Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 320, 525-539 (1997)

Previous Section Next Section Title Page Table of Contents

4. Possible problems: temporal variability and abundance variations

Before we start with the interpretation of our fitting results, we briefly discuss two effects that could in principle affect our results.

The first are X-ray flares, which are known to cause temporary variations in the spectral parameters (e.g. Ottmann & Schmitt 1994; Pan & Jordan 1995). Since we excluded all flaring or strongly variable sources from this study, our data should not be affected by, at least strong flares. However, there still might be variability on amplitudes below the noise level of the light curves. We can investigate temporal variability in the spectral parameters of four Pleiades stars and two Hyades stars, for which we have spectra from two different ROSAT observations. For none of these stars (Hz 761, Hz 1032, Hz 2147, Hz 2500, VB 71, and VB 141) significant variations of the spectral parameters can be found. Thus, we conclude that spectral variability is not expected to affect the results of our study.

Another problem might be caused by deviations in the elemental abundances of the coronal plasma from the assumed solar abundances. Recent results, especially from ASCA and EUVE observations, have revealed an under-abundance of metals relative to photospheric values by factors between 2 and 10 in the coronae of some, but not all, late type stars (e.g. Drake et al. 1996a, b; Singh et al. 1996). Furthermore, some late type stars seem to exhibit some kind of FIP-effect 1 (cf. Drake et al. 1996a), while others do not (cf. Drake et al. 1995) and still others show abundance variations with strong deviations from the usual FIP-effect (cf. Drake et al. 1994). On top of that, it should be noted that even the solar FIP-effect is neither fully understood (e.g. Henoux 1995) nor without controversy (cf. Phillips et al. 1995).

Since the abundance variations are far from being well established, the limited spectral resolution of the PSPC actually prevents the derivation of good constraints on elemental abundances, and for nearly all stars in our sample no information about possible abundance variations is available, we have assumed solar abundances for fitting the X-ray spectra. To investigate the possible impact of assuming wrong abundances we have performed spectral fitting simulations (see Appendix). Our simulations show that the assumption of wrong abundances is not expected to cause serious errors in the derived coronal temperatures. Over- or underestimations of the temperatures up to a factor of two might occur, what merely exceeds the uncertainties in our fitted maximum temperatures.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: June 30, 1998