by Lee Billings
The shuttles had been conceived in the technological ferment of the successful Apollo missions to the Moon. As part of a larger twenty-year plan to build lunar outposts and send humans to explore Mars, NASA’s top brass had pushed for funding to develop spacecraft that could launch like a rocket, rendezvous in orbit with robotic satellites and crewed space stations, then glide back to Earth through the fires of reentry to land like a plane at spaceports around the world. Unlike the gargantuan Saturn V rockets that were used once then discarded on the way to the Moon, such a system would in theory be fully reusable, creating an economy of scale to reduce launch costs, which at the time were higher than $10,000 per kilogram. The space shuttle was sold as a revolutionary spacecraft that would fly again and again, perhaps once a week, making space travel cheap, frequent, and routine. The wide, wondrous expanse of the solar system would be thrown open to human curiosity and ingenuity. Lunar bases and human explorations of Mars would be only the beginning of an incredible journey to the stars.
Instead, President Richard Nixon balked at the projected expense of NASA’s grand plans. To ensure the agency would not pursue bases on the Moon and bootprints on Mars, he dramatically cut its funding and scrapped the Saturn rocket program. Perhaps because its focus was on making space travel more economical, the shuttle was the only piece of NASA’s plan to survive, albeit in diminished form. Its goal of full reusability was scaled down to a “semi-reusable” design that was cheaper to develop but more expensive to operate. Even so, the shuttle’s development required more money than a reduced NASA could provide, money that NASA found by appealing to the military’s need for launching and intercepting spy satellites. In exchange for its support, the Pentagon insisted on changes—a more spacious cargo bay, a heavier thermal protection system, and larger delta-shaped wings—all of which ramped up the shuttle’s complexity, cost, and risk.
The chimeric vehicles that finally emerged were elegant, versatile, and irreparably flawed. Instead of achieving 50 flights per year as originally projected, the entire shuttle fleet collectively flew 135 times during the program’s thirty-year lifetime. The shuttles lofted payloads to orbit at a cost estimated anywhere between $18,000 and $60,000 per kilogram—more expensive than the expendable launchers they were built to replace. The shuttle program’s failures came in part from the fact that many of its “reusable” components required extensive refurbishment by a small standing army of technicians after each flight. They also came from the shuttle’s inescapable operational risks, which led to the tragic losses of two orbiters and crews. Space shuttle Challenger exploded shortly after launch in 1986 due to a sealant failure in one of its boosters, and the Columbia disintegrated during reentry in 2003 after a piece of foam insulation punctured a wing. Politically driven compromises made early in the shuttle design process proved to be major factors in both disasters.
The total cost of the program has been estimated at $150 billion, with a similar amount spent on its signature achievement, the ISS, a massive orbital laboratory that the vast majority of Earth’s scientists did not want and could not use. For a time, the shuttles and the ISS collectively consumed nearly half of NASA’s total budget, all while offering only the slimmest fraction of scientific returns in comparison to drastically less expensive robotic exploration. The most useful research to emerge directly from NASA’s shuttle-era human spaceflight program only involved astronauts as test subjects, in experiments measuring the effects of prolonged space flight on human beings. Of course, the value of such research vastly diminishes without a capability to visit new destinations and perform meaningful work once off-world. Saddled with the financial strain of operating the shuttles and building the ISS, NASA saw its once-bold vision for a human future in space become a literal path to nowhere, in which astronauts looped endlessly around low Earth orbit waiting for their bones and muscles to waste away in microgravity. In nearly every aspect, the shuttle program was a ruinous white elephant that failed to deliver on its most crucial promises.
The lone exception was arguably the shuttle program’s role in the Hubble Space Telescope, a school bus–size robotic observatory that was delivered into low Earth orbit by the space shuttle Discovery in 1990. After initially being proposed in the 1940s by the American astronomer Lyman Spitzer, the Hubble had been conceived and funded around the same time as the shuttles, and had taken decades and upwards of $2 billion to construct and launch. Its snug fit in a shuttle’s cargo bay was not coincidental, as the telescope’s design was derived from some of the same spy satellites the shuttle had been built to carry. It was not the first space telescope, but it was for its time by far the largest, with a 2.4-meter primary mirror of aluminized, precision-polished ultra-low-expansion glass—an almost eight-foot-wide eye above the sky. Above the Earth’s atmosphere, Hubble’s big eye would not have to contend with shimmering layers of turbulent air that muddied and distorted celestial light. It promised to revolutionize all of astronomy with the unheralded clarity and sensitivity of its observations.
Except, once in orbit, Hubble began beaming back blurry images. Its mirror had been polished into an extremely precise but ever-so-slightly incorrect figure, so that it deviated from ideal curvature by some 2 microns, less than one-third the width of a human red blood cell. The polishing had been a laborious two-year process that could not be readily replicated in orbit, and there was no way to swap out Hubble’s mirror for a new one. Orbiting more than 550 kilometers (342 miles) overhead, the entire telescope was poised to be a multibillion-dollar boondoggle. That it instead became the most celebrated and productive observatory in history was due to the space shuttle’s unique capabilities, and the good fortune that Hubble, uniquely among all space telescopes past and since, had been designed with astronautic upgrades and repairs in mind.
In December 1993, a solution devised by NASA’s Hubble team was launched on the shuttle Endeavour. The space telescope would be given a set of smaller mirrors to refocus the starlight bouncing off the flawed primary mirror, rather like sliding a pair of glasses over myopic eyes. In the first-ever servicing mission of its kind, a crack team of seven astronauts clad in bulky spacesuits spent a total of thirty-five hours stretched across ten days installing the corrective mirrors and performing additional upgrades to Hubble. The marathon feat was rather like performing delicate eye surgery while wearing a welder’s helmet and oven mitts—all in an unforgiving environment where even minor equipment failures could kill in an instant. Within weeks, the space telescope was delivering images unmatched by any ground-based observatory. Shuttles ferried new instruments and equipment to Hubble four more times during its life, with the telescope becoming even more powerful after each visit. Critics of NASA’s human spaceflight program noted that for the estimated cost of each shuttle servicing mission, an entirely new Hubble could have been built and launched via expendable rockets, all without risking human lives, but even they couldn’t complain about the new transformative vistas accessed by the shuttle-borne upgrades.
Within our solar system, Hubble’s view was sharp enough to discern changing weather on Mars, explosive comet impacts on Jupiter, and Saturn’s ghostly polar aurorae. Targeting Pluto, its keen sight revealed new moons and crudely mapped the distant body’s surface. Gazing out to nearby star-forming regions, the telescope spied young stars cocooned in swirling disks of gas and dust that were midway through the process of forming planets. Measuring the motion of our nearest spiral galaxy, Andromeda, Hubble conclusively showed that the galaxy would collide and merge with our Milky Way in about four billion years. Inspecting surrounding galaxies, it found that nearly all of them had supermassive black holes at their centers, each squeezing the equivalent of hundreds of millions of Suns into a space smaller than the breadth of the solar system. Practically anywhere it looked, Hubble took gorgeous pictures that enthralled the public and enabled profound discoveries. In September 2012, astronomers released a “deep field” image a decade in the making, generated from two million seconds of Hubble o
bservations. From time to time, they had pointed Hubble at what looked to be a blank patch of sky empty of any celestial object. The patch was small, with a field of view less than what you’d get by gazing at the heavens through a soda straw. Hubble’s scrutiny revealed thousands of far-distant galaxies strewn like gems within that seeming void: yellow ellipticals, blue spirals, and jumbled “irregulars” in tangled skeins of color. The oldest and most distant galaxies manifested as small ruby points that gave no hint of structure—they had sent their light streaming toward us only a half billion years after the Big Bang, from stars that burned out long before the birth of our solar system. When the ancient photons arrived at Hubble’s mirror, each was some ten billion times fainter than a human eye could detect.
In the years that followed Hubble’s launch and repair, NASA expanded on the space telescope’s success by constructing three additional massive space-based “Great Observatories,” each devoted to a different wavelength of light and built at an average cost of about a billion dollars each. There was Compton, gazing at gamma rays from explosions at the edge of the universe. There was Chandra, watching with X-ray eyes as massive stars detonated as supernovae and supermassive black holes fed on molecular gas clouds. There was Spitzer, capturing the birth of stars and measuring the atmospheres of large, hot, transiting exoplanets in infrared light. All but one were launched by shuttles, and each added its own subset of breakthrough discoveries to what scientists soon began calling a “golden age” of astronomy. In addition to NASA’s four “flagship” space telescopes, the agency also built and launched an armada of smaller, more specialized observatories at costs of a few hundred million dollars apiece.
The high costs of the flagships and their accompanying fleet of smaller telescopes were still primarily caused by the staggering expense of reaching orbit—an expense that the shuttles had failed to reduce. A launch cost of tens of thousands of dollars per kilogram cascaded into greater expenditures in the design, fabrication, and testing of each telescope, which had to be made simultaneously lightweight and rugged, as reliably as technology could allow. At the time, the expenditures weren’t so worrisome—this was the America of the mid-1990s, a robust post–Cold War hyperpower with low unemployment, high productivity, a nation headed toward a trillion-dollar federal surplus, with a GDP and stock market soaring into orbit. Projecting forward, NASA’s leaders thought they saw a bright future in which the agency’s budget would steadily rise year after year, allowing the construction of even more-ambitious space telescopes, as well as sample returns from Mars and, ultimately, the rekindling of human exploration beyond low Earth orbit. After Hubble reached the end of its useful life sometime in the first or second decade of the twenty-first century, it would be de-orbited into the Pacific Ocean, and a new, even more revolutionary observatory would take its place. Hubble’s successor was announced in 1996 as the Next Generation Space Telescope before being renamed the James Webb Space Telescope (JWST) in 2002 in honor of the NASA administrator who had guided the agency in the glory days of Apollo. Its mission would be to fully unveil the universe’s very first galaxies, the objects that manifested only as tiny red blobs in Hubble’s deepest images. JWST would be only the beginning—the U.S. astronomy community rapidly made plans to pursue many additional big, ambitious space telescopes, like a hungry diner selecting not one but several gut-busting entrées from a menu.
Just as NASA threw its weight behind JWST, the field of exoplanetology began its meteoric rise. Astronomers could for the first time rationally discuss the possibility of finding other Earth-like planets, to considerable public interest and acclaim. Over interstellar distances, the planet hunters calculated, our world would appear slightly fainter than a typical galaxy in a Hubble deep field image. In theory, that would be something JWST could detect, and indeed the telescope would excel at imaging hot young gas-giant planets far from their stars. But in practice, habitable planets would lie far too near their much brighter host stars—the planned telescope would not possess the high dynamic range required to satiate planet hunters or their suddenly adoring public. Our own Earth, for instance, is some ten billion times fainter than our Sun in visible light—for every photon reflected off our planet out to space, our star blasts out ten billion more. In infrared, the contrast ratio becomes more favorable—at those wavelengths, the Sun is only some ten million times brighter than our world. Astronomers like to compare imaging another Earth around a Sun-like star to photographing a firefly hovering near a bright spotlight from a viewpoint thousands of miles away, but the simple reality is more powerful. To image a rocky planet around a star is to capture a dim fleck of dust practically hugging the cusp of a thermonuclear fireball, like photographing an unlit match adjacent to a detonating hydrogen bomb. To do that, you would have to somehow block out all those thermonuclear photons in their millions or billions so that even a single photon reflected from a planet could be seen. For nearly all the stars in the sky, the blurring interference of Earth’s atmosphere ruled out such precise measurements from the ground—only a space-based observatory could deliver the light of any potentially habitable planets orbiting many other stars.
At an early 1996 meeting of the American Astronomical Society in San Antonio, Texas, shortly after Geoff Marcy unveiled his team’s first discoveries of hot Jupiters, NASA’s then-administrator, Dan Goldin, took the stage to present an alluring vision of what the agency would do immediately post-JWST to support the search for other living worlds. Goldin intended to reshape NASA’s entire science program around astrobiology, with new life-finding space telescopes as the glittering centerpiece. “About ten years from now,” he explained, the agency could be ready to launch a “Planet Finder,” an observatory that would locate potentially habitable worlds and, via various starlight-suppression techniques, take their low-resolution pictures. It would look for atmospheric biosignatures in the spectrum of each small clump of planetary pixels. This was one of the earliest public mentions of what would become NASA’s “Terrestrial Planet Finder” (TPF) mission concept. If a TPF found promising worlds around nearby stars, Goldin told his rapt audience, then “perhaps in twenty-five years” even more ambitious telescopes could be built that could image those planets “with a resolution to see ocean, clouds, continents and mountain ranges.” Goldin laid out a not-too-distant future in which maps of alien Earths would grace the walls of school classrooms around the world, courtesy of America’s wealth and ingenuity. Sometime in the twenty-first century, he later said, those revealed as living worlds could become prime targets for robotic interstellar probes. By Goldin’s rosy estimates, a TPF might fly as early as 2006, serving as the predecessor to another observatory that would arrive in the early 2020s to begin practicing Rand McNally cartography on any nearby terrestrial exoplanets.
Unfortunately, JWST’s development proved more difficult than planned. To image the earliest stars and galaxies, the telescope would need a much larger primary mirror than Hubble’s, one optimized for the infrared wavelengths where molecular clouds, giant planets, and the earliest galaxies shine the brightest. It would also need to be cryogenically cooled so that its own internal heat would not wash out the fragile light of cosmic dawn. Finally, it could not operate in low Earth orbit, because the lightbulb-like glow of our planet at infrared wavelengths would contaminate the delicate observations. Over several years and iterations, a design was nailed down: JWST would possess a 6.5-meter mirror with nearly seven times Hubble’s light-collecting area and would be located at a point of stability between our planet and the Sun, almost a million miles from Earth, some four times farther away than the Moon. Nearly every aspect of the telescope would require major new technologies. A multilayered “sunshield” as wide and long as a Boeing 737 jet would protect the telescope and its suite of custom-built, state-of-the-art instruments and detectors. The entire assembly would be far too large for any existing rocket, so for launch the observatory would be folded up like origami, like a butterfly in a chrysalis, before unfurling in
the depths of space. To fold, JWST’s mirror would be divided into eighteen adjustable gold-coated hexagons, each chiseled from featherweight, highly toxic beryllium metal.
Various international partners signed on to construct instruments or to provide a launch vehicle, but NASA would bear the brunt of the cost, which early estimates pegged at approximately $1.5 billion. Launch was tentatively scheduled for circa 2010. As the project’s true complexity and scale became clear, cost estimates were revised ever upward, but little in the way of boosted funding appeared. Instead, money for JWST would have to come from NASA’s other space-science programs. In the end, more than $2 billion would be required for technology development alone. JWST’s schedule began to slip, ballooning the project’s total cost and shifting more and more of the major expenditures into the future. By 2012, JWST’s construction, testing, launch, and first five years of operation were estimated to cost nearly $9 billion, with a launch date no earlier than 2018.
JWST’s birthing pains were exacerbated by repeated national and global economic calamities, culminating in the Great Recession that began in 2008, in which the U.S. government spent trillions of dollars to prevent the total collapse of its major banks and other financial institutions. NASA’s budget, once projected to steadily grow, now was fortunate to simply stay flat, and even then did not keep up with monetary inflation. The trillion-dollar surplus built up in the 1990s under President Bill Clinton had become a multitrillion-dollar deficit in the 2000s under the tax cuts and runaway spending of his successor, President George W. Bush. After the Columbia shuttle disaster, Bush had delivered a bold new mandate to NASA that harked back to the agency’s original post-Apollo plans: build new heavy-lift rockets, then use them to return to the Moon and to deliver humans to Mars. It would be called the Constellation program. Alas, Bush did not deliver sufficient funding or deep congressional support, and scarcely mentioned the program after his initial announcement. As was typical of so many government projects begun during Bush’s administration, the only thing Constellation seemed to excel at was transferring billions of dollars of public, federal money into the coffers of well-connected private contractors who too often delivered precious little in return.
