Rapid Imaging of Large Earthquake Rupture Zones with P waves: Application to the 28 March 2005 Sumatra Mw 8.7 Earthquake Suggests Bilateral Rupture - Figure 2 Figure 2. P-waves used to create the image in Figure 1. These seismograms were recorded on vertical component stations of the Global Seismic Network. Traces are sorted by station azimuth from the epicenter and aligned with a waveform cross-correlation method using a 15 s window beginning at the predicted IASP91 travel time.
We image the rupture zone of the March 28, 2005, Sumatra Mw 8.7 earthquake by directly back-projecting teleseismic P waves to their source. We use broadband, vertical-component seismograms recorded by the Global Seismic Network and the Japanese Hi-net network. Our resulting images agree favorably with the aftershock distribution and the location of the Harvard centroid moment tensor. The back-projected energy suggests that the rupture was 120 s long and propagated at 2.9-3.3 km/s from the hypocenter in two directions delayed by about 50 s: first towards the northeast for about 100 km and then toward the southeast for about 200 km. The seismic radiation throughout the rupture zone is characterized by periods between 2-10 s. However, the southern half of the rupture zone generated additional longer-period energy between 10-30 s. The rupture occurred over a surface area of about 70,000 km2, which is fairly consistent with a Mw 8.7 event on a nearly horizontal fault plane. It appears that this earthquake is similar to the 1861 Mw 8.3-8.5 thrust event in its location, size, and geometry.