Sampling dune sand

Back to main page

Drag the 'scoop' in the figure below to 'take a sample' of the dune's slipface sand.

Grain diameter (microns)

Include a surficial pinstripe?

Fig 1. Histograms of grain-size distributions in full scoop (white) and partial (blue).

The histogram shows the grain-size distribution of dune sand in the partial scoop (blue) compared with the distribution in a 'full scoop' (white). This illustrates that distribution measurements can depend on how dune sand is sampled.

Although the slipface and dunes look homogeneous, there are actually textural variations created by avalanching and wind, as illustrated in figure 2; layers with centimeter-scale thickness, separated by millimeter-scale pinstripes (the darker lines in figure 2).

These layered, textural differences can give rise to the sample-dependent variations illustrated by the model.

Fig 2. Grainflow/pinstripe layers created by an advancing slipface (from Aeolian dunes and sandstone: Overview).

How the model works

Fig 3. Grains (black) from a thin section of Navajo Sandstone. The section exposes three pinstripes; the top-most pinstripe is delineated by yellow and blue lines. The figure is tilted to align with the slipface slope in figure 1, which uses data obtained from the thin section by particle analysis. The model's dune surface is aligned with either the yellow or blue line, depending on the checkbox beneath figure 1.

The data for figure 1 comes from a particle analysis of a thin section of an outcrop of Navajo Sandstone, shown in figure 3. Three pinstripes (layers of finer grains) are visible in figure 3. The upper pinstripe is delineated by yellow and blue lines. The top part (greyed out) is not used by the model.

As the scoop goes into the dune, a corresponding fraction of the thin section data is included in the blue histogram of figure 1. When the scoop is fully inserted, the sample includes two pinstripes (or three, if the surficial pinstripe checkbox is selected) and three 'between pinstripes' layers.

The model can simulate the dune surface being at either of the two lines (yellow and blue) in figure 3, controlled by the checkbox at the bottom of figure 1. Of course the distribution in a shallow sample is significantly affected by whether there is a surficial layer of finer grains (eg., via grainfall or traction).

The length of the scoop is a bit misleading; the thin section data comes from only about 3 cm of rock (which, before compaction, was about 4 cm of dune sand). The full stroke of the scoop is mapped to that 3cm length of the thin section data.

The grain diameters are derived from grain areas, ie., are 'round grain equivalents', with a low-end cut-off of 10 microns. The grain-size distribution from the thin section is probably short a few larger grains; the many voids (white) within the areas between pinstripes are probably left by grains that pulled out during production of the thin section. That's not a problem for this web page, as the actual distribution isn't important; the point being that textural inhomogeneities have an effect and thus should be considered when sampling.

Related pages

Method: Making pinstripe laminae visible. A quick, easy way to get a feel for textural differences within dry sand.

Video: Dry sand grainflows/avalanching, and
Video: Lee-side processes, Coral Pink Sand Dunes. These videos give a feel for what happens on the lee side of a dune. Even a single grainflow layer may have a complex internal textural structure.

Particle analysis: Obtaining representative subsamples. Getting a representative sample into a bag is just the start.