The most intuitive method used by astronomers to determine whether an exoplanet has an atmosphere is to observe its transit - passage in front of its host star (by partially blocking the star’s light for around 30 minutes) - in different wavelengths to detect which parts of the light are transmitted and which are not, indicating the presence of molecules.
However, stellar contamination problems arise for planets orbiting very cold stars. This is because these stars are not homogeneous, and have hot and cold spots on their surface. These spots have their own spectra, which can pollute the transmission spectrum of transiting planets and mimic atmospheric features. Such a phenomenon has been observed on several occasions with the JWST when observing transits of planets around cool stars.
Emission to the rescue
One way to overcome this stellar contamination and obtain information on the presence (or absence) of an atmosphere is to directly measure the planet’s heat by observing a drop in flux as the planet passes behind the star (an event called occultation). By observing the star just before and during occultation, we can deduce the amount of infrared light coming from the planet.
The JWST is particularly well suited to this type of detailed spectroscopic study of small, rocky planets orbiting red dwarf stars. In this context, the red dwarf star TRAPPIST-1, home to seven Earth-sized rocky planets, including three located in the star’s habitable zone (Gillon et al., 2017), proves an ideal target. In particular, its closest planet, TRAPPIST-1 b, has been widely observed in the mid-infrared, at 15 microns wavelength, by JWST (October 2022, November 2022, July 2023, November 2023).
The unresolved case of TRAPPIST-1 b
From this first cohort of data, a 2023 study by Greene et al. had suggested that a thick, CO2-rich atmosphere was unlikely on TRAPPIST-1 b.
But these conclusions were qualified by the same team, in the light of new data now available at 12.8 microns on the planet’s flux. On December 16, 2024, she reported in the journal Nature Astronomy on a comprehensive analysis of all the infrared data collected on TRAPPIST-1 b. In this new study, led by Elsa Ducrot, then a postdoctoral fellow at Observatoire de Paris-PSL (and currently an astronomer at CEA), the authors carried out a global analysis of all available JWST data and compared these observations with surface and atmospheric models to identify the scenario that best matched the data.
In the “dark bare rock” scenario proposed by Greene et al. (2023), the expected temperature at 12.8 microns was around 227°C. However, the actual measurement showed a lower temperature of 150°C.
To explain such a discrepancy, the authors explored various surface and atmospheric models. They found that a bare surface composed of ultramafic rocks (mineral-rich volcanic rocks) could explain the observations. Indeed, ultramafic rocks emit less thermal radiation at 12.8 microns than a conventional dark surface.
The authors also found that an atmosphere rich in CO2 and mists could explain the observations. Mists are tiny particles or droplets suspended in a planet’s atmosphere, often created by chemical reactions, volcanic activity or solar radiation. These particles can scatter and absorb light, affecting the atmosphere’s appearance and temperature. For example, mists are present in the atmosphere of Titan, Saturn’s famous moon.
It’s surprising that a hazy, CO2-rich atmosphere matches the data, as CO2 was thought to be incompatible with the high emission observed at 15 microns. However, haze can be a game-changer. They reflect a lot of light and can make the upper atmosphere warmer than the lower layers, creating a thermal inversion similar to that of the Earth’s stratosphere. This causes CO2 to emit radiation instead of absorbing it, resulting in a higher flux at 15 microns compared to 12.8 microns-an unexpected result compared to CO2 behavior on Earth or Venus.
The authors note, however, that this atmospheric model, while consistent with the data, remains less likely than the bare rock scenario. Its complexity and questions about mist formation and long-term climate stability on TRAPPIST-1 b make it a difficult fit.
Future research, including advanced 3D modeling, will be needed to explore these issues. More generally, the team stresses the difficulty of definitively determining the composition of a planet’s surface or atmosphere using measurements with broadband filters alone, while highlighting two compelling scenarios that will be explored in more detail with future observations of TRAPPIST-1 b’s phase curve, which represents the variation in an exoplanet’s brightness during its orbit, caused by changes in the illuminated portion visible from Earth. This provides information on the planet’s atmosphere, surface properties and temperature distribution.
Now what ?
While both scenarios remain viable, the authors explain that recent observations of TRAPPIST-1 b’s phase curve-which tracks the planet’s flux throughout its orbit-could help solve the mystery. By analyzing the efficiency with which heat is redistributed on the planet, astronomers can deduce whether an atmosphere is present. If an atmosphere exists, heat should be distributed from the planet’s day side to its night side ; without an atmosphere, heat redistribution would be minimal. Investigations are therefore set to continue.
| A method adopted for STScI’s “Rocky Worlds” observation program The Space Telescope Science Institute (STScI) recently approved a 500-hour Director’s Discretionary Time (DDT) program called “Rocky Worlds” to investigate the atmospheres of a large number of terrestrial exoplanets orbiting red dwarf stars. Note that this program uses exactly the same approach as the authors, via occultation observations, but at 15 microns only. |