Ocean heat transport and its relationship to ocean circulation in the CMIP coupled models

The Coupled Model Intercomparison Project was set up to examine climate variability and predictability as simulated by coupled atmosphere–ocean general circulation models, and to evaluate the model results against available observations. This study considers the results from a subproject that concerns the ocean heat transport and its relationship to ocean circulation in these coupled models. In particular, the relative roles of the meridional overturning circulation and horizontal circulation in determining the ocean heat transport are assessed and contrasted between the major ocean basins. Many consistent patterns are found among the models and with observations. A number of discrepancies and uncertainties are also identified. Some of the discrepancies have already been seen in ocean only models, and many of the uncertainties are also present in observational studies. In the Atlantic, the meridional overturning circulation that consists of the southward transport of the North Atlantic Deep Water and the northward transport of the upper layer water masses is captured by all the models, and so is the associated northward heat transport, although quantitatively there are large differences. In all the models, the North Atlantic Deep Water is too warm compared with that derived from direct oceanographic measurements, an indication that high-latitude processes responsible for the production and transport of the North Atlantic Deep Water are not well represented. In the Pacific, most of the models produced the upper layer wind-driven circulation. However, none of the models produced a deep circulation pattern close to that deduced from direct oceanographic measurements and consequently the heat transport exhibits large differences between the models. In the Indian Ocean, all models agree on a southward heat transport resulting from a meridional overturning circulation of warm southward flow of upper layer water and cold northward flow of deep water, but disagree on the magnitude of the heat transport and the strength of the overturning. Observational studies agree and disagree in a similar way. All the models produced poleward heat transport by the horizontal circulation in the subtropical regions of the South Atlantic and the Pacific Ocean basins. In the southern subtropical Indian Ocean a number of the models produced equatorward heat transport by the horizontal circulation, contrary to the common belief of a warm western boundary current. In the North Atlantic, a negligible contribution from the horizontal circulation is expected but several models produced significant equatorward heat transport.