A Three-Dimensional Velocity of an Erupting Prominence Prior to a Coronal Mass Ejection
Maria V. Gutierrez, Kenichi Otsuji, Ayumi Asai, Raul Terrazas, Mutsumi Ishitsuka, Jose Ishitsuka, Naoki Nakamura, Yusuke Yoshinaga, Satoshi Morita, Takako T. Ishii, Satoru UeNo, Reizaburo Kitai, Kazunari Shibata
aa r X i v : . [ a s t r o - ph . S R ] J a n A Three-Dimensional Velocity of an EruptingProminence Prior to a Coronal Mass Ejection
Maria V.
GUTIERREZ ∗ , Kenichi OTSUJI , Ayumi
ASAI , Raul TERRAZAS ,Mutsumi ISHITSUKA , Jose ISHITSUKA † ,6 , Naoki NAKAMURA , Yusuke YOSHINAGA , Satoshi MORITA , Takako T. ISHII , Satoru UENO , Reizaburo KITAI and Kazunari
SHIBATA Institute for Research in Astronomy and Astrophysics (IAFE CONICET-UBA), Buenos Aires,Argentina Geophysical Institute of Peru, Lima, Peru National Institute of Information and Communications Technology (NICT), Koganei, Tokyo,184-8795, Japan Astronomical Observatory, Kyoto University, Sakyo, Kyoto, 606-8502, Japan Ica Solar Station, Department of Physics, San Luis Gonzaga National University of Ica, Ica,Peru National University of the Center of Peru, Huancayo, Peru Department of Astronomy, Kyoto University, Sakyo, Kyoto, 606-8502, Japan National Astronomical Observatory of Japan, Osawa, Mitaka, Tokyo, 181-8588, Japan Bukkyo University, Kita, Kyoto, 603-8301, Japan Ritsumeikan University, Kita, Kyoto, 603-8577, Japan Doshisha University, Kyotanabe, Kyoto, 610-0394, Japan ∗ E-mail: [email protected]
Received ; Accepted 2021 January 8
Abstract
We present a detailed three-dimensional (3D) view of a prominence eruption, coronal loop ex-pansion, and coronal mass ejections (CMEs) associated with an M4.4 flare that occurred on2011 March 8 in the active region NOAA 11165. Full-disk H α images of the flare and filamentejection were successfully obtained by the Flare Monitoring Telescope (FMT) following its re- ocation to Ica University, Peru. Multiwavelength observation around the H α line enabled us toderive the 3D velocity field of the H α prominence eruption. Features in extreme ultraviolet werealso obtained by the Atmospheric Imager Assembly onboard the Solar Dynamic Observatory andthe Extreme Ultraviolet Imager on board the
Solar Terrestrial Relations Observatory - Ahead satel-lite. We found that, following collision of the erupted filament with the coronal magnetic field,some coronal loops began to expand, leading to the growth of a clear CME. We also discussthe succeeding activities of CME driven by multiple interactions between the expanding loopsand the surrounding coronal magnetic field.
Key words:
Sun: chromosphere — Sun: flares — Sun: filaments, prominences — Sun: coronal massejections (CMEs)
Solar flares are explosive events occurring in the solar corona that release extremely large quantities( – erg) in radiative, kinetic, thermal and non-thermal form. They are considered to be oneof the most attractive scientific objects in solar physics, with emissions across the electromagneticspectrum originating from atmospheric layers extending from the chromosphere, in extreme cases thephotosphere, to the outer regions of the solar corona. Solar flares can also produce high-speed coronalmass ejections (CMEs) that travel through interplanetary space with occasional serious impacts on thesolar-terrestrial environment.A filament, relatively cool, dens plasma supported by magnetic field and floated in the corona,sometimes erupts and cause or be part of a CME. Therefore, the relations among filament eruptions,flares and CMEs have been earnestly studied (e.g., Schmieder et al. 2013; Schmieder et al. 2015;Schmieder et al. 2020). Munro et al. (1979) reported that more than 70 % of CMEs are associatedwith eruptive prominences of filament disappearance, both with and without large soft X-ray (SXR)enhancement, that are accordingly recognized as flares. Gopalswamy et al. (2003) also reportedthat 72 % of prominence eruptions observed in microwaves are associated with CMEs. Filamentseruptions are thought to be caused by loss of equilibrium and/or magnetohydrodynamic instability(D´emoulin and Aulanier 2010; Kliem and T¨or¨ok 2006; Mackay et al. 2010).Finding the origin of a CME in the solar lower atmosphere and following its evolution to bea CME are, however, sometimes difficult owing to a gap between the coronagraphic observational † This is the former affiliation. α data with the goal of detecting signatures preceding the launch of CMEs. Recently, athree-dimensional (3D) prominence reconstruction was done by using IRIS data (Schmieder et al.2017). The authors succeeded to demonstrate that an helical structure of the prominence seen inthe IRIS images consists in fact of horizontal field lines parallel to the solar surface. Spectroscopicobservations of eruptive filaments have become more important to reveal the relation between CMEsand filament eruptions.To this end, we examined a set of solar phenomena that occurred on 2011 March 8, in ActiveRegion (AR) NOAA 11165. We examined the erupted prominence in detail and followed its 3Dtemporal evolution and relation to the overlying coronal magnetic field. The remainder of this paperis organized as follows. We describe the observations in Section 2 and present analysis and results onthe prominence eruption and coronal loops prior to the appearance of the CMEs in Sections 3 and 4,respectively. Finally, a summary and discussion are given in Section 5. On 2011 March 8, an M4.4 solar flare with a soft X-ray (SXR) classification was recorded at ARNOAA 11165, which is located near the southwest solar limb (S17 ◦ , W88 ◦ ), by the GeostationaryOperational Environmental Satellite ( GOES ). Figure 1 shows the SXR light curves taken by
GOES .The flare began at 18:08 UT and peaked at 18:28 UT (shown in the figure with arrow 2), with lightcurves featuring several small humps and subpeaks. This flare was followed by another gradual M1.5flare that began and peaked at 19:35 and 20:16 UT (arrow 3), respectively.A filament/prominence eruption associated with the flare can be seen in H α images obtainedby the Flare Monitoring Telescope (FMT; Kurokawa et al. 1995), which was originally installed atHida Observatory, Kyoto University, Japan, and relocated to Ica University, Peru in 2010 under theinternational collaboration of the Continuous H-Alpha Imaging Network (CHAIN) Project (UeNo et3 OES 15 X-Rays : 2011- Mar - 08
16 18 20 22
Time [UT] -9 -8 -7 -6 -5 -4 -3 F l u x [ W m - ] Fig. 1.
GOES
X-ray emission at 1.0 – 8.0 ˚A (red) and 0.5 – 4.0 ˚A (blue) channels of the 2011 March 8 flare. Arrow 2 indicates the peak of the M4.4 flare at18:28 UT. The peak at 20:16 UT corresponds to an M1.5 flare that occurred in the same AR NOAA 11165 (shown with arrow 3). al. 2007). FMT provides full-disk images in five wavelengths: H α line center (6562.8 ˚A); wings at +0 . ˚A and − . ˚A; continuum (6100 ˚A) images; and another H α line center image to detect limbprominences. The spatial and temporal samplings of FMT are about ′′ . , and 20 s, respectively. Inthis paper, we will use the term ‘FMT-Peru’ to clarify that the data we used in this study were obtainedfollowing the relocation.The telescope is capable of measuring the 3D velocity fields of active filaments/prominences(Morimoto & Kurokawa 2003a; Morimoto & Kurokawa 2003b; Cabezas et al. 2017), and Moretonwaves (Eto et al. 2002; Narukage et al. 2002; Narukage et al. 2008; Asai et al. 2012; Cabezas et al.2019). Figure 2 shows a full-disk H α center image of the flare taken by FMT-Peru, with the eruptingfeature visible in the region marked with the white box on the southwest limb.To view the spatial distribution of prominence gases irrespective to line-of-sight motion, wecreated ‘combined’ images of three wavelengths (center and ± . ˚A). Prominences/filaments movingrapidly in the line-of-sight direction are often invisible with images of only one wavelength due tothe Doppler shift of the H α line. The images of three wavelengths were, therefore, combined intotime series to enable tracking of the evolution of overall prominence gas. At each wavelength, aspecific intensity level was set and the brighter regions above that level were extracted to reveal theprominences involved in the selected region. Finally, we superimposed the images extracted at thethree wavelengths at each time step to produce a time sequence of combined H α images, as shownin the leftmost column of Figure 3. See also the supplementary movie 1 for the temporal evolutionof the H α images. It shows images at H α − . ˚A (top left), center (top right), +0 . ˚A (bottom left),respectively. The bottom right panel is the combined images.Flare and eruption features were also captured in extreme ultraviolet (EUV) images taken4 MT H a center 6562.8A -1000 -500 0 500 1000X (arcsecs)-1000-50005001000 Y ( a rc s e c s ) Fig. 2.
Full-disk solar image (solar north is up and west is to the right) of the flare in H α center taken at 18:14 UT by FMT-Peru. The white box shows theflaring region of AR NOAA 11165. by the Atmospheric Imaging Assembly (AIA; Lemen et al. 2012) onboard the Solar DynamicsObservatory ( SDO ; Pesnell et al. 2012). AIA takes full-disk solar images with a temporal resolutionof 12 s, while the images of filters 1600 and 1700 ˚A are taken every 24 s. The pixel size of the imagesis ′′ . . In this study, we used images taken with 304, 171, 193, and 94 ˚A filter, which are mainlyattributable to the He II (80,000K), Fe IX (0.6MK), Fe XII (1.2MK), and Fe
XVIII lines (6.3MK),respectively. The second to fifth columns from the left in Figure 3 show the temporal evolution of theflare and the prominence eruption at 304, 171, 193, and 94 ˚A, respectively.The prominence eruption associated with the flare is seen as a bright feature in the H α imagesthat moves to the southwest from the flare site. It is also seen as bright features (moving blobs) in the5 DO/AIA 94
SDO/AIA 193
SDO/AIA 171
SDO/AIA 304
FMT H a : U T : U T : U T : U T : U T
15 Mm
Fig. 3.
Multiwavelength time sequences of the 2011 March 8 flare. Columns from left to right are FMT-Peru H α combined, and SDO /AIA at 304, 171, 193, and94 ˚A images, respectively. The red solid line in the bottom left panel indicates the slit position used in Figure 7.
AIA 304 and 171 ˚A images. In particular, common features are seen on the erupting blobs in the H α ,EUV 304 and 171 ˚A images. By contrast,in the AIA 193 and 94 ˚A images outer coronal loops withhigher temperature are predominant. As is often observed (Chifor et al. 2006), the erupting filamentsuffers heating at the outer edge and is therefore visible even in EUV bands such as 171 and 193 ˚A,which are sensitive to higher temperatures. Here, we have to note that the AIA channels have broadtemperature response. It is often difficult to know whether a structure is really hot or cool. After18:16 UT, the prominence becomes unclear in both H α and 304 ˚A, while the coronal loops visible in193 and 94 ˚A are still evolving and brightening.The Extreme-Ultraviolet Imager (EUVI; Wuelser et al. 2004), part of the Sun-EarthConnection Corona and Heliospheric Investigation (SECCHI; Howard et al. 2008) onboard the SolarTerrestrial Relations Observatory - Ahead ( STEREO-A ; Kaiser et al. 2008), also observed the flareand prominence eruption (filament eruption in the view of
STEREO-A ) in the EUV bands. At the timeof observation,
STEREO-A was located . ◦ ahead of the Earth on a heliocentric elliptical orbit inthe ecliptic plane and, therefore, was able to capture a top view of the flaring region (NOAA 11165).We used 171 ˚A EUVI images in which the emission line of Fe IX (0.6MK) was dominant to comparethe temporal evolution of the erupting blobs. 6he flare and prominence eruption were followed by a CME, which was first seen at 19:00 UTin the field of view of coronagraph C2 on the Large Angle and Spectrometric COronagraph (LASCO;Brueckner et al. 1995) onboard SOHO . We will discuss this CME further in Section 4.To examine the temporal evolution and dynamics of the flare and prominence eruption, weused hard X-ray (HXR) data obtained by the
Reuven Ramaty High Energy Solar Spectroscopic Imager ( RHESSI ; Lin & Rhessi Team 2002) and microwave data taken at the Sagamore Hill Solar Observatoryunder the Radio Solar Telescope Network (RSTN; Guidice & Eadon 1981). We focused in particularon 25 – 50 and 50 – 100 keV HXR emissions measured by
RHESSI and microwave emissions atthe 4.99, 8.8, and 15.4 GHz channels measured by RSTN. At these wavelengths (in this case given,respectively, as energy bands and frequencies) it is possible to observe nonthermal Bremsstrahlungand gyrosynchrotron emissions associated with accelerated electrons, allowing for the examination ofthe energetic features of accelerated electrons in the flare.
There have been some reports on the derivation of the Doppler (line-of-sight) velocities of H α disk fil-aments observed using FMT (Morimoto & Kurokawa 2003a; Morimoto & Kurokawa 2003b; Cabezaset al. 2017). The methods used in these studies are based on the ‘cloud model’, which was originallysuggested by Beckers (1964) and Mein & Mein (1988) and modified for FMT data. By contrast, inthis study we observed an off-limb bright prominence with a dark background, which meant that wecould not use the cloud-model-based methods.Instead, to derive Doppler velocities we assumed that the observed profiles were emitted froma slab with a uniform and constant source function. Based on the three wavelength data points (H α center and ± . ˚A) and an additional assumption that the H α intensities at the ± . ˚A wings are equalto zero, profile fitting produced four unknown parameters (i.e., the source function, optical depth,Doppler velocity, and Doppler width) with an error of about ± km s − . Owing to misalignment ofthe wavelengths of Peru-FMT filters from their nominal values, the actual error in determined Dopplervelocity was potentially worse at roughly ± km s − . Furthermore, FMT-Peru cannot detect blobswith velocities greater than 50 km s − . We, in this paper, fitted the erupting filament only with imagesof the H α center and the wings with the band position of ± α line witha maximum FWHM between 0.6 and 0.8 ˚A (Ruan et al. 2018). In our case the wide bandpass of thefilters around 0.5 ˚A allows us to integrate the prominence emission from the inflexion point of theline profile (0.45 ˚A) to the far wing. Therefore, the prominences are visible in our three filters, while7here is a large uncertainty in the Doppler shift values.The left panels of Figure 4 show temporal evolution maps of Doppler velocity with blue andred colors indicating blue- and red-shifted Doppler velocities, respectively. The erupting blob initiallyhas a red-shift with a velocity of about 25 km s − that changes at around 18:14 UT to a blue-shiftwith a velocity of about 30 km s − . This timing (18:14 UT) corresponds to arrow 1 in Figure 1.By assuming that the observed H α prominence is optically thin, it is possible to combine theoptical thickness and the source function into one unknown parameter – the emissivity – and reducethe number of unknown physical parameters to three. We derived the Doppler velocities by fitting thedata under this assumption and then confirmed that the resulting velocities were consistent with thosederived with the former method. The Doppler velocities in the remainder of the study were thereforederive using the former method.We derived the tangential velocity of the ejected filament (i.e., the velocity of the filament inthe plane of the sky) by applying the local correlation tracking (LCT) method to the H α combinedimage sequence. The right panels of Figure 4 show the temporal evolution maps of the tangentialvelocity, with the green arrows showing the direction and amplitude of the tangential velocity. Thetangential velocity of the erupting filament is roughly 80 km s − and rapidly decelerating. The di-rection of the tangential velocity changes from southwest to slightly northwest at around 18:14 UT,corresponding temporally to the sudden direction change of the Doppler velocity.Because the flare site was nearly on the solar limb (W88 ◦ ), STEREO-A , which was located . ◦ ahead of the Earth at the time of observation, was able to capture a top view of the flare and thefilament eruption. As discussed above, the erupting H α prominence was co-spatial with a bright blobseen in the AIA 171 ˚A images (see Figure 3), and therefore we investigated the temporal evolution ofthe filament eruption in the STEREO-A /SECCHI/EUVI 171 ˚A images. Figure 5 shows the temporalevolution of the erupting filament in the EUVI 171 ˚A images. The contours outline a bright feature(bright blob) associated with the erupting filament, with different colors indicating different times.It is seen that the filament predominantly moves southward, with a slight westward motion before18:14 UT, followed by eastward motion afterwards. These motions correspond to motion away fromand toward the Earth, respectively. The centroid of the contoured region moves at roughly 40 km s − between 18:12:15 UT (red) and 18:13:30 UT (yellow) and at 40 km s − between 18:14:45 UT (green)and 18:16:00 UT (blue).Analysis of the blob dynamics in the prominence eruption reveal an interesting behavior inwhich the blobs are deflected at specific heights within the corona, indicating an interaction betweenthe blobs and the coronal environment. In the next section, we examine the reactions consequentialto this interaction. 8 MT Doppler Velocity -360-340-320-300-280-260-240 -40-2002040 ( k m / s ) FMT Tangential Velocity -360-340-320-300-280-260-240 : : U T -360-340-320-300-280-260-240 -40-2002040 ( k m / s ) -360-340-320-300-280-260-240 : : U T -360-340-320-300-280-260-240 -40-2002040 ( k m / s ) -360-340-320-300-280-260-240 : : U T
860 880 900 920 940 960 980(arcsecs)-360-340-320-300-280-260-240 -40-2002040 ( k m / s )
860 880 900 920 940 960 980(arcsecs)-360-340-320-300-280-260-240 : : U T
100 km/s100 km/s100 km/s100 km/s
Fig. 4.
Temporal evolution of 3D velocity of erupting prominence.
Left : Doppler velocity shown with blue- and red-shifted features colored blue and red,respectively.
Right : Tangential velocity with direction and magnitude indicated with green arrows. The background images are FMT-Peru H α combinedimages. The sign of the Doppler shift changes from red to blue at around 18:14 UT; at the same time, the direction of the tangential velocity changes fromsouthwest to slightly northwest. TEREO-A EUVI 171A -100 -50 0 50 100X (arcsecs)-400-350-300-250-200 Y ( a rc s e c s ) UT18:14:45 UT18:16:00 UT
Fig. 5.
Temporal evolution of filament eruption seen in the
STEREO-A /SECCHI/EUVI 171 ˚A images. The contours show the bright feature associated with theerupting filament, with different colors indicating different times. The background is an EUVI 171 ˚A image taken at 18:11 UT . Following the disappearance of the cool H α prominence at around 18:18 UT, further activity com-mences in the higher corona, in the form of expansion of the outer coronal loops seen in the SDO/AIA94 ˚A band. This activity appears to lead to a CME. Figure 6 shows the temporal evolution of the ex-panding loops. From 18:10 UT onward, the expansion of a coronal loop is seen (Loop1). After18:19 UT, another loop (Loop2) begins to expand from just south of Loop1. Loop1 remains in placeor expand very slowly, and disappears at 18:35 UT. On the other hand, Loop2 rapidly expands until iteventually leaves the field of view.To study the temporal evolution of the vertical motion of these activities in more detail, weconstructed the time slice diagrams (time-sequenced images along slit lines) shown in the bottompanels of Figure 7. Panels (c) and (d) show time slice diagrams for the coronal loops of AIA 94 ˚A(negative images) and the H α prominence, respectively. The zero of the vertical axis on each panelcorresponds to the lower edge of the slit line (i.e., the edge closer to the solar limb). The positionsof the slits are also shown in Figure 6(i) and in the bottom left panel of Figure 3. As described inSection 3, the H α prominence erupts and decelerates after 18:14 UT prior to fading out at around18:18 UT. The vertical motion of the EUV 94 ˚A loop (Loop1) is seen to be associated with the H α prominence eruption. Loop1 also shows rapid acceleration and sudden deceleration correspondingto motions in the H α prominence eruption. The Loop1 and H α prominence ascending velocities of10 Y ( a rc s e c s ) (b) 18:12:02UT (c) 18:14:02UT(d) 18:18:02UT
880 920 960 1000-350-300-250 Y ( a rc s e c s ) (e) 18:19:14UT
880 920 960 1000 (f) 18:24:26UT
880 920 960 1000 (g) 18:28:26UT
800 850 900 950 1000 1050X (arcsecs)-450-350-250-150 Y ( a rc s e c s ) Loop2 Loop1 (h) 18:31:29UT
800 850 900 950 1000 1050X (arcsecs) (i) 18:34:41UT
800 850 900 950 1000 1050X (arcsecs) (a) 18:10:02UT
Loop1
Fig. 6.
Temporal evolution of expanding loops seen in the
SDO /AIA 94 ˚A images. Note that the field of view of panels (g) to (i) is wider that from (a) to (f). Thegray solid line in panel (i) indicates the slit position used in Figure 7. about 130 and 100 km s − , respectively. After 18:19 UT, another vertical motion associated withLoop2 is seen in the time slice diagram of the EUV 94 ˚A images. The ascending velocity in this caseis about 120 km − .Figures 7(a) and (b) show the light curves at SXRs (1.0 – 8.0 and 0.5 – 4.0 ˚A) taken by GOES ,HXRs (25 – 50 and 50 – 100 keV) taken by
RHESSI , and microwaves (4.99, 8.8, and 15.4 GHz)taken by RSTN. In the HXR and microwave light curves two bursts, with peaks at 18:12 – 18:14 andat 18:19 – 18:20 UT, are seen. The short-lived nature of these features and the simultaneity of theemissions suggest that both the microwave and HXR emissions are of nonthermal origin. The timingsof these nonthermal bursts are associated with the rapid eruptions of H α prominence and EUV coronalloops. The temporal association between the vertical motion (i.e., outward rapid acceleration) and the11 H e i g h t [ M m ] (c) AIA 94A timeslice (a) GOES X-Rays -8 -7 -6 -5 F l u x [ W m ] - R H E SS I C o un t R a t e [ s d e t ] RSTN m i cr o w a v e f l u x [ W m H z ] - - (b) RHESSI - - - (nega) H e i g h t [ M m ]
100 km/s 130 km/s 60 km/s 18:10 18:20 18:30
Time [UT]
120 km/s
Fig. 7. (a) Soft X-ray light curves in the
GOES
RHESSI at 25 – 50(green) and 50 – 100 keV (purple) and microwave flux taken with RSTN in the 4.99 (orange), 8.8 (blue), and 15.4 (red) channels. (c) Time-sequenced EUV(94 ˚A) image (time slice diagram) obtained with
SDO /AIA along the slit line shown in Figure 6(i). (d) Time slice diagram of H α combined image obtained withFMT-Peru along the slit line shown in the bottom left panel of Figure 3. The zeros of the vertical axis for both time-slice diagrams correspond to the loweredges of the slit lines. nonthermal emissions indicates that plasmoids are being ejected with the release of high amounts ofmagnetic energy as has previously been reported (e.g., Kahler et al. 1988; Ohyama & Shibata 1998;Asai et al. 2004; Asai et al. 2006; Nishizuka et al. 2010; Takasao et al. 2016). These results supportthe so-called ‘plasmoid-induced-reconnection’ model (Shibata 1999; Shibata & Tanuma 2001).Figure 8 shows spatial distributions of emission sources in HXR for the nonthermal burstsaround at 18:13 UR (Fig. 8a) and 18:19 UT (Fig. 8b). The backgrounds are AIA 94 ˚A images aretaken at 18:13:02 and 18:19:02 UT, respectively. We overlaid HXR contour images observed with RHESSI in 12 – 25 keV (red) and 40 – 80 keV (yellow). The levels of the contours are 40%, 60%,80%, and 95% of the peak intensity. We synthesized the HXR images by using grids 3 -– 8 andintegrating over 40 seconds for the first peak and 80 seconds for the second peak. We recognize a12 b) AIA 94A 2011-Mar-08 18:19:02
900 920 940 960X (arcsec)-320-300-280-260 Y ( a rc s e c ) (a) AIA 94A 2011-Mar-08 18:13:02
900 920 940 960X (arcsec)-320-300-280-260 Y ( a rc s e c ) Fig. 8.
Spatial distribution of the emission sources. The backgrounds are AIA 94 ˚A images taken at 18:13:02 (a) and 18:19:02 UT (b). The contours showthe
RHESSI
12 – 25 keV intensity (red), and the 40 -– 60 keV intensity (yellow), respectively. The contour levels are 40, 60, 80, and 95% of the peak intensity.
HXR emission source in the low energy band (12 – 25 keV) appear above the flare loops seen in the94 ˚A images in the both times. These emission sources extend in the vertical direction, i.e. along thecurrent sheet. The HXR emission in the high energy band (40 – 80 keV), on the other hand, mainlycomes from the footpoint of the flare loops. These are consistent with the previous study (Su et al.2012). We can also see a “loop-top” HXR emission source (Masuda et al. 1994) or an extension ofthe contour line in the high energy HXR band. These loop-top HXR emission sources are located justabove the top of the bright flare loops seen in the 94 ˚A, and are probably associated with the interactionbetween the reconnection outflow jet and/or plasmoid ejected from the reconnection region and theflare loops (Takasao et al. 2016).The
SOHO /LASCO C2 coronagraph detected the appearance of a faint and slow CME at19:06 UT with a linear velocity of 280 km s − (see the SOHO /LASCO CME online catalog,http://cdaw.gsfc.nasa.gov/CME list/) (Yashiro et al. 2004). From the timing and direction of expan-sion, this CME appears to be associated with the flare occurring at 18:30 UT at AR NOAA 11165.Figures 9(a–f) show LASCO C2 running difference images of the evolution of the CME, which isreferred to hereafter as CME1. After 20:12 UT, another, faster CME (CME2) with a velocity of700 km s − appears and follows nearly the same traveling path as that of CME1. From Figures 9(d–f),which show CME2, it is seen that it catches up to and interacts with CME1 in an act of ‘cannibalism’(Gopalswamy et al. 2001)). Figure 9(g) shows a time-distance diagram of both CMEs and the EUVcoronal loop (Loop2). CME2 appears to be associated with the flare starting at 19:45 UT on the sameAR. 13 D i s t a n c e [ R _ s ] + ++ ++ ++ + + + + + AIA 94A Loops GOES X-rayCME1 CME2
CME1CME2 (a)(d) (b)(e) (c)(f)(g)
Fig. 9. (a – f) Time sequences of
SOHO /LASCO C2 running difference images. Each image is overlaid with a corresponding EUV 193 ˚A running differenceimage obtained from
SDO /AIA. (g) Time-distance diagram of CMEs and EUV coronal loops. The distance is measured from the solar limb. The CME frontdistances for CME1 (CME2) at a given time are represented by black (gray) plus + signs, while the circles (cid:13) mark the position of the EUV coronal loop(Loop2). R s is the solar radius, ( ≈ GOES
In this study, our goal was to investigate the initial phases of CME-related filament eruptions andto detect signatures preceding the launch of a CME. We performed multi-wavelength analyses of thetemporal behaviors of the prominence eruption, coronal loop expansion, and CME associated with theM4.4 flare that occurred on 2011 March 8 in AR NOAA 11165. The 3D velocity field of the eruptedprominence was derived using the H α data taken by FMT-Peru. The H α prominence accelerated whenhigh amounts of magnetic energy were released, and nonthermal bursts were seen in the microwave(RSTN) and HXR ( RHESSI ) from 18:12 to 18:14 UT, after which the prominence rapidly deceleratedin the vertical direction and changed its Doppler velocity direction. Finally, the prominence faded14ut. This temporal evolution indicates that the cool material seen in the H α line interacted with theoverlying magnetic environment and failed to eject into space. In AIA 304 ˚A images we can seematerial falling from there. Therefore, the disappearance of the H α prominence is caused due todecrease of plasma density. The change of direction of the erupted prominence was also captured asa deflection of bright blobs in STEREO-A /EUVI 171 ˚A images, which represents another feature of aconfined energy release.Another energy release associated with the nonthermal emissions occurred just after the con-fined event (from 18:19 to 18:20 UT). This energy release caused an expansion of coronal loops seenin the
SDO /AIA 94 ˚A images (Loop2) and led to a faint CME (CME1). Hours later, another gradualflare and CME (CME2) occurred. Using
SDO /AIA images, Su et al. (2012) investigated in detailthe temporal evolution of the EUV coronal loops associated with the flare in question (i.e., the flareat around 18:10 UT) and the ones associated with another energy release at around 19:45 UT, andinterpret the two flares as two stages of a single associated event.By contrast, we found that the first flare comprised further two-stage energy releases andfurthermore, that the initial energy release represented by the H α prominence eruption appears tobe confined. Despite this confinement, the energy release triggered another magnetic reconnectionfollowed by the coronal loop expansion and the CME. As Kliem et al. (2020) and Gou et al. (2020)reported, even a confined flare can cause the eruption of an unstable flux rope. In our case, we couldnot confirm the formation of a flux rope as we were observing a limb flare. Nevertheless, it is possiblethat dynamic disturbances of the plasmoids and magnetic field triggered further energy release duringthe course of the event, resulting in multi-stage behavior. On the other hand, as some authors havereported, erupting filaments are often split into sections (Morimoto & Kurokawa 2003b; Tripathi etal. 2006; Guo et al. 2010; Cheng et al. 2018). The blob we observed in H α may be a part of such splitfilaments and may not be involved at all with the flux rope involved in the CME. Acknowledgments
The authors acknowledge anonymous referees for their comments and suggestions. The authors are very grateful with all the staff members of theKwasan and Hida Observatories, Kyoto University of Japan, for all the supports and discussions during the FMT-workshops and working-group meet-ings conducted in Japan and in Peru. They thank Dr. A. Hillier for his contribution to the FMT-workshops. They also are grateful to
SDO /AIAand
STEREO /EUVI teams for providing high quality data used in this study. This work was also supported by the international program “ClimateAnd Weather of the Sun-Earth System - II (CAWSES-II) : Towards Solar Maximum” sponsored by SCOSTEP. This work was also supported by the“UCHUGAKU” project of the Unit of Synergetic Studies for Space, Kyoto University. This work was also supported by JSPS KAKENHI GrantNumbers 25287039, 26400235, 15K17772, and 16H03955. A.A. was supported by a Shiseido Female Researcher Science Grant. The authors wouldlike to thank Enago for the English language review. eferences Asai, A., Yokoyama, T., Shimojo, M., & Shibata, K. 2004, ApJL, 605, L77–L80Asai, A., Nakajima, H., Shimojo, M., White S. M., Hudson, H. S., Lin, R. P. 2006, PASJ, 58, L1–L5Asai, A., Ishii, T. T., Isobe, H., et al.. 2012, ApJL, 745, L18Beckers, J. M. 1964, PhD Thesis, Univ. UtrechtBrueckner, G. E., Howard, R. A., Koomen, M. J., et al. 1995, Sol. Phys., 162, 357Cabezas, D. P., Mart´ınez, L. M., Buleje, Y. J., et al. 2017, ApJ, 836, id.33Cabezas, D. P., Asai, A., Ichimoto, K., et al. 2019, ApJ, 883, id.32Chen, P. F., & Shibata, K. 2000, ApJ, 545, 524Cheng, X., Kliem, B., Ding, M. D. 2018, ApJ, 856, id.48Chifor, C., Mason, H. E., Tripathi, D., et al. 2006, A&A, 458, 965–973D´emoulin, P., and Aulanier, G.. 2010, ApJ, 718, 1388–1399Eto, S., Isobe, H., Narukage, N., et al. 2002, PASJ, 54, 481–491Gopalswamy, N., Yashiro, S., Kaiser, M. L., et al. 2001, ApJL, 548, L91Gopalswamy, N., Shimojo, M., Lu, W., et al. 2003, ApJ, 586, 562–578Gopalswamy, N., Miki´c, Z., Maia, D., et al. 2006, Space Sci. Rev., 123, 303Guidice, D. A., & Eadon, E. J. 1981, BAAS, 13, 400Gou, T., Liu, R., Kliem, B., et al. 2019, Sci. Adv., 5, eaau7004Guo, Y., Ding, M. D., Schmieder, B., et al. 2010, ApJL, 725, L38–L42Howard, R. A., Moses, J. D., Vourlidas, A., et al. 2008, Space Sci. Rev., 136, 67Kahler, S. W., Moore, R. L., Kane, S. R., & Zirin, H. 1988, ApJ, 328, 824–829Kaiser, M. L., Kucera, T. A., Davila, J. M., et al. 2008, Space Sci. Rev., 163, 5Kliem, B., and T¨or¨ok, T. 2006, Phys. Rev. Lett., 96, 255002Kliem, B., Lee, J., White, S. M., et al. 2020, ApJ, submittedKurokawa, H., Kimura, G., Nakai, Y., Kitai, R., Funakoshi, Y., Shinkawa, T. 1995, J. Geomag. Geoelectr., 47,1043-1052Lemen, J. R., Title, A. M., Akin, D. J., et al. 2012, Sol. Phys., 275, 17Lin, R. P., & Rhessi Team 2002, Solar Variability: From Core to Outer Frontiers, 506, 1035Mackay, D. H., Karpen, J. T., Ballester, J. L., et al. 2010 Space Sci. Rev., 151, 333–399Masuda, S., Kosugi, T., Hara, H., Tsuneta, S., Ogawara, Y. 1994, Nature, 371, 495McAllister, A. H., Dryer, M., McIntosh, P., Singer, H., Weiss, L. 1996, J. Geophys. Res., 101, 13497-13516Mein, P., & Mein, N. 1988, A&A, 203, 162Moore, R. L., Sterling, A. C., Hudson, H. S., Lemen, J. R. 2001, ApJ, 552, 833-848 orimoto, T., & Kurokawa, H. 2003a, PASJ, 55, 503-508Morimoto, T., & Kurokawa, H. 2003b, PASJ, 55, 1141-1151Munro, R. H., Gosling, J. T., Hildner, E., et al. 1979, Sol. Phys., 61, 201Narukage, N., Hudson, H. S., Morimoto, T., et al. 2002, ApJL, 572, L109-L112Narukage, N., Ishii, T. T., Nagata, S., et al. 2008, ApJL, 684, id.L45Nishizuka, N., Takasaki, H., Asai, A., & Shibata, K. 2010, ApJ, 711, 1062–1072Ohyama, M., & Shibata, K. 1998, ApJ, 499, 934–944Pesnell, W. D., Thompson, B. J., & Chamberlin, P. C. 2012, Sol. Phys., 275, 3Ruan, G., Schmieder, B., Mein, P., et al. 2018, ApJ, 865, id.123Schmieder, B., D´emoulin, P., Aulanier, G. 2013, Adv. Space Res., 51, 1967–1980Schmieder, B., Aulanier, G., Vrˇsnak, B. 2015, Sol. Phys., 290, 3457–3486Schmieder, B., Zapi´or, M., L´opez Ariste, A., et al. 2017, A&A, 606, id.A30Schmieder, B., Kim, R.-S., Grison, B. et al. 2020, J. Geophys. Res., 125, A027529Shibata, K. 1999, Ap&SS, 264, 129Shibata, K., & Tanuma, S. 2001, Earth Planets Space, 53, 473Takasao, S., Asai, A., Isobe, H., & Shibata, K. 2016, ApJ, 828, id.103Tripathi, D., Isobe, H., Mason, H. E. 2006, A&A, 453, 1111–1116Su, Y., Dennis, B. R., Holman, G. D., et al. 2012, ApJL, 746, L5UeNo, S., Shibata, K., Kimura, G., et al. 2007, BASI, 35, 697Wuelser, J.-P., Lemen, J. R., Tarbell, T. D., et al. 2004, Proc. SPIE, 5171, 111Yashiro, S., Gopalswamy, N., Michalek, G., et al. 2004, J. Geophys. Res., 109, A07105orimoto, T., & Kurokawa, H. 2003a, PASJ, 55, 503-508Morimoto, T., & Kurokawa, H. 2003b, PASJ, 55, 1141-1151Munro, R. H., Gosling, J. T., Hildner, E., et al. 1979, Sol. Phys., 61, 201Narukage, N., Hudson, H. S., Morimoto, T., et al. 2002, ApJL, 572, L109-L112Narukage, N., Ishii, T. T., Nagata, S., et al. 2008, ApJL, 684, id.L45Nishizuka, N., Takasaki, H., Asai, A., & Shibata, K. 2010, ApJ, 711, 1062–1072Ohyama, M., & Shibata, K. 1998, ApJ, 499, 934–944Pesnell, W. D., Thompson, B. J., & Chamberlin, P. C. 2012, Sol. Phys., 275, 3Ruan, G., Schmieder, B., Mein, P., et al. 2018, ApJ, 865, id.123Schmieder, B., D´emoulin, P., Aulanier, G. 2013, Adv. Space Res., 51, 1967–1980Schmieder, B., Aulanier, G., Vrˇsnak, B. 2015, Sol. Phys., 290, 3457–3486Schmieder, B., Zapi´or, M., L´opez Ariste, A., et al. 2017, A&A, 606, id.A30Schmieder, B., Kim, R.-S., Grison, B. et al. 2020, J. Geophys. Res., 125, A027529Shibata, K. 1999, Ap&SS, 264, 129Shibata, K., & Tanuma, S. 2001, Earth Planets Space, 53, 473Takasao, S., Asai, A., Isobe, H., & Shibata, K. 2016, ApJ, 828, id.103Tripathi, D., Isobe, H., Mason, H. E. 2006, A&A, 453, 1111–1116Su, Y., Dennis, B. R., Holman, G. D., et al. 2012, ApJL, 746, L5UeNo, S., Shibata, K., Kimura, G., et al. 2007, BASI, 35, 697Wuelser, J.-P., Lemen, J. R., Tarbell, T. D., et al. 2004, Proc. SPIE, 5171, 111Yashiro, S., Gopalswamy, N., Michalek, G., et al. 2004, J. Geophys. Res., 109, A07105