Without turbulence and the mixing it causes, we would not have the same ocean that we do now, nor indeed the same climate. Turbulent mixing brings nutrients into the surface layer from below so that plankton can grow. Turbulence near the surface, driven by surface winds and cooling, transmits heat in and out of the ocean to create the reservoir of heat that governs climate. It is turbulence that diffuses the permanent pycnocline separating the cold bottom waters of polar origin from the atmosphere-connected upper ocean, either directly, or indirectly by local mixing followed by distribution along isopycnals (surfaces of constant density). Turbulence in the bottom layer affects the deposition, resuspension and movement of sediments. Turbulence creates micro-environments for the small creatures that form the basis for life in the oceans.
Turbulence is the irregular, random component of fluid motion. Its spatial scales are usually the smallest scales of the flow. The turbulence we shall discuss is the 0.01 to 10-meter overturning motions that result in vertical transport and mixing. The complex nature of the flow makes deterministic analyses impossible and kinematic observations generally unsatisfying; our conceptual framework for understanding turbulence is largely statistical. Turbulent motions in a stratified fluid intermingle fluid parcels from different parts of the flow, thereby enhancing mixing across concentration gradients by molecular diffusion. Oceanographers want to know the end result: the rate at which the ocean is mixed. But to understand the mixing, we have to deal with the messiness of turbulence. Further, since turbulent motions are rapidly dissipated in the absence of an energy source, it is probably rare that we observe steady-state turbulence. This makes it critical that we identify and quantify the sources of turbulence. The goals are clear. The issues are as complex as the motion.
Understanding turbulence and mixing in the ocean is important for many reasons. Most importantly perhaps, ocean models that can predict global circulation, climate change, pollutant dispersal and primary productivity, will provide reliable predictions only when we have the capability to quantify the subgrid-scale effects of turbulence.
In this report we will attempt to cover some of the significant developments of the past eight years. (No report on this subject was included 4 years ago.) Because this report must cover more time than usual, we are forced to neglect some important subjects (and the bibliography is not as complete as we would like).
Recent reviews of topics relevant to turbulence and mixing have dealt with ocean turbulence [ Gargett, 1989], air-sea interactions [ Donelan, 1990], mixing in stratified fluids [ Fernando, 1991], atmospheric turbulence [ Wyngaard, 1992], mixing [ Ottino, 1990], the history of microstructure work [ Gregg, 1991], Arctic processes [ Mc-Phee, 1990], boundary mixing [ Garrett et al., 1994], ocean mixing [ Garrett, 1994], salt fingering [ Schmitt, 1994], and applications of the Kolmogoroff spectrum to the ocean [ Phillips, 1991].
In this review, we will cover progress in some areas that have seen important advances in the past eight years. These include mixing in the near-surface region, in fronts, in the main thermocline, and near boundaries. Issues concerning the nature of turbulence will be considered: how can it be parameterized in terms of larger-scale quantities? What are its statistics? Is the assumption of isotropy on small scales justified? How can we estimate vertical fluxes? We will also say a bit about the very interesting subject of the effect of turbulence on small organisms.
We will neglect some important subjects: We will not deal with double diffusive phenomena (effects resulting from the difference between the diffusivity of heat and that of salt) because a review has been published recently [ Schmitt, 1994]. Although potentially important to mixing, coherent structures in the mixed layer, such as Langmuir circulations (roll structures aligned along the wind [ Weller and Price, 1988]) and roll vortices aligned across the wind [ Soloviev, 1990], are not discussed here. Also bottom layer studies must be neglected; a recent review of cross-shelf transport may be helpful [ Nittrouer and Wright, 1994]. Results from equatorial mixing experiments can be accessed via Moum et al. [1992].