turbulence. The analysis confirms the differences sketched above between upslope
cooling phase and downslope warming phase of the carrier wave. The former is
consistent with passive scalar statistics but for the bottom half of a 100 m tall
T-sensor array, while in the upper half also evidence is found of active scalar
statistics. Likewise, the downslope phase is not exclusively dominated by passive
scalar statistics. These observations confirm the notion that shear- and convection
turbulence co-exist in the ocean, at least above sloping topography. In general
strong intermittency is found during the downslope phase, while a clear inertial
subrange of −5/3 spectral slope is observed during the upslope phase in particular
in the lower 50 m above the bottom (cf., Fig. 6).
From extended statistics using nearly 500 sensors in a small-scale 3D mooring
array of five lines 100 m long and 4 m apart horizontally, the inertial subrange is
observed superposed by larger variability than expected from chi-squared statistics
[67]. The observations suggest localized patches that are more coherent than their
spectral surroundings (Fig. 6), and which are not found in open-ocean spectra. They
may be due to intermittency, to convection in a shear-dominated environment, or to
an interaction between turbulence and small-scale thin-layer internal wave motions:
The precise mechanism is not known as the pattern is not yet resolved, except that
their energy levels are found to be proportional to the tidal energy level as their
peaks fall along a −2 spectral slope to within the 95% statistical significance level.
Fig. 6 Temperature variance spectrum, an average of 442 near-raw periodograms for the Kaiser
window tapered time series of a four-day turbulent period observed by independent T-sensors at a
small-scale 3D (five line) mooring array above a super-critical slope of Mount Josephine. The data
are sub-sampled at 0.5 Hz for computational reasons. Besides inertial (f) and semidiurnal lunar
tidal (M 2 ) frequencies several buoyancy frequencies are indicated including four-day large-scale
mean N and the maximum small-scale N s,max . The number −5/3 indicates the spectral slope σ
−5/3
for frequency σ, −2 the spectral slope σ
−2
. The 95% significance level is indicated by the small
vertical bar and by the width of the two −2 slopes. After [67]
140
H. van Haren
cooling phase and downslope warming phase of the carrier wave. The former is
consistent with passive scalar statistics but for the bottom half of a 100 m tall
T-sensor array, while in the upper half also evidence is found of active scalar
statistics. Likewise, the downslope phase is not exclusively dominated by passive
scalar statistics. These observations confirm the notion that shear- and convection
turbulence co-exist in the ocean, at least above sloping topography. In general
strong intermittency is found during the downslope phase, while a clear inertial
subrange of −5/3 spectral slope is observed during the upslope phase in particular
in the lower 50 m above the bottom (cf., Fig. 6).
From extended statistics using nearly 500 sensors in a small-scale 3D mooring
array of five lines 100 m long and 4 m apart horizontally, the inertial subrange is
observed superposed by larger variability than expected from chi-squared statistics
[67]. The observations suggest localized patches that are more coherent than their
spectral surroundings (Fig. 6), and which are not found in open-ocean spectra. They
may be due to intermittency, to convection in a shear-dominated environment, or to
an interaction between turbulence and small-scale thin-layer internal wave motions:
The precise mechanism is not known as the pattern is not yet resolved, except that
their energy levels are found to be proportional to the tidal energy level as their
peaks fall along a −2 spectral slope to within the 95% statistical significance level.
Fig. 6 Temperature variance spectrum, an average of 442 near-raw periodograms for the Kaiser
window tapered time series of a four-day turbulent period observed by independent T-sensors at a
small-scale 3D (five line) mooring array above a super-critical slope of Mount Josephine. The data
are sub-sampled at 0.5 Hz for computational reasons. Besides inertial (f) and semidiurnal lunar
tidal (M 2 ) frequencies several buoyancy frequencies are indicated including four-day large-scale
mean N and the maximum small-scale N s,max . The number −5/3 indicates the spectral slope σ
−5/3
for frequency σ, −2 the spectral slope σ
−2
. The 95% significance level is indicated by the small
vertical bar and by the width of the two −2 slopes. After [67]
140
H. van Haren