seasonal time scale could not be improved by tuning, (C) 2002 Elsevier Science B.V All rights reserved.