The Space Reviewin association with SpaceNews

Kepler illustration
The first six weeks of scientific data returned by Kepler has already turned up five exoplanets, with many more candidates waiting to be studied. (credit: NASA)

The exoplanet explosion

Bookmark and Share

Fifteen years ago, there were virtually no known planets beyond the (then nine) planets in our own solar system: just a few found by chance orbiting a pulsar. Then, in late 1995 and 1996, came the initial discovery of planets orbiting main sequence stars like the Sun. That slow trickle of discoveries became a steady stream as astronomers refined their instruments and techniques, as well as increased both the number of stars studied and their period of time observed. By the beginning of 2010 astronomers reported finding over 400 such extrasolar planets, or exoplanets.

That steady stream of discoveries, though, could become a torrent in the next few years. Last year NASA launched the Kepler spacecraft on a mission to look for extrasolar planets in one region of the sky. Now, with the first exoplanet discoveries by Kepler announced, and with the promise of many more to come, astronomers hope that it and other search efforts will help them answer a more fundamental question: how many planets like the Earth exist in our galaxy.

Kepler’s worlds

The 215th meeting of the American Astronomical Society (AAS), held last week in Washington, DC, was something of a coming out party for Kepler. Project scientists used the meeting to announce the first scientific results from the mission, including the first exoplanets detected by the spacecraft, and their expectations for the future of the mission.

Throughout the AAS meeting astronomers heaped praise on Kepler, using words like “amazing”, “fantastic”, “really, amazingly good”, and even “strikingly beautiful data” in presentations about mission science.

Kepler, launched by NASA in March 2009, has a simple mission: observe a single region of the sky continuously for several years. Kepler monitors over 150,000 stars in that region of the sky, looking for minute, periodic decreases in brightness caused when a planet crosses the disc of the star, blocking a tiny fraction of the star’s light. Such “transits” account for only a small fraction of the exoplanets discovered to date—most have been found by the “radial velocity” technique, where astronomers look for Doppler shifts in the spectral lines in stars caused by the gravitational tug of an orbiting planet—but this approach has to detect planets far smaller and in more distant orbits than the most sensitive radial velocity survey, including, astronomers hope, planets similar in size and orbit to the Earth.

While that is the ultimate goal of the Kepler mission, the first planets found by astronomers sifting though Kepler data are similar to the dozens of “hot Jupiters”—gas giants in close-in orbits—found in radial velocity searches. The five planets, designated Kepler-4b through -8b, all orbit close to their stars, with periods of 3.2 to 4.9 days. One of the five, Kepler-4b, is similar in size to Neptune, while the other four are all 30 to nearly 50 percent larger than Jupiter.

Of the five, perhaps the most interesting is Kepler-7b. It is the largest of the five in terms of radius, but has a mass just over two-fifths that of Jupiter. That gives the planet an average density of just 0.17 grams per cubic centimeter, or one-sixth the density of water. “It has the density of Styrofoam,” said William Borucki, principal investigator of the Kepler mission, in a plenary talk at the AAS meeting on January 4.

Why is the planet’s density so low? “The only way you can keep a planet like that less dense is by puffing it up, and the way you can puff up a planet is by heating it”, said Dimitar Sasselov, a professor of astronomy at the Harvard-Smithsonian Center for Astrophysics and a member of the Kepler science team. He added that there are several potential sources of heating, ranging from heating by the star to tidal friction, that could account for the planet’s low density, although astronomers had yet to identify a single key source.

Kepler’s promise

While the discoveries announced last week were interesting in and of themselves, their real significance could be as a harbinger of future exoplanets. The discoveries demonstrated that Kepler is meeting astronomers’ expectations in its ability to detect transits and perform other science. Throughout the AAS meeting astronomers heaped praise on Kepler, using words like “amazing”, “fantastic”, “really, amazingly good”, and even “strikingly beautiful data” in presentations about mission science.

After one presentation about planet-sized objects hotter than their parent stars, a session moderator half-jokingly asked, “So, does anybody know what they are?”

The five exoplanets found so far may be just the beginning of a windfall of worlds to come. The discoveries came in data from only the first six weeks of science observations by Kepler, starting last May. Analysis of that data turned up 175 “planet candidates”, or stars with lightcurves that contained transits, said Natalie Batalha of San Jose State University. Many of those candidates, though, are “false positives”: astrophysical phenomena, such as binary stars where one star performs a grazing eclipse of the other, as well as an eclipsing binary stars in the background near a brighter foreground star.

Some of the candidates, Batalha said, could be dismissed by a closer analysis of the Kepler data, but others required follow-up observations by a number of groundbased telescopes, including the Keck Observatory. Those observations started in August and continued to November, taking up 11 nights on Keck and 85 nights on smaller telescopes. It was those follow-up observations that confirmed the five exoplanets announced at the meeting.

Those observations, though, got to only about 50 of the 175 planet candidates found in that initial set of Kepler data. “So there are still 125 planet candidates waiting in the wings,” she said. “And while these candidates wait in the wings, we’ve got this tidal wave of new data that’s approaching us” as more Kepler data waits to be analyzed. “We expect to have many, many new candidates for the upcoming [observing] season.”

Some of the other candidates are interesting in their own right. For a couple of Kepler “objects of interest”, astronomers found a deeper drop in light when the object passed behind the star rather than in front of it. That means, astronomers said, that the orbiting objects are hotter than their stars, by up to several thousand degrees, even though they have sizes similar to gas giant planets. After one presentation about these objects at the AAS meeting, a session moderator half-jokingly asked, “So, does anybody know what they are?”

Ronald Gilliland of the Space Telescope Science Institute suspects that the objects might be white dwarfs that somehow lost much of their mass. “A white dwarf with a mass about 0.15 of the Sun would have about the right radius, and could have the right temperature, to be these other objects,” he said. Additional observations, including ultraviolet observations by the Hubble Space Telescope, could explain their origin.

“We can say with a fair amount of confidence that solar systems are neither very rare nor are they likely to be in the majority,” Gaudi said of the microlensing survey data.

Kepler is proving useful for more than just planet searches. Astronomers like Gilliland are examining the minute variations in light from the stars to perform astroseismology, mapping those variations to the structure of stellar interiors. Others are using Kepler to perform precision astrometry, measuring the position of the stars in the spacecraft’s field to accuracies of ultimately up to a tenth of a milliarcsecond. Batalha is using Kepler data to characterize the activity of Sun-like stars: an initial study of 43,000 G-class stars found that two-thirds were no more active that the Sun, and only 10 percent were twice as active or more. “This is very good news,” she said, in part because it makes it easier to detect the transits by exoplanets orbiting those stars, but also from an astrobiological standpoint, since more active stars, with more frequent and more severe solar storms, may be more hostile to life.

All this clearly gratified—and relieved—project scientists and NASA officials. Kepler proponents fought for years to win approval for the mission from NASA, battling skepticism that such transits could be detected. “The mission success here was about persistence,” said Pete Worden, an astronomer and director of NASA’s Ames Research Center, which manages the mission. He admitted that when he first heard about the idea “several decades ago” he dismissed it as “rubbish”, but credited Borucki and his team for continuing to push the concept after previous proposals were rejected, eventually winning over scientists and the agency, and then overcoming delays and cost overruns during development that at one point threatened the mission with cancellation.

How many and when

As astronomers wait the bounty of exoplanets that Kepler promises to find, they are also weighing some related questions. One is how long it will take to discover an “Earth-like” planet—at least one that is about the same size as the Earth and in a similar orbit. The second is how many stars out there have solar systems at least somewhat similar to our own, with gas giants in relatively distant orbits.

For the latter question, one astronomer already has an estimate. In an address at the AAS meeting to accept the society’s Warner Prize, Scott Gaudi of Ohio State University noted that it’s difficult to study the “demographics” of extrasolar systems because of several obstacles, including the small sample size of stars surveyed and selection biases. Moreover, the two leading planet-detection techniques, radial velocity and transits, have tended to detect two different populations of planets with little overlap.

As an alternative, Gaudi has been using another technique, called microlensing, taking advantage of the gravitation lensing of an intermediate star to magnify the light from a more distant, and perhaps otherwise unobservable, star. The lightcurve created as the foreground star lenses the more distant one can show spikes created by the light of one or more planets orbiting the more distant star. Or, as Gaudi recalls one astronomer describing it, “you’re looking for planets you can’t see around stars you can’t see.”

However, microlensing can detect smaller planets—potentially down to the size of Mars—and also at greater distances from its star. The disadvantage, though, is that these observations can’t be planned in advance: microlensing events can take place without warning as existing surveys for such events, looking at wide regions of the sky, detect them. Once a potential microlensing event is detected, Gaudi uses a network of “enthusiastic amateurs” called the Microlensing Follow-Up Network (MicroFUN) to carry out additional observations of the event and look for any features in the lightcurve that could be caused by planets.

“I have to plead with everybody to have patience,” said NASA’s Jon Morse, dismissing claims Kepler was in a race to find the first Earth-like exoplanet.

“This might seem like the absolute worst way to go about an experiment to measure the frequency of planets,” he said. That apparent disadvantage, though, is actually an advantage. “It’s because of the chaos that makes this almost a perfectly controlled experiment. Because we have no clue what’s going on as these high-magnification [microlensing] events are happening, there’s no way we can be biased” to focus or not focus on a particular event, removing a key selection bias found in other surveys.

While the sample size of planets is still small in this study, the lack of selection bias gives Gaudi more confidence to generalize on the frequency of solar systems. He concludes that about 15 percent of stars have solar systems at least roughly similar to our own, with multiple gas giants. “We can say with a fair amount of confidence that solar systems are neither very rare nor are they likely to be in the majority,” he said.

As for the first question—how long until an Earth-like planet is detected—astronomers and others advised patience. In the case of Kepler, scientists need to see at least three transits in order to identify a planet candidate, which can mean up to three years’ of observations for planets with periods like the Earth’s. The need for follow-up observations, and the potential bottlenecks in analyzing data and then performing those additional observations, could mean it’s even longer before astronomers can confirm any discovery Kepler makes.

Meanwhile, more advanced groundbased observations, as well as the French spacecraft CoRoT, are also making headway towards detecting Earthlike planets: at the AAS meeting, for example, astronomers reported the discovery of the second smallest exoplanet found to date, with a mass just over four times that of the Earth, found in observations by the Keck. Some have claimed that Kepler is in a race with these observatories to be the first to find an Earthlike planet.

“I have to plead with everybody to have patience, that we will walk through this experiment and we will announce all the new discoveries after we have verified them according to some very meticulous criteria,” said Jon Morse, head of the astrophysics division in NASA’s Science Mission Directorate. In the meantime, though, Kepler and other observatories, in space and on the ground, promise to provide astronomers with a bounty of other exoplanets in a variety of sizes and orbits that will help them better understand just how common—or how rare—solar systems like our own might be in the galaxy.