Return of the Flat Slab
Hollywood, it seems, has been in an exceptionally strong recycling mode, with reboots or re-envisionings or extensions of hits (and not-so-hits) past (think Star Wars, Star Trek, live-action versions of classic Disney animations, etc.). So maybe it isn’t a surprise that we are seeing the same thing to some degree in tectonics.
There are two problems manifest in the late Mesozoic to early Cenozoic geology of western North America that face those of us interested in understanding how everything was made: the creation of the southern Rocky Mountains in the Laramide orogeny, and the assembly of exotic terranes into western Canada. That these two events overlap in time inspires many to look for a common cause. So let’s do a quick review and update and see where we stand. (“Quick” by the ways, does not mean short–sorry).
First let’s tackle the exotic terrane mess in Canada. There is agreement that much of what is now in British Columbia, southwestern Yukon and southern Alaska came from somewhere other than pre-Jurassic North America. There’s a fair bit of evidence that many of the pieces were jostling near one another by the Jurassic. And things had pretty well closed up against North America by the Eocene. It is the history between the Jurassic and Eocene that fetches up troubles: paleomagnetic studies of early Cretaceous rocks place most of these rocks south of the US border near modern Baja California (in North American coordinates), but many geologic relationships suggest things were pretty close to their modern positions by the early Cretaceous. Paleomag has had to deal with inclination flattening in sedimentary rocks, paleohorizontal problems in intrusive rocks, and secular variation averaging in volcanic rocks, all of which have biases that will skew towards lower paleomagnetic inclinations and this suggest a greater southward displacement. Geologic ties suffer from a great deal of along-strike similarity of depositional environments. The Gravina belt sediments, for instance, have been taken to represent suturing of the Insular Superterrane (stuff along the western Canadian coast) to the Intermontane Superterrane just inland, but at least one recent paper allows that the eastern part of the “basin” probably formed a great distance from the western part, with the two being juxtaposed at a later time. The introduction of detrital zircon analyses was supposed to get this all fixed, but in fact initial applications could be interpreted both as confirming Baja-BC and refuting it.
There are signs this long-standing discrepancy is starting to close. Most Cordilleran geologists are aware of the large strike-slip faults present in western Canada; S. J. Wyld and others (in Geol. Assoc. Canada Special Paper 46) compile and restore recognized strike-slip faults to push much of western Canada south as far as northern California. In the same volume, Enkin’s contribution seems to move from the 3000 km displacement range down towards 2000 km as paleomagnetic results are winnowed and cleaned. Detrital zircon studies are getting more focused: a presentation by Brian Mahoney and others at the 2016 GSA Meeting stood out for being a multipronged attack on the placement of the Nanaimo Basin, a Late Cretaceous collection of sediments in part overlying the Insular Superterrane; their work seems to get around the along-strike issues by not only looking at detrital zircons with a limited north-south source but examining coarser materials present in these sediments. In this case, they have strong ties to Idaho and Montana (however, it is worth noting that Dumitru et al. have suggested materials from near this source area were transported all up and down the Cordilleran margin at this time, so this too might run into problems).
Another, rather gleeful, presentation at GSA, this by Basil Tipoff (and hats off to Basil who, as a co-convenor, could have put his talk earlier than the last talk on the last day) pointed out the pivot between the exotic terranes to the north and the Laramide orogeny to the south exists somewhere in Idaho, where he and others have documented profound Late Cretaceous deformation that, he argued, reflects northward motion of the exotic terranes impinging on a salient of North America. Although GG will dispute his interpretation of the significance of these observations, the point that this deformation must be incorporated into whatever story gets created is well taken.
There are still many things that are perhaps under-appreciated or under-discussed. In all this mobilist chaos, the Sierra is often taken to be unmoved. Admittedly there is paleomag supporting that, but it has not been remotely as well vetted as that to the north, and the presence of fairly large dextral faults during the culminating stages of Sierra magmatism ~85 Ma indicates that at least the Great Valley and associated Franciscan melanges were farther south than today; the likely presence of some strike-slip faulting to the east would move all of this more to the south–probably not greatly, but perhaps enough to get out of the way of the terranes from the north. Another point that Cordilleran geologists know but often try not to invoke is that the shattered pieces of crust associated with a single terrane might well have moved quite a bit relative to one another; assuming relative positions to be fixed both in distance and polarity might well be producing inferences that are not required by observation. This is in part illustrated by the messy pre-Cretaceous geology derived from restoration of Cretaceous and Cenozoic strike-slip faulting as done by Wyld et al. mentioned above; this indicates (as the authors note) that there was some scrambling of things prior to that series of faults. And finally, we have little direct evidence of what oceanic plates were present. An extreme example of what might be possible was offered by Wilson et al. (Marine Geology, 1991); parsimony might not be appropriate in this environment.
Enough of the problems for Canadians and let us look again at the Laramide.
The basic problem of the Laramide is getting sufficiently high stresses far into the continent to create the basement-cored uplifts found in Wyoming and Colorado. Associated with this is the requirement that the continental lithosphere not be totally removed; enough has to remain to satisfy geochemical observations and xenoliths showing that pre-Tertiary lithosphere survived the Laramide across much of the southwestern U.S. The initial inference was that a subducting slab flattened and was coupled to North America; the basal traction is balanced by normal stresses farther inboard, thus creating high stresses in the interior while not producing mountains to the west, in the Colorado Plateau in particular. A numerical simulation of this by Peter Bird in 1988 predicted the wholesale removal of mantle lithosphere to the west of the Rockies; the failure to agree with observation has led to other ideas on how to solve this. Even if the particular parameters Bird chose were off some, the feeling was that hydration of the mantle as the downgoing slab dewatered would prevent significant shear stress from being transmitted into the overriding lithosphere. Mechanically, it seemed, the flat slab as the generator of force was dead.
A recent paper by Whitney Behr and Doug Smith seems to show that the burial of this idea might be premature (not that a lot of people acknowledged its burial; the flat slab has been a zombie hypothesis for quite awhile). GG has yet to read this in detail, but the gist of this is interesting: basically, analysis of hydrated and deformed xenoliths from the lithosphere under the Colorado Plateau indicates that these rocks could have had effective viscosities far higher than many of us would have guessed (up to 1023 Pa s). They infer that instead of the weakest components determining the strength of the rocks, the strongest component (“wet” olivine) determined the rheology, and with the cold temperatures (below 750° C) present, a stiff mantle lithosphere could have been present despite the effects of hydration. Furthermore, they infer differential stress could have exceeded 400 MPa from the fabrics in the rocks. Somebody probably needs to revisit Peter Bird’s old calculations and see what happens if you put this material in the mantle; Bird’s maximum viscosity in the mantle was 3 x 1024 and his maximum basal shear stress was 100 MPa, so perhaps this changes nothing except to add some evidence for direct contact of the slab and mantle lithosphere at some time under the Colorado Plateau.
Anyways, another part of Tikoff’s presentation seemed to be resurrecting some variety of a collision hypothesis he championed some years back. His title, though, seems to auger a renewed fight: “The recurring myth of Farallon flat-slab subduction” (it is a title GG can wholeheartedly agree with even if disagreeing on what replaces it). Perhaps the quantitative observations of Behr and Smith will bring back a more substantive analysis of flat-slab physics that might finally push this idea into permanent mythdom or bring it back from the grave.