Cosmic scaffolding. The "cosmic web" of dark matter as simulated on a computer. A new technique could enable scientists to detect much more of the enormous structure than before.

Nickolay Y. Gnedin, Nature, 435 (2 June 2005)

A New Way to Map the Universe

A new technique might soon enable cosmologists to map the universe even when they can't pick out individual galaxies. If it works, researchers would be able to probe the structure of 500 times as much of the universe as they have studied so far. "This is the first pioneering experiment that shows that this can be done," says Avi Loeb, a theoretical astrophysicist at Harvard University, who was not involved in the work.

Our galaxy resides within a huge "cosmic web" of dark matter, mysterious stuff that so far has revealed itself only through its gravity. Galaxies lie inside the dark-matter filaments, and scientists have made great strides in tracing this "large-scale structure" by mapping those luminous beacons. Astronomers working with the Sloan Digital Sky Survey have used a 2.5-meter telescope at the Apache Point Observatory in Sunspot, New Mexico, to map the location of more than 930,000 nearby galaxies, determining the distance to each by how much the expansion of the universe has stretched, or "redshifted," the wavelength of the galaxy's light.

Unfortunately, galaxies more than 5 billion light-years away are so dim that the Sloan survey can't spot them. In contrast, the afterglow of the big bang provides a kind of celestial wallpaper that's about 45 billion light-years away. So scientists have surveyed only 0.1% of the observable universe, Loeb says.

But that could change, thanks to work led by cosmologist Tzu-Ching Chang of Academia Sinica in Taipei and the University of Toronto in Canada. Instead of searching for the light from individual galaxies with an optical telescope, the team stalked a different quarry, red-shifted radio waves emitted by hydrogen atoms floating in huge clouds within the galaxies. That distinctive "21 centimeter" radiation can be traced with a radio telescope even if the scope does not have enough angular resolution to resolve individual galaxies, Chang says.

To prove it, Chang and colleagues trained the 100-meter Robert C. Byrd Green Bank Telescope in West Virginia on certain patches of the sky. Subtracting out radio signals 100,000 times stronger from our own galaxy and from television broadcasts, they detected the blurred 21-centimeter signals from galaxies about 6 billion to 12 billion light-years away. They couldn't quite resolve the spatial variations in the glow well enough to prove that they weren't just seeing some sort of random noise in their system. But they could tell that the variations they saw matched a map of galaxies made by a previous optical survey, as the team reports in tomorrow's issue of Nature. That cross-correlation confirms that the radio emissions do in fact trace the large-scale structure of the universe, Chang says.

To be useful in mapping the universe, the method needs to be refined so that researchers need study only the variations in the 21-centimeter radiation alone. "We’re working on that right now," Chang says. With a purpose-built radio telescope, the approach could map as much as 50% of the observable universe far faster and cheaper than galaxy surveys can, Loeb says.

In fact, Chang and colleagues are developing plans for just such a radio telescope. Chang says it would cost about $20 million, a tiny fraction of the $2 billion radio astronomers want for the proposed Square Kilometre Array (SKA) of radio telescopes, which aims to trace large-scale structure by locating individual galaxies. If the new technique works, then in that regard "it does somewhat obviate the need for an SKA," says Chris Carilli of the National Radio Astronomy Observatory in Socorro, New Mexico. But Chang is quick to point out that SKA has other objectives as well.