Orientational effects in geophysical fluid dynamics (Q795637)

The authors present an attempt to model wind-driven global ocean circulations. Numerical experiments are carried out for one and two layer models of the real world ocean basins with a five-degree horizontal resolution. The hydrodynamical modeling begins with the Boussinesq approximation of the Navier-Stokes equations for stratified, laminar, and incompressible fluid motions, which are assumed to govern also turbulent motions. Using results of earlier work [the fourth author, \textit{D. Sh. Iskenderov} and \textit{E. N. Koszhof}, Trudy III Vsesoysnogo Seminara po Modelyam Sploshnoy Sredy, Novosibirsk, 36-42 (1976)], the equations of motion are volume averaged over cells of grid mesh size, which enclose quasi-two-dimensional synoptic (mesoscale) eddies. Since these subgrid circulations are not resolved by the large-scale model, their regular and orientational effects on the averaged flow variables is modeled by an asymmetric Reynolds stress tensor consisting of additive symmetric and antisymmetric parts. Using Boussinesq's approximation, the symmetric Reynolds stress tensor is replaced by the laminar strain-rate tensor of the averaged flow with lateral and horizontal eddy viscosities replacing the molecular viscosity. The antisymmetric Reynolds stress tensor is assumed to be linearly related to the surplus angular velocity of the unresolved subscale eddies with an unknown angular turbulent viscosity coefficient. The angular velocity is defined by averaged angular momentum equations as derived earlier by one of the authors. This equation introduces three additional unknown parameterization constants. The computed results display the anticipated increased westward intensification of subtropical gyre transports as observed, for instance, in the Gulf Stream and Kuroshio.