Assimilating High-resolution Sea Surface Temperature Data Improves the Ocean Forecast in the Baltic Sea
Abstract. We assess the impact of assimilating the satellite sea surface temperature (SST) data on the Baltic forecast, practically on the forecast of ocean variables related to SST. For this purpose, a multivariable data assimilation (DA) system has been developed based on a Nordic version of the Nucleus for European Modelling of the Ocean (NEMO-Nordic). We use a localized Singular Evolutive Interpolated Kalman (SEIK) filter to characterize correlation scales in the coastal regions. High resolution SST from OSISAF is assimilated to verify the performance of DA system. The assimilation run shows very stable improvements on the model simulation as compared with both independent and dependent observations. The SST prediction of NEMO-Nordic is significantly enhanced by the DA system. Temperatures are also closer to observation in the DA system than the model results in the water above 100 m in the Baltic Sea. In the deeper layers, salinity is also slightly improved. Besides, we find that Sea level anomaly (SLA) is improved with the SST assimilation. Comparison with independent tide gauge data show that overall root mean square error (RMSE) is reduced by 1.8 % and overall correlation coefficient is increased by 0.4 %. Moreover, the sea ice concentration forecast is improved considerably in the Baltic proper, the Gulf of Finland and the Bothnian Sea, respectively.