How do astronomers get dates?

By Phil Plait | July 10, 2008 9:45 am

Q: How do astronomers get dates?

A: Ask a lot of heavenly bodies.

Go ahead and ogle NGC 6791:

Hubble view of cluster NGC 6791

Whoa. That’s an open cluster, a collection of thousands of stars that are (in general) gravitationally bound to one another. In reality, over millions of years, the stars interact with each gravitationally, and a lot of the stars get flung out of the cluster, becoming loners. But a large fraction of the stars stick around, aging and eventually dying while still in the cluster. They’re like city dwellers who never feel the need to leave town.

In this newly released Hubble image, you can see thousands of stars in just this one small patch of the cluster. You can see far more distant background galaxies, too (I love that kind of stuff).

But how old are these clusters, and the stars in them? Those are good questions, and important ones. The age tells us a lot about the environment of the cluster. For example, more massive stars tend to "sink" down to the center, and less massive ones move out away from the middle. How long does that take? The age of the cluster can tell us about how it moves around the Milky Way, and how stars behave in a cluster. All kinds of cool stuff can be figured out if we just know how long this guy has been around.

One advantage we have is that we’re pretty sure all the stars in the cluster formed at about the same time. Not exactly, but probably not off by that much. So if we can find the age of any of the stars, then we should know the age of all of them. Still, it turns out that’s not easy to determine. One way to is to look at the stars that have died already. We know that stars with more mass live their lives more quickly than low mass stars, eventually either exploding (if they are really massive) or blowing off their outer layers and leaving behind a white dwarf, a dense hot cinder.

So, if you want to date the cluster, look at the white dwarfs. Once formed, they don’t generate any more heat, so they simply sit there and cool off like a chunk of charcoal. We know how that works, so we can work backwards to get the age of the cluster.

Astronomers used Hubble to observe NGC 6791, a cluster that sits a little over 13,000 light years away toward the summer constellation of Lyra. They made that gorgeous image above, and looked for white dwarfs. They found a bunch, got their ages… and immediately had a problem: they got two different ages. Some of the dead stars appeared to be 4 billions years old (a little younger than the Sun), and others appeared to be 6 billion years old. Ouch. Worse, another technique used to get the ages of normal stars showed them to be 8 billion years old.

Uh oh.

Zoom in on NGC 6791 showing white dwarfs

In the image above, a zoom of the previous image, the younger appearing white dwarfs are circled in blue, and the older ones in red. Why would there be two separate populations of white dwarfs?

Well, there probably aren’t! It turns out that 13,000 light years is a long way off. The younger-seeming white dwarfs actually are binary stars, white dwarfs orbiting normal low mass stars, but they’re so far away from us they look like one single star (and it’s easier to date single stars than ones in a committed relationship). The light from the normal star changes the color we see, making us think the star is younger, when in fact it’s not.

So that fixes the 4 and 6 billion year issue; the white dwarfs are probably all 6 billion years old (why do stars want to look younger all the time?). But there’s still the problem that the normal stars in the cluster look like they’re 8 billion years old. Why would the dwarfs look younger?

Maybe they evolved differently than we expected while they were still alive. Maybe there’s something about the cool-down rates of white dwarfs we don’t understand. Maybe there’s something about the normal stars in the cluster that make them look older. It’s hard to say.

My suspicion is that the white dwarfs cool more slowly than we think. It takes them longer to get to a lower temperature, so when we look at them now they’re warmer than we expect, so we think they’re younger. What could do that? It might be that they have an odd chemical composition which affects their cooling rate (the presence or absence of some elements can affect how well a star radiates away its heat). I wonder if stellar encounters might play a role as well: stars are densely distributed in clusters, and there are more encounters between stars than out here in the suburbs of space. I don’t know how that might play a part… but it usually pays to look at the environment. How is a cluster different than other parts of the galaxy? More stars, more encounters, more binaries… somewhere in there is the key to the mystery of the discrepant cluster star ages.

Only by studying more stars and more clusters are astronomers going to get answers to these questions. Happily there are lots of clusters to observe, and lots of stars in them. So truly, I was right before: how do astronomers get dates? Volume.


Comments (38)

Links to this Post

  1. » How do astronomers get dates? | July 11, 2008
  1. firemancarl

    Wow, pretty cool stuff Phil. I like the backround galaxies in the first picture.

  2. firemancarl

    Oh yeah, what two galaxies are those in the backround?

  3. madge

    What a drop dead gorgeous picture! Great post too! Raises lots of interesting questions. Maybe all those stars are huddled together to keep warm so that’s why they lose less heat! Bless ’em : )

  4. Quiet Desperation

    Too many stars. Bad joke. Head hurt. Go play Boom Blox now. Oh, wait. I’m at work. :-(

  5. Keith Thompson

    Is there any chance that we’re looking at two clusters, one in front of the other? That would explain two sets of ages. Though I suppose it wouldn’t explain why all the white dwarfs are 6 billion years old, and all the normal stars are 8 billion years old. Unless one cluster is mostly white dwarfs and the other is mostly normal stars; is that at all plausible? Or might there be some systematic bias in which stars were measured? Or could the cluster have been formed by a merger of two clusters of different ages?

    Hmm. Any more straws I can grasp at?

  6. I think if I tried dating a star cluster I’d get burned… ok, that was a really bad joke.

    Now for a serious comment: This definitely helps sort out one of the things I think about before I go to sleep at night. Thanks for the great post!

  7. Thomas Siefert

    Engineers don’t get a lot of dates, we prefer quality over quantity. :-)

  8. Nicole

    Not to be prejudiced or anything, but isn’t 8 billion a little on the old side for an open cluster? I don’t think there are many in that age range, so that alone is pretty fascinating if confirmed.

    Multiple dating methods can be fun, but may lead to confusion. In clusters and in life I suppose!

  9. Todd W.

    Hmm…looks like a star curtain that someone hung a couple moving bits of cotton in front of to fake the galaxies. Clearly a hoax, especially with the supposed ages in there. Nothing can be older than about 6,000-10,000 years old. And the discrepancies in ages just demonstrate how flimsy the data are that astronomers are using. [/ snark]

    Beautiful picture, though I still think it looks like a star curtain for a theatre.

  10. Ha! While reading your description, the theme song from Cheers popped into my head.

    “But a large fraction of the stars stick around, aging and eventually dying while still in the cluster. They’re like city dwellers who never feel the need to leave town.”

    Ha! Just like the gang at Cheers!

  11. bo

    lol at todd w,

    that is EXACTLY what i was thinking.

  12. Curtis P

    How do we know that all the stars in the cluster are the same age? Is this based on a theory of how the stars form, or is it simply based on prior observation?

    My assumption is that, as a scientist, when you get conflicting information, you assume that the data is incorrect, but know there is a slight chance that the underlying premise is incorrect.

  13. Nicole

    I wonder about that, too, Curtis. I think that’s theory of how starts form in clusters, and observations of age-dating clusters. In the simple model in my head (yay) it’s difficult for a cluster to have stages of star formation separated by billions of years. The supernovae from the earliest big stars go off after some millions (10s of millions?) of years, and they can do a pretty good job of blowing any star-forming gas right out of the cluster. But there’s a chance this cluster is saying otherwise. Or just that the dating methods are incorrect (insert witty comment here).

  14. Hail the new font size! Hurrah!

  15. Tavi

    Very interesting article, Phil. I like your take on the white dwarfs’ cool-down rates.

    Even in its current state, Hubble has brought us so much. I can’t imagine how much better it will be after Atlantis’ long over-due servicing mission in October.

    I just ordered Robert Zimmerman’s, “The Universe in a Mirror: The Saga of the Hubble Space Telescope and the Visionaries Who Built It”, about an hour ago and am excited to be listening to a live interview with Zimmerman later this evening.

  16. “(and it’s easier to date single stars than ones in a committed relationship)”

    Now THAT’S funny!

  17. It seems plausible that a globular cluster could pass through a gas cloud, create some new stars, and capture some of them gravitationally. Or has this been totally ruled out?

  18. ShavenYak

    Just a layman’s thought, but it seems like with the stars clustered together, the outer layers blown off the dying stars would end up impacting the other white dwarfs pretty frequently, and that could possibly keep the dwarfs warm longer. I was also thinking tidal effects briefly, but I’m guessing the separation between any two stars is such that tidal forces aren’t big enough to have any measurable heating effect.

  19. TMB

    The problem with the different cooling rates is that the predicted rates seem to work pretty well in most clusters, so if their rates are different it would have to be something particular to this cluster (the idea of different elemental abundances is pretty good) rather than something generic to the fact that they’re in a cluster.

    As for multiple episodes of star formation, there’s two problems:
    1. The “old” stars are normal main sequence/giant stars, while the “new” stars are white dwarfs (and therefore started life as more massive stars). So you would have to have an old bout of star formation that only produced low-mass stars and then a new bout of star formation that only produced high-mass stars.
    2. An object that small can’t hold on gravitationally to enough gas to form new stars. The very very biggest globular clusters have multiple episodes of star formation (omega Centauri is the classic exxample), but they are much bigger and were probably even larger (ie. dwarf galaxy sized) when the stars formed, so they had much more mass to hold on to gas.


  20. Francis

    In the feedback to one of my classes, a student said the following:

    Q: What does an astronomer use as a contraceptive?

    A: Their personality!!

  21. Bruce

    Um, there is nothing to demand that all stars in an open cluster are of the same age. It may start out that way, if our understanding of stellar formation is anything to go on, but the gravitational “web” could possibly capture stars wandering by and other stars may leave to go home early. There may have been an interaction with another cluster over the course of the past little while. Is there another cluster nearby? And if so, what are their relative radial velocities and prober motions? Just asking.

  22. quasidog

    Top post BA. Learned something new today. :) Thanks

  23. good article. terrible joke though :)

  24. I feel like Peter Davidson should be looking out at me.

  25. BJ Nemeth

    It’s not only beautiful, but it’ll make a glorious background screen for my iPhone. :-)

  26. Walabio

    Ibid Says:

    July 10th, 2008 at 2:43 pm

    > “I feel like Peter Davidson should be looking out at me.”



    ¡The 5th doctor!

  27. OtherRob

    BJ Nemeth:

    I don’t have an iPhone, but that image is now the background screen on my computer. :-)

    I love those background galaxies. (After all of the images I’ve seen on this blog, I think I may be becoming a “Galaxy Man”.)

  28. I like this post, Phil! It meshes nicely with the podcasts I have been listening to of Richard Pogge’s Intro to Astronomy classes. I was just listening in my car earlier today about globular and open clusters.

  29. Josh

    I can count about 6, no…
    7 galaxies that I can see in that photo

  30. Josh

    And Before I forget:

    My god, it’s full of stars.

  31. Kevin McLin

    Hey Dr. Phil,

    I thought most open clusters are not gravitationally bound… that once the massive stars blow away the gas cloud that formed the cluster, there is no longer enough mass to bind them. Globulars don’t work this way, but I thought open clusters mostly did. There could be exceptions to this, I guess.

    Another point: does chemical composition really matter for the cooling rate of a WD? They are “black bodies,” no? So if the energy doesn’t come out in the blue, say because of absorption by some metal lines, then it should come out someplace else, probably the red.

    Just thinking out loud here…


  32. That picture is giving me a hard on!

    HEHE! If there is a God, I bet this is his stack of Playboys! HAHA

  33. StevoR

    Sayeth the BadAstronomer :

    “But how old are these clusters, and the stars in them? Those are good questions, and important ones. The age tells us a lot about the environment of the cluster. For example, more massive stars tend to “sink” down to the center, and less massive ones move out away from the middle.”

    But wait a second .. What if the more massive stars actually form in the middle of the cluster where most of the gas cloud’s mass is concentrated instead? I can sort of understand the massive stars falling inwards too but arewe sure there aren’t many formed at the heart of these clusters to begin with?

    Eg. The Rosette nebula star cluster (in Monoceros) – as I grok it has
    a heap of massive stars at its core and lighter ones further out but is
    extremely young – maybe even still forming? Not 100 % sure with that exact
    example but do you get my drift?

    Anyway another puzzler like the mystery planetary with the two young A-type stars that shouldn’t have been there – good thought-provoking article Thanks Dr Phil Plait

    BTW. Did anything more come of that puzzling planetary investigation?
    (“One ring” ? thread title I think.)

  34. Ronn! Blankenship

    Phil asked:

    Q: How do astronomers get dates?

    A. We seldom do because it’s hard to find women who know what we are talking about when we ask if they want to go to a movie that starts at JD 2454659.53125 . . .

  35. quasidog
  36. TMB

    Kevin: White dwarfs aren’t perfect blackbodies… here’s a typical spectrum of a white dwarf:
    (first link in google images for “white dwarf spectrum”)

    The real white dwarf has all of the Hydrogen absorption lines, while the black body is overplotted. So it’s not *that* dramatically different from a blackbody, but the spectrum still depends on things like composition and magnetic field strength.

  37. Torbjörn Larsson, OM

    Q: How do astronomers get dates?

    A: Ask a lot of heavenly bodies.

    Oh, for the love of … babes.

    Q: How do physicists get dates?

    A: Multiple dating methods. [“Multiple” is funnier than “various”, hat tip Nicole.]

    Q: How do engineers get dates?

    A: By way of binary addition.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!


See More

Collapse bottom bar