Astron. Astrophys. 328, 565-570 (1997)

4. Discussion and summary

With a total of erg this X-ray flare on HU Virginis released within 1.5 days % of the energy of the exceptional 9 day long flare on CF Tuc (1.4 erg, Kürster & Schmitt 1996). Previous observations of X-ray flares on RS CVn-type stars already included events of even greater magnitude: Pye & McHardy (1983) discovered a strong flare on the (presumable) RS CVn-binary 3A1431-409, that released a total of 3 erg in the significantly wider (2 - 10 keV) bandpass of the ARIEL-V SSI. Another very energetic flare was seen by Charles et al. (1979), who estimated a total energy release of 4 10³⁷ erg in the 0.2 - 2.8 keV for the flare on DM UMa.

In contrast to previous ROSAT observations of long duration flares, our well covered onset and peak of the flare on HU Virginis allowed a more detailed comparison with flare models. In the case of the X-ray flare on EV Lac the onset of the flare was not covered at all, hence no decision in favour of a particular two-ribbon model could be made (Schmitt 1994). The highly structured light curve of the very long flare on CF Tuc, perhaps including multiple flaring events and geometrical effects (CF Tuc is a partially eclipsing system), also did not enable a specific solution for both the quasi-static cooling loop- or the two-ribbon flare model (Kürster & Schmitt 1996). Table 2 summarizes the results for the loop sizes we obtained by using these two models. In the case of the two-ribbon model the given loop size corresponds to the maximum height which the flaring loops reach in the corona. Note that both models, although they describe the flare event completely differently, yield similar and very large loop sizes. Despite that we are able to favour a specific solution within the framework of each flare model, the current data do not allow a firm decision between these two best-fit models.

Table 2. Range of loop heights derived by flare modelling

It is not surprising that we cannot detect any rotational modulation during the flare light curve since the flare duration of 1.5 days is small compared to the rotational period of 10.4 days and furthermore we seem to deal with a spatially extended flare, perhaps even a flare occuring in the polar region of HU Virginis where previous Doppler images showed a major polar spot (Strassmeier 1994).

Our main results can be summarized as follows:

1. We detected a long duration flare on HU Virginis lasting more than 1.5 days and a total energy output of erg in the 0.1 - 2.4 keV bandpass.
2. The well covered flare onset and maximum phase allowed detailed comparison with two solar flare models and provided an estimated size of the flaring region. In both cases the resulting size of the flaring region is very extended with the (maximum) height of the coronal loop being comparable to the stellar radius.
3. A continuous 8-day observation revealed variability of the X-ray flux on time scales shorter than the rotational/orbital period. The mean emission ("quiescent") level of the corona of HU Virginis did not change over a year. Within the precision of our data we found no evidence for very short-term variability which could be interpreted as microflaring.

© European Southern Observatory (ESO) 1997

Online publication: March 26, 1998

helpdesk.link@springer.de