A study of flareassociated Xray plasma ejections. I. Association with coronal mass ejections

1
A study of flare-associated X-ray plasma
ejections.I. Association with coronal mass
ejections
2005?6?6? ????? ??

• Yeon-Han Kim, Y.-J. Moon, K.-S. Cho, Kap-Sung
  Kim, and Y. D. Park,
• 2005, ApJ, 622, 1240-1250

2
Abstract

• The authors have made a statistical study of the
  relation ship between flare-associated X-ray
  plasma ejections and coronal mass ejections
  (CMEs).
• In 279 limb flares observed by Yohkoh/SXT,
• 69 of the events with plasmoid ejections are
  associated with CMEs, observed by SOHO/LASCO,
• 84 of the events without plasmoid ejections have
  no related CMEs.
• X-ray plasma ejections occur nearly
  simultaneously with HXR peak.
• 80 of the CMEs are preceded X-ray plasma
  ejections, by approximately 20 minutes on average.

3
1. Introduction
4
X-ray plasma ejections

• Outside main flare loops
• Around the impulsive phase of flares
• Bloblike, looplike, jetlike, or complex in shape
• Found in both LDE and impulsive flares

5
Three stages of kinematic evolution

• Ohyama Shibata (1997) found that the ejected
  material was already heated to 10MK before the
  start o the ejection and that its temperature was
  nearly the same as that of the flare loop.
• The three stages of kinematic evolution preflare
  rise, main rise, and gradual propagation.
• This is very similar to the kinematic evolution
  of CMEs. (Zhang et al. 2001a)
• Owing to their kinematic and morphological
  likeness to CMES, X-ray plasma ejections have
  sometimes been regarded as possibly being direct
  signatures of CMEs.

6
Onset of CMEs

• The onset and origin of CMEs are still not well
  understood.
• In order to understand the mechanism of CME
  launch, one needs to observe their early
  signature in the low corona.
• There are several candidate CME early signatures,
  such as filament eruptions, sigmoid-to-arcade
  events, flare-associated X-ray plasma ejections,
  and coronal dimmings.

7
Relation between flare-associated X-ray ejections
and CMEs

• Nitta Akiyama (1999) made the first attempt to
  correlate flare-associated plasma ejections and
  CMEs, using 17 limb flares.
• No CME around the flare time ? No X-ray ejection
• Eight flares with CMEs (all but one) ? X-ray
  ejection
• The authors made statistical extension of Nitta
  Akiyama (1999).
• In addition, they examine the difference in onset
  time between X-ray plasma ejections and their
  associated phenomena and pay special attention to
  comparing the event times between the X-ray
  plasma ejections and the CMEs when the CMEs are
  extrapolated into the Yohkoh field of view.

8
2. Data and analysis
9
2.1. Data

• For the identification of X-ray plasma ejections,
  they used all flare-mode data of Yohkoh/SXT with
  high temporal resolution.
• The flare time was taken by GOES and Yohkoh/HXT.
• The flare location comes from the list of optical
  flares at NGDC or SXT images.
• They used CME catalogue (Yashiro et al. 2004) and
  raw LASCO data observed in order to identify the
  position and speed of the CMEs.
• They also used the 195Å (Fe XII) SOHO/EIT images,
  since this channel allows one to see both the
  prominence and coronal structures in CMEs (Dere
  et al. 1997).

10
Figure 1
11
2.2. Event selection

• They consider 279 limb flares whose longitudes
  are larger than 60 from all flare-mode data in
  Yohkoh/SXT from 1999 April to 2001 March.
• The identification procedure whether each flares
  accompanied an X-ray plasma ejection
• Make movie files of SXT.
• Identify large-scale plasma ejection around the
  impulsive phase looking half- and
  quarter-resolution movies.
• For the events without large-scale plasma
  ejections, they examined full-resolution movies.
• They found many confined ejections that did not
  show any eruptive motion in the half- and
  quarter-resolution images but did appear in the
  full-resolution images.
• As a result of this analysis, they found a total
  of 137 flares with X-ray plasma ejections.
• The identified event times for some X-ray plasma
  ejections may be a little later than the real
  onset times, because of their apparently being
  hidden by flare loops.

12
2.2. Event selection

• They established associations between the flares
  and the LASCO CMEs according to temporal and
  spatial proximity that is, the flare start time
  is within 1 hr of the CME onset time extrapolated
  at 1.1 R? using the constant-speed method, and
  the flare position angle is within the angular
  extent of the associated CME.
• We excluded data from the period when LASCO made
  no observations because of its operational
  condition.

13
2.3. Morphological classification

• They classified the X-ray plasma jections into
  five groups according to their shape
• Loop-type (60 events)
• Shape of loops.
• Spray-type (40)
• Continuous stream of plasma without any typical
  shape.
• Jet-type (11)
• Collimated motions of plasma.
• Confined ejection (18)
• Limited plasma motion near the flaring site that
  is usually seen only in the full-resolution
  flare-mode movie.
• Other (8)

14
Figure 2

• Typical example of a loop-type X-ray plasma
  ejection, associated with an M2.4 X-ray flare on
  1999 July 25.
• Initial speed about 112 km/s

15
Figure 3

• Running-difference images of the CME associated
  with the plasma ejection shown in Fig. 2.
• The central circle drawn in the top left panel
  indicates the solar disk, and the small box
  represents the Yohkoh field of view, which
  corresponds to 10'4 10'4.
• X-ray plasma ejection is quite similar to that of
  the associated CME.
• ?
• Such a similarity may argue for the possibility
  that X-ray plasma ejections are early signatures
  of CMEs.

16
Figure 4

• A jet-type X-ray plasma ejection and its
  associated CME on 2000 October 26.
• Top Half-resolution SXT flare-mode images.
• Bottom The corresponding LASCO C2 and EIT
  images.
• It is interesting that this ejection was
  accompanied by quite a narrow CME compared with
  the one associated with the loop-type ejection on
  1999 July 25.

17
Table 1
Continue
18
3. Results and discussion
19
3.1. CME Association

• About half (137/279) of the flares have
  associated plasma ejections within the
  sensitivity of the Yohkoh SXT.
• While 69 (95/137) of the X-ray plasma ejections
  are associated with CMEs, 31 (42/137) have no
  CMEs.
• Of the events without plasma ejections, only 16
  (23/142) are related to CMEs, and 84 (119/142)
  do not have associated CMEs.
• On the other hand, it is also found that 81
  (95/118) of the flares associated with LASCO CMEs
  are related to X-ray plasma ejections.
• Our results support Nitta Akiyama (1999), who
  found a close correlation between the presence or
  absence of X-ray plasma ejections and CMEs using
  a sample of 17 limb flares.

20
Flare-strength dependence of the association
between X-ray plasma ejections and CMEs

• Stronger flares with CMEs are more closely
  associated with X-ray plasma ejections.
• It is also found that for LDEs, all
  flare-associated CMEs have associated X-ray
  plasma ejections regardless of their strength. It
  it known that LDEs are highly correlated with
  CMEs and filament eruptions.
• The association of non-LDE flares with CMEs
  varies with flare strength.
• Thus, Nitta (2002) proposed that LDEs may be a
  part of the CME process that commence as a result
  of large-scale instability or loss of equilibrium
  and that non-LDE flares are something else that
  could occur without CMEs.

21
Morphological dependence of the CME associations

• Loop-type, spray-type, and jet-type of plasma
  ejections show a relatively high association with
  CMEs in paticular, the jet type plasma ejections
  are all associated with CMEs.
• The morphology of an X-ray plasma ejection will
  be affected by the magnetic topology of the
  flaring site.
• Such a close correlation may imply that an open
  field structure near a flaring site makes a
  better environment or producing a CME.

22
3.2. Temporal relationship
Figure 5 Time differences between X-ray plasma
ejection start and HXR flare peak (93 events).

• The time differences fall within 10 minutes for
  most events.
• The mean time difference is only about -2
  minutes, implying that X-ray plasma ejections are
  nearly coincident with HXR flare peak times.
• Our results also support the theory that X-ray
  plasma ejections are probably due to magnetic
  reconnection.

23
3.2. Temporal relationship
Figure 6 Comparison of the event times of the
X-ray plasma ejections and the CME event times
extrapolated into the Yohkoh field of view. (43
events).

• They compared the event times of the X-ray plasma
  ejections with the extrapolated CME-front times
  at the same location in the Yohkoh field of view.
• They find that the extrapolated CME fronts in
  most cases (35/43) preceded the expanding fronts
  of the X-ray plasma ejections, by about 20
  minutes on average.
• In addition, for about 28 of the events (12/43)
  both fronts are coincident to within 10 minutes.

24
3.2. Temporal relationship

• They note several reports of strong accelerations
  in the lower corona (e.g., Zhang et al. 2001a)
• If we were to consider such accelerations, the
  CME event times would be even earlier.
• From Figure 7, they find that the CME was
  strongly accelerated below 2R?.
• As a result, the real onset time is found to be
  much earlier than the onset time predicted by the
  constant-speed method.
• Statistically speaking, the fronts of X-ray
  plasma ejections seem to represent the CME the
  CMEs internal structures rather than early
  signatures of CME fronts.

GOES
CME (C2, C3)
CME (C1)
Extrapolated CME time
Figure 7 Height-time behavior of a well-observed
CME on 1998 June 11 from the lower corona to the
higher corona.
25
3.3. Flare association

• Shibata (1995) have argued that X-ray plasma
  ejections are a universal phenomenon in solar
  flares.
• Several observations of each LDEs and impulsive
  flares are consistent with the predictions of
  CSHKP-type flare models, in which magnetic
  reconnections occur in the vertical current sheet
  above flare loops.

26
3.3. Flare association

• Occurrence rate of flare associated X-ray plasma
  ejections are 35-40 (Nitta 1996), 20-35
  (Akiyama), and 43-46 (Ohyama Shibata 2000).
• Although X-ray plasma ejections were originally
  found around the flare impulsive phase (shibata
  et al. 1995), SXT flare-mode observations
  occasionally started too late to catch this
  phase.
• 63-70 with observations that started before the
  HXR peak time (Ohyama Shibata 2000)
• It is difficult to detect X-ray plasmoids in the
  weaker flares and proposed that X-ray plasma
  ejections are a general phenomenon associated
  with solar flares.
• The authors results supports it.

27
4. Summary and conclusion
28
Summary and conclusion

• In this work, we have carried out a comprehensive
  statistical study in order to understand the
  relationship between flare associated X-ray
  plasma ejections and CMEs, using data from 1999
  April to 2001 March.
• There have been several studies of the close
  relationship between flares and CMEs.
• A detailed discussion will be presented in a
  separate paper.

29
Nitta Akiyama (1999)