Australia: The Land Where Time Began

A biography of the Australian continent 

Scraped by flat bed Subduction

Material that is scraped off the base of the continental mantle lithosphere builds a migrating keel. Features of the Laramide Orogeny are recreated by this testable mechanism for flat subduction.

The oceanic plate descends below an overriding continental plate from a deep-sea trench and down into the mantle in most subduction zones. The oceanic plate descends at an angle, and the dip of the slab usually increases from a few degrees at the trench to 30^o or more at depth. The slab dip can reach extremes of nearly vertical or horizontal. In cases where the dip is almost horizontal, after reaching 20^o to 30^o, there is decrease of the slab dip and the subducting plate proceeds for several hundred kilometres almost horizontally, sliding along the underside of the overriding plate (Gutscher, 2000). Axen et al. (2018) present numerical simulations that show that during the horizontal phase, the flat slab scrapes off 20 to 50 km of the overriding continental mantle lithosphere.

The landscape and geology of the western United States has been strongly affected by an episode of flat-slab subduction that occurred during the Laramide Orogeny about 80 – 50 Ma (Liu & Currie, 2016; Dickenson & Snyder, 1978).  Migration of arc magmatism 1,500-2,000 km in extent inboard from the trench and later retreat back towards the trench provides support for Laramide flat subduction. The distance between the volcanic arc and the trench of 200 to 300 km is typical for normal slab dip, with the arc forming when the slab reaches a depth of about 100 km. Arc magmatism about 1,000 km from the trench, therefore, indicates a slab dip that is extremely shallow that is achieved only by flat subduction. Also, tomographic images of the upper mantle beneath North America document currently subducting and fossil slabs where the lab has resubducted (Sigloch et al., 2008). Following the end of this flat slab phase, the underlying slab peeled away and sank into the mantle. This triggers widespread volcanic eruptions, pyroclastic flows and ignimbrite flare ups in the Basin and Range (Humphreys 1995), and the formation of porphyry copper deposits that are economically important.

Numerical simulations of the initiation and evolution of flat subduction have been presented by Axen et al. (Axen et al., 2018). They found that the flattened slab advances and scrapes the base of the overriding plate. The mantle lithosphere that is scraped off accumulates at the nose of the advancing flat slab, in the form of a bulldozed keel to fill the asthenospheric wedge. Convection in the asthenospheric wedge is inhibited by this filling, where water would typically have been released from the descending slab which would cause partial melting. Thus, arc magmatism ends. The model of Axen et al. is specifically applied to the Laramide Orogeny of the Late Cretaceous in North America and show that it can recreate features such as block fault uplift in the overriding continental plate.

In modern times flat subduction is relatively rare, being observed in about 10% of the subduction zones in the world, such as Peru and central Chile. There has long been debate about the causes of flat subduction. It has been contributed to a combination of factors such as slab buoyancy, kinematics of plates and dynamics of the asthenosphere that combine to produce favourable conditions (Gutscher, 2000; Dickinson & Snyder, 1978; van Hunen, 2002). Flat slab subduction beneath North America has previously been simulated by numerical models (Bird, 1988), though some of the key observations were difficult to reproduce. E.g., there is a belt of compressional structures, block fault uplifts, the age of which progresses from the Colorado Plateau in the west to the high plains, such as the Big Horn Range of South Dakota to the east. Previously, these were attributed to compressional deformation throughout the overlying portion of the upper plate that was driven by basal shear that was exerted from the flat slab below. The lack of compressional deformation above the flat portion of the slab, which has been demonstrated by active normal faulting that was observed in field studies in the cordillera Blanca in Peru (Margirier et al., 2016), proved difficult to reconcile with the basal shear model (Gutscher, 2000).

The simulations by Axen et al. (Axen et al., 2018) actually predict that behind the compressional deformation belt that is observed and modelled there is a zone in which extensional stress is elevated in the upper plate. Examination of stress field indicators above flat slab regions can test this prediction. Included among these in situ measurements, e.g., are borehole elongations during breakouts, field observations and analysis of earthquake focal mechanisms. Above flat slab zones in Central and Southern Peru initial assessment of the focal measurements for crustal earthquakes reveals a low level of seismicity in general, with 4 normal faulting mechanisms out of 6. Ongoing extension is indicated by focal mechanisms and mapped normal faults and they are therefore supportive of the predictions of the model by Axen (Axen et al., 2018). A large-scale seismology survey was carried out above the Peruvian flat slab which succeeded in imaging the deep slab, where deep subduction occurs (Scire, 2016), and the impact on the upper plate (Fitzcarrald Arch) uplift was also explored and explained tentatively by basal shear (Bishop et al., 2018). Gutscher suggested that this interpretation may need to be reconsidered in light of the Axen et al. study. The interior portion of the upper plate could be targeted by future passive seismological work and direct images of the bulldozed keel be obtained.

An elegant explanation has been offered by Axen et al. (Axen et al., 2018) of the observations that are specific to the Laramide Orogeny of western North America. Though beyond this regional significance, the mechanism that is proposed of a scraped continental mantle lithosphere and regional extension may be important to flat subduction zones more generally, at the present as well as in the past. The Peruvian Andes of the present may yet prove to be an analogue for the Basin and Range of 50-60 Ma.

Sources & Further reading

Gutscher, M.-A. (2018). "Scraped by flat-slab subduction." Nature Geoscience 11(12): 890-891.