A biogeochemical forecasting system of the North and Baltic Seas is developed based on the HIROMB-BOOS circulation Model (HBM) coupled with the ERGOM ecosystem model and augmented by data assimilation (DA). The DA system is built within the Parallel Data Assimilation Framework (Nerger et al., 2005, Nerger and Hiller, 2013) and has been validated by the German Federal Maritime and Hydrographic Agency (BSH) for sea surface temperature assimilation into the operated BSHcmod with the Singular Evolutive Interpolated Kalman (SEIK) filter (Pham, 1998). The DA system is further extended by assimilating chlorophyll concentrations. In the frame of the ensemble based DA techniques- SEIK and a sequential Importance Resampling (SIR) filter,- we consider various aspects and strategies of the biogeochemical state and parameter estimation when assimilating MODIS satellite chlorophyll “a” and NOAA’s sea surface temperature observations. In particular, we identify crucial ecosystem parameters, investigate possible impacts of the assumed stoichiometry and scaling biogeochemical variables in the presence of non-Gaussianity on the forecasting system performance.