THE SENSITIVITY OF AN EDDY-RESOLVING MODEL TO THE SURFACE THERMAL-BOUNDARY CONDITIONS

We compare the effect of using four different formulations for the sur face thermal boundary condition on a primitive equation eddy-resolving model. The first formulation is the conventional restoring boundary c ondition. This relaxes the surface temperature of the model to a speci fied ''restoring temperature'' on a timescale of 30 to 60 days. The se cond formulation calculates the surface heat flux interactively by cou pling the ocean model to a simple atmospheric model with an effective restoring time of several hundred days. The third formulation (Rahmsto rf and Willebrand, 1995) (RW0) is a simplified energy balance model wi thout atmospheric heat transport. The fourth formulation (Rahmstorf an d Willebrand, 1995) (RW1) is a linearized energy balance model with at mospheric heat transport parameterized as a diffusion term. The bigges t impact is on the vertical structure of the temperature variance. Und er the restoring condition the maxima in this variance always occur be neath the surface. Under the other three boundary conditions, maxima a re found at the surface and/or subsurface levels, depending on geograp hical location and in closer agreement with observations. There is als o an increase in the magnitude and eastward extension of both the eddy and mean kinetic energy at midlatitudes and in the subpolar gyre regi on with the use of less constraining surface boundary conditions. We s uggest that the use of a conventional restoring surface boundary condi tion acts to suppress mesoscale variability in eddy-resolving models. The northward heat transport is also increased by using the RW0 and RW 1 formulations. The main reason for the enhancement of eddy variabilit y and northward heat transport using the RW0 and RW1 surface boundary conditions is the release of the mean state.