Merging of two-layer baroclinic vortices

Abstract

Studying the interaction of isolated vorticity regions is of primary importance for the understanding of turbulence in general and the behaviors of many geophysical situations in particular. In the context of shear flows it has received much attention from experimentalists (Winant and Browand, 1974, Brown and Roshko, 1974) as well as from theoretical and computational points of view (Zabusky and Deem, 1971, Christiansen, 1973, Christiansen and Zabusky, 1973). Isolated vortex structures have also been studied in detail theoretically and numerically (Kida, 1981, Overman and Zabusky, 1982). However it is only recently that progress has been made in the understanding as to why and how merging occurs. Axisymmetrization due to filament formation appears to be the chief merger mechanism (Melander, Zabusky and McWilliams, 1989). The problem of the interaction of baroclinic vortex structures is even more difficult. Point vortices in the simplest configuration, namely a two-layer stratified fluid, assuming geostrophy, have been recently investigated by Hogg and Stommel (1985a). In this case there is no merging but there is cluster formation formation (Hogg and Stommel, 1985b). Griffiths and Hopfinger (1987, hereafter GH) studied experimentally the interaction and merging conditions of real, finite core vortices in a two layer stratified rotating fluid. A strong dependance of merging conditions on stratification was observed. Contour dynamics calculations of the same problem by Polvani, Zabusky and Flierl (1989, hereafter PZF) indicated however that merging conditions are not altered by stratification except when the two layers were not of equal depth.