2008A&A...488.1069A

The solar corona is a template to understand stellar activity. The Sun is a moderately active star, and its corona differs from that of active stars: for instance, active stellar coronae have a double-peaked emission measure distribution EM(T) with a hot peak at 8-20MK, while the non-flaring solar corona has one peak at 1-2MK and, typically, much cooler plasma. We study the average contribution of flares to the solar emission measure distribution to investigate indirectly the hypothesis that the hot peak in the EM(T) of active stellar coronae is due to a large number of unresolved solar-like flares, and to infer properties about the flare distribution from nano- to macro-flares. We measure the disk-integrated time-averaged emission measure, EM_F(T), of an unbiased sample of solar flares, analyzing uninterrupted GOES/XRS light curves over time intervals of one month. We obtain the EM_Q(T) of quiescent corona for the same time intervals from Yohkoh/SXT data. To investigate how EM_F(T) and EM_Q(T) vary during the solar cycle, we evaluate them at different phases of the cycle between December 1991 and April 1998. Irrespective of the solar cycle phase, EM_F(T) appears as a peak in the distribution, and it is significantly larger than the values of EM_Q(T) for T∼5-10MK. As a result, the time-averaged EM(T) of the entire solar corona is double-peaked, with the hot peak, due to time-averaged flares, being located at temperatures similar to those of active stars, but less enhanced. The EM_F(T) shape supports the hypothesis that the hot EM(T) peak of active coronae is due to unresolved solar-like flares. If this is the case, quiescent and flare components should follow different scaling laws for increasing stellar activity. In the assumption that the heating of the corona is entirely due to flares, from nano- to macro-flares, then either the flare distribution or the confined plasma response to flares, or both, are bimodal.