This is a tale of three men who traveled a long distance to follow a star.

Carnegie Mellon University photo Carnegie Mellon University astrophysicist Jeff Peterson is one of three principal investigators on the radioastronomy project. Click photo for larger image. Map: Ulastai valley in China's Xinjiang province, in .pdf format

Sure, it sounds familiar, but this tale doesn't end with men kneeling before the Christ child.

Rather, this has to do with three astronomers whose quest landed them in a remote valley of northwestern China. Instead of gold, frankincense and myrrh, they brought with them 10,000 television antennas ---- the makings of a revolutionary radiotelescope that they hope will reveal the universe's very first stars.

The instrument, called the Primeval Structure Telescope, or PAST, was designed specifically to look for stars that astrophysicists suspect came into existence as early as 200 million years after the Big Bang.

Unlike today's stars, these were made exclusively of light elements such as hydrogen and helium and, some scientists speculate, likely were giants that burned out rapidly and violently. But just two years ago, no one even suspected that these stars might have existed so early in the developing universe.

"It's a fishing expedition," said Jeff Peterson, an astrophysicist at Carnegie Mellon University and one of the three principal investigators for the new telescope. "That's one of the fun parts of this project ---- we really don't know what we're going to see."

Peterson, who returned a couple of weeks ago from his most recent visit to China, is directing the project along with Ue-Li Pen of the University of Toronto and Xiang-Ping Wu of China's National Astronomical Observatories in Beijing. And he'll be off to the South Pole in February to evaluate it as a site for a possible future version of the telescope.

Since construction began in earnest in August, about 3,000 antennas have been erected in the Ulastai valley, which is south of Urumqi, capital of the Xinjiang province. All 10,000 antennas ---- providing 70,000 square meters of collecting area ---- are to be in place by next fall.

The impetus for the project came a little less than two years ago, when NASA reported results from the Wilkinson Microwave Anisotropy Probe, a satellite that scanned the sky to produce a detailed map of the cosmic microwave background, the so-called afterglow of the Big Bang.

The cosmic microwave background is the earliest light in the universe and was generated about a third of a million years after the Big Bang, when the hot soup of particles from that event cooled enough to form neutrally charged atoms.

Cosmologists suspected that atoms began to condense into stars or other objects sometime in the first billion years after the Big Bang. But the Wilkinson probe data was surprising because it indicated that these objects may have begun much earlier.

"There's quite a bit of uncertainty about what these objects are," said Don Backer, astronomy chair at the University of California, Berkeley, who hesitates to call them stars. They could just as easily be black holes, or maybe something else.

The mystery of the Dark Ages

	Carnegie Mellon University photo The PAST antennas -- just ordinary television antennas -- of the Primeval Structure Telescope stand silhouetted against the mountains in Ulastai. Click photo for larger image.

A number of groups now are rushing to figure out what happened in those first billion years of the universe -- an epoch known as the Dark Ages because of its supposed lack of light sources. Backer was at the National Radio Astronomy Observatory offices in Charlottesville, Va., last week to test several elements of an instrument he hopes to deploy, initially in West Virginia but perhaps later in Australia.

LOFAR, a project based in the Netherlands, similarly has announced plans to construct a large array of antennas. MIT's Haystack Observatory, once part of the LOFAR effort, is now proposing an array to be built with Australian collaborators.

Peterson and his colleagues opted for a fast, low-tech approach and hope to beat the other groups by discovering the first stars, or whatever these early objects might be. That could come as early as next month, Toronto's Pen said last week from Beijing, though only if everything works out right. It's more likely, he added, that they'll find something a year from now, when the telescope is fully up and running.

"I think that this type of astronomy will be going on for a decade," Peterson said.

He and Pen were only acquaintances two years ago. But at an astrophysics conference in Aspen, Colo., in the summer of 2003, the two discussed the Wilkinson probe data and wondered why, if the findings were correct, no one had seen any of these early objects long ago.

"There's no particular reason why the Dark Ages would be all that hard to observe at radio frequencies," Pen said. Hydrogen would have been the most abundant element back then and these early objects would have produced copious amounts of ultraviolet radiation when they formed. The UV light would have stripped electrons off of surrounding hydrogen atoms; this ionized hydrogen would produce a distinctive radio signal.

The longest wavelength of this signal, Peterson said, would be 21 centimeters, or 1420 megahertz. But these signals originated more than 12 billion years ago and thus are receding from us at 95 percent of the speed of light as the universe is continually expanding. This recession results in a so-called "redshift," a lengthening of wavelengths.

At these high redshifts, Peterson explained, the original 21 centimeter waves will have stretched to 1.5 to 7 meters by the time they would reach Earth. Those wavelengths, he added, are equivalent to frequencies of 50 to 200 megahertz ---- smack dab in the FM radio and VHF television spectrum.

So the reason no one has detected the weak signals from the universe's earliest stars is because they are swamped by broadcasts of "I Love Lucy," Monday Night Football and Howard Stern's conversations with strippers.

"You couldn't do [this type of astronomy] in Pittsburgh," Peterson said. Even his attempts to test the antenna technology from atop CMU's Wean Hall came to naught, thanks to the WRCT transmitter on nearby Warner Hall, not to mention WQED.

Good, cheap tech

The good news is that TV antenna technology is cheap and well-developed. Astronomers just need to find a quiet place to listen. Peterson, who has been operating telescopes at the South Pole for almost two decades, immediately thought of Antarctica. Toronto's Pen immediately thought of the remote Canadian Arctic.

Later that summer, however, Pen was in China for a scientific conference and began discussing the subject with Wu, from the Chinese observatories. Wu also had been pondering this problem and suggested China's western deserts as a site.

Western China is sparsely populated and its mountains provide a natural shield from radio and TV transmitters. The environment is not nearly so harsh as Antarctica and, in contrast to the expensive labor costs required to work in the Canadian Arctic, Chinese workers could be had for $5 a day.

And the director of the observatory, recognizing an opportunity for Chinese astronomy to shine, even put up seed money.

So a year ago, Peterson, Pen and Wu began a grand tour of the western provinces and settled on Ulastai, a valley with a Mongolian name that means "no trees." The barren valley floor could accommodate an array that eventually will cover three square kilometers. But the site also was served by both roads and the railroad.

The Chinese railroad has installed optical fiber along its right of ways, Peterson said, so the site had high-speed Internet access. Yet the Mongolian herdsmen who populated the area, riding their horses and camels, do not have computers that might interfere with radio signals. The astronomers have the only computers in the area and they've placed those in a shielded farmhouse.

The Chinese national observatories have provided about $3 million to build the telescope. The antennas are practically off-the-shelf TV antennas, built for less than $20 apiece. Other elements are literally off the shelf; signal splitters, used in Pittsburgh homes to connect one TV cable to two TV sets, are used "backward" to combine the signals of two antennas.

Unlike the giant dish antennas, such as the NRAO's Greenbank, W.Va., telescope, that radio astronomers typically use, the Primeval Structure Telescope antennas can't readily be moved. So they are all pointed at the celestial north pole, the one spot in the sky that doesn't move.

The antennas thus can continually gather signals from that patch of sky, allowing the telescope to look five times closer to the Big Bang than can the Hubble Space Telescope.

"Astronomy this sensitive has never been done," Peterson said, even within the narrow band of the radio-TV spectrum that historically has been reserved for radio astronomy.

In the Dark Ages of the universe, galaxies had yet to develop. But Peterson said he and his colleagues couldn't hope to see anything as small as a single star with the Ulastai telescope. It might be able to detect a billion stars, he suggested.

Pen said the easiest thing for it to see might be a quasar. Quasars, powered by black holes, are the most luminous objects in the universe, capable of radiating a trillion times more energy than the sun. Several quasars already have been found near the end of the Dark Ages.

But it's hard to know what to expect. The stars or objects or whatever they were in the early universe were composed much differently than today's stars, which were formed with debris recycled from earlier generations of stars.

For now, Peterson and his colleagues are concentrating their energies on getting their telescope assembled in Ulastai and gathering data.

"This is somewhat higher risk" than most astronomy projects, Peterson said. "But I think tenured faculty members should take risks. Who else in society can?"