Risk and Avoidance - Tsunami - page 199

conducive to tsunami, and no major wave propagated into
the Pacific Ocean.

America. Messages are disseminated by satellite, teletype,
e-mail, the internet, and phone to a number of crucial
people locally. Once a warning has been issued, over 90 tide
gauges are monitored to confirm the existence of a tsunami,
and its degree of severity. The Center also conducts com-
munity preparedness programs to educate the public on how
to avoid tsunami if they are caught in the middle of a violent
earthquake. Follow-up visits are made to the communities
that have experienced a false alarm. The purpose of these
visits is to explain why a warning was issued and to stress
the

10.3.3 Localized Tsunami Warning Systems

A tsunami originates in, or near, the area of the earthquake
that creates it. It propagates outwards in all directions at a
speed that depends upon ocean depth. In the deep ocean,
this speed may exceed 600 km s -1 . In these circumstances,
the need for rapid data handling and communication
becomes obvious if warnings are to be issued in sufficient
time for local evacuation. Because of the time spent in
collecting seismic and tidal data, the warnings issued by the
PTWC cannot protect areas against local tsunami in the first
hour after generation. For this purpose, regional warning
systems have been established. Local systems generally
have data from a number of seismic and tidal stations
telemetered to a central headquarters. Nearby earthquakes
have to be detected within 15 min or less, and a warning
issued soon afterwards to be of any benefit to the nearby
population. Because warnings are based solely upon a
seismic signature, false warnings are common. At present,
warning systems tend to err on the side of caution to the
detriment of human life.
One of the first local warning systems was establishment
of the West Coast/Alaska Tsunami Warning Center (WC/
ATWC) in Palmer, Alaska, in 1967 (Sokolowski 1999 ).
This filled a gap in the Pacific Tsunami Warning System
made apparent in Alaska by the Alaskan earthquake of
March 27, 1964. Here, tsunami were of three types: local-
ized, landslide-induced, and ocean-wide. Only the latter was
processed by the Pacific Tsunami Warning System. Not
only did warnings from Honolulu reach Alaska after the
arrival of all three types of tsunami, they also went through
a process that delayed dissemination to the public along the
west coast of the United States. In 1982, the Center's
mandate was extended to include the coasts of California,
Oregon, Washington, and British Columbia. In 1996, the
Center's responsibility was expanded to include all Pacific-
wide tsunamigenic sources that could affect these coasts.
The objectives of the West Coast and Alaska Tsunami
Warning Center are to provide immediate warning of
earthquakes in the region to government agencies, the
media, and the public; and to accelerate the broadcast of
warnings to the wider community along the west coasts of
Alaska, Canada, and the United States. Alarms are triggered
automatically by any sustained, large earthquake monitored
at eight seismometers positioned along the west coast of
North America and 23 short- and long-period seismometers
in Alaska. A warning can be issued automatically within
15 min of the event together with the estimated arrival time
of the tsunami at 24 sites along the west coast of North

continued need to respond to emergency tsunami

warnings.
Other tsunamigenic source areas in the Pacific Ocean
have developed localized warning systems. Separate
warning systems also exist for Hawaii, Russia, French
Polynesia, Japan, and Chile. The Russian warning system
was developed for the Kuril-Kamchatka region of north-
eastern Russia following the devastating Kamchatka Tsu-
nami of 1952 (International Oceanographic Commission
1999 ). This system operates from three centers at Petro-
pavlovsk-Kamchatskiy, Kurilskiye, and Sakhalinsk. It is
geared towards the rapid detection of the epicenter of
coastal tsunamigenic earthquakes because some tsunami
here take only 20-30 min to reach shore. In French Poly-
nesia, an automated system was developed in 1987, for both
near- and far-field tsunami, by the Polynesian Tsunami
Warning Center at Papeete, Tahiti (Okal et al. 1991 ; Rey-
mond et al. 1993 ). The system uses the automated algorithm
TREMORS (Tsunami Risk Evaluation through seismic
MOment in a Real time System) to analyze in real time
seismic data for any earthquake in the Pacific Ocean. Rather
than using P and S waves to calculate earthquake magni-
tude, TREMORS uses the magnitude of seismic waves
traveling through the mantle. This mantle magnitude, M m ,is
calculated from Rayleigh or Love waves having periods
between 30 and 300 s. These long wave periods are virtu-
ally independent of the focal geometry and depth of any
earthquake. Surprisingly good forecasts of tsunami wave
heights have been achieved for 17 tsunami that reached
Papeete between 1958 and 1986, including the Chilean
Tsunami of 1960. Because the TREMORS system is not
site-specific and the underlying equipment is inexpensive,
there is no reason why the system could not be installed in
any country bordering the Pacific Ocean.
In Japan, a number of systems are used for local tsunami
prediction (Shuto et al. 1991 ; Furumoto et al. 1999 ; Murata
et al. 2010 ). Tsunami warning began in Japan in 1941 under
the auspices of the Japan Meteorological Agency. Origi-
nally, coverage was only for the northeast Pacific Ocean
coast, but this was extended nationwide in 1952. The Japan
Meteorological Agency has a national office in Tokyo and
six regional observatories—at Sapporo, Sendai, Tokyo,
Osaka, Fukuoka, and Naha, with each responsible for local