Analysis of the hot stellar population of the globular cluster ω Centauri

We analyze the far-UV and Stromgren u photometric data of the globular cluster omega Cen presented in an earlier paper. The color-magnitude diagram of the cluster from these two bands shows that omega Cen's horizontal-branch (HE) consists of a group of cooler intermediate blue HE (IBHB) and a group of extreme HE (EHB) stars, together with a large population of post-HE stars. Unexpected features in the diagram are a discontinuity between the EHB and IBHB objects lying at (FUV - 3500)(0) similar to - 1.5, an unusually large population of stars below the EHB, and a number of sources bluer than an infinite temperature blackbody. No adjustment of the assumed reddening or distance modulus parameters satisfactorily explains either the sub-HE or very hot star components observed. The radial distributions of the IBHB and EHB subpopulations are similar after corrections for completeness and crowding are made. This result, as well as the fact that omega Cen's core is not dynamically evolved, implies that dynamical effects are not required for the production of EHB stars in globular clusters. To compare the observations to theory we use Hess diagrams, which describe the density distribution of stars in the color-magnitude diagram. To simulate the known abundance spread in omega Cen, we use image-processing "morphing" techniques to create a composite Hess diagram for [Fe/H] = -2.2, -1.5, and -0.5. We populate the zero-age HE (ZAHB) in our simulations assuming either flat or Gaussian distributions in total stellar mass or, alternatively, distributions in the red giant branch mass-loss efficiency parameter eta in Reimers's formula. Our ZAHB's extend to the lowest possible ZAHB mass as determined by evolving models along the red giant branch with extreme mass-loss rates. Neither flat nor Gaussian distributions in ZAHB mass reproduce the observed HB gap or the sub-HB population. However, the eta distribution models can crudely reproduce the gap as well as the sub-HB population while simultaneously fitting the rest of the HB and post-HB population.