A high resolution data assimilation system for the Aegean Sea hydrodynamics

• Published in: Journal of marine systems 2009-05, Vol.77 (3), p.325-340
• Main Authors: Korres, G., Nittis, K., Hoteit, I., Triantafyllou, G.
• Format: Article
• Language: English
• Subjects: Aegean Sea; Data assimilation; Ferrybox data; Marine; POM model; SEEK filter
• ISSN: 0924-7963; 1879-1573
• DOI: 10.1016/j.jmarsys.2007.12.014

We present the development and validation of an eddy-resolving sequential data assimilation system for the Aegean Sea hydrodynamics that has been developed as part of the Poseidon operational system. The assimilation scheme is based on the Singular Evolutive Extended Kalman (SEEK) filter which is an error subspace Extended Kalman filter that operates with low-rank error covariance matrices as a way to reduce the computational burden. The filter was used to correct the forecast state of a 1/20° Princeton Ocean Model (POM) of the Aegean Sea every time new observations were available. The model is forced with hourly fluxes from a 1/10° ETA regional atmospheric model and is one-way nested to a 1/10° POM model of the Eastern Mediterranean Sea. The assimilated data set is multivariate including weekly AVISO sea level anomalies, weekly AVHRR sea surface temperature, and daily Ferrybox sea surface salinity (SSS) along the ferry boat route from Piraeus to Heraklion. Data assimilation experiments are performed to validate the system over a 6-month period (January–June, 2004). In one of the assimilation experiments, where the Ferrybox SSS data were not assimilated and used as independent observations, the multivariate assimilation system was able to reduce the error on the SSS prediction. Additional assimilation of the Ferrybox data on a daily basis shows their positive impact on other predicted variables of the system and their significant local effect on the SSS prediction within the southern Aegean Sea.