92 Physicsof theEarth andPlanetaryInteriors, 38 (1985)92—109ElsevierSciencePublishersBY., Amsterdam— Printedin TheNetherlands

Thenear-sourcestrong-motionaccelerogramsrecordedby anexperimentalarrayin Tangshan,China

KezhongPeng1, Liii Xie 1, ShabaiLi 1, D.M. Boore 2 W.D. Iwan3 andT.L. Teng”

‘InstituteofEngineeringMechanics,AcademiaSinica (People’sRepublicof China)2 US. Geologicalsurvey,MenloPark, CA (USA.)

~California Instituteof Technology,Pasadena,CA (USA.)~UniversityofSouthernCalifornia, Los Angeles,CA (USA.)

(ReceivedNovember23, 1983; revisionacceptedAugust 16, 1984)

Peng,K., Xie, L., Li, S., Boore,D.M., Iwan, W.D. andTeng,T.L., 1985. Thenear-sourcestrong-motionaccelerogramsrecordedby anexperimentalarrayin Tangshan,China.Phys.Earth Planet. Inter., 38: 92—109.

A joint researchprojecton strong-motionearthquakestudiesbetweenthePeople’sRepublicof ChinaandtheUnitedStatesis in progress.As a partof this project,an experimentalstrong-motionarray,consistingof twelve KinemetricsPDR-1 Digital Event Recorders,wasdeployedin the meizoseismalareaof the M~= 7.8 Tangshanearthquakeof July28, 1976. These instrumentshave automaticgain ranging, a specifieddynamic rangeof 102 dB, a 2.5 s pre-eventmemory,programmabletriggering,andareequippedwith TCG-1B TimeCodeGeneratorswith a stabilityof 3 partsini0~over a range of 0—50°C.In 2 y of operation beginning July, 1982 a total of 603 near-source3-componentaccelerogramswere gatheredfrom 243 earthquakesof magnitudeML = 1.2—5.3. Most of these accelerogramshaverecordedtheinitial P-wave.

Theconfigurationof theexperimentalarrayanda representativesetof near-sourcestrong-motionaccelerogramsarepresentedin this paper. The set of accelerogramsexhibitedwereobtainedduring theML = 5.3 Lulong earthquakeofOctober19, 1982, whendigital eventrecordersweretriggered.Theepicentraldistancesrangedfrom 4 to 41 km andthecorrespondingrangeof peakhorizontalaccelerationswas0.232g to 0.009g.A preliminaryanalysisof thedataindicatesthat comparedto motions in the westernUnited States,the peakaccelerationattenuatesmuch more rapidly in theTangshanarea.The scalingof peakaccelerationwith magnitude,however,is similar in thetwo regions.Dataatmoredistantsitesareneededto confirm themorerapidattenuation.

1. Description of array strong-motion array was deployed in the mei-

zoseismalarea of the M~= 7.8 Tangshanearth-A cooperativeresearchprojecton strong-earth- quakeof July 28, 1976. Temporaryinstallationof

quakegroundmotion studiesbetweenthePeople’s instrumentsin an area still experiencingafter-Republic of China and the United States of shockswas expectedto afford the opportunitytoAmericastartedin 1981 (Booreet al., 1982) andis gain operationalfield experiencewith the systemin progressnow. Theoverall goal of this projectis and also yield usefuldata.to increasethe databaseof stronggroundmotion The networkwas installedat an intersectionofof large earthquakes.Of particular interest are several active faulting structures(the Tangshan,near-sourcedatafrom great(M � 8) earthquakes, Guye, Luanxianand Taoyuanfault structures)infor which dataarestill lacking. the northeastpart of the aftershockarea of the

As a part of this project, an experimental Tangshanmainshock.Four digital accelerographs

0031-9201/85/$03.30 © 1985 ElsevierSciencePublishersB.V.

93

g J j :1 i Jo J~ ~ .~ ~

~ZU~E~ ~ VEU~h

~‘Quo 2 2222222 22 2

~° ~, ‘.0 ~.0 ‘. ‘.0 ‘.0 ‘.0 ‘.0

00 N — — — — — — — — — fl —

E0

uI> ~ —

° ‘~ .a.~ ~ a ~eue~u N u

-~ ~ 2~.~E~II ~ ~~ ~ .~ ~

~ F

i~~LL~I ~u’.s ~ ~ ~~ ~

~ ~i ~0 O~

EC0

N Q~ N ‘0 N N N O\ ~ 00 ‘I~ N N \0 ~ ~ ‘0 0 ~ ‘~ N0’. 00 0’ — ‘0 ~ N N N ‘fl 0 ~ (~ — 0’. 00 N N N ‘1~ — C’~~i — N m — m 0 1~ 0 ‘.(‘. ~ 0 I~ N m 0 ‘.4”. ,.~ ~ ~ 0 ~ ~ —‘.4’. ~ 00 1~ ~ ‘4’. 0 4’. 0 ~ ,.‘. ~ .-~ .-‘. ~ m .r’. ~n0 ~ — N N

~N ~N ~ ~“.‘.“. .4’.”..0 0 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

X O’.oO 0’00 0’000’.o00’000’.000’.000s000’.00 0’000’.OO 0’000, m~ m,-4 m~m~m,-’.~’.,.om,-. ~ ~C L) ,_. ,~ — — ,., ,_, — — — ,.., ,., —

‘.5

Q4~ L)00~ -~ -~C c,i C

00 .~ 00 .‘.0, ‘.5 ~0 •~ ti 0 2 2 2 . C 50 50~ ~.~ !~

94

118° 118°30’ 119°

o M1:2.9--3.9 __ Io L~.O__L~.9 LI M~. 5.3

0 City• Station(PDR-1) Qianan

— 0 6 12 18 2Lf Km :.‘ ‘::‘“ . 0 — Lf0

0

“ :1 ~.: ~ .‘ ~, ‘ Lülong— i_I ~ 11

Fen;run / / 12

2 ~ .10,...

/ o0 ~ 6\9 Luanxian Changli

Tangshnn f~ 00 \~•

Luannan o

—~---—/ /0 39°30’

‘.4—Fig. 1. Strong-motionaccelerographexperimentalarrayin Tangshanregionanddistributionof epicenters(1982.7—1983.3).TS-01 Zhaogezhuangseismologicalstation;TS-02 Beijiadianrefractorymaterialfactory;TS-03 Leizhuangarmyoffice;TS-04 Office of oil depotatTuozitou;TS-05 Businessagencyof oil depotatTuozitou;TS-06 Luanxianschoolof CommunistPartyatTuozitou;TS-07 Luanxiannurses’ schoolatTuozitou;TS-08 Xiangtangseismologicalstation;TS-09 Luanxianseismologicaloffice;TS-10 Shimentown office;TS-11 Lulong seismologicaloffice;TS-12 Lulong goodsbureau.

were installed along the Luanxian and Taoyuan with the FBA-13 Force Balance Accelerometerfaults where moderateearthquakeseismicity was (Table II). A metal pinch roller wasusedto allowhigh; eight digital accelerographsweresitedalong recordingat cold temperatures.a line normal to the fault. Four of those eight To reducethe effectsof buildings on the re-instrumentswere concentratedin a small areaof corded ground motion, all the instrumentswereabout 0.3 km2, near the town of Tuozitou, to installedin small, light shelters(Fig. 2). Forinstal-observethe variationof groundmotion in a small lation of instruments,the FBA-13 tn-axial accel-area. Figure 1 shows the primary tectonic struc- erometerpackagewasbolted to a steelplatewhichtures at the array site and the distribution of was thenattachedto the cementfloor with eitherinstruments.A brief descriptionof the stationsites epoxyresinor anchorbars(Fig. 3). Thesemethodsis givenin TableI. of attachmentareconvenientandreliable.

The instrumentsare servicedat 10- to 43-dayintervals.Problemsassociatedwith the effectsof

2. Description of instrument used in array the cold winter temperatureson the batterieshavebeenfound in 57 out of 367 sitevisits. No mecha-

The twelve digital accelerographsinstalledare nical problemsdue to the cold havebeen noted,KinemetricsPDR-1Digital EventRecorders,fitted althoughtemperatureshavedroppedto — 20°C.

95

TABLEII

Specificationsof PDR-1/FBA-13

Numberof channels Selectable2.5, 12.5,25, ________ ______

50 Hz; 12 dB/octroll-off ,,,

Gain-ranging ±36 dB up anddownduringeventDynamicrange 102 dB _____

Resolution 12-bitFormat Phase-encoded4-track;

threedata,oneparity ~ ______

Bit density 1280 bits/inch/trackSamplerate 100 or200 sampless’~’

1 -

perchannelselectable 0Tapespeed 1.25 or 2.Sins~Start-uptime 150 msTrigger STA/LTA selectableratios

ordifferencesPre-eventmemory StandardPEM is 2.56s

with 200 spsor 5 swith 100 sps _________

Temperaturestability •~.

ofTCG-1B ±3X107(0—50°C) ________________________Full scale ±2g - ________

M if ______________________ ________

atura requency _______________________________________

of FBA-13 . 50Hz _______________________

DampingratioofFBA-13 ~i-~12Vand—12VDC ~4”~ ~ -~

Fig. 2. (a) Exterior of a typical station(TS-10).(b) Installationof instruments(PDR-1/113andFBA-13).

3. Near-sourceaccelerogramstamed during 133 earthquakesof magnitudeML

In the 9 months from July, 1982 to March, from 1.2—5.3 (Table III). Two years of recordings1983, a total of 285 accelerogramshavebeenob- haveproduced603 recordsfrom 243 events(25 of

Seal washer

Nut, 1/4-20

\\ Stud,1/4-20x3”

Washer,

II ii I flat

/ FBA-13

I Anchor bar6mm plate 6mm plate IAdhesive

________________ (or anchor bar) Adhesive ... ..,. ...

4 I I Cement floor 4 4.

1 ~ “~ ~ soil ground) ‘ ‘~~‘A~ 0 Floor(cement or soil)

Installation of FBA-13 Section A-A

Fig. 3. Installationof FBA-13.

96

S~ ~ CH. 1 COMP. N-S

0.1G

~ ,

0 5 lOSecW I Cli. 3 COMP. E-W

~0 I~ II~i~II~4?~~ ‘.1 ~ 0’ “ 0,

Et

Fig. 4. Ninedigital accelerographstiggeredduringtheLulong earthquake.Event: 19/10/1982.20:46’(BJT), M1 = 5.3, Depth:9.6 km,

Lat. 39°52.6’(N), Long. 118°55.4’(E).Station: Lulong City Goods Bureau (TS-12), Lat. 39°52’41”(N),Long. 118°52’14”(E).PDR-1/109andFBA-13/16204,05,06(Kinemetrics).

97

which hadmagnitudes> 4.0). Figure 1 shows the Ninedigital accelerographstriggeredduring theepicentersof some earthquakes(ML = 2.9—5.3) Lulong earthquakeof magnitude ML = 5.3 ofduring the period. October19, 1982 are shown in Fig. 4. The epi-

OCT. 19. 1982 STAT. 2

:

~0. 00 4.00 8.00 12.00 18.00 20.00 24.00 28.00 32.00— TIME ~SEC)

OCT. 19, 1982 STAT.2cm

~ ______

~0. 00 4.00 8. CO 12. CO 16. CO 20. CC 24.00 28. CO 32.00cm TIME (SEC)

OCT. 19, 1982 STA1’. 2

~ ~

00 4. 00 8. 00 12. 00 16.00 20. 00 24. 00 28. 00 32. 00cm TIME (SEC.)

98

central distanceof the nearest station (Lulong peakhorizontalaccelerationof 0.232g andverticalstation,No. TS-12) is 4.2 km; the fartheststa- accelerationof 0.128gwererecordedat the nearesttion is 40.5 km from the epicenter.The maximum station. For this earthquake,the epicenterareais

cDOCT. 19. 1982 STAT. 3

CD’ I I I I ~‘‘‘I’~’’’’’’’ I I

~0. CO 4.00 8.00 12.00 18.00 20.00 24.00 28.00 32.00TIME (SEE)

I OCT. 19, 1982 STAT. 3

__ __ ID

ID

U

NJ I T~~I I I 1~C. 00 4. 00 8. 00 12. 00 10. 00 20. 00 24. CO 28. CO 32. 00CD 1 IME HEC)

CD

CD OC1. 19. 1982 STAT.3NJJJ

I~IIIII

~0. 00 4.00 8.00 12.00 16.00 20.00 24.00 28.00 32.00TIME (SEE)

99

at Xiazhai Town where the rural masonrybuild- 4. A preliminary analysisof analogaccelerogramsings sufferedmoderatedamage.

A completeanalysisof the datawashamperedby the lack of adequateplaybackequipment.The

~JJ.00 4. 00 8.00 12. 00 16. 00 20. 00 24 00 28. 00 32. 00TIME (SEE)

0

CD OCT. 19. 1982 STAr. 40.)

II~II~

~p.00 4. 00 8. OlD 12. 00 16. 00 20. 00 24. 00 28. 00 32. 00TIME (SEC)

0 OCT. 19, 1982 STAT. 4

~p.00 4. 00 8.00 12. 00 16.00 20. 30 24. 00 28.00 32. 00TIME (SEC)

100

necessaryequipmenthasnow beenobtained,and resultsobtainedthrougha preliminaryanalysisofa report describingthe analysiswill be published the visualplaybackedrecords.in the future. In this paperwe provide several

C

D OCT. 19. 1982 STAT. 5

~,C.00 4. 00 8. 00 12. 00 16. 00 20. 00 24.03 28.03 32. 00TIME (SEE)

ID

CCT. 19, 1982 STAT. 5~1 L II ~zt I i D

~ 00 4. 00 8. 00 12. 00 16. 00 2C. 00 24. 00 28.00 32. 00NJ TIME (SEE)

C

OCT. 19.1982 STAT.S

~ 00 4. 00 8. 00 12. 00 16. CC 20. 00 24. 00 28. 00 32. 30TIME (SEE)

101

4.1. Determinationof hypocenters natesof the earthquakeshavebeendeterminedbyusing the chord(Ishigawa)method.The locations

From absolutearrival times measuredon the are illustrated in Table IV and Fig. 5, and aaccelerograms,the origin timesandsourcecoordi- comparisonwith the locations provided by the

OCT. 19,1982 STAT.6

r~IIIThI

~f. 00 4. 00 8. DO 12. 00 16. 00 20. 00 24. 00 28.00 32. 00TIME (SEE)

0 OCT. 19, 1982 STAT.6

N.)

~_____ ______00 4. DO 8. 00 12. 00 16. 00 23. 00 24. 00 28. 00 32. 00

TIME (SEE)

ID OCT. 19, 1982 STRT.6

~~I:II

00 4. 00 8. 00 12. 00 16.00 23. DO 24.00 28.00 32. 00F’.) TIME (SEE)

102

Instituteof Geophysics(IGP), StateSeismological It is worth noting that the epicenterwe locatedBureauof China(1977), usingmoredistant data, for the Lulong earthquakeis much closer to theis also given, field investigatedepicenterthan is the IGP epi-

C

OCT. 19, 1982 STAT. 7‘F-

C-)

~

ID I I - I I I‘COO 4.00 8.00 12.00 16.00 20.00 24.00 28.00 32.00

TIME (SEE)

ID

OCT. 19,1982 STAT.7NJ

‘0.00 4.00 8.00 12.00 16.00 20.00 24.00 28.00 32.00TIME (SEE)

ID

OCT. 19, 1982 STAT. 7

ID I I I I I I I —

0. 00 4. 00 8. 00 22. 00 16.00 20. 00 24.00 28, 00 32. 00

TIME ISEC)

103

center.The improvementin the earthquakeloca- standardseismic stations close to large earth-tion obtained by using the strong-motion data quakesare driven offscaleandinformation aboutemphasizesthe utility of digital accelerographs; the arrival time of later seismicphasesis lost.

ID

OCT. 19, 1982 STAT. 9

JJ. 00 4. 00 8.00 12. 00 16. 00 20. 00 24. 00 28. 00 32. 00TIME (SEE)

po OCT. 29. 1982 STAT. 9

C-)

ID F — I I I I~0. 00 4. 00 8.00 22. OC 16. 00 20. 00 24. 00 28. 00 32. 00

TIME (SEE)

ID

OCT. 19.1982 STAT.9

.~H~IIIII~II

~j3. 00 4. 00 8. 00 12.00 16.00 20.00 24. 00 28.00 32.00TiME (SEE)

104

TABLE III

Numberof EarthquakeRecordsin TangshanArray(1982.7—1983.3)

MagnitudeM, 1—1.9 2—2.9 3—3.9 4—4.9 5.3 TotalNo. of events 3 102 16 11 1 133No. of records 3 183 51 39 9 285

In addition to the hypocenterdetermination, parameters.a is the standarddeviation of thethe P- and S-wavevelocitiesin the areaare calcu- logarithmof a predictedvalue.Ninety-threemea-latedas 5.58 km ~_1 and3.23 km ~ respectively, suredhorizontal accelerations(larger of two corn-from P- and S-wavearrival times on the Lulong ponents)from 19 earthquakesranging in magni-earthquakeaccelerograms(Table V). tude from 2.9 to 5.3 were used in the regression

analysis;for the vertical acceleration,87 measure-

4.2. TheattenuationofpeakgroundaccelerationAm mentsfrom 19 eventswereused.Events recordedon only one stationwereexcludedfrom the analy-

A two-stage regressionanalysis (Joyner and sis.Boore,1981) hasbeenused to fit the peakacceler- The magnitudesused were provided by theationsto the following regressionmodel Instituteof Geophysics(IGP), as determinedfrom

A = a~10(/~M+~~± (1) the network of instrumentsin the Beijing-Tang-m shanarea. The IGP usesRichter’s correction for

where distance(Richter,1958) and, for earthquakeswithR — + h

2 M < 3.2, the peak motion from once-integrated— ‘I 1-Hz velocity transducers.For largerearthquakes

and M is the local magnitudeand ~ is the epi- which saturate the 1-Hz instruments,measure-central distance. a, /3, y, and h are regression ments are madeon seismogramsobtainedon a

118° 118°30’ 119°

o M1:’2.9--3.9 I ... ‘....

o Ls.O——k.9 Q M1: 5.3• Station(PDR—1) 0 1 ~‘ Qianan

— 06 12 18 2k Km .: ‘~,‘ . —

(IGP)

/ / ~

‘ 12 (BSMOC)Fengrun ‘‘‘‘ : / I :

///~ ~ Lua~xian ChL1I

Tangshan , \\~

Luannan o

— — —/ / o ~— 39°30’

Fig. 5. Comparisonof earthquakelocationsobtainedfrom IGP and BSMOC/IEM.

105

TABLE IV

Comparisonof earthquakeparametersbetweenIGP andBSMOC

Data Timeof earthquake Location Depth Magrn.source (Date,h:min’s”) (km) M

1E N

IGP 82.07.13,05:42’29.9” 118°42’ 39°46’ 2.9

BSMOC 05:42’30.0” 118°45.8’ 39°44.7’ 14.7IGP 82.07.17,12:42’55.9” 118°38’ 39045~ 44BSMOC 12:42’56.7” 118°36.0’ 39°44.7’ 8.4IGP 82.07.25,10:10’59.4” 118°47’ 39047~ 3.1BSMOC 10:10’59.3” 118°47.7’ 39.48.4’ 8.4IGP 82.10.19,20:45’59.2” 119°0.4’ 39057~ 53 *

BSMOC 20:46’00.3” 118°55.4’ 39°52.6’ 9.6IGP 82.10.25,17:44’35.3” 118°26’ 39043~ 43BSMOC 17:44’35.5” 118°26.5’ 39°44.9’ 12.0IGP 82.11.21,01:01’36.5” 118°28’ 39043~ 4.2

BSMOC 01:01’36.5” 118°28.8’ 39°45.5’ 12.9IGP 82.12.30,22:34’46.0” 118°22’ 39°42’ 4.4BSMOC 22:34’46.5” 118°26.3’ 39045,9~ 12.0IGP 83.03.04,04:03’20.5” 118°32’ 39047~ 44BSMOC 04:03’20.9” 118°31.3’ 39°46.9’ 7.7

IGP 83.03.08,08:35’19.3” 118°30’ 39°43’ 4.0BSMOC 08:35’19.3” 118°28.1’ 39°46.4’ 9.9

Notes: * This valueof magnitude is from Hebei SeismologicalBureau. IGP—Institute of Geophysics/SSB. BSMOC—Beijing StrongMotion ObservationCenter/IEM.

lower-gain12sinstrumentlocatedin the basement however,reportedmagnitudesgreaterthan3.2haveof the IGP recordingbuilding (in thesemeasure- been reducedby 0.5. Not doing so leads to aments, the longer period surfacewave trains are severebias in the results.The only magnitudenotignored, even if they representthe peakmotion). correctedis that for the Lulong earthquake.TheThis recordingsitehasa pronouncedamplification IGP reported M = 6.2; the value of 5.3 was ob-factor, however,correspondingon the averageto tamedfrom a 12sinstrumentat anotherobserva-+0.5 magnitudeunits. Becausethis factor is still tion site in the Beijing areanot used in routinebeing studied,it is not appliedto the magnitudes analysis.reportedby the IGP (including thoselisted in the The analysis yielded the following empiricaltablesin this paper).For the regressionanalysis, equations(Fig. 6a, b) for peakaccelerationin g,

TABLE V

Dataof V~,V~and from Lulong earthquakeof Oct. 19th, 1982

No. of station p travel-time p wavevelocity s travel-time swavevelocity(s) (km s_i) (s) (kin s_I)

TS-04 5.070 5.550 8.710 3.231TS-06 5.267 5.635 9.207 3.223TS-07 4.868 5.620 8.408 3.254TS-09 4.391 5.518 7.571 3.200TS-10 3.805 5.566 6.585 3.216

Average 5.578 3.225Standard deviation a 0.049 0.020

Velocity of virtual wave V~,= 7.645 (km s —

106

0- -

a 40 Average ratio 0.682

C)I Samples 88

a ...tandard deviation 0.292-1 + F..

± ± ~30±+ ++ C)

+ 0ID + 20

a: ±C) -2 ±± ±+ 10

ID0

3 ______ _______A5(V)

0 ______________________________________________ A!I~(H)

0 1.0 2.0

A- Fig.7. Ratiobetweenvertical andhorizontal maximum acceler-

0 1 2 ation.LOOt EP. 01ST.

Fig. 6. (A) Data from eventswith more than 1 recordhorizon-

tal data,reducedto M = 4.These equations are preliminary; more refined

0- equationsawait further studiesof the magnitude

anddistancesused in the analysis.In an analysis of strong-motion data from

- 1 earthquakesin westernNorth America,Joyneret

— + al. (1981) found— ±± logCD ~ + +a: ++ +* ±ID —2a- ± —1.97 + O.

41ML — log R — 0.0026R±0.20 (4a)++

— +ID

÷-“- g-+ h9.2 (4b)0-J

for 6.4� ML � 7.2. Themagnitudescalingfactor ishigher than found from the Tangshan-areadata,but the simulationmodel of Boore(1983) predic-

B-14 ted that the magnitudescaling factor for large

0 1 2 earthquakesshouldbe — 0.1 units larger than forLOOt EP. 01ST. 1

earthquakeswith 3 <ML < 5. For this reason,theFig. 6. (B) Data from eventswith morethan 1 record vertical magnitude scaling factor found from the Tang-data, reducedto M = 4. shan-areadata is in agreementwith that found

from westernNorth Americandata(assumingthatthe magnitudeM in eqns.2 and3 is equivalentto

with distancesin km ML). The attenuationfactor,however,is ten times

Horizontal: log Am(H) = larger for the Tangshanregion. Whether this at-tenuationfactor is ruly representativeof the actual

—1.49 + 0.31M— log R — 0.0248R±0.32 (2a) attenuationin the regionwill require the analysisof datafrom greaterdistancesandan investigation

h=9.4 (2b) into possiblebiasesintroducedby the distributionVertical: log Am ( V) = of data in magnitude-distancespace (as well as

improvedmagnitudeanddistanceestimates).—1.92+ 0.29M— log R — 0.0146R±0.36 (3a) In Fig. 7, a distribution of the ratio Am(V)/Am

h = 6.7 (3b) (H) of peakvertical accelerationto horizontalis

107

shown and the mean ratio is 0.682 (88 data are at Tuozitou Town. The areasurroundedby thoseused for statisticalevaluation) and the standard four instrumentsis about 0.3 km2. Although thedeviationa is 0.292. groundhere is flat, upon comparingthe recorded

peakaccelerationsat four stations,a distinct dif-4.3. Thedifferenceof groundmotionin a smallarea ferencein both peakvalue of the samedirection

andorientationof groundmotionrecordedat eachIn the experimental array, 4 PDR-1 digital station is found for the same event. The biggest

accelerographs(No. TS-04—TS-07)were installed differencebetweenpeakaccelerationsin the same

UCI) IO”3GN(I) $O”3G 0 / \\ 0 - -30--30 /

0 0- -20 0 -~Oo / 0 - ~08 8 N

jO 3r 30 2f~ 0 0N’ of SO UC.D o 0

0 0 / 0 0 000 % / 0 0 00 0 -

°o/ 000 0/ 8o\ 0

-20 0 o 0 0 / -20 00 /

00 COMP U-DN-S SW -30 DCI) _______________

- -30

0 ED I~3G / REMARK--30 / _____ ______~ 0 I NO OF STATION0 /

0 0 - -20 - ‘‘c” C~°1BI- AXIS 6 4 ~ 6

0 NATION jx AXIS 5 6 I 7 I 6 I I0

NOOFDATA 115 14128111113112

- 0 0 I _________08 0 0

N —~

EU) ~ wLI)c~P~00 0 0 ~30 20 ~“q~~(0 110 20 31D ~ZITOU~~~

6S[

H~P.EW ____ ______ 5 __0 ~ _____________________

Fig. 8. Thedifferenceof accelerationin a local area.

108

TABLE VI

Variation of peakgroundaccelerationover a small area

Timeof events Magnitude Average Component PeakaccelerationAm (10~G) Average Standard a/A,,,(Date,h—min—s) M

1 epicentral TS-04 TS-05 TS-06 TS-07 peak deviationdistance acc. a (10~G)~ (kin) Am

1982.07.17 N—S 20.0 26.0 29.0 33.0 27.0 5.5 0.20312-42-56.7 4.4 7.1 U—D 19.0 20.0 29.0 22.0 22.5 4.5 0.200

E—W 30.0 32.0 30.0 24.0 29.0 3.5 0.1191982.07.17 N—S 13.0 10.0 10.0 18.0 12.8 3.8 0.29612-43-19.2 3.7 7.1 U—D 8.0 10.0 11.0 8.0 9.3 1.5 0.162

E—W 14.0 13.0 35.0 19.0 20.3 10.2 0.5031982.07.25 N—S 13.0 12.0 22.8 15.0 15.7 4.9 0.31210-10-59.3 3.1 11.9 U—D 19.0 21.0 15.6 17.7 18.3 2.3 0.124

E—W 13.0 6.6 11.6 11.0 10.6 2.8 0.2621982.08.09 N—S 9.6 7.1 6.5 12.5 8.9 2.7 0.30601-23-41.7 2.3 2.0 U—D 4.8 9.0 5.4 4.4 5.9 2.1 0.357

E—W 7.9 11.2 6.9 14.7 10.2 3.5 0.3471982.08.31 N—S 8.8 24.0 18.0 32.0 20.7 9.8 0.47319-58-40.9 2.5 7.7 U—D 8.8 12.0 13.7 12.5 11.8 2.1 0.178

E—W 14.4 16.0 12.4 13.5 14.1 1.5 0.1081982.09.14 N—S 8.1 10.0 13.7 16.3 12.0 3.7 0.30612-42-09.9 2.4 8.4 U—D 7.0 10.0 10.6 9.4 9.3 1.6 0.171

E—W 13.8 21.0 19.0 23.8 19.4 4.2 0.218

1982.10.19 N—S 19.0 29.0 23.0 23.0 23.5 4.1 0.17520-46-00.3 5.3 24.9 U—D 15.0 14.0 13.0 11.0 13.3 1.7 0.129

E—W 28.0 20.0 16.0 19.0 20.8 5.1 0.247

1982.11.30 N—S 8.5 12.0 13.0 6.0 9.9 3.2 0.32622-53-48.4 2.7 14.4 U—D 7.8 6.3 10.5 4.1 7.2 2.7 0.375

E—W 11.0 10.2 11.5 11.0 10.9 0.5 0.049

direction at four stationsis sometimesas greatas tions at severalpointsin a small areaas a measure200—300%(Fig. 8). To illustrate this phenomenon, of the intensityof strongground motion.a schemefor suchcomparisonis shownin Fig. 8.In 8a, b, the x and y coordinates,respectively,representthe peakhorizontal accelerationin the Acknowledgmentssamedirection for the sameearthquakerecorded The writerswishto acknowledgethe supportofat any two different stations.Figure 8c showsthe the StateSeismologicalBureauof Chinaand U.S.samerelation for the peakvertical accelerations. NationalScienceFoundation.ThewritersalsowishThe spatial variation in peakmotionsis given by to expresstheir deepestthanksto An Yanru, Dingthe scatter of the points about the two straight Feng,Du Meiqi, Du Ming, Ren Zhengyui~,Wanglineswith slopesof ±1. Tiehua, Wu Weilian, Yan Jianwu, Yu Shuangjiu,

Additionally, thereare eightearthquakesduring Yu Shuqin and ZhaoShufangfor their c~ntribu-which all the four stationsobtainedthe records. tions of operating the TangshanExperimentalThe meanvalueAm andastandarddeviationaof Strong-Motion Array and processingthe data.thosepeak accelerationsand the ratio a/Am are Information regarding magnitude deteripinationlisted in TableVI. Theresultsverify thatconsider- was obtainedthrough discussionswith Professorabledifferencescan exist in peakgroundaccelera- Qin Xinling and Mr. Fu Yu of the Institute oftions in a small area. For this reason,it seems Geophysics.Art McGarr of the U.S.G.S.providedreasonableto considerthe meanof peakaccelera- a critical review of the paper.

109

References Analysis of Near-Earthquake.EarthquakePress,Beijing.(inChinese).

Boore, D.M., 1983. Stochasticsimulation of high-frequency Joyner,W.B. and Boore,D.M., 1981. Peakhorizontalaccelera-groundmotionsbasedon seismologicalmodelsof the radia- tion and velocity from strong-motion records includingtedspectra.Bull. Seismol.Soc. Am., 73: 1865—1894. recordsfrom the 1979 Imperial Valley, California, earth-

Boore, D.M., Hsieh, L.L., Iwan, W.D., Peng,K.Z. and Teng, quake.Bull. Seismol.Soc. Am., 71: 2011—2038.T.L., 1982. U.S.A.—P.R.C. cooperativeproject on strong Joyner, W.B., Boore, D.M. and Porcella, R.L., 1981. Peakground motion. Proceedings of U.S.—P.R.C. Bilateral horizontal accelerationand velocity from strong-motionWorkshopon EarthquakeEngineering,vol. 1, A 1, Harbin, records. Earthquake Notes, 52: 80—81 (Abstract).China. Richter, CF., 1958. ElementarySeismology.W.H. Freeman

Institute of Geophysicsof State SeismologicalBureau, 1977. and Co., SanFrancisco,Chap.22: 338—343.