by either wind, buoyancy loss, or mixing along
isopycnals, depending on the author. McCartney
(1982) presented an alternative view, that AAIW is
made up of the densest SAMW produced in the
Pacific, the end-product of gradual cooling, freshening and loss of buoyancy by air–sea fluxes along
the long SAMW circulation path across the Indian
and Pacific Oceans. England et al. (1993) showed
a similar mechanism operating in a GCM. From
the southeast Pacific, AAIW spreads north and
west in the Pacific and through Drake Passage.
Further modification by air–sea fluxes and mixing
in the southwest Atlantic produces a cooler and
fresher variety of AAIW in the southwest Atlantic
(Molinelli, 1981; Piola and Georgi, 1982; Piola
and Gordon, 1989; Talley, 1996a).
High-quality hydrographic and tracer sections
collected during WOCE have focused attention on
the distribution and circulation of SAMW and
AAIW. The SAMW becomes progressively cooler,
fresher and denser across the Indian and Pacific
basins (McCartney, 1982). Within each basin, the
lighter SAMW varieties are injected further west,
and are restricted to the southwest corner of the
subtropical gyre. Dense varieties entering the subtropical gyres on their eastern sides travel around
the gyres and extend to lower latitudes. While the
circulation of the denser AAIW shares some similarities with that of the overlying SAMW, there are
important differences as well. In contrast to the
near-circumpolar formation of SAMW, the formation of ‘new’, well-ventilated AAIW is limited to
the southeast Pacific and southwest Atlantic, as
indicated by the gradual decrease in AAIW oxygen
and potential vorticity across the Atlantic and
Indian basins (Gordon and Molinelli, 1982;
Talley, 1996a). For example, at the WOCE SR3
section south of Tasmania (the eastern limit of the
Atlantic–Indian variety of AAIW), the oxygen and
chlorofluorocarbon (CFC) saturations of AAIW
are only 65% and 10–20%, respectively; the ‘ventilation age’ of AAIW based on CFC-11 is 22–26
years (Rintoul and Bullister, 1999). AAIW enters
the lower thermocline of the Atlantic and Pacific
subtropical gyres mainly in the southeast quadrant; in contrast, oxygen-rich AAIW enters the
Indian basin preferentially in the southwest (Fine,
1993; Toole and Warren, 1993). In the western
boundary currents of the subtropical gyres, modified ‘older’ AAIW returns to the south, resulting in
multiple varieties of AAIW with similar densities
at some longitudes (e.g. 40°E, Read and Pollard,
1993; 140°E, Rintoul and Bullister, 1999), and a
shift in AAIW properties between basins (Piola
and Georgi, 1982).
Few estimates have been made of the formation
rate of SAMW and AAIW. Even less attention has
been paid to the rest of the circulation loop associated with export of SAMW/AAIW to lower latitudes: what is the fate of SAMW/AAIW exported
from the Southern Ocean? What water masses are
modified to supply the SAMW/AAIW, and where
and how does this occur?
Sloyan and Rintoul (2000, 2001a) have used a
box inverse model, which explicitly includes
air–sea buoyancy fluxes and the water mass transformations they drive (e.g. Walin, 1982; Speer and
Tziperman, 1992), to examine the formation and
circulation of SAMW and AAIW. Comparing the
transport in density layers at each Southern Ocean
chokepoint shows that 18 Sv more SAMW/AAIW
leaves the Indian Ocean sector south of Australia
than enters the basin south of Africa (Fig. 4.6.12).
This production rate represents the net effect of
air–sea buoyancy fluxes, diapycnal mixing and
meridional exchange with the subtropical gyre.
The net production in the Indian sector is balanced
by net consumption in the other basins. SAMW is
formed in the Indian sector both by cooling and
freshening of subtropical water carried south in
the Agulhas Current and its extension, and by
water flowing northward across the ACC (and
gaining buoyancy) in the Ekman layer. SAMW
is carried from the Indian to the Pacific by the
ACC, where it is modified by mixing and air–sea
exchange, and ultimately returns to the Indian
basin via the Indonesian passages (Gordon, Chapter
4.7). The SAMW therefore participates in an
Indian–Pacific ‘throughflow gyre’ in which warm
throughflow water is converted to cool SAMW in
the Indian Ocean, and SAMW entering the Pacific
basin is converted back to water warm enough to
supply the throughflow.
In the zonal integral, 34 Sv of Upper Circumpolar Deep Water (UCDW) is upwelled south of the
ACC and converted to IW densities by air–sea
buoyancy flux (Fig. 4.6.12; see also the discusion in
Section 4.6.5 and Fig. 4.6.14). This transformation
by air–sea fluxes of heat and fresh water is largely
compensated by diapycnal mixing south of 40°S.
However, the compensation is not immediate.
While the zonally integrated export of SAMW/
4.6 The Antarctic Circumpolar Current System
293
Rintoul, Hughes and Olbers
isopycnals, depending on the author. McCartney
(1982) presented an alternative view, that AAIW is
made up of the densest SAMW produced in the
Pacific, the end-product of gradual cooling, freshening and loss of buoyancy by air–sea fluxes along
the long SAMW circulation path across the Indian
and Pacific Oceans. England et al. (1993) showed
a similar mechanism operating in a GCM. From
the southeast Pacific, AAIW spreads north and
west in the Pacific and through Drake Passage.
Further modification by air–sea fluxes and mixing
in the southwest Atlantic produces a cooler and
fresher variety of AAIW in the southwest Atlantic
(Molinelli, 1981; Piola and Georgi, 1982; Piola
and Gordon, 1989; Talley, 1996a).
High-quality hydrographic and tracer sections
collected during WOCE have focused attention on
the distribution and circulation of SAMW and
AAIW. The SAMW becomes progressively cooler,
fresher and denser across the Indian and Pacific
basins (McCartney, 1982). Within each basin, the
lighter SAMW varieties are injected further west,
and are restricted to the southwest corner of the
subtropical gyre. Dense varieties entering the subtropical gyres on their eastern sides travel around
the gyres and extend to lower latitudes. While the
circulation of the denser AAIW shares some similarities with that of the overlying SAMW, there are
important differences as well. In contrast to the
near-circumpolar formation of SAMW, the formation of ‘new’, well-ventilated AAIW is limited to
the southeast Pacific and southwest Atlantic, as
indicated by the gradual decrease in AAIW oxygen
and potential vorticity across the Atlantic and
Indian basins (Gordon and Molinelli, 1982;
Talley, 1996a). For example, at the WOCE SR3
section south of Tasmania (the eastern limit of the
Atlantic–Indian variety of AAIW), the oxygen and
chlorofluorocarbon (CFC) saturations of AAIW
are only 65% and 10–20%, respectively; the ‘ventilation age’ of AAIW based on CFC-11 is 22–26
years (Rintoul and Bullister, 1999). AAIW enters
the lower thermocline of the Atlantic and Pacific
subtropical gyres mainly in the southeast quadrant; in contrast, oxygen-rich AAIW enters the
Indian basin preferentially in the southwest (Fine,
1993; Toole and Warren, 1993). In the western
boundary currents of the subtropical gyres, modified ‘older’ AAIW returns to the south, resulting in
multiple varieties of AAIW with similar densities
at some longitudes (e.g. 40°E, Read and Pollard,
1993; 140°E, Rintoul and Bullister, 1999), and a
shift in AAIW properties between basins (Piola
and Georgi, 1982).
Few estimates have been made of the formation
rate of SAMW and AAIW. Even less attention has
been paid to the rest of the circulation loop associated with export of SAMW/AAIW to lower latitudes: what is the fate of SAMW/AAIW exported
from the Southern Ocean? What water masses are
modified to supply the SAMW/AAIW, and where
and how does this occur?
Sloyan and Rintoul (2000, 2001a) have used a
box inverse model, which explicitly includes
air–sea buoyancy fluxes and the water mass transformations they drive (e.g. Walin, 1982; Speer and
Tziperman, 1992), to examine the formation and
circulation of SAMW and AAIW. Comparing the
transport in density layers at each Southern Ocean
chokepoint shows that 18 Sv more SAMW/AAIW
leaves the Indian Ocean sector south of Australia
than enters the basin south of Africa (Fig. 4.6.12).
This production rate represents the net effect of
air–sea buoyancy fluxes, diapycnal mixing and
meridional exchange with the subtropical gyre.
The net production in the Indian sector is balanced
by net consumption in the other basins. SAMW is
formed in the Indian sector both by cooling and
freshening of subtropical water carried south in
the Agulhas Current and its extension, and by
water flowing northward across the ACC (and
gaining buoyancy) in the Ekman layer. SAMW
is carried from the Indian to the Pacific by the
ACC, where it is modified by mixing and air–sea
exchange, and ultimately returns to the Indian
basin via the Indonesian passages (Gordon, Chapter
4.7). The SAMW therefore participates in an
Indian–Pacific ‘throughflow gyre’ in which warm
throughflow water is converted to cool SAMW in
the Indian Ocean, and SAMW entering the Pacific
basin is converted back to water warm enough to
supply the throughflow.
In the zonal integral, 34 Sv of Upper Circumpolar Deep Water (UCDW) is upwelled south of the
ACC and converted to IW densities by air–sea
buoyancy flux (Fig. 4.6.12; see also the discusion in
Section 4.6.5 and Fig. 4.6.14). This transformation
by air–sea fluxes of heat and fresh water is largely
compensated by diapycnal mixing south of 40°S.
However, the compensation is not immediate.
While the zonally integrated export of SAMW/
4.6 The Antarctic Circumpolar Current System
293
Rintoul, Hughes and Olbers