Stars likely began with a bang, not a whimper

If Kenneth Lanzetta's observations about the rise of stars in our universe are correct, astronomers may soon be able to witness the ultimate blast from the past.

Drawing on data from thousands of distant galaxies taken with the Hubble Space Telescope, Dr. Lanzetta and colleagues suggest that the earliest stars burst onto the scene with an intensity unmatched by anything since. That picture stands in contrast to one painted six years ago by researchers who drew on Hubble data to suggest that star formation gradually peaked, then began an inexorable decline.

Lanzetta's estimate holds out the promise that, when a new generation of more powerful telescopes has been constructed - hopefully within the next decade - observers will be able to look back in time to see the Galaxy's brightest hour.

"We think on the order of 50 percent of the stars were formed in this early epoch," says Lanzetta, an associate professor of physics and astronomy at the State University of New York at Stony Brook.

Lanzetta's work, unveiled yesterday at a briefing at the National Aeronautics and Space Administration's headquarters, is the latest attempt to shed light on how the universe emerged from what Cambridge University astronomer Martin Reese dubs the cosmic 'dark age' to become the more luminous universe astronomers see today.

Dr. Reese notes that roughly 500,000 years after the Big Bang, the universe must have seen a remarkable burst of star formation after a period of extended darkness. This has prompted astronomers to try to determine what the starformation rates were like during the universe's first billion to billion-and-a-half years.

Star formation can be estimated from the amount of ultraviolet light a galaxy gives off. Ultraviolet light, however, can appear as visible light and even as infrared light with distance, thanks to an expanding universe.

But astronomers can adjust their calculations for this "redshift," yielding a UV estimate even for the most distant galaxies observable - if the light is bright enough to be detected by the current crop of telescopes.

In 1996, a team led by Piero Madau, then with the Space Telescope Science Institute, reported the results of its effort to estimate star formation rates by analyzing the deepest views of the universe the Hubble telescope had to offer - back to about 10 percent of the universe's current age.

Based on the UV brightness of galaxies in the Hubble deep field, the team found that stars burst onto the scene at an increasing rate that peaked at about half the current age of the universe. Then the rate tapered off. That peak period, the team concluded, gave rise to most of the stars astronomers see today. Subsequent studies by Dr. Piero and others revised the picture, suggesting that the star formation rate prior to half the age of the universe was high but fairly constant.

Yet Dr. Piero, now at the University of California at Santa Cruz, acknowledges that his team's '96 survey did not try to take into account the overall dimming effect the expansion of the universe has on light from distant objects. Depending on the assumptions one makes about galaxies and star formation at earlier epochs, that dimming effect could lead researchers to underestimate the starformation rate.

"We know so little about stars," he says. "It's very difficult to estimate the effect of changes in surface brightness with distant galaxies. It's difficult to correct for what you can't see."

Nevertheless, that is what Lanzetta and his colleagues have tried to do. Using Hubble deep-field surveys plus observations from ground-based telescopes, the team carefully measured the brightness of nearby 5,000 galaxies and derived distance estimates for them.

As the team analyzed the data, "We began to realize that there was an effect that had been neglected by previous analyses," Lanzetta says, referring to the dimming effect the expansion of the universe has on light from distant galaxies. It became important to quantify how much light they were "missing."

To estimate the missing light, "we looked at all the intensities of all galaxies at all different epochs and compared what we saw with some knowledge of what we could and couldn't have seen at any epoch," he says.

Based on those estimates, as well as on two other sets of observations, the team concludes that it's plausible that the farther back in time one goes, the higher the rate of star formation will be.

Dr. Madau notes that one of the challenges to this approach is the Lanzetta's team's "big assumption" that the brightness and distribution of light within galaxies does not change with time.

The test of Lanzetta's estimates will depend on more-detailed observational tests. For that, his team will have to wait for telescopes geared to infrared studies of fainter, more-distant objects - the next generation of space telescopes - to be built.