Heaven Can Wait

A flawed mirror blurs the vision of the $1.5 billion Hubble telescope, imperiling its hopes of seeing to the edge of space and the beginning of time

The dozen or so space jocks hunched over their computers were a study in scientific calm. As the magnificent 43-foot-long Hubble Space Telescope (MST) orbited 380 miles above Earth, they controlled it as precisely as a jockey does a prize filly, using as reins nothing more than weightless radio signals. With one sequence this spring they tilted one of the telescope's mirrors and reoriented a lens in one of its cameras. The image of a galactic cluster 5,000 light-years distant looked like cotton balls--no worse than they expected from an instrument still being fine-tuned after its launch from the space shuttle in April. Early last month, with another radio barrage, they moved another of the telescope's mirrors and moved it again, and again. Six times--and still the picture of the bright star Iota Carina was blurred. Calm grew into anxiety. Repeated commands shifting the small mirror failed to sharpen the image. Finally, last week the horrified scientists deduced that the blurs were there to stay: the telescope had a bad case of "spherical aberration." Of all the complicated gizmos on the $1.5 billion craft, the flaw lay in a mirror hardly different from that in the first telescope, fashioned by Galileo in 1609. The Hubble, which had promised to give man a crystal-clear view of the heavens, was seeing the cosmos as if through a cataract.

Two days after the devastating announcement about the Hubble came another bad-news bulletin from the space agency. Because of hydrogen leaks in the fuel lines, which might have caused another Challenger-type tragedy, NASA was grounding the space-shuttle fleet indefinitely. That move doomed the best shot for fixing the Hubble: ferrying up replacement parts on a shuttle. The one-two punch to America's space program called into question yet again the nation's ability to carry out costly and expensive megascience projects, which at best starve other research and at worst fail spectacularly. The staggering setbacks come at a time when NASA is planning a small flotilla of space probes, many designed to be lofted aboard a space shuttle (page 52). The scientific community was ready to riot. Said Harvard University astronomer Clifford Stoll (author of the best-selling tale of computer hacking, "The Cuckoo's Egg"): "This is one of the worst things to happen to astronomy since the pope strung up Galileo." Politicians agreed. At hearings on the HST last week, Sen. Albert Gore Jr. excoriated NASA for having "eyes bigger than its stomach. We owe it to the taxpayers to find out what went wrong, how to fix it and make sure it never happens again."

That's up to the investigating commission, chaired by Jet Propulsion Laboratory director Gen. Lew Allen, that NASA quickly set up to examine the snafu. The panel, which will include non-NASA scientists, will begin by examining the mirror records impounded last week from the contractor and from NASA's own labs, and is expected to report back within three months. The investigation will center on the design, fabrication and testing of the telescope's primary mirror, a 94.5-inch, 1,800-pound ring resembling the shiniest Life Saver in the universe, and its 12.2-inch secondary mirror, which is a solid disc. The mirrors function as middle sprinters in a four-man relay. The primary mirror takes the light raining in from the heavens and hands it to the secondary mirror (diagram, page 50). The light then ricochets through the 26inch hole in the primary mirror. At last it burrows into an array of photometers, spectrographs and cameras poised to examine cosmic mysteries ranging from the source of stardust that forms living organisms to the very fate of the universe.

For the telescope that would see farther and more clearly into the black night of space than anything ever built, it was a stunningly symmetrical deformation, in keeping with all the superlatives: the most perfect flaw ever created by the mind and hand of man. Spherical aberration is common in telescopes. The defect means that light reflected from any two spots lying at different distances from the mirror's center do not focus at the same point. Light entering the Hubble that should narrow to a pinpoint covers a two-inch-round area. As a result, instead of capturing pictures of the cosmic vault as painted by Seurat--brilliant individual points of light from galaxies, quasars and nebulae--the Hubble produces blurry images more reminiscent of a de Kooning smear. Says space watcher John Pike of the Federation of American Scientists, "It's clear Murphy's law doesn't stop where space begins."

Public record: Almost as soon as the flaw was announced the search for culprits turned nasty. NASA threatened not to testify at Gore's hearing if the contractors also attended; mirror maker Hughes Danbury Optical Systems, a subsidiary of General Motors, conveniently declined the senator's invitation, saying it had not had time to analyze the focusing defect. It was a matter of public record, though, that Hughes (then known as Perkin-Elmer Corp.) had actually bettered some of NASA's specifications. The agency required that the primary mirror not deviate from a concave shape called a hyperboloid by more than one sixty-fourth the wavelength of neon light (.000005 inch); Hughes came 21.9 percent closer to perfect than that. Despite the opticians' success, some scientists suspect the mirror fault arose from the contractor's basic design. NASA issued only broad guidelines for the telescope's specifications, leaving to the contractor a design for achieving them. Speculated one congressional aide, "Somewhere there is a memo saying this sucker won't work. The challenge is finding the memo." The likeliest source of error: a mistake in the computer program that ran the glassgrinding machine at Hughes.

Whatever caused the defect, the real mystery is "how this level of trouble was not caught well in advance of launch," says astronomer Marcia Rieke of the Stewart Observatory in Arizona. One answer is that, in the early '80s, cutbacks in congressional largesse forced NASA to cut its budget for quality control by 70 percent and reduce the number of HST tests it ran. Oversight, too, may have been lax because of national-security concerns. Robert Smith, author of the recent history "The Space Telescope," notes that the Pentagon didn't want too many NASA people "crawling around [Perkin-Elmer], one of its contractors." One astronomer close to NASA told NEWSWEEK that with Perkin-Elmer's experience in building spy satellites es sentially telescopes that peer down rather than up--it could easily have tested HST in superior Defense Department facilities and perhaps would have caught the focal defect. NASA officials claimed that they never knew such mirror-testing equipment was available. That may well be, but it is also possible that the space agency resisted the military's help because of its fear, in the early 1980s, that the Pentagon was attempting a hostile takeover of the nation's space program. In the end, the testing equipment that NASA did use may have had the wrong shape and hence failed to pick up defects in the HST mirror.

NASA also tested the two mirrors separately. Agency officials maintained last week that it would have cost "hundreds of millions of dollars" to test them together in their intended configuration But many optics experts disagreed. Roger Angel, director of the mirrors lab at the Stewart, called the distortion so gross that "a test to detect it could have been conducted at almost no cost." Back then, the agency figured that since EST was designed to be as serviceable as a vacuum cleaner, with regular flights by the shuttle bringing everything from new cameras to wrench wielding astronauts who could haul the whole assembly back to Earth, it was a reasonable gamble to skimp on testing. The tragedy is that the flaw, if found, could have been corrected merely by repolishing part of the problem mirror.

Fine-tuning:The fiasco cripples the two cameras that promised to answer such questions as whether we are alone in the cosmos and where we came from--and answer them in pictures, no less. The Wide Field/Planetary Camera, designed to capture great swaths of the universe as well as close-ups of the planets, is hardly better than the best telescope still stuck on Earth. The Faint Object Camera (FOC), which had 60 or so projects, including a search for planetary systems that might harbor life, lost 10 experiments completely and 40 partially. "We had hoped to get a whole new view of the universe," says astronomer Sidney Wolff, director of the National Optical Astronomy Observatories in Tucson, Ariz.

The focal problem cannot be corrected from the ground; it is too large to be erased with the fine-tuning possible through radio signals. Replacing a mirror through a house call to space is also out of the question: although Eastman Kodak built a backup primary mirror. astronauts could never wrestle the glass into place in orbit. Returning HST to Earth risks contamination: a single speck of dust or fingerprint would ruin the mirrors.

Still, NASA held out hope that 95 percent of the science scheduled for HST would eventually, if belatedly get done. "You will marvel at the discoveries that are going to come from the observations Hubble is capable of today," NASA's chief scientist, Lennard Fisk, promised Congress. "Do not assume that Hubble is crippled from doing first-class, worldclass science." Indeed, some instruments are not hurt at all by the mirror defect. The spectrographs, for instance, analyze the wavelengths of light and don't depend on sharp focus. Sending a slightly fuzzy beam of light into a spectrometer says Bruce Margon of the University of Washington, is no different than using a cracked egg to make an omelet. And instruments that collect ultraviolet light will work as well as planned, receiving rays from intergalactic plasma and galactic nuclei and deducing what they're made of.

It may even be too soon to schedule a requiem for the Hubble's cameras. They might see clearly again with the help of something like a contact lens that would correct the smeared images. In a shuttle flight scheduled for June 1993, astronauts had planned to lift out the wide-field camera and replace it with a second-generation model; with the addition of a quarter-size mirror within the camera, the pinch hitter should see just as clearly and as far as if the mirrors hadn't been defective. Two other instruments were to have been replaced by updated devices sometime in the 1990s. (The Hubble has modular components, new instruments can slide in and out like tapes into a VCR.) "The secondgeneration instruments now become doubly important," says Stewart's Angel. "Any of the new instruments could be redesigned to compensate for the error." That includes Stewart's high-resolution spectrograph, scheduled for replacement in about 1996. By measuring the light spectra emanating from newborn stars, it promises to disclose their composition and thus their origins.

Instruments without scheduled replacements are simply out of luck. For instance, the high-speed photometer, which analyzes individual particles of light, will accomplish only half to three quarters of what astronomers planned. "We took a big hit," sighs Robert Bless of the University of Wisconsin, the principal investigator. He and his colleagues had hoped to discover exactly how much the universe's 800 pound gorillas--black holes--really weigh. These collapsed stars are so dense that not even light can escape their gravitational clutches. Bless hoped to glimpse the cosmic belches--pulses of light--emitted by a black hole gulping down matter spinning into it. The light pulses reflect the speed of the spinning matter, which is proportional to the mass of the black hole.

Realizing that such secrets would remain outside their grasp because of the Hubble's troubles, astronomers reacted with bitter disappointment and even rage. "Some people have spent virtually their entire careers on this project. It's a horror," says Marcia Reike of Stewart. Astrophysicist Edward Kolb of Fermi National Accelerator Laboratory in Illinois was especially worried about the effect on the vitality of astronomy: "Bright young scientists don't want to work on something that's not going to fly until they retire." Furious astronomers burned up the phone lines last week with electronic messages calling for a lynch mob to get whoever failed to test the mirrors before sending them into orbit.

Although some astronomers all along questioned whether money would not be better spent on dozens of smaller, Earthbased observatories, no one doubts the benefits of sending a telescope beyond the distorting atmosphere of Earth. The Hubble, in the clear emptiness of orbit, promised a wealth of discoveries as unlimited as the universe itself, for the telescope would act as an optical alchemist. It would transmute quasars, which appear to the most powerful telescopes on Earth as blurry patches of faraway light, into sharp galactic cores brimming with the first fire of youth. It would magically resolve stars perched at the brink of time and space, which to Earthbound astronomers look as fuzzy as a pussy willow, into a colony of blazing points of light. It would work this magic through a technique as ancient as the first human who turned his retina to the sky and allowed photons from a distant star to hit his optic nerve. It would collect light. For light is the ink in which the past, and the future, of the universe is writ.

Light emitted soon after the big bang continues to stream through the universe a cosmic courier bringing to earthlings the only news they will ever receive of events just moments after genesis. Light that began its journey 12 billion light-years (48 trillion miles) away by definition began to travel 12 billion years ago. Thus detecting light from far away is equivalent to seeing the universe as it was when young. Yet like any aging messenger, light that has been on the cosmic road for billions of years has lost so much of its power that astronomers require an instrument as sensitive as the Hubble to capture it. The Hubble can, for instance, detect light 50 times fainter than existing telescopes can. It therefore offered the hope of revealing how old and how big the universe is, what it is made of, how it will end and whether we are the only forms of life that have taken hold in an otherwise barren cosmos.

Unless other forms of life take pity on our technological frailty and reveal their presence, the search for extraterrestrial life will get no help from the faulty Hubble. The Faint Object Camera, which had only a slim chance of detecting planetary systems hospitable to life anyway, will never detect such a dim target with the blurry light now reaching it courtesy of the flawed mirror. "The chances for seeing planets before was small," says Duccio Macchetto of the European Space Agency. "I think now the chances are zero."

Perhaps the highest priority of the Hubble mission had been to determine the size of the universe. The wide-field camera was supposed to measure distance by focusing on a standard star called a cepheid variable. These stars have a constant brightness wherever they are in space. By measuring how bright they look from the Hubble, astronomers would have been able to determine their distance, much as measuring the brightness of a nine-volt flash light reveals how far away it is. "With out very good resolution," says David Schramm of the University of Chicago, "you can't pick out a single star in a distant galaxy because they all blur together."

Those who wonder about the stardust we're made of are in better luck. The only chemical elements cooked up in the big bang were hydrogen and helium; everything else, even the carbon that is the basis i of life, was baked in stars and then released in the blasts of supernovae. Astronomer Gordon MacAlpine of the University of Michigan hopes to train the Hubble's Faint Object Spectrograph (FOS) on the Crab I nebula, the remnant of a supernova, "to analyze the stuflf left over from the explosion." That might teach us what makes novae blow up and what elements they send into the cosmic void. The FOS might also reveal, by studies of dim stars in distent galaxies, whether the expansion of the universe is slowing down or holding steady. If the expansion underway ever since the big bang shows no sign of letting up, the cosmos will fly apart forever, eventually dissipating into nothingness. If the universe stops growing, however, it will begin to contract, rushing headlong toward a fiery big crunch.

Bizarre objects: If the bleary-eyed mirrors scuttle attempts to answer the big cosmic questions, at least the remaining instruments may reveal what bizarre objects we share with the night sky. Quasars, which burn with so much energy that they defy the laws of conventional physics, may be fueled by black holes, or by something not even dreamed of. They may be the kernels of nascent galaxies, for they speckle the universe at its farthest, and hence oldest, edges, where the first galaxies would have formed. Exactly how the 100 billion galaxies in the universe arose, and arranged themselves into the celestial mosaic observed today, remains a central puzzle. Scientists' best answer is that the kernel of space-time that exploded in the big bang was packed with galactic seeds in such a way as to throw them across space into just the pattern seen today. Astronomers prefer a less ad hoc explanation, and HST might have provided one. The most astonishing answers from the Hubble, of course, will be to questions not even imagined.

They may yet come. For as long as man has turned his eyes toward the glittering sky, he has sought in it answers to puzzles as practical as how nuclear fires burn and as spiritual as where one's ancestors are sleeping tonight. The stars have yielded some of their secrets, the Hubble would have teased out more. Those we expected to come rushing down in a rain of radio signals over the next few months will remain locked in the cosmos awhile longer. The heavens will still shine with the same mystery they do today; the universe will continue its headlong rush toward death by fire or ice. Yes, the heavens can wait. The greatest danger from the troubles of the space telescope comes not from postponing discoveries. The risk is that those who fund such research may think twice before again indulging the curious who, in the dark shadows of the celestial vault, would seek their origins and their destiny--and ours.

Cannot be replaced in space, and returning the Hubble to Earth is too risky. Verdict: irreparable

Astronauts can compensate for the flawed mirror by inserting new 'second generation' models

Scientists might use computers to counter some of the distortion. Computers can also sharpen the pictures sent back. Radio signals can move the secondary mirror slightly, but not enough to overcome the flaw entirely.

Join the Discussion