HR: 0800h
AN: U11B-0833    [Abstracts]
TI: Four Hypotheses on the Indian Ocean Tsunami Caused by The 26 December 2004 Earthquake
AU: * Galvin, C
EM: galvincoastal@juno.com
AF: Cyril Galvin, Coastal Engineer, Box 623, Springfield, VA 22150 United States
AB: The following ideas come from a close reading of reports on the Sumatra earthquake of 26 Dec 04 and its resulting tsunami, and from experimenting with and observing water waves: 1. A widely reproduced illustration combines a computer model of the tsunami in the Indian Ocean with the water surface profile across that ocean sensed by the altimeter on the Jason 1 satellite. The satellite track crosses the radiating tsunami twice in ten minutes, first where the tsunami moves south of the latitude of Sri Lanka and then where it moves north into the upper Bay of Bengal. There is difficulty fitting the model to a large trough in the Bay of Bengal. The ocean surface profiles from the satellite altimeter appear to be plotted to balance crest and trough displacements about a mean sea level, but if the tsunami is a solitary wave, the displaced water will be entirely above existing sea level. If so, the troughs, which are somewhat flat-bottomed as plotted, should be considered sea level. Then the problem with the anomalous trough in the Bay of Bengal goes away. 2. The vertical component of quick seafloor motion generates the tsunami. If that component is down, a trough is created; if up, a crest. Water surface profiles recorded at shore sites show that the Dec 04 tsunami locally began with troughs or crests. Two notable initial troughs were at Banda Aceh and Phuket, both east of the fault. Most of the best recorded records west of the fault began with crests. This is consistent with field observations in the Nicobar and Andaman Islands which show tilting to the east, that is, relative to mean island level, the seafloor usually went down on the east and up on the west. It also is consistent with seismic wave analyses which show that the major quick horizontal seafloor displacement occurred in the Sumatra segment of the fault toward the west and the southwest. 3. Reports on the Dec 04 tsunami often estimate for it a wave period or wave length. Analysis of the wave form of an arriving tsunami will produce a wave period or wave length, but such analysis forecloses consideration of the essential nonlinearities in a tsunami, which is a shoaling solitary wave. Upon entering 'shallower' water, the phase speed of a solitary wave decreases. Conservation of power transmitted with the wave builds up the crest until the maximum stable elevation of that solitary wave in local water depth is reached. At this point, the solitary wave separates into two or more solitons, arranged in order of decreasing height, traveling at speeds determined by the depths below their respective soliton crests. Laboratory experiments show that the largest separating soliton has a crest elevation actually higher than the solitary wave from which it separates. The observer on shore sees the incoming tsunami as a train of unequal waves. 4. Analyses of the seismic signals show that the initial rupture, after a slow start, unzipped the Sumatra segment of the fault in a south-to-north direction, the rupture traveling north at a relatively constant 2.5km/s for at least 600km, and eventually to 1200km. This rupture traveled far more swiftly than the phase speed of the tsunami. Energy radiated from this moving rupture might act as a wave maker which adds to the the wave produced by the quick vertical displacement of the seafloor. Given the great difference in speeds of rupture and tsunami, the unzipping of the fault would radiate a tsunami moving perpendicular to the fault.
DE: 4217 Coastal processes
SC: Union [U]
MN: Fall Meeting 2005