| ABSTRACT | The Adriatic and the Black Seas are two semi-enclosed seas both connected to the Mediterranean sea by the Otranto strait and the very narrow Bosphorus strait respectively. The purpose of this work is to investigate the main sources of sea level variability during the 20th century and the eventual presence of a common forcing acting on both the basins depending by dynamical mechanism active in the Mediterranean Sea. A seamless sea level time series is constructed using 7 tide gauge records from stations located along the Adriatic coast and 5 from the Black Sea coast from 1900 to 2009. Data have been collected and provided by PSMSL. The reconstruction procedure is carried out by a statistical tool based on PCA and Least Square Method. Satellite sea level anomaly data confirm that sea level time series reconstructed from coastal stations are reliable representation of the Adriatic and Black Sea level (correlations are 0.87 and 0.72 respectively). The sea level has been decomposed in various components in order to investigate the role of the factors responsible for its interannual variability. The sea level annual cycle has been compared with the annual cycle of the inverse barometer component, the effect of the wind and the steric component. For the Adriatic sea the major role is played by wind, the inverse barometer effect plays a minor role whilst steric effect results to be quite negligible. On the contrary, for the Black Sea, the main contribution to interannual variability is due to the Danube river, which is estimated from the available discharge data of Sulina (one of the exits of the Danube delta). Inverse barometer and steric effect play both a minor role. A linear regression model in which a SLP gradient, identified after a correlation analysis, is used as predictor, is capable to explain a further percentage (about 20-25%) of variability of the sea level after subtracting all the factors considered above. Finally, residual sea levels show a positive correlation (0.42 about) revealing the likely action of a common boundary forcing associated to the mass exchange with Mediterranean sea. Anyway the present analysis is still unable to explain a non-negligible fraction of interannual variability of sea level, in particular for Black Sea. This is likely to a substantial extent due to uncertainties of hydrographic data caused by their irregular distribution in space and time and on the lack of regular records of past river discharge. This study is part of the activities of RISES-AM project (FP7-EU-603396). |