A neat technique, applied to the brightest transiting extrasolar planet known, identifies an indisputable signature of water vapour in the planet's atmosphere. The method might be used to probe a nearby habitable world.
When the first discovery1 of a planet orbiting a Sun-type star was announced in 1995, few could have imagined that we would stand where we do today: more than 300 planets outside our Solar System (exoplanets) are now known. The presence of most of these exoplanets has been inferred either through detecting their parent star's 'wobble', which is caused by the gravitational tug of the planet, or through measuring a dip in the host star's light as the planet passes in front of it (transits). The focus of exoplanet research is now shifting towards finding planets that are the most promising candidates for supporting life, and searching for indicators of biological processes (biomarkers) in their visible and infrared spectra. In this issue (page 767), Grillmair et al.2 report that, by using long-duration observations from NASA's Spitzer Space Telescope, they have found an unequivocal signature of water vapour in the atmosphere of the brightest transiting exoplanet yet detected, the gas giant HD 189733b.
The transit method of detecting exoplanets has been especially fruitful for studying these objects because it allows the planet's light to be measured using a geometric trick. Almost all planets that transit their stars will also pass behind the star at a known time. Measuring the system before or after this 'secondary' eclipse gives the light of the star plus its planet, but a measurement of the system when the planet is behind the star gives the star's light in isolation. Subtracting one measurement from the other reveals the contribution of the planet.
The secondary-eclipse method was the technique used to measure the first crude spectra of HD 189733b and another extrasolar planet, based on Spitzer observations3,4. But these pioneering measurements raised more questions than they answered. We know that oxygen is the most abundant element in the cosmos after hydrogen and helium, and that oxygen atoms should combine with hydrogen in gas-giant planets to produce water vapour. But the first spectra of HD 189733b did not show the expected signature of water vapour. Was the water missing, variable in abundance, or just hidden by high cloud layers? The reason for its absence in the first observations is still unclear. Other investigations have found signs of water in giant exoplanets5,6,7, but doubts have been slow to dissipate.
Grillmair et al.2 blast away any remaining uncertainty by pushing the secondary-eclipse technique to its limit — averaging the results from ten eclipses of HD 189733b. The fruit of their persistence is a characteristic peak in the planet's spectrum that arises from a specific bending mode of the water molecule. The authors show conclusively that water vapour is present in the atmosphere of this hot gas giant world, validating a key aspect of our current models for these kinds of planets.
Giant hot exoplanets are fascinating, but what we ultimately want to find are life-bearing worlds. That search should target small, rocky or icy exoplanets, not gas giants, because the thick hydrogen–helium envelopes of gas giants are not where we expect to find life. The detection of atmospheric biomarkers will be crucial to finding life. However, although water is deemed to be necessary for life, atmospheric water vapour is not a biomarker; it can be seen in a wide variety of astrophysical environments, even in sunspots8. Nevertheless, we must first learn how to detect abundant molecules such as water before we can advance to identifying the more subtle signatures that scarcer molecules such as molecular oxygen leave in exoplanet spectra. Grillmair and colleagues have taken that first step.
But how do we proceed to find biomarkers in the atmosphere of a habitable exoplanet? The most obvious way would be to image a planet that is a twin of Earth — that is, one with an Earth-like mass and distance from its parent star — orbiting a nearby Sun-type star. This could be achieved by using high-contrast imaging observations to separate the planet's light from the bright glare of its star. Astronomers have successfully imaged massive planets orbiting far from their stars9,10, but imaging a twin of our Earth will be a much more difficult task because terrestrial planets are orders of magnitude fainter and closer to the glare of their star. Using imaging would also allow us to obtain spectra and thereby find a potentially habitable world, but such a find seems many years away.
Fortunately, there is a quicker way to identify potentially life-bearing worlds in our close cosmic neighbourhood. Our nearest stellar neighbours are predominantly red-dwarf stars. These small, cool stars seem to lack gas-giant planets, but harbour rocky 'super-Earth' planets in abundance. Moreover, the habitable zone — the area around a star where water is in liquid form and life is potentially sustainable — is tucked in close to these stars. This is exactly the region where, as viewed from Earth, planets are more likely to transit the star. The odds seem good that in the next few years we will find a super-Earth exoplanet that transits a nearby red-dwarf star in, or close to, its star's habitable zone.
The search for a transiting super-Earth has already begun. David Charbonneau and his collaborators recently inaugurated the MEarth (pronounced 'mirth') Project11. MEarth uses a collection of 0.4-metre-diameter robotic ground-based telescopes to monitor the nearest 2,000 red dwarfs visible from the Northern Hemisphere in the search for the transit of a rocky planet. MEarth is fully operational at the Fred Lawrence Whipple Observatory in Arizona, and the robotic telescopes survey our nearest stellar neighbours every clear night.
Also, NASA's prospective Transiting Exoplanet Survey Satellite mission could potentially find a rocky habitable planet transiting a nearby small star. Once we find such a planet, we will turn to the forthcoming James Webb Space Telescope to measure its spectrum, using the technique that Grillmair et al.2 have pioneered. Orbiting a red dwarf, this habitable super-Earth may be a bizarre and unfamiliar world (Fig. 1), hosting 'life as we don't know it'. Many investigators think that this prospect is as exciting as finding a twin of our Earth, whose rarity is still an open question.