Teleseismic P wave imaging of the 26 December 2004 Sumatra-Andaman and 28 March 2005 Sumatra earthquake ruptures using the Hi-net array

摘要

Seismograms from a dense, high-quality seismic network in Japan are used to investigate the characteristics of the 26 December 2004 Sumatra-Andaman and the 28 March 2005 Sumatran earthquakes. The onset of the P waveforms are aligned through cross correlation, and a simple concept of back-projecting seismic energy to a grid of potential source locations is applied. The waveform alignment removes the effects due to lateral variations in wave speed between the hypocenter and each station. To better approximate the effects of three-dimensional heterogeneity for paths originating from grid points away from the hypocenter, cross-correlation results of the P waveforms from aftershocks are introduced. This additional information leads to improved resolution of smaller-scale features near many of the aftershocks by reducing wavefront distortion. The back-projection analysis provides a quick assessment of the spatiotemporal extent and variability of relative high-frequency energy release, which can be translated into an estimate of the moment magnitude, as well as an unparalleled view of high-frequency rupture propagation. The results are, in general, consistent with those obtained from more involved source inversion methods. The 2004 Sumatra-Andaman earthquake released most energy in a region northwest of the Sumatra island and the rupture extended to the northern Andaman islands, about 1300 km from the epicenter. This northern portion of the rupture radiated a considerable amount of energy, but there is little evidence of slow slip. The 2005 event is imaged to have bilateral rupture with northwestern slip occurring for about 50 s before it moved to the southeast of the epicenter.