October 22, 2019

Google Earth and a simple model explain a weird pattern seen in LiDAR hillshade

Posted by larryohanlon

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

High-resolution surface imagery derived from LiDAR data reveals landforms and geologic features in stunning detail, but sometimes the forms and patterns that appear are somewhat unusual looking if you don’t spend your time studying landscape evolution and structural geology. A particular feature in Pendleton County, West Virginia recently caught my eye because I think it looks exactly like a set of frog legs. From one “knee” to the other, the feature is about 1,300 ft (380 m) across.

This visual association may be a reflection of my southeastern US origins, so I offer the following edited image to inspire the imagination:

Real talent, and a damn big frog. If you Google “Face on Mars,” you will find a similar body reconstruction for the famous Martian feature. Actually, here is a link.

Is this a bizarre and gigantic fossil discovery? An unsolved mystery akin to the face on Mars? Unfortunately, this is just another set of compressional folds within the Appalachian Valley and Ridge, but they do stand out in the hillshade due to their interesting topographic pattern. While I can tell what this feature really is because of my structural geology background (I will still always see frog legs, regardless), it makes a good example of why I like to turn LiDAR hillshades into .kmz files for use in Google Earth. The Google Earth interface allows the image to be tilted and rotated so the 3-D context of the “frog legs gorge” can be appreciated. Google Earth is also free and fairly ubiquitous these days, so I can email the .kmz overlay to anyone. The above image is taken directly overhead, in plan view. The image below shows how things change when the view is rotated and tilted to look at one “leg.”

Viewed from this angle, the “leg” is shown to be a pair of folds affecting a few prominent sandstone layers encased by shale above and below. This type of structure is exactly what you would expect to find in this portion of the Appalachian Valley and Ridge, but you really won’t ever see it in this way without high-resolution digital land surface imagery. The feature is also too large to fully appreciate from the ground, but too small to be a major topography-affecting structure at map scale.

Once you can see the folds involved, the “frog legs” shape is easy to recreate with a clay (or Play Doh) model. Creating the anticline-syncline fold pair and cutting a V-shaped valley through it gives a very nice explanation of how such a strange and eye-catching landscape pattern can form. All that is needed is different colored Play Doh (strongly contrasting rock types in the real world) and a way to cut the “gorge” in the Play Doh (a steep creek in the real world.)

The finished product, produced by making a Play Doh fold set and cutting a V-shaped “gorge” into it.


The V-shaped cut into the block


Tilted view, from the side. The cross section shape is apparent here, as seen in the hillshade a few pictures up.


Cross section view of the outside of the block.

The Google Earth interface allows you to do something similar with the overlay, at least in terms of viewing it from different angles to get an idea of what causes the frog legs pattern to occur in plan view. Some thing is lost in trying to narrate this, so the following video link shows what the overlay looks like in Google Earth and how changing perspective helps create context for the fold structure and gorge that produce the “legs.”


The frog legs gorge is developed into Silurian-age sedimentary rocks of the McKenzie Formation and Clinton Group, which includes both sandstones and shales. Shales above and below the strong sandstones allow the sandstones to create folds with very small wavelengths in the context of the greater Valley and Ridge fold-thrust belt. The Google Earth geologic map for West Virginia can be opened in conjunction with the hillshade to determine the formations involved in the frog legs.

This is actually a very minor feature in the Valley and Ridge fold-thrust belt; the sandstones that form the “leg” pattern are not even mapped as a distinct unit here. Even so, the feature is big at the human scale. Geo map from here.

The terrain model that comes with Google Earth (if you have Terrain turned on) is also useful for getting a sense of the size of the feature, how steep the gorge walls are, and whether or not you would appreciate the “legs” if you were field mapping in the area. The gorge is not tremendously deep, but it produces about 500 ft (150 m) of relief between the stream and the surrounding hilltops. If you walked this stream while mapping, you would likely feel less than inclined to follow the sandstone bed up the slope, particularly with the amount of vegetation, talus, and timber rattlers that are likely to be there.

Both images are the same perspective and scale; the overlay is just turned off on the bottom image. If I were mapping this area, I’m not sure if I would chase the sandstone bed up the hill or not.

While the frog legs are hundreds of feet across and hundreds of feet tall, they are actually an incredibly small structure within the greater Valley and Ridge. The folds that make the “legs” pattern represent deformation in a sandstone sequence that is disconnected from the layer sequence above and below by shale, allowing small wavelength folds to form.

The frog legs are a tiny structure in the grand scheme of things…the tiny black line showing the scale of the folds is almost hard to see here. The Will Mountain Anticline creates a valley 3.2 miles across; it involves kilometers of Paleozoic sedimentary rock (Cambrian to Siluro-Devonian at this erosional level. The frog legs fold pair is only 1,100 feet across, and affects only a few beds within the Silurian part of the section. Frog legs are at 38.4915N 79.473W.

The Wills Mountain Anticline, immediately west of the frog legs area, involves the whole Paleozoic sedimentary section (or what has not yet been eroded away at this point). The valley created by differential weathering and erosion of carbonate rocks in the core of the anticline is just a bit over 3 miles (5.1 km) across; by comparison, the exposed portion of the anticline-syncline pair that creates the frog legs is only 1,100 feet across. This comparatively small fold is, however, positioned on the limb of a larger, kilometer-scale fold that is still difficult to discern in the landscape because most of it remains buried. More on this in a follow-up post to come…

This is all supposed to be a bit funny, but I genuinely do consider this a very good example of why I process hillshades into .kmz files. You can certainly view them in 3-D in Arc and other applications, but I think the ease and availability of Google Earth is hard to beat. A few key points from this cool Appalachian feature:

–Good surface imagery puts a whole new spin on geologic mapping. Even though the frog legs don’t involve several formations, the ability to see the fold pair in such crisp detail is useful for understanding structural style within the Silurian section in this area.

–Being able to see the small size of the fold pair gives some context for its place within a large-scale cross section. Folds this small don’t involve much section, so you can’t project them downward. This means the sandstone layers involved are detached from section above and below by shale layers.

–I think water well drilling in this area might be interesting. The type of structure that makes the frog legs also affects the whole Siluro-Devonian sandstone section (albeit at a larger wavelength due to the greater thickness). In valleys where this unit is still buried beneath Devonian black shale, I think water wells might hit the sandstone and associated carbonates in some places while others would remain in undesirable black shale nearby. Knowing the scale and style of folding in the sandstone horizon could be useful in siting wells, as the sandstone folds do appear extensive along strike, allowing their position to be approximated even if they are still buried.

This post was originally published on The Geo Models blog