Caribbean Sea is still the subject of some debate.
Schmitz and Richardson (1991) estimated, based
on earlier data, that on the order of 13 Sv of
southern hemisphere water was entrained into the
northern subtropical gyre above 1000 m. At least
half this transport can be traced through the
passages of the lesser Antilles (Wilson and Johns,
1997). Sections off French Guiana (Bourlès et al.,
1999a,b) indicate that in the boreal spring, part
of the NBC near the surface, and even in the subthermocline, continues along the continental slope.
The presence of a northwestward flow was also
found near 10°N in March 1996 (Schott et al.,
1998), but it is arguable whether this water is
really of equatorial origin. The seasonal poleward
continuation of the NBC along the slope has been
a feature consistently present in numerical simulations of the equatorial Atlantic (Philander and
Pacanowski, 1986; Schott and Böning, 1991;
Böning and Schott, 1993), although probably to a
larger extent than observed. There is also a flow
present most of the year on the Guiana shelf going
toward the northwest and carrying 3–5 Sv (Johns
et al., 1998; Bourles et al., personal communication). There are other means of transporting
southern water to the north: by retroflection
eddies spun off by the NBC in October–April, and
through the interior circulation (in particular,
via the NECC and around the dome of Guinea:
Siedler et al., 1992; Arhan et al., 1998). There is
also some indication of poleward flow along
Africa from the thermocline down to the AAIW
layer (Fratantoni and Richardson, 1999; Stramma
and Schott, 1999). However, the relative importance of these different pathways is not yet known,
nor understood.
The previous discussion has focused on the
northward flow of southern water, the direction
required to supply the Atlantic conveyor belt circulation. An interesting question is to what degree
northern water penetrates into the equatorial area.
Based on the water mass characteristics, this is not
a dominant effect. However, there is indication of
a penetration of northern water towards the equator in the thermocline. During boreal spring, some
surveys identified a remnant of the NECC west of
40°W, which is directly fed by thermocline water
of northern Atlantic origin (Bourles et al., 1999b).
There is clearly a recirculation of part of the North
Equatorial Current towards the NECC in all
seasons. In the thermocline and below, at 4°N
and further north, there is an indication of a subsurface equatorward flow (the Western Boundary
Undercurrent; Fig. 4.3.6) along the continental
slope, which is strongest in late boreal winter
and spring (Johns et al., 1990, 1998; Colin and
Bourles, 1994; Wilson et al., 1994; Bourlès et al.,
1999a). Numerical experiments (Schott and Böning,
1991) suggested that this could contribute to the
EUC, although the presence of this water has not
yet been identified there. This return current could
fit the circulation expected from Sverdrupian budgets (Mayer and Weisberg, 1993), but these budgets do not take into account the thermohaline
component of the flow and are not expected to
hold for the mean (although there is more evidence
for the annual variability).
4.3.3.3 Circulation below 1000 m
Transient tracers (for example CFCs) illustrate the
arrival to the equatorial area of recently formed
waters, in particular in the upper North Atlantic
Deep Water (uNADW near 1500–2000 m: Andrié
et al., 1998, 1999; Fig. 4.3.7). This shows a flow
path near the western boundary, with a tongue of
high concentration towards the east in the vicinity
of the equator. The simultaneous current measurements have indicated that this is associated with
particular current directions, both near the western
boundary and along the equator (usually, towards
the southern hemisphere and towards the east,
respectively). In particular, along 35°W, flows
associated with high CFCs are consistently eastward near 2–3°S in the uNADW (Fig. 4.3.8),
and near 1.5°S in the lNADW. There are also
occasional eastward zonal currents in the uNADW
along the equator, but with large differences from
cruise to cruise (Rhein et al., 1995; Fischer and
Schott, 1997). Similar to what is observed in the
Pacific Ocean, all the current profiles within 1° of
the equator indicate the presence of jets with a
short vertical scale of 400–600 m, in particular
between 1000 and 2500 m depth (Gouriou et al.,
1999). These jets are found to last many months
and have a large zonal coherence. The flow north
of the equator can also be quite variable, in particular away from the slope, as has been shown from
two repeat sections off French Guiana separated
by 45 days (Andrié et al., 1999).
The deep water flow along the continental slope
is steadier than is found in the interior. Moorings
near 44°W indicate a current trapped to within
SECTION 4 THE GLOBAL FLOW FIELD
230
Schmitz and Richardson (1991) estimated, based
on earlier data, that on the order of 13 Sv of
southern hemisphere water was entrained into the
northern subtropical gyre above 1000 m. At least
half this transport can be traced through the
passages of the lesser Antilles (Wilson and Johns,
1997). Sections off French Guiana (Bourlès et al.,
1999a,b) indicate that in the boreal spring, part
of the NBC near the surface, and even in the subthermocline, continues along the continental slope.
The presence of a northwestward flow was also
found near 10°N in March 1996 (Schott et al.,
1998), but it is arguable whether this water is
really of equatorial origin. The seasonal poleward
continuation of the NBC along the slope has been
a feature consistently present in numerical simulations of the equatorial Atlantic (Philander and
Pacanowski, 1986; Schott and Böning, 1991;
Böning and Schott, 1993), although probably to a
larger extent than observed. There is also a flow
present most of the year on the Guiana shelf going
toward the northwest and carrying 3–5 Sv (Johns
et al., 1998; Bourles et al., personal communication). There are other means of transporting
southern water to the north: by retroflection
eddies spun off by the NBC in October–April, and
through the interior circulation (in particular,
via the NECC and around the dome of Guinea:
Siedler et al., 1992; Arhan et al., 1998). There is
also some indication of poleward flow along
Africa from the thermocline down to the AAIW
layer (Fratantoni and Richardson, 1999; Stramma
and Schott, 1999). However, the relative importance of these different pathways is not yet known,
nor understood.
The previous discussion has focused on the
northward flow of southern water, the direction
required to supply the Atlantic conveyor belt circulation. An interesting question is to what degree
northern water penetrates into the equatorial area.
Based on the water mass characteristics, this is not
a dominant effect. However, there is indication of
a penetration of northern water towards the equator in the thermocline. During boreal spring, some
surveys identified a remnant of the NECC west of
40°W, which is directly fed by thermocline water
of northern Atlantic origin (Bourles et al., 1999b).
There is clearly a recirculation of part of the North
Equatorial Current towards the NECC in all
seasons. In the thermocline and below, at 4°N
and further north, there is an indication of a subsurface equatorward flow (the Western Boundary
Undercurrent; Fig. 4.3.6) along the continental
slope, which is strongest in late boreal winter
and spring (Johns et al., 1990, 1998; Colin and
Bourles, 1994; Wilson et al., 1994; Bourlès et al.,
1999a). Numerical experiments (Schott and Böning,
1991) suggested that this could contribute to the
EUC, although the presence of this water has not
yet been identified there. This return current could
fit the circulation expected from Sverdrupian budgets (Mayer and Weisberg, 1993), but these budgets do not take into account the thermohaline
component of the flow and are not expected to
hold for the mean (although there is more evidence
for the annual variability).
4.3.3.3 Circulation below 1000 m
Transient tracers (for example CFCs) illustrate the
arrival to the equatorial area of recently formed
waters, in particular in the upper North Atlantic
Deep Water (uNADW near 1500–2000 m: Andrié
et al., 1998, 1999; Fig. 4.3.7). This shows a flow
path near the western boundary, with a tongue of
high concentration towards the east in the vicinity
of the equator. The simultaneous current measurements have indicated that this is associated with
particular current directions, both near the western
boundary and along the equator (usually, towards
the southern hemisphere and towards the east,
respectively). In particular, along 35°W, flows
associated with high CFCs are consistently eastward near 2–3°S in the uNADW (Fig. 4.3.8),
and near 1.5°S in the lNADW. There are also
occasional eastward zonal currents in the uNADW
along the equator, but with large differences from
cruise to cruise (Rhein et al., 1995; Fischer and
Schott, 1997). Similar to what is observed in the
Pacific Ocean, all the current profiles within 1° of
the equator indicate the presence of jets with a
short vertical scale of 400–600 m, in particular
between 1000 and 2500 m depth (Gouriou et al.,
1999). These jets are found to last many months
and have a large zonal coherence. The flow north
of the equator can also be quite variable, in particular away from the slope, as has been shown from
two repeat sections off French Guiana separated
by 45 days (Andrié et al., 1999).
The deep water flow along the continental slope
is steadier than is found in the interior. Moorings
near 44°W indicate a current trapped to within
SECTION 4 THE GLOBAL FLOW FIELD
230