Opacities for stellar envelopes

We define stellar envelopes to be those regions of stellar interiors in which atoms exist and are not markedly perturbed by the plasma environment. Availability of accurate and extensive atomic data is a prime requirement for the calculation of envelope opacities. For envelopes we adopt the criterion of mass density $$\rho \le 0.01\ {\rm g} \ {\rm cm}^{-3}$$. We present radiative Rosseland mean opacities for envelopes obtained using atomic data calculated in an international collaboration referred to as the Opacity Project, or OP. Equations of state are calculated using an occupation-probability formalism. To a good approximation, ionization equilibria and level populations in envelopes depend only on the temperature T and electron density $$N_{\rm e}$$ and are insensitive to chemical mixtures. Monochromatic opacities for all abundant chemical elements are therefore calculated on a grid of ($$T,\ N_{\rm e}$$) values and are archived. Rosseland mean opacities are then readily calculated for any chemical mixture. Tables of Rosseland means, for any required mixtures and as functions of $$\rho$$ and T, are available on request in computer-readable form. The present, op, results are compared with those from another recent study, referred to as opal, by C. A. Iglesias and F. A. Rogers at the Lawrence Livermore National Laboratory. The agreement between the op and opal calculations is generally good, although there are some differences. Both calculations give results larger than those obtained in earlier work, by factors of up to 3 or more.