American Scientist

One of the most exciting potential uses of the James Webb Space Telescope (JWST), which is scheduled to launch in 2018, is to hunt for habitable exoplanets—something that was beyond imagining at its inception. In the 1970s, no one even knew whether exoplanets existed. In the 1990s, when JWST was conceived as the successor to the Hubble and Spitzer space telescopes, the notion that the atmospheres of alien worlds could be studied seemed faintly ludicrous. Part of the early motivation was to build a telescope that would be powerful enough to detect the earliest stars and galaxies. Because the universe is expanding, which reddens light as it travels across space, this new eye on the cosmos would need to be built for the infrared spectrum. Fast forward to 2017, and the measurement of atmospheric properties of exoplanets is now fairly routine. Humanity’s most expensive telescope, originally intended for scanning the outer reaches of the cosmos, is turning out to be a decisive instrument for exploring alien worlds and—if we are lucky—will find ones that are habitable.

When JWST was conceived, studying the atmospheres of exoplanets was not on the minds of its developers. Then in 2005, photons from the atmosphere of an exoplanet were detected for the first time using the Spitzer Space Telescope. Later, astronomers learned how to record signals from these atmospheres at different colors and interpret them to identify the presence of atoms and molecules, using both space- and ground-based telescopes. To date, water, carbon monoxide, hydrogen, magnesium, methane, sodium, and potassium have been robustly detected. Nowadays, the Hubble Space Telescope is routinely used to check whether an exoatmosphere contains water. Astronomers have also made crude temperature maps of these atmospheres. Exoatmospheric science tells us about the general climate conditions of an exoplanet, including chemistry and temperature. As technology has advanced, enabling us to probe cooler (and fainter) exoatmospheres, these discoveries have opened a potential window into studying an exoplanet’s habitability.

These recent advances are prompting changes to JWST—both in terms of tweaks to the hardware and the telescope’s operation—while it undergoes testing in preparation for its scheduled launch in October of 2018. Given the limited lifetime of JWST, which may be as short as 5.5 years, astronomers and astrophysicists are focusing on the best targets for advancing our understanding of exoplanetary atmospheres: gas and ice giants first, and a selected sample of smaller exoplanets second.

Prioritizing Space Time

The Hubble and Spitzer space telescopes have already demonstrated the exoplanet science that is possible from space. JWST will go further by incorporating the capabilities and heritage of both Hubble and Spitzer. Its larger mirror will collect more light in a shorter amount of time, enabling the study of faint targets, and the sensitivity and range of frequencies covered by its instruments exceed both predecessors. JWST follows in the footsteps of NASA’s Great Observatories—space telescopes built to serve the entire astronomical community.

The easiest targets to study are the gas and ice giants, with sizes ranging from that of Neptune to slightly larger than Jupiter, and which orbit close to their stars. Because of their relatively large sizes, the dip in light that occurs when these exoplanets transit their stars is easy to detect. Measuring the change in the size of the exoplanet across frequency is equivalent to constructing a spectrum—splitting light into colors. For example, if an exoplanet with an atmosphere consisting purely of water is observed at wavelengths that are opaque to water, it will appear slightly bigger; at wavelengths at which the atmosphere is transparent to water, the exoplanet will appear slightly smaller. In this way, we may determine whether water is present in an exoplanetary atmosphere. The same technique may be generalized to infer whether other atoms and molecules are present. Astronomers realized that it will be possible to study the atmospheres of a couple of hundred gas and ice giants using JWST. Such a sample of giant exoplanets would allow statistical trends in the properties of their atmospheres to be quantified and would lead to breakthroughs in understanding how they formed. Detections of carbon- and oxygen-bearing molecules would allow us to infer the ratio of carbon to oxygen contained within the atmosphere and thus to tell whether the exoplanet formed close to or far away from its star.

For smaller exoplanets, closer in size to Earth, we are not so lucky. Because they are smaller, their transit signals are weaker and harder to detect. If the planets of the Solar System are any indication, these smaller exoplanets may have secondary, rather than primary, atmospheres. Primary atmospheres are composed of the remnant gas of the protostellar nebula used to construct the star and its orbiting exoplanets and are predominantly made of hydrogen. In the Solar System, Jupiter and Saturn have primary atmospheres that reflect the composition of the Sun. Secondary atmospheres originate from the geology (or biology) of an exoplanet. Their chemical abundances are markedly different from those of their stars. Earth, Mars, and Venus all have secondary atmospheres with chemistries that differ from that of the Sun—Earth’s atmosphere is predominantly composed of nitrogen, whereas those of Mars and Venus are dominated by carbon dioxide. Generally, we expect secondary atmospheres to be made up of heavier elements; if so, they would be more compact and thus more difficult to detect. Cooler atmospheres are expected to be cloudy, which would further complicate any detection.

All these factors combined imply that to study the atmospheres of smaller exoplanets that have sizes between that of Earth and Neptune, we have to invest much larger amounts of telescope time. Therefore, we will probably only be able to study about a dozen of these exoplanets. Our chances of extracting statistical trends from the atmospheres of these super Earths and mini-Neptunes with JWST are quite bleak. However, measuring the atmospheric properties of a dozen such objects with JWST would allow those exoplanets to serve as important benchmarks for the future. Exoplanets in this size range have no analog in our Solar System, and any new knowledge regarding their atmospheric conditions would be groundbreaking.

Getting Lucky: Probing a Second Earth

Studying Earth-like exoplanets with JWST will be even more elusive. An exoplanet referred to as “Earth-like” is usually one that is nearly the same size as Earth, has roughly the same mass, and orbits its star within a range of distances that would allow for liquid water to exist on the surface of the exoplanet. Until we develop a more general understanding of what life is, looking for exoplanets is our way of searching for life as we know it.

The prohibitive amount of telescope time needed to measure the light spectrum of an Earth-like exoplanet would restrict the exoplanet community to only a handful of targets. Because the signals are expected to be weak, we would ideally want to study one that is orbiting a nearby, bright star. The recently discovered exoplanet Proxima Centauri b, which at about four light years away is practically in our cosmic backyard, would have been an interesting target, but it does not appear to transit its star—the size of this exoplanet is unknown. Recording tiny dips in light as an exoplanet transits its star is easier if the star is bright. However, the star should not be so bright that it would saturate the instruments of JWST. Such a target currently does not exist in the catalogs of astronomers.

Three upcoming space missions are designed to address this shortcoming: CHEOPS (CHaracterizing ExOPlanet Satellites), TESS (Transiting Exoplanet Survey Satellite), and PLATO (PLAnetary Transits and Oscillations). After its launch in 2018, TESS will scan the entire sky to detect transiting small exoplanets around bright stars. Also launching in 2018, CHEOPS will take a more targeted approach and focus on one individual star at a time to search for transits. PLATO, because it will not launch until 2024 (or later), will probably not come soon enough to influence our use of JWST, but it could be decisive in detecting true Earth analogs—Earth-sized, Earth-mass exoplanets orbiting a twin of the Sun. These missions are expected to deliver a catalog of small transiting exoplanets that are amenable to atmospheric characterization by JWST. Even if a true Earth analog is found orbiting a bright star, using JWST to characterize its atmosphere will be a formidable task, because doing so involves measuring multiple transits to build a spectrum with a high degree of confidence—and one would have to wait a year between transits!

Another strategy is to build ground-based telescopes dedicated to searching for small exoplanets orbiting stars smaller than the Sun. These, too, could find good targets for JWST. Because the transit signal depends on the ratio of the size of the exoplanet to that of its star, a smaller star would imply a larger signal. Several of these ground-based surveys have sprung up in the past few years. The Belgian-led TRAPPIST (TRAnsiting Planets and PlanetesImals Small Telescope) survey targets transiting exoplanets orbiting ultracool stars that are cooler, redder, and much less massive than our Sun. The first discovery was astonishing—despite having a mass barely a tenth that of our Sun, the TRAPPIST-1 star hosts three Earth-sized exoplanets within orbits comparable to that of Mercury. Despite the closeness of these three exoplanets to the star, one of them may have atmospheric conditions that allow for water to exist at its surface. The closer distance also means that it is possible to record multiple transits within a reasonable amount of time.

A follow-up campaign with the Spitzer Space Telescope revealed the presence of four more Earth-sized exoplanets around the TRAPPIST-1 star. This discovery reaffirms a key finding of the Kepler mission: Rocky exoplanets are common, and finding one of them orbiting a star indicates that there are usually more. The Swiss-led SAINT-Ex project aims to harvest the same types of systems by building a similar ground-based telescope in the northern hemisphere. Since these exoplanets orbit stars that are less luminous, their surface conditions are expected to be more hospitable for life—in some cases, they may have Earth-like temperatures. Deciphering the atmospheres of Earth-like exoplanets is no longer a science-fiction fantasy, but it remains to be seen how many of these small exoplanets near red dwarfs will be suitable targets for JWST.

Digging into the Noise

To make exciting discoveries with JWST, astronomers must do the mundane but important work of thoroughly understanding its instruments—indeed, studying the limitations of this new equipment is at the frontier of the field. The four scientific instruments of JWST are NIRSpec (Near-Infrared Spectrograph), NIRISS (Near-Infrared Imager and Slitless Spectrograph), NIRCam (Near-Infrared Camera), and MIRI (Mid-Infrared Instrument). Collectively, they cover the range of wavelengths from 0.6 to 28 micrometers, going beyond what the Hubble and Spitzer were able to do. This versatile suite of instruments allows both imaging (taking photographs) and spectroscopy (splitting the light into different colors) of astronomical objects. There are various ways that each instrument may be used, depending on whether one values spectral resolution (how finely the light is split into colors) or signal-to-noise (how confidently the signal is detected at each color or frequency). These two aspects are related—the more finely light is split across frequency, the less confidently the signal is detected at each specific frequency.

Another important task is to understand the peculiarities of the telescope and its instruments during operation—known as the systematics. When trying to detect a weak signal, one must identify all sources of noise and confusion. The pixels on a detector may not all have the same sensitivity—and the sensitivity may even vary within a pixel. Pointing the telescope at the target and keeping it perfectly still are essentially impossible. Some kind of jitter will exist and introduce noise into the signal. Even the movement of secondary components of the telescope, such as an antenna, may introduce small but measurable disturbances into the signal. Thermal gradients may wreak havoc by introducing flexure into the telescope.

Astronomers using the Hubble and Spitzer space telescopes to study exoplanetary atmospheres experienced these issues in one way or another. In particular, the Hubble Space Telescope experienced something known as thermal breathing—a flexure of the telescope, recurring every 30 minutes, that resulted from it entering and exiting the Earth’s shadow. According to Harvard astronomer Laura Kreidberg, thermal breathing affected the accuracy of the near-infrared detectors on Hubble via a phenomenon known as charge trapping, in which impurities in the detectors capture and release photo-electrons. Intrapixel sensitivity was a central issue when using the Spitzer Space Telescope and prevented early studies of exoatmospheres from being definitive. It took a decade to hone the techniques for dealing with intrapixel sensitivity and to obtain results that were consistent with one another. Because JWST uses detectors that share a heritage with both Hubble and Spitzer, the lessons learned will be invaluable when JWST data start streaming in. The improved orbit of JWST (compared with Hubble’s) and its better pointing stability (compared with Spitzer’s) should attenuate these issues.

Astronomers at the Space Telescope Science Institute, which is the headquarters for the operation of Hubble and JWST, recently initiated an Early Release Science program. Its goal is to identify an exoplanet that has a high ecliptic latitude (such that it lies within the continuous viewing zone of JWST), a short orbital period (such that many transits may be recorded within a reasonable amount of time), and a known mass and radius, and that is orbiting a relatively bright, quiet star. After scrutinizing a list of targets, the hot Jupiter WASP-62b was determined to be the best target known so far. By observing WASP-62b with the four instruments of JWST in various configurations, we can build up a base of knowledge of the systematics and learn how to use each instrument optimally for transit spectroscopy. Even before the launch of JWST, the exoplanetary atmospheres community is poised for action.

Inferring Atmospheric Characteristics

Once the infrared spectrum has been distinguished from any other noise, astronomers will still need to figure out how to interpret it to deduce the characteristics of an exoplanet’s atmosphere. Fundamentally, a measured spectrum is a series of peaks and troughs of different shapes and sizes. Our ability to interpret a spectrum is based on the notion that the laws of physics and chemistry are universal—the spectrum of a water molecule measured in a laboratory on Earth should look the same as one measured from another location in the universe.

Atmospheres are nearly in hydrostatic equilibrium, in which gravity is balanced by pressure gradients, leading to an exponential variation of pressure (and matter) within an atmosphere. Light from the parent star of the exoplanet impinges on the atmosphere, and how it is absorbed depends on how the various atoms and molecules are distributed. The exact nature of these atoms and molecules matters as well—one would not expect a water molecule to absorb at the same frequencies, and with the same strengths, as a methane molecule. As starlight is absorbed, it sets up a temperature gradient in the atmosphere. The spectrum of an exoplanetary atmosphere that is measured by an astronomer is the radiation that escapes from this complex system. The astrophysicist’s task is to translate this spectrum back into knowledge of what types of atoms and molecules exist in the atmosphere, what their relative quantities are, and what the thermal structure of the atmosphere is. Some features of the climate, such as the geology of the surface, cannot be directly inferred from the spectrum.

Astrophysicists have borrowed a technique known as atmospheric retrieval that was honed in the atmospheric and climate sciences. It is used to retrieve the chemical and thermal properties of the exoplanetary atmosphere, given the measured spectrum. Several groups around the world have developed expertise in atmospheric retrieval, and are cutting their teeth applying this technique to study hot Jupiters. We now know that these hot Jupiters contain water and that some of them are cloudy; we even know that one of them glows blue in reflected light. We can make temperature maps of their daysides and nightsides. As the technology advances, we expect to obtain measured spectra of cooler exoplanetary atmospheres and to work our way toward a more Earth-like regime.

Ultimately, the technique may be applied to an Earth-like atmosphere to identify the types of atoms and molecules it contains, and to infer whether the signatures of life are hidden within the measured spectrum. If we get lucky, JWST may be able to find chemical hints of life on an exoplanet within a decade, although the challenge will be to uniquely identify a signal tied to the presence of life (rather than one that is consistent with the presence of life).

Beyond JWST

Always thinking ahead, NASA is already formulating plans for the successor to JWST. It currently has the generic placeholder name of the Large Ultraviolet/Optical/Infrared Surveyor (LUVOIR). (As with its predecessors, LUVOIR will be given a more catchy name as the mission progresses.) It has the ambitious goal of measuring spectra from the ultraviolet to the infrared, effectively combining the capabilities of Hubble, Spitzer, and JWST. One of its purposes is to characterize the atmospheres of habitable—or even inhabited—exoplanets. It is expected to fly in the 2030s—not anytime soon, but certainly within most of our lifetimes. This future telescope will be designed right from the start with exoplanets in mind. In the distant future, spectra from exoplanets indicating the presence of continents, oceans, and biosignature gasses may well be commonplace.