Even though Kepler’s primary mission ended three years ago, the data it collected just revealed a mother lode: 1,284 newly confirmed planets.
Over the course of four years, the Kepler spacecraft stared at a tiny patch of sky teeming with stars and tallied up 4,696 exoplanet candidates. But for years now, many have remained just that: candidates. To confirm any one planet requires difficult, time-consuming follow-up observations. And the smaller candidates are out of reach of even the largest ground-based telescopes. After years of follow-up observations, the list of confirmed planets stood at 1,041 — nothing to sneeze at, but nowhere near complete.
Now that list has just doubled in size. With the newest study from the Kepler team, the confirmed planets now number 2,325.
Timothy Morton (Princeton University) and colleagues accomplished that feat by taking a different tack than their predecessors. Acknowledging that follow-up observations wouldn’t be possible for all planet candidates, the team didn’t look for additional evidence supporting a planet’s existence. Instead, they searched for signs that the planet wasn’t there. Then they calculated the chance that what looked like a planet was actually an imposter. Any candidate with less than a 1% chance of being fake is now considered real.
When astronomers look for transiting planets, they never actually see the planet itself. Instead they see a dip in the host star’s light as the planet passes in front of it. But other astrophysical sources can mimic that dip. The most common imposter is an edge-on pair of stars whose mutual orbit blinks on and off in the background — an eclipsing binary.
Morton and his team took a two-pronged, fully automated approach to determining whether a signal was an imposter or a real planet. Their code first examined the signal itself — a real planet transit will block the star’s light in a certain way that can sometimes be distinguished from imposters. Then the code factored in how common imposters might be. For example, how common are binary stars, and how many lie in the direction Kepler was looking?
The algorithm studied each planet candidate for several minutes, then spit out a probability that the signal is a fake. And as long as that probability was less than 1%, the candidate moved to the confirmed list.
All in all, 1,935 planets were confirmed in this way, including 651 that had already been confirmed by some other method. And 428 candidates (generally those with large radii, whose signals are more easily faked by other astrophysical sources) were flagged as imposters.
Several thousand candidates remain in limbo, not quite flagrant enough to be flagged as fakes, but not quite convincing enough to be confirmed either. These candidates await improvements to Morton’s algorithm, follow-up observations, or both.
In the Habitable Zone
Of the newly confirmed planets, nine orbit in their star’s habitable zone, the region in which a planet with an Earth-like atmosphere could sustain liquid water on its surface. Add that to the dozen already known, and astronomers have assembled a decent size of potentially habitable planets.
It’s simply amazing that the Kepler team is still returning a bonanza of exoplanet results almost three years after two reaction wheel failures blew the primary mission. But now we’re nearing the end. The mission will close out in a year and a half, says mission scientist Natalie Batalha (NASA Ames), though we can expect one last Kepler catalog sometime next year. The secondary K2 mission is still going strong and will likely end in mid-2018, when the spacecraft runs out of fuel.
“We’re getting ready to pass the baton to future missions,” Batalha adds, namely, the dynamic duo: the Transiting Exoplanet Survey Satellite (TESS) and the James Webb Space Telescope (JWST).