The Last Total Solar Eclipse of the Millennium in Turkey
ASP Conference Series, Vol. 205, 2000
uc, eds.
W. Livingston, and A. Ozg
The Occurrence Probability of X-ray Plasma Ejections by
Solar Flares
S. Akiyama
Department of Astronomical Science, Graduate University for Advanced
Studies, Mitaka, Tokyo 181-8588, Japan
H. Hara
National Astronomical Observatory, Mitaka, Tokyo 181-8588, Japan
Abstract. We have investigated X-ray ejecta above are loops, which
were observed before soft X-ray peak time, using the Yohkoh soft X-ray
observations in 1996 when the background corona was very weak. We
found 35 plasma ejecta out of 84 ares and studied soft X-ray data to
discover the probability to observe the plasma ejecta. On the other hand,
we searched for prominence eruptions using Solar-Geophysical Data in
the same are series and found 16 prominence eruptions out of 29 ares
having soft x-ray plasma ejecta. As a result, 76% of ares have plasma
ejecta observed in either H or soft X-ray wavelength. It follows from
this that there are many are associated ejecta.
1. Introduction
Solar ares observed with Yohkoh soft X-ray telescope occasionally showed looplike or blob-like ejecta even from relatively short duration events (Shibata et al.
1995 Ohyama & Shibata 1997 Akiyama & Hara 1999). Since X-ray plasma
ejecta were observed high above are loops, X-ray plasma ejecta associated with
ares were considered to be a sign that ares were the result of magnetic reconnection. Those plasma ejecta associated with ares, however, were not always
observed in the Yohkoh soft X-ray observations. Of course there are various observing conditions for the Yohkoh soft X-ray telescope and there are some ejecta
over the wide wavelength. For this reason at rst we would like to focus on the
characteristic of ejecta, and some observing conditions which could prevent the
X-ray ejecta associated with ares from being detectable. Secondly we examine
prominence eruptions for the H wavelength. Finally we conclude that ejecta
associated with ares are common structures.
2. Data & Analysis
We used the partial frame images (PFIs) obtained in 1996 with Yohkoh/SXT,
provided that the PFIs included the QUIET mode and the FLARE mode. The
QUIET mode is usual observing conditions and the FLARE mode is automatically triggered for the are observations by the soft and hard X-ray spectrometers
137
138
S. Akiyama & H. Hara
on board Yohkoh. Though the data in the QUIET mode have been barely used
for are studies, these data needed for weaker are events below about C level
in the GOES classication. We dened are as increase in X-ray above A1 level
in the GOES classication, using the GOES satellite. Weaker ares and faint
X-ray plasma ejecta around are loops were easily detected because of very weak
background emission during solar minimum. The selection of ares was based
on the criterion that SXT observation started before the soft X-ray peak time
because many X-ray plasma ejecta could be observed during the impulsive phase
(Ohyama & Shibata 1997). In addition we counted only ejecta that expanded
to at least twice size of the are loops. As a result, we pick up only obvious
ejecta, throwing away marginal ones. For other wavelengths, we used the H
solar are report at the Solar-Geophysical Data.
3. Results & Discussion
We examined 84 ares and found 35 X-ray plasma ejecta associated with those
ares. At rst we checked the characteristic of ejecta. We compared intensity of
ejecta with ares and background. We dened each region using the uctuation
() in a background model. The background model was interpolated from only
non-are pixels. Flare and background regions were dened as the region which
had an intensity more than 30 and less than 3 , respectively. X-ray ejecta
were visually identied in time-sequence images played as movie and in dierence
images, where the background was eliminated. Figure 1a constitutes the average
X-ray intensity of are ({), ejecta ( ), and background ( and 4), as a function
of the average intensity of ares. For background we distinguish between events
observing ejecta () and the others (4). This gure indicates that the intensity
of the background is proportional to the that of the are. We suppose that
this condition is caused by scattering. For ejecta, the intensity is lower than for
ares.
The ratio of the peak intensity of ejecta to are are set out in Figure 1b.
We see here that the intensity of ejecta is distributed from 2 to 60 percentage
of the peak intensity of a are. The intensity of many ejecta is roughly 10% of
peak intensity of a are.
Figure 1c shows the ratio of the intensity of ejecta to the intensity of background as a function of the average intensity of ares. The dashed line shows
the minimum ratio ( 0.4). The error bars are 1 of the background model.
We found that the intensity enhancement of ejecta exceeds 40 % of background
because the minimum ratio is 0.4.
We checked the dierence of background between the events associated
with ejecta and the others. In the Figure 1a, we cannot see any clear dierence
between events associated with ejecta () and the others (4). Figure 1d shows
the distribution of the intensity of the background. The top histogram and the
bottom one are the events associated with ejecta and the others, respectively.
The average backgrounds are indicated by the black triangles on each histogram.
We see from Figure 1d that the distribution of the intensity of the background
is about the same. Therefore, there is no reason to think that ejecta could not
observed due to high intensity of background.
Occurrence Probability of X-ray Plasma Ejections
(a) Intensity
(b) Intensity of Ejecta for Flare
Flare
Ejecta
Background without Ejecta
Background with Ejecta
104
6
103
Number
X-ray Intensity [DN/sec/pixel]
105
102
4
2
101
100
102
103
104
Intensity of Flare [DN/sec/pixel]
0
0.0 1
105
0.1
Ejecta / Flare
1
(c) Intensity of Ejecta for Background
1000.0
(d) Intensity of Background
20
100.0
10.0
10
5
101.58
0
1.0
5
0.1
102
Flare without
Ejecta
15
Number
( Ejecta - Background ) / Background
139
103
104
Intensity of Flare [DN/sec/pixel]
105
0
1 0 0.5 10 1
Flare with
Ejecta
1.62
10
10 1.5 10 2 10 2.5 10 3
[DN/sec/pixel]
Figure 1. (a): X-ray intensity as a function of the intensity of are.
Solid line ({), star ( ), cross () and triangle (4) symbols stand for
the intensity of are, ejecta, background associated with ejecta and
background associated with others, respectively. (b): The Distribution
of the ratio of the peak intensity ( ejecta : are ). (c): Ratio of the
intensity of ejecta to the background as a function of the intensity of a
are. Dashed line shows the minimum ratio ( 0.4). The error bars is
the in a calculated the model background. (d): The distribution of
the intensity of background. The top histogram and the bottom one
are the events associated with ejecta and the others, respectively. The
averages are indicated by the black triangles on each histogram.
Second we examined a few observation conditions. The following factors
might aect the detectability of ejecta, namely exposure time, eld of view and
location of are site. If exposure time, i.e., limited dynamic range of the (CCD)
detector, became shorter, it was probable that the plasma ejecta couldn't be
distinguished from background. We found that the detectability of ejecta was
25 % for short exposure times (1{38 msec]), and 67% for long exposure times
(7548 msec]). For limited eld-of-view we considered the extent of plasma ejecta.
If the ejecta was much larger than the main are region, we could not observe
ejecta. We found that wider elds-of-view were advantageous to observe plasma
140
S. Akiyama & H. Hara
ejecta. We supposed that many X-ray ejecta were relatively large structures.
The point about location of are site was the direction of X-ray ejecta projected
on the image plane. As an extreme case, if X-ray ejecta were collimated and
parallel to the line of sight, we would not see the motion of ejecta. As a result
we found that there is weak correlation between the location and ratio of X-ray
ejecta.
As mentioned above, the detectability of ejecta depend on observation conditions. However, it must be noted that nothing was a necessary and sucient
condition for capability to observe X-ray ejecta. In other words, we found that
there were no denitive conditions for the capability to observe X-ray ejecta.
Consequently, even if we considered the occurrence probability of X-ray ejecta
in lack of each SXT observation conditions, it was reasonable to suppose that
our analysis was trustworthy.
Finally, using the H Solar Flare Report at the Solar-Geophysical Data, we
searched for prominence eruptions associated with ares. These ares were the
same events which we examined to analyze the occurrence probability of X-ray
ejecta with SXT. In our 84 events observed by SXT, we found that 29 events
were observed by H . For 29 ares, there were soft X-ray ejecta in 12 ares, H
prominence eruptions in 16 ares, both X-ray ejecta and prominence eruption
in 6 ares and nothing in 7 ares. As a result, for 76% of the are we found
plasma ejecta observed by either X-ray or H . It follows that there are many
ares associated with ejecta.
4. Summary
We found 35 X-ray ejecta in 84 events studied. The intensity of all of X-ray
ejecta were lower than that of are. On the other hand we found 16 prominence
eruptions out of 29 ares observed by both H and X-ray wavelength. We found
22 plasma ejecta observed at either wavelengths, as a result for 76% ares plasma
ejecta were observed by either H or X-ray.
References
Akiyama, S. & Hara, H. 1999, Adv. Space Sci. in press
Shibata, K. et al. 1995, ApJ, 451, L83
Ohyama, M. & Shibata, K. 1997, PASJ, 49, 249