I read an article this morning on the new study of "red and dead" galaxies. It contained a statement that blue globulars were composed of metal poor stars and red globulars were composed of metal rich stars. In general old stars have lower metallicity and newer stars higher. Blue stars generally have short life spans and red stars have much longer ones. Therefore the statement about the metallicity of blue globulars vs. red globulars seems to be a contradiction. I can think of two possible reasons why it may not be. The first is that the stars in the globular are so old that there is a concentration on the left side of the asymptotic giant branch and/or the cluster is so old that significant merging of old red stars has taken place. However, this is just a guess on my part. Does anyone know the real reasons astronomers think the blue globulars have lower metallicity?
Brad Walter
In this case, I suspect that the authors of that "red and dead" article are thinking only of the old populations found in globular clusters such as we find in the Milky Way galaxy. If one compares the Herzsprung Russell diagram of a metal-rich globular cluster (close to but a little below solar metal abundance) to that of a metal-poor globular cluster (with 1/100th the solar heavy element abundance) with equal ages, theory leads us to expect distinct differences. The HR diagram of the main sequence and, especially, the red giant branch is cooler (and thus redder) for the metal-rich globular cluster than for the metal poor globular cluster. That is also generally true for the helium burning horizontal branch stars in the clusters, although other perameters may play a role there in addition to heavy element abundance. Many examples of old globular clusters with different metal-abundances are known in the Milky Way where this expectation has been confirmed. For example, 47 Tuc is a relatively metal-rich globular cluster and M15 is a metal-poor globular cluster (confirmed spectroscopically). They may not have identical ages, but both are about 11 or 12 billion years old. As a consequence, the average color of the cluster as a whole is redder for the metal-rich globular than for the metal-poor globular, even if they are both equally old. Thus, after correcting for any interstellar reddening, we find that the integrated color of 47 Tuc is redder than that of M15. You are quite right that things are different if very young and very old clusters are considered, where the young cluster might have blue main sequence stars that are no longer present in the old cluster.
Horace Smith