Hydrogen line wings are one of the strongest tests of model stellar atmosphere structure. In the majority of stars hydrogen is a dominant source of continuous opacity and thus for strong hydrogen lines the abundance parameter is excluded. In hot stars, where the broadening of hydrogen lines is dominated by protons and electrons produced by the ionisation of hydrogen, the hydrogen line profiles are dependent only on atmospheric structure and properties of the hydrogen atom. In cool stars the metallicity is important as ionisation of metals is the principal source of ions and electrons which contribute to the Stark broadening of the lines while hydrogen atoms in their ground state produce self-broadening of the lines. Further, hydrogen lines can be observed in all stars, unlike for example limb darkening curves which are one of the strongest tests for solar photosphere models. The large range of opacities within a single line means that many different depths are probed.
If the behaviour of the hydrogen atom in the conditions of stellar atmospheres is understood, the hydrogen absorption lines can be a powerful diagnostic. The transition probabilities are known with extremely high accuracy. Furthermore, the line profile shape is particularly sensitive to atmospheric structure, due to the unique situation of the broadening which derives from the "accidental degeneracy" of states in the hydrogen atom. However, this degeneracy means that the broadening is extremely complex by comparison with metallic lines.
Recently there have been a number of applications of hydrogen lines, particularly lower Balmer lines, to the analysis of stellar photospheric models, and in particular models of stellar convection (Fuhrmann et al. 1993 , 1994; Van't Veer-Menneret & Mégessier 1996; Castelli et al. 1997; Gardiner et al. 1999). Of course, such analyses will be dependent on the accuracy of the theories describing the hydrogen atom properties. Of particular importance for analyses of photospheres is the broadening of the wings in stellar photospheric conditions. This is especially true when testing convection theories, as convection affects the atmosphere in the deeper layers which do not contribute to the core of the line.
It was pointed out by Lortet & Roueff (1969) that the effect of neglecting dispersive-inductive forces should be significant, however, this seems to have gone largely unnoticed. They showed that relative to resonance broadening, dispersive interactions make a significant contribution to the self-broadening of Balmer lines. For Paschen lines they demonstrated that dispersive interactions should dominate the self-broadening. However, in their analysis they used an inadequate theory of the dispersive-inductive interaction which is known to underestimate this type of broadening by typically a factor of two. In this paper a theory of self-broadening of hydrogen lines which includes a better treatment of both resonance and dispersive-inductive interactions, which was announced in an earlier letter (Barklem et al. 2000, hereafter Paper I), is presented. It is shown that the inclusion of these interactions has a significant effect on the predicted profiles of Balmer lines in cool stars.
© European Southern Observatory (ESO) 2000
Online publication: December 5, 2000