16 
Negative Surges in the Southern Baltic Sea 
3. Seasonal and long term 
variation 
Seasonal and annual frequency distributions and 
long-term variations in the occurrence of negative 
surges provide important information about this 
hydrological phenomenon. 
3.1. Interannual variation 
Fig. 3.1. a shows the measured minimum sea 
level in each calendar year, annual mean and 
median sea levels, and the values below which 
sea levels fell for 24, 72, and 240 consecutive 
hours, respectively. In general, sea level values 
during negative surges increase from west to 
east, with the exception of Sassnitz. The differ 
ence at Sassnitz as compared to the other 
gauges, which are located on a straight coast or 
in a bay, is its geographic location on a land spit. 
The mean sea level at all gauges has risen over 
time at a rate of about 15-17 cm per century (see 
Table 3.1. a), again with the exception of Sassnitz 
where the rate is only about 6-7 cm per century. 
This change observed at the tide gauges has 
several causes, the most important being global 
sea level rise and local land rise or subsidence. 
On a larger scale, the land rises in the north due 
to post-glacial rebound following melting of the 
ice shield, and subsides in the south. This differ 
ence in land rise may be an explanation for the 
lower sea level rise at Sassnitz as compared to 
the other gauges. Observed year-to-year varia 
tions of mean sea level are attributable mainly to 
longer-term wind conditions, which influence 
water exchange with the North Sea, and to fluc 
tuations of precipitation and riverine runoff. 
With mean sea level rising, it would appear logi 
cal to expect low sea level values to rise as well. 
The 240-hour values have in fact increased over 
time, the correlation coefficient with mean sea 
level changing from 0.57 in the west to 0.85 in 
the east. However, in the west, the correlation 
coefficient drops to low levels for the 72-hour 
(0.28) and 24-hour (0.03) values, and the annual 
minimum low sea level even shows a decrease 
for all stations except Kotobrzeg. This clearly 
shows that the direct causes of low sea levels 
are not the same as those leading to the ob 
served changes of mean sea level. Extremely low 
sea levels are caused by single, strong wind 
events, and a better correlation of values with the 
annual mean (calculated from 365*24 values) is 
only obtained by combining several values into 
an integral number (e.g. 72 hourly values to cal 
culate the 72-hour value). Negative correlation 
may also find an easy explanation: westerly winds 
continuing for an extended period of time push 
North Sea water into the Baltic Sea, thus increas 
ing annual mean sea level. But this also increases 
the possibility of strong south/southwesterly 
winds causing low sea levels in the western and 
southern Baltic to drop to extreme values. 
Table 3.1. a Statistical indicators for mean and low sea level: coefficient of linear regression and correlation with time 
series of annual MSL values 
Wismar 
Warnemünde 
Sassnitz 
Swinoujscie 
Kotobrzeg 
A 
B 
C 
A 
B 
C 
A 
B 
C 
A 
B 
C 
A 
B 
C 
MSL (in cm) 
503 
17,4 
504 
16,80 
505 
6,5 
502 
14,6 
504 
15,1 
Level which is 
undercut for more 
than the given 
hours in one year 
240 h 
456 
12,9 
0,57 
462 
10,20 
0,69 
469 
4,0 
0,83 
465 
9,9 
0,79 
469 
16,5 
0,85 
72 h 
436 
6,8 
0,28 
445 
5,60 
0,38 
455 
0,6 
0,58 
450 
1,6 
0,61 
458 
13,2 
0,75 
24 h 
415 
-6,4 
0,03 
429 
2,40 
0,09 
441 
-2,9 
0,32 
436 
-6,3 
0,38 
448 
14,7 
0,60 
Minimum sea level 
383 
-24,7 
-0,16 
400 
-0,12 
-0,12 
416 
-6,3 
0,04 
419 
-3,3 
0,04 
436 
23,2 
0,41 
A - Constant (for year 2000) from linear regression 
B - Trend (in 2000 year) for linear regression 
C - Coefficient of correlation with annual MSL