1. Hard X-ray Production in a Failed Filament Eruption David, Alexander, Rui Liu and Holly R., Gilbert 2006 ApJ 653, L719 Related Paper: Ji. H. et al., 2003 ApJ, 595, L135 CSTR Journal Club 2007 Nov.8

2. What is so-called failed eruption? Failed filament eruptions are defined by the dynamical evolution of the filament, which displays an initially eruptive-like acceleration prior to a period in which the filament decelerates. Such events can produce solar flares, but generally are not associated with CMEs. A filament eruption on 2002 May 27 following an M2-class flare.

3. Data SetsRHESSI HXR Imaging algorithm Observations Ji et al. (2003) RHESSI TRACE BBSO H  1.3A CLEAN integration time: 40s Energy band: 12-25 keV A coronal source above the filament at the onset of the activation Alexander et al. (2007) RHESSI TRACE Chromospheric Helium Imaging Photometer (CHIP) He I at the Mauna Loa Solar Observatory (MLSO) Pixon reconstruction algorithm (Metcalf et al. 1996) integration time: 12- 40s Energy band: 12-25 keV A coronal source above the filament at the onset of the activation A second coronal source lying below the erupted filament structure at a later time

4. Fig. 1. Temporal evolution of the filament eruption seen in TRACE 195 Å images on 2002 May 27. The contours show the 12-25 keV hard X-ray emission from RHESSI.

5. Fig. 2 Coronal hard X-ray sources, marked by arrows, in relation to the filament. (a) Source S1 lies above the filament prior to the filament eruption. (b) Source S2, situated at the projected crossing point of the kinking filament legs.

6. Fig. 3 Light curves of RHESSI hard X-ray emission. The dotted and solid lines show the time evolution of the 12–25 and 25–50 keV emission, respectively, from the whole flare (in photons). The diamonds indicate the time evolution of the 12–25 keV coronal source (S2; photons cm-2 s-1). The arrows delimit the time over which the hard X-ray coronal source (S1) was detected. The gaps in the whole flare light curves are due to attenuator switches in RHESSI. S1

7. Fig. 4 Correspondence between MLSO CHIP He 1083 nm line-of-sight velocity data (top row) and the TRACE 195 Å observations (bottom row) of the kinking filament. In the velocity data, black (white) indicates velocity toward (away from) the observer. In the TRACE data black (white) indicates absorbing (emitting) plasma. expansion untwisting S2S1

8. Summary and Discussion S1: lying above the filament, prior to the rapid expansion phase of the filament S2: lying at the crossing point of the kinking filament legs, prior to the untwisting phase of the filament Reconnection above the filament Kink motion Reconnection between two filament legs

9. Left: TRACE images of the failed filament eruption on 2002 May 27. Right: Magnetic field lines outlining the core of the kink-unstable flux rope at t=0, 24,and 37. A favorable model: The kink-unstable flux rope model of Torok & Kliem (2005) successfully reproduces both the observed development of the kink structure and the time profiles of the height and velocity of the filament. Comparison of height and velocity of the flux rope in the simulation (solid lines) with the corresponding values of the filament (diamond, data from Ji et al. 2003) Torok & Kliem (2005)