The Sailing Stones of Racetrack Playa
Later explainedSummary
On a dry lakebed in Death Valley, rocks move. Some weigh three hundred kilograms. They leave long trails behind them in the cracked mud, hundreds of metres in some cases, with turns and reversals in them.
Nobody had ever seen one move. For sixty years people proposed hurricane winds, magnetic fields, slime, pranksters and aliens.
In December 2013, two men who had set up cameras and put GPS units inside the rocks stood on the playa and watched it happen. The answer is a sheet of ice three millimetres thick and a wind you could walk into without leaning.
What is documented
The trails. Rocks on the Racetrack, ranging from small stones to boulders of several hundred kilograms, leave furrows in the playa surface. The trails run for tens and sometimes hundreds of metres. Some are straight. Some curve. Some contain sharp turns, and some rocks reverse direction. Parallel trails from different rocks sometimes turn in unison.
Nobody saw it. This is the central fact of the case, and it is what made it durable. The trails were documented from the 1940s onward. The rocks were photographed, measured, mapped and marked. Not one person, in sixty years, ever witnessed a rock move.
The proposals. Hurricane-force winds. A film of algae reducing friction. Ice sheets. Magnetic fields. Dust devils. Practical jokers. Extraterrestrials. Some were serious hypotheses and some were not, and none could be tested, because the event had never been observed.
The experiment. Richard Norris, a palaeobiologist at the Scripps Institution of Oceanography, and his cousin James Norris, an engineer, obtained permission to instrument the playa. They installed a weather station, set up time-lapse cameras, and placed fifteen limestone rocks fitted with GPS units on the surface.
They expected to wait years. They said as much: they anticipated it would be the most boring experiment they had ever run.
They saw it. In December 2013, the playa was covered by a shallow pond a few centimetres deep. Overnight, a thin sheet of ice formed on it. In the morning sun the ice began to break up, and the Norrises watched, and photographed, and recorded by GPS, rocks moving across the playa in front of them.
The results were published in PLOS ONE in August 2014.
Leading explanations
Floating ice, and a light wind. Established, and observed.
The mechanism requires four things to happen in sequence, and the reason nobody had solved it is that all four are needed and the combination is rare.
Water. The playa floods to a depth of a few centimetres. Not enough to submerge the rocks; enough to cover the ground.
Ice. A cold night freezes a layer on the surface. It is thin. The ice observed by the Norrises was on the order of three to six millimetres, what they called windowpane ice.
Breakup. Sun in the morning melts the ice enough that it fractures into large floating panels, metres across, still rigid.
Wind. A light breeze pushes the ice panels across the pond. The panels press against the rocks. The rocks, sitting on a slick saturated mud bottom, slide.
The wind is the part everyone got wrong. For sixty years the assumption was that moving a three-hundred-kilogram boulder requires an enormous force, so the theories reached for hurricane-force gusts. They do not. The Norrises recorded movement driven by winds of a few metres per second, around ten miles an hour. A large panel of ice has an enormous surface area for the wind to push against, and it transmits that force to the rock, and the rock is sitting on wet mud with almost no friction.
Why nobody ever saw it. The rocks move at a few centimetres per second. That is slow enough that, standing on the playa, you would have to be watching a particular rock with attention to notice that it was moving at all. The motion is nearly silent, lasts a few minutes, and happens on a remote lakebed on the rare mornings when four separate conditions align.
The phenomenon was not hiding. It was just extremely boring to look at, and almost nobody was there.
What the popular version gets wrong
"It took powerful winds." It took about ten miles an hour. The single most persistent assumption in the case was that a heavy rock needs a strong force, and it is wrong: the ice does the work, and the mud removes the friction.
"It was never explained." It was explained in 2014, by two men who put cameras on it and waited.
"Ice sheets had already been proposed, so nothing new was found." Ice had been proposed, and the version usually proposed was wrong: a thick raft of ice buoying the rock up and carrying it. What actually happens is a thin panel of ice pushing the rock along the ground. The distinction matters, and only observation could have settled it.
"The mystery lasted because it was hard." The mystery lasted because for sixty years, nobody put a camera on it. The Norrises solved a sixty-year problem by going to the place where it happened and waiting for it to happen.
Current status
Later explained. Observed directly, recorded by GPS, photographed, published, and reproducible in principle by anyone willing to spend a winter on a lakebed in Death Valley.
The registry keeps this record, and keeps it alongside AR-0087, because the two are the same story. Both were mysteries for decades. Both were solved when somebody finally went and measured the thing instead of theorising about it.
Sources
- Norris, R. D., Norris, J. M., Lorenz, R. D., Ray, J. and Jackson, B. (2014). "Sliding Rocks on Racetrack Playa, Death Valley National Park: First Observation of Rocks in Motion." PLOS ONE 9(8): e105948.
- GPS instrumentation, time-lapse photography and weather station data, Racetrack Playa, 2011 to 2014.
- Death Valley National Park records on the trails and the playa surface.
Last reviewed: July 2026. Records are provisional. Where the evidence changes, the entry changes. Found an error? Tell us.