Multiple stellar populations in globular clusters (overview).
like this in NGC2808 (paper)
or even good old M4, and and NGC1851, and of course Omega Cen.
the original view of globular clusters is that they are the quintessential coeval sets of stellar populations - that all the stars in any given cluster ought to have formed more or less simultaneously, with an age spread of no more than a few million years or so
but, now that we have very good colour-magnitude data from wide field Hubble images, with proper-motion selected field decontamination, it turns out that many, maybe most, clusters have two or three distinct stellar populations - either with a age spread of tens or even hundreds of millions of years, or significant compositional differences, or both.
Compositional differences are interesting, with suggestion of either extreme helium abundance variations (how?), or significant alpha element variations.
Possibilities include self-enrichment, multiple epochs of star formation or mergers of two or more clusters.
Each explanation has problems.
Real issue is the degeneracy - there is clearly a compositional difference, we can measure giant abundances; question is whether the CMD split is only due to composition and evolution differences, which pushes conclusions like the extreme Helium abundance variation or a rapidly rotation sub-population; or is some of the difference due to age difference, and is that difference tens or hundreds of millions of years.
Large age differences are problematic, unless we're looking at cluster mergers, but it seems implausible that all clusters have undergone mergers.
The Na/O anti-correlation is intriguing, and how exactly we get sodium enriched, but oxygen depleted stars is worrying - models invoking extreme helium enriched main sequence stars just seem wrong, IMNSHO, even if they work from modeling perspective.
- Log in to post comments
About how many clusters have data of sufficient quality to suggest multiple populations? I am curious to know if there are any (again, with similar quality data) CMDs that show no clear indication of multiple populations?
This is extremely interesting stuff...
That is a good question, since the data are quite heterogenous.
I think there are only a handful of clusters with split isochrones "by eye", but about 20 that have hi-res spectra and as I recall essentially all of them have broad Na-O anticorrelations suggesting at least few % mixed population.
But I don't have hard numbers at hand.
Null results are hard to pin down of course.
As is often the case might be easier to pin this down with non-MW objects. Apparently a fairly high fraction of multiple populations in the LMC. Although the CMDs are nothing like the one above. Seem to be mostly age differences, and not the multiple main sequences seen in some MW clusters. Might very well be Na-O (and other) oddities though.
Steinn, this is probably a naive question. But do globular clusters contain dark matter
and do they affect the dynamics of globular clusters?
Also is there an explanation within the G.C. community of Figure 4 of 0806.2585
as to why the characteristic acceleration of globular clusters (assuming I am interpreting this plot correctly) is 10^-10 /sec ^2
Most globulars for which there is good enough kinematic data have no or negligble dark matter in them.
The exception is Omega Centauri, which seems to have some dark matter (seen in the outer region kinematics) and maybe a couple of others which are consistent with being tidally stripped dwarf galactic nuclei (I wanna say G1).
There are some intriguing models where most or all globulars are embedded in very extended dark matter halos, so the central DM density is low, but the total mass is large.
I was asked about the characteristic acceleration a few years ago - it is one of my oldest "unanswered" e-mails in my inbox.
I don't know how reliable that number is, and I don't have an obvious answer for the magnitude. Intriguing question that may either be profound or trivial. Not sure which.
Need to think about it.
Ok, thanks for the detailed reply.