On the temperatures, colours, and ages of metal-poor stars predicted by stellar models

Abstract

Most (but not all) of the investigations that have derived the effective temperatures of metal-poor, solar-neighbourhood field stars, from analyses of their spectra or from the infrared flux method, favour a Teff scale that is ∼100–120 K cooler than that given by stellar evolutionary models. This seems to be at odds with photometric results, given that the application of current colour–Teff relations to the observed subdwarf colours suggests a preference for hotter temperatures. Moreover, the predicted temperatures for main-sequence stars at the lowest metallicities ([Fe/H] <-2) appear to be quite robust; i.e. it may not be possible to produce significantly reduced values of Teff for them unless some fundamental modification is made to the adopted physics. No such problems are found if the temperatures of metal-poor field stars are ∼100–120 K warmer than most determinations. In this case, stellar models would appear to provide consistent interpretations of both field and globular cluster (GC) stars of low metallicity. However, this would imply, e.g. that M 92 has an [Fe/H] value of approximately - 2.2, which is obtained from analyses of Fe I lines, instead of approximately equal to - 2.4, as derived from Fe II lines (and favoured by studies of three-dimensional model atmospheres). Finally, the age of the local, Population II subgiant HD 140283 (and GCs having similar metal abundances) is estimated to be ≈13 Gyr, if diffusive processes are taken into account.