By Eric Hand

A decade ago Faith Vilas, director of the Multiple Mirror Telescope in Arizona, developed a sideline obsession with the Moon. Perusing archived data from the Galileo mission to Jupiter, she saw something odd in the pictures taken of the Moon. When she filtered the pictures for certain infrared wavelengths, a telling signal popped out at a few spots near the Moon's south pole. The signal, at least in asteroids, is associated with phyllosilicates, which are minerals that need two things to form: heat and water. Was this a clue pointing to all the water ice that many think hides within the Moon's polar craters?

She thought so, and submitted an abstract to the 1999 Lunar and Planetary Science Conference. But for years she couldn't get the idea published. "It kept getting roundly slapped down," she says.

Now, she's being vindicated. Results soon to be published from two other spacecraft will show detailed spectra confirming that, indeed, the polar regions of the moon are chock full of water-altered minerals.

That's not all. Early results from NASA's Lunar Reconnaissance Orbiter (LRO), launched on 18 June, are offering a wide array of watery signals. Increasingly, lunar scientists are confident that the decades-long debate is over. The Moon, in fact, has water in all sorts of places: not just locked up in minerals, but scattered throughout the broken-up surface, and, potentially, in blocks or sheets of ice at depth.

"We are on the verge of a renaissance in our thinking about the poles of the Moon, including how water ice gets there," says Anthony Colaprete, principal investigator for the Lunar Crater Observation and Sensing Satellite (LCROSS), which on 9 October will slam into a polar crater with the intention of ploughing up a plume of water ice for many telescopic eyes to see. "Our simplistic ideas were just that: simplistic."

The new evidence has scientists scratching their heads, not only to explain the origin and movements of the water, but also at how a tantalizing signal first seen a decade ago could have been left for so long. "No one really took [Vilas' work] seriously," says one lunar scientist with knowledge of the new studies, which are to be published in Science. "It wasn't until word got out that people suspected and went and looked."

Hydrogen excess

The initial LRO results, released Thursday, confirm what was long suspected as a way for ice to stay trapped on the Moon for billions of years. A thermal mapping instrument showed that permanently shadowed regions within deep polar craters are as cold as 35o Kelvin (-238o Celsius). Project scientist Richard Vondrak says they are the coldest spots in the Solar System -- even colder than the surface of Pluto.

But the surprise comes from a different instrument on LRO, which counts slow-moving neutrons as a way of measuring hydrogen abundance in the top metre or so of the surface. This hydrogen is often interpreted as a proxy for water ice, although it could also be molecular hydrogen or hydrogen trapped in other molecules. Like the earlier Lunar Prospector mission, the LRO instrument has already found a significant excess of hydrogen at the poles. But with added resolution, it is seeing surprising variability within the polar regions. Some of the craters appear enriched in hydrogen. Others are not. Stranger still, some areas outside the crater walls--which were thought to get too hot for water to linger--show an excess of hydrogen. Vondrak says this shows that the water could have arrived more recently, or that it can persist if buried as impacts till the lunar soil.

The radar instrument on LRO, which probes for chunks or blocks of ice, rather than the more diffuse signal of the neutron detector, is showing similar variability, says Stewart Nozette, principal investigator for the radar instrument. At the south pole, he is seeing strong ice-like signals in the floors of a few deep craters. But unfortunately, he says, there is no strong signal for Cabeus A, the crater that NASA announced on 9 September as the crash site for LCROSS.

Colaprete is worried about the weak radar signal at Cabeus A, and in light of the new data has delayed a final LCROSS manoeuvre. He will make a final decision on Monday or Tuesday in advance of a 26 September deadline for altering the spacecraft's trajectory. The main Cabeus crater is the current best alternative to Cabeus A, he says.

Deep impact

If the LCROSS impact spews up ice, it will eliminate the last vestiges of doubt about water on the Moon. It could also start a new hunt: to find a record of impact events, such as water-rich comet strikes, that put the ice there in the first place. These, if they were ever dated, Vondrak says, could offer similar value to ice cores in Antarctica--a way to understand an ancient bombardment history that has been erased on Earth but could live on in the craters of the Moon.

With the detection of water ice will come debate over the mechanisms that put it there. There are two popular theories. One suggests comets from the ancient heavy bombardment that peaked 3.9 billion years ago. These impacts would have buried large doses of water deep under the surface. But others think that a more continuous, shallow system of water deposition could be happening--one water source in this system could be micrometeorites, which impact frequently and carry small amounts of water.

Another source could be the solar wind, which sends a steady stream of protons into the surface, where they can combine with oxygen in the lunar soil to make water. Most of that water quickly erodes into space, but some of it can travel, via random molecular walks, to the polar traps, where it can persist. The variability in the hydrogen and radar signals between craters suggests the sporadic impacts of comets long ago, but the signals outside the crater walls could suggest a more continuous shallow emplacement. Vondrak says both theories could be partially right. "It's not either-or," he says.

Meanwhile, lunar scientists are eagerly awaiting data from two groups investigating the hydrated minerals in polar areas outside the permanently shadowed craters (the only place where instruments that depend on reflected light can see). Observations made during the extended mission of the Deep Impact probe, and from the Moon Mineralogy Mapper, an instrument aboard India's recently ended Chandrayaan-1 spacecraft, will be published in Science, and will show more detailed spectroscopic evidence for the types of watery minerals that Vilas saw in the Galileo pictures so long ago.

Vilas herself finally published her result last year in a Japanese journal. "I'm annoyed that it was ignored some years ago," she says, "but I'm really thrilled that it's being proven the case."