Planet finding:
Detecting the atmosphere
of a faraway world

As astronomers find more and more planets orbiting distant stars, an NCAR scientist and his colleagues have for the first time detected the atmosphere of an extrasolar planet. Now they plan to analyze the atmosphere in more depth.

It was more than 200 years ago--on June 6, 1761, to be precise--that a Russian scientist named Mikhail Lomonosov observed what looked like a black dot crossing the surface of the Sun. The dot was Venus, making a rare direct passage between the Sun and Earth, and Lomonosov noticed something odd about the planet's edge: it appeared to be enveloped in a thin haze. This marked the first time a scientist detected an atmosphere on another planet.

Last year, NCAR's Timothy Brown and a team of fellow astronomers also detected an atmosphere by observing light passing by the edge of a planet. Their discovery made scientific history because the planet isn't in our solar system. Instead, the massive sphere--known prosaically as HD 209458 b--orbits a star some 150 light years from Earth (one light year is equivalent to about 6 trillion miles, or more than 10 trillion kilometers).

Building on this breakthrough, scientists believe they can analyze distant atmospheres to learn more about the formation of other solar systems--and potentially even gather clues about life elsewhere in the galaxy. Brown and his collaborators are turning to the next step in their research: determining the molecular composition of the atmosphere of HD 209458 b.

Brown's colleagues include Robert Noyes (Harvard-Smithsonian Center for Astrophysics), Ronald Gilliland (Space Telescope Science Institute), David Charbonneau (California Institute of Technology), Edward Dunham (Lowell Observatory), and Juan-Antonio Belmonte (IAC, Astrophysics Institute of the Canary Islands). Despite the enormous difficulties in making observations from a distance of 150 light years, Brown has little doubt of success. "We should be able to detect common molecules in the atmosphere," he says.





Brown has been interested in the stars ever since he was a child, reading about the launch of Sputnik and other satellites in the 1950s. "I can't remember wanting to be anything but an astronomer," he says.


With his father, James Brown (former chancellor of Southern Illinois University), Brown built a telescope when he was 14 years old--then built a small observatory for the telescope so he and his friends could be out of the wind as they watched the stars.

After earning a doctorate in astrophysics from the University of Colorado, Brown joined NCAR's High Altitude Observatory to pursue solar research. His work took a dramatic turn in 1992, when he and a team of scientists at NCAR and the Smithsonian Astrophysical Observatory mounted a spectrograph they'd designed and built on a 60-inch (1.5-meter) telescope at the Fred Lawrence Whipple Observatory on Mt. Hopkins in Arizona. They'd created the spectrograph to study subtle oscillations in the light coming from the Sun and other stars. But Brown soon realized it could be used for planet hunting as well, and he began his search, just as others were launching similar quests. At that time, no planets in other solar systems had ever been detected--but scientists such as Brown were certain of their existence.

Such planets were too far away and too small to be observed directly by even the most powerful telescopes. Instead, Brown and other scientists turned to a technique involving radial velocity. This method uses Doppler spectroscopy to look for tiny shifts in spectral lines that indicate motion toward (blue) or away (red) from the observer. While stars exert gravitational pull on planets, the reverse is also true. The detected radial velocity is a faint wobble in a star created by the gravitational pull of its orbiting planet. This method yields three kinds of data: how close the planet is to its star, how massive the planet is, and the shape of its orbit.

It wasn't until 1995 that Michel Mayor and Didier Queloz of the Geneva Observatory found the first extrasolar planet, circling the star 51 Pegasi. Brown's team, which had also been investigating the star, was one of the first to confirm the groundbreaking observations.

Four years later, teams at Harvard-Smithsonian (which included Brown as a member) and San Francisco State University revealed that not one, but three planets were orbiting the star Upsilon Andromedae. This was the first time an entire solar system had been detected--and it provided strong evidence that the galaxy may well be teeming with planets. "When I look up at the stars now at night, I can imagine easily that every one of them has planets around [it]," says Debra Fischer, a physicist and astronomer who was a member of the San Francisco State team.


©Lynette R. Cook

This artist's depiction shows three extrasolar planets revolving around the star Upsilon Andromedae.


A distant STARE

Under the remote and starry skies of the Canary Islands, a group of NCAR researchers is probing the galaxy for hints of distant planets. The project, known appropriately as STARE (Stellar Astrophysics and Research on Exoplanets), consists of viewing thousands of stars for weeks or months at a time to detect the faint dips of light that could indicate an orbiting planet.

Using a telescope that NCAR scientist Timothy Brown built, the researchers (including Donald Kolinski of NCAR and Roi Alonso Sobrino of IAC) take timed exposures of the same field of view, then analyze the resulting data for light curves.

Timothy Brown

The STARE telescope was installed in a newly constructed dome on the island of Tenerife by a team from NCAR's High Altitude Obsevatory. There's no guarantee of success—Brown and Ronald Gilliland of the Space Telescope Science Institute in Baltimore failed to detect planets after a survey of 35,000 stars in 2000. But STARE will provide other scientific payoffs, such as cataloging stars that have significant variations.


In the summer of 1999, Brown and Charbonneau (who was then an HAO Newkirk Fellow and a graduate student at Harvard) set up a small telescope in the parking lot of NCAR's Foothills Lab and pointed it skyward. The telescope (with lenses that Brown ground himself) used precise time-series photometry to detect the dimming caused when a planet transits (crosses in front of) its parent star. When the two men trained the instrument on the star known as HD 209458 in September 1999, they found the classic dimming pattern that indicated the existence of a planet.

Brown turned to the transit method to augment the data gleaned from radial velocity. The method relies on the alignment of extrasolar systems. If a system is oriented so that Earth lies near the plane of the distant planet's orbit, then once per orbit the planet passes between its star and Earth, making the star appear dimmer. A Jupiter-sized planet transiting a Sun-sized star, for example, would cause the star's light to dim about 1% for several hours.

By combining the data on diameter with data on a planet's mass from radial velocity measurements, "you can determine its density," Brown explains. "And if you know its density, then--eureka!--you know what it's made of."

The transit method also can be used for determining whether a distant planet has an atmosphere. That's because an atmosphere, if it existed, would distort or leave a chemical signature on the small amount of the star's light that passed through it (as Lomonosov saw when he detected the atmosphere of Venus by noticing a slight haze.) Different molecules absorb starlight in different ways, and Brown's team decided to look for sodium, which leaves a distinctive signature that shows up in the yellow-green part of the spectrum. "Sodium has a strong spectroscopic signature," Brown explains. "The analogy I like to use is that it's like skunk. You don't have to have a lot of it in the air to know that it's there."

Last year, Brown and his team observed several transits using the imaging spectrograph on NASA's Hubble Space Telescope. They compared these data with spectra taken at other times. The result: traces of sodium in the spectra of the transits that could only have come from an atmosphere around HD 209458 b.


Timothy Brown and David Charbonneau detected a planet crossing in front of its star, HD 209458, as shown in this artist's rendition.  
©Lynette R. Cook

The discovery itself didn't come as a major surprise. Scientific models predict that gas-giant planets such as HD 209458 b are surrounded by atmospheres that contain sodium. But the finding is an astronomical landmark because it points the way to observing atmospheres of other distant planets. "This opens up an exciting new phase of extrasolar planet exploration, where we can begin to compare and contrast the atmospheres of planets around other stars," says Charbonneau.

Within the next few years, scientists will have powerful new space- and ground-based tools to hunt for extrasolar planets. Astronomers believe they can use spectrographs to look for signs of life elsewhere in the galaxy. If they detected oxygen in an atmosphere, for example, that would be an indication of plant life. That's because oxygen, which easily combines with other elements, is rarely present in pure form unless it is being regularly replenished.

Brown doubts that HD 209458 b harbors life as we know it. The planet is so hot that any clouds in the atmosphere would probably be composed of liquid minerals, such as molten drops of iron. Given its rapid rotation and proximity to its sun, the atmosphere is probably extremely turbulent. "It's probably crazy by Earth standards," he says. "You can bet that there are winds on that planet, and they may even be supersonic."

Brown believes the data gathered by his team and other astronomers will yield valuable insights into the formation of planets and enable scientists to determine whether the composition of our own solar system is the exception or the rule in the galaxy. "For a long time we're going to be quite limited in what we can detect," he warns, because the planets are so far away. But, he adds, "the proposition is so exciting we ought to see what we can deduce."



On the Web

NASA PlanetQuesty
UCAR > Communications > Highlights > 2002 Search
Highlights 2002