Artist's conception of an Earth-sized Kepler planet around a cool star. Image Credit: NASA/Ames/JPL-Caltech.

Aug 21, 2015 EPIC-206: A nearby cool star with two Earth-sized planets

Sen—Among other amazing discoveries, the Kepler mission has shown us that small planets are extremely common. Of the stars that Kepler and other missions targeted, astronomers have found that 26 per cent of Sun-like stars have an Earth-sized planet that orbits close to the host star. The occurrence rate of small planets on short orbits shoots up for the even smaller and cooler M-dwarf stars, with an average of one to two of these planets per star. Theoretically, finding planets that orbit M-dwarf stars should be easy: Because the stars are dimmer and less massive, transit and radial velocity signals will appear larger, which allows discovery of smaller planets at better precision.

In practice, however, finding planets around M-dwarfs is rather challenging, for the simple fact that M-dwarfs are generally very faint. While Kepler surveyed thousands of M-dwarfs, the mission only found about 160 planet candidates around those stars despite the fact that they should all have small planets. Stars in the Kepler field are, on average, over 1,000 light-years away so the stars appear very faint and many potential transit signatures would have been too small to be detected by Kepler. That also means that the few planets found around M-dwarf Kepler stars are very hard to follow up with measurements from other telescopes. The obvious solution—targeting closer M-dwarfs so that they are brighter—has its challenges as well, as nearby M-dwarfs are spread very thin across the sky.

That is precisely what makes EPIC-20601691 (EPIC-206) such an exciting discovery. EPIC-206 is an M-dwarf star only 212 light-years away that hosts two transiting Earth-sized planet candidates. This discovery is part of a larger campaign proposed by lead author on the discovery, Dr Erik Petigura of Caltech, to measure the occurrence rate of small planets using the Kepler telescope in its K2 format. The system is named after ESA's European Photon Imaging Camera (EPIC) onboard the XMM-Newton satellite that first imaged it, and the star just happened to fall into the observing field of K2's third campaign which lasted from Nov. 14, 2014 until Feb. 3, 2015.

The observations of EPIC-206 fulfil all of the necessary requirements to accurately measure small planets: Around a cool star, nearby, observed with the best planet-finding telescope ever made, and capable of being followed up by either ground- or space-based telescopes. EPIC-206 has two planets, both in orbits too close their star to be habitable. EPIC-206 b, only 60 per cent larger than Earth, orbits in 9.3 days, and EPIC-206 c, 90 per cent larger than Earth, orbits in 15.5 days.


The top shows the total transit light curve of EPIC-206, with transits from planets b and c marked by the red and blue ticks, respectively. The bottom shows the folded transit light curves of EPIC-206 b (left) and EPIC-206 c (right) in black points, with the best fitting transit model overlaid (red and blue curves). The transit models indicate that the two planets are super-Earths or mini-Neptunes and orbit well inside the habitable zone.

Detecting these planet candidates with Kepler's K2 allowed for precise measurements of the properties of the planets in relation to the star. Since most Kepler stars are too far away to accurately measure the stellar properties, the planetary properties are dependent on the best guess for some of the star values. But as EPIC-206 is so nearby, Petigura and his collaborators were able to follow up on this system with additional measurements from three other instruments as well as going back into data archives to get even more information. This allows them to get very accurate values for the star EPIC-206. While EPIC-206 b and c are still only planet candidates, they are some of the most well-measured planet candidates on record.

The first set of additional measurements were taken using the SpeX spectrograph on the 3.0 meter NASA Infrared Telescope Facility (IRTF). This instrument allowed them to measure the spectrum of EPIC-206 in the near-infrared. M-dwarfs emit light primarily in the infrared, so they could observe the most interesting and telling features of the star's spectrum to determine its exact properties. Next, they obtained additional stellar spectra using the HIRES spectrograph at the Keck Observatory. HIRES observes the entire visible part of the electromagnetic spectrum, from just past violet to just past red, and so was able to obtain information on a different part of EPIC-206's spectrum.

The SpeX and HIRES spectra were combined and then compared to stellar spectra from "standard" stars; that is, stars whose spectra have already been well measured and matched to a set of stellar parameters like spectral type, mass, radius, and temperature. The best fitting "standard" spectra shows that EPIC-206 is an M0 star with a radius 60 per cent of the Sun, a mass 64 per cent of the Sun, and a temperature over 2,500 Kelvin cooler. Comparing the intrinsic brightness of EPIC-206 to the observed brightness revealed that the planetary system is only 212 light-years from us.

A common problem when detecting planets via the transit method are other light sources near a star mimicking a transit signal or blending in with the target star. The first scenario is often caused by a background eclipsing binary star system, where a binary star system behind the target (often very far behind) has the stars eclipsing each other, causing periodic dimming and brightening when the cooler star blocks light from the hotter star in the background system. If the eclipsing binary system appears close to the target star the light from the two systems can blend together, making it appear as if the target system has a planet. Kepler in its primary mission detected 2,165 eclipsing binary star systems. It takes a lot of follow-up observations to rule out these false-positives and makes the process of confirming a planet much harder.

Petigura and collaborators rule out the possibility of a false positive mimicking the transits of EPIC-206 b and c with a combination of new observations taken in the infrared with the NIRC2 camera at Keck Observatory and archival images from the Digitized Sky Survey taken in 1954 and 1991. In both the new and old images they searched for any nearby stars and determined that there were no nearby stars that could have produced a fake transit signal like the transits seen of EPIC-206 b and c. From this they are able to conclude that the planets are real.

While EPIC-206 b and c are much too close to their host star to be habitable, this system is notable for other reasons. EPIC-206 emits a relatively low amount of light in the ultraviolet, so even though the planets are close to the star they do not receive all that much stellar irradiation. In fact, EPIC-206 is the brightest star known to host a planet smaller than twice the Earth's size that orbits in less than 10 days and receives less than 20 times the stellar irradiation of Earth. Usually small planets with orbital periods less than 10 days are at risk for the intense starlight blasting away surface material, so EPIC-206 b is a rare case where mass-loss is not a factor, and so should be studied further .

The exact orbital periods of EPIC-206 b and c are also interesting and may contain clues to the planetary formation and orbital evolution. The inner and outer planets orbit near a 5:3 resonance, that is, the inner planet orbits five times for every three times the outer planet orbits. When two planets are in resonance they are capable of strong gravitational interactions that can lead to changes in the planetary orbits. We see the effects of orbital resonance in our Solar System in the main asteroid belt: a few orbital locations in the asteroid belt are in resonance with Jupiter and if an asteroid is nudged into resonance the gravitational interactions quickly destabilize the orbit and fling the asteroid away. This is how we get a lot of our near-Earth asteroids.

In the case of planets, and EPIC-206 b and c in particular, the near-resonance of the orbits means that the planets could interact very strongly with each other despite their small sizes and low masses. This could lead to transit timing variations that are observable from the ground and allow us to calculate the actual masses of the planets. We could also obtain mass measurements from ground-based radial velocity instruments, and these planets are perfect targets for the James Webb Space Telescope to take spectra of the planets' atmospheres if they exist.

As you can probably tell, a lot of work goes on in the process of detecting and following up on planet candidates. Most of this happens after the initial discovery rather than in tandem with it. Now that Petigura and collaborators have gotten most of the additional observations out of the way, it should only be a short time until EPIC-206 b and c are confirmed planets and join the growing ranks of planets discovered by K2.


Two common types of astrophysical phenomena that can masquerade as a planetary transit are grazing eclipsing binaries (left), where a pair of stars orbit each other, and background eclipsing binaries (right), where a distant binary star system is aligned very close to the star of interest. These require significant amount of ground-based observations to rule-out using radial velocity techniques. Image credit: NASA/Kepler