Racetrack Playa in Death Valley is famous for it's "living stones"—photographed here by BoingBoing reader Rajesh Vijayarajan. In this patch of desert, rocks, ranging in size from pebbles to boulders, have left trails in the dry clay, as if they moved across the mostly flat Playa by themselves.
Based on analysis of the trails, scientists assume the rocks do move, but they still don't know exactly how. To date, nobody's ever seen the process in action. But there is a good theory. It starts with the fact that Racetrack Playa is also, occasionally, a very shallow lake.
Researchers noticed that although some trails change direction, most trend in a generally southwest to northeast direction. This is consistent with the direction of the prevailing winds. ... After analyzing their rock trail map, researchers found that the longest, straightest trails are concentrated in the southeastern part of Racetrack Playa. In this area, wind is channelled through a low point in the mountains, forming a natural wind tunnel ...
The evidence suggests that strong gusts of wind and swirling dust devils, in combination with a slick playa surface may set even the heaviest the rocks in motion. Off they go, scooting along downwind until friction slows them down and they come to rest.
That's currently the accepted professional explanation.
But, based on what he observed on a trip to the Play, blogger Brian Dunning thinks the scientists are missing two key factors: Ice, and a lake that moves.
In the early spring of 2002, I made one of my many trips to Racetrack Playa with two friends, Dan Bocek and John Countryman. The surrounding mountains were still covered with snow, and the playa itself was firm but had a large lake covering about a fifth of its surface, perhaps an inch or two deep at its edges, concentrated at the playa's south end where it's lowest. We ventured out, armed with cameras, shortly before sunrise. The temperature was just above freezing. The wind, from the south, was quite stiff and very cold. When we reached the lake, we found to our great surprise that the entire lake was moving with the wind, at a speed we estimated at about one half of a mile per hour.
The sun was on the lake by now and we could see a few very thin ice sheets that were now dissolving back into water. This whole procession was washing past many of the famous rocks. It's easy to imagine that if it were only few degrees colder when we were there -- as it probably had been a couple of hours earlier -- the whole surface would be great sheets of thin ice. Solid ice, moving with the surface of the lake and with the inertia of a whole surrounding ice sheet, would have no trouble pushing a rock along the slick muddy floor. Certainly a lot more horsepower than wind alone, as has been proposed. The wind was gusty and moved around some, and since the surface is not perfectly flat and with rocks and various obstructions, the water didn't flow straight; rather it swapped around as it moved generally forward. Ice sheets driven by the water would move in the same way, accounting for the turns and curves found in many of the rock trails.
I didn't totally follow the description the first time I read it, but about half way through this video, you can see the phenomenon he's talking about. It's pretty fascinating.
Scientists have thought about the role ice might play in moving the stones, and they've mostly ruled it out. Why? Because fresh trails have been spotted during periods when the Playa was too warm for ice to form. (Check out Chapter 7 of this 1998 dissertation. It also gets into more detail about what makes the surface of the Playa slick enough that rocks could move over it when it was wet.)
That paper doesn't mention the action of the lake, itself, moving across the valley, though. And I haven't found any other scientific references that do. I wonder if this is something researchers haven't observed? Or if it's just interesting, but not necessary to the movement of the rocks.