The wind bloweth where it listeth, and the rocks fall into pits

Dave Ricks posted the link to Emily Lakdawalla’s post on why rocks are evenly spaced on Mars so I read it so now I have to share it More.

On Mars there are rocks everywhere. The difference is that Mars’ landscape is shaped in large part by impact processes. Far-away impacts can toss rocks for miles, and they fall where they land. So it’s not particularly surprising that you see rocks everywhere, even in flat places on Mars. What is a bit surprising is their even spacing. Here’s an example of a rock-strewn landscape selected more or less at random from the early part of Spirit’s mission, when it was dashing across the flat plains to the east of its landing site toward the Columbia Hills.

Plains near the Columbia Hills, sol 149

NASA / JPL / Cornell / calibrated color by Daniel Crotty

She adds a closeup of some rocks and then explains how they get to be evenly spread.

…the wind doesn’t move the rocks, at least not directly. What the wind does do is lift sand; sand particles jump (or “saltate”) along the ground, knocking into each other and launching more sand particles. When the wind runs into a rock, it loops and whirls, scouring the area right in front of the rock. Over time, it digs a pit in front of the rock. At the same time, the sand that was scoured from in front of the rock gets deposited in the wind shadow behind the rock. Do this for long enough, dig a steep-sided enough pit, and one random day the rock will tip forward, rolling in the upwind direction, into the pit. Rinse and repeat, and you get rocks trooping across the landscape over time. (The press release didn’t give a time scale for this process.)

That explains how rocks can move, but how does it explain an even spacing? Well, according to the release, when you have a cluster of rocks, “those in the front of the group shield those in the middle or on the edges from the wind, Pelletier said. Because the middle and outer rocks are not directly hit by the wind, the wind creates pits to the sides of those rocks. Therefore, they roll to the side, not directly into the wind, and the cluster begins to spread out.”

Well how stinkin’ cool is that? Not to mention the photo.

We live in interesting times.


  1. lorn says

    Looking at pictures of the rocks on mars I’ve often felt that the spacing of rocks was somehow something other than what I had expected. It all seemed to be slightly more manipulated, possibly artificial, than what I expected.

    I was, before this, unable to put my finger on exactly what I meant by that. Now I know. The dispersion of the rocks is far too even to be entirely random. Like what people produce when people arrange dots in what they think of as a random manner. It usually comes out far too even when a more random arrangement would show lots of concentrations and empty spaces.

    Now I know that the rocks are rearranged by the wind to be much more evenly spaced. A nagging hunch now has an explanation.


  2. Dave Ricks says

    Emily Lakdawalla’s post was my favorite story of the press releases and blog posts I found on the topic, and she posted her story the day of the press release from the University of Arizona (versus other blogs I saw posted later). I didn’t know about her before; I’ll be happy to read more of her work.

    I also like that the physics of blown sand is a topic within classical physics (which we might expect to be simple) but with enough complexity that a wind tunnel is still a relevant tool versus computer simulation. This work involved the University of Arizona, University of Calgary, and University of Wyoming, where UW professor James Steidtmann showed pebbles in sand moving into the wind 30 years earlier in a wind tunnel. For the present work, the UC press release showed a photo of rocks in the UW wind tunnel again, but this time spreading out to explain the distribution of rocks on Mars.

    The wind tunnel relates to an interest of mine: in movies, they call physical in-camera effects “practicals” (versus computer-generated images), and practicals still give a particular desirable look. Which reminds me of a Hollywood joke: a full symphony orchestra was setting up to record a soundtrack, and someone remarked, “Damn, this is putting two keyboard players out of work.”  I mean, simulations are great, but physical vibrations still have complexity we want.

  3. Dave Ricks says

    On the theme of surprising physical effects still being discovered in classical physics some 300 years after Newton’s Principia, in 1971 the retired race car driver turned Indy 500 crew chief Dan Gurney empirically discovered the Gurney flap that we now understand can help keep airflow attached on the suction side of an airfoil.

    On the one hand, the Navier-Stokes equations that model the fluid flow are well known, but on the other hand, our understanding about solutions to the equations is so limited that the Clay Mathematics Institute will award you one million dollars if you can prove any one of these four claims true or false.

    Seriously, we understand so little, that proving true or false for any of those four claims would be worth a million dollars. Blows my mind.

  4. says


    On the one hand, the Navier-Stokes equations that model the fluid flow are well known, but on the other hand, our understanding about solutions to the equations is so limited that the Clay Mathematics Institute will award you one million dollars if you can prove any one of these four claims true or false.

    The Navier-Stokes equations are only known to have 3 dimensional solutions in very special cases. That’s not to say that these solutions are not useful, but it would be nice to have more general ones. The claims on the Clay Institute site are about the existence or otherwise of such solutions.

    The four claims aren’t independent:

    (A) false => (C) true
    (B) false => (D) true
    (B) false => (A) false
    (D) true => (C) true.

    Although, conceivably, (A) and (C) could both be true, or (D) could be false and (C) be true etc. My own guess is that all four claims are true. For (A) and (B), this is based on the observation that the conditions in the first two claims that f be identically zero and u° be smooth are very strong indeed. Unfortunately the most interesting applications of Navier-Stokes are to turbulent flow, where you wouldn’t expect u° to be smooth. So even proving (A) or (B) would only be a step in the right direction. For (C) and (D) it seems to me that if you relax the condition that f be identically zero then why would anyone expect the equations to have a solution in general? Mathematics isn’t normally that nice!

    In general solving partial differential equations isn’t easy. You can write computer programs that attempt to find approximate solutions. However they’re always working on the assumption that there’s an actual solution to approximate; if there isn’t then they just spew out rubbish. On the other hand I wouldn’t want that to put physicists developing theories that simply assume various equations have solutions. Rigorous mathematical justification of theoretical work in physics often only comes after the theories have been formulated and successfully used for a period of time.

    As to whether a solution to any of the four problems is worth a million dollars, I find it difficult to see how any business could benefit in the short term from such a solution. So from the free market point of view the whole exercise is pointless. However in the long term, provided we don’t nuke ourselves or globally warm ourselves out of existence, the research that the challenge will stimulate will be worth incalculably more than a million dollars even if none of the problems is ever solved.

Leave a Reply

Your email address will not be published. Required fields are marked *