How, and Why, Do Astronomers Take Pictures of Exoplanets?

It's just a bright yellow blip in a mottled field of blue, but this picture of the far-flung planet 51 Eridani b has astronomers abuzz because it is just that: a picture. Released this week by the Gemini Planet Imager, this view lets us gaze directly at a young Jupiter-like world that's about 100 light-years away.

Despite countless announcements of new and exotic exoplanets, including many that supposedly look a lot like Earth, the vast majority of worlds found beyond our solar system have been detected only via indirect means. Any ideas about their atmospheres, surfaces and ability to support life are, for now, educated speculation.

Bruce Macintosh at Stanford University and his colleagues hope to change all that. They are pushing the boundaries of planetary picture-taking with the Gemini Planet Imager (GPI), an instrument installed in 2013 on the Gemini South telescope in Chile. Actually seeing the light from a whole planet allows scientists to tease out chemical clues to its composition and temperature, helping paint a clearer picture of the alien world.

"Direct imaging is really the technique of the future," says study co-author Sasha Hinkley, an astronomer at the University of Exeter. "To get an understanding of what these atmospheres are like, you need spectroscopy, and direct imaging is suited to that."

Exoplanets today are usually found in one of two ways. When the planet moves across the face of its host star as seen from Earth, it alters the incoming starlight slightly—this is called a transit. Alternatively, the radial velocity method looks for a star that wobbles slightly in response to the pull of an orbiting planet. Such indirect evidence accounts for most of the almost 2,000 confirmed exoplanets found so far.

Only about a dozen exoplanets have been seen in images, and all of these are very large gassy worlds that are far from their stars. For instance, the planetary companion to GU Piscium, discovered in 2014, is 9 to 13 times the mass of Jupiter and 2,000 times as far from its star as Earth is from the sun, taking some 163,000 years to complete an orbit. Meanwhile, the controversial world Fomalhaut b is on an extremely elliptical orbit that takes it from 4.5 billion miles from its star to a whopping 27 billion miles out.

GPI was designed to see planets that are smaller and closer to their stars. It uses adaptive optics, in which tiny motors alter the surface of telescope's mirror up to a thousand times per second. The changes in shape compensate for blurring that happens as light from distant objects passes through Earth's atmosphere, helping it spot smaller targets. The instrument also has a coronagraph, a device that blocks out the light of a star to make it easier to see any nearby planets.

In this case, GPI looked at the star 51 Eridani and was able to see a planet orbiting at about 13 Astronomical Units, more than twice the distance between Jupiter and our sun. The planet's surface temperature is about 800 degrees Fahrenheit. It's so hot because the star system is only 20 million years old, and the planet is still glowing with the heat of formation. The team was also able to see that its atmosphere is mostly methane, just like that of Jupiter.

Studying images of worlds like 51 Eridani b could help solve mysteries of planet formation, notes Macintosh. "At 20 million years old, it still 'remembers' the process," he says. One big question is whether Jupiter-sized planets accrete quickly—on the scale of thousands of years—or if it's a more slow and steady process of millions or tens of millions of years. Because Jupiter is so large and uses up so much mass, figuring out how it came to be and how typical it is could affect models of how other types of planets form.

While direct imaging can give a sense of size, it is not as good at judging the mass of a planet, and it can't yet resolve anything much smaller than our own Jupiter unless the star is relatively dim and the planet is unusually bright. "It's not going to get you rocky planets," Macintosh says. "That's for the next generation [of telescopes]."

In the meantime, GPI and a related instrument, the Spectro-Polarimetric High-contrast Exoplanet Research (SPHERE) at the Very Large Telescope in Chile, are refining the technique and looking for more new worlds that are ready for their closeups.

While GPI sees only in infrared, SPHERE will also look at nearby stars to see if it can resolve planets in visible light, says Julien Girard, operations staff astronomer at the VLT. It won't be able to see another Earth—that's most likely a job for a space telescope—but it will prove that resolving such planets is possible, especially as future technologies achieve better contrast in the light reaching the telescopes' detectors, Girard says.

Hinkley, however, thinks there's a good chance that a next-generation telescope on the ground might be the first to snap a picture of a rocky planet. "The very large telescopes that come online in ten years or so, the 30- and 40-meter class, might do it," he says.

Getting to that stage may depend on improvements in adaptive optics, but it might also mean focusing on the coronagraph and improving the ability to block the light of the star, says Ben Montet, a Ph.D. candidate at the Center for Astrophysics at Harvard. "The challenge isn't imaging the faint thing, but blocking out the bright thing right next to it," he says.

As these expected improvements come online, a nearby star system such as Tau Ceti, which is similar to our sun and only 11 light-years away, would be a good candidate for taking a peek. "It's one of the first things I would turn my telescope towards," Hinkley says. 

Jesse Emspak | | READ MORE

Jesse Emspak is a freelance science writer based in New York City. His work has appeared in Scientific American, The Economist, New Scientist, Livescience.com, The Christian Science Monitor and Astronomy Magazine.