Knowledge Bank

In order to study the importance of molecules in the spectra of cool giant stars, we have computed model atmospheres for several such objects. The models have been computed with the Indiana University Stellar Atmospheres Code, which uses a highly modified version of the Strom code for temperature correction. For purposes of calculating molecular dissociative equilibrium, we include 25 diatomic and 26 polyatomic molecules of H, C, N, O, and Si. Opacity sources include both absorption and scattering due to H., $H2+$, $H2-$, H, $H-$, $H2O$, CO, and CN. The molecular opacities are added as straight mean opacities. We employ about 35 points in frequency and 60 in optical depth, and a relative flux accuracy of at least 0.0001 is usually required at the worst point in the atmosphere. We include both radiative and convective transport of energy, the latter using a mixing-length theory with the ratio of mixing length to scale height equal to 1.0. Convection never carries more than 0.01 of the flux for any of the models described here. We have calculated a series of model atmospheres centered around $Teff=3000$ K, log (surface gravity) = 1.0; and a chemical composition of H, He, C, N, and O by number of $1.0,0.1,5.3\times 10-5$, and $9.6\times 10-5$, and $6.0\times 10-5$. That is, we have lowered the abundance of carbon relative to hydrogen by an order of magnitude from the solar value, and then lowered the oxygen abundance to give a ratio C/O = 0.88. This composition approximates that which might result from the mixing of nuclear processed material from the stellar interior with the original matter in the envelop. We have then varied $Teff$, log g, and C/O. For abundances similar to the sun (C/O = 0.58), the atmospheres in this temperature range are dominated by the opacities of H---, and $H2O$. As the C/O ratio is changed toward 1.0, the principal effect is the removal of $H2O$ from the atmosphere, which results in a larger fraction of scattering (relative absorption) and a lower temperature gradient throughout the atmosphere. For a specific model ($Teff=3000$ K, log g = 1.0, and C/O = 0.95), we predict stronger CO and $H2O$ features than are observed in the star $\alpha$ Ori, for which these parameters have been suggested as representative. Better agreement could be obtained, we predict, if some additional sources of continuous opacity could be added. In fact, we suggest that such molecules as HCN, $C2H$, and $NH2$ might be important in these objects and urge that work be done to provide the needed wave-lengths and strengths of lines for these molecules