(Reid and Nowlin, 1971), has sometimes been
interpreted to mean that the variability at Drake
Passage is entirely barotropic. This is not supported by the ISOS measurements: the range of
observed baroclinic transport in the upper 2500 m
(70–100 Sv) is comparable to the range of absolute
transport (105–140 Sv), and the respective standard
deviations are 5.5 Sv and 8.5 Sv. The barotropic
variability is larger and of higher frequency, but the
baroclinic variability is also significant.
Prior to WOCE, there were no measurements of
either baroclinic or barotropic variability of the
ACC at other locations. Repeats of WOCE section
SR3 south of Australia (140°E) show that the
baroclinic variability there is similar in magnitude
to that measured by dynamic height moorings
deployed for 1 year in Drake Passage during
ISOS. The variability is dominated by changes in
dynamic height at the northern end, as also found
during ISOS. The SR3 section south of Australia
extends both further south (into colder water, with
lower dynamic height) and north (warmer water,
higher dynamic height) than the Drake Passage
section, and so the dynamic height difference and
baroclinic transport is larger at SR3. For example,
the mean transport in Drake Passage above and
relative to 2500 m is 87 Sv (Nowlin and Clifford,
1982; Whitworth, 1983); at SR3, the mean of six
CTD (Conductivity-Temperature-Depth) sections
is 107 Sv (Rintoul and Sokolov, 2000), while
the mean based on 36 summer XBT (expendable
bathythermograph) sections is 109 Sv (Rintoul
et al., 2001). The mean baroclinic transport south
of Australia (relative to a ‘best guess’ reference
level: at the bottom except near the Antarctic margin, where a shallower level is used consistent with
westward flow over the continental slope and rise;
see Rintoul and Sokolov, 2000) is 147<10 Sv
(mean<1 standard deviation), about 13 Sv larger
than the ISOS estimate of absolute transport
through Drake Passage. The transport south
of Australia must be larger than that at Drake
Passage to balance the Indonesian Throughflow,
which is believed to be of O(10 Sv) (Gordon,
Chapter 4.7; Cresswell et al., 1993; Meyers et al.,
1995). However, given the large remaining uncertainty in the barotropic flow at both chokepoints,
the agreement is likely to be fortuitous.
Rintoul and Sokolov’s (2000) estimates of the
baroclinic transport variability south of Tasmania
show that the ACC itself is surprisingly steady
with time (Fig. 4.6.5). Variations in net transport
largely reflect variations in westward flow across
the northern end of the section, rather than
changes in transport of the ACC fronts. Because
the water flowing to the west south of Tasmania is
warm relative to the rest of the section, the
changes in this current branch have a relatively
large impact on the net interbasin exchange of
heat. In other words, while the ACC is undoubtedly the primary means of interbasin exchange,
variations in the transport between the Indian and
Pacific basins are dominated by changes in the
flow north of the ACC.
While our picture of the circumpolar baroclinic
structure of the ACC has become more complete in
recent years, progress in determining the barotropic
flow, and hence improving our estimates of the
absolute transport, has been slower. The WOCE
strategy to determine the transport of the ACC
relied on several elements: repeat hydrographic sections across each of the Southern Ocean ‘chokepoints’, pairs of deep pressure gauges spanning
each chokepoint, and direct velocity measurements
from shipboard and lowered ADCPs (Acoustic
Doppler Current Profilers). Because of the width of
the sections, directly monitoring the absolute transport with a coherent array of traditional moored
instruments is not feasible.
Shipboard and lowered ADCPs combined with
much more accurate navigation and heading
measurements are likely to provide valuable constraints on the barotropic component of the ACC.
For example, Donohue et al. (2000b) use ADCP
observations in the Pacific to infer that the flow at
the bottom beneath the SAF is significant, and in
the same direction as the near-surface flow, thus
enhancing the transport of the SAF over that estimated from the thermal wind alone. At least at the
present time, however, both SADCP (Shipboard
ADCP) and LADCP (Lowered ADCP) measurements are subject to uncertainties that are large
enough to prevent their direct use as a reference
velocity for estimating transports across long
sections (Donohue et al., 2000).
However, it is feasible to monitor a portion of
the current directly, and such a strategy was
adopted at the Australian chokepoint, where a
425-km-long coherent array of current meters,
Inverted Echo Sounders (IESs), and seafloor electrometers (HEMs) was deployed across the
main axis of the ACC to measure both absolute
4.6 The Antarctic Circumpolar Current System
277
Rintoul, Hughes and Olbers
interpreted to mean that the variability at Drake
Passage is entirely barotropic. This is not supported by the ISOS measurements: the range of
observed baroclinic transport in the upper 2500 m
(70–100 Sv) is comparable to the range of absolute
transport (105–140 Sv), and the respective standard
deviations are 5.5 Sv and 8.5 Sv. The barotropic
variability is larger and of higher frequency, but the
baroclinic variability is also significant.
Prior to WOCE, there were no measurements of
either baroclinic or barotropic variability of the
ACC at other locations. Repeats of WOCE section
SR3 south of Australia (140°E) show that the
baroclinic variability there is similar in magnitude
to that measured by dynamic height moorings
deployed for 1 year in Drake Passage during
ISOS. The variability is dominated by changes in
dynamic height at the northern end, as also found
during ISOS. The SR3 section south of Australia
extends both further south (into colder water, with
lower dynamic height) and north (warmer water,
higher dynamic height) than the Drake Passage
section, and so the dynamic height difference and
baroclinic transport is larger at SR3. For example,
the mean transport in Drake Passage above and
relative to 2500 m is 87 Sv (Nowlin and Clifford,
1982; Whitworth, 1983); at SR3, the mean of six
CTD (Conductivity-Temperature-Depth) sections
is 107 Sv (Rintoul and Sokolov, 2000), while
the mean based on 36 summer XBT (expendable
bathythermograph) sections is 109 Sv (Rintoul
et al., 2001). The mean baroclinic transport south
of Australia (relative to a ‘best guess’ reference
level: at the bottom except near the Antarctic margin, where a shallower level is used consistent with
westward flow over the continental slope and rise;
see Rintoul and Sokolov, 2000) is 147<10 Sv
(mean<1 standard deviation), about 13 Sv larger
than the ISOS estimate of absolute transport
through Drake Passage. The transport south
of Australia must be larger than that at Drake
Passage to balance the Indonesian Throughflow,
which is believed to be of O(10 Sv) (Gordon,
Chapter 4.7; Cresswell et al., 1993; Meyers et al.,
1995). However, given the large remaining uncertainty in the barotropic flow at both chokepoints,
the agreement is likely to be fortuitous.
Rintoul and Sokolov’s (2000) estimates of the
baroclinic transport variability south of Tasmania
show that the ACC itself is surprisingly steady
with time (Fig. 4.6.5). Variations in net transport
largely reflect variations in westward flow across
the northern end of the section, rather than
changes in transport of the ACC fronts. Because
the water flowing to the west south of Tasmania is
warm relative to the rest of the section, the
changes in this current branch have a relatively
large impact on the net interbasin exchange of
heat. In other words, while the ACC is undoubtedly the primary means of interbasin exchange,
variations in the transport between the Indian and
Pacific basins are dominated by changes in the
flow north of the ACC.
While our picture of the circumpolar baroclinic
structure of the ACC has become more complete in
recent years, progress in determining the barotropic
flow, and hence improving our estimates of the
absolute transport, has been slower. The WOCE
strategy to determine the transport of the ACC
relied on several elements: repeat hydrographic sections across each of the Southern Ocean ‘chokepoints’, pairs of deep pressure gauges spanning
each chokepoint, and direct velocity measurements
from shipboard and lowered ADCPs (Acoustic
Doppler Current Profilers). Because of the width of
the sections, directly monitoring the absolute transport with a coherent array of traditional moored
instruments is not feasible.
Shipboard and lowered ADCPs combined with
much more accurate navigation and heading
measurements are likely to provide valuable constraints on the barotropic component of the ACC.
For example, Donohue et al. (2000b) use ADCP
observations in the Pacific to infer that the flow at
the bottom beneath the SAF is significant, and in
the same direction as the near-surface flow, thus
enhancing the transport of the SAF over that estimated from the thermal wind alone. At least at the
present time, however, both SADCP (Shipboard
ADCP) and LADCP (Lowered ADCP) measurements are subject to uncertainties that are large
enough to prevent their direct use as a reference
velocity for estimating transports across long
sections (Donohue et al., 2000).
However, it is feasible to monitor a portion of
the current directly, and such a strategy was
adopted at the Australian chokepoint, where a
425-km-long coherent array of current meters,
Inverted Echo Sounders (IESs), and seafloor electrometers (HEMs) was deployed across the
main axis of the ACC to measure both absolute
4.6 The Antarctic Circumpolar Current System
277
Rintoul, Hughes and Olbers