The First Year of Kepler’s K2 Mission

Without a doubt, the most prolific planet hunter has got to be NASA’s Kepler mission. Launched into solar orbit on March 7, 2009, Kepler spent four years continuously observing the brightness of almost 200,000 stars in a single patch of sky covering 115 square degrees which straddled the border of the constellations Cygnus and Lyra looking for small dips in star brightness indicative of planetary transits. To date, there have been over one thousand confirmed extrasolar planet detections by Kepler with many more unconfirmed “planet candidates” currently the subject of follow up observations. Analysis of the full data set from Kepler’s primary mission suggests that the final tally of extrasolar planets may reach 20,000 (see “First Look at Kepler’s Complete Primary Mission Data Set”). Unfortunately, after the second of Kepler’s four reaction wheels stop functioning in May of 2013, Kepler was no longer able to point accurately at its target star field bringing its primary mission to an end.


Kepler’s Extended Mission

To get around the loss of the two reaction wheels, a new extended mission with a different mode of operation was devised by Kepler’s team of scientists and engineers. Instead of staring at a single patch of sky, Kepler now observes a sequence of star fields along the ecliptic. Each “campaign” lasts about 80 days before Kepler moves on to the next star field. With this new observation strategy, Kepler can track each star field with sufficient accuracy to continue planet hunting using its remaining pair of operating reaction wheels in conjunction with the slight pressure of sunlight reflecting off of the spacecraft. Although this approach is not as sensitive as that used in the primary mission and can only detect planets with orbital periods shorter than about a month instead of about a year or more, this approach promises to sample a broader range of stars looking for extrasolar planets. In addition, Kepler can now be directed towards star fields that contain brighter stars that are more amenable to follow up observations than existed in the primary mission’s field. Called “K2”, Kepler’s extended mission with its new observation strategy officially started in May 2014.

Diagram showing the major components of NASA’s Kepler spacecraft. (NASA/Kepler Mission/Ball Aerospace)

By the beginning of 2015, members of the Kepler mission science team started announcing discoveries from the K2 mission. Among the first was a trio of planets orbiting the red dwarf called EPIC 201367065 announced on January 16, 2015 by a team led by Ian Crossfield (Lunar & Planetary Laboratory – University of Arizona). Crossfield et al. believed that the outermost planet they found in this system is potentially habitable, although an independent analysis suggests that this claim may be somewhat overstated (see “Habitable Planet Reality Check: Kepler’s New K2 Finds”). Another find that some have claimed may be among the most Earth-like extrasolar planets known is EPIC 201912552b announced in a paper with Benjamin T. Montet (Harvard-Smithsonian Center for Astrophysics) a the lead author. Unfortunately given the measured size of this world, even Montet et al. believe it is much more likely that this planet is a mini-Neptune (see “Habitable Planet Reality Check: EPIC 201912552b”). Despite the dubious prospects for these worlds being habitable, future study promises to shed much light on the limits of planetary habitability and demonstrate the promise of the K2 mission.

Diagram showing the star fields for Kepler new K2 extended mission. Click on image to enlarge. (NASA/Ames Research Center)

While there have been scattered discoveries from the K2 mission announced over the past several months, there were also systematic efforts to analyze the new K2 data set for planet candidate. In a paper recently accepted for publication in The Astrophysical Journal Supplement with Andrew Vanderburg (Harvard-Smithsonian Center for Astrophysics) as the lead author, an initial analysis of the Kepler data set from the first year of the K2 extended mission is presented.


The First K2 Campaigns

For this paper, a total of four campaigns of data acquired from March 2014 to February 2015 were analyzed. What is called “Campaign 0” was actually designed to be a full length shakedown of the proposed K2 operating mode conducted prior to the official start of the extended mission. Kepler observed a star field near the galactic anti-center (i.e. the direction opposite of the galactic center as viewed from the Earth) centered at right ascension 6:33:11.14, declination +21:35:16.40 in the constellation Gemini. Included in this field were the open clusters M35 and NGC 2158. Kepler observed this field for a total of 80 days between March and May 2014. Unfortunately the bright planet Jupiter passed near this field causing fine pointing control issues during the first half of the campaign affecting the photometric accuracy in the process. As a result, Vanderburg et al. focused their initial analysis on just the last 33 days of data. The concentration of giant stars and the higher than average temperatures of the typical stars in this field resulted in more false positives than usual for the candidates in this star field.

The star field used for Campaign 0 which proved the operating concepts of the K2 mission. The ecliptic and positions of M35 and NGC 2304 are indicated. Click on image to enlarge. (NASA/Ames Research Center)

Campaign 1 was the first official observing run of the K2 mission. During this time Kepler observed a star field centered at right ascension 11:35:45.51, declination +01:25:02.28 near the border of the constellations Leo and Virgo. This field was near the northern galactic pole and was observed for a total of 83 days between June and August 2014. Vanderburg et al. ignored the first two days of data from this campaign as Kepler settled into its final observation attitude. There was also a gap of three days in the middle of the campaign as Kepler turned its antenna back towards the Earth to transmit data. Because of the low density of stars in this field, the photometric data were of high quality with a low false alarm rate.

The star field used for Campaign 1 of Kepler’s extended K2 mission with the position of the ecliptic indicated. Click on image to enlarge. (NASA/Ames Research Center)

The second star field observed during the K2 mission was centered at right ascension 16:24:30.34, declination -22:26:50.28. This star field for Campaign 2, which covered 79 days between August and November 2014, straddled the constellations Libra and Scorpius near the galactic center. This crowded star field included the globular clusters M4, M19 and M80, the Upper Scorpius subgroup of the 11-million year old Scorpius-Centaurus association as well as the ρ Ophiuchus cloud complex – the closest star forming region to our Solar System at a distance of about 430 light years. Unlike the earlier K2 campaigns, there was no mid-campaign data download. Only 14 hours of data near the middle of Campaign 2 were omitted by Vanderburg et al. from their analysis as Kepler shifted and settled into its new roll direction. Mars passed though Kepler’s field of view during this campaign but did not cause any major problems as had happened earlier with Jupiter. The reddening of stars caused by dust associated with the ρ Ophiuchus cloud complex complicated target selection somewhat.

The star field used for Campaign 2 of Kepler’s extended K2 mission. The position of the ecliptic as well as selected bright stars and members of the Upper Scorpius subgroup are indicated. Click on image to enlarge. (NASA/Ames Research Center)

Campaign 3 was the final K2 observation run analyzed by Vanderburg et al. and was centered at right ascension 22:26:39.68, declination -11:05:47.99. Observed for 69 days between November 2014 and February 2015, it viewed the region near the south galactic pole near the constellations of Aquarius and Capricornus. As before, Vanderburg et al. ignored the first day of data as Kepler settled into its new attitude. As in Campaign 2, there was no mid-campaign data download and this time the change in roll direction was gentle enough that Vanderburg et al. did not need to exclude any data from their analysis. Changes in the way attitude control adjustments were made led to a noticeable improvement in the photometric precision of Kepler’s measurements during Campaign 3.

The star field used for Campaign 3 of Kepler’s extended K2 mission. The positions of Neptune and its moon, Nereid, are indicated as are other objects of interest. Click on image to enlarge. (NASA/Ames Research Center)

During this campaign, Neptune also passed through the field but caused no major issues due to its relative dimness. Data from Neptune’s passage have been assembled into a video which was publicly released in May 2015 and is in the Related Video section near the end of this article. Neptune appears on Day 15 along with its closely orbiting moon, Triton, which has an orbital period of 5.8 days. Appearing from the left at Day 24 is the tiny moon Nereid in its slow 360-day orbit around the planet. A few asteroids appear as fast-moving streaks across the K2 field of view.

Closeup of a Kepler image showing Neptune and its two brightest moon, Triton and Nereid. (NASA Ames/SETI Institute/J. Rowe)


Summary of Results

After processing the data from a total of 59,174 targets observed in the first four K2 campaigns and vetting them for false positives as well as other issues, Vanderburg et al. found a total of 234 planetary candidates orbiting 203 stars. The team performed follow up observations including conducting initial “reconnaissance spectroscopy” of the brightest 68 targets using TRES (Tillinghast Reflector Echelle Spectrograph) on the 1.5-meter reflector at the Whipple Observatory on Mt. Hopkins, Arizona. These data allowed the stars’ properties to be more accurately determined (thus improving the derived planetary properties) as well as eliminate the possibility these stars were orbited by stellar companions. The candidate planets ranged in size from gas giants to sub-Earths with orbital periods measuring from hours to just over a month.

A detailed statistical comparison of this initial batch of K2 finds with planets found during Kepler’s primary mission show that the latter tend to have somewhat more small planets. This is likely to be due to the longer observing time during the primary mission which made it easier to detect smaller planets. The distribution of orbital periods is also similar except for fewer planets with very short orbital periods in the K2 finds. This is likely due to the processing techniques employed by Vanderburg et al. on the K2 data set which tends to make it more difficult to detect planets with orbital periods less than about 2.5 days.

Histograms comparing the K2 and Kepler primary mission finds as a function of planet radius (top) and orbital period (bottom). Click on image to enlarge. (Vanderburg et al.)

This initial survey by Vanderburg et al. found a total of 26 sub-Neptune size planet candidates in the 1 to 4 Earth radii (or RE) range orbiting stars with Kepler magnitudes brighter than 12. Ten of these candidates have radii greater than 1.6 RE and are likely to be volatile-rich mini-Neptunes. It should prove possible for the masses of many of these two dozen or so planets to be determined or at very least constrained by future precision radial velocity measurements. Bright stars like these may also be good targets for future spectroscopic detection of the atmospheres of their transiting planets. A total of 8 sub-Earths with radii estimated to be as small as 0.75  RE were identified. Future study of these sub-Neptune size planets will be useful in more accurately determining the mass-radius relationship of planets in this size range as well as the properties of these worlds (see “Habitable Planet Reality Check: Terrestrial Planet Size Limit”).

In addition to new finds, Vanderburg et al. detected five planets that had been found earlier by the WASP (Wide Angle Search for Planets) and HATNet (Hungarian Automated Telescope Network) ground-based surveys. The planets discovered earlier by Crossfield et al. and Montet et al. were also detected along with a previously announced hot Jupiter, EPIC 204129699, along with other interesting objects. The promising start of the K2 extended mission and the improvements made in the photometric accuracy of the measurements starting with Campaign 3 bodes well for what future K2 campaigns will find.


Follow Drew Ex Machina on Facebook.


Related Video

Here is a brief video created from Kepler data showing Neptune with its two brightest moons, Triton and Nereid, moving through the field of view.


Related Reading

“Habitable Planet Reality Check: Kepler’s New K2 Finds”, Drew Ex Machina, January 20, 2015 [Post]

“Habitable Planet Reality Check: EPIC 201912552b”, Drew Ex Machina, May 12, 2015 [Post]

“First Look at Kepler’s Complete Primary Mission Data Set”, Drew Ex Machina, January 26, 2015 [Post]


General References

Ian J. Crossfield et al., “A Nearby M Star with Three Transiting Super-Earths Discovered by K2”, The Astrophysical Journal, Vol. 804, No. 1, ID 10, May 2015

Benjamin T. Montet et al., “Stellar and Planetary Properties of K2 Campaign 1 Candidates and Validation of 17 Planets, Including a Planet Receiving Earth-like Insolation”, The Astrophysical Journal, Vol. 809, No. 1, ID 25, August 2015

Andrew Vanderburg et al., “Planetary Candidates from the First Year of the K2 Mission”, arXiv 1511.07820 (accepted for publication in The Astrophysical Journal Supplement), November 24, 2015 [Preprint]

“Kepler Observes Neptune Dance with Its Moons”, NASA Press Release, May 14, 2015 [Press Release]