continental slope, near 3°S, and then flowed across
the equator to about 2°N before separating from
the slope. Like similar ALACE trajectories shown
by Davis (1998b), these floats crossed the equator
without notable flow perturbations like one might
expect from the zonal flows generally felt to dominate equatorial flow system.
The largest regional float experiment was in the
South Atlantic as part of the Deep Basin Experiment and used RAFOS floats from several nations.
Most of these floats drifted at intermediate depths
(600–1100 m) but Hogg and Owens (1999) concentrated their seedings on two deeper levels of the
Basin. Their preliminary findings were based on
complete tracks for only about one-half of their
returning floats. They had launch and surfacing
positions for the floats that had not yet been
tracked and these produced a set of net displacement vectors over a mission length of 800 days.
Several features were clearly evident. First, the
Deep Western Boundary Current (DWBC) at
2000 m was observed advecting North Atlantic
Deep Water southward along the Brazilian continental rise down to about 20°S. Second, the
steadily southward flow of the DWBC is interrupted by the Vitória-Trindade Seamount Chain
deflecting all deep floats to the east with a westward return flow to the south of the seamount
chain. This seems to confirm an advective balance
of the DWBC tongue as argued by Zangenberg
and Siedler (1998) on potential-vorticity and massconservation grounds. Third, away from the
boundary, floats at 2500 m and within the Antarctic Bottom Water (4000 m) reveal a primarily
zonal pattern with astonishingly short space- and
long time-scales, as shown graphically in Fig. 4.5.7
(see Plate 4.5.7, p. 300) of this volume. Variancepreserving spectra from simultaneous deep Eulerian observations show energy to be concentrated
at periods shorter than 100 days for the meridional
component while spectra of zonal flow continue to
increase with time scale.
Meddies have been the subject of various float
studies, summarized by Käse and Zenk (1996),
with emphasis on Lagrangian behaviour. These
intermediate-depth anticyclonic eddies have
azimuthal speeds up to 30 cm s
91 and translation
speeds of 2.5–5 cm s
91
. They carry warm salty
water of Mediterranean origin deep into the North
Atlantic and are implicated in the formation of the
Mediterranean salt tongue. Indeed, the longevity
and long distance travelled by these floats raises
again Crease’s question of the relative importance
of mean flow and eddies in determining ocean
transport. Bower et al. (1997) examined the formation, near the Strait of Gibraltar, of these small
coherent eddies by seeding the Mediterranean
undercurrent with 49 RAFOS floats. They found
that Meddies are formed along the continental
slope south of Spain and west of Portugal at a rate
of 15–20 per year in a process that apparently
involves roll-up of the boundary layer along the
slope. Richardson et al. (2000) re-analysed all
RAFOS data from the Iberian and Canary Basins
in the early 1990s for evidence of Meddies. They
found approximately 29 coexisting eddies that
seem to decay primarily through collision with seamounts. Remarkably, it appears that in February
of 1994 (toward the end of the Bower et al. study)
about 50% of the Meddies in the northeastern
Atlantic were tagged by floats.
In an example of how regional experiments can
clarify processes lost to the global-scale coverage,
an international consortium mounted the Cape of
Good Hope Experiment (KAPEX) to examine interbasin communication at the tip of Africa. Using a
combination of hydrography, surface drifters and
RAFOS floats (Boebel et al., 1998), KAPEX focused
on the processes by which intermediate-depth water
is exchanged between the Indian and South Atlantic
Oceans. Analysis of the 1996–99 observations, plus
ALACE data described below, will provide new
insights into the complex flow and mixing patterns
in this region and address the hypothesis that there
is a significant transport from the Indian to Atlantic
Oceans carried by Agulhas eddies.
An extensive study with RAFOS floats with
tuned compressibility addressed the Lagrangian
behaviour of the North Atlantic Current (NAC)
using nearly 100 quasi-isopycnal RAFOS floats
(Rossby, 1996; Dutkiewicz et al., 1999; Zhang
et al., 1999a). The mean velocity of the NAC in
the Newfoundland Basin was seen to consist of a
sequence of large-amplitude meanders with wavelength near 400 km (not dissimilar to those in the
deeper flow shown in Fig. 3.2.8 – see Plate 3.2.8,
p. 172). While the NAC front was found to approximate a material stream-tube, eddy transport with
the surrounding waters was quantified and found
to be asymmetric, favouring transport from the
subpolar gyre to the subtropical side. Rossby et al.
(2000) have expanded their studies of the NAC
SECTION 3 NEW WAYS OF OBSERVING THE OCEAN
136
the equator to about 2°N before separating from
the slope. Like similar ALACE trajectories shown
by Davis (1998b), these floats crossed the equator
without notable flow perturbations like one might
expect from the zonal flows generally felt to dominate equatorial flow system.
The largest regional float experiment was in the
South Atlantic as part of the Deep Basin Experiment and used RAFOS floats from several nations.
Most of these floats drifted at intermediate depths
(600–1100 m) but Hogg and Owens (1999) concentrated their seedings on two deeper levels of the
Basin. Their preliminary findings were based on
complete tracks for only about one-half of their
returning floats. They had launch and surfacing
positions for the floats that had not yet been
tracked and these produced a set of net displacement vectors over a mission length of 800 days.
Several features were clearly evident. First, the
Deep Western Boundary Current (DWBC) at
2000 m was observed advecting North Atlantic
Deep Water southward along the Brazilian continental rise down to about 20°S. Second, the
steadily southward flow of the DWBC is interrupted by the Vitória-Trindade Seamount Chain
deflecting all deep floats to the east with a westward return flow to the south of the seamount
chain. This seems to confirm an advective balance
of the DWBC tongue as argued by Zangenberg
and Siedler (1998) on potential-vorticity and massconservation grounds. Third, away from the
boundary, floats at 2500 m and within the Antarctic Bottom Water (4000 m) reveal a primarily
zonal pattern with astonishingly short space- and
long time-scales, as shown graphically in Fig. 4.5.7
(see Plate 4.5.7, p. 300) of this volume. Variancepreserving spectra from simultaneous deep Eulerian observations show energy to be concentrated
at periods shorter than 100 days for the meridional
component while spectra of zonal flow continue to
increase with time scale.
Meddies have been the subject of various float
studies, summarized by Käse and Zenk (1996),
with emphasis on Lagrangian behaviour. These
intermediate-depth anticyclonic eddies have
azimuthal speeds up to 30 cm s
91 and translation
speeds of 2.5–5 cm s
91
. They carry warm salty
water of Mediterranean origin deep into the North
Atlantic and are implicated in the formation of the
Mediterranean salt tongue. Indeed, the longevity
and long distance travelled by these floats raises
again Crease’s question of the relative importance
of mean flow and eddies in determining ocean
transport. Bower et al. (1997) examined the formation, near the Strait of Gibraltar, of these small
coherent eddies by seeding the Mediterranean
undercurrent with 49 RAFOS floats. They found
that Meddies are formed along the continental
slope south of Spain and west of Portugal at a rate
of 15–20 per year in a process that apparently
involves roll-up of the boundary layer along the
slope. Richardson et al. (2000) re-analysed all
RAFOS data from the Iberian and Canary Basins
in the early 1990s for evidence of Meddies. They
found approximately 29 coexisting eddies that
seem to decay primarily through collision with seamounts. Remarkably, it appears that in February
of 1994 (toward the end of the Bower et al. study)
about 50% of the Meddies in the northeastern
Atlantic were tagged by floats.
In an example of how regional experiments can
clarify processes lost to the global-scale coverage,
an international consortium mounted the Cape of
Good Hope Experiment (KAPEX) to examine interbasin communication at the tip of Africa. Using a
combination of hydrography, surface drifters and
RAFOS floats (Boebel et al., 1998), KAPEX focused
on the processes by which intermediate-depth water
is exchanged between the Indian and South Atlantic
Oceans. Analysis of the 1996–99 observations, plus
ALACE data described below, will provide new
insights into the complex flow and mixing patterns
in this region and address the hypothesis that there
is a significant transport from the Indian to Atlantic
Oceans carried by Agulhas eddies.
An extensive study with RAFOS floats with
tuned compressibility addressed the Lagrangian
behaviour of the North Atlantic Current (NAC)
using nearly 100 quasi-isopycnal RAFOS floats
(Rossby, 1996; Dutkiewicz et al., 1999; Zhang
et al., 1999a). The mean velocity of the NAC in
the Newfoundland Basin was seen to consist of a
sequence of large-amplitude meanders with wavelength near 400 km (not dissimilar to those in the
deeper flow shown in Fig. 3.2.8 – see Plate 3.2.8,
p. 172). While the NAC front was found to approximate a material stream-tube, eddy transport with
the surrounding waters was quantified and found
to be asymmetric, favouring transport from the
subpolar gyre to the subtropical side. Rossby et al.
(2000) have expanded their studies of the NAC
SECTION 3 NEW WAYS OF OBSERVING THE OCEAN
136