Craters gouged into the ruddy Martian terrain have revealed subsurface water ice closer to the red planet's equator than would be expected, new orbiter images show.

The ice also seems to be 99 percent pure, instead of the dirty dust and ice mixture some scientists expected to see, scientists said today.

And while numerous surface features on Mars suggest that water once flowed on the red planet in the past, the new discovery - detailed in the Sept. 25 issue of the journal Science - adds to the evidence that has been piling up in recent years that water exists on present-day Mars, in the form of subsurface ice. It also gives scientists a way to further probe the Martian surface for signs of water ice.

Because water is essential to life as we know it, any findings of potentially once-liquid water has implications for the search for evidence of possible past Martian life.

The new finding comes just one day after scientists announced new evidence for water ice on Earth's moon.

Found just in time

In August 2008, members of the Mars Reconnaissance Orbiter's (MRO) Context camera team examined images of the northern Martian mid-latitudes taken by the camera for any dark spots or other changes not seen in earlier images. These dark marks are signs of meteorites that have recently crashed into the dust-covered Martian terrain.

They found several, and the following month, members of MRO's HiRISE camera team followed up by snapping high-resolution images of these suspected impact craters.

"We saw something very unusual when we followed up on the first of these impact craters," said HiRISE team member Shane Byrne of the University of Arizona, "and that was this bright blue material poking up from the bottom of the crater. It looked a lot like water ice."

A few days later, MRO's Compact Reconnaissance Imaging Spectrometer (CRISM) was used to take the spectrum of the material and, sure enough, it found the spectral signature of water ice.

The behavior of the material over the ensuing days also helped clinch its identity: "When we started monitoring the material, it faded away like you'd expect water ice to fade, because water ice is unstable on Mars' surface and turns directly into water vapor in the atmosphere," Byrne explained.

The relatively quick disappearance of the ice means the MRO teams were fortunate to have spotted the craters when they did.

"All of this had to happen very quickly because 200 days after we first saw the ice, it was gone, it was the color of dirt," Byrne said. "If we had taken HiRISE images just a few months later, we wouldn't have noticed anything unusual. This discovery would have just passed us by."

Further evidence

The evidence of these ice layers exposed by meteorite impacts stacks on top of other recently uncovered clues, including the excavation of a shallow water ice layer by NASA's Phoenix Mars Lander last year.

But Phoenix was at a more northerly locale than the new craters, so this fresh evidence shows that the subsurface water ice extends further south than previously thought.

"We knew there was ice below the surface at high latitudes of Mars, but we find that it extends far closer to the equator than you would think," Byrne said.

Also surprising was how clean the water ice was (something Phoenix also observed, along with the gradual sublimation of exposed ice).

"The thinking before was that ice accumulates below the surface between soil grains, so there would be a 50-50 mix of dirt and ice," Byrne said. "We were able to figure out, given how long it took that ice to fade from view, that the mixture is about one percent dirt and 99 percent ice."

The craters, which ranged from 1.5 to 8 feet (about 0.5 to 2.4 meters) deep, were located at five Martian sites.

Though the MRO researchers had identified 80 to 90 craters around the Martian globe before, this was the first time the spotted ice in the bottoms, likely because most of the others were more southerly and outside of the likely area of subsurface water ice.

Byrne told SPACE.com that it was surprising to the team to find the bluish ice, though "in retrospect maybe it shouldn't have been." Scientists knew of the existence of underground ice and had been monitoring craters as they formed, but "I guess we didn't put the two together," he said.

Several of the craters were also near the landing site of the Viking Lander 2. Viking also looked for water ice on Mars, but was only able to dig down about 6 inches (15 cm) below the surface — about 4 inches (10 cm) shy of where Byrne and his colleagues think the ice table sits.

"It's a shame that didn't happen," Byrne said. "You might have been having this conversation 30 years ago."

How the ice got there

There are several theories as to how such pure ice could form under the Martian surface. Byrne thinks that one of the most promising explanations is that the ice formed in the same way that so-called pure ice lenses form on Earth.

"That's where you have very thin films of liquid water around ice grains and soil grains and they migrate around to form clear ice lenses on top of the ice table, even at temperatures well below zero," he explained. "This process is called 'frost heave' on Earth, and it's considered a nuisance in most places because it cracks up roads and tilts walls and destroys the foundations of houses."

However the water ice got there, it tells scientists something about Mars recent climate. The ice is essentially "a remnant of a previous climate," Byrne said, one which likely existed around 10,000 years ago.

As the climate changes and becomes drier, the ice is expected to retreat, though based on estimates of its current extent, it hasn't done so quite as quickly as expected.

"The climate has changes but the ice is still there," Byrne said. Just why that is isn't clear yet.

These ice lenses are likely to be a source of interest to those studying the possibility of life on Mars as well, though Byrne said he's "not entirely sure if this is enough water to be interesting to a microbe."

Byrne and his colleagues suggest that fresh impact craters can be used as a new tool to probe the depth and extent of Mars' subsurface water ice.

"These impacts are really very useful," Byrne said.

And this time around, Byrne and the rest of the MRO team will be ready. Mars' northern hemisphere is heading into summer, and Byrne hopes to see about 10 more craters over this a subsequent seasons, building up a map of where known subsurface ice exists. Of course, the observations depend on MRO's successful reboot out of its current safe mode, which has temporarily suspended all science operations.

"This is I hope the start of a promising new method" of looking for water ice, he said.