Interior mixing and phytoplankton survival in stratified environments - Introduction to the Physical and Biological Oceanography of Shelf Seas - page 213

the winds were light, the model indicates only slight dissipation in the surface
boundary which in this case was too thin to be observed by the profiler. The rates
of dissipation in the bottom boundary layer are reasonably well simulated, as is the
fall-off to low levels in the interior region of the water column. There is, however, a
conspicuous failure of the model to reproduce the small but important episodes of
dissipation in the interior which are apparent in the observations when e rises, for
short periods, by two orders of magnitude to 10 3.5 Wm 3 . This intermittent

dissipation is responsible for mixing properties into/out of pycnoclines. It is crucial
in relation to the supply of nutrients to the upper part of the water column to drive
primary production. It is also important when assessing the potential for directed
swimming that some phytoplankton are capable of, and for the export of phyto-
plankton from a pycnocline into deeper water. In the present form of the TC model,
this vital contribution of midwater mixing is missing since modelled e falls several
orders of magnitude below the observations. This interior turbulence and its associ-
ated diapycnal fluxes are missing from the model because the turbulence closure
scheme is relying solely on energy inputs due to boundary stresses at the seabed and
the sea surface. Other physical mechanisms can generate velocity shear, and hence
mixing, away from the boundaries; these mechanisms need to be either fully resolved
and simulated by the model, or at the very least parameterised within the model in a
way that leads to realistic mixing.

∼

Box 7.2 How turbulent are the shelf seas?

Rates of turbulent energy dissipation vary widely in shelf seas. In regions of strong
flows (

2ms 1 ) in shallow water, dissipation near the bottom boundary may exceed
1Wm 3 . In less energetic regions where the water column stratifies, as in Fig. 7.5a ,
there is a large contrast in dissipation of 2-3 orders of magnitude between the
tranquil pycnocline region and the boundary layers near the surface and bottom.
You can get an idea of the strengths of turbulent mixing in terms of something more
familiar by making a comparison with, say, the mixing generated by a typical kitchen
mixer. For instance, mean dissipation in the pycnocline may be as low as 10 5 Wm 3 .
This is roughly equivalent to two 100 W hand-held kitchen mixers operating in a
square kilometer of pycnocline 20 metres thick (see Problem 7.1).

>

Before we continue, let us briefly convince ourselves that this 'missing mixing' that
appears in our observations but not in our model of the interior of the water column
is real. Estimates of the diffusivity in the water column for the SWIS time series of
Fig. 7.5a using the Osborn relation ( Equation 7.11 ) and the observed values of e
and N 2 are shown in Fig. 7.6 . The average mixing rate corresponds to K z ∼ 10 5

m 2 s 1 . We might ask if this mixing rate through the pycnocline is sufficient to
account for the observed rate of temperature rise in the deep water of the stratified

3