N out to the mid-Atlantic Ridge, and along
two sections at 6
and 4
N normal to the continental
slope. Their observations showed the basic structure of two CFC
maxima. Unlike Abaco, in the tropics the CFC and high velocity
cores are coincident. Molinari et al. [1992] found that the
shallow CFC and velocity cores are located over bottom depths
between 1500 and 2500 m, whereas the deep cores are located further
offshore over bottom depths of 4000 m. Although the deep CFC core
correlates with high salinity, high oxygen and low vertical density
gradient, the shallow core does not. Just below the shallow core
they identified a layer of high oxygen saturation and low density
gradient at 3.2
to 3.6
C as LSW. Molinari et al.
[1992] showed that the water masses at the three sections have
similar properties, and that the total transport of the DWBC at
each section is 26 Sv. Of the total, approximately 8.5 Sv is
transported by the DWBC in the CFC minimum layer of recirculated
water. Some of this recirculated water is of Southern Hemisphere
origin [ Reid, 1994]. The transport in the LNADW was 13 Sv,
with 4.5 Sv in the SLSW/LSW, giving 17 Sv of recently ventilated
water and a total of 26 Sv approaching the equator in the DWBC.
The 26 Sv total agrees with the estimate of Friedrichs and
Hall [1993] at 11
N. Possibly due to different vertical
limits, other estimates of transport using geostrophy, floats and
direct velocities [e.g., Speer and McCartney, 1991;
Richardson and Schmitz, 1993; Johns et al., 1993] yielded
transports that are three times higher for the upper core.
A comparison by Molinari et al. [1992] of CFC
concentrations measured in 1989 with the 1982-83 TTO data [
Weiss et al., 1985] showed large concentration increases with
time, not only in the DWBC, but also in the interior. The elevated
tracer concentrations in the interior are significant, in that they
are consistent with recirculation of the DWBC into an interior gyre
in the Guiana Basin [ Johns et al., 1990; McCartney,
1993; Johns et al., 1993; Friedrichs and Hall, 1993;
Friedrichs et al., 1994]. Also consistent with this
recirculation is the Molinari et al. [1992] conclusion that
the majority of the bottom flow is steered north by the Ceara Rise.
They found the water below 1.8
C does not exit through the
44
W section, and that low salinity bottom water flows
poleward north of the Ceara Rise.
The observations of Rhein et al. [1994] included a
combination of CFCs, hydrographic data, lowered ADCP and PEGASUS
profiling done annually between 1990-92 along 35
W (from the
boundary to 2
N) and 5
S (from the boundary to
30
W). They were the first to measure detectable CFC
concentrations in the AABW in the Northern Hemisphere.
Rhein et al. [1994] pointed out that the LSW should
have lower tracer concentrations than the SLSW, because LSW
convects to greater depths. They estimated that 7 Sv of the
SLSW crosses the equator in the DWBC (none in the interior),
as compared with only 1 Sv of the LNADW (2 Sv in the interior).
In the DWBC along 35
W highest transports are associated
with both CFC maxima cores. This is in contrast to the DWBC at
5
S, where the high velocity cores extend over a wider depth
range than the CFC cores.
In addition to mean transports, Rhein et al. [1994]
presented evidence of variability in the direction of the SLSW flow
at 35
W on the equator. In November 1992, the flow of the
SLSW was eastward, but in October 1990 and May-June 1991 the flow
was westward. Flow reversals along the equator are also observed
in the intermediate depth float tracks [ Richardson and
Schmitz, 1993], and are consistent with model calculations [
Boning and Schott, 1993]. Whereas, in all of the Rhein et
al. [1994] observations, the LNADW flows eastward along the
equator and is guided by the Parnaiba Ridge. However, an important
finding of Rhein et al. [1994] was that the LNADW may not
follow a continuous path south of the equator. The tracer and
velocity cores associated with the LNADW are considerably weaker at
5
S, and the 
-S properties are significantly
different. This is consistent with the conclusions of Speer
and McCartney [1991] and Friedrichs et al. [1994].
Friedrichs et al. [1994] suggested LNADW flows eastward to upwell
in the eastern tropical Atlantic, and appears in weak form in the
South Atlantic DWBC (see discussion of Wallace et al. [1994],
below). Thus, the structure that is emerging from the recent work
is that of a three part branching of the DWBC upon reaching the
equator: circulation back northwards as part of the Guiana Basin
gyre, flow along the equator (which for SLSW is temporally
variable), and flow into the Southern Hemisphere (with eastward
branching of the LNADW).
Data collected by Wallace et al. [1994] extended the
structure of the DWBC to 19
S (the WOCE A9 section). The full
suite of CFCs, including CCl
and F113 (present for about twenty-
five years) were measured in 1991. The compound F113 was
undetectable below 500 m. The SLSW in the DWBC of the Brazil Basin
had measurable concentrations of F11, F12 and CCl
. The SLSW had
an F11/F12 age of thirty-five to forty years and a CCl
/F11 age
of thirty-two years, with a 2-7 fold dilution factor. Both the
ages and dilution factors are consistent with the observations of
Weiss et al. [1985] at the equator. The SLSW appears to be
older and less diluted than the northward flowing AABW. The
northward flowing bottom waters had measurable concentrations of
F11 and F12, and a ratio age of thirty years. The distribution of
CCl
is more extensive than that of F11. Unlike F11,
anthropogenic CCl
is detectable in the LNADW, and throughout the
Brazil Basin. The compound CCl
is also present on the eastern
flank of the mid-Atlantic Ridge in the bottom waters of the Angola
Basin. Differences in the CFC compound distributions highlight the
potential of CCl
as a tracer for processes on time scales of up
to eighty years.