Thermal and nonthermal hard X-ray source sizes in solar flares obtained from RHESSI observations

Abstract

Context. In solar flares, a large amount of thermal and nonthermal energy is released impulsively in the form of heated plasma and accelerated particles. These processes can be studied via hard X-ray (HXR) diagnostics. In addition to spectroscopic observations, a thorough understanding of the thermal and nonthermal particle populations requires the knowledge of the HXR source sizes.Aims. We derive the geometric source parameters of both thermal coronal sources and the nonthermal HXR footpoints in solar flares. We compare and evaluate four different methods for obtaining source sizes, and then derive the most reliable source sizes, as well as the systematic uncertainties.Methods. We obtained time series of HXR images for 24 flares from GOES class C3.4 to X17.2 using the RHESSI instrument. The four imaging techniques employed are CLEAN, Pixon, visibility forward fit, and MEM_NJIT. From this data set, we derived the geometric parameters of the thermal HXR sources and the nonthermal footpoints. Using the different imaging techniques allowed us to quantify systematic measurement uncertainties.Results. We find that the different methods give consistent results on HXR source sizes. The correlations are very good for the thermal sources, and somewhat lower for the footpoints. The MEM_NJIT algorithm gives systematically smaller sizes than the other methods, possibly a result of over-resolution. Thermal source volumes are in the range of 2 × 1025 −1.2 × 1028 cm3 (with a median relative uncertainty of 30%), and nonthermal footpoint areas in the range of 2 × 1016 −6 × 1017 cm2 (median relative uncertainty: 40%). The thermal volumes of our sample are in the same range as those derived for microflares, which would imply that source size is not an important parameter for flare energetics.Conclusions. Using different imaging algorithms for determining HXR source sizes offers the advantage that uncertainties can be better quantified thus making the derived parameters more reliable. Combined with geometric parameters that were derived as time series in a larger number of flares, this will allow the study of the scaling relations and the temporal evolution of thermal and nonthermal HXR sources.