Published November 1, 2025 | Version Supplemental material
Journal Article Open

Constraints on nonlinear mantle rheology from multi-scale geodetic observations of plate motion and post-seismic enhanced landward motion

  • 1. ROR icon California Institute of Technology
  • 2. ROR icon Jet Propulsion Lab
  • 3. ROR icon Earth Observatory of Singapore

Abstract

Plate motions and post-seismic deformation are surface expressions of Earth's nonlinear dynamics shaped by a visco-elasto-plastic rheology. A particularly exciting avenue of investigation lies in the development of numerical models that can self-consistently resolve these processes across spatial and temporal scales. We formulate a three-dimensional model of the Chilean subduction zone that incorporates buoyancy forces, nonlinear rheology, shear modulus variability and friction on the megathrust. When dislocation creep parameters are varied, computations produce the ≈ 7 cm/year motion of the Nazca Plate but different vertical viscosity profiles in the upper mantle. The model predicts enhanced landward motion (ELM) in the non-ruptured segment on the overriding plate adjacent to the 2010 Maule earthquake (M_w = 8.8), with amplitudes inversely related to the asthenosphere viscosity. Our preferred model shows a pre-seismic viscosity of 5–9 x 10¹ Pa⋅s in the asthenosphere and ~10²¹ Pa⋅s close to the 660 km interface, and is able to reproduce the ∼5 mm/year ELM observed geodetically. The range of permissible nonlinear rheological parameters is wide when constrained by the long-term motion of the Nazca Plate alone, but becomes well posed when post-seismic observations are simultaneously considered. The model reconciles experimentally-constrained dislocation creep parameters, long-term plate velocities and far-field post-seismic deformation and advances our understanding of nonlinear mantle rheology.

Copyright and License

© 2025 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.

Acknowledgement

This work is supported by the National Science Foundation (NSF) through award EAR-2343864. The authors acknowledge use of the Anvil supercomputer at Purdue University, supported by the NSF ACCESS program under the allocation EAR160027. This work was done as an outside activity and not in R.M.'s capacity as an employee of the Jet Propulsion Laboratory, California Institute of Technology. J.F. would like to thank Simone Puel, Yuan-Kai Liu, Leonid Pereiaslov and Linxuan Li for helpful discussions. The authors thank Rob Govers and an anonymous reviewer for constructive reviews.

Data Availability

The computations were completed using the open-source package Underworld2 version v2.10.0b (Mansour et al., 2020). The code used for developing this model and producing figures is available on Zenodo (https://doi.org/10.5281/zenodo.15844027). The key model outputs can be downloaded from CaltechDATA (https://doi.org/10.22002/e9rnc-k7z23). The deformation rates before and after the 2010 Malue earthquake are from the Supporting Information of Melnick et al. (2017).

Supplemental Material

MMC. The supplementary material includes text detailing the thermal structure used in this study (Text S1), dislocation and diffusion creep laws (Text S2), and additional computations predicting similar enhanced landward motion (ELM) but different plate motions (Text S3). Table S1 contains key rheological parameters in the model. Figure S1 shows the thermal structure zoomed in near the megathrust. Figure S2 shows the shear modulus structure used in the reference model. Figure S3 shows the co-seismic stress changes in the megathrust shear zone. Figure S4 shows the plate motion and co-seismic slip for different viscosity profiles for computations discussed in the main text. Figure S5 shows plate motions, viscosity profiles and ELM produced by computations discussed in Text S3. Figure S6 shows parameters required for similar ELM and plate motion. Figure S7 shows the force balance in the adjacent non-ruptured segment where ELM is observed. Figure S8 shows average post-seismic slip velocity in the megathrust shear zone for cases with different rates of friction increase after the earthquake. Figure S9 shows post-seismic velocity changes from additional computations with different rates of friction increase after the earthquake. Figure S10 shows post-seismic velocity changes from additional computations with different time step sizes.

Files

Files (1.4 MB)

Name Size Download all
md5:63b810e954874d2fe1d65246a5ee720d
1.4 MB Download

Additional details

Related works

Funding

National Science Foundation
EAR-2343864
Purdue University West Lafayette
EAR160027

Dates

Accepted
2025-07-30
Available
2025-08-08
Available online
Available
2025-08-08
Version of record

Caltech Custom Metadata

Caltech groups
Seismological Laboratory, Division of Geological and Planetary Sciences (GPS)
Publication Status
Published