DEATH VALLEY NATIONAL PARK, Calif. — The cracking sounds were ferocious. An ankle-deep, frozen lake in Death Valley National Park was breaking apart under sunny skies.

As cousins Richard Norris and James Norris watched, a light wind began moving huge floes of ice across the surface of the water and into rocks weighing up to 200 pounds. Propelled by the ice masses, the rocks began to slide across the slick, muddy bottom of the normally dry lake bed known as “the Racetrack Playa.”

“My god, Jim, it’s happening,” Richard yelled.

James Norris grabbed a camera.

Their photos last Dec. 21 provided the final evidence in solving a mystery of the Racetrack Playa that has long puzzled visitors and scientists: What mechanism moves rocks across flat dirt in the heart of the hottest, driest place on Earth?

Rocks of various heft — some weighing 600 pounds or more — leave trails that wiggle like snakes or form complete loops or even rectangles. The trails are cut sharply into the dirt but no other tracks are visible.

Theories over the decades have included sporadic hurricane-force winds when the surface is covered with rain water, or rocks carried across the mud by small rafts of ice, or UFOs.

But until the Norrises had an incredible stroke of luck that December day, no one had verified the phenomenon. The findings were formally presented Wednesday in the online scientific journal PLOS ONE.

“I’m amazed by the irony of it all,” James Norris said, nodding toward the glistening playa this month. “In a place where rainfall averages 2 inches a year, rocks are being shoved around by mechanisms typically seen in arctic climes.”

“And the movement is incredibly slow,” he added. “These rocks clock in at about 15 feet per minute.”

Geologists have been studying the moving rocks since 1948, when the first study suggested they were driven by dust devils. One reason the mystery endured is that the movements are episodic, often with no motion for decades until a precise series of natural events occurs.

The first requirement is rain in a parched climate. Next, temperatures must fall low enough to freeze the water before it evaporates. Then the sun has to come out and thaw the ice.

Finally, wind has to blow to break the ice into floes and move it across shallow water underneath. Even a light wind will do.

Ralph Lorenz, a researcher at the Johns Hopkins University Applied Physics Laboratory who had investigated playa rock movement for a decade, believed strongly enough that ice floes were the cause that he erected time-lapse cameras in the area about seven years ago. But they failed to record the phenomenon.

The Norrises subscribed to a different theory, believing hurricane-force winds were the cause.

Richard, 55, a paleobiologist at the Scripps Institution of Oceanography, and James, 59, a research engineer, launched their “Slithering Stones Research Initiative” in 2011.

During the next two years, friends and relatives armed with permits from the National Park Service helped them install a weather station in the area and place 15 stones equipped with GPS devices on its pancake-flat surface.

The “GPS stones,” which were engineered to record movement and velocity, were stationed at the southern end of the playa where rocks begin their strange journeys.

On Dec. 20, 2013, Richard and James Norris returned to inspect the instruments. “We found the playa covered with ice,” Richard recalled. “We also noticed fresh rock trails near shards of thin ice stacked up along the shoreline.”

The next afternoon, “We were sitting on a mountainside and admiring the view when a light wind kicked up and the ice started cracking,” he said. “Suddenly, the whole process unfolded before our eyes.”

“There was a side of me that was wistful,” James Norris added, “because the mystery was no more.”

A review of the weather data showed that a rare winter storm had dropped about 1½ inches of rain and 7 inches of snow on the region in late November. The playa was transformed into a shallow lake where the GPS stones recorded movements on sunny days with light winds after nights of subfreezing temperatures.

James Norris’ photographs put it all in perspective. Panes of ice hundreds of feet across and as thin as ¼ -inch thick blew into rocks. The rocks slid along the slushy, slippery mud on trajectories determined by the direction and velocity of the winds.

The cousins first shared the news with Lorenz, whose specialty is the Saturn moon of Titan and who became one of the five authors of the PLOS ONE study.

“While it takes away the mystery, it also underscores what an amazingly rare and wonderful mechanism is at work there,” Lorenz said.