Three new projects aim to capture a first: an image of a planet orbiting a star outside the solar system

11:19am, June 20, 2008
Magazine issue: Vol. 174 #1, July 5, 2008
Sponsor Message
In the 13 years since the first discovery of a planet orbiting a sunlike star outside our solar system, astronomers have found about 300 such “extrasolar” planets, but still have no pictures of any of them.

These 300 orbs have only been detected indirectly: by the wobble of a parent star as an orbiting planet tugs on it, for example, or by minieclipses a planet generates as it passes in front of its star. But none of the current methods allow an astronomer to actually see the planet. With the first optical system devoted to extrasolar imaging set to begin surveying the heavens this summer —
and with two other systems scheduled to come online by early 2011 — astronomers could get their first real image of such a planet within the next three years, and perhaps much sooner. “The pace is accelerating,” says Michael Liu of the University of Hawaii at Manoa.

Searching for planets among a variety of stars is critical for understanding where and how planets form, Liu says. “We expect to find a lot of stars that don’t have planets around them, and that’s part of the answer.”

Astronomers already know that most, if not all, stars are born with protoplanetary disksthe reservoirs of material from which planets coalesceand they know how many millions of years these disks last.

“But that doesn’t really tell you if the disk went away because it formed planets or if it simply fell into the star,” Liu says. The pioneering efforts of three new research programs will begin to map out the places where planets are most likely to reside, providing key information for the ultimate of planet quests: finding a place just like home.

Even when a candidate planet is found, it is difficult to tell whether it is an actual planet or merely some background objecta faint star, for examplethat lies in the same part of the sky. An image that made headlines in 1998, identified as the first photograph of a planet, turned out to be nothing more than a background star (SN: 4/22/2000, p. 271).

In 2004, a team led by Gaël Chauvin, now at the astrophysics laboratory at Grenoble Observatory in France, used the Very Large Telescope in Paranal, Chile, to image a faint, red dot of light orbiting the brown dwarf 2M1207. Brown dwarfs form like stars, from the collapse of a cloud of gas and dust. But unlike stars, these lightweight bodies can’t sustain nuclear burning.

In this case the red dot turned out to be an object two to five times as massive as Jupiter, located at a distance from the brown dwarf that is farther than Pluto’s average separation from the sun (SN: 9/18/2004, p. 179). That is heavy enough to qualify as a giant planet, but the object almost certainly doesn’t meet what many astronomers consider to be an equally important criterion for planethood: formation from a disk of gas and dust that surrounds a young star. Brown dwarfs simply don’t have the heft to create a disk with enough material to make Jupiter-like planetsespecially at distances as remote from the brown dwarf as the body imaged by Chauvin’s team.

Because of these objections, some researchers don’t think of Chauvin’s image as a picture of a bona fide planet. In any event, astronomers have yet to take an image of a planet orbiting an ordinary, full-fledged star.

False signals

Astronomers always knew it would be a challenge to take a picture of an extrasolar planet, or exoplanet. Even a young orb, still warm and relatively bright from its birth inside a swirling, circumstellar disk of gas and dust, is only one-hundred-thousandth to one-millionth as luminous as its parent star. Imaging such a planet is like trying to find a firefly caught in the glare of a nearby searchlight. Observers have used coronagraphsmasks on telescopes that block the light of a starto search for faint planets orbiting the star. But these masks aren’t tailored to the search for extrasolar planets, and the optics may not be precise enough to create a sufficiently sharp image. As a result, stray starlight can scatter or spread out from the mask, creating a lumpy halo of light. In the image, the lumps would appear similar in size to the planet.

“A small bit of light that happened to land on the detector will produce a false signal” of a planet, says Christian Marois of the Herzberg Institute of Astrophysics in Victoria, Canada.

Planet hunters must also contend with the limits of ground-based optics to take sharp images of tiny, faint objects. The blurriness induced by Earth’s turbulent atmosphere has been the bane of astronomers ever since the invention of the telescope. Over the past decade, researchers have fought back by using adaptive opticsmirrors that flex hundreds of times a second to correct for Earth’s atmosphere. Attached to the back of a mirror, tiny electronic devices called actuators exert gentle pressures, reconfiguring the mirror’s shape.

But among current adaptive optics systems, the number of actuators and their ability to respond rapidly enough to atmospheric distortions may not suffice to photograph a faint planet.

Efforts to image a planet beyond the solar system are heating up on two mountaintop observatories in Chile. Early last year, a new instrument arrived at the Gemini South Observatory atop Cerro Pachon. The Near-Infrared Coronagraphic Imager is the first adaptive optics system designed solely to image planets. “We’ve had general purpose adaptive optics instruments, but NICI is the first built from end to end for this express goal,” Liu say.

NICI is scheduled to begin its search this summer, and Liu’s team has been granted a whopping 50 observing nights over the next two years to conduct its survey. In the quest to image an extrasolar planet, “it’s the biggest program that’s ever been done,” Liu says.

NICI features a specially designed coronagraph along with two cameras, which will simultaneously image a star and its immediate surroundings at two different infrared wavelengths. The two-camera strategy takes advantage of a way in which stars differ from brown dwarfs and massive, Jupiter-like planets. Atmospheres of the dwarfs and planets contain an abundance of methane, which absorbs light at certain infrared wavelengths. One camera will take a picture of a star and its environs through a methane filter, while the other camera records the same view through a different infrared wavelength. A planet will look dim in the methane filter but bright in the other, while the star ought to look the same at both wavelengths.

Subtracting the two images, “the star goes away but the planet pops out,” Liu says. The technique will aid astronomers in distinguishing giant planets from background stars and speckles caused by stray starlight, he says.

To differentiate a faint planet from an artifact created by a camera’s imperfect optics, NICI’s developers rely on another trick. The technique, known as angular differential imaging, takes advantage of the fact that most large telescopes are built to rotate about an axis that differs from Earth's rotation axis. Because of that difference in rotation, these telescopes must employ built-in rotators to keep a celestial target fixed in the field of view of the camera observing it. As a result, any imperfection imparted by the telescope will appear to move on the image.

In the new techniqueindependently developed by Liu at the W. M. Keck Observatory and by Marois at the Gemini North telescope, both on Hawaii’s Mauna Keaa telescope’s rotator is turned off. That ensures that any blob of light generated by an optical imperfection always falls in the same place on an image. In contrast, an image of a bona fide planet will slowly rotate from one image to the next.

Subtracting the stationary blobs and carefully adding together the recordings of the moving target removes the optical aberration and enhances the image of an orbiting planet.

In searching for massive planets, NICI will focus on newborn stars. Lying several hundred to 1,000 light-years from Earth, these youngsters harbor the brightest planets, those that haven’t cooled down since they coalesced. The largest population of newborns happens to lie in the southern sky, which is why Liu’s team was eager to use the Gemini South observatory.

NICI will also examine a collection of stars that are closer, though not quite as young. Nearby stars offer astronomers the best chance of imaging a massive planet that is close to its parent star.

NICI’s optics will primarily look for planets with a separation roughly equal to Neptune’s distance from the sun. That’s in contrast to the highly successful technique of hunting planets indirectly, through the wobble a planet’s gravitational tug induces in its parent star. The wobble method favors close-in planets since they exert the greatest pull. It also favors more mature stars because it’s much easier to track their wobbles. Younger stars favored by NICI are more tempestuous, producing violent bursts of activitystarspotsthat can confound wobble measurements.

Two successors to NICI are now in development. The Gemini Planet Imager is expected to begin operation at Gemini South by 2011, around the same time that a similar device, called SPHERE, for Spectro-Polarimetric High-contrast Exoplanet Research, is installed at another Chilean observatory, the European Southern Observatory’s Very Large Telescope atop Paranal.

Like NICI, GPI has a tailored coronagraph, but one with a more sophisticated adaptive optics system. GPI also has a group of coronagraphic masks designed to minimize scattered light from the star. A built-in interferometer will further aid in canceling out unwanted starlight and speckle patterns. A spectrograph will not only help discriminate planets from background stars and stray starlight, but also help reveal the composition of these orbiting bodies.

Sharper view

Researchers for the first time will have a chance to image a Jupiter-like planet at a smaller, Jupiter-like distance from its stara true replica of what our solar system’s biggest planet might have looked like in its youth.

An hour-long exposure with GPI should enable astronomers to record planets one-ten-millionth as faint as their parent stars, says Bruce Macintosh of the Lawrence Livermore National Laboratory in Livermore, Calif. The system won’t just look for youngsters, but also for orbs that are up to 1 billion years old, ranging in separation from roughly Jupiter’s distance from the sun to twice Pluto’s distance.

Using a miniature set of 1,600 actuators etched like a microchip, GPI’s main deformable mirror can be flexed at the finest of scales. This mirror will be combined with a second, more conventional deformable mirror that flexes more coarsely but more rapidly. Together, the mirrors will produce a sharper view of the heavens than any other adaptive optics system now in operation, says Macintosh

SPHERE will hunt for planets among young stars, the nearest stars and those ranging in age from 100 years to 1 billion years, says Jean-Luc Beuzit of the Grenoble Observatory.

SPHERE will also use polarizing filters. Polarized light is radiation whose electric field vibrates in one specific direction in the plane perpendicular to the direction that the light wave travels, rather than in random directions.

About 50 percent of light reflected from planets may be polarizedcompared to almost none from direct starlight. So looking through the telescopic equivalent of Polaroid sunglasses is yet another way to pick out the dim firefly from the stellar searchlight.

More from this issue of Science News