July 10, 2019

Using sand models to explain the concept of geologic mapping

Posted by larryohanlon

By Philip Prince, Virginia Tech Active Tectonics and Geomorphology Lab

Geologic maps can be very visually engaging, but non-geologists may find it difficult to extract the information that a map is supposed to communicate. Without at least some experience in visualizing Earth process and how subsurface features intersect with the land surface, a non-specialist user may end up enjoying the aesthetic qualities of the map and not much more. Looking good is certainly important to make the map document appealing to users, but I personally want folks to appreciate both the map’s appearance and what it says about sub-surface interpretation, which is a focus and skill set unique to geology. Cross sections included with a map can help, but it can still be tough to pull it all together if you don’t look at this sort of material all the time. The Virginia state geologic map (1:500,000 scale) shown below as a Google Earth overlay is a good example.

This overlay was much more appealing with its previous colors. The red arrow is for comparison to the model setup image a few pictures down–it indicates Alleghanian movement of the Appalachian foreland basin against the fixed (relatively speaking) “backstop” at left to create the Valley and Ridge sedimentary fold-thrust belt. The “backstop” is today’s metamorphic and plutonic igneous Appalachian Blue Ridge (green colors at left) and Piedmont (just out of view to the left); the blue and tan zone is folded and faulted sedimentary rock accreted against the backstop.

There are some neat color patterns here (they were better before the colors recently changed, but whatever), but their significance as an expression of geologic process and history, as well as how the colors imply what sort of structures are under the land surface, aren’t exactly spelled out here. One of the reasons I enjoy physical models so much is that they can bridge the gap between fixed visual patterns and the ability to visualize process and motion. While sand models are typically focused on cross section alone, it’s possible to make a “take-apart” geologic map by deforming a sandpack, eroding it as desired, and then gelling and slicing it up. I think it makes for a good visual reference that allows someone to hold a piece of a geologic map in their hand. The outcrop pattern shown below looks interesting and bears resemblance to real outcrop patterns in fold-thrust belt settings, but figuring out what causes the patterns is not immediately obvious unless you’ve spent a good bit of time around geology.

A homemade geologic map that you can hold in your hand. The red arrow indicates the same movement setup as the first image of the post. At left, just off the image, is a fixed backstop. The layer pack is pulled against it, right to left, causing the layers to thrust fault and fold. Faults are marked in the bottom image, with teeth to hanging wall…but the question remains–what’s underneath the cool surface pattern?

Once the model is gelled and sliced, however, the connection between surface patterns and subsurface structure becomes more tangible.

Cutaway view. Now the surface color zones can be directly connected to subsurface features that result from brittle deformation of a layer sequence.


The other side of the model, with the very different outcrop pattern…it results from very different subsurface structure.

This model does not scale particularly well to a real world feature set, although it broadly represents the transition from crystalline (metamorphic or igneous) thrust sheets (the gray stuff that is the deepest layer) to sedimentary thrust sheets (yellow, white, and blue) of a fold-thrust belt. This is the general zone of the Appalachians shown in the first image in the post. The model does get some overall ideas across, such as the relationship between fold wavelength, the thickness of the layer sequence involved in folding, and outcrop zone width.

I made this model while experimenting with different positions for decollement layers, the weak slip layers that allow the different thrust sheets to slide on top of each other. All portions of the model were shortened the same amount and at the same rate, and all of the sand used is the same (meaning yellow sand on one end of the model is the same yellow sand on the other, etc.). The different faulting patterns are controlled by different decollement positions. The side of the model with broad outcrop belts has a deep decollement and a shallow decollement, causing the yellow and white sequence to stack on itself and create a broad ramp anticline and two high-displacement thrust faults.

Placing a decollement (slip layer) below yellow and above white causes the entire yellow and white sequence to stack on top of itself. Flattening of the fault on top of the white layer (not much flattening, but enough) just barely produces a broad ramp anticline (the forelimb does dip to the foreland, but only slightly) that creates a huge outcrop footprint. Note that yellow-white has stacked on top of itself twice in this portion of the model.

The portion of the model with the more complicated outcrop patterns has a deep decollement and a second decollement on top of the yellow layer. This causes the yellow layer to stack on top of itself, while the white and blue layers are scraped off the top of yellow, forming their own thinner fold belt. The result is a lot of thrust faults, all of which have smaller displacements, and shorter wavelength folds resulting from faulting in thinner packages of layers.

More faults, each with less displacement. Collectively, they accommodate the same amount of movement of the layer pack as the fewer faults with higher displacement in the previous image. 2 faults with 5 units of displacement and 5 faults with 2 units of displacement both give you 10 units of displacement–that’s the idea shown here. Note that the blue and white layers were faulted before the underlying yellow layer faulted to create the frontal structure.

The model setup is basic. Layers are poured onto a broad strip of paper on a flat board. The paper moves under a fixed backstop to which a “pre-wedge” of sand is added to initiate the wedge shape without requiring it to grow through large amounts of shortening. You could say the pre-wedge represents a large crystalline thrust sheet that becomes “indenter” against which the sedimentary layers fault and fold. As the wedge grew, I scraped material away (erosion) to maintain a steady wedge shape and maintain some consistency with the processes active on an actual fold-thrust belt undergoing erosion during its growth.

Pulling the paper strip (the white thing dangling out from underneath the backstop) shortens the layer pack and forces thrust faulting. The layer pack would move right to left, the orientation indicated by the red arrows in the images above. The backstop and pre-wedge represent thicker crystalline sheets of a growing mountain belt; the colorful layers represent sedimentary units of the foreland basin and any underlying sedimentary cover on the continental margin that can detach and slide freely.

Note that the faulting style interacting with erosion to limit the thickening of the model controls where portions of the “pre-wedge” are preserved. In the portion of the model where the whole yellow and white sequence stacks on itself (top slice in the picture below), uplift is rapid and widely distributed as the thrust sheets move up their large ramps. This substantial, widespread uplift coupled with steady erosion has caused the pre-wedge to be almost completely eroded away. Note also that this decollement/fault pattern causes the white layer to be pulled deep into the thrust stack because it is below the upper decollement; the yellow layers are thrust on top of it.


Three slices from the model show the different structural styles that correspond to the different outcrop patterns. Numbers on the sections in the top image correspond to section lines shown in the bottom image. In section 1, the top decollement is above the white layer, meaning the white layer can easily be overthrust by the deep yellow layer. In sections 2 and 3, the top decollement is deeper, below the white layer. This allows the white layer to be pushed out of the way and ride above the slices of yellow layer that stack beneath it.

In the portion of the model with the larger number of faults, substantial but localized uplift occurred where slices or duplexes of the yellow layer stacked up against the backstop (left end of slice 3, above). Uplift due to this underplating caused the pre-wedge to be eroded away on top of the growing duplex stack, but a good chunk of it remains intact ahead of the stack where the wedge only thickened by one slice of yellow. Also note that the white layer doesn’t get pulled into the deeper guts of the wedge in this case. The decollement beneath it allows it to be thrust out of the way of, or to ride above, the yellow thrust slices.

Geologic survey mapping done at the 1:24,000 or even 1:100,000 scale would only address small surface portions and thicknesses of a system like the one represented by the model. Even so, I think this type of model provides a way to show what geologic mapping–and the cool map patterns–are all about.

This post was originally posted on The GeoModels blog