Instructions for use - HUSCAP...inteRsity attenuation with distances was deduced and its anomaly map...
Transcript of Instructions for use - HUSCAP...inteRsity attenuation with distances was deduced and its anomaly map...
Instructions for use
Title Dense Survey of Seismic Intensity for Wider Application to Earthquake and Engineering Problems
Author(s) Ohta, Yutaka; Kagami, Hiroshi; Okada, Shigeyuki; Murakami, Hitomi Ohashi
Citation 北海道大學工學部研究報告, 145, 165-176
Issue Date 1988-12-27
Doc URL http://hdl.handle.net/2115/42171
Type bulletin (article)
File Information 145_165-176.pdf
Hokkaido University Collection of Scholarly and Academic Papers : HUSCAP
北海道大学工学部研究報告
第145号(昭和63年)
Bulletin of the Faculty of Engineering,
Hokl〈aido University, No. 145 (1988)
Demse Sewvey of Se蓋senre玉。 IRteems晦fo罫W量“e罫A舩量ca樋。晒。
Eartkquake and ERgineering Prebgeitas
Yutaka OHTA, }liroshi KAGAMI, Shigeyuki OKADA
and Hitomi Ohashi MURAKAMI
(Received September 10, 1988)
Abstract
Motivated by a belief that quaiified seismic intensity data should have a wider
application capability in engineering seismology and earthquake engineering, we have
been continuing a series of investigations, based essentially upon a questionnaire
method advanced in both of survey scheme and subsequent data processing. ln this
paper we demonstrate a few applications to drawing minute isoseismal maps, to
exploriRg seismic source and path characteristics, and further to eiucidate rr}icrozoning
characteristics in urban and rural areas. The exemplified incidences are 1982 Off-
Urakawa, 1983 Central Japan Sea, 1981 Corinthos, 1984 Morgan Hill and 1986 Holiister
earthquakes. The obtained results are well examined through various comparisons
with instrumental and other data.
1. lntroduction
For more than a decade we have been condueting a dense survey on seismic intensity
for almost all the moderate-to-large earthquakes occurring in and around Japan and also
have been applying it to foreign earthquakes. We had developed the method by means of
questionnaire in order to estimate seismic iRtensities precisely and densely’). The major
reasons are as foliows. The officially announced seismic intensities by JMA (Japan
Meteorological Agency) have serious demerits in that its determination is made in a limited
number as 2-3 locations per prefectural unit with the average area of 10,00e km2 and ehat its
accuracy is rather low due to the summary determination. Therefore the anneunced inten-
sities are insufficient to be basic data for succeeding studies in seismology and engineering.
High-precision and dense survey has iong been desired in this respect.
In this paper, maj or subjects covered and results obtained through this series of studies
are summarized in its applications to engineering seismological and earthquake engineering
problems, demonserating the significaRce of dense and qualified seismic intensity data.
Department of Architectural Engineering
166 Yutaka OHTA, Hirtshi KAGAMi, Shigeyuki Oi〈ADA and Hitomi Ohashi MuRAKAMi2
2. i982 Off-Urakawa Earthquake
On March 21, 1982 a large earthquake with M = 7.1 occurred off Urakawa, Hokkaido
district iR northern most Japan and caused a significant amount of damage iR several
town-ships along the Pacific coast and a few inland cities. Although the maximum intensity
by the JMA was VI at Urakawa, Ro continuation to V was reported at any affected areas.
We performed dense questionnaire surveys for this earthquake with special objectives of
obtaining a precise isoseismal map, elucidating intensity distribution along the coastal line
passing through epicentral area, aRd charaeterizing microzoning Sapporo city.
2.1. lsoseismal map
The delivery of 6,325 questionnaires was made in relation of population as 100 for a
city with 100,eOO population or higher, 50 for the other cities and 25 for a town or a village,
and we received 86.5% answers. Figure 1 illustrates O.25 increment-isoseismal maps drawn
by use of obtained intensities by municipali£ies. The actual drawing was made autoix}atl-
cally by Ohta and Kagami’s techRique2). The observed isoseismals are somewhat irreguiar
than circular. Contour liRes seem to be extended in E-W direction, but contracted in N-S
dlrection. This is a comrnon character of isoseismals due to earthquakes in Hokkaido
district. To know more in detail about this peculiarity, the deviatlon from the standard
inteRsity attenuation with distances was deduced and its anomaly map was drawn as is shown
in Figure 2 and compared with the so-called Bouguer anomaly map compiled by Hagiwara3).
As is seen in this figure we find a significant correlation between two maps. lt is evident
that in a region where intensity anomaly is positive deep seated soft soils deposit widely and
iR its negative region geologically old or volcanic rocks are distributed. This clearly shows
that the obtained isoseismal map reflects well the regiona}ity in seismic wave propagation.
1ノ
O
o’
,
.
.
.
’i.e
Sapporo
.
御Tokyo
×・
一
3.0:tt
邑 ●●
.
K ;
.賠
汐・
x4.0
。二k:’k・‘
;:.
.
.一覧
.
.
げ.
f.
;一 一
一 一 一
,
.
,
サ の
覧” .’ ・∫..
● ●
噂
.
.
.
曜
N
.
.
O 100 km.
Fig. 1 Obtained isoseismal map of 1982 Off Urakawa earthquake.
3 167
Intensity anomaly
g塗
.毒
過
遡
。
歪..
霧,
薯.i17rSo
きel
隔
争働。
O 100km
Bouguer anomaly (mgal )
Fig. 2 Comparison between intensity (1982 Off Urakawa eq.)
and Beuguer anomalies.
2.2. Epicentral regioR
A dense questionnaire survey was performed in severai towns located on tke Pacific
coastal iine in epicentral region. Figure 3 summarizes observed and calculated in£ensity
distributions, together with other physical data. The upper most curve shows relative ground
deformatioR between this earthquake. The second one is PGA(peak ground acceleration)
distribution curve derived by Kobayashi et al‘〉. The lowest oRes are from our caiculations
Hekkaid
騒
(15旨
S10窟ラ
85琶P G t
の ほ
/v\
◎
.
一.一一 一 ” 一 一hny 一
4eo
300
200
100
80
60
5,5
一 observed
e estimatecl
.
o
/./eXexxxe
.
5,0
4.5
Tom
4.0
Mitsuishi
leo so 60 40 20 o 20 40 60 Distance £rom Mitsuishi [km]
Fig. 3 Comparison of intensity (1982 Off Urakawa eq.) with other physical data.
168 Yutaka OHTA, Hirtshi KAGAMI, Shigeyuki OKADA and Hitomi Ohashi MuRAi〈AMi 4
and observations. The calculated intensities are based upon the empirical equation pr. oposed
by us5). All the dis#ributions cited here are in remarkable agreemeRt. A comparison of the
ground deformation curve with that of observed ingensity tells that the occurreRce of
dominant ground failures is limited within the region where the intensity is equal to or higher
than the lower bound of VI.
2.3. UrbaR area
A similar but more dense survey was performed in Sapporo city distribu£ing IO,eOO
questionnaires so as to cover the whole residential area, and we obtained 83 %o answers.
Figure 4 shows the histogram through which one can see that the observed intensities spread
from II to V in JMA scale, although the average intensity of 3.8 is iR good agreement with
that by JMA Sapporo station. Areal distribution of 1’ntensities are shown in Figure 5
together with site plots where coRsiderable damages occurred. Thus obtained microzoning
characteristic reflects well the seismic severity distribution due to this earthquake.
Next, we examined the physical reason of such mlcrozoning characteristics. Since
Sapporo city is in an epicenerai distance range of (i45 ± 15)km, the distance effect seems
small. Source and path effects are also less influential because, at this distance, source
4,25
2000. 一 3・75, 4,0 3,5 axX>(lt{k,,. ,..,...±il.Jl’i
E. 1 Pg lj,1,i i.}’llllSXilCl”1・tsf,一一。,1,1,1’1’1’l i, l l
Ihtensity 3.’s”’一”51’i’s?”lillj’:/S 4’O e Darnaged houses
Fig. 4 Histogram of intensities obtained in Fig. 5 lntensity microzonig map in Sapporo city
Sapporo city (1982 Off Urakawa eq.〉. (1982 Off Urakawa eq).
sandy and clayey silt
Ale.H
偽vi
o’pri
写¢
で
眺.‡ミ
の
sH”
一一 〇
1 ∬ 皿 Iv V WJMA豆皿 IV V VI V旺圃 IX MSK
1.5
1.O
O.5
o
一〇,5
一1.O
一1,5
talus
an’р?sit?
gravel
volcanic ash
river deposit ×
sandy silt
一v一
s
peat
silt
/
Fig. 6 Kistograms of intensity deviation by soil types in Sapporo city
(1982 Off Urakawa eq.) .
5
Dense Survey of Seismic lntensity for Wider Application to
Earthquake and Engineering Probiems 169
finiteness can be out of consideration
and all of the paths to Sapporo are
approximately identical. lf these are
true, only the physical possibiiity of
such spatiai intensity distributioR
should be explored in the so-cailed site
effect due to surface soils. Figure 6
illustrates a iine-up of histograms of
intensity deviation (61) from the aver-
age value by surface soil types. The
right side 4 histograms are for alluviai
soils, which indicates rather large
intensities. and the remainder are ’
mostly for older soils and rock siees.
that it seems to be saturated.
chimney
ext. wall
ground
heavy injured
slight injured
water tube
kerosene leakage
gas leakage
e
.
e e
一 e - e 一 t 一 一 一
禽 ●●
一ee e 一e一 e
一 t
- t e te 一
. , .e e e e .
一 t 一
=Ut II-vr 1] 〔工MSK intensity
mxg. 7 Relation between intensity and various
disasters in Sapporo city (1982 Off Urak-
awa eq.).
The soil layer thickness is effective up to 10m but after
Based upon this preliminary analysis we can develop an empirical equation to estima.te
seisrRic intensities in Sapporo city. The intensity at a given site,
Ii=f(M,A) 十 O.256
0.182
0.050
0.040
-O.051
-O.156
-O.278
-0396
-O.406
silt
peat
voicanic ash
sandy silt
sandy and clayey silt
river deposlt
grave圭
andesite(weathered)
talus
Ii, is written as
十〇.04 H,
where
f(M,A) == 2 M-4.601 ’ log A-O.OO166A-O.32 (Kawasumi6))
and M the JMA magnitude, A epicentral distance in km, H thickness of surface soil in m.
Figure 7 summarizes various disasters observed in this city in reiatioR with seismic
intensities.
3. 1983 Central Japan Sea Earthquake
A great earthquake occurred on May 26, 1983 and brought serious damages to northern
distrlcts in mainland Japan. Because of its large magnitude ef 7.7, the affected areas were
so wide coveriRg Hokkaido, Tohoku and Kanto districts of neariy IO prefectures. For this
earthquake a dense survey was performed similariy to the 1982 Off-Urakawa earthquake,
including 435 municipaiities. The total number of questionnaires deiivered exceeds weli 10,
170 Ytttal〈a OHTA, Hirtshi KAGAMi, Shigeyuki OKADA and Hitomi Ohashi MuRAKAMi 6
OOO, by which almost all the areas with intensity of III and higher were iRvestigated. The
answers arrived to a total of about 800/o. By use of these data several analyses were
conducted.
3’1●亙タごs濃紫重器、、、yd。,,_。,、。n一∴..・黒黒
was made at 403 unit areas. Althougk the
maximum intensity by JIV[A is reported as V
at 3 stations, we could reveal that there are
several areas where it reached VI. The
isoseisma} map drawn with e.25 iRcrement is
illustrated in Figure 8. From this we find
several significant features. One is that the
area with intensity V and higher spreads
widely in NS direction than in EW direction.
This seems due to configurational charac-
teristics of the earthquake fault. The other
contours have also the similar tendency of
extendiRg in a NS direction. And this ten一
Akita
x
.憾
嚢
/
ノ
;,cPN
の ゲ
TokyQ
O 50 100 150km
鴫,5
Lom
Fig. 8
篇
瓢
、臥墜嚇i一
Obtained isoseismal map of 1983 Central
Japan Sea earthquake.
dency is rather common eo most isoseismal maps for shallow earthquakes occurring in the
northerR part of Japan.
3.2. Source amdi patk effects
The N-S directionally extended isoseismal contours in Figure 8 can not be explaiRed
by a simple attenuatioR equation assuming a point source and seems to reflect well the
finiteness ef the seismic source. We have already proposed an empirical equation appropri-
ate for near-distance intensity distribution enc}osing a seismic source area (Ohta5)). Let us
apply the same equation to this earthquake. AccordiRg to the equation, the intensity at any
point on the ground is expressed as
I=:cilogDi‘fLfWivatlir:r;s-p)dg’drp ,
where D is the average resultant slip measured by the coerdinate system(e,ij) oR the fault
surface, k is the coefficient related to the slip contribution upon intensity, and p is the
coefficien£ related to the attenuation due to the distance, and the appropriate values are
empirically derived for k and p as O.25 and 2-3 respectively. Here, we perform the following
examinations. Namely, through the nuTnerical comparison of calculated and observed
intensities at near-distances we attempted to determine the optimum configurational solu-
tions in terms of fault leRgth and strike directioR. The ranges we examined are for fault
length L frorn 40 to 200km and for strike direction from N300 W to N300 E, assuming thatthe
fault wid£h is knowR as W= 40km and the dip angle as 300 E . The actuai calculations were
done in such a way as to evaluate the correlation between the observed and the calculated
intensities in ranges higher than III. Figure 9 illustrates contours of the thus obtained
7
Dense Survey of Seismic lntensity for Wider Application to
Earthquake and Engineering Problems 171
correiaeion coefficient. From this we know that the optimum values are L=120km and N20e
E, though the probable va}ues for fauit length and strike direction are of a sornewhat wlder
range. These va’lues are in fairiy good agreement with those derived from instrurneRtal
data. This clearly tells that the obtained isoseismals well reflect the source finiteness and
configurational characteristics.
148. 750E・
50e
伊、
\}/紛V,〉・i/ .!c
・ 一v」賢:一・一・一・ 一一一 40eN
ε
ど
遣bρ
£
当
一 の 葡囎 曜 一 「 一 ・ 榊 . }
! ’ N7r-Nf. !
ivGL翌窒R
W-401〈m
200
180
160
140
120
100
80
60
40
× .80
XN .85 -N
.87 〈
.85 v N30ew 20e loe oo Nlo“E 20“ 30e
Strike angle
Fig. 9 Estimation of fault length and stril〈e direction
(1983 Central Japan Sea eq,) .
40“ soe
3.3. Local site effect
A case study in a town located aloRg the Japan Sea coastal line with a popuiation of
le,OeO, a whole survey for ali the families was performed and compared damage and other
data in order to eiucidate local changes of seismic effect. Figure 10 illustrates the observed
inteRsities by 18 sub-areas iR comparison with strttctural damage rate and periods necessary
for recovery of daily life and for restoration of dwelling. We understaRd that all of them in
V VI JMA
dune
coastal
piain
terrace
coastal
plain
[
Tこ
E
ユ
2
3
45
6
7
89
1011
12
13
14
15
16
ユ7
18
工
〉
/e x
e/
\曾 ノ
置
1 逸
ef/
X.e
e./
Xe
.2:::1二:盲
嫉臨dera・・
X一×’)s.
〉〕ヲll、
舞6i・・1・
ts’
TopographyFig. IO
¥s
tt-i-L一一一.L一一一一一i一一一一一一一一i
ve M MSK O 20 40 60 80 1001ntensity Dwelling damage ifoO
/o 曳 o. e
’N. t /1/二・・
,.’“H/e”一’一一@houses
.t
メfP 1
’〕レ・t
oな /o ’tee/e
璽’\・
鮫
L.一:.fse”“ daily life
btu“al...L-uny
Comparison of intensity with damage and aftermath in Wakami town
Akita prefecture (1983 Central Japan Sea eq.) .
10 100Restoration length(day)
’
172 Yutal〈a OHTA, Hirtshi KAGAMi, Shigeyuki OKADA and Hitomi Ohashi MuRAKAMi 8
sub-areai distribution are in dominant correlation and therefore the seismic severity mea-
sured in terms of seismic intensity is the most fundamental factor oR direct and indirect
earthquake disasters. lf we compare this iRteRsity change with local geological data, we
find that the sub-area wlth highest intensity is composed of the latest sand duRes and the
second highest areas are of alluvial soils and in the areas with surface soil older than alluvium
the intensities are clearly low.
4. 1981 Corinthos, Centra1 Greece Earthquake
In 1981, February 24th a shallow earthquake of magnitude Ms==6.7 occurred in the
eastern part of Corinthian Gulf in Central Greece, which is accompaRied by a largest
aftershock in February 25th of rnagnitude M[s=6.4, causing widespread damage to an area
larger than 1,500 km2. The focal depths of the two events were about 15 km under the sea,
nearby the tiny Alkionides island7) (Figure 11).
The questionnaire survey was planned in order to evaluate the seismic intensity and
also e}ucidate the direct eo indirect damages. After considering the affected region, the
capital city of Corinthos having 45,000 population was chosen as one of the survey sites. The
city lies on the eastern end of the Gulf oR alluvial deposlts where damages were concentrated.
The other survey site was sought in a rural area so that a comparative study versus urbanized
area could be performed. The village of Perahora (populatioR 2,20e) was selected where the
most extensive damages were seen. The village had a £ot’ ≠戟@of 470 houses and lies on
rRountaiRous firm ground.
The questionnaire form was prepared translated from Japanese to Greek and some
more questions inquiring about social impacts weye added. The first part of questionnaire
sheet, with 29 items, is arranged so as to correspond to the definition of MSK intensity scale.
The second part, with 47 items, asks about various aspects of earthquake damage as well as,
the restoration process, economical impact on household income and some mental effects on
people. The survey was taken place in July 1985 and by directly interviewing the local
people with queseionRaire sheets. We obtained 137 aRd 66 effective aRswers by Corinthos
Cer主nthian Gu}f
×
ゆ Perahora
ゆ
・⊥・
㌔・灘981
つ × Feb.25,工98王
M=6.4.
“㌔黛
Ci ( ,pt
Saronic Gulf
Athens
p
A 30蓉
雲2・
き
kpa
C’ 10
o
CorinthosPerahora
Fig. 11 Epicenters and surveyed sites for 1981
Corinthos, Central Greece earthquake.
V V至 W ~張’ 茎X
Intensity(MSK)
Fig. 12 Histograms of obtained intensities for Cor-
inthos and Perahola (1981 Corinthos eq.).
9
Dense Survey of Seismie lntensity for Wider Application to
Earthquake and Engineering Problems 173
and Perahora iRhabitants.
In Corinthos we had沁te!lsities ra漁ging from 5.6 to 8ユ, with an average intensity of 6.
8 when rounded out is the equivalent to the intensity VII in MSK scale. ln Perahora we had
intensities ranging from 6.6 to 8.4, with an average of 7.6 which rounded out is equivalent to
VIII. Thus, a clear difference between the intensities in the two sites was obtained as shown
in Figure !2.
Through the survey the following secia} effects are also revealed. The whole eco-
nomic activities of the region were decreased and are not yet compietely recovered at the
time of answering the questionaires. ln Corinthos more than 25% of the families liviRg in R.
C. framed buildings, have changed dwelling place to seek fer a safer residence althougk the
damage to their buildings was not so heavy. ln Perahora, where 70% of the houses become
uninhabitable, extensive reconstruction took place and at the time of this survey was in i£s
final stages. But still many of the families live in shelter houses of very low standards
because they cannot a£ford the reconstruction cost.
5. 1984 Morgan Hill and 1986 Hollister Eartkquake of CaliforRia,USA.
In order to examine the effectiveness of questionnaire survey and intensity estimates
in MM scale, a trial survey was planRed in California, USA and the following conditions were
considered in the choice of target earthquakes. a) The earthquake is recent so that people
can remember its effects and damages correctly. b) The affected area is properly populated
and several towns are distributed. c) lnstruinental records were obtained. Thus the 1984
Morgan Kill earthquake (ML=6.2) in California was choselt. While preparing for the survey,
the Hollister earthquake (ML=5.5) occurred on January 26, 1986 in the same region, about 65
Table 1 Studied earthquakes
1984Morgan Hlll 1986Hollister
Data Apr,24, Tue, Jan.26, Sun.
Time 1:15PM 11:20AMMagnitude(ML> 6.2 5.5
Max intensity(MM) VI王1:Limited area V王1:Cienega VaHey
V王1:Mor墓an Hill VI:Hollister
Strong motion record L29g;Coyote Dam 0.31g:Hollister
(PGA) 0.469:Gilroy
031g:Holllster
km apart. We decided to add another set of the same questions for the second earthquake
for further comparison. Tabie 1 shows the attributes of the two earthquakes8).
We contacted county boards of education and school districts asking cooperation to
distribute and collect questionRaires through schools to parents. We chose i6 public school
174 Yutaka OHTA, Hirtshi KAGAMi, Shigeyuki OKADA and Kitomi Ohashi MuRAKAMi 10
San Francisco
5∵2\諜1984 ㌔~9
⑧ 7.4\ ㊤ 4,1
5.6 ⑨ 、 4.8 ⑲
7.3⑩ ㊥ ⑩ 4、1⑧
・ξ。6.37’0 41i7。・84・3
ゆ
6.1 @ 4・6
6.4
0 20km O 20 k∬}
」一一一_』 」_一__鱒一
Morgan Hill EQ. Hollister EQ.
F孟g,13 0btained intensities for two earthquakes in California.
ゆ
6.4 ×
Jan.26,1986
ML=5.5
』わ’お
9.日
零
沿
ヒ
お
8
7
6
5
4
3
@ 84 Morgan Hill
o 86 Hollister
/
/
/
/
o /o /
/
/ o/o
.曹/
/蟹/
曾 /
⑳ 吻
/
/
/PG
lg
O,5
O.l
O.05
o
Coyote dam
aso @
Halls Valley
㈱霧 ⑱
345678 45678 Reported intensity intensity
Fig. 14 Comparison of obtained intensities with Fig. 15 Comparison of obtained intensities with
reported ones for the two earthquakes in PGA of gtrong motion records for two
California. earthquakes in Califomia.
in Santa Clara county, 5 in Santa Cruz county, 3 in Monterey county and 2 in San Benito
county within the MM 6 isoseismal of Morgan Hill earthquake. Hundreds questionnaires
11
Dense Survey of Seismic lntensity for Wider Application to
Earthqual〈e and Engineering Problems 175
were mai}ed to each ef 26 schools in late April, !986. The rnaps in Figure 13 show the
surveyed area and the epiceRters.
Calculated intensities and observed intensities are compared in Figure !4. Observed
IRteRsities of Morgan }{[ill earthquake are from Toppozada and those of the Hollister event
are assigned based on USGS earthquake report and news articles. . Calculated intensities
agree very well with the observation by a seismologist through the range of 4 to 7.
CDMG9) and USGCie) have an extensive network of strong motioR instruments in the
surveyed area. Calc雛lated i就ensity is compared with PGA ill Figure 15。 Good correlation
between intensities and PGAs is found.
6.Conc畳ud重魏g Remarks
Through this series of studies for developing seismic intensity stirvey by means of
questionnaire method we found that dense and qualified intensities are rather easily obtained
aRd such intensity data are well applied for advanced analysis on the seismic source and path
characteristics and for comparative study of damage and other data. ln conclusion we
should iil〈e to say tkat, since this questionnaire method is sknple and sufficintly systematic
to be performed with no much special technique, it should have wider application capability
to any other countries than Japan.
Aclmowledgment
The authors wish to acknowledge that these surveys have been accomplished with the
cooperation of hundreds of thousaRds of people who deigned to answer this questionRaire.
The survey in Greece was performed by cooperation of Mr. Antonios Pononis, Fereign
Student of Hokkaido University, and they express their thanks to him.
RefereRees
1 ) Yutaka Ohta, Noritoshi Goto and Hitomi Ohashi (1979). A Questionnaire Survey for Estimating Seismic
lntensities, Bull, Fuc. Eng, Hokkaido Univ., 92, 117-128 (in Japanese).
2) Yutaka Ohta (editor) (1983), A Comprehensive Study oii Earthqual〈e Disasters in Turkey in View of
Seismic Risk Reduction, Hokkaido Univ,, Sapporo, Japan.
3) Yul〈io Hagiwara (1983). Seismic lntensity of the 1982 Urakawa-oki Earthquake as Related to the
Bouguer Gravity Anomaly, Jour. Geodetic Soc. 3apan, 29, 164-165 (in japanese).
4) Hiroyoshi Kobayashi, Kazuoh Seo and Saburo Midorikawa (1982). Damage Distribution due to 1982
Urakawa-oki Earthquake, Proc. Natural Disaster Sci., 19, 159-162 (ln Japanese).
5) Yutal〈a Ohta and Hiroshi Kagami (1982). An Automatic Drawing Technique of Contour Maps of
Seismic lntensity and Other Spatially Distributed Earthquake Engineering Data, Proc. lnter. Microzona-
tion Conf. 3, 1405-1416.
6 〉 Hiroshi Kawasumi (1954). lntensity and Magnitude of Shallow Earthquaks, Travaux Scientifique, Publ.
BCSI, Ser. A, 19, 99-/14.
7 ) Panayotis G., Carydis et al. (1982) The Central Greece Earthqual〈e of February-March 1981一 A Recon-
naissance and Engineering Reportm, National Academy Press, Washington, D. C,
176 Yutaka OHTA, Hirtshi KAGAMi, Shigeyul〈i OKADA and Hitorni Ohashi MuRAKAMi 12
8) CaHfornia Department of Conservation, Division of Mines and Geology (1984>. The 1984 Morgan Hill,
California Earthquake, Special Publicatien 68, 271.
9) California Department of Conservation, Division of rVlines and Geology (1984). CDMG Strong-Motion
Records from the Morgan Hill, California Earthquake of 24 April 1984, CDMG report.
10) A. G. Brady et al. (1986). Preliminary Report on Records from the USGS-maintained Strong-Motion
Network in the Hollister Area January 26, 1986, Open File Report, 86-156.