Planetary atmospheres in 2009


An overview of the scientific literature on exoclimes in 2009 (if you notice some glaring omission, give us a shout in the comment section).




The HAT-Net and WASP surveys churned out a large amount of new transiting planets this year, including some interesting extremes.

The heaviest is WASP-14 b (7 Jupiter masses, Joshi et al. 2009), the hottest WASP-12 b (Teq=2516K

with P=1.09d: Hebb et al. 2009). Also of interest is HAT-P-12 b (Hartman et al. 2009), the lowest- mass gas giant found to date (the composition is inferred from theoretical models by Baraffe et al. 2008 and Fortney et al. 2007). HAT-P-13 (Bakos et al. 2009), seems to have another, non-transiting companion detected by a long-term trend in the radial-velocity data. And then there is WASP-18b (Hellier et al. 2009), with a period of only 0.94 days, suggesting that tidal interactions will soon make short work of this planet as it spirals into the stellar surface. Whether or not this is the case may become clear from observations over the next decade (Hellier et al. 2009).

The HAT survey is extending into the southern hemisphere (Bakos et al. 2009).

The first forays are made into the “super-Earth” domain this year with the transit detections of GJ 1214 b (Charbonneau et al. 2009) and CoRoT-7 b (Léger et al. 2009). GJ 1214 b was found around an M- dwarf host by the MEarth project. With R=2.7 REarth and M=6.6 MEarth, it should be mainly composed of rocks and ices but with an hydrogen envelope. CoRoT-7 b has a radius around 1.7 that of Earth. The host star is quite active, and the HARPS radial-velocity signal is dominated by the effect if stellar activity. Queloz et al. (2009) think they see not one but two planets in the residuals, but this requires a considerable leap of faith.

Twelve other “bog-standard” hot/warm Jupiters are also discovered this year, see for their properties.

Directly Imaged Planets

The imaging of two long-period planets around the bright star HR 8799 by Marois et al. (2008) is a great milestone in exoplanet studies, and apparently the evidence of a planetary companion in the system has been there since 1998! Lafreniere et al. (2009) spot HR 8799 b in archive HST data from 1998. Metchev et al. (2009) and Fukagawa et al. (2009) use 2007 archive Keck data and 2002 archive Subaru data to confirm the detection.

Thalman et al. (2009) reports the discovery of a sub-stellar object (10-40 MJ) around GJ 758 using HiCIAO on Subaru, with the possible sighting of a second, longer period, companion.


The “two sisters” HD 209458 and HD 189733 keep gathering the most attention from observers.


Knutson et al. (2009) present a Spitzer lightcurve at 24 microns, showing both primary and secondary eclipses as well as a clear phase curve along the orbit, with a hottest point shifted with respect to the substellar point by 20-30 degrees. As in Knutson et al. 2007 (Spitzer @ 8 microns) this indicates an efficient thermal energy transport across the surface by eastwards jets. They also find that the features are unchanged between 8 and 24 microns, indicating that the “temperature map” found at 8 microns is a robust result.

Spitzer observations of the primary transit of HD 189733 b at 4.5 and 8 micron, and updated reductions of the observations at 3.6 and 5.8 microns (Désert et al. 2009) seem to invalidate the water detection of Tinetti et al. (2007). The Spitzer data is compatible with an extension of the featureless Rayleigh signature seen below 1 microns with ACS/HST (Pont et al. 2008), except for an enhancement at 4.5 microns, that Désert et al. (2009) tentatively attribute to CO. The transit depth was measured in two NICMOS channels, 1.66 and 1.87 microns, by Sing et al. (2009), in an attempt to confirm the claimed water detection of Swain et al. (2008). The result though, is not compatible with the presence of water absorption, but with an extension of the Rayleigh scattering feature to longer wavelengths.

Boisse et al. (2009) observed HD 189733 regularly with the SOPHIE spectrograph during on month, simultaneously with monitoring by the MOST satellite. They study the relation between the photometry and spectroscopic signal in relation to different activity indicators (HeI, Hα, CaIIH & K emission features, and the RV line bisector).


A HST/NICMOS dayside spectrum of HD 209458 b is measured between 1.5 and 2.5 microns by Swain et al. 2009. Taken at face value, the radius variations that they observe over this wavelength range are very constraining for the abundance of methane, carbon dioxide and water vapour. But Pont et al. (2009), using NICMOS to measure the transit of GJ 436 b, see the same kind of variations, that they attribute to detector systematics. It all rests on the trust that can be placed in the decorrelation of external factors like the detector temperature and telescope pointing.

Langland-Shula et al. (2009) use HIRES on Keck to attempt a measurement the sodium features in the atmosphere of HD 209458 b during transit (as first reported in Charbonneau et al. 2002). They do not find compatible results and evoke possible variations in the atmosphere. The question, of course, is the atmosphere of which planet, ’209 or Earth?

Others transiting planets

XO-5 bHebrard et al., Winn et al.

TrES-2 b  TrES-3 b De Mooij & SnellenTrES-4 b Knutson et al.WASP-4 b

Object Spin-orbit angle from RM effect Secondary eclipse
HD17156 b
HD149026 b Knutson et al.
HD80606 b Pont et al., Winn et al. Laughlin et al.
GJ436 b
XO-1 b
XO-2 b Machalek et al.
XO-3 b Hebrard et al., Winn et al.
XO-5 b
TrES-1 b
TrES-2 b
TrES-3 b De Mooij & Snellen
TrES-4 b Knutson et al.
WASP-4 b
WASP-5 b
WASP-10 b
WASP-11 b
WASP-12 b
WASP-14 b Johnson et al.
WASP-18 b
CoRoT-1 b Snellen et al., Alonso et al., Gillon et al., Rogers et al.
CoRoT-2 b Alonso et al.
CoRoT-3 b Traiud et al.
HAT-P-7 b Narita et al., Winn et al.
HAT-P-11 b
OGLE-TR- 10 b
OGLE2- TR-L9 b
OGLE2- TR-56 b Sing & Lopez-Morales

The extremely eccentric HD 80606 b (e = 0.93) is observed during its secondary eclipse by Laughlin et al. (2009) with Spitzer at 8 microns. They observe a change in surface temperature from 800K to 1500K over only a six-hour period, as expected from the “flash heating” at periastron.

The secondary eclipse of TrES-4 b is observed at 3.6, 4.5, 5.8 and 8.0 microns by Knuston et al. (2009a) using Spitzer. As for HD 209458 b, the data is best fit with an atmospheric thermal inversion near the infrared photosphere.

Machalek et al. (2009) using Spitzer who measure the planet-to-star flux ratios (0.1-0.01%) of XO-2 b at 4 different wavelengths (3.6, 4.5, 5.8 and 8 microns) also consistent with an upper atmosphere temperature inversion.

Knutson et al. (2009) use Spitzer at 8 microns to measure the secondary eclipse of HD 149026 b. They find half the depth reported by Harrrington et al. (2007), corresponding to a dayside brightness temperature of 1440K as opposed to Harrington et al.’s 2300K.

A transmission spectrum of GJ 436 is taken with HST/NICMOS by Pont et al. (2009) from 1.1-1.9 microns. They find no evidence for any transit timing variations due to a further companion’s perturbation, and measure a fairly flat transmission spectrum with no clear water feature at 1.4 microns.

Sing & Lopez-Morales (2009) detect the secondary eclipse of OGLE-TR-56 in the z band with the VLT and Magellan telescopes, making it the first confident ground-based detection of an exoplanet thermal emission. This is a great step forward in characterising atmospheres as means that future observations won’t have to rely on cryogen-free Spitzer observations. They find an extremely high brightness temperature of 2718K, implying a very low albedo.

Very shortly after, De Mooij et al. (2009) detect the secondary eclipse of TrES-3 b from ground-based observations at 2.2 microns.

The secondary eclipse of CoRoT-1b is measured in the visual by Snellen et al. (2009) in CoRoT data, as well as the out-of-transit corresponding to the phases of the planet. The amplitude of the signal seems to indicate a completely black night side and a geometric albedo lower than 0.2. Rogers et al. (2009) also measure a secondary eclipse using NICFPS on the ARC 3.5m and combine their results with those mentioned above to infer a planetary temperature of 2460K and a negligible Bond albedo.

A very shallow secondary eclipse of CoRoT-2b is reported by Alonso et al. (2009) from CoRoT observations, implying a surface brightness temperature of 1910K, and a geometric albedo lower than 0.12.

Seager & Deming (2009) attempt to observe the intensity changes in the light curve of the non- transiting planet host GJ 876, but are thwarted by the photometric variation of its active M-type host star.

Lucas et al. (2009) unsuccessfully attempt to detect reflected light from the planets around 55 Cnc and Tau Boo using WHT.

Early initial results from the Kepler space telescope start coming through, showing the transit and eclipse of HAT-P-7 b (Borucki et al. 2009). The excellent quality of the data is evident, and the changes in the baseline flux due to day and night flux variations clearly seen. Kepler has indeed made an excellent first impression.

An interesting experiment was performed by Palle et al. (2009): they measured the transmission spectra of Earth during a lunar eclipse by looking at light reflected from the Moon. They observe signatures of ozone, molecular oxygen, water, methane and carbon dioxide signatures in the optical and near infrared.

In a similar vein, Cowan et al. (2009) use Deep Impact to observe the light curve of Earth as if it were an exoplanet, and find it possible to reconstruct the gross distribution of landmasses and oceans from the two-colour lightcurve. Somewhat surprisingly, it turns out that clouds can be factored out, but they caution that this only works if the rotation period of the planet is known a priori. Oakley & Cash (2009) similarly simulate Earth as an exoplanet being observed by the proposed New Worlds Observer and conclude it “may be” able to discern the presence of surface water.

The neighbours

(not on planetary atmospheres directly but still good to know)

NStED (NASA Star and Exoplanet Database; launch is reported in Ramirez et al. (2009) containing information on exoplanets and their host stars.

Safronov number

Southworth (2009) looks at the total planet population to date, finding correlations between periods, masses and surface gravities, and finds a much weaker separation by Safronov number (the ratio of escape velocity to orbital shear) than in Hansen and Barman (2007) showing that the original correlation was likely an artefact of a small sample size.

Spin-orbit alignment

XO-3 b is the first confirm case of a planetary system with a highly tilted orbit (high angle between the stellar spin and orbital plane). The misalignment is tentatively detected by Hébrard et al. (2009) and confirmed by Winn et al. (2009). XO-3 b is a very heavy planet, possibly a light brown dwarf (12 MJ), so that it was not clear at that stage what the implications were for more regular hot Jupiters. But in short order came the detection of spin-orbit misalignment in HD 80606 (Pont et al. 2009, Winn et al. 2009), WASP-14 (Johnson et al. 2009), and HAT-P-7 (Winn et al. 2009b, Narita et al. 2009). While the first system can also be construed as an exception because of its extreme eccentricity (0.93), there is nothing obviously special about the other two. Note that the orbit of HAT-P-7 b is not merely tilted, but retrograde.

Their results are difficult to explain with disc migration, and seem to herald a return to favour of scenarios involving dynamical interactions between several planets or with companion stars (planet- planet scattering, Kozai mechanism).

The orbit of HD 17156 b, on the other hand, is found to be aligned after all (Narita et al. 2009), contrarily to the initial claim (Narita et al. 2008) for which the authors blame low-precision data and understated uncertainties.


Winds and circulation

Two studies of atmospheric circulation of hot Jupiters were published in the summer of 2009. Menou & Rauscher (2009) and Showman et al. (2009). They confront the prediction to the observations of ‘209 and ‘189. Both model predict the observed eastward shift of the hottest point observed in ‘189, due to equatorial jets, although some observed cannot be reproduced in detail, such as the emission spectrum of ‘209.


Fortney et al. (2008) and Burrows et al. (2008) proposed that the temperature inversion observed in hot Jupiter atmospheres was caused by TiO and VO vapour. Spiegel et al. (2009) find however that these compounds will not persist in the upper atmosphere.

Madhusudhan & Seager 2009 construct a pressure-temperature and elemental abundance model and test this against observational data from HD 189733 b and HD 209458 b. Mousis et al. 2009 try to model the oxygen and carbon abundances observed in HD 189733 b by Tinetti et al. 2007, Swain et al. 2008 and Charbonneau et al. 2008. They find their formation model abundances to be much higher than those observed and attribute this to gravitational settling of heavy elements.

Helling (2009) discusses the different cloud formation models for brown dwarfs and exoplanets, and draws to the reader’s attention the difficulty in modelling large-scale turbulent dust formation, and the effect of coupling of the atmosphere with the planetary magnetic field.

Zahnle et al. (2009) creates a model for photochemical hazes from sulphur species in hot Jupiter atmospheres. These species should absorb strongly at 300-460nm. A study of the primarily CO2 atmosphere of Super-Earths and its escape for planets around M stars is made by Tian 2009, finding that the atmospheres are mostly stable against escape.

The HITRAN spectroscopic line database receives its first major public update since 2004, now containing various data for 42 different molecules that can be used for modelling theoretical exoplanet atmospheres (detailed in Rothman et al. 2009).


Koskinen et al. (2009) model the atmosphere of the transiting planet HD 17156 b, and find that the atmosphere is unlikely to experience escape during its eccentric orbit (e = 0.67).

The atmospheric mass loss of planets such as HD 209458 b is modelled by Murray-Clay et al. 2009. They conclude that UV radiation can cause strong winds but not drive mass loss, and claim that the observed Lymna-alpha Doppler shift is due to charge exchange between winds rather than atmospheric mass loss.

Stone & Proga (2009) find that hot Jupiters can experience mass loss from 5-25% (the extreme is WASP-12 b) in their lifetime, while others a negligible <2%. This mass loss can be from radiation pressure, charge exchange and solar wind interactions. Johansson et al. (2009) simulate the magnetosphere of hot Jupiters.

Cecchi-Pestellini et al. (2009) look at the effect of solar UV and X-ray radiation on a planetary atmospheres composed of hydrogen.

Super Earths

Can we study super-Earths in transmission spectroscopy? Miller-Ricci et al. (2009) find that although the signal is weak, due to the relatively small atmospheric scale height compared to hot Jupiters, is should be possible to discriminate between hydrogen-dominated atmospheres (with a large scale height) and atmospheres composed of heavier elements, such as water or CO2.

The silicate atmospheres of very hot super-Earths are modelled by Schaefer, Fegley et al. (2009).

Radius inflation

Gu & Ogilvie 2009 show that stellar irradiation can excite internal waves that can in some cases propagate through the planetary interior.

Tidal effects caused by atmospheric insolation (“thermal tides”) on hot Jupiters are modelled by Arras & Socrates 2009. They find that the radius increases strongly as it nears the host star. Goodman 2009 argues that their study has a fatal flaw.

Ibgui & Burrows (2009) propose that tidal heating could have caused the inflation of hot Jupiters, and reproduce the observed radius of HD 209548 b with a tailor-made tidal evolution history. The contraction and expansion of hot Jupiters as a result of tidal evolution is studied by Miller et al. (2009).

Things for the Future

Arnold (2009) discusses how the sharp reflectance at 700nm due to vegetation on Earth could be resolved in an exoplanet with future space based “hypertelescopes”. A study of theoretical transmission spectra of Earth by Kaltenegger & Traub (2009) shows the spectral features one should be looking for when hunting habitable worlds. Their theoretical spectra show that O2 and O3 are detectable in the NIR, CO2, H2O and CH4 in the mid IR and that the lower atmosphere is hard to probe in transmission due to Rayleigh scattering and aerosol absorption.

A visible to IR spectral library for 18 of the largest solar system bodies is provided by Lundock et al. (2009), as a reference for exoplanet spectra.


About Author

I am a PhD student in the Exeter astrophysics department. I studied as an undergraduate in Durham university. I specialise in statistical properties of radial velocity datasets and theoretical atmospheric line features. (More)