The Sun 89
a) b) c)d) e) f)FIGURE 17 Left: A version of the standard 2D X-type reconnection model for two-ribbon flares,
pioneered by Carmichael, Sturrock, Hirayama, and Kopp-Pneumann (CSHKP), which also
includes the slow and fast shocks in the outflow region, the upward-ejected plasmoid, and the
locations of the soft X-ray bright flare loops. (Courtesy of Saku Tsuneta.) Right: 3D version of the
two-ribbon flare model, based on the observed evolution during the Bastille Day (July 14, 2000)
flare: (a) low-lying, highly sheared loops above the neutral line first become unstable; (b) after
loss of magnetic equilibrium the filament jumps upward and forms a current sheet according to
the model by Forbes and Priest. When the current sheet becomes stretched, magnetic islands
form and coalescence of islands occurs at locations of enhanced resistivity, initiating particle
acceleration and plasma heating; (c) the lowest lying loops relax after reconnection and become
filled due to chromospheric evaporation (loops with thick linestyle); (d) reconnection proceeds
upward and involves higher lying, less sheared loops; (e) the arcade gradually fills up with filled
loops; (f) the last reconnecting loops have no shear and are oriented perpendicular to the neutral
line. At some point, the filament disconnects completely from the flare arcade and escapes into
interplanetary space.
reconnecting field lines (Fig. 18, top), while EUV images
invariably display the relaxed postreconnection field lines
after the flare loops cooled down to EUV temperatures in
the postflare phase (Fig. 18, middle and bottom).
6.4 Flare Plasma Dynamics
The flare plasma dynamics and associated thermal evo-
lution during a flare consists of a number of sequential
processes: plasma heating in coronal reconnection sites,
chromospheric flare plasma heating (either by precipitating
nonthermal particles and/or downward propagating heat
conduction fronts), chromospheric evaporation in the form
of upflowing heated plasma, and cooling of postflare loops.
The initial heating of the coronal plasma requires anoma-
lous resistivity because Joule heating with classical resis-
tivity is unable to explain the observed densities, tempera-
tures, and rapid timescales in flare plasmas. Other forms of
coronal flare plasma heating, such as slow shocks, electron
beams, proton beams, or inductive currents, are difficult
to constrain with currently available observables. The sec-
ond stage of chromospheric heating is more thoroughly ex-
plored, based on the theory of the thick-target model, with
numeric hydrodynamic simulations, and with particle-in-
cell simulations. Important diagnostics on chromospheric
heating are also available from Hα, white light, and UV
emission, but quantitative modeling is still quite difficult