News & Views | Published:

Earth science

Lithosphere today ...

Nature volume 472, pages 420421 (28 April 2011) | Download Citation

Seismic images of the Colorado plateau region reveal a mantle 'drip' forming under the Grand Canyon area. This hidden process may be responsible for the puzzling uplift of the plateau. See Letter p.461

Large-scale uplift, erosion and volcanic activity are standard features of mountain belts that are subject to the horizontal tectonic forces associated with plate movements. But somehow, in the past 70 million years, possibly quite recently, the virtually undeformed Colorado plateau of the southwestern United States was uplifted about 2 kilometres, intruded by diverse magmas, and eroded and deeply incised, creating a dramatic topography that includes the 1.8-km-deep Grand Canyon. This enigmatic history has perplexed geologists for more than a century.

On page 461 of this issue, Levander et al.1 provide a possible explanation that relates uplift, erosion and volcanism on the surface to deep features on the underside of the plateau, where it sits atop the convecting upper mantle. Understanding these features and their surface manifestations is a timely challenge as we begin to recognize the critical part that deeper mantle processes must play in the evolution of the continents.

Using new seismic data from the Earthscope Transportable Array, the authors combine different seismic techniques to get a better view of the lithospheric structure of the Colorado plateau. Continental lithosphere is the strong, long-lived crust and upper mantle, extending to about 150 km depth and comprising less-dense crustal rocks down to about 40 km and denser mantle rocks below that. This lithospheric package sits on the convecting asthenosphere, the part of the mantle below the lithosphere that is hotter, weaker and perhaps less dense — a potentially unstable configuration.

The seismic images reveal an anomalous region in the lower crust and upper mantle under the Grand Canyon and underlying much of the western half of the plateau (Fig. 1). In this region, the mantle part of the lithosphere seems to form a vertically elongated viscous drip sinking into the Earth, pulling off the lower part of the crust above it, a process called convective instability2 or delamination3. Once the drip detaches and is replaced by upwelling hot asthenosphere, the remaining crust above it bobs up and is often intruded by magmas. This deep process may be manifesting itself at the surface through uplift, volcanism and erosion in and around the Grand Canyon. Levander et al.1 hypothesize that this lithospheric drip formed in the past 6 million years and is just the latest of several such events that occurred around the edges of the Colorado plateau, producing rock and surface uplift over the past 20 million to 30 million years.

Figure 1: Over and under the Colorado plateau and surrounding regions.
Figure 1

a, Low-angle topographic view of the area, which lies within the states of Colorado, New Mexico, Arizona and Utah. Dashed line, the approximate location of the plateau's physiographic boundary; blue, rivers, lakes and reservoirs; GC, Grand Canyon. b, Cross-section of the lithospheric structure along the transect shown in a. Levander et al.1 used multiple seismic imaging techniques to identify an active mantle drip peeling off the lower crust over a large area centred near the Grand Canyon. This subsurface process may be responsible for geologically recent uplift in the area. A series of such drips, occurring over tens of millions of years, may account for the regional uplift of the entire Colorado plateau.

Drips are turning up in many places. In the western United States alone, geophysicists also see them under the Sierra Nevada4, Wallowa Mountains5 and central Great Basin6. They seem to be telling us something important about the upper mantle and how it affects the surface of continents. But exactly what they are telling us, other than that the upper mantle is often drippy, requires the integrated perspective offered by deep seismic imaging and clues to the palaeoelevation and erosion of the surface through time.

The work1 is supported by recent studies of volcanic rocks and mantle xenoliths — rocks carried by magmas from as deep as 120 km below the surface. Volcanism seems to have been encroaching on the plateau for the past 25 million years or so, and the geochemistry of the volcanic rocks shows increasing contributions from the convecting asthenosphere, rather than the ancient lithospheric mantle beneath the crust7. Levander and colleagues' main observation of one major lithospheric drip is thus superimposed on the longer-term picture of progressive erosion of the lithosphere by asthenospheric convection on the edges8. Decreased strength and increased density leading to dripping of the formerly buoyant lithosphere are thought to have resulted from its viscous weakening by hydration9 from fluids released by an unusually flat subducting slab, followed by infiltration by relatively dense, iron-rich melts from the asthenosphere after removal of the flat slab about 25 million years ago. These are reasonable arguments, based on convincing observations from volcanics and xenoliths. But relatively little is known about how much slab-derived hydration and magmatic infiltration would actually be required to counteract the initial buoyancy and strength of the ancient lithosphere, shaking it loose from its overlying crust and producing uplift.

Erosion of the Colorado plateau occurs not only from below by dripping, but also from above by surface processes such as those that carved the Grand Canyon. Because surface erosion is driven in large part by elevation and topographic relief, its spatial–temporal patterns provide clues about past uplift events. Levander et al.1 use evidence for increased erosion rates about 6 million years ago in the Grand Canyon area — the approximate epicentre of the drip — to suggest that their drip is actually a delamination peeling off from northeast to southwest.

But the history and spatial pattern of elevation gain are probably more complex than this, and much of it may have occurred much earlier. Evidence from regional drainage patterns10, stable isotopes11 and palaeohydrology12, as well as a growing body of thermochronological data13, provide strong indications that much of the Colorado plateau was quite high much earlier. Although still controversial, several lines of evidence indicate that, by 20 million to 60 million years ago, at least parts of the Grand Canyon were already almost as deep as they are today, as rivers flowed into what is now the Colorado plateau from highlands to the southwest and the Rockies to the northeast.

Also complicating a simple model of a single recent drip is evidence that large increases in the erosion rate over the past 6 million years or so are widespread and most pronounced tens to hundreds of kilometres upstream, in the Colorado River drainage basin. Whether this requires multiple drips, broader asthenospheric upwelling14 or geomorphic effects of recent river integration, it is likely that the story of one of geologists' favourite natural laboratories and playgrounds is far from fully told.


  1. 1.

    et al. Nature 472, 461–465 (2011).

  2. 2.

    , & , J. Geophys. Res. 86, 6115–6132 (1981).

  3. 3.

    J. Geophys. Res. 84, 7561–7571 (1979).

  4. 4.

    et al. Nature 431, 41–46 (2004).

  5. 5.

    , , & . Nature 438, 842–845 (2005).

  6. 6.

    , , & Nature Geosci. 2, 439–444 (2009).

  7. 7.

    et al. Geology 39, 27–30 (2011).

  8. 8.

    et al. Geology 38, 611–614 (2010).

  9. 9.

    et al. J. Geophys. Res. 113, B09210 (2008).

  10. 10.

    Geol. Soc. Am. Bull. doi:10.1130/B30274.1 (2011).

  11. 11.

    , & Tectonics 29, TC3005 (2010).

  12. 12.

    , & Science 319, 1377–1380 (2008).

  13. 13.

    , & , Geol. Soc. Am. Bull. 120, 571–587 (2008).

  14. 14.

    et al. Geophys. Res. Lett. 36, L19310 (2009).

Download references

Author information


  1. George Zandt and Peter Reiners are in the Department of Geosciences, University of Arizona, Tucson, Arizona 85721, USA.

    • George Zandt
    •  & Peter Reiners


  1. Search for George Zandt in:

  2. Search for Peter Reiners in:

Corresponding author

Correspondence to George Zandt.

About this article

Publication history



Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing