Document Type


Publication Date



We present thermal phase curve measurements for the hot Jupiter WASP-103b observed with Hubble/WFC3 and Spitzer/IRAC. The phase curves have large amplitudes and negligible hotspot offsets, indicative of poor heat redistribution to the nightside. We fit the phase variation with a range of climate maps and find that a spherical harmonics model generally provides the best fit. The phase-resolved spectra are consistent with blackbodies in the WFC3 bandpass, with brightness temperatures ranging from 1880 ± 40 K on the nightside to 2930 ± 40 K on the dayside. The dayside spectrum has a significantly higher brightness temperature in the Spitzer bands, likely due to CO emission and a thermal inversion. The inversion is not present on the nightside. We retrieved the atmospheric composition and found that it is moderately metal-enriched ($[{\rm{M}}/{\rm{H}}]={23}_{-13}^{+29}\times \mathrm{solar}$) and the carbon-to-oxygen ratio is below 0.9 at 3σ confidence. In contrast to cooler hot Jupiters, we do not detect spectral features from water, which we attribute to partial H2O dissociation. We compare the phase curves to 3D general circulation models and find that magnetic drag effects are needed to match the data. We also compare the WASP-103b spectra to brown dwarfs and young, directly imaged companions. We find that these objects have significantly larger water features, indicating that surface gravity and irradiation environment play an important role in shaping the spectra of hot Jupiters. These results highlight the 3D structure of exoplanet atmospheres and illustrate the importance of phase curve observations for understanding their complex chemistry and physics.

Available for download on Saturday, January 01, 2022