and hence Antarctic Intermediate Water (AAIW)
west of Chile. On the other hand, McCartney
(1982) showed substantial northward circulation
of the SAMW in each ocean basin. The thick
SAMW in the southeast Indian Ocean is singled
out and called the SEISAMW because of its major
impact on ventilation of the Indian Ocean’s subtropical gyre thermocline through subduction and
northward advection. The SEISAMW advection
leads to a thick layer of high oxygen that persists
even into the tropics (McCarthy and Talley, 1999).
A less dense SAMW is subducted and advected
northwards in the western Indian Ocean (McCartney, 1982). Likewise, the northward circulation of
the densest SAMW in the east, near Chile, provides
a major ventilation signal of high oxygen for the
South Pacific, associated with the low salinity. This
subducted SAMW is the Antarctic Intermediate
Water (AAIW) of the Pacific. SAMW subducted
and advected northward in the western South
Pacific is less dense than the AAIW and provides
an oxygen-rich layer in the western South Pacific.
It is likely that, as with the North Atlantic
SPMW, both hypotheses – of continuous eastward
flow and of substantial equatorward advection –
are correct. Thus the SAMW ventilates the Indian
as SEISAMW and the Pacific as both SAMW and
AAIW through subduction, while at the same time
a portion of the SAMW continues eastward and is
gradually transformed into colder, denser SAMW,
with the final portion flowing through Drake Passage to become the main core of the AAIW in the
Atlantic and Indian Oceans (Talley, 1997). The
SAMW is modified somewhat during this passage,
and then at the confluence of the Falkland and
Brazil Currents enters the South Atlantic and
Indian gyres as a salinity minimum (Atlantic/Indian
Antarctic Intermediate Water; Talley, 1996a).
There is a wide variation in the thickness of
SAMW (reflected in mixed-layer thickness in
Figure 5.4.2, see Plate 5.4.2, p. 428), with the
thickest layers being found in the eastern Indian
Ocean (SEISAMW) and across the South Pacific,
with somewhat shallower layers in the South
Atlantic and western Indian Ocean (Piola and
Georgi, 1982). The transition to thick layers in the
central Indian Ocean is sudden and occurs just east
of Kerguelen Plateau; a cause has not been established. Thompson and Edwards (1981) observed
winter formation of SEISAMW with approximate
properties of 8°C and 34.55 psu, confirming
McCartney’s (1977) finding for SAMW properties in this region. They pointed out, as did
McCartney, that unlike in the South Atlantic,
SAMW in this area cannot contribute locally to
the formation of AAIW.
A major open question regarding SAMW is the
role of northward Ekman transport across the SAF
in maintaining the volume and thickness of the
SAMW. The SAF lies near the maximum westerly
wind stress (and hence zero wind-stress curl), and
so northward Ekman transport is largest there.
Various estimates place the net northward Ekman
transport at about 10 Sv (1 Sv:10
6 m
3 s
91
) around
the circumpolar belt. This is a good fraction of
the net 5 Sv and 14 Sv of SEISAMW and AAIW,
respectively, that move northward into the subtropical gyres (Talley, 1999a for SEISAMW, and
Schmitz, 1995 for AAIW). The remaining subducted transport would then originate in the subtropical gyres. Using a high-resolution numerical
model, Ribbe and Tomczak (1997b) and Ribbe
(1997) examined the effect of northward Ekman
transport of the Antarctic surface water on water
properties of SAMW. They noted that this cold
and fresh water might also drive mid-latitude convection itself to form SAMW. Using air–sea fluxes,
Speer et al. (1997) estimated a formation rate of
25 Sv of SAMW in the Indian Ocean in the density
range 26.5–27.2 , with a peak formation rate at
26.9 ; geostrophic estimates from their inverse
model suggest that half remains in the Indian and
half is exported to the Pacific. This large SAMW
formation rate is reflected in the proxy mixedlayer map of Figure 5.4.2 (see Plate 5.4.2, p. 428)
where an abrupt change in mixed-layer depths is
apparent in the central Indian Ocean.
Antarctic Intermediate Water (AAIW) is the
low-salinity intermediate-depth layer of the southern hemisphere (distribution in Figure 5.4.3b, see
Plate 5.4.3b, p. 428). As noted above, it is very
closely associated with SAMW. AAIW in the
Pacific Ocean originates in the southeast near
Chile, where it is identical with the local SAMW
(McCartney, 1977). Convection in this region
reaches to about 600 m based on oxygen profiles
(Tsuchiya and Talley, 1998). SAMW/AAIW is
subducted northward into the South Pacific subtropical gyre. The resulting low-salinity layer is
apparent throughout the Pacific up to the southern
boundary of the North Pacific’s subtropical gyre.
Part of the convected water near Chile flows
5.4 Mode Waters
383
Hanawa and Talley
west of Chile. On the other hand, McCartney
(1982) showed substantial northward circulation
of the SAMW in each ocean basin. The thick
SAMW in the southeast Indian Ocean is singled
out and called the SEISAMW because of its major
impact on ventilation of the Indian Ocean’s subtropical gyre thermocline through subduction and
northward advection. The SEISAMW advection
leads to a thick layer of high oxygen that persists
even into the tropics (McCarthy and Talley, 1999).
A less dense SAMW is subducted and advected
northwards in the western Indian Ocean (McCartney, 1982). Likewise, the northward circulation of
the densest SAMW in the east, near Chile, provides
a major ventilation signal of high oxygen for the
South Pacific, associated with the low salinity. This
subducted SAMW is the Antarctic Intermediate
Water (AAIW) of the Pacific. SAMW subducted
and advected northward in the western South
Pacific is less dense than the AAIW and provides
an oxygen-rich layer in the western South Pacific.
It is likely that, as with the North Atlantic
SPMW, both hypotheses – of continuous eastward
flow and of substantial equatorward advection –
are correct. Thus the SAMW ventilates the Indian
as SEISAMW and the Pacific as both SAMW and
AAIW through subduction, while at the same time
a portion of the SAMW continues eastward and is
gradually transformed into colder, denser SAMW,
with the final portion flowing through Drake Passage to become the main core of the AAIW in the
Atlantic and Indian Oceans (Talley, 1997). The
SAMW is modified somewhat during this passage,
and then at the confluence of the Falkland and
Brazil Currents enters the South Atlantic and
Indian gyres as a salinity minimum (Atlantic/Indian
Antarctic Intermediate Water; Talley, 1996a).
There is a wide variation in the thickness of
SAMW (reflected in mixed-layer thickness in
Figure 5.4.2, see Plate 5.4.2, p. 428), with the
thickest layers being found in the eastern Indian
Ocean (SEISAMW) and across the South Pacific,
with somewhat shallower layers in the South
Atlantic and western Indian Ocean (Piola and
Georgi, 1982). The transition to thick layers in the
central Indian Ocean is sudden and occurs just east
of Kerguelen Plateau; a cause has not been established. Thompson and Edwards (1981) observed
winter formation of SEISAMW with approximate
properties of 8°C and 34.55 psu, confirming
McCartney’s (1977) finding for SAMW properties in this region. They pointed out, as did
McCartney, that unlike in the South Atlantic,
SAMW in this area cannot contribute locally to
the formation of AAIW.
A major open question regarding SAMW is the
role of northward Ekman transport across the SAF
in maintaining the volume and thickness of the
SAMW. The SAF lies near the maximum westerly
wind stress (and hence zero wind-stress curl), and
so northward Ekman transport is largest there.
Various estimates place the net northward Ekman
transport at about 10 Sv (1 Sv:10
6 m
3 s
91
) around
the circumpolar belt. This is a good fraction of
the net 5 Sv and 14 Sv of SEISAMW and AAIW,
respectively, that move northward into the subtropical gyres (Talley, 1999a for SEISAMW, and
Schmitz, 1995 for AAIW). The remaining subducted transport would then originate in the subtropical gyres. Using a high-resolution numerical
model, Ribbe and Tomczak (1997b) and Ribbe
(1997) examined the effect of northward Ekman
transport of the Antarctic surface water on water
properties of SAMW. They noted that this cold
and fresh water might also drive mid-latitude convection itself to form SAMW. Using air–sea fluxes,
Speer et al. (1997) estimated a formation rate of
25 Sv of SAMW in the Indian Ocean in the density
range 26.5–27.2 , with a peak formation rate at
26.9 ; geostrophic estimates from their inverse
model suggest that half remains in the Indian and
half is exported to the Pacific. This large SAMW
formation rate is reflected in the proxy mixedlayer map of Figure 5.4.2 (see Plate 5.4.2, p. 428)
where an abrupt change in mixed-layer depths is
apparent in the central Indian Ocean.
Antarctic Intermediate Water (AAIW) is the
low-salinity intermediate-depth layer of the southern hemisphere (distribution in Figure 5.4.3b, see
Plate 5.4.3b, p. 428). As noted above, it is very
closely associated with SAMW. AAIW in the
Pacific Ocean originates in the southeast near
Chile, where it is identical with the local SAMW
(McCartney, 1977). Convection in this region
reaches to about 600 m based on oxygen profiles
(Tsuchiya and Talley, 1998). SAMW/AAIW is
subducted northward into the South Pacific subtropical gyre. The resulting low-salinity layer is
apparent throughout the Pacific up to the southern
boundary of the North Pacific’s subtropical gyre.
Part of the convected water near Chile flows
5.4 Mode Waters
383
Hanawa and Talley