PORTLAND, Ore.—In a matter of months after its launch two years ago, the Kepler spacecraft had made more planet sightings than in the entire history of astronomy—1,235 and counting. So far, only 17 have been confirmed, but scientists at the National Aeronautics and Space Administration (NASA) are confident that 80 percent will eventually be verified.
That's 988 planets spotted in just 1/400th of the sky by NASA's first mission dedicated to finding planets—almost twice as many as have been confirmed in the entire history of astronomy, 564.
Before Kepler, astronomers could only infer the presence of planets orbiting distant stars (called "extra-solar planets" or exoplanets for short), usually by measuring their gravitational influence on the motion of their parent suns. A few were also spotted because of their unique characteristics, such as being unimaginably large, or super-hot, or because they were located in an area that is easier to see, such as at the focal point of a gravitational lens. Most planets, however, are beyond the capabilities of any ground-based telescope, and even those observable from orbiting telescopes like the Hubble can only be seen a specific times of day and year—making a comprehensive search impossible.
This artist's conception depicts the Kepler-10 star system, located about 560 light-years away near the Cygnus and Lyra constellations.
Kepler was specifically designed to overcome all these problems, which it has done beautifully, despite the enforced austerity of its Discovery-class budget ($500 million). A few newer methods of detecting planets are being developed at NASA, such as a technique for blocking a star's light with a vector-vortex coronagraph and an analytical method that cancels out a stars glare using nulling interferometry, but for now the best way to detect exoplanets is Kepler.
"Kepler is the most exclusive photometer that has ever been built," said Jon Jenkins, an EE who is the Kepler Mission's co-investigator for data analysis at NASA. "We have been able to a pull out these very weak signals by creating a very low noise environment with a combination of software algorithms that we use both on-board and on the ground to process and analyze the data."