Earthquake Prediction, Seismic Hazard And Vulnerability

SeiyaUyeda

Earthquake Prediction Research Center, Tokai University, Shizuoka, Japan

Kimiro Meguro

International Center for Urban Safety Engineering, Institute of Industrial Science,

The University of Tokyo, Tokyo, Japan

Mitigation o f seismic hazards requires integration of science and human action,

namely the science o f earthquakes, anti-seismic engineering and socio-political
measures. The public, media, policy makers and funding agencies must be con­
stantly reminded that seismic disasters rapidly escalate with civilization's growth and
that disasters c o m e when the last tragedy has been "forgotten". Loss of human life
is caused overwhelmingly by the collapse o f houses and other buildings within less
than a few minutes of the main shocks. The most urgent countermeasure is the rein­
forcement o f weak structures. When structural damage is reduced, most other seis­
mic hazards will correspondingly be greatly reduced. If short-term prediction is
made, casualties will be further reduced dramatically. Despite general pessimism,
short-term prediction research needs to be enhanced because recent research shows
real promise. Thus, the reinforcement o f existing structures and enhancement of
short-term prediction research are the two keys for seismic hazard mitigation.

INTRODUCTION
Earthquakes are caused by sudden fault motion. It has long
been known that the global distribution of earthquakes is far
from uniform, as shown in Fig. 1. While mid-oceanic ridges
are characterized by linear distribution of relatively small
earthquakes, large earthquakes occur mainly in the circum-
Pacific belt and in the wide zone between Eurasia and the
southern continents. The reason why they are distributed in this
manner is explained by plate tectonics [e. g., Stein andKlosko,
2002]; large earthquakes occur due to plate interactions at
convergent and transform plate boundaries. Convergent plate
boundaries consist of subduction and collision zones. Statis­
tical properties of earthquake occurrences in these seismic
zones are well-known thanks to many decades of seismic
The State of the Planet: Frontiers and Challenges in Geophysics
Geophysical Monograph 150, IUGG Volume 19
Copyright 2004 by the International Union of Geodesy and Geophysics
and the American Geophysical Union.
10.1029/150GM27
observations around the globe, and many seismic hazard maps
have been compiled both globally and regionally [e. g., GSHAP
program, Giardini, 1999].
Plate tectonics has shown with reasonable certainty that
plate motions have been essentially steady for the time-scales
of a few millions of years, although they have changed over
much longer time scales during the earth's history [e.g.,
Kumazawa and Maruyama, 1994; Uyeda, 2002]. However,
the seismicity of the globe also displays variation on finer
time-scales of tens of years, as shown in the upper panel of Fig.
2. The detailed mechanisms for these short-term secular
changes are not clear, but it seems plausible that there can be
fluctuations even under steady global plate motions. It seems
significant that the variations in the loss of life shown in the
lower panel of Fig. 2 are very different from variations of
seismicity. The reason for this is quite clear. Earthquakes in
densely populated regions cause greater disasters [Utsu, 2002].
For instance, the two largest giant earthquakes, the 1960 M9.5
Chilean and 1964 M9.2 Alaskan earthquakes, caused much less

349
350 EARTHQUAKE PREDICTION, SEISMIC HAZARD AND VULNERABILITY
Figure 1. Epicenter distribution of the world for magnitude greater than 4.0 in 1980-2000, after USGS PDE.
loss of life compared to the much smaller 1923 M7.9 Kanto
and 1976 M7.8 Tangshan earthquakes. (Hereafter, M stands
for the magnitude of an earthquake).
Here we review briefly the state of the art on seismic haz­
ard and present personal views on the priority items related to
earthquake hazard mitigation. Specific matters taken up in
the text are focused on Japan. However, it is anticipated that
what we present will be generally applicable to all earthquake-
prone regions of the world.
As well as loss of life, earthquakes cause disasters of all
kinds. In terms of monetary loss, this can amount to a sub­
stantial fraction of the Gross Domestic Product (GDP) of a
nation. For the 1995 Kobe earthquake, monetary loss was
estimated at 10 billion US dollars and for the Tonankai-Nankai
event, which is a major anticipated future earthquake, it may
amount to at least 80 billion US dollars, which would be more
than 15% of the GDP according to the Central Disaster Man­
agement Council of Japan. Since buildings in Japan are now
more seismic resistant than in the past, loss of life in large
cities may be less than those in the past. However, huge dam­
age of complicated life-lines and infra-structures can still be
expected in modern mega-cities when they are hit by major
quakes. Further many mega-cities in the developing world do
not have good adherence to building codes and so there can
be huge loss of life as illustrated by the 2004 Bam earthquake
in Iran. Seismic risk rapidly escalates with population growth
and the disasters come when the last experience is "forgotten".
FIRST PRIORITY IS THE REINFORCEMENT
OF OUR HOUSES
Many problems arise at and after a disastrous earthquake
[Meguro and Takahashi, 2001], including:
1) Loss of life at the time of the main shock,
2) Further victims due to fires,
3) Psychological instability of people in the affected areas,
4) Disruption of the community,
5) Building refugee camps for large number of displaced peo­
ple,
6) Demolition of damaged structures and related environ­
mental effects,
7) Economic, business and societal disruption.
At the time of the main shock, rapid delivery of information
on ground motions and damages to the authorities and public
are critically important for rescue activity. Thanks to major
advances in information technology, there is an emerging new
category of hazard mitigation; real-time seismology that has
been developed in several areas, including Japan, Mexico,
California and Taiwan [e. g., Nakamura, 1988; EpinosaAranda
et al, 1995; Kanamori et al, 1997; Shin et al., 2000]. The
Seismic Alert System in Mexico made use of the gap of the
arrival times of the P- and S-waves. The authorities were able
to issue timely alarms so that the subways in the Mexico City
were stopped 50 seconds before the arrival of the destructive
S-waves from the 1995 M7.3 Guerrero earthquake located at
the Pacific coast. However, many of the problems listed above
will last for a long time after the event. To cope with the evo­
lution of disaster situations in hours, days, weeks and even
months, innovative engineering strategies will be needed. A
new kind of strategy can be called the real-time earthquake
engineering. After the main shock, continuing data collection
of structure damage distribution, rescue activities, lifeline
interruptions, debris removal, refugee camps, and power
demand fluctuations and other parameters will be critical
because of the utmost importance for disaster relief agencies
and local and national governments to allocate resources in the
 UYEDA AND MEGURO 351

dyne-cm
J
10 n 1 r~ ~! ! i 1 s 1 r 1 1— Mw
erg S ; Sumbawa A : Aleutian I : India Kr: Kurile Me: Mexico
0
1026
Co: Colombia A k : Alaska J : Japan M : Mongolia
9.5
10* P: Peru C: Chile K: Kamchatka
c %
9,0
10"
2
10
8.5
10*
10 8.0
1810 1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
Year

1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
Year
Figure 2. Secular variations of seismic energy release (upper panel) and human loss (lower panel). Ordinate of the upper
panel represents the size of earthquakes in seismic moment M0, seismic energy E and the moment magnitude Mw. Ordi­
nate of the lower panel shows the number of victims. In both panels, the vertical bars are for individual event and the
solid curve shows the annual average (untagged 5-year running average) (after H. Kanamori, private communication).

most prompt and efficient manner [Noda and Meguro, 1995; that, almost without exception, the majority of victims are
Hada and Meguro, 2002]. killed by the collapse of buildings right at the moment of the
We now focus items 1) and 2) in the above list, namely on main shock (Plate 1, a & b). An example is shown in Fig. 3 for
how to save human life. Experiences all over the world show the 1995 M7.3 Hyogoken Nanbu Earthquake, which took over
5,500 human lives. (This earthquake is commonly called the
Kobe earthquake after the name of the city, where the damage
was most intense.) Almost 90 % of the victims were killed in
No record
Hospital their own houses. Moreover, medical examinations indicate that
1.0 % 3.8% 92 % of casualties were killed within less than 15 minutes
Other places after the quake before any organized rescue operation had
started [Nishimura et. al., 1997a, b].
8.5%
Many victims of earthquakes can be killed by fires.
Although fires can be caused by shaking, most of big fires
start from collapsed houses. Fires that started in non-col­
lapsed houses were commonly extinguished by the resi­
dents, while those that started in collapsed houses were not.
Residents under debris could not fight the fires and the first
Own houses priority of outside rescuers was to get them out from the
fallen houses. Furthermore, roads were often blocked to
86.6 % fire engines by collapsed houses. In Kobe, the Fire Depart­
ment had a capacity to handle only three to four fires at a
time. When many tens of fires started simultaneously, they
were completely overwhelmed. To prevent fires from devel­
Figure 3. Places where people died (in Kobe City), (after Nishimura oping to uncontrollable size, the best way is to prevent
et al., 1997b) houses from collapsing.
3 52 EARTHQUAKE PREDICTION, SEISMIC HAZARD AND VULNERABILITY
Plate 1. Damage to low earthquake resistant structures is major cause of casualties and many problems generated after
the earthquake in the world (photo by K. Meguro) (a) Damage to old timber houses due to the Kobe earthquake (M7.3),
1995 in Japan (b)Damage to masonry houses due to the Qayen earthquake (M7.1), 1997 in Northeast Iran.
If the government of Japan were to fund 10 billion US dol­
lars annually, all the necessary reinforcement of houses in
Japan would be completed in ten years [Ohtake, 2003]. In
order to implement the house reinforcement program effec­
tively, the following system, called the Retrofitting Promo­
tion System, has been proposed [Meguro and Takahashi, 2001;
Yoshimura and Meguro, 2003]. The main concept of this sys­
tem is that the government guarantees to pay a portion of the
repair and reconstruction expenses of damaged houses pro­
vided that proper retrofitting had been implemented by the
owners before the earthquake. With such a system, the over­
all financial burden of both government and residents can be
greatly reduced. The feasibility of this system has been veri­
fied for cases in which the retrofitting cost is 10 to 15% of the
new construction cost, as is the case in Japan. However, for
countries like Turkey, where the retrofitting cost amounts to
approximately 75% of the new construction cost, the system
is not viable. In such cases, efficient and economic techniques
for retrofitting are needed, especially for masonry structures,
using locally available and inexpensive materials [Mayorca and
Meguro, 2003].
If the structural damage is reduced, loss of human life is
reduced. Thus, the most effective and highest priority coun-
termeasure against the loss of human life, and all kinds of
related seismic hazards and consequences, is the structural
issue. In the recent December 26, 2003, M6.7 Bam Earth­
quake in Iran, there were more than 30,000 casualties due to
the collapse of adobe and brick structures which lacked rein­
forcement. After the disaster, international and local agen­
cies rushed to the affected sites trying to rescue survivors.
Although these were valuable efforts, only a few people sur­
vived and were rescued, again highlighting the quality of
buildings as a key issue.
FIRST PRIORITY IN SCIENCE IS SHORT-TERM
PREDICTION
Engineering seismology and earthquake prediction consti­
tute the two key measures that need implementation for major
seismic hazard mitigation. It is customary to classify earth­
quake prediction into three categories: long-term, intermedi­
ate-term and short-term predictions. They are different in
methodology, accuracy and purposes. For the purpose of sav­
ing human life, the short-term (less than a day to months)
prediction is most effective. If people had been warned in
advance and had escaped adobe buildings in time, casualties
in the recent Bam earthquake could have been many times
smaller. The impact of short-term prediction on casualties
would also be dramatic for cases of large earthquakes that
generate tsunamis.
Long-term prediction deals with the probability of earth­
quake occurrence on time-scales of 10 to 100 years, based
mainly on geologic studies of faults and historic records of
seismicity, while intermediate-term (1 to 10 years) prediction
uses more recent data including seismological observations.
One of the most advanced of such efforts is the M8 algorithm
which is based on the non-linear dynamics (Keilis-Borok and
Kossobokov, 1990). It is a pattern recognition approach based
on monitoring seismicity and its fluctuations. Long and inter­
mediate-term predictions are typically statistical likelihood
estimates, while short-term prediction is based on some def­
inite precursors.
Systematic short-term earthquake prediction research started
in the 1960's in several countries including Japan, USA, Soviet
Union, and China [Rikitake, 1976]. In the 1970's, optimism pre­
vailed due to the encouraging developments, such as dila-
tancy models [e. g., Scholz et al, 1973] and the successful
 UYEDA AND MEGURO 353

prediction of the 1975 M7.2 Haisheng earthquake in China
[Chen et al., 1990; Li Hui 1996]. However, no further gener­
ally recognized successes followed, causing the community to
become pessimistic. Apparently, the failure of predicting the
earthquake in Parkfield, California [Andrews, 1992], where the
World's best monitoring system was in operation, discour­
aged American researchers. This is understandable but today's
widespread pessimism seems to be unjustified because such
a view misses the point that both science and technology are
making rapid progress.
In 1978, the Earthquake Prediction Program of Japan des­
ignated eight "areas of special observation" and two "areas
of intensified observation" (inset of Fig. 4), based on his­
torically disastrous earthquakes, active faults, high seismic­
ity and socio-economic importance [Hamada, 1992]. This
can be taken as an example of set of nationwide intermedi­
ate-term predictions. Comparing the two maps in Fig. 4 indi­
cates that the intermediate-term predictions have been more
or less fulfilled, because most of the large earthquakes after
1978 have occurred in or near the "areas of special observa­
tion". However, none of these earthquakes, including the
Kobe earthquake, was predicted in the short-term. The main
reason for this was that the methods for short-term prediction
were not applied in any of these "areas of special observation".
In fact, the methods, such as densely distributed tilt-meter
and strain-meter monitoring, have been applied only in the two
"areas of intensified observation" where major earthquakes
have not yet occurred. The two "areas of intensified obser­
vation" were selected because a great earthquake exceeding
magnitude 8 is expected to occur there soon, based on long
term predictions. This future earthquake has even been given
a name: the Tokai (meaning East Sea in Japanese) earth­
quake. The postulated epicentral area has lately been extended
westward as in the main map of Fig. 4, so that it is now often
called the Tonankai/Nankai (East South/South Sea) earth­
quake (see Fig. 6).
When people assert that the search for earthquake precur­
sors has proven useless, they often overlook the fact that sci­
entific and effective precursor research has seldom been
carried out. Certainly, seismic networks have been consider­
ably upgraded, so that even micro-earthquakes can now be
detected and located precisely. P r e c u r s o r y changes in
local/regional seismicity, such as fore-shock activity and qui­
escence, may be detected [e. g., Sobolev et al., 2002], but
seismic networks are not suited, by definition, for detecting
non-seismic precursors.
When an earthquake occurs, it is a mechanical vibration of
the ground but since it is caused by breaking of the earth's
crust, which, unlike flawless piece of glass, has highly het­
erogeneous structures, it is reasonable to expect that its prepara­
tory process has various facets which may be observed before
the final catastrophe. Therefore, the science of earthquake
prediction should be multidisciplinary as depicted in Fig. 5.
In seismology, the study of the physical mechanisms of earth­
quake generation has made significant advances, which include
the earthquake dynamics models with pre-seismic slips,

M7.8, 1993

Areas of
I \: special observation
• : intensified observation

M7.5-8 Off Miyagi-ken

M7.3, 2000

M6.4, 2001
M8 Tokai

M6.5, 1997
MS Nankai/Tonankai

Figure 4. Inset shows the eight "areas for special observation" (empty rectangles) and two "areas for intensified observation"
(filled rectangles), selected by the Japanese Earthquake Prediction Program in 1978. Main figure shows the roughly esti­
mated source regions of major earthquakes which occurred afterwards (smaller letters) and expected earthquakes (larger
letters).
354 EARTHQUAKE PREDICTION, SEISMIC HAZARD AND VULNERABILITY
Seismology Electro-Magnetism
Geology Earthquake
Active faults Social science
J Prediction
Geodesy Sc ience ) against the north-westward subduction of the Philippine Sea
GPS
Macro-anomaly
Hydro-geochemistry Animal behavior
Figure 5. Multidisciplinary nature of earthquake prediction science.
although pre-seismic slip for real earthquakes has yet to be ver­
ified, (e. g., Aochi and Matsu 'ura, 2002).
Pre-seismic changes in levels and chemistry of ground water
are providing potentially useful data for short-term prediction
[e. g., Silver and Wakita, 1996; Koizumi et al, 1999]. Macro-
anomalies, such as anomalous animal behaviors may also be use­
ful [ e. g., Ikeya et al, 1997], although it may take some time
for the scientific community to take them up seriously. Social
science should work on establishing better ways for timely dis­
semination of prediction information to the public without caus­
ing undue social unrest. The rapid progress of information
technology will play a major role in all these issues.
The introduction of the global positioning system (GPS) has
revolutionized geodetic science. Movements of the earth's sur­
face can now be tracked with the precision of 1-0.1 cm. Fig. 6
Figure 6. GPS observation of the displacement of central Japan relative to Point 940241 (After website of Geodetic Sur­
vey Institute of Japan). A. Displacement in cm from March 27,2001 to January 24,2004. B. Average speed in cm/year for
1998-2000.
is an example to illustrate the occurrence of anomalous crustal
movements detected by GPS in the Tonankai area, Japan. The
displacement after March 2001 shown in (A) has been drasti­
cally different from the longer term average shown in (B). The
central coastal area, in particular, has moved south-eastward
Plate. This might indicate a pre-slip of the feared great Tonankai
earthquake cited above or a manifestation of a slow earthquake.
Such observations were not possible before GPS.
Another novel direction of research into the mechanics of
earthquake generation is focused on the possibility of elec­
tromagnetic precursors. Since earthquakes are sudden fault
motion that involves fracture of rocks and/or factional move­
ments of rock masses, it is anticipated that the seismogenic
process involves electromagnetic phenomena. The main prob­
lem is whether observable electromagnetic pre-seismic man­
ifestations exist or not. There have been a number of positive
reports during the last decade. Evidence has been found for
electromagnetic phenomena in a wide frequency range from
many parts of the world, including Greece, Japan, Russia,
China, Taiwan, Armenia and Italy [e. g., Lighthill (Ed.) 1996;
Hayakawa and Molchanov (Eds.), 2002; Uyeda and Park
(Eds.), 2002, Balassanian et al, 1997 and references therein.].
Pioneering work has been documented in Greece (see for
instance, Varotsos et al, 1984: Lighthill (Ed.), 1996). A group
of Athens University physicists claim that short-term predic­
tion can be made by detecting anomalous pre-seismic transient
134'E 136°E 138'E 140°E
 UYEDA AND MEGURO 355

mic changes also occur in Japan [Uyeda et al, 2000]. They

were observed before M > 5 earthquakes that occurred
within 20 km or so of our stations (Fig. 8). Clear co-seismic
signals, synchronized with the arrival of seismic waves,
have also been observed for nearby earthquakes [Nagao et
al, 2000].
Pioneering work has also been made on the precursory phe­
nomena in the ultra low frequency (ULF: 0.01 Hz) geomag­
netic field variations. Notable examples are the 1988 M6.9
Spitak [Kopitenko et al, 1993], 1989 M7.1 Loma Prieta
[Fraser-Smith et al, 1990], 1993 M8.0 Guam [Hayakawa et
al, 1996] and 1997 M6.5 Kagoshima earthquakes [Hattori
etal, 2002].
Fig. 9 is the global summary of the ULF investigations,
showing the empirical relationship between earthquake mag­
nitude and epicentral distance. White and black marks show
the earthquakes with and without observed ULF anomalies,
respectively. The slant line indicates the empirical estimate
of the threshold for the appearance of ULF signals. For
36 -! 1 1 1 1 1 h instance, ULF signals of M7 earthquake will be detected
20 21 22 23 24 25 26E within 100 or so km from the epicenter.
Figure 7. Evaluation of VAN prediction. All the earthquakes with In the summer of 2000, there was a large swarm activity
USGS PDE magnitude larger than 5.5 for 1985-2003. Earthquakes in the Izu island region south of Tokyo. In this case, both elec­
are specified next to each circle. For example, 950513 mb 6.1 means tric field and magnetic field measured at far apart stations
year (1995), month (May), day (13th) and magnitude 6.1. The sym­
bol mb is body wave magnitude.
Shaded circles: "successfully" predicted. 130°E 140°E 150°E
White circle with triangle: unsuccessfully predicted.
White circles: missed.

electric currents in the earth. Fig. 7 shows some results of

their method, now called the VAN method (after the initials of
the members of the founding team) [e. g., Nagao et al, 1996;
Uyeda andAl-Damegh, 1999, Uyeda etal, 1999; Kondo etal,
2002]. For the period 1985-2003, 12 out of 16 M=5.5 earth­
quakes in the Greek region were successfully predicted. Suc­
cessful prediction here means that the errors were less than 0.5
for earthquake magnitude, 100 km in epicentral position and
a few weeks in date. The public impact of VAN's predictions
has been quite large because citizens' lives have been saved at
disastrous earthquakes [Uyeda, 2000].
In Japan, several research groups have worked on earth­
quake related electromagnetic phenomena since the early
1990's. At the time of the January 17, 1995 Kobe earth­
quake, anomalous electromagnetic phenomena at frequen­
2 3
cies ranging from extremely low frequency (ELF; 10 —10
2
Hz) to very high frequency (VHF; ~10 MHz) were observed 130°E 140°E 150°E
at many localities [Nagao et al, 2002]. Encouraged by these
Figure 8. Distribution of M > 5 earthquakes with pre-seismic signals
results, seismo-electromagnetic research in Japan has during the period before March 1999. Diamonds and circles showed
become very active. Through deploying about 40 monitor­ pre-seismic electric signals and ULF magnetic signals, respectively.
ing stations for the geoelectric potential changes over the Stars showed both electric and magnetic signatures. EQ 98/09/03
country, it was demonstrated that the VAN-type pre-seis- M6.1 means M6.1 earthquake which occurred on September 3,1998.
356 EARTHQUAKE PREDICTION, SEISMIC HAZARD AND VULNERABILITY

350
predicted. It was also found that even in unsuccessful pre­

300
: Ovnth ULF anomaly
atHouTU IFanbmi
• dictions, apparently meaningful signals were detected although
the interpretation was not correct. The method is still far from
try - Okushiri EQ
perfect and its physical basis is uncertain [Pilipenko et al.,
— Empir cal Thres hold
2001]. However, the performance of the method justifies fur­
——
ther investigation. Active cross-check experiments are now

200 •
Northric ge EQ
underway by several independent groups [e. g., Moriya et al.,
2003, Takano et al., 2003, Kamogawa et al., 2003].
a Pre-seisimic anomalous transmission of electromagnetic
waves has been intensively investigated in the VLF (~ 10 kHz)
8
c
150

Izu Swi
He(

o • band also [e. g., Molchanov and Hayakawa, 1998; Hayakawa

h u Swam rm Spi
a k E O ^ ^ C h i - j C h i EQ and Molchanov (eds.), 2002]. Moreover, the research interest
200007
100 )1 is extended to space science. A micro-satellite dedicated to the
~ ' 2T 1000709
Izu S ° o orthrige m EQ
investigation of the ionospheric perturbations, DEMETER
Erimo E( } 20000 tfiyagi B
Matsusl iro goshima ^Kagosh ma EQ (Detection of Electro-Magnetic Emissions Transmitted from
EQ " E<)1997051 3l 99703: :6
Earthquake Regions) is planned to be launched in 2004 by
O
Ij:u Swarrr Olwate EQ QLoma Prieta E(
, - , ,1
4.5 5.0 5.5 6.0 6.5 7.0 7.5
Magnitude
Figure 9. Magnitude—Epicentral distance plot for earthquakes (EQ)
with and without ULF signatures. (Revised from Hattori et al., in
press). 06/26 08/29 ~ "

showed distinct anomalous precursory changes as shown in

Fig. 10 [Uyeda et al., 2002]. Remarkably these changes started
a few months before the beginning of the swarm activity.
Three major earthquakes in this swarm followed distinct pre­
cursory signals as indicated in Fig. 10 (b). These earthquakes
correspond to the star symbol to the south of Tokyo in the
inset of Fig. 8 and those in the central part of Fig. 9.
Research on electromagnetic phenomena has investigated
a wide frequency range. For instance, Kushida and Kushida,
at the Yatsugatake South Base Observatory, central Japan,
were monitoring the transient reflections of VHF FM radio
waves beyond the line of sight for meteor detection. They
accidentally noted that their system recorded unusual signals
on the night before the Kobe earthquake [Nagao et al., 2002].
They proposed that VHF FM radio waves travelling over the
focal zone of an imminent earthquake are scattered to reach
over-the-horizon distances. They devised a method of short-
term prediction, now called the Kushida method. Since the
mid-1990's, the Kushidas have been practicing actual short-
term prediction experiments [Kushida and Kushida, 2002]. Figure 10. Time change of the 0.01 Hz spectral intensity, in arbitrary
The performance of the Kushida method during 2000-2003 unit, of the geoelectric potential at a station in Niijima Island and that
has been evaluated for M=5.5 earthquakes by checking their of the geomagnetic field at an array station in Izu Peninsula. The
data shown are after some data processing process, (a) Three year
predictions against the actual seismicity [Uyeda et al., 2004].
records of electric (upper panel) and magnetic (lower panel) data. Both
The criteria for successful prediction are as follows: the error
plots show the anomalous signals starting a few months before the
in the date of the earthquake is less than ~10 days, the error onset of the swarm activity (June 26 to August 29,2000). Data gaps
in the epicenter is less than ~50 km, and the error in Magni­ are due to system failure by earthquakes and typhoons, (b) Enlarged
tude is less than 1 Richter unit. About 40% of their predictions for January through October, 2000. Three M > 6 earthquakes shown
were successful and about 30% of M=5.5 earthquakes were in Figs. 8 and 9 are indicated, (c) Seismicity of the region.

CNES (French National Space Agency) [Parrot, 2002]. Thus,
the seismo-electromagnetic research involves scientists from
diverse branches; not only the conventional earthquake science,
such as seismology, geo-electromagnetism, but also the radio,
ionosphere and space science. In view of the urgent need for
promoting interdisciplinary cooperation and for assisting this
new science in the developing world, IUGG established, in
2002, an IAGA/IASPEI/IAVCEI Inter-Association Working
Group on Electromagnetic Studies of Earthquakes and Vol­
canoes (EMSEV: S. Uyeda: Chair).
CONCLUSION
Reinforcement of buildings is the highest priority in the
fight against the seismic hazard and to save human life, from
the engineering aspect. Earthquake science needs to make
utmost efforts in short-term earthquake prediction to further
reduce casualties. Disaster prevention measures and prediction
research are complementary activities.
Acknowledgments. The authors thank Professors Stephen Sparks,
University of Bristol, and Zhongliang Wu, Graduate School of Chi­
nese Academy of Sciences, and Dr. Masashi Kamogawa, Tokyo
Gakugei University for their critical and constructive comments.
Thanks are also due to Drs. Tom Beer, CSIRO, Australia and
G. Heiken, Los Alamos National Laboratory for inviting the authors
to Symposium U4 at the 23rd General Assembly of IUGG (2003) in
Sapporo. Assistance of Mr. Yoichi Noda, Tokai University in prepar­
ing figures is highly acknowledged.
REFERENCES
Andrews, R. A., The Parkfield earthquake prediction of October
1992: the Emergency Services response, Earthquakes & Volca­
noes, 23, No. 4, 170-174, 1992.
Aochi, H. and M. Matsu'ura, Slip- and time-dependent fault con­
stitutive law and its significance in earthquake generation cycles,
PureAppl. Geophys., 159, 2029-2047, 2002.
Aranda, E. et al., Mexico City seismic alert system, Seimol Res.
Lett., 66, 42-53, 1995.
Balassanian, S., A. Mouradian, A. Sahakian, S. Kalinin, M. Babayan and
A. Pogossian, The invesitgation of electromagnetic precursors to
earthquakes in Armenia, Annali di Geofisica, XL, 209-225, 1997.
Chen,Y. T., Z. L. Chen and B. Q. Wang, Seismological studies of
earthquake prediction in China: a review, In 'Earthquake Pre­
diction" (E. Boschi and M. Dragoni, Eds.), 71-109, IL CIGNO
GALILEO GALILEI, 1992.
Fraser-Smith, A. C , A. Bernardi, P. R. McGill, M. E. Ladd, R. A. Hel-
liwell, and O. G. Villard, Jr., Low-frequency magnetic field meas­
urements near the epicenter of the Ms7.1 Loma Prieta earthquake,
Geophys. Res. Lett, 17, 1465-1468, 1990.
Giardini, D. (Ed.), The Global Seismic Hazard Assessment Program
(SHAP), Annali di Geofisica, 42, No.6, 1999.
UYEDA AND MEGURO 357
Hada, Y. and K. Meguro, Application of Power Demand Changes to
Evaluate Building and Dwelling Damages due to Earthquake,
Bulletin of Earthquake Resistant Structure Research Center—
Institute of Industrial Science, 35, 135-144, 2002.
Hamada, K., Present state of earthquake prediction system in Japan,
In "Earthquake Prediction " (E. Boschi and M. Dragoni, Eds.),
33-69, IL CIGNO GALILEO GALILEI, 1992.
Hattori, K., Y. Akinaga, M. Hayakawa, K. Yumoto, T. Nagao, and
S. Uyeda, ULF magnetic anomaly preceding the 1997 Kagoshima
Earthquakes, In Seismo Electromagnetics, 19-28, TERRAPUB,
Tokyo, 2002.
Hattori, K., I. Takahashi, C. Yoshino, N. Isezaki, H. Iwasaki,
M. Harada, K. Kawabata, E. Kopytenko, Y. Kopytenko, P. Malt-
sev, V Korepanov, O., Molchanov, M. Hayakawa, Y. Noda, T.
Nagao, and S. Uyeda, ULF Geomagnetic Field Measurements
in Japan and some recent results associated with Iwateken Nairiku
Hokubu earthquake in 1998, Physics and Chemistry of the Earth,
in press.
Hayakawa, M., R. Kawate and O.A. Molchanov, Ultra-low- frequency
signatures of the Guam earthquake on 8 August, 1993 and their
implication, J. Atmos. Electr, 16, 193-198, 1996.
Hayakawa, M. and O. Molchanov (Eds.), "Seismo Electromagnetics ",
pp. 477, TERRAPUB, Tokyo, 2002.
Ikeya, M., T. Komatsu, Y. Kinoshita, K. Teramoto, K. Inoue, M. Gon-
dou and T. Yamamoto, Pulsed electric field before Kobe and Izu
earthquakes from seismically-induced anomalous animal behav­
ior (SAAB), Episodes, 20, 253-260, 1997.
Kamogawa, M., H. Fujiwara, H. Sakata, J. Y. Liu, H. Ofuruton,
Y. J. Chuo, Seismo-atmospheric disturbances observed by anom­
alous transmission of VHF electromagnetic waves, IUGG 2003
Sapporo Abstracts Week A, A. 189, JWS01/02P/D-034, 2003.
Kanamori, H., E. Hauksson and T. Heaton, Real-time seismology
and earthquake hazard mitigation, Nature, 390, 461-464, 1997.
Keilis-Borok, V I. and V G. Kossobokov, Premonitory activation of
earthquake flow: algorithm M8, Phys. Earth and Planet. Inter,
61, 73-83, 1990.
Koizumi, N., E. Tsukuda, O. Kamigaichi, N. Matsumoto, M. Taka­
hashi and T. Sato, Preseismic changes in ground water level and
volumetric strain associated with earthquake swarms off the east
coast of the Izu Peninsula, Japan, Geophys. Res. Lett, 26,
3509-3512, 1999.
Kondo, S., S. Uyeda and T. Nagao, The selectivity of the Ioannina
VAN station, J. Geodynamics, 33, 433-461, 2002.
Kopytenko, Y.A., T.G. Matishvili, P.M. Voronov, E.A. Kopytenko,
and O.A. Molchanov, Detection of ultra-low-frequency emissions
connected with the Spitak earthquake and its aftershock activity,
based on geomagnetic pulsations data at Dusheti and Vardzia
observatories. Phys. Earth Planet. Inter., 1 1 , 85-95, 1993.
Kumazawa, M. and S. Maruyama, Whole Earth tectonics, J. Geol. Soc.
Japan, 100, 81-102, 1994.
Kushida, Y. and R. Kushida, Possibility of earthquake forecast by
radio observations in the VHF band, J. Atmosph. Electricity, 22,
239-255, 2002.
Hui, L., China's campaign to predict earthquakes, Science, 273,
1484-1486, 1996.
358 EARTHQUAKE PREDICTION, SEISMIC HAZARD AND VULNERABILITY
Lighthill, Sir James (Ed.), "A Critical Review of VAN", pp. 376,
World Scientific, 1996.
Mayorca, P., and K., Meguro, Proposal of a new economic retro­
fitting method for masonry structures, Proc. 27th Earthquake
Engineering Symposium, CD-ROM, Japan Soc. Civil Engineers,
2003.
Meguro, K., Dynamic evaluation model for large earthquake disas­
ter, J. Geography, 110, 900-914, 2001. (In Japanese with Eng­
lish abstract).
Meguro, K, and T. Takahashi, System for promotion of retrofitting
of existing pre code-revision weak structures, J. Social Safety Sci­
ence, 3, 81-86, 2001. (In Japanese with English abstract)
Molchanov, O. A. and M. Hayakawa, Subionospheric VLF signal
perturbations possibly related to earthquakes, J. Geophys. Res.,
103, No. A8, 17,489-17,504, 1998.
Moriya, T., T. Mogi, M. Takada and M. Kasahara, Modification of
Kushida's method and preliminary result - high possibility method
for earthquake forecast based on the exploration for VHF scat-
terer in the atmosphere, IUGG 2003 Sapporo Abstracts Week A,
A. 188, JWS01/02P/D-032, 2003.
Nakamura, Y, On the urgent earthquake detection and alarm sys­
tem (UrEDAS), Proc. Ninth World Conf. Earthq. Engineering,
17, 673-678, 1988.
Nagao, T., Y Enomoto, Y. Fujinawa, M. Hata, M. Hayakawa,
Q. Huang, J. Izutsu, Y Kushida, K. Maeda, K. Oike, S. Uyeda,
T. Yoshino, Electromagnetic anomalies associated with 1995 Kobe
earthquake, J. Geodynamics, 33, 401^111, 2002.
Nagao, T, Y Orihara, T. Yamaguchi, I. Takahashi, K. Hattori, Y. Noda,
K. Sayanagi and S. Uyeda, Geophys. Res. Lett, 27, 1535-1538,
2000.
Nagao, T, M. Uyeshima and S. Uyeda, An independent check of
VAN's criteria for signal recognition, Geophys. Res. Lett., 23,
1441-1444, 1996.
Nishimura, A., Y. Ueno, S. Fujiwara, I. Ijiri, T. Fukunaga, S. Hishida,
K. Hatake, A. Tanegashima, H. Kinoshita, Y. Mizoi, and Y Tatsuno,
Medical examination report on the Great Hanshin Earthquake, J.
Advance in Legal Medicine, 3, 234—238, 1997a.
Nishimura, A., Y Ueno, S. Fujiwara, I. Ijiri, T. Fukunaga, S. Hishida,
K. Hatake, A. Tanegashima, H. Kinoshita, Y. Mizoi, andY. Tatsuno,
Statistical Report on Casualty of the Great Hanshin Earthquake,
J. Advance in Legal Medicine, 3, 346-349, 1997b.
Ohtake, M., Reinforcement is urgently needed for wooden houses,
Japan Society for Disaster Information Studies News Letter, No.
13, 1, 2003 (in Japanese)
Parrot, M., The miceo-satellite DEMETER, J. Geodynamics, 33,
535-541,2002.
Rikitake, T., 'EarthquakePrediction'', pp. 357, Elsevier, 1976.
Scholz, C. H., L. R. Sykes andY. P. Aggawal, Earthquake Prediction:
a physical basis, Science, 181, 803-809, 1973.
Shin, T. C , K. W. Kuo et al., A preliminary report on the 1999 Chi-
Chi (Taiwan) earthquake, Seismol. Res. Lett., 71, 24-30, 2000.
Silver, P. G.. and H. Wakita, A search for earthquake precursors, Sci­
ence, 273, 77-78, 1996.
Sobolev, G. A., Q. Huang and T. Nagao, Phases of earthquake's
preparation and by chance test of seismic quiescence anomaly, J.
Geodynamics, 33, 413^124, 2002.
Stein, S. and E. Klosko, Earthquake mechanisms and plate tecton­
ics, in " International Handbook of Earthquake and Engineer­
ing Seismology " (W. H. K. Lee, H. Kanamori, P. C. Jennings and
C. Kisslinger (Eds.), Academic Press, 69-78, 2002.
Takano, T., K. Sakai, A. Yamada, H. Higasa and S. Shimakura, Elec­
tromagnetic phenomena possibly associated with earthquakes
obtained with broadband observations, IUGG 2003 Sapporo
Abstracts Week A, A. 188, JWS01/02P/D-030, 2003.
Utsu, T, A list of deadly earthquakes in the world: 1500-2000, in
"International Handbook of Earthquake and Engineering Seis­
mology" (W. H. K. Lee, H. Kanamori, P. C. Jennings and
C. Kisslinger (Eds.), Academic Press, 691-717, 2002.
Uyeda, S., In defense of VAN's earthquake predictions, EOS, Trans.
Amer. Geophys. Un., 61, no.l, Jan. 4, 3 & 6, 2000.
Uyeda, S., Continental drift, sea-floor spreading and plate tecton­
ics/Plume tectonics, in 'International Handbook ofEarthquake and
Engineering Seismology" (W. H. K. Lee, H. Kanamori, P. C. Jen­
nings and C. Kisslinger (Eds.), Academic Press, 51-67, 2002.
Uyeda, S. and K. Al-Damegh, Evaluation of VAN method, In:
"Seismo-electromagnetic Phenomena ", Hayakawa (Ed.), Terra
Scientific, Tokyo, 58-69, 1999.
Uyeda, S., K. Al-Damegh, E. Dologlou and T. Nagao, Some rela­
tionship between VAN seismic electric signals (SES) and earth­
quake parameters, Tectonophys., 304, 41-55, 1999.
Uyeda, S., M. Hayakawa, T. Nagao, O. Molchanov, K. Hattori, Y. Ori­
hara, K. Gotoh, Y. Akinaga and H. Tanaka, Electric and magnetic
phenomena observed before the volcano-seismic activity in 2000
in the Izu Island region, Japan, Proc. Nat. Acad. Sci., 99,
7352-7355, 2002.
Uyeda, S. and A. Kumamoto, Evaluation of the Kushida method of
short-term earthquake prediction, Proc. Japan Acad., Ser. B, Phys.
Biolog. Sciences, 80, 140-147, 2004.
Uyeda, S., T. Nagao, Y. Orihara, Y Yamaguchi and T. Takahashi,
Geoelectric potential changes: Possible precursors to earthquakes
in Japan, Proc. Nat. Acad. Sci., 97, 4561^4566. Uyeda, S. and
S. Park (Eds.), "Recent Investigations of electromagnetic Vari­
ations Related to Earthquakes", J. Geodynamics, 33, 377-570,
2002.
Seiya Uyeda, Earthquake Prediction Research Center Tokai Uni­
versity, Shizuoka 424-8610, Japan (suyeda@st.rim.or.jp)
Kimiro Meguro, International Center for Urban Safety Engineer­
ing, Institute of Industrial Science, The University of Tokyo, Meguro-
ku, Tokyo, 153-8505, Japan