On April 24, 1990, the Hubble Space Telescope rode a space shuttle into low Earth orbit to become the most productive observatory in history. A quarter-century on, the universe may be the same but our understanding of it is not, forever transformed by the pristine ultraviolet, visible and near-infrared vistas revealed by the Hubble’s 2.4-meter mirror high above Earth’s atmosphere. Peering across the cosmos, Hubble mapped dark matter and helped discover dark energy, the mysterious force driving our universe’s accelerating expansion. Closer to home, it snapped pictures of giant exoplanets orbiting other stars, found new moons around Pluto and spied watery plumes bursting from the subsurface ocean of Jupiter's moon Europa. Almost everywhere it looked, Hubble made major discoveries. It became NASA’s premiere “flagship” observatory, and the agency supported it by replacing and upgrading obsolete components during five space shuttle servicing missions.

But Hubble’s time is running out; the space shuttles are no longer flying, and no more servicing missions are planned. Sooner or later Hubble’s crucial components will degrade and fail. Eventually its orbit will decay, turning the multibillion-dollar telescope into tumbling chunks of slag that burn in the atmosphere and splash into the ocean. Astronomers, ever hopeful, plan to continue using Hubble for several years to come, potentially into the 2020s, but know all too well that its days are numbered.

“When Hubble goes, it goes,” says John Mather, a Nobel laureate astrophysicist at NASA Goddard Space Flight Center. “And we don’t have anything else on the books that does what it does.” In particular, Hubble’s ultraviolet observations are crucial because Earth’s atmosphere filters out most of those wavelengths—they are accessible only from space.

Mather is also the senior project scientist for the project often billed as Hubble’s successor—NASA’s next flagship observatory, the James Webb Space Telescope. Once launched, Webb will soar some 1.5 million kilometers away from Earth to unfold a giant 6.5-meter segmented mirror. Operating at cryogenic temperatures and using a wealth of newly developed technology, the telescope will gaze out at the universe with exquisitely sensitive infrared eyes designed to study the very first galaxies and stars. That large size and technological complexity have come at a steep price, however. Originally targeted for a 2011 launch and an estimated cost of $1.6 billion, Webb’s price tag has swelled to nearly $9 billion and its launch is slated for no earlier than October 2018. In 2011 the observatory narrowly escaped cancellation when frustrated members of Congress threatened to terminate its funding.

Partially in response to those troubles, NASA’s post-Webb plans are to launch a far less ambitious observatory in the 2020s, a repurposed infrared spy satellite called WFIRST. Although Hubble-size, WFIRST will have a field of view a hundred times greater than Hubble’s, and will survey nearly the entire sky to study dark energy.

Both telescopes will be phenomenal tools for discovery, Mather says, and will open “whole new territories in astronomy,” but because they will not observe in ultraviolet or in most of the visible spectrum, they cannot be considered true successors or replacements to Hubble. Furthermore, neither Webb nor WFIRST is designed to directly address one of the most broadly appealing and fastest growing research areas in all of astronomy, the search for life on Earth-like exoplanets.

According to Mather and other leading astronomers now working on a report to be released this summer by the Association of Universities for Research in Astronomy (AURA), that quest and others require an even bigger space telescope that would observe, as Hubble does, at optical, ultraviolet and near-infrared wavelengths. Provisionally called the “High-Definition Space Telescope,” or HDST, the proposed super-size successor would have a mirror 10 to 12 meters in diameter—four to five times larger than Hubble’s, and roughly twice the size of Webb’s. Although neither Webb nor WFIRST are yet off the ground, the decade-spanning timelines required for planning flagship-scale space telescopes is forcing forward-thinking astronomers to consider these futuristic projects as things of the past.

Much like Webb, HDST would be segmented and deployable, so that it could be folded up and packed into the tight confines of a rocket. Unlike Webb, however, it would not need to operate at cryogenic temperatures, potentially mitigating some of its cost and complexity. The observatory could launch as soon as the early 2030s, the AURA committee says, but only if NASA and the U.S. astronomy community begin planning for it now.

“NASA’s gotten more conservative since we started the Webb,” Mather says. “After the telescope was nearly killed due to cost overruns, no one wants to think big anymore. But on the other hand, if everyone is too timid and terrified of big projects, then we’re never going to find those new worlds to explore. Columbus didn’t cross the Atlantic in a fleet of rowboats. Once in awhile, you need bigger stuff.”

One size fits all
This is not the first time astronomers have lobbied for such a large space telescopeVarious proposals calling for 10-meter-class space-based observatories have continually circulated through the community since the early 1990s, when the first exoplanets were discovered. Today, two decades later, the astronomers’ calls are growing louder, in part because thousands of exoplanets are now known, and more are thought to orbit as yet unseen around essentially every star in the sky. There are many conceivable uses for a gargantuan 10-meter mirror in space, but taking pictures of rocky, potentially habitable worlds—“direct imaging,” in astronomer lingo—is the killer app.

Viewed from Earth's vicinity an exo-Earth a dozen light-years away might be about as bright as a 40-watt lightbulb on the surface of Mars, but it would also be adjacent to its 10 billion-times-brighter star. A 10-meter mirror on something like the HDST would provide enough surface area to efficiently gather the faint light of dozens of such exoplanets as well as provide higher resolution to spatially distinguish them from the glares of their suns. HDST would also be equipped with a starlight-blocking device called a coronagraph, allowing it to more easily gather the faint photons from its planetary quarry. Astronomers could then examine each targeted world for spectroscopic signatures of water vapor, greenhouse gases and “biosignatures” such as atmospheric oxygen, which on Earth is produced by photosynthesizing plants.

Critics of ambitious proposals like HDST note that smaller, more modest space observatories could seek signs of life on a few potentially habitable exoplanets much sooner and for less money. In coming years NASA’s Webb and WFIRST telescopes will probably attempt such observations on a handful of the nearest, brightest promising exoplanets, as will a new generation of ground-based 30-meter-class observatories now under construction.

But those more modest approaches are unlikely to deliver the answers everyone wants, says Marc Postman, an astronomer and HDST report co-author at the Space Telescope Science Institute. Trapped beneath Earth’s ocean of air, ground-based observatories will be stymied by starlight-warping turbulence, and by airglow—faint light emitted by atmospheric chemical reactions that can corrupt delicate observations. Further, neither they nor Webb can directly image and investigate large numbers of potentially habitable exoplanets. “You really want to measure dozens of candidates, not just two or three, not even just 10,” Postman says. “If you look at 10 exo-Earths and see no evidence for water or biosignatures, you can’t say much. Maybe you got unlucky. But if you look at 50, even if you don’t find anything, that null result still says something very interesting about whether or not life is rare out there…. To cross that threshold, you need a highly stable, high-dynamic-range space telescope in the 10-meter range.”

Focusing on Earth-like exoplanets, however, did not pay off for previously proposed giant space telescopes. Those proposals were scuttled by their astronomical estimated costs, combined with the perception that they were boutique projects serving just one subsection of scientists rather than the astronomy community as a whole. HDST, the report’s authors say, will be different.

At a meeting of the American Astronomical Society in January, the AURA committee presented a grand vision for HDST as a revolutionary general-purpose observatory of broad appeal. Besides imaging exoplanets, the observatory would also excel at studying planets in our own solar system as well as far-distant galaxies and the webs of rarefied gas that flow around and through them. Befitting its “high definition” moniker, HDST’s huge mirror could capture features the size of Manhattan in the cloudscapes of Jupiter and track the motions of individual sunlike stars in galaxies up to 30 million light-years away. It could, in fact, resolve structures about 300 light-years across in galaxies on the opposite side of the visible universe—something useful for understanding the history of star formation as well as the nature of dark matter and dark energy.

All of these capabilities, the AURA team says, will synergize with the observations of Webb, WFIRST and the next-generation 30-meter observatories on the ground. And they can only come from building big.

Breaking paradigms—and banks
No one involved with AURA’s HDST report will publicly hazard a guess at just how much a telescope capable of all this would cost—only that it would be costly. The most difficult technological challenge will be building a lightweight, segmented mirror even bigger than Webb’s that is optimized for observing a wide variety of targets in ultraviolet, visible and near-infrared light. Skeptical of the financial feasibility of HDST, some astronomers suggest a smaller, Webb-size broadband telescope or even a series of Hubble-size telescopes would better serve the community.

HDST’s proponents say those ideas are shortsighted—too modest and incremental to remain relevant two or more decades hence, when any of these proposed telescopes would fly. “Someone once told me that, for any mission above a billion dollars, science is a necessary but not a sufficient condition,” Postman says. “You need scientists, policy makers, the public and industry to all support a mission of this scale. If Congress doesn’t like it, if NASA isn’t onboard, if the public finds it uninspiring, if industry doesn’t think it can be built—well, any one of those would be a big problem…. That’s why you have to think big, and consider the landscape that will exist in 2030 and beyond…. You don’t make revolutionary changes in our understanding of the cosmos by taking small incremental steps.”

As a result, proposing a giant space telescope like HDST is rather like walking a tightrope, balancing between controlling costs and making ambitious scientific and technological commitments. Tackle too few big questions, each needing its own set of advanced technologies, and your support from scientists and the public may dry up. Promise too much, and your political and industrial taskmasters will balk at the projected costs and difficulty or your telescope will go vastly over-budget.

The quandary is familiar to Alan Dressler, an astronomer at the Carnegie Institute for Science. In the early 1990s Dressler headed an influential committee that looked beyond Hubble to envision its successor, the “Next Generation Space Telescope” that would become Webb. Dressler and his team proposed building an infrared observatory with a four-meter mirror, the sort of thing that could snugly fit into a rocket without needing to be folded up. NASA's leadership insisted on being bolder and going bigger, pushing for an eight-meter mirror; Webb’s segmented 6.5-meter emerged as the compromise, although it was still far larger and more complex than any mirror ever flown in space before. Webb would require a host of entirely new technologies, and unprecedented amounts of exhaustive testing to ensure it would work as planned once it reached the cold vacuum of space.

At the time, Dressler recalls, a senior NASA official explained the agency “would rather build the first of the next generation of telescopes rather than the last of the previous one.” But now, with Webb years behind schedule and billions of dollars over budget, Dressler says choosing such a big, complex mirror for an already ambitious cryogenic telescope was “a bridge too far,” caused by “trying to make too much innovation in one step.” With Webb, NASA broke the paradigm of how space telescopes are designed, built and tested, but it also broke the bank.

Webb’s legacy has mixed implications for HDST, Dressler says. On one hand, the HDST proposal incorporates many of Webb’s innovations, potentially reducing the investments required for developing new technology. On the other, memories of Webb’s woes could prevent the HDST project from seeking out new paradigms necessary for building such a large space telescope. Instead of constructing and testing such a giant observatory on the ground, Dressler speculates, it might make more sense to launch HDST in modular pieces to be assembled in space by astronauts or robots.

“The trouble is, no one wants to wait the long time for that to happen for this particular telescope,” Dressler says. “They’re all sort of anxious to do this sooner, to do it in their lifetimes as scientists.”

According to Julianne Dalcanton, an astronomer at the University of Washington who is co-chair of the AURA committee, however Hubble’s super-size successor may someday come to be, what’s more important is for scientists and policy makers to realize how vital it is for astronomy’s future. “This is the obvious next step,” Dalcanton says. “And it will continue to be the obvious next step. The science won’t become any less compelling as time goes on. You can’t get around the physics. The things astronomers want to see are faint, and they want resolution. They want access to ultraviolet, which you can’t get on the ground. All that sends you to space, and makes you build big telescopes. The case for this will be just as relevant in 20 years as it is now. So why not begin to manage and plan for this and figure out where to make our investments and minimize the difficulties?”

There is, of course, one other thing besides physics driving HDST, Dalcanton admits. “It’s a lot of work getting another space telescope up—endless committee meetings, authoring reports. If I’m going to do that, I want the result to be really exciting. I would love to have a bunch of smaller telescopes like Hubble, and I would certainly use them, but I think we’re allowed to think bigger than that. And we should.”