8. Discussion and conclusions
From the IRAS faint source catalog we selected 29 potential AGB stars in the direction of the SMC. Near-infrared photometry was obtained for that sample and two foreground late-type giants. One additional red object was discovered serendipitously. Optical spectroscopy was obtained for a sub-set of stars including two AGB stars in the LMC. Improved IRAS fluxes were obtained using the GIPSY package.
With this sample, it is possible to make a first comparison with AGB stars in the LMC. It turns out that the reddest stars in the SMC are as red as the stars so far found in the LMC. For example, the reddest two LMC AGB stars in Reid (1991) have of 3.00 and 2.85, those in Wood et al. (1992) have = 2.30 and 2.81, and those in van Loon et al. (1997) have = 4.9 () and 2.9. The reddest objects in our sample have = 5.0 and 2.8. This is somewhat surprising as one might have expected that stars in the SMC lose mass at a lower rate because of the lower metallicity. It may be a consequence of the fact that only the most heavily mass-losing stars in the more distant SMC will have been detected by IRAS.
On the other hand, the stars are not as red as the reddest Galactic AGB stars, which reach colours up to = 6.0, or = 10 (e.g. Fig. 5 in Groenewegen et al. 1997).
Of the 30 AGB star candidates, 5 are known or confirmed as carbon stars (S2,12,14,20b,30) and 6 as oxygen-rich stars (S9,10,15,18,20a,22). Three of these (S2,14,30) carbon stars are peculiar in the sense that the C2 and CN bands are unusually weak. One of these (S2) had previously been suggested to be in an interacting binary system, or a post-AGB star. This could also be the case for the other two objects. Based on colour-colour diagrams an additional 4 carbon- and 2 oxygen-rich stars are suggested. In three cases no red object could be found near the IRAS position. In one case we have obtained multiwavelength ISOCAM images that show an extremely red object, that indeed could not have been found in K. There always remains the possibility that some of the non-confirmed C-/O-objects are red sources unrelated to the IRAS source. This suspicion is confirmed in a few cases by the unusual position of some objects in the colour-colour diagrams (e.g. S8, 26) which indicate that the NIR photometry does not connect very well with the IRAS fluxes. In order to study this sample further, and in particular to establish the nature of these sources, we are in the process of obtaining multiwavelength ISOCAM images and ISOCAM CVF spectra.
© European Southern Observatory (ESO) 1998
Online publication: March 10, 1998