7 .

@CE0 PHYSICAL STUDIES IN SUPPORT OF SEISMIC HAZARDS ASSESSMENT OF

( SEATTLE AND OLYMPIA, VASHINGTON.

by

K. King, A.Tarr, D. Carver, R.Villiams and D.VorleyU. S. Geological Survey

INTRODUCTION

According to plate tectonics theories, the Juan de Fuca oceanic plateand the continental North American plate are converging at about 3 4 cm/yr ina subduction zone more or less parallel to the Pacific Northwest coast. Queconsequence of this convergence is the occurrence of earthquakes, activevolcanism, and tectonic deformation. The potential exists in the Puget Soundarea for (1) large, shallow earthquakes along the interface of the underthrustzone, (2) major earthquakes within the cold,. brittle Juan da Fuca slab, (3)moderate earthquakes associated with active volcanism in the' Cascade Range, i

and (4) major shallow earthquakes within the North American plate landward of {the underthrust zone. The Pacific Northwest cities of Seattle, Tacoma, and!

Olympia, Washington and Portland, Oregon are urban centers with significantrisk from the occurrence of large earthquakes.

Major earthquakes occurred in the Puget Sound area in 1946, 1949, and1965 (Fig. 1). The 1949 shock caused major damage to high rise structures inOlympia; highest intensities were VII. The 1965 shock caused widespreaddamage in both Seattle and Tacoma, and intensity VII affects in Olympia; the '

highest intensity effects (VIII) were observed in West Seatcle and Harbor I( Island. '

t

URBAN RAZARDS INVESTICATION

!The USGS is engaged in a regional earthquake hazards assessment program

in the States of Washington and Oregon, concentrating on the Puget Sound andPortland urban areas. The program is a partnership among governmental(Federal, State, and local), academic, and private entities to study how theNorthwest would be impacted if a lar5e, potentially dama6tng earthquake wereto occur in the region. The program is the outgrowth of two earthquakehazards workshops (USGS, 1983; USGS, 1986) and is divided into five,interrelated components:

(1) Information Systems(2) Synthesis of geological and geophysical data(3) Cround motion modeling(4) Loss estimation models(5) ImplementationThe studies described in this report were conducted by the Urban Hazards

Field Investigations project in support of the ground motion modslingcomponent of the urban hazards program. The objective of the ground motionmodeling component is to produce deterministic and probabilistic ground motionmodels and to produce maps of ground shaking hazard. One element of theground motion modeling component is predicting relative ground response tostrong vibratory motion from a model earthquake. Relative ground response is

' determined from observations of ground motion a..d from extrapolations of those

kmeasurements into areas where ground motion data are not available. The

,

G808010130 8G061'5 ~PDR ADOCK 05C00344P PDO

, .

.

geophysical studies described in this report are designed to provide seismicdata from which relative ground response values are determined and to collectgeotechnical data which assist in making the extrapolations.

OBJECTIVES

The o ;ectives of the Urban Hazards Field Investigations project are to:(1) Directly record, in digital form, seismogr.as of actual vibratcry

ground motion at sites in urban areas where ground response data are desired.(2) Collect; geotechnical (geological and engineering) data from sites

near where the ground motion measurements were made.(3) Correlate geotechnical data with seismic response data by clustering

sites of similar geotechnical parameters.

METHODS

Relative ground response is determined by comparing the ground responseat a site with a standard or reference response site. In this study, relativeground response is obtained by dividing the Fourier amplituds spectrum of asite by the amplitude spectrum of the reference location. The resultingspectral ratio may then be smoothed or averaged over any number of bandwidths;the average spectral ratio in the value of relative ground response withinthat band. In the relative ground response method, it is assumed that thegrcund response is due to ground conditions only, that is, the seismic inputsto the crust under the response site and reference site are essentiallyidentical for a specified event and that any changes in response are due todifference between reference and response sites. If it is possible tocorrelate those difference in terms of physical parameters and geologicaldescription characterizing the sites, it may be possible to predict groundresponse from geotechnical data in locations where seismic observations arenot available.

Desirable seismic waveforms for the study may occasionally be masked by i

seismic noise of natural and manmade origin. Ic is important to know thecharacteristics of noise to minimize the effects of contaminating a desirablesignal and to help identify the frequencies of interest for site responsestudies.

Recordings of vibratory ground motion and geotechnical data of specificsites are the principal kinds of data which are required for predictive ground

,

response studies. Ground motions resulting from various seismic sources were(recorded by portable digital seismic systems for this study. The seismic data l

used in the studies described in this report were recorded by calibratedportable digital seismic systems especially for the urban hazards program.The sources of seismic energy were both natural (microcarthquakes andmicroseisms) and artificial (nuclear explosions and mining explosions).

Digital recordings of induced ground vibratiens in the Seattle andOlympia areas were successfully acquired for four nuclear explosions at theNevada Test Site (approximately 1,100 km distance) and seven mining explosionsat an open pit coal mine near Centralia, Washington (about 80 km from Seattle)(Figs. 1,2,and 3). The general procedure for recording these explosions was tomanually start all recorders within 15 min of the expected arrival time of theseismic waves and record up to one hour for each,

Five to eight seismic stations were installed at temporary locations inthe Vest Seattle Area. The recorders at the stations continuously store

i

1( approximately 15 seconds of data in a digital memory and will permanently

2

!i

_ _ _ _ _ _ _ _ _ _ - - _ - _ _ _ - - _ _ - _ - - - - - - - - i

_

.

store the digital data for longer duration when activated by a radio signal(~ from a master station or by a local vibratory ground motion which coincides

with a pre-set algrorithm. Two microcarthquakes have been recorded to date.The time history data for one of the earthquakes and for one of the nuclearevents are shown in Figs. 4,end 5.

Three separate experiments were conducted to evaluate the nature ofseismic noise and microseisms in Seattle. The first experiment consisted ofrecording 10 min of background vibrations at five sites at three times (at 1a.m., 2 a.m., and 3 a.m.) on two successive dsys. The second experimentconsisted of recording 15 min of background noise at five stations during aSaturday afternoon in the Brighton district of Saattle. Two of the sites werelocated on bedrock while the other three were located on varying thickness ofsediment. The third experiment consisted of record bg 15 min of backgroundnoise at five stations on a windy, rainy Sunday morning in Vest Seattle,extending inland along a line about a mile long.

Seismic refraction and high resolution reflection are geophysical anthodsused to determine the subsurface structural details at a site, based on theacoustic contrast across interfaces. The two methods are complementary:Seismic refraction is most accurate in determining average velocities oflayers whereas seismic re.tlection is most accurate in determining layerthickness. In this study, seismic refraction experiments were run todetermine layer velocities and approximate layer thickness at several responsosites. Seismic reflection experiments were run, using an approximate velocitymodel determined from refraction, to determine more accurate thicknesa andscructure of the shallow layers. In addition, longer reflection lines wererun to detect deep interfaces which were not accessible to the refractionexperiments.

( In all cases, refraction lines were run in both forward and reversedirections, and the reflection methods used the "push pull" technique to givea minimum 18-fold summsry of the common depth points. Several data reductiontechniques were used to derive the velocity model and depths frors the travel-time data. Fourteen refraction / reflection lines were run in the Seattle area(Fig.2). Examples of the high resolution reflection profiles are shown inFig.6.

The method used to determine building response parameters is toartificially force the structure into oscillation and to record the vibrationusing a seismograph. An impulse delivered to the structure produces a damped j

vibration whose waveform allows a damping constant to be measured. i

A spectrum !

of the waveform indicates the predominant resonant period of the structure.Ten one story single family dwellings, with brick chimneys in the West Seat::learea were tested for building response. In all cases, both the dwelling andthe chimney were tested (Fig.7).

SUMMARY

Only preliminary conclusions can be derived from the present data set. Thecomparison of the derived spectra from the ground motions induced by theNevada Test Site nuclear explosions and the spectra derived from the groundmotions induced by small earthquakes suggest that the spectra derived from thelarge nuclear event ground motions are comparable and therefore useful atfrequencies less than approximately 1.5 Hz. i

The ground motions at Seattle and Portland induced by the quarry and mineblasts at the Centralia, Washington coal area are too small to be used for

.( site response studies. The study has shown that ground motions induced by

3

j

-

.

- local earthquakes are the only technically acceptable source at this time for(- ground motion studies in these areas. The ground motions induced by the quarry

blasts at Centralia are less than desirable, but are adequate for siteresponse studies in Olympia, Washington.

The derived site response values in the Olympia, Washington area from themotions indu:ed by the Centralia quarry blasts cerrespond favorably with.theMM intensities from the 1965 earthquake; that is, the higher response sitesare located at areas of higher intensities, the medium response values are atsites of medium intensities and low response values are located at areas where !

ro intensities or damage was reported Fig.9. The few site response values.

derived thus far for the sites in the West Seattle area from the motionsinduced by the nuclear blasts and local earthquakes do not seem to agrec aswell with the 1965 MM intensity values as the Olympia data do except in a verygeneral sense; that is, the highest response value derived for the WestSeattle area from the limite,' data available is at Harbor Island whichexperienced higher shaking damage in the 1965 earthquake than did the West

* Seattle Area Figs.10,and 11..;

In many cases, the noise spectrum at a response site was amplified abovethe corresponding reference spectrum. Prominent peaks were apparent at :

'

several sites and remained prominent at different times of day and en,

;

different days. These results suggest that microseisms are a possiMe fourth| source of seismic energy for ground response measurements.i

Refraction and reflection lines were run at sites to help determine 1ccali

,

subsurface geologic structure, to determine near surfacs variability of !seismic velocities, and to establish the velocity of bedrock. The highestvelocity ebserved, about 8,500 ft/sec, was also the velocity of the exposedbedrock unit at Seward Park. The velocity of the surface soil layer was less

( than 1000 ft/see and the velocity of the intermediate till 1syers ranged from'

2,400 to 5,100 ft/sec.(Fig.6).The low rise building testing established the range of the period and

damping paramet.ers of one story houses and chimineys in the West Seattle area.The predotinant frequency of the dweilings ranged from 5.4 Hz to 14.8 Hz. andthe chimneys ranged from 6.2 Hz. to 13.7 Hz. The building dampings varied from2.54 to a high of 64 of critical,

i The microseismic data, reflection data, refraction data and past intensitydata all suggest interesting correlations to the site respons,e values;however, the amount of site response data and the number of sites under studyare too small to make any but preliminary conclusions or trends. A basicconclusion is that a larger dar.a set from ground motions induced byearthquakes are needed to continue and complete the study.

4

f

1

ii!

!

s) 4

h

. - - . - , _ ~ - - , . - - - - - - . - . - - . - . - - . - - - . - ~ - - _ ~ - _ . - . - - _ - - - - - - - - - - - , - - - - , . . - - -

1 ,

| j,

==

( Y.

/ '

Seante

,

I*,

i O

%n ' ...:

. . . . . . .

n y

I Au m ae

{e

l. c.w~

\WASHINGTON

'. '

i

l

||

A.Aw$~~ w.

I

1

OREGON

c ..,

| L.bn.

s#- ..

i

i f. . i ) Portland...

/1

:~.; . x = . .

|) '

)Location map showingk1946,1949 and 1965 earthquake locations;( Figure 1.

j $recentlocalmicrocarthquakes;andenergysourcelocationc.1 ,

1

i

I. - - - _ . . _ _ . _ - _ _ . _ . . . -_._

'

\ \

N ID b>

V Ja

n '

1N,

p)) i

I c ' A " '- '" \ i-

O 3 M .. ... 3 A,

LakeWashington i

Elliot Bay I

f f I' 190'

i. ,

A! - .I4Ag " '|.

\s '.I .,

'

N ('.,

1-

I

e i eu n. .

Figure 2. l.ocations of reflection / refraction lines = m ; and siteresponse status . .

I

, _ _ _ - _ _ _ ___ - _ - . _ _ _ _ , - _ . . _ _ - - - - . - - - - - - - _ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ---

.. . . _ _

-,:-

yt - - - - .,

l |

(~

i I

Io

l* |,- BUDDI ~ ~ ~ ~ ~ - - ,

, l|

.,. ^

'.

,1NLET'

Jr llI .

~i

M g|i t*ii

| I(. \'*I i

I *I l

|

|*' ' ' e G lile ' ***

!-~t---- '~

% l - =r e e_a m >

|~l

OLYMPIA|

I

'- - ~;1 '~' 7

|i

| '& lol y

t * Ie l ;

| : , l. i

.| ,J I

t e Io

'

I~ -ame ,,?. .I ,

--. ,.

n--.- ,-, = ,~,-I 1

I

O * &I i'

k il \

R o*c k' 'I \ I\

I 's ~~

| |I \' I o-,

9 Mlle| 1 l----, I

<

,

!MWATER ,' ' '-

Tigure 3, to s of refl'etion/ refraction 11"'8 * W i and site( response status

__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .

. i i i . . i , i i i i

SEWARD PARK (ROCK)HO RIZ. (N- S ), h4),":Mrj, % 9 * W , w w # : = .--v. + :: -_

, v -:= - - - -

HORIZ (E-W)-+g,w & A9 p v=w;=Q==:= :.. - 3 - - = = ^ -x

BRIGHTON PLYFLDHORIZ. (N-S)

-- j:y;$y),hwQ>&=:f W, u =h%m = :: , ',':-=;.e, : =r. : :

7 7

" "'*- ("%wam:a:= = w = . -, -

,

OW GENESSEE PLYFLDb HORIZ, (N-S)j mm , w1 'i? e'/,hh'$MAf thMk %d'/MwW'M.WAW4bseO HORIZ (E-W)3 - = =ccW.Q%h ' g ),g f,q q 6 , % eg GENESSEE HILL SCHOOL

I i}| | A'

I( I

U HORIZ. (E-W)> "A 3W-IF.h

.

IMfhk|thhMMWehhf4Mft'/pW$>-O HlAWATHA PLYFLDO HORIZ. (N i

g +w t4 g % ,' t M'p H W,1 % + p * y e w m *>

H ORIZ, (E-W)

4^-W'WMNgh' e' h, i % %' h . M M dih,k19$ % W,

W. SEA TTLE FIRE STAHORIZ. (N S)| ;

hl j h b M d f W M / M N 'I*'i i

HORIZ. (E-W)

TNY|Y M///kh|h.W$f.%hgkfr'6)'hMWNN."e'

-- - -i

i l i i f I t i ___1 i t t |

0 32 64 96 128 160 192*

SECONDSiFigure 4. Time M 4 tory of ground motions recorded in West Seattle.{( Induced source is a nuclear explosion at the Nevada Test Site.

i i i i i i ie i i i i ...iie i i i i i

SEVARD PARK (R0rk)

( d 'IZ '" 5'iagggy.,.. .

warz. ic w_ x , , m,,,,, ,, , . _

__

.mi. . , . - --

'

-

b] CENTRAL VEST SEATTLE<

'

_L | | | |gs ,3.., . , _ n , ,p_. _ , .; _.; _g _,,_ ,,

I-

BeEU *

[ uvccc*

f,I !, N t M q l,'g, m i m: . cP, 1 fl 6,la; ;c n( t i I

U

$g HARBOR ISLAND

if,f,/f %y[l/ih'Sk$p,%\l.W.AMMwg4y*yh !

ki)

i

I h fgi' l I | |

Mj y

i t i i i l i i i i f f i t 1 i i f i _1 I i i i0 4 8 12 16 20 24 28 32 36 40 44 49

SECONDSFigure 5. 712ne history of ground motions recorded in k'est seattle.Induced source is a local microaarthquake.

J

(,

|

|

ALKl PARKI SOUTH NORTH

''

........ . . . . . . . ........

"""4(6 h,7,-.e%.i=-----m,w,,.e.d,,M,..,g~g,pn.,tBIt 4- r - P2 i s-

a < e,a,a > e~x,. ---- pe, .

I , , u n , , o , ; n . n i n n n o u n , u n n . m . 2 0_0 f t , , , o , , , , o m o . . m o . , o

4. jJ . . . . . . . ., Q(! @@,*,34n,a $, w$ - ~ ~ ~"-g P7,, v-------

ym 4. . b Q\ {,,4. .. y pa . g, e-- -

........ . . . . . .. . . . .,

t coi i-,u.,m

HUTCHINSON PARKI SOUTH NORTH,

l I| | {

'

...... . ... ... . . . . . . . . . . .. .. .. . . . . . . . . .. . . . . . . . . . . . . .

hh[9jiWiqphidh'.5$$~I$b5NE:$$2F N ....-!$-- --- W'Q.i@)p@[EhgAg;g,pW+'sip:n

4 il# m,5* "O "M. .. . 20 0.

..

[. ... ......,

N m n o n ! n o n i n o n n o n f i n fii n o .d0.0 f t o o , in h o i n o n n o u n n o n o n -tooj

i

Pisure 6.--Shows :he 24-fold high-resolut d on seismic-reflection stackedprofiles from Alki Park, West Seattle (top), and Hutchinson Park,a

'

southeast Seattle (bottom). Dotted reflection on both profiles! indicates probable top of the 8500 ft/s bedrock seen in the refrac-

tion records at Seward Park. Reflection data were collected usinga 4 f t geophone interval (2 f t CDP interval), single 100 Ha geo-

. (' phones, 220 Ha low-cut recording filters, and a 30-06 rifle for the'

seismic source.

.

. - - _ _ _ .-- - -._.-- . ,-- , - - - - . - - - - - - - - - - - - - - - - - - - - - ~ ~ - - - - ~ - - - - - - - - - - - ~ ~ ' ~ - - ~

1

,

BLDG. 03-CHIMNEY BLDG. 04-CHIMNEY.

WW.^:: W ::::= %, - -.

O 1 2 3 4 5 0 1 2 3 4 6SECONOS SECONDS

8.8,

No

1 .0>

* \,

5 he=

A1 4 4 12 18 20 1 4 8 12 14 20 26

FREOVENCY (HD FREQUENCY (HD

BLDG. 03-1 STORY BLDG. 04-1 STORY~'

||OU[#W**~~~4P^~~+|WWI

1 2 3 4 6 0 1 2 3 4 6SECONDSSECOtOS

5.4 7.0

mGW. .

. ,

.E .

E i!

$ '

1

1 4 8 13 18 20 1 4 8 12 18 20 28FREQtENCY (HD FREQUENCY (HD

Tigure 7. k'est Seattle building response studies.

_. . _ _ _ _ _ _ _ _ _

Nr_ ----'i-

i\ |

i,

>

IBUDD i,_ i________,1

i t

| flNLET #

QJpii

V 1.8 i7 !'

! 7>,

l'-V 2.8 2.3' Yll 5.2 . 2.3 i

ii- , e v 'iI

~~t.,--, i

V '' .' '

o-.) M' 3 y il s.5 4.8 '.

DOt.YMP,lA |

,

ii_ _ _ i

| i--Vil 4.3 4.1 i

i a 'tol gt ,i

' o 8k, '$ Vll 2.8 4.8I '

ei ia| i

'g i->_,

---

}____ ',O ,-- r

' , 1

's 'O fi

,i_ ss iflo c k | %

g .d;

1 ||-., wi. i 3 |-----| '

,

i li TUMWATER ' I,J 8 ( - -

Tigure 9. Stations of ground motion site response studies. Box !is >bdified Mercalli Intensity. Boxes 2 and 3 are spectral ratios

1of 0.5-1 Hz and 1-2 Hz, respectively.|

( )i|

|

|1

- -

.

's

F

.

,

SEATTLE (.

nI

j { Lake'

,

! Washington! ElLiot Bay

J| . vh.84 (i

-

<9 x- e<

&\'

_

,} vil li t 6.7 5.9 |

k7j vI 3.6 % j1.3*

N/ .IV 3.1

3.5'

k) J.

(1o 1 2 3.1,, ,

) [ Miles N \ g

Figure 10. Stations of ground motion site response studies. Box I is Modified Mercalli Intens1*;.Boxer 2, 3 and 4 are Spectral Ratios of 0.5 to 2 IIz, 2 to 6 hz, and 6 to 10 liz, respectively. Signalsource is earthquake Q-3.

i

l

I_ _ _ . . - _ _ . _ _ . . . - _ . . - -

. - . . . - - - - . . - ~ _ - - - - _ - - - _ - - - - - , , - . . . , . - . _ . . - . - . _ . . . _- . --.- - ..- . ---

A ^ ,o

CENTER VEST SEATTLE HARBOR ISLAND

HORIZONTAL COMPONENTS HORIZONTAL COMPONENTS*\ 14.1

_

| 30 - -~'~

-4 3,3 .-

|. 78

^ I,

a.3;_ ** i u

: 3.,4.8. : ;_

| 8_ 3.s .

'

o u 22:2 : =: :"3 : :

a.. 3.s

.

ur : ts :< - : ue 3,0 - '

--

~ .

e-- .

,<>x

>

10 - - -

0.5 2 4 6 8 to 0.5 2 4 3 3 30

FREQUENCY (Hz) FREQUENCY (Hz)

Figure 11. West Seattle spectral' ratios from local microcarthquake Q-3.

!

__ . _ _ _ _ _ _ _ - ~ _ _ _ _ - - _ _ _ . - . _ _ _ . _ - _ _ _ _ _ - _ _ _ _ _ - - _ _ _ . - . - _ _ . _ _ _ _ - - _ - _ . - - _ - . - - - - - - _ _ . , _ - _ _ . - - , _ . _ _ _.-___ _-

_