Over the years I've seen dozens of partial lunar eclipses. They're exciting, they're pretty — but scientifically useful? "No way," I used to say.

Then I read an article in June 11th's Nature by an observing team led by Enric Pallé (Instituto de Astrofísica de Canarias). He and four colleagues did something that, apparently, hasn't been attempted in more than 80 years: they recorded the lunar spectrum during last August's deep partial eclipse.

So why do that? Because, as they explain, the dim light reflected off the lunar surface within the umbra is sunlight that has passed through Earth's atmosphere. It's literally the spectrum of Earth transiting the Sun, and that could prove critical as astronomers attempt to seek habitable Earthlike worlds elsewhere in space.

Of the 353 exoplanets tallied to date, 59 of them were discovered because they transit their host stars. In a few of these cases, astronomers have been able to divine hints about composition from the altered spectrum of starlight passing through the planet's atmosphere.

By recording Earth's visible and near-infrared spectrum as mirrored in the Moon, Pallé and his team hit the jackpot. Not only did they detect the absorption bands of biologically important gases (water vapor, oxygen, carbon dioxide, and others), but these spectral fingerprints proved more obvious that expected and, critically, stronger they appear in sunlight reflected directly off Earth's atmosphere.

Most interesting to me was the detection of both oxygen and methane. These two gases can't coexist without reacting, so methane's presence must be continually replenished somehow. On Earth, the principal sources of atmospheric methane are decaying organic matter and flatulent livestock. What manner of strange organisms might be exhaling methane on some alien Earth?

So the next time you gaze upon the lunar disk swallowed by Earth's shadow (which won't occur until December 2010), just remember that it's not just a pretty sky show — there's real science in that dim, ruddy light.