Operational oceanography requires state of the art numerical modelling, efficient data assimilation methods and optimal use of observations. My main interest is to apply, test and port promising theoretical methods both from the data assimilation and modelling field to realistic applications of operational oceanography. Application to realistic applications often reveals the limitations of theoretical methods and motivate new theoretical challenges.

PhD achievements highlighted the capabilities and the limitations of a computationally cheap ensemble data assimilation method. Inspiration from the meteorology community led us to complement our forecast with an additional predictability index, and develop for realistic application a novel data assimilation method that makes use of this information to improve the prediction.

Postdoc efforts were directed towards the development of the new TOPAZ forecasting system used as core of the Arctic Marine Forecasting Center of the European MyOcean system. In addition, collaborative research with Chinese and American colleagues deals with the assimilation of altimetry data in a model that includes tides.

More recent efforts cover the problem of data assimilation into Earth system model for seasonal to decadal prediction within the new Center for Climate Dynamics in Bergen.