Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 323, 909-922 (1997)

Previous Section Next Section Title Page Table of Contents

1. Introduction

The physics of stellar photospheres aims at both, the detailed understanding of all intrinsic processes that generate the observed spectrum, and - even more - at deriving the stellar input parameters for numerous applications in e.g. galactic research. In that respect, the Sun is a unique laboratory and reference source. Albeit we learn from d irect observations and the extremely resolved spectrum that our means of modeling are not at all sophisticated, one may nevertheless state that we understand at least the basic processes in the solar photosphere. Along with its precisely known effective temperature and surface gravity, standard model atmosphere analyses for instance result in a remarkably good match of the elemental abundance pattern compared to the meteoritic one derived from carbonaceous C1 chondrites in which most of the abundances of the primeval solar nebula are recorded. At this confidence level there is justified hope to a similar understanding of stellar photospheres, at least for solar-type stars and as long as we intend to work differentially. It seems reasonable to restrict the sample to spectral types F and G on or close to the main sequence and ask whether there are other single-lined standards that provide accurate stellar parameters from direct methods. The answer to this question is "yes", but there is currently only one - the F5 star Procyon. It is close by and therefore allows the direct measurement of the stellar diameter. Because of its white dwarf companion it is also possible to derive the stellar mass from astrometric methods. Steffen (1985), among others, has analysed this star at length and found discrepant results from the standard model atmosphere approach in that the surface gravity derived from the ionization equilibrium turns out to be [FORMULA], compared to the direct astrometric value [FORMULA].

A discrepancy of this amount is without doubt not acceptable and has led to the notion that the temperature stratification of the standard hydrostatic model atmosphere approach is erroneous, or that the assumption of local thermodynamic equilibrium is not fulfilled, or both.

Irrespective of this major shortcoming, F stars are of course not excluded from stellar analyses. But this seems risky, unless we can solve or circumvent the contradicting problems we have with Procyon. One might conjecture that many spectroscopists (as well as photometrists) are not aware of these fundamental problems and take model atmospheres and results of so-called standard stars for granted. Kurucz (1995a) e.g. is a good reference for a recent discussion of probable errors that can arise following the "black box" approach.

On the other hand, we intend to demonstrate in what follows that the spectrum of Procyon nevertheless contains the unequivocal fingerprints of the correct surface gravity value in the framework of standard model atmospheres.

It is well-known since long that strong lines with pronounced wings can be good tracers of the gravity parameter, since stars with extended atmospheres provide much less support for collisional broadening (e.g. Gray 1992, and references therein). Cayrel & Cayrel (1963), for instance, employed the pressure-dependent Mg Ib lines [FORMULA] 5172 and [FORMULA] 5183 in [FORMULA]  Virginis (G8III) as gravity indicators, and, more specific, Cayrel de Strobel (1969) presented the Mg Ib triplet lines as one of the best gravity criteria for late-type stars. Later, Blackwell & Willis (1977) derived the [FORMULA] parameter of Arcturus (K2 III) from the strong Fe I line [FORMULA] 5269. In combination with weak Fe I lines they proved to be fairly independent of the effective temperature with this method. More recently, Smith, Edvardsson & Frisk (1986) determined the surface gravity parameter from the pressure-broadened Ca I line [FORMULA] 6162 in their analysis of [FORMULA]  CenA (G2 V) and [FORMULA]  CenB (K0 V), followed by [FORMULA]  Ceti (G8 V) and [FORMULA]  CasA (G0 V) in Smith & Drake (1987), the K0 giant Pollux (Drake & Smith 1991), the G8 subdwarf Groombridge 1830 (Smith et al. 1992) and the K2 dwarf [FORMULA]  Eridani (Drake & Smith 1993).

Most remarkable, however, is the work of Edvardsson (1988a, 1988b, and references therein), who compared the strong line method to the ionization equilibrium method. His elaborate study made use of strong lines from Fe I and Ca I [FORMULA] 6162 applied to a sample of 8 nearby subgiants with [FORMULA]  CenA+B and Arcturus included. As a result, the surface gravities derived from the ionization equilibria of iron and silicon are found to be systematically lower than the strong line gravities. This, as Edvardsson proposes, may be an effect of errors in the model atmospheres, or departures from LTE in the ionization equilibria. Since then many abundance analyses of late stars have been done. But except for few investigations - such as those mentioned above - most of them do not take advantage of the wealth of information stored in the wings of strong lines.

By means of the strong Mg Ib lines and in conjunction with other, rather weak, Mg I lines we endorse in what follows that the strong line method is able to earmark the surface gravity in a very reliable way in most F and G dwarfs. This especially holds true since we proceed differentially and therefore a precise knowledge of neither the effective temperature nor metal abundance is required. As a consequence it is also possible to derive a very well established iron abundance from Fe II, the dominant stage of ionization in our temperature range.

In Sect. 2 we give a short description of the observations. Sect. 3 discusses the model atmospheres and line formation. We then proceed in Sect. 4 to exemplify with respect to Procyon how we intend to derive stellar parameters in future analyses. Sect. 5 gives some results for other, predominantly metal-poor stars, followed by the conclusions in the final section.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: May 26, 1998