Plasma flows and magnetic field interplay during the formation of a pore

2.2 MB·

by  I. Ermolli

Astronomy&Astrophysicsmanuscriptno.ibis2012˙arxiv (cid:13)c ESO2017
January24,2017
Plasma flows and magnetic field interplay
during the formation of a pore
I.Ermolli1,A.Cristaldi1,F.Giorgi1,F.Giannattasio1⋆,M.Stangalini1,P.Romano2,A.Tritschler3,andF.Zuccarello4
1 INAFIstitutoNazionalediAstrofisica–OsservatorioAstronomicodiRoma,ViaFrascati33,I-00040MontePorzioCatone,Italy
2 INAFIstitutoNazionalediAstrofisica–OsservatorioAstrofisicodiCatania,ViaS.Sofia78,I-95123Catania,Italy
3 NSONationalSolarObservatory,SacramentoPeakBox62,Sunspot,NM88349,USA
4 DipartimentodiFisicaeAstronomia–SezioneAstrofisica,Universita`diCatania,ViaS.Sofia78,I-95123Catania,Italy
7
1 Received...;accepted...
0
2 ABSTRACT

n Aims. Recentsimulationsofsolarmagneto-convectionhasofferednewlevelsofunderstandingoftheinterplaybetweenplasmamo-
a tionsandmagneticfieldsinevolvingactiveregions.Weaimatverifyingsomeaspectsoftheformationofmagneticregionsderived
J
fromrecentnumericalstudiesinobservationaldata.

3 Methods. We studied the formation of a pore in the active region (AR) NOAA 11462. We analysed data obtained with the
2 InterferometricBidimensional Spectrometer(IBIS)attheDunnSolarTelescopeonApril17, 2012, consisting offullStokesmea-
surementsoftheFeI617.3nmlines.Furthermore,weanalysedSDO/HMIobservationsinthecontinuumandvectormagnetograms

] derivedfromtheFeI617.3nmlinedatatakenfromApril15to19,2012.Weestimatedthemagneticfieldstrengthandvectorcom-
R
ponentsandtheline-of-sight(LOS)andhorizontal motionsinthephotospheric regionhostingtheporeformation. Wediscussour
S resultsinlightofotherobservationalstudiesandrecentadvancesofnumericalsimulations.
. Results. Theporeformationoccursinlessthan1hourintheleadingregionoftheAR.Weobservethattheevolutionoftheflux
h
patchintheleadingpartoftheARisfaster(¡12hour) thantheevolution(20-30 hour)of themorediffuseandsmallerscaleflux
p
patchesinthetrailingregion.Duringtheporeformation,theratiobetweenmagneticanddarkareadecreasesfrom5to2.Weobserve
-
o strongdownflowsattheformingporeboundaryanddivergingpropermotionsofplasmainthevicinityoftheevolvingfeaturethatare
r directedtowardstheformingpore.TheaveragevaluesandtrendsofthevariousquantitiesestimatedintheARareinagreementwith
t resultsofformerobservationalstudiesofsteadyporesandwiththeirmodelledcounterparts,asseeninrecentnumericalsimulations
s
a ofarising-tubeprocess.Theagreementwiththeoutcomesofthenumericalstudiesholdsforboththesignaturesofthefluxemergence
[ process(e.g.appearanceofsmall-scalemixedpolaritypatternsandelongatedgranules)andtheevolutionoftheregion.Theprocesses
drivingtheformationoftheporeareidentifiedwiththeemergenceofamagneticfluxconcentrationandthesubsequentreorganization

1 oftheemergedflux,bythecombinedeffectofvelocityandmagneticfield,inandaroundtheevolvingstructure.
v
0 Keywords.Sun:activity-Sun:photosphere-Sun:sunspots-Techniques:highangularresolution
4
4
6 1. Introduction pores(e.g.Choetal.2010,2013;VargasDom´ınguezetal.2010;
0
Toriumietal.2014),showthatthesefeaturesaresmallsunspots
. Sunspots represent the best-known manifestation of so-
1 withoutpenumbraandwith a prevailingverticalmagneticfield
lar magnetism (for a review see e.g. Solanki 2003;
0 that can reach up to 1-2 kG strength in the photosphere (e.g.
7 Rempel&Schlichenmaier 2011). The structure and dy- Sobotkaetal. 2012). The periphery of pores displays strong
1 namics of sunspots have been investigatedfor a long time, but downflowswith plasma velocitiesthat decrease with the atmo-
: their evolution after emergence in the solar atmosphere cannot
v sphericheight;supersonicflowswerealsoreportedinthechro-
bepredictedbycurrentknowledgeasyet.Inadditiontoadvance
i mosphere(e.g.Laggetal. 2007;Sobotkaetal.2012;Choetal.
X science,theabilitytopredicttheevolutionofsunspotspresently
2013; Sobotkaetal. 2013, and references therein). Converging
entailseconomicandethicalconsequences,byallowingefficient
r horizontal motions appear around pores with velocities twice
a protectionof technologicalsystems andhumanlife fromspace
as high as those found inside them and the highest values
weatherevents(Hapgood2012).
neartheborderofthepores(e.g.VargasDom´ınguezetal.2012;
Pores constitute the first stage of the evolution of sunspots Verma&Denker 2014, and references therein). Like sunspots,
from which they mainly differ in size, and in strength and pores host fine bright features and several types of waves (e.g.
orientation of the magnetic field. High-resolution observations
Balthasaretal. 2000; Bogdan&Judge 2006; Stangalinietal.
of isolated pores in the photosphere (e.g. SainzDaldaetal. 2011,2012).
2012; Sobotkaetal. 2012; Guglielmino&Zuccarello 2011;
Giordanoetal. 2008, and references therein) and chromo- High-resolutionobservationsshowthatporesareformedby
sphere (e.g. Sobotkaetal. 2013), and the study of samples of the merging of small magnetic elements dragged together by
plasma motions (e.g. Sobotka (2003); Yangetal. (2003), and
Sendoffprintrequeststo:I.Ermolli references therein). However, until recent times, limited diag-
⋆ Currentaddress:INAFIstitutoNazionalediAstrofisica-Istitutodi nostic capabilities have impeded the understandingof whether
Astrofisica e Planetologia Spaziali, Via del Fosso del Cavaliere, 100, the above process results from emergence through the photo-
00133Roma,Italy sphereofamagneticfield generatedbyglobaldynamomecha-
1
I.Ermollietal.:Velocityandmagneticfieldsinaformingpore
nismsinthesolarinterior,orfromamplificationandstructuring
ofamagneticfieldgeneratedbylocalconvectivedynamodueto
plasma motions. Current spectro-polarimetric instruments and
methods,andrecentnumericalsimulationspromisetooffernew
levelsofunderstandingof theprocessesinvolvedin theforma-
tion of pores and larger scale magnetic structures; see for ex-
ampletheobservationalstudiesbySchlichenmaieretal.(2010),
Rezaeietal.(2012),BelloGonza´lezetal.(2012),Romanoetal.
(2013, 2014), and Watanabeetal. (2014) concerning the con-
ditions that lead to the formation of penumbral regions, and
the three-dimensional (3D) radiative magneto-hydro-dynamic
(MHD) simulations of flux emergence by Rempel&Cheung
(2014).
Inthisframework,weherebystudytheevolutionofamag-
neticregionthatledadiffusedipolarfluxpatchtoformapore,
and, at a later stage, dipolar sunspot groups in AR 11462. We
analyseauniqueground-basedspectro-polarimetricdatasetac-
quiredduringtheporeformation,andcomplementtheseobser-
vationswithspace-bornedatafromtheSDOmission.Wefocus
on the photosphericprocesses related to the pore formationby
investigating the physical conditions in the whole evolving re-
gion as observed on timescales longer than about 10 minutes.
ThedataandmethodsemployedarepresentedinSect.2,while
theresultsobtainedaredescribedinSect.3.Wediscussourfind-
ings in light of other observationalstudies and the minute out-
comes of recentsimulations, and then draw our conclusionsin
Sect.4.
2. Dataandmethods
2.1.Observations
The region analysed in our study was observed with the
Interferometric Bidimensional Spectrometer (IBIS; Cavallini
2006) at the Dunn Solar Telescope of the National Solar
Observatory(NSO/DST) on April 17, 2012, from 13:58UT to
20:43 UT, within the AR NOAA 11462 (hereafter referred to
as AR), at initial disk position [S24, E0.3]. The observations
were assisted by the adaptive optics system of the NSO/DST
(Rimmeleetal.2004),underfairconditionsofatmosphericsee-
ing. No data were acquired between 16:30 UT and 18:30 UT
becauseofaworseningoftheseeingstartedatabout16:15UT.
The IBIS data consist of 223 sequences, each contain-
ingnarrowbandfiltergramsderivedfroma24-,30-,and25-
point scan of the Fe I 617.3 nm, Fe I 630.2 nm, and Ca II
854.2 nm lines, respectively, over a field of view (FOV) of
≈ 40×90 arcsec2 with a cadence of 67 s. The Fe I data in-
cludesequentialmeasurementsofthesixpolarizationstates
[I+Q, I+V, I-Q, I-V, I-U, I+U] at each wavelength position
ofthelinesampling(FWHM2pm,step2pm),whiletheCa
II data (FWHM 4.4 pm, step 4.4 pm) include only Stokes I
measurements. Each measurement consists of a single filter-
gram taken with an integration time of 60 ms and pixel scale
of≈0.09arcsec.Theabovedataarecomplementedwithsimul-
taneousbroadbandfiltergramsobtainedat 633.32± 5 nm with
thesameexposuretimeandFOVofthenarrowbanddataforthe
post-factoimagerestoration.
In this study we focus on the available Fe I 617.3 nm line
data;accordingtoNortonetal. (2006),theexcitationpotential,
effectiveLande´ factor,andaverageline-formationheightofthe
Fe I 617.3 nm line are 2.22 eV, 2.5, and 250-350 km, respec-
Fig.1.AR11462asseenintheSDO/HMIcontinuumfiltergrams(left)
tively.
andLOSmagnetograms(right)fromApril16,2012,12:00UT,toApril
The AR evolution was also studied by analysing the
19,2012,10:48UT.TheboxintheApril17,2012,14:00UTdatashow
data obtained with the Helioseismic and Magnetic Imager
theFOVofIBISdata.MoredetailsinSect.3.1.Themagneticfieldin
thebackgroundmagnetogramisshownintherangeofvalues[-1.5,1.5]
kG.Thefulltemporalevolutionoftheanalyseddataisshowninamovie
2 availableonline.
I.Ermollietal.:Velocityandmagneticfieldsinaformingpore
(HMI;Scherreretal. 2012;Schouetal.2012)aboardtheSolar opposite. At each iteration, the synthetic profiles derived
Dynamics Observatory (SDO; Pesnelletal. 2012). In particu- from the solution of the radiative transfer equation were
lar,weanalysedtheLevel1.5SDO/HMIphotosphericfull-disk also convolved with the spectral response function of IBIS
filtergrams and magnetogramsat the Fe I 617.3 nm line taken (Reardon&Cavallini 2008) and weighted by considering
from April 15 to April 19, 2012, when the AR longitudinal thestray-lightcontaminationonthedata.Weestimatedthe
distance was within ± 30◦ of the central meridian. The data latter quantity by averaging Stokes I spectra in a region
consist of ≈ 500 images, each 4096 × 4096 pixels, with pixel characterized by low polarization degree over the inverted
size of 0.505 arcsec and cadence of 720 s. Additional infor- sub-array,asinforexampleBellotRubioetal.(2000).
mation about the AR was derived from the SDO/HMI Space- We tested the accuracy of the results obtained for different
weatherActiveRegionPatches(SHARP;Hoeksemaetal.2014; initializationsof the inversionsand chose the initialization that
Bobraetal.2014) mapsobtainedduringthesametimeinterval producedthe best fit between the synthesized profile and mea-
oftheotherSDOdata. surementovereachpixelandthelargestphysicalconsistencyof
theestimatedquantitiesoverthewholeFOVinverted.Wefound
that increasing the number of inversion cycles did not further
2.2.Methods
improve the result of our calculation; on average less than 12-
The IBIS observations were processed with the standard re- 15iterationsallowedthecomputationalconvergence.Examples
duction pipeline1 to compensate data for the dark and flat- of the results obtained and comparisons between the observed
field response of the CCD devices, instrumental blueshift, and invertedprofilesat20 positionson the analysedsub-FOVs
and instrument- and telescope-induced polarizations. Besides, are given in Figs. B.1-B.8, which are available online. Values
they were also restored for seeing-induced degradations, us- ofthestray-lightfraction,LOSmagneticfieldstrength,fieldin-
ing the Multi-Frame Blind Deconvolution technique (MFBD; clinationandazimuth,andLOSvelocityderivedfromourdata
vanNoortetal.2005,andreferencestherein). analysisatthesamepositionsofthecomparedprofilesarelisted
To get quantitative estimates of the physical parameters in inTablesB.1-B.7,whichareavailableonline.
the observed region, we performed spectro-polarimetric inver- We then transformedthe magnetic field inclination and az-
sions of a subset of the IBIS measurementswith the SIR code imuthderivedfromthedatainversionsintothelocalsolarframe
(RuizCobo&delToroIniesta 1992; BellotRubio 2003). We (LSF).Weresolvedthe180degreeambiguityoftheazimuthan-
selected20timeintervalsduringtheIBISobservationsthattrack gleviatheNPFCcode(Georgoulis2005).
theevolutionoftheregionundergoodandstableseeingcondi- InordertodescribethehorizontalpropermotionsintheIBIS
tions.Weprocessed,withtheSIRcode,asub-arrayof360×350 FOVandestimatetheirvelocity,vH,we appliedtheFourierlo-
pixelsextractedfromtheselecteddataandcentredontheevolv- calcorrelationtrackingmethod(FLCT; Fisher&Welsch 2008,
ingfeature. andreferencestherein)totheavailableline-continuumdata.We
Following common approaches (see e.g. set the FWHM of the Gaussian tracking window to 0.5 arcsec
AsensioRamosetal. (2012); Requereyetal. (2015); toproperlytrackmagneticstructureswithspatialscalessmaller
Buehleretal.(2016),andreferencestherein),weperformedthe thanthetypicalgranularsize;wemadethetemporalintegration
data inversionby consideringonecomponentplusa stray-light overa13minutetimeinterval.Finally,wecomputedtheplasma
component of unspecified amplitude. This latter component LOSvelocity,vLOS,bytheDopplershiftsoflinecoreswithre-
thus acts as a free parameter. Depending on the amount of spect to the average quiet Sun line centre position in the IBIS
polarizationineachpixel,weconsideredthefirstcomponentto FOV. We computed the reference value for each filtergram of
bemagneticorquiet.Wedefinedthemagnetizedpixelsasthose theanalysedseries.Theserieswerepreviouslyprocessedwitha
in which the total circular polarization signal is ≥ 2 times the subsonicfilteringwithaphase−velocitycut-offsetto5km/s.
standard deviation of the signal estimated over the sub-array. We processed the time series of the SDO/HMI data ac-
Examplesofthe identifiedregionsaregivenin Fig.B.1,which cording to Ermollietal. (2014), by extracting from each im-
isavailableonline.WeconsideredtheHarvard−Smithsonian agethe512×512pixel2sub-arraycentredontheARbaricentre.
Reference Atmosphere (HSRA; Gingerichetal. 1971) as Besides, we produced photospheric velocity maps of the hori-
an initial guess model for the quiet regions and the same zontal plasma motions via the differential affine velocity esti-
model, but modified with an initial value for the magnetic mator method for vector magnetograms (DAVE4VM; Schuck
field strength of 0.2 kG, for the magnetized regions. We 2006) on the SDO/HMI SHARP data. In particular, following
performed the data inversion by applying two computa- Schuck (2008), we derived persistent plasma motions by com-
tional cycles, each one including up to 30 iterations. At paring magnetograms taken 24 minute apart with a 5.5 arcsec
the first cycle, we considered all modelled quantities to be FWHMapodizationwindow.
constantalongtheLOSandassignedthemonenode.Atthe
second cycle, we added one node in the temperature. The
3. Results
magnetic field strength, inclination, azimuth, LOS velocity,
and micro-turbulent velocity, however, were considered to
3.1.ARevolutionandporeformation
be constant with height, i.e. the temperature was assigned
two nodes, and the other quantities were given one node. The AR was observedon the solar disk fromApril16 to April
Thetwonodesweresetatlog(τ)=1.4andlog(τ)=-4.Since 22,2012,whenitreachedthewesternlimb.Figure1showsthe
we performed one-component inversions, the magnetic AR as seen in the SDO/HMI observationstaken at giventimes
filling factor is equal to unity; we set the macroturbulent fromApril16,2012,12:00UT,toApril19,2012,10:48UT.In
velocity to 0.75 km/s. For quiet Sun region pixels, we gave eachpanel,weshowthesubfieldof≈160×170arcsec2,centred
the I measurements four times the statistical weight of the on the AR, that was analysed to describe the evolution of the
Q, U, V profiles; for magnetized region pixels, we did the magneticfieldandradiativefluxintheAR.Wealsoshow(Fig.
1column2)thetwoparts(sFOV)oftheabovesubfieldthatwere
1 http://nsosp.nso.edu/dst−pipelines consideredtodescribetheevolutionofthetrailing(negative)and
3
I.Ermollietal.:Velocityandmagneticfieldsinaformingpore
Fig.2. Evolution of the magnetic flux (top, middle) and flux deriva-
tive(bottom)inAR11462fromtheSDO/HMILOSmagnetogramsac-
quiredfromApril16,2012,00:00UT,toApril18,2012,12:00UT,by
analysingthesubfieldshowninFigs.1(solidline),andtheleadingand
trailingpartsofthesamesubfieldseparately(dottedanddashedlines,
respectively).Theblack,red,andbluelinesindicatethetotal,positive,
and negative magnetic flux, respectively. The vertical lines show the
timeintervaloftheIBISobservations(solid),thetimesthedatashown
inFig.3weretaken(dashed),andthetimesof00:00UTfromApril16
toApril18,2012(dotted).Theadditionalaxisindicatescalendardays
at 00:00 UT. For clarity, flux values are only shown from data taken
withacadenceof36minute.
leading(positive)polarityregionsintheAR,sFOV andsFOV,
t l
respectively.Thefulltemporalevolutionoftheanalysedsubfield
isshowninthemovieattachedtoFig.1.
According to the NOAA/USAF active region summary, in
the early stages of its evolution, the AR consists of seven tiny
field features organized to form a diffuse dipolar flux region,
which is seen for example in Fig. 1 (lines a-b). The outermost
availableobservationsoftheARconsistsofseveralsunspotsand
pores,whicharealreadyseenintheobservationstakenonApril
19,2012andshown,forexampleinFig.1(lineh).
Figure 2 shows the evolution of the magnetic flux (top and
Fig.3. ExamplesoftheIBISFeI617.3nmlinedataanalysedinour
middle panels) and flux derivative (bottom panel) in the AR,
study;letters(numbers)betweenbracketsindicatetheline(column)la-
from the SDO/HMI LOS magnetograms taken from April 16 bel.(a)StokesIinthecontinuumneartheline,(b)StokesIintheline
00:00UTtoApril1812:00UT,2012,overthesubfieldandthe core,(c)StokesVinthebluewing,nearthelinecore,atfourstagesof
two sFOVs shownin Fig. 1 and in the attachedmovie.The to- poreevolution,fromdatatakenat(1)13:58UT,(2)16:09UT,(3)19:45
talunsignedmagneticflux(B )emergedintheAR(Fig.2,top UT,and(4)20:35UT.North(west)isatthetop(right).Theblackbox
tot
panel) remains almost constant with values below 4×1021 Mx in the top panels shows the subfield inverted with the SIR code and
tillApril17,2012,≈12:00UT,thenitundergoesanincrease. showninFigs.5and6.Thecontoursoverplottedineachpanelindicate
thelocationoftheevolvingstructure,assingledoutbyapplyinganin-
Thefluxevolutionpointsoutanincreasedfluxofcompara-
tensitythresholdcriterion,I ¡0.9I ,whereI istheaveragequietSun
blemagnitudeinboththesFOV andsFOV onApril17,2012, c qs qs
l t intensity.MoredetailsareprovidedinSect.3.1.
between 9:30 UT and 12:30 UT, before the start of the IBIS
measurements(Fig.2,middleandbottompanels).Thisfluxin-
crease,whichisconsistentwiththefluxincreasebyarising-tube
process,precedestheformationofafilamentary,weakS-shaped right end of the initial S-shaped structure breaks away at
structureandoccurredinsFOV between≈12:30UTand13:30 roughly one-third of the length of the structure. After the
l
UT.TheIBISmeasurementstargetedtheevolutionthatleadsthe break,therightendofthestructureevolvestoformthepore,
abovefilamentary structure, of positive polarity flux, to forma which at some first stages resembles a tiny U-shaped (e.g.
pore, on April 17, 2012 from 13:58 UT to 20:43 UT. Figure 1 SDO/HMIobservationsat15:36UT),thenatrident-shaped
(line c) shows the FOV of the IBIS data, whose examples are structure (e.g. 16:09 UT, IBIS data in Fig. 3, column 2). In
giveninFig.3. the early stages of its evolution, the S-shaped structure is
ThetimeintervaloftheIBISmeasurementsischaracter- aligned to about 10 degree to the east-west direction (Fig.
izedbyasteepincreaseoffluxintheARlastinguntilApril 3, column 1). The IBIS observations show elongated gran-
17,2012,≈17:00UT(Fig.2,middleandbottompanels)and ules near the evolving structure, mostly located east of this
small flux changes afterwards.Available observationsshow structureinagreementwith,forinstanceCenteno(2012)and
that onApril 17,2012between15:15UT and15:30UT the Vermaetal.(2016);these granulesareorientedalmostper-
4
I.Ermollietal.:Velocityandmagneticfieldsinaformingpore
pendicularly to the axis of the evolving feature at the early
stagesofitsevolutionandparalleltoitlateron(seee.g.the
IBIS observationsinFig.3linea,granulesindicatedbythe
redbarineachpanel).
The formation of the positive polarity pore is accompa-
niedbytheemergenceofsmall-scalemagneticfeaturesofop-
positepolarity.Indeed,sincetheearlystagesoftheporefor-
mation,i.e.between10:00UTand12:00UTandmoreexten-
sivelyafter14:00UT,theSDO/HMIobservationsshowatiny
magneticpatchofnegativepolarityfluxthatemergesnorth-
eastoftheevolving,positivefluxstructure,atabout100de-
greetotheeast-westdirection(seee.g.themovieattachedto
Fig. 1). This diffuse negative polarity patch mostly lies out-
side the FOV of the IBIS observations. It nearly preserves
the same size during the pore formation, but similar sized
negativeflux featuresappearsouthofit (e.g.theSDO/HMI
observations at 16:12 UT in the movie attached to Fig. 1).
Thenegativepolarityfeaturesshowclearlinkstotheevolv-
ing, positive polarity region. This suggests that they belong
tothesameemergingfluxloopfromwhichtheARevolution
could be ensued. The negative patches are clearly seen on
theavailabledataatthetimetheporehasalreadyincreased
significantly in size. Near the evolving positive flux struc-
ture, the SDO/HMI magnetograms also show small-scale,
mixed polarity features that counterstream; as seen in the
movie attached to Fig. 1, the negative flux moves towards
thesFOV,whilethepositivepolarityfluxmovestowardsthe
t
sFOV.Thisarrangementofoppositepolarityfluxresembles
l
the footpoints of an emerging loop. The SDO/HMI magne-
tograms show the growth of flux regions by coalescence of
smallerscale,samepolarityfeatures.Thisprocessalsoshows
up as short-lived, small-scale light features observed in the
photospheric available observations, in both the SDO/HMI
andIBISdata,neartheeastwardside(leftmost)oftheevolv-
ingstructure.
BoththeSDO/HMIandIBISobservationsrevealacounter-
clockwise rotation of the growing pore, which is clearly seen
as a swirling motion of the plasma immediately outside the
western (rightmost) border of the evolving structure. Figure 4
shows some maps of the horizontal motions derived from the
SDO/HMI data; the full evolution of all the computedmaps is
showninthemovieattachedtoFig.4.Inthesemaps,thearrows
wereplottedtoshowpersistentmotions(¿12min)atmoderate
resolution (¿ 2 arcsec). Figure 4 and the attached movie show
theabove-mentionedrotationonlargersizeregionsofbothpo-
larities in the AR with velocities up to about 1 km/s (see e.g.
themapsat16:00UTand19:12UT).Inaddition,theyalsodis-
playanoutwardsmotionoftheformedpore,withrespecttothe
primary flux patch, and a drift of the negative polarity flux in
theoppositedirection(seee.g.themapsat19:12UTand20:48
UT).As seenonthe moviesavailableonline,all togetherthese
plasmamotionsseemtofostertheaccumulationofflux.Figure
4alsoshowsthat,aftertheporeformation,theaveragevelocity
ofthehorizontalplasmamotionsattheevolvingregionreduces
toabout0.3-0.4km/s(Fig.4,bottompanel).
ThroughoutthedurationoftheIBISobservations,thecoher-
entstructureresultingfromtheevolutionoftheS-shapedfeature
displaysafragmented,changingcoreintheIBIS617.3nmdata
Fig.4.Examplesofthephotospherichorizontalvelocitymapsderived
and lack of penumbra around its entire perimeter. The formed
fromtheSDO/HMISHARPdatatakenfromApr.16,2012,12:48UT
poreshows a rathersymmetricfunnel-shapedstructure(Fig. 3,
toApr.18,2012,11:24UT,whichisfrom24hourbeforethestartofthe
poreformationtoseveralhourafter.Thearrowsindicatethehorizontal column4)thatlastsabout9hours.Duringthesametimeinter-
velocity;red(lightblue)showstheleading(trailing),positive(negative) val,thefieldaggregationinthetrailing(negative)polarityregion
polarity.Inallpanelsbutthefirst,thestudiedporecoincideswiththe formsinasmallerscale(¡10arcsec)andmorediffusefeatures.
largest size structure in the leading region. The magnetic field in the While the pore is formed in the leading region in less than 1
backgroundmagnetogramisshownintherangeofvaluesspecifiedin
the colour bars. The full temporal evolution of all computed maps is
showninamovieavailableonline. 5
I.Ermollietal.:Velocityandmagneticfieldsinaformingpore
hour,themorediffuseandsmallerscalemagneticstructureson merfeaturesareseentocoalesceandformthepore,theyshow
the trailing regionshow a more gradualincrease in flux overa upastheleadingfootpointsofanemergingdipole.
timeintervalof20-30hours.
On April 18-22, 2012, the total magnetic flux estimated in
3.2.2. Line-of-sightandhorizontalmotions
theevolvingregionslightlyincreasesbyremainingalmostcon-
stantforabout24hour,andthenitincreasesuptovaluesclose Figure6showsexamplesofthev andv mapsderivedfrom
LOS H
to 9×1021 Mx (notshown in Fig. 2). This second flux increase theIBISdataatthesamestagesoftheporeformationpresented
followsthe formationofa partialpenumbraaroundthe formed inFig.5.
pore (Fig. 1, line g). This flux leads, in about 10 hours, to the
The v maps(Fig. 6, column1) show downflows(corre-
LOS
finalconfigurationoftheregion,whichconsistsofaspotinthe
spondingto positivevaluesin the velocitymaps) characterized
leadingpolarityregionoftheAR,twosmallerspotsinthesame by velocitiesup to 2 km/s in the area of the evolving structure
area,andseveralporesinthetrailingfluxregionoftheAR(Fig.
anditsperiphery.Thesedownflowsoccurduringthewholepore
1,lineh).Theleadingspotshowsapartialpenumbraontheside
formation;asthefeatureevolves,theybecomestronger,whichis
oppositefromthefollowingpolarityoftheAR,asreported,for likelyasaconsequenceofthefluxcoalescence.Theyaremostly
examplebySchlichenmaieretal.(2010).Thesteepestchangeof
locatedatthenorthernedgeofthefeatureduringtheinitialand
thetotalfluxduringtheanalysedtimeintervaloccursduringthe
intermediate evolutionary stages (Fig. 6, lines a-d), and at its
formation of coherentmagnetic structures eastward of the ma- eastern and southern sides duringthe final stages (Fig. 6, lines
tureleadingspotintheleadingpolarityregion.
e-g).Upflowsaresuppressedintheevolvingregionduringmost
of its evolution (Fig. 6, lines a-d), while they appear after the
poreformationmostlyatitswesternregion(Fig.6,lineg).The
3.2.Magneticandvelocityfieldsduringtheporeformation
mapsalso showconvectiveupflowsaroundtheevolvingregion
andlocalizeddownflows,wheremagneticfluxaccumulates(e.g.
3.2.1. Magneticfield
Fig.6,linesc,d,g,totheleftsideofthemap).Theseplasmamo-
Figure5showsexamplesoftheresultsderivedfromtheSIRin- tionsare also characterizedby velocitiesup to about1-2 km/s.
versionoftheIBISdata.Inparticular,weshowthemapsofthe It is worth noting that the positive (negative) values of plasma
magneticfieldstrength(B),andofthefieldinclination(θ),trans- velocityshownwithred(blue)coloursinthevLOS mapsdonot
verse(B)andlongitudinal(B)fieldcomponentsintheLSF,at correspondtopuredownflows(upflows)sincetheywerederived
t l
sevenrepresentativestagesoftheporeformation. fromanalysesofobservationstakenoffdisccentre.
The v maps (Fig. 6, column 2) show plasma motions
The B maps (Fig. 5, column 2) indicate that the field ex- H
in agreement with those inferred from the analysis of the
tends beyondthe visible outline of the evolving feature during
SDO/HMI observations (Fig. 4). However, with respect to the
thewholeporeformation;thefieldreachesvaluesupto1-2kG
latter maps, those in Fig. 6 more clearly represent the diverg-
during the entire interval analysed and slightly increases over
ing motions of plasma seen from the visual inspection of the
time.Attheinitialstages,themostintensefieldconcentrations
availabledata in theclose proximityof theevolvingregion.At
arefoundattheedgesoftheevolvingregion;thesestrongfields
theinitialstagesoftheporeformation,therearehorizontalmo-
are located near magnetic fields with lower strength and oppo-
tions at both sides of the elongated evolving structure (Fig. 6,
sitepolaritythanthosefoundintheevolvingfeature(Fig.5lines
lines a-b), inwards to its upper (northern) edge and outwards
b,c),fromwhichtheyareseentomoveaway(Fig.5linese-g).
beyond its lower (southern) boundary; these motions push the
At the initial stages, the magnetic area is 4-5 times larger than
evolvingfeatureforward.Atthesestages,thev mapsobtained
the photometric area, as reported from analysis of an evolving H
fromtheIBISdataalsoclearlyshowtheswirlingmotionofthe
small AR, for example by Vermaetal. (2016), while after the
plasmaintheareaoftheevolvingfeature(seee.g.Fig.6linesa,
formationofthecoherentfeature,themagneticregionisabout2
d). After the pore formation, the maps show coherent motions
timeslargerthanthedarkstructure.Atthelatterstages,themost
inwards directed around most of the eastward visible outline
intense fields are detected in the central-southernregion of the
of the pore (Fig. 6, line g). After the formation of the funnel-
pore.Thefieldisalmostverticalwithrespecttothephotosphere
shapedstructure,thehighestvaluesinthev mapsarefoundto
inthecentralsectionoftheevolvingfeature,andratherinclined H
belocatedfarfromthevisibleoutlineoftheevolvingstructure,
outsideit(e.g. B and B mapsinFig.5,column4).Indeed,the
l t
θ maps (Fig. 5, column 3) show a large patch characterizedby at a distance of about 5-10 arcsec, as reported for example by
valuesof ≈ 45◦ overan areaextendingwellbeyondthatof the VargasDom´ınguezetal.(2010).
evolvingfeature,whichhostsfieldswithθ¡≈25◦.Theθmaps
also show small patches of magnetic field characterized by an
3.2.3. Temporalevolution
inclinationof about90-180◦, which correspondto the opposite
polarityfeaturesobservedoutsidetheevolvingpore,north-east Figure7showstheevolutionofthephysicalquantitiesdiscussed
ofit. above,asderivedfromtheSIRinversionandothermethodsap-
After the formation of the coherent, funnel-shapedfeature, plied to the IBIS data. Each panel represents the average and
the B maps(Fig. 5, column4, lines e-g) show magneticfields standarddeviationofthevariousquantitiesestimatedinsideand
t
that are aligned along the direction of the opposite polarity aroundtheevolvingfeature.
patchesintheevolvingregionfacingeachother.Comparisonbe- Insidetheevolvingfeature,Bvariesfromabout1kGto2kG,
tweenthemapsderivedfromthedatatakenattheinitial(Fig.5, byreachingaveragevalueslargerthan1.5kGaftertheformation
linesa-d)andfinal(Fig.5,linese-g)evolutionarystagesshows ofthefunnel-shapedpore(Fig.7,toppanel).Themaximum(not
the same azimuth pattern on both the fragmented positive po- shown in Fig. 7) and average values of B increase in time, as
larity patches of the forming pore and the later formed pore. well asthe same quantitiesforboththe B and B components.
t l
Therefore,thefieldsinthesmallerscale,thesamepolarityfea- Outsidetheformingpore,theaveragevalueofBrangesbetween
tures,andthoseintheformedporearecoaligned.Sincethefor- 0kGand0.1kG,butthereareregionsinthisareawithmagnetic
6
I.Ermollietal.:Velocityandmagneticfieldsinaformingpore
Fig.5.Fromlefttoright,toptobottom,numbersandletterswithinbracketsindicatethecolumnandlinelabels,respectively.Examplesof(1)
continuumintensity,(2)magneticfieldstrength,(3)inclination,(4)B longitudinal(backgroundimage),andB transverse(overplottedvectorfield)
l t
componentsofthemagneticfieldintheLSFderivedfromtheSIRinversionoftheIBISFeI617.3nmlinedata,atopticaldepthlogτ =1,at
500
sevenstagesoftheporeformation,at(a)13:58UT,(b)14:26UT,(c)15:08UT,(d)16:08UT,(e)18:44UT,(f)19:29UT,and(g)20:35UT.North
isatthetop,andwestistotheright.Thearrowatthebottomleftonpanel4arepresentsahorizontalfieldof1kG;transversefieldcomponents
lowerthan0.2kGarenotshown.Thecontourlineineachpanelshowsthelocationoftheevolvingstructuresingledoutinthecontinuumdata,as
specifiedinFig.3.
7
I.Ermollietal.:Velocityandmagneticfieldsinaformingpore
Fig.7. Variation of the magnetic field strength (B strength), trans-
verse(B tran),andlongitudinal (B long) components intheLSF(top
panel), field inclination, and azimuth in the LSF (middle panel), and
LOSplasmavelocityv (bottompanel)derivedfromtheIBISphoto-
LOS
sphericFeI617.3nmlinedataduringtheporeformation,inside(black
lines)andoutsidetheevolvingfeature(bluelines).Eachplotshowsthe
meanandstandarddeviationofthevaluesderivedfromthedatainver-
sionandothermethodsdescribedinSect.2.Field,angle,andvelocity
values are given in kG, degree, and km/s units, respectively. For the
sakeofclarity,thestandarddeviationisshownonlyformagneticfield
strength,inclination,andv estimates,insideandaroundtheevolving
LOS
structure.TheverticaldottedlinesshowthetimeintervallackingIBIS
observations.Thenumbersintheupperpanelindicatetheseventimes
correspondingtotheevolutionarystagesshowninFigs.5,6.
field strength ≥ 1 kG. The average value of B over the whole
area does not change significantly during the pore formation;
the same holds for both the mean value of B over the stronger
fieldelementsandthe maximumvalueof B inthe wholearea,
t
whilethemaximumvaluesofB andBoutsidetheevolvingpore
l
slightlydecrease(≤10%)overtime.
During the interval of the IBIS observations, θ does not
changesignificantly(Fig.7,middlepanel).Theaverageandthe
rangeofv valuesmeasuredoverthemagneticregionslightly
LOS
decreaseafter theporeformation,comparedto thoseestimated
in previous stages (Fig. 7, bottom panel); outside the evolving
pore,thevelocityoftheplasmamotionsdoesnotchangesignif-
icantlyovertheanalysedperiod.
4. Discussionandconclusions
The results derived from our analysis agree with the outcomes
offormerstudiesofsteadyporesmentionedabove;specifically,
withthemagneticfieldpatternsandstrengthspresented,forex-
amplebySobotkaetal.(2012),andthev andv reportedby
LOS H
Choetal. (2010), Sobotkaetal. (2012), Sobotkaetal. (2013),
andVermaetal.(2016),forexample.However,unlikeprevious
studies, our data also allow us to investigate the properties of
the photosphericplasma in theevolvingregionduringthe pore
formation,andthusprovidefurtherobservationalconstraintsto
numericalmodelsoftheARevolution.
Magneticfields and fluid motionsare coupledaccordingto
Fig.6. Example of the vLOS (column 1) and vH (column 2) plasma the equations of the magnetohydrodynamics in 3D space and
velocityfieldsintheevolvingregionderivedfromtheIBISFeI617.3 time.Cheungetal.(2008)performedsimulationstoinvestigate
nmdataatthesevenstages(linesa-g)oftheporeformationshownin theinteractionbetweenconvectionandamagneticfluxtuberis-
Fig.5andspecifiedineachpanelofcolumn1.Theblue(negative)and
ing into the photosphere, and discussed the outcomes with re-
red(positive) v indicateupflowsand downflows, respectively. The
LOS specttoobservationsfromtheHINODEmission.Theyreported
intensity background inthe v maps shows the average image of the
H
representativeseries.Inpanel 2a,thehorizontal bluebar indicatesv
H
plasmavelocityof1km/s.
8
I.Ermollietal.:Velocityandmagneticfieldsinaformingpore
thattherisingfluxtubeexpandsbecauseofthestrongstratifica- flux accumulationat distinctsites (e.g. movie available on-
tionoftheconvectivezonebyformingamagneticsheetthatacts lineandFig.4).
as a reservoir for small-scale flux emergence events occurring – Atthattime,elongatedgranulesappearincloseproximityof
atthescale ofgranulation.Theyalso foundthattheinteraction theevolvingfeature,mostlylocatedintheregiontheoppo-
of the convective downflows and the rising magnetic flux tube sitepolaritiesoftheevolvingARfacingeachother(e.g.Fig.
undulates it to form serpentine field lines that emerge into the 3);
photosphere. – The pore increases in size, while the sites of flux accumu-
Later on, from an in-depth progression of the above simu- lationmoveaway from each otherwith clear horizontaldi-
lationstotheformationofanAR,Cheungetal.(2010)showed vergingmotions and a rather small increase of the average
thattheaboveserpentinefieldsgraduallycoalescetoformlarger fieldintheformingpore(e.g.Figs.4,5,6).
magneticconcentrationsthateventuallyformapairofopposite – Horizontal diverging motions seem to produce further ag-
polarity spots. They also pointed out that correlations between gregationoffieldofthesamepolarity(e.g.moviesavailable
themagneticfieldandvelocityfieldfluctuationsallowthespots onlineandFigs.4,6);theplasmavelocityisupto0.4km/s
to accumulate flux by Lorentz-force-driven,counter streaming intheformingporeandupto1km/soutsideit.
motionofoppositepolarityregions. – Strongdownflows,with plasma velocity¿ 1.5km/s,appear
nearthe peripheryoftheformingporeandwheremagnetic
Fromrecent3DMHDinvestigationsoftheeffectsofinflows
fluxaccumulates(e.g.Fig.6).
intheevolutionofARs,Rempel&Cheung(2014)reportedthat,
– Mostintensefieldconcentrationsoccurneartheedgesofthe
in theirsimulatedphotosphere,the flux appearsorganizedona
magneticregionsin evolution(e.g.Figs. 5, 6) as dueto the
granularscalewithmostlymixedpolaritywithmagneticfluxof
confinementofthefieldbytheambientplasmamotions.
the order 1021 Mx and average value lower than 0.1 kG. After
– Theanalyseddatashowthattheporeformationinthelead-
emergence,the simulated dipolesundergohorizontaldiverging
ingregionoftheARoccursrapidly(¡1hour);theevolution
flows, reachingan amplitudeof up to 2 km/s. In the numerical
ofthefluxpatchintheleadingpartisfaster(¡12hour)than
domain, these motions produce a progressive separation of the
the evolution (20-30 hour) of the more diffuse and smaller
polarityofthedipolesthatmigrateintheoppositedirection,by
scale flux patches in the trailing region (e.g. supplemental
movingtheoppositepolarityfluxawayfromtheemergencere-
moviesandFig.1).
gions.
– AtthefinalstagesoftheAR evolution,about48hourafter
The above simulations have contributed to a long series of
the pore formation,the evolution of the region leads to the
numericalstudiesaimedattheidentificationofthemechanisms
formation of a large-scale AR with a magnetic flux of the
responsible for the formation and evolution of solar magnetic orderof1022Mx.
structures (see e.g. Cameronetal. (2007, 2011), Cheungetal.
(2008, 2010), Fangetal. (2012b,a, 2014), Kitiashvilietal.
FromanalysisoftworelativelyisolatedARsobservedfrom
(2010), Martinez-Sykora (2012); Mart´ınez-Sykoraetal.
the SDO mission, Centeno (2012) already reported some ob-
(2015), Steinetal. (2011); Stein&Nordlund (2012),
servational signatures of AR formation consistent with the 3D
Toriumi&Yokoyama (2012, 2013), to mention those that
MHD numerical simulation of a rising-tube process, such as
have been presented during the last decade). Although these
evidence of a connection between horizontal field patches and
studiescanstronglydifferintermsoftheinitialconditionsand
strongupflows,elongatedgranulationaroundtheevolvingARs,
scales of the simulated processes, they have all reproduced
and a mass discharge process through magnetic reconnection,
someflux evolutionsignaturesinagreementwith observations.
as envisaged in the simulations of Cheungetal. (2010) and
Hence the question arises on which processes unveiled by
Rempel&Cheung (2014), for example. In contrast, from a re-
the numerical studies can be considered robust with both the
cent study of HINODE spectro-polarimetric observations of a
observationsandmodelassumptions.
youngdipolarsubregiondevelopingwithinanAR,Getlingetal.
In this regard, the data analysed in our study show several (2016) presented observational results that are considered to
observationalfactsthatare consistentwith theoutcomesofthe conflictwiththesignaturesexpectedbytheemergenceofaflux-
MHD simulations presented by Rempel&Cheung (2014). In tube loop. In particular, they reported a fountain-like 3D mag-
particular,wefoundasfollows: netic structure of the studied features and lack of large-scale
(horizontalandvertical)flowsovertheevolvingarea,whichare
– AttheinitialstagesoftheARevolution,theanalysedobser- seen as two pieces of evidence of magnetic region generation
vations display mixed polarity flux patches organized on a bylocalconvectivedynamomechanismsasenvisagedinthe3D
granular scale; the flux patches in the trailing regionof the MHDsimulations,forexamplebyStein&Nordlund(2012).
formingARaremoreclearlyseenthanthoseintheleading Comparedtothe HINODEandSDO observationsanal-
part(e.g.Fig.1andattachedmovie). ysed, for example by Getlingetal. (2016) and Centeno
– Afterabout24hours,thepatchesintheleadingpartbecome (2012), the IBIS data considered in our study have spatial
strongerthanthoseinthetrailingregionandformafilamen- and temporal resolution that are higher at least by factors
tary,sheet-like,coherentstructure(e.g.movieavailableon- 3.5and11,respectively.Byanalysingtheavailabledata,we
lineandFigs.1,2,3);thesearetheobservedinitialstagesof foundthatthesimulationsoftherising-tubeprocesssuccess-
poreformation. fully reproduce both the average properties of the physical
– Atthattime,thesmall-scalemixedpolaritypatchesholdflux quantities estimated in the studied region and the mecha-
oftheorder1021 Mxinanevolvingregionwithanaverage nismsdrivingtheobservedporeformation.Inparticular,the
fieldbelow0.1kG(e.g.Figs.1,2,5). studied pore seems to result from the emergence into the
– Thefluxthenshowsanincreaseofcomparablemagnitudein photosphere of a strong field formed in the solar interior,
bothpolaritypatchesoftheevolvingregion(e.g.Fig.2). with some amplification and structuring effects of the ini-
– Atalatertime,thedataclearlydisplaysmall-scaleopposite tially emerged field by surface plasma motions, as evinced
polarityfeaturesthatcounterstream,coalesce,andreinforce from the simulations for example by Rempel&Cheung
9
I.Ermollietal.:Velocityandmagneticfieldsinaformingpore
(2014).Theabovesimulationsalsodescribetheevolutionof Rezaei,R.,BelloGonza´lez,N.,&Schlichenmaier,R.2012,A&A,537,A19
the studied region at different spatial and temporal scales Rimmele, T. R., Richards, K., Hegwer, S., et al. 2004, in Society of Photo-
fairlywell.Thesignaturesobservedinthestudiedregion,in Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 5171,
Telescopes and Instrumentation for Solar Astrophysics, ed. S. Fineschi &
contrast,differfromthosepresentedbyGetlingetal.(2016),
M.A.Gummin,179–186
whichsupportthescenarioofporeandlargerscalemagnetic Romano,P.,Frasca,D.,Guglielmino,S.L.,etal.2013,ApJ,771,L3
regiongenerationbylocalconvectivedynamomechanisms. Romano,P.,Zuccarello,F.P.,Guglielmino,S.L.,&Zuccarello,F.2014,ApJ,
794,118
Acknowledgements. The authors wish to thank Serena Criscuoli, Han RuizCobo,B.&delToroIniesta,J.C.1992,ApJ,398,375
Uitenbroek, and the whole DST staff for its support during the observing SainzDalda,A.,VargasDom´ınguez,S.,&Tarbell,T.D.2012,ApJ,746,L13
campaign anddatareduction. Thisstudyreceived fundingfromtheEuropean Scherrer,P.H.,Schou,J.,Bush,R.I.,etal.2012,Sol.Phys.,275,207
Unions Seventh Programme for Research, Technological Development and Schlichenmaier, R.,Rezaei,R.,BelloGonza´lez,N.,&Waldmann,T.A.2010,
Demonstration, under the Grant Agreements of the eHEROES (n 284461, A&A,512,L1
www.eheroes.eu),SOLARNET(n312495,www.solarnet-east.eu),andSOLID Schou,J.,Scherrer,P.H.,Bush,R.I.,etal.2012,Sol.Phys.,275,229
(n313188,projects.pmodwrc.ch/solid/)projects.Thisworkwasalsosupported Schuck,P.W.2006,ApJ,646,1358
bytheIstituto Nazionale diAstrofisica (PRIN-INAF-2014)andItalian MIUR Schuck,P.W.2008,ApJ,683,1134
(PRIN-2012).Wethanktheanonymousrefereeforusefulcommentsandsug- Sobotka,M.2003,AstronomischeNachrichten,324,369
gestions. Sobotka,M.,DelMoro,D.,Jurcˇa´k,J.,&Berrilli,F.2012,A&A,537,A85
The National Solar Observatory is operated by the Association of Sobotka,M.,Sˇvanda,M.,Jurcˇa´k,J.,etal.2013,A&A,560,A84
UniversitiesforResearchinAstronomyunderacooperativeagreementwiththe Solanki,S.K.2003,A&ARev.,11,153
NationalScienceFoundation. Stangalini,M.,DelMoro,D.,Berrilli,F.,&Jefferies,S.M.2011,A&A,534,
A65
Stangalini,M.,Giannattasio,F.,DelMoro,D.,&Berrilli,F.2012,A&A,539,
References L4
Stein,R.F.,Lagerfja¨rd,A.,Nordlund,Å.,&Georgobiani,D.2011,Sol.Phys.,
AsensioRamos,A.,MansoSainz,R.,Mart´ınezGonza´lez,M.J.,etal.2012,ApJ, 268,271
748,83 Stein,R.F.&Nordlund,Å.2012,ApJ,753,L13
Balthasar,H.,Collados,M.,&Muglach,K.2000,AstronomischeNachrichten, Toriumi,S.,Hayashi,K.,&Yokoyama,T.2014,ApJ,794,19
321,121 Toriumi,S.&Yokoyama,T.2012,A&A,539,A22
BelloGonza´lez,N.,Kneer,F.,&Schlichenmaier,R.2012,A&A,538,A62 Toriumi,S.&Yokoyama,T.2013,A&A,553,A55
Bellot Rubio, L. R. 2003, in Astronomical Society ofthe Pacific Conference vanNoort,M.,RouppevanderVoort,L.,&Lo¨fdahl,M.G.2005,Sol.Phys.,
Series, Vol. 307, Solar Polarization, ed. J. Trujillo-Bueno & J. Sanchez 228,191
Almeida,301 VargasDom´ınguez,S.,deVicente,A.,Bonet,J.A.,&Mart´ınezPillet,V.2010,
BellotRubio,L.R.,RuizCobo,B.,&Collados,M.2000,ApJ,535,489 A&A,516,A91
Bobra,M.G.,Sun,X.,Hoeksema,J.T.,etal.2014,Sol.Phys.,289,3549 VargasDom´ınguez, S.,MacTaggart,D.,Green,L.,vanDriel-Gesztelyi, L.,&
Bogdan, T. J. & Judge, P. G. 2006, Royal Society of London Philosophical Hood,A.W.2012,Sol.Phys.,278,33
TransactionsSeriesA,364,313 Verma,M.&Denker,C.2014,A&A,563,A112
Buehler,D.,Lagg,A.,vanNoort,M.,&Solanki,S.K.2016,A&A,589,A31 Verma,M.,Denker,C.,Balthasar,H.,etal.2016,ArXive-prints
Cameron,R.,Schu¨ssler,M.,Vo¨gler,A.,&Zakharov,V.2007,A&A,474,261 Watanabe,H.,Kitai,R.,&Otsuji,K.2014,ApJ,796,77
Cameron,R.,Vo¨gler,A.,&Schu¨ssler,M.2011,A&A,533,A86 Yang,G.,Xu,Y.,Wang,H.,&Denker,C.2003,ApJ,597,1190
Cavallini,F.2006,Sol.Phys.,236,415
Centeno,R.2012,ApJ,759,72
Cheung,M.C.M.,Rempel,M.,Title,A.M.,&Schu¨ssler,M.2010,ApJ,720,
233 AppendixA: Movies
Cheung,M.C.M.,Schu¨ssler,M.,Tarbell,T.D.,&Title,A.M.2008,ApJ,687,
1373 AppendixB: Examplesdatainversionresults
Cho,K.-S.,Bong,S.-C.,Chae,J.,Kim,Y.-H.,&Park,Y.-D.2010,ApJ,723,440
Cho,K.-S.,Bong,S.-C.,Chae,J.,etal.2013,Sol.Phys.,288,23
Ermolli,I.,Giorgi,F.,Romano,P.,etal.2014,Sol.Phys.,289,2525
Fang,F.,Fan,Y.,&McIntosh,S.W.2014,ApJ,789,L19
Fang,F.,Manchester,IV,W.,Abbett,W.P.,&vanderHolst,B.2012a,ApJ,754,
15
Fang,F.,Manchester, IV,W.,Abbett, W.P.,&vanderHolst,B.2012b,ApJ,
745,37
Fisher, G. H. & Welsch, B. T. 2008, in Astronomical Society of the Pacific
ConferenceSeries,Vol.383,SubsurfaceandAtmosphericInfluencesonSolar
Activity, ed. R. Howe, R. W.Komm,K. S. Balasubramaniam, & G. J. D.
Petrie,373
Georgoulis,M.K.2005,ApJ,629,L69
Getling,A.V.,Ishikawa,R.,&Buchnev,A.A.2016,Sol.Phys.,291,371
Gingerich,O.,Noyes,R.W.,Kalkofen,W.,&Cuny,Y.1971,Sol.Phys.,18,347
Giordano,S.,Berrilli,F.,DelMoro,D.,&Penza,V.2008,A&A,489,747
Guglielmino,S.L.&Zuccarello,F.2011,ApJ,743,L9
Hapgood,M.2012,Nature,484,311
Hoeksema,J.T.,Liu,Y.,Hayashi,K.,etal.2014,Sol.Phys.,289,3483
Kitiashvili,I.N.,Kosovichev,A.G.,Wray,A.A.,&Mansour,N.N.2010,ApJ,
719,307
Lagg,A.,Woch,J.,Solanki,S.K.,&Krupp,N.2007,A&A,462,1147
Martinez-Sykora,J.2012,IAUSpecialSession,6,105
Mart´ınez-Sykora,J.,Moreno-Insertis,F.,&Cheung,M.C.M.2015,ApJ,814,
2
Norton,A.A.,Graham,J.P.,Ulrich,R.K.,etal.2006,Sol.Phys.,239,69
Pesnell,W.D.,Thompson,B.J.,&Chamberlin,P.C.2012,Sol.Phys.,275,3
Reardon,K.P.&Cavallini,F.2008,A&A,481,897
Rempel,M.&Cheung,M.C.M.2014,ApJ,785,90
Rempel,M.&Schlichenmaier,R.2011,LivingReviewsinSolarPhysics,8,3
Requerey,I.S.,DelToroIniesta,J.C.,BellotRubio,L.R.,etal.2015,ApJ,810,
79
10