The sea conditions (water level and SST) were represented by data from Port Pionerskiy, which is located at the open Baltic sea coast between the Vistula and Curonian Lagoons, and by SST measurements on the sea shore at Zingst, and Klaipėda. As historical data has been used, both the current and the historical names of the locations are given in the plot legends and tables: Klaipėda/Memel, Baltiysk/Pillau,

Krasnoflotskoye/Rosenberg, Nida/Nidden, Pionerskiy/Neukuhren. We analysed the variations in the annual mean water level without specifically revealing their eustatic and isostatic components, for the periods of 1840–2008 for Baltiysk/Pillau, 1898–2008 for Klaipėda/Memel, 1937–2008 for Zingst, and 1961–2008 for all the other points. It is remarkable

that all the lagoons lie on the periphery of the Fennoscandian land uplift, and that all had the same rate of land subsidence: Proteases inhibitor 0 mm year−1 (Ekman 2003, 2009) and –1 mm year−1 (Vestøl 2006). This information is taken into account in the Discussion and Conclusions. The rate of water level [mm year−1] and SST [°C year−1] changes at the various stations were evaluated using linear regression, which expresses unidirectional tendencies (trends) of water level and temperature changes over time. To eliminate irregular fluctuations in the illustrations of longterm trends, yearly mean values were smoothed by using the 11-year moving average (band width). The information on the quality of the regression was assessed by the R2 determination coefficient, which gives the square of the correlation coefficient, and by buy 17-AAG Student’s t-test. As enough the atmospheric conditions in the Baltic region were driven by the inflow of air masses from the west, the annual mean water level changes in the CL, VL and DZBC were compared with values of the North Atlantic Oscillation index (NAO index). The NAO index is associated with changes in the oceanic and atmospheric heat flux towards Europe and changes

in the atmospheric moisture and oceanic freshwater fluxes (Hurrell 1995); it is therefore an important indicator of climate changes. We used the winter (December to March) NAO index based on the difference in normalized sea level pressure between Lisbon (Portugal) and Stykkisholmur/Reykjavik (Iceland) when analysing the relation between the sea level and NAO index variability. Positive trends in water level variations were found for the three lagoons (Figure 2), but the trend rate differs. Water levels in the CL and VL rose significantly by 18 cm in the period between 1961 and 2008 (Table 2), while in the DZBC the water level increase was three times less (by 6 cm). The maximum rate during 1961–2008 was ~ 4 mm year−1, recorded in the CL and the VL, and the minimum (approximately 1 mm year−1) was in the DZBC (Table 2).