Circulation models and JWST observations of inflated ultra-hot Jupiters

Introduction: Recent advances in observation with the JWST and high-resolution ground-based instruments have enabled the study of exoplanets to progress towards atmospheric characterisation, as opposed to merely detection. Hot and ultra-hot Jupiters remain among the best characterised and studied class of exoplanet, due to their large sizes and close orbits, however how the internal heating and resulting radius inflation of bloated ultra-hot Jupiters and related coupling to the internal magnetic field impacts their atmospheric circulation remains an open question. Moreover, the impact of atmospheric dynamics on observable properties can now be studied in detail. This study investigates the effect of varying both atmospheric drag and internal heat flux on the observable properties of WASP-76b, with comparisons made to JWST NIRSpec white-light phase curves. In addition, we perform a broader parameter sweep using the SPARC/MITgcm to investigate the influence of internal heating and inflated radii on the observable properties of hot and ultra-hot Jupiters.

Methods: A suite of general circulation models are run, which solve the primitive equations of meteorology coupled to non-grey correlated-k radiative transfer with the SPARC/MITgcm [1]. The effect of Lorentz forces is represented by changing a spatially constant drag timescale , and for WASP-76b we consider two different internal heat fluxes for comparison, across the range of predicted values for hot and ultra-hot Jupiters [2]. We then will perform a broader parameter sweep, exploring the space of inflated-radii hot and ultra-hot Jupiters by covering a range of irradiation levels from zero-albedo full-redistribution equilibrium temperatures of 1000 – 3200K, again using the SPARC/MITgcm. This parameter space is inclusive of most inflated gas-giant planets, excluding KELT-9b, and will allow us to study the impact of internal heating on atmospheric circulation, with interior heating and evolution modelled using MESA [3]. We then use the gCMCRT radiative transfer code [4] to post-process the GCM results to produce simulated phase curves.

Results: The key result from this study is shown in Figure 1, with simulated phase curves shown in comparison to Spitzer telescope data [5] at 3.6mm. We make the comparison to Spitzer data here as a placeholder for the comparison to JWST NIRSpec data, as the JWST data is not yet published. Figure 3 shows the impact of the interior heat flux on the internal temperature structure of WASP-76b. There is no observable difference between the interior heat flux scenarios. Figures 2 and 4 show characteristics of the atmospheric dynamics and temperature structure. Strong drag acts to suppress all winds throughout the atmosphere, as is expected, while intermediate drag removes the offset of the hot spot due to the suppression of the deep equatorial jet. There is a strong equatorial jet within the deep atmosphere, and the T-P profile implies that cloud species Al₂O₄ and Mg₂SiO₄ can form on the night-side and terminators of WASP-76b, and within its deep atmosphere.

Conclusions: Spitzer data is best matched by a strong () drag case. There is no potentially observable difference between the hot interior flux and cold interior flux. The comparisons of these simulated phase curve to JWST NIRSpec white-light phase curves will help further constrain drag in the ultra-hot regime, which will be a useful point of comparison to other ultra-hot Jupiters. Other ultra-hot Jupiters with Spitzer phase-curves (WASP-18b [6], WASP-103b [7], WASP-121b [8]) also show high dayside-nightside temperature differences. This may imply that the drag mechanisms are similar in each planet in the ultra-hot regime (~2000-2500 K). New JWST NIRSpec/G395H phase-curve data (JWST GO proposal 5268) will also constrain metallicity, breaking the drag/metallicity degeneracy. The similarity in deep-atmosphere temperature shown by Figure 3 motivates the need for a parameter sweep where the temperature at the bottom boundary is varied, as opposed to an interior heat flux, in order to speed up convergence. Likewise, the T-P profile in Figure 4 motivates the need for longer simulation runs, as the model has not reached equilibrium within the deep atmosphere.

References:

[1] Showman, A.P. et al. (2009), The Astrophysical Journal, 699(1), pp. 564–584.

[2] Thorngren, D. et al. (2019), ApJL (Vol. 884, Issue 1)

[3] Jermyn, A.S. et al. (2023), The Astrophysical Journal Supplement Series, 265, p. 15.

[4] Lee, E.K. et al. (2022), The Astrophysical Journal, 929(2), p. 180

[5] May, E.M. et al. (2021), The Astronomical Journal, 162(4), p. 158.

[6] Maxted, P.F. et al. (2012), Monthly Notices of the Royal Astronomical Society, 428(3), pp. 2645–2660

[7] Kreidberg, L. et al. (2018), The Astronomical Journal, 156(1), p. 17

[8] Davenport, B. et al. (2025),  Available at: https://arxiv.org/abs/2503.12521 (Accessed: 20 March 2025).