JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 100, NO. C3, PAGES 4593-4612, MARCH 15, 1995
Ice pack and lead surface energy budgets during LEADEX 1992 DominiqueRuffieux and P. Ola G. Persson Cooperative Institutefor Researchin Environmental Sciences, Universityof Colorado,Boulder C. W. Fairall
and Daniel E. Wolfe
EnvironmentalTechnologyLaboratory,NOAA, Boulder,Colorado
Abstract. During a 1-monthdeploymentfor the Arctic LeadsExperiment(LEADEX) in March and April 1992 on the Arctic ice cap roughly200 km northof PrudhoeBay, Alaska, surface-based mean meteorologicaland flux instrumentsplus a variety of remote sensorswere operatedat the main basecamp. Identical systemswere also deployedby helicopteron the upwindand downwindedgesof severalArctic leads,two of which we describein this paper. At the basecampthe diurnalamplitudefor sensibleheat flux
was+10 W m-2abouta meanof-3 W m-2,net radiationwas+30 W m-2abouta meanof
-15 W m'2,andnetsurface energyflux was+20 W m-2abouta meanof-12 W m-2. The mean latent heat flux was +1 W m-2 with a diurnal variation of about +1.5 W m-2. Mean
valuesfor the momentumand sensibleheat transfercoefficientswere Co = (1.20 + 0.20) x
10'3andCn= (0.75+ 0.25)x 10-3at a 10-mreference heightwithonlymodest diurnal variations. Two lead deploymentswere examined. Lead 3 was approximately1 km across. Only limited meteorologicaldata were obtainedfor about6 hoursat the end of April 7 andbeginningof April 8 whenthe lead wascoveredwith about10 cm of ice. Downwindof the lead,the sensibleheatflux increasedto about170 W m-2andthe stress
doubled, suggesting anice-covered lead10-mdragcoefficient of 2.2 x 10-3. Morethan 36 hoursof data were obtainedupwind and downwindof lead 4, which varied in width from 80 to 120 m. Doppler minisodarsupwind and downwindof the lead indicateda doublingin the depth(5 to 10 m) of the shear-driventurbulentsurfacelayer downwind of the lead and an intensificationof intermittentwave interactionsexceeding60 m (sodar rangemaximum). Three prominentwaveswith strongdownwardmotion were observed in this period,apparentlycausingincreasesin the downwindstressmagnitude. Various sourcesof data were usedto computeestimatesover a 36-hourperiodof the net surface
heatfluxQgoverthelead,theadjacent packice,andanyopenwaterthatmighthave occurredin the lead. The resultsindicatethat oncesignificantice forms,the sunis increasingly moreeffectivein reducingthe surfacefreezingrate andin shuttingoff convectivemixing in the oceanunderthe lead. Over the periodof observations the
average netsurface heatflux was-75 W m-2overthepackice,-130W m-2overthelead, and-250W m'2overtheopenwater. 1.
Introduction
Atmosphericforcingplaysa majorrole in themaintenance anddynamicsof the Arctic packice. This includesthe basic heatbalanceof the packice, the synopticice motion,andthe creationof local ice stressthat opensand closesleads. Our lack of knowledge about Arctic atmosphericphysical phenomenaaffects our ability to forecastArctic weather, in general,andatmospheric boundarylayer(ABL) processes, in particular. Climate generalcirculationmodels (GCMs)
and synoptic,regiona!,and mesoscalenumericalweather
modelgrid size. In somecasesthe subgrid-scale processes are as important as the explicitly resolved processes. Of particularconcernfor the Arctic are (1) the influenceof low-level clouds and moisture at low temperature;(2)the importanceof radiative transferprocesses;(3) the effect of intermittentturbulent transportin the ABL, the inversion layer, and above; and (4)coupling of the ABL dynamics with the intermittentforcingprovidedby leads. Detailed sea ice modelsare still in the early stagesof development[e.g., Holland et al., 1993;Ebert and Curry, 1993], andfundamental knowledgeaboutthe physicalprocesses influencingthe surfaceheat budget of the Arctic is still inadequate. For
forecastmodels all contain parameterizationsof physical processes that take placeon smallscalescomparedwith the example, we do notknowif the•tverage radiativeeffectof cloudsis to cool or heat the ice cap. Anothermajor source of uncertaintyis the overalleffectsof leads. Copyright1995 by the AmericanGeophysicalUnion. Leads representa major sourceof nonuniformityin the surfaceforcing of the ABL in the Arctic. This is in sharp Papernumber94JC02485. 0148-0227/95/94JC-02485 $05.00
contrast to the much more uniform surface of the ocean and 4593
4594
RUFFIEUX
ET AL.: ICE PACK AND LEAD SURFACE ENERGY BUDGETS
(designatedleads 3 and 4) provideduseful meteorological data. We also maintaineda continuouslyoperatingsurface air and lead water of 20ø-40øC can result in as much as and ABL sensingstationat the main LEADEX basecamp. 1000W m'2 of sensibleand latentheatexchange to the Thesedata are beingusedto investigatethe physicsof the atmosphere[Badgley, 1966; Vowinckeland Orvig, 1973]. surfaceenergyand stressbudgetof the pack ice, including Such rates of exchangeare globally quite rare. Although the role of clouds,and of Arctic ABL physicalprocessesand leadsrepresentonly a few percentof the Arctic surfacearea, their impactin mesoscalenumericalmodels[Thompsonand they contributeabout50% of the sensibleheattransferfrom Burk, 1991]. the Arctic Ocean to the atmosphere. This energy transfer Details and backgroundmaterial about the LEADEX warms and moistensthe atmosphereand coolsthe oceanin experimentcan be found in the literature[LEADEX Group, a highly localizedprocess.By triggeringfog and clouds,the 1993], so here we discussonly the ETL effort. We begin lead can severelyperturbthe net radiativebalance;the lead with a generaldiscussion of experimentaldetails(section2). is also a sourceof buoyancythat generatesturbulenceand The meanand diurnalpropertiesof the meteorologyand the' mixing. The evolution of the plume of warmer, moister, componentsof the surfaceenergy budget are discussedin more turbulent air as it is advected over the downwind ice section3. In section4 we describeextensivelythe resultsof surfaceand interactswith the overlyingboundarylayer has, the deploymentto lead 4; resultsfrom lead 3 are muchmore sofar, beenstudiedin detail only by usingnumericalmodels limited and are discussedonly briefly. In section5 we [e.g., Glendeningand Burk, 1992]. examinethe implicationsof the lead 4 resultsand contrast Severalaircraft and ice-basedfield programshave been the heat budgetover the lead versusthat over the pack ice executedto studyleadsand polynyasfrom the atmospheric and openwater (if it existed). Our conclusions are given in perspective(seeSmithet al. [ 1990] for an extensivereview). section 6. Simple bulk parameterizationshave been developed to representthe sensibleand latent turbulentheat fluxes over the lead in terms of the wind speed,air temperature,and 2. Experimental Background humidity upwind of the lead [Andreaset al., 1979; Andreas, 2.1. SensorDescriptionsand Deployment 1980; Smith et al., 1983]. The growth of the internal Mean winds, turbulent stress,and sensibleheat flux were boundarylayer (IBL) as a functionof fetchover the lead has been examinedboth experimentally[Andreaset al., 1979] measuredwith special ruggedized sonic anemometer/ and with a turbulent closure model [Lo, 1986]. However, thermometers [Kairnal and Gaynor, 1991]. Becausethe theseflux parameterizations are only valid while the lead is atmospheric humidityconcentration is solow, correlations of composedof totally exposedocean water. It turns out that verticalvelocityandsonictemperature fluctuations provide moreheatmay be lostfrom the typicalleadin the few days a direct measurement of the sensible heat flux. Sonic following the first surfaceice formation than in the brief anemometers have been used in the Arctic since the 1970s period it is completelyopen water. This issueis uncertain [Banke et al., 1976; Anderson, 1987], but the newer instrubecausemostestimatesof lead heatlossusethe openwater mentsprovide continuous,unattendedoperationover long parameterizations.To our knowledge,directmeasurementsperiods. The use of sonic temperature[Schotanuset al., of heat transferover the entire life cycle of a lead have not 1983], ratherthana fragile thermistoror thermocouple, the been done. Furthermore, the latent and sensible heat fluxes relative lack of transducerdrift effects, and the general are not the only factorsin the surfaceenergybudgetof the advantagesof digital processing all contribute to this used lead. Bothlong-waveandsolarradiativefluxesare quite improvedreliability. For example,sonicanemometers relevant. The higherskin temperature of the lead increases in the multiyear Arctic Ice Dynamics Joint Experiment thermal radiative heat loss, and the much lower albedo (AIDJEX) program[Bankeet al., 1980] producedabout100 greatlyincreasessolarinput in the nonwinterseasons.The stressand 25 sensibleheat flux values (nominally 1-hour greatdiversity of iceforms,thelackof horizontal homogene- averages),whereasthe 1-monthdeploymentat thebasecamp ity, andthecomplexinteractions of ice,horizontal drag,and duringLEADEX producedmore than 600 valuesof each. to the more typical forms of surface inhomogeneityover land. During the winter, temperaturedifferencesbetweenthe
turbulent mixingin leadscertainlydoomsthesimplemodels. Radiative fluxes were measured with conventional Simplystated,muchmoreinformation thantheupwindbulk commercialradiometers(pyranometersand pyrgeometers); parameters is requiredto characterize the surfaceenergy mean air temperatureand relative humidity (RH) were budgetof a real lead.
measuredwith commercialtemperature-RHunits(HMP35C
Followingseveralyearsof instrumentation andtechnique probes) specifically designedfor operation over a wide development,the polar researchgroup in the National temperaturerange. Three identicalflux/bulk meteorology Oceanic and AtmosphericAdministrationEnvironmental systemswere fabricated,consistingof a sonicanemometer,
Technology Laboratory (ETL) (namedtheWavePropagationfour radiometers,and a mean temperature-humidity sensor Laboratory at the time of the experiment) participated in a with a data acquisitionsystem,one for the basecampand large cooperative effort to investigate both atmospherictwo (one upwindand one downwind)for the lead deployandoceanographic processes as part of the Arctic Leads ments[Wolfeet al., 1992a]. The radiometersweretwo pairs Experiment (LEADEX). The ETL effort involved the investigation of the heatbudgetof bothactiveleadsand the
of upwardanddownwardpointingshort-waveandlong-wave sensors. The short-wavesensorswere Eppley precision background packice. Our approachincludeda "lead-scale" spectralpyranometers with a windowof 0.285-2.8 grn. The deploymentof sensorsat selectedArctic leadsto determine long-wavesensorswere Eppley precisioninfraredpyrgeo(asdirectlyasfeasible) thefluxesat theleadandto studythe meterswith silicondomesand with a window of 4-50 [an. atmospheric plumegeneratedby the lead. Four leadswere Pyrgeometerdata are postcorrected for variationsin the case examinedby the LEADEX team, but only the last two and dome temperatures[Albrechtand Cox, 1977].
RUFFIEUX
ET AL.:
ICE PACK AND LEAD
Three portable, high-frequencyDoppler sodars(minisodars)for measuringprofilesof verticalwind and smallscaletemperature turbulencefrom 5- to 100-m heightswere alsodevelopedwith a similardeploymentstrategy[MurschRadlgruber and Wolfe, 1993; Mursch-Radlgruberet al., 1994]. Sodar used in previousArctic experiments[e.g., Carsey, 1980; Chueng, 1991] typically lack the vertical resolutionto resolvethe importantnear-surface phenomena. At the basecampwe alsodeployeda UHF Dopplerradar wind profiler with radio acousticsoundingsystem(RASS) [see Strauchet al., 1988] for continuoussamplingof wind and temperatureprofiles at ABL scalesand a rawinsonde systemfor samplingthefull troposphere [Wolfeet al., 1992b; Persson et al., 1992]. More information on instruments is
providedin Table 1. The lead instrumentswere deployed by helicopter to selectedsites. A completeinstrumenthut was deployedon
SURFACE
ENERGY
BUDGETS
4595
the downwindedge of a lead, and one packageof micrometeorologicalinstrumentswas placedupwind on the ice. Sufficient cable length was provided to allow the datalogging computersto be kept in the hut. For lead 3 the sonicanemometer wassiteddirectlyat the ice edge,whereas for lead 4 it was about20 m from the ice edge. The second instrument packagewas deployedon the upwindsideof the lead with the majority of the oceanographic huts. In this casethe computerswere in a speciallyconstructed insulated
container thatwasjust placedon the ice andoccasionally caredfor by cooperativeoceanographic colleagues.Both upwind and downwind sonic anemometerswere situatedat
2 m abovethe ice; the temperature-humidity sensorwason the sametripodat a heightof 1.9 m. For lead 3 therewas a 0.8-m dropfrom the ice to the lead;for lead4 the drop was negligible. Lead 3 was approximately1 km in width, and lead 4 was approximately0.1 km in width.
Table 1. National Oceanic Atmospheric Administration Environmental Technology Laboratory LEADEX Instrumentation
Instrument
Profiler,915 MHz RASS, 2 kHz Laser ceilometer
pressure
sensor
Sodar,*6.5 kHz
Sonicanemometer*
Parameters
U, V, W, C2n, zi
HeightRange (RangeResolution)
AveragingTime (SampleInterval)
0.1-4km
1 hour
(0.1, 0.4 km)
(1.5 min)
Tv
0.1-1.5 km (0.1 km)
5 min eachhour (30 s)
Cloud-baseheight aerosolbscat
0-6 km (15 m)
30 s (30 s)
P'
surface
1s
C2t,W, zi
0-0.1km
5 s- 5 min
acousticbscat
(0.4 m)
(5 s)
U, V, W, Tv
2-3 min
15min (10 Hz)
Pyranometer,* 0.285-2.8pm
R•., R•,
0.6m
(shortwave)
Pyrgeometer,* 4-50 pm Pressure sensor
5 min (10 s)
Ri?, Ri$ (long wave)
0.6 m
P
surface
5 min (10 s) 5 min
(10s)
HMP35Cprobe*
T, RH
2 m
5 min
(10 s) Rawinsonde
WS, WD, T, RH, P
0-12 km
(2-6 d-1)
(5-12 m for T, RH; 100-200 m for WS, WD) Anemometer
W S, WD
2-3 m
5 min
(10 s)
Abbreviations areRASS,radioacoustic sounding system; RH, relativehumidity; WS,windspeed; WD, wind direction. Seetext for descriptions of variables.
*Instruments werealsodeployed to leads.
RUFFIEUXET AL.: ICE PACKANDLEADSURFACE ENERGYBUDGETS
4596
2.2. Intercomparisons and Measurement Accuracies
Followingthe deploymentto lead 4, the two lead micrometeorological packagesweresetup andoperatednextto the base camp system for approximately2 days (April 14
of_+3W rn'2. (3) Solarradiationsensors agreedwithin_+2%. No adjustments weremadeto individualsensors.
3. Meteorologyat the Base Camp
and15). All leadradiometer sensors werePointedupward
The instrumentation deployedat thebasecampis listedin for comparisonwith the two normallyupwardpointingbase Table 1. Wolfe et al. [1992b]previouslydiscussed results campradiometers.Becauseof a tape failure, the west side obtainedfrom remotesensingsystems,and Perssonet al. lead sonic anemometer data could not be recovered. A comparisonof basecampandeastsidesonicanemometer fluxes [1992] characterizedthe lower troposphereby analyzing is given in Figuresl a and lb. The small scatterof the rawinsondedata from the base camp, from Deadhorse, individual15-minvaluesandthe absenceof significantmean Alaska, and from Barrow, Alaska. They found a median
depthof 560-660m anda median temperature bias demonstratethe validity of the upwind-downwind inversion turbulent flux comparisonsthat are critical to the lead differenceacrosstheinversionvaryingfrom 6.2øCto 10.5øC. Here we focuson a characterization of the Arctic packice studies. These data indicate that the sonic anemometer surfaceparameters as recordedby the meteorological tower about_+2W m'2, andthe stressvaluesare accurateto about at the basecampduringthe entireexperiment.The first set of daily averagedtime series(Figures2a-2c) illustrates the _+0.003N m'2. derived values for mean sensible heat flux are accurate to
In the interestof savingspacewe will simplysummarize the comparisonresults for the other sensors. (1) The temperature-humidity sensorswere accuratein a differential senseto within about 0.2øC for temperatureand 5% for relative humidity. (2) The long-wave sensorsshowed
temporal evolution of temperature, moisture, andwindat the base camp. These data were recordedon data loggers sampledat 10 s and then averagedto 5-min values. The data from various radiation sensors(Figures3a and 3b)
coupledwith sonicanemometer data(Figure3c) give us a good estimate of the background surfaceenergybudget disagreements aslargeas8 W m'2. Thedifferences between components during LEADEX. The average diurnalcyclesof sensors variedoverthe period,but someaveragebiaseswere resolvable. An averageof all sensorsfor the period was the thesevariablesare alsoexamined(Figures4-6). computed,and a singlemeancorrectionwas appliedto each 3.1. Time Series sensorto make it agree, on average,with this consensus. Threedailyaveragedtemperature time seriesareshownin Thesecorrectionswere appliedto the sensorsover the entire experimentalperiod,resultingin a singlesensoruncertainty Figure2a. The skin temperatureis estimatedfrom the
upwardlong-waveradiationflux (measured by a downward pointingpyrgeometer), usingtherelationship
15 ....
I ....
I ....
I ....
! ....
I ....
(a)
Rlt- (1 -œ)Rl,l,
lo
(1)
œG /
ß
/
,.
.
where T, is skin temperature, G is the Stefan-Boltzmann constant,œis the surfaceemissivityof 0.97 [Hollandet al.,
ß
1993], and Rl is measuredlong-waveirradianceand the arrowspoint in the directionof the flux. This formula impliesthatthemeasured upwardR11'long-waveflux results
_
,'
ß
/
•,e el
-10
-15 • -15
-
..... -10
• .... • .... • .... i .... -5
0
5
10
15
BaseCamp•m '=)
(b)
,'
¾"-0.05 E z
'..•I
v
c: -0.10
-
/
0//; ..
/
o -0.15
ß
i
e.'
ß
'
from the surfaceemissionand the reflecteddownwardlongwave flux and assumeszero surfacetransmissivityof the downwardlong-waveflux. The sky temperatureis calculatedfrom the downward long-waveradiationflux with an emissivityof 1.0. It is
representative of the air temperature "seen"by the sensor. In thepresence of opticallythickcloudsthe skytemperature tendsto represent thecloud-base temperature; in theabsence of clouds the sky temperatureis representativeof the convolvedemissivityand temperatureprofiles of the ice particlesandwatervaporin the atmosphere.The clearsky valuetypicallycoincideswith the temperature recordedby the rawinsondes between 5000- and 7000-m
altitude.
In Figure2a the 2-m temperature time seriesshowsa minimumin the middle of the experiment,with v'aluesof ', , , ,i , , , , I , , , , I , , , , -0.20 5øC below average(-21.9øC) and a strongwarmingat the -0.20 -0.15 -0.10 -0.05 0.00 end of the measurement period,startingat Julianday 106. BaseCamp(Nm'2) The daily averagedskintemperature followsthe sametrend, Figure 1. Intercomparison of basecampanddownwind lead with minor variationsfrom day to day. Becauseof the sonicanemometers for eddycorrelation valuesof (a) sensible effectsof clouds,the sky temperaturevaries much more. The strongtemperature increaseobservedat the end of the heat flux and (b) stress.
RUFFIEUX ET AL.: ICE PACK AND LEAD SURFACE ENERGY BUDGETS
4597
oC
B.•,K -• -•,7- - -•-• ................... -30
c
f
o
o
)- .....
o
o o
-,
--
o
-
ß
-40• '50 -
ß m i
,
, ßß m ,
m I
I
I
i
i
m m m i
i
,
•
,
m •
,
,
,
,
,
000000030303030303030303030000000000.,.-.,--..,--.::3
%
100 r.
L
80'-
o o
o
B.•,=,..L-
o
o
0'-0'
70 )-
0
0
0
0 0 0
0
0
'0-
0
0
0
-
0 0
6O r 50 •* m m t t ,
0
i
i
!
i
i
, , ,
, , m! ! t t ! mmt ,
m, ,
!
i
i
!
i
i
t
i
i
i
I
i
i
I
I
I
i
I
i
-
000000030303030303030303030000000000','-','-.,--.::3
Figure2. Basecamptimeseriesof dailyaverages, fromJulianday87 to Julianday112. Thedashed lines representthe differentmeansfor the period. (a) 2-m temperature (triangles,lineA = mean), skin temperature (opencircles,line B = mean),andskytemperature (solidcircles,line C = mean);(b) relative
humidity withrespect toice(solidcircles, lineA = mean),andrelative humidity withrespect towater(open circles,line B = mean);(c) wind speed.
experiment wasdueto thearrivalof a warmersynoptic air massoverthesite. With a meantemperature of about-22øC the dataset collectedat basecampis representative of an early springmeteorology over the ice [Ebertand Curry,
biasedby the increasedspeedsbetweenJulian days 107 and 111. During this perioda surfacelow from the Gulf of Alaska pushed northward into the interior of Alaska. North of thislow, a strongpressuregradientwasestablished 1993]. andproduceda strongeasterlyjet at the basecamp. Thedailyaveraged timeseriesof RH (Figure2b)doesnot Figures3a-3c describethe temporal evolution of the showthesametrendasthetemperature. The meanvalueof radiativecomponentsmeasuredat the base camp. The RH with respectto waterwas 73%, with no periodsof upwardand downwardlong-waveradiationshow a similar saturation,and only a few cases with RH above 80%. temporalevolution(Figure3a). The averageflux values However,whenwe examinetheRH with respectto ice,the were 220 and 183W m'2, respectively.The difference meanjumpsto 90%,withcases of saturation quitecommon. betweenthe two fluxes increasedduring the experiment, Lightwindsbetween 0 and5 ms-' predominated during exceptfor the last 2 days. This gradualincreasein the net LEADEX (Figure2c). The meanwindspeedof 4 m s-• is long-waveenergylossappearsto be dueto a slowlydecreas-
4598
RUFFIEUX
ET AL.'
ICE PACK AND LEAD SURFACE ENERGY BUDGETS
W m'2
300 • (a) 250
A'--'> fo-'....... ß
200
ß
o
O
o
ß
o
o
o
ß
ß
ß ß
o
ß
ß
ß
ß ß
o
ß
o
o
o
o
•
o
• •
o
150-
'
i
'
i
i
i
i
i
i
i o 1ol o i
]
i
i
lot
t
iø1
•
•
•
•••••••OOOOOOOOOO•
•
W m'2 200,
,
,
I
!
!
i
i
i
!
i
i
i
i
i
i
i
i
i
i
i
0
,
(b)
o
o
ß
ß
i
0
!
0
i
0
o o _(3. 0
B•[.
.......
100 •0o
0
D- 0 - -
ß
ß
'
50
-
•0
,
,
(•0 (•00•
ß
o ß o o ß ß ß
ß
ß
l ß
C•I-
ß
-O_-•
A-A•-•--• 3•AA,A A• AAAAAAA >'
,
c o•
O•
O•
O•
O•
O•
O•
O•
O•
O•
O
O
O
O
O
O
O
O
O
O
w-- w-- w-- •
W m'• (Sensible Flux)
(Stress) Nm
20• '*'*' ' '*'*'' .................. ii0.0 15:i (C) ß
*
10•'"'.... . ....... i© 0 00
*
' ø 0 • 0 c•0
o ß
5 ' t-0.1 0 r. •-.............. ¾............. o""ø' ........ '•......... :........ •.................. ;;................................. 'o'
A-.-•10
o
ø0.2
-15
.
-20 -25r- , ,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
o ,
,
,
,
c
•0 (•0 00 O• O• O• O• O• O• O• O• O• O• O O O O O O O O O O w" w" w",_•
Figure 3. Same as Figure 2, but for (a) upwardlong-waveradiation(solid circles,line A = mean) and downwardlong-waveradiation(opencircles,line B = mean);(b) downwardshort-waveradiation(open circles,line A = mean), upwardshort-waveradiation(solidcircles,line B = mean), and total net radiation (triangles,line C = mean);(c) sensibleheatflux (opencircles,line A = mean),andfrictionalstress(solid
circles,lineB = mean). The dottedlinesin Figures3b and3c show0 W m-2.
ing, downward long-wave flux and an upward flux that is decreasingless, resulting from a smaller decreaseof the surface temperaturethan of the sky temperature. When heavycloudsappearedduringthe last 2 days,the downward long-wave radiation increasedgreatly, resultingin a nearzero differencein upward and downwardlong-waveradiation. For the short-wave radiation flux (Figure 3b) the
portionof the solarflux is reflectedback to the atmosphere, and the energyavailablefor heatingthe surfaceis relatively
small(22 W m-2,on average).Thisfactpartiallyexplains
the counterintuitivesteadydecreaseof net radiation(sum of the short-waveand long-wavenet fluxes) over the observation period (Juliandays 87 to 105). The slightincreasein availableshort-waveradiationis more than compensated for downwardcomponent had a meanof 140W m-2,andthe by the increasein the long-wave loss of energy. The downwardflux graduallyincreasedfrom thebeginningto the presenceof heavycloudsthe last2 days,whichincreasedthe end of the experiment. The heavy cloudsat the very end downwardflux of long-wave radiation, suddenlyreversed produceda large drop in both componentsof short-wave this tendency. Note that the magnitudesof both the mean radiation but only a slight decreasein the net short-wave netsolar(22 W m-2)andnetlong-wave (-37 W m-2)radiation radiation. Becauseof the high surface albedo, a major areabout5 W m-2greater thantheAprilclimatology values
RUFFIEUX ET AL.' ICE PACK AND LEAD SURFACE ENERGY BUDGETS i
i
i
i
i
!
i
i
[
[
i
i
i
i
4599
-
-20
-25
[
-30
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hours (local Time) -15
I
i
I
i
i
i
i
I
i
4
5
6
7
8
i
i
i
i
i
i
i
i
!
i
i
i
i
i
i
(b) -20
-25
-30
0
I
2
3
' 11' 12' 13' 14' 15' 16....... 9' 10 17 18 19 20 21 22 23 24
i i i .r--r--T ....I--q-----T-I i 1----r----r----•
1'01'11'21'314151617181920212 Hours (Local Time)
Figure4. Meandiurnaltimeseries.Theverticalbarsarethehourlystandard deviations, andthedashed
linesrepresent thedifferent means fortheperiod asfollows'(a)skintemperature, (b)2-mtemperature, (c) windspeed, (d)relativehumidity withrespect to ice(thinline)andwithrespect to water(thickline), and (e) mixing ratio.
given by Maykut [1978], but the combinednet radiation periodsof largerpositive(10-12W m-2) and negative
(-15 W m-e)is the same. (-20W m-•)values didoccur, withtheperiod of significant Thesonicanemometers measured thethreecomponents of negativeturbulent heattransport (transport of heatfromthe the wind(U, V, and W) alongwith the virtualtemperature atmosphere to the ice) on Juliandays 106-109associated Tv at a samplingrate of 10 Hz. Gaynor et al. [1992] with the strong,warmeasterly jet mentioned previously. analyzed the frequency spectraof a sonic anemometer This4-dayperiod determined thesignof theaverage sensible operatedthe previousyear under similar conditionsand heat flux for the experiment,which was -3.4 W m-:. As found that the sonic anemometers operatedwell with expectedfor over an ice surface,the downwardstresseswere turbulence intensities that follow unstable and near-neutral usually smallin magnitude, withanaverage of-0.07 N m'•. surfacelayer similaritytheory. Fifteen-minute sensible Largedownward stresses only occurred with the strong
heatflux W'T' (approximately equalto virtualtemperaturewindson Juliandays 106-109. flux) and frictional stress U'W' were calculated at all three
sites. The daily averagesensible heat fluxesgenerally fluctuated between _+5W m'2,witha slighttendency to be
3.2. Diurnal Cycles
By averaging hourlydatafor the studyperiod,we can describe thediurnalpatterns of themeteorological variables. theatmosphere) moreoftenthannegative (Figure3c). Some Thehourlyaverage valuesgivea general viewof thecycle,
weaklypositive(turbulenttransportof heatfrom the ice to
4600
RUFFIEUXET AL.: ICE PACKAND LEAD SURFACEENERGYBUDGETS !
i
!
!
!
!
i
•
i
!
!
i
T 90
.............
•80 70
60
• • • • • • • • • • 1•1'11•1•1•1'51• Hours (Local Time)
1.0
i i i i i i [ ! i i ! i 1' ! ]
0.9 0.8
0.7 0.6
LCnO. o)0.4 0.3 0.2 0.1 0.0
1'01'11'21'31' 1'51'6 '71'81'9 2'0 2'12'2 2'3
Hours (Local Time)
Figure 4. (continued)
angle(between 20ø and30ø at noon),theamount while the standarddeviationdescribes the dispersion of the elevation the surfacevariesconsiderably from dataaroundthe mean. LEADEX occurredin a seasonwhen of energyreaching
tonoonandbacktotheevening. Themajorpartof the amplitude of the solarradiation diurnalcycleis very morning backto theatmosphere (Figure5b) similarto that typicallyobserved in midlatitudes during thisenergyis reflected of the highsnowalbedo(Figure5c). The albedo winter,although thenumber of hoursof insolation is longer because werecalculated onlyattimeswhenRs,> 25W m-2. for LEADEX. Examinationof diurnal cyclesgives some values The downward and upward long-wave fluxes(Figure5d) indications of the response time of atmospheric boundary havea diurnaltrend,with largervaluesduringtheday. The layerandsubsurface icethermalprocesses. The skintemperature diurnalcycle(Figure4a) is well small relative decreaseof flux around noon can be seen on
pronounced witha maximum between 1200and1500local both curves. This middaycoolingremainsunexplained. thedownward long-wave flux is notin phase standard time (LST) and a minimumin the middleof the Surprisingly,
(Figure4b) but actuallypeaks night.Notetheunexplained relative cooling between 1200 with the2-m air temperature
and1400LST. At 2 m aboveground(Figure4b), thecycle earlierthanthe upwardlong-waveflux. This is probably causedby a maximumin low-levelcloudiness in the in agreement withthediurnalcycleof cloudiness smaller,and both the maximumand minimumvaluesare morning, by Perssonet al. [1992, Figure5]. The net shifted2 hourslater in the day. The meandiurnalrange suggested of temperature wasmorethan8øC(skintemperature) and radiationhasa strongdiurnalpattern,with positivevalues 5øC at 2 m. The diurnalcycle of the skin-airtemperature between1000 and 1600LST (Figure5e). The time delay between the minimum of downwardlong-wave radiation differencehasan amplitudeof about_+2.5øC. radiation Wind speeddid not vary muchon a daily basis (2000LST)andtheminimumof upwardlong-wave betweenmorningand (Figure4c),evenwitha largerandom variability (standard(0200LST) explainsthe asymmetry deviations> 2.5 m s-1). It is interestingto note that the afternoon net radiation. slightdiurnalcycleis in phasewith the 2-m temperature The sensibleheatflux diurnalcycle (Figure6a) showsa strongdiurnalvariationwith positivevalues(turbulent cycle. of heatfromthesurface totheatmosphere) between As expected, thediurnalcycleof relativehumidity with transport = 10W m-2)andnegative respect to ice followedthe 2-m temperature cycle,with 1000and 1600LST (maximum = -10 W m-2). minima at the warmesttime of the day (Figure4d). The valuestherestof thetime(minimum is smootherthan at the interface,the standarddeviationis
mixingratiocurve(Figure4e) shows thesametrendasthe
2-mtemperature, withnodailymean values over0.6g kg-1.
The latent heat flux was not measured but was estimated
usingthe equation
Because of the weak surface humidity fluxes, relative
humidityvariations tendto bedominated by airtemperature. Downwardshort-wave radiation(Figure5a) is responsible
for drivingthe diurnalcycles. Even with a low solar
L-r
RUFFIEUX ET AL.' ICE PACK AND LEAD SURFACE ENERGY BUDGETS 5OO
(a) ß
400
3OO
200
ß
lOO
0
I
I
I
I
2
3
I
4
5
6
7
8
I
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hours ('Local Time)
500•: , ,
• (b) 400
300
200
lOO
o • ;• :• 4 5 6 7 8 9 lO11121314151617181920212'22'324 Hours ('Local Time)
1,o
i
i
i
i
i
0.9(c)
i
'_
0.8
....
0.7 0.6 0.5 0.4 0.3 0.2
0.1
0'00 • • :• ,• • • '• I• fi 1'01'11'21'31'4 1'51'61'71'81'92'02'1 2'22'324 Hours ('Local Time)
t
!
i
i
I
240
220
......
200 180
160 i
i
i
!
i
i
!
0 • 2 3 • 5 s 7 8 9 •0 • •2 •'3•'• •'5•'• •'7•'8•'92'o2'•2'22'3 Hours (Local Time)
Figure5. SameasFigure4, butfor (a) downward short-wave radiation, (b) upwardshort-wave radiation, (c) surface short-wave albedo,(d) downward (thinline)andupward(thickline)long-wave radiation, and (e) totalnetradiation.The dottedline in Figure5e shows0 W m'2.
4601
4602
RUFFIEUXET AL.: ICEPACKANDLEADSURFACE ENERGYBUDGETS 5o 40
(e)
3o 2o
lO E
0 -10 -20
-30 -40 •$0
0
1
2
3
4
5
6
7
8
-
9 1011 12 13 14 15 16 17 18 19 20 21 22 23 24 Hours (Local Time)
Figure 5. (continued)
whereH l is the latentheatflux, Hs is the sensibleheatflux, The lasttermof thesurfaceenergybudgetis theground qsat.i• (Ts) is the saturated vapormixingratiofor T•, q is the flux or "conductivity flux" Qg,whichin the absence of mixingratio at 2 m, T• is the skintemperature, and T is the meltingandfreezingat the surface canbe estimated by the 2-m air temperature.The factor2.5 is the ratio of the latent flux balance at the air-snow interface heatof vaporization of waterto theheatcapacityof air when q is expressed in gramsper kilogram. Data were excluded Qg= Q, _Hi- • (3) for IT• - TI < 0.2øC.Because of theverylow vaporpressure of ice at thesetemperatures, the latentheatflux (Figure6b) whereQ* is the net radiation. Here we use the sign is small and fluctuatesaround 0 with an averageof convention thata positive Qgvalueimpliesthatheatistrans0.6W m'2. The two minimain the morning andat the porteddownwardinto the snow/ice. If the popularforceafternoonare artifactscausedby the measurement uncertain- restoremethodfor predictingthe surfacetemperature is used ties for the changein the temperaturegradientbetween [Blackadar,1976], the right side of (3) will also containa surfaceand2 m. The Bowenratioat the afternoon peakis heat storageterm. The use of this storageterm impliesthat about 5.
the calculated groundflux, givenby Q;, represents the
20•- ....................... ß
lO
-10
'200 1 2 3 4 5 6 7 8 9 101112131415161718192021222324 Hours (Local Time)
5I (b) • o :-:•
:--•--•----p--
-5
i
i
i
i
i
i
i
i
i
!
i
i
i
i
i
!
i
i
i
i
i
i
-100 1 2 3 4 5 6 7 8 9 101112131415161718192021222324 Hours (Local Time)
Figure6. SameasFigure 4, butfor (a)sensible heatflux,(b)latentheatflux,(c)ground fluxwithout (solid)andwith(dash-dotted) thestorage term,(d)frictional stress, (e)dragcoefficient, and(f) transfer coefficient.The dottedlinesin Figures6a, 6b, and6c show0 W m-2.
RUFFIEUX ET AL..
50
! , (c)
40
!
ICE PACK AND LEAD SURFACE ENERGY BUDGETS
!
i
!
!
!
!
i
!
!
!
i
!
!
!
!
!
!
!
4603
!
.%
/!
30
'-,
20
10 E
• o -10
: ..................
-20
•;-.........
-30 -40 -50
.....
0 I
'•"--•..•""v';,, ........
, , , ,
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2.122 23 24 Hours (LocalTime)
0.00
................
-0.05
E-0.10
z
-0.15
(d)
-0.200{ •' ;• •, .• • • • • 1'01'11'2. 1'31'41'51'61'71'81'92'02'12'22'324 Hours ('Local Time)
0.003-
,
i
i
i
i
I
--I'
i
i
(e)
0.002 0.001 ß
0.000! i•'••,•I•'•••1'0 1'1 1'2. 1,31•41,51,61,71 Hours ('Local Time)
0.0030 .r'-,
,
,
-,- ,
,
,---•
,- ....
,--•
0.0025 ! (f)
0.0015 IT•J 00010 0.0005•-
I
i
i
i
0.00006 • •' ;• • ,• I• • • • 101'11'21'314151'61'718192021222324 Hours (Local Time)
Figure 6. (continued)
4604
RUFFIEUX
ET AL.:
ICE PACK
AND
LEAD
SURFACE
ENERGY
BUDGETS
ground flux at the base of an assumedisothermalsurface
wind speed. Thus thesedata do not provideconvincing
slab, rather than at the air-snow interface, and that the
evidencethat CH is differentfrom
temporalchangesin Ts used to calculatethe storageterm
3.3.
occurthroughout this slab. Both Qg (withoutthe storage term)andQg(withthestorage termandanassumed surface slab depth given by the diurnal thermal wave [Arya, 1988]) were calculated,and Figure 6c describesthe diurnalcycle of
Discussion
The study of boundary layer meteorology during LEADEX implies the considerationof at least two main timescales, the annual and diurnal timescales. It is not clear
the groundflux. On average,Qgtendsto coolthe surface that we have been able to observethe springtimeportionof
the annual cycle during this 1-month observationperiod. During the period a gradual increase in surface or 2-m temperaturewas not observed, nor were there any Q•,especially in theearlymorning, suggest thattheassump- obviousgradualtrendsin mostotherparameters.The steady tion of an isothermalsurfaceslab is suspect,sinceone would increasein the incomingsolarradiationthroughmostof the expect that the groundfluxes shoulddecreasewith depth, perioddueto the rapidlyincreasingnumberof daylighthours and the increasingsolar elevationangle did not producean during the late afternoonand heat it during the rest of the daily cycle. If the heat storageterm is considered,then the diurnalpatternis similar,but the much largeramplitudesof
rather thanincrease. Theaverage valueof Qgis-12W m'2
increase in the net radiation
because most of the additional
withoutthe storage termand-16 W m'2 with it compared with a meanvaluefor April of-6 W m-2givenby Maykut energywas lost due to the large surfacealbedoand because
the incominglong-waveradiationtendedto decreaseslightly during most of the period. Instead, variationstend to be related to synopticevents,suchas thoseon Juliandays91, The magnitudeof the momentumflux or frictionalstress is largest during the evening and smallest near dawn 97, and 106-109. Ratherthan a gradualseasonaltransition, (Figure 6d), partiallybecauseof slightlylargerdaytimewind a sharp, distinct transition to more summerlike Arctic speeds and hydrostatic stability effects. The 3-m drag conditionsoccurredduringthe last 2 days of the observing period (Juliandays 110 and 111). These days had surface coefficientCt>was calculatedby usingthe relationship and 2-m temperatures much warmerthan the rest, incoming solarradiationmuch smaller,and, most importantly,incoming long-waveradiationmuch larger due to increasedlowS' W • Ca= _ U 2 (4) level cloudiness.The net radiationwasthereforethe largest observedduringthe entireperiod. The strongsynopticevent during Julian days 106-109 undoubtedlyproducedthese whereU is the streamwise fluctuatingcomponent of thewind changes,althoughit is not clear whetherthe observedlarge, speed,and W is the verticalcomponent of the wind speed. downwardturbulentheat fluxes, the presumednear-surface Calculatedvaluesof Ct>excludeperiodsof low wind speeds horizontal warm air advection, the increase in downward
[1978]. Ebert and Curry [1993] give an April valuefor the
net heat flux over the ice of-5 W m-2.
( observedwarm-up. It is quitelikely thatthey all contributed. of 1.54x 10-3(meanuncertainty about15%)at a height of Althoughit is unknownif the springof 1992 was climato3 m givesa 10-mvalueof 1.2x 10-3,whichisequivalent to logicallytypicalin this respect,theseobservations suggest a roughnesslength of 0.1 mm. This is a bit lower than that synopticconditionstend to dominatethe degreeof typical10-mvalues quoted forthepackice(e.g.,1.58x 10-3 warmingand coolingof the Arctic atmosphereand that the from Banke et al. [1980]) but is consistent with the seasonaltransitionto summerconditionsmay occur with rather smooth, featureless snow cover [Leavitt, 1980] that very strongsynopticevents,rather than throughgradual characterizedthe upwind footprint of the base camp heatingprocesses.Hencethe synoptictimescaleappearsto anemometer. be an importanttimescalein the Arctic as well. The sensibleheatturbulenttransfercoefficient(Figure 6f) The apparentparadoxthat the skin temperaturedid not was calculatedusingthe relationship cool during the LEADEX observationperiod despitean
average surface heatlossof 12 W m-2canbe explained by C•-
W'T' U
T)
(5)
recallingthatwe actuallyhavea thin, insulating snowlayer on top of a thick ice slab. Thus 12 W m'2 of heat mustbe suppliedby the ice layerto the bottomof the snowlayer. This impliesthat the snow-iceinterfaceis significantly
Data were excludedfor ITs- TI < 0.5øC,wind speedsU less warmerthanthe meanskin temperature(-22øC). In fact, we
than 1 m s-•, and IH•I < 5 W m-2. Here we plottedan note that on March 27 the net flux was aboutzero, while the averageonly whenthe numberof observations meetingthose skintemperaturewas-19øC;thereforewe deducethat on that
of about-19øC. criteriaexceeded 20. The meanvalueof 1.0 x 10-3(uncer- datethe snow-iceinterfacehada temperature occurred on April 20, suggesting a snowtaintyabout25%)givesa 10-mvalueof 7.5 x 10-4. The A similarsituation uncertaintyin CH is muchgreaterthan that for Ca because ice interface temperatureof -14øC. These results are the 3 W m-2 uncertaintyin the upwardlong-waveflux consistentwith thoseof Maykut and Untersteiner[1971], of about-18.5øC translatesto +_1 øC uncertaintyin T•. This leadsto a signifi- who foundsnow-iceinterfacetemperatures cantlylargererrorin the typicalvaluesfor T• - T thanfor the and - 15øC on these dates.
RUFFIEUX
ET AL.'
ICE PACK AND LEAD SURFACE ENERGY
BUDGETS
4605
Variationson the diurnal timescaleare clearly presentin 12 W m'2duringtheperiodof thedownwind observations at thedata,andsomeof theparameters, suchasthe surfaceand the lead. At the lead the winds were 6.6 m s'• from 330 ø 2-m temperatures, havesurprisingly largediurnalamplitudes. which was abouta 45 ø angleacrossthe lead. Becauseof the Hence the sun drives a significantdiurnal cycle of most of greatwidthof thislead,theleadinternalboundary layerwas the meteorologicalparameters. Even with low sun angles of sufficientdepth that our sonic anemometerprovided and high ice albedo,the temperaturetime seriesshowsa accurate measurement of the actual surface fluxes at the substantialheatingduring the day and cooling during the downwind edge of the lead. The stress was about from 180 to night; even the wind speedshowsa measurablediurnal -0.18 N m-:, andthesensibleheatflux decreased cycle,with higherwindsduringthe "warm"periodof the 130W m': duringthe 6-hourperiod. The factorof 2 day. The diurnalcyclereachesa heightof 100-200m, with increasein stressover the backgroundpack ice value is frequent superadiabaticlayers in the lowest 30-60m puzzlingbecauseit cannotbe explainedby stabilityeffects. [Perssonet al., 1992]. This diurnalsensitivityis primarily This implies that the ice-coveredlead had substantially due to the thermodynamicpropertiesof the snow layer on greaterroughness (2 mm versus0.1 mm), whichimpliesan topof thepackice at thebasecamplocation.The amplitude increase in the 10-mneutraldragcoefficient from 1.2 x 10-3 of the diurnalcycle of sensibleheat flux is abouthalf that to 2.2 x 10-3. observedin April at camp BrassMonkey during AIDJEX The remainder of this section is devoted to a discussion of [Thorpeet al., 1973]. resultsfrom lead4, which openedon April 11 and where When studyingthe leadeffectson the planetaryboundary more completemeasurementswere obtained. layer in the vicinity of a newly formedlead, the very low averagesensibleandlatentheatfluxes(their sumis lessthan -3.0 W m'2) shouldbe considered asreferences.Of course, 4.1. Air Temperature, SensibleFlux, and Frictional to properlyinterpretthe contextof the lead measurements, Stress at Lead 4 we mustalsoconsiderthe typicalpeak diurnalvaluesof the At basecamp,duringApril 11-13, 1992,boththeskinand surface sensibleheat flux of +_10W m'2.
the2-m temperature showa diurnalcyclewith a tendency to havean unstable layercloseto theice nearnoon(Figure7). This superadiabatic temperature profile is in phasewith a 4. Fluxes and Atmospheric Boundary Layer slight increase in wind speed. The sky temperature variaEvolution in the Vicinity of a Lead tionsrecordedduringthese2 daysindicatepartly cloudy On April 7, 8, and 9, the LEADEX group deployedto conditions.Laser ceilometerbackscatter data (not shown) lead 3. Operationallimitationscausedby the extremewidth indicatesomeperiodsof low-levelice cloudsandfog. (1 km) and distancefrom the basecamp of this lead greatly Sonicanemometers andtemperature sensors wereplaced reduced the usable data we obtained.
Flux and sodar data
at 2 m abovegroundon both the west and eastsidesof the lead,thewestsideanemometer beinginstalledat 2100 UTC (1200 LST) on April 12, 18 hours after the east side anemometer. The westsideinstruments wereplacedwestof theportablehuts(Figure8). Because of thechanging wind direction,theeastsidesiterepresented thedownwindsitefor the first 20 hours(0100-2100UTC on April 12), and the
were obtainedon the downwindsideof the lead in the period from 2100 UTC on April 7 to 0200 UTC April 8 when the lead was 2 days old and essentiallyice coveredbut quite smoky. No simultaneousupwind flux or sodardata were obtained (although mean bulk meteorology data were obtained); upwind measurementsmade the day before indicatedthat the upwindice roughness wasnot significantly different from that at the base camp. At the base camp, winds were about6 m s'• from 345ø, the stresswas about -0.09 N m'2, and the sensibleheatflux increasedfrom 5 to
-10•:
west side site was the downwind site for the last 9 hours
(0300-1200UTC on April 13). Datafrom the basecamp sonicanemometer, at 2-m height10 km away,are usedto
characterize thebackground lead-free sensible fluxesduring
,
10
-15i
9
-25 -30 - -
'.
-35
., , ' ....
'"•'• "'
,,-. • ".--:".":' 0.85 aroundthe noon hours), substantialdiurnal mentsat 2-m heightundersampled the true surfacefluxesby cycles were observedin the surfaceenergy budget at base as much as a factor of 2. The increase in surface stress over camp. The diurnal amplitude for sensibleheat flux was both leads is too large to be explainedby conventional +_10W m-2 about a mean of-3 W m-2, net radiationwas hydrostaticstabilityeffects. For lead 3 this suggests an ice+_30W m-2 about a mean of -15 W m-2, and net surface covered lead10-mdragcoefficient of 2.2 x 10-3versus1.2 x For the entire experimentthe mean air temperaturewas
energy fluxwas+_20W m-2abouta meanof- 12W m-2.The 10-3observed atbasecampandvalues of 1.5x 10-3typically mean latent heat flux was +1 W m-2 with a diurnal variation
reportedin the literaturefor pack ice. At the time of these
of about+_ 1.5W m-2.Thedifference in the2-mairtempera- observations,lead 3 was alreadycoveredwith 5-10 cm of ture and the skin temperaturehad a diurnal variation of +_2.5K. We presumethatthe 15-cm snowcoveringthe pack ice was responsiblefor the large diurnalcyclesof sensible heat flux and skin temperature. The base camp sonic anemometerdata were usedto computetransfercoefficients for momentumand sensibleheat. Mean valueswere Co =
(1.20 +_0.20)x 10-3 and CH= (0.75 +_0.25)x 10-3 at a 10-m referenceheight with only modestdiurnal variations. Thesevaluesare somewhatlower than typical averagesfor the Arctic pack ice [Thorpe et al., 1973; Banke et al., 1980] but are consistentwith the valuesfor the relativelyfeaturelesssnownearthe basecamp. As surfaceroughness and/or wind speedincrease,we expect Co to becomegreaterthan CH[Joffre,1982]. However,we suspectthatthe experimental difficulty in measuringCn at thesesmallheatfluxesand
deformedandraftedice, whichwouldcertainlycontributeto increasedroughness. Doppler minisodarsupwind and downwind of the lead indicatedsignificantlead effectson the near-surfaceturbulence. We investigateda limited periodbetween0900 and
1500 UTC on April 12, a nighttimeregimecharacterized by strongsurfacestability. The upwindsodarindicateda sheardriven turbulentsurfacelayer of about 5-m depth with intermittentwave interactionsup to heightsof 40 m. The downwindsodarindicateda surfacelayerof 10-mdepthwith muchstrongerwaveinteractions extendingto heightsgreater than 60 m (sodarrangemaximum). Three prominentwaves with strongdownwardmotion were observedin this period. Significantincreases in the downwindstressmagnitudewere coincidentwith the lowest downward penetrationof these
RUFFIEUX
ET AL.:
ICE PACK
AND
LEAD
SURFACE
ENERGY
turbulentwaves, presumablydue to the import of higher sonic anemometer data.
BUDGETS
We
4611
also thank Tim Stanton,
Because this increased stress
JamieMorison, and Miles McPhee for assistancewith our
was on the downwindside of the lead, it tendsto pull the lead apart.
upwindlead sensors and the entiresupportstafffrom the Universityof Washington PolarScience Centerfor makinga
momentum flow from aloft.
Using(6) anddatafrom the basecampandfrom upwind difficult environmentalmostfun. This work is supportedby the Office of Naval ResearchgrantN00014-91-J-1557.
anddownwindof the lead, we computedestimatesof the net
surface heattransfer Qgoverthelead,theadjacent packice, and any openwater that might have occurredin the lead. References We assumedthatRs, andR l,L were the sameover the lead and the adjacentice. For open water we assumedTs= Albrecht,B., and S.K. Cox, Procedures for improving -1.6øC and •x = 0.06; Hs and H i were computed with pyrgeometerperformance,J. Appl. Meteorol., 16, 188-197, 1977. McBean's[1986] bulk parameterization.Surfacevaluesfor Hs over the ice-coveredlead were deducedfrom the Anderson, R. J., Wind stress measurementsover rough ice during the 1984 Marginal Ice Zone Experiment,J. Geophys. 2-m sonicdata usinga fetch-dependent correctionbasedon Res., 92, 6933-6941, 1987. the numericalcomputations of Lo [1986]. Thesecomputations indicate that the ice on the lead reduced the sensible
heatflux by abouta factorof 5. We thenusedthe bulk flux formula to estimate the effective skin temperatureof the lead T•'. Finally, we estimatedthat the albedoof the lead increasedfrom 0.3 to 0.6 over the periodof interest(0.3-0.7 was also used with little effect on the results).
Andreas, E. L., Estimation of heat and mass fluxes over Arctic leads, Mon. Weather Rev., 108, 2057-2063, 1980.
Andreas,E. L., C. A. Paulson,R. M. Williams, R. W. Lindsay, andJ. A. Businger,The turbulentheatflux from Arctic leads, Boundary-LayerMeteorol., 17, 57-91, 1979. Arya, S.P., Introduction to Micrometeorology, 307 pp., Academic, San Diego, Calif., 1988. Badgley,F. J., Heat budgetat the surfaceof the Arctic Ocean, Proc., Symposium on the Arctic Heat Budget and Atmospheric Circulation, edited by J.O. Fletcher, Memo. RM-5233-NSF, pp. 267-277, Rand Corporation, Santa Monica, Calif., 1966. (Available as PB-182433 from Natl. Tech. Inf. Serv., Springfield, Va.)
The contrastof net flux over pack ice, the lead, and open water (where it existed)is impressive(see Figure 13). The short-livedpositiveportionof the openwater flux at about localnoonis a bit misleadingbecausesomeof the solarflux not reflectedis absorbedbelow the oceanmixed layer, rather than in it. On the other hand, most of the solar energy not reflectedis absorbedby the ice covering the lead. Thus Banke, E.G., S.D. Smith, and R.J. Anderson, Recent measurements of wind stress on Arctic sea ice, J. Fish. Res. whensignificantice forms,the sunbecomesmoreeffective • Board Can., 33, 2307-2317, 1976. in reducingthe surfacefreezing rate and in shuttingoff convectivemixing in the oceanunder the ice in the lead. Banke,E.G., S. D. Smith,andR. J. Anderson,Drag coefficients at AIDJEX from sonic anemometer measurements, in Sea Ice Over the period of observations the averagenet heat flux Processes and Models, edited by R.S. Pritchard, over the pack ice was -75 W m-2, over the lead was pp. 430-442, Universityof WashingtonPress,Seattle,1980. -130W m-2,andovertheopenwaterwas-250W m-2. Baumann, R., An analysis of one year of surface layer
A 1-monthsampleis far too shortto draw any definitive meteorologicaldata from the Arctic pack ice, M.S. thesis, conclusionsaboutthe springclimateof the Arctic ice pack, 52 pp., Schoolof Oceanogr.,OregonStateUniv., Corvallis, 1978. particularlybecausethisis a transitionperiodwith considerable variability. However, our average results are well Beran, D. W., C. G. Little, and B.C. Willmarth, Acoustic Doppler measurements of vertical velocities in the within the envelopeof valuesfrom variousexperimentsand atmosphere,Nature, 230, 160-163, 1971. models. We were surprisedby the magnitudeof the diurnal range in skin temperature,8øC, and in sensibleheat flux, Blackadar,A. K., Modeling the nocturnalboundarylayer, paper presentedat Third Symposiumon AtmosphericTurbulence, 20 W m-2,but we thinkit is fairly typicalfor areaswith
significantsnow cover. A few aspectsof this study are
Diffusion and Air Quality, Am. Meteorol. Soc., Boston, Mass., 1976.
puzzling, (1) a steady increase from0 to 40 W m'2in thenet Carsey, F.D., The boundary layer height in air stress radiativecoolingoverthe first 23 daysof the experimentand (2) the unusuallylarge drag coefficient for lead ice. In contrast to most previous studies of leads, this project emphasizedleads in the secondand third day of their life cycles when ice covering varied from 5 to 30 cm. In one casethe thin lead ice reducedaverageheat lossby more than
100W m-2 compared with openwaterbut allowedmuch greaterdiurnalheatflux variationsthan are presentover the
measurements,in Sea Ice Processesand Models, edited by R. S. Pritchard,pp. 443-451, Universityof WashingtonPress, Seattle, 1980.
Cheung, T.K., Sodar observations of the stable lower atmosphericboundary layer at Barrow, Alaska, BoundaryLayer Meteorol., 57, 251-274, 1991. Ebert, E. E., and J. A. Curry, An intermediateone-dimensional thermodynamic sea ice model for investigating iceatmosphereinteractions,J. Geophys.Res.,98, 10,085-10,109,
packice. The leadalsoenhancedsignificantturbulentand 1993. wave turbulenceactivityon the downwindsideof the lead. Fett, R. W., S. D. Burk, W. T. Thompson,and T. L. Kozo, Thesestudiessuggestthat manymore sourcesof data (e.g., Environmental phenomena of the Beaufort Sea observed multiple-heightflux measurements near the lead and skin during the Leads Experiment,Bull. Am. Meteorol. Soc., 75, temperature andice thickness measurements acrossthelead) 2131-2145, 1994. are necessary to understand the heatflux over real leads. Gaynor, J.E., D.E. Wolfe, and Y. Jing-Ping, Turbulence Acknowledgments. The authors wish to recognize the contributions of Dave Gregg, ScottAbbott, Jay Palmer, JesseLeach,NorbertSzczepczynski, and Karen Martin of ETL. Specialthanksto Ye JingPing,whoprocessed an abundance of
structureof the atmosphericsurfacelayer over Arctic ice near a lead, paper presented at Third Conference on Polar Meteorology and Oceanography, Am. Meteorol. Soc., Portland, Ore., 1992.
Glendening, J. W., andS. D. Burk,Turbulenttransport froman
4612
RUFFIEUX
ET AL.:
ICE PACK
AND
LEAD
SURFACE
ENERGY
BUDGETS
Arctic lead:
A large-eddy simulation, Boundary-Layer Smith,S. D., R. J. Anderson,G. den Hartog,D. R. Topham,and R. G. Perkin,An investigationof a polynyain the Canadian Holland D. M., L. A. Mysak, and D. K. Manak, Sensitivity archipelago,2, Structureof turbulenceand sensibleheatflux, J. Geophys.Res., 88, 2900-2910, 1983. study of the dynamic thermodynamicsea ice model, J. Geophys.Res.,98, 2561-2585, 1993. Smith, S. D., R. D. Muench, and C. H. Pease, Polynyas Joffre S. M., Momentum and heat transfersin the surfacelayer and leads: An overview of physical processesand over a frozen sea, Boundary-LayerMeteorol., 24, 211-229, environment,J. Geophys.Res., 95, 9461-9479, 1990. 1982. Strauch R. G., K. P. Moran, P. T. May, A. J. Bedar, and Kaimal, J.C., and J.E. Gaynor, Another look at sonic W. L. Ecklund, RASS temperaturesounding techniques, thermometry, Boundary-Layer Meteorol.,56, 401-410, 1991. Tech. Memo. ERL WPL-158, 12 pp., Natl. Oceanic and Atmos. Admin., Environ. Res. Labs., Boulder, Colo., 1988. LEADEX Group, The LEADEX experiment,Eos Trans.AGU, 74, 393-397, 1993. Thompson,W. T., and S. D. Burk, An investigationof an Arctic Leavitt, E., Surface-basedair stressmeasurements made during front with a verticallynestedmesoscalemodel,Mon. Weather AIDJEX, in Sea Ice Processesand Models, edited by Rev., 119, 233-261, 1991. R. S. Pritchard,pp. 419-429, Universityof WashingtonPress, Thorpe,M. R., E.G. Banke,and S. D. Smith,Eddy correlation Seattle, 1980. measurementsof evaporationand sensibleheat flux over Lo, A. K.-F., On the boundarylayer flow over a Canadian Arctic sea ice, J. Geophys.Res., 78, 3573-3584, 1973. Archipelagopolynya,Boundary-LayerMeteorol.,35, 53-71, Vowinckel, E., and S. Orvig, Synopticenergybudgetsfrom the Meteorol., 59, 315-319, 1992.
1986.
Mastrantonio G., and G. Fiocco, Accuracy of wind velocity determinations with Doppler sodars,J. Appl. Meteorol.,21, 823-830, 1981.
Maykut, G. A., Energy exchangeover young sea-icein the centralArctic, J. Geophys.Res., 83, 3646-3658, 1978. Maykut, G. A., and N. Untersteiner,Someresultsfrom a timedependent thermodynamic modelof seaice,J. Geophys.Res.,
Beaufort Sea, in Energy Fluxes Over Polar Surfaces,edited
by S. Orvig, 299 pp., World MeteorologicalOrganization, Geneva, 1973.
Walter, B. A., J. E. Overland, and P. Turet, A comparisonof satellite-derived and aircraft-measured regional surface sensibleheat fluxes over the Beaufort Sea, J. Geophys.Res., this issue.
Wolfe,
D.E.,
C.W.
Fairall,
and
D. Ruffieux,
76, 1550-1575, 1971.
Surface
energymeasurements on the Arctic ice pack,paperpresented McBean, G., The atmospheric boundary layer, in The at Third Conference on Polar Meteorology and Geophysics of Sea Ice, edited by N. Untersteiner, Oceanography,Am. Meteorol. Soc., Portland,Ore., 1992a. pp. 283-337, Plenum,New York, 1986. Wolfe, D. E., C. W. Fairall, J. R. Jordan, and D. W. Gregg, Mursch-Radlgruber, E., andD. E. Wolfe, Mobile high-frequency Remotesensingof theArctic boundarylayer, paperpresented mini-SODAR and its potential for boundary-layerstudies, at Third Conference on Polar Meteorology and Appl. Phys., B, 57, 57-63, 1993. Oceanography,Am. Meteorol. Soc., Portland,Ore., 1992b. Mursch-Radlgruber, E., D. E. Wolfe, D. W. Gregg,C. W. King, W. D. Neff, K. A. H. Sharp, and D. Ruffieux, NOAA's portable high-frequencyminisodar-designand first results, C. W. Fairall and D. E. Wolfe, EnvironmentalTechnology lnt. J. Remote Sens., 15, 325-332, 1994. Laboratory,NOAA/ERL, R/E/ET7, 325 Broadway,Boulder,CO Payne, R. E., Albedo of the sea surface,J. Atmos. Sci., 29, 80303-3328. (e-mail: c'fairall @gateway'ømnet'cøm; 959-970, 1972.
[email protected]) Persson,P.O. G., D. Ruffieux, and K. Davidson, Characteristics P.O. G. Perssonand D. Ruffieux, CooperativeInstitutefor of the lower troposphere during LEADEX 92, paper Researchin EnvironmentalSciences,University of Colorado, presentedat Third Conferenceon Polar Meteorology and Campus Box 216, Boulder, CO 80309. (e-mail: Oceanography,Am. Meteorol. Soc., Portland,Ore., 1992.
[email protected];
[email protected]. ch) Schotanus, P., F. T. M. Nieuwstadt, and R. A. R. de Bruin,
Temperaturemeasurementwith a sonic anemometerand its application to heat and moisture fluxes, Boundary-Layer (ReceivedJanuary18, 1994; revisedJuly 27, 1994; Meteorol., 26, 81-93, 1983. acceptedAugust2, 1994.)
