Can We Find Another Earth?

NASA is betting that we can, and a team of Princeton astronomers has a clever design for a telescope that could do it within 20 years

By Michael D. Lemonick|Friday, March 01, 2002

One sunlike star just 19.5 light-years away that may harbor an Earth-like planet is Eta Cassiopeia, the brightest light in this swatch of the northern sky.
Based on Photographic data obtained using Oschin Schmidt Telescope on Palomar Mountain. The Palomar Observatory Sky Survey was funded by the National Geographic Society. The Oschin Schmidt Telescope is operated by the California Institute of Technology and Palomar Observatory. The plates were processed into the present compressed digital format with their permission. The Digitized Sky Survey was produced at the Space Telescope Science Institute (ST ScI) under U.S. Government grant NAG W-2166.
Astronomers have searched for, and found, some truly cosmic phenomena over the past few decades: black holes gobbling up a million stars, quasars shining from the very edge of the universe, even neutron stars so dense that a teaspoonful would weigh a hundred tons. But one type of celestial object—the most important one—has eluded stargazers. Despite ground-based telescopes now so powerful they can see galaxies billions of light-years away; despite the Hubble, with its clear vision unhindered by atmospheric distortions; despite X-ray telescopes and infrared telescopes, no one has ever seen a planet anything like ours—another Earth orbiting another star.

That search, high on the wish list of astronomers, is incredibly difficult. From even just a few light-years away in our own little corner of the Milky Way, a planet in an orbit comparable to Earth's would be too close to its star for even the Hubble to see them as two distinct objects. A star is 10 billion times brighter than a planet. Picking out a tiny little Earth in the nearby glare of its heat source is like trying to see a firefly hovering in front of a searchlight—only more difficult. Still, NASA decided several years ago that the prize was worthy of the challenge and began working on a mission known as the Terrestrial Planet Finder. Researchers had always assumed that the project would cost billions, require launching powerful and technologically advanced space-based telescopes, and take a quarter of a century before anyone might actually see another Earth out there.

Or maybe not.

While astronomers have been mired in plans for an exotic array of space-based telescopes, a small, creative team of scientists and engineers based at Princeton University has come out of intellectual left field with a new idea that could cut years from NASA's schedule and cost far less than anyone had believed possible. The key is a revolutionary kind of telescope, invented by Princeton astrophysicist David Spergel, a theorist who didn't know much at all about telescopes until he taught himself optics from a textbook two summers ago. "This is a completely new idea," says Michael Littman, an eminent Princeton optical engineer, "and yet once you see it, you realize how simple and elegant it is. I'm kicking myself that I didn't think of it first."

Simplicity and elegance were the last things astronomers could hope for as they began planning the Finder back in 1996, only six months after planets of any kind were found outside our solar system. The first few planets discovered were huge alien gas balls, much bigger than Saturn or Jupiter, and clearly unfriendly to life. But where there are huge gas planets, astronomers reasoned, there may also be other Earths.

Scientists initially consulted by NASA envisioned launching a fleet of four or five telescopes, each bigger than the Hubble, and sending them out to Jupiter, where our solar system's dust clouds thin out. Flying in perfect formation and combining their light into a single, large, supersharp image, the telescopes would be able to pick out Earth-like planets. The cost would be in the billions and the technology was still undeveloped, but scientists believed it to be the only hope.

Princeton's dynamic planet-hunting duo—Jeremy Kasdin, center, and David Spergel—share a moment of levity in the lab. The black instrument is a borrowed camera they used in developing a revolutionary new telescope design.
Photograph by Amy Eckert
Jeremy Kasdin didn't even know about NASA's quest to find another Earth when he arrived at Princeton in 1999 to take up a position as an assistant professor of aerospace engineering. Kasdin had spent several years at Stanford University working on an Earth-orbiting satellite intended to test Einstein's theory of general relativity. Now he needed new research. As luck would have it, he soon learned that a conference was planned in Pasadena, California, to discuss NASA's Finder mission. It was exactly what he wanted, a space-science project with terrific engineering challenges. "I thought, 'I've gotta go,'" Kasdin says, and he went, hat in hand, to his department head for travel funds.

Meanwhile, NASA administrators, under perennial budget pressure, were having second thoughts about what direction to take with the Finder. They had already asked major aerospace companies, including Ball Aerospace, Lockheed Martin, and TRW, to propose alternative approaches. When representatives of those firms convened in Pasadena, Kasdin turned out to be one of the few academic scientists there.

Princeton didn't have an extensive track record in space-systems work, but Kasdin figured he could put together a strong team that blended the school's engineering talent with its celebrated astrophysicists. When he talked to people from Ball Aerospace, he recalls, "they asked me if I could get Dave Spergel."

He knew that Spergel had been a key member of the team that was building a satellite to explore the afterglow of the Big Bang. But they'd met only briefly at a new-members brunch. Nevertheless, he said, "Sure I can get him," and then came home, fingers crossed, to see if Spergel was interested.

Luckily, Spergel was, and so were a half-dozen other astronomers and engineers, including Ed Turner from Princeton's astrophysics department, Sara Seager from the nearby Institute for Advanced Study, and Michael Littman and Dick Miles from engineering. Within a few weeks Kasdin's team had established weekly discussions that continue to this day. Just about every Friday, whoever's available drifts over to the faculty lounge at the mechanical and aerospace engineering department on campus and joins in a high-level bull session on how to find planets. "Right from the start, these meetings were really fun," says Kasdin. "You've got a roomful of very smart people throwing out creative ideas and then arguing about which ones are crazy and which ones make sense."

The first places where astronomers will look for another earth: From our perch at the edge of the Milky Way, astronomers have identified at least a half-dozen stars within a distance of 20 light-years that might support an Earth-like planet. Each of the stars is similar to our sun in size and mass and is at a stable midpoint in its life cycle: Shown above on a 3-D grid, the stars fan out in several directions from Earth. Apart from Eta Cassiopeia, which is visible from North America, all the stars on this short list are in the southern sky. Alpha Centauri is the closest. Epsilon Eridani is already known to have a giant Jupiter-like planet in orbit around it. Both Epsilon Eridani and Tau Ceti have long been favorite targets of radio astronomers searching for signs of intelligent life.
Graphic by Matt Zang
The best way to find small planets, everyone agreed, was to move away from conventional telescopes and build an interferometer, a series of telescopes that has tremendous power by taking advantage of a principle of optics. For example, if several telescopes are positioned 10 miles apart and the images gathered by each are digitized and fed together at the same time, the final image would have the resolution of a telescope with a single mirror measuring 10 miles across. The largest telescope mirror on Earth is 33 feet across. About a year ago NASA funded the conversion of the twin Keck telescopes—the world's largest—into an interferometer with an image area about 300 feet across that will ultimately be able to find planets as small as Jupiter.

But building an interferometer to find even smaller planets is problematic. First, the engineering behind combining light signals simultaneously is daunting. Second, to find a planet like Earth, and especially to see it directly rather than just detecting it by the wobble caused by its gravitational pull on its sun, the array of telescopes used to create the interferometer would have to be launched into space to get them above our planet's murky atmosphere. Third, the telescopes would still be blinded by a small planet's nearby star. Fourth, interferometry is easier to pull off if the instruments see in infrared, not visible, light.

The pluses and minuses of infrared light illustrate the complexities that must be dealt with in designing an observatory to see small planets. Infrared wavelengths are longer than those of visible light, so the optics don't have to be as precise. And infrared lessens the problem of being blinded by the light of the sun. A typical star outshines a planet in visible light by a factor of 10 billion to one, but in infrared light the ratio drops to 10 million to one, making the planet 1,000 times easier to see. There's another bonus to infrared, too: When scientists take a picture of an Earth-like planet, the light reflected from its atmosphere will bear the telltale signature of the planet's chemistry—and maybe even its biochemistry. The signature of livable planets, Carl Sagan and others have shown, is likely to be most prominent in the infrared part of the spectrum.

One downside is that a dust cloud that extends beyond Mars interferes with infrared transmissions, so the array of space telescopes that NASA envisioned launching to create the Terrestrial Planet Finder would still have to be sent out toward Jupiter. And by the time Kasdin's group got involved in the design, NASA had decided going to Jupiter was too expensive. "It was clear," says Charles Beichman, chief scientist for physics and astronomy at the Jet Propulsion Laboratory and the head of the Finder project, "that going out to three or four or five astronomical units [500 million miles from Earth] was just too hard."

YOU ARE HERE: Young stars that might support terrestrial planets are in greater abundance on the outer edge of the Milky Way than at the center of the galaxy, some 28,000 light-years away from Earth.
Graphic by Matt Zang
One alternative would be to stay closer to Earth and beat the dust problem by building bigger telescopes. Early on, the Princeton team came up with a novel orbit that would keep the Finder relatively close to Earth but send it swinging alternately far above and below the solar system's dusty plane. But solving the problems of infrared wasn't enough; some members of Ball's science team remained disenchanted. It turned out, first of all, that Sagan hadn't had the last word on detecting the biochemistry of a planet. A little more research by scientists like Wes Traub, at Harvard University, revealed that there were plenty of life-related molecules that would show up in visible light as well, including water, oxygen, and ozone. Thanks to the smaller wavelengths involved, a visible-light interferometer would be smaller and thus easier to launch.

Still, the biggest problem with visible light remained: How could the blinding glare of a sun be nullified? The only idea anyone had tried was to use a special filter that blocks out light and grows progressively darker toward its center. It should blot out the star. But no one has ever been able to make one work well.

As interest turned back to visible light, many of the scientists at other institutions, still committed to infrared, began to drop out of the group. Then David Spergel made an intellectual leap so great that team member Ed Turner, himself one of the world's most creative astrophysicists, recalls, "My jaw just dropped when I heard it." Spergel is a theorist, not an instrument designer. He's an expert on galaxies and cosmology, not planets. But, he says, "I had a few months with nothing special to do, so I figured it would be fun to think about something new. I took home a standard textbook, and every night after the kids went to sleep, I'd spend an hour or two reading it."

For most scientists it would have been a woefully inadequate education in optics. Spergel, however, was undaunted and began pondering what was already known about nullifying a star's glare. Other astronomers had deduced that while spacing two telescopes just the right distance apart would suppress light at the center of the field of view, the suppression became even deeper if you flanked the telescopes with two more, smaller scopes, one on each side.

Laser light focused through a cat's-eye mask creates an interference pattern that mimics a close-up telescopic image of a star. In theory, an Earth-like planet would be visible in one of the two dark wedge-shaped regions on either side of the star.
Photograph courtesy of David Spergel
What would happen, Spergel wondered, if you added yet another two, even smaller? "That seemed to help," he says, "and so we tried adding another pair, even smaller, to make eight." Working with other members of the group, Spergel found that each additional set of mirrors helped suppress starlight more. In principle, an infinite set of mirror pairs, large at the center and becoming smaller and smaller at the sides, would be ideal. In the simple four-mirror interferometers others had designed, a little bit of starlight would still spill out in all directions from the central blind spot, making any close-in planets harder to see. Spergel's multiple-mirror setup was designed to funnel the spillover into two fan-shaped regions of light above and below the star, leaving the areas to the sides utterly dark.

An infinite set of mirrors was out of the question, of course, and even three or four pairs of scopes would be too complicated and too expensive. But the multiple mirrors gave Spergel an idea: If you trace the overall shape described by that idealized, infinite series of mirrors, it looks something like a cat's eye—a bulge in the center that tapers off at points on the right and left. So instead of building a lot of individual mirrors that approximate this shape, why not, he wondered, just make a single mirror with that shape? Even better: Why go to the trouble of making an oddly shaped mirror? You could achieve precisely the same effect simply by putting a mask over the opening of an ordinary telescope, making an opening shaped like a cat's eye.

It worked, as Littman made clear in a small lab at Princeton's engineering school in the fall of 2000. He arranged a couple of lenses and an eyepiece so that the light of a small laser passed through them in succession. Then he added a photographic slide, opaque except for a clear circle in the center, to simulate the round opening of a telescope. As the light emerged from the eyepiece, he focused it on a piece of paper. The result was a monochromatic pinpoint of light, surrounded by dimmer, alternating ripples of light and dark called an Airy pattern. Similarly, what's seen at the focus of any telescope when the image of a star is magnified enough is the sum of the Airy patterns corresponding to all colors in the starlight. Then Littman replaced that slide with a second slide, also opaque, but this time with a cat's-eye opening instead of a round one. The pinpoint appears on the paper as before. This time, though, the ripples of light and dark are confined to wedges aimed up and down. To the right and left of the pinpoint is nothing but darkness. It's in that darkness that Spergel, Kasdin, Littman, and the rest of the team plan to see planets.

"This demonstration basically took me an hour to set up," Littman says, "and we can suppress the light of the laser 'star' by a factor of 100,000. That's nowhere near the 10 billion we need, but if we can do this so easily on a shoestring, I'm confident we can make it." If they can, then the Terrestrial Planet Finder will look very different from what NASA envisioned only a few years ago. It will consist of just one satellite, not four, and it will sit somewhere near Earth—probably a million miles away at one of the gravitationally stable points where several satellites are already parked. And it could find Earths decades sooner than the space agency's original target date.

Kasdin and Spergel use a tabletop setup created by optical engineer Michael Littman to test their design for a planet-finding telescope. To simulate starlight, a laser beam is bounced off multiple mirrors. Ultimately, the beam passes through the cat's-eye opening before reaching a curved focusing mirror. Because of the way telescopes are built, a star appears as streaks rather than as a point of light. "To see a planet close to a star, the trick is to get rid of that spillover," Kasdin says.
Photograph by Amy Eckert
To move ahead even sooner, the Princeton-Ball team has proposed an interim Finder with a mirror four meters in diameter that could be launched as early as 2010. Armed with the mask designed by Spergel and refined by Kasdin, the interim telescope would be powerful enough to find Earth-like planets around our closest 20 or 30 stars. If there aren't any, the telescope will still be extraordinarily valuable: Rotate the cat's-eye mask out of the way and you've got a conventional telescope with a mirror roughly four times as powerful as Hubble's.

In addition, the team is proposing a separate Discovery-class mission, NASA's special designation for satellites that cost $350 million or less. This interim telescope would have a smaller, Hubble-size mirror. It would have trouble finding small Earth-like planets, but it would find larger planets like crazy. "In an ideal world," says Kasdin, "we'd be doing the bigger telescope, then eventually moving on to the full-size Terrestrial Planet Finder. But first we have to see if NASA wants to spend the money."

They also have to clear some serious technical and political hurdles. "The Princeton design does have the advantage of not having to make multiple satellites work together," says Beichman. "But their mirror has to be a lot closer to perfect." It may be beyond current engineering to make a mirror with the requisite surface—no bump or dimple bigger than a thousandth of the wavelength of light. The Princeton team has an answer to that: Correct imperfections by bouncing the light off a second, flexible mirror that could be adjusted to compensate precisely. They also have to deal with variations in reflectivity in different parts of the mirror's surface. "That's what I stay up nights worrying about," says Littman.

All these concerns must be addressed before NASA chooses a winning design—and there's plenty of competition. Lockheed Martin is still pushing a multisatellite infrared interferometer; other teams are working on single-telescope optical planet finders with a combination of filters and a crude version of Spergel's mask. "There are no front-runners and no back-runners at this point," says Beichman.

This month NASA will decide which of the competing designs should be funded for further development. If the Princeton-Ball team's telescope ultimately gets built, it will be an achievement so unexpected nobody could even have imagined it a few years ago and so powerful it could reveal Earth's sister planets by the score, bringing us closer to answering a question that has plagued us for most of our existence: Are we alone?

Learn more about the mission at JPL's Terrestrial Planet Finder site:

Princeton's Terrestrial Planet Finder project site: EngAnniv/sld023.htm.

An excellent hands-on site listing known extrasolar planets: ~mariob/astro/exoplan/intro-e.htm.

The Extrasolar Planets Encyclopaedia includes news and background information on planet-hunting:

Comment on this article

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar

Log in to your account

Email address:
Remember me
Forgot your password?
No problem. Click here to have it emailed to you.

Not registered yet?

Register now for FREE. It takes only a few seconds to complete. Register now »