Representation of convective plumes by vertical adjustment

Open-ocean deep-water formation involves the interplay of two dynamical processes; plumes (≤1 km wide), driven by “upright” convection, and geostrophic eddies (≥5 km wide), driven by baroclinic instability. Numerical “twin” experiments are used to address two questions about the plumes: Can they be represented by a simple mixing process in large-scale models? If so, is it important that the mixing occurs over a finite time tmix, or would instantaneous mixing produce the same effect on large-scale properties? In numerical simulations which resolve the geostrophic eddies, we represent the plumes with a “slow” convective adjustment algorithm which is broadly equivalent to an enhanced vertical diffusivity of density in statically unstable regions. The diffusivity κ depends on tmix, the mixing timescale. The fidelity of the plume parameterization is then evaluated by comparison with plume-resolving simulations of open-ocean deep convection. Integral properties of the plumes, such as the temperature census of the convected water and the strength of the rim current that encircles the convecting region, are all accurately reproduced by the slow adjustment scheme. The importance of choosing an appropriate finite value for tmix is explored by setting tmix = 12 hours in some experiments, in accordance with scaling considerations, and tmix = 0 in others, corresponding to instantaneous adjustment, the conventional assumption. In the case of convection into a moderately or strongly stratified ocean the behavior does not significantly depend on tmix. However, in neutral conditions the slow adjustment does improve the parametric representation. Our experiments confirm the picture of plumes homogenizing the water column over a time tmix.

[1]  F. Schott,et al.  Hydrographic Structure of the Convection Regime in the Gulf of Lions: Winter 1987 , 1991 .

[2]  B. M. Boubnov,et al.  Temperature and velocity field regimes of convective motions in a rotating plane fluid layer , 1990, Journal of Fluid Mechanics.

[3]  Kevin D. Leaman,et al.  Observations with Moored Acoustic Doppler Current Profilers in the Convection Regime in the Golfe du Lion , 1991 .

[4]  F. Schott,et al.  Observation of a deep convection regime with acoustic tomography , 1995 .

[5]  M. Visbeck,et al.  Open ocean deep convection, Mediterranean and Greenland Seas , 1994 .

[6]  H. Fernando,et al.  Effects of rotation on convective turbulence , 1991, Journal of Fluid Mechanics.

[7]  T. Yates,et al.  Laboratory and Numerical Experiments in Oceanic Convection , 1994 .

[8]  B. M. Boubnov,et al.  Experimental study of convective structures in rotating fluids , 1986, Journal of Fluid Mechanics.

[9]  John Marshall,et al.  Regimes and scaling laws for rotating deep convection in the ocean , 1995 .

[10]  E. Sarachik,et al.  An Efficient Convective Adjustment Scheme for Ocean General Circulation Models , 1994 .

[11]  Medoc Group Observation of Formation of Deep Water in the Mediterranean Sea, 1969 , 1970, Nature.

[12]  K. Bryan A Numerical Method for the Study of the Circulation of the World Ocean , 1997 .

[13]  J. Marshall,et al.  Integral Effects of Deep Convection , 1995 .

[14]  T. Maxworthy,et al.  Unsteady, Turbulent Convection into a Homogeneous, Rotating Fluid,with Oceanographic Applications , 1994 .

[15]  J. Marotzke Influence of Convective Adjustment on the Stability of the Thermohaline Circulation , 1991 .

[16]  D. Olbers,et al.  Numerical studies of open ocean deep convection , 1995 .

[17]  John Marshall,et al.  Convection with Rotation in a Neutral Ocean: A Study of Open-Ocean Deep Convection , 1993 .