Astron. Astrophys. 356, 1067-1075 (2000)

5. Conclusions

In this paper we have performed a detailed comparison of H line center imaging observations with microwave and HXR/GR spectral measurements, obtained with 1 s time resolution during a large X1.3/2B flare. This data have been used to infer the magnetic field structure which connects the hot flare corona to the chromosphere and to discriminate between alternate forms of energy transport within this structure. For the flare under study, the main results of our analysis can be summarized as follows:

• Energy transport takes place in four loop systems with sizes ranging from approximately 10⁴ to 2.5 10⁴ km. These loop systems are not static but expand in the course of the flare as the two flare ribbons move apart from each other.

• The four H kernels, which materialize the foot points of the four loop systems, show a similar time evolution. During the two successive impulsive microwave and ( 73 keV) HXR bursts (referred to as B1 and B2 in Sect. 3), this time evolution is well fitted by a linear combination of the time profile of the HXR count rate and its time integral (see Eq. 1). Such a close relationship between the H and the HXR emission strongly suggest that non-thermal electrons are the dominant energy transport process within the different loop systems during both B1 and B2. An alternative energy transport may be possible by low energy ( 500 keV) protons provided that they are accelerated in close synchronism with 20 keV electrons. In any case, our results indicate that slower transport processes, such as by conduction fronts, which have been found to be effective at some sites in less energetic flares (see Sect. 1), do not seem to play a major role, if any, in this large GR flare. This suggests that during large GR flares most of the primary energy release goes into particle acceleration while in less energetic flares the relative amounts of energy which go into direct heating and acceleration may vary from one loop system to the other (e.g. Benz et al. 1994and references therein).

• According to Eq. 1 the temporal response of a given H kernel to accelerated particles consists of a fast and a slowly varying component. The latter, which evolves like the time integral of the HXR count rate, i.e. like the soft X-ray emission of the loop systems associated to that kernel, shows approximately the same amplitude for the four kernels during both B1 and B2. Although the present data set did not allow us to fully understand the physical origin of this slow response (continuous heat flux from the hot corona or chromospheric evaporation) its intensity appears, surprisingly, to depend only weakly on the loop model parameters which are probably different from one loop system to the other. In contrast, the fast H response, which matches the HXR count rate, is similar for the four kernels during B1, but varies from one kernel to the other during B2 where the largest responses are found for the largest loop systems. This indicates that the energy transport is predominant in the largest loops, i.e. when the number of accelerated particles is largest. Furthermore during B2, our results suggest that the relative amounts of accelerated particles injected into the four loop systems vary from the first to the second injection, i.e. when the number of accelerated electrons starts to increase dramatically (see Sect. 4.3). This is consistent whith former multi-wavelength analysis of GR flares which indicate that changes in the characteristics of the accelerated particles are associated to changes of the magnetic pattern traced by these particles (e.g. Chupp et al 1993; Trottet et al. 1994; Trottet et al. 1998).

• The present data provide for the first time evidence for nearly simultaneous H and HXR pulses with typical rise times in the range of 0.4 to 1.5 s. This is broadly consistent with the essential features of models simulating the H response to non-thermal electrons which heat a loop atmosphere (e.g. Canfield & Gayley 1987; Heinzel 1991). However, it is premature to compare our results with these models which apply to a much more simple field geometry, idealized electron injection functions and to much less energetic electron populations than those involved in the present flare.

© European Southern Observatory (ESO) 2000

Online publication: April 17, 2000
helpdesk.link@springer.de