A CATALOG OF SOLAR X-RAY PLASMA EJECTIONS OBSERVED BY THE SOFT X-RAY TELESCOPE ON BOARD YOHKOH

M. Tomczak; E. Chmielewska

doi:10.1088/0067-0049/199/1/10

1. INTRODUCTION

X-ray plasma ejections (XPEs) are sudden expulsions of hot magnetized plasma in the solar corona seen in X-rays. They establish a wide range of macroscopic motions showing different morphology, kinematics, and physical conditions. XPEs occur usually during the impulsive phase of flares, but their connection with other solar-activity phenomena, such as coronal mass ejections (CMEs), prominences, radio bursts, coronal dimmings, and global waves, is also known. There are some restrictions in calling any motions in the corona around the flare times XPEs. The restrictions regard the size, duration, brightness, speed, etc., and are introduced mainly by spatial, temporal, and spectral resolutions of imaging instruments and their operational schemes.

XPEs have been systematically observed since 1991 when the Yohkoh satellite began to operate. They became commonly known since the paper written by Shibata et al. (1995) was published. However, we note earlier articles on essentially the same phenomena from the Solar Maximum Mission (Harrison et al. 1985) and from Yohkoh (Klimchuk et al. 1994). Until now images recorded by the Yohkoh Soft X-ray Telescope (SXT; Tsuneta et al. 1991) have been the largest database of XPEs, even though the newer solar X-ray imaging instruments operate, e.g., GOES Solar X-ray Imager, RHESSI, and Hinode X-Ray Telescope.

Detailed analyses of individual XPEs were performed first by Tsuneta (1997) and Ohyama & Shibata (1997, 1998). In these papers, the authors determined values of physical parameters describing an XPE using temperature and emission measure maps obtained from SXT images. The maps allowed them to investigate overall magnetic configuration including a flare loop and a reconnection region. They also used hard X-ray light curves, derived by the Yohkoh Hard X-ray Telescope (HXT; Kosugi et al. 1991), for a detailed description of reconnection timing.

Nitta & Akiyama (1999) made the first attempt to correlate XPEs and CMEs. For 17 well-observed limb flares they found that flares associated with CMEs show XPEs and vice versa—flares not associated with CMEs also lack XPEs. A more extensive investigation of association between XPEs and flares was performed by Ohyama & Shibata (2000). For 57 well-observed limb flares they found that almost 70% show XPEs. They also reported dependence on X-ray class, namely, the association is larger for stronger flares, but it could be caused by observational biases.

To investigate interesting examples of XPEs, other Yohkoh instruments have also been used, namely, the HXT (Hudson et al. 2001) and the Bragg Crystal Spectrometer (BCS; Tomczak 2005). In both papers, a special location of investigated events has been chosen. These XPEs occurred far behind the solar limb and owing to their fast expansion came into view of an instrument before brighter flares, which expand slower. It is virtually the only way to use full-Sun instruments like the BCS to resolve faint soft X-ray emission of XPEs. The behind-the-limb location also protects against strong emission of footpoint hard X-ray sources of flares, which usually dominate fainter coronal emission. The obtained results proved that an XPE can contain energetic non-thermal electrons (Hudson et al. 2001) and superhot thermal plasma (Tomczak 2005).

An important progress in investigation of XPEs led to a trilogy by Kim et al. (2004, 2005a, 2005b). They investigated systematically SXT observations obtained during a two-year interval and found 137 XPEs. The events were the subject of multipurpose analysis—the authors introduced a morphological classification of XPEs, investigated their kinematics, and specified the association with flares and CMEs. The present name of XPEs also comes from these papers. We recapitulate the results of Kim et al. in detail in further sections, where we compare them with our results.

More recently, an association between XPEs and radio events and prominences has been investigated. In statistical surveys Shanmugaraju et al. (2006) studied type II radio bursts, whereas Kołomański et al. (2007) studied drifting pulsating structures (DPS). Both surveys suggest a kind of connection between XPEs and radio events, but further examinations are needed to establish the connection. The relationship between hot (XPEs) and cold (prominences) ejections was discussed by Ohyama & Shibata (2008) and Kim et al. (2009) for single events, which was followed up with statistical studies (Chmielewska & Tomczak 2012).

Finally, in our short review illustrating research progress we would like to recall the following two papers. First, the results of a quantitative analysis of SXT images describing time evolution of basic physical parameters for 12 XPEs were given by Tomczak & Ronowicz (2007). Second, after extensive analysis of a complex XPE that consisted of several recurrent episodes, Nishizuka et al. (2010) reported a close connection between sequential ejections and successive hard X-ray bursts.

The most commonly accepted physical explanation of XPEs connects these phenomena directly with flare magnetic reconnection. Shibata et al. (1995) regarded XPEs as proof of the presence of plasmoids driven by magnetic reconnection occurring above a soft X-ray loop in short-duration, compact-loop flares similar to the canonical two-dimensional CSHKP model (Švestka & Cliver 1992 and references therein), which was proposed for long-term, two-ribbon flares. In this way, Shibata et al. (1995) postulated a unification of two observationally distinct classes of flares, i.e., two-ribbon flares and compact-loop flares, by a single mechanism of magnetic reconnection called the plasmoid-induced-reconnection model.

The first qualitative studies of individual events (Ohyama & Shibata 1997, 1998) reported that the measured velocities of XPEs are much smaller than the velocity of reconnection outflow expected from the model to be about the Alfvén speed. To reconcile this discrepancy, the authors suggested (1) the high density of the XPEs, (2) the time evolution effect (i.e., the plasmoid should be accelerated as it propagates, thus the investigated XPEs have not yet reached the maximum velocity), or (3) an interaction with coronal magnetic fields overlying the XPEs.

Although more recently two-dimensional resistive-MHD numerical simulations of the reconnection have explained kinematical properties of various observational features attributed to the current-sheet plasmoids (Bárta et al. 2008), it has been expected that three-dimensional reconnection renders a more realistic description of eruptive phenomena. For example, Nitta et al. (2010) suggested three-dimensional quadrupolar reconnection of two loop systems that appear to exchange their footpoints as a result of loop–loop interaction (Aschwanden et al. 1999).

On the other hand, in some cases the XPEs seem to play the same role as phenomena called precursors of CMEs (Cheng et al. 2011). This opinion is supported observationally by common kinematical evolution of XPEs and CMEs (Gallagher et al. 2003; Dauphin et al. 2006; Bak-Steślicka et al. 2011), as well as their morphological resemblance (Kim et al. 2005b). If so, loss of equilibrium or MHD instability, commonly accepted as one of the CME triggering mechanisms (Forbes 2000), should also be taken seriously into consideration as a cause of XPEs.

Reports concerning XPE observations in the SXT database were scattered until now across many different sources: refereed articles, conference communications, electronic bulletins, etc. An exception was the survey given by Kim et al. (2005b), which includes almost all the XPEs associated with limb flares for a two-year interval. Our motivation was to ingest all available reports in one catalog and organize them in a uniform way for easy usage. We have examined SXT images in those time intervals in which any systematic searches of XPEs did not perform.

Knowledge about XPEs has so far been shaped by a limited number of events that have repeatedly appeared in the literature. Our catalog is meant to serve as a convenient tool for every scientist who wants to better understand the nature of XPEs.

2. DESCRIPTION OF THE CATALOG

2.1. General Contents

The catalog contains all the XPEs we know that were observed by the SXT during the entire Yohkoh operations, i.e., between 1991 October 1 and 2001 December 14. There are three main surveys of events that we used in our catalog.

1.
Kim et al. (2005b), which contain 137 limb events, observed between 1999 April and 2001 March.
2.
M. Ohyama (2009, private communication), with 53 limb events that occurred between 1991 October and 1998 August. The survey was prepared for the aim of statistical research (Ohyama & Shibata 2000) but was not published.
3.
Chmielewska (2010), which reports 113 events, observed basically within two time intervals, 1998 September–1999 March and 2001 April–2001 December, that were not systematically searched before.

We also incorporated 65 XPEs reported in other scientific papers and in the electronic bulletin Yohkoh SXT Science Nuggets.¹

Keeping in mind the examination of SXT images made by different authors, we can conclude that the list of XPEs associated with limb flares (defined as |λ| > 60°, where λ is the heliographic longitude) is almost complete. On the other hand, the list of XPEs associated with disk flares is largely incomplete, with the exception of time intervals examined by Chmielewska (2010). Occasional reports of XPEs not associated with any flares (Klimchuk et al. 1994) teach us that the SXT images made without any flares should also be examined, and this work still awaits to be done.

In summary, our catalog contains 368 events. Time frequency of XPEs' occurrence during the Yohkoh mission is given in Figure 1, where sizes of bins are six and three months for years 1991–1997 and 1998–2001, respectively. This traces variability of general solar activity. A larger occurrence rate in cycle 23 in comparison with cycle 22 may be attributable to the revision of the SXT flare mode observing sequences. Indeed, the ratio of the number of XPEs and that of flares taken from the Solar-Geophysical Data (SGD) is 3.6 times greater for solar cycle 23 than for cycle 22.

Before 1997, a routine scheme of observations during the flare mode was dominated by images in which the exposure time and the position of the field of view (typically 2.5×2.5 arcmin²) were automatically adjusted by the signals and locations of the brightest pixels. XPEs are distinctly fainter and located higher in the corona than flares. Thus, they had usually too poor statistics during short exposure times and owing to a fast expansion left immediately the narrow field of view. Under these circumstances, XPEs were rarely well observed.

In 1997, the frequency of images with sufficiently long and constant exposures and broader field of view (5.2 × 5.2 arcmin² and 10.5 × 10.5 arcmin²) was increased to every 10–20 s (Nitta & Akiyama 1999). This observational scheme worked more favorably for the XPEs identification; however, flare structures seen in those images often suffer from heavy saturation that manifests itself as vertical spikes disturbing a picture of XPEs.

We registered events to the catalog on the basis of the SXT observations exclusively. For this reason, we omitted some X-ray ejections from years 1991–2001 identified using observations made with other instruments alone, like the HXT, e.g., Hudson et al. (2001).

The online catalog resides at http://www.astro.uni.wroc.pl/XPE/catalogue.html since 2010 October 22. It is also linked from the Yohkoh Legacy Data Archive² (Takeda et al. 2009). The general arrangement of the catalog as a matrix of years and months of observation is presented in Figure 2(a). After clicking the month, each XPE is identified by a chronological catalog number, date, and time of occurrence. The letter (a) added to a start time means that the XPE began earlier than shown in the available movies. The letter (b) added to an end time means that the XPE finished later than shown in the available movies. The letter (c) added to an end time means a time interval of available movies in which we cannot identify the XPE reported earlier by other authors. Each event has links to five entries that provide detailed information on the XPE, flare (SXR and HXR), CME, and references on the XPE (see an example in Figure 2(b)).

**Figure 2.** Overview of the *Yohkoh* SXT XPE Catalog, which resides online at http://www.astro.uni.wroc.pl/XPE/catalogue.html. (a) The main entry into the catalog as a matrix of years and months of observations. (b) A few of the entries in the catalog for 1997 November. (c) A screenshot for the XPE entry of event no. 62 on 1997 November 14.
Download figure:
Standard image High-resolution image

2.2. The "XPE" Entry

This entry contains eight columns labeled as follows: (1) event ID, (2) date, (3) time, (4) quality, (5) classification, (6) movies, (7) results of analysis, and (8) references (see an example in Figure 2(c)). The first three columns are replicated from the higher entry.

In Column 4, we indicate the quality of available SXT observations by assigning one letter between (A) and (D). The letter (A) means the highest quality: an XPE is clearly seen and only slightly disturbed by flare saturation, observations have almost full spatial and time coverage, images are made by at least two different filters. In conclusion, events with this letter are a good source for any kind of quantitative analysis including plasma diagnostics on the basis of the filter-ratio method (Hara et al. 1992). The letter (B) also means quite good quality of observations, but the usage of only one filter in some cases makes plasma diagnostics unavailable. Nevertheless, XPEs marked with this letter are always good for kinematical studies. The letter (C) means poor quality for some of the following reasons: the brightness of the XPE only marginally above the background, short observation window, inadequate field of view, or strong effect from flare saturation. For events with this letter only limited analyses are usually possible, e.g., a description with our three-parameters classification. The letter (D) is designed for XPEs, which were mentioned by other authors but whose presence is not confirmed in the movies that we made.

In Column 5, we characterize general observational features of XPEs using a new classification scheme that we have developed in this catalog. In our classification we define three criteria considering (a) morphology of an XPE, (b) its kinematics, and (c) recurrence. Examining each criterion, we distinguish two subclasses of events only: (a) 1—collimated, 2—loop-like; (b) 1—confined, 2—eruptive; (c) 1—single, 2—recurrent. In consequence, our classification can resolve 2³ = 8 subclasses.

Our motivation should be addressed in the context of the earlier classification made by Kim et al. (2004), who proposed five morphological groups of XPEs: a loop-type, spray-type, jet-type, confined, and other. In our opinion, the classification that is too "hair-splitting" may be uncomfortable in practical usage because it is easy to make a wrong assignment in the case of poor quality of the observational data or their limited coverage. We may recall attempts of organizing properties of CMEs as observed by different coronagraphs (Munro & Sime 1985; Howard et al. 1985; Burkepile & St. Cyr 1993; Gopalswamy et al. 2009). They have never worked out any commonly accepted classification scheme for CMEs on the basis of morphological features only.

Our morphological criterion resolves only a direction of soft X-ray plasma movement in comparison with the direction of local magnetic field. Roughly speaking, in the case of subclass 1 the direction is parallel, i.e., along the already existing magnetic field lines, in the case of subclass 2—perpendicular, i.e., across the already existing lines (or strictly speaking, together with them). XPEs from the first morphological subclass usually take the form of a blob or a column of matter propagating within a bundle of magnetic lines without any serious modification of their structure. Therefore, these events are more collimated (hence its name) and less energetic. The direction of their motions depends on the configuration of guiding lines. Our subclass 1 comprises the majority of events classified by Kim et al. (2004) as the spray-type and jet-type events. XPEs from the second morphological subclass take the form of a rising loop or a system of loops. Our subclass 2 is very similar to the loop-type events proposed by Kim et al. (2004). Some events showed features of both morphological subclasses, 1 and 2; in this case we classified them according to a more evident feature.

For the XPE assignment into one of the kinematical subclasses we have chosen height increase rate above the chromosphere, $\dot{h}$ . A negative value, $\dot{h} < 0$ , means subclass 1; the opposite case, $\dot{h} \ge 0$ , means subclass 2. There are several papers presenting plots h(t) for events belonging to both subclasses (Klimchuk et al. 1994; Tsuneta 1997; Ohyama & Shibata 1997, 1998, 2008; Nitta & Akiyama 1999; Kundu et al. 2001; Alexander et al. 2002; Tomczak 2003, 2004; Kim et al. 2005a, 2005b, 2009; Nishizuka et al. 2010). XPEs from the first kinematical subclass can be connected with plasma motion within closed magnetic structures as well as with some changes in a plasma situation or in the local magnetic field structure that do not evacuate any mass from the Sun. In summary, XPEs from the first kinematical subclass suggest the presence of a kind of magnetic or gravitational confinement of X-ray plasma. For XPEs from the second kinematical subclass, an increasing velocity in the radial direction in the field of view of the SXT allows us to anticipate further expansion leading to irreversible changes (eruption) of the local magnetic field. In consequence, at least a part of the plasma escapes from the Sun.

For many weak XPEs construction of the diagram h versus t was impossible. Therefore, we estimated dh/dt qualitatively by watching the expansion rate of XPEs in movies made with images uniformly spaced in time. In some cases the classification was problematic because of a limited coverage of available observations; hence, we added a question mark to the digit for this criterion.

According to our third criterion, we separate disposable, unique XPEs that occurred once (subclass 1) from recurrent events for which following expanding structures can be seen with time (subclass 2). The majority of XPEs described in the literature belong to subclass 1; however, samples of subclass 2 were already presented (Nitta & Akiyama 1999; Tomczak 2003; Nishizuka et al. 2010). A partial time coverage of the available observations probably introduces a bias toward single XPEs because a narrow observational window allows us to resolve only a single feature even for recurrent XPEs.

Especially important is Column 6, in which all available movies illustrating evolution of the XPE are collected. The movies consist of images obtained by the SXT and are written in the MPEG format. Images made by using particular filters and spatial resolutions are collected in separate movies. We label the movies to indicate their contents, e.g., AlMg/HN marks images obtained with the AlMg filter of half resolution. We use the following standard annotations of filters and spatial resolutions applied in the Yohkoh software (Morrison 1994): the filter Al.1—the wavelength range 2.5–36 Å, AlMg—2.4–32 Å, Mg3—2.4–23 Å, Al12—2.4–13 Å, Be119—2.3–10 Å; the full resolution, FN, 2.45 arcsec, half-resolution, HN, 4.9 arcsec, quarter resolution, QN, 9.8 arcsec. A particular resolution means a specific field of view: 2.6 × 2.6 arcmin², 5.2 × 5.2 arcmin², and 10.5 × 10.5 arcmin² for the FN, HN, and QN resolutions, respectively. Sometimes we divided images made with the same filter and the same spatial resolution into separate movies consisting of images made with the same time exposition. In that case, the labels contain additionally successive roman digits.

The movies consist of images that we previously processed using the standard Yohkoh routine SXT_PREP, allowing us to reduce the influence of typical instrumental biases, e.g., telemetry compression, electronic offset, dark current, straylight, de-jittering. In the images the heliospheric coordinates are overwritten by using the SolarSoft routine PLOT_MAP. For better identification of faint features slightly above the background, we represented a signal distribution with nonlinear color table nos. 16 ("Haze"), 33 ("Blue-red"), or 3 ("Red temperature") available in the Interactive Data Language. Images that form movies in the catalog are sometimes non-uniformly spaced in time; therefore, it is strongly recommended to watch a time print that is present in each image.

The XPEs for which a more detailed analysis has already been performed show in Column 7 an entry with a concise report concerning results. Inside the report, the obtained values of investigated parameters such as velocity, acceleration, temperature, emission measure, electron density, pressure, and secondaries, as well as references, are given. For 14 events (nos. 29, 30, 34, 53, 62, 67, 72, 126, 144, 169, 252, 293, 303, and 330) a more complete set of results is presented in the form of plots and tables illustrating the whole evolution (Ronowicz 2007).

Finally, in Column 8 references to all the reports (also in electronic form) in chronological order are given.

2.3. Other Catalog Entries

The "SXR Flare" entry contains basic information about a flare that was associated with the given XPE. Finding the associated flare for the majority of XPEs in the catalog was very easy. A flare is seen usually in movies illustrating evolution of an XPE as heavy saturation due to its much stronger soft X-ray radiation. In several cases, a flare occurred simultaneously with an XPE but in another active region. We considered this flare as the associated event only when some distinct magnetic loops connecting both active regions were seen. Finally, there are several XPEs that occurred when no flare was observed on the Sun.

Each record describes the following attributes: date, time of start, maximum, and end defined on the basis of the Geostationary Operational Environmental Satellites (GOES) 1–8 Å light curve, GOES class, location in heliographic coordinates, and NOAA active region number. By clicking on the GOES class, one can view the GOES light curves in two wavelength ranges: 1–8 Å (upper) and 0.5–4 Å (lower). Time span of plots is always two hours and includes the occurrence of an XPE, which is marked by vertical lines. The hatched area on the plot represents Yohkoh nights.

Records presented in this entry are generally adopted from the SGD; however, some clarifications and supplements were necessary. For example, the lacking locations were completed on the basis of SXT images as a place of flare bright loop-top kernels. The values obtained in this way are given in parentheses. Coordinates of events that occurred behind the solar limb are taken basically from Tomczak (2009).

If no flare was associated with an XPE, the tags devoted to flare characteristics are empty. Exceptions are GOES light curves, heliographic coordinates, and NOAA active region number. The last two tags describe then an XPE.

The "HXR Flare" entry presents some attributes of hard X-rays emitted by a flare that was associated with a given XPE. We used data from the HXT on board Yohkoh. This telescope measured the hard X-ray flux in four energy bands: 14–23 (L), 23–33 (M1), 33–53 (M2), and 53–93 keV (H). Each record contains peak time and peak count rate (together with the background), inferred for the energy band M1. By clicking on the peak count rate, one can view the HXT light curves in all energy bands. Time span of plots usually includes the maximum of hard X-ray flux and the occurrence of an XPE, which is marked by vertical lines. Wherever available, we also include the event ID from the Yohkoh Flare Catalog (HXT/SXT/SXS/HXS).³

If no flare was associated with an XPE, the tags devoted to flare characteristics are empty. If the hard X-ray flux in the energy band M1 was below the doubled value of the background, we left the tags describing peak time and peak time rate empty.

The "CME" entry contains some attributes of a CME that was associated with a given XPE. The observations are derived by the Large Angle and Spectrometric Coronagraph (LASCO) on board the Solar and Heliospheric Observatory (SOHO). From the SOHO LASCO CME Catalog⁴ (Gopalswamy et al. 2009) values of the following parameters are given: date and time of the first appearance in the C2 coronagraph field of view, central position angle, angular width, speed from linear fit to the h(t) measurements, and acceleration inferred from the quadratic fit. The first appearance time is the link to the beginning of the list of events in the SOHO LASCO CME Catalog for a given year and month. By clicking on the entry "Related links," one can view a javascript movie of the CMEs within the C2 field of view for a given day. Movies reside at the home page of the SOHO LASCO CME Catalog.

According to Yashiro et al. (2008), we consider a pair XPE–CME as physically connected if the XPE occurred within position angles defined by the CME angular width increased by 10° from both sides. Moreover, time of the XPE occurrence had to fall within a three hour interval centered around extrapolated time of the CME start for h = 1 R_☉. For extrapolation we used time of the first appearance in the LASCO/C2 field of view and linear velocity taken from the SOHO LASCO CME Catalog. The "CME" entry is empty when the XPE occurred during a LASCO gap. XPEs not associated with a CME are labeled "No related event."

In the "References" entry the references to all reports (also in electronic form), known for us, that mentioned a particular XPE are given in chronological order.

2.4. Statistics of the Catalog Content

Some useful characteristics of XPEs included in the catalog are extracted in Table 1. In this subsection, we discuss general statistics of the catalog content.

Table 1. XPEs Presented in the Catalog

No.	Date	Time	Class.	Q.	AR	GOES	Coordinates	CME	References
001	1991 Oct 22	06:42.2–07:28.8^a	2,1,2	B	6891	M1.2	S11 E85	⋅⋅⋅	19
002	1991 Nov 2	16:31.2–16:57.3	2,2,1	B	6891	M4.8	S10 W84	⋅⋅⋅	19
003	1991 Nov 17	18:32.9–18:41.8^a	1,1,2	C	6929	M1.9	S12 E78	⋅⋅⋅	19
004	1991 Dec 2	04:50.6–05:21.0	2,2,1	A	6952	M3.6	N18 E92	⋅⋅⋅	18,19,27,28,54
005	1991 Dec 3	16:35.0–17:04.5	2,1,1	B	6952	X2.2	N17 E72	⋅⋅⋅	19
006	1991 Dec 9	02:02.7–02:06.5^a	2,1,1	C	6966	M1.4	S06 E91	⋅⋅⋅	19
007	1991 Dec 10	04:03.0–04:10.1^a	1,1,1	B	6968	C9.3	S14 E93	⋅⋅⋅	19
008	1992 Jan 13	17:27.9–17:35.2^a	2,2,2	C	6994	M2.0	(S15 W89)	⋅⋅⋅	18,19,27,28
009	1992 Jan 13	19:04.1–19:13.8^a	1,2,1	C	7012	M1.3	S10 E95	⋅⋅⋅	19
010	1992 Jan 14	19:29.0^b–19:32.6^a	2,2,1	B	7012	M1.7	S11 E89	⋅⋅⋅	19
011	1992 Jan 15	18:56.1–19:04.5	2,2,1	B	7012	M2.0	(S09 E72)	⋅⋅⋅	19
012	1992 Jan 30	17:07.6–17:17.6^a	2,1,1	C	7042	M1.6	S13 E84	⋅⋅⋅	19
013	1992 Feb 6	03:17.4–03:36.6^a	2,2,1	B	7030	M7.6	N05 W82	⋅⋅⋅	18,19,27,28
014	1992 Feb 6	20:52.7–21:24.8^a	2,2,1	B	7030	M4.1	N05 W94	⋅⋅⋅	19
015	1992 Feb 9	03:01.0–03:10.7^a	2,1,1	C	7035	M1.2	S17 W74	⋅⋅⋅	19
016	1992 Feb 17	15:41.8–16:25.0	1,2,2	C	7050	M1.9	N16 W81	⋅⋅⋅	18,19,27,28
017	1992 Feb 18	18:00.1–18:29.9	2,1,1	C	7067	+^d	(N05 E89)	⋅⋅⋅	13
018	1992 Feb 19	14:45.4–15:39.1^a	2,2,1	B	7067	M1.2	N06 E94	⋅⋅⋅	19
019	1992 Feb 21	03:11.9–03:19.2^a	2,1,1	C	7070	M3.2	(N09 E80)	⋅⋅⋅	19
020	1992 Feb 21	22:04.6–22:08.0^a	2,1?,1	C	7070	M2.2	N05 E65	⋅⋅⋅	19
021	1992 Apr 1	10:12.8–10:22.5	2,2,1	B	7123	M2.3	(S03 E89)	⋅⋅⋅	18,27,28
022	1992 Jun 5	18:08.7–19:08.9	2,1,1,	C	7186	C2.6	N07 E28	⋅⋅⋅	13
023	1992 Jun 7	01:41.0–01:50.0^a	2,2,1	B	7186	M2.7	N09 E10	⋅⋅⋅	4
024	1992 Jul 20	17:18.1^b–17:49.6	1,1,1	C	7222	—	(S06 W88)	⋅⋅⋅	13
025	1992 Jul 29	20:19.7–21:09.8	2,2,2	B	7236	—	(N19 W88)	⋅⋅⋅	13
026	1992 Aug 25	19:02.6–19:36.4^a	1,1,1	B	7260	C8.7	N13 W98	⋅⋅⋅	19
027	1992 Sep 9	02:06.2–02:18.6	1,1,1	C	7270	M3.1	S10 W72	⋅⋅⋅	19
028	1992 Sep 9	17:57.9–18:07.2	1,2?,1	C	7270	M1.9	S11 W78	⋅⋅⋅	19
029	1992 Oct 4	22:14.0–22:32.4	2,2,2	B	7293	M2.4	S05 W90	⋅⋅⋅	18,19,24,27,28,53
030	1992 Oct 5	09:24.3–09:52.0	1,2,1	B	7293	M2.0	S08 W90	⋅⋅⋅	10,12,18,19,21,24,53
031	1992 Nov 5	06:19.0–06:40.7^c	⋅⋅⋅	D	7323	M2.0	S16 W90	⋅⋅⋅	18,19,27,28
032	1992 Nov 5	20:30.1–21:08.6^c	⋅⋅⋅	D	7323	C8.7	S17 W92	⋅⋅⋅	19
033	1993 Feb 14	12:51.9–12:59.4	2,2,1	B	7427	M2.0	S22 E78	⋅⋅⋅	19
034	1993 Feb 17	10:35.4–10:53.5^a	1,2,1	B	7420	M5.8	S07 W87	⋅⋅⋅	14,18,19,24,27,28,53
035	1993 Feb 21	00:31.2–00:45.2^a	2,2,1	B	7433	M1.4	N13 E75	⋅⋅⋅	19
036	1993 Mar 15	20:31.9–21:15.2	2,2,1	B	7440	M2.9	S03 W93	⋅⋅⋅	4
037	1993 Mar 23	01:21.0^b–01:29.5^a	2,2,1	B	7448	M2.3	N18 W78	⋅⋅⋅	4
038	1993 May 7	20:56.6–21:30.6^a	2,2,1	B	7500	M1.6	N14 E41	⋅⋅⋅	4
039	1993 May 14	22:00.1–22:10.3	2,2,1	A	7500	M4.4	N19 W48	⋅⋅⋅	23
040	1993 Jun 25	03:13.8^b–03:40.3	2,2,1	B	7530	M5.1	S09 E88	⋅⋅⋅	19
041	1993 Jun 28	01:06.7–01:23.0^a	2,2,2	B	7535	C6.5	N03 E69	⋅⋅⋅	6
042	1993 Sep 26	17:26.2–17:28.3^a	2,1,1	B	7590	C3.4	N14 E94	⋅⋅⋅	19
043	1993 Sep 27	12:07.5–12:17.8	2,2,1	B	7590	M1.8	N08 E90	⋅⋅⋅	18,19
044	1993 Oct 1	23:51.3–00:01.3^a	2,1,1	C	7592	C8.5	S14 E69	⋅⋅⋅	19
045	1993 Nov 11	11:15.4–11:31.8	1,2,1	B	7618	C9.7	N10 E95	⋅⋅⋅	10,14,20
046	1993 Nov 13	06:38.6–06:48.9	1,2?,1	B	7618	M2.1	N08 E73	⋅⋅⋅	19
047	1994 Jan 5	06:49.4–06:59.6^a	2,2?,1	B	7647	M1.0	S13 W23	⋅⋅⋅	4
048	1994 Jan 16	23:09.6–23:22.9^a	2,2,1	B	7654	M6.1	N05 E71	⋅⋅⋅	19
049	1994 Jan 27	03:47.8–03:59.0^a	1,1,1	B	7654	C4.6	N08 W68	⋅⋅⋅	4
050	1994 Jan 28	16:53.2–17:26.5^a	2,2,2	B	7654	M1.8	N08 W85	⋅⋅⋅	19
051	1994 Feb 27	09:02.7–09:18.9^a	2,2,1	B	7671	M2.8	N08 W98	⋅⋅⋅	4
052	1994 Aug 30	08:20.5–08:41.4^a	1,1,1	B	7773	M1.1	S06 E82	⋅⋅⋅	19
053	1996 Apr 20	06:51.7–07:04.4	1,2,1	A	7956	B2.9	N04 W68	⋅⋅⋅	24,36,53
054	1996 Aug 22	07:42.5–07:52.1^a	2,2,1	B	7986	C4.5	S14 E107	+	3,50,52
055	1997 Feb 23	01:30.2–02:15.2^a	2,2?,1	B	8019	B7.2	(N31 E90+)	—	2
056	1997 Feb 23	02:58.2–03:31.0^a	2,1,1	C	8019	B7.2	(N33 E81)	+	4
057	1997 May 16	11:44.0–12:10.7^a	1,2,1	C	8038	—	(N20 W73)	+	4
058	1997 Aug 9	16:32.9–16:35.3^a	1,2?,1	C	8069	C8.5	N19 W85	+	16,19
059	1997 Sep 17	11:38.4–11:49.2	1,2,1	B	8084	M1.7	N21 W82	+	16,19
060	1997 Sep 17	17:48.4–18:24.6^a	2,2,2	B	8084	M1.0	N21 W84	+	19
061	1997 Nov 6	11:50.8–12:06.3	2,2,1	C	8100	X9.4	S18 W63	+	17,48
062	1997 Nov 14	09:11.0–09:20.8^a	2,1,1	B	8108	C2.5	N21 E70	+	24,48,53
063	1997 Nov 27	13:10.4–13:25.2	2,2,2	A	8113	X2.6	N17 E63	+	29,48
064	1998 Mar 23	02:45.9–03:12.1	2,2,2	B	8179	M2.3	S22 W99	+	16,19,50
065	1998 Mar 25	13:04.1–13:17.1	2,2,1	C	8180	C5.3	(S37 W90+)	+	2
066	1998 Apr 20	09:43.8–10:12.5^a	2,2?,2	C	8194	M1.4	S30 W90	+	17
067	1998 Apr 23	05:29.4–05:46.8	2,2,2	A	8210	X1.2	S18 E104	+	1,10,16,18,19,24,26,
									50,51,52,53
068	1998 Apr 24	08:47.6–08:55.2	2,1,1	C	8210	C8.9	S20 E91	+	19
069	1998 Apr 25	14:21.9–14:49.2	2,2,2	B	8210	C3.6	S19 E73	+	2
070	1998 Apr 27	08:50.0–08:55.3^a	2,2?,1	B	8210	X1.0	S16 E50	+	34
071	1998 May 3	21:17.4–21:25.5	2,2,2	A	8210	M1.4	S13 W34	+	30
072	1998 May 6	07:54.3–08:11.0	2,2,2	A	8210	X2.7	S11 W65	+	17,24,35,48,53
073	1998 May 8	01:50.4–02:20.1	2,2,2	A	8210	M3.1	(S16 W90+)	+	16,18,19
074	1998 May 8	14:21.2–14:49.4	2,2,2	B	8210	M1.8	S17 W95	+	19
075	1998 May 9	02:04.7–02:20.1	1,2,1	C	8210	C7.0	(S15 W90+)	—	16
076	1998 May 9	03:17.7–03:37.3	2,2,2	A	8210	M7.7	S17 W102	+	16,18,19,50
077	1998 May 10	13:25.4–13:33.2	2,2?,1	B	8220	M3.9	S27 E89	—	19
078	1998 May 28	06:22.5–06:28.2	1,1,1	B	8226	C1.2	N20 W78	—	19
079	1998 May 28	19:02.4–19:17.6	2,2,2	B	8226	C8.7	(N16 W87)	+	16,19
080	1998 May 29	01:02.8–01:12.0	1,2,1	C	8226	M6.7	(N15 W89)	+	19
081	1998 Jun 11	10:00.8–10:18.3^a	2,2,2	C	8243	M1.4	(N21 E89)	+	16,50
082	1998 Jun 15	07:26.8–07:34.3^a	2,2,1	C	8232	C1.4	(S16 W87)	+	2
083	1998 Jul 3	01:06.5–01:12.7	2,2?,1	C	8256	M1.2	S26 W14	⋅⋅⋅	4
084	1998 Aug 14	08:25.5–08:31.7	2,2,1	B	8293	M3.1	S23 W74	⋅⋅⋅	19
085	1998 Aug 18	08:17.5–08:26.5	2,2,2	B	8307	X2.8	N33 E68	⋅⋅⋅	18,19,48
086	1998 Aug 18	22:15.1–22:20.7	2,2,1	B	8307	X4.9	N33 E87	⋅⋅⋅	18,19,31,48
087	1998 Sep 6	02:12.3–02:18.5	1,1?,1	B	8323	C2.3	S22 W35	⋅⋅⋅	4
088	1998 Sep 20	02:38.5–03:19.1	2,2,2	B	8340	M1.8	N22 E62	⋅⋅⋅	4
089	1998 Sep 23	00:30.2–00:47.6	2,2,1	B	8344	C9.3	S20 E22	⋅⋅⋅	4
090	1998 Sep 23	06:45.3–07:05.2	2,2,2	B	8340	M7.1	N18 E09	⋅⋅⋅	40
091	1998 Sep 28	16:07.9–16:11.7	2,2,1	C	8339	C6.8	(S15 W87)	⋅⋅⋅	4
092	1998 Sep 30	13:16.4–13:40.5	1,2,2	B	8340	M2.8	N23 W81	⋅⋅⋅	4
093	1998 Oct 7	17:08.6–17:44.8^a	2,2,1	A	8355	M2.3	S23 E68	⋅⋅⋅	32
094	1998 Oct 20	20:34.5–21:16.1	2,2,2	C	8360	C7.4	(S21 W86)	⋅⋅⋅	4
095	1998 Nov 3	19:14.3–19:36.3	2,1?,2	B	8375	M1.0	N21 E02	+	4
096	1998 Nov 5	01:03.4–01:12.2	2,1,2	C	8375	C7.1	(N20 W14)	+	33
097	1998 Nov 6	02:42.1–02:51.8^a	1,1,1	C	8375	C4.4	N19 W24	—	4
098	1998 Nov 6	09:09.9–09:16.6	1,1?,1	C	8375	C4.2	N19 W27	+	4
099	1998 Nov 7	17:52.1–18:01.1	1,2?,1	B	8375	C5.3	N19 W48	—	4
100	1998 Nov 16	23:11.4–23:18.7	2,2,2	B	8385	C6.8	(N24 W83)	⋅⋅⋅	4
101	1998 Nov 22	06:39.6–06:47.2	2,2,2	C	8384	X3.7	S27 W82	⋅⋅⋅	28,34
102	1998 Nov 22	16:21.5–16:31.1	1,2,2	B	8384	X2.5	S30 W89	⋅⋅⋅	28,34
103	1998 Nov 24	02:19.6–02:28.3	2,2,1	C	8384	X1.0	S30 W103	+	4
104	1998 Nov 25	14:00.8^b–14:07.5	2,2,1	C	8395	C6.4	N18 E68	+	4
105	1998 Nov 28	05:29.5–05:49.6	2,2,2	B	8395	X3.3	N17 E32	+	34,48
106	1998 Dec 18	17:17.5–17:27.9^a	2,2,1	C	8414	M8.0	(N34 E40)	+	4
107	1998 Dec 23	05:53.5^b–06:02.9^a	2,1,1	C	8414	M2.3	(N28 E70)	⋅⋅⋅	4
108	1998 Dec 24	01:22.2–01:29.1^a	2,1,1	C	8421	C6.2	N29 E85	⋅⋅⋅	4
109	1998 Dec 25	06:18.4–06:35.0^a	2,1?,2	B	8421	M1.2	N30 E66	⋅⋅⋅	4
110	1998 Dec 28	05:16.2–05:31.9	2,1,2	B	8416	M1.4	N28 E26	⋅⋅⋅	4
111	1999 Jan 3	08:15.4–09:03.8^a	2,2,2	A	8420	C8.2	N16 W70	⋅⋅⋅	4
112	1999 Jan 6	23:59.7–00:05.4^a	1,2,1	B	8422	C8.0	(S23 W90)	⋅⋅⋅	4
113	1999 Jan 14	10:06.9–10:17.0	2,2,1	B	8439	M3.0	N18 E64	⋅⋅⋅	4
114	1999 Feb 10	23:09.1–23:25.3^a	2,2?,1	C	8457	C3.7	(N18 E57)	+	4
115	1999 Feb 12	15:23.7–15:32.1^a	2,2?,1	B	8456	C5.7	N12 E27	—	4
116	1999 Feb 16	21:19.7–21:27.8	2,2,2	C	8462	C5.3	N19 W12	⋅⋅⋅	4
117	1999 Feb 21	13:16.0–13:41.5^a	1,1?,2	B	8462	M1.3	N24 W81	—	4
118	1999 Mar 7	04:05.5–04:11.9	2,2,1	C	8477	C2.9	N24 W03	+	4
119	1999 Apr 3	23:02.0–23:10.5	1,1,1	B	8508	M4.3	N29 E81	+	9
120	1999 May 7	04:31.5–04:45.0	2,2,2	A	8535	M3.2	N20 E87	+	9,10
121	1999 May 8	14:33.0–14:52.1	1,1,1	C	8526	M4.6	N23 W75	+	9
122	1999 May 9	12:34.7–13:05.5	2,1?,2	C	8526	M1.0	(N22 W84)	+	9
123	1999 May 9	18:04.2–18:17.3	1,2,1	B	8526	M7.6	(N22 W86)	+	8,9
124	1999 May 11	21:46.9–22:02.7	2,1,1	C	8542	C4.7	S19 E79	+	9
125	1999 May 16	17:26.2–17:32.9^a	1,1,1	B	8534	M1.1	S17 W76	—	9
126	1999 May 17	03:43.2–03:51.6	1,1,2	B	8534	C3.9	(S16 W82)	—	8,9,24,37,53
127	1999 May 17	04:57.5–05:01.0^a	1,1,1	C	8534	M2.3	(S15 W82)	—	9
128	1999 May 29	03:08.1–03:14.2^a	2,2,1	B	8557	M1.6	(S21 E65)	+	9
129	1999 Jun 11	11:25.4–11:38.4^a	2,2,1	B	8585	C8.8	(N46 E90)	+	9
130	1999 Jun 19	22:31.8–22:42.8^a	2,1,1	C	8592	C4.0	(N25 E86)	+	9
131	1999 Jun 23	00:41.2–00:51.0^a	2,1,1	B	8583	C7.9	S12 W78	+	9
132	1999 Jun 30	03:11.0–03:18.3	1,1,1	C	8613	C4.5	(N19 E90)	+	9
133	1999 Jun 30	04:36.2–05:00.8^a	2,2,2	B	8611	M2.1	S26 E28	+	4
134	1999 Jul 7	06:20.7–06:30.3	1,2?,1	C	8611	C1.2	S26 W76	+	9
135	1999 Jul 7	09:19.1^b–09:24.4^a	1,1,1	C	8611	C1.2	(S26 W76)	+	9
136	1999 Jul 9	22:38.5–22:48.2	1,1,1	C	8629	C6.9	N22 W69	—	9
137	1999 Jul 12	23:31.9–23:38.0	1,1,1	B	8626	C2.1	(S19 W76)	—	9
138	1999 Jul 16	15:50.8–15:55.3^a	1,1,1	C	8635	M3.1	N43 W71	+	9
139	1999 Jul 23	04:45.9–05:03.8^a	2,1,1	B	8645	C9.3	S23 E97	—	9
140	1999 Jul 23	15:56.1–16:26.3^a	1,1,1	C	8645	M1.0	S26 W92	—	9
141	1999 Jul 23	22:58.1–23:10.9	2,1,1	C	8644	M1.2	S26 E87	—	9
142	1999 Jul 24	03:31.7–04:06.6	2,1,1	C	8636	M1.7	S29 E87	—	9
143	1999 Jul 24	08:00.4–08:16.7^a	2,1?,1	B	8636	M3.3	S28 E78	+	9
144	1999 Jul 25	13:08.8^b–13:14.7^a	2,2,1	B	8639	M2.4	N38 W81	+	8,9,10,24,38,53
145	1999 Aug 4	05:58.5–06:11.5	2,1,1	C	8647	M6.0	S16 W64	+	9
146	1999 Aug 4	21:54.2–22:03.2^a	1,1,1	C	8647	C3.3	(S18 W80)	—	9
147	1999 Aug 6	10:05.3–10:18.0^a	2,1,1	C	8647	C7.1	S28 W81	+	9
148	1999 Aug 7	20:52.0–21:10.9	2,2,2	B	8645	M1.7	S29 W104	+	9
149	1999 Aug 10	06:02.4–06:13.2^a	1,1,1	C	8656	C4.8	N15 W74	—	9
150	1999 Aug 13	15:07.9–15:16.2^a	1,1,1	B	8668	C5.3	(N17 E90)	—	9
151	1999 Aug 14	12:06.0–12:16.9^a	1,1,1	C	8668	C7.2	N23 E72	—	9
152	1999 Aug 20	12:40.4–13:36.6^a	2,2?,2	C	8674	M1.8	S28 E76	+	9
153	1999 Aug 28	17:54.1–18:04.6^a	2,2,2	B	8674	X1.1	S26 W14	+	48
154	1999 Sep 8	12:13.6–12:14.5^a	2,2?,1	C	8690	M1.4	N12 E53	+	25
155	1999 Sep 21	10:16.0–10:34.6^a	2,2,1	B	8692	C6.4	(S25 W84)	+	9
156	1999 Oct 25	19:37.8–19:44.1	2,2,1	C	8737	C4.8	S19 W69	+	9
157	1999 Oct 26	21:12.0–21:27.6^a	2,2,2	B	8737	M3.7	(S16 W86)	+	9,10,11,18,28,39
158	1999 Oct 27	09:11.2–09:13.9^a	2,2,1	B	8737	M1.0	S12 W88	—	9
159	1999 Oct 27	13:35.2–14:04.9^a	2,2,1	B	8737	M1.8	S15 W90	+	9
160	1999 Oct 27	15:21.1–15:43.6^a	2,1,1	B	8737	M1.4	S14 W92	+	9
161	1999 Nov 5	18:22.5–18:48.6	2,2?,1	B	8759	M3.0	N12 E96	+	9,10
162	1999 Nov 6	06:32.5–06:35.1	1,1,1	C	8759	C4.6	(N12 E86)	—	8,9
163	1999 Nov 6	17:04.4–17:14.4	1,1,1	B	8759	C5.0	(N12 E86)	+	9
164	1999 Nov 8	06:06.0–06:16.0^c	⋅⋅⋅	D	8749	C5.9	S18 W82	+	9
165	1999 Nov 13	02:22.7–03:09.4	2,2?,1	C	8763	M1.3	S15 E44	+	9
166	1999 Nov 27	12:12.1–12:23.3	1,1,1	C	8771	X1.4	S15 W68	+	9,48,49
167	1999 Dec 18	01:26.8–01:39.1	2,2,1	C	8806	C9.4	N19 E82	—	9
168	2000 Jan 12	20:50.0–20:56.9	2,2,2	B	8829	M1.1	N13 E67	+	9
169	2000 Jan 18	09:36.0–09:58.8^a	1,2,1	A	8827	M1.2	S15 W106	+	9,10,24,53
170	2000 Jan 18	17:12.1–17:18.1	2,2?,1	C	8831	M3.9	S19 E11	+	4
171	2000 Jan 22	17:58.9^b–18:08.8^a	2,1,2	C	8831	M1.0	S23 W50	+	4
172	2000 Feb 4	09:14.6–09:57.4	2,1,1	C	8858	M3.0	N25 E71	—	28
173	2000 Feb 4	19:28.3–19:43.5^a	1,2,2	B	8858	C7.0	N25 E71	+	9
174	2000 Feb 5	19:33.9–19:43.9	2,1?,2	B	8858	X1.2	N26 E52	+	48
175	2000 Feb 22	20:26.2–21:24.0	2,1,1	C	8882	C9.2	(S18 E90+)	+	9
176	2000 Feb 26	23:38.5–23:43.8^a	2,2,2	B	8889	M1.0	N29 E50	—	4
177	2000 Mar 2	08:23.7–08:28.7	2,2,1	B	8882	X1.1	(S18 W55)	+	48,49
178	2000 Mar 2	13:12.5–13:20.5	1,2,1	C	8882	C5.5	S19 W60	+	4
179	2000 Mar 3	02:11.6–02:14.6^a	2,2,1	B	8882	M3.8	S15 W60	+	4
180	2000 Mar 6	10:47.9–10:50.0^a	1,1,1	B	8889	C4.5	(N20 W75)	—	9
181	2000 Mar 6	16:20.0–16:28.4^a	2,1,1	B	8889	C3.9	N20 W78	—	9
182	2000 Mar 18	20:50.7–20:57.3^a	2,2,1	B	8906	M2.1	(S15 W68)	+	4
183	2000 Mar 18	21:53.5–21:59.0	1,1,1	B	8906	C4.2	S19 W67	+	4
184	2000 Mar 18	23:16.9–23:27.4^a	2,1,1	B	8902	+^d	(S18 E90+)	+	9
185	2000 Mar 27	13:59.1–14:25.5^c	⋅⋅⋅	D	8926	C8.4	S09 W69	—	9
186	2000 Mar 27	15:32.8^b–15:41.9	2,1,1	C	8926	C8.9	S10 W69	—	24
187	2000 Mar 31	06:27.8–06:50.8	2,2?,1	B	8936	M1.2	S15 E55	+	9
188	2000 Apr 6	02:22.4–02:31.9	2,2,1	B	8948	M1.8	S15 E53	+	4
189	2000 Apr 8	02:37.5–02:47.5	2,2,1	C	8948	M2.0	S15 E26	+	4
190	2000 May 2	14:53.2–15:03.0	1,1,2	C	8971	M2.8	N22 W68	+	9
191	2000 May 5	15:18.3–15:47.5	2,2,2	B	8977	M1.5	S18 W110	+	9
192	2000 May 12	08:40.9–08:50.7^c	⋅⋅⋅	D	8998	C8.1	S14 E90	—	9,41
193	2000 May 12	21:32.8–21:39.8^c	⋅⋅⋅	D	8998	C9.8	S15 E81	—	9
194	2000 May 13	01:46.6–02:10.1	2,2,1	C	9002	M1.1	N22 E109	—	9
195	2000 May 13	23:12.4–23:21.8^a	2,1,1	C	9002	C7.4	N22 E96	+	9
196	2000 May 15	08:26.8–08:53.1	2,2,1	C	9002	M4.4	(N23 E87)	+	9
197	2000 May 18	15:55.6–16:05.0	2,2?,1	C	9002	M2.7	N23 E30	+	4
198	2000 May 23	17:50.1–18:00.1	1,1,1	C	8996	C4.3	S22 W80	+	9
199	2000 May 24	00:09.2–00:19.1	1,1,1	C	9017	C6.8	(S12 E90+)	—	9
200	2000 May 24	03:14.3–04:02.8	1,1,1	C	9017	C7.0	S12 E93	—	9
201	2000 May 24	11:43.3–12:13.0	1,1,1	C	9017	M1.1	(S12 E90)	—	9
202	2000 May 24	21:05.6–21:51.1^a	1,1,1	C	9017	C9.7	(S12 E90)	—	9
203	2000 May 26	11:33.1–11.37.6^a	2,2?,1	B	8998	C6.1	S13 W90	+	9
204	2000 May 28	10:23.6–10:34.6	1,1,1	C	9002	C8.6	(N21 W88)	+	9
205	2000 Jun 2	03:44.6–03:50.8^a	2,1,1	B	9026	M1.2	(N22 E77)	+	9
206	2000 Jun 2	20:32.7–21:10.3	2,1,1	B	9026	M3.1	(N21 E67)	+	9
207	2000 Jun 6	15:35.6–15:45.6	1,1,1	C	9026	X2.3	(N20 E18)	+	4
208	2000 Jun 7	15:42.9–15:52.9^a	2,2?,1	B	9026	X1.2	N23 E03	+	4
209	2000 Jun 12	12:24.4–12:57.5	2,1,1	C	9042	C6.4	N16 E87	—	9
210	2000 Jun 17	02:29.6–02:45.3	1,1,2	C	9033	M3.5	N22 W72	+	9
211	2000 Jun 18	02:04.4–02:14.4^a	1,2,1	C	9033	X1.0	N23 W85	+	9
212	2000 Jun 19	04:18.1–04:23.0	1,1,1	C	9033	C1.7	(N21 W89)	+	9
213	2000 Jun 21	09:24.5–09:34.1	1,1,1	C	9042	M1.3	N24 W42	+	4
214	2000 Jun 23	14:22.7–14:37.9	2,2,1	B	9042	M3.0	N26 W72	+	9
215	2000 Jun 23	22:14.9–22:24.8	1,1?,2	C	9042	C7.7	N22 W74	+	9
216	2000 Jun 28	12:17.4–12:19.5^a	2,1?,1	B	9064	C6.1	(S22 W73)	—	9
217	2000 Jun 29	10:28.6–10:38.7	1,1,1	B	9064	C4.8	(S17 W82)	—	9
218	2000 Jul 1	12:37.6–12:43.1^a	1,1,1	B	9054	C6.0	(N15 W75)	—	9
219	2000 Jul 1	23:21.4–23:31.3	2,1,1	C	9054	M1.5	N07 W88	—	9
220	2000 Jul 10	21:14.7–21:22.7	2,2,2	B	9077	M5.7	N18 E49	+	4
221	2000 Jul 12	03:32.6–03:37.5	1,1,1	C	9078	M1.4	(S11 E87)	+	9
222	2000 Jul 12	10:26.4–10:38.7	2,2,1	C	9077	X1.9	N17 E27	+	48
223	2000 Jul 12	16:24.9–16:32.7	2,1,1	C	9070	M1.0	N17 W68	+	9
224	2000 Jul 12	19:46.1–20:18.1	2,2,1	C	9077	M1.5	(N17 W72)	+	9
225	2000 Jul 13	02:04.5–02:14.2^a	1,1,1	B	9069	C6.1	S16 W71	—	9
226	2000 Jul 13	06:59.8–07:09.3	2,1,1	C	9070	C6.8	(N16 W80)	—	9
227	2000 Jul 13	18:14.1–19:04.8^c	⋅⋅⋅	D	9070	M1.3	(N16 W83)	—	9
228	2000 Jul 14	00:42.0–00:44.6	1,1,1	B	9070	M1.5	(N17 W86)	—	9
229	2000 Jul 14	10:20.0–10:27.2	2,2,2	B	9077	X5.7	N22 W07	+	48,49
230	2000 Jul 14	13:45.1–13:55.0	2,2,1	B	9077	M3.7	N20 W08	—	4
231	2000 Jul 16	01:24.5–01:32.5	2,1,1	B	9087	C6.3	S11 E53	—	4
232	2000 Jul 18	05:02.9–05:10.5^a	2,2,2	B	9077	M1.8	N17 W58	⋅⋅⋅	4
233	2000 Jul 20	09:46.1–10:18.7	2,2?,2	C	9087	M3.6	S12 W08	⋅⋅⋅	24
234	2000 Jul 21	14:33.3–14:40.6	2,1,2	C	9090	M5.5	(N10 E12)	—	4
235	2000 Jul 22	11:21.1–11:27.1	2,2,1	B	9085	M3.7	N14 W56	+	4
236	2000 Jul 25	02:46.8–02:52.4	2,2,1	B	9097	M8.0	N06 W08	+	42
237	2000 Jul 26	03:54.6–04:13.6	1,1,1	C	9087	C8.9	S13 W89	—	9
238	2000 Jul 27	04:08.1–04:17.9	2,1,1	B	9090	M2.4	N10 W72	—	9
239	2000 Jul 27	16:46.0–16:52.6^a	2,1,1	B	9087	M1.5	S09 W105	—	9
240	2000 Aug 2	08:17.8–08:27.1	1,1,1	B	9114	C7.9	(N10 E85)	+	9
241	2000 Aug 12	09:49.6–10:00.5	2,2,2	B	9119	M1.1	(S15 W84)	+	9
242	2000 Aug 14	05:01.1–05:10.8^a	1,1,1	C	9126	C8.1	N06 W75	+	9
243	2000 Aug 24	09:01.8–09:11.6^a	2,1,1	C	?	C6.2	(N27 W88)	—	9
244	2000 Aug 25	14:27.7–14:32.5	2,2,1	B	9143	M1.4	S15 E67	+	7,9,25,43
245	2000 Sep 7	20:38.0–20:47.6	1,1,2	B	9151	C7.2	N06 W47	+	44
246	2000 Sep 15	14:41.2–14:43.9^a	1,2?,1	C	9165	M2.0	N12 E07	+	24
247	2000 Sep 22	23:46.9–23:56.5^a	1,1,1	C	9165	C8.5	N14 W94	—	9
248	2000 Sep 30	17:53.4–18:21.1	2,1,1	B	?	M1.0	(S29 E85)	+	9
249	2000 Sep 30	23:17.3–23:29.0	2,2,1	B	9169	X1.2	N07 W90+	—	9,18,28,45,48
250	2000 Oct 1	07:01.7–07:14.2^a	2,1,1	C	9169	M5.0	N08 W97	—	9
251	2000 Oct 1	13:59.2–14:06.5^a	2,1,1	A	9169	M2.2	N09 W101	—	9
252	2000 Oct 16	05:35.1–06:03.9	2,2,1	A	9182	C7.0	N04 W107	+	9,24,53
253	2000 Oct 16	06:42.5^b–06:46.9^a	2,2?,1	B	9182	M2.5	N03 W108	+	9
254	2000 Oct 26	05:10.2–05:18.2^a	1,1,1	B	9199	C3.7	(N14 W79)	+	8,9
255	2000 Oct 26	06:11.2–06:20.7^a	2,1,1	C	9209	C4.3	S25 E71	—	9
256	2000 Oct 26	11:43.9–11:46.9^c	⋅⋅⋅	D	9203	C6.9	N17 W77	+	9
257	2000 Oct 26	15:59.1–16:35.9	2,1?,2	B	9209	C8.5	S20 E64	+	9
258	2000 Oct 29	01:32.8–01:57.4	2,2,2	B	9209	M4.4	S25 E35	⋅⋅⋅	4
259	2000 Nov 1	12:04.5^b–12:21.1	1,1,1	C	9212	—	(N10 E29)	—	24
260	2000 Nov 8	23:18.2–23:44.0	2,1,1	C	9213	M7.4	(N10 W77)	+	9,17
261	2000 Nov 9	03:03.2–03:12.9^a	2,1?,1	C	9213	M1.2	(N08 W86)	—	9
262	2000 Nov 9	06:28.9–06:38.6^a	1,1,1	C	9213	M1.2	(N12 W88)	—	9
263	2000 Nov 14	16:26.2–16:35.9	2,2,2	B	9232	M1.0	(N14 E90)	+	9
264	2000 Nov 18	16:48.3–16:57.9^a	2,1,1	C	9227	C5.9	(S16 W79)	—	9
265	2000 Nov 24	14:55.0–15:20.5	2,2,2	A	9236	X2.3	N22 W07	+	46,48
266	2000 Nov 24	21:50.2–22:03.5	2,2,2	B	9236	X1.8	N21 W14	+	15,48
267	2000 Nov 25	01:00.1–01:10.4	2,2,2	A	9240	M8.2	N07 E50	+	4
268	2000 Nov 25	09:10.0–09:21.0	2,2,2	B	9236	M3.5	N18 W24	+	4
269	2000 Nov 25	18:36.5–18:46.4	2,1,1	C	9236	X1.9	N20 W23	+	48
270	2000 Nov 26	16:38.3–16:47.4	2,1,2	B	9236	X4.0	N18 W38	+	48
271	2000 Nov 30	08:58.0–09:23.5^a	2,2,1	A	9236	M1.0	(N17 W88)	⋅⋅⋅	4
272	2000 Dec 6	22:22.0–22:47.5^a	2,2?,1	C	9246	M1.6	S10 W66	+	9
273	2000 Dec 18	08:58.6–09:06.2^a	1,1,1	C	9276	C5.2	S14 W76	—	9
274	2000 Dec 19	10:23.5–11:05.1	1,1,1	C	9276	+^d	(S12 W85)	+	9
275	2000 Dec 23	04:57.9–05:07.5^a	1,1,1	C	9283	C3.4	S12 E76	—	9
276	2000 Dec 23	08:20.2–08:23.1^a	1,1,1	C	9283	C3.7	S13 E75	—	9
277	2000 Dec 24	01:06.7–01:15.0	2,1,1	C	9283	C7.0	S15 E66	+	9
278	2000 Dec 26	23:47.9–23:56.1	1,1,1	B	9289	C4.0	(S06 E87)	—	9
279	2000 Dec 27	15:52.0–16:05.3	2,1,1	C	9289	M4.3	S07 E73	—	9
280	2001 Jan 4	08:56.9–09:06.0^a	2,1,1	B	9302	C4.5	N25 E87	—	9
281	2001 Jan 5	18:29.5–18:39.3^a	2,1,1	A	9302	C5.8	N20 E72	—	9
282	2001 Jan 8	10:36.5–10:53.3^a	2,1?,1	B	9302	C5.1	(N21 E36)	+	24
283	2001 Jan 9	08:50.0–08:59.5^a	2,1,1	B	9297	C5.1	(N21 W87)	—	9
284	2001 Jan 19	17:09.2–18:08.4	1,1,1	C	9313	M1.0	S07 E61	+	9
285	2001 Jan 24	14:42.7–15:11.4^c	⋅⋅⋅	D	9311	M1.0	N06 W77	—	9
286	2001 Jan 25	07:10.8–07:18.5^a	2,2,1	C	9325	C7.4	(N10 E74)	+	9
287	2001 Jan 30	00:57.1–01:03.2^a	2,2,1	C	9313	C3.7	(S08 W89)	+	9
288	2001 Feb 19	20:55.4–21:05.1^a	2,1,1	B	9360	C5.4	(S09 E89)	+	9
289	2001 Mar 6	10:10.1–10:19.7	2,2,1	B	9364	C6.7	(S11 W89)	+	9
290	2001 Mar 7	14:49.2–14:59.1^a	2,1,1	C	9371	C5.8	N23 W75	+	9
291	2001 Mar 9	19:54.4–20:00.5	1,1,1	C	9371	C4.2	(N22 W84)	+	9
292	2001 Mar 10	17:14.3–17:23.8^a	1,1,1	C	9365	C5.9	(S12 W90)	+	9
293	2001 Mar 11	08:47.4–08:57.2^a	2,1,1	B	9376	C5.0	S14 E87	+	9,24
294	2001 Mar 21	02:36.6–02:44.4	2,2?,2	B	9373	M1.8	S05 W65	+	9
295	2001 Mar 21	11:25.2–11:31.5	1,1,1	B	9373	C9.8	S05 W70	+	9
296	2001 Mar 24	01:36.5–01:40.5	2,2?,1	C	9376	M1.2	S14 W82	+	9
297	2001 Mar 24	23:32.6–00:08.8	1,1,1	C	9393	M1.1	N19 E60	—	9
298	2001 Mar 28	10:56.0–11:05.7^a	2,1,1	B	9397	M4.3	S09 E29	+	24
299	2001 Mar 29	01:37.4–01:58.0^a	1,1,1	C	9393	—	(N15 W09)	—	24
300	1991 Mar 29	12:47.1–13:02.3	2,1,1	C	9393	C7.6	(N16 W11)	—	24
301	2001 Apr 1	10:59.8–11:03.5^a	1,2,1	B	9415	M5.5	S21 E107	+	4
302	2001 Apr 3	03:37.6–03:49.3	1,1?,1	C	9415	X1.2	S21 E83	+	4
303	2001 Apr 5	08:30.7–08:49.5	2,2,1	B	9393	M8.4	N14 W103	+	24,53
304	2001 Apr 6	19:12.5–19:23.5^a	2,2,2	B	9415	X5.6	S21 E31	+	48
305	2001 Apr 9	15:23.6–15:24.8^a	2,1,1	C	9415	M7.9	S21 W04	+	24
306	2001 Apr 10	05:19.2–05:33.9	1,1,1	C	9415	X2.3	S23 W09	+	4
307	2001 Apr 12	10:12.8–10:22.0	2,2,1	B	9415	X2.0	S19 W43	+	47,48
308	2001 Apr 15	13:29.7–13:53.0	2,2,2	B	9415	Y1.4	S20 W85	+	17,18,48
309	2001 Apr 17	12:15.4–12:20.7^a	2,2?,1	C	9415	C1.9	(S21 W90)	—	4
310	2001 Apr 20	21:30.7–21:35.5^a	2,2,2	B	9433	C8.0	(N17 E45)	⋅⋅⋅	4
311	2001 Apr 23	10:15.7–10:23.8^a	2,1?,1	B	9433	C9.1	N17 E12	—	4
312	2001 Apr 24	05:36.1–05:52.0	2,1,2	B	9433	M2.1	N18 E01	—	4
313	2001 Apr 25	13:43.5–13:56.7	1,1,2	B	9433	M2.7	N18 W09	+	4
314	2001 Apr 26	13:10.7^b–13:18.2^a	1,1,1	C	9433	M7.8	N17 W31	+	4
315	2001 May 8	00:41.8–01:09.9^a	2,1,1	B	9445	C9.9	N23 W43	—	4
316	2001 May 12	23:26.8–23:39.7	2,2,2	B	9455	M3.0	S17 E00	—	4
317	2001 May 15	02:57.4–03:04.4	2,2,1	B	9455	M1.0	S17 W29	+	4
318	2001 Jun 19	23:19.8–23:29.5^a	2,1,1	B	9501	C4.2	S10 W37	+	4
319	2001 Jun 22	20:25.5–20:35.1	2,2,1	B	9511	C5.5	N09 E28	—	4
320	2001 Jul 16	03:17.0–03:23.3^a	2,2?,1	C	9539	M1.2	S18 W20	+	4
321	2001 Jul 23	06:22.7–06:29.5^a	2,1,1	B	9545	C5.0	N10 W65	⋅⋅⋅	4
322	2001 Jul 30	20:40.6–20:47.7	2,2,1	B	9562	C6.0	(N05 E79)	⋅⋅⋅	4
323	2001 Jul 31	04:02.5–04:09.4	1,1,1	B	9562	C6.0	(N05 E76)	⋅⋅⋅	4
324	2001 Aug 8	07:09.4–07:17.7	1,1,1	B	9557	C3.9	(S18 W86)	+	4
325	2001 Aug 9	18:27.7–18:37.4	2,2?,2	B	9570	C7.8	S17 E19	—	4
326	2001 Aug 11	01:19.3–01:28.0	1,1,1	B	9563	C5.2	(N20 W83)	+	4
327	2001 Aug 25	16:25.8^b–16:34.6	1,1,2	B	9591	X5.3	S17 E34	+	5
328	2001 Aug 26	13:18.9^b–13:58.9^a	2,1,1	C	?	M1.3	(N16 E89)	+	4
329	2001 Aug 31	10:38.1–10:47.8^a	2,2,1	B	9601	M1.6	N15 E37	+	4
330	2001 Sep 2	13:44.8–14:04.5^a	2,1,2	B	9591	M3.0	S20 W53	—	24,53
331	2001 Sep 3	18:19.1–18:34.0	2,2,2	B	9608	M2.5	S22 E96	+	4
332	2001 Sep 7	15:28.9–15:33.8^a	1,1,1	C	9601	M1.2	N19 W65	+	4
333	2001 Sep 8	16:43.6–16:49.7	2,1,1	B	9608	C5.1	S23 E32	—	4
334	2001 Sep 12	21:44.0–21:48.6^a	1,2,1	C	9606	C9.6	(S17 W63)	+	4
335	2001 Sep 13	19:50.7^b–19:54.2	1,2?,1	B	9606	C5.8	(S18 W79)	—	4
336	2001 Sep 14	21:44.6–21:45.9^a	2,2,1	B	9616	M3.7	(S14 E39)	+	4
337	2001 Sep 17	08:21.0–08:29.8	2,2,2	B	9616	M1.5	S14 E04	+	4
338	2001 Sep 17	21:04.7–21:10.2	2,2,1	C	9616	M1.0	S11 W06	—	4
339	2001 Sep 20	18:15.3–18:18.7	2,2,2	B	9631	M1.5	N09 W11	+	4
340	2001 Sep 21	04:52.8–05:01.5	2,2,1	B	9620	C4.1	N10 E12	—	4
341	2001 Sep 22	18:09.9–18:15.6^a	2,2?,1	C	9633	C5.4	(N14 E88)	—	4
342	2001 Sep 24	09:50.1–10:13.7^a	2,1,1,	C	9632	X2.6	S18 E27	+	17
343	2001 Sep 30	11:34.1–11:41.0^a	1,1,1	B	9628	M1.0	S20 W75	—	28
344	2001 Oct 1	04:45.8–04:50.8^a	1,1,2	B	?	C4.7	(N19 W89)	—	28
345	2001 Oct 1	04:57.9–05:17.9^a	2,2,2	B	9628	M9.1	(S21 W84)	+	28
346	2001 Oct 2	17:11.5–17:50.5	2,1,2	A	9628	C4.7	(S21 W85)	—	4
347	2001 Oct 3	06:42.5–06:47.4	1,2,2	B	9636	C6.1	N19 W46	+	4
348	2001 Oct 9	07:36.8–07:46.7	2,2,2	B	9645	C7.0	(S19 W86)	+	4
349	2001 Oct 19	16:25.2–16:35.9	2,2,2	B	9661	X1.6	N15 W29	+	4
350	2001 Oct 20	21:11.2–21:17.6	2,2,1	B	9674	C4.6	S09 E24	—	4
351	2001 Oct 22	00:38.0–00:43.4	1,1,1	C	9658	M1.0	S16 W86	+	4
352	2001 Oct 22	14:40.6–14:58.5	2,2,2	B	9672	M6.7	S21 E18	+	4
353	2001 Oct 22	17:52.1–18:01.8	2,2,2	C	9672	X1.2	S18 E16	+	4
354	2001 Oct 25	22:52.6–23:00.9	1,2,2	B	9678	C6.1	(N08 E24)	—	4
355	2001 Oct 29	08:14.6–08:21.7	1,1,1	C	9672	M1.0	S18 W82	+	4
356	2001 Nov 1	14:03.6–14:29.6	2,2?,2	B	9687	M1.7	S19 E77	+	2
357	2001 Nov 4	16:05.3–16:19.8	2,2,2	B	9684	X1.0	N06 W18	+	17
358	2001 Nov 6	03:01.3^b–03:03.6^a	2,1,1	B	9687	M2.0	S19 E10	—	4
359	2001 Nov 8	15:12.9–15:45.1	2,1?,1	B	9690	M4.2	S17 E36	⋅⋅⋅	4
360	2001 Nov 9	18:32.3–18:42.1	1,1,1	C	9687	M1.9	S21 W42	⋅⋅⋅	4
361	2001 Nov 17	05:17.3–05:49.3	2,2,1	B	9704	M2.8	S13 E42	+	4
362	2001 Nov 28	15:41.4–15:44.4^a	1,2,1	C	9715	C7.7	(N05 E17)	+	4
363	2001 Nov 28	15:51.9–16:03.1	1,1,1	C	9715	C2.1	(N05 E17)	+	4
364	2001 Nov 28	16:14.9–16:22.3^a	1,1,1	C	9715	M6.9	N04 E16	+	4
365	2001 Nov 29	01:45.7–01:49.0^a	2,2,1	C	9715	M1.1	N04 E12	—	4
366	2001 Nov 30	01:03.7–01:08.8	2,1,2	B	9718	M3.5	S06 E57	—	4
367	2001 Dec 2	21:38.0–22:25.5	2,2,2	B	9714	M2.0	(S09 W88)	+	4
368	2001 Dec 10	22:46.1–22:55.9	2,1,2	B	9733	C7.0	N10 E52	—	4

Notes. ^aLater event end. ^bEarlier event start. ^cNo identification, time interval of available SXT observations. ^dGOES class disturbed by a flare in another active region. References. (1) Alexander et al. 2002; (2) U. Bak-Steślicka 2010, private communication; (3) Chertok 2000; (4) Chmielewska 2010; (5) Falewicz et al. 2002; (6) Hori 1999; (7) Khan et al. 2002; (8) Kim et al. 2004; (9) Kim et al. 2005b; (10) Kim et al. 2005a; (11) Kim et al. 2009; (12) Kliem et al. 2000; (13) Klimchuk et al. 1994; (14) Kundu et al. 2001; (15) Nishizuka et al. 2010; (16) Nitta & Akiyama 1999; (17) Nitta et al. 2003; (18) Nitta et al. 2010; (19) M. Ohyama 2009, private communication; (20) Ohyama & Shibata 1997; (21) Ohyama & Shibata 1998; (22) Ohyama & Shibata 2000; (23) Ohyama & Shibata 2008; (24) Ronowicz 2007; (25) Saint-Hilaire & Benz 2003; (26) Shanmugaraju et al. 2006; (27) Shibata et al. 1995; (28) Shimizu et al. 2008; (29) SXT SN 1997/11/28 (Nitta); (30) SXT SN 1998/05/08 (McKenzie & Hudson); (31) SXT SN 1998/08/22 (Alexander); (32) SXT SN 1998/10/09 (McKenzie); (33) SXT SN 1998/11/06 (Hudson); (34) SXT SN 1998/11/27 (Hudson); (35) SXT SN 1998/12/25 (Hudson & Akiyama); (36) SXT SN 1999/04/04 (Akiyama); (37) SXT SN 1999/05/21 (McKenzie); (38) SXT SN 1999/09/11 (Nitta); (39) SXT SN 1999/10/29 (McKenzie & Fletcher); (40) SXT SN 2000/04/18 (Hudson); (41) SXT SN 2000/05/12 (Hudson); (42) SXT SN 2000/07/28 (Hudson); (43) SXT SN 2000/09/01 (Fletcher & Hudson); (44) SXT SN 2000/09/08 (Hudson); (45) SXT SN 2000/10/06 (Handy); (46) SXT SN 2000/12/22 (Nitta); (47) SXT SN 2001/04/13 (Hudson); (48) SXT SN 2001/08/17 (Nitta); (49) SXT SN 2002/10/04 (Nitta); (50) Tomczak 2003; (51) Tomczak 2004; (52) Tomczak 2005; (53) Tomczak & Ronowicz 2007; (54) Tsuneta 1997.

Download table as: ASCIITypeset images: 1 2 3 4 5 6

In Table 2, the quality of XPE observations from the catalog is summarized. The most frequent are events that we categorized as (B) and (C). Contribution of remaining categories is marginal. In Sections 3 and 4, we present results of a statistical analysis that was performed for two different populations of events: from (A) to (C) and from (A) to (B). The first population is more frequent, which offers some advantages in statistical approach; however, for events categorized as (C) observations are often not complete enough to give confidence in our classification choices. In consequence, in the first population an additional bias can be introduced, which is inadvisable. We expect that this problem is overcome for the less frequent, second population of XPEs, which was better observed.

Table 2. Quality of the Observed XPEs

Quality	Quantity
A (excellent)	20/368 (5.4%)
B (good)	190/368 (51.6%)
C (poor)	149/368 (40.5%)
D (problematic)	9/368 (2.5%)

Download table as: ASCII Typeset image

In Table 3, we summarize heliographic longitudes of XPEs in the catalog. A distinct concentration of the XPEs around the solar limb is seen. This is an artificial effect caused by observational constraints. XPEs are easier to detect when we observe them against the dark background sky than when we observe them between plenty of different features seen on the solar disk. Moreover, in two out of three main surveys that we used in our catalog (Kim et al. 2005b; M. Ohyama 2009, private communication), only flares that occurred close to the solar limb (|λ| > 60°) were systematically reviewed.

Table 3. Location of the Observed XPEs

Heliographic	Quantity
Longitude (\|λ\|)
<60°	106/368 (28.8%)
60°–90°	218/368 (59.2%)
>90°	44/368 (12.0%)

Download table as: ASCII Typeset image

Is it possible to estimate the actual number of XPEs that occurred on the Sun during the Yohkoh years? Assuming their uniform distribution with heliographic longitude and taking the number for the interval 60° ⩽ |λ| ⩽ 90° as the most representative, we obtain a value 6 × 218 ≈ 1300. Including a duty time of Yohkoh to be about 0.65 (ratio of satellite day to the total orbital period), we obtain a number 2 × 10³.

However, even this huge number could be significantly lower than actual, for several reasons. First, we do not include the influence of worse detection conditions before 1997. Second, for strong flares, especially in 2001, the conditions for detecting XPEs were quite bad because of too long exposures. Third, the estimated number is roughly representative for flares stronger than the GOES class C5–C6. Only for those events was the flare mode initiated in the Yohkoh operation (Tsuneta et al. 1991), and this mode guarantees a sufficient time resolution of an image cadence for a successful detection of XPEs. XPEs associated with weaker flares are only known accidentally, since no systematic examination of images recorded during the quiet mode of the Yohkoh satellite has been performed yet.

In conclusion, XPEs should be considered as very frequent events occurring in the solar corona. XPEs described in the catalog are only a minor representation of a countless population of events, which are typical for the hot solar corona.

In Table 4, we present a time coverage of the observed XPEs. Evolution of an important fraction of events (42.6%) is illustrated only partially. This limits its detailed investigation. Even relatively simple activities like classification can be meaningless. For example, XPEs classified as single and observed only partially can be actually recurrent.

Table 4. Time Coverage of the Observed XPEs

	Quantity
Full	206/359 (57.4%)
Partial	153/359 (42.6%)

Download table as: ASCII Typeset image

In Table 5, we present a number of XPEs that were classified in one of eight subclasses defined under the following three criteria: morphological, kinematical, and recurrence. We organize the results twofold: for the total population and for carefully selected events. In the second case we omitted events categorized as (C), untrustworthy assignments of kinematical criterion as shown in the catalog with a question mark, and examples classified as single in the case of partial time coverage of observations. It reduces the whole population almost three times and for particular subclasses even more, but we believe that numbers less affected by observational limits, seen in the last column of Table 5, are more representative for real conditions.

Table 5. Population of Particular Subclasses of XPEs

		Quantity
Subclass	Description	Total	Special Selection
		(T)	(SS)
1,1,1	Collimated, confined, single	74	14
1,1,2	Collimated, confined, recurrent	10	5
1,2,1	Collimated, eruptive, single	24	6
1,2,2	Collimated, eruptive, recurrent	6	5
2,1,1	Loop-like, confined, single	72	7
2,1,2	Loop-like, confined, recurrent	15	7
2,2,1	Loop-like, eruptive, single	94	30
2,2,2	Loop-like, eruptive, recurrent	64	52

	Total	359	126

Download table as: ASCII Typeset image

What do these numbers tell us about XPEs? At first sight, loop-like XPEs seem to be more frequent than collimated XPEs by a factor of 2.1 or 3.2 for total population and special selection, respectively. However, it can be caused by the effect of observational selection. On average, loop-like XPEs are more massive than collimated ones (Tomczak & Ronowicz 2007); thus, they are easier to detect above the background. Indeed, the exclusion of faint events categorized as (C) increases the relative contribution of loop-like XPEs. It is interesting that in Kim et al. (2005b) loop-type XPEs (60) only slightly outnumber the sum of spray-type and jet-type events (51).

For other relations, the carefully selected events seem to be less affected by observational constraints than those of the entire populations. Therefore, we conclude that the former events adequately characterize intrinsic features of particular subclasses. For example, for loop-like XPEs the relation between eruptive and confined or between recurrent and single events is distinctly different from that for collimated XPEs. Loop-like XPEs are dominantly eruptive (82 to 14) and recurrent (59 to 37), whereas collimated XPEs are more frequently confined (19 to 11) and single (20 to 10).

We would like to stress that subclasses of XPEs defined by us resemble some types of classical prominences observed in Hα line (Tandberg-Hanssen 1995). Namely, a surge is a prominence that is collimated and confined, a spray is a prominence that is collimated and eruptive, a loop-like and confined event we call an activation of a prominence, and a loop-like and eruptive event is an eruptive prominence or "disparition brusque." Classifications of prominences do not distinguish the recurrence criterion; nevertheless, there are known observations in which aforementioned types of prominences were observed as single or recurrent events (B. Rompolt 2011, private communication). This similarity between XPEs and prominences suggests a close association between hot and cold components of active regions. This relation has not been investigated in detail so far, except for Ohyama & Shibata (2008).

3. XPE ASSOCIATION WITH SOLAR FLARE

Using the classification, we were able to separate several subclasses of XPEs looking more homogeneous than the full population. Unfortunately, we are not sure if particular subclasses of XPEs refer to events that are physically different. A quantitative analysis of soft X-ray images would give closer confirmation; however, for the majority of XPEs in the catalog this kind of analysis is practically unreliable, owing to minor signal and other observational limits. Therefore, the main motivation of this section is to justify the presence of physically different subclasses of XPEs by a comparison of properties of other solar-activity phenomena associated with particular XPEs. Basic characteristics of flares and CMEs are well known. In this section we present the association of XPEs with flares, and in Section 4 we present the association of XPEs with CMEs.

3.1. Soft X-Rays

It has been commonly agreed that an XPE is a consequence of a flare occurrence. As a matter of fact, there are five XPEs in our catalog, for which we could not find any associated flare. However, this sample is too small to justify the existence of flareless XPEs. Moreover, three of the five XPEs occurred close enough to the solar limb that they might have come from flares from the back side. Are they just the tip of the iceberg? The answer may depend on extensive and careful examination of SXT images made in the quiet mode.

3.1.1. Time Coincidence

For better insight in time coincidence between XPEs and flares, we mark the time of an XPE on the GOES light curve of an associated flare. In Figure 3, we present a histogram of time differences between start times of flares and XPEs for 330 pairs of events. In 311 cases out of 330 (94.2%), an increase of soft X-ray emission occurred earlier than the XPE. The time difference is very often several minutes only (see the maximum and median of the histogram); however, higher values also occur. Similar conclusions can be given regarding better-observed XPEs of qualities A and B (compare the gray bins in Figure 3). Similar histograms made for particular subclasses of the XPEs introduced in our classification do not show any important differences.

In Figure 4, we present a histogram of time differences between the end of XPEs and the peak of the associated flares as determined from the GOES light curves. In about 20% of investigated samples (41 out of 198) any XPEs were completed before the flare peak, i.e., within the rising phase of a flare. For almost 80% of events the final evolution of XPEs is seen after the maximum of soft X-ray emission, very often no longer than 10 minutes (108 examples out of 198, 54.5%). Similar conclusions can be given regarding better-observed XPEs of qualities A and B (compare the gray bins in Figure 4). Similar histograms made for particular subclasses of XPEs introduced in our classification do not show any important differences.

The above results should be normalized to the timescales of flares. Therefore, we have prepared counterparts of Figures 3 and 4 in which we normalize time differences with the flare rising-phase duration. In Figure 5 we illustrate occurrences of the XPE start: negative values mean that an XPE preceded its flare, the value 0—simultaneous start, the value 1—start of an XPE at the maximum of its flare, values grater than 1—later XPE start. As we see, the majority of XPEs (282 out of 330, 85.5%) start within the rising phase of flares. This rule is fulfilled even stronger for better-observed XPEs (gray bins)—176 out of 198, 88.8%.

Interesting results are revealed by further versions of Figure 5, in which particular subclasses of XPEs are separated: collimated and loop-like, confined and eruptive, single and recurrent, in Figures 6–8, respectively. In these figures, we show side by side the distributions of the XPEs that have contrasting properties. For example, in Figure 6 loop-like XPEs show a tendency to start earlier in the rising phase of flares than collimated ones: the difference for medians is more than 0.2 of the rising-phase duration. A similar tendency is seen in Figure 7, where eruptive XPEs start earlier in the rising phase of associated flares than confined ones, and in Figure 8, where recurrent XPEs precede, on average, single ones.

**Figure 6.** Histogram of time differences between the SXR start of the associated flare (from *GOES* light curve) and the XPE start normalized with the flare rising-phase duration made for collimated and loop-like XPEs separately. Black or gray and hatched or white bins represent the better-observed XPEs (qualities A and B, so-called selected XPEs) and the rest of XPEs (quality C), respectively. The size of bins is 0.1 of the flare rising-phase duration with an exception of outermost ones. Numbers of all the considered XPEs and the selected XPEs, as well as their medians, are given.
Download figure:
Standard image High-resolution image

**Figure 7.** Histogram of time differences between the SXR start of the associated flare (from *GOES* light curve) and the XPE start normalized with the flare rising-phase duration made for confined and eruptive XPEs separately. Black or gray and hatched or white bins represent the better-observed XPEs (qualities A and B, so-called selected XPEs) and the rest of XPEs (quality C), respectively. The size of bins is 0.1 of the flare rising-phase duration with an exception of outermost ones. Numbers of all the considered XPEs and the selected XPEs, as well as their medians, are given.
Download figure:
Standard image High-resolution image

**Figure 8.** Histogram of time differences between the SXR start of the associated flare (from *GOES* light curve) and the XPE start normalized with the flare rising-phase duration made for single and recurrent XPEs separately. Black or gray and hatched or white bins represent the better-observed XPEs (qualities A and B, so-called selected XPEs) and the rest of XPEs (quality C), respectively. The size of bins is 0.1 of the flare rising-phase duration with an exception of outermost ones. Numbers of all the considered XPEs and the selected XPEs, as well as their medians, are given.
Download figure:
Standard image High-resolution image

In Figure 9, we have normalized time differences between the XPE end and the associated flare start with the rising-phase duration of a flare. In this scale the value 1 means that the XPE end occurred exactly at the maximum of the associated flare. The histogram is rather gradual with two maxima between 1.2–1.4 and 1.6–1.8. The number of XPEs lower than the first maximum and greater than the second one decreases systematically with a marginal contribution of those that are lower than 0.4 and greater than 3. It means that all soft X-ray plasma motions are limited within a relatively narrow part of the total duration of associated flares.

The variants of Figure 9, in which particular subclasses of XPEs are separated, i.e., collimated and loop-like, confined and eruptive, single and recurrent, are presented in Figures 10–12, respectively. In these figures, as in Figures 6–8, we show side by side the distributions of the XPEs that have contrasting properties. In Figure 10 the collimated XPEs seem to last longer, on average, than the loop-like ones: medians of both distributions differ by 0.4 of the rising-phase duration. Similarly, the confined XPEs seem to last longer than the eruptive ones (Figure 11)—medians differ by about 0.6 of the rising-phase duration, in the case of better-observed XPEs. In Figure 12 the single XPEs last longer, on average, than the recurrent ones—medians differ by about 0.35 of the rising-phase duration, in the case of better-observed XPEs.

**Figure 10.** Histogram of time differences between the SXR start of the associated flare (from *GOES* light curve) and the XPE end normalized with the flare rising-phase duration made for collimated and loop-like XPEs separately. Black or gray and hatched or white bins represent the better-observed XPEs (qualities A and B, so-called selected XPEs) and the rest of XPEs (quality C), respectively. The size of bins is 0.2 of the flare rising-phase duration with an exception of outermost ones. Numbers of all the considered XPEs and the selected XPEs, as well as their medians, are given.
Download figure:
Standard image High-resolution image

**Figure 11.** Histogram of time differences between the SXR start of the associated flare (from *GOES* light curve) and the XPE end normalized with the flare rising-phase duration made for confined and eruptive XPEs separately. Black or gray and hatched or white bins represent the better-observed XPEs (qualities A and B, so-called selected XPEs) and the rest of XPEs (quality C), respectively. The size of bins is 0.2 of the flare rising-phase duration with an exception of outermost ones. Numbers of all the considered XPEs and the selected XPEs, as well as their medians, are given.
Download figure:
Standard image High-resolution image

**Figure 12.** Histogram of time differences between the SXR start of the associated flare (from *GOES* light curve) and the XPE end normalized with the flare rising-phase duration made for single and recurrent XPEs separately. Black or gray and hatched or white bins represent the better-observed XPEs (qualities A and B, so-called selected XPEs) and the rest of XPEs (quality C), respectively. The size of bins is 0.2 of the flare rising-phase duration with an exception of outermost ones. Numbers of all the considered XPEs and the selected XPEs, as well as their medians, are given.
Download figure:
Standard image High-resolution image

3.1.2. Flare Class and Total Duration

For each associated flare, we determined X-ray class and total duration based on light curves recorded by GOES, in the wavelength range of 1–8 Å. We defined the total duration as the interval between a constant level of the solar soft X-ray flux before and after a flare; therefore, our values of this parameter are larger than intervals between a start time and an end time that are routinely reported in the SGD. In some cases we could not estimate the total duration properly. This is the reason why the number of considered events in this paragraph is slightly lower than the number of XPEs associated with flares.

In Figure 13, we present a scatter plot of X-ray class versus total duration for flares associated with morphological subclasses of XPEs, i.e., collimated and loop-like XPEs. All points are marked with dots. Additionally, we emphasized well-observed XPEs (quality A or B) and flares that are non-occulted by the solar disk. These flares associated with well-observed collimated and loop-like XPEs are marked with boxes and stars, respectively. Both groups of flares are mixed in the plot; however, some shifts toward higher X-ray class and longer duration can be seen for flares associated with loop-like XPEs.

Similar scatter plots of X-ray class versus total duration for flares associated with kinematical and recurrence subclasses of XPEs are given in Figures 14 and 15. All points are marked with dots. Again, additionally we emphasized well-observed XPEs (quality A or B) and flares that are non-occulted by the solar disk. Moreover, we excluded events for which the assignment of kinematical subclasses for XPEs was uncertain (Figure 14) and events associated with XPEs that were classified as single in the case of partial time coverage of observations (Figure 15). In both figures, flares associated with well-observed XPEs classified as subclass 1 (confined and single, respectively) are marked with boxes, whereas flares associated with XPEs classified as subclass 2 (eruptive and recurrent, respectively) are marked with stars. Similarly to Figure 13, both groups of flares are mixed in the plots and some shifts toward higher X-ray class and longer duration are seen for flares associated with XPEs of subclasses 2.

**Figure 14.** Scatter plot of flare X-ray class vs. flare total duration. This plot compares flares associated with XPEs classified according to the kinematical criterion. All points are marked with dots. A subset of well-observed (see the text) flares is additionally marked with boxes and stars for confined and eruptive XPEs of high quality (A and B), respectively.
Download figure:
Standard image High-resolution image

**Figure 15.** Scatter plot of flare X-ray class vs. flare total duration. This plot compares flares associated with XPEs classified according to the recurrence criterion. All points are marked with dots. A subset of well-observed (see the text) flares is additionally marked with boxes and stars for single and recurrent XPEs of high quality (A and B), respectively.
Download figure:
Standard image High-resolution image

The shifts seen in Figures 13–15 are confirmed by medians calculated separately for both groups of flares for each classification criterion. As seen in Table 6 (bold-faced columns), medians for flares associated with XPEs of subclass 2 are 1.5–3.8 times and 2.1–2.7 times greater than medians for flares associated with XPEs of subclass 1 for flare X-ray class and flare duration, respectively. Higher X-ray class and longer duration mean a more energetic flare; thus, we can conclude that more energetic XPEs are, on average, associated with more energetic flares and less energetic XPEs rather prefer less energetic flares.

Table 6. Properties of Flares Associated with Particular Subclasses of XPEs

			Flare		Flare
XPE	Number		Class		Duration
Subclass	of HXR		(Median)		(Median)
	Events		(W m⁻²)		(minutes)
	Tot.	SS	Tot.	SS	Tot.	SS
Morphological criterion
1 (collimated)	67/114	24/42	C8.7	C6.1	70	51
2 (loop-like)	168/245	102/136	M1.6	M1.8	120	110
2 to 1 ratio			1.8	3.0	1.7	2.2
Kinematical criterion
1 (confined)	97/171	32/56	C9.3	C6.1	75	45
2 (eruptive)	138/188	80/98	M1.8	M2.3	125	120
2 to 1 ratio			1.9	3.8	1.7	2.7
Recurrence criterion
1 (single)	165/264	43/57	M1.1	M1.4	90	75
2 (recurrent)	70/95	53/69	M1.9	M2.1	135	155
2 to 1 ratio			1.7	1.5	1.5	2.1
Extreme differences
(1,1,1)	42/74	8/14	C7.9	C5.2	45	42
(2,2,2)	50/64	38/44	M2.1	M3.1	160	155
(2,2,2) to			2.7	6.0	3.6	3.7
(1,1,1) ratio

Download table as: ASCII Typeset image

One can expect that the difference between characteristics describing associated flares should be even higher for two subclasses of XPEs defined by combining our three criteria simultaneously. Indeed, medians in Table 6 for flares associated with subclass (1,1,1)—collimated, confined, single XPEs—and subclass (2,2,2)—loop-like, eruptive, recurrent XPEs—show extreme differences (a factor of 6.0 and 3.7 for X-ray class and duration, respectively). As seen in Figure 16, in the diagram X-ray class versus duration, flares associated with subclasses (1,1,1) and (2,2,2) of well-observed XPEs are almost separated.

**Figure 16.** Scatter plot of flare X-ray class vs. flare total duration. This plot compares flares associated with XPEs classified according to different criteria that are employed simultaneously. All points are marked with dots. A subset of well-observed (see the text) flares is additionally marked with boxes and stars for collimated, confined, single (1,1,1) and loop-like, eruptive, recurrent (2,2,2) XPEs of high quality (A and B), respectively.
Download figure:
Standard image High-resolution image

In unbold-faced columns in Table 6 we present medians for flares associated with different subclasses of XPEs that were defined less strictly, i.e., by including quality C events and without excluding any doubtful examples. Ratios of medians for subclasses 2 and subclasses 1 that were constituted more liberally are usually lower in comparison with the more strictly defined bold-faced values. It shows how some physical differences can be masked by observational constraints.

3.2. Hard X-Rays

We included in the catalog hard X-ray light curves of associated flares, recorded by Yohkoh HXT, for investigating the relation between XPEs and non-thermal electron signatures. We considered light curves in the energy band M1 (23–33 keV) and interpreted a signal above the doubled value of the background as proof that in a particular flare an acceleration of an appropriate number of non-thermal electrons occurred. For 353 flares that were associated with XPEs we found that 235 events, i.e., 66.6%, showed this signature.

In the second and third columns of Table 6, we present detailed results of this relation for both populations of events and particular subclasses of XPEs. A percentage of associated flares showing non-thermal electrons depends on how energetic is the subclass, with higher values (75%–82%) for subclasses 2 and lower values (57%–75%) for subclasses 1. The difference is the highest for the kinematical criterion, but for the recurrence criterion percentages for both subclasses are almost the same. After applying all three criteria, we found that the difference between the least energetic subclass (1,1,1)—collimated, confined, single—and the most energetic (2,2,2)—loop-like, eruptive, recurrent—is maximal: 57% and 86% of associated flares indicating non-thermal electrons, respectively.

We also investigated time coincidence between macroscopic X-ray plasma motions (XPEs) and non-thermal electron signatures (HXRs) in detail. In this aim, we measured time differences between the XPE start and the HXT/M1 flare peak. The results are presented as a histogram in Figure 17. In 185 of 227 cases (81.5%) XPEs started before the HXR flare peak; in 18.5% of cases the chronology was opposite. However, the most frequent bin, 0–2 minutes in about 45% of cases, suggests that both considered processes, i.e., soft X-ray plasma motion and non-thermal electron acceleration, are strongly coupled.

Similar investigation was performed by Kim et al. (2005b). At first glance our Figure 17 and their Figure 5 are different. However, it is important to know that the values in our histogram have opposite sign and we used the HXR peak time for higher energy band M1, 23–33 keV, than Kim et al., who used energy band L (14–23 keV). In flares with a strong contribution of the non-thermal component, the peak time in those energy bands is close, but in flares with a stronger contribution of the superhot component in the L band, the M1 peaks tend to occur earlier than the L peaks. Keeping in mind the above-mentioned differences in data organization, we can conclude that our results are consistent.

We also prepared variants of Figure 17 for particular subclasses of XPEs. However, we did not find any evident differences between the considered distributions, namely, each of them shares the common peak bin.

4. XPE ASSOCIATION WITH CORONAL MASS EJECTIONS

In order to associate out XPEs with CMEs, we used the SOHO LASCO CME Catalog (Gopalswamy et al. 2009). Only 275 XPEs occurred when the LASCO coronagraphs were operational. We found that 182 XPEs (66.2%) were associated with CMEs. This is slightly less than the 69% (95 of 137 events) obtained by Kim et al. (2005b). For particular subclasses of XPEs, the association was between 44% and 88% (see Table 7). According to Yashiro et al. (2008), we consider an XPE–CME pair as physically connected if the XPE occurred within the position angle range defined by the CME angular width increased by 10° from either side. Moreover, the time of the XPE had to fall within a three hour interval centered around the extrapolated time of the CME front start at h = 1 R_☉. For the extrapolation we used the time of the first appearance in the LASCO/C2 field of view and the linear velocity taken from the CME catalog.

4.1. Time Coincidence

The histogram of the time differences between the extrapolated CME front onset and the XPE start is presented in Figure 18. As we can see, there are more events with negative values, i.e., those in which the CME starts before the XPE, than those with the opposite chronology. The frequencies are 73.7% (87 of 118) and 26.3% (31 of 118), respectively. The carefully selected subgroup (well-observed XPEs of quality A or B that occurred close to the solar limb, |λ| > 60°) shows slightly different proportions: 66.1% (41 of 62) and 33.9% (21 of 62), respectively. Both distributions, all the XPEs and the selected XPEs, are quite gradual with slightly different medians: −17.3 minutes and −10.6 minutes, respectively.

Kim et al. (2005b) performed similar analysis for XPEs from the two-year interval 1999–2001. Their Figure 6 containing 43 events was made under slightly different assumptions: (1) the CME front times were extrapolated at individual locations of XPEs in the Yohkoh field of view, and (2) the CME speed was determined from the first two observing times and heights. Despite these differences, our histogram looks quite similar to those of Kim et al. Therefore, we conclude that, at least in a statistical sense, different ways of extrapolating the CME onset time do not seriously affect the temporal relation between XPEs and CMEs.

As in the analysis in Section 3.1.1, we give further versions of Figure 18, in which particular subclasses of XPEs, collimated and loop-like, confined and eruptive, and single and recurrent, are separated in Figures 19, 20, and 21, respectively. In Figure 19, the histogram for loop-like XPEs shows a relatively narrow maximum located close to the zero point. It means that a large fraction of XPEs (∼50%) start almost simultaneously with the CME onset. Collimated XPEs are shifted toward negative values in this figure, and their maximum is distinctly broader. The histogram made for the selected subgroup of better-observed events (gray and black bins in Figure 19) shows a similar trend.

**Figure 19.** Histogram of time differences between the extrapolated CME front onset and the XPE start made for collimated and loop-like XPEs separately. Black or gray and hatched or white bins represent the selected subgroup of better-observed XPEs (qualities A and B, |λ| > 60°) and the rest of XPEs, respectively. The size of bins is 10 minutes with an exception of the outermost right one. Numbers of all the considered XPEs and the selected XPEs, as well as their medians, are given.
Download figure:
Standard image High-resolution image

A similar pattern can be seen in Figure 20. In this figure, the histogram made for eruptive XPEs is narrower and centered closer to the zero point than the histogram made for confined XPEs. The distribution for confined XPEs is much broader, especially in the plot for the selected subgroup of better-observed events (black bins), which makes an impression of a random occurrence within almost the whole time window of the CME onset.

In Figure 21 both histograms made for single and recurrent XPEs show a similar width, but recurrent XPEs tend to start earlier (almost simultaneously with the CME) than single ones. The difference between medians is about 15 minutes for all events, as well as for the selected subgroup of better-observed events.

4.2. CME Angular Width and Velocity

In Figure 22, we present a scatter plot of an angular width versus a linear velocity for CMEs associated with morphological subclasses of XPEs, i.e., collimated and loop-like XPEs. All points are marked with dots. Additionally, we emphasized well-observed XPEs (quality A or B) that occurred close to the solar limb (|λ| > 60°). CMEs associated with well-observed collimated and loop-like XPEs are marked with boxes and stars, respectively. Both groups of CMEs are mixed in the plot; however, some shifts toward wider and faster events are seen for CMEs associated with loop-like XPEs.

Similar scatter plots of an angular width versus a linear velocity for CMEs associated with kinematical and recurrence subclasses of XPEs are given in Figures 23 and 24. All points are marked with dots. Again, additionally we emphasized well-observed XPEs (quality A or B) that occurred close to the solar limb (|λ| > 60°). Moreover, we excluded events for which the assignment of kinematical subclasses for XPEs was uncertain (Figure 23) and events associated with XPEs that were classified as single in the case of partial time coverage of observations (Figure 24). In both figures, CMEs associated with well-observed XPEs classified as subclass 1 (confined and single, respectively) are marked with boxes, whereas CMEs associated with well-observed XPEs classified as subclass 2 (eruptive and recurrent, respectively) are marked with stars. Similarly to Figure 22, both groups of CMEs are mixed in the plots and some shifts toward wider and faster events are seen for CMEs associated with XPEs of subclasses 2.

**Figure 23.** Scatter plot of CME angular width vs. CME linear velocity. This plot compares CMEs associated with XPEs classified according to the kinematical criterion. All points are marked with dots. A subset of well-observed (see the text) CMEs is additionally marked with boxes and stars for confined and eruptive XPEs of high quality (A and B), respectively. Values are taken from the *SOHO* LASCO CME Catalog (Gopalswamy et al. 2009).
Download figure:
Standard image High-resolution image

**Figure 24.** Scatter plot of CME angular width vs. CME linear velocity. This plot compares CMEs associated with XPEs classified according to the recurrence criterion. All points are marked with dots. A subset of well-observed (see the text) CMEs is additionally marked with boxes and stars for single and recurrent XPEs of high quality (A and B), respectively. Values are taken from the *SOHO* LASCO CME Catalog (Gopalswamy et al. 2009).
Download figure:
Standard image High-resolution image

The shifts seen in Figures 22–24 are confirmed by medians calculated separately for both groups of CMEs for each classification criterion. As seen in Table 7 (bold-faced columns for well-observed events), the medians for CMEs associated with XPEs of subclass 2 are 1.2–2.0 times and 1.2–1.4 times greater than those for CMEs associated with XPEs of subclass 1 for CME angular width and linear velocity, respectively. A higher angular width and velocity mean a more energetic CME; thus, we can conclude that more energetic XPEs are, on average, associated with more energetic CMEs and less energetic XPEs rather prefer less energetic CMEs.

Table 7. Properties of CMEs Associated with Particular Subclasses of XPEs

			CME		CME
XPE			Angular		Velocity
Subclass	Number		Width		(Median)
	of Events		(Median)		(km s⁻¹)
	Tot.	SS	Tot.	SS	Tot.	SS
Morphological criterion
1 (collimated)	54/90	13/26	89°	61°	522	444
2 (loop-like)	128/185	49/67	113°	125°	649	547
2 to 1 ratio			1.3	2.0	1.2	1.2
Kinematical criterion
1 (confined)	75/142	17/39	90°	83°	522	450
2 (eruptive)	107/133	38/43	125°	126°	629	642
2 to 1 ratio			1.4	1.5	1.2	1.4
Recurrence criterion
1 (single)	125/200	18/27	93°	104°	528	518
2 (recurrent)	57/75	24/30	155°	126°	662	602
2 to 1 ratio			1.7	1.2	1.3	1.2
Extreme differences
(1,1,1)	33/63	7/12	83°	61°	500	357
(2,2,2)	42/52	20/23	168°	126°	718	613
(2,2,2) to (1,1,1)			2.0	2.1	1.4	1.7
ratio

Download table as: ASCII Typeset image

One can expect that the difference between characteristics describing associated CMEs should be even higher for two subclasses of XPEs that we define by applying our three criteria simultaneously. Indeed, medians in Table 7 for CMEs associated with subclass (1,1,1)—collimated, confined, single XPEs—and subclass (2,2,2)—loop-like, eruptive, recurrent XPEs—show extreme differences (factors 2.1 and 1.7 for angular width and linear velocity, respectively). The difference between characteristics describing associated CMEs is also seen in Figure 25.

**Figure 25.** Scatter plot of CME angular width vs. CME linear velocity. This plot compares CMEs associated with XPEs classified according to different criteria that are employed simultaneously. All points are marked with dots. A subset of well-observed (see the text) CMEs is additionally marked with boxes and stars for collimated, confined, single (1,1,1) and loop-like, eruptive, recurrent (2,2,2) XPEs of high quality (A and B), respectively. Values are taken from the *SOHO* LASCO CME Catalog (Gopalswamy et al. 2009).
Download figure:
Standard image High-resolution image

In unbold-faced columns in Table 7 we present medians for CMEs associated with different subclasses of XPEs that were defined less strictly, i.e., by including quality C, for |λ| ⩽ 60° and without excluding any doubtful examples. Ratios of medians for subclasses 2 to medians for subclasses 1 that were constituted more liberally are often comparable to the more strictly defined bold-faced values. It is opposite to flares for which bigger differences between unbold-faced and bold-faced values are evident (see Table 6). We suggest that the main reason for this is the condition |λ| > 60° that we constituted for specially selected (bold-faced) events. It excludes the majority of halo CMEs being systematically wider and faster than ordinary CMEs (Michałek et al. 2003). In other words, more strict selection criteria undoubtedly limit a scatter of values in two physically different groups; however, it is compensated with the bias introduced by halo CMEs.

5. DISCUSSION

The XPEs collected in the catalog confirm the strong association with flares. Starts of XPEs observed since their very beginning fall usually within the rising phase of associated flares (Figures 3 and 5) and well coincide with the HXR peaks (Figure 17). It means that symptoms of SXR plasma motions occur when magnetic energy conversion in flares—via reconnection—is most vigorous (Benz 2008). A small number of exceptions are connected mainly with complex events in which X-ray enhancements, recorded by GOES and Yohkoh/HXT, are accumulated from at least two different positions on the Sun.

We would like to stress a lower correlation between XPEs and HXR flares than between XPEs and SXR ones. As approximately one-third of the flares associated with XPEs did not show any clear signatures of non-thermal electrons, we can conclude that some macroscopic motion of SXR plasma is a more obvious characteristic of reconnection than acceleration of non-thermal electrons. It can be caused by the limited sensitivity of the HXT. As we can see in Table 6, the more energetic subclasses 2 in our classification scheme of XPEs show a stronger correlation with HXR flares. Thus, under the assumption that the more energetic XPEs are associated with more energetic flares, we can expect a larger fraction of them to be able to produce the HXR emission above the threshold of the HXT.

Another proof of close association between processes responsible for XPEs and flares is similarity between their durations. A comparison between the medians in Figures 5 and 9 shows that an XPE lasts, on average, as long as the rising phase of an associated flare. Histograms presented for particular subclasses of XPEs (Figures 6–8 and 10–12) show that the more energetic subclasses 2 occurs earlier and lasts shorter than the less energetic subclasses 1. This difference probably reflects some differences in reconnection processes occurring in both subclasses. At first sight, this result is in contradiction with Figures 13–15, in which XPEs from subclasses 2 seem to prefer flares of longer duration; however, we should remember that in Figures 6–8 and 10–12 the time is normalized with the flare rising-phase duration.

Some interesting hints concerning hierarchy and chronology of processes occurring in restructuring active regions can be found in histograms of time differences between the XPE start and the extrapolated CME onset for particular subclasses of XPEs (Figures 19–21). The more energetic subclasses 2 of XPEs shows close relationship with its associated CMEs. They seem to start almost simultaneously, and small deviations from the zero value are probably caused by unrealistic extrapolation of the CME onsets. The start of XPEs of the less energetic subclasses 1 shows a much looser connection with the CME start. Very often the CME seems to occur first. Keeping in mind that XPEs are usually caused by magnetic reconnection, we suggest that in the case of subclasses 2, the reconnection and loss of equilibrium of magnetic structure, thus a CME development, occur almost simultaneously. On the other hand, in the case of subclasses 1, the reconnection is usually a consequence of destabilization of magnetic structure, which may occur earlier.

The results summarized in Tables 6 and 7 strongly suggest that total amount of energy, converted from the magnetic field in an active region during its magnetic reconfiguration, determines characteristics of events including flares, CMEs, and XPEs, which are thought to be consequences of this common reconfiguration. Thus, more energetic XPEs are associated with more energetic flares and CMEs, while less energetic ones seem to occur commonly. This statistically averaged picture does not exclude, for sure, exceptions in partitioning of magnetic energy. For example, there are X-class confined flares completely devoid of any CME (Wang & Zhang 2007; Cheng et al. 2011). These flares are probably also devoid of XPEs, but this research is beyond the scope of this work.

Our investigation shows that characteristics of flares and CMEs associated with particular subclasses of XPEs are different. We found that a scale of differences is higher for flares than for CMEs. We also found that the recurrence criterion proposed in our XPE classification scheme does not separate the associated events as strongly as the morphological and kinematical criteria.

6. CONCLUSIONS

In our catalog we have collected the most extensive database of XPEs so far. Images from the SXT on board Yohkoh have been organized into movies in the MPEG format. The events have been classified on the basis of elementary and uniform criteria. The catalog also gives a piece of information concerning the associated flares and CMEs by using entries to the Yohkoh Flare Catalog (HXT/SXT/SXS/HXS) and the SOHO LASCO CME Catalog, respectively. The collected data allow us to study XPEs more comprehensively as separate solar-activity phenomena and also as elements of more complex processes occurring in the solar corona.

XPEs constitute a strongly inhomogeneous group of events. Their appearances include expanding loop structures, moving blobs, rising columns, and so on. Their strong inhomogeneity is supported by a wide range of values of basic parameters: altitude (10⁸–10¹⁰ cm), volume (10²⁶–10³⁰ cm⁻³), duration (10¹–10³ s), velocity (10⁰–10³ km s⁻¹), acceleration (−10³–10⁴ m s⁻²), mass (10¹²–10¹⁵ g), and energy (10²⁵–10³¹ erg).

It is difficult to point out a universal mechanism responsible for all the events presented in the catalog. There is no doubt that the majority of XPEs are connected somehow with magnetic reconnection. However, in many events, the evolution is far from what may be expected from the canonical CSHKP model, suggesting the existence of more complex three-dimensional quadrupolar reconnection (Nitta et al. 2010). We often observe an XPE as a result of magnetic reconnection that leads to chromospheric evaporation as a hydrodynamic response of intensified plasma heating or non-thermal electron beams in a flare magnetic structure.

On the other hand, the close morphological and kinematical connection of some XPEs with CMEs, together with the similar start time, suggests a mechanism of loss-of-equilibrium type common for CMEs. (Indeed, movies illustrating evolution of some XPEs resemble cartoons presenting the tether release model or the tether straining model leading to the magnetic breakout model.) Finally, some movies in the catalog give an impression that SXR plasma leaks out from the magnetic structure probably under low-β-plasma conditions.

For proper interpretation of the data, we need to identify the mechanism responsible for the observed XPE. An inappropriate choice of the mechanism can lead to meaningless and erroneous conclusions regarding processes occurring in the solar corona. In the context of strong inhomogeneity of XPEs and several possible mechanisms of their origin, it is not advised to routinely interpret all XPEs in terms of a single and same mechanism. A similar conclusion was given by Nitta et al. (2010), who criticized the tendency to employ the CSHKP model for the description of all "Masuda-type" flares (Masuda et al. 1994).

If we consider one particular XPE, it is basically difficult to decide which mechanism is responsible for its occurrence without the complete quantitative analysis including plasma diagnostics and the modeling of magnetic field structure. These conditions were unreachable in practice for the majority of events in the catalog. Therefore, in advent of new observations of XPEs derived by modern instruments on board Hinode, Solar–Terrestrial Relations Observatory (STEREO), and the Solar Dynamics Observatory (SDO), we have been trying to give some solutions that would be correct at least in a statistical sense.

We have shown that the subclasses of XPEs separated on the basis of our simple observational criteria have different levels of correlation with other solar-activity phenomena. The difference is also seen if we consider basic parameters describing these flares and CMEs. However, the association of XPEs with different flares or CMEs does not mean a specific physical mechanism as far as these flares or CMEs represent physically different groups. In the meantime, discussions concerning a difference in observational characteristics to justify separate physical mechanisms responsible for flares or CMEs are still open. Are there two different classes of flares (Pallavicini et al. 1977), or can all flares be explained by only one mechanism (Shibata et al. 1995)? Are there two kinematically different classes of CMEs (Sheeley et al. 1999), or is the division artificial (Vršnak et al. 2005)?

We have found that more energetic XPEs are better correlated with flares and CMEs and that more energetic XPEs correlate with more energetic flares and CMEs. Virtually the effect of observational conditions works in the same way, and we cannot resolve correctly the influence of the effect on our conclusion.

The most promising way in the investigation of XPEs is to deal with them as an element of a larger ensemble. The use of observations made in temporal, spatial, and spectral ranges broader than those needed for direct monitoring of XPEs allows better understanding of processes in which XPEs participate. Recently, a similar picture of flares as global events was presented by Hudson (2011). Our experience is that XPEs are strongly coupled with flare HXR quasi-periodic oscillations (Nakariakov & Melnikov 2009), probably because the reconnection rate is controlled by plasmoid generation (Nishida et al. 2009). We also found that XPEs are somehow associated with progressive spectral hardening in HXRs (Tomczak 2008) and thus with solar energetic particles (Kiplinger 1995; Grayson et al. 2009).

In the future, we are going to upgrade the XPE catalog by adding entries devoted to associated prominences and radio bursts. Moreover, the TRACE movies will be added, if available. We are also going to perform the comprehensive analysis of several very interesting events from the catalog that have been omitted by other Yohkoh researchers.

Yohkoh is a project of the Institute for Space and Astronautical Sciences, Japan, with substantial participation from other institutions within Japan and with important contributions from the research groups in the US and the UK under the support of NASA and SERC. The SXT instrument was jointly developed by the Lockheed Palo Alto Research Laboratory and the National Astronomical Observatory of Japan. Collaborators include the University of Tokyo, Stanford University, the University of California at Berkeley, and the University of Hawaii. The CME catalog used in this work is generated and maintained at the CDAW Data Center by NASA and The Catholic University of America in cooperation with the Naval Research Laboratory. SOHO is a project of international cooperation between ESA and NASA. The Yohkoh Flare Catalog (HXT/SXT/SXS/HXS) used in this work was developed by J. Sato, K. Yoshimura, T. Watanabe, M. Sawa, M. Yoshimori, Y. Matsumoto, S. Masuda, and T. Kosugi. We acknowledge solar data collected and distributed by U.S. National Geophysical Data Center. We thank Dr. Masamitsu Ohyama for providing us his list of XPEs from years 1991 to 1998. We appreciate the valuable remarks of the referee, which helped us to improve this paper. This work was supported by Polish Ministry of Science and High Education grant N N203 1937 33.

A CATALOG OF SOLAR X-RAY PLASMA EJECTIONS OBSERVED BY THE SOFT X-RAY TELESCOPE ON BOARD YOHKOH

Article metrics

Permissions

Author e-mails

Author affiliations

Dates

ABSTRACT

1. INTRODUCTION