Developing and Testing a Novel Physical Model for Ultra-hot Jupiter Atmospheres

Over the past few years, the exoplanet research community has been discovering that hot Jupiters with dayside temperatures above ~2500 K exhibit distinct characteristics compared to their colder peers. My thesis work provided new insights into the atmospheric composition and circulation of these “ultra-hot” Jupiters through physical modelling and observations of their eclipses and phase variations.

With the second-ever spectrally resolved optical eclipse spectrum of a hot Jupiter, I showed that the ~3000 K dayside of the ultra-hot Jupiter WASP-12b was devoid of clouds and consisted primarily of atomic hydrogen and helium. This showed that the dayside atmospheres of ultra-hot Jupiters are similar to those of stars, with most molecules—including the dominant constituent H₂—becoming thermally dissociated. Realizing that the enormous day-night temperature contrast in ultra-hot Jupiter atmospheres would result in atomic hydrogen from the dayside recombining into molecules nearer the nightside, I developed a semi-analytical model accounting for the thermodynamic impacts of this H₂↔2H process. This model qualitatively explained the unusually large phase offsets of WASP-12b and WASP-33b known at the time and successfully predicted the large phase offset and extremely hot nightside temperature of KELT-9b.

I then co-wrote SPCA, an open-source pipeline for Spitzer phase curve observations, which I used to show that the highly unusual phase curve observations of WASP-12b were reproducible and were best explained by CO emission from a stream of gas being stripped from the planet. SPCA analyses of the phase curves of hot Jupiter Qatar-1b and ultra-hot Jupiters KELT-16b and MASCARA-1b measured the impact of differing Coriolis forces on hot Jupiter atmospheric circulation. Finally, I used SPCA to perform the first comprehensive reanalysis of 4.5 μm Spitzer phase curves. I found that phase curve parameters are usually independent of the detector model used and my reanalyses were often consistent with the literature. Using my uniform reanalyses, I also confirmed numerous population-level trends such as that of Bond albedo with planetary surface gravity which is likely driven by differences in cloud sedimentation rates.

In summary, my thesis work provides new insights into the atmospheres of ultra-hot Jupiters through numerous observations and novel physical modelling. This body of work provides a strong foundation for future observational studies of atmospheric phenomena, with an emphasis on improving reproducibility and progressing towards studies spanning a wider range of planetary properties.