Seeing red: in distant galaxies that shine with a ruddy glow are stars that look older than the universe that begat them

Natural History, March, 2005 by Charles Liu

To an astronomer, color is just as important as it is to an interior designer--though in quite a different way. To both, what the eye perceives as red is light of relatively long wavelength; the wavelength of the light the eye perceives as blue is relatively short. The designer, however, seeks the complex balance of wavelengths that, like the notes in a musical chord, gives a unified color tone to create a mood--a crimson, say, a scarlet, or a cardinal. The astronomer's colors are equally complex, but here it is the parts that are important, the individual, single-wavelength colors into which the spectrum of a distant star or galaxy can be analyzed.

The many colors of starlight, it turns out, can reveal a great deal about a star--including its age. Statistically speaking, long-lived stars emit more long-wavelength light than short-lived stars do. The most massive stars in the cosmos are also the bluest and brightest, and they tend to explode, as supernovae, after at most a few million years. Stars of lower mass and luminosity, however, glow dimly with red light for billions of years: the redder the star is, the older it is. So if, for instance, a large population of stars forms in a relatively short time, with a broad mix of stellar masses and luminosities, the combined light from those stars is relatively blue at first and then reddens gradually with age.

Astronomers have long exploited this correlation between age and color to study the ages of stellar populations in star clusters and galaxies. Time and again, such study has led to new and fascinating scientific puzzles. In the past few years, for instance, observations of numerous distant galaxies whose light dates to the earliest years of the universe have posed a bewildering paradox: some of these galaxies appear to be older than the universe that begat them. Yet there's no way these child galaxies can be older than their parent universe.

So why do we astronomers think these galaxies are so old? For one thing, their starlight is very, very red.

If the light emitted from a galaxy looks red, it's a safe bet that most of its stars are long-lived--and at least some of them are billions of years old. Moreover, if such a galaxy is also billions of light-years from Earth, the effects of cosmological redshift make the galaxy look redder still [see "A Desert No More," by Charles Liu, June 2004]. Such doubly red galaxies are known as "extremely red objects," or EROs.

Imagine observing extremely red objects between 9 billion and 12 billion light-years away. Among the nearer EROs, based on their redshift, are stars at least 4 billion years old, as indicated by their starlight colors. Among the farther EROs are stars at least 2 billion years old. Now do the math: stars in the closer EROs must have started forming 9 billion plus 4 billion, or 13 billion, years ago; the more distant EROs must have formed 12 billion plus 2 billion, or 14 billion, years ago.

That's the problem. The universe is about 13.7 billion years old, according to the best current measurements. How can distant EROs include stars that were born before the cosmos itself?.

Well, of course, they can't. There must be another explanation. One possibility is as clear as a city sunset. Particulates and dust in Earth's atmosphere along the line of sight tend to absorb blue light more effectively than red light. That's why the setting sun looks redder over a polluted urban skyline than it does over a pristine seascape. For the same reason, EROs might be made up simply of younger stars, heavily enshrouded by dusty gas. The dust would redden their outgoing starlight, thereby making them look like an older stellar population.

In some cases, at least, dust may be the solution. A recent study of 275 EROs between 6 billion and 10 billion light-years away, led by Leonidas A. Moustakas of the Space Telescope Science Institute in Baltimore, showed that there's no way to distinguish "dust-free old-star" EROs from "dusty young-star" EROs by looking at galaxy colors alone. Dust, the investigators report, could be causing the EROs' extremely red colors. Furthermore, their study demonstrated that EROs can take any shape, from the ellipsoid traditionally associated with old-star galaxies to the disky or irregular shapes typical of young-star galaxies.

Another recent study, however, led by Natascha M. Forster Schreiber of the Max Planck Institute for Extraterrestrial Physics in Garching, Germany, reached a different conclusion. Schreiber's group examined thirty-four EROs between 10 billion and 12 billion light-years away, comparing ERO colors with color models of starlight and dust reddening. They determined that dust alone cannot account for the extreme redness of many of the EROs in their sample. These galaxies, they report, really do include old stars, and the measured ages of the stars, though imprecise, are still high enough to make cosmologists sweat just a little.

Extremely red objects in the distant universe are reminders of how much astronomers still don't understand about the birth and aging of stars and galaxies. Everyone agrees that EROs probably play a big part in that story. The study by Moustakas and his colleagues, for instance, noted that the number of ellipsoidal EROs in the distant universe is about the same as the number of giant elliptical galaxies in the nearby universe. Are many EROs, then, juvenile ellipticals, destined to grow bigger? Just as children seem ruddier than their parents, some galaxies may show the same tendency until they, too, grow old.