MAule eaRthquake: Integration of Seismic Cycle Observations and Structural investigations

Megathrust earthquakes at subduction zones dominate the global moment release and present a grave danger to coastal populations from both shaking and from the tsunamis they generate. These subduction zone earthquakes (M_w>~8) usually break the whole width of the seismogenic zone in dip direction, and exhibit several local maxima of coseismic slip (‘asperities’) in the strike-direction. Slip in these earthquakes releases not only critical levels of shear stress at the plate interface, but also imposes stress changes in adjacent fault segments and the surrounding volume, triggering aftershocks and post-seismic processes such as afterslip, and visco- and poroelastic relaxation. The equivalent moment of these processes can exceed 50% of the seismic moment release of the main shock, but their relative importance is an open issue.

Although there are indications that there is some consistency between asperities in multiple seismic cycles and some correlation with the level of interseismic locking, the degree to which these patterns are stable has not been quantified. It is also unclear under what circumstances the gaps between asperities act as barriers to rupture propagation, or alternatively act as nucleation points. The way barriers operate is also still a matter of debate: are they points of weakness, where strain is continuously released in the interseismic period, such that no slip deficit can build up? Or do they correspond to heterogeneous stress levels, preventing great earthquakes from rupturing through, and instead breaking in large but not quite great earthquakes (M_w=7-8), or does their material state foreclose velocity weakening thereby preventing unstable rupture propagation?

Given the uncertainty about the way barriers operate it is no surprise that we also have no real understanding of the underlying causes which control the distribution of barriers and asperities. Candidate hypotheses are varying degrees of geometric complexity of the plate interface dependent on incoming seafloor relief, the nature and thickness of the sedimentary pile, perturbations to the amount of fluid release at different depths due to varying incoming plate structures, structures within the overriding plate. All these determine the stress, mechanics and hydraulic properties of the subduction channel itself in terms of frictional strength and stability. One of the main reasons for our limited understanding is the lack of sufficient dynamic (e.g. stress, seismicity) and kinematic data (e.g. GPS, InSAR) from all three phases of the seismic cycle, i.e., the pre-, co- and postseismic stage.

The M_w=8.8 Maule earthquake was the first megathrust event for which such a large amount of high-quality data is available for all stages and thus is a unique opportunity to deepen our understanding of the geodynamic processes that are involved in the seismic cycle (the 2011 Tohuku earthquake represents another extremely well recorded earthquake, but the main slip occurred further offshore). The goal of the project is to reveal structural details of the Maule subduction zone, detect tempo-spatial changes of material properties and quantify the relative importance of the processes that are ascribed to the seismic cycle by means of 3D geomechanical numerical models.

2. Key Objectives

In order to make progress in the questions raised in the introduction it is necessary

• to determine how the strain-deficit between the asperities and the up- and downdip ends of the coseismic rupture is accommodated in the pre-, inter- and postseismic period,

• to link the short-term geodetic observations to the medium and long-term trends represented in the geomorphological and paleoseismological record,

• to map seismic properties of the plate interface with as high a resolution as possible and link them to the kinematics of motion along the plate interface and within the plates,

• to look for markers of an evolving stress field or fluid motion in the form of time-dependent variations of seismic properties or focal mechanisms,

• to develop predictive models for competing hypotheses that are tested against geological, geodetic and seismological data to identify and quantify physical processes occurring in the seismic cycle.

As most rapid changes occur during the coseismic and the early postseismic phase, observations made just before, during and shortly after the major earthquake are absolutely crucial for our understanding.

3. The Maule Earthquake

The earthquake nucleated near the mouth of the Maule river in south-central Chile near 35.91°S, 72.73°W at ~35 km depth on February 27, 2010 06:34:14 UTC. Rupture propagated trenchwards and bilaterally north and south along strike for a total length of 500-550 km, with most preliminary rupture models showing two distinct asperities to the north and south of the epicentre, with peak slip exceeding 10 m (e.g., Delouis et al., 2010). The global CMT solution indicates a dip of ~18° and a magnitude of M_W=8.8. The ground deformation associated with the earthquake triggered a tsunami, which devastated large portions of the coastline with mean run-ups of 10 m, and caused significant property damage and loss of life as far away as the Juan-Fernández-archipelago 700 km from the rupture zone. In total, the earthquake caused over 500 fatalities, mostly due to the tsunami. Aftershocks occur along or near the plate interface, in the outer rise, and in the overriding continental crust, which hosted the largest aftershock, a shallow normal faulting event with M_W=6.9 (Fig. 1). Few aftershocks occur at depths greater than 50 km, the most active area in the pre-seismic period (Fig. 1). The Maule earthquake ruptured the megathrust segment north of the 1960 Chile earthquake (Fig. 2), but with ~100 km of overlap (Farías et al., 2010). It filled a known seismic gap, most of which has remained unbroken since 1835 and had been inferred to be locked (Ruegg et al., 2009, Moreno et al., 2010).

The Maule earthquake is the fifth largest instrumentally recorded earthquake, and the second largest (after the Aceh-Andaman earthquake of 26/12/2004) since modern space geodetic techniques (GPS, InSAR) have become available. Unlike the Aceh-Andaman earthquake, the whole length of the rupture is adjacent to an accessible coastline with a well-functioning infrastructure. The international scientific community therefore quickly realised the potential of detailed observations of the aftershock sequence and the co- and postseismic deformation. Consequently, a rapid response was mounted involving deployment of additional seismological and geodetic instrumentation, surveying of coastal uplift and surface ruptures of secondary faults, and tsunami run-up heights. In total, ~142 seismic stations were installed along the whole length of the rupture by Chilean, US, German, French and British groups; 40 continuous GPS stations were installed and 42 campaign GPS sites were reoccupied to obtain coseismic displacement (Fig. 3). PI’s of this proposal were heavily involved in these rapid-response efforts and most of the data acquired by other groups are available either through open-data policies or established bilateral arrangements. (Ocean bottom instrumentation will be deployed by Grevemeyer and Flüh in September 2010, Fig. 3.) First results from these efforts are beginning to appear in the literature (Farías et al., 2010; Moreno et al., 2010). Finally, several of the PIs of this proposal were involved in three major interlinked research programs (TIPTEQ – From the Incoming Plate to megathrust Earthquakes - and SFB267 – Deformation processes in the Andes; SFB 574 – Volatiles and Fluids in Subduction Zones) that have studied the South Central Chilean margin acquiring high-quality data from a large number of large-scale geophysical, geodetic and geological studies between 33° and 42°S – i.e. encompassing the 1960 Chile earthquake rupture zone and large parts of the 2010 Maule earthquake. This provides us the unparalleled opportunity to have at our hands detailed constraints on the geometry of the margin system and of its physical and kinematic state prior to the earthquake.

4. Scientific Approach and Aims of the Proposal

A key scientific challenge is to identify and quantify the physical and geological processes that contribute to the seismic cycle at subduction zones and in particular control the segmentation of megathrust events. Two major efforts have to be accomplished: (1) Assembly of high-quality and high resolution data sets that reveal small-scale structural details of the plate interface as well as the detailed seismicity distribution and deformation pattern during the inter-, co- and post-seismic phases. We also need to bridge the gap between the short time-scale of geodetic observations to the long time-scale observations of geological markers. (2) Testing various hypotheses that are candidates for explaining kinematic signals by means of numerical models using forward and inversion techniques. These models must aim at a quantitative understanding of the main processes and assess their relative importance in time and space. An important aspect of this proposal is that the different model concepts presented will rely on the same geometrical and physical data as well as using the same kinematic and dynamic observations to test their validity and to constrain the individual model parameter space. This guarantees that the model results can be compared and that the joint interpretation can rule out deviation due to different data sets used. The available data quantity and quality is unique and we have full access to these data.

In particular we aim to achieve the following:

1. Produce a detailed image of aftershock activity and seismic velocities along the whole rupture zone.

2. Map changes to seismic properties, both coseismically and in the postseismic phase.

3. Identify what features of the current earthquake are repeated in past earthquake cycles (e.g. location of asperities, correlation between coseismic and permanent deformation gradients).

4. Invert the coseismic slip combining GPS, InSAR, tsunami and coastal uplift data, including a quantitative understanding of the uncertainties of the solution.

5. Determine the evolution of strain in the postseismic phase by a joint consideration of geodetic constraints and seismicity.

6. Identify the processes of stress and strain accumulation and release that are involved in the seismic cycle and quantify their relative importance, including upper plate deformation, and perturbations of the stress field in the outer rise.

7. Assess the role of fluid migration throughout the seismic cycle.

8. Determine to which extent kinematic and seismic observations are diagnostic of dynamic processes that drive the seismic cycle, and pinpoint observables that allow discriminating between remaining competing hypotheses.

Aktuelles

Neueste Veröffentlichung:

Lange, D., Tilmann, F., Barrientos, S. E., Contreras-Reyes, E., Methe, P., Moreno, M., Heit, B., Agurto, H., Bernard, P., Vilotte, J.-P., Beck, S. (2012):

Aftershock Seismicity of the 27 February 2010 Mw 8.8 Maule Earthquake Rupture Zone,

Earth and Planetary Science Letters, 317-318C, pp. 413-425, doi:10.1016/j.epsl.2011.11.034.

Moreno, M., Rosenau, M., Oncken, O.:
2010 Maule earthquake slip correlates with pre-seismic locking of Andean subduction zone

Nature, Vol 467, issue 7312, pp 198-202; DOI: 10.1038/nature09349