Atmosphere 2022, 13, 1634 
6 of 21 
from [42]. As part of the classification, the gale index is then classified as NUL (“No Gale”), 
G (“Gale”), SG (“Severe Gale”) and VSG (“Very Severe Gale”) if it exceeds the 90th, 98th 
and 99.73rd percentile of the climatological G* distribution based on a reference period of 
‘971-2000 [43]. 
2.2.4. Effective Wind 
A combination of large scale wind direction (as implicitly marked by the LWT) and 
the gale strength can be derived by defining an effective wind. The effective wind is the wind 
component (or the part of the wind vector) that has the strongest effect on the water level 
at a specific location. In the case of the NSSs considered here, this is the wind direction 
driving the water out of the Elbe estuary resulting in an ELW. 
In order to identify the wind direction for the effective wind at Cuxhaven, a composite 
analysis for the 24 h before the respective 271 ELWs was conducted by averaging the 
respective geostrophic wind components (see also Figure A1). This procedure yields the 
effective wind direction as 142° (south-easterly wind). Now the daily geostrophic wind 
vectors for the entire period were projected onto the wind direction of 142° to the absolute 
effective wind, hereafter referred to as “v_eff”, comprising both wind direction and strength 
'nto a scalar metric for simple further analysis. We would like to emphasize that the 
effective wind calculated here is specific to ELWs at Cuxhaven. This approach has already 
been used in several studies related to (positive) storm tides (see, e.g., [45,46]). 
2.2.5. Hydrodynamic-Numerical Simulation 
In order to investigate how ELWSs in the Elbe estuary will be influenced by future 
SLR, the extreme ELW chain-event from 2018 is simulated in combination with several SLR 
scenarios using a hydrodynamic-numerical model. 
For this study, the three-dimensional hydrodynamic numerical model UnTRIM2 [47] 
is used, which solves the three-dimensional shallow water equations on an orthogonal 
anstructured grid. By using the subgrid technology described by [47] the model bathymetry 
can be discretized with a much finer resolution than the computational grid, which allows 
a better description of wetting and drying on the large intertidal areas in the German 
Wadden Sea. The effect of wind forcing is implemented in the model according to [48]. 
The generation of wind waves is not included in the model. We assume, that the effect 
of wind waves on the water level during NSS (wave-setup and wave-setdown) can be 
neglected because of the offshore directed wind and wave propagation and due to the 
limited fetch length and wind speed over the Elbe estuary. The model domain covers 
the German Bight from the island Terschelling in the Netherlands to Hvide Sande in 
Denmark including the German estuaries Elbe, Weser and Ems with their main tributaries 
(see Figure 1b). The topography data of the year 2016 implemented into the model was 
generated in the EasyGSH-DB project [49]. The most recent measurements additionally 
taken into account in the Elbe estuary are from 2019, taken by the Federal Waterways and 
Shipping Administration (WSV). 
The ELW chain-event 2018 is simulated using surface wind speed (10 m height) and 
surface pressure of COSMO-REAG6 [50] data as meteorological forcing over the model do- 
main. The data is generated and made available by the Hans Ertel Center of the University 
of Bonn in cooperation with the German Weather Service (DWD) [51]. Since a comparison 
of reanalysis data with observational data at Helgoland and Cuxhaven shows that wind 
speeds are underestimated during the storm we performed a simple bias correction by 
increasing the reanalysis wind speeds by 10% for the simulation. An underestimation of 
wind speed in the reanalysis data, especially in the mouth of the Elbe estuary, is most likely 
caused by the resolution, which is 6 km in our study area. For COSMO-EU (resolution 
of 7 km) [52] calculated correction factors for the Elbe estuary in the range of 1.1 to 1.5. 
An evaluation of the performance of COSMO-REA6 wind speed data over the North Sea 
can be found in [53]. The discharge into the Elbe estuary required for the hydrodynamic- 
numerical simulation of the ELW event 2018, is derived from measured discharge data