We analyze hard X-ray imaging observations of three flares, showing widely different characteristics, in order to try and discriminate the relative efficiency of heating and acceleration in the primary energy release. Using a simplified approach, we compute the hard X-ray distribution and energy deposition due to accelerated electrons, with beam and ambient plasma parameters appropriate for each of the observed events. The results are convolved with the Hard X-Ray Imaging Spectrometer (HXIS) instrumental response and compared with observations. We find that: (a) Many observations are compatible with thick target processes, and with the possibility that flares may have high (>20%) acceleration efficiency. (b) Single hard X-ray sources should be very common in the data available at present (HXIS and HINOTORI), as it is the case, as well as a transition from chromospheric footpoints to single source structures. The latter cannot then be directly interpreted as thermal sources. (c) In the particular case of a limb flare, associated with a rather weak high energy burst, we show that the spatial and spectral behavior of the hard X-ray emission is incompatible with pure nonthermal processes. We thus propose that the observed emission was principally due to the strong heating intrinsic to a reconnection process within the region of interaction between two magnetic structures which are seen in the soft X-ray data. (d) We also study the heating effect of a beam, due to Coulomb losses, during its passage through the flare loops. In some cases, rather large and localized temperature increases can be expected to appear within short timescales (≈ 1 s), leading to a combination of nonthermal plus thermal output in the hard X-ray spectrum, which renders virtually impossible the determination of the underlying beam parameters. We finally discuss the extent to which our conclusions are valid, considering the instrumental limitations as well as the simple physical treatment that we apply.
Read full abstract