How a "Planet Score" Helped NASA Identify 1,284 New Exoplanets in One Fell Swoop

It's all about astrophysical false positive probability calculations.


Before Tuesday, there were no shortage of theories about what NASA’s discovery announcement would entail. (Full disclosure: I was responsible for much of that speculation.) Then Tuesday hit and we found out exactly what the big news was: NASA scientists just confirmed the identify of 1,284 new exoplanets in the universe — including nine planets that have the potential to be habitable to life.

It’s an announcement that has already inspired scientists and ordinary individuals around the world to ponder whether we might seriously find extraterrestrial life soon enough. But the new study raises an interesting question: what changed between the last few years and now that allowed scientists to identify so many new exoplanets all at once? Did all of these planets just show up at once? Did we develop better technology? Did the Kepler Space Telescope miraculously get better (after weirdly almost breaking down)? What gives?

The answer: It all comes down to a new method of validating exoplanet candidates that provides “astrophysical false positive probability calculations” for such objects, according to a new paper published in the latest issue of The Astrophysical Journal. Basically, the new method ascribes a number to every object found by Kepler that determines the likelihood that object is an exoplanet, and not an “imposter.” Call it a planet score. The higher the number, the more likely it’s a planet.

The new method only allows an object to move from the “candidate” category to “exoplanet” if Kepler researchers can say so with 99 percent reliability or higher.

This is an artist's conception of Kepler-20e, the first planet smaller than the Earth discovered to orbit a star other than the sun. A year on Kepler-20e only lasts six days, as it is much closer to its host star than the Earth is to the sun.


We should slow down at this point and expound on exactly how astronomers find and evaluate potential exoplanets. Basically, through Kepler and a few other instruments, scientists stare at distant stars and measure the brightness of light emitting from those balls of fiery energy. When a star has a planet in orbit, its brightness will dim as that planet transits past it in relation to the telescope we’re using to watch it (a recent, albeit small, example is Mercury passing in front of the sun). As long as that dimming isn’t just a technical error, it’s a sign that something is passing through the neighborhood. A consistent dimming occurring regularly over time is further evidence it might be a planet.

In the past, scientists had to pore over the brightness numbers along with assessing a variety of different data that might be attainable, like radio velocity observation or high-resolution imaging. Unfortunately, doing that kind of work is extremely time consuming, and we don’t always have the resources to find what we need.

So in this day-and-age, we turn to computers for help. Timothy Morton, a Princeton researcher who studies exoplanets, developed a new method for exoplanet validation that combines previous exoplanet observations and the current brightness measurements scientists are gathering with Kepler.

There are two kinds of simulations. The first looks at how the dimming compares to that from known exoplanets and imposter objects. The second goes a step further and deduces whether dimming is indicative of exoplanet behavior given what we already about how exoplanets are distributed and laid around the Milky Way.

The two simulations are used to determine the statistical likelihood the object in question is an exoplanet. It’s a faster way of doing this work — and by all accounts, it’s even more accurate. In fact, the method is actually being used to verify previously confirmed exoplanets and determine whether they might actually be false-positives.

This is crucial for the direction of future exoplanet research. The work accomplished since Kepler’s launch in 2009 has been huge in illustrating just how many other worlds exist in the universe — and it has given humans a staggering amount of hope we may find another habitable planet, or even alien life.

NASA is already getting ready to launch the Transiting Exoplanet Survey Satellite (TESS) in late 2017, and the James Webb Space Telescope in 2018. Both will play a pivotal role in exoplanet investigations by acquiring lots more data that we’ve ever dealt with. Morton’s model will help our scientists on the ground sift through that data and identify potentially habitable exoplanets faster than we could have hoped.

Related Tags