Five Billion Years of Solitude

by Lee Billings

  In 2006, NASA chose to siphon billions of dollars from its science budget to prop up Bush’s failing plan, throwing JWST’s development into disarray and dashing hopes for a prompt development and launch of a TPF, which was officially “deferred indefinitely.” Not everyone mourned the loss—many astronomers not studying exoplanets had come to view the narrow focus and projected cost of a TPF as an almost existential threat to their less-glamorous subfields that also required space telescopes. Indeed, some had actively lobbied against it within influential study groups and planning committees.

  After years of middling results and more than $10 billion in expenditures, Constellation was canceled in 2010 by President Barack Obama, but the damage to NASA’s science programs had already been done. To adequately fund JWST, the agency was forced to scale back, delay, or cancel nearly all its other major next-generation astrophysics and planetary-science missions; if the observatory was to succeed, it could only do so at the great expense of effectively eliminating most of NASA’s space-science portfolio. As the prior generation of aging space telescopes wore out and broke down one by one, it seemed likely that whenever JWST finally launched, it would find itself almost alone, gazing out to the edge of the universe, the beginning of time, in a realm suddenly emptied of other major U.S. observatories. For want of money and strong institutional support, a TPF seemed almost as far distant and out of reach as the stars themselves. Congress repeatedly threatened to defund JWST due to the program’s incessant delays and overruns—there was a chance Hubble’s replacement would not fly at all. Even if it did, the telescope’s feasible lifetime was only ten years, after which its fuel reserves would be exhausted and its instruments degraded. Astronomers began murmuring that, with JWST, the golden era Hubble had initiated was perhaps drawing to a close.

  The thought weighed heavy on John Grunsfeld, a jovial, mustachioed astrophysicist and NASA astronaut who had flown on five space shuttle missions, including three to visit the Hubble. The telescope’s success had come in no small part from Grunsfeld’s space-suited dexterity, which he exercised in a record-setting fifty-eight and a half hours of spacewalking across his three Hubble servicing missions. The press hailed Grunsfeld as a hero, and called him “Dr. Hubble.” Riding the shuttle into orbit to repair the most productive space telescope in history, then using that same telescope to study binary pulsars and other exotic celestial phenomena, Grunsfeld directly experienced the powerful synergistic benefits that could exist between NASA’s human and scientific space programs. He mulled over the hundreds of billions of dollars spent on the ISS and the shuttles, and the comparative sliver of funding required to sustain the golden age of space telescopes. He wondered how, as with the shuttle and the Great Observatories, NASA’s brawny human exploration program might once again forge a powerful partnership with the agency’s purely scientific side, to great mutual benefit. During 2003 and 2004 he had served as NASA’s chief scientist and had helped develop science applications for Bush’s Constellation program. Big rockets, it turns out, are just as useful for launching extremely large telescopes as they are for hurling astronauts to the Moon. Such a rocket could, for instance, launch JWST without the expensive hassle of segmenting and folding the mirror. It could conceivably make larger TPF-style observatories cheaper as well. But the planning had backfired when Constellation cast its hungry shadow over NASA’s science budget.

  After completing the final Hubble servicing mission, in early 2010 Grunsfeld left NASA to serve as deputy director of the Space Telescope Science Institute in Baltimore, Maryland, the pulsing nexus of operations for Hubble and, someday, for JWST. For nearly two years, he worked closely with the Institute’s director, the astronomer Matt Mountain, laying the groundwork for a future TPF-style telescope that the Institute might someday manage, too. Their preferred in-house design was aptly named ATLAST, the Advanced Technology Large-Aperture Space Telescope, and was intended as an astronomical workhorse that would, among other things, deliver images of potentially habitable exoplanets. Dr. Hubble had become Dr. TPF, Dr. ATLAST.

  In his new role, freed from the stagecraft of being a high-profile NASA public servant, Grunsfeld spoke ardently and at length, often unbidden, about the importance and value of building new observatories to find other worlds and seek out new life. Then, in late 2011, Grunsfeld’s phone rang. It was a friend at NASA. The agency wanted him to come back, to serve as the associate administrator of its Science Mission Directorate—a role that would place Grunsfeld at the helm of the largest pure science budget in the world, albeit one struggling to meet all its myriad obligations. He accepted, and upon returning tamped down his past vocal advocacy for life-finding space telescopes in favor of a more carefully managed public persona that emphasized balance across all of NASA’s science programs. There was no announcement of bold new funding to search for alien Earths, but Grunsfeld’s closest friends and former confidants could not forget his former fervor. After nearly a year’s worth of fruitless e-mail exchanges with NASA’s press team attempting to secure an interview with Associate Administrator Grunsfeld for this book, I took solace in what Deputy Director Grunsfeld had freely told me once during an earlier interview.

  “Hubble and Webb will probably leave us hanging on the question of whether there is life elsewhere in the universe,” he said. “What we need in the next generation of great space telescopes—and what we can achieve—is the capability to observe the atmospheres and surface features of every single habitable planet around the nearest thousand stars. We could finally find out we’re not alone. We could finally find other habitable worlds, each of which, in principle, could be visited by humans. That’s the big picture, and I want to be able to convince the public and the Congress and the future administrations that it’s worth investing in this next step.” Clearly, Grunsfeld had read his Tsiolkovsky.

  • • •

  I visited the Space Telescope Science Institute on a cold, foggy morning in early 2012, less than a month after Grunsfeld had departed to take the reins of NASA’s science programs. The Institute occupies a nondescript building of tinted glass and dun-colored brick on the Johns Hopkins University campus and employs some five hundred scientists, engineers, and support staff. In the director’s office, amid glossy posters of stellar nurseries, scale models of space telescopes, and Hubble mementos flown back from orbit by the shuttle, Matt Mountain warmly shook my hand and, in keeping with his British upbringing, offered me tea. Mountain is middle-aged, a wry, owlish figure with searching eyes beneath a mop of sandy curls. His standard uniform—a dapper suit—looked baggy over a once-stout build made vestigial by “a ruthless doctor.” He had become director of the Institute in 2005, after spending a few years as JWST’s telescope scientist and a long, successful stint supervising the development, construction, and operation of the twin 8-meter Gemini infrared telescopes. When he spoke, it was with the brisk but calm cadence of someone well versed in elevator-pitching complex and expensive projects to impatient politicians and technocrats—powerful people whose busy schedules did not afford the luxury of long attention spans. Early on, he unpacked a well-polished quip.

  “In the twenty-first century, the discovery of life around another star will probably be as important a step for mankind as Neil Armstrong’s was for the twentieth,” Mountain articulated in plummy tones. “Finding life that has independently formed in another place, regardless of its progression or not to intelligence, will be like putting Copernicus and Darwin into the same bottle and giving it a good shake. What happens then? You look in the bottle. Maybe you revolutionize the world. I think this is a legitimate direction for NASA.”

  He produced an iPad from his desk and summoned imagery to accompany his words. “By the time we get to 2020, Earth-mass planets in habitable zones will be boring, you know. Because we won’t have a clue what we’re really looking at unless we have a spectrum.” He flashed an image of six boxes, each filled with a substantially different squiggly line. The lines were simulated atmospheric spectra
of the Earth at six different points in its geological history, from a lifeless atmosphere of carbon dioxide and nitrogen, to an Archean atmosphere filled with biogenic methane, up through the planet’s incremental oxygenation. “There’s only been a very small period in Earth’s history where the action looks like this,” he said, tapping the box containing Earth’s present-day spectral squiggle. “We know an awful lot about the nearby stars—where they are, how old they are. Most of them are actually younger than the Sun, so maybe an Earth there will look more like these earlier epochs. But to know, we will need spectra. And once you admit that, suddenly you’re in trouble, because it means you need a significant aperture in space—a telescope with a big mirror.

  “I’m going to really upset everyone and just say that suppressing starlight to one part in ten billion is just an engineering problem,” he went on. “Let’s just assume we are really smart and solve that. Well, we still need a high enough angular resolution to spatially separate a planet from its star. And we still have the problem that Earths are bloody faint, fainter than a galaxy in a Hubble deep field! We’re talking about something so faint, you can almost count its individual photons off on your fingers as they arrive at the mirror. It takes time to build up a spectrum, but in any reasonable mission we can’t spend more than a million seconds staring at one of these things, because the telescope will need to be used for other research, too, and we may need to search many stars to find what we’re looking for.”

  “How many stars are we talking about here?” I asked.

  “Ahh!” Mountain exclaimed, tapping at his iPad. The screen went black, then what looked like a slowly revolving cloud of sparkling ruby, topaz, and sapphire spun into view.

  “These are all the stars within two hundred light-years of the Sun,” he said, then tapped again. The rubies and sapphires disappeared, leaving behind only orange, yellow, and white orbs. “Here are all the stars like our Sun, the ones we think are most likely to be habitable. In a computer, we can put an Earth in the habitable zone around every single one of these and ask, ‘How many can we see with a telescope of a given diameter?’ A telescope’s resolution scales as its diameter, and its light-gathering power, its collecting area, scales by the square of its diameter. You need both to find something small and faint, so you put those factors together, and the number of candidates you can observe scales as the cube of the telescope’s diameter. With a 4-meter, you can get . . .” He tapped the screen, and nearly all the stars vanished, leaving behind about two dozen in a central core clustered about the Sun. “A few.”

  An 8-meter would bring hundreds within reach, Mountain said between taps. A diffuse shell of stars materialized around the small core. “With a 16-meter you get thousands.” He tapped a final time. The spinning swarm was shining, most of its original stars restored.

  “Keep in mind this assumes every one of the nearby stars has an Earth in the habitable zone. We now know from Kepler that’s probably too optimistic, and that somewhere between one and three out of every ten stars probably has a potentially habitable planet. We of course have no idea how often life gets started out there. So the question is, do you feel lucky? If you’re very lucky, you can build a 4-meter and get away with it, because one of the handful of stars you’ll look at will give you what you want. But what if you’re unlucky? If you don’t find anything in the nearest ten, well, you’re not sure what you’ve learned. You may have just been dealt a bum hand. If you look at the nearest thousand and come up empty, then for all practical purposes we’re probably alone. To have any rational chance of an answer, you should really look at hundreds of stars, and the physics of doing that thrusts you into this realm of 8-meter, 16-meter telescopes.”

  I nodded, picturing a large, flat, silvery disk slewing and tipping between targets in deep space, slowly building spectra from accumulated trickles of photons. That seemed simple enough. Why not go big and make it 16 meters wide?

  Later, I did the math and fully grasped the difficulty. A 16-meter mirror, more than 50 feet in diameter, would have slightly more surface area than a regulation singles tennis court—quite a lot to fold into a rocket. Even if it were fashioned from lightweight beryllium segments like JWST’s, the mirror and its support structure alone, sans instrumentation, would weigh more than 45,000 kilograms—about 50 tons—somewhat heavier than the Apollo spacecraft that had required a governmental crash program and the world’s biggest rocket to reach lunar orbit. Much of that weight would serve no other purpose than to ensure that the mirror’s micron-scale figuring could endure the intense vibrations and g-forces of launch as well as the frigid vacuum and zero gravity of deep space. A mirror to find alien Earths would require at least one of three things, each quite expensive: a rocket even larger and more powerful than Apollo’s Saturn V; piece-by-piece orbital assembly like that which built the ISS; or dramatic reductions in mirror weight, cost, and performance tolerances.

  “If you asked me today how much this 8- to-16-meter telescope would cost, I’d tell you I haven’t a clue,” Mountain said after I did ask. “I don’t want to go there. But that’s not the question you should be asking. The question is not how much does it cost, it’s what technology do you need to make it affordable? We must avoid the past mistakes of some other communities, like the particle physicists. Their entire field slowed right down in the 1970s and ’80s because it was asking for technologies that no one else wanted or needed. They had to build almost everything from scratch, component by component, and it got very expensive. To get this done, we need to look at our industrial base and stick with technologies that other people want.”

  During the Constellation program’s heyday, NASA had wanted a hulking rocket, the Ares V, a rocket so large it would surpass even the Saturn V. Shortly after departing the Gemini Observatory to become director of the Institute, Mountain got a phone call from Phil Stahl, an optical physicist at NASA’s Marshall Space Flight Center. Stahl was looking for science applications for the Ares V, and wanted to know how much the Gemini 8-meter mirrors weighed. Mountain recalled that he told Stahl the mirrors were roughly twenty tons apiece.

  “Is that all?” Stahl asked.

  “Is that all?!” Mountain was flabbergasted.

  “Matt, the Ares V will be so damn big we could just put a Gemini in it,” Stahl explained. Soon, the Institute was concocting plans for ATLAST.

  “That was the ‘holy shit’ moment, when we understood major progress in space science could occur simply by NASA getting new rockets,” Mountain recalled. “A Gemini mirror is big, stiff, and you can easily test it on the ground. It’s a much simpler system than something lightweight and segmented like JWST. A Gemini mirror only costs $20 million! You can just pack the whole thing into the rocket, light up the blue touch paper, and suddenly you have an 8-meter in space!”

  Mountain’s optimism for the “big rocket” approach had dimmed, however, with Constellation’s constant overruns and eventual cancellation. Out of the ashes, in 2011 Congress had assembled plans for a nearly identical rocket, the Space Launch System, but he was not convinced the replacement would ever fly. NASA optimistically estimated its development costs at nearly $20 billion. Operating the rocket over a program’s lifetime would probably require tens upon tens of billions more, yet it would most likely launch only once per year. Critics called it the Senate Launch System. Like its predecessor, it appeared to be a pork-barrel project designed less for affordable orbital transit and more for delivering jobs to the districts of influential members of Congress. The nascent rocket’s fate would be determined by politics, not science and engineering.

  As Constellation fell into its death spiral, Mountain and others at the Institute had sought a cheaper, alternate path to big mirrors in space. Perhaps there were ready ways to make large telescopes lighter and less expensive while also preserving their ultraprecise figuring. On the ground, newer observatories had decades earlier abandoned thick, rigid, monolithic mirrors for ones that were thin, flexible, and segmented. The new mi
rrors were cheaper but also floppy and easily deformed by shifting winds, changing temperatures, and a telescope’s pointing slews. The secret to their success was “active optics,” arrays of computer-controlled actuators mounted on each mirror’s back that performed what astronomers call “wavefront sensing and control.” Monitoring waves of light as they propagated across the mirror, a computer could manipulate the actuators to change the mirror’s shape and orientation, precisely counteracting any detected deformities. The JWST’s segmented mirror of beryllium hexagons already incorporated active optics to a limited degree in its design, allowing for periodic on-orbit adjustments on a timescale of days or weeks. To make larger mirrors that would still fit within the size and weight limits of existing launch vehicles, mission planners would have to use flimsier, lighter materials that would develop deformities even in the vacuum of space, requiring constant correction from complex active optics systems.

  Lightweight, flexible, large mirrors in space stabilized by active optics seemed to be a game-changing idea, except for one crucial detail: they were unproven. To Mountain’s knowledge, no one had ever flown such systems in space before. Developing and flight-testing the necessary technologies could prove very expensive and time-consuming, potentially eliminating the immediate benefit of using lighter mirrors. Mountain and his peers tempered their enthusiasm, until they noticed something curious: a shopping spree by major defense contractors such as Northrop Grumman and Lockheed Martin. In recent years the aerospace giants had bought up a string of smaller companies that specialized in manufacturing either lightweight mirrors or active optics systems.