
The outer cliffs in the image are Jurassic-aged Kayenta and Wingate sandstone, and the red talus slopes are formed of debris eroded from them. The grayish-whitish slopes toward the center are Triassic-era Chinle and Moenkopi Formation rocks; the brown rocks in the center are also part of the Moenkopi. In the mishmash of rocks in the center are small exposures of Permian White Rim sandstone, the oldest rock exposed here. The White Rim sandstone has undergone a lot of violence; in essence, the stratum has been broken into little pieces and then reglued together.
So the obvious question is: what did it? What pushed up the rocks so much, just in this one specific spot, and deformed them so greatly while doing so?
The prevailing opinion for most of the 20th century was that Upheaval Dome is the eroded-away remnant of a salt diapir; salt domes and other salt-related structures are very common throughout this part of eastern Utah and western Colorado. The idea was that a plume of salt had pushed completely through the sedimentary strata here, bulging the rocks upwards and greatly warping the rocks at the center of the plume. Then, erosion stripped away the overlying layers of rock and salt, exposing the rock layers disrupted by the salt plume's passage. There have always been some problems with the salt-origin hypothesis. None of the other salt structures in the region resemble Upheaval Dome, in size, shape, or geology; rather than being eroded away, the salt-bearing strata responsible for creating domes and diapirs appear to be at least 1800 feet below the surface of the Dome, and show no evidence of having produced a plume here; and there's no evidence in the center of the Dome of any, like, salt.
The possibility that the Dome might have been created by other forces was proposed in the 1930's, but evidence for those ideas was not seen as very compelling until the 1980's and 90's. The growing recognition of the large number of impact crater sites on and below Earth's surface has lead many geologists to reinterpret the Dome as being the remnant of an impact crater, and there are features of the Dome that are more characteristic of impact craters than other explanations of its origin, such as the presence of a distinct magnetic anomaly in the area, (rare) blobs of melted and recrystallized sandstone, and unusual fracture patterns in the rock that are commonly found in rocks that have undergone impact shock.

Here's a closer view of the center of the Dome. Notice how the strata are folded nearly vertical in many places. Harder to see (curse my amateurish photography!) are the pinnacles formed by these warped strata.
Most of these notes shamelessly cribbed from Kriens, Shoemaker, & Herkenhoff, "Geology of the Upheaval Dome Impact Structure, southeastern Utah" (yes, that Shoemaker). A more general overview can be found here. The best images I've found of Upheaval Dome, bar none, are here.
Update: More links! Good images, diagrams, and descriptions here, close-ups of the central rocks here, Google Map satellite image here, more Serious Geology with some nice images here.
When I lived in southern California, someone was advancing the theory that the mountains north of L.A. and the rocks along the coast (Palos Verdes, for instance) had at one time been right next to one another instead of thirty or forty miles apart, and that some sort of seismic action had basically turned a whole lot of interior rock inside out and forced it to the surface, creating the Los Angeles basin as we know it.
Which doesn't explain this, necessarily, but does suggest that immovable objects are a lot more movable than we think.
Posted by: CGHill on October 1, 2006 03:28 PM
Mmmm...looks like chocolate fudge swirl ice cream.
Also looks a lot like bits of Death Valley, but I couldn't find a good picture of the place I'm thinking of.
Presumably, if it took this long to get people to think of the dome as an impact crater, it must differ from structures that are clearly impact craters. How does it differ? How long ago was this impact supposed to have occurred?
Some cool aerial photos here.
Posted by: Angie Schultz on October 2, 2006 03:55 PM
Fascinating.
Posted by: Jonathan on October 2, 2006 09:27 PM
A-
That bottom image is very cool indeed.
How does it differ? I think the central uplift of strata is unusual, especially for small craters. (One of the points of the Kriens, Shoemaker, & Herkenhoff paper is that once impacts exceed a certain size, this sort of central 'rebounding' is to be expected.) This sort of strata-bending is pretty common around salt features, though. Obvious impact products (fused quartz and tektite-like objects) are apparently very rare.
But I think the main reason the impact hypothesis didn't get much traction until the 80's is mainly that the salt diapir hypothesis was just 'good enough'. It seemed plausible to most geologists, because there are lots of salt-related features in the area, and I don't think that until people like Shoemaker and Alvarez really started promoting the idea that the earth has been hit – a lot – by bolides, that people started looking for evidence of impacts and re-evaluating old evidence. Even Meteor Crater wasn't generally accepted as being just that (a meteor crater) until the 50's or 60's.
(That's just my guess - I'm no historian of geology.)
Posted by: David Fleck on October 2, 2006 09:28 PM
...once impacts exceed a certain size, this sort of central 'rebounding' is to be expected.
Right. Take for example Mimas. There's an uplift in the middle of the crater. But that was a big impact! Perhaps you don't need such a big impact if it occurs in softer rock.
Posted by: Angie Schultz on October 3, 2006 09:43 AM