The age of the oldest stars in the local galactic disk from Hipparcos parallaxes of G and K subgiants

We review the history of the discovery of field subgiant stars, the role that they played in the development of the early understanding of stellar evolution, and their importance in the age dating of the Galactic disk. We use the cataloged data from the Hipparcos satellite in this latter capacity. Based on Hipparcos parallaxes with relative accuracies of sigmapi/pi less than or equal to 0.10, the absolute magnitude of the lower envelope of the nearly horizontal subgiant sequence for field stars in the H-R diagram for B-V colors between 0.85 and 1.05 is measured to be M-v = 4.03 +/- 0.06. New stellar evolutionary tracks calculated for metal abundances in the range -0.29 less than or equal to [Fe/H] less than or equal to +0.37 are fitted to the main-sequence, subgiant, and giant-star distributions in the Hipparcos H-R diagram. Isochrones for [Fe/H] = +0.37 provide the best fit to the reddest giants between +3 greater than or similar to M-v greater than or similar to 0, as well as to the envelope of the reddest main-sequence stars at M-v greater than or similar to 5. The red edge of the densest part of the distribution of field giants is, however, most readily matched by isochrones having [Fe/H] approximate to +0.23. Such high metal abundances are evidently confirmed by the spectroscopically observed high metallicity (between [Fe/H] = +0.2 and +0.4) of the old thick-disk Galactic cluster NGC 6791, whose color-magnitude diagram can be made to fit either of these red boundaries by adopting suitable values for the reddening and distance modulus ( from within the observed ranges of uncertainties of these quantities). The age of the field stars in the solar neighborhood is found to be 7.9 +/- 0.7 Gyr (or 7.4 +/- 0.7 Gyr if the stellar models allow for the effects of diffusive processes) by fitting the theoretical isochrones for [Fe/H = 0.37 to the lower envelope of the Hipparcos subgiants. However, this age is a function of metallicity. The models show a dependence of deltat = -3.99([Fe/H] - 0.37) at M-v similar to +4 for metallicities between 0.00 and +0.37. The same grid of isochrones yields ages, in turn, of, 4.0 +/- 0.2 6.2 +/- 0.5, and 7.5 to 10 Gyr (depending on the assumed reddening) for the old Galactic clusters M67, NGC 188, and NGC 6791 using metal abundances, distance moduli, and reddenings adopted in the text. Although the distance ( and hence age) of NGC 6791 is somewhat uncertain, the ages of both the Galactic disk in the solar neighborhood and of NGC 6791 are, nevertheless, likely between 3 and 5 Gyr younger than the oldest halo globular clusters, which have ages of similar to 13.5 Gyr. The conclusion is the same as that reached earlier by Liu & Chaboyer, who used Hipparcos parallax stars that were mainly near the main-sequence turnoff rather than at the lower bound of the subgiant luminosity that we determine here. The most significant results are (1) the supermetallicity of the oldest local disk stars and (2) the large age difference between the most metal poor component of the halo and the thick and thin disk in the solar neighborhood, confirming Liu & Chaboyer. These facts are undoubtedly related and pose again the problem of the proper scenario for the timing of events in the formation of the halo and the Galactic disk in the solar neighborhood.