The Life in Our Stars

Still, curiosity about life out there simmered among scientists, albeit quietly. Rather than search for life that has evolved beyond us, maybe we’d have more luck looking for life at its roots. That meant searching for life’s incubators—planets.

One of the first people to take up this search was Harvard astrophysicist David Latham. Speaking to me one autumn morning via Skype from his home in the town of Harvard, down the road from Harvard University’s Oak Ridge Observatory, Latham recalled the exact moment that he decided to start looking for astronomy’s Holy Grail: It was Friday, September 21, 1984. He was in the astronomy department’s computing room. A few days earlier, Latham had received a call out of the blue from an Israeli scientist named Tsevi Mazeh, who was visiting astronomers in California. Mazeh was in his late thirties, just starting out as an associate professor of astronomy at Tel Aviv University.

At the time, Latham was 44 years old and the associate director of Harvard’s optical and infrared astronomy program, studying stars and the evolution of galaxies. Mazeh told Latham that he wanted to use Harvard’s telescopes and digital speedometers to search for exoplanets. And he had a new idea about how to do it: by looking for wobbles in starlight caused by the gravitational tug of an orbiting planet. Latham was skeptical. His telescopes were only sensitive enough to detect a wobble if it was caused by something really big—a planet on the scale of Jupiter or bigger, orbiting close to its star. Prevailing theory at the time suggested that this was a paradox: A really big planet would be made largely of gas and ice crystals, and thus would have to orbit farther away from the star, where it’s cold enough for gases to condense. In other words, the planet Mazeh proposed to find was one that “everybody knew” could not exist, Latham said.

“Well, you know,” Mazeh told him, “the theoreticians could be wrong.”

Latham thought the idea was crazy. But he kind of liked crazy. As a young astronomer in the early 1970s, Latham had moonlighted as a professional motorcycle racer, and his rule of thumb was that if he didn’t fall off his bike at least once a day, he wasn’t going fast enough. And in truth, he’d long been fascinated by the possibility that he could find planets—maybe even one that looked and acted something like Earth. “For many years, I co-taught a course where in the spring term we talked about life in the universe and the prospects for finding it,” Latham told me. “The field didn’t have a very good reputation at the time. It was a little far out.”

As unlikely as Mazeh’s plan seemed, Latham decided it was worth a look. The pair hunted for four years and found…nothing. But then, on the night of March 31, 1988, while working from home over a telephone link to the computers at Harvard, Latham detected a wobble from a star about 132 light-years away. He excitedly emailed his collaborators at other observatories, asking them to take a look, too. “By the time I hit send for the email, it was well into April Fools Day,” Latham wrote in an essay about the find. “But I was not fooling.”

As exciting as the evidence was, Latham recalled, it was difficult to pin down exactly what had caused the wobble. The object in question was at least 11 times as big as Jupiter, on the outer edge of how massive a planet could be, according to theory. Plus, whatever this huge object was, it orbited the visible star in about 84 days, a sign that it was circling too close—and hence too hot—for a gas giant to theoretically exist. He and Mazeh couldn’t be sure what it was, so they hedged, describing the discovery as the star’s “unseen companion” in a May 1989 Nature article. Only in their abstract did they dare suggest that it “may even be a giant planet.” Still, that discovery kick-started what became a massive wave of exoplanet investigation.

In 1995, the Swiss astronomers Michel Mayor and Didier Queloz found another promising wobble around a star called 51 Pegasi. His data indicated that the orbiting mass was about half that of Jupiter’s—certainly within the realm of a real planet. Excited by the find, two other exoplanet hunters, Berkeley astronomer Geoffrey Marcy and R. Paul Butler, now at the Carnegie Institution for Science in Washington, DC, hurried to the Lick Observatory on California’s Mount Hamilton and confirmed the wobble. New theories arose to explain these massive, close-orbiting “hot Jupiters”: Maybe the gas giants had formed in a distant orbit but then migrated closer to their sun as their solar system evolved and changed. Within a few years, astronomers led by Marcy and Butler, along with the Swiss group, had discovered about two dozen new planets.

Or so they claimed. Many astronomers pooh-poohed their finds. A mere light wobble wasn’t enough evidence, they said. They needed more proof. The next breakthrough would come from an enterprising Harvard grad student stationed in a wooden shed, staring at a bright star about 150 light-years from Earth.

On a warm day in mid-September, 1999, 25-year-old David Charbonneau was racing through Boulder, Colorado, gripping the wheel of the beat-up Ford Escort he’d driven out from Cambridge, where he was a Harvard doctoral student. He’d spent August making observations at Boulder’s High Altitude Observatory, part of the National Center for Atmospheric Research (NCAR). Now he was headed to the Boulder home of his adviser, Tim Brown, an NCAR astronomer. He had news—unexpected, history-making news.

Three years earlier, Charbonneau had planned to enter the well-established and respected field of cosmology—the study of how the universe has evolved since the big bang. But when he arrived in Cambridge in the fall of 1996, the first “hot Jupiters” were being detected and debated, and Charbonneau was intrigued. “[Exoplanets] just seemed so practical and tangible to me,” Charbonneau, now an astronomy professor at Harvard, told me in his office near the Radcliffe Quad.

David Charbonneau is among the world’s most prominent exoplanet researchers.
Photograph by Shawn G. Henry

At the time, scientists were still debating whether light wobbles were really exoplanets—maybe the stars were just pulsating. Or the data was showing two stars orbiting each another—a quite common phenomenon, known as a binary star system. In response, Charbonneau’s master’s project adviser, astronomer Robert Noyes, had cooked up another way to detect exoplanets: look for light reflecting off the planet.

Easier said than done. A planet’s reflected light would be about a billion times dimmer than its star—it’d be like trying to find a single Christmas-light bulb in front of a searchlight from a thousand miles away. Regardless, Charbonneau agreed to take up the pursuit. “The question of whether or not there are other inhabited worlds is the biggest question in all of science. That yearning has been with us for a long time.” What’s different, he said, is that for the first time in history, mankind has “the technology to actually answer that question.”

For two years, Charbonneau searched for signs of light bouncing off an exoplanet—and found absolutely nothing, much to the amusement of the undergraduates in the off-campus co-op where he was a resident tutor. “Every day I’d go home, and they’d say, ‘So, did you find the planet?’ I’d say, ‘No.’ Then they’d all laugh and we’d make dinner,” Charbonneau recalled. “Looking back, it could have been crushing. But it was pretty funny.”

Charbonneau changed course for his doctorate: Rather than search for the light of distant planets, he would chase their shadows. If an exoplanet passed between its star and Earth’s line of sight—an occurrence known as a transit—the planet would block a tiny fraction of that star’s light, causing a teeny tiny eclipse. Combining a transit with a wobble would remove all doubt of a planet’s existence.

Before the wobbles were discovered, hunting for transits seemed nearly as foolish as looking for planetary light. First, you’d have to have a planet with the correct orbit, relative to Earth. Then you’d need to stare at its star for months, or possibly years, hoping to spot a barely perceptible dip in its light. (A planet the size of Earth might dim its star by only about 80 parts per million—and for just a few hours. You’d have to be awfully lucky to catch that.) But now that he knew of a few stars that wobbled, Charbonneau knew exactly where to point his telescope. Plus, these planets were huge, so they would block much more of their sun than an Earth would—and thanks to their close orbits, they’d do so as often as every few days.

Noyes had recommended that Charbonneau head out to Colorado to work with Brown, who’d been searching for transits for more than a decade. “So I drove out there, and literally, it’s like a garden shed,” Charbonneau told me of Brown’s research digs. Inside was a small telescope. He motioned across his office to a boxy black amateur scope, perched on a tripod about 6 feet tall, which he explained is almost identical to the one he used in Colorado. “And I’m like, Oh wow, that’s my thesis.”

At the end of that summer, Charbonneau aimed the telescope at a star called HD 209458, located near the Great Square of Pegasus. Back at Harvard, David Latham had recently detected a wobble in that star’s light, suggesting a Jupiter-sized planet in a tight orbit. But when Charbonneau went looking for the transit, he couldn’t find it.

“I plotted the light curve [of HD 209458] and it was flat, of course. It’s extremely unlikely that there would be a transiting planet. But then I realized I had the indexing off. Literally, I forgot that in this software we count zero as the first number, not one.” It was like forgetting to carry the one. He had made a minor computational error.

Charbonneau went back to the computer and re-indexed the data. “I replotted the light curve and all of a sudden, there was this dip. I thought, Oh shit!”

That eureka moment was both exciting and nerve-wracking. “It’s actually very uncomfortable, because nobody’s ever seen it before,” he said. “You’re really worried that maybe it’s not really there.”

He spent most of the day double-checking the data, and then he felt ready to tell Brown. Except his adviser wasn’t there. Brown’s elderly mother had fallen and broken her hip, and he was at the hospital waiting for her to come out of surgery. Tragically, she died on the operating table. A couple of days later, Brown had recovered enough to invite Charbonneau over to his house to hear the news. When Brown saw the data, he poured Charbonneau a congratulatory glass of scotch.

“In this very sad time, I showed him something that for him had been this scientific quest,” Charbonneau said. “It was extremely emotional.”

Within a couple months of Charbonneau’s discovery of a transit across HD 209458, Marcy and other colleagues confirmed their findings from the Keck Observatory on Mauna Kea in Hawaii. The revelation had a transformative effect on astronomical research. After that, most of Harvard’s astronomers shifted their exoplanet efforts: Instead of looking for stars that wobbled, they focused on stars that blinked. “That was a strategic decision, and it really paid off,” Harvard’s Sasselov says. The transit method, which didn’t require such powerful telescopes, would deliver an avalanche of new exoplanet discoveries, leading astronomers to conclude that in the universe, there are likely at least as many planets as there are stars. But did any of those planets contain life? Two Harvard researchers were determined to find out.

Transit discoveries didn’t just confirm the existence of exoplanets; they also gave astronomers critical clues about a planet’s composition. Transits could help scientists estimate a planet’s size (based on how much light it blocked), temperature (using its orbit to calculate how close it was to its star), and density (when its size was combined with the measure of its wobble). Best of all, by analyzing the spectrum of the light skimming around a transiting planet, astronomers could deduce the elements in its atmosphere.

That’s what Sara Seager suggested in a paper she published in 2000, a year after earning her doctorate at Harvard. Seager argued that gaps in the spectrum of light skimming a transiting planet could reveal gases swirling in its atmosphere. Charbonneau and Seager had been friends since their undergraduate days at the University of Toronto. Charbonneau knew about Seager’s theoretical work, and the next year, he validated it by using observations from the Hubble Space Telescope to identify sodium in the atmosphere of the giant planet he’d found transiting HD 209458. The sodium itself wasn’t anything special. But at least they now knew that the technique could work.