Helium escape in context: Comparative signatures of four close-in exoplanets

Technical Summary: Helium Escape in Context: Comparative Signatures of Four Close-in Exoplanets

Problem Statement
Observations of escaping atmospheres on close-in exoplanets reveal a wide range in the strength and morphology of He I 10830 Å and H I Balmer absorption. While empirical scaling relations attempt to link He I signals to XUV irradiation, mass loss rates, and bulk planetary parameters, substantial scatter and null trends with global parameters (such as age or baseline transit depth) suggest that simplified parameterizations are insufficient. These relations often omit critical physics, including composition, multi-species transport, detailed energy balance, and 3D circulation. Furthermore, degeneracies between temperature, velocity structure, and mass loss rates in frequently used Parker wind models can lead to incomplete interpretations. The authors aim to test these scaling relations and identify the general drivers of transit depth variability by analyzing four representative systems: the hot Jupiters HD 209458b, HD 189733b, and HD 149026b, and the warm sub-Neptune GJ 1214b.

Methodology
The authors employ a unified, self-consistent 1D hydrodynamic, multi-species, full-atmosphere escape model (based on Taylor et al. 2025) to simulate the atmospheres of the four target planets. The methodology involves:

• Lower/Middle Atmosphere: Using a photochemical model (Lavvas & Arfaux 2021) to calculate thermal and compositional structures from 1000 bar to 10⁻⁶ bar, providing consistent lower boundary conditions (temperature and altitude at 1 µbar) for the upper atmosphere simulations.
• Upper Atmosphere Escape: Solving coupled, time-dependent equations for mass, momentum, and energy conservation. The model includes photoionization, thermal ionization, recombination, charge exchange, advection, molecular/atomic diffusion, and radiative cooling (via recombination, H I line emission, and H₃⁺ cooling).
• Chemical Networks: Two classes of models are utilized. For HD 209458b and HD 189733b, the primary focus is on H/He species (including H, H⁺, He, He⁺, He²⁺, He(2³S), and electrons), with H₂ included for HD 189733b to match lower-atmosphere conditions. For GJ 1214b, the authors compare an H/He-only model against a model including molecular hydrogen (H₂) and associated ions (H₂⁺, H₃⁺, HeH⁺) to assess the impact of molecular chemistry on ionization balance and cooling. HD 149026b is modeled with an atomic-only upper atmosphere due to predicted high H₂ dissociation at the boundary.
• Stellar Fluxes: The models utilize specific spectral energy distributions (SEDs) for each host star (G0V, K2V, G0IV, and M4.5), incorporating XUV fluxes derived from MUSCLES Treasury Survey data or scaled solar proxies.
• Non-LTE Populations: H I level populations (specifically H(n=2)) are calculated using a non-LTE framework coupled with a Lyα Monte Carlo resonant scattering radiative transfer model to determine Hα absorption.
• Transit Depth Calculation: Synthetic transit depths are computed by ray-tracing through the modeled pressure-temperature-density structures, incorporating Voigt profiles with natural, thermal, rotational, and line-of-sight outflow broadening.

Key Results

1. HD 209458b (Benchmark): The model reproduces observed He I and Hα transit depths without imposing ad-hoc composition constraints (e.g., sub-solar He/H). The solution naturally exhibits diffusive separation, with the He/H ratio declining from 8% at the base to ~2.5% at high altitudes.
2. HD 189733b: The model reproduces the He I line core depth comparable to HD 209458b but finds that the broadest reported profiles require an additional ~12 km s⁻¹ of non-thermal broadening beyond thermal, bulk-flow, and rotational effects. However, more recent, narrower observations (Orell-Miquel et al. 2025) are consistent with the model predictions without extra broadening. The system also exhibits diffusive separation (He/H dropping to ~2%).
3. HD 149026b: Despite similar system properties to HD 209458b, the model predicts extremely weak He I absorption (~0.01%) and no Hα excess. This is attributed to the planet's higher surface gravity, which compresses the thermosphere, suppresses hydrodynamic expansion, and enhances diffusive separation, depleting helium at observable altitudes. The model prediction falls significantly below the trend of proposed scaling relations.
4. GJ 1214b: H/He-only models overestimate He I absorption. Including H₂ and its ions (H₂⁺, H₃⁺, HeH⁺) lowers the escape rate and modifies the ion/electron balance, reducing metastable helium densities by shielding helium from ionizing photons and altering cooling mechanisms. The model predicts a significantly shallower He I signal (~7% vs. ~14% in the atom-only case) and no detectable Hα transit.
5. Scaling Relations: When compared against the scaling relation $EW R^2_\star \propto F_{XUV,He} R^2_p / \Phi_p$, HD 189733b and the model prediction for HD 149026b fall below the trend, while some observations of HD 209458b and GJ 1214b are consistent. The authors note that observed transit depths are variable across epochs and instruments.

Significance and Claims
The paper claims that interpreting He I and Hα absorption requires first-principles models that include self-consistent temperature and velocity profiles, multi-species transport, and molecular chemistry, rather than relying solely on empirical scaling relations.

• Diffusive Separation: Across all targets, the models find significant diffusive separation of helium and hydrogen. The authors suggest this mechanism explains why simplified models often require sub-solar He/H ratios to match observations.
• Gravity and Composition: The study demonstrates that high surface gravity (HD 149026b) and molecular composition (GJ 1214b) are critical drivers that can suppress escape signatures and deviate from simple XUV-driven scaling laws.
• Broadening Mechanisms: The discrepancy in HD 189733b's line width suggests that 1D radial escape models may be insufficient to explain all observed features, potentially pointing to magnetically mediated dynamics or turbulence, though the authors note recent narrower observations align with 1D predictions.
• Model Robustness: The work highlights that while scaling relations offer population-level intuition, they fail to capture the scatter in individual systems. Physics-rich models are essential for understanding the underlying atmospheric physics (thermal structure, ionization state, and transport vs. chemistry) that set the metastable helium population.

The authors conclude that their comparative analysis places recent helium-scaling prescriptions in a broader physical context, showing that individual systems depart from trends for well-defined physical reasons related to gravity, diffusion, and composition.