ebook img

The Confluence Zone of the Intense Katabatic Winds at Terra Nova PDF

15 Pages·2007·1.64 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview The Confluence Zone of the Intense Katabatic Winds at Terra Nova

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 95, NO. D5, PAGES 5495-5509, APRIL 20, 1990 The ConfluenceZ one of the IntenseK atabaticW inds at Terra Nova Bay, Antarctica, as Derived From Airborne SastrugiS urveys and MesoscaleN umerical Modeling DAVID H. BROMWICH Byrd Polar Research Center, Ohio State University, Columbus THOMAS R. PARISH Departmento f AtmosphericS cience, Universityo f Wyoming,L aramie CHRISTIAN A. ZORMAN(cid:127)' Byrd Polar Research Center, Ohio State University, Columbus The surfacew ind field inlando f the intensec oastakl ataba!,(cid:127)cw ind regimea t Terra Nova Bay, Antarctica,h as been studied observationallayn d numerically. Airbornes urveyso f wind- inducedf eatureso n the snows urfaceh aveb eenu sedt o constructth e time-averagewd inters urface airflow pattern. The surfacem otionf ield hasa lso beens imulatedb y a mesoscalper imitivee qua- tionm odeul singte rrains lopews itha horizontrael solutioonf 3 2 km. Bothm ethodosf a nalys!(cid:127)s demonstrateth at the intensek atabatica irstreama t Terra Nova Bay is forced by converginga ir currentsi n the continentailn terior. The broadscalec onfluencez one becomeso rganizedin to two regionsw ithin about 180 km of the coast. The primaryr outef or katabaticm asst ransporitn to the Terra Nova Bay area is ReevesG lacier but an importants econdarys ourcei s providedb y airflowd ownD avidG lacier. The former is generatedb y the Coriolis-inducecdo ncentratioonf mass transporto n the left side (lookingd ownwind)o f the broadscalec onfluencez one as well as by the near-coastaml ountaind eflectiono f airflow into the valley head. The confluencez one feeding into DavidG lacierv alleys tretcheosv er 100 km into the interiora nd is forcedb y the broadscale terrain configurationo f the ice sheet. Source areas for the two airstreamsd iffer with Reeves Glacier valley being fed by cold air formedw ell into the interior, whereast he airflow downD avid Glacier is sustainedb y radiativoev ceoro loliwnge r partso f the ice sheetw hich are much closer to the RossS ea. Evaluationo f the observationaaln d modelingr esultsr evealsb otht he strengths and weaknesseosf eacha pproach. Airborne surveyso f sastrugoi rientationsa re a highlys uccess- ful methodf or establishingth e detailedp atterno f surfacea irflow. However, a systematiecx ami- nationo f sastrugid imensionss uggeststh at such work could also be carried out using SPOT-type satellite observations,w hich is a more cost-effectivea pproacht han aircraft surveying. The double-jets tructureo f katabatica irflow into the Terra Nova Bay area is well representedb y the primitivee quationm odeling,b ut much finer grids are neetdoe rde solvet he near-coastadl rainage details. 1. INTRODUCTION toted by these units continuest o expand, with the result that surface climatic processesa re being diagnosedm ore The continental-scale surface wind field over Antarctica completely[ Stearnsa nd Wendlet,1 988]. is a keyc limaticc omponenitn the southernh emispherew hose The secondd evelopmenits that the ice sheett opography effects are manifested well into the middle latitudes has been determinedw ith unprecedentedp recision [Drewry, [James,1 988] and throughouta substantiadl epth of the 1983]; elevationsa re knownw ith an uncertaintyo f about3 0 atmosphereo verlyingt he continent[ Polar ResearchB oard, m for the entire continent. Based upon a realization that 1984, p. 10]. Yet descriptiona nd studyo f this phenomenon the prime forcing of Antarctic drainage winds is gravity have been severely limited by the continent's remote acting on radiatively-cooleda ir over sloping terrain location and harsh climate. Two recent developmentsa re [Parish, 1982], Parisha nd Bromwich[ 1987] have diagnosed drastically altering this situation. Automatic weather the pattern of surface airflow over Antarctica from the stations( AWS) which transmito bservationsto NOAA polar- simple steadys tatem odelo f Ball [1960]. Input data con- orbiting meteorological satellites can be deployed in remote sisted of terrain slopes at 50-kin resolution and estimates areas and have a high probabilityo f operatings uccessfully of the boundaryl ayer stratification[ Phillpot and Zillman, throughoutth e year. The fractiono f ice sheeta reasm oni- 1970]. Surfacea ir does not blow radially and uniformly away from the highestp arts of the ice sheet, but ratihse r highlyi rregularw ith regionso f pronouncecdo nfluencaen d difiuence. Where negativelyb uoyanta ir from a large in- terior section of the ice sheet converges,t he supply of cold air to sustaink atabaticw inds blowing down the mar- Now at Departmenot f Physics,C aseW esternR eserveU niversity, ginal ice slopesi s considerablye nhanced[ Parish, 1984]. Cleveland, Ohio. Such confluencer egions in the interior wind field are believed to be the prime cause of the intense katabatic Copyright1 990 by the AmericanG eophysicaUl nion. winds monitoreda t Cape Denison and Port Martin in Adelie Paper number 89JD03027. Land [Parish, 1981]. The resulting highly irregular trans- 0148-0227/90/89JD-03027505.00 port of katabatica ir across the Antarctic coastlinei s of 5495 5496 BROMWICH ET AL.: TERRA NOVA BAY KATABATICC ONFLUENCEZ ONE supremei mportancein determiningt he effectso f the ice age-hardened(s intered)c an persistf or months[ Watanabe, sheetb oundaryla yerc irculationu pont hea tmospheraen dt he 1978; Okuhiraa nd Narita, 1978, Figure 5]. Becausem ost oceans urroundingth e continent. East Antarctic precipitation( approximatesd eposition)f alls Terra NovaB ayi s locatedd ownwindo f a pronouncecdo n- during winter [Bromwich1. 988] a sastrugsi urveyc onducted fluenee region in the simulated interior wind field of at the end of winter should record sastrugio rientations Parish and Bromwich[1 987]. Consistenwt ith the arguments that formed as a result of many past depositionael vents. outlineda bove,t he coastalk atabaticw indsa re knownf rom As is demonstrateidn section2 .3, analyseso f sucho bserva- the historicalr ecordso f Scott'sN orthernP arty. who were tions allow the time-averagedai rflow patternt o be con- forcedt o winteri n this area during 1912, to rival thoseo f structeda, techniqupe ioneerebdy Mathera ndM iller[ 1967]. CapeD enisona ndP ort Martin in termso f persistencaen d Somev ariabilityo f sastrugdi irectionsw ill inevitablyb e strength[ Bromwicha nd Kurtz, 1982, 1984]. RecentA WS present in such data sets as a result of atypical but observati9nhsa ves hownt hatt hes urfacew inds peeda verages infrequents ynopticf orcing [Mather, 1962]. 17 ms- for thef all monthosf February_Atpp rilw ith To construct the time-averagedw inter surface airflow speedsm ostlyv aryingb etween1 0 and 30 m s [Bromwich, patterno vert he ice slopesw esto f Terra Nova Bay, airborne 1989]. A joint study,d irectedb y the first two co-authors, photographics urveyso f sastrugio rientationsw ere under- is being conductedto describet he kinematicsa nd dynamics takeno n November6 , 1986, and November8 . 1987. Early of this intense katabatica irstream. The investigationi s November was chosen as the desirable time period because directed toward obtaininga n enhancedu nderstandingo f lighting conditionsa re favorable for aerial photography, coastalk atabaticw ind regimesw hich are forcedb y interior but winter wind conditions still prevail. Both missions confluence zones; these features seem to dominate the boun- were planneda nd coordinatedb y Jerry Mullins of the U.S. dary layerc irculationo vert he ice sheetso f Antarctica. GeologicaSl urvey( USGS)a ndf lownb y U.S. Navyp ersonnel This paperc ombineso bservationaaln alysisa nd numerical using an LC-130 transporta ircraft. For both missionsth e modeling to explore the interior confluencez one which LC-130wasequippedwithaWildRC-8camerawitha152.3mm sustainst he intensek atabaticw inds at Terra Nova Bay; a focal length. The camera was set at a shutter speedo f preliminaryv ersiono f this work was presentedb y Bromwich 1/300 sec., an ilstop of t:8 and was loadedw ith Kodak et al. [1988]. The snow surface over the plateaut o the Panatomic-Xa erographicfi lm 2412. The film wasd eveloped west of Terra Nova Bay has been extensivelyp hotographed and processedi nto a series of 9 x 9 in. black and white from the air. Section2 describest hese surveyso f wind- contactp rints which formed the data base for the sastrugi forced features at the snow-air interface which allow the analyses.P hotographsw ere takenw ith 60% along-tracko ver- time-averageds treamlineso f surfacea ir motiont o be estab- lap betweens uccessivep hotographs. Thus inspectiono f lished with unprecedenteds patial resolution. The interior every other print provided complete examination of the wind field has also been simulatedb y the three-dimensionalp hotographedre gion. version of the mesoscale numerical model of Parish and The objectiveo f the missionsw asn ot to providec omplete Waight[ 1987]. and the resultsa re presentedi n section3 . spatialc overageo f the Terra Nova Bay confluencez one, but The last section compares and contrastst he observational rather to sample the region in order to capturet he overall and modelings tudieso f the confluencez one, and concludes patterno f the surfacea irflow. Hence the missionsw ere that the discrepanciesa re primarily due to the grid spacing flown in discrete,w e!l-separatefdli ght legs. The sampling used in the model calculation. area differed between 1986 and 1987 but there was sufficient overlap to evaluate the interannual variability of snow 2. SASTRUGI-DERIVED STREAMLINE FIELDS surface features. The 1986 missionw asd esignedto capturet hed rainageo f 2.1. Flight Surveys David. Reevesa nd Priestleyg laciers. Figure ! showst he Sastrugair e wind-formemd inorr elieff eattireast the 1986f lightd omain.I t consisteodf six north-souftlhig ht snow surface whoseo rientationsm irror the directiono f the lines approximately 18 km apart and three east-westt ie wind that generatedth em. The literaturec ontainstw o lines. The tie linesw eret o aid in the positioninogf the definitionso f what entitiesc onstitutes astrugi. Roscoen. orth-southfl ight lines and to fill in detailsl ying be- [1953] and Alley [1989] define all aeolian-forcefde atures tweent he north-soutlhin es. The missionw asf lownb etween to be sastrugib, ut Mather[1962],D oumani[1967]F, ujiwara 1858 UTConNovemberSand0314U TC onN ovembe6r in andE ndo[ 1 97]1, andW atana[b1e9 78]d iscriminabteet ween1 9116T. hea ircrafgt rounsdp eerd.a 4agferodm 4 35t o5 14k m alepositio(nsnalo wdun(=e sw halbea cks)andbarchanoidhs)-a nwdit ha na veragoef 480k mh - '. Thea ircrafftl ewa t a erosional( calleds astrugi)f orms. The varyingd efinition constanht eighto f 5390 m aboves ea level. Becauseth e appearst o arise in part becauseo f areal variationsi n terrain elevationr angedf rom 200 to 2200 m. the aircraft typical alepositioncaol nditions. Around Siple Coast,f or altitudev ariedf rom 3190 to 5190 m. Sincet he scaleo f examplem, osts nows eemsto bed epositeidn conjunctiowni th eachp hotograpdhe pendsu pont he aircrafth eighta bovet he light windso r downslopset ormw inds( M. Strobel,p ersonal !?hotographreedg ion thes calec hangedth roughoutht er egion communic1a9ti8o9n)T,. hescea nth ebne e rodbeydd ownf-r om1 :20,95a0t3 '!90-amlt itudtoe 1 :34,0a8t05 190-m slope drainagew inds so that alepositionaaln d erosional altitude. Consequentlya,t 3190 m ab9vet he terraint he forms are parallelt o the prevailingw ind direction. By sampledarea:ppehro tograpwha s22.9k in-, whilea t5190mit contrastin lando f SyowaS tation,p recipitatioins usually was6 0.7 kin'. The smallesrte solutionfo. undb yc alculating accompaniebdy strongc ross-slopcey clonicw indsa nds now the resolutiono f the photograpwh ith the smallesst cale, dunesf orm by forcedd eposition[W atanabe1, 978]: subse- was 5 m. Positionalu ncertaintyw as calculatedto be a quentlyt hesea re carvedb y drainagew inds. The typical maximumo f q-2 kin, andw asg reatesint then orthwesste ction angulard eviationb etweend epositionaaln d erosionaflo rms of the flight domaind ue to the lack of locallyp rominent on MizuhoP lateauis 50ø [FujiwaraandEndo1. 971]. Auto- terrain featuresw ith which to tie the flightlines. This mariew eathers tationo bservationfsr om inlando f Terra Nova positional uncertainty is no problem, however, because Bay (discussedin section2 .3)show that almosta ll winds changesi n sastrugi directionsi n this region are large- comef roma narrowa ngularr ange. Depositionaclo nditionss cale,o ccurringo vert enso f kilometers.T he totaln umber inlando f Terra Nova Bay thusa ppeart o be similart o those of photographisn the 1986 missionw as9 45. at Siple Coast. Becauset he predominanlti near depositional The 1987 missionw asd esignedto resolvet he broadscale and erosionaflo rmsm ay not be readilyd istinguishabflreo m structureo f the Terra NovaB ayc onfluenczeo ne,a nd is also the air, the sastrugdi efinitiono f Roscoe[1 953] and Alley showni n Figure 1. Becauseth e 1987 and 1986 flight legs [1989] is adoptedh ere. overlappeda round David and Reevesg laciers,i nterannual Sastrugif orm rapidly in hours to days [Mather, 1962; variability can be investigated. The 1987 flight domain I.M. Whi!!ans.p ersonacl ommunication1. 989]. and once consistedo f five flight legs, four comprisingth e sideso f BROMWICH ET AL.' TERRA NOVA BAY KATABATIC CONFLUENCE ZONE 5497 / 1900 CapeA dare J J J J .---.% (cid:127).i',.'., ..-. ½.C ;aYHp.a.e l lelt /...-- 2300 TalDoosm e4 I . . :. / I TuckerG I. ! I ,;.':- .(cid:127)?'.;. -' ) (cid:127), 23OO .. -/oIsu.l man J Gl i/ rGI. \ ß ' '''(cid:127):'.i..C ampbGelIl. TERRA NOVA BAY 75'os - / Ice Ton ;(cid:127)........(cid:127)'.(cid:127) ........... ROSS SEA \ \ ------- 1986 , 1987 /2500' (cid:127)jx (cid:127)McMurdo $ta. [..(cid:127) .;: I .-,1i 1(cid:127) ..;..?...:. .:. ., (cid:127) ?8ø5 i (cid:127)%_ (cid:127)'(cid:127)' '(cid:127):(cid:127) (cid:127):(cid:127),.. ROSS SkerG(cid:127) lto r-'--] Faiscte ''//(cid:127)/(cid:127)(cid:127) Routegrhra in 145øE 150øE 155øE 160øE 165øE 170øE Fig. I. Regionall ocationm ap. Ice sheete levationc ontoursi n meters( solid and dashed)w ere obtainedb y merging Drewry's[ 1983] interiora nalysisw ith USGS 1:250,000 reconnaissancmea psf or the coastarl egions.T he 1986 and 1987 sastrugis urveyf light legs are shownb y the heavyd asheda nd heavys olid lines, respectively. a box, the fifth being a tie line connectingt he northwest 2.2. Data Collectiona ndA nalysis and southeast corners. The mission was flown between 1600 an2d2 0U0T Con N ovem8,b1 e9rrh T8h.7ae i rcrfalefawt at n 2.2i.. Lavootftf tli ghlitn esU. porne ceoipft th e averagger ounsdp eeodf 432k m at a heighot f 5390m photograpfrhdsm th eU SGSt,h ep hotograpfohrbs othth e1 986 aboves eal evel. The terraine levationra ngedf rom 400 to and 1987 missionws ere laid out in their flightl eg configu- 2250 m; thereforet, he aircrafta ltituder angedf rom 3140 to rationsa nd checkedf or prominentt errain featuress uch as 4990m , thep hotographsicca lefr om! :20,620t o 1:32,76(cid:127)), mountaings.l acierv alleysa ,d nunataksa,s showno n the andt he sampledar eap er photograpfhro m2 2.2 to 56. I km . USGS !-250,000 reconnaissancmea ps eries. Using these As in the 1986 mission,t he resolutionp er photographw as 5 terrain featuresa s referencep ointsa nd the missionp lans m andl arger. Themaximumpositionaluncertaintywas+4km,a s guides, the flight and tie lines were plotted on the and the numbero f photographwsa s5 75. detailedU SGS maps. The flightd omainf or the 1986m ission 5498 BROMWICH ET AL.' TERRA NOVA BAY KATABATIC CONFLUENCE ZONE was characterizedb y prominentt errain featureso nly in the (i) Linear Sastrugi (1, 2, 3 or 4) southern and eastern portions. Thus the confidencei n the position of the north-southf light lines decreasedi n the Wind direction a northerna nd westernp ortionso f the domain. By beginning with the easternmosftl ight line and workingw est, the tie Type AverageL ength Rangeo f Lengths lines not only positionedt he north-southf light lines in relation to the terrain, but also in relation to the other 1 13m 10m to 16m north-south lines. This method worked well in all but the 2 43 rn 21 rn to 73 rn northern section of the westernmosftl ightline where the 3 51 rn 21 rn to 73m exact intersection of the tie line with the flight line 4 114m 80mto 181m could not be found. However, it wasp ossibleto determinea 4 km square containing the intersectionb y looking for NoteT: ype2 lineasr astrugisi a well-definesdin, gulaurn itw hile slight changesi n sastrugis ize and orientation. type3 occursin d ensely-packebdu nches. Unlike the 1986 mission,t he flightd onlainf or the 1987 mission was characterizedb y identifiablet errain features (ii) Barchanoid Sastrugi (B) only in the northeasts ection. Thus only two flight lines could be accuratelyp ositionedu singt errain features. The positiono f the other flight lines and the diagonalt ie line were fixed by searchingf or changesi n sastrugis ize and Wind direction orientation. While the exact intersectionp oints could not be found, 8 km squaresc ontainingt he intersectionsw ere located. .,---- 25 m ---,- 2.2.2. Extractiono f directionali nformation. Sastrugi were readily identifiable throughoutb oth flight domains. (iii) Ripples (R) Sample air photographsa re not presentedh ere becauset he prints do not contains ufficientc ontrastf or sastrugit o be clearly resolvedi n gray-shader eproductions.R oscoe[1 953, pp. 76-77] and Wendler[1 988] givee xampleso f photographs of sastrugtia kenf rom muchl owera ltitudes.M ost sastrugi Wind direction (Figure 2) were of the linear .type[ Roscoe,1 953, p. 76], but barchanoids[D oumani,1 967]p redominatetdo the northo f the ReevesG lacier entrance. Resolvableli near sastrugi generallyr angedi n size from 10 to 181 m. In partso f the 1986 flight domain, some very large linear featuresw ere presentw hich hada n averagele ngtho f 600 m and a maximum (iv) Drift Plumes (DP) lengtho f about ! km. Thesef eaturess eemt o disappeawr hen the photographasr e examinedw ith a magnifyingg lass,a nd -,- 113 m---- f Collapsedb ridge often have regular-sizeds astrugoi n top of them. Previous (cid:127)' (cid:127)_Wdirinedct ion aerial surveysh aver eportedm uch longers astrugti han seen from theg round: up to 240 m [Roscoe1, 953,p . 78] compared with typical surface estimateso f a few meters [e.g., (cid:127)- Bridcgreedv asse Roscoe,1 953; Doumani, 1967]. It is likely that the spatial organizationo f these relatively low-amplitudef eatures is obscuredb y oblique observationfr om the surface[ compare Roscoe,1 953, p. 78]. It seemsf rom the commentso f Mather [1962, p. 167] that many of the sastrugis een on the air Fig. 2. Primaryw ind-generatesdn ows urfacef eaturesw ithint he con- photographsa re alepositionailn origin. Airborne photo- fluence zone. (a) Classifications cheme.( b) Spatialo rganization; graphic surveys offer the distinct advantageo f spatial data for every tenthp hotographa re plottedf or both 1986 and 1987 flight legs. integration,a nd provided ataw hichc an be reanalyzed.T he primary spatialo rganizationi n Figure 2 is the increasei n length of linear sastrugiw ith elevation. This must result from the interactionb etweent he atmospheriacn d ice surface characteristics that govern sastrugi evolution [e.g., In many photographs,l ess prominent sastrugi with Roscoe, 1953, p. 76], but is not a simple function of the markedlyd ifferent orientationsto those of the primary surfacew ind field dynamics. sastrugi were also present. In no instancew as the dis- Sastrugi orientationsw ere easily discernablei n most tinction betweent heset wo groupsi n doubt:t his is another areas, and the following strategyw as employedf or both consequencoef the spatiala veragingp rovidedb y airborne missionst o extractt hesed ata. Each independenpth otograph photographics urveys. It is probable that a surface-based was dividedi nto quadrantsa nd the directiono f the primary observerw ith his limited field of view is more likely to sastrugi( called the primary direction)w as found for each confusep rimary and secondarys astrugi. The latter are quadrant. Primary sastrugi were thosew hose spatial fre- probablyg eneratedb y short-term,s ynopticallfyo rcedm odi- quency( numberp er unit area) wasg enerallytw icea s high as ficationso f the time-averageadi rflow. Primarys astrugi those of sastrugi with different orientations. If the directionsa re presumedt o follow the gravity-drivenw inter angularr angeo f quadrandt irectionwsa sl esst han2 0ø , then drainageo f surfacea ir, and the generallyd isorganized all quadrandt irectionsw ere averagedy,i eldinga single secondardyi rectionws ereo mittedfr omt hep resenatn alysis. primaryd irectionfo r the photographI.f the angularr ange In ordert o successfulrlye presentth e surfacew indf ield exceeded2 0 ø, then the quadrantd irectionsw ere organized within the confluencez one, the primary sastrugdi irections into groupsw heret he angularr angew asl esst han2 0ø . Then from the 1986 and 1987 missionsh adt o be merged. But in the memberso f eachg roupw erea veragedy, ieldingm ultiple order to justify such a combinationi,t was necessaryth at primary directionsf or the photograph. Primary sastrugi the interannualv ariability of primary sastrugid irections directions were assigned to the centerpoint of the for regionsc ommont o bothm issionsb e insignificantB. e- respectivep hotographs. cause the overlap of flight domainsp rovidedm any such BROMWlCHE T AL.' TERRAN OVAB AYK ATABATICC ONFLUENCZEO NE 5499 b NO',: .....:. os ßS AY / / DPe \ {(cid:127),,, e3 %.,% 3. %!3 .3 ß (cid:127), e3 '' 1(cid:127)3 e .......-..(cid:127)½...;.'.3 (cid:127)} ,, - (cid:127).(cid:127) ..................... / / / !.50:E -,5; 5 ø: (cid:127). 60 ø E Fig. 2. (continued) regions,t he variabilityo f primarys astrugdi irectionws as dary. The resultingp atternis well establishebde, ingo nly examineda nd the resultst abulatedin Table 1. The compari- slightly influencedb y the multiple sastrugid irectionse n- son indicateda meana bsoluted ifferenceo f 6.2 ø with a max- counteredi n the transitiona reas betweend rainager egions imum differenceo f + 12ø Becauseth e meana bsoluted iffer- for particularg laciers. cncies s macllo mparve(cid:127)itdh t hec riteriouns etdo d etermineF igur3e showths er esultintigm e-averaagirefldo wpa t- a primarys astrugdi irection.d ifferencebse tweenth e two tern. The domaini s basicallyd ividedin to two partsw ith yearsw erei nsignificantH. encet he methoduss edi n oh- windsto t hen ortho f 74ø40'Sc onverginingt oR eeveGs lacier tainingt he streamlinep atterna re justified,a nd it will be valleya ndt hoset o the southd rainingd ownD avidG lacier. assumedth at the streamlinep atternr epresenttsh e surface The steepm ountainsto the southwesotf PriestleyG lacier wind field conditionsf or late wintertimer egardlesso f the (with peakst o 2800 m) block the approachings tablea ir- year. streama ndd eflecitt to the rightd ownt het roughle ading to ReevesG lacier. This barrier effect for drainagew inds, 2.3. StreamlinAen alysis whichw as inferredb y Brormvicahn d Kurtz[ 1984] from the Manuasl treamlinaen alysiws asp erformeodn t hef ieldo f theoreticaslt udyo f Ball [1956], playsa key role in set- primarys astrugoir ientationsS. treamlinewse res tartedat tingu pt hea irflowd ownR eeveGsl acier.F igure4 a provides equidistainntt ervalasr oundth ew esterann ds outherend ges a more detailedlo ok at this barrierj et. If the flight of the photographaerde aa ndf ollowedto thee asternb oun- lineo rientedin thed irectioonf WNW fromS kinneRr idgeis 5500 BROMWICH ET AL.: TERRA NOVA BAY KATABATIC CONFLUENCE ZONE TABLE 1. Variabilityo f PrimaryS astrugDi irectionsIn lando f Terra NovaB ay Between1 986a nd 1987 Iax:ationo f Centerpoint of Photol(cid:127)raph AveragSe astrugDi irectiond, eg Differenced, eg Latitude.d el(cid:127) S Longituded. egE 1986 1987 (1986- 1987) Remarks 74ø30.77 f 161 ø37.94f 337 339 -2 74ø29.81 (cid:127) 161 o 10.81 341 341 0 74ø22.50 (cid:127) 160ø23.71 276 283 -7 74ø18.02 f 159ø48.00 302 290 + 12 296 292 +4 74 ø 13.29 (cid:127) 158ø23.57 320 } 325 } -5 } transitionz one 74 ø 6.59' 159ø 12.58 304 314 -10 75ø49.98 ' 159ø48.00 183 195 -12 complext opography 75ø35.69 f 160ø23.71 224 213 + 11 complext opography 75ø 14.80 ' 161 ø'9.82' 283 282 +1 complext opography 75 2.83 (cid:127) 161ø37.94' 293 297 -4 complext opography Mean difference -1.1 Standard deviation 7.8 Mean absolute difference 6.2 taken to be representativeth, e jet is about3 0 km wide, and 3. MODELING OF THE TERRA NOVA BAY the shift from airflow toward the mountains to mountain-de- KATABATIC WIND REGIME flecteda irflowt akesp lacea bruptlyin about5 km. Surface wind observations from four AWS between Feb- In addition to the observational evidence for the conflu- ruary and June in 1988 are available to check the winter encez onep aradigmo fferedb y thes astrugai nalysisn, umeri- streamline analysis (Figure 4a). The directionso f the cal simulationws erec onducteidn ordert o providep hysical vector-averagew inds for all observationsa nd for those insighti nto the mesoscalceo nfluencefe aturea nd attendant speeds(cid:127)h ich exceedt he lowestt hresholdfo r driftings now katabaticw indsn earT erra NovaB ay. The numericaml odel (8 ms-' [B.ryazgi1n9, 86.F igureI ]) agreet o within2 ø . used is the three-dimensionvael rsiono f the primitive This arises because,w ith rare exceptionst.h e winds blow equatiomno dedl iscussined th ek atabawticin ds tudoy f from one direction. Resultantd irectionsa t stations0 9, Parish and Waight[ 1987] and is a modified version or the 23, and 27 lie along the streamlines.T he apparendt is- modeld escribedb y Anthesa nd Warner[1 978]. The modeli s crepancy between the direction at station 21 and the writteni n terrain-followinsgi gmac oordinateasn d includes streamlinesp robablya risesb ecause2 1 is locatedv ery close 10 verticall evels. The first sigmal evelc orrespondtso a to the edge of the barrier jet where there is a sharpd i- heighto f approximatel2y0 m aboveg roundl evel:t he re- rectionalg radient. It can be concludedth at the sastrugi maining vertical levels are distributed such that the analysis accuratelyr eproducest he winter surface-airflow highestr esolutioni s found in the lower portiono f the pattern. atmosphereT.h ise nabletsh eb oundarlya yert o bee xplicigt The broadscalea irflow into David Glacier valley is well resolvedF. luxeso f momentuman dh eatw ithint heb oundary definedw ith mostc olda ir comingf rom plateaua reasa dja- layer are modeledu singa first-orderc losures chemea s cent to the coast( Figure 3). By contrastt. he cold air describebdy Parish[ 1984]a ndt 'arisha ndW aigh[t1 987]. A whichb lowsd ownR eevesG lacierc omesfr om muchf arther horizontagl rid consistingo f 25 grid pointsi n the x direc- inland. The prominenst treamlinec onfluenczeo nei nlando f tion and 21 points in the y directionw ere usedf or the DavidG lacieri s associatewdi tha troughs hownb y the 1800- simulationO. penb oundarcyo nditionass d escribebdy A nthes and2 000-me levatiocno ntoursA. morea ccuratteo pographica ndW arner[1978a]r ee mployeidn whicha ll prognostvica ri- surveym ayr evealt hatt he valleyi s evenm orew ell-defineda blesa re extrapolateodu twardto the boundariesP. rognos- than appearsa t present. The shift from southerlya irflow tic equationsin the modeli ncludet he two horizontaml otion toward, to westerlya irflow within, the confluencez one componentst,e mperaturet,h e surfacee nergy budgeto f (Figure 4b) takesp lace abruptlya long the 1987 tie line Blackada[r1 978] and presstire. Explicitr epresentatioonf within about4 km. Figure 4b suggesttsh at the southern the longwaver adiativep rocessesis incorporatedin the edgeo f thec onfluenczeo nec anm igratea bout2 5 km to the modefl ollowingC ernia ndP arish[1 984]t o providea realis- northa nds outho f its time-averagelodc ation.T he areat o tic numericatl reatmenot f the coolingp rocesseisn the the southa nd southwesot f David Glacier (Figure 3) has lowera tmosphere.T his allowst he katabaticw ind to evolve numerouns unataksT. hisc omplicatedtopographydisruptsinth ae physicallyc onsistenmt annerw ithoutt he needf or ad surfaced rainagein to the southerhna lf of DavidG lacier hoca ssumptiornesg ardintgh ec oolingra te. valley. The ice topographfoy r the Terra NovaB ay regionw as The 1986 flight linesc apturedth e airflowi nto Priestley digitizedt o a 32-kin grid scalef rom the contourm ap of Glacierv alleya ndt he southernfr ingeso f the drainagere - Drewry[ 1983]. This grid representatiorens olvetsh e large gion for RennickG lacier( Figure3 ). The col betweenth ese scale confluencez one structure in the interior of the two drainagea reasi s locatedb y the sastrugai nalysisju st continenta nd is sufficientt o depictt he broadscalset ruc- to the northo f the heado f PriestleyG lacier. The source ture of the katabaticfl ow. but individuagl laciero utlets regionf or air flowingd ownP riestleyG lacierc oversth e and irregularm ountainoutosp ographayr e not explicitly southerpna rto f TalosD omea nds omep lateauar easfa rther representedT.h e terraini n thev icinigot f TerraN ovaB ay west. Airflowi ntoR ennickG lacierv alleyi s discusseidn is extremelcyo mplebx ecausoef thep roximigotf theT rans- the nexts ection. antarctiMc ountainasn dn umeroumso untaipne akasd jacentot BROMWICEHT AL.:T ERRAN OVAB AYK ATABATCICO NFLUENCZEO NE 5501 '55øE ß / / / 1.67E Fig. 3. Time-averagweidn tesr treamlinoefs s urfacaei rflowo vert hep lateatuo t hew esot f TerraN ovaB ay( heavcyo n- tinuousli nes)a s analyzefdr omt hea irbornes astrugsiu rveys.D irectionosf the primarys astrug(ai rrowsa) re plotted at the centerpoinot f everyt enthp hotograpfhr om 1986 and 1987. TerraN ovaB ay. Ther esulttsh ereforew,'h iler epresentative The strongr adiativefl ux divergencrea pidlyc oolst he of the mesoscalkea tabaticw ind structureu pslopeo f indi- ice surfacea ndd evelopmeonft the katabatiwc indi s abrupt. viduagl lacierssu cha sR eeveGs lacierc. annobt e expectedW ithinf ourh oursfr omt hes tarto f the modeiln tegration, to providefi ne-scalein formatioonv er the very complexs ignificandtr ainagfelo wsa ppeaarl ongR eeveGs lacier.T he terrain to the north and southwesat s well as within narrow result has been observedi n other simulations[ Parish, 1984; glacierss uch as PriestleyG lacier. The modelh as been Parisha ndW aight1, 987]a ndi s supportebdy d ataf romA WS; initalized about a state of rest to isolate the terrain- the onseto f strongk atabatiwc indsa t InexpressibIlsel and inducedd rainagef lowsw ithoutc omplicationisn troducedb y adjacentto Terra NovaB ay ofteno ccursw ith suddenness. the large scale synopticf orcing. The initial temperature Resultsf rom the numericasl imulationssu ggestth atm osto f profile usedi n the initializationp rocessw asa daptedfr om the adjustmenotf the katabatiwc indsi s completeb y eight the soundingp resentebdy Schwerdtfeg[e1r9 84, Figure6 .9] hours and relativelym inor changesta ke placed uring the and extrapolatedto the sea level and is consistenwt ith the final few hourso f model integration. Figure 5 illustrates initial state of rest in the wind field. Model equations the streamlineos f colda ir drainagea fter the 12-hours imu- have beeni ntegratedtb r a 12-hourp eriodb y whicht ime the lation for the first and seconds igma levelsc orresponding katabaticw inds are well developeda nd have reacheda near to heightso f 20 and 60 m abovet he ice surface. The steadys tate. terrain-inducedfo rcingo f the airflow is pronouncedt;h e 5502 BROMWICH ET AL.' TERRA NOVA BAY KATABATIC CONFLUENCE ZONE a 160 ø E 74 ø 30'S (cid:127)1800 I I 0 10 Km 600 I 160 ø E 157 ø E I b 75øS ! 75øS (cid:127)) l(cid:127)Okm I---- I 157øE Fig.4 . Detailedde scriptioonf sso mep rominefneta tureosf thei nferrewd intesr treamlinePsr.i marsya strudgiir ec- tionsa re plotteda t the centerpoinotf everys econdp hotographC. omplexte rraini s stippled.( a) Barrierj et feeding into the northernh alf of ReevesG lacierv alley. (b) Confluenczeo neu pwindo f DavidG lacier. In Fig. 4a, vector- averages urface-winddi rectionfsro m AWSs ites0 9, 21, 23, and2 7 for February-Jun1e9 88a rep lottedas h eavyd ashed lines. Dots mark the AWS locations. katabaticw ind patterna t both levels is stronglyi nfluenced the gently sloping interior to the steeper ice slopes near by the orientationo f the topographyin the continentali n- the coast. Streamlinesa re directeda t only slight angles terior. Owing to the larger friction term in the lower part to the ice contoursi n the continentahl interland,b ecoming of the atmospheret,h e streamlinesfo r the first sigmal evel directed in a more downslope direction near the steep ß O are directeds ome2 0 more downslopeth an correspondingc oastal ice slopesi n responset o the suddene nhancementi n streamlinesfo r the seconds igmal evel. A clear tendencyin the katabatic acceleration. Two distinct confluence zones the downslopeo rientationo f streamlinesi s also seen from are delineatedi n each streamline depiction upslopef rom BROMWICH ET AL.: TERRA NOVA BAY KATABATIC CONFLUENCE ZONE 5503 a) STREAMLINES cr LEVEL - I 715 I I 596 477 358 238 119 o o III 223 334 446 557 669 78O x DISTANCE (km) b) STREAMLINES LEVEL : 2 715 I 596 477 358 238 119 0 o III 223 334 446 557 669 780 X DISTANCE (km) Fig. 5. Models treamlineasf ier 12-hourti mei ntegratiofno r (a) firsts igmale vel( 20 m aboveth e ices urfacea) nd (b) seconds igma level (60 m abovet he ice surface). Reevesa nd David glaciers. The analysess uggestt hat cold Two additionalc onfluencef eaturesc an be identified to air producedo ver a broad area in the Antarctic interior the northwesot f Terra Nova Bay on the seconds igmal eveli n drainst hrougho nly two glacierv alleys. The larges ource Figure 5b. The confluenczeo nea t the top left cornerr ep- area and relativen arrownesosf the outflowr egions uggestsr esentst he flow convergencweh ich feedst he Adelie Land that a nearlyu nlimitedc old air supplyi s presentu pslopec oast. This zone is clearly identifiedi n the large scale of the glaciersa nd katabaticw ind enhancemenist to be streamlines imulationo f ParisIta ndB romwic[h1 987]. The expected. seconda dditionaclo nfluencfee atureis situatedso me3 00 km 5504 BROMWlCH ET AL.: TERRA NOVA BAY KATABATIC CONFLUENCE ZONE a) VELOCITY FIELD (m/s) o- LEVEL = I 715 i 2400 596 477 358 238 119 ZOO0 1600 1200 o o iii 22:5 ;534 446 557 669 780 x DISTANCE (km) b) VELOCITY FIELD (m/s) o- LEVEL = 2 715 I i I 2400 596 477 358 238 119 2400 2000 1600 1200 0 (cid:127) 0 III 223 334 446 557 669 780 X DISTANCE (km) Fig. 6. Model katabatiwc inds peed(sm /s) after 12-hourt imei ntegratiofno r (a) firsta nd( b) seconds igmale vels. northwesotf TerraN ovaB aya ndp robablyre presentcso lda ir in Figure 6. In general,t he strongeskt atabaticw indsa t drainagein to RennickG lacierv alley. A polynyai s fre- bothl evelsa re foundn eart hec oasat nda re associatewdi th quentlyp resenjut st offshorfer omt he footo f Rennickt hes _t(cid:127)epeiscte s lopesN. ear-surfakcaet abatwicin dos f 12 Glacier[ comparKe urtzcazdBromwi1c9h8, 3,F igure2c;K (cid:127)r/z m s are founda long nearly the entire coastals tretch. andBrornwich19, 85,F igure4 dJ,a ndm ayb eg eneratebdy o ff- The influenceo f the confluencez onesi s seeni n the local shore katabaticw indsw hen the katabatica irstreamf ed by katabaticw ind enhancemennte ar the Reevesa nd David gla- thee levateindt eriocro nfluenzcoen eis forceddo wnto sea ci_(cid:127)rs.W indm aximaat thel oweslet veiln excesosf 16 m level. s are situatzca _cdr ostsh el owerp ortiono f bothg laciers' The correspondwinign ds peedast thef irsta nds econda smal2l 0 m s katabatwici ndm aximumca n(cid:127) seenn ear levels in the model after the 12-hour simulation are shown Re:evesG lacier. At the secondl evel, the 16 m s and 20 m

Description:
demonstrate that the intense katabatic airstream at Terra Nova Bay is forced by . to be sastrugi, but Mather[1962], Doumani[1967], Fujiwara 1858
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.