First author

Hot Jupiters are a class of giant extrasolar planet that reach very high temperatures owing to their close orbits about their stars. Theoretical models of hot Jupiters predict that water vapour should be abundant in their atmospheres. Surprisingly, Carl Grillmair, an astrophysicist at the Spitzer Science Center in Pasadena, California, and his colleagues found no water in the atmosphere of HD 189733b, an extrasolar planet 63 light years from Earth, when they used NASA's Spitzer Space Telescope to measure the planet's emission spectrum two years ago. On page 767, Grillmair's team present, from additional Spitzer data, the highest quality emission spectrum of an extrasolar planet so far. This enabled them to detect the predicted water absorption signature. Grillmair tells Nature more.

Was Spitzer designed to monitor extrasolar planets?

No. It was designed to explore a range of other objects. It happened to be launched around the time of an explosion of extrasolar-planet discoveries. With its infrared detectors, Spitzer is uniquely equipped to measure the atmospheres of extrasolar planets.

Did the initial results further your interest in this planet?

It was a bit disturbing to find no evidence of the water that atmospheric specialists were convinced should be there. Having published the initial results, we submitted a proposal to make more extensive spectral measurements. After monitoring 10 separate orbits of the planet, we now have the highest signal-to-noise ratio spectrum currently achievable.

Why did you miss the water initially?

Even after reanalysing the original data we still don't know. Unfortunately we can't directly compare our data sets because in the latest work we increased the exposure times to get better sampling. However, other data suggest that changes in the weather on these planets may occur on a global scale, which could explain the differences between our measurements.

Do the data shed light on HD 189733b's atmospheric conditions?

Yes. Theoreticians expect these kinds of planet, which have the same side always facing their star, to be covered with rapidly changing, intense storms able to transport huge amounts of energy. Our data show that energy transport from the dayside to the nightside is weak. However, other data suggest much stronger energy transport. Reconciling this inconsistency might provide clues to what's happening in the deeper atmospheric layers.