A Red Haze on Distant Exoplanet

by Paul Gilster on December 12, 2007

We’re learning more and more about HD 189733b, an extrasolar planet some 63 light years from Earth in the direction of the constellation Vulpecula. This transiting ‘hot Jupiter’ orbits once every two days about three million miles out from its primary. David Charbonneau (Harvard-Smithsonian Center for Astrophysics) and team recently measured an unusual spectrum from the planet’s atmosphere using the Spitzer Space Telescope. Looking for water, carbon dioxide and methane, they found instead a flat spectrum that Charbonneau thinks may indicate the presence of dark silicate clouds.

Now we have further work, this time using Hubble Space Telescope data, that points to the presence of haze in the atmosphere of HD 189733b. That’s an interesting finding to which we can add another result: Studying how light varies when the planet makes its transit indicates that this world has neither Earth-sized moons or a discernible ring system. Moreover, we’ve got a fairly good read on the temperature of its atmosphere, a toasty seven hundred degrees Celsius. That’s a lot to learn about a place we can’t image with the best equipment we possess.

Like Charbonneau’s team, the group studying the Hubble data, led by Frédéric Pont (Geneva University Observatory), knew what they were looking for. But there were no signatures characteristic of sodium, water or potassium. Weighing what they found in the entire planetary spectrum, the scientists infer that high level hazes about 1000 kilometers in altitude are present. Likely haze constituents are tiny particles of condensates of iron, silicates and aluminum oxide dust. Another finding: A starspot on HD 189733b’s surface thought to be over 80,000 kilometers across.

Pont points to the significance of the work in terms of future goals:

“One of the long-term goals of studying extrasolar planets is to measure the atmosphere of an Earth-like planet. This present result is a step in this direction. HD 189733b is the first extrasolar planet for which we are piecing together a complete idea of what it really looks like.”

True enough, but let’s be cautious. I can recall when we had a pretty good idea of what the Jovian moons looked like. Then Voyager changed everything, and we realized that the outer planets weren’t simply orbited by small worlds much like our Moon. The stunning variety that exists around Jupiter alone reminds us to hedge our bets when weighing data from so much further out. Still, this work, which studies starlight passing through the atmosphere of a giant planet in transit, is a marker of how we’re learning to use existing equipment to begin filling in planetary details. Bit by bit, HD 189733b is getting to be known. The unanswered question is, how many surprises does it hold in store for the future?

Addendum: The lack of detected sodium in the atmosphere of this world, as one reader has already pointed out (thanks, Luis!), seems to contradict the recent findings from the HET instrument in Texas, discussed in this earlier Centauri Dreams article. Comments?

luis December 12, 2007 at 17:34

Hi! this result is curious in the light of another recent result (last week) for this system which used ground based observations with the HET (Hobby-Eberly Telescope) to make the first ground based detection of sodium in the visible spectrum the atmosphere of HD189733b. The results with the HST would be inline with those obtained earlier for HD209458b, which also failed to detect alkaline metals in the atmosphere. It seems to me that one of the teams got it wrong, or is there a way to reconcile the observations ? Cheers, Luis

andy December 15, 2007 at 10:45

Perhaps the two studies probed different regions of the planet’s atmosphere, or maybe the weather systems are very dynamic and the planet’s climate oscillates?

Maybe it’s time to drag out my copy of POV-ray and have a go at rendering this planet.

luis December 17, 2007 at 6:10


Well, that is what I thought at first (different layers of the atmosphere) but the wavelengths probed by both experiments overlap to a great extent, namely in the visible at the wavelength of Na I. The data from the HET was taken over a year so maybe the Charbonneau team just got some unlucky hazy weather.

These guys are stretching the limits of current technology and techniques. It is not too far fetched to think that, despite their careful analysis, one of the teams has got it wrong. This is the price you pay for making cutting edge science. If so, the important thing is to identify the source of the error.



ljk December 18, 2007 at 10:49

The composition of transiting giant extrasolar planets

Authors: Tristan Guillot (OCA)

(Submitted on 15 Dec 2007)

Abstract: In principle, the combined measurements of the mass and radius a giant exoplanet allow one to determine the relative fraction of hydrogen and helium and of heavy elements in the planet. However, uncertainties on the underlying physics imply that some known transiting planets appear anomalously large, and this generally prevent any firm conclusion when a planet is considered on an individual basis.

On the basis of a sample of 9 transiting planets known at the time, Guillot et al. A&A 453, L21 (1996), concluded that all planets could be explained with the same set of hypotheses, either by large but plausible modifications of the equations of state, opacities, or by the addition of an energy source, probably related to the dissipation of kinetic energy by tides. On this basis, they concluded that the amount of heavy elements in close-in giant planets is correlated with the metallicity of the parent star.

Furthermore they showed that planets around metal-rich stars can possess large amounts of heavy elements, up to 100 Earth masses. These results are confirmed by studying the present sample of 18 transiting planets with masses between that of Saturn and twice the mass of Jupiter.

Comments: 13 pages, 6 figures

Subjects: Astrophysics (astro-ph)
Journal reference: Dans Physica Scripta – Physics of Planetary Systems, Nobel Symposium 135, Stockholm : Su\`ede (2007)

Cite as: arXiv:0712.2500v1 [astro-ph]

Submission history

From: Tristan Guillot [view email] [via CCSD proxy]

[v1] Sat, 15 Dec 2007 07:47:56 GMT (74kb)


ljk March 14, 2008 at 14:37


WASHINGTON – NASA will hold a media teleconference at 2 p.m. EDT on
Wednesday, March 19, to report on the first-ever detection of the
organic molecule methane in the atmosphere of a planet orbiting a
distant star.

Though the planet is too hot to support life as we know it, the
finding demonstrates the ability to detect organic molecules
spectroscopically around Earth-like planets in habitable zones around

Briefing participants are:

– Dr. Mark Swain, NASA Jet Propulsion Laboratory, Pasadena, Calif.
– Dr. Sara Seager, Massachusetts Institute of Technology, Cambridge

This unique discovery, made with Hubble’s Near Infrared Camera and
Multi-Object Spectrometer (NICMOS), will be featured in the March 20
issue of the journal Nature.

To participate in the teleconference, reporters must contact Ray
Villard at 410-338-4514 or Cheryl Gundy at 1-410-338-4707 at the Space
Telescope Science Institute by noon on March 19 for the call-in
number and passcode. At the start of the briefing, images and
supporting graphics will be posted on the Web at:


Audio of the teleconference will be streamed live on NASA’s Web site


For more information about NASA’s Hubble Space Telescope on the Web,


Ron S March 14, 2008 at 15:29

Ok, I guess methane isn’t a surprise in any gas giant, but as they say the interesting thing is the ability to identify compounds. Now, if they would only detect atmospheric oxygen…

Comments on this entry are closed.

{ 1 trackback }