Astron. Astrophys. 336, 654-661 (1998)

1. Introduction

The semiregular (types SRa and SRb) and the cool irregular variables (Lb) - quite numerous groups of objects among the stars on the Asymptotic Giant Branch (AGB) - have been almost neglected although their role within the evolution on the AGB and their overall properties are far from being understood. They can provide important constraints for theoretical models due to their different pulsational behaviour compared to the more frequently studied Mira variables. Fortunately, this situation has changed during the last years - at least in the case of the semiregular variables (SRVs). In Kerschbaum & Hron (1992, 1994, 1995; hereafter SRI, SRIIa, and SRIII, respectively) and Kerschbaum (1995; hereafter SR IIb) stellar properties of SRVs, derived from the GCVS4 (Kholopov et al. 1985-88), the IRAS-PSC (1988), the IRAS-LRS (1986), as well as new near infrared photometry, were used to divide this inhomogeneous group of objects into physically distinct classes and to probe their evolutionary status. One of the main outcomes was the definition of O-rich subclasses namley `blue'-, `red'-, and `Mira'-SRVs with the first having short periods (typically below 100 days) and showing no mass-loss. The border between `red'- and `Mira'-SRVs is mainly defined by the period (P200<sup>d</sup> for `red', and P200^d for `Mira'). The `blue'-SRVs seem to be on the early and the other two groups on the thermally pulsing AGB. A study by Jura & Kleinmann (1992) came to similar conclusions concerning the galactic distribution of these stars.

Complementary work on the cool irregular variables of type Lb (IRVs) have to be done in an analogous way, but a major problem is that the IRVs may also act as a kind of "trash group" for poorly observed late-type variables as is partly the case for the SRVs (see Lebzelter et al. 1995). From what is known about the luminosities of IRVs, and since this group contains also Carbon-stars, a significant number of them should be on the thermally-pulsing AGB (Querci 1986). However, Little et al. (1987) found no Technetium, which would be an evidence for a recent thermal pulse, in the O-rich IRVs of their sample.

Peters (1991) used data from the IRAS-mission for an analysis of the space distribution and the mass-loss of the main three AGB-variables: the Miras, the SRVs and the IRVs. The main outcome concerning the latter is that they seem to have mass-loss rates comparable to those of SRVs but smaller than those of Miras. He arrived at similar scale heights for all three groups except for the long-period Miras which were more concentrated to the disk. In a study mainly devoted to SRVs, Jura & Kleinmann (1992) derived a galactic distribution of the IRVs comparable to that of "thin disk" Miras. Unfortunately, their sample was a mixture of IRVs and a significant number of SRVs and Miras with unknown periods. Moreover, it was limited to objects with having GCVS and IRAS-data. All these restrictions did not allow a more detailed analysis.

Recently, Kerschbaum et al. (1996b, hereafter LbI) found that their sample of visually bright IRVs displays infrared properties very similar to the SRVs. As judged from NIR two-colour diagrams the O-rich IRVs seem to have intermediate atmospheric conditions between Miras and normal giants. There may be a slightly larger "contamination" with non-AGB giants than in the case of the SRVs but the AGB objects seem to resemble the `blue' and the `red' SRV groups. We note that using only IR-colours the S- and the Carbon-stars among the Lbs are indistinguishable from SRVs of the same chemistry. In the paper by Kerschbaum et al. (1996a), dealing with NIR- and IRAS-photometry, spectral energy distributions, mass-loss rates from IRAS-data, galactic distribution and the evolutionary status of O-rich AGB-variables, the similarity between the IRVs and SRVs was demonstrated in many more aspects.

Besides the dust mass-loss derived from IRAS photometry circumstellar CO emission has proven to be a good measure of the gas mass-loss rate. Whereas quite a lot has been done for evolved AGB objects like Carbon-stars (Kastner et al. 1993; Nyman et al. 1992; Olofsson et al. 1993; Groenewegen et al. 1996), OH/IR-stars (Heske et al. 1990), and more recently S-stars (Bieging & Latter 1994; Sahai & Liechti 1995; Groenewegen & de Jong 1998), the low end of the mass-loss rate distribution, populated by objects situated on the early AGB, or at the beginning of the thermally pulsing AGB, has not been studied in any detail.

Kahane & Jura (1994) published high quality observations of 11 nearby O-rich SRVs in the CO (J=1-0 and 2-1) lines. They came to the conclusion that these short-period objects have mass-loss properties quite similar to those of longer-period Miras. Kerschbaum et al. (1996c, hereafter SRIV) extended this work considerably in their study of O-rich SRVs. The majority of their detected objects, covering both small and longer periods, are weak in CO, i.e., they are low mass-loss rate objects (), and have in general envelopes with small expansion velocities (the mean value is 8 km/s).

A recent CO (J=3-2) study was devoted to nearby, optically bright O-rich Miras (Young 1995, Y95), whose periods are 2-3 times higher than those of the SRVs (they are probably pulsating in a lower pulsational mode). Among the results we note that the average circumstellar gas expansion velocity of this volume-limited sample is small compared to that of bright infrared Miras (see e.g., Nyman et al. 1992) and Carbon-stars (Olofsson et al. 1993), and that the mass-loss rates are low and increase with decreasing effective temperature.

These studies suggest that the mass-loss properties of the bright O-rich SRVs and Miras are very similar and hence support the view that the mass-loss is not strongly affected by the pulsational mode, at least not in this mass-loss rate regime ( few).

© European Southern Observatory (ESO) 1998

Online publication: July 20, 1998
helpdesk.link@springer.de