The 0 to 150 db temperature is obtained by
subtracting the depth-weighted 150 to 400 db temperature time series from the full-depth temperature time series estimated from the travel time
measured by Inverted Echo Sounders. The 1997
Makassar Strait internal energy transport is
0.50 PW (petawatts). If the Makassar temperatures
are referenced to 3.72°C, as in Schiller et al.
(1998), the 1997 internal energy transport is
reduced to 0.39 PW. The ENSO influence on the
internal energy transport may be estimated:
0.63 PW during the La Niña months of December
1996 through February 1997, and 0.39 PW during
the El Niño months of December 1997 through
February 1998.
The average transport-weighted temperature
determined by the Arlindo moorings is approximately 11.5°C with an average salinity of 34.45
(K. Vranes, Columbia graduate student). This is
cooler and saltier than the ITF characteristics envisioned by Piola and Gordon (1984; ITF of 33.6)
and Toole and Warren (1993; ITF temperature of
24°C). Toole and Warren (1993) deduce a
throughflow of 6.7 Sv with salinity of 34.5. They
assign it with a temperature of 24°C, which yields
a heat-flux divergence between Indonesia and 32°S
of 0.98 PW. The heat and freshwater divergence
within the Indian Ocean north of 32°S depends on
the temperature and salinity of the net poleward
flow across 32°S. Using the Toole and Raymer
(1985) Fig. 3 and their mean temperature and
salinity along 32°S across the Indian Ocean of
5.71°C and 34.80, together with the ITF characteristics found by the Arlindo moorings, yields an
ITF heat flux of roughly 0.27 PW and a net precipitation minus evaporation plus runoff over the
Indian Ocean (north of 32°S) of nearly 90.3 Sv.
As the ITF water is not likely to be cooled to
5.71°C, this value represents a maximum heat
flux. An estimate of minimum heat flux may be
calculated by assuming the ITF water eventually
passing poleward within the Agulhas Current
across 32°S confined to roughly the same density
interval as the ITF in the Indonesian passages. This
yields a value statistically equal to zero. This
means that the ITF heat may be eventually lost
south of 32°S, consistent with the model results of
Hirst and Godfrey (1993) and of Allan et al.
(1995), who find much enhanced ocean-to-atmosphere heat flux along 40°S when the ITF is
included.
4.7.4 The Agulhas Retroflection
4.7.4.1 Introduction
A third interocean exchange route considered as
important to larger-scale thermohaline circulation
lies not within the confines of a channel, but rather
in the gap between the southern shores of Africa
and the ACC, a gap occupied by the Agulhas
Retroflection (see Lutjeharms, 1996, Fig. 1d). The
Agulhas Current flows westward along the southern
rim of Africa, with a transport approaching 100 Sv.
Rather than continuing into the South Atlantic,
Agulhas water curls back to the Indian Ocean, feeding the eastward-flowing Agulhas Return Current
near 40°S, the ACC main axis falling further south,
near 48°S (Read and Pollard, 1993). However, not
all of the Agulhas water turns back to the Indian
Ocean. That part that does not turn back passes
into the South Atlantic, in what is often called Agulhas leakage. Of all of the interocean exchanges, the
Agulhas leakage into the Atlantic has drawn the
most attention (and controversy) in terms of its role
in NADW formation (Gordon, 1996a).
Recent WOCE-generated advances in our
understanding of the Indian–South Atlantic
exchanges are reported in two recent collections:
Wefer et al. (1996) and a special section of the
Journal of Geophysical Research (Gordon et al.,
1999b). The articles by de Ruijter et al. (1999),
Witter and Gordon (1999), Garzoli et al. (1999),
Arhan et al. (1999) and McDonagh and Heywood
(1999) are particularly relevant to the topic of
invasion of Indian Ocean water into the Atlantic.
Also see Rintoul et al., Chapter 4.6 and the model
results of Semtner and Chervin (1992), Boddem
and Schlitzer (1995), Cai and Greatbatch (1995),
Döös (1995), Lutjeharms and Webb (1995) and
Florenchie and Verron (1998). How models ‘handle’ the Agulhas leakage depends very much on
their eddy-resolving characteristics (Marsh et al.,
2000). The de Ruijter et al. (1999) paper provides
an excellent review of interocean exchange
afforded by the Agulhas Retroflection.
4.7.4.2 Transport
Benguela Current
Gordon et al. (1992), using oxygen and CFC data
collected during the South Atlantic Ventilation
Experiment cruise no. 4 (SAVE-4) from 7 December 1988 to 15 January 1989 (not 1989 and 1990,
as stated in Gordon et al., 1992), find that of the
SECTION 4 THE GLOBAL FLOW FIELD
310
subtracting the depth-weighted 150 to 400 db temperature time series from the full-depth temperature time series estimated from the travel time
measured by Inverted Echo Sounders. The 1997
Makassar Strait internal energy transport is
0.50 PW (petawatts). If the Makassar temperatures
are referenced to 3.72°C, as in Schiller et al.
(1998), the 1997 internal energy transport is
reduced to 0.39 PW. The ENSO influence on the
internal energy transport may be estimated:
0.63 PW during the La Niña months of December
1996 through February 1997, and 0.39 PW during
the El Niño months of December 1997 through
February 1998.
The average transport-weighted temperature
determined by the Arlindo moorings is approximately 11.5°C with an average salinity of 34.45
(K. Vranes, Columbia graduate student). This is
cooler and saltier than the ITF characteristics envisioned by Piola and Gordon (1984; ITF of 33.6)
and Toole and Warren (1993; ITF temperature of
24°C). Toole and Warren (1993) deduce a
throughflow of 6.7 Sv with salinity of 34.5. They
assign it with a temperature of 24°C, which yields
a heat-flux divergence between Indonesia and 32°S
of 0.98 PW. The heat and freshwater divergence
within the Indian Ocean north of 32°S depends on
the temperature and salinity of the net poleward
flow across 32°S. Using the Toole and Raymer
(1985) Fig. 3 and their mean temperature and
salinity along 32°S across the Indian Ocean of
5.71°C and 34.80, together with the ITF characteristics found by the Arlindo moorings, yields an
ITF heat flux of roughly 0.27 PW and a net precipitation minus evaporation plus runoff over the
Indian Ocean (north of 32°S) of nearly 90.3 Sv.
As the ITF water is not likely to be cooled to
5.71°C, this value represents a maximum heat
flux. An estimate of minimum heat flux may be
calculated by assuming the ITF water eventually
passing poleward within the Agulhas Current
across 32°S confined to roughly the same density
interval as the ITF in the Indonesian passages. This
yields a value statistically equal to zero. This
means that the ITF heat may be eventually lost
south of 32°S, consistent with the model results of
Hirst and Godfrey (1993) and of Allan et al.
(1995), who find much enhanced ocean-to-atmosphere heat flux along 40°S when the ITF is
included.
4.7.4 The Agulhas Retroflection
4.7.4.1 Introduction
A third interocean exchange route considered as
important to larger-scale thermohaline circulation
lies not within the confines of a channel, but rather
in the gap between the southern shores of Africa
and the ACC, a gap occupied by the Agulhas
Retroflection (see Lutjeharms, 1996, Fig. 1d). The
Agulhas Current flows westward along the southern
rim of Africa, with a transport approaching 100 Sv.
Rather than continuing into the South Atlantic,
Agulhas water curls back to the Indian Ocean, feeding the eastward-flowing Agulhas Return Current
near 40°S, the ACC main axis falling further south,
near 48°S (Read and Pollard, 1993). However, not
all of the Agulhas water turns back to the Indian
Ocean. That part that does not turn back passes
into the South Atlantic, in what is often called Agulhas leakage. Of all of the interocean exchanges, the
Agulhas leakage into the Atlantic has drawn the
most attention (and controversy) in terms of its role
in NADW formation (Gordon, 1996a).
Recent WOCE-generated advances in our
understanding of the Indian–South Atlantic
exchanges are reported in two recent collections:
Wefer et al. (1996) and a special section of the
Journal of Geophysical Research (Gordon et al.,
1999b). The articles by de Ruijter et al. (1999),
Witter and Gordon (1999), Garzoli et al. (1999),
Arhan et al. (1999) and McDonagh and Heywood
(1999) are particularly relevant to the topic of
invasion of Indian Ocean water into the Atlantic.
Also see Rintoul et al., Chapter 4.6 and the model
results of Semtner and Chervin (1992), Boddem
and Schlitzer (1995), Cai and Greatbatch (1995),
Döös (1995), Lutjeharms and Webb (1995) and
Florenchie and Verron (1998). How models ‘handle’ the Agulhas leakage depends very much on
their eddy-resolving characteristics (Marsh et al.,
2000). The de Ruijter et al. (1999) paper provides
an excellent review of interocean exchange
afforded by the Agulhas Retroflection.
4.7.4.2 Transport
Benguela Current
Gordon et al. (1992), using oxygen and CFC data
collected during the South Atlantic Ventilation
Experiment cruise no. 4 (SAVE-4) from 7 December 1988 to 15 January 1989 (not 1989 and 1990,
as stated in Gordon et al., 1992), find that of the
SECTION 4 THE GLOBAL FLOW FIELD
310