Environmental Engineering Reference
In-Depth Information
world's second strongest recorded earthquake had its
epicenter just west of Aceh in northern Sumatra, where
the Indian Ocean plate subducts the Asian continental
plate (Lay et al. 2005).
Stick-slip frictional instability, long recognized as
the prime mover of earthquakes, explains not only seis-
mogenesis but also pre- and post-seismic phenomena
(Scholz 1998). During an earthquake a rupture usually
propagates along the fault plane in just a few tens of sec-
onds. During the Sumatra-Andaman earthquake the rup-
ture extended (at@2.8 km/s) more than 1300 km along
the Andaman trough for about 8 min, the longest on
record (Ammon et al. 2005; Ishii et al. 2005). But there
are also very slow-moving ruptures, which cause silent
earthquakes without any sudden, strong tremors or
waves on seismometers (Hirn and Laigle 2004). As for
the temporal variation, the twentieth-century record
shows a clear cluster of events during the 1950s and
1960s, but a detailed analysis of wait times for all of the
40 great interplate events of the twentieth century shows
them to conform to a model of random occurrence, with
average wait time of 2.3 years between the successive
events.
Energies dissipated during an earthquake include heat
produced by friction and kinetic energy of new cracks in
the crust and seismic waves radiated through the Earth.
Only the waves, felt as the earthquake tremors and
recorded by seismographs, can be measured. The earliest
earthquake intensity scales (De Rossi in 1874, Sekiya in
1885, Murashi in 1887) were highly subjective. The first
fundamental approach to strength-based classification of
tremors was taken by Richter's (1935) magnitude scale
defined by decadic logarithms of the largest trace ampli-
tude (in mm) recorded with a standard Wood-Anderson
torsion seismograph at a distance of 100 km from the
epicenter. This measure eventually became known as
local magnitude (M L ). Other measures, including surface
wave magnitude (M S ), followed during the next 50 years.
Earthquake magnitude was first correlated with the re-
lease of seismic energy (E S ) by Gutenberg and Richter
(1942). The conversion (E S originally in ergs) had the
form
log 10 E S ¼ A þ BM L :
A and B, two empirical coefficients, were initially 11.3
and 1.8, and later various authors used values between
5.8 and 14.2 for A and between 1.21 and 4 for B
(Howell 1990). This conversion is a gross oversimplifica-
tion that yields only very approximate equivalents. The
same is true for the conversion of M S into seismic energy
because M S is calculated from a bandwidth that is too
narrow to capture all the radiated frequencies. Inherent
inaccuracies aside, relative differences remain unchanged:
a magnitude increase of 0.2 doubles the seismic energy
release, and a unit increase multiplies it 32 times.
The M L and M S scales were eventually replaced by
a more consistent scale, not based on an instrumental
recording but on the measurement of the ruptured area
and of the average slip across the affected fault:
M 0 ¼ mAD ;
where M 0 is moment, m rigidity of the material around
the rupture zone, A the area displaced, and D the aver-
age displacement across the fault. Moment magnitude
(M W ), introduced by Hanks and Kanamori (1979), is
defined (in SI units) as
M W ¼ 3 ð log 10 M 0 9 : 1 Þ ;
Search WWH ::




Custom Search