Evidence for H$_{2}$ Dissociation and Recombination Heat Transport in the Atmosphere of KELT-9b

Phase curve observations provide an opportunity to study the full energy budgets of exoplanets by quantifying the amount of heat redistributed from their daysides to their nightsides. Theories explaining the properties of phase curves for hot Jupiters have focused on the balance between radiation and dynamics as the primary parameter controlling heat redistribution. However, recent phase curves have shown deviations from the trends that emerge from this theory, which has led to work on additional processes that may affect hot Jupiter energy budgets. One such process, molecular hydrogen dissociation and recombination, can enhance energy redistribution on ultra-hot Jupiters with temperatures above $\sim2000$ K. In order to study the impact of H$_{2}$ dissociation on ultra-hot Jupiters, we present a phase curve of KELT-9b observed with the Spitzer Space Telescope at 4.5 $\mu$m. KELT-9b is the hottest known transiting planet, with a 4.5-$\mu$m dayside brightness temperature of $4566^{+140}_{-136}$ K and a nightside temperature of $2556^{+101}_{-97}$ K. We observe a phase curve amplitude of $0.609 \pm 0.020$ and a hot spot offset of $18.7^{+2.1}_{-2.3}$ degrees. The observed amplitude is too small to be explained by a simple balance between radiation and advection. General circulation models (GCMs) and an energy balance model that include the effects of H$_{2}$ dissociation and recombination provide a better match to the data. The GCMs, however, predict a maximum hot spot offset of $5$ degrees, which disagrees with our observations at $>5\sigma$ confidence. This discrepancy may be due to magnetic effects in the planet's highly ionized atmosphere.