pers. comm.), others are puzzling. While laboratory experiments over uniform
slopes suggest wave breaking on ‘critical’ bottom slopes that match the angle of
internal wave propagation [20, 21], wave-breaking in the ocean is also observed
away from critical slopes of the major internal wave frequency, mostly tides, e.g.
[22]. Examples of the latter wave breaking is also observed to occur in areas where
near-horizontal inertial motions are the dominant carrier wave like in the
Mediterranean, or sub-inertial motions like in the Faeroe-Shetland Channel [23], or
trapped waves like along the Rockall Bank. This suggests a sloshing motion
importance for the generation of nonlinear waves leading to breaking [19].
Even if the precise process leading to nonlinear wave formation is unknown, or
not known for every particular topographic region, the importance of internal wave
breaking upon topography associates with the importance of ‘boundary mixing’ for
the maintenance of ocean stratification, as suggested by Munk [24] and Munk and
Wunsch [3]. Assuming a layer of 100 m of enhanced turbulence over topography in
an ocean of average depth of 3,000 m, Armi [25] suggested a turbulent diffusivity
of K z = 3 × 10
−3 m
2 s
−1 sufficient to maintain the necessary heat flux in the ocean
interior. This was debated by Garrett [26] as a boundary layer in the classic sense of
Ekman [27], even over sloping bottoms [28], is homogeneous and thus very
inefficient in its mixing, unless a complex restratification mechanism is invoked.
The debate was based on steady flows having time scales much longer than the
inertial period, or actually t ≫ 1/f. Microstructure profiler observations by, e.g.,
[29, 30] to within 0.3 m from the bottom above sloping topography in shallow seas
and [31] in the deep ocean showed that the turbulent flux (∝turbulence dissipation
rate) does increase towards the bottom and, indirectly, evidences effective mixing or
a mechanism to restratify the sloping ‘bottom boundary layer’. Detailed moored
temperature sensor measurements (e.g., [32]) indicate that internal wave breaking
dominates bottom friction in turbulence generation over sloping topography. It is
also the key mechanism in rapidly restratifying the near-bottom area, by transporting the mixed waters into the interior and by pushing the stratification to within
a meter or so from the bottom on time scales of the order of the inertial time scale,
or shorter. They alternate with sudden upslope propagating frontal bores [23, 33]
reaching up to 100 m above and actually touching the bottom thereby creating large
sediment resuspension resembling a desert dust storm. In essence, a bottom
boundary layer is not observed over sloping topography, at least not in the classic
sense.
A challenge to this view is the requirement of a near-bottom reduction of the
heat flux in order to have an upslope motion close to the bottom balancing the
interior downward turbulent heat transport [34]. Whilst the concept is well-posed, it
remains to be established how close to the bottom the heat flux should be reduced
for the mechanism to work. Perhaps, the adoption of the theorem of Prandtl that the
length-scales of the largest turbulent overturns can never be larger than the distance
to the bottom in a mixing length theory may not be adequate for flows with more
than one characteristic velocity [35]. Prandtl did not consider wave breaking over
sloping topography, but frictional shear-driven turbulent overturning over a solid
boundary.
High-Resolution Observations of Internal Wave Turbulence …
129
slopes suggest wave breaking on ‘critical’ bottom slopes that match the angle of
internal wave propagation [20, 21], wave-breaking in the ocean is also observed
away from critical slopes of the major internal wave frequency, mostly tides, e.g.
[22]. Examples of the latter wave breaking is also observed to occur in areas where
near-horizontal inertial motions are the dominant carrier wave like in the
Mediterranean, or sub-inertial motions like in the Faeroe-Shetland Channel [23], or
trapped waves like along the Rockall Bank. This suggests a sloshing motion
importance for the generation of nonlinear waves leading to breaking [19].
Even if the precise process leading to nonlinear wave formation is unknown, or
not known for every particular topographic region, the importance of internal wave
breaking upon topography associates with the importance of ‘boundary mixing’ for
the maintenance of ocean stratification, as suggested by Munk [24] and Munk and
Wunsch [3]. Assuming a layer of 100 m of enhanced turbulence over topography in
an ocean of average depth of 3,000 m, Armi [25] suggested a turbulent diffusivity
of K z = 3 × 10
−3 m
2 s
−1 sufficient to maintain the necessary heat flux in the ocean
interior. This was debated by Garrett [26] as a boundary layer in the classic sense of
Ekman [27], even over sloping bottoms [28], is homogeneous and thus very
inefficient in its mixing, unless a complex restratification mechanism is invoked.
The debate was based on steady flows having time scales much longer than the
inertial period, or actually t ≫ 1/f. Microstructure profiler observations by, e.g.,
[29, 30] to within 0.3 m from the bottom above sloping topography in shallow seas
and [31] in the deep ocean showed that the turbulent flux (∝turbulence dissipation
rate) does increase towards the bottom and, indirectly, evidences effective mixing or
a mechanism to restratify the sloping ‘bottom boundary layer’. Detailed moored
temperature sensor measurements (e.g., [32]) indicate that internal wave breaking
dominates bottom friction in turbulence generation over sloping topography. It is
also the key mechanism in rapidly restratifying the near-bottom area, by transporting the mixed waters into the interior and by pushing the stratification to within
a meter or so from the bottom on time scales of the order of the inertial time scale,
or shorter. They alternate with sudden upslope propagating frontal bores [23, 33]
reaching up to 100 m above and actually touching the bottom thereby creating large
sediment resuspension resembling a desert dust storm. In essence, a bottom
boundary layer is not observed over sloping topography, at least not in the classic
sense.
A challenge to this view is the requirement of a near-bottom reduction of the
heat flux in order to have an upslope motion close to the bottom balancing the
interior downward turbulent heat transport [34]. Whilst the concept is well-posed, it
remains to be established how close to the bottom the heat flux should be reduced
for the mechanism to work. Perhaps, the adoption of the theorem of Prandtl that the
length-scales of the largest turbulent overturns can never be larger than the distance
to the bottom in a mixing length theory may not be adequate for flows with more
than one characteristic velocity [35]. Prandtl did not consider wave breaking over
sloping topography, but frictional shear-driven turbulent overturning over a solid
boundary.
High-Resolution Observations of Internal Wave Turbulence …
129