JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 89, NO. C1, PAGES 641-648, JANUARY 20, 1984
Thermohaline
Stratification
Below the Southern
Ocean Sea Ice
ARNOLD L. GORDON AND BRUCE A. HUBER
Lamont-DohertyGeologicalObservatory
The end of winter stratificationwithin the cold cyclonictrough of the Weddell gyre near 60øSbetween
5øE and the Greenwichmeridian is resolvedwith the Mikhail Somovdata set.The temperaturemaximum
of the Weddell Deep Water (WDW) is, for the most part, lessthan 0.5øC,but warmer cellsof WDW are
found. These warm WDW cellshave temperature,salinity,and oxygenpropertiessimilar to the WDW
characteristicof the Weddell gyre inflow, which is situatedto the southeastof the Mikhail Somovstudy
region.The warm WDW cellsare accompaniedby domesin the pycnoclineof 40 m amplitudeover the
surroundingpycnocline,while deeperisopycnalsare depressed.Anticyclonicshear below the 27.83 a-0
isopycnalwithin the warm WDW cells is compensatedby the cyclonic shear associatedwith the pycnocline dome. The pycnoclinedomes are exposedto about 50% greater entrainment by the turbulently
active winter mixed layer, relative to the regionalentrainmentrate. This entrainmentcan significantly
erode the warm cells in a singlewinter season,introducingexcessheat and salt into the mixed layer.
While the heat is lost to the atmosphere,the excesssalt is not necessarily
compensated
by increasedfresh
water introduction.It is hypothesizedthat the warm WDW cells within the Weddell gyre trough are
derivedfrom instabilitywithin the frontal zone which extendsfrom Maud Rise to the northeast,separating the Weddell warm regimefrom the cold regime.Greater than normal injectionof warm WDW cells
into the Weddell gyre trough would increasethe surfacesalinity, which would tend to destabilizethe
pycnocline,increasingthe probabilityof deepconvectionand polynyaevents.
1.
INTRODUCTION
During October and November 1981 the joint U.S.-U.S.S.R.
Weddell Polynya Expedition was carried out within the
SouthernOcean seaice, aboard the Sovietship Mikhail $omov
1979' Gordonet al., 1981' Gordonand MolineIll,1982,plates
231-233].This?egio
n is characterized
by a relativelycold(less
than 0.5øC)Weddell Deep Water (WDW) and is here designa-
ted astheWeddellcol•.regime(Figure2). The northernsta-
of the Arctic-Antarctic Research Institute of Leningrad tions fall withinihe eastward flowing limb of the Weddell
[Gordon and Sarukhanyan,1982; Gordon, 1982]. The objec- gyre,with stations36a0d 37 positioned
northof the northern
tives of the expeditionwere to obtain a comprehensiveinter- boundary
of theWedd•11
gyre.Theboundary
is markedby a
disciplinarydata set near the Greenwichmeridian (0øE),well slightincrease
in barOclinicity
and a transitioninto deepwater
wRhin the seasonalsea ice zone and to investigatean active far toowarm(wellin excess
of 1øCat thetemperature
maxiopen oceanpolynya. A polynya developednear 65øSand 0øE mum' Figure 3) to have participatedin the upstreamsegment
in the 1974-1976 winters [Carsey, 1980]; however,no poly- of the Weddellgyre.This warmerdeepwater is calledCircum-
nyadeveloped
in 1981.Theobservations
extendacross
theice polarDeepWater(CDW).
edgezone into the interior pack about 590 km from the mean
positionof the outer fringesof seaice during the expedition.
The Mikhail Somovhydrographicdata set resolvesthe thermohaline stratification during the transition phase from
waxing to waning of the Southern Ocean sea ice cover. Sea ice
covers the region near 60øS between 0¸ and 5øE in June,
reachinga maximum northward extent during Septemberand
October. The region becomesice free in mid-December (Antarctic Ice Charts, Navy-NOAA Joint Ice Center, Naval Polar
Oceans Center, Suitland, Maryland). In 1981 the ice edge
overtook the region between June 18 and 25, remained near
56øSfrom the end of July to early November, and retreated
from the region during the third week of December.The Mikhail Somov data set reveals the cumulative
effects of the 1981
South of the Mikhail Somovdata set, historical (summer)
data reveal the westward flowing limb of the Weddell gyre,
with warmer WDW (temperaturemaximum of up to 1.1øC),
here designatedthe Weddell warm regime.The warmer WDW
is derived from the main mass of CDW in the vicinity of
20ø-30øE [Deacon, 1979]. It appearsas a wedge of warm
WDW with its apex near Maud Rise (Figure 2). A rather
abrupt transition to the Weddell cold regime occursto the
west of Maud Rise (Figure 4) and extendsto the northeast
from Maud Rise. The transition from cold to warm regimes
occurswithin a distanceof 100 km (e.g.,betweenlslas Orcadas
stations95 and 96; Figure 5). The increasein WDW temperature of over 0.4øCis accompaniedby a decreasein depth of
the temperaturemaximumfrom 425 m to 240 m and shallowing of the pycnoclineby approximately50 m. An lslas Or-
austral winter and thus depictsthe end of winter stratification,
althoughthe expeditiontook placein the austral spring.
cadassection
along20øE(Figure.
6) shows
a similartransition
The positions of the CTD-O2/Rosette hydrographic sta- occuringbetweenstations75 and76.
tions and the expendablebathythermograph(XBT) observaThus, there is a relatively sharp transition from warm to
tions [Huber et al., 1983], as well as ice core sites [Ackley et cold regime coincidingwith southwestflow within the Wedal., 1982] and the ice edge position at the time of entry and dell gyre(Figure 2). It is markedby the 0.6¸ to 0.8øCisotherms
of the temperaturemaximum, extendingfrom 58øS20øE to
exitfromthepac•k
icearegivenasFigure1.
63øS2øE.The historicaldata setsuggests
that this is a climatic
2. OCEANOGRAPHICSETTING
feature(e.g.,see350 m temperaturesectionplate 23 of Gordon
Most of the Mikhail Sotnov hydrographic stations fall and MolineIll [1982]).
within the trough of the cyclonicWeddell gyre [Deacon,1976,
The primary advectivepath of warm WDW into the western hemisphereoccurs south of Maud Rise, over the contiCopyright1984by the AmericanGeophysicalUnion.
nental slopewhere a relatively strongwestwardcurrent exists
(Figure 2) ITchernia, 1977; Foster and Carmack,1976]. OutPaper number 3C1642.
sidethe:boundary
currents
theWeddell
gyreis'sluggish
with
0148-0227/84/003C- 1642505.00
641
642
GORDON AND HUBER' STRATIFICATION BELOW SOUTHERN OCEAN SEA ICE
IOøW
5ø
0ø
5ø
I0 øE
cumpolar
DeepWater(CDW)occurs
near56ø45'S
Onboth
XBT
55øS -
-55os
sections.
The anomalouslycold WDW observedat station 35 (temperaturemaximumof only 0.19øCat 600 m) and at XBT 103,
109, and 113 surroundsthe CDW eddy.This cold deepwater
is probablyan eastwardextensionof the Weddell-Scotiaconfluence[Pattersonand Sievers,1980].
3.
60øS
60øS
WARM WDW
CELLS
The warm WDW cells have not previouslybeen observed.
within the Weddell cold regime,thoughthe historicaldata set
is too sparseto excludethe possibilityof their existenceduring
the summer.
....
CTD
Station
•
Ice
Care
X
XBT
Ice Edge
_
Wedde Abyssal Plain
•
65øS -
-65os
IOøW
5ø
Oø
5ø
IO ø E
Fig. 1. Cruise track and station sitesof the Mikhail Somovduring
the U.S.-U.S.S.R. Weddell Polynya Expedition, October-November
1981.
The warm cellsare representedin temperature-salinity
(0/S)
space(Figure 9) as nearly isohalineextensionsabove the cold
WDW. The WDW of the Weddell warm regime and CDW
just north of the Weddell gyre have the samenearly isohaline
0/S form above 0.4øC I-Huber et al., 1983; Gordon and Molinelli, 1982]. This indicates,as expected,that the sourceof the
warm WDW cells is the deep water surroundingthe Weddell
cold regime.The warm cellsappear within or to the southof a
trough in dynamic topography as determinedfrom the Mikhail Sotnov data and have characteristics
similar to the WDW
of the Weddell warm regimefound eastof Maud Rise.
The Mikhail Sotnovpotential temperature-oxygen(0/02)
characteristicvelocityof only 1 cm/s [Gordonet al., 1981].
points(Figure 10) within the warm cellsfollow the 0/02 trend
The hydrographicdata along the two north-south sections of both the South Atlantic CDW and Weddell warm regime.
obtained by Mikhail Sotnov(Figures 3 and 7) reveal three Oxygenlevelswithin the cellsare higher than thosewithin the
structuresassociatedwith lateral variationsof WDW temper- cold WDW temperaturemaximum. While it is possiblethat
ature and pycnocline depth. These structuresare (1) the the warm cellsare spawnedat the Weddell gyre-CDW front, it
warmer deep water observedat CTD stations36 and 37 and is probable that the sourceis the Weddell warm regimewhich
by the seriesof expendablebathythermograph(XBT) observa- is closerto the observedwarm cells.The warm cellsrepresent
tions north of the Mikhail Sotnovhydrographicstation grid a transferof propertiesacrossthe transitionalzone separating
(Figure 8); (2) very cold WDW at station 35; and (3) isolated the Weddell cold from the warm deepwater regimes.
cellsof warm WDW within the Weddell cold regime(above
The other major characteristicof the warm cellsis doming
0.6øCat the temperaturemaximum).
of the pycnoclineabove the cells (Figures 7 and 11). Within
The warm feature from 57ø40'S to 58ø40'S on the 0 ø XBT
the warm cells, where the potential temperaturemaximum
section(lower panel of Figure 8) is apparently an eddy of (O-max)is above 0.5øC,the top of the pycnocline(marked by
CDW. The CDW characteristicswithin the eddy as revealed the depth at which the density increasesby more than 0.01
at CTD station36 and 37 extendto 1100m beforeyieldingto sigma-0units in 10 m) occursat an averagedepth of 74 m. At
a local oxygenminimumwith WDW characteristics
[Huber et sitesof the cold WDW (O-maxof lessthan 0.5øC)more typical
al., 1983]. The primary transition from Weddell gyre to Cirof the region,the top of the pycnoclineoccursat 114 m.
Vl/eatlell
Worm
/
Fig. 2. Schematic
of theWeddellgyre.Dashed
linesshowtheisotherms
0,.6
øand1.0øC
withintheoxygen
minimum
corelayer [Gordonand Molinelli, 1982,plate 206)whichis coincidentwith the temperaturemaximumcore layer within
the Weddellgyre. Solid lines showthe 0/2500dbar isopleths0.65 and 0.70 dynamicmeters[Gordonand Molinelli, 1982,
plate 230). The Mikhail Somov sections(line a) and schematicrepresentationof the warm WDW cells (solid circles)
appear north of Maud Rise (large open circle near 65øSand the Greenwich meridian).
GORDONAND HUBER:STRATIFICATION
BELOWSOUTHERN
OCEANSEAICE
I
Y
I00
o '
•o
2
?
3
Y
5
?
643
7
Y
9
II 151721
?
ß vvv
._,g½
200
300
t
I\,
i
i
i
/
400
\
/
.-.'•'44'6øJ
/
/.69'•'x
//
60O
700
8OO
•DO
S%o
IOOC
3
I00
200
300
400
500
0
I00
37:56
200
:35
34
300
400
33 32. 31 29
260
500
28
3bo
4bo
5bo
26 23
,, ivv vvvv ß vv
30O
400
500
7OO
I00 200 300 400 500
0
I00 200 300 400 500
0
I00 200 300 400 5•X)
Distance (kin)
Fig. 3.
Potential
temperature,
salinity,
andpotential
density
alongthemeridional
hydrographic
sections
obtained
by
Mikhail Somov(Figure 1).
Similarly,the depth of the 0øCisothermincreasesby about 50
m asthe O-maxtemperatureis reducedfrom 0.7ø to 0.4øC.The
samerelationis foundin the regionalIslas Orcadasdata set.
The warm WDW cells(representedby station 28 and 32)
dbar relative to 5000 dbar (Mikhail Somovstation pair 29-32)
is 0.4 cm/sin the samesenseas the O-maxvelocitymaximum;
thus the surface current relative to 5000 dbar is close to zero.
Hence,the pycnoclinedomeeffectivelycompensates
the baro-
containan excess
of 20.5Kcal/cm
2 of heat(49Joules/m
2)and clinicity of the rest of the water column. This situation is
2.3 gm/cm
2 (23 kg/m2) of salt relativeto the Weddellcold similar to, though substantiallylessenergeticthan, the pycregime water column (representedby stations 31 and 33)
nocline eddies observed in the Arctic [Newton et al., 1974;
above 500 m (Figure 11).
Manley, 1981]. The shearreversalwithin the upper layer of
the Arctic is believed to be induced by secondaryEkman
circulation in the ocean boundary layer below the sea ice
The isopycnalinterval 27.82 to 27.83 near 200 m, which
coincideswith the O-max,is nearly horizontal, while deeper
isopycnalsdip below the warm cells(Figure 7). This induces cover [Manley, 1981].
The secondaryEkman circulationassociated
with the warm
anticyclonicgeostrophicshearbelow27.83 and cyclonicshear
above 27.82. The baroclicity is not great. A maximum geo- WDW cell is extremely weak, inducing a maximum pycof 10-5 to 10-6 cm/s.Thisassumes
a circustrophievelocity of only 0.5 cm/s relative to 1900 decibars noclineupwelling
(dbar)occursin the 27.82to 27.83sigma-0interval.A second- lar form of the warm cells,a drag coefficientat the sea iceary velocitymaximumof the samemagnitudebut of opposite waterinterfaceof 5 x 10-3 [Langleben,1982;Andreas,1983],
velocitywithin the cell of 0.5 to
signoccursat the seasurface.The geostrophic
velocityat 1900 and a tangentialanticyclonic
644
GORDONAND HUBER:STRATIFICATION
BELOWSOUTHERN
OCEANSEAICE
126
125
124
122
lib
•
2OO
105
-----•oø.'o
o
.•4oo
./
BOO
BOO
IOOO
I
•4•'•
20
•
601
80
•5•/'•
I00 I
D•stonce (•)
120
•/'•
140
160
180
•4•'•
200
220
•5•'•
240
Fig. 4. East-west
potential
temperature
section
(lineb ofFigure2)fromlslasOrcagas
cruise12[flubete•al.,1981].
1.0cm/sas determined
fromthe geostrophic
velocityrelative surroundingdeeper pycnocline.Within the warm cells the
to 1900 and 5000 dbar, respectively.
During the ice-covered pycnocline
intensityis stronger(Figure 11), whichmay be
periodthiswouldinducea pycnocline
domeof onlya meter takenas evidence
for enhanced
entrainment
[Phillips,1977].
or two, not the 40 m pycnoclinerelief observed.It is more
likely that the shallowpycnoclineassociatedwith a warm
WDW cellis a characteristic
transferred
into the regionfrom
the Weddellwarm regime.
4.
ENTRAINMENTOF THE WARM WDW
Niiler andKraus [1977], in a reviewof one-dimensional
mixed
layer models,offer the followingequationfor mixedlayer
deepening
rate,WE(cms- 1)in a twolayersystem:
2mpU, 3 npBo
WE- •
CELLS
h gap
gap
(1)
The pycnocline
domeof the warmcellsis exposed
to more whereU, isthefrictionvelocity
defined
by U, = (CdU2)
1/2,p
intenseentrainmentby mixed layer turbulencethan is the is the waterdensity,Cais the dragcoefficient
at the ice-water
98
97
0
96
95
93
92
91
90
89
88
V
V
!
V
V
• kV • L.__•c
V
i.o
IOO
/
20O
•x
---_._
//
--'•o.•_•
• .//
/
/
/
/
I
3OO
I
/
I
I
II
/
//
I
/
/
i
/
i
i
/
400
\
II
/
\\k //
//
/
/I
I
\\
i
\\
/I
/
/
/
/
/
/
k
60øS
61
'
D,stance
///
65 ø
'
/•--•-o,••[
68"
69 ø
'
70øS
(km)
Fig.5. Potential
temperature
section
along
10øE
(linecofFigure
2)fromIslasOrcadas
cruise
12[Huber
etal.,1981].
645
GORDONAND HUBER: STRATIFICATION
BELOWSOUTHERNOCEAN SEAICE
73
74
V
75
!
!
/
76
L
77
V
•
_
_
78
79
V
•
•._L.••o__.-•
80
81
f
•
•
82
83
•
V
_
_
84
85
V
•
:_ •
-tO
86
--oo -
,N• '"-i,•• •
600
7OO
800
90O
[ a• •o[
I•o
•o
53øS
56 ø
I
'•
57ø
•o
••
59ø
I
•1 620
I •
61ø
Distance (kin)
63ø
I '
1•4oo •oI
•o
•67 ø
65ø
I
680S
Fig. 6. Potentialtemperaturesectionalong 20øE(line d of Figure 2) from Islas Orcagascruise12 [Euher et M., 1981].
interface(a valueof 5 x 10-3 is used[Langleben,1982;An-
dissipationand rangesfrom 0 to 1 (approaching
zero as the
dreas,1983]); U is the magnitudeof the oceancurrent relative
to the ice; h is the mixed layer thickness,taken as 74 m for the
warm cells and 114 m for regional value; Ap is the density
differenceacrossthe pycnocline,taken as 0.2 a-0 units; Bo is
the buoyancy flux into the ocean (at the ice-oceaninterface
Bo - g[/•SF] where/• is the expansioncoefficientfor salinity,
S is the salinity of the mixed layer, F is the freshwaterinput to
the ocean);g is the accelerationof gravity.The coefficients
m
and n are proportionalityfactors.The value of m is taken as
1.25, given by Kato and Phillips [1969] though a range of
0.9-2.8 has been suggested[Cushrnan-Roisin,1981]. The
factor n is proportional to the loss of convectiveenergy to
mixed layer deepens).
A maximumnegativeBo value would occurif all the observed75 cm of sea ice [Ackley et al., 1982] were formed
locally.Usingn = 0.036(Farmer[1975] for frozenlakesmixed
layer) the entrainmentinducedby the secondterm on the
rightof equation(1) is 2.8x 10-5 cm s-•. It is likelythat
muchof the seaice is transportedinto the region,rather than
formedlocally(by the endof wintertheremay be net melting
[Gordonet al., 1983]);thus,thisrepresents
a maximumvalue.
From the first term on the right of equation (1) and the
reliefof the pycnoclineover the warm cellsthe entrainment
rateat the pycnocline
domeis expected
to be 1.54timeslarger
ioo
2OO
300
400
5OO
I0
20
30
40
50
60
70
80
90
I0
20
30
40
50
60
70
80
90
I0
20
30
40
50
60
70
80
Distance (kin)
Fig. 7. Expandedscalehydrographicsectionof the stationswithin the southeastcorner of the Mikhail $omovstation
array (Figure 1).
90
I00
646
GORDONANDHUBER'STRATIFICATION
BELOWSOUTHERN
OCEANSEAICE
-5O
.x<a-• .,.•/"•:'%
/,o-,\ •
• :.-,,
,oat--,
/
x
// •'11 I1.\
:.: ......
i ..-•';-,).
-IOO
¾'"•' '-'•-,•'"•/)"7x•
', ,
\o
I
.;,
!I
:1 •.."... ..:' "..t //
.-.,/
\
z•o•
•
'-,
/- I1',.•I [
',
,.;
I I
II I/l
I
Ill
II Ilk•
I
Ill
t
o :
/
• ,I /
[
-15o
.:
ß.....
, [L/; .........
-•/
I, ,::
I
I
:
..'
-25o
"
O l:.
-300
/i
350 '
I
,
,
!
,
, '
•eo '
•'o' '
5'90 '
•o'
'
350
6o0
,, ,fo,, ,'ffi,,'f•,,,f,
,, I,•,
, , ,f,, ,•,,ff
,,, ,f,,
,v•,,,v?,
• • ,ßl,' T• • TI
• ITo
•
off'
:
III
II
..;..l
BIB
_
lB
50-
I00-
150-
\øe•
- 200
250-
-250
-300
550
53ø
•o'
•'
57 ø
Latitude
goo
go'
35O
6,0
øS
Fig. 8. Temperature
sections
across
theiceedgeforthetwomeridional
sections
obtained
byMikhailSornov
(Figure1).
Toppanelis theeastern
section.
Thevertical
ticksareXBT observations,
thetemperature
datafromCTD hydrographic
stations
aremarked
bytriangles.
Theapproximate
seaicecoverisdenoted
bythethicksolidlinealongtheseasurface
[see
AckleyandSmith,1982].
20
Ix'xx
..........
' ' , ,•.•
+, ,
•'•
e*
CDWx•x
• ,0
'•:,'
•
ß
•
oo
•'•'....
- '•'•
WDW
.:.:',:•:"
-,
.
•....,_:.•...
o.•
+ Islas Orcados
X DrakePassage
•-I0
•,•,
"' M'f'•?L""
"• ' '
-20
I.C
ßSomov
Z.O
3.0
,
•
40
50
60
7.0
8.0
9.0
Oxygen (ml/I)
34 2
343
344
34 5
346
34 7
Salinity(%0)
Fig. 9. Potentialtemperature-salinity
relationof the MiAh•il $omo•
CTD
data.
Fig. 10. Potentialtemperature-oxygen
distributionfrom the Mikhail Sornovdata set, superimposedon selectstations from the Islas
Orcadascruise12 and Drake Passagestationfrom FDRAKE-75 data
[Nowlin et al., 1977].
GORDON AND HUBER: STRATIFICATION BELOW SOUTHERN OCEAN SEA ICE
200
---'"
-
Potenhal
(eøc)___
o
2
3
be diluted with sea ice melt, but since all the sea ice melts
regionally each spring anyway, the excesssalt of the warm
WDW cellsmust ultimately be compensatedby an increasein
atmosphericwater and glacier melt or by net convergenceof
sea ice. If the salinity anomaly of the warm WDW cells is
introducedinto the overlyingmixed layer within 1 year, then a
67 cm increasein the yearly fresh water input is required
b
Salin•ty(%0)
400
600
34.25
i
34 35
i
34 45
i
3455
i
I
34 65
I
34 75
200
400
__
_ Density
(or
e)],
600
:>74
.
275
salinity observedfrom Mikhail Somovwas -1.844øC and
34.2879/oo,
respectively,which is 0.035øC above the freezing
point (Figure 9). Pre- and post-cruisecalibration of the CTD
and independent calibration of the reversing thermometers
[Huber et al., 1983] support the significanceof the abovefreezingmixed layer temperature,which is believeda product
of the upward flux of deep water heat. The excessheat would
either be lost to the atmospherein leads and thin ice or be
usedin the melting of seaice [Gordon,1981; Gordonet al., this
issue].Presumably,local seaice productionoccursonly when
the atmosphereremovesheat faster than deep water heat is
entrainedinto the mixedlayer.
The excesssalt introducedto the mixed layer initially may
Tmperature
•--200-•
•
647
276
277
278
279
Fig. 11. (a) Potential temperature,(b) salinity, and (c) potential
densityand verticalgradientof densityof threeMikhail Somovhydrographicstationsacrossa warm WDW cell (seeFigures1 and 3). The
scalebar for densitygradientrepresents
a valueof 2 x 10-3 % units
per meter.
than .theregion.alrate. Using a value for U of 17 cm/s, based
on a 6 hour Soviet current meter record at 60 m [Huber et al.,
1983], an estimate of the entrainment rate at the dome is
32.1 x 10-'• cm/s(82 m/month)versus20.1 x 10-'• cm/s(54
locallyto producethe meanannualsurfacewater salinityof
34.259/0o
[Gordon et al., this issue]. This would more than
double the normal annual freshwater input of 45 cm [Gordon,
1981].
Naturally, the area over which extra freshwater is required
to maintain static stability dependson the degreeto which the
salinity anomaly spreadslaterally within the mixed layer. It
may spread rapidly into the entire regional mixed layer, in
which casethe required extra fresh water demand is not significant, though this ultimately depends on the number of
warm cells. It is more likely that the excesssalt introduced
into the mixed layer is confined to the cell scale of tens of
kilometers.Thus the local requirementfor freshwater is large
and may not be met. In this situation, the warm cell might
convert to a convective chimney [Gordon, 1978; Killworth,
1979]. This interestingpossibilitydeservesfurther attention.
What relation might the warm WDW cells have to the
polynya?The normal route of warm WDW into the Weddell
cold regime seemsto be by way of a westwardflow over the
continental slope [Foster and Carmack, 1976]. Instability of
the frontal zone from Maud Riseto the northeastwould inject
warm WDW directly into the axial trough of the Weddell cold
regime(to the 0.65 dy m isopleth;Figure 2) whereit would be
advecteddirectlyinto the deepoceanwestof Maud Rise,thus
transferingsalt into the central region or "hub" of the Weddell
gyre. An increasein the transfer of warm WDW cellsinto the
central region of the gyre would increasethe salinity of the
surfacewater and decreasethe pycnoclinestability, unlessan
m/month) for the regionaldeeperpycnoclinevalue.
However, this U value is most likely not representative.In
addition, the relative ocean-icemotion at a depth of 1 m,
within the logarithmicboundarylayer, would be the appropriate value to use with Ca values [Langleben,1982]. Such a
value would be significantlylessthan U at 60 m. For example,
calculating U from the winter entrainment rate determined
increase
in freshwaterfluxaccompanies
thec•epwatertransfrom Mikhail Somovoxygen data [Gordon et al., this issue]
fer. This is unlikely sincethesetwo processes
are not coupled.
yieldsa value of 8 cm/s.
The
decrease
of
pycnocline
stability
increases
the probability
A reasonable conclusion of the entrainment estimates is that
of convectiveeventsand the initiation of an open oceanpolythe 40 m pycnoclinedomesof the warm WDW cellswould be nya.
expectedto be significantlyattenuated within a single iceWhat would induce instability and generation of warm
coveredseason,introducingmuch of their heat and salt anom- WDW cells ? An increase in the curl of the wind stress would
aly into the mixed layer. Warm cells that survivethe winter spin-upthe Weddell gyre [Gordonet al., 1981] which would
would be subjectto much lessintenseentrainmentduring the draw more warm saline CDW into the gyre thus increasing
summer,when the densitydifferenceacrossthe seasonalpyc- the frontal zone intensitybetweenthe Weddellwarm and cold
noelincincreasesto greaterthan 0.5 a-0 units.
regimes.Perhapsthis would make the front more prone to
5.
DISCUSSION
As the warm cells are entrained,injecting excessheat and
salt into the mixed layer, someimpact on mixed layer characteristicsis expected.The averagemixed layer temperatureand
instability,increasingwarm WDW cellgeneration.
6.
CONCLUSION
Instability of the frontal zone separating relatively warm
deepwater of the Weddell gyre inflow from colderdeepwater
648
GORDON AND HUBER: STRATIFICATION BELOW SOUTHERN OCEAN SEA ICE
of the Weddell gyre outflow injectsWDW cellsinto cyclonic Gordon, A. L., and E. M. M olinelli, Southern Ocean Atlas:
Thermohaline-Chemical Distributions and the Atlas Data Set,
trough of the Weddell gyre. The excessheat and salt of these
ColumbiaUniversityPress,NewYork, 1982.
cells eventually (of the order of 1 year) enters the surface
Gordon, A. L., and E. I. Sarukhanyan, American and Soviet Exwater. Any increasein the injectionrate of thesecells,perhaps
pedition•iniothe SouthernOcean sea ice in October and Novemdue to spinupof the Weddell gyre,would forcemore heat and
ber 1981, Eos Trans. AGU, 63, 2, 1982.
salt into the central region or hub of the Weddell gyre, which Gordon, A. L., D. G. Martinson, and H. W. Taylor, The wind-driven
circulationin the Weddell-EnderbyBasin,Deep Sea Res.,28A, 151ultimately would enter the surface layer. The excessheat
wouldbe lostto theatmosphere,
but the excess
saltwould
163, 1981.
Gordon, A. L., C. T. A. Chen, and W. G. Metcalf, Winter mixed layer
accumulate
unless a commensurate
increase in fresh water
entrainmentof Weddell Deep Water, J. Geophys.Res.,this issue.
inputoccurs,
whichis unlikely.Thusgreaterproduction
of Huber, B., S. Rennie,D. Georgi, S. Jacobs,and A. Gordon, Ara Islas
Orcadas Data Report, Cruise 12, Ref. CU-2-81-TR.2, Lamontwarm WDW cellswould tend to lower the pycnoclinestability
Doherty
Geol. Observ.,Palisades,New York, 1981.
making the central region of the Weddell gyre more suscep-
tible to complete pycnoclinebreak down and polynya occurrence.
Huber,B. A., J. jennings,C.-T. Chen,J. Marra, S. Rennie,P. Mele,
and A. Gordon, Reports of the US-USSR Weddell Polynya Ex-
pedition,Hydrographicdata,vol. 2, L-DGO-83-1,Lamont-Doherty
Geol. Observ., Palisades,New York, 1983.
Acknowledgments.The funding for the U.S.-U.S.S.R. Weddell
Polynya Expedition was•providedby the Division of Polar Programs
of the National ScienceFoundation. The physicaloceanographyresearchis supportedby grant DPP 80-05765. Lamont-Doherty Geological Observatory contribution number 3555. Discussionswith
Hsien Wang Ou were particularly useful.We thank S.S. Jacobsfor
review of the manuscript.
REFERENCES
Ackley, S. F., and S. J. Smith, Reports of the Us-ussR Weddell
Polynya Expeditions,Sea Ice Observations,vol. 5, CRREL Spec.
Rep. 83-2, U.S. Army Corps of Eng., Hanover, N.H., 1983.
Ackley, S. F., D. B. Clarke, and S. J. Smith, Weddell Polynya Expeditionpreliminarydata report: Physical,chemicaland biological
propertiesof ice cores,technicalnote, U.S. Army Cold RegionsRes.
and Eng. Lab., Hanover, N.H., 1982.
Andreas,E. L., Comment on "Water drag coefficientof first-yearsea
ice" by M.P. Langleben,d. Geophys.Res.,88, 779-780, 1983.
Carsey,F. D., Microwave observationof the Weddell Polynya, Mon.
Weather Rev., 108, 2032-2044, 1980.
Cushman-Roisin,B., Deepeningof the wind-mixedlayer: A model of
the vertical structure, Tellus, 33, 564-582, 1981.
Deacon, G. E. R., The cycloniccirculationin the Weddell Sea, Deep
Sea Res., 23, 125-126, 1976:
Deacon, G. E. R., The Weddell Gyre, Deep Sea Res., 26A, 981-995,
1979.
Farmer,
D. M., Penetrat•ve
convection
in theabsence
ofmeanshear,
Q. J. R. Meteorol. Soc.,101, 869-891, 1975.
Foster,T. D., and E. C. Carmack, Temperatureand salinty structure
Kato, H., and O. M. Phillips,On the penetrationof a turbulentlayer
into stratifiedfluid, J. Fluid Mech., 37, 643-655, 1969.
Killworth, P. D., On "chimney" formations in the ocean, J. Phys.
Oceanogr.,
9, 531-554, 1979.
Langleben,M.P., Water drag coefficientof first-yearseaice, J. Geophys.Res.,87, 573-578, 1982.
Manley, T. O., Eddies of the westernArctic Ocean: Their characteristicsandimportanceto theenergy,heatand saltbalance,CU-1-81,
Tech. Rep. 1, Lamont-DohertyGeol. Obs., Palisades,New York,
1981.
Newton,J. L., K. Aagaard,and L. K. Coachman,Barocliniceddiesin
the Arctic Ocean, Deep Sea Res.,21, 707-719, 1974.
Niiler, P. P., and E. B. Kraus, One-dimensionalmodels of the upper
ocean, in Modelling and Prediction of the Upper Layers of the
Ocean,editedby E. B. Kraus, pp. 143-172, Pergamon,New York,
1977.
Nowlin, W. D. Jr., T. Whitworth III, L. I. Gordon, and G. C. Anderson, Oceanographicstation data collectedaboard R/V Melville
during FDRAKE 75, Ref 77-2-D, Texas A&M Res. Foundation,
CollegeStation, 1977.
Patterson,S. L., and H. A. Sievers,The Weddell-Scotiaconfluence,J.
Phys.Oceanogr.,10, 1584-1610, 1980.
Phillips,O. M., Entrainmentin ModellingandPredictionof the Upper
Layersof the Ocean,editedby E. B. Kraus, pp. 92-101, Pergamon,
New York, 1977.
Tchernia, P., Etude de la Derive AntarctiqueEst-Ouestau Moyen
d'Ice-bergsSuiviopar le SatelliteEole, in Polar Oceans,editedby
M. J. Dunbar, Arctic Institute of North America, Montreal,
Quebec, 1977.
intheWeddell
Sea,J.Phys.
Oceanogr.,
6,36-44,1976.
A.L. GordonandB. A. Huber,Lamont-Doherty
Geological
ObGordon,A. L., DeepAntarcticconvection
westof Maud Rise,J.• servatory,
Columbia
University,
Palisades,
NY 10964.
Phys. Oceanogr.,8, 600-612, 1978.
Gordon, A. L., Seasonalityof Southern Ocean sea ice, J. Geophys.
Res.,85, 4193-4197,1981.
Gordon, A. L., The US-USSR Weddell Polynya Expedition,Antarctic
J., 17, 96-98, 1982.
(ReceivedApril 28, 1983;
revisedOctober 6, 1983;
acceptedOctober6, 1983.)