Two Massive Jupiters in Eccentric Orbits from the TESS Full Frame Images
Mma Ikwut-Ukwa, Joseph E. Rodriguez, Samuel N. Quinn, George Zhou, Andrew Vanderburg, Asma Ali, Katya Bunten, B. Scott Gaudi, David W. Latham, Steve B. Howell, Chelsea X. Huang, Allyson Bieryla, Karen A. Collins, Theron W. Carmichael, Markus Rabus, Jason D. Eastman, Kevin I. Collins, Thiam-Guan Tan, Richard P. Schwarz, Gordon Myers, Chris Stockdale, John F. Kielkopf, Don J. Radford, Ryan J. Oelkers, Jon M. Jenkins, George R. Ricker, Sara Seager, Roland K. Vanderspek, Joshua N. Winn, Jennifer Burt, R. Paul Butler, Michael L. Calkins, Jeffrey D. Crane, Crystal L. Gnilka, Gilbert A. Esquerdo, Wlliam Fong, Laura Kreidberg, Jessica Mink, David R. Rodriguez, Joshua E. Schlieder, Stephen Schectman, Avi Shporer, Johanna Teske, Eric B. Ting, Jesus Noel Villasenor, Daniel A. Yahalomi
DD RAFT VERSION F EBRUARY
5, 2021Typeset using L A TEX twocolumn style in AASTeX63
Two Massive Jupiters in Eccentric Orbits from the TESS Full Frame Images M MA I KWUT -U KWA , J OSEPH
E. R
ODRIGUEZ ,
2, 1 S AMUEL
N. Q
UINN , G EORGE Z HOU , A NDREW V ANDERBURG , A SMA A LI ,
1, 4 K ATYA B UNTEN ,
1, 5
B. S
COTT G AUDI , D AVID
W. L
ATHAM , S TEVE
B. H
OWELL , C HELSEA
X. H
UANG , ∗ A LLYSON B IERYLA , K AREN
A. C
OLLINS , T HERON
W. C
ARMICHAEL ,
9, 1 M ARKUS R ABUS ,
10, 11, 12 J ASON
D. E
ASTMAN , K EVIN
I. C
OLLINS , T HIAM -G UAN T AN , R ICHARD
P. S
CHWARZ , G ORDON M YERS , C HRIS S TOCKDALE , J OHN
F. K
IELKOPF , D ON J. R
ADFORD , R YAN
J. O
ELKERS , J ON M. J
ENKINS , G EORGE
R. R
ICKER , S ARA S EAGER ,
8, 21, 22 R OLAND
K. V
ANDERSPEK , J OSHUA
N. W
INN , J ENNIFER B URT , R. P
AUL B UTLER , M ICHAEL
L. C
ALKINS , J EFFREY
D. C
RANE , C RYSTAL
L. G
NILKA , G ILBERT
A. E
SQUERDO , W ILLIAM F ONG , L AURA K REIDBERG , J ESSICA M INK , D AVID
R. R
ODRIGUEZ , J OSHUA
E. S
CHLIEDER ,
28, 29 S TEPHEN S HECTMAN , A VI S HPORER , J OHANNA T ESKE , E RIC
B. T
ING , J ESUS N OEL V ILLASEÑOR , D ANIEL
A. Y
AHALOMI
30, 1 Center for Astrophysics | Harvard & Smithsonian, 60 Garden St, Cambridge, MA 02138, USA Department of Physics and Astronomy, Michigan State University, East Lansing, MI 48824, USA Department of Astronomy, University of Wisconsin-Madison, Madison, WI 53706, USA Algonquin Regional High School, MA, USA Cambridge Rindge and Latin High School, MA, USA Department of Astronomy, The Ohio State University, 140 West 18th Avenue, Columbus, OH 43210, USA NASA Ames Research Center, Moffett Field, CA, 94035, USA Department of Physics and Kavli Institute for Astrophysics and Space Research, Massachusetts Institute of Technology, Cambridge, MA 02139, USA Harvard University, Cambridge, MA 02138, USA Las Cumbres Observatory Global Telescope, 6740 Cortona Dr., Suite 102, Goleta, CA 93111, USA Department of Physics, University of California, Santa Barbara, CA 93106-9530, USA Departamento de Matemática y Fı’sica Aplicadas, Universidad Católica de la Santísima Concepción, Alonso de Rivera 2850, Concepción, Chile George Mason University, 4400 University Drive MS 3F3, Fairfax, VA 22030, USA Perth Exoplanet Survey Telescope, Perth, Australia Patashnick Voorheesville Observatory, Voorheesville, NY 12186, USA American Association of Variable Star Observers, 49 Bay State Road, Cambridge, MA 02138, USA Hazelwood Observatory, Churchill, Victoria, Australia Department of Physics and Astronomy, University of Louisville, Louisville, KY 40292, USA Brierfield Observatory, New South Wales, Australia Department of Physics and Astronomy, Vanderbilt University, Nashville, TN 37235, USA Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, USA Department of Aeronautics and Astronautics, MIT, 77 Massachusetts Avenue, Cambridge, MA 02139, USA Department of Astrophysical Sciences, Princeton University, 4 Ivy Lane, Princeton, NJ, 08544, USA Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA 91109, USA Earth & Planets Laboratory, Carnegie Institution for Science, 5241 Broad Branch Road, NW, Washington, DC 20015, USA The Observatories of the Carnegie Institution for Science, 813 Santa Barbara St., Pasadena, CA 91101, USA Space Telescope Science Institute, Baltimore, MD 21218, USA Exoplanets and Stellar Astrophysics Laboratory, Code 667, NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA GSFC Sellers Exoplanet Environments Collaboration, NASA Goddard Space Flight Center, Greenbelt, MD 20771 Department of Astronomy, Columbia University, 550 West 120th Street, New York, NY 10027, USA
ABSTRACTWe report the discovery of two short-period massive giant planets from NASA’s Transiting Exoplanet SurveySatellite (TESS). Both systems, TOI-558 (TIC 207110080) and TOI-559 (TIC 209459275), were identified fromthe 30-minute cadence Full Frame Images and confirmed using ground-based photometric and spectroscopicfollow-up observations from TESS’s Follow-up Observing Program Working Group. We find that TOI-558 b,
Corresponding author: Mma [email protected] a r X i v : . [ a s t r o - ph . E P ] F e b I KWUT -U KWA ET AL .which transits an F-dwarf ( M ∗ =1.349 + − M (cid:12) , R ∗ =1.496 + − R (cid:12) , T eff =6466 + − K, age 1.79 + − Gyr) withan orbital period of 14.574 days, has a mass of 3.61 ± M J , a radius of 1.086 + − R J , and an eccentric(e=0.300 + − ) orbit. TOI-559 b transits a G-dwarf ( M ∗ =1.026 ± M (cid:12) , R ∗ =1.233 + − R (cid:12) , T eff =5925 + − K, age 1.79 + − Gyr) in an eccentric (e=0.151 ± + − M J and a radiusof 1.091 + − R J . Our spectroscopic follow-up also reveals a long-term radial velocity trend for TOI-559,indicating a long-period companion. The statistically significant orbital eccentricity measured for each systemsuggests that these planets migrated to their current location through dynamical interactions. Interestingly, bothplanets are also massive ( > M J ), adding to the population of massive hot Jupiters identified by TESS. Promptedby these new detections of high-mass planets, we analyzed the known mass distribution of hot Jupiters but findno significant evidence for multiple populations. TESS should provide a near magnitude-limited sample oftransiting hot Jupiters, allowing for future detailed population studies. INTRODUCTIONThe formation and migration of giant planets in close or-bits has been debated extensively. Hot Jupiters (with or-bital periods less than 10 days) could theoretically form ina number of ways, with three main formation and migrationschemes dominating the literature. It has traditionally beenthought that short-period giant planets must form at largerorbital radii and migrate inwards over time (Lin et al. 1996;Rafikov 2006; Dawson & Johnson 2018). In order for the for-mation outcome to be a giant planet, the core needs to formrapidly enough to accrete gas within the lifetime of the proto-planetary disk (Bodenheimer & Pollack 1986). Core accre-tion theories suggest that this atmospheric accretion can onlyoccur in a region of the disk where the core can coalesceenough material to grow to ∼
10 Earth masses–this criticalmass declines as a function of semi-major axis (Piso et al.2015). This assumes that the mass of the gaseous envelopebecomes greater than the mass of the core (Pollack et al.1996). After formation, a giant planet could migrate to aclose-in orbit through either gentle migration through the gasdisk (Goldreich & Tremaine 1980; Lin & Papaloizou 1986;Lin et al. 1996) or more dynamical migration caused by in-teraction with another planet or star (Rasio & Ford 1996; Wu& Murray 2003; Fabrycky & Tremaine 2007; Nagasawa &Ida 2011; Wu & Lithwick 2011), after which the planet’s or-bit could be circularized and shrunk by tidal forces (Naozet al. 2011; Beaugé & Nesvorný 2012). However, more re-cent models have suggested that hot Jupiters may also formin-situ (Batygin et al. 2016) and show that the period-massdistribution and inner boundary of short-period giant plan-ets could be consistent with predictions for in-situ formation(Bailey & Batygin 2018). Other efforts have shown this massdistribution of giant planets to be consistent with high ec-centricity migration from dynamical interaction (Matsakos &Königl 2016). The dominance of each of these three forma-tion and migration scenarios remains an open question, and itis likely that a combination of these methods have shaped thehot Jupiter population seen today. Atmospheric characteriza- ∗ Juan Carlos Torres Fellow tion is one frontier that may constrain hot Jupiter migration;the measurement of carbon and oxygen abundances in hotJupiters can be used to trace migration histories (Madhusud-han et al. 2014).The discovery of very massive giant planets (> 6 M J ), has raised the question of whether there are meaningful massboundaries separating giant planets, brown dwarfs, and low-mass stars–specifically, whether there is a particular massrange in which the dominant formation mechanism changesfrom core accretion to gravitational instability and fragmen-tation of giant molecular gas clouds. Some studies (e.g.Schlaufman 2018; Moe & Kratter 2019) have argued thatcore accretion is the dominant formation mechanism for gi-ant planet companions with masses M P < M J . Addition-ally, Schlaufman (2018) notes that higher host star metallicityis the property associated with core accretion and may indi-cate that M P < M J giant planets may preferentially formvia core accretion around metal-rich stars. There also existsa gap in the mass distribution of giant planets very near thethreshold of M P = M J that Moe & Kratter (2019) claim tobe a feasible lower mass boundary for disk fragmentation toform an object. Moe & Kratter (2019) also highlight that rel-atively metal-poor host stars seem to preferentially host ob-jects at masses at and above this M P = M J threshold. Thediscovery and characterization of massive giant planets andlow-mass brown dwarfs may enable a better understandingof the transition between these formation mechanisms.The observed parameters of a planet and its orbit may beindicative of its formation and migration mechanism. Onepossible path to determining the dominant mechanism ofgiant planet migration is to create a complete sample ofhot Jupiters with well characterized fundamental parame-ters (masses and radii, and orbital periods and eccentricities).Statistical population studies of such a sample may provideinsight into the dominant evolutionary pathways for giantplanets; this type of analysis led to the discovery of the radiusvalley in small planets (Fulton et al. 2017; Fulton & Petigura2018), supporting the prediction due to photo-evaporation of exoplanetarchive.ipac.caltech.edu/ OI-558 B & 559 B Table 1.
Literature and Measured Properties for TOI-558 and TOI-559Other identifiers TOI-558 TOI-559TIC 207110080 TIC 209459275TYC 8497-00028-1 TYC 7019-00191-1Parameter Description Value Value Source α J . . . . . . . . Right Ascension (RA) . . . . . . . . . . . . . . . 02:49:09.9601 03:07:16.4958 1 δ J . . . . . . . . . Declination (Dec) . . . . . . . . . . . . . . . . . . . -58:01:28.9180 -31:09:45.7019 1B T . . . . . . . . . . . Tycho B T mag. . . . . . . . . . . . . . . . . . . . . . 12.049 ± ± T . . . . . . . . . . . Tycho V T mag. . . . . . . . . . . . . . . . . . . . . . 11.309 ± ± G mag. . . . . . . . . . . . . . . . . . . . . . . . . 11.33 ± ± P . . . . . . . . . . . . Gaia B P mag. . . . . . . . . . . . . . . . . . . . . . . . 11.58 ± ± P . . . . . . . . . . . . Gaia R P mag. . . . . . . . . . . . . . . . . . . . . . . . 11.576 ± ± ± ± ± ± ± ± S . . . . . . . . . . . . 2MASS K S mag. . . . . . . . . . . . . . . . . . . . . 10.262 ± ± WISE1 . . . . . . . .
WISE1 mag. . . . . . . . . . . . . . . . . . . . . . . . . 10.216 ± ± WISE2 . . . . . . . .
WISE2 mag. . . . . . . . . . . . . . . . . . . . . . . . . 10.248 ± ± WISE3 . . . . . . . .
WISE3 mag. . . . . . . . . . . . . . . . . . . . . . . . . 10.236 ± ± WISE4 . . . . . . . .
WISE4 mag. . . . . . . . . . . . . . . . . . . . . . . . . — 9.33 ± µ α . . . . . . . . . . . Gaia DR2 proper motion . . . . . . . . . . . . . 1.071 ± ± − ) µ δ . . . . . . . . . . . . Gaia DR2 proper motion . . . . . . . . . . . . . 3.859 ± ± − ) π † . . . . . . . . . . . . Gaia Parallax (mas) . . . . . . . . . . . . . . . . . 2.4850 ± † ± v sin i (cid:63) . . . . . . . . . Rotational velocity ( km s − ) . . . . . . . 7.8 ± ± mac . . . . . . . . . . macroturbulent broadening ( km s − ) 5.9 ± ± U ∗ . . . . . . . . . . . Space Velocity ( km s − ) . . . . . . . . . . . . . 3.8 ± ± V . . . . . . . . . . . . . Space Velocity ( km s − ) . . . . . . . . . . . . . − ± − ± W . . . . . . . . . . . . Space Velocity ( km s − ) . . . . . . . . . . . . . − ± ± NOTES: † Values have been corrected for the 30 µ as offset as reported by Lindegren et al. (2018). ∗ U is in the direction of the Galactic center.See §D in the appendix of Collins et al. (2017) for a description of each detrending parameter.Sources are: Gaia Collaboration et al. (2018), Høg et al. (2000), Stassun et al. (2018), Cutri et al. (2003), Zacharias et al. (2017) volatiles (Yelle 2004; Tian et al. 2005; Murray-Clay et al.2009; Owen & Jackson 2012; Lopez & Fortney 2013).NASA’s Transiting Exoplanet Survey Satellite (
TESS ),launched in 2018, is an all-sky photometric survey with thegoal of discovering thousands of new planets around bright,nearby stars (Ricker et al. 2015). The
TESS mission hasalready discovered over a dozen new hot Jupiters, includ-ing a few massive systems ( > M J Rodriguez et al. 2019a;Nielsen et al. 2020; Rodriguez et al. 2021), and is expected to be largely complete for giant planets with periods up to 10days around bright stars (Zhou et al. 2019). Detailed charac-terization of new discoveries from
TESS will help completethe sample of known short-period giant planets, setting thefoundation for more robust population studies.In this paper, we confirm and characterize two short-periodgiant planets from
TESS , TOI-558 b and TOI-559 b. Wepresent the photometric and spectroscopic observations from
TESS and ground-based facilities in §2, which we globally I
KWUT -U KWA ET AL .model using
EXOFASTv2 (Eastman et al. 2019) in §3. Fur-ther, we examine the existing population of hot Jupiters,studying existing trends in the mass-period distribution anddiscussing the contribution of
TESS discoveries (§4). Ourconclusions are summarized in §5. OBSERVATIONS AND ARCHIVAL DATAWe confirm and characterize TOI-558 and TOI-559 asplanetary systems using combined
TESS observations withground-based photometric and spectroscopic follow-up ob-servations from the
TESS
Follow-up Observing Program(TFOP) Working Group. Table 1 provides a list of the lit-erature identifiers, magnitudes, and kinematics for TOI-558and TOI-599. 2.1.
TESS Photometry
In the two-year primary mission,
TESS completed 26 ob-servation sectors, each of approximate length ∼
27 days, cov-ering the southern hemisphere in the first year-long cycle andthe northern hemisphere in the second (Ricker et al. 2015).
TESS recently began its first extended mission with a similarobservation footprint that will cover over 90% of the sky intotal, including a large part of the ecliptic plane. As of UT2021 January 1,
TESS has yielded 91 confirmed planets and2440 planet candidates, or
TESS
Objects of Interest (TOIs) . TESS used four wide-field cameras, each with an f/1.4aperture, 21 arcsecond pixel scale, and field of view of24 ◦ × ◦ , comprising a total field of view of 24 ◦ × ◦ foreach observing sector. Neither TOI-558 nor TOI-559 wasone of the stars pre-selected for short-cadence observationsduring the prime mission; both were observed only in theFull Frame Images (FFIs), which cover the entire field ofview at a 30-minute cadence. TOI-558 (TIC 207110080) wasobserved by Camera 3 in both Sector 2, from UT 2018 Au-gust 22 to UT 2018 September 20, and Sector 3, from UT2018 September 20 to UT 2018 October 18, during TESS ’sfirst year of the primary mission. TOI-559 (TIC 209459275)was observed by Camera 2 in Sector 4, from UT 2018 Oc-tober 18 to UT 2018 November 15. We identified TOI-558and TOI-559 as planet candidates through a search indepen-dent of the
TESS planet search pipeline, using a standardBox Least Squares algorithm (Kovács et al. 2002) and visualexamination of candidates, and both candidates were desig-nated as pre-selected targets for Cycle 3. TOI-558 was thenreobserved by
TESS again during Cycle 3, the first year of theextended mission, in Sector 29 (UT 2020 August 26 to UT2020 September 22) and Sector 30 (UT 2020 September 22to UT 2020 October 21) at a cadence of 2 minutes. TOI-559was reobserved during Sector 31 from UT 2020 October 21to UT 2020 November 19. The 2-minute observations were https://tev.mit.edu/data/collection/193/ inspected by the SPOC team and did not indicate a false pos-itive transit detection (Twicken et al. 2018; Li et al. 2019).We extracted and processed light curves from the FFIs us-ing Tesscut and the
Lightkurve package for Python(Lightkurve Collaboration et al. 2018; Brasseur et al. 2019).The
TESS
Science Processing Operations Center (SPOC)(Jenkins et al. 2016) at NASA Ames Research Center pro-cessed the raw FFIs through a pipeline that calibrated the pix-els and mapped world coordinate system (WCS) informationfor each image frame. Our selected apertures included pixelswith a mean flux of 80th percentile or greater within a 3-pixelradius of the target’s center. We subtracted background scat-tered light and deblended contamination from nearby starsusing a simple target star model. We removed spacecraft sys-tematics by decorrelating against the scattered backgroundlight and the standard deviation of the quaternion time seriesfollowing Vanderburg et al. (2019). We performed the decor-relation using
Lightkurve ’s RegressionCorrector utility. We used the spline-fitting routine
Keplerspline (Van-derburg & Johnson 2014; Shallue & Vanderburg 2018) onthese light curves to remove any remaining stellar variability,resulting in a flattened light curve. For TOI-559, the baselinefluxes observed during the two orbits of the TESS spacecraftin Sector 4 had a significant offset, so we detrended the twoorbits separately. We omitted from further consideration allof the data obtained long before or after a transit, leavingroughly one full transit duration prior to each ingress and af-ter each egress (including the full transit). These light curveswere then used for the global modeling described in §3.The 2-minute cadence
TESS light curve for TOI-558 insectors 29 and 30 was extracted by the Science Process-ing Operations Center (SPOC) pipeline, based at the NASAAmes Research Center (Jenkins et al. 2016). Specifically, thedata were downloaded, reduced and analyzed by the SPOCpipeline, which included pixel-level calibrations, optimiza-tion of photometric aperture, estimation of the total flux con-tamination from other nearby stars, and extraction of the lightcurve. To remove systematics and instrumental artifacts,the Presearch Data Conditioning (PDC, Smith et al. 2012;Stumpe et al. 2014) module was applied to the extractedSPOC light curve. The resulting processed light curve wasrun through the SPOC Transiting Planet Search (TPS, Jenk-ins 2002) to identify any known or additional planet candi-dates. To remove any remaining low-frequency out-of-transitastrophysical or instrumental variability in the light curves,we use
Keplerspline . We simultaneously fit the spline witha transit model to ensure that the transits were not distortedby the removal of low-frequency variability (see Vanderburget al. 2016 and Pepper et al. 2019) https://github.com/avanderburg/keplersplinev2 OI-558 B & 559 B Figure 1.
The full corrected light curves from TESS. The discovery light curves (left) extracted from the Full Frame Images are at a 30-minutecadence, from Sectors 2 and 3 for TOI-558 and Sector 4 for TOI-559. The additional light curves (right) extracted by the SPOC pipeline are at2-minute cadence, from Sectors 29-31 of the first extended mission (see §2.1).
Table 2.
Ground-based photometry observations from TFOP for TOI-558 and TOI-559 used in the global analysis.Date (UT) Facility size (m) Filter FOV Pixel Scale Exp (s) Additive Detrending
TOI-558 (cid:48) (cid:48) × (cid:48) (cid:48)(cid:48)
55 airmass, Width T12019 Oct. 26 LCO CTIO 1m B (cid:48) × (cid:48) (cid:48)(cid:48)
34 airmass2020 Nov. 09 LCO CTIO 1m i 27 (cid:48) × (cid:48) (cid:48)(cid:48)
25 airmass
TOI-559 (cid:48) × (cid:48) (cid:48)(cid:48)
60 None2019 Oct. 18 LCO SSO 1m Sloan z (cid:48) (cid:48) × (cid:48) (cid:48)(cid:48)
35 airmass2020 Aug. 20 LCO SSO 1m Sloan i (cid:48) (cid:48) × (cid:48) (cid:48)(cid:48)
25 none2020 Aug. 27 LCO SSO 1m Sloan i (cid:48) (cid:48) × (cid:48) (cid:48)(cid:48)
25 airmass, sky/pixel T1
Ground-based Photometry from the TESS Follow-upObserving Program Working Group
To rule out any astrophysical false positives or systemat-ics causing the transit events and to refine the timing andtransit parameters, we obtained photometric transit follow-upfrom ground-based telescopes. The
TESS
Follow-up Observ- ing Program (TFOP) Sub Group 1 (SG1), which specializesin ground-based time-series photometry, observed transits ofboth TOI-558 and TOI-559 with the Las Cumbres Observa-tory Global Telescope (LCOGT) network of 1-meter tele- https://tess.mit.edu/followup/ I KWUT -U KWA ET AL . Figure 2.
The phase-folded transit light curves for (left) TOI-558 and (right) TOI-559 from
TESS and the TFOP working group. The solidcolored lines correspond to the best-fit model from our global fit (see §3). scopes (Brown et al. 2013) and the Perth Exoplanet Sur-vey Telescope (PEST) . The observations were scheduledusing the TAPIR software package (Jensen 2013), and allobservations but the ones taken by PEST were reduced andlightcurves were extracted using
AstroImageJ (Collinset al. 2017). PEST uses a custom software suite to reducethe images and extract light curves, the
PEST Pipeline .These transit observations and facilities are listed in Table 2.These observations not only extended the baseline, but alsoprovided an independent check on the depth and duration ofthe transit as compared to what was observed by TESS.2.3. TRES Spectroscopy (TOI-559)
Reconnaissance spectroscopic follow-up observations ofTOI-559 were taken on three separate epochs at a resolv- https://lco.global/observatory/telescopes/1-m/ http://pestobservatory.com http://pestobservatory.com/the-pest-pipeline/ ing power of R ∼ . TRES islocated at the the Fred L. Whipple Observatory (FLWO) onMt. Hopkins, AZ. The reduction and RV extraction pipelinedetails are described in Buchhave et al. (2010) and Quinnet al. (2012). With only three observations, we do not includethese RVs in the global analysis (see §3). Nevertheless theextracted RVs yielded a semi-amplitude consistent with theglobal analysis. The TRES spectra were also used as an in-dependent check on the metallicity from CHIRON. Using theStellar Parameter Classification (SPC) package (Buchhaveet al. 2012), we derive a metallicity for TOI-559 of [ Fe / H ] =-0.24 ± ± PFS Spectroscopy (TOI-558) OI-558 B & 559 B -2000200 R V ( m / s ) TOI-558
TDB - 2458500-31031 O - C ( m / s ) -50005001000 R V ( m / s ) TOI-559
500 550 600 650 700 750BJD
TDB - 2458000-51051 O - C ( m / s ) -2000200 R V ( m / s ) TOI-558
PFS P - T C )/P + 0.25-31031 O - C ( m / s ) -50005001000 R V ( m / s ) TOI-559
CHIRON P - T C )/P + 0.25-51051 O - C ( m / s ) Figure 3.
Top) The radial velocity observations over time for (left) TOI-558 b from PFS and (right) TOI-559 b from CHIRON. The RVsphase-folded to the best-fit periods are shown above. The EXOFASTv2 model fit is shown in red.
TOI-558 was observed using the Planet Finder Spectro-graph (PFS) on the 6.5-meter Magellan Clay Telescope atLas Campanas Observatory in Chile (Crane et al. 2006, 2008,2010), which has been extensively used to follow up and con-firm TOIs (e.g. Teske et al. 2020). We obtained 14 radial ve-locity (RV) measurements from UT 2019 January 19 to UT2019 February 18, which are shown in Table 3. PFS is a highresolution optical (391nm to 734nm) spectrograph that uti-lizes an iodine cell to achieve highly precise RV ( < ×
10k CCD,with a 0.3 (cid:48)(cid:48) slit and resolving power of (R ∼ − precision, we chose shorter ex-posures with a typical RV precision of ∼ − since ourtargets have very large semi-amplitudes ( >
30 m s − ).We derived stellar parameters, specifically the host star’smetallicity, for TOI-558 from the iodine free template spec-trum obtained with PFS. The spectrum, in the region of 5000- 5500 Å, was analyzed with the ZASPE package (Brahmet al. 2017), which performs a model comparison betweenthe observed spectrum and a grid of the PHOENIX stellaratmospheres synthetic spectra (Husser et al. 2013). ZASPEweights spectral regions based on their pre-determined im-portance to the stellar parameter determination, and varies the depths of those spectral regions with a Monte Carlo anal-ysis to determine the uncertainties and covariance of the de-rived stellar parameters. The resulting best fit metallicity was [ Fe / H ] = -0.020 ± v sin i (cid:63) and macroturbulent broadening for TOI-558 fol-lowing the methodology in Zhou et al. (2018). We measured v sin i (cid:63) for TOI-558 to be be 4.1 ± − and v mac to be4.4 ± − .2.5. CHIRON Spectroscopy (TOI-559)
TOI-559 was observed with the CTIO High Resolutionspectrometer (CHIRON) on the CTIO 1.5-meter telescope(Tokovinin et al. 2013). CHIRON covers a wavelength rangeof 420nm to 880nm, with a resolving power of R ∼ Extracted radial velocities were obtained from standardized PFS and CHI-RON pipelines. I KWUT -U KWA ET AL . Table 3.
Radial Velocities for TOI-558 and TOI-559Target
BJD
TDB ( days ) RV ( ms − ) σ RV ( ms − ) FacilityTOI-558 2458502.57047 137.8 4.6 PFSTOI-558 2458503.60715 151.3 4.4 PFSTOI-558 2458504.61284 120.7 5.0 PFSTOI-558 2458505.57474 37.0 5.2 PFSTOI-558 2458506.56937 -93.4 4.6 PFSTOI-558 2458507.57311 -270.0 5.0 PFSTOI-558 2458508.56806 -368.9 4.7 PFSTOI-558 2458509.56851 -317.2 6.2 PFSTOI-558 2458526.58483 -115.2 5.1 PFSTOI-558 2458527.53828 -54.8 6.8 PFSTOI-558 2458528.53871 0.0 4.8 PFSTOI-558 2458529.54633 59.0 5.4 PFSTOI-558 2458531.61868 165.5 5.7 PFSTOI-558 2458532.54853 152.1 5.2 PFSTOI-559 2458510.54861 -15924.9 33.2 CHIRONTOI-559 2458511.60839 -16208.5 21.6 CHIRONTOI-559 2458512.54877 -16119.7 22.1 CHIRONTOI-559 2458526.54291 -16162.9 17.0 CHIRONTOI-559 2458527.57189 -15526.1 33.6 CHIRONTOI-559 2458529.53872 -15143.5 28.0 CHIRONTOI-559 2458531.55750 -16018.5 16.7 CHIRONTOI-559 2458537.57692 -15573.6 13.7 CHIRONTOI-559 2458539.57232 -16224.7 22.9 CHIRONTOI-559 2458541.51279 -15590.0 16.4 CHIRONTOI-559 2458542.51358 -15008.6 13.2 CHIRONTOI-559 2458543.51021 -15169.4 23.8 CHIRONTOI-559 2458550.51777 -15165.0 25.0 CHIRONTOI-559 2458551.50301 -15612.5 25.4 CHIRONTOI-559 2458552.51271 -15998.8 26.6 CHIRONTOI-559 2458553.49918 -16231.8 28.1 CHIRONTOI-559 2458554.49890 -16175.8 54.2 CHIRONTOI-559 2458741.86403 -16340.3 21.7 CHIRONTOI-559 2458742.79259 -16371.4 33.3 CHIRONTOI-559 2458743.87250 -15812.6 23.1 CHIRONTOI-559 2458744.81891 -15203.6 24.5 CHIRONTOI-559 2458745.85030 -15210.3 30.6 CHIRONTOI-559 2458746.84768 -15634.3 21.6 CHIRONTOI-559 2458511.60506 -896.80 20.02 TRESTOI-559 2458515.62512 218.75 20.29 TRESTOI-559 2458738.98177 104.78 34.18 TRES
We also use the CHIRON spectra to determine some con-straints on the host star’s metallicity and v sin i (cid:63) . The spectrawere matched against an interpolated grid of ∼
10, 0000 ob-served spectra from the TRES database, previously classifiedusing the Spectral Classification Pipeline (Buchhave et al.2012). This library is interpolated using a gradient boost clas- sifier algorithm in the scikit-learn machine learning package.The CHIRON observed spectrum is then convolved against aGaussian profile such that it matches the spectral resolutionof observations in this library ( R =
44, 000). We measure themetallicity of TOI-559 to be [ Fe / H ] = -0.22 ± T eff = 5784 ±
50 K, and the sur-face gravity to be log g = 4.18 ± v sin i (cid:63) for TOI-559 to be 7.8 ± − and v mac to be 5.9 ± − following Zhou et al. (2018).2.6. High Resolution Speckle Imaging
It is difficult to rule out the possibility of blended compan-ion stars using TESS data alone given the size of the pixels.Contamination from blended stars can cause a false positivetransit signal on the planetary candidate host star or affectthe derived planetary radius (Ciardi et al. 2015; Ziegler et al.2018). To check for very nearby stars not resolved by seeing-limited images, we obtained high-resolution speckle imagingof TOI-558 and TOI-559 from the Southern AstrophysicalResearch Telescope (SOAR) (Tokovinin 2018). TOI-558 andTOI-559 were both observed on UT 2019 February 18, andhad a sensitivity of ∆ Mag = 6.7 and 7.2 at 1 (cid:48)(cid:48) , respectively.Figure 4 displays the reconstructed images as well as the lim-iting magnitude difference versus on-sky distance from thecenter of the target star. We see no signs of any nearby close(within 3 arcseconds) companions in the SOAR observationsof TOI-558 or TOI-559. For a detailed description of the ob-serving strategy for
TESS targets see Ziegler et al. (2020).Using the ‘Alopeke instrument mounted on the 8-meterGemini North telescope (located on Mauna Kea, Hawai’i),we observed TOI-558 on UT 2020 December 23 and 29. Thefirst observation had poor seeing, so we show the December29th observation in Figure 4. TOI-559 was observed usingthe ‘Alopeke instrument on UT 2019 October 09. ‘Alopekesimultaneously observes in blue ( λ ∆ λ = 562/54 nm) and red( λ ∆ λ = 832/40 nm) band passes, with inner working anglesof 0.026 (cid:48)(cid:48) for the blue and 0.017 (cid:48)(cid:48) for the red. The instru-ment has a pixel scale of 0.01 (cid:48)(cid:48) . Three thousand 0.06-secondimages were obtained and combined for each star, and theFourier analysis described in Howell et al. (2011) was per-formed on the combined image. The ‘Alopeke observationsconfirm and extend to smaller inner working angles the re-sults seen by SOAR, in that TOI-559 is a single star with nosigns of any previously unknown companions to within the5-sigma contrast limits obtained (see Figure 4). The obser-vations had a sensitivity of ∆ mag = 5.557 for the blue and7.375 for the red, at 1 (cid:48)(cid:48) for TOI-559 and ∆ mag = 4.355 and6.394 for TOI-558.OI-558 B & 559 B arcsec m a g n i t u d e ( I - b a n d ) [arcsec] -202 [ a r c s e c ] SOAR Speckle ACF
TIC207110080 m
562 nm832 nm
TOI558 arcsec m a g n i t u d e ( I - b a n d ) [arcsec] -202 [ a r c s e c ] SOAR Speckle ACF
TIC209459275 m
562 nm832 nm
TOI559
Figure 4.
The (left) Speckle interferometric observations for TOI-558 and TOI-559 of the two targets from the Southern Astrophysical ResearchTelescope (SOAR). The autocorrelation function is shown inset the contrast curve from SOAR. The (right) Gemini ‘Alopeke speckle imaging5-sigma contrast curves are shown along with the reconstructed images (embedded) of TOI-558 and TOI-559.
Galactic Locations, Kinematics, Orbits, andPopulations
We used the parallaxes, proper motions, radial velocities,and associated uncertainties of TOI-558 and TOI-559 fromthe Gaia DR2 catalog (Gaia Collaboration et al. 2018) todetermine the location, kinematics, orbits, and associationsof each system with known stellar populations. We cor-rected the DR2 parallaxes and uncertainties following Linde-gren et al. (2018). We then used these parallaxes to estimatethe distances to the systems. These distances and their uncer-tainties where then used in combination with the DR2 propermotions and radial velocities to determine the heliocentricUVW velocities of the host stars. We determined the UVWvelocities with respect to the Local Standard of Rest (LSR)using the determination of the sun’s motion relative to the For part of this section, we follow the analysis methodology performed byBurt et al. (2020).
LSR by Co¸skunoˇglu et al. (2011). These UVW values areshown in Table 1.For each system, we estimated its Z height relative to thesun, and then corrected for the Z (cid:12) (cid:39)
30 pc offset of the sunfrom the Galactic plane as determined by Bovy (2017) basedon local giants. We use the UVW velocities (with respect tothe LSR) to estimate the likelihood that the star belongs tothin disk, thick disk, halo, or Hercules stream, using the cate-gorization criteria of Bensby et al. (2014). We use the Galac-tic orbits estimated by Mackereth & Bovy (2018), and re-port estimates of the orbital parameters (apogalacticon, peri-galaciton, eccentricity, and maximum excursion perpendicu-lar to the plane). We estimated the spectral type of each TOIusing their effective temperatures (as given in Table 4) andthe relations of Pecaut & Mamajek (2013). We then com- We adopt a coordinate system such that positive U is toward the Galacticcenter. KWUT -U KWA ET AL .pared the position and orbits of the two systems to the scaleheight h Z of stars of similar spectral type as determined byBovy (2017).We also considered whether either of the systems belongto any of the known nearby young associations using theBANYAN Σ (Bayesian Analysis for Nearby Young Associa-tioNs Σ ) tool (Gagné et al. 2018). The BANYAN Σ estimatorassigned both hosts to be ‘field’ stars.TOI-558 is at a distance of d = ± Z + Z (cid:12) (cid:39) −
291 pc. It has Galactic velocities with respect to the LSR of ( U , V , W ) = ( ± − ± − ± ) km s − . Ac-cording to the categorization of Bensby et al. (2014), the sys-tem has a ∼
98% probability of belonging to the thin disk.The Galactic orbit has a perigalacticon of R p = R a = e = Z excursion from the Galactic plane of Z max =
460 pc. Thus, the orbit is consistent with the currentlocation of the system. The scale height of stars of similarspectral type (F5.5V) is only 85 pc. Nevertheless, there is anon-negligible probability that a star belonging to this popu-lation can have a maximum excursion above the plane that isseveral scale heights.TOI-559 is at distance of d = ± Z + Z (cid:12) (cid:39) −
172 pc. It hasGalactic velocities with respect to the LSR of ( U , V , W ) =( ± − ± ± ) km s − . According tothe categorization of Bensby et al. (2014), the system has a ∼
92% probability of belonging to the thin disk, and an ∼ R p = R a = e = Z excursion from the Galactic plane of Z max =
320 pc. Thusthe orbit is consistent with the current location of the systemand with the scale height of 108 pc for stars of similar spec-tral type (G0V). Although this system has a non-negligibleprobability of belonging to the thick disk, it is neverthelessmore likely to be a member of the thin disk. We estimate anage of ∼ EXOFASTv2 GLOBAL FIT FOR TOI-558 ANDTOI-559In order to characterize the planetary systems, we mod-eled the observations obtained in §2 with
EXOFASTv2 , aglobal fitting suite for exoplanets (Eastman et al. 2013, 2019)to simultaneously fit the
TESS and TFOP SG1 photometryand the PFS and CHIRON RVs.
EXOFASTv2 uses a dif-ferential evolution Markov Chain Monte Carlo (MCMC) to simultaneously model the star and planet globally and self-consistently. For our fits of TOI-558 and TOI-559, we con-ducted a fit of the Spectral Energy Distribution (SED) ofthe host star (see Table 1 for a list of the broadband pho-tometric measurements used in the SED analysis) simulta-neously with the available radial velocities and photometry.We imposed Gaussian priors on the
Gaia parallaxes (GaiaCollaboration et al. 2018) (accounting for the 30 µ offset asreported by Lindegren et al. 2018) and the stellar metallici-ties obtained from spectroscopy ( [ Fe / H ] = -0.020 ± − ± R (cid:63) ).Within the fit, we placed a lower bound on the precision( ∼ bol ) for the SED, which cor-responds to the variations in F bol from different calculationtechniques (Zinn et al. 2019). EXOFASTv2 uses the MESAIsochrones and Stellar Tracks (MIST) stellar evolution mod-els (Paxton et al. 2011, 2013, 2015; Choi et al. 2016; Dot-ter 2016), thereby encoding the physics of stellar evolution,where the global model is penalized for large differencesfrom MIST-predicted stellar values. We ran MCMC fits forboth systems, with strict convergence criteria of a GelmanRubin statistic of less than 1.01 and at least 1000 indepen-dent draws in each parameter. We also fit for a dilution termon the
TESS observations. Specifically, we adopt a Gaus-sian prior on the contamination ratio equal to that reportedby the TESS Input Catalog (TIC, Stassun et al. 2018), with adispersion of 10%. This assumes that the
TESS light curveshave been corrected for known companions in the aperture tobetter than 10%. Although we deblend the FFI light curve,the SPOC pipeline corrects the 2-minute light curve, and nounknown companions were detected in our high-resolutionimaging (see §2.6). This provides an independent check onthose corrections and properly propagate uncertainties. Inboth cases, the fitted dilution found by
EXOFASTv2 is con-sistent with zero. The TFOP SG1 photometry for each sys-tem was detrended within the full fit using an additive modeland the detrending parameters seen in Table 1. See Collinset al. (2017) appendix D for a description of each detrend-ing parameter listed. The fitted transit data for TOI-558 andTOI-559 are shown in Figure 2, the RV fit is shown in Figure3, and the resulting median values and 1-sigma uncertaintiesfor all fitted stellar and planetary parameters are displayed inTables 4 and 5. At the top of Table 4 is a list of the priorsused in the fit. See Eastman et al. (2019) for a full list ofthe fitted and derived parameters from
EXOFASTv2 and anybounds on fitted parameters.OI-558 B & 559 B Table 4.
Median values and 68% confidence interval for global models
Priors: TOI-558 b TOI-559 bGaussian π Gaia Parallax (mas) . . . . . . . . . . . . 2.53691 ± ± [ Fe / H ] Metallicity (dex) -0.02 ± ± A V V-band extinction (mag) 0.0617 0.0415Gaussian (cid:48) D T Dilution in
Tess . . . . . . . . . . . . . . . 0.00000 ± ± M ∗ . . . . . Mass ( M (cid:12) ) . . . . . . . . . . . . . . . . . . . . . . 1.349 + − ± R ∗ . . . . . . Radius ( R (cid:12) ) . . . . . . . . . . . . . . . . . . . . . 1.496 + − + − L ∗ . . . . . . Luminosity ( L (cid:12) ) . . . . . . . . . . . . . . . . . 3.52 + − + − F Bol . . . . . Bolometric Flux × − (cgs) . . . . . . 6.99 + − + − ρ ∗ . . . . . . Density (g cm − ). . . . . . . . . . . . . . . . . . 0.568 + − + − log g . . . . . Surface gravity (cgs) . . . . . . . . . . . . . . 4.218 + − + − T eff . . . . . . Effective Temperature (K) . . . . . . . . . 6466 + − + − [ Fe / H ] . . Metallicity (dex) . . . . . . . . . . . . . . . . . . − + − − + − [ Fe / H ] . Initial Metallicity . . . . . . . . . . . . . . . . . 0.137 + − − + − Age . . . . . . Age (Gyr). . . . . . . . . . . . . . . . . . . . . . . . 1.79 + − + − EEP ‡ . . . Equal Evolutionary Phase . . . . . . . . . 345 + − + − A V . . . . . . V-band extinction (mag) . . . . . . . . . . . 0.033 + − + − σ SED . . . . SED photometry error scaling 1.02 + − + − ϖ . . . . . . . Parallax (mas) . . . . . . . . . . . . . . . . . . . . 2.491 ± + − d . . . . . . . Distance (pc) . . . . . . . . . . . . . . . . . . . . . 401.4 + − ± ˙ γ . . . . . . . RV slope (m/s/day) . . . . . . . . . . . . . . . — − + − Planetary Parameters: P . . . . . . . Period (days) . . . . . . . . . . . . . . . . . . . . . 14.574071 ± ± R P . . . . . . Radius ( R J ) . . . . . . . . . . . . . . . . . . . . . . 1.086 + − + − M P . . . . . . Mass ( M J ) . . . . . . . . . . . . . . . . . . . . . . . 3.61 ± + − T (cid:63) . . . . . . Optimal conjunction Time (BJD TDB ) 2458871.07253 ± ± a . . . . . . . Semi-major axis (AU) . . . . . . . . . . . . . 0.1291 + − ± i . . . . . . . . Inclination (Degrees) . . . . . . . . . . . . . . 86.24 + − + − e . . . . . . . . Eccentricity . . . . . . . . . . . . . . . . . . . . . . 0.298 + − + − τ π circ . . . . . Tidal circularization timescale (Gyr) 347 + − + − ω ∗ . . . . . . Argument of Periastron (Degrees) . . 132.3 + − − + − T eq . . . . . . Equilibrium temperature (K) . . . . . . . 1061 + − + − K . . . . . . . RV semi-amplitude (m/s) . . . . . . . . . . 257.1 ± + − R P / R ∗ . . Radius of planet in stellar radii . . . . 0.0746 + − + − a / R ∗ . . . Semi-major axis in stellar radii . . . . 18.56 + − + − Depth . . . Flux decrement at mid transit . . . . . . 0.00557 + − + − τ . . . . . . . Ingress/egress transit duration (days) 0.0385 + − + − T . . . . . . Total transit duration (days) . . . . . . . . 0.1127 + − + − b . . . . . . . Transit Impact parameter . . . . . . . . . . 0.9073 + − + − T S ,14 . . . . Total eclipse duration (days) . . . . . . . 0.00 ± + − ρ P . . . . . . Density (g cm − ). . . . . . . . . . . . . . . . . . 3.50 + − + − logg P . . . Surface gravity . . . . . . . . . . . . . . . . . . . 1.16358088 ± + − T S . . . . . . Time of eclipse (BJD TDB ) . . . . . . . . . 2458366.38 ± + − e cos ω ∗ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . − ± + − e sin ω ∗ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.221 + − − ± d / R ∗ . . . Separation at mid transit . . . . . . . . . . 13.85 + − + − NOTES: † The initial metallicity is the metallicity of the star when it was formed. ‡ The Equal Evolutionary Point corresponds to static points in a stars evolutionary history when using the MIST isochrones and can be a proxy for age. See §2 inDotter (2016) for a more detailed description of EEP. (cid:63)
Optimal time of conjunction minimizes the covariance between T C and Period. (cid:48) In our analysis, we assume the TESS correction for blending should be better than 10%. Therefore, we adopt a 10% prior on the blending determined fromTICv8 (Stassun et al. 2018). π The tidal quality factor (Q S ) is assumed to be 10 and is calculated using Equation 2 from Adams & Laughlin (2006). KWUT -U KWA ET AL . Table 5.
Median values and 68% confidence intervals for the global models
TOI-558Wavelength Parameters: B i’ z’ TESS u + − ± ± + − u ± ± + − ± AD .. Dilution from neighboring stars – – – 0.00000 ± γ rel . Relative RV Offset (m/s) ...... − + − σ J .. RV Jitter (m/s) ............... 15.4 + − σ J .. RV Jitter Variance ............ 240 + − TESS TESS TESS
LCOSAAO (z’) LCOCTIO (B) LCOCTIO (i’)Sector 2 Sector 3 Sectors 29+30 UT 2019-09-28 UT 2019-10-26 UT 2020-11-06 σ + − + − − + − + − + − + − F ± + − + − + − ± ± C + − − ± + − C + − u ± ± ± + − u + − + − + − + − AD .. Dilution from neighboring stars – – – 0.00000 ± γ rel . Relative RV Offset4 (m/s)..... − ± σ J .. RV Jitter (m/s) ............... 13.6 + − σ J .. RV Jitter Variance ............ 180 + − σ + − + − + − F + − ± + − σ + − + − + − + − F ± ± ± + − C − + − − + − C + − OI-558 B & 559 B DISCUSSIONOur global model shows that TOI-558 is an F-type starwith a mass of 1.349 + − M (cid:12) and a radius of 1.496 + − R (cid:12) . TOI-558 b is a 3.61 ± M J planet on a 14.57-dayorbit with an eccentricity of 0.298 + − . We characterizeTOI-559 as a G dwarf with a stellar mass of 1.026 ± M (cid:12) and radius 1.233 + − R (cid:12) ; TOI-559 b is 6.01 + − M J and its orbital period is 6.98 days with an eccentricity of0.151 + − . Although both planets’ masses are likely consis-tent with core accretion, the mass for TOI-559 b is near thetheoretical lower limit for disk fragmentation (Moe & Kratter2019).We note that we detect a significant long-term RV trendin the multi-year radial velocities of TOI-559. The trend iswell fit by a linear velocity variation at a rate of 0.65 m/day.Assuming a circularly bound orbit for the companion, such atrend would correspond to a substellar mass companion witha semi-major axis less than ∼ >
20 AU are unlikely, as they wouldneed to be of significant mass, and therefore luminosity, toinduce our observed trend. TOI-559 is worthy of long-termRV monitoring to unveil the nature of its companion.With high planetary masses and significant orbital eccen-tricities, TOI-558 b and TOI-559 b occupy a parameter spacewith few known planets. Only around two dozen previouslyconfirmed transiting giant planets with periods between 5 and15 days show eccentricity that differs from zero by more than1 sigma (see Figure 5) . Most ground-based surveys havehad poor completeness for planets with periods longer than5 days (Gaudi et al. 2005), though TESS , which has near-complete sensitivity to hot Jupiters across the main-sequence(Zhou et al. 2019), will yield many more discoveries in thisparameter space. In addition to being particularly massive,TOI-558 b and TOI-559 b have relatively high orbital eccen-tricities (0.3 and 0.15), indicating that these planets may havemigrated to their current orbits through dynamical interac-tions. Based on the ages of the host stars and our estimatesof their respective tidal circularization timescales (see Table4 π ), we expect that neither of these systems has had sufficienttime to circularize.4.1. Period-Mass Distribution
Like eccentricity, the masses of hot Jupiters may also holdclues to their evolutionary processes. The known populationof transiting giant planets with reported masses greater than0.4 Jupiter masses and orbital periods <15 days is shown inFigure 5 (we exclude planets that do not have reported un- as of UT 1 November 2020, https://exoplanetarchive.ipac.caltech.edu/ certainties on the mass in the NASA Exoplanet Archive).With TESS expected to eventually be magnitude-limited forall transiting hot Jupiters ( P <10 days), we can test whetherpossible trends may already exist in the mass distribution ofhot Jupiters (Rodriguez et al. 2019b). To probe this ques-tion, we include TOI-558 b and TOI-559 b in a study ofthe known population of hot Jupiters with periods shorterthan 10 days, evaluating the potential existence of multi-ple populations. We use the Scipy implementations of thetwo-sample Kolmogorov-Smirnov (K-S) test (Massey 1951;Grover 1977) and a two-sample Anderson-Darling (A-D) test(Scholz & Stephens 1987) to qualitatively identify possiblesplits in the total population. Across a range of orbital pe-riod values, we divide the population into two samples, onewith periods shorter than the given value and one with peri-ods longer, and apply the K-S and A-D tests to those two dis-tributions. As shown in Figure 7, we find a minimum p-valuewhen the population split occurs between 5 and 5.5 days, atroughly 5.2 days with the K-S test and 5.4 days with the A-Dtest. In order to limit the influence of detection bias againstlower-mass giant planets at longer periods, we include onlytransiting planets and only those with masses greater than 0.4 M J (with reported mass uncertainties). Given the presence ofdetection biases at long periods and low masses, it is possiblethat our sample selection criteria ( M p > M J ) could affectthe result. We therefore rerun the test using 0.3 and 0.5 M J for the minimum mass cutoff for the sample, but we find noqualitative change in the location of the minimum p-value.The presence of this p-value valley may suggest that thereare two distributions separated near 5.2 days drawn from dis-tinct parent distributions. The short-period (with 311 plan-ets) and long-period (with 42 planets) samples have meanmasses and standard errors of approximately 1.59 ± M J and 2.31 ± M J , respectively. The mass distributions andcumulative mass distributions of the two samples are shownin Figure 7.We caution that the current population of hot giant planetsis a heterogeneous sample that comes from a variety of sur-veys. There are a number of possible biases in the presentsample. For example, ground-based surveys have yieldedfewer discoveries at longer periods, and there may also be adetection bias against the lowest masses among them. Phys-ical factors, including the effect of tidal evolution on short-period planets (Jackson et al. 2009), influence the primor-dial mass-period distribution. Even for an unbiased sample,it is also possible that an apparent minimum in the p-valuelike the one we observe could be equally well described bya single, continuous model (e.g. Schlaufman 2015). Futureinvestigation is warranted, as a more careful characteriza-tion of the population may provide constraints on hot Jupiterformation channels. The presence (and characteristics) oftwo separate hot Jupiter populations—or a single, continuous4 I KWUT -U KWA ET AL . Figure 5.
The population of transiting giant planets with periods less than 15 days and mass > 0.4 M J , shown as a function of orbital periodversus planet mass, as of UT 1 November 2020. Color and size indicate 1-sigma detection of orbital eccentricity; planets shown in gray do nothave significant eccentricity. Figure 6.
The two-sample Kolmogorov-Smirnov (K-S) andAnderson-Darling (A-D) tests applied to the short-period ( P <10days) giant planet ( M p > M J ) population split at orbital peri-ods ranging from 1 to 9 days. The x-axis is the period at whichthe population is split into two samples, and the y-axis is the result-ing p-value. Tidal forces influence the distribution at short periods,possibly shaping the broad minimum between ∼ ∼ ∼ relationship—in the mass-period plane could be comparedto model predictions and simulations of different formation Figure 7.
The mass distributions (solid) and cumulative mass dis-tributions (dashed) of all known hot Jupiters with measured masses>0.4 M J , split at the position of the K-S p-value valley from Figure6 (P ≈ ∼ ∼ processes and migration mechanisms. There are many con-founding variables to consider, such as host star properties,metallicity, system architectures, likely disk conditions, andmore, all of which may affect planetary properties and the ef-ficiency of migration mechanisms, and in turn, the expectedresulting mass-period distribution. Simply identifying thebroad characteristics of the population in mass-period spacewill require additional discoveries, so a large ensemble–likeOI-558 B & 559 B TESS –will likely be re-quired to draw firm conclusions. TOI-558 b and TOI-559 brepresent two examples of planets that can contribute to thesetypes of investigations. CONCLUSIONWe present the discovery and detailed characterization oftwo short-period massive giant planets from the
TESS
FFIs.Globally modeling photometric and spectroscopic observa-tions from
TESS and ground-based facilities using EXO-FASTv2, we confirm TOI-558 b as a 3.62 ± M J planet ina 14.574076 ± + − M J planet in 6.9839115 + − -day orbit around an early G-dwarf. Additionally, both plan-ets are on eccentric orbits, (e = 0.298 + − for TOI-558 band 0.151 + − for TOI-559 b). The measured eccentricitiesmay be remnant from their evolutionary history since tidalforces at these periods would not have had enough time tocircularize their orbits. A long-term RV trend suggests thepresence of an exterior companion to TOI-559, which we donot detect in high-resolution ( ∼ > TESS willprovide a near magnitude-complete sample of transiting hotJupters (Zhou et al. 2019), enabling more robust future stud-ies of the population, possibly yielding signatures of migra-tion. Such future work may help illuminate the evolutionarypathways of hot Jupiter, a question that has persisted sincethe first exoplanet discoveries.
Software:
EXOFASTv2 (Eastman et al. 2013; East-man 2017), AstroImageJ (Collins et al. 2017) , Lightkurve(Lightkurve Collaboration et al. 2018), Tesscut (Brasseuret al. 2019), Keplerspline (Vanderburg & Johnson 2014;Shallue & Vanderburg 2018)
Facilities:
TESS , FLWO 1.5m (Tillinghast ReflectorEchelle Spectrograph), 4.1-m Southern Astrophysical Re-search (SOAR), LCOGT 0.4m, LCOGT 1.0m, 6.5m Mag-ellan Telescope ACKNOWLEDGMENTSThis research has made use of SAO/NASA’s AstrophysicsData System Bibliographic Services. This research has madeuse of the SIMBAD database, operated at CDS, Strasbourg,France. This work has made use of data from the Eu-ropean Space Agency (ESA) mission
Gaia
Gaia
Gaia
Multilateral Agreement.This work makes use of observations from the LCOGT net-work. B.S.G. was supported by a Thomas Jefferson Grant forSpace Exploration from the Ohio State State University.Funding for the
TESS mission is provided by NASA’s Sci-ence Mission directorate. We acknowledge the use of pub-lic
TESS
Alert data from pipelines at the
TESS
Science Of-fice and at the
TESS
Science Processing Operations Center.This research has made use of the NASA Exoplanet Archiveand the Exoplanet Follow-up Observation Program website,which are operated by the California Institute of Technology,under contract with the National Aeronautics and Space Ad-ministration under the Exoplanet Exploration Program. Thispaper includes data collected by the
TESS mission, whichare publicly available from the Mikulski Archive for SpaceTelescopes (MAST). Resources supporting this work wereprovided by the NASA High-End Computing (HEC) Pro-gram through the NASA Advanced Supercomputing (NAS)Division at Ames Research Center for the production of theSPOC data products. Part of this research was carried out atthe Jet Propulsion Laboratory, California Institute of Tech-nology, under a contract with NASA.This paper includes data gathered with the 6.5 meterMagellan Telescopes located at Las Campanas Observa-tory, Chile. This paper includes observations obtained un-der Gemini programs GN-2018B-LP-101 and GN-2020B-LP-105. Some of the observations in the paper made use ofthe High-Resolution Imaging instrument ‘Alopeke. ‘Alopekewas funded by the NASA Exoplanet Exploration Programand built at the NASA Ames Research Center by Steve B.Howell, Nic Scott, Elliott P. Horch, and Emmett Quigley.‘Alopeke was mounted on the Gemini North telescope ofthe international Gemini Observatory, a program of NSF’sOIR Lab, which is managed by the Association of Universi-ties for Research in Astronomy (AURA) under a cooperativeagreement with the National Science Foundation. on behalfof the Gemini partnership: the National Science Foundation(United States), National Research Council (Canada), Agen-cia Nacional de Investigación y Desarrollo (Chile), Ministe-rio de Ciencia, Tecnología e Innovación (Argentina), Min-istério da Ciência, Tecnologia, Inovações e Comunicações6 I
KWUT -U KWA ET AL .(Brazil), and Korea Astronomy and Space Science Institute(Republic of Korea).REFERENCES
Adams, F. C., & Laughlin, G. 2006, ApJ, 649, 1004Bailey, E., & Batygin, K. 2018, ApJL, 866, L2Batygin, K., Bodenheimer, P. H., & Laughlin, G. P. 2016, ApJ,829, 114Beaugé, C., & Nesvorný, D. 2012, ApJ, 751, 119Bensby, T., Feltzing, S., & Oey, M. S. 2014, A&A, 562, A71Bodenheimer, P., & Pollack, J. B. 1986, Icarus, 67, 391Bovy, J. 2017, MNRAS, 470, 1360Brahm, R., Jordán, A., & Espinoza, N. 2017, PASP, 129, 034002Brasseur, C. E., Phillip, C., Fleming, S. W., Mullally, S. E., &White, R. L. 2019, Astrocut: Tools for creating cutouts of TESSimages, ascl:1905.007Brown, T. M., Baliber, N., Bianco, F. B., et al. 2013, PASP, 125,1031Buchhave, L. A., Bakos, G. Á., Hartman, J. D., et al. 2010, ApJ,720, 1118Buchhave, L. A., Latham, D. W., Johansen, A., et al. 2012, Nature,486, 375Burt, J. A., Nielsen, L. D., Quinn, S. N., et al. 2020, AJ, 160, 153Butler, R. P., Marcy, G. W., Williams, E., et al. 1996, PASP, 108,500Choi, J., Dotter, A., Conroy, C., et al. 2016, ApJ, 823, 102Ciardi, D. R., Beichman, C. A., Horch, E. P., & Howell, S. B.2015, ApJ, 805, 16Co¸skunoˇglu, B., Ak, S., Bilir, S., et al. 2011, MNRAS, 412, 1237Collins, K. A., Kielkopf, J. F., Stassun, K. G., & Hessman, F. V.2017, AJ, 153, 77Crane, J. D., Shectman, S. A., & Butler, R. P. 2006, in Society ofPhoto-Optical Instrumentation Engineers (SPIE) ConferenceSeries, Vol. 6269, Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series, ed. I. S. McLean & M. Iye,626931Crane, J. D., Shectman, S. A., Butler, R. P., et al. 2010, in Societyof Photo-Optical Instrumentation Engineers (SPIE) ConferenceSeries, Vol. 7735, Ground-based and Airborne Instrumentationfor Astronomy III, ed. I. S. McLean, S. K. Ramsay, &H. Takami, 773553Crane, J. D., Shectman, S. A., Butler, R. P., Thompson, I. B., &Burley, G. S. 2008, in Society of Photo-Optical InstrumentationEngineers (SPIE) Conference Series, Vol. 7014, Ground-basedand Airborne Instrumentation for Astronomy II, ed. I. S.McLean & M. M. Casali, 701479Cutri, R. M., Skrutskie, M. F., van Dyk, S., et al. 2003, VizieROnline Data Catalog, 2246, 0 Dawson, R. I., & Johnson, J. A. 2018, Annual Review ofAstronomy and Astrophysics, 56, 175Donati, J.-F., Semel, M., Carter, B. D., Rees, D. E., & CollierCameron, A. 1997, MNRAS, 291, 658Dotter, A. 2016, ApJS, 222, 8Eastman, J. 2017, EXOFASTv2: Generalized publication-qualityexoplanet modeling code, Astrophysics Source Code Library,ascl:1710.003Eastman, J., Gaudi, B. S., & Agol, E. 2013, PASP, 125, 83Eastman, J. D., Rodriguez, J. E., Agol, E., et al. 2019, arXive-prints, arXiv:1907.09480Fabrycky, D., & Tremaine, S. 2007, ApJ, 669, 1298F˝urész, G. 2008, PhD thesis, University of Szeged, HungaryFulton, B. J., & Petigura, E. A. 2018, AJ, 156, 264Fulton, B. J., Petigura, E. A., Howard, A. W., et al. 2017, AJ, 154,109Gagné, J., Mamajek, E. E., Malo, L., et al. 2018, ApJ, 856, 23Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al. 2018,A&A, 616, A1Gaudi, B. S., Seager, S., & Mallen-Ornelas, G. 2005, ApJ, 623, 472Goldreich, P., & Tremaine, S. 1980, ApJ, 241, 425Grover, N. B. 1977, Computer Programs in Biomedicine, 7, 247Høg, E., Fabricius, C., Makarov, V. V., et al. 2000, A&A, 355, L27Howell, S. B., Everett, M. E., Sherry, W., Horch, E., & Ciardi,D. R. 2011, AJ, 142, 19Husser, T. O., Wende-von Berg, S., Dreizler, S., et al. 2013, A&A,553, A6Jackson, B., Barnes, R., & Greenberg, R. 2009, ApJ, 698, 1357Jenkins, J. M. 2002, ApJ, 575, 493Jenkins, J. M., Twicken, J. D., McCauliff, S., et al. 2016, inProc. SPIE, Vol. 9913, Software and Cyberinfrastructure forAstronomy IV, 99133EJensen, E. 2013, Tapir: A web interface for transit/eclipseobservability, Astrophysics Source Code Library, ascl:1306.007Kovács, G., Zucker, S., & Mazeh, T. 2002, A&A, 391, 369Li, J., Tenenbaum, P., Twicken, J. D., et al. 2019, PASP, 131,024506Lightkurve Collaboration, Cardoso, J. V. d. M., Hedges, C., et al.2018, Lightkurve: Kepler and TESS time series analysis inPython, ascl:1812.013Lin, D. N. C., Bodenheimer, P., & Richardson, D. C. 1996, Nature,380, 606Lin, D. N. C., & Papaloizou, J. 1986, ApJ, 309, 846Lindegren, L., Hernández, J., Bombrun, A., et al. 2018, A&A, 616,A2
OI-558 B & 559 B17