Whatever, the 3D mooring reveals the transition from squashed stratified turbulence
having an aspect ratio of ≤ 0.5 near the large-scale buoyancy frequency to fully
developed turbulence of aspect ratio equal to 1 before coherence is lost around a
frequency of about twice the maximum (resolved) small-scale buoyancy frequency.
Discussion and Conclusions
Turbulence values are high in the lower 100 m above sloping topography (e.g., [8,
32]). They are 100–1000 times larger than observed in the open ocean, away from
boundaries (e.g., [52, 31). Considering the rapid restratification process by the
(sloshing or propagating) carrier waves, the turbulent mixing is found to be effective.
Thus, the interplay between the large-scale internal waves and the small-scale waves
near the buoyancy frequency acts to create both the diapycnal turbulent mixing and
the isopycnal transport of the mixed waters into the interior. The T-string observed
high mean turbulent diffusivity values O(10
−3
− 10
−2 ) m
2 s
−1 are confirmed by
various other Mid-Atlantic Ridge shipborne data (e.g., [68]) and Jim Ledwell’s
recent Mid-Atlantic Ridge tracer release experiments: These diffusivities are
(observed to be) required to balance a near-bottom upslope motion.
Following Armi’s [25] suggestion that about 3% of the ocean is occupied by
such 100 m tall layer of large turbulence, the high mean turbulent diffusivities
observed over sloping topography will yield an overall mean ocean-basin-interior
diffusivity of K z ≈ 1–3 × 10
−4 m
2 s
−1 , sufficient to maintain the ocean stratification [3, 24]. Above particular steep seamount slopes of Mount Josephine (NE
Atlantic Ocean) that are supercritical for internal tides, the local large-scale carrier
wave, high local diffusivity values suggest a relatively narrow range of only 450 m
depth interval sufficient to supply this overall ocean-basin-interior diffusivity if the
turbulence observed over these particular slopes is extrapolated to other ocean
topography [18]. This large turbulence above tidally supercritical slopes is confirmed by numerical modelling of Winters [19] and Sarkar (2016, pers. comm.). It
supports earlier conjectures (e.g., [24, 25]) for the importance of sloping boundary
mixing.
Previous and present data suggest that such mixing is quite universal, independent of forcing (carrier wave) mechanism (frequency), but strongly dependent
on slope with respect to that of the carrier wave. The turbulence above sloping
topography well (>300 m) below the summits of seamounts is determined by slope
steepness and nonlinear wave evolution, but not by bottom-friction, and not necessarily by ‘critical’ internal tide reflection, internal hydraulic jumps or lee-wave
generation. Lee-waves are estimated to carry about one-tenth of the energy of
internal tides, although this may vary for certain locations [69]. Tides are not the
only forcing mechanism, as turbulent overturns are also observed in seas like the
Baltic and the Mediterranean Sea where tides are weak. Bottom slopes need not be
‘critical’ for sediment resuspension as suggested for the occurrence of intermediate
nepheloid layers [70]: Vigorous bore-like motions have also been observed at a
High-Resolution Observations of Internal Wave Turbulence …
141
having an aspect ratio of ≤ 0.5 near the large-scale buoyancy frequency to fully
developed turbulence of aspect ratio equal to 1 before coherence is lost around a
frequency of about twice the maximum (resolved) small-scale buoyancy frequency.
Discussion and Conclusions
Turbulence values are high in the lower 100 m above sloping topography (e.g., [8,
32]). They are 100–1000 times larger than observed in the open ocean, away from
boundaries (e.g., [52, 31). Considering the rapid restratification process by the
(sloshing or propagating) carrier waves, the turbulent mixing is found to be effective.
Thus, the interplay between the large-scale internal waves and the small-scale waves
near the buoyancy frequency acts to create both the diapycnal turbulent mixing and
the isopycnal transport of the mixed waters into the interior. The T-string observed
high mean turbulent diffusivity values O(10
−3
− 10
−2 ) m
2 s
−1 are confirmed by
various other Mid-Atlantic Ridge shipborne data (e.g., [68]) and Jim Ledwell’s
recent Mid-Atlantic Ridge tracer release experiments: These diffusivities are
(observed to be) required to balance a near-bottom upslope motion.
Following Armi’s [25] suggestion that about 3% of the ocean is occupied by
such 100 m tall layer of large turbulence, the high mean turbulent diffusivities
observed over sloping topography will yield an overall mean ocean-basin-interior
diffusivity of K z ≈ 1–3 × 10
−4 m
2 s
−1 , sufficient to maintain the ocean stratification [3, 24]. Above particular steep seamount slopes of Mount Josephine (NE
Atlantic Ocean) that are supercritical for internal tides, the local large-scale carrier
wave, high local diffusivity values suggest a relatively narrow range of only 450 m
depth interval sufficient to supply this overall ocean-basin-interior diffusivity if the
turbulence observed over these particular slopes is extrapolated to other ocean
topography [18]. This large turbulence above tidally supercritical slopes is confirmed by numerical modelling of Winters [19] and Sarkar (2016, pers. comm.). It
supports earlier conjectures (e.g., [24, 25]) for the importance of sloping boundary
mixing.
Previous and present data suggest that such mixing is quite universal, independent of forcing (carrier wave) mechanism (frequency), but strongly dependent
on slope with respect to that of the carrier wave. The turbulence above sloping
topography well (>300 m) below the summits of seamounts is determined by slope
steepness and nonlinear wave evolution, but not by bottom-friction, and not necessarily by ‘critical’ internal tide reflection, internal hydraulic jumps or lee-wave
generation. Lee-waves are estimated to carry about one-tenth of the energy of
internal tides, although this may vary for certain locations [69]. Tides are not the
only forcing mechanism, as turbulent overturns are also observed in seas like the
Baltic and the Mediterranean Sea where tides are weak. Bottom slopes need not be
‘critical’ for sediment resuspension as suggested for the occurrence of intermediate
nepheloid layers [70]: Vigorous bore-like motions have also been observed at a
High-Resolution Observations of Internal Wave Turbulence …
141