Vivien Parmentier

Abstract:

We present one of the most precise emission spectra of an exoplanet observed so far. We combine five secondary eclipses of the hot Jupiter WASP-18b (T day ~ 2900 K) that we secured between 1.1 and 1.7 μm with the Wide Field Camera 3 instrument on board the Hubble Space Telescope. Our extracted spectrum (S/N = 50, R ~ 40) does not exhibit clearly identifiable molecular features but is poorly matched by a blackbody spectrum. We complement this data with previously published Spitzer/Infrared Array Camera observations of this target and interpret the combined spectrum by computing a grid of self-consistent, 1D forward models, varying the composition and energy budget. At these high temperatures, we find there are important contributions to the overall opacity from H− ions, as well as the removal of major molecules by thermal dissociation (including water), and thermal ionization of metals. These effects were omitted in previous spectral retrievals for very hot gas giants, and we argue that they must be included to properly interpret the spectra of these objects. We infer a new metallicity and C/O ratio for WASP-18b, and find them well constrained to be solar ([M/H] = −0.01 ± 0.35, C/O ≺ 0.85 at 3σ confidence level), unlike previous work but in line with expectations for giant planets. The best-fitting self-consistent temperature–pressure profiles are inverted, resulting in an emission feature at 4.5 μm seen in the Spitzer photometry. These results further strengthen the evidence that the family of very hot gas giant exoplanets commonly exhibit thermal inversions.

Abstract:

ABSTRACT The determination of atmospheric structure and molecular abundances of planetary atmospheres via spectroscopy involves direct comparisons between models and data. While varying in sophistication, most model spectra comparisons fundamentally assume one-dimensional (1D) model physics. However, knowledge from general circulation models and of solar system planets suggests that planetary atmospheres are inherently three-dimensional in their structure and composition. We explore the potential biases resulting from standard “1D” assumptions within a Bayesian atmospheric retrieval framework. Specifically, we show how the assumption of a single 1D thermal profile can bias our interpretation of the thermal emission spectrum of a hot Jupiter atmosphere that is composed of two thermal profiles. We retrieve spectra of unresolved model planets as observed with a combination of the Hubble Space Telescope Wide Field Camera 3 (WFC3)+Spitzer Infrared Array Camera (IRAC) as well as the James Webb Space Telescope (JWST) under varying differences in the two thermal profiles. For WFC3+IRAC, there is a significantly biased estimate of CH4 abundance using a 1D model when the contrast is 80%. For JWST, two thermal profiles are required to adequately interpret the data and estimate the abundances when contrast is greater than 40%. We also apply this preliminary concept to the recent WFC3+IRAC phase curve data of the hot Jupiter WASP-43b. We see similar behavior as present in our simulated data: while the abundance determination is robust, CH4 is artificially well-constrained to incorrect values under the 1D assumption. Our work demonstrates the need to evaluate model assumptions in order to extract meaningful constraints from atmospheric spectra and motivates exploration of optimal observational setups.