waves and dissipation move this part of the mixing
parameterization problem to one of internal wave
prediction. Showing the most promise here is the
abyssal mixing topic where flow over irregular
bathymetry has been linked to the profile of diapycnal diffusivity via baroclinic tide and internal lee
wave generation, propagation and interaction
(Polzin, 1999, 2000). One of the weak links in this
modelling is the wave generation at the bottom;
work to date builds on models that assume smallamplitude bathymetry or neglect of non-linear
advection effects. Mid-ocean ridges, continental
slopes and seamounts (where much of the deepocean mixing is found) are certainly not smallamplitude structures. It also remains to sort out
the relative roles of bottom-generated and bottomreflected internal waves in heightened abyssal
mixing.
Over sizeable ocean areas, the thermohaline
stratification supports double diffusion and here
we also anticipate enhanced diapycnal fluxes.
Double-diffusive flux laws based on laboratory
work have been developed that predict the diapycnal heat and solute diffusivities in terms of the
thermohaline stratification (e.g. Schmitt, 1988;
Kelley, 1984, 1990), and initial efforts to explore
these in ocean circulation models have been made.
Yet to be made are definitive ocean observations
to verify these laws at low density ratio where the
fluxes are believed large to complement those
acquired in the North Atlantic Tracer Release
Experiment at moderate density ratio.
Last we come to mixing in and about the bottom boundary layer, a topic we admittedly gave
limited attention to here. Albeit often inefficient
with turbulence acting on previously homogenized
water, we are loath to discount completely the
bottom boundary layer from an account of oceanmixing processes. Often more problematic than
parameterizing physical bottom-boundary-layer
processes in ocean models is the elimination of
spurious mixing deriving from the model formulation. Beckmann (1998) gives a nice review of these
issues.
Although there has been headway developing
understanding of ocean stirring and mixing, fundamental questions remain. In what ways does
stirring and mixing (mesoscale, turbulent, doublediffusive) control the mean meridional overturning
circulation and its associated redistribution of heat
and fresh water? What is the interplay here
between mixing in the surface boundary layer and
that in the interior? Is there need for significant
mixing across thermocline-depth isopycnals beyond
that occurring in and about the time-varying surface mixed layer? If yes, is near-boundary mixing
above continental slopes and seamounts sufficient
in addition to the O(10
95 m
2 s
91 ) diffusivity supported by the background internal wave field?
How representative are the Brazil Basin deep mixing results to the global ocean? What does a global
abyssal circulation driven by spatially varying
mixing over rough bathymetry look like, and is it
consistent with the distribution of abyssal water
properties and the handful of transport estimates
that are available? Though daunting, these are
all approachable questions that define an active
research area today. Growth in understanding
ocean mixing and stirring will continue.
Acknowledgements
JMT wishes to acknowledge support from the US
National Science Foundation and Office of Naval
Research for his own research efforts on ocean
mixing and for the preparation of this chapter.
This article is a contribution to the CSIRO Climate Change Program.
5.2 Mixing and Stirring in the Ocean Interior
355
Toole and McDougall
parameterization problem to one of internal wave
prediction. Showing the most promise here is the
abyssal mixing topic where flow over irregular
bathymetry has been linked to the profile of diapycnal diffusivity via baroclinic tide and internal lee
wave generation, propagation and interaction
(Polzin, 1999, 2000). One of the weak links in this
modelling is the wave generation at the bottom;
work to date builds on models that assume smallamplitude bathymetry or neglect of non-linear
advection effects. Mid-ocean ridges, continental
slopes and seamounts (where much of the deepocean mixing is found) are certainly not smallamplitude structures. It also remains to sort out
the relative roles of bottom-generated and bottomreflected internal waves in heightened abyssal
mixing.
Over sizeable ocean areas, the thermohaline
stratification supports double diffusion and here
we also anticipate enhanced diapycnal fluxes.
Double-diffusive flux laws based on laboratory
work have been developed that predict the diapycnal heat and solute diffusivities in terms of the
thermohaline stratification (e.g. Schmitt, 1988;
Kelley, 1984, 1990), and initial efforts to explore
these in ocean circulation models have been made.
Yet to be made are definitive ocean observations
to verify these laws at low density ratio where the
fluxes are believed large to complement those
acquired in the North Atlantic Tracer Release
Experiment at moderate density ratio.
Last we come to mixing in and about the bottom boundary layer, a topic we admittedly gave
limited attention to here. Albeit often inefficient
with turbulence acting on previously homogenized
water, we are loath to discount completely the
bottom boundary layer from an account of oceanmixing processes. Often more problematic than
parameterizing physical bottom-boundary-layer
processes in ocean models is the elimination of
spurious mixing deriving from the model formulation. Beckmann (1998) gives a nice review of these
issues.
Although there has been headway developing
understanding of ocean stirring and mixing, fundamental questions remain. In what ways does
stirring and mixing (mesoscale, turbulent, doublediffusive) control the mean meridional overturning
circulation and its associated redistribution of heat
and fresh water? What is the interplay here
between mixing in the surface boundary layer and
that in the interior? Is there need for significant
mixing across thermocline-depth isopycnals beyond
that occurring in and about the time-varying surface mixed layer? If yes, is near-boundary mixing
above continental slopes and seamounts sufficient
in addition to the O(10
95 m
2 s
91 ) diffusivity supported by the background internal wave field?
How representative are the Brazil Basin deep mixing results to the global ocean? What does a global
abyssal circulation driven by spatially varying
mixing over rough bathymetry look like, and is it
consistent with the distribution of abyssal water
properties and the handful of transport estimates
that are available? Though daunting, these are
all approachable questions that define an active
research area today. Growth in understanding
ocean mixing and stirring will continue.
Acknowledgements
JMT wishes to acknowledge support from the US
National Science Foundation and Office of Naval
Research for his own research efforts on ocean
mixing and for the preparation of this chapter.
This article is a contribution to the CSIRO Climate Change Program.
5.2 Mixing and Stirring in the Ocean Interior
355
Toole and McDougall