The hot Jupiter HD 189733 b


The best-known hot Jupiter

The transiting hot Jupiter HD 189733 b is one of the most favourable targets we have for measuring the structure and composition of an exoplanet’s atmosphere for a number of reasons. To start with, with a radius of around 1.1 Jupiter radii, it’s a large planet, so when it transits its 0.8 solar-radius K host star, it causes an easy-to-detect 2.5% drop in the system’s flux. Similarly, when it passes behind the star during secondary eclipse, the system’s overall brightness drops by around 0.5% in the near-infrared. Furthermore, with a K-band magnitude of 5.5, the host star is bright, providing the large number of photons that are required to perform these high precision measurements.

HD 189733 b is one of only two targets so favourable for planetary atmosphere studies, the other being HD 209458 b, the first transiting planet discovered.


Stellar V magnitude (photometric signal) vs. atmospheric signal. The further to the top right the planets are, the easier it is to measure their atmospheric features with transmission spectroscopy. The dotted lines show curves of constant S/N. The planets to the right of the top curve have an atmosphere detectable with a 90 minute integration (∼ 1 transit) with the Hubble Space Telescope. The planets to the right of the remaining two lines would have detectable atmospheres if the amount of incident photons were increased by a factor 5 and 10 respectively.



Host star brightness vs planet-to-star brightness contrast. The emission spectrum of the day side is easiest to measure during secondary eclipse for planets towards the top left of the diagram. Dotted lines show constant S/N. The planet-to-star ratio is normalised to 1 for HD 209458b.


For these reasons, HD 189733 b has been targeted by numerous high signal-to-noise, multi-wavelength measurements in an effort to unveil the structure and composition of its atmosphere. Most of these measurements have been made at optical and near-infrared wavelengths using the various instruments of the Hubble Space Telescope and the Spitzer Space Telescope.



‘189 is a relatively ordinary member of the hot Jupiter family, slightly off the centre of the main group in parameter space. The clump in the hot-Jupiter distribution is at around half a Jupiter mass, three days period and a size significantly larger than Jupiter (with HD 209458b being the prototype). ‘189 has a slightly closer orbit (2.2-days period), a correspondingly larger mass (1.1 MJup, hot Jupiters follow a mass-radius relation with heavier ones being closer), and an uninflated radius (1.1 RJup). Its host star is an active K0 dwarf, situated in the sky close to the picturesque Dumbell Nebula.


Phase curve

As of 2011 the most spectacular observation of  ‘189 is the complete transit-to-eclipse phase curve obtained with Spitzer at 8 microns and 24 microns (Knutson et al. 2007, 2009), that shows not only the transit and secondary eclipse, but also the small intensity changes as the planet presents different phases to us. The longitudinal temperature map of the planet can be reconstructed from this phase curve.



Spitzer phase curve for HD 189733b at 8 microns, with transit and secondary eclipse. From Knutson et al. 2008.



These observations showed that the hottest point on the planet is shifted eastward of the sub-stellar point, as predicted by atmospheric circulation models. Intriguingly, the light curve also indicates that the coldest point is nearer to the dusk limb than to the dawn limb, which is very difficult to explain with any realistic circulation pattern.



The emission spectrum was measured in broadband filters by Spitzer photometry, and at higher spectral resolution in some wavelength ranges. The evolution of brightness temperature at different wavelengths can be interpreted with a “normal” temperature profile (i.e. temperature decreasing with height), and an efficient day-night energy redistribution. Absorption by water bands is required to fit models to the spectrum. CO is also needed to closely match the NICMOS data in the 1-2 micron range, but if our discussion of NICMOS data analysis in Gibson et al. (2011) is correct, then the actual uncertainties are too large for these data to constrain spectral molecular signatures in that way.


Dayside emisison spectrum of HD 189733 b. Figure adapted from Lee et al. (2011).



Limb and transmission spectrum

Observations of the transmission spectrum of ‘189 have been numerous, covering wavelengths from the UV to the mid-IR. The figure below summarises our present best guess of what the mean atmospheric transmission spectrum at the planetary limb looks like.

Overall, the dominant feature is a steep decrease in opacity from blue to red wavelength, all the way from the UV to the mid IR. The core of the sodium and potassium lines are visible above this “haze” absorption level. Evidence for a haze on a hot-Jupiter was a big surprise, as theory had predicted that these objects would be free from haze or clouds. The slope of the continuum, level of the alkali line and absence of broad pressure-broadened lines wings can all be explained with Rayleigh diffusion high in an in an atmosphere with mean temperature around 1300K near the visible photosphere, then increasing to 2000 K in the upper parts. A reasonable guess is to attribute this diffusion to silicate dust (as observed in brown dwarfs), although there are more exotic possibilities (such as Na2S, sodium sulphite).

In the infrared, detection of several molecular features were claimed, but abundant new data has lead us to conclude that all detections were due to underestimated uncertainties (although since we are part of this debate, you shouldn’t take our word for it). On the contrary, the more data is gathered, the more the initial tentative features have disappeared.

The overall transmission spectrum suggests a pervasive dust layer covering the whole limb of the planet over several atmospheric scale heights (several hundred km), with narrow alkali metal absorption lines above the dust level. This dust scatters the incoming stellar light in the UV and visible, and covers it with a bluish haze. The dust would have to be made of elements abundant enough in the planetary atmosphere to be efficient at scattering the star light high in the atmosphere, which suggests candidates such as silicate grains and sodium sulphite (Na2S). But since the only signature at present is Rayleigh scattering in the transmission spectrum, the actual composition is largely unconstrained.



The overall transmission spectrum of HD 189733b (slide from my talk at the “Extreme Solar Systems II” conference in September)


Detecting the eclipse of ‘189 in visible light – thereby measuring its albedo – would provide a key piece of the puzzle. Dust-free hot Jupiters have very low albedos (a few percent), whereas an atmosphere dominated by a silicate haze would be much more reflective: at least 20% according to our simplified treatment of scattering in Heng et al. (2011) and up to 60% if the scattering is as dominant as in the atmosphere of Venus or Neptune.

Rumor has it that the albedo was measured with the MOST satellite, which monitored ‘189 for weeks over two seasons, but the measurement is made very delicate by the fact that the host star is highly variable.


Last but not least; The activity of HD 189733

That’s right, we left this for the end. HD 189733 is quite active, varying in flux by 1 to 3 percent over its 12-day rotation period under the effect of numerous starspots. This makes all measurements of the atmosphere of ‘189 more difficult, especially at visible wavelengths, where correcting for the effect of starspots requires many complementary observations. On the plus side, we now know quite a bit about the behaviour of star spots on an average K dwarf main-sequence star, using the planetary transits as a “scanner” to read them like bar codes at a cashier.

Light curve of HD189733 over several years measured by G. Henry with the APT photometer, and Gaussian-Process variability model.



Feature Image: NASA/JPL-Caltech


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.