Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 363, 958-969 (2000)

Previous Section Next Section Title Page Table of Contents

4. Discussion

4.1. Cluster membership

Table 6 summarizes the criteria we have used in this paper in order to study the membership of our candidates in the Coma Berenices stellar cluster. These criteria are the following: (1) fitting the photometric and spectroscopic cluster sequence (Columns 2 and 3, respectively), (2) the presence of H[FORMULA] in emission and/or H[FORMULA] variability in the optical spectra (Column 4), (3) radial velocity consistent with the mean value of the cluster (Column 5), and (4) proper motion in agreement with the Coma star group (Column 6). We have assigned "Y" or "N" (followed by "?" in the case of marginal consistency) on the basis of what have been discussed in previous sections. The last criterion, proper motions, is the most decisive one when concluding on the membership status of our program stars.


Table 6. Membership criteria.

As the table shows, none of the cluster X-ray candidates was confirmed as a cluster member by our analysis. This result contrasts with those for other clusters where, as mentioned in the Introduction, X-ray surveys and optical follow-up observations were instead very successful in identifying previously unknown cluster members. This, in turn, suggests two possible conclusions: (1) since Coma is considerably older than the clusters where new members were identified through ROSAT surveys, its population of low mass stars, if existent, must be considerably X-ray fainter. Therefore, possible late-type cluster members may have not been detected (discovered) in X-rays because their X-ray luminosities are below the sensitivity of the RSP ROSAT observations; (2) alternatively, Coma sequence does really truncate at [FORMULA], at least in the cluster core which was covered by the X-ray survey. Since the XLDF of solar-type stars in the Hyades and Coma are very similar (the Coma population of solar-type stars actually appears slightly more X-ray active), we can make the plausible assumption that the same occurs for lower mass stars. Under this assumption, the sensitivity of the RPS X-ray survey should have allowed detecting part of these low mass stars; the fact that we do not detect them argues in favour of a drastically reduced population of low mass stars in Coma, consistently with the results of the optical surveys. Unfortunately, the limited cluster coverage of the X-ray survey, which did not extend much beyond the cluster core, does not allow us to say much about the existence (and the spatial distribution) of K/M dwarfs outside the cluster core and, in particular, about the existence of extratidal low mass stars.

4.2. Nature of the X-ray detected sources

Whereas our analysis indicated that the X-ray candidates are not cluster members, we believe that, on the basis of our photometric and spectroscopic data in the visible, the stars #1a, #2a, #3, #4, #7, #10, #11 and #12 can still be considered as the optical counterparts of the X-ray sources of Table 1 (RSP). These optical counterparts are located within a distance of 30" from the X-ray coordinates. They comply with the expected properties of being late-type chromospherically active objects, showing H[FORMULA] emission for most of the late-K/M cases. Only the G-type stars #5 and #8 show a potentially low chromospheric activity. As it can be seen in the finding charts shown by RSP, there are two other non-stellar and fainter objects located close to the star #5, and the star #8 lies slightly further than 30" from the corresponding X-ray coordinates. These facts could be related to the stars not being the optical counterparts of the X-ray sources.

RXJ 1212.0+2232 (#1a) has already been discussed in Sect. 3.4. We divide the remaining sources in two groups on the basis of their location on the colour-magnitude diagrams, namely, those lying above the Coma sequence (group a) and those lying below it (group b). If, as supported by the good matching between the colours and the spectral-types, the reddening towards these stars is negligible, stars in group a) are likely to be slightly closer to us than Coma. The absence of detectable lithium lines in the spectra of these late-type stars suggests that, if they are single objects, they could be as young as 50-70 Myr (a total Li preservation, which takes place until about 10-20 Myr, would have produced features clearly observable with our spectral resolution; Pavlenko et al. 1995). Therefore, these objects cannot be very young stars still above the main sequence and located beyond the Coma cluster. Furthermore, H[FORMULA] EWs of stars in group a) are comparable to those of Hyades stars with the same colours and/or to those of active field dMe dwarfs, and these stars show a very high level of chromospheric activity. Therefore, they are most probably "normal" foreground young-to-intermediate age M dwarfs.

Stars in group b), instead, should have larger distances than Coma. This sample is very probably constituted by a combination of rather young stars and active binaries. The radial velocities we infer, with changes in magnitude and sign, suggest that at least three of the objects in this group (#4, #11, and #12 -plus #2a in group a) could in fact be binary systems: these objects have radial velocity measurements diverging by more than 3[FORMULA] the error bar and thus could be rapid-rotating spectroscopic binaries (note that our spectral resolution is [FORMULA] and 150 km s-1 for the blue and red spectra, respectively). However, observations in the R-band performed to search for photometric variability of these stars on February 29, 2000 (using the IAC80 telescope and the instrumental setup described in Sect. 2.2), showed that there was no detectable variation within 0.05 mag for any of these objects. On the other hand, there is no indication of binarity for stars #5, #8, #10. Having only one red spectrum for the star #7 does not allow us to go further than providing its spectral type and radial velocity at the time of the observation. The lower limits to the X-ray luminosities that we derive for the two G-type stars, if they are the optical counterparts of the X-ray sources, are consistent with an age comparable to the Hyades or slightly younger; vice versa, both the lower limit to its X-ray luminosity and its H[FORMULA] EW suggest that star #10 should be at least as young as the Pleiades (120 Myr).

Our conclusion that this sample of stellar X-ray sources not belonging to Coma is composed by a mixture of young to intermediate-age stars and close active binaries is consistent with the results of the optical follow-up studies of Einstein and EXOSAT serendipitous sources (Favata et al. 1993; Tagliaferri et al. 1994). Furthermore, we have compared the number of detected stellar X-ray sources (including both the 10 objects from the present study and those already known in the SIMBAD database -see RPS) with the expected number based on the models of the distribution of X-ray sources in the Galaxy constructed by Guillout et al. (1996) and on the sensitivity of the RPS X-ray survey. We estimate that no more than one stellar X-ray source (not belonging to Coma) per square degree has been detected in the ROSAT survey and that the predicted number of stellar X-ray sources per square degree is of the order of 2. The sensitivity of the Coma X-ray survey is not uniform, which introduces additional uncertainties in our estimate. It is important, however, the fact that we detect less stellar X-ray sources than predicted and not more: this provides an additional support to the conclusion that the objects studied by us are not members of the Coma Berenices open cluster.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: December 5, 2000