Mean dynamic topography: Inter-comparisons and errors

Knowledge of the height of the sea surface above the geoid would allow one to determine the associated geostrophic surface currents and help to quantify subsurface currents and heat transports that regulate the Earth's climate. Such a determination is presently hampered by our inability to accurately measure the geoid, especially over the open ocean. In anticipation of a future mean dynamic topography (MDT) provided by combining altimetry and a GOCE geoid, this paper compares the MDTs of several ocean models in the important North Atlantic and Nordic Seas region (focus area of the EU GOCINA project). Each of the models used in this study assimilates hydrography, which is an essential requirement for a model to accurately interpolate the data and provide a realistic MDT. The available MDTs differ in resolution and time period; both these issues are addressed. A composite MDT is formed from the individual MDTs, and we assess the uncertainty in the composite MDT, and in the associated geostrophic surface currents. We find a remarkable level of agreement between this suite of model MDTs. The uncertainty in the composite MDT is less than 4cm over most of the study region, and the uncertainty in the associated geostrophic surface velocities is of the order 2cm/s. These uncertainties can be used as error estimates when comparing with satellite derived MDTs, to judge consistency.