• PHYSICAL PROBLEM: Can we explain the solar flares by invoking shear flows at the solar surface? Do flares occur preferentially in the  neighbourhood of neutral points of the magnetic field ("breakout" scenario)?

 
• OBSERVATIONS: solar flares (mov7,5 Mo, TRACE mission, extreme UV, 171 Angströms)

 
• EQUATIONS: Magnetohydrodynamic

 
• NUMERICAL EXPERIMENT: reconstruction of the local 3-dimensional magnetic field above the solar surface just before a flare, (time-independant, force-free, current-free approximation), using observations of the photospheric field made at Kitt Peak

 
• CONCLUSION: (conjectures!) a) to generate a flare, the shear flow needs not to be just below the neutral point of the magnetic structure b) the break of the magnetic structure (by reconnection of lines at the neutral point) apparently occurs about 3 minutes before the peak of the flare, and can thus trigger the opening of the field lines necessary for the mass ejection. The flare would be a consequence of the opening of the field lines.

 
• REFERENCE: Aulanier, DeLuca, Antiochos, McMullen, Golub, The topology and evolution of the Bastille day flare, Astrophysical Journal, 540, 1126, 2000.

1. The breakout model: an interrupted numerical experiment
The breakout scenario proposes that flares with mass ejection occur by first opening a (special) magnetic structure at the neutral point, and then eject the plasma (Antiochos, Devore, Klimchuk, Astrophys.J. 510, 485). This is illustrated by a very short FILM (mov, 0.7 Mo; gif, 1,1 Mo) showing a time-dependent solution of the MHD equations.
The initial magnetic structure shown has a weak (neutral) point at the top of it. This weak point is supposed, via a shear at the basis (photosphere) to lead to the opening of the structure, and then to the flare.
The film stops, just before the magnetic field lines completely open, because, due to strong gradients, the plasma density becomes too low. This stops the integration, so the question remains open...

2. Analysis of a particular case: reconstructing the magnetic structure of the Bastille Day flare (1998)
The author has looked in detail at a given flare, that of 14 July 1998. He has "reconstructed" the three-dimensional (volume) structure just before the flare, extrapolating the observed surface field. He has shown (see film below) that the structure indeed contains a neutral, weak point, and observed a partial reconnection of the field lines there 3 minutes before the flare, as proposed by the breakout scenario. He also has found that the sheared region is not exactly located below the neutral point, but shifted aside. The FILM (mov, 10 Mb; gif, 18 Mb) has three successive parts:
a) a view of the sun followed by a close view of the flare region (the viewpoint changes but the scene is static)
b) the superposed reconstruction of the magnetic field and of the detailed flare region (still static)
c) a time-dependent sequence of pictures showing the real evolution of the flare as observed by TRACE

3. Continued...
The author is presently working on the true time-dependent integration of the MHD equation (first paragraph), but taking into account the shift between the shear and the neutral point.