Inverse Modelling Methods for the Characterisation of Exoplanetary Atmospheres

Since the launch of the Kepler Space Telescope in 2009, results now suggest that 1 in 5 Sun-like stars may have habitable planets (Petigura et al., 2013), which potentially holds startling implications for the field of Astrobiology and the search for life elsewhere in the Universe.

In quick succession over the past 20 years, we have gone from planetary detection, to measuring bulk properties (such as density), to, most recently, characterisation of the atmospheres of a selection of these planets. Such atmospheric studies are vital to the long term goal of assessing habitability, as even within our own solar system the case studies of Mars and Venus demonstrate how otherwise 'habitable' planets can prove inhospitable due to their atmospheric makeup. We now face a continual drive to lower mass worlds, bringing the prospect of probing the atmosphere of a true Earth-analogue tantalisingly close to fruition.

One of the issues that has plagued many past attempts to truly characterise exoplanet atmospheres is poorly understood degeneracies. For instance, when we observe the light from a transiting planet-star system, certain spectral features that have been interpreted as absorption due to a given molecule can also be explained by a different atmospheric temperature structure. In the past, single models were run until one fitting the data was found, without a throughout investigation of other combinations of atmospheric parameters that could also explain the data. To circumvent this, so called 'atmospheric retrieval' models have been developed (originally by my supervisor, Dr. Madhusudhan), which run millions of models to fully explore all the possible explanations for the observed behaviour. These take in observational data and use it to infer the properties of the atmosphere (often called 'inverse modelling').

My current research focuses on theoretical modelling of exoplanet transmission spectra using the aforementioned methodology of atmospheric retrieval. My work involves investigating new physical phenomena that can be included in such models, such that a greater information content can be extracted from existing exoplanet transmission spectra. This will prove increasingly vital in the upcoming era of high-precision exoplanet spectroscopy (for instance, from the James Webb Space Telescope), as the quality of data will be sufficient to characterise such worlds in unprecedented detail. This, however, requires extensive theoretical exploration of previously neglected physical effects to ensure the maximum extraction possibility of potential discoveries.