Characterizing K2 Candidate Planetary Systems Orbiting Low-mass Stars. II. Planetary Systems Observed During Campaigns 1–7

Since 2014, the NASA K2 mission has been using the Kepler spacecraft to search for transiting planets in 100 sq. deg. fields along the ecliptic plane. K2 observes 10,000–30,000 stars in each field for roughly 80 days before moving onto the next field. The placement of each field is driven by the three primary Requirements of the extended mission design: (1) K2 must look along the ecliptic plane so that the torque from solar radiation pressure is balanced, (2) sunlight must illuminate the solar panels, and (3) K2 cannot look so close to the Sun that sunlight illuminates the detectors (Howell et al. 2014; Van Cleve et al. 2016).

During the main Kepler mission, the planet search targets were primarily selected by the Kepler Science Office with a small contribution from Guest Observer proposals. In contrast, all K2 targets are nominated by members of the community. Although some of the selected target stars are well-characterized, many have poorly estimated properties constrained by only photometry, proper motion, and (when lucky) parallax. The problem of inadequate stellar characterization is particularly dire for the smallest, coolest target stars around which K2 has the highest sensitivity to transiting planets. In order to improve the characterization of low-mass K2 target stars,²⁰ we are conducting an extensive spectroscopic survey of potentially low-mass stars. In the first paper in this series (Dressing et al. 2017, hereafter D17), we presented near-infrared (NIR) spectra and determined stellar parameters for 144 potentially low-mass stars observed by K2 during Campaigns 0–7 (2014 May 30–2015 December 26).

In this paper, we use our previously determined stellar parameters combined with planet candidate lists to generate a catalog of K2 planetary systems orbiting low-mass stars. The structure of the paper is as follows.

In Section 2, we explain the origins of our planet candidate sample and describe the K2 planet candidate lists from which we selected our targets. We then consult the catalog of stellar parameters presented in D17 and update the host star parameters accordingly in Section 3. Next, we fit the K2 photometry in Section 4 to determine the transit parameters for each candidate. In Section 5, we use the open-source VESPA software (Morton 2015b) to run a false positive analysis to statistically validate planets and identify likely false positives. In Section 6, we combine the information from our various analyses to revise parameters for K2 planet candidates and validated planets orbiting low-mass dwarfs. We highlight particularly noteworthy individual systems in Section 7 before concluding in Section 8.

We obtained NIR spectroscopy of all stars in our sample because they were initially believed to host transiting planets or because they were close enough to the candidate host star that they might have been responsible for the transit-like signal observed in the light curve. In some cases, subsequent detailed analyses of the K2 photometry or ground-based follow-up observations revealed that the putative transit signals were actually due to false positives.

The majority of the 74 systems in the cool dwarf sample were selected from unpublished candidate lists provided by A. Vanderburg (53 stars hosting 59 candidates) or the K2 California Consortium (K2C2; 43 stars hosting 54 candidates); several stars appear on both lists and many of the K2C2 targets were later published in Crossfield et al. (2016). The cool dwarf sample also includes 20 single-planet systems from Barros et al. (2016), 16 singles from Pope et al. (2016), 16 stars hosting 18 candidates reported by Vanderburg et al. (2016b), three singles from Adams et al. (2016), two singles from Mann et al. (2017), and five singles from Montet et al. (2015), who refined the properties of the planet candidates reported by Foreman-Mackey et al. (2015). Two of the stars in the sample (EPIC 212773309 and EPIC 211694226) are in close proximity to other stars and EPIC 212773309 also displays a clear secondary eclipse.

The 74 systems in our cool dwarf sample host 79 unique K2 Objects of Interest (K2OIs). As of 2016 September 7, 24 of those K2OIs had been confirmed as bona fide planets using the VESPA false positive probability tool (Morton 2012, 2015b, see Section 5), 23 had been previously published as planet candidates, two were classified as false positives, and 30 were new detections. Twenty seven of the new detections were identified by A. Vanderburg, eight were found by K2C2, and five were discovered by both collaborations. In total, our cool dwarf planet sample consists of 60 single-planet systems, six double-planet systems (EPIC 201549860 = K2-35, EPIC 206011691 = K2-21, EPIC 210508766 = K2-83, EPIC 210968143 = K2-90, EPIC 211305568, and EPIC 211331236 = K2-117), one triple-planet system (EPIC 211428897) and one quadruple-planet system (EPIC 206209135 = K2-72).

In D17, we applied empirical relations to NIR spectra acquired at IRTF/SpeX and Palomar/TSPEC to revise the classifications of putative low-mass stars harboring potential K2 planet candidates. Of the 144 K2 targets we observed, 49% were actually contaminating giant stars or hotter dwarfs (typically reddened by interstellar extinction) and 74 (51%) were bona fide low-mass dwarfs. For the cool dwarfs, we measured a series of equivalent widths and spectral indices and applied empirically based relations from Newton et al. (2015) and Mann et al. (2013a, 2015) to estimate effective temperatures, radii, masses, metallicities, and luminosities. Our results agree well with those of Martinez et al. (2017), who used lower-resolution NIR spectra from NTT/SOFI to improve the characterization of late-type dwarfs hosting K2 planet candidates.

In general, we found that our new radii were typically 0.13 R_⊙ (39%) larger than the original estimates provided in the Ecliptic Plane Input Catalog (EPIC, Huber et al. 2016). These changes are unsurprising because, as noted by Huber et al. (2016), the EPIC values were determined by comparing photometry to stellar models that have been shown to systematically underestimate the temperatures and radii of cool stars (e.g., Kraus et al. 2011; Boyajian et al. 2012; Feiden & Chaboyer 2012; Spada et al. 2013; von Braun et al. 2014; Newton et al. 2015). In addition, our quest to find potentially habitable planets likely biased our stellar sample toward stars with underestimated radii at the expense of stars with overestimated radii. We adopt the revised stellar classifications from D17 in this paper.

Although the planet candidate catalogs we consulted typically provided their own estimates of stellar properties, we found that the original stellar classifications provided therein also tended to underestimate the radii and temperatures of the cool dwarfs. For most candidates, the amplitude of the suggested radius change is similar to the 15% radius increase found by Newton et al. (2015) for Kepler planet candidates orbiting M dwarfs with previous radius estimates based on fits to stellar models (e.g., Muirhead et al. 2012, 2014; Dressing & Charbonneau 2013; Mann et al. 2013b; Huber et al. 2014).

The exceptions to the general trend of underestimated stellar radii are the parameters provided in Vanderburg et al. (2016b) and the associated unpublished Vanderburg lists. Indeed, their estimates are based on empirical relations (Casagrande et al. 2008; González Hernández & Bonifacio 2009; Boyajian et al. 2013; Pecaut & Mamajek 2013). For those two sources, the discrepancies between our revised values and their initial values are likely due to the fact that spectroscopic observations help break the degeneracy between stars that are intrinsically red and stars that appear red due to interstellar extinction.

Some of the planets in our target list have well-determined properties because they were previously published in other catalogs, but others are new. In order to provide a uniform catalog of properties for all of the planets in our sample, we perform our own fits to the K2 photometry to determine updated properties and errors for the full planet sample.

We began by downloading the K2 Self Flat Fielding (K2SFF) photometry provided by A. Vanderburg.²¹ The K2SFF pipeline processes K2 photometry by recording the roll angle of the spacecraft during each cadence and removing the correlation between flux and roll angle (Vanderburg & Johnson 2014; Vanderburg et al. 2016b). This procedure yields precision within 60% of that achieved during the baseline Kepler mission for faint stars (Kp > 12.5) and Kepler-like performance for brighter stars. Prior to fitting the transits, we re-processed the K2SFF data by simultaneously fitting for the transit light curves, long-term variability, and K2 pointing systematics using the procedure described by Vanderburg et al. (2016b).

Next, we ran a Markov chain Monte Carlo (MCMC) analysis to constrain the time of transit T₀, orbital period P, planet/star radius ratio R_p/R_*, semimajor axis/stellar radius ratio a/R_*, inclination i, quadratic limb darkening, eccentricity e, and longitude of periastron ω. We ran our fits in Python and used the emcee package (Foreman-Mackey et al. 2013) to determine the errors on transit parameters. We assessed convergence by measuring the integrated autocorrelation time of our chains and requiring that the MCMC ran for at least ten times longer than the estimated autocorrelation time.

In order to efficiently sample the full allowed parameter space, we fit the limb darkening parameters using the q₁, q₂ coordinate-space defined by Kipping (2013a). We also assumed that the eccentricity distribution of transiting planets followed the Beta distribution found by Kipping (2013b) for short-period planets (P < 382 days, P_β(e; α, β) = $\tfrac{{\rm{\Gamma }}(\alpha +\beta )}{{\rm{\Gamma }}(\alpha ){\rm{\Gamma }}(\beta )}$e^α−1(1 − e)^β−1 where α = 0.697, β = 3.27, and Γ is the Gamma function) and fit for $\sqrt{e}\cos \omega$ and $\sqrt{e}\sin \omega$ to enable more efficient sampling of low-eccentricity orbits (e.g., Eastman et al. 2013).

At each point in the analysis, we computed the likelihood of our transit model by using the BATMAN package written by Kreidberg (2015) to solve the equations of Mandel & Agol (2002) and generate a model transit lightcurve. Our K2 light curves were obtained in "long-cadence" mode using 30 min integration times, which is relatively long compared to total durations of the transits we model. Accordingly, we employed the "supersample" feature of BATMAN to generate sample long-cadence light curves by modeling the brightness of the star at 1 min cadence and recording the average of 30 consecutive modeled fluxes.

We restricted our fits to consider 70° < i < 90°, 0 < R_p/R_* < 1, a/R_* > 1, 0 ≤ q₁ ≤ 1, 0 ≤ q₂ ≤ 1, $0\lt \sqrt{e}\cos \omega \lt 1$, and $0\lt \sqrt{e}\sin \omega \lt 1$. We further required that K2OIs transit their host stars (i.e., impact parameter b ≤ 1 + R_P/R_⋆) and imposed priors on the limb-darkening coefficients by interpolating the tables produced by Claret et al. (2012) at the temperatures and surface gravities of the host stars.²² Specifically, we assumed that u₁ and u₂ were drawn from Gaussian distributions with dispersions set by propagating the errors in our stellar parameter estimates. Despite our attention to limb darkening, we note that the dominant contribution to the shape of ingress and egress is the smearing due to the lengthy exposure times used for long-cadence K2 data.

All of our target stars were spectroscopically characterized by D17. We incorporated our knowledge of the host stars into our transit fits by using Kepler's third law and the estimated host star masses to determine the orbital semimajor axes of planets with the observed orbital periods. We then compared the resulting distances to the estimated stellar radii and set Gaussian priors on the expected a/R_* ratio for each planet (Seager & Mallén-Ornelas 2003; Sozzetti et al. 2007; Torres et al. 2008). The widths of these priors reflect the uncertainties in the stellar radii and masses,²³ but by imposing this prior we implicitly assume that the candidate transit event is indeed caused by the transit of a planet across the target star rather than a blended transit or eclipse of a contaminating star. The VESPA false positive analysis discussed in Section 5 should reveal such scenarios.

For a handful of targets, the transit depths were shallow enough that the MCMC sometimes wandered away from the transit signal under consideration. Accordingly, we required that the transit center must be within 6% of the initial guess (up to a maximum difference of 6 hr) and that the orbital period must be within 0.025 days (36 min) of the initial guess. As discussed in Section 6.3, we later repeated the transit fitting process using k2phot and K2SC photometry to check for systematic offsets in planet parameters.

Early in the Kepler mission, transiting planet candidates were "confirmed" by conducting radial velocity or transit timing variation studies to measure planet masses. However, mass measurements are expensive and not feasible for all systems. As a result, tools like BLENDER (e.g., Fressin et al. 2011; Torres et al. 2011), VESPA (Morton 2012, 2015b), and PASTIS (Díaz et al. 2014) have been adopted to "validate" planet candidates in a statistical sense. These tools simulate the range of astrophysical configurations that could generate the observed lightcurve subject to the constraints placed by in-depth analyses of the transit photometry (e.g., shifts in the photocenter during transit, presence or absence of secondary eclipse, variations in the depths of odd and even transits) and subsequent follow-up investigations (e.g., high-contrast imaging, radial velocity searches for additional stars, achromaticity of transit events, analyses of archival observations).

5.1. Identifying Clear False Positives

5.2. Assessing False Positive Probabilities

For the more promising K2OIs, we used the VESPA framework to assess the probabilities that each was truly a transiting planet. We first fit the transit photometry as described in Section 4 and rescaled the photometric errors so that the adopted transit model had a reduced chi squared of 1 for the segment of the light curve centered on the transit event. Prior to rescaling the errors, we clipped the light curves and kept only the points within six durations of the transit center. We then searched for secondary eclipses by phase-folding the full data set to the orbital period of the planet in question and measuring the "eclipse depth" at multiple points in the light curve. We assumed that the eclipse has the same duration as the primary transit, but we allowed the phase of secondary eclipse to vary from phase = 0.3 to phase = 0.7 (i.e., we did not assume that the orbit is circular). We recorded the depth of the deepest event as the maximum allowed secondary eclipse depth ("secthresh" in the VESPA fpp.ini file).

Next, we referred back to the K2SFF photometry and recorded the radius of the selected photometric aperture as the maximum allowed separation between the target and the source of the transit event ("maxrad"). We then consulted previously published K2 papers and the Exo-FOP K2 follow-up website to check whether there are extant speckle or high-contrast images placing limits on the allowed brightness of nearby stars. If so, we included those contrast curves as additional constraints. For reference, we list the adaptive optics observations used in our false positive analysis in Table 1 and display the composite set of contrast curves in Figure 1. These observations were obtained with NIRC2 on the 10 m Keck II telescope, NIRI (Hodapp et al. 2003) on the 8 m Gemini-N telescope, PHARO (Hayward et al. 2001) on the 200 inch Palomar Hale telescope, and DSSI (Howell et al. 2011; Horch et al. 2012) on the 8 m Gemini-N and Gemini-S telescopes and the 3.5 m WIYN²⁵ telescope.

Zoom In Zoom Out Reset image size

Table 1.  Speckle and AO Observations Used in VESPA Analysis

				Pixel	PSF	Nearby Stara			Contrast	Observation	Uploaded
EPIC	Telescope	Instrument	Filter	Scale	('')	Det?	Δmag	Sep ('')	Achievedb	Date	By
201205469	Keck2_10m	NIRC2	K	0.009942	0.066557	no	⋯	⋯	Δ7.91 mag at 05	4/7/15	Ciardi
201208431	Keck2_10m	NIRC2	K	0.009942	0.079626	no	⋯	⋯	Δ5.86 mag at 05	2/19/16	Ciardi
201549860	Keck2_10m	NIRC2	K	0.009942	0.060571	no	⋯	⋯	Δ8.10 mag at 05	4/1/15	Ciardi
201617985	Keck2_10m	NIRC2	K	0.009942	0.091839	no	⋯	⋯	Δ7.84 mag at 05	4/1/15	Ciardi
201637175	GeminiN_8m	DSSI	692 nm	0.011410	0.020000	noc	⋯	⋯	Δ4.977 mag at 05	1/15/16	Ciardi
201637175	GeminiN_8m	DSSI	880 nm	0.011410	0.020000	noc	⋯	⋯	Δ4.666 mag at 05	1/15/16	Ciardi
201855371	Keck2_10m	NIRC2	K	0.009942	0.053694	no	⋯	⋯	Δ8.49 mag at 05	4/7/15	Ciardi
205924614	Keck2_10m	NIRC2	K	0.009942	0.058868	no	⋯	⋯	Δ8.17 mag at 05	8/7/15	Ciardi
206011691	Keck2_10m	NIRC2	K	0.009942	0.050896	no	⋯	⋯	Δ7.45 mag at 05	7/25/15	Ciardi
206209135	Keck2_10m	NIRC2	K	0.009942	0.062750	no	⋯	⋯	Δ7.74 mag at 05	8/21/15	Ciardi
210448987	GeminiN_8m	NIRI	K	0.021400	0.102616	no	⋯	⋯	Δ6.86 mag at 05	12/14/15	Ciardi
210489231	Keck2_10m	NIRC2	K	0.009942	0.101436	no	⋯	⋯	Δ5.78 mag at 05	10/28/15	Ciardi
210508766	Keck2_10m	NIRC2	K	0.009942	0.054223	no	⋯	⋯	Δ7.40 mag at 05	10/28/15	Ciardi
210508766	GeminiN_8m	DSSI	692 nm	0.011410	0.020000	no	⋯	⋯	Δ4.994 mag at 05	1/13/16	Ciardi
210508766	GeminiN_8m	DSSI	880 nm	0.011410	0.020000	no	⋯	⋯	Δ5.248 mag at 05	1/13/16	Ciardi
210508766	Palomar-5 m	PHARO-AO	K_short	0.025000	0.159000	no	⋯	⋯	Δ4.90 mag at 05	10/20/16	Ciardi
210558622	Keck2_10m	NIRC2	K	0.009942	0.049441	nod	⋯	⋯	Δ7.86 mag at 05	10/28/15	Ciardi
210558622	Keck2_10m	NIRC2	K	0.009942	0.050778	nod	⋯	⋯	Δ8.24 mag at 05	2/17/16	Ciardi
210558622	GeminiN_8m	DSSI	692 nm	0.011410	0.020000	nod	⋯	⋯	Δ5.286 mag at 05	1/13/16	Ciardi
210558622	GeminiN_8m	DSSI	880 nm	0.011410	0.020000	nod	⋯	⋯	Δ5.248 mag at 05	1/13/16	Ciardi
210558622	WIYN_3.5 m	DSSI	692 nm	0.022000	0.050000	nod	⋯	⋯	Δ3.5 mag at 02	10/24/15	Everett
210558622	WIYN_3.5 m	DSSI	880 nm	0.022000	0.063000	nod	⋯	⋯	Δ3.5 mag at 02	10/24/15	Everett
210707130	Keck2_10m	NIRC2	K	0.009942	0.051244	no	⋯	⋯	Δ7.92 mag at 05	10/28/15	Ciardi
210707130	Keck2_10m	NIRC2	K	0.009942	0.053366	no	⋯	⋯	Δ6.62 mag at 05	2/19/16	Ciardi
210707130	GeminiN_8m	DSSI	692 nm	0.011410	0.020000	no	⋯	⋯	Δ5.650 mag at 05	1/13/16	Ciardi
210707130	GeminiN_8m	DSSI	880 nm	0.011410	0.020000	no	⋯	⋯	Δ5.719 mag at 05	1/13/16	Ciardi
210750726	Keck2_10m	NIRC2	K	0.009942	0.111601	no	⋯	⋯	Δ5.36 mag at 05	10/28/15	Ciardi
210750726	Keck2_10m	NIRC2	K	0.009942	0.080981	no	⋯	⋯	Δ6.06 mag at 05	2/19/16	Ciardi
210750726	GeminiN_8m	DSSI	692 nm	0.011410	0.020000	no	⋯	⋯	Δ4.901 mag at 05	1/15/16	Ciardi
210750726	GeminiN_8m	DSSI	880 nm	0.011410	0.020000	no	⋯	⋯	Δ5.002 mag at 05	1/15/16	Ciardi
210838726	GeminiN_8m	NIRI	K	0.021400	0.109663	no	⋯	⋯	Δ6.12 mag at 05	11/2/15	Ciardi
210838726	GeminiN_8m	NIRI	K	0.021400	0.099683	no	⋯	⋯	Δ6.83 mag at 05	12/14/15	Ciardi
210968143	Keck2_10m	NIRC2	K	0.009942	0.053846	no	⋯	⋯	Δ7.59 mag at 05	10/28/15	Ciardi
210968143	Keck2_10m	NIRC2	K	0.009942	0.051648	no	⋯	⋯	Δ7.91 mag at 05	2/17/16	Ciardi
210968143	GeminiN_8m	DSSI	692 nm	0.011410	0.020000	no	⋯	⋯	Δ5.279 mag at 05	1/13/16	Ciardi
210968143	GeminiN_8m	DSSI	880 nm	0.011410	0.020000	no	⋯	⋯	Δ5.089 mag at 05	1/13/16	Ciardi
210968143	Palomar-5 m	PHARO-AO	K_short	0.025000	0.165000	no	⋯	⋯	Δ4.73 mag at 05	10/20/16	Ciardi
211077024	Keck2_10m	NIRC2	K	0.009942	0.061843	no	⋯	⋯	Δ7.41 mag at 05	2/19/16	Ciardi
211077024	GeminiN_8m	DSSI	692 nm	0.011410	0.020000	no	⋯	⋯	Δ4.924 mag at 05	1/15/16	Ciardi
211077024	GeminiN_8m	DSSI	880 nm	0.011410	0.020000	no	⋯	⋯	Δ5.371 mag at 05	1/15/16	Ciardi
211331236	Keck2_10m	NIRC2	K	0.009942	0.079162	no	⋯	⋯	Δ6.62 mag at 05	1/21/16	Ciardi
211428897	Keck2_10m	NIRC2	J	0.009942	0.114165	yes	⋯	⋯	Δ5.41 mag at 05	1/21/16	Ciardi
211428897	Keck2_10m	NIRC2	K	0.009942	0.060653	yes	⋯	⋯	Δ6.86 mag at 05	1/21/16	Ciardi
211428897	GeminiN_8m	DSSI	692 nm	0.011410	0.020000	yes	1.8	1.1	Δ4.442 mag at 05	1/15/16	Ciardi
211428897	GeminiN_8m	DSSI	880 nm	0.011410	0.020000	yes	1.2	1.1	Δ4.869 mag at 05	1/15/16	Ciardi
211509553	GeminiN-8 m	NIRI	open	0.021400	0.098000	yes	3.3	1.9	Δ5.84 mag at 05	2/20/16	Ciardi
211770795	Keck2_10m	NIRC2	K	0.009942	0.065018	no	⋯	⋯	Δ7.36 mag at 05	2/19/16	Ciardi
211770795	GeminiN-8 m	NIRI	open	0.021400	0.112000	no	⋯	⋯	Δ5.57 mag at 05	2/20/16	Ciardi
211799258	GeminiN-8 m	NIRI	open	0.021400	0.107000	no	⋯	⋯	Δ6.61 mag at 05	2/20/16	Ciardi
211818569	Palomar-5 m	PHARO-AO	Ks	0.025000	0.122000	no	⋯	⋯	Δ5.22 mag at 05	10/20/16	Ciardi
211831378	GeminiN_8m	NIRI	K	0.021400	0.125410	no	⋯	⋯	Δ6.59 mag at 05	1/28/16	Ciardi
211924657	GeminiN-8 m	NIRI	Br-gamma	0.021400	0.101000	no	⋯	⋯	Δ6.26 mag at 05	2/20/16	Ciardi
211970234	GeminiN-8 m	NIRI	open	0.021400	0.113000	no	⋯	⋯	Δ5.60 mag at 05	2/20/16	Ciardi
212006344	Keck2_10m	NIRC2	K	0.009942	0.053540	no	⋯	⋯	Δ8.06 mag at 05	1/21/16	Ciardi
212006344	GeminiN_8m	DSSI	692 nm	0.011410	0.020000	no	⋯	⋯	Δ5.281 mag at 05	1/14/16	Ciardi
212006344	GeminiN_8m	DSSI	880 nm	0.011410	0.020000	no	⋯	⋯	Δ5.800 mag at 05	1/14/16	Ciardi
212069861	Keck2_10m	NIRC2	J	0.009942	0.092989	no	⋯	⋯	Δ6.17 mag at 05	2/17/16	Ciardi
212069861	Keck2_10m	NIRC2	K	0.009942	0.096097	no	⋯	⋯	Δ7.63 mag at 05	2/17/16	Ciardi
212069861	GeminiN-8 m	NIRI	Br-gamma	0.021400	0.131000	no	⋯	⋯	Δ4.93 mag at 05	2/20/16	Ciardi
212154564	Keck2_10m	NIRC2	K	0.009942	0.111359	no	⋯	⋯	Δ6.06 mag at 05	2/19/16	Ciardi
212154564	GeminiN-8 m	NIRI	open	0.021400	0.106000	no	⋯	⋯	Δ5.52 mag at 05	2/20/16	Ciardi
212354731	GeminiS_8m	DSSI	692 nm	0.011000	0.021000	no	⋯	⋯	Δ5 mag at 02	6/29/16	Everett
212354731	GeminiS_8m	DSSI	880 nm	0.011000	0.027000	no	⋯	⋯	Δ5 mag at 02	6/29/16	Everett
212460519	WIYN_3.5 m	DSSI	692 nm	0.022000	0.050000	no	⋯	⋯	Δ3.5 mag at 02	4/21/16	Everett
212460519	WIYN_3.5 m	DSSI	880 nm	0.022000	0.063000	no	⋯	⋯	Δ3.5 mag at 02	4/21/16	Everett
212554013	GeminiS_8m	DSSI	692 nm	0.011000	0.021000	no	⋯	⋯	Δ5 mag at 02	6/22/16	Everett
212554013	GeminiS_8m	DSSI	880 nm	0.011000	0.027000	no	⋯	⋯	Δ5 mag at 02	6/22/16	Everett
212565386	GeminiS_8m	DSSI	692 nm	0.011000	0.021000	no	⋯	⋯	Δ5 mag at 02	6/22/16	Everett
212565386	GeminiS_8m	DSSI	880 nm	0.011000	0.027000	no	⋯	⋯	Δ5 mag at 02	6/22/16	Everett
212572452	GeminiS_8m	DSSI	692 nm	0.011000	0.021000	no	⋯	⋯	Δ5 mag at 02	6/22/16	Everett
212572452	GeminiS_8m	DSSI	880 nm	0.011000	0.027000	no	⋯	⋯	Δ5 mag at 02	6/22/16	Everett
212628098	GeminiS_8m	DSSI	692 nm	0.011000	0.021000	yes	⋯	⋯	Δ5 mag at 02	6/22/16	Everett
212628098	GeminiS_8m	DSSI	880 nm	0.011000	0.027000	yes	3.8	1.3	Δ5 mag at 02	6/22/16	Everett
212628098	WIYN_3.5 m	DSSI	692 nm	0.022000	0.050000	yes	⋯	⋯	Δ3.5 mag at 02	4/20/16	Everett
212628098	WIYN_3.5 m	DSSI	880 nm	0.022000	0.063000	yes	⋯	⋯	Δ3.5 mag at 02	4/20/16	Everett
212628098	GeminiN-8 m	NIRI	Br-gamma	0.021400	0.107000	yes	⋯	⋯	Δ7.01 mag at 05	6/20/16	Ciardi
212679181	GeminiS_8m	DSSI	692 nm	0.011000	0.021000	yes	1.1	1.2	Δ5 mag at 02	6/21/16	Everett
212679181	GeminiS_8m	DSSI	880 nm	0.011000	0.027000	yes	1.1	1.3	Δ5 mag at 02	6/21/16	Everett
212679181	WIYN_3.5 m	DSSI	692 nm	0.022000	0.050000	yes	1.5	1.5	Δ3.5 mag at 02	4/17/16	Everett
212679181	WIYN_3.5 m	DSSI	880 nm	0.022000	0.063000	yes	1.2	1.5	Δ3.5 mag at 02	4/17/16	Everett
212679798	GeminiS_8m	DSSI	692 nm	0.011000	0.021000	yes	⋯	⋯	Δ5 mag at 02	6/22/16	Everett
212679798	GeminiS_8m	DSSI	880 nm	0.011000	0.027000	yes	2.6	0.1	Δ5 mag at 02	6/22/16	Everett
212679798	GeminiN-8 m	NIRI	open	0.021400	0.110000	yes	⋯	⋯	Δ6.62 mag at 05	6/20/16	Ciardi
212686205	WIYN_3.5 m	DSSI	692 nm	0.022000	0.050000	no	⋯	⋯	Δ3.5 mag at 02	4/20/16	Everett
212686205	WIYN_3.5 m	DSSI	880 nm	0.022000	0.063000	no	⋯	⋯	Δ3.5 mag at 02	4/20/16	Everett
212773272	GeminiS_8m	DSSI	692 nm	0.011000	0.021000	no	⋯	⋯	Δ5 mag at 02	6/21/16	Everett
212773272	GeminiS_8m	DSSI	880 nm	0.011000	0.027000	no	⋯	⋯	Δ5 mag at 02	6/21/16	Everett
212773272	GeminiN-8 m	NIRI	Br-gamma	0.021400	0.109000	no	⋯	⋯	Δ7.09 mag at 05	6/20/16	Ciardi
212773309	GeminiS_8m	DSSI	692 nm	0.011000	0.021000	yes	2.8	1.0	Δ5 mag at 02	6/21/16	Everett
212773309	GeminiS_8m	DSSI	880 nm	0.011000	0.027000	yes	2.0	1.0	Δ5 mag at 02	6/21/16	Everett
212773309	WIYN_3.5 m	DSSI	692 nm	0.022000	0.050000	yes	2.8	1.2	Δ3.5 mag at 02	4/24/16	Everett
212773309	WIYN_3.5 m	DSSI	880 nm	0.022000	0.063000	no	2.0	1.2	Δ3.5 mag at 02	4/24/16	Everett
213951550	GeminiN-8 m	NIRI	open	0.021400	0.116000	yes	⋯	0.2	Δ6.41 mag at 05	7/15/16	Ciardi
216892056	GeminiS_8m	DSSI	692 nm	0.011000	0.021000	no	⋯	⋯	Δ5 mag at 02	6/29/16	Everett
216892056	GeminiS_8m	DSSI	880 nm	0.011000	0.027000	no	⋯	⋯	Δ5 mag at 02	6/29/16	Everett
216892056	GeminiN-8 m	NIRI	Br-gamma	0.021400	0.110000	no	⋯	⋯	Δ6.86 mag at 05	6/15/16	Ciardi
217941732	GeminiS_8m	DSSI	692 nm	0.011000	0.021000	no	⋯	⋯	Δ5 mag at 02	6/19/16	Everett
217941732	GeminiS_8m	DSSI	880 nm	0.011000	0.027000	no	⋯	⋯	Δ5 mag at 02	6/19/16	Everett

Notes.

^aWe use the "Nearby Star Det?" column to indicate whether any follow-up image revealed a companion to the star. The values in the "Δmag" and "Sep" columns refer to the magnitude difference and separation measured in the specific image described on the corresponding line of the table. ^bThese point sensitivity estimates provide a rough view of whether the image provides deep or shallow limits on the presence of nearby companions. As shown in Figure 1, the contrast achieved improves with increasing separation from the target star. We use the full separation-dependent contrast curves for our false positive probability estimates. ^cThe Gemini/DSSI speckle images of EPIC 201637175 did not reveal any companions, but Subaru/HSC imaging displayed a second star roughly 12% as bright as EPIC 201637175. The separation of the two stars is approximately 2'' (Sanchis-Ojeda et al. 2015). ^dThe follow-up images of EPIC 210558622 did not reveal any companions, but a second set of spectral lines was detected in the Keck/HIRES reconnaissance spectrum (Crossfield et al. 2016).

Download table as:  ASCIITypeset images: 1 2 3

Finally, we filled in the host star properties (coordinates, magnitudes, and spectroscopic fits from D17) and ran VESPA to compute the false positive probability (FPP). Recognizing that VESPA FPPs are statistical and depend on the assumed planet radius, we ran the analysis twice for each planet using R_p/R_⋆ ratios set to the 16th and 84th percentiles of the posterior distribution. As in Crossfield et al. (2016), we adopted a threshold of FPP < 1% for validation, but we adopted a less forgiving FPP cut of FPP > 90%. When classifying K2OIs, we required that both FPP estimates were below 1% or above 90% in order to label K2OIs as validated planets or false positives, respectively. We classified the remaining systems as planet candidates.

We summarize our new K2OI dispositions in Table 2. In total, 44 K2OIs met the formal criteria for validation, but seven orbit stars with nearby companions and therefore cannot be validated with VESPA. Of the remaining 37 K2OIs with low FPPs, 20 were previously validated by Crossfield et al. (2016), eight are new detections, and nine were previously classified as planet candidates. As discussed in Section 5.2, we rejected eight K2OIs as false positives based on visual inspection of their light curves. No additional K2OIs were classified as false positives based on VESPA FPPs alone. The remaining 27 K2OIs had ambiguous FPPs between 1% and 90%. The ambiguous sample included three previously confirmed planets, nine previously announced planet candidates, and 15 new detections. The three confirmed planets that failed to meet our 1% FPP threshold are EPIC 201345483.01 (R_p = 10.4 R_⊕, FPP = 15%), EPIC 201635569.01 (R_p = 7.5 R_⊕, FPP = 4%–7%), and EPIC 210508766.02 (R_p = 2.1 R_⊕, FPP = 1.9%). One other previously confirmed planet (EPIC 201637175.01 = K2-22b, Sanchis-Ojeda et al. 2015) met our FPP cut for validation, but is listed as a planet candidate in Table 2 due to the presence of a nearby star. We do not dispute the previous validation of K2-22b, but our VESPA analysis is not sufficiently sophisticated to validate planets orbiting stars with nearby companions.

Table 2.  Breakdown of K2OI Dispositions

Previous	Updated Dispositiona
Disposition	CP	PC	FP	All
CP	22	3b	0	25
PC	9	13	1c	23
FP	0	0	2	2
UK	7	17	5	29
All	38d	33	8	79

Notes.

^aCP = Confirmed Planet, PC = Planet Candidate, FP = False Positive, UK = Unknown. ^bThe previously confirmed planets that we cannot validate with VESPA are EPIC 201345483.01 (FPP = 15%), EPIC 201635569.01 (FPP = 4%–7%), and EPIC 201637175.01 (=K2-22b, nearby star detected). See Section 5.2 for details. ^cThe planet candidate rejected as a false positive is EPIC 212572452.01, which was announced by Pope et al. (2016) as a candidate with R_p/R_⋆ = 0.174 and an orbital period of 2.6 days. As discussed in Section 5.2, the K2 photometry for this target is contaminated by light from the nearby, brighter star EPIC 212572439. ^dBold values indicate summations.

Download table as:  ASCIITypeset image

The most likely explanation for why we were unable to validate EPIC 201345483.01 is that our estimates of the stellar radius (${0.69}_{-0.04}^{+0.06}\,{R}_{\odot }$, D17) and planet radius (${10.4}_{-0.7}^{+0.9}\,{R}_{\oplus }$) are much larger than the values of 0.445 ± 0.066 R_⊙ and 6.71 R_⊕ assumed by Crossfield et al. (2016) when validating the system. In contrast, our estimates agree well with those adopted by Vanderburg et al. (2016b) in their discovery paper (0.66 R_⊙, 11 R_⊕). Because larger planets are rarer than smaller planets, our larger planet radius estimate led us to adopt a smaller prior probability of planethood than Crossfield et al. (2016) based on their smaller planet radius estimate. In the future, including AO or speckle imaging would be useful for discriminating between the planetary interpretation and remaining false positive scenarios.

EPIC 201635569.01 was previously validated by Montet et al. (2015) as a 4.81 ± 0.42 R_⊕ planet with FPP = 4.9 × 10⁻³and by Crossfield et al. (2016) as a 4.48 ± 0.52 R_⊕ planet with FPP = 0.6%. Our inability to confirm the validation of this planet may be due to our larger estimate of the planet radius, which is in turn caused by the larger stellar radius found by D17. We estimated a revised radius of 0.62 ± 0.03 R_⊙, which is significantly larger than the values of 0.45 ± 0.01 R_⊙ and 0.39 ± 0.04 R_⊙ assumed by Montet et al. (2015) and Crossfield et al. (2016), respectively. Our new FPP estimate of 4%–7% is still consistent with the planetary interpretation of the transit-like event, but indicates that additional observations such as AO imaging would be useful to rule out the remaining false positive scenarios.

For EPIC 210508766.02, our initial FPP estimate of 1.9% is only slightly above our validation threshold of 1% and does not consider the fact that transit-like events detected in candidate multiple-planet systems are more likely to be bona fide planets (Lissauer et al. 2012). Given that EPIC 210508766 also hosts 210508766.01, we can apply a multiplicity boost to reduce the FPP for 210508766.02 by a factor of 30 (Sinukoff et al. 2016; Vanderburg et al. 2016a). We therefore support the previous validation of EPIC 210508766.02 (K2-83c) by Crossfield et al. (2016) and classify that K2OI as a validated planet while categorizing all of the other K2OIs with FPP between 1% and 90% as planet candidates. The final disposition breakdown for our K2OI sample is 38 validated planets, eight false positives, and 33 planet candidates.

We list the estimated FPPs and resulting dispositions for individual K2OIs in Table 3. For conciseness, Table 3 also includes estimates of the orbital periods and mid-points of transit-like events. We present the remaining transit parameters and the corresponding physical parameters in Table 4. As part of our classification and transit fitting process, we produced an array of vetting plots for each candidate. We have uploaded all of these plots to the ExoFOP-K2 website and provide examples in the Appendix.

Table 3.  K2OI False Positive Probabilities and Dispositions

		Disposition		VESPA FPPa		Nearby	Rp/R⋆			P (days)c			t0 (BKJD)c
EPIC	K2OI	Old	New	Small Rp	Big Rp	Star?b	K2SFF	k2phot	K2SC	Val	−Err	+Err	Val	−Err	+Err
201205469	1	CP	CP	3.6e–08	3.6e–08	no	0.074	0.072	⋯	3.47134	0.00016	0.00016	1976.881	0.002	0.002
201208431	1	CP	CP	3.7e–07	2.0e–06	no	0.034	0.035	⋯	10.00342	0.00100	0.00102	1982.524	0.004	0.004
201345483	1	CP	PC	1.5e–01	1.5e–01	⋯	0.140	0.136	⋯	1.72926	0.00001	0.00001	1976.526	0.000	0.000
201549860	1	CP	CP	2.3e–05	3.4e–05	no	0.028	0.028	⋯	5.60835	0.00034	0.00034	1979.119	0.002	0.002
201549860	2	CP	CP	1.5e–09	6.9e–08	no	0.019	0.019	⋯	2.39996	0.00020	0.00020	1977.584	0.004	0.004
201617985	1	PC	PC	5.0e–02	8.3e–02	no	0.033	0.033	⋯	7.28118	0.00057	0.00055	1979.641	0.004	0.004
201635569	1	CP	PC	3.6e–02	6.7e–02	⋯	0.110	0.104	⋯	8.36879	0.00019	0.00019	1978.447	0.001	0.001
201637175	1	CP	PC	2.4e–03	9.1e–03	yes	0.075	0.074	⋯	0.38108	0.00000	0.00000	2034.139	0.000	0.000
201717274	1	PC	PC	1.1e–01	9.3e–02	⋯	0.038	0.039	⋯	3.52675	0.00030	0.00030	1976.915	0.004	0.004
201855371	1	CP	CP	6.2e–06	1.4e–05	no	0.030	0.030	⋯	17.96903	0.00142	0.00138	1984.944	0.003	0.003
205924614	1	CP	CP	6.0e–11	7.9e–11	no	0.056	0.055	0.056	2.84927	0.00003	0.00003	2150.423	0.000	0.000
206011691	1	CP	CP	4.1e–07	1.0e–08	no	0.026	0.026	0.026	9.32504	0.00040	0.00038	2156.422	0.001	0.001
206011691	2	CP	CP	2.5e–08	8.0e–13	no	0.035	0.033	0.034	15.50192	0.00093	0.00092	2155.471	0.002	0.002
206119924d	1	PC	CP	2.6e–03	2.2e–03	⋯	0.009	0.009	0.009	4.65541	0.00047	0.00049	2146.948	0.004	0.004
206209135	1	CP	CP	1.2e–04	1.2e–05	no	0.030	0.030	0.030	5.57721	0.00042	0.00042	2177.376	0.002	0.002
206209135	2	CP	CP	8.4e–05	1.1e–04	no	0.032	0.031	0.029	15.18903	0.00315	0.00313	2156.465	0.005	0.005
206209135	3	CP	CP	7.2e–04	5.1e–04	no	0.028	0.026	0.030	7.76018	0.00150	0.00150	2151.788	0.007	0.008
206209135	4	CP	CP	8.4e–05	1.1e–05	no	0.036	0.038	0.034	24.15887	0.00385	0.00373	2154.054	0.005	0.005
206312951	1	PC	PC	2.3e–02	2.1e–02	⋯	0.022	0.021	0.396	1.53402	0.00017	0.00018	2147.160	0.005	0.005
206318379	1	PC	PC	1.0e–01	4.6e–02	⋯	0.075	0.076	0.076	2.26044	0.00003	0.00003	2147.250	0.000	0.000
210448987	1	CP	CP	2.8e–07	7.5e–08	no	0.029	0.028	0.027	6.10233	0.00041	0.00039	2237.440	0.002	0.002
210508766	1	CP	CP	5.1e–04	3.9e–04	no	0.029	0.028	0.028	2.74723	0.00013	0.00013	2234.060	0.002	0.002
210508766	2	CP	CP	1.8e–02	1.9e–02	no	0.035	0.034	0.034	9.99744	0.00061	0.00062	2233.271	0.002	0.002
210558622	1	PC	PC	2.2e–06	2.1e–10	no	0.035	0.033	0.033	19.56324	0.00237	0.00255	2250.779	0.003	0.003
210564155	1	UK	PC	2.0e–02	1.5e–02	⋯	0.035	0.034	0.034	4.86407	0.00027	0.00028	2263.759	0.001	0.001
210707130	1	CP	CP	1.4e–05	3.4e–05	no	0.019	0.019	0.019	0.68456	0.00002	0.00002	2232.367	0.001	0.001
210750726	1	CP	CP	5.6e–04	3.8e–04	no	0.046	0.045	0.046	4.61228	0.00017	0.00017	2233.218	0.001	0.001
210838726	1	CP	CP	9.8e–04	8.1e–04	no	0.020	0.018	0.020	1.09598	0.00006	0.00006	2233.008	0.002	0.002
210968143	1	CP	CP	2.8e–13	9.8e–12	no	0.037	0.036	0.037	13.73491	0.00080	0.00080	2245.659	0.001	0.001
210968143	2	CP	CP	2.6e–04	1.5e–04	no	0.019	0.017	0.017	2.90071	0.00032	0.00033	2233.740	0.004	0.004
211077024	1	CP	CP	1.9e–05	4.1e–06	no	0.035	0.032	0.032	1.41960	0.00006	0.00006	2232.430	0.002	0.002
211305568	1	UK	PC	5.6e–01	1.0e+00	⋯	0.041	0.041	0.038	11.55063	0.00112	0.00116	2336.344	0.002	0.002
211305568e	2	UK	PC	⋯	⋯	⋯	0.015	0.016	0.017	0.19785	0.00001	0.00001	2343.970	0.002	0.002
211331236f	1	PC	CP	4.7e–08	2.8e–08	no	0.037	0.037	0.036	1.29151	0.00004	0.00004	2309.776	0.001	0.001
211331236g	2	UK	CP	2.2e–06	3.5e–06	no	0.038	0.037	0.037	5.44482	0.00042	0.00040	2310.560	0.003	0.004
211336288	1	UK	PC	1.9e–01	1.2e–01	⋯	0.018	0.015	0.017	0.22181	0.00002	0.00002	2344.090	0.002	0.002
211357309	1	PC	PC	7.6e–02	6.7e–02	⋯	0.017	0.018	0.018	0.46394	0.00002	0.00002	2368.458	0.001	0.001
211428897	1	PC	PC	7.0e–06	5.5e–08	yes	0.024	0.025	0.023	1.61093	0.00006	0.00006	2309.275	0.001	0.001
211428897	2	UK	PC	7.4e–04	2.5e–05	yes	0.021	0.019	0.021	2.17807	0.00012	0.00012	2310.647	0.002	0.002
211428897	3	UK	PC	1.2e–03	5.2e–04	yes	0.021	0.021	0.020	4.96883	0.00037	0.00037	2340.523	0.002	0.002
211509553	1	PC	PC	1.0e–03	1.3e–03	yes	0.181	0.176	0.177	20.35954	0.00032	0.00033	2318.412	0.001	0.001
211680698h	1	UK	CP	5.6e–03	7.6e–03	⋯	0.032	0.029	0.031	50.92092	0.00525	0.00558	2327.473	0.004	0.004
211694226	1	UK	PC	2.9e–01	6.2e–01	yes	0.020	0.017	0.368	1.91828	0.00017	0.00019	2342.946	0.002	0.002
211762841	1	UK	PC	9.8e–02	1.2e–01	⋯	0.031	0.223	0.185	1.56494	0.00009	0.00009	2343.261	0.001	0.001
211770795i	1	PC	CP	1.2e–05	7.6e–05	no	0.031	0.031	0.029	7.72858	0.00070	0.00073	2315.826	0.003	0.003
211791178j	1	UK	CP	1.7e–03	1.5e–03	⋯	0.028	0.028	0.029	9.56274	0.00067	0.00067	2342.604	0.002	0.002
211799258	1	UK	PC	9.0e–01	9.4e–01	no	0.264	0.258	0.348	19.53406	0.00077	0.00079	2320.146	0.001	0.001
211817229	1	UK	PC	3.9e–01	9.5e–01	⋯	0.276	0.266	0.176	2.17693	0.00005	0.00005	2342.525	0.001	0.001
211818569k	1	PC	CP	1.4e–04	6.1e–04	no	0.109	0.108	0.108	5.18575	0.00020	0.00020	2310.561	0.001	0.001
211822797	1	CP	CP	6.7e–04	7.6e–04	⋯	0.032	0.030	0.029	21.16986	0.00169	0.00172	2332.577	0.002	0.002
211826814	1	UK	PC	5.6e–01	9.6e–01	⋯	0.082	0.260	0.028	1.53453	0.00012	0.00013	2343.198	0.002	0.002
211831378	1	UK	FP	2.4e–05	1.9e–06	no	0.015	0.074	0.579	3.48929	0.00050	0.00048	2310.755	0.005	0.005
211924657l	1	PC	PC	5.3e-01	6.4e-01	no	0.052	0.051	0.051	2.64484	0.00011	0.00011	2311.641	0.001	0.001
211965883	1	PC	PC	3.4e–01	3.8e–01	⋯	0.043	0.039	0.039	10.55630	0.00066	0.00064	2334.605	0.001	0.001
211969807	1	PC	PC	8.6e–02	9.5e–02	⋯	0.037	0.036	0.037	1.97424	0.00011	0.00011	2342.915	0.001	0.001
211970234	1	UK	FP	1.6e–01	1.9e–01	no	0.069	0.104	0.212	1.48350	0.00004	0.00003	2310.371	0.001	0.001
211988320	1	UK	PC	7.5e–02	7.2e–02	⋯	0.041	0.036	0.034	63.84808	0.00588	0.00560	2309.721	0.004	0.005
212006344m	1	PC	CP	1.5e–04	1.7e–04	no	0.020	0.019	0.020	2.21940	0.00007	0.00007	2311.048	0.001	0.001
212069861n	1	PC	CP	1.1e–04	2.9e–06	no	0.043	0.042	0.040	30.95676	0.00272	0.00259	2314.492	0.003	0.003
212154564o	1	PC	CP	6.8e–07	9.7e–07	no	0.070	0.067	0.069	6.41354	0.00026	0.00025	2309.183	0.002	0.002
212398486p	1	UK	CP	7.5e–03	1.7e–03	⋯	0.050	0.050	0.045	21.75026	0.00196	0.00199	2410.088	0.002	0.002
212443973	1	PC	PC	1.6e–01	8.9e–01	⋯	0.024	0.295	0.096	0.77923	0.00005	0.00006	2423.710	0.002	0.002
212460519q	1	PC	CP	4.2e–13	1.1e–11	no	0.027	0.028	0.027	7.38707	0.00025	0.00025	2390.794	0.001	0.001
212554013r	1	PC	CP	1.0e–03	1.6e–03	no	0.117	0.115	0.117	3.58816	0.00002	0.00002	2390.926	0.000	0.000
212572452	1	PC	FP	3.3e–07	4.3e–08	no	0.072	0.067	0.176	2.58148	0.00001	0.00001	2390.028	0.000	0.000
212628098	1	FP	FP	9.2e–02	3.0e–02	yes	0.228	0.278	0.286	4.35244	0.00003	0.00003	2390.348	0.000	0.000
212634172	1	UK	PC	3.0e–01	1.0e+00	⋯	0.093	0.081	0.069	2.85177	0.00009	0.00010	2421.669	0.001	0.001
212679181	1	PC	PC	4.2e–04	6.4e–04	yes	0.026	0.024	0.026	1.05459	0.00002	0.00001	2423.570	0.000	0.000
212679798	1	UK	FP	2.1e–01	9.5e–01	yes	0.482	0.678	0.158	1.83473	0.00002	0.00002	2389.389	0.002	0.002
212686205s	1	UK	CP	1.1e–07	4.2e–06	no	0.017	0.016	0.014	5.67582	0.00041	0.00041	2422.503	0.002	0.002
212690867	1	UK	PC	2.9e–02	3.5e–02	⋯	0.049	0.047	0.049	25.86125	0.00309	0.00298	2422.464	0.003	0.003
212773272	1	UK	FP	9.4e–01	9.8e–01	no	0.623	0.203	0.205	4.68189	0.00005	0.00005	2389.666	0.001	0.001
212773309	1	FP	FP	8.5e–01	9.4e–01	yes	0.732	0.738	0.728	4.68199	0.00007	0.00006	2389.665	0.001	0.001
213951550	1	UK	FP	1.0e+00	1.0e+00	yes	0.683	0.625	⋯	1.11704	0.00002	0.00002	2478.230	0.001	0.001
214254518	1	UK	PC	9.8e–03	1.2e–02	⋯	0.015	0.015	⋯	5.05899	0.00054	0.00053	2506.630	0.003	0.003
214522613	1	UK	PC	5.1e–02	4.9e–02	⋯	0.024	0.022	⋯	10.99041	0.00126	0.00124	2506.314	0.002	0.002
214787262t	1	UK	CP	6.1e–04	1.4e–03	⋯	0.027	0.026	⋯	8.23949	0.00029	0.00029	2502.009	0.001	0.001
216892056	1	UK	PC	6.1e–01	9.8e–01	no	0.279	0.083	⋯	2.78592	0.00005	0.00005	2478.406	0.001	0.001
217941732u	1	UK	CP	1.2e–06	2.9e–07	no	0.015	0.015	⋯	2.49412	0.00013	0.00013	2507.612	0.001	0.001

Notes.

^aThe "Small R_p" and "Big R_p" values refer to VESPA FPP estimates made using the 16th and 84th percentiles of the R_p/R_⋆ posterior probability distribution, respectively. Note that VESPA operates in a statistical fashion and that the reported FPP may occasionally be higher for the small R_p case than for the big R_p case due to changes in the simulated population of stars. ^bIndicates whether a nearby star was revealed in the follow-up AO and speckle imagery (see Table 1). Rows without definitive answers mark stars without follow-up images posted to the ExoFOP-K2 website. Note that planet radius estimates are not corrected for flux dilution due to the presence of nearby stars. ^cValues and errors for orbital period and transit center are from fits using K2SFF photometry. ^dNow validated as K2-116b. ^eVESPA was unable to calculate an FPP for EPIC 211305568.02. See Section 7.2.5. ^fNow validated as K2-117b. ^gNow validated as K2-117c. ^hNow validated as K2-118b. ⁱNow validated as K2-119b. ^jNow validated as K2-120b. ^kNow validated as K2-121b. ^lThis K2OI displays transit timing variations, but our fits assumed a linear ephemeris. ^mNow validated as K2-122b. ⁿNow validated as K2-123b. ^oNow validated as K2-124b. ^pNow validated as K2-125b. ^qNow validated as K2-126b. ^rNow validated as K2-127b. ^sNow validated as K2-128b. ^tNow validated as K2-129b. ^uNow validated as K2-130b.

Download table as:  ASCIITypeset images: 1 2

Table 4.  Planet Properties^a

		Pb	Rp/R⋆					Limb Darkening				Rp (R⊕)			a	Fp (F⊕)
EPIC	K2OI	(days)	Val	−Err	+Err	a/R⋆	i	q1	q2	e	ω	Val	−Err	+Err	(au)	Val	−Err	+Err
201205469	1	3.47134	0.074	0.002	0.002	13.51	88.90	0.48	0.30	0.07	−25.11	4.75	0.33	0.34	0.038	46.4	16.3	23.0
201208431	1	10.00342	0.034	0.001	0.001	27.68	89.54	0.45	0.33	0.08	−41.44	2.10	0.19	0.20	0.078	15.6	6.2	9.2
201345483	1	1.72926	0.140	0.002	0.003	8.02	87.75	0.45	0.43	0.10	59.61	10.44	0.70	0.90	0.025	380.5	150.1	219.7
201549860	1	5.60835	0.028	0.001	0.001	19.10	88.58	0.55	0.44	0.12	41.16	1.93	0.11	0.13	0.055	68.0	10.5	12.1
201549860	2	2.39996	0.019	0.001	0.001	10.53	88.98	0.54	0.45	0.13	−74.55	1.32	0.08	0.08	0.031	211.0	32.5	37.4
201617985	1	7.28118	0.033	0.002	0.003	26.16	88.30	0.47	0.27	0.23	62.91	1.77	0.17	0.21	0.060	9.2	2.6	3.3
201635569	1	8.36879	0.110	0.004	0.004	22.94	88.19	0.50	0.33	0.14	−70.72	7.49	0.47	0.48	0.069	5.5	3.2	6.1
201637175	1	0.38108	0.075	0.004	0.004	3.19	77.48	0.48	0.30	0.19	46.70	4.75	0.36	0.35	0.009	737.9	196.7	242.8
201717274	1	3.52675	0.038	0.002	0.004	15.20	88.23	0.59	0.29	0.13	23.29	1.32	0.25	0.26	0.026	15.0	3.3	4.1
201855371	1	17.96903	0.030	0.002	0.002	40.17	89.08	0.51	0.37	0.18	52.06	2.03	0.16	0.18	0.117	10.5	2.9	3.7
205924614	1	2.84927	0.056	0.001	0.002	10.46	88.21	0.53	0.44	0.07	23.94	4.38	0.25	0.29	0.035	141.3	23.5	28.8
206011691	1	9.32504	0.026	0.001	0.001	25.37	88.98	0.52	0.41	0.10	34.48	1.84	0.10	0.10	0.076	10.0	1.7	2.1
206011691	2	15.50192	0.035	0.002	0.002	35.50	88.85	0.50	0.41	0.21	59.96	2.49	0.19	0.17	0.107	5.1	0.9	1.1
206119924	1	4.65541	0.009	0.000	0.000	15.37	89.09	0.54	0.44	0.06	−23.44	0.69	0.04	0.04	0.048	78.6	10.5	12.2
206209135	1	5.57721	0.030	0.001	0.002	24.16	89.15	0.54	0.27	0.11	7.49	1.08	0.11	0.11	0.040	8.5	1.1	1.2
206209135	2	15.18903	0.032	0.002	0.002	47.82	89.54	0.54	0.27	0.11	16.83	1.16	0.13	0.13	0.078	2.2	0.3	0.3
206209135	3	7.76018	0.028	0.002	0.002	29.91	89.26	0.54	0.27	0.11	14.28	1.01	0.12	0.12	0.050	5.4	0.7	0.8
206209135	4	24.15887	0.036	0.002	0.002	64.53	89.68	0.54	0.27	0.11	11.39	1.29	0.13	0.14	0.106	1.2	0.2	0.2
206312951	1	1.53402	0.022	0.002	0.002	9.41	87.23	0.47	0.27	0.11	32.40	1.13	0.10	0.11	0.021	120.2	16.8	19.0
206318379	1	2.26044	0.075	0.002	0.004	16.21	88.35	0.54	0.27	0.13	42.80	2.30	0.26	0.28	0.020	30.1	3.8	4.5
210448987	1	6.10233	0.029	0.001	0.001	19.77	89.29	0.54	0.45	0.06	−27.89	1.97	0.11	0.13	0.059	62.9	8.4	9.2
210508766	1	2.74723	0.029	0.001	0.001	12.59	88.72	0.46	0.29	0.07	−5.43	1.71	0.10	0.10	0.032	38.8	5.9	6.3
210508766	2	9.99744	0.035	0.001	0.001	29.67	89.51	0.47	0.29	0.06	−19.35	2.11	0.12	0.12	0.076	6.9	1.0	1.1
210558622	1	19.56324	0.035	0.001	0.001	38.50	89.75	0.53	0.43	0.48	−90.15	2.59	0.15	0.17	0.125	13.2	2.1	2.3
210564155	1	4.86407	0.035	0.002	0.002	26.74	89.00	0.53	0.26	0.13	36.10	1.09	0.13	0.13	0.036	7.7	1.0	1.2
210707130	1	0.68456	0.019	0.001	0.001	4.30	80.21	0.56	0.45	0.24	67.59	1.37	0.11	0.11	0.013	1069.1	144.7	163.5
210750726	1	4.61228	0.046	0.003	0.003	19.98	87.75	0.49	0.27	0.26	72.02	2.30	0.21	0.23	0.042	16.4	2.0	2.2
210838726	1	1.09598	0.020	0.001	0.001	7.36	85.77	0.47	0.28	0.15	51.29	1.08	0.08	0.09	0.017	144.4	18.0	20.2
210968143	1	13.73491	0.037	0.001	0.001	33.90	89.40	0.54	0.45	0.07	16.40	2.53	0.13	0.14	0.100	10.2	1.4	1.7
210968143	2	2.90071	0.019	0.001	0.002	12.02	86.94	0.55	0.45	0.19	56.76	1.30	0.11	0.13	0.035	80.8	11.1	13.2
211077024	1	1.41960	0.035	0.001	0.001	11.19	88.62	0.43	0.25	0.09	−23.36	1.22	0.12	0.12	0.018	56.1	6.6	7.4
211305568	1	11.55063	0.041	0.004	0.805	36.12	88.64	0.49	0.26	0.35	10.33	1.99	0.25	39.17	0.078	5.7	0.7	0.8
211305568	2	0.19785	0.015	0.001	0.002	2.50	79.26	0.48	0.26	0.13	33.20	0.72	0.08	0.11	0.005	1292.5	158.3	178.2
211331236	1	1.29151	0.037	0.001	0.001	8.32	88.36	0.46	0.28	0.06	−22.63	1.96	0.12	0.12	0.019	145.8	18.9	21.5
211331236	2	5.44482	0.038	0.001	0.001	20.93	89.49	0.46	0.28	0.15	−77.06	2.03	0.13	0.13	0.051	21.4	2.8	3.2
211336288	1	0.22181	0.018	0.002	0.002	2.28	77.48	0.47	0.33	0.17	54.63	1.14	0.13	0.15	0.006	1140.9	151.6	172.3
211357309	1	0.46394	0.017	0.001	0.001	4.28	86.80	0.47	0.27	0.08	−36.30	0.84	0.06	0.06	0.010	437.3	56.8	63.9
211428897	1	1.61093	0.024	0.001	0.001	12.64	89.08	0.47	0.25	0.15	−71.99	0.75	0.08	0.08	0.021	48.0	5.9	6.8
211428897	2	2.17807	0.021	0.001	0.001	16.36	89.21	0.47	0.25	0.11	−54.04	0.65	0.07	0.07	0.025	32.1	3.9	4.6
211428897	3	4.96883	0.021	0.001	0.001	29.12	89.52	0.47	0.25	0.10	−40.14	0.67	0.08	0.08	0.044	10.7	1.3	1.5
211509553	1	20.35954	0.181	0.002	0.003	47.46	89.61	0.48	0.28	0.09	43.86	10.82	0.58	0.59	0.119	1.8	0.3	0.4
211680698	1	50.92092	0.032	0.002	0.002	71.82	89.50	0.54	0.45	0.20	58.67	2.54	0.20	0.23	0.245	4.3	0.6	0.6
211694226	1	1.91828	0.020	0.002	0.002	10.51	86.33	0.49	0.26	0.19	59.28	0.99	0.11	0.13	0.021	75.8	12.1	13.8
211762841	1	1.56494	0.031	0.004	0.007	7.91	83.60	0.51	0.27	0.17	55.97	2.14	0.30	0.49	0.023	157.6	25.9	29.7
211770795	1	7.72858	0.031	0.001	0.001	21.58	89.40	0.51	0.27	0.07	−37.71	2.29	0.13	0.15	0.070	54.9	8.8	9.6
211791178	1	9.56274	0.028	0.001	0.001	24.05	89.57	0.49	0.38	0.14	−81.19	2.01	0.12	0.13	0.078	35.1	5.1	5.9
211799258	1	19.53406	0.264	0.021	0.168	77.55	89.26	0.54	0.45	0.48	89.44	9.46	1.93	6.33	0.087	1.0	0.6	1.4
211817229	1	2.17693	0.276	0.208	0.494	17.03	85.82	0.55	0.44	0.08	−4.12	7.13	5.51	12.83	0.019	15.1	11.9	56.6
211818569	1	5.18575	0.109	0.005	0.005	14.71	87.25	0.50	0.27	0.22	72.97	9.12	0.68	0.66	0.052	89.5	11.2	12.5
211822797	1	21.16986	0.032	0.001	0.001	48.67	89.73	0.45	0.25	0.06	−21.51	1.98	0.11	0.11	0.131	3.5	0.5	0.5
211826814	1	1.53453	0.082	0.056	0.520	12.15	85.00	0.56	0.47	0.09	−8.50	2.34	1.65	14.85	0.015	26.0	18.5	63.9
211831378	1	3.48929	0.015	0.001	0.001	14.01	88.89	0.49	0.37	0.08	−22.01	0.91	0.07	0.07	0.037	24.6	7.1	10.5
211924657	1c	2.64484	0.052	0.001	0.002	13.85	88.77	0.48	0.28	0.10	−26.38	1.83	0.21	0.24	0.026	18.8	2.6	2.9
211965883	1	10.55630	0.043	0.004	0.004	29.73	88.32	0.43	0.25	0.23	62.02	2.79	0.30	0.27	0.083	11.4	1.5	1.7
211969807	1	1.97424	0.037	0.001	0.002	9.93	88.00	0.51	0.26	0.09	6.24	1.96	0.15	0.15	0.023	62.1	13.8	16.0
211970234	1	1.48350	0.069	0.003	0.015	12.68	87.70	0.51	0.40	0.12	10.96	1.42	0.30	0.41	0.015	19.2	4.3	5.0
211988320	1	63.84808	0.041	0.001	0.001	92.33	89.80	0.50	0.27	0.07	3.10	2.86	0.15	0.16	0.276	0.9	0.1	0.1
212006344	1	2.21940	0.020	0.001	0.001	10.42	86.39	0.46	0.25	0.21	60.80	1.28	0.08	0.08	0.029	79.9	11.1	13.1
212069861	1	30.95676	0.043	0.001	0.001	61.89	89.79	0.53	0.42	0.06	−33.53	2.65	0.15	0.15	0.167	2.9	0.4	0.5
212154564	1	6.41354	0.070	0.002	0.002	30.09	89.51	0.47	0.33	0.08	−26.80	2.64	0.24	0.24	0.051	8.7	1.1	1.2
212398486	1	21.75026	0.050	0.002	0.002	63.31	89.72	0.45	0.25	0.08	−3.74	2.18	0.18	0.19	0.121	2.0	0.3	0.3
212443973	1	0.77923	0.024	0.005	0.183	7.18	83.37	0.48	0.26	0.12	23.59	0.90	0.19	6.84	0.011	98.7	11.5	13.0
212460519	1	7.38707	0.027	0.000	0.000	22.38	89.40	0.47	0.26	0.06	−33.72	1.84	0.10	0.11	0.066	35.3	5.9	6.6
212554013	1	3.58816	0.117	0.002	0.002	12.02	87.22	0.55	0.27	0.12	−65.74	8.66	0.59	0.68	0.041	105.6	17.8	20.7
212572452	1	2.58148	0.072	0.003	0.003	9.92	86.28	0.58	0.45	0.15	−76.47	5.30	0.37	0.40	0.033	132.0	33.7	41.8
212628098	1	4.35244	0.228	0.006	0.006	17.15	87.29	0.55	0.42	0.18	91.75	14.06	0.78	0.78	0.044	81.2	11.7	13.1
212634172	1	2.85177	0.093	0.028	0.528	14.47	86.15	0.56	0.44	0.12	6.67	3.51	1.12	20.03	0.027	18.8	2.6	2.9
212679181	1	1.05459	0.026	0.003	0.003	8.02	83.74	0.47	0.31	0.18	64.17	1.25	0.18	0.15	0.016	114.7	13.9	16.5
212679798	1	1.83473	0.482	0.231	0.335	8.98	84.65	0.53	0.26	0.52	−91.97	29.52	14.20	20.58	0.024	164.7	28.0	35.0
212686205	1	5.67582	0.017	0.001	0.001	15.46	87.40	0.48	0.26	0.23	62.55	1.42	0.13	0.16	0.056	68.8	9.7	11.1
212690867	1	25.86125	0.049	0.001	0.002	64.68	89.76	0.48	0.29	0.09	−26.75	2.20	0.18	0.19	0.133	1.4	0.2	0.3
212773272	1	4.68189	0.623	0.281	0.259	15.60	85.55	0.56	0.47	0.07	−36.21	29.05	13.26	12.25	0.036	13.8	2.0	2.4
212773309	1	4.68199	0.732	0.348	0.201	16.74	85.66	0.48	0.26	0.14	−79.60	46.96	22.42	13.10	0.048	69.5	8.4	9.8
213951550	1	1.11704	0.683	0.314	0.229	6.71	79.52	0.57	0.28	0.08	−44.84	35.08	16.27	11.98	0.016	166.0	26.5	31.7
214254518	1	5.05899	0.015	0.001	0.001	16.18	89.17	0.50	0.38	0.07	−30.08	1.12	0.07	0.08	0.051	56.1	7.9	9.2
214522613	1	10.99041	0.024	0.002	0.002	36.38	88.90	0.49	0.26	0.22	63.16	1.15	0.12	0.13	0.075	6.9	1.2	1.4
214787262	1	8.23949	0.027	0.001	0.001	34.37	89.21	0.49	0.27	0.13	39.89	1.04	0.09	0.10	0.057	4.5	0.5	0.6
216892056	1	2.78592	0.279	0.232	0.412	15.87	85.50	0.55	0.44	0.06	4.45	12.11	10.13	17.91	0.028	25.5	3.1	3.5
217941732	1	2.49412	0.015	0.001	0.001	9.26	86.53	0.49	0.26	0.17	49.80	1.25	0.14	0.20	0.032	136.4	35.8	45.8

Notes.

^aFor presentation purposes, this table contains only a subset of the available columns. See the machine-readable version for transit parameters based on fits to k2phot and K2SC photometry as well as errors on all parameters. ^bSee Table 3 for errors on P and estimates of the transit center. ^cThis K2OI displays transit timing variations, but our fits assumed a linear ephemeris.

Only a portion of this table is shown here to demonstrate its form and content. A machine-readable version of the full table is available.

Download table as:  DataTypeset images: 1 2

As indicated in Figure 1 and Table 3, six of the planet candidates are associated with K2 targets for which our follow-up imaging observations revealed nearby companions and Keck/HIRES observations revealed an additional set of stellar lines in the spectrum of EPIC 210558622. The imaged companions might be physically associated with the target star or are simply background stars that fall along the same line of sight. Regardless, the close proximity of additional stars dilutes the depths of the transit-like events in the K2 light curves and causes the planets to appear smaller than their true size. In general, the radii of planet candidates orbiting stars with stellar companions are underestimated by a few percent if they orbit the target stars and by a factor of three if they orbit the companion stars (Furlan et al. 2017). Assessing whether the companion stars are bound to the target stars and determining the source of the transit events will require additional scrutiny of the K2 photometry and follow-up imagery (see Furlan et al. 2017, and references therein). We will discuss these systems in more detail in E. J. González et al. (2017, in preparation), an upcoming catalog paper describing the results of our follow-up images of candidate K2 planet host stars of all spectral types.

After refitting the transit photometry and calculating false positive probabilities, we combined our new transit parameters with updated stellar characterizations from D17 to determine the physical properties of each K2OI. We display the revised planet properties in Figure 2. The panels include the full population of K2 planet candidate, validated planets, and false positives as well as planet candidates and confirmed or validated planets identified during the original Kepler mission. In general, the left panel demonstrates that the planet size and orbital period distribution of our K2 planet candidates and validated planets is similar to the distribution of short-period Kepler planet candidates. The majority of planets and planet candidates have radii ≲3 R_⊕, but both the Kepler and K2 radius distributions have tails extending to larger planet radii. As expected, Figure 2 also reveals that K2OIs with high FPPs tend to be larger than those with lower FPPs. Martinez et al. (2017) noted a similar size difference between the radii of their planet candidates and validated planets (see their Figure 9) and remarked that the radius estimates for candidates tend to be more uncertain.

Zoom In Zoom Out Reset image size

The primary difference between our K2 planet candidates and the Kepler sample is that the K2 sample is biased toward brighter host stars. Considering only stars hosting planet candidates or validated planets, the median Kp of our cool dwarf sample is 14.1. As shown in the right panel of Figure 2, 77% of Kepler planets and planet candidates orbiting stars cooler than 4800K have fainter host stars. The Kepler cool dwarf sample has a median host star brightness of Kp = 15.2, fainter than 93% of our K2 host stars. Interestingly, we note that our false positive sample is biased toward fainter host stars relative to the overall sample: the median Kp of our false positive sample is 14.8. A possible explanation for this trend is that the reduced photon counts for these faint stars cause the signal-to-noise ratio of the putative transit events to be below the threshold required to distinguish between bona fide transits and grazing eclipsing binaries.

In both the Kepler and K2 samples, we note a deficit of large planets at short orbital periods. The sole K2 planet with a period shorter than one day and a plotted radius larger than 1.5 R_⊕ is EPIC 201637175.01 (K2-22b), which is reported to be in the process of evaporating (Sanchis-Ojeda et al. 2015). Although our best-fit transit model uses a radius of ${4.75}_{-0.36}^{+0.35}\,{R}_{\oplus }$, Sanchis-Ojeda et al. (2015) argued that the planet itself is likely significantly smaller and surrounded by large dust clouds. In that interpretation, the deep transits are caused by the clouds of debris and do not reflect the underlying radius of the planet (Sanchis-Ojeda et al. 2015).

In general, the relative lack of larger short-period planets is consistent with results from the Kepler mission: Sanchis-Ojeda et al. (2014) observed that all ultra-short-period planets orbiting G, K, or M dwarfs have radii ≲2 R_⊕. The lack of large planets on ultra-short orbital periods may indicate that the envelopes of highly irradiated planets are highly vulnerable to photoevaporation (e.g., Watson et al. 1981; Lammer et al. 2003; Baraffe et al. 2004; Murray-Clay et al. 2009; Valencia et al. 2010; Sanz-Forcada et al. 2011; Lopez et al. 2012; Kurokawa & Kaltenegger 2013; Lopez & Fortney 2013; Owen & Wu 2013). At longer orbital periods, our sample includes planet candidates and validated planets with estimated radii between 0.7 R_⊕ and 12.1 R_⊕.

6.1. Biases in the Planet Sample

Adopting the same planet size categories as Fressin et al. (2013), our sample contains 21 Earths (<1.25 R_⊕; 8 validated), 18 super-Earths (1.25–2 R_⊕; 13 validated), 21 small Neptunes (2–4 R_⊕; 13 validated), three large Neptunes (4–6 R_⊕; two validated), and eight giant planets (>6 R_⊕; two validated). The size distribution depicted in Figure 3 is not representative of the larger population of planets orbiting low-mass stars. Rather, our sample, like most K2 planet catalogs, is shaped by strong selection biases due to the increased detectability of planets on short-period orbits compared to planets with longer periods and our interest in identifying compelling small planets orbiting bright stars for future follow-up observations.

Zoom In Zoom Out Reset image size

Looking at Figure 4, the smallest planets are predominantly detected around the coolest target stars (median T_eff = 3595 K for R_p < 1.25 R_⊕ compared to T_eff = 3842 K for the full sample of planet candidates and validated planets), but the observed correlation of planet radius and stellar effective temperature is likely a selection effect due to the 1/(R_*)² scaling of transit depth with stellar radius rather than a reflection of the true underlying occurrence rate of small planets. We further investigate the role of selection biases in Figure 5 by comparing the host star magnitudes and radii for different sizes of planets.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

As would be expected if the minimum planet radius were a function of search sensitivity, we find that our smallest planet candidates and validated planets are preferentially associated with the brightest and smallest host stars. We also note that the Neptunes and giant planets in the sample fall along the lower right edge of the distribution, indicating that they tend to orbit stars that are fainter and/or larger than the majority of the stars in our target sample. We interpret both of these results as evidence that our planet sample contains significant selection effects and therefore cannot be used to estimate planet occurrence rates unless the detection biases are considered as part of the analysis.

6.2. Planet Radii versus Insolation Flux: Photoevaporation and Potentially Habitable Systems

In Figure 6 we plot the insolation flux received by the planet candidates and validated planets in our sample as a function of the effective temperatures of their host stars. We calculate the insolation flux from the stellar luminosities and the orbital semimajor axes, which we derive from the orbital periods and the stellar masses.

Zoom In Zoom Out Reset image size

Due to the relatively short, 80 day durations of K2 campaigns and the bias of the transit method toward short-period planets, the majority of the planets have very short orbital periods and are therefore highly irradiated. The most heavily irradiated planets in our sample receive over 1000 times the flux received by the Earth (F_⊕) and 21% of the planets receive at least 100 F_⊕.

Although our ability to discern relationships between planet radii and insolation flux environment is complicated by selection effects, Figure 6 reveals that the population of highly irradiated planets is dominated by smaller planets. As previously discussed, the shortage of highly irradiated intermediate-sized planets may be due to photoevaporation. Of the seven validated planets and planet candidates receiving at least 200 F_⊕, the only object with a radius between 2 R_⊕ and 10 R_⊕ is the evaporating planet EPIC 201637175.01 (K2-22b Sanchis-Ojeda et al. 2015).

At the opposite extreme of the insolation flux distribution, our cool dwarf sample contains 21 planets or planet candidates receiving <10 F_⊕. In order to assess whether any of these planets might be habitable, we used the polynomial relations from Kopparapu et al. (2013) to determine the insolation flux boundaries corresponding to conservative habitable zone (HZ) limits of the moist greenhouse inner edge and the maximum greenhouse outer edge and to more optimistic limits of recent-Venus inner limit and early-Mars outer limit. A more sophisticated choice would have been to use the planet-mass-dependent relations from Kopparapu et al. (2014), but we do not know the masses of these planets.

Given the temperature distribution of the host stars in our cool dwarf sample, the median HZ boundaries are 0.26–0.89 F_⊕ (0.22–0.40 au) for the conservative case and 0.23–1.55 F_⊕ (0.16–0.42 au) for the more optimistic case. As shown in Figure 6, four of the planets and planet candidates in our sample fall within the optimistic HZ limits (EPIC 206209135.04, 211799258.01, 211988320.01, and 212690867.01). We discuss each of these K2OIs individually in Section 7.

6.3. Comparison of K2 Photometric Pipelines

In order to investigate the influence of systematic effects in the reduced K2 photometry on the derived planet properties, we repeated the transit fits using photometry processed by the K2 Systematics Correction Pipeline^{26 ,27} (Aigrain et al. 2015, 2016) and the k2phot pipeline (Petigura et al. 2013). We note that lightcurves produced using the k2varcat (Armstrong et al. 2014, 2015, 2016) and EVEREST (Luger et al. 2016) pipelines are also available on the MAST, but we did not perform fits using those lightcurves.

Figure 7 compares the resulting planet/star radius ratios found by fitting photometry from the K2SFF, K2SC, and k2phot pipelines. For K2OIs with small R_p/R_⋆, we find generally consistent planet properties regardless of our choice of photometric pipeline, but the estimates for K2OIs with R_p/R_⋆ > 0.05 can be quite discrepant. In general, the K2OIs with the largest radius disagreement are false positives for which the transit model provides a poor fit to the data. In contrast, nearly all of the K2OIs with R_p/R_⋆ > 0.05 appear well-fit by a transit model and many were classified as confirmed planets in Section 5 based on their K2SFF lightcurves.

Zoom In Zoom Out Reset image size

Specifically, the median absolute difference in R_p/R_⋆,KSFF and R_p/R_⋆,k2phot is ΔR_p/R_{⋆,K2SFF−k2phot} = 0.001 for validated planets, ΔR_p/R_{⋆,K2SFF−k2phot} = 0.002 for planet candidates, and ΔR_p/R_{⋆,K2SFF−k2phot} = 0.05 for false positives. Comparing the KSFF and K2SC fits, we find similar median differences of 0.001 for validated planets and 0.003 for planet candidates, but a higher median difference of 0.12 for false positives. These values are comparable to the median differences of 0.001, 0.002, and 0.06 found when comparing the R_p/R_⋆ fits from the k2phot and K2SC pipelines for validated planets, planet candidates, and false positives, respectively.

Despite the overall agreement between the transit fits produced using distinct sets of photometry, there are a few K2OIs with large R_p/R_⋆ differences across pipelines. For instance, we noticed that the K2SC pipeline appeared to remove the transits of EPIC 211799258.01. Furthermore, our k2phot fits for EPIC 211826814.01 and EPIC 212443973.01 and our K2SC fits for EPIC 206312951.01 and EPIC 211694226.01 failed to converge on the transit midpoint. In general, we found that comparing fits from different pipelines was a convenient way to bolster confidence in borderline detections and reject astrophysical false positives due to blended photometry.

6.4. Updates to Transit Parameters

As shown in Table 2, many of the K2OIs analyzed in this paper were previously announced in other publications. Consulting the NASA Exoplanet Archive (Akeson et al. 2013), we found that 27, 17, 5, and 3 of our K2OIs were included in the candidate catalogs published by Crossfield et al. (2016), Vanderburg et al. (2016b), Montet et al. (2015), and Adams et al. (2016), respectively. In addition, 18 K2OIs were included in the Barros et al. (2016) catalog, 15 appeared in the Pope et al. (2016) catalog, and one (EPIC 205924614.01 = K2-55b) was part of the Schmitt et al. (2016) catalog. While the first set of catalogs included both transit parameters and physical properties like planetary and stellar radii, the second set did not convert transit parameters to physical properties. We compare the available fits for all K2OIs in Figure 8.

Zoom In Zoom Out Reset image size

As shown in the top left panel, our period and planet/star radius ratio estimates are generally in agreement with past results. The median differences (R_p/R_⋆)_revised − (R_p/R_⋆)_literature between our revised planet/star radius ratios and previously published values are 0.0005 (Adams et al. 2016, three K2OIs), −0.001 (Barros et al. 2016, 18 K2OIs, standard deviation σ = 0.003), 0.002 (Crossfield et al. 2016, 27 K2OIs, σ = 0.072), 0.002 (Pope et al. 2016, 15 K2OIs, σ = 0.003), −0.0007 (Montet et al. 2015, five K2OIs, σ = 0.005), −0.001 (Schmitt et al. 2016, EPIC 205924614.01), and −0.0006 (Vanderburg et al. 2016b, 17 K2OIs, σ = 0.002).

The right panels of Figure 8 reveal that our stellar radius estimates tend to be larger than the values used in previous studies. Specifically, we find median differences of 0.04 R_⊙, 0.13 R_⊙, −0.03 R_⊙, and 0.01 R_⊙ compared to the radius estimates from Adams et al. (2016, three K2OIs), Crossfield et al. (2016, 27 K2OIs, σ = 0.07), Montet et al. (2015, five K2OIs, σ = 0.09), and Vanderburg et al. (2016a, 17 K2OIs, σ = 0.10). Primarily due to the large difference between our stellar radius estimates, our planet radius estimates are typically 0.5 R_⊕ (35%) larger than those from Crossfield et al. (2016, σ = 3 R_⊕). We measure smaller differences of −0.01 R_⊕ (−1%), −0.20 R_⊕ (−9%), and −0.13 (−5%) between our values and those from Adams et al. (2016, three K2OIs), Montet et al. (2015, five K2OIs, σ = 1 R_⊕), and Vanderburg et al. (2016a, 17 K2OIs, σ = 0.5 R_⊕), respectively. In comparison, our typical errors on planet radii are 6%–13%.

The two K2OIs with large changes are EPIC 201617985.01 and EPIC 201637175.01 (K2-22b). Crossfield et al. (2016) reported P = 1.143201 ± 0.000012 days for EPIC 201637175.01 and ${R}_{p}/{R}_{\star }={0.41}_{-0.34}^{+0.34}$ for EPIC 201617985.01 whereas we found P = 0.38 days and ${R}_{p}/{R}_{\star }={0.033}_{-0.024}^{+0.032}$. Our shorter orbital period for EPIC 201637175.01 is one third of the value reported by Crossfield et al. (2016) and matches earlier estimates by Sanchis-Ojeda et al. (2015), Vanderburg et al. (2016b), and Adams et al. (2016). Similarly, our revised planet/star radius estimate for EPIC 201617985.01 is in agreement with previous estimates by Montet et al. (2015, ${R}_{p}/{R}_{\star }=0.033$) and Vanderburg et al. (2016a, R_p/R_⋆ = 0.032).

We attribute the period error for EPIC 201637175.01 to the fact that the TERRA search did not consider threshold-crossing events with periods shorter than a day (Crossfield et al. 2016). For EPIC 201617985.01, the Crossfield et al. (2016) fit favored a large impact parameter ($b={1.38}_{-0.31}^{+0.41}$), which led to a broad posterior distribution of allowed planet radii. Excluding those two candidates does not alter the median differences between our revised values and those published by Crossfield et al. (2016), but considerably shrinks the standard deviations of the differences from 0.072 to 0.002 for the planet/star radius ratios and from 3 R_⊕ to 0.8 R_⊕ for the planet radii.

Several of the planets in our sample are particularly interesting because they might be habitable, reside in systems with multiple transiting planets, or orbit particularly bright stars. For reference, we present schematics of all of the multi-planet systems and potentially habitable systems in Figure 9. We also discuss individual systems in the following sections.

Zoom In Zoom Out Reset image size

For the purpose of planning potential follow-up observations, we first employed the planetary mass–radius relation presented in Weiss & Marcy (2014) to assign masses to the K2OIs and computed the expected radial velocity (RV) signal due to each planet. We display the resulting estimates in Figure 10. We note that the full range of small planet compositions is not well captured by a one-to-one mass–radius relation (Wolfgang & Lopez 2015; Chen & Kipping 2017), but these coarse mass estimates are sufficient for assessing the feasibility of future RV investigations.

Zoom In Zoom Out Reset image size

In general, we find that these planets would be challenging targets for current RV spectrographs due to the small expected signal (median value = 2.3 m s⁻¹) and the faintness of their host stars at optical wavelengths. In addition, several of the easiest candidates in Figure 10 are unsuitable for RV observations: EPIC 210558622 was identified as an spectroscopic binary (Crossfield et al. 2016), EPIC 212679181 is accompanied by a fainter star 15 away,²⁸ and EPIC 216892056.01 is almost certainly a false positive.

In contrast, the small Neptune EPIC 205924614.01 (K2-55b) is an ideal follow-up target and has already been observed by both Keck/HIRES and Spitzer. The validated small planets EPIC 210707130.01 and 212460519.01 are also attractive RV targets. We discuss K2-55b in detail in a separate publication (C. D. Dressing et al. 2017, in preparation) and consider the two smaller planets in Section 7.3.

Although our targets are generally poorly suited for RV mass measurements, they are more compelling candidates for transmission spectroscopy. The combination of the small sizes and red colors of their host stars produces larger transmission signals than would be expected for similar planets orbiting Sun-like stars with similar brightnesses in the Kepler bandpass. While our targets are moderately faint at visible wavelengths, they are modestly bright in the NIR. As shown in Figure 11, several planets should yield detectable transmission signals if our assumptions regarding planet masses and atmospheric compositions are correct.

Zoom In Zoom Out Reset image size

We approximated the transmission signals shown in Figure 11 by assuming that the planetary atmospheres extend for five scale heights. When calculating the latter, we adopted the same planet masses as for the RV predictions and determined planetary equilibrium temperatures using a fixed planetary albedo of 0.3. We also assumed that each planet possesses a Jupiter-like atmosphere with a mean molecular weight of 2.2m_H; planets with water-dominated atmospheres would generate transmission signals an order of magnitude smaller. While exciting from a habitability standpoint, the smallest planets in our sample may be bare rocky cores without substantial atmospheres and would therefore be poor targets for transmission spectroscopy.

Consulting Figure 11, the most intriguing targets for atmospheric characterization are EPIC 201205469.01 (K2-43b, Crossfield et al. 2016), 201345483.01 (K2-45b, Crossfield et al. 2016), 201635569.01 (K2-14b, Montet et al. 2015; Crossfield et al. 2016), 205924614.01 (K2-55b, Crossfield et al. 2016), 206318379.01 (K2-28b, Hirano et al. 2016; G. Chen et al. 2017, in preparation), 211509553.01, 211818569.01 (K2-121b), and 212554013.01 (K2-127b). These planets span a broad range in both size and equilibrium temperature: EPIC 211509553.01 is a cool giant planet (R_p = 10.8 ± 0.6 R_⊕, T_eq = 355 K); EPIC 201345483.01 is a hot giant planet (${R}_{p}={10.4}_{-0.7}^{+0.9}\,{R}_{\oplus }$, T_eq = 988 K); EPIC 201635569.01, EPIC 211818569.01, and EPIC 212554013.01 are warm giant planets (R_p = 7.5 ± 0.5 R_⊕, T_eq = 527 K; R_p = 9.2 ± 0.7 R_⊕, T_eq = 756 K; and ${R}_{p}={8.7}_{-0.6}^{+0.7}\,{R}_{\oplus }$, T_eq = 789 K, respectively); EPIC 201205469.01 and EPIC 205924614.01 are warm large Neptunes (R_p = 4.8 ± 0.3 R_⊕, T_eq = 676 K and R_p = 4.4 ± 0.3 R_⊕, T_eq = 861 K, respectively); and EPIC 206318379.01 is a warm small Neptune (R_p = 2.3 ± 0.3 R_⊕, T_eq =547 K). Of these planets, EPIC 205924614.01, 20618379.01, and 211818569.01 have the brightest host stars and are therefore the most favorable targets.

7.1. Systems with Planets in or Near the HZ

7.2. Systems with Multiple Transiting Planets

7.3. Systems Orbiting Bright Stars

Due to the dependence of planet radius estimates on host star characterization, our recent work improving the classification of cool dwarfs observed by K2 significantly affected the assumed properties of the associated planet candidates. In general, we found that the host star radii were underestimated by 8%–40% (D17), implying that the initial planet radius estimates were also undersized. In this paper, we investigated 79 candidate transit signals in the lightcurves of 74 low-mass K2 target stars and provided an updated catalog of planet properties. Our revisions to the system properties include both improved classifications of the host stars from D17 and our new transit fits. As part of the analysis, we also assessed the credibility of each planet candidate by using the VESPA framework developed by Morton (2012, 2015b) to calculate the probability that each transit-like event was caused by an astrophysical false positive rather than a genuine planetary transit.

In total, we considered 79 putative transit events. We rejected six as false positives, validated 16 as bona fide planets, and classified 17 new planet candidates. We also upheld the previous classifications of two false positives and provided updated planet radius estimates for 16 planet candidates and 22 validated planets announced in earlier publications.

Our cool dwarf planet sample is dominated by small worlds: 55% (39) are smaller than 2 R_⊕ and 85% (60) are smaller than Neptune. Compared to the planets detected during the prime Kepler mission, our candidates tend to orbit brighter stars that are more amenable to follow-up observations. In particular, the thirteen small planets with radii of 1.5–3 R_⊕ and host stars brighter than Ks = 11 are likely targets for atmospheric characterization studies with Spitzer, HST, and JWST. Furthermore, radial velocity observations may be feasible for the brightest systems, particularly given the advent of red spectrographs like CARMENES (Quirrenbach et al. 2010), the Habitable Zone Planet Finder (Mahadevan et al. 2010), the Infrared Doppler instrument (Tamura et al. 2012; Kotani et al. 2014), iSHELL (Rayner et al. 2012), and SpiROU (Delfosse et al. 2013; Artigau et al. 2014).

As the K2 mission continues, additional planets will be detected around stars across the ecliptic. We will continue to conduct follow-up observations for future targets and eagerly await the opportunity to characterize the best systems with the JWST. Beginning in 2018, we will expand our follow-up program to include planet candidates detected by the Transiting Exoplanet Survey Satellite (Ricker et al. 2014; Sullivan et al. 2015), which is scheduled to launch in 2018 March.

Many of our targets were provided by the K2 California Consortium (K2C2). We thank K2C2 for sharing their candidate lists and vetting products. We are grateful to Tim Morton for making VESPA publicly available and to Jennifer Winters, Nic Scott, and Lea Hirsch for assisting with speckle imaging. We also acknowledge helpful conversations with Chas Beichman, Eric Gaidos, Jessie Christiansen, Michael Werner, and Arturo Martinez. We thank the anonymous referee for providing helpful suggestions to improve the paper.

This work was performed under contract with the Jet Propulsion Laboratory (JPL) funded by NASA through the Sagan Fellowship Program executed by the NASA Exoplanet Science Institute. A.V. is supported by the NSF Graduate Research Fellowship, grant No. DGE 1144152. This publication was made possible through the support of a grant from the John Templeton Foundation. The opinions expressed here are those of the authors and do not necessarily reflect the views of the John Templeton Foundation. This paper includes data collected by the K2 mission, which is funded by the NASA Science Mission directorate. This research has made use of the NASA Exoplanet Archive, and the Exoplanet Follow-up Observation Program website, which are operated by the California Institute of Technology, under contract with the National Aeronautics and Space Administration under the Exoplanet Exploration Program. We appreciate the willingness of ExoFOP users to share their follow-up observations with the broader K2 community.

Our stellar characterization follow-up observations were obtained at the Infrared Telescope Facility, which is operated by the University of Hawaii under contract NNH14CK55B with the National Aeronautics and Space Administration and at Palomar Observatory. We thank the staff at both observatories and the Caltech Remote Observing Facilities staff for supporting us during our many observing runs. We are grateful to the IRTF and Caltech TACs for awarding us telescope time.

Our contrast curves were based on observations obtained at Gemini Observatory, Kitt Peak National Observatory via the NN-EXPLORE program time allocation at the WIYN telescope, and the W.M. Keck Observatory. Kitt Peak National Observatory, National Optical Astronomy Observatory, is operated by the Association of Universities for Research in Astronomy (AURA) under a cooperative agreement with the National Science Foundation. Gemini Observatory is operated by the Association of Universities for Research in Astronomy, Inc., under a cooperative agreement with the NSF on behalf of the Gemini partnership: the National Science Foundation (United States), the National Research Council (Canada), CONICYT (Chile), Ministerio de Ciencia, Tecnología e Innovación Productiva (Argentina), and Ministério da Ciência, Tecnologia e Inovação (Brazil). The W.M. Keck Observatory is operated as a scientific partnership among the California Institute of Technology, the University of California and the National Aeronautics and Space Administration. The Observatory was made possible by the generous financial support of the W.M. Keck Foundation.

The authors wish to recognize and acknowledge the very significant cultural role and reverence that the summit of Maunakea has always had within the indigenous Hawaiian community. We are most fortunate to have the opportunity to conduct observations from this mountain.

Facilities: Kepler - The Kepler Mission, K2 - , IRTF (SpeX) - , Palomar:Hale (TripleSpec - , PALM-3000/PHARO) - , Gemini-N (DSSI - , NIRI) - , Gemini-S (DSSI) - , Keck:II (NIRC2) - , WIYN (DSSI) - .

As discussed in Section 5, we have posted transit fits and false positive assessments for all K2OIs on the ExoFOP-K2 website. Figures 12–14 provide an example set of vetting data products. The K2OIs featured are a newly validated planet (EPIC211680698.01 = K2-118b, Figure 12), a newly detected planet candidate (EPIC 212634172.01, Figure 13), and a newly rejected false positive (EPIC 212679798.01, Figure 14).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size