Exploring the Climate of HD 189733 b with Spitzer


An outline of the paper: The Climate of HD189733b From Fourteen Transits and Eclipses Measured by Spitzer by Eric Agol et al.

Of all the instruments that have contributed to our knowledge of exoplanet atmospheres, the Spitzer Space Telescope has truly been one of the standouts. With its exquisite photometric precision over the wavelength range 3.6-8.0 microns, the Infrared Array Camera (IRAC) has measured dozens of secondary eclipses, as well as thermal phase variations for a number of hot Jupiters.

In this 2010 paper, Eric Agol and collaborators presented an analysis of 7 primary transit and 7 secondary eclipse measurements of the old favourite, HD 189733 b, made in the 8.0 micron passband of IRAC. A significant feature of this analysis is the way in which the authors deal with the effect of star spots on their measurements. This is particularly important when it comes to HD 189733 b, because the overall brightness of the system undergoes variations at the ~1% level as spotted regions move into and out of view on the surface of the K1V host star, which is rotating on its axis every ~12-13 days. Disentangling such a signal from planetary effects, such as a ~0.1% secondary eclipse, can prove to be very tricky, as we are never able to directly observe what the spots are doing on the stellar disk.

Faced with the challenge of removing the effect of star spots from their Spitzer measurements, the team made use of regular ground-based photometric monitoring data for HD 189733 that was obtained using an automated 0.8m telescope at Fairborn Observatory in Arizona. By interpolating this ground-based photometry to the times of the Spitzer observations, Agol et al were able to get a reasonable idea of what the spot coverages were, and so account for this in their analysis.

With these spot corrections applied, the results of the transit and eclipse fits were not all that surprising. Although significant variation in the measured primary transit depths was observed, the most plausible explanation for this is that the path traced out by the planet as it transited the stellar disk sometimes passed over brighter regions and at other times passed over duller regions. Unfortunately, more interesting hypotheses, such as big changes in the atmosphere’s nightside flux, seem much less likely. Meanwhile, the eclipse depths were all found to be constant at the precision of the measurements, ruling out things like extremely large-scale weather events in the dayside atmosphere during the times of observation.

Agol et al also had another look at the 8.0 micron IRAC phase curve that was originally presented by Knutson et al (2007). The latter study had strikingly revealed that the hottest part of the dayside atmosphere was offset in longitude from the substellar point by about 16 degrees, presumably due to the advection of thermal energy by super-rotating winds. The motivation for revisiting the phase curve was that the new star spot corrections used by Agol et al were significantly better than those available to Knutson et al back in 2007, due to the accumulation of more ground-based photometry in the intervening period. Indeed, it turned out to be well worth double checking the previous result with the improved star spot corrections, because Agol et al found that the offset is actually closer to 30 degrees – uncorrected star spot contamination had resulted in an underestimate back in 2007 (Figure 1).

Figure 1. Small dots show the measured flux from the planet (normalised to the stellar flux) as a function of the orbital phase measured in the 8 micron passband of IRAC, after the improved corrections for star spots has been applied. The solid line shows a best-fit model for the planetary phase curve. The peak in brightness occurs 3.5 hours before secondary eclipse, indicating that it is offset from the substellar point by ~30 degrees in longitude. The larger dot with an error bar on the left shows the estimate of the nightside brightness obtained by comparing the depths of the primary transits and secondary eclipses. The other larger dot without an errorbar indicates where the peak of the planetary phase function would occur if the hot spot was located at the substellar point (i.e. orbital phase = 0.5 since the orbit is circular).

The larger offset for the hot spot is if anything more consistent with the predictions from state-of-the-art general circulation models (GCMs), such as those of Showman et al (2009) (Figure 2). These models couple 3D atmospheric dynamics with non-grey radiative transfer, and will play an increasingly important role in developing our understanding how planetary atmospheres work, especially as we continue to discover exoplanets with a seemingly endless diversity of properties. The basic consistency between the first generation GCMs of Showman et al and the HD 189733 b Spitzer data presented by Knutson et al and Agol et al offers encouraging signs that the marriage between the observations and theory of exoplanet atmospheres is entering an exciting new regime.

Figure 2. Temperature in degrees Kelvin as a function of longitude and latitude for a global circulation model simulation of HD 189733 b, assuming solar metallicity and a planetary rotation rate synchronous with the orbital period. The region shown corresponds to a pressure of 30mbar, where most of the emitted infrared radiation is expected to originate. At this pressure, the hottest part of the atmosphere is offset from the substellar point (i.e. latitude,longitude=0,0 degrees) by about 60 degrees in longitude. However, when a metallicity of 5x solar is used in the simulations, the resulting hot spot is found to be more like 30 degrees, in good agreement with the 8 micron IRAC observations presented by Knutson et al and updated by Agol et al. Image taken from Showman et al 2009.

Feature image: The SOHO/MDI Team, SOHO is a project of international cooperation between ESA and NASA, found here.


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.