The research suggests that even though Mars is now very cold, flowing water may have carved the gullies in the middle latitudes within the past 500,000 years.
Liquid water is of interest because it could provide a home for life. Some scientists had proposed that Mars has been cold and dry for a billion years or more.
Many explanations have been proposed for the gullies.
Some suggest they were formed by water from springs or the melting of subsurface ice. Others trace the gullies to frozen carbon dioxide from the atmosphere.
In a study to be published online Thursday by the journal Nature, Arizona State University geologist Philip Christensen proposes another source: snow created and then melted by a wobbling planet.
His theory depends on the behavior of the planet's axis, the imaginary line that runs through both its poles. The axis of Mars wobbles slowly. Over the course of 100,000 to 1 million years, its tilt can change by more than 20 degrees.
Christensen's theory goes like this:
When the axis is tilted far toward the Sun, the sun's heat falls more intensely than usual on one of the poles, depending on the planet's position in its orbit. The heat creates water vapor from the polar ice and the vapor later turns to snow that falls in the middle latitudes, closer to the equator.
Later, when the axis is more upright and the sun shines more directly on these middle latitudes, that snow melts, creating the liquid water that formed the gullies.
Christensen said he developed the theory after examining images taken five years ago by the Mars Global Surveyor.
"Our thinking was all of the water activity on Mars had occurred really early and today it's cold and frozen. But then these pictures came down and we said `Whoa, liquid water is flowing in really recent times,"' Christensen said. "It's an active process, this isn't a dead process. It isn't a dead planet."
Liquid water on Mars would be important not only as a habitat for life, but also as a source of hydrogen and oxygen for future astronauts, Christensen said.
Christensen believes some flat areas on Mars contain dust-covered snow. Space probes could be targeted to those areas, he suggests.
Dr. Bruce M. Jakosky, director of the center for astrobiology at the University of Colorado in Boulder, said Christensen's theory is "much more intriguing" than others that have been proposed.
"I think this is on the right track," Jakosky said.
The theory is exciting because of the implication the snow could harbor life, Jakosky said.
Two robotic rovers being sent to Mars this year should gather data that will help prove whether Christensen's theory is correct, said John Mustard, a Brown University geology professor who studies Mars.
By focusing on the middle latitudes, Mustard said the theory puts water "in a place that has far greater implication for life than other ones and makes it a target for exploration."
By Alex Dominguez