Jovian Temperature and Cloud Variability during the 2009-2010 Fade of the South Equatorial Belt LeighN.Fletchera,b,G.S.Ortonc,J.H.Rogersd,A.A.Simon-Millere,I.dePaterf,M.H.Wongf,O.Mousisg,P.G.J.Irwina,M. Jacquessonh,P.A.Yanamandra-Fisherc aAtmospheric,Oceanic&PlanetaryPhysics,DepartmentofPhysics,UniversityofOxford,ClarendonLaboratory,ParksRoad,Oxford,OX13PU,UK bDepartmentofPhysics&Astronomy,UniversityofLeicester,UniversityRoad,Leicester,LE17RH,UK cJetPropulsionLaboratory,CaliforniaInstituteofTechnology,4800OakGroveDrive,Pasadena,CA,91109,USA dBritishAstronomicalAssociation,BurlingtonHouse,Piccadilly,LondonW1JODU,UK eNASA/GoddardSpaceflightCenter,Greenbelt,Maryland,20771,USA fUniversityofCalifornia,Berkeley,AstronomyDept.,601CampbellHall,Berkeley,CA94720-3411,USA gInstitutUTINAM,CNRS-UMR6213,ObservatoiredeBesanc¸on,Universite´deFranche-Comte´,Besanc¸on,France hJUPOSTeam,C/OBritishAstronomicalAssociation,BurlingtonHouse,Piccadilly,LondonW1JODU,UK. 7 1 0 2 Abstract n a Mid-infrared7-20µmimagingofJupiterfromESO’sVeryLargeTelescope(VLT/VISIR)demonstratethattheincreasedalbedo J of Jupiter’s South Equatorial Belt (SEB) during the ‘fade’ (whitening) event of 2009-2010 was correlated with changes to atmo- 4 spherictemperatureandaerosolopacity. Theopacityofthetroposphericcondensationclouddeckatpressureslessthan800mbar ] increasedby80%betweenMay2008andJuly2010, makingtheSEB(7-17◦S)asopaqueinthethermalinfraredastheadjacent Pequatorialzone. AfterthecessationofdiscreteconvectiveactivitywithintheSEBinMay2009,acoolbandofhighaerosolopacity E(the SEB zone at 11-15◦S) was observed separating the cloud-free northern and southern SEB components. The cooling of the h.SEBZ(withpeak-to-peakcontrastsof1.0±0.5K),aswellastheincreasedaerosolopacityat4.8and8.6µm,precededthevisible pwhitening of the belt by several months. A chain of five warm, cloud-free ‘brown barges’ (subsiding airmasses) were observed -regularlyintheSEBbetweenJune2009andJune2010,bywhichtimetheytoohadbeenobscuredbytheenhancedaerosolopacity o rof the SEB, although the underlying warm circulation was still present in July 2010. Upper tropospheric temperatures (150-300 stmbar) remained largely unchanged during the fade, but the cool SEBZ formation was detected at deeper levels (p > 300 mbar) awithintheconvectivelyunstableregionofthetroposphere. TheSEBZformationcausedthemeridionaltemperaturegradientofthe [ SEBtodecreasebetween2008and2010,reducingtheverticalthermalwindshearonthezonaljetsboundingtheSEB.Thesouthern 1 SEBhadfullyfadedbyJuly2010andwascharacterisedbyshort-waveundulationsat19-20◦S.ThenorthernSEBpersistedasa vnarrow grey lane of cloud-free conditions throughout the fade process. The cool temperatures and enhanced aerosol opacity of 7 the SEBZ after July 2009 are consistent with an upward flux of volatiles (e.g., ammonia-laden air) and enhanced condensation, 5 obscuring the blue-absorbing chromophore and whitening the SEB by April 2010. These changes occurred within cloud decks 9 0in the convective troposphere, and not in the radiatively-controlled upper troposphere. NH3 ice coatings on aerosols at p < 800 0mbarareplausiblesourcesofthesuppressed4.8and8.6-µmemission,althoughdifferencesinthespatialdistributionofopacityat .thesetwowavelengthssuggestthatenhancedattenuationbyadeepercloud(p>800mbar)alsooccurredduringthefade. Revival 1 0ofthedarkSEBcolorationinthecomingmonthswillultimatelyrequiresublimationoftheseicesbysubsidenceandwarmingof 7volatile-depletedair. 1 :Keywords: Jupiter,Atmospheres,composition,Atmospheres,structure v i X r1. Introduction GreatRedSpot(GRS)atitssouthernedge, istypicallyadark a brownstripeencirclingtheglobebetweentwoopposingzonal Jupiter’s axisymmetric structure, consisting of bright zones flows (Fig. 1a): a prograde (eastward) jet at 7◦S (SEBn) and and dark brown belts, can undergo dramatic visible changes theplanet’sfastestretrograde(westward)jetat17◦S(SEBs,all over short time scales. The most impressive of these is the latitudes are given as planetocentric). The SEB is a site of in- variability of the South Equatorial Belt (SEB, 7-17◦S), which tenseconvectiveactivityandlightningstorms(Ingersolletal., can change from the broadest and darkest belt on the planet 2004),andisoneofthefewlocationswherespectroscopically to a white zone-like appearance over a matter of months. The identifiableammoniaclouds(SIACs, Bainesetal.,2002)have SEB, which lies in Jupiter’s southern tropics and contains the been observed. However, this activity and the dark coloura- tionoftheSEBwerecompletelyabsentwhenJupiteremerged from behind the Sun during the 2010 apparition, replaced by Emailaddress:[email protected](LeighN. the pale ‘faded’ state (Fig. 2c). This missing jovian belt cap- Fletcher) PreprintsubmittedtoIcarus January5,2017 turedtheimaginationofamateurandprofessionalastronomers niques used to determine temperatures and aerosol opacities. alike, and it prompted a program of thermal infrared imaging Section3presentsatimelineforthe2009-2010fade,whichis of Jupiter’s faded SEB from the ESO Very Large Telescope usedtoprovideinsightsintotheunderlyingmechanismsforthe (VLT) at Cerro Paranal in Chile. These data, along with sup- SEBfadeinSection4. portingobservationsfromtheNASAInfraredTelescopeFacil- ity(IRTF)andamateurobservers,willbeusedtodeterminethe 2. ObservationsandAnalysis variations in temperature and aerosol opacity within the SEB between 2008 and 2010 and provide insights into the under- 2.1. VLTImaging lying physicochemical mechanisms responsible for these dra- High spatial resolution imaging of Jupiter in the Q (17-20 maticmodificationstoJupiter’sappearance. µm) and N (8-14 µm) bands from the ESO Very Large Tele- The SEB fade and revival cycles appear to follow a repeat- scope (VLT) mid-infrared camera/spectrograph (VISIR, La- able pattern, albeit at irregular and unpredictable intervals be- gageetal.,2004)waspreviouslyusedtoprobetheatmospheric cause the underlying physical causes are unknown. Excellent structure of Jupiter’s Great Red Spot (Fletcher et al., 2010b) historicalaccountsoftheSEBlifecycleatvisiblewavelengths andtostudytheaftermathofthe2009asteroidalcollision(Or- can be found in Peek (1958), Rogers (1995) and Sanchez- tonetal.,2011;Fletcheretal.,2010a). Wewereawardedfour Lavega and Gomez (1996). The 2009-2010 fade is the start hoursofDirectorsDiscretionaryTime(program285.C-5024A) of the fifth SEB life cycle since the first spacecraft encounter tocharacterisethestructureandcompositionofJupiter’sfaded withJupiter(Pioneer10inDecember1973),andthefirsttobe SEBinJuly2010,whichwecomparetopreviousVISIRobser- investigatedindetailinthethermalinfraredusingthehighspa- vationsacquiredbetweenMay2008andNovember2009(Table tialresolutionsandbroadwavelengthcoverageofmoderntele- 1)usingidenticalimagingtechniques. scopes and instrumentation. Pioneer 10 and 11 visited Jupiter Jupiter is too large to fit entirely within the VISIR field of duringthe1972-1975fadedstatepriortotheJuly1975revival view (32”×32”), so northern and southern hemispheres were (Rogers,1995;Ortonetal.,1981). Aftera14-yearhiatus, the imagedseparately,oftenondifferentdates. Furthermore,ther- SEBfadedandrevivedin1989-1990(Yanamandra-Fisheretal., mal imaging requires chopping between the target and an off- 1992;KuehnandBeebe,1993;SatohandKawabata,1994)and source position to detect the jovian flux on top of the back- 1992-1993(Sanchez-Lavegaetal.,1996;Morenoetal.,1997). ground telluric emission. However, the maximum chopping The next ‘partial’ fade began in 2007: New Horizons obser- amplitudeofVLT/VISIRislimitedto25”,meaningthatsome vationsinJanuary-February2007revealedtheabsenceofboth ofthe‘sky’imageisobscuredbytheplanet,preventingtheuse the chaotic turbulence and fresh NH3 ice clouds northwest of ofsomeregionsofthechopped-differencedimage(e.g., white theGRS(Reuteretal.,2007;Bainesetal.,2007),butretrievals regionsremovedfrom8.6-µmimagesinFig. 1). Furtherback- of cloud opacity from VLT/VISIR still demonstrated a cloud- ground stability was achieved by offsetting the telescope to a freeSEB(Fletcheretal.,2010b). Afadehadstarted,butavi- skyposition60”fromJupiter(nodding). olent revival began much earlier than expected, restoring dark An imaging sequence typically featured eight filters (Table coloration and turbulent activity by the end of 2007 (Rogers, 2) sensitive to (a) atmospheric temperatures via the collision- 2007a,b). This study concerns the fading event between May induced hydrogen continuum and stratospheric hydrocarbon 2009andJuly2010. emission;and(b)troposphericammoniaandaerosols(8.59and ThegeneralpatternoftheSEBlifecyclecanbesummarised 10.77 µm). Each filter was observed twice, with a small 1-2” as follows. After the cessation of turbulent rifting and con- dither to fill in bad pixels on the detector for all eight filters. vective events to the northwest of the GRS, the SEB (7-17◦S A full imaging sequence required approximately 40 minutes. planetocentric latitude) fades to a pale colour over a mat- TheESOdatapipelinewasusedforinitialreductionandbad- ter of months, obscuring the southern component of the SEB pixel removal via its front-end interface, GASGANO (version (SEB(S), 15-17◦S) and leaving a narrow northern component 2.3.01). Imagesweregeometricallyregistered,cylindricallyre- (SEB(N),7-10◦S)whichhasalsobeenobservedtofadeinsome projectedandabsolutelycalibratedusingthetechniquesdevel- years.DuringthisunusualphasetheGRSappearsasaconspic- oped in Fletcher et al. (2009b). Radiometric calibration was uous red oval surrounded by white aerosols. The faded state achieved by scaling the observations to match Cassini/CIRS can persist for 1-3 years before a spectacular revival begins measurements of Jupiter’s radiance acquired during the 2000 withasingle,localiseddisturbance(theSEBD).Vigorouserup- flyby (Flasar et al., 2004). Consequently we cannot investi- tionsgeneratecomplexpatternswithbrightanddarkcoloration gate absolute changes in Jupiter’s infrared emission, but we throughouttheSEB,encirclingtheplanetandultimatelyrestor- canstudyrelativevariabilitybetweendifferentlatitudinalbands ingthetypicalbrowncolour. (see Fletcher et al., 2009b, for a complete discussion). Mea- Theaimofthisresearchistoreconcilephysicochemicalvari- surementerrorsofapproximately4-6%foreachfilterwereesti- ability (temperatures, composition, clouds) derived from 7-20 matedfromthevariabilityoftheskybackgroundineachimage. µmVLTimagingwithvisiblechangesinthealbedooftheSEB. Alllatitudesinthisstudyareplanetocentric,alllongitudesare The vertical temperature structure and spatial distribution of quoted for Jupiter’s System III West (using the standard IAU aerosolsduringthefadeareusedtodifferentiatebetweendiffer- entmechanismsforthe‘disappearance’oftheSEB.Section2 introducesthesourcesofinfraredandvisibledataandthetech- 1http://www.eso.org/sci/data-processing/software/gasgano/ 2 SEB Fade Sequence I: May 2008 - July 2009 VLT/VISIR 8.6 μm Visible Light IRTF/NSFCAM2 4.8 μm 1100 e (a) 2008-05-17 atitud 00 2008-07-10 L c --1100 ntri ce--2200 o et an--3300 Pl --4400 Oval BA --5500 1 80 1 70 1 60 1 50 1 40 1 30 1 20 1 10 1 00 9 0 8 0 2008-05-13 M. Salway 1 40 1 30 1 20 1 10 1 00 9 0 8 0 7 0 6 0 5 0 System III West Longitude System III West L ongitu de 1100 e (b) 2008-09-03 SEBn EZ Latitud 00 2008-09-24 SEBs SEB centric ----21210000 STrZ o et an--3300 Pl --4400 Oval BA T (K) --55001 90 1 80 1 70 1 60 1 50 1 40 1 30 1 20 1 10 1 00 9 0 2008-09-01/02 A. Wesley 1 8 0 1 70 1 60 1 50 1 40 1 30 1 20 1 10 1 00 9 0 8 0 150 System III West Longitude Sy stem III W est Lo ngitu de 1100 148 e (c) 2009-07-24 d 146 atitu 00 2009-07-20 L 144 c --1100 142 centri--2200 SEBZ B1 140 eto 138 Plan--3300 136 --4400 --5500 31 0 3 00 2 90 2 80 2 70 2 60 2 50 2 40 2 30 2 20 2 10 2009-07-24 M. Salway/A. Wesley 8 0 7 0 6 0 5 0 4 0 3 0 2 0 1 0 0 -1 0 -2 0 System III West Longitude System III West Longitude Figure1:TheSEBfadesequenceMay2008-July2009,observedbyVLT/VISIRat8.6µm(left,sensitivetoaerosolopacityatp<800mbar);visiblelightfrom amateurobservers(centre);andIRTF/NSFCAM2observationsat4.8µm(right,sensitivetoaerosolsabovethe2-3barlevel). Visibleimagestakenascloseas possibletotheVISIRobservationshavebeenprovidedbyM.SalwayandA.Wesley. Suppressedemissionatboth4.8and8.6µmiscausedbyexcessaerosol opacity.NSFCAM2imagesweretakenondifferentnightstotheVISIRimages,anddonotshowthesamelongituderange.Turbulentactivityisevidentinrows(a) and(b)(the‘wake’oftheGRS),andthechainofthreebarges(B1-B3)arelabelledintheSEB(S)inrow(c).The2008conjunctionofOvalBAandtheGRScanbe seeninrows(a)and(b).Largewhitearcsseenequatorwardof5◦Sinthe8.6-µmimagesareduetotheremovalofnegative-beamartefactscausedbythesmall20” choppingamplitude. 3 SEB Fade Sequence II: August 2009-July 2010 VLT/VISIR 8.6 μm Visible Light IRTF/NSFCAM2 4.8 μm 1100 e ud 0 (a) 2009-08-12 2009-08-06 atit c L--1100 entri--2200 B1 B1 c o net--3300 a Pl --4400 --5500 2 90 2 80 2 70 2 60 2 50 2 40 2 30 2 20 2 10 2 00 1 90 2009-08-12 T. Barry 3 30 3 20 3 10 3 00 2 90 2 80 2 70 2 60 2 50 2 40 2 30 System III West Longitude S ystem III W est L ongit ude 1100 EZ ude 00 (b) 2009-11-12 2009-08-27 SEBn atit SEBs SEB entric L----21210000 B3 B2 B1 STrZ oc net--3300 a Pl --4400 --5500 T (K) 3 50 3 40 3 30 3 20 3 10 3 00 2 90 2 80 2 70 2 60 2009-11-14 C. Go 3 80 3 70 3 60 3 50 3 40 3 30 3 20 3 10 3 00 2 90 System III West Longitude Sy stem III West Lo ngitu de 150 1100 148 ude 00 (c) 2010-07-13 GRS White Spot 2010-07-01 146 atit L--1100 144 c 142 centri--2200 SEBZ o 140 net--3300 a 138 Pl --4400 Oval BA 136 --5500 6 0 5 0 4 0 3 0 2 0 1 0 3 60 3 50 3 40 3 30 3 20 2010-07-14 J.P. Prost 3 50 3 40 3 30 3 20 3 10 3 00 2 90 2 80 2 70 2 60 2 50 System III West Longitude System III West Longitude Figure2: TheSEBfadesequenceAugust2009-July2010,observedbyVLT/VISIRat8.6µm(left,sensitivetoaerosolopacityat p < 800mbar);visiblelight fromamateurobservers(centre);andIRTF/NSFCAM2observationsat4.8µm(right,sensitivetoaerosolsabovethe2-3barlevel).Visibleimagestakenascloseas possibletotheVISIRobservationshavebeenprovidedbyT.Barry,C.GoandJ.P.Prost. Suppressedemissionatboth4.8and8.6µmiscausedbyexcessaerosol opacity.NSFCAM2imagesweretakenondifferentnightstotheVISIRimages,anddonotshowthesamelongituderange.Thechainofthreebarges(B1-B3)are labelledintheSEB(S)inrow(b).The2010conjunctionofOvalBAandtheGRScanbeseeninrow(c).Largewhitearcsseenequatorwardof5◦Sinthe8.6-µm imagesofrowaareduetotheremovalofnegative-beamartefactscausedbythesmall20”choppingamplitude,whichwasincreasedto25”inNovember2009. 4 definition of System III rotation (1965), recently reviewed by maps in eight filters (7.9, 8.6, 10.8, 12.3, 13.0, 17.7, 18.7 and RussellandDougherty,2010). 19.5µm)werestackedtoformrudimentaryimagecubesusing Examples of the VISIR imaging at 8.6 µm between May the procedure described by Fletcher et al. (2009b). The tem- 2008 and July 2010 (where low fluxes indicate either elevated perature structure, composition and aerosol distribution were aerosolopacitiesorcoolatmospherictemperatures)areshown derived from the 8-point spectra using an optimal estimation in Figs. 1 and 2. These are compared to examples of ama- retrieval algorithm, NEMESIS (Irwin et al., 2008). Sources teur images at visible wavelengths and Jupiter’s 4.8-µm emis- of spectral line data; conversion of line data to k-distributions sionondatesclosetothoseoftheVISIRobservationsobtained and Jupiter’s a priori atmospheric structure were described in by NASA’s Infrared Telescope Facility NSFCAM2 instrument detail by Fletcher et al. (2011). Two different retrieval ap- (Table 3). Jupiter’s 4.8-µm emission is attenuated by aerosols proachesweretested:(i)asingle-stageprocessretrievingT(p), above the 2-3 bar level, so dark regions indicate locations of NH ,C H andaerosolopacitysimultaneously,and(ii)atwo- 3 2 6 higheropacity. Althoughthedatesandspatialcoverageofthe stageprocessderivingtemperaturesfirst(fromthe7.9µmand mid-IRand4.8-µmimagesdiffer,theirzonalpropertiescanbe 13.0-19.5 µm filters) before the retrieval of composition. At- compared to provide insights into the vertical distribution of mospheric temperatures were derived as a vertical profile de- cloudopacityduringthefadesequence(Section3). fined on a grid of 80 layers between 10 bar and 0.1 mbar, whereas Cassini-derived profiles of aerosol opacity, NH and 3 C H (Fletcher et al., 2009a; Nixon et al., 2007) were simply Table1: VLT/VISIRObservationsofJupiter’sBelts. UTtimesareapproxi- 2 6 scaled to reproducethe VISIR spectra. The two-stage process mate,aseachfiltersequencerequiredapproximately45minutestoexecute. was found to produce unreliable results because the compet- Date Hemisphere Time(UT) ProgrammeID ingeffectsoftemperatureandaerosolopacitywereinseparable 2010-07-13 S 08:10 285.C-5024(A) withoutusingall8filterssimultaneously. Thusthesingle-stage 2010-06-23 N 08:50 285.C-5024(A) retrievalwasusedforeachoftheobservationsinTable1. 2009-11-12 S 01:30 084.C-0206(B) Radiances from each observation in Table 1 were averaged 2009-08-10 N 03:00 383.C-0161(A) over a 10◦-longitude range surrounding the central meridian, 2009-08-11 S 02:30 383.C-0161(A) andinterpolatedontoa1◦latitudegrid. Carewastakentoavoid 2009-08-12 S 03:00 383.C-0161(A) largeovals(theGRS,OvalBA)topreventspuriousresults,such 2009-07-24 S 03:30 283.C-5043(A) thatcentralmeridianscansarerepresentativeofthezonalmean 2008-09-03 S 04:00 081.C-0137(C) radiances on each date. Atmospheric temperatures, aerosol 2008-09-01 N 02:00 081.C-0137(B) opacity and the distributions of tropospheric NH and strato- 3 2008-05-17 S 10:00 381.C-0134(A) sphericC H wereretrievedsimultaneouslyforeachlatitudeto 2 6 determinemeridionalvariationsineachquantity. Examplesof thequalityofthe8-pointspectralfitsateachlatitudeareshown in Fig. 3, which compares synthetic radiances with measure- mentsatthecentreoftheSEB(12.5◦S)onthreedifferentdates. Table3:NSFCAM24.8-µmdatausedinthisstudy Calibratedradiancesweresimilarforallthreeepochs,withev- Date Time(UT) CentralMeridian idence of a decrease in emission at 13.0 and 8.6 µm between 2008-05-27 13:53:39 334.9 2008and2010. Conversely,variationsofzonalmeanradiances 2008-06-16 13:45:26 103.3 at 10.8 and 12.3 µm (sensitive to NH3 and C2H6 respectively) 2008-07-10 13:06:44 96.1 werenegligible,suggestingthattheVISIRphotometrydoesnot 2008-07-27 12:52:05 224.6 showevidenceforNH3 orC2H6 variabilityduringthefade. It 2008-08-07 07:22:50 145.9 islikelythatdetectionofgaseousvariabilityduringanSEBlife 2008-09-24 06:22:15 133.2 cyclewouldrequirespectroscopyratherthanphotometry. Nev- 2009-07-20 14:09:23 27.6 ertheless,atmospherictemperaturesandaerosolscouldbereli- 2009-08-06 10:30:41 296.9 ably separated using the 8-point spectra and will be discussed 2009-08-18 13:19:15 46.9 inSection3. 2009-08-27 13:45:42 338.8 The retrieved haze opacity is driven by the 8.6-µm bright- 2010-06-24 15:50:41 319.9 nesstemperaturesplottedontheleftofFigs. 1-2,wherelower 2010-06-30 14:15:45 85.9 brightness temperatures are the result of higher aerosol opaci- 2010-07-01 15:47:43 292.1 ties. However,the8-pointspectralacktheinformationcontent 2010-09-03 08:26:22 305.4 required to determine the vertical distribution of this aerosol 2010-09-05 10:04:33 306.1 opacity. Cassini/CIRS spectral analyses (Wong et al., 2004; Matchevaetal.,2005;Fletcheretal.,2009a)demonstratedthat the8-11µmwavelengthrangewasbestreproducedbyacom- pact cloud layer of 10-µm radius NH ice particles near 800 3 2.2. Mid-IRRetrievals mbar,althoughneitherthecompositionnorthealtitudeofthese To separate the effects of temperature, aerosol opacity and aerosols were particularly well constrained. Using the CIRS- compositionontheVISIRfilteredimaging,cylindricalradiance derived aerosol distribution, we scale the cumulative optical 5 Table2:VLT/VISIRFiltersusedinthisstudy.ApproximatepeaksofthefiltercontributionfunctionsarebasedonFletcheretal.(2009b). Name Wavelength(µm) Sensitivity Approx. Pressure(mbar) J7.9 7.90 StratosphericT(p) 5 PAH1 8.59 AerosolsandT(p) 650 SIV2 10.77 NH ,aerosols,T(p) 400 3 NeII 1 12.27 StratosphericT(p)andC H 6 2 6 NeII 2 13.04 TroposphericT(p) 460 Q1 17.65 TroposphericT(p) 200 Q2 18.72 TroposphericT(p) 270 Q3 19.50 TroposphericT(p) 400 depth of the aerosols above the 800-mbar level to reproduce thisinhibitionofconvectiveactivitycannotbeidentifiedinthe the8-pointspectraateachlatitude. timesequencepresentedhere,itsignalledthestartofaremark- abletransformationintheSEBoverthefollowing12months. 3. Results: TheSEBFadeTimeline 3.2. FormationoftheSEBZ(July-August2009): Acomparisonofthevisible, 8.6and4.8-µmimagesinJuly This following sections compare temperature and aerosol 2009(Fig. 1c)showsasubstantialalterationtothedistribution distributions derived from VLT 7-20 µm filtered imaging to of aerosol opacity within the SEB at a time when the visible both IRTF 4.8-µm images of deep cloud opacity and ama- colours of the SEB were largely unaltered. At this time, the teur imaging of the visible coloration to reveal the sequence SEB had a pale interior separating a narrow brown northern of changes occurring during the SEB fade. The timeline of component (the SEB(N), 7-10◦S) and a broad brown southern SEBchangesbetween2008and2010issummarisedinTable4, component (the SEB(S), 14-18◦S). Because there is minimal alongwithaselectionofkeycharacteristicsofthefadeinTable gas opacity at 4.8 µm, strong thermal emission implies a rela- 5. tive dearth of aerosol opacity above the 2-3 bar pressure level (e.g., Roos-Serote et al., 1998); whereas 8.6-µm is sensitive 3.1. InitialStateoftheSEB(Pre-May2009) only to clouds and hazes above the 800-mbar level. An en- Filamentary turbulent convective activity usually dominates hancementincloudopacityabovethe2-3barlevelinJune-July theSEBtothenorthwestoftheGreatRedSpot(GRS)inare- 2009 caused a substantial reduction in 4.8-µm emission from gion known as the ‘GRS wake.’ The chaotic activity is driven the SEB, restricting cloud-free conditions to the SEB(N) and bytheconvergenceofasystemofcomplexatmosphericflows, SEB(S). The transformation at 8.6 µm was equally dramatic - whereretrogradeflow(fromtheeast)at17◦Sisdeflectednorth- thefinestructuresthatcharacterisedtheVISIR8.6-µmimaging wardaroundtheperipheryoftheGRStomeettheprogradeflow in2008hadbeenreplacedbyamorediffuseappearance,witha (fromthewest)at7◦S.Thistypicalstatewasobservedthrough- narrowdarklane(8-13◦S)ofelevatedopacityatthe800-mbar out 2008 (Fig. 1a-b), where the SEB between 7-17◦S has a level.ThislaneofelevatedaerosolopacityisknownastheSEB higher temperature than the equatorial zone (EZ) to the north zone(SEBZ). andtheSouthTropicalZone(STrZ)tothesouth. Cloudopac- Fig. 4 shows how the establishment of the SEBZ modified itymeasuredatboth4.8and8.6µmistypicallylow,sothatthe central-meridian brightness temperatures across the SEB be- SEBappearsrelativelybrightatbothofthesewavelengths.This tween 2008 and 2010. These brightness temperatures are av- is consistent with the typical view of upper tropospheric belt- eraged within 10◦ longitude of the central meridian for each zone circulation (e.g., Ingersoll et al., 2004) whereby air rises filter, so are not true zonal averages. Hence, some of the ap- attheequatorandsubsidesovertheneighbouringbelts, creat- parentvariabilityresultsfromthepresenceofdiscretefeatures ingwarm,cloud-freeandvolatiledepletedconditionswithinthe (waves,vortices)closetothecentralmeridian. Changesassoci- SEBandNEB(NorthEquatorialBelt). Smallwhitespotsand atedwiththeSEBfadebetween7and20◦Saremostdramaticat riftingobservedinthevisibleinFig. 1a-bcoincidewithhigher 8.6 µm, with the whole region becoming darker in 2009-2010 aerosol opacities at 8.6 µm and cooler tropospheric tempera- than the south temperate region poleward of 25◦S. Variations tures, indicating localised upwelling within the generally sub- at other wavelengths are more subtle: filters with contribution siding belt. Such upwelling transports spectroscopically iden- functions sensitive to upper tropospheric temperatures (150- tifiable NH ice (SIACs, Baines et al., 2002) upwards into the 300 mbar, approximately, for 17.6 and 18.7 µm, Table 2) re- 3 GRSwakeregion. mainedlargelyunchangedthroughoutthefadesequencewhen Table 4 shows that this SEB activity continued until May- compared to mid-IR variability at other latitudes. Conversely, June2009,whenamateurimagingrevealedtheabsenceofany close inspection of images probing higher pressures (p > 300 convective white spots (Rogers, 2010a), just prior to the first mbar: 8.6, 10.8, 13.0 and 19.5 µm) demonstrate the develop- mid-IR images in July 2009 (Fig. 1c). Although the cause of mentofamulti-peakedstructure,withradiancemaximaat9◦S 6 Table4:Timelineofeventsinthe2009-2010Fade Date Event PossibleImplication 2008-Sep Typical state of SEB (Fig. 1a): visibly brown, warm troposphere and Subsidence over much of the SEB, lo- cloud-free;turbulentactivityNWoftheGRS calisedconvectiveupwelling. 2009-May-26 TurbulentconvectionNWoftheGRShadlargelyceased(Fig. 1c) Inhibitionofdiscreteconvectiveupwelling bysomemechanism. 2009-Jun-05 Final bright convective spots observed NW of the GRS, SEB still ap- CessationofSEBturbulencecomplete. pearsvisiblybrown 2009-Jun-15 First dark brown cyclonic barge (B1) appeared west of the GRS BargesareanewfeatureofthefadedSEB; (Rogers,2010a)inpalebrownSEB(S)(e.g.,Fig. 1c) four more formed at progressively more westerlylongitudes. 2009-Jul-20 High-opacitySEBZobservedat4.8µm(deepclouds)separatingcloud- Zone-likeconditionsbeginningatdepth. freeSEB(N)andSEB(S),visiblybrowncolourstillpresent(Fig. 1c). 2009-Jul-24 NarrowlaneofSEBZhigheropacityobservedat8.6-µm(upperclouds); AerosolopacityinnewSEBZappearsdif- SEB(N)andSEB(S)appeardiffuseandcloud-free(Fig. 1c) ferentin4.8µm(deepclouds)and8.6µm (uppercloud)images. 2009-Aug-04 Fifthbrownbarge(B5)formsfurthestwestoftheGRS Circulation forming/revealing brown bargeshaspropagatedwestward. 2009-Aug-06 Opacity of SEB(S) and SEB(N) equal at 8.6-µm, but 4.8-µm opacity SEB(S)fadewestofGRShadbegun. considerablyhigheroverSEB(S)thanSEB(N);SEB(S)8.6-µmopacity is higher west of the GRS than to the east (Fig. 2a); brown colour of SEB(S)fadingwestofGRS 2009-Oct BrowncolouroftheSEB(S)eastoftheGRSbeginstofade,bulkofthe Fade had progressed around the planet SEBacquiredapaleorangetint. fromitsstartingpointwestoftheGRS. 2009-Oct-04 BrilliantwhitespotobservednorthoftheGRS(Rogers,2010a) Uncertainconnectiontofadeprocess 2009-Nov All barges persist in faint red-brown southern SEB, only SEB(N) re- Fadehasnotyetcompleted,butcoolSEBZ mains relatively cloud-free; SEB(S) has high cloud opacity at 4.8 and isfullyestablished. 8.6µmandcannotbedistinguishedfromSEB(Fig. 2a) 2010-Jan End of 2009 apparition; SEB(N) narrow and dark; most of SEB pale Fadehadnotyetcompleted. orange;5bargesdarkandconspicuous 2010-Apr Start of 2010 apparition, all of SEB is pale, brown barges still faintly SEBfadeproceededalmosttocompletion visible whileJupiterwasobscuredbytheSun. 2010-Jul-01 SEB(S) is completely absent in 4.8-µm imaging, SEB(N) can still be Deep cloud opacity completely covers seen (Fig. 2c). Visible images show brown barges have almost com- SEB pletelyfaded,butstillpartiallyoutlinedbyfaintblue-greypatches. 2010-Jul-13 SEB(S)(narrowandundulating)andSEB(N)(broad)aredetectableat SEBfadehascompleted. 8.6 µm, barges cannot be distinctly observed (except at 10.8 µm, Fig. 5);centralSEBcoveredbyhighopacitycloud(Fig. 2c) 7 Table5:CharacteristicsofnoteduringtheSEBfade Characteristic TypicalLatitude TemporalBehaviour GRSWake 7-20◦S Filamentary turbulence normally dominates region NW of GRS (Fig. 1a-b; became quiescentinMay-June2009(Fig. 1c);preludetotheformationofSEBZandthefade. SEB(N) 7-12◦S Dark, narrow brown lane separating SEB and EZ (Fig. 1a); warm cloud-free condi- tionspersistedthroughoutfade;opacityincreasedbutdidnotobscure4.8and8.6-µm emissioncompletely(Fig. 2c). SEB(S) 15-18◦S Broad dark band of variable width separating SEB and STrZ; began to fade west of GRSinJune-July2009; eastoftheGRSbyOctober-November2009. ByJuly2010 deepcloudopacitycompletelyobscuredSEB(S)4.8-µmemission;whereasanarrow laneoflowupper-cloudopacity(8.6µm)wasvisible(Fig. 2c). SEBZ 11-15◦S Coolzone-liketemperaturesdevelopedJuly-August2009for p > 300mbarandper- sisted to July 2010 (Fig. 5; Fig. 9); coincides with increased opacity of SEB centre andreductionofzonalwindshearinuppertroposphere(Fig. 6). SEB(S)Undulations 17-19◦S Unique characteristic of fully-faded state in July 2010 (Fig. 2c); zonal oscillations ofcloudopacitywith5-6◦longitudewavelengthandretrogrademotion;furthersouth thanthetypicalSEB(S)edge. SEB(N)Projections 6-9◦S Persistent features of SEB(N), wave activity at the boundary between clouds of the EZandSEBmixescloudopacity. Nochangeduringthefade. BrownBarges 14-16◦S FivedarkbrownbargesintheSEB(S)(B1-B5)formedwestoftheGRSbetweenJune- August2009(astheSEBZwasforming,Fig. 2). Allwerecloud-freewithwarmcore temperatures. Not observed in previous fades (Sanchez-Lavega and Gomez, 1996). BargesfadedbyJuly2010,butwarmcycloniccirculationwasstillpresent(Fig. 5). GRSWhiteSpot 9-11◦S Whitecloud(vigorousconvectiveplume)nearNedgeofGRSinJuly2010;high8.6- µmopacity(Fig. 10c),givesrisetoablue-greystreaknorthwestoftheGRS(Rogers, 2010a). Uncertain connection to the fade process, but may have also been present during1989-1993fades(Rogers,2010a). SEBD 7-20◦S ExpectedvigorousconvectivedisturbancethatwillsignalthestartoftheSEBrevival. 8 (SEB(N))and15-17◦S(SEB(S)),andazone-likeminimumat 12◦S(theSEBZ).ThetransitionfromtheturbulentSEBtothe cool SEBZ by late 2009 is clearly shown in maps of 10.8-µm brightnesstemperatures(Fig. 5),sensitivetotemperaturesand (a) VISIR Radiances ammonianearthe400-mbarlevel. Wavelength (μm) 20.017.0 14.0 12.0 10.0 9.0 8.0 3.2.1. SEBZTemperatures 400 Thechangesinthemid-IRradianceinFig. 4manifestthem- 2008 Sep selvesasvariationsintheretrievedtropospherictemperaturesin 2009 Nov Fig. 6, particularlythoseathighpressures(480and630mbar ) -1m 2010 Jul inthepanelsfandh), indicatingadifferenceinthelatitudinal c r/ temperature gradient between 2008 and 2009-10. Uncertain- s 2m/ 100 tiesintheabsolutecalibrationoftheVISIRimages,combined c with the broad pressure range covered by the weighting func- W/ tion for each filter (Fletcher et al., 2009b) and the degenera- n ( cies between temperature and composition, lead to consider- e c ableretrievaluncertaintiesfortemperaturesat p>500mbarin n a Fig. 6e-h, so that absolute cooling is difficult to detect at the di a SEB latitude. On the other hand, the presence of the distinct R cool zone (SEBZ) from 2009 onwards, relative to the warmer SEB(N)andSEB(S),canbeseenat480and630mbarinFigs. 10 6f-h. Temperaturecontrastsof1.0±0.5KbetweentheSEBZ 600 800 1000 1200 and the northern and southern components were measured at Wavenumbers (cm-1) 630mbarinJuly2010(Fig. 6h). OfthethreeQ-bandfiltersin (b) VISIR Brightness Temperatures Fig.4,onlythedeep-sensing19.5-µmfilter(whichhasthelow- Wavelength (μm) est diffraction-limited resolution of all the images used in this 20.017.0 14.0 12.0 10.0 9.0 8.0 study)detectedtheSEBZ,confirmingthattheSEBZformation occurred at depth. The cool SEBZ was not observed at lower 150 2008 Sep pressures (150-300 mbar) sensed by 17.6- and 18.7-µm filters 2009 Nov ) wheretheSEBretaineditsusualwarmbelt-likeconditions. K 2010 Jul ( Indeed, the temperature fluctuations in the stable upper tro- e r posphere between 15-300 mbar were small compared to those u 140 at at other latitudes, particularly those associated with the NEB r e p (see Appendix A). The SEBZ formation at depth did have a m subtle effect at these lower pressures - both the retrieved tem- e T s 130 peraturesnear240mbar(Fig. 6d)andtherawradiances(Fig. s 4)demonstratea‘flattening’ofthemeridionaltemperaturegra- e htn dient (dT/dy) between 7-20◦S as the fade progressed. Instead g of having a belt with a peak temperature in the centre (12◦S), ri B the240-mbartemperaturesduringthefadedstatebecamemore 120 homogenised with latitude (dT/dy tends to zero between 10- 15◦S at 240 mbar, Fig. 6d). As vertical shears on the zonal 600 800 1000 1200 wind (du/dz) are related to dT/dy by the geostrophic thermal Wavenumbers (cm-1) windshear equation, this reduction of dT/dy in 2009-10 com- pared to the normal state of the SEB suggests a reduction in Figure3: Comparisonofsynthetic8-pointspectra(lines)tomeasuredzonal the windshear on both the prograde SEBn jet at 7◦S and the mean radiances (points) at the centre of the SEB (14.5◦S) at three different retrogradeSEBsjetat17◦Sinthetroposphere. Zonalflowas- epochs-September2008(normalstateoftheSEB);November2009(midway sociatedwiththesetwoopposingjetscouldthereforepersistto throughthefade)andJuly2010(oncethefadehadcompleted). TheNEME- SISretrievalmodeliscapableofreproducingthemeasuredradiancesinall8 higheraltitudes(i.e., nearthetropopause)duringafadedstate filters.Therewerefewdifferencesinthecalibratedradiancesbetweenthethree thanduringthe‘normal’state. epochs,exceptacoolingtrendat8.6and13.1µmasthefadeprogressedand Finally,stratospherictemperaturesabovetheSEBshowcon- 7.9-µmvariabilityduetoJupiter’squasi-quadrennialoscillation(QQO).Radi- siderable variability between 2008 and 2010 (Fig. 6a) as part ancesinthetoppanelwereconvertedtobrightnesstemperaturesinthebottom panel.ThesensitivityofeachoftheeightfiltersisdescribedinTable2. ofJupiter’sQuasiQuadrennialOscillation(QQO,Leovyetal., 1991), which affects the stratospheric temperatures between 20◦Nand20◦Swithaperiodof4-5years. However,nocausal connection between the fades and the stratospheric oscillation is found: (i) the regularity of the QQO means that it cannot 9 (a) 8.6 μm (b) 10.8μm K) K) e ( e ( 150 atur 170 atur 145 2008 Sep per per 140 2009 Jul m 160 m e e 2009 Aug ess T 150 ess T 113305 22001009 JNuonv/Jul n n ght 140 ght 125 Bri 60 50 40 30 20 10 0 -10 -20 -30 -40 -50 -60 Bri 60 50 40 30 20 10 0 -10 -20 -30 -40 -50 -60 Planetocentric Latitude Planetocentric Latitude (c) 13.0μm (d) 17.6μm K) K) ure ( 140 ure ( perat 135 perat 125 m m e e s T 130 s T 120 s s e e n n ght 125 ght 115 Bri 60 50 40 30 20 10 0 -10 -20 -30 -40 -50 -60 Bri 60 50 40 30 20 10 0 -10 -20 -30 -40 -50 -60 Planetocentric Latitude Planetocentric Latitude (e) 18.7μm (f) 19.5μm K) K) e ( 130 e ( atur atur 130 2008 Sep er er p 125 p 2009 Jul m m 125 e e 2009 Aug T T ss 120 ss 2009 Nov e e 120 2010 Jun/Jul n n ht ht g g Bri 60 50 40 30 20 10 0 -10 -20 -30 -40 -50 -60 Bri 60 50 40 30 20 10 0 -10 -20 -30 -40 -50 -60 Planetocentric Latitude Planetocentric Latitude Figure4:Centralmeridianbrightnesstemperaturescansfrom2008to2010insixoftheeightfiltersobservedbyVISIR.Northernandsouthernhemisphereimages werenotacquiredsimultaneously,soagapispresentattheequatorfortheearliestdates(Table1). ThelocationoftheSEBisdenotedbyverticaldottedlines. Thebrightnesstemperatureshavebeenoffsetfromthe2010measurements(blackcurve)byarbitraryamountsforclarity.Themostdramaticchangeoccuredat8.6 µm(decreasedemissionduetotheincreasedaerosolopacityovertheSEB).TheformationofthecoolSEBZisalsoapparentinfilterssensitivetotemperaturesat p>300mbar(10.8,13.0and19.5µm),butonlyasubtle‘flattening’ofthemeridionaltemperaturescanbeobservedat17.6and18.7µm(sensitiveto150-300 mbar).Brightnesstemperatureshavenotbeencorrectedforemissionangle,sothegeneraldecreasefromequatortopoleineachfilteristheeffectoflimbdarkening. 10