The old debate about how many loops are involved in a flare appears to have no end. The main reason is because there are roughly as many observed flaring configurations in soft X-rays having one loop as having a multiple loop structure. After the first results from the Soft X-ray Telescope (SXT, Tsuneta et al. 1991) the configuration with one single loop and two footpoints in soft X-rays became more popular (Dennis et al. 1994), but at the same time the evidence for flares with interacting multiple loops (Hanaoka 1996; Yoshida & Tsuneta 1996; Liu et al. 1998) and with multiple sources in hard X-rays (Sakao et al. 1994) has increased as well.
Khodachenko and Zaitsev (1998) modeled energetic processes in solar ARs. They concluded that energetic processes in individual magnetic loops lead to a flare-like heating of nearby loops. That means, a multiple loop structure is considered. Hori et al. (1998) used hydrodynamic simulations to reproduce spectral profiles observed in the early phase of solar flares. They found that only multiple-loop systems are compatible with their models. In Démoulin et al. (1997) the authors show that in all the flares they have studied more than one single loop is involved. Ozaki et al. (1997) studied by means of a three-dimensional magneto-hydrodynamic simulation the interaction of magnetic loops and arcades. They found that a single loop or arcade does not exhibit any significant energy release, even though it is highly twisted and expanded. These authors also concluded that a flare-like energy release can occur only when two flux tubes collide and reconnect with each other.
The idea of energy release in flares driven by reconnection in separatrices is intimately related with a multiple-loop structure. In separatrices, the field-line linkage is discontinuous. There, field lines can reconnect and the magnetic field can relax to a new, less energetic configuration. The generalization of the concept of separatrices to configurations without field-line linkage discontinuities was put forward by Priest & Démoulin (1995). They proposed that magnetic reconnection may occur in 3D in the absence of null points at "quasi-separatrix layers" (QSLs), which are regions where there is a steep gradient or a drastic change in the mapping of field lines from one boundary to another of a given magnetic volume. They proved that in a special configuration (a sheared X-type magnetic field), nearly any smooth and weak flow imposed on the boundary produces strong flows at the QSLs. Démoulin et al. (1996) have determined the location of QSLs and studied their properties in simple theoretical flaring configurations. To do that, they have developed a numerical algorithm called QSLM, for Quasi-Separatrix Layers Method. Their results have been extended successfully to observed configurations (e.g. Démoulin et al. 1997; Mandrini et al. 1997; Schmieder et al. 1997; Gaizauskas et al. 1998).
Large-scale structures in the solar corona have been observed since 1970, starting with the Skylab X-ray instrument, only in soft X-rays. These early observations had a poor temporal resolution (see e.g. Chase et al. 1976; Fárník & Svestka 1986), but nevertheless, more than a hundred interconnecting loops were observed. HXIS (on board the Solar Maximum Mission), with better timing but poorer spatial resolution, has also observed interacting ARs connected through a large scale arc (De Jager & Svestka 1985; Poletto et al. 1993). Howard & Svestka (1977) and Svestka & Howard (1979, 1981), studying Skylab data, concluded that brightness variations of the loops are always caused by magnetic field changes near their footpoints. These changes can also produce flares in the connected regions, but the flare occurrence and the loop brightenings are independent consequences of one common trigger action (e.g. the emergence of a new bipole). SXT, on board the Yohkoh satellite, has better temporal and spatial resolutions. At present, over 100 interconnecting loops measured with SXT have been analyzed. The first results (Manoharan et al. 1996; Tsuneta 1996; Pevtsov et al. 1996; Fárník et al. 1999) suggest strongly that magnetic reconnection is the trigger for the development and/or the brightening of the interconnecting loops. Several studies proposed also that different loops can interact in such a way that an instability in one loop system can be transmitted to another loop system rooted nearby (Rompolt & Svestka 1996). The example described here suggests that both possibilities are present at different times.
We study the magnetic evolution of AR 7031 throughout a week, and compare this evolution with the location of flare emission in H and soft X-rays. As we have observed the brightening of a loop connecting AR 7031 to AR 7038, we have analyzed this arc and we have generated a magnetogram combining the two ARs. The magnetic field evolution is described in Sect. 2. The flare data are analyzed in Sect. 3. After this, we show the results of the magnetic field model and of the computed topology for the studied regions, and compare the location of QSLs to observed features (Sect. 4). We conclude, first that energy is released by magnetic reconnection, and second that the magnetic evolution (e.g. emergence of flux, displacement of the photospheric polarities and changes in the configuration due to nearby magnetic reconnection) is what determines the location of the energy release sites at QSLs for the flares and the interconnecting arc (Sect. 5).
© European Southern Observatory (ESO) 2000
Online publication: December 11, 2000