Inverting tsunami deposits to their source via tsunami modeling: Examples from
the Kuril-Kamchatka region.

Joanne Bourgeois, [with Tatiana Pinegina, Robert Weiss, Breanyn MacInnes, M.
Elizabeth Martin, Vasily Titov, Heidi Houston and others]

Tsunami deposits are a resource for determining earthquake rupture
characteristics. The most basic information derived from a tsunami deposit is
1) a tsunami must have reached the location of the deposit, and 2) the
deposit’s elevation is a minimum estimate of peak tsunami amplitude at that
location. Via tsunami modeling, the size of the tsunami can then be inverted to
initial sea-floor disturbance and rupture characteristics of the earthquake.
Deposits and other physical evidence in the tsunami nearfield also provide
important information for modelers because whereas tsunami models can routinely
and reliably predict amplitudes of farfield tsunamis, nearfield tsunami
amplitudes are strongly affected by earthquake characteristics such as slip
distribution along the rupture length. Farfield modeling can simply use
location and seismic moment of an earthquake because with increasing distance
from the source, perturbations in the waveform caused by the rupture are erased
by bathymetry. However, in the nearfield, as every earthquake is different, so
also will every tsunami be different. 

Starting with tsunami deposits, we are using the tsunami model MOST [Method of
Splitting Tsunamis] to elucidate several earthquake sources in the Russian Far
East, including 1952 Kamchatka-Kurils, 1969 Ozernoi, 1971 Kamchatskiy, 1997
Kronotsky and 2006 middle Kurils. For example, the high elevation of certain
tsunami deposits from the Mw 9.0 1952 Kamchatka earthquake lead us to conclude
via modeling that there was a region of high slip off south Kamchatka. Modeling
of 1969 and 1971 northern Kamchatka tsunamis permits us to distinguish deposits
from each in the field, and to elucidate both earthquakes and tsunamis.
Deposits from the 1997 Kronotsky tsunami point to a larger local tsunami than
catalogued and require a region of high slip at the northern end of the rupture
zone. 

Deposits from and modeling of the 2006 mid-Kurils tsunami will provide a
further test of our approach. For this event we have immediate “before and
after” observations of the tsunami’s effects, GPS measurements from the
middle Kurils, extensive far-field observations, and several seismologic
models. Moreover, even before the great subduction-zone earthquake of 15 Nov
2006, we had field evidence of large prehistoric tsunamis in the middle Kuril
Islands [a historical seismic gap till 2006], indicating that this segment of
the subduction zone has been seismically very active, comparable to other parts
of the Kuril-Kamchatka system. Our next step is to quantify the 2006 indicators
so that we can estimate the sizes and sources of the prehistoric events.