sign in sign up
<<
Home , Southern Ocean

SOUTHERN - CIRCUMPOLAR CURRENT 1

SOUTHERN OCEAN-ANTARCTIC CIRCUMPOLAR CURRENT

characteristics of the water, not land boundaries. John Klinck, Old Dominion University, Center for The northern edge of the Southern Ocean is marked Coastal , Crittenton Hall, by the Subtropical Front (Figure 1), where water Norfolk, VA 23529, USA near the surface changes from warm and salty W. D. Nowlin Jr., Department of Oceanography, (123C, 35.0; characteristic of lower ) to Texas A&M University, College Station, TX, USA cold and fresh (103C, 34.6; characteristic of polar Copyright ^ 2001 Academic Press latitudes). The southern edge of the Southern Ocean is marked by . doi:10.1006/rwos.2001.0370 Within the Southern Ocean is a large eastward 0004 Sowing current, the Antarctic Circumpolar Current 0001 The Antarctic Circumpolar Current (ACC) Sows (ACC), which Sows unbroken around the globe. eastward around the globe in the Southern Ocean, S driven by the strong eastward winds characteristic The narrowest constriction to this ow is Drake of southern polar latitudes. Direct and indirect Passage (about 700 km across) at the southern tip of . By convention, any Sow through measurements of the total transport of this current S are consistent with the idea that average winds drive this passage is part of the ACC. The eastward ow the average Sow. However, abrupt changes in trans- associated with the Front marks the port do not correspond to changes in local winds, northern boundary of the ACC. The southern nor do changes occur consistently around the globe. boundary of the ACC is less dramatic, being re- cently deRned (middle 1990s) based on water prop- The path of the ACC is controlled by ocean depth R through the tendency for large-scale ocean Sow to erties (speci cally, the surfacing of water originating in the north Atlantic). South of the southern bound- be along lines of constant planetary vorticity (Co- R riolis parameter divided by depth). is ary of the ACC are polar gyres lling the Weddell the narrowest constriction to this Sow (about and Ross , which are not part of the ACC. 700 km in width). Strong current extends through- Drake Passage opened about 30 million years ago 0005 out the related to an upward tilt of as South America and Antarctica separated allowing the ACC to form. This current formation is thought the constant density, temperature, and surfa- S ces to the south. The strongly tilted property surfa- to have a profound in uence on Antarctica and ces in the ACC allow deep (3}4 km deep) water, global as there was a simultaneous accumu- originating in the polar North Atlantic, to reach the lation of ice over Antarctica and a global decline in surface where it is driven northward by the winds, level. The ACC also isolated Antarctica biolo- thus completing the circuit. The ACC is composed gically allowing a unique marine to of three circumpolar, frontal jets, each having about evolve. three times the speed of the Sow between the fronts. An early numerical model study of the dynamical 0006 Dynamic instability of these jets creates eddies effect of Drake Passage found that as the Passage (about 150 km in diameter) which redistribute mo- deepened export of dense Antarctic water declined mentum and water properties. and the speed of the ACC increased. A more recent study using a more realistic global ocean model conRrmed these inSuences of an open passage, and S Introduction showed that reduced out ow of Antarctic dense water increased deep-water formation in the North 0002 The Southern Ocean, that part of the global ocean Atlantic. covering the higher latitudes of the southern hemi- The ACC was discovered by European mariners 0007 sphere, is unique in being continuous around the in the late seventeenth century with the Rrst re- globe. This allows exchange of mass, heat, fresh ported crossing by on the HMS water, carbon, and other properties, including living Paramore (1699}1700). After this time, a number of material, among the three major : Atlantic, mariners explored this for the purposes of Indian, and PaciRc. commerce and science. Notable explorers were 0003 The speciRc boundary between the Southern (1772}1775), Thaddeus Bellingshausen Ocean and the rest of the global ocean depends on (1819}1821), and (1839}1843).

RWOS 0370 Editor: Indira Operator: Indira Scan: Ramakrishna 2 SOUTHERN OCEAN-ANTARCTIC CIRCUMPOLAR CURRENT

Many unnamed sailors came looking for seals and and form drag (pressure acting on bottom depth whales and so kept their knowledge secret. The late variations) retards the deep Sow. Alternative force nineteenth and early twentieth centuries saw an ex- balances that balance the wind with friction (either plosion of oceanographic research by various coun- bottom friction or horizontal friction against the tries which constituted the Rrst large-scale survey of continental margins) are not plausible due to the the Southern Ocean, providing measurements of enormous level of turbulence that would be required water properties and current speeds in the ACC. to provide the retarding force. Such turbulence would not allow the observed water properties to Large-Scale Structure exist. The force balance in the ACC is different from 0012 0008 The ACC is 23 000 km long (at 553S) and up to that in the rest of the ocean, which is also driven 2000 km wide (about 203 of ) in some re- largely by surface wind stress. In most of the ocean, gions away from Drake Passage (Figure 1). The large-scale pressure gradients, due to land bound- general eastward Sow of the ACC is strongest near aries, balance the effect of the wind stress. In the the surface with speeds between 0.25 and 0.4m s\1. Southern Ocean, however, Drake Passage is a wide Unlike lower latitude currents, such as the Gulf region that is not blocked by land, so no large-scale, Stream, ACC currents extend throughout the water east}west, sea surface tilt can develop. However, the column, declining monotonically with depth to wide continental shelves of Drake Passage and the a few centimeters per second near the bottom submarine ridges east of the Passage close off essen- (2.5 km or greater). Transport of water by the ACC tially all of the pathways beneath 1 km along each is 100}150 ;106 m3 s\1, several times larger than latitude. Therefore, only the near-surface Sow is other strong currents in the ocean (such as the Gulf unblocked and deep pressure gradients may develop. Stream or Kuroshio). In spite of this early explanation, dynamicists dis- 0013 0009 Distributions of potential temperature, salinity, agree about the details. One issue is the mechanism and density across the ACC (Figure 2) reveal the by which the momentum added by the wind near characteristic southward upward tilt of surfaces on the surface is transferred to the deep ACC where it which property values are nearly uniform. The tilt is can be removed by the ocean bottom. Transient so strong that the salinity maximum and the 23C mesoscale eddies (about 150 km in diameter) and isotherm found 3 km deep on the north side of the Rxed pressure distributions (standing eddies) each ACC are within a few hundred meters of the surface transfer momentum towards the bottom. A second on the south side. This tilt is a consequence of the issue involves the importance of structure of the Sow dynamics and is a reason for the importance of wind (the wind stress curl) which differentially the ACC in global transport of heat and other prop- transfers near-surface water equatorward and pro- erties. Tilting density surfaces create horizontal duces part of the vertical overturning in the South- pressure changes which, coupled with the Coriolis ern Ocean. It is argued that, similar to other ocean force, drive the eastward Sow in the ACC. More basins, the wind curl drives the ocean (called the speciRcally, the north to south increase in density at Sverdrup Balance) and leads to good estimates of every depth (Figure 2C) produces a vertical change ACC location and total transport. Bottom topogra- in the Sow, which is slower with depth, that is phy in this view allows deep circulation across the responsible for the monotonic decline in the speed ACC. However, another argument is that this view of the ACC with depth (Figure 2D). misses the momentum balance and ignores the effect 0010 ACC Sow obtained from density changes by geos- of water pressure pushing on the solid thereby trophic estimates, with the choice of no Sow near retarding the ocean Sow. the bottom, compare well to direct current measure- In either view, the circulation in the Southern 0014 ments in Drake Passage as well as to surface Sow Ocean responds strongly to variations in bottom speed measured by surface drifters. This good com- depth, unlike lower latitude oceanic . The parison validates the estimates of the slope of the near surface Sow (Figure 3) is along submarine ocean surface (Figure 3) based on all density obser- ridges and tends to cross ridges at gaps (for vations. From these estimates, as well as Sow esti- example, in the south-western PaciRc between mates from surface drifters and the surface slope 1403W and 1603E). Because of the rotation of the measured from satellites, it is clear that the ACC earth, a column of water that stretches will spin Sows continuously around the Southern Ocean. faster in the direction of the Earth’s spin (counter- 0011 The basic force balance for the ACC was Rrst clockwise looking down on the ocean in the south- identiRed in the middle of the twentieth century. ern hemisphere). To avoid these changes, ocean Sow Wind stress accelerates the water near the surface tends to be along lines of constant depth (speciR-

RWOS 0370 Editor: Indira Operator: Indira Scan: Ramakrishna SOUTHERN OCEAN-ANTARCTIC CIRCUMPOLAR CURRENT 3

cally, along lines of constant planetary vorticity ACC fronts (Figures 1 and 2). The water density (PIV) deRned as the Coriolis parameter divided by increases southward across each front, creating depth). a pressure gradient that is balanced by the from the strong eastward Sow. Stronger gradi- ents in the fronts give rise to stronger currents, so Role In Global Thermohaline the fronts are associated with jets (high speed, nar- Circulation row currents). Each of these fronts is observed in every section across the ACC independent of longi- 0015 The circulation of the global ocean is driven by two tude. In some places, however, they are so close basic mechanisms: surface wind stress and dense together that they form a single front. Downstream, water creation. Surface cooling and ice formation at the fronts again separate with no indicated change high latitudes (in the , Norwegian, and in characteristics. Labrador seas and in the Weddell and Ross Seas) The Subantarctic Front (SAF) occurs where cold 0018 creates dense water which sinks into the deep ocean. near the surface is denser than water to These waters move along a variety of paths, mix the north and thus descends into the interior, cre- with surrounding water, become less dense, and ating Antarctic Intermediate Water indicated by eventually return to the surface. The ACC partici- a distinct minimum in salinity at 400 m just north of pates in this circulation by mixing water among the the front (Figure 2B). Additional indicators are major ocean basins and by providing a path for a northward temperature increase (warmer than dense water (North Atlantic Deep Water) to return 43C) at 400 m and a southward increase in near- to the surface (along the tilting density surfaces seen 1 surface dissolved oxygen (to greater than 7 ml l\ ). in Figure 2C, as shown schematically in Figure 4). The Polar Front (PF) is located where cold near- Less extreme surface cooling over and north of the surface water (Antarctic Surface Water) moves ACC creates water that is driven northward by the northward and sinks beneath less-dense surface surface winds and pushed below the surface to in- waters. The minimum temperature above 200 m is termediate depths (1 km or so) driving additional colder than 23C south of the PF (Figure 2A). The circulation. Cold water near the surface in the depth of the temperature minimum increases from southern ACC is driven north by the wind and 125 m or less to greater than 200 m in crossing the eventually sinks under the warmer water to the front from south to north. The warm Circumpolar north creating Antarctic Intermediate Water (Figure Deep Water is absorbed into the surface mixed layer 4). at the Southern ACC Front (SACCF). North of the 0016 This secondary, north}south circulation is weak SACCF, the temperature maximum below 500 m is compared to the stronger eastward Sow of the ACC, warmer than 1.83C, whereas the deeper salinity but is clearly indicated in the diagram (Figure 4) maximum is above 34.73 (Figure 2A and B). A dis- showing water masses and bounding density surfa- tinct dissolved oxygen minimum is observed below ces in vertical section across Drake Passage. It plays 500 m and the value at the minimum is below an important role in the southward transport of 1 4.2 ml l\ north of the SACCF. heat and salt needed to balance the heat lost to the The processes responsible for the ACC fronts may 0019 atmosphere and the freshening due to precipitation be considered from two points of view. The water and ice melting. It also distributes carbon dioxide mass point of view focuses on the secondary and other trace gases entering the ocean at the (north}south) circulation of the ACC driven north- surface throughout the interior ocean. Deep water ward near the surface by winds and southward at that rises near the surface provides a source of depth by density differences. The dynamic point of nutrients that are used by which view recognizes that the ACC is unstable to distur- form the basis of the Southern Ocean ecosystem. bances in which mesoscale eddies are ejected from the frontal jets by baroclinic instability. The result is Circumpolar Fronts that momentum is transferred to the frontal jets, increasing their speed and increasing the tilt of the 0017 Ocean fronts are locations at which water proper- density surfaces. Bottom topography inSuences this ties change over short distances (a few tens of km), process either by creating the disturbances that grow and are usually associated with boundaries between into eddies or by changing the way in which the different types of water. The ACC is known to have eddies interact. These points of view are not incon- three distinct fronts that are continuous around the sistent because the acceleration of the ACC jets by Southern Ocean. These fronts are labeled, from eddies tilts the background density, thus increasing north to south, Subantarctic, Polar and Southern its horizontal gradient. The location at which the

RWOS 0370 Editor: Indira Operator: Indira Scan: Ramakrishna 4 SOUTHERN OCEAN-ANTARCTIC CIRCUMPOLAR CURRENT

water is subducted is controlled by the density of dinally coherent changes in the ACC Sow. the surface water and the location of the matching An analysis of dynamics in a longitudinally un- 0024 interior density surface. The eddy effect collects bounded ocean, combined with results from two more density surfaces in a given place making it the different realistic ocean models representing the likely location of subduction. Southern Ocean provide an explanation for the character of variability of the ACC. Transport vari- Circulation Variability ations with periods of 1}8 months occur by changes in the Sow that is independent of depth. However, 0020 Estimates of the total transport of the ACC based these changes follow lines of constant planetary vor- on density measurements across Drake Passage, ticity (PlV), creating a global free circulation path. with the choice of zero Sow at the bottom, are More correctly, Sow follows PlV lines except in relatively constant at about 90 Sv (1 Sv is 1 million a few places, like Drake Passage, where the Sow m3 s\1). This method only estimates the transport jumps from one PlV line to a nearby line of the associated with those currents that change with same PlV value. Wind stress variations parallel to depth. An unknown, and potentially large, depth- PlV lines drive transport variations in the ACC. independent Sow must also be estimated to obtain Wind stress was strongly correlated with trans- 0025 the total transport. port changes in two realistic ocean models and with 0021 The total transport of the ACC was Rrst estimated measurements of bottom pressure. Wind stress at Drake Passage in the late 1970s and early 1980s along PlV lines was most strongly correlated with using measurements of density, current, and bottom model changes, but various area and longitudinal pressure over time. The bottom pressure difference averages of wind stress or curl of the wind stress across the passage was related to total transport (tendency of the wind to spin the ocean) were also thereby extending the transport time series from signiRcantly correlated with transport changes. 1977 to 1982 (with 1980 missing). The average However, pressure variations due to density changes transport over the cross-sectional area westward may mask any relationship between wind and pres- during this four-year record is 123$10 Sv, only sure, making it difRcult to measure the total trans- slightly different from the estimated value over the port of the ACC with bottom pressure best 14 months with direct current estimates (Figure measurements. Furthermore, Suctuations of the 5). To this should be added the estimate of 11 Sv ACC transport were not coherent around the ACC transport through the nonmeasured part of the for periods shorter than semiannnual. Drake Passage section, giving a total of 134 Sv. ACC variability also occurs through the creation 0026 Most transport variation is due to changes in the of mesoscale eddies (see ⅷⅷⅷ), in which a strong slope of the free surface while the internal density current wraps around itself due to unstable lateral structure remains relatively constant and supports meandering of ocean jets. The average kinetic en- 70% of the total transport. Observed changes in net ergy due to eddies is higher downstream (east) of sea surface slope across the passage result in 20}40 large topographic features that block the Sow Sv changes of the ACC transport over periods of (Drake Passage, Kerguelen Plateau). Locations of weeks to months. the PF for a period of 7 years were analyzed, reveal- 0022 A semiannual cycle in the transport is driven by ing an average frontal width of around 45 km and a semiannual change in the wind stress, with two a displacement from the average position of main transport peaks per year (mostly spring and 120}150km. In some places, the bottom topogra- fall) but with lesser peaks in many months (Figure phy limits the variability, basically requiring the 5). Lunar monthly and fortnightly tides in Drake ACC to Sow through certain narrow gaps. Thus, Passage create small variations in transport and eddies are an integral part of the dynamics of the have a weak inSuence on the surface tilt. ACC in providing a mechanism to narrow and ac- 0023 The applicability of the Drake Passage transport celerate frontal jets and acting as a mixing mecha- studies to other parts of the ACC has not been nism to redistribute water properties, including demonstrated. The presumption has been that the momentum. ACC changes speed globally as the average strength At the end of the twentieth century, a number of 0027 of the Southern Ocean winds changes. Variations of laboratories in various countries were creating real- sea surface height, measured from satellites, are less istic numerical models of global ocean circulation related with increasing distance between measure- for the purpose of testing ideas of ocean dynamics ments or increasing time difference. Furthermore, and to answer questions about the role of the ocean global changes in the ACC were not found, making in the global climate. These models of necessity it unlikely that global wind variations drive longitu- contain the Southern Ocean and are largely success-

RWOS 0370 Editor: Indira Operator: Indira Scan: Ramakrishna SOUTHERN OCEAN-ANTARCTIC CIRCUMPOLAR CURRENT 5

ful in creating an ACC-like Sow in the observed Gordon AL, Molinelli E and Baker T (1978) Large-scale location. However, all of these simulations overesti- relative dynamic topography of the Southern Ocean. mate the ACC transport by about 50%. Those mod- Journal of Geophysical Research 83(C): 3023}3032. els which allow mesoscale eddies to occur (although Hofmann EE (1985) The large-scale horizontal structure the dynamics of these small features are only mar- of the Antarctic Circumpolar Current from FGGE drifters. Journal of Geophysical Research 90(C4): ginally represented) have higher eddy energy in the 7087}7097. proper places, compared to satellite measurements, Hughes CW, Meredith MP and Heywood KJ (1999) but at magnitudes that are low by a factor of two or Wind-driven transport Suctuations through Drake Pas- more. sage: a southern mode. Journal of Physical Oceanogra- phy 29(8 part 2): 1971}1992. Mikolajewicz U, Maier-Reimer E, Crowley TJ and Kim Conclusions K-Y (1993) Effect of Drake and Panamanian gateways S on the circulation of an ocean model. Paleoceanogra- 0028 The ACC is a large, variable ow around the South- phy 8(4): 409}426. ern Ocean connecting the three major ocean basins. Moore JK, Abbott MR and Richman JG (1999) Location This current exchanges properties in the global and dynamics of the Antarctic Polar Front from satel- ocean at the same time that it isolates Antarctica. It, lite sea surface temperature data. Journal of Geophysi- thus, plays an integral role in the vertical overturn- cal Research 104(C2): 3059}3073. ing of the global ocean. Munk WH and Palmen E (1951) Note on the dynamics of 0029 The large-scale structure of the ACC is clear in the Antarctic Circumpolar Current. Tellus 3: 53}55. current observations and the details of the dynam- Nowlin Jr, WD and Klinck JM (1986) The physics of the ical balance are largely understood, with a few dis- Antarctic Circumpolar Current. Review of putes lingering. The time variability of the ACC is and Space Physics 24: 469}491. Olbers D (1998) Comments on ‘On the obscurantist phys- not well observed nor is the spatial structure of ics of ‘form drag’ in theorizing about the Circumpolar large-scale change known. Observation programs Current’. Journal of Physical Oceanography, 28(8): and numerical model studies are addressing these 1647}1654. issues of variability. Orsi AH, Whitworth T III and Nowlin WD Jr (1995) On the meridional extent and fronts of the Antarctic Cir- See also cumpolar Current. Deep-Sea Research I 42(5): 641}673. 0030 Circulation (371). Mesoscale Eddies Peterson RG (1988) On the transport of the Antarctic (143). Wind Driven Circulation (110). Wind and Circumpolar Current through Drake Passage and its Buoyancy-forced Upper Ocean (157). Thermoha- relation to wind. Journal of Geophysical Research line Circulation (111). 158. 93(C11): 13 993}14 004. Semtner AJ Jr and Chervin RM (1992) Ocean general circulation from a global eddy-resolving model. Jour- Further Reading nal of Geophysical Research 97: 5493}5550. Sievers HA and Nowlin WD Jr (1984) The stratiRcation Bakerl DJ Jr (1982) A note on Sverdrup balance in the and water masses at Drake Passage. Journal of Geo- Southern Ocean. Journal of Marine Research physical Research 89: 10 489}10415. 40(suppl): 21}26. Sinha B and Richards KJ (1999) Jet structure and scaling Deacon G (1984) The Antartic Circumpolar Ocean. Cam- in southern ocea models. Journal of Physical Oceano- bridge: Cambridge University Press. graphy 29(6) 1143}1155. Fogg GE (1992) A History of Antarctic Science. Cam- Tomczak M and Godfrey JS (1994) Regional Oceano- bridge: Cambridge University Press. graphy: An Introduction. Pergamon Press. Group (1991) An eddy-resolving model of the Treguier AM and Panetta RL (1994) Multiple zonal jets Southern Ocean. EOS, Transactions of the American in a quasigeostrophic model of the Antarctic Circum- Geophysical Unions 72(17): 169}175. polar Current. Journal of Physical Oceanography Gill AE and Bryan K (1971) Effects of geometry on the 24(11): 2263}2277. circulation of a three-dimensional van Loon H (1972) Half-yearly oscillations in the Drake ocean model. Deep-Sea Research 18: 685}721. Passage. Deep-Sea Research 19: 525}527. Gille ST (1994) Mean sea surface height of the Antarctic Warren BA, LaCasce JH and Robbins PE (1996) On the Circumpolar Current from GEOSAT data; method and Obscurantist physics of ‘form drag’ in theorizing about application. Journal of Geophysical Research 99(C9): the circumpolar current. Journal of Physical Oceano- 18 225}18 273. graphy 26(10): 2297}230l. Gille ST and Kelly KA (1996) Scales of spatial and tem- Whitworth T III (1983) Monitoring the transport of the poral variability in the Southern Ocean. Journals of Antarctic Circumpolar Current at Drake Passage. Jour- Geophysical Research 101(C4): 8759}8773. nal of Physical Oceanography 13: 2045}2057.

RWOS 0370 Editor: Indira Operator: Indira Scan: Ramakrishna 6 SOUTHERN OCEAN-ANTARCTIC CIRCUMPOLAR CURRENT

Whitworth T III and Peterson RG (1985) Volume tr nsport of the Antarctic Circumpolar Current from bot- tom pressure measurements. Journal of Physical Oceanography 15(6): 810}816. Whitworth T III and Nowlin WD Jr (1987) Water masses and currents of the Southern Ocean at the Greenwich Meridian. Journal of Geophysical Research 92(C6): 6462}6476.

RWOS 0370 Editor: Indira Operator: Indira Scan: Ramakrishna SOUTHERN OCEAN-ANTARCTIC CIRCUMPOLAR CURRENT 7

0°

40°E

40°W

40°S 80°E 50°S

0°W 70°S 8

120°W

120°E

160°W 160°E a0370fig0001 Figure 1 Polar view of Southern Ocean with climatological locations of circumpolar fronts. From north to south, these fronts are Subtropical Front, Subantarctic Front, Polar Front, and Southern ACC Front. The southern boundary of the ACC also is shown. Black outlines show land. Gray shading indicates regions where water is shallower than 3000 m. (Courtesy of Alex Orsi.)

RWOS 0370 Editor: Indira Operator: Indira Scan: Ramakrishna a0370fig0002 oun r etadfo.(oiidfo htot n Now- Shaded and pair. Whitworth 1987.) station from lin, each (Modified for flow. depth westward deepest are s the columns the (cm at flow flow along geostrophic zero (D) ing or ACC and maxima ( density; temperature indicate the potential potential o (C) (A) across or property: x a m of figure 3000 minima each to On Meridian. relative Greenwich flow trophic 2 Figure 8 45 STF 49 64 SAF 70 PF 75 80 45 STF 49 64 SAF 70 PF 75 80 45STF 49 64 SAF 70 PF 75 80 45 49 64 SAF 70 PF 75 80

0 0 34.0 0 0 0 20 2 10 CURRENT CIRCUMPOLAR OCEAN-ANTARCTIC SOUTHERN 8 10 3 3 2 12 35 26.6 10 6 34.6 27.0 3 10 27.4 STF 20 5 34.6 34.2 27.2 15 4 2 5 34.4 27 5 2 27.6 34.3 5

etoso eprtr,slnt,dniy n geos- and density, salinity, temperature, of Sections 27.7 3 34.25 15 10 2.0 1.6 27 3 3 3 1 2.8 1 34.3 1 1 2.4 2.2 1 2 3 3 34.4 3 36.7 2 10 1.2 2 2.6 34.7 36.9 1 5 1 2 5 0.8 34.6 68 36.8 5 2 2.4 .9 1 5 34.7 36 2 1 0.74 2 2 2 7.0 2

WS07 dtr niaOeao:Idr cn Ramakrishna Scan: Indira Operator: Indira Editor: 0370 RWOS 3 3 34.8 0.76 37.125 3 1 3 0.82 1 2 2.0 37.0 2.2 37.1 3 1 1.2 3 3 3 0.84 0.72 3 37.05 3 2 0.8 2 2 km 2.0 km km km 0.4 2 0.82 1 1.6 34.8 34.7 1 1 0.3 46.02

1 MID-OCEAN BRIDGE 1.2 0.78 45.95 46.0 1 MID-OCEAN BRIDGE

4 4 4 45.95 MID-OCEAN BRIDGE 4 0.8

MID-OCEAN BRIDGE 0.74 0.7 45.98 SEAMOUNT

DISCOVERY SEAMOUNT 45.99 3 DISCOVERY SEAMOUNT DISCOVERY SEAMOUNT ) B salinity; (B) C); 5 5 5 5 \ 1 assum- ) CAPE BASIN CAPE BASIN CAPE BASIN h S r CAPE BASIN 40°S 50° 40°S 50° 40°S 50° 40°S 50° 6 6 6 6 0km 1000 2000 0km 1000 2000 0km 1000 2000 0km 1000 2000 (A) (B) (C) (D) SOUTHERN OCEAN-ANTARCTIC CIRCUMPOLAR CURRENT 9

0°

1.2 1.3 1.3 1.1 1.2 1.3 40°E 1.0 1.2 40°W 1 1.5 1.7 1.2 1.6 1.5 1.4 0.8 0.7 1.3 0.9 0.5 0.8 1.1 0.5 1.4 1 0.4 35 0.9 0.4 0.6

0.3 1 0.3 0.35 0.7 1.2 80°E

1.1 0.5 1.3 80°W 0.9 0.3 1 35 1.2 0.7 0.3 0.5 1.1 1.3 0.8 1.1 0.4 0.8 0.8 1.3 0.6 0.4 0.35 0.6 0.3 1.2 1 0.5

0.4 0.6 0.4 1.2 0.9 120°W 0.7 0.8 1 1.1 1.1 120°E

1.2 1.3 1.2 1.3 1.2 1.3 1.6

1.4 160°W 160°E a0370fig0003 Figure 3 Dynamic topography of the 50-m depth surface relative to 1000 m. The surface slope, with the choice on no pressure gradient at 1000 m, is estimated from measurements of water density. Curves represent contours of equal surface height (m) and approximate the streamlines of surface currents. Black outlines show land. Gray shading indicates areas with depths less than 3000 m. (Courtesy of Alex Orsi.)

RWOS 0370 Editor: Indira Operator: Indira Scan: Ramakrishna 10 SOUTHERN OCEAN-ANTARCTIC CIRCUMPOLAR CURRENT

SAF PF SACCF Subantarctic zone Polar frontal zone Antarctic zone Subantarctic Surface Water Antarctic Water Surface Water Subantarctic Mode Water mediate

Antarctic Inter Water

Water Upper Circumpolar Deep

Water Lower Circumpolar Deep

Pacific Deep

Deep South-east Weddell Sea Water

a0370fig0004 Figure 4 Diagram showing water masses and bounding den- sity surfaces in the Antarctic Circumpolar Current (ACC) at Drake Passage. Arrows indicate the direction of secondary flow along density surfaces across the ACC. Approximate locations of fronts are indicated. (Modified after Sievers and Nowlin, 1984.) SAF, Subantarctic Front; PF, Polar Front; SACCF, Southern ACC Front.

RWOS 0370 Editor: Indira Operator: Indira Scan: Ramakrishna 11

81_82 July 78

79_80 77

160

150

1 140 _

63 130

120 SOUTHERN OCEAN-ANTARCTIC CIRCUMPOLAR CURRENT

Transport (10Transport m110 s )

100

90 Jan July Jan July Jan July Jan Jan July Jan 1982) of volume transport through Drake Passage estimated from direct measurements of current,

1977 1978 1979 1980 1981 1982 } RWOS 0370 Editor: Indira Operator: Indira Scan: Ramakrishna A time series (1977

Figure 5 density, and bottomPeterson, pressure. 1985.) The insert allows easy comparison of transport variation in different years. (After Whitworth and a0370fig0005