Premium
Simulated halocline variability in the Baltic Sea and its impact on hypoxia during 1961–2007
Author(s) -
Väli Germo,
Meier H. E. Markus,
Elken Jüri
Publication year - 2013
Publication title -
journal of geophysical research: oceans
Language(s) - English
Resource type - Journals
eISSN - 2169-9291
pISSN - 2169-9275
DOI - 10.1002/2013jc009192
Subject(s) - halocline , anoxic waters , oceanography , stratification (seeds) , geology , salinity , hypoxia (environmental) , wind stress , environmental science , oxygen , chemistry , seed dormancy , botany , germination , organic chemistry , dormancy , biology
Salinity and halocline depth variations in the Baltic Sea during 1961–2007 are studied using a three‐dimensional ocean circulation model. Significant interannual and interdecadal variations in the halocline depth are found, together with distinct periods characterized either by shallow (1970–1975) or deep halocline (1990–1995). The model simulation indicates that the mean top layer salinity in the Baltic Sea is mainly controlled by the accumulated river runoff, while the mean below halocline salinity in the Baltic proper (which comprises Bornholm and Gotland basins) is more dependent on the low‐pass filtered zonal wind stress, with cutoff period of 4 years, henceforth called the mean zonal wind stress. The halocline depth and stratification strength in the Baltic Sea are significantly affected by the mean zonal wind stress, while the impact of runoff is smaller. The ventilation of the halocline from bottom layers is stronger during the shallow and from surface layers during the deep halocline period. Due to changes in ventilation variations in halocline depth systematically affect bottom oxygen concentrations on seasonal and decadal, but not on interannual time scales. For instance, a deeper halocline reduces hypoxic (oxygen concentration in bottom water below 2 mL/L) and anoxic (anoxic conditions in bottom water) areas and increases the bottom oxygen concentrations in the Gulf of Finland but decreases them in the deeper parts of the Baltic proper. Model results suggest that due to undersampling during 1961–2007 mean hypoxic and anoxic areas calculated from observed profiles are underestimated by 41% and 43%, respectively.