Tsunamis - Extreme Events: A Physical Reconstruction and Risk Assessment - page 118

larger diameter scours occurred away from trees (Fig 5.5). The form and location
of these erosional features suggests that vortices developed within the tsunami
flow. The long scarp was probably developed by a horizontal 'roller' type vortice,
whereas the scour pits appear to have developed by vertical vortices occurring
both within the flow, and also as a function of flow separation around trees
where eddies developed within a zone of lower pressure on the lee side of the
trees.
Sato et al .(1995)also report scour holes between 0.5 and 1.5 m deep in a
gravel terrace following the 1993 SW Hokkaido tsunami. The tsunami scour
holes developed where a 50 m length of breakwater collapsed concentrating the
outgoing tsunami through a narrow opening. About 1500 m 3 of gravel from a
Holocene coastal gravel terrace was eroded to produce the bowl shaped depres-
sion. Other erosional features were noted by Sato et al .(1995)onOkushiri Island.
Here the tsunami eroded grooves several metres in diameter trending parallel to
theshore. These grooves occurred in two zones both oriented parallel to shore
with a maximum depth of 40 cm. They were located about 20--40 m inland of a
shore protection wall. One of the groups of grooves formed as water swept over
thewall, presumably due to turbulence generated as the tsunami flow separated
from the wall itself.
palaeotsunami
Atwater (1987)undertook one of the earliest studies of a prehistoric
tsunamic deposit. He noted a small forest of dead trees behind a coastal barrier
on the coast of Washington State, USA (Atwater and Moore, 1992), where the
tree bases are now standing within the intertidal zone. These trees do not nor-
mally grow within this zone and Atwater hypothesised that the trees had been
drowned after subsidence of the coastal land most likely during an earthquake.
He also noted a sand layer in the sediments surrounding the trees. The sand
layer occurred within much finer-grained sediments and it became clear that
thesand had been deposited from a marine inundation into an area not nor-
mally subject to encroachment by the sea. The Cascadia subduction zone, where
thePacific tectonic plate is being subducted below the North American tectonic
plate occurs offshore. Such a geological structure is very likely to have produced
large earthquakes. Atwater concluded that the sand layer was deposited by a
tsunami generated by an earthquake with its epicentre in the Cascadia subduc-
tion zone. He was able to determine the age of the earthquake, and tsunami,
by analysing the chronology of the drowned trees through radiocarbon dating
and tree ring analysis. Interestingly, nearly ten years after Atwater published
his study, Satake et al .(1996)showed that the same tsunami probably travelled