A comparison of two methods for the assimilation of altimeter data into a shallow-water model

Two simple methods for assimilating ocean surface pressure data into a three-layer adiabatic primitive equation model are tested and compared using a twin-experiment approach. One of the methods is based on a recently presented direct insertion scheme using a quasi-geostrophic model. The method explicitly avoids making changes to the potential vorticity fields in the lower layers and the vertical current structure after assimilation is determined by this criterion. The modification required for a primitive equation framework are discussed and a comparison is made with the more familiar ‘nudging’ method of assimilation. We use root mean square errors to quantify the response of the model to the assimilation of a single complete surface pressure field and also to repeated or intermittent data available every 20 days for a period of 1 year. The two methods are almost equally effective in the short term, although the direct insertion scheme appears to be more effective in the longer term. It is suggested that this is because insertion maintains a geostrophic balance at all levels and thus avoids the generation of internal gravity waves which are needed when nudging is used to restore geostrophic currents in the deeper layers.