Observations of conduction driven evaporation in the early rise phase of solar flares

Context. The classical flare picture features a beam of electrons, which were accelerated in a site in the corona, hitting the chromosphere. The electrons are stopped in the dense chromospheric plasma, emitting bremsstrahlung in hard X-rays. The ambient material is heated by the deposited energy and expands into the magnetic flare loops, a process termed chromospheric evaporation. In this view hard X-ray emission from the chromosphere is succeeded by soft-X-ray emission from the hot plasma in the flare loop, the soft X-ray emission being a direct consequence of the impact of the non-thermal particle beam. However, observations of events exist in which a pronounced increase in soft X-ray emission is observed minutes before the onset of the hard X-ray emission. Such pre-flare emission clearly contradicts the classical flare picture. Aims. For the first time, the pre-flare phase of such solar flares is studied in detail. The aim is to understand the early rise phase of these events. We want to explain the time evolution of the observed emission by means of alternative energy transport mechanisms such as heat conduction. Methods. RHESSI events displaying pronounced pre-flare emission were analyzed in imaging and spectroscopy. The time evolution of images and full sun spectra was investigated and compared to the theoretical expectations from conduction driven chromospheric evaporation. Results. The pre-flare phase is characterized by purely thermal emission from a coronal source with increasing emission measure and density. After this earliest phase, a small non-thermal tail to higher energies appears in the spectra, becoming more and more pronounced. However, images still only display one X-ray source, implying that this non-thermal emission is coronal. The increase of emission measure and density indicates that material is added to the coronal region. The most plausible origin is evaporated material from the chromosphere. Energy provided by a heat flux is capable of driving chromospheric evaporation. We show that the often used classical Spitzer treatment of the conductive flux is not applicable. The conductive flux is saturated. During the preflare-phase, the temperature of the coronal source remains constant or increases. Continuous heating in the corona is necessary to explain this observation. Conclusions. The observations of the pre-flare phase of four solar flares are consistent with chromospheric evaporation driven by a saturated heat flux. Additionally, continuous heating in the corona is necessary to sustain the observed temperature.