Extreme Ultraviolet Explorer Photometry of RS Canum Venaticorum Systems: Four Flaring Megaseconds

Abstract

We present an analysis of 12.2 Ms of Extreme Ultraviolet Explorer (EUV E) photometry of 16 RS CVn systems (including four Ñaring megaseconds). Our study attempts a systematic categorization of stellar coronal emission in RS CVn binary systems. The temporal resolution of the EUV E satellite and the opportunity for long observations enables us to study separately the quiescent and Ñaring states of RS CVn systems. Thirty of 31 observations are statistically variable. We examine light curves, characterizing both phase-dependent variations and large-and small-scale Ñaring. There is evidence for small-scale sto-chastic variability on the one short-period day) system in our sample (ER Vul), which we inter-(P orb \ 1 pret as small-scale Ñaring. CF Tuc is the only system for which phase-dependent variations over multiple orbital periods are seen. We analyze 30 Ñares on nine systems, Ðtting rise and decay times. Several Ñares have durations and thus cannot originate from compact emission regions close to the stellar [0.5P orb surfaces. Many Ñares have unusual Ñare morphologies, with rise times comparable to or greater than the decay time, and emission plateaus. In addition, eight Ñares have decay phases that are Ðt better by a broken power law than a single power law. The decay times for the second power law are statistically di †erent than the single decay times and are correlated with the Ñare rise times. This observed connection points to the potential importance of the change in decay rate as a means of examining the physical processes operating during the Ñare. The distribution of orbital phases of Ñare onsets on V711 Tau (HR 1099), the best-studied RS CVn in our sample, is consistent with a Poisson distribution that is random in time. In two systems that are partially eclipsing (CF Tuc and ER Vul), we Ðnd no evidence for an eclipse in the light curve, indicating that the emission region is large compared to the stellar radius. We param-eterize the distribution of Ñare energies as a power law, Ðnding a cut-o † energy of B1033 ergs and a slope of [0.6. Flare energy increases with Ñare duration as E P *t1.42, conÐrming the long-duration nature of high-energy Ñares on RS CVn systems. The integrated Ñare luminosity depends on the quies-cent luminosity as Flare frequencies range from 0.1 day~1 to 1.5 day~1. For obser-L flare P L quiescent 1.05. vations not a †ected by the dead spot on the Deep Survey detector, 40% of the observed time was spent in a Ñaring state, indicating that Ñaring is a "" normal ÏÏ state for the coronae of these systems. We note the presence of a quasi-periodicity in EUV E photometric data on the order of 1 day, which is not corrected for in the standard reduction and analysis software.