California scientist has a knack for finding alien worlds like our own
BERKELEY, Calif. — Last summer, a homely room in the basement of a math building on the University of California-Berkeley campus was ground zero in the epic quest to end cosmic loneliness.
This is the lair of Geoffrey W. Marcy, outside a certain robot spacecraft named Kepler, the most prolific American discoverer of alien worlds, so-called exoplanets circling stars beyond the sun.
An August evening found Marcy crouched as usual in a corner in an old wooden desk chair. In front of him were computer screens and a video display connecting him to Mauna Kea, Hawaii, home of the twin Keck telescopes, the two largest in the world.
Queued up for observation from Mauna Kea that night were a few dozen of the most promising objects yet found by NASA’s vaunted planet-hunting Kepler spacecraft.
“These are Earths,” Marcy said, gesturing to the screens. “All my life, I’ve pointed telescopes at stars not knowing if planets were there or not. Now, we know.”
In fact, astronomers now have evidence that there are more planets than stars out there— a signature shift in cosmic perspective, in which Marcy played a leading role.
Kepler was launched in 2009 into an Earth-trailing orbit around the sun. The grand goal was to find Earth-like planets.
The fraction of stars with such planets is known as eta-Earth; it is a key factor in the so-called Drake Equation, used to calculate the number of intelligent civilizations in the galaxy.
Kepler shook the sky as if it were a tree. More than 1,000 possible planets fell out in the first year.
Marcy had been a member of Kepler’s science team from the beginning, in 2001. He and his colleagues were on the verge of being able to say how common Earth-like worlds were in the galaxy. Marcy was being mentioned as a contender for the Nobel Prize.
But Kepler broke down after four years of planet-hunting glory, and plans collapsed for a space mission known as the Terrestrial Planet Finder, which could produce images of distant planets, sniff their atmospheres and perhaps map their geography to determine whether they were habitable or inhabited.
The field, Marcy feared, was approaching a lull.
But then last fall, Erik Petigura, one of Marcy’s graduate students, announced, based on his own analysis of Kepler data, that about a fifth of the 100 billion sunlike stars in the galaxy had potentially habitable Earthsize planets. In effect, he beat the Kepler team to the first estimate of eta-Earth.
Under Marcy’s direction, Petigura had spent the previous two years building and testing his own version of the computer pipeline by which Kepler data was analyzed.
“Learning the occurrence of Earth-like planets can be done only once,” Marcy told him. “Erik, you’re the one; you can sleep later.”
The announcement overshadowed a major exoplanet meeting at NASA’s Ames Research Center, even as astronomers agreed that it was only the first of what would be many tries at getting eta-Earth right. Petigura’s analysis was full of assumptions and extrapolations that would be tested and retested in the coming years, astronomers said.
Marcy nevertheless pronounced himself “tingly,” saying it was the most important work he had been involved in. The National Academy of Sciences recently named the paper as the best on the physical sciences published last year in The Proceedings of the National Academy of Sciences.
One thing Kepler couldn’t do without outside help was to say what these putative planets were like. It could measure the sizes of planets but not their masses and densities. Thus there was no way to know whether these worlds were bags of gas or rocks like Earth.
That’s where Marcy comes in, along with the Keck telescope array and its ability to measure wobbles and masses.
“We’re pouring all our effort into planets roughly the size of Earth,” Marcy said, “not just discovering but measuring the properties of Earth-size planets.”
He was particularly interested in learning at what size a planet went from being a rock with water on it, and possibly habitable like Earth, to being gas, like Neptune.
If Petigura’s analysis was right, he said, the nearest Earthlike planets could be as close as 10 or 12 light-years away, within reach of a moderate-size telescope.
“If you do TPF, you will not come up empty,” he said, referring to a Terrestrial Planet Finder. “You’ll have a handful of them. So we have our homework.”