A planet’s atmosphere contains different chemical elements. The Earth’s atmosphere, for example, consists of about 78% nitrogen, 21% oxygen, 0.9% argon, and just 0.01% other gases like carbon dioxide, methane, and ozone. Earth’s atmosphere also contains varying amounts of water vapour in different regions and at different heights. That catch-all “other” category may be small, but trace gases can have an outsized influence on the climate–as we know from the example of carbon dioxide.
For comparison, the atmosphere of Venus consists of 96.5% carbon dioxide, 3.5% nitrogen, and only trace amounts of other gases. The huge amount of carbon dioxide on Venus causes a greenhouse effect that makes it the hottest planet in the Solar System, even hotter than Mercury, the closest planet to the Sun. The Venusian atmosphere contains a small amount of sulphur dioxide–around 0.015%–but this is enough to create sulphuric acid clouds that blanket the planet and rain down acid through the cloud layers. The intense heat evaporates the acid rain before it reaches the surface.
Clearly, the detailed recipe for a planet’s atmosphere has a huge effect on the environment… On top of that, to know whether life exists on distant planets we’ll have to rely on detections of “biosignatures” in the atmosphere. These are gases believed to be produced only (or only in the detected quantities) by living beings, like the phosphine reported last year on Venus. Even though Venus is right next door, the detection has been fiercely disputed and remains a topic of debate and a target of further study. We can expect the same difficulties when we start exploring the atmospheres of far more distant worlds.
We know what the Earth’s atmosphere is made of from direct measurements. We have studied the atmosphere of Venus through telescope observations and spacecraft we sent there, like the Pioneer Venus mission. But how can we know what elements are found in the atmospheres of exoplanets billions of miles away? These objects are so dim that interfering light from their host stars outshines them by a factor of a billion. They’re so small that most of them can’t be photographed separately from their host stars, even by the most sensitive telescopes we have. And they’re so far away that sending a probe to take atmospheric measurements is firmly in the realm of science fiction.
The most promising way of studying and describing exoplanet atmospheres is called transmission spectroscopy. To understand it, we need to know a little about how light interacts with matter.
You may know that white light consists of electromagnetic waves with different wavelengths. For visible light, each of these wavelengths corresponds to a different colour. Infrared light, microwaves, X-rays, and radio waves also contain a range of wavelengths, even though we can’t see them or perceive them as colours. If a photon travelling through space collides with an atom or molecule, the molecule can absorb it if the photon has the right wavelength. For example, water vapour absorbs infrared light with wavelengths of 1.4, 1.8, 2.7, and 6 micrometres.
Transmission spectroscopy takes advantage of these tiny quantum interactions to draw the big picture conclusions about atmosphere and climate that we want.
This method only works on a fraction of the planets that exist in the galaxy. One of the main ways of finding exoplanets is called the transit method. If a planet is orbiting around a star in our line of sight from Earth, it will pass in front of the star for part of its orbit. The planet blocks part of the star’s light. This is called the primary eclipse. In a different part of its orbit, the planet disappears behind the star: the secondary eclipse. If we observe the star for whole orbit, we will see the star grow dimmer twice, during both eclipses. This pattern repeats regularly over multiple orbits and allows us to infer the presence of an exoplanet even though we can’t see it directly.
Most of the exoplanets we know of have been detected using the transit method. Transmission spectroscopy takes advantage of the planet-star eclipses to gather data about the planet’s atmosphere. During the primary eclipse, when the planet passes over the star’s face, it is backlit by starlight. Remember that although planets are very faint, their host stars are a billion times brighter–the eclipsing planet blocks only a small portion of the starlight, so we still get a strong signal. The starlight shines through the thin layer of the exoplanet’s atmosphere and some of it is absorbed by atmospheric gases. If a planet has water in its atmosphere, it will absorb the light with wavelengths of 1.4, 1.8, 2.7, and 6 micrometres.
Say we’re observing TRAPPIST 1-e, an Earth-like exoplanet in the habitable zone of a star 39.5 lightyears away. First we observe the star, TRAPPIST 1, during the secondary eclipse. The starlight isn’t just featureless white light because elements in the upper layers of the stellar atmosphere also absorb photons travelling from deeper inside the star. We use this observation as a base of comparison. Then, we keep observing the star-planet duo during the primary eclipse. We see that, at the wavelengths 1.4, 1.8, 2.7, and 6 micrometres, our telescope collects less light than at neighbouring wavelengths. The eclipse looks “bigger” at these wavelengths because not only the planet, but also the absorbing atmosphere is blocking light. We conclude that there is water vapour in the atmosphere of TRAPPIST 1-e–and therefore water on the planet! (In reality, we have to remove many kinds of noise from our data, compare it to computer simulations of how we expect a planetary spectrum with water vapour features to look, and quantify how good the match is, using all sorts of technical wizardry and dealing with numerous sources of error and bias.)
Transmission spectroscopy requires extremely precise measurements of light, but has been successfully applied to several exoplanets already (check out WASP-17b or HD 189733b, both of which have water vapour in their atmospheres). This method works best if the planet is large compared to the star and has a close-in orbit, which means eclipses happen frequently and astronomers can make multiple observations in a short time. So far the planets observed with transmission spectroscopy are what are called “hot Jupiters,” gas giants orbiting very close to their host stars. But Earth-like planets are a target of several upcoming space missions with ultra-sensitive instruments. We may start getting data from the TRAPPIST 1 system using the James Webb Space Telescope as early as 2023. Until then, computer models and artist impressions will have to tide us over.