Trappist-1b and c: atmospheres or not?

After planet b a few months ago, it is now the turn of planet c in the TRAPPIST-1 system to be the target of the JWST space telescope.  An international research group, including Michaël Gillon, FNRS Research Director and astrophysicist at ULiège, the system's principal discoverer, has just presented its detection of the thermal emission from planet TRAPPIST-1c with the JWST. Their analysis of this measurement shows that the presence of a dense atmosphere rich in carbon dioxide around the planet, like that of Venus, is unlikely, a result similar to that obtained for planet B a little earlier. Can we, therefore, conclude that these two planets are large balls of rock with no atmosphere? Michaël Gillon explains in this interview that "this is largely premature".

M.Gillon rond ULliege : M.Houet

Michaël Gillon is an astrophysicist, Director of Research at the F.R.S.-FNRS and head of the EXOTIC laboratory (Astrobiology Research Unit/Faculty of Science). He heads the SPECULOOS consortium, which brings together several scientists and institutions renowned for their knowledge of exoplanets. Michaël Gillon is, above all, responsible for the discovery 2015 of the TRAPPIST-1 system with the TRAPPIST telescope of the University of Liège and, among others, NASA's Spitzer space telescope. This unique system is now one of the prime targets of the JWST telescope.


What is TRAPPIST-1?

Michaël Gillon: TRAPPIST -1 is a very low-mass star 40 light-years from our solar system that hosts a very unique planetary system made up of seven Earth-sized rocky planets, at least three of which are potentially habitable. Its typically Belgian name is a reference to the Liège telescope TRAPPIST (itself a reference to Belgian beers), which my team and I used to discover the system in 2015. The scientific importance of this system is major, as it contains the only potentially habitable Earth-sized exoplanets for which the detection and in-depth study of an atmosphere is possible, including the possible detection of chemical traces of life (biosignatures).

So astronomers are looking for traces of life on the TRAPPIST-1 planets?

Michaël Gillon: At this stage, the main question we want to answer is whether there is an atmosphere around the planets. This is an essential prerequisite for the possible presence of life on their surfaces. The TRAPPIST-1 planets orbit extremely close to their small star, on the order of 1% to 6% of the Earth-Sun distance. What's more, very low-mass stars like TRAPPIST-1 emit a lot of high-energy radiation (X-rays, UV rays, charged particles).  Under such constant bombardment, it is possible that the planets' atmospheres have been totally eroded, transforming them into large balls of rock like Mercury or the Moon. But it is also possible that they have managed to retain significant atmospheres that could be similar to those of Venus, Earth or Mars... or totally different.

What methods do astronomers hope to use to detect an atmosphere around the TRAPPIST -1 planets?

Michaël Gillon: First of all, we need a very powerful space telescope operating in the infrared, a type of light to which our eyes are not sensitive. We now have such a telescope, the JWST, which has been in operation for almost a year. We now have access to several methods for trying to detect these possible atmospheres. The first is based on what is known as the transit method, the passage of planets in front of their star. It was in fact these transits that enabled us to discover the Trappist-1 planets. When they pass in front of the star, they hide part of it, so that the star appears less luminous and we can deduce that the planet exists.  What's more, during a transit, a small part of the light emitted by the star in our direction is filtered out by the planet's atmosphere (if it has one). By separating the star's light into different colours (known as spectroscopy) and comparing measurements taken during and outside the transit, we can hope to detect traces of this filtration and obtain information about the composition of the planet's atmosphere.  This method has been or will shortly be applied to all the TRAPPIST-1 planets with the JWST, and the first results should be published soon.

However, there is an alternative method, the one used in this new publication, which is based on the planet passing behind the star. This is called an occultation. When this happens, the small luminous contribution of the planet to the brightness of the system disappears, so it can be measured negatively by comparing the brightness of the system measured during and before or after the occultation. This gives us a measure of the planet's emission, its heat flux. The JWST is very powerful, but the TRAPPIST-1 planets are much smaller and cooler than their star, so we can only hope to make this kind of measurement for the three or four innermost planets, which are the hottest because they are the most irradiated by the star. And even then, we can only do this at one wavelength at a time, without any spectroscopic information.

ESA trappist

This image illustrates how a star illuminates and heats the day side of a synchronously rotating planet and the phenomena of transit and occultation (eclipse). The bottom of the figure shows the evolution of the measured flux as a function of the orbital phase of the planet. When the planet passes in front of the star, it hides part of it, and the system appears less bright (transit). When the planet passes behind the star, its small contribution to the brightness of the system disappears (occultation/eclipse).  Finally, the system's brightness is modulated by the phase of the planet. If its day side is hotter than its night side, the system will appear brighter at a time close to occultation than at a time close to transit.  Credit: ESA, CC BY-SA 3.0 IGO

What does this new result, published in Nature, involve?

Michaël Gillon: A few months ago, another team used the JWST and this occultation method to measure the thermal emission from the innermost planet, TRAPPIST-1b, at a wavelength of 15 micrometres corresponding to a strong reaction of light with the carbon dioxide molecule. The idea was as follows: if the planet has a dense atmosphere rich in carbon dioxide, like Venus, the light emitted by the surface should be absorbed by the carbon dioxide before reaching space, and the planet should then appear very faint at this wavelength. However, the opposite has been observed. This high brightness can be explained by a very simple hypothesis: the planet has no atmosphere. Since it always faces the star the same way, if it has no atmosphere to transport heat to the night side, its day side should be very hot and very bright at the wavelength in question. This is the hypothesis that NASA has decided to put forward, with some NASA posts even mentioning "no atmosphere for TRAPPIST-1b".

In this new study (1), which has just been published in the journal Nature, a team including myself and led by Sebastian Zieba from the Max Planck Astronomy Institute in Heidelberg, used the JWST to make an identical measurement for TRAPPIST-1c, the second innermost planet in the system. Our measurement shows that planet c is again quite bright at 15 microns, which, according to our models, makes a dense carbon dioxide-rich atmosphere similar to that of Venus unlikely. This is a very important result, as TRAPPIST-1c is the coldest Earth-sized exoplanet on which such a measurement of thermal emission has been carried out.

So these results show that TRAPPIST-1b and c have no atmosphere?

Michaël Gillon: Not at all! If we assume that our theoretical models are correct, these measurements just rule out atmospheres rich in carbon dioxide. But they could have dense atmospheres that are very low in carbon dioxide. For example, if these planets were originally much richer in water than the Earth, they could very well have atmospheres rich in water and oxygen. In fact, the intense high-energy radiation they experience could break down the water molecules in the upper atmosphere. The hydrogen, being very light, would easily escape, while the oxygen would remain. We can also imagine that under the star's intense ultraviolet radiation, a complex photochemistry could be at work in the atmosphere of these planets, greatly depleting them of carbon dioxide and enriching them with so-called reducing molecules such as methane, carbon monoxide or ammonia. Alternatively, the high brightness measured could be explained by the presence of very hot carbon dioxide in the upper atmosphere of the planets.

Taking a step back, I must point out an essential element: science is a method based on constant comparison between theories and observations. In other words, our theories, even the most elegant ones, are worth nothing at the scientific level without this comparison with observations. However, our models of the atmospheres of the terrestrial planets were built on the basis of observations of just three planets: Venus, Earth and Mars. These are the only rocky planets for which we have access to high-quality atmospheric data. But these three planets are just three out of hundreds of billions of rocky planets in our galaxy. In a different environment, or with a different initial composition, or with a different evolution, a rocky planet could very well be or have been the site of processes completely neglected by our current theories. When we write in our publications that "our models show", we are overlooking the very likely possibility that our models may be incomplete, or even totally wrong, when applied to objects that are very different from the planets in the Solar System. One of the major objectives of the study of exoplanets is to build and refine these models, to obtain a theory applicable to all the planets in the Universe. We are a long way from achieving this, especially for rocky planets, for the simple reason that we still have virtually no data, no measurements.

To illustrate this point, I'll take the example of 51 Pegasi b, the first exoplanet detected in 1995 by Michel Mayor and Didier Queloz. At the time, some people did not want to believe in its existence because it was a giant planet with an extremely short orbit, something that was not predicted by the theories of planetary formation at the time. The detection of many more of these hot Jupiters and the confirmation of their planetary nature showed that these theories were incomplete. They have since been greatly refined, and our understanding of the mechanisms of planetary formation has come a long way.

So, if I had to sum up ... we need more observations before we can draw any conclusions!

What kind of observation do you need?

Michaël Gillon: We could imagine reproducing the observations published for TRAPPIST-1b and c at many other wavelengths. Such spectroscopic data could lead to the detection of unmistakable signs of an atmosphere. For example, next month the JWST will again observe TRAPPIST-1b, but this time at 12.8 microns instead of 15 microns. If the brightness of the planet at 12.8 microns turns out to be much lower than at 15 microns, it will be very difficult to explain the observations without the presence of an atmosphere. On the other hand, if we accumulate measurements at a large number of wavelengths and they all correspond to what we would expect for a planet with no atmosphere, then we will have to accept what the observations tell us and concentrate on the five outermost planets in the system.

However, there is a simpler and more effective method for confirming or rejecting the presence of an atmosphere around planets b and c. As I said, these planets rotate synchronously in relation to their star, like the Moon in relation to the Earth, meaning that they always show the same face to the star. When we use the occultation method, we measure the brightness of the day side of the planet, the side facing the star. But if we observe the brightness of the system during a complete orbit of the planet around the star with the JWST, what we call a phase curve, we can hope to detect a modulation that will tell us about the variation in longitude of the thermal emission of the planet, and in particular tell us to what extent the night side is cooler than the day side. This is crucial information, because if a planet in synchronous rotation has a dense enough atmosphere, it will efficiently transport heat from the day side to the night side, whatever its composition. Without an atmosphere, this is not the case. Let's take the example of Mercury, which has no atmosphere. On its day side, the temperature can rise to over 400° Celsius, while on its night side it can drop to minus 180°! Based on this principle, colleagues and I proposed to NASA that we observe the phase curve of planets b and c using the JWST. This has been accepted, and the observations should be made in November. Once analysed, they should enable us to determine once and for all whether or not these two planets have an atmosphere. 

So an answer is just around the corner?

Michaël Gillon: For planets b and c, yes, with observations scheduled for next July and November. For the other five planets, it's going to take more time and a lot more observations. The best thing at this stage, I think, is to wait for the results for b and c before stepping up observations of the other five planets. Indeed, if we can demonstrate that b and/or c have an atmosphere, then it will be much more likely that the other planets, which are more external and less bombarded by the star's high-energy radiation, have also managed to retain an atmosphere. This will put us in an excellent position to convince NASA to set up a very ambitious JWST programme to detect these atmospheres and search for possible traces of life. Such a programme could also be launched if b and c turn out to be planets without atmospheres, but this would be much more difficult. The pressure on the JWST is enormous, and TRAPPIST-1 is far from being its only target...

Scientific references

  1. No thick carbon dioxide atmosphere on the rocky exoplanet TRAPPIST-1c, Sebastian Zieba et al. 2023, Nature.
  2. Thermal emission from the Earth-sized exoplanet TRAPPIST-1b using JWST, Thomas P. Greene et al. 2023, Nature, 618, 39-42.

Share this news