Deep-Ocean Tides in the South-West Indian Ocean . . .
157
a low-frequency, deep and noticeable excursion while during the non-marked events
large-amplitude waves are of high-frequency.
In the EM data, blow-downs were removed by replacing downward displacements
in excess of 10 m by NaNs.
Whether large tidal expressions during blow-downs are artificial, resulting from
drag forces on buoys exerted by tidal currents superimposed on low-frequency, eddyrelated currents, or genuine, expressing internal tides trapped within eddies [2, 17],
is not clear and will not be settled here. We note, that HP-filtering also filters out
the slow (large-scale) atmospheric pressure contribution to the BPRs. But, as [34]
noticed, BPRs may still contain atmospheric tides that would not show up in altimetric measurements, and can be one of the causes for discrepancies between the
two.
Time Series of Pressure Measurements
Mozambique Channel
At moorings located along the MC transect in Fig. 2b, Fig. 4 shows time series of both
bottom as well as mid-water column pressure recorders. Panels a-h are from LOCO
4 (periods listed in Table 2), panel i from LOCO 5, and panels j-n from LOCO 6. We
combine measurements from different LOCO periods to account for the best spatial
and temporal coverage of instrumentation across the channel (both along the sea bottom and throughout the water column). The ‘string of pearls’, visible in all records,
represents the spring-neap tidal cycle, produced by superposition of the two dominating surface tidal components, M2 and S2. This supports the notion that amplitude
and phase of this tidal interference pattern is about the same for each of these records.
Analysis of semidiurnal and diurnal tidal amplitudes and phases, discussed below,
will confirm this in more detail.
Most instruments higher up in the water column frequently suffer from large vertical excursions, sometimes reaching apparent displacements of thirty meter or more
(see Fig. 4, panels a–c and j–m). Interestingly, despite the long duration of these
events (typically lasting a week or longer) these excursions decrease in size with
depth. This is odd, as their long duration suggests them to be due to large scale
features that should, if these were of a barotropic nature as the surface tide, penetrate throughout the whole water column. The weakening of these vertical displacements with depth suggests these events are instead due to frequent blow-down of
buoys, caused by drag forces due to eddy and tide-related currents, that we aim to
eliminate. The buoys apparently often fail to keep their mooring lines from pointing
straight upwards. During previous observational periods in MC, tilt sensors indeed
revealed tilts of over 30
◦ that were moreover tidally modulated with another 5
◦ [23].
Even BPRs sometimes suffer from vertical displacements of 6 m or more that are
obviously not produced by surface elevations (see Fig. 4d).
157
a low-frequency, deep and noticeable excursion while during the non-marked events
large-amplitude waves are of high-frequency.
In the EM data, blow-downs were removed by replacing downward displacements
in excess of 10 m by NaNs.
Whether large tidal expressions during blow-downs are artificial, resulting from
drag forces on buoys exerted by tidal currents superimposed on low-frequency, eddyrelated currents, or genuine, expressing internal tides trapped within eddies [2, 17],
is not clear and will not be settled here. We note, that HP-filtering also filters out
the slow (large-scale) atmospheric pressure contribution to the BPRs. But, as [34]
noticed, BPRs may still contain atmospheric tides that would not show up in altimetric measurements, and can be one of the causes for discrepancies between the
two.
Time Series of Pressure Measurements
Mozambique Channel
At moorings located along the MC transect in Fig. 2b, Fig. 4 shows time series of both
bottom as well as mid-water column pressure recorders. Panels a-h are from LOCO
4 (periods listed in Table 2), panel i from LOCO 5, and panels j-n from LOCO 6. We
combine measurements from different LOCO periods to account for the best spatial
and temporal coverage of instrumentation across the channel (both along the sea bottom and throughout the water column). The ‘string of pearls’, visible in all records,
represents the spring-neap tidal cycle, produced by superposition of the two dominating surface tidal components, M2 and S2. This supports the notion that amplitude
and phase of this tidal interference pattern is about the same for each of these records.
Analysis of semidiurnal and diurnal tidal amplitudes and phases, discussed below,
will confirm this in more detail.
Most instruments higher up in the water column frequently suffer from large vertical excursions, sometimes reaching apparent displacements of thirty meter or more
(see Fig. 4, panels a–c and j–m). Interestingly, despite the long duration of these
events (typically lasting a week or longer) these excursions decrease in size with
depth. This is odd, as their long duration suggests them to be due to large scale
features that should, if these were of a barotropic nature as the surface tide, penetrate throughout the whole water column. The weakening of these vertical displacements with depth suggests these events are instead due to frequent blow-down of
buoys, caused by drag forces due to eddy and tide-related currents, that we aim to
eliminate. The buoys apparently often fail to keep their mooring lines from pointing
straight upwards. During previous observational periods in MC, tilt sensors indeed
revealed tilts of over 30
◦ that were moreover tidally modulated with another 5
◦ [23].
Even BPRs sometimes suffer from vertical displacements of 6 m or more that are
obviously not produced by surface elevations (see Fig. 4d).