
Fig. 1
Map of the Baltic and North seas. (a) Five tide gauge stations for long-term sea level observations. Their results have been used in this study. (b) The map of Danish straits (the Oresund, the Great Belt and the Little Belt) shows with red marks the sea level sites used.
Table 1. Hourly sea level datasets characteristic
1Cuxhaven53.878.72Germany1917–19872Göteborg57.6811.79Sweden1967–20063Stockholm59.3218.08Sweden1889–20124Gorniy Institute59.9330.28Russia1977–20065Wladyslawowo54.8018.42Poland1992–2006

Fig. 2
Sea level records in Stockholm (a) annual mean sea level data, dashed line shows a linear trend; (b) monthly mean sea level data, dashed line shows climatologic seasonal cycle; (c) daily mean sea level, dashed line shows climatologic seasonal cycle; (d) hourly sea level, dashed line shows the predicted tide. Data are referenced to 0-BSH+500 cm.

Fig. 3
Spectra of the sea level variations f·S(f): (a) in Göteborg (the Kattegat, Sweden), and (b) in Stockholm (the Baltic Sea, Sweden). Dashed lines show frequencies of tidal harmonics: a 18.6-yr nodal one (Mn), a month one (Mm), half-month ones (Msf, Mf), and seasonal ones (Sa, Ssa – annual and semi-annual) and a 14-month pole tide (P14).

Fig. 4
Wavelet diagrams of the sea level variations in Stockholm. Dashed lines show pole tide (P14).

Fig. 5
Synchronous daily mean sea level records in Cuxhaven (the North Sea) and Stockholm (the Baltic Sea). Data are referenced to 0-BSH+500 cm.

Fig. 6
Cross-spectral characteristics calculated for four pairs of synchronous sea level records: Cuxhaven – Stockholm (a), Göteborg – Stockholm (b), Cuxhaven – Gorniy (c), Stockholm – Wladyslawowo (d). Graphs of coherence are given in the upper panel; dashed line shows the 95% confidence interval. The central panel presents graphs of the normalised frequency magnitude response function. The red line traces the approximation of the response changes by analytical relationship [eq. (7)]. A frequency phase response function is given in the bottom panel. The red line shows the approximation of the phase changes by analytical relationship [eq. (8)].

Fig. 7
Geometry of the problem: A is the surface of the bay, L is the channel length and its width is W.
