The Next Great Exoplanet

What strange new worlds will our next-generation telescopes find? Kevin Heng is an assistant professor of astrophysics at the University of Bern, Switzerland (Twitter: @KevinHeng1). He is a core science team member of the CHEOPS mission and is involved in the PLATO mission. Joshua Winn is associate professor of physics at the Massachusetts Institute of Technology, was a member of the Kepler science team, and is currently deputy science director of the TESS mission. Edited by Katie Burke. www.americanscientist.org One of the most stunning scientific advances of our generation has been the discovery of planets around distant stars. Fewer than three decades ago, astronomers could only speculate on the probability for a star to host a system of exoplanets. Now we know the Universe is teeming with exoplanets, many of which have properties quite unlike the planets of our Solar System. This has given birth to the new field of exoplanetary science, one of the most active areas of astronomy, pursued by many research groups around the world. Although there are many ways of detecting exoplanets, the most successful in recent years has been the transit method, in which the exoplanet reveals itself by transiting (passing directly in front of) its host star, causing a miniature eclipse. These eclipses are detected by telescopes that are capable of precisely tracking the brightnesses of many stars at the same time. Since its launch in 2009, NASA’s Kepler space telescope used this technique to resounding effect, finding more than 1,000 confirmed transiting exoplanets. The success of the Kepler mission has inspired the next generation of exoplanet transit-hunting machines: a fleet of new space telescopes and complementary ground-based telescopes. They will expand our inventory of strange new worlds and further our quest to find Earth-like exoplanets and search them for signs of life. The long-term strategy of exoplanet hunting is easy to state: find exoplanets, characterize their atmospheres and, ultimately, search for chemical signatures of life (biomarkers). These efforts strive to answer questions such as: Which molecules are the most abundant? What is the typical cloud coverage, temperature and wind speed? Is there a solid surface beneath all the gas? The answers to these questions help establish whether an exoplanet is potentially habitable. And unless we can detect radio broadcasts from an intelligent civilization or build starCorresponding authors: [email protected] (Kevin Heng), [email protected] (Joshua Winn) ships to visit the exoplanet itself, these atmospheric characterizations seem the most promising—and to some, the only—way of discerning whether a given exoplanet is actually inhabited. Molecular oxygen, or ozone, for example, would be circumstantial, but not definitive, evidence for life, because there are plausible ways of producing it abiotically using the known laws of physics and chemistry. Subsurface life may exist on an exoplanet, but we probably have no chance of detecting it with astronomical observations. This logical sequence of exploration is largely mirrored in the series of space missions that NASA and the European Space Agency (ESA) have planned, approved, and constructed—or will construct. Exoplanet detection is poised to make the transition from cottage industry to big science, while atmospheric characterization is inexorably finding its feet. Bright, Twinkling Stars Early efforts to detect transiting exoplanets in the early 2000s used ground-based telescopes, which face daunting obstacles. One obstacle is the annoying tendency of the Sun to rise every morning. This foils the attempt to monitor stars as continuously as possible. Transits that occur during daytime are missed. Another obstacle is the Earth’s atmosphere, a turbulent and constantly varying screen that causes apparent fluctuations in starlight much stronger than the transit signals we seek to detect. Even on a seemingly clear night, the atmosphere alters the passage of starlight in a manner that depends on both wavelength and brightness, through the phenomena of extinction and scintillation. Extinction is why the setting sun is red and dim; scintillation is why stars twinkle. To correct for these effects, astronomers point their telescopes at groups of stars with similar colors and brightness, trusting that the atmosphere will affect them all equally—whereas a transiting exoplanet would cause only one of them to temporarily dim. Using these reference stars, exoplanet transits can be flagged and confirmed. Published in American Scientist: Volume 103, Number 3, Pages 196–203 (May-June 2015 Issue) July 1, 2015 ar X iv :1 50 4. 04 01 7v 2 [ as tr oph .E P] 3 0 Ju n 20 15 Heng & Winn: The Next Great Exoplanet Hunt (in American Scientist) 2 0.996 0.998 1.000 1.002 Re la tiv e br ig ht ne ss Ground−based 1.2m telescope (FLWO) 0 2 4 6 8 Time [hours] 0.996 0.998 1.000 1.002 Re la tiv e br ig ht ne ss Space−based 1.0m telescope (Kepler) Figure 1: Why we need space telescopes to observe transiting exoplanets. The top panel shows a time series of brightness measurements of the exoplanet HAT-P-11b, using an Arizona-based telescope with a 1.2-meter diameter. The dip of 0.4%, lasting a few hours, is due to the miniature eclipse of the star by an exoplanet a shade larger than Neptune. The large scatter in the measurements is due to failure to correct adequately for atmospheric effects and the interruption just after the transit is due to morning twilight. The bottom panel shows an observation of the same exoplanet with the 1.0-meter Kepler space telescope. Despite being a smaller telescope, the absence of atmospheric effects and the uninterrupted observations enable a vastly improved view of the transit. It is even possible to tell that the exoplanet crossed over a dark spot on the surface of the star (known as a “star spot”), based on the slight upward flicker that was seen in the second half of the transit (at a time coordinate of about 4.5 hours). From our vantage point on Earth, bright stars are rarer than faint ones, simply because of geometry—casting our net farther into space would create a celestial sphere that encompasses more stars, which tend to be fainter on average because of the greater distances involved. The brightest stars are distributed uniformly across the entire sky and are therefore widely separated in angle. Finding reference stars to confirm an exoplanet transit becomes harder for bright stars: the only comparable reference stars may be located so far away on the sky that one can no longer assume they are affected equally by the atmosphere. This helps to explain the otherwise paradoxical fact that the “best and brightest” few thousand stars in the sky are relatively unexplored for transiting exoplanets. Astronomers realized that these difficulties may be circumvented by launching telescopes into space. Although he was not specifically thinking about exoplanet transits, the late Princeton astronomer Lyman Spitzer lobbied tirelessly in the 1960s and 1970s for what would later become known as the Hubble Space Telescope. Placed above Earth’s atmosphere eliminates some of the obstacles to precise measurements. For exoplanet hunters, there is no need to find bright reference stars and the precision is limited only by photon-counting noise (sometimes termed “shot noise”)—the inevitable fluctuations in the signal caused by the discrete and random nature of photon emission. Even with a space telescope, detecting transiting exoplanets is a fight against long odds. The vast majority of exoplanets do not transit. The exoplanet’s orbit must be aligned nearly edgeon, as we see it from Earth, in order for transits to occur. The probability for transits is equal to the stellar diameter divided by the orbital diameter, which is only 0.1% for an Earth-like orbit around a Sun-like star. For this reason, a meaningful transit survey must include tens of thousands of stars, or more. Because faint stars far outnumber bright ones in any given region of the sky, a practical strategy is to monitor a rich field of relatively faint stars. This is precisely what the Kepler space telescope did, staring at about 150,000 stars in a small, 115-square-degree patch of sky in the constellations of Cygnus and Lyra. Unfortunately, most of the Kepler discoveries involve stars that are too faint for the atmospheric characterization of their exoplanets. Transiting exoplanets offer the potential for atmospheric characterization. During a transit, a small portion of the starlight is filtered through the exoplanet’s atmosphere. By comparing the spectrum of starlight during a transit to the starlight just before or after the transit, we can identify absorption features due to atoms and molecules in the exoplanet’s atmosphere. Another approach is to measure the light emitted from the exoplanet, by detecting the drop in brightness when the exoplanet is hidden behind the star, an event known as an “occultation”. The practical value of transits and occultations depends critically on the brightness of the star. The higher the rate of photons (particles of light) that a star delivers to Earth, the faster we can accumulate information about its exoplanets. This is because many astronomical observations are limited by photoncounting noise. The greater the number of photons we collect, the smaller the effects of shot noise and the higher the signalto-noise ratio. This in turn allows us to search for smaller exoplanets, which produce smaller transit signals. To put the smallness in perspective, an Earth-sized exoplanet transiting a Sunlike star produces a total dimming of only 84 parts per million and its atmospheric signal would be smaller by at least an order of magnitude. To measure the spectrum of an exoplanetary atmosphere, one needs to divide up the starlight according to wavelength. In essence, a small signal must be split into even smaller ones, an endeavour only feasible if the signal-to-noise ratio is very high at the outset. And this is only possible for bright stars. The Kepler stars are typically more than a million times fainter than the brightest naked-eye stars such as Sirius, Vega and Alpha Centauri. These challenges are easier to overcome when the exoplanet is large, is located close to its star and has a “light” (low mean molecular mass) or nonmetallic atmosphere. This is why hot Jupiters, hydrogen-dominated gas giants orbiting their stars closer than Mercury is to our Sun, were the first to have their atmospheres characterized by astronomers. To appreciate how demanding the observations can be, consider the transiting super-Earth known as GJ 1214b, the most favorable system for atmospheric characterization among all the known transiting exoplanets smaller than Neptune, because it is located relatively close to us and resides in a 1.6-day orbit. Even in this most favorable case, characterizing GJ 1214b’s atmosphere required observing with the Hubble Space Telescope for a record-breaking 60 orbits, which translates roughly into a cost of about $12 million. (Observing time on the Hubble Space Telescope is quantified by the number of 90-minute orbits of the spacecraft around the Earth and is awarded competitively. Heng & Winn: The Next Great Exoplanet Hunt (in American Scientist) 3