Assimilation de données dans un modèle d'écosystème marin de la mer Ligure

The objective is to explore the potentialities of sequential statistical estimation methods to assimilate observations in a primary production biological model coupled to a vertical 1D hydrodynamical model characterised by a k-l turbulent closure. The assimilation method is derived from the SEEK filter (Singular Evolutive Extended Kalman filter), which uses an error subspace represented by multivariate empirical orthogonal functions (EOFs). Real data assimilation experiments collected at sea have been realised to reconstruct the variability of the Ligurian Sea ecosystem during the FRONTAL field experiment.