Astron. Astrophys. 342, 799-808 (1999)

1. Introduction

Together with H₂ and CO, SiO is one of the most abundant molecules in the atmospheres of oxygen-rich AGB stars. It is also regarded to be the prime condensate for the silicate dust that is formed around these objects (Gail & Sedlmayr 1986). Several features in the infrared and radio spectra of late-type stars are produced by SiO, probing quite different regions in their atmospheres and circumstellar shells.

The band system generated by the first overtone rotation-vibration transitions is situated in the near infrared. It is formed in the "photosphere" (= deeper atmospheric layers with T K), which is, especially in cool Mira variables dominated by stellar pulsations and related shock waves. This band system is particularly interesting, since its short wavelength end at m can easily be observed with ground based telescopes (e.g. Rinsland & Wing 1982, Aringer et al. 1995) and is not contaminated with emission or absorption caused by other molecules (except some water absorption in very cool stars). For some of the brighter AGB stars there are also high resolution Fourier-transform spectra available (Ridgway et al. 1984, Tsuji et al. 1994). In contrast to the fundamental band system situated around m the first overtone transitions do not interfere with the strong dust features produced by the circumstellar shells of many AGB stars.

In Paper I of this series (Aringer et al. 1997a, Paper I) we have published a large grid of synthetic SiO spectra calculated from hydrostatic MARCS atmospheres. They cover the range between 2.0 and 12.5 µm at a very low resolution of about 50 and the region of the first overtone bandheads around 4 µm at a medium resolution of 4000. We have investigated the behaviour of the SiO features as a function of different stellar parameters like temperature, log (g), stellar mass and metallicity, and we have compared our results to existing observations. One of the conclusions of Paper I was that the appearance of the SiO bands in giants with (K and early M) could be very well explained by our models, while they tend to predict too intense absorption features for the cooler and more extended stars. This problem has also been discussed by Tsuji et al. (1994), who proposed a "warm molecular envelope" (Tsuji et al. 1997) in order to remove the discrepancies. However, it should be mentioned in this context that one may not expect that the atmospheres of AGB variables showing such phenomena as pulsation connected with shock waves, heavy mass loss and dust formation can be described by hydrostatic models.

This topic will be discussed in the current paper where we will focus on our own observations of AGB stars (see also Aringer et al. 1995) and synthetic spectra based on dynamical model atmospheres as they have been computed by Höfner & Dorfi (1997). A similar method to reproduce the observations of carbon-rich objects has been applied by Loidl et al. (1997) and Hron et al. (1998).

© European Southern Observatory (ESO) 1999

Online publication: February 23, 1999
helpdesk.link@springer.de