The universe and globular clusters: An age conflict?

Chaboyer (1995) has recently shown that the range of possible ages of globular clusters, tau similar to 14-18 Gyr, corresponding to different choices of distance scales, is further enlarged mainly by uncertainties in modeling convection. Within the mixing-length theory (MLT) the result is tau = 11-21 Gyr. Bolte & Hogan (1995) adopted the held subdwarfs as indicators of the distance scale and, assuming that tau is insensitive to the treatment of convection, found that tau = (15.8 +/- 2.1) Gyr. Within Chaboyer's findings, it is marginally possible to avoid an ''age conflict'' with the expansion age of the universe, even if this latter is t(0) = 8-11 Gyr, while this is not the case with the Bolte & Hogan (1995) conclusion. Mazzitelli, D'Antona, & Caloi (1995) have implemented updated macro- and microphysics inputs and determined the distance scale by fitting the horizontal-branch models to the horizontal branch of metal-poor globular clusters. Within the MLT, they find tau similar or equal to 14 Gyr. Switching from MLT to a full spectrum of turbulence (FST) model of convection (Canuto & Mazzitelli 1991) lowers the value of tau by 1-2 Cyr. Since the Mazzitelli et al. (1995) results would alleviate the age conflict to the point of making it disappear, it is important to assess their validity by answering the following questions. Independently of the assumptions on the distance scale: 1. What is the role of convection near turnoff on both its shape and luminosity? Chaboyer (1995) and Mazzitelli et al. (1995) find an appreciable effect, while Bolte & Hogan (1995) a priori assume no such effect. 2. How meaningful is the fit of the red giant branch in the determination of tau? In this paper, we answer question (1) carrying out extensive new stellar model calculations. When we assume the long distance scale, the FST results differ from the MLT values both in the shape and in the luminosity of the turnoff, while there is no appreciable difference if one employs the short distance scale. Helium sedimentation also affects the shape of isochrones and age determination. The long distance scale is also shown to be consistent with the subdwarf location. As for (2), we find that, depending on the correlations between T-eff and colors, both FST and MLT results can be made compatible with both distance scales. Models including sedimentation of helium improve the agreement with the red giant-branch location for low ages. At present, red giant branches can even be misleading as age indicators. In conclusion, we choose the long distance scale on the basis of both horizontal-branch luminosities and the present data on subdwarf location, with resulting ages of about 12 Gyr for the most metal-poor globular clusters.