Full orbit Warm Spitzer observations of WASP-12b


An outline of the paper: Thermal Phase Variations of WASP-12b: Defying Predictions by Nicolas Cowan et al.

A couple of months ago, Frédéric gave an update on recent observations of the extreme hot Jupiter WASP-12b. Since then, Cowan et al (2012) have published additional Warm Spitzer observations that were obtained in the 3.6 and 4.5 micron passbands of the IRAC camera. Each set of observations consists of around 31 hours of continuous monitoring, with infrared snapshots taken every 2 seconds. Since the planetary orbital period is roughly 26 hours, this means that both the 3.6 and 4.5 micron lightcurves encompass two secondary eclipses and a primary transit (top panel of Figure 1).

Figure 1. The 4.5 micron observations, taken from the Cowan et al paper. (Top panel) The light curve after instrumental systematics have been removed. Clearly visible is a primary transit and two shallower secondary eclipses. Also evident are the out-of-transit flux modulations caused by the planetary phase variations. (Middle panel) Best-fit ellipsoidal variations, which, if real, are caused by the distortion of the planet by the star. (Bottom panel) Residuals after the best-fit model has been subtracted from the data.

Measurements like these have the potential to reveal a lot of information about the planet, and in particular, the properties of its atmosphere. For starters, the depth of the primary transit at different wavelengths can reveal how the opacity of the day-night terminator varies with wavelength, giving us clues about what kinds of absorbers might be present. The depths of the secondary eclipse reveal how bright the planet’s dayside surface is at different wavelengths, and this can help us put constraints on the vertical pressure-temperature profile of the atmosphere, as well as how its composition varies with depth (see for example, the recent analysis of Lee et al for HD189733b). Furthermore, if our measurements are sensitive enough to detect the secondary eclipse signal, there’s also a good chance that we could detect the changes in the system’s overall flux caused by the phase variations of the planet as it moves between inferior and superior conjunction (top panel of Figure 1). The amplitude of these phase variations gives us an idea of the day-night temperature contrast, revealing information about the atmospheric circulation. And finally, because WASP-12b is so puffed up and close to its host star, the uppermost layers of its atmosphere could well be filling its Roche lobe, so that it probably has more of an ellipsoidal, rather than spherical, shape. As a result, at the times of inferior and superior conjunction, when we see the ellipsoid “front on”, the planet will have a smaller cross-sectional area than at other times when we see the ellipsoid “side on”. This would cause an additional modulation of the observed flux from the planet, which could in principle be disentangled from the phase variations as it would repeat itself twice as often (middle panel of Figure 1).

Turning back to the observations that Cowan et al have made with Spitzer, however, it seems that rather than making the picture clearer, they have perhaps raised more questions than they have answered! First of all, the 3.6 micron transit depth is measured to be significantly deeper than the 4.5 micron transit depth (see the red points in Figure 2). This is difficult to reconcile with current models that assume either a solar composition for the planetary atmosphere or those that allow for a non-solar C/O ratio, as has recently been proposed by Madhusudhan et al (2011) for WASP-12b. If the measurements are accurate then, it might suggest that there is some unknown absorber in the atmosphere that is more opaque at 3.6 microns than it is at 4.5 microns.

Figure 2. Transmission spectrum, taken from the Cowan et al paper. Red points indicate the best-fit 3.6 and 4.5 micron transit depths. Blue point indicates the adjusted best-fit 4.5 micron transit depth if it is assumed that the inferred ellipsoidal variations are actually caused by instrumental systematics rather than a bona fide astrophysical signal. Black points to the left indicate best-fit transit depths obtained in previous studies (namely, Hebb et al 2009, Maciejewski et al 2011 and Chan et al 2011). The solid black line indicates a solar composition atmospheric model with night-like temperature-pressure profile, and dotted line indicates a solar composition model with a day-like temperature pressure profile.

But it would be premature to fully trust these results just yet. For one thing, Cowan et al discuss the very real possibility that there could be instrumental systematics remaining in the data, despite their best efforts to remove them, and that these could potentially muddle up the inferred transit depths. They note, for instance, that their best-fit ellipsoidal variations (caused by the distortion of the planet by the star) are in line with predictions for the 3.6 micron light curve, but much larger than expected at 4.5 microns. That would be particularly puzzling, because it would seem to imply that the highest layers of the planetary atmosphere are more opaque at 4.5 microns than at 3.6 microns, which is precisely the opposite conclusion to be drawn from the measured transit depths if they are taken at face value. On the other hand, if it is assumed that the ellipsoidal variations detected at 4.5 microns are somehow caused by uncorrected systematic effects, and they are removed from the data, then refitting the light curves gives transit depths that are consistent with solar composition models (see the blue point in Figure 2).

The measured amplitudes for the planetary phase variations are also not so clear cut. This is because Cowan et al tried two different methods for removing systematics from the data, and get significantly different values for the fitted planetary phase signals depending on which one they use. However, for reasons that I won’t go into, it seems that one of these methods might be suppressing the planetary phase signal in addition to removing the instrumental systematics, whereas the other one is more likely to preserve it. If we accept this scenario, then the results obtained using the latter of the two would seem to suggest that WASP-12b has a high day-night temperature contrast, indicating a low recirculation efficiency for the atmosphere.

Finally, the secondary eclipse depths don’t shed much additional light on the situation. This is simply because the error bars are too large to make a clear distinction between different model predictions, such as an atmosphere with the enhanced C/O ratio proposed by Madhusudhan et al as opposed to one with a solar composition. And besides, it’s not yet obvious if these models are sufficient for describing the planetary atmosphere, or if there are some key ingredients missing, given the problems noted above with matching them to the measured primary transit depths. Whatever the case, there still remains a lot to be learned about the atmosphere of WASP-12b!


About Author

I'm a PhD student at the University of Oxford. My work focuses on transiting exoplanets and, in particular, what we can learn about the atmospheres of these systems. A large part of this involves getting a better handle on the various instrumental systematics that contaminate the small signals we're trying to measure, and devising methods to remove them from the data. I'm also investigating ways of correcting for the effect of star spots on planetary transmission and emission spectroscopy measurements. My supervisor is Suzanne Aigrain.