An assessment of the Geophysical Fluid Dynamics Laboratory ocean model with coarse resolution: Annual‐mean climatology

Publisher Website
Google Scholar
Abstract
The Geophysical Fluid Dynamics Laboratory Modular Ocean Model 2.2 code with coarse resolution (4° × 3°) is assessed by performing three experiments and comparing their equilibrated solutions with recent observationally based analyses (OBAs). The first experiment (E1) uses subgrid‐scale horizontal diffusion and surface boundary conditions which relax surface temperature and salinity toward observations. The second (E2) replaces the physically incorrect heat and moisture flux boundary conditions of E1 by flux conditions taken from OBAs, plus a term relaxing surface temperatures toward observations. The third (E3) uses the same surface boundary conditions as E2 but replaces the horizontal diffusion by the Gent‐McWilliams (GM) parameterization of isopycnal diffusion. Under the restoring surface boundary conditions (E1), the North Atlantic overturning rate is about 17 Sv, smaller than in OBAs, the maximum poleward heat transport in the Northern Hemisphere is 1.2 Petawatts (PW), also smaller than in OBAs, and in the Antarctic Circumpolar Current (ACC) region the poleward heat transport is 1.3 PW, much larger than in OBAs. Under the more realistic flux boundary condition (E2) the overturning rate increases to an unrealistically large level of 40 Sv, and the poleward heat transports are only slightly improved. When the GM parameterization is employed (E3), the overturning is reduced to 28 Sv, and the poleward heat transport in the ACC region is reduced to 0.3 PW; both results are consistent with OBAs. However, there is only a slight further improvement in the poleward heat transport in the Northern Hemisphere, which now has a peak value of 1.6 PW, still about 0.5 PW less than in OBAs. The sea surface temperature errors in E3 are consistent with the conclusion that the heat transport in the Northern Hemisphere is still being underestimated. All the experiments show strong systematic biases in the salinity field.