Planetary atmospheres in 2011 (1/2)


An overview of the scientific literature on exoclimes in 2011 (if you notice some glaring omission, give us a shout in the comment section).



2011 will be remembered in the field as the year of the “Kepler Revolution“. The Kepler mission revealed hundreds of firm transiting planet candidates (Borucki et al. (2011), more than enough to measure the abundance of planets down to the size of the Earth. According to Morton & Johnson (2011), the rate of false positives in the current list of candidates is less than one in ten.

The key result is that the abundance of planets around normal stars keep increasing as the mass of planets gets smaller, broadly following a power law, down to sizes of 2 Earth radii, and probably even down to 1 Earth radius (Youdin 2011). Also a large fraction of stars have planetary systems, and multi-planet systems are very common (Lissauer et al. 2011).

Some of the most stunning Kepler detections deserve individual papers, with a dizzying list of seminal results: Kepler-9 (Batalha et al. 2011), Kepler-11 (Lissauer et al. 2011) and Kepler-18 (Cochran et al. 2011) are systems with from three to … six planets transiting the same star! Kepler-10 (Batalha et al. 2011) was the smallest measured radius yet, with 2.2 RE and the first confirmed rocky planet  (if you believe our study the radial-velocity data of CoRoT-7 – (Pont et al. 2011)  showing how stellar variability and instrumental systematics make the claim of a well-measured mass for that object largely wishful thinking). Kepler-19c is the first planet detected purely from transit timing variations (Ballard et al. 2011). Cherry on the cake, Kepler-16b is orbiting a binary stellar system (Doyle et al. 2011), eclipsing both component, ringing a strong cultural bell for some (Star Wars’ Tatooine).

But the Kepler targets are too faint for most follow-up studies, and transiting planets around bright stars remain important. In this context, Winn et al. (2011) found that one of the planets around 55 Cnc is transiting (55 Cnc e, 1.5 RE).

Wright et al. (2011) introduce the Exoplanet Orbit Database at (, a slightly more edited counterpart to the classical Exoplanet Encyclopaedia, specially focussed on accurate RV orbital parameters.

Finally, Sumi et al. (2011) examine the statistical implication of microlensing detections for the abundance of distant and unbound planets, lest we forget about this once-trail-blazing technique that is now being left behind by the new kids in the block.

Oh, and Kepler also detects transiting brown dwarfs (Johnson et al. 2011).


Exoplanet atmosphere Observations

The patient collection of data on the atmospheric emission, transmission, and phase curve of transiting planets continues apace. A key result,  from Cowan & Agol (2011) is that the ensemble of all current measurements seem to indicate two things:

(a) the redistribution of heat from the day side to the night side of hot Jupiters break down around equilibrium temperatures of 2300K. For hot Jupiters hotter than this, the day-night temperature contrast is huge,

(b) some hot Jupiters are dark and others have higher albedos. It’s not clear yet what this depends on.

Now to the most studied individual objects:

GJ 1214 b  – the favourite super-Earth

The super-Earth GJ 1214 b is stealing the show from the “two sisters” this year. The big question is whether the transmission spectrum in the near infrared shows strong features, which would indicate a cloud-free hydrogen atmosphere, or a flat spectrum, due to clouds or, more enticingly, a water-vapour atmosphere. Bean et al. (2011), Désert et al. (2011) and Crossfield et al. (2011) find the second with Spitzer and ground-based data at a large range of infrared wavelengths, but Croll et al. (2011) measure some discrepant points indicative of large features. The balance of evidence leans towards a water or cloudy atmosphere, but “MMN” (more measurements needed).

Two more papers, Berta et al. (2011) and Carter et al. (2011) discuss the light curve of GJ 1214. Netterlman et al. (2011) run simulations for the structure and evolution of the planet. They find that a hydrogen envelope is required to explain the observed radius, unless the water-to-rocks ratio of the interior of the planet is an implausible 6:1. This is potentially in conflict with the water-vapour atmosphere required by the transit spectrum. Interesting…

WASP-12 b  – the favourite very hot Jupiter

The extreme “very-hot Jupiter” WASP-12 b is another object that has been gathering attention – although some of it sound like last-year news already. Campo et al. (2011) and Husnoo et al. (2011) show that the very surprising eccentricity measured by Hebb et al. (2010) and Lopez-Morales et al. (2010) is spurious, and that the orbit is circular, as expected with an orbit so close to the star.

Croll et al. (2011b) measure the occultations in J, H and K from the ground, and Madhusudhan et al. (2011), by fitting zillions of possible spectral models to the Spitzer occultation data, find that the C/O ratio in the atmosphere should be above unity, with a host of interesting consequences, detailed in another paper (Madhusudhan et al. 2011b). The issue though is that this conclusion is derived in a Bayesian way, and rests almost entirely on a single low point in the emission spectrum, thus it is heavily dependent on the completeness of the space of model spectrum sampled, as well as the accuracy of the error bars (at conferences, Cowan and collaborators already show new Spitzer measurements that throw doubts on the picture).

 HR 8799 – the favourite directly imaged system

Currie et al. (2011) show some marvellous spectroscopic data at 1-5 micron from Subaru for the four planets orbiting HR 8799. They find that the planets do not behave like brown dwarfs of the same temperature in a colour-magnitude diagram. Cooler that the L-T transition, brown dwarfs become much bluer, which is interpreted as due to the disappearance of dust clouds. Planets, however, seem to keep getting redder. Madhusudhan et al. (2011c) and Barman et al. (2011) calculate spectrum models and show that silicate or iron clouds can do this. I found this very exciting, but we were told since then by Mark Marley that this is just what is expected from the gravity effect on clouds: planets should keep clouds at lower temperatures because of the lower gravity, so no surprise there, but another flying-colours test for extra-solar clouds.

Barman et al. (2011b) also model the spectrum of 2M1207b, which they find to be cloudy as well.

Meanwhile, there are rumors that the planet around Formalhaut has not been confirmed by new images (to be continued in 2012).

Of course, the poster kids of the transiting planets didn’t go unobserved in 2011:

‘189 – the favourite hot Jupiter

Sing et al. (2011) collected a STIS transmission spectrum from 3000 to 6000 Å for ‘189, the counterpart of the classic study on ‘209. This is long overdue, but STIS was broken and only recently repaired by “Servicing Mission 4″ on the HST (the good old days when we had Space Shuttles, remember?). The STIS data connects beautifully with the 6000 Å – 1 μm ACS measurements, showing a featureless bluewards slope compatible with Rayleigh scattering by a haze of silicate grains.

The re-analysis of NICMOS data in Gibson et al. (2011) suggests that the striking results of the group detecting all those molecular features in transmission spectra (see earlier years) rest on overoptimistic assumptions on instrumental systematics. If you relax the assumption that systematics are entirely described by a few linear parameters, then all observed features can be attributed to instrumental systematics. Désert et al. (2011) show that this may also be the case for Spitzer data. The fact that none of the claimed molecular feature has been reproduced in two different data sets or by different group is disturbing. Mandell et al. (2011) also attempt to verify the surprising “fluorescent methane” detection of Swain et al. (2010) and find that it can be excluded by their data in the L band, and was probably due to telluric line contamination. Thus, no detection of features in transmission spectroscopy in the infrared has so far stood the test of independent confirmation.

 ‘209 – the other favourite hot Jupiter

Vidal-Madjar et al. (2011) measure some details of the shape of the Sodium doublet in the transmission spectrum of ‘209 with STIS. The observed shape is best explained by a temperature increase in the very high part of the atmosphere (a “thermosphere”, not to be mixed up with the temperature inversion of the “stratosphere”, the core of the sodium doublet probes much higher regions in the atmosphere than the Spitzer infrared day-side photometry, ~10e-5 bars as opposed to ~0.5 bars).

Other planets

The phase curve of Kepler-7b is clearly detected in the Kepler lightcurve (Demory et al. 2011, Kipping & Bakos 2011), showing that the visible albedo is high (~0.4). This is the first solid measurement of a high albedo for a hot Jupiter (cloud-free models predict very low albedos in the visible because of alkali-metal absorption).

At the other extremes, Kipping & Spiegel (2011) find that the low amplitude of the Kepler phase curve for TrES-2 implies an incredibly low albedo (<1%) for the planet.

Knutson et al. (2011) and Beaulieu et al. (2011) present Spitzer data for the hot Neptune GJ 436 b. Instrumental systematics and star spots make any interpretation weak, and even the usually assertive second group include a question mark in their title. The prolific Madhusudhan et al. (2011d) model the occultation data in terms of molecular abundances.

Spitzer emission in the near infrared are collected for Kepler-5, Kepler-6, CoRoT-1, CoRoT-2, WASP-4, WASP-17 and WASP-18 (sorry if I missed some, difficult to keep up). Altogether the Spitzer occultation data, presented in meeting by Joe Harrington, suggests a transition between hot Jupiters and “even hotter Jupiters”, the latter with very strong day-night contrasts and a day-side temperature close to the sub-stellar equilibrium temperature.

Sing et al. (2011) present the first detection of the 7670 Å potassium line in an exoplanet atmosphere (this line is the second strongest expected from the models, after the Sodium doublet at 5890 Å), with the 10.4-m GTC telescope using a tuneable filter. Wood et al. (2011) obtain a 4-sigma detection of the Sodium doublet for WASP-17 with a multi-fiber spectrograph on the VLT.

Not atmospheres, but good to know

Orbital Tilt (problem solved?)

The game-changing discovery that most hot Jupiters have tilted orbits generated a number of theoretical studies. Naoz et al. (2011) and Wu & Lithwick (2011) study the possibility of bringing Jupiters close to their star through secular orbital interactions with other planets – and find that it can be done, while Lai et al. (2011) show that it is possible for the star and disc to be misaligned after formation, but Watson et al. (2011) study the observed distribution of inclinations for debris discs and  find that such misaligned systems are rare or absent. Morton & Johnson (2011) find that the observed distribution of orbital angles strongly favours planet-planet scattering as the single dominant mechanism to produce hot Jupiters, rather than Kozai or disc migration (provided one accepts the Winn et al. hypothesis of tidal re-alignment for G dwarfs). Problem solved*.

*I may be pushing it a bit here, and as usual there may be dissenting opinions far into the future, but this is rare enough in our business to be stressed. Here we have a mystery, which emerged 16 years ago with the discovery of 51 Peg, for which we now have a solid, coherent and convincing explanations. Close-in planets form by architectural re-arrangement of closely packed planetary systems, with subsequent tidal circularization of the orbit of the close planet by tidal interaction with the star. The 3-day pile up and its mass dependence, the mass distribution of hot Jupiters, and their tilted orbits around F dwarfs are all neatly explained.

Radius inflation problem (continued)

Laughlin et al. (2011) calculate the “normal” radius for coreless gas giants as a function of their mass and temperature, and show that the excess grows with temperature in an orderly manner, and decreases with metallicity – a possible sign of heavies cores from more metal-rich protoplanetary discs (see the earlier Guillot et al . 2006 with the same conclusion). Batygin et. al. (2011) calculate the evolution of hot Jupiters with Ohmic dissipation, and find that the observed trend can be reproduced – although looking into the details it seems to me that Ohmic dissipation should be producing a more abrupt dependence with temperature than is observed.

Ibgui et al. (2011) calculate the evolution of radius with tidal heating included, and find that it can explain some cases, and not others.  Now that the trend of radius excess with temperature is so clear, it seems that tidal heating requires too much fine-tuning to work as a general explanation.

Kepler-12b (Fortney et al. 2011) is an interesting case in this context, because it seems to be substantially more inflated than even others hot Jupiters at the same temperature, maybe that’s a clue to something – although it is also one of the lightest inflated hot Jupiters, so all we may be seeing is that it is easier to inflate lighter planets.

continue on to part 2/2


About Author

I am a professor of planetary science at the University of Exeter. My specialty is the study of exoplanets, in particular the observation and modelling of exoplanet atmospheres. I have done my PhD a the University of Geneva and worked in Chile, France and Switzerland.