Scientists have long known water is currently frozen in Mars’ polar caps. But on a planet with an average temperature that NASA puts at minus 64 degrees Fahrenheit, the difference between water and ice stops being trivial.

Today, scientists agree Mars is cold, dry, geologically inactive, and nearly airless.

But the planet’s surface boasts a complicated network of channels and features that geologists say appear to have been carved by flowing water in the distant past.

What researchers are looking for from the rovers, the EU’s Mars Express orbiter, and future missions, is clues that will resolve the debate over how the water got there.

Anthony Colaprete, a planetary atmospheric scientist at NASA Ames research center, calls the mystery of the ancient “warm, wet mars” one of the most vexing problems in planetary science.

Nick Hoffman, a planetary scientist at the University of Melbourne in Australia, agrees.

“We don’t understand the context in which those fluids existed,” he said.

In 1996, Raymond Pierrehumbert, a climatologist at the University of Chicago, and colleague François Forget produced weather models indicating that clouds of carbon dioxide could create a greenhouse effect, trapping sunlight that could heat Mars sufficiently to sustain liquid water on the surface.

Greenhouse effect models have been embraced by the scientific community as an explanation for how there could have been a warm, wet mars where rainfall would form channels like those seen in the southern highlands, Colaprete said.

The only problem, he adds, is that the models don’t work.

“The crux of the problem with early mars is it’s just too damn cold,” he said. When the channels formed, over 3.5 billion years ago, astronomical observations clearly show that the sun was about 25 percent cooler than it is today.

To get temperatures warm enough to sustain liquid water would have required copious amounts of carbon dioxide in the atmosphere, Colaprete said, and greenhouse effects can’t warm an atmosphere indefinitely. Eventually the gases that trap solar heat will condense out into clouds that reflect that heat back into space.

Since 1996, Colaprete and Pierrehumbert have dueled in the pages of leading scientific journals over whether a greenhouse effect could warm Mars.

“The community accepts this paradigm, but cannot come up with a model to show that it works,” Colaprete said.

In 2002, Colaprete and Teresa Segura, a Ph.D. student at the University of Colorado, Boulder, advanced an alternate idea.

Instead of a continuous warm climate, Mars’ channels may have been formed by brief periods of catastrophic flooding following major meteor impacts.

“Impacts occurred, you know this because there are craters,” Segura said. She said she began her model with the intent of seeing how the bombardment of Mars could have affected temperature.

When it became apparent that impacts could cause periods of rainfall lasting months or years, she saw a potential explanation for Mars’ mysterious features.

Hoffman, who advocates a similar model based on meteor impacts releasing trapped carbon dioxide, said catastrophic climate changes could explain why the network channels on Mars look more like young streams than the mature river systems one would expect to develop in a sustained warm climate.

Obviously, the two different possibilities would have vastly different consequences for the possibility of life having once existed on the red planet.

Other competing theories include a “cold, wet” theory, in which tiny trickles of water slowly carved Mars’ channels over eons, and a number of atmospheric models relying on the presence of other greenhouse gases like methane, scientists said.

“When you have very limited data, you find you can tell almost any kind of story,” Colaprete said. “Physics can help you not to get too out of hand.”