The last Sunday Sand post returned to the ever-fascinating work and exploits of Ralph Bagnold; after writing it, I was catching up on arenaceous research news and there he was again – on Mars, so to speak. A recent press release from the Planetary Science Institute and the Mars HiRise imaging team announces “to their surprise” that seasonally repeated imagery reveals that the polar dunes of the planet are constantly changing. Why should this be a surprise? After all, dune systems on Earth are among the most dynamic landscapes on the planet – why should Mars be any different? Well, there are important differences, as we shall see, but that Martian dunes have been “long thought to be frozen in time” is the real surprise – and is shown to be unfounded. Yes, seasonal icing stabilises the dunes, but melting and degassing causes instability – and the wind re-sculpts the surface.
Nor should we be surprised that Bagnold enters the arena of this debate for, while his seminal work on the physics of windblown sand was first published seventy years ago, it remains the foundation of aeolian research today, his basic equations and analysis of processes refined but essentially unchanged. And we should also not be surprised that his work applies as well to Mars as it does to Earth – after all, he was called in as an advisor to NASA during the planning of the early Mars missions and, in 1974, he published a paper with Carl Sagan comparing transport on the two planets:
In this paper, as apparent in the abstract, the interest is in threshold velocities of the wind – how strong a wind is necessary to move sand? This is a complex topic, but Bagnold’s work on Earth reveals the keys to understanding sand dynamics on Mars, as revealed by this recent press release – particularly when combined with research results published last year.
But first, back to Bagnold basics. During his extraordinary expeditions, he closely observed the characteristics of dune architecture and movement, and came to appreciate that there were a number of fundamental questions that, at that point, had not been answered. One of the most basic of these questions was why do dunes form at all? Why is the sand not spread evenly over the desert floor? Whether on Earth or Mars (or, indeed, Venus or Titan), dunes appear to be self-accumulating, seeming to vacuum up sand from the bare stony areas between them – they grow by attracting more sand. “Why did they absorb nourishment and continue to grow instead of allowing the sand to spread out evenly over the desert as finer dust grains do?” was one of Bagnold’s questions. This was, he thought, something that “could be explored at home in England under laboratory-controlled conditions” - and so began his rigorous science. Two of the most important revelations of Bagnold’s work are the process of saltation and the role of two different threshold velocities for the wind.
First, saltation. From the Latin verb “to jump,” this is the process whereby sand grains move in the wind by individual leaps, and, landing on a hard surface, bounce off again; if a grain lands amongst other grains on the surface of a dune, the impact kicks some of them up into the wind and the crowd of flying grains grows. It is these two contrasting behaviours – bouncing versus splashing – that explain the self-accumulating nature of dunes. Over a hard surface of rocks and pebbles, the trajectories of individual grains are high into the air, and they keep on bouncing. As soon as they hit a soft surface of a dune, they kick off more grains, but the trajectories are lower and shorter – the dune grows. Here’s Bagnold’s original illustration of this:
So, saltation is the key activity of windblown sand. But how does it start? Clearly, if the wind blowing over a surface of sand is strong enough, it will nudge, roll, and pick up grains until the self-sustaining process of saltation begins. The wind speed at which this starts Bagnold referred to as the “fluid threshold” – and it represents a pretty strong wind. But, once grains are saltating and being kicked up into the wind, it only needs a slower velocity to keep the process going – lower the wind speed to the point where all grain motion stops, and that is the “impact threshold,” the minimum velocity to keep sand in motion – and it’s much lower than the fluid threshold.
So, back to Mars. The three sequential images at the head of this post show clear changes in the dune as the thawing of winter carbon dioxide ice destabilises the structure. The caption is as follows:
Three images of the same location taken at different times on Mars show seasonal activity causing sand avalanches and ripple changes on a Martian dune. The High Resolution Imaging Science Experiment (HiRISE) camera on NASA’s Mars Reconnaissance Orbiter took these images, centered at 84 degrees north latitude and 233.2 degrees east longitude. Dune fields at high latitudes are covered every year by a seasonal polar cap of condensed CO2 (dry ice).
The sequential images, which each show an area 285 meters by 140 meters, depict the before and after morphology of the dune in one Mars year, with new alcoves and extension of the debris apron on the slipface, or steeply sloping leeward surface, of the dune caused by the grainfall, and new wind ripples on the debris apron.
The top image was taken first, in the Martian summer when the dunes were free of seasonal dry ice. The middle image was acquired in the spring when the region was covered by a layer of seasonal ice. Spring evaporation of the seasonal layer of ice is manifested as dark streaks of fine particles carried to the top of the ice layer by escaping gas. Gas flow under the ice as the ice sublimates – changes from solid to gas – from the bottom destabilizes the sand on the dune, and causes the sand to avalanche down the dune slipface.
The third image shows the resulting changes revealed the following summer after the frozen layer of ice was gone. Comparison of the middle and lower images shows the correlation of seasonal activity with locations of change of dune morphology.
The emphasis here is on the avalanches down the side of the dune. But these are – hardly surprising – gravitational effects, even under the relatively low gravity of Mars; what is perhaps more surprising is the brief mention of “new wind ripples on the debris apron.” Conventional wisdom and standard Martian climate models had held that wind speeds on Mars were rarely adequate to cause sand movement; measurements from landers confirmed relatively modest winds – and yet sand grains have accumulated on the deck of Spirit, the stuck rover, and now we see dynamic ripples.
Enter, firmly in the footsteps of Ralph Bagnold, Jasper Kok, an atmospheric physicist at the National Center for Atmospheric Research in Boulder, Colorado (previously at the University of Michigan). Last year, he published the results of his work on sand transport on Mars and the key roles of different threshold velocities. He pointed out that the focus of Martian modelling had been on the fluid threshold, the velocity required to start saltation, but that little attention had been paid to the impact threshold, above which any saltation already happening would be sustained. His work demonstrated that “"While it is very difficult for the wind to lift sand grains, once the wind does become strong enough to start blowing sand on Mars, the sand will keep bouncing, even when the wind speed drops by up to a factor of 10." Because of the thin atmosphere (and coupled with the low gravity), while Martian sand grains need hurricane-strength wind speeds of 150 km/h to start moving, they will keep bouncing over the surface at wind speeds of just 15 km/hour. The conspiracy of the difference in parameters between Earth and Mars means that the fluid thresholds are little different – but the impact threshold is far lower of the red planet: windblown sand processes are alive and well – and, now, observable. Take into account different grain sizes and differing saltation trajectories, and Kok’s work (see references at the end of this post) also begins to explain the smaller dunes apparently typical of Mars, and the complex relationships between sand and dust movement, not to mention dust devils.
So, once again, conventional wisdom is out the window, but the wisdom of Ralph Bagnold endures; in Kok’s papers, the bibliography includes citations of Bagnold’s work from seventy years ago – how often do you see that?
[Read Jasper Kok’s papers here and here, and reports of his work at Physics World and Wired Science; for summaries of the recent HiRise sequential imaging, see, for example, Mars Daily and Wired. Images and more at http://hirise.lpl.arizona.edu/]