348
CiUNNAR KULLENBERG
consisting of acetic acid, methanol, dye, and heat. The differences in molecular diffusivities are present. The corresponding stability ratios (Turner, 1965)
are calculated for some cases and found to be of the order 10’. The rnaximum thickness of the convecting layers between successive interfaces
estimated by the relation of Stern and Turner (1969) is at least two orders of
magnitude less than the observed step thickness. Also Stem’s (1960) criteria
imply that salt-fingers cannot develop in the present conditions. These considerations show that doubkdiflusive phenomena can be ruled out. It is
furthermore evident that the natural density variations in the water are
much larger than the perturbation introduced by the tracer solution, at least
after an initial period.
Possible layer-generating mechanisms are then advective processes, overturning instabilities and wave induced shear instabilities.
(1) Advective processes may generate a small-scale structure of
interleafing water types reflected in the temperature-salinity distributions.
There are always more or less pronounced natural horizontal inhomogeneities present in the water body which will cause advective motion possibly
generating intermediate to small-scale vertical: variations. A sfrong enough
horizontal variation of temperature or salinity can give rise to horizontal
layered convection as pointed out by Stern (1%7) and demonstrated for
instance by Wirtz et 01. (1972). It is, however, not likely that advective
processes are the direct cause of microstructure (Gregg and Cox. 1972).
(2) Overturning instabilities as described by Orlanski and Bryan (1969)
are possible when the density distribution shows maxima or minima. An
examination of the Lake Ontario records shows that the requirement is at
hand in some cases, and the pictures suggest that overtuning instabilities
can occur.
Orlanski and Bryan neglect rotational effects, which, however, are
included in Frankignoul’s (1972) analysis of the stability of finite amplitude
waves. He demonstrates the importance of the gravitational instability
showing also that with rotation it will develop before the shear instability,
except for waves of small amplitude.
Considering the records it is not likely that this is the dominating process
in the present observations.
(3) The wave induced shear instability is believed to be of central importance. The model described by Woods and Wiley (1972) can explain the
occurrence of interleafing structures (Figs. 6 and 7).
Recently Garrett and Munk (1972a) proposed a universal small amplitude
internal wave spectrum which they applied to the mixing problem (Garrett
and Munk. 1972b). In particular they consider the inertial frequency. They
find that for this case, the wave-induced shear instability is much more likely
to occur than overturning. This is especially the case in the presence of
CiUNNAR KULLENBERG
consisting of acetic acid, methanol, dye, and heat. The differences in molecular diffusivities are present. The corresponding stability ratios (Turner, 1965)
are calculated for some cases and found to be of the order 10’. The rnaximum thickness of the convecting layers between successive interfaces
estimated by the relation of Stern and Turner (1969) is at least two orders of
magnitude less than the observed step thickness. Also Stem’s (1960) criteria
imply that salt-fingers cannot develop in the present conditions. These considerations show that doubkdiflusive phenomena can be ruled out. It is
furthermore evident that the natural density variations in the water are
much larger than the perturbation introduced by the tracer solution, at least
after an initial period.
Possible layer-generating mechanisms are then advective processes, overturning instabilities and wave induced shear instabilities.
(1) Advective processes may generate a small-scale structure of
interleafing water types reflected in the temperature-salinity distributions.
There are always more or less pronounced natural horizontal inhomogeneities present in the water body which will cause advective motion possibly
generating intermediate to small-scale vertical: variations. A sfrong enough
horizontal variation of temperature or salinity can give rise to horizontal
layered convection as pointed out by Stern (1%7) and demonstrated for
instance by Wirtz et 01. (1972). It is, however, not likely that advective
processes are the direct cause of microstructure (Gregg and Cox. 1972).
(2) Overturning instabilities as described by Orlanski and Bryan (1969)
are possible when the density distribution shows maxima or minima. An
examination of the Lake Ontario records shows that the requirement is at
hand in some cases, and the pictures suggest that overtuning instabilities
can occur.
Orlanski and Bryan neglect rotational effects, which, however, are
included in Frankignoul’s (1972) analysis of the stability of finite amplitude
waves. He demonstrates the importance of the gravitational instability
showing also that with rotation it will develop before the shear instability,
except for waves of small amplitude.
Considering the records it is not likely that this is the dominating process
in the present observations.
(3) The wave induced shear instability is believed to be of central importance. The model described by Woods and Wiley (1972) can explain the
occurrence of interleafing structures (Figs. 6 and 7).
Recently Garrett and Munk (1972a) proposed a universal small amplitude
internal wave spectrum which they applied to the mixing problem (Garrett
and Munk. 1972b). In particular they consider the inertial frequency. They
find that for this case, the wave-induced shear instability is much more likely
to occur than overturning. This is especially the case in the presence of