Advancements of operational oceanography in the Baltic Sea

Managing the sea environment is a complicated interdisciplinary task. To understand changes in the sea, knowledge of the present state is essential. Many variables are monitored constantly, and long historical data sets exist. However, the spatial and temporal data coverage varies widely over the Baltic Sea. The preparation for the emerging circumstances demands the ability to forecast the future marine conditions. Thus, improved modelling and forecasting systems are needed. In this thesis, methods were developed to 1) understand the present state of the sea and 2) predict future conditions. The study areas were the Bothnian Sea and the Eastern Gotland Basin. Argo floats are a common tool in the oceans, but so far they have not been used in shallow marginal seas, such as the Baltic Sea. The autonomous measurement device brings possibilities to fill the gaps in the existing observation network (e.g. research vessels, moorings) as well as to enable new scientific experiments. To better understand the present state of the Northern Baltic Sea, methods were developed using the Argo floats. The salinity, temperature and GPS data collected with these floats from the area is analysed in this thesis for the first time and its applicability for studying the different physical phenomena, such as currents at the float diving depth and wind induced mixing, are evaluated. The usability of Argo data was compared with the ship-borne CTD data. Due to the higher frequency of the Argo data, the seasonal variations can be studied in detail with this method. However, the spatial coverage of the Argo data is not as good as the CTD data collected with a research vessel due to the fact that the floats only operate near the deep areas of the Baltic Sea. To be able to predict the future conditions of the Baltic Sea, monthly ensemble forecasting system was developed. A 3D biogeochemical model was forced with monthly ensembles of the atmospheric forcing and the results were applied to forecast upwelling events and harmful algal blooms. The monthly ensemble forecasts for upwelling events were evaluated. The result was that the upwelling events could be forecasted on a weekly scale. This enables, for example, better planning of the scientific study of upwelling events or the improvement of local-scale weather forecasts. The same probability-based ensemble prediction system was used to produce harmful algal bloom forecasts. The forecasts showed the effects of the weather scenarios on marine biogeochemistry. In the future, it will be possible to interconnect the observations and forecasts better than before. The more dense observations can be used to improve the computational methods, for example, by assimilation. The probability-based forecasts can help, for example, to mitigate the environmental risks.