Sen—A few weeks back I attended the 225th American Astronomical Society (AAS) meeting. Over 2000 astronomers descended upon Seattle, gathered to present the latest in the field. One of the talks that stood out for me this year was the one presenting the results from a recent paper on the confirmation of the first planet discovery from NASA's Kepler spacecraft now in its second life as the K2 mission.
Kepler is not new to the exoplanet scene. It launched in March 2009 and had stared nearly continuously until March 2013 at one patch of the sky, monitoring about 170,000 stars for the signature of exoplanets. If the orbit of a planet outside our Solar System is aligned just so to the Earth, then when the planet moves in front of its parent star or transits, it blocks a small fraction of the star's light detectable from Earth. Kepler was capable of detecting the 0.01% dip in starlight produced by an Earth-sized planet transiting a Sun-like star. The Kepler mission has been wildly successful, detecting over 3000 planet candidates (likely planets but the transit method only gives a size estimate so to be a bonafide planet requires a second independent discovery/validation technique). In fact at AAS, the Kepler team just announced the 1,000th Kepler planet confirmation.
But in May 2013, Kepler suffered a mechanical failure of the second of its four reaction wheels that kept it pointed with the exquisite precision necessary to detect the transits of Earth-sized planets. This malfunction ended Kepler's primary mission, and thus the K2 mission was born. In something that reads a bit like a movie script, the engineers came up with an idea to save Kepler's exoplanet hunting days. Engineers found that the gimpy spacecraft could use sunlight as a proxy for its failed reaction wheels if it pointed along the plane of the Solar System (known as the ecliptic). Kepler can't quite achieve the photometric precision it had before the reaction wheel failures, but this new observing mode gets close with the prospect of detecting small rocky exoplanets around M-dwarf stars, smaller and cooler than the our Sun.
Image credit: NASA Ames/W Stenzel
Now on top of the natural changes in the brightnesses of stars observed, the data from Kepler has bumps and wiggles in a saw tooth-like pattern repeating every 6-ish hours due to the spacecraft becoming unstable (the new observing configuration is quasi-stable), rotating and drifting from its desired pointing, with then Kepler's thrusters kicking in to the move the spacecraft back into position. Andrew Vanderburg, a graduate student based at Harvard, has developed a technique to try and take these systematic effects out of the Kepler observations. After applying their method to 9 days worth of test K2 engineering data taken last February, the Harvard group spotted a solitary dip in the observations of star HIP 116454 consistent with that of a 2.53 Earth-radii planet orbiting the star less than every 20 days. But with a single transit-like event, you can't be sure that it is caused by planet and not a false positive of one kind or another. They couldn't go back and watch the star again with Kepler to see if the signal repeats as it should if there is indeed a planet. At this point Kepler was already observing new stars. The nature of the Kepler's K2 observing scheme is that the spacecraft observes a new portion of sky for 75-80 days before moving on to the next field along the ecliptic to keep the solar panels illuminated and prevent sunlight from hitting Kepler's mirror.
To confirm what that they had seen was due to an exoplanet, the Harvard group set out to use HARPS-N (High Accuracy Radial velocity Planet Searcher for the Northern hemisphere) spectrograph among other ground-based telescope follow-up. If there was a large enough planet or a star orbiting HIP 116454, HARPS-N should detect it and measure its mass. One of the upsides of the K2 mission is that Kepler can now observe many very bright stars whilst ground-based telescopes can do follow-up for the largest of these planets to measure masses, confirm discoveries, and potentially detect and characterize atmospheres. Most of the original Kepler mission stars are too faint or the planet candidates are too small to be able to do these observations with the current technology. If there is indeed a body in orbit around HIP 116454, it should gravitationally tug on the star as it orbits, moving it to and fro just a little bit. That motion is observed in radial velocity observations as the red shifting and blue shifting of the star's light that can be measured by HARPS-N. And indeed HARPS-N detected a nearly 12 Earth-mass planet orbiting HIP 116454 every 9.1 days!
So there is a planet present, but how do you know that's the culprit transiting in front of the star that K2 observed? There are known exoplanets detected in radial velocity observations that do not transit their star. Vanderburg and collaborators turned to the space-based MOST (Microvaribiltiy and Osciltations of STars) telescope, a suitcase-sized Canadian satellite that is capable of monitoring a single star for transits. Because the star is so bright, they were able to point MOST at HIP 116454 and wait to see a transit at the times predicted from the K2 observations and the radial velocity measurements. They had to observe a few transits, but the there is a weak signal. They detected the signature of a super-Earth sized body as predicted! The single solitary dip spotted in 9 days worth of K2 engineering data is real and is the same planet detected by HARPS-N, making this K2's first confirmed planet discovery.
This detective story hints at the exciting future prospects and capabilities of the K2 mission in the field of exoplanets and planetary science. Kepler is going strong with K2 observations continuing right now. This discovery is the first of many new planets and planet candidates to come in the K2 era, and just the other day the first multi-planet system discovered by K2 was announced. Long live K2!