Comments on global models of the oceanic general circulation

Abstract

So far most of the numerical models of the world ocean are not very successful in simulating the meridional heat transport which is one of the key factors in climate, because the calculated transport is too small. Table I shows the maximum of the annual mean of the northward heat transport across a parallel in the northern hemisphere by three numerical studies together with results based on observations. Sellers' (1965) estimate (A) is based on Budyko's surface climatological data. The other two (B) and (C) are obtained from satellite radiation data and atmospheric and oceanic data over the northern hemisphere. Model (ii) is a joint model of the ocean and atmosphere, where the boundary conditions at the ocean surface are not specified a priori but determined by the ocean--atmosphere interaction. Since no observational data are used for the external forcing driving the ocean circulation, the external forcing generated in the model could be far from the actual one and deteriorate the calculated circulation. However, the general aspect of the circulation is surprisingly realistic, though the northward heat transport is a little too small. The result of model (i), which is not yet published, is in good agreement with Sellers' estimate but still small compared with the other two