The absolute sea level trend from May 1995 to May 2019 in the Baltic Sea is analyzed by means of a regional monthly gridded dataset based on a dedicated processing of satellite altimetry data. In addition, we evaluate the role of the North Atlantic Oscillation and the wind patterns in shaping differences in sea level trend and variability at a sub-basin scale. To compile the altimetry dataset, we use information collected in coastal areas and from leads within sea-ice. The dataset is validated by comparison with tide gauges and the available global gridded altimetry products. The agreement between trends computed from satellite altimetry and tide gauges improves by 9%. The rise in sea level is statistically significant in the entire region of study and higher in winter than in summer. A gradient of over 3 mm/yr in sea level rise is observed, with the north and east of the basin rising more than the south-west. Part of this gradient (about 1 mm/yr) is directly explained by a regression analysis of the wind contribution on the sea level time series. A sub-basin analysis comparing the northernmost part (Bay of Bothnia) with the south-west reveals that the differences in winter sea level anomalies are related to different phases of the North-Atlantic Oscillation (0.71 correlation coefficient). Sea level anomalies are higher in the Bay of Bothnia when winter wind forcing pushes waters through Ekman transport from the south-west toward east and north. The study also demonstrates the maturity of enhanced satellite altimetry products to support local sea level studies in areas characterized by complex coastlines or sea-ice coverage. The processing chain used in this study can be exported to other regions, in particular to test the applicability in regions affected by larger ocean tides.
|Journal||Frontiers in Marine Science|
|Publication status||Published - 2021|
Bibliographical noteFunding Information:
This study is a contribution to the ESA Baltic+ Sea Level project (ESA AO/1-9172/17/I-BG - BALTIC +, contract number 4000126590/19/I/BG). We use the python distribution eofs (Dawson, 2016) for computation of the EOF. We used the Hector software (Bos et al., 2013) to estimate the trend uncertainties from MLE and study the impact of different noise models, as described in section 2.3.1. We thank Samantha Royston for the useful discussion concerning noise models in time series analysis.
© Copyright © 2021 Passaro, Müller, Oelsmann, Rautiainen, Dettmering, Hart-Davis, Abulaitijiang, Andersen, Høyer, Madsen, Ringgaard, Särkkä, Scarrott, Schwatke, Seitz, Tuomi, Restano and Benveniste.
- Baltic Sea
- coastal altimetry
- North Atlantic Oscillation (NAO index)
- satellite altimetry
- sea level