The lack of absorption features in the transmission spectrum of exoplanet GJ 1214b rules out a hydrogen-rich atmosphere for the planet. It is consistent with an atmosphere rich in water vapour or abundant in clouds. See Letter p.669
Sometimes the most telling evidence comes not from what is observed but from what is not observed. On page 669 of this issue, Bean and colleagues1 report results of the latter type for the transiting 'super-Earth' exoplanet GJ 1214b.
This nearby world2 (it is only about 13 parsecs away from Earth) belongs to the special category of transiting planets. When a planet transits — passes in front of its star as seen from the vantage point of Earth — we can measure its radius from the amount of stellar light that it blocks. By adding precision spectroscopic data, we can also determine its mass from the Doppler 'wobble' that it induces in the parent star's motion. Knowing the mass and radius of an exoplanet is a major step towards characterizing its nature. The mass and radius of GJ 1214b imply that it almost certainly has a massive atmosphere3.
In their study, Bean et al.1 have pushed the methodology even farther than measuring the mass and radius of of GJ 1214b. Their measurements offer the first direct probe of the atmosphere of an extrasolar super-Earth. Super-Earths are planets two to ten times more massive than Earth, and GJ 1214b weighs in at 6.5 Earth masses. Specifically, the authors measured the amplitude of the transit — the amount of starlight that the planet blocks — as a function of wavelength. Molecules such as water vapour in the planet's atmosphere can absorb starlight during the transit, and can do so more strongly at some wavelengths than at others, making the amplitude of the transit wavelength-dependent. The pattern of absorption potentially allows specific molecules to be identified.
Using data of exquisite sensitivity, Bean et al. show that the transmission spectrum of GJ 1214b is a smooth function of wavelength, with no bumps or wiggles that can be attributed to absorption by atmospheric molecules. It is this absence of specific spectral features that makes the results so intriguing. The simplest molecule, molecular hydrogen, is the easiest to measure, albeit indirectly. Molecular hydrogen produces no absorption features of its own at readily measured wavelengths, but its low molecular mass allows the putative atmosphere to extend to high altitudes. This spreads all of the constituents of the atmosphere over a greater height range, and allows transmitted starlight to interact with many absorbing atoms and molecules (Fig. 1a). Transmission spectra of gas-giant exoplanets4,5 show detectable spectral features largely for this reason. One signature of a hydrogen-rich atmosphere surrounding a transiting super-Earth will therefore be the ease with which the absorption features are detected. Bean and colleagues' high-precision spectral data1 rule out a hydrogen-rich atmosphere for GJ 1214b — a significant advance in the field of exoplanetary atmospheric science.
An irony of transit spectroscopy is that atmospheres rich in strongly absorbing complex molecules but poor in weakly absorbing hydrogen will not necessarily lead to a strong absorption signal. Paradoxically, they will tend to produce an absence of spectral absorption features. Hydrogen-poor atmospheres, having greater average molecular masses than hydrogen-rich atmospheres, are pulled by a planet's gravity to lower altitudes, where they intercept relatively few photons from the parent star (Fig. 1b). These low-lying atmospheres, even if they are rich in complex molecules, produce very weak absorption features6.
One possible interpretation of Bean and colleagues' results1 showing a lack of absorption features is that this extrasolar super-Earth has an atmosphere rich in molecules heavier than hydrogen. Among molecules heavier than molecular hydrogen, the most cosmically abundant possibility is water. Hence, one particularly intriguing explanation for the authors' results is that the planet is surrounded by an atmosphere rich in water vapour. However, another — and at the moment equally valid — interpretation of their data puts clouds in view for exoplanet transit spectroscopy, both literally and figuratively.
Bean and colleagues' observations are consistent with abundant clouds and haze in the atmosphere of GJ 1214b. Clouds in the atmospheres of giant exoplanets were inferred from the first detection of an exoplanet atmosphere4, and other results for giant exoplanets have conclusively demonstrated the existence of hazy atmospheres7. Clouds and haze intercept and block starlight as it passes through the atmospheres of transiting planets (Fig. 1c), weakening or totally obscuring absorption features. In addition to real clouds, figurative clouds have recently gathered over exoplanet transit spectroscopy: detections of molecular absorptions in data from the Hubble Space Telescope for several giant exoplanets have recently been challenged, and attributed to uncorrectable instrumental effects8.
Fortunately, both the literal and figurative clouds should clear for transit spectroscopy of exoplanets. Spectroscopy of GJ 1214b in the near infrared has already been scheduled for Hubble's Wide Field Camera 3 (WFC3). WFC3 observations will probe a longer wavelength than was available to Bean et al., and hazy atmospheres can often be clearer at longer wavelengths. As for giant planets, a new and extensive Hubble programme using WFC3 should clarify many questions concerning their molecular absorptions.
On a longer timescale, astronomers await the advent of the James Webb Space Telescope, which will not only provide excellent sensitivity, but also operate at long infrared wavelengths. At sufficiently long infrared wavelengths, haze and clouds tend to become transparent. Moreover, many molecules have their strongest absorption bands in the long-wavelength infrared region. Sufficiently strong bands can imprint detectable signals on the small portion of the transmitted light that misses the clouds. The James Webb Space Telescope should blow away any remaining clouds surrounding exoplanet spectroscopy, and give us the clearest view yet.
Bean, J. L., Kempton, E. M.-R. & Homeier, D. Nature 468, 669–672 (2010).
Charbonneau, D. et al. Nature 462, 891–894 (2009).
Miller-Ricci, E. & Fortney, J. J. Astrophys. J. 716, L74–L79 (2010).
Charbonneau, D., Brown, T. M., Noyes, R. W. & Gilliland, R. W. Astrophys. J. 568, 377–384 (2002).
Redfield, S., Endl, M., Cochran, W. D. & Koesterke, L. Astrophys. J. 673, L87–L90 (2008).
Miller-Ricci, E., Seager, S. & Sasselov, D. Astrophys. J. 690, 1056–1067 (2009).
Pont, F., Knutson, H., Gilliland, R. L., Moutou, C. & Charbonneau, D. Mon. Not. R. Astron. Soc. 385, 109–118 (2008).
Gibson, N. P., Pont, F. & Aigrain, S. Mon. Not. R. Astron. Soc. (in the press); preprint at http://arxiv.org/abs/1010.1753 (2010).
About this article
The Astrophysical Journal (2014)