Monday, October 09, 2006

Astronomical Detective Work - Pinning Down the Stragglers

Recently, I've been spending a lot of time posting on bad science. Any time I've been sufficiently interested in something enough to consider writing a post on it, it seems Phil Plait beats me to the punch. That rascal.

But today something caught my eye that's not something that really deserves comment, but is still sufficiently interesting that I feel like posting about it.

For those of you that are new to this blog, I spent my summer in sunny San Diego doing some work on researching an open cluster (NGC 7142 to be specific). While doing so, I had to do a bit of reading on clusters, learning their properties and some of the peculiarities about them. One of the things that made NGC 7142 unusual was that it had some features in common with what are known as globular clusters.

Globular clusters are typically much older than open clusters, and as such, don't generally have nearly as many blue stars, which are more massive and die off quickly unless there's a source of new ones being made. Since globular clusters lack the gas to form new stars, the standard model for these clusters predicts that we shouldn't see any blue stars.

But we do.

Thus, astronomers have needed to form a model for these "blue stragglers". There's been two main theories behind it thus far.

The first was that, since globular clusters are pretty tightly packed, perhaps stars collided. In that case, it would form a new star. In the process, it would make the resulting star more massive (bluer), as well as mixing material throughout the star.

The second predicted that, perhaps stars didn't collide, but just came fairly close together. If one of them was expanding into its red giant phase, it could pass material from its photosphere to the nearby companion, again adding fresh mass and making the star bluer.

So the problem is that there are two competing theories. But to be really good science, we now need to test each model against some real world obesrvations and see which can hold up to scrutiny.

Finally that task has been achieved.

By looking at the spectra of some of these blue stragglers in the globular cluster 47 Tucanae. What the group found was that six of the blue stragglers observed had substantial carbon and oxygen depletion in their spectra. This strongly supports the second theory I mentioned above, involving a mass transfer. The reason for this is that, if the straggler was the result of two stars colliding, hydrodynamic simulations show that there shouldn't be much mixing and that concentrations of elements should remain about the same as other stars in the cluster.

However, if the straggler formation was a result of a companion star piling on mass, it would pile on mass from its outer layers, which would add a fresh layer of hydrogen to the "surface" of the star and mask the presence of the carbon and oxygen which would get buried.

But wait! There's more! If these stragglers really were formed by mass transfer from a close binary, it might be possible to check to see if the star is in a known binary system. It turned out that half of the carbon and oxygen depleted stragglers are known to be binaries. The other three aren't known to be, but determining whether or not a star is a binary in such a crowded field of stars relies on some tricky methodology that doesn't always work depending on the orientation of the system.

So here we have two competing theories (to be more accurate, I should say hypothesies), both of which make testable predictions. When one is tested, it is empirically confirmed.

But does this mean the other is wrong? Unlike what the ID crowd would like you to believe, science isn't an all or nothing venture. As the authors of the paper state, it's entirely possible that each one contributes in some way.

In fact, the end of the paper suggests that it's quite likely that this is the case. It also turns out that a few of the stars they observed are a class of stars called "W UMa" stars which are thought to be close binary stars that are slowly spiraling into eachother. So, perhaps the blue stragglers first arise through a mass transfer before the stars finally merge, but are eventually merged stars.

All in all, it's an interesting article. It's also a nice reminder that science is still very much an ongoing process. There's a lot we know, but a whole universe left to figure out.

4 comments:

John A said...

Can I ask some obvious questions (to me at least)? Why do globalar clusters exist at all?

Why do globular clusters manage to keep so many stars in such close proximity for so long even though they are composed of massive stars that should have gone kablooey long ago?

How do globular clusters keep so many stars in such close proximity when they should shatter as stars get flung away and others coalesce?

mollishka said...

1. Difficult question, but: globular clusters are believed to exist because it's simply how stars form... you start with a large cloud of gas, which has about the same mass as a globular cluster. It then cools and condenses, and individual stars form when individual lumps of gas become dense enough to start deuterium burning. So, if there's nothing to disrupt the cluster, then the initial large cloud will remain fairly intact, but as a cluster of stars instead of as a cloud of gas.

2 & 3. Not all of the stars are really massive and die young. The less massive stars live for a really long time. When a star explodes, it's really only going to affect those stars very close to it (like a close binary companion). The other stars will be affected like rocks in a pond: the explosion will just wash over them.


You might also want to keep your eyes open for a Nature article coming out soon-ish (probably) about the brightnesses of stars coalescing ... not as bright as supernovae, but also much brigher than the brightest single-stars.

Jon Voisey said...

Way to beat me to the punch. I was behind in grading today and wanted to get it done before my lab, so I wasn't able to get the time to respond.

But I'll just quickly add that there's also evidence that globular clusters are slowly stripped of stars as they orbit the center of the galaxy. Each time it passes through the disk, some of the outer stars are pulled off and trails of such stars have been observed in some instances if I recall correctly.

And thanks for the heads up about that paper. It sounds interesting.

Generally stars by themselves aren't my thing. I don't really get exicted until you get a good number of them hanging out together.

Stephen said...

If you have a star in a binary accreting matter from the other, shouldn't it spin up? That would be detectable by looking at how 'wide' the spectral lines are. You get some red shift and blue shift in the one object from the spin.

I haven't read the paper.