2019ApJ...880..105A

We search for outliers in extreme events of statistical size distributions of astrophysical data sets, motivated by the Dragon-King hypothesis of Sornette, which suggests that the most extreme events in a statistical distribution may belong to a different population, and thus may be generated by a different physical mechanism, in contrast to the strict power-law behavior of self-organized criticality models. Identifying such disparate outliers is important for space weather predictions. Possible physical mechanisms to produce such outliers could be generated by sympathetic flaring. However, we find that Dragon-King events are not common in solar and stellar flares, identified in 4 out of 25 solar and stellar flare data sets only. Consequently, small, large, and extreme flares are essentially scale-free and can be modeled with a single physical mechanism. In very large data sets (N >= 10⁴) we find significant deviations from ideal power laws in almost all data sets. Nevertheless, the fitted power-law slopes constrain physical scaling laws in terms of flare areas and volumes, which have the highest nonlinearity in their scaling laws.