Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 322, 256-265 (1997)

Previous Section Next Section Title Page Table of Contents

8. Summary

We have investigated the influence of NLTE effects on the analysis of hot white dwarfs and subdwarf B stars. Present day NLTE atmospheres are of similar sophistication as up-to-date LTE atmospheres. Thus there exists no more a fundamental reason to prefer LTE atmospheres as in former times, when drastic simplifications were necessary to perform NLTE computations. Nevertheless, the numerical effort required to perform those calculations is still very large. For economical reasons one is therefore interested to use LTE model atmospheres whenever that is possible without a loss of accuracy. To establish when loss of accuracy is to be expected we have compared profiles from self-consistent NLTE and LTE models representative for hot white dwarfs and sdB stars.

Model atmospheres for hydrogen-rich white dwarfs were calculated for [FORMULA], typical for hot white dwarfs with various traces of helium. The pure hydrogen models show moderate NLTE deviations only for [FORMULA] K. NLTE and LTE models are in perfect agreement for temperatures up to 70000 K. This means that nearly all DA white dwarfs, except a few extraordinary hot ones, are accessible to LTE techniques. The situations becomes completely different, however, as soon as traces of helium are present in the atmosphere. As Bergeron et al. (1994) have shown, the Balmer lines are strongly modified in LTE atmospheres with [FORMULA] or [FORMULA] and temperatures not much higher than 40000 K. We have shown that this is an artifact caused by the assumption of LTE. The influence of small traces of helium on the Balmer lines vanishes for NLTE atmospheres. If a LTE analysis of the hydrogen lines is to be performed we recommend to use pure hydrogen models, even if it is likely that trace elements are present in the atmosphere.

For DAO white dwarfs with [FORMULA] K there is reasonable agreement between LTE and NLTE Balmer line profiles. Nevertheless, one should always be aware of moderate NLTE deviations. They only vanish for [FORMULA] K, the regime of the DAB white dwarfs. Drastic effects are present for hot DAO white dwarfs ([FORMULA] K). The situation becomes even worse, because the hot DAO white dwarfs tend to have lower gravities, which amplifies deviations from LTE. Thus the use of NLTE atmospheres is mandatory for reliable results. Generally speaking the result for the helium lines is similar. Moderate NLTE effects are present for cooler DAO atmospheres. They become quite drastic for hot (lower gravity) DAO stars.

For the hot, helium-rich DO white dwarfs we found large differences for many important lines even at relatively low temperatures. This prevents accurate LTE analysis for virtually all DO white dwarfs. The use of NLTE calculations is strongly recommended.

Subdwarf B stars have lower gravities than white dwarfs. Thus the NLTE effects are more pronounced. LTE results are in good agreement with NLTE results for [FORMULA] K. However, for higher temperatures (the sdOB regime) the deviations become larger very quickly and should be considered for accurate analyses. Additionally, we have checked the effect of metal line blanketing on sdB atmospheres and shown that the influence is only slightly smaller than that of NLTE effects for the hot sdOB stars, but for the cooler "classical" sdB stars the influence of metals is certainly the dominant effect of both.

We further investigated whether LTE results could be improved by NLTE line formation calculations. For the sdB model this worked well. Also the hot DAO models were substantially improved, although notable deviations remained. No substantial improvement could be obtained for the DA model with trace helium nor the DO model atmosphere. The reason is a modification of the atmospheric structure by NLTE effects in the latter two cases. We conclude that NLTE line formation on LTE atmospheres can be an useful tool, but a careful verification of the basic assumption (no repercussion on the atmosphere) is necessary beforehand.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: June 30, 1998