≡ Menu

Rethinking Stellar Populations

Back in April a paper appeared in the Astrophysical Journal that drew into question our view of star populations. We’ve assumed since the 1950s that we could count the stars in a particular area of sky by looking at the light from the brightest and most massive stars. In making this assumption, we were tapping the initial mass function, a way of describing the mass distribution of a group of stars in terms of their initial mass.

We could, then, estimate the total number of stars based on a sample of the stars that were the easiest to see, assuming that a set number of smaller stars ought to have been created in the same region. Every star twenty or more times as massive as the Sun should be accompanied, in this thinking, by about 500 stars of solar mass or less. But Gerhardt R. Meurer (Johns Hopkins University) and team used data from the Galaxy Evolution Explorer to challenge these proportions.

The numbers, it turns out, don’t work out as consistently as we had thought. Says Meurer:

“What this paper is showing is that some of the standard assumptions that we’ve had – that the brightest stars tell you about the whole population of stars – this doesn’t seem to work, at least not in a constant way.”

Many galaxies, this work shows, fail to form large numbers of massive stars, but continue to produce large numbers of their less massive counterparts. Indeed, for every massive star there may be as many as 2000 lower-mass ones, depending on the galaxy or region of sky being considered.


Image: Images taken with the Galaxy Evolution Explorer at shorter ultraviolet wavelengths are dark blue, while longer ultraviolet wavelengths are light blue. The optical images are colored red and yellow; red light is shown in yellow, while specially filtered red light from a type of hydrogen emission called H-alpha is colored red. By combining the data, astronomers were able to learn that not all galaxies make stars of different sizes in the same quantities, as was previously assumed. In other words, the proportion of small to big stars can differ from galaxy to galaxy. Credit: NASA/JPL-Caltech/JHU.

Ultraviolet images from the Galaxy Evolution Explorer were contrasted with filtered optical images from the Cerro Tololo International Observatory in Chile to obtain this result, the latter sensitive to the largest stars, the former to stars three times or more massive than the Sun. We thus gain a new way of estimating stellar populations, and are forced to re-evaluate galaxies that may be more crowded than we thought.

The paper is Meurer et al., “Evidence for a Nonuniform Initial Mass Function in the Local Universe,” Astrophysical Journal 695 (10 April, 2009), pp. 765-780 (abstract). For more, see this feature on the JPL site.


Comments on this entry are closed.

  • Don Elder August 21, 2009, 16:25

    So does this mean that since there may be way more stars than thought, then we don’t need that hard to prove dark matter theory?

  • Daniel Fischer August 22, 2009, 8:59

    The full paper is freely available here – as are most astrophyical publications these days: It’s actually rare when an important paper has not been put on the Arxiv server.

  • Mike August 22, 2009, 12:17

    Don Elder is thinking just what I was. How these new findings may influence dark matter theory. Also as low mass stars are considered to be much more
    congenial to support life the more the better.
    For our own Milky Way galaxy I think the stellar class ratios are pretty well
    understood but if someone can dispute that please do.

  • philw1776 August 22, 2009, 14:04

    “the IGIMF deviates from the canonical IMF down toMsun of just a fewMsun
    and that standard conversion factors between LH and SFR
    could underestimate the SFR by orders of magnitude in such
    cases. However, at the low mass end, below 0.8M sun, the
    IMF appears to be constant (Feltzing et al. 1999; Wyse et al.
    2002). More work needs to be done to determine where between
    the masses of O stars and the sub-solar range that the
    IMF switches from being variable to constant.”

    So depending on the galactic environs you get more or less O and B stars. They surmise that this may also hold true down to say F5 or thereabouts.

    And yes, I think our home galaxy’s stellar class % populations is reasonably well understood, excepting possibly the Brown Dwarfs. It’s extragalactic % stellar class differences they’re examining.

  • Mark August 22, 2009, 22:22

    I find it profoundly disturbing that we do not 1) actually know how many stars there are, and 2) understand the true nature of dark matter or for that matter dark energy.

    I think it is utterly creepy that this dark matter and energy is all around us, possibly pervading our space without our being able to detect it. I imagine I see it swirling around my room at night, taking on the shapes of hideously deformed monsters which, when the room light is switched on, are instantly revealed to be a bathrobe hanging on a doorknob, or a towel casually tossed over the back of a chair. Perhaps dark matter is nothing more than our imagination, a shadow in the darkness of our ignorance, waiting for someone –perhaps Einstein’s successor– to throw the light switch?

    I don’t know what to make of the star counting. I can’t imagine who would do this for a living, wouldn’t it be terribly tedious work? Or is it all computerized by now?

  • Tulse August 23, 2009, 11:13

    “I think it is utterly creepy that this dark matter and energy is all around us, possibly pervading our space without our being able to detect it. I imagine I see it swirling around my room at night, taking on the shapes of hideously deformed monsters ”

    Sounds like you’ve been reading Lovecraft, specifically “From Beyond”.

  • keith August 24, 2009, 8:13

    I wonder how this is reconciled with the Tully-Fisher relation? In this, the luminosity of a galaxy is correlated with its rotation speed very closely, at least within a certain mass range. Since the luminosity is very dependent on the few high mass stars, this new finding would seem to be discrepant with Tully-Fisher.

    I’m surprised that people think the IMF is “well-established in our own galaxy”. The RECONS project has been finding new M-class stars in our own neighbourhood right up to this decade, and probably more will be found yet. Having tried to research this topic, it’s quite clear that we have little idea how many stars there are in our own galaxy. However, it’s also clear we can’t explain the flat outer rotation curves with stellar mass, so it’s not the end of dark matter.

  • philw1776 August 24, 2009, 14:30

    RECONS has done a great job establishing what you question. As I stated, now (thanks to RECONS) we DO understand the M star population to a much higher degree of accuracy. Brwon Dwarfs are NOT well understood but they in any number don’t contribute to a galaxy’s luminosity.

    As to Tully-Fisher these studies will either bring greater accuracy or contrary, will weaken the relationship. Science is cool. The O,B,A populations of galaxies now cry out for other researchers to confirm or falsify this environmental contention.

  • Ronald August 25, 2009, 5:32

    See also out other very interesting thread here, ‘How Many Stars in the Galaxy’:


    Does this also imply that he number of stars in our MW galaxy could be some 4x as large after all or is this indeed well established thanks to RECONS and other surveys, as some have suggested above? This I think also depends on whether those results have been extrapolated to the entire MW galaxy or not yet.
    Fact is, as I have mentioned in the other thread, that, as the radius of the surveys increases, the % of red dwarfs decreases significantly, undoubtedly as a result of observational bias.

  • Ronald August 28, 2009, 8:27

    Please also see the above-mentioned thread ‘How Many Stars in the Galaxy’ where I finally present my promised check and analysis of stellar luminosity and color index (B-V) data, to verify the observed peaks in certain stellar spectral subtypes (G0, G5, G8 and K0).