Using Coronal Cells to Infer the Magnetic Field Structure and Chirality of Filament Channels
aa r X i v : . [ a s t r o - ph . S R ] J un Using Coronal Cells to Infer the Magnetic Field Structure andChirality of Filament Channels
N. R. Sheeley, Jr. , S. F. Martin , O. Panasenco , and H. P. Warren Space Science Division, Naval Research Laboratory, Washington DC 20375-5352, USA Helio Research, La Crescenta, CA 91214, USA
ABSTRACT
Coronal cells are visible at temperatures of ∼ α observations. However, the coronal cell observations are easier to use and provideclear inferences of the horizontal field direction for heights up to ∼
50 Mm intothe corona.
Subject headings:
Sun: corona — Sun: UV radiation — Sun: magnetic fields —Sun: filaments
1. INTRODUCTION
In a previous paper, Sheeley & Warren (2012) described cellular features that werevisible in 193 ˚A solar images, obtained with the Atmospheric Imaging Assembly (AIA) onthe Solar Dynamics Observatory (SDO) (Lemen et al. 2011) and in 195 ˚A images obtainedwith the Extreme Ultraviolet Imagers (EUVI) on the Solar Terrestrial Relations Observatory Now at Advanced Heliophysics, Pasadena, CA 91106 ∼
20 km s − ) relative to the average speed at the centers of the cells.The cells occurred in regions of predominantly unipolar magnetic network between coro-nal holes and filament channels. Individual cells had diameters ∼
30 Mm that were cen-tered on the flux concentrations, not between them like supergranules in the photosphere(Simon & Leighton 1964). As solar rotation carried the cellular regions across the solardisk, the individual structures were visible as cells only when they were near central merid-ian. At other locations, these features appeared as tapered plumes projecting ∼
100 Mmtoward the closest limb. Moreover, when the cells were located at a solar longitude thatwas midway between the locations of the SDO and STEREO-B (or STEREO-A) spacecraft,their plumes appeared to lean toward the west limb as seen from one spacecraft and theeast limb as seen from the other spacecraft. Thus, these simultaneous orthogonal viewswere showing sky-plane projections of structures that extend radially outward from the Sun.Sheeley & Warren (2012) concluded that the cells are discrete plumes extending upwardfrom flux concentrations in unipolar magnetic regions like candles on a cake, and are onlyvisible as bright cells separated by dark intercellular lanes when observed from above (or atleast inward along their nearly radial magnetic field lines).In this paper, we show that as one approaches the polarity reversal boundary within afilament channel, the cellular plumes bend horizontally along the channel with their taperedends pointing in opposite directions on the two sides of the boundary. Because the plumesare rooted in flux elements of majority polarity, these oppositely directed features provide aconsistent indication of the magnetic field direction along the channel, which therefore canbe determined from an Fe XII 193 ˚A image (or Fe XII 195 ˚A image) of the cells and a mapof their line-of-sight magnetic fields.This method is similar to the one that is often used to infer the horizontal field di-rection in H α filament channels from the orientation of the adjacent chromospheric fibrilsand their associated line-of-sight fields (Foukal 1971; Martin et al. 1994; Martin 1998a,b;Martin et al. 2008). However, the Fe XII observations have several advantages over the H α observations. First, the coronal plumes are much larger than the chromospheric fibrils andshow the horizontal field structure without the need for the relatively high spatial resolution 3 –required using H α observations. Second, the coronal observations show the bent plumes atmuch greater heights in the atmosphere and provide three-dimensional information whensolar rotation and views from multiple spacecraft (like STEREO-A and STEREO-B) areincluded. By comparison, chromospheric fibrils can only indicate the structure underneaththese plumes. Third, the Fe XII observations are available on a nearly continuous basis fromthe AIA instrument on SDO and the EUVI instruments on STEREO-A and STEREO-B.Panasenco et al. (2012) have recently used AIA 193 ˚A observations of coronal cells as anaid for interpreting the origin of twisting and rotating motions in prominence eruptions,and Wang et al. (2013) have used the Fe XII images to study the origin of the horizontalmagnetic field in filament channels. Of course, high-resolution chromospheric observationsin H α Hinode
X-Ray Telescope (XRT) images as wellas STEREO/EUVI images to study the structure and dynamics of filament channels priorto the launch of the SDO spacecraft. Not only did they see cellular plumes curving along thepolarity reversal boundary, but also they noticed that the plume arrangement was differentfor east-west filament channels (like the polar crown filament channels) and for north-southoriented filament channels (like the channels that stretch poleward from the sunspot belts).For a sample of six north-south channels observed on twelve solar rotations, those authorsfound that the plumes on the east side of the channel curved toward the polarity reversalboundary, but the plumes on the west side were straight and oriented nearly perpendicularto the boundary. More recently, Su & van Ballegooijen (2012) combined SDO/AIA imagesand STEREO/EUVI images with a flux-rope model to study the structure of a polar crownprominence.In the next section, we describe the magnetic structure and chirality (handedness) offilament channels by combining AIA 193 ˚A coronal images with virtually simultaneous (aver-aged over the same 5-minute interval) photospheric magnetograms, obtained in Fe I 6173 ˚Awith the Helioseismic and Magnetic Imager (HMI) on SDO. In the first three figures, we infera rotational magnetic topology from downward views of the filament channel. Then, in thenext three figures, we look for evidence of this rotational structure in nearly broadside viewsof a long lived filament channel. The last figure of this section is a time-lapse sequence afterthe eruption of a filament in this channel. In the final section of this paper, we summarizethe observations and discuss their implications. 4 –
2. OBSERVATIONS
Figure 1 compares an AIA 193 ˚A image (top) with an HMI photospheric magnetogram(bottom) obtained when a filament channel and its adjacent coronal holes were at the centralmeridian on 2011 June 17. Like most of the 193 ˚A images in this paper, this one has beenconstructed by averaging images over a 5-minute interval (to increase the signal-to-noiseratio) and then by applying a minimal amount of sharpening. The magnetic picture wasconstructed by averaging magnetograms obtained over the same 5-minute interval and thenapplying a minimal amount of smoothing (to reduce the fine-grain noise and some mixed-polarity field).In this figure, the filament channel is the elongated dark region of Fe XII intensityrunning approximately north-south along the polarity reversal boundary of the photosphericfield. The channel width is not precisely defined, but probably extends beyond the dark areato include cellular plumes that are affected by the axial magnetic field along the polarityreversal boundary, as evidenced by the curved tails of those tapered plumes. By extendingthe width to include those plumes, we obtain agreement with the definition used previouslyfor H α filament channels whose widths include the zones where chromospheric fibrils beginto bend along the polarity reversal boundary (Smith 1968; Martin 1998a; Wang & Muglach2007). We shall use this definition of Fe XII filament channel in the remainder of our paper.In Figure 1, coronal cells are visible on the positive and negative sides of the filamentchannel, extending from the polarity reversal boundary inside the channel to the distantcoronal holes on each side of the channel. In this paper, we will focus our attention on thecellular plumes that lie close to the polarity reversal boundary and show the influence of theaxial magnetic field in the filament channel.On the positive (white)-polarity side of the channel, several cells are stretched intotapered plumes whose tails bend southward along the channel. Because these plumes arerooted in magnetic flux elements of positive polarity, their tapered tails point along thedirection of the magnetic field. Thus, based on the directions of these plumes, we concludethat the field changes from quasi-radial near the positive-polarity coronal hole to horizontaland pointing southward along the polarity reversal boundary in the middle of the filamentchannel. This corresponds to a leftward rotation of the field and a sinistral chirality, astypically occurs for such southern-hemisphere filament channels (Martin 1998a,b). The senseof this rotation continues on the opposite side of the channel where some fainter cells arevisible with their tails pointing northward. Because these cells are rooted in elements ofnegative polarity, their fields are directed southward, consistent with the southward directionthat we inferred from the orientations of the plumes on the positive-polarity side of thechannel. 5 –At this point, we recall the observations of Su et al. (2010) who found that the X-Rayand XUV structures on the two sides of north-south filament channels were asymmetricwith curved features on the east side of the channel and straight features oriented nearlyperpendicular to the channel on the west side. At first glance, Figure 1 does seem to possesssuch an asymmetry, but with the curved plumes on the west (right) side and the straighterplumes on the east (left) side, opposite to what Su et al. (2010) found for their sample ofnorth-south filament channels.The asymmetry is interesting regardless of whether the curved plumes lie preferentiallyon the east side or west side of the channel. There is no doubt that the plumes on the two sidesof the filament channel are oriented so that that their tails point in opposite directions alongthe polarity reversal boundary, giving a consistent southward direction for the horizontalfield along the boundary. However, a close inspection of the plumes in Figure 1 reveals thattheir components normal to the boundary are pointed in the same direction on the two sidesof the boundary, corresponding to magnetic fields that point toward the boundary from bothsides. However, if we look closely at the faint plumes in the dark part of the channel, we cansee that these plumes become increasingly coaligned along the polarity reversal boundary,suggesting that the normal components of field vanish here. Thus, Figure 1 supports theidea that there is an asymmetry across the channel, but that the strength of this asymmetryweakens toward the polarity reversal boundary and the sign of this asymmetry is oppositeto what Su et al. (2010) found for their selection of north-south filament channels.Figure 2 shows this southern-hemisphere region 28 days later on 2011 July 15 when itwas again at central meridian. Differential rotation has caused the filament channel to bemore inclined relative to the north-south direction than it was a rotation earlier in Figure 1.Cellular plumes are still visible, oriented with their tails pointing southward on the positive-polarity side of the channel and northward on the negative-polarity side. Because theseplumes are rooted in flux concentrations of majority polarity, their tapered ends point alongthe field on the positive-polarity side of the channel and opposite the field on the negative-polarity side. Again, this indicates a southward-directed field along the channel and asinistral chirality. The normal component of field is still oppositely directed on the two sidesof the channel, but some of the negative-polarity plumes are now oriented at smaller anglesto the polarity reversal boundary than Figure 1 showed 28 days earlier.Looking closely at the images in Figure 2, we find that we can determine the location ofthe polarity reversal boundary more accurately from the Fe XII coronal image than from thephotospheric magnetogram. In attempting to follow the polarity reversal boundary throughweak-field areas in the magnetogram, we become lost in the fine-scale maze of mixed polari-ties. For this reason, it is often necessary to spatially smooth photospheric magnetograms in 6 –order to determine the location of the larger-scale polarity reversal boundaries (Wang et al.2011). However, in this Fe XII image, the cellular plumes clearly point the way. Martin(1998a,b) noticed the same effect in her comparison of photospheric magnetograms andH α images whose chromospheric fibrils provided a precise definition of the polarity reversalboundary.Figure 3 shows a northern-hemisphere region on 2011 September 8. Here, the tails ofthe negative-polarity plumes are stretched southward along the filament channel, pointingopposite to the direction of the horizontal field. This orientation corresponds to a dextralchirality, as expected for such northern-hemisphere regions. A few plumes are visible withtheir tails pointing northward on the positive-polarity side of the channel. These orientationsare consistent with the dextral chirality that we inferred from plumes on the negative-polarityside.Also, we see in Figure 3 that the tails of the plumes are oriented so that their componentsalong and normal to the channel have different properties on the two sides of the channel.Whereas they are oppositely directed along the channel (corresponding to the same horizontaldirection of the magnetic field), the tails point in the same direction normal to the channel(corresponding to fields directed away from the polarity reversal boundary on both sides ofthe channel). Recall that the normal components were directed toward the polarity reversalboundary in Figures 1 and 2. It is easy to see that this direction, toward or away from theboundary, depends on the polarity of the bent plumes. The fields point toward the boundaryif the bent plumes are rooted in positive-polarity field, and the fields point away from theboundary if the bent plumes are rooted in negative-polarity field, independent of the chiralityof the field. Also, we see in Figure 3 that on 2011 September 8, the bent plumes lay on theeast side of the channel, consistent with the result that Su et al. (2010) found.In this analysis, the magnetogram was necessary to infer the direction of the horizontalfield in the filament channel. However, note that the Fe XII image is sufficient to determinethe chirality of the filament channel without the need for a magnetogram. For example, ifwe look from south to north along the filament channel in the Fe XII image in Figure 3,we see the tapered ends of the plumes pointing toward us on the left side of the channeland away from us on the right side. If we reverse our direction and look from north tosouth, the tapered ends of the plumes still point toward us on our left side and away fromus on our right side. So ‘toward us on the left’ corresponds to a dextral chirality. For thesinistral channels in Figures 1 and 2, the plumes have the opposite handedness, pointing‘toward us on the right’ regardless of which way we look along the channel. This sense ofchirality corresponds to left-skewed arcades over dextral channels and right-skewed arcadesover sinistral channels (Martin 1998a,b). 7 –The images in Figures 1–3 are typical of SDO and STEREO images obtained in thepast few years during the rising phase of sunspot cycle 24. The observed chiralities, dextralin the north and sinistral in the south, satisfy the hemisphere chirality tendencies that werederived from orientations of chromospheric fibrils (Martin et al. 1994; Martin 1998b). Also,the inference of an axial field along the channel suggests that the transition from an upward-pointing field on one side of the channel to a downward-pointing field on the other side isproduced by a rotation of the field as described by Martin et al. (2008, 1994), rather than bycurrent-free loops that directly join opposite-polarity fields on the two sides of the channel. Itis this inferred rotation and its accompanying chiral pattern that distinguishes the magneticstructure of filament channels from potential field configurations.In most of the Fe XII images that we have examined, the tails of cellular plumes becometapered and fade out as they approach the polarity reversal boundary from one side ofa filament channel . An exception is visible in the lower right corner of Figure 3 wherethe southward-pointing plumes make a sharp counterclockwise bend and line up with brightthreads that point directly across the polarity reversal boundary near the right edge of thepanel. These threads link to a growing active region outside the field of view. As we shallsee later, eruptive events produce ‘post flare loops’ that extend across the polarity reversalboundary, and it is possible that these threads in Figure 3 are remnants of intermittenteruptions that occurred in the growing active region.Next, we look at SDO observations to see if we can find examples of this rotationalstructure in broadside views. Figure 4 shows a three-month old filament channel in thenorthern-hemisphere on 2012 April 23. In this oblique view, the cellular plumes are inclinedin opposite directions, tipping like crossed swords, toward the east on the north side of thechannel where the field is positive and toward the west on the south side where the field isnegative. Thus, for an observer standing on the positive-polarity side of the channel, thefield would tilt to the right, corresponding to a dextral chirality. This tilt seems to changewith distance to the channel, with plumes that are nearly horizontal close to the channel andmore vertical farther away, as we have already seen at lower latitudes. The clockwise (right-handed) sense of this rotation is also consistent with a dextral chirality. Finally, we note thatfrom this perspective, the plumes appear more curved on the east side of the channel thanon the west side, and the normal component of field points away from the polarity reversalboundary.Looking eastward and poleward along the filament channel in Figure 4, we see that thechannel bends southward around the trailing end of the negative-polarity region to form alower-latitude extension of the channel. Along the positive-polarity side of this ‘switchback’,the cellular plumes are directed to the west, corresponding to a southward directed horizontal 8 –field. Thus, the horizontal field of the filament channel changed from northward-directed tosouthward-directed as it passed around the bend, and preserved the dextral chirality ofthe field. It is difficult to characterize the plumes in this switchback, but if we regard thepositive-polarity plumes to be the curved ones, then the eastward asymmetry would also bepreserved around the bend.Our potential-field source-surface calculations (not shown here) indicate that abovethis switchback pattern, the magnetic field has a pseudostreamer geometry, produced whenopen fields from the like-polarity coronal holes meet high in the corona (Wang et al. 2007).It is similar to the pseudostreamer configuration that occurred above a switchback con-figuration of dextral filaments on 2010 August 1, as shown in Figure 12 of the paper byPanasenco et al. (2012). Such switchback configurations of dextral (sinistral) chirality arecommon in the northern (southern) hemisphere during the rising phase of the sunspot cyclewhen large unipolar magnetic regions trail poleward from their sources in active regions (seeMackay et al. (2008) and references contained therein).Figure 5 shows the same filament channel 28 days later. This image is dominated bypositive-polarity plumes whose relatively long tails project outward toward the limb. Closerto the polarity reversal boundary, the plumes become progressively more inclined to the east,corresponding to a dextral chirality and a right-handed rotation of the field. On the southside of the channel, the plumes are shorter and bend sharply to the west, indicating a dextralchirality and the eastward asymmetry seen 28 days earlier in Figure 4. Near the negative-polarity coronal hole, the tails of several plumes extend upward and are faintly visible inprojection across the channel. Based on their proximity to the coronal hole, we assume thatthese plumes point along magnetic field lines that extend to great heights before returningto the Sun.This region had a different appearance on 2012 February 27 about a month after itformed. As shown in Figure 6, the positive-polarity region contained many cellular plumeswith their tails combed in the northeast direction along the polarity reversal boundary. Onthe south side of the channel, the tails of the negative-polarity plumes bend southward alongthe channel, indicating a dextral chirality and an eastward asymmetry. Near the negative-polarity coronal hole, several plumes point directly westward across the channel. These loopsare fading remnants of the ‘post-flare loops’ that formed after a filament eruption on 2012February 24.This filament channel and its aftermath of successively forming loops are visible in thetime-lapse sequence of AIA 193 ˚A images in Figure 7. Panel (a) shows the filament at 2300UT on February 23 when a distant segment had erupted, producing the mound of loopsseen under the X-shaped darkening at the east limb. Panel (b) shows that by 0700 UT on 9 –February 24, the rest of that long filament had erupted and was replaced by loops that wereskewed relative to the axis of the channel. Panels (c) and (d) show that as time passed,new loops formed with less skew and at greater heights over the polarity reversal boundary(assuming that their heights scale roughly as their lengths).In effect, the time sequence provides a cut-away view of the field, revealing the well-known weakening of the axial component with height. Because the majority polarity ispositive on the west side of the channel (cf. Figure 6), this field has a dextral chirality. Asshown in Panel (f), only a few loops remained visible across the channel at 1200 UT onFebruary 25, but many loop legs survived to produce the ‘sea of cellular plumes’ on the westside of the channel. Their common alignment (and a potential-field calculation that is notshown here) suggests that they are linked to negative-polarity flux farther to the north.
3. SUMMARY AND DISCUSSION
In the previous section, we combined AIA Fe XII 193 ˚A images with HMI Fe I 6173 ˚Amagnetograms to infer the direction of the horizontal magnetic field in filament channels.The technique is the same as that used previously for chromospheric structures; the positive-polarity plumes point in the direction of the field and the negative-polarity plumes point inthe opposite direction. If the plumes point to the right when you are looking from thepositive-polarity side of the channel, then the field has a dextral chirality. If they point tothe left, the chirality is sinistral. Once this handedness has been determined for one featurewithin a chiral system, the chiralities of all the other features within the system are known.For a dextral filament channel, the field components of the overlying arches will be skewedto the left and the threads of the associated filament barbs will project downward to theright (into the minority-polarity flux elements of negative-polarity). For a sinistral channel,these quantities will be reversed.We have seen that the cellular plumes within filament channels turn sharply sideways,and point along the polarity reversal boundary, rather than directly across the boundaryto their nearby counterparts of opposite polarity, as would occur for a potential magneticfield. Consequently, these observations are consistent with the general consensus that themagnetic field of a filament channel is non-potential.In this paper, we have concentrated on the systematic properties of the componentof field along the filament channel. However, it is also important to consider the normalcomponent, which is the starting orientation for plumes before they curve inward towardthe polarity reversal boundary. As described previously by Su et al. (2010), the plume 10 –orientations are not symmetric across the channel. On the two sides of the channel, the tailsof the plumes point in opposite directions along the polarity reversal boundary, but in thesame direction normal to this boundary. Because these two sets of plumes are rooted infields of opposite polarity, the normal component of field must point in opposite directionson the two sides of the polarity reversal boundary, either toward the boundary if the bentplumes originate on the positive-polarity side of the channel or away from the boundary ifthe bent plumes originate on the negative-polarity side.These oppositely directed fields raise the question of how the fields on one side of thechannel can join their opposite-polarity counterparts on the other side. Su et al. (2010)suggested that the fields on the ‘curved-plume’ side of the channel merge into a flux rope inthe channel and that the fields on the ‘straight-plume’ side go much farther away, either toremote footpoints across the channel or to their opposite-polarity counterparts elsewhere onthe Sun. Regardless of whether there is a flux rope in the channel, we have no doubt that thebent plumes indicate fields that curve sideways to become part of the horizontal field in thechannel. The idea that the fields of the straight plumes extend to great distances is supportedby Figures 6 and 7, which show a ‘sea of positive-polarity plumes’, whose fields reached overthe polarity reversal boundary immediately after the filament eruption (as indicated by thepost flare loops) as well as to regions farther to the northeast (as suggested by potential fieldcalculations not shown here.Vertical views of filament channels like those in Figures 1–3 show that on one side ofthe channel the tails of the plumes become narrow and fade out as they bend asymptoticallyalong the polarity reversal boundary. It is significant that the tails fade out while pointingalong the boundary rather than directly toward the boundary. This implies that the fieldlines do not cross the polarity reversal boundary directly in a simple arcade of loops. Ifthey cross the boundary at all, they do it incognito by first merging with the horizontal fieldon one side of the channel, and then after passing some distance along the channel, peelingaway from the horizontal field on the other side of the channel. Wang et al. (2013) arguedthat this scenario is consistent with the diffusive annihilation of flux at the polarity reversalboundary, but they did not consider an asymmetry across the channel.As mentioned in the previous section, the curved plumes were on the west side of thefilament channel in Figures 1 and 2, rather than the east side as Su et al. (2010) found. Thiscounter example prompted us to search for more examples, and we found several in a briefsurvey of SDO/AIA 193 ˚A images during June–November 2011. We found all combinations— curved plumes on the east side of the channel, curved plumes on the west side, and curvedplumes on both sides at different locations along the channel. We obtained the impressionthat the specific geometry depends on the shape of the channel as well as the surrounding 11 –distribution of flux on the Sun. Nevertheless, the asymmetry of the field normal to thechannel seemed to persist with plumes curving toward the polarity reversal boundary on oneside of the channel and straighter plumes pointing away from the boundary on the otherside.The converging plumes may also be telling us that the field lines do not cross the filamentchannel at low heights, but instead become linked like barbs to elements of minority polarityon the same side of the polarity reversal boundary or in opposite-polarity flux at the end ofthe filament channel (Martin et al. 1994; Martin 1998a,b). Sheeley & Warren (2012) foundthat the amount of minority-polarity flux in the cellular regions is only ∼ ∼ Thus, the cellular plumes are similar to the classic polarplumes modeled by Newkirk & Harvey (1968), except that the cellular plumes are linked to http://umbra.nascom.nasa.gov/images 2013 May 29
12 –regions of opposite polarity and are therefore brighter, hotter, and more closely spaced thanpolar plumes.The filament channels in this paper all had chiralities that were consistent with the‘hemisphere chirality tendency’, with dextral in the north and sinistral in the south (Leroy et al.1983; Martin et al. 1994; Martin 1998b; Zirker et al. 1998). Thus, their fields made a right-handed (clockwise) rotation across the northern hemisphere channels and a left-handed(counter-clockwise) rotation across the southern-hemisphere channels. When these samechannels are described in terms of magnetic helicity and its associated rotation, the signsare reversed, with negative helicity and left-handed rotation in the north and positive helic-ity and right-handed rotation in the south (DeVore 2000; Wang et al. 2013; Berger & Field1984). This is essentially a matter of definition with negative helicity occurring when thecurrent is directed opposite to the magnetic field (as happens for a dextral configuration)and with rotation corresponding to the leftward skew of the overlying field lines. However,numerous real exceptions to the hemispheric chirality ‘rule’ do exist and have been studiedby many authors (van Ballegooijen et al. 1998, 2000; Martin & Wen 2009; Yeates & Mackay2009; Martin et al. 2012; Wang et al. 2013).It is interesting to consider whether coronal cells (or cellular plumes) can be regarded asindividual domains of force-free magnetic structure that have spread out from larger and moreconcentrated sources in active regions. The Fe XII images do not show spiral patterns withinindividual cells like those that have been calculated previously for sunspots and other discreteflux distributions (Nakagawa et al. 1971; Nakagawa & Raadu 1972; Sheeley & Harvey 1975;Chiu & Hilton 1977). Rather, the images in this paper show teardrop-shaped tails bendingsideways under the influence of the horizontal field. Thus, it is plausible that the magneticfield of the filament channel consists of force-free domains, each centered on a magnetic fluxconcentration and exhibiting writhe rather than twist. Where these domains come in contact,their azimuthal fields would either oppose each other and reconnect, or reinforce each other,depending on the chirality and polarity of the fields. Thus, in the northern hemisphere wherethe fields are dextral, we would expect horizontal fields to circulate clockwise (seen fromabove) around large-scale regions of positive polarity and counter clockwise around regions ofnegative polarity. These circulations would be reversed in the southern hemisphere. Similarapplications of Ampere’s Law have been proposed recently as an inverse cascade process fortransporting helicity Antiochos (2009, 2012) and as a way of calculating the evolution of theaxial magnetic field in filament channels (Wang et al. 2013).In this paper, we have concentrated on observations of cellular plumes in filament chan-nels, and have given little attention to prominences and their eruptions. However, a popularscenario is that solar magnetic fields in filament channels involve stresses and helicities that 13 –accumulate over time and are released during filament eruptions. In the future, we mayexpect to obtain observational constraints on such hypotheses by tracking the evolution ofcoronal plumes in these Fe XII images.We are grateful to the AIA and HMI science teams for providing observations from theNASA SDO spacecraft. We are also grateful to the SECCHI science team for providingEUVI images of coronal cells and to Nathan Rich (NRL) for his continuing help in the devel-opment of software for observing and analyzing SOHO, STEREO, and SDO images. Also,we thank the referee for helpful suggestions and for pointing out the interesting asymmetryacross filament channels, as described by Su et al. (2010). One of us (NRS) is pleased toacknowledge Y.-M. Wang (NRL) for a variety of help and advice, including his suggestionthat the normal component of field be used in future numerical studies of filament channels.NRS is also indebted to C. R. Devore (NASA/GSFC) and S. K. Antiochos (NASA/GSFC)for useful discussions of force-free fields. OP is grateful to M. Velli (JPL) for pointing outthat coronal cells in a filament channel follow the pattern of chromospheric fibrils. OP andSFM were supported under NASA grant NNX09AG27G. SFM was supported under Predic-tive Science, Inc. contract NNH 12CF37C. NRS and HPW acknowledge financial supportfrom NASA and the Office of Naval Research.
REFERENCES
Antiochos, S. K. 2009, in AAS/Solar Physics Division Meeting, Vol. 40, AAS/Solar PhysicsDivision Meeting
This preprint was prepared with the AAS L A TEX macros v5.2.
16 –Fig. 1.— An AIA 193 ˚A image and an HMI magnetogram (white positive and black negative),showing a filament channel between two coronal holes in the southern hemisphere on 2011June 17. On each side of the channel, coronal cells are distorted into tadpole shapes with theirheads rooted in elements of majority polarity and their tails pointing in opposite directionsalong the channel. This corresponds to a southward directed field along the channel and asinistral chirality. In all figures of this paper, solar north is up and east is to the left. 17 –Fig. 2.— SDO observations of the region in Fig. 1, obtained 28 days later on 2011 July15. Cells are still distorted into tadpoles with their heads rooted in elements of majoritypolarity and their tails oriented in opposite directions on the two sides of the channel. Thiscorresponds to a southward directed field along the channel and a sinistral chirality. 18 –Fig. 3.— SDO images of a northern-hemisphere region on 2011 September 08. Cells, rootedin elements of negative polarity, are distorted into tadpole shapes with their tails pointingsouthward along the channel. This corresponds to a northward field along the channel anda dextral chirality. 19 –Fig. 4.— SDO images of a northern-hemisphere filament channel on 2012 April 23, showingcellular plumes leaning in opposite directions on the two sides of the channel, correspondingto a dextral chirality. This chirality is preserved around the switchback at the trailing endof the channel where positive-polarity plumes at lower latitude point to the west. 20 –Fig. 5.— The region in Fig. 4, 28 days later on 2012 May 21 when the plumes on thepositive-polarity side of the channel appear to be lined up vertically like a row of teeth.A close examination reveals that these positive-polarity plumes occur in several rows withfainter, more tilted plumes in the foreground next to the channel and brighter, more verticalplumes in the back - suggesting a rotation of the field and a dextral chirality. 21 –Fig. 6.— The filament channel from Figs. 4 and 5, but seen on 2012 February 27, whenmany positive-polarity plumes pointed northward parallel to the channel and the tails ofsome plumes near the negative-polarity coronal hole stretched westward across the channel. 22 – c U T b U T ( / ) a U T ( / / ) f U T ( / ) e U T d U T ( / ))