Urgent assessment of the earthquake and tsunami hazard in western Sumatra following the 09/07 Mentawai Islands earthquake sequence.

Starting with the great Sumatra-Andaman earthquake M=9.2 of 26 December 2004 over 2000km of the Sunda megathrust, where the Indian ocean plate is being forced under the Asian plate, has been ruptured in a series of great M>8 earthquakes. The most recent, of magnitude 8.4, occurred on 12 September 2007. Of the more than 2500km of the megathrust, from the Andaman Islands to Enganno, only 300km remains un-ruptured in the last three years. This segment, under the Island of Siberut is still capable of an earthquake of greater than magnitude 8. It has not failed since 1797, from examination of the way the earth's crust is being deformed in the area we know that it has not experienced any slow (aseismic) slip. Simple calculations, using the known rate of convergence of the plates here, show that all the strain released in 1797 has been regained and it seems clear that it is ripe for another earthquake. This could happen at any time. In addition, it is clear from earlier work and from preliminary investigation of the recent events, that the segment under Siberut is now being stressed from the north by the 2005 event and from the south by the 12/09/07 event. It is likely that such an earthquake might cause a tsunami. The maximum height of Padang, broadside on to the Siberut segment, is only 10m above sea level so a tsunami is a real threat. However, it does not have to be as destructive as the 2004 tsunami. Indeed recent work we have done, funded by NERC in 2006, has shown that the presence of the Islands offshore Padang make the threat of a massive tsunami recede significantly. Our forecasts of smaller waves generated by western Sumatran megathrust earthquakes would appear to have been borne out by preliminary data from tsunami generated by the recent M8.4 earthquake which only reached 4m in height and apparently killed no one. Recently, we have made predictions relating the height of a tsunami observed on the Sumatran Coast to the vertical movement experienced there. When one of these earthquakes occurs the offshore islands are forced upward by the earthquake while the coast of Sumatra always drops / sometimes by metres. By a detailed series of coupled earthquake and tsunami simulations we have shown that the wave height experienced onshore is directly proportional to the size of this drop. The beauty is that the size of the vertical movement can be measured up to 30 minutes before the wave arrives. The recent earthquake should allow us to test this relationship. In this project we aim to do three things. Firstly, we plan to reassess the stresses under the island of Siberut accounting for the entire earthquake sequence. These calculations, particularly the time varying components, are complex and require sophisticated computations which we will finalise during the project. Secondly, we will test the predictions of our tsunami simulation methods against data collected by the recent M8.4 event and investigate the relationship between wave height and vertical displacement. This will involve a 2 week field campaign to collect information about the wave heights near the city of Bengkulu which was hit by the recent tsunami. Scars left by the tsunami will still be clearly visible for the next couple of months and can be measured accurately in the field. Using high resolution GPS equipment, which we will borrow from the NERC Geophysical Equipment Facility, we will accurately locate the positions of wave height estimates and also measure the present height of the coast. These heights will allow us to estimate the change in height during the earthquake. Finally, we will compute another suite of tsunami scenarios taking into account the recent earthquakes and the results of our validation work. Western Sumatra is currently one of highest earthquake risk areas in the world. This project will not stop the earthquake but it will contribute to a body of scientific knowledge which will help to prepare people for it.