A comparison of the thick-target model with stereo data on the height structure of solar hard x-ray bursts

Abstract

The height structure of a thick-target solar hard X-ray source is predicted for a beam injected vertically downward with a power-law spectrum and dominated by Coulomb collisional energy losses. This structure is characterised by the ratio of hard X-ray flux from an upper part of the source to that from the entire source, and is essentially a function only of the atmospheric column density ΔN (cm−2) in the upper region. These predictions are compared with the flux ratios at 150 keV and 350 keV observed by two spacecraft for five events in which the solar limb occults part of the source for one spacecraft. In three events the occulting levels h ranged from 0 to 2500 km. For these the theoretical and observed ratios are found to be comparable for values of ΔN in reasonable accord with those found at these altitudes by optical and UV spectroscopic modelling of flare chromospheres. In one event the occultation ratio was observed to rise after the burst peak and it is found that this rise is consistent with an increase in ΔN due to conductively driven chromospheric evaporation. However the energy dependence of the occultation ratio is not consistent with that predicted by the model and it is concluded that non-collisional losses must be significant in beam dynamics. In the other two events, the occultation level h was ≳ 25 000 km. For these the energy dependence of the occultation ratio is comparable with the model predictions. However the values of ΔN required demand extremely high coronal densities and/or acceleration altitudes. Furthermore, the one observed evolution of the occultation ratio is entirely inconsistent with the model. It is concluded that in these, bremsstrahlung emissions other than that from a beam must be important.