Elsevier

Earth and Planetary Science Letters

Volume 498, 15 September 2018, Pages 169-184
Earth and Planetary Science Letters

SKS splitting in the Western Indian Ocean from land and seafloor seismometers: Plume, plate and ridge signatures

https://doi.org/10.1016/j.epsl.2018.06.033Get rights and content

Highlights

  • We find strong evidence for an asthenospheric Réunion plume-Central Indian Ridge connection.

  • At the Southwest Indian Ridge mantle may flow along-axis, channeled by lithospheric walls.

  • E–W anisotropy in the Mozambique Channel results likely from Africa's separation from Gondwana.

  • At mid-ocean ridges, mantle flows may generally orient as a function of the ridges' spreading rate.

  • Somali plate motion does not produce significant asthenospheric anisotropy due to its slow motion of <2.6 cm/yr.

Abstract

We present SKS splitting measurements in the Western Indian Ocean, recorded on 20 land and 57 seafloor seismometers deployed by the RHUM-RUM experiment (Réunion Hotspot and Upper Mantle – Réunions Unterer Mantel). We discuss our splitting observations within their geodynamic settings and compare them to SKS splitting parameters predicted from an azimuthally anisotropic Rayleigh wave tomography model that includes the RHUM-RUM data. We find that anisotropic directions poorly correlate with the present-day motion of the Somali plate, which at <2.6 cm/yr may be too slow to cause strongly sheared fabric in the asthenosphere. Fast split directions (Φ) between La Réunion and the Central Indian Ridge (CIR) trend E–W and provide strong, first seismological evidence for near-horizontal flow in the asthenosphere that connects the Réunion mantle upwelling with the CIR, supporting a long-standing hypothesis on plume–ridge interaction. In the vicinity of the Réunion hotspot, we observe a seismic anisotropy pattern indicative of a parabolic asthenospheric flow controlled by the Réunion mantle upwelling and its consecutive asthenospheric spreading. We furthermore observe ridge-normal Φ along the CIR and ridge-parallel Φ along the Southwest Indian Ridge (SWIR), both mainly attributed to asthenospheric mantle flows. In the Mozambique Channel between East-Africa and Madagascar, we attribute E–W trending Φ to frozen lithospheric structures, recording the paleo-orientation of the spreading ridges that enabled Madagascar's separation away from Africa. Based on the synopsis of this and previous SKS splitting studies at mid-ocean ridges, we propose that ridge-normal Φ may develop at fast and intermediate spreading ridges (e.g., CIR and East Pacific Rise) and ridge-parallel Φ could be characteristic to slow spreading ridges (e.g., SWIR, Mid-Atlantic Ridge and the paleo-ridges in the Mozambique Channel).

Introduction

The Réunion hotspot in the Western Indian Ocean feeds the Piton de la Fournaise, one of the most active volcanoes in the world. Its age-progressive hotspot track is formed by La Réunion Island, Mauritius Island and the Mascarene Plateau on the Somali plate, and the Chagos, Maldive and Laccadive alignment on the Indian plate (Duncan, 1990, Duncan et al., 1990). The track leads to the Deccan Traps of India, one of the largest flood basalt provinces on Earth that erupted 65 Ma ago (Courtillot et al., 1986) and is likely linked to the Cretaceous–Paleogene extinction event (Richards et al., 2015).

The Western Indian Ocean presents an unusual variety of upper mantle phenomena to investigate. The Réunion volcanic hotspot has been proposed to be fed by a “primary” (Courtillot et al., 2003) mantle plume (Morgan, 1972) – a deep rooted upwelling of mantle material that may be connected to the South-African Superswell (Forte et al., 2010). A recent, regional Rayleigh wave tomography study indicates that the Réunion hotspot could also be an expression of mantle material rising from beneath the Mascarene Basin, where a broad low-shear wave velocity anomaly at asthenospheric depths is observed (Mazzullo et al., 2017). Morgan (1978) also hypothesized that some of the hot material rising beneath La Réunion may be feeding the nearest spreading ridge, the Central Indian Ridge (CIR) at 1000 km distance, through a sub-lithospheric, channeled mantle flow. The Southwest Indian Ridge (SWIR) is the other nearby spreading center. Despite its ultra-slow spreading rate and magma-starved dynamics, it also could be influenced by adjacent hotspots/plumes (La Réunion, Marion and/or Crozet; Sauter et al., 2009) and/or the South-African Superswell. Finally, in the regional context of the East African Rift System (EARS), the location of the diffuse plate boundary that connects the southern EARS to the SWIR remains subject to discussion (e.g., Kusky et al., 2010, Stamps et al., 2015), together with the synchronous volcanism occurring from the EARS to the Mascarene Basin at 10–20 Ma ago (Michon, 2016) that could suggest episodic, large-scale events of mantle upwelling.

To address these questions of upper mantle structures and dynamics, we analyzed seismic anisotropy via the splitting of the teleseismic, core-refracted shear waves such as SKS, SKKS, and pSKS phases (hereafter called XKS). Seismic anisotropy is accepted to result mostly from lattice preferred orientation (LPO) of rock-forming minerals in response to tectonic strain. In the upper mantle, olivine is the dominating phase. It is intrinsically anisotropic to P and S-waves (e.g., Mainprice et al., 2000) and controls large-scale patterns of seismic anisotropy (Nicolas and Christensen, 1987). In the lithosphere, LPO may record past tectonic episodes that produced deformation such as faults and shear zones, tectono-thermal interactions with the asthenosphere such as plume head arrivals, and/or plate accretion at mid-ocean ridges (e.g., Wolfe and Silver, 1998). In the latter scenario, rock fabrics acquired through ridge-parallel or ridge-normal mantle flow (i.e., ridge-parallel or ridge-normal LPO) could become “frozen-in” by lithospheric cooling and preserved during the seafloor's entire lifetime. In the asthenosphere, LPO may reflect present-day mantle flow, the subducting of mantle slabs, the shearing caused by motion of the overlying plate, and/or the flow induced by rising plumes spreading horizontally beneath the lithosphere (Morgan et al., 1995). In addition to LPO (or “intrinsic” anisotropy), shape preferred orientation (SPO, or “extrinsic” anisotropy) can contribute to observed shear wave splitting patterns. SPO can be generated by (liquid filled) cracks, oriented melt pockets, dipping discontinuities, and/or fine layering (e.g., Wang et al., 2013).

Seismic anisotropy may be also present within the D layer in the lowermost mantle (e.g., Kendall and Silver, 1996), and this region is also sampled by XKS waves. There are, however, several seismological arguments why observed XKS splitting is dominantly caused by upper mantle anisotropy: i) XKS splitting parameters often display short-scale variations indicative of rather superficial causes of anisotropy (e.g., Alsina and Snieder, 1994); and ii) anisotropy measurements from XKS and (local) S-phases yield similar splitting parameters, putting an upper bound of δtlower_mantle0.2 s on the splitting contribution from the lower mantle (e.g., Vinnik et al., 1995, Savage, 1999, Long, 2009). XKS splitting is hence a suitable tool to investigate seismic anisotropy in the upper mantle.

Section snippets

Data set

Seismic data analyzed in this study were recorded during the RHUM-RUM experiment (Réunion Hotspot and Upper Mantle – Réunions Unterer Mantel; Barruol and Sigloch, 2013). This French–German experiment in the Western Indian Ocean (Fig. 1) deployed 20 broadband, three-component land seismometers between 2011 and 2016, and 57 broad- and wideband, three-component ocean-bottom seismometers (OBSs) between October 2012 and December 2013. Detailed station information is provided in the on-line

Methodology

We refer to our measurements as XKS splittings, meaning we recorded splitting mostly on SKS phases but occasionally on SKKS and pSKS phases, too.

Results and interpretation

20 land seismometers and 40 usable ocean-bottom seismometers (OBSs) yielded 74 non-null and 205 null XKS splitting measurements from 101 earthquakes (Fig. 1). The signal-to-noise ratio for all these measurements averages 9.9, the dominant frequency 0.1 Hz. The smallest earthquake magnitudes yielding splitting were MW=6.1 on land station MAYO, and MW=6.3 on OBS RR29 (Fig. 2). Individual measurements can be found in the supporting material and in our on-line XKS splitting data-base (Wüstefeld et

Discussion

Generally, for the Western Indian Ocean our XKS splitting observations largely coincide with the azimuthal anisotropies determined by global waveform studies (e.g., Debayle et al., 2016, Schaeffer et al., 2016) and mantle flow computations (e.g., Becker et al., 2008, Conrad and Behn, 2010 – the latter with constraints on the Mozambique Channel for example).

In Table 2, we classified our interpretations of our XKS splitting measurements performed on active or fossil mid-ocean ridges in the

Conclusions

As part of the RHUM-RUM project that investigates whole-mantle structure beneath the Réunion hotspot in the Western Indian Ocean, we presented XKS splitting measurements for 20 terrestrial and 40 ocean-bottom seismometers (Fig. 1; Table 1), installed temporarily between 2011–2016 (land stations) and 2012–2013 (seafloor stations). We compared measured XKS splitting parameters with predicted XKS splitting parameters computed from a regional, azimuthally anisotropic Rayleigh wave tomography (

Acknowledgments

The presented XKS splitting measurements can be found in the on-line supplements and in our splitting data-base (http://splitting.gm.univ-montp2.fr/DB/public/searchdatabase.html), which is also mirrored at IRIS (Incorporated Research Institutions for Seismology, https://ds.iris.edu/spud/swsmeasurement). The RHUM-RUM project (http://www.rhum-rum.net) was funded by ANR (Agence Nationale de la Recherche) in France (project ANR-11-BS56-0013), and by DFG (Deutsche Forschungsgemeinschaft) in Germany

References (86)

  • J.-E. Martelat et al.

    Strain pattern and late Precambrian deformation history in southern Madagascar

    Precambrian Res.

    (2000)
  • M.J. Pratt et al.

    Shear velocity structure of the crust and upper mantle of Madagascar derived from surface wave tomography

    Earth Planet. Sci. Lett.

    (2017)
  • E.J. Rindraharisaona et al.

    Earth structure and instrumental seismicity of Madagascar: implications on the seismotectonics

    Tectonophysics

    (2013)
  • A. Tommasi

    Forward modeling of the development of seismic anisotropy in the upper mantle

    Earth Planet. Sci. Lett.

    (1998)
  • A. Wüstefeld et al.

    Identifying global seismic anisotropy patterns by correlating shear-wave splitting and surface–wave data

    Phys. Earth Planet. Inter.

    (2009)
  • A. Wüstefeld et al.

    SplitLab: a shear-wave splitting environment in Matlab

    Comput. Geosci.

    (2008)
  • D. Alsina et al.

    Small-scale sublithospheric continental mantle deformation: constraints from SKS splitting observations

    Geophys. J. Int.

    (1994)
  • C. Amante et al.

    NOAA Technical Memorandum NESDIS NGDC-24

    (2009)
  • D.F. Argus et al.

    Geologically current motion of 56 plates relative to the no-net-rotation reference frame

    Geochem. Geophys. Geosyst.

    (2011)
  • Barruol, G., 2014. RHUM-RUM Marion Dufresne MD192 cruise report Sept.–Oct. 2012....
  • G. Barruol et al.

    Investigating La Réunion hot spot from crust to core

    Eos

    (2013)
  • G. Barruol et al.

    Mapping upper mantle flow beneath French Polynesia from broadband ocean bottom seismic observations

    Geophys. Res. Lett.

    (2009)
  • T.W. Becker et al.

    Toward a generalized plate motion reference frame

    Geophys. Res. Lett.

    (2015)
  • M. Beyreuther et al.

    ObsPy: a Python toolbox for seismology

    Seismol. Res. Lett.

    (2010)
  • P. Bird

    An updated digital model of plate boundaries

    Geochem. Geophys. Geosyst.

    (2003)
  • P.D. Bissessur

    Structure, Age and Evolution of the Mascarene Basin, Western Indian Ocean

    (2011)
  • I.T. Bjarnason et al.

    Shear wave splitting across the Iceland hot spot: results from the ICEMELT experiment

    J. Geophys. Res.

    (2002)
  • M. Bodmer et al.

    Seismic anisotropy beneath the Juan de Fuca plate system: evidence for heterogeneous mantle flow

    Geology

    (2015)
  • F. Brenguier et al.

    First results from the UnderVolc High Resolution Seismic and GPS network deployed on Piton de la Fournaise volcano

    Seismol. Res. Lett.

    (2012)
  • G. Burgos et al.

    Oceanic lithosphere–asthenosphere boundary from surface wave dispersion data

    J. Geophys. Res.

    (2014)
  • M. Cannat et al.

    Formation of the axial relief at the very slow spreading Southwest Indian Ridge (49° to 69°E)

    J. Geophys. Res.

    (1999)
  • J.A. Collins et al.

    Shear wave splitting at the Hawaiian hot spot from the PLUME land and ocean bottom seismometer deployments

    Geochem. Geophys. Geosyst.

    (2012)
  • C.P. Conrad et al.

    Constraints on lithosphere net rotation and asthenospheric viscosity from global mantle flow models and seismic anisotropy

    Geochem. Geophys. Geosyst.

    (2010)
  • E. Debayle et al.

    An automatically updated S-wave model of the upper mantle and the depth extent of azimuthal anisotropy

    Geophys. Res. Lett.

    (2016)
  • R.A. Duncan

    40Ar/39Ar geochronology of basement rocks from the Mascarene Plateau, the Chagos Bank, and the Maldives Ridge

    Proc. Ocean Drill. Program Sci. Results

    (1990)
  • R.A. Duncan et al.

    The volcanic record of the Reunion Hotspot

    Proc. Ocean Drill. Program Sci. Results

    (1990)
  • F.R. Fontaine et al.

    Crustal and uppermost mantle structure variation beneath La Réunion hotspot track

    Geophys. J. Int.

    (2015)
  • F.R. Fontaine et al.

    Upper-mantle flow beneath French Polynesia from shear wave splitting

    Geophys. J. Int.

    (2007)
  • F.R. Fontaine et al.

    Shear-wave splitting beneath the Galápagos Archipelago

    Geophys. Res. Lett.

    (2005)
  • A.E. Gripp et al.

    Young tracks of hotspots and current plate velocities

    Geophys. J. Int.

    (2002)
  • N. Harmon et al.

    Variations in shear-wave splitting in young Pacific seafloor

    Geophys. Res. Lett.

    (2004)
  • É. Kaminski et al.

    Timescales for the evolution of seismic anisotropy in mantle flow

    Geochem. Geophys. Geosyst.

    (2002)
  • J.-M. Kendall et al.

    Mantle upwellings, melt migration and the rifting of Africa: insights from seismic anisotropy

    Geol. Soc. (Lond.) Spec. Publ.

    (2006)
  • Cited by (17)

    • Mantle flow under the Central Alps: Constraints from shear-wave splitting for non-vertically-incident SKS waves

      2022, Physics of the Earth and Planetary Interiors
      Citation Excerpt :

      Such a contribution has been recently revealed by Barruol et al. (2019) investigating a possible plume-ridge interaction, linked by a channel flow near Réunion, as proposed by Morgan (1978, 1971). Revealing a coincidence between SWS (based on Scholz et al., 2018 and surface-wave fast orientations (based on Mazzullo et al., 2017, this study has found a roughly 100-150 km thick asthenospheric flow narrowing eastwards toward the ridge. The related depth distribution of fast orientation showed a spindle-shaped behavior.

    • Nature of the crust beneath the islands of the Mozambique Channel: Constraints from receiver functions

      2021, Journal of African Earth Sciences
      Citation Excerpt :

      From the Middle Jurassic to the Early Cretaceous, the Davie Ridge accommodated the opening of two basins (Coffin and Rabinowitz, 1987). In the north, the Somali Basin resulted from the southward drift of Madagascar leading to oceanic accretion between ∼150 and 120 Ma (Ségoufin and Patriat, 1981) with N–S crustal extension along E-W trending ridges (Davis et al., 2016) supported by magnetic anomalies (Phethean et al., 2016) and SKS splitting measurements (Scholz et al., 2018) compatible with ridge-parallel asthenospheric flow. In the south, the Mozambique Basin is due to the separation of Africa and Antarctica that started at the end of the Jurassic (König and Jokat, 2010; Leinweber et al., 2013) and finished during the Early Cretaceous (e.g. Rabinowitz et al., 1983).

    • Depth dependent azimuthal anisotropy in Madagascar island from ambient noise tomography

      2020, Tectonophysics
      Citation Excerpt :

      Using Rayleigh and Love wave dispersion measurements extracted from seismic ambient noise correlations, they obtained radial anisotropy patterns that broadly reflect the different geodynamic histories of the Morondava basin and the Precambrian shield. Seismic anisotropy have also been investigated for some of Madagascar's neighbouring regions like Seychelles (e.g., Barruol and Ismail, 2001), La Reunion (e.g., Barruol and Hoffmann, 1999; Barruol and Fontaine, 2013; Mazzullo et al., 2017), western Indian ocean (e.g., Scholz et al., 2018), Ethiopia (e.g., Hammond et al., 2005), East Africa (e.g., Bagley and Nyblade, 2013), etc. In East Africa, a NE-SW aligned fast polarization direction consistent with mantle flow from the African superplume was observed.

    • A trace of recycled continental crust in the Réunion hotspot

      2019, Chemical Geology
      Citation Excerpt :

      Small variations in the historical volcanic activity of the Réunion hotspot have been interpreted as being due to interaction of Réunion magmas with the oceanic crust or the volcanic edifice (Pietruszka et al., 2009). In contrast with this apparently simple geochemical model, recent seismic tomography and SKS splitting results depict the Réunion hotspot as a complex low-velocity structure interacting with the Central Indian Ridge (Mazzullo et al., 2017; Scholz et al., 2018). In addition, the recent discovery of Archean zircons in Mauritius has raised the possibility that plume magmas may also be contaminated by fragments of continental crust left behind when India separated from Madagascar (Torsvik et al., 2013; Ashwal et al., 2017).

    View all citing articles on Scopus
    View full text