and Gordon (1995) find that three Agulhas eddies
found within the Cape Basin (east of Walvis Ridge)
each transferred between 0.45 and 0.90 Sv of Indian
Ocean thermocline and intermediate water. As there
are on average six such eddies per year, the total
flux of Indian Ocean water above 1000 m introduced into the Atlantic by the eddy mechanism may
be from 5 to 10 Sv. Van Ballegooyen et al. (1994)
finds that Agulhas eddies contribute salt and heat to
the South Atlantic thermocline waters at a rate of up
to 2.510
6 kg s
91 (salt) and 0.045 PW (heat),
assuming a yearly total of six eddies. Agulhas filaments may provide only 13% of the total interocean
salt flux, the heat being lost quickly to the atmosphere (Lutjeharms and Cooper, 1996). Garzoli
et al. (1996) and Duncombe Rae et al. (1996) find
that the eddies measured during the BEST
experiment (1992–93) contributed 0.007 PW,
4.5–6.710
5 kg s
91 and 2.6–3.8 Sv of heat, salt and
volume flux, respectively, into the South Atlantic
Ocean.
Garzoli et al. (1999) concluded from a detailed
survey of three Agulhas eddies that a reasonable
estimate for mass and heat flux from the Indian
Ocean to the Atlantic is 6–10 Sv and 0.006 PW.
The decrease of Indian Ocean core water observed
as an Agulhas eddy dissipates indicate a slow transfer of Indian Ocean water from the confines of the
eddy to the Benguela Current. In this way Indian
Ocean water injected into the South Atlantic within
an eddy may not accompany the dynamic feature
of the eddy across the South Atlantic near 30°S
(Byrne et al., 1995; Witter and Gordon, 1999), but
rather may be advected to the northwest within the
general circulation of the Benguela Current and
South Equatorial Current, as shown schematically
by Gordon et al. (1992). The interaction of an
Agulhas eddy with the atmosphere, seafloor topography (Kamenkovich et al., 1996) and surrounding South Atlantic water strongly control its
trajectory and modification (Arhan et al., 1999).
Weijer et al. (1999) find that heat and salt
flux associated with South Atlantic interocean
exchanges, and in particular the ratio of thermocline
to intermediate components (lateral boundary buoyancy profile), has a strong influence on the Atlantic
meridional overturning circulation. An important
aspect of the injection of Indian Ocean salt into the
South Atlantic at the Agulhas Retroflection may be
related to the ‘… shape of the salt flux profile …’,
but the more important impact of Agulhas leakage
to Atlantic meridional overturning circulation may
be due to the Indian Ocean heat flux, which will
drive further evaporation in the South Atlantic.
There is ambiguity in the heat flux calculations
associated with Agulhas leakage, as authors use
different reference temperatures. To determine
interocean heat flux one needs to identify the temperature of the water that leaves the Atlantic
Ocean, balancing the mass injection of the
Agulhas leakage. There are two reasonable options
(Gordon, 1985): on the horizontal plane, with the
transfer of cooler South Atlantic Current upperlayer water into the Indian Ocean south of the
Agulhas Return Current; or on the vertical plane,
by export of (2°C) NADW into the Indian and
Pacific Oceans. Transport-weighted temperature
(relative to 1500 m) was estimated from the SAVE4 section across the South Atlantic Current crossing 10°W (used by Gordon et al., 1992) as 10°C.
The transport-weighted Agulhas transport (relative
to 1500 m) taken between the station pair 49 and
50 in Gordon et al., (1987) is calculated as 15°C.
The heat flux from the Indian Ocean to the
Atlantic for an Agulhas leakage of 15 Sv (upper
1500 m, Gordon et al., 1992) for closure within
the upper 1500 m is 0.3 PW (this is a larger value
than calculated by Gordon (1985), where the
smaller temperatures difference from opposing
sides of a detached Agulhas eddy was used). For
closure by the colder NADW the heat flux is
0.8 PW. The corresponding values for 5 Sv and
10 Sv Agulhas leakage, respectively are 0.1,
0.3 PW and 0.2, 0.6 PW. Estimates of the northward heat flux across the South Atlantic subtropics (18°S to 30°S) ranges from 0.1 to 0.9 PW, with
error estimated in excess of 0.2 PW (MacDonald
and Wunsch, 1996; Ganachaud, 1999; Holfort
and Siedler, 2001). Heat flux values of less than
0.3 PW indicate that Agulhas leakage is predominately balanced by export of Atlantic water
within the upper 1500 m, while heat flux values
greater than 0.3 PW suggest significant involvement with the Atlantic meridional overturning circulation. The two methods for closure are not
mutually exclusive and their relative importance
may be time variable.
Within 50 km of the coast the Agulhas Current
intermediate water, in the 4–8°C interval, is more
saline than at points further offshore (Beal and
Bryden, 1999). Remnants of Red Sea water flow
southward are observed hugging the continental
SECTION 4 THE GLOBAL FLOW FIELD
312
found within the Cape Basin (east of Walvis Ridge)
each transferred between 0.45 and 0.90 Sv of Indian
Ocean thermocline and intermediate water. As there
are on average six such eddies per year, the total
flux of Indian Ocean water above 1000 m introduced into the Atlantic by the eddy mechanism may
be from 5 to 10 Sv. Van Ballegooyen et al. (1994)
finds that Agulhas eddies contribute salt and heat to
the South Atlantic thermocline waters at a rate of up
to 2.510
6 kg s
91 (salt) and 0.045 PW (heat),
assuming a yearly total of six eddies. Agulhas filaments may provide only 13% of the total interocean
salt flux, the heat being lost quickly to the atmosphere (Lutjeharms and Cooper, 1996). Garzoli
et al. (1996) and Duncombe Rae et al. (1996) find
that the eddies measured during the BEST
experiment (1992–93) contributed 0.007 PW,
4.5–6.710
5 kg s
91 and 2.6–3.8 Sv of heat, salt and
volume flux, respectively, into the South Atlantic
Ocean.
Garzoli et al. (1999) concluded from a detailed
survey of three Agulhas eddies that a reasonable
estimate for mass and heat flux from the Indian
Ocean to the Atlantic is 6–10 Sv and 0.006 PW.
The decrease of Indian Ocean core water observed
as an Agulhas eddy dissipates indicate a slow transfer of Indian Ocean water from the confines of the
eddy to the Benguela Current. In this way Indian
Ocean water injected into the South Atlantic within
an eddy may not accompany the dynamic feature
of the eddy across the South Atlantic near 30°S
(Byrne et al., 1995; Witter and Gordon, 1999), but
rather may be advected to the northwest within the
general circulation of the Benguela Current and
South Equatorial Current, as shown schematically
by Gordon et al. (1992). The interaction of an
Agulhas eddy with the atmosphere, seafloor topography (Kamenkovich et al., 1996) and surrounding South Atlantic water strongly control its
trajectory and modification (Arhan et al., 1999).
Weijer et al. (1999) find that heat and salt
flux associated with South Atlantic interocean
exchanges, and in particular the ratio of thermocline
to intermediate components (lateral boundary buoyancy profile), has a strong influence on the Atlantic
meridional overturning circulation. An important
aspect of the injection of Indian Ocean salt into the
South Atlantic at the Agulhas Retroflection may be
related to the ‘… shape of the salt flux profile …’,
but the more important impact of Agulhas leakage
to Atlantic meridional overturning circulation may
be due to the Indian Ocean heat flux, which will
drive further evaporation in the South Atlantic.
There is ambiguity in the heat flux calculations
associated with Agulhas leakage, as authors use
different reference temperatures. To determine
interocean heat flux one needs to identify the temperature of the water that leaves the Atlantic
Ocean, balancing the mass injection of the
Agulhas leakage. There are two reasonable options
(Gordon, 1985): on the horizontal plane, with the
transfer of cooler South Atlantic Current upperlayer water into the Indian Ocean south of the
Agulhas Return Current; or on the vertical plane,
by export of (2°C) NADW into the Indian and
Pacific Oceans. Transport-weighted temperature
(relative to 1500 m) was estimated from the SAVE4 section across the South Atlantic Current crossing 10°W (used by Gordon et al., 1992) as 10°C.
The transport-weighted Agulhas transport (relative
to 1500 m) taken between the station pair 49 and
50 in Gordon et al., (1987) is calculated as 15°C.
The heat flux from the Indian Ocean to the
Atlantic for an Agulhas leakage of 15 Sv (upper
1500 m, Gordon et al., 1992) for closure within
the upper 1500 m is 0.3 PW (this is a larger value
than calculated by Gordon (1985), where the
smaller temperatures difference from opposing
sides of a detached Agulhas eddy was used). For
closure by the colder NADW the heat flux is
0.8 PW. The corresponding values for 5 Sv and
10 Sv Agulhas leakage, respectively are 0.1,
0.3 PW and 0.2, 0.6 PW. Estimates of the northward heat flux across the South Atlantic subtropics (18°S to 30°S) ranges from 0.1 to 0.9 PW, with
error estimated in excess of 0.2 PW (MacDonald
and Wunsch, 1996; Ganachaud, 1999; Holfort
and Siedler, 2001). Heat flux values of less than
0.3 PW indicate that Agulhas leakage is predominately balanced by export of Atlantic water
within the upper 1500 m, while heat flux values
greater than 0.3 PW suggest significant involvement with the Atlantic meridional overturning circulation. The two methods for closure are not
mutually exclusive and their relative importance
may be time variable.
Within 50 km of the coast the Agulhas Current
intermediate water, in the 4–8°C interval, is more
saline than at points further offshore (Beal and
Bryden, 1999). Remnants of Red Sea water flow
southward are observed hugging the continental
SECTION 4 THE GLOBAL FLOW FIELD
312