Variational data assimilation in ocean model: A simulation experiment

Abstract

Variational data assimilation approach using adjoint equations has been used in a reduced gravity model of the Northwestern Indian Ocean. The method minimises a cost function defining the misfit between the model and the data with the model equations acting as constraints. Several numerical experiments have been performed with simulated data generated from a control run of the model with analysed wind forcing data. In the case of perfect data, assimilation procedure is able to recover Sea Surface Elevation and current. The algorithm is found to have good potential and will be very much applicable for assimilation of spaceborne altimeter data.