January 23, 2020

Discount inverted basin models: You get what you pay for (mostly)

Posted by larryohanlon

By Philip S. Prince, Virginia Tech Active Tectonics and Geomorphology Lab

Inversion of extensional basins seems to be a very popular subject within the geologic literature. A basic Google search of “inversion geology” will produce a tremendous number of results, including conceptual illustrations, analog model results, and actual cross sections generated from subsurface imaging and drilling exploration. A particular point of interest in the overall basin inversion conversation is the role that existing normal faults play during compression of the basin. The conceptual images below emphasize reactivation of the normal faults at the margins of the basin as reverse faults during compression.

Both images sourced here. The first model shown in the video (cross section labeled with “steep limb” a few images down) resembles the geometry immediately above, with asymmetric anticlines developed at the margins of the inverting basin.

I have messed around with various attempts at illustrative models of the inversion process, and a common takeaway is that the pre-existing normal faults are too steep to move as thrust faults during compression. The result is “shortcut” faults which initiate at the base of the basin and cut through the basin walls, reaching the surface well outside of the margins of the basin and passively uplifting it in the center of a pop-up structure. The image below from this very cool paper illustrates the concept nicely.

Click the link just above this image to get to the source paper. It’s a really cool read, and the clay-cake models are particularly impressive. The image above shows how the fault geometry within the basin of a dry sand model doesn’t do much at all during inversion. The models made for this post were intended to see if making the basin fill mechanically variable and distinct from surrounding material altered this behavior. Overall, my results are minimally different from what is seen here.

I wanted to see if filling the extensional basin with a granular material mechanically distinct from the surrounding “basement” would alter faulting during the early phases of inversion. I produced a few small, extremely simple (hence “discount”) models using a mixture of microbead and slightly cohesive sand layers for growth layers in the basin. The models used focused extension, meaning a rigid baseplate was moved to deform the overlying layers (see caption a few photos down). The goal of this setup was to film the whole basin growth and inversion process under low angle light. Once the model was complete, screen shots from the growth phase could be laid under screenshots of the early stages of inversion to compare normal fault and reverse fault scarp location. The GIF below shows the basic idea.

Locations of the reverse/thrust scarps seem to be more than coincidentally related to normal fault location. Obviously it would be great to run this model in a box with a clear side wall, but I think wall drag would skew results, particularly where features may be extremely subtle.

The video linked below shows the extension and compression phases, sped up significantly. Black lines mark normal faults at the end of the extension; these are superimposed on the inverted models to compare scarps to underlying basin geometry. I think there is a relationship, but slicing models does not indicate it is one of pure fault reversal.


These models were only inverted a modest portion of their extensional displacement; in other words, a model extended 4 cm was only inverted 2 cm or even a bit less. If the models are inverted anywhere near 100% of the extensional displacement, big thrusts will cut the whole section and the passive uplift scenario will develop. I tried to stop short of this to emphasize what happened to the mechanically distinct (and internally varied) basin fill just as inversion began.

The actual cross section views of these models are particularly disappointing, as the magnitude of inversion is too low at the physical size of the models to make compressional features readily obvious.

Using focused extension above a moving rigid baseplate makes short, squat basins. Distributed extension using an elastic sheet beneath the section makes a different geometry, and reversing the distributed motion makes an equally distinct inversion pattern. I tried to keep it as simple as possible with the focused setup here. In this case, the side of the model with three normal faults moved to produce inversion (the right side of the basin moved; the left side was fixed).

I think the videos of inversion show that normal fault geometry may indeed impact initial reverse fault position, but no clear offset or fault reversal motion can be seen in the model sections due to their size, limited deformation, and my poor color choices (each of my sand colors is mechanically distinct, so I do have to choose carefully).

This trial (2nd in video) produced three distinct scarps above the end of the basin with several normal faults. While obvious from the top, they are essentially undetectable in the cross section. Note that in this trial, the side of the model with the single normal fault moved to produce inversion.

Numbers match the surface model above.

Faults are mostly conjecture, although beds do seem to be rotated counterclockwise beneath “1.”

The basin walls clearly act as rigid buttresses during the initiation of inversion, but I think the faults that reach the surface above the basin walls start in the core of the basin and deflect onto shallower dips moving upward, as marked below.

One cool aspect of these models is that the basin fill can “absorb” a moderate amount of shortening without producing distinct fault structures. I think this shortening is probably visible as thickening of microbead layers, etc., but I can’t really see its effects in the section. The cross sectional deformation of the models is also quite disappointing compared to the appearance of the model surface during inversion.

This trial is not shown in the video, and shortening was kept at a minimum. Note that the basin does not extend to the base of the grayish “basement” material. This model is a good example of how some dry granular media “abosrb” shortening without producing faults; the shortening creates distributed thickening of layers, etc.

I think a very large-scale version of this setup might produce interesting small-scale structures around the basin margins during inversion, where the basin fill tends to produce anticlines above the margins of the basin. Using lots of microbeads to create an overall weaker basin fill seems to create more small-scale structure (1st and 3rd experiments in the video), and a model several times the size of these might make the small-scale structures apparent. The 3rd experiment in the video (screenshot below) would be a good candidate for scaling up; its basin fill was nearly pure microbead with a few marker layers. This model also developed a thrust through the entire basement section as shown near the beginning of the post; its scarp is well away from the basin margin.

Scarps suggest the inverted basin fill would have made a cool outcrop pattern here.

There is enough structural relief in the models to create an outcrop pattern in the model slices. The first section below was cut to a flat surface, and the second had a “landscape” carved into with post-extension beds creating a caprock appearance to show dip of layers away from the basin core.

At the end of the day, I think the distributed extension/shortening setup is better.

Top view. The dip of the purple and yellow beds away from the basin core is apparent in the “V” shaped outcrop pattern produced when a tiny valley is carved into them. Analagous “V’s” can be seen around the rim of the Weald below.

Inverted basin systems occur around the world, and some very interesting examples are found on the southern coast of England. The topographic outline of the Weald is particularly cool, and there is no shortage of material written about it. As stated above, I think a distributed extension/shortening setup makes a better analog to the Weald, but some aspects of the models shown here offer a bit of context for Wealdian topography and rock type distribution.

Image sourced here. The strong diamond-shaped outline of the Weald is produced by resistant layers dipping away from the anticline core, around and just south of Turnbridge Wells.

This post was originally published on The Geo Models blog.