One of the curious problems in tectonics is seeing how ancient mountains worked. You can look at the guts of such mountains and see faults and folds and such not, but figuring out the timing of these features can be tricky. In contrast, next door to many mountains are piles of sediment that accumulate as the mountains grow. Some of these basins are bounded by normal faults, which have their own problems, but a lot of them are bounded by thrusts and the basins themselves are considered to be flexural in origin: the weight of the mountains pushes down the plate, making a hole next to the load–a convenient place to store sediment being eroded off the mountains (the foredeep, a common feature of compressional mountains). Plate flexure is cute in that while the stuff near the load goes down, at some distance farther away things go up (which is called the forebulge) and even farther out the surface should go down a little (the backbulge basin).
Now where this idea works really well is in the oceans. Loads from volcanic chains make beautiful moats around them. At trenches, where ocean floor is bent down under the overriding plate, we see the forebulge and get earthquakes consistent with the development of these features. But where this idea gets applied most in geology is in trying to interpret the foredeep. Issues with this are the subject of a new Geology paper by Guy Simpson.
In this paper, Simpson argues in essence that the flexure model does a terrible job at predicting the height of mountains associated with a foredeep. When a model is run with a thrust fault being active for tens of millions of years, the basin depth derived here is greater than that predicted from the surface topography. And so the author concludes that application of simple plate flexure is, essentially, wrong.
At first you might think that this breaks a lot of earlier work. That is not so, and GG is grumpy that reviewers were not a bit more adept at disassembling this paper. (GG is really grumpy with a letter journal that seems focused on having papers that are extended abstracts for material in the supplementary materials). The problem is that the load on the lower plate as usually used is NOT the topography, is it the weight of everything added between the fault and the surface (this was probably most cleanly drawn by Teresa Jordan back in 1980 but has been true of most flexural analyses since). Thus the factor of three difference between a flexural basin’s depth and that predicted by its topography is largely because the paper here has grossly understated the load on the plate. [There are some other problems: the total shortening listed is about 100 km for the figures in the paper, but it appears there is only about 10 km of shortening suggesting that something was misstated.]
However there is an aspect of this paper of great interest if the models were in fact run correctly (the problem above is a misinterpretation, not necessarily an incorrect application of a modeling code). That has to do with the forebulge in the model with efficient erosion. Now a lot of sedimentary geologists try to identify the position of the forebulge and use that to estimate changes in the position of the load, elastic plate thickness, etc. But what has happened here is that the forebulge got wiped out early on by erosion and then was buried by the sediments from the rapidly eroding mountain belt. Now the forebulge should NOT have moved more than a few kilometers given how this paper was set up, yet it appears to have migrated substantially away from the load owing to something–perhaps this erosional issue. Now a lot of times the forebulge is not a positive but remains underwater, so if this is solely caused by erosion, perhaps this isn’t significant, but anything that messes with the forebulge is worth examination.
Too bad this paper didn’t focus on that instead of focusing on what is largely a misstatement of application of plate theory.