No access

Reconstructing Farallon Plate Subduction Beneath North America Back to the Late Cretaceous

7 Nov 2008
Vol 322, Issue 5903
pp. 934-938


Using an inverse mantle convection model that assimilates seismic structure and plate motions, we reconstruct Farallon plate subduction back to 100 million years ago. Models consistent with stratigraphy constrain the depth dependence of mantle viscosity and buoyancy, requiring that the Farallon slab was flat lying in the Late Cretaceous, consistent with geological reconstructions. The simulation predicts that an extensive zone of shallow-dipping subduction extended beyond the flat-lying slab farther east and north by up to 1000 kilometers. The limited region of flat subduction is consistent with the notion that subduction of an oceanic plateau caused the slab to flatten. The results imply that seismic images of the current mantle provide more constraints on past tectonic events than previously recognized.

Get full access to this article

View all available purchase options and get full access to this article.

Supplementary Material

File (liu_som.pdf)

References and Notes

G. Schubert, D. L. Turcotte, P. Olson, in Mantle Convection in the Earth and Planets (Cambridge Univ. Press, Cambridge, 2001), pp. 417–498.
G. F. Davies, in Dynamic Earth: Plates Plumes and Mantle Convection (Cambridge Univ. Press, Cambridge, 1999), pp. 262–323.
S. Zhong, M. Gurnis, Science267, 838 (1995).
S. P. Grand, Philos. Trans. R. Soc. London A360, 2475 (2002).
R. D. Van der Hilst, S. Widiyantoro, E. R. Engdahl, Nature386, 578 (1997).
Y. Ren, E. Stutzmann, R. D. Van der Hilst, J. Besse, J. Geophys. Res.112, B01302 (2007).
Both shear (4) and compressional (5, 6) wave tomography show a prominent positive slablike seismic anomaly at mid-mantle depths below the east coast of North America. The high-velocity component of the shear wave seismic tomography (4) is converted to effective temperature (21) with a scaling factor of 2 × 103 °C km–1 s. The upper 250-km signal associated with the North American craton is removed because it is likely neutrally buoyant (28). We also remove structures below 2400-km depth, where there is a clear gap in the tomographic image. For the rest, we assume a constant seismic-to-temperature scaling that we constrain by fitting our models to stratigraphic data.
H.-P. Bunge, C. R. Hagelberg, B. J. Travis, Geophys. J. Int.152, 280 (2003).
L. Liu, M. Gurnis, J. Geophys. Res.113, B08405 (2008).
The adjoint method, a gradient-based inversion, optimizes the initial condition by minimizing the mismatch between a model prediction and observation. For the nonlinear mantle convection, the observational constraint is seismic tomography. A simple backward integration of present-day mantle structures to the initial time provides a first-order estimate of the initial condition (9). In a forward-adjoint looping scheme, the forward run—which solves for conservation of mass (ui,i = 0, where u is velocity), momentum [–P,i +ui,j + ηuj,i),j + ρoαΔTgδir = 0, where P is dynamic pressure, η viscosity, ΔT effective temperature, α thermal expansion, ρo reference density, and g gravitational acceleration], and energy (T,t + uiT,i = κT,ii, where T is temperature, t time, and κ thermal diffusivity) for an incompressible Newtonian fluid—predicts the present-day mantle thermal structure, whereas the adjoint run, which solves for the kinematic adjoint equation (λ,t + ui λ,i + κλ,ii = 0, where λ is adjoint temperature) with u stored from the forward run—back-propagates the mismatch to the initial time. The past mantle thermal state is iteratively updated until the mismatch with the observation (tomography) is small. The adjoint method has been incorporated (9) in a finite-element incompressible model of mantle convection (17).
We use GPlates reconstructions of global plate motions at 1-million-year intervals, in which the plate margins continuously evolve with self-consistent velocities between the plates and plate margins (29). The rotation poles of (30) are used in the GPlates reconstructions, implemented in a moving hotspot reference frame.
M. Gurnis, R. D. Müller, L. Moresi, Science279, 1499 (1998).
A. G. Smith, D. G. Smith, B. M. Funnel, Atlas of Mesozoic and Cenozoic Coastlines (Cambridge Univ. Press, Cambridge, 1994).
T. D. Cook, A. W. Bally, Stratigraphic Atlas of North and Central America (Princeton, Univ. Press, Princeton, NJ, 1975).
M. Pang, D. Nummedal, Geology23, 173 (1995).
S. Liu, D. Nummedal, P.-G. Yin, H.-J. Luo, Basin Res.17, 487 (2005).
We use CitcomS, a finite-element model (31) on a global mesh with a 129-by-129-by-65 grid on each of the 12 caps. We assume a free slip and isothermal core-mantle boundary and an isothermal surface. A Rayleigh number (based on thickness of the mantle) of 9.4 × 106 is used (21). The three-layer viscosity includes a lithosphere with constant 5 × 1022 Pa s viscosity, an upper mantle, and a lower mantle with varying viscosities to be constrained. Laterally, viscosity increases exponentially by one order of magnitude for a temperature drop of 200°C.
J. X. Mitrovica, C. Beaumont, G. T. Jarvis, Tectonics8, 1079 (1989).
P. M. Burgess, M. Gurnis, L. Moresi, Bull. Geol. Soc. Am.109, 1515 (1997).
H.-P. Bunge, S. P. Grand, Nature405, 337 (2000).
Materials and methods are available as supporting material on Science Online.
G. A. Bond, Geology4, 557 (1976).
G. A. Milneet al., J. Geophys. Res.109, B02412 (2004).
J. Saleeby, Bull. Geol. Soc. Am.115, 655 (2003).
M.-A. Gutscher, W. Spakman, H. Bijwaard, E. R. Engdahl, Tectonics19, 814 (2000).
R. F. Livaccari, K. Burke, A. M. C. Sengor, Nature289, 276 (1981).
R. M. Flowers, Eos52, abstr. V34C-06 (2007).
S. Goes, S. van der Lee, J. Geophys. Res.107, 2050 (2002).
M. Gurnis et al., Eos52, abstr. DI14A-07 (2007).
R. D. Mülleret al., Geochem. Geophys. Geosys.9,2008).
S. Zhong, M. T. Zuber, L. Moresi, M. Gurnis, J. Geophys. Res.105, 11,063 (2000).
This is contribution no. TO 88 of the Caltech Tectonics Observatory. The work was partially supported by the Gordon and Betty Moore Foundation through the Tectonics Observatory and the NSF through EAR-0609707. We appreciate discussions with J. Saleeby and R. D. Müller. The original CitcomS software was obtained from the Computational Infrastructure for Geodynamics (CIG) (


eLetters is an online forum for ongoing peer review. Submission of eLetters are open to all. eLetters are not edited, proofread, or indexed. Please read our Terms of Service before submitting your own eLetter.

Log In to Submit a Response

No eLetters have been published for this article yet.

Information & Authors


Published In

Volume 322 | Issue 5903
7 November 2008

Submission history

Received: 8 July 2008
Accepted: 8 October 2008
Published in print: 7 November 2008


Request permissions for this article.


Supporting Online Material
Materials and Methods
Figs. S1 to S3



Lijun Liu*
Seismological Laboratory, California Institute of Technology, Pasadena, CA 91125, USA.
Sonja Spasojević
Seismological Laboratory, California Institute of Technology, Pasadena, CA 91125, USA.
Michael Gurnis
Seismological Laboratory, California Institute of Technology, Pasadena, CA 91125, USA.


To whom correspondence should be addressed. E-mail: [email protected]

Metrics & Citations


Article Usage


Export citation

Select the format you want to export the citation of this publication.

Cited by

  1. The Mesozoic Amdo micro-block and East Asian superconvergent tectonic system, Gondwana Research, 101, (257-277), (2022).
  2. Controls on the present-day dynamic topography predicted from mantle flow models since 410 Ma, Geophysical Journal International, 225, 3, (1637-1652), (2021).
  3. Early Cretaceous tectonics across the North Pacific: New insights from multiphase tectonic extension in Eastern Eurasia, Earth-Science Reviews, 217, (103552), (2021).
  4. Fresh Outlook on Numerical Methods for Geodynamics. Part 2: Big Data, HPC, Education, Encyclopedia of Geology, (841-855), (2021).
  5. Global mantle flow retrodictions for the early Cenozoic using an adjoint method: evolving dynamic topographies, deep mantle structures, flow trajectories and sublithospheric stresses, Geophysical Journal International, 226, 2, (1432-1460), (2021).
  6. Mesozoic intraplate tectonism of East Asia due to flat subduction of a composite terrane slab, Earth-Science Reviews, 214, (103505), (2021).
  7. East Asian lithospheric evolution dictated by multistage Mesozoic flat-slab subduction, Earth-Science Reviews, 217, (103621), (2021).
  8. Observational Estimates of Dynamic Topography Through Space and Time, Mantle Convection and Surface Expressions, (371-411), (2021).
  9. Extending full-plate tectonic models into deep time: Linking the Neoproterozoic and the Phanerozoic, Earth-Science Reviews, 214, (103477), (2021).
  10. Mountain building, mantle convection, and supercontinents: revisited, Earth and Planetary Science Letters, 564, (116905), (2021).
  11. See more

View Options

Check Access

Log in to view the full text


AAAS login provides access to Science for AAAS Members, and access to other journals in the Science family to users who have purchased individual subscriptions.

Log in via OpenAthens.
Log in via Shibboleth.
More options

Purchase digital access to this article

Download and print this article for your personal scholarly, research, and educational use.

Purchase this issue in print

Buy a single issue of Science for just $15 USD.

View options

PDF format

Download this article as a PDF file

Download PDF

Full Text








Share article link

Share on social media