Nat. Hazards Earth Syst. Sci., 16, 2373–2389, 2016 www.nat-hazards-earth-syst-sci.net/16/2373/2016/ doi:10.5194/nhess-16-2373-2016 © Author(s) 2016. CC Attribution 3.0 License. Coastal flooding: impact of waves on storm surge during extremes – a case study for the German Bight Joanna Staneva1, Kathrin Wahle1, Wolfgang Koch1, Arno Behrens1, Luciana Fenoglio-Marc2, and Emil V. Stanev1 1Institute for Coastal Research, HZG, Max-Planck-Strasse 1, 21502 Geesthacht, Germany 2Institute of Geodesy and Geoinformation, University of Bonn, Nussallee 17, 53115 Bonn, Germany Correspondence to: Joanna Staneva ([email protected]) Received: 23 June 2016 – Published in Nat. Hazards Earth Syst. Sci. Discuss.: 15 July 2016 Revised: 23 October 2016 – Accepted: 31 October 2016 – Published: 21 November 2016 Abstract. This study addresses the impact of wind, waves, 1 Introduction tidal forcing and baroclinicity on the sea level of the German Bight during extreme storm events. The role of wave-induced processes, tides and baroclinicity is quantified, and the results A challenging topic in coastal flooding research is the ac- are compared with in situ measurements and satellite data. A curate prediction of sea surface elevation and wave heights. coupled high-resolution modelling system is used to simu- This is highly relevant over the European shelf, which is late wind waves, the water level and the three-dimensional characterized by vast near-coastal shallow areas and a large hydrodynamics. The models used are the wave model WAM near-coastal urban population. The increased demand to im- and the circulation model GETM. The two-way coupling prove wave and storm predictions requires further develop- is performed via the OASIS3-MCT coupler. The effects of ment and improved representation of the physical processes wind waves on sea level variability are studied, accounting in ocean models. The wind-induced surface stress over the for wave-dependent stress, wave-breaking parameterization ocean plays an important role in enhancing sea surface eleva- and wave-induced effects on vertical mixing. The analyses tion (e.g. Flather, 2001). The importance of wind and wave- of the coupled model results reveal a closer match with ob- induced turbulence for the ocean surface layer was demon- servations than for the stand-alone circulation model, espe- strated by Davies et al. (2000), and it was further demon- cially during the extreme storm Xaver in December 2013. strated for the bottom layer by Jones and Davies (1998) and The predicted surge of the coupled model is significantly en- for the wave-induced mixing by Babanin (2006) and Huang hanced during extreme storm events when considering wave– et al. (2011). Craig and Banner (1994) and Mellor (2003, current interaction processes. This wave-dependent approach 2005, 2008) suggested that surface waves can enhance mix- yields a contribution of more than 30 % in some coastal ar- ing in the upper ocean. Qiao et al. (2004) developed a param- eas during extreme storm events. The contribution of a fully eterization of wave-induced mixing from the Reynolds stress three-dimensional model compared with a two-dimensional induced by wave orbital motion and coupled this mixing with barotropic model showed up to 20 % differences in the water a circulation model. They found that wave-induced mixing level of the coastal areas of the German Bight during Xaver. can greatly enhance vertical mixing in the upper ocean. The improved skill resulting from the new developments jus- Understanding the wave–current interaction processes is tifies further use of the coupled-wave and three-dimensional important for the coupling between the ocean, atmosphere circulation models in coastal flooding predictions. and waves in numerical models. Longuet-Higgins and Stew- art (1964) showed that wave and dissipation-induced gradi- ents of radiation stress account for a transfer of wave mo- mentum to the water column, changing the mean water level. The effects of waves on the lower marine atmospheric bound- ary layer have been demonstrated by a number of studies: Janssen (2004), Donelan et al. (2012) and Fan et al. (2009), Published by Copernicus Publications on behalf of the European Geosciences Union. 2374 J. Staneva et al.: Coastal flooding: impact of waves on storm surge during extremes and for the light wind regimes Veiga and Queiroz (2015) 2002). The nested-grid model setup for the German Bight and Sun et al., (2015). The effects of wave–current interac- has a horizontal resolution of 1 km and 21 σ layers (Fig. 1) tions caused by radiation stress have also been addressed by (Stanev et al., 2011). GETM uses the k-" turbulence clo- Brown and Wolf (2009) and Wolf and Prandle (1999). A dif- sure to solve for the turbulent kinetic energy k and its dis- ferent approach, i.e. the vortex force formulation, was used sipation rate ". The data for temperature, salinity, veloc- by Bennis and Ardhuin (2011), McWilliams et al. (2004) ity and sea surface elevation at the open boundary are ob- and Kumar et al. (2012). The comparison of both meth- tained from the coarser resolution (approximately 5 km and ods by Moghimi et al. (2013) showed that the results are 21 σ layers) North Sea–Baltic Sea GETM model configura- similar for longshore circulations, but radiation stress en- tion (Staneva et al., 2009). The sea surface elevation at the hanced offshore-directed transport in wave shoaling regions. open boundary of the outer (North Sea–Baltic Sea) model Many other studies addressed the role of the interaction be- was prescribed using 13 tidal constituents obtained from the tween wind waves and circulation in the model simulations satellite altimetry via OSU Tidal Inversion Software (Eg- (Michaud et al., 2012; Barbariol et al. 2013; Brown et al., bert and Erofeeva, 2002). Both models were forced by atmo- 2011; Katsafados et al., 2016; Bolaños et al., 2011, 2014; spheric fluxes computed from bulk aerodynamic formulas. Röhrs et al., 2012). These formulas used model-simulated sea surface tempera- Storm surges are meteorologically driven, typically by ture, 2 m air temperature, relative humidity and 10 m winds wind and atmospheric pressure. As shown by Holleman and from atmospheric analysis data. This information was de- Stacey (2014), an increasing water level decreases the fric- rived from the COSMO-EU regional model operated by the tional effects in the basin interior, which alters tidal amplifi- German Weather Service (DWD, Deutscher Wetter Dienst) cation. Waves combined with higher water levels may break with a horizontal resolution of 7 km. River run-off data were dykes, cause flooding, destroy construction and erode coasts provided by the German Federal Maritime and Hydrographic (Pullen et al., 2007). Waves can also modify the sediment Agency (BSH, Bundesamt für Seeschifffahrt und Hydrogra- dynamics (Grashorn et al., 2015; Lettman et al., 2009). phie). The German Bight is dominated by strong north-westerly winds and high waves due to north-eastern Atlantic low- 2.2 Wave model pressure systems (Rossiter, 1958; Fenoglio-Marc et al., 2015). Extratropical cyclones in the area present a consid- Ocean surface waves are described by the two-dimensional erable hazard, especially in the shallow coastal Wadden Sea wave action density spectrum N(σ , θ, ', λ, t) as a function of areas (Jensen and Mueller-Navarra, 2008). Coastal flooding the relative angular frequency σ, wave direction θ, latitude ', can be caused by the combined effects of wind waves, high longitude λ and time t. The appropriate tool to solve the bal- tides and storm surges in response to fluctuations in local ance equation is the advanced third-generation spectral wave and remote winds and atmospheric pressure. The role of model WAM (WAMDI group, 1988; ECMWF, 2014). The these processes can be assessed using high-resolution cou- use of the wave action density spectrum N is required if cur- pled models. However, in the frame of forecasting and cli- rents are taken into account. In that case, the action density mate modelling studies, the processes of wave and current in- is conserved, in contrast to the energy density, which is nor- teractions are not sufficiently exploited. In this study, we ad- mally used in the absence of time-dependent water depths dress wave–current interaction to assess the impact of waves and currents. The action density spectrum is defined as the on the sea level of the German Bight during extremes. We energy density spectrum E(σ, θ, ', λ, t) divided by σ ob- quantify their individual and collective roles and compare served in a frame moving with the ocean current velocity the model results with observational data that include various (Whitham, 1974; Komen et al., 1994): in situ and remote sensing measurements. The wave model E(σ;θ/ N(σ;θ/ D : (1) (WAM), circulation model (GETM), study period and model σ experiments are presented in Sect. 2. The observational data are described in Sect. 3, followed by model–data compar- The wave action balance equation in Cartesian coordinates is isons in Sect. 4. Finally, Sect. 5 addresses the effects of the given as different physical processes on sea level variability, followed @N @cσ N @cθ N by concluding remarks in Sect. 6. C .c C U/r N C C @t g xy @σ @θ S C S C S C S C S D wind nl4 wc bot br : (2) σ 2 Models The first term on the left side of Eq. (2) represents the local 2.1 Hydrodynamic model rate of change of wave-energy density and the second term describes the propagation of wave energy in two-dimensional The circulation model used in this study is the General Es- geographical space, where cg is the group velocity vector and tuarine Transport Model (GETM, Burchard and Bolding, U is the corresponding current vector.