Search

The sensitivity of the Baltic Sea salinity to the freshwater supply is investigated using a 3-dimensional (3D) coupled sea-ice-ocean model. Today's climate is characterized by an average salinity of about 7.4 parts per thousand. and a freshwater supply, including river runoff and net precipitation, of about 16 000 m(3) s(-1). As recent results of some regional climate models have suggested a significant increase in precipitation in the Baltic catchment area due to anthropogenic climate change, in this study the response of salinity in the Baltic Sea to changing freshwater inflow is investigated. Of special interest is the possibility of the Baltic Sea becoming a freshwater sea with 0 parts per thousand salinity in the future. Therefore, model simulations with modified river runoff and precipitation for 1902-1998 were performed. The model is forced with daily sea-level observations in the Kattegat, monthly basin-wide discharge data, and reconstructed atmospheric surface data. The reconstruction utilizes a statistical model to calculate daily sea-level pressure, and monthly surface-air temperature, dew-point temperature, precipitation, and cloud-cover fields. It is assumed that the Kattegat deepwater salinity of about 33 parts per thousand. will not change regardless of the changed freshwater supply. In most of the experiments the final stratification is almost in a steady state after 100 yr. We found that even for a freshwater supply increased by 100% compared to 1902-1998 the Baltic Sea cannot be classified as a freshwater sea. A pronounced halocline still separates the upper and lower layers in the Baltic Proper, limiting the impact of direct wind mixing to the surface layer. A calculated phase diagram suggests that the relationship between freshwater supply and average salinity of the final steady state is non-linear. The results of the 3D model are in agreement with an analytical steady-state model assumed to work for freshwater changes smaller than 30 %. The latter model was applied in scenarios for the average salinity of the Baltic Sea.

Two models of the North Sea are compared with respect to their performance in the assessment of sea level changes in a changing climate: the OPYC model, originally designed as a global ocean model and run in a regional version, and the HAMSOM, which was built as a shelf sea model and applied to an extended region in a simplified version. Both models agree very well in their hindcast skill as measured by correlation and explained variance. The reaction of both models to the '2 〈 CO2' and 'control' runs of the time slice experiments of the DKRZ (German Climate Computing Center) is also very similar and indicates an increase of up to 30 cm of winter mean sea levels and no significant (compared with the variability in the past) impact on the intra-monthly 90% quantiles. It is concluded that sea level is a rather robust parameter that could still be assessed with some confidence if both models were simplified further.