Fast-Paced Algorithms Help Identify Nearly 80 Potential Exoplanets in Record Time

The planet sits relatively close to Earth (by astronomical standards) at 35 parsecs/114 light years away. The researchers hope this proximity as well as it orbiting an extremely bright star could make it the subject of more intensive studies. 

“We think it would probably be more like a smaller, hotter version of Uranus or Neptune,” said Ian Crossfield, an assistant professor of physics at MIT. Crossfield co-led the study with graduate student Liang Yu.

This discovery is impressive both in the amount of information gathered about potential exoplanets and also in the speed with which this information was gathered. The researchers used tools already developed by other MIT researchers to quickly scour graps of light intensity and analyze lightcurves from the 50,000 stars monitored by K2. Normally, this analysis takes anywhere from several months to an entire year. The MIT team gathered the information in just weeks. 

The MIT team hopes to apply this speed when they start getting data from NASA's Transiting Exoplanet Survey Satellite -- otherwise known as TESS, 'the Exoplanet Hunter.' TESS will monitor nearby stars in month-long stretches and ultimately cover the entirety of the night sky.

[see-also]

“When the TESS data come down, there’ll be a few months before all of the stars that TESS looked at for that month ‘set’ for the year,” Crossfield said. “If we get candidates out quickly to the community, everyone can start immediately observing systems discovered by TESS, and do a lot of great planetary science. So this [analysis] was really a dress rehearsal for TESS.”

There were two major campaigns studied by the MIT team. The C16 and C17 campaigns, both analyzed a single patch of sky for 80 days. The telescope itself orbits as the Earth travels around the sun. K2 has been referred to as a "rear-facing" orientation, acting as a kind of rear-view mirror for seeing stars. 

When the campaigns wrapped up in February 2018, the mission released the data to astronomy's community of researchers. Crossfield and Yu set to work sifting through the data with algorithms designed to narrow down the high volume of stars into the ones most interesting. 

Once the team whittled down the stars, they then analyzed the images themselves. They looked for dips in starlight that could indicate a planet passed through the light. The C16 analysis gave them a potential 30 "highest-quality" planetary candidates, and C17 proved even more fruitful. 

“Our experience with four years of K2 data leads us to believe that most of these are indeed real planets, ready to be confirmed or statistically validated,” the researchers wrote in their paper.

Crossfield and others in the astronomy community hope these algorithms can further be improved to become even more specific to various stars and potential planets.

“You want to observe [candidates] again relatively soon so you don’t lose the transit altogether,” Crossfield said. “You might be able to say, ‘I know there’s a planet around that star, but I’m no longer at all certain when the transits will happen.’ That’s another motivation for following these things up more quickly.”