Neural posterior estimation for exoplanetary atmospheric retrieval

Abstract

Retrieving the physical parameters from spectroscopic observations of exoplanets is key to understanding their atmospheric properties. Exoplanetary atmospheric retrievals are usually based on approximate Bayesian inference and rely on sampling-based approaches to compute parameter posterior distributions. Accurate or repeated retrievals, however, can result in very long computation times due to the sequential nature of sampling-based algorithms. We aim to amortize exoplanetary atmospheric retrieval using neural posterior estimation (NPE), a simulation-based inference algorithm based on variational inference and normalizing flows. In this way, we aim (i) to strongly reduce inference time, (ii) to scale inference to complex simulation models with many nuisance parameters or intractable likelihood functions, and (iii) to enable the statistical validation of the inference results. We evaluate NPE on a radiative transfer model for exoplanet spectra petitRADTRANS, including the effects of scattering and clouds. We train a neural autoregressive flow to quickly estimate posteriors and compare against retrievals computed with MultiNest. NPE produces accurate posterior approximations while reducing inference time down to a few seconds. We demonstrate the computational faithfulness of our posterior approximations using inference diagnostics including posterior predictive checks and coverage, taking advantage of the quasi-instantaneous inference time of NPE. Our analysis confirms the reliability of the approximate posteriors produced by NPE. The accuracy and reliability of the inference results produced by NPE establishes it as a promising approach for atmospheric retrievals. Amortization of the posterior inference makes repeated inference on several observations computationally inexpensive since it does not require on-the-fly simulations, making the retrieval efficient, scalable, and testable.

arXiv commentThe paper has been submitted to AandA after a final revision

References in corpus (39)

• MultiNest: an efficient and robust Bayesian inference tool for cosmology and particle physics
• Multimodal nested sampling: an efficient and robust alternative to MCMC methods for astronomical data analysis
• X-ray spectral modelling of the AGN obscuring region in the CDFS: Bayesian model selection and catalogue
• Importance Nested Sampling and the MultiNest Algorithm
• Precipitating Condensation Clouds in Substellar Atmospheres
• petitRADTRANS: a Python radiative transfer package for exoplanet characterization and retrieval
• A Systematic Retrieval Analysis of Secondary Eclipse Spectra I: A Comparison of Atmospheric Retrieval Techniques
• High C/O Ratio and Weak Thermal Inversion in the Very Hot Atmosphere of Exoplanet WASP-12b
• How to Distinguish between Cloudy Mini-Neptunes and Water/Volatile-Dominated Super-Earths
• HD 209458b in New Light: Evidence of Nitrogen Chemistry, Patchy Clouds and Sub-Solar Water
• Tau-REx I: A next generation retrieval code for exoplanetary atmospheres
• An ultrahot gas-giant exoplanet with a stratosphere
• Real-time gravitational-wave science with neural posterior estimation
• Model atmospheres of irradiated exoplanets: The influence of stellar parameters, metallicity, and the C/O ratio
• Retrieving scattering clouds and disequilibrium chemistry in the atmosphere of HR 8799e
• A Systematic Retrieval Analysis of Secondary Eclipse Spectra II: A Uniform Analysis of Eight Planets and Their C to O Ratios
• HAT-P-26b: A Neptune-Mass Exoplanet with a Well Constrained Heavy Element Abundance
• Observing transiting planets with JWST -- Prime targets and their synthetic spectral observations
• Tau-REx II: Retrieval of emission spectra
• Massive optimal data compression and density estimation for scalable, likelihood-free inference in cosmology
• H2O abundances in the atmospheres of three hot Jupiters
• HELIOS-Retrieval: An Open-source, Nested Sampling Atmospheric Retrieval Code, Application to the HR 8799 Exoplanets and Inferred Constraints for Planet Formation
• Retrieval of atmospheric properties of cloudy L dwarfs
• The Kepler-19 System: A Transiting 2.2 R_Earth Planet and a Second Planet Detected via Transit Timing Variations
• Treatment of overlapping gaseous absorption with the correlated-k method in hot Jupiter and brown dwarf atmosphere models
• Retrieval of Exoplanet Emission Spectra with HyDRA
• ExoGAN: Retrieving Exoplanetary Atmospheres Using Deep Convolutional Generative Adversarial Networks
• An Ensemble of Bayesian Neural Networks for Exoplanetary Atmospheric Retrieval
• Fast and Credible Likelihood-Free Cosmology with Truncated Marginal Neural Ratio Estimation
• The Water Abundance of the Directly Imaged Substellar Companion κ And b Retrieved from a Near Infrared Spectrum
• A neural simulation-based inference approach for characterizing the Galactic Center $\gamma$-ray excess
• Accurate Machine Learning Atmospheric Retrieval via a Neural Network Surrogate Model for Radiative Transfer
• Assessment of Supervised Machine Learning for Atmospheric Retrieval of Exoplanets
• Dynamical mass inference of galaxy clusters with neural flows
• Inference of cosmic-ray source properties by conditional invertible neural networks
• Dust in brown dwarfs and extra-solar planets. VII. Cloud formation in diffusive atmospheres
• Optimizing exoplanet atmosphere retrieval using unsupervised machine-learning classification
• Retrieval Analysis of the Emission Spectrum of WASP-12b: Sensitivity of Outcomes to Prior Assumptions and Implications for Formation History
• Convolutional neural networks as an alternative to Bayesian retrievals

Cited by in corpus (22)

• The ESO SupJup Survey I: Chemical and isotopic characterisation of the late L-dwarf DENIS J0255-4700 with CRIRES${}^{+}$
• LtU-ILI: An All-in-One Framework for Implicit Inference in Astrophysics and Cosmology
• Evidence for SiO cloud nucleation in the rogue planet PSO J318
• The JWST Weather Report: retrieving temperature variations, auroral heating, and static cloud coverage on SIMP-0136
• Isolated pulsar population synthesis with simulation-based inference
• Machine Learning for Exoplanet Detection in High-Contrast Spectroscopy: Revealing Exoplanets by Leveraging Hidden Molecular Signatures in Cross-Correlated Spectra with Convolutional Neural Networks
• Grid-Based Atmospheric Retrievals for Reflected-Light Spectra of Exoplanets using PSGnest
• Parameterizing pressure-temperature profiles of exoplanet atmospheres with neural networks
• Flow Matching for Atmospheric Retrieval of Exoplanets: Where Reliability meets Adaptive Noise Levels
• Full-waveform earthquake source inversion using simulation-based inference
• Near-instantaneous Atmospheric Retrievals and Model Comparison with FASTER
• The Galactic population of magnetars : a simulation-based inference study
• Disentangling disc and atmospheric signatures of young brown dwarfs with JWST/NIRSpec
• Radiative Transfer and Inversion codes for characterizing planetary atmospheres: an overview
• Differentiable Modeling of Planet and Substellar Atmosphere: High-Resolution Emission, Transmission, and Reflection Spectroscopy with ExoJAX2
• Radio pulsar population synthesis with consistent flux measurements using simulation-based inference
• A normalizing flow approach for the inference of star cluster properties from unresolved broadband photometry I: Comparison to spectral energy distribution fitting
• Physically Interpretable Probabilistic Domain Characterization
• JWST high-contrast spectroscopy with speckle modelling: Atmospheric retrievals of the T dwarf companion HD 19467 B
• $\text{Exoformer}$: Accelerating Bayesian atmospheric retrievals with transformer neural networks
• ASBI: Leveraging Informative Real-World Data for Active Black-Box Simulator Tuning
• Panchromatic characterization of the Y0 brown dwarf WISEP J173835.52+273258.9 using JWST/MIRI