Astron. Astrophys. 349, 553-572 (1999)

2. Observational data

The optical spectra for the stars listed in Table 1 have been discussed in Papers I and II. In the former paper, details are presented of selection criteria and observing and data reduction techniques, whilst in the latter, equivalent widths of selected lines are tabulated.

Table 1. Galactic supergiants and their non-LTE atmospheric parameters. Random errors are typically 1 000 K for , 0.2 dex for , 20 kms^-1 for and 5 kms^-1 for . Note that the effective temperatures and logarithmic gravities are based on standard non-LTE techniques and they may not be physically realistic - see text for further discussion. Also listed are the stars' CN status, where  is `normal/moderate',  is `processed?' and  is `highly processed' - see Sect. 7.1. The final two columns contain estimates of distance moduli and luminosities, these are discussed further in Sect. 2.

Table 1 also presents estimates of each star's distance and luminosity, which are discussed below. Each object has already been assigned membership of a galactic association or open cluster in Paper I. We have taken distance estimates for these stellar aggregates from the literature, using the catalogue of Rupprecht (1966) as a starting point, but using more recent estimates if these were judged to be more reliable. These additional sources were; de Zeeuw et al. (1999) for distance moduli for OB associations and clusters derived from Hipparcos data, the Hipparcos catalogue itself (ESA SP-1200 1997), Massey et al. (1995) for recent spectroscopic parallax results, plus catalogues of Becker & Fenkart (1971), Nicolet (1981) and Humphreys & McElroy (1984). In Paper I the star HD12301 was not assigned membership of any association and we have therefore used the luminosity published by Zorec & Briot (1991) which is derived from a Balmer decrement calibration. We have also reassigned membership of HD213087 from Cep OB1 to Cep OB2, and used the Hipparcos association distance rather than individual Hipparcos distances for all Cep OB2 stars. HD190603 was previously assigned membership of Vul OB2, but upon closer examination it is not obvious which association, if any, this star belongs to. There are stellar aggregates at approximately 1, 2 and 4 kpc in this direction and we have arbitrarily taken a distance of 1.5kpc. Its distance must be considered highly uncertain.

In fact, it should be emphasized that both the distances to individual associations and memberships of those associations are in many cases problematic. As an example we consider HD53138 ( CMa) and HD58350 ( CMa), both of which were considered secure members of Coll 121. Hipparcos data for Coll 121 was discussed in detail by de Zeeuw et al. (1999) who concluded from parallax and proper motion studies that neither of these stars are members. Per OB1 is another important example, since many of our targets are in this association; we have adopted a distance modulus of 11.9 (corresponding to a parrallax of approximayely 0.4 mas) but two stars (HD13841 and HD13866) have Hipparcos parallaxes of 1.83 and 2.92 mas (though with standard errors of 0.92 and 0.83 mas respectively). We have chosen to adopt a uniform distance for all Per OB1 stars however, since these smaller distances would put these stars on the main-sequence while their spectra are clearly those of supergiants (the extinctions are also consistent with Per OB1 membership). Finally, even the Hipparcos distance of  Ori of 4.520.77 mas, corresponding to a distance modulus of 6.7, is controversial and should be compared with Hipparcos distance moduli ranging from 7.6 to 8.5 for the different Orion subgroups (Brown et al. 1999).

© European Southern Observatory (ESO) 1999

Online publication: September 2, 1999
helpdesk.link@springer.de