Giants in old open clusters - Temperatures, luminosities, and abundances from infrared photometry

We present infrared observations for more than 100 red giants and clump stars in eight old open clusters. We have assembled the best available optical photometry for these stars and determined a consistent set of integrated physical parameters (reddening, distance modulus, abundance, and age) for the clusters. From color-color and color-magnitude diagrams that make use of both the infrared and optical data, we are able to identify quite a few stars as probable field stars rather than cluster members. However, because of a general scarcity of bright stars, it is often difficult to distinguish between cluster members on the asymptotic giant branch and field interlopers. In a ( U-V)_0, ( V-K)_0 plot, stars from the most metal-poor open clusters tend to lie between the relations defined by field and globular cluster giants. On the other hand, nearly all of the open cluster stars lie near the field giant line in a (H-K)_0 plot. The mean CO strengths of the giants in each open cluster show a range consistent with the optically determined range in [Fe/H], but the correlation between these two quantities is weak, probably because of the small total range of each and the significant uncertainties in [Fe/H]. The results for the open clusters, though, are consistent with the relation between CO and [Fe/H] established for globular clusters and considerably strengthen that relation near the solar metallicity end. For these eight open clusters, there is a modest linear correlation between [Fe/H] and age which shows a gradient in metallicity of about —0.1 dex per Gyr and gives [Fe/H] ~ -0.6 for τ = 5 Gyr, the age of the sun. The relation is quite similar in slope and zero point to that exhibited by clusters in the Large Magellanic Cloud. If the open cluster data are adjusted for the galactic radial metallicity gradient, the age-metallicity relation becomes intermediate to those of the Large Magellanic Cloud and the solar neighborhood. This could be an indication that the old open clusters are representative of the stellar population of the galactic thick disk. The temperatures of the cluster giant branches determined from infrared observations are tightly correlated with the optically derived values for age and [Fe/H]. Also, there is general agreement between the location of the cluster giant branches in an H-R diagram determined from infrared photometry and the predictions of the Revised Yale Isochrones. Specific differences that exist between the theoretical and semiempirical parameters can at least partially be attributed to uncertainties in the (primarily) optical data and/or the presence of convective overshootin