Atmospheric Net Transport of Water Vapor and

2 downloads 0 Views 1MB Size Report
Jan 20, 1992 - Kobayashi, 1978; Wendler, 1989a], but uncertainties in the interpolation and ..... mean rate of approximately 531 kg m '2 a -1 and a total.
JOURNAL OF GEOPHYSICAL

RESEARCH, VOL. 97, NO. D1, PAGES 917-930, JANUARY

20, 1992

Atmospheric NetTransport of WaterVaporandLatentHeatAcross70øS M. B. GIOVINETrO Department of Geography, University of Calgary,Calgary,Alberta,Canada D. H. BROMWICH ByrdPolarResearch Center,OhioStateUniversity, Columbus

G. WENDLER Geophysical Institute,Universityof Alaska,Fairbanks

Theannual netatmospheric transports of watervaporandlatentheatpoleward across 70øSareestimated usingthelatestcompilation of surface massbalance for theAntarctic ice sheetandnewestimates of precipitation andevaporation in sectors of thesouthern oceans andof seaward driftingsnowtransport in particular sectors of theicesheet.Themass andenergy exchange ratesattheicesheet-atmosphere and ocean-atmosphere interfaces areintegratedstrictlyfor areaswithinthatlatitude.Theestimates of net

southward water vapor transport (6.6+ 1.3kgm-1s-1)andlatent heat transport (18.9 + 3.6MJm-1s-1)are largerthanreported in allpreceding studies, based onatmospheric advection andmoisture datacollected at stations locatedbetween66øSand80øS,andaregenerally in agreement withthosebasedon surface mass

balance dataandseaward drifting snow transport across theiceterminus whichextends between 65øSand 79øS.

INTRODUCTION

sectors 61øW- 76øW and 1øW (eastward) - 160øE, and

Thewater andenergy budgets ofthe area poleward of70øSincluding ocean areas lyingsouth ofthereference latitude inthe are important elements of modelsusedin atmospheric, sectors 1øW- 61øWand76øW(westward) - 160øE. It isbased on glaciological, andoceanographic studies. Twobasictermsin estimates of themeanannual netdifference between surface estimates of those budgets aretheatmospheric nettransports of accumulation andablation,or surfacemassbalance(herein watervapor andlatent heat,thelattertreated asa function of the referred asbalance), in theice sheet,andon thedifference

former. In thelastthree decades, estimates ofnetvapor transport between precipitation andevaporation, oroutflow, intheocean [e.g.,BryanandOort,1984]havebeenmadeusingdatafor areas. Therefore drifting snow transport needs tobeconsidered Antarctica, excluding Graham Land(Figure1).Mostestimates in twocontexts: as an outputtermin thecaseof transport havebeenbasedon upperair data,hereafterreferredto as northward across 70øSontheicesheet andasaninputtermin atmospheric data,fromdatasetsof approximately 10stations [e.g., thecaseof transport across theperiphery of theicesheet, orice Peixoto andOort,1983] located at relatively shortandregularterminus, to the oceanareas lyingwithinthereference latitude distances fromeachotherandpresenting an acceptable fit relative(Figure 2). to the referencelatitudefrom 45øW (eastward)to 160øW [Rubin

andWeyant, 1963].Suchis notthecasein theremaining sector,

BACKGROUND

where onlydatafromByrdStation atalatitude of80øShave been

Theavailable estimates ofpoleward moisture transport from

used. A few studieshaveincludedcomparisons of estimatesatmospheric measurements resultgenerally in underestimates. basedon atmosphericdata with thosebasedon various Bromwich [1990]compared themoisture transport convergence combinations of surface accumulation andablation termsandother values for Antarctica from two recent climatological massbudget termssuchascalving, hereafter referred to assurfaceatmospheric analyseswith thoseinferablefrom the balance data, integrated for theice sheet[e.g.,Rubin,1962;Bromwich,compilation of Giovinetto andBentley[1985].Forthecontinent 1990].Thusthe atmospheric dataaredirectlyrepresentative of asa wholetheatmospheric-derived estimates wereonlyonethird vaportransport across theboundary of an areaof approximatelyto onehalf the surface-based values;for 80øS- 90øSthe two

12x 106km2,which boundary extends between 66øSand80øS,approaches yieldedcomparable results. Bromwich[1990] andsurface dataareintegrated foranareaof approximately 13.7x concluded thattheprimarycauses of thelargediscrepancy for

106km2,which boundary extends between 65øS andalmost 79øS.thecontinent weredeficiencies in atmospheric diagnoses of the Theseestimates arethenextrapolated to applyto theareawithin transport contributions of cyclones andsurface windsnearthe 70øS or 15.51 x 106 km 2.

coast.A secondarycausecontributingto the discrepancywas the small systematic underestimate by radiosondes of the is madespecifically for theareawithin70øS(Table1) excluding atmosphericmoisturecontentduring the cold polar night. the areasof the ice sheetlying north of the referencelatitudein A chronological overview of estimates of water vapor The assessment of net vapor transportpresentedin this work

transportacross70øS basedon either atmosphericor surface Copyright1992by the AmericanGeophysical Union.

data is presented in Table 2. The values are arranged by publicationdate, and only analysesdrawing upon data collected during or after the InternationalGeophysicalYear 1957-1958

Papernumber911D02485. 0148-0227/92/91 JD-02485505.00 917

918

GIOVINETTOET AL.: ATMOSPHERIC WATERVAPORTRANSPORT ACROSS70oS

oC• !

Novolazarevskay•

'• •

.....&Norway

Roi

Halley

Bay.:/F-"'"

½•

'"...GRAHAM .. LAND

..

.•_ -',80'S ....... .," ............... ; "" Ellsworth

ß'" •

FILCHNER I.S.

RONNE I,S.{• >

90øW-]-

-,; •

,-

,o,s•

/• \

.'.,..e

..• \\

c..

_ •\ •



0

%•



O•

-',-

/

.•Y'•



•o /\

:,,, •

u• 4,,'...o.• •,,, • .......•. ..

.• ,

South Pole Sta.T,

S T

F' 2(c) s;;;s. .....

J-

.... '¾2

-'-

'%/ • -,-

_.•:•A ' N T A R C T • C A o'r

'•.•,• '_-'.,

EA

d :/

,,,,• • T

,, ':::',• :.\

O\ "

-:;

.os

-',-

•,

r'x,I

.••



I

t'"' --90"E



'!,,,,,r•,

/ /

l'•...... : •t'.'"

J 'GF o8'),.• •f:

/

"-

/ •

7o'0s



%.\.

. soøs •' --

/



...'•.. ,.•.?••(• •.•j,,•....d L.•i. ' .............. •.McMurdo ......... Ameri• V ..........• % •

-

/-

A N T A R C T I C A//

,ga,,e,

/



?.Casey

,.,2'

•e .._•..'

/



J.;

/

,,o,

,,-

Fig.1. Antarctica, showing thelocation ofplaces mentioned inthetext,icesheet andocean areas within the70thparallel (full circle),stations forwhichupper airdatahavebeenused in studies ofvapor transport (PortMartin,D 80,andGF08,arestations discussed in thecontext of surface windspeed), andtransects (a)-(c)used asexamples inFigure 2.TheSouthern Ocean surrounds

the continent. Areas are listed in Table 1.

(IGY) are considered. Estimatesprecedingthe IGY were smallequatorward to a poleward transport of 5 kg m-1s-1'The primarilyconjectural because of verylimitedobservations. There trend in atmosphericestimatesparallels the degree of is a markedcontrast between theatmospheric andsurface-based monitoringof atmospheric processes over the oceanareas estimates. The latterreveallittle trendwith publication year, surroundingthe continent,ranging from a few oceanic fluctuating arounda poleward transport of 6 kgm-1s-1;whereas,observations duringthe IGY winter to the comprehensive the atmospheric estimates showa strongtrend,changingfrom drifting buoy programin the SouthernOcean during 1979 TABLE1. TermsUsedin theEstimate of NetWaterVaporTransport Southward Across 70øS(V) AgT•rms

V Terms,

•m '2a'1

1015 t•ga'1

•otS•g,4

124 ...

1.48 0.12

1.48 0.12

531 170

1.89 0.61

...

...

11 i,•,/,

',terminus

,

I

I

1

! Ice shelf barrier

-,',-,, %

......... [

70ø

5



' ice I Grounded

V

919

I

[

80ø

I [

90øS



I ,

I

0 '-'

•,V

, I

•' ,I

[

80ø

% I

70ø



Fig.2. Schematic cross sections ofAntarctica (along lines (a)-(c) shown inFigure 1)depicting thedisposition ofterms discussed in the text.

[Hollingsworth, 1989].It appears thata substantial fraction of Antarctica as a wholethe balancedataofferbetterareal andat present aremorereliablethantheatmospheric the polewardmoisturetransportshortfallin early atmospheric coverage useof stable-and diagnoses wasdue to limited datain the easternPacificsector, data.In the last two decadesthe widespread radioactive-isotope methods, the development of sampling where approximately 40% of the moisture transport into that incorporate deeperstratarepresenting longer Antarcticaoccurs[Lettau, 1969] and which was well monitored techniques

andtheproliferation of stakenetworks haveconlxibuted only during 1979. Interannual variability complicatesthis periods

comparison because atmospheric estimates tendtobebased on to substantial reductions in theerrorof pointbalance rate datafora single yearandsurface estimates arebased ondata determinations [e.g.,Giovinetto andBull,1987].Moreover, thataregenerally themean ofoverlapping multiyear periods. It increases in thenumber of pointdata,whichin thelasttwo 350 to over 1200, and can be concludedthat, althoughthe atmosphericanalysesshow decadesincreasedfrom approximately

steadyimprovement, they arenot yet suitablefor continent-wide areal coverage,togetherwith the productionof detailed

maps,havecontributed to substantial reductions in studies of theatmospheric waterbalance. However,for 80øS- topographic the error of interpolation and extrapolation of those rates and 90øS andperhapsa largefractionof EastAntarcticatheymay thusof areallyintegrated means,whichmaynowbe estimated be usedfor thispurtx)se[Bromwich,1990]. with an error of approximately +_.10% [Giovinetto andBentley, The underestimate may be explainedin part by the paucityof atmospheric datain thecoastal sector 70øW- 160øW,for which 1985].

Estimates of netvaporlxansportbasedexclusivelyon surface eddyandtime-averaged circulationvaporlxansport. It hasbeen data for the ice sheetare a direct assessment of particular knownfor sometime that the vaportransportinto that sectoris accumulationand ablationterms,typically combinedunderthe large basedon both meteorologicalstudies[Alt et al., 1959; headingsof precipitationand evaporationin the productionof

Lettau [1969] indicated the probability of large tropospheric

Wexler, 1959; Rubin and Giovinetto, 1962] and balance compilations. On the basis of the latter, Giovinetto [1961] suggested that the largestbalanceratesin Antarcticawouldbe found in the coastal zone of the BellingshausenSea sector (Figure 3). This inference was later supportedby field observations of Shimizu[1964] andBehrendt[1965], asstudied by Lettau[1969],andsubstantiated by Bishopand Walton[1981] andPotteret al. [1985].For thissectorin particularaswell asfor

water budget atlassesof the world showing continental and oceanic summaries[e.g., Baumgartner and Reichel, 1975; Korzounet al., 1977].Estimates of net vaporlxansport basedin part on assessments of drifting snow seawardacrossthe ice terminusmust be made with caution becausethe range of estimatesof drifting snowreachingthe sea as reportedin the literatureis very large [e.g., Rubin, 1962; Budd et al., 1966]. Methods to improve snow drift measurements at particular

TABLE 2. Estimates ofMeanAnnual NetWaterVaporTransport Across 70øS(V)UsingDataCollected During and AftertheInternational Geophysical Year 1957-1958

Data Base,periodor

V *,

number ofyears

kgm'1s'1

1958 1963-1973 1973-78 and1980-84 1979

(+)0.73 (-)3.0 (-)3.7 (-)5.3

Rubin[1962]

multiyear

(-)5.6to (-)7.2•'

Bromwich [1990]

multiyear

(-)5.8to (-)6.0t

Basedon atmospheric data

Starretal. [1969] Peixoto andOort[1983] Howarth [1983]; Howarth andRaynet[1986] Masuda[1990]

Based on surface data

Baumgartner andReichel [1975]

Thisstudy

multiyear

multiyear

* Negativesouthward (followingstandard meteorological convention).

(-)5.4

(-)6.6+ 1.3

t Transport across thecoast ofthecontinent converted forthisstudy totransport across 70øSusing theratioofthe twovaluesgivenby Baumgartner andReichel[1975].

920

GIOVINETTOET AL.' ATMOSPHERIC WATERVAPORTRANSPORT ACROSS?0oS

3

o..

3

0.5

-0.5 0

-I-

I

iiiII I

.,!

!

I',. /

'

i

t

Fig. 3. Distribution of surfacebalanceon theice sheet(modifiedafterGiovinetto and Bentley[1985])andits extrapolation to

show aprobable distribution ofprecipitation intheocean areas poleward of70øS. Units arein100kgm'2a-1'

locationsand for different snow surface and boundarylayer conditions have been developed [e.g., Budd et al., 1966; Kobayashi, 1978; Wendler, 1989a], but uncertaintiesin the interpolationand extrapolationof the mean rate, as well as of the actuallengthof the peripheryto which the meanrate may be assumedto apply, remain the principal causeof the wide rangeof estimates. Typically, measurements made at locations in the coastalareaare qualifiedfor the effectsof topography on surfacewinds,as thesemay be deducedfor areasof the orderof

to the groundedice terminuscould be excluded,particularly as they contain the steepest parts of terminal slopes where katabaticflow is sustainedby surfaceinversionconditionsin the interior [Lettau, 1966; Lettau and Schwerdtfeger,1967]; these conditionslead to the highestfrequenciesof strongwinds [e.g., Schwerdtfeger,1970;Parish and Waight, 1987] and largestrams of drifting snow transport.We obtainthat objectiveby limiting our studyto the area strictly within the 70th parallel becauseit excludesa sectorapproximatelyfrom 30oE to 160oE, wherethe 103 km2 but not for the effectsof broadridgeandtrough groundedice sheet extends northward of that latitude (an area

1.76x 106km2) andwheremorethan3/4 of topographyextendingupslope,respectively,of capesandbaysat of approximately or near the groundedice terminus. For a detailedtopography, the periphery of the ice sheet characterizedby steepterminal see Drewry [1983]. The broad ridge and trough topography slopesandstrongkatabaticwindsis found. induce fanning and channelingof the airflow in areasof the APPROACH

orderof 105 km2 and consequently large uncertaintyin

Our approachto the estimateof net vapor transportacross interpolation and extrapolation. Furthermore, the varying 70øS (V), basedon massexchange datafor the surfaceof both availability of fine-graineddry snow, both as precipitationand surfacematerial dislodgedby saltation and corrasionand the the ice sheet and ocean areassouthof that latitude (Figure 2), frequencyof strongsurfacewinds sustainedby air drainageover includes: 1. Estimatesof balancein the ice sheet(B) and of drifting long slopesof diverse steepnessare difficult phenomenaand snow losses fromtheicesheetnorthward across 70øS(Do). elementsto synthesize.Thus, a reductionin the compositeerror 2. Estimatesof the differencebetweenprecipitation(Ps) and would be achievedif all or substantialpartsof zonescontiguous

GIOVINETTO ET AL.: ATMOSPHERIC WATERVAPORTRANSPORT ACROSS 70ø5

evaporation(Es) in the oceanareas,or outflow (F), and of

921

In climatologicalstudiesof water balanceat the air-ocean

driftingsnowtransport fromtheicesheetacross theiceterminusinterfacefor wholeoceansor theirhemispheric partsandat

totheocean areas lyingwithin 70øS (Di).Then V = (B+ Do)+ (F + Di)

smaller scales for mostseasit is customary to showthe distributionof the precipitationrate by extending isohyets from

(1) landto seaareas.In thecaseof theoceanareaslyingpoleward

of 70øS the distributionof precipitationis inferredby seaward In the ice sheet,B is obtainedby integratingcontoured extrapolations of balanceisopleths asdetermined for thecoastal

distributions of the balance rateasdetermined for a particularzoneof theice sheet(Figure3); Ps is obtained by integrating location(b). The specificratesof massexchange processes at the thosecontoureddistributions [e.g., Baurngartnerand Reichel,

surface arenotintegrated separately; rather,b is determined asa 1975;Korzounet al., 1977]. The extrapolation of balance net value on the basis of one or more annual accumulation isoplethsasisohyetsdeservesome qualifications,whichfollow.

layers:

It is well knownthatin thecoastalareasof thegrounded ice sheetthe balancerate is smaller than the pr•ipitation rate [e.g.,

b=(Pl+Svsl+Cl+Dpl+I1)-(Sl+El+Dfl+M1)

(2) Bromwich,1990]. Overdistances of the orderof l00 km from

the groundedice terminusthe annualbalanceis positivebecause whereP1is precipitation,Svs1 is vaporto solidsublimation,C1is the summationof the rates of precipitationand drifting snow condensation, Dpl is driftingsnowdeposition, I 1is superimposed deposition is greater than the rate of snow deflation. This ice,S1is sublimation, E1is evaporation, Dfl is snowdeflation,and situation changes in a relatively short distance across the M1 is melt runoff. groundedice terminus, since over the relatively flat sea ice, An updateof a previousreviewof snowstratigraphyin the ice wherewind speedtypicallyd•reases beyondthe zonecontiguous sheet[Giovinetto,1964a] showsthat southwardof 70øS thereis to the terminus with a width of approximately15 km [e.g.,

nomeltingat thesurface in approximately 90%of thearea.In a Schwerdtfeger, 1970],it maybeassumed thattheramsof drifting narrowzone alongthe coastof West Antarctica,as well as in the RossIce Shelf and the Siple Coast[Alley and Bentley, 1988], surfacemelting occurs sporadicallyin summer;this water percolatesinto the underlyingwinter accumulationlayer and refreezes.In relativelyfew locationsandonly for a few yearsin longseriesdoesthe stratigraphy showthatpercolation hastaken

snowdepositionand snowdeflationare approximatelyequal,and hence balance and precipitation rates should also be nearly equal. (It shouldbe noted that complexdistributionsof ratesof drifting snow deposition and deflation, and balance, over distancesof the order of 10 km on either side of groundedand floating ice termini [e.g., Giovinetto, 1964b] involve relatively

placeinto otherannualaccumulation layers.Thusthe potential small areas and do not affect our estimate of terms such as B or for percolationto introduceerrorsin the determinationof the Ps.) It remainsto determinea correctionto bring the (smaller) balancerate is restrictedto a relativelysmall area,and it may be totallyignoredbecausemostdeterminations of theratearemade on the basis of several annual accumulation layers. In the remainingareaenclosed by the dry snowline, somesnowandice surfacesmay be exposedfor one or more years[e.g., Fujiwara

balancerate, as estimatedor measuredover the terminal slopes to a value that approximates the actual (larger) rate of precipitation.

type andbalancemay be in all or part of someof thoseregions [e.g.,Mcintyre, 1985;Allisonet al., 1985;Sekoet al., in press]. However,the regionsare relativelysmall anddifferencesin the estimateof b for their areasare not significantin the estimateof B. From this brief discussionof snowstratigraphyand surface

s•tors which lie outside the area of interest, such as those where

A

correction

for

the

difference

between

the

rates

of

precipitation and balance in the terminal slopesis incidentally and Endo, 1971; Picciotto et al., 1972; Nishio et al., 1985]. These applied when the balance isopleths are extrapolated as typesof surfaceshave been observedin partsof someof the precipitationisohyets.The compilationsof balanceare produced regions depicted inFigure 3 bythezeroand0.2x 102kgm-2a-1 on small-scalemapson which the complexityof the balancerate balanceisoplcths; the regionslie well abovethe dry snowline. distributionin the terminal slopescannotbe shown [Giovinetto Therearesomeconflictingreportsas to what the actualsurface and Bull, 1987], with the exceptionof a few relatively small stationsSyowa,Mawson, Davis, andCaseyare located(Figures1 and 3). In the areal integrationof balancefrom small-scalemaps a deflation correctionis applied to most sectorscharacterizedby steep terminal slopes where deflation is much greater than typesfoundin theicesheetsouthward of 70øS,it maybe stated deposition[Giovinetto,1964b; Giovinettoand Bentley, 1985], but that in the area of interest the rates of vapor to solid the balance isoplethsas mapped do not include the deflation sublimation,condensation,superimposed ice, evaporation,and correction, so that their balance rates approximate the melt runoff are negligibleand that the ratesof sublimationand precipitationrates. Moreover, precipitationis large in the zone snow deflation are particularlylarge in some regions. In the of terminal slopes due to orographically induced uplift of context of the latter term it may be noted that an increasein polewardmoving air masses[e.g., Rubin and Giovinetto, 1962; driftingsnowleadsto an increasein the densityof air bothby Oerlemans, 1982; Brornwich, 1988], a precipitationenhancing

coolingfromsublimation of driftingsnowandby the phenomenon thatis absent beyond approximately 100km incorporation ofparticles, thus increasing thespeed ofkatabatic offshore, where formuch oftheyearsensible heatfluxfromair flow [Wentiler, 1989a]. This suggestsa potentialfor positive to either water or sea ice promotes stability in the lower layersinhibitingprecipitation[e.g., Zillman, 1972; feedbackepisodesthat,howevershort, may producelargerrates tropospheric

Andreas and Makshtas, 1985]. We assume that these previous studies. Nevertheless,these losses are implicitly overestimate and underestimate biasses introduced by the extrapolationprocedurecancel each other out. accounted for in the estimate of balance. Es is estimatedfor openwater and seaice areasas a function In the ocean areas, as stated above in 2, F is estimatedas the of Le (latentheat eddy flux) which is in turn estimatedin several differencebetween precipitationand evaporation: studiesusingmodificationsof the basicequation,for exampleas F=Ps- Es (3) givenby Zillman [1972]: of sublimation

and snow deflation

than have been assessed in

922

GIOVINETTO ET AL.: ATMOSPHERIC WATER VAPOR TRANSPORT ACROSS 70oS

Le= k (dL/p)(vpsf- vph)(u)

(4)

where k is a factor determinedby the units used and levels at which the observationsare made, d is densityof air, L is latent

heatof vaporization, p is atmospheric pressure, vps f is vapor pressureat the surface,vph is vaporpressureat level h, and u is wind speedat a particularlevel. Thus Le may be integratedto estimatethe net heat energyflux (positiveupwards),which at a point and time includes all processesinvolving latent heat transfer,such as condensationand vapor to solid sublimationon

sea ice in winter, evaporationfrom oceanand (wet) sea ice surfacesin summer[e.g., Gow and Tucker, 1990]. We estimate Es usingexistingestimatesof Le, as describedin the following section.

ESTIMATES OF N•:'r WATER VAPOR AND LATENT HEAT TRANSPORTS

Estimatesof B and F

The compilationof balancefor the ice sheetby G iovinetto and Bentley [1985] is used to estimate B (Figure 3); that compilationindicatesa balancefor the area southof 70øS of

approximately 1.48x1015 kga'1or 124kgm-2a-1 (Table1). The projectionsof balanceisoplethsfrom that compilation over the ocean areas, where technically they becomeisohyets, are shown in Figure 3. They are drawn showingorientations

parallel to trends in thefrequency andarealdistributions of cyclones [Taljaard,1972;Kep,1984]andcloudvortices [Streten

andTroup, 1973; Carleton, 1981], andlarger rates downwind autumn through spring [Streten, 1973; Carsey, 1980; Zwally in et from areas where large polynyas and leads form frequently

al., 1983;ComisoandGordon,1987]aswell asicefreeareasin

summer[Naval OceanographyCommandDetachment (NOCD), 1985]. The isohyets patternindicates a probable

distribution ofprecipitation intheocean areas south of70øS, themain features ofwhich are:(1)Thelowprecipitation pockets in areasadjacent to thecoastin thesouthern WeddellSeaand

southwestern Ross Sea inferred from well-substantiated low

balance values on the ice sheet; (2) The relatively low precipitationin the westernareasof both seasinferredfrom high

elevation in Palmer Land and Victoria Land oriented

perpendicular tothecyclonic tracks, inpart deflecting cyclones and in part inducing"rain shadow"effects on the oceanareas

downtrack; (3) Thehighprecipitation in thenortheastern Weddell Seainferred fromtherelatively highfrequency of cyclolysisand the probabilityof polynya formationnorth and

eastofthe600kgm-2a-1isohyet; and(4) Thehighprecipitation in the easternAmundsenSea- Bellingshausen Seaarea,inferred from the well-substantiated high balancevalueson the ice sheet, as well as meteorologicalanalysesapplicableto this sectoras mentionedin a precedingsection. The distributionof precipitationindicatesmeanratesfor the

ocean areas of approximately 544kgm'2 a'1inthe Pacific sector and504kgm'2 a-1in theAtlanticsector, fora combined-sector mean rate of approximately 531 kg m'2 a-1 anda total precipitation of approximately 1.89x 1015kga'1 (Table1). The estimateof Es is derivedin tabularform usingexisting estimatesof Le for ice-freeocean,openwaterwhenice is present, and sea ice (Tables 3a and3b). In our estimatethe particular surfacetypesjust describedare each assumedto be relatively homogeneousregarding surface parameters such as albedo, roughnesslength, and substratethermal diffusivity becauseof their roughlycircumpolardistribution.

GIOVINETTO ET AL.' ATMOSPHERICWATER VAPOR TRANSPORT ACROSS70oS

923

TABLE 3b.Estimate ofLatent Heat ofEvaporation (Le,Wm-2)intheOcean Sectors South of70øS Summer

Autumn

Winter

Spring

VI-VIH

IX-XI

XII-H

III-V

Mean for water*

20•'

40

100:[:

40

Meanfor ice*

17õ



(nil)•

9

Mean for area

18

16

10

13

37

14

* Latentheatflux valuesarefactoredby the weightedindexeslistedin Table3a. t Extrapolated from Zillman [1972]. $ Interpolated fromAndreaset al. [1979]andBrown[1990].

õ Adopted withmodifications fromAllison[1972]andMaykut[1978]. • Adopted withmodifications fromMaykut[1978]andWeller[1980].

Thefirststepin theestimate ofLe isanexamination of the Thesecond stepistoestimate Leforparticular surface types arealandtemporal distribution of ice-free ocean, openwaterandseasons toensure reliable interpolations andextrapolations when iceis present, andseaice.Using semimonthly chartsapplicable tothemean latitude fortheocean areasouth ofthe prepared bytheNOCD[1985] fortheperiod 1973-1982 based on 70thparallel, i.e.,approximately 73øS.Weller [1980] estimated passive-microwave datafromNimbus 5 andNimbus 7,aswellas latentheatfluxforseaicein midwinter in thesector 20øEvisibleandinfrared imagery particularly duringsummer months,90øE,between 58øSand70øSwithconcentrations of 15-85%

it ispossible tosummarize theoccurrence of eachof thethreeand> 85%(37and28W m-2,respectively). Allison[1972] typesof surfaces(Figure4). It is necessary, however,to estimated latentheatflux at 67.5øSduringfreeze-uptime in

differentiate between particular seaiceconcentration categories earlyautumn, andearlystages of iceformation to a thickness of (.g,. 20%,20-50%, 50-80%, and>80%)toestimate thefrequency0.5 m (valuesrangingfromapproximately 80 to 20 W m-2, andareaof openwater.In thiscontext it should benotedthat respectively). Zillman[1972]estimated latentheatfluxforiceboththe rangeof seaice concentration and openwaterare freeoceanin summer in thesector110øE- 160øE,between 58øS

simplified forthisstudy byassuming a middle range value,and68øS, listing 6-hour mean fluxes forbands of4øoflatitude

although thedata compiled bytheNOCD only represent theoverlapped every 2ø (values between 15and17.5Wm-2, range ofseaiceconcentration ineach category. Theseaiceincreasing southward with length ofday). concentration data summarized inFigure 4 areused toobtain Estimates oflatent heat fluxforsea iceathigh latitude inthe seasonal areaandtime-weighted indicesmeasuring the areaof central Arctic in winter show small negative values [Maykut,

each particular seaiceconcentration ineach particular season 1978]. Estimates of latent heatfluxforopen water leads in relativeto the total grapharea(Table 3a). For brevity,the winterat approximately 71øN nearPointBarrow,Alaska,have resultsof this work are shown for the Pacific and Atlantic sectors

combined. Moreover, foreasy comparison with existing estimates been estimated at50Wm-2[Andreas etal.,1979], and elsewhere

ofLe(particularly those ofZillman [1972]) the results areinthe northem polar region at150 Wm-2[Brown, 1990].

summarized onthebasis offourseasons, each consisting ofthree Maykut [1978] hasshown thatlatent heatflux,among other

calendar

months.

SUMMER

I-

AUTUMN

I

WINTER



SPRING

• SU.-"

100

termsof the energybudgetof seaice, is highly dependenton ice thickness in the range of from a few millimeters to approximately0.5 m. This and othermodelsshowlittle change in latentheat flux after thicknessincreasesbeyond 1 m [Weller,

1980].In theoceanareapolewardof 70øS,icethickness exceeds 1 m in at least 9 monthsof the year [Washingtonet al., 1976; Hiikkinen, 1990]. Thereforethe valuesobtainedby Weller [ 1980]

90



8o

for midwinterat a latitudelower than70øS andrelativelythin ice should be drastically reducedto reflect the lower values

• 70

• 50

obtainedby Maykut [1978] for higher latitudesand thicker ice, particularly as earlier studiessuggestedthat net condensation andvaporto solidsublimationmay occurin winteron seaice in the SouthernOcean [Fletcher, 1969]. Also, the values obtained

_> 80

byAllison[1972]forearlyautumn at a latitude lowerthan70øS

40

and relatively thin ice, should be reduced, although not as drasticallyasWeller's,in lightof Maykut'sresults. The valueswe adoptedfor our estimateof latentheatflux are shownin Table 3b and are implicitly the values assignedto a

'• 3o

Z•20

50-80

o

lO



o

j

F

J

A

S

O

N

D

latitudeof 73øS.Forice-freeoceanandopenwaterwhenice is

present theyrangebetween 20 W m-2in summer (extrapolated fromZillman[1972])and100W m-2 in winter(adopted from Arcticdatastudies by Andreaset al. [1979]andBrown[ 1990]).

Fig.4. Composite annual distribution ofmeanseaiceconcentration in the For sea ice they rangebetweennegligiblein winter (adopted ocean areaspoleward of70øSbased onananalysis of semimonthly charts (modified from for the period 1973-1982.The seasonallabelscorrespond to the fromMaykut[1978])and17W m-2 in summer distribution of latent heat flux discussed in the text. Allison [1972] allowing for differencesin time of the year,

924

GIOVINETTO ET AL.: ATMOSPHERICWATER VAPOR TRANSPORTACROSS70oS

latitude, and ice thickness). In each of the surface types we Even though a fixed relationship between wind speed and arbitrarily assignedintermediatevaluesfor autumnand spring, transportcannotbe expected,we assumedsucha relationshipto althoughthere are studiesthat suggestdifferent valuesfor the obtain order-of-magnitudevalues for the amount of drifting

transitional seasons [e.g.,Fletcher,1965,1969;Washington et al., 1976; Maykut, 1986;Brown, 1990; Gow and Tucker, 1990]; however,theirdiscussion is beyondthe scopeof ourpresentwork. The weightedmeansfor each season,as well as the weighted

snowfrom estimatesof wind speed.For Antarctica,wheremost of thezonealong70øSlieswithintheareacoveredby dry snow

ice (increased to account for sublimation or melting), we

to 70øS there is an insufficient

[Giovinetto,1964a], the errorsintroducedby this assumption are minimized,as hardnessof the snowcoverandsizedistributionof annual means obtained for water and sea ice surfaces, 37 and 8 snowgrainsvary lessthan in coastalareas. Wind velocity observationsare easier to perform than those W m-2,respectively, indicate a meanlatentheatfluxforthearea of 14 W m'2. Usingvaluesof latentheatof vaporization for of drifting snow, and furthermore, they can be carried out phase changes at0 C, i.e.,2495J g-1forwater,and2830Jg-1for remotely [Stearnsand Wendlet, 1988]. Nevertheless,at or close number of stations to infer a

estimate Es tobeapproximately 170kgm-2a-1or0.61x 1015 kg reasonabledistributionof the drifting snow transport.In part, a'1.ThusF is estimated to be approximately 361kg m-2 a-1or this is due to the pronouncedspatial variability of the average surface-windspeeds[Parishand Bromwich,1991]. 1.28x 1015kga-1(Table1). We resorted to a numerical

model of katabatic

airflow

over

Antarctica [Parish and Bromwich, 1991] to provide the mean Estimatesof Do and Di wind speeds.This approachis justifiedbecausethe surfacewinds To completethe estimateof V, it is necessaryto estimateDo in Antarctica are mostly of a katabaticnature. Schwerdtfeger northward across 70øS in the ice sheet sectors 61øW - 76øW and [ 1970] pointed out that in no other continentis surfaceclimate 1øW (eastward)- 160øE, and Di from the ice sheetto ocean so stronglydominatedby a single meteorologicalparameteras areasin the sectors1øW - 61øW and76øW (westward)- 160øE. that of Antarcticais by the katabaticwind. Mather and Miller Ideally, theseestimatesof drifting snowtransportshouldbe based [1967] usedthe observedwindsand sastrugiorientations to map on a large number of actual transportmeasurements.However, the time-averaged surface winds over Antarctica, and later

few suchdeterminations exist and as measurements of drifting Parish [1984] and Parish and Bromwich[1987, 1991] used

snow are difficult to perform, they cannotbe carriedout numerical modelsof katabafic airflowto providea moredetailed remotely. depiction of thesewinds.Thestreamline mapsderivedfromthe Observations of driftingsnowhavebeencarried outoverthe observations and from the modelsshowedgood overall yearsat severallocations in Antarctica usingbothmechanicalagreement. Furthermore, theconstancy of winddirection is very traps[e.g.,MellorandRadok,1960;Budd,1966; Loewe,1970; high;meanmonthly valuesarearound 0.9, whichconfirmthe Naruse,1970;Kobayashi,1972,1978]andelectronic devices overwhelming predominance of katabatic surfacewinds.The [e.g.,Landon-Smith andWoodberry, 1965'Tag,1988;Wendler, observedwinds exhibit an annualdirectionalrange of

1989a,b];theformer catch grains andcrystals, whilethelatter approximately 10ø withmoredownslope flow in winter eitherdetectindividual snowparticles by theirshadows on [Wendler, 1991]'neglect ofthiseffect willintroduce onlyaminor photosensitive semiconductors [Schmidt, 1977]ordepend onthe change in thecalculated transport ofdrifting snow. pulsecounting of anoscillating crystal.SuchmeasurementsTheprimitive equation model[Parish andBromwich, 1991] relatewindspeedto thefluxof driftingsnowat oneor more wasintegrated froma stateof resttonearsteady conditions after heights; integration withrespect toheight yieldsthetransport of 48 model hours.The resultingthree-dimensional mass driftingsnow. Normally,the logarithm of the transport is circulation describes thegravity-driven drainage of cold,nearexpressed asa function of windspeed orequivalently thecube of surface airata spatial resolution of 100km(Figure 5),aswellas thewindspeed is related to thedrifting snowtransport. This thetropospheric response tothisboundary layerdivergence. The means thatrelatively minorerrors in windspeed canleadto simulation represents well-developed drainage during thepolar

substantial errors in thecalculated transport of drifting snow.night(March-September) without thecomplications introduced Furthermore, theheight variation of fluxabove analtitude of a bycyclones andclouds. Comparisons between gridpoint surface fewmeters isnotwellknown, assuch measurements aredifficultwinds andobserved annual resultant winds atadjacent stations

to obtain. This makes the relationship between transportand wind speeduncertainfor very strongwind speeds,whenthe flux at greateraltitudesbecomesimportant. Also, a single relationship between drifting snow transport and wind speedcannotbe fixed because,in addition to wind speed,the sizedistributionof snowgrains(a grainmay consistof one or more crystals),the surfaceroughness,and the availability of loose snow are important.For example,Loewe [1970, p. 5], who made observationsin Ad61ieLand where extremely strong winds are experienced,states:

Occasionally,the density of the drift can becomevery small, even with very strong winds. After a prolonged period of gale from the samedirection,the snowsurface will become firmly compressed and so perfectly

showedthat directionswere well simulated,but modeled speeds (as expected) tended to be stronger than the observed annual values.Model wind speedswere multiplied by 0.62 (averageof seven ratios of observedspeed to model speed, with a mean deviation of 0.16) to more closely approximateclimatological

conditions. The mean annualwind speedis not well suitedto estimatethe amountof drifting snow. The wind-speeddistributionbecomes very important,as the transportis highly sensitiveto wind-speed

changes.Close to 70øS, there are two automatic weather stations,which report via satellite [Stearns, 1984]: GF 08 in westernWilkes Land at 68.5øS, 102.2øE at an elevationof 2118 m [Allison and Morrissey, 1983]; and D 80 in Ad61ieLand at

casewas experiencedat Port-Martinon 8 March, 1951, when the drift almostceasedalthoughthe wind remained

70.0øS,134.7øEat an elevationof 2500m [Wendleret al., 1983; Andrdet al., 1986]. Hourly valuesfrom GF 08 for 1987 andfrom D 80 for 1987 and 1988 were used.The model gaveonly mean annualwind velocitiesand no informationon the wind speed

of theorderof 40 m s-1withunchanging direction.

distribution.

streamlined

that no more snow can be removed.

Such a

Hence

we used these actual

measurements

to

GIOVINETTO ETAL.' ATMOSPHERIC WATER VAPOR TRANSPORT ACROSS 70oS

925

Fig. 5.Combined streamline and isotach depictions ofsurface winds (March-September) onthe icesheet simulated bythe model of Parish and Bromwich [ 1991].

describe thewindspeed distribution asa function ofthemeantheslopes anddownslope forcing onthecoldairgoes tozero

annual wind speed. Afrequency distribution isnotwelldescribed [Parish andBromwich, 1991]. In reality, theadjustment of byasingle parameter, butthisapproach wasdeemed acceptable katabatic winds beyond thegrounding linesappears tobea inviewoftheapproximate transport estimates being sought. function of theupstream boundary layermassflux [e.g., To summarize, thefollowing simple method wasusedto Bromwich, 1989],andcanhavea length scale ofuptoseveral

provide estimates ofdrifting snow transport. First, thecorrected hundred kilometers. These types ofmodels appear torealistically wind-speed data from themodel ofParish andBromwich [1991]represent many aspects of Antarctic katabatic windbehavior wereapplied toobtain anapproximation of themean annual[e.g., Bromwich etal.,1990] andqualitatively exhibit theabove windspeeds. Second, hourly datafromtheautomatic weather behavior beyond theslope break; therefore it isassumed thatthe stations near70øSwereutilizedto obtainthewindspeedmodel results provide afirst-order depiction ofthewindfield near

distribution in termsof the meanannualwind speed.Third, thebarrierof thoseiceshelves.

drifting snow models from three sources were used: namely Budd Thecorrected model winds were also used toestimate Di by etal. [1966]asreported byRadok[1970], Lister[1960]as prescribing thevariation ofannual speeds along thecoast lying reported byLoewe[1970],andKobayashi [1978];these modelssouthward of 70øS,withtheexception of sectors alongthe predict thedriftingsnowtransport giventhewindspeed. From barriers of theRoss, Ron_ne, andFilchner iceshelves. These large these datathenettransport ofdrifting snow wascalculated. ic, e shelves haverelatively flatareas withlength scales ofseveral In thecalculation of Do thecorrected modelwindswereused hundredkilometers,and a model that simulatesdownslope

as estimates of the annualwindsalong70øS from 1øW airflowcannot supply evenanapproximate description of typical

(eastward) to160øE andfrom 61øW to76øW.Theuseofmodelwindconditions. Theobserved winds along thebarriers ofthose winds between 1øWand30øE, approximately along thebarriericeshelves aregenerally directed offshore withannual resultant oficeshelves some 100kmtothenorth ofthegrounding linesspeeds being much less than 5 ms-1[Mather andMiller,1967].

where surface slopes characteristic ofgrounded icetopography There are notable exceptions tothis inthewestem Ross IceShelf end, isbelieved tobereliable because inthemodel theairstream [Savage andStearns, 1985], andintheFilchnet IceShelf, where

undergoes marked deceleration once drainage winds blow beyond marked offshore airflow withvector resultant speeds of

926

GIOVINETTOET AL.: ATMOSPHERIC WATERVAPORTRANSPORT ACROSS70øS

approximately 6 m s'1 axeobserved, paxtlyfed by inland

203

katabatic airflow [Bromwich, 1989]. More exceptionsmay becomeknown as the surfaceairflow alongthe barrier of the

18

Ronne Ice Shelf is studied in more detail. In this context our

estimateof Di is a minimum that may be compensated to an uncertainextentby the overestimates discussed below. There are someapproximations which mustbe madewith the

16 ¾

14

method describedabove. In constructingthe wind speed distribution, all measurements with windspeeds smallerthanthe E annualmean or laxget than three times the annualmean were ignored,i.e., the distribution is zerooutsidethisrange.The low •o wind speedsare unimportantbecausethey havelittle influence on driftingsnowtransport. The highwindspeeds, however,cause • problemsfor two reasons:first, theyarerare,thusit is difficultto fit a distributionto them;second,the driftingsnowmodelsare •-o not valid at thesehigh speeds.To fit the data, we produceda histogram according to the logarithm of the number of -J observations(Figure 6). A linear least squaresfit was then appliedand convertedto give a probabilitydistributionof the

12 10 8 6

4

form

p(u)= (wW•-(b/W) u

(5)

for thewindspeed,whereW is themeanannualwindspeed,u is thewindspeed,anda andb areconstants. Whenintegrated from W to 3W, the value obtainedrepresents the fractionof wind speedswhich fall in this interval. The drifting snowmodelswe usedare all of the form

0

10

20

30

40

50

Windspeed( ms-1) Fig. 7. Comparison of driftingsnowtransportasa functionof wind

In(D) = cu + d

(6)

speedpredicted by themodelsdiscussed in thetext(a: Buddet al. model from Radok [1970]; b: Lister modelfrom Loewe [1970]; c and d: Kobayashi models1 and2 fromKobayashi [1978]).

whereD is thetransport in gramspermeterpersecond, u is the wind speedin metersper second,and c andd are constants. madein a fairlysmallrangeof windspeeds, thesemodels donot These models are all derived from actual measurements of

drifting snow. Because there are considerabledifficulties in measuringdrifting snow and most of the measurementswere

showcloseagreement, particularly atwindspeeds above 25m s-1 (Figure7). For consistency, the originalformulaewereall converted to usethenaturallogarithm andunitsof gramsper meterper secondfor transportandmetersper secondfor wind

-2

speed. Windspeeds wereadjusted to aheightof 3 m, assuming a

roughness lengthof z0 = 1 mm.

Kobayashi [1978]givesa rangeratherthana singleformula; we processed the extremesof the rangeandrefer to themas

m -3

Kobayashi 1 and2. Lister's formula isgivenbyLoewe[1970],and the formulafromBuddet al. is givenby Radok [1970].The

ß

adjustedformulaeare as follows: Budd et al.

In(D)= 0.235u+ 2.72

(7)

In(D)= 0.407u- 0.345

(8)

Lister

Kobayashi1

In(D)= 0.397u- 0.276 Kobayashi2 In(D) = 0.397u - 2.12

(9) (10)

The estimates of Do across70øS arepresented in Table4.

Valuesforeachof thedriftingsnowtransport models aregiven

forsegments of 10ø of longitude. Thetransports areall northward,exceptin the northernPalmer Land. The valuesfor Kobayashi 1 and 2 indicate that this model would result in a

middle valueof 10TM kga'l, andtherefore theestimates range

-7100

,

150

,

200

,

250

ß .Xx300 between themodels of Buddet al.andLister,respectively 43 and

-•

206x 1012 kga'1.Thislaxge range emphasizes theorder-ofmagnitudenatureof the estimates.It was indicatedabovethat

% of mean annual windspeed themodels giveunrealistically largetransports at veryhigh Fig.6. Truncated windspeed distribution usedin theestimate ofdrifting speeds, andthusthemethodprobablyoverestimates the actual snowtransport.Data are from stationsD 80 and GF 08.

transports.Another aspect that leads to overestimationis the

GIOVINETTO ET AL.' ATMOSPHERICWATER VAPOR TRANSPORTACROSS?0oS

927

TABLE4. MeridionalDriftingSnowTransportAcross70øS(Do) for the Sectors0ø (Eastward)- 160øEand60øW (Westward) - 80øW(MeanValuesfor Subsectors of 10ø of Longitude)

Longitude

Budd et al. p

Transport in1012 k•a'1(Ne•,ative Southward)* Lister.•

Kobayashi1õ

Kobayashi2õ

0ø-10øE

0.69

0.39

0.36

0.057

10ø-20øE

0.59

0.24

0.22

0.035

20ø-30øE

0.13

0.11

0.10

0.016

30ø-40øE

0.61

1.10

0.89

0.140

40ø-50øE

0.98

1.50

1.30

0.210

50ø-60øE

2.00

Z20

1.90

0.300

60ø-70øE

2.80

4.10

3.50

0.560

70ø-80øE

0.10

0.02

0.02

0.003

80ø-90øE

3.90

14.00

11.00

1.800

90 ø-100øE

8.70

65.00

51.00

8.000

100ø-110øE

4.90

28.00

22.00

3.500

110ø-120øE

3.00

8.90

7.30

1.200

120ø-130øE

3.80

17.00

14.00

2.200

130ø-140øE

1.90

4.00

3.30

0.530

140o. 150øE

8.80

59.00

46.00

7.300

150ø-160øE

0.87

0.93

0.80

0.130

60ø-70øW

(-)0.59

(-)0.29

(-)0.26

(-)0.042

70ø-80øW

(-)0.35

(-)0.13

(-)0.12

(-)0.019

42.83

206.07

16331

25.92

Totals Adopted 43

206

95

* Usingwinddatafromthemodelof ParishandBromwich[ 1991]. '• Frommodeldescribed byRadok[1970]. $ Frommodeldescribed by Loewe[1970].

õ Maximumandminimumtransports froma rangedescribed byKobayashi [1978].

observed limitedsupplyof snowat the surface,alreadydescribed. distributionswith height and throughthe seasonsindicatesthat The preceding caveats withstanding,we estimate Do at a the meantemperature at condensation levelsandup to the 300-

midrange value ofthemodels, i.e.,approximately 0.12x 1015kgmbarlevelis approximately -30C foratleast9 months of the a~1(Table1). The sameprocedures wereusedto estimate Di. year[e.g.,Weyant,1966;Taljaardet al., 1969].Condensation Thevalues obtained usingthefourmodels were< 2 x 1012 kga'1 occurs at a relatively largetemperature rangeandthelatentheat andarethereforeignoredin this study. release(Lc) increasesas the temperatureat which condensation It shouldbenotedthatLoewe[1970],usingwinddatafrom 17 occursdecreases followinga simplerelationship [e.g.,Hare and stations,estimatedsnowdrift lossacrossthe ice terminusfor the Thomas,1979]:

wholeicesheetat 0.11x 1015kga-1. However, ourestimate excludes the ice sheet sector north of 70øS where it should be

expectedthat drifting snowlossis much greater.

Atmospheric Net Water Vapor and Latent Heat Transport

Lc = 3075- 2.13 T

(11 )

whereLc is in Joulesper gram andT is the air temperaturein kelvin at the moment of condensation. Assuming that condensation

occurs at 0 C would

result in a minimum

estimate

of Lc; this conservative estimate of Lc needs to be increased,

Across 70øS

however, because south of 70øSmostprecipitation occurs in solid Thesummation of surface balance in theicesheet,outflowin form(i.e., thereis a needto addthelatentheatreleased at 'the

theocean, anddrifting snowasdiscussed in preceding sections moment of freezing, or 335J g-l). Usinga latentheatrelease

results inanetsouthward vapor transport of2.88x 1015 kga-1or figure ofapproximately 2830J g-1results inanettransport of 6.6 kg m-1 s-1 (Table1). An examination of air temperatureapproximately 8.2x1021 Ja-1or 18.9x106jm-1s-1.

928

GIOVINETTO ET AL.' ATMOSPHERIC WATERVAPORTRANSPORT ACROSS 70oS

A first approximationof the compositeerror in the estimates This researchwas supportedin part by the Natural Sciencesand may be assessed consideringthe relative standarderrorsin each Engineering Research Councilof CanadaundergrantOGP0036595 to the of the principalterms: Universityof Calgary and by the National ScienceFoundationunder

1. An errorof 10%in theestimate of B (.'k0.15x 1015kg

grantsDPP-8916134 to the Ohio State Universityand DPP-8714828to

a-1) as was assessed for the wholeice sheet [Giovinettoand the University of Alaska. Byrd Polar ResearchCenter, Ohio State Bentley, 1985]. University,contribution770. 2. An error of 70% in the estimateof Do, as indicatedby the rangeof driftingsnowtransportusingthemodelsdiscussed above REFERENCES

(• 0.08x 1015kg

3. An error in the estimateof F that is largerthan that for B

Alley, R. B., and C. R. Bentley, Siple Coast fire and ice studies'

andsmaller thanthatforDo,say40%(ñ 0.51x 1015 kga-l). Conclusion and prospects, Antarc. J. U.S.,XXIII, 5,58-59,1988. Thisadmittedly cursory approach indicates a relative error Allison, I. F.,A sample study of theenergy fluxes preceding and nearlyequalto the errorthat wouldbe derivedfrom accompanying theformation ofAntarctic sea ice,Energy Fluxes over Polar Surfaces,Tech. Note 129, 115-132, World Meteorol. Org. , uncertaintiesdiscussedin the estimatesof precipitationover Geneva,1972. large ocean areas [e.g., Baurngartner and Reichel, 1975] and Allison, I. F., and J. V. Morrissey,Automaticweatherstationsin the latent heat flux over ice-free ocean and sea ice areas [e.g.,

Zillrnan, 1972; Weller, 1980]. The summation of errors in the estimates of B, Do, and F,

treatedas standarderrors,indicatesa compositeerrorof + 0.54 x

Antarctic,Aust.Meteorol.Mag., 31,71-76, 1983. Allison, I. F., N. W. Young, and T. Medhurst,On reassessment of the massbalanceof the LambertGlacier drainagebasin,Antarctica,J.

1015 kga'1or:t:19%forthebulkrateestimate ofnetwatervapor

Glaciol.,31,378-381, 1985.

transport. Thiscomposite errorissmaller thanthatwhich couldAlt,J.,P.D.Astapenko, and N.J.Ropar, Some Aspects oftheAntarctic beassessed, atpresent, foranestimate of V using atmospheric Circulation in 1958, Int.Geophys. YearGen.Rep.Set.,vol.4, NationalAcademyof Sciences, Washington, D.C., 1959. Andrt, J., G. Wendler, and M. Zephoris, The IAGO KatabaticProgramme, errorsof + 1.6x 1021J a'l for thebulkrateestimate of netlatent in Antarctic Climate Research, vol. l, 17-18, SCAR Group of heattransport andof + 1.3kg m'l s-1 and+ 3.6x 106jm-1s'l Specialistson Antarctic Climate Research, Antarctic Division, for the correspondingrate-over-lengthestimates. Kingston,Australia, 1986. In the comparisonwith previousestimatesof V listedin Table Andreas,E. L., andA. P. Makshtas,EnergyexchangeoverAntarcticseaice 2 the compositeerror indicatesa rangebetween- 5.3 and - 7.9 kg Res.,90, 7199-7212,1985. m-1 s-1' The lowervalueof thisrangeis equalto thelatest in thespring,J. Geophys. Andreas, E. L., C. A. Paulson,R. M. Williams, R. W. Lindsay, and estimateusing atmosphericdata [Masuda, 1990], and the upper J. A. Businger,The turbulentheat flux from Arctic leads,Boundary value is approximately10% larger than the uppervalue of the Layer Meteorol.,17, 57-91, 1979. first estimate using surface data [Rubin, 1962]. Overall, the Baumgarmer,A., and E. Reichel, The World Water Balance, Elsevier,

data. Applied to the other estimates,the relative error indicates

resultsof our study fit the chronological(as opposedto real or physical)trendof increasingvaluesin the estimatesof net water

New York, 1975.

Behrendt,J. C., Antarctic Peninsulatraverse,geophysicalresultsof vapor transport across70øS reportedin the last two decades, glaciologicaland geologicalstudies,Res.Rep. 64-1, Geophys.and

whichwere basedon atmospheric data.In the contextof physical PolarRes. Cent. Univ. of Wis., Madison, 1965. trends, however, it appears that in the last decadessurface Bishop,J. F., andJ. L. W. Walton,BottommeltingunderGeorgeVI Ice balancehas increasedsignificanfiyat a numberof locationsin Shelf, Antarctica,J. Glaciol.,27, 429-447, 1981. the ice sheet [Pourchet et al., 1983' Peel and Davison, 1988; D.

Bromwich,D. H., Snowfall in high southernlatitudes,Rev. Geophys.,

M. Ethridge, personalcommunication,1991]. Since the balance 26, 149-168, 1988. datausedto draw isoplethsasshownin Figure3 consistof means Bromwich,D. H., Satelliteanalysesof Antarctickatabaticwindbehavior, of from a few to severalyearsdistributedbetween1940 and 1980 Bull. Am. Meteorol. Soc., 70,738-749, 1989. [Giovinettoand Bull, 19871, the possibilityof an actualincrease

in the annualwater vapor transportacross70øS shouldbe

Bromwich,D. H., Estimates of Antarcticprecipitation, Nature,343,627629, 1990.

examinedin the future. At presentwe note that the resultsof our Bromwich, D. H., T. R. Parish,and C. A. Zorman, The confluencezoneof studyare in generalagreementwith estimatesof net water vapor the intensekatabaticwinds at Terra Nova Bay, Antarctica,as derived

transportacross70øS reportedin the last threedecadeswhich were basedon surfacedata.This agreementis explained,in part,

fromairbornesastrugi surveys andmesoscale numericalmodeling, J. Geophys. Res.,95(D5), 5495-5509,1990.

bytheusein all estimates of largely thesame surface balanceBrown. R.A.,Meteorology, inPolarOceanography, Part A,Physical data base, as statedin a precedingsection,and in part by our extrapolationof the values of V estimated for Antarctica by

Science, edited byW. O.SmithJr.,1-46,Academic, SanDiego,Calif., 1990.

Rubin[1962]andBrornwich[1990]to theareapolewardof 70ø - Bryan,F., andA. Oort,Seasonal variation of theglobalwaterbalance 90øS, usingratiosobtainedfrom the analysisof Baumgartner based onaerological data,J.Geophys,Res., 89, 11,717-11,730, 1984. snowparticles, in Studies in and Reichel [1975]. Moreover,the uppervalueof V given by Budd,W. F., Thedriftingof nonuniform Antarctic Meteorology,Antarc. Res. Ser., vol. 9, editedby Rubin [1962] is basedon the largestestimateof net northward M. J. Rubin,59-70,AGU, Washington, D.C., 1966. transportof driftingsnowavailableat that time;thisuppervalue wouldbe greatlyreducedif the driftingsnowtransportmodels Budd,W. F.,W. R. J.Dingle,andU. Radok,TheByrdSnowDriftProject: usedin our studywere applied.

outlineand basicresults, in Studiesin Antarctic Meteorology,

Antarc.Res.Serv.,vol. 9, editedby M. J. Rubin,71-134,AGU, Acknowledgments.The authorswish to thank C. R. Bentley, U. Radok, and C. R. Steams for information leading to sources, T. R. Parish for assistancewith extractionof wind outputfrom the

numericalmodel,and threeanonymous reviewersfor their suggestions.

Washington, D.C., 1966.

Carleton,A.M., Ice-ocean-atmosphere interactions at high southern latitudesin winter from satelliteobservation,Aust. Meteorol. Mag., 29, 183-195,1981.

GIOVINETTO ET AL.: ATMOSPHERIC WATER VAPOR TRANSPORT ACROSS 70oS

Carsey,F., Microwaveobservations of the WeddellSea polynya,Mon.

929

Lettau, B., The transportof moistureinto the Antarcticinterior, Tellus, 21,331-340, 1969.

Weather Rev., 108, 2032-2044, 1980.

Comiso,J. C., andA. L. Gordon,Recurringpolynyasover the Cosmonaut Lettau,H. H., A casestudyof katabaticflow on the southpolarplateau, SeaandtheMaudRise,J. Geophys. Res.,92(C3), 2819-2833,1987. in Studiesin AntarcticMeteorology,Antarc. Res.Serv.,vol. 9, edited Drewry, D. J., The surfaceof the Antarctic ice sheet,in Antarctica: by M. J. Rubin, 1-11,AGU, Washington,D.C., 1966. Glaciologicaland GeophysicalFolio, editedby D. J. Drewry, sheet2, Lettau, H. H., and W. Schwerdtfeger,Dynamicsof the surface-wind ScottPolarResearchInstitute,Universityof Cambridge,Cambridge, regime over the interior of Antarctica,Antarc. J. U.S., 2, 155-158, 1967. England,1983. Fletcher,J. O., The heat budgetof the Arctic Basin and its relation to Lister,H., Glaciology,PartI: Solidprecipitationanddrift snow,Transclimate, R-444-PR, The Rand Corporation,Santa Monica, Calif., AntarcticExpedition,1955-1958,Scient.Rep.No. 5, Trans-Antarctic 1965. ExpeditionCommittee,London,1960. Fletcher,J. O., The heat budgetof the SouthemOceanand its relationto Loewe,F., The transportof snowon ice sheetsby the wind, in Studieson climate,RM-5793-NSF,The RandCorporation, SantaMonica,Calif., Drifting Snow,publ. 13, 1-69, Departmentof Meteorology,Univ. of 1969.

Melboume, Australia, 1970.

Fujiwara,K., andY. Endo,Preliminaryreportof glaciologicalstudies,in Reportof theJapaneseTraverseSyowa-South Pole 1968-69, editedby M. Murayama, 68-109, Polar Research Center, National Science Museum,Tokyo, 1971.

Masuda,K., Atmosphericheat and water budgetsof polar regions: Analysisof FGGE data,in Proceedings of theNationalInstituteof Polar ResarchSymposium on Polar Meteorologyand Glaciology,vol. 3, 79-88,NationalInstituteof PolarResearch, Tokyo,1990.

Giovinetto, M. B., Mass accumulation in West Antarctica, Eos Trans.,

Mather,K. B., andG. S. Miller, Noteson topographic factorsaffecting the surfacewind in Antarctica,with specialreferenceto katabatic winds,andbibliography, Rep. UAG-189,Geophys. Inst.,Univ. of

AGU, 42,386-389, 1961.

Giovinetto,M. B., Distributionof diageneticsnowfacies in Antarctica andin Greenland,Arctic, 17, 3240, 1964a.

Alaska, Fairbanks,1967.

Giovinetto,M. B., The drainagesystemsof Antarctica:accumulation,in Antarctic Snowand Ice Studies,Antarc. Res. Serv., vol. 2, edited by M. Melior, 127-155, AGU, Washington,D.C., 1964b. Giovinetto, M. B., and C. R. Bentley, Surface balancein ice drainage systemsof Antarctica,Antarc.J. U.S., 20 (4), 6-13, 1985. Giovinetto, M. B., and C. Bull, Summaryand analysisof surfacemass balancecompilationsfor Antarctica,1960-1985,Rep. 1, Byrd Polar Res. Cent., Ohio StateUniv., Columbus,1987.

Gow, A. J., andW. B. TuckerllI, Seaice in the polarregions,in Polar Oceanography,Part A, PhysicalScience,editedby W. O. Smith Jr., 47-122, Academic,SanDiego,Calif., 1990. H•ikkinen,S., Models and their applicationsto polar oceanography,in Polar Oceanography, PartA, PhysicalScience,editedby W. O. Smith Jr.,335-384, Academic,SanDiego,Calif., 1990. Hare,F. K., andM. K. Thomas,ClimateCanada,JohnWiley, 2nd.ed.,New York, 1979.

Hollingsworth, A., The global weather experiment - 10 years on, Weather,44,278-285, 1989. Howarth, D. A., Seasonalvariations in the vertically integrated water

vaportransportfields over the southernhemisphere, Mon. Weather Rev., 111, 1259-1272, 1983. Howarth, D. A., and J. N. Rayner, Estimatesof sourcesand sinksof atmospheric moisture in the southern hemisphere, in Second International Conferenceon SouthernHemisphereMeteorology, 163-166,AmericanMeteorologicalSociety,Boston,Mass.,1986. Kep, S. L., A climatologyof cyclogenesis, cyclonetracksandcyclolysisin the Southern Hemispherefor the period 1972-1981, Publ. 25, Meteorol.Dep., Univ. of Melboume,Parkville,Victoria, 1984. Kobayashi,D., Studiesof snowtransportin low-level drifting snow, Contrib. Inst. Low Temp. Sci., Hokkaido Univ., Set. A (24), 1-58, 197Z

Kobayashi,S., Snowtransportby katabaficwindsin Mizuho Camparea, EastAntarctica,J. Meteorol. Soc.Jpn., 56, 130-139, 1978. Korzoun, V. J., A. A. Sokolov, M. I. Budyko, K. P. Voskresensky, G. P. Kalinin, A. A. Konoplyantsev, E. S. Korotkevich, and M. I. Lvovich (Eds.), Atlas of the Worm Water Balance, USSR Committee for the International Hydrological Decade, HydrometeorologicalPublishingHouse,UNESCO, Paris, 1977. Landon-Smith,I. H., and B. Woodberry,The photo-electricmeteringof windblown

snow,

Australian

National

Antarctic

Research

ExpeditionsInter. Rep. Ser. A(IV), Glaciol.Publ., 79, 1-21, 1965.

Maykut,G. A., Energyexchange overyoungseicein thecentralArctic,J. Geophys.Res., 83, 3646-3658, 1978.

Maykut,G. A., Thesurface heatandmassbalance, in TheGeophysics of Sea Ice, edited by N. Untersteiner,385-463, Plenum,New York, 1986.

Mellor, M., and U. Radok,Somepropertiesof drifting snow,in AntarcticMeteorology,333-346,Pergamon, New York, 1960. Mcintyre, N., A re-assessment of the massbalanceof the LambertGlacier drainagebasin,Antarctica,J. Glaciol.,31,381-382, 1985.

Naruse,R., Measurement of driftingsnowon the coastalregionof Antarctica,nearSyowaStation(in Japanese with Englishabstract), Contrib.Inst.Low Temp.Sci.,HokkaidoUniv. Set.A (28), 147-154, 1970.

NavalOceanography Command Detachment, Seaiceclimaticarias,vol.I, Antarctic,Rep.NAVAIR 50-1C-540, 1985. Nishio, F., T. Katusushima,and H. Ohmae,Volcanicashlayersin bareice areasnear the Yamato Mountains,DronningMaud Land and AUan Hills, Victoria Land, Antarctica,Ann. Glaciol., 7, 34-41, 1985.

Oefiemans,J., Responseof the Antarcticice sheetto a climaticwarming: a modelstudy,J. Clirnatol.,2, 1-11, 1982. Parish, T. R., A numerical study of strong katabatic winds over Antarctica, Mon. Weather Rev., 112,545-554, 1984. Parish, T. R., and D. H. Bromwich, The surface windfield over the

Antarcticice sheets,Nature, 328, 51-54, 1987. Parish, T. R., and D. H. Bromwich, Continental-scale simulation of the

Antarctickatabaticwind regime,J. Climate,4, 135-146, 1991. Parish,T. R., andK. T. Waight,The forcingof Antarctickatabaticwinds, Mon. Weather Rev., 115, 2214-2226, 1987.

Peel,D. A., andB. M. Davison,Ice coresandclimate,Ice, 88, 23-25, 1988.

Peixoto,J.P. andA. H. Oort,Theatmospheric branchof thehydrological cycleandclimate,in Variationsin theGlobalWaterBudget,editedby A. Street-Perrott, 5-65, D. Reidel,Hingham,Mass.,1983. Picciotto,E., G. Crozaz, and W. DeBreuck,Accumulationon the South Pole-QueenMaud Land traverse,1964-1968,in AntarcticSnowand Ice StudiesH, Antarc.Res.Set., vol. 16, editedby A. P. Crary,257315, AGU, Washington,D.C., 1972. Potter,J. R., J. G. Paren,andJ. Loynes,Glaciological andoceanographic calculations of themassbalanceandoxygen-isotope ratioof a melting ice shelf,J. Glaciol., 30, 161-170, 1985. Pourchet, M., F. Pinglot,andC. Lorius,Somemeteorological applications

930

GIOVINETTO ET AL.: ATMOSPHERICWATER VAPOR TRANSPORTACROSS70oS

of radioactivefallout in Antarcticsnows,J. Geophys.Res., 88 (C10),

Taljaard,J. J., H. van Loon, H. L. Cmtcher,andR. L. Jenne,Climate of the

6013-6020, 1983.

upperair: southernhemisphere, vol. I, Temperatures, dewpoints,and heightsat selectedpressurelevels, Rep. NAVAIR 50-1C-55, Nay. WeatherServ.Command,Washington, D.C., 1969. Trig, H., A pulse-countingtechniquefor the measurement of drifting

Radok,U., Boundaryprocesses of driftingsnow,in Studieson Drifting Snow, Department of Meteorology, University of Melbourne, Australia,publ. 13, 1-20, 1970. Rubin, M. J., Atmosphericadvectionand the Antarcticmassand heat budget, in Antarctic Restarch: The Matthew Fontaine Maury Memorial Symposium,Geophys. Monogr. Ser., vol. 7, edited by H. Wexler, et al., 149-159, AGU, Washington,D.C., 1962. Rubin, M. J., and M. B. Giovinetto, Snow accumulationin central West

snow, Ann. Glaciol., 11, 184-186, 1988.

Washington,W. M., A. J. SemtnerJr., C. Parkinson,and L. Morrison,On the developmentof a seasonalchange sea-ice model, J. Phys. Oceanog.,6, 679-685, 1976.

Weller, G., Spatialand temporalvariationsin the southpolar surface energybalance,Mon. WeatherRev.,108, 2006-2014, 1980. Antarctica as related to atmosphericand topographicfactors,J. Wendler,G., On the blowingsnowin Adtlie Land, EastemAntarctica,in Geophys.Res.,67, 5163-5170,1962. Proceedingsof Symposiumon Glacier Fluctuations and Climatic Rubin,M. J., andW. S. Weyant,The massandheatbudgetof the Antarctic Change,Amsterdam,June 1987, edited by J. Oedemans,261-279, atmosphere, Mon. WeatherRev.,91,487-493, 1963. Kluwer Academic, Boston,Mass., 1989a. Savage,M. L., and C. R. Steams,Climate in the vicinity of RossIsland, Wendler, G., Measuring blowing snow with a photoelectricparticle Antarctica,Antarc. J. U.S., 20 (1), 1-9, 1985. counterat PoleStation,Antarctica,Polarforschung, 59, 9-16, 1989b. Seko, K., T. Furukawa, and O. Watanabe, The variation of surface Wendler, G., Stronggravity flow observedalong the slopeof eastern conditions on the Antarctic ice sheet, in Proceedings of the Antarctica,Meteorol.Atmos.Phys.,43, 127-135, 1991. International Conferenceon the Role of the Polar Regionsin Global Wendlet,G., Y. Kodama,and A. Poggi,Katabaticwind in Ad61ieLand, Change, June 11-15, 1990, Univ. of Alaska, Fairbanks,Alaska, in press.

Shimizu, H., Glaciological studiesin West Antarctica, 1960-1962, in Antarctic Snowand Ice Studies,Antarc. Res. Serv., vol. 2, editedby M. Mellor, 37-64, AGU, Washington,D.C., 1964. Schmidt,R. A., A systemthat measuresblowing snow, Res.Paper, RM194, U.S. Dep.of Agric.ForestServ.,Washington, D.C., 1977.

Antarc.J. U.S., 18(5), 236-238, 1983. Wexler, H., Seasonaland other temperaturechangesin the Antarctic atmosphere, Q. J. R. Meteorol.Soc.,85, 196-208,1959. Weyant,W. S., The Antarcticatmosphere: Climatologyof the surface environment, in Antarctic Map Folio Series, Folio 8, edited by V. Bushnell,AmericanGeographical Society,Washington, D.C., 1966.

Schwerdtfeger, W.,Theclimate oftheAntarctic, inClimates ofthePolar Zillman, J.W., Solarradiation andsea-air interaction south of Australia, Regions,World Surv.Climatol.,vol. 14, editedby S. Orvig, 253-355, Elsevier, New York, 1970.

Starr, V. P., J.P. Peixoto, and R. G. McKean, Pole-to-polemoisture conditionsfor the IGY, Pure Appl. Geophysi.,75,300-331, 1969.

in Antarctic Oceanology H, The Australian-New Zealand Sector, Antarc. Res. Ser., vol. 19, edited by D. E. Hayes, 11-40, AGU, Washington, D.C., 1972.

Zwally,H. J., J. C. Comiso,C. L. Parkinson, W. J. Campbell,F. D. Carsey, and P. Gloersen,AntarcticSea Ice, 1973-1976: Satellite passive1983-1984,Antarc. J. U.S., •9(5), 189-191, 1984. Steams,C. R., and G. Wendler, Researchresultsfrom Antarcticautomatic microwaveobservations, NASA Spec.Publ.,NASASP-459,1983. weatherstations,Rev. Geophys.,26, 45-61, 1988. Streten, N. A., Satellite observationsof the summer decay of the D. H. Bromwich,Byrd PolarResearchCenter,Ohio StateUniversity, Antarcticsea-ice,Arch. Meteorol. Geophys.Bioklimatol.,Ser. A, 22, Steams, C. R., Antarctic automatic weather stations, austral summer

119-134, 1973.

Streten,N. A., andA. J. Troup,A synopticclimatologyof satellitecloud vorticesoverthesouthemhemisphere, Q. J. R. Meteorol.Soc.,99, 5672, 1973.

Columbus, OH 43210.

M. B. Giovinetto,EarthSciences356, Universityof Calgary,Calgary, Alberta, Canada T2N 1N4.

G. Wendlet, GeophysicalInstitute,Universityof Alaska, Fairbanks, Alaska 99701.

Taljaard,J. J., Synopticmeteorologyof the southernhemisphere,in Meteorology of the SouthernHemisphere,Meteorol. Monogr., vol. 13, no. 35, edited by C. W. Newton, 139-213, American MeteorologicalSociety,Boston,Mass., 1972.

(Received April25, 1991; revisedSeptember 13, 1991; accepted September 17, 1991.)