0711.3952/ms.tex
1: \documentclass[namedreferences]{SolarPhysics}
2: \usepackage{epsfig}
3: \usepackage[optionalrh]{spr-sola-addons} 
4: \usepackage{url}
5: \urlstyle{sf}
6: 
7: 
8: \newcommand{\etal}{et al. }
9: \newcommand{\be}{\begin{equation}}
10: \newcommand{\ee}{\end{equation}}
11: \newcommand{\beq}{\begin{eqnarray}}
12: \newcommand{\eeq}{\end{eqnarray}}
13: \newcommand{\ietal}{{\it et al.}}
14: \newcommand{\eg}{{\it e.g.}}
15: \newcommand{\ie}{{\it i.e.}}
16: 
17: 
18: \newcommand{\adv}[3]{   #1, {\it Adv. Space Res.}    {\bf  #2}, #3.}
19: \newcommand{\aeta}[3]{  #1, {\it Astron. Astrophys.} {\bf  #2}, #3.}
20: \newcommand{\aetap}[1]{ #1, {\it Astron. Astrophys.} in press.}
21: \newcommand{\aetas}[3]{ #1, {\it Astron. Astrophys. Suppl. series} {\bf  #2}, #3.}
22: \newcommand{\aetasu}[1]{#1, {\it Astron. Astrophys.} submitted.}
23: \newcommand{\aspj}[3]{  #1, {\it Astrophys. J.} {\bf  #2}, #3.}
24: \newcommand{\aspjp}[1]{ #1, {\it Astrophys. J.} in press.}
25: \newcommand{\aspjsu}[1]{#1, {\it Astrophys. J.} submitted.}
26: \newcommand{\gafd}[3]{  #1, {\it Geophys. Astrophys. Fluid Dyn.} {\bf  #2}, #3.}
27: \newcommand{\grl}[3]{   #1, {\it Geophys. Res. Lett.} {\bf  #2}, #3.}
28: \newcommand{\jogr}[3]{  #1, {\it J. Geophys. Res.} {\bf  #2}, #3.}
29: \newcommand{\jogrsu}[1]{#1, {\it J. Geophys. Res.} submitted.}
30: \newcommand{\jogrp}[1]{ #1, {\it J. Geophys. Res.} in press.}
31: \newcommand{\jfm}[3]{   #1, {\it J. Fluid. Mech.} {\bf  #2}, #3.}
32: \newcommand{\pfl}[3]{   #1, {\it Phys. Fluids} {\bf  #2},  #3.}
33: \newcommand{\sph}[3]{   #1, {\it Solar Phys.} {\bf  #2}, #3.}
34: \newcommand{\sphp}[1]{  #1, {\it Solar Phys.} in press.}
35: \newcommand{\sphsu}[1]{ #1, {\it Solar Phys.} submitted.}
36: \newcommand{\SpaceS}[3]{#1, {\it Space Sci Rev.} {\bf  #2}, #3.}
37: \newcommand{\pasj}[3]{#1, {\it Publ. Astron. Soc. Japan} {\bf  #2}, #3.}
38: 
39: \begin{document}
40: \begin{article}
41: \begin{opening}
42: 
43: \title{Large-Amplitude Oscillation of an Erupting 
44: Filament as Seen in EUV, H$\alpha$ and Microwave Observations}
45: 
46: \author{H. \surname{Isobe}$^{1}$
47:         D.   \surname{Tripathi}$^{2}$
48:         A. \surname{Asai}$^{3}$
49:         R. \surname{Jain}$^{4}$}        
50: 
51: 
52: \runningauthor{H. Isobe et al.}
53: \runningtitle{}
54: 
55: 
56: \institute{$^{1}$ Department of Earth and Planetary Science, University 
57:              of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
58:                   \email{isobe@eps.s.u-tokyo.ac.jp}\\
59:            $^{2}$ Department of Applied Mathematics and Theoretical 
60:                   Physics, University of Cambridge, 
61:                   Wilberforce Road, Cambirdge CB3 0WA, UK\\
62:            $^{3}$ Nobeyama Radio Observatory, National Astronomical 
63:                   Observatory of Japan, Minamimaki, Minamisaku, 
64:                   Nagano 384-1305, Japan
65:                   \\
66:            $^{4}$ Department of Applied Mathematics, University 
67:                   of Sheffield, Hicks building, Hounsfield Road, 
68:                   Sheffield S3 7RH, UK
69:              }
70: 
71: \date{Received ; accepted }
72: 
73: \begin{abstract}
74: We present multiwavelength observations of a large-amplitude 
75: oscillation of a polar crown filament on 15 October 2002, 
76: which has been reported by Isobe and Tripathi 
77: ({\it Astron. Astrophys.} {\bf 449}, L17, 2006). 
78: The oscillation occurred during the slow rise ($\approx$ 1 km s$^{-1}$)
79: of the filament. It completed three cycles 
80: before sudden acceleration and eruption.  
81: The oscillation and following eruption were clearly seen 
82: in observations recorded by the Extreme-Ultraviolet Imaging 
83: Telescope (EIT) onboard the Solar and Heliospheric Observatory (SOHO). 
84: The oscillation was seen only in a part of the filament, 
85: and it appears to be a standing oscillation rather than 
86: a propagating wave.  
87: The amplitudes of velocity and spatial displacement of the 
88: oscillation in the plane of the sky were about 5 km s$^{-1}$ 
89: and 15,000 km, respectively. The period of oscillation was about two hours 
90: and did not change significantly during the oscillation. 
91: The oscillation was also observed in H$\alpha$ by the 
92: Flare Monitoring Telescope at Hida Observatory. We determine the 
93: three-dimensional motion of the oscillation from the 
94: H$\alpha$ wing images. The maximum line-of-sight velocity 
95: was estimated to be a few tens of km s$^{-1}$, though the uncertainty 
96: is large owing to the lack of the line-profile information. 
97: Furthermore, we also identified the spatial displacement of 
98: the oscillation in 17 GHz microwave images from 
99: Nobeyama Radio Heliograph (NoRH). 
100: The filament oscillation seems to be triggered by 
101: magnetic reconnection between a filament barb and 
102: nearby emerging magnetic flux as was evident from the 
103: MDI magnetogram observations. 
104: No flare was observed to be associated with the onset of the oscillation.
105: We also discuss possible implications of the oscillation 
106: as a diagnostic tool for the eruption mechanisms. 
107: We suggest that in the early phase of eruption a part of the filament lost 
108: its equilibrium first, while the remaining part was still in an equilibrium and 
109: oscillated. 
110: \end{abstract}
111: \keywords{Prominences, Dynamics; Oscillations; Coronal Seismology}
112: 
113: \end{opening}
114: 
115: \section{Introduction}
116: \label{Introduction} 
117: 
118: A variety of oscillatory motions of prominences and filaments have been 
119: observed and used to infer their structure and physical parameters 
120: (see the review by Ballester, 2006). 
121: In particular, global oscillations of filaments initiated 
122: by flare-associated disturbances (Ramsey and Smith, 1966) 
123: may provide useful information on the properties of filaments 
124: as oscillators, and hence on their stability. 
125: 
126: Isobe and Tripathi (2006; hereafter paper I) reported an event 
127: in which a polar-crown filament exhibited large amplitude oscillation 
128: in its pre-eruptive slow rise phase. 
129: This event provides a rare opportunity to use so-called 
130: coronal seismology or prominence seismology 
131: in order to study the eruption mechanism of filaments. 
132: In paper I we reported that the oscillation repeated for about three 
133: cycles without any significant damping, until the onset of rapid acceleration 
134: and eruption of the filament. The period and the velocity amplitude 
135: measured in the plane of sky 
136: of the oscillation were about 2 hours and 4.2 km s$^{-1}$, respectively. 
137: It was suggested that the large-amplitude oscillation in the 
138: filament was evidence for it retaining a non-linear stable 
139: equilibrium in its pre-eruptive, slow-rise phase. The 
140: transition from this equilibrium to eruption, \ie, loss of 
141: equilibrium (LOE) or instability, was triggered by rapid ($\approx$ 
142: Alfv\'{e}n time scale) mechanism such as magnetic reconnection. 
143: 
144: In this paper we present further detailed analyses of this event using 
145: multiwavelength observations. 
146: In addition to the detection of the oscillation in extreme ultraviolet  
147: and H$\alpha$, we discovered that the oscillation was seen 
148: also in microwave. 
149: To the best of our knowledge, this is the first evidence of spatially resolved, 
150: large-amplitude oscillation of a filament in microwave imaging. 
151: One of the fundamental issues that remained unknown in paper I was 
152: the physical cause behind excitation of the oscillation. 
153: It is known that large amplitude filament oscillations are 
154: usually excited by flare-associated disturbances 
155:  (see \eg, Ramsey and Smith,  1966; Eto \ietal,  2002; Jing \ietal, 2003, 2006; 
156: Okamoto \ietal, 2004; Vr\v{s}nak \ietal, 2007). 
157: However, no flare associated with the onset of oscillation in this event was observed. 
158: In this paper we examine the magnetic-field data to study the mechanism 
159: of the excitation of the oscillation.  
160: 
161: The paper is organized as follows: In Section 2 
162: we briefly summarize the observational data. In Section 3 
163: the oscillation seen in various wavelengths is presented. 
164: We discuss the possible excitation mechanisms responsible for the 
165: filament oscillation and their implication to eruption mechanisms in Section 4. 
166: Conclusions are summarized in Section 5. 
167: 
168: 
169: \section{Observations}
170: \label{observations}
171: The oscillation and subsequent eruption of a filament was  
172: observed on 15 October 2002. The filament was a large polar 
173: crown in the southern polar region. 
174: The oscillation was most clearly seen in 195 \AA \, images 
175: taken by the Extreme-Ultraviolet Imaging Telescope 
176: (EIT; Delaboudini\`{e}re \ietal, 1995) onboard the 
177: {\it Solar and Heliospheric Observatory} (SOHO).
178: The cadence and spatial resolution of EIT are  
179: 12 minutes and 2.6 arcsec, respectively. 
180: The 195~{\AA} passband of the EIT is dominated by an 
181: Fe \, {\sc xii}  
182: line formed at 1\,--\,2 MK, but also 
183: contains an Fe \,{\sc xxii} line at 192~{\AA} formed at around 
184: 20 \,MK, which is usually much weaker in the quiet Sun 
185: region but highly significant in flaring active region (Tripathi \ietal, 2006). 
186: The filament (cold plasma in the corona) is seen as a dark feature due to 
187: absorption in the H\,{\sc i} Lyman continuum (Kucera, Andretta and Poland, 1998; 
188: Engvold \ietal, 2001). 
189: We also used the line-of-sight magnetograms recorded by the Michelson Doppler Imager 
190: (MDI: Scherrer \ietal, 1995) onboard SOHO to study the photospheric magnetic-field configuration. 
191: 
192: The filament oscillation was also observed by the Flare 
193: Monitoring Telescope (FMT) at Hida Observatory, Kyoto University.
194: FMT observes the full disk of the Sun in five channels including 
195: H$\alpha$ center and wings at $\pm 0.8$ \AA \, with the time 
196: cadence of two seconds and a pixel sampling of 4.25 arcsec 
197: (Kurokawa \ietal, 1995). The advantage of FMT is its continuous, 
198: full disk observations in H$\alpha$ wings. It has been used 
199: for the studies of transient and dynamic events 
200: such as filament oscillation (Eto \ietal, 2002; Okamoto \ietal, 2004), 
201: filament eruptions (Morimoto and Kurokawa, 2003b) and surges 
202: (Liu \ietal, 2005). 
203: 
204: The spatial displacement associated with the oscillation was also identified 
205: in the microwave observation at 
206: 17 GHz from Nobeyama Radio Heliograph (NoRH; Nakajima \ietal, 1994). 
207: The advantage of microwave observation is that it is insensitive 
208: to temperature and velocity of filaments/prominences, \ie, 
209: the visibility of the filament does not change even when 
210: heating or motion of the filament occur. Hence the NoRH 
211: data have been used for the studies of prominence eruptions 
212: (Gopalswamy \ietal, 2003; Shimojo \ietal, 2006)
213: 
214: All of the images shown in this paper were rotated to the 
215: reference time of 06:00 UT to compensate for the effect 
216: of solar rotation. We used the \url{DROT_MAP} procedure in the Solar SoftWare 
217: to correct the rotation, and neglect the projection effect caused by 
218: the height of the filament above the photosphere. 
219: 
220: 
221: \section{Results}
222: \subsection{Oscillation and eruption as seen by EIT}
223: \label{subsec:eitosc}
224: 
225: Figure 1 shows the EIT images of the filament during  
226: the oscillation (left), around the time of the onset of eruption (middle) 
227: and after the eruption (right). The $x$ and $y$-axes are 
228: the distances from the disk center. 
229: Before the eruption, the eastern half ($x < -250$ arcsec) 
230: of the filament was seen above the limb. 
231: (Therefore, part of the filament was actually seen as a ``prominence", 
232: though we use the term of ``filament" to refer to the whole entity.) 
233: The western half ($x > -250$ arcsec) was seen as a dark filament on the disk. 
234: The apparent morphology of the filament is different 
235: between the eastern and western parts, likely because of projection effects. 
236: It seems that the eastern 
237: part is a vertical slab seen from the side, and the western  
238: part has similar structure but is seen from the top. 
239: After the eruption, brightenings at the eastern 
240: and western ends of the filament (right panel: 
241: [$x,y$]=[-650, -500] and [-100, -500]) are evident. 
242: 
243: The oscillatory motion of the filament can be visualized by 
244: the temporal slices along the slits perpendicular to the filament. 
245: Figure 2 shows the temporal slices along 
246: seven different slits shown in the left panel of Figure 1. 
247: The right-bottom panel shows the time slice for slit three 
248: but for the microwave image; see Section 3.2.  
249: The oscillation starts at around 02:30 UT and completes 
250: three cycles with a period of about two hours. 
251: During the fourth cycle the filament is suddenly accelerated 
252: and erupted. 
253: 
254: The oscillation is most prominent at slit three. 
255: The amplitude of the oscillation is smaller in the slits 
256: away from slit three in both directions and is almost unrecognizable in slit seven.  
257: The vertical black lines indicate the times of the 
258: first peak (04:00 UT) and the second peak (06:00 UT) 
259: as determined by visual inspection from the time slice of slit three.  
260: Although there are some ambiguities in determining the peaks, 
261: the oscillatory motion at the different slit position  
262: appears to be in the same phase. Therefore, the oscillatory 
263: motion can be interpreted as a standing oscillation. 
264: The amplitude of velocity and spatial displacement are 
265: about 5 km s$^{-1}$ and 15,000 km, respectively. 
266: Note that, however, these values are measured in the plane of the sky. 
267: 
268: In Paper I it was suggested that there may be an increase in 
269: the period based on the visual inspection of the time slices, 
270: although this was not conclusive. To study the 
271: change of period  accurately, a wavelet analysis was 
272: performed on the EIT data; no substantial increase in the period 
273: was found (Pint\'{e}r \ietal, 2007). 
274: This result has a significant influence on the diagnosis of the trigger mechanism for the 
275: filament eruption (see Section \ref{dis-erup}).
276: 
277: The remarkable point of this event is that the oscillation 
278: occurred while the filament is slowly rising toward the eruption. 
279: The slow rise motion is seen in Figure 2 as a general trend 
280: toward the southwest with a velocity of about 1 km s$^{-1}$ 
281: (See also Figure 2 of paper I). Let us examine the effect of the 
282: correction of rotation on the velocity of the filament motion. 
283: As previously mentioned, we neglected the height of the filament 
284: when correcting the solar rotation. The error caused by this simplification 
285: is $\lesssim H\theta$, where $H$ is the filament height and $\theta$ is 
286: the rotated angle. During the 12 hours in which we made the time slices, 
287: the photosphere at this latitude rotates about 5\,--\,6 degrees. If $H=10^5$ km, 
288: which seems reasonable, the error in terms of the velocity is about 0.25 km s$^{-1}$ 
289: or less. Thus the error is smaller than the velocities of the slow rise and 
290: the oscillation. 
291: 
292: 
293: \subsection{NoRH and FMT observation of oscillation}
294: Figure 3 shows the H$\alpha$ line center image from FMT (left panel) 
295: and the 17 GHz image from NoRH (right panel). 
296: Unfortunately the eruption occurred 
297: during the night for FMT and NoRH, but the first period of the oscillation 
298: was within the observation time of these instruments. 
299: The white line in the microwave image (right panel) shows 
300: the position of slit three shown in Figure 1 The time slice of 
301: this slit is shown in Figure 2 (panel h) which reveals a   
302: clear oscillatory motion. 
303: 
304: Figure 4 shows the time series of H$\alpha$ images at line center 
305: and blue (-0.8 \AA) and red (+0.8 \AA) wings. The field of 
306: view is shown by the rectangle in Figure 3. 
307: The irregular 
308: time cadence is due to frequent interruptions by clouds. 
309: During 04:00 -- 05:00 UT the filament 
310: appears in the blue wing images as a dark feature. The filament 
311: fades away in the blue wing around 05:00 UT, and then it appears 
312: in the red wing (05:02 and 05:27 UT). This is the well-known 
313: feature of oscillating filaments called ``winking filament''
314: (Hyder, 1966). 
315: 
316: The three-dimensional velocity field can be inferred 
317: by comparing the H$\alpha$ wing images and the motion  
318: in the plane of the sky. When the filament appears in the blue (red)  wing, 
319: which means it is moving toward (away from) the Earth, 
320: it is moving toward Northeast (Southwest) in the plane of the sky 
321: (see Figure 2). 
322: Given that the filament is located near the southern limb 
323: of the Sun, if the direction of the oscillatory motion is vertical, 
324: the blue (red) shift should accompany southward (northward) motion. 
325: Since in this case the sense of oscillation appears to be opposite, 
326: we conclude that the oscillatory motion is horizontal. 
327: The contrast of the wing images is largest (\ie, the filament is dark)  
328: in the central area of the field of view, where the plane-of-the-sky 
329: velocity is also largest; see the position of slit 3 shown 
330: in the right panel of Figure 4. This is consistent with the 
331: interpretation as a standing oscillation. 
332: 
333: The line-of-sight velocity of the filament can be inferred
334: from the contrast in the wing images. Morimoto and Kurokawa (2003a) 
335: developed a method to determine the line-of-sight velocity 
336: from FMT data based on a cloud model. Using their method, 
337: we estimated the line-of-sight velocity of the 
338: oscillation to be about 20\,--\,30 km s$^{-1}$ (see Paper I). 
339: This value depends on the line profile of the filament, 
340: which is unknown here (whereas in paper I we assumed a typical line profile 
341: presented in Morimoto and Kurokawa, 2003a). 
342: Therefore the uncertainty in the line-of-sight velocity is large. 
343: 
344: 
345: \section{Discussion}
346: \subsection{What is the trigger of the oscillation?}
347: \label{efr}
348: It is known that large-amplitude oscillations of filaments like 
349: this event are excited for flare-associated disturbances, such as 
350: Moreton waves (Ramsey and Smith, 1966; Eto \ietal, 2002), 
351: EIT waves (Okamoto \ietal, 2004), and nearby subflares (Jing \ietal, 2003, 2006). 
352: There was no observed flare or brightening associated with the present event that 
353: can account for the excitation of the oscillation. 
354: However, we identify an emerging flux region that may play a role 
355: in triggering the oscillation, and possibly in triggering the eruption as well. 
356: 
357: Figure 5 shows EIT images taken just before the onset of the oscillation. 
358: An MDI magnetogram taken at 03:12 UT is overlaid on 
359: the EIT image taken at 02:48 UT shown in the left panel. 
360: The blue and red contours correspond to +25 and -25 gauss, respectively. 
361: As indicated in the figure, there is a bipole whose opposite polarities are 
362: connected by a bright EUV loop.
363: The lower panels of Figure 5 show temporal evolution of the 
364: magnetogram in the black box shown in the upper left panel. 
365: The bipole is already visible at 19:15 UT, October 14, and evolves 
366: with time as see in Figure 5. The separation of the two polarities 
367: is the characteristic of an emerging flux. 
368: 
369: Also indicated in Figure 5 is a filament 
370: barb connected to the preexisting positive polarity region 
371: near the negative polarity of the emerging flux (02:48). 
372: This barb was seen for hours before the onset of oscillation. 
373: At 03:12 UT it suddenly disappears from the EIT images (see the middle panel of Figure 5) 
374: and a very faint jet-like ejection was seen at 03:36 UT toward the southwest. 
375: This ejection is co-temporal with the 
376: onset of oscillation and also co-spatial with the part of the 
377: filament where the amplitude of the oscillation is largest.  
378: We interpret these observations as follows. 
379: As the new flux emerged and expanded into the corona,  it interacted 
380: with the filament barb and magnetic reconnection occurred. 
381: This reconnection produced the plasma ejection, which blew the filament 
382: from the side and excited the oscillation. 
383: We note that since we could not identify the polarity of the footpoint of the barb in 
384: the magnetogram, we do not know whether the direction of the 
385: magnetic field was favorable for reconnection or not. 
386: However, three-dimensional MHD simulations suggest that 
387: reconnection between an emerging flux and a coronal field can occur 
388: regardless of their mutual orientation, though the resultant dynamics and 
389: magnetic field configuration vary 
390: (Archontis \ietal, 2005; Galsgaard \ietal, 2007). 
391: 
392: There was no brightenings in EIT images associated with the 
393: ejection, probably because the released energy is too small to 
394: brighten EUV or the reconnection occurred in the chromosphere. 
395: It is also possible that the time scale of brightening associated with the 
396: reconnection was shorter than the EIT cadence, and the brightening 
397: was not recorded by the EIT. 
398: 
399: 
400: \subsection{Internal structure of the filament}
401: So far we considered the filament as a single entity. However, the filament 
402: has some internal structure, as seen in the EIT images in Figure 5. One can 
403: also see internal structure of the filament structure in Figure 3,  which may indicate the 
404: presence of substructures that could be oscillating a little out of phase. 
405: Indeed, a movie of EIT images shows some motion along the filament. 
406: This is probably due to the direction of the perturbation that excited the 
407: oscillation (Jing \ietal, 2006; Vr\v{s}nak \ietal, 2007). As indicated by the long arrow in Figure 5,   
408: the direction of the ejection that blows the filament was oblique to the filament axis, 
409: which possibly excited both lateral and longitudinal motions in the filament.
410: 
411: 
412: \subsection{What is the implication for eruption mechanisms?}
413: \label{dis-erup}
414: The large-amplitude oscillation occurred while the filament is 
415: slowly rising before eruption. Slow-rise motions before 
416: the fast acceleration and eruption have been commonly observed 
417: in various types of solar plasma ejections, such as filament 
418: eruption (Sterling and Moore, 2003, 2005; Chifor \ietal, 2006), 
419: X-ray plasmoid ejections (Ohyama and Shibata, 1997), 
420: and coronal mass ejections (Zhang \ietal, 2001). 
421: A fast-acceleration phase following the slow-rise phase has also been found 
422: in numerical simulations (\eg, Magara, Shibata, and Yokoyama, 1997; 
423: Chen and Shibata, 2000). 
424: The temporal evolution of the height of the ejecta seen in observations is similar 
425: to those in simulations where the ejecta already lost their stable equilibrium 
426: (for comparison of observation  and simulation, see, \eg, 
427: Magara, Shibata, and Yokoyama, 1997; 
428: Sterling and Moore, 2005). 
429: This is suggestive of an initial growth of an instability or LOE as 
430: an explanation of the slow rise. 
431: Such slow rise usually lasts several tens of the Alfv\'{e}n time of the system
432: (\ie, time in which Alfv\'{e}n waves travel across the system), 
433: typically $\approx 10^3$ seconds for active region events and 
434: several hours for quiet region filaments. 
435: 
436: However, a slow rising motion of a filament that lasted much longer 
437: (period $\approx$ days) before its eruption has been reported by Nagashima \ietal (2007).  
438: They interpreted  the motion as a quasi-static evolution toward the critical height 
439: at which the filament lost its equilibrium (Forbes and Isenberg, 1991).  
440: The quasi-static evolution (a series of MHD equilibria) may be driven by 
441: photospheric motions (Forbes and Priest, 1995), an emerging flux 
442: (Lin, Forbes, and Isenberg, 2001), 
443: or continuous small scale reconnection in the lower atmosphere (Nagashima \ietal, 2007). 
444: This kind of slow rise is physically different from the former type,  
445: that is, initial growing phase of the LOE or instability. 
446: 
447: The oscillation during the slow-rise phase provides strong evidence 
448: that the prominence retained its equilibrium during that time, which 
449: indicates that the slow rise found in our event is of the latter type: 
450: the quasi-static evolution {\it before}  the LOE or instability sets in. 
451: However, our event also shows characteristics to those of the former type, 
452: the initial growing phase {\it after} the LOE or instability sets in. 
453: The height-time plot of the filament is quite similar to those  
454: of the former type; it shows a smooth transition from slow rise with constant 
455: velocity to the fast acceleration (see time slice for slit seven in Figure 3 and Figure 2 
456: of Paper I).  The duration of slow rise in our event ($>$ 12 hours) seems 
457: quite long as that of the former type, but it is probably due to the long Alfv\'{e}n time 
458: ($\sim$ tow hours, see Paper I). 
459: 
460: The oscillation continued until the onset of the fast acceleration. 
461: The amplitude of the oscillation was large; the displacement in the plane of the 
462: sky ($\approx$ 20000 km) is as large as the width of the filament. Therefore, during 
463: the oscillation the equilibrium of the filament, or at least its oscillating part, 
464: was nonlinearly stable to the large perturbation. 
465: However, if the eruption occurs through the exchange of stability, we expect that the period 
466: of the oscillation becomes longer and longer as the system changes its equilibrium state 
467: and approaches the eruption. 
468: As mentioned in Section \ref{subsec:eitosc}, the wavelet analysis showed 
469: no evidence for an increase in period (Pint\'{e}r \ietal, 2007). 
470: Therefore it is unlikely that the slow rise is due to the change of the equilibrium state 
471: toward the critical point at which the LOE or instability sets in. 
472: Moreover, the transition from the nonlinearly 
473: stable equilibrium to eruption occurred within a period of the oscillation or, in other words, 
474: in the Alfv\'{e}n time scale. 
475: We note that emerging flux and its reconnection with the filament barb that 
476: triggered the oscillation may have played a role in triggering the eruption, too
477: (Feynman and Martin, 1995; Chen and Shibata, 2000; 
478: Lin, Forbes, and Isenberg, 2001; Tripathi, 2005). 
479: 
480: However, the following question remains:  
481: Does the oscillation found in our event suggest that the similar slow rises found commonly in 
482: other events (Ohyama and Shibata, 1997, Zhang \ietal, 2001; Sterling and Moore, 2003, 2005; 
483: Chifor \ietal, 2006) also indicate quasi-static evolution, in contrast to the previous interpretation 
484: as the initial phase of LOE or instability? This may be the case, but 
485: here we propose another scenario. 
486: 
487: We note that the oscillation occurred only in a part of the filament. The size of 
488: the oscillating part of the filament is about $2\times10^5$ km (approximately 
489: the size of the box in Figure 3), whereas the size of the complete erupting filament 
490: is as large as the solar radius. In such a long filament, it is probably possible that 
491: the filament loses its equilibrium in a ``global" sense, but a part of the 
492: filament still retains ``local" equilibrium that can support ``local" oscillation. 
493: Indeed, in many filament eruption events the whole filament does not erupt 
494: simultaneously, but often a part of the filament erupts first and the remaining 
495: parts seem to be dragged by the leading part (a process called ``asymmetric eruption" by 
496: Tripathi, Isobe, and Mason, 2006. See also Isobe, Shibata, and Machida, 2002). 
497: Chifor \ietal (2006) analyzed X-ray, EUV, and microwave brightenings of 
498: such an asymmetric eruption and found  (i) that the filament started to rise 
499: from one end where an X-ray precursor was found, and (ii) there was a weak heating of the 
500: filament during the slow rise. They concluded that at first a part of the filament 
501: started to rise owing to the  localized magnetic reconnection below the filament.  
502: Then the rise of the filament induced successive magnetic reconnection 
503: along the filament, and eventually the whole filament was destabilized and erupted 
504: (see  also Chifor \ietal, 2007). This scenario naturally explain how a localized 
505: perturbation (such as an emerging flux) eventually leads to the destabilization 
506: of the whole filament. 
507: 
508: We suggest that the observation presented in this paper 
509: further supports the model of Chifor \ietal, (2006). 
510: The oscillation in a part of the filament during the slow rise is evidence 
511: that that part of the filament still retained an equilibrium. However, the other 
512: nonoscillating part had already lost the equilibrium and started to rise, dragging 
513: the oscillating part.  Thus the oscillating part was also forced rise, and 
514: eventually erupted along with the other parts of the prominence.
515: 
516: 
517: 
518: \section{Conclusions}
519: In this paper we presented the observations of a large amplitude 
520: oscillation of the polar crown filament in its pre-eruption, 
521: slowly rising phase. The oscillation was seen only in a part of the 
522: filament, and seems to be a standing oscillation rather than 
523: a propagating wave. 
524: The spatial displacement of the oscillation was also identified 
525: in 17 GHz microwave images from NoRH. 
526: Such an oscillation has not been reported during the more than 
527: ten years of NoRH observations (although we have not done a systematic survey 
528: for such events, so this does not mean that ours is the only such event), 
529: and we believe that this is the first observation of spatially resolved 
530: filament oscillation by microwave imaging. 
531: 
532: In Contrast to previous observations of large-amplitude filament 
533: oscillations, there were no flares or even microflares associated 
534: with the excitation of the filament. However, we found that a 
535: filament barb connected near an emerging flux disappeared 
536: and seems to be ejected at the onset of the oscillation. 
537: We believe that reconnection between the emerging flux 
538: and the barb caused the ejection that blew the filament 
539: to excite the oscillation. 
540: 
541: The large-amplitude oscillation occurred in the slow-rise phase of the filament, 
542: and only a part of the filament exhibited the oscillation. 
543: The oscillation is evidence that the oscillating part of the filament still retained 
544: a nonlinearly stable equilibrium during the slow rise. 
545: It supports the model by Chifor \ietal (2006) that 
546: at first a part of filament loses equilibrium and starts to rise, and the slow rise 
547: induces LOE of the remaining part, leading to the 
548: destabilization of the whole filament. 
549: 
550: Filament and prominence seismology is a powerful tool, not only for 
551: measuring the physical parameters, but also for diagnosing 
552: the stability and hence the trigger mechanism of filament eruption. 
553: It is worthwhile to search for more events in which 
554: filaments exhibit large-amplitude oscillation in their 
555: pre-eruption phase. 
556: From a theoretical point of view, there are many analytical and 
557: numerical models of pre-eruption filaments or flux ropes.
558: Investigation of the behavior of these model filaments 
559: as oscillators may provide some physical insight into 
560: the mechanisms of filament eruptions. 
561: 
562: 
563: \acknowledgements
564: We thank M. Shimojo for fruitful discussions. 
565: This work was supported in part by a Grant-in-Aid for 
566: Creative Scientific Research of the Ministry of Education, 
567: Culture, Sports, Science and Technology
568: (MEXT) of Japan ``The Basic Study of Space Weather Prediction"
569: (17GS0208, K. Shibata, Principal Investigator).
570: HI is supported by a Research Fellowship from the Japan 
571: Society for the Promotion of Science for Young Scientists. 
572: DT acknowledges the support from PPARC. 
573: RJ acknowledges The Nuffield Foundation (UK) for NUF-NAL 04 award.
574: 
575: 
576: \begin{thebibliography}{}
577: 
578: \bibitem[2005]{archontis05} Archontis, V., Moreno-Insertis, F., Galsgaard, K., 
579:  Hood, A.W.: \aspj{2005}{635}{1299}
580: 
581: \bibitem[\protect\citeauthoryear{Ballester}{2006}]{balle06}
582: Ballester, J.L.: 2006, {\it Phil. Trans. Roy. Soc. A} {\bf 364}, 405
583: 
584: \bibitem[2000]{chenp00} Chen, P.F., Shibata, K.: \aspj{2000}{545}{524}
585: 
586: \bibitem[2006]{chifo06} Chifor, C., Mason, H.E., Tripahi, D., 
587:  Isobe, H., Asai, A.: \aeta{2006}{458}{965}
588:  
589: \bibitem[2007]{chifor07} Chifor, C., Tripathi, D., Mason, H.E., Dennis, B.R.: 
590:  \aeta{2007}{472}{967}
591:  
592: \bibitem[1995]{delab95} Delaboudiniere, J.-P. \ietal: \sph{1995}{162}{291}
593: 
594: \bibitem[2001]{engvold01} Engvold, O., Jakobsson, H., Tandberg-Hanssen, E., 
595: Gurman, J.B., Moses, D. \sph{2001}{202}{293}
596: 
597: \bibitem[2002]{eto04} Eto, S., Isobe, H., Narukage, N. \ietal:
598: \pasj{2002}{54}{481}
599: 
600: \bibitem[1995]{feynm95} Feynman, J., Martin, S.F.: 
601:  \jogr{1995}{10}{3355}
602: 
603: \bibitem[1991]{forbes91} Forbes, T.G., Isenberg, P.A.: 
604:  \aspj{1991}{373}{294}
605: 
606: \bibitem[1991]{forbes91} Forbes, T.G., Priest, E.R.: 
607:  \aspj{1995}{446}{377}
608: 
609: \bibitem[2007]{galsgaard07} Galsgaard, K., Archontis, V., Moreno-Insertis, F. 
610:  Hood, A.W.: \aspj{2007}{666}{516}
611: 
612: \bibitem[2003]{gopal03} Gopalswamy, N., Shimojo, M., Lu, W., 
613:  Yashiro, S., Shibasaki, K., Howard, R.A.: 
614:  \aspj{2003}{586}{562}
615: 
616: \bibitem[1966]{hyder66} Hyder, C. L.: 1966, 
617:  {\it Zeit. Astrophys.}, {\bf 63}, 78
618: 
619: \bibitem[2002]{isobe02} Isobe, H., Shibata, K., Machida, S.: 2002, 
620:  {\it Geophys. Res. Lett.,} 29(21), 2014, \url{doi:10.1029/2001GL013816}
621: 
622: \bibitem[2006]{isobe06} Isobe, H., Tripathi, D.: \aeta{2006}{449}{L17}
623: 
624: \bibitem[2006]{jing06} Jing, J., Lee, J., Spirock, T.J., Wang, H.: 
625:   \sph{2006}{236}{97}
626: 
627: \bibitem[2003]{jing03} Jing, J., Lee, J., Spirock, T.J., Xu, Y., Wang, H., 
628:   Choe, G. S.: \aspj{2003}{584}{L103}
629: 
630: \bibitem[1998]{kucera98} Kucera, T.A., Andretta, V., Poland, A.I.: 
631:  \sph{1998}{183}{107}
632: 
633: \bibitem[1995]{kurok95} Kurokawa, H., Ishiura, K., Kimura, G. \ietal: 
634:   1995, {\it Geomagn. Geoelectr.} {\bf 47}, 1043
635: 
636: \bibitem[2001]{lin01} Lin, J., Forbes, T.G., Isenberg, P.A.: 
637:  \jogr{2001}{106}{25053}
638: 
639: \bibitem[2005]{liu05} Liu, Y., Kurokawa, H., Kitai, R. Ueno, S., 
640:  Su, J.T.: \sph{2005}{228}{149}
641: 
642: \bibitem[1996]{magara06} Magara, T., Shibata, K., Yokoyama, T.: 
643:  \aspj{1997}{487}{437}
644: 
645: \bibitem[2003a]{morim03a} Morimoto, T., Kurokawa, H.:
646:  \pasj{2003a}{55}{503}
647: 
648: \bibitem[2003b]{morim03b} Morimoto, T., Kurokawa, H.: 
649:  \pasj{2003b}{55}{1141}
650: 
651: \bibitem[2007]{nagashima07} Nagashima, K., Isobe, H., Yokoyama, T., 
652:   Ishii, T. T., Okamoto, T.J., Shibata, K.: \aspj{2007}{668}{533}
653: 
654: \bibitem[1994]{nakaj04} Nakajima, H. \ietal: {\it Proc. IEEE}, 1994, 
655:  {\bf 82}, 705
656: 
657: \bibitem[1997]{ohyam97} Ohyama, M., Shibata, K.: 
658:  \pasj{1997}{49}{249}
659: 
660: \bibitem[2004]{okamoto04} Okamoto, T.J., Nakai, H., Keiyama, A., 
661: Narukage, N., UeNo, S., Kitai, R., Kurokawa, H., Shibata, K.:  
662: \aspj{2004}{608}{1124}
663: 
664: \bibitem[2007]{pinter07} Pint\'{e}r, B., Jain, R., Tripathi, D., Isobe, H.  
665: {\it Astrophys. J.} submitted. 
666: 
667: \bibitem[1966]{ramse66} Ramsey, H., Smith, S.F.: 1966, 
668:  {\it Astron. J.}, {\bf 71}, 197
669: 
670: \bibitem[\protect\citeauthoryear{Scherrer et al.}{1995}]{Scherrer95}
671: Scherrer, P.H., \ietal, 1995, {\it Solar Phys.} {\bf 162}, 129.
672: 
673: \bibitem[2006]{shimojo06}Shimojo, M., Yokoyama, T., Asai, A., Nakajima, H., 
674:  Shibasaki, K.: \pasj{2006}{58}{85}
675: 
676: \bibitem[2003]{sterl03} Sterling, A.C., Moore, R.L.: 
677:  \aspj{2003}{599}{1418}
678: 
679: \bibitem[2005]{sterl05} Sterling, A.C., Moore, R.L.: 
680:  \aspj{2005}{630}{1148}
681: 
682: \bibitem[2005]{tripa05} Tripathi, D.: 2005, 
683:  Ph.D. Thesis, University of G\"{o}ttingen, Copernicus, GmBH,
684:  ISBN 3-936586-35-7
685: 
686: \bibitem[2006]{tripa06} Tripathi, D., Del Zanna, G., 
687:  Mason, H.E., Chifor, C.: \aeta{2006}{460}{L53}
688: 
689: \bibitem[2006]{tripa06} Tripathi, D., Isobe, H., Mason, H.E.: 
690: \aeta{2006}{453}{1111}
691: 
692: \bibitem[2007]{vrsnak07} Vr\v{s}nak, B., Veronig, A.M., Thalmann, J.K., 
693:    \v{Z}ic, T.: \aeta{2007}{295}{299}
694: 
695: \bibitem[2001]{zhang01} Zhang, J., Dere, K.P., Howard, R.A., 
696:  Kundu, M.R., White, S.M.: \aspj{2001}{559}{406}
697: 
698: \end{thebibliography}
699: 
700: 
701:  
702: 
703: \begin{figure}    %%%%%%%%%%%%%%%%%% FIGURE 1
704: \begin{center}
705: \psfig{file=f1.eps,width=12.cm}
706: \caption{EIT images of the filament during the oscillation 
707: (left), at the onset of eruption (middle) and after the eruption (right). 
708: The $x$ and $y$ axes are the distances from the disk center. North is up, 
709: and East is to the left. 
710: The solid lines in the left panel with numbers indicate the locations 
711: of the slits for the time slices shown in Figure 2. 
712: }
713: \label{mdi-evol}
714: \end{center}  
715: \end{figure}
716: 
717: \begin{figure}    %%%%%%%%%%%%%%%%%% FIGURE 2
718: \begin{center}
719: \psfig{file=f2.eps,width=12.cm}
720: \caption{Time slices of EIT (a-g) and NoRH (h) images. 
721: The positions of the slits are indicated in Figure 1.
722: The solid and dashed lines in panel (c) indicate 
723: velocities of 5 and 1 km s $^{-1}$, respectively. 
724: The black vertical lines indicate 04:00 UT and 06:00 UT, 
725: as the reference of the first and second peaks of 
726: the oscillation determined by visual inspection of 
727: panel (c). 
728: }
729: \label{overview}
730: \end{center}  
731: \end{figure}
732:  
733: \begin{figure}    %%%%%%%%%%%%%%%%%% FIGURE 3
734: \begin{center}
735: \psfig{file=f3.eps,width=12.cm}
736: \caption{H$\alpha$ center image of the filament 
737: taken by FMT (left) and 17 GHz microwave image 
738: taken by NoRH. The field of view is the same as in Figure 1. 
739: The rectangle shows the field of view of Figure 4. The 
740: white line on the microwave image indicates the slit position 
741: of the time slice shown in panel (h) of Figure 2, which 
742: is the same as that of slit 3. 
743: }
744: 
745: \label{rfmaps}
746: \end{center}  
747: \end{figure}
748: 
749: \begin{figure}    %%%%%%%%%%%%%%%%%% FIGURE 4
750: \begin{center}
751: \psfig{file=f4.eps,width=12.cm}
752: \caption{Time series of H$\alpha$ images at the 
753: line center (bottom row), blue wing at -0.8 \AA \, (middle row), 
754: and red wing at +0.8 \AA \, (top row). 
755: The field of view is shown in Figure 3. 
756: }
757: \label{fmtimages}
758: \end{center}  
759: \end{figure}
760: 
761: \begin{figure}    %%%%%%%%%%%%%%%%%% FIGURE 5
762: \begin{center}
763: \psfig{file=f5.eps,width=12.cm}
764: \caption{Upper panels: EIT images just before the onset of the oscillation. 
765: Overlaid on the left panel is the MDI magnetogram taken at 
766: 03:12 UT. The red and blue contours are +25 and -25 gauss, 
767: respectively. 
768: Lower panels: Temporal evolution of the MDI magnetograms. The field of view is 
769: the same as the box in the upper left panel. The pixels in which $|B| < 25$ G 
770: are set to zero to aid visualization. 
771: }
772: \label{efr}
773: \end{center}  
774: \end{figure}
775: 
776: 
777: 
778: 
779: \end{article} 
780: \end{document} 
781: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
782: 
783: