Odds of Finding Earth-Size Exoplanets Are 1-in-4

Nearly one in four sun-like stars should host an Earth-mass planet, according to a new census. The finding is the first quantitative measurement of the frequency of planets of various masses in the galaxy. “It’s a landmark paper,” said exoplanet expert Josh Winn of MIT, who was not involved in the new study. “There’s been […]

Nearly one in four sun-like stars should host an Earth-mass planet, according to a new census.

The finding is the first quantitative measurement of the frequency of planets of various masses in the galaxy.

"It's a landmark paper," said exoplanet expert Josh Winn of MIT, who was not involved in the new study. "There's been all this talk, that low-mass planets like the Earth are very common. But this is the first time it's been documented."

The study, published in the Oct. 29 Science, also found plenty of planets in a mass range that astronomers expected to be empty, which may prompt an overhaul of planet-formation models.

Using the Keck Observatory in Hawaii, astronomer Andrew Howard of the University of California, Berkeley and colleagues watched 166 sun-like stars for the telltale wobbles induced by a planet's pull.

Unlike previous surveys, Howard and colleagues were just as interested in stars that lack planets as stars that host them. To avoid biasing the study toward planet-bearing stars, the team selected the nearest and brightest stars in the 120,000-star Hipparcos Catalog.

"It's hard to write a telescope proposal that says, 'We want to look at these stars because we think they don't have planets,'" Howard said. "But that's what we had to do."

Over the course of five years, the team observed each star at least 20 times in search of planets that have masses between 3 and 1,000 times the Earth's mass and that orbit close to their stars.

They found 33 planets around 22 of the stars, some of which had already been discovered and reported by other groups, and 12 candidate planets that still need to be confirmed. Because some stars were observed more often than others, the team included a "missing-planet correction" to account statistically for planets that would probably show up with more observations.

None of these planets was actually the same mass as Earth. Astronomers' instruments aren't yet sensitive enough to detect such small worlds.

"But what we can do is extrapolate," Howard said. "It involves a little bit of speculation, but we're comfortable with that uncertainty."

In general, small planets turned out to be much more common than large ones. The researchers extended that trend down to planets about half Earth's mass.

They found that about 23 percent (give or take about 10 percent) of sun-like stars should have a planet between half and twice the Earth's mass orbiting very close in, about a quarter of the distance from the Earth to the sun. That distance would make the planets far too warm for liquid water. But because planets tend to be more abundant farther from their stars, Howard thinks there should be even more Earth-mass planets in cooler orbits where liquid water is stable.

"I wouldn't be surprised if the true number [of stars hosting Earth-mass planets] is one in two, or one in eight -- but I'm almost sure it's not one in 100," he said. "That's a really big improvement on our knowledge."

Surprisingly, the observations also showed a lot of planets between 5 and 30 times Earth's mass, a range that theoretical models of planet formation predicted should be so empty it earned the name, "the planet desert."

"We showed that the desert is in fact closer to a tropical rain forest," Howard said.

The new numbers are a windfall to researches like Winn, who are involved in designing the next generation of planet-hunting telescopes.

"It sets our expectations much more clearly than they were last week," he said. "We were just guessing, to see how to design the instrument. Now we have some much more solid numbers to put in."

Exoplanet expert Sara Seager of MIT, who was not involved in the new study, noted that this is one of the first exoplanet papers that doesn't focus on just one planet or one system of planets.

"Exoplanet [research] is moving from single-planet characterization to statistics," she said. Thanks to the statistical treatment, Seager is more willing to accept uncertainties in Howard's analysis.

"You might not have confirmation of every individual planet, but that's OK," she said. "That shouldn't stop you from making general statements."

The study also prepares the way for the deluge of planets that should come from the Kepler Space Telescope in the next few years, she says. "We want the world to accept things from Kepler, but they have to be primed to believe that statistics are OK."

Image: 1) WMKO 2) NASA/JPL-Caltech/UC Berkeley

See Also:

Follow us on Twitter @astrolisa and @wiredscience, and on Facebook.