Transmission spectroscopy of exoplanets has the potential to provide precise measurements of atmospheric chemical abundances, in particular of hot Jupiters whose large sizes and high temperatures make them conducive to such observations. To date, several transmission spectra of hot Jupiters have revealed low amplitude features of water vapour compared to expectations from cloud-free atmospheres of solar metallicity. The low spectral amplitudes in such atmospheres could either be due to the presence of aerosols that obscure part of the atmosphere or due to inherently low abundances of H 2 O in the atmospheres. A recent survey of transmission spectra of ten hot Jupiters used empirical metrics to suggest atmospheres with a range of cloud/haze properties but with no evidence for H 2 O depletion. Here, we conduct a detailed and homogeneous atmospheric retrieval analysis of the entire sample and report the H 2 O abundances, cloud properties, terminator temperature profiles, and detection significances of the chemical species. Our present study finds that the majority of hot Jupiters have atmospheres consistent with sub-solar H 2 O abundances at their day-night terminators. The best constrained abundances range from log(H 2 O) of −5.04 +0.46 −0.30 to −3.16 +0.66 −0.69 , which compared to expectations from solar-abundance equilibrium chemistry correspond to 0.018 +0.035 −0.009 × solar to 1.40 +4.97 −1.11 × solar. Besides H 2 O we report statistical constraints on other chemical species and cloud/haze properties, including cloud/haze coverage fractions which range from 0.18 +0.26 −0.12 to 0.76 +0.13 −0.15 . The retrieved H 2 O abundances suggest sub-solar oxygen and/or super-solar C/O ratios, and can provide important constraints on the formation and migration pathways of hot giant exoplanets. clouds/hazes (e.g., Deming et al. 2013) or due to inherently low H 2 O abundances (e.g., Madhusudhan et al. 2014a) Sing et al. (2016) interpreted the observed atmospheric spectra using empirical metrics based on chemical equilibrium atmospheric models to constrain the prominence of clouds/hazes vis-à-vis the H 2 O abundances in the atmospheres. The contribution of H 2 O was represented by the amplitude of the H 2 O feature at ∼1.4 µm, while the difference between the optical and infrared planetary radii was taken as representative of the cloud/haze contribution. These empirical metrics of the observations were evaluated with a grid of theoretical chemical equilibrium models to suggest that the atmospheres spanned a continuum from clear to cloudy with no evidence for sub-solar H 2 O abundances. However, a complementary study of the original Sing et al. (2016) survey by Barstow et al. (2017) suggested a general range of H 2 O abundances indicating sub-solar abundances in most of the targets.

As an exoplanet transits its host star, some of the light from the star is absorbed by the atoms and molecules in the planets atmosphere, causing the planet to seem bigger; plotting the planets observed size as a function of the wavelength of the light produces a transmission spectrum 1 . Measuring the tiny variations in the transmission spectrum, together with atmospheric modelling, then gives clues to the properties of the exoplanets atmosphere. Chemical species composed of light elementssuch as hydrogen, oxygen, carbon, sodium and potassiumhave in this way been detected in the atmospheres of several hot giant exoplanets [2][3][4][5] , but molecules composed of heavier elements have thus far proved elusive. Nonetheless, it has been predicted that metal oxides such as titanium oxide (TiO) and vanadium oxide occur in the observable regions of the very hottest exoplanetary atmospheres, causing thermal inversions on the dayside 6, 7 . Here we report the detection of TiO in the atmosphere of the hot-Jupiter planet WASP-19b. Our combined spectrum, with its wide spectral coverage, reveals the presence of TiO (to a confidence level of 7.7σ), a strongly scattering haze (7.4σ) and sodium (3.4σ), and confirms the presence of water (7.9σ) in the atmosphere 5,8 .Hot Jupiters are gas-giant exoplanets with sizes like that of Jupiter but much shorter orbital periods. WASP-19b is the shortest-period hot Jupiter to be discovered so far 9 , and has an excessively bloated radius, owing to the extreme radiation that it receives from its host star; as a result of this radiation, the planets effective temperature is more than 2,000 K (obtained via secondaryeclipse measurements 10 ). It is thought that high atmospheric temperatures imply the presence of metal oxides such as TiO, but despite extensive searches 11, 12 a definitive detection of metal oxides in exoplanetary atmospheres has proved elusive.We observed three transits of European Southern Observatorys Very Large Telescope (VLT), using the low-resolution FORS2 spectrograph. By using three of FORS2s grisms600B (blue), 600RI (green) and 600z (red), thereby covering the entire visible-wavelength domain (0.431.04 µm)together with the multi-object spectroscopy configuration, we were able to obtain relatively high-resolution, precise, broadband transmission spectra. Such results were made possible through optimized observing strategies 13 and careful design of the observing mask used for the multi-object observations: this has slits about 30 wide, which minimized differential losses owing to variations in telescope guiding and seeing conditions. The observations presented here were made between 11 November 2014 and 29 February 2016.For each set of observations, we obtained a series of spectra for the main target (WASP-19), as well as for several comparison stars. After standard data-reduction steps, we integrated those spectra for the largest common wavelength domain and 10-nm bins, to produce the 'white' and 'spectrophotometric' light curves, respectively. To correct for the imp...

Interpretations of exoplanetary transmission spectra have been undermined by apparent obscuration due to clouds/hazes. Debate rages on whether weak H 2 O features seen in exoplanet spectra are due to clouds or inherently depleted oxygen. Assertions of solar H 2 O abundances have relied on making a priori model assumptions, e.g. chemical/radiative equilibrium. In this work, we attempt to address this problem with a new retrieval paradigm for transmission spectra. We introduce POSEIDON, a two-dimensional atmospheric retrieval algorithm including generalised inhomogeneous clouds. We demonstrate that this prescription allows one to break vital degeneracies between clouds and prominent molecular abundances. We apply POSEIDON to the best transmission spectrum presently available, for the hot Jupiter HD 209458b, uncovering new insights into its atmosphere at the day-night terminator. We extensively explore the parameter space with an unprecedented 10 8 models, spanning the continuum from fully cloudy to cloud-free atmospheres, in a fully Bayesian retrieval framework. We report the first detection of nitrogen chemistry (NH 3 and/or HCN) in an exoplanet atmosphere at 3.7-7.7σ confidence, non-uniform cloud coverage at 4.5-5.4σ, high-altitude hazes at >3σ, and sub-solar H 2 O at 3-5σ, depending on the assumed cloud distribution. We detect NH 3 at 3.3σ and 4.9σ for fully cloudy and cloud-free scenarios, respectively. For the model with the highest Bayesian evidence, we constrain H 2 O at 5-15 ppm (0.01-0.03× solar) and NH 3 at 0.01 − 2.7 ppm, strongly suggesting disequilibrium chemistry and cautioning against equilibrium assumptions. Our results herald new promise for retrieving cloudy atmospheres using high-precision HST and JWST spectra.

Transmission spectroscopy is a powerful technique widely used to probe exoplanet terminators. Atmospheric retrievals of transmission spectra are enabling comparative studies of exoplanet atmospheres. However, the atmospheric properties inferred by retrieval techniques display a significant anomaly: most retrieved temperatures are far colder than expected. In some cases, retrieved temperatures are ∼ 1000 K colder than T eq . Here, we provide an explanation for this conundrum. We demonstrate that erroneously cold temperatures result when 1D atmospheric models are applied to spectra of planets with differing morning-evening terminator compositions. Despite providing an acceptable fit, 1D retrieval techniques artificially tune atmospheric parameters away from terminator-averaged properties. Retrieved temperature profiles are hundreds of degrees cooler and have weaker temperature gradients than reality. Retrieved abundances are mostly biased by > 1σ and sometimes by > 3σ, with the most extreme biases for ultra-hot Jupiters. When morning-evening compositional differences manifest for prominent opacity sources, H 2 O abundances retrieved by 1D models can be biased by over an order of magnitude. Finally, we demonstrate that these biases provide an explanation for the cold retrieved temperatures reported for WASP-17b and WASP-12b. To overcome biases associated with 1D atmospheric models, there is an urgent need to develop multidimensional retrieval techniques.

We present a new ground-based optical transmission spectrum of the ultrahot Jupiter WASP-103b ( T eq = 2484 K). Our transmission spectrum is the result of combining five new transits from the ACCESS survey and two new transits from the LRG-BEASTS survey with a reanalysis of three archival Gemini/GMOS transits and one VLT/FORS2 transit. Our combined 11-transit transmission spectrum covers a wavelength range of 3900–9450 Å with a median uncertainty in the transit depth of 148 parts per million, which is less than one atmospheric scale height of the planet. In our retrieval analysis of WASP-103b’s combined optical and infrared transmission spectrum, we find strong evidence for unocculted bright regions (4.3σ) and weak evidence for H2O ( 1.9 σ ), HCN ( 1.7 σ ), and TiO ( 2.1 σ ), which could be responsible for WASP-103b’s observed temperature inversion. Our optical transmission spectrum shows significant structure that is in excellent agreement with the extensively studied ultrahot Jupiter WASP-121b, for which the presence of VO has been inferred. For WASP-103b, we find that VO can only provide a reasonable fit to the data if its abundance is implausibly high and we do not account for stellar activity. Our results highlight the precision that can be achieved by ground-based observations and the impacts that stellar activity from F-type stars can have on the interpretation of exoplanet transmission spectra.

Measuring the abundances of carbon and oxygen in exoplanet atmospheres is considered a crucial avenue for unlocking the formation and evolution of exoplanetary systems1,2. Access to the chemical inventory of an exoplanet requires high-precision observations, often inferred from individual molecular detections with low-resolution space-based3–5 and high-resolution ground-based6–8 facilities. Here we report the medium-resolution (R ≈ 600) transmission spectrum of an exoplanet atmosphere between 3 and 5 μm covering several absorption features for the Saturn-mass exoplanet WASP-39b (ref. 9), obtained with the Near Infrared Spectrograph (NIRSpec) G395H grating of JWST. Our observations achieve 1.46 times photon precision, providing an average transit depth uncertainty of 221 ppm per spectroscopic bin, and present minimal impacts from systematic effects. We detect significant absorption from CO2 (28.5σ) and H2O (21.5σ), and identify SO2 as the source of absorption at 4.1 μm (4.8σ). Best-fit atmospheric models range between 3 and 10 times solar metallicity, with sub-solar to solar C/O ratios. These results, including the detection of SO2, underscore the importance of characterizing the chemistry in exoplanet atmospheres and showcase NIRSpec G395H as an excellent mode for time-series observations over this critical wavelength range10.

No abstract

Context. Of the presently known ≈ 3900 exoplanets, sparse spectral observations are available for ≈ 100. Ultra-hot Jupiters have recently attracted interest from observers and theoreticians alike, as they provide observationally accessible test cases. Aims. We aim to study cloud formation on the ultra-hot Jupiter HAT-P-7b, the resulting composition of the local gas phase, and how their global changes affect wavelength-dependent observations utilised to derive fundamental properties of the planet. Methods. We apply a hierarchical modelling approach as a virtual laboratory to study cloud formation and gas-phase chemistry. We utilise 97 vertical 1D profiles of a 3D GCM for HAT-P-7b to evaluate our kinetic cloud formation model consistently with the local equilibrium gas-phase composition. We use maps and slice views to provide a global understanding of the cloud and gas chemistry. Results. The day/night temperature difference on HAT-P-7b (∆T ≈ 2500 K) causes clouds to form on the nightside (dominated by H 2 /He) while the dayside (dominated by H/He) retains cloud-free equatorial regions. The cloud particles vary in composition and size throughout the vertical extension of the cloud, but also globally. TiO 2 [s]/Al 2 O 3 [s]/CaTiO 3 [s]-particles of cm-sized radii occur in the higher dayside-latitudes, resulting in a dayside dominated by gas-phase opacity. The opacity on the nightside, however, is dominated by 0.01 . . . 0.1 µm particles made of a material mix dominated by silicates. The gas pressure at which the atmosphere becomes optically thick is ∼ 10 −4 bar in cloudy regions, and ∼ 0.1 bar in cloud-free regions. Conclusions. HAT-P-7b features strong morning/evening terminator asymmetries, providing an example of patchy clouds and azimuthally-inhomogeneous chemistry. Variable terminator properties may be accessible by ingress/egress transmission photometry (e.g., CHEOPS and PLATO) or spectroscopy. The large temperature differences of ≈2500 K result in an increasing geometrical extension from the night-to the dayside. The chemcial equilibrium H 2 O abundance at the terminator changes by < 1 dex with altitude and 0.3 dex (a factor of 2) across the terminator for a given pressure, indicating that H 2 O abundances derived from transmission spectra can be representative of the well-mixed metallicity at P 10 bar. We suggest the atmospheric C/O as an important tool to trace the presence and location of clouds in exoplanet atmospheres. The atmospheric C/O can be sub-and supersolar due to cloud formation. Phase curve variability of HAT-P-7b is unlikely to be caused by dayside clouds.