KELT-12b: A Day, Highly Inflated Hot Jupiter Transiting a Mildly Evolved Hot Star
We report the discovery of KELT-12b, a highly inflated Jupiter-mass planet transiting a mildly evolved host star. We identified the initial transit signal in the KELT-North survey data and established the planetary nature of the companion through precise follow-up photometry, high-resolution spectroscopy, precise radial velocity measurements, and high-resolution adaptive optics imaging. Our preferred best-fit model indicates that the host, TYC 2619-1057-1, has = K, = , and [Fe/H] = , with an inferred mass = and radius = . The planetary companion has = , = , = , and density = g cm, making it one of the most inflated giant planets known. The time of inferior conjunction in is and the period is days. Despite the relatively large separation of AU implied by its -day orbital period, KELT-12b receives significant flux of erg s cm from its host. We compare the radii and insolations of transiting gas-giant planets around hot ( K) and cool stars, noting that the observed paucity of known transiting giants around hot stars with low insolation is likely due to selection effects. We underscore the significance of long-term ground-based monitoring of hot stars and space-based targeting of hot stars with the Transiting Exoplanet Survey Satellite (TESS) to search for inflated gas giants in longer-period orbits.
The discovery of transiting exoplanets is generally partitioned into two regimes: giant planets on short-period orbits around bright stars and smaller planets around fainter stars. Ground-based transit surveys are most sensitive to the former due to design and selection biases (Pepper et al., 2003; Pepper & Gaudi, 2005; Gaudi, 2005; Beatty & Gaudi, 2008; Pont et al., 2006; Fressin et al., 2007), while space-based surveys such as CoRoT (Rouan et al., 1998) and Kepler (Borucki et al., 2010) specialize in the latter; the two-wheeled Kepler mission, K2, explores the intermediate regime (Howell et al., 2014).
In addition, the Transiting Exoplanet Survey Satellite (TESS; Ricker et al. 2015) will be sensitive to the same transiting systems to which the ground-based surveys are sensitive. However, many ground-based surveys – including HAT (Bakos et al., 2004), the Kilodegree Extremely Little Telescope (KELT; Pepper et al. 2007, 2012), and SuperWASP (Pollacco et al., 2006), and their Northern Hemisphere components specifically – have been taking continuous observations of the night sky for approximately one decade. Thus, in the era of TESS, it may be possible to combine TESS data with that from ground-based surveys to discover and characterize longer-period giant planets than those that can be found with TESS data alone.
The KELT survey consists of two similar telescopes – one in Sonoita, Arizona (KELT-North; Pepper et al. 2007) and the other in Sutherland, South Africa (KELT-South; Pepper et al. 2012) – which are primarily sensitive to 1 flux changes in stars of band brightness . KELT-North has found nine transiting substellar companions since starting in late 2006, while KELT-South has independently discovered four planets since starting operations in 2010, with a fourteenth planet found by both in an overlap survey field monitored by both telescopes (Zhou et al., 2016). KELT’s continued monitoring of the same fields throughout its lifetime increases its sensitivity to long-duration and longer-period ( days) systems such as KELT-6b, which orbits its host once every days (Collins et al., 2014).
Moreover, due to the KELT telescopes’ sensitivity to giant planets around bright stars (which tend to be hot), the survey has discovered a few inflated planets: these include the giant planets KELT-4Ab (Eastman et al., 2016), KELT-6b (Collins et al., 2014), KELT-8b (Fulton et al., 2015), and KELT-11b (Pepper2016), as well as the highly irradiated and massive brown dwarf, KELT-1b (Siverd et al., 2012). Such companions are ideal targets for atmospheric characterization (e.g. Beatty et al. 2014) due to both their large radii and the brightness of their hosts; most planets with studied atmospheres have (Sing et al., 2016; Seager & Deming, 2010). They also provide clues about which environmental parameters (such as incident flux; Demory & Seager 2011) may drive exoplanetary radius inflation.
In this paper, we present the discovery and characterization of KELT-12b, an inflated hot Jupiter on a long (by ground-based transit standards), 5-day orbit around the hot star TYC 2619-1057-1, which is towards the end of its main sequence lifetime. We place KELT-12b’s extremely inflated radius in context, discuss radius inflation in hot Jupiters, and investigate its connection to incident flux and host star temperature.
2 Discovery and Follow-Up Observations
Section 2.1 provides a summary of the pertinent KELT-North survey data, its reduction, and the light curve processing. We detail the follow-up photometry in Section 2.2, radial velocity observations in Section 2.3, and adaptive optics imaging in Section 2.4.
2.1 KELT-North Photometry
KELT-12 is in KELT-North survey field 10, which is centered on (, ; J2000). We monitored field 10 from 2007 January to 2013 June, collecting a total of 8,150 observations. Our image reduction and light curve processing is described in detail in Siverd et al. (2012), but we summarize the salient features here. In short, we reduced the raw survey data using a custom implementation of the ISIS image subtraction package (Alard & Lupton, 1998; Alard, 2000), combined with point-spread fitting photometry using DAOPHOT (Stetson, 1987). To select likely dwarf and subgiant stars within the field for further analysis, we implemented a reduced proper motion cut (Gould & Morgan, 2003) based on the specific implementation of Collier Cameron et al. (2007); we used proper motions from the Tycho-2 catalog (Høg et al., 2000) and J and H magnitudes from 2MASS (Skrutskie et al. 2006; Cutri et al. 2003).
In an update to the Siverd et al. (2012) procedure, we window-smoothed the stellar light curves with a 90-day window prior to applying both the Trend Filtering Algorithm (TFA; Kovács et al. 2005) to remove systematics common to nearby stars and the Box-Least Squares algorithm (BLS; Kovács et al. 2002) to search the light curves for periodic boxcar-shaped transit signals. We used the TFA and BLS routines as implemented in the VARTOOLS package (Hartman, 2012).
One of the candidates from field 10 that passed our selection criteria was TYC 2619-1057-1 at (). The KELT-North discovery light curve exhibits a transit-like signal at a period of about 5.031 days with a depth of 4 mmag. The light curve contains 7,497 observations – bad observations were removed during the image reduction stage – and is shown in Figure 1. The broadband magnitudes and other stellar properties are listed in Table 1.
|Proper Motion in RA (mas yr)||NOMAD||7|
|Proper Motion in Dec. (mas yr)||NOMAD||7|
|Absolute Systemic RV ()||This PaperaaThe absolute RV uncertainty is due to the systematic uncertainties in the absolute velocities of the RV standard stars.|
|Distance (pc)||This Paper|
|Age||(Gyr)||This PaperbbThe uncertainty does not include possible systematic errors in the adopted evolutionary tracks.|
|Visual Extinction||This Paper|
|()||Galactic Space Velocities ()||This PaperddSee §3.2|
Note. – Magnitudes are on the AB system. 2MASS and WISE uncertainties were increased to 0.03 mag and 0.05 mag, respectively, to account for systematic uncertainties. 1=Høg et al. (2000), 2=Martin et al. (2005), 3=Richmond et al. (2000), 4=Henden et al. (2015) 5=Skrutskie et al. (2006); Cutri et al. (2003), 6=Wright et al. (2010). 7=Zacharias et al. (2004)
2.2 Follow-Up Time-Series Photometry
To improve the precision of the transit-derived parameters and to check against a false positive (e.g. blended eclipsing binary), we acquired several high-cadence, high-precision light curves from our global follow-up network of observers and small telescopes. We obtained a total of 15 partial and full transits between August 2014 and August 2015. The -day period and -hour duration made observing opportunities for full transits scarce. Figure 2 shows the follow-up light curves used in the global fit and analysis, and Table 2 gives a summary of the follow-up observations. Figure 3 shows all primary transit follow-up light curves from Figure 2 combined in five-minute bins. We do not use this light curve for analysis, but we include it to illustrate the statistical power of the full suite of follow-up light curves.
We scheduled the follow-up observations using the Tapir software package (Jensen, 2013) and reduced the follow-up photometric data with the AstroImageJ (AIJ) software package111http://www.astro.louisville.edu/software/astroimagej/ (Collins & Kielkopf 2013; Collins et al. 2016). We also used AIJ to identify the best detrending parameters, and we included these parameters in the global fit (see Section 4.1).
|Date (UT)||Observatory||Filter||FOV||Pixel Scale||Exposure (s)||Detrending Parameters|
|2015-07-10||ZRO||0.92||200||atm. lossaaA representation of losses due to atmospheric changes. Calculated as airmass minus a scaled version of total comp star counts., y-positionbby-centroid pixel value.|
|2015-08-15||PvdKO||0.76||60||airmass, sky background|
2.2.1 Canela’s Robotic Observatory (CROW)
On UT 2014 August 7, we observed one partial transit of KELT-12b at CROW in Portalegre, Portugal. We observed the ingress in the filter with a 12in Schmidt-Cassegrain telescope and a KAF-3200E CCD, which gives a field of view and 0.84 arcsec/pixel resolution.
2.2.2 Peter van de Kamp Observatory (PvdKO)
We obtained two partial transits and one full transit at PvdKO at Swarthmore College. We used the 0.6m Ritchey-Chrétien optical (RCOS) telescope and Apogee U16M CCD, which give a field of view and 0.76 arcsec/pixel resolution with binning. We observed ingress in alternating and filters on UT 2014 September 27, and we observed a full transit in on 2015 August 15.
2.2.3 Kutztown Observatory (Kutztown)
With the Kutztown University Observatory 0.6m RCOS telescope, we observed most of a transit of KELT-12b in and bandpasses on UT 2014 September 27. This system employs a CCD that achieves a image at 0.38 arcsec/pixel.
We used KeplerCam on the 1.2m telescope at the Fred Lawrence Whipple Observatory (FLWO) to observe a partial -band transit on UT 2015 April 06. KeplerCam has a single Fairchild CCD with 0.366 arcsec/pixel and a field of view of .
2.2.5 Salerno University Observatory (Salerno)
We obtained an ingress in on UT 2015 July 5 as well as a nearly-full transit (sans egress) in on UT 2015 July 10 from the Salerno University Observatory in Fisciano Salerno, Italy. The observing setup consists of a 14in Celestron C14 SCT and an SBIG ST2000XM CCD, yielding a resolution of 0.54 arcsec/pixel.
2.2.6 Canis Mayor Observatory (ZRO)
From ZRO in Italy, we observed one nearly complete transit (missing only the ingress) on UT 2015 July 5; the full transit on 2015 July 10; and two separate ingresses on UT 2015 July 15 and UT 2015 July 20. All observations are -band. ZRO uses a 12in Meade LX 200 with an SBIG ST8XME pixel CCD, which gives a resolution of 0.92 arcsec/pixel over a field of view.
2.2.7 Manner-Vanderbilt Ritchey-Chrétien (MVRC) Observatory
We observed one complete transit of KELT-12b on UT 2015 August 20 using the 0.6m MVRC telescope at Mt. Lemmon Observatory in Arizona. The RCOS telescope is equipped with an SBIG STX camera, giving a field of view and 0.39 arcsec/pixel resolution. The transit was observed in both the and bands by alternating filters from one exposure to the next.
2.3 Radial Velocity Observations
We conducted radial velocity (RV) observations of KELT-12 to rule out false positives and to determine the RV orbit. We obtained data using both the Tillinghast Reflector Echelle Spectrograph222http://tdc-www.harvard.edu/instruments/tres (TRES) on the 1.5m Tillinghast Reflector at the Fred L. Whipple Observatory (FLWO) on Mt. Hopkins, Arizona, and the Levy high-resolution optical spectrograph on the 2.4m Automated Planet Finder333http://www.ucolick.org/public/telescopes/apf.html (APF) at Lick Observatory on Mt. Hamilton, California.
We observed KELT-12 with TRES over five months, from UT 2014 June 12 to UT 2014 November 17. We obtained 17 R = 44,000 spectra that were extracted following Buchhave et al. (2010). The first two observations, taken at quadrature, showed single-lined spectra (ruling out a double-lined eclipsing binary) and a low velocity variation suggestive of a substellar companion. The additional high-precision observations were taken to obtain an RV orbit.
We then observed KELT-12 with APF over two months from UT 2015 May 28 to UT 2015 July 21. We obtained 21 R = 100,000 spectra that were extracted in a manner similar to that detailed in Section 3.2 of Fulton et al. (2015); for KELT-12, however, the iodine-free template was observed using the slit, giving a resolution of .
Initial fits to the RV data suggested a linear trend in addition to the periodic orbital motion. The TRES and APF data do not overlap in time, however: the first APF observation was taken after the UT 2014 November 17 TRES observation. To determine whether the linear trend is physical or due to a systematic velocity offset between the TRES and APF data, we obtained four additional TRES observations from UT 2015 December 04 to UT 2016 February 14; thus, the APF data are bracketed in time by TRES observations. Table 3 lists the full set of RV observations from TRES and APF.
Our global fits presented in Section 4.2 all show that the RV linear trend persists at the level. Thus, the linear trend is not due to a systematic offset between the APF and TRES data, but it is not significant enough for us to claim a physical cause (e.g. a massive outer companion) for the linear trend. Long-term RV monitoring of the KELT-12 system will elucidate the origin of this trend.
Bisector spans for both APF and TRES observations were calculated following the prescription of Buchhave et al. (2010) and are also listed in Table 3. We use the bisector spans as part of the false-positive analysis in Section 5, and we show them in Figure 4.
|Rel||Rel||BSccspectral line bisector spans (m s)||ddspectral line bisector span errors (m s)||Source|
|RVaarelative RVs (m s)||bbunrescaled relative RV errors (m s)|
Note. – The relative RV values reported are on the native system for each instrument and cannot be directly compared to values from a different instrument. The bisector spans (BS) from the TRES spectra are computed as described in the text.
2.4 High-resolution Imaging
We obtained speckle imaging of KELT-12 from the Differential Speckle Survey Instrument (DSSI; Horch et al. 2009) on the WIYN 3.5m telescope on UT 2015 October 25. DSSI is a speckle imaging camera which takes images in two bands simultaneously. Images are taken as sets of 1000 40ms speckle frames and then later combined using the method detailed in Howell et al. (2011). The top two panels of Figure 5 show KELT-12 in narrow bands centered on 692 nm () and 880 nm (); each image consists of multiple frame sets stacked into one reconstructed image. Observing conditions were worse than median for the WIYN site, with roughly 1 seeing. No companions were detected down to a contrast limit of 3.31 mag in and 2.78 mag in . The bottom half of Figure 5 show the and contrast curves. These curves are estimated using the method of Horch et al. (2011).
We also obtained adaptive optics imaging of KELT-12 from NIRC2 on Keck II in April 2016. Figure 6 shows the -band AO image and the contrast curve. With 0.49 seeing and an airmass of 1.1, we achieved a contrast of approximately 9 mag at an angular separation of 1; no companions were detected.
|(YY circular; =0 fixed)||(YY eccentric)||(Torres circular; =0 fixed)||(Torres eccentric)|
|Surface gravity (cgs)|
|Effective temperature (K)|
|Argument of periastron (degrees)|
|Semi-major axis (AU)|
|Equilibrium temperature (K)|
|Incident flux (10 erg s cm)|
|Time of inferior conjunction ()|
|Time of periastron ()|
|RV semi-amplitude (m/s)|
|Minimum mass ()|
|RM linear limb darkening|
|RV slope (m/s/day)|
|Mass function ()|
|(YY circular; =0 fixed)||(YY eccentric)||(Torres circular; =0 fixed)||(Torres eccentric)|
|Primary Transit Parameters:|
|Radius of the planet in stellar radii|
|Semi-major axis in stellar radii|
|Time of inferior conjunctionaaFrom the best-fit linear ephemeris. ()|
|PeriodaaFrom the best-fit linear ephemeris. (days)|
|FWHM duration (days)|
|Ingress/egress duration (days)|
|Total duration (days)|
|A priori non-grazing transit probability|
|A priori transit probability|
|Secondary Eclipse Parameters:|
|Time of eclipse ()|
|FWHM duration (days)|
|Ingress/egress duration (days)|
|Total duration (days)|
|A priori non-grazing eclipse probability|
|A priori eclipse probability|
3 Host Star Properties
3.1 Properties from the Literature
Table 1 contains various measurements of KELT-12 collected from the literature or derived in this work. The literature information includes FUV and NUV fluxes from GALEX (Martin et al., 2005); and fluxes from the Tycho-2 catalog (Høg et al., 2000); and from The Amateur Sky Survey (TASS; Richmond et al. 2000); , , and Sloan , , and fluxes from the AAVSO APASS catalogue (Henden et al., 2015); near-infrared fluxes in the , , and bands from the 2MASS Point Source Catalog (Cutri et al. 2003; Skrutskie et al. 2006); near- and mid-IR fluxes in three WISE passbands (Wright et al., 2010); and proper motions from the NOMAD catalog (Zacharias et al., 2004).
3.2 UVW Space Motion
We determine the motion of KELT-12 through the Galaxy to determine its membership among the Galactic stellar populations. We adopt an absolute RV of , calculated as the error-weighted mean of the TRES and APF mean absolute RVs. The individual absolute RVs are listed in Table 3; the uncertainty is due to the systematic uncertainties in the absolute RVs of the RV standard stars. We combine the adopted absolute RV with the NOMAD proper motions (Zacharias et al., 2004) and the distance that we estimate from fitting the spectral energy distribution (SED; Section 3.3) to calculate U, V, and W space velocities. We adopt the Coşkunoǧlu et al. (2011) solar velocity with respect to the Local Standard of Rest for this calculation.
We find that – all in units of – where positive points toward the Galactic Center. We find a probability that KELT-12 is a thin disk star, according to Bensby et al. (2003).
3.3 SED Analysis
We construct an empirical spectral energy distribution (SED) of KELT-12 using the available broadband photometry in the literature, which is listed in Table 1 and in Section 3.1. We fit this SED to NextGen models from Hauschildt et al. (1999) by fixing the values of , , and to the values inferred from the global fit to the light curve, RV, and spectroscopic data; these parameters are listed in Table 4. We then find the values of the visual extinction and distance that minimize the of the fit. The best-fit model has a reduced of 1.99 for 12 degrees of freedom, suggesting that the photometric uncertainties are underestimated. We find and = pc.
We note that the quoted statistical uncertainties on and are likely to be underestimated because we have not accounted for the uncertainties in values of , , and [Fe/H]used to derive the model SED. Furthermore, it is likely that alternate model atmospheres would predict somewhat different SEDs and thus values of the extinction and distance.
3.4 Spectroscopic Analysis
We derive KELT-12’s stellar properties from both the APF and TRES spectra. To analyze the APF spectra, we use SpecMatch (Petigura, 2015). This analysis yields K, , , and km s.
To analyze the TRES spectra, we use the Spectral Parameter Classification (SPC) procedure, version 2.2 (Buchhave et al., 2012). We ran SPC initially with , , [m/H], and as free parameters. We took the error-weighted mean value for each stellar parameter and adopted the mean error for each parameter. From this initial run, we found that K, , [m/H], and km s. Only the surface gravity agrees with the APF SpecMatch value within ; of note, differs by . Additionally, an initial analysis of the transit data with stellar models and with empirical relations using the APF and [Fe/H] values as priors resulted in , inconsistent with both APF and TRES values at .
Because the gravity from the transit data and stellar models is expected to be more accurate than the spectroscopic gravities, we re-ran SPC on the TRES data with the surface gravity fixed at = 3.9, giving us K, , [m/H] , and km s and bringing , , and metallicity into agreement () with the APF SpecMatch values. We adopt these , [m/H], and as priors for the final analysis in Section 4.2.
4 Planetary Characterization
To determine the physical and observable properties of the KELT-12 system, we perform a global fit of the photometric and spectroscopic data using a modified version of the IDL exoplanet fitting tool EXOFAST (Eastman et al., 2013). In short, we run simultaneous Markov Chain Monte Carlo analyses on the RV data and follow-up photometry to determine the posterior probability distribution of each parameter; the technique is described in detail in Siverd et al. (2012). EXOFAST constrains the stellar mass and radius by using either the Yonsei-Yale stellar evolution models (Demarque et al., 2004) or the empirical relations of Torres et al. (2010) (hereafter ”the Torres relations”). We include the raw follow-up photometry and the relevant detrending parameters (see Section 4.1) in the fits. We set a prior on the orbital period days from analysis of the KELT-North discovery light curve and the follow-up photometry.
4.1 Light Curve Detrending
Owing to the KELT-12 system’s nearly six-hour transit duration, our analysis relies extensively on partial transits. As a result of this and the shallow, mmag transit depth, the shape of the transit and the inferred transit parameters can thus be heavily influenced by our choice of detrending parameters. As described in Collins et al. (2014), we use AIJ to determine the detrending parameters that best improve the individual light curve fits, as including all possible detrending parameters for all light curves in the EXOFAST global fit would be prohibitively expensive from a computational point-of-view. We list the included detrending parameters for each fitted data set in Table 2.
4.2 Global Fit
To determine the prior values of , [Fe/H], and that we imposed on the final global fits, we performed an iterative SPC analysis as described in Section 3.4 and adopted the final TRES SPC values for , [Fe/H], and as spectroscopic priors.
We ran the global fits using either the YY isochrones or the Torres relations, and we either forced circular orbits or allowed for eccentric orbits; permutation of these choices yielded four global fits. In all four fits, we allowed for a non-zero RV slope.
Tables 4 and 5 list the best-fit parameters and their 68% confidence intervals for the four cases. While the two eccentric fits report eccentricities that are formally inconsistent with zero at the level, eccentricity measurements are biased to artificially large values due to the hard boundary at 0, so a significance of is generally required to claim an eccentric orbit (Lucy & Sweeney, 1971). Because the eccentrities do not meet this significance threshold and because the other parameters agree across all four scenarios within , we adopt the YY circular global fit for our analyses in this paper.
We note that all four cases exhibit a best-fit RV slope that is inconsistent with zero at the level. While we do not claim a strong detection of an RV slope given this low significance, we note that long-term RV monitoring would determine whether or not the RV slope is physical and, if so, if it is due to a massive outer companion. Figure 8 shows the RV slope for the adopted best-fit model.
We searched for transit timing variations (TTVs) in the system by allowing the transit times for each follow-up light curve to vary. The ephemeris is constrained by the RV data and a prior imposed from the KELT-North discovery light curve and the follow-up photometry. The transit times are listed in Table 6 and Figure 9. We find only one TTV, on epoch 25, but a deviation from a different observatory during the same epoch suggests that this TTV is likely spurious. Hence, we do not claim evidence for TTVs in the KELT-12 system.
Finally, we report a high-precision ephemeris for the KELT-12 system. The time of inferior conjunction in is and the period is days. These are also included in Table 5.
5 False-positive Analysis
We perform several analyses to exclude possible false-positive scenarios. First, we find that the depths found in each of our follow-up light curves are consistent with each other, even across different photometric filters. Moreover, the follow-up observations are well-modeled by a dark companion occulting a star, and the limb darkening effects on the light curves from the host star are consistent with the and determined from the spectrum. We can therefore rule out a blended EB scenario in which the blended stars have significantly different colors, as such a blend would effect detectable differences in the measured depths across our photometric filters.
We also investigate the possibility that the RV variations are caused by stellar activity or a nearby, unresolved eclipsing binary: in these cases, spectral line asymmetries will induce bisector span (BS) variations that correlate with radial velocity. We calculate the APF BS measurements as described in Section 5.2 of Fulton et al. (2015), and we follow Torres et al. (2007) to calculate the TRES BS measurements. Analyzing both the APF and TRES measurements, we calculate a Spearman rank correlation coefficient of -0.28 (), which does not indicate a significant correlation between BS and RV. The BS measurements and uncertainties are listed in Table 3 and are plotted versus RV in Figure 10.
6 Evolutionary Analysis
6.1 Stellar Models and Age
To estimate KELT-12’s age, we use the , , , and [Fe/H] values from the adopted global fit (see Table 4) along with evolutionary tracks from the Yonsei-Yale stellar models (Demarque et al., 2004). We assume uniform priors on , , and [Fe/H], resulting in a non-uniform prior on the stellar age. Figure 11 shows the best-fit theoretical HR diagram for KELT-12 along with evolutionary tracks that correspond to the uncertainties in and . We infer that KELT-12 is Gyr old (Table 1), approaching the main-sequence turn-off but not yet a subgiant; we note that this age is model-dependent.
6.2 Insolation Evolution
Demory & Seager (2011) found that planets receiving more than insolation from their host stars will have inflated radii compared to planets receiving insolation below this threshold. As listed in Table 4, KELT-12b receives over 10 times as much flux, with an insolation of 10 erg s cm. Along with a density and a mass , KELT-12b is an inflated hot Jupiter that follows the insolation-inflation trend of Demory & Seager (2011). It is worth investigating KELT-12b’s insolation history to determine whether or not its incident flux always exceeded the Demory & Seager (2011) threshold. An understanding of KELT-12b’s insolation evolution enables us to examine the timescales of planetary inflation mechanisms (cf. Assef et al. 2009 and Spiegel & Madhusudhan 2012).
To infer KELT-12b’s insolation history, we simulate the evolution of the KELT-12 system. We impose the adopted global fit parameters (see Tables 4 and 5) as the present-day boundary conditions. We assume solid-body rotation for KELT-12 and that tidal torques exerted by the planet are the only physical influence on the stellar rotation. We test three stellar tidal quality factors : = 5, 6, and 7. Figure 12 shows the results of our simulation. The top panel shows that KELT-12b’s incident flux has exceeded the Demory & Seager (2011) threshold throughout KELT-12’s main-sequence lifetime, despite its large orbital separation (bottom panel); as a result, KELT-12b has always received an amount of stellar insolation that is greater than the boundary suggested by Demory & Seager (2011) for inflated hot Jupiters. Moreover, the insolation is insensitive to our choice of for the system parameters that we have adopted.
As mentioned in Section 6.2, Demory & Seager (2011) found that planets irradiated at levels above are inflated relative to less-irradiated planets; additionally, Weiss et al. (2013) found that for planets with , whereas for less-massive planets. Since all transiting gas giant planets discovered by KELT (along with the brown dwarf KELT-1b) receive stellar flux in excess of this amount, a significant fraction of KELT planets exhibit inflated radii. Since most KELT planet hosts are also hotter than the Kraft break at K, we investigate associations between planet radius, insolation, and host star effective temperature to check whether or not those system parameters are typical of transiting hot Jupiters.
Figure 13 shows the planet radius as a function of insolation for transiting planets in the Extrasolar Planet Encyclopedia444http://www.exoplanet.eu, accessed 2016 July 17. (Schneider et al., 2011). To within the uncertainties, KELT-12b is one of the largest – if not the largest – transiting hot Jupiters known. In this analysis, we restrict ourselves to the KELT-12 system (red diamond) plus the 339 transiting systems with listed host radii, companion radii , semimajor axes, and host effective temperatures.
As the top panel of Figure 13 shows, transit surveys have found giant planets with insolation below , but few of those have radii above . Above this threshold, the radii of known giant planets increase; above 10 erg s cm, the overwhelming majority of planets have , with only four planets having smaller radii.
To examine whether or not this distribution changes with stellar effective temperature, we divided the sample into transiting giant planets around hot stars (blue points) and cool stars (grey points), using the Kraft break ( K) as the partition. We chose this effective temperature for its physical significance: stars above this temperature have largely radiative envelopes, with very thin or even absent convective envelopes, whereas stars cooler than 6250 K have increasingly larger convective envelopes (Kraft, 1967). As a result, stars cooler than the Kraft break have magnetic fields that cause them to spin down with time, whereas hotter stars largely retain their primordial spin rates. While both samples show radius inflation above the Demory & Seager (2011) threshold, none of the 64 planets in our ”hot” sample have radii below , and only two receive less than incident flux.
Figure 14 shows planet radius as a function of period for the same two populations of giant planets. For giant planets around cool stars, the planet radius decreases with increasing orbital period (hence decreasing incident flux). However, all but a couple known giant planets around hot stars are on short-period orbits; only three systems orbit on -day periods, and all three have radii that are distinctly larger than the radii of the giant planets around cool stars at comparable periods.
We note that the paucity of giant planets with low insolation around hot stars is most likely a selection effect. The bottom panel of Figure 13 shows the planet radius versus insolation for the subsample of 77 giant planets discovered by Kepler and K2. This subsample includes the bulk of systems below the Demory & Seager (2011) threshold. Only six of the Kepler systems orbit hot stars, including the two receiving low incident flux. Kepler avoided searching for planets around hot stars (Batalha et al., 2010), which explains the dearth of such systems. Conversely, ground-based transit surveys are biased towards discovering large planets on short orbits (Beatty & Gaudi, 2008), and thus are biased towards planets receiving high amounts of radiation from their hosts.
From our available data, we cannot support the hypothesis that giant planets around hotter stars tend to be more inflated than giant planets around cooler stars until the selection effects of ground- and space-based surveys are taken into account. The TESS target sample includes hot stars and will recover some longer-period systems (Sullivan et al., 2015), and this will be complemented by ground-based transit surveys’ increasing sensitivity to longer-period transiting systems (due to the increasing baseline of observations). In the coming years, we will extend the sample of hot Jupiters around hot stars to longer periods (and thus lower insolations), putting us in a better position to investigation any differences in giant planet inflation caused by the stellar effective temperature.
We announce the discovery of KELT-12b, an inflated hot Jupiter on a 5.03-day period around a mildly evolved star. KELT-12 appears to be a single-star system as AO imaging has revealed no companions beyond within nine magnitudes in apparent brightness. With a mass of and a radius of , KELT-12b is one of the most inflated hot Jupiters known, despite its relatively long orbital period.
The majority of giant planets transiting hot ( K) stars have radii exceeding 1 and receive stellar flux exceeding – the threshold above which giant planets appear inflated, as found by Demory & Seager (2011). However, the lack of giant planets around hot stars on long-period orbits (and therefore receiving less stellar radiation) is likely due to selection biases in both ground- and space-based transit surveys. Determining whether giant planets around hot stars are systematically more inflated than giant planets around cooler stars hinges on both the inclusion of hot stars in the TESS survey sample and the longevity of ongoing ground-based transit surveys such as HAT, KELT, and SuperWASP.
- Alard (2000) Alard, C. 2000, A&AS, 144, 363
- Alard & Lupton (1998) Alard, C., & Lupton, R. H. 1998, ApJ, 503, 325
- Assef et al. (2009) Assef, R. J., Gaudi, B. S., & Stanek, K. Z. 2009, ApJ, 701, 1616
- Bakos et al. (2004) Bakos, G., Noyes, R. W., Kovács, G., et al. 2004, PASP, 116, 266
- Batalha et al. (2010) Batalha, N. M., Borucki, W. J., Koch, D. G., et al. 2010, ApJ, 713, L109
- Beatty & Gaudi (2008) Beatty, T. G., & Gaudi, B. S. 2008, ApJ, 686, 1302
- Beatty et al. (2014) Beatty, T. G., Collins, K. A., Fortney, J., et al. 2014, ApJ, 783, 112
- Bensby et al. (2003) Bensby, T., Feltzing, S., & Lundström, I. 2003, A&A, 410, 527
- Borucki et al. (2010) Borucki, W. J., Koch, D., Basri, G., et al. 2010, Science, 327, 977
- Brown (2003) Brown, T. M. 2003, ApJ, 593, L125
- Buchhave et al. (2010) Buchhave, L. A., Bakos, G. Á., Hartman, J. D., et al. 2010, ApJ, 720, 1118
- Buchhave et al. (2012) Buchhave, L. A., Latham, D. W., Johansen, A., et al. 2012, Nature, 486, 375
- Coşkunoǧlu et al. (2011) Coşkunoǧlu, B., Ak, S., Bilir, S., et al. 2011, MNRAS, 412, 1237
- Collier Cameron et al. (2007) Collier Cameron, A., Wilson, D. M., West, R. G., et al. 2007, MNRAS, 380, 1230
- Collins & Kielkopf (2013) Collins, K., & Kielkopf, J. 2013, AstroImageJ: ImageJ for Astronomy, Astrophysics Source Code Library
- Collins et al. (2016) Collins, K. A., Kielkopf, J. F., & Stassun, K. G. 2016, ArXiv e-prints
- Collins et al. (2014) Collins, K. A., Eastman, J. D., Beatty, T. G., et al. 2014, AJ, 147, 39
- Cutri et al. (2003) Cutri, R. M., Skrutskie, M. F., van Dyk, S., et al. 2003, VizieR Online Data Catalog, 2246, 0
- Demarque et al. (2004) Demarque, P., Woo, J.-H., Kim, Y.-C., & Yi, S. K. 2004, ApJS, 155, 667
- Demory & Seager (2011) Demory, B.-O., & Seager, S. 2011, ApJS, 197, 12
- Eastman et al. (2013) Eastman, J., Gaudi, B. S., & Agol, E. 2013, PASP, 125, 83
- Eastman et al. (2016) Eastman, J. D., Beatty, T. G., Siverd, R. J., et al. 2016, AJ, 151, 45
- Fressin et al. (2007) Fressin, F., Guillot, T., Morello, V., & Pont, F. 2007, A&A, 475, 729
- Fulton et al. (2015) Fulton, B. J., Weiss, L. M., Sinukoff, E., et al. 2015, ApJ, 805, 175
- Gaudi (2005) Gaudi, B. S. 2005, ApJ, 628, L73
- Gould & Morgan (2003) Gould, A., & Morgan, C. W. 2003, ApJ, 585, 1056
- Hartman (2012) Hartman, J. 2012, VARTOOLS: Light Curve Analysis Program, Astrophysics Source Code Library
- Hauschildt et al. (1999) Hauschildt, P. H., Allard, F., Ferguson, J., Baron, E., & Alexander, D. R. 1999, ApJ, 525, 871
- Henden et al. (2015) Henden, A. A., Levine, S., Terrell, D., & Welch, D. L. 2015, in American Astronomical Society Meeting Abstracts, Vol. 225, American Astronomical Society Meeting Abstracts, 336.16
- Høg et al. (2000) Høg, E., Fabricius, C., Makarov, V. V., et al. 2000, A&A, 355, L27
- Horch et al. (2011) Horch, E. P., Gomez, S. C., Sherry, W. H., et al. 2011, AJ, 141, 45
- Horch et al. (2009) Horch, E. P., Veillette, D. R., Baena Gallé, R., et al. 2009, AJ, 137, 5057
- Howell et al. (2011) Howell, S. B., Everett, M. E., Sherry, W., Horch, E., & Ciardi, D. R. 2011, AJ, 142, 19
- Howell et al. (2014) Howell, S. B., Sobeck, C., Haas, M., et al. 2014, PASP, 126, 398
- Jensen (2013) Jensen, E. 2013, Tapir: A web interface for transit/eclipse observability, Astrophysics Source Code Library
- Kovács et al. (2005) Kovács, G., Bakos, G., & Noyes, R. W. 2005, MNRAS, 356, 557
- Kovács et al. (2002) Kovács, G., Zucker, S., & Mazeh, T. 2002, A&A, 391, 369
- Kraft (1967) Kraft, R. P. 1967, ApJ, 150, 551
- Lucy & Sweeney (1971) Lucy, L. B., & Sweeney, M. A. 1971, AJ, 76, 544
- Martin et al. (2005) Martin, D. C., Fanson, J., Schiminovich, D., et al. 2005, ApJ, 619, L1
- Ochsenbein et al. (2000) Ochsenbein, F., Bauer, P., & Marcout, J. 2000, A&AS, 143, 23
- Pepper & Gaudi (2005) Pepper, J., & Gaudi, B. S. 2005, ApJ, 631, 581
- Pepper et al. (2003) Pepper, J., Gould, A., & Depoy, D. L. 2003, Acta Astron., 53, 213
- Pepper et al. (2012) Pepper, J., Kuhn, R. B., Siverd, R., James, D., & Stassun, K. 2012, PASP, 124, 230
- Pepper et al. (2007) Pepper, J., Pogge, R. W., DePoy, D. L., et al. 2007, PASP, 119, 923
- Petigura (2015) Petigura, E. A. 2015, PhD thesis, University of California, Berkeley
- Pollacco et al. (2006) Pollacco, D. L., Skillen, I., Collier Cameron, A., et al. 2006, PASP, 118, 1407
- Pont et al. (2006) Pont, F., Zucker, S., & Queloz, D. 2006, MNRAS, 373, 231
- Richmond et al. (2000) Richmond, M. W., Droege, T. F., Gombert, G., et al. 2000, PASP, 112, 397
- Ricker et al. (2015) Ricker, G. R., Winn, J. N., Vanderspek, R., et al. 2015, Journal of Astronomical Telescopes, Instruments, and Systems, 1, 014003
- Rouan et al. (1998) Rouan, D., Baglin, A., Copet, E., et al. 1998, Earth Moon and Planets, 81, 79
- Schneider et al. (2011) Schneider, J., Dedieu, C., Le Sidaner, P., Savalle, R., & Zolotukhin, I. 2011, A&A, 532, A79
- Seager & Deming (2010) Seager, S., & Deming, D. 2010, ARA&A, 48, 631
- Sing et al. (2016) Sing, D. K., Fortney, J. J., Nikolov, N., et al. 2016, Nature, 529, 59
- Siverd et al. (2012) Siverd, R. J., Beatty, T. G., Pepper, J., et al. 2012, ApJ, 761, 123
- Skrutskie et al. (2006) Skrutskie, M. F., Cutri, R. M., Stiening, R., et al. 2006, AJ, 131, 1163
- Spiegel & Madhusudhan (2012) Spiegel, D. S., & Madhusudhan, N. 2012, ApJ, 756, 132
- Stetson (1987) Stetson, P. B. 1987, PASP, 99, 191
- Sullivan et al. (2015) Sullivan, P. W., Winn, J. N., Berta-Thompson, Z. K., et al. 2015, ApJ, 809, 77
- Torres et al. (2010) Torres, G., Andersen, J., & Giménez, A. 2010, A&A Rev., 18, 67
- Torres et al. (2005) Torres, G., Konacki, M., Sasselov, D. D., & Jha, S. 2005, ApJ, 619, 558
- Torres et al. (2007) Torres, G., Bakos, G. Á., Kovács, G., et al. 2007, ApJ, 666, L121
- Weiss et al. (2013) Weiss, L. M., Marcy, G. W., Rowe, J. F., et al. 2013, ApJ, 768, 14
- Wright et al. (2010) Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al. 2010, AJ, 140, 1868
- Zacharias et al. (2004) Zacharias, N., Monet, D. G., Levine, S. E., et al. 2004, in Bulletin of the American Astronomical Society, Vol. 36, American Astronomical Society Meeting Abstracts, 1418
- Zhou et al. (2016) Zhou, G., Rodriguez, J. E., Collins, K. A., et al. 2016, ArXiv e-prints