Non-local Ocean Mixing Model and a New Plume Model for Deep Convection

V. M. Canuto,Y. Cheng,A. M. Howard
DOI: https://doi.org/10.1016/j.ocemod.2006.07.003
IF: 3.293
2007-01-01
Ocean Modelling
Abstract:Turbulent fluxes can be represented by a diffusivity tensor, the symmetric part of which describes "turbulent diffusion" while the anti-symmetric part describes "advection". Diffusion is a local process in the sense that it depends only on the local gradients of the mean fields while advection is non-local for it is represented by an integral over all length scales (all eddies) that can "fit" from say the bottom of the physical domain to the where the fluxes are computed.In the ocean, there are two main regimes where non-local transport is important. One regime is where storms release a sudden burst of mechanical energy to the ocean surface that is then transported downward by energetic eddies that deepen the mixed layer. Even relatively simple non-local models yield results considerably more realistic than those of local models. The second regime is deep convection (DC) caused by loss of surface buoyancy, the description of which is required for a reliable assessment of water masses formation.At present, there is no reliable model for either of these non-local regimes individually or much less a formalism capable of accounting for both regimes simultaneously. The goal of this paper is to present a formalism that provides the expressions for the non-local fluxes for momentum, heat and salinity encompassing both cases. Since the resulting number of dynamic equations involves is however large, we work out two sub-models, one when only shear must be treated non-locally (e.g., when storms release mechanical energy) and one when only buoyancy is to be treated non-locally (the DC case).We employ the Reynolds Stress formalism in which non-locality is represented by the third-order moments which in turn depend on the fourth-order moments for which we employ a new model that has been tested against LES data, aircraft data and a full PBL simulation.For the DC case, we rewrite the non-local model in terms of Plumes since thus far the only non-local model used to treat oceanic DC has been the "plume model" of Morton, Taylor and Turner (MTT model). We show that the MTT model has two key limitations, (1) an important physical process such as the rate of entrainment cannot be determined by the model and remains an adjustable parameter and (2) MTT is purely advective and thus only applicable to the initial stages of DC but not to the whole process which is both advective and diffusive. The model we derive bypasses these limitations, is a generalization of the MTT model and is applicable to the entire development of deep convection. Published by Elsevier Ltd.
What problem does this paper attempt to address?