Reading the Environment in the Sand

December 21, 2004

Doug Jerolmack Doug Jerolmack talks to students and scientists about wind-driven processes on Mars at an American Geophysical Meeting in San Francisco. Photo credit: NASA/JPL Link to Full Res Sometimes a scientist can see a whole world in a grain of sand. Take Doug Jerolmack, for instance. A graduate student in geophysics at the Massachusetts Institute of Technology, Jerolmack has gone from counting grains of sand on Mars and plotting their size and distribution to programming computers to predict the amount of wind it takes to move them around on the surface. That's any wind. Not just on Mars. Not just on Earth. Any wind, on any planet, anywhere. In fact, he says, the equations used to estimate the movement of sand should be valid independent of particular variables such as gravity or atmospheric density. "We can look at a grain of sand and measure it to get its size and mass. From that, we can calculate the wind strength it takes to pick up the grains of sand and move them," he explains, delighted to have an interested audience. "Basically, instead of looking forward and predicting where the sand will end up, we're looking backward and figuring out how it got there." Jerolmack's research was the subject of a poster presentation at the annual American Geophysical Union meeting in San Francisco, where an estimated 11,000 scientists convened in December to discuss their latest findings. For several weeks during the previous summer, Jerolmack collaborated with an entire team of Mars Exploration Rover scientists, including MIT geology professor John Grotzinger, Cornell geologists Robert Sullivan and Steve Squyres, and several others. Their poster was titled "Aeolian Processes at Meridiani Planum." (Aeolian means wind.) Sand Grains Up Close Studying wind processes in Mars involves counting grains of sand in microscopic images such as this one to get statistical estimates of grain size and distribution. Image credit: NASA/JPL/Cornell Link to Full Res Careful Observations Are Key Jerolmack's contribution is fairly representative of the kind of work graduate students, and ultimately scientists, do. It involves measuring, counting, making detailed observations, and keeping copious notes. For Jerolmack, it began when he realized that statistical averages of grain sizes in martian sand could be used to estimate wind strength at the surface. To make these estimates, he helped examine microscopic images sent to Earth by the Opportunity rover and counted the grains of sand. A typical image might have a couple of dozen larger spherules (affectionately called "blueberries" by the rover team) a few millimeters across, along with several dozen smaller grains big enough to count. By counting grains repeatedly, team members were able to estimate the relative frequency of different-size grains of sand. Evidence of Wind Erosion Prevailing wind direction is indicated by wind tails, bedrock protected from wind erosion behind a spherule protruding from the rock surface. Opportunity took this snapshot with the microscopic imager on martian day, or sol, 85 (April 20, 2004). Image credit: NASA/JPL/Cornell/USGS Link to Full Res Distribution was also important. They noted, for example, that on occasion dark and light grains of sand formed lines or other patterns that stood out against the background. Sometimes grains of sand formed tiny ridges or ripples, while others were distributed randomly across an entire surface. On a larger scale, they charted the geographic alignment of sand dunes in images from the rover's panoramic cameras. In close-up images, they measured the direction of wind tails - the flaring out of bedrock material protected in the wind shadow behind a spherule protruding from a rock surface, similar to the wake of seawater that forms behind a moving boat. From such directional clues, they determined that the predominant winds blew from either the northwest or southeast. As with many scientific undertakings, they also observed some things by chance. One pleasant summer day, Jerolmack and some MIT colleagues were at the beach beside a Cape Cod dune field. One of them wondered aloud if larger sand grains could roll over the top of finer sand grains. "We got down on our hands and knees and started blowing," he said. "Sure enough, the larger sand grains moved while the finer grains remained in place. We could crush the sand surface with our fingertips, yet it was cemented enough by seawater to remain in place." Ripples on Meridiani Planum Computer simulations can calculate the wind speed required to produce aeolian features such as these ripples on Meridiani Planum. The Mars Exploration Rover took this image with its panoramic camera in April, 2004. Image credit: NASA/JPL/Cornell Link to Full Res Dust in the Wind Once in place, very fine material such as dust can be just as hard to move as large grains, because of electrostatic attraction between individual grains. A similar process likely occurs on Mars, he adds. Early in the Mars rover mission, scientists noticed that sand ripples on the red planet were covered by an armor of larger sand grains on top of the finer stuff. The larger grains formed a resistant barrier, harder to blow away. Ultimately team members calculated that on average, a surface wind velocity of 80 meters per second would be needed to move a grain of sand at Opportunity's landing site on Mars. At that point, sand grains begin to roll. Occasionally, they collide and bounce into the air. Over time, this colliding and bouncing, known as saltation, abrades the grains so they become highly rounded, losing their sharp edges and reducing their size. Given the wind velocity, the scientists may also be able to estimate the length of time needed to form martian dunes. But the real payoff, in Jerolmack's judgment, is that, given the laws of physics, the same computer model can be used again and again to describe what will happen on any sandy surface anywhere in the universe. "Once you know the relationships between atmospheric density, gravity, wind speed, grain size, and other variables, all you do is plug in the values you have and the sediment transport equations tell you everything else. This allows us to reconstruct environmental conditions on the surface of any planet where we can get good enough data." Though it can take weeks or months or, in some cases, even years to collect enough data, Jerolmack derives great satisfaction from being able to put all the pieces together. "What I am really interested in is the detective story -- that we can use these techniques to unravel past climate conditions on another planet," he says. "Sometimes the sediments are all that we have left in the geologic record."