Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.izmiran.rssi.ru/~obridko/papers/53381.pdf Äàòà èçìåíåíèÿ: Mon Mar 1 17:58:46 2004 Äàòà èíäåêñèðîâàíèÿ: Sun Apr 10 01:04:20 2016 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: molecular cloud

THE ASTROPHYSICAL JOURNAL, 567 : 1225 õ 1233, 2002 March 10
( 2002. The American Astronomical Society. All rights reserved. Printed in U.S.A.

SOLAR DISAPPEARING FILAMENT INSIDE A CORONAL HOLE I. M. CHERTOK, E. I. MOGILEVSKY, V. N. OBRIDKO, AND N. S. SHILOVA
IZMIRAN, Troitsk, Moscow Region, 142190, Russia ; ichertok=izmiran.troitsk.ru

AND H. S. HUDSON
Solar Physics Research Corporation, ISAS, Sagamihara-shi, Kanagawa 229, Japan Received 2001 January 25 ; accepted 2001 November 9

ABSTRACT Based on Y ohkoh/SXT, SOHO/EIT images and movies, as well as on Ha, He I 10830 A heliograms and other relevant data, we analyze an event of 1999 December 28, which is interesting in at least two aspects. (1) A major horseshoe-shaped Ha ïlament appeared to be located within a large transequatorial coronal hole (CH) in the eastern hemisphere. (2) This ïlament subsequently disappeared, with its eruption combined with a number of dynamic phenomena, including large-scale ones. The probable location of the ïlament inside the CH was conïrmed in detail by calculations of the open ïeld regions and, for the ïrst time, the quasi-separatrix layers in the global solar magnetic ïeld. The ïlament eruption was accompanied by signiïcant evolution of the soft X-ray and EUV-emitting structures inside the CH as well as by a coronal mass ejection. The analysis indicates that CHs need not have the simple and uniform structure normally assumed and can sometimes contain local areas with low-altitude closed magnetic ïelds. It demonstrates also that the erupting ïlament inside the CH was a part of a much more global, evolving magnetic structure associated with activity extending through at least the entire eastern half of the disk. Subject headings : Sun : corona õ Sun : ïlaments õ Sun : UV radiation
1

. INTRODUCTION

It is usually believed that coronal holes (CHs) represent areas of relatively cool, low-density (and hence X-ray faint) plasma with open, predominantly unipolar magnetic ïelds and without signiïcant internal activity (e.g., Zirker 1977 ; Harvey 1995 ; Mogilevsky, Obridko, & Shilova 1997 ; Obridko & Shelting 1999 ; Zhao, Hoeksema, & Scherrer 1999). In soft X-rays and high-temperature EUV lines, they are visible as extended structures with diminished emission. They are bright in the He I 10830 A chromospheric absorp tion line, modiïed by overlying coronal radiation with wavelengths smaller than 500 A and have reduced internal network intensity and contrast. Cold ïlaments and hot active regions both appear dark. The most common type of the small-scale activity observed inside CHs appears to be soft X-ray bright points and the corresponding dark points visible in the He I 10830 A line (e.g., Nolte et al. 1978 ; Kahler & Moses 1990 ; Harvey 1995). Among the medium-scale features inside CHs, emerging magnetic ÿuxes (Stepanyan & Malanushenko 2001) and compact active regions capable of producing large soft X-ray coronal jets (e.g., Shibata et al. 1994) sometimes appear. In this paper, we present the event of 1999 December 28õ29, which is of interest due to two essential peculiarities (Fig. 1). In this case, a reasonably large ïlament appeared to be located within an extended transequatorial CH. Moreover, a disappearance/eruption of this ïlament occurred. The disappearance was accompanied by a number of largescale dynamic phenomena such as a coronal mass ejection (CME), EUV-emitting structures inside the CH, a soft X-ray arcade, and some other eects. These features seem to mean that CH structures need not be so simple and quiet as previously assumed. In ° 2, the data and approach used in the analysis are outlined. Diverse imaging and magnetic ïeld data on the 1225

ïlament position relative to the CH are presented in ° 3. As a whole, these data indicate the probable location of the ïlament inside a CH. The activity preceding the ïlament eruption, as well as the eruption itself and the accompanying phenomena are described and analyzed in ° 4. The discussion and conclusion are given in ° 5.
2.

DATA AND APPROACH

The majority of the full-disk images and movies used in this analysis are accessible through the Internet. The Ha heliograms of the Big Bear Solar Observatory (BBSO), Mauna Loa Solar Observatory/High Altitude Observatory (MLSO/HAO), and National Astronomical Observatory (Mitaka) are used to search dynamics of the ïlament. The He I 10830 A spectroheliograms of the MLSO/HAO and National Solar Observatory (Kitt Peak) allowed us to have information on the visibility, location, and conïguration of the CH in this line as well as on the position and evolution of the ïlament. The imaging data from the Y ohkoh/SXT (Tsuneta et al. 1991) are known to be valuable material both for locating CHs and observing posteruption structures ; for example, arcades. Observations from the EUV Imaging Telescope (EIT ; ` Delaboudiniere et al. 1995) aboard the Solar and Heliospheric Observatory (SOHO) were also widely used. In particular, the images in the high-temperature (2.0 MK) Fe XV 284 A line are almost as informative about CHs as the soft X-ray heliograms. Two other coronal lines, Fe IX, X 171 A and Fe XII 195 A, sensitive to the coronal plasma with the temperature of 1.1 and 1.5 MK, are the most suitable for detection of the emitting coronal features particularly associated with the ïlament activity. To make these relatively faint features visible and pronounced, the modiïed EUV images with a limited range of intensities were formed. On some of them, the brighter sources associated with active regions prove to be saturated. The white-light infor-


1226

CHERTOK ET AL.

Vol. 567

FIG. 1.õT op row : BBSO (a), MLSO (b), and Mitaka (c) Ha images of 1999 December 28 illustrating presence, evolution, and disappearance of the horseshoe-shaped ïlament 1-2-3. Blob D on the frame (c) is an artifact. Bottom row : Negative modiïed SOHO/EIT 284 A (d ) and Y ohkoh/SXT ( f ) images with the large CH on the eastern half of the disk and inside emitting element 4-1 as well as the positive Kitt Peak He I 10830 A spectroheliogram (e), in which both the ïlament 1-2-3 and emitting element 4-1 are visible.

mation on CME occurrence comes from the Large Spectroscopic Coronagraph (LASCO ; Brueckner 1995) on the SOHO spacecraft. The most complete observations covering the event under consideration were made by SOHO/EIT

Angle et al. whole at the

195 A wavelength with a 12-minute cadence. These data indicate that the main ïlament eruption occurred on December 28 around 22õ23 UT. Unfortunately, the available Ha data have a gap just at this time between 21 : 51 and 23 : 31 UT. The last MLSO/HAO spectroheliogram in the

FIG. 2.õ(a) Calculated synoptic magnetic maps of the radial ïeld at the heliocentric distances of R \ 1.0, 1.1, and 1.4 R centered at 1999 December 28. _ Thick line marks the neutral lines of the large-scale magnetic ïelds. Small circles are bases of open ïeld lines at each level. (b) Projection of the computed synoptic map at 1.1 R with the open ïeld area on the solar sphere superimposed on the BBSO Ha heliogram (Fig. 1a) demonstrating location of the ïlament _ 1-2-3 inside the CH, identiïed with the open ïeld line area. (c) Distribution of the footpoint separation factor N on the synoptic chart of the eastern half of the disk at 1.1 R , displaying the QLSs particularly along the CH boundaries. _


No. 2, 2002

DISAPPEARING FILAMENT INSIDE A CORONAL HOLE

1227

FIG. 3.õModiïed SOHO/EIT 195 A images of 1999 December 28 at the pre-eruption stage illustrating development of the inside-CH curved emitting element 4-1, as well as some other emitting features, passing through the ïlament region and outlining the CH boundaries visible particularly in the 284 A line and soft X rays (Figs. 1d and 1 f ).

He I 10830 A line was from 22 : 08 UT, i.e., before the main eruption. Nevertheless, the available multiwavelength data are sufficient to study the peculiar features of the event. It should be kept in mind that in this paper all Y ohkoh/SXT and SOHO/EIT images are presented in the inverted (negative) form, i.e., emitting features appear dark and absorption structures appear bright. At the same time, the Ha, He I 10830 A, and SOHO/LASCO images are shown in the normal (positive) form. The Stanford digital synoptic magnetic ïeld map for 1999 December 28 and the corresponding IZMIRAN techniques were used to calculate the open ïeld areas and, most interestingly, the quasi-separatrix layers in the global magnetic ïeld (Obridko, Kharshiladze, & Ivanov 2002). These calculations help us to outline the location, conïguration, and boundaries of the CH. The short deïnition and explanation

of the quasi-separatrix layers with some relevant references are given in ° 3.1
3.

ON LOCATION OF THE FILAMENT RELATIVE TO THE CORONAL HOLE

In Figures 1a and 1b, particularly from the Ha heliograms of 1999 December 28, one can see that the horseshoeshaped ïlament 1-2-3 of our interest was reasonably long (about 250 Mm) and located in the southeast quadrant. On this day, the clear, large, transequatorial CH, best visible on the SOHO/EIT 284 A (Fig. 1d) and Y ohkoh/SXT (Fig. 1f )
1 Digital versions of the relevant illustrations, including not only ïgures but also movies in Javascript and MPEG formats, can be found at the following web site : http ://helios.izmiran.troitsk.ru/lars/Chertok/991228/ title.html.


1228

CHERTOK ET AL.

Vol. 567

FIG. 4.õModiïed SOHO/EIT 171 A images illustrating development of the inside-CH bright element 4-1 and emitting features before ïlament eruption (a, b) and at the post-eruption stage (c, d ).

images, is present in the central sector of the eastern half of the disk. On the Kitt Peak He I 10830 A spectroheliogram (Fig. 1e), as well as on the similar MLSO/HAO spectroheliograms (see Fig. 5), the CH and its boundaries are not sharply deïned. Nevertheless, it seems that the ïlament 1-23 appears to be located near the southern boundary of the He I 10830 A CH. It is important to note that the curved emitting element 1-4, visible as dark on the negative SOHO/ EIT 284 A (Fig. 1d), Y ohkoh/SXT (Fig. 1 f ) images, and on the positive He I image (Fig. 1e), is located southeastward relative to the ïlament. Taking this point into account, one can conclude from the SOHO/EIT 284 A (Fig. 1d) image and especially from the Y ohkoh/SXT (Fig. 1f ) images (see also Fig. 6) that the ïlament 1-2-3 appears to be positioned inside the CH visible in the soft X-rays and hightemperature 284 A line. Unfortunately, at present there is no objective or quantitative method for determining the CH boundaries on soft X-ray, EUV, and He I 10830 A heliograms (e.g., Zhao et al.

1999). The Kitt Peak CH maps routinely published in the Solar-Geophysical Data (SGD) are constructed from the full-disk He I 10830 A spectroheliograms and magneto grams, based on qualitative criteria (e.g., Recely & Harvey 1986) ; these criteria exclude a priori any nonunipolar featuresõparticularly ïlamentsõinside CHs. For our case of 1999 December 28 (SGC 2000), the Kitt Peak map classiïes as a CH only the "" bright îî northeastern part of the CH visible on the soft X-ray and 284 A images as well as two small patches located near the central meridian, in the southeast sector of this CH. The area adjoining the ïlament 1-2-3 is not included in the CH on the Kitt Peak map. Therefore, one can note that in our case, as well as in some other cases, the He I 10830 A spectroheliograms display mainly the most intense and pronounced areas of the real CH (see below). Let us consider now the information on the CH which can be inferred from some magnetic ïeld data. From the physical point of view, perhaps the most reliable and valid


No. 2, 2002

DISAPPEARING FILAMENT INSIDE A CORONAL HOLE

1229

FIG. 5.õFragments of the MLSO/HAO Ha (aõc) and He I 10830 A (dõf ) images displaying partial disappearance of the southern branch of the ïlament 1-2-3 before the main eruption of the whole ïlament. Adjacent emitting element 4-1 is also visible in the He I line.

approach to determination of a CH conïguration is its identiïcation with the computed footpoint area of open ïeld lines (Obridko & Shelting 1999 ; Zhao et al. 1999). Shown in Figure 2a are the synoptic magnetic ïeld maps of the radial ïeld at the heliocentric distances of R \ 1.0, 1.1, and 1.4 R calculated using the Stanford synoptic chart and _ IZMIRAN modiïcation of the potential ïeld-source surface model (Obridko & Shelting 1999 ; Obridko et al. 2002). All maps are centered at 1999 December 28. The thick lines mark the neutral lines of the large-scale magnetic ïelds. The small circles are the bases of the open ïeld lines at each level. One can see that the location and conïguration of the eastern open ïeld area corresponds well to the CH visible on the soft X-ray and 284 A images. (Also, the northwest narrow open ïeld region matches another CH in the northwest quadrant.) These data show additionally that the CH under consideration spans a region of complex magnetic ïeld. At the photospheric level, this includes a near-equatorial region of closed ïeld and even a ïeld of

opposite polarity. The magnetic ïeld structure, however, simpliïes signiïcantly with height and becomes almost simply connected already at R \ 1.1 R . Figure 2b shows the computed synoptic map at R \ 1.1_ , with the open R _ ïeld area projected on the solar sphere, superimposed on the BBSO Ha heliogram shown in Figure 1a. This composite picture demonstrates clearly that the ïlament 1-2-3 of our interest is really located inside the calculated open ïeld area, i.e., inside the CH at this altitude. An additional novel aspect of the magnetic ïeld under consideration is the comparison of the CH and other largescale features with calculated quasi-separatrix layers of the three-dimensional global coronal magnetic ïelds. Accord ing, for example, to Demoulin et al. (1997), Titov (1999), and Titov & Hornig (2001), the quasi-separatrix layers are the regions of drastic change in the ïeld-line linkage formed on the boundaries of dierent interacting magnetic ÿux systems. The quasi-separatrix layers are of interest because magnetic reconnection and energy release may occur there.

FIG. 6.õY ohkoh/SXT images illustrating development of the inside-CH soft X-ray arcade before (a), during (b), and after (c) the main ïlament eruption. (See also Fig. 1f.)


1230

CHERTOK ET AL.

Vol. 567

FIG. 7.õModiïed SOHO/EIT 195 A images at the eruption (aõd ) and posteruption (eõf ) stages illustrating development of curved element 4-1 and oval emitting feature 4-15-16-17-4 around the ïlament region as well as some other associated emitting structures. (See text.)

So far, the separators, the quasi-separatrix layers, and other similar peculiarities of the three-dimensional magnetic ïeld topology have been modeled for the local magnetic ïeld systems above isolated active regions, where they could be identiïed with ÿare-related features and bright loops observed in the Ha, soft X-ray, and EUV ranges (e.g., Gorb achev & Somov 1988 ; Demoulin et al. 1997 ; Wang et al. 2000). Now the technique has been further developed in IZMIRAN (Obridko et al. 2002), allowing us to compute large-scale quasi-separatrix layers in the global coronal magnetic ïeld. It is reasonable to expect, in particular, that some quasi-separatrix layers arise on the CH boundaries because here the open and closed magnetic ÿuxes interact with each other. The location of the quasi-separatrix layers is calculated by the so-called footpoint separation factor N, which characterizes a ratio of the starting point distances of the ïeld lines to their end-point distances. Figure 2c shows the spatial distribution of the separation factor N at a heliocentric level of 1.1 R on the synoptic chart corresponding _ to the eastern half of the disk of 1999 December 28. It can be seen that the biggest values of the factor N, i.e., the quasi-separatrix layers, outline the boundaries of the largescale structure, which appears to coincide with the CH, visible in the 284 A line and soft X-rays shown in Figures 1d and 1f, as well as with the calculated open ïeld area presented in Figure 2a. Thus, although the position of the ïlament 1-2-3 relative to the CH on the He I 10830 A spectroheliograms is ques tionable, other important data, such as the soft X-ray and

284 A images, computed open ïeld area, and quasi separatrix layers indicate the location of the ïlament inside the CH.
4

. MANIFESTATIONS OF THE FILAMENT ERUPTION

Similar to usual ïlament eruption events occurring outside CHs, in the case under consideration, noticeable activity began long before the eruption. At the 195 A wave length (Fig. 3), the bright curved emitting element 4-1, mentioned already in the ïrst paragraph of ° 3, develops from bright point 4 near the ïlament between 09 : 12 UT (Fig. 3a) and 13 : 25 UT (Fig. 3b). During several following hours, this feature continues to develop (Fig. 3c) and its intensity increases signiïcantly to 18 : 24 UT (Fig. 3d ). Some other emitting features are present in and near the ïlament area within the CH at this stage. On the 171 A SOHO/EIT pre eruption images (Fig. 4, top row), a similar but somewhat dierent ïlament-associated activity, particularly an analog of the near-the-ïlament emitting curved element 4-1, is visible. The MLSO/HAO Ha and He I 10830 A heliograms (Fig. 5) illustrate the pre-eruption evolution of the ïlament 1-2-3 at the time interval between 18 and 22 UT. The Ha brightening, positioned southeastward of the horseshoe-shaped ïlament 1-2-3, appears to be a partial counterpart of the curved emitting element 4-1 mentioned above (see also Figs. 1, 3, and 4). Comparison of two pairs of the images (Figs. 5a and 5d with Figs. 5b and 5e) shows that an almost absolute disappearance of the southern branch 3-2 of the ïlament


No. 2, 2002

DISAPPEARING FILAMENT INSIDE A CORONAL HOLE

1231

FIG. 8.õDierence (relative to 22 : 24 UT) SOHO/EIT 195 A images near the eruption stage displaying development of the curved element and oval emitting structure around the eruptive ïlament within the CH. (See also the caption and numbering of Fig. 7.)

occurred at about of 21 UT in both the Ha and He I 10830 A ranges. Just at this time, a signiïcant intensiïcation and enlargement of the soft X-ray emission in the near-the-ïlament region inside the CH takes place (Figs. 1f and 6a). Nearer to the main eruption time, the southern ïlament branch 3-2 becomes visible partly again on the Ha and He I images (Fig. 5c and 5f ; see also Fig. 1b). The occurrence of the main eruption of the whole ïlament at about 23 UT follows not only from a comparison of the nearest of the available Ha heliograms (Figs. 1b and 1c) but also from the detailed SOHO/EIT data at 195 A (Fig. 7). In addition to the various emitting features appearing

southward of the bright curved element 4-1 near point 17, the east (4-15) and north (15-16) diuse emitting branches are visible at 22 : 36 UT (Fig. 7a). Then at 23 : 00 UT (Fig. 7b), these branches transform rather sharply into a closed, oval, emitting structure 4-15-16-17-4 surrounding the ïlament region. Simultaneously, a few transient bright points appear near element 4-1. The oval structure remains pronounced and occupies a more or less stable position during at least several tens of minutes (Fig. 7bõ7d ), accompanied by some enhancement and swelling of the curved element 4-1. These features of the eruption stage of the event are more prominently visible on the SOHO/EIT 195 A dierence

FIG. 9.õRunning dierence SOHO/LASCO images of 1999 December 29 displaying the southeastern CME associated with the CH ïlament eruption. (Courtesy of S. Plunkett.)


1232

CHERTOK ET AL.

Vol. 567

FIG. 10.õPhenomena accompanying the eruption of the southeast ïlament (see Fig. 1aõ1c and Fig. 5) on 1999 December 29. (a) Huge soft X-ray cusp structure 18 with three legs extending to the northeast region 7 along the eastern CH boundary to the still-illuminated inside-CH ïlament area 4 and some southward of the erupted south ïlament 19. (b) 195 A counterparts of these structures. (c) Southeast CME registered with the SOHO/LASCO.

images (Fig. 8), where the 22 : 24 UT image was used as a background. It is reasonable to note that the CH, coinciding with the soft X-ray and 284 A one, is clearly distin guished on these images. In soft X-rays, the ïlament eruption manifested itself by signiïcant intensiïcation and development of an arcadelike structure (Fig. 6b). It is essential that this large-scale structure also occurred inside the soft X-ray CH and was located mainly south-southeast relative to the ïlament region, consistent with the projection of the local vertical. It seems that this direction was somewhat distinguished, because according to the SOHO/LASCO data (Fig. 9), a rather noticeable CME was observed just over the southsoutheastern limb after 23 : 30 UT. By our estimation, at the heliocentric distance range of 2.5õ7.0 R , the projected _ velocity of the CME increased from 270 to 520 km s~1. It is reasonable to suggest that this CME was associated with the ïlament erupted from within the CH under consideration. During all phases of the event, i.e., before, at the time of, and after the ïlament eruption on the 195 A images, one can see some thin emitting structures outlining the area, corresponding to the CH observed in soft X-rays and 284 A (see, for example, the contour 5-6-7-8-9-10-5 in Fig. 3c). According to the ïndings of Chertok (1999, 2001), these so-called large-scale chains are observed rather often, particularly at the boundaries of CHs. It should be noted that in this case, the CH boundary-associated chains appear to coincide also with the calculated quasi-separatrix layers of the global coronal magnetic ïelds (Fig. 2c). Such an association of the chains with energy release in large-scale quasiseparatrix layers was supposed by Chertok (1999, 2001). One more meridional transequatorial chain was observed at both 171 and 195 A lines and passed through the ïlament region 1-4 inside the CH extending at ïrst (Figs. 3d and 4b) from environs of the northeastern active complex 11 and then (Figs. 4c, 4d, and 7f ) to the southeastern CH boundary 13-5 and to the southeast limb edge 12 of the long south ïlament visible in the Ha and He I 10830 A lines (see Figs. 1 and 5). This ïlament-associated transient chain seems to resemble the patchy ribbons observed particularly in the He I 10830 A line on either side of the channel from which a ïlament has erupted (e.g., Harvey et al. 1987, 1996). The

development of the chains in the event under consideration will be discussed in more detail in a separate paper. The large-scale features described above suggest a close physical connection between the CH, the ïlament interior to it, the neighboring active region complexes 9 and 11, and the southeastern ïlament region 5 (e.g., Figs. 3, 4, and 7). The illumination and development of the large-scale emitting structures, particularly going to the southeastern limb, indicate strong evolution of this part of the global solar magnetic ïelds. It appears not to be a coincidence that on December 29 at about 15 UT, the southeastern ïlament also erupted. This resulted in a relatively large, very slow (182 km s~1), loop-like CME (Fig. 10c) as well as the huge over-the-limb soft X-ray cusp structure (18) with three diuse, leg branches (Fig. 10a). One leg, 18-7, visible also at 195 A (Fig. 10b), extends to the northeastern region along the east CH boundary, the second (18-4) stretches to the still-illuminated ïlament area within the CH, and the third (18-19) is positioned along but somewhat southward of the southern erupted ïlament. This implies close connections existing between all these structures and tends to conïrm their involvement in the large-scale eruptive events considered.
5

. DISCUSSION AND CONCLUSION

Our multirange analysis of the event of 1999 December 28õ29 demonstrates that sometimes large CHs may dier from usual ones with respect to the presence of such internal large-scale activity as a moderately long and eruptive ïlament. A number of arguments speak well of the probable location of the ïlament inside the CH inferred at the heliocentric distance 1.1 R . Although their relative position on _ the Kitt Peak and MLSO/HAO He I 10830 A images is debatable, the Y ohkoh/SXT and SOHO/EIT 284 A helio grams clearly show the ïlament lying within the CH. This conclusion is conïrmed also by calculations of two characteristics of the large-scale magnetic ïelds outlining particularly the CH boundaries : the open magnetic ïeld line area and, for the ïrst time, the quasi-separatrix layer patterns of the global coronal magnetic ïelds. The inferred probable location of the ïlament inside the CH means that some CHs may not be so simple and


No. 2, 2002

DISAPPEARING FILAMENT INSIDE A CORONAL HOLE

1233

uniform a structure as normally believed and can contain local areas of closed large-scale magnetic ïelds, at least at low altitudes. If the CH is to remain unipolar, another possible interpretation is that the ïlament and the bipolar region which it separates carve out an interior hole of more complicated magnetic topology that is not a part of the CH. The fact that this ïlament within a CH also erupted is one more peculiarity of the event under consideration. Noticeable pre-eruption ïlament activity, typical for usual ïlament eruptions occurring outside CHs, was observed in this case as well. The main eruption of the whole ïlament was accompanied by an obvious white-light CME, an appreciable restructuring of the EUV emission over the large area around the ïlament, a remarkable soft X-ray brightening and arcade formation, and some other phenomena. The observed features of the event suggest that the CH, internal and external ïlaments, and adjacent active regions are combined into a united large-scale active complex extending almost throughout the whole eastern half of the disk. The conception of such large-scale active complexes connected by large-scale magnetic ïelds was suggested by Mogilevsky et al. (1997). In our case, the observations show that a ïlament internal to a CH can play an important role in an evolving large-scale active complex, indicated above,

and that all components of the complex were strongly involved in the pre-eruption activity, the eruption itself, and post-eruption processes. At last, consideration of the Y ohkoh/SXT and SOHO/EIT data for other periods shows that similar but not so spectacular CH-interior ïlaments and extended accompanying activity can be found in some other events. The authors thank the anonymous referees for their valuable remarks, comments, and suggestions. We are grateful to members of the SOHO/EIT and LASCO, Y ohkoh/SXT, MLSO/HAO, BBSO, NAO (Mitaka), and NSO (Kitt Peak) teams for solar images and movies used in this study. We especially thank Simon Plunkett for the LASCO running dierence movie of the CME, as well as Yoshinori Suematsu and Andrew Stanger for the additional Mitaka and MLSO Ha images. The helpful consultations of Roman V. Gorgutsa and Victor V. Grechnev on the computer processing of various solar data are also acknowledged. This work was supported by the Russian Foundation of Basic Research (RFBR), the Russian Federal Program on Astronomy, and the INTAS/RFBR grant for the Joint Research Project on Support for the SOHO Solar Maximum Science Mission.

REFERENCES Brueckner, G. E., et al. 1995, Sol. Phys., 162, 357 Nolte, J. T., Davis, J. M., Gerassimenko, M., Krieger, A. S., Solodyna, Chertok, I. M. 1999, in Proc. 8th SOHO Workshop, ed. J.-C. Vial & C. V., & Golub, L. 1978, Sol. Phys., 60, 143 B. Kaldeich-Schurmann (ESA SP-446 ; Noordwijk : ESA), 229 Obridko, V. N., Kharshiladze, A. F., & Ivanov, K. G. 2002, in preparation õõõ. 2001, Sol. Phys., 198, 367 Obridko, V. N., & Shelting, B. D. 1999, Sol. Phys., 187, 185 ` Delaboudiniere, J.-P., et al. 1995, Sol. Phys., 162, 291 Recely, F., & Harvey, K. L. 1986, Solar-Terrestrial Predictions, ed. P. A. Demoulin, P., Bagala, L. G., Mandrini, C. H., Henoux, J. C., & Rovira, Simon, G. Heckman, & M. A. Shea (Boulder : NOAA), 204 M. G. 1997, A&A, 325, 305 Shibata, K., Nitta, N., Strong, K. T., Matsumoto, R., Yokoyama, T., HirayGorbachev, V. S., & Somov, B. V. 1988, Sol. Phys., 117, 77 ama, T., Hudson, H., & Ogawara, Y. 1994, ApJ, 431, L51 Harvey, K. L. 1995, in ASP Conf. Ser. 111, Magnetic Reconnection in the Solar-Geophysical Data 2000, Part I. (Boulder : NOAA), 666, 86 Solar Atmosphere, ed. R. D. Bentley & J. T. Mariska (San Francisco : Stepanyan, N. N., & Malanushenko, E. V. 2001, Izv. Krimskoi Astroïz. ASP), 9 Obs., 97, 63 Harvey, K. L., McAllister, A., Hudson, H., Alexander, D., Lemen, J. R., & Titov, V. S. 1999, Bull. Russian Acad. Sci. & Phys., 63, 1497 Jones, H. P. 1996, in ASP Conf. Ser. 95, Solar Drivers of Interplanetary Titov, V. S., & Hornig, G. 2001, Adv. Space Res., in press and Terrestrial Disturbances, ed. K. S. Balasubramaniam, S. L. Keil, & Tsuneta, S., et al. 1991, Sol. Phys., 136, 37 R. N. Smartt (San Francisco : ASP), 100 Wang, H., Yan, Y., Sakurai, T., & Zhang M. 2000, Sol. Phys., 197, 263 Harvey, K. L., Sheeley, N. R., Jr., & Harvey, J. W. 1987, in Proc. Solar Zhao, X. P., Hoeksema, J. T., & Scherrer, P. N. 1999, J. Geophys. Res., 104, Terrestrial Predictions Workshop, ed. P. A. Simon, G. Heckman, & 9735 M. A. Shea (Boulder : NOAA), 198 Zirker, J. B., ed. 1977, Coronal Holes and High-Speed Wind Streams Kahler, S. W., & Moses, D. 1990, ApJ, 362, 728 (Boulder : Colorado Associated Press) Mogilevsky, E. I., Obridko, V. N., & Shilova, N. S. 1997, Sol. Phys., 176, 107