HR: 0800h
AN: U11A-0819    [Abstracts]
TI: Imaging of Large Earthquake Rupture Processes Using Multiple Teleseismic Arrays: Application to the Sumatra-Andaman Islands Earthquake
AU: * Ohrnberger, M
EM: mao@geo.uni-potsdam.de
AF: Institute of Geosciences, University of Potsdam, Karl-Liebknecht-Str. 24/25, Golm, 14476 Germany
AU: Krüger, F
EM: kruegerf@geo.uni-potsdam.de
AF: Institute of Geosciences, University of Potsdam, Karl-Liebknecht-Str. 24/25, Golm, 14476 Germany
AB: The spatial extent of large earthquake ruptures is usually indirectly inferred from aftershock distributions or by waveform inversion techniques. In this work we present a method which allows the direct estimation of the spatio-temporal characteristics of large earthquake rupture processes. The technique exploits the high-quality records from the stations of the global broadband network using a simple, yet efficient, migration technique. In particular, we combine coherency and beam-power measures which are obtained from curved wavefront stacking of the direct P wave at multiple large aperture arrays surrounding the source region at tele-seismic distances. Applying this method to the M_w=9.3 Sumatra earthquake from 26/12/2004 and the subsequent Nias earthquake from 28/03/2005 (M_w=8.7), we show that it is possible to track the focus of the most coherent/largest energy release in space and time. For the Sumatra event, we confirm the overall extent of the rupture length being in the order of 1150 km. The rupture front propagated during a time span of at least 480-500 s following the trench geometry from the northern tip of Sumatra to the Andaman Islands region. A visualization of the coherent energy accumulation over time suggests the existence of slow after-slip in the northern part of the rupture after the main rupture front has passed. However, due to the interference of large later phases it is not possible to determine whether this afterslipping event persists much longer then the overall duration of the rupture. The final areal estimate of cumulative energy release is in full agreement with the aftershock distribution observed in the months following this earthquake. Including a number of additional seismic phases (e.g. pP, sP) into the migration scheme, it seems for this event feasible to constrain the depth extent of the rupture. For the Nias earthquake we observe unilateral propagation of the rupture in south-eastern direction starting from an area south-east to the epicentral region of the Sumatra event. The rupture extent and duration of this event is about 25% of its predecessor. The estimation of the rupture duration for these events and their corresponding spatial extent allows deriving approximate source energy estimates during the rupture process and finally provides an appropriate estimate of the total earthquake size within minutes after the rupture onset. Considering the urgent need for early-warning systems for the detection of tsunami-genic earthquakes, we feel that this technique has the potential to become an integral part of such a system.
DE: 7203 Body waves
DE: 7215 Earthquake source observations (1240)
DE: 9340 Indian Ocean
SC: Union [U]
MN: Fall Meeting 2005