| Name: Tonmoy Deka |
| Affiliation: National Institute of Science Education and Research, Bhubaneswar |
| Conference ID: ASI2026_109 |
| Title: NEXOTRANS: A Next-Generation Exoplanet Atmospheric Retrieval Framework Applied to the 0.6–12 μm JWST Spectrum of WASP-39 b |
| Abstract Type: Poster |
| Abstract Category: Sun, Solar System, Exoplanets, and Astrobiology |
| Author(s) and Co-Author(s) with Affiliation: Tonmoy Deka(National Institute of Science Education and Research, Bhubaneswar - 752050, India), Tasneem Basra Khan(National Institute of Science Education and Research, Bhubaneswar - 752050, India), Swastik Dewan(National Institute of Science Education and Research, Bhubaneswar - 752050, India), Priyankush Ghosh(National Institute of Science Education and Research, Bhubaneswar - 752050, India), Debayan Das(National Institute of Science Education and Research, Bhubaneswar - 752050, India), Liton Majumdar(National Institute of Science Education and Research, Bhubaneswar - 752050, India) |
| Abstract: The James Webb Space Telescope (JWST) has transformed exoplanet atmospheric studies with its broad wavelength coverage and high spectral precision compared to previous space-based observatories. Accurate characterization of chemical composition and atmospheric structure in planets such as WASP-39 b remains challenging due to complex spectral features and high dimensional parameter spaces, requiring novel data analysis techniques and robust models. In this work, we present NEXOTRANS, a next-generation atmospheric retrieval framework that combines Bayesian nested sampling with machine learning models to analyze JWST transmission spectra efficiently and robustly. NEXOTRANS integrates Bayesian inference (e.g., PyMultiNest) with ensemble machine learning algorithms (Random Forest, Gradient Boosting, and k-Nearest Neighbors) to enable fast exploration of parameter space. We apply NEXOTRANS to JWST NIRISS, NIRSpec PRISM, and MIRI transmission spectra observations of the hot-Saturn exoplanet WASP-39 b spanning 0.6-12 μm. Multiple chemistry models (free, equilibrium, hybrid, and equilibrium-offset) were compared to interpret the data. Our retrievals robustly constrain key molecular abundances (H₂O, CO₂, CO, H₂S) and reveal evidence for disequilibrium chemistry through retrieved SO₂ signatures, consistent with photochemical processes inferred in previous JWST studies. High-altitude aerosols (MgSiO₃, ZnS) influencing the spectrum are also characterized, and we infer supersolar C/O ratios and metallicity. Machine learning retrievals closely match Bayesian posterior constraints, offering a promising avenue for rapid retrievals, particularly valuable for population studies. NEXOTRANS demonstrates an efficient, flexible approach for detailed atmospheric insights and supports future exoplanet studies with JWST and upcoming missions. |
