Water on the Moon?

Recent headlines have announced a raging controversy among scientists about whether there is actually water ice in the permanently shadowed craters near the lunar poles. Because these permanently shadowed regions are extremely cold (~100K) water ice is expected to be stable there ? even in the vacuum of space. If water is present, it will dramatically reduce the cost of a lunar base. The Lunar Crater Observing and Sensing Satellite (LCROSS) mission is intended to test for this water by impacting the lunar surface with its empty rocket upper stage, and looking for water in the ejected plume.

At issue is the evidence for water ice from neutron counts and radar observations of the polar regions. In the mid-1990s, Earth-based radar observations of the polar regions detected areas of higher circular polarization ratio (CPR). This is the ratio of the reflected radar signal having the same circular polarization as the incident beam to the reflected signal having the opposite polarization. High values of CPR tend to indicate blocky, irregular surfaces such as boulder-strewn fields or crater ejecta, but can also result from low-loss scattering by water ice entrained in smoother surface material such as soil or regolith. Large rocks and boulders mixed into the regolith also can have a distinct signature. The high CPRs detected at the lunar poles seemed to correlate with crater floors in these regions, where geometry and orbital mechanics conspire to produce permanently shadowed regions. This effect has been observed on Mercury, where permanently shadowed craters exhibit CPRs which resemble those from the icy Galilean satellites. The radar results for the moon were thus interpreted to be due to similar effects.

The neutron detector on Lunar Prospector subsequently found low counts of epithermal neutrons over the polar regions as well, which are typically indicative of hydrogen. However, it is not clear whether this is hydrogen in the form of pure water (H₂O), hydroxyl (OH^-, likely bound to minerals), or hydrogen (H⁺) implanted by the solar wind. If the hydrogen is in the form of H₂O or OH^- bound to minerals, it could still be useful, but extraction will be more difficult and costly.

The current controversy centers around recent radar observations at higher resolution than the previous studies. These observations show high CPR's in crater walls and floors, but they are not correlated with the permanently shadowed regions. They are correlated with the rocky surfaces and ejecta blankets around young craters, and with the inner walls of some of the larger craters. They definitely do not indicate large deposits of pure ice — there are no skating rinks on the Moon, as one scientist put it. And the neutron studies detect hydrogen, not water — so there are other ways to explain those results too. The results also depend on how much solar wind implanted hydrogen is assumed to be present, as well as the exact locations and sizes of the permanently shadowed regions — which aren't actually very well known, since we cannot see them all from Earth. The two areas where the shadowed regions are best constrained are within the craters Shackleton and Shoemaker.

Water ice remains consistent with the neutron studies, but only in concentrations of around 0.4 to 1 percent by weight, as small chunks mixed into the regolith. Or, if the ice is trapped in small pores in the soil minerals, it could be as much as 10 to 20 percent. Surface roughness and neutron counts both vary on spatial scales much smaller than the instruments can resolve, creating additional uncertainty. So the interpretation depends on certain assumptions, and can vary — it is still consistent with water ice mixed into the soil, and possibly even large amounts, but more likely very small amounts, if any at all.

The LCROSS mission is being optimized for 0.5 to 2 percent by weight of water in the soil — consistent with the neutron results. The general consensus at present is that the radar polarization in the polar regions is consistent with soil containing small amounts of ice, or could be due to surface roughness effects. Because the polarization is not actually correlated with the permanently shadowed regions, many are now leaning toward the surface roughness hypothesis, but there are insufficient data at present to conclusively determine whether there is ice. A dry regolith is virtually indistinguishable from regolith containing tiny amounts of ice within and between mineral grains. This is why the radar experiments on Chandrayan and Lunar Reconnaissance Orbiter (LRO) are being designed to understand the relationship between CPR and ice on the moon, and why the LCROSS mission is being sent to directly probe the surface and test these hypotheses through observation and experiment.

And that is how the scientific method deals with controversy.

• VIDEO: Moon 2.0: Join the Revolution
• The Greatest Lunar Crashes Ever
• All About the Moon