Searching for ever-clearer images, astronomers develop a machine to take the twinkle out of starlight.

By BONNIE WALLACE
Illustrations by Bridget Keimel

MOST PEOPLE NOTICE TURBULENCE only in airplanes, at 10,000 feet when a series of jolts shakes vows from passengers to take the train next time, should they survive the flight. Those promises--and the turbulence--are safely forgotten when the plane's wheels hit the runway.

But astronomers often worry about turbulence, especially when standing on the ground in front of their telescopes. The reason for their concern is that the plane-bouncing pockets of hot and cold air disturb starlight, too. Turbulence warps light traveling through the atmosphere, making it waver on Earth and causing the pinpoints of starlight, which astronomers want to measure precisely, to twinkle into fuzzy dots.

That earthbound fuzziness annoyed UC Santa Cruz astronomer Jerry Nelson, designer of the world's largest optical telescope at Hawaii's Keck Observatory. But he couldn't do anything about the fuzziness until the Cold War ended a few years ago. Before that time, the military had been pouring money into classified research to solve the problem of atmospheric distortion, hoping not only to take detailed photos of orbiting satellites, but also, in the Reagan-era "Star Wars" project, to use lasers to burn holes into incoming ICBMs.

With the Cold War over, the military experts were finally allowed to talk about their top-secret work. Getting the new data was like Christmas for the civilian scientists. Astronomers had known the theory behind what's called adaptive optics for nearly twenty years, says Nelson, but now they had a chance to try it.

An experimental system of adaptive optics is now being constructed, and by 1998, Nelson and his colleages want to have a $7.5 million working model on the ten-meter telescope at Keck. The system will add basically two things to the telescope: a laser to measure turbulence and a flexible mirror that can reshape light hundreds of times each second through the night to compensate for the turbulence.

Getting around the atmosphere and the problems it creates won't be easy.

Skies above Keck are relatively calm, owing to the even temperature of the surrounding ocean. At an elevation of nearly 14,000 feet, the summit of Mauna Kea, Keck's mountain home, stands above about 40 percent of the Earth's atmosphere. Yet even there the stars twinkle.

Light rays from a star reach the outer atmosphere of Earth somewhat like a falling disk of dim light. The disk is parallel to the Earth's surface and to the telescope's mirror. If the earth had no air, that disk of starlight would sail on down to land flat and undisturbed on the mirror--that is, as much of the disk as the ten-meter-wide mirror could catch.

But the earth does have air, and Nelson says the plane of light "gets all ripply" going through it. The air contains cool, dense pockets that slow the starlight that passes through them. But those pockets slow some parts of the starlight in the imaginary disk more than others, so the disk crinkles like a ridged potato chip as it heads for the mirror, causing the astronomers to see a fuzzy image.

In sharpening the focus, the first half of the task of adaptive optics is measuring exactly how turbulence disrupts an image--that is, finding the shape of the theoretical potato chip of light. "If a star happens to be by, you can look at the light from the star," says Claire Max, one of the newly declassified physicists at Lawrence Livermore National Labs. "But if your favorite object doesn't happen to have a star nearby, you're out of luck."

Most astronomers want to look at dim and distant objects without nearby guide stars, so Max is helping them create an "artificial star" to measure the turbulence right where they're looking. For this task she will use a laser that emits a yellow shaft of light the same color as sodium streetlights. The laser beam will shoot up more than 60 miles into the air to make glow the sodium atoms left by micrometeorites. Of the elements up there, Max says, "sodium glows the best." The artificial guide star created by the energized sodium will be too dim to see with the naked eye, but any small telescope could detect it. It will "tell" the telescope mirror just how much to flex to correct for turbulence.

When light from the laser-made sodium star reaches the telescope, a filter will separate the laser light from the scientifically interesting starlight of the dimmer object behind it and deflect the sodium light through a grid of tiny lenses onto a display. That display will measure exactly how the light ripples each fraction of a second.

Once the shape of the rippling is known, it's time for the second step: correcting the ripple. The lens display will send instructions to a deformable mirror in the path of the light from the real star. Tiny pips will poke the rubbery reflector from behind, shaping an instant cure for the incoming uneven light. As the flexed mirror reflects the light from the star, it unwarps it and makes it a flat disk once more. The starlight then reaches the astronomers and the scientific instruments at the other end of the telescope, smoothed and twinkle-free.

The process sounds complicated, but a similar system is undergoing testing now at UCSC's Lick Observatory's on Mount Hamilton. "Lick is the guinea pig" for Keck, says Max. She's looking forward to making the Keck laser. "A lot of what's going to be interesting will be taking it from the stage where it's a curiosity to something everyone uses."

Whether everyone will use it remains to be seen, but Keck Telescope designer Nelson is cautiously optimistic. "Nobody's certain we'll be successful, but we hope it's the future," he says. If the system works well, he and the scientists and engineers he's working with hope to have another adaptive optics setup on the second Keck telescope. And he'll be able to stop thinking so much about turbulence, needing to worry only about the occasional bump in the air on his flights to his telescope in Hawaii.