In this paper the HXT observations of 14 behind-the-limb flares with occulted footpoints has been analyzed. The basic conclusions can be summarized as follow.
Obscuring of strong footpoint HXR emission sources allows to investigate the loop-top sources in more detail. Reconstructed images provide more reliable photometry of such sources.
All investigated flares showed relatively soft, quasi-thermal HXR radiation which was concentrated in the low energy channels (below 33 keV). The HXR light curves had generally smooth shape (the smooth component) with some randomly occurring impulsive spikes (the impulsive component) which were more energetic. Both components were emitted from the same emission source. For half of the events the smooth component consisted of two or three parts and similar number of emission sources was observed.
Two different types of the loop-top HXR emission sources have been found. Type A sources, detected in each event, are co-spatial with the SXR bright loop-top kernels and have quasi-circular shape. Type B sources have strongly elongated shape, are co-spatial with very weak SXR structures, and occur in complex active regions. No other characteristics, except morphological, differentiate between the sources of types A and B.
The turbulent flare kernel model explains main observational characteristics of the loop-top HXR emission sources of type A: random motions seen as the non-thermal line broadening of SXR lines, confinement of electrons which generate hard X-rays, the smooth and the impulsive component of the HXR light curves. The sources of type B have probably a different origin. Any explanation of type B sources should account for the following features: an additional acceleration mechanism of electrons, the magnetic mirrors in the upper chromosphere, the prolonged evolution of flares showing this kind of sources.
The proposed classification of the loop-top HXR emission sources seems to be more physical than that of Masuda who actually resolved two phases of evolution of the same loop-top source. Moreover, their time variability and energy spectra - the additional criteria of the classification - were probably deformed by strong footpoint sources.
Further studies of the loop-top HXR emission sources are urgently needed, including a more complete survey of the behind-the-limb flares with occulted footpoints, improvement of HXR image reconstruction techniques as well as new sources of data e.g. HESSI.
Acknowledgements
The Yohkoh satellite is a project of the Institute of Space and Astronautical Science of Japan. I thank Professor J. Jakimiec for many useful comments and discussions. I thank also Dr. A. Fludra for comments on the manuscript. This work was supported by the KBN grant No. 2 P03D 016 14.
Copyright ESO 2001