Search

Ariel, the Atmospheric Remote-sensing Infrared Exoplanet Large-survey, was adopted as the fourth medium-class mission in ESA's Cosmic Vision programme to be launched in 2029. During its 4-year mission, Ariel will study what exoplanets are made of, how they formed and how they evolve, by surveying a diverse sample of about 1000 extrasolar planets, simultaneously in visible and infrared wavelengths. It is the first mission dedicated to measuring the chemical composition and thermal structures of hundreds of transiting exoplanets, enabling planetary science far beyond the boundaries of the Solar System. The payload consists of an off-axis Cassegrain telescope (primary mirror 1100 mm x 730 mm ellipse) and two separate instruments (FGS and AIRS) covering simultaneously 0.5-7.8 micron spectral range. The satellite is best placed into an L2 orbit to maximise the thermal stability and the field of regard. The payload module is passively cooled via a series of V-Groove radiators; the detectors for the AIRS are the only items that require active cooling via an active Ne JT cooler. The Ariel payload is developed by a consortium of more than 50 institutes from 16 ESA countries, which include the UK, France, Italy, Belgium, Poland, Spain, Austria, Denmark, Ireland, Portugal, Czech Republic, Hungary, the Netherlands, Sweden, Norway, Estonia, and a NASA contribution. This presentation provides an overall summary of the science and instrument design for Ariel and presents the many activities that the Ariel team have planned to engage the science community at large and the public prior to launch. These include the Ariel Dry-Run program and citizen-science programs such as ExoClock and the Ariel Data Challenges.

The use of machine and deep learning is prevalent in many fields of science and industry and is now becoming more widespread in extrasolar planet and solar system sciences. Deep learning holds many potential advantages when it comes to modelling highly non-linear data, as well as speed improvements when compared to traditional analysis and modelling techniques. One such problem is the identification and de-trending of stellar and systematic instrument noise in exoplanet lightcurves and in particular time-resolved spectroscopy of exoplanet atmospheres. As part of the ESA Ariel Space mission and the European Conference on Machine Learning (ECML-PKDD), we have organised two very successful machine learning challenges in 2019 and 2021 (https://www.ariel-datachallenge.space). The aim was to provide new solutions to traditionally intractable problems and to foster closer collaboration between the exoplanet and machine learning communities. Often interdisciplinary approaches to long-standing problems are thwarted by jargon and a lack of familiarity. Data challenges are an excellent way to break down existing barriers and establish new links and collaborations. The top-ranked approaches range from deep-learning to gradient boosted extreme learning machines. In this presentation, I will discuss their pros and cons of solving the set challenge and show how new ideas generated by the data challenges can achieve real progress in the field.

The study of extra-solar planets, or simply, exoplanets, planets outside our own Solar System, is fundamentally a grand quest to understand our place in the Universe. Discoveries in the last two decades have re-defined what we know about planets, and helped us comprehend the uniqueness of our very own Earth. In recent years, however, the focus has shifted from planet detection to planet characterisation, where key planetary properties are inferred from telescope observations using Monte Carlo-based methods. However, the efficiency of sampling-based methodologies is put under strain by the high-resolution observational data from next generation telescopes, such as the James Webb Space Telescope and the Ariel Space Mission. We propose to host a regular competition with the goal of identifying a reliable and scalable method to perform planetary characterisation. Depending on the chosen track, participants will provide either quartile estimates or the approximate distribution of key planetary properties. They will have access to synthetic spectroscopic data generated from the official simulators for the ESA Ariel Space Mission. The aims of the competition are three-fold. 1) To offer a challenging application for comparing and advancing conditional density estimation methods. 2) To provide a valuable contribution towards reliable and efficient analysis of spectroscopic data, enabling astronomers to build a better picture of planetary demographics, and 3) To promote the interaction between ML and exoplanetary science. The competition is open for all and is expected to run from July to October. We will provide a brief introduction to the competition, its aim and the different tracks available for participants. We will also be sharing preliminary results from the competition in this session.

Current endeavours in exoplanet characterisation rely on atmospheric retrieval to quantify crucial physical properties of remote exoplanets from observations. However, the scalability and efficiency of the technique are under strain with increasing spectroscopic resolution and forward model complexity. The situation becomes more acute with the recent launch of the James Webb Space Telescope and other upcoming missions. Recent advances in Machine Learning provide optimisation-based Variational Inference as an alternative approach to perform approximate Bayesian Posterior Inference. In this investigation we combined Normalising Flow-based neural network with our newly developed differentiable forward model, Diff-tau, to perform Bayesian Inference in the context of atmospheric retrieval. We demonstrated, with examples from real and simulated spectroscopic data, several advantages of our proposed framework: 1.) Our neural network does not require a large library of spectra, all it takes is a single observation 2.) It is able to produce distributions similar to sampling-based retrieval and 3.) It requires much less forward model computation to converge. Our proposed framework contribute towards the latest development of a neural-powered atmospheric retrieval. Its flexibility and speed hold the potential to complement sampling-based approaches in large and complex data sets in the future.

Recent analysis of the planet K2-18b has shown the presence of water vapour in its atmosphere. While the H2O detection is significant, the Hubble Space Telescope (HST) WFC3 spectrum suggests three possible solutions of very different nature which can equally match the data. The three solutions are a primary cloudy atmosphere with traces of water vapour (cloudy sub-Neptune), a secondary atmosphere with a substantial amount (up to 50% Volume Mixing Ratio) of H2O (icy/water world) and/or an undetectable gas such as N2 (super-Earth). Additionally, the atmospheric pressure and the possible presence of a liquid/solid surface cannot be investigated with currently available observations. In this paper we used the best fit parameters from Tsiaras et al. (2019) to build James Webb Space Telescope (JWST) and Ariel simulations of the three scenarios. We have investigated 18 retrieval cases, which encompass the three scenarios and different observational strategies with the two observatories. Retrieval results show that twenty combined transits should be enough for the Ariel mission to disentangle the three scenarios, while JWST would require only two transits if combining NIRISS and NIRSpec data. This makes K2-18b an ideal target for atmospheric follow-ups by both facilities and highlights the capabilities of the next generation of space-based infrared observatories to provide a complete picture of low mass planets., Comment: Accepted in Experimental Astronomy, 18 pages, 13 figures

Since its tentative detection in 2003[1][2], the nature of methane, CH4 (a possible bio-signature), on Mars has caused controversy [3][4]. The discrepancy between surface and orbital CH4 measurements, significantly constrains the mechanisms of production and destruction, that are needed to corroborate the measurements [5][6][7][8]. Abiotic and biotic sources of CH4 having been suggested to explain the detection, ranging from Fischer-Tropsch-type (FTT) reactions to methanogenesis by methanogenic archaea respectively [9][10]. Here we present the application of the TauREx Mars model [11][12][13] performing Nested Sampling to attempt retrievals of CH4, H2O and HDO, for Trace Gas Orbiter (TGO) Nadir and Occultation for MArs Discovery (NOMAD) instrument Solar Occultation (SO) channel measurements. This framework, through producing marginalised and conditional posterior distributions of Mars forward model parameters (e.g. trace gas abundances), can be used to map correlations between these parameters. These results may give a greater understanding of the nature of CH4 and the deuterium to hydrogen ratio (D/H) in water in the Martian atmosphere. Tests, validations, and comparisons against NASA’s Planetary Spectrum Generator (PSG) [14] are also presented. Figure 1: An average of 3552 NOMAD SO diffraction order 121 transmittance minus baseline vs wavenumber vs tangent height colourmaps, derived from NOMAD SO PSA L3 calibrated transmittance data from April 2018 to September 2019. Additionally, as part of a broad search for Martian trace gases, all publicly available ESA Planetary Science Archive (PSA) NOMAD SO data have been processed to generate over 60,000 transmittances minus polynomial baseline vs tangent height vs wavenumber maps, covering all available NOMAD diffraction orders, from April 2018 to September 2019. These representations give insight into the Martian atmosphere; with HCl [15] and the magnetic dipole of CO2 [16] identified. Furthermore, this open dataset may be useful to identify new trace gas species, as these representations act as visual tools that can be useful in suggesting which spectra are worthy of performing retrievals and/or contain systematic errors, as well as identifying missing absorption lines in forward models. Figure 2: A plot of diffraction order averaged TGO NOMAD SO transmittance minus polynomial baseline vs wavenumber vs tangent height maps for 39 diffraction orders, using all NOMAD SO PSA L3 calibrated transmittance data from April 2018 to September 2019. Acknowledgements: We would like to thank the UK Space Agency for their support of this Aurora Science studentship STFC: 535385. We would also like to recognise that the development of TauREx has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 Research and Innovation Programme (Grant agreement No. 758892, ExoAI). References: 1.Krasnopolsky-et-al.(2004),Icarus.[https://doi.org/10.1016/j.icarus.2004.07.004]. 2.Formisano-et-al.(2004),Science.[https://doi.org/10.1126/science.1101732]. 3.Mumma-et-al.(2009),Science.[https://doi.org/10.1126/science.1165243]. 4.Lefèvre-and-Forget(2009),Nature.[https://doi.org/10.1038/nature08228]. 5.Webster-et-al.(2015),Science.[https://doi.org/10.1126/science.1261713]. 6.Korablev-et-al.(2019),Nature.[https://doi.org/10.1038/s41586-019-1096-4]. 7.Knutsen-et-al.(2021),Icarus.[https://doi.org/10.1016/j.icarus.2020.114266]. 8.Montmessin-et-al.(2021),A&A.[https://doi.org/10.1051/0004-6361/202140389]. 9.Morozova-et-al.(2007),Orig.Life.Evol.Biosph.[https://doi.org/10.1007/s11084-006-9024-7]. 10.Allen-et-al.(2006),Eos.[https://doi.org/10.1029/2006EO410001]. 11.Cann-et-al.-(2020),[https://arxiv.org/abs/2007.13458]. 12.Waldmann-et-al.(2015),ApJ.[https://doi.org/10.1088/0004-637X/802/2/107]. 13.Al-Refaie-et-al.(2019),ApJ.[https://arxiv.org/pdf/1912.07759.pdf]. 14.Villanueva-et-al.(2013),Icarus.[http://dx.doi.org/10.1016/j.icarus.2012.11.013.]. 15.Korablev-et-al.(2021),ScienceAdvances.[https://doi.org/10.1126/sciadv.abe4386]. 16.Trokhimovskiy-et-al.(2020),A&A.[https://doi.org/10.1051/0004-6361/202038134].

The Twinkle space telescope has been designed for the characterisation of exoplanets and Solar System objects. Operating in a low Earth, Sun-synchronous orbit, Twinkle is equipped with a 45 cm telescope and visible (0.4 – 1 μm) and infrared (1.3 – 4.5 μm) spectrometers which can be operated simultaneously. Twinkle is a general observatory which will provide on-demand observations of a wide variety of targets within wavelength ranges that are currently not accessible using other space telescopes or accessible only to oversubscribed observatories in the short-term future. Here we explore the ability of Twinkle’s spectrometers to characterise the currently-known exoplanets. We study the spectral resolution achievable by combining multiple observations for various planetary and stellar types. We also simulate spectral retrievals for some well-known planets (HD 209458 b, GJ 3470 b and 55 Cnc e). From the exoplanets known today, we find that with a single transit or eclipse, Twinkle could probe 89 planets at low spectral resolution (R < 20) as well as 12 planets at higher resolution (R > 20) in channel 1 (1.3 – 4.5 μm). With 10 observations, the atmospheres of 144 planets could be characterised with R 20. By stacking 10 transits, there are 1185 potential targets for study at R < 20 as well as 388 planets at higher resolutions. The majority of targets are found to be large gaseous planets although by stacking multiple observations smaller planets around bright stars (e.g. 55 Cnc e) could be observed with Twinkle. Photometry and low resolution spectroscopy with Twinkle will be useful to refine planetary, stellar and orbital parameters, monitor stellar activity through time and search for transit time and duration variations (TTVs and TDVs). Refinement of these parameters could be used to in the planning of observations with larger space-based observatories such as JWST and ARIEL. For planets orbiting very bright stars, Twinkle observations at higher spectral resolution will enable us to probe the chemical and thermal properties of an atmosphere. Simultaneous coverage across a wide wavelength range will reduce the degeneracies seen with Hubble and provide access to detections of a wide range molecules. There is the potential to revisit them many times over the mission lifetime to detect variations in cloud cover.

We present a study on the spatially scanned spectroscopic observations of the transit of GJ 1132 b, a warm ($\sim$500 K) Super-Earth (1.13 R$_\oplus$) that was obtained with the G141 grism (1.125 - 1.650 $\mu$m) of the Wide Field Camera 3 (WFC3) onboard the Hubble Space Telescope. We used the publicly available Iraclis pipeline to extract the planetary transmission spectra from the five visits and produce a precise transmission spectrum. We analysed the spectrum using the TauREx3 atmospheric retrieval code with which we show that the measurements do not contain molecular signatures in the investigated wavelength range and are best-fit with a flat-line model. Our results suggest that the planet does not have a clear primordial, hydrogen-dominated atmosphere. Instead, GJ 1132 b could have a cloudy hydrogen-dominated envelope, a very enriched secondary atmosphere, be airless, or have a tenuous atmosphere that has not been detected. Due to the narrow wavelength coverage of WFC3, these scenarios cannot be distinguished yet but the James Webb Space Telescope may be capable of detecting atmospheric features, although several observations may be required to provide useful constraints., Comment: 17 pages, 11 figures. Accepted for publication in AJ

A publicly available database of opacities for molecules of astrophysical interest, ExoMolOP, has been compiled for over 80 species, based on the latest line list data from the ExoMol, HITEMP and MoLLIST databases. These data are generally suitable for characterising high temperature exoplanet or cool stellar/substellar atmospheres, and have been computed at a variety of pressures and temperatures, with a few molecules included at room-temperature only from the HITRAN database. The data are formatted in different ways for four different exoplanet atmosphere retrieval codes; ARCiS, TauREx, NEMESIS and petitRADTRANS, and include both cross-sections (at R~=~$\frac{\lambda}{\Delta \lambda}$~=~15,000) and k-tables (at R~=~$\frac{\lambda}{\Delta \lambda}$~=~1000) for the 0.3~-~50$\mu$m wavelength region. Opacity files can be downloaded and used directly for these codes. Atomic data for alkali metals Na and K are also included, using data from the NIST database and the latest line shapes for the resonance lines. Broadening parameters have been taken from the literature where available, or from those for a known molecule with similar molecular properties where no broadening data are available. The data are available from www.exomol.com., Comment: 25 pages, 9 figures. Accepted for publication in A&A

Thermochemical equilibrium is one of the most commonly used assumptions in current exoplanet retrievals. As the James Webb Space Telescope (JWST) and Ariel launch draw near, assessing the underlying biases and assumptions made when applying self-consistent chemistry into spectral retrievals is crucial. Here we use the flexibility of TauREx 3.1 to cross-compare three state of the art chemical equilibrium codes: ACE, FastChem and GGchem. We simulate JWST spectra for ACE, FastChem, GGchem and GGchem+condensation containing only C, H, O, N elements and spectra for FastChem, GGchem and GGchem+condensation with a more extensive range of elements giving seven simulated JWST spectra in total and then cross-retrieve giving a total of 49 retrievals. Our analysis demonstrates that like-for-like, all chemical codes retrieve the correct parameters to, 31 pages, 15 figures

We present an atmospheric characterization study of two medium sized planets bracketing the radius of Neptune: HD 106315 c (R$_{\rm{P}}$=4.98 $\pm$ 0.23 R$_{\oplus}$) and HD 3167 c (R$_{\rm{P}}$=2.740$_{-0.100}^{+0.106}$ R$_{\oplus}$). We analyse spatially scanned spectroscopic observations obtained with the G141 grism (1.125 - 1.650 $\mu$m) of the Wide Field Camera 3 (WFC3) onboard the Hubble Space Telescope. We use the publicly available Iraclis pipeline and TauREx3 atmospheric retrieval code and we detect water vapor in the atmosphere of both planets with an abundance of $\log_{10}[\mathrm{H_2O}]=-2.1^{+0.7}_{-1.3}$ ($\sim$5.68$\sigma$) and $\log_{10}[\mathrm{H_2O}]=-4.1^{+0.9}_{-0.9}$ ($\sim$3.17$\sigma$) for HD 106315 c and HD 3167 c, respectively. The transmission spectrum of HD 106315 c shows also a possible evidence of ammonia absorption ($\log_{10}[\mathrm {NH_3}]=-4.3^{+0.7}_{-2.0}$, $\sim$1.97$\sigma$ -even if it is not significant-), whilst carbon dioxide absorption features may be present in the atmosphere of HD 3167 c in the $\sim$1.1-1.6~$\mu$m wavelength range ($\log_{10}[\mathrm{CO_{2}}]= -2.4^{+0.7}_{-1.0}$, $\sim$3.28$\sigma$). However the CO$_2$ detection appears significant, it must be considered carefully and put into perspective. Indeed, CO$_2$ presence is not explained by 1D equilibrium chemistry models, and it could be due to possible systematics. The additional contribution of clouds, CO and CH$_4$ are discussed. HD 106315 c and HD 3167 c will be interesting targets for upcoming telescopes such as the James Webb Space Telescope (JWST) and the Atmospheric Remote-Sensing Infrared Exoplanet Large-Survey (Ariel)., Comment: Accepted for publication in AJ

The Hubble Space Telescope’s Wide Field Camera 3 (WFC3) has been widely used for transmission and emission spectroscopy of exoplanet atmospheres, identifying the main molecular constituents, detecting the presence of clouds and probing their thermal structure. Hubble observations of the emission spectra of a number of ultra-hot Jupiters have led to somewhat surprising results. Initially, these very hot planets were predicted to have inverted temperature pressure profiles due to strong optical absorption by TiO/VO in the upper atmospheres. However, observations of their emission spectra have been inconclusive on their thermal structure and composition. While some datasets show rich spectral features, others can be fit with simple blackbody models. We will present the analysis of Hubble WFC3 transmission and emission spectra for two ultra-hot Jupiters: WASP-76 b and KELT-7 b. In each case, the data was reduced and fitted using the open-source codes Iraclis and Taurex3. Previous studies of the WFC3 transmission spectra of WASP-76 b found hints of TiO and VO or non-grey clouds. Accounting for a fainter stellar companion to WASP-76, we reanalyse this data and show that removing the effects of this background star changes the slope of the spectrum, resulting in these visible absorbers no longer being detected, removing the need for a non-grey cloud model to adequately fit the data but maintaining the strong water feature previously seen. However, our analysis of the emission spectrum suggests the presence of titanium oxide (TiO) and an atmospheric thermal inversion. Meanwhile, our study of KELT-7 b uncovers a rich transmission spectrum which suggests the presence of water and H-. In contrast, the extracted emission spectrum does not contain strong absorption features and, although it is not consistent with a simple blackbody, it can be explained by a varying temperature-pressure profile, collision induced absorption (CIA) and H-. These finding bring new insights into the nature of this intriguing class of planets but more data is required to fully understand them and thus we will also present the anticipated results of further characterisation.

The precise derivation of transit depths from stellar light curves is a key component in the construction of exoplanet transit spectra, and thereby for the characterization of exoplanet atmospheres. However, it is still deeply affected by various kinds of complex systematic errors and noises taking their source from host stars’ or instruments’ variability. On the other hand, as the volume of exoplanetary data is quickly increasing, a new way is being opened up for using machine learning as part of the data processing pipeline. By training a recurrent neural network to model the temporal dependencies in stellar light curves, our results on both real on simulated light curves highlight that it is possible to: Model accurately the compound of trends and periodic effects with few or no assumptions about the instrument, star, or planetary signals Improve the understanding of each instrument’s systematic behaviour Optimise a deep detrending model jointly with a transit fit Leverage the cross-light curves and cross-instruments information Such an approach therefore paves the way for a global, flexible and efficient noise-correction pipeline which will be of paramount importance to make the most of exoplanets observations and provide high precision spectra to subsequent atmospheric retrieval pipelines.

Ares is a novel retrieval framework system, designed for the European Space Agency's (ESA) Trace Gas Orbiter (TGO) Nadir and Occultation for MArs Discovery (NOMAD) instrument, Solar Occultation (SO) channel. Ares provides unique insights into the chemical composition of the Martian atmosphere by applying methods utilised in exoplanetary atmospheric retrievals[1][2]. This insight may help unravel the nature of CH4 on Mars. The Ares model is here described including; the NOMAD SO channel instrument function model, Martian atmospheric molecular absorption cross-sections, geometry models and a NOMAD noise model. Ares atmospheric priors and forward models are presented, (i.e. simulated NOMAD observations), and are analysed, compared and validated against NASA’s Planetary Spectrum Generator (PSG)[3]. Introduction In 2003, methane, CH4, was tentatively detected in the Martian atmosphere, at 10±5ppbv, varying by 0-30ppbv globally[4], and 10±3ppbv[5]. CH4 has, at most, a predicted photochemical lifetime of 300 years[6][7], the short lifetime of CH4 in the Martian atmosphere implies that CH4 should be uniformly distributed over Mars. However, non-uniform distributions have been observed[8]. This raises questions with regard to the source(s) and/or sink(s) of CH4. Abiotic and biotic sources have been suggested to explain the detection, ranging from Fischer-Tropsch-type (FTT) reactions[9] to methanogenesis by methanogenic archaea[10]. The NOMAD instrument[11][12], on-board the ESA’s Exomars TGO was designed to investigate the nature of methane, on Mars[13][8]. However, the arrival of the TGO and subsequent science mission has detected no CH4, with an upper limit of 0.05ppbv above 5km[14]. In contrast, NASA’s Curiosity Sample Analysis at Mars Tunable Laser Spectrometer instrument (SAM-TLS)[15] has made multiple measurements of CH4, including measuring an elevated CH4 background of 7.2±2.1ppbv CH4 over a 60-sol period in 2013[16]. Subsequently, using SAM-TLS, a mean CH4 abundance of 0.41±0.16ppbv was determined, as well as a repeatable seasonal variation from 0.24-0.65ppbv[17]. Moreover, on the 20th June 2019 a 19.5±0.18ppbv CH4 emission in Gale Crater was reported by SAM-TLS [18]. Furthermore, the Planetary Fourier Spectrometer (PFS)[19] on-board Mars Express detected 15.5±2.5ppbv of CH4, above Gale Crater on 16th June 2013, one day after SAM-TLS independently detected a CH4 spike of 5.78±2.27ppbv[20]. Since then PFS has detected no CH4[21]. The discrepancy between surface measurements by SAM-TLS and orbital measurements from NOMAD and PFS, combined with the independent confirmation of detection of CH4 by PFS, significantly constrains the mechanisms to corroborate the measurements. Here we present a novel retrieval framework model, Ares, designed for TGO NOMAD SO channel solar occultation measurements. Ares provides a unique insight into the Martian atmosphere by applying methods utilised in exoplanetary atmospheric retrievals. This insight could help unravel the nature of CH4 on Mars. Methods Ares is the Mars version of TauREx3[2], the 3rd generation of TauREx[1] (Tau Retrieval for Exoplanets). TauREx3 is a fully Bayesian atmospheric retrieval framework that uses Nested Sampling[22][23], and Markov chain Monte Carlo methods to sample the full likelihood space of possible solutions. This allows TauREx3 to produce marginalised and conditional posterior distributions of forward model parameters, which can be used to map correlations between these parameters. This is advantageous over other planetary atmospheric retrieval frameworks that only find the maximum-a-posteriori (MAP) solution through Optimal Estimation[24]. A NOMAD SO channel blaze function, acousto-optic tunable filter (AOTF) module has been incorporated into the Ares forward model. Furthermore, Martian ellipsoidal and spherical geometry modules have been included in order to calculate the atmospheric layer line-of-sight (LOS) intersection points and corresponding path lengths[25]. A Mars Climate Database (MCD) Python library, pymcd, has been written and linked to Ares, providing access to Martian vertical temperature, pressure and volume-mixing-ratio (VMR) profiles[26][27]. Moreover, Martian absorption cross-sections have been generated using HITRAN2016[28], accounting for expected Martian pressures, temperatures and CO2 broadening, due to the CO2-rich Martian atmosphere. Overall this enables Ares to produce simulated NOMAD SO channel transmission spectra. Results In order to validate Ares, the Ares forward model has been compared against output from NASA’s PSG[3]. The PSG is an online radiative transfer suite capable of synthesising Martian NOMAD SO channel spectra. Like Ares, the PSG extracts atmospheric priors from the MCD[26][27]. Ares Martian transmittance spectra, simulated using NOMAD observation metadata, compared against the PSG are shown to be in strong agreement, through running Ares and PSG forward models with mutual input variables, for wavenumber range 3011-3035cm-1, i.e. NOMAD SO channel diffraction order 134. Figure 1: A schematic diagram of the TGO, Martian atmospheric layers and corresponding Ares posterior distributions, mapping correlations between atmospheric parameters, for diffraction order 134, at three tangent heights, in the lower, middle and upper atmosphere. Conclusions This research has described a novel retrieval framework, Ares, the Mars version of TauREx3[2], designed for TGO NOMAD SO channel solar occultation measurements. The Ares forward model can simulate NOMAD spectra, including instrumental effects due to the blaze function, AOTF, as well as AOTF temperature dependence. In this research Ares has been applied to simulate and perform self-retrieval of typical Martian mixing ratio concentration vertical profiles of CH4, H2O, and CO2. Simulated Ares and PSG NOMAD SO channel solar occultation observations have been compared and are in strong agreement, with marginalised and conditional posterior distributions subsequently generated. Showing that Ares offers a new method of research in Martian and solar system atmospheric science. Acknowledgements We would like to thank the UK Space Agency for their support of this Aurora Science studentship STFC:535385. We would also like to thank BIRA-IASB for providing test NOMAD spectra, in particular Ian Thomas for help in understanding the intricacies of NOMAD. Finally, we would like to thank the LMD and Ehouarn Millour for help in providing the MCDv5.3. References 1.Waldmann et al.(2015),ApJ.[https://doi.org/10.1088/0004-637X/802/2/107]. 2.Al-Refaie et al.(2019),Submitted to ApJ.[https://arxiv.org/pdf/1912.07759.pdf]. 3.Villanueva et al.(2018),JQSRT.[https://doi.org/10.1016/j.jqsrt.2018.05.023]. 4.Formisano et al.(2004),Science.[https://doi.org/10.1126/science.1101732]. 5.Krasnopolsky et al.(2004),Icarus.[https://doi.org/10.1016/j.icarus.2004.07.004]. 6.Summers et al.(2002),GRL.[https://doi.org/10.1029/2002GL015377]. 7.Lefèvre and Forget(2009),Nature.[https://doi.org/10.1038/nature08228]. 8.Mumma et al.(2009),Science.[https://doi.org/10.1126/science.1165243]. 9.Morozova et al.(2007),Orig.Life.Evol.Biosph.[https://doi.org/10.1007/s11084-006-9024-7]. 10.Allen et al.(2006),Eos.[https://doi.org/10.1029/2006EO410001]. 11.Vandaele et al.(2015),Planet.SpaceSci.[https://doi.org/10.1016/j.pss.2015.10.003]. 12.Robert et al.(2016),Planet.SpaceSci.[https://doi.org/10.1016/j.pss.2016.03.003]. 13.Liuzzi et al.(2019),Icarus.[https://10.1016/j.icarus.2018.09.021]. 14.Korablev et al.(2019),Nature.[https://doi.org/10.1038/s41586-019-1096-4]. 15.Mahaffy et al.(2012),SpaceSci.Rev.[https://doi.org/10.1007/s11214-012-9879-z]. 16.Webster et al.(2015),Science.[https://doi.org/10.1126/science.1261713]. 17.Webster et al.(2018),Science.[https://doi.org/10.1126/science.aaq0131]. 18.Moores et al.(2019),GRL.[https://doi.org/10.1029/2019GL083800]. 19.Formisano et al.(1997),Adv.SpaceRes.[https://doi.org/10.1016/S0273-1177(97)00282-2]. 20.Giuranna et al.(2019),Nature-Geoscience.[https://doi.org/10.1038/s41561-019-0331-9]. 21.ESA(2019),[http://www.esa.int/Science_Exploration/Human_and_Robotic_Exploration/Exploration/ExoMars/ESA_s_Mars_orbiters_did_not_see_latest_Curiosity_methane_burst]. 22.Skilling(2006),Bayesian-Anal.[https://projecteuclid.org/euclid.ba/1340370944]. 23.Feroz et al.(2009),MNRAS,[http://dx.doi.org/10.1111/j.1365-2966.2009.14548.x]. 24.Rodgers(2000),IMFAS,World-Scientific. 25.Thomas et al.(2018),[http://mars.aeronomie.be/en/exomars/observations/pages/EXM-NO-ICD-AER-00001-iss1rev11-EAICD_NOMAD_181026.pdf]. 26.Forget et al.(1999),JGR.[https://doi.org/10.1029/1999JE001025]. 27.Millour et al. (2018),[https://www.cosmos.esa.int/documents/1499429/1583871/Millour_E.pdf]. 28.Gordon et al.(2016),JQSRT.[https://doi.org/10.1016/j.jqsrt.2017.06.038].

TauREx 3.1 is the next version of the open-source python retrieval framework TauREx 3[1], which is backward-compatible with the previous version but offers a swathe of improvements and optimizations to the overall architecture. This version upgrade includes support for k-tables, non-uniform priors, and H- opacities. The most major inclusion in this version of TauREx is the plugin system, which allows any developer to imbue TauREx 3 with new features without touching the main codebase. Plugins are installable additions that TauREx 3 will detect and automatically include in its Bayesian retrieval pipeline. They can be installed from PyPI through pip install or directly from a git repository. They can consist of collections of new chemistries, temperature profiles, contribution functions, stellar and planetary models, non-uniform prior functions, opacity formats, and forward models, to name a few. Anyone can host and develop them, and they are designed so that all are interoperable with each other. Presented are the CUDA and OpenCL plugins, which provide GPU accelerated forward models. Fastchem[2], GGchem[3], and disequilibrium[4] chemical plugins and petitRADTRANS[5] and phase curve forward model plugins. Presented are Forward models and retrievals that exploit the plugins Introduction Exoplanetary atmospheres are multi-faceted physical phenomena that touch many scientific fields. Chemical modeling, cloud physics, fluid dynamics, orbital mechanics and molecular spectroscopy are some of the physical processes that need to be appropriately handled in a retrieval framework to characterize these complex environments correctly. With the upcoming JWST and Ariel telescope expected to bring a higher density of information, it is vital that the many codes and contributions of a wide variety of fields can be exploited for characterization. Plugins TauREx 3 allowed for the inclusion of custom codes in the pipeline. TauREx 3.1 provides a means for which a developer can develop and distribute their custom codes through plugins. Plugins exploit the python packaging system to allow developers to make their codes installable and easily used in retrievals without modification of the main TauREx 3 codebase. The installation is an essential aspect as this allows many FORTRAN and C++ codes to be automatically compiled or binaries distributed. TauREx will then automatically detect these plugins and integrate them into its retrieval pipeline. Plugins can provide replacements for all components in the TauREx 3 framework. For example, if a user wishes to use the GGchem chemistry code in atmospheric retrievals, they can simply write pip install taurex_ggchem, which will automatically download a precompiled GGchem library and its data files and install it alongside a new TauREx chemistry component. A user can then immediately include GGchem for both forward models and retrievals in input files or python scripts. A single plugin can consist of any number of new replacement atmospheric components and can make use of FORTRAN, C++, and python codes/libraries. TauREx-CUDA/OpenCL TauREx-CUDA and TauREx-OpenCL are plugins that, when installed, provide replacement forward models that take advantage of heterogeneous computing. These replacement forward models allow the optical depth calculation to execute on hardware accelerators such as Nvidia, AMD, and Intel GPUs to be easily exploited for a 25--50x speed up in both forward models and retrievals. TauREx-GGchem/Fastchem/ACE/Disequilibrium The four plugins TauREx-ACE, TauREx-Fastchem, and TauREx-GGchem, are installable components that provide the ACE[6], Fastchem[2] and GGchem[3] equilibrium codes for use in both forward modeling and retrievals. These can be easily installed and used in Windows, Mac, and Linux through PyPi and provide their full capabilities for both forward modeling (Figure 1) and retrievals (Figure 2). Also included is a new plugin for a photochemical kinetic solver[4] that provides disequilibrium chemistry with retrieval capabilities (Figure 3). Figure 1. Plots of simulated JWST transmission spectra from different chemistry schemes. Figure 2. Posteriors of ACE (blue), Fastchem (red), GGchem (green) and GGchem with condensation (orange) Figure 3. Active molecular profiles using a photochemical kinetic solver[4] with retrieval sampling uncertainties. Forward Model plugins Plugins can include entirely new forward models. The TauREx-petitRADTRANS plugin that utilizes petitRADTRANS[5] code to compute the transmission and emission spectra. It demonstrates the interoperable nature of plugins. Running petitRADTRANS under the TauREx 3.1 framework gives it the ability to use all available chemistries, including ACE, Fastchem, GGchem, and disequilibrium codes, as well as retrievals utilizing nested sampling with non-uniform priors. The TauREx-phase plugin implements a phase curve forward model for retrievals and can exploit all plugins, including the equilibrium chemistry and CUDA, for self-consistent accelerated optimizations. Summary The plugin system in TauREx 3.1 aims to simplify the inclusion of new atmospheric parameters and external codes. TauREx 3.1 is available at http://github.com/ucl-exoplanets/TauREx3_public or through PyPi. Each plugin is also available through both in the ucl-exoplanets GitHub page and PyPi. References [1] Ahmed F. Al-Refaie, Quentin Changeat, Ingo P. Waldmann, and Giovanna Tinetti. Taurex III: A fast, dynamic and extendable framework for retrievals, 2019. [2] Joachim W Stock, Daniel Kitzmann, A Beate C Patzer, and Erwin Sedlmayr. FastChem: A computer program for efficient complex chemical equilibrium calculations in the neutral/ionized gas phase with applications to stellar and planetary atmospheres .Monthly Notices of the Royal Astronomical Society,479(1):865–874, 06 2018. [3] Woitke, P., Helling, Ch., Hunter, G. H., Millard, J. D., Turner, G. E.,Worters, M., Blecic, J., and Stock, J. W. Equilibrium chemistry down to 100 k - impact of silicates and phyllosilicates on the carbon to oxygen ratio..A&A, 614:A1, 2018. [4] Venot, O., Hebrard, E., Agundez, M., Dobrijevic, M., Selsis, F., Hersant,F., Iro, N., and Bounaceur, R. A chemical model for the atmosphere of hot jupiters. A&A, 546:A43, 2012. [5] Molliere, P., Wardenier, J. P., van Boekel, R., Henning, Th., Molaverdikhani, K., and Snellen, I. A. G. petitRADTRANS - a python radiative transfer package for exoplanet characterization and retrieval.A&A, 627:A67, 2019. [6] M. Agundez, O. Venot, N. Iro, F. Selsis, F. Hersant, E. Hebrard, and M. Do-brijevic. The impact of atmospheric circulation on the chemistry of the hotJupiter HD 209458b.A&A, 548:A73, Dec 2012.

Deep learning algorithms are growing in popularity in the field of exoplanetary science due to their ability to model highly non-linear relations and solve interesting problems in a data-driven manner. Several works have attempted to perform fast retrieval of atmospheric parameters with the use of machine learning algorithms or deep neural networks (DNNs). Yet, despite their high predictive power, DNNs are also infamous for being `black boxes’. It is their apparent lack of explainability that makes the astrophysics community reluctant to adopt them. What are their predictions based on? How confident should we be in them? When are they wrong and how wrong can they be? In this work, we present a number of general evaluation methodologies that can be applied to any trained model and answer questions like these. In particular, we train 3 different popular DNN architectures to retrieve atmospheric parameters from exoplanet spectra and show that all 3 achieve good predictive performance. We then present an extensive analysis of the predictions of DNNs, which can inform us —among other things — of the credibility limit for atmospheric parameters for a given instrument and model. Finally, we perform a sensitivity analysis to identify to which features of the spectrum the outcome of the retrieval is most sensitive. We conclude that for different molecules, the wavelength ranges to which the DNN’s predictions are most sensitive, indeed coincide with their characteristic absorption regions. The methodologies presented in this work help to improve the evaluation of DNNs and to grant interpretability to their predictions.

Abstract Knowledge of the constituents of the Martian surface and their distributions over the planet informs us about Mars’ geomorphological formation and evolutionary history. In this research, an unsupervised end-to-end unmixing model using autoencoders is proposed, that can produce physically plausible abundance maps of the surface mineralology from hyperspectral imaging data. 1 Introduction Mars’ formationary and evolutionary history can be inferred by studying the mineralogical constituents of the Martian surface and their distributions. Knowledge of the specific minerals present and their abundances constrain the possible histories of the planet’s surface (Liu et al., 2016). Hyperspectral imaging of Mars from the Compact Reconnaissance Imaging Spectrometer for Mars (CRISM; Murchie et al., 2007) provides unprecedented insight into the distribution of surface mineralogy. Usually, minerals are identified from spectral absorption features and comparison to spectral libraries (see Viviano-Beck et al., 2014). This can be incredibly time consuming, and relies upon detailed a priori knowledge of the surface chemistry and geology. In order to obtain constituent mineralogical spectra (endmembers) and their fractional abundances one must perform spectral unmixing. Machine learning methods have been used extensively for Earth based hyperspectral imaging analysis, but only a few works to our knowledge have attempted to produce end-to-end unmixing models: Guo et al. (2015); Palsson et al. (2018, 2017); Zhang et al. (2018). Data obtained from hyperspectral imaging of Mars is essentially of the same format as that obtained from Earth, so similar kinds of analysis can be performed. 2 Method Signal decomposition techniques, namely principal component analysis (PCA) and independent component analysis (ICA) are used in order to investigate the presence - and the approximate number - of distinct, clearly detectable signals (component spectra) within CRISM imaging data. Clustering algorithms are then used to obtain an alternative estimate of the number of distinct identifiable mineralogical components present within the spectra. The intuition is that the optimal number of clusters identified should match the number of components determined through the use of signal decomposition techniques. Autoencoders (a type of neural network consisting of an encoder and decoder function that are trained to reproduce their input to their output) are then trained on the spectra from the image. They learn an approximation of the LMM whereby the weights of the decoder correspond to the endmember spectra and the activations of the encoder are the fractional abundances of the endmembers for a given spectrum (Palsson et al., 2018, 2017). The number of endmembers in the learned approximation to the LMM are then set as the estimates of the number of component signals from PCA and clustering. Figure 1 shows a schematic of this. Learned endmembers are extracted from the networks and abundance maps are produced by projecting the activation of each neuron in the decoder to the pixel location of the input spectrum for all of the spectra in the image. These techniques all get around the requirement of detailed a priori knowledge by making no assumptions about the state of the surface. To our knowledge, this is the first work to address the use of unsupervised neural networks for an end-to-end spectral unmixing of Martian hyperspectral images. Figure 1: Schematic view of an autoencoder used for spectral unmixing. 3 Results and Conclusion The signal decomposition techniques provide strong evidence for the presence of detectable mineralogical signals within the data, and the number of signal components estimated from PCA are consistent with the estimates from clustering. PCA, ICA, and the abundance maps produced by the trained autoencoders all appear to convincingly replicate features of the surface topography. Figure 2 shows an example of this, comparing the quicklook image for phyllosilicates (a false colour RGB image constructed from wavelengths using spectral parameters; Viviano-Beck et al., 2014) to one of the produced abundance maps. This provides strong evidence that these methods have learnt mineralogically significant spectra, and that the autoencoders have learnt a physically plausible approximation to the LMM. This is despite the fact that at no spatial information is provided at any point to any of these methods – their results are purely based on spectral information. Hence, when combined with methods to estimate the number of endmembers present, one can construct an unsupervised end-to-end linear unmixing model using autoencoders, that learns to identify physically plausible mineralogical endmembers. Once trained, this pipeline will allow the large number of CRISM data sets to be characterised and abundances mapped over the surface of the planet. This method is also extendable to any kind of planetary hyperspectral imaging, allowing simple mineralogical mapping on an unprecedented scale. Figure 2: Quicklook image for phyllosilicates (a false colour RGB image made using spectral parameters; Viviano-Beck et al., 2014) alongside one of the produced abundance maps. References Guo, R., et al. (2015). Hyperspectral image unmixing using autoencoder cascade. In 2015 7th Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHISPERS), volume 2015-June, pages 1–4. IEEE. Liu, Y., et al. (2016). End-member identi- fication and spectral mixture analysis of CRISM hy- perspectral data: A case study on southwest Melas Chasma, Mars. Journal of Geophysical Research: Planets, 121(10):2004–2036. Murchie, S., et al. (2007). Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) on Mars Reconnaissance Orbiter (MRO). Journal of Geophysical Re- search, 112(E5):E05S03. Palsson, B., et al. (2018). Hyperspectral Unmixing Using a Neural Network Autoencoder. IEEE Access, 6:25646–25656. Palsson, F., et al. (2017). Neural network hyperspec- tral unmixing with spectral information divergence objective. In 2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), pages 755–758. IEEE. Viviano-Beck, et al. (2014). Revised CRISM spectral parameters and summary products based on the currently detected mineral diversity on Mars. Journal of Geophysical Research E: Planets, 119(6):1403–1431. Zhang, X., et al. (2018). Hyperspectral Unmixing via Deep Convolutional Neural Networks. IEEE Geoscience and Remote Sensing Letters, 15(11):1755–1759.

International audience; This paper presents the atmospheric characterization of three large, gaseous planets: WASP-127 b, WASP-79 b, and WASP-62 b. We analyzed spectroscopic data obtained with the G141 grism (1.088-1.68 μm) of the Wide Field Camera 3 on board the Hubble Space Telescope using the Iraclis pipeline and the TauREx3 retrieval code, both of which are publicly available. For WASP-127 b, which is the least dense planet discovered so far and is located in the short-period Neptune desert, our retrieval results found strong water absorption corresponding to an abundance of log(H2O) = -2.71 +0.78−1.05 and absorption compatible with an iron hydride abundance of log(FeH) = −5.25+0.88−1.10, with an extended cloudy atmosphere. We also detected water vapor in the atmospheres of WASP-79 b and WASP-62 b, with best-fit models indicating the presence of iron hydride, too. We used the Atmospheric Detectability Index as well as Bayesian log evidence to quantify the strength of the detection and compared our results to the hot Jupiter population study by Tsiaras et al. While all the planets studied here are suitable targets for characterization with upcoming facilities such as the James Webb Space Telescope and Ariel, WASP-127 b is of particular interest due to its low density, and a thorough atmospheric study would develop our understanding of planet formation and migration. * ARES: Ariel Retrieval of Exoplanets School.

We analyse the transmission and emission spectra of the ultra-hot Jupiter WASP-76b, observed with the G141 grism of the Hubble Space Telescope's Wide Field Camera 3 (WFC3). We reduce and fit the raw data for each observation using the open-source software Iraclis before performing a fully Bayesian retrieval using the publicly available analysis suite TauRex 3. Previous studies of the WFC3 transmission spectra of WASP-76 b found hints of titanium oxide (TiO) and vanadium oxide (VO) or non-grey clouds. Accounting for a fainter stellar companion to WASP-76, we reanalyse this data and show that removing the effects of this background star changes the slope of the spectrum, resulting in these visible absorbers no longer being detected, eliminating the need for a non-grey cloud model to adequately fit the data but maintaining the strong water feature previously seen. However, our analysis of the emission spectrum suggests the presence of TiO and an atmospheric thermal inversion, along with a significant amount of water. Given the brightness of the host star and the size of the atmospheric features, WASP-76 b is an excellent target for further characterisation with HST, or with future facilities, to better understand the nature of its atmosphere, to confirm the presence of TiO and to search for other optical absorbers., Accepted for publication in AJ

The James Webb Space Telescope (JWST) is expected to revolutionize the field of exoplanets. The broad wavelength coverage and the high sensitivity of its instruments will allow characterization of exoplanetary atmospheres with unprecedented precision. Following the Call for the Cycle 1 Early Release Science Program, the Transiting Exoplanet Community was awarded time to observe several targets, including WASP-43b. The atmosphere of this hot Jupiter has been intensively observed but still harbors some mysteries, especially concerning the day-night temperature gradient, the efficiency of the atmospheric circulation, and the presence of nightside clouds. We will constrain these properties by observing a full orbit of the planet and extracting its spectroscopic phase curve in the 5--12 $\mu$m range with JWST/MIRI. To prepare for these observations, we performed an extensive modeling work with various codes: radiative transfer, chemical kinetics, cloud microphysics, global circulation models, JWST simulators, and spectral retrieval. Our JWST simulations show that we should achieve a precision of 210 ppm per 0.1 $\mu$m spectral bin on average, which will allow us to measure the variations of the spectrum in longitude and measure the night-side emission spectrum for the first time. If the atmosphere of WASP-43b is clear, our observations will permit us to determine if its atmosphere has an equilibrium or disequilibrium chemical composition, providing eventually the first conclusive evidence of chemical quenching in a hot Jupiter atmosphere. If the atmosphere is cloudy, a careful retrieval analysis will allow us to identify the cloud composition., Comment: 26 pages, 16 figures, accepted in The Astrophysical Journal

A new generation of exoplanet research beckons and with it the need for simulation tools that accurately predict signal and noise in transit spectroscopy observations. We developed ExoSim: an end-to-end simulator that models noise and systematics in a dynamical simulation. ExoSim improves on previous simulators in the complexity of its simulation, versatility of use and its ability to be generically applied to different instruments. It performs a dynamical simulation that can capture temporal effects, such as correlated noise and systematics on the light curve. It has also been extensively validated, including against real results from the Hubble WFC3 instrument. We find ExoSim is accurate to within 5% in most comparisons. ExoSim can interact with other models which simulate specific time-dependent processes. A dedicated star spot simulator allows ExoSim to produce simulated observations that include spot and facula contamination. ExoSim has been used extensively in the Phase A and B design studies of the ARIEL mission, and has many potential applications in the field of transit spectroscopy., 23 pages, 14 figures

Over the last several years, spectroscopic observations of transiting exoplanets have begun to uncover information about their atmospheres, including atmospheric composition and indications of the presence of clouds and hazes. Spectral retrieval is the leading technique for interpretation of transmission spectra and is employed by several teams using a variety of forward models and parameter estimation algorithms. However, different model suites have mostly been used in isolation and so it is unknown whether the results from each are comparable. As we approach the launch of the James Webb Space Telescope we anticipate advances in wavelength coverage, precision, and resolution of transit spectroscopic data, so it is important that the tools that will be used to interpret these information rich spectra are validated. To this end, we present an inter-model comparison of three retrieval suites: TauREx, NEMESIS and CHIMERA. We demonstrate that the forward model spectra are in good agreement (residual deviations on the order of 20 - 40 ppm), and discuss the results of cross retrievals between the three tools. Generally, the constraints from the cross retrievals are consistent with each other and with input values to within 1 sigma However, for high precision scenarios with error envelopes of order 30 ppm, subtle differences in the simulated spectra result in discrepancies between the different retrieval suites, and inaccuracies in retrieved values of several sigma. This can be considered analogous to substantial systematic/astrophysical noise in a real observation, or errors/omissions in a forward model such as molecular linelist incompleteness or missing absorbers., 25 pages, 21 figures. Accepted in MNRAS

The ExoMol database ( www.exomol.com ) provides molecular data for spectroscopic studies of hot atmospheres. While the data are intended for studies of exoplanets and other astronomical bodies, the dataset is widely applicable. The basic form of the database is extensive line lists; these are supplemented with partition functions, state lifetimes, cooling functions, Lande g-factors, temperature-dependent cross sections, opacities, pressure broadening parameters, k-coefficients and dipoles. This paper presents the latest release of the database which has been expanded to consider 80 molecules and 190 isotopologues totaling over 700 billion transitions. While the spectroscopic data are concentrated at infrared and visible wavelengths, ultraviolet transitions are being increasingly considered in response to requests from observers. The core of the database comes from the ExoMol project which primarily uses theoretical methods, albeit usually fine-tuned to reproduce laboratory spectra, to generate very extensive line lists for studies of hot bodies. The data have recently been supplemented by line lists derived from direct laboratory observations, albeit usually with the use of ab initiotransition intensities. A major push in the new release is towards accurate characterisation of transition frequencies for use in high resolution studies of exoplanets and other bodies.

The precise derivation of transit depths from transit light curves is a key component for measuring exoplanet transit spectra, and henceforth for the study of exoplanet atmospheres. However, it is still deeply affected by various kinds of systematic errors and noise. In this paper we propose a new detrending method by reconstructing the stellar flux baseline during transit time. We train a probabilistic Long Short-Term Memory (LSTM) network to predict the next data point of the light curve during the out-of-transit, and use this model to reconstruct a transit-free light curve - i.e. including only the systematics - during the in-transit. By making no assumption about the instrument, and using only the transit ephemeris, this provides a general way to correct the systematics and perform a subsequent transit fit. The name of the proposed model is TLCD-LSTM, standing for Transit Light Curve Detrending LSTM. Here we present the first results on data from six transit observations of HD 189733b with the IRAC camera on board the Spitzer Space Telescope, and discuss some of its possible further applications., Comment: 12 pages, 10 figures, 4 tables, accepted for publication in The Astronomical Journal

We have conducted a re-analysis of publicly available Hubble Space Telescope Wide Field Camera 3 (HST WFC3) transmission data for the hot-Jupiter exoplanet WASP-43b, using the Bayesian retrieval package Tau-REx. We report evidence of AlO in transmission to a high level of statistical significance (> 5-sigma in comparison to a flat model, and 3.4-sigma in comparison to a model with H2O only). We find no evidence of the presence of CO, CO2, or CH4 based on the available HST WFC3 data or on Spitzer IRAC data. We demonstrate that AlO is the molecule that fits the data to the highest level of confidence out of all molecules for which high-temperature opacity data currently exists in the infrared region covered by the HST WFC3 instrument, and that the subsequent inclusion of Spitzer IRAC data points in our retrieval further supports the presence of AlO. H2O is the only other molecule we find to be statistically significant in this region. AlO is not expected from the equilibrium chemistry at the temperatures and pressures of the atmospheric layer that is being probed by the observed data. Its presence therefore implies direct evidence of some disequilibrium processes with links to atmospheric dynamics. Implications for future study using instruments such as the James Webb Space Telescope (JWST) are discussed, along with future opacity needs. Comparisons are made with previous studies into WASP-43b., Accepted for publication in A&A. 16 pages, 7 figures

In the past decade, the analysis of exoplanet atmospheric spectra has revealed the presence of water vapour in almost all the planets observed, with the exception of a fraction of overcast planets. Indeed, water vapour presents a large absorption signature in the wavelength coverage of the Hubble Space Telescope's (HST) Wide Field Camera 3 (WFC3), which is the main space-based observatory for atmospheric studies of exoplanets, making its detection very robust. However, while carbon-bearing species such as methane, carbon monoxide and carbon dioxide are also predicted from current chemical models, their direct detection and abundance characterisation has remained a challenge. Here we analyse the transmission spectrum of the puffy, clear hot-Jupiter KELT-11 b from the HST WFC3 camera. We find that the spectrum is consistent with the presence of water vapor and an additional absorption at longer wavelengths than 1.5um, which could well be explained by a mix of carbon bearing molecules. CO2, when included is systematically detected. One of the main difficulties to constrain the abundance of those molecules is their weak signatures across the HST WFC3 wavelength coverage, particularly when compared to those of water. Through a comprehensive retrieval analysis, we attempt to explain the main degeneracies present in this dataset and explore some of the recurrent challenges that are occurring in retrieval studies (e.g: the impact of model selection, the use of free vs self-consistent chemistry and the combination of instrument observations). Our results make this planet an exceptional example of chemical laboratory where to test current physical and chemical models of hot-Jupiters' atmospheres., Comment: 24 pages, 14 figures, Accepted in AJ

The H2020 Europlanet-2020 programme, which ended on Aug 31st, 2019, included an activity called VESPA (Virtual European Solar and Planetary Access), which focused on adapting Virtual Observatory (VO) techniques to handle Planetary Science data [1] [2]. The outcome of this activity is a contributive data distribution system where data services are located and maintained in research institutes, declared in a registry, and accessed by several clients based on a specific access protocol. During Europlanet-2020, 52 data services were installed, including the complete ESA Planetary Science Archive, and the outcome of several EU funded projects. Data are described using the EPN-TAP protocol, which parameters describe acquisition and observing conditions as well as data characteristics (physical quantity, data type, etc). A main search portal has been developed to optimize the user experience, which queries all services together. Compliance with VO standards ensures that existing tools can be used as well, either to access or visualize the data. In addition, a bridge linking the VO and Geographic Information Systems (GIS) has been installed to address formats and tools used to study planetary surfaces; several large data infrastructures were also installed or upgraded (SSHADE for lab spectroscopy, PVOL for amateurs images, AMDA for plasma-related data).In the framework of the starting Europlanet-2024 programme, the VESPA activity will complete this system even further: 30-50 new data services will be installed, focusing on derived data, and experimental data produced in other Work Packages of Europlanet-2024; connections between PDS4 and EPN-TAP dictionaries will make PDS metadata searchable from the VESPA portal and vice versa; Solar System data present in astronomical VO catalogues will be made accessible, e.g. from the VizieR database. The search system will be connected with more powerful display and analysing tools: a run-on-demand platform will be installed, as well as Machine Learning capacities to process the available content. Finally, long-term sustainability will be improved by setting VESPA hubs to assist data providers in maintaining their services, and by using the new EU-funded European Open Science Cloud (EOSC). In addition to favoring data exploitation, VESPA will provide a handy and economical solution to Open Science challenges in the field.The Europlanet 2020 & 2024 Research Infrastructure project have received funding from the European Union's Horizon 2020 research and innovation programme under grant agreements No 654208 & 871149.[1] Erard et al 2018, Planet. Space Sci. 150, 65-85. 10.1016/j.pss.2017.05.013. ArXiv 1705.09727 [2] Erard et al. 2020, Data Science Journal 19, 22. doi: 10.5334/dsj-2020-022.

Deep-learning algorithms are growing in popularity in the field of exoplanetary science due to their ability to model highly nonlinear relations and solve interesting problems in a data-driven manner. Several works have attempted to perform fast retrievals of atmospheric parameters with the use of machine-learning algorithms like deep neural networks (DNNs). Yet, despite their high predictive power, DNNs are also infamous for being "black boxes." It is their apparent lack of explainability that makes the astrophysics community reluctant to adopt them. What are their predictions based on? How confident should we be in them? When are they wrong, and how wrong can they be? In this work, we present a number of general evaluation methodologies that can be applied to any trained model and answer questions like these. In particular, we train three different popular DNN architectures to retrieve atmospheric parameters from exoplanet spectra and show that all three achieve good predictive performance. We then present an extensive analysis of the predictions of DNNs, which can inform us–among other things–of the credibility limits for atmospheric parameters for a given instrument and model. Finally, we perform a perturbation-based sensitivity analysis to identify to which features of the spectrum the outcome of the retrieval is most sensitive. We conclude that, for different molecules, the wavelength ranges to which the DNNs predictions are most sensitive do indeed coincide with their characteristic absorption regions. The methodologies presented in this work help to improve the evaluation of DNNs and to grant interpretability to their predictions.

TauREx 3 is the next generation of the TauREx exoplanet atmospheric retrieval framework for Windows, Mac and Linux. It is a complete rewrite with a full Python stack that makes it simple to use, high performance and dynamic/flexible. The new main taurex program is extremely modular, allowing the user to augment taurex functionalities with their own code and easily perform retrievals on their own parameters. This is achieved by dynamic determination of fitting parameters where TauREx 3 can detect new parameters for retrieval from the user code though a simple interface. TauREx 3 can act as a library with a simple 'import taurex' providing a rich set of classes and functions related to atmospheric modelling. A 10x speed-up in forward model computations is achieved compared to the previous version with a six-fold reduction in retrieval times whilst maintaining robust results. TauREx 3 intends to act as a standalone, all in one package for retrievals whilst the TauREx 3 python library can be used by the user to easily build or augment their own data pipelines.

The chemical environments of young planets are assumed to be largely influenced by impacts of bodies lingering on unstable trajectories after the dissolution of the protoplanetary disk. We explore the chemical consequences of impacts within the context of reducing planetary atmospheres dominated by carbon monoxide, methane and molecular nitrogen. A terawatt high-power laser was selected in order to simulate the airglow plasma and blast wave surrounding the impactor. The chemical results of these experiments are then applied to a theoretical atmospheric model. The impact simulation results in substantial volume mixing ratios within the reactor of 5% hydrogen cyanide (HCN), 8% acetylene (C$_2$H$_2$), 5% cyanoacetylene (HC$_3$N) and 1% ammonia (NH$_3$). These yields are combined with estimated impact rates for the Early Earth to predict surface boundary conditions for an atmospheric model. We show that impacts might have served as sources of energy that would have led to steady-state surface quantities of 0.4% C$_2$H$_2$, 400 ppm HCN and 40 ppm NH$_3$. We provide simulated transit spectra for an Earth-like exoplanet with this reducing atmosphere during and shortly after eras of intense impacts. We predict that acetylene is as observable as other molecular features on exoplanets with reducing atmospheres that have recently gone through their own `Heavy Bombardments', with prominent features at 3.05 $\mu$m and 10.5 $\mu$m., Comment: 26 pages, 12 figures, 2 tables, accepted for publication by The Astrophysical Journal

Further advances in exoplanet detection and characterisation require sampling a diverse population of extrasolar planets. One technique to detect these distant worlds is through the direct detection of their thermal emission. The so-called direct imaging technique, is suitable for observing young planets far from their star. These are very low signal-to-noise-ratio (SNR) measurements and limited ground truth hinders the use of supervised learning approaches. In this paper, we combine deep generative and discriminative models to bypass the issues arising when directly training on real data. We use a Generative Adversarial Network to obtain a suitable dataset for training Convolutional Neural Network classifiers to detect and locate planets across a wide range of SNRs. Tested on artificial data, our detectors exhibit good predictive performance and robustness across SNRs. To demonstrate the limits of the detectors, we provide maps of the precision and recall of the model per pixel of the input image. On real data, the models can re-confirm bright source detections., 16 Pages, 6 Figures, 3 Tables, Presented in ECML-PKDD 2019

Transmission spectroscopy provides us with information on the atmospheric properties at the limb, which is often intuitively assumed to be a narrow annulus around the planet. Consequently, studies have focused on the effect of atmospheric horizontal heterogeneities along the limb. Here we demonstrate that the region probed in transmission -- the limb -- extends significantly toward the day and night sides of the planet. We show that the strong day-night thermal and compositional gradients expected on synchronous exoplanets create sufficient heterogeneities across the limb to result in important systematic effects on the spectrum and bias its interpretation. To quantify these effects, we developed a 3D radiative transfer model able to generate transmission spectra of atmospheres based on 3D atmospheric structures. We first produce synthetic JWST observations from a simulation of GJ 1214b and demonstrate the necessity of a real 3D approach to model data for such precise observatories. Second, we investigate how day-night temperature gradients cause a systematic bias in retrieval analysis performed with 1D forward models. For that purpose we synthesize a large set of forward spectra for prototypical HD 209458 b and GJ 1214 b type planets varying the temperatures of the day and night sides as well as the width of the transition region. We then perform typical retrieval analyses and compare the retrieved parameters to the ground truth of the input model. This study reveals systematic biases on the retrieved temperature (found to be higher than the terminator temperature) and abundances. This is due to the fact that the hotter dayside is more extended vertically and screens the nightside. These biases are difficult to detect as the 1D profiles used in the retrieval are found to provide an excellent match to the observed spectra. This needs to be kept in mind when interpreting real data., Comment: Accepted for publication at Astronomy and Astrophysics. Abstract abridged to meet ArXiv size limit

Although there exists a large sample of known exoplanets, little spectroscopic data exists that can be used to study their global atmospheric properties. This deficiency can be addressed by performing phase-resolved spectroscopy -- continuous spectroscopic observations of a planet's entire orbit about its host star -- of transiting exoplanets. Planets with characteristics suitable for atmospheric characterization have orbits of several days, thus phase curve observations are highly resource intensive, especially for shared use facilities. In this work, we show that an infrared spectrograph operating from a high altitude balloon platform can perform phase-resolved spectroscopy of hot Jupiter-type exoplanets with performance comparable to a space-based telescope. Using the EXoplanet Climate Infrared TElescope (EXCITE) experiment as an example, we quantify the impact of the most important systematic effects that we expect to encounter from a balloon platform. We show an instrument like EXCITE will have the stability and sensitivity to significantly advance our understanding of exoplanet atmospheres. Such an instrument will both complement and serve as a critical bridge between current and future space-based near infrared spectroscopic instruments., Comment: 18 pages, 9 figures. Accepted for publication by the Journal of Astronomical Instrumentation

State of the art spectral retrieval models of exoplanet atmospheres assume constant chemical profiles with altitude. This assumption is justified by the information content of current data sets which do not allow, in most cases, for the molecular abundances as a function of pressure to be constrained. In the context of the next generation of telescopes, a more accurate description of chemical profiles may become crucial to interpret observations and gain new insights into atmospheric physics. We explore here the possibility of retrieving pressure-dependent chemical profiles from transit spectra, without injecting any priors from theoretical chemical models in our retrievals. The “two-layer” parameterization presented here allows for the independent extraction of molecular abundances above and below a certain atmospheric pressure. By simulating various cases, we demonstrate that this evolution from constant chemical abundances is justified by the information content of spectra provided by future space instruments. Comparisons with traditional retrieval models show that assumptions made on chemical profiles may significantly impact retrieved parameters, such as the atmospheric temperature, and justify the attention we give here to this issue. We find that the two-layer retrieval accurately captures discontinuities in the vertical chemical profiles, which could be caused by disequilibrium processes—such as photochemistry—or the presence of clouds/hazes. The two-layer retrieval could also help to constrain the composition of clouds and hazes by exploring the correlation between the chemical changes in the gaseous phase and the pressure at which the condensed phase occurs. The two-layer retrieval presented here therefore represents an important step forward in our ability to constrain theoretical chemical models and cloud/haze composition from the analysis of future observations.

Clouds and aerosols provide unique insight into the chemical and physical processes of gas-giant planets. Mapping and characterizing the spectral features indicative of the cloud structure and composition enables an understand-ing of a planet's energy budget, chemistry and atmospheric dynamics. Current space missions to Solar System planets produce high-quality datasets, yet the sheer amount of data obtained often prohibits detailed 'by hand' analyses. Current techniques rely mainly on two approaches: identifying the existence of spectral features by dividing the fluxes of two or more spectral channels, and performing full radiative transfer calculations for individual spectra. The first method is not sufficiently accurate and the second is not easily scalable to the entire planetary surface. Here we have developed a deep learning algorithm, PlanetNet, that is able to quickly and accurately map spatial and spectral features across large, heterogeneous areas of a planet. We use PlanetNet to delineate the major components of the 2008 storm on Saturn, enhancing the scope of the area previously studied and indicating regions that can be probed more deeply with radiative transfer models. Our spectral-component maps indicate compositional and cloud variations of the vast region affected by the storm, showing regions of vertical upwelling, and diminished clouds at the centre of compact substorms, Comment: Nature Astronomy, 28 pages 18 figures

We present a new open source python package, based on PyLightcurve and PyTorch, tailored for efficient computation and automatic differentiation of exoplanetary transits. The classes and functions implemented are fully vectorised, natively GPU-compatible and differentiable with respect to the stellar and planetary parameters. This makes PyLightcurve-torch suitable for traditional forward computation of transits, but also extends the range of possible applications with inference and optimisation algorithms requiring access to the gradients of the physical model. This endeavour is aimed at fostering the use of deep learning in exoplanets research, motivated by an ever increasing amount of stellar light curves data and various incentives for the improvement of detection and characterisation techniques., Comment: 7 pages, 3 figures; submission status updated, fig 2 caption added

Atmospheric characterisation of temperate, rocky planets is the holy grail of exoplanet studies. These worlds are at the limits of our capabilities with current instrumentation in transmission spectroscopy and challenge our state-of-the-art statistical techniques. Here we present the transmission spectrum of the temperate Super-Earth LHS 1140b using the Hubble Space Telescope (HST). The Wide Field Camera 3 (WFC3) G141 grism data of this habitable zone (T$_{\rm{eq}}$ = 235 K) Super-Earth (R = 1.7 $R_\oplus$), shows tentative evidence of water. However, the signal-to-noise ratio, and thus the significance of the detection, is low and stellar contamination models can cause modulation over the spectral band probed. We attempt to correct for contamination using these models and find that, while many still lead to evidence for water, some could provide reasonable fits to the data without the need for molecular absorption although most of these cause also features in the visible ground-based data which are nonphysical. Future observations with the James Webb Space Telescope (JWST) would be capable of confirming, or refuting, this atmospheric detection., Accepted for publication in AJ on 30th October 2020

Spectral retrieval techniques are currently our best tool to interpret the observed exoplanet atmospheric data. Said techniques retrieve the optimal atmospheric components and parameters by identifying the best fit to an observed transmission/emission spectrum. Over the past decade, our understanding of remote worlds in our galaxy has flourished thanks to the use of increasingly sophisticated spectral retrieval techniques and the collective effort of the community working on exoplanet atmospheric models. A new generation of instruments in space and from the ground is expected to deliver higher quality data in the next decade, it is therefore paramount to upgrade current models and improve their reliability, completeness and numerical speed with which they can be run. In this paper, we address the issue of reliability of the results provided by retrieval models in the presence of systematics of unknown origin. More specifically, we demonstrate that if we fit directly individual light-curves at different wavelengths (L-retrieval), instead of fitting transit or eclipse depths, as it is currently done (S-retrieval), the results obtained are more robust against astrophysical and instrumental noise. This new approach is tested, in particular, when discrepant simulated observations from HST/WFC3 and Spitzer/IRAC are combined. We find that while S-retrievals converge to an incorrect solution without any warning, L-retrievals are able to identify potential discrepancies between the data-sets., 14 pages, 15 figures,Submitted to ApJ

Although there are a large number of known exoplanets, there is little data on their global atmospheric properties.Phase-resolved spectroscopy of transiting planets - continuous spectroscopic observation of planets during their full orbits - probes varied depths and longitudes in the atmospheres thus measuring their three-dimensional thermal and chemical structure and contributing to our understanding of their global circulation. Planets with characteristics suitable for atmospheric characterization have orbits of several days, so phase curve observations are highly resource intensive, especially for shared use facilities. The Exoplanet Climate Infrared TElescope (EXCITE) is a balloon-borne near-infrared spectrometer designed to observe from 1 to 5 meters to perform phase-resolved spectroscopy of hot Jupiters. Flying from a long duration balloon (LDB) platform, EXCITE will have the stability to continuously stare at targets for days at a time and the sensitivity to produce data of the quality and quantity needed to signicantly advance our understanding of exoplanet atmospheres. We describe the EXCITE design and show results of analytic and numerical calculations of the instrument sensitivity. We show that an instrument like EXCITE will produce a wealth of quality data, both complementing and serving as a critical bridge between current and future space-based near infrared spectroscopic instruments.

Most exoplanets detected so far have atmospheric T significantly higher than 300K. Often close to their star, they receive an intense UV photons flux that triggers important photodissociation processes. The T dependency of VUV absorption cross sections are poorly known, leading to an undefined uncertainty in atmospheric models. Similarly, data measured at low T similar to that of the high atmosphere of Mars, Venus, and Titan are often lacking. Our aim is to quantify the T dependency of the abs. cross section of important molecules in planetary atmospheres. We want to provide both high-resolution data at T prevailing in these media and a simple parameterization of the absorption in order to simplify its use in photochemical models. This study focuses on carbon dioxide. We performed experimental measurements of CO$_2$ absorption cross section with synchrotron radiation for the wavelength range (115--200nm). For longer wavelengths (195--230nm), we used a deuterium lamp and a 1.5m Jobin-Yvon spectrometer. We used these data in our 1D thermo-photochemical model in order to study their impact on the predicted atmospheric compositions. The cross section of CO$_2$ increases with T. It can be separated in two parts: a continuum and a fine structure superimposed on the continuum. The variation of the continuum of absorption can be represented by the sum of three gaussian functions. Using data at high T in thermo-photochemical models modifies significantly the abundance and the photodissociation rates of many species, in addition to CO$_2$, such as methane and ammonia. These deviations have an impact on synthetic transmission spectra, leading to variations of up to 5 ppm. We present a full set of HR ($\Delta \lambda$=0.03nm) absorption cross sections of CO$_2$ from 115 to 230nm for T ranging from 150 to 800K., Comment: 14 pages, 17 figures, accepted for pulication in A&A

The Exoplanet Characterisation Observatory (EChO) has been studied as a space mission concept by the European Space Agency in the context of the M3 selection process. Through direct measurement of the atmospheric chemical composition of hundreds of exoplanets, EChO would address fundamental questions such as: What are exoplanets made of? How do planets form and evolve? What is the origin of exoplanet diversity? More specifically, EChO is a dedicated survey mission for transit and eclipse spectroscopy capable of observing a large, diverse and well-defined planetary sample within its four to six year mission lifetime. In this paper we use the end-to-end instrument simulator EChOSim to model the currently discovered targets, to gauge which targets are observable and assess the EChO performances obtainable for each observing tier and time. We show that EChO would be capable of observing over 170 relativity diverse planets if it were launched today, and the wealth of optimal targets for EChO expected to be discovered in the next 10 years by space and ground-based facilities is simply overwhelming. In addition, we build on previous molecular detectability studies to show what molecules and abundances will be detectable by EChO for a selection of real targets with various molecular compositions and abundances. EChO's unique contribution to exoplanetary science will be in identifying the main constituents of hundreds of exoplanets in various mass/temperature regimes, meaning that we will be looking no longer at individual cases but at populations. Such a universal view is critical if we truly want to understand the processes of planet formation and evolution in various environments. In this paper we present a selection of key results. The full results are available online (http://www.ucl.ac.uk/exoplanets/echotargetlist/)., Comment: Accepted for publication in Experimental Astronomy, 20 pages, 10 figures, 3 tables

We present observations of WASP-63b by the Hubble Space Telescope (HST) as part of "A Preparatory Program to Identify the Single Best Transiting Exoplanet for JWST Early Release Science". WASP-63b is one of the community targets under consideration for the James Webb Space Telescope (JWST) Early Release Science (ERS) program. We present a spectrum derived from a single observation by HST Wide Field Camera 3 in the near infrared. We engaged groups across the transiting exoplanet community to participate in the analysis of the data and present results from each. There is general agreement amongst all results that we find an H2O absorption feature with 3.5-4.0 sigma significance. However, the feature is muted in comparison to a clear atmosphere at solar composition. Although the detection of the water feature is robust, the reasons for the muting of this feature are ambiguous due to a degeneracy between clouds and composition. The data does not yield robust detections of any molecular species other than H2O. The group was motivated to perform an additional set of retrieval exercises to investigate an apparent bump in the spectrum at ~ 1.55 um. We explore possible disequilibrium chemistry and find this feature is consistent with super-solar HCN abundance but it is questionable if the required mixing ratio of HCN is chemically and physically plausible. The ultimate goal of this study is to vet WASP-63b as a potential community target to best demonstrate the capabilities and systematics of JWST instruments for transiting exoplanet science. In the case of WASP-63b, the presence of a detectable water feature indicates that WASP-63b remains a plausible target for ERS observations., repaired broken citations

International audience; The Atmospheric Remote-Sensing Infrared Exoplanet Large-survey (ARIEL) is one of the three candidate missions selected by the European Space Agency (ESA) for its next medium-class science mission due for launch in 2026. The goal of the ARIEL mission is to investigate the atmospheres of several hundred planets orbiting distant stars in order to address the fundamental questions on how planetary systems form and evolve. During its four (with a potential extension to six) years mission ARIEL will observe 500+ exoplanets in the visible and the infrared with its meter-class telescope in L2. ARIEL targets will include gaseous and rocky planets down to the Earth-size around different types of stars. The main focus of the mission will be on hot and warm planets orbiting close to their star, as they represent a natural laboratory in which to study the chemistry and formation of exoplanets. The ARIEL mission concept has been developed by a consortium of more than 50 institutes from 12 countries, which include UK, France, Italy, Germany, the Netherlands, Poland, Spain, Belgium, Austria, Denmark, Ireland and Portugal. The analysis of the ARIEL spectra and photometric data in the 0.5-7.8 micron range will allow to extract the chemical fingerprints of gases and condensates in the planets' atmospheres, including the elemental composition for the most favorable targets. It will also enable the study of thermal and scattering properties of the atmosphere as the planet orbit around the star. ARIEL will have an open data policy, enabling rapid access by the general community to the high-quality exoplanet spectra that the core survey will deliver.