circulation, particularly that which is in response
to NADW formation (Gordon, 1986, 1996a,b;
Broecker, 1991; Rintoul, 1991; Gordon et al.,
1992; Schmitz, 1995; MacDonald and Wunsch,
1996). The reader is directed to the excellent twovolume report of Schmitz (1996a,b).
NADW is exported into the Indian and Pacific
Oceans by the ACC, becoming entangled in the
overturning circulation of the Southern Ocean and
associated AAIW and AABW formation. NADW
export from the Atlantic Ocean must be balanced
by import of Indian and Pacific Ocean waters
within the water column shallower than NADW –
but how? From the warm saline Indian Ocean
thermocline and intermediate waters around the
southern rim of Africa (the warm route), or from
cooler, fresher subpolar Pacific water through the
Drake Passage (the cold route)? Or, as more likely,
both? Does the Indonesian Throughflow have anything to do with NADW formation? Do the ratio
or efficiencies of these return routes vary in time?
Might such variability be coupled to NADW formation and climate variability? Though there has
been a plethora of papers on these subjects, some
favouring the warm route (a recent example is that
of Holfort and Siedler, 2001) and others the cold
route (Rintoul, 1991; Schlitzer, 1996), definitive
answers as to the climate importance of interocean
exchange are still evolving.
MacDonald and Wunsch (1996) investigated
the nature of the global thermohaline circulation
pattern using a set of 23 (mostly WOCE) sections
and the statistical guidelines of the inverse box
model approach. Within the limits of the non-synoptic data set, two independent large-scale, global
integrated circulation cells emerge. The Atlantic–
Southern Ocean cell carries NADW into the Southern Ocean, where it is integrated into the Southern
Ocean overturning cell and spreads into the Indian
and Pacific Oceans. The second cell is confined
more to a horizontal plane, linking the Pacific and
Indian Oceans by westward flow within the
Indonesian Seas and eastward flow (presumably
within the subpolar zone) south of Australia. The
two cells are linked through the highly time-dependent Agulhas Retroflection south of Africa and
through upwelling of deep waters in the Pacific
Ocean (Gordon, 1996b).
Interannual to decadal SST anomalies may be
transferred between ocean basins by the interocean
links (e.g. the ACC; Peterson and White, 1998).
Also, one can envision that SST anomalies could
be generated within an ocean basin by variability
in the interocean transport (e.g. variability in any
of the three interocean channels discussed below:
Bering Strait, Indonesian Seas and Agulhas Retroflection). Millennium-scale changes in wind and
sea level associated with glacial epochs may also
be expected to alter the form of the interocean
exchange, which would alter global thermohaline
circulation with feedback to the climate system
(Seidov and Haupt, 1999).
While the ACC is by far the largest conduit
for interocean exchange, water mass differences
between the major ocean basins would be much
larger were it not for various smaller interocean
links, which may be linked into a global chain.
The objective of this chapter is to present the
current state of knowledge about those small but
important interocean links. The regional oceanography of these areas is not developed (see Tomczak
and Godfrey, 1994), but rather only those aspects
directly associated with interocean fluxes. The
interocean links discussed below are: the Bering
Strait (Fig. 4.7.1c, see Plate 4.7.1c, p. 300) and
Indonesian Seas (Fig. 4.7.1d, see Plate 4.7.1d,
p. 300), which allow for the export of low-salinity
North Pacific upper layer water to the Atlantic
(Arctic) and Indian Oceans, respectively; and the
Agulhas leakage of Indian Ocean thermocline and
intermediate water into the Atlantic at the southern rim of Africa (Fig. 4.7.1e, see Plate 4.7.1e,
p. 300). Before discussing these interocean links, it
is worthwhile mentioning the presence of westward flow of Pacific water into the Indian Ocean
immediately south of Australia.
4.7.1.4 South of Australia
Observations (Fine, 1993; Reid, 1997; Rintoul and
Bullister, 1999) and models (Semtner and Chervin,
1992; Speich et al., 2000) suggest flow of Pacific
water into the Indian Ocean along the southern
coast of Australia. While much of this may be
part of a closed anticyclonic gyre in the Great
Australian Bight, there is a possibility of waters
from the Tasman Sea flowing into the Indian
Ocean. Rintoul and Bullister (1999) find 2–3 Sv of
Tasman water flowing westward south of Tasmania within the 800 and 3000 m depth interval;
Speich et al. (2000) in their model study find
3.2 Sv of Pacific water entering the Indian Ocean
in the upper 1200 m, and the model shows this
4.7 Interocean Exchange
305
Gordon
to NADW formation (Gordon, 1986, 1996a,b;
Broecker, 1991; Rintoul, 1991; Gordon et al.,
1992; Schmitz, 1995; MacDonald and Wunsch,
1996). The reader is directed to the excellent twovolume report of Schmitz (1996a,b).
NADW is exported into the Indian and Pacific
Oceans by the ACC, becoming entangled in the
overturning circulation of the Southern Ocean and
associated AAIW and AABW formation. NADW
export from the Atlantic Ocean must be balanced
by import of Indian and Pacific Ocean waters
within the water column shallower than NADW –
but how? From the warm saline Indian Ocean
thermocline and intermediate waters around the
southern rim of Africa (the warm route), or from
cooler, fresher subpolar Pacific water through the
Drake Passage (the cold route)? Or, as more likely,
both? Does the Indonesian Throughflow have anything to do with NADW formation? Do the ratio
or efficiencies of these return routes vary in time?
Might such variability be coupled to NADW formation and climate variability? Though there has
been a plethora of papers on these subjects, some
favouring the warm route (a recent example is that
of Holfort and Siedler, 2001) and others the cold
route (Rintoul, 1991; Schlitzer, 1996), definitive
answers as to the climate importance of interocean
exchange are still evolving.
MacDonald and Wunsch (1996) investigated
the nature of the global thermohaline circulation
pattern using a set of 23 (mostly WOCE) sections
and the statistical guidelines of the inverse box
model approach. Within the limits of the non-synoptic data set, two independent large-scale, global
integrated circulation cells emerge. The Atlantic–
Southern Ocean cell carries NADW into the Southern Ocean, where it is integrated into the Southern
Ocean overturning cell and spreads into the Indian
and Pacific Oceans. The second cell is confined
more to a horizontal plane, linking the Pacific and
Indian Oceans by westward flow within the
Indonesian Seas and eastward flow (presumably
within the subpolar zone) south of Australia. The
two cells are linked through the highly time-dependent Agulhas Retroflection south of Africa and
through upwelling of deep waters in the Pacific
Ocean (Gordon, 1996b).
Interannual to decadal SST anomalies may be
transferred between ocean basins by the interocean
links (e.g. the ACC; Peterson and White, 1998).
Also, one can envision that SST anomalies could
be generated within an ocean basin by variability
in the interocean transport (e.g. variability in any
of the three interocean channels discussed below:
Bering Strait, Indonesian Seas and Agulhas Retroflection). Millennium-scale changes in wind and
sea level associated with glacial epochs may also
be expected to alter the form of the interocean
exchange, which would alter global thermohaline
circulation with feedback to the climate system
(Seidov and Haupt, 1999).
While the ACC is by far the largest conduit
for interocean exchange, water mass differences
between the major ocean basins would be much
larger were it not for various smaller interocean
links, which may be linked into a global chain.
The objective of this chapter is to present the
current state of knowledge about those small but
important interocean links. The regional oceanography of these areas is not developed (see Tomczak
and Godfrey, 1994), but rather only those aspects
directly associated with interocean fluxes. The
interocean links discussed below are: the Bering
Strait (Fig. 4.7.1c, see Plate 4.7.1c, p. 300) and
Indonesian Seas (Fig. 4.7.1d, see Plate 4.7.1d,
p. 300), which allow for the export of low-salinity
North Pacific upper layer water to the Atlantic
(Arctic) and Indian Oceans, respectively; and the
Agulhas leakage of Indian Ocean thermocline and
intermediate water into the Atlantic at the southern rim of Africa (Fig. 4.7.1e, see Plate 4.7.1e,
p. 300). Before discussing these interocean links, it
is worthwhile mentioning the presence of westward flow of Pacific water into the Indian Ocean
immediately south of Australia.
4.7.1.4 South of Australia
Observations (Fine, 1993; Reid, 1997; Rintoul and
Bullister, 1999) and models (Semtner and Chervin,
1992; Speich et al., 2000) suggest flow of Pacific
water into the Indian Ocean along the southern
coast of Australia. While much of this may be
part of a closed anticyclonic gyre in the Great
Australian Bight, there is a possibility of waters
from the Tasman Sea flowing into the Indian
Ocean. Rintoul and Bullister (1999) find 2–3 Sv of
Tasman water flowing westward south of Tasmania within the 800 and 3000 m depth interval;
Speich et al. (2000) in their model study find
3.2 Sv of Pacific water entering the Indian Ocean
in the upper 1200 m, and the model shows this
4.7 Interocean Exchange
305
Gordon