Atmosphere 2022, 13, 1634 
12 of 21 
Cuxhaven (271 events below 0.55th percentile) 
Climatology 24h before negative storm surge 
435% 
90.9 %, 
Nu 
2% 
A% 
‚4% 
\ 
0.4 % 
74% 
48. 
Figure 10. Left: Mean distribution of gale classes determined from daily means of ERA5 sea-level 
pressure for the entire period from 1950 to 2019. Right: Mean distribution of gale classes 24 h before 
ELW at Cuxhaven (LW below 0.55th percentile of all LW) between 1950 and 2019. 
While a storm (G, SG and VSG) occurs on average with a probability of about 9%, it 
occurs with a probability of about 56.5% during the 24 h before an ELW. This significant 
difference (p < 0.001 tested via 10,000-fold bootstrapping) confirms the expectation that 
comparatively high wind speeds on a larger spatial scale are necessary to produce ELWs. 
3.4.4. Effective Wind 
Figure 11 shows the empirical probability distribution function of the effective wind 
(Section 2.2.4.) as a climatology of the daily mean for the entire period 1950-2019 and for 
the 24 h before ELWs. A clear distinction between the two distributions is obvious. Before 
ELWSs, the effective wind was always positive and in more than 90% of all cases higher than 
10 m/s which in turn approx. matches the 95th percentile of the climatological distribution. 
[he mean of the ELW-sample is 15.6 m/s, while the climatological probability distribution 
of the effective wind spreads from about —30 m/s to 20 m/s with a mean value of approx. 
2 m/s. The difference between these two mean values is statistically significant with 
pvp < 0.001 (based on 10,000-fold bootstrapping). 
12 
Effective wind at 142° 
Climatology 
24 h before negative storm surge 
0 
8 
6 
2 
A 
4 
, 
+ 
— 30 
—— LITT LT TTTT TAT 11T Ta 
—20 —10 0 10 20 
Velocity [m/s] 
2m um 0 m 
. — 
30 
Figure 11. Distribution of the effective wind at 142° for the whole time period 1950-2019 (climatology, 
blue) and for the 24 h before the respective 271 EI Ws (orange).