Physical Oceanography 
System Nordsee 
27 
early summer of the following year. This longevity arises from the heat storage capac 
ity of the ocean and is supported by the lack of counteracting atmospheric forcing in 
winter (neutral NAO mode). The substantial increase in summer temperature, which 
may be related to a poleward shift of the Hadley Circulation (Hu and Fu 2007), has 
dominated the entire seasonal temperature cycle since 1997. This is reflected in a high 
correlation between annual (July - June) and summer temperatures, replacing earlier 
years' strong dependence of (regular) annual temperatures on winter temperatures. As 
to biological and ecological impacts, the described changes are relevant at least inso 
far as they render cause-effect chains of the type NAO -> temperature -> ... useless. 
Finally, it is worth noting that these changes have taken place concurrently with accel 
erated global warming (Trenberth et al. 2007). 
Strong ice winters in the German Bight in the second half of the 20th century have re 
curred quasi-periodically, at intervals of 7 - 9 years. The last winter of this type even 
tuated in 1996, during a prolonged phase of easterly circulation resulting from a per 
sistently negative state of the NAO. The absence of such winters is attributable to high 
ocean heat content at the onset of winter and insufficient cooling rates in the course of 
winter due to neutral NAO modes. The threshold temperature for the freezing of sea 
water is reached late in the season or not at all, with effective ice formation rendered 
impossible by rapidly increasing insolation from mid-February. With an accumulated 
areal ice volume of just 0.03 m, the winter of 2005 was virtually free of sea ice. Also 
this year's winter (2009), which would appear cold, only produced normal temperature 
conditions in the North Sea and German Bight as compared to climatological means 
of 1971 - 1993. Obviously, such temperatures are insufficient to rupture the current 
scenario. 
Salinity (p. 135 sqs.) 
In winter 2005, high-salinity Atlantic water extended as far south as the Dogger Bank. 
Bottom-water salinities > 35 occurred in 63 % of the North Sea area, which was un 
precedented since the beginning of the time series in 1999. The strong southward 
propagation of Atlantic water was attributable to persistently high salinities in the 
North-East Atlantic Ocean, major storm activity in December and January, and sub 
stantial volume transports across the northwestern boundary. Distinct vertical salinity 
stratification was observed in the northeastern part of the North Sea in August 2005. 
This whole sea area, which is influenced by the Baltic outflow, had mixed-layer salini 
ties < 34 and extended much farther south (57° N) and west (2° E) than in the preced 
ing summer. 
The time series of salinity in the >Fair Isle Current< in the northwestern inflow area of 
Atlantic water is characterised by 6- to 9-year quasi-cycles whose negative phases are 
associated with the >Great Salinity Anomalies« in the northern North Atlantic Ocean. 
In comparison with Gaussian processes, major deviations from the mean (anomalies) 
are much more common with such random variables that are dominated by cyclic vari 
ations. The idea of excessiveness or rarity associated with the concept of anomaly is 
inapplicable insofar as the (most probable) normal oscillatory state is not represented 
by the mean state but by strong deviations from it. 
A continuous phase of low water flow since April 2003 ended in January 2005. Total 
runoff of the river Elbe, at 21 km 3 /a, matched the long-term mean, and monthly devi 
ations from climatological values also were inconspicuously small. Clearly elevated 
spring and autumn salinities of Helgoland waters, that are typically - though not per