ON THE MIDLATITUDE CIRCULATION IN A HIGH-RESOLUTION MODEL OF THE NORTH-ATLANTIC

This paper describes, and establishes the dynamical mechanisms respons ible for, the large-scale, time-mean, midlatitude circulation in a hig h-resolution model of the North Atlantic basin. The model solution is compared with recently proposed transport schemes and interpretations of the dynamical balances operating in the subtropical gyre. In partic ular, the question of the degree to which Sverdrup balance holds for t he subtropical gyre is addressed. At 25 degrees N, thermohaline-driven bottom flows cause strong local departures from the Sverdrup solution for the vertically integrated meridional mass transport, but these ne arly integrate to zero across the interior of the basin. In the northw estern region of the subtropical gyre, in the vicinity of the Gulf Str eam, higher-order dynamics become important, and linear vorticity dyna mics is unable to explain the model's vertically integrated transport. In the subpolar gyre, the model transport bears little resemblance to the Sverdrup prediction, and higher-order dynamics are important acro ss the entire longitudinal extent of the basin. The sensitivity of the model transport amplitudes, patterns, and dynamical balances are esti mated by examining the solutions under a range of parameter choices an d for four different wind stress forcing specifications. Taking into a ccount a deficit of 7-10 Sv (Sv = 10(6) m(3) s(-1)) in the contributio n of the model thermohaline circulation to the meridional transports a t 25 degrees N, the wind stress climatology of Isemer and Hasse appear s to yield too strong of a circulation, while that derived from the NC AR Community Climate Model yields too weak of a circulation. The Helle rman and Rosenstein and ECMWF climatologies result in wind-driven tran sports close to observational estimates at 25 degrees N. The range bet ween cases for the annual mean southward transport in the interior abo ve 1000 m is 14 Sv, which is 40%-70% of the mean transport itself. The re is little sensitivity to the model closure parameters at this latit ude. At 55 degrees N, in the subpolar gyre, there is little sensitivit y of the model solution to the choice of either closure parameters or wind climatology, despite large differences in the Sverdrup transports implied by the different wind stress datasets. Large year to year var iability of the meridional transport east of the Bahamas makes it diff icult to provide robust estimates of the sensitivity of the Antilles a nd deep western boundary current systems to forcing and parameter chan ges.