Simulating Baltic Sea climate for the period 1902-1998 with the Rossby Centre coupled ice-ocean model

Abstract

Hindcast simulations for the period 1992-1998 have been performed using a 3D coupled ice-ocean model for the Baltic Sea. Daily sea level obervations in Kattegat, monthly basin-wide discharge data, and reconstructed atmospheric surface data have been used to force the Baltic Sea model. The reconstruction utilizes a statistical model to calculate daily sea level pressure and monthly surface air temperature, dewpoint temperature, precipitation, and cloud cover fields on a 1o x 1o regular horizontal grid for the Baltic Sea Region. An improved turbilence scheme has been implemented into the Baltic Sea model to simulate saltwater inflows realisticly. The results are validated agaist avaliable observational datasets for sea level, salinity, saltwater inflow, volume transport, and sea ice. in addition, a comparison is performed with simulations for the period 1980-1993 using 3-hourly gridded atmospheric obervations from synoptic stations. It is shown that the reults of the Baltic Sea model forced with the reconstructed data are satisfactory. Sensitivity experiments have been performed to explore the impact of internal mixing, fresh- and saltwater inflows, sea ice, and the sea level in Kattegat on the salinity og the Baltic sea. It is found that the decadal variability of mean salinity is explained partly by decadal volume variations of the accumulated freshwater inflow from river runoff and net precipitation and partly by decadal variations of the large-scale sea level presure over scandinavia. During the last century two exceptionally long stagnation periods are found, the 1920s to the 1930s and the 1980s to the mid 1990s. During these periods precipitation, runoff and westerly winds were stronger than normal. Stronger westerly winds caused increased eastward surface-layer transports. Consequently, the mean eastward lower-layer transports through the Stolpe Channel is reduced. The response time scale of the Baltic Sea is of the order of 30-40 years. The large inter-annual variability of the freshwater supply has no significant impact on the decadal variability of the mean salinity. Also the impact of river regulation which changes the discharge seasonality is only minor. Finally, the response of the Baltic Sea salinity to extreme changes of the freshwater inflow is investigated. We found that even with a 100% increased freshwater supply compared to the period 1902-1998 the Baltic Sea cannot be classified as a lake. The relationship between freshwater supply and mean salinity of the final steady-state is non-linear.