Science - USA (2018-12-21)

SCIENCE sciencemag.org 21 DECEMBER 2018 • VOL 362 ISSUE 6421 1361

GRAPHIC: M. BROGI

exoplanets. As the lightest element, it can

be stripped off very effectively, forming an

extended cloud of gas around the planet,

possibly with an elongated shape similar

to cometary tails. If the planet transits its

parent star as seen from Earth, this escap-

ing gas can absorb substantially more star-

light than the planet itself, revealing its

presence during transit observations. This

method has allowed scientists to detect

massive “tails” of hydrogen around several

exoplanets; the most dramatic example is

the Neptune-size exoplanet GJ 436 b ( 8 ).

As neutral hydrogen absorbs at ultra-

violet (UV) wavelengths, it is only possible

to observe it from space, where UV light

is not shielded by ozone in Earth’s atmo-

sphere. However, even at the altitude of

the Hubble Space Telescope (about 568

km), hydrogen in Earth’s exosphere (outer-

most region of the atmosphere) produces a

contaminating signal (the so-

called geo-coronal emission).

Moreover, hydrogen gas per-

meates interstellar space, mak-

ing it difficult to distinguish

whether the absorption occurs

around the planet or anywhere

else along our line of sight. In

addition, stellar activity alters

the strength of hydrogen spec-

tral lines, complicating the

interpretation of variable ab-

sorption ( 9 ). As a result, only

the “wings” of hydrogen spec-

tral lines can be reliably as-

sociated with exoplanet mass

loss. This corresponds to gas

moving at tens to hundreds of

kilometers per second, already

quite far from the planet.

Helium is the second most

abundant element in the atmo-

spheres of giant planets, and it

is still light enough to be eas-

ily stripped off. In the near in-

frared, around a wavelength of 10,830 Å,

there is a particular helium spectral line

that can form by absorbing radiation from

a metastable state (a state where electrons

decay but with very long time scales—

hours in this case). Whereas this line can

form at the temperatures and densities of

planetary exospheres, interstellar helium

absorption is 1000 times less likely than

for hydrogen. This means that excess ab-

sorption due to a transiting planetary at-

mosphere can be sought even in the very

core of the helium line, allowing small ve-

locities to be probed. These give us direct

access to the region where atmospheric es-

cape originates.

Helium absorption was predicted at the

dawn of exoplanet science ( 10 ). Absorp-

tions up to a few percent in the core of the

helium line—more than 10 times as strong

as typical features in exoplanet spectra—

were estimated for two well-known transit-

ing exoplanets, the hot giant HD 209458

b and the hot Neptune-size GJ 436 b ( 11 ).

Observations with the WFC3 instrument

onboard the Hubble Space Telescope ( 12 )

detected helium in the transmission spec-

trum of the exoplanet WASP-107 b. How-

ever, owing to the low spectral resolution,

this excess absorption could only be seen

in one of the spectral channels, prevent-

ing the unambiguous identification of the

planetary origin of the signal.

Fortunately, infrared radiation around

11,000 Å penetrates Earth’s atmosphere

and therefore helium measurements can be

attempted with ground-based telescopes,

leveraging larger mirrors and higher spec-

tral resolution. Allart et al. and Nortmann

et al. resolved the detailed shape of the he-

lium spectral line and detected its chang-

ing velocity due to the radial component of

the planet’s orbital motion. This “Doppler”

effect causes a shift in the characteristic

absorption frequency of a spectral line for

nonzero velocity.

Allart et al. used the line shape informa-

tion to model the physical environment

around the warm Neptune-size planet

HAT-P-11 b. They derived a spherical shape

for the cloud of helium around the planet,

but no sizeable tail trailing the orbit. This

is in stark contrast to the extended hydro-

gen comas such as the one detected around

GJ 436 b, but in line with expectations

from interactions with the stellar radiation

field and winds. Nortmann et al. report a

strong helium absorption surrounding the

Saturn-size exoplanet WASP-69 b. By com-

paring their detection with a small sample

of exoplanets, they propose that the pres-

ence of escaping helium depends on the in-

tensity of x-rays and extreme-UV radiation

from the parent star, which in turn regu-

lates the population of the metastable state

of helium responsible for the 10,830 Å line.

Thus, only planets hosted by moderately

active stars should show helium absorp-

tion, in line with observations to date.

The findings of Allart et al. and Nort-

mann et al. reinforce the importance of

stellar environments in shaping the atmo-

spheric evolution of exoplanets. Attention

of the scientific community is turning to-

ward potentially habitable planets around

M-dwarf stars, which are particularly ac-

tive. Studying atmospheric escape across

a wide class of planetary systems will be

crucial for assessing how hab-

itability is influenced by en-

ergetic stellar radiation ( 13 ).

The measurements presented

by Allart et al. and Nortmann

et al. were obtained with a

new spectrograph called CAR-

MENES (Calar Alto high-reso-

lution search for M dwarfs with

exoEarths with near-infrared

and optical echelle spectro-

graphs) originally designed to

look at minute radial velocity

variations of stars due to the

gravitational influence of or-

biting exoplanets, but also ex-

celling in frontier atmospheric

characterization. As more

spectrographs like CARMENES

will soon become available, it

is expected that the synergy

between space and ground-

based observatories will inten-

sify, with the ultimate aim of

unveiling the atmospheres of

temperate exoplanets in the not-too-dis-

tant future. j

REFERENCES

1. R. Allart et al., Science 362 , 1384 (2018).


2. L. Nortmann et al., Science 362 , 1388 (2018).


3. H. Lammer et al., Astrophys. J. 598 , 121 (2003).


4. R. A. Murray-Clay, E. I. Chiang, N. Murray, Astrophys. J. 693 ,
   23 (2009).


5. B. J. Fulton et al., Astron. J. 154 , 109 (2017).


6. J. E. Owen, Y. Wu, Astrophys. J. 847 , 29 (2017).


7. A. Vidal-Madjar et al., Nature 422 , 143 (2003).


8. D. Ehrenreich et al., Nature52 2, 459 (2015).


9. P. W. Cauley, S. Redfield, A. G. Jensen, Astron. J. 153 , 185
   (2017).


10. S. Seager, D. D. Sasselov, Astrophys. J. 537 , 916 (2000).


11. A. Oklopčić, C. M. Hirata, Astrophys. J. 855 , L1 (2018).


12. J. J. Spake et al., Nature 557 , 68 (2018).


13. S. Rugheimer, L. Kaltenegger, A. Segura, J. Linsky,
    S. Mohanty, Astrophys. J. 809 , 57 (2015).
    10.1126/science.aav7010

Hydrogen-Lyman a-1,216Å

Planet

Helium-10,830Å

UV

Infrared

Star

Escaping gases

Atmospheric hydrogen and helium gases of an exoplanet that orbits closely

to its parent star can escape under strong irradiation. The escaping gas

can absorb more starlight than the exoplanet and can be detected by remote,

ground-based spectroscopy.

Published by AAAS

on December 21, 2018^

http://science.sciencemag.org/

Downloaded from