Deep convection and deepwater formation: Progress and new directions



Focused interest in deep convection and deepwater formation was evidenced as early as 1972, when an international colloquium entirely dedicated to oceanic deepwater formation was convened by the Laboratoire d'Oceanographie du Museum in Paris. Discussions concentrated heavily, but not exclusively, on observations taken in the northwest Mediterranean Sea by the Medoc Group [1970], who proposed that processes leading to deepwater formation could be divided into three phases: preconditioning, violent mixing, and sinking and spreading. A new feature—the chimney—was identified, which tends to define the horizontal extent of an area where deep convection occurs to great depths and leads to deepwater formation.

In terms of mechanisms for destroying stratification, double diffusion was the prevalent choice at that time among most participants. However, the effect of advection was also considered. Stommel [1972] claimed that the deep convection observed in the Mediterranean Sea was of a nonpenetrative type, since the density of the mixed layer increased steadily with thickness and time instead of decreasing as for penetrative convection. Most of these conclusions were based on the fact that the problem was believed to be truly one dimensional. Stommel recognized that the small size of the sinking area was not related to a hydrodynamical phenomenon such as the one he described in 1962 [Stommel, 1962] nor even directly forced by the atmosphere, which acts at a much larger scale. Rather, he indicated that the small size of a deep convective area was preconditioned by the general circulation in the basin. The major discovery during the following 10-year period (1973–1983) was realization of the important role played by baroclinic instability, active during ALL phases including the initial preconditioning.