Atmosphere 2022, 13, 1634 
40o0f 21 
to NCEP and 20CR is likely to minimise the interference of mixed sea/land grid points. Fur- 
ther, ERA5 shows similar results regarding the occurrence of atmospheric circulation types 
as ERA40 and ERA-Interim (not shown). Differences (statistically insignificant) between 
both, the NCEP and ERA groups, can be seen for A and NW circulation, but are negligible 
for our investigations since negative storm surges occur almost exclusively during easterly 
situations. Therefore, the use of ERA5 for our investigations seems highly appropriate. 
2.1.3. Climate Model Simulations 
In order to analyse the potential future change of meteorological conditions favourable 
for ELWs in the Elbe estuary, we make use of the large ensemble produced by the Swedish 
Meteorological and Hydrological Institute (SMHI-LENS; [34]), employing the atmosphere- 
ocean general circulation model EC-Earth3 ([35]). EC-Earth3 includes a model component 
‚or the global atmosphere that is based on ECMWF’s IFS (cy36r4) as well as NEMO3.6 and 
LIM3 for the ocean and sea-ice, respectively. EC-Earth3 features a spectral resolution of 
I’L255 (approx. 80 km grid spacing) and 91 levels in the vertical for the atmosphere and 
the tripolar ORCA1-grid with a nominal resolution of 1°—with mesh refinement along the 
Equator—and 75 levels for the ocean and sea-ice components. 
SMHI-LENS comprises 50 simulations for each, the historical period 1970-2014 and 
different scenarios covering the period 2015-2100. Details about the initialisation and 
ensemble generation can be found in [34]. All simulations of a given experiment are 
subject to identical external forcings—such as greenhouse gases and aerosols—following 
the protocol for the historical-experiment of the Coupled Model Intercomparison Project in 
its sixth phase (CMIP6) [36] and ScenarioMIP [37], respectively. For the analyses presented 
in this study, we examined the four Tier1-scenarios of ScenarioMIP that is SSP1-2.6, SSP2- 
4.5, SSP3-7.0, and SSP5-8.5. These span a wide range of socio-economic narratives and 
greenhouse gas concentration pathways for the remainder of the 21st century. The large 
number of simulations available for SMHI-LENS make it an excellent tool to derive robust 
estimates of climate change signals projected by this model. While every single simulation 
is subject to natural variability superimposed on potential climate change signals, the joint 
analysis of the whole ensemble enables us to “average out” the internal fluctuations of the 
individual runs. 
Our analyses were based on daily mean sea-level pressure fields taken from 
chese simulations. 
2.2. Methods 
2.2.1. Definition Negative Storm Tide 
In contrast to wind-induced extreme high water levels, i.e., (positive) storm tides, there 
is no consistent definition for wind-induced extreme low water levels. We decided to define 
a threshold in this respect that yields a similar frequency of events as for the (positive) 
storm tides. For the German North Sea coast, (positive) storm tides are defined as water 
levels 1.5 m above the mean high water (mHW) [38]. Based on data from Gerber et al. [39], 
246 storm tides were recorded at the gauge Cuxhaven in the period 1949-2012, which 
corresponds to an average of about 4 storm tides per year. With an average of 706 tides per 
year this threshold translates into the 99.45th percentile of all high waters (HWs). For the 
definition of ELWs, we accordingly chose the 0.55th Percentile of all low waters (LWs) at 
Cuxhaven which corresponds to approx. 2.5 m below the mean low water (mLW). 
It should be emphasized here that these events are not necessarily independent of 
each other. A large-scale synoptic weather situation may lead to a succession of several 
LWSs that meet the criterion of an ELW given here. Such a series of ELWSs is referred to as a 
‘chain of ELWs” in the remainder of this manuscript. 
2.2.2. Classification of Large-Scale Atmospheric Circulation 
A coherent description of the large-scale atmospheric circulation is possible using 
defined circulation patterns or weather types (e.g., [40]). At the German Federal Maritime