Abstract
We present 44 validated planets from the 10th observing campaign of the NASA K2 mission, as well as high-resolution spectroscopy and speckle imaging follow-up observations. These 44 planets come from an initial set of 72 vetted candidates, which we subjected to a validation process incorporating pixel-level analyses, light curve analyses, observational constraints, and statistical false positive probabilities. Our validated planet sample has median values of = , Porb = days, = K, and J = mag. Of particular interest are four ultra-short period planets ( day), 16 planets smaller than 2 , and two planets with large predicted amplitude atmospheric transmission features orbiting infrared-bright stars. We also present 27 planet candidates, most of which are likely to be real and worthy of further observations. Our validated planet sample includes 24 new discoveries and has enhanced the number of currently known super-Earths ( ≈ 1–2), sub-Neptunes ( ≈ 2–4), and sub-Saturns ( ≈ 4–8) orbiting bright stars (J = 8–10 mag) by ∼4%, ∼17%, and ∼11%, respectively.
Export citation and abstract BibTeX RIS
1. Introduction
The K2 mission (Howell et al. 2014) is extending the Kepler legacy to a survey of the ecliptic plane, enabling the detection of transiting planets orbiting a wider range of host stars. The increased sky coverage of K2 has enabled the detection of planets orbiting brighter host stars, as well as a larger selection of M dwarfs (Crossfield et al. 2016; Dressing et al. 2017; Hirano et al. 2018a). As a result, K2 is yielding a large number of promising targets for follow-up studies (e.g., Crossfield et al. 2015; Montet et al. 2015; Petigura et al. 2015; Vanderburg et al. 2015, 2016a, 2016b, 2016c; Crossfield et al. 2017). K2 has also discovered planets in stellar cluster environments (David et al. 2016a; Mann et al. 2016a, 2017; Obermeier et al. 2016; Gaidos et al. 2017; Pepper et al. 2017; Ciardi et al. 2018), including one possibly still undergoing radial contraction (David et al. 2016b; Mann et al. 2016b).
We present here the results of our analysis of the K2 photometric data collected during Campaign 10 (C10), along with a coordinated campaign of follow-up observations to better characterize the host stars and rule out false positive scenarios. Because of C10's relatively high galactic latitude, blending within the photometric apertures is less significant than for other fields, and contamination from background eclipsing binaries is low. We detect 72 planet candidates and validate 44 of them as bona fide planets using our observational constraints, 24 of which have not previously been reported in the literature. Our sample contains a remainder of 27 planet candidates, many of which are likely real planets.
The transit detections and follow-up observations that led to these discoveries were the result of an international collaboration called KESPRINT. Formed from the merger of two previously separate collaborations (KEST and ESPRINT), KESPRINT is focused on detecting and characterizing interesting new planet candidates from the K2 mission (e.g., Dai et al. 2017; Fridlund et al. 2017; Gandolfi et al. 2017; Guenther et al. 2017; Niraula et al. 2017; Hirano et al. 2018b; Livingston et al. 2018; Smith et al. 2018; Van Eylen et al. 2018b).
The rest of the paper is structured as follows. In Section 2, we describe our K2 photometry and transit search. In Sections 3 and 4, we describe our follow-up speckle imaging and high-resolution spectroscopy of the candidates from our detection and vetting procedures. In Section 5, we describe our statistical validation framework and results. In Section 6, we discuss particular systems of interest, and we conclude with a summary in Section 7.
2. K2 Photometry and Transit Search
Here we describe how we produce a list of vetted planet candidates from the pixel data telemetered from the Kepler spacecraft, as well as detailed light curve analyses. Throughout this paper, we refer to stars by their nine-digit EPIC IDs, and we concatenate these with two-digit numbers to refer to planet candidates (ordered by orbital period).
2.1. Photometry
In C10, K2 observed a ∼110 square degree field near the North Galactic cap from 2016 July 06 to 2016 September 20. Long-cadence (30 minute) exposures of 28345 target stars were downlinked from the spacecraft, and the data were calibrated and subsequently made available on the Mikulski Archive for Space Telescopes23 (MAST). During the beginning of the campaign, a 3.5-pixel pointing error was detected and subsequently corrected six days after the start of observations. The data during this time is of substantially lower quality than the rest of the campaign, so we discard it in our analysis. An additional data gap was the result of the failure of detector module 4, which caused the photometer to power off for 14 days.
2.2. Systematics
Following the loss of two of its four reaction wheels, the Kepler spacecraft has been operating as K2 (Howell et al. 2014). The dominant systematic signal in K2 light curves is caused by the rolling motion of the spacecraft along its bore sight coupled with inter- and intra-pixel sensitivity variations. We used a method similar to that described by Vanderburg & Johnson (2014) to reduce this systematic flux variation. Our light curve production pipeline is as follows. We first downloaded the target pixel files from MAST. We laid circular apertures around the brightest pixel within the "postage stamp" (the set of pixels of the Kepler photometer corresponding to a given source). To obtain the centroid position of the image, we fitted a two-dimensional (2D) Gaussian function to the in-aperture flux distribution. We then fitted a piecewise linear function between the flux variation and the centroid motion of target. The fitted piecewise linear function was then detrended from the observed flux variation.
2.3. Transit Search
Before searching the light curve for transits, we first removed any long-term systematic or instrumental flux variations by fitting a cubic spline to the reduced light curve from the previous section. To look for periodic transit signals, we employed the box-least-squares algorithm (BLS; Kovács et al. 2002). We improved the efficiency of the original BLS algorithm by using a nonlinear frequency grid that takes into account the scaling of transit duration with orbital period (Ofir 2014). We also adopted the signal detection efficiency (SDE; Ofir 2014), which quantifies the significance of a detection. SDE is defined by the amplitude of peak in the BLS spectrum normalized by the local standard deviation. We empirically set a threshold of SDE > 6.5 for the balance between completeness and false alarm rate. In order to identify all the transiting planets in the same system, we progressively re-ran BLS after removing the transit signal detected in the previous iteration.
To search for additional transit signals that may have been missed by the transit search method described above, we used two separate pipelines: one based on the DST code (Cabrera et al. 2012), and one based on the wavelet-based filter routines VARLET and PHALET (Grziwa & Pätzold 2016). This helps to ensure higher detection rates, and the number of false positives is potentially reduced by utilizing multiple diagnostics. The DST code is optimized for space-based photometry and has been successfully applied to data from CoRoT and Kepler; we ran it on the light curves extracted by Vanderburg & Johnson (2014), which are publicly available from MAST. In the wavelet-based search, we first used VARLET to remove long-term stellar variability in the light curves and then searched for transits using a modified version of the BLS algorithm. Detected transit-like signals were then removed using PHALET, which combines phase-folding and a wavelet basis to approximate periodic features. In similar fashion to the above approach, we iterate this process of feature detection and removal to enable the detection of multi-planet systems.
2.4. Candidate Vetting
We performed a quick initial vetting to identify obvious false positives among the transiting signals identified in the previous section. Planetary candidates that survived the various tests were followed up with speckle imaging and reconnaissance spectra for proper statistical validation. We tested for the presence of any "odd–even" variations and significant secondary eclipse, both of which are likely signatures of eclipsing binaries. The odd–even effect is the variation of the eclipse depth between the primary and secondary eclipse of an eclipsing binary. If mistaken for planetary transits, the primary and secondary eclipses will be the odd and even numbered transits.
We fitted Mandel & Agol (2002) model to the odd and even transits separately. If a system shows odd–even variations with more than 3σ significance, it is flagged as a false positive. We also looked for any secondary eclipse in the light curve, using the Mandel & Agol (2002) model fit of the transits as a template for the occultation. After fitting the primary transits, we searched for secondary eclipses via an additional MCMC fitting step. We set the limb-darkening coefficients to zero and fixed all transit parameters except for two: the time of secondary eclipse and the depth of the eclipse. The resulting posterior distributions of these two parameters were then used to quantify the significance and phase of any putative secondary eclipses. For non-detections, we use the 3σ upper limit derived from the eclipse depth posterior to set the "maximum allowed secondary eclipse" constraint in our vespa analyses. If a system shows a secondary eclipse with more than 3σ significance, we calculated the geometric albedo using the depth of secondary eclipse. The object is likely self-luminous, hence likely a false positive, if the albedo is much greater than 1.
2.5. Stellar Rotation Periods
We also measured stellar rotation periods Prot from the variability in the light curves induced by starspot modulation. About half of the light curves of our candidates exhibited a lack of rotational modulation, or the K2 C10 time baseline was not long enough to constrain the period. For the rest, we used the autocorrelation function (ACF; e.g., McQuillan et al. 2014) to measure the rotational period, and we include these results in Table 1 along with initial estimates of the basic transit parameters of each candidate. To help ensure the validity of these measurements, we also used the Lomb–Scargle periodogram (Lomb 1976; Scargle 1982) to measure the rotational periods, and the results were in good agreement.
Table 1. Candidate Planets Detected in K2 C10
EPIC | Kp | Porb | T0 | T14 | Depth | SDE | Prot |
---|---|---|---|---|---|---|---|
(mag) | (days) | (BKJD) | (hr) | (days) | |||
201092629 | 11.9 | 26.810 | 2751.22 | 4.1 | 0.00090 | 13.2 | |
201102594 | 15.6 | 6.514 | 2753.24 | 2.0 | 0.00624 | 8.2 | 25 ± 3 |
201110617 | 12.9 | 0.813 | 2750.14 | 1.3 | 0.00029 | 16.2 | 16.8 ± 2.5 |
201111557 | 11.4 | 2.302 | 2750.17 | 1.9 | 0.02268 | 7.6 | 12.0 ± 1.8 |
201127519 | 11.6 | 6.179 | 2752.55 | 2.5 | 0.01303 | 11.6 | ⋯ |
201128338 | 13.1 | 32.655 | 2775.62 | 4.0 | 0.00159 | 6.7 | 15.6 ± 2.2 |
201132684 | 11.7 | 10.061 | 2757.49 | 3.8 | 0.00070 | 8.7 | 13.8 ± 1.3 |
201132684 | 11.7 | 5.906 | 2750.82 | 5.0 | 0.00015 | 9.7 | 13.8 ± 1.3 |
201164625 | 11.9 | 2.711 | 2750.15 | 3.1 | 0.00020 | 6.7 | 12.5 ± 1.5 |
201166680 | 10.9 | 24.941 | 2751.51 | 5.2 | 0.00019 | 6.6 | ⋯ |
201166680 | 10.9 | 11.540 | 2760.22 | 3.7 | 0.00016 | 7.8 | ⋯ |
201180665 | 13.1 | 17.773 | 2753.50 | 2.9 | 0.03662 | 11.2 | ⋯ |
201211526 | 11.7 | 21.070 | 2755.48 | 3.9 | 0.00030 | 8.3 | ⋯ |
201225286 | 11.7 | 12.420 | 2753.52 | 3.3 | 0.00065 | 11.6 | 20.8 ± 1.6 |
201274010 | 13.9 | 13.008 | 2756.51 | 2.2 | 0.00065 | 7.7 | ⋯ |
201352100 | 12.8 | 13.383 | 2761.79 | 2.2 | 0.00120 | 12.5 | 36 ± 11 |
201357643 | 12.0 | 11.893 | 2754.55 | 4.2 | 0.00107 | 12.3 | ⋯ |
201386739 | 14.4 | 5.767 | 2750.70 | 3.4 | 0.00134 | 11.1 | 35 ± 6 |
201390048 | 12.0 | 9.455 | 2750.92 | 3.0 | 0.02669 | 7.7 | ⋯ |
201390927 | 14.2 | 2.638 | 2750.34 | 1.7 | 0.00110 | 12.9 | ⋯ |
201392505 | 13.4 | 27.463 | 2759.08 | 5.5 | 0.00150 | 9.3 | ⋯ |
201437844 | 9.2 | 21.057 | 2757.07 | 4.4 | 0.00100 | 10.0 | ⋯ |
201437844 | 9.2 | 9.560 | 2753.52 | 3.5 | 0.00030 | 9.8 | ⋯ |
201595106 | 11.7 | 0.877 | 2750.05 | 1.0 | 0.00025 | 9.4 | ⋯ |
201598502 | 14.3 | 7.515 | 2755.43 | 2.3 | 0.00129 | 7.5 | ⋯ |
201615463 | 12.0 | 8.527 | 2753.77 | 3.7 | 0.00016 | 7.2 | ⋯ |
228707509 | 14.8 | 15.351 | 2752.51 | 3.6 | 0.02386 | 13.6 | ⋯ |
228720681 | 13.8 | 15.782 | 2753.42 | 3.4 | 0.01028 | 14.3 | 9.8 ± 1.1 |
228721452 | 11.3 | 4.563 | 2749.98 | 2.8 | 0.00020 | 12.6 | ⋯ |
228721452 | 11.3 | 0.506 | 2750.56 | 0.9 | 0.00010 | 9.6 | ⋯ |
228724899 | 13.3 | 5.203 | 2753.45 | 1.4 | 0.00113 | 12.3 | ⋯ |
228725791 | 14.3 | 6.492 | 2755.15 | 1.7 | 0.00110 | 9.8 | 32 ± 3 |
228725791 | 14.3 | 2.251 | 2749.97 | 1.2 | 0.00100 | 7.3 | 32 ± 3 |
228725972 | 12.5 | 4.477 | 2752.69 | 2.4 | 0.03270 | 11.5 | ⋯ |
228725972 | 12.5 | 10.096 | 2755.41 | 3.6 | 0.05928 | 13.0 | ⋯ |
228729473 | 11.5 | 16.773 | 2752.76 | 12.4 | 0.00199 | 11.6 | |
228732031 | 11.9 | 0.369 | 2749.93 | 1.0 | 0.00040 | 15.1 | 9.4 ± 1.9 |
228734900 | 11.5 | 15.872 | 2754.37 | 4.6 | 0.00034 | 8.0 | ⋯ |
228735255 | 12.5 | 6.569 | 2755.29 | 3.3 | 0.01280 | 12.6 | 31.1 ± 2.0 |
228736155 | 12.0 | 3.271 | 2751.02 | 2.4 | 0.00027 | 9.3 | ⋯ |
228739306 | 13.3 | 7.172 | 2755.11 | 2.8 | 0.00070 | 8.1 | ⋯ |
228748383 | 12.5 | 12.409 | 2750.04 | 5.9 | 0.00024 | 8.0 | ⋯ |
228748826 | 13.9 | 4.014 | 2751.13 | 2.4 | 0.00102 | 13.2 | |
228753871 | 13.2 | 18.693 | 2757.74 | 2.2 | 0.00082 | 7.7 | 16.4 ± 2.3 |
228758778 | 14.8 | 9.301 | 2756.07 | 2.7 | 0.00214 | 7.8 | ⋯ |
228758948 | 12.9 | 12.203 | 2753.83 | 4.0 | 0.00128 | 12.4 | 11.3 ± 1.7 |
228763938 | 12.6 | 13.814 | 2763.19 | 3.6 | 0.00036 | 8.8 | ⋯ |
228784812 | 12.6 | 4.189 | 2751.02 | 2.2 | 0.00014 | 8.9 | ⋯ |
228798746 | 12.7 | 2.697 | 2750.20 | 1.5 | 0.02587 | 14.1 | ⋯ |
228801451 | 11.0 | 8.325 | 2753.35 | 2.5 | 0.05325 | 12.9 | 19.5 ± 2.7 |
228801451 | 11.0 | 0.584 | 2750.46 | 1.5 | 0.01625 | 10.0 | 19.5 ± 2.7 |
228804845 | 12.6 | 2.860 | 2749.60 | 2.6 | 0.00020 | 7.3 | ⋯ |
228809391 | 12.6 | 19.580 | 2763.80 | 2.6 | 0.00100 | 8.3 | ⋯ |
228809550 | 14.7 | 4.002 | 2751.00 | 2.1 | 0.01259 | 12.5 | ⋯ |
228834632 | 14.9 | 11.730 | 2758.63 | 2.1 | 0.00111 | 8.6 | 23.6 ± 2.1 |
228836835 | 14.9 | 0.728 | 2750.26 | 0.8 | 0.00068 | 15.4 | ⋯ |
228846243 | 14.5 | 25.554 | 2756.93 | 5.4 | 0.00220 | 10.5 | ⋯ |
228849382 | 13.8 | 12.120 | 2757.61 | 2.4 | 0.00120 | 7.6 | ⋯ |
228849382 | 13.8 | 4.097 | 2749.96 | 1.6 | 0.00052 | 8.8 | ⋯ |
228888935 | 14.1 | 5.691 | 2751.67 | 3.3 | 0.00533 | 10.3 | 7.2 ± 1.1 |
228894622 | 13.3 | 1.964 | 2750.31 | 1.1 | 0.00183 | 16.3 | 20.8 ± 2.4 |
228934525 | 13.4 | 3.676 | 2752.05 | 1.7 | 0.00110 | 14.2 | 28.3 ± 3.1 |
228934525 | 13.4 | 7.955 | 2751.34 | 2.1 | 0.00110 | 11.4 | 28.3 ± 3.1 |
228964773 | 14.9 | 37.209 | 2776.76 | 3.1 | 0.00280 | 6.9 | ⋯ |
228968232 | 14.7 | 5.520 | 2753.52 | 3.6 | 0.00097 | 8.6 | ⋯ |
228974324 | 12.9 | 1.606 | 2750.29 | 1.3 | 0.00034 | 13.1 | 22.0 ± 2.3 |
228974907 | 9.3 | 20.782 | 2759.64 | 5.0 | 0.00010 | 7.2 | ⋯ |
229004835 | 10.2 | 16.138 | 2764.63 | 2.1 | 0.00036 | 10.6 | 22.2 ± 2.5 |
229017395 | 13.2 | 19.099 | 2753.28 | 6.0 | 0.00049 | 8.1 | ⋯ |
229103251 | 13.7 | 11.667 | 2756.72 | 3.1 | 0.00114 | 9.9 | ⋯ |
229131722 | 12.5 | 15.480 | 2752.71 | 4.2 | 0.00037 | 8.3 | ⋯ |
229133720 | 11.5 | 4.037 | 2750.96 | 1.5 | 0.00091 | 12.4 | 11.8 ± 1.3 |
Note. Kp denotes magnitude in the Kepler bandpass.
2.6. Transit Modeling
We used the orbital period, mid-transit time, transit depth, and transit duration identified by BLS as the starting points for more detailed transit modeling. The transit light curve was generated by the Python package batman (Kreidberg 2015). To reduce the data volume, we only use the light curve in a window centered on the mid-transit times. We first tested if any of the systems showed strong transit timing variations (TTVs). We used the Python interface to the Levenberg–Marquardt nonlinear least squares algorithm lmfit (Newville et al. 2014) to find the best-fit model of the phase-folded transit, and then fit this template to each transit separately to identify individual transit times of each candidate. Since none of the system presented in this work showed significant TTVs within the K2 C10 observations, we assumed linear ephemerides in subsequent analyses.
The transit parameters in our linear ephemeris model include the orbital period Porb, the mid-transit time T0, the planet-to-star radius ratio /, the scaled orbital distance a/, the impact parameter , and the transformed quadratic limb-darkening coefficients q1 and q2. Instead of fixing the parameters of the quadratic limb-darkening law to theoretical values based on stellar models, in this work we opt to allow these parameters to vary, as this allows for error propagation from stellar uncertainties. We utilize the available stellar parameters and their uncertainties to impose Gaussian priors on the limb-darkening coefficients (i.e., in the non-transformed parameter space, u1 and u2). To determine the location and width of these priors, we used a Monte Carlo method to sample the stellar parameters of each candidate host star (, , and ), and then used these to derive distributions of u1 and u2 from an interpolated grid based on the limb-darkening coefficients for the Kepler bandpass tabulated by Claret et al. (2012). We used the median and standard deviation of these distributions to define the Gaussian limb-darkening priors, and used uniform priors for all other parameters. Depending on the uncertainty in the stellar parameters, the limb-darkening priors determined in this way have typical widths of ∼10%, which is comparable to the uncertainty in the models used to predict them (e.g., Csizmadia et al. 2013; Müller et al. 2013). In addition, when the stars are active we do not expect agreement between theoretical and observed limb darkening because the tabulated theoretical values do not take into account the effects of stellar spots and faculae (Csizmadia et al. 2013). To account for the 30 minutes integration time of long-cadence K2 photometry, we used the built-in feature of batman to super-sample the model light curve by a factor of 16 before averaging every 3 minutes window (Kipping 2010).
We adopted a Gaussian likelihood function and found the maximum likelihood solution using scipy.optimize (Jones et al. 2001). We then sampled the joint posterior distribution using emcee (Foreman-Mackey et al. 2013), a Python implementation of the affine-invariant Markov Chain Monte Carlo ensemble sampler (Goodman & Weare 2010). We assumed the errors to be Gaussian, independent, and identically distributed, and thus described by a single parameter. In the maximum likelihood fits, we fixed the value of this parameter to the standard deviation of the out of transit flux, and during MCMC we fit for this value as a free parameter. We launched 100 walkers in the vicinity of the maximum likelihood solution and ran the sampler for 5000 steps, discarding the first 1000 as "burn-in." To ensure that the resultant marginalized posterior distributions consisted of 1000's of independent samples (enough for negligible sampling error) we computed the autocorrelation time of each parameter, and visual inspection revealed the posteriors to be smooth and unimodal. We summarize the transit parameter posterior distributions in Table 5 using the 16th, 50th, and 84th percentiles, and we use the posterior samples to compute other quantities of interest throughout this work (i.e., , ). The phase-folded light curves of the candidates are shown in Figure 1, with best-fitting transit model and 1σ (68%) credible region overplotted.
3. Speckle Imaging
We observed candidate host stars with the NASA Exoplanet Star and Speckle Imager (NESSI) on the 3.5 m WIYN telescope at the Kitt Peak National Observatory. NESSI is a new instrument that uses high-speed electron-multiplying CCDs (EMCCDs) to capture sequences of 40 ms exposures simultaneously in two bands (Scott et al. 2016, Scott et al. 2018). Data were collected following the procedures described by Howell et al. (2011). We conducted all observations in two bands simultaneously: a "blue" band centered at 562 nm with a width of 44 nm, and a "red" band centered at 832 nm with a width of 40 nm. The pixel scales of the "blue" and "red" EMCCDs are 00175649 and 00181887 per pixel, respectively. We make all of our speckle imaging data publicly available via the community portal ExoFOP.24 We list the individual NESSI data products used in this work in Table 9.
Speckle imaging data were reduced following the procedures described by Howell et al. (2011), resulting in diffraction limited reconstructed images (256 × 256 pixels) of each target star. The methodology has been described in detail in previous works (e.g., Horch et al. 2009, 2012, 2017), but we provide a brief review here for convenience.
First, the ACF of each 40 ms exposure is summed and Fourier transformed, resulting in the average spatial frequency power spectrum. The speckle transfer function is then deconvolved by dividing the target's power spectrum by that of the corresponding point source calibrator, yielding the square of the modulus estimate of the target's Fourier transform. The phase information can then be recovered from bispectral analysis, as first described by Lohmann et al. (1983). This is accomplished by computing the Fourier transform of the summed triple correlation function of the exposures, which in combination with the modulus estimate yields the complex Fourier transform of the target. This is then filtered with a low-pass 2D Gaussian before being inverse transformed, yielding the reconstructed image.
We extract background sensitivity limits from the reconstructed images by computing the mean and standard deviation of a series of concentric annuli centered on the target star, as described by Howell et al. (2011). We then compute contrast curves by fitting a cubic spline to the kernel-smoothed 5σ sensitivity limits, expressed as a magnitude difference relative to the target star as a function of radius. For stars of moderate brightness (V = 10–12 mag), we typically achieve contrasts of ∼4 mag at 02. See Figure 2 for a plot showing all of the contrast curves obtained in this work. We detect four candidate host stars with secondaries, see Table 2.
Download figure:
Standard image High-resolution imageTable 2. Stars with Detected Companions
EPIC | Δarcsec | Δmag | θ [deg. E of N] | Note |
---|---|---|---|---|
201352100 | 0.387 | 3.37 | 312.054 | a |
201390927 | 0.883 | 1.14 | 341.286 | a |
201392505 | 0.242 | 3.68 | 42.491 | b |
228964773 | 0.332 | 2.08 | 43.499 | b |
Note. All Detections made in the 832 nm Band. (a) The quadrant of the position angle is ambiguous, meaning it could be off by exactly 180°. (b) The binary model fit is of poor quality, so uncertainty may be larger than typical.
Download table as: ASCIITypeset image
4. High-resolution Spectroscopy
4.1. McDonald/Tull
Most of the high-resolution spectra presented in this paper were obtained with the Tull Coudé cross-dispersed echelle spectrograph (Tull et al. 1995) at the Harlan J. Smith 2.7 m telescope at McDonald Observatory. Observations were conducted with the 1.2 × 82 slit, yielding a resolving power of R ∼ 60000. The spectra cover 375–1020 nm, with increasingly larger inter-order gaps long-ward of 570 nm. For each target star, we obtained three successive short exposures in order to allow removal of energetic particle hits on the CCD detector. We used an exposure meter to obtain an accurate flux-weighted barycentric correction and to give an exposure length that resulted in a signal/noise ratio of about 30 per pixel. Bracketing exposures of a Th–Ar hollow cathode lamp were obtained in order to generate a wavelength calibration and to remove spectrograph drifts. This enabled calculation of absolute radial velocities from the spectra. The raw data were processed using IRAF routines to remove the bias level, inter-order scattered light, and pixel-to-pixel ("flat field") CCD sensitivity variations. We traced the apertures for each spectral order and used an optimal extraction algorithm to obtain the detected stellar flux as a function of wavelength.
We computed stellar parameters from our reconnaissance Tull spectra using Kea (Endl & Cochran 2016). In brief, we used standard IRAF routines to perform flat fielding, bias subtraction, and order extraction, and we used a blaze function determined from high signal-to-noise ratio (S/N) flat field exposures to correct for curvature induced by the blaze. Kea uses a large grid of synthetic model stellar spectra to compute stellar effective temperatures, surface gravities, and metallicities. See Table 6 for the stellar parameters used in this work. From a comparison with higher S/N spectra obtained with Keck/HIRES, we found typical uncertainties of 100 K in , 0.12 dex in , and 0.18 dex in . For a detailed description of Kea, see Endl & Cochran (2016).
4.2. NOT/FIES
We also used the FIber-fed Échelle Spectrograph (FIES; Frandsen & Lindberg 1999; Telting et al. 2014) on the 2.56 m Nordic Optical Telescope (NOT) of Roque de los Muchachos Observatory (La Palma, Spain) to collect high-resolution () spectra of four C10 candidate host stars: 228729473, 228735255 (K2-140; Giles et al. 2018, Korth et al., submitted to MNRAS), 201127519, and 228732031 (K2-131; Dai et al. 2017). The observations were carried out between 2017 February 15 and May 23 UTC, within observing programs 54-027, 55-019, and 55-202. We followed the same strategy as in Gandolfi et al. (2013) and traced the RV drift of the instrument by bracketing the science exposures with 90 s ThAr spectra. We reduced the data using standard IRAF routines and extracted the RVs via multi-order cross-correlations using different RV standard stars observed with the same instrument.
4.3. TNG/HARPS-N
We observed the stars 228801451, 228732031 (K2-131; Dai et al. 2017), 201595106, and 201437844 (HD 106315; Crossfield et al. 2017; Rodriguez et al. 2017) with the HARPS-N spectrograph ( 115,000; Cosentino et al. 2012) mounted at the 3.58 m Telescopio Nazionale Galileo (TNG) of Roque de los Muchachos Observatory (La Palma, Spain). The observations were performed in 2017 January as part of observing programs A34TAC_10 and A34TAC_44. We reduced the data using the dedicated off-line pipeline and extracted the RVs by cross-correlating the échelle spectra with a G2 numerical mask. The HARPS-N data of 228732031 have been published by our team in Dai et al. (2017). We refer the reader to that paper for a detailed description and analysis of the data. We list the results of our analysis of these spectra in Table 10.
4.4. Stellar Properties
We obtained spectra for 27 candidate host stars in this work, from which we derived , , , and , as described in Section 4.1. We augment this set of spectroscopic stellar parameters with values from the literature for an additional 14 candidate host stars (Rodriguez et al. 2017; Hirano et al. 2018a; Mayo et al. 2018). To maximize both the quality and uniformity of the final set of stellar parameters we use in this work, we adopted the following strategy. First, we gathered 2MASS JHK photometry and Gaia DR2 parallaxes for all stars; 2MASS photometry is available in the EPIC, and we cross-matched to Gaia DR2 using both position and optical magnitude agreement (Kp and Gaia G band). We then used the isochrones (Morton 2015a) interface to the Dartmouth stellar model grid (Dotter et al. 2008) to estimate stellar parameters and their uncertainties using the MultiNest sampling algorithm (Feroz et al. 2013). For those stars with parameters from spectroscopic analyses, we imposed Gaussian priors on , , and , with mean and standard deviation set by the spectroscopically derived values and their uncertainties. We also ran the same analysis without including parallax, as a check on the quality of the parameters derived in this manner without any distance information; unsurprisingly, we found that including parallax yielded the biggest improvement for stars lacking spectroscopy. This is perhaps most important for the M dwarfs in our sample, which suffer from systematically underestimated radii in the EPIC (see, e.g., Dressing et al. 2017).
As an additional quality check, we also performed spectral analyses for the targets 201127519, 201437844, 201595106, and 228801451, using spectra from FIES and HARPS-N and SpecMatch-emp (Yee et al. 2017). SpecMatch-emp fits the input spectra to hundreds of library template spectra collected by the California Planet Search, and the stellar parameters (, , and ) are estimated based on the interpolation of the parameters for best-matched library stars. Among them, 201127519, 201595106, and 228801451 were also observed with the Tull spectrograph, and the resulting parameters by SpecMatch-emp are in agreement within ∼1.5σ with those estimated from the Tull spectra by the Kea code. For HD 106315, we obtained = 6326 ± 110 K, = 1.86 ± 0.30 R⊙, and = −0.20 ± 0.08. While and agrees within 1σ with the literature values (Crossfield et al. 2017; Rodriguez et al. 2017), exhibits a moderate disagreement with that in the literature ( = R⊙ Rodriguez et al. 2017). This is probably due to the small number of library stars in SpecMatch-emp in the region with > 6300 K, but this disagreement does not have any impact on our results.
5. Planet Validation
5.1. Statistical Framework
We use the open source vespa software package (Morton 2012, 2015b) to compute the false positive probabilities (FPPs) of each planet candidate. vespa uses the TRILEGAL Galaxy model (Girardi et al. 2005) to compute the posterior probabilities of both planetary and non-planetary scenarios given the observational constraints, and considers false positive scenarios involving simple eclipsing binaries, blended background eclipsing binaries, and hierarchical triple systems. vespa models the physical properties of the host star, taking into account any available broadband photometry and spectroscopic stellar parameters, and compares a large number of simulated scenarios to the observed phase-folded light curve. Both the size of the photometric aperture and contrast curve constraints are accounted for in the calculations, as well as any other observational constraints such as the maximum depth of secondary eclipses allowed by the data. We adopt a fiducial validation criterion of FPP < 0.01, which is reasonably conservative and also consistent with the literature (e.g., Montet et al. 2015; Crossfield et al. 2016; Morton et al. 2016). vespa utilizes the contrast curves derived from the observations listed in Table 9 and described in Section 3. To minimize the possibility of errors in the vespa calculations induced by zero-point offsets or underestimated uncertainties in broadband photometry, we opt to use only the well-calibrated 2MASS JHK magnitudes and their uncertainties, taken from the EPIC, in addition to the Kepler band magnitude required by vespa. The stellar parameters used as input to vespa are identical to those used in our uniform isochrones analysis (see Section 4.4). In addition to stellar parameters, vespa utilizes basic system properties (i.e., R.A., decl., Porb, /), as well as contrast curves (see Section 3) and constraints on secondary eclipse depth and maximum exclusion radii (see Table 8). We tabulate candidate parameters along with their FPPs and final dispositions in Table 5, and the full vespa likelihoods are listed in Table 7. We denote final dispositions as follows: "VP" = validated planet; "PC" = planet candidate; "FP" = false positive.
All of the candidates we detect in multi-planet systems meet the fiducial validation criterion of FPP < 1%. However, FPPs computed with vespa treat only the individual planet candidates in isolation and thus do not take into account any multiplicity in each system. Stars with multiple transiting planet candidates have been shown to exhibit a lower false positive rate by an order of magnitude (Lissauer et al. 2011, 2012, 2014). For this reason, we apply a "multiplicity boost" factor to the planet probability appropriate for each candidate in a multi-planet system. Lissauer et al. (2012) estimated a multiplicity boost factor of 25 for systems containing two planet candidates in the Kepler field, and we apply the same factor in this work. To check that this factor is appropriate for K2 C10, we follow Sinukoff et al. (2016) and utilize Equations (2) and (4) of Lissauer et al. (2012) to estimate the sample purity P from the integrated FPP of our sample and the number of planet candidates we detect (72). This estimate of P is quite high, perhaps due to a lack of contamination from background stars due to the high galactic latitude of the field, or due to our team's vetting procedures. The fraction of detected planet candidates in multi-systems (18/72) in conjunction with the high sample purity yields a multiplicity boost which is significantly higher than the factor of 25 estimated by Lissauer et al. (2012) for the Kepler field. Although the true value is likely to be higher, we conservatively apply only a factor of 25, consistent with Lissauer et al. (2012), and the FPPs in Table 5 reflect this accordingly.
5.2. Stellar Companions
To ensure that the FPPs computed by vespa are reliable, we take into account the presence of any nearby stars detected in speckle or archival imaging. Table 2 lists the nearby stars we detected via speckle imagine, along with their separations and delta-magnitudes relative to the primary stars. Figure 3 shows the reconstructed speckle images for these stars, and Figure 2 shows these detections relative to the ensemble of contrast curves from all of our speckle images. Table 3 lists those stars found in the EPIC to be near and bright enough to be the source of the observed transit signals.
Download figure:
Standard image High-resolution imageTable 3. EPIC Sources within the Photometric Apertures which are Bright Enough to Produce the Observed Transit-like Signals
EPIC | Contaminant | ρ (arcsec) | ΔKp (mag) |
---|---|---|---|
201111557 | 201111694 | 15.90 | 5.187 |
201164625 | 201164669 | 17.58 | 3.228 |
201595106 | 201595004 | 13.62 | 5.839 |
228707509 | 228707572 | 12.48 | 1.563 |
228720681 | 228720649 | 7.86 | 2.905 |
228758948 | 228758983 | 9.00 | 3.267 |
Download table as: ASCIITypeset image
5.2.1. Companions Detected in High-resolution Imaging
On the nights of 2017 March 15, 17, and 18 we acquired speckle imaging of the stars 201352100, 201390927, 201392505, and 228964773 (see Table 9). We detected companions in the reconstructed images (see Figure 3), so we assessed the possibility that the transit signal might not originate from the primary stars. We used the following relation between the observed transit depth and the true transit depth δ in the presence of dilution from a companion magnitudes fainter than the primary star:
Assuming a maximum eclipse depth of 100% (i.e., a brown dwarf—M dwarf binary) we can potentially rule out the secondary star as the source of the observed signal. For shallower transits the maximum allowed dilution from the primary is larger, and therefore even a relatively faint secondary source cannot be ruled out as the host. For each of these four candidates, the secondary source is bright enough (given the observed transit depth) that we cannot rule out the possibility they are the source of the signal (see Table 2). For this reason, we do not validate any of these candidates as planets, as we do not know the true source of the signal (and therefore the true planet size), even though they all have low FPPs.
5.2.2. Companions in the EPIC
In addition to analyzing the scenarios involving companions detected in high-resolution speckle imaging, we also performed a search of the EPIC for any additional stars within the photometric apertures which could be the source of the observed signals. Most of these queries yielded no stars within the aperture other than the primary, but there were some cases in which the query yielded a star bright enough to be the source of the observed transit signal; we list these cases in Table 3. Despite their low FPPs, we do not validate these candidates because we do not know which star is the true host. As we expect most of these candidates to be genuine planets, they present good validation opportunities via higher angular resolution follow-up transit observations, either from the ground or from space (i.e., with Spitzer or CHEOPS).
5.2.3. Archival Imaging
As a check on the accuracy of the sources comprising the EPIC, we also queried 1' × 1' Pan-STARRS-125 grizy images centered at the position of each candidate host star. We found good agreement with the catalog query: nearby stars found by the catalog query were clearly visible in the images, and no nearby bright sources were seen in the images that were not previously found by the catalog query. We show these images in Figure 4, with overplotted circular regions illustrating the size and location of the apertures used to extract photometry from the K2 pixel data.
Download figure:
Standard image High-resolution image5.3. Multi-aperture Light Curve Analysis
In light of several recent cases of contamination from false positives in statistically validated planet samples (Cabrera et al. 2017; Shporer et al. 2017), we also scrutinized our candidates at the pixel level. To do so, we extracted light curves from different sized apertures and looked for signs of a dependence of transit depth on aperture radius. In some cases, these light curves are too noisy to draw conclusions from, as they are extracted from "non-optimal" apertures. However, this analysis is especially important when there are widely separated neighboring stars (i.e., several Kepler pixels away) that still contribute flux to the K2 apertures, in which case it may be possible to determine the origin of the transit-like signal by this method. Based on these analyses we found that the transit signal associated with the candidate 201164625.01 most likely originates from the neighboring star, 201164669 (see Table 3 and Figure 4). We also detected suspicious transit depth behavior in the light curves of 201392505.01 and 228964773.01, both of which have nearby companions detected in speckle imaging. Intriguingly, these companions are well within a Kepler pixel of the target star, so even the smallest aperture possible (one Kepler pixel) should contain light from both the primary and secondary stars. This result may indicate the presence of another (undetected) star further away, and suggests that such multi-aperture analyses should be useful for ranking the quality of candidates when high-resolution imaging is unavailable.
5.4. Transit S/N
As a final step in the validation process, we compute the transit S/N for each candidate in order to enforce a minimum transit quality standard for all planets in the validated sample. We compute the transit S/N using the simple approximation that the signal scales with the transit depth and the square root of the number of transits (e.g., Bouma et al. 2017). We estimate the noise by computing the standard deviation of the out-of-transit photometry used in our light curve fits and scaling it from the K2 observing cadence to the transit duration of each candidate. We find median S/N values of 17.1 and 17.6 for the validated and candidate samples, respectively. The slightly lower S/N of the validated sample is likely attributable to the fact that candidates with higher FPPs are typically larger and have correspondingly deeper transits, whereas the vast majority of our validated planets are sub-Neptunes (see Figure 5). Our validated sample consists of planets with S/N > 10, with the exception of K2-254 b and K2-247 c, which have S/N values of 6.7 and 8.9, respectively. However, these are both in multi-planet systems, which increases our confidence in the veracity of the transit signals. We argue that candidates with relatively low S/N found in systems with multiple validated candidates need not be regarded with as much suspicion as similarly low-S/N candidates in single-candidate systems; this is related to, but more qualitative than, the "multi-boost" argument of Lissauer et al. (2012). Indeed, many interesting planets with low S/N likely remain to be found in both the Kepler and K2 data (e.g., Shallue & Vanderburg 2018).
Download figure:
Standard image High-resolution image5.5. Pipeline Comparison
To check the quality of our light curves and provide an additional layer of confidence in our candidates, we performed a parallel analysis using light curves from an independent K2 pipeline. We first downloaded the light curves of Vanderburg & Johnson (2014) from MAST for all the targets listed in Table 1, then detrended the light curves by fitting a second order polynomial to the out-of-transit data using exotrending (Barragán & Gandolfi 2017). To explore the transit model parameter space with MCMC, we used pyaneti (Barragán et al. 2017) to fit the detrended light curves with uniform priors for all parameters; more description of the pyaneti MCMC evolution and parameter estimation can be found in Barragán et al. (2018) and Gandolfi et al. (2017). For the majority of candidates, the main transit parameters of interest (Porb, /, b, and a/) are consistent within 1σ between our two independent analyses, although there are some cases in which marginally significant differences were found. These differences are likely to be the result of different handling of the K2 systematics and/or the stellar variability in the light curves. The overall good agreement between these two independently derived sets of transit parameters provides an additional layer of confidence in the quality of the candidates. The results of this comparison are listed in Table 12.
6. Discussion
6.1. Validated Planets
We validate 44 planets out of our sample of 72 candidates, and tabulate the FPPs along with parameter estimates of interest in Table 5. Of the 44 validated planets we report here, 20 of them have been previously statistically validated or confirmed: 201598502.01, 228934525.01, and 228934525.02 (K2-153 b, K2-154 bc; Hirano et al. 2018a); 228735255.01 (K2-140 b; Giles et al. 2018, Korth et al., submitted to MNRAS); 201437844.01 and 201437844.02 (HD 106315bc; Crossfield et al. 2017; Rodriguez et al. 2017); 228732031.01 (K2-131 b; Dai et al. 2017); and 13 others were recently validated by Mayo et al. (2018). In the left panel of Figure 5 we plot the planetary radii, orbital periods, and equilibrium temperatures of the validated planets in the sample.
We investigated the impact of these new planets to the population of currently known planets by querying the NASA Exoplanet Archive26 (Akeson et al. 2013). We computed the fractional enhancement to the known population due to the 44 planets as a function of planet size and host star brightness (see Figure 6). As of 2018 June 12, the populations of super-Earths ( ≈ 1–2), sub-Neptunes ( ≈ 2–4), and sub-Saturns ( ≈ 4–8) orbiting bright stars (J = 8–10 mag) are enhanced by ∼4%, ∼17%, and ∼11%, respectively. Because of the brightness of the host stars, many of these planets are ideal for detailed characterization studies via precision Doppler and transmission spectroscopy, which we discuss in greater detail in Section 6.4.
Download figure:
Standard image High-resolution image6.2. Candidates
Out of the 72 planet candidates we present here, 27 are not validated. Most cannot be validated due to the FPP being above our fiducial validation criterion of 1% or the presence of a contaminating star within the photometric aperture. See Table 7 for the likelihoods of various false positive scenarios and the planet scenario, as computed by vespa. There are several candidates which we do not validate for other reasons, which we discuss below. In the right panel of Figure 5, we plot the planetary radii, orbital periods, and equilibrium temperatures of the non-validated candidates.
The candidate 228729473.01 exhibits a long transit duration, and subsequent spectroscopic analyses revealed large RV variations which are consistent with the candidate being a false positive involving an M dwarf eclipsing a sub-giant, see S. Csizmadia et al. (2018, in preparation) for more details. The light curve of 229133720.01 exhibits low levels of variability in phase with the transit signal, which could be due to ellipsoidal variations; thus we do not validate the candidate in spite of its low FPP. Although 201390048.01 was recently validated (K2-162 b; Mayo et al. 2018), we found marginal evidence of odd–even variations in the light curve of this candidate, which could be an indication that the signal is actually caused by an eclipsing binary at twice the estimated orbital period. Although vespa accounts for this scenario in its FPP calculation, we do not validate the candidate even though its FPP is below 1%. The candidate 201180665.01 has a relatively high FPP (∼64%), and also a suspiciously large radius estimate (∼26 ). Although spectroscopic characterization could yield a different radius estimate for the host star (and thus also for the candidate), we conclude that this is most likely an eclipsing M dwarf companion. The candidates 228974907.01, and 228846243.01 do not have particularly low FPPs, but they may be interesting targets for further observations due to their relatively long orbital periods. The candidate 201128338.01 was statistically validated previously in the literature (K2-152 b; Hirano et al. 2018a); we find a similarly low FPP, but we do not validate it simply because it has fewer than three transits in the K2 photometry (and thus odd/even variations in transit depth cannot be robustly ruled out). Further observations will shed light on the true nature of these candidates, either by measuring RV variations with precision spectrographs or via simultaneous multi-band transit observations with instruments such as MuSCAT (Narita et al. 2015) and MuSCAT2 (a griz clone of MuSCAT now in operation at Teide Observatory).
The integrated FPP is ∼2.1 for the full set of 72 candidates, which implies the existence of two false positives in the sample. We have already confirmed that 228729473.01 is a false positive via RV observations (see S. Csizmadia et al. 2018, in preparation), and we suspect 229133720.01, 201390048.01, and 201180665.01 of being false positives, as described above. Therefore, we expect no false positives among the remainder of the sample, and most of the 27 unvalidated candidates could be statistically validated or confirmed by future observations.
6.3. Interesting New Systems
6.3.1. Ultra-short Period Planets (USPs)
USPs are defined by having orbital periods less than one day (e.g., Sanchis-Ojeda et al. 2013, 2015). Our validated planet sample contains four USPs: K2-131 b (Dai et al. 2017); K2-156 b and K2-223 b (Mayo et al. 2018); and K2-229 b (Mayo et al. 2018; Santerne et al. 2018). These planets join a growing list of USPs discovered by K2 (e.g., Vanderburg et al. 2016b; Adams et al. 2017; Barragán et al. 2018; Christiansen et al. 2017; Dai et al. 2017; Gandolfi et al. 2017; Malavolta et al. 2018). The radii of these USPs place all three of them below the recently observed gap in the radius distribution (Fulton et al. 2017; Van Eylen et al. 2018a) which was predicted as a consequence of photoevaporation (e.g., Owen & Wu 2013; Lopez & Fortney 2014). These three USPs are therefore likely to be rocky and have high densities, consistent with having lost any primordial or secondary atmospheres they might once have had. Of these validated USPs, we measured the metallicity of the host stars spectroscopically for three of them; K2-229 appears to have only a modestly sub-solar metallicity of −0.09 ± 0.02 , but K2-131 and K2-156 have more significantly sub-solar metallicities of −0.17 ± 0.03 and −0.25 ± 0.06 , respectively (see Table 6). Due to their small size, these USPs are likely to have a mass less than 5–6 , and thus the sub-solar metallicity of their host stars would be consistent with the USP mass–metallicity trend noted by Sinukoff et al. (2017) (i.e., similar to Kepler-78 b and Kepler-10 b).
The G dwarf K2-223 and K dwarf K2-229 are both relatively bright (Kp ∼ 11 mag), and host planets with predicted masses and Doppler semi-amplitudes well within the reach of current precision spectrographs, such as HARPS or HIRES. K2-156 b orbits a slightly fainter star and has a slightly smaller predicted mass and Doppler semi-amplitude, but is also a viable target for characterization with today's instrumentation. Such mass measurements would yield densities and constrain the bulk compositions of these USPs, which would enable tests of USP formation theories.
In addition to the four validated USPs mentioned above, we also note that our sample contains two USP candidates: 201595106.01 and 228836835.01. We do not validate 201595106.01 because of the presence of a faint star in the EPIC with a of 5.839 and a separation of 1362 (see Table 3), which is within the photometric aperture we used to extract the K2 light curve. We do not validate 228836835.01 because it has a FPP of ∼4% and thus does not meet our validation criterion. Future observations could potentially rule out false positive scenarios for both of these candidates, resulting in the validation of two more USPs from K2 C10.
6.3.2. Multi-planet Systems
Of the 44 validated planets in our sample, 18 of them were found in two-planet systems, which enables the study of their orbital architectures and evolution. Four of these systems have orbital architectures with period ratios just wide of a 2:1 commensurability, and two are close to a 3:1 commensurability. The pairs closest to 2:1 are K2-243 bc and K2-154 bc, which both have . The relatively large fraction of multi-planet systems (4/9) in our sample with period ratios just wide of a 2:1 commensurability is reminiscent of the distribution of orbital architectures observed with Kepler (Fabrycky et al. 2014). K2-254 bc and K2-247 bc are both just inside a 3:1 commensurability, with period ratios of and , respectively. Although we did not detect any significant TTVs in the K2 data, some of these systems may have TTVs which could be detected with higher cadence transit observations.
Intriguingly, two of the four validated USPs in the sample were found in two-planet systems with large period ratios, similar to the Kepler-10 system: K2-223 bc has , and K2-229 bc has . The presence of an additional transiting planet decreases the likelihood that these USPs reached their current orbits via dynamical scattering, as this would increase the chances of higher mutual inclinations; even after tidal circularization, the geometric transit probability would be decreased by a higher likelihood of non-coplanarity. This is consistent with previous analyses in which USP systems have been noted to be dynamically cold (e.g., Dai et al. 2017).
6.4. Characterization Targets
We predicted the masses of the candidates using the probabilistic mass–radius relation of Wolfgang et al. (2016) 27 (see Table 5). The predicted masses enabled us to compute other quantities of interest, which we then used to identify potentially interesting targets for follow-up characterization via Doppler and transmission spectroscopy.
6.4.1. Doppler Targets
We computed the expected Doppler semi-amplitude due to the reflex motion of the host star induced by each planet (see Table 5). We used these expected semi-amplitudes in conjunction with the brightness of the host stars to identify planets in the sample which are good targets for radial velocity (RV) follow-up study using current and future facilities. Such RV observations will reveal the planets' densities and constrain their bulk compositions. This is of particular interest for relatively small planets with radii in the range 1.5–2.5 because such measurements could enable tests of planet formation theories and post-processes, such as the photoevaporation (e.g., Owen & Wu 2013; Lopez & Fortney 2014), which has been proposed to explain the observed gap in the radius distribution (Fulton et al. 2017; Van Eylen et al. 2018a). However, because of the difficulty of detecting the small Doppler signals of such planets, it is especially important to identify such planets which are orbiting relatively bright stars, for which the RV precision required to measure their masses is more readily obtainable. Table 4 lists validated planets with predicted Doppler semi-amplitudes greater than 1 m s−1 orbiting stars brighter than Kp = 12 mag. For convenience, we also list planetary orbital periods and stellar rotational periods (when available); potentially confounding quasi-periodic RV signals produced by stellar magnetic activity are less likely to present a challenge for mass measurement when the orbital period is far from the stellar rotational period (or a harmonic). We note that 228732031.01 (K2-131 b) and 228801451.01 (K2-229 b) both already have measured masses via precision RVs (Dai et al. 2017; Santerne et al. 2018).
Table 4. Validated Planets with Predicted Doppler Semi-amplitudes Greater than 1 m s−1 Orbiting Stars Brighter than Kp = 12 mag
EPIC | Kp | Kpred | Porb | Prot | |
---|---|---|---|---|---|
(mag) | (m s−1) | () | (days) | (days) | |
201092629.01 | 11.858 | 2.55 | 26.8199 | ||
201132684.01 | 11.678 | 2.64 | 10.0605 | 13.8 ± 1.3 | |
201132684.02 | 11.678 | 1.28 | 5.9028 | 13.8 ± 1.3 | |
201166680.02 | 10.897 | 2.17 | 11.5418 | ⋯ | |
201166680.03 | 10.897 | 2.01 | 24.9460 | ⋯ | |
201211526.01 | 11.696 | 1.75 | 21.0688 | ⋯ | |
201225286.01 | 11.729 | 2.26 | 12.4220 | 20.8 ± 1.6 | |
201357643.01 | 11.998 | 4.34 | 11.8931 | ⋯ | |
201437844.01 | 9.234 | 2.32 | 9.5580 | ⋯ | |
201437844.02 | 9.234 | 4.31 | 21.0579 | ⋯ | |
201615463.01 | 11.964 | 2.19 | 8.5270 | ⋯ | |
228721452.02 | 11.325 | 1.57 | 4.5633 | ⋯ | |
228732031.01 | 11.937 | 1.70 | 0.3693 | 9.4 ± 1.9 | |
228734900.01 | 11.535 | 3.49 | 15.8721 | ⋯ | |
228801451.01 | 10.955 | 1.14 | 0.5843 | 19.5 ± 2.7 | |
228801451.02 | 10.955 | 2.03 | 8.3273 | 19.5 ± 2.7 |
Note. 228721452.01 is not listed here because it does not meet these criteria, but RV measurements to constrain the mass of 228721452.02 could also reveal the inner planet's mass, as both Keplerian signals would need to be accounted for in the RV analysis.
Download table as: ASCIITypeset image
Table 5. Planet and Candidate Parameters
EPIC | Name | Porb | T0 | a | b | a | MP | Kpred | FPP | Disposition | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
(days) | (BKJD) | () | () | (au) | () | (K) | () | (m s−1) | ||||||
201092629.01 | K2-241 b | VP | ||||||||||||
201102594.01 | K2-242 b | VP | ||||||||||||
201110617.01 | K2- 156 b | VP | ||||||||||||
201111557.01 | PC | |||||||||||||
201127519.01 | PC | |||||||||||||
201128338.01 | K2-152 b | PC | ||||||||||||
201132684.01 | K2-158 b | VP | ||||||||||||
201132684.02 | K2-158 c | VP | ||||||||||||
201164625.01 | PC | |||||||||||||
201166680.02 | K2-243 b | VP | ||||||||||||
201166680.03 | K2-243 c | VP | ||||||||||||
201180665.01 | PC | |||||||||||||
201211526.01 | K2-244 b | VP | ||||||||||||
201225286.01 | K2-159 b | VP | ||||||||||||
201274010.01 | PC | |||||||||||||
201352100.01 | PC | |||||||||||||
201357643.01 | K2-245 b | VP | ||||||||||||
201386739.01 | K2-246 b | VP | ||||||||||||
201390048.01 | K2-162 b | PC | ||||||||||||
201390927.01 | PC | |||||||||||||
201392505.01 | PC | |||||||||||||
201437844.01 | HD 106315 b | VP | ||||||||||||
201437844.02 | HD 106315 c | VP | ||||||||||||
201595106.01 | PC | |||||||||||||
201598502.01 | K2-153 b | VP | ||||||||||||
201615463.01 | K2-166 b | VP | ||||||||||||
228707509.01 | PC | |||||||||||||
228720681.01 | PC | |||||||||||||
228721452.01 | K2- 223 b | VP | ||||||||||||
228721452.02 | K2-223 c | VP | ||||||||||||
228724899.01 | PC | |||||||||||||
228725791.01 | K2-247 b | VP | ||||||||||||
228725791.02 | K2-247 c | VP | ||||||||||||
228725972.01 | K2-224 b | VP | ||||||||||||
228725972.02 | K2-224 c | VP | ||||||||||||
228729473.01 | FP | |||||||||||||
228732031.01 | K2-131 b | VP | ||||||||||||
228734900.01 | K2-225 b | VP | ||||||||||||
228735255.01 | K2-140 b | VP | ||||||||||||
228736155.01 | K2-226 b | VP | ||||||||||||
228739306.01 | K2-248 b | VP | ||||||||||||
228748383.01 | K2-249 b | VP | ||||||||||||
228748826.01 | K2-250 b | VP | ||||||||||||
228753871.01 | PC | |||||||||||||
228758778.01 | K2-251 b | VP | ||||||||||||
228758948.01 | PC | |||||||||||||
228763938.01 | K2-252 b | VP | ||||||||||||
228784812.01 | PC | |||||||||||||
228798746.01 | K2-228 b | VP | ||||||||||||
228801451.01 | K2-229 b | VP | ||||||||||||
228801451.02 | K2-229 c | VP | ||||||||||||
228804845.01 | K2-230 b | VP | ||||||||||||
228809391.01 | PC | |||||||||||||
228809550.01 | K2-253 b | VP | ||||||||||||
228834632.01 | PC | |||||||||||||
228836835.01 | PC | |||||||||||||
228846243.01 | PC | |||||||||||||
228849382.01 | K2-254 b | VP | ||||||||||||
228849382.02 | K2-254 c | VP | ||||||||||||
228888935.01 | PC | |||||||||||||
228894622.01 | K2-255 b | VP | ||||||||||||
228934525.01 | K2-154 b | VP | ||||||||||||
228934525.02 | K2-154 c | VP | ||||||||||||
228964773.01 | PC | |||||||||||||
228968232.01 | K2-256 b | VP | ||||||||||||
228974324.01 | K2-257 b | VP | ||||||||||||
228974907.01 | PC | |||||||||||||
229004835.01 | PC | |||||||||||||
229017395.01 | K2-258 b | VP | ||||||||||||
229103251.01 | PC | |||||||||||||
229131722.01 | K2-259 b | VP | ||||||||||||
229133720.01 | PC |
Note. MP is the mass predicted using the mass–radius relation of Wolfgang et al. (2016) (see Section 6.4). We note that this mass–radius relation was calibrated with sub-Neptunes similar in size to the vast majority of the planets in our validated sample; the predictions may not be accurate for larger candidates, but we report them here anyway for the sake of uniformity. The "Disposition" column indicates the final validation status of each candidate: "VP" = validated planet; "PC" = planet candidate; "FP" = false positive.
Table 6. Stellar Parameters
EPIC | log g | [Fe/H] | Mass | Radius | Distance | Provenance | ||
---|---|---|---|---|---|---|---|---|
(K) | (cgs) | (dex) | (M⊙) | (R⊙) | (pc) | (km s−1) | ||
201092629 | 4.54 ± 0.01 | 0.71 ± 0.01 | 0.75 ± 0.01 | 2.08 ± 0.29 | This work | |||
201102594 | 4.89 ± 0.01 | 0.38 ± 0.01 | 0.37 ± 0.01 | 109.6 ± 0.6 | ⋯ | ⋯ | ||
201110617 | 4597 ± 50 | −0.25 ± 0.06 | 0.66 ± 0.01 | 150.4 ± 1.0 | 1.80 ± 0.30 | This work | ||
201111557 | 4.62 ± 0.01 | −0.23 ± 0.23 | 0.71 ± 0.01 | 97.5 ± 0.5 | ⋯ | ⋯ | ||
201127519 | 0.03 ± 0.05 | 0.81 ± 0.03 | 0.77 ± 0.01 | 118.0 ± 0.7 | 1.85 ± 0.22 | This work | ||
201128338 | 4.67 ± 0.01 | 0.63 ± 0.01 | 0.61 ± 0.01 | 108.7 ± 0.4 | ⋯ | Hirano et al. (2018a) | ||
201132684 | 4.45 ± 0.02 | 0.07 ± 0.07 | 0.95 ± 0.02 | 198.2 ± 2.1 | ⋯ | Mayo et al. (2018) | ||
201164625 | 3.66 ± 0.03 | 11.50 ± 0.46 | This work | |||||
201166680 | 1.39 ± 0.03 | ⋯ | ⋯ | |||||
201180665 | 1.28 ± 0.03 | ⋯ | ⋯ | |||||
201211526 | 4.44 ± 0.02 | −0.29 ± 0.03 | 0.86 ± 0.02 | 3.69 ± 0.24 | This work | |||
201225286 | 5425 ± 44 | −0.07 ± 0.06 | 0.83 ± 0.01 | ⋯ | Mayo et al. (2018) | |||
201274010 | 0.88 ± 0.02 | ⋯ | ⋯ | |||||
201352100 | 4.60 ± 0.01 | −0.08 ± 0.04 | 0.76 ± 0.01 | 203.5 ± 3.3 | 2.15 ± 0.30 | This work | ||
201357643 | 4.16 ± 0.02 | −0.45 ± 0.02 | 1.25 ± 0.03 | 4.62 ± 0.24 | This work | |||
201386739 | −0.19 ± 0.03 | 0.87 ± 0.02 | 0.94 ± 0.03 | 2.90 ± 0.30 | This work | |||
201390048 | 4.63 ± 0.01 | 0.75 ± 0.02 | 0.69 ± 0.01 | 125.1 ± 0.6 | ⋯ | Mayo et al. (2018) | ||
201390927 | 0.97 ± 0.09 | ⋯ | ⋯ | |||||
201392505 | −0.16 ± 0.06 | 0.75 ± 0.01 | 2.46 ± 0.29 | This work | ||||
201437844 | 4.25 ± 0.02 | −0.22 ± 0.07 | 1.29 ± 0.02 | 109.7 ± 0.7 | 12.90 ± 0.40 | Rodriguez et al. (2017) | ||
201595106 | 5823 ± 19 | −0.00 ± 0.03 | 3.62 ± 0.18 | This work | ||||
201598502 | 3845 ± 37 | 4.73 ± 0.01 | −0.09 ± 0.08 | 0.55 ± 0.01 | 0.53 ± 0.01 | 143.6 ± 0.9 | ⋯ | Hirano et al. (2018a) |
201615463 | 4.09 ± 0.02 | 0.09 ± 0.06 | 1.61 ± 0.04 | ⋯ | Mayo et al. (2018) | |||
228707509 | 1.00 ± 0.04 | ⋯ | ⋯ | |||||
228720681 | −0.27 ± 0.04 | 0.88 ± 0.03 | 12.00 ± 0.60 | This work | ||||
228721452 | 4.48 ± 0.01 | 0.12 ± 0.06 | 1.06 ± 0.02 | 0.99 ± 0.01 | 201.1 ± 2.5 | ⋯ | Mayo et al. (2018) | |
228724899 | 5533 ± 52 | 4.44 ± 0.02 | 0.17 ± 0.04 | 0.97 ± 0.02 | 2.90 ± 0.30 | This work | ||
228725791 | −0.08 ± 0.22 | 0.74 ± 0.03 | 260.2 ± 2.1 | ⋯ | ⋯ | |||
228725972 | −0.19 ± 0.06 | 0.91 ± 0.02 | 0.84 ± 0.01 | 277.7 ± 2.9 | ⋯ | Mayo et al. (2018) | ||
228729473 | 3.32 ± 0.04 | −0.05 ± 0.02 | 3.46 ± 0.27 | This work | ||||
228732031 | 4.61 ± 0.01 | −0.17 ± 0.03 | 0.84 ± 0.01 | 0.75 ± 0.01 | 4.30 ± 0.20 | This work | ||
228734900 | 4.08 ± 0.02 | 0.38 ± 0.06 | 1.70 ± 0.04 | ⋯ | Mayo et al. (2018) | |||
228735255 | 4.45 ± 0.01 | 0.13 ± 0.04 | 1.00 ± 0.02 | 0.99 ± 0.01 | 3.80 ± 0.20 | Giles et al. 2017 | ||
228736155 | 4.47 ± 0.03 | −0.03 ± 0.07 | 0.88 ± 0.03 | 0.91 ± 0.02 | ⋯ | Mayo et al. (2018) | ||
228739306 | 4.45 ± 0.03 | −0.10 ± 0.04 | 0.92 ± 0.02 | 410.5 ± 4.9 | 2.60 ± 0.20 | This work | ||
228748383 | 1.57 ± 0.06 | ⋯ | ⋯ | |||||
228748826 | 4.53 ± 0.03 | −0.09 ± 0.04 | 0.81 ± 0.02 | 2.20 ± 0.30 | This work | |||
228753871 | 0.77 ± 0.01 | 295.9 ± 2.0 | ⋯ | ⋯ | ||||
228758778 | 4.76 ± 0.01 | 0.52 ± 0.01 | 0.49 ± 0.01 | 147.7 ± 1.0 | ⋯ | ⋯ | ||
228758948 | 0.11 ± 0.03 | 1.09 ± 0.02 | 1.03 ± 0.02 | 4.40 ± 0.20 | This work | |||
228763938 | −0.09 ± 0.04 | 0.82 ± 0.01 | 2.20 ± 0.20 | This work | ||||
228784812 | ⋯ | ⋯ | ||||||
228798746 | 4.66 ± 0.01 | −0.25 ± 0.06 | 0.71 ± 0.02 | 0.65 ± 0.01 | 142.4 ± 3.2 | ⋯ | Mayo et al. (2018) | |
228801451 | −0.09 ± 0.02 | 0.87 ± 0.01 | 0.79 ± 0.01 | 103.8 ± 1.0 | 2.46 ± 0.22 | This work | ||
228804845 | 4.20 ± 0.02 | 0.10 ± 0.04 | 1.11 ± 0.02 | 1.39 ± 0.04 | 5.20 ± 0.20 | This work | ||
228809391 | −0.01 ± 0.03 | 0.97 ± 0.03 | 0.93 ± 0.02 | 3.20 ± 0.20 | This work | |||
228809550 | 1.10 ± 0.04 | ⋯ | ⋯ | |||||
228834632 | 4.67 ± 0.01 | 0.65 ± 0.02 | 0.62 ± 0.01 | 285.3 ± 3.0 | ⋯ | ⋯ | ||
228836835 | 4.83 ± 0.01 | 0.45 ± 0.01 | 0.43 ± 0.01 | ⋯ | ⋯ | |||
228846243 | ⋯ | ⋯ | ||||||
228849382 | 4.64 ± 0.01 | 0.67 ± 0.01 | 229.4 ± 1.3 | ⋯ | ⋯ | |||
228888935 | 4.01 ± 0.05 | 1.45 ± 0.11 | 1.97 ± 0.09 | ⋯ | ⋯ | |||
228894622 | 4.62 ± 0.02 | 0.71 ± 0.03 | 0.69 ± 0.01 | 192.4 ± 1.5 | 2.46 ± 0.27 | This work | ||
228934525 | 4.65 ± 0.01 | 0.63 ± 0.01 | 129.8 ± 0.4 | ⋯ | Hirano et al. (2018a) | |||
228964773 | −0.18 ± 0.03 | 0.89 ± 0.03 | 3.40 ± 0.20 | This work | ||||
228968232 | −0.10 ± 0.17 | 0.78 ± 0.02 | ⋯ | ⋯ | ||||
228974324 | 4.76 ± 0.01 | 0.52 ± 0.01 | 0.50 ± 0.01 | 64.1 ± 0.3 | ⋯ | ⋯ | ||
228974907 | 3.86 ± 0.03 | ⋯ | ⋯ | |||||
229004835 | 4.40 ± 0.02 | −0.22 ± 0.01 | 0.92 ± 0.01 | 1.00 ± 0.01 | 3.77 ± 0.12 | This work | ||
229017395 | 1.31 ± 0.03 | ⋯ | ⋯ | |||||
229103251 | ⋯ | ⋯ | ||||||
229131722 | 0.16 ± 0.04 | 1.14 ± 0.03 | 5.23 ± 0.20 | This work | ||||
229133720 | 0.72 ± 0.01 | 105.2 ± 0.5 | ⋯ | ⋯ |
Note. "Provenance" indicates the source of the spectroscopic parameters used as priors in our analysis (see Section 4.4). The uncertainties are internal to the Kea pipeline and do not account for other types of line broadening; thus they are likely to be underestimated.
Table 7. Individual False Positive Scenario Likelihoods Computed by vespa
EPIC | L_beba | L_beb_P × 2a | L_ebb | L_eb_P × 2b | L_hebc | L_heb_P × 2c | L_pld | FPP |
---|---|---|---|---|---|---|---|---|
201092629.01 | 0 | 0 | ||||||
201102594.01 | 0 | 0 | ||||||
201110617.01 | 0 | 0 | ||||||
201111557.01 | 0 | 0 | ||||||
201127519.01 | 0 | 0 | ||||||
201128338.01 | 0 | 0 | ||||||
201132684.01 | 0 | 0 | ||||||
201132684.02 | 0 | 0 | ||||||
201164625.01 | 0 | 0 | ||||||
201166680.02 | ||||||||
201166680.03 | ||||||||
201180665.01 | 0 | 0 | ||||||
201211526.01 | 0 | 0 | ||||||
201225286.01 | 0 | 0 | ||||||
201274010.01 | ||||||||
201352100.01 | 0 | 0 | ||||||
201357643.01 | 0 | 0 | ||||||
201386739.01 | 0 | 0 | ||||||
201390048.01 | ||||||||
201390927.01 | 0 | 0 | ||||||
201392505.01 | 0 | 0 | ||||||
201437844.01 | 0 | 0 | ||||||
201437844.02 | 0 | 0 | ||||||
201595106.01 | 0 | 0 | ||||||
201598502.01 | 0 | 0 | ||||||
201615463.01 | ||||||||
228707509.01 | 0 | 0 | ||||||
228720681.01 | 0 | 0 | ||||||
228721452.01 | 0 | 0 | ||||||
228721452.02 | 0 | 0 | ||||||
228724899.01 | 0 | 0 | ||||||
228725791.01 | 0 | 0 | ||||||
228725791.02 | 0 | 0 | ||||||
228725972.01 | 0 | 0 | ||||||
228725972.02 | 0 | 0 | ||||||
228729473.01 | 0 | 0 | ||||||
228732031.01 | 0 | 0 | ||||||
228734900.01 | ||||||||
228735255.01 | 0 | 0 | ||||||
228736155.01 | 0 | 0 | ||||||
228739306.01 | 0 | 0 | ||||||
228748383.01 | 0 | 0 | ||||||
228748826.01 | 0 | 0 | ||||||
228753871.01 | ||||||||
228758778.01 | 0 | 0 | ||||||
228758948.01 | 0 | 0 | ||||||
228763938.01 | 0 | 0 | ||||||
228784812.01 | ||||||||
228798746.01 | ||||||||
228801451.01 | 0 | 0 | ||||||
228801451.02 | 0 | 0 | ||||||
228804845.01 | 0 | 0 | ||||||
228809391.01 | 0 | 0 | ||||||
228809550.01 | 0 | 0 | ||||||
228834632.01 | ||||||||
228836835.01 | ||||||||
228846243.01 | 0 | 0 | ||||||
228849382.01 | 0 | 0 | ||||||
228849382.02 | 0 | 0 | ||||||
228888935.01 | 0 | 0 | ||||||
228894622.01 | 0 | 0 | ||||||
228934525.01 | 0 | 0 | ||||||
228934525.02 | 0 | 0 | ||||||
228964773.01 | 0 | 0 | ||||||
228968232.01 | 0 | 0 | ||||||
228974324.01 | 0 | 0 | ||||||
228974907.01 | 0 | 0 | ||||||
229004835.01 | 0 | 0 | ||||||
229017395.01 | 0 | 0 | ||||||
229103251.01 | 0 | 0 | ||||||
229131722.01 | 0 | 0 | ||||||
229133720.01 | 0 | 0 |
Notes.
aLikelihood that the signal is due to a background eclipsing binary, at the measured period or twice that. bLikelihood that the signal is due to an eclipsing binary, at the measured period or twice that. cLikelihood that the signal is due to a hierarchical star system with an eclipsing component, at the measured period or twice that. dLikelihood that the signal is due to a planet.Table 8. Additional Constraints to vespa
EPIC | Maxrad (arcsec) | Secthresh |
---|---|---|
201092629.01 | 20.7 | |
201102594.01 | 15.1 | |
201110617.01 | 15.9 | |
201111557.01 | 17.5 | |
201127519.01 | 10.3 | |
201128338.01 | 18.3 | |
201132684.01 | 21.5 | |
201132684.02 | 21.5 | |
201164625.01 | 19.1 | |
201166680.02 | 22.3 | |
201166680.03 | 22.3 | |
201180665.01 | 8.8 | |
201211526.01 | 13.5 | |
201225286.01 | 12.7 | |
201274010.01 | 16.7 | |
201352100.01 | 10.3 | |
201357643.01 | 6.4 | |
201386739.01 | 16.7 | |
201390048.01 | 10.3 | |
201390927.01 | 14.3 | |
201392505.01 | 8.0 | |
201437844.01 | 31.8 | |
201437844.02 | 31.8 | |
201595106.01 | 17.5 | |
201598502.01 | 15.1 | |
201615463.01 | 16.7 | |
228707509.01 | 14.3 | |
228720681.01 | 10.3 | |
228721452.01 | 14.3 | |
228721452.02 | 14.3 | |
228724899.01 | 11.1 | |
228725791.01 | 11.9 | |
228725791.02 | 11.9 | |
228725972.01 | 14.3 | |
228725972.02 | 14.3 | |
228729473.01 | 19.1 | |
228732031.01 | 21.5 | |
228734900.01 | 15.9 | |
228735255.01 | 16.7 | |
228736155.01 | 15.1 | |
228739306.01 | 15.1 | |
228748383.01 | 15.9 | |
228748826.01 | 15.1 | |
228753871.01 | 13.5 | |
228758778.01 | 12.7 | |
228758948.01 | 17.5 | |
228763938.01 | 14.3 | |
228784812.01 | 8.8 | |
228798746.01 | 9.6 | |
228801451.01 | 18.3 | |
228801451.02 | 18.3 | |
228804845.01 | 19.1 | |
228809391.01 | 9.6 | |
228809550.01 | 11.1 | |
228834632.01 | 11.9 | |
228836835.01 | 8.0 | |
228846243.01 | 9.6 | |
228849382.01 | 8.0 | |
228849382.02 | 8.0 | |
228888935.01 | 8.8 | |
228894622.01 | 12.7 | |
228934525.01 | 9.6 | |
228934525.02 | 9.6 | |
228964773.01 | 7.2 | |
228968232.01 | 10.3 | |
228974324.01 | 11.9 | |
228974907.01 | 32.6 | |
229004835.01 | 11.1 | |
229017395.01 | 15.9 | |
229103251.01 | 16.7 | |
229131722.01 | 12.7 | |
229133720.01 | 18.3 |
Notes. The columns "maxrad" and "secthresh" refer to the maximum radius (the angular size of the photometric aperture) and the secondary eclipse threshold (the maximum secondary eclipse depth allowed by the light curve), respectively.
Table 9. WIYN/NESSI Data Sets Used in This Work
EPIC | Filter Center | Filter Width | Obs. Date |
---|---|---|---|
(nm) | (nm) | ||
201092629 | 562 nm | 44 nm | 2017 May 15 |
201092629 | 832 nm | 40 nm | 2017 May 15 |
201092629 | 562 nm | 44 nm | 2017 Mar 18 |
201092629 | 832 nm | 40 nm | 2017 Mar 18 |
201102594 | 562 nm | 44 nm | 2017 Apr 05 |
201102594 | 832 nm | 40 nm | 2017 Apr 05 |
201110617 | 832 nm | 40 nm | 2017 Mar 10 |
201110617 | 562 nm | 44 nm | 2017 Mar 10 |
201111557 | 562 nm | 44 nm | 2017 Mar 15 |
201111557 | 832 nm | 40 nm | 2017 Mar 15 |
201127519 | 562 nm | 44 nm | 2017 Mar 11 |
201127519 | 832 nm | 40 nm | 2017 Mar 11 |
201128338 | 832 nm | 40 nm | 2017 Mar 10 |
201128338 | 562 nm | 44 nm | 2017 Mar 10 |
201132684 | 832 nm | 40 nm | 2017 May 12 |
201132684 | 562 nm | 44 nm | 2017 May 12 |
201132684 | 562 nm | 44 nm | 2017 Mar 15 |
201132684 | 832 nm | 40 nm | 2017 Mar 15 |
201164625 | 562 nm | 44 nm | 2017 Mar 18 |
201164625 | 832 nm | 40 nm | 2017 Mar 18 |
201164625 | 562 nm | 44 nm | 2017 May 12 |
201164625 | 832 nm | 40 nm | 2017 May 12 |
201180665 | 832 nm | 40 nm | 2017 Mar 18 |
201180665 | 562 nm | 44 nm | 2017 Mar 18 |
201211526 | 832 nm | 40 nm | 2017 Mar 18 |
201211526 | 562 nm | 44 nm | 2017 Mar 18 |
201225286 | 562 nm | 44 nm | 2017 Apr 03 |
201225286 | 832 nm | 40 nm | 2017 Apr 03 |
201352100 | 562 nm | 44 nm | 2017 Mar 15 |
201352100 | 832 nm | 40 nm | 2017 Mar 15 |
201357643 | 562 nm | 44 nm | 2017 Mar 18 |
201357643 | 832 nm | 40 nm | 2017 Mar 18 |
201386739 | 562 nm | 44 nm | 2017 Mar 17 |
201386739 | 832 nm | 40 nm | 2017 Mar 17 |
201390927 | 832 nm | 40 nm | 2017 Mar 17 |
201390927 | 562 nm | 44 nm | 2017 Mar 17 |
201392505 | 832 nm | 40 nm | 2017 Mar 18 |
201392505 | 562 nm | 44 nm | 2017 Mar 18 |
201437844 | 562 nm | 44 nm | 2017 Mar 11 |
201437844 | 832 nm | 40 nm | 2017 Mar 11 |
201595106 | 832 nm | 40 nm | 2017 Mar 18 |
201595106 | 562 nm | 44 nm | 2017 Mar 18 |
201598502 | 832 nm | 40 nm | 2017 Mar 18 |
201598502 | 562 nm | 44 nm | 2017 Mar 18 |
228707509 | 562 nm | 44 nm | 2017 Apr 08 |
228707509 | 832 nm | 40 nm | 2017 Apr 08 |
228720681 | 832 nm | 40 nm | 2017 Mar 14 |
228720681 | 562 nm | 44 nm | 2017 Mar 14 |
228721452 | 562 nm | 44 nm | 2017 Mar 11 |
228721452 | 832 nm | 40 nm | 2017 Mar 11 |
228724899 | 562 nm | 44 nm | 2017 Mar 14 |
228724899 | 832 nm | 40 nm | 2017 Mar 14 |
228725791 | 562 nm | 44 nm | 2017 Mar 17 |
228725791 | 832 nm | 40 nm | 2017 Mar 17 |
228725972 | 832 nm | 40 nm | 2017 Mar 17 |
228725972 | 562 nm | 44 nm | 2017 Mar 17 |
228729473 | 832 nm | 40 nm | 2017 Apr 03 |
228729473 | 832 nm | 40 nm | 2017 May 19 |
228729473 | 562 nm | 44 nm | 2017 Apr 03 |
228729473 | 562 nm | 44 nm | 2017 May 19 |
228732031 | 832 nm | 40 nm | 2017 Apr 05 |
228732031 | 562 nm | 44 nm | 2017 Apr 05 |
228735255 | 832 nm | 40 nm | 2017 Mar 10 |
228735255 | 562 nm | 44 nm | 2017 Mar 10 |
228736155 | 562 nm | 44 nm | 2017 Apr 05 |
228736155 | 832 nm | 40 nm | 2017 Apr 05 |
228739306 | 562 nm | 44 nm | 2017 Mar 09 |
228739306 | 832 nm | 40 nm | 2017 Mar 09 |
228748383 | 832 nm | 40 nm | 2017 Mar 18 |
228748383 | 562 nm | 44 nm | 2017 May 19 |
228748383 | 832 nm | 40 nm | 2017 May 19 |
228748383 | 562 nm | 44 nm | 2017 Mar 18 |
228748826 | 562 nm | 44 nm | 2017 Mar 09 |
228748826 | 832 nm | 40 nm | 2017 Mar 09 |
228758778 | 562 nm | 44 nm | 2017 Apr 08 |
228758778 | 832 nm | 40 nm | 2017 Apr 08 |
228758948 | 832 nm | 40 nm | 2017 Mar 10 |
228758948 | 562 nm | 44 nm | 2017 Mar 10 |
228763938 | 562 nm | 44 nm | 2017 May 19 |
228763938 | 832 nm | 40 nm | 2017 May 19 |
228763938 | 562 nm | 44 nm | 2017 Mar 18 |
228763938 | 832 nm | 40 nm | 2017 Mar 18 |
228801451 | 832 nm | 40 nm | 2017 Mar 11 |
228801451 | 562 nm | 44 nm | 2017 Mar 11 |
228804845 | 562 nm | 44 nm | 2017 Mar 10 |
228804845 | 832 nm | 40 nm | 2017 Mar 10 |
228809391 | 562 nm | 44 nm | 2017 Mar 10 |
228809391 | 832 nm | 40 nm | 2017 Mar 10 |
228809550 | 832 nm | 40 nm | 2017 Mar 18 |
228809550 | 562 nm | 44 nm | 2017 Mar 18 |
228846243 | 832 nm | 40 nm | 2017 Mar 17 |
228846243 | 562 nm | 44 nm | 2017 Mar 17 |
228849382 | 832 nm | 40 nm | 2017 May 20 |
228849382 | 562 nm | 44 nm | 2017 May 20 |
228888935 | 832 nm | 40 nm | 2017 Mar 17 |
228888935 | 562 nm | 44 nm | 2017 Mar 17 |
228894622 | 832 nm | 40 nm | 2017 Mar 09 |
228894622 | 562 nm | 44 nm | 2017 Mar 09 |
228934525 | 562 nm | 44 nm | 2017 Mar 09 |
228934525 | 832 nm | 40 nm | 2017 Mar 09 |
228964773 | 562 nm | 44 nm | 2017 Mar 18 |
228964773 | 832 nm | 40 nm | 2017 Mar 18 |
228968232 | 832 nm | 40 nm | 2017 Mar 18 |
228968232 | 562 nm | 44 nm | 2017 Mar 18 |
228974324 | 832 nm | 40 nm | 2017 Mar 10 |
228974324 | 562 nm | 44 nm | 2017 Mar 10 |
228974907 | 562 nm | 44 nm | 2017 Mar 18 |
228974907 | 832 nm | 40 nm | 2017 Mar 18 |
229004835 | 562 nm | 44 nm | 2017 Mar 11 |
229004835 | 832 nm | 40 nm | 2017 Mar 11 |
229017395 | 832 nm | 40 nm | 2017 Mar 18 |
229017395 | 562 nm | 44 nm | 2017 Mar 18 |
229103251 | 832 nm | 40 nm | 2017 Mar 09 |
229103251 | 562 nm | 44 nm | 2017 Mar 09 |
229131722 | 832 nm | 40 nm | 2017 May 19 |
229131722 | 832 nm | 40 nm | 2017 Mar 10 |
229131722 | 562 nm | 44 nm | 2017 May 19 |
229131722 | 562 nm | 44 nm | 2017 Mar 10 |
229133720 | 562 nm | 44 nm | 2017 Mar 11 |
229133720 | 832 nm | 40 nm | 2017 Mar 11 |
Table 10. TNG/HARPS-N Results
EPIC | Tobs | RV | BIS | FWHM | log(RHK) | B − V | Texp | S/N |
---|---|---|---|---|---|---|---|---|
(BJDTDB) | (km s−1) | (km s−1) | (km s−1) | (mag) | (s) | (5500 nm) | ||
228801451 | 2457782.629699 | 22.960809 ± 0.001844 | −0.012789 | 7.175241 | −4.5707 ± 0.0098 | 0.873 | 1800.0 | 48.8 |
201595106 | 2457782.687224 | 0.692781 ± 0.002263 | −0.022588 | 6.965865 | −4.9714 ± 0.0273 | 0.703 | 2400.0 | 45.0 |
201437844 | 2457762.701586 | −3.449696 ± 0.005740 | 0.037015 | 20.649605 | −4.8647 ± 0.0058 | 0.451 | 1200.0 | 101.6 |
201437844 | 2457774.738143 | −3.441043 ± 0.005931 | 0.045533 | 20.699375 | −4.8584 ± 0.0060 | 0.451 | 1800.0 | 98.7 |
201437844 | 2457774.759707 | −3.441611 ± 0.006562 | 0.073457 | 20.632207 | −4.8629 ± 0.0071 | 0.451 | 1800.0 | 90.0 |
Download table as: ASCIITypeset image
Another possibly interesting RV target is K2-257 b, a sub-Earth-size planet orbiting a nearby M dwarf. Although the planet's radius is only , the Doppler semi-amplitude could be as high as ∼1 m s−1 due to the low mass of the host star and the planet's short orbital period. The host star is moderately bright (Kp = 12.873, J = 10.477 mag), so this presents an opportunity to directly measure the mass of a sub-Earth with one of today's high-precision optical or NIR spectrographs. Such a measurement would yield the planet's density and constrain its composition, as well as improve our knowledge of the mass–radius relation for small planets. The only other sub-Earth-size planet known to transit a similarly bright M dwarf is Kepler-138 b, for which the mass has been measured only via TTVs (Jontof-Hutter et al. 2015; Almenara et al. 2018).
6.4.2. Atmospheric Targets
In order to identify viable new targets for atmospheric studies via transmission spectroscopy, we used the properties of the host stars and planets to predict atmospheric scale heights and the amplitudes of the wavelength dependence of transit depth (δTS). Following Miller-Ricci et al. (2009), we calculated the atmospheric scale height H and δTS for each validated planet by
where μ, Teq, and g are the mean molecular weight, planet equilibrium temperature, and planet surface gravity, respectively. We used the predicted planet mass estimated in Section 6.4 to predict the surface gravity, and assumed a bond albedo of 0.3 and a mean molecular weight μ = 2 (hydrogen-dominated atmosphere) for each planet (see Table 11). We note that this assumption for μ is likely to be invalid for the smaller planets in our sample (i.e., ≲ 1.5–2 ), as they are not likely to have substantial hydrogen-dominated atmospheres; these smaller planets likely have higher mean molecular weight atmospheres, which would make their characterization via transmission spectroscopy more challenging. The validated planets K2-140 b and K2-255 b both orbit relatively bright host stars ( mag) and have large expected transmission spectroscopy signals ( ppm), and thus could be interesting targets for future atmospheric characterization.
Table 11. Predicted Atmospheric Characteristics, where g is Surface Gravity, H is Atmospheric Scale Height, and δTS is the Expected Amplitude of Atmospheric Spectral Features
EPIC | g | H | |
---|---|---|---|
(g⊕) | (km) | (ppm) | |
201092629.01 | 1.38 | 156 | 94 |
201102594.01 | 1.38 | 128 | 318 |
201110617.01 | 1.92 | 298 | 107 |
201111557.01 | 1.80 | 246 | 73 |
201127519.01 | 0.57 | 575 | 1146 |
201128338.01 | 1.49 | 96 | 78 |
201132684.01 | 1.34 | 252 | 98 |
201132684.02 | 2.08 | 194 | 37 |
201164625.01 | 0.94 | 1050 | 55 |
201166680.02 | 1.54 | 285 | 43 |
201166680.03 | 1.62 | 209 | 29 |
201180665.01 | 0.26 | 1359 | 2885 |
201211526.01 | 1.80 | 150 | 41 |
201225286.01 | 1.50 | 193 | 85 |
201274010.01 | 1.41 | 214 | 90 |
201352100.01 | 1.32 | 195 | 122 |
201357643.01 | 0.94 | 414 | 152 |
201386739.01 | 1.10 | 377 | 197 |
201390048.01 | 2.09 | 128 | 51 |
201390927.01 | 1.25 | 447 | 169 |
201392505.01 | 1.13 | 180 | 143 |
201437844.01 | 1.46 | 304 | 56 |
201437844.02 | 0.95 | 357 | 122 |
201595106.01 | 2.00 | 396 | 68 |
201598502.01 | 1.63 | 129 | 124 |
201615463.01 | 1.54 | 313 | 35 |
228707509.01 | 0.37 | 854 | 1850 |
228720681.01 | 0.49 | 641 | 890 |
228721452.01 | 1.08 | 887 | 108 |
228721452.02 | 1.97 | 235 | 50 |
228724899.01 | 1.08 | 393 | 197 |
228725791.01 | 1.57 | 265 | 158 |
228725791.02 | 1.54 | 190 | 117 |
228725972.01 | 2.00 | 213 | 62 |
228725972.02 | 1.44 | 225 | 101 |
228729473.01 | 0.33 | 1524 | 248 |
228732031.01 | 1.84 | 476 | 190 |
228734900.01 | 1.10 | 348 | 56 |
228735255.01 | 0.45 | 895 | 1463 |
228736155.01 | 2.01 | 236 | 59 |
228739306.01 | 1.37 | 275 | 109 |
228748383.01 | 1.29 | 348 | 52 |
228748826.01 | 1.41 | 287 | 141 |
228753871.01 | 1.46 | 165 | 86 |
228758778.01 | 1.46 | 127 | 162 |
228758948.01 | 0.97 | 354 | 184 |
228763938.01 | 1.81 | 150 | 51 |
228784812.01 | 2.02 | 240 | 43 |
228798746.01 | 2.12 | 183 | 70 |
228801451.01 | 2.00 | 384 | 95 |
228801451.02 | 1.61 | 197 | 86 |
228804845.01 | 1.69 | 383 | 52 |
228809391.01 | 1.29 | 212 | 89 |
228809550.01 | 0.44 | 1192 | 1642 |
228834632.01 | 1.76 | 124 | 78 |
228836835.01 | 1.99 | 199 | 188 |
228846243.01 | 0.59 | 655 | 202 |
228849382.01 | 1.91 | 176 | 85 |
228849382.02 | 1.54 | 152 | 97 |
228888935.01 | 0.33 | 1918 | 1219 |
228894622.01 | 1.24 | 354 | 288 |
228934525.01 | 1.64 | 185 | 122 |
228934525.02 | 1.60 | 146 | 100 |
228964773.01 | 0.84 | 251 | 227 |
228968232.01 | 1.35 | 267 | 152 |
228974324.01 | 0.97 | 342 | 154 |
228974907.01 | 1.21 | 470 | 26 |
229004835.01 | 1.64 | 191 | 51 |
229017395.01 | 1.22 | 288 | 66 |
229103251.01 | 1.02 | 394 | 127 |
229131722.01 | 1.47 | 229 | 55 |
229133720.01 | 1.51 | 244 | 140 |
Table 12. Comparison of Parameters between K2 Pipelines
EPIC | Porb | Δ | b | a | ||||
---|---|---|---|---|---|---|---|---|
(days) | (σ) | () | (σ) | (σ) | () | (σ) | ||
201092629.01 | 3.7 | 3.0 | 0.4 | 0.7 | ||||
201102594.01 | 0.0 | 0.3 | 0.3 | 0.1 | ||||
201110617.01 | 0.5 | 0.1 | 0.0 | 0.4 | ||||
201111557.01 | 0.8 | 0.1 | 0.0 | 0.1 | ||||
201127519.01 | 0.6 | 1.1 | 0.3 | 0.6 | ||||
201128338.01 | 0.6 | 1.3 | 0.1 | 0.2 | ||||
201132684.01 | 1.3 | 0.9 | 0.1 | 0.0 | ||||
201132684.02 | 1.5 | 0.7 | 0.3 | 0.6 | ||||
201164625.01 | 0.6 | 0.5 | 0.1 | 1.2 | ||||
201166680.02 | 0.4 | 0.8 | 0.1 | 0.1 | ||||
201166680.03 | 0.6 | 1.4 | 0.5 | 0.1 | ||||
201180665.01 | 1.1 | 0.4 | 0.8 | 0.5 | ||||
201211526.01 | 1.2 | 0.6 | 0.1 | 0.0 | ||||
201225286.01 | 1.0 | 0.0 | 0.1 | 0.3 | ||||
201274010.01 | 0.6 | 0.8 | 0.0 | 0.1 | ||||
201352100.01 | 0.1 | 0.9 | 0.1 | 0.3 | ||||
201357643.01 | 0.2 | 0.1 | 0.0 | 0.0 | ||||
201386739.01 | 0.8 | 1.1 | 0.1 | 0.0 | ||||
201390048.01 | 1.6 | 0.9 | 0.0 | 0.1 | ||||
201390927.01 | 0.0 | 0.9 | 0.0 | 0.1 | ||||
201392505.01 | 2.9 | 5.3 | 0.3 | 1.4 | ||||
201437844.01 | 2.4 | 1.8 | 0.5 | 0.7 | ||||
201437844.02 | 0.0 | 0.3 | 0.5 | 0.5 | ||||
201595106.01 | 1.2 | 1.2 | 0.0 | 0.3 | ||||
201598502.01 | 0.5 | 0.9 | 0.1 | 0.2 | ||||
201615463.01 | 0.2 | 1.1 | 0.0 | 0.1 | ||||
228707509.01 | 3.7 | 2.8 | 0.8 | 0.9 | ||||
228720681.01 | 0.3 | 0.9 | 0.6 | 0.6 | ||||
228721452.01 | 1.0 | 0.6 | 0.9 | 1.0 | ||||
228721452.02 | 2.1 | 3.0 | 0.3 | 0.8 | ||||
228724899.01 | 0.0 | 0.3 | 0.2 | 0.7 | ||||
228725791.01 | 0.6 | 0.8 | 0.1 | 0.0 | ||||
228725791.02 | 0.5 | 0.6 | 0.3 | 0.2 | ||||
228725972.01 | 0.2 | 0.7 | 0.5 | 0.2 | ||||
228725972.02 | 0.8 | 0.2 | 0.4 | 0.3 | ||||
228729473.01 | 1.0 | 2.0 | 0.8 | 0.9 | ||||
228732031.01 | 1.2 | 0.6 | 0.1 | 0.6 | ||||
228734900.01 | 0.3 | 0.5 | 0.0 | 0.3 | ||||
228735255.01 | 0.3 | 0.4 | 0.5 | 0.8 | ||||
228736155.01 | 0.4 | 0.1 | 0.0 | 0.3 | ||||
228739306.01 | 0.0 | 0.9 | 0.1 | 0.0 | ||||
228748383.01 | 1.5 | 1.0 | 0.0 | 0.1 | ||||
228748826.01 | 0.3 | 1.0 | 0.2 | 0.1 | ||||
228753871.01 | 0.5 | 0.4 | 0.1 | 0.2 | ||||
228758778.01 | 1.1 | 0.7 | 0.0 | 0.1 | ||||
228758948.01 | 0.4 | 0.6 | 0.0 | 0.1 | ||||
228763938.01 | 0.1 | 0.3 | 0.0 | 0.2 | ||||
228784812.01 | 0.4 | 0.1 | 0.3 | 0.2 | ||||
228798746.01 | 0.3 | 0.4 | 0.0 | 0.0 | ||||
228801451.01 | 0.2 | 0.7 | 0.2 | 0.9 | ||||
228801451.02 | 3.0 | 4.8 | 0.9 | 0.4 | ||||
228804845.01 | 0.3 | 1.6 | 0.0 | 0.0 | ||||
228809391.01 | 0.7 | 0.4 | 0.1 | 0.0 | ||||
228809550.01 | 1.1 | 0.7 | 0.2 | 0.3 | ||||
228834632.01 | 0.9 | 2.2 | 0.5 | 0.4 | ||||
228836835.01 | 0.3 | 0.0 | 0.2 | 0.7 | ||||
228846243.01 | 1.7 | 0.7 | 0.2 | 0.0 | ||||
228849382.01 | 1.0 | 0.9 | 1.2 | 0.0 | ||||
228849382.02 | 0.1 | 0.8 | 0.5 | 0.2 | ||||
228888935.01 | 1.1 | 0.2 | 0.8 | 0.7 | ||||
228894622.01 | 5.9 | 0.1 | 0.1 | 0.1 | ||||
228934525.01 | 0.6 | 1.3 | 0.3 | 0.1 | ||||
228934525.02 | 0.2 | 0.4 | 0.4 | 0.2 | ||||
228964773.01 | 2.5 | 1.2 | 1.2 | 0.1 | ||||
228968232.01 | 1.2 | 3.5 | 0.0 | 0.7 | ||||
228974324.01 | 0.1 | 0.3 | 0.1 | 0.0 | ||||
228974907.01 | 2.6 | 2.7 | 0.0 | 0.2 | ||||
229004835.01 | 2.1 | 0.4 | 0.0 | 0.0 | ||||
229017395.01 | 0.2 | 0.5 | 0.1 | 0.0 | ||||
229103251.01 | 2.0 | 1.3 | 0.0 | 0.0 | ||||
229131722.01 | 0.9 | 0.9 | 0.0 | 0.0 | ||||
229133720.01 | 0.4 | 0.0 | 0.0 | 0.1 |
7. Summary
We detected 72 planet candidates in K2 Campaign 10 and obtained high-resolution imaging and spectroscopy follow-up observations to characterize the host stars. We performed detailed modeling of the light curves and used the resulting transit parameters to compute physical planet properties. We used the planet and host star properties to predict masses and atmospheric signals, which enabled us to identify good targets for future characterization via Doppler and transmission spectroscopy. We statistically validated 44 planets, leaving a remainder of 27 candidates and one false positive. We expect nearly all of these remaining candidates to be real planets, which could potentially be validated via further observations and analysis.
This work was carried out as part of the KESPRINT consortium. The WIYN/NESSI observations were conducted as part of an approved NOAO observing program (PI Livingston, proposal ID 2017A-0377). Data presented herein were obtained at the WIYN Observatory from telescope time allocated to NN-EXPLORE through the scientific partnership of the National Aeronautics and Space Administration, the National Science Foundation, and the National Optical Astronomy Observatory. This work was supported by a NASA WIYN PI Data Award, administered by the NASA Exoplanet Science Institute. NESSI was funded by the NASA Exoplanet Exploration Program and the NASA Ames Research Center. NESSI was built at the Ames Research Center by Steve B Howell, Nic Scott, Elliott P Horch, and Emmett Quigley. The authors are honored to be permitted to conduct observations on Iolkam Du'ag (Kitt Peak), a mountain within the Tohono O'odham Nation with particular significance to the Tohono O'odham people. J.H.L. gratefully acknowledges the support of the Japan Society for the Promotion of Science (JSPS) Research Fellowship for Young Scientists. This work was supported by Japan Society for Promotion of Science (JSPS) KAKENHI Grant Number JP16K17660. M.E. and W.D.C. were supported by NASA grant NNX16AJ11G to The University of Texas. This paper includes data collected by the Kepler mission. Funding for the Kepler mission is provided by the NASA Science Mission directorate.
Facilities: Kepler - The Kepler Mission, WIYN (NESSI) - Wisconsin-Indiana-Yale-NOAO Telescope, McDonald (Tull) - , NOT (FIES) - Nordic Optical Telescope, TNG (HARPS-N) - Telescopio Nazionale Galileo.
Software: scipy, emcee, batman, vespa, IRAF, pyaneti, exotrending.
Footnotes
- 23
- 24
- 25
Data release 1, dated 2016 December 19, available at http://ps1images.stsci.edu/cgi-bin/ps1cutouts.
- 26
- 27