Extrasolar planets

Remote climes

A distant planet traversing its orbit shows variations in its infrared brightness, providing the first map of its climate. These variations paint a picture of a dynamic world, with efficient redistribution of stellar heat.

Here's a startling fact for those not up with the latest planetary news: we now know of more than 20 times as many planets outside our Solar System as in it. Most of these extrasolar planets are gas giants like Jupiter; almost all were discovered indirectly by the slight wobble they induce in the orbit of their bright primary star. To transform the study of extrasolar planets into a true physical science, we require the direct detection of the planets' light — that is, remote sensing. On page 183 of this issue, Knutson et al.1 sketch the first directly obtained infrared map of the surface of a giant planet outside the Solar System. They also use the data, obtained using NASA's Spitzer Space Telescope2, to explore the atmosphere of this planet, known as HD 189733b.

It had been thought that such remote sensing of an extrasolar planet would necessarily entail the separation of images of the planet and star, and their independent spectroscopic characterization. For Jupiter-like planets in Jupiter-like orbits, seeing the dim planet from under its bright stellar lamp-post demands extremely high-contrast imaging3. One favoured idea, yet to be tested in practice, involves the artificial occultation of a planet's star by a 'coronagraphic disk' built into the imaging telescope. Knutson and colleagues take a different approach, one that works for a transiting planet such as HD 189733b.

A transiting planet is one for which Earth, the planet itself and its star all coincidentally lie in the same plane. So far, 17 transiting giant planets have been found. All are extremely close to their central star — just a few per cent of the Earth–Sun distance. Jupiter, by contrast, the archetypal giant planet in our Solar System, is five times farther away from the Sun than is Earth. What sets apart a transiting giant planet is that, when it passes in front of its star's disk (as the star appears to us), the flux of stellar light at Earth is diminished by a fraction that depends on the planet's projected area. The magnitude of this diminution (generally about one part in 100) translates into an estimate of the planet's radius. Using low-tech, ground-based telescopes, working at optical-light wavelengths, it is easy to make photometric observations that can discern an effect of this magnitude. The radii of all 17 known transiting extrasolar giant planets have thus already been measured4.

But this is not the whole story. In the case of transiting planets, things gets interesting half an orbital period after transit, when the planet goes behind its star and is itself eclipsed (a so-called secondary eclipse). Because the planets are so close to their stars (generally around 10 solar radii away), they are severely irradiated and are incandescent5. Unlike the stellar flux, which is mainly at optical wavelengths, most of the planetary flux is in the near- to mid-infrared range. At these longer wavelengths (from around 3 to 30 µm) the planet-to-star contrast ratio becomes much more favourable, reaching values of a few thousandths that are within reach of Spitzer's capabilities.

When the planet enters secondary eclipse, its contribution to the combined star–planet infrared signal suddenly disappears. From this abrupt change, the emissions of the planet's dayside — the side facing the star — can be determined. Effectively, the star itself is being used as a coronagraphic occulting disk. This technique has now been used to constrain the dayside atmospheric properties of a handful of transiting extrasolar giant planets6,7,8,9,10. Even though some of the inferred compositions are dubious, this collection of results represents a milestone in research into extrasolar planets.

As Knutson et al.1 show, there is more. The planet is locked gravitationally into its orbit: it always presents the same side to the star, and only this side is heated. Thus, the detected infrared signal should vary periodically as the contributions of the day and the night sides of the planet to the summed light vary in the course of the planet's completion of an orbit (Fig. 1). Furthermore, the degree to which the night side can be seen at infrared wavelengths at all is a measure of how efficiently the stellar heat is transported by the circulating jet streams and zonal winds in the planet's surface layers. In other words, measurements of the day–night contrast provide constraints on the climate of the extrasolar planet.

Figure 1: Planet in transit.

Knutson and colleagues study1 the transiting giant planet HD 189733b at infrared wavelengths, where the contrast between the emissions of its parent star, which are mainly at optical wavelengths, and its own emissions — starlight absorbed and reradiated as infrared light — is greatest. The side of the planet facing the star has the fiercest emissions, and what the authors observe are 'phases' of the planet, akin to the phases of the Moon: the planet is at its brightest as it moves into secondary eclipse behind the star, and at its faintest as it transits across the star's disk, with its irradiated dayside face concealed. But the very fact that the planet is visible at all in transit speaks for the efficiency of heat transfer processes in the planet's surface layer, and tells us something about the planet's climate.

Using Spitzer's Infrared Array Camera at a wavelength of 8 µm, Knutson and colleagues have measured the emissions of the planet HD 189733b from just before transit to just after secondary eclipse. The day-to-night variation they see is small: the day–night brightness difference is only about a third of the full dayside brightness, implying that the redistribution of stellar heat to the night side is quite efficient. This value is, incidentally, much higher than the redistribution efficiency inferred last year11, from a much sparser data set and at a higher infrared wavelength, for the non-transiting giant planet υ Andromedae b. The reasons for this stark difference are not yet clear.

Knutson et al.1 also partition HD 189733b into latitudinal strips and, using a crude model and their measured light curve, fit for the brightness as a function of longitude. The upshot is, in a sense, an image of the planet itself at a wavelength of 8 µm. From the timing of the phases of the light curve in the course of their measurements, the authors find a slight eastward shift in the hottest spot and a slight westward shift in the coolest spot, curiously putting both spots on the same hemisphere.

These startling data will clearly exercise theorists for some time to come. The euphoria is tempered only by the realization that the HD 189733 system boasts the most favourable planet–star contrast ratio of the known transiting extrasolar planets, and is one of the closest systems to Earth. We therefore might not obtain better infrared light curves for a while, perhaps not until NASA's James Webb Space Telescope comes online in five to seven years' time. Spitzer's mission will soon end, as its cryogenic sources are exhausted, and its design and programme lives have been reached. The emphasis must now be on getting as much data, on as many close-in giant planets and in as many wavebands, as we can. The prize is mankind's first direct glimpse of the exotic worlds beyond the narrow confines of our isolated Solar System.


  1. 1

    Knutson, H. et al. Nature 447, 183–186 (2007).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Werner, M. W. & Fanson, J. L. Proc. SPIE 2475, 418–427 (1995).

    ADS  Article  Google Scholar 

  3. 3

    Burrows, A. Nature 433, 261–268 (2005).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Charbonneau, D., Brown, T. M., Burrows, A. & Laughlin, G. in Protostars and Planets V (eds Reipurth, B. & Jewitt, D.) 701–716 (Univ. Arizona Press, Tucson, 2007).

    Google Scholar 

  5. 5

    Burrows, A., Sudarsky, D. & Hubeny, I. Astrophys. J. Lett. 650, 1140–1149 (2006).

    ADS  CAS  Article  Google Scholar 

  6. 6

    Deming, D., Seager, S., Richardson, L. J. & Harrington, J. Nature 434, 740–743 (2005).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Deming, D., Harrington, J., Seager, S. & Richardson, L. R. Astrophys. J. Lett. 644, 560–564 (2006).

    ADS  CAS  Article  Google Scholar 

  8. 8

    Charbonneau, D. et al. 626, 523–529 (2005).

    CAS  Google Scholar 

  9. 9

    Richardson, L. J., Deming, D., Horning, K., Seager, S. & Harrington, J. Nature 445, 892–895 (2007).

    ADS  CAS  Article  Google Scholar 

  10. 10

    Grillmair, C. J. et al. Astrophys. J. Lett. 658, 115–118 (2007).

    ADS  Article  Google Scholar 

  11. 11

    Harrington, J. et al. Science 314, 623–626 (2006).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Burrows, A. Remote climes. Nature 447, 155–156 (2007). https://doi.org/10.1038/447155a

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing