IODP Exp. 362 Sumatra Seismogenic Zone - Post-cruise research

The giant 2004 Mw 9.2 Sumatra-Andaman earthquake ruptured the greatest fault length of any recorded earthquake spanning a distance of 1300 km, longer than the UK from the Shetlands to Cornwall. The rupture initiated about 30 km below the ocean floor along the megathrust that separates the Indian plate from the overlying Burma-Sunda plate. It propagated northwards for 10 minutes at a maximum speed of 2.5 km/sec. Fault slip was as large as 15 m to the west with fault-slip-linked uplift reaching 2 m. The displacement of the ocean floor was a sudden 'shock' that created the devastating tsunami waves.
A comprehensive view of the variable megathrust slip characteristics and distribution was obtained by combining tsunami, seismologic and geodetic observations. In the southern Sumatra-Andaman region co-seismic slip reached the near-trench portion of the plate boundary. Three months later, in March 2005, the Mw 8.7 Nias earthquake ruptured the central portion of the Indian/Burma-Sunda megathrust. This earthquake produced a smaller, but still deadly tsunami. In the Nias segment earthquake slip concentrated below the outer forearc.
The potential for seismic near-trench slip in the Sumatra-Andaman segment and the more general internal variability in rupture speed along the plate boundary has been correlated with different frictional properties of the material hosting the plate boundary interface. Megathrusts form part of the subduction conveyor belt, along which incoming sediment is ultimately transported into the zone associated with Earth's most destructive earthquakes. To assess the potential for rupture propagation near the trench the geological, physical and mechanical properties of incoming plate material should be considered together with the distribution of excess fluid pressure, the complexity of the rough plate interface geometry, and the slip deficit rate at the megathrust interface.
In this study, new samples recovered by IODP Exp. 362 from incoming Indian plate sediments will be used to investigate the frictional behavior of these sediments en-route to the plate boundary interface.
The UKIODP Moratorium award will be used to achieve 5 Strategic Objectives:
SO 1 - identify the characteristics of the sedimentary sections to optimize the sampling strategy for friction experiments based upon onboard sedimentology, petrology, geochemistry and in-situ logging profiles. This objective will benefit from the Exp. 362 Science Party work carried out during the cruise.
SO 2 - acquire a quantitative data set of the mineral assemblages, fluid permeability, and porosity for each sample.
SO 3 - measure their frictional dependence on slip, slip rate, slip velocity, and normal stress by performing experiments on the collected samples under deformation conditions typical of earthquakes using the high velocity rock friction apparatus SHIVA.
SO 4 - monitor gas emission, humidity, and temperature variations during friction experiments using mass spectrometry, temperature and humidity measurements with the sensors installed on the rotary shear apparatus.
SO 5 - analyse the experimental fault rock material using a multidisciplinary approach that involves microstructural analysis, mineralogy, and petrology so that proxy records may be reconstructed for plate interface seismic slip. Validate these against the seismological record.
The ultimate goal is to incorporate the actual physical properties of the Sumatra-Andaman incoming sedimentary section within an improved theoretical earthquake rupture propagation model.
This research will develop a new approach to the assessment of extreme near-trench tsunamogenic slip based on the analysis of incoming plate sediments. This approach is also applicable to other plate-boundary megathrusts (e.g. Japan Trench, Barbados). Future studies can also consider possible lateral variations in the lithological composition of the incoming plate/subduction plate boundary material.

Since 1900 about half million people have lost their lives from earthquake-generated tsunamis (source: USGS). The 2004 Sumatra-Andaman tsunami had a severe humanitarian and political impact not only on the countries directly affected by the earthquake and tsunami, but also on countries far away. The UK, for example, had 149 victims. Sweden, the hardest hit country outside Asia, had 543 victims. Great earthquakes such as the Sumatra-Andaman event are produced by megathrust faults slipping in subduction zones. The 1960 M 9.5 Great Chilean EQ, the 1964 M 9.2 Great Alaska EQ, the 1952 M 9.0 Kamchatka EQ and the 2011 Tohoku EQ together with the Sumatra-Andaman EQ account for more than half of the total earthquake moment in the last 100 years. Each of these very large megathrust earthquakes also spawned tsunamis.

This research will contribute to the global questions about when an earthquake will potentially develop a devastating tsunami. What is special about the sediments entering in the subduction system that can allow co-seismic propagation of rupture to the trench in some places, but not others? How frequently has this happened in the past? Through the integration of this study with companion studies addressing the mechanisms of earthquake nucleation in Sumatra and in other regions around the world, we can add new insights to earthquake and tsunami forecasting and mitigation.

The results of IODP Exp. 362 and this proposed research will expand our current knowledge of rupture propagation and slip during megathrust earthquakes at the very moment when the exceptional datasets acquired from the 2004 Sumatra and 2011 Tohoku earthquakes are showing for the first time that the seismogenic zone of a subduction megathrust can be very shallow and even reach the trench, a domain that was considered, until recently, as being able to only slip aseismically. Although this is not the only mechanism to create a tsunami, shallow rupture during megathrust earthquakes can exponentially multiply the volume of water mobilized to generate tsunami waves. This project will make significant scientific advances towards our understanding of the role of the subducting sediments for the seismogenic characteristics of plate boundaries.
Here we present a new approach to earthquake and tsunami forecasting. While the immediate beneficiary of outputs from this work will be the academic community, IODP and UKIODP have a track record of successfully feeding into research agendas and government policy in relation to hazard prevention. This project will build on the experience gained through research in Costa Rica and Japan, where informing local interest groups in our research has been a high priority. The SEARG group at RHUL already works in South East Asia and members of this group have a keen interest in this project. Through them and their network built in >20 years, we will also work with local communities (local councils, tourist office, schools, landowners and guides) in South-East Asia.
IODP has already assembled an international team of leading experts in subduction systems and megathrust mechanics, providing an exceptional opportunity to learn and cross fertilise ideas between these diverse scientific communities. We will employ our innovative ideas using INGV's world-leading laboratory facilities. We have pioneered the concept of weak megathrusts being able to propagate seismic rupture to the trench (Vannucchi and Leoni, EPSL, 2007) and we will build on this, for example highlighting the value of petrography and microstructural stratigraphy as an essential step to generate robust rupture dynamic reconstructions.
The academic impacts of this research are far reaching. The skills learnt by the PI and collaborators will feed back into the newly launched "Hazard Group" at RHUL, and improve its teaching and learning.