[Paper Review] Potential Atmospheric Compositions of TRAPPIST-1 c constrained by JWST/MIRI Observations at 15 $μ$m

Andrew Lincowski, Victoria Meadows+14 more|arXiv (Cornell University)|Aug 11, 2023

Astro and Planetary SciencePhysics and Astronomy3 citations

TL;DR

This study uses a coupled climate-photochemistry model to evaluate potential atmospheric compositions for TRAPPIST-1 c using JWST/MIRI secondary eclipse observations at 15 μm. It finds that thin O2-dominated atmospheres with low CO2 (≤100 ppm) and up to 10% water vapor are consistent with the observed 421 ± 94 ppm eclipse depth within 1σ, while thicker O2 or steam atmospheres (≤3 bar) are also plausible within 1.8σ, though Venus-like atmospheres are excluded at 3σ confidence.

ABSTRACT

The first JWST observations of TRAPPIST-1 c showed a secondary eclipse depth of 421+/-94 ppm at 15 um, which is consistent with a bare rock surface or a thin, O2-dominated, low CO2 atmosphere (Zieba et al. 2023). Here, we further explore potential atmospheres for TRAPPIST-1 c by comparing the observed secondary eclipse depth to synthetic spectra of a broader range of plausible environments. To self-consistently incorporate the impact of photochemistry and atmospheric composition on atmospheric thermal structure and predicted eclipse depth, we use a two-column climate model coupled to a photochemical model, and simulate O2-dominated, Venus-like, and steam atmospheres. We find that a broader suite of plausible atmospheric compositions are also consistent with the data. For lower pressure atmospheres (0.1 bar), our O2-CO2 atmospheres produce eclipse depths within 1$σ$ of the data, consistent with the modeling results of Zieba et al. (2023). However, for higher-pressure atmospheres, our models produce different temperature-pressure profiles and are less pessimistic, with 1-10 bar O2, 100 ppm CO2 models within 2.0-2.2$σ$ of the measured secondary eclipse depth, and up to 0.5% CO2 within 2.9$σ$. Venus-like atmospheres are still unlikely. For thin O2 atmospheres of 0.1 bar with a low abundance of CO2 ($\sim$100 ppm), up to 10% water vapor can be present and still provide an eclipse depth within 1$σ$ of the data. We compared the TRAPPIST-1 c data to modeled steam atmospheres of $\leq$ 3 bar, which are 1.7-1.8$σ$ from the data and not conclusively ruled out. More data will be required to discriminate between possible atmospheres, or to more definitively support the bare rock hypothesis.

Motivation & Objective

• To assess the compatibility of various atmospheric compositions with the observed 15 μm secondary eclipse depth of TRAPPIST-1 c from JWST/MIRI.
• To investigate how photochemistry and thermal structure affect atmospheric opacity and eclipse depth predictions.
• To determine whether the observed eclipse depth supports a bare rocky surface or a range of possible atmospheres.
• To evaluate the plausibility of O2-dominated, Venus-like, and steam-rich atmospheres under self-consistent radiative-convective and photochemical modeling.
• To identify observational pathways for distinguishing between a rocky surface and different atmospheric states in future observations.

Proposed method

• Employed a two-column energy balance model coupled with a photochemical model to simulate atmospheric thermal structure and composition.
• Generated synthetic secondary eclipse spectra for O2-dominated, Venus-like, and steam atmospheres across a range of surface pressures (0.1–10 bar).
• Incorporated radiative transfer using LBLABC and DISORT to compute atmospheric opacity and thermal emission at 15 μm.
• Compared model-predicted eclipse depths to the observed 421 ± 94 ppm from JWST/MIRI F1500W observations.
• Accounted for molecular radiative transfer and heat redistribution effects in the climate model to improve accuracy of eclipse depth predictions.
• Used statistical confidence levels (σ) to assess the compatibility of each atmospheric model with the observed data.

Figure 1: Day-side hemisphere temperature structures for all modeled atmospheres: Venus-like (left panel), steam atmospheres (middle panel) and \ce O2- \ce CO2 atmospheres (right panel). For our modeled atmospheres, we have used a thicker line to show the layers over which the 15 $\upmu$ m band reac

Experimental results

Research questions

• RQ1 Which atmospheric compositions for TRAPPIST-1 c are consistent with the observed 15 μm secondary eclipse depth of 421 ± 94 ppm?
• RQ2 How do photochemical processes and thermal structure influence the predicted eclipse depth for O2-rich or water-vapor-rich atmospheres?
• RQ3 Can thin, O2-dominated atmospheres with low CO2 and water vapor be consistent with the data, and what is the maximum allowable water vapor mixing ratio?
• RQ4 Are thick, Venus-like atmospheres still plausible given the observed eclipse depth, or are they excluded at high confidence?
• RQ5 What future observations or spectral bands could best distinguish between a bare rocky surface and different types of atmospheres on TRAPPIST-1 c?

Key findings

• O2-dominated atmospheres at 0.1 bar with CO2 abundances ≤100 ppm are consistent with the observed eclipse depth within 1σ, confirming earlier findings.
• Up to 10% water vapor mixing ratio is consistent with the data within 1σ, provided the atmosphere remains optically thin at 15 μm.
• Atmospheres with 1–10 bar O2 and 100 ppm CO2 produce eclipse depths within 2.0–2.2σ of the observed value, indicating they are plausible but less likely than thinner atmospheres.
• Atmospheres with 0.5% CO2 are consistent within 2.9σ, suggesting higher CO2 levels are still possible but less probable.
• Steam atmospheres with surface pressures ≤3 bar are within 1.7–1.8σ of the observed depth and are not conclusively ruled out.
• Venus-like atmospheres (≥0.1 bar) are excluded at 2.6–3.1σ confidence, indicating they are highly unlikely to explain the observed eclipse depth.

Figure 2: Brightness temperature spectra for the dayside hemisphere of all modeled environments, with points corresponding to the model spectra convolved to the F1500W filter band over the band’s wavelength extent (horizontal error bars show the FWHM of the filter band). We also plot lines for 340 K