Geography Reference
In-Depth Information
prevailing volcanic processes that constitute the
hazard (Martinelli 1991). Thus hazard maps have
been made of ash falls on Etna and Vesuvius in
Italy (Gasparini 1993), and of lahars, pyroclastic
blasts and ash falls on Nevado del Ruiz, Colombia
(Parra and Cepeda 1990). General hazards were
mapped at the Soufriere Hills volcano on
Montserrat in the Caribbean almost a decade
before the repeated eruptions of 1997, which were
dominated by pyroclastic flows, vertical blasts and
ash falls (Wadge and Isaacs 1988). However, the
mere presence of a hazard map does not
necessarily mean that there will be an adequate
response in terms of land use and civil protection,
as the awful case of the November 1985 eruption
of Nevado del Ruíz demonstrated so graphically:
23,000 people were killed by lahars, and yet an
accurate and comprehensive hazard map had
existed for some months previously (Voight 1990).
Remote sensing has proved to be invaluable as
a means of generating a mappable overview of
seismic and volcanic hazards (Murphy 1994).
Further advances have been made using
geographic information systems (GIS). For
example, Emmi and Horton (1993) used a GIS of
Salt Lake County, Utah, to assess the spatial
variation of earthquake risk in terms of exposure
period, intensity of ground shaking and
probability of earthquake occurrence. They related
these factors to the vulnerability of the built
environment (by constructing an inventory of
buildings and a series of earthquake engineering
damage functions), the pattern of building
occupation and the expected nature of casualties.
Simulation of spatial patterns enabled a sensitivity
analysis to be conducted in order to assess the
model's limitations (Emmi and Horton 1995).
Mediterranean (Soloviev 1990). These enable
hazard assessments to be refined as new data are
added, especially with respect to the recurrence
intervals of tsunamis of given magnitudes. For
instance, in eastern Honshu (Japan), run-up studies
have been carried out since the 1930s, and it is
known that 10m high tsunamis have a return
period of only ten years. Even for areas with
limited and infrequent tsunamis, such as the Italian
peninsula, it has been possible to map the hazard
quite successfully (Tinti 1991).
The other pertinent aspect of tsunami research
is warning, which is well developed only in the
Pacific basin, through a regional mechanism, the
Pacific Tsunami Warning System (Pararas-
Carayannis 1986), and a rapid reaction local
system, Project THRUST (Bernard 1991). Several
geographical problems have been experienced
with the PTWS. It must cover enormous areas and
be able to monitor the progress of waves that travel
through open ocean water at the speed of a
cruising jet liner but which are not detectable to
the naked eye until they make landfall. For
adequate monitoring, instruments must be
deployed in remote and widely scattered locations,
and the network must be dense enough to detect
tsunamigenic events rapidly and efficiently in the
widest possible variety of seismic, volcanic and
tectonic settings. Perhaps it was the apparent
remoteness of the problem that led Canada to
withdraw briefly from among the twenty-three
nations that participate in the PTWS. This proved
to be a false economy, and it later rejoined.
However, Pacific-wide warning is not particularly
effective in the case of near-field tsunamis (i.e.
when the interval between tsunamigenesis and the
arrival of destructive waves is less than, say, twenty
minutes), and for these, Project THRUST was
developed on the basis of satellite and
microcomputer technology. This is a particularly
important development: worldwide, of the 53,000
coastal residents who died in ninety-four tsunamis
that occurred over the period 1900-94, 99 per
cent were located within 400 km of the point of
tsunamigenesis, which was usually an earthquake
epicentre. The PTWS gives a minimum alert time
of one hour throughout the Pacific basin and ten
TSUNAMIS
Tsunamis are included here because they are
mainly caused by earthquakes and volcanic
eruptions. One aspect of tsunami research has been
the compilation of catalogues that list known
events in the world's marine basins, for example
the Pacific (Lockridge 1988) and the
