The 26 December 2004 Sumatra earthquake (Mw = 9.1) initiated around 30 km depth and ruptured 1300 km of the Indo-Australian Sunda plate boundary. During the Sumatra OBS (ocean bottom seismometer) survey, a wide angle seismic profile was acquired across the epicentral region. A seismic velocity model was obtained from combined travel time tomography and forward modeling. Together with reflection seismic data from the SeaCause II cruise, the deep structure of the source region of the great earthquake is revealed. Four to five kilometers of sediments overlie the oceanic crust at the trench, and the subducting slab can be imaged down to a depth of 35 km. We find a crystalline backstop 120 km from the trench axis, below the fore arc basin. A high velocity zone at the lower landward limit of the raycovered domain, at 22 km depth, marks a shallow continental Moho, 170 km from the trench. The deep structure obtained from the seismic data was used to construct a thermal model of the fore arc in order to predict the limits of the seismogenic zone along the plate boundary fault. Assuming 100C-150C as its updip limit, the seismogenic zone is predicted to begin 530 km from the trench. The downdip limit of the 2004 rupture as inferred from aftershocks is within the 350C 450C temperature range, but this limit is 210-250 km from the trench axis and is much deeper than the fore arc Moho. The deeper part of the rupture occurred along the contact between the mantle wedge and the downgoing plate.
Subduction megathrusts produce by far the largest earthquakes on earth (Ruff and Kanamori, 1972;Stein and Okal, 2005). These large ruptures also cause significant vertical motions that can generate devastating tsunamis (Savage, 1983;Satake, 1993;Johnson et al., 1996;Fuji and Satake, 2007). Together these phenomena pose a substantial threat to 1 populations and infrastructure located in coastal regions, such as most of the Pacific rim. The surface area of the earthquake fault plane, together with the amount of slip, controls the magnitude of the earthquake (Wells and Coppersmith, 1994). The portion of the fault plane, which ruptures, is said to be "seismogenic" and is bounded by an updip and a downdip limit (Tichelaar and Ruff, 1993;Oleskevich et al., 1999). The position of the updip limit exerts a strong control on tsunami generation and the location of the downdip limit which commonly lies close to the coast, influences the intensity of ground shaking here. Therefore, it is of critical importance for the estimation and mitigation of these natural hazards to have reliable estimates of these updip and downdip limits. However, there is still no consensus on the physical processes which control these limits, and thus large uncertainties still exist for most zones which have not experienced a great earthquake in recent times, e.g. Cascadia (Hyndman and Wang, 1995;Khazaradze et al., 1999;Stanley and Villasenor, 2000). This work focuses on the SE end of the rupture zone of the great SumatraAndaman earthquake of 26 Dec. 2004, where the event initiated. We present new deeppenetration seismic data to constrain the geometry and structure of the crust and upper mantle. Additionally, numerical modeling of forearc thermal structure is performed in order to calculate the thermally expected limits of the seismogenic zone. These results are compared to the observed distribution of aftershocks and to published sourcerupture models in order to determine where the earthquake rupture initiated. The implications for the control of seismogenic rupture by the lithologies and temperatures along the plate contact are discussed.
One of the first deep seismic studies on the Sumatran subduction zone consisted of a two ship survey in 1966/67 (Curray et al., 1977). Their preferred interpretation of a profile off central Java shows minor thickening of the oceanic crust at the trench and slightly seawards which they propose to be caused by faulting. Later studies were carried out by Kiekhefer, et al., 1980 using free floating sonobuoys, airguns and explosives on 5 marginparallel lines located offshore Nias and in the Nias basin. Based on the analysis of this dataset they propose a shallow Moho depth beneath the continental crust of only about 20 km along this segment of the Sumatra forearc.
During the SUMENTA I and II cruises in 1992 over 5000 km of 6 channel seismic data were acquired offshore Sumatra (Izart et al., 1994), which helped to constrain the existence of the Mentawai microplate formed by partitioning of the oblique convergence of the subduction and strikeslip deformation of the upper plate (Malod and Kemal, 1996).
More recent multichannel and wideangle reflection seismic studies of the SESumatra to Java portion of the Sunda margin were carried during the GINCO project with the German research vessel Sonne in 1998 and 1999. Based on these data two main phases of forearc basin and accretionary prism evolution were identified (Schlüter et al, 2002). Analysis of the wideangle data provided images of the subducting plate to a depth of 20 km and additionally the crustal structure was modeled to a depth of 30 km using gravity data (Kopp et al., 2001).
Their preferred model includes a shallow Moho at only 15 km depth beneath the Java margin (Kopp et al., 2002).
The 26th December 2004 earthquake (Mw = 9.19.3) is among the four largest earthquakes ever recorded and the largest of the last 40 years (Lay et al., 2005;Stein and Okal, 2005). The earthquake initiated off the NW Sumatra margin in the vicinity of Simeulue Island, and ruptured northwards past the Nicobar and Andaman islands, along 1300 km of the IndoAustralian/Sunda plate boundary (Lay et al., 2005). Source time studies (Ammon et al. 2005) as well as geodetic (Vigny et al., 2005) and tsunami inversions (Fuji and Satake, 2007), indicate that the rupture zone was widest (up to 200 km), coseismic slip was largest (locally over 20m) and thus seismic moment release was greatest, in the region directly west of the NW tip of Sumatra, near Banda Aceh (Figure 1).
Several marine geophysical surveys were conducted after the earthquake on the Sumatran margin. During the first cruise onboard the British military vessel HMS Scott multibeam bathymetry data were acquired on the southern part of the 2004 earthquake rupture zone (Henstock et al., 2006). From February March 2005 a Japanese oceanbottom seismometer array was deployed to record the numerous aftershocks of the earthquake. The a
This content is AI-processed based on open access ArXiv data.
