However, Kepler’s greatest legacy in the public imagination is its census of exoplanets: planets orbiting other stars. When a planet transits its star — passing between the star and Earth — it blocks a tiny amount of light on each orbit. Patient monitoring of the Kepler kind reveals both the length of the explanet’s year and its size. As with other planet-hunting methods, detection using transits is easiest for very large planets orbiting close in to their stars, but Kepler was both sensitive enough and observed the same stars long enough to identify a number of much smaller worlds.
Researchers will continue analyzing Kepler data for some time, but a paper published yesterday by Erik Petigura, Andrew Howard, and Geoffrey W. Marcy provided a summary of what we’ve learned about the census of alien worlds. Specifically, they focused on the possibility of exoplanets somewhat like Earth orbiting within their host star’s habitable zone: the region in which liquid surface water could conceivably exist.
The results: out of the approximately 150,000 stars in Kepler’s field of view, astronomers identified over 3000 possible planets. Each of those requires follow-up observation to confirm they actually are planets, something Petigura (a graduate student at University of California, Berkeley) and others worked on using the Keck telescope in Hawaii. They found that out of 603 exoplanets in their sample, 10 were roughly Earth-sized (with radius between 1 and 2 times Earth’s radius) and received approximately 1/4 to 4 times the amount of radiation Earth gets. (I’ll come back to the meaning of those values shortly.)
The researchers estimated that perhaps as many as 22 percent of Sun-like stars harbor Earth-class planets with orbits in a similar range to Earth’s: more than 200 but less than 400 days. These figures were obtained both from the observed planets and from models of star systems, to compensate for the fact that Kepler isn’t very good at spotting Earth-sized planets that far out from their host stars. (Lest you think that’s a criticism, I’d say it’s necessary to make that kind of model. Nobody in their right mind would claim we’ve identified every planet in Kepler’s field of view: there are some we can’t detect yet, but we can infer their existence based on other factors.)
The central conclusion is worth repeating: if these estimates are correct, then more than one in five Sun-like stars harbors an Earth-size planet in its habitable zone. Of course, this says absolutely nothing about whether the planet is actually habitable. Venus is an Earth-size planet in the Sun’s habitable zone, yet it is much hotter and rains sulfuric acid. Conversely, Europa and Enceladus are icy moons orbiting Jupiter and Saturn (respectively), but they could be reasonable candidates for life despite being very unlike Earth. Mars falls in a middle category: smaller than the planets Petigura and collaborators considered, but showing signs of a wetter and warmer past.
Nevertheless, Earth is the only planet where we know life exists, so it’s our best template when we look at other star systems.
What do we mean by Earth-like?
Most of the planets we’ve found that are roughly Earth-size are still noticeably bigger. (Sometimes we call them super-Earths, though don’t think of capes.) The paper I’m discussed considers planets with radius 1 to 2 times Earth’s, which conceivably have similar density and composition to our planet. The weakness of the transit method is that, while we can determine sizes of planets, we can’t find masses without more data. That means we can’t be very clear on what the planet is made of, though there are many good reasons to think that small planets are rocky.
Beyond the examples of Mercury, Venus, Earth, the Moon, and Mars, I wrote about a newly identified exoplanet where we do have density data last week. The exoplanet Kepler-78b has a radius about 20 percent larger than Earth’s, and has about 70 percent more mass, matching Earth in density, but it orbits so close to its star that no atmosphere is possible…and it probably has a surface of liquid rock.
A planet with double Earth’s radius might have similar composition and somewhat comparable surface gravity, so it might still qualify as Earth-like even with four times the surface area. However, the key aspect that makes Earth habitable apart from its distance from the Sun is its atmosphere: we have enough greenhouse gases to trap heat, but not so much to turn us into Venus. We only have partial atmospheric data on a small number of exoplanets, so we have to make assumptions. That gets complicated, which is why this paper used a simplified model.
The total amount of light a planet receives is called the flux. The flux depends on the amount of light emitted by the star, the distance between the star and the planet, and the size of the planet itself. If you moved Earth out twice the distance from the Sun, it would receive 1/4 the amount of flux. However, if you moved Earth out that far and doubled its radius, you’d get the same amount of light. (The diagram at right might help clarify that; see also my earlier post on the inverse square law.)
That’s why the exoplanet census paper considered flux as a proxy for habitability: they could take into account the size of the planet and distance from the star, while glossing over stuff we can’t know yet, like atmospheric composition.
If these estimates are right and 22 percent of Sun-like stars harbor Earth-size planets at a hospitable distance, then future observations should find them. As Phil Plait pointed out, this could mean a huge number of Earth-like planets, not least since Sun-like stars aren’t the most common type out there. The census of exoplanets is just beginning, and who knows what we might yet find?
- Now that it’s unstable, Kepler now tumbles slowly, in part because it’s pushed by light from the Sun. The telescope can still be used because all its scientific instruments are still in order, but the days of steadily watching one part of the sky for exoplanets are over.