Passages

Revealing the nature of exoplanets

Eleven years ago, David Charbonneau was a new graduate student at Harvard University’s astronomy department, eager to explore the birth of the universe. “Then I learned of the incredible first discoveries that had just been announced in exoplanets,” he recalls. Those objects, the first planets found outside the solar system, prompted Charbonneau to drop the Big Bang like a hot potato. He’s been hunting for exoplanets ever since.

A giant, Jupiterlike planet closely orbits its fiery star in this illustration. When such planets pass in front of their stars, astronomers can not only detect the planet but also measure its size and composition. The method is revolutionizing exoplanet astronomy. T. Pyle, JPL-Caltech/NASA
MAP QUEST. The first map of an exoplanet shows temperature variations across the cloud tops of a hot, Jupiterlike orb called HD 189733b. The map reveals a hot spot on the side of the planet that always faces its star. Knutson et al.
PLANETARY FINGERPRINTS. When an exoplanet passes in front of its parent star, starlight filters through the planet’s atmosphere (small halo in illustration) and reveals the gases there. The change in infrared brightness of the planet-star system when the planet dives behind the star and then reappears indicates the planet’s heat. Brown

Yet the orbs that piqued the imagination of Charbonneau and so many other astronomers in the mid-1990s were then little more than phantoms. Too small to be seen, each planet revealed its presence only because its gravitational pull made its parent star wobble a little. Astronomers could ascertain just two basic properties of each of these elusive planets: a minimum value for its mass and the time it takes to orbit its star.

Seven years ago, Charbonneau and his colleagues brought the first of these alien worlds out of the shadows. They measured how much light the planet blocked when it passed in front of, or transited, its parent star. Such minieclipses reveal a planet’s true size and mass, while the filtering of starlight by its atmosphere shows what gases cling to the alien world.

Scientists have now observed the transits of 20 planets. Astronomers have been able to measure the heat emitted by four of the planets and in one case have gone even further, constructing the first temperature map of an exoplanet.

Now, a mother lode of new transit observations is within reach. Two space missions, one launched late last year and the other scheduled for takeoff in late 2008, are likely to find hundreds of planets showing transits. Next year’s mission offers the prospect of finding the first exoplanet that’s the same size as Earth.

“Transits . . . offer insights that cannot be gained by [the wobble method] alone,” says theorist Alan Boss of the Carnegie Observatories of Washington (D.C.).

“We want to know what planets are made of—whether they’re made of rock or [are] giant balls of gas,” says Charbonneau. Transits provide that information.

Planetary passages

In the standard technique for finding an exosolar planet, astronomers analyze the light from stars, searching for periodic shifts in wavelength that result from the star wobbling ever so slightly as it moves through space. Those wobbles betray the presence of an unseen planet that’s pulling the star to and fro. The method has a natural bias for finding massive planets that lie close to their stars. Nevertheless, in the late 1990s, astronomers were astonished to find dozens of planets nearly as heavy as Jupiter whipping around their stars in orbits less than one-tenth the size of Mercury’s orbit around the sun.

Such close-in planets, lost in the glare of their parent stars, can’t be photographed. But a massive, tightly orbiting planet makes an ideal candidate for transit studies, realized astronomer Tim Brown, now director of the Las Cumbres Observatory Global Telescope Network in Goleta, Calif.

To create a minieclipse, a planet must pass in front of its star, from the viewpoint of an observer on or near Earth. The closer a planet lies to its star, the less precise that alignment needs to be. Moreover, a fatter planet will block more starlight than a smaller one, making a transit easier to spot. Still, the effect is tiny: A large, close-in planet might blot out about 1 percent of a star’s light, producing an effect comparable to a mosquito flying in front of a flashlight a few thousand kilometers away.

With that in mind, Brown and a colleague cobbled together a tiny, 3.5-inch telescope from spare parts and installed the device in a converted chicken coop on a friend’s farm just north of Boulder, Colo. By the time graduate student Charbonneau joined him in 1999, Brown had built a new 4-inch telescope, this one set up in a parking lot at the National Center for Atmospheric Research in Boulder.

Charbonneau came not only with enthusiasm but also with some intriguing information: An international team of astronomers had just used the wobble method to find tentative evidence of a giant, Jupiterlike planet circling close to a near-Earth star called HD 209458. Since Brown’s telescope needed to be tested anyway, why not point it at the star and hope that its planet might transit?

Brown and Charbonneau monitored the star for 2 months but were too busy with another project to immediately analyze the data. When they did, they found that their telescope had indeed found a transiting planet. It blocked about 1 percent of HD 209458’s light.

“We were really startled,” says Brown. “The first time you use a telescope, you wind up a winner—that just doesn’t happen.” Information from the transit, combined with the data from the wobble method, revealed that the planet around HD 209458 is two-thirds as massive as Jupiter and that it has a diameter 32 percent larger.

Next, Charbonneau used a spectrograph aboard the Hubble Space Telescope to record the wavelengths of starlight absorbed when the planet passed in front of HD 209458. The observations showed that while some starlight was entirely blocked by the planet, another portion was merely attenuated. That proved what astronomers had always suspected: The planet wasn’t a solid body, but, like Jupiter, was a puffy ball of gas with an extended atmosphere.

Starlight filtering through the planet’s atmosphere indicated a small population of sodium atoms. It was the first time that anyone had detected a planetary atmosphere beyond the solar system and gotten a whiff of its composition (SN: 12/1/01, p. 340).

“Everyone had assumed that if you wanted to [detect] the atmosphere of an extrasolar planet, you’d have to image it,” says Charbonneau. But the transit method provided that information without the need of a snapshot.

It took several years for astronomers to detect another transiting planet. There was no dearth of intriguing signals, notes Brown, but most of them proved to be spurious—eclipses of one star by another star, not by a planet. Now, with several arrays of telescopes set up across the globe—the Trans-Atlantic Exoplanet Survey in Arizona, California, and the Canary Islands; the Optical Gravitational Lensing Experiment in Chile; the Hungarian Automated Telescope Network in Arizona and atop Mauna Kea in Hawaii; the Wide Angle Search Experiment in the Canary Islands; and the XO project in Maui—transit information is flooding in.

These small, automated telescopes stare at large fields of stars over as many consecutive nights as possible, usually for about 2 months at a time. Because the devices precisely measure a star’s brightness, astronomers can detect the telltale, periodic dimming that’s due to a transiting planet. The surveys have now discovered 16 transits among stars not previously known to have planets orbiting them.

Hot topics

The 2003 launch of NASA’s infrared Spitzer Space Telescope dramatically increased the information that astronomers can glean from transits. Planets emit most of their light in the infrared part of the spectrum. When detectable, that radiation provides a measure of a planet’s surface temperature.

The Spitzer telescope can’t directly image a planet, so researchers employed a trick to separate the planet’s infrared emissions from those of the star. First, they used Spitzer to measure emissions from the star and planet when the two were side by side. They measured again when the planet went behind the star. By subtracting the second measurement from the first, researchers determined how much infrared light was emitted by the planet alone.

Having used an elaboration of this technique, Heather Knutson of Harvard University and her colleagues report in the May 10 Nature the first temperature map of the atmosphere of an exoplanet.

The astronomers focused their attention on a Jupiterlike gas giant that circles the star HD 189733 every 2.2 days. During 33 hours of Spitzer telescope observations, Knutson and her colleagues recorded the infrared light emitted by the planet just after it emerged from behind the star. Infrared brightness varied slightly because the planet rotates, Knutson explains. The team used those variations to define a series of longitudinal strips, from the orb’s north pole to south pole, depicting the uneven infrared brightness of the planet’s upper atmosphere.

Among the cloud-top features, the map reveals a hot spot about 19,000 kilometers wide, or roughly 1.5 times the diameter of Earth. Intriguingly, the hot spot is offset by about 30° in longitude from the region of the planet closest to the parent star, the spot that would receive the highest intensity of radiation.

Knutson and her colleague suggest that the offset has come about because strong winds redistribute the planet’s heat. Blowing with speeds up to 10,000 kilometers per hour, the proposed winds would be about 30 times as strong as any wind on Earth.

Other measurements, reported by Giovanna Tinetti of the Institute of Astrophysics in Paris and her colleagues in the July 12 Nature, suggest that the planet’s atmosphere contains water vapor.

In the spring of 2009, the Spitzer telescope will run out of coolant and will no longer be able to make observations at the longest infrared wavelengths, where emissions from planets are brightest. However, the telescope will still register radiation from transiting planets at shorter infrared wavelengths.

Spitzer can examine only worlds that emit copious infrared radiation and are far too hot to sustain liquid water or life. However, the proposed successor to Hubble, the James Webb Space Telescope, scheduled for launch in 2013, will have the capability to record the much fainter infrared emissions from Earthlike worlds.

Getting small

Astronomers have now found 13 exoplanets that, according to wobble measurements, may be no more massive than Neptune. For those small orbs that happen to produce transits, astronomers hope to determine whether the bodies are composed mostly of gas like Jupiter, contain mostly ice like Neptune, or are rocky, giant versions of our own planet.

Indeed, Swiss researchers announced this spring that they had discovered the smallest transiting planet yet, a body only 22 times as heavy as Earth, similar in mass to Neptune. Every 2.6 days, the planet whips about GJ 436, a dwarf star considerably dimmer than the sun. These transit observations prove that there are exoplanets similar in structure to Neptune but warmer, Charbonneau says. The planet isn’t habitable: Water would be steam at its surface and a compressed solid below (SN: 5/19/07, p. 308).

But the parent star’s dimness also means that the close-in planet isn’t as blisteringly hot as it would be around a sunlike star. Searching for transiting planets around dwarf stars could reveal a truly habitable planet, says theorist Sara Seager of the Massachusetts Institute of Technology.

Two new missions are likely to aid in that search. Last December, the European Space Agency launched a spacecraft called COROT, which will survey 120,000 stars to look for transiting planets with sizes down to twice that of Earth.

Scheduled for launch in early 2009, NASA’s Kepler mission will feature a larger telescope, just under a meter in diameter, that will have the capability to detect and study hundreds of transiting planets that are Earth’s size or even smaller. With a large field of view and an orbit around the sun rather than Earth, Kepler will have an unobstructed view of the heavens, enabling it to repeatedly monitor the brightness of 100,000 stars during its 4-year mission.

Kepler scientists calculate that the telescope should be able to detect the transit of 50 planets about the size of Earth. If most of these star-eclipsing bodies have a diameter 30 percent larger than Earth’s, the number of detections could rise to 185 planets. The latter number would more than triple if the typical transiting planet had a diameter about double that of Earth.

The researchers base their predictions on the assumption that most stars have planets but that intrinsic variability in the brightness of some stars would make transits difficult to detect. To have confidence in a detection, the team will require that a planet make at least four transits during Kepler’s 4-year mission.

“We may be witnessing the start of a cosmic real estate boom,” says Jill Tarter of the SETI Institute in Mountain View, Calif. “Hopefully, Kepler will be able to tell us about the frequency of truly Earth-size planets around stars of varying types within the next few years.”