Constraining exoplanet interiors using observations of their atmospheres

Abstract

Astronomical surveys have identified numerous exoplanets with bulk compositions that are unlike the planets of the Solar System, including rocky super-Earths and gas-enveloped sub-Neptunes. Observing the atmospheres of these objects provides information on the geological processes that influence their climates and surfaces. In this Review, we summarize the current understanding of these planets, including insights into the interaction between the atmosphere and interior based on observations made with the James Webb Space Telescope (JWST). We describe the expected climatic and interior planetary regimes for planets with different density and stellar flux and how those regimes might be observationally distinguished. We also identify the observational, experimental, and theoretical innovations that will be required to characterize Earth-like exoplanets.

Figures (6)

• Figure 0: Predicted interior structures of sub-Neptune exoplanets. Observed density and atmospheric constraints are interpreted in three scenarios, each with a gaseous envelope. (Left) Water worlds have a small metal core, rocky mantle, high-pressure ice layer, and possible liquid water ocean. (Middle) Gas dwarfs have a larger metal core and a magma ocean. (Right) Global supercritical regimes have no clear boundaries between layers. Image credit: Mark A. Garlick.
• Figure 1: Timeline of exoplanet detection and atmospheric characterization. (A) Symbol sizes schematically illustrate the sizes of exoplanets detected by each facility and the increasing numbers of super-Earths and sub-Neptune exoplanets found by using Kepler and TESS. (B) An equivalent illustration of the atmospheric characterization of exoplanets by use of space telescopes. (C) The conceptual interior structures of sub-Neptune exoplanets, with increasing complexity owing to improved characterization.
• Figure 2: Observed exoplanet masses and radii compared with compositional models. Radii [in Earth radii ($R_{\rm{Earth}}$)] are plotted as a function of mass [in Earth masses ($M_{\rm{Earth}}$)]. Colored symbols are observed exoplanets; we selected those with mass uncertainties $<$50% and radius uncertainties $<$30%, from Parc2024. Symbol color indicates the equilibrium temperature, assuming full heat redistribution and zero albedo. Temperature thresholds in the color bar were chosen to reflect the regimes identified in Fig. \ref{['fig:regimes']}. Curved lines are theoretical mass-radius relations for different compositions, as labeled Zeng2019Dorn2021Rogers2023. Diamonds are planets with high transmission spectroscopy metric (TSM) or emission spectroscopy metric (ESM), making them a high priority for observational characterization; we selected those with TSM $>$ 90 if the planet radius $\geq 1.5$$R_{\rm{Earth}}$, or TSM $>$ 10 for smaller radii, or ESM $>$ 7.5 2018PASP..130k4401K. Some specific exoplanets are labeled; those in boldface are shown in Figs. \ref{['fig:transmission_spectra']} and \ref{['fig:brightness_temperature']}. The shaded region is a gas dwarf model of sub-Neptunes Rogers2023, which assumes an Earth-like silicate mantle and iron core composition, surrounded by a H/He envelope accreted from the protoplanetary disk.
• Figure 3: Transmission spectra and atmospheric models for selected small exoplanets observed with JWST. Data points are the observed spectra of five example exoplanets discussed in the text, with vertical bars indicating the 1-$\sigma$ uncertainties. Thick solid lines indicate the best-fitting atmospheric model, as reported by each study. Colors indicate different exoplanets: green is TOI-421 b Davenport2025, cyan is GJ 9827 d Piaulet-Ghorayeb2024, blue is TOI-270 d Benneke2024, purple is K2-18 b Madhusudhan2023, and gray is LHS 1140 b (JWST/NIRISS in dark gray and JWST/NIRSpec in light gray) Cadieux2024Damiano2024. The planets are ordered from hottest ($\sim$920 K) to coldest ($\sim$ 225 K) equilibrium temperature, assuming zero albedo; they also have different masses and radii (Fig. \ref{['fig:M-R']}). For comparison, the spectra are offset by a constant value and have been converted to atmospheric scale height, assuming a mean molecular weight ($\mu$) of $2.3$ g/mol. This weight is unlikely to be valid for all these exoplanets, given their diversity of compositions. Labels and thin horizontal lines above each spectrum identify the molecules responsible for absorption features (which extend upward in this representation). The wavelength scale is logarithmic.
• Figure 4: Brightness temperature ratio from eclipse observations compared with theoretical models. Colored bars are the 1-$\sigma$ ranges of brightness temperatures ($T_\mathrm{day}$) derived from eclipse observations, normalized to the theoretical maximum disk-integrated dayside temperature for a zero-albedo, zero-heat redistribution planet (solid black line). They are plotted as a function of the substellar temperature of a tidally locked planet with zero albedo at all wavelengths ($T_\mathrm{irr}$). Bar colors indicate the equilibrium temperature categories defined in Fig. \ref{['fig:M-R']}, calculated by using published methods ParkCoy2024 applied to eclipse depths and uncertainties from JWST and Spitzer emission observations as reported in the literature. Data sources are TRAPPIST-1 c 2021PSJ.....2....1A2023Natur.620..746Z, TRAPPIST-1 b 2021PSJ.....2....1A2023Natur.618...39G2024NatAs.tmp..292D, LTT 1445 A b Wachiraphan2025, LHS 3844 b Kreidberg20192019ApJ...871L..24V, GJ 1132 b 2024ApJ...973L...8X, GJ 486 b 2024ApJ...975L..22W, GJ 1252 b 2020ApJ...890L...7S2022ApJ...937L..17C, TOI-1685 b Luque20252023ApJ...955L...3G, GJ 367 b 2024ApJ...961L..44Z, 55 Cancri e 2024Natur.630..609HPatel2024, and K2-141 b 2022AA...664A..79Z. Shaded regions indicate the ranges predicted by theoretical models for the specific exoplanets with red and blue bars Hammond2025; gray is a model of bare rock surfaces with no atmosphere. Cyan is a model of atmospheres with N$_2$-CO$_2$-H$_2$O compositions with 1 to 10 bar total surface pressure, with (top) no (light cyan) or (bottom) complete (dark cyan) day-to-night heat redistribution through atmospheric circulation.