Discovery of a young low-mass brown dwarf transiting a fast-rotating F-type star by the Galactic Plane eXoplanet (GPX) survey
P. Benni, A. Y. Burdanov, V. V. Krushinsky, A. Bonfanti, G. Hébrard, J. M. Almenara, S. Dalal, O. D. S. Demangeon, M. Tsantaki, J. Pepper, K. G. Stassun, A. Vanderburg, A. Belinski, F. Kashaev, K. Barkaoui, T. Kim, W. Kang, K. Antonyuk, V. V. Dyachenko, D. A. Rastegaev, A. Beskakotov, A. A. Mitrofanova, F. J. Pozuelos, E. D. Kuznetsov, A. Popov, F. Kiefer, P. A. Wilson, G. Ricker, R. Vanderspek, D. W. Latham, S. Seager, J. M. Jenkins, E. Sokov, I. Sokova, A. Marchini, R. Papini, F. Salvaggio, M. Banfi, Ö. Baştürk, Ş. Torun, S. Yalçınkaya, K. Ivanov, G. Valyavin, E. Jehin, M. Gillon, E. Pakštienė, V.-P. Hentunen, S. Shadick, M. Bretton, A. Wünsche, J. Garlitz, Y. Jongen, D. Molina, E. Girardin, F. Grau Horta, R. Naves, Z. Benkhaldoun, M. D. Joner, M. Spencer, A. Bieryla, D. J. Stevens, E. L. N. Jensen, K. A. Collins, D. Charbonneau, E. V. Quintana, S. E. Mullally, C. E. Henze
MMNRAS , 1–13 (2020) Preprint 28 September 2020 Compiled using MNRAS L A TEX style file v3.0
Discovery of a young low-mass brown dwarf transiting afast-rotating F-type star by the Galactic Plane eXoplanet(GPX) survey
P. Benni (cid:63) , A. Y. Burdanov , (cid:63) , V. V. Krushinsky , A. Bonfanti , , G. H´ebrard , ,J. M. Almenara , S. Dalal , O. D. S. Demangeon , M. Tsantaki , J. Pepper ,K. G. Stassun , A. Vanderburg , , A. Belinski , F. Kashaev , K. Barkaoui , ,T. Kim , W. Kang , K. Antonyuk , V. V. Dyachenko , D. A. Rastegaev ,A. Beskakotov , , A. A. Mitrofanova , F. J. Pozuelos , , A. Popov , F. Kiefer ,P. A. Wilson , , G. Ricker , R. Vanderspek , D. W. Latham , S. Seager , , ,J. M. Jenkins , E. Sokov , , I. Sokova , , A. Marchini , R. Papini ,F. Salvaggio , M. Banfi , ¨O. Ba¸st¨urk , ¸S. Torun , S. Yal¸cınkaya , K. Ivanov ,G. Valyavin , , , E. Jehin , M. Gillon , E. Pakˇstien ˙e , V.-P. Hentunen ,S. Shadick , M. Bretton , A. W¨unsche , J. Garlitz , Y. Jongen , D. Molina ,E. Girardin , F. Grau Horta , R. Naves , Z. Benkhaldoun , M. D. Joner ,M. Spencer , A. Bieryla , D. J. Stevens , , , E. L. N. Jensen , K. A. Collins ,D. Charbonneau , E. V. Quintana , S. E. Mullally , C. E. Henze Affiliations can be found after the references. Submitted to MNRAS on 4 July 2020
ABSTRACT
We announce the discovery of GPX-1 b, a transiting brown dwarf with a mass of . ± . M Jup and a radius of . ± . R Jup , the first sub-stellar object discoveredby the Galactic Plane eXoplanet (GPX) survey. The brown dwarf transits a mod-erately bright ( V = 12.3 mag) fast-rotating F-type star with a projected rotationalvelocity v sin i ∗ ∼
40 km/s, effective temperature ± K, mass . ± . M (cid:12) ,radius . ± . R (cid:12) and approximate age . + . − . Gyr. GPX-1 b has an orbital pe-riod of . ± . d, mid-transit time T = . ± . TDB and a transit depth of . ± . %. We describe the GPX transit detection observa-tions, subsequent photometric and speckle-interferometric follow-up observations, andSOPHIE spectroscopic measurements, which allowed us to establish the presence of asub-stellar object around the host star. GPX-1 was observed at 30-min integrations byTESS in Sector 18, but the data is affected by blending with a 3.4 mag brighter star42 arcsec away. GPX-1 b is one of about two dozen transiting brown dwarfs known todate, with a mass close to the theoretical brown dwarf/gas giant planet mass transi-tion boundary. Since GPX-1 is a moderately bright and fast-rotating star, it can befollowed-up by the means of Doppler tomography. Key words: stars: brown dwarfs – stars: rotation – stars: individual: GPX-1 (cid:63)
E-mail: [email protected] (PB); [email protected](AB)
Brown dwarfs (BDs) are sub-stellar objects with masses inthe range of ∼ M Jup . Objects within these mass lim-its are below the hydrogen-burning minimum mass of 0.07-0.08 M (cid:12) ( ∼ M Jup ) and cannot sustain thermonuclear fu- © a r X i v : . [ a s t r o - ph . S R ] S e p Benni et al. sion of hydrogen or helium (Kumar 1963; Hayashi & Nakano1963), but reactions with deuterium and lithium are possi-ble (Burrows et al. 1997; Chabrier et al. 2000, 2014). Theexact mass boundaries depend on the chemical compositionof a sub-stellar object (Baraffe et al. 2002; Spiegel et al.2011). The formation processes of BDs are still debated, butsub-stellar objects above ∼ M Jup are generally consideredBDs regardless of formation mechanism or current location(Boss et al. 2007). It is still not well understood whetherBDs are a product of protostellar cloud fragmentation orif they are formed in protoplanetary discs around youngstars. Current statistics of the observed properties of BDsindicates that two distinct populations exist: massive BDs( (cid:38) M Jup ), which are formed similar to binary stars, andlow-mass BDs, which are formed as planets (e.g., Ma & Ge2014; Wilson et al. 2016). However, this division of BDs intotwo populations is being challenged with the new discover-ies of intermediate-mass BDs (Carmichael et al. 2019, 2020;Jackman et al. 2019; ˇSubjak et al. 2020).The relatively large masses and radii of BDs (comparedto most planets) should make them readily detectable in ra-dial velocity (RV) and photometric surveys. However, mostknown BDs have been found as isolated objects, and farfewer BDs are known to be in multiple systems (Sahlmannet al. 2011; Kiefer et al. 2019). Transiting BDs are evenrarer, but these systems provide a unique opportunity toprobe the properties of these objects by making possible themeasurement of their radii, masses, and obliquities, whichcan give initial glimpses into their dynamical history (Zhouet al. 2019). The lack of detections of BDs with orbital peri-ods shorter than 10 yr around main-sequence stars has beenknown as the brown dwarf desert (Marcy & Butler 2000;Halbwachs et al. 2000; Grether & Lineweaver 2006). It isassociated with the different formation mechanisms of low-and high-mass BDs and with instability of their orbits, wheretidal interaction causes the loss of angular momentum andorbit shrinking until the engulfment of a BD by its hoststar (Armitage & Bonnell 2002; P¨atzold & Rauer 2002). ForBDs orbiting late-type stars with a deep convective layer,magnetic braking should speed up the merging. However,more and more detections ”populate” the desert (e.g., Wil-son et al. 2016; Kiefer et al. 2019), while BDs with orbitalperiods shorter than 100 d still remain quite rare. A detailedunderstanding of the formation and evolution of BDs orbit-ing stars is still difficult due to the lack of statistical data.Exoplanet RV surveys keep expanding the current pop-ulation of BDs and exoplanet transit surveys have dis-covered most of the known transiting BDs (see the Exo-planet.eu database; Schneider et al. 2011). Despite the factthat roughly half of the known transiting BDs were discov-ered with the use of space telescopes, ground-based exo-planet surveys detected a number of the BDs known to tran-sit relatively bright host stars in the V ≤ range: KELT-1(Siverd et al. 2012), HATS-70 (Zhou et al. 2019), WASP-30 (Anderson et al. 2011), and WASP-128 (Hodˇzi´c et al.2018). Such systems allow more in-depth follow-up studiescompared to systems with fainter host stars.The ground-based transit surveys, which discovered the above-mentioned BDs transiting bright stars, have observeda substantial portion of the sky in an attempt to find newtransiting planets. However, most of these surveys do notobserve very dense parts of the Galactic plane in orderto avoid problems associated with blending of the stars.This is especially true for wide-field surveys like WASPthat had poor spatial resolution (the WASP image scaleis 13.7 arcsec pixel − ). Blending complicates the detection oftransit signals and can significantly increase the rate of false-positive exoplanet candidates. The Kepler/K2 space mis-sions (Borucki et al. 2010; Howell et al. 2014) had better spa-tial resolution (4 arcsec pixel − ) than most ground-based ex-oplanet transit surveys, and brought a significant contribu-tion to the known sample by discovering 40 hot Jupiters and6 BDs (Schneider et al. 2011), but Kepler/K2 was limitedby the fact that it observed only some parts of the sky. TheTransiting Exoplanet Survey Satellite survey (TESS; Rickeret al. 2014) that has been operating since 2018 will observealmost the whole sky, but it also has quite low spatial reso-lution (20.25 arcsec pixel − ). The future PLAnetary Transitsand Oscillation of stars mission (PLATO; Rauer et al. 2014)has a resolution of 15 arcsec pixel − and it will perform long-duration observations of only two sky fields. OGLE (Opti-cal Gravitational Lens Experiment; Udalski 2003) is a largesouthern sky variability survey, which observes the Galac-tic Bulge and Disk with an image scale of 0.26 arcsec pixel − that allows detection of transiting exoplanets in the densestellar fields (e.g., Konacki et al. 2003; Pont et al. 2008).However, only small parts of the Galactic Bulge and Diskare observed with a cadence that allows transit detections(Udalski et al. 2015).Therefore, there is an opportunity for a dedicated exo-planet survey that will explore the Galactic plane with suf-ficient spatial resolution and cadence to find new transit-ing exoplanets. Motivated by this, we initiated the Galac-tic Plane eXoplanet (GPX) survey. GPX is a multinationalproject involving a collaboration of amateur and professionalastronomers from Europe, Asia, and North America. GPXevolved from the prototype KPS survey (Burdanov et al.2016), which resulted in a discovery of the transiting hotJupiter KPS-1 b (Burdanov et al. 2018). The main goalof GPX is to survey high-density star fields of the Galac-tic plane with moderately high image resolution of 2 arc-sec pixel − in order to find new transiting gas giants. Inthis paper, we present the discovery of a transiting BD witha mass of . ± . M Jup and a radius of . ± . R Jup orbiting F2 star 2MASS 02332859+5601325 (henceforth re-ferred to as the GPX-1) with a period of ∼ V ∼ MNRAS , 1–13 (2020) iscovery of GPX-1 b Table 1.
General information about GPX-1. Position, distanceand motion information are based on Gaia DR2 data (Bailer-Jones et al. 2018). B , V , g (cid:48) , r (cid:48) and i (cid:48) magnitudes are from theAPASS catalog (Henden et al. 2016), J , H , K s magnitudes – fromthe 2MASS catalog (Cutri et al. 2003), W , W , W – from the WISE catalog (Cutri & et al. 2012).Identifiers GPX-1 bGPX-TF8A-1859 (GPX input catalog ID)2MASS 02332859+5601325Gaia DR2 4573175348800811521SWASP J023328.42+560133.4GSC 3691:0475TESS TIC 245392284RA (J2000.0) 02h 33m 28.606sDEC (J2000.0) +56 ◦ (cid:48) (cid:48)(cid:48) Gal l ◦ Gal b -4.055172 ◦ Distance ± pcpmRA − . ± . mas/ypmDEC . ± . mas/yRV − . ± . km/sTESS mag . ± . APASS B mag . ± . APASS V mag . ± . APASS g (cid:48) mag . ± . APASS r (cid:48) mag . ± . APASS i (cid:48) mag . ± . J mag . ± . H mag . ± . K s mag . ± . WISE W mag . ± . WISE W mag . ± . WISE W mag . ± . The GPX survey telescope was built from readily avail-able commercial equipment: a RASA 11” wide-field telescope(279 mm, f/2.2; Celestron, Torrance, CA USA), based at theprivate observatory Acton Sky Portal in Acton, MA, USA.The GPX survey covered 10 star fields with a Field of View(FoV) of . × . deg and 54 fields with a . × . deg FoV in 2015-2019, which were observed for 3-5 months with150-180 hours of image data collected per field. A detaileddescription of the GPX survey observations is presented inBurdanov et al. (2018).The star GPX-1 (see Table 1) was observed with theRASA telescope from September to December 2016 with theJohnson-Cousins R c band. Survey images were treated withthe K-pipe pipeline described in Burdanov et al. (2016). Inshort, the pipeline performs a calibration of the FITS im-ages (bias, dark and flat-field corrections), extracts fluxesof all the stars in the FoV, conducts differential photome-try, and performs a period search with the Box-fitting LeastSquares method (BLS; Kov´acs et al. 2002). Stars with low-noise light curves were selected as high-priority and theirphase-folded light curves were examined visually. Using BLS,a strong peak corresponding to a ∼ We conducted a photometric follow-up campaign to observetransits of GPX-1 b with several goals: to confirm that thesignals found in the GPX wide-field data were real; to re-fine the transit ephemeris; and to compare the depths of thetransits in different filters to check their chromaticity to ver-ify that the signal was not caused by an eclipsing binary. Adozen different telescopes participated in the follow-up cam-paign. The majority of those observations were performedby the small- and middle-aperture telescopes, which partic-ipate in the EXPANSION project (EXoPlanetary trANsitSearch with an International Observational Network; Sokovet al. 2018). In particular, multi-colour observations of twoGPX-1 b transits in 2017 with the 1-m telescope T100 of theT ¨UB˙ITAK National Observatory of Turkey (TUG) helpedus to discard evident eclipsing binary scenario and allowed torequest spectroscopic observations with the SOPHIE spec-trograph (see sub-section 2.5). We used all available data forthe initial characterisation of the system, but here we willpresent only the most precise light curves from our observa-tional sets, which were used to derive the system parameters.High-precision light curves of GPX-1 were obtainedwith four telescopes (see Table 2). Five transits were ob-served with the RC600 telescope of the Caucasian Moun-tain Observatory (CMO) of Sternberg Astronomical Insti-tute (SAI) of Moscow State University (MSU) in g (cid:48) , r (cid:48) , and R c filters (Berdnikov et al. 2020). Three partial transits wereobserved with the TRAPPIST-North telescope (Barkaouiet al. 2017; Jehin et al. 2011; Gillon et al. 2011) at theOuka¨ımeden Observatory in z (cid:48) filter and in a custom near-infrared filter I + z (cid:48) (transmittance > from 700 nm).One full transit in R c and B bands were obtained with theAZT-11 telescope of the Crimean Astrophysical Observa-tory (CrAO) and 1-m telescope of the Deokheung OpticalAstronomy Observatory (DOAO). The MASTER telescopeat the Kourovka Observatory was used to observe GPX-1in out-of-transit phase in B , V , R , I filters to construct aSpectral Energy Distribution (SED).All obtained images were treated similarly: data reduc-tion consisted of dark and flat-field corrections and aper-ture photometry with IRAF/DAOPHOT (Tody 1986). Pho-tometric extraction was performed with different aperturesizes. Selection of the best aperture and comparison starswas made based on the minimization of the out-of-transitscatter of the light curve. The resulting light curves are pre-sented in Fig. 3, after period-folding and being multipliedby baseline polynomials to correct for systematic errors (seesub-section 3.2 for details). Achromatic transits with depths0.9% are clearly visible, which strengthens the hypothesisthat the transiting body is not a star.For the SED fit, we measured instrumental magnitudesof GPX-1 and nearby stars in the FoV and then used theAPASS catalog (Henden et al. 2016) to convert instrumentalmagnitudes to the standard ones. MNRAS000
General information about GPX-1. Position, distanceand motion information are based on Gaia DR2 data (Bailer-Jones et al. 2018). B , V , g (cid:48) , r (cid:48) and i (cid:48) magnitudes are from theAPASS catalog (Henden et al. 2016), J , H , K s magnitudes – fromthe 2MASS catalog (Cutri et al. 2003), W , W , W – from the WISE catalog (Cutri & et al. 2012).Identifiers GPX-1 bGPX-TF8A-1859 (GPX input catalog ID)2MASS 02332859+5601325Gaia DR2 4573175348800811521SWASP J023328.42+560133.4GSC 3691:0475TESS TIC 245392284RA (J2000.0) 02h 33m 28.606sDEC (J2000.0) +56 ◦ (cid:48) (cid:48)(cid:48) Gal l ◦ Gal b -4.055172 ◦ Distance ± pcpmRA − . ± . mas/ypmDEC . ± . mas/yRV − . ± . km/sTESS mag . ± . APASS B mag . ± . APASS V mag . ± . APASS g (cid:48) mag . ± . APASS r (cid:48) mag . ± . APASS i (cid:48) mag . ± . J mag . ± . H mag . ± . K s mag . ± . WISE W mag . ± . WISE W mag . ± . WISE W mag . ± . The GPX survey telescope was built from readily avail-able commercial equipment: a RASA 11” wide-field telescope(279 mm, f/2.2; Celestron, Torrance, CA USA), based at theprivate observatory Acton Sky Portal in Acton, MA, USA.The GPX survey covered 10 star fields with a Field of View(FoV) of . × . deg and 54 fields with a . × . deg FoV in 2015-2019, which were observed for 3-5 months with150-180 hours of image data collected per field. A detaileddescription of the GPX survey observations is presented inBurdanov et al. (2018).The star GPX-1 (see Table 1) was observed with theRASA telescope from September to December 2016 with theJohnson-Cousins R c band. Survey images were treated withthe K-pipe pipeline described in Burdanov et al. (2016). Inshort, the pipeline performs a calibration of the FITS im-ages (bias, dark and flat-field corrections), extracts fluxesof all the stars in the FoV, conducts differential photome-try, and performs a period search with the Box-fitting LeastSquares method (BLS; Kov´acs et al. 2002). Stars with low-noise light curves were selected as high-priority and theirphase-folded light curves were examined visually. Using BLS,a strong peak corresponding to a ∼ We conducted a photometric follow-up campaign to observetransits of GPX-1 b with several goals: to confirm that thesignals found in the GPX wide-field data were real; to re-fine the transit ephemeris; and to compare the depths of thetransits in different filters to check their chromaticity to ver-ify that the signal was not caused by an eclipsing binary. Adozen different telescopes participated in the follow-up cam-paign. The majority of those observations were performedby the small- and middle-aperture telescopes, which partic-ipate in the EXPANSION project (EXoPlanetary trANsitSearch with an International Observational Network; Sokovet al. 2018). In particular, multi-colour observations of twoGPX-1 b transits in 2017 with the 1-m telescope T100 of theT ¨UB˙ITAK National Observatory of Turkey (TUG) helpedus to discard evident eclipsing binary scenario and allowed torequest spectroscopic observations with the SOPHIE spec-trograph (see sub-section 2.5). We used all available data forthe initial characterisation of the system, but here we willpresent only the most precise light curves from our observa-tional sets, which were used to derive the system parameters.High-precision light curves of GPX-1 were obtainedwith four telescopes (see Table 2). Five transits were ob-served with the RC600 telescope of the Caucasian Moun-tain Observatory (CMO) of Sternberg Astronomical Insti-tute (SAI) of Moscow State University (MSU) in g (cid:48) , r (cid:48) , and R c filters (Berdnikov et al. 2020). Three partial transits wereobserved with the TRAPPIST-North telescope (Barkaouiet al. 2017; Jehin et al. 2011; Gillon et al. 2011) at theOuka¨ımeden Observatory in z (cid:48) filter and in a custom near-infrared filter I + z (cid:48) (transmittance > from 700 nm).One full transit in R c and B bands were obtained with theAZT-11 telescope of the Crimean Astrophysical Observa-tory (CrAO) and 1-m telescope of the Deokheung OpticalAstronomy Observatory (DOAO). The MASTER telescopeat the Kourovka Observatory was used to observe GPX-1in out-of-transit phase in B , V , R , I filters to construct aSpectral Energy Distribution (SED).All obtained images were treated similarly: data reduc-tion consisted of dark and flat-field corrections and aper-ture photometry with IRAF/DAOPHOT (Tody 1986). Pho-tometric extraction was performed with different aperturesizes. Selection of the best aperture and comparison starswas made based on the minimization of the out-of-transitscatter of the light curve. The resulting light curves are pre-sented in Fig. 3, after period-folding and being multipliedby baseline polynomials to correct for systematic errors (seesub-section 3.2 for details). Achromatic transits with depths0.9% are clearly visible, which strengthens the hypothesisthat the transiting body is not a star.For the SED fit, we measured instrumental magnitudesof GPX-1 and nearby stars in the FoV and then used theAPASS catalog (Henden et al. 2016) to convert instrumentalmagnitudes to the standard ones. MNRAS000 , 1–13 (2020)
Benni et al.
105 52.5 0 -52.5 -105Offset (arcsec)10552.50-52.5-105 O ff s e t ( a r c s e c ) NE 0 2 4 6 8Pixels02468 P i x e l s NE105 52.5 0 -52.5 -105Offset (arcsec)10552.50-52.5-105 O ff s e t ( a r c s e c ) NE Figure 1.
Left: Pan-Starrs i (cid:48) image of a ×
210 arcsec region around GPX-1 ( V ∼
12 mag), obtained with an image scale of0.25 arcsec pixel − . Note the bright star HD 15691 ( V ∼ − . Right: TESS ×
210 arcsec ( ×
10 pixel ) image of the sameField of View (FoV). Table 2.
Photometric follow-up observations log.Observatory Telescope Detector Date (filter)CMO SAI MSU RC600, D=60 cm, F/7 Andor iKon-L BV 15 Sept 2019 ( r (cid:48) ), 13 Oct 2019 ( r (cid:48) )8 Nov 2019 ( R c ), 22 Nov 2019 ( r (cid:48) )06 Dec 2019 ( g (cid:48) )Ouka¨ımeden TRAPPIST-N, D=60 cm, F/8 Andor iKon-L BEX2 DD 16 Aug 2019 ( z (cid:48) ), 23 Aug 2019 ( z (cid:48) )13 Sept 2019 ( I + z (cid:48) )CrAO AZT-11, D=125 cm, F/13 FLI ProLine PL230 8 Nov 2019 ( R c )DOAO 1-m, D=100 cm, F/8 FLI ProLine PL-16803 21 Jan 2019 ( B ) N o r m a li z e d f l u x Figure 2.
GPX discovery light curve as obtained with the RASA11” wide-field telescope and folded with ∼ GPX-1 (TIC 245392284) was observed at 30-min integra-tions by TESS in Sector 18 in November 2019, which spans13 orbital periods of GPX-1 b. Since GPX-1 was not ob-served with 2-min-cadence, the Science Processing Opera-tions Center pipeline (SPOC; Jenkins et al. 2016) did not ex- tract photometric flux nor attempted a transit search. Lightcurve of GPX-1 was extracted by the Quick-Look Pipeline(QLP; Huang et al. 2018, Huang et al. 2020 in prep.), but itwas not included for candidate vetting procedures as in thevast majority of cases vetting is carried out for exoplanetcandidates brighter than TESS mag of 10.5 (and TESS magof GPX-1 is 11.9). However, even if GPX-1 light curve wasincluded for vetting, it would not pass predetermined thresh-old for a transit signal.We obtained the light curve from the TESS Full FrameImages (FFIs) using a 2-pixel aperture. The raw lightcurve (not corrected for systematics) is presented in theupper panel of Fig. 4. We removed the systematic trendsin the data by performing decorrelation with quaternions(high-cadence vector time-series that describe TESS atti-tude based on observations of guide stars; Vanderburg et al.2019) and the background flux outside the aperture. Then,we made use of the SHERLOCK pipeline (Pozuelos et al.2020) to search for transit signals in the data. The pipelineperformed transit search by means of the Transit LeastSquares package (TLS; Hippke & Heller 2019). We success-fully recovered the ∼ > > MNRAS , 1–13 (2020) iscovery of GPX-1 b N o r m a li z e d f l u x B band g band r band Rc band I + z band z band Figure 3.
Transits of GPX-1 b in different bands as obtainedwith RC600, TRAPPIST-North, AZT-11 and DOAO 1-m tele-scopes. The observations are phase-folded, multiplied by the base-line polynomials (detrended), and are shown in different colours(depending on the filter used). For each transit, the solid blackline represents the best-fit model. All transits were normalisedand shifted vertically for visual clarity. with a dilution term, i.e. allowing the transit depth to beset by the ground-based observations. The resulting period-folded light curve is presented in Fig. 5, where a ∼ σ upper limiton occultation depth of 0.04 % (which corresponds to ∼ ). The WASP telescope hada better image scale than TESS, but blending with afore-mentioned nearby star HD 15691 coupled with a poorer pho-tometric precision (compared to TESS), prevented robusttransit detections in the data. WASP time-series viewer is available at https://exoplanetarchive.ipac.caltech.edu
Figure 4.
Upper panel: TESS data for GPX-1 (TIC 245392284)from Sector 18. The teal line corresponds to the light curve fromthe Full Frame Images using a 2-pixel aperture, and black lineis the final clean light curve. The red triangles mark the ∼ ∼ N o r m a li z e d f l u x Figure 5.
Phase-folded light curve of GPX-1 b as obtained usingthe TESS data. Due to a blending with a nearby star, obtainedtransit depth is ∼ ∼ We carried out speckle-interferometric observations of GPX-1 to identify any nearby companions, which might affectour transit photometry and spectroscopy, or potentially bethe true source of the transit signal. Observations were per-formed with the use of the 6-m telescope of the Special As-trophysical Observatory (SAO) of the Russian Academy ofSciences (RAS) on 8 October and 1 December 2017. Both ob-servations were performed using SAO speckle-interferometer(Maksimov et al. 2009) with . × . FoV with a100 nm-wide filter centred on the wavelength of 800 nm,which corresponds to the photometric I band. The lengthof the series was 2000 frames, and the exposure time perframe was 100 ms and 50 ms for the nights of 8 October and1 December 2017, respectively. Image processing includedthe calculation of the average power spectra of the seriesand construction of the corresponding auto-correlation func-tions. The search for companions was done by analysingauto-correlation functions. The calculated limits for the MNRAS000
Phase-folded light curve of GPX-1 b as obtained usingthe TESS data. Due to a blending with a nearby star, obtainedtransit depth is ∼ ∼ We carried out speckle-interferometric observations of GPX-1 to identify any nearby companions, which might affectour transit photometry and spectroscopy, or potentially bethe true source of the transit signal. Observations were per-formed with the use of the 6-m telescope of the Special As-trophysical Observatory (SAO) of the Russian Academy ofSciences (RAS) on 8 October and 1 December 2017. Both ob-servations were performed using SAO speckle-interferometer(Maksimov et al. 2009) with . × . FoV with a100 nm-wide filter centred on the wavelength of 800 nm,which corresponds to the photometric I band. The lengthof the series was 2000 frames, and the exposure time perframe was 100 ms and 50 ms for the nights of 8 October and1 December 2017, respectively. Image processing includedthe calculation of the average power spectra of the seriesand construction of the corresponding auto-correlation func-tions. The search for companions was done by analysingauto-correlation functions. The calculated limits for the MNRAS000 , 1–13 (2020)
Benni et al. m m m m m m m m m † † † † † † † † † † † Radius, arcsec M agn i t ude d i ff e r en c e Figure 6.
Results of SAO speckle-interferometer observations ofGPX-1. The 3- σ detection limit is plotted as a function of radius.The limits are 0.03 arcsec for a brightness difference ∆ m = 0 mag,0.05 arcsec for ∆ m = 3 mag and 0.1 arcsec for ∆ m = 4.5 mag. presence of components near GPX-1 are shown in Fig. 6.Based on the analysis of the SNR by series, we were unableto detect a visual companion of GPX-1, and we provide con-servative limits on the presence of a secondary companion.The limits are 0.03 arcsec for a brightness difference ∆ m =0 mag, 0.05 arcsec for ∆ m = 3 mag and 0.1 arcsec for ∆ m =4.5 mag. These limits are consistent with the data obtainedin both observational sets. GPX-1 was observed with the SOPHIE spectrograph to ob-tain RV data to measure the mass of the transiting body.SOPHIE is dedicated to high-precision RV measurementswith the 1.93-m telescope of the Haute-Provence Observa-tory (Perruchot et al. 2008; Bouchy et al. 2009). We usedSOPHIE in High-Efficiency mode with a resolving power R =
40 000 and slow readout mode. We obtained 16 observa-tions between October 2017 and March 2018. Depending onobserving conditions, exposure times ranged between 9 and32 min in order to maintain an SNR of ∼
20 as constant aspossible throughout the observations. This SNR was chosenas a compromise between accuracy and exposure time.The spectra were extracted using the standard SO-PHIE data reduction pipeline. We used a weighted cross-correlation function (CCF) with a G2-type numerical mask(Baranne et al. 1996; Pepe et al. 2002). The CCF has con-trasts representing only ∼ ± km/s. This corresponds to a high projected rotationalvelocity v sin i ∗ ∼
40 km/s according to the calibration ofBoisse et al. (2010). We fitted the CCFs with Gaussians tomeasure the RVs, their errors bars, and the bisector spans(Table 3).Such broad and shallow CCFs allowed us to determineRVs with only a relatively low precision of ± m/s. Still,the RVs show significant variations, which are in agreementwith the period and phase obtained from the photometrictransits. The semi-amplitude of a simple fit to the RVs is of Table 3.
SOPHIE measurements of GPX-1.BJD
UTC
RV 1- σ bisect. ∗ exp. SNR † -2 458 000 (km/s) (km/s) (km/s) (sec)038.5049 -12.76 0.52 0.42 792 23.0041.4569 -16.32 0.49 1.55 536 23.5053.5725 -16.43 0.53 -0.77 795 22.7054.4488 -10.49 0.51 0.56 978 21.9055.6264 -15.21 0.51 0.61 1207 22.4057.5347 -13.81 0.47 -0.27 723 22.9083.5412 -14.69 0.57 -2.12 1800 18.0084.5573 -13.51 0.58 0.06 1800 19.3085.5799 -11.45 0.49 -0.78 844 23.1121.4556 -13.44 0.74 -5.93 1004 19.7123.4021 -15.70 0.59 -1.52 875 22.3171.3555 -11.81 0.53 0.70 1362 23.0172.2931 -14.66 0.50 -1.77 1307 22.0177.3268 -15.51 0.51 -0.37 1919 23.4204.2990 -11.32 0.64 2.80 1800 18.6206.2974 -10.32 0.51 1.17 1100 23.0 ∗ : bisector spans; error bars are twice those of the RVs. † : SNR per pixel at 550 nm.
17 16 15 14 13 12 11 10
Radial velocity [Km/s] B i s e c t o r s p a n [ K m / s ] Figure 7.
GPX-1 bisector span as a function of the radial veloc-ities (RVs) with 1- σ error bars. order 2.5 km/s which would correspond to a ∼ M Jup com-panion.To check if RV variations were caused by a blending sce-nario of stars with different spectral types, we measured RVswith a set of different stellar masks. In all cases, we obtainedRV variations with similar amplitudes and we can concludethat the blending scenario is very unlikely. To assess possiblechange of the shape of spectral lines due to stellar activityor blends, we measured CCF bisector spans. Our measure-ments have low precision due to the broad, shallow CCFs,but they show no correlations with the RV variations (aSpearman correlation coefficient is 0.33; see Fig. 7). There-fore, we can conclude that the observed RV variations arecaused by a sub-stellar companion.
GPX-1 has a spectral type F2 and is a fast rotator witha line broadening parameter v sin i ∗ ∼
40 km/s. This combi-
MNRAS , 1–13 (2020) iscovery of GPX-1 b dT [days] R V [ m / s ] dT [days] O - C [ m / s ] Figure 8.
Top: SOPHIE radial velocity (RV) measurements ofGPX-1 phase-folded at the 1.75 days period of the BD with theimposed best-fit Keplerian model. Bottom: best-fit residuals. nation makes the inference of stellar parameters, such asthe effective temperature T eff , metallicity [Fe/H], and sur-face gravity log g (cid:63) from spectroscopy, challenging and unre-liable due to the small number of spectral absorption linesand their broadening. However, knowledge of a host star iscrucial as all the parameters of a companion are obtainedrelative to the host star.To address this problem, we used the MCMCI tool toperform an integrated modelling of GPX-1 and its transit-ing BD. MCMCI is a code that utilises Markov Chain MonteCarlo (MCMC) stochastic simulations of the RV and photo-metric data to analyse the posterior probability distributionfunction (PDF) of the model parameters coupled with theisochrone placement algorithm. The MCMCI code and thelogic behind it are described in details in Bonfanti & Gillon(2020) and here we outline the main aspects of our analysis.The MCMCI tool performs an integrated three-step analy-sis:(i) Photometric light curves are analysed to infer a meanstellar density ρ (cid:63) ;(ii) The stellar density ρ (cid:63) together with the stellar metal-licity [Fe/H], and effective temperature T eff (computed fromphotometric colour) are used to infer the stellar mass M (cid:63) ,radius R (cid:63) and age by the isochrone placement algorithm,which considers pre-computed grids of isochrones and tracksbased on the PAdova and TRieste Stellar Evolutionary Code(PARSEC, v1.2S; Marigo et al. 2017). The isochrone place-ment algorithm is presented in detail in Bonfanti et al. (2015,2016);(iii) Inferred stellar parameters are used to recover com-panion’s radius R BD and mass M BD using the photometricand RV data sets.Similar to the MCMC code described in Gillon et al.(2012), the MCMCI tool utilises a Keplerian model by Mur-ray & Correia (2010) and a transit model by Mandel & Agol(2002) to jointly fit the RV and photometric data. Wavelength, (um)10 F l u x , e r g Å s c m Figure 9.
The APASS ( B , V , g (cid:48) , r (cid:48) , i (cid:48) ), 2MASS ( J , H , and K s ), WISE ( W , W ), MASTER ( B , V , R , I ) and TRAPPIST-North( B , V , R c , I c ) spectral points are shown. The blue curve is amodel for the star with T eff = K, R (cid:63) = . R (cid:12) , log g (cid:63) = . and [ Fe / H ] = . at 655 pc and E ( B − V ) = . . We would like to note that interstellar dust extinctionplays a major role when observing a distant object withlow galactic latitudes by attenuating and reddening observedphotometric magnitudes. In the case of GPX-1, its distanceis 655 ±
17 pc (Bailer-Jones et al. 2018) and its galacticlatitude is b ∼ -4 deg. We constructed a SED of GPX-1(see Fig. 9) using available photometric data from catalogsand observations made with the MASTER telescope (Gor-bovskoy et al. 2013) and with the TRAPPIST-North tele-scope. We performed a fit using a grid of ATLAS9 stellaratmosphere models (Castelli & Kurucz 2003) with a con-stant Gaia DR2 distance, preliminary estimation of log g (cid:63) in the range . − . and [ Fe / H ] in the range . − . , todetermine the amount of interstellar reddening and stellarparameters. Our best-fit parameters are: T eff = ± K, R (cid:63) = . ± R (cid:12) , [ Fe / H ] = . ± . and E ( B − V ) = . ± . .The derived value of reddening is in a good agreement withthe interpolated values from Chen et al. (2019). Thus, GPX-1 colour and magnitude corrected for interstellar extinctionare: ( B − V ) = . ± . mag and V = . ± . mag.The stellar metallicity [Fe/H] from the SED fit, thedereddened ( B − V ) colour index, and the stellar luminosity L (cid:63) (computed from the dereddened V magnitude and thedistance from Gaia DR2 catalog (Gaia Collaboration et al.2018) were used as priors. The stellar effective temperature T eff (computed from ( B − V ) colour), stellar luminosity L (cid:63) (computed from the V magnitude and the distance), andmean density ρ (cid:63) (computed from light curves) were the in-put parameters for computing M (cid:63) , R (cid:63) , and the age fromisochrones. Before performing a global MCMC analysis of the availablehigh-precision photometric and RV data, we performed a
MNRAS000
MNRAS000 , 1–13 (2020)
Benni et al. preliminary analysis of each light curve. The goal of thisanalysis was to account for correlations of the extractedphotometric fluxes with the external environmental and/orinstrumental parameters by obtaining a proper baseline cor-rection model (see Gillon et al. 2012 and Burdanov et al.2018 for a more detailed description of the applied method).In most cases, corrections for position drift of stars on aCCD and rapid changes of FWHM were applied.For our global analysis of all the data sets, the followingparameters were used as jump parameters in our MCMCsimulations and were allowed to vary: • the luminosity of the host star L (cid:63) ; • the brown dwarf (BD) orbital period P ; • the BD transit duration W ; • the BD mid-transit time T ; • the ratio of the BD and host star areas ( R BD / R (cid:63) ) ,where R BD is the BD radius and R (cid:63) is the stellar radius; • the occultation depth dF occ in TESS filter; • the impact parameter b (cid:48) = a cos i BD / R (cid:63) for a circularorbit, where a BD is the semi-major axis and i BD is the orbitalinclination of the BD; • the parameter K = K √ − e P / , where K is the radialvelocity orbital semi-amplitude; • √ e sin ω and √ e cos ω parameters, where ω is the argu-ment of periastron; • the effective temperature T eff of the host star computedfrom ( B − V ) colour and metallicity [Fe/H] from the SEDfit; • the combinations c = u + u and c = u − u of thequadratic limb-darkening coefficients u and u .All jump parameters, for which we had no prior con-straints had uniform non-informative prior distributions.Quadratic limb-darkening (LD) law coefficients u and u with normal priors distributions were interpolated from thepaper by Claret et al. (2012) and applied to the data in eachphotometric band. We converted all time-stamps of our mea-surements from HJD
UTC to BJD
TDB (Eastman et al. 2010).Then, we ran one relatively short MCMC chain of 20 000steps to obtain a correction factor (CF) for every light curve,which was used to multiply initial photometric errors. Thisprocedure is done to rescale the photometric errors and ac-count for over- or underestimation of photometric noise (seeGillon et al. 2012 and Burdanov et al. 2018 for the details).Once proper CFs were obtained, five 100 000-step chainswere executed with starting points inferred from perturbedBLS solutions and with a 20% burn-in phase. Convergenceof the chains was checked with the use of Gelman-Rubinstatistical test (Gelman & Rubin 1992). Parameters of theBD were deduced from the set of jump parameters and stel-lar parameters ( M (cid:63) , R (cid:63) ) inferred from isochrones. Deducedparameters of the system for a circular orbit are presentedin Table 4 (median values of the posterior PDFs and theirrespective 1- σ limits) and the results of the analysis arepresented in the next section. We also performed a secondanalysis of the data where eccentricity was allowed to float.We inferred a 3- σ upper limit on orbital eccentricity of 0.1. Table 4.
Inferred parameters of GPX-1 system: median values ofthe posterior PDFs and their respective 1- σ limits.Output parameters from the global MCMC analysisJump parameters ValueOrbital period P [d] . ± . Transit width W [d] . ± . Mid-transit time T [ BJD
TDB ] . ± . b (cid:48) = a cos i BD / R (cid:63) [ R (cid:63) ] . + . − . BD/star area ratio ( R BD / R (cid:63) ) [%] . ± . dF occ (TESS band) [%] < σ ) K [m/s] ± √ e sin ω √ e cos ω T eff [K] ± [Fe/H] [dex] . ± . ( B − V ) colour [mag] . ± . Luminosity L (cid:63) [ L (cid:12) ] . ± . Deduced stellar parameters ValueMass M (cid:63) [ M (cid:12) ] . ± . Radius R (cid:63) [ R (cid:12) ] . ± . Mean density ρ (cid:63) [ ρ (cid:12) ] . ± . Surface gravity log g (cid:63) [cgs] . ± . Age [Gyr] . + . − . Deduced BD parameters Value K [m/s] ± BD/star radius ratio R BD / R (cid:63) . ± . Orbital semi-major axis a [au] . ± . Orbital eccentricity e i BD [deg] . + . − . Argument of periastron ω [deg] -Surface gravity log g p [cgs] . ± . Mean density ρ p [ ρ Jup ] . + . − . Mass M BD [ M Jup ] . ± . Radius R BD [ R Jup ] . ± . Roche limit a R [au] . ± . a / a R . ± . Equilibrium temperature T eq [K] ± Irradiation I BD [ I Earth ] ± We obtained a mass and a radius of the host star of . ± . M (cid:12) and . ± . R (cid:12) respectively. Given anRV semi-amplitude of ± m/s and a transit depthof . ± . %, we estimate GPX-1 b to have a mass of . ± . M Jup and a radius of . ± . R Jup . Based on theseresults from the data modelling, achromaticity of the transitdepths, the lack of a visual companion near GPX-1, and alack of bisector variations, we are confident that the resultsdemonstrate that the F2 star GPX-1 has a BD companion.With its ∼ ∼
80 deg inclination, GPX-1 b enters a small set of transiting and short-period BDs,where similar BDs are HATS-70 b ( P ∼ P ∼ ± greater than that of the Earth, and its equi-librium temperature, assuming complete heat redistributionand zero albedo, is ± K. However, its radius is 7%larger than that of the M Jup
HATS-70 b ( . + . − . R Jup ),which receives almost twice as much irradiation as GPX-1 b.One important difference between these two young systems
MNRAS , 1–13 (2020) iscovery of GPX-1 b is the age, with HATS-70 b, estimated to be . + . − . Gyrand for GPX-1 b to be only . + . − . Gyr (or + − Myr;median value of the posterior PDF and its respective 1- σ limits).Unfortunately, our age estimate from the MCMCI codehas large uncertainties and GPX-1 is too hot for the ap-plication of gyrochronology relations, through which an agecould be independently estimated from the stellar rotationrate. Also, GPX-1 is very likely not a member of the nearbyGalactic open cluster Trumpler 2, which has an age estimateof ∼
89 Myr obtained by Frolov et al. (2006) and ∼
84 Myr byKharchenko et al. (2013). The most recent age estimationof ∼
92 Myr is made by Bossini et al. (2019) using Gaia DR2data. We used the basic physical properties of GPX-1 b andits host star together to infer the evolutionary state of thesystem. Figure 10 shows the expected evolution of both thestar and BD in radius, T eff , and luminosity, based on theirmasses and metallicity. The stellar track is from the PAR-SEC models (v1.2S; Marigo et al. 2017), whereas the BDtrack is from the sub-stellar models of Phillips et al. (2020).They are compared with the measured/inferred propertiesat three different ages: 200 Myr (close to the median estimatefrom our global solution), 75 Myr (close to the literature es-timated age for Trumpler 2), and 25 Myr (chosen to matchthe observed properties of both the star and BD simultane-ously).Taken at face value, all of the BD’s properties are consis-tent with it being in an early stage of contraction at ∼
25 Myr,requiring no inflation. At that age, the properties of the hoststar are also simultaneously consistent with expectation. Tobe clear, the T eff adopted here for the BD is its equilibriumtemperature T eq from the global solution, which might notbe the same as the T eff that is predicted by the models. Al-ternatively, an age of 75 Myr could be correct, in which casethe BD radius is somewhat inflated, but much less so thanfor an age of 200 Myr. We checked some other possible signsof a star’s youth, such as H-alpha and X-rays emissions, butfound nothing to help clarify its age. Some methods, such aslithium abundances, require additional high signal-to-noisespectra.We checked whether the GPX-1 system was a mem-ber of the nearby open cluster Trumpler 2 (Kharchenkoet al. 2012a), which has an age of ∼
92 Myr, an angularradius of 0.45 deg and which is located at the same dis-tance as GPX-1 (670 pc). Kharchenko et al. (2012b) ratedthe probability of GPX-1 membership in that cluster ac-cording to its angular distance as 0%, according to propermotion as 4.7%, and according to 2MASS photometry as90% and 100% for J − K s and J − H respectively. In addi-tion, we tested if GPX-1 was a member of Trumpler 2 us-ing Gaia DR2 parallaxes, proper motions and RVs. We de-rived cluster parameters as a local maximum in a 6-D space(Ra, Dec, distance, pmRa, pmDec, RV): Ra = 39.32634 ◦ ,Dec = 55.93867 ◦ , distance = 711 pc, pmRa = .
60 mas / y , pmDec = − .
27 mas / y , RV = − .
77 km / s . Correspondingparameters of GPX-1 are presented in Table 1. There is aseparation of 60 pc at the present time, and the minimumseparation of 27 pc will occur in 4.5 Myr, whereas a linearradius of Trumpler 2 for its distance and angular radius is5.6 pc. Thus, we believe that GPX-1 was never a memberof Trumpler 2. The same conclusion about membership of the GPX-1 in Trumpler 2 was obtained in (Cantat-Gaudin& Anders 2020), using UPMASK procedure applied to theGAIA DR2 astrometric data (Cantat-Gaudin et al. 2018).Regarding its position on the mass-radius diagram ofknown transiting BDs (see Fig. 11), GPX-1 b is among thelargest and youngest BDs that happen to transit their hoststars. If its true age is close to a value of . + . − . Gyr ob-tained from our global modelling, then it is significantly in-flated. Then, such an object could serve as a testbed forradius inflation theories.We expect GPX-1 b to have a low projected obliquityas massive exoplanets and BDs tend to be aligned (H´ebrardet al. 2011; Triaud 2018), but Rossiter-McLaughlin observa-tions should confirm this. Such a BD with measured obliq-uity could be important for further understanding of theorigins of short-period BDs.Assuming that the host star and the BD are black-bodies, and that the thermal energy dominates, measuringthe observed occultation depth can reveal the BD’s effec-tive temperature. Alternatively, we can assess expected oc-cultation depth if T eff of GPX-1 b is close to its T Eq fromthe global solution. In this case, occultation depth (in theRayleigh-Jeans limit) is expected to be 0.02%, 0.05% and0.08% for J , H and K s bands respectively. In the case ofno heat redistribution, occultation depth is expected to be0.08%, 0.12% and 0.17% for J , H and K s bands respectively. We presented here the discovery of GPX-1 b, a transitingBD on a short circular orbit with a mass of . ± . M Jup and a radius of . ± . R Jup . The BD transits a moder-ately bright fast-rotating F-type star with a projected ro-tational velocity v sin i ∗ ∼
40 km/s. Due to the small num-ber of spectral absorption lines and their broadening, weused an isochrone placement algorithm (Bonfanti et al. 2015,2016) to perform stellar characterization of GPX-1. We ob-tained stellar effective temperature T eff = ± K, mass . ± . M (cid:12) , radius . ± . R (cid:12) and approximate age + − Myr. We also used the basic physical properties ofGPX-1 b and its host star together to infer the evolutionarystate of the system. An age of ∼
25 Myr matches the observedproperties of both the star and BD simultaneously. In thiscase, no radius inflation is required. If the true age is 75 Myr(close to the estimated age for the nearby Galactic open clus-ter Trumpler 2), then the BD radius is inflated, but muchless than that for an age of 200 Myr. We checked whetherthe GPX-1 system was a member of the nearby open clusterTrumpler 2 (Kharchenko et al. 2012a) and we believe thatGPX-1 was never a member of Trumpler 2.Since GPX-1 was not observed by TESS with 2-min in-tegrations, the SPOC pipeline did not extract photometricflux nor attempted a transit search. Light curve of GPX-1was extracted by the QLP, but no transit search was carriedout as QLP reports exoplanet candidates only to TESS magof 10.5 (and TESS mag of GPX-1 is 11.9). Transit recoveryof GPX-1 b in the TESS data is complicated by the blendingby the nearby bright star HD 15691. The same star blendingissue is applicable to the WASP photometry, which preventsrobust transit detections in the data. These facts provide aproof of concept of the GPX survey’s scientific value and
MNRAS , 1–13 (2020) Benni et al.
10 100Age (Myr)0.11.0 R a d i u s ( R s un )
10 100Age (Myr)100010000 S u rf ace t e m p e r a t u r e ( K )
10 100Age (Myr)-4-3-2-101 l og L u m i no s it y ( L s un ) Figure 10.
Evolutionary state of the GPX-1 system. In each panel, the blue curve represents the PARSEC (v1.2S) evolutionary trackfor the GPX-1 primary star (Marigo et al. 2017), assuming the mass and [Fe/H] from the global solution; red curves represent the samefor the GPX-1 b brown dwarf, except using the sub-stellar evolutionary models of Phillips et al. (2020). Blue (red) symbols represent theobserved stellar (brown dwarf) properties at three ages: 200 Myr (close to the median estimate from our global solution); 75 Myr (closeto the literature estimated age for Trumpler 2); and 25 Myr (representing the best fit to both the components simultaneously).
10 20 30 40 50 60 70 80Mass ( M Jup sin i )1.01.52.02.53.03.5 R a d i u s ( R J u p ) GPX-1b A g e ( M y r ) Figure 11.
Mass-radius relationship of a set of known transitingbrown dwarfs ( ∼ M Jup ) as of May 2020 and isochrones fromthe evolutionary models of Phillips et al. (2020). motivates us to continue its operations. The TESS missionbrings exoplanet hunting into a new era with the delivery ofvery high-precision photometry of bright stars across the sky,beyond the ability of most ground-based telescopes. TESSwill survey most of the sky over two years with thousandsof expected planet discoveries (Barclay et al. 2018). How-ever, a number of transiting gas giants and BDs may still bediscovered in crowded fields (including open clusters) whenobserved by ground-based surveys, and even in the TESSdata, they might be difficult to detect (see Figure 3 and 4 inBarclay et al. 2018). Thus, the GPX survey and similar setups might find a niche in the TESS era and contribute bydiscovering new transiting gas giants and BDs in crowdedfields.
ACKNOWLEDGEMENTS
This research has made use of the Exoplanet OrbitDatabase, the Exoplanet Data Explorer at exoplanets.org,Extrasolar Planets Encyclopaedia at exoplanets.eu and theNASA Exoplanet Archive, which is operated by the Cali-fornia Institute of Technology under contract with the Na-tional Aeronautics and Space Administration under the Ex-oplanet Exploration Program. This research made use of Al-adin (Bonnarel et al. 2000). IRAF is distributed by the Na-tional Optical Astronomy Observatory, which is operatedby the Association of Universities for Research in Astron-omy, Inc., under cooperative agreement with the NationalScience Foundation. This research made use of Astropy, a community-developed core Python package for Astron-omy (Astropy Collaboration et al. 2013; Price-Whelan et al.2018).We acknowledge the use of TESScut.MAST data fromfull frame time series images (FFI) collected by the TESSmission, which are publicly available from the MikulskiArchive for Space Telescopes (MAST). Funding for theTESS mission is provided by NASA’s Science Mission direc-torate. Resources supporting this work were provided by theNASA High-End Computing (HEC) Program through theNASA Advanced Supercomputing (NAS) Division at AmesResearch Center for the production of the SPOC data prod-ucts.P. Benni thanks Bruce Gary, the XO survey, and theKELT survey for furthering his education in exoplanetresearch. A.Y.B would like to thank Catarina Fernandesand Julien de Wit for helpful discussions about the sys-tem. Organization of the EXPANSION project (E. Sokov),follow-up campaign of the photometry observations, speckle-interferometry observations with 6-m telescope BTA weresupported by the Russian Science Foundation grant 19-72-10023. The work of V.K. was supported by the Ministry of , 1–13 (2020) iscovery of GPX-1 b science and higher education of Russian Federation, topicFEUZ-0836-2020-0038. This work was partly supported bythe Ministry of Science and High Education (project no.FZZE-2020-0024) and Irkutsk State University (project no.111-14-306). This work was supported by the Ministry ofScience and Higher Education of the Russian Federation(projects no. FEUZ-2020-0030 and no. 13.1902.21.0039).TRAPPIST-North is a project funded by the University ofLiege, in collaboration with Cadi Ayyad University of Mar-rakech (Morocco). E.J. and M.G. are F.R.S.-FNRS SeniorResearch Associates. The research leading to these resultshas received funding from the ARC grant for Concerted Re-search Actions financed by the Wallonia-Brussels Federa-tion and from the Balzan Prize Foundation. TRAPPIST isfunded by the Belgian Fund for Scientific Research (FondNational de la Recherche Scientifique, FNRS) under thegrant FRFC 2.5.594.09.F. Erika Pakˇstien˙e acknowledges theEuroplanet 2024 RI project funded by the European Union’sHorizon 2020 Research and Innovation Programme (Grantagreement No. 871149). A. Belinski acknowledges the sup-port from the Program of development of M.V. LomonosovMoscow State University (Leading Scientific School ’Physicsof stars, relativistic objects and galaxies’). O.B. thanksT ¨UB˙ITAK National Observatory for a partial support inusing the T100 telescope with the project number 19AT100-1346. O.D.S.D. is supported by Portuguese national fundsthrough Funda¸c˜ao para a Ciˆencia e Tecnologia (FCT) inthe form of a work contract (DL 57/2016/CP1364/CT0004),institutional funds UIDB/04434/2020, UIDP/04434/2020and scientific projects funds PTDC/FIS-AST/28953/2017,POCI-01-0145-FEDER-028953. DATA AVAILABILITY
Data available on request. The data underlying this articlewill be shared on reasonable request to the correspondingauthor.
REFERENCES
Anderson D. R., et al., 2011, ApJ, 726, L19Armitage P. J., Bonnell I. A., 2002, MNRAS, 330, L11Astropy Collaboration et al., 2013, A&A, 558, A33Bailer-Jones C. A. L., Rybizki J., Fouesneau M., Mantelet G.,Andrae R., 2018, AJ, 156, 58Baraffe I., Chabrier G., Allard F., Hauschildt P. H., 2002, A&A,382, 563Baranne A., et al., 1996, A&AS, 119, 373Barclay T., Pepper J., Quintana E. V., 2018, ApJS, 239, 2Barkaoui K., Gillon M., Benkhaldoun Z., Emmanuel J., Elhalk-ouj T., Daassou A., Burdanov A., Delrez L., 2017, in Jour-nal of Physics Conference Series. p. 012073, doi:10.1088/1742-6596/869/1/012073Berdnikov L. N., Belinskii A. A., Shatskii N. I., Burlak M. A.,Ikonnikova N. P., Mishin E. O., Cheryasov D. V., Zhuiko S. V.,2020, Astronomy Reports, 64, 310Boisse I., et al., 2010, A&A, 523, A88Bonfanti A., Gillon M., 2020, A&A, 635, A6Bonfanti A., Ortolani S., Piotto G., Nascimbeni V., 2015, A&A,575, A18Bonfanti A., Ortolani S., Nascimbeni V., 2016, A&A, 585, A5Bonnarel F., et al., 2000, A&AS, 143, 33 Borucki W. J., et al., 2010, Science, 327, 977Boss A. P., et al., 2007, Transactions of the International Astro-nomical Union, Series A, 26A, 183Bossini D., et al., 2019, A&A, 623, A108Bouchy F., et al., 2009, A&A, 505, 853Burdanov A. Y., et al., 2016, MNRAS, 461, 3854Burdanov A., et al., 2018, PASP, 130, 074401Burrows A., et al., 1997, ApJ, 491, 856Cantat-Gaudin T., Anders F., 2020, A&A, 633, A99Cantat-Gaudin T., et al., 2018, A&A, 618, A93Carmichael T. W., Latham D. W., Vand erburg A. M., 2019, AJ,158, 38Carmichael T. W., et al., 2020, arXiv e-prints, p.arXiv:2002.01943Castelli F., Kurucz R. L., 2003, in Piskunov N., Weiss W. W.,Gray D. F., eds, IAU Symposium Vol. 210, Modelling of Stel-lar Atmospheres. p. A20 ( arXiv:astro-ph/0405087 )Chabrier G., Baraffe I., Allard F., Hauschildt P., 2000, ApJ, 542,464Chabrier G., Johansen A., Janson M., Rafikov R., 2014, inBeuther H., Klessen R. S., Dullemond C. P., Henning T.,eds, Protostars and Planets VI. p. 619 ( arXiv:1401.7559 ),doi:10.2458/azu uapress 9780816531240-ch027Chen B. Q., et al., 2019, MNRAS, 483, 4277Claret A., Hauschildt P. H., Witte S., 2012, A&A, 546, A14Cutri R. M., et al. 2012, VizieR Online Data Catalog, p. II/311Cutri R. M., et al., 2003, VizieR Online Data Catalog, 2246Eastman J., Siverd R., Gaudi B. S., 2010, PASP, 122, 935Frolov V. N., Ananjevskaja J. K., Jilinski E. G., Gorshanov D. L.,Bronnikova N. M., 2006, A&A, 451, 901Gaia Collaboration et al., 2018, A&A, 616, A1Gelman A., Rubin D. B., 1992, Statistical Science, 7, 457Gillon M., Jehin E., Magain P., Chantry V., Hutsem´ekers D.,Manfroid J., Queloz D., Udry S., 2011, in European Physi-cal Journal Web of Conferences. p. 06002 ( arXiv:1101.5807 ),doi:10.1051/epjconf/20101106002Gillon M., et al., 2012, A&A, 542, A4Gorbovskoy E. S., et al., 2013, Astronomy Reports, 57, 233Grether D., Lineweaver C. H., 2006, ApJ, 640, 1051Halbwachs J. L., Arenou F., Mayor M., Udry S., Queloz D., 2000,A&A, 355, 581Hayashi C., Nakano T., 1963, Progress of Theoretical Physics, 30,460H´ebrard G., et al., 2011, A&A, 533, A130Henden A. A., Templeton M., Terrell D., Smith T. C., Levine S.,Welch D., 2016, VizieR Online Data Catalog, 2336Hippke M., Heller R., 2019, A&A, 623, A39Hodˇzi´c et al., 2018, MNRAS, 481, 5091Howell S. B., et al., 2014, PASP, 126, 398Huang C. X., et al., 2018, ApJ, 868, L39Jackman J. A. G., et al., 2019, MNRAS, 489, 5146Jehin E., et al., 2011, The Messenger, 145, 2Jenkins J. M., et al., 2016, in Proc. SPIE. p. 99133E,doi:10.1117/12.2233418Kharchenko N. V., Piskunov A. E., Roeser S., Schilbach E.,Scholz R. D., 2012a, VizieR Online Data Catalog, ppJ/A+A/543/A156Kharchenko N. V., Piskunov A. E., Schilbach E., R¨oser S., ScholzR. D., 2012b, A&A, 543, A156Kharchenko N. V., Piskunov A. E., Schilbach E., R¨oser S., ScholzR. D., 2013, A&A, 558, A53Kiefer F., et al., 2019, A&A, 631, A125Konacki M., Torres G., Jha S., Sasselov D. D., 2003, Nature, 421,507Kov´acs G., Zucker S., Mazeh T., 2002, A&A, 391, 369Kumar S. S., 1963, ApJ, 137, 1121Ma B., Ge J., 2014, MNRAS, 439, 2781Maksimov A. F., Balega Y. Y., Dyachenko V. V., MalogolovetsMNRAS000
Anderson D. R., et al., 2011, ApJ, 726, L19Armitage P. J., Bonnell I. A., 2002, MNRAS, 330, L11Astropy Collaboration et al., 2013, A&A, 558, A33Bailer-Jones C. A. L., Rybizki J., Fouesneau M., Mantelet G.,Andrae R., 2018, AJ, 156, 58Baraffe I., Chabrier G., Allard F., Hauschildt P. H., 2002, A&A,382, 563Baranne A., et al., 1996, A&AS, 119, 373Barclay T., Pepper J., Quintana E. V., 2018, ApJS, 239, 2Barkaoui K., Gillon M., Benkhaldoun Z., Emmanuel J., Elhalk-ouj T., Daassou A., Burdanov A., Delrez L., 2017, in Jour-nal of Physics Conference Series. p. 012073, doi:10.1088/1742-6596/869/1/012073Berdnikov L. N., Belinskii A. A., Shatskii N. I., Burlak M. A.,Ikonnikova N. P., Mishin E. O., Cheryasov D. V., Zhuiko S. V.,2020, Astronomy Reports, 64, 310Boisse I., et al., 2010, A&A, 523, A88Bonfanti A., Gillon M., 2020, A&A, 635, A6Bonfanti A., Ortolani S., Piotto G., Nascimbeni V., 2015, A&A,575, A18Bonfanti A., Ortolani S., Nascimbeni V., 2016, A&A, 585, A5Bonnarel F., et al., 2000, A&AS, 143, 33 Borucki W. J., et al., 2010, Science, 327, 977Boss A. P., et al., 2007, Transactions of the International Astro-nomical Union, Series A, 26A, 183Bossini D., et al., 2019, A&A, 623, A108Bouchy F., et al., 2009, A&A, 505, 853Burdanov A. Y., et al., 2016, MNRAS, 461, 3854Burdanov A., et al., 2018, PASP, 130, 074401Burrows A., et al., 1997, ApJ, 491, 856Cantat-Gaudin T., Anders F., 2020, A&A, 633, A99Cantat-Gaudin T., et al., 2018, A&A, 618, A93Carmichael T. W., Latham D. W., Vand erburg A. M., 2019, AJ,158, 38Carmichael T. W., et al., 2020, arXiv e-prints, p.arXiv:2002.01943Castelli F., Kurucz R. L., 2003, in Piskunov N., Weiss W. W.,Gray D. F., eds, IAU Symposium Vol. 210, Modelling of Stel-lar Atmospheres. p. A20 ( arXiv:astro-ph/0405087 )Chabrier G., Baraffe I., Allard F., Hauschildt P., 2000, ApJ, 542,464Chabrier G., Johansen A., Janson M., Rafikov R., 2014, inBeuther H., Klessen R. S., Dullemond C. P., Henning T.,eds, Protostars and Planets VI. p. 619 ( arXiv:1401.7559 ),doi:10.2458/azu uapress 9780816531240-ch027Chen B. Q., et al., 2019, MNRAS, 483, 4277Claret A., Hauschildt P. H., Witte S., 2012, A&A, 546, A14Cutri R. M., et al. 2012, VizieR Online Data Catalog, p. II/311Cutri R. M., et al., 2003, VizieR Online Data Catalog, 2246Eastman J., Siverd R., Gaudi B. S., 2010, PASP, 122, 935Frolov V. N., Ananjevskaja J. K., Jilinski E. G., Gorshanov D. L.,Bronnikova N. M., 2006, A&A, 451, 901Gaia Collaboration et al., 2018, A&A, 616, A1Gelman A., Rubin D. B., 1992, Statistical Science, 7, 457Gillon M., Jehin E., Magain P., Chantry V., Hutsem´ekers D.,Manfroid J., Queloz D., Udry S., 2011, in European Physi-cal Journal Web of Conferences. p. 06002 ( arXiv:1101.5807 ),doi:10.1051/epjconf/20101106002Gillon M., et al., 2012, A&A, 542, A4Gorbovskoy E. S., et al., 2013, Astronomy Reports, 57, 233Grether D., Lineweaver C. H., 2006, ApJ, 640, 1051Halbwachs J. L., Arenou F., Mayor M., Udry S., Queloz D., 2000,A&A, 355, 581Hayashi C., Nakano T., 1963, Progress of Theoretical Physics, 30,460H´ebrard G., et al., 2011, A&A, 533, A130Henden A. A., Templeton M., Terrell D., Smith T. C., Levine S.,Welch D., 2016, VizieR Online Data Catalog, 2336Hippke M., Heller R., 2019, A&A, 623, A39Hodˇzi´c et al., 2018, MNRAS, 481, 5091Howell S. B., et al., 2014, PASP, 126, 398Huang C. X., et al., 2018, ApJ, 868, L39Jackman J. A. G., et al., 2019, MNRAS, 489, 5146Jehin E., et al., 2011, The Messenger, 145, 2Jenkins J. M., et al., 2016, in Proc. SPIE. p. 99133E,doi:10.1117/12.2233418Kharchenko N. V., Piskunov A. E., Roeser S., Schilbach E.,Scholz R. D., 2012a, VizieR Online Data Catalog, ppJ/A+A/543/A156Kharchenko N. V., Piskunov A. E., Schilbach E., R¨oser S., ScholzR. D., 2012b, A&A, 543, A156Kharchenko N. V., Piskunov A. E., Schilbach E., R¨oser S., ScholzR. D., 2013, A&A, 558, A53Kiefer F., et al., 2019, A&A, 631, A125Konacki M., Torres G., Jha S., Sasselov D. D., 2003, Nature, 421,507Kov´acs G., Zucker S., Mazeh T., 2002, A&A, 391, 369Kumar S. S., 1963, ApJ, 137, 1121Ma B., Ge J., 2014, MNRAS, 439, 2781Maksimov A. F., Balega Y. Y., Dyachenko V. V., MalogolovetsMNRAS000 , 1–13 (2020) Benni et al.
E. V., Rastegaev D. A., Semernikov E. A., 2009, AstrophysicalBulletin, 64, 296Mandel K., Agol E., 2002, ApJ, 580, L171Marcy G. W., Butler R. P., 2000, PASP, 112, 137Marigo P., et al., 2017, ApJ, 835, 77Murray C. D., Correia A. C. M., 2010, Keplerian Orbits and Dy-namics of Exoplanets. pp 15–23P¨atzold M., Rauer H., 2002, ApJ, 568, L117Pepe F., Mayor M., Galland F., Naef D., Queloz D., Santos N. C.,Udry S., Burnet M., 2002, A&A, 388, 632Perruchot S., et al., 2008, The SOPHIE spectrograph: design andtechnical key-points for high throughput and high stability.p. 70140J, doi:10.1117/12.787379Phillips M. W., et al., 2020, A&A, 637, A38Pollacco D. L., et al., 2006, PASP, 118, 1407Pont F., et al., 2008, A&A, 487, 749Pozuelos F. J., et al., 2020, arXiv e-prints, p. arXiv:2006.09403Price-Whelan A. M., et al., 2018, AJ, 156, 123Rauer H., et al., 2014, Experimental Astronomy, 38, 249Ricker G. R., et al., 2014, in Space Telescopes and Instrumenta-tion 2014: Optical, Infrared, and Millimeter Wave. p. 914320( arXiv:1406.0151 ), doi:10.1117/12.2063489Sahlmann J., et al., 2011, A&A, 525, A95Schneider J., Dedieu C., Le Sidaner P., Savalle R., Zolotukhin I.,2011, A&A, 532, A79Siverd R. J., et al., 2012, ApJ, 761, 123Sokov E. N., et al., 2018, MNRAS, 480, 291Spiegel D. S., Burrows A., Milsom J. A., 2011, ApJ, 727, 57Tody D., 1986, in Crawford D. L., ed., Proceedings of the Meet-ing, Tucson, AZ, March 4-8, 1986 Vol. 627, Instrumentationin astronomy VI. Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series, Bellingham, WA, p. 733Triaud A. H. M. J., 2018, The Rossiter-McLaughlin Effect in Ex-oplanet Research. p. 2, doi:10.1007/978-3-319-55333-7 2Udalski A., 2003, Acta Astron., 53, 291Udalski A., Szyma´nski M. K., Szyma´nski G., 2015, Acta Astron.,65, 1Vanderburg A., et al., 2019, ApJ, 881, L19Wilson P. A., et al., 2016, A&A, 588, A144Zhou G., et al., 2019, AJ, 157, 31ˇSubjak J., et al., 2020, AJ, 159, 151
Affiliations Acton Sky Portal (Private Observatory), Acton, MA,USA Department of Earth, Atmospheric and Planetary Sciences,Massachusetts Institute of Technology, 77 MassachusettsAvenue, Cambridge, MA 02139, USA Instituto de Astrof´ısica de Canarias, V´ıa L´actea s/n, 38205La Laguna, Tenerife, Spain Laboratory of Astrochemical Research, Ural Federal Uni-versity, Ekaterinburg, Russia, ul. Mira d. 19, Yekaterinburg,Russia, 620002 Space Research Institute, Austrian Academy of Sciences,Schmiedlstrasse 6, A-8042 Graz, Austria Space sciences, Technologies and Astrophysics Research(STAR) Institute, Universit´e de Li`ege, All´ee du 6 Aoˆut 17,4000 Li`ege, Belgium Institut d’Astrophysique de Paris, UMR7095 CNRS,Universit´e Pierre & Marie Curie, 98bis boulevard Arago,75014 Paris, France Observatoire de Haute-Provence, CNRS, Universit´e d’Aix-Marseille, 04870 Saint-Michel-l’Observatoire, France Universit´e Grenoble Alpes, CNRS, IPAG, 38000 Grenoble,France Instituto de Astrof´ısica e Ciˆencias do Espa¸co, Universi-dade do Porto, CAUP, Rua das Estrelas, 4150-762 Porto,Portugal INAF – Osservatorio Astrofisico di Arcetri, Largo E.Fermi 5, 50125, Firenze, Italy Department of Physics, Lehigh University, 16 MemorialDrive East, Bethlehem, PA 18015, USA Department of Physics and Astronomy, Vanderbilt Uni-versity, 6301 Stevenson Center Ln., Nashville, TN 37235,USA Department of Astronomy, The University of Texas atAustin, Austin, TX 78712, USA NASA Sagan Fellow Sternberg Astronomical Institute, M.V. LomonosovMoscow State University, 13, Universitetskij pr., 119234Moscow, Russia Faculty of Physics, M.V. Lomonosov Moscow StateUniversity, Leninskie Gory, 1, 119991, Moscow, Russia Astrobiology Research Unit, Universit´e de Li`ege, All´ee du6 Aoˆut 19C, 4000 Li`ege, Belgium Oukaimeden Observatory, High Energy Physics and As-trophysics Laboratory, Cadi Ayyad University, Marrakech,Morocco National Youth Space Center, Goheung, Jeollanam-do,59567, S. Korea Federal State Budget Scientific Institution Crimean As-trophysical Observatory of RAS, Nauchny, 298409, Crimea,Russia Special Astrophysical Observatory, Russian Academy ofSciences, Nizhnij Arkhyz, Russia, 369167 Saint Petersburg State University, Faculty of Mathemat-ics & Mechanics, Universitetskij pr. 28, Petrodvorets, St.Petersburg 198504, Russia Ural Federal University, 620002, Mira Street, 19, Yekater-inburg, Russian Federation Centre for Exoplanets and Habitability, University ofWarwick, Gibbet Hill Road, Coventry CV4 7AL, UnitedKingdom Department of Physics, University of Warwick, GibbetHill Road, Coventry CV4 7AL, United Kingdom Department of Physics, and Kavli Institute for Astro-physics and Space Research, Massachusetts Institute ofTechnology, Cambridge, MA 02139, USA Center for Astrophysics | Harvard & Smithsonian, 60Garden St., Cambridge, MA 02138, USA Department of Aeronautical and Astronautical Engineer-ing, Massachusetts Institute of Technology, Cambridge,MA, 02139 NASA Ames Research Center, Moffett Field, CA 94035,USA Central Astronomical Observatory at Pulkovo of RussianAcademy of Sciences, Pulkovskoje shosse d. 65, St. Peters-burg, Russia, 196140 Astronomical Observatory - DSFTA, University of Siena,Via Roma 56, 53100 Siena, Italy Wild Boar Remote Observatory, San Casciano in Val diPesa (FI), Italy Ankara University, Faculty of Science, Department of As-tronomy and Space Science, TR-06100 Tandogan, Ankara,Turkey Irkutsk State University, K. Marx str., 1, Irkutsk, 664003,Russia
MNRAS , 1–13 (2020) iscovery of GPX-1 b Institute of Theoretical Physics and Astronomy, VilniusUniversity, Saul˙etekio al. 3, Vilnius, LT-10257, Lithuania Taurus Hill Observatory, H¨ark¨am¨aentie 88, 79480Varkaus, Finland Physics and Engineering Physics Department, Universityof Saskatchewan, Saskatoon, SK, Canada, S7N 5E2 Baronnies Proven¸cales Observatory, Hautes Alpes -Parc Naturel R´egional des Baronnies Proven¸cales, 05150Moydans, France GJP private observatory, Elgin, OR, USA Rasteau Observatory, 84110 Rasteau, France Anunaki Observatory, Manzanares El Real, Spain Grand-Pra private observatory, Switzerland Observatory Ca lˆa ˘A´ZOu, Sant Mart´ı Sesgueioles, Spain Observatori Montcabrer, Spain Department of Physics and Astronomy, Brigham YoungUniversity, Provo, UT 84602 USA Center for Exoplanets and Habitable Worlds, The Penn-sylvania State University, 525 Davey Lab, University Park,PA 16802, USA Department of Astronomy & Astrophysics, The Pennsyl-vania State University, 525 Davey Lab, University Park, PA16802, USA Eberly Research Fellow Swarthmore College Dept. of Physics & Astronomy, 500College Ave., Swarthmore PA 19081 USA NASA Goddard Space Flight Center, 8800 Greenbelt Rd,Greenbelt, MD 20771 Space Telescope Science Institute, 3700 San Martin Drive,Baltimore, MD, 21218, USA
This paper has been typeset from a TEX/L A TEX file prepared bythe author.MNRAS000