Parameterizing pressure-temperature profiles of exoplanet atmospheres with neural networks

Abstract

Atmospheric retrievals (AR) of exoplanets typically rely on a combination of a Bayesian inference technique and a forward simulator to estimate atmospheric properties from an observed spectrum. A key component in simulating spectra is the pressure-temperature (PT) profile, which describes the thermal structure of the atmosphere. Current AR pipelines commonly use ad hoc fitting functions here that limit the retrieved PT profiles to simple approximations, but still use a relatively large number of parameters. In this work, we introduce a conceptually new, data-driven parameterization scheme for physically consistent PT profiles that does not require explicit assumptions about the functional form of the PT profiles and uses fewer parameters than existing methods. Our approach consists of a latent variable model (based on a neural network) that learns a distribution over functions (PT profiles). Each profile is represented by a low-dimensional vector that can be used to condition a decoder network that maps $P$ to $T$. When training and evaluating our method on two publicly available datasets of self-consistent PT profiles, we find that our method achieves, on average, better fit quality than existing baseline methods, despite using fewer parameters. In an AR based on existing literature, our model (using two parameters) produces a tighter, more accurate posterior for the PT profile than the five-parameter polynomial baseline, while also speeding up the retrieval by more than a factor of three. By providing parametric access to physically consistent PT profiles, and by reducing the number of parameters required to describe a PT profile (thereby reducing computational cost or freeing resources for additional parameters of interest), our method can help improve AR and thus our understanding of exoplanet atmospheres and their habitability.

arXiv commentAccepted for publication in Astronomy & Astrophysics

References in corpus (42)

• SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python
• Array Programming with NumPy
• X-ray spectral modelling of the AGN obscuring region in the CDFS: Bayesian model selection and catalogue
• Normalizing Flows: An Introduction and Review of Current Methods
• A Temperature and Abundance Retrieval Method for Exoplanet Atmospheres
• petitRADTRANS: a Python radiative transfer package for exoplanet characterization and retrieval
• On the radiative equilibrium of irradiated planetary atmospheres
• A Possible Bifurcation in Atmospheres of Strongly Irradiated Stars and Planets
• A Systematic Retrieval Analysis of Secondary Eclipse Spectra I: A Comparison of Atmospheric Retrieval Techniques
• The Pale Orange Dot: The Spectrum and Habitability of Hazy Archean Earth
• Transits of Earth-Like Planets
• Fingering convection and cloudless models for cool brown dwarf atmospheres
• Synthetic Spectra and Colors of Young Giant Planet Atmospheres: Effects of Initial Conditions and Atmospheric Metallicity
• Large Interferometer For Exoplanets (LIFE): I. Improved exoplanet detection yield estimates for a large mid-infrared space-interferometer mission
• Cloudless atmospheres for L/T dwarfs and extra-solar giant planets
• Accuracy tests of radiation schemes used in hot Jupiter global circulation models
• The Effects of Consistent Chemical Kinetics Calculations on the Pressure-Temperature Profiles and Emission Spectra of Hot Jupiters
• On the Absorption and Redistribution of Energy in Irradiated Planets
• The Habitability of Proxima Centauri b: II: Environmental States and Observational Discriminants
• The 13O-rich atmosphere of a young accreting super-Jupiter
• An Analytic Radiative-Convective Model for Planetary Atmospheres
• Detecting planetary geochemical cycles on exoplanets: Atmospheric signatures and the case of SO2
• On the effects of clouds and hazes in the atmospheres of hot Jupiters: semi-analytical temperature-pressure profiles
• Spectra of Earth-like Planets Through Geological Evolution Around FGKM Stars
• An Ensemble of Bayesian Neural Networks for Exoplanetary Atmospheric Retrieval
• The Implications of 3D Thermal Structure on 1D Atmospheric Retrieval
• A Library of Self-Consistent Simulated Exoplanet Atmospheres
• Large Interferometer For Exoplanets (LIFE): V. Diagnostic potential of a mid-infrared space-interferometer for studying Earth analogs
• Large Interferometer For Exoplanets (LIFE): III. Spectral resolution, wavelength range and sensitivity requirements based on atmospheric retrieval analyses of an exo-Earth
• Large Interferometer For Exoplanets (LIFE): II. Signal simulation, signal extraction and fundamental exoplanet parameters from single epoch observations
• Outstanding Challenges of Exoplanet Atmospheric Retrievals
• Neural posterior estimation for exoplanetary atmospheric retrieval
• Exoplanet atmosphere retrievals in 3D using phase curve data with ARCiS: application to WASP-43b
• Effects of UV stellar spectral uncertainty on the chemistry of terrestrial atmospheres
• Aura-3D: A Three-dimensional Atmospheric Retrieval Framework for Exoplanet Transmission Spectra
• Assessing Spectra and Thermal Inversions due to TiO in Hot Jupiter Atmospheres
• Convolutional neural networks as an alternative to Bayesian retrievals
• Large Interferometer For Exoplanets (LIFE): IX. Assessing the Impact of Clouds on Atmospheric Retrievals at Mid-Infrared Wavelengths with a Venus-Twin Exoplanet
• Toward a multidimensional analysis of transmission spectroscopy. Part III: Modelling 2D effects in retrievals with TauREx
• Conditional Neural Process for non-parametric modeling of AGN light curve
• Photoreverberation mapping of quasars in the context of LSST observing strategies
• SVEEEETIES: Singular vector expansion to estimate Earth-like exoplanet temperatures from infrared emission spectra

Cited by in corpus (3)

• Earth as an Exoplanet. III. Using Empirical Thermal Emission Spectra as an Input for Atmospheric Retrieval of an Earth-twin Exoplanet
• Flow Matching for Atmospheric Retrieval of Exoplanets: Where Reliability meets Adaptive Noise Levels
• Pursuing Truth: Improving Retrievals on Mid-Infrared Exo-Earth Spectra with Physically Motivated Water Abundance Profiles and Cloud Models