The main tsunami was produced by a virtually instantaneous elevation and subsidence pattern on the sea floor caused by slip on the interface between the Indian plate and the microplate to the north and east. This can be emulated in a computer as a series of pistons placed on the sea floor that suddenly heave the water from its original position at rest, to the new sea floor shape caused by the earthquake. The computer then calculates the propagation path of the resulting wave-form using the depth of the shallow sea and deep ocean floor, and the beach shelf data, or harbour depth data where these are known. The tsunami travels at a velocity that is proportional to the square root of the depth of the water column. By adjusting the location and amplitude of the source region it is possible to match the observed data and the computer-modeled output. A preliminary model has been completed by Dr. Modesto Ortiz (ortizf@cicese.mx), and is summarized here.

An estimate of the initial northernmost extent of the sea floor deformation that caused the tsunami can be obtained using the closest tide gauge in India at Vishakapatnan. Modesto Ortiz' model shows that reasonable agreement occurs if the source extends 600-650 km northward from the epicenter. This initial rupture apparently stopped just south of Car Nicobar, near the southernmost edge of the Mw=7.9 1881 earthquake. This tsunami model is consistent with subsidence in the Nicobar Islands. It is possible that the enigmatic slow propagation of the rupture to the north, suggested by the earthquake aftershock distribution, occurred too slowly to produce a coherent tsunami, although details of its propagation may be contained in the later arriving tsnami data.