Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.tesis.lebedev.ru/docs/1581.pdf Äàòà èçìåíåíèÿ: Thu Nov 26 16:46:48 2015 Äàòà èíäåêñèðîâàíèÿ: Sat Apr 9 23:49:33 2016 Êîäèðîâêà: Windows-1251 Ïîèñêîâûå ñëîâà: earth's atmosphere

Draft version September 28, 2015
A Preprint typ eset using L TEX style emulateap j v. 01/23/15

INITIATION AND EARLY EVOLUTION OF THE CORONAL MASS EJECTION ON MAY 13, 2009 FROM EUV AND WHITE-LIGHT OBSERVATIONS
Reva A.A., Ulyanov A.S., Bogachev S.V., Kuzin S.V.
Lebedev Physical Institute, Russian Academy of Sciences Draft version September 28, 2015

arXiv:1509.07713v1 [astro-ph.SR] 25 Sep 2015

Abstract We present the results of the observations of a coronal mass ejection (CME), which o ccurred on May 13, 2009. The most important feature of these observations is that the CME was observed from the very early stage (the solar surface) up to a distance of 15 solar radii (R ). Below 2 R , we used the data from the TESIS EUV telescopes obtained in the Fe 171 œ and He 304 œ lines, and above 2 R , A A we used the observations of the LASCO C2 and C3 coronagraphs. The CME was formed at a distance of 0.2 0.5 R from the Sun's surface as a U-shaped structure, which was observed both in the 171 œ A images and in white-light. Observations in the He 304 œ line showed that the CME was asso ciated A with an erupting prominence, which was lo cated not above as predicts the standard mo del but in the lowest part of the U-shaped structure close to the magnetic X-point. The prominence lo cation can be explained with the CME breakout mo del. Estimates showed that CME mass increased with time. The CME tra jectory was curved its helio-latitude decreased with time. The CME started at latitude of 50 and reached the ecliptic plane at distances of 2.5 R . The CME kinematics can be divided into three phases: initial acceleration, main acceleration, and propagation with constant velo city. After the CME onset GOES registered a sub-A-class flare. Keywords: Sun:corona Sun: coronal mass ejections (CMEs)
1. INTRODUCTION

Coronal Mass Ejections (CME) are ejections of coronal plasma into the interplanetary space. CME mass lies in the interval 1012 - 1016 g (Vourlidas et al. 2010), and CME size could be larger than solar radii (R ). CMEs o ccur due to the large-scale pro cesses of energy release, and that is why CME investigations are important for solar physics. CME investigations are also critically important for solar-terrestrial issues because CMEs, which reach the Earth, influence Earth's magnetosphere and satellite operation (Schwenn 2006; Pulkkinen 2007). Approximately 30% of CMEs have a three-part structure: bright core, dark cavity, and bright frontal lo op (Illing & Hundhausen 1985; Webb & Hundhausen 1987). Perhaps more CMEs have such structure, but due to pro jection effects, it is unobservable. It is considered that the bright core is the prominence, which erupted with the CME (House et al. 1981); the dark cavity is the less dense plasma surrounding the prominence; and the bright frontal lo op is the coronal plasma, which is amassed during the CME motion (Forbes 2000). Primary CME acceleration o ccurs at distances of 1.4- 4.5 R (Zhang et al. 2001, 2004; Gallagher et al. 2003). At distances higher than 5 R , CMEs can accelerate, decelerate, or move with constant velo city. In general, CMEs with velo city greater than 400 km s-1 decelerate, and with less than 400 km s-1 accelerate. It is considered that fast CMEs move faster than solar wind, and the wind decelerates them; and slow CMEs move slower than solar wind, and the wind accelerates them (Yashiro et al. 2004). CME velo city lies in the range 20-3500 km s-1 (Gopalswamy et al. 2010). Non-impulsive CME acceleration lies in the range -60-+40 m s2 (Yashiro et al. 2004),
reva.antoine@gmail.com

and impulsive acceleration is of an order of 500 m s-2 (Zhang et al. 2001). CME behavior depends on the phase of the solar cycle. In the solar maximum CMEs o ccur more often than in the minimum (Robbrecht et al. 2009). In the maximum, average CME velo city is about 500 km s-1 and in the minimum 250 km s-1 . Moreover, in the maximum, CMEs are observed at all latitudes, and in the minimum, they are close to the ecliptic plane (Yashiro et al. 2004). There is no clear connection between CMEs and other solar phenomena. Some CMEs are asso ciated with flares, some with erupting prominences, some with both flares and erupting prominences (Webb & Howard 2012), and some CMEs, known as "stealth CMEs", are not asso ciated with any visible changes on the solar surface (Ma et al. 2010). For the first time, CMEs were observed with whitelight coronagraphs on board the satellite OSO-7 in 1971 (Tousey 1973). Early CME investigations were carried out with white-light coronagraphs on board Skylab (1973-1974, Munro et al. (1979)), P78-1/Solwind (19791985, Sheeley et al. (1980)), and Solar Maximum Mission (1980; 1984-1989; Hundhausen (1999)). The mo dern era of CME investigations started with the launch of the SOHO satellite (1995; Domingo et al. (1995)). SOHO carries white-light coronagraphs LASCO/C1, C2, C3, which take images of the solar corona at distances 1.1-30 R (C1: 1.1-3 R ; C2: 2- 6 R ; C3: 4-30 R ). Coronagraph C1 operated from 1995 till 1998, and coronagraphs C2 and C3 still work. LASCO coronagraphs have obtained large volumes of data, and most of the knowledge about CME behavior we have thanks to these instruments. The next step in experimental CME investigation was the launch of STEREO satellites (2006; Howard et al.


2

A. Reva, A. Ulyanov et al. blue stripes to the north from the Sun (see Figure 1a,b). 2. We co-aligned the images using correlation techniques. 3. We scaled the intensities of the 0.1 s and 3.1 s images to the intensity of the 31 s image. The scaling co efficient is the same for all pixels on the same image, but different for the 0.1 s and 3.1 s images. 4. We made a composite image out of these three images. If pixel is not saturated on the 31 s image, we pick its intensity as the composite image intensity. If pixel is saturated on the 31 s image, but not saturated on the 3.1 s image, we pick 3.1 s image intensity. In other cases we pick 0.1 s image intensity (see Table 1). 5. We removed the trace. During the read-out the CCD-matrix was exposed to the Sun, and this additional exposure generated trace. The trace is a linear function of the signal from the Sun, and we can calculate it analytically. The trace removal algorithm is described in Kuzin et al. (2011a). The result is shown in Figure 1d. We must stress that this metho d do esn't recover the far corona information for all pixels on the composite image. The pixels on the 31 s image to the north and the south form the Sun are saturated, and we used the intensity from the 3.1 s image for these regions. As a consequence, the noise dominates over signal at closer distances from the Sun in the north and the south regions than in the west and east ones. We were lucky that the CME, analyzed in this work, o ccurred at the eastern edge of the image. The intensity of the far corona is significantly lower than the solar disk intensity. In order to show all layers of the solar corona on one image, we need to artificially increase the intensity of far corona. In this work, we have applied a radial filter for Fe 171 œ images A we have multiplied the signal by a function, that depends on the distance from the Sun's center. Inside the solar disk, this function equals unity. Outside the solar disk, we picked the function values to make the far corona intensity to be the same order of magnitude as the solar disk intensity. The radial filter function is shown in Figure 2. The image after applying the radial filter is shown in Figure 3. In this work, we investigate a CME, which o ccurred on May 13, 2009 at 00:10 UT close to the solar limb. The CME was observed with the TESIS Fe 171 œ telescope in A the far corona observation mo de. Below 2 R , we used the data from TESIS EUV telescopes obtained in the Fe œ 171 A and He 304 œ lines, and above 2 R , we used the A data from LASCO C2 and C3 coronagraphs. We made a movie of these observations, and we strongly recommend the reader to watch it before pro ceeding further (see Figure 4). STEREO coronographs also observed this CME. The main fo cus of our work is EUV observations of the CME in the low corona. LASCO better complements TESIS observations than STEREO, because LASCO has the same viewpoint as TESIS. That is why we do not use STEREO data in our research.

(2008)). STEREO satellites move along the Earth's orbit: one ahead of the Earth (STEREO-A), and the other in the opposite direction (STEREO-B ). Both STEREO satellites carry white-light coronagraps: COR1 and COR2. COR1 builds solar corona images at distances 1.3 - 4 R and COR2 at distances 2 - 15 R . LASCO and STEREO coronagraphs observe the Sun from three different viewpoints, allowing reconstruction of three-dimensional CME tra jectory and structure. Furthermore, STEREO coronagraphs can see CMEs, that are not seen with LASCO. More details about CME observations can be read in the review Webb & Howard (2012). To day, the CME behavior is studied best at distances greater than 2 R from the Sun's center. There are a small number of observations in the distance range 1.2- 2 R , and this distance range is studied less. However, this distance range is the most interesting because it is the range where CME forms, accelerates, and evolves. In this work, we present results of the observations of a CME, which o ccurred on May 13, 2009. The most important feature of these observations is that the CME was observed from the very early stage (the solar surface) up to the distance of 15 R . Below 2 R , we used the data from TESIS EUV telescopes obtained in Fe 171 œ A and He 304 œ lines (Kuzin et al. 2011b), and above 2 R , A we used the data from LASCO C2 and C3 coronagraphs. The aim of this work is to describe the formation, early evolution, and propagation of this CME.
2. EXPERIMENTAL DATA

TESIS is an instrument assembly employed for solar corona investigations in the extreme ultra-violet (EUV) and soft X-ray wavelength ranges. TESIS was launched on board the CORONAS-PHOTON satellite in 2009 (Kotov 2011). TESIS included EUV telescopes that built œ images in Fe 171 œ and He 304 A lines. The telescopes A had 1 field of view and 1.7 angular resolution. In the synoptic observation mo de, the telescope's cadence was 10-15 minutes. The main difficulties in the construction of far corona EUV images are small field of view and limited dynamic range of the telescopes. The field of view of TESIS EUV telescopes reached distances of 2 R from the Sun's center, but their dynamic range was not high enough to build an image of the far corona. On images with long exposure, the far corona is visible, but the solar disk is overexposed. Conversely, on images with short exposure, the solar disk is not overexposed, but the far corona is to o faint. In order to see the far corona with the TESIS Fe 171 œ A telescope, we developed special observational program. During this program, the telescope to ok images with different exposure times: short (0.1 s), medium (3.1 s), and long (31 s) (see Figure 1a,b,c). As you can see a lot of the pixels on the 31 s image are saturated. The idea of our metho d is to substitute saturated pixels on the 31 s image with non-saturated pixels from the 0.1 s or 3.1 s images. We do this with the following algorithm: 1. For each image we performed an initial pro cessing: removing of hot pixels, background, flat-field adjustment, etc. For more details see Kuzin et al. (2011a). At this step we do not remove "the trace"


Initiation and early evolution of the CME from EUV and white-light observations

3

a)

b)

0.1 s c)

3.1 s d)

31 s
œ Figure 1. Processing of the TESIS 171 A images. a, b, c raw TESIS images with different exposure times (a 0.1 s, b 3.1 s, c 31 s). d Composite image. We used logarithmic scale for all four images in this figure.

Table 1 Formation of the composite image. Pixel on 0.1 s image Saturated Not saturated Not saturated Not saturated Pixel on 3.1 s image Saturated Saturated Not saturated Not saturated Pixel on 31 s image Saturated Saturated Saturated Not saturated Comp osi 0.1 0.1 3.1 31 te s s s s image pixel image image image image

3. RESULTS 3.1. General Properties The observed CME is relatively slow it takes it 15 hours to travel from the solar surface to the LASCO/C2 field of view. Its angular size is 15 . The CME started at 50 latitude and then moved along a curved tra jectory to the ecliptic plane. The observed CME do es not have the common threepart structure (bright core, dark cavity, bright frontal lo op). In all stages of its evolution, the CME had a U-shaped structure, which formed at the height of 0.2- 0.5 R from the solar surface. The CME's U-shaped structure extended to the LASCO field of view (see Figure 4a), which means that the CME had an open magnetic field structure. In the early stages of the evolution, the CME was located at high latitudes, and its `leg' the magnetic field

line that connects the CME to the Sun connected with the solar surface at high latitudes (see Figure 4a). In the late stages of the evolution, the CME reached the ecliptic plane, and its `leg' connected with the solar surface at the equator (see Figure 4b,c). Somewhere at distances 1.8 - 2.3 R from the Sun's center the CME changed its `leg'. Unfortunately, at this distance range, TESIS images are to o blurry, and we can not distinguish this pro cess. When the CME passes from TESIS Fe 171 œ to A LASCO C2 field of view, we see complete correspondence between the bright structures on the TESIS and LASCO images (see Figure 4b,c). Since bright structures on LASCO images are density structure, then the CME on TESIS Fe 171 œ images is also a density structure. A After the CME started, a flare o ccurred at 00:30 UT (see Figure 5). GOES flux in the 0.5-4 œ channel did not A A exceed A-level, the flux in the 1-8 œ channel was below


4

A. Reva, A. Ulyanov et al.

Figure 2. Radial filter function.

Figure 3. CME in the low corona, which occurred on May 13, 2009. Left TESIS He 304 œ image. A

TESIS Fe 171 œ image with applied radial filter; right A

the sensitivity threshold. We cannot prove that this flare is related to the CME, so the flare could o ccur in some other remote place on the Sun. 3.2. Prominence The CME was asso ciated with an erupting prominence, which was observed in TESIS He 304 œ images. The A prominence was below the U-shaped structure, close to the magnetic X-point (see Figure 6). Just before the eruption, the prominence broke into two parts: one fell on the Sun; the other left the Sun with the CME (see Figure 7). The prominence was observed up to the height of 0.4 R above the solar surface, where it faded away. 3.3. CME mass In order to estimate the CME mass, we measured CME intensity and area on TESIS Fe 171 œ images. The inA tensity (I ) decreased by 3.6 times, and the area (A) increased by 6.5 times (see Figure 8).

The CME intensity on the TESIS Fe 171 œ images A equals I= G(T )DE M (T ) dT , (1)

where G(T ) the temperature response of the TESIS Fe 171 œ telescope, and DE M (T ) the CME differenA tial emission measure. For an estimation, we assume that the CME was isothermal and its temperature was constant. In this ca s e, I = G(T0 )E M n2 V = ( e where T0 the sion measure; CME volume; The CME mas trons N : N2 N2 )V= , (2) V V CME temperature; E M the CME emisne the CME electron density; V the N the number of electrons in the CME. s M is proportional to the number of elec (3) M N IV


Initiation and early evolution of the CME from EUV and white-light observations

5

a)

b)

c)

d)

Figure 4. Composite LASCO C2 (red) and TESIS Fe 171 œ (blue) images. A

10

-7

GOES 1.0 - 8.0 A GOES 0.5 - 4.0 A

Flux, Watt m

-2

10

-8

A level

10

-9

May 12 20:00

May 13 00:00

May 13 04:00

May 13 08:00 Time, UT

May 13 12:00

May 13 16:00

May 13 20:00

Figure 5. GOES flux. Red: the GOES 1.0-8.0 œ channel; blue: the GOES 0.5-4.0 œ channel. A A

Let us consider two cases of the relationship between CME volume and area: 1. V A the CME did not expand in the direction perpendicular to the image plane; 2. V A rections.
3/2

In the first (two-dimensional) case: M2D I V I A

(4)

the CME expanded evenly in all di-

For each of these cases, we calculated the CME mass using measured values of I and A. In the two-

In the second (three-dimensional) case: M3D I V I A3/2

(5)


6

A. Reva, A. Ulyanov et al.

prominence

œ Figure 6. Prominence location relative to U-shape structure. Left: TESIS Fe 171 A images with applied radial filter. The prominence is overlayed with contours. Right: TESIS He 304 œ images. We put an artificial occulting disk on He 304 œ image in order to improve the A A prominence visibility.

dimensional case the CME mass increased by 30 %, in the three-dimensional case by 150 % (see Figure 9). In reality, we deal with an intermediate case, and the real CME mass dynamics is somewhere between these two cases. Since in both cases the mass increased, the real CME mass should increased by 30-150 %. 3.4. Kinematics We measured the co ordinates of the CME's frontal bright edge on TESIS and LASCO images. For LASCO C3 images, we measured co ordinates until the CME could not be distinguished from the background. The CME had curved tra jectory its helio-latitude decreased with time (see Figure 10). The CME originated at 50 latitude and reached ecliptic plane at 2.5 R . Figure 11 shows the dependencies of the CME's distance from the Sun's center, velo city, and acceleration on time. The CME kinematics could be divided into three phases: initial acceleration, main acceleration, and steady movement. During the first phase, the CME had an acceleration 0.5 m s-2 , velo city 20 km s-1 , and reached the height of 1 R above the solar surface within 14 hours. During the second phase, the CME acceleration increased to 5 m s-2 , and the velo city increased from 20 to 170 km s-1 . The second phase to ok place at distances 2-6 R over 12 hours. During the third phase, the CME moved with a constant velo city of 170 km s-1 .
4. DISCUSSION

the prominence and forms the observed in the 171 œ line A U-shaped structure. So, in the standard CME mo del, the prominence is lo cated above the U-shaped structure. This contradicts the presented observations, in which the prominence is lo cated in the lowest part of the U-shaped structure close to the magnetic X-point. Although the observations do not fit into the standard mo del, they fit into one of its mo dification breakout mo del (Antio chos et al. 1999). In the breakout mo del the standard CME magnetic configuration is confined by the quadrupolar magnetic structure (see Figure 13). As the CME moves up, the overlying closed field lines reconnect and free the way for the CME. This mo del A fits the observations, if we interpret the 171 œ U-shaped structure as the field lines, which lie above the standard CME magnetic structure (see Figure 13). In this case the prominence will be below the U-shaped structure. 4.2. Scattered Emission For the estimation of the CME mass, we assumed that the CME emission in the Fe 171 œ is thermal (emitting A ions are excited by collisions with thermal electrons). We neglected the contribution of the scattered emission because it is small in the low corona. But, as the CME go es away from the Sun, its electron density decreases, and the contribution of the scattered emission increases. At some distance, the scattered and thermal emissions could equalize, and the conclusion about the CME mass dynamics could be incorrect. In this section, we will estimate the contribution of the scattered emission to the CME intensity in the Fe 171 œ line and its impact on the A estimate of the CME mass. In the coronal approximation, the number of the emitted photons in Fe 171 œ line equals the number of the A electrons excited to the upper level of the corresponding transition. The number of excitations due to the collision with thermal electrons is dNth =N dt
IX Fe

4.1. Prominence According to the standard CME mo del, the CME has structure, which is shown in Figure 12 (Carmichael 1964; Sturro ck 1966; Hirayama 1974; Kopp & Pneuman 1976). The prominence should be lo cated above the current sheet and be surrounded by circular magnetic field lines. Under the prominence in the current sheet the reconnection happens. Plasma outflow from the reconnection region pushes the prominence up, and the CME erupts. This outflow also fills magnetic field lines under

ne Ccoll ,

(6)


Initiation and early evolution of the CME from EUV and white-light observations

7

Figure 7. Prominence evolution. The coordinates are measured in arc seconds. We put an artificial occulting disk on the He 304 œ A images in order to improve the prominence visibility.
I where NFX the number of Fe IX ions; ne electron e density; Ccoll collision strength of the transition. According to CHIANTI database (Dere et al. 1997) for the transition corresponding to Fe 171 œ line (3s2 3p6 1 S0 A 3s2 3p5 3d 1 P1 ), Ccoll = 1.36 ž 10-8 cm3 s-1 . The number of excitations due to the resonance scattering is

dNsc dt

att

=N

IX Fe

F,

(7)

where resonance scattering cross section; F photons flux in the Fe 171 œ line. A We have calculated F with the CHIANTI package: F = 1013 photons s-1 cm-2 . The F decreases by a factor of 4 if the distance from the Sun's center increases from 1 to 2 R . For an estimation, we used the constant value of F since the change in the electron density is much higher. We calculated with the formula (Sylwester et al.


8
1ž10
8

A. Reva, A. Ulyanov et al.

8ž10 Intensity, DN s-1

7

6ž10

7

4ž10

7

2ž10

7

0
May 12 22:00 May 13 00:00 May 13 02:00 May 13 04:00 May 13 06:00 Time, UT May 13 08:00 May 13 10:00 May 13 12:00

2.0ž105

1.5ž105 Area, Mm2

1.0ž105

5.0ž104

0
May 12 22:00 May 13 00:00 May 13 02:00 May 13 04:00 May 13 06:00 Time, UT May 13 08:00 May 13 10:00 May 13 12:00

Figure 8. Top panel: the CME intensity on TESIS Fe 171 œ images. Bottom panel: CME area on TESIS Fe 171 œ images. A A

1986), 2 e = f me c MF e = 5 ž 10 2kB T

one (Ith ) equals dNscatt /dt F Iscatt = 3 ž 10 = Ith dNth /dt Ccoll ne
-3

(9)

-16

cm

-2

(8)

where e electron charge; me electron mass; c speed of light; f = 2.94 transition oscillator strength; = 171 œ transition wavelength; kB Boltzmann conA stant; T = 1 MK plasma temperature; MF e ion Fe IX mass. For low corona electron density ne = 108 cm-3 , the ratio of the scattered emission (Iscatt ) to the thermal

This means that in the low corona, the CME emission is indeed thermal. Let us estimate at what electron density the thermal and scattered emissions will equalize: ne F 3 ž 105 cm- Ccoll
3

(10)

According to Vourlidas et al. (2010), CMEs have such electron densities at a distance of 2 R from the Sun's


Initiation and early evolution of the CME from EUV and white-light observations

9

1ž106 8ž105

M2D, a.u. M3D, a.u.

6ž105

4ž105

2ž105 0
May 12 22:00 May 13 00:00 May 13 02:00 May 13 04:00 May 13 06:00 May 13 08:00 May 13 10:00 May 13 12:00 Time, UT

1.2ž10 1.0ž10 8.0ž10 6.0ž10 4.0ž10 2.0ž10

7

7

6

6

6

6

0
May 12 22:00 May 13 00:00 May 13 02:00 May 13 04:00 May 13 06:00 May 13 08:00 May 13 10:00 May 13 12:00 Time, UT
Figure 9. Top panel: the CME mass in two-dimensional case; bottom panel: the CME mass in three-dimensional case.

center. This means that our assumption about the thermal nature of the CME emission in the Fe 171 œ line A is violated only at the boundaries of the TESIS field of view, which do es not influence our conclusion about the CME mass dynamics. 4.3. CME Mass We considered two cases of CME geometry behavior: two-dimensional and three-dimensional. As previously mentioned, in reality we deal with some intermediate case, and the real mass dynamics is somewhere between these two cases. Since in both cases, the CME mass increased with time, we conclude that the real CME mass

also increased with time. This implies that the CME beside the low corona plasma to ok away the plasma from the far corona. However, the effect could be explained not only by the mass increase by accumulation of the far corona's plasma. œ First, the observed in the Fe 171 A line mass increase could be caused by the prominence heating. In this case, total CME mass is constant, but the CME mass is redistributing among hot and co ol temperatures. This explanation do es not contradict the observations: the prominence erupted, the prominence eventually faded away, and the prominence disappearance could o ccur due to its heating.


10

A. Reva, A. Ulyanov et al.

Figure 10. CME tra jectory. Left: LASCO C2 image; right: TESIS Fe 171 œ image. Green crosses designate CME's position at different A times.

Second, we assumed that the CME temperature was constant. If initially the CME temperature was 1 MK or higher, and the CME temperature decreased with time, then the temperature would have moved towards the temperature response maximum of the TESIS Fe 171 œ A telescope (0.7 MK). This would have lead to an increase in CME intensity and therefore to higher values of the CME mass estimation. Third, we did not calculate the CME mass; we estimated it. The CME mass increased by 30 - 150 %; and therefore, the effect could be caused by the roughness of the mo del. So, although the estimations tell us that the CME mass increased with time, the effect could be caused by prominence heating, CME co oling, or roughness of the mo del. 4.4. Kinematics The main CME acceleration o ccurred at distances 2- 6 R from the Sun's center. This is consistent with other results on the CME kinematics, according to which the main acceleration o ccurs at distances 1.4-4.5 R (Zhang et al. 2001, 2004; Gallagher et al. 2003). After the CME erupted, the flare below A-level o ccurred. The lag of the flare behind the CME is common (Zhang et al. 2001). During the CME main acceleration phase, the flux in GOES 0.5-4 œ slightly increased. We A cannot confirm whether this flux is related to the CME or not. If it is related to the CME, then it could be a sign of some reconnection pro cesses taking place during the CME acceleration. In the solar minimum, LASCO C2 observes CMEs in the ecliptic plane (Yashiro et al. 2004). However, erupting prominences asso ciated with the CMEs in the solar minimum have bi-mo dal latitude distribution with maximums at the latitudes of +30 (Plunkett et al. 2002). It is considered, that this difference in the latitude distributions is caused by the presence of complicated multi-polar structures, which make the CME move along a curved tra jectory (Webb & Howard 2012; Antio chos et al. 1999). The results presented in this work show that CMEs indeed could move along a curved tra jectory and that the high-latitude erupting promi-

nence and the low-latitude LASCO CMEs could be connected with curved tra jectory.
5. CONCLUSION In this work, we presented results of observations of the CME, which o ccurred on May 13, 2009. The observations were carried out with TESIS EUV telescopes and LASCO coronagraphs. We observed the CME from the solar surface up to 15 R . The main results of the work are the prominence lo cation, the CME mass dynamics, and the detailly measured kinematics. The CME was asso ciated with an erupting prominence. We proved that the prominence was lo cated in the lowest part of the U-shaped structure (observed in the Fe 171 œ A line) close to the magnetic X-point. This fact do es not fit into the standard CME mo del, which expects the prominence to be above the U-shaped structure. We explained the observations with the CME breakout mo del. We interpreted the U-shaped structure as closed field lines, which lie above the standard CME magnetic structure. We have estimated the CME mass for two different behaviors of CME geometry. The CME mass had increased during its motion. Although the effect could be explained by other reasons the prominence heating, the CME co oling, or the roughness of the mo del we think that question of the CME mass dynamics deserves more attention. Thanks to the large field of view of TESIS EUV telescopes, we observed the CME from the very beginning up to 15 R without losing it from sight. Due to the relatively low CME velo city, we have a lot of the CME images. These two factors allowed us to measure the CME kinematics in detail. The CME kinematics is consistent with literature: three acceleration phases; distance range of the acceleration; the lag of the flare behind the CME. Moreover, we found out that the CME moved along a curved tra jectory. Observation of the curved tra jectory confirms the hypothesis of connection between highlatitude erupting prominences, observed with EIT 304 œ, A and low-latitudes CMEs, observed with LASCO C2. The observations presented in this work show that EUV telescopes with wide fields of view are very effective


Initiation and early evolution of the CME from EUV and white-light observations
12

11

10

TESIS LASCO C2 LASCO C3

8
Sun

r, R

6

4

2

0 May 12 20:00

May 13 00:00

May 13 04:00

May 13 08:00

May 13 12:00 Time, UT

May 13 16:00

May 13 20:00

May 14 00:00

May 14 04:00

200

150

v, km s-1

100

50

0 May 12 20:00

May 13 00:00

May 13 04:00

May 13 08:00

May 13 12:00 Time, UT

May 13 16:00

May 13 20:00

May 14 00:00

May 14 04:00

8

6

GOES 1.0-8.0 A GOES 0.5-4.0 A

4 a, m s 2 0 -2 May 12 20:00
-2

May 13 00:00

May 13 04:00

May 13 08:00

May 13 12:00 Time, UT

May 13 16:00

May 13 20:00

May 14 00:00

May 14 04:00

Figure 11. CME kinematics: r : distance to the Sun's center; v: radial CME velocity, a: radial CME acceleration; RS On the acceleration plot (bottom panel), GOES flux is over plotted.

un

: solar radius.

in CME investigations. Thanks to the wide field of view, we can observe CMEs in the far corona, and thanks to the ability to observe in different wavelengths (for example in Fe 171 œ and He 304 œ lines), we can study how A A CME plasma with different temperatures interact with each other. We hope that in the future more "EUV coronagraphs" will be created and that they will shed some light on the nature of CMEs.

We are grateful to Ivan Lobo da for his invaluable help. This work was supported by a grant from the Russian Foundation of Basic Research (grant 14-02-00945), and by the Program No. 22 for fundamental research of the Presidium of the Russian Academy of Sciences. REFERENCES
Antiochos, S. K., DeVore, C. R., & Klimchuk, J. A. 1999, ApJ, 510, 485 Carmichael, H. 1964, NASA Special Publication, 50, 451


12
U-shaped structure observed in 171 A

A. Reva, A. Ulyanov et al.
where prominence should be

current sheet

prominence

solar surface
Figure 12. CME standard model and the real prominence location.

U-shaped structure observed in 171 A

prominence

solar surface
Figure 13. CME breakout model and the observations. Dere, K. P., Landi, E., Mason, H. E., Monsignori Fossi, B. C., & Young, P. R. 1997, A&AS, 125, 149 Domingo, V., Fleck, B., & Poland, A. I. 1995, Sol. Phys., 162, 1 Forbes, T. G. 2000, J. Geophys. Res., 105, 23153 Gallagher, P. T., Lawrence, G. R., & Dennis, B. R. 2003, ApJL, 588, L53 Gopalswamy, N., Akiyama, S., Yashiro, S., & Makela, P. 2010, in ÅÅ Magnetic Coupling between the Interior and Atmosphere of the Sun, ed. S. S. Hasan & R. J. Rutten, 289-307 Hirayama, T. 1974, Sol. Phys., 34, 323 House, L. L., Wagner, W. J., Hildner, E., Sawyer, C., & Schmidt, H. U. 1981, ApJL, 244, L117 Howard, R. A., Moses, J. D., Vourlidas, A., et al. 2008, Space Sci. Rev., 136, 67 Hundhausen, A. 1999, in The many faces of the sun: a summary of the results from NASA's Solar Maximum Mission., ed. K. T. Strong, J. L. R. Saba, B. M. Haisch, & J. T. Schmelz, 143 Illing, R. M. E., & Hundhausen, A. J. 1985, J. Geophys. Res., 90, 275 Kopp, R. A., & Pneuman, G. W. 1976, Sol. Phys., 50, 85 Kotov, Y. D. 2011, Solar System Research, 45, 93 Kuzin, S. V., Shestov, S. V., Bogachev, S. A., et al. 2011a, Solar System Research, 45, 174 Kuzin, S. V., Zhitnik, I. A., Shestov, S. V., et al. 2011b, Solar System Research, 45, 162 Ma, S., Attrill, G. D. R., Golub, L., & Lin, J. 2010, ApJ, 722, 289 Munro, R. H., Gosling, J. T., Hildner, E., et al. 1979, Sol. Phys., 61, 201 Plunkett, S. P., Michels, D. J., Howard, R. A., et al. 2002, Advances in Space Research, 29, 1473 Pulkkinen, T. 2007, Living Reviews in Solar Physics, 4 Robbrecht, E., Berghmans, D., & Van der Linden, R. A. M. 2009, ApJ, 691, 1222 Schwenn, R. 2006, Living Reviews in Solar Physics, 3 Sheeley, Jr., N. R., Michels, D. J., Howard, R. A., & Koomen, M. J. 1980, ApJL, 237, L99 Sturrock, P. A. 1966, Nature, 211, 695 Sylwester, B., Faucher, P., Jakimiec, J., Krutov, V. V., & McWhirter, R. W. P. 1986, Sol. Phys., 103, 67 Tousey, R. 1973, in Space Research Conference, ed. M. J. Rycroft & S. K. Runcorn, 713-730 Vourlidas, A., Howard, R. A., Esfandiari, E., et al. 2010, ApJ, 722, 1522 Webb, D. F., & Howard, T. A. 2012, Living Reviews in Solar Physics, 9 Webb, D. F., & Hundhausen, A. J. 1987, Sol. Phys., 108, 383 Yashiro, S., Gopalswamy, N., Michalek, G., et al. 2004, Journal of Geophysical Research (Space Physics), 109, 7105 Zhang, J., Dere, K. P., Howard, R. A., Kundu, M. R., & White, S. M. 2001, ApJ, 559, 452 Zhang, J., Dere, K. P., Howard, R. A., & Vourlidas, A. 2004, ApJ, 604, 420