Seasonal stratification and the spring bloom - Introduction to the Physical and Biological Oceanography of Shelf Seas - page 175

Figure 6.2 The action of (a) a
surface grid only, and
(b) surface and bottom grids,
on an initial linear density
profile in the experimental
tank. The initial density
profile is the dashed line; the
solid line is the density
profile some time after the
grids have been oscillating. U F
is an entrainment velocity of
fluid driven by the stirring.

(a)

r

U F

(b)

r

U F

bottom mixed layers separated by a sharp density interface (a pycnocline), as shown
in Fig. 6.2b . The interface develops at a point where the turbulent intensities from the
two grids are the same and upward and downward entrainment rates are equal; if the
two grids are identical and driven at the same rate, this balance will occur halfway
between them. If there is no further addition of buoyancy, mixing across this interface
slowly erodes the density contrast and the associated stability until mixing is complete
and the water column is vertically uniform. This is the first of two important conclu-
sions to draw from this type of experiment: turbulent mixing by stress at a boundary
will produce a homogeneous layer associated with that boundary, the thickness of
which is dependent on the strength of the turbulence. Measurement of the
entrainment velocity, the effective velocity at which fluid is being transferred across
the pycnocline by the stirring, provides further insight into how turbulence acts
on density gradients. Figure 6.3 shows the results from experiments carried out by
J. S. Turner using grid-generated turbulence in a laboratory tank.
Notice in Fig. 6.3 that the ability of turbulence to transfer fluid across the density
interface is significantly reduced as the stability and the Richardson number at
the interface increase. This is a second important result from such experiments,
which is fundamental to understanding how a water column responds to a buoyancy
supply. If buoyancy is added to the surface layer at a sufficiently high rate, it may
out-compete the reduction of density contrast by stirring and stratification may be
sustained or even enhanced. Indeed, the laboratory results show that once stratifica-
tion starts to increase, it acts to reduce the 'efficiency' of mixing and makes it easier
for stratification to develop further. We met the theoretical basis for these laboratory
results, encapsulated in the Richardson number,

in Section 4.4.1 . This positive