Astronomy&Astrophysicsmanuscriptno.aa20078311 (cid:13)c ESO2008 February2,2008 Analysis and interpretation of a fast limb CME with eruptive prominence, C-flare and EUV dimming S.Koutchmy1,V.Slemzin2,B.Filippov3,J.-C.Noens4,D.Romeuf5,andL.Golub6. 1 Institutd’AstrophysiquedeParis,CNRSandUniv.P.&M.Curie,98BisBoulevardArago,F-75014Paris,France 8 e-mail:[email protected] 0 2 P.N.LebedevPhysicalInstitute,Leninskypr.53,Moscow,119991,Russia 0 e-mail:[email protected] 2 3 PushkovInstituteofTerrestrialMagnetism,IonosphereandRadioWavePropagation,RussianAcademyofSciences(IZMIRAN), TroitskMoscowRegion,142190,Russia n e-mail:[email protected] a 4 OMPandPicduMidiObservatory,France J e-mail:[email protected] 7 5 C.R.I.ClaudeBernardLyonIUniversity,O.A.-Fiducial,France 1 e-mail:[email protected] 6 Harvard-SmithsonianCenterforAstrophysics,60GardenStreetMS58,Cambridge,MA02138USA ] e-mail:[email protected] h p ReceivedJuly00,2007;accepted - o r ABSTRACT t s a Aims.Coronal mass ejections or CMEs are largedynamical solar-corona events. The mass balance and kinematics of afast limb [ CME,includingitsprominenceprogenitorandtheassociatedflare,willbecomparedwithcomputedmagneticstructurestolookfor theiroriginandeffect. 1 Methods.Multi-wavelengthground-basedandspaceborneobservationsareusedtostudyafastW-limbCMEeventofDecember2, v 2003,takingintoaccountbothonandoffdiskobservations.ItseruptingprominenceismeasuredathighcadencewiththePicduMidi 6 full Hαline-fluximaging coronagraph. EUVimages fromSOHO/EITand CORONAS-F/SPIRITspace instruments areprocessed 4 includingdifferenceimaging.SOHO/LASCOimagesareusedtostudythemassexcessandmotions.Computedcoronalstructures 7 fromextrapolatedsurfacemagneticfieldsarecomparedtoobservations. 2 Results.Afastbrightexpandingcoronal loopisidentifiedintheregionrecordedslightlylaterbyGOESasaC7.2flare,followed . byabrighteningandanaccelerationphaseoftheeruptingmaterialwithbothcoolandhotcomponents.Thetotalcoronalradiative 1 fluxdroppedby∼7%inthe19.5nmchannelandby4%inthe17.5nmchannel,revealingalargedimmingeffectatandabovethe 0 limbovera2hour interval.Thetypical3-partstructureobserved1hourlaterbytheLascoC2andC3coronagraphs showsacore 8 shapedsimilarlytotheeruptivefilament/prominence.ThetotalmeasuredmassoftheescapingCME(∼1.5·1016gfromC2LASCO 0 observations)definitelyexceedstheestimatedmassoftheescapingcoolprominencematerialalthoughassumptionsmadetoanalyze : v theHαeruptingprominence, aswellasthecorresponding EUVdarkeningofthefilamentobservedseveraldaysbefore,madethis i evaluation uncertain byafactor of 2. Thismass budget suggests that theevent isnot confined totheeruption region alone. From X thecurrentfreeextrapolationwediscusstheshapeofthemagneticneutralsurfaceandapossiblescenarioleadingtoaninstability, r includingthesmallscaledynamicsinsideandaroundthefilament. a Keywords.Sun:activity–Sun:filaments–Sun:prominences–Sun:coronalmassejections(CMEs)–Sun:flares 1. Introduction 2000;Cremades&Bothmer2004;Gopalswamy2006)thisques- tion has not been systematically considered and no clear an- CoronalMassEjections(CMEs)wereoriginallydefinedaslarge swerexists.OnepossibilityisthatthemassoftheCMEcomes coronaldynamicalphenomena(coronaltransients)propagating fromarelatederuptingprominence(EP)anditsimmediatesur- outwardinthefieldofviewofwhite-light(W-L)externallyoc- roundings. This assumption was seriously considered recently cultedspace-bornecoronagraphsandoftenshowingthreeparts andpromptedanotherquestion:whatisthemassoftheEP(see (seeWagner1984).Theirprimarypropertyisthemasstransport Gilbert et al. 2005, 2006)? However, the general consensus is ofapresumablycoronalmassoriginallysituatedbehindtheoc- thatonlythecoreofthe3-partCMEcorrespondstothepromi- culting device, toward the externalpart, without specifyingthe nence. originofthismassrevealedbytheW-Leventduetothescatter- ingonfreeelectrons.Modelershadtoaddressthefundamental Adistinctionhasbeenmadein consideringthepresumably question:fromwheredoesthislargeamountofcoronalplasma more importantfast CMEs, usually related to a flare and more originate?AlthoughmanystatisticalstudiesofCMEswerepub- likely to be geoeffective and with space weather implications lished in recent decades (among the latest, see Vourlidas et al. (seee.g.Baoetal.2007).TheEPconnectionisthensometimes omittedin favor ofa discussionof theoriginof solar energetic Sendoffprintrequeststo:S.Koutchmy particlesthatcontributeonlyalittletothemassbudget. 2 S.Koutchmyetal.:FastlimbCMEwitheruptiveprominence CMEsareoftenassociatedwitheruptingfilamentsorpromi- nencesevenwhenconsideredfarfromtheSun(seeBothmer& Schwenn1994).Inthelastfewyearsmanyinvestigationswere undertakentostudytherelationshipbetweenprominenceactiv- ity and CMEs in different wavelengths. Based on Hα observa- tionsattheMaunaLoaSolarObservatory,Gilbertetal.(2000) proposedaclassificationofactiveanderuptiveprominencesac- cording to their kinematical properties (radial height, velocity and acceleration). A separation between escaping prominence materialliftingawayfromtheSunandmaterialreturningtoward thesolar surfacewouldoccurin theheightrangefrom1.20R ⊙ to1.35R whichcouldberelatedtotheformationofanX-type ⊙ neutrallineinthisregion. Gopalswamy et al. (2003) studied the temporal and spatial relations between the signatures of prominence eruptions de- tected in microwaveswith the NobeyamaRadioheliographand white-light CMEs. It was shown that in most of the studied eventstheprominencemovedpredominantlyintheradialdirec- tion,withCMEsandEPsstartingatnearlythesametime.Jing et al. (2004) presented a statistical study of filament eruptions observed in Hα at the Big Bear Solar Observatory with other phenomenaofsolaractivity.Theyfoundthatmosteruptions(> 50%)wereassociatedwithwhite-lightCMEs,activeregionfila- menteruptionsbeingmoreassociatedwithflaresthanquiescent filamenteruptions.BecauseoftheircloserelationshiptoCMEs, EPs were regarded as one of the main sources of material in themassbalanceofCMEs.Indeed,fromthetheoreticalmodel- ingpointofview,themassoftheprominencecouldbeanissue when consideringthe launchof the CME (Low et al. 2003) or, more generally, its associated magnetic structure (see Chen et al. 2006).Furtherout,therearestrongindicationsthatthe core inthethree-partCMEstructureisassociatedwithaprominence butthisrelationisnotdirect;multi-wavelengthobservationsare needed. EvaluationsoftheEPmassaretraditionallybasedonobser- vations of the Hα absorption of the correspondingfilament on the disk prior to eruption,where they are seen as dark objects. Thedestabilizationoffilamentsappearsasasuddendisappear- ance in Hα or a disparition brusque (DB) (Tandberg-Hanssen 1974;Soru-Escaut&Mouradian1990)whichcanbecausedby an increase of the plasma ionization degree (thermal effect) or which can also be a result of Doppler- Fizeau effects (dynam- icalDB)includingadarkeningproducedbyrapidmassmotion (Heinzel& Rompolt1987).With TRACE usuallya large,high butfaint∼1MKloopmovesoutwardbeforeanythingelsehap- pens. Theapproachtoestimatingthehydrogendensityinfilaments from Hα absorption was developed by Mein et al. (1996) and Heinzel et al. (1999) who used the non-LTE cloud model with optical thickness, Doppler width and velocity as input param- eters. Because the hydrogen ionization degree in the non-LTE modelisafreeparameter,thetotalmassofthefilamentcannot bedeterminedunambiguously.Whentheprominenceisseenin emission outside the disk, a more direct evaluation can be at- tempted using both narrow passband W-L coronagraphicmea- surementsandHαline-fluxemissionimaging(seeKoutchmy& Nikolsky 1981). To make filtergrams, a narrow passband Lyot filterisusuallyusedandthedataneedadditionalcalibrationsto takeintoaccountthebroaderlineprofile. Fig.1. TheBBSO Hαimage(a)showingthezoomboxcorre- This problem was partly overcome by combining Hα ob- spondingtotheMDImagneticmap(b)oftheAR10508region servationswithobservationsoffilamentsincoronalEUVlines, with superimposedcontoursof filaments (red) from the BBSO where filaments are seen as dark objects on the backgroundof the underlying corona, on the disk or even outside the disk. Hαimage;(c)EITimageoftheboxregionat195Å.Allimages EUVobservationsoffilamentswerestartedwiththeSkylabmis- aretakenonNovember26,2003. S.Koutchmyetal.:FastlimbCMEwitheruptiveprominence 3 HerewereportonalargeW-limbevent,whichoccurredon December 2, 2003, 9-14h UT, in the vicinity of the old multi- polarregionAR10508whichwaswellobservedearlierbutwas still on thedisk at the time ofthe CME. We analyse the obser- vationsoftheprominenceeruptiontakeninHαfull-lineimages with the HACO coronagraph of the Pic du Midi Observatory (Romeufetal.2007),imagesofthesurroundingcoronainEUV withSOHO/EIT(Delaboudiniereetal.1995),andCORONAS- F/SPIRIT telescopes (Zhitnik et al. 2002), and W-L images of theassociatedCMEtakenwiththeLASCO(SOHO)C2andC3 coronagraphs.We discussindetailthe spatialandthetemporal developmentofthe eruption,evaluatethekinematicsandespe- ciallythemassbudgetoftheCME-relatedphenomenonandwe proposeapossiblescenariooftheevent.Thisapproachisquite similar to the early work of Plunkett et al. (2000) where a fast limbCMEeventwithanEPwasstudied,althoughnoflarewas reported (possibly because the region was already beyond the limb).Herewepayspecialattentiontothemassbudget,includ- ing the measurementof the EUV dimming effect, and we pro- poseamodelapplicabletothisparticulareventbasedonacom- Fig.2.PartialframefromanHαfiltergramtakenbeforetheerup- putedmagneticfieldcontexttodescribethewholeevent,instead tionfromtheInternationalHαNetworkand,atleft,thededuced ofproposingauniversal2.5Dmodel. setofcalibratedisophotesusedto evaluatethe massoftheoff- diskprominence.Notethelevelsofscatteredparasiticlightout- side the solar disc (some typicalintensities are shown near the bottompart). 2. Observations 2.1.Thecontext sionandthencontinuedbySOHO(EITandCDS)andTRACE. Batchelor&Schmahl(1994)interpretedtheEUVabsorptionof Agroupofthreemassivefilamentswasobservedoneweekbe- coolobjectsasacontinuumabsorptionofthecoronalradiation fore the eventby the Global High-ResolutionHα Network ob- by neutral H, and by neutral and ionized He dependingon the servatories (BBSO image in Fig. 1a) in the vicinity of NOAA wavelength. Kucera et al. (1998) and Penn (2000) used CDS AR 10508, which is one rotation after the famous AR 10486 dataforprominencediagnosticsandobtainedacolumndensity thatproducedextremelypowerfulX17flaresonOctober29and of neutral hydrogenH of ∼ 4·1017 −1018 cm−2. Mein et al. X28 on November 4 during the preceding Carrington rotation. (2001)haveshownthatacombinationofHα,Ca(8542Å)and Thelargestfilament(whichwewillexaminefurther)locatedto TRACE 171 Å data may yield a more self-consistent determi- thenorthofthecenteroftheAR,wasdividedintotwopartsby nation of the neutral hydrogen column density, of the electron astrongnegativemagneticfieldpatchseenwithMDI(Fig.1b). densityandoftheelectrontemperature. The filament was seen in all EUV wavelengths with EIT and Heinzeletal. (2001)andSchmiederetal.(2003, 2004)no- with SPIRIT in the 175 Å and 304 Å bands as a dark fea- ticed that filaments are more extended in EUV spectral lines ture.Incoronallinesthefilamentwascoveredbyoverlayinghot loopsoftheactiveregion.Duringaweek,fromNovember25to with wavelengths below the hydrogen Lyman-continuum edge December2,GOESobservedmorethan20X-rayflaresinthis (912Å)thaninHαandtheyintroducedtheterm“EUVfilament active region and the filament was not yet destabilized by the channel”(EFC) to describetheir observations.A spectroscopic flares. modeloftheEUVabsorptionoffilamentsandprominenceswas suggested by Heinzel et al. (2003a, 2003b), Anzer & Heinzel We estimatethegeometricalparameters,intensitiesandde- (2005), and Schwartz et al. (2006). The method takes the ab- rived physical values of two parts of the filament marked in sorption of EUV-line radiation by the Lyman continuum self- figure 1b (top half) as f1 and f2, from its Hα image taken on consistentlyintoaccount,aswellasthevolume-blockingeffect November 26, 2003 at 17:57:25 UT. The data are shown in that is potentially important for coronal lines. A similar tech- Table 1 (the thickness D was not measured and was taken as niqueforderivingprominencemassfromEUVabsorptionfea- equaltotheaveragevisiblewidth).Takingintoaccountthatthe tures was developed by Gilbert et al. (2005, 2006). However, measuredintensityisthesumofpartlyabsorbedbackgroundand when the EUV filament is seen as a prominence on the off- foreground,weestimatedtheopticalthicknessinHα.Assuming diskcoronalbackground,oftenanedge-brighteningaroundthe the mean temperature of the filament body (8000 K) and fol- prominenceisobserved,stronglysuggestingthatheatedcoronal lowingtheapproachofHeinzeletal.(2003a),weestimateτ 912 materialisconcentratedattheperipheryoftheprominence.De which gives the neutral hydrogen density n , electron (H-ion) 1 Boeretal.(1998)analyzedtheperipheryofaprominenceusing densityn andthetotalhydrogendensityn .Thisgivesthemass e H transitionregion(TR)lineemissionanddescribesthisenhanced ofbothpartsoftheHαfilament M = 1.3·1015 g.According Hα emissionwithnon-thermalbroadeningsandvelocitiesreaching toAulanier&Schmieder(2002),andHeinzeletal.(2003b),the valuesupto45km/s,whichisconsiderablyhigherthanwhatis EFCdefinedabovecontainsanadditionalmassof50-100%of observedusingtheprofilesofcoollinesinsidetheprominence. theHαfilamentmass,althoughwearenotsureweseetheEFC DuringtheonsetofourEPweobserveda similarcoronaledge in our case, see Fig. 1. The total mass of the filament material brightening. (Hα+EFC)isthenestimatedasM =(2.3±.3)·1015g. F 4 S.Koutchmyetal.:FastlimbCMEwitheruptiveprominence Fig.3. Active phase of the prominence eruption seen with HACO in Hα (a)-(c) and in EIT 195 Å (d)-(f). The arrow marks the possibleplaceofemergenceofthemagneticloop.Inthebottompanels,thetimevariationsofthetotalfluxinHαmeasuredwith HACO(g)and,overthefulldaytimeinterval,inX-rays1-8ÅmeasuredwithGOES-12(h). 2.2.Theeruptingprominenceandtheflare inthefullHαlinewiththeHACO(H-AlphaCOronagraph)im- ageratPic duMidiobservatory(seeRomeufetal. 2007foran On December 1 all day and on December 2 before 09:00 UT, exhaustive description of HACO observations) with a cadence a filament previously studied on the disk was observed on the of15 s. Abovethe limb a prominenceis seenin Hα asa lumi- westlimbinHαandinEUVlines(withSPIRITandEIT)asa nous object because of its large opacity to UV and EUV radi- quiescentprominencewithacenterofmasslyingataheightof ation from the corona, the TR and the chromosphere (Heinzel ∼63Mm.Lateritbecameaneruptingprominence(EP)thatwe &Rompolt1987).HACOcollectsthefullfluxproducedbythe studiedindetail.After09:10UTtheprominencewasobserved Hαline(spontaneoustransition3-2afterlevel3isexcited),in- S.Koutchmyetal.:FastlimbCMEwitheruptiveprominence 5 Table1.Geometricalandphysicalparametersofthefilament. Filamentpart D,cm S,cm2 V,cm3 τ τ n ,cm−3 n ,cm−3 n ,cm−3 M ,g Hα 912 1 e H Hα f1 1.1·109 2.2·1019 2.2·1028 0.48 49 7.9·109 1.2·1010 2.0·1010 1.0·1015 f2 1.1·109 3.9·1018 3.9·1027 0.85 96 1.5·1010 1.6·1010 3.1·1010 2.9·1014 wellwithinthepass-bandofthefilter,providedtheirtransverse componentsdonotexceed70km/s. We looked into the details of these short brightenings us- ing a simultaneous histogram analysis of each consecutive im- age taken at high rate (15 s cadence;not shown here),keeping inmindallpossiblevariationsofinstrumentalandatmospheric origin(seeing),to concludethatthe prominencewas reallyag- itated “inside”. But the details of this analysis are beyond the scopeofthisworkandwepresenthereonlythesummarycurves shown in Fig. 4 with some small scale rapid variations which couldpartlybeproducedbyseeingeffects.Wealsonotethatthe increase ofHα flux is difficultto attribute to a suddenincrease ofthe densityorofthe pressureinsidethe prominence:neither themorphologynorthekinematicsoftheEPsuggeststhis. AnincreaseoftheturbulenceinsidethebodyoftheEPgiv- inganincreaseinHαemission,precedinganyaccelerationup- Fig.4.Temporalvariationsinrelativeunitsofi)thefullHαline ward(ordownward)ofthemainpartsoftheEP(seeFig.4),is flux, ii) the effective surface area of the prominence (inside a anovelfeaturewhichtellsussomethingabouttheheatingpro- closed contour drawn at half intensity) and iii) the computed cessesoftheEPbeforeitslaunch.Themeasuredbehaviorofthe heights of the center of gravity and of the upper edge of the variationsof the effective surface (Fig. 4) of the prominenceis prominence;after 10:30UT the valuesare less meaningfulbe- an additionalargumentin favor of the “turbulent”scenario be- causetheintensitiesoftheEPdrasticallydrop.Thephaseshifts cause of the observedphase shift between the differentcurves: ofthetemporalvariationsofthesedifferentparametersareillus- i) variationoffluxes,ii) variationsofthe effectivesurfacesand trated. iii) themotionofthe centerofgravityoftheprominence,plot- ted in Fig. 4. One possibility to explain this increase of turbu- lenceinside the prominenceis to considerprocessesleadingto shortperiod(3to6min)oscillationswhichwouldputitsthreads dependentlyof the dynamicstate of the emitting elements(see intorapidtransversemotion.Subsequently,asignificantrateof Fig. 3) thanksto the broad width (0.25nm) of the interference dissipation(thermalheatingduetowavedissipation,forexam- filterusedbythisinstrument.Thisisincontrasttomostinstru- ple)wouldthenincreasetheionizationratioandultimately,will ments taking Hα filtergrams(see Fig. 2) which are affected by maketheEPdisappearinHαandmakeitthenappearinemis- theDopplershiftsinvariablyoccurringduringtheonsetofanEP. sionin the hotterEUV lines. Anargumentin favorofthispro- Inadditionthelevelofscatteredparasiticlightisgreatlyreduced cess comesfromthe examinationof the TRACE moviesof fil- inHACO,comparedtousingafiltergram(seeFig.3a-c). aments seen in absorption in the 171 Å coronalchannel. They At 09:13 UT we observed a new bright small expanding illustrate several cases of agitation phenomena occurring be- coronalloopemerginginthecenterofAR10508(indicatedby fore the activation (rising) leading to ionization and eruption anarrowinFig.3d),”striking”theprominenceanddestabilizing (see the filament movies produced by the Lockheed group; A. it.Figure3a-fshowsthemostimportantstagesofthisprocessin Title,2006,personalcommunication)ofafilament/prominence, Hα (HACO) and using EIT 195 Å observations. EUV images althoughthelimitedfieldofviewdoesnotalwayspermitafull synchronouslyshow the expanding loop and increased heating evaluation of dynamical events linked to a CME. In TRACE, ofthe gasaroundthe prominencebody(also calledthe periph- bright emission (Fe/ and Fe) is usually seen intermixed ery,seedeBoeretal.1998),whichpossiblycorrespondstoapart withthedarkabsorptionduringtheeruption,notalongtimebe- oftheEUVfilamentchannel.Atthesametimethebrightnessof fore(seealsoHerantetal.1991Figs.4,5and6). theprominencebodyincreasedinHαtoitsmaximumvalue(at 09:35UT,seeFig.4).Thisispossiblyduetothegrowthofturbu- After09:41UTthebrightnessofourprominencedecreased lence inside, which would then producea Doppler-brightening inHαandincreasedinEUVduetoheatingandionizationpro- effectaffectingitssmallelements(threads),asGontikakisetal. cesses and it started to rapidly move outward, with the hottest (1997)showedusingadetailedcalculationforthecaseofanEP. partdirectedtowardstheSun.At09:48UTtheGOESX-rayin- Followingtheseauthors,thefullfluxinHαcouldbemagnified tensityreacheditsmaximumvaluecorrespondingtoaC7.2flare. by a factor of 2 to 3, which corresponds to what we observed Figure3e,gshowsthelightcurvesofthetotalfluxinHα,mea- in the body of the prominence for a short time, with turbulent suredwithHACO andinX-rays(1-8Å)measuredwith GOES velocitieswhichkeepthecorrespondingshiftedlineprofilesstill 12. 6 S.Koutchmyetal.:FastlimbCMEwitheruptiveprominence Fig.5.EstimationoftheCMEmassfromLASCOC2images:(left)-theintegratedregionscorrespondingtothefrontalstructure, thecoreoftheCMEandtothestreamer;(right)-theequivalentmasscorrespondingtotheshownintegratedintensities. UT)aweakstreamerwasseennearbyat225◦.Asaresultofthe CME development, the streamer at the periphery of the CME wasdeflectedtotheSouthanditsbrightnessgraduallyincreased withtime.TheCMEhadatypicalthree-partstructure-afrontal bright loop, a dark cavity and a bright core (see Fig. 5). The shapeofthecoreresembledtheshapeoftheheatedEFCin195 Å (Fig. 3f). After 11:06 UT the frontal structure and, 20 min later,thecoreoftheCMElefttheC2fieldofview.InC3thefull CMEwasseenonlyinoneframeat12:18UT. AccordingtotheLASCOcatalogue,theCME(frontalstruc- ture) moved in the C2 and C3 from 4 to 25 R with a nearly ⊙ constantvelocity of 1393 km/s. At 6R the core movedwith a ⊙ velocityof942km/s.Thepointscorrespondingtothecoreinthe height-timegraphmatchwellthesamelineasthelastpointfor the prominence, which suggests a close relationship. The acti- vationandthemotionoftheprominencestartedatleasthalfan hourbeforetheCMEonsetdeterminedfromalinearinterpola- tionoftheLASCOpoints. 3. Massloadingandkinematicsoftheeruptionand theCME Thefirstcomponenttoconsideristhemassdirectlyinjectedor loaded by the EP and its surroundings. We discussed in Sect. 2.1thedeterminationofthemassoftheHαfilamentwhichwas observed well before in the active region underlying the event (Fig. 1) and evaluated its mass, including its immediate sur- rounding,as(2.3±.3)·1015g(thisvaluepossiblymaybedoubled Fig.6.Height-timeprofileoftheprominence,thefrontalstruc- duetoturbulence). ture and the core of the CME (a); profile of the velocity and Weusedseveralmethodstoevaluatethemorerelevantbulk accelerationoftheCMEcore(b).Theshadedareacorresponds massoftheprominencebecomingtheEPjustpriortoitslaunch tothetimeintervalwheretheeruptivematerialwasoutoffields (Fig. 2 and Fig. 3a). The effective surface area of the promi- ofviewoftheinstruments. nence is easy to measure. Then, we compared its average in- tensity and morphologywith those of typicalprominencesthat were measured with the aim of determining typical masses of 2.3.TheCMEinW-L prominenceswith the large 52 cm aperture coronagraphof the TheCMEassociatedwiththeeruptionappearedintheLASCO high altitude station near Kislovodsk. Narrow pass-band filters C2fieldofviewafter10:50UT.Beforethistime(e.g.at10:26 (FWHM of .5 nm) in W-L (a continuum window near 607.3 S.Koutchmyetal.:FastlimbCMEwitheruptiveprominence 7 nm) and around Hα were used to perform a photometric anal- Theestimatedvaluesofmassescorrespondingtothetimeof ysis(Koutchmy&Nikolsky1981).Wefoundatypicalelectron thefullCMEexpansionintheC2fieldofview(11:26UT)are: density integrated along the l.o.s. (or “column” density) of the for the frontalstructure 8·1015 g, forthe core6.6·1015 g, for order of 2.5· 1038 cm−2. Assuming an ionisation rate of 50% thestreamer2·1015 g;forthetotalamountincludingallCME overthe103Mm2effectivesurfacewehaveamassof1.2·1015 parts and the streamer 1.1 · 1016 g. The values for the frontal g.Insidetheprominencetheionisationratiocouldbelowerand structure and for the core partly include the moving streamer the estimated mass of the prominencewill increase by a factor componentbecauseitis difficulttoclearlydistinguishbetween of2to4. them. A more realistic estimate is obtainedafter subtractionof thestreamermassfromeachcomponent,whichgivesamassof The second method we used takes advantage of intensity the frontalstructureof 6·1015 g,for the core4.6·1015 g. The measurementsdoneonthe filtergramshownin Fig.2, bycom- sumofthesevaluesgivesafullCMEmassforC2(withoutthe paringthe isophote intensity levels overthe prominenceimage streamer)of10.6·1015 g.ThetotalCME massvalueforC3at with the simultaneouslyobserveddisk intensities. We then can 13:42UT(R = 25R )isestimatedas1.5·1016 g,showingthat ⊙ obtaintheintegratedl.o.s.amountofplasmawithneutralhydro- theprocessofCMEdevelopmenteffectivelycontinued. geninlevel2,andthenassumptionshavetobemadeconcerning Thetotaleruption/CMEprocesscanberoughlydividedinto thethermalequilibriumamongthelevelstodeterminetheabun- threestages:activation,eruptionoftheprominence,andpropa- danceofHatgroundlevel1,wheredensitiesareseveralorders gationoftheCMEintheLASCOfieldofview.Betweenthelast ofmagnitudehigher,etc.Atlocalthermo-dynamicalequilibrium momentwheretheprominencewasseen inEUVor Hα(10:24 (LTE),itwilldependontheexcitationtemperature(Gouttebroze UT)andthefirstappearanceoftheCMEintheLASCOC2field etal.1993)whichwetaketobeoftheorderof6000K.Thede- ofview(10:50UT)thereisa’greyzone’wherethepositionof ducedcolumndensityfromtheaveragedvaluesofHαfluxesof the erupting matter and the parameters of its evolution are not Fig. 2 gives an overallmass, taking into account the extension possibletodetermine. shownbyisophotes,quiteclosetothemassdeducedfromW-L Wemeasuredthepositionsofthemovingprominenceseenin coronagraphicobservations,when a 50%ionisation ratio is as- absorptionfromtheimagestakenwiththeEITandtheSPIRIT sumed(seealsoKoutchmy&Nikolsky1981).Itcouldmeanthat telescopes (from 09:00 UT to 10:24 UT) and combine it with thisassumptionisnottoofarfromreality.However,theionisa- the measured positions of the CME core in the LASCO C2 tion ratio is a matter of discussion: it is certainly not uniform images. The height-time profiles of the prominence/core and inside the prominence (and even across the section of a single frontalstructureareshowninFig.6a,thederivedvaluesofve- thread) and at least at the peripherywhere Hα emission disap- locityandaccelerationofthecore-inFig.6b.Thedottedlinesin pears,theratioiscloseto100%.Further,theassumptionofLTE thesefigurescorrespondtotheprojectedCMEonsettimefrom is certainlynotsatisfied, andfinally the temperatureis notuni- the LASCO CME catalogue (http://cdaw.gsfc.nasa.gov/CME form(seeWiehretal.2007).Inthisrespect,whattheEPregion list) where it was derived from the height-time profile for the lookslikeusingdifferentEITchannelsisimportant.The304Å frontalstructureusingalinearapproximation. image of He+, taken several hours before the eruption, shows the prominence in absorption but surrounded by rather strong emission;itissimilarbutweakerinthe171Å(Fe/)channel. 4. EUVdimmings In the 284 Å channel where the coronal temperature is much higher(Fe),itisnotlikethis:althoughtheprominenceisalso In general, dimmings (areas of temporarily reduced emission seeninabsorption,noenhancementisseenattheperipheryand measure) in EUV or soft X-ray images appear at the places of thecoronalemissionismoreextended.Thistellsusthattheex- thecoronalplasmaoutflowduringaneruptionleadingtoaCME tendedsurroundingofthe EPisanimportantpartofthewhole (e.g. Zhukov & Auchere 2007). It is interpreted as a density eruptionandthatthe massof a largepartofthis areahasto be depletionassociated with theCME initiation(Hudson& Webb takenintoaccount(seethediscussionofthedimmingeffectbe- 1997;Zarroetal.1999andothers).Theprocedureofimagepre- low). processingandofselectionofthedimmingregionsweusedhere wasdescribedindetailelsewhere(Chertoketal.2004;Chertok In view of the large dispersion foundwhen trying to deter- &Grechnev2005;Slemzinetal.2005,2006).A detailedanal- minethemassoftheprominencefromtheobservedHαflux(we ysis of dimmings and their relationship with CMEs have been donotdiscusserrorsproducedbyusingaveragedvalueswithout carriedoutinmanypapers(e.g.Harra&Sterling2001;Harrison goingintothedetailsoftheopticalthicknessvariationsandvari- etal.2003;Grechnevetal.2006andsoon).Mapsofdimming ationsofthe profileofthe line),werefrainfromgivingformu- regionswe obtained in 195 Å and 175 Å are shown in Fig. 7a laeandnumericalresults,asidefromtheroughestimatesgiven and Fig. 7b. Graphs in Fig. 7c display the light curves of the above.Our conclusionis thatmorecoronagraphicobservations total dimming intensities relative to the total intensities in the of prominences are needed, such as observing simultaneously original(notdifferenced)nominalimages. The lightcurvesfor theHαfluxandtheKCafluxtodeduceaparametersensitive bothcasesaresimilar:theyreachedamaximumat∼09:30UT- to theionisationratio of H. Themethodpromotedin EUVby 10:00UTbeforetheCMEonset,thendecreasedtoaminimumat Gilbertetal.(2005,2006)shouldalsobedeveloped(uptonow, 11:30UT-12:00UT.Thedropofintensityinthedimmingarea theirsampleofprominencesgavemassesnotexceeding1015g). relative to the total solar intensity before the eruption is large: We also evaluated the mass of the CME using the LASCO in195Åitisabout7%,in175Åitisabout4.5%.Thespectral imagesandthestandardSolarSoftprocedures,whichincludethe responsefunctionsofbothbandsarequitesimilarexceptthata allocationof regionsof interestoverwhich intensitiesare inte- small,pronouncedhotcomponent(10-15MK)existsinthe195 grated.Ineachimageweseparatelytreatedthreeregions,which Åcase. correspond to the frontal structure, the core and the streamer Wenotethatthebeginningofthegradualdimmingintensity (Fig. 5a). The plots of the calculated masses are presented in decrease in 195 Å (Fig. 7c) corresponds to the moment of the Fig.5b. launchoftheprominence(dash-dotlineinFig.7c)andthemax- 8 S.Koutchmyetal.:FastlimbCMEwitheruptiveprominence Fig.7. Left and center panels - EUV dimmings in the SPIRIT 175 Å and EIT 195 Å difference images. Right panel - the total intensityinthedimmingarearelativetothetotalintensityofthenominalimages.DashedlinecorrespondstotheprojectedLASCO CMEonset(10:15:50UT). imumofintensityoftheX-rayflare.IfweassumethattheEUV where B is the normalmagnetic-fieldcomponenton the plane n dimmingisassociatedwiththeformationofthefrontalstructure, S and r is the radius vector from some pointon the surface to onecanconcludethatboththefrontalstructureandthecoreof agivenpointinthecorona.Zagnetkoetal. (2005) showedthat theCMEwereinitiatedverycloseintime(within5-10min). the prominence material is mainly concentrated at the surface that passes through the apex of the field line arches which is a magnetic neutral surface. Recall that the systematic inclina- 5. Interpretation:themagneticfieldstructureand tion of the bodies of large quiescent prominences at moderate thefilamentequilibrium latitudestoward the west, which had beenwidely describedby d’Azambuja&d’Azambuja(1948),wasexplainedbythepecu- Thereisnodoubtthatthefilamentdestabilizationanditserup- liarities of the structure of the large-scale magnetic fields. We tionwere relatedto the propertiesandthe changesof the mag- calculated the coordinates of the neutral surface in AR 10508. neticfieldintheactiveregion.WefirstnotethattheCMEmor- Figure 8 shows the projection of the magnetic neutral surface, phology (see Fig. 5) seems to fit one of those identified by asitshouldlooknearthelimb,calculatedabovetheneutralline Cremades& Bothmer (2004), where the observedmorphology relatedto thestudiedfilamentuptothe heightof150Mm.We seemstobeconsistentwithwhatisexpectedfortheneutralline cancomparetheshapeofanearlysemicircularlooplyinginthe locationandthepositionoftheeruptionnearthelimb.However, neutralsurface(Fig.8, right)withthe shapeoftheprominence thecoronalmagneticfieldisnotmeasurednorevenqualitatively attheearlystageoftheeruption(Fig.3b-c)whentheloopofthe observed.Accordingly,thispartoftheworkismainlybasedon eruptiveprominenceroughlyconservesitsinitialform.Atequi- extrapolationsmadefromthesurfaceoftheSunandonguesses librium,theprominenceloopshouldbelocatedneartheneutral of the role of the magnetic field. The position of the active re- surface,hencethefrontlegoftheprominenceshouldbecurved gion,close to the limbon December2, wasfavorableformea- andinthemiddlepartnearlyparalleltothesolarsurface,while suringtheplasmadensitydistributionandthevelocityofcoronal thebacklegshouldbemorestraightandtiltedfromthevertical structuresbutmadeitalmostimpossibletoseethechangesinthe directionatanangleofabout30◦.Thisisqualitativelythesame photosphericmagneticfield.Inordertoguessthemagneticcon- aswhatweseeinFig.3b-c. figurationintheeruptingregionwestartedfromthetimewhen Then we found the limiting height of the filament equilib- theregionwasnotfarfromthecenterofthedisconNovember riumh bysolvingtheequation(Filippov&Den2000,2001) 25. c We calculatedthe current-freecoronalmagneticfield1 B in dB B(h ) thevicinityofthefilamentusingthephotosphericmagneticfield =− c . (2) dh(cid:12) h measurementsfrom SOHO/MDI. As the regionof interest was (cid:12)(cid:12)hc c (cid:12) notlargecomparedwiththewholesurfaceofthehemisphere,we (cid:12) This parameter is derived from the vertical stability condition consideredtheboundaryasaflatsurfacerestrictedbythesizeof in the inverse polarity filament model (Van Tend & Kuperus the active region containing the main magnetic concentrations. 1978; Molodensky & Filippov 1987; Forbes & Isenberg 1991) We used the well-known solution for half-space with a plane andreflectsthescaleheightofthemagneticfieldofphotospheric boundary: sources.WecalculatedthelimitingheightusingtheSOHO/MDI magnetogramsofNovember25,27,and29.Theresultsforthe 1 B (x′,y′,0)r B= n dx′dy′, (1) twoformerdaysareverysimilar.Figure9showsthecalculated 2πZ Z r3 neutralline attheheightof6Mm,wherethe fieldis smoothed S enough, overlaid on the photospheric magnetogram. The fila- 1 Generallyspeaking, thepresence ofanon-potential fluxropeisa ment that erupts on December 2 is also shown as a white con- necessarypartofthemodel.Aquitesimilarapproachwasrecentlyat- tour.Inthe middlepartofthe filamentlength,whereitsaxisis temptedinthecontextofanactiveregion(vanBallegooijen&Mackay ratherstraight,thelimitingheightishigherthantheverticalsize 2007). of our calculation domain (150 Mm). Near the filament ends, S.Koutchmyetal.:FastlimbCMEwitheruptiveprominence 9 Fig.8. Projection of the magnetic neutral surface in the AR 10508 on November 25, as it should look near the limb, cal- culatedabovetheneutrallinerelatedtotheeruptiveprominence Fig.9. The neutral line calculated at a level of 6 Mm above (left),andthesideviewofalooplyingintheneutralsurfacethat thephotosphere(yellowcontours)overlaidonthephotospheric modelstheeruptiveprominence(right). MDImagnetogramobtainedon25November2003at19:12UT withintheareausedastheboundaryconditionforcoronalmag- netic field calculations. The filament position taken from the the curvature of the neutral line increases. This means that the BBSOHαfiltergramfor16:45UT25November2003isshown local scale of the photospheric magnetic field is smaller there. asaredcontour. Inaccordancewiththereducedfieldscale,thelimitingheightis lowerand reaches∼ 60Mm. However,nearthe filamentends fieldlinesareanchoredinthephotosphereandthemagneticten- dance with the shape of the neutralsurface, the fore leg of the sionplaysanimportantrole,whereasEquation2 doesnottake archismorecurved.Thearchapexwasslowlyascendingduring intoaccounttheanchoringofthefilamentlegs. severaldaysbeforetheeruptionandreachedaveryhighaltitude On November29 the AR 10508 was close to the limb and byDecember2.Wenotethataquitesimilarschemahasbeenre- the magnetic field measurements became unreliable. We then centlyproposedbyAlexanderetal.(2006)todescribewhathe usedaspecialproceduretoobtainthecomponentnormaltothe calleda“failed”filamenteruptionwithaflareandRHESSIhard solar surface from the line-of-sightmeasurements (Den 2002). X-rayemissionsbutnoCME,sothatweshouldstillbecautious Calculationsshowasmallloweringofthelimitingheightatthe withtheassociationofanEPandaCME.Ontheotherhand,this filamentendsandasignificantloweringinthemiddleofthefil- scenarioagreesratherwellwiththerecentmodelofChenetal. amentdownto40Mm.Thisvaluepossiblyreflectsthetendency (2006),proposedtodescribetheCME-EPphenomenon. ofaloweringofthelimitingheight;however,ashasbeenmen- Thedarkfeaturevisiblebefore09UTinEUVimages,which tioned,the initialphotosphericmagneticfield measurementsin coincides in space with the Hα prominence, overlaps the mid- thisregionarenotreliable. dle part and the far leg of the filament arch along the line of Observationsofthefilament(andthecorrespondingpromi- sight.Buttheactivationandtheascentoftheprominencebegan nence)nearthelimbonNovember30andDecember1demon- near its fore leg. The rising part becomes bright in EUV as is stratethatthefilamentwasstableuptoarathergreatheight.On observedinmanyeruptiveprominences(Filippov&Koutchmy November30theheightofthefilamenttopcanbeestimatedas 2002),whilethefarpartofthearchisstillatrest.Thatiswhywe 35Mm,whileonDecember1-as50Mm.Thegreatestheight see risingbrightfeaturesonthebackgroundoftheratherstatic on December 2, before the filament begins to ascend, was 120 darkformation.Thisfrontalpartofthefilamentbecomesclearly Mm. This value is smaller than the limiting height on 25-27 visibleintheEITimagesinemissionat09:48UTandlater.Itis November (more than 150 Mm) but greater than the limiting alsoclearlyseenintheHαimagesoftheeruptiveprominenceas height on November 29 (40 Mm, possibly underestimated). It acurvedfeature(Fig.3b-c). is not unlikely that the height of the filament reaches a limit- Theappearanceofmovingbrightfeatureswithinthebodyof ing value not long before the eruption, as has been shown for anEPisconsistentwiththemechanismofheatingofthepromi- severaltensoffilamentsbyFilippov&Zagnetko(2007).Webe- nencematerialproposedbyFilippov& Koutchmy(2002). The lieve that the rising of the filament, possibly due to the growth sourceofenergyinthismechanismisthemagneticenergyofthe ofanelectriccurrent(thatwedonotdiscusshere),withinaflux pre-eruptivemagneticconfigurationstoredinthefluxropeelec- rope containingthe filament, up to the threshold of stability, is tric current and its interaction with the surrounding fields and the main cause of the eruption. On the other hand, changes in plasma. The magnetic energy transforms into the macroscopic the active regionmagneticfield can lower the threshold.When kineticenergyofthefilamentmotion,includingitsthreads.Part the heightof thefilamentreachesthe limitingheightanysmall of this energy turns into heat throughcollisions inside the flux disturbanceisabletoinitiatethebeginningoftheeruption. tubeafteranincreaseoftheturbulenceinside.Arathercomplex The sequence of events observed on the west limb on structureofthefluxtubeisnecessaryforthispurpose.Someseg- December 2 between 07:30 UT and 11 UT can then be inter- mentsof thetubeshouldhaveanupwardconcaveformto pro- pretedintheframeworkofamodelofaneruptingfilamentloop. videthecounterflowacceleration.Wenotethatatypicaltwist- Weassumethatthefilamentistheinternalpartofatwistedflux ing flux rope easily meets this requirementand that such twist rope,whichhastheshapeofahalftorus,withbothendsstaying wasindeedobservedinseveralinstances(seee.g.Plunkettetal. anchored in the chromosphere. The loop does not belong to a 2000;Koutchmyetal.2004;Sterling&Moore2004;Anderson flatplanebutliesinthecurvedneutralsurface(Fig.8).Inaccor- etal. ?). Itis also the requirementdeducedfromthe latest the- 10 S.Koutchmyetal.:FastlimbCMEwitheruptiveprominence oreticalanalysis ofthe flux ropemodelproposedfor CMEsby sightintegrationreachesa2-foldincreasedvaluerightabove Chenetal.(2006). the limb with a correspondingincrease of the dimming ef- fect. The dimming is in phase with the onset of the CME. Approximately 5% or more of the corona is depleted, ne- 6. Conclusions glectingthedepletionofthemassofthecoronaatverylow (TR)and/orveryhighcoronaltemperatures.Thecoronaseen We documenteda largeand fast coronallimb event,which oc- curred on December 2, 2003, in the interval 9-14h UT, in the in the EIT 195 Å channel should indeed correspond to the bulkofthemassbecauseitcorrespondstothemostprobable vicinity of an old multi-polar region. Using observations ob- temperatureofthecorona,sothatourestimateshouldnotbe tainedonthegroundandinspacewithtwoEUVtelescopes,an farfromthetruth. HαcoronagraphandtheLASCO white-lightcoronagraphs,we analysedaspectaculardynamicalphenomenonpassingthrough Wealsofoundthattheeruptionledtoaspatialdeflectionand differentstages:aquiescentfilamentinitially,thenaprominence a brightening of the neighboring South-West streamer, which withEUVbrighteningatitsperiphery,itsdynamicalandthermal maygiveriseto anoverestimateofthe CME massby∼ 20%. activation,a coronaleruptionwith a small flare and the launch Further out toward both polar regions the deflections of struc- of the eruptive prominence(EP) and finally the formation of a turessuchasextendedplumesareclearlyvisibleasabackand classicalthree-partfastCME.ThisCMEfitsratherwellwithin forthquasi-coherenttransversemotion.Thisphenomenonwould the frameworkof the so-called “standard model”, (see e.g. Lin needspecialstudywhichisbeyondthescopeofthispaper. 2004).Herewehaveaddedseveralfeaturesthatwebelievecould Finally our results suggest that greater attention should be shed some light on the whole process, emphasizing the mass paidtoregionsaroundtheEP,bringingaddedjustificationtothe transport processes related to the EP and its coronal surround- theoretical studies trying to describe the behavior of the coro- ings. nalmagneticfieldleadingtoCMEs.Howeveronlythefieldin- From what we saw, the eruption process apparently started sideprominencescanbeeasilymeasuredwithcoollinesandthe withtheemergenceofanewmagneticloopstrikingthepromi- Hanleeffect;animprovedprecisionHαcoronagraphcouldrou- nenceanddestabilizingit.Thisledtoarapidheatingofthegas tinelybeusedtoperformsuchmeasurementsathighspeed. around the prominence body, which spatially corresponds to a part of the EUV filament channel. The brightness of the body oftheprominenceinHαatfirstincreasedtoamaximumvalue Acknowledgements. TheauthorsaregratefultoO.G.Denforcoronalmagnetic fieldcalculations,toJ.-C.VialforclarificationsconcerningtheDopplerbright- (at 09:35 UT) possibly due to the growth of turbulence inside eningeffect,toG.Stellmacherforin-depthdiscussionsonthedeterminationof whichproducesalargeDopplerbrighteningeffectand,accord- theprominencemassandtheinterpretation ofprominenceobservations, toC. ingly, magnifies the Hα flux without the need to increase the Delanne´e,G.LawrencefordiscussionsonCMEsandtoA.Zhukovforcarefully densities.Thisbrighteningevenprecedestheliftingofthecool looking atthe 1stdraft ofthe paper. This work was supported inpart bythe RussianFoundationforBasicResearch(grants06-02-16424and05-02-17415) prominence,whichseemstobe animportantnew featureto be andinpartbytheNATOCRG940291.SOHOisaspaceprojectofinternational takenintoaccountinfutureworkonCMEs.Wenotethatthisob- cooperationbetweenESAandNASAandwespeciallythanktheteamssupport- servationwasmadepossiblebytheuseofacoronagraphhaving ingtheEITandtheLASCOobservationsforexcellentworkandprovidingtheir a broad band filter and not a narrow band Hα filter as is usu- datatothecommunity.S.K.alsobenefitedfromCNESincollaboratingwiththe LASCOteaminMarseille.TheHACOground-basedcoronalHαobservations allydonewithasolartelescopetorejecttheparasiticlight.After aresupportedbythe“ObservateursAssocie´s”teamsponsoredbyFiducialLTD 09:41UTthebrightnessdecreasedinHαandincreasedinEUV, andbythe“ObservatoireMidi-Pyre´ne´es”.WearegratefultoM.Audejeanand duetotheenhancedionizationprocessandpossiblyadissipation J.Guignardforcollectingthesequenceusedinthispaper.Wearealsoindebted of the mechanicalenergyfromwaves or counter-flowmotions. totheinternationalHαNetworkforprovidingfiltergramstocoverthechromo- Theprominencebegantomoveoutwardat09:48UT,whichco- sphericactivityin2003. incides with the X-ray flare peak and would precede the CME extrapolatedonsettime. References Atemporalanalysisoftheeruptiveprominencemotionand of the dimming intensity light curves has shown that both the Alexander,D.,Lin,R.,&Gilbert,H.R.2006,ApJ,653,719 frontalstructureandthe core of a CME wereinitiated simulta- Aulanier,G.&Schmieder,B.2002,A&A,386,1106 neouslywithin5-10minaroundthepeakofanX-rayflare. Anzer,U.&Heinzel,P.2005,ApJ,622,714 d’Azambuja,M.&d’Azambuja,L.1948,Ann.Obs.Paris,Meudon,V.6,Fasc. EvaluationsbasedonphotometricdatacomingfromtheW- VII L coronagraphiccoverageofthe CME andfromthe photomet- Bao,X.Zhang,H.Lin,J.,&Stenborg,G.A.2007,A&A,463,321 ricdataprovidedbytheHαline-fluximagesoftheprominence Batchelor,D.A.&Schmahl,E.J.1994,SolarDynamicPhenomenaandSolar before its launch show that the total mass of the CME almost WindConsequences,(Ed.)J.J.Hunt,ESASP-373,203 certainlysurpassesthemassoftheejectedprominencematerial deBoer,C.R.,Stellmacher,G.,&Wiehr,E.1998,A&A,334,280 Bothmer,V.&Schwenn,R.1994,SpaceSci.Rev.,70,215 seeninHαbeforeitseruption: Chen,J.,Marque´,C.Vourlidas,A.Krall,J.,&Schuck,P.W.2006,ApJ,649, 452 – the mass of the CME core is 2-2.5 times greater than the Chertok,I.M.,Slemzin,V.A.,Kuzin,S.V.,etal.2004,AstronomyReports,48, massofthequiescentprominence,includingthemassofan 407 EUV channel(althoughtaking into accountpossible turbu- Chertok,I.M.&Grechnev,V.V.2005,Sol.Phys.,229,95 lence this difference may vanish). This clearly shows that Cremades,H.&Bothmer,V.2004,A&A,422,307 Delaboudiniere,J.-P.,Artzner,G.E.,Brunaud,J.etal.1995,Sol.Phys.,162,291 themassofafastCMEismainlyduetothesurroundinglow Delanne´e,C.,Hochedez,J-F.,&Aulanier,G.2007,A&A,inpress. corona swept out during the eruption, as confirmed by the Den,O.G.,2002,AstronomyLetters,28,345 largedimmingeffectwemeasured.ThispartoftheCMEis Filippov,B.P.&Den,O.G.2000,AstronomyLetters26,322 nottreatedinthispaperbutnewwaysalreadyexistto con- Filippov,B.P.&Den,O.G.,2001,J.Geophys.Res.,106,25177 Filippov,B.&Koutchmy,S.2002,Sol.Phys.,208,283 siderit,e.g.Delanne´eetal.(2007). Filippov,B.&Zagnetko,A.2007,JournalofAtmosphericandSolar-Terrestrial – the dimming effect is observed over a large part of the Physics,inpress. corona,onthediskandalsooutsidethediskwhenthelineof Forbes,T.G.&Isenberg,P.A.1991,ApJ,373,294