Reports of rotational or tornado-like behaviour of solar prominences and/or their associated cavities have been made for almost a century. These include ground-based observations of rotational eddies within prominences [5,3], spectroscopic evidence of filament rotation , tornado-like jets within prominences, and rotation of prominence cavities . The Solar Dynamics Observatory (SDO) provides the necessary temporal and spatial resolution, and long-term systematic full-disk observations to view and interpret such events in detail.
On 25th September 2011 a small and unremarkable crown polar prominence rotated into view on the south-west limb. As is typical, this prominence was cradled within a dark extended cavity. At 02:30 on the 25th, the whole structure experienced a large-scale wobble, and small bright blobs appear in the cavity above the prominence. Dark fibrils at the base of the prominence became very active, and by 06:00 the filament and cavity developed a distinct tornado-like appearance, with a large circular structure atop a narrower pillar.
From 10:00, there was a large new injection of material into the filament base and for the next few hours spectacular dynamics occurred within the prominence and cavity (See Fig. 1). Streaks and blobs of varying brightness followed circular paths counter-clockwise around the top of the filament pillar inside what was formerly a dark cavity. The blobs of material flowing into this area traced out magnetic structures that were previously invisible.
At first material moved along a thin channel, but by 10:10 the thin channel had already widened and a helix-like structure with at least seven turns was very obvious. The injection of such tightly wound helices was repeated at about 11:00, and of less tightly wound helices at 11:45. The very core of the tornado head was bright and complex, with strange slow rotation and movements of filamentary structure. By 18:00 the head of the tornado dimmed, the rotational movement stopped and by 00:00 on 26th September the tornado disappeared, leaving wispy strands extending at obtuse angles relative to the radial into the region previously occupied by the filament pillar. The main period of coherent rotation lasted for approximately 3 hours.
Properties of the tornado plasma
Local Correlation Tracking (LCT) and manual methods reveals speeds of as high as 55km/s to 95km/s within the tornado – smaller than the sound speed. The flow gains speed substantially, even when it ascends against gravity, suggesting that magnetic tension forces play an important role in accelerating the flow. From viewing the tornado in several channels of the AIA instrument, and its appearance in ground-based H-alpha observations, the material flowing within the structure contains ions at a large range of formation temperatures. Throughout the whole period of tornado formation and rotation, the emission in the 304A channel is almost identical to that of the hotter 171A channel.
Discussion and Conclusions
The complex appearance of the coherent rotation can be interpreted as injections of helical magnetic fluxes into the filament and cavity, and subsequent sporadic injections of material along the helical structures. When an injected helical flux tube is tightly wound, it may be unable to maintain stability and is eventually injected or simply untwists into the surrounding cavity. The filament and cavity are extended helical flux ropes along the line of sight, and as material moves along the helical structures, the observed motion is that of rotation. The innermost prominence looks narrow and tangled whilst the surrounding cavity is more loosely wound. This supports the view that tangled flux tubes and flow give favourable conditions for the mass supply and maintenance of a prominence. The line-of-sight integration of emission from such tangled helices gives a very complicated appearance in observations.
Observations of the large circular motions and flows originating from a narrow channel may shed some light to the question why a cavity exists above a prominence. If most of the magnetic field flux in the cavity is rooted in a small region in the lower atmosphere, then the supply of plasma may simply be insufficient to fill the cavity unless there a dramatic injection of material caused by some catastrophic event at the prominence base. The general (quiescent) case would therefore be of a dark cavity devoid of plasma due to the restrictive geometry of the flux tube at low heights.
The dynamics and shape of this prominence and cavity are significantly larger, more complex and coherent than several other eruptive features reported: erupting prominences , rotational spicules , helical `EUV sprays’ , and emerging helical prominences . This prominence contains plasma at both cool (104 K) and hot coronal temperatures suggesting that the flows are driven by the same mechanism as in emerging prominences , which are quite cool.
Questions for the future
Interesting questions which arise from this event are:
- What mechanism drives the injection of magnetic helices and/or material into the prominence and cavity?
- Are the brightness enhancements which we see flowing within the confines of the helical structures blobs of material or density waves?
- What exactly is the structural relationship between a prominence and associated cavity? Do current models account for all cases?
- Why does this structure not erupt?
- Will the observations of SDO show such tornadoes to be common events? We have found several tornadoes during the past few months, but none as coherent and spectacular as the 25th September event.
-  Harrison, R. A., Bryans, P., & Bingham, R. 2001, A&A, 379, 324
-  Kurokawa, H., Hanaoka, Y., Shibata, K., & Uchida, Y. 1987, Sol. Phys., 108, 251
-  Liggett, M., & Zirin, H. 1984, Sol. Phys., 91, 259
-  Öhman, Y. 1969, Sol. Phys., 9, 427
-  Okamoto, T. J., Tsuneta, S., & Berger, T. E. 2010, ApJ, 719, 583
-  Pettit, E. 1925, Publications of the Yerkes Observatory, 3, 4
-  Pike, C. D., & Mason, H. E. 1998, Sol. Phys., 182, 333
-  Wang, Y., et al. 2010, ApJ, 717, 973