Earth science

When an oceanic tectonic plate cracks

Article metrics

Analyses of two recent earthquakes of great magnitude show how complex the breaking of the oceanic lithosphere can be, how it is linked to earlier great events and how it triggers seismicity worldwide. See Letters p.240, p.245 & p.250

On 11 April 2012, two great earthquakes of magnitude (Mw) 8.6 and 8.2 struck the northeastern Indian Ocean, a few hundred kilometres off the Sunda Trench that lies just west of Indonesia (Fig. 1). The two earthquakes were close in time and space (2 hours and 185 km apart), and occurred not far from the epicentres of the devastating megathrust earthquakes that hit Sumatra's Aceh region in 2004 (Mw 9.2) and Nias Island in 2005 (Mw 8.7). Fortunately, the 2012 intra-oceanic events were not damaging to human activities, but they will be marked in the annals of seismology for several reasons: first, because of their unusually high magnitudes given that they occurred in the interior of an oceanic tectonic plate, rather than at a plate boundary; second, because the 'strike-slip' faulting mechanism of the events is unexpected for earthquakes of such magnitude; and finally, because of the complexity of the rupture of the intervening plate. These exceptional properties put the Mw-8.6 mainshock into the top-ten list of the world's largest earthquakes since 1900, and promptly aroused the curiosity of seismologists1,2 and those working on plate tectonics, because the events took place in one of the largest and most complex deforming zones in the oceanic lithosphere (Earth's crust and uppermost mantle). Three papers in this issue3,4,5 investigate these unprecedented events.

Figure 1: Seismo-tectonic chart of the India–Australia plate.
figure1

Earthquakes in this region have either strike-slip faulting mechanisms (green dots and stars) with faults oriented parallel to fracture zones, or thrust faulting mechanisms (red dots and stars) with faults generally perpendicular to the fracture zones6,13. Two large earthquakes (green stars) that occurred on 11 April 2012 have now been investigated3,4,5; aftershocks of those earthquakes are shown in white circles14. The magnitudes (specifically the 'moment' magnitudes, Mw) of earthquakes equal to or greater than 6.9 are indicated. Seismic events related to the subduction zone (east of the Sunda Trench, which is depicted as a white, broken line) are not shown, except for the megathrust earthquakes (red stars) at Aceh in 2004 (Mw 9.2) and at Nias in 2005 (Mw 8.7). Events with unknown fault mechanisms within the India–Australia plate boundary are also not shown. Black dashed lines define the limits of the active boundary between the Indian and Australian plates11. Gravity anomalies (measured in mgal) reflect the ocean basin's topography beneath sediments, and are indicated by colours on the map: troughs are associated with negative anomalies, and ridges with positive ones. There is a close relationship between the distribution of earthquakes, particularly those that occurred on 11 April 2012, and the extinct seafloor fabric of the eastern Indian Ocean, such as the long north–south oriented fracture zones (lineated troughs) in the Wharton Basin.

Yue et al.4 (page 245) unravel the extraordinary complexity of the underlying mechanisms of the 2012 events, which display perpendicular (orthogonal) and discontinuous (en échelon) fault ruptures. By analysing short-period and long-period seismic waves recorded by stations in Europe and Japan, the authors identified a sequence of shocks along four fault planes, and estimated the rupture length and amount of slip along the faults. The deduced radiation pattern of the seismic waves outlines the interplay between multiple orthogonal and en échelon faults, whose spatial distribution is consistent with the epicentres of the first-week aftershocks. The orientations of the faults reflect the seafloor's tectonic fabric2,4, with ruptures occurring alternately along faults that extend bilaterally west-northwest and east-southeast (parallel to the abyssal hills) or along faults extending north-northeast and south-southwest (parallel to the fracture zone's orientation), the latter being perpendicular to the former. The high-level seismicity in the Wharton Basin (Fig. 1), as well as seafloor mapping, had already provided direct evidence that the fossil fracture zones — seismically inactive, long linear faults generated by seafloor spreading — in this region are continually reactivated6,7,8,9. But the question remains as to why so much stress can accumulate and provoke such large-magnitude earthquakes, and furthermore in an area that could have been activated by the nearby Aceh megathrust earthquake.

This is exactly what Delescluse et al.3 (page 240) attempt to answer, by investigating the causal relationship between the high-level seismicity along the Sunda Trench (Fig. 1) — in particular the Aceh and Nias megathrust events — and the off-trench intra-oceanic seismicity. From Delescluse and colleagues' Figure 2 (page 241), there is no doubt that the Aceh and Nias events have shaken the Indo–Australian lithosphere which is plunging (subducting) beneath the Sunda plate, and that they triggered a long-lasting, large increase in seismic activity off-trench. By looking at how stresses are released in the vicinity of large ruptures (Coulomb stress changes), the authors show that co-seismic slips of the Aceh and Nias megathrusts can promote left-lateral strike-slip earthquakes with a direction parallel to the orientation of the Wharton Basin fracture zones, similar to the mechanisms observed in the April 2012 events. However, such an approach assumes an elastic rheology, for which stress changes decrease immediately after the earthquake, as a new interseismic stress-loading cycle starts along ruptured faults. Delescluse et al. explain the seven-year lag between the Aceh and April 2012 events by a buffering effect of the viscoelastic properties of the asthenosphere, the region of the mantle that underlies the lithosphere. The authors' model predicts that the maximum post-seismic stresses generated by the Aceh megathrust could have reached the area of the April 2012 earthquakes seven to ten years later.

In turn, the two April 2012 events may have triggered seismicity worldwide. Pollitz et al.5 (page 250) show that the number of remote earthquakes with magnitudes larger than 5.5 and up to 7 increased worldwide nearly fivefold in the six days following the April 2012 events. The authors attribute this effect to the powerful radiation pattern of surface seismic waves generated by the strike-slip mechanisms, but cannot exclude the possibility that the 2012 earthquakes struck during a particularly quiet seismic period, and so triggered an unusually large number of earthquakes that were already on the brink of occurring.

Interestingly, the papers by Delescluse et al. and by Pollitz et al. reveal contrasting short-term and long-term 'domino effects' of major seismic events: the April 2012 events may have triggered quasi-immediate aftershocks far away, whereas the 2004 and 2005 events possibly triggered aftershocks nearby several years later. The two studies emphasize how the rupture mechanism and the tectonic setting of earthquakes govern their ability to trigger further seismic activity through seismic-wave excitation and short-term or long-term stress relaxation

The occurrence off-trench of such major events is perhaps better understood in the context of the high lithospheric stresses caused by differential motion between the Indian plate, which is slowed down in its northward drift by collision with the Eurasian plate, and the Australian plate, which is rapidly subducting underneath Indonesia along the Sunda Trench. As a result, the two plates move towards each other at a rate of 15 millimetres per year in an east-southeast direction10, and have been doing so for at least 10 million years11. The April 2012 events have the largest magnitudes ever recorded in the diffuse (as opposed to narrow) India–Australia plate boundary, where earthquakes of magnitude 7 and greater are not uncommon (Fig. 1).

Do these major intraplate earthquakes indicate the birth of a new, discrete plate boundary? Probably not, because this unusually high-level seismicity occurs over large areas (diffuse boundaries) bounded by otherwise stable and 'quiet' oceanic plate interiors (rigid plates), with all these pieces constituting the composite, deforming Indo–Australian lithosphere11. The April 2012 events illustrate how oceanic lithosphere can fail in complex ways2,4,9 as a result of the fabric structure it has inherited from seafloor spreading12, and they shed new light on its mechanical behaviour and strength, and on earthquake physics.

References

  1. 1

    McGuire, J. J. & Beroza, G. C. Science 336, 1118–1119 (2012).

  2. 2

    Meng, L. et al. Science 337, 724–726 (2012).

  3. 3

    Delescluse, M. et al. Nature 490, 240–244 (2012).

  4. 4

    Yue, H., Lay, T. & Koper, K. D. Nature 490, 245–249 (2012).

  5. 5

    Pollitz, F. F. et al. Nature 490, 250–253 (2012).

  6. 6

    Petroy, D. E. & Wiens, D. A. J. Geophys. Res. 94, 12301–12319 (1989).

  7. 7

    Deplus, C. et al. Geology 26, 131–134 (1998).

  8. 8

    Robinson, D. P. et al. Science 292, 1145–1148 (2001).

  9. 9

    Abercrombie, R. E., Antolik, M. & Ekström, G. J. Geophys. Res. 108, 16–31 (2003).

  10. 10

    Gordon, R. G. et al. Geology 36, 827–830 (2008).

  11. 11

    Royer, J.-Y. & Gordon, R. G. Science 277, 1268–1274 (1997).

  12. 12

    Delescluse, M. et al. Geophys. Res. Lett. 35, L16312 (2008).

  13. 13

    www.globalcmt.org

  14. 14

    http://earthquake.usgs.gov

Download references

Author information

Correspondence to Jean-Yves Royer.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Royer, J. When an oceanic tectonic plate cracks. Nature 490, 183–185 (2012) doi:10.1038/490183a

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.