JWST is peering into the atmospheres of giant exoplanets by measuring their thermal glow, but current models cannot properly interpret the 3D information hidden in these high-precision observations. The problem is that existing atmospheric models treat these planets as simple, uniform spheres. In reality, a giant exoplanet’s dayside has dramatic variations in temperature and chemistry from equator to pole, and from the upper atmosphere down to deeper layers. When models ignore these 3D effects, they introduce systematic errors that bias measurements of what the planet is made of and how its heat is distributed. Initial JWST datasets have already been limited by this model uncertainty, leaving a gap between the telescope’s capability and our ability to extract the full story. This project will develop and test a hierarchy of multi-dimensional retrieval models—from simple to complex—to determine how much 3D detail is needed for different types of planets. It will also create a new method tailored to eclipse mapping, building on an existing algorithm called 'Eigenspectra' that identifies distinct spectral regions across a planet’s dayside. If successful, the work will unlock the full potential of JWST’s groundbreaking observations, revealing how giant exoplanets circulate heat, mix chemicals, and structure their atmospheres in three dimensions. This is fundamental science: it will not change daily life tomorrow, but it will transform our understanding of planetary physics across the galaxy.
View original technical description
JWST is revolutionising the study of giant exoplanet atmospheres, and recent observations have allowed unprecedentedly precise measurements of their chemical compositions and thermal structures. In particular, thermal emission measurements of these exoplanets are extremely sensitive to their physicochemical properties. In contrast to other techniques such as transmission spectroscopy, thermal emission observations are not significantly inhibited by stellar contamination effects and clouds/hazes. However, current models are not sufficient to interpret the high-precision emission observations of giant exoplanets made possible by JWST. These spectra are sensitive to 3D effects, which can in turn bias atmospheric inferences unless correctly modelled. Indeed, the analysis of some initial JWST datasets has been limited by model uncertainty, highlighting a critical need for improved models to capitalise on the 3D information in these rich datasets. In this STFC project, we will investigate and develop methods to extract accurate and multidimensional information from emission spectrum observations of giant exoplanets with JWST. Our first objective is to test a hierarchy of multi-dimensional atmospheric retrieval models against real and simulated JWST observations, assessing the degree of multidimensionality required when interpreting observations in different thermal and chemical regimes. In particular, we will investigate different parameterisations for inhomogeneous dayside temperature profiles and varying-with-depth chemical abundance profiles. Our second objective is to develop a novel atmospheric retrieval method tailored to eclipse mapping observations. This method will map variations in chemistry and vertical temperature profile across the daysides of giant exoplanets. We will build on an existing eclipse mapping algorithm, 'Eigenspectra', which is able to identify distinct regions of the planet which have similar spectra. We will develop atmospheric retrieval models which account for the relevant geometric factors in order to measure the atmospheric compositions and thermal structures of these regions across the planetary dayside. Modelling 3D effects in giant exoplanet atmospheres is essential in the era of JWST. The proposed work will unlock the full potential of these groundbreaking observations, revealing new insights into the atmospheric physics of giant exoplanets.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know