Resumen:
Although the analysis of exoplanet atmospheres has become one of the most pertinent topics within planetary sciences, characterising these objects directly from their spectra is still a challenge. To interpret the observed spectrum of an exo-atmosphere, one can apply a technique known as atmospheric retrieval, i.e. fitting a model to this data in order to infer the properties of the atmosphere, such as temperature, chemical composition, and presence of clouds. This work considers a retrieval framework which includes H2O as the main molecular opacity source, and optional opacity features such as additional molecules (e.g. CH4, CO2, CO), collision-induced absorption (CIA), Rayleigh scattering, and clouds. Furthermore, our retrieval code accounts for a non-isobaric transit chord, setting atmospheric opacity sources to a full pressure dependency. We perform non-isobaric retrievals in 38 Hubble Space Telescope Wide Field Camera 3 (WFC3) near-infrared transmission spectra to establish how the variation of pressure affects the estimation of atmospheric parameters. Our results show that, for WFC3 wavelength range and resolution, mainly H2O-only cloud-free and constant/grey cloud atmospheres are retrieved. We compare our findings with previous analyses in the literature, concluding that Rayleigh scattering is negligible in most of our retrievals, except in the ones where shorter-wavelength WFC3 data is available. On the other hand, CIA is strongly dependent on pressure, therefore helping set the H2O abundance, which is unclear in the former isobaric studies. Additionally, we acknowledge the degeneracy between molecular abundances and the reference parameter (i.e. where the atmosphere is optically thick), as pointed out by previous studies, may be broken for cloud-free fits, but is still indirectly present in cloudy results. Finally, we suggest new approaches that could help identify additional atmospheric features imprinted in the spectra, considering data in complementary wavelengths, as well as retrieval analyses using higher-quality James Webb Space Telescope spectra.