Numerical modeling of wave-current interaction with the use of a two-way coupled system

Abstract

The Dutch Flood Defense Act prescribes the assessment of the safety level of Dutch primary sea and flood defenses every six years. To asses if the required level of security of the sea and flood defenses is guaranteed, first the Hydraulic Boundary Conditions (HBC) which the sea and flood defenses are exposed to, have to be determined. At the eastern Wadden Sea, nonlinear wave-current interaction is important in the determination of wave fields and the wave set-up contributions to the surge elevation. The Hydraulic Boundary Conditions of the primary sea and flood defenses adjacent to the Wadden Sea are therefore determined with the use of a two-way coupled modeling system, existing of the circulation model Delft3D-FLOW and the wind wave model SWAN, to account for wave-current interaction. An inextricable consequence of modeling is the introduction of errors and the aim always should be to minimize these errors. The goal of this study is to improve the understanding of the two-way coupled modeling system and to increase the knowledge on the reliability of the result. A literature survey is performed with the goal to improve the understanding of the two-way coupled modeling system and to identify the largest sources of uncertainty. The relative importance of all sorts of physical processes is highly dependent on the set-up of the model schematization and the location. The effect on the result due to the applied coupling interval, wind drag parameterization in both Delft3D-FLOW and SWAN and different depth-induced wave breaking parameterizations is investigated at the eastern Wadden Sea, for the storm of 9 November 2007. Results showed that increasing the coupling interval mainly influenced the predicted water level, while the wave conditions were hardly influenced. The advice is to not increase the coupling interval to larger intervals than 30 minutes during storm conditions. The 'fit' wind drag parameterization Zijlema et al. (2012), is compared with Wu (1982) in SWAN and Charnock (1955) in Delft3D-FLOW. The largest differences are observed due to different wind drag parameterization in Delft3D-FLOW, though the effect of wind drag parameterization in SWAN possibly also might have an effect at higher wind speeds. The effect of applying the depth-induced wave breaking by Salmon et al. (2013) is investigated by comparing the results with the results generated when Battjes and Janssen (1978) is applied. Differences in wave set-up were in the order of 8 cm, while wave conditions showed better agreement with measurements in case of Battjes and Janssen (1978), compared with Salmon et al. (2013) with exception of low-frequency wave energy.