THE EVOLUTION OF FRONTAL-GEOSTROPHIC VORTICES IN A 2-LAYER OCEAN

The properties of large amplitude vortices are numerically investigated using a set of two-layer primitive equations. Numerical experiments are systematically conducted for the f and beta planes, quasigeostrophic (small amplitude) and frontal-geostrophic (large amplitude) dynamics, cyclone and anticyclone, and vortex sizes comparable to the radius of deformation and larger. In particular, the evolution and migration of frontal-geostrophic vortices are discussed with reference to the evolution equations in a two-layer ocean. It is shown that the anticyclonic vortex with large amplitude displacement evolves differently from the quasigeostrophic vortex and that it is extremely long-lived. The evolution of anticyclonic vortices, in which interfacial displacement is large and the size is greater than the radius of deformation, is different from that of cyclonic vortices. While the shapes of anticyclonic vortices change from circle to ellipse because of instability, the large amplitude effect of interfacial displacement restores the vortices' shape and causes them to vacillate based on me conservation of potential vorticity and angular momentum. On the other hand, while the shapes of cyclonic vortices change from circle to ellipse, the effect makes their shape more slenderly elliptical. The splitting phenomenon of anticyclonic vortices is also different from that with cyclonic vortices. The anticyclonic vortices are robust in their cores, although they divide their marginal part into several vortices: The cores of cyclonic vortices stretch slenderly and split several into vortices whose size is the radius of deformation. The structure of potential vorticity causes this different behavior between cyclonic and anticyclonic vortices due to the large amplitude of interfacial displacement in the vortices. An upper-ocean vortex generates a tripolar vortex in the lower ocean through baroclinic instability while changing its shape from circular to elliptical on the f plane. Generating a tripolar vortex in the lower ocean, the upper-ocean vortex, moreover, releases the available potential energy to the lower ocean as Rossby waves on the beta plane. The generated vortex in the lower ocean affects the migration of the upper-ocean vortex due to the dispersion of Rossby waves and nonlinear modal coupling. These results show that the two-layer vortex with large amplitude evolves differently from the reduced-gravity vortex. Finally, some arguments are presented to explain the observation of splitting cyclonic eddies and the robustness of large anticyclonic eddies.