The MODE/POLYMODE experiments conclusively established that the ocean has an energetic mesoscale, which is a major contributor to the northward transport of heat and the mixing of salinity and other conserved properties. The early process models for the mesoscale emphasized long-lived isolated vortices. Flierl (Ann. Rev. Fluid Mech., 1987) and Olsen (Ann. Rev. Earth Planet. Sci, 1991), present comprehensive reviews of these models. The ultimate examples of long-lived vortices are modons (Larichev and Reznik, Rep. USSR Acad. Sci., 1976). Modons are exact solutions to the nonlinear Euler equations with both wave-like and vortex-like properties. All isolated vortex models share a common characteristic: a substantial amount of core fluid trapped by bounding surfaces of potential vorticity. Climatologically forced general circulation models also are consistent with the notion of long-lived mesoscale vortices containing a core of trapped fluid. See for example Kantha et al (in Circ in Gulf of Mexico: Observations and Models, AGU Press 2005). In contrast, we have studied mesoscale eddies in a variety of data assimilating models and have found that all exhibit considerable exchange between the eddy cores and the ambient fluid outside the eddies. A detailed blob calculation for one ring, Fourchon, from the Gulf of Mexico illustrates the leaky property. This calculation illustrates the fundamental dilemma: does data assimilation corrupt mesoscale processes, or does it delineate new physics? Both possibilities are discussed.
This is a joint work with H. S. Huntley and B. L. Lipphardt, Jr.