During the last decade, the Aceh (2004, Mw>9.1), Maule (2010, Mw 8.8) and Tohoku-Oki (2011, Mw 9) earthquakes dramatically shed light on the seismogenic and tsunamigenic hazard due to subduction megathrusts. The recent Japan event revealed fundamental misunderstandings on the mechanics of the subduction and on the way mega-events may repeat in time and space, leading to a failure in correctly assessing and forecasting hazards. To significantly move ahead on the understanding of the mechanics of megathrust earthquakes at different time scales we propose to study the Mw 8.8 Maule mega earthquake of February 2010 in every possible detail and to put it into the perspective of the geodynamics of the Andean margin.
|The Maule earthquake of 27 february 2010. Map showing coupling inverted from pre-seismic GPS data (Métois et al. 2011), slip distribution inverted from co-seismic geodetic (GPS+INSAR) data (Vigny et al. 2011), and historical earthquake segmentation.
Past major earthquakes on the
Chilean subduction trench
For the specific study of the Maule earthquake we have acquired a substantial coseismic data: campaign GPS, high dynamic range accelerograms, classical strong motion instruments, 1Hz continuously recording GPS antennas. This unique data set will be used to study the rupture process of the earthquake. We hope to explain why such a large event produced moderate strong motion. Is this a unique feature of this event, or is it typical of most megathrust earthquakes? We need to study the slip distribution of the main event and understand why the aftershocks stretched over an area substantially longer than the main rupture. The aftershock series did not contain any events larger than Mw 7.1, a puzzling feature indeed. We dispose of aftershock recordings obtained by the CNRS-INSU researchers and our foreign and Chilean colleagues. Are these aftershock uniformly distributed or highly concentrated in a few asperities as preliminary results seem to suggest?
With the 2011 Japan event, Maule 2010 is the only mega-earthquake that occurs in a closely surveyed area. Published studies pointed out that the interplate zone was completely locked not only in the so-called «Darwin gap» but well outside it, suggesting that historical earthquakes can serve to identify gaps, but not to determine their actual size. Our rupture model (Vigny et al., Science, 2011) supports this conclusion and shows that the whole subduction plane broke for about 500km along strike, and from 40km depth to its very shallow parts near the trench. A conclusion also reached for Japan, which requires changes in our understanding of shallow rupture propagation and coupling. We also anticipate new advances on the comprehension of after slip that will be studied with a rich set of cGPS data: where does it occur and what are the properties of faults and bulk rheology that produce long episodes of silent deformation around earthquakes. This problem is closely related to that of segmentation; there is substantial evidence that the Chilean margin has long standing segments associated both with features of the oceanic plate (seamounts, ridges) and of the overriding plate (faults, uplifted terraces...). Do they actually play a role in stopping earthquakes or at least in fragmenting them? The Mejillones, Arauco and Talinay peninsulas are long standing features where earthquakes seem to stop or slow down as they reach them. We plan to explore their role using tectonic, geological and geodetic methods.
The Maule earthquake is a rare event; the lessons to be learned from the study of this earthquake will most certainly have a lasting influence in the way geoscientists and engineers approach the problem of the occurrence of these events.