The bulk metal content of WASP-80 b from joint interior-atmosphere retrievals: Breaking degeneracies and exploring biases with panchromatic spectra

TL;DR

This work presents a joint interior-atmosphere retrieval framework combining GASTLI interior models with petitRADTRANS atmospheres to extract WASP-80 b’s bulk metal mass fraction and atmospheric composition from panchromatic JWST/HST data. By simultaneously fitting mass, radius, and age with transmission and emission spectra, the study reveals degeneracies between envelope chemistry and internal structure, and demonstrates that including emission data and planetary age tightens constraints on $Z_{planet}$ and $M/H$. Two fiducial joint scenarios emerge: secular cooling with modest atmospheric metallicity ($M/H ≈ 2.75×$ solar, $C/O ≈ 0.12$) and a heating scenario with much higher metallicity ($M/H ≈ 10×$ solar, $M_{core} ≈ 32 M_{igoplus}$), illustrating how internal heating can reconcile bulk density with metal enrichment. The results imply late accretion of solids, potential Ohmic heating, and degeneracies that require careful modeling of clouds and chemistry; these findings highlight the value of comprehensive, multi-geometry data to inform planet formation theories.

Abstract

WASP-80 b is an unusually low-density exoplanet in tension with the metal-rich composition expected for a planet of its mass. We aim to derive precise constraints on WASP-80 b's bulk metal mass fraction, atmospheric composition, and thermal structure. We conducted a suite of retrievals using three approaches: traditional interior-only, atmosphere-only, and joint interior-atmosphere retrievals. We coupled the open-source models GASTLI and petitRADTRANS, which describe planetary structure and thermal evolution, and atmospheric chemistry and clouds, respectively. Our retrievals combine mass and age with panchromatic spectra from JWST and HST in both transmission (0.5-4 $μ$m) and emission (1-12 $μ$m) as observational constraints. We identify two fiducial scenarios. In the first, WASP-80 b has an internal temperature consistent with its age in the absence of external heating sources, and its atmosphere is in chemical equilibrium, with an atmospheric metallicity M/H = 2.75$^{+0.88}_{-0.56}$x solar, a bulk metal mass fraction $Z_{planet}=0.12\pm0.02$, and a core mass $M_{core}=3.49^{+3.49}_{-1.59} \ M_{\oplus}$. In the second scenario, WASP-80 b may be inflated by an additional heat source - possibly induced by magnetic fields - with an atmospheric metallicity M/H = 10.00$^{+8.20}_{-4.75}$x solar, $Z_{planet}=0.28\pm0.11$, and $M_{core}=31.8^{+21.3}_{-17.5} \ M_{\oplus}$. The super-solar M/H and sub-solar C/O ratios in both scenarios suggest late pebble or planetesimal accretion, while additional heating is required to reconcile the data with the more massive core predicted by the core accretion paradigm. In general, joint retrievals are inherently affected by a degeneracy between atmospheric chemistry and internal structure. Together with flexible cloud treatment and an unweighted likelihood, this leads to larger uncertainties in bulk and atmospheric compositions than previously claimed.

Figures (13)

• Figure 1: Comparison between our atmospheric composition estimates and previous work from transmission-only (upper panel) and emission-only (lower panel) retrievals. Previous work include Bell23 (B23, orange), Wong22 (W22, dark gold) and Wiser25 (Wi25, violet). The shaded area indicates the region forbidden by our interior-only retrievals at 4$\sigma$ (see Sect. \ref{['sec:interior_only_retrievals']}).
• Figure 2: Pressure-Temperature (PT) profile of WASP-80 b as constrained by our two emission-only retrievals. The shaded region indicates the 1$\sigma$ area of the retrieval, while solid lines correspond to the mean. We show self-consistent 1D models from Molliere17 for comparison, including cloudy model 1 in their table 2 (see text). The grey dotted line shows the contribution function from our emission-only retrievals.
• Figure 3: Summary of the mean forward model from one of our two fiducial interior-atmosphere joint retrievals, JR6, of WASP-80 b. This model corresponds to a core mass fraction (CMF) of 0.02, an atmospheric metallicity of M/H = 2.75 $\times$ solar, C/O = 0.12, and $T_{\rm int}$ = 134 K. (a) Schematic cross-section of the planet; layer thicknesses are to scale. The atmosphere corresponds to pressures below 1000 bar. (b) Pressure-temperature (P-T) profile of the atmosphere (light blue) and interior (blue and teal). (c) Transmission spectrum for this model, with the three observed transmission datasets shown for comparison. d) Emission spectrum for this model, compared against the four observed emission datasets. (e) Interior metal mass fraction profile ($Z$; black) and H/He mass fraction as functions of normalized radius. The atmospheric region ($P<$ 1000 bar) is indicated in light blue. (f) Atmospheric mass mixing ratios of the molecular species comprising the metals ($P<$ 1000 bar). The region highlighted in orange indicates the pressure range probed by emission spectra, as calculated from the contribution function.
• Figure 4: Top panel: Core and bulk metal mass fraction 1-$\sigma$ estimates obtained by our joint retrievals (JR1-JR6). Joint retrievals that take into account the age have triangle markers. Interior-only retrievals (R1-R3 and R5) are shown as colored boxes for comparison. Bottom panel: Atmospheric metallicity and C/O ratio estimates from our suite of joint retrievals. Spectra-only estimates are displayed as colored boxes for comparison.
• Figure 5: Pressure-Temperature (PT) profiles of WASP-80 b as constrained by four of our joint retrievals. The shaded regions indicate the 1, 2, and 3$\sigma$ regions of the retrieval, while solid lines correspond to the mean. Black lines highlight self-consistent models from Molliere17Acuna24_gastli_science_paper of a clear atmosphere with solar composition at $T_{\rm int}$ = 100 (dashed) and 500 K (dotted) for comparison. Left: The difference between JR3 and JR5 shows the effect of adding the transmission spectrum to JR3, which incorporates emission spectra plus the mass and radius. The transmission spectrum reduces significantly the uncertainties in thermal structure. Center: Comparing JR6 and JR5 allows us to study the effect of including the age to the retrieval. JR6 is the most-data complete of the joint retrievals, since it incorporates both transmission and emission, plus the mass and age. If the age is removed (JR5), the temperature in the deep atmosphere increases. Right: The comparison between JR5 and JR5$^{\ast}$ illustrates the effect of assuming equilibrium chemistry in the joint retrieval. If we relax the assumption of equilibrium chemistry in JR5 by changing to a free chemistry model (JR5$^{\ast}$), the temperature decreases in the deep interior.