You are on page 1of 11

2

The Nature of Earthquakes

2.1 Introduction
An earthquake is a spasm of ground shaking caused by a sudden release of energy
in the earths lithosphere (i.e. the crust plus part of the upper mantle). This energy
arises mainly from stresses built up during tectonic processes, which consist of interaction between the crust and the interior of the earth. In some parts of the world,
earthquakes are associated with volcanic activity. For example, in Guatemala such
earthquakes occur in swarms, with an average duration of three to four months, the
largest having a magnitude normally under 6.5. These events are of shallow focus
and cause considerable damage within a radius of about 30 km from the epicentre.
Human activity also sometimes modifies crustal stresses enough to trigger small or
even moderate earthquakes, such as the swarms of minor tremors resulting from
mining in the Midlands of England, or the sometimes larger events induced by the
impounding of large amounts of water behind dams, such as the earthquakes associated with the construction of the Koyna dam in central India in 1967 (Chopra and
Chakrabarti, 1973).
While the design provisions of this book apply to all earthquakes regardless of origin,
any discussion of earthquakes themselves is generally confined to events derived from
the main cause of seismicity, i.e. tectonic activity.
As most earthquakes arise from stress build-up due to deformation of the earths
crust, understanding of seismicity depends heavily on aspects of geology, which is the
science of the earths crust, and also calls upon knowledge of the physics of the earth
as a whole, i.e. geophysics. The particular aspect of geology which sheds most light on
the source of earthquakes is tectonics, which concerns the structure and deformations
of the crust and the processes which accompany it; the relevant aspect of tectonics is
now often referred to as seismotectonics.
Geology tells us the overall underlying level of seismic hazard which may differ
from the available evidence of historical seismicity, notably in areas experiencing
present day quiescent periods.

Earthquake Risk Reduction D.J. Dowrick


2003 John Wiley & Sons, Ltd ISBN: 0-471-49688-X (HB)

16

The nature of earthquakes

2.2 Global Seismotectonics


On a global scale, the present-day seismicity pattern of the world is illustrated in
general terms by the seismic events plotted in Figure 2.1. Most of these events can be
seen to follow clearly defined belts which form a map of the boundaries of segments of
the earths crust known as tectonic plates. This may be seen by comparing Figure 2.1
with Figure 2.2, which is a world map of the main tectonic plates taken from the highly
understandable book on the theory of continental drift by Stevens (1980). According
to the latter, the earths crust is composed of at least 15 virtually undistorted plates of
lithosphere. The lithosphere moves differentially on the weaker asthenosphere which
starts at the Low-Velocity Layer in the Upper Mantle at a depth of about 50 km.
Boundaries of plates are of four principal types;
(1) Divergent zones, where new plate material is added from the interior of the earth.
(2) Subduction zones, where plates converge and the under-thrusting one is consumed.
(3) Collision zones, former subduction zones where continents riding on plates are
colliding.
(4) Transform faults, where two plates are simply gliding past one another, with no
addition or destruction of plate material.
Almost all the earthquake, volcanic and mountain-building activity which marks the
active zones of the earths crust closely follows the plate boundaries, and is related to
movements between them.
Divergent boundaries are found at the oceanic sea-floor ridges, affecting scattered
islands of volcanic origin, such as Iceland and Tristan da Cunha, which are located
on these ridges. As these zones involve lower stress levels, they generate somewhat
smaller earthquakes than the other types of plate boundary.
As can be seen in Figure 2.2, subduction zones occur in various highly populated
regions, notably Japan and the western side of Central and South America. Figure 2.3
shows the cross-section of the likely structure of the subduction zone formed by the
Pacific plate thrusting under the Indian-Australian plate beneath the North Island of
New Zealand. The seismic cross-section corresponding to Figure 2.3 is in Figure 2.4,
and gives earthquakes located under the shaded region of the key map (during a period
of time when no events shallower than c. 40 km occurred). The zone of diffuse seismic
activity which exists down to a depth of over 300 km is believed to be related both
to volcanic activity (movement of magma in the crust and upper mantle and related
expansion and contraction), and to faulting within the volcanic belt and the New
Zealand Shear Belt. The latter is a continuation of the major Alpine Fault of the
Southern Alps (Figure 2.5). Below 100 km and down to about 250 km, the pattern of
earthquakes tends to lie on a well-defined plane known as a Benioff Zone, dipping
50 degrees to the north-west. This is the contact plane between the Indian-Australian
plate and the Pacific plate. The isolated group of earthquakes about 600 km deep in
Figure 2.4 have been conjectured to be caused by a piece of lithosphere that has become
detached and has moved deeper into the mantle, as illustrated in Figure 2.3.
The progressive movement of the Pacific plate, subducting under the overlying
plate, caused shear stresses to develop, as illustrated by Walcotts (1981) geophysical

60

60

60

80

180

Figure 2.1

120

Seismicity map of the world. The dots indicate the distribution of seismic events in the mid-twentieth century (after Barazangi and
Dorman (1969))

Antarctic circle

60

60

20

40

Tropic of capricorn

Equator

20

40

20

20

120

120

40
Tropic of cancer

180

40

Arctic circle

120

60

60

60

80

Global seismotectonics
17

18

The nature of earthquakes

EURASIAN
PLATE

EURASIAN PLATE
AMERICAN
PLATE

PACIFIC PLATE

INDIAN
AUSTRALIAN
PLATE

Nazca
plate

AFRICAN
PLATE

ANTARTIC PLATE
Subduction zone

Figure 2.2

Direction of plate motion

Tectonic plate map of the world, showing names of the seven largest plates and
indicating subduction zones and the directions of plate movement (reproduced with
permission from G.R. Stevens, 1980)
Tongariro
Ngauruhoe
Egmont Ruapehu

Indian- A
Australion
plate
0

Hikurangi trench

B
Pacific
plate

100

200
A
300

400

500

200 km

600
km

Figure 2.3

The likely structure of the subduction zone beneath the North Island of
New Zealand inferred from figure 2.4 (reproduced with permission from
G.R. Stevens, 1980)

Global seismotectonics

Egmont

19

Tongariro
Ngauruhoe
Ruapehu
Hikurangi trench

0
300

200

100

100

200

300
Edge of
continental
slope

400

500

200 km

600
km

Figure 2.4

Seismic cross-section through the North Island of New Zealand, showing locations
of earthquake foci (reproduced with permission from G.R. Stevens, 1980)

study (Figure 2.5), which relates the shear strains of the shear belt referred to above
to large historical earthquakes. It is believed that the fault forming the plate boundary
periodically locks together, and this leads to an accumulation of shear and compressional strain until it is in part relieved by a large thrust type of earthquake. The sudden
release of strain (when the shear resistance is overcome) signals the recommencement
of movement of the subducting plate in a further cycle of aseismic slip, then another
locking of the fault leads to the next plate interface earthquake.
As well as the 15 or so main plates shown in Figure 2.2, studies of seismic activity
need to consider the smaller buffer plates or sub-plates which in certain areas tend
to ease the relative movements of the worlds giant plates. Buffer plates have been
recognized in Tibet and China, the western USA, and at the complex junction of the
African, Arabian, Iranian, and Eurasian plates, where eight Mediterranean buffer plates
have been identified (Stevens, 1980).
In the foregoing discussion, tectonic plates have been described as rigid, virtually
undistorted plates and the worlds principal zones of seismicity have been shown to
be associated with the interaction between the plates. However, occasional damaging
intra-plate earthquakes also occur, well within the interior of the plates that clearly
are not associated with plate boundary conditions, and so far their origins are illunderstood. The uncertainties associated with intra-plate seismicity are much greater
than is the case for interplate regions of high seismicity.

20

The nature of earthquakes

200

0m

rd
Lo

1995

H
e
ow
R
e
is

1932

>3 x 107/yr

1931 1931 h
c
1843
1863
en
1897
r
T
1904 i
1868
g
1893 1934
n
a
1942
ur
k
1855
i
H
1848
1968
1929

>5 x 107/yr

1994

Strain Rate

40 S

k
oo it
C tra
S

1888
1929

1901

PS

RN

Chatham Rise

AL

TH

U
SO

1938

00

45

20

1993

200 km

1979
1945
170E

Figure 2.5

Campbell

Plateau

180

Generalized shear strain rates and large (MW 6.8) shallow (hC < 50 km) New
Zealand earthquakes from 18402000 (adapted from Walcott, 1981)

2.3 The Strength of EarthquakesMagnitude and Intensity


During earthquakes the release of crustal stresses is believed generally to involve
the fracturing of the rock along a plane which passes through the point of origin
(the hypocentre or focus) of the event (Figure 4.25). Sometimes, especially in larger
shallower earthquakes, this rupture plane, called a fault, breaks through to the ground
surface, where it is known as a fault trace (Figure 4.41).

The strength of earthquakesmagnitude and intensity

21

The cause and nature of earthquakes is the subject of study of the science of seismology, and further background may be obtained from the books by Richter (1958),
Bolt (1999) and Lay and Wallace (1995).
Unfortunately, for non-seismologists at least, understanding the general literature
related to earthquakes is impeded by the difficulty of finding precise definitions of
fundamental seismological terms. For assistance in the use of this book, definitions of
some basic terms are set out below. Further definitions may be found elsewhere in this
book or in the references given above.
The strength of an earthquake is not an official technical term, but is used in the
normal language sense of How strong was that earthquake? Earthquake strength
is defined in two ways: first the strength of shaking at any given place (called the
intensity) and second, the total strength (or size) of the event itself (called magnitude,
seismic moment, or moment magnitude). These entities are described below.
Intensity is a qualitative or quantitative measure of the severity of seismic ground
motion at a specific site. Over the years, various subjective scales of what is often called
felt intensity have been devised, notably the European Macroseismic and the Mercalli
scale, which are very similar. The most widely used in the English speaking world
is the Modified Mercalli scale (commonly denoted MM), which has twelve grades
denoted by Roman numerals IXII. A detailed description of this intensity scale is
given in Appendix A, taken from Dowrick (1996).
Quantitative instrumental measures of intensity include engineering parameters such
as peak ground acceleration, peak ground velocity, the Housner spectral intensity,
and response spectra in general. Because of the high variability of both subjective
and instrumental scales, the correlation between these two approaches to describing
intensity is inherently weak (Figure 4.23).
Magnitude is a quantitative measure of the size of an earthquake, related indirectly
to the energy released, which is independent of the place of observation. It is calculated
from amplitude measurements on seismograms, and is on a logarithmic scale expressed
in ordinary numbers and decimals. Unfortunately several magnitude scales exist, of
which the four most common ones are described here (ML , MS , Mb and MW ).
The most commonly used magnitude scale is that devised by and named after
Richter, and is denoted M or ML . It is defined as
ML = log A log A0

(2.1)

where A is the maximum recorded trace amplitude for a given earthquake at a given
distance as written by a WoodAnderson instrument, and A0 is that for a particular
earthquake selected as standard.
The WoodAnderson seismograph ceases to be useful for shocks at distances beyond
about 1000 km, and hence Richter magnitude is now more precisely called local magnitude (ML ) to distinguish it from magnitude measured in the same way but from
recordings on long-period instruments, which are suitable for more distant events. When
these latter magnitudes are measured from surface wave impulses they are denoted by
MS . Gutenburg proposed what he called unified magnitude, denoted m or mb , which
is dependent on body waves, and is now generally named body wave magnitude (mb ).
This magnitude scale is particularly appropriate for events with a focal depth greater
than c. 45 km. All three scales ML , mb and MS suffer from saturation at higher values.

22

The nature of earthquakes

The most reliable and generally preferred magnitude scale is moment magnitude,
MW . This is derived from seismic moment, M0 , which measures the size of an
earthquake directly from the energy released, Wyss and Brune (1968), through the
expression
(2.2)
M0 = AD
where is the shear modulus of the medium (and is usually taken as 3 1010 Nm), A
is the area of the dislocation or fault surface, and D is the average displacement or slip
on that surface. Seismic moment is a modern alternative to magnitude, which avoids
the shortcomings of the latter but is not so readily determined. Up to 1985, seismic
moment had generally only been used by seismologists.
Moment magnitude is a relatively recent magnitude scale from Kanamori (1977)
and Hanks and Kanamori (1979), which overcomes the above-mentioned saturation
problem of other magnitude scales by incorporating seismic moment into its definition,
such that moment magnitude
MW =

2
log M0 6.03 (M0 in Nm)
3

(2.3)

Local magnitude ML is inherently a poor magnitude scale, as shown by the plot


in Figure 2.6 of against of New Zealand data from Dowrick and Rhoades (1998),
who found that the best fit relationship for estimating MW from ML and depth,
hc , was
MW = 0.96[0.49] + 0.84(0.08]ML 0.0055[0.0015](hc 25)

(2.4)

The regression explains only 59% of the variance and has a residual standard error of
0.31. The ML scale as estimated in other parts of the world, as well as New Zealand,
is similarly unreliable.
The relation between moment magnitude MW , surface-wave magnitude MS and
centroid depth hc , using data restricted to modern MW determinations (i.e. from 1964
March 8 onwards), is shown in Figure 2.7. For earthquakes of hc 30 km, MS and MW
are close to being equal above magnitude 6.5. At lower magnitudes MS is consistently
smaller than MW , and is as much as a quarter-unit smaller between magnitude 5.0
and 5.5. Depth also influences the discrepancy between MS and MW ; for deep New
Zealand earthquakes (hc > 50 km) MS is about a half-unit smaller than MW between
magnitude 5.0 and 5.5. This results from the tendency for MS to decrease with depth
for earthquakes of a given seismic moment. Karnik (1969) first dealt with this effect
by proposing a focal depth correction term for MS in relation to mb for various parts
of Europe, while Ambraseys and Free (1997) more recently estimated a focal depth
correction term for in relation to log M0 for European earthquakes. Considering the
New Zealand data (Figure 2.7) Dowrick and Rhoades (1998) found the best fit for
finding MW in terms of MS and hc was the quadratic expression:
MW = 1.27[0.16] + 0.80(0.03]MS + 0.087[0.031](MS 6)2 + 0.0031
[0.0006](hc 25)

(2.5)

The strength of earthquakesmagnitude and intensity

Figure 2.6

23

Scatter plot of local magnitude ML against moment magnitude MW for New


Zealand earthquakes distinguishing events in different classes of centroid depth
hC . Also shown are the linear and quadratic regression fits for ML evaluated at
hC = 25 km and a local regression trend curve of ML on MS for events with
hC 50 km (from Dowrick and Rhoades, 1998)

The above expression explains 93% of the variance. In Eq. (2.5) it can be seen that the
quadratic term contributes significantly to the regression because the coefficient of this
term is more than twice its standard error. It is of interest to note that, although their
expression is different from ours, Ambraseys and Free obtained a coefficient for their
depth term of 0.0036, which is very similar to the coefficient of 0.003 in Eq. (2.5).
Also shown in Figure 2.7 is the relation of Ekstrom and Dziewonski (1988), derived
from global data, between log MO and MS for events with h < 50 km. In terms of MW ,
this relation is

2
MS < 5.3

2.13 + 3 MS

(2.6)
MW = 9.40 41.09 5.07MS
5.3 MS 6.8

0.03 + MS
MS > 6.8
As seen in Figure 2.7, there is no great difference between this relation and the linear
and quadratic fits for shallow New Zealand events over the magnitude range of the
data, but the latter also describe the effect of depth.

24

The nature of earthquakes

Figure 2.7

Scatter plot of moment magnitude MW against surface wave magnitude MS for


earthquakes distinguishing events in different classes of centroid depth hC . Also
shown are the linear and quadratic (Eq. (2.5)) regression fits for MW evaluated at
hC = 25 km and a local regression trend curve of MW on MS for events with hC
50 km, and the relation of Eckstrom and Dziewonski (Eq. (2.4) (from Dowrick and
Rhoades, 1998)

References
Ambraseys NN and Free MW (1997) Surface-wave magnitude calibration for European region
earthquakes. J Earthq Eng 1(1): 122.
Barazangi M and Dorman J (1969) World seismicity map of ESSA coast and geodetic survey
epicentre data for 196167. Bull Seism Soc Amer 59: 36980.
Bolt B (1999) Earthquakes. WH Freeman & Co, New York, 4th Edn.
Chopra AK and Chakrabarti P (1973) The Koyna earthquake and damage to Koyna dam. Bull
Seism Soc Amer 63(2): 38197.
Dowrick DJ (1996) The Modified Mercalli earthquake intensity scaleRevisions arising from
recent studies of New Zealand earthquakes. Bull NZ Nat Soc Earthq Eng 29(2): 92106.
Dowrick DJ and Rhoades DA (1998) Magnitudes of New Zealand earthquakes, 19011993.
Bull NZ Soc Earthq Eng 31(4): 26080.
Ekstrom G and Dziewonski AM (1988) Evidence of bias in estimations of earthquake size.
Nature 332: 319323.
Hanks TC and Kanamori H (1979) A moment magnitude scale. J Geophys Res B 84: 234850.
Kanamori H (1977) The energy release in great earthquakes. J Geophys Res 82: 298187.

References

25

Karnik V (1969) Seismicity of the European Area, Part 1. Reidel, Dordrecht, Holland.
Lay T and Wallace TC (1995) Modern Global Seismology. Academic Press, San Diego.
Richter CF (1958) Elementary Seismology. Freeman, San Francisco.
Stevens GR (1980) New Zealand Adrift. AH & AW Reed, Wellington.
Walcott RI (1981) The gates of stress and strain. In: Large Earthquakes in New Zealand. The
Royal Society of New Zealand, Miscellaneous Series No 5.
Wyss M and Brune J (1968) Seismic moment, stress and source dimensions. J Geophys Res 73:
468194.

You might also like