A continuum from clear to cloudy hot-Jupiter exoplanets without
primordial water depletion
Sing, D. K., Fortney, J. J., Nikolov, N., Wakeford, H. R., Kataria, T., Evans, T. M., Aigrain, S., Ballester, G. E.,
Burrows, A. S., Deming, D., Désert, J-M., Gibson, N. P., Henry, G. W., Huitson, C. M., Knutson, H. A., Etangs, A.
L. D., Pont, F., Showman, A. P., Vidal-Madjar, A., ... Wilson, P. A. (2016). A continuum from clear to cloudy hot-
Jupiter exoplanets without primordial water depletion.
Nature
,
529
(7584), 59-62.
https://doi.org/10.1038/nature16068
Published in:
Nature
Document Version:
Peer reviewed version
Queen's University Belfast - Research Portal:
Link to publication record in Queen's University Belfast Research Portal
Publisher rights
© 2016 Macmillan Publishers Limited.
General rights
Copyright for the publications made accessible via the Queen's University Belfast Research Portal is retained by the author(s) and / or other
copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated
with these rights.
Take down policy
The Research Portal is Queen's institutional repository that provides access to Queen's research output. Every effort has been made to
ensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in the
Research Portal that you believe breaches copyright or violates any law, please contact openaccess@qub.ac.uk.
Download date:09. Aug. 2022
A continuum from clear to cloudy hot-Jupiter exoplanets
without primordial water depletion
David K. Sing
1
, Jonathan J. Fortney
2
, Nikolay Nikolov
1
, Hannah R. Wakeford
1
, Tiffany
Kataria
1
, Thomas M. Evans
1
, Suzanne Aigrain
3
, Gilda E. Ballester
4
, Adam S. Burrows
5
,
Drake Deming
6
, Jean-Michel Désert
7
, Neale P. Gibson
8
, Gregory W. Henry
9
, Catherine M.
Huitson
7
, Heather A. Knutson
10
, Alain Lecavelier des Etangs
11
, Frederic Pont
1
, Adam P.
Showman
4
, Alfred Vidal-Madjar
11
, Michael H. Williamson
9
, Paul A. Wilson
11
!
1
Astrophysics Group, School of Physics, University of Exeter, Stocker Road, Exeter, EX4 4QL, UK. !
2
Department of Astronomy and Astrophysics, University of California, Santa Cruz, CA 95064, USA.!
3
Department of Physics, University of Oxford, Keble Road, Oxford OX1 3RH, UK.!
4
Lunar and Planetary Laboratory, University of Arizona, Tucson, Arizona 85721, USA. !
5
Department of Astrophysical Sciences, Peyton Hall, Princeton University, Princeton, NJ 08544, USA.!
6
Department of Astronomy, University of Maryland, College Park, MD 20742 USA.
!
7
Department of Astrophysical and Planetary Sciences, University of Colorado, Boulder, CO 80309, USA.!
8
European Southern Observatory, Karl-Schwarzschild-Str. 2, D-85748 Garching bei Munchen, Germany.
!
9
Center of Excellence in Information Systems, Tennessee State University, Nashville, TN 37209, USA.
!
10
Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA 91125 USA.!
11
CNRS, Institut dAstrophysique de Paris, UMR 7095, 98bis boulevard Arago, 75014 Paris, France.!
!
!
Thousands of transiting exoplanets have been discovered, but spectral analysis of their
atmospheres has so far been dominated by a small number of exoplanets and data
spanning relatively narrow wavelength ranges (such as 1.1 to 1.7 µm). Recent studies
show that some hot- Jupiter exoplanets have much weaker water absorption features in
their near-infrared spectra than predicted
1–5
. The low amplitude of water signatures
could be explained by very low water abundances
6–8
, which may be a sign that water was
depleted in the protoplanetary disk at the planet’s formation location
9
, but it is unclear
whether this level of depletion can actually occur. Alternatively, these weak signals could
be the result of obscuration by clouds or hazes
1–4
, as found in some optical spectra
3,4,10,11
.
Here we report results from a comparative study of ten hot Jupiters covering the
wavelength range 0.3–5 micrometres, which allows us to resolve both the optical
scattering and infrared molecular absorption spectroscopically. Our results reveal a
diverse group of hot Jupiters that exhibit a continuum from clear to cloudy atmospheres.
We find that the difference between the planetary radius measured at optical and infrared
wavelengths is an effective metric for distinguishing different atmosphere types. The
difference correlates with the spectral strength of water, so that strong water absorption
lines are seen in clear-atmosphere planets and the weakest features are associated with
clouds and hazes. This result strongly suggests that primordial water depletion during
formation is unlikely and that clouds and hazes are the cause of weaker spectral
signatures.
A continuum from clear to cloudy hot-Jupiter exoplanets without primordial water depletion 2
We observed the transits of eight hot Jupiters as part of a spectral survey of exoplanet
atmospheres with the Hubble Space Telescope (HST). The eight planets covered in our survey
(WASP-6b, WASP-12b, WASP-17b, WASP-19b, WASP-31b, WASP-39b, HAT-P-1b, HAT-
P-12b) span a large range of planetary temperature, surface gravity, mass and radii, allowing
for an exploration of hot-Jupiter atmospheres across a broad range of physical parameters (see
Table 1). In this survey, we observed all eight planets in the full optical wavelength range (0.3-
1.01 µm) using the Space Telescope Imaging Spectrograph (STIS) instrument. We also used
the Wide Field Camera 3 (WFC3) instrument to observe transits of WASP-31b and HAT-P-1b
in the near-infrared (1.1-1.7 µm), and used additional WFC3 programs to observe transits of
four other survey targets (WASP-12b, WASP-17b, WASP-19b and HAT-P-12b). The HST
survey was complemented by photometric transit observations of all eight targets at 3.6 and 4.5
µm using the Spitzer Space Telescope Infrared Array Camera (IRAC) instrument. We analyzed
the survey targets in conjunction with HST and Spitzer data from the two best-studied hot
Jupiters to date, HD 209458b
1
and HD 189733b
5
, giving a total of ten exoplanets in our
comparative study with transmission spectra between 0.3 and 5 µm (see Extended Data Table
1 for a detailed list of the observations).
Our data reduction methods followed those in our previous studies
3,4,11-14
where the
transmission spectra of WASP-19b, WASP-12b, HAT-P-1b, WASP-6b, and WASP-31b have
been presented (see Methods for further details). The transit light curves
15
of the band-
integrated spectra were fit simultaneously with detector systematics, with all HST and Spitzer
transit data used to determine the planets’ orbital system parameters (inclination, stellar density,
and transit ephemeris), which were then fixed to the weighted mean values in the subsequent
analysis measuring the transmission spectra. To create the broad-band transmission spectrum,
we extracted various wavelength bins for the HST STIS and WFC3 spectra and separately fit
each bin for the planet-to-star radius ratio R
p
/R
*
and detector systematics. The uncertainties for
each data point were rescaled based on the standard deviation of the residuals, and any
systematic errors correlated in time were measured using the binned residuals
16
.
The resulting transmission spectra are shown in Fig. 1 and exhibit a variety of spectral
absorption features due to Na, K, and H
2
O, as well as strong optical scattering slopes (e.g.
WASP-6b and HAT-P-12b). Planets such as WASP-39b show prominent alkali absorption
lines with pressure-broadened wings, whereas other planets such as WASP-31b show strong
but narrow alkali features, implying they are limited to lower atmospheric pressures. H
2
O
vapour has been predicted to be a significant source of opacity for hot Jupiter atmospheres
17-19
,
and it is detected in five of the eight exoplanets where WFC3 spectra are available
1-5,13,14
.
However, the amplitude of the H
2
O absorption varies significantly across the ten planets,
ranging from features that are very pronounced (as in WASP-19b)
14
to those that are
significantly smaller than expected (HD 209458b)
1
or even absent (WASP-31b)
4
.
Previous studies using HST/WFC3 spectra have shown that HD 209458b, HD 189733b,
and WASP-12b have low-amplitude water features
1,3,5
which can be attributed to a severe
depletion of atmospheric H
2
O abundance relative to solar values
6-8
. Any such depletion would
be a remnant of planet formation, as H
2
O is expected to be well-mixed in a hot atmosphere,
such that currently measured molecular abundances would be consistent with primordial values.
The depletion of water vapour can occur beyond a protoplanetary disk’s snow line
9
, where
water is found predominantly as solid ice. Therefore, a hot Jupiter with a large depletion in H
2
O
A continuum from clear to cloudy hot-Jupiter exoplanets without primordial water depletion 3
gas would imply the planet formed at large orbital distances beyond the snow line and, during
its inward orbital migration, avoided accretion and dissolution of icy planetesimals as well as
the subsequent accretion of appreciable H
2
O-rich gas. Such scenarios have been proposed for
Jupiter
20,21
based on Galileo probe measurements
22
that indicate it is a water-poor gas giant,
though the measurements were affected by local meteorology
22
.
However, it is possible that these weak water absorption bands could be attributed to
cloud opacity, which have yielded featureless transmission spectra for a number of transiting
exoplanets
23,24
. For simplicity we define a cloud as a grey opacity source, while a haze as one
that yields a Rayleigh-scattering-like opacity, which could be due to small (sub-µm size)
particles. Silicate or higher temperature cloud condensates are expected to dominate the hotter
atmospheres, like those observed for brown dwarfs, while in cooler atmospheres sulphur-
bearing compounds are expected to play a significant role in the condensation chemistry
25,26
.
In Fig. 2, we plot model atmospheric pressure-temperature (P-T) profiles for the planets in our
comparative study and compare them to condensation curves for expected cloud-forming
molecules. The base, or bottom, of condensate clouds is expected to form where the planetary
P-T profiles cross the condensation curve; in this case, Cr, MnS, MgSiO
3
, Mg
2
SiO
4
and Fe are
possible condensates. For example, the spectra of WASP-31b shows clouds
4
, which likely form
at pressures of ~10 mbar and can be explained by Fe or MgSiO
4
condensates. However, the
curves alone cannot explain cloud versus cloud-free planets, as hazy planets such as HAT-P-
12b and WASP-12b do not cross condensation curves at observable pressures. Therefore,
atmospheric circulation must also play a role, as vertical mixing allows for particles to be lofted
and maintained at pressures probed in transmission at the terminators. Additionally, equatorial
eastward superrotation arising from day-night temperature variations can allow for clouds that
form on the nightside to be transported to the terminator
27
.
Table 1 | Physical parameters of hot Jupiters and associated spectral results.
Name
T
eq
K
g
m/s
2
R
p
R
J
M
p
M
J
P
day
log R'
HK
ΔZ
UB-LM
/H
eq
ΔZ
J-LM
/H
eq
H
2
O amp
(%)
Features
Ref
WASP-17b
1740
3.6
1.89
0.51
3.73
−5.531
−0.80±0.36
−1.48±0.71
94±29
Na, H
2
O
WASP-39b
1120
4.1
1.27
0.28
4.06
−4.994
0.10±0.41
Na, K
HD209458b
1450
9.4
1.36
0.69
3.52
−4.970
0.73±0.36
−0.49±0.36
32±5
Aer, Na, H
2
O
WASP-19b
2050
14.2
1.41
1.14
0.79
−4.660
1.04±1.79
−1.97±1.32
105±20
H
2
O
14
HAT-P-1b
1320
7.5
1.32
0.53
4.46
−4.984
2.01±0.81
0.19±0.93
68±19
Na, H
2
O
12,13
WASP-31b
1580
4.6
1.55
0.48
3.40
−5.225
2.15±0.77
1.25±0.77
31±12
Aer, K
4
WASP-12b
2510
11.6
1.73
1.40
1.09
−5.500
3.76±1.59
1.65±1.47
38±34
Aer
3
HAT-P-12b
960
5.6
0.96
0.21
3.21
−5.104
4.14±0.77
1.37±0.79
17±23
Aer, K
HD189733b
1200
21.4
1.14
1.14
2.22
−4.501
5.52±0.50
−0.56±0.50
53.6±9.6
Aer, Na, H
2
O
5,10
WASP-6b
1150
8.7
1.22
0.50
3.36
−4.741
8.49±1.33
Aer, K
11
The listed physical parameters are based on data compiled from our HST and Spitzer results
3,4,10-14
and online databases.
Sources for published spectral results are also listed. Atmospheric features detected of cloud or haze aerosols, sodium,
potassium and water are listed (Aer, Na, K, H
2
O respectively). The equilibrium temperature T
eq
assumes zero albedo and
uniform redistribution. Also listed are the surface gravity, g; radius of the planet, R
p
; planet mass, M
p
; orbital period, P; and
Ca II H&K stellar activity index log(R
'
HK
). R
J
is the radius of Jupiter and M
J
the mass of Jupiter. ΔZ
UB-LM
/H
eq
gives the
difference in pressure scale heights between the optical and mid-infrared transmission spectra, while ΔZ
J-LM
/H
eq
is the
difference between the near- and mid-infrared (see Methods). The atmospheric scale height, H
eq
=kT
eq
/(µg), is estimated using
the planet-specific equilibrium temperature and assuming a H/He atmosphere with a mean molecular weight µ = 2.3 amu. The
H
2
O amplitude is measured using the WFC3 data, taking the average radii from 1.34 to 1.49 µm and subtracting it from the
average value between 1.22 to 1.33 µm, then dividing that value by the theoretical difference as calculated by models
16
assuming clear atmospheres and solar-abundances. !
A continuum from clear to cloudy hot-Jupiter exoplanets without primordial water depletion 4
Figure 1 | HST/Spitzer transmission spectral sequence of hot-Jupiter survey targets. Solid
coloured lines show fitted atmospheric models with prominent spectral features indicated. The spectra
have been offset, ordered by values of ΔZ
UB-LM
(the altitude difference between the blue-optical and
mid-infrared,
Table 1). Horizontal and vertical error bars indicate the wavelength spectral bin and 1σ
measurement uncertainties, respectively. Planets with predominantly clear atmospheres (top) show
prominent alkali and H
2
O absorption, with infrared radii values commensurate or higher than the
.3 .4 .5 .6 .7 .8 .9 1 1.5 2 2.5 3 3.5 4 5
Wavelength (µm)
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
Relative altitude in scale heights z(λ)/H
eq
Na K
H
2
O
WASP-19b
HAT-P-1b
WASP-6b
WASP-12b
HAT-P-12b
HD189733b
WASP-31b
HD209458b
WASP-17b
WASP-39b