War of The Worlds
By Lee Billings
ESO/L. Calcada
No one knows what the planet Gliese 667Cc looks like. We know that it is about 22 light-years from Earth, a journey of lifetimes upon lifetimes. But no one can say whether it is a world like ours, with oceans and life, cities and single-malt Scotch. Only a hint of a to-and-fro oscillation in the star it orbits, detectable by Earth's most sensitive telescopes and spectrographs, lets astronomers say the planet exists at all. The planet is bigger than our world, perhaps made of rocks instead of gas, and within its star's “habitable zone”—at a Goldilocks distance that ensures enough starlight to make liquid water possible but not so much as to nuke the planet clean.
That's enough to fill the scientists who hunt for worlds outside our own solar system—so-called exoplanets—with wonder. Gliese 667Cc is, if not a sibling to our world, at least a cousin out there amid the stars. No one knows if it is a place we humans could someday live, breathe, and watch triple sunsets. No one knows whether barely imagined natives are right now pointing their most sensitive and far-seeing technology at Earth, wondering the same things. Yet regardless, to be the person who found Gliese 667Cc is to be the person who changes the quest for life beyond our world, to be remembered as long as humans exist to remember—by the light of the sun or a distant, unknown star.
Which is a problem. Because another thing no one knows about Gliese 667Cc is who should get credit for discovering it.
Gliese 667Cc is at the center of an epic controversy in astronomy—a fight over the validity of data, the nature of scientific discovery, and the ever-important question of who got there first.
In late 1995 Swiss astronomer Michel Mayor and his student Didier Queloz found 51 Pegasi b, the first known exoplanet orbiting a sunlike star. It was orbiting far too close to its sun to allow the formation of water, but the discovery made Mayor's European team world famous anyway.
Soon, though, they lost their lead in the planet-hunting race to a pair of American researchers, Geoff Marcy and Paul Butler. The two men had been looking for exoplanets for almost a decade; they bagged their first two worlds a couple of months after Mayor's announcement.
The two teams evolved into fiercely competitive dynasties, fighting to have the most—and most tantalizing—worlds to their names. Their rivalry was good for science; within a decade, each had found on the order of a hundred planets around a wide variety of stars. Soon the hunt narrowed to a bigger prize. The teams went searching for smaller, rocky planets they could crown “Earth-like.”
The Spectral fluctuations of a star with an exoplanet create a sine wave.
Most planet hunters aren't looking for exoplanets, per se. Those worlds are too small and dim to easily see. They're looking instead for telltale shifts in the light of a star, “wobbles” in its spectral identity caused by the gravitational pull of an unseen orbiting exoplanet. When that force tugs a star toward Earth, the Doppler effect ever so slightly compresses the waves of light it emits, shifting them toward the blue end of the spectrum. When the star moves away from Earth, its waves of starlight stretch to reach us, shifting toward the red. You can't see those shifts with the naked eye. Only a spectrograph can, and the more stable and precise it is, the smaller the wobbles—and planets—you can find.
By late 2003 the European team had a very precise instrument, the High Accuracy Radial velocity Planet Searcher, or Harps. Mounted to a 3.6-meter telescope on a mountaintop in Chile, Harps could detect wobbles of less than a meter per second. (Earth moves the sun just a tenth that amount.) The Americans had to make do with an older instrument called the High Resolution Echelle Spectrometer, or Hires—less precise but paired with a more powerful telescope.
Want more WIRED? Subscribe now to get 6 months for $5
As the two teams continued to fight for preeminence, trouble was brewing among the Americans. Marcy, a natural showman as well as a brilliant scientist, regularly appeared on magazine covers, newspaper front pages, and even David Letterman's late-night show. His far more taciturn partner, Butler, preferred the gritty tasks of refining data pipelines and calibration techniques. Having devoted years of their lives to the planet-hunting cause, Butler and another member of the team, Marcy's PhD adviser, Steve Vogt (the mastermind behind Hires), began to feel marginalized and diminished by Marcy's growing fame. The relationships hit a low in 2005, when Marcy split a $1 million award with their archrival Mayor. Marcy credited Butler and Vogt in his acceptance speech and donated most of the money to his home institutions, the University of California and San Francisco State University, but the damage was done. Two years later, the relationship disintegrated. Butler and Vogt formed their own splinter group; Butler and Marcy have barely spoken since.
It was a risky move. Harps and Hires remained the best planet-hunting spectrographs available, and Butler and Vogt now lacked easy access to fresh data from either one. The American dynasty was shattered, and Marcy was forced to find new collaborators. Meanwhile, the ever-expanding European team continued to wring planets from Harps even though Mayor had formally retired in 2007. The search for Earth 2.0, long seen as a struggle between two teams, became a more crowded and open contest.
Then a seeming breakthrough: In the spring of 2007, the Europeans announced that they'd spotted a potentially habitable world, Gliese 581d.1 It was a blockbuster—a “super-Earth”—on the outer edge of the habitable zone, eight times more massive than our own world.
Three years later, in 2010, Butler and Vogt scored their own big find around the same star—Gliese 581g. It was smack in the center of the habitable zone and only three or four times the bulk of Earth, so idyllic-seeming that Vogt poetically called it Zarmina's World, after his wife, and said he thought the chances for life there were “100 percent.” Butler beamed too, in his own subdued way, saying “the planet is the right distance from the star to have water and the right mass to hold an atmosphere.” They had beaten Marcy, laid some claim to the first potentially Earth-like world, and bested their European competitors.
But to a chorus of skeptics, Zarmina's World seemed too good to be true. The European group said the signals the Americans had seen were too weak to be taken seriously. The fight was getting ugly; entire worlds were at stake.
Plotted on a computer screen, a stellar wobble caused by a single planet looks like a sine wave, though real measurements are rarely so clear. A centimeters-per-second wobble in a million-kilometer-wide ball of seething, roiling plasma isn't exactly a bright beacon across light-years. Spotting it takes hundreds to thousands of observations, spanning years, and even then it registers as a fractional offset of a single pixel in a detector. Sometimes a signal in one state-of-the-art spectrograph will fail to manifest in another. Researchers can chase promising blips for years, only to see their planetary dreams evaporate. Finding a stellar wobble caused by a habitable world requires a volatile mix of scientific acumen and slow-simmering personal obsession.
He Claims he discoveredGliese 667Cc in archive data. 
Joe McKendry
A Spanish astronomer named Guillem Anglada-Escudé certainly meets that description. Now a lecturer at Queen Mary University in London, he began working with the American breakaways Butler (a friend and collaborator) and Vogt not long after they announced Gliese 581g.
Today, Anglada-Escudé's name is on the books next to between 20 and 30 exoplanets, many found by scraping public archives in search of weak, borderline wobbles. The European Southern Observatory, which funds Harps, mandates that the spectrograph's overlords release its data after a proprietary period of a year or two. That gives other researchers access to high-quality observations and potential discoveries that the Harps team might have missed. Scavenging scraps from the European table, it turns out, can be almost as worthwhile as being invited to the meal.
In the summer of 2011, Anglada-Escudé was a 32-year-old postdoc at the end of a fellowship, looking for a steady research position in academia. With Butler's help he had developed alternative analytic techniques that he used to scour public Harps data. In fact, Anglada-Escudé argued that his approach treated planetary data sets more thoroughly and efficiently, harvesting more significant signals from the noise.
One late night that August, he picked a new target: nearly 150 observations of a star called Gliese 667C2 taken by the Harps team between 2004 and 2008. He sat before his laptop in a darkened room, waiting impatiently as his custom software slowly crunched through possible physically stable configurations of planets within the data.
The first wobble to appear suggested a world in a seven-day orbit—the faster the orbit, the closer to the star the planet must be. A weeklong year is about enough time to get roasted to an inhospitable cinder—and anyway the Harps team had announced that one in 2009, as the planet Gliese 667Cb. But Anglada-Escudé spied what looked suspiciously like structure in the residuals of the stellar sine wave snaking across his screen. He ran his software again and another signal emerged, a strong oscillation with a 91-day period—possibly a planet, possibly a pulsation related to the estimated 105-day rotation period of the star itself.
Want more WIRED? Subscribe now to get 6 months for $5
He decided to try once more before bed, nulling the seven- and 91-day signals. When the final fit was complete, he stared for a long moment at the results on his laptop's glowing screen, then smoked a cigarette to calm his fraying nerves. The software had uncovered what looked to be another planet. But this one had a 28-day orbit, in Gliese 667C's habitable zone. It looked rocky too, because it was just over four times the mass of Earth. This was what came to be known as Gliese 667Cc. If the data held, it would be the third Earth-like world ever discovered. “It was very strange,” Anglada-Escudé told me in 2012, when I was reporting the story of Gliese 667Cc for my book, Five Billion Years of Solitude, “to find an unpublished, unclaimed, potentially habitable planet in a three-year-old public data set.” Yet there it was.
Anglada-Escudé took his numbers to his mentors, Butler and Vogt. Butler took 21 new measurements of Gliese 667C with the Carnegie Planet Finder Spectrograph in Chile, and Vogt retrieved 20 more measurements from archived Hires data. It all only bolstered the finding, and the team began drafting a discovery paper.
But Anglada-Escudé wanted to be absolutely certain the signal was genuine. The best plan: Get new data from Harps, the European spectrograph. Since Germany was a member of the consortium that funds the instrument, and Anglada-Escudé was at the time working at a German institution, he could apply to use it. Worried it would tip off the European team, Butler and Vogt cautioned against it. But Anglada-Escudé wanted those numbers to cement the discovery. On September 28, 2011, he submitted a proposal for 20 nights on Harps with Gliese 667C on his target list and mentioned the existence of a 28-day signal detected in its vicinity.
In the following weeks, Anglada-Escudé monitored his personal website for visits by members of the Harps review committee. He figured if they were checking his credentials, that meant they were reviewing his proposal. In mid-November he saw a spike in traffic from Garching, Germany, where the review committee was based, as well as from other cities around Europe with Harps team members.
And then his Harps proposal was rejected.
His European team says it found Gliese 667Cc first. Joe McKendry
About two months after Anglada-Escudé's initial application, the Harps researchers uploaded a paper to a public online repository of preprints—unreviewed but otherwise rigorous research reports. The preprint summarized Harps observations of red dwarf stars from 2003 to 2009, and the team had already submitted it to a prominent journal. Grenoble, France-based Xavier Bonfils—head of the Harps campaign to find planets around red dwarf stars3—was its lead author. Within its 77 pages, one paragraph and a short entry in a data table discussed the team's detection of a super-Earth in a 28-day orbit around Gliese 667C, stating that a more detailed discovery paper was in preparation.
Vogt was first to see the Harps team's preprint. He sent a terse email to Butler and Anglada-Escudé: “We've been scooped.”
“I was very upset,” Anglada-Escudé said. “So I reread the preprint again and started cataloging strange things.”
He found what he believed to be a smoking gun: Gliese 667Cc's orbital period was correctly included as 28 days, but another datum for the planet—separation from its star—erroneously matched an orbit of about 91 days. The entry, it occurred to Anglada-Escudé, might at one time have concerned a different signal, before perhaps being hastily changed. He started to wonder if the Harps team had seen his proposal and quashed it so they could take credit for the blockbuster planet they had overlooked. “It could have all been a coincidence,” he said. “But I couldn't help feeling suspicious.”
Anglada-Escudé pushed harder to claim what he felt was rightfully his. He and his collaborators wrote their discovery paper and submitted it to Astrophysical Journal Letters, an influential publication that reviewed and published it before the Harps team's paper went to press.
The Europeans cried foul. They got a paper into Astronomy & Astrophysics and began a modest PR campaign to strengthen their case as Gliese 667Cc's true discoverers.
That June, Bonfils and Anglada-Escudé attended an astronomy conference in Barcelona, and the two men met privately to hash out their differences. Over espresso in a café, they talked quietly for an hour. But as they conversed, their words and attitudes hardened. Neither would back down from claiming to have found the faraway world first.
Was Gliese 581g the second Earth-like planet ever discovered? Or would it turn out to be just a ghost?
Lynette Cook/NASA
The fight would be little more than an academic footnote if it weren't for another up-and-coming astronomer, Paul Robertson, a postdoc at Penn State University. His contribution began, as so many momentous discoveries do, with an expletive: “Oh shit.”
Robertson was sitting at his computer on a chilly afternoon this past February, looking at data from the star the Americans and Europeans first started fighting about, Gliese 581. He was using specialized analytical techniques to try to correlate the magnetic activity of the star to the supposed wobbles, all in an attempt to settle once and for all whether Gliese 581g—Zarmina's World—was real or just noise. As he delved into the data, the wobble from Gliese 581g indeed dissipated into static, falling far below the threshold of statistical significance. It seemed the American team had only seen a ghost. The second ever Earth-like exoplanet would have to be crossed off the star charts.
But the plots scrolling across Robertson's screen revealed something much more shocking: The Europeans' find, Gliese 581d, long considered an unimpeachable discovery and the first plausibly Earth-like planet ever found, vanished as well.
There was no avoiding it. Robertson published a paper explaining that some of Earth's foremost astronomers had gotten it wrong. “Busting other people's planets was never really my intention—I'm more interested in finding them,” Robertson says. But even though he hadn't asked for a reputation as a ruthless destroyer of worlds, Robertson couldn't shake the feeling that more false-positive planets were out there, calling like Sirens to scuttle unsuspecting planet hunters in deep, turbulent seas of noise. With the potential Earths of Gliese 581 now dead as Alderaan, Robertson's team moved to the next world on the list: Gliese 667Cc. But even under Robertson's scrutiny, it wouldn't go away.
In other words, the planet whose discovery was so highly contested became, officially, the first potentially Earth-like world ever detected. The person who found it should be able to bask in the achievement ever after—if he can make his claim stick.
Not long after the failed coffee shop détente, I talked to an exasperated Bonfils. The whole thing was, he said, a misunderstanding, compounded by speed bumps typical of the publication process. The European team had submitted its survey for peer-reviewed publication in the spring of 2009, Bonfils explained, long before Anglada-Escudé's research. Feedback from a fellow researcher had delayed the preprint's release until late 2011. He dismissed the minor inconsistencies it contained as simple mistakes, nothing more. In his view, the most pertinent fact was that others were attempting to take credit for a discovery they simply had not made. “Harps was built by our team, and the scientific program and observations were done by our team,” Bonfils said. “Most of the data reduction was already done and provided in our public data.”
In other words, as Bonfils sees it, his team had already done the hard work of analyzing the data. (Vogt says this is no longer true. The Europeans, he wrote in an email—not to me, but since published—process and archive data in a way that specifically makes it “not reliable by any other than Harps team individuals.” So Vogt says his team now reprocesses the public data and derives their own measurements of wobbles.)
What was at stake here was more than just the discovery. It was about access to and treatment of data. “It would be a pity if the guys who made the instruments, and designed and performed the program of observations, did not receive credit for their work,” Bonfils said. “I'm a supporter of public data, but I had long feared someone would try to publish our data before us, and it has now happened. Right now this community rests on good behavior and gentlemen's agreements.” Bonfils said he senses “anger, an aggression” in Butler, Vogt, and their collaborators—including Anglada-Escudé. “You see it in their papers, in the language and accusations,” Bonfils said. “I think it has become more difficult for them to get the observing time they need.” Without access to the right machines, in other words, the planet-hungry renegade group had no choice but to go a little rogue.
In the research community, who an astronomer believes actually discovered the planet seems to depend to some degree on which side that astronomer is closer to—the established, old-guard teams of Marcy and the Harps superstars, or Anglada-Escudé and a few other upstarts. Meanwhile, those upstarts still unabashedly seek overlooked planets in the public data from Harps and other instruments. As a result, the fight between Bonfils and Anglada-Escudé seethes beneath the cordial formalities of professional astronomy. Bonfils now calls Anglada-Escudé's behavior “unethical,” citing several further clashes over questionable planets drawn from Harps data. Anglada-Escudé still believes the Harps team is stonewalling him—a claim Bonfils says is ridiculous. “We don't have this power,” he says.
In February 2014 NASA announced it had discovered 715 new exoplanets.
Chris Philpot
In 2009 nasa launched the Kepler space telescope on a three-and-a-half-year planet-finding mission. Chasing Earth in orbit around the sun, Kepler looks for exoplanets by searching for worlds that “transit” across their stars as seen from our solar system, revealing themselves in silhouette. The half-billion-dollar telescope has found almost 1,000 exoplanets this way, including—as of April 2014—an Earth-sized world called Kepler 186f, orbiting in the habitable zone of a star about 500 light-years away. That's too far off for an easy follow-up investigation, of course. But Kepler's scientists would argue that's not really the point. Their discoveries have moved the field well past the point where any individual planet or person holds the same allure as the heroic, obsessive world seekers of years past. Not that long ago, the discovery of a single exoplanet was an international media sensation. In February, by contrast, the Kepler mission announced 715 new exoplanets. Outside the astronomy community, nobody really cared.
Kepler's results strongly suggest that planets of all types—including ones identical to Earth in their broad description—are common. The search for Earth 2.0 is, in a sense, already over, even though it has barely begun. Now astronomers want a galaxy-encompassing statistical planetary census, accounting for the profound diversity of planetary systems along the way. Such an approach doesn't focus on individual exoplanets. But it can tell you, as Kepler's rich statistics already have, that about one in five sunlike stars should harbor an approximately Earth-sized planet in the habitable zone, placing the nearest life-friendly world somewhere within perhaps a dozen light-years of our solar system. That calculation actually comes from Geoff Marcy's group—even the central figure from planet-finding's Heroic Age has changed with the times.
Yet when it comes to discovery, being first still matters. It's built into the kind of exploration that planet hunting embodies. It comes from the same insatiable urge that drove our ancestors to climb down from the trees, then race from horizon to horizon, until they finally ran out of frontiers. Even if we have exhausted most new places to explore on Earth, our desire to discover, to make the unknown known with our names, our dreams, our stories, is inexhaustible. We call Jupiter's four largest moons Galilean because Galileo saw them first. We look to the stars because we see in them a future, and being first gives someone the power, however mystical and irrational, to make that future manifest and give it life and meaning that could echo through generations. Even if history doesn't always choose the right person.
Bonfils and Anglada-Escudé actually agree on a lot. They both think planetary data should be open. They both want more and better instruments to monitor each and every nearby sun for planet-induced tremors. They both hope for a grand space telescope that will actually see all the worlds within a hundred light-years of Earth—and anyone who lives on them. Like Robertson, they don't want to destroy other people's planets. All they really want is to find new ones—faraway pale blue dots, kin to our own, pirouetting silently in a typical habitable zone, all tumbling together through endless night.
