MNRAS 000 1ndash12 (2017) Preprint 11 December 2017 Compiled using MNRAS LATEX style file v30
K2-139 b a low-mass warm Jupiter on a 29-day orbit transiting anactive K0 V star
O Barragaacuten1 D Gandolfi1 A M S Smith2 H J Deeg34 M C V Fridlund56C M Persson5 P Donati7 M Endl8 Sz Csizmadia2 S Grziwa9 D Nespral34A P Hatzes10 W D Cochran8 L Fossati11 S S Brems12 J Cabrera2 F Cusano7Ph Eigmuumlller1 C Eiroa13 A Erikson2 E Guenther10 J Korth9 D Lorenzo-Oliveira14L Mancini151617 M Paumltzold9 J Prieto-Arranz34 H Rauer218 I Rebollido13J Saario19 and OV Zakhozhay1520
1Dipartimento di Fisica Universitaacute di Torino via P Giuria 1 10125 Torino Italy2Institute of Planetary Research German Aerospace Center Rutherfordstrasse 2 12489 Berlin Germany3Instituto de Astrofiacutesica de Canarias 38205 La Laguna Tenerife Spain4Departamento de Astrofiacutesica Universidad de La Laguna 38206 La Laguna Tenerife Spain5Department of Earth and Space Sciences Chalmers University of Technology Onsala Space Observatory 439 92 Onsala Sweden6Leiden Observatory University of Leiden PO Box 9513 2300 RA Leiden The Netherlands7INAF - Osservatorio Astronomico di Bologna Via Ranzani 1 20127 Bologna Italy8Department of Astronomy and McDonald Observatory University of Texas at Austin 2515 Speedway Stop C1400 Austin TX 78712 USA9Rheinisches Institut fuumlr Umweltforschung an der Universitaumlt zu Koumlln Aachener Strasse 209 50931 Koumlln Germany10Thuumlringer Landessternwarte Tautenburg Sternwarte 5 07778 Tautenburg Germany11Space Research Institute Austrian Academy of Sciences Schmiedlstrasse 6 A-8041 Graz Austria12Landessternwarte Zentrum fuumlr Astronomie der Universitaumlt Heidelberg Koumlnigstuhl 12 69117 Heidelberg Germany13Departamento Fiacutesica Teoacuterica Unversidad Autoacutenoma de Madrid Cantoblanco 28049 Madrid Spain14Universidade de Satildeo Paulo Departamento de Astronomia do IAGUSP Rua do Matatildeo 1226 Cidade Universitaacuteria 05508-900 Satildeo Paulo SP Brazil15Max-Planck-Institut fuumlr Astronomie Koumlnigstuhl 17 D-69117 Heidelberg Germany16Department of Physics University of Rome Tor Vergata Rome17INAF - Astrophysical Observatory of Turin Turin18Center for Astronomy and Astrophysics TU Berlin Hardenbergstr 36 10623 Berlin Germany19Nordic Optical Telescope Apartado 474 E-38700 Santa Cruz de La Palma Spain20Main Astronomical Observatory National Academy of Sciences of the Ukraine 27 Akademika Zabolotnoho St 03143 Kyiv Ukraine
Last updated in original form
ABSTRACTWe announce the discovery of K2-139 b (EPIC 218916923 b) a transiting warm-Jupiter(Teq=547plusmn25K) on a 29-day orbit around an active (log RprimeHK =minus446plusmn 006) K0V star inK2 Campaign 7 We derive the systemrsquos parameters by combining the K2 photometry withground-based follow-up observations With a mass of 0387+0083
minus0075 MJ and radius of 0808+0034minus0033
RJ K2-139 b is one of the transiting warm Jupiters with the lowest mass known to date Theplanetary mean density of 091+024
minus020 g cmminus3can be explained with a core of sim50 Moplus Given thebrightness of the host star (V = 11653mag) the relatively short transit duration (sim5 hours)and the expected amplitude of the Rossiter-McLaughlin effect (sim25 m sminus1) K2-139 is an idealtarget to measure the spin-orbit angle of a planetary system hosting a warm Jupiter
Key words planetary systems mdash planets and satellites detection mdash planets and satellitesindividual K2-139 b (EPIC 218916923 b) mdash stars fundamental parameters
E-mail oscarbarraganvileduunitoit
1 INTRODUCTION
Gas-giant planets (Mp amp 03 MJup Hatzes amp Rauer 2015) with or-bital periods ranging between sim10 and 100 days are called warm
copy 2017 The Authors
arX
iv1
702
0069
1v2
[as
tro-
phE
P] 8
Dec
201
7
2 O Barragaacuten et al
Jupiters Theymark the transition between hot Jupiters (giant planetswith orbital period between sim1 and 10 days) and Jupiter analogues(orbital period longer than 100 days) They seem to be less com-mon than hot Jupiters and their formation scenario is still underdebate (eg Frewen amp Hansen 2016 Boley et al 2016) Whereasit is commonly accepted that hot Jupiters did not form in situ (egKley amp Nelson 2012) but rather formed beyond the snow line andthen migrated inwards to their current position it has been recentlyproposed that warm Jupiters might have formed in situ (eg Boleyet al 2016 Huang et al 2016)
Eighty warm Jupiters have been discovered so far from bothground- (eg da Silva et al 2007 Brahm et al 2016 Jenkins et al2017) and space-based surveys (eg Deeg et al 2010 Saad-Oliveraet al 2017 Smith et al 2017) About thirty are known to transittheir parent star and only thirteen have masses and radii known witha precision better than 251 They have been detected both in low-eccentricity orbits (e 04 eg Brahm et al 2016 Niedzielski et al2016 Smith et al 2017) as well as in highly eccentric orbits (egDawson et al 2012 Ortiz et al 2015) Dong et al (2014) foundthat warm Jupiters with high eccentricities (e amp 04) tend to havea massive planetarystellar companion in a long period orbit Thearchitectures of these systems suggest that eccentric warm Jupitersmight have reached their current positions via high-eccentricitymigration excited by the outer companion (Dong et al 2014) On theother hand warm Jupiters with no detected Jovian companion tendto have lower eccentricities peaked around 02 This suggests thattwo different types of warm Jupiters might exist those formed viahigh-eccentricity migration and those formed in situ Alternativelywarm Jupiters in low-eccentricity orbits can also result from disc-drivenmigration from the outer region of the system (KleyampNelson2012)
Petrovich amp Tremaine (2016) studied the possibility that warmJupiters are undergoing secular eccentricity oscillations induced byan outer companion in an eccentric andor mutually inclined orbitTheir model suggests that high-eccentricity migration can accountfor most of the hot Jupiters as well as for most of the warm Jupiterswith e amp 04 However it cannot account for the remaining popula-tion of low-eccentricity warm Jupiters which must have undergonea different formation mechanism The low efficiency to generatewarm Jupiters in nearly circular orbits via high-eccentricity migra-tion has been corroborated by Hamers et al (2016) and Antoniniet al (2016) using numerical simulations
In order to test different planet formationmechanisms we needto characterize the population of warm Jupiters in terms of planetarymass radius and orbital parametersWe herein present the discoveryof K2-139 b (EPIC 218916923 b) a transiting warm Jupiter (Mp =
0387+0083minus0075 MJ Rp = 0808+0034
minus0033 RJ) in a 29-day orbit aroundan active K0V star that has been photometrically monitored bythe K2 space-mission during its Campaign 7 We combine the K2photometry with ground-based imaging and high-precision radialvelocity measurements to confirm the planet and derive the mainparameters of the system
2 K2 PHOTOMETRY
K2 Campaign 7 was performed between 2015 October 04 UT and2015 December 26 UT2 The Kepler spacecraft was pointed at
1 Source httpexoplaneteu as of January 20172 See httpkeplersciencearcnasagovk2-fieldshtml
Table 1 Main identifiers coordinates optical and infrared magnitudes andproper motion of K2-139
Parameter Value Source
Main IdentifiersTYC 6300-2008-1 TychoEPIC 218916923 EPICUCAC 361-185490 EPIC2MASS 19161596-1754384 EPIC
Note ndash Values of fields marked with EPIC are taken from the EclipticPlane Input Catalog available at httparchivestscieduk2epicsearchphp Values marked with Gaia 2MASS and WISE arefrom Fabricius et al (2016) Cutri et al (2003) and Cutri et al (2012)respectively The WISE W4 magnitude is an upper limit
coordinates α = 19h11m19s δ = minus2321prime36primeprime K2 observed si-multaneously 13 469 sources in long cadence mode (sim30 minuteintegration time) and 72 objects in short cadence mode (sim1 minuteintegration time) leading to a total of 13 541 light curves
For the detection of transiting planet candidates we used theK2 Campaign 7 light curves3 extracted by Vanderburg amp John-son (2014) We analyzed the light curves using the DST algorithm(Cabrera et al 2012) and the EXOTRANS pipeline (Grziwa et al 2012Grziwa amp Paumltzold 2016) Both codes have been used extensively onCoRoT (Carpano et al 2009 Cabrera et al 2009 Erikson et al2012 Carone et al 2012 Cavarroc et al 2012) andKepler (Cabreraet al 2014 Grziwa amp Paumltzold 2016) data These search algorithmsdetect periodic patterns in time series photometric data DST usesan optimized transit shape with the same number of free parametersas for the BLS algorithm (Box-fitting Least Squares Kovaacutecs et al2002) and it also implements better statistics for signal detectionEXOTRANS uses the BLS algorithm combined with the wavelet-basedfilter technique VARLET (Grziwa amp Paumltzold 2016) diminishing theeffects of stellar variability and data discontinuities
We detected a periodic transit-like signal associated with
3 Publicly available at httpswwwcfaharvardedu~avanderballk2c7obshtml
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 3
2480 2500 2520 2540BJD shy 2454833
099
100
101
Rel
ativ
e flu
x
Figure 1 K2 Light curve for K2-139 as extracted by Luger et al (2016) The positions of the 3 observed transits are marked with vertical dashed lines
the star EPIC 218916923 with both DST and EXOTRANS Asa sanity check we downloaded the EVEREST light curve ofEPIC 218916923 (Luger et al 2016) and detected the same sig-nal We note that Vanderburg amp Johnson (2014) and Luger et al(2016) used the same mask to extract the time-series data from therawK2 images EPIC 218916923 was proposed forK2 observationsby programs GO7086 (PI Thompson) GO7030 (PI Howard) andGO7087 (PI Dragomir) We will hereafter refer to the star and itstransiting planet as K2-139 and K2-139 b respectively
We searched the Vanderburg amp Johnson (2014)rsquos light curvefor odd-even transit depth variation and secondary eclipse thatmighthint to a binary scenario making the system a likely false positiveNone of themwere significantly detected The depth of the oddeventransits agrees within the 1-σ uncertainty of 16 times 10minus3 whereasthe 3-σ upper limit of the occultation depth is 79 times 10minus5 both re-spect to the normalized flux We proceeded to more detailed fittingof the light curve as well as ground-based imaging (Sect 3) andspectroscopic observations (Sect 4) The main identifiers coordi-nates optical and infrared magnitudes and proper motions of thestar are listed in Table 1 We display the EVEREST K2 light curveof K2-139 in Fig 1
3 ALFOSC IMAGING
K2 Campaign 7 is projected close to the galactic center and thus ina relatively crowded stellar region In order to estimate the contam-ination factor arising from sources whose light leaks into the photo-metricmasks used byVanderburgamp Johnson (2014) and Luger et al(2016) we observed K2-139 on 13 September 2016 (UT) with theALFOSC camera mounted at the Nordic Optical Telescope (NOT)of Roque de los Muchachos Observatory (La Palma Spain) Thesky conditions were photometric with excellent seeing conditions(sim06primeprime) We used the Bessel R-filter and acquired 16 images of 6sec 2 images of 20 sec and 1 image of 120 sec The data were biassubtracted and flat-fielded using dusk sky flats Aperture photom-etry was then performed on all stars within the mask used in theextraction of the light curve by Vanderburg amp Johnson (2014) andLuger et al (2016)
Several fainter stars can be identified inside the photometricmask (Fig 2) of which the two brightest sources are also in theEPIC catalog with Kepler band magnitudes of 168 and 184 Theclosest detected source is a 68-mag fainter star at 38primeprime South of K2-
Figure 2 ALFOSC Bessel R-band image of the sky region around K2-139North is up and East is to the left The target star is the brightest sourcein the middle The solid black polygon marks the EVEREST photometricmask (Luger et al 2016)
139 We can exclude stars as faint as sim20mag at an angular distancelarger than sim06primeprime from K2-139 It is worth noting that the fainteststar whose flux could account for the sim1 deep transit of K2-139cannot be more than sim5 mag fainter than our target The summedflux of these faint stars amounts to 14plusmn03 of the total off-transitflux within the apertureWe subtracted this contamination flux fromthe EVEREST K2 light curve prior to performing the joint analysispresented in Sect 6
4 HIGH-RESOLUTION SPECTROSCOPY
In June and August 2016 we obtained two reconnaissance spectraof K2-139 with the Tull spectrograph (Tull et al 1995) at the 27-mtelescope at McDonald Observatory (Texas USA) The high res-olution (R asymp 60 000) spectra have a signal-to-noise ratio of sim30per pixel at 5500Aring We reduced the data using standard IRAF rou-tines and derived preliminary spectroscopic parameters using ourcode Kea (Endl amp Cochran 2016) The results from both spec-
MNRAS 000 1ndash12 (2017)
4 O Barragaacuten et al
Table 2 Radial velocity measurements and activity indexes of K2-139
tra are nearly identical and reveal a star with Teff = 5500plusmn 100 Klog g = 465plusmn 012 (cgs) [FeH]=+011plusmn 012 dex and a slow pro-jected rotational velocity of v sin iasymp 2 km sminus1
The high-precision radial velocity follow-up of K2-139 wasstarted in June 2016with the FIbre-fed Eacutechelle Spectrograph (FIESFrandsenampLindberg 1999 Telting et al 2014)mounted at the 256-m Nordic Optical Telescope (NOT) The observations were carriedout as part of the OPTICON andCAT observing programs 16A055P53-201 and P53-203 We used the high-resmode which providesa resolving power of Rasymp 67 000 in the whole visible spectral range(3700 minus 7300Aring) The exposure time was set to 2100 ndash 3600 secbased on sky conditions and observing scheduling constraints Fol-lowing the observing strategy outlined in Buchhave et al (2010)and Gandolfi et al (2015) we traced the RV drift of the instrumentby acquiring long-exposed (Texp asymp 35 sec) ThAr spectra immedi-ately before and after the target observations The typical RV driftmeasured between two ThAr spectra bracketing a 2100 ndash 3600 secscience exposure is about 50 ndash 80 m sminus1 A linear interpolation ofthe RV drift to the mid-time of the science exposure allows us toachieve a radial velocity zero-point stability of about 5 ndash 6 m sminus1which is 2 ndash 3 times smaller than the nominal error bars listed in Ta-ble 2 The data reduction uses standard IRAF and IDL routines Thesignal-to-noise (SN) ratio of the extracted spectra is sim30 ndash 40 perpixel at 5500 Aring Radial velocity measurements were extracted viamulti-order cross-correlation with the RV standard star HD182572observed with the same instrument set-up as K2-139
We also observed K2-139 in July August and September 2016with the HARPS (Mayor et al 2003) and HARPS-N (Cosentinoet al 2012) spectrographs mounted at the ESO 36-m Telescope ofLa SillaObservatory (Chile) and at the 358-mTelescopioNazionaleGalileo (TNG) of Roque de los Muchachos observatory (La PalmaSpain) respectively Both instruments provide a resolving powerof Rasymp 115 000 in the wavelength range sim3800 ndash 6900 Aring The ob-
servations were performed as part of the ESO and TNG observingprograms 097C-0948 and A33TAC_15 respectively The exposuretime was set to 1800 sec leading to a SN ratio of sim35 on the ex-tracted spectra We reduced the data using the dedicated HARPSand HARPS-N pipelines and extracted the RVs by cross-correlationwith a G2 numerical mask
The FIES HARPS and HARPS-N RVs are listed in Ta-ble 2 along with the bisector span (BIS) and the full width at halfmaximum (FWHM) of the cross-correlation function (CCF) Timestamps are given in barycentric Julian date in barycentric dynam-ical time (BJDTDB) For the HARPS and HARPS-N data we alsoprovide the Ca ii HampK chromospheric activity index log RprimeHK Wedid not measure log RprimeHK from the FIES spectra because of the poorSN ratio at wavelengths shorter than 4000Aring
5 STELLAR PARAMETERS
51 Spectral analysis
We derived the spectroscopic parameters of K2-139 from the co-added FIES spectra The stacked FIES data have a SN ratio of sim110per pixel at 5500 Aring We adopted three different methods For eachmethod results are reported in Table 3
First method The technique fits spectral features that are sen-sitive to different photospheric parameters It uses the stellar spec-tral synthesis program Spectrum (Gray 1999) to compute syntheticspectra from ATLAS 9model atmospheres (Castelli ampKurucz 2004)Microturbulent (vmic) and macroturbulent (vmac) velocities are de-rived from the calibration equations of Bruntt et al (2010) andDoyle et al (2014) We used the wings of the Hα and Hβ lines toestimate the effective temperature (Teff) and the Mg i 5167 5173and 5184 Aring Ca i 6162 and 6439 Aring and the Na iD lines to determine
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 5
Table 3 Spectroscopic parameters of K2-139 as derived using the three methods described in Sect 5
Method Teff (K) log g (cgs) [FeH] (dex) vmic ( km sminus1) vmac ( km sminus1) v sin i ( km sminus1)
the surface gravity log g We simultaneously fitted different spec-tral regions to measure the iron abundance [FeH] The projectedrotational velocity v sin i was determined by fitting the profile ofmany isolated and unblended metal lines
Second method It relies on the use of the spectral analysispackage Spectroscopy Made Easy (SME Valenti amp Piskunov 1996Valenti amp Fischer 2005) For a set of given stellar parameters SMEcalculates synthetic spectra and fits them to high-resolution ob-served spectra using a chi-squared minimization procedure Weused SME version 443 and ATLAS 12model spectra (Kurucz 2013)We adopted the same calibration equation as described in the firstmethod to determine vmic and vmac Effective temperature is de-rived from the Hα wings log g from the Ca i 6102 6122 6162and 6439 Aring lines [FeH] and v sin i from isolated iron lines
Third method It uses the classical equivalent width (EW)method adopting the following criteria i) Teff is obtained by re-moving trends between abundance of the chemical elements andthe respective excitation potentials ii) log g is optimised by as-suming the ionisation equilibrium condition ie by requiring thatfor a given species the same abundance (within the uncertainties)is obtained from lines of two ionisation states (typically neutral andsingly ionised lines) iii) vmic is set by minimising the slope of therelationship between abundance and the logarithm of the reducedEWs The equivalent widths of Fe i and Fe ii lines are measuredusing the code DOOp (Cantat-Gaudin et al 2014) a wrapper ofDAOSPEC (Stetson amp Pancino 2008) The stellar atmosphere param-eters are derived with the program FAMA (Magrini et al 2013) awrapper of MOOG (Sneden et al 2012) We used the public versionof the atomic data prepared for the Gaia-ESO Survey (Heiter et al2015) and based on the VALD3 data (Ryabchikova et al 2011) Weused sim200 Fe i lines and sim10 Fe ii lines for the determination of thestellar parameters
The three methods provide consistent results within the 1-σerror bars (Table 5) While we have no reason to prefer one tech-nique over the other we adopted the parameter estimates of thefirst method ie Teff = 5340 plusmn 110 K log g = 450 plusmn 009 (cgs)[FeH] = 022 plusmn 008 dex vmic = 09 plusmn 01 km sminus1 vmac = 25 plusmn06 km sminus1 and v sin i = 28 plusmn 06 km sminus1 As a sanity check wealso analyzed the HARPS and HARPS-N data and obtained con-sistent results but with larger error bars owing to the lower SNratio of the co-added HARPS and HARPS-N spectra compared tothat of the co-added FIES data Using the Boyajian et al (2013)rsquoscalibration (see their Table 6) the effective temperature of K2-139defines the spectral type of the host star as K0V
52 Interstellar extinction
We measured the visual reddening (AV) of K2-139 following thetechnique described in Gandolfi et al (2008) We fitted the spectral
energy distribution of the star to synthetic colors extracted from theBT-NEXTGEN model spectrum (Allard et al 2011) with the samephotospheric parameters as the star We adopted the extinction lawof Cardelli et al (1989) and assumed a normal value for the total-to-selective extinction ie RV = AVE(B minus V)= 31 We measureda visual extinction of AV = 007plusmn 005 mag This value is belowthe upper limit of AV 03 mag extracted from the Schlegel et al(1998)rsquos all-sky extinction map corroborating our result
53 Rotational period
The K2 light curve of K2-139 displays periodic and quasi-periodicvariations with a peak-to-peak photometric amplitude of sim2(Fig 1) The late-type spectral type of the star suggests that theobserved variability is due to Sun-like spots appearing and disap-pearing from the visible stellar disc as the star rotates around its axisThis is corroborated by the fact that K2-139 is a chromosphericallyactive star The HARPS and HARPS-N spectra show clear emis-sion components in the cores of the Ca ii HampK lines from whichwe measured an average activity index of log RprimeHK =minus446plusmn 0064
The out-of-transit photometric variability observed in the lightcurve of K2-139 is mainly due to two active regions located at op-posite stellar longitudes whose lifetime is longer than the durationof the K2 observations Using the spots as tracers of stellar rotationand following the auto correlation function (ACF) technique de-scribed in McQuillan et al (2014) we estimated that the rotationalperiod of the star is Prot = 1724 plusmn 012 days The Lomb-Scargleperiodogram of the light curve shows its strongest peak at the sameperiod confirming our results
It is worth noting that the rotation period (Prot = 1724 plusmn012 days) and radius (R= 0862 plusmn 0032 R see next section)of the host star translate into a maximum value for the projected ro-tational velocity of v sin imax = 253plusmn 010 km sminus1 which agreeswith the spectroscopically derived v sin i = 28 plusmn 06 km sminus1 sug-gesting that the star is seen nearly equator-on (i asymp 90) and thatthe system might be aligned along the line-of-sight
54 Stellar mass radius and age
We derived the stellar mass radius and age using the online inter-face for Bayesian estimation of stellar parameters available at thefollowing web page httpstevoapdinafitcgi-binparam Briefly the web tool interpolates onto PARSEC modelisochrones (Bressan et al 2012) the V-band apparent magnitude
4 This value is corrected for the interstellar medium absorption followingthe procedure described in Fossati et al (2017) and using themeasured stellarparameters and reddening The correction is +006 The star is thereforeslightly more active than what measured from the spectra
MNRAS 000 1ndash12 (2017)
6 O Barragaacuten et al
effective temperature metal content and parallax We used the V-band magnitude reported in Table 1 ndash after correcting for interstel-lar reddening (Sect 52) ndash along with the effective temperature andmetal content we derived in Sect 5 The parallax was retrieved fromthe Gaiarsquos first data release (px = 656 plusmn 043mas d = 152plusmn 10 pcFabricius et al 2016) We adopted the log-normal initial mass func-tion from Chabrier (2001)
K2-139 has a mass of M= 0919 plusmn 0033 M and radiusof R= 0862 plusmn 0032 R corresponding to a surface gravity oflog g = 4503plusmn0035 (cgs) in excellent agreement with the spec-troscopically derived value of log g = 450plusmn009 (cgs see Sect 5)The derivedmean density ρ = 202plusmn024 g cmminus3 ofK2-139 is alsoconsistent within 1-σ with the density estimated by the modelingof the transit light curve (ρ = 211+074
minus081 g cmminus3 see Sect 6)The isochrones provide an age of 36plusmn34 Gyr for K2-139 Us-
ing the equations given in Barnes amp Kim (2010) and Barnes (2010)the rotation period of 173 days (Sect 53) implies a gyrochronolog-ical age of 18plusmn 03 Gyr
6 JOINT RV-TRANSIT FIT
We performed the joint fit to the photometric and RV data using thecode pyaneti5 (Barragaacuten et al 2017) a PythonFortran softwaresuite based on Markov Chain Monte Carlo (MCMC) methods
The photometric data included in the joint analysis are subsetsof the whole EVEREST K2 light curve We used the EVERESTlight curve because it provides a slightly better rms over the Vander-burg amp Johnson (2014)rsquos dataWe selectedsim10 hours of data-pointsaround each of the 3 transits which have a duration of sim5 hours Wede-trended each individual transits with the code exotrending6
(Barragaacuten amp Gandolfi 2017) using a second-order polynomial fit-ted to the out-of-transit points The fitted data include 12 pointsimmediately before and after each transit with the exception of thelast transit for which only 9 data points are available We removedthe data points that are affected by stellar spot crossing events (seeSect 71 for more details)
We fitted the RV data using a Keplerian model for the planetalong with two sine-like curves to account for the activity-inducedRV (see next section for details) We adopted the limb-darkenedquadratic law of Mandel amp Agol (2002) for the transit model Weadopted the Gaussian likelihood
L =[
nprodi=1
2π
(σ2i + σ
2j
)minus12]
exp
minus
nsumi=1
12(Di minus Mi)2
σ2i+ σ2
j
(1)
where n is the number of data points σi is the error associatedto each data point Di Mi is the model associated to a given Di andσj is an extra noise term sometime referred as jitter
The sampling method and fitted parameters are the same as inBarragaacuten et al (2016) Details on the adopted priors are given in Ta-ble 5 Following Kipping (2010) we super-sampled the light curvemodel using 10 subsamples perK2 exposure to account for the long-cadence acquisition The parameter space was explored with 500independent chains created randomly inside the prior ranges Thechain convergence was analyzed using the Gelman-Rubin statisticsThe burn-in phase uses 25 000 more iterations with a thin factor
5 Available at httpsgithubcomoscaribvpyaneti6 Available at httpsgithubcomoscaribvexotrending
of 50 The posterior distribution of each parameter has 250 000independent data points
7 RESULTS AND DISCUSSION
71 Stellar activity modeling
A simple Keplerian model provides a poor fit to the RV measure-ments with χ2dof = 61 (Table 4) suggesting that additionalsignals might be present in our Doppler data Activity-induced RVvariation is expected given the 2 peak-to-peak photometric vari-ability observed in the K2 light curve of K2-139 (Fig 1) and theCa ii HampK activity index of log RprimeHK =minus446 plusmn 006 (Sect 53)The K2 photometric variation corresponds to a spot filling factorof approximately 2 if this variation is due to cool starspots Wecan use the empirical relationship relating spot coverage to RV am-plitude from Saar amp Donahue (1997) or Hatzes (2002) to estimatethe RV amplitude expected from spots Using the projected rota-tional velocity of 28 km sminus1 results in an RV semi-amplitude ofasymp20ndash30 m sminus1 The code SOAP2 designed to estimate the effectof active regions on photometric and spectroscopic measurements(Dumusque et al 2014) provides consistent results
In order to look for additional signals in our Doppler data weperformed a frequency analysis of theRVmeasurements and activityindicators On one occasion7 K2-139 was observed with FIES andHARPS-S nearly simultaneously (within less than 25 minutes) Weused the two sets of measurements to estimate the RV FWHMand BIS offsets between the two instruments We assumed no offsetbetween HARPS-N and HARPS While we acknowledge that thisassumption is arbitrary we note that the modeling of the RV datagives an offset of ∆RV(HNminusH) = 0002plusmn 00158 km sminus1 (Table 5)which is consistent with zero
Figure 3 displays the generalized Lomb-Scargle periodograms(Zechmeister amp Kuumlrster 2009) of the combined datasets From topto bottom the RVdata (first panel) the RV residuals after subtractingthe transiting planet signal (second panel) and the BIS (third panel)and FWHM (fourth panel) of the cross correlation function Theperiodogram of the window function is shown in the lower panelThe dotted vertical blue lines mark the frequency at the orbitalperiod of the planet (0035 cd) as well as the frequencies at therotation period of the star (0058 cd) and its first two harmonics(0116 and 0174 cd)
The periodogram of the RV data (upper panel) shows a peak atthe orbital frequency of the planet alongwith two additional peaks at0095 and 0130 cd Since the periodogram of the window functionshows two peaks atsim0060 andsim0095 cd (lower panel red arrows)we interpreted the 0095 and 0130 cd peaks as the aliases of theorbital frequency8 We note also that periodogram of the BIS ofthe CCF displays peaks whose frequencies are close to the stellarrotation frequency and its first two harmonics However none of thepeaks visible in the GLS periodograms of Fig 3 has a false alarmprobability (FAP)9 lower than 5 Although our spectroscopic datashow neither additional signals we note that the semi-amplitudevariation of the BIS and FWHM is expected to be 10-15 m sminus1
(Dumusque et al 2014) which is comparable with the uncertaintiesof most of ourmeasurements (Table 2) The lack of significant peaks
7 Epoch BJD=24575898 0095 = 0035 + 0060 cd and 0130 = 0035 + 0095 cd9 We determined the FAP following the Monte Carlo bootstrap methoddescribed in Kuerster et al (1997)
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 7
Figure 3 Generalized Lomb-Scargle periodogram of the combined FIESHARPS and HARPS-N Doppler datasets From top to bottom the RV datathe RV residuals after subtracting the transiting planet signal the BIS andFWHM of the CCF and the window function The dotted vertical blue linesmark the frequencies at the orbital period as well as at the stellar rotationperiod and its first two harmonics The dashed vertical red lines mark the5 false alarm probabilities as derived using the bootstrap method The redarrows in the lower panel mark the two peaks presented in the main text
in the periodogram of the RV data and RV residuals as well as inthe periodogram of the activity indicators could be explained by thelimited number of available measurements and their uncertaintiesWe conclude that we cannot exclude the existence of spot-inducedsignals in our RV measurements
Photometric and radial velocity variations due to rotationalmodulation can be complex with not only the rotational period Protpresent but also its harmonics eg Prot2 Prot3 Assuming thatthe surface structures responsible for this modulation (eg coolspots) are not evolving rapidly then the simplest representation ofthe rotationalmodulation is through the Fourier components definedby the rotation period and its harmonics Figure 1 shows that theevolution time-scale of the active regions in the stellar surface islonger than the 80-day duration of the K2 campaign Since our RVfollow-up spans 55 days we can assume that any activity-inducedRV signal is coherent within our observing window This approachhas been used previously for other planetary systems orbiting activestars (eg Pepe et al 2013)
The Fourier analysis of the K2 light curve is the best way tomeasure the contribution of the rotation period and its harmonics tothe quasi-periodic photometric variability of the star We thereforeanalyzed the K2 light curve using a pre-whitening procedure Thatis the dominant period was found a sine-fit made to the data andsubtracted and additional periods searched in the residual data
We used the program Period04 (Lenz amp Breger 2005) for thisprocedure
The dominant periods are sim172 days ie the rotation periodof the star (Sect 53) and roughly the first four harmonics (ie 8657 43 and 34 days) The 172- and 86-day periods have about thesame amplitude while the 57-day period (Prot3) has 10 of themain amplitude The Prot4 signal has only about 4 of the mainamplitude The light curve analysis indicates that the signal due torotational modulation can largely be represented by the rotationalperiod (Prot) and its first harmonic (Prot2)
In order to test if the addition of RV sinusoidal signals atthe stellar rotation period and its harmonics can account for theadditional variation seen in our RV measurements we compareddifferent models by adding signals one by one The first model (P0)includes only the planet signal ie a Keplerian model fitted to theRV data using the same priors given in Table 5 but fixing epoch andperiod to the values derived by the transit modeling The next model(P1) is obtained from P0 by adding a sinusoidal signal at the rotationperiod of the star (Prot) Models P2 includes the first harmonic ofthe rotation period (Prot2) whereas model P3 account for the first(Prot2) and second (Prot3) harmonics While adding sinusoidalsignals we fitted for their amplitudes phases and periods We usedflat priors for the phases and amplitudes (details in Table 5) Weused a Gaussian prior for Prot using the value and its uncertaintyderived in Sect 53 The periods of the harmonic signals were leftfree to vary depending on the value assumed by Prot at each stepof the MCMC chains In order to check if the RV variation inducedby the planet is significant in our data set we also performed the fitusing models where the planetary signal was not included (modelsNP1 and NP2 see Table 4)
Table 4 shows the goodness of the fit for each model Thepreferred model is P2 (planet plus 2 sinusoidal signals at Prot andProt2) with the lowest Akaike Information Criteria (AIC) and max-imum likelihood This result is consistent with the Fourier analysisof the K2 light curve which suggests that the major contributionto the photometric variations arises from the stellar rotation periodand its first harmonic Our analysis provides also additional evi-dence that the Doppler motion induced by the planet is present inour RV data set First the planet signal does not significantly varyfor the P0 P1 P2 and P3 models (Table 4) Second the modelswith no planetary signal (NP1 and NP2) provide a poor fit to theRV measurements (Table 4)
To account for additional instrumental noise not included inthe nominal RV error bars andor imperfect treatment of the varioussources of RV variations we fitted for a jitter term for each instru-ment The final parameter estimates and their error bars are listedin Table 5 They are defined as the median and the 68 credibleinterval of the final posterior distributions The best fitting transitand RVmodels are displayed in Figure 4 along with the photometricand RV data points
72 Additional companion
Huang et al (2016) found that warm Jupiters with low eccentricities(e 04) have inner low-mass companions They used this evidenceas an argument in favour of the in situ formation since the planetmigration would have cleaned the warm Jupiter neighborhood Wesearched the light curve for additional transit signals but foundno evidence for an additional transiting planet in the system Asdescribed in the previous paragraph the periodogram of the RVresiduals showno significant peakwith false alarmprobability lowerthan 5
Figure 4 Top FIES (blue circles) HARPS-N (green diamonds) and HARPS (red squares) RV measurements versus time following the subtraction of thesystemic velocities for each instrument The 1σ uncertainties are marked using the same color used for each data-set The vertical gray lines mark the errorbars including jitter The solid line represents the best fitting RV model which includes the planet signal and the activity signal at the stellar rotation periodand its first harmonic The dashed dash-dotted and dotted lines show the RV contribution of K2-139 b stellar rotation and first harmonic respectively Lowerleft panel Transit light curve folded to the orbital period of K2-139 b and residuals The red points mark the K2 data and their error bars The solid line markthe re-binned best-fitting transit model Lower right panel Phase-folded RV curve of K2-139 b and best fitting Keplerian solution (solid line) following thesubtraction of the two additional sinusoidal signals used to account for the stellar activity The FIES HARPS and HARPS-N are corrected for the instrumentoffsets as derived from the global analysis
Table 4 Model comparison
Model Comment Npars Kb (m sminus1) χ2dof(a) ln L AIC(b)
P0 Planet signal 6 291 plusmn 20 61 356 -60P1 Planet signal + 1 sine-curve at Prot 9 294 plusmn 24 34 581 -98
P2 Planet signal + 2 sine-curves at Prot and Prot2 11 273+26minus25 38 601 -98
P3 Planet signal + 3 sine-curves at Prot Prot2 and Prot3 13 278+27minus26 53 593 -93
NP1 1 sine-curve at Prot (No planet signal) 6 0 185 -448 101NP2 2 sine-curves at Prot and Prot2 (No planet signal) 8 0 159 -120 40
Note ndash (a) χ2 value assuming no jitter (b)We used the Akaike Information Criteria (AIC = 2Nparsminus ln 2L) instead of the widely used Bayesian informationcriteria (BIC) because our RV data sample is small (19 data points) and BIC performs better for large samples (Burnham amp Anderson 2002)
73 Spot-crossing events
The passage of a planet in front of a spot can be detected as abump in the transit light curve (see eg Sanchis-Ojeda amp Winn2011) Spot-crossings events are clearly visible in the EVERESTtransit light curves (Fig 4) The same features appear at the sametimes and with consistent amplitudes in the Vanderburg amp Johnson(2014) data confirming that the bumps are real and not due to
systematics To assess whether the bumps significantly affect theparameter estimates we performed the joint analysis as describedin Sect 6 including all the transit data points We found that the finalparameters are consistent within 1-σ with those reported in Table 5
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 9
01 05 1 2 3 4 5 10Mass (MJup)
06
08
10
12
14
Rad
ius
(RJu
p)
H 0 M core 10 M core 25 M core 50 M core100 M core
Figure 5 Warm Jupiters (black squares Mp gt 03 MJup and10 6 Porb 6 100 days) whose mass and radius have been estimated with aprecision of at least 25 (as of January 2017 exoplaneteu) K2-139 b isshown with a filled red circle The solid line corresponds to a planet with apure hydrogen composition (Seager et al 2007) The dashed lines representthe Fortney et al (2007) models for planet core masses of 0 10 25 50and 100 Moplus The vertical dotted line marks the giant planet lower limit asdefined by Hatzes amp Rauer (2015)
74 Planetrsquos composition and formation scenario
With amass of Mp = 0387+0083minus0075 MJ and radius of Rp = 0808+0034
minus0033RJ (resulting in a mean density of ρp=091+024
minus020 g cmminus3) K2-139 bjoins the small group of well characterized warm Jupiters Fig 5shows the position of K2-139 b in themass-radius diagram for warmJupiters (Mp gt 03 MJup 10 6 Porb 6 100 days) whose massand radius have been determined with a precision better than 25(14 objects) Notably K2-139 b is the transiting warm Jupiter withthe lowest mass known to date if the definition of giant planetsgiven by Hatzes amp Rauer (2015) is adopted Fig 5 displays also theplanetary models of Fortney et al (2007) for different core massesand age between 10 and 45Gyrs The planet radius of K2-139 bcan be explained if the planet has a core10 of 49+19
minus17 Moplus containingsim40 of the total planetary mass We expect that K2-139 b has asolid core surrounded by a gaseous envelope
Rafikov (2006) found that a core of mass 5 ndash 20 Moplus at a semi-major axis between 01 and 10AU would be able to start the run-away accretion phase to form a gas giant planet in situ Howeveraccording to his models these kind of cores are unlikely to formowing to the high irradiation coming from the star Boley et al(2016) suggested instead that more massive cores (Mcore amp 20Moplus)can be built up from the merging of tightly packed inner planetsformed at the early stages of the circumstellar disc Batygin et al(2016) found a similar result and argued that the massive core ofHD149026b (Mcore asymp 100Moplus) could be explained by one or moresuper-Earths which merged and accreted the surrounding gas toform a gas-giant planet Huang et al (2016) suggested that thesecores can initiate runaway accretion if they are formed in a re-gion with enough gas around them while those without enoughvolatiles remain super-Earths and represent the population of mas-sive rocky planets unveiled by Kepler around solar-like stars (egDemory 2014) Based on these studies and given the semi-majoraxis of 0179+0021
minus0027 AU the 48 plusmn 14 Moplus core of K2-139 b could
10 Calculated by interpolating Fortney et al (2007)rsquos models
have formed the planet in situ We note that the metallicity of K2-139 is relatively high ([FeH] = 021 plusmn 005) suggesting that theprimordial circumstellar disc had a relatively high content of dustwhich would have enhanced the formation of the core of K2-139 b(see eg Johnson amp Li 2012) Alternatively the planet might haveformed beyond the snow line and migrated inwards via planet-discinteraction (see eg Baruteau et al 2014)
8 CONCLUSIONS
We confirmed the planetary nature and derived the orbital and mainphysical parameters of K2-139 b a warm Jupiter (Teq = 565+48
minus32 K)transiting an active (log RprimeHK = minus446 plusmn 006) K0V star every 29days We measured a planetary mass of Mp = 0387+0083
minus0075 MJ andradius of Rp = 0808+0034
minus0033 RJ At a separation of ap = 0179+0021minus0027
AU the mean density of ρp = 091+024minus020 g cmminus3 implies that the
planet has a core of 49+19minus17 Moplus according to the evolutionary mod-
els of Fortney et al (2007) K2-139 b joins the small group ofwell-characterized warm Jupiters whose mass and radius have beendetermined with a precision better than 25
The spin-orbit angle ie the angle between the spin axis ofthe star and the angular momentum vector of the orbit can pro-vide us with valuable information on the migration mechanismsof exoplanets (see eg Winn 2010 Morton amp Johnson 2011 Al-brecht et al 2012 Gandolfi et al 2012) Currently there are only4 warm Jupiters (Mp gt 03 MJup and 10 6 Porb 6 100 days)with measured obliquity11 From this perspective K2-139 is anideal target to measure the sky-project spin-orbit angle via obser-vations of the Rossiter-McLaughlin (RM) effect Assuming spin-orbit alignement the expected amplitude of the RM anomaly is∆RV asymp
radic1 minus b2 (RpR)2 v sin iasymp 25 m sminus1 (Winn 2010) Given
the brightness of the host star (V = 11653 mag) this amplitudecan easily be measured using state-of-the-art spectrographs suchas HARPSESO-36m Moreover the transit duration (sim5 hours)is shorter than the visibility of K2-139 which is sim9 hours fromLa Silla observatory (altitude higher than 30 above the horizon)
Alternatively the spin-orbit angle could be measured fromthe analysis of the spot-crossing events as described in Sanchis-Ojeda et al (2011) and Sanchis-Ojeda et al (2012) Anomaliesascribable to the passage of K2-139b in front of stellar spots arevisible in the 3 transit light curves observed by K2 Unfortunatelythe limited number of transits and the K2 long cadence data donot allow us to perform a meaningful quantitative analysis of thespot-crossing events Given the amplitude of the detected anoma-lies (sim01) space-based high-precision photometry is needed todetect the spot-crossing events Observations performed with theupcoming CHaracterising ExOPlanets Satellite (CHEOPS Broeget al 2013) would allow us to photometrically determine the spin-orbit angle of this system
ACKNOWLEDGEMENTS
We warmly thank the NOT ESO TNG staff members for theirunique support during the observations We are very thankful toXavier Bonfils Franccedilois Bouchy Martin Kuumlrster Tsevi MazehJorge Melendez and Nuno Santos who kindly agreed to exchange
11 Source httpwww2mpsmpgdehomeshellercontentmain_HRMhtml as of January 2017
MNRAS 000 1ndash12 (2017)
10 O Barragaacuten et al
HARPSandFIES timewith us Special thanks go toAntoninoLanzafor assisting us with the calculation of the gyro-age of the star Wealso greatly thank the anonymous referee for herhis careful re-view and suggestions which helped us to improve the manuscriptD Gandolfi gratefully acknowledges the financial support of theProgramma Giovani Ricercatori ndash Rita Levi Montalcini ndash Rien-tro dei Cervelli (2012) awarded by the Italian Ministry of Edu-cation Universities and Research (MIUR) Sz Csizmadia thanksthe Hungarian OTKA Grant K113117 H J Deeg and D Nespralacknowledge support by grant ESP2015-65712-C5-4-R of the Span-ish Secretary of State for RampDampi (MINECO) D Lorenzo-Oliveiraacknowledges the support from FAPESP (201620667-8) This re-search was supported by the Ministerio de Economia y Competi-tividad under project FIS2012-31079 The research leading to theseresults has received funding from the European Union SeventhFramework Programme (FP72013-2016) under grant agreementNo 312430 (OPTICON) Based on observations obtained a) withthe Nordic Optical Telescope (NOT) operated on the island of LaPalma jointly by Denmark Finland Iceland Norway and Swedenin the Spanish Observatorio del Roque de los Muchachos (ORM)of the Instituto de Astrofiacutesica de Canarias (IAC) b) with the Ital-ian Telescopio Nazionale Galileo (TNG) also operated at the ORM(IAC) on the island of La Palma by the INAF - Fundacioacuten GalileoGalilei c) the 36m ESO telescope at La Silla Observatory underprogramme ID 097C-0948 The data presented here were obtainedin part with ALFOSC which is provided by the Instituto de As-trofisica de Andalucia (IAA) under a joint agreement with the Uni-versity of Copenhagen and NOTSA This paper includes data col-lected by the Kepler mission Funding for the Kepler mission is pro-vided by the NASA Science Mission directorate Some of the datapresented in this paper were obtained from theMikulski Archive forSpace Telescopes (MAST) STScI is operated by the Association ofUniversities for Research in Astronomy Inc under NASA contractNAS5-26555 Support for MAST for non-HST data is provided bythe NASA Office of Space Science via grant NNX09AF08G and byother grants and contracts MF and CMP acknowledge generoussupport from the Swedish National Space Board C Eiroa and IRebollido are supported by Spanish grant AYA2014-55840-P PDacknowledge the support from INAF and Ministero dellrsquoIstruzionedellrsquoUniversitagrave e della Ricerca (MIUR) in the form of the grantldquoPremiale VLT 2012rdquo and ldquoThe Chemical and Dynamical Evolu-tion of the Milky Way and Local Group Galaxiesrdquo This work hasmade use of data from the European Space Agency (ESA) missionGaia (httpwwwcosmosesaintgaia) processed by theGaia Data Processing and Analysis Consortium (DPAC httpwwwcosmosesaintwebgaiadpacconsortium) Fundingfor the DPAC has been provided by national institutions in particu-lar the institutions participating in theGaiaMultilateral Agreement
REFERENCES
Albrecht S et al 2012 ApJ 757 18Allard F Homeier D Freytag B 2011 in Johns-Krull C BrowningM K
West A A eds Astronomical Society of the Pacific Conference SeriesVol 448 16th Cambridge Workshop on Cool Stars Stellar Systemsand the Sun p 91 (arXiv10115405)
Antonini F Hamers A S Lithwick Y 2016 preprint(arXiv160401781)
Barnes S A 2010 ApJ 722 222Barnes S A Kim Y-C 2010 ApJ 721 675Barragaacuten O Gandolfi D 2017 Exotrending Astrophysics Source Code
Library (ascl1706001)
Barragaacuten O et al 2016 AJ 152 193Barragaacuten O Gandolfi D Antoniciello G 2017 pyaneti Astrophysics
Source Code Library (ascl1707003)Baruteau C et al 2014 Protostars and Planets VI pp 667ndash689Batygin K Bodenheimer P H Laughlin G P 2016 ApJ 829 114Boley A C Granados Contreras A P Gladman B 2016 ApJ 817 L17Boyajian T S et al 2013 ApJ 771 40Brahm R et al 2016 AJ 151 89Bressan A Marigo P Girardi L Salasnich B Dal Cero C Rubele S
Nanni A 2012 MNRAS 427 127Broeg C et al 2013 in European Physical Journal Web of Conferences p
03005 (arXiv13052270) doi101051epjconf20134703005Bruntt H et al 2010 MNRAS 405 1907Buchhave L A et al 2010 ApJ 720 1118Burnham K Anderson D 2002 Model Selection and Multimodel Infer-
ence A Practical Information-Theoretic Approach NewYork Springer-Verlag
Cabrera J et al 2009 AampA 506 501Cabrera J Csizmadia S Erikson A Rauer H Kirste S 2012 AampA 548
A44Cabrera J et al 2014 ApJ 781 18Cantat-Gaudin T et al 2014 AampA 562 A10Cardelli J A Clayton G C Mathis J S 1989 ApJ 345 245Carone L et al 2012 AampA 538 A112Carpano S et al 2009 AampA 506 491Castelli F Kurucz R L 2004 preprintCavarroc C et al 2012 ApampSS 337 511Chabrier G 2001 ApJ 554 1274Cosentino R et al 2012 in Ground-based and Airborne Instrumentation
for Astronomy IV p 84461V doi10111712925738Cutri R M et al 2003 2MASS All Sky Catalog of point sourcesCutri R M et al 2012 Technical report Explanatory Supplement to the
WISE All-Sky Data Release ProductsDawson R I Johnson J A Morton T D Crepp J R Fabrycky D C
Murray-Clay R A Howard A W 2012 ApJ 761 163Deeg H J et al 2010 Nature 464 384Demory B-O 2014 ApJ 789 L20Dong S Katz B Socrates A 2014 ApJ 781 L5Doyle A P Davies G R Smalley B Chaplin W J Elsworth Y 2014
MNRAS 444 3592Dumusque X Boisse I Santos N C 2014 ApJ 796 132Endl M Cochran W D 2016 PASP 128 094502Erikson A et al 2012 AampA 539 A14Fabricius C et al 2016 AampA 595 A3Fortney J J Marley M S Barnes J W 2007 ApJ 659 1661Fossati et al 2017 AampA submittedFrandsenS LindbergB 1999 inKarttunenH PiirolaV edsAstrophysics
with the NOT p 71Frewen S F N Hansen B M S 2016 MNRAS 455 1538Gandolfi D et al 2008 ApJ 687 1303Gandolfi D et al 2012 AampA 543 L5Gandolfi D et al 2015 AampA 576 A11Gray R O 1999 SPECTRUM A stellar spectral synthesis program As-
trophysics Source Code Library (ascl9910002)Grziwa S Paumltzold M 2016 preprint (arXiv160708417)Grziwa S Paumltzold M Carone L 2012 MNRAS 420 1045Hamers A S Antonini F Lithwick Y Perets H B Portegies Zwart S F
2016 preprint (arXiv160607438)Hatzes A P 2002 Astronomische Nachrichten 323 392Hatzes A P Rauer H 2015 ApJ 810 L25Heiter U et al 2015 Phys Scr 90 054010Huang C Wu Y Triaud A H M J 2016 ApJ 825 98Jenkins J S et al 2017 MNRAS 466 443Johnson J L Li H 2012 ApJ 751 81Kipping D M 2010 MNRAS 408 1758Kley W Nelson R P 2012 ARAampA 50 211Kovaacutecs G Zucker S Mazeh T 2002 AampA 391 369
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 11
Table 5 K2-139 system parameters
Parameter Prior(a) Final value
Stellar parametersStar mass M (M) middot middot middot 0919 plusmn 0033Star radius R (R) middot middot middot 0862 plusmn 0032Star density ρ (from spectroscopy g cmminus3) middot middot middot 202+025
minus022
Star density ρ (from light curve g cmminus3) middot middot middot 211+074minus081
Effective Temperature Teff (K) middot middot middot 5340 plusmn 110Surface gravity log g (cgs) middot middot middot 450 plusmn 009Iron abundance [FeH] (dex) middot middot middot 022 plusmn 008Microturbulent velocity vmic ( km sminus1) middot middot middot 09 plusmn 01Macroturbulent velocity vmac ( km sminus1) middot middot middot 25 plusmn 06Projected rotational velocity v sin i ( km sminus1) middot middot middot 28 plusmn 06Rotational period Prot (days) middot middot middot 1724 plusmn 012Activity index(b) logRprimeHK middot middot middot minus446 plusmn 006Gyrochronological age (Gyr) middot middot middot 18 plusmn 03Interstellar extinction AV (mag) middot middot middot 007 plusmn 005Star distance d (pc) middot middot middot 152 plusmn 10
Model parameters of K2-139 bOrbital period Porb (days) U[283773 283873] 2838236 plusmn 000026Transit epoch T0 (BJDTDBminus2 450 000) U[73258120 73258220] 732581714 plusmn 000033Scaled semi-major axis aR U[12 100] 448+47
minus67Planet-to-star radius ratio RpR U[0 02] 00961+00023
minus00015Impact parameter b U[0 12] 030+021
minus019radice sinω U[minus1 1](c) 010+029
minus030radice cosω U[minus1 1](c) 006+024
minus027Radial velocity semi-amplitude variation K ( m sminus1) U[0 200] 277+60
Note ndash The adopted Sun and Jupiter units follow the recommendations from the International Astronomical Union (Prša et al 2016) (a) U[a b] refersto uniform priors between a and b N[a b] means Gaussian priors with mean a and standard deviation b and F[a] to a fixed a value (b) Corrected forinterstellar reddening following Fossati et al (2017) The correction is +006 (c) The code always ensures that e lt 1 (d) Assuming albedo = 0
MNRAS 000 1ndash12 (2017)
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
MNRAS 000 1ndash12 (2017)
1 Introduction
2 K2 photometry
3 ALFOSC imaging
4 High-resolution spectroscopy
5 Stellar parameters
51 Spectral analysis
52 Interstellar extinction
53 Rotational period
54 Stellar mass radius and age
6 Joint RV-transit fit
7 Results and discussion
71 Stellar activity modeling
72 Additional companion
73 Spot-crossing events
74 Planets composition and formation scenario
8 Conclusions
Acknowledgements
2 O Barragaacuten et al
Jupiters Theymark the transition between hot Jupiters (giant planetswith orbital period between sim1 and 10 days) and Jupiter analogues(orbital period longer than 100 days) They seem to be less com-mon than hot Jupiters and their formation scenario is still underdebate (eg Frewen amp Hansen 2016 Boley et al 2016) Whereasit is commonly accepted that hot Jupiters did not form in situ (egKley amp Nelson 2012) but rather formed beyond the snow line andthen migrated inwards to their current position it has been recentlyproposed that warm Jupiters might have formed in situ (eg Boleyet al 2016 Huang et al 2016)
Eighty warm Jupiters have been discovered so far from bothground- (eg da Silva et al 2007 Brahm et al 2016 Jenkins et al2017) and space-based surveys (eg Deeg et al 2010 Saad-Oliveraet al 2017 Smith et al 2017) About thirty are known to transittheir parent star and only thirteen have masses and radii known witha precision better than 251 They have been detected both in low-eccentricity orbits (e 04 eg Brahm et al 2016 Niedzielski et al2016 Smith et al 2017) as well as in highly eccentric orbits (egDawson et al 2012 Ortiz et al 2015) Dong et al (2014) foundthat warm Jupiters with high eccentricities (e amp 04) tend to havea massive planetarystellar companion in a long period orbit Thearchitectures of these systems suggest that eccentric warm Jupitersmight have reached their current positions via high-eccentricitymigration excited by the outer companion (Dong et al 2014) On theother hand warm Jupiters with no detected Jovian companion tendto have lower eccentricities peaked around 02 This suggests thattwo different types of warm Jupiters might exist those formed viahigh-eccentricity migration and those formed in situ Alternativelywarm Jupiters in low-eccentricity orbits can also result from disc-drivenmigration from the outer region of the system (KleyampNelson2012)
Petrovich amp Tremaine (2016) studied the possibility that warmJupiters are undergoing secular eccentricity oscillations induced byan outer companion in an eccentric andor mutually inclined orbitTheir model suggests that high-eccentricity migration can accountfor most of the hot Jupiters as well as for most of the warm Jupiterswith e amp 04 However it cannot account for the remaining popula-tion of low-eccentricity warm Jupiters which must have undergonea different formation mechanism The low efficiency to generatewarm Jupiters in nearly circular orbits via high-eccentricity migra-tion has been corroborated by Hamers et al (2016) and Antoniniet al (2016) using numerical simulations
In order to test different planet formationmechanisms we needto characterize the population of warm Jupiters in terms of planetarymass radius and orbital parametersWe herein present the discoveryof K2-139 b (EPIC 218916923 b) a transiting warm Jupiter (Mp =
0387+0083minus0075 MJ Rp = 0808+0034
minus0033 RJ) in a 29-day orbit aroundan active K0V star that has been photometrically monitored bythe K2 space-mission during its Campaign 7 We combine the K2photometry with ground-based imaging and high-precision radialvelocity measurements to confirm the planet and derive the mainparameters of the system
2 K2 PHOTOMETRY
K2 Campaign 7 was performed between 2015 October 04 UT and2015 December 26 UT2 The Kepler spacecraft was pointed at
1 Source httpexoplaneteu as of January 20172 See httpkeplersciencearcnasagovk2-fieldshtml
Table 1 Main identifiers coordinates optical and infrared magnitudes andproper motion of K2-139
Parameter Value Source
Main IdentifiersTYC 6300-2008-1 TychoEPIC 218916923 EPICUCAC 361-185490 EPIC2MASS 19161596-1754384 EPIC
Note ndash Values of fields marked with EPIC are taken from the EclipticPlane Input Catalog available at httparchivestscieduk2epicsearchphp Values marked with Gaia 2MASS and WISE arefrom Fabricius et al (2016) Cutri et al (2003) and Cutri et al (2012)respectively The WISE W4 magnitude is an upper limit
coordinates α = 19h11m19s δ = minus2321prime36primeprime K2 observed si-multaneously 13 469 sources in long cadence mode (sim30 minuteintegration time) and 72 objects in short cadence mode (sim1 minuteintegration time) leading to a total of 13 541 light curves
For the detection of transiting planet candidates we used theK2 Campaign 7 light curves3 extracted by Vanderburg amp John-son (2014) We analyzed the light curves using the DST algorithm(Cabrera et al 2012) and the EXOTRANS pipeline (Grziwa et al 2012Grziwa amp Paumltzold 2016) Both codes have been used extensively onCoRoT (Carpano et al 2009 Cabrera et al 2009 Erikson et al2012 Carone et al 2012 Cavarroc et al 2012) andKepler (Cabreraet al 2014 Grziwa amp Paumltzold 2016) data These search algorithmsdetect periodic patterns in time series photometric data DST usesan optimized transit shape with the same number of free parametersas for the BLS algorithm (Box-fitting Least Squares Kovaacutecs et al2002) and it also implements better statistics for signal detectionEXOTRANS uses the BLS algorithm combined with the wavelet-basedfilter technique VARLET (Grziwa amp Paumltzold 2016) diminishing theeffects of stellar variability and data discontinuities
We detected a periodic transit-like signal associated with
3 Publicly available at httpswwwcfaharvardedu~avanderballk2c7obshtml
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 3
2480 2500 2520 2540BJD shy 2454833
099
100
101
Rel
ativ
e flu
x
Figure 1 K2 Light curve for K2-139 as extracted by Luger et al (2016) The positions of the 3 observed transits are marked with vertical dashed lines
the star EPIC 218916923 with both DST and EXOTRANS Asa sanity check we downloaded the EVEREST light curve ofEPIC 218916923 (Luger et al 2016) and detected the same sig-nal We note that Vanderburg amp Johnson (2014) and Luger et al(2016) used the same mask to extract the time-series data from therawK2 images EPIC 218916923 was proposed forK2 observationsby programs GO7086 (PI Thompson) GO7030 (PI Howard) andGO7087 (PI Dragomir) We will hereafter refer to the star and itstransiting planet as K2-139 and K2-139 b respectively
We searched the Vanderburg amp Johnson (2014)rsquos light curvefor odd-even transit depth variation and secondary eclipse thatmighthint to a binary scenario making the system a likely false positiveNone of themwere significantly detected The depth of the oddeventransits agrees within the 1-σ uncertainty of 16 times 10minus3 whereasthe 3-σ upper limit of the occultation depth is 79 times 10minus5 both re-spect to the normalized flux We proceeded to more detailed fittingof the light curve as well as ground-based imaging (Sect 3) andspectroscopic observations (Sect 4) The main identifiers coordi-nates optical and infrared magnitudes and proper motions of thestar are listed in Table 1 We display the EVEREST K2 light curveof K2-139 in Fig 1
3 ALFOSC IMAGING
K2 Campaign 7 is projected close to the galactic center and thus ina relatively crowded stellar region In order to estimate the contam-ination factor arising from sources whose light leaks into the photo-metricmasks used byVanderburgamp Johnson (2014) and Luger et al(2016) we observed K2-139 on 13 September 2016 (UT) with theALFOSC camera mounted at the Nordic Optical Telescope (NOT)of Roque de los Muchachos Observatory (La Palma Spain) Thesky conditions were photometric with excellent seeing conditions(sim06primeprime) We used the Bessel R-filter and acquired 16 images of 6sec 2 images of 20 sec and 1 image of 120 sec The data were biassubtracted and flat-fielded using dusk sky flats Aperture photom-etry was then performed on all stars within the mask used in theextraction of the light curve by Vanderburg amp Johnson (2014) andLuger et al (2016)
Several fainter stars can be identified inside the photometricmask (Fig 2) of which the two brightest sources are also in theEPIC catalog with Kepler band magnitudes of 168 and 184 Theclosest detected source is a 68-mag fainter star at 38primeprime South of K2-
Figure 2 ALFOSC Bessel R-band image of the sky region around K2-139North is up and East is to the left The target star is the brightest sourcein the middle The solid black polygon marks the EVEREST photometricmask (Luger et al 2016)
139 We can exclude stars as faint as sim20mag at an angular distancelarger than sim06primeprime from K2-139 It is worth noting that the fainteststar whose flux could account for the sim1 deep transit of K2-139cannot be more than sim5 mag fainter than our target The summedflux of these faint stars amounts to 14plusmn03 of the total off-transitflux within the apertureWe subtracted this contamination flux fromthe EVEREST K2 light curve prior to performing the joint analysispresented in Sect 6
4 HIGH-RESOLUTION SPECTROSCOPY
In June and August 2016 we obtained two reconnaissance spectraof K2-139 with the Tull spectrograph (Tull et al 1995) at the 27-mtelescope at McDonald Observatory (Texas USA) The high res-olution (R asymp 60 000) spectra have a signal-to-noise ratio of sim30per pixel at 5500Aring We reduced the data using standard IRAF rou-tines and derived preliminary spectroscopic parameters using ourcode Kea (Endl amp Cochran 2016) The results from both spec-
MNRAS 000 1ndash12 (2017)
4 O Barragaacuten et al
Table 2 Radial velocity measurements and activity indexes of K2-139
tra are nearly identical and reveal a star with Teff = 5500plusmn 100 Klog g = 465plusmn 012 (cgs) [FeH]=+011plusmn 012 dex and a slow pro-jected rotational velocity of v sin iasymp 2 km sminus1
The high-precision radial velocity follow-up of K2-139 wasstarted in June 2016with the FIbre-fed Eacutechelle Spectrograph (FIESFrandsenampLindberg 1999 Telting et al 2014)mounted at the 256-m Nordic Optical Telescope (NOT) The observations were carriedout as part of the OPTICON andCAT observing programs 16A055P53-201 and P53-203 We used the high-resmode which providesa resolving power of Rasymp 67 000 in the whole visible spectral range(3700 minus 7300Aring) The exposure time was set to 2100 ndash 3600 secbased on sky conditions and observing scheduling constraints Fol-lowing the observing strategy outlined in Buchhave et al (2010)and Gandolfi et al (2015) we traced the RV drift of the instrumentby acquiring long-exposed (Texp asymp 35 sec) ThAr spectra immedi-ately before and after the target observations The typical RV driftmeasured between two ThAr spectra bracketing a 2100 ndash 3600 secscience exposure is about 50 ndash 80 m sminus1 A linear interpolation ofthe RV drift to the mid-time of the science exposure allows us toachieve a radial velocity zero-point stability of about 5 ndash 6 m sminus1which is 2 ndash 3 times smaller than the nominal error bars listed in Ta-ble 2 The data reduction uses standard IRAF and IDL routines Thesignal-to-noise (SN) ratio of the extracted spectra is sim30 ndash 40 perpixel at 5500 Aring Radial velocity measurements were extracted viamulti-order cross-correlation with the RV standard star HD182572observed with the same instrument set-up as K2-139
We also observed K2-139 in July August and September 2016with the HARPS (Mayor et al 2003) and HARPS-N (Cosentinoet al 2012) spectrographs mounted at the ESO 36-m Telescope ofLa SillaObservatory (Chile) and at the 358-mTelescopioNazionaleGalileo (TNG) of Roque de los Muchachos observatory (La PalmaSpain) respectively Both instruments provide a resolving powerof Rasymp 115 000 in the wavelength range sim3800 ndash 6900 Aring The ob-
servations were performed as part of the ESO and TNG observingprograms 097C-0948 and A33TAC_15 respectively The exposuretime was set to 1800 sec leading to a SN ratio of sim35 on the ex-tracted spectra We reduced the data using the dedicated HARPSand HARPS-N pipelines and extracted the RVs by cross-correlationwith a G2 numerical mask
The FIES HARPS and HARPS-N RVs are listed in Ta-ble 2 along with the bisector span (BIS) and the full width at halfmaximum (FWHM) of the cross-correlation function (CCF) Timestamps are given in barycentric Julian date in barycentric dynam-ical time (BJDTDB) For the HARPS and HARPS-N data we alsoprovide the Ca ii HampK chromospheric activity index log RprimeHK Wedid not measure log RprimeHK from the FIES spectra because of the poorSN ratio at wavelengths shorter than 4000Aring
5 STELLAR PARAMETERS
51 Spectral analysis
We derived the spectroscopic parameters of K2-139 from the co-added FIES spectra The stacked FIES data have a SN ratio of sim110per pixel at 5500 Aring We adopted three different methods For eachmethod results are reported in Table 3
First method The technique fits spectral features that are sen-sitive to different photospheric parameters It uses the stellar spec-tral synthesis program Spectrum (Gray 1999) to compute syntheticspectra from ATLAS 9model atmospheres (Castelli ampKurucz 2004)Microturbulent (vmic) and macroturbulent (vmac) velocities are de-rived from the calibration equations of Bruntt et al (2010) andDoyle et al (2014) We used the wings of the Hα and Hβ lines toestimate the effective temperature (Teff) and the Mg i 5167 5173and 5184 Aring Ca i 6162 and 6439 Aring and the Na iD lines to determine
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 5
Table 3 Spectroscopic parameters of K2-139 as derived using the three methods described in Sect 5
Method Teff (K) log g (cgs) [FeH] (dex) vmic ( km sminus1) vmac ( km sminus1) v sin i ( km sminus1)
the surface gravity log g We simultaneously fitted different spec-tral regions to measure the iron abundance [FeH] The projectedrotational velocity v sin i was determined by fitting the profile ofmany isolated and unblended metal lines
Second method It relies on the use of the spectral analysispackage Spectroscopy Made Easy (SME Valenti amp Piskunov 1996Valenti amp Fischer 2005) For a set of given stellar parameters SMEcalculates synthetic spectra and fits them to high-resolution ob-served spectra using a chi-squared minimization procedure Weused SME version 443 and ATLAS 12model spectra (Kurucz 2013)We adopted the same calibration equation as described in the firstmethod to determine vmic and vmac Effective temperature is de-rived from the Hα wings log g from the Ca i 6102 6122 6162and 6439 Aring lines [FeH] and v sin i from isolated iron lines
Third method It uses the classical equivalent width (EW)method adopting the following criteria i) Teff is obtained by re-moving trends between abundance of the chemical elements andthe respective excitation potentials ii) log g is optimised by as-suming the ionisation equilibrium condition ie by requiring thatfor a given species the same abundance (within the uncertainties)is obtained from lines of two ionisation states (typically neutral andsingly ionised lines) iii) vmic is set by minimising the slope of therelationship between abundance and the logarithm of the reducedEWs The equivalent widths of Fe i and Fe ii lines are measuredusing the code DOOp (Cantat-Gaudin et al 2014) a wrapper ofDAOSPEC (Stetson amp Pancino 2008) The stellar atmosphere param-eters are derived with the program FAMA (Magrini et al 2013) awrapper of MOOG (Sneden et al 2012) We used the public versionof the atomic data prepared for the Gaia-ESO Survey (Heiter et al2015) and based on the VALD3 data (Ryabchikova et al 2011) Weused sim200 Fe i lines and sim10 Fe ii lines for the determination of thestellar parameters
The three methods provide consistent results within the 1-σerror bars (Table 5) While we have no reason to prefer one tech-nique over the other we adopted the parameter estimates of thefirst method ie Teff = 5340 plusmn 110 K log g = 450 plusmn 009 (cgs)[FeH] = 022 plusmn 008 dex vmic = 09 plusmn 01 km sminus1 vmac = 25 plusmn06 km sminus1 and v sin i = 28 plusmn 06 km sminus1 As a sanity check wealso analyzed the HARPS and HARPS-N data and obtained con-sistent results but with larger error bars owing to the lower SNratio of the co-added HARPS and HARPS-N spectra compared tothat of the co-added FIES data Using the Boyajian et al (2013)rsquoscalibration (see their Table 6) the effective temperature of K2-139defines the spectral type of the host star as K0V
52 Interstellar extinction
We measured the visual reddening (AV) of K2-139 following thetechnique described in Gandolfi et al (2008) We fitted the spectral
energy distribution of the star to synthetic colors extracted from theBT-NEXTGEN model spectrum (Allard et al 2011) with the samephotospheric parameters as the star We adopted the extinction lawof Cardelli et al (1989) and assumed a normal value for the total-to-selective extinction ie RV = AVE(B minus V)= 31 We measureda visual extinction of AV = 007plusmn 005 mag This value is belowthe upper limit of AV 03 mag extracted from the Schlegel et al(1998)rsquos all-sky extinction map corroborating our result
53 Rotational period
The K2 light curve of K2-139 displays periodic and quasi-periodicvariations with a peak-to-peak photometric amplitude of sim2(Fig 1) The late-type spectral type of the star suggests that theobserved variability is due to Sun-like spots appearing and disap-pearing from the visible stellar disc as the star rotates around its axisThis is corroborated by the fact that K2-139 is a chromosphericallyactive star The HARPS and HARPS-N spectra show clear emis-sion components in the cores of the Ca ii HampK lines from whichwe measured an average activity index of log RprimeHK =minus446plusmn 0064
The out-of-transit photometric variability observed in the lightcurve of K2-139 is mainly due to two active regions located at op-posite stellar longitudes whose lifetime is longer than the durationof the K2 observations Using the spots as tracers of stellar rotationand following the auto correlation function (ACF) technique de-scribed in McQuillan et al (2014) we estimated that the rotationalperiod of the star is Prot = 1724 plusmn 012 days The Lomb-Scargleperiodogram of the light curve shows its strongest peak at the sameperiod confirming our results
It is worth noting that the rotation period (Prot = 1724 plusmn012 days) and radius (R= 0862 plusmn 0032 R see next section)of the host star translate into a maximum value for the projected ro-tational velocity of v sin imax = 253plusmn 010 km sminus1 which agreeswith the spectroscopically derived v sin i = 28 plusmn 06 km sminus1 sug-gesting that the star is seen nearly equator-on (i asymp 90) and thatthe system might be aligned along the line-of-sight
54 Stellar mass radius and age
We derived the stellar mass radius and age using the online inter-face for Bayesian estimation of stellar parameters available at thefollowing web page httpstevoapdinafitcgi-binparam Briefly the web tool interpolates onto PARSEC modelisochrones (Bressan et al 2012) the V-band apparent magnitude
4 This value is corrected for the interstellar medium absorption followingthe procedure described in Fossati et al (2017) and using themeasured stellarparameters and reddening The correction is +006 The star is thereforeslightly more active than what measured from the spectra
MNRAS 000 1ndash12 (2017)
6 O Barragaacuten et al
effective temperature metal content and parallax We used the V-band magnitude reported in Table 1 ndash after correcting for interstel-lar reddening (Sect 52) ndash along with the effective temperature andmetal content we derived in Sect 5 The parallax was retrieved fromthe Gaiarsquos first data release (px = 656 plusmn 043mas d = 152plusmn 10 pcFabricius et al 2016) We adopted the log-normal initial mass func-tion from Chabrier (2001)
K2-139 has a mass of M= 0919 plusmn 0033 M and radiusof R= 0862 plusmn 0032 R corresponding to a surface gravity oflog g = 4503plusmn0035 (cgs) in excellent agreement with the spec-troscopically derived value of log g = 450plusmn009 (cgs see Sect 5)The derivedmean density ρ = 202plusmn024 g cmminus3 ofK2-139 is alsoconsistent within 1-σ with the density estimated by the modelingof the transit light curve (ρ = 211+074
minus081 g cmminus3 see Sect 6)The isochrones provide an age of 36plusmn34 Gyr for K2-139 Us-
ing the equations given in Barnes amp Kim (2010) and Barnes (2010)the rotation period of 173 days (Sect 53) implies a gyrochronolog-ical age of 18plusmn 03 Gyr
6 JOINT RV-TRANSIT FIT
We performed the joint fit to the photometric and RV data using thecode pyaneti5 (Barragaacuten et al 2017) a PythonFortran softwaresuite based on Markov Chain Monte Carlo (MCMC) methods
The photometric data included in the joint analysis are subsetsof the whole EVEREST K2 light curve We used the EVERESTlight curve because it provides a slightly better rms over the Vander-burg amp Johnson (2014)rsquos dataWe selectedsim10 hours of data-pointsaround each of the 3 transits which have a duration of sim5 hours Wede-trended each individual transits with the code exotrending6
(Barragaacuten amp Gandolfi 2017) using a second-order polynomial fit-ted to the out-of-transit points The fitted data include 12 pointsimmediately before and after each transit with the exception of thelast transit for which only 9 data points are available We removedthe data points that are affected by stellar spot crossing events (seeSect 71 for more details)
We fitted the RV data using a Keplerian model for the planetalong with two sine-like curves to account for the activity-inducedRV (see next section for details) We adopted the limb-darkenedquadratic law of Mandel amp Agol (2002) for the transit model Weadopted the Gaussian likelihood
L =[
nprodi=1
2π
(σ2i + σ
2j
)minus12]
exp
minus
nsumi=1
12(Di minus Mi)2
σ2i+ σ2
j
(1)
where n is the number of data points σi is the error associatedto each data point Di Mi is the model associated to a given Di andσj is an extra noise term sometime referred as jitter
The sampling method and fitted parameters are the same as inBarragaacuten et al (2016) Details on the adopted priors are given in Ta-ble 5 Following Kipping (2010) we super-sampled the light curvemodel using 10 subsamples perK2 exposure to account for the long-cadence acquisition The parameter space was explored with 500independent chains created randomly inside the prior ranges Thechain convergence was analyzed using the Gelman-Rubin statisticsThe burn-in phase uses 25 000 more iterations with a thin factor
5 Available at httpsgithubcomoscaribvpyaneti6 Available at httpsgithubcomoscaribvexotrending
of 50 The posterior distribution of each parameter has 250 000independent data points
7 RESULTS AND DISCUSSION
71 Stellar activity modeling
A simple Keplerian model provides a poor fit to the RV measure-ments with χ2dof = 61 (Table 4) suggesting that additionalsignals might be present in our Doppler data Activity-induced RVvariation is expected given the 2 peak-to-peak photometric vari-ability observed in the K2 light curve of K2-139 (Fig 1) and theCa ii HampK activity index of log RprimeHK =minus446 plusmn 006 (Sect 53)The K2 photometric variation corresponds to a spot filling factorof approximately 2 if this variation is due to cool starspots Wecan use the empirical relationship relating spot coverage to RV am-plitude from Saar amp Donahue (1997) or Hatzes (2002) to estimatethe RV amplitude expected from spots Using the projected rota-tional velocity of 28 km sminus1 results in an RV semi-amplitude ofasymp20ndash30 m sminus1 The code SOAP2 designed to estimate the effectof active regions on photometric and spectroscopic measurements(Dumusque et al 2014) provides consistent results
In order to look for additional signals in our Doppler data weperformed a frequency analysis of theRVmeasurements and activityindicators On one occasion7 K2-139 was observed with FIES andHARPS-S nearly simultaneously (within less than 25 minutes) Weused the two sets of measurements to estimate the RV FWHMand BIS offsets between the two instruments We assumed no offsetbetween HARPS-N and HARPS While we acknowledge that thisassumption is arbitrary we note that the modeling of the RV datagives an offset of ∆RV(HNminusH) = 0002plusmn 00158 km sminus1 (Table 5)which is consistent with zero
Figure 3 displays the generalized Lomb-Scargle periodograms(Zechmeister amp Kuumlrster 2009) of the combined datasets From topto bottom the RVdata (first panel) the RV residuals after subtractingthe transiting planet signal (second panel) and the BIS (third panel)and FWHM (fourth panel) of the cross correlation function Theperiodogram of the window function is shown in the lower panelThe dotted vertical blue lines mark the frequency at the orbitalperiod of the planet (0035 cd) as well as the frequencies at therotation period of the star (0058 cd) and its first two harmonics(0116 and 0174 cd)
The periodogram of the RV data (upper panel) shows a peak atthe orbital frequency of the planet alongwith two additional peaks at0095 and 0130 cd Since the periodogram of the window functionshows two peaks atsim0060 andsim0095 cd (lower panel red arrows)we interpreted the 0095 and 0130 cd peaks as the aliases of theorbital frequency8 We note also that periodogram of the BIS ofthe CCF displays peaks whose frequencies are close to the stellarrotation frequency and its first two harmonics However none of thepeaks visible in the GLS periodograms of Fig 3 has a false alarmprobability (FAP)9 lower than 5 Although our spectroscopic datashow neither additional signals we note that the semi-amplitudevariation of the BIS and FWHM is expected to be 10-15 m sminus1
(Dumusque et al 2014) which is comparable with the uncertaintiesof most of ourmeasurements (Table 2) The lack of significant peaks
7 Epoch BJD=24575898 0095 = 0035 + 0060 cd and 0130 = 0035 + 0095 cd9 We determined the FAP following the Monte Carlo bootstrap methoddescribed in Kuerster et al (1997)
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 7
Figure 3 Generalized Lomb-Scargle periodogram of the combined FIESHARPS and HARPS-N Doppler datasets From top to bottom the RV datathe RV residuals after subtracting the transiting planet signal the BIS andFWHM of the CCF and the window function The dotted vertical blue linesmark the frequencies at the orbital period as well as at the stellar rotationperiod and its first two harmonics The dashed vertical red lines mark the5 false alarm probabilities as derived using the bootstrap method The redarrows in the lower panel mark the two peaks presented in the main text
in the periodogram of the RV data and RV residuals as well as inthe periodogram of the activity indicators could be explained by thelimited number of available measurements and their uncertaintiesWe conclude that we cannot exclude the existence of spot-inducedsignals in our RV measurements
Photometric and radial velocity variations due to rotationalmodulation can be complex with not only the rotational period Protpresent but also its harmonics eg Prot2 Prot3 Assuming thatthe surface structures responsible for this modulation (eg coolspots) are not evolving rapidly then the simplest representation ofthe rotationalmodulation is through the Fourier components definedby the rotation period and its harmonics Figure 1 shows that theevolution time-scale of the active regions in the stellar surface islonger than the 80-day duration of the K2 campaign Since our RVfollow-up spans 55 days we can assume that any activity-inducedRV signal is coherent within our observing window This approachhas been used previously for other planetary systems orbiting activestars (eg Pepe et al 2013)
The Fourier analysis of the K2 light curve is the best way tomeasure the contribution of the rotation period and its harmonics tothe quasi-periodic photometric variability of the star We thereforeanalyzed the K2 light curve using a pre-whitening procedure Thatis the dominant period was found a sine-fit made to the data andsubtracted and additional periods searched in the residual data
We used the program Period04 (Lenz amp Breger 2005) for thisprocedure
The dominant periods are sim172 days ie the rotation periodof the star (Sect 53) and roughly the first four harmonics (ie 8657 43 and 34 days) The 172- and 86-day periods have about thesame amplitude while the 57-day period (Prot3) has 10 of themain amplitude The Prot4 signal has only about 4 of the mainamplitude The light curve analysis indicates that the signal due torotational modulation can largely be represented by the rotationalperiod (Prot) and its first harmonic (Prot2)
In order to test if the addition of RV sinusoidal signals atthe stellar rotation period and its harmonics can account for theadditional variation seen in our RV measurements we compareddifferent models by adding signals one by one The first model (P0)includes only the planet signal ie a Keplerian model fitted to theRV data using the same priors given in Table 5 but fixing epoch andperiod to the values derived by the transit modeling The next model(P1) is obtained from P0 by adding a sinusoidal signal at the rotationperiod of the star (Prot) Models P2 includes the first harmonic ofthe rotation period (Prot2) whereas model P3 account for the first(Prot2) and second (Prot3) harmonics While adding sinusoidalsignals we fitted for their amplitudes phases and periods We usedflat priors for the phases and amplitudes (details in Table 5) Weused a Gaussian prior for Prot using the value and its uncertaintyderived in Sect 53 The periods of the harmonic signals were leftfree to vary depending on the value assumed by Prot at each stepof the MCMC chains In order to check if the RV variation inducedby the planet is significant in our data set we also performed the fitusing models where the planetary signal was not included (modelsNP1 and NP2 see Table 4)
Table 4 shows the goodness of the fit for each model Thepreferred model is P2 (planet plus 2 sinusoidal signals at Prot andProt2) with the lowest Akaike Information Criteria (AIC) and max-imum likelihood This result is consistent with the Fourier analysisof the K2 light curve which suggests that the major contributionto the photometric variations arises from the stellar rotation periodand its first harmonic Our analysis provides also additional evi-dence that the Doppler motion induced by the planet is present inour RV data set First the planet signal does not significantly varyfor the P0 P1 P2 and P3 models (Table 4) Second the modelswith no planetary signal (NP1 and NP2) provide a poor fit to theRV measurements (Table 4)
To account for additional instrumental noise not included inthe nominal RV error bars andor imperfect treatment of the varioussources of RV variations we fitted for a jitter term for each instru-ment The final parameter estimates and their error bars are listedin Table 5 They are defined as the median and the 68 credibleinterval of the final posterior distributions The best fitting transitand RVmodels are displayed in Figure 4 along with the photometricand RV data points
72 Additional companion
Huang et al (2016) found that warm Jupiters with low eccentricities(e 04) have inner low-mass companions They used this evidenceas an argument in favour of the in situ formation since the planetmigration would have cleaned the warm Jupiter neighborhood Wesearched the light curve for additional transit signals but foundno evidence for an additional transiting planet in the system Asdescribed in the previous paragraph the periodogram of the RVresiduals showno significant peakwith false alarmprobability lowerthan 5
Figure 4 Top FIES (blue circles) HARPS-N (green diamonds) and HARPS (red squares) RV measurements versus time following the subtraction of thesystemic velocities for each instrument The 1σ uncertainties are marked using the same color used for each data-set The vertical gray lines mark the errorbars including jitter The solid line represents the best fitting RV model which includes the planet signal and the activity signal at the stellar rotation periodand its first harmonic The dashed dash-dotted and dotted lines show the RV contribution of K2-139 b stellar rotation and first harmonic respectively Lowerleft panel Transit light curve folded to the orbital period of K2-139 b and residuals The red points mark the K2 data and their error bars The solid line markthe re-binned best-fitting transit model Lower right panel Phase-folded RV curve of K2-139 b and best fitting Keplerian solution (solid line) following thesubtraction of the two additional sinusoidal signals used to account for the stellar activity The FIES HARPS and HARPS-N are corrected for the instrumentoffsets as derived from the global analysis
Table 4 Model comparison
Model Comment Npars Kb (m sminus1) χ2dof(a) ln L AIC(b)
P0 Planet signal 6 291 plusmn 20 61 356 -60P1 Planet signal + 1 sine-curve at Prot 9 294 plusmn 24 34 581 -98
P2 Planet signal + 2 sine-curves at Prot and Prot2 11 273+26minus25 38 601 -98
P3 Planet signal + 3 sine-curves at Prot Prot2 and Prot3 13 278+27minus26 53 593 -93
NP1 1 sine-curve at Prot (No planet signal) 6 0 185 -448 101NP2 2 sine-curves at Prot and Prot2 (No planet signal) 8 0 159 -120 40
Note ndash (a) χ2 value assuming no jitter (b)We used the Akaike Information Criteria (AIC = 2Nparsminus ln 2L) instead of the widely used Bayesian informationcriteria (BIC) because our RV data sample is small (19 data points) and BIC performs better for large samples (Burnham amp Anderson 2002)
73 Spot-crossing events
The passage of a planet in front of a spot can be detected as abump in the transit light curve (see eg Sanchis-Ojeda amp Winn2011) Spot-crossings events are clearly visible in the EVERESTtransit light curves (Fig 4) The same features appear at the sametimes and with consistent amplitudes in the Vanderburg amp Johnson(2014) data confirming that the bumps are real and not due to
systematics To assess whether the bumps significantly affect theparameter estimates we performed the joint analysis as describedin Sect 6 including all the transit data points We found that the finalparameters are consistent within 1-σ with those reported in Table 5
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 9
01 05 1 2 3 4 5 10Mass (MJup)
06
08
10
12
14
Rad
ius
(RJu
p)
H 0 M core 10 M core 25 M core 50 M core100 M core
Figure 5 Warm Jupiters (black squares Mp gt 03 MJup and10 6 Porb 6 100 days) whose mass and radius have been estimated with aprecision of at least 25 (as of January 2017 exoplaneteu) K2-139 b isshown with a filled red circle The solid line corresponds to a planet with apure hydrogen composition (Seager et al 2007) The dashed lines representthe Fortney et al (2007) models for planet core masses of 0 10 25 50and 100 Moplus The vertical dotted line marks the giant planet lower limit asdefined by Hatzes amp Rauer (2015)
74 Planetrsquos composition and formation scenario
With amass of Mp = 0387+0083minus0075 MJ and radius of Rp = 0808+0034
minus0033RJ (resulting in a mean density of ρp=091+024
minus020 g cmminus3) K2-139 bjoins the small group of well characterized warm Jupiters Fig 5shows the position of K2-139 b in themass-radius diagram for warmJupiters (Mp gt 03 MJup 10 6 Porb 6 100 days) whose massand radius have been determined with a precision better than 25(14 objects) Notably K2-139 b is the transiting warm Jupiter withthe lowest mass known to date if the definition of giant planetsgiven by Hatzes amp Rauer (2015) is adopted Fig 5 displays also theplanetary models of Fortney et al (2007) for different core massesand age between 10 and 45Gyrs The planet radius of K2-139 bcan be explained if the planet has a core10 of 49+19
minus17 Moplus containingsim40 of the total planetary mass We expect that K2-139 b has asolid core surrounded by a gaseous envelope
Rafikov (2006) found that a core of mass 5 ndash 20 Moplus at a semi-major axis between 01 and 10AU would be able to start the run-away accretion phase to form a gas giant planet in situ Howeveraccording to his models these kind of cores are unlikely to formowing to the high irradiation coming from the star Boley et al(2016) suggested instead that more massive cores (Mcore amp 20Moplus)can be built up from the merging of tightly packed inner planetsformed at the early stages of the circumstellar disc Batygin et al(2016) found a similar result and argued that the massive core ofHD149026b (Mcore asymp 100Moplus) could be explained by one or moresuper-Earths which merged and accreted the surrounding gas toform a gas-giant planet Huang et al (2016) suggested that thesecores can initiate runaway accretion if they are formed in a re-gion with enough gas around them while those without enoughvolatiles remain super-Earths and represent the population of mas-sive rocky planets unveiled by Kepler around solar-like stars (egDemory 2014) Based on these studies and given the semi-majoraxis of 0179+0021
minus0027 AU the 48 plusmn 14 Moplus core of K2-139 b could
10 Calculated by interpolating Fortney et al (2007)rsquos models
have formed the planet in situ We note that the metallicity of K2-139 is relatively high ([FeH] = 021 plusmn 005) suggesting that theprimordial circumstellar disc had a relatively high content of dustwhich would have enhanced the formation of the core of K2-139 b(see eg Johnson amp Li 2012) Alternatively the planet might haveformed beyond the snow line and migrated inwards via planet-discinteraction (see eg Baruteau et al 2014)
8 CONCLUSIONS
We confirmed the planetary nature and derived the orbital and mainphysical parameters of K2-139 b a warm Jupiter (Teq = 565+48
minus32 K)transiting an active (log RprimeHK = minus446 plusmn 006) K0V star every 29days We measured a planetary mass of Mp = 0387+0083
minus0075 MJ andradius of Rp = 0808+0034
minus0033 RJ At a separation of ap = 0179+0021minus0027
AU the mean density of ρp = 091+024minus020 g cmminus3 implies that the
planet has a core of 49+19minus17 Moplus according to the evolutionary mod-
els of Fortney et al (2007) K2-139 b joins the small group ofwell-characterized warm Jupiters whose mass and radius have beendetermined with a precision better than 25
The spin-orbit angle ie the angle between the spin axis ofthe star and the angular momentum vector of the orbit can pro-vide us with valuable information on the migration mechanismsof exoplanets (see eg Winn 2010 Morton amp Johnson 2011 Al-brecht et al 2012 Gandolfi et al 2012) Currently there are only4 warm Jupiters (Mp gt 03 MJup and 10 6 Porb 6 100 days)with measured obliquity11 From this perspective K2-139 is anideal target to measure the sky-project spin-orbit angle via obser-vations of the Rossiter-McLaughlin (RM) effect Assuming spin-orbit alignement the expected amplitude of the RM anomaly is∆RV asymp
radic1 minus b2 (RpR)2 v sin iasymp 25 m sminus1 (Winn 2010) Given
the brightness of the host star (V = 11653 mag) this amplitudecan easily be measured using state-of-the-art spectrographs suchas HARPSESO-36m Moreover the transit duration (sim5 hours)is shorter than the visibility of K2-139 which is sim9 hours fromLa Silla observatory (altitude higher than 30 above the horizon)
Alternatively the spin-orbit angle could be measured fromthe analysis of the spot-crossing events as described in Sanchis-Ojeda et al (2011) and Sanchis-Ojeda et al (2012) Anomaliesascribable to the passage of K2-139b in front of stellar spots arevisible in the 3 transit light curves observed by K2 Unfortunatelythe limited number of transits and the K2 long cadence data donot allow us to perform a meaningful quantitative analysis of thespot-crossing events Given the amplitude of the detected anoma-lies (sim01) space-based high-precision photometry is needed todetect the spot-crossing events Observations performed with theupcoming CHaracterising ExOPlanets Satellite (CHEOPS Broeget al 2013) would allow us to photometrically determine the spin-orbit angle of this system
ACKNOWLEDGEMENTS
We warmly thank the NOT ESO TNG staff members for theirunique support during the observations We are very thankful toXavier Bonfils Franccedilois Bouchy Martin Kuumlrster Tsevi MazehJorge Melendez and Nuno Santos who kindly agreed to exchange
11 Source httpwww2mpsmpgdehomeshellercontentmain_HRMhtml as of January 2017
MNRAS 000 1ndash12 (2017)
10 O Barragaacuten et al
HARPSandFIES timewith us Special thanks go toAntoninoLanzafor assisting us with the calculation of the gyro-age of the star Wealso greatly thank the anonymous referee for herhis careful re-view and suggestions which helped us to improve the manuscriptD Gandolfi gratefully acknowledges the financial support of theProgramma Giovani Ricercatori ndash Rita Levi Montalcini ndash Rien-tro dei Cervelli (2012) awarded by the Italian Ministry of Edu-cation Universities and Research (MIUR) Sz Csizmadia thanksthe Hungarian OTKA Grant K113117 H J Deeg and D Nespralacknowledge support by grant ESP2015-65712-C5-4-R of the Span-ish Secretary of State for RampDampi (MINECO) D Lorenzo-Oliveiraacknowledges the support from FAPESP (201620667-8) This re-search was supported by the Ministerio de Economia y Competi-tividad under project FIS2012-31079 The research leading to theseresults has received funding from the European Union SeventhFramework Programme (FP72013-2016) under grant agreementNo 312430 (OPTICON) Based on observations obtained a) withthe Nordic Optical Telescope (NOT) operated on the island of LaPalma jointly by Denmark Finland Iceland Norway and Swedenin the Spanish Observatorio del Roque de los Muchachos (ORM)of the Instituto de Astrofiacutesica de Canarias (IAC) b) with the Ital-ian Telescopio Nazionale Galileo (TNG) also operated at the ORM(IAC) on the island of La Palma by the INAF - Fundacioacuten GalileoGalilei c) the 36m ESO telescope at La Silla Observatory underprogramme ID 097C-0948 The data presented here were obtainedin part with ALFOSC which is provided by the Instituto de As-trofisica de Andalucia (IAA) under a joint agreement with the Uni-versity of Copenhagen and NOTSA This paper includes data col-lected by the Kepler mission Funding for the Kepler mission is pro-vided by the NASA Science Mission directorate Some of the datapresented in this paper were obtained from theMikulski Archive forSpace Telescopes (MAST) STScI is operated by the Association ofUniversities for Research in Astronomy Inc under NASA contractNAS5-26555 Support for MAST for non-HST data is provided bythe NASA Office of Space Science via grant NNX09AF08G and byother grants and contracts MF and CMP acknowledge generoussupport from the Swedish National Space Board C Eiroa and IRebollido are supported by Spanish grant AYA2014-55840-P PDacknowledge the support from INAF and Ministero dellrsquoIstruzionedellrsquoUniversitagrave e della Ricerca (MIUR) in the form of the grantldquoPremiale VLT 2012rdquo and ldquoThe Chemical and Dynamical Evolu-tion of the Milky Way and Local Group Galaxiesrdquo This work hasmade use of data from the European Space Agency (ESA) missionGaia (httpwwwcosmosesaintgaia) processed by theGaia Data Processing and Analysis Consortium (DPAC httpwwwcosmosesaintwebgaiadpacconsortium) Fundingfor the DPAC has been provided by national institutions in particu-lar the institutions participating in theGaiaMultilateral Agreement
REFERENCES
Albrecht S et al 2012 ApJ 757 18Allard F Homeier D Freytag B 2011 in Johns-Krull C BrowningM K
West A A eds Astronomical Society of the Pacific Conference SeriesVol 448 16th Cambridge Workshop on Cool Stars Stellar Systemsand the Sun p 91 (arXiv10115405)
Antonini F Hamers A S Lithwick Y 2016 preprint(arXiv160401781)
Barnes S A 2010 ApJ 722 222Barnes S A Kim Y-C 2010 ApJ 721 675Barragaacuten O Gandolfi D 2017 Exotrending Astrophysics Source Code
Library (ascl1706001)
Barragaacuten O et al 2016 AJ 152 193Barragaacuten O Gandolfi D Antoniciello G 2017 pyaneti Astrophysics
Source Code Library (ascl1707003)Baruteau C et al 2014 Protostars and Planets VI pp 667ndash689Batygin K Bodenheimer P H Laughlin G P 2016 ApJ 829 114Boley A C Granados Contreras A P Gladman B 2016 ApJ 817 L17Boyajian T S et al 2013 ApJ 771 40Brahm R et al 2016 AJ 151 89Bressan A Marigo P Girardi L Salasnich B Dal Cero C Rubele S
Nanni A 2012 MNRAS 427 127Broeg C et al 2013 in European Physical Journal Web of Conferences p
03005 (arXiv13052270) doi101051epjconf20134703005Bruntt H et al 2010 MNRAS 405 1907Buchhave L A et al 2010 ApJ 720 1118Burnham K Anderson D 2002 Model Selection and Multimodel Infer-
ence A Practical Information-Theoretic Approach NewYork Springer-Verlag
Cabrera J et al 2009 AampA 506 501Cabrera J Csizmadia S Erikson A Rauer H Kirste S 2012 AampA 548
A44Cabrera J et al 2014 ApJ 781 18Cantat-Gaudin T et al 2014 AampA 562 A10Cardelli J A Clayton G C Mathis J S 1989 ApJ 345 245Carone L et al 2012 AampA 538 A112Carpano S et al 2009 AampA 506 491Castelli F Kurucz R L 2004 preprintCavarroc C et al 2012 ApampSS 337 511Chabrier G 2001 ApJ 554 1274Cosentino R et al 2012 in Ground-based and Airborne Instrumentation
for Astronomy IV p 84461V doi10111712925738Cutri R M et al 2003 2MASS All Sky Catalog of point sourcesCutri R M et al 2012 Technical report Explanatory Supplement to the
WISE All-Sky Data Release ProductsDawson R I Johnson J A Morton T D Crepp J R Fabrycky D C
Murray-Clay R A Howard A W 2012 ApJ 761 163Deeg H J et al 2010 Nature 464 384Demory B-O 2014 ApJ 789 L20Dong S Katz B Socrates A 2014 ApJ 781 L5Doyle A P Davies G R Smalley B Chaplin W J Elsworth Y 2014
MNRAS 444 3592Dumusque X Boisse I Santos N C 2014 ApJ 796 132Endl M Cochran W D 2016 PASP 128 094502Erikson A et al 2012 AampA 539 A14Fabricius C et al 2016 AampA 595 A3Fortney J J Marley M S Barnes J W 2007 ApJ 659 1661Fossati et al 2017 AampA submittedFrandsenS LindbergB 1999 inKarttunenH PiirolaV edsAstrophysics
with the NOT p 71Frewen S F N Hansen B M S 2016 MNRAS 455 1538Gandolfi D et al 2008 ApJ 687 1303Gandolfi D et al 2012 AampA 543 L5Gandolfi D et al 2015 AampA 576 A11Gray R O 1999 SPECTRUM A stellar spectral synthesis program As-
trophysics Source Code Library (ascl9910002)Grziwa S Paumltzold M 2016 preprint (arXiv160708417)Grziwa S Paumltzold M Carone L 2012 MNRAS 420 1045Hamers A S Antonini F Lithwick Y Perets H B Portegies Zwart S F
2016 preprint (arXiv160607438)Hatzes A P 2002 Astronomische Nachrichten 323 392Hatzes A P Rauer H 2015 ApJ 810 L25Heiter U et al 2015 Phys Scr 90 054010Huang C Wu Y Triaud A H M J 2016 ApJ 825 98Jenkins J S et al 2017 MNRAS 466 443Johnson J L Li H 2012 ApJ 751 81Kipping D M 2010 MNRAS 408 1758Kley W Nelson R P 2012 ARAampA 50 211Kovaacutecs G Zucker S Mazeh T 2002 AampA 391 369
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 11
Table 5 K2-139 system parameters
Parameter Prior(a) Final value
Stellar parametersStar mass M (M) middot middot middot 0919 plusmn 0033Star radius R (R) middot middot middot 0862 plusmn 0032Star density ρ (from spectroscopy g cmminus3) middot middot middot 202+025
minus022
Star density ρ (from light curve g cmminus3) middot middot middot 211+074minus081
Effective Temperature Teff (K) middot middot middot 5340 plusmn 110Surface gravity log g (cgs) middot middot middot 450 plusmn 009Iron abundance [FeH] (dex) middot middot middot 022 plusmn 008Microturbulent velocity vmic ( km sminus1) middot middot middot 09 plusmn 01Macroturbulent velocity vmac ( km sminus1) middot middot middot 25 plusmn 06Projected rotational velocity v sin i ( km sminus1) middot middot middot 28 plusmn 06Rotational period Prot (days) middot middot middot 1724 plusmn 012Activity index(b) logRprimeHK middot middot middot minus446 plusmn 006Gyrochronological age (Gyr) middot middot middot 18 plusmn 03Interstellar extinction AV (mag) middot middot middot 007 plusmn 005Star distance d (pc) middot middot middot 152 plusmn 10
Model parameters of K2-139 bOrbital period Porb (days) U[283773 283873] 2838236 plusmn 000026Transit epoch T0 (BJDTDBminus2 450 000) U[73258120 73258220] 732581714 plusmn 000033Scaled semi-major axis aR U[12 100] 448+47
minus67Planet-to-star radius ratio RpR U[0 02] 00961+00023
minus00015Impact parameter b U[0 12] 030+021
minus019radice sinω U[minus1 1](c) 010+029
minus030radice cosω U[minus1 1](c) 006+024
minus027Radial velocity semi-amplitude variation K ( m sminus1) U[0 200] 277+60
Note ndash The adopted Sun and Jupiter units follow the recommendations from the International Astronomical Union (Prša et al 2016) (a) U[a b] refersto uniform priors between a and b N[a b] means Gaussian priors with mean a and standard deviation b and F[a] to a fixed a value (b) Corrected forinterstellar reddening following Fossati et al (2017) The correction is +006 (c) The code always ensures that e lt 1 (d) Assuming albedo = 0
MNRAS 000 1ndash12 (2017)
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
MNRAS 000 1ndash12 (2017)
1 Introduction
2 K2 photometry
3 ALFOSC imaging
4 High-resolution spectroscopy
5 Stellar parameters
51 Spectral analysis
52 Interstellar extinction
53 Rotational period
54 Stellar mass radius and age
6 Joint RV-transit fit
7 Results and discussion
71 Stellar activity modeling
72 Additional companion
73 Spot-crossing events
74 Planets composition and formation scenario
8 Conclusions
Acknowledgements
The transiting warm Jupiter K2-139 b 3
2480 2500 2520 2540BJD shy 2454833
099
100
101
Rel
ativ
e flu
x
Figure 1 K2 Light curve for K2-139 as extracted by Luger et al (2016) The positions of the 3 observed transits are marked with vertical dashed lines
the star EPIC 218916923 with both DST and EXOTRANS Asa sanity check we downloaded the EVEREST light curve ofEPIC 218916923 (Luger et al 2016) and detected the same sig-nal We note that Vanderburg amp Johnson (2014) and Luger et al(2016) used the same mask to extract the time-series data from therawK2 images EPIC 218916923 was proposed forK2 observationsby programs GO7086 (PI Thompson) GO7030 (PI Howard) andGO7087 (PI Dragomir) We will hereafter refer to the star and itstransiting planet as K2-139 and K2-139 b respectively
We searched the Vanderburg amp Johnson (2014)rsquos light curvefor odd-even transit depth variation and secondary eclipse thatmighthint to a binary scenario making the system a likely false positiveNone of themwere significantly detected The depth of the oddeventransits agrees within the 1-σ uncertainty of 16 times 10minus3 whereasthe 3-σ upper limit of the occultation depth is 79 times 10minus5 both re-spect to the normalized flux We proceeded to more detailed fittingof the light curve as well as ground-based imaging (Sect 3) andspectroscopic observations (Sect 4) The main identifiers coordi-nates optical and infrared magnitudes and proper motions of thestar are listed in Table 1 We display the EVEREST K2 light curveof K2-139 in Fig 1
3 ALFOSC IMAGING
K2 Campaign 7 is projected close to the galactic center and thus ina relatively crowded stellar region In order to estimate the contam-ination factor arising from sources whose light leaks into the photo-metricmasks used byVanderburgamp Johnson (2014) and Luger et al(2016) we observed K2-139 on 13 September 2016 (UT) with theALFOSC camera mounted at the Nordic Optical Telescope (NOT)of Roque de los Muchachos Observatory (La Palma Spain) Thesky conditions were photometric with excellent seeing conditions(sim06primeprime) We used the Bessel R-filter and acquired 16 images of 6sec 2 images of 20 sec and 1 image of 120 sec The data were biassubtracted and flat-fielded using dusk sky flats Aperture photom-etry was then performed on all stars within the mask used in theextraction of the light curve by Vanderburg amp Johnson (2014) andLuger et al (2016)
Several fainter stars can be identified inside the photometricmask (Fig 2) of which the two brightest sources are also in theEPIC catalog with Kepler band magnitudes of 168 and 184 Theclosest detected source is a 68-mag fainter star at 38primeprime South of K2-
Figure 2 ALFOSC Bessel R-band image of the sky region around K2-139North is up and East is to the left The target star is the brightest sourcein the middle The solid black polygon marks the EVEREST photometricmask (Luger et al 2016)
139 We can exclude stars as faint as sim20mag at an angular distancelarger than sim06primeprime from K2-139 It is worth noting that the fainteststar whose flux could account for the sim1 deep transit of K2-139cannot be more than sim5 mag fainter than our target The summedflux of these faint stars amounts to 14plusmn03 of the total off-transitflux within the apertureWe subtracted this contamination flux fromthe EVEREST K2 light curve prior to performing the joint analysispresented in Sect 6
4 HIGH-RESOLUTION SPECTROSCOPY
In June and August 2016 we obtained two reconnaissance spectraof K2-139 with the Tull spectrograph (Tull et al 1995) at the 27-mtelescope at McDonald Observatory (Texas USA) The high res-olution (R asymp 60 000) spectra have a signal-to-noise ratio of sim30per pixel at 5500Aring We reduced the data using standard IRAF rou-tines and derived preliminary spectroscopic parameters using ourcode Kea (Endl amp Cochran 2016) The results from both spec-
MNRAS 000 1ndash12 (2017)
4 O Barragaacuten et al
Table 2 Radial velocity measurements and activity indexes of K2-139
tra are nearly identical and reveal a star with Teff = 5500plusmn 100 Klog g = 465plusmn 012 (cgs) [FeH]=+011plusmn 012 dex and a slow pro-jected rotational velocity of v sin iasymp 2 km sminus1
The high-precision radial velocity follow-up of K2-139 wasstarted in June 2016with the FIbre-fed Eacutechelle Spectrograph (FIESFrandsenampLindberg 1999 Telting et al 2014)mounted at the 256-m Nordic Optical Telescope (NOT) The observations were carriedout as part of the OPTICON andCAT observing programs 16A055P53-201 and P53-203 We used the high-resmode which providesa resolving power of Rasymp 67 000 in the whole visible spectral range(3700 minus 7300Aring) The exposure time was set to 2100 ndash 3600 secbased on sky conditions and observing scheduling constraints Fol-lowing the observing strategy outlined in Buchhave et al (2010)and Gandolfi et al (2015) we traced the RV drift of the instrumentby acquiring long-exposed (Texp asymp 35 sec) ThAr spectra immedi-ately before and after the target observations The typical RV driftmeasured between two ThAr spectra bracketing a 2100 ndash 3600 secscience exposure is about 50 ndash 80 m sminus1 A linear interpolation ofthe RV drift to the mid-time of the science exposure allows us toachieve a radial velocity zero-point stability of about 5 ndash 6 m sminus1which is 2 ndash 3 times smaller than the nominal error bars listed in Ta-ble 2 The data reduction uses standard IRAF and IDL routines Thesignal-to-noise (SN) ratio of the extracted spectra is sim30 ndash 40 perpixel at 5500 Aring Radial velocity measurements were extracted viamulti-order cross-correlation with the RV standard star HD182572observed with the same instrument set-up as K2-139
We also observed K2-139 in July August and September 2016with the HARPS (Mayor et al 2003) and HARPS-N (Cosentinoet al 2012) spectrographs mounted at the ESO 36-m Telescope ofLa SillaObservatory (Chile) and at the 358-mTelescopioNazionaleGalileo (TNG) of Roque de los Muchachos observatory (La PalmaSpain) respectively Both instruments provide a resolving powerof Rasymp 115 000 in the wavelength range sim3800 ndash 6900 Aring The ob-
servations were performed as part of the ESO and TNG observingprograms 097C-0948 and A33TAC_15 respectively The exposuretime was set to 1800 sec leading to a SN ratio of sim35 on the ex-tracted spectra We reduced the data using the dedicated HARPSand HARPS-N pipelines and extracted the RVs by cross-correlationwith a G2 numerical mask
The FIES HARPS and HARPS-N RVs are listed in Ta-ble 2 along with the bisector span (BIS) and the full width at halfmaximum (FWHM) of the cross-correlation function (CCF) Timestamps are given in barycentric Julian date in barycentric dynam-ical time (BJDTDB) For the HARPS and HARPS-N data we alsoprovide the Ca ii HampK chromospheric activity index log RprimeHK Wedid not measure log RprimeHK from the FIES spectra because of the poorSN ratio at wavelengths shorter than 4000Aring
5 STELLAR PARAMETERS
51 Spectral analysis
We derived the spectroscopic parameters of K2-139 from the co-added FIES spectra The stacked FIES data have a SN ratio of sim110per pixel at 5500 Aring We adopted three different methods For eachmethod results are reported in Table 3
First method The technique fits spectral features that are sen-sitive to different photospheric parameters It uses the stellar spec-tral synthesis program Spectrum (Gray 1999) to compute syntheticspectra from ATLAS 9model atmospheres (Castelli ampKurucz 2004)Microturbulent (vmic) and macroturbulent (vmac) velocities are de-rived from the calibration equations of Bruntt et al (2010) andDoyle et al (2014) We used the wings of the Hα and Hβ lines toestimate the effective temperature (Teff) and the Mg i 5167 5173and 5184 Aring Ca i 6162 and 6439 Aring and the Na iD lines to determine
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 5
Table 3 Spectroscopic parameters of K2-139 as derived using the three methods described in Sect 5
Method Teff (K) log g (cgs) [FeH] (dex) vmic ( km sminus1) vmac ( km sminus1) v sin i ( km sminus1)
the surface gravity log g We simultaneously fitted different spec-tral regions to measure the iron abundance [FeH] The projectedrotational velocity v sin i was determined by fitting the profile ofmany isolated and unblended metal lines
Second method It relies on the use of the spectral analysispackage Spectroscopy Made Easy (SME Valenti amp Piskunov 1996Valenti amp Fischer 2005) For a set of given stellar parameters SMEcalculates synthetic spectra and fits them to high-resolution ob-served spectra using a chi-squared minimization procedure Weused SME version 443 and ATLAS 12model spectra (Kurucz 2013)We adopted the same calibration equation as described in the firstmethod to determine vmic and vmac Effective temperature is de-rived from the Hα wings log g from the Ca i 6102 6122 6162and 6439 Aring lines [FeH] and v sin i from isolated iron lines
Third method It uses the classical equivalent width (EW)method adopting the following criteria i) Teff is obtained by re-moving trends between abundance of the chemical elements andthe respective excitation potentials ii) log g is optimised by as-suming the ionisation equilibrium condition ie by requiring thatfor a given species the same abundance (within the uncertainties)is obtained from lines of two ionisation states (typically neutral andsingly ionised lines) iii) vmic is set by minimising the slope of therelationship between abundance and the logarithm of the reducedEWs The equivalent widths of Fe i and Fe ii lines are measuredusing the code DOOp (Cantat-Gaudin et al 2014) a wrapper ofDAOSPEC (Stetson amp Pancino 2008) The stellar atmosphere param-eters are derived with the program FAMA (Magrini et al 2013) awrapper of MOOG (Sneden et al 2012) We used the public versionof the atomic data prepared for the Gaia-ESO Survey (Heiter et al2015) and based on the VALD3 data (Ryabchikova et al 2011) Weused sim200 Fe i lines and sim10 Fe ii lines for the determination of thestellar parameters
The three methods provide consistent results within the 1-σerror bars (Table 5) While we have no reason to prefer one tech-nique over the other we adopted the parameter estimates of thefirst method ie Teff = 5340 plusmn 110 K log g = 450 plusmn 009 (cgs)[FeH] = 022 plusmn 008 dex vmic = 09 plusmn 01 km sminus1 vmac = 25 plusmn06 km sminus1 and v sin i = 28 plusmn 06 km sminus1 As a sanity check wealso analyzed the HARPS and HARPS-N data and obtained con-sistent results but with larger error bars owing to the lower SNratio of the co-added HARPS and HARPS-N spectra compared tothat of the co-added FIES data Using the Boyajian et al (2013)rsquoscalibration (see their Table 6) the effective temperature of K2-139defines the spectral type of the host star as K0V
52 Interstellar extinction
We measured the visual reddening (AV) of K2-139 following thetechnique described in Gandolfi et al (2008) We fitted the spectral
energy distribution of the star to synthetic colors extracted from theBT-NEXTGEN model spectrum (Allard et al 2011) with the samephotospheric parameters as the star We adopted the extinction lawof Cardelli et al (1989) and assumed a normal value for the total-to-selective extinction ie RV = AVE(B minus V)= 31 We measureda visual extinction of AV = 007plusmn 005 mag This value is belowthe upper limit of AV 03 mag extracted from the Schlegel et al(1998)rsquos all-sky extinction map corroborating our result
53 Rotational period
The K2 light curve of K2-139 displays periodic and quasi-periodicvariations with a peak-to-peak photometric amplitude of sim2(Fig 1) The late-type spectral type of the star suggests that theobserved variability is due to Sun-like spots appearing and disap-pearing from the visible stellar disc as the star rotates around its axisThis is corroborated by the fact that K2-139 is a chromosphericallyactive star The HARPS and HARPS-N spectra show clear emis-sion components in the cores of the Ca ii HampK lines from whichwe measured an average activity index of log RprimeHK =minus446plusmn 0064
The out-of-transit photometric variability observed in the lightcurve of K2-139 is mainly due to two active regions located at op-posite stellar longitudes whose lifetime is longer than the durationof the K2 observations Using the spots as tracers of stellar rotationand following the auto correlation function (ACF) technique de-scribed in McQuillan et al (2014) we estimated that the rotationalperiod of the star is Prot = 1724 plusmn 012 days The Lomb-Scargleperiodogram of the light curve shows its strongest peak at the sameperiod confirming our results
It is worth noting that the rotation period (Prot = 1724 plusmn012 days) and radius (R= 0862 plusmn 0032 R see next section)of the host star translate into a maximum value for the projected ro-tational velocity of v sin imax = 253plusmn 010 km sminus1 which agreeswith the spectroscopically derived v sin i = 28 plusmn 06 km sminus1 sug-gesting that the star is seen nearly equator-on (i asymp 90) and thatthe system might be aligned along the line-of-sight
54 Stellar mass radius and age
We derived the stellar mass radius and age using the online inter-face for Bayesian estimation of stellar parameters available at thefollowing web page httpstevoapdinafitcgi-binparam Briefly the web tool interpolates onto PARSEC modelisochrones (Bressan et al 2012) the V-band apparent magnitude
4 This value is corrected for the interstellar medium absorption followingthe procedure described in Fossati et al (2017) and using themeasured stellarparameters and reddening The correction is +006 The star is thereforeslightly more active than what measured from the spectra
MNRAS 000 1ndash12 (2017)
6 O Barragaacuten et al
effective temperature metal content and parallax We used the V-band magnitude reported in Table 1 ndash after correcting for interstel-lar reddening (Sect 52) ndash along with the effective temperature andmetal content we derived in Sect 5 The parallax was retrieved fromthe Gaiarsquos first data release (px = 656 plusmn 043mas d = 152plusmn 10 pcFabricius et al 2016) We adopted the log-normal initial mass func-tion from Chabrier (2001)
K2-139 has a mass of M= 0919 plusmn 0033 M and radiusof R= 0862 plusmn 0032 R corresponding to a surface gravity oflog g = 4503plusmn0035 (cgs) in excellent agreement with the spec-troscopically derived value of log g = 450plusmn009 (cgs see Sect 5)The derivedmean density ρ = 202plusmn024 g cmminus3 ofK2-139 is alsoconsistent within 1-σ with the density estimated by the modelingof the transit light curve (ρ = 211+074
minus081 g cmminus3 see Sect 6)The isochrones provide an age of 36plusmn34 Gyr for K2-139 Us-
ing the equations given in Barnes amp Kim (2010) and Barnes (2010)the rotation period of 173 days (Sect 53) implies a gyrochronolog-ical age of 18plusmn 03 Gyr
6 JOINT RV-TRANSIT FIT
We performed the joint fit to the photometric and RV data using thecode pyaneti5 (Barragaacuten et al 2017) a PythonFortran softwaresuite based on Markov Chain Monte Carlo (MCMC) methods
The photometric data included in the joint analysis are subsetsof the whole EVEREST K2 light curve We used the EVERESTlight curve because it provides a slightly better rms over the Vander-burg amp Johnson (2014)rsquos dataWe selectedsim10 hours of data-pointsaround each of the 3 transits which have a duration of sim5 hours Wede-trended each individual transits with the code exotrending6
(Barragaacuten amp Gandolfi 2017) using a second-order polynomial fit-ted to the out-of-transit points The fitted data include 12 pointsimmediately before and after each transit with the exception of thelast transit for which only 9 data points are available We removedthe data points that are affected by stellar spot crossing events (seeSect 71 for more details)
We fitted the RV data using a Keplerian model for the planetalong with two sine-like curves to account for the activity-inducedRV (see next section for details) We adopted the limb-darkenedquadratic law of Mandel amp Agol (2002) for the transit model Weadopted the Gaussian likelihood
L =[
nprodi=1
2π
(σ2i + σ
2j
)minus12]
exp
minus
nsumi=1
12(Di minus Mi)2
σ2i+ σ2
j
(1)
where n is the number of data points σi is the error associatedto each data point Di Mi is the model associated to a given Di andσj is an extra noise term sometime referred as jitter
The sampling method and fitted parameters are the same as inBarragaacuten et al (2016) Details on the adopted priors are given in Ta-ble 5 Following Kipping (2010) we super-sampled the light curvemodel using 10 subsamples perK2 exposure to account for the long-cadence acquisition The parameter space was explored with 500independent chains created randomly inside the prior ranges Thechain convergence was analyzed using the Gelman-Rubin statisticsThe burn-in phase uses 25 000 more iterations with a thin factor
5 Available at httpsgithubcomoscaribvpyaneti6 Available at httpsgithubcomoscaribvexotrending
of 50 The posterior distribution of each parameter has 250 000independent data points
7 RESULTS AND DISCUSSION
71 Stellar activity modeling
A simple Keplerian model provides a poor fit to the RV measure-ments with χ2dof = 61 (Table 4) suggesting that additionalsignals might be present in our Doppler data Activity-induced RVvariation is expected given the 2 peak-to-peak photometric vari-ability observed in the K2 light curve of K2-139 (Fig 1) and theCa ii HampK activity index of log RprimeHK =minus446 plusmn 006 (Sect 53)The K2 photometric variation corresponds to a spot filling factorof approximately 2 if this variation is due to cool starspots Wecan use the empirical relationship relating spot coverage to RV am-plitude from Saar amp Donahue (1997) or Hatzes (2002) to estimatethe RV amplitude expected from spots Using the projected rota-tional velocity of 28 km sminus1 results in an RV semi-amplitude ofasymp20ndash30 m sminus1 The code SOAP2 designed to estimate the effectof active regions on photometric and spectroscopic measurements(Dumusque et al 2014) provides consistent results
In order to look for additional signals in our Doppler data weperformed a frequency analysis of theRVmeasurements and activityindicators On one occasion7 K2-139 was observed with FIES andHARPS-S nearly simultaneously (within less than 25 minutes) Weused the two sets of measurements to estimate the RV FWHMand BIS offsets between the two instruments We assumed no offsetbetween HARPS-N and HARPS While we acknowledge that thisassumption is arbitrary we note that the modeling of the RV datagives an offset of ∆RV(HNminusH) = 0002plusmn 00158 km sminus1 (Table 5)which is consistent with zero
Figure 3 displays the generalized Lomb-Scargle periodograms(Zechmeister amp Kuumlrster 2009) of the combined datasets From topto bottom the RVdata (first panel) the RV residuals after subtractingthe transiting planet signal (second panel) and the BIS (third panel)and FWHM (fourth panel) of the cross correlation function Theperiodogram of the window function is shown in the lower panelThe dotted vertical blue lines mark the frequency at the orbitalperiod of the planet (0035 cd) as well as the frequencies at therotation period of the star (0058 cd) and its first two harmonics(0116 and 0174 cd)
The periodogram of the RV data (upper panel) shows a peak atthe orbital frequency of the planet alongwith two additional peaks at0095 and 0130 cd Since the periodogram of the window functionshows two peaks atsim0060 andsim0095 cd (lower panel red arrows)we interpreted the 0095 and 0130 cd peaks as the aliases of theorbital frequency8 We note also that periodogram of the BIS ofthe CCF displays peaks whose frequencies are close to the stellarrotation frequency and its first two harmonics However none of thepeaks visible in the GLS periodograms of Fig 3 has a false alarmprobability (FAP)9 lower than 5 Although our spectroscopic datashow neither additional signals we note that the semi-amplitudevariation of the BIS and FWHM is expected to be 10-15 m sminus1
(Dumusque et al 2014) which is comparable with the uncertaintiesof most of ourmeasurements (Table 2) The lack of significant peaks
7 Epoch BJD=24575898 0095 = 0035 + 0060 cd and 0130 = 0035 + 0095 cd9 We determined the FAP following the Monte Carlo bootstrap methoddescribed in Kuerster et al (1997)
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 7
Figure 3 Generalized Lomb-Scargle periodogram of the combined FIESHARPS and HARPS-N Doppler datasets From top to bottom the RV datathe RV residuals after subtracting the transiting planet signal the BIS andFWHM of the CCF and the window function The dotted vertical blue linesmark the frequencies at the orbital period as well as at the stellar rotationperiod and its first two harmonics The dashed vertical red lines mark the5 false alarm probabilities as derived using the bootstrap method The redarrows in the lower panel mark the two peaks presented in the main text
in the periodogram of the RV data and RV residuals as well as inthe periodogram of the activity indicators could be explained by thelimited number of available measurements and their uncertaintiesWe conclude that we cannot exclude the existence of spot-inducedsignals in our RV measurements
Photometric and radial velocity variations due to rotationalmodulation can be complex with not only the rotational period Protpresent but also its harmonics eg Prot2 Prot3 Assuming thatthe surface structures responsible for this modulation (eg coolspots) are not evolving rapidly then the simplest representation ofthe rotationalmodulation is through the Fourier components definedby the rotation period and its harmonics Figure 1 shows that theevolution time-scale of the active regions in the stellar surface islonger than the 80-day duration of the K2 campaign Since our RVfollow-up spans 55 days we can assume that any activity-inducedRV signal is coherent within our observing window This approachhas been used previously for other planetary systems orbiting activestars (eg Pepe et al 2013)
The Fourier analysis of the K2 light curve is the best way tomeasure the contribution of the rotation period and its harmonics tothe quasi-periodic photometric variability of the star We thereforeanalyzed the K2 light curve using a pre-whitening procedure Thatis the dominant period was found a sine-fit made to the data andsubtracted and additional periods searched in the residual data
We used the program Period04 (Lenz amp Breger 2005) for thisprocedure
The dominant periods are sim172 days ie the rotation periodof the star (Sect 53) and roughly the first four harmonics (ie 8657 43 and 34 days) The 172- and 86-day periods have about thesame amplitude while the 57-day period (Prot3) has 10 of themain amplitude The Prot4 signal has only about 4 of the mainamplitude The light curve analysis indicates that the signal due torotational modulation can largely be represented by the rotationalperiod (Prot) and its first harmonic (Prot2)
In order to test if the addition of RV sinusoidal signals atthe stellar rotation period and its harmonics can account for theadditional variation seen in our RV measurements we compareddifferent models by adding signals one by one The first model (P0)includes only the planet signal ie a Keplerian model fitted to theRV data using the same priors given in Table 5 but fixing epoch andperiod to the values derived by the transit modeling The next model(P1) is obtained from P0 by adding a sinusoidal signal at the rotationperiod of the star (Prot) Models P2 includes the first harmonic ofthe rotation period (Prot2) whereas model P3 account for the first(Prot2) and second (Prot3) harmonics While adding sinusoidalsignals we fitted for their amplitudes phases and periods We usedflat priors for the phases and amplitudes (details in Table 5) Weused a Gaussian prior for Prot using the value and its uncertaintyderived in Sect 53 The periods of the harmonic signals were leftfree to vary depending on the value assumed by Prot at each stepof the MCMC chains In order to check if the RV variation inducedby the planet is significant in our data set we also performed the fitusing models where the planetary signal was not included (modelsNP1 and NP2 see Table 4)
Table 4 shows the goodness of the fit for each model Thepreferred model is P2 (planet plus 2 sinusoidal signals at Prot andProt2) with the lowest Akaike Information Criteria (AIC) and max-imum likelihood This result is consistent with the Fourier analysisof the K2 light curve which suggests that the major contributionto the photometric variations arises from the stellar rotation periodand its first harmonic Our analysis provides also additional evi-dence that the Doppler motion induced by the planet is present inour RV data set First the planet signal does not significantly varyfor the P0 P1 P2 and P3 models (Table 4) Second the modelswith no planetary signal (NP1 and NP2) provide a poor fit to theRV measurements (Table 4)
To account for additional instrumental noise not included inthe nominal RV error bars andor imperfect treatment of the varioussources of RV variations we fitted for a jitter term for each instru-ment The final parameter estimates and their error bars are listedin Table 5 They are defined as the median and the 68 credibleinterval of the final posterior distributions The best fitting transitand RVmodels are displayed in Figure 4 along with the photometricand RV data points
72 Additional companion
Huang et al (2016) found that warm Jupiters with low eccentricities(e 04) have inner low-mass companions They used this evidenceas an argument in favour of the in situ formation since the planetmigration would have cleaned the warm Jupiter neighborhood Wesearched the light curve for additional transit signals but foundno evidence for an additional transiting planet in the system Asdescribed in the previous paragraph the periodogram of the RVresiduals showno significant peakwith false alarmprobability lowerthan 5
Figure 4 Top FIES (blue circles) HARPS-N (green diamonds) and HARPS (red squares) RV measurements versus time following the subtraction of thesystemic velocities for each instrument The 1σ uncertainties are marked using the same color used for each data-set The vertical gray lines mark the errorbars including jitter The solid line represents the best fitting RV model which includes the planet signal and the activity signal at the stellar rotation periodand its first harmonic The dashed dash-dotted and dotted lines show the RV contribution of K2-139 b stellar rotation and first harmonic respectively Lowerleft panel Transit light curve folded to the orbital period of K2-139 b and residuals The red points mark the K2 data and their error bars The solid line markthe re-binned best-fitting transit model Lower right panel Phase-folded RV curve of K2-139 b and best fitting Keplerian solution (solid line) following thesubtraction of the two additional sinusoidal signals used to account for the stellar activity The FIES HARPS and HARPS-N are corrected for the instrumentoffsets as derived from the global analysis
Table 4 Model comparison
Model Comment Npars Kb (m sminus1) χ2dof(a) ln L AIC(b)
P0 Planet signal 6 291 plusmn 20 61 356 -60P1 Planet signal + 1 sine-curve at Prot 9 294 plusmn 24 34 581 -98
P2 Planet signal + 2 sine-curves at Prot and Prot2 11 273+26minus25 38 601 -98
P3 Planet signal + 3 sine-curves at Prot Prot2 and Prot3 13 278+27minus26 53 593 -93
NP1 1 sine-curve at Prot (No planet signal) 6 0 185 -448 101NP2 2 sine-curves at Prot and Prot2 (No planet signal) 8 0 159 -120 40
Note ndash (a) χ2 value assuming no jitter (b)We used the Akaike Information Criteria (AIC = 2Nparsminus ln 2L) instead of the widely used Bayesian informationcriteria (BIC) because our RV data sample is small (19 data points) and BIC performs better for large samples (Burnham amp Anderson 2002)
73 Spot-crossing events
The passage of a planet in front of a spot can be detected as abump in the transit light curve (see eg Sanchis-Ojeda amp Winn2011) Spot-crossings events are clearly visible in the EVERESTtransit light curves (Fig 4) The same features appear at the sametimes and with consistent amplitudes in the Vanderburg amp Johnson(2014) data confirming that the bumps are real and not due to
systematics To assess whether the bumps significantly affect theparameter estimates we performed the joint analysis as describedin Sect 6 including all the transit data points We found that the finalparameters are consistent within 1-σ with those reported in Table 5
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 9
01 05 1 2 3 4 5 10Mass (MJup)
06
08
10
12
14
Rad
ius
(RJu
p)
H 0 M core 10 M core 25 M core 50 M core100 M core
Figure 5 Warm Jupiters (black squares Mp gt 03 MJup and10 6 Porb 6 100 days) whose mass and radius have been estimated with aprecision of at least 25 (as of January 2017 exoplaneteu) K2-139 b isshown with a filled red circle The solid line corresponds to a planet with apure hydrogen composition (Seager et al 2007) The dashed lines representthe Fortney et al (2007) models for planet core masses of 0 10 25 50and 100 Moplus The vertical dotted line marks the giant planet lower limit asdefined by Hatzes amp Rauer (2015)
74 Planetrsquos composition and formation scenario
With amass of Mp = 0387+0083minus0075 MJ and radius of Rp = 0808+0034
minus0033RJ (resulting in a mean density of ρp=091+024
minus020 g cmminus3) K2-139 bjoins the small group of well characterized warm Jupiters Fig 5shows the position of K2-139 b in themass-radius diagram for warmJupiters (Mp gt 03 MJup 10 6 Porb 6 100 days) whose massand radius have been determined with a precision better than 25(14 objects) Notably K2-139 b is the transiting warm Jupiter withthe lowest mass known to date if the definition of giant planetsgiven by Hatzes amp Rauer (2015) is adopted Fig 5 displays also theplanetary models of Fortney et al (2007) for different core massesand age between 10 and 45Gyrs The planet radius of K2-139 bcan be explained if the planet has a core10 of 49+19
minus17 Moplus containingsim40 of the total planetary mass We expect that K2-139 b has asolid core surrounded by a gaseous envelope
Rafikov (2006) found that a core of mass 5 ndash 20 Moplus at a semi-major axis between 01 and 10AU would be able to start the run-away accretion phase to form a gas giant planet in situ Howeveraccording to his models these kind of cores are unlikely to formowing to the high irradiation coming from the star Boley et al(2016) suggested instead that more massive cores (Mcore amp 20Moplus)can be built up from the merging of tightly packed inner planetsformed at the early stages of the circumstellar disc Batygin et al(2016) found a similar result and argued that the massive core ofHD149026b (Mcore asymp 100Moplus) could be explained by one or moresuper-Earths which merged and accreted the surrounding gas toform a gas-giant planet Huang et al (2016) suggested that thesecores can initiate runaway accretion if they are formed in a re-gion with enough gas around them while those without enoughvolatiles remain super-Earths and represent the population of mas-sive rocky planets unveiled by Kepler around solar-like stars (egDemory 2014) Based on these studies and given the semi-majoraxis of 0179+0021
minus0027 AU the 48 plusmn 14 Moplus core of K2-139 b could
10 Calculated by interpolating Fortney et al (2007)rsquos models
have formed the planet in situ We note that the metallicity of K2-139 is relatively high ([FeH] = 021 plusmn 005) suggesting that theprimordial circumstellar disc had a relatively high content of dustwhich would have enhanced the formation of the core of K2-139 b(see eg Johnson amp Li 2012) Alternatively the planet might haveformed beyond the snow line and migrated inwards via planet-discinteraction (see eg Baruteau et al 2014)
8 CONCLUSIONS
We confirmed the planetary nature and derived the orbital and mainphysical parameters of K2-139 b a warm Jupiter (Teq = 565+48
minus32 K)transiting an active (log RprimeHK = minus446 plusmn 006) K0V star every 29days We measured a planetary mass of Mp = 0387+0083
minus0075 MJ andradius of Rp = 0808+0034
minus0033 RJ At a separation of ap = 0179+0021minus0027
AU the mean density of ρp = 091+024minus020 g cmminus3 implies that the
planet has a core of 49+19minus17 Moplus according to the evolutionary mod-
els of Fortney et al (2007) K2-139 b joins the small group ofwell-characterized warm Jupiters whose mass and radius have beendetermined with a precision better than 25
The spin-orbit angle ie the angle between the spin axis ofthe star and the angular momentum vector of the orbit can pro-vide us with valuable information on the migration mechanismsof exoplanets (see eg Winn 2010 Morton amp Johnson 2011 Al-brecht et al 2012 Gandolfi et al 2012) Currently there are only4 warm Jupiters (Mp gt 03 MJup and 10 6 Porb 6 100 days)with measured obliquity11 From this perspective K2-139 is anideal target to measure the sky-project spin-orbit angle via obser-vations of the Rossiter-McLaughlin (RM) effect Assuming spin-orbit alignement the expected amplitude of the RM anomaly is∆RV asymp
radic1 minus b2 (RpR)2 v sin iasymp 25 m sminus1 (Winn 2010) Given
the brightness of the host star (V = 11653 mag) this amplitudecan easily be measured using state-of-the-art spectrographs suchas HARPSESO-36m Moreover the transit duration (sim5 hours)is shorter than the visibility of K2-139 which is sim9 hours fromLa Silla observatory (altitude higher than 30 above the horizon)
Alternatively the spin-orbit angle could be measured fromthe analysis of the spot-crossing events as described in Sanchis-Ojeda et al (2011) and Sanchis-Ojeda et al (2012) Anomaliesascribable to the passage of K2-139b in front of stellar spots arevisible in the 3 transit light curves observed by K2 Unfortunatelythe limited number of transits and the K2 long cadence data donot allow us to perform a meaningful quantitative analysis of thespot-crossing events Given the amplitude of the detected anoma-lies (sim01) space-based high-precision photometry is needed todetect the spot-crossing events Observations performed with theupcoming CHaracterising ExOPlanets Satellite (CHEOPS Broeget al 2013) would allow us to photometrically determine the spin-orbit angle of this system
ACKNOWLEDGEMENTS
We warmly thank the NOT ESO TNG staff members for theirunique support during the observations We are very thankful toXavier Bonfils Franccedilois Bouchy Martin Kuumlrster Tsevi MazehJorge Melendez and Nuno Santos who kindly agreed to exchange
11 Source httpwww2mpsmpgdehomeshellercontentmain_HRMhtml as of January 2017
MNRAS 000 1ndash12 (2017)
10 O Barragaacuten et al
HARPSandFIES timewith us Special thanks go toAntoninoLanzafor assisting us with the calculation of the gyro-age of the star Wealso greatly thank the anonymous referee for herhis careful re-view and suggestions which helped us to improve the manuscriptD Gandolfi gratefully acknowledges the financial support of theProgramma Giovani Ricercatori ndash Rita Levi Montalcini ndash Rien-tro dei Cervelli (2012) awarded by the Italian Ministry of Edu-cation Universities and Research (MIUR) Sz Csizmadia thanksthe Hungarian OTKA Grant K113117 H J Deeg and D Nespralacknowledge support by grant ESP2015-65712-C5-4-R of the Span-ish Secretary of State for RampDampi (MINECO) D Lorenzo-Oliveiraacknowledges the support from FAPESP (201620667-8) This re-search was supported by the Ministerio de Economia y Competi-tividad under project FIS2012-31079 The research leading to theseresults has received funding from the European Union SeventhFramework Programme (FP72013-2016) under grant agreementNo 312430 (OPTICON) Based on observations obtained a) withthe Nordic Optical Telescope (NOT) operated on the island of LaPalma jointly by Denmark Finland Iceland Norway and Swedenin the Spanish Observatorio del Roque de los Muchachos (ORM)of the Instituto de Astrofiacutesica de Canarias (IAC) b) with the Ital-ian Telescopio Nazionale Galileo (TNG) also operated at the ORM(IAC) on the island of La Palma by the INAF - Fundacioacuten GalileoGalilei c) the 36m ESO telescope at La Silla Observatory underprogramme ID 097C-0948 The data presented here were obtainedin part with ALFOSC which is provided by the Instituto de As-trofisica de Andalucia (IAA) under a joint agreement with the Uni-versity of Copenhagen and NOTSA This paper includes data col-lected by the Kepler mission Funding for the Kepler mission is pro-vided by the NASA Science Mission directorate Some of the datapresented in this paper were obtained from theMikulski Archive forSpace Telescopes (MAST) STScI is operated by the Association ofUniversities for Research in Astronomy Inc under NASA contractNAS5-26555 Support for MAST for non-HST data is provided bythe NASA Office of Space Science via grant NNX09AF08G and byother grants and contracts MF and CMP acknowledge generoussupport from the Swedish National Space Board C Eiroa and IRebollido are supported by Spanish grant AYA2014-55840-P PDacknowledge the support from INAF and Ministero dellrsquoIstruzionedellrsquoUniversitagrave e della Ricerca (MIUR) in the form of the grantldquoPremiale VLT 2012rdquo and ldquoThe Chemical and Dynamical Evolu-tion of the Milky Way and Local Group Galaxiesrdquo This work hasmade use of data from the European Space Agency (ESA) missionGaia (httpwwwcosmosesaintgaia) processed by theGaia Data Processing and Analysis Consortium (DPAC httpwwwcosmosesaintwebgaiadpacconsortium) Fundingfor the DPAC has been provided by national institutions in particu-lar the institutions participating in theGaiaMultilateral Agreement
REFERENCES
Albrecht S et al 2012 ApJ 757 18Allard F Homeier D Freytag B 2011 in Johns-Krull C BrowningM K
West A A eds Astronomical Society of the Pacific Conference SeriesVol 448 16th Cambridge Workshop on Cool Stars Stellar Systemsand the Sun p 91 (arXiv10115405)
Antonini F Hamers A S Lithwick Y 2016 preprint(arXiv160401781)
Barnes S A 2010 ApJ 722 222Barnes S A Kim Y-C 2010 ApJ 721 675Barragaacuten O Gandolfi D 2017 Exotrending Astrophysics Source Code
Library (ascl1706001)
Barragaacuten O et al 2016 AJ 152 193Barragaacuten O Gandolfi D Antoniciello G 2017 pyaneti Astrophysics
Source Code Library (ascl1707003)Baruteau C et al 2014 Protostars and Planets VI pp 667ndash689Batygin K Bodenheimer P H Laughlin G P 2016 ApJ 829 114Boley A C Granados Contreras A P Gladman B 2016 ApJ 817 L17Boyajian T S et al 2013 ApJ 771 40Brahm R et al 2016 AJ 151 89Bressan A Marigo P Girardi L Salasnich B Dal Cero C Rubele S
Nanni A 2012 MNRAS 427 127Broeg C et al 2013 in European Physical Journal Web of Conferences p
03005 (arXiv13052270) doi101051epjconf20134703005Bruntt H et al 2010 MNRAS 405 1907Buchhave L A et al 2010 ApJ 720 1118Burnham K Anderson D 2002 Model Selection and Multimodel Infer-
ence A Practical Information-Theoretic Approach NewYork Springer-Verlag
Cabrera J et al 2009 AampA 506 501Cabrera J Csizmadia S Erikson A Rauer H Kirste S 2012 AampA 548
A44Cabrera J et al 2014 ApJ 781 18Cantat-Gaudin T et al 2014 AampA 562 A10Cardelli J A Clayton G C Mathis J S 1989 ApJ 345 245Carone L et al 2012 AampA 538 A112Carpano S et al 2009 AampA 506 491Castelli F Kurucz R L 2004 preprintCavarroc C et al 2012 ApampSS 337 511Chabrier G 2001 ApJ 554 1274Cosentino R et al 2012 in Ground-based and Airborne Instrumentation
for Astronomy IV p 84461V doi10111712925738Cutri R M et al 2003 2MASS All Sky Catalog of point sourcesCutri R M et al 2012 Technical report Explanatory Supplement to the
WISE All-Sky Data Release ProductsDawson R I Johnson J A Morton T D Crepp J R Fabrycky D C
Murray-Clay R A Howard A W 2012 ApJ 761 163Deeg H J et al 2010 Nature 464 384Demory B-O 2014 ApJ 789 L20Dong S Katz B Socrates A 2014 ApJ 781 L5Doyle A P Davies G R Smalley B Chaplin W J Elsworth Y 2014
MNRAS 444 3592Dumusque X Boisse I Santos N C 2014 ApJ 796 132Endl M Cochran W D 2016 PASP 128 094502Erikson A et al 2012 AampA 539 A14Fabricius C et al 2016 AampA 595 A3Fortney J J Marley M S Barnes J W 2007 ApJ 659 1661Fossati et al 2017 AampA submittedFrandsenS LindbergB 1999 inKarttunenH PiirolaV edsAstrophysics
with the NOT p 71Frewen S F N Hansen B M S 2016 MNRAS 455 1538Gandolfi D et al 2008 ApJ 687 1303Gandolfi D et al 2012 AampA 543 L5Gandolfi D et al 2015 AampA 576 A11Gray R O 1999 SPECTRUM A stellar spectral synthesis program As-
trophysics Source Code Library (ascl9910002)Grziwa S Paumltzold M 2016 preprint (arXiv160708417)Grziwa S Paumltzold M Carone L 2012 MNRAS 420 1045Hamers A S Antonini F Lithwick Y Perets H B Portegies Zwart S F
2016 preprint (arXiv160607438)Hatzes A P 2002 Astronomische Nachrichten 323 392Hatzes A P Rauer H 2015 ApJ 810 L25Heiter U et al 2015 Phys Scr 90 054010Huang C Wu Y Triaud A H M J 2016 ApJ 825 98Jenkins J S et al 2017 MNRAS 466 443Johnson J L Li H 2012 ApJ 751 81Kipping D M 2010 MNRAS 408 1758Kley W Nelson R P 2012 ARAampA 50 211Kovaacutecs G Zucker S Mazeh T 2002 AampA 391 369
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 11
Table 5 K2-139 system parameters
Parameter Prior(a) Final value
Stellar parametersStar mass M (M) middot middot middot 0919 plusmn 0033Star radius R (R) middot middot middot 0862 plusmn 0032Star density ρ (from spectroscopy g cmminus3) middot middot middot 202+025
minus022
Star density ρ (from light curve g cmminus3) middot middot middot 211+074minus081
Effective Temperature Teff (K) middot middot middot 5340 plusmn 110Surface gravity log g (cgs) middot middot middot 450 plusmn 009Iron abundance [FeH] (dex) middot middot middot 022 plusmn 008Microturbulent velocity vmic ( km sminus1) middot middot middot 09 plusmn 01Macroturbulent velocity vmac ( km sminus1) middot middot middot 25 plusmn 06Projected rotational velocity v sin i ( km sminus1) middot middot middot 28 plusmn 06Rotational period Prot (days) middot middot middot 1724 plusmn 012Activity index(b) logRprimeHK middot middot middot minus446 plusmn 006Gyrochronological age (Gyr) middot middot middot 18 plusmn 03Interstellar extinction AV (mag) middot middot middot 007 plusmn 005Star distance d (pc) middot middot middot 152 plusmn 10
Model parameters of K2-139 bOrbital period Porb (days) U[283773 283873] 2838236 plusmn 000026Transit epoch T0 (BJDTDBminus2 450 000) U[73258120 73258220] 732581714 plusmn 000033Scaled semi-major axis aR U[12 100] 448+47
minus67Planet-to-star radius ratio RpR U[0 02] 00961+00023
minus00015Impact parameter b U[0 12] 030+021
minus019radice sinω U[minus1 1](c) 010+029
minus030radice cosω U[minus1 1](c) 006+024
minus027Radial velocity semi-amplitude variation K ( m sminus1) U[0 200] 277+60
Note ndash The adopted Sun and Jupiter units follow the recommendations from the International Astronomical Union (Prša et al 2016) (a) U[a b] refersto uniform priors between a and b N[a b] means Gaussian priors with mean a and standard deviation b and F[a] to a fixed a value (b) Corrected forinterstellar reddening following Fossati et al (2017) The correction is +006 (c) The code always ensures that e lt 1 (d) Assuming albedo = 0
MNRAS 000 1ndash12 (2017)
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
MNRAS 000 1ndash12 (2017)
1 Introduction
2 K2 photometry
3 ALFOSC imaging
4 High-resolution spectroscopy
5 Stellar parameters
51 Spectral analysis
52 Interstellar extinction
53 Rotational period
54 Stellar mass radius and age
6 Joint RV-transit fit
7 Results and discussion
71 Stellar activity modeling
72 Additional companion
73 Spot-crossing events
74 Planets composition and formation scenario
8 Conclusions
Acknowledgements
4 O Barragaacuten et al
Table 2 Radial velocity measurements and activity indexes of K2-139
tra are nearly identical and reveal a star with Teff = 5500plusmn 100 Klog g = 465plusmn 012 (cgs) [FeH]=+011plusmn 012 dex and a slow pro-jected rotational velocity of v sin iasymp 2 km sminus1
The high-precision radial velocity follow-up of K2-139 wasstarted in June 2016with the FIbre-fed Eacutechelle Spectrograph (FIESFrandsenampLindberg 1999 Telting et al 2014)mounted at the 256-m Nordic Optical Telescope (NOT) The observations were carriedout as part of the OPTICON andCAT observing programs 16A055P53-201 and P53-203 We used the high-resmode which providesa resolving power of Rasymp 67 000 in the whole visible spectral range(3700 minus 7300Aring) The exposure time was set to 2100 ndash 3600 secbased on sky conditions and observing scheduling constraints Fol-lowing the observing strategy outlined in Buchhave et al (2010)and Gandolfi et al (2015) we traced the RV drift of the instrumentby acquiring long-exposed (Texp asymp 35 sec) ThAr spectra immedi-ately before and after the target observations The typical RV driftmeasured between two ThAr spectra bracketing a 2100 ndash 3600 secscience exposure is about 50 ndash 80 m sminus1 A linear interpolation ofthe RV drift to the mid-time of the science exposure allows us toachieve a radial velocity zero-point stability of about 5 ndash 6 m sminus1which is 2 ndash 3 times smaller than the nominal error bars listed in Ta-ble 2 The data reduction uses standard IRAF and IDL routines Thesignal-to-noise (SN) ratio of the extracted spectra is sim30 ndash 40 perpixel at 5500 Aring Radial velocity measurements were extracted viamulti-order cross-correlation with the RV standard star HD182572observed with the same instrument set-up as K2-139
We also observed K2-139 in July August and September 2016with the HARPS (Mayor et al 2003) and HARPS-N (Cosentinoet al 2012) spectrographs mounted at the ESO 36-m Telescope ofLa SillaObservatory (Chile) and at the 358-mTelescopioNazionaleGalileo (TNG) of Roque de los Muchachos observatory (La PalmaSpain) respectively Both instruments provide a resolving powerof Rasymp 115 000 in the wavelength range sim3800 ndash 6900 Aring The ob-
servations were performed as part of the ESO and TNG observingprograms 097C-0948 and A33TAC_15 respectively The exposuretime was set to 1800 sec leading to a SN ratio of sim35 on the ex-tracted spectra We reduced the data using the dedicated HARPSand HARPS-N pipelines and extracted the RVs by cross-correlationwith a G2 numerical mask
The FIES HARPS and HARPS-N RVs are listed in Ta-ble 2 along with the bisector span (BIS) and the full width at halfmaximum (FWHM) of the cross-correlation function (CCF) Timestamps are given in barycentric Julian date in barycentric dynam-ical time (BJDTDB) For the HARPS and HARPS-N data we alsoprovide the Ca ii HampK chromospheric activity index log RprimeHK Wedid not measure log RprimeHK from the FIES spectra because of the poorSN ratio at wavelengths shorter than 4000Aring
5 STELLAR PARAMETERS
51 Spectral analysis
We derived the spectroscopic parameters of K2-139 from the co-added FIES spectra The stacked FIES data have a SN ratio of sim110per pixel at 5500 Aring We adopted three different methods For eachmethod results are reported in Table 3
First method The technique fits spectral features that are sen-sitive to different photospheric parameters It uses the stellar spec-tral synthesis program Spectrum (Gray 1999) to compute syntheticspectra from ATLAS 9model atmospheres (Castelli ampKurucz 2004)Microturbulent (vmic) and macroturbulent (vmac) velocities are de-rived from the calibration equations of Bruntt et al (2010) andDoyle et al (2014) We used the wings of the Hα and Hβ lines toestimate the effective temperature (Teff) and the Mg i 5167 5173and 5184 Aring Ca i 6162 and 6439 Aring and the Na iD lines to determine
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 5
Table 3 Spectroscopic parameters of K2-139 as derived using the three methods described in Sect 5
Method Teff (K) log g (cgs) [FeH] (dex) vmic ( km sminus1) vmac ( km sminus1) v sin i ( km sminus1)
the surface gravity log g We simultaneously fitted different spec-tral regions to measure the iron abundance [FeH] The projectedrotational velocity v sin i was determined by fitting the profile ofmany isolated and unblended metal lines
Second method It relies on the use of the spectral analysispackage Spectroscopy Made Easy (SME Valenti amp Piskunov 1996Valenti amp Fischer 2005) For a set of given stellar parameters SMEcalculates synthetic spectra and fits them to high-resolution ob-served spectra using a chi-squared minimization procedure Weused SME version 443 and ATLAS 12model spectra (Kurucz 2013)We adopted the same calibration equation as described in the firstmethod to determine vmic and vmac Effective temperature is de-rived from the Hα wings log g from the Ca i 6102 6122 6162and 6439 Aring lines [FeH] and v sin i from isolated iron lines
Third method It uses the classical equivalent width (EW)method adopting the following criteria i) Teff is obtained by re-moving trends between abundance of the chemical elements andthe respective excitation potentials ii) log g is optimised by as-suming the ionisation equilibrium condition ie by requiring thatfor a given species the same abundance (within the uncertainties)is obtained from lines of two ionisation states (typically neutral andsingly ionised lines) iii) vmic is set by minimising the slope of therelationship between abundance and the logarithm of the reducedEWs The equivalent widths of Fe i and Fe ii lines are measuredusing the code DOOp (Cantat-Gaudin et al 2014) a wrapper ofDAOSPEC (Stetson amp Pancino 2008) The stellar atmosphere param-eters are derived with the program FAMA (Magrini et al 2013) awrapper of MOOG (Sneden et al 2012) We used the public versionof the atomic data prepared for the Gaia-ESO Survey (Heiter et al2015) and based on the VALD3 data (Ryabchikova et al 2011) Weused sim200 Fe i lines and sim10 Fe ii lines for the determination of thestellar parameters
The three methods provide consistent results within the 1-σerror bars (Table 5) While we have no reason to prefer one tech-nique over the other we adopted the parameter estimates of thefirst method ie Teff = 5340 plusmn 110 K log g = 450 plusmn 009 (cgs)[FeH] = 022 plusmn 008 dex vmic = 09 plusmn 01 km sminus1 vmac = 25 plusmn06 km sminus1 and v sin i = 28 plusmn 06 km sminus1 As a sanity check wealso analyzed the HARPS and HARPS-N data and obtained con-sistent results but with larger error bars owing to the lower SNratio of the co-added HARPS and HARPS-N spectra compared tothat of the co-added FIES data Using the Boyajian et al (2013)rsquoscalibration (see their Table 6) the effective temperature of K2-139defines the spectral type of the host star as K0V
52 Interstellar extinction
We measured the visual reddening (AV) of K2-139 following thetechnique described in Gandolfi et al (2008) We fitted the spectral
energy distribution of the star to synthetic colors extracted from theBT-NEXTGEN model spectrum (Allard et al 2011) with the samephotospheric parameters as the star We adopted the extinction lawof Cardelli et al (1989) and assumed a normal value for the total-to-selective extinction ie RV = AVE(B minus V)= 31 We measureda visual extinction of AV = 007plusmn 005 mag This value is belowthe upper limit of AV 03 mag extracted from the Schlegel et al(1998)rsquos all-sky extinction map corroborating our result
53 Rotational period
The K2 light curve of K2-139 displays periodic and quasi-periodicvariations with a peak-to-peak photometric amplitude of sim2(Fig 1) The late-type spectral type of the star suggests that theobserved variability is due to Sun-like spots appearing and disap-pearing from the visible stellar disc as the star rotates around its axisThis is corroborated by the fact that K2-139 is a chromosphericallyactive star The HARPS and HARPS-N spectra show clear emis-sion components in the cores of the Ca ii HampK lines from whichwe measured an average activity index of log RprimeHK =minus446plusmn 0064
The out-of-transit photometric variability observed in the lightcurve of K2-139 is mainly due to two active regions located at op-posite stellar longitudes whose lifetime is longer than the durationof the K2 observations Using the spots as tracers of stellar rotationand following the auto correlation function (ACF) technique de-scribed in McQuillan et al (2014) we estimated that the rotationalperiod of the star is Prot = 1724 plusmn 012 days The Lomb-Scargleperiodogram of the light curve shows its strongest peak at the sameperiod confirming our results
It is worth noting that the rotation period (Prot = 1724 plusmn012 days) and radius (R= 0862 plusmn 0032 R see next section)of the host star translate into a maximum value for the projected ro-tational velocity of v sin imax = 253plusmn 010 km sminus1 which agreeswith the spectroscopically derived v sin i = 28 plusmn 06 km sminus1 sug-gesting that the star is seen nearly equator-on (i asymp 90) and thatthe system might be aligned along the line-of-sight
54 Stellar mass radius and age
We derived the stellar mass radius and age using the online inter-face for Bayesian estimation of stellar parameters available at thefollowing web page httpstevoapdinafitcgi-binparam Briefly the web tool interpolates onto PARSEC modelisochrones (Bressan et al 2012) the V-band apparent magnitude
4 This value is corrected for the interstellar medium absorption followingthe procedure described in Fossati et al (2017) and using themeasured stellarparameters and reddening The correction is +006 The star is thereforeslightly more active than what measured from the spectra
MNRAS 000 1ndash12 (2017)
6 O Barragaacuten et al
effective temperature metal content and parallax We used the V-band magnitude reported in Table 1 ndash after correcting for interstel-lar reddening (Sect 52) ndash along with the effective temperature andmetal content we derived in Sect 5 The parallax was retrieved fromthe Gaiarsquos first data release (px = 656 plusmn 043mas d = 152plusmn 10 pcFabricius et al 2016) We adopted the log-normal initial mass func-tion from Chabrier (2001)
K2-139 has a mass of M= 0919 plusmn 0033 M and radiusof R= 0862 plusmn 0032 R corresponding to a surface gravity oflog g = 4503plusmn0035 (cgs) in excellent agreement with the spec-troscopically derived value of log g = 450plusmn009 (cgs see Sect 5)The derivedmean density ρ = 202plusmn024 g cmminus3 ofK2-139 is alsoconsistent within 1-σ with the density estimated by the modelingof the transit light curve (ρ = 211+074
minus081 g cmminus3 see Sect 6)The isochrones provide an age of 36plusmn34 Gyr for K2-139 Us-
ing the equations given in Barnes amp Kim (2010) and Barnes (2010)the rotation period of 173 days (Sect 53) implies a gyrochronolog-ical age of 18plusmn 03 Gyr
6 JOINT RV-TRANSIT FIT
We performed the joint fit to the photometric and RV data using thecode pyaneti5 (Barragaacuten et al 2017) a PythonFortran softwaresuite based on Markov Chain Monte Carlo (MCMC) methods
The photometric data included in the joint analysis are subsetsof the whole EVEREST K2 light curve We used the EVERESTlight curve because it provides a slightly better rms over the Vander-burg amp Johnson (2014)rsquos dataWe selectedsim10 hours of data-pointsaround each of the 3 transits which have a duration of sim5 hours Wede-trended each individual transits with the code exotrending6
(Barragaacuten amp Gandolfi 2017) using a second-order polynomial fit-ted to the out-of-transit points The fitted data include 12 pointsimmediately before and after each transit with the exception of thelast transit for which only 9 data points are available We removedthe data points that are affected by stellar spot crossing events (seeSect 71 for more details)
We fitted the RV data using a Keplerian model for the planetalong with two sine-like curves to account for the activity-inducedRV (see next section for details) We adopted the limb-darkenedquadratic law of Mandel amp Agol (2002) for the transit model Weadopted the Gaussian likelihood
L =[
nprodi=1
2π
(σ2i + σ
2j
)minus12]
exp
minus
nsumi=1
12(Di minus Mi)2
σ2i+ σ2
j
(1)
where n is the number of data points σi is the error associatedto each data point Di Mi is the model associated to a given Di andσj is an extra noise term sometime referred as jitter
The sampling method and fitted parameters are the same as inBarragaacuten et al (2016) Details on the adopted priors are given in Ta-ble 5 Following Kipping (2010) we super-sampled the light curvemodel using 10 subsamples perK2 exposure to account for the long-cadence acquisition The parameter space was explored with 500independent chains created randomly inside the prior ranges Thechain convergence was analyzed using the Gelman-Rubin statisticsThe burn-in phase uses 25 000 more iterations with a thin factor
5 Available at httpsgithubcomoscaribvpyaneti6 Available at httpsgithubcomoscaribvexotrending
of 50 The posterior distribution of each parameter has 250 000independent data points
7 RESULTS AND DISCUSSION
71 Stellar activity modeling
A simple Keplerian model provides a poor fit to the RV measure-ments with χ2dof = 61 (Table 4) suggesting that additionalsignals might be present in our Doppler data Activity-induced RVvariation is expected given the 2 peak-to-peak photometric vari-ability observed in the K2 light curve of K2-139 (Fig 1) and theCa ii HampK activity index of log RprimeHK =minus446 plusmn 006 (Sect 53)The K2 photometric variation corresponds to a spot filling factorof approximately 2 if this variation is due to cool starspots Wecan use the empirical relationship relating spot coverage to RV am-plitude from Saar amp Donahue (1997) or Hatzes (2002) to estimatethe RV amplitude expected from spots Using the projected rota-tional velocity of 28 km sminus1 results in an RV semi-amplitude ofasymp20ndash30 m sminus1 The code SOAP2 designed to estimate the effectof active regions on photometric and spectroscopic measurements(Dumusque et al 2014) provides consistent results
In order to look for additional signals in our Doppler data weperformed a frequency analysis of theRVmeasurements and activityindicators On one occasion7 K2-139 was observed with FIES andHARPS-S nearly simultaneously (within less than 25 minutes) Weused the two sets of measurements to estimate the RV FWHMand BIS offsets between the two instruments We assumed no offsetbetween HARPS-N and HARPS While we acknowledge that thisassumption is arbitrary we note that the modeling of the RV datagives an offset of ∆RV(HNminusH) = 0002plusmn 00158 km sminus1 (Table 5)which is consistent with zero
Figure 3 displays the generalized Lomb-Scargle periodograms(Zechmeister amp Kuumlrster 2009) of the combined datasets From topto bottom the RVdata (first panel) the RV residuals after subtractingthe transiting planet signal (second panel) and the BIS (third panel)and FWHM (fourth panel) of the cross correlation function Theperiodogram of the window function is shown in the lower panelThe dotted vertical blue lines mark the frequency at the orbitalperiod of the planet (0035 cd) as well as the frequencies at therotation period of the star (0058 cd) and its first two harmonics(0116 and 0174 cd)
The periodogram of the RV data (upper panel) shows a peak atthe orbital frequency of the planet alongwith two additional peaks at0095 and 0130 cd Since the periodogram of the window functionshows two peaks atsim0060 andsim0095 cd (lower panel red arrows)we interpreted the 0095 and 0130 cd peaks as the aliases of theorbital frequency8 We note also that periodogram of the BIS ofthe CCF displays peaks whose frequencies are close to the stellarrotation frequency and its first two harmonics However none of thepeaks visible in the GLS periodograms of Fig 3 has a false alarmprobability (FAP)9 lower than 5 Although our spectroscopic datashow neither additional signals we note that the semi-amplitudevariation of the BIS and FWHM is expected to be 10-15 m sminus1
(Dumusque et al 2014) which is comparable with the uncertaintiesof most of ourmeasurements (Table 2) The lack of significant peaks
7 Epoch BJD=24575898 0095 = 0035 + 0060 cd and 0130 = 0035 + 0095 cd9 We determined the FAP following the Monte Carlo bootstrap methoddescribed in Kuerster et al (1997)
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 7
Figure 3 Generalized Lomb-Scargle periodogram of the combined FIESHARPS and HARPS-N Doppler datasets From top to bottom the RV datathe RV residuals after subtracting the transiting planet signal the BIS andFWHM of the CCF and the window function The dotted vertical blue linesmark the frequencies at the orbital period as well as at the stellar rotationperiod and its first two harmonics The dashed vertical red lines mark the5 false alarm probabilities as derived using the bootstrap method The redarrows in the lower panel mark the two peaks presented in the main text
in the periodogram of the RV data and RV residuals as well as inthe periodogram of the activity indicators could be explained by thelimited number of available measurements and their uncertaintiesWe conclude that we cannot exclude the existence of spot-inducedsignals in our RV measurements
Photometric and radial velocity variations due to rotationalmodulation can be complex with not only the rotational period Protpresent but also its harmonics eg Prot2 Prot3 Assuming thatthe surface structures responsible for this modulation (eg coolspots) are not evolving rapidly then the simplest representation ofthe rotationalmodulation is through the Fourier components definedby the rotation period and its harmonics Figure 1 shows that theevolution time-scale of the active regions in the stellar surface islonger than the 80-day duration of the K2 campaign Since our RVfollow-up spans 55 days we can assume that any activity-inducedRV signal is coherent within our observing window This approachhas been used previously for other planetary systems orbiting activestars (eg Pepe et al 2013)
The Fourier analysis of the K2 light curve is the best way tomeasure the contribution of the rotation period and its harmonics tothe quasi-periodic photometric variability of the star We thereforeanalyzed the K2 light curve using a pre-whitening procedure Thatis the dominant period was found a sine-fit made to the data andsubtracted and additional periods searched in the residual data
We used the program Period04 (Lenz amp Breger 2005) for thisprocedure
The dominant periods are sim172 days ie the rotation periodof the star (Sect 53) and roughly the first four harmonics (ie 8657 43 and 34 days) The 172- and 86-day periods have about thesame amplitude while the 57-day period (Prot3) has 10 of themain amplitude The Prot4 signal has only about 4 of the mainamplitude The light curve analysis indicates that the signal due torotational modulation can largely be represented by the rotationalperiod (Prot) and its first harmonic (Prot2)
In order to test if the addition of RV sinusoidal signals atthe stellar rotation period and its harmonics can account for theadditional variation seen in our RV measurements we compareddifferent models by adding signals one by one The first model (P0)includes only the planet signal ie a Keplerian model fitted to theRV data using the same priors given in Table 5 but fixing epoch andperiod to the values derived by the transit modeling The next model(P1) is obtained from P0 by adding a sinusoidal signal at the rotationperiod of the star (Prot) Models P2 includes the first harmonic ofthe rotation period (Prot2) whereas model P3 account for the first(Prot2) and second (Prot3) harmonics While adding sinusoidalsignals we fitted for their amplitudes phases and periods We usedflat priors for the phases and amplitudes (details in Table 5) Weused a Gaussian prior for Prot using the value and its uncertaintyderived in Sect 53 The periods of the harmonic signals were leftfree to vary depending on the value assumed by Prot at each stepof the MCMC chains In order to check if the RV variation inducedby the planet is significant in our data set we also performed the fitusing models where the planetary signal was not included (modelsNP1 and NP2 see Table 4)
Table 4 shows the goodness of the fit for each model Thepreferred model is P2 (planet plus 2 sinusoidal signals at Prot andProt2) with the lowest Akaike Information Criteria (AIC) and max-imum likelihood This result is consistent with the Fourier analysisof the K2 light curve which suggests that the major contributionto the photometric variations arises from the stellar rotation periodand its first harmonic Our analysis provides also additional evi-dence that the Doppler motion induced by the planet is present inour RV data set First the planet signal does not significantly varyfor the P0 P1 P2 and P3 models (Table 4) Second the modelswith no planetary signal (NP1 and NP2) provide a poor fit to theRV measurements (Table 4)
To account for additional instrumental noise not included inthe nominal RV error bars andor imperfect treatment of the varioussources of RV variations we fitted for a jitter term for each instru-ment The final parameter estimates and their error bars are listedin Table 5 They are defined as the median and the 68 credibleinterval of the final posterior distributions The best fitting transitand RVmodels are displayed in Figure 4 along with the photometricand RV data points
72 Additional companion
Huang et al (2016) found that warm Jupiters with low eccentricities(e 04) have inner low-mass companions They used this evidenceas an argument in favour of the in situ formation since the planetmigration would have cleaned the warm Jupiter neighborhood Wesearched the light curve for additional transit signals but foundno evidence for an additional transiting planet in the system Asdescribed in the previous paragraph the periodogram of the RVresiduals showno significant peakwith false alarmprobability lowerthan 5
Figure 4 Top FIES (blue circles) HARPS-N (green diamonds) and HARPS (red squares) RV measurements versus time following the subtraction of thesystemic velocities for each instrument The 1σ uncertainties are marked using the same color used for each data-set The vertical gray lines mark the errorbars including jitter The solid line represents the best fitting RV model which includes the planet signal and the activity signal at the stellar rotation periodand its first harmonic The dashed dash-dotted and dotted lines show the RV contribution of K2-139 b stellar rotation and first harmonic respectively Lowerleft panel Transit light curve folded to the orbital period of K2-139 b and residuals The red points mark the K2 data and their error bars The solid line markthe re-binned best-fitting transit model Lower right panel Phase-folded RV curve of K2-139 b and best fitting Keplerian solution (solid line) following thesubtraction of the two additional sinusoidal signals used to account for the stellar activity The FIES HARPS and HARPS-N are corrected for the instrumentoffsets as derived from the global analysis
Table 4 Model comparison
Model Comment Npars Kb (m sminus1) χ2dof(a) ln L AIC(b)
P0 Planet signal 6 291 plusmn 20 61 356 -60P1 Planet signal + 1 sine-curve at Prot 9 294 plusmn 24 34 581 -98
P2 Planet signal + 2 sine-curves at Prot and Prot2 11 273+26minus25 38 601 -98
P3 Planet signal + 3 sine-curves at Prot Prot2 and Prot3 13 278+27minus26 53 593 -93
NP1 1 sine-curve at Prot (No planet signal) 6 0 185 -448 101NP2 2 sine-curves at Prot and Prot2 (No planet signal) 8 0 159 -120 40
Note ndash (a) χ2 value assuming no jitter (b)We used the Akaike Information Criteria (AIC = 2Nparsminus ln 2L) instead of the widely used Bayesian informationcriteria (BIC) because our RV data sample is small (19 data points) and BIC performs better for large samples (Burnham amp Anderson 2002)
73 Spot-crossing events
The passage of a planet in front of a spot can be detected as abump in the transit light curve (see eg Sanchis-Ojeda amp Winn2011) Spot-crossings events are clearly visible in the EVERESTtransit light curves (Fig 4) The same features appear at the sametimes and with consistent amplitudes in the Vanderburg amp Johnson(2014) data confirming that the bumps are real and not due to
systematics To assess whether the bumps significantly affect theparameter estimates we performed the joint analysis as describedin Sect 6 including all the transit data points We found that the finalparameters are consistent within 1-σ with those reported in Table 5
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 9
01 05 1 2 3 4 5 10Mass (MJup)
06
08
10
12
14
Rad
ius
(RJu
p)
H 0 M core 10 M core 25 M core 50 M core100 M core
Figure 5 Warm Jupiters (black squares Mp gt 03 MJup and10 6 Porb 6 100 days) whose mass and radius have been estimated with aprecision of at least 25 (as of January 2017 exoplaneteu) K2-139 b isshown with a filled red circle The solid line corresponds to a planet with apure hydrogen composition (Seager et al 2007) The dashed lines representthe Fortney et al (2007) models for planet core masses of 0 10 25 50and 100 Moplus The vertical dotted line marks the giant planet lower limit asdefined by Hatzes amp Rauer (2015)
74 Planetrsquos composition and formation scenario
With amass of Mp = 0387+0083minus0075 MJ and radius of Rp = 0808+0034
minus0033RJ (resulting in a mean density of ρp=091+024
minus020 g cmminus3) K2-139 bjoins the small group of well characterized warm Jupiters Fig 5shows the position of K2-139 b in themass-radius diagram for warmJupiters (Mp gt 03 MJup 10 6 Porb 6 100 days) whose massand radius have been determined with a precision better than 25(14 objects) Notably K2-139 b is the transiting warm Jupiter withthe lowest mass known to date if the definition of giant planetsgiven by Hatzes amp Rauer (2015) is adopted Fig 5 displays also theplanetary models of Fortney et al (2007) for different core massesand age between 10 and 45Gyrs The planet radius of K2-139 bcan be explained if the planet has a core10 of 49+19
minus17 Moplus containingsim40 of the total planetary mass We expect that K2-139 b has asolid core surrounded by a gaseous envelope
Rafikov (2006) found that a core of mass 5 ndash 20 Moplus at a semi-major axis between 01 and 10AU would be able to start the run-away accretion phase to form a gas giant planet in situ Howeveraccording to his models these kind of cores are unlikely to formowing to the high irradiation coming from the star Boley et al(2016) suggested instead that more massive cores (Mcore amp 20Moplus)can be built up from the merging of tightly packed inner planetsformed at the early stages of the circumstellar disc Batygin et al(2016) found a similar result and argued that the massive core ofHD149026b (Mcore asymp 100Moplus) could be explained by one or moresuper-Earths which merged and accreted the surrounding gas toform a gas-giant planet Huang et al (2016) suggested that thesecores can initiate runaway accretion if they are formed in a re-gion with enough gas around them while those without enoughvolatiles remain super-Earths and represent the population of mas-sive rocky planets unveiled by Kepler around solar-like stars (egDemory 2014) Based on these studies and given the semi-majoraxis of 0179+0021
minus0027 AU the 48 plusmn 14 Moplus core of K2-139 b could
10 Calculated by interpolating Fortney et al (2007)rsquos models
have formed the planet in situ We note that the metallicity of K2-139 is relatively high ([FeH] = 021 plusmn 005) suggesting that theprimordial circumstellar disc had a relatively high content of dustwhich would have enhanced the formation of the core of K2-139 b(see eg Johnson amp Li 2012) Alternatively the planet might haveformed beyond the snow line and migrated inwards via planet-discinteraction (see eg Baruteau et al 2014)
8 CONCLUSIONS
We confirmed the planetary nature and derived the orbital and mainphysical parameters of K2-139 b a warm Jupiter (Teq = 565+48
minus32 K)transiting an active (log RprimeHK = minus446 plusmn 006) K0V star every 29days We measured a planetary mass of Mp = 0387+0083
minus0075 MJ andradius of Rp = 0808+0034
minus0033 RJ At a separation of ap = 0179+0021minus0027
AU the mean density of ρp = 091+024minus020 g cmminus3 implies that the
planet has a core of 49+19minus17 Moplus according to the evolutionary mod-
els of Fortney et al (2007) K2-139 b joins the small group ofwell-characterized warm Jupiters whose mass and radius have beendetermined with a precision better than 25
The spin-orbit angle ie the angle between the spin axis ofthe star and the angular momentum vector of the orbit can pro-vide us with valuable information on the migration mechanismsof exoplanets (see eg Winn 2010 Morton amp Johnson 2011 Al-brecht et al 2012 Gandolfi et al 2012) Currently there are only4 warm Jupiters (Mp gt 03 MJup and 10 6 Porb 6 100 days)with measured obliquity11 From this perspective K2-139 is anideal target to measure the sky-project spin-orbit angle via obser-vations of the Rossiter-McLaughlin (RM) effect Assuming spin-orbit alignement the expected amplitude of the RM anomaly is∆RV asymp
radic1 minus b2 (RpR)2 v sin iasymp 25 m sminus1 (Winn 2010) Given
the brightness of the host star (V = 11653 mag) this amplitudecan easily be measured using state-of-the-art spectrographs suchas HARPSESO-36m Moreover the transit duration (sim5 hours)is shorter than the visibility of K2-139 which is sim9 hours fromLa Silla observatory (altitude higher than 30 above the horizon)
Alternatively the spin-orbit angle could be measured fromthe analysis of the spot-crossing events as described in Sanchis-Ojeda et al (2011) and Sanchis-Ojeda et al (2012) Anomaliesascribable to the passage of K2-139b in front of stellar spots arevisible in the 3 transit light curves observed by K2 Unfortunatelythe limited number of transits and the K2 long cadence data donot allow us to perform a meaningful quantitative analysis of thespot-crossing events Given the amplitude of the detected anoma-lies (sim01) space-based high-precision photometry is needed todetect the spot-crossing events Observations performed with theupcoming CHaracterising ExOPlanets Satellite (CHEOPS Broeget al 2013) would allow us to photometrically determine the spin-orbit angle of this system
ACKNOWLEDGEMENTS
We warmly thank the NOT ESO TNG staff members for theirunique support during the observations We are very thankful toXavier Bonfils Franccedilois Bouchy Martin Kuumlrster Tsevi MazehJorge Melendez and Nuno Santos who kindly agreed to exchange
11 Source httpwww2mpsmpgdehomeshellercontentmain_HRMhtml as of January 2017
MNRAS 000 1ndash12 (2017)
10 O Barragaacuten et al
HARPSandFIES timewith us Special thanks go toAntoninoLanzafor assisting us with the calculation of the gyro-age of the star Wealso greatly thank the anonymous referee for herhis careful re-view and suggestions which helped us to improve the manuscriptD Gandolfi gratefully acknowledges the financial support of theProgramma Giovani Ricercatori ndash Rita Levi Montalcini ndash Rien-tro dei Cervelli (2012) awarded by the Italian Ministry of Edu-cation Universities and Research (MIUR) Sz Csizmadia thanksthe Hungarian OTKA Grant K113117 H J Deeg and D Nespralacknowledge support by grant ESP2015-65712-C5-4-R of the Span-ish Secretary of State for RampDampi (MINECO) D Lorenzo-Oliveiraacknowledges the support from FAPESP (201620667-8) This re-search was supported by the Ministerio de Economia y Competi-tividad under project FIS2012-31079 The research leading to theseresults has received funding from the European Union SeventhFramework Programme (FP72013-2016) under grant agreementNo 312430 (OPTICON) Based on observations obtained a) withthe Nordic Optical Telescope (NOT) operated on the island of LaPalma jointly by Denmark Finland Iceland Norway and Swedenin the Spanish Observatorio del Roque de los Muchachos (ORM)of the Instituto de Astrofiacutesica de Canarias (IAC) b) with the Ital-ian Telescopio Nazionale Galileo (TNG) also operated at the ORM(IAC) on the island of La Palma by the INAF - Fundacioacuten GalileoGalilei c) the 36m ESO telescope at La Silla Observatory underprogramme ID 097C-0948 The data presented here were obtainedin part with ALFOSC which is provided by the Instituto de As-trofisica de Andalucia (IAA) under a joint agreement with the Uni-versity of Copenhagen and NOTSA This paper includes data col-lected by the Kepler mission Funding for the Kepler mission is pro-vided by the NASA Science Mission directorate Some of the datapresented in this paper were obtained from theMikulski Archive forSpace Telescopes (MAST) STScI is operated by the Association ofUniversities for Research in Astronomy Inc under NASA contractNAS5-26555 Support for MAST for non-HST data is provided bythe NASA Office of Space Science via grant NNX09AF08G and byother grants and contracts MF and CMP acknowledge generoussupport from the Swedish National Space Board C Eiroa and IRebollido are supported by Spanish grant AYA2014-55840-P PDacknowledge the support from INAF and Ministero dellrsquoIstruzionedellrsquoUniversitagrave e della Ricerca (MIUR) in the form of the grantldquoPremiale VLT 2012rdquo and ldquoThe Chemical and Dynamical Evolu-tion of the Milky Way and Local Group Galaxiesrdquo This work hasmade use of data from the European Space Agency (ESA) missionGaia (httpwwwcosmosesaintgaia) processed by theGaia Data Processing and Analysis Consortium (DPAC httpwwwcosmosesaintwebgaiadpacconsortium) Fundingfor the DPAC has been provided by national institutions in particu-lar the institutions participating in theGaiaMultilateral Agreement
REFERENCES
Albrecht S et al 2012 ApJ 757 18Allard F Homeier D Freytag B 2011 in Johns-Krull C BrowningM K
West A A eds Astronomical Society of the Pacific Conference SeriesVol 448 16th Cambridge Workshop on Cool Stars Stellar Systemsand the Sun p 91 (arXiv10115405)
Antonini F Hamers A S Lithwick Y 2016 preprint(arXiv160401781)
Barnes S A 2010 ApJ 722 222Barnes S A Kim Y-C 2010 ApJ 721 675Barragaacuten O Gandolfi D 2017 Exotrending Astrophysics Source Code
Library (ascl1706001)
Barragaacuten O et al 2016 AJ 152 193Barragaacuten O Gandolfi D Antoniciello G 2017 pyaneti Astrophysics
Source Code Library (ascl1707003)Baruteau C et al 2014 Protostars and Planets VI pp 667ndash689Batygin K Bodenheimer P H Laughlin G P 2016 ApJ 829 114Boley A C Granados Contreras A P Gladman B 2016 ApJ 817 L17Boyajian T S et al 2013 ApJ 771 40Brahm R et al 2016 AJ 151 89Bressan A Marigo P Girardi L Salasnich B Dal Cero C Rubele S
Nanni A 2012 MNRAS 427 127Broeg C et al 2013 in European Physical Journal Web of Conferences p
03005 (arXiv13052270) doi101051epjconf20134703005Bruntt H et al 2010 MNRAS 405 1907Buchhave L A et al 2010 ApJ 720 1118Burnham K Anderson D 2002 Model Selection and Multimodel Infer-
ence A Practical Information-Theoretic Approach NewYork Springer-Verlag
Cabrera J et al 2009 AampA 506 501Cabrera J Csizmadia S Erikson A Rauer H Kirste S 2012 AampA 548
A44Cabrera J et al 2014 ApJ 781 18Cantat-Gaudin T et al 2014 AampA 562 A10Cardelli J A Clayton G C Mathis J S 1989 ApJ 345 245Carone L et al 2012 AampA 538 A112Carpano S et al 2009 AampA 506 491Castelli F Kurucz R L 2004 preprintCavarroc C et al 2012 ApampSS 337 511Chabrier G 2001 ApJ 554 1274Cosentino R et al 2012 in Ground-based and Airborne Instrumentation
for Astronomy IV p 84461V doi10111712925738Cutri R M et al 2003 2MASS All Sky Catalog of point sourcesCutri R M et al 2012 Technical report Explanatory Supplement to the
WISE All-Sky Data Release ProductsDawson R I Johnson J A Morton T D Crepp J R Fabrycky D C
Murray-Clay R A Howard A W 2012 ApJ 761 163Deeg H J et al 2010 Nature 464 384Demory B-O 2014 ApJ 789 L20Dong S Katz B Socrates A 2014 ApJ 781 L5Doyle A P Davies G R Smalley B Chaplin W J Elsworth Y 2014
MNRAS 444 3592Dumusque X Boisse I Santos N C 2014 ApJ 796 132Endl M Cochran W D 2016 PASP 128 094502Erikson A et al 2012 AampA 539 A14Fabricius C et al 2016 AampA 595 A3Fortney J J Marley M S Barnes J W 2007 ApJ 659 1661Fossati et al 2017 AampA submittedFrandsenS LindbergB 1999 inKarttunenH PiirolaV edsAstrophysics
with the NOT p 71Frewen S F N Hansen B M S 2016 MNRAS 455 1538Gandolfi D et al 2008 ApJ 687 1303Gandolfi D et al 2012 AampA 543 L5Gandolfi D et al 2015 AampA 576 A11Gray R O 1999 SPECTRUM A stellar spectral synthesis program As-
trophysics Source Code Library (ascl9910002)Grziwa S Paumltzold M 2016 preprint (arXiv160708417)Grziwa S Paumltzold M Carone L 2012 MNRAS 420 1045Hamers A S Antonini F Lithwick Y Perets H B Portegies Zwart S F
2016 preprint (arXiv160607438)Hatzes A P 2002 Astronomische Nachrichten 323 392Hatzes A P Rauer H 2015 ApJ 810 L25Heiter U et al 2015 Phys Scr 90 054010Huang C Wu Y Triaud A H M J 2016 ApJ 825 98Jenkins J S et al 2017 MNRAS 466 443Johnson J L Li H 2012 ApJ 751 81Kipping D M 2010 MNRAS 408 1758Kley W Nelson R P 2012 ARAampA 50 211Kovaacutecs G Zucker S Mazeh T 2002 AampA 391 369
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 11
Table 5 K2-139 system parameters
Parameter Prior(a) Final value
Stellar parametersStar mass M (M) middot middot middot 0919 plusmn 0033Star radius R (R) middot middot middot 0862 plusmn 0032Star density ρ (from spectroscopy g cmminus3) middot middot middot 202+025
minus022
Star density ρ (from light curve g cmminus3) middot middot middot 211+074minus081
Effective Temperature Teff (K) middot middot middot 5340 plusmn 110Surface gravity log g (cgs) middot middot middot 450 plusmn 009Iron abundance [FeH] (dex) middot middot middot 022 plusmn 008Microturbulent velocity vmic ( km sminus1) middot middot middot 09 plusmn 01Macroturbulent velocity vmac ( km sminus1) middot middot middot 25 plusmn 06Projected rotational velocity v sin i ( km sminus1) middot middot middot 28 plusmn 06Rotational period Prot (days) middot middot middot 1724 plusmn 012Activity index(b) logRprimeHK middot middot middot minus446 plusmn 006Gyrochronological age (Gyr) middot middot middot 18 plusmn 03Interstellar extinction AV (mag) middot middot middot 007 plusmn 005Star distance d (pc) middot middot middot 152 plusmn 10
Model parameters of K2-139 bOrbital period Porb (days) U[283773 283873] 2838236 plusmn 000026Transit epoch T0 (BJDTDBminus2 450 000) U[73258120 73258220] 732581714 plusmn 000033Scaled semi-major axis aR U[12 100] 448+47
minus67Planet-to-star radius ratio RpR U[0 02] 00961+00023
minus00015Impact parameter b U[0 12] 030+021
minus019radice sinω U[minus1 1](c) 010+029
minus030radice cosω U[minus1 1](c) 006+024
minus027Radial velocity semi-amplitude variation K ( m sminus1) U[0 200] 277+60
Note ndash The adopted Sun and Jupiter units follow the recommendations from the International Astronomical Union (Prša et al 2016) (a) U[a b] refersto uniform priors between a and b N[a b] means Gaussian priors with mean a and standard deviation b and F[a] to a fixed a value (b) Corrected forinterstellar reddening following Fossati et al (2017) The correction is +006 (c) The code always ensures that e lt 1 (d) Assuming albedo = 0
MNRAS 000 1ndash12 (2017)
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
MNRAS 000 1ndash12 (2017)
1 Introduction
2 K2 photometry
3 ALFOSC imaging
4 High-resolution spectroscopy
5 Stellar parameters
51 Spectral analysis
52 Interstellar extinction
53 Rotational period
54 Stellar mass radius and age
6 Joint RV-transit fit
7 Results and discussion
71 Stellar activity modeling
72 Additional companion
73 Spot-crossing events
74 Planets composition and formation scenario
8 Conclusions
Acknowledgements
The transiting warm Jupiter K2-139 b 5
Table 3 Spectroscopic parameters of K2-139 as derived using the three methods described in Sect 5
Method Teff (K) log g (cgs) [FeH] (dex) vmic ( km sminus1) vmac ( km sminus1) v sin i ( km sminus1)
the surface gravity log g We simultaneously fitted different spec-tral regions to measure the iron abundance [FeH] The projectedrotational velocity v sin i was determined by fitting the profile ofmany isolated and unblended metal lines
Second method It relies on the use of the spectral analysispackage Spectroscopy Made Easy (SME Valenti amp Piskunov 1996Valenti amp Fischer 2005) For a set of given stellar parameters SMEcalculates synthetic spectra and fits them to high-resolution ob-served spectra using a chi-squared minimization procedure Weused SME version 443 and ATLAS 12model spectra (Kurucz 2013)We adopted the same calibration equation as described in the firstmethod to determine vmic and vmac Effective temperature is de-rived from the Hα wings log g from the Ca i 6102 6122 6162and 6439 Aring lines [FeH] and v sin i from isolated iron lines
Third method It uses the classical equivalent width (EW)method adopting the following criteria i) Teff is obtained by re-moving trends between abundance of the chemical elements andthe respective excitation potentials ii) log g is optimised by as-suming the ionisation equilibrium condition ie by requiring thatfor a given species the same abundance (within the uncertainties)is obtained from lines of two ionisation states (typically neutral andsingly ionised lines) iii) vmic is set by minimising the slope of therelationship between abundance and the logarithm of the reducedEWs The equivalent widths of Fe i and Fe ii lines are measuredusing the code DOOp (Cantat-Gaudin et al 2014) a wrapper ofDAOSPEC (Stetson amp Pancino 2008) The stellar atmosphere param-eters are derived with the program FAMA (Magrini et al 2013) awrapper of MOOG (Sneden et al 2012) We used the public versionof the atomic data prepared for the Gaia-ESO Survey (Heiter et al2015) and based on the VALD3 data (Ryabchikova et al 2011) Weused sim200 Fe i lines and sim10 Fe ii lines for the determination of thestellar parameters
The three methods provide consistent results within the 1-σerror bars (Table 5) While we have no reason to prefer one tech-nique over the other we adopted the parameter estimates of thefirst method ie Teff = 5340 plusmn 110 K log g = 450 plusmn 009 (cgs)[FeH] = 022 plusmn 008 dex vmic = 09 plusmn 01 km sminus1 vmac = 25 plusmn06 km sminus1 and v sin i = 28 plusmn 06 km sminus1 As a sanity check wealso analyzed the HARPS and HARPS-N data and obtained con-sistent results but with larger error bars owing to the lower SNratio of the co-added HARPS and HARPS-N spectra compared tothat of the co-added FIES data Using the Boyajian et al (2013)rsquoscalibration (see their Table 6) the effective temperature of K2-139defines the spectral type of the host star as K0V
52 Interstellar extinction
We measured the visual reddening (AV) of K2-139 following thetechnique described in Gandolfi et al (2008) We fitted the spectral
energy distribution of the star to synthetic colors extracted from theBT-NEXTGEN model spectrum (Allard et al 2011) with the samephotospheric parameters as the star We adopted the extinction lawof Cardelli et al (1989) and assumed a normal value for the total-to-selective extinction ie RV = AVE(B minus V)= 31 We measureda visual extinction of AV = 007plusmn 005 mag This value is belowthe upper limit of AV 03 mag extracted from the Schlegel et al(1998)rsquos all-sky extinction map corroborating our result
53 Rotational period
The K2 light curve of K2-139 displays periodic and quasi-periodicvariations with a peak-to-peak photometric amplitude of sim2(Fig 1) The late-type spectral type of the star suggests that theobserved variability is due to Sun-like spots appearing and disap-pearing from the visible stellar disc as the star rotates around its axisThis is corroborated by the fact that K2-139 is a chromosphericallyactive star The HARPS and HARPS-N spectra show clear emis-sion components in the cores of the Ca ii HampK lines from whichwe measured an average activity index of log RprimeHK =minus446plusmn 0064
The out-of-transit photometric variability observed in the lightcurve of K2-139 is mainly due to two active regions located at op-posite stellar longitudes whose lifetime is longer than the durationof the K2 observations Using the spots as tracers of stellar rotationand following the auto correlation function (ACF) technique de-scribed in McQuillan et al (2014) we estimated that the rotationalperiod of the star is Prot = 1724 plusmn 012 days The Lomb-Scargleperiodogram of the light curve shows its strongest peak at the sameperiod confirming our results
It is worth noting that the rotation period (Prot = 1724 plusmn012 days) and radius (R= 0862 plusmn 0032 R see next section)of the host star translate into a maximum value for the projected ro-tational velocity of v sin imax = 253plusmn 010 km sminus1 which agreeswith the spectroscopically derived v sin i = 28 plusmn 06 km sminus1 sug-gesting that the star is seen nearly equator-on (i asymp 90) and thatthe system might be aligned along the line-of-sight
54 Stellar mass radius and age
We derived the stellar mass radius and age using the online inter-face for Bayesian estimation of stellar parameters available at thefollowing web page httpstevoapdinafitcgi-binparam Briefly the web tool interpolates onto PARSEC modelisochrones (Bressan et al 2012) the V-band apparent magnitude
4 This value is corrected for the interstellar medium absorption followingthe procedure described in Fossati et al (2017) and using themeasured stellarparameters and reddening The correction is +006 The star is thereforeslightly more active than what measured from the spectra
MNRAS 000 1ndash12 (2017)
6 O Barragaacuten et al
effective temperature metal content and parallax We used the V-band magnitude reported in Table 1 ndash after correcting for interstel-lar reddening (Sect 52) ndash along with the effective temperature andmetal content we derived in Sect 5 The parallax was retrieved fromthe Gaiarsquos first data release (px = 656 plusmn 043mas d = 152plusmn 10 pcFabricius et al 2016) We adopted the log-normal initial mass func-tion from Chabrier (2001)
K2-139 has a mass of M= 0919 plusmn 0033 M and radiusof R= 0862 plusmn 0032 R corresponding to a surface gravity oflog g = 4503plusmn0035 (cgs) in excellent agreement with the spec-troscopically derived value of log g = 450plusmn009 (cgs see Sect 5)The derivedmean density ρ = 202plusmn024 g cmminus3 ofK2-139 is alsoconsistent within 1-σ with the density estimated by the modelingof the transit light curve (ρ = 211+074
minus081 g cmminus3 see Sect 6)The isochrones provide an age of 36plusmn34 Gyr for K2-139 Us-
ing the equations given in Barnes amp Kim (2010) and Barnes (2010)the rotation period of 173 days (Sect 53) implies a gyrochronolog-ical age of 18plusmn 03 Gyr
6 JOINT RV-TRANSIT FIT
We performed the joint fit to the photometric and RV data using thecode pyaneti5 (Barragaacuten et al 2017) a PythonFortran softwaresuite based on Markov Chain Monte Carlo (MCMC) methods
The photometric data included in the joint analysis are subsetsof the whole EVEREST K2 light curve We used the EVERESTlight curve because it provides a slightly better rms over the Vander-burg amp Johnson (2014)rsquos dataWe selectedsim10 hours of data-pointsaround each of the 3 transits which have a duration of sim5 hours Wede-trended each individual transits with the code exotrending6
(Barragaacuten amp Gandolfi 2017) using a second-order polynomial fit-ted to the out-of-transit points The fitted data include 12 pointsimmediately before and after each transit with the exception of thelast transit for which only 9 data points are available We removedthe data points that are affected by stellar spot crossing events (seeSect 71 for more details)
We fitted the RV data using a Keplerian model for the planetalong with two sine-like curves to account for the activity-inducedRV (see next section for details) We adopted the limb-darkenedquadratic law of Mandel amp Agol (2002) for the transit model Weadopted the Gaussian likelihood
L =[
nprodi=1
2π
(σ2i + σ
2j
)minus12]
exp
minus
nsumi=1
12(Di minus Mi)2
σ2i+ σ2
j
(1)
where n is the number of data points σi is the error associatedto each data point Di Mi is the model associated to a given Di andσj is an extra noise term sometime referred as jitter
The sampling method and fitted parameters are the same as inBarragaacuten et al (2016) Details on the adopted priors are given in Ta-ble 5 Following Kipping (2010) we super-sampled the light curvemodel using 10 subsamples perK2 exposure to account for the long-cadence acquisition The parameter space was explored with 500independent chains created randomly inside the prior ranges Thechain convergence was analyzed using the Gelman-Rubin statisticsThe burn-in phase uses 25 000 more iterations with a thin factor
5 Available at httpsgithubcomoscaribvpyaneti6 Available at httpsgithubcomoscaribvexotrending
of 50 The posterior distribution of each parameter has 250 000independent data points
7 RESULTS AND DISCUSSION
71 Stellar activity modeling
A simple Keplerian model provides a poor fit to the RV measure-ments with χ2dof = 61 (Table 4) suggesting that additionalsignals might be present in our Doppler data Activity-induced RVvariation is expected given the 2 peak-to-peak photometric vari-ability observed in the K2 light curve of K2-139 (Fig 1) and theCa ii HampK activity index of log RprimeHK =minus446 plusmn 006 (Sect 53)The K2 photometric variation corresponds to a spot filling factorof approximately 2 if this variation is due to cool starspots Wecan use the empirical relationship relating spot coverage to RV am-plitude from Saar amp Donahue (1997) or Hatzes (2002) to estimatethe RV amplitude expected from spots Using the projected rota-tional velocity of 28 km sminus1 results in an RV semi-amplitude ofasymp20ndash30 m sminus1 The code SOAP2 designed to estimate the effectof active regions on photometric and spectroscopic measurements(Dumusque et al 2014) provides consistent results
In order to look for additional signals in our Doppler data weperformed a frequency analysis of theRVmeasurements and activityindicators On one occasion7 K2-139 was observed with FIES andHARPS-S nearly simultaneously (within less than 25 minutes) Weused the two sets of measurements to estimate the RV FWHMand BIS offsets between the two instruments We assumed no offsetbetween HARPS-N and HARPS While we acknowledge that thisassumption is arbitrary we note that the modeling of the RV datagives an offset of ∆RV(HNminusH) = 0002plusmn 00158 km sminus1 (Table 5)which is consistent with zero
Figure 3 displays the generalized Lomb-Scargle periodograms(Zechmeister amp Kuumlrster 2009) of the combined datasets From topto bottom the RVdata (first panel) the RV residuals after subtractingthe transiting planet signal (second panel) and the BIS (third panel)and FWHM (fourth panel) of the cross correlation function Theperiodogram of the window function is shown in the lower panelThe dotted vertical blue lines mark the frequency at the orbitalperiod of the planet (0035 cd) as well as the frequencies at therotation period of the star (0058 cd) and its first two harmonics(0116 and 0174 cd)
The periodogram of the RV data (upper panel) shows a peak atthe orbital frequency of the planet alongwith two additional peaks at0095 and 0130 cd Since the periodogram of the window functionshows two peaks atsim0060 andsim0095 cd (lower panel red arrows)we interpreted the 0095 and 0130 cd peaks as the aliases of theorbital frequency8 We note also that periodogram of the BIS ofthe CCF displays peaks whose frequencies are close to the stellarrotation frequency and its first two harmonics However none of thepeaks visible in the GLS periodograms of Fig 3 has a false alarmprobability (FAP)9 lower than 5 Although our spectroscopic datashow neither additional signals we note that the semi-amplitudevariation of the BIS and FWHM is expected to be 10-15 m sminus1
(Dumusque et al 2014) which is comparable with the uncertaintiesof most of ourmeasurements (Table 2) The lack of significant peaks
7 Epoch BJD=24575898 0095 = 0035 + 0060 cd and 0130 = 0035 + 0095 cd9 We determined the FAP following the Monte Carlo bootstrap methoddescribed in Kuerster et al (1997)
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 7
Figure 3 Generalized Lomb-Scargle periodogram of the combined FIESHARPS and HARPS-N Doppler datasets From top to bottom the RV datathe RV residuals after subtracting the transiting planet signal the BIS andFWHM of the CCF and the window function The dotted vertical blue linesmark the frequencies at the orbital period as well as at the stellar rotationperiod and its first two harmonics The dashed vertical red lines mark the5 false alarm probabilities as derived using the bootstrap method The redarrows in the lower panel mark the two peaks presented in the main text
in the periodogram of the RV data and RV residuals as well as inthe periodogram of the activity indicators could be explained by thelimited number of available measurements and their uncertaintiesWe conclude that we cannot exclude the existence of spot-inducedsignals in our RV measurements
Photometric and radial velocity variations due to rotationalmodulation can be complex with not only the rotational period Protpresent but also its harmonics eg Prot2 Prot3 Assuming thatthe surface structures responsible for this modulation (eg coolspots) are not evolving rapidly then the simplest representation ofthe rotationalmodulation is through the Fourier components definedby the rotation period and its harmonics Figure 1 shows that theevolution time-scale of the active regions in the stellar surface islonger than the 80-day duration of the K2 campaign Since our RVfollow-up spans 55 days we can assume that any activity-inducedRV signal is coherent within our observing window This approachhas been used previously for other planetary systems orbiting activestars (eg Pepe et al 2013)
The Fourier analysis of the K2 light curve is the best way tomeasure the contribution of the rotation period and its harmonics tothe quasi-periodic photometric variability of the star We thereforeanalyzed the K2 light curve using a pre-whitening procedure Thatis the dominant period was found a sine-fit made to the data andsubtracted and additional periods searched in the residual data
We used the program Period04 (Lenz amp Breger 2005) for thisprocedure
The dominant periods are sim172 days ie the rotation periodof the star (Sect 53) and roughly the first four harmonics (ie 8657 43 and 34 days) The 172- and 86-day periods have about thesame amplitude while the 57-day period (Prot3) has 10 of themain amplitude The Prot4 signal has only about 4 of the mainamplitude The light curve analysis indicates that the signal due torotational modulation can largely be represented by the rotationalperiod (Prot) and its first harmonic (Prot2)
In order to test if the addition of RV sinusoidal signals atthe stellar rotation period and its harmonics can account for theadditional variation seen in our RV measurements we compareddifferent models by adding signals one by one The first model (P0)includes only the planet signal ie a Keplerian model fitted to theRV data using the same priors given in Table 5 but fixing epoch andperiod to the values derived by the transit modeling The next model(P1) is obtained from P0 by adding a sinusoidal signal at the rotationperiod of the star (Prot) Models P2 includes the first harmonic ofthe rotation period (Prot2) whereas model P3 account for the first(Prot2) and second (Prot3) harmonics While adding sinusoidalsignals we fitted for their amplitudes phases and periods We usedflat priors for the phases and amplitudes (details in Table 5) Weused a Gaussian prior for Prot using the value and its uncertaintyderived in Sect 53 The periods of the harmonic signals were leftfree to vary depending on the value assumed by Prot at each stepof the MCMC chains In order to check if the RV variation inducedby the planet is significant in our data set we also performed the fitusing models where the planetary signal was not included (modelsNP1 and NP2 see Table 4)
Table 4 shows the goodness of the fit for each model Thepreferred model is P2 (planet plus 2 sinusoidal signals at Prot andProt2) with the lowest Akaike Information Criteria (AIC) and max-imum likelihood This result is consistent with the Fourier analysisof the K2 light curve which suggests that the major contributionto the photometric variations arises from the stellar rotation periodand its first harmonic Our analysis provides also additional evi-dence that the Doppler motion induced by the planet is present inour RV data set First the planet signal does not significantly varyfor the P0 P1 P2 and P3 models (Table 4) Second the modelswith no planetary signal (NP1 and NP2) provide a poor fit to theRV measurements (Table 4)
To account for additional instrumental noise not included inthe nominal RV error bars andor imperfect treatment of the varioussources of RV variations we fitted for a jitter term for each instru-ment The final parameter estimates and their error bars are listedin Table 5 They are defined as the median and the 68 credibleinterval of the final posterior distributions The best fitting transitand RVmodels are displayed in Figure 4 along with the photometricand RV data points
72 Additional companion
Huang et al (2016) found that warm Jupiters with low eccentricities(e 04) have inner low-mass companions They used this evidenceas an argument in favour of the in situ formation since the planetmigration would have cleaned the warm Jupiter neighborhood Wesearched the light curve for additional transit signals but foundno evidence for an additional transiting planet in the system Asdescribed in the previous paragraph the periodogram of the RVresiduals showno significant peakwith false alarmprobability lowerthan 5
Figure 4 Top FIES (blue circles) HARPS-N (green diamonds) and HARPS (red squares) RV measurements versus time following the subtraction of thesystemic velocities for each instrument The 1σ uncertainties are marked using the same color used for each data-set The vertical gray lines mark the errorbars including jitter The solid line represents the best fitting RV model which includes the planet signal and the activity signal at the stellar rotation periodand its first harmonic The dashed dash-dotted and dotted lines show the RV contribution of K2-139 b stellar rotation and first harmonic respectively Lowerleft panel Transit light curve folded to the orbital period of K2-139 b and residuals The red points mark the K2 data and their error bars The solid line markthe re-binned best-fitting transit model Lower right panel Phase-folded RV curve of K2-139 b and best fitting Keplerian solution (solid line) following thesubtraction of the two additional sinusoidal signals used to account for the stellar activity The FIES HARPS and HARPS-N are corrected for the instrumentoffsets as derived from the global analysis
Table 4 Model comparison
Model Comment Npars Kb (m sminus1) χ2dof(a) ln L AIC(b)
P0 Planet signal 6 291 plusmn 20 61 356 -60P1 Planet signal + 1 sine-curve at Prot 9 294 plusmn 24 34 581 -98
P2 Planet signal + 2 sine-curves at Prot and Prot2 11 273+26minus25 38 601 -98
P3 Planet signal + 3 sine-curves at Prot Prot2 and Prot3 13 278+27minus26 53 593 -93
NP1 1 sine-curve at Prot (No planet signal) 6 0 185 -448 101NP2 2 sine-curves at Prot and Prot2 (No planet signal) 8 0 159 -120 40
Note ndash (a) χ2 value assuming no jitter (b)We used the Akaike Information Criteria (AIC = 2Nparsminus ln 2L) instead of the widely used Bayesian informationcriteria (BIC) because our RV data sample is small (19 data points) and BIC performs better for large samples (Burnham amp Anderson 2002)
73 Spot-crossing events
The passage of a planet in front of a spot can be detected as abump in the transit light curve (see eg Sanchis-Ojeda amp Winn2011) Spot-crossings events are clearly visible in the EVERESTtransit light curves (Fig 4) The same features appear at the sametimes and with consistent amplitudes in the Vanderburg amp Johnson(2014) data confirming that the bumps are real and not due to
systematics To assess whether the bumps significantly affect theparameter estimates we performed the joint analysis as describedin Sect 6 including all the transit data points We found that the finalparameters are consistent within 1-σ with those reported in Table 5
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 9
01 05 1 2 3 4 5 10Mass (MJup)
06
08
10
12
14
Rad
ius
(RJu
p)
H 0 M core 10 M core 25 M core 50 M core100 M core
Figure 5 Warm Jupiters (black squares Mp gt 03 MJup and10 6 Porb 6 100 days) whose mass and radius have been estimated with aprecision of at least 25 (as of January 2017 exoplaneteu) K2-139 b isshown with a filled red circle The solid line corresponds to a planet with apure hydrogen composition (Seager et al 2007) The dashed lines representthe Fortney et al (2007) models for planet core masses of 0 10 25 50and 100 Moplus The vertical dotted line marks the giant planet lower limit asdefined by Hatzes amp Rauer (2015)
74 Planetrsquos composition and formation scenario
With amass of Mp = 0387+0083minus0075 MJ and radius of Rp = 0808+0034
minus0033RJ (resulting in a mean density of ρp=091+024
minus020 g cmminus3) K2-139 bjoins the small group of well characterized warm Jupiters Fig 5shows the position of K2-139 b in themass-radius diagram for warmJupiters (Mp gt 03 MJup 10 6 Porb 6 100 days) whose massand radius have been determined with a precision better than 25(14 objects) Notably K2-139 b is the transiting warm Jupiter withthe lowest mass known to date if the definition of giant planetsgiven by Hatzes amp Rauer (2015) is adopted Fig 5 displays also theplanetary models of Fortney et al (2007) for different core massesand age between 10 and 45Gyrs The planet radius of K2-139 bcan be explained if the planet has a core10 of 49+19
minus17 Moplus containingsim40 of the total planetary mass We expect that K2-139 b has asolid core surrounded by a gaseous envelope
Rafikov (2006) found that a core of mass 5 ndash 20 Moplus at a semi-major axis between 01 and 10AU would be able to start the run-away accretion phase to form a gas giant planet in situ Howeveraccording to his models these kind of cores are unlikely to formowing to the high irradiation coming from the star Boley et al(2016) suggested instead that more massive cores (Mcore amp 20Moplus)can be built up from the merging of tightly packed inner planetsformed at the early stages of the circumstellar disc Batygin et al(2016) found a similar result and argued that the massive core ofHD149026b (Mcore asymp 100Moplus) could be explained by one or moresuper-Earths which merged and accreted the surrounding gas toform a gas-giant planet Huang et al (2016) suggested that thesecores can initiate runaway accretion if they are formed in a re-gion with enough gas around them while those without enoughvolatiles remain super-Earths and represent the population of mas-sive rocky planets unveiled by Kepler around solar-like stars (egDemory 2014) Based on these studies and given the semi-majoraxis of 0179+0021
minus0027 AU the 48 plusmn 14 Moplus core of K2-139 b could
10 Calculated by interpolating Fortney et al (2007)rsquos models
have formed the planet in situ We note that the metallicity of K2-139 is relatively high ([FeH] = 021 plusmn 005) suggesting that theprimordial circumstellar disc had a relatively high content of dustwhich would have enhanced the formation of the core of K2-139 b(see eg Johnson amp Li 2012) Alternatively the planet might haveformed beyond the snow line and migrated inwards via planet-discinteraction (see eg Baruteau et al 2014)
8 CONCLUSIONS
We confirmed the planetary nature and derived the orbital and mainphysical parameters of K2-139 b a warm Jupiter (Teq = 565+48
minus32 K)transiting an active (log RprimeHK = minus446 plusmn 006) K0V star every 29days We measured a planetary mass of Mp = 0387+0083
minus0075 MJ andradius of Rp = 0808+0034
minus0033 RJ At a separation of ap = 0179+0021minus0027
AU the mean density of ρp = 091+024minus020 g cmminus3 implies that the
planet has a core of 49+19minus17 Moplus according to the evolutionary mod-
els of Fortney et al (2007) K2-139 b joins the small group ofwell-characterized warm Jupiters whose mass and radius have beendetermined with a precision better than 25
The spin-orbit angle ie the angle between the spin axis ofthe star and the angular momentum vector of the orbit can pro-vide us with valuable information on the migration mechanismsof exoplanets (see eg Winn 2010 Morton amp Johnson 2011 Al-brecht et al 2012 Gandolfi et al 2012) Currently there are only4 warm Jupiters (Mp gt 03 MJup and 10 6 Porb 6 100 days)with measured obliquity11 From this perspective K2-139 is anideal target to measure the sky-project spin-orbit angle via obser-vations of the Rossiter-McLaughlin (RM) effect Assuming spin-orbit alignement the expected amplitude of the RM anomaly is∆RV asymp
radic1 minus b2 (RpR)2 v sin iasymp 25 m sminus1 (Winn 2010) Given
the brightness of the host star (V = 11653 mag) this amplitudecan easily be measured using state-of-the-art spectrographs suchas HARPSESO-36m Moreover the transit duration (sim5 hours)is shorter than the visibility of K2-139 which is sim9 hours fromLa Silla observatory (altitude higher than 30 above the horizon)
Alternatively the spin-orbit angle could be measured fromthe analysis of the spot-crossing events as described in Sanchis-Ojeda et al (2011) and Sanchis-Ojeda et al (2012) Anomaliesascribable to the passage of K2-139b in front of stellar spots arevisible in the 3 transit light curves observed by K2 Unfortunatelythe limited number of transits and the K2 long cadence data donot allow us to perform a meaningful quantitative analysis of thespot-crossing events Given the amplitude of the detected anoma-lies (sim01) space-based high-precision photometry is needed todetect the spot-crossing events Observations performed with theupcoming CHaracterising ExOPlanets Satellite (CHEOPS Broeget al 2013) would allow us to photometrically determine the spin-orbit angle of this system
ACKNOWLEDGEMENTS
We warmly thank the NOT ESO TNG staff members for theirunique support during the observations We are very thankful toXavier Bonfils Franccedilois Bouchy Martin Kuumlrster Tsevi MazehJorge Melendez and Nuno Santos who kindly agreed to exchange
11 Source httpwww2mpsmpgdehomeshellercontentmain_HRMhtml as of January 2017
MNRAS 000 1ndash12 (2017)
10 O Barragaacuten et al
HARPSandFIES timewith us Special thanks go toAntoninoLanzafor assisting us with the calculation of the gyro-age of the star Wealso greatly thank the anonymous referee for herhis careful re-view and suggestions which helped us to improve the manuscriptD Gandolfi gratefully acknowledges the financial support of theProgramma Giovani Ricercatori ndash Rita Levi Montalcini ndash Rien-tro dei Cervelli (2012) awarded by the Italian Ministry of Edu-cation Universities and Research (MIUR) Sz Csizmadia thanksthe Hungarian OTKA Grant K113117 H J Deeg and D Nespralacknowledge support by grant ESP2015-65712-C5-4-R of the Span-ish Secretary of State for RampDampi (MINECO) D Lorenzo-Oliveiraacknowledges the support from FAPESP (201620667-8) This re-search was supported by the Ministerio de Economia y Competi-tividad under project FIS2012-31079 The research leading to theseresults has received funding from the European Union SeventhFramework Programme (FP72013-2016) under grant agreementNo 312430 (OPTICON) Based on observations obtained a) withthe Nordic Optical Telescope (NOT) operated on the island of LaPalma jointly by Denmark Finland Iceland Norway and Swedenin the Spanish Observatorio del Roque de los Muchachos (ORM)of the Instituto de Astrofiacutesica de Canarias (IAC) b) with the Ital-ian Telescopio Nazionale Galileo (TNG) also operated at the ORM(IAC) on the island of La Palma by the INAF - Fundacioacuten GalileoGalilei c) the 36m ESO telescope at La Silla Observatory underprogramme ID 097C-0948 The data presented here were obtainedin part with ALFOSC which is provided by the Instituto de As-trofisica de Andalucia (IAA) under a joint agreement with the Uni-versity of Copenhagen and NOTSA This paper includes data col-lected by the Kepler mission Funding for the Kepler mission is pro-vided by the NASA Science Mission directorate Some of the datapresented in this paper were obtained from theMikulski Archive forSpace Telescopes (MAST) STScI is operated by the Association ofUniversities for Research in Astronomy Inc under NASA contractNAS5-26555 Support for MAST for non-HST data is provided bythe NASA Office of Space Science via grant NNX09AF08G and byother grants and contracts MF and CMP acknowledge generoussupport from the Swedish National Space Board C Eiroa and IRebollido are supported by Spanish grant AYA2014-55840-P PDacknowledge the support from INAF and Ministero dellrsquoIstruzionedellrsquoUniversitagrave e della Ricerca (MIUR) in the form of the grantldquoPremiale VLT 2012rdquo and ldquoThe Chemical and Dynamical Evolu-tion of the Milky Way and Local Group Galaxiesrdquo This work hasmade use of data from the European Space Agency (ESA) missionGaia (httpwwwcosmosesaintgaia) processed by theGaia Data Processing and Analysis Consortium (DPAC httpwwwcosmosesaintwebgaiadpacconsortium) Fundingfor the DPAC has been provided by national institutions in particu-lar the institutions participating in theGaiaMultilateral Agreement
REFERENCES
Albrecht S et al 2012 ApJ 757 18Allard F Homeier D Freytag B 2011 in Johns-Krull C BrowningM K
West A A eds Astronomical Society of the Pacific Conference SeriesVol 448 16th Cambridge Workshop on Cool Stars Stellar Systemsand the Sun p 91 (arXiv10115405)
Antonini F Hamers A S Lithwick Y 2016 preprint(arXiv160401781)
Barnes S A 2010 ApJ 722 222Barnes S A Kim Y-C 2010 ApJ 721 675Barragaacuten O Gandolfi D 2017 Exotrending Astrophysics Source Code
Library (ascl1706001)
Barragaacuten O et al 2016 AJ 152 193Barragaacuten O Gandolfi D Antoniciello G 2017 pyaneti Astrophysics
Source Code Library (ascl1707003)Baruteau C et al 2014 Protostars and Planets VI pp 667ndash689Batygin K Bodenheimer P H Laughlin G P 2016 ApJ 829 114Boley A C Granados Contreras A P Gladman B 2016 ApJ 817 L17Boyajian T S et al 2013 ApJ 771 40Brahm R et al 2016 AJ 151 89Bressan A Marigo P Girardi L Salasnich B Dal Cero C Rubele S
Nanni A 2012 MNRAS 427 127Broeg C et al 2013 in European Physical Journal Web of Conferences p
03005 (arXiv13052270) doi101051epjconf20134703005Bruntt H et al 2010 MNRAS 405 1907Buchhave L A et al 2010 ApJ 720 1118Burnham K Anderson D 2002 Model Selection and Multimodel Infer-
ence A Practical Information-Theoretic Approach NewYork Springer-Verlag
Cabrera J et al 2009 AampA 506 501Cabrera J Csizmadia S Erikson A Rauer H Kirste S 2012 AampA 548
A44Cabrera J et al 2014 ApJ 781 18Cantat-Gaudin T et al 2014 AampA 562 A10Cardelli J A Clayton G C Mathis J S 1989 ApJ 345 245Carone L et al 2012 AampA 538 A112Carpano S et al 2009 AampA 506 491Castelli F Kurucz R L 2004 preprintCavarroc C et al 2012 ApampSS 337 511Chabrier G 2001 ApJ 554 1274Cosentino R et al 2012 in Ground-based and Airborne Instrumentation
for Astronomy IV p 84461V doi10111712925738Cutri R M et al 2003 2MASS All Sky Catalog of point sourcesCutri R M et al 2012 Technical report Explanatory Supplement to the
WISE All-Sky Data Release ProductsDawson R I Johnson J A Morton T D Crepp J R Fabrycky D C
Murray-Clay R A Howard A W 2012 ApJ 761 163Deeg H J et al 2010 Nature 464 384Demory B-O 2014 ApJ 789 L20Dong S Katz B Socrates A 2014 ApJ 781 L5Doyle A P Davies G R Smalley B Chaplin W J Elsworth Y 2014
MNRAS 444 3592Dumusque X Boisse I Santos N C 2014 ApJ 796 132Endl M Cochran W D 2016 PASP 128 094502Erikson A et al 2012 AampA 539 A14Fabricius C et al 2016 AampA 595 A3Fortney J J Marley M S Barnes J W 2007 ApJ 659 1661Fossati et al 2017 AampA submittedFrandsenS LindbergB 1999 inKarttunenH PiirolaV edsAstrophysics
with the NOT p 71Frewen S F N Hansen B M S 2016 MNRAS 455 1538Gandolfi D et al 2008 ApJ 687 1303Gandolfi D et al 2012 AampA 543 L5Gandolfi D et al 2015 AampA 576 A11Gray R O 1999 SPECTRUM A stellar spectral synthesis program As-
trophysics Source Code Library (ascl9910002)Grziwa S Paumltzold M 2016 preprint (arXiv160708417)Grziwa S Paumltzold M Carone L 2012 MNRAS 420 1045Hamers A S Antonini F Lithwick Y Perets H B Portegies Zwart S F
2016 preprint (arXiv160607438)Hatzes A P 2002 Astronomische Nachrichten 323 392Hatzes A P Rauer H 2015 ApJ 810 L25Heiter U et al 2015 Phys Scr 90 054010Huang C Wu Y Triaud A H M J 2016 ApJ 825 98Jenkins J S et al 2017 MNRAS 466 443Johnson J L Li H 2012 ApJ 751 81Kipping D M 2010 MNRAS 408 1758Kley W Nelson R P 2012 ARAampA 50 211Kovaacutecs G Zucker S Mazeh T 2002 AampA 391 369
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 11
Table 5 K2-139 system parameters
Parameter Prior(a) Final value
Stellar parametersStar mass M (M) middot middot middot 0919 plusmn 0033Star radius R (R) middot middot middot 0862 plusmn 0032Star density ρ (from spectroscopy g cmminus3) middot middot middot 202+025
minus022
Star density ρ (from light curve g cmminus3) middot middot middot 211+074minus081
Effective Temperature Teff (K) middot middot middot 5340 plusmn 110Surface gravity log g (cgs) middot middot middot 450 plusmn 009Iron abundance [FeH] (dex) middot middot middot 022 plusmn 008Microturbulent velocity vmic ( km sminus1) middot middot middot 09 plusmn 01Macroturbulent velocity vmac ( km sminus1) middot middot middot 25 plusmn 06Projected rotational velocity v sin i ( km sminus1) middot middot middot 28 plusmn 06Rotational period Prot (days) middot middot middot 1724 plusmn 012Activity index(b) logRprimeHK middot middot middot minus446 plusmn 006Gyrochronological age (Gyr) middot middot middot 18 plusmn 03Interstellar extinction AV (mag) middot middot middot 007 plusmn 005Star distance d (pc) middot middot middot 152 plusmn 10
Model parameters of K2-139 bOrbital period Porb (days) U[283773 283873] 2838236 plusmn 000026Transit epoch T0 (BJDTDBminus2 450 000) U[73258120 73258220] 732581714 plusmn 000033Scaled semi-major axis aR U[12 100] 448+47
minus67Planet-to-star radius ratio RpR U[0 02] 00961+00023
minus00015Impact parameter b U[0 12] 030+021
minus019radice sinω U[minus1 1](c) 010+029
minus030radice cosω U[minus1 1](c) 006+024
minus027Radial velocity semi-amplitude variation K ( m sminus1) U[0 200] 277+60
Note ndash The adopted Sun and Jupiter units follow the recommendations from the International Astronomical Union (Prša et al 2016) (a) U[a b] refersto uniform priors between a and b N[a b] means Gaussian priors with mean a and standard deviation b and F[a] to a fixed a value (b) Corrected forinterstellar reddening following Fossati et al (2017) The correction is +006 (c) The code always ensures that e lt 1 (d) Assuming albedo = 0
MNRAS 000 1ndash12 (2017)
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
MNRAS 000 1ndash12 (2017)
1 Introduction
2 K2 photometry
3 ALFOSC imaging
4 High-resolution spectroscopy
5 Stellar parameters
51 Spectral analysis
52 Interstellar extinction
53 Rotational period
54 Stellar mass radius and age
6 Joint RV-transit fit
7 Results and discussion
71 Stellar activity modeling
72 Additional companion
73 Spot-crossing events
74 Planets composition and formation scenario
8 Conclusions
Acknowledgements
6 O Barragaacuten et al
effective temperature metal content and parallax We used the V-band magnitude reported in Table 1 ndash after correcting for interstel-lar reddening (Sect 52) ndash along with the effective temperature andmetal content we derived in Sect 5 The parallax was retrieved fromthe Gaiarsquos first data release (px = 656 plusmn 043mas d = 152plusmn 10 pcFabricius et al 2016) We adopted the log-normal initial mass func-tion from Chabrier (2001)
K2-139 has a mass of M= 0919 plusmn 0033 M and radiusof R= 0862 plusmn 0032 R corresponding to a surface gravity oflog g = 4503plusmn0035 (cgs) in excellent agreement with the spec-troscopically derived value of log g = 450plusmn009 (cgs see Sect 5)The derivedmean density ρ = 202plusmn024 g cmminus3 ofK2-139 is alsoconsistent within 1-σ with the density estimated by the modelingof the transit light curve (ρ = 211+074
minus081 g cmminus3 see Sect 6)The isochrones provide an age of 36plusmn34 Gyr for K2-139 Us-
ing the equations given in Barnes amp Kim (2010) and Barnes (2010)the rotation period of 173 days (Sect 53) implies a gyrochronolog-ical age of 18plusmn 03 Gyr
6 JOINT RV-TRANSIT FIT
We performed the joint fit to the photometric and RV data using thecode pyaneti5 (Barragaacuten et al 2017) a PythonFortran softwaresuite based on Markov Chain Monte Carlo (MCMC) methods
The photometric data included in the joint analysis are subsetsof the whole EVEREST K2 light curve We used the EVERESTlight curve because it provides a slightly better rms over the Vander-burg amp Johnson (2014)rsquos dataWe selectedsim10 hours of data-pointsaround each of the 3 transits which have a duration of sim5 hours Wede-trended each individual transits with the code exotrending6
(Barragaacuten amp Gandolfi 2017) using a second-order polynomial fit-ted to the out-of-transit points The fitted data include 12 pointsimmediately before and after each transit with the exception of thelast transit for which only 9 data points are available We removedthe data points that are affected by stellar spot crossing events (seeSect 71 for more details)
We fitted the RV data using a Keplerian model for the planetalong with two sine-like curves to account for the activity-inducedRV (see next section for details) We adopted the limb-darkenedquadratic law of Mandel amp Agol (2002) for the transit model Weadopted the Gaussian likelihood
L =[
nprodi=1
2π
(σ2i + σ
2j
)minus12]
exp
minus
nsumi=1
12(Di minus Mi)2
σ2i+ σ2
j
(1)
where n is the number of data points σi is the error associatedto each data point Di Mi is the model associated to a given Di andσj is an extra noise term sometime referred as jitter
The sampling method and fitted parameters are the same as inBarragaacuten et al (2016) Details on the adopted priors are given in Ta-ble 5 Following Kipping (2010) we super-sampled the light curvemodel using 10 subsamples perK2 exposure to account for the long-cadence acquisition The parameter space was explored with 500independent chains created randomly inside the prior ranges Thechain convergence was analyzed using the Gelman-Rubin statisticsThe burn-in phase uses 25 000 more iterations with a thin factor
5 Available at httpsgithubcomoscaribvpyaneti6 Available at httpsgithubcomoscaribvexotrending
of 50 The posterior distribution of each parameter has 250 000independent data points
7 RESULTS AND DISCUSSION
71 Stellar activity modeling
A simple Keplerian model provides a poor fit to the RV measure-ments with χ2dof = 61 (Table 4) suggesting that additionalsignals might be present in our Doppler data Activity-induced RVvariation is expected given the 2 peak-to-peak photometric vari-ability observed in the K2 light curve of K2-139 (Fig 1) and theCa ii HampK activity index of log RprimeHK =minus446 plusmn 006 (Sect 53)The K2 photometric variation corresponds to a spot filling factorof approximately 2 if this variation is due to cool starspots Wecan use the empirical relationship relating spot coverage to RV am-plitude from Saar amp Donahue (1997) or Hatzes (2002) to estimatethe RV amplitude expected from spots Using the projected rota-tional velocity of 28 km sminus1 results in an RV semi-amplitude ofasymp20ndash30 m sminus1 The code SOAP2 designed to estimate the effectof active regions on photometric and spectroscopic measurements(Dumusque et al 2014) provides consistent results
In order to look for additional signals in our Doppler data weperformed a frequency analysis of theRVmeasurements and activityindicators On one occasion7 K2-139 was observed with FIES andHARPS-S nearly simultaneously (within less than 25 minutes) Weused the two sets of measurements to estimate the RV FWHMand BIS offsets between the two instruments We assumed no offsetbetween HARPS-N and HARPS While we acknowledge that thisassumption is arbitrary we note that the modeling of the RV datagives an offset of ∆RV(HNminusH) = 0002plusmn 00158 km sminus1 (Table 5)which is consistent with zero
Figure 3 displays the generalized Lomb-Scargle periodograms(Zechmeister amp Kuumlrster 2009) of the combined datasets From topto bottom the RVdata (first panel) the RV residuals after subtractingthe transiting planet signal (second panel) and the BIS (third panel)and FWHM (fourth panel) of the cross correlation function Theperiodogram of the window function is shown in the lower panelThe dotted vertical blue lines mark the frequency at the orbitalperiod of the planet (0035 cd) as well as the frequencies at therotation period of the star (0058 cd) and its first two harmonics(0116 and 0174 cd)
The periodogram of the RV data (upper panel) shows a peak atthe orbital frequency of the planet alongwith two additional peaks at0095 and 0130 cd Since the periodogram of the window functionshows two peaks atsim0060 andsim0095 cd (lower panel red arrows)we interpreted the 0095 and 0130 cd peaks as the aliases of theorbital frequency8 We note also that periodogram of the BIS ofthe CCF displays peaks whose frequencies are close to the stellarrotation frequency and its first two harmonics However none of thepeaks visible in the GLS periodograms of Fig 3 has a false alarmprobability (FAP)9 lower than 5 Although our spectroscopic datashow neither additional signals we note that the semi-amplitudevariation of the BIS and FWHM is expected to be 10-15 m sminus1
(Dumusque et al 2014) which is comparable with the uncertaintiesof most of ourmeasurements (Table 2) The lack of significant peaks
7 Epoch BJD=24575898 0095 = 0035 + 0060 cd and 0130 = 0035 + 0095 cd9 We determined the FAP following the Monte Carlo bootstrap methoddescribed in Kuerster et al (1997)
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 7
Figure 3 Generalized Lomb-Scargle periodogram of the combined FIESHARPS and HARPS-N Doppler datasets From top to bottom the RV datathe RV residuals after subtracting the transiting planet signal the BIS andFWHM of the CCF and the window function The dotted vertical blue linesmark the frequencies at the orbital period as well as at the stellar rotationperiod and its first two harmonics The dashed vertical red lines mark the5 false alarm probabilities as derived using the bootstrap method The redarrows in the lower panel mark the two peaks presented in the main text
in the periodogram of the RV data and RV residuals as well as inthe periodogram of the activity indicators could be explained by thelimited number of available measurements and their uncertaintiesWe conclude that we cannot exclude the existence of spot-inducedsignals in our RV measurements
Photometric and radial velocity variations due to rotationalmodulation can be complex with not only the rotational period Protpresent but also its harmonics eg Prot2 Prot3 Assuming thatthe surface structures responsible for this modulation (eg coolspots) are not evolving rapidly then the simplest representation ofthe rotationalmodulation is through the Fourier components definedby the rotation period and its harmonics Figure 1 shows that theevolution time-scale of the active regions in the stellar surface islonger than the 80-day duration of the K2 campaign Since our RVfollow-up spans 55 days we can assume that any activity-inducedRV signal is coherent within our observing window This approachhas been used previously for other planetary systems orbiting activestars (eg Pepe et al 2013)
The Fourier analysis of the K2 light curve is the best way tomeasure the contribution of the rotation period and its harmonics tothe quasi-periodic photometric variability of the star We thereforeanalyzed the K2 light curve using a pre-whitening procedure Thatis the dominant period was found a sine-fit made to the data andsubtracted and additional periods searched in the residual data
We used the program Period04 (Lenz amp Breger 2005) for thisprocedure
The dominant periods are sim172 days ie the rotation periodof the star (Sect 53) and roughly the first four harmonics (ie 8657 43 and 34 days) The 172- and 86-day periods have about thesame amplitude while the 57-day period (Prot3) has 10 of themain amplitude The Prot4 signal has only about 4 of the mainamplitude The light curve analysis indicates that the signal due torotational modulation can largely be represented by the rotationalperiod (Prot) and its first harmonic (Prot2)
In order to test if the addition of RV sinusoidal signals atthe stellar rotation period and its harmonics can account for theadditional variation seen in our RV measurements we compareddifferent models by adding signals one by one The first model (P0)includes only the planet signal ie a Keplerian model fitted to theRV data using the same priors given in Table 5 but fixing epoch andperiod to the values derived by the transit modeling The next model(P1) is obtained from P0 by adding a sinusoidal signal at the rotationperiod of the star (Prot) Models P2 includes the first harmonic ofthe rotation period (Prot2) whereas model P3 account for the first(Prot2) and second (Prot3) harmonics While adding sinusoidalsignals we fitted for their amplitudes phases and periods We usedflat priors for the phases and amplitudes (details in Table 5) Weused a Gaussian prior for Prot using the value and its uncertaintyderived in Sect 53 The periods of the harmonic signals were leftfree to vary depending on the value assumed by Prot at each stepof the MCMC chains In order to check if the RV variation inducedby the planet is significant in our data set we also performed the fitusing models where the planetary signal was not included (modelsNP1 and NP2 see Table 4)
Table 4 shows the goodness of the fit for each model Thepreferred model is P2 (planet plus 2 sinusoidal signals at Prot andProt2) with the lowest Akaike Information Criteria (AIC) and max-imum likelihood This result is consistent with the Fourier analysisof the K2 light curve which suggests that the major contributionto the photometric variations arises from the stellar rotation periodand its first harmonic Our analysis provides also additional evi-dence that the Doppler motion induced by the planet is present inour RV data set First the planet signal does not significantly varyfor the P0 P1 P2 and P3 models (Table 4) Second the modelswith no planetary signal (NP1 and NP2) provide a poor fit to theRV measurements (Table 4)
To account for additional instrumental noise not included inthe nominal RV error bars andor imperfect treatment of the varioussources of RV variations we fitted for a jitter term for each instru-ment The final parameter estimates and their error bars are listedin Table 5 They are defined as the median and the 68 credibleinterval of the final posterior distributions The best fitting transitand RVmodels are displayed in Figure 4 along with the photometricand RV data points
72 Additional companion
Huang et al (2016) found that warm Jupiters with low eccentricities(e 04) have inner low-mass companions They used this evidenceas an argument in favour of the in situ formation since the planetmigration would have cleaned the warm Jupiter neighborhood Wesearched the light curve for additional transit signals but foundno evidence for an additional transiting planet in the system Asdescribed in the previous paragraph the periodogram of the RVresiduals showno significant peakwith false alarmprobability lowerthan 5
Figure 4 Top FIES (blue circles) HARPS-N (green diamonds) and HARPS (red squares) RV measurements versus time following the subtraction of thesystemic velocities for each instrument The 1σ uncertainties are marked using the same color used for each data-set The vertical gray lines mark the errorbars including jitter The solid line represents the best fitting RV model which includes the planet signal and the activity signal at the stellar rotation periodand its first harmonic The dashed dash-dotted and dotted lines show the RV contribution of K2-139 b stellar rotation and first harmonic respectively Lowerleft panel Transit light curve folded to the orbital period of K2-139 b and residuals The red points mark the K2 data and their error bars The solid line markthe re-binned best-fitting transit model Lower right panel Phase-folded RV curve of K2-139 b and best fitting Keplerian solution (solid line) following thesubtraction of the two additional sinusoidal signals used to account for the stellar activity The FIES HARPS and HARPS-N are corrected for the instrumentoffsets as derived from the global analysis
Table 4 Model comparison
Model Comment Npars Kb (m sminus1) χ2dof(a) ln L AIC(b)
P0 Planet signal 6 291 plusmn 20 61 356 -60P1 Planet signal + 1 sine-curve at Prot 9 294 plusmn 24 34 581 -98
P2 Planet signal + 2 sine-curves at Prot and Prot2 11 273+26minus25 38 601 -98
P3 Planet signal + 3 sine-curves at Prot Prot2 and Prot3 13 278+27minus26 53 593 -93
NP1 1 sine-curve at Prot (No planet signal) 6 0 185 -448 101NP2 2 sine-curves at Prot and Prot2 (No planet signal) 8 0 159 -120 40
Note ndash (a) χ2 value assuming no jitter (b)We used the Akaike Information Criteria (AIC = 2Nparsminus ln 2L) instead of the widely used Bayesian informationcriteria (BIC) because our RV data sample is small (19 data points) and BIC performs better for large samples (Burnham amp Anderson 2002)
73 Spot-crossing events
The passage of a planet in front of a spot can be detected as abump in the transit light curve (see eg Sanchis-Ojeda amp Winn2011) Spot-crossings events are clearly visible in the EVERESTtransit light curves (Fig 4) The same features appear at the sametimes and with consistent amplitudes in the Vanderburg amp Johnson(2014) data confirming that the bumps are real and not due to
systematics To assess whether the bumps significantly affect theparameter estimates we performed the joint analysis as describedin Sect 6 including all the transit data points We found that the finalparameters are consistent within 1-σ with those reported in Table 5
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 9
01 05 1 2 3 4 5 10Mass (MJup)
06
08
10
12
14
Rad
ius
(RJu
p)
H 0 M core 10 M core 25 M core 50 M core100 M core
Figure 5 Warm Jupiters (black squares Mp gt 03 MJup and10 6 Porb 6 100 days) whose mass and radius have been estimated with aprecision of at least 25 (as of January 2017 exoplaneteu) K2-139 b isshown with a filled red circle The solid line corresponds to a planet with apure hydrogen composition (Seager et al 2007) The dashed lines representthe Fortney et al (2007) models for planet core masses of 0 10 25 50and 100 Moplus The vertical dotted line marks the giant planet lower limit asdefined by Hatzes amp Rauer (2015)
74 Planetrsquos composition and formation scenario
With amass of Mp = 0387+0083minus0075 MJ and radius of Rp = 0808+0034
minus0033RJ (resulting in a mean density of ρp=091+024
minus020 g cmminus3) K2-139 bjoins the small group of well characterized warm Jupiters Fig 5shows the position of K2-139 b in themass-radius diagram for warmJupiters (Mp gt 03 MJup 10 6 Porb 6 100 days) whose massand radius have been determined with a precision better than 25(14 objects) Notably K2-139 b is the transiting warm Jupiter withthe lowest mass known to date if the definition of giant planetsgiven by Hatzes amp Rauer (2015) is adopted Fig 5 displays also theplanetary models of Fortney et al (2007) for different core massesand age between 10 and 45Gyrs The planet radius of K2-139 bcan be explained if the planet has a core10 of 49+19
minus17 Moplus containingsim40 of the total planetary mass We expect that K2-139 b has asolid core surrounded by a gaseous envelope
Rafikov (2006) found that a core of mass 5 ndash 20 Moplus at a semi-major axis between 01 and 10AU would be able to start the run-away accretion phase to form a gas giant planet in situ Howeveraccording to his models these kind of cores are unlikely to formowing to the high irradiation coming from the star Boley et al(2016) suggested instead that more massive cores (Mcore amp 20Moplus)can be built up from the merging of tightly packed inner planetsformed at the early stages of the circumstellar disc Batygin et al(2016) found a similar result and argued that the massive core ofHD149026b (Mcore asymp 100Moplus) could be explained by one or moresuper-Earths which merged and accreted the surrounding gas toform a gas-giant planet Huang et al (2016) suggested that thesecores can initiate runaway accretion if they are formed in a re-gion with enough gas around them while those without enoughvolatiles remain super-Earths and represent the population of mas-sive rocky planets unveiled by Kepler around solar-like stars (egDemory 2014) Based on these studies and given the semi-majoraxis of 0179+0021
minus0027 AU the 48 plusmn 14 Moplus core of K2-139 b could
10 Calculated by interpolating Fortney et al (2007)rsquos models
have formed the planet in situ We note that the metallicity of K2-139 is relatively high ([FeH] = 021 plusmn 005) suggesting that theprimordial circumstellar disc had a relatively high content of dustwhich would have enhanced the formation of the core of K2-139 b(see eg Johnson amp Li 2012) Alternatively the planet might haveformed beyond the snow line and migrated inwards via planet-discinteraction (see eg Baruteau et al 2014)
8 CONCLUSIONS
We confirmed the planetary nature and derived the orbital and mainphysical parameters of K2-139 b a warm Jupiter (Teq = 565+48
minus32 K)transiting an active (log RprimeHK = minus446 plusmn 006) K0V star every 29days We measured a planetary mass of Mp = 0387+0083
minus0075 MJ andradius of Rp = 0808+0034
minus0033 RJ At a separation of ap = 0179+0021minus0027
AU the mean density of ρp = 091+024minus020 g cmminus3 implies that the
planet has a core of 49+19minus17 Moplus according to the evolutionary mod-
els of Fortney et al (2007) K2-139 b joins the small group ofwell-characterized warm Jupiters whose mass and radius have beendetermined with a precision better than 25
The spin-orbit angle ie the angle between the spin axis ofthe star and the angular momentum vector of the orbit can pro-vide us with valuable information on the migration mechanismsof exoplanets (see eg Winn 2010 Morton amp Johnson 2011 Al-brecht et al 2012 Gandolfi et al 2012) Currently there are only4 warm Jupiters (Mp gt 03 MJup and 10 6 Porb 6 100 days)with measured obliquity11 From this perspective K2-139 is anideal target to measure the sky-project spin-orbit angle via obser-vations of the Rossiter-McLaughlin (RM) effect Assuming spin-orbit alignement the expected amplitude of the RM anomaly is∆RV asymp
radic1 minus b2 (RpR)2 v sin iasymp 25 m sminus1 (Winn 2010) Given
the brightness of the host star (V = 11653 mag) this amplitudecan easily be measured using state-of-the-art spectrographs suchas HARPSESO-36m Moreover the transit duration (sim5 hours)is shorter than the visibility of K2-139 which is sim9 hours fromLa Silla observatory (altitude higher than 30 above the horizon)
Alternatively the spin-orbit angle could be measured fromthe analysis of the spot-crossing events as described in Sanchis-Ojeda et al (2011) and Sanchis-Ojeda et al (2012) Anomaliesascribable to the passage of K2-139b in front of stellar spots arevisible in the 3 transit light curves observed by K2 Unfortunatelythe limited number of transits and the K2 long cadence data donot allow us to perform a meaningful quantitative analysis of thespot-crossing events Given the amplitude of the detected anoma-lies (sim01) space-based high-precision photometry is needed todetect the spot-crossing events Observations performed with theupcoming CHaracterising ExOPlanets Satellite (CHEOPS Broeget al 2013) would allow us to photometrically determine the spin-orbit angle of this system
ACKNOWLEDGEMENTS
We warmly thank the NOT ESO TNG staff members for theirunique support during the observations We are very thankful toXavier Bonfils Franccedilois Bouchy Martin Kuumlrster Tsevi MazehJorge Melendez and Nuno Santos who kindly agreed to exchange
11 Source httpwww2mpsmpgdehomeshellercontentmain_HRMhtml as of January 2017
MNRAS 000 1ndash12 (2017)
10 O Barragaacuten et al
HARPSandFIES timewith us Special thanks go toAntoninoLanzafor assisting us with the calculation of the gyro-age of the star Wealso greatly thank the anonymous referee for herhis careful re-view and suggestions which helped us to improve the manuscriptD Gandolfi gratefully acknowledges the financial support of theProgramma Giovani Ricercatori ndash Rita Levi Montalcini ndash Rien-tro dei Cervelli (2012) awarded by the Italian Ministry of Edu-cation Universities and Research (MIUR) Sz Csizmadia thanksthe Hungarian OTKA Grant K113117 H J Deeg and D Nespralacknowledge support by grant ESP2015-65712-C5-4-R of the Span-ish Secretary of State for RampDampi (MINECO) D Lorenzo-Oliveiraacknowledges the support from FAPESP (201620667-8) This re-search was supported by the Ministerio de Economia y Competi-tividad under project FIS2012-31079 The research leading to theseresults has received funding from the European Union SeventhFramework Programme (FP72013-2016) under grant agreementNo 312430 (OPTICON) Based on observations obtained a) withthe Nordic Optical Telescope (NOT) operated on the island of LaPalma jointly by Denmark Finland Iceland Norway and Swedenin the Spanish Observatorio del Roque de los Muchachos (ORM)of the Instituto de Astrofiacutesica de Canarias (IAC) b) with the Ital-ian Telescopio Nazionale Galileo (TNG) also operated at the ORM(IAC) on the island of La Palma by the INAF - Fundacioacuten GalileoGalilei c) the 36m ESO telescope at La Silla Observatory underprogramme ID 097C-0948 The data presented here were obtainedin part with ALFOSC which is provided by the Instituto de As-trofisica de Andalucia (IAA) under a joint agreement with the Uni-versity of Copenhagen and NOTSA This paper includes data col-lected by the Kepler mission Funding for the Kepler mission is pro-vided by the NASA Science Mission directorate Some of the datapresented in this paper were obtained from theMikulski Archive forSpace Telescopes (MAST) STScI is operated by the Association ofUniversities for Research in Astronomy Inc under NASA contractNAS5-26555 Support for MAST for non-HST data is provided bythe NASA Office of Space Science via grant NNX09AF08G and byother grants and contracts MF and CMP acknowledge generoussupport from the Swedish National Space Board C Eiroa and IRebollido are supported by Spanish grant AYA2014-55840-P PDacknowledge the support from INAF and Ministero dellrsquoIstruzionedellrsquoUniversitagrave e della Ricerca (MIUR) in the form of the grantldquoPremiale VLT 2012rdquo and ldquoThe Chemical and Dynamical Evolu-tion of the Milky Way and Local Group Galaxiesrdquo This work hasmade use of data from the European Space Agency (ESA) missionGaia (httpwwwcosmosesaintgaia) processed by theGaia Data Processing and Analysis Consortium (DPAC httpwwwcosmosesaintwebgaiadpacconsortium) Fundingfor the DPAC has been provided by national institutions in particu-lar the institutions participating in theGaiaMultilateral Agreement
REFERENCES
Albrecht S et al 2012 ApJ 757 18Allard F Homeier D Freytag B 2011 in Johns-Krull C BrowningM K
West A A eds Astronomical Society of the Pacific Conference SeriesVol 448 16th Cambridge Workshop on Cool Stars Stellar Systemsand the Sun p 91 (arXiv10115405)
Antonini F Hamers A S Lithwick Y 2016 preprint(arXiv160401781)
Barnes S A 2010 ApJ 722 222Barnes S A Kim Y-C 2010 ApJ 721 675Barragaacuten O Gandolfi D 2017 Exotrending Astrophysics Source Code
Library (ascl1706001)
Barragaacuten O et al 2016 AJ 152 193Barragaacuten O Gandolfi D Antoniciello G 2017 pyaneti Astrophysics
Source Code Library (ascl1707003)Baruteau C et al 2014 Protostars and Planets VI pp 667ndash689Batygin K Bodenheimer P H Laughlin G P 2016 ApJ 829 114Boley A C Granados Contreras A P Gladman B 2016 ApJ 817 L17Boyajian T S et al 2013 ApJ 771 40Brahm R et al 2016 AJ 151 89Bressan A Marigo P Girardi L Salasnich B Dal Cero C Rubele S
Nanni A 2012 MNRAS 427 127Broeg C et al 2013 in European Physical Journal Web of Conferences p
03005 (arXiv13052270) doi101051epjconf20134703005Bruntt H et al 2010 MNRAS 405 1907Buchhave L A et al 2010 ApJ 720 1118Burnham K Anderson D 2002 Model Selection and Multimodel Infer-
ence A Practical Information-Theoretic Approach NewYork Springer-Verlag
Cabrera J et al 2009 AampA 506 501Cabrera J Csizmadia S Erikson A Rauer H Kirste S 2012 AampA 548
A44Cabrera J et al 2014 ApJ 781 18Cantat-Gaudin T et al 2014 AampA 562 A10Cardelli J A Clayton G C Mathis J S 1989 ApJ 345 245Carone L et al 2012 AampA 538 A112Carpano S et al 2009 AampA 506 491Castelli F Kurucz R L 2004 preprintCavarroc C et al 2012 ApampSS 337 511Chabrier G 2001 ApJ 554 1274Cosentino R et al 2012 in Ground-based and Airborne Instrumentation
for Astronomy IV p 84461V doi10111712925738Cutri R M et al 2003 2MASS All Sky Catalog of point sourcesCutri R M et al 2012 Technical report Explanatory Supplement to the
WISE All-Sky Data Release ProductsDawson R I Johnson J A Morton T D Crepp J R Fabrycky D C
Murray-Clay R A Howard A W 2012 ApJ 761 163Deeg H J et al 2010 Nature 464 384Demory B-O 2014 ApJ 789 L20Dong S Katz B Socrates A 2014 ApJ 781 L5Doyle A P Davies G R Smalley B Chaplin W J Elsworth Y 2014
MNRAS 444 3592Dumusque X Boisse I Santos N C 2014 ApJ 796 132Endl M Cochran W D 2016 PASP 128 094502Erikson A et al 2012 AampA 539 A14Fabricius C et al 2016 AampA 595 A3Fortney J J Marley M S Barnes J W 2007 ApJ 659 1661Fossati et al 2017 AampA submittedFrandsenS LindbergB 1999 inKarttunenH PiirolaV edsAstrophysics
with the NOT p 71Frewen S F N Hansen B M S 2016 MNRAS 455 1538Gandolfi D et al 2008 ApJ 687 1303Gandolfi D et al 2012 AampA 543 L5Gandolfi D et al 2015 AampA 576 A11Gray R O 1999 SPECTRUM A stellar spectral synthesis program As-
trophysics Source Code Library (ascl9910002)Grziwa S Paumltzold M 2016 preprint (arXiv160708417)Grziwa S Paumltzold M Carone L 2012 MNRAS 420 1045Hamers A S Antonini F Lithwick Y Perets H B Portegies Zwart S F
2016 preprint (arXiv160607438)Hatzes A P 2002 Astronomische Nachrichten 323 392Hatzes A P Rauer H 2015 ApJ 810 L25Heiter U et al 2015 Phys Scr 90 054010Huang C Wu Y Triaud A H M J 2016 ApJ 825 98Jenkins J S et al 2017 MNRAS 466 443Johnson J L Li H 2012 ApJ 751 81Kipping D M 2010 MNRAS 408 1758Kley W Nelson R P 2012 ARAampA 50 211Kovaacutecs G Zucker S Mazeh T 2002 AampA 391 369
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 11
Table 5 K2-139 system parameters
Parameter Prior(a) Final value
Stellar parametersStar mass M (M) middot middot middot 0919 plusmn 0033Star radius R (R) middot middot middot 0862 plusmn 0032Star density ρ (from spectroscopy g cmminus3) middot middot middot 202+025
minus022
Star density ρ (from light curve g cmminus3) middot middot middot 211+074minus081
Effective Temperature Teff (K) middot middot middot 5340 plusmn 110Surface gravity log g (cgs) middot middot middot 450 plusmn 009Iron abundance [FeH] (dex) middot middot middot 022 plusmn 008Microturbulent velocity vmic ( km sminus1) middot middot middot 09 plusmn 01Macroturbulent velocity vmac ( km sminus1) middot middot middot 25 plusmn 06Projected rotational velocity v sin i ( km sminus1) middot middot middot 28 plusmn 06Rotational period Prot (days) middot middot middot 1724 plusmn 012Activity index(b) logRprimeHK middot middot middot minus446 plusmn 006Gyrochronological age (Gyr) middot middot middot 18 plusmn 03Interstellar extinction AV (mag) middot middot middot 007 plusmn 005Star distance d (pc) middot middot middot 152 plusmn 10
Model parameters of K2-139 bOrbital period Porb (days) U[283773 283873] 2838236 plusmn 000026Transit epoch T0 (BJDTDBminus2 450 000) U[73258120 73258220] 732581714 plusmn 000033Scaled semi-major axis aR U[12 100] 448+47
minus67Planet-to-star radius ratio RpR U[0 02] 00961+00023
minus00015Impact parameter b U[0 12] 030+021
minus019radice sinω U[minus1 1](c) 010+029
minus030radice cosω U[minus1 1](c) 006+024
minus027Radial velocity semi-amplitude variation K ( m sminus1) U[0 200] 277+60
Note ndash The adopted Sun and Jupiter units follow the recommendations from the International Astronomical Union (Prša et al 2016) (a) U[a b] refersto uniform priors between a and b N[a b] means Gaussian priors with mean a and standard deviation b and F[a] to a fixed a value (b) Corrected forinterstellar reddening following Fossati et al (2017) The correction is +006 (c) The code always ensures that e lt 1 (d) Assuming albedo = 0
MNRAS 000 1ndash12 (2017)
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
MNRAS 000 1ndash12 (2017)
1 Introduction
2 K2 photometry
3 ALFOSC imaging
4 High-resolution spectroscopy
5 Stellar parameters
51 Spectral analysis
52 Interstellar extinction
53 Rotational period
54 Stellar mass radius and age
6 Joint RV-transit fit
7 Results and discussion
71 Stellar activity modeling
72 Additional companion
73 Spot-crossing events
74 Planets composition and formation scenario
8 Conclusions
Acknowledgements
The transiting warm Jupiter K2-139 b 7
Figure 3 Generalized Lomb-Scargle periodogram of the combined FIESHARPS and HARPS-N Doppler datasets From top to bottom the RV datathe RV residuals after subtracting the transiting planet signal the BIS andFWHM of the CCF and the window function The dotted vertical blue linesmark the frequencies at the orbital period as well as at the stellar rotationperiod and its first two harmonics The dashed vertical red lines mark the5 false alarm probabilities as derived using the bootstrap method The redarrows in the lower panel mark the two peaks presented in the main text
in the periodogram of the RV data and RV residuals as well as inthe periodogram of the activity indicators could be explained by thelimited number of available measurements and their uncertaintiesWe conclude that we cannot exclude the existence of spot-inducedsignals in our RV measurements
Photometric and radial velocity variations due to rotationalmodulation can be complex with not only the rotational period Protpresent but also its harmonics eg Prot2 Prot3 Assuming thatthe surface structures responsible for this modulation (eg coolspots) are not evolving rapidly then the simplest representation ofthe rotationalmodulation is through the Fourier components definedby the rotation period and its harmonics Figure 1 shows that theevolution time-scale of the active regions in the stellar surface islonger than the 80-day duration of the K2 campaign Since our RVfollow-up spans 55 days we can assume that any activity-inducedRV signal is coherent within our observing window This approachhas been used previously for other planetary systems orbiting activestars (eg Pepe et al 2013)
The Fourier analysis of the K2 light curve is the best way tomeasure the contribution of the rotation period and its harmonics tothe quasi-periodic photometric variability of the star We thereforeanalyzed the K2 light curve using a pre-whitening procedure Thatis the dominant period was found a sine-fit made to the data andsubtracted and additional periods searched in the residual data
We used the program Period04 (Lenz amp Breger 2005) for thisprocedure
The dominant periods are sim172 days ie the rotation periodof the star (Sect 53) and roughly the first four harmonics (ie 8657 43 and 34 days) The 172- and 86-day periods have about thesame amplitude while the 57-day period (Prot3) has 10 of themain amplitude The Prot4 signal has only about 4 of the mainamplitude The light curve analysis indicates that the signal due torotational modulation can largely be represented by the rotationalperiod (Prot) and its first harmonic (Prot2)
In order to test if the addition of RV sinusoidal signals atthe stellar rotation period and its harmonics can account for theadditional variation seen in our RV measurements we compareddifferent models by adding signals one by one The first model (P0)includes only the planet signal ie a Keplerian model fitted to theRV data using the same priors given in Table 5 but fixing epoch andperiod to the values derived by the transit modeling The next model(P1) is obtained from P0 by adding a sinusoidal signal at the rotationperiod of the star (Prot) Models P2 includes the first harmonic ofthe rotation period (Prot2) whereas model P3 account for the first(Prot2) and second (Prot3) harmonics While adding sinusoidalsignals we fitted for their amplitudes phases and periods We usedflat priors for the phases and amplitudes (details in Table 5) Weused a Gaussian prior for Prot using the value and its uncertaintyderived in Sect 53 The periods of the harmonic signals were leftfree to vary depending on the value assumed by Prot at each stepof the MCMC chains In order to check if the RV variation inducedby the planet is significant in our data set we also performed the fitusing models where the planetary signal was not included (modelsNP1 and NP2 see Table 4)
Table 4 shows the goodness of the fit for each model Thepreferred model is P2 (planet plus 2 sinusoidal signals at Prot andProt2) with the lowest Akaike Information Criteria (AIC) and max-imum likelihood This result is consistent with the Fourier analysisof the K2 light curve which suggests that the major contributionto the photometric variations arises from the stellar rotation periodand its first harmonic Our analysis provides also additional evi-dence that the Doppler motion induced by the planet is present inour RV data set First the planet signal does not significantly varyfor the P0 P1 P2 and P3 models (Table 4) Second the modelswith no planetary signal (NP1 and NP2) provide a poor fit to theRV measurements (Table 4)
To account for additional instrumental noise not included inthe nominal RV error bars andor imperfect treatment of the varioussources of RV variations we fitted for a jitter term for each instru-ment The final parameter estimates and their error bars are listedin Table 5 They are defined as the median and the 68 credibleinterval of the final posterior distributions The best fitting transitand RVmodels are displayed in Figure 4 along with the photometricand RV data points
72 Additional companion
Huang et al (2016) found that warm Jupiters with low eccentricities(e 04) have inner low-mass companions They used this evidenceas an argument in favour of the in situ formation since the planetmigration would have cleaned the warm Jupiter neighborhood Wesearched the light curve for additional transit signals but foundno evidence for an additional transiting planet in the system Asdescribed in the previous paragraph the periodogram of the RVresiduals showno significant peakwith false alarmprobability lowerthan 5
Figure 4 Top FIES (blue circles) HARPS-N (green diamonds) and HARPS (red squares) RV measurements versus time following the subtraction of thesystemic velocities for each instrument The 1σ uncertainties are marked using the same color used for each data-set The vertical gray lines mark the errorbars including jitter The solid line represents the best fitting RV model which includes the planet signal and the activity signal at the stellar rotation periodand its first harmonic The dashed dash-dotted and dotted lines show the RV contribution of K2-139 b stellar rotation and first harmonic respectively Lowerleft panel Transit light curve folded to the orbital period of K2-139 b and residuals The red points mark the K2 data and their error bars The solid line markthe re-binned best-fitting transit model Lower right panel Phase-folded RV curve of K2-139 b and best fitting Keplerian solution (solid line) following thesubtraction of the two additional sinusoidal signals used to account for the stellar activity The FIES HARPS and HARPS-N are corrected for the instrumentoffsets as derived from the global analysis
Table 4 Model comparison
Model Comment Npars Kb (m sminus1) χ2dof(a) ln L AIC(b)
P0 Planet signal 6 291 plusmn 20 61 356 -60P1 Planet signal + 1 sine-curve at Prot 9 294 plusmn 24 34 581 -98
P2 Planet signal + 2 sine-curves at Prot and Prot2 11 273+26minus25 38 601 -98
P3 Planet signal + 3 sine-curves at Prot Prot2 and Prot3 13 278+27minus26 53 593 -93
NP1 1 sine-curve at Prot (No planet signal) 6 0 185 -448 101NP2 2 sine-curves at Prot and Prot2 (No planet signal) 8 0 159 -120 40
Note ndash (a) χ2 value assuming no jitter (b)We used the Akaike Information Criteria (AIC = 2Nparsminus ln 2L) instead of the widely used Bayesian informationcriteria (BIC) because our RV data sample is small (19 data points) and BIC performs better for large samples (Burnham amp Anderson 2002)
73 Spot-crossing events
The passage of a planet in front of a spot can be detected as abump in the transit light curve (see eg Sanchis-Ojeda amp Winn2011) Spot-crossings events are clearly visible in the EVERESTtransit light curves (Fig 4) The same features appear at the sametimes and with consistent amplitudes in the Vanderburg amp Johnson(2014) data confirming that the bumps are real and not due to
systematics To assess whether the bumps significantly affect theparameter estimates we performed the joint analysis as describedin Sect 6 including all the transit data points We found that the finalparameters are consistent within 1-σ with those reported in Table 5
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 9
01 05 1 2 3 4 5 10Mass (MJup)
06
08
10
12
14
Rad
ius
(RJu
p)
H 0 M core 10 M core 25 M core 50 M core100 M core
Figure 5 Warm Jupiters (black squares Mp gt 03 MJup and10 6 Porb 6 100 days) whose mass and radius have been estimated with aprecision of at least 25 (as of January 2017 exoplaneteu) K2-139 b isshown with a filled red circle The solid line corresponds to a planet with apure hydrogen composition (Seager et al 2007) The dashed lines representthe Fortney et al (2007) models for planet core masses of 0 10 25 50and 100 Moplus The vertical dotted line marks the giant planet lower limit asdefined by Hatzes amp Rauer (2015)
74 Planetrsquos composition and formation scenario
With amass of Mp = 0387+0083minus0075 MJ and radius of Rp = 0808+0034
minus0033RJ (resulting in a mean density of ρp=091+024
minus020 g cmminus3) K2-139 bjoins the small group of well characterized warm Jupiters Fig 5shows the position of K2-139 b in themass-radius diagram for warmJupiters (Mp gt 03 MJup 10 6 Porb 6 100 days) whose massand radius have been determined with a precision better than 25(14 objects) Notably K2-139 b is the transiting warm Jupiter withthe lowest mass known to date if the definition of giant planetsgiven by Hatzes amp Rauer (2015) is adopted Fig 5 displays also theplanetary models of Fortney et al (2007) for different core massesand age between 10 and 45Gyrs The planet radius of K2-139 bcan be explained if the planet has a core10 of 49+19
minus17 Moplus containingsim40 of the total planetary mass We expect that K2-139 b has asolid core surrounded by a gaseous envelope
Rafikov (2006) found that a core of mass 5 ndash 20 Moplus at a semi-major axis between 01 and 10AU would be able to start the run-away accretion phase to form a gas giant planet in situ Howeveraccording to his models these kind of cores are unlikely to formowing to the high irradiation coming from the star Boley et al(2016) suggested instead that more massive cores (Mcore amp 20Moplus)can be built up from the merging of tightly packed inner planetsformed at the early stages of the circumstellar disc Batygin et al(2016) found a similar result and argued that the massive core ofHD149026b (Mcore asymp 100Moplus) could be explained by one or moresuper-Earths which merged and accreted the surrounding gas toform a gas-giant planet Huang et al (2016) suggested that thesecores can initiate runaway accretion if they are formed in a re-gion with enough gas around them while those without enoughvolatiles remain super-Earths and represent the population of mas-sive rocky planets unveiled by Kepler around solar-like stars (egDemory 2014) Based on these studies and given the semi-majoraxis of 0179+0021
minus0027 AU the 48 plusmn 14 Moplus core of K2-139 b could
10 Calculated by interpolating Fortney et al (2007)rsquos models
have formed the planet in situ We note that the metallicity of K2-139 is relatively high ([FeH] = 021 plusmn 005) suggesting that theprimordial circumstellar disc had a relatively high content of dustwhich would have enhanced the formation of the core of K2-139 b(see eg Johnson amp Li 2012) Alternatively the planet might haveformed beyond the snow line and migrated inwards via planet-discinteraction (see eg Baruteau et al 2014)
8 CONCLUSIONS
We confirmed the planetary nature and derived the orbital and mainphysical parameters of K2-139 b a warm Jupiter (Teq = 565+48
minus32 K)transiting an active (log RprimeHK = minus446 plusmn 006) K0V star every 29days We measured a planetary mass of Mp = 0387+0083
minus0075 MJ andradius of Rp = 0808+0034
minus0033 RJ At a separation of ap = 0179+0021minus0027
AU the mean density of ρp = 091+024minus020 g cmminus3 implies that the
planet has a core of 49+19minus17 Moplus according to the evolutionary mod-
els of Fortney et al (2007) K2-139 b joins the small group ofwell-characterized warm Jupiters whose mass and radius have beendetermined with a precision better than 25
The spin-orbit angle ie the angle between the spin axis ofthe star and the angular momentum vector of the orbit can pro-vide us with valuable information on the migration mechanismsof exoplanets (see eg Winn 2010 Morton amp Johnson 2011 Al-brecht et al 2012 Gandolfi et al 2012) Currently there are only4 warm Jupiters (Mp gt 03 MJup and 10 6 Porb 6 100 days)with measured obliquity11 From this perspective K2-139 is anideal target to measure the sky-project spin-orbit angle via obser-vations of the Rossiter-McLaughlin (RM) effect Assuming spin-orbit alignement the expected amplitude of the RM anomaly is∆RV asymp
radic1 minus b2 (RpR)2 v sin iasymp 25 m sminus1 (Winn 2010) Given
the brightness of the host star (V = 11653 mag) this amplitudecan easily be measured using state-of-the-art spectrographs suchas HARPSESO-36m Moreover the transit duration (sim5 hours)is shorter than the visibility of K2-139 which is sim9 hours fromLa Silla observatory (altitude higher than 30 above the horizon)
Alternatively the spin-orbit angle could be measured fromthe analysis of the spot-crossing events as described in Sanchis-Ojeda et al (2011) and Sanchis-Ojeda et al (2012) Anomaliesascribable to the passage of K2-139b in front of stellar spots arevisible in the 3 transit light curves observed by K2 Unfortunatelythe limited number of transits and the K2 long cadence data donot allow us to perform a meaningful quantitative analysis of thespot-crossing events Given the amplitude of the detected anoma-lies (sim01) space-based high-precision photometry is needed todetect the spot-crossing events Observations performed with theupcoming CHaracterising ExOPlanets Satellite (CHEOPS Broeget al 2013) would allow us to photometrically determine the spin-orbit angle of this system
ACKNOWLEDGEMENTS
We warmly thank the NOT ESO TNG staff members for theirunique support during the observations We are very thankful toXavier Bonfils Franccedilois Bouchy Martin Kuumlrster Tsevi MazehJorge Melendez and Nuno Santos who kindly agreed to exchange
11 Source httpwww2mpsmpgdehomeshellercontentmain_HRMhtml as of January 2017
MNRAS 000 1ndash12 (2017)
10 O Barragaacuten et al
HARPSandFIES timewith us Special thanks go toAntoninoLanzafor assisting us with the calculation of the gyro-age of the star Wealso greatly thank the anonymous referee for herhis careful re-view and suggestions which helped us to improve the manuscriptD Gandolfi gratefully acknowledges the financial support of theProgramma Giovani Ricercatori ndash Rita Levi Montalcini ndash Rien-tro dei Cervelli (2012) awarded by the Italian Ministry of Edu-cation Universities and Research (MIUR) Sz Csizmadia thanksthe Hungarian OTKA Grant K113117 H J Deeg and D Nespralacknowledge support by grant ESP2015-65712-C5-4-R of the Span-ish Secretary of State for RampDampi (MINECO) D Lorenzo-Oliveiraacknowledges the support from FAPESP (201620667-8) This re-search was supported by the Ministerio de Economia y Competi-tividad under project FIS2012-31079 The research leading to theseresults has received funding from the European Union SeventhFramework Programme (FP72013-2016) under grant agreementNo 312430 (OPTICON) Based on observations obtained a) withthe Nordic Optical Telescope (NOT) operated on the island of LaPalma jointly by Denmark Finland Iceland Norway and Swedenin the Spanish Observatorio del Roque de los Muchachos (ORM)of the Instituto de Astrofiacutesica de Canarias (IAC) b) with the Ital-ian Telescopio Nazionale Galileo (TNG) also operated at the ORM(IAC) on the island of La Palma by the INAF - Fundacioacuten GalileoGalilei c) the 36m ESO telescope at La Silla Observatory underprogramme ID 097C-0948 The data presented here were obtainedin part with ALFOSC which is provided by the Instituto de As-trofisica de Andalucia (IAA) under a joint agreement with the Uni-versity of Copenhagen and NOTSA This paper includes data col-lected by the Kepler mission Funding for the Kepler mission is pro-vided by the NASA Science Mission directorate Some of the datapresented in this paper were obtained from theMikulski Archive forSpace Telescopes (MAST) STScI is operated by the Association ofUniversities for Research in Astronomy Inc under NASA contractNAS5-26555 Support for MAST for non-HST data is provided bythe NASA Office of Space Science via grant NNX09AF08G and byother grants and contracts MF and CMP acknowledge generoussupport from the Swedish National Space Board C Eiroa and IRebollido are supported by Spanish grant AYA2014-55840-P PDacknowledge the support from INAF and Ministero dellrsquoIstruzionedellrsquoUniversitagrave e della Ricerca (MIUR) in the form of the grantldquoPremiale VLT 2012rdquo and ldquoThe Chemical and Dynamical Evolu-tion of the Milky Way and Local Group Galaxiesrdquo This work hasmade use of data from the European Space Agency (ESA) missionGaia (httpwwwcosmosesaintgaia) processed by theGaia Data Processing and Analysis Consortium (DPAC httpwwwcosmosesaintwebgaiadpacconsortium) Fundingfor the DPAC has been provided by national institutions in particu-lar the institutions participating in theGaiaMultilateral Agreement
REFERENCES
Albrecht S et al 2012 ApJ 757 18Allard F Homeier D Freytag B 2011 in Johns-Krull C BrowningM K
West A A eds Astronomical Society of the Pacific Conference SeriesVol 448 16th Cambridge Workshop on Cool Stars Stellar Systemsand the Sun p 91 (arXiv10115405)
Antonini F Hamers A S Lithwick Y 2016 preprint(arXiv160401781)
Barnes S A 2010 ApJ 722 222Barnes S A Kim Y-C 2010 ApJ 721 675Barragaacuten O Gandolfi D 2017 Exotrending Astrophysics Source Code
Library (ascl1706001)
Barragaacuten O et al 2016 AJ 152 193Barragaacuten O Gandolfi D Antoniciello G 2017 pyaneti Astrophysics
Source Code Library (ascl1707003)Baruteau C et al 2014 Protostars and Planets VI pp 667ndash689Batygin K Bodenheimer P H Laughlin G P 2016 ApJ 829 114Boley A C Granados Contreras A P Gladman B 2016 ApJ 817 L17Boyajian T S et al 2013 ApJ 771 40Brahm R et al 2016 AJ 151 89Bressan A Marigo P Girardi L Salasnich B Dal Cero C Rubele S
Nanni A 2012 MNRAS 427 127Broeg C et al 2013 in European Physical Journal Web of Conferences p
03005 (arXiv13052270) doi101051epjconf20134703005Bruntt H et al 2010 MNRAS 405 1907Buchhave L A et al 2010 ApJ 720 1118Burnham K Anderson D 2002 Model Selection and Multimodel Infer-
ence A Practical Information-Theoretic Approach NewYork Springer-Verlag
Cabrera J et al 2009 AampA 506 501Cabrera J Csizmadia S Erikson A Rauer H Kirste S 2012 AampA 548
A44Cabrera J et al 2014 ApJ 781 18Cantat-Gaudin T et al 2014 AampA 562 A10Cardelli J A Clayton G C Mathis J S 1989 ApJ 345 245Carone L et al 2012 AampA 538 A112Carpano S et al 2009 AampA 506 491Castelli F Kurucz R L 2004 preprintCavarroc C et al 2012 ApampSS 337 511Chabrier G 2001 ApJ 554 1274Cosentino R et al 2012 in Ground-based and Airborne Instrumentation
for Astronomy IV p 84461V doi10111712925738Cutri R M et al 2003 2MASS All Sky Catalog of point sourcesCutri R M et al 2012 Technical report Explanatory Supplement to the
WISE All-Sky Data Release ProductsDawson R I Johnson J A Morton T D Crepp J R Fabrycky D C
Murray-Clay R A Howard A W 2012 ApJ 761 163Deeg H J et al 2010 Nature 464 384Demory B-O 2014 ApJ 789 L20Dong S Katz B Socrates A 2014 ApJ 781 L5Doyle A P Davies G R Smalley B Chaplin W J Elsworth Y 2014
MNRAS 444 3592Dumusque X Boisse I Santos N C 2014 ApJ 796 132Endl M Cochran W D 2016 PASP 128 094502Erikson A et al 2012 AampA 539 A14Fabricius C et al 2016 AampA 595 A3Fortney J J Marley M S Barnes J W 2007 ApJ 659 1661Fossati et al 2017 AampA submittedFrandsenS LindbergB 1999 inKarttunenH PiirolaV edsAstrophysics
with the NOT p 71Frewen S F N Hansen B M S 2016 MNRAS 455 1538Gandolfi D et al 2008 ApJ 687 1303Gandolfi D et al 2012 AampA 543 L5Gandolfi D et al 2015 AampA 576 A11Gray R O 1999 SPECTRUM A stellar spectral synthesis program As-
trophysics Source Code Library (ascl9910002)Grziwa S Paumltzold M 2016 preprint (arXiv160708417)Grziwa S Paumltzold M Carone L 2012 MNRAS 420 1045Hamers A S Antonini F Lithwick Y Perets H B Portegies Zwart S F
2016 preprint (arXiv160607438)Hatzes A P 2002 Astronomische Nachrichten 323 392Hatzes A P Rauer H 2015 ApJ 810 L25Heiter U et al 2015 Phys Scr 90 054010Huang C Wu Y Triaud A H M J 2016 ApJ 825 98Jenkins J S et al 2017 MNRAS 466 443Johnson J L Li H 2012 ApJ 751 81Kipping D M 2010 MNRAS 408 1758Kley W Nelson R P 2012 ARAampA 50 211Kovaacutecs G Zucker S Mazeh T 2002 AampA 391 369
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 11
Table 5 K2-139 system parameters
Parameter Prior(a) Final value
Stellar parametersStar mass M (M) middot middot middot 0919 plusmn 0033Star radius R (R) middot middot middot 0862 plusmn 0032Star density ρ (from spectroscopy g cmminus3) middot middot middot 202+025
minus022
Star density ρ (from light curve g cmminus3) middot middot middot 211+074minus081
Effective Temperature Teff (K) middot middot middot 5340 plusmn 110Surface gravity log g (cgs) middot middot middot 450 plusmn 009Iron abundance [FeH] (dex) middot middot middot 022 plusmn 008Microturbulent velocity vmic ( km sminus1) middot middot middot 09 plusmn 01Macroturbulent velocity vmac ( km sminus1) middot middot middot 25 plusmn 06Projected rotational velocity v sin i ( km sminus1) middot middot middot 28 plusmn 06Rotational period Prot (days) middot middot middot 1724 plusmn 012Activity index(b) logRprimeHK middot middot middot minus446 plusmn 006Gyrochronological age (Gyr) middot middot middot 18 plusmn 03Interstellar extinction AV (mag) middot middot middot 007 plusmn 005Star distance d (pc) middot middot middot 152 plusmn 10
Model parameters of K2-139 bOrbital period Porb (days) U[283773 283873] 2838236 plusmn 000026Transit epoch T0 (BJDTDBminus2 450 000) U[73258120 73258220] 732581714 plusmn 000033Scaled semi-major axis aR U[12 100] 448+47
minus67Planet-to-star radius ratio RpR U[0 02] 00961+00023
minus00015Impact parameter b U[0 12] 030+021
minus019radice sinω U[minus1 1](c) 010+029
minus030radice cosω U[minus1 1](c) 006+024
minus027Radial velocity semi-amplitude variation K ( m sminus1) U[0 200] 277+60
Note ndash The adopted Sun and Jupiter units follow the recommendations from the International Astronomical Union (Prša et al 2016) (a) U[a b] refersto uniform priors between a and b N[a b] means Gaussian priors with mean a and standard deviation b and F[a] to a fixed a value (b) Corrected forinterstellar reddening following Fossati et al (2017) The correction is +006 (c) The code always ensures that e lt 1 (d) Assuming albedo = 0
MNRAS 000 1ndash12 (2017)
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
Figure 4 Top FIES (blue circles) HARPS-N (green diamonds) and HARPS (red squares) RV measurements versus time following the subtraction of thesystemic velocities for each instrument The 1σ uncertainties are marked using the same color used for each data-set The vertical gray lines mark the errorbars including jitter The solid line represents the best fitting RV model which includes the planet signal and the activity signal at the stellar rotation periodand its first harmonic The dashed dash-dotted and dotted lines show the RV contribution of K2-139 b stellar rotation and first harmonic respectively Lowerleft panel Transit light curve folded to the orbital period of K2-139 b and residuals The red points mark the K2 data and their error bars The solid line markthe re-binned best-fitting transit model Lower right panel Phase-folded RV curve of K2-139 b and best fitting Keplerian solution (solid line) following thesubtraction of the two additional sinusoidal signals used to account for the stellar activity The FIES HARPS and HARPS-N are corrected for the instrumentoffsets as derived from the global analysis
Table 4 Model comparison
Model Comment Npars Kb (m sminus1) χ2dof(a) ln L AIC(b)
P0 Planet signal 6 291 plusmn 20 61 356 -60P1 Planet signal + 1 sine-curve at Prot 9 294 plusmn 24 34 581 -98
P2 Planet signal + 2 sine-curves at Prot and Prot2 11 273+26minus25 38 601 -98
P3 Planet signal + 3 sine-curves at Prot Prot2 and Prot3 13 278+27minus26 53 593 -93
NP1 1 sine-curve at Prot (No planet signal) 6 0 185 -448 101NP2 2 sine-curves at Prot and Prot2 (No planet signal) 8 0 159 -120 40
Note ndash (a) χ2 value assuming no jitter (b)We used the Akaike Information Criteria (AIC = 2Nparsminus ln 2L) instead of the widely used Bayesian informationcriteria (BIC) because our RV data sample is small (19 data points) and BIC performs better for large samples (Burnham amp Anderson 2002)
73 Spot-crossing events
The passage of a planet in front of a spot can be detected as abump in the transit light curve (see eg Sanchis-Ojeda amp Winn2011) Spot-crossings events are clearly visible in the EVERESTtransit light curves (Fig 4) The same features appear at the sametimes and with consistent amplitudes in the Vanderburg amp Johnson(2014) data confirming that the bumps are real and not due to
systematics To assess whether the bumps significantly affect theparameter estimates we performed the joint analysis as describedin Sect 6 including all the transit data points We found that the finalparameters are consistent within 1-σ with those reported in Table 5
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 9
01 05 1 2 3 4 5 10Mass (MJup)
06
08
10
12
14
Rad
ius
(RJu
p)
H 0 M core 10 M core 25 M core 50 M core100 M core
Figure 5 Warm Jupiters (black squares Mp gt 03 MJup and10 6 Porb 6 100 days) whose mass and radius have been estimated with aprecision of at least 25 (as of January 2017 exoplaneteu) K2-139 b isshown with a filled red circle The solid line corresponds to a planet with apure hydrogen composition (Seager et al 2007) The dashed lines representthe Fortney et al (2007) models for planet core masses of 0 10 25 50and 100 Moplus The vertical dotted line marks the giant planet lower limit asdefined by Hatzes amp Rauer (2015)
74 Planetrsquos composition and formation scenario
With amass of Mp = 0387+0083minus0075 MJ and radius of Rp = 0808+0034
minus0033RJ (resulting in a mean density of ρp=091+024
minus020 g cmminus3) K2-139 bjoins the small group of well characterized warm Jupiters Fig 5shows the position of K2-139 b in themass-radius diagram for warmJupiters (Mp gt 03 MJup 10 6 Porb 6 100 days) whose massand radius have been determined with a precision better than 25(14 objects) Notably K2-139 b is the transiting warm Jupiter withthe lowest mass known to date if the definition of giant planetsgiven by Hatzes amp Rauer (2015) is adopted Fig 5 displays also theplanetary models of Fortney et al (2007) for different core massesand age between 10 and 45Gyrs The planet radius of K2-139 bcan be explained if the planet has a core10 of 49+19
minus17 Moplus containingsim40 of the total planetary mass We expect that K2-139 b has asolid core surrounded by a gaseous envelope
Rafikov (2006) found that a core of mass 5 ndash 20 Moplus at a semi-major axis between 01 and 10AU would be able to start the run-away accretion phase to form a gas giant planet in situ Howeveraccording to his models these kind of cores are unlikely to formowing to the high irradiation coming from the star Boley et al(2016) suggested instead that more massive cores (Mcore amp 20Moplus)can be built up from the merging of tightly packed inner planetsformed at the early stages of the circumstellar disc Batygin et al(2016) found a similar result and argued that the massive core ofHD149026b (Mcore asymp 100Moplus) could be explained by one or moresuper-Earths which merged and accreted the surrounding gas toform a gas-giant planet Huang et al (2016) suggested that thesecores can initiate runaway accretion if they are formed in a re-gion with enough gas around them while those without enoughvolatiles remain super-Earths and represent the population of mas-sive rocky planets unveiled by Kepler around solar-like stars (egDemory 2014) Based on these studies and given the semi-majoraxis of 0179+0021
minus0027 AU the 48 plusmn 14 Moplus core of K2-139 b could
10 Calculated by interpolating Fortney et al (2007)rsquos models
have formed the planet in situ We note that the metallicity of K2-139 is relatively high ([FeH] = 021 plusmn 005) suggesting that theprimordial circumstellar disc had a relatively high content of dustwhich would have enhanced the formation of the core of K2-139 b(see eg Johnson amp Li 2012) Alternatively the planet might haveformed beyond the snow line and migrated inwards via planet-discinteraction (see eg Baruteau et al 2014)
8 CONCLUSIONS
We confirmed the planetary nature and derived the orbital and mainphysical parameters of K2-139 b a warm Jupiter (Teq = 565+48
minus32 K)transiting an active (log RprimeHK = minus446 plusmn 006) K0V star every 29days We measured a planetary mass of Mp = 0387+0083
minus0075 MJ andradius of Rp = 0808+0034
minus0033 RJ At a separation of ap = 0179+0021minus0027
AU the mean density of ρp = 091+024minus020 g cmminus3 implies that the
planet has a core of 49+19minus17 Moplus according to the evolutionary mod-
els of Fortney et al (2007) K2-139 b joins the small group ofwell-characterized warm Jupiters whose mass and radius have beendetermined with a precision better than 25
The spin-orbit angle ie the angle between the spin axis ofthe star and the angular momentum vector of the orbit can pro-vide us with valuable information on the migration mechanismsof exoplanets (see eg Winn 2010 Morton amp Johnson 2011 Al-brecht et al 2012 Gandolfi et al 2012) Currently there are only4 warm Jupiters (Mp gt 03 MJup and 10 6 Porb 6 100 days)with measured obliquity11 From this perspective K2-139 is anideal target to measure the sky-project spin-orbit angle via obser-vations of the Rossiter-McLaughlin (RM) effect Assuming spin-orbit alignement the expected amplitude of the RM anomaly is∆RV asymp
radic1 minus b2 (RpR)2 v sin iasymp 25 m sminus1 (Winn 2010) Given
the brightness of the host star (V = 11653 mag) this amplitudecan easily be measured using state-of-the-art spectrographs suchas HARPSESO-36m Moreover the transit duration (sim5 hours)is shorter than the visibility of K2-139 which is sim9 hours fromLa Silla observatory (altitude higher than 30 above the horizon)
Alternatively the spin-orbit angle could be measured fromthe analysis of the spot-crossing events as described in Sanchis-Ojeda et al (2011) and Sanchis-Ojeda et al (2012) Anomaliesascribable to the passage of K2-139b in front of stellar spots arevisible in the 3 transit light curves observed by K2 Unfortunatelythe limited number of transits and the K2 long cadence data donot allow us to perform a meaningful quantitative analysis of thespot-crossing events Given the amplitude of the detected anoma-lies (sim01) space-based high-precision photometry is needed todetect the spot-crossing events Observations performed with theupcoming CHaracterising ExOPlanets Satellite (CHEOPS Broeget al 2013) would allow us to photometrically determine the spin-orbit angle of this system
ACKNOWLEDGEMENTS
We warmly thank the NOT ESO TNG staff members for theirunique support during the observations We are very thankful toXavier Bonfils Franccedilois Bouchy Martin Kuumlrster Tsevi MazehJorge Melendez and Nuno Santos who kindly agreed to exchange
11 Source httpwww2mpsmpgdehomeshellercontentmain_HRMhtml as of January 2017
MNRAS 000 1ndash12 (2017)
10 O Barragaacuten et al
HARPSandFIES timewith us Special thanks go toAntoninoLanzafor assisting us with the calculation of the gyro-age of the star Wealso greatly thank the anonymous referee for herhis careful re-view and suggestions which helped us to improve the manuscriptD Gandolfi gratefully acknowledges the financial support of theProgramma Giovani Ricercatori ndash Rita Levi Montalcini ndash Rien-tro dei Cervelli (2012) awarded by the Italian Ministry of Edu-cation Universities and Research (MIUR) Sz Csizmadia thanksthe Hungarian OTKA Grant K113117 H J Deeg and D Nespralacknowledge support by grant ESP2015-65712-C5-4-R of the Span-ish Secretary of State for RampDampi (MINECO) D Lorenzo-Oliveiraacknowledges the support from FAPESP (201620667-8) This re-search was supported by the Ministerio de Economia y Competi-tividad under project FIS2012-31079 The research leading to theseresults has received funding from the European Union SeventhFramework Programme (FP72013-2016) under grant agreementNo 312430 (OPTICON) Based on observations obtained a) withthe Nordic Optical Telescope (NOT) operated on the island of LaPalma jointly by Denmark Finland Iceland Norway and Swedenin the Spanish Observatorio del Roque de los Muchachos (ORM)of the Instituto de Astrofiacutesica de Canarias (IAC) b) with the Ital-ian Telescopio Nazionale Galileo (TNG) also operated at the ORM(IAC) on the island of La Palma by the INAF - Fundacioacuten GalileoGalilei c) the 36m ESO telescope at La Silla Observatory underprogramme ID 097C-0948 The data presented here were obtainedin part with ALFOSC which is provided by the Instituto de As-trofisica de Andalucia (IAA) under a joint agreement with the Uni-versity of Copenhagen and NOTSA This paper includes data col-lected by the Kepler mission Funding for the Kepler mission is pro-vided by the NASA Science Mission directorate Some of the datapresented in this paper were obtained from theMikulski Archive forSpace Telescopes (MAST) STScI is operated by the Association ofUniversities for Research in Astronomy Inc under NASA contractNAS5-26555 Support for MAST for non-HST data is provided bythe NASA Office of Space Science via grant NNX09AF08G and byother grants and contracts MF and CMP acknowledge generoussupport from the Swedish National Space Board C Eiroa and IRebollido are supported by Spanish grant AYA2014-55840-P PDacknowledge the support from INAF and Ministero dellrsquoIstruzionedellrsquoUniversitagrave e della Ricerca (MIUR) in the form of the grantldquoPremiale VLT 2012rdquo and ldquoThe Chemical and Dynamical Evolu-tion of the Milky Way and Local Group Galaxiesrdquo This work hasmade use of data from the European Space Agency (ESA) missionGaia (httpwwwcosmosesaintgaia) processed by theGaia Data Processing and Analysis Consortium (DPAC httpwwwcosmosesaintwebgaiadpacconsortium) Fundingfor the DPAC has been provided by national institutions in particu-lar the institutions participating in theGaiaMultilateral Agreement
REFERENCES
Albrecht S et al 2012 ApJ 757 18Allard F Homeier D Freytag B 2011 in Johns-Krull C BrowningM K
West A A eds Astronomical Society of the Pacific Conference SeriesVol 448 16th Cambridge Workshop on Cool Stars Stellar Systemsand the Sun p 91 (arXiv10115405)
Antonini F Hamers A S Lithwick Y 2016 preprint(arXiv160401781)
Barnes S A 2010 ApJ 722 222Barnes S A Kim Y-C 2010 ApJ 721 675Barragaacuten O Gandolfi D 2017 Exotrending Astrophysics Source Code
Library (ascl1706001)
Barragaacuten O et al 2016 AJ 152 193Barragaacuten O Gandolfi D Antoniciello G 2017 pyaneti Astrophysics
Source Code Library (ascl1707003)Baruteau C et al 2014 Protostars and Planets VI pp 667ndash689Batygin K Bodenheimer P H Laughlin G P 2016 ApJ 829 114Boley A C Granados Contreras A P Gladman B 2016 ApJ 817 L17Boyajian T S et al 2013 ApJ 771 40Brahm R et al 2016 AJ 151 89Bressan A Marigo P Girardi L Salasnich B Dal Cero C Rubele S
Nanni A 2012 MNRAS 427 127Broeg C et al 2013 in European Physical Journal Web of Conferences p
03005 (arXiv13052270) doi101051epjconf20134703005Bruntt H et al 2010 MNRAS 405 1907Buchhave L A et al 2010 ApJ 720 1118Burnham K Anderson D 2002 Model Selection and Multimodel Infer-
ence A Practical Information-Theoretic Approach NewYork Springer-Verlag
Cabrera J et al 2009 AampA 506 501Cabrera J Csizmadia S Erikson A Rauer H Kirste S 2012 AampA 548
A44Cabrera J et al 2014 ApJ 781 18Cantat-Gaudin T et al 2014 AampA 562 A10Cardelli J A Clayton G C Mathis J S 1989 ApJ 345 245Carone L et al 2012 AampA 538 A112Carpano S et al 2009 AampA 506 491Castelli F Kurucz R L 2004 preprintCavarroc C et al 2012 ApampSS 337 511Chabrier G 2001 ApJ 554 1274Cosentino R et al 2012 in Ground-based and Airborne Instrumentation
for Astronomy IV p 84461V doi10111712925738Cutri R M et al 2003 2MASS All Sky Catalog of point sourcesCutri R M et al 2012 Technical report Explanatory Supplement to the
WISE All-Sky Data Release ProductsDawson R I Johnson J A Morton T D Crepp J R Fabrycky D C
Murray-Clay R A Howard A W 2012 ApJ 761 163Deeg H J et al 2010 Nature 464 384Demory B-O 2014 ApJ 789 L20Dong S Katz B Socrates A 2014 ApJ 781 L5Doyle A P Davies G R Smalley B Chaplin W J Elsworth Y 2014
MNRAS 444 3592Dumusque X Boisse I Santos N C 2014 ApJ 796 132Endl M Cochran W D 2016 PASP 128 094502Erikson A et al 2012 AampA 539 A14Fabricius C et al 2016 AampA 595 A3Fortney J J Marley M S Barnes J W 2007 ApJ 659 1661Fossati et al 2017 AampA submittedFrandsenS LindbergB 1999 inKarttunenH PiirolaV edsAstrophysics
with the NOT p 71Frewen S F N Hansen B M S 2016 MNRAS 455 1538Gandolfi D et al 2008 ApJ 687 1303Gandolfi D et al 2012 AampA 543 L5Gandolfi D et al 2015 AampA 576 A11Gray R O 1999 SPECTRUM A stellar spectral synthesis program As-
trophysics Source Code Library (ascl9910002)Grziwa S Paumltzold M 2016 preprint (arXiv160708417)Grziwa S Paumltzold M Carone L 2012 MNRAS 420 1045Hamers A S Antonini F Lithwick Y Perets H B Portegies Zwart S F
2016 preprint (arXiv160607438)Hatzes A P 2002 Astronomische Nachrichten 323 392Hatzes A P Rauer H 2015 ApJ 810 L25Heiter U et al 2015 Phys Scr 90 054010Huang C Wu Y Triaud A H M J 2016 ApJ 825 98Jenkins J S et al 2017 MNRAS 466 443Johnson J L Li H 2012 ApJ 751 81Kipping D M 2010 MNRAS 408 1758Kley W Nelson R P 2012 ARAampA 50 211Kovaacutecs G Zucker S Mazeh T 2002 AampA 391 369
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 11
Table 5 K2-139 system parameters
Parameter Prior(a) Final value
Stellar parametersStar mass M (M) middot middot middot 0919 plusmn 0033Star radius R (R) middot middot middot 0862 plusmn 0032Star density ρ (from spectroscopy g cmminus3) middot middot middot 202+025
minus022
Star density ρ (from light curve g cmminus3) middot middot middot 211+074minus081
Effective Temperature Teff (K) middot middot middot 5340 plusmn 110Surface gravity log g (cgs) middot middot middot 450 plusmn 009Iron abundance [FeH] (dex) middot middot middot 022 plusmn 008Microturbulent velocity vmic ( km sminus1) middot middot middot 09 plusmn 01Macroturbulent velocity vmac ( km sminus1) middot middot middot 25 plusmn 06Projected rotational velocity v sin i ( km sminus1) middot middot middot 28 plusmn 06Rotational period Prot (days) middot middot middot 1724 plusmn 012Activity index(b) logRprimeHK middot middot middot minus446 plusmn 006Gyrochronological age (Gyr) middot middot middot 18 plusmn 03Interstellar extinction AV (mag) middot middot middot 007 plusmn 005Star distance d (pc) middot middot middot 152 plusmn 10
Model parameters of K2-139 bOrbital period Porb (days) U[283773 283873] 2838236 plusmn 000026Transit epoch T0 (BJDTDBminus2 450 000) U[73258120 73258220] 732581714 plusmn 000033Scaled semi-major axis aR U[12 100] 448+47
minus67Planet-to-star radius ratio RpR U[0 02] 00961+00023
minus00015Impact parameter b U[0 12] 030+021
minus019radice sinω U[minus1 1](c) 010+029
minus030radice cosω U[minus1 1](c) 006+024
minus027Radial velocity semi-amplitude variation K ( m sminus1) U[0 200] 277+60
Note ndash The adopted Sun and Jupiter units follow the recommendations from the International Astronomical Union (Prša et al 2016) (a) U[a b] refersto uniform priors between a and b N[a b] means Gaussian priors with mean a and standard deviation b and F[a] to a fixed a value (b) Corrected forinterstellar reddening following Fossati et al (2017) The correction is +006 (c) The code always ensures that e lt 1 (d) Assuming albedo = 0
MNRAS 000 1ndash12 (2017)
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
MNRAS 000 1ndash12 (2017)
1 Introduction
2 K2 photometry
3 ALFOSC imaging
4 High-resolution spectroscopy
5 Stellar parameters
51 Spectral analysis
52 Interstellar extinction
53 Rotational period
54 Stellar mass radius and age
6 Joint RV-transit fit
7 Results and discussion
71 Stellar activity modeling
72 Additional companion
73 Spot-crossing events
74 Planets composition and formation scenario
8 Conclusions
Acknowledgements
The transiting warm Jupiter K2-139 b 9
01 05 1 2 3 4 5 10Mass (MJup)
06
08
10
12
14
Rad
ius
(RJu
p)
H 0 M core 10 M core 25 M core 50 M core100 M core
Figure 5 Warm Jupiters (black squares Mp gt 03 MJup and10 6 Porb 6 100 days) whose mass and radius have been estimated with aprecision of at least 25 (as of January 2017 exoplaneteu) K2-139 b isshown with a filled red circle The solid line corresponds to a planet with apure hydrogen composition (Seager et al 2007) The dashed lines representthe Fortney et al (2007) models for planet core masses of 0 10 25 50and 100 Moplus The vertical dotted line marks the giant planet lower limit asdefined by Hatzes amp Rauer (2015)
74 Planetrsquos composition and formation scenario
With amass of Mp = 0387+0083minus0075 MJ and radius of Rp = 0808+0034
minus0033RJ (resulting in a mean density of ρp=091+024
minus020 g cmminus3) K2-139 bjoins the small group of well characterized warm Jupiters Fig 5shows the position of K2-139 b in themass-radius diagram for warmJupiters (Mp gt 03 MJup 10 6 Porb 6 100 days) whose massand radius have been determined with a precision better than 25(14 objects) Notably K2-139 b is the transiting warm Jupiter withthe lowest mass known to date if the definition of giant planetsgiven by Hatzes amp Rauer (2015) is adopted Fig 5 displays also theplanetary models of Fortney et al (2007) for different core massesand age between 10 and 45Gyrs The planet radius of K2-139 bcan be explained if the planet has a core10 of 49+19
minus17 Moplus containingsim40 of the total planetary mass We expect that K2-139 b has asolid core surrounded by a gaseous envelope
Rafikov (2006) found that a core of mass 5 ndash 20 Moplus at a semi-major axis between 01 and 10AU would be able to start the run-away accretion phase to form a gas giant planet in situ Howeveraccording to his models these kind of cores are unlikely to formowing to the high irradiation coming from the star Boley et al(2016) suggested instead that more massive cores (Mcore amp 20Moplus)can be built up from the merging of tightly packed inner planetsformed at the early stages of the circumstellar disc Batygin et al(2016) found a similar result and argued that the massive core ofHD149026b (Mcore asymp 100Moplus) could be explained by one or moresuper-Earths which merged and accreted the surrounding gas toform a gas-giant planet Huang et al (2016) suggested that thesecores can initiate runaway accretion if they are formed in a re-gion with enough gas around them while those without enoughvolatiles remain super-Earths and represent the population of mas-sive rocky planets unveiled by Kepler around solar-like stars (egDemory 2014) Based on these studies and given the semi-majoraxis of 0179+0021
minus0027 AU the 48 plusmn 14 Moplus core of K2-139 b could
10 Calculated by interpolating Fortney et al (2007)rsquos models
have formed the planet in situ We note that the metallicity of K2-139 is relatively high ([FeH] = 021 plusmn 005) suggesting that theprimordial circumstellar disc had a relatively high content of dustwhich would have enhanced the formation of the core of K2-139 b(see eg Johnson amp Li 2012) Alternatively the planet might haveformed beyond the snow line and migrated inwards via planet-discinteraction (see eg Baruteau et al 2014)
8 CONCLUSIONS
We confirmed the planetary nature and derived the orbital and mainphysical parameters of K2-139 b a warm Jupiter (Teq = 565+48
minus32 K)transiting an active (log RprimeHK = minus446 plusmn 006) K0V star every 29days We measured a planetary mass of Mp = 0387+0083
minus0075 MJ andradius of Rp = 0808+0034
minus0033 RJ At a separation of ap = 0179+0021minus0027
AU the mean density of ρp = 091+024minus020 g cmminus3 implies that the
planet has a core of 49+19minus17 Moplus according to the evolutionary mod-
els of Fortney et al (2007) K2-139 b joins the small group ofwell-characterized warm Jupiters whose mass and radius have beendetermined with a precision better than 25
The spin-orbit angle ie the angle between the spin axis ofthe star and the angular momentum vector of the orbit can pro-vide us with valuable information on the migration mechanismsof exoplanets (see eg Winn 2010 Morton amp Johnson 2011 Al-brecht et al 2012 Gandolfi et al 2012) Currently there are only4 warm Jupiters (Mp gt 03 MJup and 10 6 Porb 6 100 days)with measured obliquity11 From this perspective K2-139 is anideal target to measure the sky-project spin-orbit angle via obser-vations of the Rossiter-McLaughlin (RM) effect Assuming spin-orbit alignement the expected amplitude of the RM anomaly is∆RV asymp
radic1 minus b2 (RpR)2 v sin iasymp 25 m sminus1 (Winn 2010) Given
the brightness of the host star (V = 11653 mag) this amplitudecan easily be measured using state-of-the-art spectrographs suchas HARPSESO-36m Moreover the transit duration (sim5 hours)is shorter than the visibility of K2-139 which is sim9 hours fromLa Silla observatory (altitude higher than 30 above the horizon)
Alternatively the spin-orbit angle could be measured fromthe analysis of the spot-crossing events as described in Sanchis-Ojeda et al (2011) and Sanchis-Ojeda et al (2012) Anomaliesascribable to the passage of K2-139b in front of stellar spots arevisible in the 3 transit light curves observed by K2 Unfortunatelythe limited number of transits and the K2 long cadence data donot allow us to perform a meaningful quantitative analysis of thespot-crossing events Given the amplitude of the detected anoma-lies (sim01) space-based high-precision photometry is needed todetect the spot-crossing events Observations performed with theupcoming CHaracterising ExOPlanets Satellite (CHEOPS Broeget al 2013) would allow us to photometrically determine the spin-orbit angle of this system
ACKNOWLEDGEMENTS
We warmly thank the NOT ESO TNG staff members for theirunique support during the observations We are very thankful toXavier Bonfils Franccedilois Bouchy Martin Kuumlrster Tsevi MazehJorge Melendez and Nuno Santos who kindly agreed to exchange
11 Source httpwww2mpsmpgdehomeshellercontentmain_HRMhtml as of January 2017
MNRAS 000 1ndash12 (2017)
10 O Barragaacuten et al
HARPSandFIES timewith us Special thanks go toAntoninoLanzafor assisting us with the calculation of the gyro-age of the star Wealso greatly thank the anonymous referee for herhis careful re-view and suggestions which helped us to improve the manuscriptD Gandolfi gratefully acknowledges the financial support of theProgramma Giovani Ricercatori ndash Rita Levi Montalcini ndash Rien-tro dei Cervelli (2012) awarded by the Italian Ministry of Edu-cation Universities and Research (MIUR) Sz Csizmadia thanksthe Hungarian OTKA Grant K113117 H J Deeg and D Nespralacknowledge support by grant ESP2015-65712-C5-4-R of the Span-ish Secretary of State for RampDampi (MINECO) D Lorenzo-Oliveiraacknowledges the support from FAPESP (201620667-8) This re-search was supported by the Ministerio de Economia y Competi-tividad under project FIS2012-31079 The research leading to theseresults has received funding from the European Union SeventhFramework Programme (FP72013-2016) under grant agreementNo 312430 (OPTICON) Based on observations obtained a) withthe Nordic Optical Telescope (NOT) operated on the island of LaPalma jointly by Denmark Finland Iceland Norway and Swedenin the Spanish Observatorio del Roque de los Muchachos (ORM)of the Instituto de Astrofiacutesica de Canarias (IAC) b) with the Ital-ian Telescopio Nazionale Galileo (TNG) also operated at the ORM(IAC) on the island of La Palma by the INAF - Fundacioacuten GalileoGalilei c) the 36m ESO telescope at La Silla Observatory underprogramme ID 097C-0948 The data presented here were obtainedin part with ALFOSC which is provided by the Instituto de As-trofisica de Andalucia (IAA) under a joint agreement with the Uni-versity of Copenhagen and NOTSA This paper includes data col-lected by the Kepler mission Funding for the Kepler mission is pro-vided by the NASA Science Mission directorate Some of the datapresented in this paper were obtained from theMikulski Archive forSpace Telescopes (MAST) STScI is operated by the Association ofUniversities for Research in Astronomy Inc under NASA contractNAS5-26555 Support for MAST for non-HST data is provided bythe NASA Office of Space Science via grant NNX09AF08G and byother grants and contracts MF and CMP acknowledge generoussupport from the Swedish National Space Board C Eiroa and IRebollido are supported by Spanish grant AYA2014-55840-P PDacknowledge the support from INAF and Ministero dellrsquoIstruzionedellrsquoUniversitagrave e della Ricerca (MIUR) in the form of the grantldquoPremiale VLT 2012rdquo and ldquoThe Chemical and Dynamical Evolu-tion of the Milky Way and Local Group Galaxiesrdquo This work hasmade use of data from the European Space Agency (ESA) missionGaia (httpwwwcosmosesaintgaia) processed by theGaia Data Processing and Analysis Consortium (DPAC httpwwwcosmosesaintwebgaiadpacconsortium) Fundingfor the DPAC has been provided by national institutions in particu-lar the institutions participating in theGaiaMultilateral Agreement
REFERENCES
Albrecht S et al 2012 ApJ 757 18Allard F Homeier D Freytag B 2011 in Johns-Krull C BrowningM K
West A A eds Astronomical Society of the Pacific Conference SeriesVol 448 16th Cambridge Workshop on Cool Stars Stellar Systemsand the Sun p 91 (arXiv10115405)
Antonini F Hamers A S Lithwick Y 2016 preprint(arXiv160401781)
Barnes S A 2010 ApJ 722 222Barnes S A Kim Y-C 2010 ApJ 721 675Barragaacuten O Gandolfi D 2017 Exotrending Astrophysics Source Code
Library (ascl1706001)
Barragaacuten O et al 2016 AJ 152 193Barragaacuten O Gandolfi D Antoniciello G 2017 pyaneti Astrophysics
Source Code Library (ascl1707003)Baruteau C et al 2014 Protostars and Planets VI pp 667ndash689Batygin K Bodenheimer P H Laughlin G P 2016 ApJ 829 114Boley A C Granados Contreras A P Gladman B 2016 ApJ 817 L17Boyajian T S et al 2013 ApJ 771 40Brahm R et al 2016 AJ 151 89Bressan A Marigo P Girardi L Salasnich B Dal Cero C Rubele S
Nanni A 2012 MNRAS 427 127Broeg C et al 2013 in European Physical Journal Web of Conferences p
03005 (arXiv13052270) doi101051epjconf20134703005Bruntt H et al 2010 MNRAS 405 1907Buchhave L A et al 2010 ApJ 720 1118Burnham K Anderson D 2002 Model Selection and Multimodel Infer-
ence A Practical Information-Theoretic Approach NewYork Springer-Verlag
Cabrera J et al 2009 AampA 506 501Cabrera J Csizmadia S Erikson A Rauer H Kirste S 2012 AampA 548
A44Cabrera J et al 2014 ApJ 781 18Cantat-Gaudin T et al 2014 AampA 562 A10Cardelli J A Clayton G C Mathis J S 1989 ApJ 345 245Carone L et al 2012 AampA 538 A112Carpano S et al 2009 AampA 506 491Castelli F Kurucz R L 2004 preprintCavarroc C et al 2012 ApampSS 337 511Chabrier G 2001 ApJ 554 1274Cosentino R et al 2012 in Ground-based and Airborne Instrumentation
for Astronomy IV p 84461V doi10111712925738Cutri R M et al 2003 2MASS All Sky Catalog of point sourcesCutri R M et al 2012 Technical report Explanatory Supplement to the
WISE All-Sky Data Release ProductsDawson R I Johnson J A Morton T D Crepp J R Fabrycky D C
Murray-Clay R A Howard A W 2012 ApJ 761 163Deeg H J et al 2010 Nature 464 384Demory B-O 2014 ApJ 789 L20Dong S Katz B Socrates A 2014 ApJ 781 L5Doyle A P Davies G R Smalley B Chaplin W J Elsworth Y 2014
MNRAS 444 3592Dumusque X Boisse I Santos N C 2014 ApJ 796 132Endl M Cochran W D 2016 PASP 128 094502Erikson A et al 2012 AampA 539 A14Fabricius C et al 2016 AampA 595 A3Fortney J J Marley M S Barnes J W 2007 ApJ 659 1661Fossati et al 2017 AampA submittedFrandsenS LindbergB 1999 inKarttunenH PiirolaV edsAstrophysics
with the NOT p 71Frewen S F N Hansen B M S 2016 MNRAS 455 1538Gandolfi D et al 2008 ApJ 687 1303Gandolfi D et al 2012 AampA 543 L5Gandolfi D et al 2015 AampA 576 A11Gray R O 1999 SPECTRUM A stellar spectral synthesis program As-
trophysics Source Code Library (ascl9910002)Grziwa S Paumltzold M 2016 preprint (arXiv160708417)Grziwa S Paumltzold M Carone L 2012 MNRAS 420 1045Hamers A S Antonini F Lithwick Y Perets H B Portegies Zwart S F
2016 preprint (arXiv160607438)Hatzes A P 2002 Astronomische Nachrichten 323 392Hatzes A P Rauer H 2015 ApJ 810 L25Heiter U et al 2015 Phys Scr 90 054010Huang C Wu Y Triaud A H M J 2016 ApJ 825 98Jenkins J S et al 2017 MNRAS 466 443Johnson J L Li H 2012 ApJ 751 81Kipping D M 2010 MNRAS 408 1758Kley W Nelson R P 2012 ARAampA 50 211Kovaacutecs G Zucker S Mazeh T 2002 AampA 391 369
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 11
Table 5 K2-139 system parameters
Parameter Prior(a) Final value
Stellar parametersStar mass M (M) middot middot middot 0919 plusmn 0033Star radius R (R) middot middot middot 0862 plusmn 0032Star density ρ (from spectroscopy g cmminus3) middot middot middot 202+025
minus022
Star density ρ (from light curve g cmminus3) middot middot middot 211+074minus081
Effective Temperature Teff (K) middot middot middot 5340 plusmn 110Surface gravity log g (cgs) middot middot middot 450 plusmn 009Iron abundance [FeH] (dex) middot middot middot 022 plusmn 008Microturbulent velocity vmic ( km sminus1) middot middot middot 09 plusmn 01Macroturbulent velocity vmac ( km sminus1) middot middot middot 25 plusmn 06Projected rotational velocity v sin i ( km sminus1) middot middot middot 28 plusmn 06Rotational period Prot (days) middot middot middot 1724 plusmn 012Activity index(b) logRprimeHK middot middot middot minus446 plusmn 006Gyrochronological age (Gyr) middot middot middot 18 plusmn 03Interstellar extinction AV (mag) middot middot middot 007 plusmn 005Star distance d (pc) middot middot middot 152 plusmn 10
Model parameters of K2-139 bOrbital period Porb (days) U[283773 283873] 2838236 plusmn 000026Transit epoch T0 (BJDTDBminus2 450 000) U[73258120 73258220] 732581714 plusmn 000033Scaled semi-major axis aR U[12 100] 448+47
minus67Planet-to-star radius ratio RpR U[0 02] 00961+00023
minus00015Impact parameter b U[0 12] 030+021
minus019radice sinω U[minus1 1](c) 010+029
minus030radice cosω U[minus1 1](c) 006+024
minus027Radial velocity semi-amplitude variation K ( m sminus1) U[0 200] 277+60
Note ndash The adopted Sun and Jupiter units follow the recommendations from the International Astronomical Union (Prša et al 2016) (a) U[a b] refersto uniform priors between a and b N[a b] means Gaussian priors with mean a and standard deviation b and F[a] to a fixed a value (b) Corrected forinterstellar reddening following Fossati et al (2017) The correction is +006 (c) The code always ensures that e lt 1 (d) Assuming albedo = 0
MNRAS 000 1ndash12 (2017)
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
MNRAS 000 1ndash12 (2017)
1 Introduction
2 K2 photometry
3 ALFOSC imaging
4 High-resolution spectroscopy
5 Stellar parameters
51 Spectral analysis
52 Interstellar extinction
53 Rotational period
54 Stellar mass radius and age
6 Joint RV-transit fit
7 Results and discussion
71 Stellar activity modeling
72 Additional companion
73 Spot-crossing events
74 Planets composition and formation scenario
8 Conclusions
Acknowledgements
10 O Barragaacuten et al
HARPSandFIES timewith us Special thanks go toAntoninoLanzafor assisting us with the calculation of the gyro-age of the star Wealso greatly thank the anonymous referee for herhis careful re-view and suggestions which helped us to improve the manuscriptD Gandolfi gratefully acknowledges the financial support of theProgramma Giovani Ricercatori ndash Rita Levi Montalcini ndash Rien-tro dei Cervelli (2012) awarded by the Italian Ministry of Edu-cation Universities and Research (MIUR) Sz Csizmadia thanksthe Hungarian OTKA Grant K113117 H J Deeg and D Nespralacknowledge support by grant ESP2015-65712-C5-4-R of the Span-ish Secretary of State for RampDampi (MINECO) D Lorenzo-Oliveiraacknowledges the support from FAPESP (201620667-8) This re-search was supported by the Ministerio de Economia y Competi-tividad under project FIS2012-31079 The research leading to theseresults has received funding from the European Union SeventhFramework Programme (FP72013-2016) under grant agreementNo 312430 (OPTICON) Based on observations obtained a) withthe Nordic Optical Telescope (NOT) operated on the island of LaPalma jointly by Denmark Finland Iceland Norway and Swedenin the Spanish Observatorio del Roque de los Muchachos (ORM)of the Instituto de Astrofiacutesica de Canarias (IAC) b) with the Ital-ian Telescopio Nazionale Galileo (TNG) also operated at the ORM(IAC) on the island of La Palma by the INAF - Fundacioacuten GalileoGalilei c) the 36m ESO telescope at La Silla Observatory underprogramme ID 097C-0948 The data presented here were obtainedin part with ALFOSC which is provided by the Instituto de As-trofisica de Andalucia (IAA) under a joint agreement with the Uni-versity of Copenhagen and NOTSA This paper includes data col-lected by the Kepler mission Funding for the Kepler mission is pro-vided by the NASA Science Mission directorate Some of the datapresented in this paper were obtained from theMikulski Archive forSpace Telescopes (MAST) STScI is operated by the Association ofUniversities for Research in Astronomy Inc under NASA contractNAS5-26555 Support for MAST for non-HST data is provided bythe NASA Office of Space Science via grant NNX09AF08G and byother grants and contracts MF and CMP acknowledge generoussupport from the Swedish National Space Board C Eiroa and IRebollido are supported by Spanish grant AYA2014-55840-P PDacknowledge the support from INAF and Ministero dellrsquoIstruzionedellrsquoUniversitagrave e della Ricerca (MIUR) in the form of the grantldquoPremiale VLT 2012rdquo and ldquoThe Chemical and Dynamical Evolu-tion of the Milky Way and Local Group Galaxiesrdquo This work hasmade use of data from the European Space Agency (ESA) missionGaia (httpwwwcosmosesaintgaia) processed by theGaia Data Processing and Analysis Consortium (DPAC httpwwwcosmosesaintwebgaiadpacconsortium) Fundingfor the DPAC has been provided by national institutions in particu-lar the institutions participating in theGaiaMultilateral Agreement
REFERENCES
Albrecht S et al 2012 ApJ 757 18Allard F Homeier D Freytag B 2011 in Johns-Krull C BrowningM K
West A A eds Astronomical Society of the Pacific Conference SeriesVol 448 16th Cambridge Workshop on Cool Stars Stellar Systemsand the Sun p 91 (arXiv10115405)
Antonini F Hamers A S Lithwick Y 2016 preprint(arXiv160401781)
Barnes S A 2010 ApJ 722 222Barnes S A Kim Y-C 2010 ApJ 721 675Barragaacuten O Gandolfi D 2017 Exotrending Astrophysics Source Code
Library (ascl1706001)
Barragaacuten O et al 2016 AJ 152 193Barragaacuten O Gandolfi D Antoniciello G 2017 pyaneti Astrophysics
Source Code Library (ascl1707003)Baruteau C et al 2014 Protostars and Planets VI pp 667ndash689Batygin K Bodenheimer P H Laughlin G P 2016 ApJ 829 114Boley A C Granados Contreras A P Gladman B 2016 ApJ 817 L17Boyajian T S et al 2013 ApJ 771 40Brahm R et al 2016 AJ 151 89Bressan A Marigo P Girardi L Salasnich B Dal Cero C Rubele S
Nanni A 2012 MNRAS 427 127Broeg C et al 2013 in European Physical Journal Web of Conferences p
03005 (arXiv13052270) doi101051epjconf20134703005Bruntt H et al 2010 MNRAS 405 1907Buchhave L A et al 2010 ApJ 720 1118Burnham K Anderson D 2002 Model Selection and Multimodel Infer-
ence A Practical Information-Theoretic Approach NewYork Springer-Verlag
Cabrera J et al 2009 AampA 506 501Cabrera J Csizmadia S Erikson A Rauer H Kirste S 2012 AampA 548
A44Cabrera J et al 2014 ApJ 781 18Cantat-Gaudin T et al 2014 AampA 562 A10Cardelli J A Clayton G C Mathis J S 1989 ApJ 345 245Carone L et al 2012 AampA 538 A112Carpano S et al 2009 AampA 506 491Castelli F Kurucz R L 2004 preprintCavarroc C et al 2012 ApampSS 337 511Chabrier G 2001 ApJ 554 1274Cosentino R et al 2012 in Ground-based and Airborne Instrumentation
for Astronomy IV p 84461V doi10111712925738Cutri R M et al 2003 2MASS All Sky Catalog of point sourcesCutri R M et al 2012 Technical report Explanatory Supplement to the
WISE All-Sky Data Release ProductsDawson R I Johnson J A Morton T D Crepp J R Fabrycky D C
Murray-Clay R A Howard A W 2012 ApJ 761 163Deeg H J et al 2010 Nature 464 384Demory B-O 2014 ApJ 789 L20Dong S Katz B Socrates A 2014 ApJ 781 L5Doyle A P Davies G R Smalley B Chaplin W J Elsworth Y 2014
MNRAS 444 3592Dumusque X Boisse I Santos N C 2014 ApJ 796 132Endl M Cochran W D 2016 PASP 128 094502Erikson A et al 2012 AampA 539 A14Fabricius C et al 2016 AampA 595 A3Fortney J J Marley M S Barnes J W 2007 ApJ 659 1661Fossati et al 2017 AampA submittedFrandsenS LindbergB 1999 inKarttunenH PiirolaV edsAstrophysics
with the NOT p 71Frewen S F N Hansen B M S 2016 MNRAS 455 1538Gandolfi D et al 2008 ApJ 687 1303Gandolfi D et al 2012 AampA 543 L5Gandolfi D et al 2015 AampA 576 A11Gray R O 1999 SPECTRUM A stellar spectral synthesis program As-
trophysics Source Code Library (ascl9910002)Grziwa S Paumltzold M 2016 preprint (arXiv160708417)Grziwa S Paumltzold M Carone L 2012 MNRAS 420 1045Hamers A S Antonini F Lithwick Y Perets H B Portegies Zwart S F
2016 preprint (arXiv160607438)Hatzes A P 2002 Astronomische Nachrichten 323 392Hatzes A P Rauer H 2015 ApJ 810 L25Heiter U et al 2015 Phys Scr 90 054010Huang C Wu Y Triaud A H M J 2016 ApJ 825 98Jenkins J S et al 2017 MNRAS 466 443Johnson J L Li H 2012 ApJ 751 81Kipping D M 2010 MNRAS 408 1758Kley W Nelson R P 2012 ARAampA 50 211Kovaacutecs G Zucker S Mazeh T 2002 AampA 391 369
MNRAS 000 1ndash12 (2017)
The transiting warm Jupiter K2-139 b 11
Table 5 K2-139 system parameters
Parameter Prior(a) Final value
Stellar parametersStar mass M (M) middot middot middot 0919 plusmn 0033Star radius R (R) middot middot middot 0862 plusmn 0032Star density ρ (from spectroscopy g cmminus3) middot middot middot 202+025
minus022
Star density ρ (from light curve g cmminus3) middot middot middot 211+074minus081
Effective Temperature Teff (K) middot middot middot 5340 plusmn 110Surface gravity log g (cgs) middot middot middot 450 plusmn 009Iron abundance [FeH] (dex) middot middot middot 022 plusmn 008Microturbulent velocity vmic ( km sminus1) middot middot middot 09 plusmn 01Macroturbulent velocity vmac ( km sminus1) middot middot middot 25 plusmn 06Projected rotational velocity v sin i ( km sminus1) middot middot middot 28 plusmn 06Rotational period Prot (days) middot middot middot 1724 plusmn 012Activity index(b) logRprimeHK middot middot middot minus446 plusmn 006Gyrochronological age (Gyr) middot middot middot 18 plusmn 03Interstellar extinction AV (mag) middot middot middot 007 plusmn 005Star distance d (pc) middot middot middot 152 plusmn 10
Model parameters of K2-139 bOrbital period Porb (days) U[283773 283873] 2838236 plusmn 000026Transit epoch T0 (BJDTDBminus2 450 000) U[73258120 73258220] 732581714 plusmn 000033Scaled semi-major axis aR U[12 100] 448+47
minus67Planet-to-star radius ratio RpR U[0 02] 00961+00023
minus00015Impact parameter b U[0 12] 030+021
minus019radice sinω U[minus1 1](c) 010+029
minus030radice cosω U[minus1 1](c) 006+024
minus027Radial velocity semi-amplitude variation K ( m sminus1) U[0 200] 277+60
Note ndash The adopted Sun and Jupiter units follow the recommendations from the International Astronomical Union (Prša et al 2016) (a) U[a b] refersto uniform priors between a and b N[a b] means Gaussian priors with mean a and standard deviation b and F[a] to a fixed a value (b) Corrected forinterstellar reddening following Fossati et al (2017) The correction is +006 (c) The code always ensures that e lt 1 (d) Assuming albedo = 0
MNRAS 000 1ndash12 (2017)
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
MNRAS 000 1ndash12 (2017)
1 Introduction
2 K2 photometry
3 ALFOSC imaging
4 High-resolution spectroscopy
5 Stellar parameters
51 Spectral analysis
52 Interstellar extinction
53 Rotational period
54 Stellar mass radius and age
6 Joint RV-transit fit
7 Results and discussion
71 Stellar activity modeling
72 Additional companion
73 Spot-crossing events
74 Planets composition and formation scenario
8 Conclusions
Acknowledgements
The transiting warm Jupiter K2-139 b 11
Table 5 K2-139 system parameters
Parameter Prior(a) Final value
Stellar parametersStar mass M (M) middot middot middot 0919 plusmn 0033Star radius R (R) middot middot middot 0862 plusmn 0032Star density ρ (from spectroscopy g cmminus3) middot middot middot 202+025
minus022
Star density ρ (from light curve g cmminus3) middot middot middot 211+074minus081
Effective Temperature Teff (K) middot middot middot 5340 plusmn 110Surface gravity log g (cgs) middot middot middot 450 plusmn 009Iron abundance [FeH] (dex) middot middot middot 022 plusmn 008Microturbulent velocity vmic ( km sminus1) middot middot middot 09 plusmn 01Macroturbulent velocity vmac ( km sminus1) middot middot middot 25 plusmn 06Projected rotational velocity v sin i ( km sminus1) middot middot middot 28 plusmn 06Rotational period Prot (days) middot middot middot 1724 plusmn 012Activity index(b) logRprimeHK middot middot middot minus446 plusmn 006Gyrochronological age (Gyr) middot middot middot 18 plusmn 03Interstellar extinction AV (mag) middot middot middot 007 plusmn 005Star distance d (pc) middot middot middot 152 plusmn 10
Model parameters of K2-139 bOrbital period Porb (days) U[283773 283873] 2838236 plusmn 000026Transit epoch T0 (BJDTDBminus2 450 000) U[73258120 73258220] 732581714 plusmn 000033Scaled semi-major axis aR U[12 100] 448+47
minus67Planet-to-star radius ratio RpR U[0 02] 00961+00023
minus00015Impact parameter b U[0 12] 030+021
minus019radice sinω U[minus1 1](c) 010+029
minus030radice cosω U[minus1 1](c) 006+024
minus027Radial velocity semi-amplitude variation K ( m sminus1) U[0 200] 277+60
Note ndash The adopted Sun and Jupiter units follow the recommendations from the International Astronomical Union (Prša et al 2016) (a) U[a b] refersto uniform priors between a and b N[a b] means Gaussian priors with mean a and standard deviation b and F[a] to a fixed a value (b) Corrected forinterstellar reddening following Fossati et al (2017) The correction is +006 (c) The code always ensures that e lt 1 (d) Assuming albedo = 0
MNRAS 000 1ndash12 (2017)
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
MNRAS 000 1ndash12 (2017)
1 Introduction
2 K2 photometry
3 ALFOSC imaging
4 High-resolution spectroscopy
5 Stellar parameters
51 Spectral analysis
52 Interstellar extinction
53 Rotational period
54 Stellar mass radius and age
6 Joint RV-transit fit
7 Results and discussion
71 Stellar activity modeling
72 Additional companion
73 Spot-crossing events
74 Planets composition and formation scenario
8 Conclusions
Acknowledgements
12 O Barragaacuten et al
Kuerster M Schmitt J H M M Cutispoto G Dennerl K 1997 AampA320 831
Kurucz R L 2013 ATLAS12 Opacity sampling model atmosphere pro-gram Astrophysics Source Code Library (ascl1303024)
Lenz P Breger M 2005 Communications in Asteroseismology 146 53Luger R Agol E Kruse E Barnes R Becker A Foreman-Mackey D
Deming D 2016 AJ 152 100Magrini L et al 2013 AampA 558 A38Mandel K Agol E 2002 ApJ 580 L171Mayor M et al 2003 The Messenger 114 20McQuillan A Mazeh T Aigrain S 2014 ApJS 211 24Morton T D Johnson J A 2011 ApJ 729 138Niedzielski A et al 2016 preprint (arXiv160307581)Ortiz M et al 2015 AampA 573 L6Pepe F et al 2013 Nature 503 377Petrovich C Tremaine S 2016 preprint (arXiv160400010)Prša A et al 2016 AJ 152 41Rafikov R R 2006 ApJ 648 666Ryabchikova T A Pakhomov Y V Piskunov N E 2011 Kazan Izdatel
Kazanskogo Universiteta 153 61Saad-Olivera X Nesvornyacute D Kipping D M Roig F 2017 AJ 153 198Saar S H Donahue R A 1997 ApJ 485 319Sanchis-Ojeda R Winn J N 2011 ApJ 743 61Sanchis-Ojeda RWinn J N HolmanM J Carter J A Osip D J Fuentes
C I 2011 ApJ 733 127Sanchis-Ojeda R et al 2012 Nature 487 449Schlegel D J Finkbeiner D P Davis M 1998 ApJ 500 525Seager S Kuchner M Hier-Majumder C A Militzer B 2007 ApJ 669
1279Smith A M S et al 2017 MNRAS 464 2708Sneden C Bean J Ivans I Lucatello S Sobeck J 2012 MOOG LTE
line analysis and spectrum synthesis Astrophysics Source Code Library(ascl1202009)
Stetson P B Pancino E 2008 PASP 120 1332Telting J H et al 2014 Astronomische Nachrichten 335 41Tull R G MacQueen P J Sneden C Lambert D L 1995 PASP 107
251Valenti J A Fischer D A 2005 ApJS 159 141Valenti J A Piskunov N 1996 AampAS 118 595Vanderburg A Johnson J A 2014 PASP 126 948Winn JN 2010 Exoplanet Transits andOccultationsUniversity ofArizona
Press pp 55ndash77Zechmeister M Kuumlrster M 2009 AampA 496 577da Silva R et al 2007 AampA 473 323
This paper has been typeset from a TEXLATEX file prepared by the author
