178
L. R. M. Maas et al.
Figure 18 shows the diurnal surface (black vectors) and internal tides (coloured).
In this case, since amplitudes of the surface and internal tides lie in the same range,
the surface tide’s vector length corresponds its magnitude.
A remarkable property of these observations is that diurnal baroclinic tidal vertical displacements always seem stronger and sometimes much stronger than those
produced by the surface tide, see e.g. in Fig. 18 the amplitude increases with increasing height above the bottom at lmc5 and lmc5a for frequencies Q1 and P1. This
requires an amplification mechanism. This may be because internal tides, forced by
cross-isobath surface tidal motions near the shelf edges of the adjacent coasts are
focused near the center of the Mozambique Channel. At the shelf edge, its amplitude will be set by the local surface tide, which, due to a reduced shelf depth, will
be stronger than over the deep-sea, hence leading to stronger internal tides. Another
reason for elevated amplitude levels of internal tides may be that once internal tides
propagate beam-wise into the abyss, they reflect from sloping bottoms, at which point
they will often be focused, leading to amplification of their energy density.
It is also remarkable that nearly all diurnal internal tide vectors are co-aligned
with one another. This means they are in phase, with vectors increasing in magnitude
higher up in the water column. This suggests them to be propagating horizontally,
varying in magnitude along the vertical.
East Madagascar Coherent Internal Tides
For EM moorings we do not have collocated bottom and mid-depth instruments.
Therefore, estimates from mid-depth instruments might contain the signal of both
surface and internal tides. In Fig. 13, the harmonic vectors within the water column,
at EMC2, 4 and 5, however, dwarf those at EMC1 and 3, related to the surface tide at
the bottom. Evidently coherent internal tides are remarkably strong on the transect
East of Madagascar. Looking at the semidiurnal components (lowest two rows), curiously, tidal phases within the water column are generally in opposition to those of the
surface tide (as measured by BPRs). Even though moorings are up to 100 km apart,
for the M2-tide the phase relationship within the water column is very tight. Subtracting the surface M2 tide (measured at EMC3 or OTIS) by eye from the harmonic
vectors within the water column, leads to equivalent surface displacements due to
internal tides of O (50 cm). With the conversion employed in section “Mozambique
Channel Coherent Internal Tides” this would suggest isopycnal displacements of
O(500 m)! Such large magnitude is suspicious and should be treated with caution.
It is quite unexpected because of the weakness of the surface tide along the EM
transect and surroundings. Near sloping topography, surface and coherent internal
tides are not necessarily scale separated [22, 43]. This suggests the internal tide to
have a barotropic length scale, such as for a forced internal tide. For S2 there appear
some horizontal phase differences, suggesting the S2 coherent internal tide to have
a propagating character.
L. R. M. Maas et al.
Figure 18 shows the diurnal surface (black vectors) and internal tides (coloured).
In this case, since amplitudes of the surface and internal tides lie in the same range,
the surface tide’s vector length corresponds its magnitude.
A remarkable property of these observations is that diurnal baroclinic tidal vertical displacements always seem stronger and sometimes much stronger than those
produced by the surface tide, see e.g. in Fig. 18 the amplitude increases with increasing height above the bottom at lmc5 and lmc5a for frequencies Q1 and P1. This
requires an amplification mechanism. This may be because internal tides, forced by
cross-isobath surface tidal motions near the shelf edges of the adjacent coasts are
focused near the center of the Mozambique Channel. At the shelf edge, its amplitude will be set by the local surface tide, which, due to a reduced shelf depth, will
be stronger than over the deep-sea, hence leading to stronger internal tides. Another
reason for elevated amplitude levels of internal tides may be that once internal tides
propagate beam-wise into the abyss, they reflect from sloping bottoms, at which point
they will often be focused, leading to amplification of their energy density.
It is also remarkable that nearly all diurnal internal tide vectors are co-aligned
with one another. This means they are in phase, with vectors increasing in magnitude
higher up in the water column. This suggests them to be propagating horizontally,
varying in magnitude along the vertical.
East Madagascar Coherent Internal Tides
For EM moorings we do not have collocated bottom and mid-depth instruments.
Therefore, estimates from mid-depth instruments might contain the signal of both
surface and internal tides. In Fig. 13, the harmonic vectors within the water column,
at EMC2, 4 and 5, however, dwarf those at EMC1 and 3, related to the surface tide at
the bottom. Evidently coherent internal tides are remarkably strong on the transect
East of Madagascar. Looking at the semidiurnal components (lowest two rows), curiously, tidal phases within the water column are generally in opposition to those of the
surface tide (as measured by BPRs). Even though moorings are up to 100 km apart,
for the M2-tide the phase relationship within the water column is very tight. Subtracting the surface M2 tide (measured at EMC3 or OTIS) by eye from the harmonic
vectors within the water column, leads to equivalent surface displacements due to
internal tides of O (50 cm). With the conversion employed in section “Mozambique
Channel Coherent Internal Tides” this would suggest isopycnal displacements of
O(500 m)! Such large magnitude is suspicious and should be treated with caution.
It is quite unexpected because of the weakness of the surface tide along the EM
transect and surroundings. Near sloping topography, surface and coherent internal
tides are not necessarily scale separated [22, 43]. This suggests the internal tide to
have a barotropic length scale, such as for a forced internal tide. For S2 there appear
some horizontal phase differences, suggesting the S2 coherent internal tide to have
a propagating character.