Astron. Astrophys. 326, 647-654 (1997)

6. Discussion

We have shown, by analysis of the low-resolution spectra of a number of low-mass stars spanning a wide range of spectral types as well as of lithium abundances, and by direct comparison of their low- and high-resolution spectra, that usage of low-resolution spectra alone is likely to lead, at least in G and K stars, to high inferred lithium abundances in late-type stars. An absorption feature at 6708 Å appears to be commonly present, in the low-resolution spectra of G and K stars, independently from their actual lithium abundance. Such feature appears to be present, for stars later than M0, only in true high-lithium sources. Thus, classification of stellar counterparts to soft X-ray sources done exclusively on the basis of low-resolution optical spectroscopy is likely to significantly over-estimate the number of PMS stars present in the source population. We have also shown that, even with fully reliable Li I equivalent widths, the adoption of a single equivalent width threshold will lead to over-estimating the number of WTTS sources present in the sample.

As discussed in Sect.  1, several works have recently appeared in the literature which present the identification of stellar counterparts to soft X-ray sources based on low-resolution spectra alone. These works discuss X-ray sources in the general direction of star forming regions, but usually cover large region of the sky, extending to quite large projected distances from the SFR. A common feature to all these works is that they seem to find, in addition to the expected concentration of PMS stars in and around the SFR, a large number of widespread WTTS with no apparent immediate relationship with the SFR under investigation, which, as discussed in Sect.  1, are a challenge to current ideas of low-mass star formation. At the same time, the same samples lack the large number of young main-sequence coronal sources which are known to be present in X-ray selected samples at these flux levels.

We make the hypothesis that a non-negligible fraction of the "field WTTS" discussed in the RASS-WTTS papers are normal, active young low-mass stars, on, or very near to the main sequence. The arbitrary placement of foreground active stars at the distance of the putative parent SFR (which is common practice in the WTTS-RASS papers) will make them appear brighter then they actually are, and thus make them wrongly appear as still in a PMS contraction phase when placed on evolutionary tracks. The apparent large number of dispersed WTTS are thus most likely not the solution to the still standing puzzle of the apparent lack of the deficiency of stars older than Myr in most known star-forming regions (the so-called missing post-T Tauri problem), as discussed by Feigelson (1996). Palla & Galli (1997) have recently argued that the post-T Tauri problem is a false one, as it is based on the assumption of a constant star-formation rate in giant molecular clouds, an assumption which, based on the similarity between the molecular cloud lifetime and the ambipolar diffusion time, they show to be unlikely. Rather, they argue, star formation accelerates sharply toward the end of a cloud's lifetime, thus justifying the lack of large numbers of older PMS in SFRs.

Micela et al. (1997) have recently used the Hipparcos parallaxes of the subsample of EMSS and ESS stars which have been observed by Hipparcos to accurately position these stars in an HR diagram, showing that only one of the stars in the sample is far away from the main sequence and clearly still in a contracting phase. The rest of the population is mostly composed of main-sequence objects, with % giants. While this subsample suffers from a bias toward brighter stars, and it is thus lacking many of the fainter and more active stars (some of which are known to be PMS stars from their lithium abundance), they show that all of the stars in their sample (seven) which would have been classified as WTTS using the Wichmann et al. (1996) 100 mÅ criterium (even using high-resolution spectra, and thus reliable Li I doublet equivalent widths) are very close to or on the main-sequence. Thus, while there certainly are a number of bona fide WTT stars in the sample of active stars selected from the Einstein surveys (as for example the low-gravity, very high lithium abundance stars of Morale et al. 1996), the majority of the low-mass stars appear however to be already on or very close to the main sequence stage. Given the similar limiting sensitivity of the RASS and of the EMSS, the detected source population has to be similar, and therefore, again, a large fraction of the RASS stellar sources are expected not to be in the PMS stage but rather young main-sequence stars.

Obviously, a (perhaps considerable) fraction of the stars identified in the RASS identification programs discussed above will be true WTTS, still contracting toward the main sequence, given also the vicinity of the surveyed areas to SFR's. However, lacking accurate, high-resolution based lithium abundances and distance measurements, they cannot be separated from the normal active main-sequence stars present in the sample. Given the type of biases and their dependence on the stellar mass, it is likely that the fraction of bona fide WTTS will be higher among late K and M stars, and lower for F and G stars. Any definitive assessment of the true nature of these RASS sources and of the (statistical) properties of RASS-selected PMS populations will thus have to wait for the availability of high-resolution spectroscopic data, including measurements of the Li I doublet, which will help in screening the bona fide WTTS sources in the sample, at least for the cooler stars.

© European Southern Observatory (ESO) 1997

Online publication: October 15, 1997
helpdesk.link@springer.de