Mirror Earth (23 page)

Even in what amounted to suspended animation, however, the few scientists and engineers still working on TPF have continued to make progress. I was in Berkeley one evening for an observing run in the basement of Campbell Hall, the physics building, where Geoff Marcy was using the Keck II telescope to do radial-velocity follow-ups of Kepler candidates. When he and Paul Butler had first begun using the Keck, in 1996, they had to fly to Hawaii, make their way up to the telescope, in the cold, thin air at the summit of Mauna Kea, and try to stay awake as they fought jet lag and oxygen deprivation. Nowadays, with a super-broadband Internet connection, all of the adventure and romance is gone. I haven't yet heard anyone complain.

The observing run wouldn't start until close to midnight, so I was killing time in the hotel lobby when in walked David Spergel, the head of Princeton's astrophysics department and one of the original members of Princeton's TPF collaboration. He'd come out from Princeton for a dinner celebrating the career of his graduate school mentor, Leo Blitz. The dinner was over and he was headed for bed, but he had a few minutes to talk about his own work on TPF. “It's not completely dead,” he said. “There's something like one hundred or two hundred million dollars in technology development available.”

In fact, the Princeton TPF project, whose genesis had come during the conversations that engaged Sara Seager so deeply in the early 2000s, had turned into two parallel projects, one focused on the coronagraph idea and the other on the occulter, also known as a starshade. The first was located in the university's school of engineering, where the Princeton TPF group's principal investigator, Jeremy Kasdin, explained both projects during a visit by Michel Mayor in the fall of 2011. The coronagraph arm of the research, said Kasdin, was still using the “pupil” idea Spergel had come up with earlier in the decade, which didn't blot out starlight so much as shunt it off to the side of the image, letting a planet become visible.

The pupil had evolved from a simple cat's-eye shape to a mask with a set of elaborate, curved openings, which turn out to be even more efficient at shunting light away from parts of the image to let planets shine though. “How do you come up with these shapes?” asked Mayor. “Is it just intuition?” Not exactly, said Kasdin. “Usually, I'll get out of the shower with a great idea, and I'll call Bob.” Bob is Robert Vanderbei, chair of
Princeton's Operations Research and Financial Engineering Department, and an expert in “optimization.” That's a field of mathematics/computer science that lets you adjust anything from an investment portfolio to an airplane wing for maximum performance. Vanderbei is also an extraordinarily talented amateur astronomer and astrophotographer, so it's not surprising that he found his way into Kasdin's group. When Kasdin calls Bob, he's looking for help in refining a new idea for a pupil shape into something that reduces light in an optimal way.

The lab where Kasdin and his group tests out new pupil shapes is located inside the university's engineering building, while the lab where they test the second version of TPF, which will use a free-flying occulter, is about three miles away, on the university's satellite Forrestal Campus. Not far from the occulter lab is the Princeton Plasma Physics Laboratory, where scientists and engineers have been doing experiments in controlled nuclear fusion since the late 1950s. That bears mentioning because the man who founded the fusion-energy lab, Lyman Spitzer, was also the father of the Hubble Space Telescope, which he first thought of in the late 1940s. (Spitzer didn't get his name on the Hubble, but he's memorialized by the infrared Spitzer Space Telescope, which Dave Charbonneau now uses to look for planetary atmospheres.) Spitzer also worked on planetary-formation theory early in his career—and to top it off, he wrote an early proposal for free-flying occulters back in 1962.

The occulter Princeton is working on, in partnership with NASA's Jet Propulsion Lab (JPL), Kasdin explained, would
be 40 meters, or about 130 feet, across; it would fly about forty thousand miles away from the telescope itself, and it would have to maintain its position to within about two feet. The occulter wouldn't be just a round disk, but would rather (because of diffraction, and thanks to optimization analysis) look something like a flower, with twenty stubby petals coming out of a wide central hub. The edges of the petals have to be as sharp as razor blades, said Kasdin; if they were any thicker, they might reflect too much stray light from the Sun into the telescope, tens of thousands of miles away. When I ran into David Spergel in Berkeley, he'd come not directly from Princeton, but from JPL, in Pasadena, where engineers had built a full-size mockup of one of the petals. He'd witnessed a demonstration of how the petal would unfurl in space. It would have to unfurl, since there's no rocket big enough to contain a hundred-foot-wide flower unless the blossom is folded up on itself.

Spergel agreed that TPF-I, the original four-telescope interferometer, was probably too complicated and difficult to get off the ground. It was basically four James Webb Space Telescopes, and even one is turning out to be very hard and very expensive to build. “The occulter is a lot easier,” he said. “You're flying two objects, and you have to keep them aligned, and deploy the occulter properly, so it's still a lot of work. But the fact that I saw them deploy a mockup at JPL makes me optimistic.”

I heard the same from Jim Kasting when we met a couple of months later at an American Astronomical Society meeting. Kasting has been an exoplaneteer for nearly as long as
Geoff Marcy, Bill Borucki, Michel Mayor, and Dave Latham, which is to say he's been at it since the late 1980s. He doesn't search for planets, and never has. Instead, he's perhaps the world's leading authority on habitable zones—the orbital bands around stars where water can be liquid and life therefore stands a chance of gaining a foothold. Like Marcy and the rest of the exoplanetology community, Kasting was under the impression that NASA was actually talking about funding not one, but two versions of TPF a few years ago. “There was a brief period of extreme optimism,” he said, “where we thought we were going to get two big flagship-type missions. Somebody was crazy, basically.” (
Flagship
is the term reserved for complex, expensive missions that cost a billion dollars or more.)

Then both of them went away. “TPF has been off the drawing boards, basically nowhere, for the last five years, and then it resurfaced as a technology-development project,” he told me one evening over dinner at an inexpensive, minimally decorated, and excellent Vietnamese restaurant a couple of blocks from the Seattle waterfront. “They actually put technology development for TPF at the top of their medium-priority list. We were all disappointed that there wasn't a dedicated exo-planet mission. That hurt a lot of people who had been working on it for fifteen years or more. But we all see the bright side of it, where if we play our cards right, some form of TPF will be the next big flagship mission.”

The flagship space mission the Decadal Survey, a once-every-ten-year report from the astronomical community to NASA, did recommend was something called the Wide-Field Infrared Survey Telescope, or WFIRST. It would mostly do
cosmology, including research on the mysterious dark energy that appears to be making the universe expand faster and faster as time goes on. “It has a small exoplanet component to do gravitational microlensing, but aside from that, most of us are not that thrilled about it,” Kasting said. “But,” he continued, “they're going to be lucky to ever fly that thing.”

The reason is the James Webb Space Telescope, aka the Next Generation Space Telescope, the successor to the Hubble that NASA has been thinking about since the mid-1990s. Originally, the Webb was supposed to have a light-gathering mirror eight meters, or more than twenty-six feet, across. That proved too expensive and too hard to launch, so the mirror was eventually downsized to six and a half meters. That's still pretty big: the Hubble's mirror, by comparison, is less than two meters across.

The original launch date was supposed to be 2007, but that was never really firm. The actual launch, once NASA started funding the project seriously, was set for 2015. In the fall of 2010, however, an independent review panel requested by Congress reported that the Webb project was badly behind schedule and over budget. Under the best of assumptions, there would be no launch until 2018 at the earliest, and the money it would take to do that would come at the expense of other projects. “This is NASA's Hurricane Katrina,” said Alan Boss, a planet-formation theorist at the Carnegie Institution of Washington, to the
New York Times.
It will, he said, “leave nothing but devastation in the astrophysics division budget.”

This could actually work to TPF's advantage, however, according to Kasting, “because the Europeans will probably fly
their Euclid satellite, which does a lot of the same science as WFIRST. Many of us think it's a waste to do both WFIRST and Euclid.” If Euclid flies late in the current decade, and if WFIRST gets pushed back by Webb's budget problems, maybe WFIRST will get canceled. “So that could help us. If WFIRST went away then maybe the next flagship could be TPF. Thinking optimistically,” he added.

But if Alan Boss is right, there's also a realistic possibility that the Webb telescope could take money not just from future missions, but that it could also hurt Kepler. Originally, NASA agreed to a four-year Kepler Mission, with the possibility, but no guarantee, that the agency would spring for another four years, in what would be called an extended mission. Like the Webb, Kepler went over budget, albeit far less drastically, so the original mission was cut to three and a half years. “We're up for senior review on the extended mission in February 2012,” Natalie Batalha said, “and I'm really worried.” She was even more upset when, in the spring of 2011, a House committee voted to cancel the Webb entirely. “JWST is just a double-whammy. The whole community has sacrificed to fund it. Everyone was unhappy at how much it was costing, but we knew how valuable it could be. And now you have Congress talking about canceling it.”

In the end, the Webb wasn't canceled. Barbara Mikulski, the Maryland senator who runs the Senate Appropriations Committee, is a big supporter of the telescope, at least partly because its headquarters, like that of the Hubble, is in Baltimore. The Goddard Space Flight Center, from which the Webb will be controlled, is in Greenbelt, Maryland. She put
the Webb back in the Senate's version of the NASA budget, and that's the version that survived. It's a good thing for astronomy, if not for planet-hunting in particular: The Webb will be one of the most powerful astronomical instruments ever built, able to peer all the way back into the Dark Ages shortly after the Big Bang, when the stars first began to form. But the Webb will be useful for exoplanetology too: Its infrared-sensitive detectors will be able to analyze the reflected light from hot Jupiters and hot Neptunes to see what their atmospheres are made of—much like what Dave Charbonneau and others have been doing with the Spitzer and Hubble space telescopes, only more effectively. You could even fly an occulter along with the Webb, making it an ad hoc version of TPF.

Like the Spitzer and the Hubble, however, the Webb is a general-purpose telescope, where Kepler has an extremely narrow mission. Any planet-hunting duties would have to compete for Webb observing time with all of the other science the telescope is capable of doing. “I'm worried,” said Batalha, “that Kepler will make this amazing catalog of planets and there won't be anything to follow it up with—that we'll end up waiting decades and decades to explore these worlds we've found.”

That might well be true of the planets Kepler identifies in any case. As everyone knew from the beginning, most of the Kepler stars were too faint to follow up transit detections with radial-velocity measurements. For Earth-size planets in the habitable zone, there was pretty much no way it could happen. And if you did manage to confirm such a planet through
transit-timing variations, say, you'd still need a TPF or a Webb-plus-starshade to have a hope of studying it.

For many young exoplaneteers in grad school or doing postdocs or holding junior faculty appointments, the litany of canceled and postponed space missions is clearly discouraging. Still, a few missions, less expensive and less ambitious but still potentially exciting, continue to move forward. One of them is being cooked up in the Green Building at MIT—the same place Sara Seager works. It's a high-rise, the tallest building on the campus, topped with two white, spherical radar domes that have been there, looking down on the Charles River Basin, at least since I was in college back in the early 1970s. (As a sophomore, I had a job driving the motorboat for the freshman crew coach. On cold November afternoons, the sight of those domes, still lit by a Sun that had fallen below the horizon from where I was sitting, was a reminder that it would be at least an hour before I could get back upriver to the warm boathouse and stop shivering.)

I heard about the mission from Josh Winn, an affable young assistant professor who began his astronomical career, much like Dave Charbonneau and Sara Seager, in cosmology. As an undergraduate at Princeton, he worked on gravitational lensing, trying to measure the size of the universe by looking at the flickering of quasars. The idea is that when the gravity of a nearby galaxy distorts the light of a distant quasar, it can create a multiple image—what looks like two or three or even four quasars where there's actually just one. The light paths the images follow to our eyes vary slightly in length, so when
the actual quasar flickers, the flickering shows first in one image, then in another. This time delay (plus a bit of calculating) tells you how far away the quasar really is. “When I talk to people about lensing,” Winn said, “they listen politely, but mostly their eyes glaze over.” Now that he's part of the search for life on other planets, he said, “they get it right away. I don't have to explain why it's important.”