Earth-sized alien worlds are out there. Now, astronomers are figuring out how to detect life on them

Stephen Kane spends a lot of time staring at bad pictures of a planet. The images are just a few pixels across and nearly featureless. Yet Kane, an astronomer at the University of California, Riverside, has tracked subtle changes in the pixels over time. They are enough for him and his colleagues to conclude that the planet has oceans, continents, and clouds. That it has seasons. And that it rotates once every 24 hours.

He knows his findings are correct because the planet in question is Earth. Kane took images from the Deep Space Climate Observatory satellite, which has a camera pointing constantly at Earth from a vantage partway to the sun, and intentionally degraded them from 4 million pixels to just a handful. The images are a glimpse into a future when telescopes will be able to just make out rocky, Earth-sized planets around other stars. Kane says he and his colleagues are trying to figure out "what we can expect to see when we can finally directly image an exoplanet." Their exercise shows that even a precious few pixels can help scientists make the ultimate diagnosis: Does a planet harbor life?

Finding conclusive evidence of life, or biosignatures, on a planet light-years away might seem impossible, given that space agencies have spent billions of dollars sending robot probes to much closer bodies that might be habitable, such as Mars and the moons of Saturn, without detecting even a whiff of life. But astronomers hope that a true Earth twin, bursting with flora and fauna, would reveal its secrets to even a distant observer.

Detecting them won't be easy, considering the meager harvest of photons astronomers are likely to get from such a tiny, distant world, its signal almost swamped by its much brighter nearby star. The new generation of space telescopes heading toward the launch pad, including NASA's mammoth James Webb Space Telescope (JWST), have only an outside chance of probing an Earth twin in sufficient detail. But they will be able to sample light from a range of other planets, and astronomers are already dreaming of a space telescope that might produce an image of an Earth-like planet as good as Kane's pixelated views of Earth. To prepare for the coming flood of exoplanet data, and help telescope designers know what to look for, researchers are now compiling lists of possible biosignatures, from spectral hints of gases that might emanate from living things to pigments that could reside in alien plants or microbes.

There is unlikely to be a single smoking gun. Instead, context and multiple lines of evidence will be key to a detection of alien life. Finding a specific gas—oxygen, say—in an alien atmosphere isn't enough without figuring out how the gas could have gotten there. Knowing that the planet's average temperature supports liquid water is a start, but the length of the planet's day and seasons and its temperature extremes count, too. Even an understanding of the planet's star is imperative, to know whether it provides steady, nourishing light or unpredictable blasts of harmful radiation.

"Each [observation] will provide crucial evidence to piece together to say if there is life," says Mary Voytek, head of NASA's astrobiology program in Washington, D.C.

An image from the Deep Space Climate Observatory satellite (left), degraded to a handful of pixels (right), is a stand-in for how an Earth-like planet around another star might look through a future space telescope.

(LEFT TO RIGHT) NASA EPIC TEAM; STEPHEN KANE

In the heady early days following the discovery of the first exoplanet around a normal star in 1995, space agencies drew up plans for extremely ambitious—and expensive—missions to study Earth twins that could harbor life. Some concepts for NASA's Terrestrial Planet Finder and the European Space Agency's Darwin mission envisaged multiple giant telescopes flying in precise formation and combining their light to increase resolution. But neither mission got off the drawing board. "It was too soon," Voytek says. "We didn't have the data to plan it or build it."

Instead, efforts focused on exploring the diversity of exoplanets, using both ground-based telescopes and missions such as NASA's Kepler spacecraft. Altogether they have identified more than 3500 confirmed exoplanets, including about 30 roughly Earth-sized worlds capable of retaining liquid water. But such surveys give researchers only the most basic physical information about the planets: their orbits, size, and mass. In order to find out what the planets are like, researchers need spectra: light that has passed through the planet's atmosphere or been reflected from its surface, broken into its component wavelengths.

Most telescopes don't have the resolution to separate a tiny, dim planet from its star, which is at least a billion times brighter. But even if astronomers can't see a planet directly, they can still get a spectrum if the planet transits, or passes in front of the star, in the course of its orbit. As the planet transits, starlight shines through its atmosphere; gases there absorb particular wavelengths and leave characteristic dips in the star's spectrum.

Astronomers can also study a transiting planet by observing the star's light as the planet's orbit carries it behind the star. Before the planet is eclipsed, the spectrum will include both starlight and light reflected from the planet; afterward, the planet's contribution will disappear. Subtracting the two spectra should reveal traces of the planet.

Teasing a recognizable signal from the data is far from easy. Because only a tiny fraction of the star's light probes the atmosphere, the spectral signal is minuscule, and hard to distinguish from irregularities in the starlight itself and from absorption by Earth's own atmosphere. Most scientists would be "surprised at how horrible the data is," says exoplanet researcher Sara Seager of the Massachusetts Institute of Technology in Cambridge.

In spite of those hurdles, the Hubble and Spitzer space telescopes, plus a few others, have used these methods to detect atmospheric gases, including sodium, water, carbon monoxide and dioxide, and methane, from a handful of the easiest targets. Most are "hot Jupiters"—big planets in close-in orbits, their atmospheres puffed up by the heat of their star.

In an artist's concept, a petaled starshade flying at a distance of tens of thousands of kilometers from a space telescope blocks a star's light, opening a clear view of its planets.

NASA/JPL

The approach will pay much greater dividends after the launch of the JWST in 2019. Its 6.5-meter mirror will collect far more light from candidate stars than existing telescopes can, allowing it to tease out fainter exoplanet signatures, and its spectrographs will produce much better data. And it will be sensitive to the infrared wavelengths where the absorption lines of molecules such as water, methane, and carbon monoxide and dioxide are most prominent.

Once astronomers have such spectra, one of the main gases that they hope to find is oxygen. Not only does it have strong and distinctive absorption lines, but many believe its presence is the strongest sign that life exists on a planet.

Oxygen-producing photosynthesis made Earth what it is today. First cyanobacteria in the oceans and then other microbes and plants have pumped out oxygen for billions of years, so that it now makes up 21% of the atmosphere—an abundance that would be easily detectable from afar. Photosynthesis is evolution's "killer app," says Victoria Meadows, head of the NASA-sponsored Virtual Planet Laboratory (VPL) at the University of Washington in Seattle. It uses a prolific source of energy, sunlight, to transform two molecules thought to be common on most terrestrial planets—water and carbon dioxide—into sugary fuel for multicellular life. Meadows reckons it is a safe bet that something similar has evolved elsewhere. "Oxygen is still the first thing to go after," she says.

Fifteen years ago, when exoplanets were new and researchers started thinking about how to scan them for life, "Champagne would have flowed" if oxygen had been detected, Meadows recalls. But since then, researchers have realized that things are not that simple: Lifeless planets can have atmospheres full of oxygen, and life can proliferate without ever producing the gas. That was the case on Earth, where, for 2 billion years, microbes practiced a form of photosynthesis that did not produce oxygen or many other gases. "We've had to make ourselves more aware of how we could be fooled," Meadows says.

To learn what a genuine biosignature might look like, and what might be a false alarm, Meadows and her colleagues at the VPL explore computer models of exoplanet atmospheres, based on data from exoplanets as well as observations of more familiar planets, including Earth. They also do physical experiments in vacuum chambers. They recreate the gaseous cocktails that may surround exoplanets, illuminate them with simulated starlight of various kinds, and see what can be measured.

Over the past few years, VPL researchers have used such models to identify nonbiological processes that could make oxygen and produce a "false positive" signal. For example, a planet with abundant surface water might form around a star that, in its early years, surges in brightness, perhaps heating the young planet enough to boil off its oceans. Intense ultraviolet light from the star would bombard the resulting water vapor, perhaps splitting it into hydrogen and oxygen. The lighter hydrogen could escape into space, leaving an atmosphere rich in oxygen around a planet devoid of life. "Know thy star, know thy planet," recites Siddharth Hegde of Cornell University's Carl Sagan Institute.

Discovering methane in the same place as oxygen, however, would strengthen the case for life. Although geological processes can produce methane, without any need for life, most methane on Earth comes from microbes that live in landfill sites and in the guts of ruminants. Methane and oxygen together make a redox pair: two molecules that will readily react by exchanging electrons. If they both existed in the same atmosphere, they would quickly combine to produce carbon dioxide and water. But if they persist at levels high enough to be detectable, something must be replenishing them. "It's largely accepted that if you have redox molecules in large abundance they must be produced by life," Hegde says.

Some argue that by focusing on oxygen and methane—typical of life on Earth—researchers are ignoring other possibilities. If there is one thing astronomers have learned about exoplanets so far, it is that familiar planets are a poor guide to exoplanets' huge diversity of size and nature. And studies of extremophiles, microbes that thrive in inhospitable environments on Earth, suggest life can spring up in unlikely places. Exobiology may be entirely unlike its counterpart on Earth, and so its gaseous byproducts might be radically different, too.

But what gases to look for? Seager and her colleagues compiled a list of 14,000 compounds that might exist as a gas at "habitable" temperatures, between the freezing and boiling points of water; to keep the list manageable they restricted it to small molecules, with no more than six nonhydrogen atoms. About 2500 are made of the biogenic atoms carbon, nitrogen, oxygen, phosphorus, sulfur, and hydrogen, and about 600 are actually produced by life on Earth. Detecting high levels of any of these gases, if they can't be explained by nonbiological processes, could be a sign of alien biology, Seager and her colleagues argue.

Transit spectroscopyWhen a planet transits in front of its star, some light passes through its atmo-sphere. gases there subtly alter the star’s light. The strongest signals come from hot Jupiters, with bloated atmospheres. Secondary eclipseJust before a planet passes behind its star, the observed light combines the star’s with some reflected off the planet. When the planet disappears behind the star, the spectrum changes. Direct spectroscopyBlocking a star’s light with a coronagraph or starshade can reveal the tiny amount of light reflected from the planet itself. The light can be broken into a spectrum. Direct imagingWith the help of a corona-graph or starshade, tele-scopes may one day gather enough photons to form an image. Even if only a few pixels across, it can hold clues to surface features. Gleaning photons from far-off worlds Studying a speck of gas or rock around a star that appears as a mere point of light seems all but impossible. But astronomers have devised ways to study exoplanets without actually seeing them, and they hold out hope of one day capturing a crude image of an exoplanet. Exoplanet Eclipsed Exoplanet Hot Jupiter Planet orbit Star Starshade Starshade Reading the photons By matching the data against simulated spectra of exoplanet atmospheres, researchers tease out a planet’s signal. During the eclipse, the exoplanet’s contributions disappear, allowing astronomers to identify them. A direct spectrum of an exoplanet could reveal more details about the makeup of its atmosphere and any surface colors. How the pixels change color over time can reveal clues to oceans, continents, rotation, and weather. Star spectrum Observed spectral data Simulated planet spectrum Planet spectrum Star plus planet spectrum Planet image
A. CUADRA/SCIENCE

Light shining through the atmospheres of transiting exoplanets is likely to be the mainstay of biosignature searches for years to come. But the technique tends to sample the thin upper reaches of a planet's atmosphere; far less starlight may penetrate the thick gases that hug the surface, where most biological activity is likely to occur. The transit technique also works best for hot Jupiters, which by nature are less likely to host life than small rocky planets with thinner atmospheres. The JWST may be able to tease out atmospheric spectra from small planets if they orbit small, dim stars like red dwarfs, which won't swamp the planet's spectrum. But these red dwarfs have a habit of spewing out flares that would make it hard for life to establish itself on a nearby planet.

To look for signs of life on a terrestrial planet around a sunlike star, astronomers will probably have to capture its light directly, to form a spectrum or even an actual image. That requires blocking the overwhelming glare of the star. Ground-based telescopes equipped with "coronagraphs," which precisely mask a star so nearby objects can be seen, can now capture only the biggest exoplanets in the widest orbits. To see terrestrial planets will require a similarly equipped telescope in space, above the distorting effect of the atmosphere. NASA's Wide Field Infrared Survey Telescope (WFIRST), expected to launch in the mid-2020s, is meant to fill that need.

Even better, WFIRST could be used in concert with a "starshade"—a separate spacecraft stationed 50,000 kilometers from the telescope that unfurls a circular mask tens of meters across to block out starlight. A starshade is more effective than a coronagraph at limiting the amount of light going into the telescope. It not only blocks the star directly, but also suppresses diffraction with an elaborate petaled edge. That reduces the stray scattered light that can make it hard to spot faint planets. A starshade is a much more expensive prospect than a coronagraph, however, and aligning telescope and starshade over huge distances will be a challenge.

Direct imaging will provide much better spectra than transit observations because light will pass through the full depth of the planet's atmosphere twice, rather than skimming through its outer edges. But it also opens up the possibility of detecting life directly, instead of through its waste gases in the atmosphere. If organisms, whether they are plants, algae, or other microbes, cover a large proportion of a planet's surface, their pigments may leave a spectral imprint in the reflected light. Earthlight contains an obvious imprint of this sort. Known as the "red edge," it is the dramatic change in the reflectance of green plants at a wavelength of about 720 nanometers. Below that wavelength, plants absorb as much light as possible for photosynthesis, reflecting only a few percent. At longer wavelengths, the reflectance jumps to almost 50%, and the brightness of the spectrum rises abruptly, like a cliff. "An alien observer could easily tell if there is life on Earth," Hegde says.

There's no reason to assume that alien life will take the form of green plants. So Hegde and his colleagues are compiling a database of reflectance spectra for different types of microbes. Among the hundreds the team has logged are many extremophiles, which fill marginal niches on Earth but may be a dominant life form on an exoplanet. Many of the microbes on the list have not had their reflectance spectra measured, so the Cornell team is filling in those gaps. Detecting pigments on an exoplanet surface would be extremely challenging. But a tell-tale color in the faint light of a distant world could join other clues—spectral absorption lines from atmospheric gases, for example—to form "a jigsaw puzzle which overall gives us a picture of the planet," Hegde says.

None of the telescopes available now or in the next decade is designed specifically to directly image exoplanets, so biosignature searches must compete with other branches of astronomy for scarce observing time. What researchers really hanker after is a large space telescope purpose-built to image Earth-like alien worlds—a new incarnation of the idea behind NASA's ill-fated Terrestrial Planet Finder.

The Habitable Exoplanet Imaging Mission, or HabEx, a mission concept now being studied by NASA, could be the answer. Its telescope would have a mirror up to 6.5 meters across—as big as the JWST's—but would be armed with instruments sensitive to a broader wavelength range, from the ultraviolet to the near-infrared, to capture the widest range of spectral biosignatures. The telescope would be designed to reduce scattered light and have a coronagraph and starshade to allow direct imaging of Earth-sized exoplanets.

Such a mission would reveal Earth-like planets at a level of detail researchers can now only dream about—probing atmospheres, revealing any surface pigments, and even delivering the sort of blocky surface images that Kane has been simulating. But will that be enough to conclude we are not alone in the universe? "There's a lot of uncertainty about what would be required to put the last nail in the coffin," Kane says. "But if HabEx is built according to its current design, it should provide a pretty convincing case."