A New Way to Find Planets

Broiling exoplanet
An artist's conception of the giant planet roasting near the star OGLE-TR-56, located some 5,000 light-years away in Sagittarius.
Courtesy David A. Aguilar / CfA.
Astronomers have found the most distant extrasolar planet yet, and by doing so, have advanced a promising technique for discovering new worlds. The announcement, made Monday at the American Astronomical Society (AAS) meeting in Seattle, paves the way for what could be a massive wave of planet detections in the next few years.

The object, dubbed OGLE-TR-56b, is unique in many respects. It’s about 5,000 light-years away, compared to just tens of light-years for most known exoplanets. It orbits its star closely at the breakneck pace of once every 29 hours and at a distance of only 4.5 stellar radii (0.023 astronomical unit). It is relatively light (about 0.9 Jupiter mass) though slightly larger than Jupiter in diameter. Most importantly, it is the first planet originally detected by the "transit" method: searching for signs of an object repeatedly passing in front of a star and blocking a little of its light.

Until now, nearly all extrasolar planets have been discovered by sensitive measurements of stars' radial velocities. A giant planet imposes a slight gravitational wobble upon the star it orbits. By closely monitoring stars for these velocity wobbles, astronomers can ascertain a planet's period, orbital eccentricity, and a lower limit for its mass. Several teams have been following potentially wobbly stars for roughly a decade and have found about 100 exoplanets out of thousands of stars monitored. Two problems with this technique, however, are that it requires a great deal of time on large telescopes and that it does not tell a planet's actual mass, only a minimum value. In most cases, statistically speaking, this value will be fairly close to the true mass, but in a few cases it will be way too low — and there's no way to tell which cases are which.

In the last couple of years a new approach to planet hunting has been getting under way. More than two dozen projects are searching for planets by monitoring vast numbers of stars for transits. Numerous difficulties with the method have turned up, but one group currently leads the pack. Andrzej Udalski (Warsaw University Observatory) and eight colleagues in the OGLE III project surveyed 52,000 Sun-like stars in Sagittarius for signs of transits. They reported last year on more than 40 stars that seem to be repeatedly crossed by small, dim bodies.

Click for animation.
Astronomers can detect a giant-planet transit by monitoring a star's brightness over a long period. If the brightness dips slightly in a characteristic way, the cause may be the passing silhouette of an orbiting world. This can happen only in the rare case when the planet's orbit lies in our line of sight to the star. Click for animation.
S&T illustration by Steven Simpson.

The problem has been in telling what these objects are — giant planets, brown dwarfs, or dim red-dwarf stars, all of which have about the same physical size. In addition, grazing eclipses among normal binary stars can mimic the transit effect. So can a blend of a normal eclipsing binary with an unrelated bright star in its foreground or background. Every suspected case needs follow-up with the radial-velocity method to tell what's actually going on.

At the AAS meeting, a team led by Maciej Konacki (Caltech) and Dimitar Sasselov (Harvard-Smithsonian Center for Astrophysics) announced the first planet confirmed by radial-velocity follow-ups of the OGLE transit suspects. Using several large telescopes including one of the 10-meter Kecks atop Mauna Kea, Hawaii, the astronomers eliminated most of the OGLE suspects as binary stars. But the object orbiting OGLE-TR-56 turned out to have only about the mass of Jupiter.

Sasselov says that looking for transits first, instead of radial velocity changes, will allow astronomers to extend the region searched for exoplanets from a couple hundred light-years to 8,000. What’s more, the pool of potentially surveyable stars grows from about 40,000 to more than 100 million. However, the transit method works only in the rare cases when a planet's orbit happens to lie almost exactly in our line of sight to the star's face. That virtually eliminates all but the closest-orbiting worlds, "hot Jupiters" such as OGLE-TR-56b.

Once a transit is discovered, however, a radial-velocity followup can be done more efficiently than in a blind hunt. With the orbit known, astronomers can time their observations to catch the predicted velocity wobble when it should be greatest. Knowing exactly where and when to look makes it feasible to measure the velocities of faint stars; ordinarily, survey measurements of a faint star are too time-consuming to undertake. (OGLE-TR-56 itself is a very dim magnitude 16.6.) The astronomers describe their method in a paper released a few days ago.

Only one other clear-cut case of a transiting planet is known. The world orbiting HD 209458, 174 light-years away in Pegasus, was originally detected by the radial-velocity method, and only later were its transits spotted.

Being so close to its star, OGLE-TR-56b should be heated to a red-hot temperature of 1,900° Kelvin on its star-facing side. At that temperature, according to theoretical models, its atmosphere may contain clouds of iron mist that drop liquid-iron rain.