2010AJ....140.1020V

Isochrones for ages ≳ 4 Gyr and metallicities in the range -2.5 ≲ [Fe/H] ∼ +0.3 that take the diffusion of helium and recent advances in stellar physics into account are compared with observations in the Johnson-Cousins BV(RI)_C photometric system for several open and globular star clusters. The adopted color-T_eff relations include those which we have derived from the latest MARCS model atmospheres and the empirical transformations for dwarf and subgiant stars given by Casagrande et al. (CRMBA). Those reported by VandenBerg & Clem have also been considered, mainly to resolve some outstanding questions concerning them. Indeed, for the latter, V - I_C colors should be corrected by ~-0.02 mag, for all metal abundances, in order to obtain consistent interpretations of the observed (B - V, V), (V - R_C, V), and (V - I_C, V) diagrams for M 67 and the Hyades as well as for local subdwarfs. Remarkably, when the subdwarfs in the CRMBA data set that have σ_π/π ≤ 0.15 are superimposed on a set of 12 Gyr isochrones spanning a wide range in [Fe/H], the inferred metallicities and effective temperatures agree, in the mean, with those given by CRMBA to within ±0.05 dex and ±10 K, respectively. Thus, the hot T_eff scale derived by CRMBA is nearly identical with that predicted by stellar models; and consequently, there is excellent consistency between theory and observations on the Hertzsprung-Russell diagram and the different color-magnitude diagrams considered in this investigation. To obtain similar consistency, the colors obtained from the MARCS and VandenBerg & Clem (B - V)-T_eff relations for metal-poor dwarf stars should be adjusted to the red by 0.02-0.03 mag. In general, isochrones that employ the CRMBA transformations provide reasonably consistent fits to our BV(RI)_Cphotometry for main-sequence stars in the globular clusters 47 Tuc, M 3, M 5, M 92, and NGC 1851–but not the cluster giants (when adopting the synthetic MARCS colors). We speculate that differences between the actual heavy-element mixtures and those assumed in the theoretical models may be the primary cause of this difficulty.