Multi-instrument atmospheric characterization of the warm super-Neptune WASP-107b using GUIBRUSH®

The detection and characterization of exoplanetary atmospheres are fundamental to advancing our understanding of planetary formation and evolution, helping to understand the complex physical and chemical processes that govern how these distant worlds develop and change over time. Additionally, atmospheric analysis enables the assessment of planetary habitability and the identification of organic molecules (such as oxygen, O2 , methane, CH4 , etc.) that could indicate biological activity. WASP-107 b is a warm super-Neptune with a short orbital period around a K6 star and a remarkably low density (ρ = 30.134+0.015−0.013 g/cm , Piaulet et al. (2021)), resulting in an extended atmosphere with a large scale height that enhances atmospheric signal detectability, making it an ideal candidate for transmission spectroscopic analysis using both high-resolution (HR) and low-resolution (LR) instruments. This thesis aims to characterize the atmosphere of WASP-107 b using a multi-instrument approach that leverages both HR and LR spectroscopic data. This analysis is accomplished using GUIBRUSH®, a novel retrieval tool developed during this thesis, capable of handling both HR and LR data within a unified Bayesian framework. One set of HR observations is obtained from GIANO-B, a near-infrared spectrograph with a resolving power of R ≈ 48000 mounted on the Telescopio Nazionale Galileo (La Palma). Another set of HR data comes from IGRINS, a near-infrared spectrograph with R ≈ 45000 at the Gemini South Observatory (Chile). The LR datasets are collected with the Wide Field Camera 3 (WFC3) on the Hubble Space Telescope (HST), spanning 0.8–1.7 μm, and from two instruments on the James Webb Space Telescope (JWST): the Near InfraRed Camera (NIRCam) covering 2.4–5.0 μm and the Mid InfraRed Instrument (MIRI) extending to 5–12 μm. This thesis research is structured in three phases. First, I reanalyzed low-resolution HST and JWST spectra, confirming previous literature results and validating the functionality of GUIBRUSH® for independent analysis. Second, I reduced GIANO-B data, removed telluric and stellar contamination, and applied cross-correlation techniques to search for molecular detections, repeating the analysis for IGRINS data and conducting initial high-resolution retrievals. This phase revealed detections only for H2O and CO, highlighting significant methodological challenges, including optimizing Principal Component Analysis efficiency for stellar and telluric contamination removal and determining the minimum transit observations required for robust characterization of planets sharing similar characteristics to WASP-107 b Finally, I combined HR and LR datasets to achieve comprehensive atmospheric characterization, enabling more precise constraints on molecular abundances and cloud properties by breaking the degeneracies inherent in single-dataset analyses and revealing previously unattainable parameter correlations. This last phase demonstrated that, while HR data had limitations due to systematic noise, their inclusion helped reduce degeneracies in molecular abundance estimates.