Characterizing Jovian aerosols using NASA Juno and JWST infrared spectral data

The characterization of atmospheres, particularly their aerosol (clouds and hazes) content, is fundamental to understanding the thermal structure, circulation, and chemical composition of planets, including the Solar System's gas giants and distant exoplanets. Aerosols critically modulate the visible and near-infrared light from planetary bodies, often masking or weakening the gaseous absorption features that atmospheric retrievals rely upon. This challenge is exacerbated by the degeneracy of the inverse problem, where different physical aerosol properties (size, shape, composition) can yield similar observational spectra. This thesis addresses these challenges by performing the first detailed, efficient multiple scattering atmospheric retrievals of Jupiter's aerosols using a combined dataset from the Juno/JIRAM and JWST/NIRSpec missions. Jupiter, serving as a ``Rosetta Stone" for gas giants, presents a complex atmosphere with uncharacterized hazes, chromophores (coloring agents), and deep-seated water clouds, whose properties defy simple thermodynamic prediction. I adopted the Planetary Spectrum Generator (PSG) for accurate multiple scattering analysis of JIRAM data, specifically targeting the solar (2-3.2 $\mu$m) and thermal (3.8-5 $\mu$m) spectral ranges, including the first successful full-range retrievals. To manage the vast data volume from both JIRAM and NIRSpec, I developed \texttt{chopper.py}, an open-source machine learning tool that exploits Principal Component Analysis (PCA) and Gaussian Mixture Models (GMM) to automatically cluster similar spectra. My research focused on four key areas, yielding novel insights. (1) I identified and characterized over 25 Spectrally Identifiable Ammonia Clouds associated with a white turbulent vortex in the Northern Temperate Domain, providing direct evidence of pure ammonia ice in regions where it was previously not detected. (2) Global retrieval constrained the properties of the unknown contaminant materials (hazes and chromophores) responsible for the planet's colors and optical properties and confirmed their photochemical origin. (3) Spectral clustering of JIRAM and NIRSpec data revealed that the GRS is composed of two subregions—a halo and an inner oval—an undetected feature at optical wavelengths, providing new constraints on the storm's fluid dynamics and long-term stability. (4) Furthermore, a quantitative analysis of the GRS's turbulent surrounding ``ripples" was performed, connecting aerosol properties to Jovian fluid dynamics and identifying these areas as potential candidates for the future search for deeper water clouds. Collectively, this thesis work advances the state-of-the-art in Jupiter’s atmosphere modeling by demonstrating the efficacy of combining accurate radiative transfer with machine learning for robust aerosol characterization, paving the way for similar analyses of Saturn, Uranus, Neptune, and exoplanets.