The Atmosphere of HD 149026b: Low Metal Enrichment and Weak Energy Transport

Abstract

Recent JWST eclipse spectra of the high-density hot Saturn HD 149026b between 2.35 and 5.08 μm have allowed for in-depth study of its atmosphere. To understand its atmospheric properties, we have created a grid of 1D radiative-convective–thermochemical–equilibrium atmosphere models and spectra with PICASO 3.0. In agreement with previous work, we find that the presence of gaseous TiO creates a thermal inversion, which is inconsistent with the data. The presence of gaseous VO, however, which condenses at temperatures 200 K cooler, does not cause such inversions but alters the temperature–pressure profile of the atmosphere. We estimate an atmospheric metallicity of 14 − 8 + 12 × solar without VO and 20 − 8 + 11 × solar with VO, a factor of ∼10 times smaller than previous work from Bean et al., who relied on atmosphere retrievals. We attribute this significant difference in metallicity to a larger temperature gradient at low pressures in radiative equilibrium models. Such models with lower metallicities readily fit the strong CO2 feature at 4.3 μm. Our lower estimated metallicity makes HD 149026b more consistent with the mass–metallicity relationship for other giant planets. We find a C/O ratio of 0.67 − 0.27 + 0.06 with and without VO. The best-fit heat redistribution factor without VO is 1.17, a very high value suggesting very little dayside energy transport and no energy transport to the nightside. The heat redistribution factor shrinks to a more plausible value of 0.91 − 0.05 + 0.05 , with VO, which we regard as circumstantial evidence for the molecule in the atmosphere of HD 149026b.

Evidence of Water Vapor in the Atmosphere of a Metal-rich Hot Saturn with High-resolution Transmission Spectroscopy

“…One important thing to note is that we are probing the terminator region of the planet, unlike Bean et al (2023). Since the compositions of the dayside and nightside of such hot gas giants might be significantly different, as indicated by the planet's inefficient day-to-night heat redistribution and a temperature contrast that reaches ∼800 K (Zhang et al 2018;Gagnebin et al 2024), this means that we might not get the same C/O ratio through transmission observations as that found by Bean et al (2023). We also emphasize that our data might not be sensitive enough to detect HCN even with a C/O ratio as high as 3.00 (see Section 5.4).…”

“…A recent study from Gagnebin et al (2024) analyzed the same JWST/NIRCam emission data set as that of Bean et al (2023). They used the grid-retrieval technique (instead of free retrieval as in Bean et al 2023) 2.09 0.32 0.35 , which is likely to be caused by the difference in the (retrieved) T-P profile shapes in the two studies (Gagnebin et al 2024).…”

“…A recent study from Gagnebin et al (2024) analyzed the same JWST/NIRCam emission data set as that of Bean et al (2023). They used the grid-retrieval technique (instead of free retrieval as in Bean et al 2023) 2.09 0.32 0.35 , which is likely to be caused by the difference in the (retrieved) T-P profile shapes in the two studies (Gagnebin et al 2024). Following this, we perform a simple analysis to find the preferred model metallicity for our data by performing the grid search as discussed in Section 5.5 for the range −4 [M/H] 4 with steps of 0.1 dex and fix the isothermal temperature to 1700 K as well as K p , V rest , and α to the retrieved values as shown in Figure 13.…”