Research ArticleGEOPHYSICS

Illuminating subduction zone rheological properties in the wake of a giant earthquake

See allHide authors and affiliations

Science Advances  18 Dec 2019:
Vol. 5, no. 12, eaax6720
DOI: 10.1126/sciadv.aax6720
  • Fig. 1 Evolution of postseismic surface displacements for the first 6 years following the 27 February 2010 Mw 8.8 Maule earthquake in Chile from CGPS data.

    Background colors show interpolated CGPS-derived vertical motion for the entire observation period overlain on hill-shaded topography. The earthquake centroid location (red star) and focal mechanism, the outline of the megathrust from the model (dashed black polygon), and 5-m coseismic slip contours (black outline with gray filling) (24), which we use to calculate the initial coseismic stress changes in our inversion, are also shown. JFR, Juan Fernández Ridge; MFZ, Mocha Fracture Zone.

  • Fig. 2 Observed and modeled postseismic surface displacements.

    (Top) Cumulative horizontal CGPS displacements (black vectors) and model results reflecting the combined contribution of afterslip on the megathrust and ductile flow in the LC and upper mantle (white vectors). Cumulative vertical CGPS and model displacements are shown as nested large and small colored circles, respectively. Background colors show interpolated model vertical motions overlain on hill-shaded topography for direct comparison with Fig. 1. (Bottom) Postseismic displacement time series for example sites. Black circles and lines are the daily positions and empirical fits. Red, blue, and orange curves are afterslip, viscoelastic flow, and total model predictions, respectively (see fig. S2 for all sites). Q-Q′ marks the location of the profiles shown in Fig. 5.

  • Fig. 3 Inversion model geometry, surface displacements, and strain contributions from different portions of the lithosphere-asthenosphere system.

    (A) Perspective view of regional topography and bathymetry, CGPS site distribution (black circles), semitransparent discretized megathrust, and semitransparent polyhedral volumes colored by cumulative total deviatoric viscous strain (second invariant). (B) Afterslip contribution to the surface displacement field with total horizontal (black vectors) and vertical (colored circles) CGPS displacements. The model horizontal and vertical afterslip contributions are shown with white vectors and background colors, respectively. (C to F) Model surface displacements due to viscoelastic flow in the lithosphere-asthenosphere system. Vectors and background colors are the same as in (B). Insets show the map view, total cumulative strain with the same color scale as in (A).

  • Fig. 4 Derived afterslip and frictional parameter on the megathrust fault.

    (A) Six years of cumulative afterslip on the Maule megathrust fault, coseismic slip contours (24) in 5-m increments, the depth of the subducting slab in 20-km increments, the north-south trending thrust ridge (TR), additional offshore, upper-plate faults, and the approximate location of the backstop (34, 38). P1 to P4 correspond to example locations where system trajectories used in the frictional parameter calculation are shown in fig. S7. (B) Rate-and-state frictional parameter on the megathrust fault calculated between 20 days and 12 months after the earthquake. Red regions are velocity strengthening (VS), blue regions are velocity weakening (VW) or stable weakening, and light blue/red regions with values close to 0 are velocity neutral. Light green circles are the first week of aftershocks. (C) Depth-averaged afterslip, scaled rate-and-state frictional parameter with minimum/maximum, and coseismic slip model. Inferred correlations with structural position are labeled on the far right. Note that the afterslip curve is well defined in areas of high afterslip, primarily surrounding the seismogenic zone as our inversion penalizes slip from occurring in regions that slipped coseismically.

  • Fig. 5 Rheological and thermal properties for the Maule region from the inversion and flow law modeling.

    (A) Time evolution of effective viscosity for a few well-resolved polyhedral volumes including two MW examples that highlight the high-viscosity MW corner (3a) compared with the adjacent sub/backarc upper mantle (3b). (B) Depth-averaged initial and steady-state viscosities for the oceanic and continental regions and best-fit theoretical temperature and viscosity profiles (Supplementary Text; table S1). (C) Background stress levels. Dashed gray contours show the resolution power for the ε13 strain component related to shear motion on the megathrust (fig. S6). Black dashed line represents the continental Moho. (D) Estimated steady-state viscosities ηM. Circled numbers refer to the curves in (A). (E) Temperature estimates based on the rock properties constrained by the inverted steady-state stresses and strain rates assuming dislocation creep. White dots are 2000–2016 seismicity. (F) Activation energy (see the Supplementary Materials, tables S1 and S2, and fig. S8 for a full description of the associated calculations, best-fit rheological parameters, and additional derivatives including the initial viscosity ηK, background strain rate, and temperature anomaly). The profile location is shown in Fig. 2.

  • Fig. 6 Conceptual view of the Maule subduction zone based on our results.

    The megathrust fault exhibits heterogeneous frictional properties characterized by a patchwork of regions with variable behavior (Fig. 4). The majority of coseismic slip occurred in a UW region, although our frictional properties are not well constrained across the seismogenic zone. The shallow megathrust beneath the narrow AP, where most of the afterslip occurred, is largely VS (Fig. 4). Significant slip may have reached the trench particularly adjacent to the zone of intense outer-rise aftershocks. Vertical slices show the interpolated steady-state temperature for the lithosphere-asthenosphere system. Key features include a “wet/cold-nose” MW corner and the adjacent wet/warm upper mantle and inferred zone of partial melt beneath the volcanic arc at the base of the wet/cold LC [see Fig. 5 for rheological interpretations and fig. S8 for a comparable cross-sectional view and the associated temperature anomaly (i.e., difference from depth-averaged mean), which further highlights prominent features in our thermal model]. Vertical arrows above the topography illustrate the schematic, first-order pattern of total postseismic vertical displacements including the afterslip and viscoelastic contributions. Other labeled features (e.g., coupled/decoupled zones) are discussed in the main text.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/12/eaax6720/DC1

    Supplementary Materials and Methods

    Fig. S1. Continuous GPS network and interpolated interseismic velocity.

    Fig. S2. CGPS time series, empirical fits, and afterslip and viscoelastic flow modeling–based time series (also see subsequent pages).

    Fig. S3. Inversion parameters and sensitivity tests for select volumes.

    Fig. S4. Five checkerboard tests for slip on the megathrust and strain in the polyhedral volumes, which demonstrate our ability to resolve any up- and down-dip slip and viscous strain in the ductile region using our CGPS site distribution (white triangles).

    Fig. S5. Sensitivity tests for refining the model geometry.

    Fig. S6. Resolution power of strain components for each deformable finite volume.

    Fig. S7. Afterslip-related test and parameters for fault friction estimates.

    Fig. S8. Cross sections of derived rheological and thermal parameters from inversion and flow law modeling.

    Fig. S9. 3D, stress-driven, postseismic forward models of frictional afterslip and viscoelstic flow for comparison with inversion results.

    Table S1. Dislocation creep rheological parameter estimates for the Maule region.

    Table S2. Temperature estimates for the polyhedra.

    References (4867)

  • Supplementary Materials

    This PDF file includes:

    • Supplementary Materials and Methods
    • Fig. S1. Continuous GPS network and interpolated interseismic velocity.
    • Fig. S2. CGPS time series, empirical fits, and afterslip and viscoelastic flow modeling–based time series (also see subsequent pages).
    • Fig. S3. Inversion parameters and sensitivity tests for select volumes.
    • Fig. S4. Five checkerboard tests for slip on the megathrust and strain in the polyhedral volumes, which demonstrate our ability to resolve any up- and down-dip slip and viscous strain in the ductile region using our CGPS site distribution (white triangles).
    • Fig. S5. Sensitivity tests for refining the model geometry.
    • Fig. S6. Resolution power of strain components for each deformable finite volume.
    • Fig. S7. Afterslip-related test and parameters for fault friction estimates.
    • Fig. S8. Cross sections of derived rheological and thermal parameters from inversion and flow law modeling.
    • Fig. S9. 3D, stress-driven, postseismic forward models of frictional afterslip and viscoelstic flow for comparison with inversion results.
    • Table S1. Dislocation creep rheological parameter estimates for the Maule region.
    • Table S2. Temperature estimates for the polyhedra.
    • References (4867)

    Download PDF

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article