Uniform Reinterpretation of Rocky Exoplanet Secondary Eclipse Observations and the Impact of Stellar and Orbital Uncertainties

1. Problem Statement

The detection of atmospheres on rocky exoplanets via secondary eclipse observations relies heavily on comparing observed eclipse depths to theoretical models of "dark bare rocks" (planets with no atmosphere and zero albedo). However, current interpretations often suffer from two critical limitations:

• Model Simplifications: Many studies assume a uniform dayside temperature for bare rocks, ignoring the realistic temperature gradient that peaks at the substellar point and drops toward the terminator.
• Propagation of Uncertainties: Previous analyses often treat host star properties (effective temperature $T_*$, metallicity, surface gravity) and orbital parameters (semi-major axis $a_p$, planet radius $R_p$) as fixed values. In reality, these parameters have significant uncertainties that propagate non-linearly into the predicted eclipse depth and brightness temperature.
• Consequence: Ignoring these uncertainties leads to an underestimation of model error bars, potentially resulting in false positives for atmospheric detection or erroneous conclusions about surface composition. The authors argue that model uncertainty is becoming a fundamental precision limit comparable to, or exceeding, observational uncertainty.

2. Methodology

The authors introduce a new framework using a tool called JESTER (JESTER: an Exoplanet Surface Tool for calculating the Emissions from Rocky worlds) to uniformly reanalyze 18 rocky exoplanets ($R_p \le 2 R_\oplus$) with published secondary eclipse data.

Key Methodological Steps:

• Non-Uniform Temperature Modeling: Instead of a uniform blackbody, JESTER solves a 2D energy balance equation numerically. It divides the dayside hemisphere into annular regions centered on the substellar point, calculating local temperatures where absorbed stellar flux equals emitted thermal flux.
• Monte Carlo Parameter Sampling: To quantify model uncertainty, the authors generate 2,000 forward models for each planet. They sample astrophysical parameters ($R_p/R_*$, $a_p/R_*$, $T_*$, $\log g_*$, $[M/H]$, C/O ratio) from normal distributions based on reported observational errors.
• Stellar Flux Treatment: The framework interpolates across the SPHINX (and PHOENIX for hotter stars) stellar grids to generate the host star's Spectral Energy Distribution (SED) for each iteration. This accounts for discrepancies between different stellar models and the impact of stellar flux errors on the planet's incident energy.
• Derivation of Metrics:
  • $\mathcal{F}$: The ratio of observed eclipse depth to the median modeled dark bare rock depth.
  • $\mathcal{R}$: The ratio of the derived dayside brightness temperature ($T_{d,der}$) to the maximum theoretical brightness temperature of a dark bare rock ($T_{d,dbr}$).
  • Uncertainty Quantification: The standard deviation of the 2,000 model iterations is reported as the model uncertainty ($\Delta \mathcal{F}$ and $\Delta \mathcal{R}$), which is then convolved with observational uncertainty.

3. Key Contributions

• Unified Framework: A reproducible pipeline that accounts for non-uniform dayside temperatures and propagates stellar/orbital parameter uncertainties into emission models.
• Quantification of Model Uncertainty: The paper demonstrates that for many systems, the uncertainty introduced by imperfect knowledge of system parameters is comparable to the observational error bars.
• Linear Correlation Discovery: The authors derive a linear relationship between the relative model uncertainty in eclipse depth and the sum of relative errors in key parameters ($R_p/R_*$, $a_p/R_*$, $T_*$). This allows for rapid estimation of model uncertainty in future studies without running full Monte Carlo simulations.
• Reanalysis of 18 Targets: A comprehensive re-evaluation of all rocky exoplanets with published emission data, providing updated $\mathcal{F}$ and $\mathcal{R}$ values with robust error bars.

4. Key Results

• Consistency with Bare Rocks: When model uncertainty is included, the majority of the 18 analyzed planets are consistent with a dark bare rock composition within $1\sigma$.
• Degeneracy of Interpretations:
  • Planets with intermediate $\mathcal{R}$ values (between 0.78 and 1.0) are now shown to be consistent with both a bare rock and a fully redistributive atmosphere due to the widened error bars.
  • Specific cases like K2-141 b deviate from previous literature conclusions; with model uncertainty included, it is no longer a definitive outlier.
• Outliers and Anomalies:
  • TRAPPIST-1 b: F1280W observations show a brightness temperature $\sim 2\sigma$ below the dark bare rock model, suggesting potential atmospheric heat redistribution or high albedo, though the result is sensitive to the specific data reduction.
  • LHS 1478 b, 55 Cnc e, TOI-431 b, TOI-561 b: These planets exhibit low $\mathcal{R}$ values ($\ge 2\sigma$ below unity), strongly suggesting the presence of atmospheres or highly reflective surfaces, even after accounting for model uncertainty.
  • TOI-1468 b: Shows an unexpectedly high eclipse depth ($\mathcal{F} > 1$), potentially indicating thermal inversions, magnetic induction heating, or tidal heating.
• Population Trends: A tentative log-linear trend is observed where $\mathcal{R}$ decreases as irradiation temperature ($T_{irr}$) decreases for M-dwarf planets. This suggests colder planets may be more likely to retain atmospheres or have different surface properties (e.g., space weathering effects).
• Surface Composition Constraints: The study shows that photometric (broadband) observations cannot distinguish between different surface compositions (e.g., basalt vs. granite) due to significant overlap in model predictions. However, spectroscopy remains viable because the shape and location of spectral features are preserved across parameter variations.

5. Significance and Implications

• Redefining Precision Limits: The paper establishes that model uncertainty is a fundamental limit to characterizing exoplanet surfaces and atmospheres. Future claims of atmospheric detection must explicitly account for this uncertainty to avoid false conclusions.
• Guidance for JWST: The findings inform the design and analysis of upcoming JWST programs (e.g., the Rocky Worlds DDT). The authors recommend:
  • Prioritizing precise characterization of host star SEDs (potentially using MIRI MRS) to reduce stellar model uncertainty.
  • Using the derived linear correlation (Eq. 18) to estimate model uncertainties for new targets quickly.
  • Relying on spectroscopic features rather than broadband eclipse depths alone to break degeneracies between surface composition and atmospheric presence.
• Methodological Shift: The field must move away from "single best-fit" parameter models toward probabilistic frameworks that sample the full parameter space, ensuring that the uncertainty in the host star and orbit is not ignored when interpreting planetary emission.

In conclusion, this work provides a critical "reality check" for the field of rocky exoplanet characterization, arguing that while we are making progress, the precision of our physical models is currently limited by our knowledge of the host stars and orbital dynamics, not just the quality of the telescope data.