The Relation between EIT Waves and Coronal Mass Ejections
aa r X i v : . [ a s t r o - ph . S R ] M a y Submitted to ApJL
The Relation between EIT Waves and Coronal Mass Ejections
P. F. Chen
Department of Astronomy, Nanjing University, Nanjing 210093, China; [email protected]
ABSTRACT
More and more evidence indicates that “EIT waves” are strongly related to coronalmass ejections (CMEs). However, it is still not clear how the two phenomena arerelated to each other. We investigate a CME event on 1997 September 9, which was wellobserved by both EUV imaging telescope (EIT) and the high-cadence MK3 coronagraphat Mauna Loa Solar Observatory, and compare the spatial relation between the “EITwave” fronts and the CME leading loops. It is found that “EIT wave” fronts are co-spatial with the CME leading loops, and the expanding EUV dimmings are co-spatialwith the CME cavity. It is also found that the CME stopped near the boundary of acoronal hole, a feature common to observations of “EIT waves”. It is suggested that“EIT waves”/dimmings are the EUV counterparts of the CME leading loop/cavity,based on which we propose that, as in the case of “EIT waves”, CME leading loops areapparently-moving density enhancements that are generated by successive stretching(or opening-up) of magnetic loops.
Subject headings:
Sun: corona — Sun: coronal mass ejections (CMEs) — Sun: flares— waves
1. Introduction
As the largest-scale eruptive phenomenon on the Sun, coronal mass ejections (CMEs) are oftenaccompanied by many other phenomena, which are visible in various wavelengths, such as solarflares, filament eruptions, and “EIT waves”. Among them, “EIT waves” are extremely enigmatic.They have attracted wide attention and provoked a lot of debate concerning their physical nature.The first reports of the phenomenon appeared in Dere et al. (1997) and Moses et al. (1997).Thompson et al. (1998) analyzed the famous 1997 May 12 event in detail using the data from theEUV Imaging Telescope (EIT) on board the
Solar and Heliospheric Observatory ( SOHO ) spacecraft.In association with the CME event, the running difference images of the EIT data showed that analmost circular bright front, with an averaged intensity enhancement of ∼
20% (Thompson et al.1999), propagates away from the source active region, giving an impression of a wave. Theywere conventionally called “EIT waves” since they were discovered with the EIT telescope, though 2 –sometimes they are referred to as “coronal waves” (Warmuth et al. 2004; Attrill et al. 2007;Tripathi & Raouafi 2007).As “EIT waves” propagate outward, they are immediately followed by expanding dimmings.Therefore, the two phenomena were proposed to result from the same physical process, which wasvery controversial during the past years (e.g., Chen 2008). One debate concerns the driving sourceof the wave propagation. It was quite often claimed that “EIT waves” are generated by the pressurepulse which may be a solar flare (e.g., Wu et al. 2001; Vrˇsnak et al. 2002), whereas Plunkett et al.(1998), Delann´ee & Aulanier (1999), Chen et al. (2002), and Chen, Fang, & Shibata (2005) pro-posed that they are associated with CMEs. Biesecker et al. (2002) studied 173 events and foundthat after correcting for observational biases all EIT waves are associated with CMEs. Noticing thatabout half of the “EIT waves” observed from 1997 March to 1998 June were associated with smallsolar flares below C class, Cliver et al. (2005) pointed out that it is hard to imagine that such weakflares can produce global-scale perturbations, so that “EIT waves” should be preferentially associ-ated with CMEs. As a complementary proof, Chen (2006) selected strong (M- and X-class) flaresnear solar minimum that were not associated with CMEs, and found that none of them produced“EIT waves”. Furthermore, Chen (2006) showed that from the same active region within the sameday, a weaker flare with a CME was associated with an “EIT wave”, however, another stronger flarewithout a CME was not accompanied by an “EIT wave”. In addition, Podladchikova & Berghmans(2005) and Attrill et al. (2007) found that EIT waves rotate in senses determined by the helicityof the CME source region, a feature not expected from flare-induced waves. These pose strong evi-dence to support that “EIT waves” are purely related to CMEs. However, the physical connectionbetween “EIT waves” and CMEs is still not well established. Veronig et al. (2008) proposed that“EIT waves” are fast-mode waves driven by CME flanks, while Chen & Fang (2005) theoreticallypostulated that “EIT wave” fronts are the EUV signature of the CME leading loops, and accord-ingly, the expanding EUV dimmings are the EUV signature of the CME cavity. With the dataanalysis of the CME event on 1997 September 9, this Letter aims to clarify the connection between“EIT waves” and CMEs.
2. Observations “EIT waves” are observed by EUV imaging telescopes like EIT, with a field of view (FOV) of ∼ . R ⊙ . Before the launch of the recent STEREO satellites, CMEs were mostly observed by threecoronagraphs (C1, C2, and C3) in the Large Angle and Spectrometric Coronagraph (LASCO),which are on board the SOHO spacecraft. The white-light coronagraphs C2 and C3, which observethe coronal mass directly, have a FOV beyond 2 R ⊙ . Therefore, the comparison of the observationsbetween them and EIT relies on spatial extrapolation, which introduces uncertainties. The C1coronagraph has a FOV from 1 . − R ⊙ , which overlaps with that of EIT. However, during itsshort lifetime, it was often observing the corona with the forbidden lines, whose intensity dependsnot only on the coronal density, but also on the temperature. Therefore, their structure may not 3 –necessarily be co-spatial with the white-light CME. In order to precisely determine the spatialrelation between “EIT wave” fronts and CMEs, we study ground-based coronagraph white-lightdata with a FOV overlapping that of EIT.Thompson & Myers (2009) compiled a catalog of 176 “EIT wave” events that occurred be-tween 1997 March and 1998 June. In order to compare the “EIT wave” fronts and CMEs, weselected those “EIT wave” events that have fronts appearing above the solar limb. It is found that38 cases in the catalog are off-limb events. We then search for the white-light data from the Mark-III K-Coronameter (MK3), which was installed at the Mauna Loa Solar Observatory (MLSO). Itturned out that only the 1997 September 9 event was well observed by the MK3 coronagraph. The“EIT wave” propagation was observed by the EIT instrument, and the CME was observed by boththe ground-based coronagraph MK3 and the space-borne LASCO instrument.The EIT instrument is a normal-incidence, multilayer EUV telescope (Delaboudini`ere et al.1995). It observes the full-disk solar corona, extending up to 1 . R ⊙ with a pixel size of 2 . ′′ . Thereare four narrow bandpass EUV channels centered at 171, 195, 284, and 304 ˚A, which selectivelyobserve spectral lines formed by Fe IX/X , Fe
XII , Fe
XII , and He II , respectively. The Fe XII ∼
16 min. The Fe
XII emission line exhibits a peakemission near 1.5 MK. The LASCO instrument consists of a set of three coronagraphs, i.e., C1,C2, and C3, with overlapping and concentric FOV. C2 and C3 are traditional white-light corona-graphs that observe Thomson-scattered visible light through a broadband filter (Brueckner et al.1995). The MK3 coronagraph at MLSO began observations in 1980 (MacQueen & Fisher 1983). Itmeasures polarization brightness of photospheric radiation scattered by free electrons in the lowercorona, with a FOV of 1 . − . R ⊙ , a pixel size of 20 ′′ , and a cadence of 3 min.
3. Results
The leading loop of the CME was visible in the FOV of the MK3 coronagraph from 19:34:32UT on 1997 September 9. The data before that were not of high enough quality to clearly showthe structure of the CME. The upper panels of Figure 1 display the white-light base-differenceimages of the CME, which were observed by MK3. The pre-event intensity map at 19:04:11 UT ischosen as the base image that is subtracted from the later images. The high-cadence observationsof MK3 indicated that the CME leading loop accelerated from 146 ±
36 km s − at 19:34:32 UT to366 ±
36 km s − at 19:47:44 UT, after which the top of the leading loop went out of the FOV of theMK3. At 20:06:02 UT, the CME began to be visible in the FOV of the LASCO C2 coronagraph at2 . R ⊙ , as seen from the lower panels of Figure 1. At 20:33:29 UT, the CME leading loop movedto a heliocentric distance of 5 . R ⊙ in the plane-of-the-sky. The averaged radial speed of the CMEpropagation is estimated to be 726 ±
20 km s − in the FOV of LASCO C2, which is almost twicethat measured in the FOV of the MK3.As the CME lifted off, no flare-like brightening was visible on the solar disk, inferring that the 4 –source region of this CME event was located behind the solar limb. However, the full-disk imager,EIT, detected the propagation of an “EIT wave”, as revealed by the middle and right panels ofFigure 2. In this figure, the left panel shows the pre-event Fe XII
195 ˚A image, and the middleand right panels depict the evolution of the base-difference 195 ˚A images. It is seen that the mainpart of the “EIT wave” was above the limb, extending beyond the FOV of EIT, with some weakbrightenings on the solar disk as indicated by the white arrow in the right panel. As revealed bythe left panel of Figure 2, a polar coronal hole existed on the northern side of the “EIT wave”. Thenorthern leg of the “EIT wave” was approaching the coronal hole from 19:26:33 UT to 19:44:44UT, after which it stopped near the boundary of the coronal hole.Owing to the high cadence of the MK3 observations, the spatial positions of the white-lightCME and the “EIT wave” can be compared directly when the two phenomena were observed almostsimultaneously. The white-light images of the CME at 19:44:27 UT and 20:00:49 UT are displayedin the upper panels of Figure 3, and the corresponding EIT 195 ˚A images at 19:44:44 UT and20:00:14 UT are shown in the lower panels. For clearness, the outlines of the white-light CMEleading loops (defined by eye) in the upper left and right panels are superimposed on the EITFe
XII intensity map in the lower left and right panels, respectively, as represented by the thickblack lines. It is seen that the “EIT wave” fronts above the limb are almost co-spatial with theleading loop of the CME, with slight differences in the detailed structures. Accordingly, the EITdimmings are co-spatial with the white-light cavity of the CME. Note that there is a small timedifference in the white-light and EUV observations ( ∼
35 s). Considering that the top of the CMEleading loop was moving with a speed of 366 ±
36 km s − in the FOV of the MK3, such a timedifference corresponds to a spatial shift of 18 ′′ , which is below the spatial resolution of the MK3coronagraph.
4. Discussions4.1. Spatial Relation between “EIT Waves” and CMEs
More and more evidence tends to support that “EIT waves” are related to CMEs, ratherthan solar flares. However, exactly how “EIT waves” are related to CMEs is still not clear. Thesimultaneous observations of an “EIT wave” in Fe
XII
195 ˚A and a white-light CME in the 1997September 9 event provide a precious opportunity to tackle this question.The source region of the eruption was located behind the solar limb, hence the “EIT wave”was observed as a limb event, which allows the direct comparison with coronagraph data. Ouranalysis in Section 3 (see Figure 3) indicates that the “EIT wave” fronts were almost co-spatialwith the leading loops of the CME during the eruption, and the expanding EUV dimmings, whichimmediately followed the “EIT wave” fronts, were co-spatial with the dark cavity of the CME.The brightening of the “EIT wave” fronts in Fe
XII
195 ˚A might be due to temperature 5 –variation and/or density enhancement (Thompson et al. 1999). Therefore, it is controversialwhether the brightening is mainly contributed by temperature variation (e.g., due to Joule heating,Delann´ee et al. 2008) or by density enhancement (Wills-Davey & Thompson 1999). Consideringthat “EIT waves” are observed simultaneously in several EUV lines that have different formationtemperatures (Wills-Davey & Thompson 1999; Long et al. 2008), it is believed that the “EITwave” brightening is mainly due to density enhancement, though the adiabatic compression mayincrease the plasma temperature to some extent, resulting in some differences between detailedfeatures in different lines (Wills-Davey & Thompson 1999; Chen & Fang 2005). The approximateco-spatiality of “EIT wave” fronts and white-light CME leading loops, as revealed by Figure 3,provides direct evidence that “EIT wave” brightenings are mainly contributed by density enhance-ment, since the white-light enhancement of the CME leading loops is produced by the increasedcoronal density only. Therefore, we conclude that “EIT wave” fronts are mainly due to densityenhancement, and they are the EUV signatures of the CME leading loops, as we theoretically pro-posed in Chen & Fang (2005). It is also inferred that, similar to the frequently assumed dome-likeshape of CME leading loops, the “EIT wave” front should also be dome-like, and the circular “EITwave” fronts, sometimes observed on the solar disk, are just a projection of the three-dimensionaldome-like structure, whose skirt is much brighter than the top of the dome.
The physics behind the formation of CME leading loops is still not clear. It is often takenfor granted that we observe coronal plasma embedded in the erupting magnetic loops. However,with UV spectral observations of halo CMEs, Ciaravella et al. (2006) found that the CME frontsshow Doppler shifts significantly smaller than their apparent velocity obtained with white-lightcoronagraphs, suggesting that CME leading loops (at least for the halo events), might be fast-modeshocks rather than being plasma carried outward by erupting magnetic loops. The co-spatiality of“EIT waves” and CME leading loops found in this Letter could also shed light on the nature of theCME leading loops.“EIT waves” are often considered to be fast-mode magnetoacoustic waves in the corona (e.g.,Wang 2000; Wu et al. 2001; Vrˇsnak et al. 2002; Warmuth et al. 2004; Grechnev et al. 2008;Pomoell et al. 2008). However, the wave model cannot explain the following features of “EITwaves” (e.g., see Wills-Davey et al. 2007; Chen 2008, for details): (1) the “EIT wave” velocityis significantly smaller than those of Moreton waves. The latter are well established to be due tofast-mode waves in the corona; (2) The “EIT wave” velocities have no correlation with those of typeII radio bursts (Klassen et al. 2000); (3) The “EIT wave” fronts may stop when they meet withmagnetic separatrices (Delann´ee & Aulanier 1999); (4) The “EIT wave” velocity may be below100 km s − (e.g., the Fig. 3 of Long et al. 2008), which is even smaller than the sound speed inthe corona. These strange features provoked Delann´ee & Aulanier (1999) to relate “EIT waves” tomagnetic restructuring during CMEs. With MHD numerical simulations, Chen et al. (2002; 2005) 6 –identified the “EIT wave” features to correspond to stretching of the magnetic field. The “EITwave” disturbance was observed well behind the fast-mode piston-driven shock waves during CMEeruptions. Chen et al. (2002; 2005) proposed that “EIT waves” are apparently-moving densityenhancements, which are actually produced by successive stretching (or opening-up) of closed fieldlines, rather than being real waves. The model can account for the main characteristics of “EITwaves”, such as the low velocity, their diffuse fronts, the stationarity near magnetic separatrices,and found support in observations (e.g., Harra & Sterling 2003). The co-spatiality of “EIT waves”and CME leading loops found in this paper would infer that CME leading loops are also generatedby the successive stretching of overlying magnetic loops. As illustrated by Figure 4, as the corestructure, e.g., a magnetic flux rope, erupts, the resulting perturbation propagates outward inevery direction, with a probability of forming a piston-driven shock as indicated by the pink lines.However, different from a pressure pulse, the erupting flux rope continues to push the overlyingmagnetic field lines to move outward, so that the field lines are stretched outward one by one. Foreach field line, the stretching starts from the top, e.g., point A for the first magnetic line, and thenis transferred down to the leg (point D) with the Alfv´en speed. The deformation at point A isalso transferred upward to point B of the second magnetic line with the fast-mode speed. Such adeformation would also be transferred down to its leg (point E) with the local Alfv´en speed, bywhich the entire second magnetic line is stretched up. The stretching at any part of the magneticfield lines compresses the coronal plasma on the outer side, producing density enhancements. Allthe newly formed density enhancements form a pattern ( green ), which is observed as the CMEleading loop. Similar to “EIT wave” fronts, the legs of the CME leading loop separate initially, andmay stop when they meet with magnetic separatrices such as the boundary of coronal holes. Thisis why CMEs generally maintain a fixed angular span in their later stages. At the same time, asthe field lines are stretched outward, the enveloped volume increases, resulting in coronal dimmings(or the dark cavity) behind the CME leading loop.To conclude, with the simultaneous observations of an “EIT wave” in Fe XII
195 ˚A and aCME in white light, we found that “EIT wave” fronts are co-spatial with CME leading loops,and accordingly, expanding EUV dimmings are co-spatial with the CME cavity. We postulatethat the CME leading loops may have the same formation mechanism and physical nature as “EITwaves”, i.e., they are the apparently-moving density enhancements that are generated by successivestretching (or opening-up) of magnetic loops. It is noted that in other “EIT wave” models like thereconnection model of Attrill et al. (2007), the “EIT wave” front is also expected to be co-spatialwith the CME leading loop. The difference between these models should be explored further.The author thanks C. Fang and the referee for the constructive suggestions, and F. Gu forthe help of data analysis. The research is supported by the Chinese foundations 2006CB806302and NSFC (10403003, 10221001, and 10333040).
SOHO is a project of international cooperationbetween ESA and NASA. We are grateful to the MLSO team for making their data available. 7 –
REFERENCES
Attrill, G. D. R., Harra, L. K., van Driel-Gesztelyi, L., & D´emoulin, P. 2007, ApJ, 656, L101Biesecker, D. A., Myers, D. C., Thompson, B. J. et al. 2002, ApJ, 569, 1009Brueckner, G. E., et al. 1995, Sol. Phys., 162, 357Chen, P. F. 2006, ApJ, 641, L153Chen, P. F. 2008, J. of Astrophys. Astron., 29, 179Chen, P. F. & Fang, C. 2005, IAU Symp., 226, 55Chen, P. F., Fang, C., & Shibata, K. 2005, ApJ, 622, 1202Chen, P. F., Wu, S. T., Shibata, K., & Fang, C. 2002, ApJ, 572, L99Ciaravella, A., Raymond, J. C., & Kahler, S. W. 2006, ApJ, 652, 774Cliver, E. W., Laurenza, M., Storini, M. & Thompson, B. J. 2005, ApJ, 631, 604Delaboudini´ere, J.-P. et al. 1995, Sol. Phys., 162, 291Delann´ee, C. & Aulanier, G. 1999, Sol. Phys., 190, 107Delann´ee, C., T¨or¨ok, T., Aulanier, G., & Hochedez, J.-F. 2008, Sol. Phys., 247, 123Dere, K. P., et al. 1997, Sol. Phys., 175, 601Grechnev, V. V., Uralov, A. M., Slemzin, V. A. et al. 2008, Sol. Phys., 253, 263Harra, L. K. & Sterling, A. C. 2003, ApJ, 587, 429Klassen, A., Aurass, H., Mann, G., & Thompson, B. J. 2000, A&AS, 141, 357Long, D. M., Gallagher, P. T., McAteer, R. T. J., & Bloomfield, D. S. 2008, ApJ, 680, L81MacQueen, R. M., & Fisher, R. R. 1983, Sol. Phys., 89, 89Moses, D., et al. 1997, Sol. Phys., 175, 571Plunkett, S. P. et al. 1998, Geophys. Res. Lett., 25, 2477Podladchikova, O., & Berghmans, D. 2005, Sol. Phys., 228, 265Pomoell, J., Vainio, R., & Kissmann, R. 2008, Sol. Phys., 253, 249Thompson, B. J., Gurman, J. B., Neupert, W. M. et al. 1999, ApJ, 517, L151Thompson, B. J. & Myers, D. C. 2009, ApJS, in press 8 –Thompson, B. J., Plunkett, S. P., Gurman, J. B. et al. 1998, Geophys. Res. Lett., 25, 2465Tripathi, D., & Raouafi, N.-E. 2007, A&A, 473, 951Veronig, A. M., Temmer, M., & Vrˇsnak, B. 2008, ApJ, 681, L113Vrˇsnak, B., Warmuth, A., Brajˇsa, R., & Hanslmeier, A. 2002, A&A, 394, 299Warmuth, A., Vrˇsnak, B., Magdaleni´c, J., Hanslmeier, A., & Otruba, W. 2004, A&A, 418, 1117Wills-Davey, M. J., DeForest, C. E., & Stenflo, J. O. 2007, ApJ, 664, 556Wills-Davey M. J. & Thompson, B. J. 1999, Sol. Phys., 190, 467Wang, Y. -M. 2000, ApJ, 543, L89Wu, S. T., Zheng, H. N., Wang, S. et al. 2001, J. Geophys. Res., 106, 25089
This preprint was prepared with the AAS L A TEX macros v5.2.
Upper panels : images observed by MK3 coronagraph inMLSO.
Lower panels : images observed by LASCO C2 coronagraph. 10 –Fig. 2.— EIT 195˚A images showing the evolution of the “EIT wave” event.
Left : Pre-event imageat 19:13:32 UT.
Middle : Base difference image at 19:26:33 UT.
Right : Base difference image at19:44:44 UT. White arrow shows weak brightening of the “EIT wave” visible on the solar disk. 11 –Fig. 3.— Spatial comparison of the CME leading loop and the “EIT wave” front.
Upper panels : Twosnapshots of the white-light CME eruption observed by the MLSO MK3 coronagraph at 19:44:27UT and 20:00:49 UT.
Lower panels : The Fe
XII
195 ˚A images of the “EIT wave” propagationobserved by the EIT instrument at 19:44:44 UT and 20:00:14 UT. 12 –Fig. 4.— A schematic sketch of the formation mechanism of CME leading loops, where the CMEleading loop ( greengreen