Rupture models
Conventional seismic define the first few minutes of the rupture. Its northward progression is currently less well constrained.
The rupture appears to have propagated northward for at least 8 minutes judging from the length of the p-wave train recorded at sufficient epicentral distances for these waves not to overlap with the first s-wave arrivals. (see annotated seismograms by Lomax and Park), implying NNE rupture propagation at velocities close 2.5 km/s. Distinct large aftershocks started sooner near the mainshock (25 minutes) than at the rupture's northern end (85 minutes) possibly due to smaller amounts of static and dynamic strain released by the rupture at each end of the rupture.
Three models of the main rupture (the first 200s) can be viewed at these web sites.
Y. Yamanaki (Tokyo) http://www.eri.u-tokyo.ac.jp/sanchu/Seismo_Note/2004/EIC161e.html
Chen Ji (Caltech) http://www.gps.caltech.edu/~jichen/Earthquake/2004/aceh/aceh.html
Yagi (IISEE) http://iisee.kenken.go.jp/staff/yagi/eq/Sumatra2004/Sumatra2004.html
A rupture propagation velocity of ≈3.5 km/s would have shifted the plate boundary near Port Blair in the Andaman Islands roughly 5 minutes after the mainshock. The first large aftershock near Port Blair, however, occurred after 83 minutes. (large jpg). Currently the timing of smaller aftershocks is unavailable.
Propagation of afterslip appears to occur in a domino-like fashion northwards, and southwards, at rates typical of creep processes on the central San Andreas fault. A creep rate of 22-24 cm/day is suggested by the larger aftershocks.
Tsunami model calculated by Kenji Sataki
http://staff.aist.go.jp/kenji.satake/animation.gif
Three tsunami models of waves propagating across the Bay of Bengal have been released by NOAA
