Factors affecting heat transport in an ocean general circulation model
by Kamenkovich, I., J. Marotzke and P.H. Stone
J. of Physical Oceanography, 30(1): 175-194, 2000
A global ocean general circulation model with idealized geometry and coupled to a simple representation of atmospheric energy fluxes is used to investigate which physical factors determine meridional heat transport. A particular focus is on causes for the common underestimation of heat transport in ocean general circulation models. The model is also forced by idealized wind stress and moisture flux profiles.
The zonal average of surface heat flux is obtained from a simple radiation parameterization and the divergence of observed atmospheric heat transport. In addition, zonal mixing in the atmosphere is implied by the relaxation of the sea surface temperature (SST) to its zonal average. A finite relaxation timescale results in a substantial increase in the meridional mass overturning in the "Atlantic" basin compared to the case with "infinitely efficient" zonal atmospheric mixing, owing to the admittance of zonal SST gradients. However, heat transport changes only by a small amount. When atmospheric zonal mixing is changed to interbasin mixing, meridional heat transport increases significantly. Doubling the width of the Pacific basin leads to a large increase in the Pacific heat transport, induced by both the meridional overtuning and the horizontal gyre circulations.
If the horizontal viscosity is decreased and the zonal resolution is increased near the boundaries, the resulting larger speed of the western boundary currents causes a noticeable increase in the Atlantic basin's heat transport.
The introduction of the Gent-McWilliams eddy parameterization leads to a substantial decrease in the strength of the overturning circulation in the Atlantic basin, presumably because the overall amount of diapycnal mixing is reduced. However, the decrease in the heat transport is much smaller because the thermocline is sharper and the deep ocean colder, resulting in enhanced vertical temperature contrast. Apparent disagreements with and among previous results are explained through the different effects of diapycnal mixing in the North Atlantic and elsewhere in the model.
© 2000 American Meteorological Society
Full article available here: http://dx.doi.org/10.1175/1520-0485(2000)030<0175:FAHTIA>2.0.CO;2