Horizontal flows concurrent with an X2.2 flare in active region NOAA 11158

0.53 MB·English

by  Laurent Beauregard

Astron.Nachr./AN999,No.1,1–6(2012)/DOIpleasesetDOI!
Horizontal flows concurrent with an X2.2 flare
in active region NOAA 11158
L.Beauregard1,2,M.Verma2,andC.Denker2,⋆
1 McGillUniversity,DepartmentofPhysics,845SherbrookeSt.W.,Montreal,Quebec,CanadaH3A2T5
2 Leibniz-Institutfu¨rAstrophysikPotsdam,AnderSternwarte16,14482Potsdam,Germany
Received2011Sep5,accepted2012Jan9
Publishedonline2012xxx
Keywords Sun:activity–Sun:flares–Sun:magneticfields–Sun:photosphere–sunspots–methods:dataanalysis
2
1 Horizontalpropermotionsweremeasuredwithlocalcorrelationtracking(LCT)techniquesinactiveregionNOAA11158
0 on2011February15atatimewhenamajor(X2.2)solarflareoccurred.Themeasurementsarebasedoncontinuumimages
2 andmagnetograms oftheHelioseismicandMagneticImager onboardtheSolarDynamicsObservatory. Theobserved
shearflowsalongthepolarityinversionlinewereratherweak(afew100ms−1).Thecounter-streamingregionshifted
n
a towardthenorthaftertheflare.Asmallcircularareawithflowspeedsofupto1.2kms−1 appearedaftertheflareneara
J regionofrapidpenumbraldecay.TheLCTsignalinthisregionwasprovidedbysmall-scalephotosphericbrigthenings,
whichwereassociated withfasttravelingmovingmagnetic features.Umbral strengthening andrapidpenumbral decay
3
wasobservedaftertheflare.Bothphenomenawerecloselytiedtokernelsofwhite-lightflareemission.Thewhite-light
1
flareonlylastedforabout15minandpeaked4minearlierthantheX-rayflux.Incomparisontoothermajorflares,the

X2.2flareinactiveregionNOAA11158onlyproduceddiminutivephotosphericsignatures.
]
R
(cid:13)c2012WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim
S
.
h
p 1 Introduction ciated with flares as well. First spectroscopic observations
- withHMIofaM2.0white-lightflareon2010June12were
o
r Majorsolarflareshaveavarietyofphotosphericsignatures, reported by Mart´ınez Oliveros et al. (2011), who noticed
st e.g.,rapidandpermanentchangesofthemagneticfieldre- thattheentireFeilineprofilewasshiftedtowardstheblue.
a sulting in an increase of the transverse field and magnetic However,interpretingsparselysampledspectraldataduring
[
shear (Wang et al. 2002), penumbral decay and strength- flares (transients in magnetograms or Doppler maps) will

1 ening of umbral areas because initially inclined penum- not lead to conclusive results without detailed knowledge
v bralfield linesbecomemore vertical(Liu et al. 2005),co- ofinstrumentcharacteristics,dataprocessingpipelines,and
0
herent lateral displacement of penumbral filaments (Go- spectral line formation (cf., Maurya, Reddy & Ambastha
0
sain,Venkatakrishnan,&Tiwari2009),elongatedmagnetic 2011).Blue-shifted velocities of up to 1 km s−1 were also
8
2 structures along the polarity inversion line (PIL) (Wang identifiedasacommonprecursorofflaresintheearlywork
. etal.2008;Zirin&Wang1993),andchangesofthevector of Harvey & Harvey (1976). Hints that horizontal shear
1
0 magnetic field around the PIL on spatial scales below one motions are important for the build-up of magnetic stress
2 secondofarc(Kuboetal.2007).Thestatisticalsignificance in flare-productive active regions were likewise presented
1 of such photosphericchanges associated with major flares by Harvey & Harvey (1976). The role of shear flows as
v: wasdemonstratedintheworkofSudol&Harvey(2005). driversformagneticreconnectionwasrecentlydiscussedby
i Battaglia & Kontar (2011) presented a comparison of Yurchyshynetal.(2006)andLiuetal.(2010)inthecontext
X
theheightdependenceofflareemission.HardX-raysorig- ofthetether-cuttingmodel(Mooreetal.2001).
r
a inate from 0.7 to 1.8 Mm above the photosphere,whereas
New telescopes with advanced post-focus instruments
EUVradiationisproducedinthetoplayersofthechromo-
andnoveldataanalysistechniquesmadeitpossibletostudy
sphereatabout3.0Mmabovethephotosphere.Emissionin
such shear motions in more detail. Using local correla-
theopticalrangecoverstheintermediaterangefrom1.5to
tion tracking (LCT, November & Simon 1988), Yang et
3.0MmandwasmeasuredwiththeHelioseismicandMag-
al. (2004) measured shear flows with horizontal speeds of
neticImager(HMI)onboardtheSolarDynamicsObserva-
up to 1.6 km s−1 on both sides of the PIL. Such head-on
tory (SDO) in thecontinuumnearthe Fei λ617.3nmline.
flows were separated by less than one second of arc. The
In contrast, observations by Xu et al. (2004) in the near-
well-definedregionsofelevatedflowspeedshowedagood
infrared at 1.6 µm, i.e., at the opacity minimum, indicate
correlationwith kernelsof white-lightemission of an X10
that deep photospheric layers can exhibit emission asso-
flare – both in the visible and near-infrared continua. The
⋆ Correspondingauthor:e-mail:[email protected] heightdependenceof horizontal(shear)flows in active re-
gion NOAA 10486 was described by Deng et al. (2006).
(cid:13)c2012WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim
2 Beauregardetal.:HorizontalflowsconcurrentwithanX2.2flare
Fig.1 GOES15X-rayflux(5-minutedata)obtainedduringthe
time period from 2011 February 14–16 in the 0.1–0.8 nm (top)
and0.05–0.4nm(bottom)energychannels.Theshadedregionin-
dicates a time interval of three hours centered on the peak time
(01:56UT)oftheX2.2flareinactiveregionNOAA11158.
Shear flows are notlimited to flare-prolificregionsbutare
alsoencounteredinquietersettings(Denkeretal.2007).Ul-
timately,themagneticfieldtopologyandtheinteractionof
magneticfieldsandplasmamotionshastobeconsideredto
decide,ifshearflowscontributetothebuild-upoffreemag-
Fig.2 Limb-darkening corrected continuum image (top) and
neticenergyorleadtomorepotentialmagneticfieldconfig-
magnetogram(bottom)ofactiveregionNOAA11158,whichwere
urations.Finally,basedondataoftheHinodemission,Tan
observedat01:33UTjustbeforetheonsetoftheflare.Theblack
et al. (2009) provided crucial links between photospheric andwhitecontourlinesenclosestrong(above/below±10G)pos-
signaturesofmajorflaresandhorizontalshearflows. itive and negative flux concentrations, respectively. Strong PILs
InVerma&Denker(2011),wedevelopedaLCTalgo- existinregionswherethecontourlinesoverlap.Thecolor-coded
rithm for bulk-processing of Hinode G-band images. This contourlinesindicatekernelsofwhite-lightflareemissionatdif-
work also includes references to alternative methods for ferenttimes.Theaxesarelabeledinheliographiccoordinates.
measuringhorizontalpropermotions,e.g.,differentialaffine
velocityestimator(DAVE,Schuck2006),non-linearaffine
56C-classflaresandfiveM-classflaresincludingtwoM6.6
velocityestimator(NAVE,Chae&Sakurai2008),andball-
eventson February13 and 18. The impulsivephase of the
tracking(Potts,Barrett,&Diver2004).WerefertoWelsch
X2.2 flare started at 01:33 UT on 2011 February 15. This
etal.(2007)foranin-depthcomparisonofthevarioustech-
timewaschosenasareferenceforthesubsequentdataanal-
niques to track horizontalproper motions. In this brief re-
ysis. We divided the dataset into three parts: the 1-hour
searchnote,weadaptedthecodeofVerma&Denker(2011)
pre- and post-flare periods (00:33–01:33 UT and 02:18–
to SDO/HMI continuum images. The X2.2 flare of 2011
03:18UT)andtheintervening45-minuteperiodcoveredby
February15waschosenasapromisingtargettosearchfor
the flare. We assume that data outside of the flare period
alterations of the horizontal flow field after a major solar
is unaffected by flare emissions. The X-ray flux measured
flare.Thethree-dimensionalmagneticfieldtopologyofac-
by the Geostationary Operational Environmental Satellite
tive region NOAA 11158 (observations and MHD model-
(GOES)is shownin Fig.1 fora three-dayperiodcentered
ing), coronal emission structures, and the eruptive events
onthemajorflare.Thegraybarindicatesthethree-hourob-
associated with the X2.2 flare (coronalmass ejection, EIT
servingperiod.
wave,andcoronalfront)aremeticulouslydescribedandex-
ThisstudyisbasedonHMIcontinuumimagesandline-
plainedinSchrijveretal. (2011),so thatwe focusonlyon
of-sight magnetograms. The instrument characteristics are
thephotosphericsignaturesoftheflare.
presentedin a series of articles:Wachter et al. (2011)give
a detailed account of the image quality as measured dur-
2 Observations ingthegroundcalibrationoftheinstrument.Thecharacter-
istics of the optical filter system are laid out in Couvidat
ActiveregionNOAA11158beganitsdiskpassageon2011 et al. (2011), who also discuss them in the context of the
February 11 when it was still classified as a β-region. Be- Feiλ617.3nmline,whichwaschosenasthemostsuitable
sides the X2.2 flare, active region NOAA 11158 produced spectrallineforHMIobservations.Finally,polarizationcal-
(cid:13)c2012WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim www.an-journal.org
Astron.Nachr./AN(2012) 3
Fig.3 Horizontal proper motions in active region NOAA 11158 during the pre-flare phase (00:33–01:33 UT). Arrows indicating
magnitudeanddirectionoftheflowfieldaresuperposedonanaveragepre-flarecontinuumimage(left).Thearrowinthelowerright
cornercorrespondstoaflowspeedof1kms−1.Theflowspeed(right)isgivenaccordingtothescaleontherighttovisualizethefine
structureofthepropermotions.Thecontourlinesrepresentthegranulation/penumbraandpenumbra/umbraboundaries.Theflowmaps
havebeencorrectedforgeometricforeshortening.ThecoordinatesaregivenwithrespecttothecenteroftheROI.
ibration is described in Schou et al. (2010).Since detailed (2011).Sincewearefocussingonexcessemission,sideef-
informationisavailable,weonlypresentabriefaccountof fects of the helioseismic interpolation scheme such as an
theobservingcharacteristicsrelevanttoourstudy. apparent“black-light”flareprecursorcanbeneglected.
The LCT algorithm is an adaptation of the code de-
HMI utilizes a 14-cm telescope (F/37.4). The diffrac-
tion limit at Fei λ617.3 nm is α = λ/D = 0.9′′ pixel−1. scribedinVerma&Denker(2011),whoprovideadetailed
account of how to choose LCT input parameters and how
Thus,thedataisslightlyundersampledwithanimagescale
of 0.5′′ pixel−1. The Fei line is sampled at six line posi- that choice will affect the outcome of the LCT procedure.
HMI continuum images are corrected for limb darkening
tionscoveringarangeofλ ±17.5pm.Full-diskdatahavea
0
and geometricalforeshortening.Theyare resampledon an
size of 4096×4096 pixels. Data are captured with a 45 s
equidistantgridwithaspacingof320km.Thelocalcross-
cadence resulting in 240 images and magnetograms dur-
correlationsarecomputedoverimagetilesof16×16pixels
ingtheperiodfrom00:30to03:30UT.Aregion-of-interest
(ROI)withasizeof260′′×180′′(seeFig.2)wasextracted (5Mm×5Mm).AGaussianwithaFWHMof6.25pixels
(2Mm)isusedbothasaweightingfunctionandahigh-pass
fromthefull-diskdataforfurtherdataanalysis.Thecontin-
filter. Thecadenceforcorrelatingimagepairsis∆t = 90s
uumimageswerecorrectedforthecenter-to-limbvariation
andindividualflowmapsareaverageover∆T = 1h.Hori-
(seeDenkeretal.1999)tofacilitatefeaturerecognitionus-
zontalflowmapsbasedon80continuumimageswerecom-
ing thresholding techniques. The black and white contour
putedforthepre-andpost-flareepisodes.Theflowfieldjust
lines in Fig. 2 were computed using a 10 G threshold for
beforetheonsetoftheflareisdepictedinFig.3.
the smoothed magnetogram (Gaussian with a FWHM of
2Mm).Locations,wheretheblackandwhitecontourlines
overlap, indicate strong PILs, most prominently, the east-
3 Results
westoriented,flaringPILinthecenteroftheROI.
ThephotosphericcontinuumemissionoftheX2.2flare ActiveregionNOAA 11158containsthreemajorsunspots
isdifficulttomeasure.Thepeakcontrastofthewhite-light (seeFig.2).Thewesternspothaspositivepolarityandthe
flarekernelsislessthanafewpercentofthelocalintensity eastern one negativepolarity.The flaring PIL is located in
contrast.Averagevaluesareevenlower.Forthisreason,the the central spot separating negative (north) from positive
quiet Sun intensity contrast of about 3.5% is sufficient to (south) polarity. The PIL runs along an alignment of um-
obfuscatethefaintflareemission.UsingGaussiansmooth- bralcores,i.e.,theregionisalmostvoidofpenumbralfila-
ing (FWHM of 2.0 Mm), division by a 30-minuterunning ments.Thenomenclatureofumbralfine structuresfollows
average, and a local intensity contrast of 5% as a thresh- the definition in Sobotka, Bonet & Vazquez (1993). Basi-
old, we could determine contours of the flare kernels for cally,theactiveregioniscomposedoftwobipolarregions,
about15minstartingat01:49UT.Earlyflareemissionsare which bothappearedon 2011February11 andco-evolved
color-codedfromdarktolightblueinthetoppanelofFig.2 withtheresultthatthetrailingspotofoneregionoverlapped
and the leftand middlecolumnsof Fig. 4, whereaslighter withtheleadingspotoftheotherregion.Thecomplexmag-
to darkergradationsofred outlinelater stages ofthe flare. neticfieldtopologyofthecentralspotistheconsequenceof
TheinfluenceofthetemporalinterpolationfunctiononHMI thisinteraction.WedonotfindanyindicationsintheHMI
continuum images is discussed in Mart´ınez Oliveros et al. magnetogramsofnarrowandelongatedmagneticstructures
www.an-journal.org (cid:13)c2012WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim
4 Beauregardetal.:HorizontalflowsconcurrentwithanX2.2flare
Fig.4 ROI depicting the central region of active region NOAA 11158 where the X2.2 flare occurred. The left column shows a
continuumimage(top)andamagnetogram(bottom)atthepeakoftheflare(01:58UT).Themiddlecolumncontainsdifferencemaps
(post-minuspre-flarephase)ofthecontinuumintensity(top)andline-of-sightmagneticfluxdensity(bottom),respectively.Thesemaps
weresmoothed withaGaussian (FWHMof4.0Mm) toemphasize regionsof flare-induced changes inthephotosphere. Superposed
onallthesemapsarewhite-lightflarekernels,whicharedepictedinthesamewayasinFig.2.Finally,therightcolumncomparesthe
horizontalflowspeedsbefore(top)andafter(bottom)theflare,respectively.Thewhitecircleencompassesacircularareaofenhanced
flowspeed,whichappearedaftertheflare.Orangearrowsindicatethedirectionofthehorizontalpropermotions.Theredcurveinall
panelsreferstotheflaringPIL.
alongtheflaringPIL,whichwereoftenobservedinleague about 500 m s−1 but speeds in excess of 1.0 km s−1 are
with major solar flares (Wang et al. 2008; Zirin & Wang also commonly encountered. All other sunspots show in-
1993). Such features are also absent in maps of the lin- dications of moat flow as well – in particular, when regu-
ear polarization for the pre- and post-flare phases (Wang lar, well formedpenumbraeexist with a radialfilamentary
et al. 2011).There, only an enhancementof the linear po- structure. Using time-lapse continuum and magnetogram
larization is observedafter the flare, which is indicativeof movies(e.g.,Movie1inSchrijveretal.2011)twistingmo-
anenhancementofthetransversemagneticfieldsalongthe tionsbecomeapparentinboththetrailingandcentralspots.
PIL. Flaretransientsarevisible inthemagnetogramsfrom Spotswithnegativepolarityturncounterclockwise,whereas
01:48–02:04UTonbothsidesofthePIL,whicharerelated aclockwisetwistcanbeseeninthepositive-polarityregion
tothetwocentralflarekernels(seeFigs.2and3). ofthecentralspot.Intheselocations,wherepenumbralfila-
Before looking at flows, which are related to the X2.2 mentsarenolongerradiallyaligned,themoatflowiseither
flare,wepresentanoverviewoftheoverallhorizontalflow suppressedordoesnotexistatall. Thispatternis also im-
fieldwithintheactiveregion.Theaverageflowspeedinre- printedonLCTmapsandcanbeseen,whentheflow vec-
gionscoveredbygranulationis424±55ms−1,wherethe torsarevisualizedathighresolution.Furthermore,moving
standarddeviationdenotesthevariationinthefield-of-view magnetic features (MMFs), which are tracers of the moat
(FOV)ratherthanaformalerror.Thesevaluesareingood flowinmagnetograms,followthesecurvedtracksaswell.
agreementwithVerma&Denker(2011)(seeTab.3therein) Shear flows are difficultto detect in the flow maps de-
consideringtheimagescale of320kmpixel−1.Thevalues pictedin Fig. 3,whichencompasstheentire activeregion.
for the average flow speeds in penumbral and umbral re- Therefore, we display zoomed-in versions of the pre- and
gionsare 159±11 m s−1 and 117±5 m s−1, respectively. postflareflowfieldsintherightcolumnofFig.4.Sincethe
Thesevaluesaresignificantlylowerthanflowspeedsprevi- flowsnearthePILareverylow(afew100ms−1)andsome-
ouslyderivedfromHinodeG-bandimages.However,con- timesevenlowerthan100ms−1,wesuperposedarrowsof
sidering the vast differenceamong active regions, they are equallength,whichindicatetheflowdirection,onthehori-
stillreasonable. zontalspeedmaps.Thelowvaluesoftheshearflowscould
The moat flow is most prominent around the lead- be explained by the larger image scale of HMI continuum
ing sunspot, where it extends about 10.0 Mm beyond the imagesascomparedtohigh-resolutionHinodeG-bandim-
sunspotboundary.Typicalflowspeedswithinthemoatare ages.ThePILseparatesthelargerumbralcoresofnegative
(cid:13)c2012WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim www.an-journal.org
Astron.Nachr./AN(2012) 5
polarity to the north from the linear alignment of smaller only slightly varies in size. The other flare kernel initially
umbral cores of positive polarity to the south. These um- straddles the PIL and then moves towards the north. The
bral cores are only separated by a strong light-bridge,i.e., other two flare kernels are located on the other side of the
the intervening space is void of any penumbralfine struc- PILina regionofpositivemagneticflux.Thewesternone
ture.Asaconsequence,onlyveryfewcontrast-richfeatures shows rapid motion to the west. On average, the flare rib-
contributetothespeedmeasuredbytheLCTalgorithm. bons move with a speed of about 10 km s−1 and cover a
Shear flows are encountered along the strongest gra- distance of 5–12 Mm. Based on the continuum emission,
dients of the PIL. In the pre-flare flow map, the PIL oc- thetwo-ribbonflarecanbecharacterizedasanX-typeflare,
cupies in many location areas of zero flow speed, as ex- wheretheeasternflarekernelsareassociatedwiththehard
pected for counter-streaming flows. After the flare, shear X-ray (50–100 keV) footpoints (see Fig. 1 in Wang et al.
flows, which were originally associated with the positive 2011).TheconnectinglineoftheseX-rayfootpointsisper-
magnetic polarity, extend more into the territory of nega- pendiculartothePIL.
tive polarity.Thisshiftof the counter-streamingregionto-
wards the north is cospatial with the region of enhanced
4 Discussionand conclusions
transverse fields reported by Wang et al. (2011). Another
intersting post-flare flow featureis a small circular area of
We have presented some exploratory work adapting the
enhanced flows, which is marked by a white circle in the
LCTalgorithmofVerma&Denker(2011)toHMIcontin-
lower-rightpanelofFig.4.Here,theflowspeedreachesup
uumimages.Wewereabletodetectphotosphericsignatures
to 1.2 km s−1. Small-scale brightenings are visible in the
ofmajorsolarflaresinHMIdata,e.g.,rapidpenumbralde-
continuumimages,whichcrossthevoidwithinthepositive
cay,umbralstrengthening,shearflows,andwhite-lightflare
polarity from north to south. This motion is accompanied
kernels. However, compared to previous studies, the phe-
in magnetograms with fast traveling MMFs. The circular
nomena are rather weak, which is rather surprising for a
area ofenhancedflowsisinclose proximityto a regionof
flareofthismagnitude.
rapidpenumbraldecay.Inatime-lapsmovieofHMIcontin-
Counter-streamingmotionsalongthePILcancontribute
uum images and magnetograms,the overallshear motions
totheshearofmagneticfieldlinesstraddlingthemagnetic
areveryapparentleadingtotheimpressionthatthefluxsys-
neutral line, thus contributing to the build-up of energy,
temofoppositepolarityareslidingalongeachotheronboth
which is then released during a flare (Deng et al. 2006;
sidesofthePIL.
Denkeretal.2007;Liuetal.2010;Yurchyshynetal.2006).
As mentionend in Sect. 1, rapid penumbral decay and The counter-streaming region, which was shifted towards
umbralstrengtheningarecommonphotosphericphenomena the north after the flare, is cospatial with the area of en-
aftermajorsolarflares.Differencemapsforcontinuumand hanced transverse magnetic field identified by Wang et al.
magnetogramsarepresentedinthemiddlecolumnofFig.4. (2011).These authorstake this enhancementas an indica-
Allearlyflarekernelsarelocateddirectlyinanumbralcore tionoftether-cutting(Mooreetal.2001)atlowatmospheric
orincloseproximity.Theseregionsgetdarkeraftertheflare layers(seealsoLiuetal.2010;Yurchyshynetal.2006).We
and the magnetic flux density increases. Only for the flare concludefurtherthatarearrangementofthelow-lyinghor-
kernelwhichlastedlongest(darkredcontoursinFigs.2and izontalmagneticfield canlead to analterationofthe hori-
4),wefindadecayingpenumbralregioninwhichthemag- zontalflowfield.
netic flux density diminishes. However, these flare-related
Another signature of such effects is the localized area
changes are small compared to photospheric changes re-
withelevatedhorizontalpropermotionsnearthePIL(white
portedintheliterature.
circle in Fig. 4), which might be an indication, that after
White-light emission was observed from the initiation stress relief within the magnetic field topology, shear mo-
of the flare (01:49 UT), through its peak time (01:54 UT tions along the PIL can increase again.This notion is also
as compared to 01:56 UT for the GOES X-ray flux in the supportedbytheobservationthattheflowspeedintheprox-
0.1–0.8nmrangeshowninFig.1),untiltheemissionfaded imityofthePILincreasesbyabout5%aftertheflare.How-
away shortly thereafter (02:03 UT). At its maximum, the ever,weshouldpointoutthatthesephotosphericshearmo-
white-lightflareemissioncoveredanareaofabout75Mm2. tions as measured by LCT are still rather small (cf., Deng
TheGaussiansmoothingresultsinflarekernels,whichare etal. 2006;Yangetal. 2004).Thelowflow speedsforthe
somewhatlargerandmorecontiguousthanwhatcanbeseen shearmotionscouldbeattributedtoaPIL,whichislocated
in time-lapse moviesof the originaldata. The time profile between umbral cores so that only few features contribute
of the flare kernel contours matches well the CaiiH flare to the LCT signal. In addition,the lower spatialresolution
emissiondepictedinFig.1ofWangetal.(2011). ofHMIascomparedtoHinodedatamightalsoincreasethe
Allflarekernelsarelocatedattheborderofumbralre- difficultytoextractthesephotosphericsignatures.However,
gionsonbothsidesofthePIL.Thefirsttwoflarekernelsap- acomparisonwithhigh-spatialresolutionHinodedataisbe-
pearatthesametimeinthecentralsunspotneartwoumbral yondthescopeofthisresearchnote.
coresof negativepolarity,which are separatedby a strong Photosphericobservationscan onlyprovidelimited in-
light-bridge.The flare kernelto the east is verystable and formation about the magnetic field evolution above an ac-
www.an-journal.org (cid:13)c2012WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim
6 Beauregardetal.:HorizontalflowsconcurrentwithanX2.2flare
tive region,since therestructuringofthecoronalmagnetic shipsinScienceandEngineeringprogram.CDwassupportedby
field is only weakly linked to changes of the photospheric grantDE787/3-1oftheGermanScienceFoundation(DFG).The
magnetic field. Nowadays, EUV images of SDO’s Atmo- authorswouldliketothankDrs.HorstBalthasarandHaiminWang
sphericImagingAssembly(AIA)andthetwospacecraftsof for carefully reading the manuscript and providing ideas, which
significantlyenhancedthepaper.
theSolarTErrestrialRElationsObservatory(STEREO)as
wellascoronagraphobservationsofSTEREOandtheSolar
and Heliospheric Observatory (SoHO) provide a compre- References
hensivethree-dimensionalpicture oferuptiveeventsin the
corona.Basedondataofthesespacecraftsandinstruments Battaglia,M.,Kontar,E.P.:2011,ArXive-prints,1107.3808
Schrijveretal.(2011)presentadetailedaccountofcoronal Chae,J.,Sakurai,T.:2008,ApJ689,593
Couvidat,S.,Schou,J.,Shine,R.A.,etal.:2011,Sol.Phys.,DOI:
activityanddynamicsrelatedtotheX2.2flareinactivere-
10.1007/s11207-011-9723-8
gion NOAA 11158,which include expanding loops above
Deng,N.,Xu,Y.,Yang,G.,etal.:2006,ApJ644,1278
thePILbeforetheonsetoftheflare,anEITwave(alsoseen
Denker, C., Johannesson, A., Marquette, W., et al.: 1999, Sol.
asaMoretonwaveinHα),andanearth-directedhaloCME
Phys.184,87
with aspeedof900kms−1. Theauthorsinterpretetheob-
Denker, C., Deng, N., Tritschler, A., Yurchyshyn, V.: 2007, Sol.
servations supported by MHD modeling as an expanding Phys.245,219
volume,wherea warmingcompressionfrontmovesin ad- Gosain, S., Venkatakrishnan, P., Tiwari, S.K.: 2009, ApJ 706,
vanceoftheeruptingfluxrope. L240
Interestingly, no Hα filament was observed as often Harvey,K.L.,Harvey,J.W.:1976,Sol.Phys.47,233
found in the core of an erupting flux rope. Even though Hudson,H.S.:1972,Sol.Phys.24,414
Kubo, M., Yokoyama, T., Katsukawa, Y., et al.: 2007, PASJ 59,
filament channelscorrespondto a state of maximummag-
779
neticshear(Martin1998),gradualchangesandslowlymov-
Krucker,S.,Hudson,H.S.,Jeffrey,N.L.S.,etal.:2011,ApJ739,
ingoppositemagneticelementsareaprerequisiteforasta-
96
blefilament.Incontrast,small-scale,rapidlychangingmag-
Liu,C.,Deng,N.,Liu,Y.,etal.:2005,ApJ622,722
netic fields are notfavorable for filamentformation.Thus,
Liu,R.,Liu,C.,Wang,S.,etal.:2010,ApJL725,L84
the strong shear motion of two opposite polarity sunspot Martin,S.F.:1998,Sol.Phys.182,107
groupsslidingpasteachothermighthavepreventedthefor- Mart´ınezOliveros,J.C.,Couvidat,S.,Schou,J.,etal.:2011,Sol.
mationofafilament. Phys.269,269
Onlydiminutivewhite-lightflarekernelswereobserved Maurya, R.A., Reddy, V., Ambastha, A.: 2011, ArXiv e-prints,
fortheX2.2flareinactiveregionNOAA11158(cf.,Fig.3 1106.4166
in Xu et al. 2004). The flare kernels to the west on both Moore,R.L.,Sterling,A.C.,Hudson,H.S.,Lemen,J.R.:2001,ApJ
552,833
sidesofthePILwerecloselyrelatedtothehardX-rayemis-
November,L.J.,Simon,G.W.:1988,ApJ333,427
sion. Such a close spatial relationship was also found by
Potts,H.E.,Barrett,R.K.,Diver,D.A.:2004,A&A424,253
Kruckeretal.(2011)betweenhardX-rayandG-bandflare
Potts,H.,Hudson,H.,Fletcher,L.,Diver,D.:2010,ApJ722,1514
kernels.ConsideringthespatialresolutionofHMIdataand
Schou, J., Borrero, J.M., Norton, A.A., et al.: 2010, Sol. Phys.,
the image processing techniques used to extract the flare DOI:10.1007/s11207-010-9639-8
kernels with sizes of 2–5 Mm, our results are compatible Schrijver, C.J., Aulanier, G., Title, A.M., et al.: 2011, ApJ 738,
withtheirhigh-resolutionG-bandobservations,whichshow 167
fine structures in flares even below 1 Mm. The altitude of Schuck,P.W.:2006,ApJ646,1358
the flare continuum emission is still debated. However, in Sobotka,M.,Bonet,J.A.,Vazquez,M.:1993,ApJ415,832
transitionregionSDO/AIA160nmfiltergramsandchromo- Sudol,J.J.,Harvey,J.W.:2005,ApJ635,647
spheric Hinode CaiiH images well-defined J-shaped rib- Tan, C., Chen, P.F., Abramenko, V., Wang, H.: 2009, ApJ 690,
1820
bonsextendmuchfurtheralongbothsidesofthePIL.Potts
Verma,M.,Denker,C.:2011,A&A529,A153
etal.(2010)usedasimplifiedradiativetransfermodeltoex-
Wachter, R., Schou, J., Rabello-Soares, M.C., et al.: 2011, Sol.
plain suchobservationsin thecontextofthe “thick-target”
Phys.,DOI:10.1007/s11207-011-9709-6
model(Hudson1972)assumingthathigh-energyelectrons
Wang,H.,Spirock,T.J.,Qiu,J.,etal.:2002,ApJ576,497
inthe10–100keVrangeareresponsibleforthewhite-light
Wang,H.,Jing,J.,Tan,C.,etal.:2008,ApJ687,658
flareemissioninanopticallythin,excitedregionwithtem- Wang,S.,Liu,C.,Liu,R.,etal.:2011,ArXive-prints,1103.0027
peraturesofabout104Ksome750–1300kmabovethepho- Welsch,B.T.,Abbett,W.P.,DeRosa,M.L.,etal.:2007,ApJ670,
tosphere. 1434
Xu,Y.,Cao,W.,Liu,C.,etal.:2004,ApJ607,L131
Acknowledgements. SDO/HMIdataareprovidedbytheJointSci-
Yang,G.,Xu,Y.,Cao,W.,etal.:2004,ApJ617,L151
enceOperationsCenter–ScienceDataProcessing.TheGOESX-
Yurchyshyn,V.,Liu,C.,Abramenko,V.,Krall,J.:2006,Sol.Phys.
rayfluxmeasurementsweremadeavailablebytheNationalGeo-
239,317
physical DataCenter.MV expressesher gratitudeforthegener-
Zirin,H.,Wang,H.:1993,Nature363,426
ousfinancialsupportbytheGermanAcademicExchangeService
(DAAD)intheformofaPhDscholarship.LB’sresearchintern-
shipinGermanywasmadepossiblebyDAAD’sResearchIntern-
(cid:13)c2012WILEY-VCHVerlagGmbH&Co.KGaA,Weinheim www.an-journal.org