Seismic characterization of mantle flow in subduction systems: Can we resolve a hydrated mantle wedge?

https://doi.org/10.1016/j.epsl.2006.01.022Get rights and content

Abstract

This study provides new constraints on the resolvability of mantle flow in subduction zone settings as inferred by observations of seismic anisotropy. We are motivated by the broad range of shear wave splitting observations in subduction systems that suggest complex flow geometries, changes in the deformation state of mantle minerals, or a combination of both. While shear wave splitting fast polarization directions are typically interpreted as a proxy for flow or maximum finite extension, experimental studies suggest that olivine slip systems change under higher stress and hydration states, conditions likely appropriate for subduction systems. In this study, we predict shear wave splitting as a result of mantle silicate lattice-preferred orientation development in steady-state two-dimensional mantle flow models using a textural development theory that incorporates the combined effects of intracrystalline slip and dynamic recrystallization. We utilize the resulting textures to predict shear wave splitting for populations of seismic raypaths traversing the model within the subduction zone mantle wedge. The results of our study demonstrate that combined observations of variations in fast polarization directions and splitting times make it possible to resolve a shift from anhydrous to hydrous mantle insubduction zone settings provided very good sampling of the mantle wedge. Our models are generally consistent with observed splitting variations for several subduction zones with dense data sampling, including Tonga, Japan, and Kamchatka. The implications of our work suggest that, provided adequate data sampling, shear wave splitting measurements can provide the necessary information to evaluate potential competing effects between variations in mantle flow direction and changes in the stress and hydration states of subduction zone mantle wedges.

Introduction

As part of the convective system that mobilizes Earth's mantle, subduction zones are one of the key expressions of plate tectonics. Convergent margins represent an important setting to study the processes of arc volcanism, seismicity, and deformation, which are influenced by thermal, rheological, and compositional factors. In particular, understanding the relationship between deformation and resulting fabric in the mantle wedge between subducting and overriding lithospheric plates is essential to improving our understanding of the subduction factory. Significant progress in constraining mantle wedge processes has been made from seismological observations, mineral physics experiments, and geodynamic models [1], [2], [3], [4], [5], but linking results from these fields remains a significant challenge. Of particular importance is providing an improved understanding of the influence of mantle hydration and its effects on inferring mantle wedge deformation using seismic observations [6], [7].

Subduction zones exhibit a broad range of mantle seismic anisotropy observations which are inferred to express dominant deformational processes in this tectonic setting. A particularly well-studied manifestation of seismic anisotropy is shear wave splitting (SWS). Observations of SWS are comprised of a fast polarization direction, which reflects the orientation of fabric, and a splitting time, which reflects the organization and strength of the fabric. In many cases, a range of fast polarization directions has been documented within the same subduction system, such as in Tonga (e.g., [8], [9], [10]), Japan (e.g., [11], [12]), Kamchatka (e.g., [13], [14]), and South America (e.g., [15], [16], [17]) (Fig. 1). These sometimes discrepant observations hint at the complex array of dynamic processes that occur near subduction zones.

Seismic anisotropy in the mantle is generally assumed to be the result of lattice-preferred orientation (LPO) of mantle minerals such as olivine. For systems with little or no hydration, LPO of olivine a-axes (the seismically fastest axis for shear waves) are assumed to be aligned with the direction of flow in the dislocation creep regime (e.g., [18], [19]). Conversely, experimental studies suggest that olivine slip systems change under higher stresses and hydration states [4]. Since some subduction zones are likely regions of high stress and hydration (e.g., [20]), the development of olivine LPO in these regions may be significantly different [4], [21], [22].

An extensive range of models for subduction zone flow and deformation have been developed (e.g., [23], [24], [25], [26]). While important advances have been made to connect deformation in flow models with seismic anisotropy observations (e.g., [9], [27]), more detailed links between geodynamical models and observed seismic anisotropy for subduction zones can yield significantly improved insight into SWS observations. Additionally, understanding the role mantle hydration plays in seismic anisotropy development due to deformation can improve interpretations of shear wave splitting parameters. Here we provide important new connections between subduction zone flow models and the seismic manifestation of flow for a range of hydration states in the mantle.

In this study, we examine the connections between dynamic flow models and seismic anisotropy to interpret deformation in subduction systems. We assume that the bulk of observed seismic anisotropy in subduction zones likely results from fabric in the mantle wedge and focus on wedge deformation as imaged by seismic anisotropy observations. To provide intuition regarding how a shift from anhydrous to hydrous olivine is manifested in shear wave splitting observations, we examine simple shear models with a first order transition in olivine hydration state. We then apply this approach to determine the patterns of shear wave splitting measurements that would develop from a hydrated wedge by examining a range of models with anhydrous and hydrated mantle wedge regions. Finally, we relate our results to subduction zones around the Pacific Rim to evaluate SWS observations in these regions.

Section snippets

Causes of seismic anisotropy in subduction systems

There are several possible causes of seismic anisotropy. Shape-preferred orientation (SPO) can result from cracks in the crust, melt-filled cracks, or lenses in the mantle (e.g., [2]). SPO may exist in areas beneath mid-ocean ridges and subduction zones in the uppermost mantle, as well as in the lowermost mantle (e.g., [28]) LPO results from mineral alignment during strain (e.g., [4], [6], [19]), and is inferred to be the dominant cause of upper mantle anisotropy (e.g., [2], [8], [29], [30],

Predicting shear wave splitting from mantle flow models

In this section, we present the methodology for linking models of mantle flow with predictions of SWS patterns for subduction zone settings following the general approach of Fischer et al. [9], Kaminski and Ribe [5], Fouch et al. [35], and Hall et al. [27]. In its original form, this methodology consists of utilizing the velocity field from mantle flow models to predict LPO development and resulting SWS. These previous studies mapped LPO for mantle mineralogies by either a) orienting

Results of LPO and shear wave splitting modeling

Here we present our LPO development and predicted SWS results for flow models using anhydrous, hydrous, and anhydrous to hydrous transition slip systems. To examine anhydrous and hydrous mantle olivine LPO development, we used the critically resolved shear stresses from Kaminski and Ribe [5], [6] for anhydrous olivine LPO development models and from Jung and Karato [4] for hydrous olivine models. We note that the simple flow fields used in our modeling are not viscosity-dependent; we,

Shear wave splitting and mantle fabric in hydrated subduction systems

The results of our LPO modeling and SWS predictions place important constraints on our understanding of the links between flow dynamics and seismic observations used to infer flow. While the underlying models in this work are relatively simple and certainly do not fully represent all details of Earth dynamics, they provide an important set of guidelines in understanding LPO development and its relationship with SWS, particularly in subduction zone environments where hydrated mantle phases

Conclusions

This study demonstrates that changes in LPO geometry due to transitions from anhydrous and hydrous olivine rheologies can be resolved with shear wave splitting provided appropriate sampling coverage of the mantle wedge. In simple shear models, complete reorientation of olivine aggregates due to a transition between anhydrous and hydrous olivine rheologies requires strains on the order of 370–450%. Continued LPO development can be observed in these models via increases in predicted splitting

Acknowledgements

We express our gratitude to Karen Fischer for the predicted shear wave splitting code and David Abt and Karen Fischer for an early version of the shear wave splitting compilation presented in Fig. 1. Thanks also to Megan Anderson, Lara Wagner, George Zandt and Susan Beck for fruitful discussions. We also thank Jules Browaeys for help with the LPO plotting and Mark Stevens for computer support. Martha Savage and an anonymous reviewer provided helpful comments that strengthened the clarity and

References (68)

  • S. Karato et al.

    Water, partial melting and the origin of the seismic low velocity and high attenuation zone in the upper mantle

    Earth Planet. Sci. Lett.

    (1998)
  • K. Gledhill et al.

    SKS splitting and the seismic anisotropy of the mantle beneath the Hikurangi subduction zone, New Zealand

    Phys. Earth Planet. Inter.

    (1996)
  • J. Xie

    Shear-wave splitting near Guam

    Phys. Earth Planet. Inter.

    (1992)
  • R.M. Russo et al.

    Shear-wave splitting in Northeast Venezuela, Trinidad, and the eastern Caribbean

    Phys. Earth Planet. Inter.

    (1996)
  • M. Bystricky et al.

    High shear strain of olivine aggregates: rheological and seismological consequences

    Science

    (2000)
  • M.K. Savage

    Seismic anisotropy and mantle deformation; what have we learned from shear wave splitting?

    Rev. Geophys.

    (1999)
  • H. Jung et al.

    Water-induced fabric transitions in olivine

    Science

    (2001)
  • E. Kaminski

    The influence of water on the development of lattice preferred orientation in olivine aggregates

    Geophys. Res. Lett.

    (2002)
  • K.M. Fischer et al.

    Anisotropy and flow in Pacific subduction zone back-arcs

    Pure Appl. Geophys.

    (1998)
  • J.R. Bowman et al.

    Shear-wave splitting in the upper-mantle wedge above the Tonga subduction zone

    Geophys. J. R. Astron. Soc.

    (1987)
  • K.M. Fischer et al.

    Modeling anisotropy and plate-driven flow in the Tonga subduction zone back arc

    J. Geophys. Res.

    (2000)
  • G.P. Smith et al.

    A complex pattern of mantle flow in the Lau backarc

    Science

    (2001)
  • C.M. Bernot et al.

    Seismic anisotropy and mantle flow beneath Japan, Fall Meet. Suppl. Abstract S52A-1056

    Eos Trans. AGU

    (2002)
  • V. Peyton et al.

    Mantle flow at a slab edge: seismic anisotropy in the Kamchatka region

    Geophys. Res. Lett.

    (2001)
  • V. Levin et al.

    Detailed mapping of seismic anisotropy with local shear waves in southeastern Kamchatka

    Geophys. J. Int.

    (2004)
  • R.M. Russo et al.

    Trench-parallel flow beneath the Nazca plate from seismic anisotropy

    Science

    (1994)
  • J. Polet et al.

    Shear wave anisotropy beneath the Andes from the BANJO, SEDA, and PISCO experiments

    J. Geophys. Res.

    (2000)
  • M.L. Anderson et al.

    Anisotropy and mantle flow in the Chile–Argentina subduction zone from shear wave splitting analysis

    Geophys. Res. Lett.

    (2004)
  • S. Karato et al.

    Rheology of the upper mantle: a synthesis

    Science

    (1993)
  • S. Zhang et al.

    Lattice preferred orientation of olivine aggregates deformed in simple shear

    Nature

    (1995)
  • E. Kaminski et al.

    D-Rex, a program for calculation of seismic anisotropy due to crystal lattice preferred orientation in the convective upper mantle

    Geophys. J. Int.

    (2004)
  • L. Mehl et al.

    Arc-parallel flow within the mantle wedge: evidence from the accreted Talkeetna arc, south central Alaska

    J. Geophys. Res.

    (2003)
  • G.K. Batchelor

    An Introduction to Fluid Dynamics

    (1967)
  • D. McKenzie

    Finite deformation during fluid flow

    Geophys. J. R. Astron. Soc.

    (1979)
  • Cited by (54)

    • Inferring upper-mantle flow from seismic anisotropy: An experimental perspective

      2016, Tectonophysics
      Citation Excerpt :

      However there are many scenarios in nature where strain histories are complex, and CPO is correspondingly affected. When changes in CPO lag significantly behind changes in kinematics, interpretation of seismic anisotropy must be modified accordingly (Boneh and Skemer, 2014; Boneh et al., 2015; Kaminski and Ribe, 2002; Lassak et al., 2006; Li et al., 2014; Skemer et al., 2012). Shear zones that intersect larger bodies of previously deformed rock provide an opportunity to assess two sequential stages of deformation, effectively serving as a large-scale rock deformation experiment.

    • Subduction of oceanic asthenosphere: A critical appraisal in central Alaska

      2013, Earth and Planetary Science Letters
      Citation Excerpt :

      This estimate is certainly subject to better calibration with improved knowledge of anisotropy layering or/and variations in anisotropy strength/symmetry in the oceanic lithosphere (Song and Kim, 2012; Audet, 2013), but it seems to indicate a need for a substantial contribution of subducting asthenosphere to observations. Typically, the effect of wedge anisotropy contributes to the splitting time of about 0.5 s or less in the fore-arc due to the relatively short path length and possibly complex fabrics in the wedge (e.g., Lassak et al., 2006; Levin et al., 2007; Wiens et al., 2008). Motivated by a linear increase in splitting time at stations above the mantle wedge in the northwest back azimuth, Christensen and Abers (2010) suggested that the mantle wedge anisotropy in central Alaska could be too strong to see slab or/and sub-slab anisotropy.

    • Sub-slab mantle anisotropy beneath south-central Chile

      2012, Earth and Planetary Science Letters
    • Anomalous bathymetry, 3D edge driven convection, and dynamic topography at the western Atlantic passive margin

      2011, Journal of Geodynamics
      Citation Excerpt :

      This Rayleigh number seems high; for example, in the context of Rayleigh numbers for Earth whole mantle convection of ∼107. Our Rayleigh number was used to set up flow velocities in the modeled mantle that were consistent with established Earth upper mantle flow velocities (typically ∼cm/yr) (Silver and Holt, 2002; Behn et al., 2004; Lassak et al., 2006). Together with the large variations in model viscosity from the upper mantle to the lithosphere and a chosen reference viscosity of 1019 Pa s, this results in the Rayleigh number value given above.

    View all citing articles on Scopus
    View full text