Imaging exo planets.
In the 16th century an Italian monk named Giordano Bruno claimed that an infinity of worlds circled other stars, each inhabited with intelligent beings. In 1698 Christiaan Huygens pointed his telescope at various stars in the hope of finding extrasolar planets, but soon realized his telescope wasn't up to the task.
Fast-forward three centuries. In the past eight years, astronomers have discovered more than 100 extrasolar planets by detecting the "wobbles" of their parent stars. Still, no one has seen an exoplanet directly--no one has taken a picture of one. But thanks to new technology, astronomers are on the verge of doing just that.
With realistic prospects for success, exoplanet imaging has suddenly become a hot field. Teams led by Ray Jayawardhana (University of Michigan), Anne-Marie Lagrange (Grenoble Observatory, France), Michael Liu (University of Hawaii), Bruce Macintosh (Lawrence Livermore National Laboratory), and Ralph Neuhauser (University of Jena, Germany) are using the world's largest telescopes, such as Keck, Gemini, and Subaru, for direct-imaging forays. Another team, led by William Sparks (Space Telescope Science Institute), is using the Hubble Space Telescope's Advanced Camera for Surveys to try to image any planets that might exist around Alpha Centauri A and B, the dominant members of the nearest star system to the Sun. This project might later target other nearby stars.
Exoplanet images will help plug important gaps in our understanding of planetary systems. For example, because the currently dominant radial-velocity technique doesn't tell us how much exoplanet orbits are tilted to our line of sight, it provides only a lower limit on a planet's mass. Direct imaging would enable astronomers to watch a planet orbit for several years, allowing them to determine its orbital inclination and true mass.
Once images are obtained, spectra won't be far behind. Securing a faint planet's spectrum will be no easy task. But several instruments coming online in the near future will provide at least a crude spectrum of any planet they image, supplying a wealth of information about its composition and temperature.
Direct imaging will also enable astronomers to study a wider and more representative sample of planets. Because radial-velocity surveys depend on velocity-induced Doppler shifts in the spectra of the planets' parent stars, they have an easier time detecting objects in tight orbits. Direct imaging has the opposite bias; it does a much better job of detecting planets relatively far from their stars, where the stellar glare is less severe. As a result, direct imaging could help uncover planetary systems more like our own.
Candles Amid Searchlights
Serious proposals to image extrasolar planets date back almost 20 years. In 1986 Roger Angel, Neville Woolf, and Andrew Cheng (all then at the University of Arizona) authored a paper in Nature that discussed how to image extrasolar planets from space with a 16-meter infrared telescope. In the very same year, Richard J. Terrile (NASA/Jet Propulsion Laboratory) and Christ Ftaclas and Edward Siebert (both then at Perkin-Elmer) put together a detailed proposal for a space-based planet imager using visible light, but funding never materialized.
At that time ground-based imaging wasn't up to the task. Atmospheric turbulence smears a star's light into an arcsecond-wide blob that hides potential planets in glare. The glare is millions of times brighter than any planets would be, even on nights with the best seeing. If we were to look at our solar system from 30 light-years away, Jupiter would be just a half arcsecond from the Sun and the Sun several hundred million times brighter. Seeing Jupiter would be like spotting a birthday-cake candle a few centimeters from a searchlight at a distance of 10 kilometers.
But thanks to the development of adaptive-optics technology, astronomers now have a chance of overcoming atmospheric turbulence. A telescope equipped with adaptive optics has a flexible mirror somewhere in the light path that changes shape hundreds or even thousands of times per second in order to compensate for the distorting effects of Earth's churning atmosphere (S&T: October 2001, page 30).
In 1994, a rudimentary adaptive-optics system provided an image of Gliese 229B, a brown dwarf--a glowing body of gas that isn't quite massive enough to sustain nuclear fusion as stars do, but is more massive than a planet. As impressive as this feat was, obtaining an exoplanet image is much harder. Gliese 229B appears about 8 arcseconds from its parent star, and with a relatively hefty 40 to 60 Jupiter masses, it still retains significant heat from its formation. As a result, the discovery team could detect the brown dwarf 's own infrared glow rather than seeing it just by reflected starlight.
The discovery of Gliese 229B illustrates the reason why all of the current direct-imaging programs, and most of those being planned, are designed for infrared rather than visible-light imaging. Young planets are still warm from their formation, and they are brightest in the infrared. In contrast, most stars being targeted in planet searches are dimmer at these wavelengths than they are in visible light. Glare becomes much less of a problem, with a typical brightness ratio of about 10,000:1 rather than 300,000,000:1.
In 1999 the European Southern Observatory's 3.5-meter New Technology Telescope demonstrated that exoplanets could, in principle, be imaged from the ground with adaptive optics and near-infrared cameras. The telescope pinpointed a faint object 2.5 arcseconds from a star in Hydra called TWA-7. TWA-7 was nine magnitudes (4,000 times) brighter than the object. Unfortunately, the faint point turned out to be a background star. But if it had been a true companion to TWA-7, its apparent brightness would have implied the existence of a 3-Jupiter-mass planet in a 140-astronomical-unit (a.u.) orbit (one a.u. is the average Earth-Sun distance).
A Bias Toward Youth
It's quite possible that one of the teams has already obtained the first exoplanet image but doesn't yet know it. Jayawardhana began a survey of nearby stars in 2001 using the 10-m Keck II telescope. He already has found several candidates that are awaiting confirmation.
Determining whether these objects are true companions or just background stars requires follow-up observations to measure their motions across the sky. Nearby stars exhibit proper motion--slight positional shifts in the dome of the sky that result from their motions with respect to the solar system. A true companion will move in roughly the same direction and speed as the primary star, but it takes at least a year or two to see any appreciable motion.
The other way to characterize candidates is to obtain their spectra in order to look for molecules, such as water vapor or methane, that indicate a temperature lower than that of a cool star, about 2000[degrees] Kelvin. "It's technically possible with current technology to get a spectrum of a companion candidate that is separate from the star's spectrum, but it's very, very difficult," says Jayawardhana.
Fortunately for today's planet hunters, since 1997 astronomers have discovered about 100 stars less than 30 million years old that are within 300 light-years of Earth. Such young stars are ideal targets for direct-imaging searches because any orbiting giant planets will retain significant heat from their formation, so they should be glowing brightly in the infrared.
Even if the current surveys turn up nothing, the negative result will tell astronomers that massive planets in wide orbits are extremely rare. "If nature did not put any Jupiter-mass planets out in 20-a.u. orbits, we might come up empty-handed," explains Jayawardhana. "But if these planets exist in any substantial numbers, we should be able to find them with the current technology."
New Contenders: The Lyot Project
Although current surveys might possibly turn up exoplanet images, even more powerful searches are just starting to come online. The Lyot Project is one of five efforts that will enter the race in the next five years. It should start scanning the skies about the time this issue reaches subscribers.
The Lyot Project is led by Ben Oppenheimer (American Museum of Natural History), who codiscovered Gliese 229B. The project employs one of the world's most powerful adaptive-optics systems: the 941-actuator, 3.6-m Advanced Electro-Optical System (AEOS) telescope on the Hawaiian island of Maui, which was built by the US Air Force to track artificial satellites. The project is named after Bernard Lyot, the French astronomer who invented the coronagraph in the late 1920s and early 1930s in order to observe the Sun's corona without having to wait for a total solar eclipse.
A standard coronagraph uses two light-blocking masks, which together eliminate most of a target star's light while leaving the surrounding area visible. The first mask is an opaque disk at the center of the focal plane, which blocks 93 percent of the starlight. The remaining light, which would normally form rings around the star (the so-called Airy rings), is diffracted around the disk and forms two bright rings in the image --one that hugs the dark region at the center, and another farther out. The second mask blocks both these areas. As a result, the coronagraph reduces the original starlight by 98.5 percent while blocking relatively little light from the surrounding area, where faint planets might lurk.
Oppenheimer will observe mostly in the infrared to look for the glows of young planets, though the telescope will also take visible-light data simultaneously to look for starlight reflected by planets. Even using just a standard coronagraph, the Lyot instrument should be able to spot a faint companion 14 magnitudes (400,000 times) dimmer than its parent star just 0.2 arcsecond away--a mere 2 a.u. at 30 light-years.
Lyot can thus see planets closer to their host stars than has any previous planet imager. "Lyot is designed to see exoplanets at solar-system scales. That's what makes us stand apart," says Oppenheimer. "We have no idea what we're going to find, but there has to be something out there."
Over the next several years, Lyot will survey several hundred stars brighter than 8th magnitude out to a distance of roughly 80 light-years. Oppenheimer's team eventually hopes to upgrade the imager to allow for true spectroscopy, but for now the project can take measurements of an object's brightness at different near-infrared wavelengths, providing a crude spectrum. The team also will be able to perform astrometric measurements. Together, these capabilities should help distinguish background stars from true planetary companions. "We hope to make some of the first images of objects that most astronomers would feel comfortable calling planets," says Oppenheimer, who adds that Lyot should find a slew of circumstellar debris disks that were created in collisions between protoplanetary chunks.
New Contenders: The Large Binocular Telescope
Whereas coronagraphy is the tried-and-true method for suppressing starlight, astronomers are developing a more intricate method called nulling interferometry. This technique has been demonstrated only recently, but it holds great promise. While the wave nature of light is an annoyance in coronagraphy because it causes diffraction, interferometry harnesses that wave nature to suppress starlight. With nulling interferometry, the crests of the light waves collected by one mirror are made to coincide with the troughs of the waves from the second mirror. The combined waves cancel each other out, but only along a line that passes through the center of the image, where the star is located. Farther from this line, the waves become increasingly misaligned, diminishing the nulling effect and leaving potential planets visible.
Starting in 2007, a team led by Philip Hinz (University of Arizona) will use this technique in an attempt to image exoplanets. Hinz's group will use the Large Binocular Telescope Interferometer, or LBTI, on Mount Graham in Arizona. The LBTI was designed specifically to image exoplanets and to study circumstellar disks. The LBT will consist of two 8.4-m mirrors whose centers are separated by 14.4 meters, giving it the combined resolving power of a 22.8-m telescope.
The LBTI was designed to pave the way for NASA's Terrestrial Planet Finder (TPF) mission. Nulling interferometry will allow Hinz's team to study remnant dust disks around other stars like the solar-system dust that gives rise to the zodiacal light. These disks are tenuous and faint, but their hazy glows could obscure planets targeted by TPF. Hinz will find out how much zodiacal dust surrounds 80 nearby stars. But the same technology that will allow Hinz to image zodiacal dust close to stars will make for a powerful planet imager as well.
The LBTI will have 10 times the resolution of Hubble's near-infrared NICMOS camera, which has been used to look for substellar companions. The LBTI is capable of imaging young planets (less than 500 million years old) containing five Jupiter masses or more and orbiting stars within 30 light-years of Earth.
A nulling interferometer called BLINC has already been tested on the 6.5-m Walter Baade telescope at Las Campanas Observatory in Chile. It might snag an exoplanet before the LBTI comes online. In fact, BLINC recently detected a disk around the young star HD 100546 in Musca. It even revealed a gap in the disk that might be the signature of a giant protoplanet orbiting about 10 a.u. from the star. "It's very exciting to find a star that we think should be forming planets, and actually see evidence of that happening," says Hinz.
Beyond the Atmosphere
The ground-based programs can detect Jupiters orbiting at several astronomical units from their host stars. But to get in even closer, where terrestrial planets might lurk, requires launching telescopes above the distorting effects of Earth's atmosphere. NASA has two powerful space-based imagers on the drawing board. The agency hopes to launch TPF sometime between 2012 and 2015. TPF could image starlight reflected by Earth-size planets in Earth-like orbits around stars within 45 light-years of the solar system. Follow-up spectroscopy could reveal biomarkers such as water vapor, carbon dioxide, and free oxygen in the form of ozone ([O.sub.3]).
NASA has commissioned two studies to determine what approach to take for the mission, and it is planning to make a final decision in 2006. One option is to launch several small telescopes to fly in formation and perform nulling interferometry in the infrared. The other is to launch a large telescope --with a primary mirror three or four times larger than Hubble's and shaped 10 times more precisely--that would perform coronagraphic imaging in visible light. Both designs require the agency to overcome major technical challenges.
NASA is also studying an optical imager called Eclipse, which would comprise a 1.8-m telescope equipped with a system to correct for imperfections in the telescope's optics, as well as a coronagraphic imager. Tentatively slated for a 2009 launch, Eclipse could image starlight reflected by Jupiter-type planets as well as circumstellar disks.
The European Space Agency (ESA) is planning to launch a space-based planet imager called Darwin in 2014. In its current design, Darwin will consist of six small infrared telescopes flying together at the Lagrangian 2 point, 1.5 million kilometers from Earth in the direction away from the Sun. The telescopes will perform nulling interferometry to image Earth-like planets. To save money, scientists on both sides of the Atlantic have discussed the possibility of combining Darwin and TPF in a joint international venture.
To test the technology needed for Darwin, ESA and the European Southern Observatory (ESO) have begun design studies for a ground-based interferometer that could image exoplanets, with first light scheduled for 2006. This Ground-based European Nulling Interferometer Experiment (GENIE) will make its observations using a combination of 8.2-m and 1.8-m telescopes in the Very Large Telescope array in Chile. It will attain contrast ratios of 10,000:1 in the near-infrared, making it capable of imaging young Jupiters.
ESO also has plans for a ground-based instrument called Planet Finder, which could see first light in 2009. Two competing designs are being studied, both based on adaptive optics and coronagraphy. The winning design will be used to build an instrument for use with the VLT array.
On the Brink of Discovery
Because exoplanet imaging is so new, no one is sure what to expect. "That's what makes it exciting--we don't know what's out there," says Roger Angel. "If past experience is a guide, nature will serve up all kinds of crazy things that we haven't thought about, such as the close-in giant planets like 51 Pegasi."
Oppenheimer agrees. "What we're trying to do here is start a new field, one I would call comparative planetary science," he says. "If we are to understand planets from a truly general perspective, we need many examples. What will a planet five times the size of Earth be like? What bizarre moons might orbit large exoplanets?"
Physicist Paul Davies (Macquarie University, Australia) argues that exoplanet images will have a profound cultural impact. "The major planets in our solar system have been known since early man first looked skyward," he says. "The discovery of more of the same within the solar system, such as Neptune, did not transform our world view. But seeing an exoplanet for the first time would put our Sun and its little retinue of planets into a truly cosmic perspective." If indicators of life such as oxygen were detected, adds Davies, it would represent "the Holy Grail of astrobiology ... a place where life started from scratch independently of life on Earth. In all likelihood, we would not be alone in the vastness of the cosmos."
A Night Hunting Exoplanets at Keck
By Govert Schilling
WAIMEA, HAWAII, July 28, 2003. Hunting exoplanets at the mighty Keck Observatory sounds thrilling, doesn't it? Well, it is thrilling, but you wouldn't reach that conclusion by looking around the control room of the Keck II telescope. Four astronomers are biding their time, drinking weak coffee, and fighting drowsiness and jet lag. Still, this might turn out to be the night when the first image of an extrasolar planet is secured. Team member Eric Becklin (University of California, Los Angeles) is confident enough:"If there are massive planets in wide orbits, they'll be found shortly."
Together with his UCLA colleague Ben Zuckerman, postdoctoral researcher Inseok Song, and adaptive-optics expert Bruce Macintosh (Lawrence Livermore National Laboratory), Becklin uses one of the world's two largest telescopes to study the star Vega. With its radiance blocked by a coronagraphic mask, Vega looks like a giant alien eye on Song's monitor, with a large black pupil and a ragged orange perimeter. The weak signal of one or more orbiting planets might be buried within Vega's infrared glow. Keck II is one of the few telescopes in the world able to detect that signal.
But don't expect champagne to flow tonight, cautions Becklin. If the planet signal is there, it will take months of tedious data analysis to uncover it. Even then, it may take several years before the find is confirmed by detecting proper motion common to Vega and any companion.
Meanwhile, Macintosh is developing a new adaptive-optics system for Keck that will be fully tailored to the needs of exoplanet hunters. "With the new system, we should be able to find massive planets in Jupiter-like orbits circling young stars at 30 light-years away," he says.
Becklin's group has already observed the nearby stars Epsilon Eridani and Fomalhaut, and it also plans to study dozens of nearby young stars recently discovered by Zuckerman and Song. Since newly born exoplanets are relatively warm, they are more conspicuous at infrared wavelengths. "Our list hasn't been published yet," says Song,"so we're ahead of the competition."
The first actual image of an extrasolar planet is an astronomical Holy Grail. That's reason enough for Becklin, Song, Zuckerman, and Macintosh to drink more coffee and keep taking hundreds of half-second exposures of Vega's immediate surroundings with the Near-InfraRed Camera at Keck II. Meanwhile, this same night, R. Paul Butler (Carnegie Institution of Washington) and his colleagues are using Keck I to measure the reflex motions of solar-type stars, which is still the most successful way to detect exoplanets, albeit indirectly. In the Keck I control room, computers quietly hum and Butler is taking a well-deserved nap. It's thrilling science.
Sky & Telescope contributing editor GOVERT SCHILLING has written a book in his native Dutch about the search for exoplanets.
Exoplanet False Alarm
By Robert Naeye
"Wait a minute," you might be thinking as you read David Shiga's overview of exoplanet imaging surveys. "Didn't NASA release a Hubble Space Telescope image of an extrasolar planet several years ago?" If so, you have a good memory. But as is typical, mainstream media outlets often report sensational discoveries with fanfare, yet seldom raise a peep when those findings are retracted.
Discoverer Susan Terebey (now at California State University, Los Angeles) thought that the object in question, TMR-1C, could be a planetary-mass body because its brightness implied a mass of just a few Jupiters and because it lay at the end of a filament that seemed to connect it to a young binary star, suggesting that it had been gravitationally flung out of the system (S&T: August 1998, page 19). As many astronomers predicted, follow-up spectra taken at Keck Observatory by Terebey and her colleagues revealed that TMR-1C is in fact a background star positioned at the end of the filament by pure coincidence (S&T: July 2000, page 19).
In recent years astronomers have imaged bona fide extrasolar planetary-mass objects, some perhaps as lightweight as 5 to 10 Jupiters. But these objects float freely in interstellar space, which makes them interesting, but not as exciting as an orbiting planet. Whoever takes the first image of a confirmed planetary-mass object that is gravitationally bound to a star will be the lucky astronomer immortalized in history books.
DAVID SHIGA is a science writer based in Toronto, Ontario. He is editor in chief of The Varsity, the University of Toronto's student newspaper.
|Printer friendly Cite/link Email Feedback|
|Publication:||Sky & Telescope|
|Date:||Apr 1, 2004|
|Previous Article:||Picturing the heavens: the rise of celestial photography in the 19th century.|
|Next Article:||Curing the refractor blues.|