Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 341, 579-586 (1999)

Previous Section Next Section Title Page Table of Contents

6. Conclusions

The temperatures deduced appear not to be consistent with a model of one single temperature component. Therefore, we have suggested that the bands originate from two different components: One would be a warm and high density layer close to the star, possibly making the [FORMULA] band optically thick, another would consist of a large, cold and optically thin CO2 region extending out in the wind to hundreds of stellar radii. These two components could explain the differences in temperatures and in the number of molecules found from the observations.

The circumstellar CO2 emission bands observed in M giants are probably formed out of LTE and caution should be exercised when discussing the gas temperature and the site of formation of CO2 from observations of the width of an unresolved Q-band. However, in analysing our seven ISO spectra of M giants, we find that the gas temperatures may be estimated from the widths of the bands even if the bands are not in vibrational LTE. We note that Onaka et al. (1998) show that all bands correlate with variability phase. This could be explained by differences in the radiation field if the population is determined by radiative processes.

Since the CO2 molecules may be dominated by radiative pumping by stellar flux via one or some of the many vibrational levels, a detailed treatment of CO2 in the radiation field and wind of the star seems important for modelling the emission and verifying the hypothesis of its origin.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: December 4, 1998