Beyond Hubble

By Michael D. Lemonick/Mauna Kea

The sun is setting over the luxury resorts of Kona, on Hawaii's Big Island. Warm tropical breezes waft lazily through the palms. Honeymooning couples sip mai tais by the pool as waves break gently on white sand.

Just 30 miles inland, conditions aren't quite so pleasant. The sunset is every bit as gorgeous from here, at the summit of the long-dormant volcano Mauna Kea, but temperatures hover around 38[degrees]F, with a windchill that dips well below freezing. At an altitude of nearly 14,000 ft., the atmosphere carries barely half the oxygen it does at sea level, so the slightest exertion can leave visitors gasping. Those who travel to the summit without getting properly acclimated risk altitude sickness and even death.

But with night skies that rank among the clearest and darkest on Earth, Mauna Kea offers an unsurpassed view of the heavens--and that's why, despite the harsh conditions, astronomers can't wait to visit. Stargazers come here from around the world to answer some of the deepest mysteries of the cosmos: When in the depths of time did galaxies first flare into existence, and what made it happen? What is the elusive dark matter whose mass dominates the universe? How many stars have planets--and do those alien worlds harbor intelligent life?

These questions and more have tantalized astronomers for decades--and Mauna Kea is one of the few places where answers may finally be found. The mountain is dotted with white and silver observatory domes, sprouting like oversize mushrooms from the barren, rocky rubble that was once molten lava and, much later, a holy place of the native Hawaiian people. And although it's not obvious to the casual visitor, these domes conceal stargazing machines of unprecedented power.

For nearly a half-century, starting in 1949, the world's most powerful research-quality telescope was the Hale, on Palomar Mountain, in California. Its mirror, 5 m (17 ft.) in diameter, focused more faint starlight than anything else on the planet. But in the past few years, the Hale has been humbled. Here on Mauna Kea alone sit the Subaru telescope (no relation to the car), with a mirror more than 8 m (27 ft.) across; the Gemini North telescope, also topping 8 m; and the kings of the mountain, the twin Keck telescopes, whose light-gathering surfaces are an astonishing 10 m--33 ft.--in diameter.

The story is the same all over the world. In the high Andes of northern Chile, five more 8-m-class telescopes are either finished or nearing completion, while peaks in Arizona, Texas and South Africa too boast scopes more powerful than anything known to science just a decade ago.

That's not all. While each of these instruments trumps the Hale in light-gathering power, many are poised to outshine even the Hubble Space Telescope, which has been delivering astonishing snapshots of deepest space since it was refurbished in 1993. The orbiting observatory's nearly 2.5-m (8-ft.) mirror isn't all that powerful, but since it floats above Earth's constantly roiling atmosphere, the Hubble has been unrivaled in the sharpness of its images. No more. Using an ingenious technological trick to eliminate atmospheric blur, most of the new telescopes will soon achieve Hubble-quality focus--and even beat it under the right conditions.

This is a breakthrough of astronomical proportions. Whereas for years scientists have had only one Hubble-quality telescope, they will soon have access to more than a dozen. "What's been happening in the telescope game," says John Huchra, a veteran observer and a professor at the Harvard-Smithsonian Center for Astrophysics, "is incredible."

It has also been a long time coming. Impressive as the Hale telescope was for its day, it represented a technological dead end. The Hale, like its smaller predecessors, was powered by a mirror that's essentially a huge hockey puck of glass ground into a concave, light-focusing curve on one face and coated with reflective metal. To keep from sagging under its own weight and distorting the curve, the mirror had to be a bulky 26 in. thick, and it weighed 20 tons. That enormous heft called for an even more massive support structure to hold the whole thing up while at the same time adjusting constantly to counteract the effect of Earth's rotation. Scaling the design up any further would have been absurdly expensive.

During the 1960s, astronomers' lust for light was temporarily satisfied by the development of electronic light detectors. Because these detectors are up to 100 times as sensitive as photographic plates--the standard recording medium since the turn of the 20th century--every telescope on Earth saw its power boosted a hundredfold essentially overnight. That kept the scientists happy only for a while, however, and everyone agreed that telescopes needed some sort of radical new design. Unfortunately, says Matt Mountain, director of the Gemini Observatory, "nobody knew how to make the conceptual leap."

By the early '80s, though, telescope designers were leaping all over the place. University of Arizona astronomer Roger Angel's solution to the sagging-glass problem was to cast huge mirrors that are mostly hollow, with a honeycomb-like structure inside to guarantee stiffness. University of California at Santa Cruz astronomer Jerry Nelson opted instead to create a mirror not from a single huge slab of glass but from 36 smaller sheets that would, under a computer's control, act as one. And in Europe, design teams came up with yet another idea, the exact opposite of Angel's: instead of making the mirror hollow to save weight, let it be thin--about 8 in. thick for an 8-m mirror, in contrast to the 5-m Hale's 26 in.--and counteract the resulting floppiness with computer-controlled supports that continually readjust its shape.

"People argued at the time that it would be crazy to rely on computers because they might fail," recalls Mountain, whose Gemini telescopes in Hawaii and Chile were built on the European model. "But when you think about it, planes are controlled by onboard computers, and those computers essentially never fail."

Neither do the ones that run the telescopes. The European Southern Observatory's New Technology Telescope, built in the 1980s as a 3.5-m precursor to the Very Large Telescope (VLT), worked beautifully. So did Keck 1 when it went into operation in 1992. And so, in turn, have the other big telescopes as they've come online over the past two years. With both enormous size and smooth performance, these giant telescopes are doing science on a heroic scale--especially the Keck, which has had more than a half-decade head start on its rivals. In fact, says an astronomer who prefers to remain anonymous lest his outspoken views earn him professional enemies, "the Keck has done way more science over its lifetime than the Hubble."

He may be right. The Hubble's forte is taking brilliantly sharp pictures. But the real meat of astronomical discovery comes not so much in pretty photos of celestial objects but in the detailed analysis of their light. By smearing that light into a spectrum--the rainbow of its component colors--scientists can identify the chemical makeup of a star or galaxy, how far away it is and how fast it's rotating, among other data. If the image of a star is going to be smeared anyway, sharp pictures don't matter much, so ground-based telescopes are at no disadvantage.

So while the Hubble is good at locating faint celestial objects, the follow-up science is often done by observatories on the ground. In essence, the Hubble is like the small finder telescopes backyard astronomers use to pinpoint interesting objects for their full-size telescopes.

In many cases, though, the ground-based giants can find their own way through the universe. Geoff Marcy, for example, leader of the world's most prolific planet-hunting team, began his research at the relatively modest 3.5-m telescope at Lick Observatory in California. Then, in 1996, he moved most of his project to the Keck, with dramatic results. "We've discovered 35 planets orbiting sunlike stars so far," says Marcy, who holds joint appointments at the University of California, Berkeley, and San Francisco State University. "And the majority of them have been with the Keck."

The objects Marcy looks at aren't especially faint: he and his collaborators find planets by looking for stars that wobble under the gravitational tug of unseen companions. But the wobbles are so subtle that a lesser telescope can barely detect them. "With a 10-m telescope," says Marcy, "we can look at fainter stars and pick out the signature of smaller objects."

Indeed, Marcy announced last year that he'd found a planet the size of Saturn--the smallest yet discovered. "We think we can get down to the level of Neptunes," he says, "which are only 10 times as massive as Earth." Despite having so many planets in hand, Marcy and other astronomers haven't found anything like our home solar system: most of the planets found elsewhere are not only huge, but they career around in orbits that would fling smaller, Earth-like planets out into space--a discouraging start to the search for life in the galaxy, though it's far too early to give up.

George Djorgovski is using the Keck as well, but where Marcy's quarries are no more than 200 light-years way, Djorgovski's are closer to 10 billion. A professor at Caltech, Djorgovski has lately been concentrating on gamma-ray bursts--mysterious flashes of high-energy radiation that have baffled astronomers for nearly 40 years. If these blips of electromagnetic energy can be seen from far across the universe, as some astronomers believe, then they must briefly shine as bright as the rest of the stars in the universe put together--a seemingly preposterous assertion.

But in 1998, Djorgovski and his colleagues used the Keck to take visible-light pictures of a burst first spotted by the Compton Gamma Ray Observer satellite--and sure enough, it came from billions of light-years away. To date, the best explanation theorists have come up with is that the bursts come from "hypernovas," massive stars exploding with hitherto unsuspected power. "I feel really fortunate," says Djorgovski. "This was a world-class mystery, and the Keck allowed us to help solve it."

UCLA astronomer Andrea Ghez, meanwhile, has focused her attention on the center of our home galaxy, the Milky Way, far closer than Djorgovski's gamma-ray bursts but hundreds of times farther away than Marcy's planets. Shrouded in thick clouds of dust, the galactic core is invisible to ordinary light detectors. But among the Keck's suite of specialized instruments is an electronic camera sensitive to infrared light--the same kind of invisible light that your remote control uses to communicate with your TV. Infrared light of some wavelengths can penetrate dust as though it weren't there, giving Ghez a perfect view of the Milky Way's core.

Armed with the combination of the Keck's power and the detector's sensitivity, Ghez has been able to measure the motions of stars that lie 100 times as close to the core as the nearest star, Proxima Centauri, lies to the sun, and he finds that they're whipping around the galactic center at 1,600 miles per second, nearly 100 times as fast as Earth orbits the sun. It only takes high school physics to calculate that the object they're orbiting is as massive as 3 million suns yet packed into an area no bigger than the orbit of Mars.

The only thing that reasonably fits this description is a black hole, an object whose gravity is so strong even light can't escape from it. "We have evidence of these supermassive black holes in several other galaxies," says Ghez, "but this is the most convincing case we know of."

With their six-year head start, the Kecks have done more science than the newer telescopes, but the newcomers haven't wasted any time catching up. The European Southern Observatory's VLT, for example, built and operated by a consortium of eight countries, got the first of its four 8.2-m telescopes up and running in 1998 and achieved "first light" with the fourth in September.

But it's already doing first-rate science. Earlier this year, for example, astronomers from Sweden, Italy, Denmark and Germany used one of the scopes to help solve astronomy's so-called age paradox. In the mid-1990s, astronomers used the Hubble to measure the age of the universe at between 8 billion and 12 billion years. But other experts insisted they knew of stars that were at least 14 billion years old--obviously a problem, since stars can't be older than the cosmos. Using the VLT, though, observers have measured minute traces of radioactive uranium and thorium in the oldest stars--a technique akin to radiocarbon dating--and proved that they're more like 12 billion years old (the age of the universe, meanwhile, is now estimated at 14 billion years).

In fact, whereas the Europeans started later than their American competitors, they could pull ahead before too long. Not only do they have four giant telescopes on one site, but they've also budgeted more money than anyone else for state-of-the-art light detectors.

Still, U.S.-based telescopes remain ahead on several fronts, including the detwinkling of starlight. The technology that does this is called adaptive optics, and it was originally developed in secrecy by the Department of Defense to help military snoops take sharp pictures of Soviet spy satellites. Largely declassified in the 1980s, it's now being adapted for major telescopes everywhere. The idea is straightforward: stars and galaxies twinkle and shimmer because turbulent pockets of air act as weak, light-distorting lenses (heat rising from a car's hood or an asphalt parking lot causes a similar effect). With adaptive optics, though, a computer can measure the shimmer and cancel it out (see diagram).

Adaptive-optics systems do have limitations. To start with, they work well only with infrared radiation. That's not a huge problem, given that infrared is ideal for spotting new planets and for studying the early universe, the core of the Milky Way and the formation of stars. A bigger drawback is that adaptive optics can currently correct only for a small patch of atmosphere at the center of the telescope's field of view. But pockets of atmospheric turbulence are small enough that a slight change in viewing angle means a whole different pattern of distortions, which in turn requires a different pattern of corrections.

Even with these limitations, astronomers at both the Keck and Gemini have taken pictures that are every bit as clear as the Hubble's. Clearer, in fact, because a large telescope's images are inherently sharper than a small one's. Indeed, Ghez's latest and sharpest Keck images of the galactic center have been made with the adaptive optics.

Will adaptive optics make space telescopes obsolete? Not entirely. Space is still the best place to take supersharp pictures in ordinary light. And some radiation--ultraviolet, for example, and some wavelengths of infrared--can't penetrate the atmosphere at all. Moreover, telescopes radiate infrared light of their own, which contaminates celestial images. That's why NASA's plan to launch a Next Generation Space Telescope by 2009 still makes sense. With an 8-m mirror of its own, NGST will be able to see distant galaxies, for example, that no earthly telescope could ever see through the glare of its own heat.

Adaptive-optics systems may sound complicated, but they pale beside another technological trick that will ultimately boost telescopes' power even more. Called interferometry, it achieves the precise focus of a truly huge telescope without actually having the thing built. Instead, light is combined from widely separated telescopes--the two Kecks, say, whose observatory building was designed with a basement-level chamber for that purpose, or two or more of the four VLT telescopes in Chile. The system is dauntingly tricky and complex, but its astonishing precision will let astronomers tease out the details of galactic structures and distant solar systems as never before.

Yet even this remarkable technology could become obsolete--along with the giant telescopes on Mauna Kea, Chile and everywhere else--if the grandiose plans of the world's astronomers come to pass over the next couple of decades. Telescope designers are already thinking about the next generation of ground-based supergiant telescopes, devices that will range in size from 30 m (100 ft.) across to a staggering 100 m, or 330 ft.--a telescope mirror wider than the length of a football field. These will probably be scaled-up versions of the Kecks, using hundreds of individual mirrors aligned to make a single giant that could have up to 100 times the Kecks' light-gathering power.

Armed with a new generation of adaptive-optics systems now under development, these futuristic scopes will once again revolutionize astronomy. "When we were planning for the Keck in the early days," recalls Caltech's Djorgovski, "we laid out some of the science we expected to do with it. And we were much too conservative: we missed most of the really important stuff we've actually found. I predict the same thing will happen with these enormous telescopes. We'll almost certainly find things we never could have imagined."