Obliquity measurement and atmospheric characterization of the WASP-74 planetary system
R. Luque, N. Casasayas-Barris, H. Parviainen, G. Chen, E. Pallé, J. Livingston, V. J. S. Béjar, N. Crouzet, E. Esparza-Borges, A. Fukui, D. Hidalgo, Y. Kawashima, K. Kawauchi, P. Klagyivik, S. Kurita, N. Kusakabe, J. P. de Leon, A. Madrigal-Aguado, P. Montañés-Rodríguez, M. Mori, F. Murgas, N. Narita, T. Nishiumi, G. Nowak, M. Oshagh, M. Sánchez-Benavente, M. Stangret, M. Tamura, Y. Terada, N. Watanabe
AAstronomy & Astrophysics manuscript no. main c (cid:13)
ESO 2020July 24, 2020
Obliquity measurement and atmospheric characterization of theWASP-74 planetary system
R. Luque , , N. Casasayas-Barris , , H. Parviainen , , G. Chen , E. Pallé , , J. Livingston , V. J. S. Béjar , ,N. Crouzet , E. Esparza-Borges , A. Fukui , , D. Hidalgo , , Y. Kawashima , K. Kawauchi , P. Klagyivik , S. Kurita ,N. Kusakabe , , J. P. de Leon , A. Madrigal-Aguado , , P. Montañés-Rodríguez , , M. Mori , F. Murgas , ,N. Narita , , , , T. Nishiumi , , G. Nowak , , M. Oshagh , , M. Sánchez-Benavente , , M. Stangret , ,M. Tamura , , , Y. Terada , and N. Watanabe , Instituto de Astrofísica de Canarias (IAC), 38205 La Laguna, Tenerife, Spain; e-mail: [email protected] Departamento de Astrofísica, Universidad de La Laguna (ULL), 38206, La Laguna, Tenerife, Spain Key Laboratory of Planetary Sciences, Purple Mountain Observatory, Chinese Academy of Sciences, Nanjing 210023, China Department of Astronomy, The University of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo, 113-0033, Japan Science Support O ffi ce, Directorate of Science, European Space Research and Technology Centre (ESA / ESTEC), Keplerlaan 1,2201 AZ Noordwijk, The Netherlands Department of Earth and Planetary Science, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku,Tokyo 113-0033, Japan SRON Netherlands Institute for Space Research, Sorbonnelaan 2, 3584 CA Utrecht, The Netherlands Institute of Planetary Research, German Aerospace Center, Rutherfordstrasse 2, 12489 Berlin, Germany Astrobiology Center of NINS, 2-21-1, Osawa, Mitaka, Tokyo 181-8588, Japan National Astronomical Observatory of Japan, 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan Komaba Institute for Science, The University of Tokyo, 3-8-1 Komaba, Meguro, Tokyo 153-8902, Japan JST, PRESTO, 3-8-1 Komaba, Meguro, Tokyo 153-8902, Japan Department of Astronomical Science, The Graduated University for Advanced Studies, SOKENDAI, 2-21-1, Osawa, Mitaka,Tokyo, 181-8588 Japan
ABSTRACT
We present new transit observations of the hot Jupiter WASP-74 b ( T eq ∼ λ = . ± .
99 deg). We are not able to detect any absorption featureof H α , or any other atomic spectral features, in its high-resolution transmission spectra due to low S / N at the line cores. Despiteprevious claims regarding the presence of strong optical absorbers such TiO and VO gases in the atmosphere of WASP-74 b, the newground-based photometry combined with a reanalysis of previously reported observations from the literature shows a slope in thelow-resolution transmission spectrum steeper than expected from Rayleigh scattering alone.
Key words. planetary systems – planets and satellites: individual: WASP-74 b – planets and satellites: atmospheres – methods:observational – techniques: photometric – techniques: radial velocities – techniques: spectroscopic
1. Introduction
Metal oxides, such as TiO and VO, have been proposed to ex-ist in the atmospheres of highly irradiated hot Jupiters, intro-ducing thermal inversions in the temperature structure (Hubenyet al. 2003; Fortney et al. 2008). However, these early theoreticalpredictions have barely been confidently confirmed by observa-tions. The overall lack of TiO / VO detections in optical transmis-sion spectroscopy then triggered several alternative theoreticalinterpretations, e.g., TiO / VO condensation (Spiegel et al. 2009),stellar activity (Knutson et al. 2010), or high C / O ratio (Mad-husudhan 2012).The progress first appeared in the emission spectroscopy ofso-called ultra-hot Jupiters (UHJs, defined as gas giants withdayside temperatures hotter than ∼ / VO in the optical transmission spectroscopyremains unresolved. An alternative explanation is that Fe is re-sponsible for the thermal inversion, a claim that is gaining sup-port with the recent detection of Fe I in the emission spectra of
Article number, page 1 of 12 a r X i v : . [ a s t r o - ph . E P ] J u l & A proofs: manuscript no. main two ultra hot Jupiters, KELT-9 b (Pino et al. 2020) and WASP-189 b (Yan et al. 2020).Here we present a study of the system WASP-74 (Hellieret al. 2015) using multi-colour photometry and high-resolutionspectroscopy observations with MuSCAT2 (Multicolour Simul-taneous Camera for studying Atmospheres of Transiting exo-planets) and HARPS-N (High Accuracy Radial velocity PlanetSearcher in North hemisphere), respectively. WASP-74 b is a hotJupiter in a 2-day orbit around a F9 star with a magnitude V = .
75 mag. With an equilibrium temperature of around 1900 K,this planet remains very close to the UHJs region. Tsiaras et al.(2018) and Mancini et al. (2019) measured its transmission spec-tra using the Wide Field Camera 3 (WFC3) on board of
Hubble
Space Telescope (
HST ) and ground-based multi-band photom-etry, respectively. Their findings, although tentative, indicate awater-depleted atmosphere with strong optical absorbers such asTiO and VO. Here we revise these finding with our new obser-vations and a re-analysis of the previously published data.Moreover, the time series of high resolution data during atransit allows us to investigate the architecture of the plane-tary system. During a transit, the planet blocks a moving por-tion of the stellar disk, and the corresponding RV of that stel-lar region is masked from the integrated stellar RV. This gen-erates a RV anomaly during the transit which is known as theRossiter–McLaughlin (RM; Rossiter 1924; McLaughlin 1924)e ff ect. The RM signal has been used to estimate the projectedspin–orbit angle ( λ ), the angle between the normal vector ofthe orbital plane and the stellar rotational spin-axis. So far awide diversity of spin-orbit angles have been measured, rangingfrom aligned (Winn 2010) to highly misaligned systems (Addi-son et al. 2018), and even retrograde planets (e.g., Hébrard et al.2011). An statistically large sample of spin-orbit angle is essen-tial for examining theories on planet formation and evolution(Winn et al. 2005; Triaud et al. 2010; Albrecht et al. 2012; Tri-aud 2017, e.g.,). For instance, Winn et al. (2010) found that hotJupiters have larger spin-orbit angles if they are orbiting aroundhot stars ( T e ff > / VO and the presence of Rayleigh scattering.Finally, in Section 6 we present the summary of our results andconclusions.
2. Observations
We observed four transits (three full, one partial) of WASP-74 bwith the MuSCAT2 multi-colour imager (Narita et al. 2019) in-stalled in Telescopio Carlos Sánchez (TCS) located at the TeideObservatory in Tenerife, Spain. Observations were carried outsimultaneously in four colours ( g , r , i , z ) in the three full tran-sits (2018-07-17, 2018-08-16, and 2018-08-31) and only in threecolours ( g , r , i ) for the partial one (2019-06-24) with a pixel scaleof 0.44 (cid:48)(cid:48) pix − . The z band observations are missing in this transit due to a problem with its CCD, which was under maintenance. Asummary of the key properties for each of the nights is presentedin Table 1.The reduction of the multi-colour photometry data was per-formed with a dedicated MuSCAT2 pipeline including bias andflat-field correction. In a nutshell, it calculates aperture photom-etry for a set of comparison stars and photometry aperture sizes,and creates the final relative light curves via global optimisationof the posterior density for a model consisting of a transit model(with quadratic limb darkening coe ffi cients), apertures, compari-son stars, and a linear baseline model with the airmass, seeing, x-and y-centroid shifts, and the sky level as covariates (see Parvi-ainen et al. 2019 for details).Mancini et al. (2019) collected broad-band photometry inseveral filters of WASP-74 b in order to determine the obser-vational transmission spectrum of the planet. Their dataset com-prises a total of 18 light curves from 11 di ff erent transits between2015 and 2017 in the following passbands: Bessell U , Johnson B , Sloan g (cid:48) , r (cid:48) , i (cid:48) , and z (cid:48) , Bessell I , Cousins I , and near-infrared J , H , and K bands. Observations were carried out in the fol-lowing telescopes: Calar Alto 1.23-m telescope (one transit inJohnson B and another in Cousins I ), Danish 1.54-m telescope(seven transits in Bessell I and two more in Bessell U ), andthe GROND (Gamma-Ray Burst Optical / Near-Infrared Detec-tor) multi-colour imager at the MPG 2.2-m telescope in La Silla,Chile (one transit in g (cid:48) , r (cid:48) , i (cid:48) , z (cid:48) , J , H , and K ). Besides, WASP-74 b was observed with the HST / WFC3 camera by Tsiaras et al.(2018) for measuring the transmission spectra of a sample of hotJupiters from 1.1 to 1.7 µ m and with Spitzer / IRAC in 3.6 and4.5 µ m (PI: Deming) for a statistical study of secondary eclipsesof hot Jupiters by Garhart et al. (2020).While we use the HST observations as presented in Tsiaraset al. (2018), we perform our own photometric analysis of the
Spitzer observations. As in Livingston et al. (2019), we extractthe
Spitzer light curves following the approach taken by Knut-son et al. (2012); Beichman et al. (2016) and select the circu-lar aperture size that minimises the combined uncorrelated andcorrelated noise (2.2 pix), as measured by the standard devia-tion and β factor (Pont et al. 2006; Winn et al. 2008). Then, wemodel jointly the transit and systematics inherent to the Spitzer light curves using the pixel-level decorrelation method (Deminget al. 2015), which uses a linear combination of (normalised)pixel light curves to model the e ff ect of point-spread function(PSF) motion on the detector coupled with intrapixel gain varia-tions. The Spitzer
PLD-corrected light curves are shown in Fig. 4and fitted jointly with the remaining photometry in Sect. 5.1.
Three transits of WASP-74 b were observed using the HARPS-N spectrograph (Mayor et al. 2003; Cosentino et al. 2012),mounted on the 3 . µ m and 0.69 µ m with a spectral resolution of R =
115 000. The observations were performed exposing continu-ously before, during, and after the transit, using an exposure timeof 600 s. The signal-to-noise (S / N), calculated as an average ofthe S / N per pixel in the Na i order (590 nm), ranges from 55–88in the second night and 32–64 in the first and third nights. Inall three cases, we used fiber B to monitor possible sky emis-sion during the night. Details on the observations are presentedin Table 1. Article number, page 2 of 12. Luque et al.: Obliquity measurement and atmospheric characterization of the WASP-74 planetary system
Table 1.
Observing log of the WASP-74 b transit observations.
Tel. Instrument Date of Start End Filter T exp N obs Airmass S / N a observation UT UT [s] Na i orderTNG HARPS-N 2018-07-17 22:42 05:08 - 600 38 1.60 → → g r i z → → → → g r i z → → → → g r i z → → b g r i → → Notes. ( a ) Averaged S / N per extracted pixel calculated in the Na i order (590 nm) for each night. ( b ) Partial transit, discarded for the joint photometricanalysis.
HARPS-N observations were reduced using the HARPS-NData reduction Software (DRS), version 3 . .
01 Å intoa one-dimensional spectrum. The final spectra are referred to thebarycentric rest frame and standard air wavelengths are used.
3. Stellar parameters
We used the Zonal Atmospheric Stellar Parameters Estimator(ZASPE; Brahm et al. 2017) code to determine the atmosphericstellar parameters of WASP-74. The parameters were obtainedwith a high S / N spectrum built by co-adding all HARPS-N out-of-transit observations. In summary, ZASPE matches the ob-served stellar spectrum via least-squares minimisation againsta grid of synthetic spectra in the spectral regions most sensi-tive to changes in T e ff , log g (cid:63) , and [Fe / H]. Then, to derive thephysical parameters of the star, we used
PARAM 1.3 , a web in-terface for Bayesian estimation of stellar parameters using the PARSEC isochrones from Bressan et al. (2012). The required in-put is the e ff ective temperature and metallicity of the star, deter-mined spectroscopically, together with its apparent visual mag-nitude and parallax.We derive an e ff ective temperature of T e ff = ±
57 K, astellar mass of M (cid:63) = . ± .
026 M (cid:12) , and a radius of R (cid:63) = . ± .
044 R (cid:12) , in fairly good agreement with the most up-to-date values reported in Mancini et al. (2019). The stellar modelsconstrain the age of the star to be 3 . ± .
65 Gyr. We stress thatthe uncertainties on the derived parameters are internal to thestellar models used and do not include systematic uncertaintiesrelated to input physics. All derived values and previous onesreported in the literature can be found in Table 2.
4. Planetary obliquity
The radial velocities of the three nights were computed using serval (Zechmeister et al. 2018), which uses least-squares fit-ting with a high S / N template to compute the radial velocities.The template is created by co-adding all the out-of-transit spectra http://stev.oapd.inaf.it/cgi-bin/param_1.3 . of the star. The Rossiter-McLaughlin (RM) e ff ect is clearly ob-served in the extracted radial velocities of each individual night(Fig. 1).In order to measure the obliquity ( λ ) of the system, we fita RM model to the radial velocity (RV) data by using the theMarkov chain Monte Carlo (MCMC) algorithm implemented in emcee (Foreman-Mackey et al. 2013). We use two di ff erent RVcontributions to build our model: the RM e ff ect and a circularorbit. Both models are implemented in PyAstronomy (Czeslaet al. 2019) as modelSuite.RmcL and modelSuite.radVel ,respectively. The model containing the RM e ff ect depends on theorbital period ( P ), the transit epoch ( T ), the planet-to-star radiusratio ( R p / R (cid:63) ≡ k ), the angular rotation velocity of the host star( Ω ), the linear limb-darkening coe ffi cient ( (cid:15) ), the inclination ofthe orbit ( i ), the inclination of the stellar rotation axis ( I (cid:63) ), thesky-projected angle between the stellar rotation axis and the nor-mal of planetary orbit plane ( λ ) and the scaled semi-major axis( a / R (cid:63) ≡ a s ). On the other hand, the circular obit RV contributiondepends on P , T c , the stellar velocity semi-amplitude ( K (cid:63) ), andthe o ff set with respect to the null RV ( γ ).As presented in previous studies (e.g., Casasayas-Barris et al.2017), in the fitting procedure, I (cid:63) to 90 deg, while T , a s , k ,and R (cid:63) are fixed to the values derived in Sections 3 and 5.1.The other parameters ( Ω , (cid:15) , λ , K (cid:63) and γ ) remain free. The RVinformation from the three nights is jointly fitted, consideringthat T , Ω , (cid:15) , λ are shared parameters. On the other hand, theo ff set between the model and the data ( γ ) can vary from night tonight as, additionally to the system velocity, the RV informationis given with possible instrumental and stellar activity e ff ects. K (cid:63) could also be a ff ected by activity and become di ff erent fordi ff erent nights (Oshagh et al. 2018). For this reason, we fit onedi ff erent γ and K (cid:63) per night (called γ , γ , γ and K (cid:63), , K (cid:63), , K (cid:63), , respectively).We analyse the system using 50 walkers and a total of 10 steps and checked their convergence using the Gelman-Rubinstatistic. Adequate convergence was considered when the Gel-man–Rubin potential scale reduction factor dropped to within1.03. Each step is initialised at a random point near the measuredvalues from literature. λ is constrained to ±
180 deg, (cid:15) between0 . . ldtk (Parviainen & Aigrain 2015) we esti-mate a linear limb-darkening coe ffi cient of 0 .
71 in the HARPS-Nwavelength coverage), and Ω between 0 . . − , whichis translated to v sin I (cid:63) limited between 3 . . − Article number, page 3 of 12 & A proofs: manuscript no. main
Time [JD] +2.458317e6020406080 V e l o c i t y [ m / s ] Time [JD] +2.458332e6402002040
Time [JD] +2.458362e6402002040
Fig. 1.
Stellar radial velocities of WASP-74 for each individual night. The radial velocity measurements are shown in black dots. In cyan we showthe best fit model obtained with the MCMC procedure. -15.0-10.0-5.00.05.010.015.020.0 R a d i a l V e l o c i t y [ m / s ] Best-fit model312018-07-172018-08-012018-08-310.06 0.04 0.02 0.00 0.02 0.04 0.06
Orbital Phase -10.0-5.00.05.010.015.0 O - C [ m / s ] Fig. 2.
RM e ff ect during the transit of WASP-74 b after being detrended(top panel), and residuals between the data and model (bottom panel). Indi ff erent symbols we show the radial velocities from the three di ff erentdata sets. The dark blue line corresponds to the best-fit RM model. Inlight blue we show the 1 σ and 3 σ uncertainties of the model. (Mancini et al. (2019) measured a v sin I (cid:63) = .
03 km s − ). Themedian values of the posteriors are adopted as the best-fit val-ues, and their error bars correspond to the 1 σ statistical errorsat the corresponding percentiles. The individual RM curves canbe observed in Fig. 1. The detrended data of all nights with thebest-fit model are presented in Fig. 2.With the joint fit of the three nights, we measure a spin-orbitangle of 0 . ± . . ± .
04 rad d − and v sin I (cid:63) = . ± .
50 km s − , consistent with our spectro-scopically derived results from Table 2, and with the results ob-tained by Mancini et al. (2019). The v sin I (cid:63) value derived fromthe RM fitting di ff ers less than 2 σ from Hellier et al. (2015) re-sults. Also, the K (cid:63) values of the individual nights are consistent T eff [K] -300-200-1000100200300 O b li q u i t y [ d e g ] WASP-74b 0.250.500.751.001.251.501.752.00 P l a n e t R a d i u s [ R J ] Fig. 3.
Measurements of orbital obliquity for the known transiting ex-trasolar planetary systems and brown dwarf companions in front ofthe e ff ective temperature of the host star (dots), extracted from TEP-Cat orbital obliquity catalogue (Southworth 2011). In the colour bar wepresent the planetary radius. The star symbol corresponds to WASP-74 b’s spin-orbit measurement. The black-dashed vertical line marksthe 6250 K e ff ective temperature transition from Winn et al. (2010). among themselves and with the value reported in Hellier et al.(2015), pointing to a low level of stellar activity (Oshagh et al.2018). This is supported by the absence of spot-crossing eventsin the MuSCAT2 simultaneous multi-colour photometric obser-vations during HARPS-N first and third transits. Although wecannot assess the impact associated to un-occulted spots a ff ect-ing both RM and transit observations, we can be confident thatour λ and v sin I (cid:63) determinations are not significantly misesti-mated. There are two reasons which support this claim; first theresults obtained with the joint fit are consistent with the resultsobtained when fitting each night independently. Second combin-ing three RMs, as was demonstrated in Oshagh et al. (2018),are su ffi cient to mitigate and minimise the influence of stellaractivity on estimated λ and v sin I (cid:63) . However, as presented inCegla et al. (2016) and Bourrier et al. (2017), the spin-orbit and v sin I (cid:63) measurements performed using the classical RM couldbe significantly biased due to variations in the shape of the localcross-correlation functions (CCFs).In Figure 3 we show the obliquity measurements for knowntransiting planets (from TEPCat orbital obliquity catalogue; Article number, page 4 of 12. Luque et al.: Obliquity measurement and atmospheric characterization of the WASP-74 planetary system
Table 2.
Stellar parameters of WASP-74.Parameter Value ReferenceName WASP-74 Hellier et al. (2015)Coordinates and spectral type α Gaia
DR2 δ − Gaia
DR2Spectral type F 9 Hellier et al. (2015)Magnitudes B [mag] 10 . ± .
04 Tycho-2 V [mag] 9 . ± .
03 Tycho-2 G [mag] 9 . ± . Gaia
DR2 J [mag] 8 . ± .
037 2MASS H [mag] 8 . ± .
018 2MASS K s [mag] 8 . ± .
023 2MASSParallax and kinematics π [mas] 6 . ± . Gaia
DR2 d [pc] 149 . ± . Gaia
DR2 µ α cos δ [mas yr − ] 1 . ± . Gaia
DR2 µ δ [mas yr − ] − . ± . Gaia
DR2 V r [km s − ] − . ± . Gaia
DR2 − . ± .
35 This workPhotospheric parameters T e ff [K] 5883 ±
57 This work5984 ±
57 Mancini et al. (2019)5990 ±
110 Hellier et al. (2015)log g (cid:63) . ± .
02 This work4 . ± . ± .
004 Mancini et al. (2019)4 . ± .
018 Hellier et al. (2015)[Fe / H] 0 . ± .
03 This work0 . ± .
02 Mancini et al. (2019)0 . ± .
13 Hellier et al. (2015) v sin I (cid:63) [km s − ] 6 . ± .
21 This work6 . ± .
19 Mancini et al. (2019)4 . ± . M (cid:63) [M (cid:12) ] 1 . ± .
026 This work1 . ± . ± .
030 Mancini et al. (2019)1 . ± .
12 Hellier et al. (2015) R (cid:63) [R (cid:12) ] 1 . ± .
044 This work1 . ± . ± .
013 Mancini et al. (2019)1 . ± .
05 Hellier et al. (2015)Age [Gyr] 3 . ± .
65 This work4 . + . + . − . − . Mancini et al. (2019)2 . + . − . Hellier et al. (2015)
References.
Gaia
DR2: Gaia Collaboration et al. (2018); Tycho-2: Høget al. (2000); 2MASS: Skrutskie et al. (2006).
Southworth 2011) with respect to the e ff ective temperature oftheir host stars. As presented in Winn et al. (2010), we confirmthat most of the planets orbiting stars with e ff ective temperatureslower than ∼ Table 3.
Estimates for the system parameters derived from RM e ff ectanalysis. Parameter Unit Value λ [deg] 0 . ± . Ω [rad d − ] 0 . + . − . v sin I (cid:63) [km s − ] 5 . ± . T [JD] 2457173 . ± . (cid:15) ... 0 . + . − . K (cid:63), [m s − ] 110 . ± . K (cid:63), [m s − ] 113 . ± . K (cid:63), [m s − ] 115 . ± . γ a [m s − ] 48 . ± . γ [m s − ] − . ± . γ [m s − ] 9 . ± . Notes. ( a ) The super scripts 1, 2, and 3 refer to the results obtained fornights 2018-07-17, 2018-08-01 and 2018-08-31, respectively.
5. Atmospheric characterization
We model the three full transits from MuSCAT2 jointly with thetwo
Spitzer / IRAC (3.6 µ m and 4.5 µ m channels), seven GROND(Sloan g (cid:48) , r (cid:48) , i (cid:48) , z (cid:48) , J , H , and K passbands), seven Danish 1.54-m Telescope (Bessell I passband), and one Calar Alto 1.23-m(Cousins I passband) light curves presented in (Mancini et al.2019) . We exclude the two Danish Bessell U light curves dueto their short pre- and post-transit baselines and strong corre-lated noise that cannot be su ffi ciently accounted for using thedata available, and the CA Johnson B light curve because of par-tial transit coverage (we include partial transits only when wehave light curves with full transit coverage in the same pass-band).We carry out the light curve analysis of the data in a Bayesianframework following Parviainen (2018). First, we construct aflux model to reproduce both the transit and the light curve sys-tematics. Then, we define a noise model to incorporate possi-ble stochastic variability in the observations and combine it withthe flux model and the observations to define the likelihood. Us-ing MCMC sampling, we estimate the joint parameter posteriordistributions after defining the priors on the model parameters.The analyses were carried out with a custom Python code basedon PyTransit (Parviainen 2015),
LDTk (Parviainen & Aigrain2015), emcee (Foreman-Mackey et al. 2013), and other standardPython libraries for astrophysics and scientific computing.The four parameters describing the orbital geometry of theplanet (zero epoch T , orbital period P , stellar density ρ (cid:63) , andimpact parameter b ) are independent of passband or per-light-curve systematics, and thus all the light curves constrain the pos-terior distributions of these parameters. The radius ratio k andthe limb darkening coe ffi cients are wavelength dependent, andthus all the light curves observed in a given passband constrainthe posterior densities of these parameters in that passband. Fi-nally, systematics are modelled using a linear combination ofstate vectors where the number of covariates varies from datasetto dataset (in most cases at least airmass and x- and y- centroidshifts are available). Each covariate is associated with a free co- Kindly provided to us by L. Mancini, personal communication.Article number, page 5 of 12 & A proofs: manuscript no. main
MuSCAT2 g
MuSCAT2 r
MuSCAT2 i
MuSCAT2 z
MuSCAT2 g
MuSCAT2 r
MuSCAT2 i
MuSCAT2 z
MuSCAT2 g
MuSCAT2 r
MuSCAT2 i
MuSCAT2 z N o r m a li z e d f l u x GROND g
GROND r
GROND i
GROND z
GROND J
GROND H
GROND K
DK Bessell I
DK Bessell I
DK Bessell I
DK Bessell I
DK Bessell I
Time - T c [h] DK Bessell I
Time - T c [h] CA Cousins I
Time - T c [h] Spitzer 3.6 m
Time - T c [h] Spitzer 4.5 m
Fig. 4.
Light curves of three full transits of WASP-74 b observed with MuSCAT2, a transit observed with GROND, seven transits observed withDK, two transits observed with CA, and two transits observed with
Spitzer . A transit model in black corresponding to the median of the modelparameter posteriors. e ffi cient, and the coe ffi cient posteriors are constrained only bythe information in the light curve modelled by the covariate.We have eleven separate passbands ( g , r , i , Cousins I , John-son I , z , J , H , K , 3.6 µ m and 4.5 µ m), so we end up with elevenradius ratio parameters and 22 parameters for the quadratic stel-lar limb darkening model. The area ratios have wide uniformpriors that do not constrain the posterior, but the limb darken-ing coe ffi cients have loosely constraining normal priors createdusing LDTk (the LDTk-derived prior standard deviation is mul-tiplied by 10) for all the passbands except J , H , and K . We set uniform priors for those passband limb darkening coe ffi cientsthat constrain the coe ffi cients to smaller values than the LDTk-derived prior for the z band.The linear baseline model contributes 105 free parameters tothe model in total. The MuSCAT2 light curves have four co-variates each (airmass, x- and y- shifts, and aperture entropythat works as a proxy for FWHM) and the GROND light curveshave five covariates each (x- and y-shifts, x- and y-FWHMs, andairmass). The publicly available Danish 1.54-m and Calar Alto1.23-m light curves do not include covariate information. Article number, page 6 of 12. Luque et al.: Obliquity measurement and atmospheric characterization of the WASP-74 planetary system
We first analyse the MuSCAT2 data and GROND data sep-arately to test whether the two datasets agree with each other,and then run the analysis using the MuSCAT2, GROND, Danish,Calar Alto, and
Spitzer light curves jointly (the DK and CA lightcurves in I band cannot be directly compared with MuSCAT2and GROND i filter due to broader wavelength coverage). TheMuSCAT2 analysis agrees well with the GROND analysis, al-though the radius ratio posteriors of the GROND data have largeruncertainties that can be attributed to a passband-dependent in-strumental e ff ect (discontinuity) very close to the transit centrethat cannot be su ffi ciently modelled by the linear baseline model.We adopt the results from the full joint analysis as our finalresult and present them in Table 4, and we also plot all the lightcurves in Fig. 4. Multi-colour observations of hot Jupiters can be used to con-struct a transmission spectrum, valuable to probe their atmo-spheres at the terminator. Figure 5 shows the measured radius ra-tio in di ff erent passbands from the new MuSCAT2 observationstogether with our reanalysis of the observations from Spitzer ,Mancini et al. (2019), and the Tsiaras et al. (2018)
HST observa-tions. Our results disagree with those from Mancini et al. (2019,see their Figs. 8 and 9). We find no evidence of TiO / VO in theatmosphere of WASP-74 b, but a steep slope in the optical trans-mission spectrum.We attribute the origin of this disagreement to the di ff erentanalysis of the light curves. First, we fit all the available lightcurves as a function of passband jointly, which constrains thegeometry and limb darkening parameters better than modellingthe light curves individually. Second, some of the GROND co-variates are relatively noisy and contain a significant amountof strong outliers (especially the centroid estimates in the NIRpassbands). Linear baseline models (the approach used also byMancini et al. 2019) do not perform well with noisy covariates,so we remove the photometry points where the covariates areclearly problematic (that is, we remove part of the data based onthe covariates, but not on the photometry itself). This approachimproves the baseline model significantly, and the separate MuS-CAT2 and GROND analyses agree with each other well.We used the PLanetary Atsmopheric Transmission for Ob-server Noobs code PLATON (Zhang et al. 2019) to retrieve theatmospheric properties of WASP-74 b. PLATON is a fast, user-friendly open-source code for retrieval and forward modellingof exoplanet atmospheres written in Python. For our retrievalanalysis, we assumed an isothermal atmosphere and did nottake into account any contamination from stellar heterogeneities(e.g., Oshagh et al. 2014; Rackham et al. 2018, 2019). We fit thecombined transmission spectrum, including the low-resolutionHST / WFC3 spectrum from Tsiaras et al. (2018), reanalysedCousins I , Bessell I , and GROND measurements from Manciniet al. (2019), and our new MuSCAT2 and Spitzer measurements.The free parameters are isothermal temperature T , planet’s ra-dius at a pressure of 1 bar R , C / O ratio, atmospheric metal-licity relative to the solar value log Z / Z (cid:12) , scattering slope s , scat-tering amplitude log f scatter , cloud-top pressure log P cloud , and anerror multiple to scale measurement uncertainties. We also al-lowed the WFC3 spectrum to have an overall free o ff set, as itwas derived with a di ff erent set of i and a / R (cid:63) , which could in-troduce an o ff set due to the correlation between R p / R (cid:63) and thosetwo transit parameters. https://github.com/ideasrule/platon Table 4.
Stellar and planetary parameters derived from the multi-colourjoint transit analysis of WASP-74 b.
Parameter Unit Value (a)
Ephemeris T [BJD] 2457173 . ± × − P [d] 2 . ± . × − T [h] 2 . ± . k g ± k r ± k i ± k s ± k CI ± k BI ± k z ± k J ± k H ± k K ± k . µ m ± k . µ m ± a s [R (cid:63) ] 4 . ± . b . ± . ρ (cid:63) [g cm − ] 0 . ± . (b) R p , g [ R J ] 1.429 ± R p , r [ R J ] 1.403 ± R p , i [ R J ] 1.383 ± R p , z [ R J ] 1.348 ± R p , CI [ R J ] 1.415 ± R p , BI [ R J ] 1.312 ± R p , J [ R J ] 1.290 ± R p , H [ R J ] 1.372 ± R p , K [ R J ] 1.362 ± R p , . µ m [ R J ] 1.352 ± R p , . µ m [ R J ] 1.367 ± a [AU] 0 . ± . i [deg] 80 . ± . T eq [K] 1865 ± Notes. ( a ) The estimates correspond to the posterior median ( P ) with1 σ uncertainty estimate based on the 16th and 84th posterior percentiles( P and P , respectively) for symmetric, approximately normal poste-riors. For asymmetric, unimodal, posteriors, the estimates are P P − P P − P . ( b ) The derived planetary parameters are based on the stellar parametersshown in Table 2 .
We performed two runs of retrieval analyses: one includingall available measurements, while the other excluding Cousins I and Bessell I passbands. Figure 5 shows the combined trans-mission spectrum along with the best retrieved 1 σ confidenceregion. In both retrieval runs, the retrieved transmission spec-trum is almost featureless, except for the significant scatteringslope in the optical. The retrieved atmospheric parameters are Article number, page 7 of 12 & A proofs: manuscript no. main R p / R Median retrieved model w/o CI+BIMedian retrieved model w/ CI+BI11MuSCAT2+GROND jointWFC3 offset 2 WFC3 offset 1GRONDCA1.23mDK1.54mSpitzer N u m b e r o f s ca l e h e i gh t Fig. 5.
Radius ratios in di ff erent passbands, including MuSCAT2 and GROND joint measurements (circles), GROND near-infrared observations(squares; Mancini et al. 2019), Calar Alto 1.23 m and Danish 1.54 m (triangles; Mancini et al. 2019), HST / WFC3 observations (no symbol) fromTsiaras et al. (2018), and
Spitzer measurements (diamonds). Blue line shows the median retrieved atmospheric model based on all measurementsand the light blue band its 1 σ uncertainty. Red line and its shaded band show the retrieval without the measurements in the Cousins-I and Bessell-I bands. The small diamonds in cyan and yellow colours show the binned version of the retrieved model in the passbands with observations.HST / WFC3 measurements were shifted using the best fit values from the two retrieval runs.
Table 5.
Best fit parameters from the atmospheric retrieval. The priorlabel U represents a uniform distribution.Parameter Prior Run 1 (a) Run 2 (b) R [ R Jup ] U (0 . R p , . R p ) 1 . + . − . . + . − . T [K] U (0 . T eq , . T eq ) 1882 + − + − Scatter slope s U ( − ,
20) 14 . + . − . . + . − . log f scatter U ( − ,
10) 2 . + . − . . + . − . C / O ratio U (0 . ,
2) 0 . + . − . . + . − . log Z / Z (cid:12) U ( − ,
3) 0 . + . − . . + . − . P Cloud − top [log Pa] U ( − . ,
5) 0 . + . − . . + . − . WFC3 o ff set [ppm] U ( − , + − + − Error multiple U (0 . ,
10) 2 . + . − . . + . − . Notes. ( a ) Retrieval run on all available measurements. ( b ) Retrieval runon measurements excluding Cousins-I and Bessell-I passbands. given in Table 5, most of which are not well constrained bythe current observations. The reported errors do not account forthe errors in the input parameters, but only for the fitting proce-dure. The analysis favours a low value for the cloud-top pressure,which is consistent with the lack of water absorption feature inthe
HST / WFC3 band. The analysis tends to retrieve an uncon-strained scattering slope that always skews to the upper bound-ary. If the optical measurements are directly fitted by a linearfunction, the observed slope s obs = − d( R p / R (cid:63) ) / d(ln λ ) can beconverted to a scattering slope of s = s obs R (cid:63) / ( k B T eq /µ/g p ) = . ± . . ± . I andBessell I , respectively, using R (cid:63) from Table 2, T eq from Table 4, g p from Mancini et al. (2019) and assuming a mean molecular weight of µ = . − . However, since di ff erent temperaturesare retrieved from the two runs, the retrieved scattering slopesare correspondingly deviating from the estimates that assume theequilibrium temperature. The inclusion of Cousins I and Bessell I also degrade the goodness of fitting, which is dictated by thevery small error bar of Bessell I .The retrieved atmospheric models might be a challenge fortheories. The observed "super-Rayleigh" slope ( s >
4) could beexplained by photochemical haze particles produced in a vigor-ously mixing atmosphere, where a steep positive opacity gradi-ent relative to altitude can be achieved (Kawashima & Ikoma2019; Ohno & Kawashima 2020). However, the atmospherictemperature of this planet ( T ∼ SiO , have refractive properties that can produce super-Rayleigh slopes without such opacity gradient (Wakeford & Sing2015, e.g.,). However, it may require a somewhat extreme con-dition (e.g., small particle size, namely high nucleation rate, andhigh atmospheric di ff usivity) to reproduce the steepness of theslope as shown for the case of MgSiO clouds in Ormel & Min(2019).In addition, at the same time with the super-Rayleigh slope, itis also necessary to explain the flat spectrum in the near-infraredregion observed by HST / WFC3. It is di ffi cult to reproduce thesetwo features only with a single-aerosol layer, and two (or more)aerosol layers are probably required, where the lower one has athick grey opacity to reproduce the NIR flat spectrum and theupper one is composed of di ff used aerosols that responsible forthe super-Rayleigh slope (Ehrenreich et al. 2014; Dragomir et al.2015; Sing et al. 2015). This idea however requires very di ff erentdi ff usivities for the two layers, and it is uncertain whether sucha condition is realistic or not. Article number, page 8 of 12. Luque et al.: Obliquity measurement and atmospheric characterization of the WASP-74 planetary system
Wavelength [ Å ] -2.0-1.5-1.0-0.50.00.51.01.52.0 F i n / F o u t - [ % ] H Fig. 6.
Transmission spectrum of WASP-74 b in the H α line, combin-ing the three nights observed with the HARPS-N spectrograph. In lightgrey we show the original result. The black dots correspond to the databinned by 0 . In any case, the current observations are not adequate forfurther detailed discussions, and additional observations witha higher precision, wider wavelength coverage, and / or higherspectral resolution would be essential to confirm and furthercharacterise the enigmatic spectral features observed in thisstudy. High-resolution transmission spectroscopy observations are anexcellent tool to study the atmospheric composition of exoplan-ets orbiting bright host stars (Wyttenbach et al. 2015, 2017; Sei-del et al. 2019; Cauley et al. 2019).The one-dimensional HARPS-N spectra were correctedof the Earth telluric absorption contamination using Molecfit(Smette et al. 2015; Kausch et al. 2015), as described in Allartet al. (2017) and used in recent atmospheric studies such as Hoei-jmakers et al. (2018); Casasayas-Barris et al. (2019); Hoeijmak-ers et al. (2019). After this correction, we follow the standardmethod to extract the atmospheric transmission spectrum (seee.g., Wyttenbach et al. 2015; Casasayas-Barris et al. 2018; Yan& Henning 2018; Yan et al. 2019, for details).In particular, in order to move the spectra to the stellarrest frame, we use the stellar velocity semi amplitude ( K (cid:63) = . − ) measured by Hellier et al. (2015). The master out-of-transit spectrum is computed by combining the out-of-transitdata using the S / N of the order as weight. After computing theratio of the spectra by this master out-of-transit spectrum, theresiduals are moved to the planet rest frame using a planet ve-locity semi amplitude K p = .
92 km s − , derived from K (cid:63) ,and the planetary and stellar masses measured in this work (seeTable 2). Finally, due to the long ingress and egress duration ofthe transit, only around 4 spectra were taken between the secondand third contacts. For this reason, when computing the trans-mission spectrum, we average the spectra between the first andfourth contacts of the transit. We note that the selection of thein- and out-of-transit observations is performed using the transitepoch measured in Section 4.We apply this method to di ff erent lines of the spectrum. Inthe case of Na i , the S / N in the stellar lines core is too low toretrieve any atmospheric signature. For the first night, for ex-ample, the central core is at null counts level. Focusing on H α (6562 .
81 Å; Kramida et al. 2019), we combine the results of thethree transit observations by using the mean S / N of each nightin H α order as weights. The transmission spectrum is presentedin Fig. 6. The results from the individual nights are presented inthe Appendix for completeness (Fig. A.1).We also model the centre-to-limb variation (CLV) and RMe ff ects in order to estimate the impact of both e ff ects. For theCLV estimation, we follow Yan et al. (2017), and also includethe RM e ff ect on the stellar lines profile as presented in Yan& Henning (2018) and Casasayas-Barris et al. (2019). For thiscomputation, the stellar spectra are modelled using VALD3 linelist (Ryabchikova et al. 2015), and MARCS (Gustafsson et al.2008) models, assuming solar abundance, local thermodynamicequilibrium and the stellar parameters presented in Section 3.For the WASP-74 system these e ff ects have an impact smallerthan ∼ .
1% (in relative flux) in the H α line core, which is in-cluded in error bars of the resulting transmission spectrum (seeFig. 6).Thus, we are not able to detect any feature with atmosphericorigin. The achieved S / N of the spectra is very low at the linecores, specially for deeper lines such as Na i . For this reason,and due to the equilibrium temperature of the planet (1860 K),which is close to the ultra-hot Jupiter zone, our study is focusedon the H α line. The transmission spectrum around this spectralline does not show any clear signature from the exoplanet atmo-sphere. In order to account for possible systematic e ff ects, wecompute the Empirical Monte Carlo analysis described in Red-field et al. (2008). Figure 7 shows that the "in-in" and "out-out"distributions are centred at zero absorption depth for all nights,as expected. On the other hand, the "in-out" distribution, whichcorresponds to the absorption scenario, is centred at a di ff erentposition depending on the night. For the first and last nights, theabsorption scenario can not be disentangled from the noise leveldue to the S / N achieved during the observations. However, forthe second night where the S / N is the highest, the "in-out" dis-tribution is centred at ∼ − . .
38 h and using 600 s of integration per exposure, weare only able to measure around five spectra fully in-transit withrelatively low S / N. The magnitude of the host star and its transitduration make WASP-74 b a challenging planet for atmosphericstudies using 3 .
6. Summary
The obliquity of the WASP-74 system is measured for the firsttime, using three transits observed with the HARPS-N spectro-graph. Here, we measure an aligned system with a projectedspin-orbit angle of 0 . ± . ff ectivetemperatures lower than ∼ ff erentanalysis of the light curves. We used the PLATON code to retrievethe atmospheric properties of WASP-74 b. We fit the combinedtransmission spectrum, including the low-resolution HST / WFC3spectrum from Tsiaras et al. (2018), the GROND measurementsand our new MuSCAT2 data. The retrieved transmission spec-
Article number, page 9 of 12 & A proofs: manuscript no. main N o r m a li s e d C o un t s In-OutIn-InOut-Out 1.0 0.5 0.0 0.5 1.0 F in /F out - 1 [%] Fig. 7.
Distributions of the EMC analysis in the H α line, using 20 000 iterations and measuring the absorption depth with a bandwidth of 0 .
75 Å.Each panel corresponds to the analysis of one night. In green we present the ’out-out’ scenario, in yellow the ’in-in’, and in grey the ’in-out’,which corresponds to the atmospheric absorption scenario. The blue dashed vertical line marks the zero absorption level. trum is almost featureless, except for the significant scatteringslope in the optical.Finally, using three transit observations of WASP-74 b withthe HARPS-N spectrograph, we investigate its high-resolutiontransmission spectrum. Unfortunately, due to the low S / N of thedata, we are not able to detect any feature with atmospheric ori-gin. The magnitude of the host star and its transit duration makeWASP-74 b a challenging planet for atmospheric studies using4-m class telescopes, but it is an interesting target to be fur-ther studied using high-resolution spectrographs placed in largeraperture telescopes.
Acknowledgements.
This article is partly based on observations made with theMuSCAT2 instrument, developed by ABC, at Telescopio Carlos Sánchez oper-ated on the island of Tenerife by the IAC in the Spanish Observatorio del Teide.It is also based on observations made with the Italian Telescopio NazionaleGalileo (TNG) operated on the island of La Palma by the Fundación GalileoGalilei of the INAF (Istituto Nazionale di Astrofisica) at the Spanish Observa-torio del Roque de los Muchachos of the Instituto de Astrofisica de Canarias.R. L. has received funding from the European Union’s Horizon 2020 researchand innovation program under the Marie Skłodowska-Curie grant agreementNo. 713673 and financial support through the “la Caixa” INPhINIT FellowshipGrant LCF / BQ / IN17 / References
Addison, B. C., Wang, S., Johnson, M. C., et al. 2018, AJ, 156, 197Albrecht, S., Winn, J. N., Johnson, J. A., et al. 2012, ApJ, 757, 18Allart, R., Lovis, C., Pino, L., et al. 2017, A&A, 606, A144Beichman, C., Livingston, J., Werner, M., et al. 2016, ApJ, 822, 39Bourrier, V., Cegla, H. M., Lovis, C., & Wyttenbach, A. 2017, A&A, 599, A33Brahm, R., Jordán, A., Hartman, J., & Bakos, G. 2017, MNRAS, 467, 971Bressan, A., Marigo, P., Girardi, L., et al. 2012, MNRAS, 427, 127Casasayas-Barris, N., Palle, E., Nowak, G., et al. 2017, A&A, 608, A135Casasayas-Barris, N., Pallé, E., Yan, F., et al. 2018, A&A, 616, A151Casasayas-Barris, N., Pallé, E., Yan, F., et al. 2019, A&A, 628, A9Cauley, P. W., Shkolnik, E. L., Ilyin, I., et al. 2019, AJ, 157, 69Cegla, H. M., Lovis, C., Bourrier, V., et al. 2016, A&A, 588, A127Cosentino, R., Lovis, C., Pepe, F., et al. 2014, in Society of Photo-Optical In-strumentation Engineers (SPIE) Conference Series, Vol. 9147, Proc. SPIE,91478C Cosentino, R., Lovis, C., Pepe, F., et al. 2012, in Proc. SPIE, Vol. 8446, Ground-based and Airborne Instrumentation for Astronomy IV, 84461VCzesla, S., Schröter, S., Schneider, C. P., et al. 2019, PyA: Python astronomy-related packagesDeming, D., Knutson, H., Kammer, J., et al. 2015, ApJ, 805, 132Dragomir, D., Benneke, B., Pearson, K. A., et al. 2015, ApJ, 814, 102Ehrenreich, D., Bonfils, X., Lovis, C., et al. 2014, A&A, 570, A89Espinoza, N., Rackham, B. V., Jordán, A., et al. 2019, MNRAS, 482, 2065Evans, T. M., Sing, D. K., Goyal, J. M., et al. 2018, AJ, 156, 283Evans, T. M., Sing, D. K., Kataria, T., et al. 2017, Nature, 548, 58Evans, T. M., Sing, D. K., Wakeford, H. R., et al. 2016, ApJ, 822, L4Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. 2013, PASP, 125,306Fortney, J. J., Lodders, K., Marley, M. S., & Freedman, R. S. 2008, ApJ, 678,1419Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al. 2018, A&A, 616, A1Garhart, E., Deming, D., Mandell, A., et al. 2020, AJ, 159, 137Gustafsson, B., Edvardsson, B., Eriksson, K., et al. 2008, A&A, 486, 951Haynes, K., Mandell, A. M., Madhusudhan, N., Deming, D., & Knutson, H.2015, ApJ, 806, 146Hébrard, G., Ehrenreich, D., Bouchy, F., et al. 2011, A&A, 527, L11Hellier, C., Anderson, D. R., Collier Cameron, A., et al. 2015, AJ, 150, 18Hoeijmakers, H. J., Ehrenreich, D., Heng, K., et al. 2018, Nature, 560, 453Hoeijmakers, H. J., Ehrenreich, D., Kitzmann, D., et al. 2019, A&A, 627, A165Høg, E., Fabricius, C., Makarov, V. V., et al. 2000, A&A, 355, L27Hubeny, I., Burrows, A., & Sudarsky, D. 2003, ApJ, 594, 1011Kausch, W., Noll, S., Smette, A., et al. 2015, A&A, 576, A78Kawashima, Y. & Ikoma, M. 2019, ApJ, 877, 109Knutson, H. A., Howard, A. W., & Isaacson, H. 2010, ApJ, 720, 1569Knutson, H. A., Lewis, N., Fortney, J. J., et al. 2012, ApJ, 754, 22Kramida, A., Yu. Ralchenko, Reader, J., & and NIST ASD Team.2019, NIST Atomic Spectra Database (ver. 5.5.2), [Online]. Available: https://physics.nist.gov/asd [2019, September 15]. National Insti-tute of Standards and Technology, Gaithersburg, MD.Livingston, J. H., Crossfield, I. J. M., Werner, M. W., et al. 2019, AJ, 157, 102Madhusudhan, N. 2012, ApJ, 758, 36Mancini, L., Southworth, J., Mollière, P., et al. 2019, MNRAS, 485, 5168Mayor, M., Pepe, F., Queloz, D., et al. 2003, The Messenger, 114, 20McLaughlin, D. B. 1924, ApJ, 60, 22Narita, N., Fukui, A., Kusakabe, N., et al. 2019, Journal of Astronomical Tele-scopes, Instruments, and Systems, 5, 015001Nugroho, S. K., Kawahara, H., Masuda, K., et al. 2017, AJ, 154, 221Ohno, K. & Kawashima, Y. 2020, ApJ, 895, L47Ormel, C. W. & Min, M. 2019, A&A, 622, A121Oshagh, M., Santos, N. C., Ehrenreich, D., et al. 2014, A&A, 568, A99Oshagh, M., Triaud, A. H. M. J., Burdanov, A., et al. 2018, A&A, 619, A150Parmentier, V., Line, M. R., Bean, J. L., et al. 2018, A&A, 617, A110Parviainen, H. 2015, MNRAS, 450, 3233Parviainen, H. 2018, Bayesian Methods for Exoplanet Science, 149Parviainen, H. & Aigrain, S. 2015, MNRAS, 453, 3821Parviainen, H., Tingley, B., Deeg, H. J., et al. 2019, A&A, 630, A89Pino, L., Désert, J.-M., Brogi, M., et al. 2020, ApJ, 894, L27Pont, F., Zucker, S., & Queloz, D. 2006, MNRAS, 373, 231
Article number, page 10 of 12. Luque et al.: Obliquity measurement and atmospheric characterization of the WASP-74 planetary system
Rackham, B. V., Apai, D., & Giampapa, M. S. 2018, ApJ, 853, 122Rackham, B. V., Apai, D., & Giampapa, M. S. 2019, AJ, 157, 96Redfield, S., Endl, M., Cochran, W. D., & Koesterke, L. 2008, ApJ, 673, L87Rossiter, R. A. 1924, ApJ, 60, 15Ryabchikova, T., Piskunov, N., Kurucz, R. L., et al. 2015, Phys. Scr, 90, 054005Sedaghati, E., Bo ffi n, H. M. J., MacDonald, R. J., et al. 2017, Nature, 549, 238Seidel, J. V., Ehrenreich, D., Wyttenbach, A., et al. 2019, arXiv e-prints[ arXiv:1902.00001 ]Sing, D. K., Wakeford, H. R., Showman, A. P., et al. 2015, MNRAS, 446, 2428Skrutskie, M. F., Cutri, R. M., Stiening, R., et al. 2006, AJ, 131, 1163Smareglia, R., Bignamini, A., Knapic, C., Molinaro, M., & GAPS Collaboration.2014, in Astronomical Society of the Pacific Conference Series, Vol. 485,Astronomical Data Analysis Software and Systems XXIII, ed. N. Manset &P. Forshay, 435Smette, A., Sana, H., Noll, S., et al. 2015, A&A, 576, A77Southworth, J. 2011, Monthly Notices of the Royal Astronomical Society, 417,2166Spiegel, D. S., Silverio, K., & Burrows, A. 2009, ApJ, 699, 1487Triaud, A. H. M. J. 2017, The Rossiter-McLaughlin E ff ect in Exoplanet Research(Springer Living Reference Work), 2Triaud, A. H. M. J., Collier Cameron, A., Queloz, D., et al. 2010, A&A, 524,A25Tsiaras, A., Waldmann, I. P., Zingales, T., et al. 2018, AJ, 155, 156Wakeford, H. R. & Sing, D. K. 2015, A&A, 573, A122Winn, J. N. 2010, Exoplanet Transits and Occultations (University of ArizonaPress), 55–77Winn, J. N., Fabrycky, D., Albrecht, S., & Johnson, J. A. 2010, ApJ, 718, L145Winn, J. N., Fabrycky, D., Albrecht, S., & Johnson, J. A. 2010, The AstrophysicalJournal, 718, L145Winn, J. N., Holman, M. J., Torres, G., et al. 2008, ApJ, 683, 1076Winn, J. N., Noyes, R. W., Holman, M. J., et al. 2005, ApJ, 631, 1215Wyttenbach, A., Ehrenreich, D., Lovis, C., Udry, S., & Pepe, F. 2015, A&A, 577,A62Wyttenbach, A., Lovis, C., Ehrenreich, D., et al. 2017, A&A, 602, A36Yan, F., Casasayas-Barris, N., Molaverdikhani, K., et al. 2019, arXiv e-prints,arXiv:1911.00380Yan, F. & Henning, T. 2018, Nature Astronomy, 2, 714Yan, F., Pallé, E., Fosbury, R. A. E., Petr-Gotzens, M. G., & Henning, T. 2017,A&A, 603, A73Yan, F., Pallé, E., Reiners, A., et al. 2020, arXiv e-prints, arXiv:2007.02716Zechmeister, M., Reiners, A., Amado, P. J., et al. 2018, A&A, 609, A12Zhang, M., Chachan, Y., Kempton, E. M. R., & Knutson, H. A. 2019, PASP, 131,034501 Article number, page 11 of 12 & A proofs: manuscript no. main
Appendix A: High-resolution transmission spectroscopy. Additional results
Wavelength [ Å ] -3-2-10123 F i n / F o u t - [ % ] H 2018-07-17
Wavelength [ Å ] -3-2-10123 F i n / F o u t - [ % ] H 2018-08-01
Wavelength [ Å ] -3-2-10123 F i n / F o u t - [ % ] H 2018-08-31
Fig. A.1.
Individual transmission spectra around H α line. In light gray we show the original data, while in black dots the data is binned by 0 . αα