Assimilating ocean tide determined data into global tidal models

Abstract

A data assimilation procedure, which has successfully been applied to fictive and realistic scenarios, is applied to a 1 °-model making use of an effective iterative method for the solution of the minimization problem. Two sets of ocean tide determined data are used for the purpose of assimilation, the more extensive one mainly comprising pelagic together with some coastal sea surface elevation data, and the other one consisting of loading gravity data. The computed O1 and M2 global tidal oscillation systems, namely the fields of tidal elevation and loading gravity, are compared with numerous additional pelagic and coastal elevation data and with a selected number of proper gravity data, respectively. The assimilation of the two sets of data leads to an enormous reduction of the errors of the model results in all oceans. Assimilating this altogether still restricted number of data, allows studying the generation of realistic tidal oscillation phenomena by individual data and comparing these data effects with those having been obtained by previous data assimilation experiments using a model with coarser grid spacing. The field of dynamical residuals resulting from data assimilation reflects the far reaching influence of the data, and it is shown that the spatially integrated work done by the residuals contributes to the in all reduced rate of dissipation in the tidal power balance.