Ground Motion Values for Use in the Seismic Design of the Trans-Alaska Pipeline System

Home , Peak ground acceleration

By Robert A. Page, David M. Boore, William 8. Joyner, and Henry W. Coulter

------

GEOLOGICAL SURVEY CIRCULAR 672

Washington 1972

I United States Department of the Interior ROGERS C. 8. MORTON, Secretary

Geological Survey V. E. McKelvey, Director

Free on application to the U.S. Geological Survey, Washington, D.C. 20242

I I CONTENTS

Page Abstract ...... I Introduction ...... I Seismic potential ...... 2 Design approach ...... 2 Ground motion values ...... 3 Acceleration ...... 4 Velocity ...... 9 Displacement ...... 10 Duration ...... I I References ...... 13 Appendix A . Recurrence intervals ...... 14 Appendix 0, Procedure of Newmark and Hall for determination of response spectra ...... 15 Appendix C, Ground motion data ...... I5

ILLUSTRATIONS

Figure I. Map of proposed route of trans-Alaska oil pipeline showing seismic zonation and magnitudes of design earthquakes...... 2 2 . Accelerogram from 19bb Parkfield earthquake illustrating definition of ground motion parameters ...... 4 3 .6 Graphs showing: 3. Peak horizontal acceleration versus distance to slipped fault as a function of magnitude ...... 5 4 . Peak horizontal acceleration versus distance to slipped fault (or epicentral distance) for magnitude 5 earthquakes...... b 5 . Peak horizontal accelerations and 0.05g horizontal dura fault (or epicentral distance) for magnitude b earthquakeS...... 7 6 . Peak horizontal acceleration versus distance to slipped Fernando earthquake ...... 8 7. Unfiltered and filtered accelerograms of the 1971 Santions versus distance to slipped fault for 1966 Parkfield earthquake 9 8- 11 Graphs showing: 8 . Peak horizontal velocity versus distance to slipped fault (or epicentral distance) for magnitudes 5, 6, and 7 ...... 10 9 . Peak horizontal dynamic displacement versus distance to slipped fault (or epicedral distance) for magnitudes 6, 6, and 7 ...... 12 Duration of shaking versus distance to slipped fault (or epicentral distance) for magnitudes 5, 6, and 7 ...... 13 Examples of tripartite logarithmic ground and respon;e spectre ...... 45

TABLES

Page

Table I. Design earthquakes along the pipeline route...... 2 2 . Near-fault horizontal ground motion...... 3 3 . Peak horizontal ground accelerations from filtered and unfiltered accelerograrns at Pacoirna dam...... 8 4 . Peak ground acceleration data for which distances to the causative fault are most acourately known...... I6 5 . Strong motion data plotted on graphs showing peak horizontal acceleration, velocity, and dynamic displacement and duration of shaking as a function of distance to slipped fault [or epicentral distance) ...... 19 Ground Motion Values for Use in the Seismic Design of the Trans-Alaska Pipeline System

By Robert A. Page, David M. Boore, William B. Joyner, and Henry W. Coulter

ABSTRACT paction, and liquefaction. This report is con- The proposed trans-Alaska oil pipeline, which would cerned only with seismic shaking that, if not traverse the state north to south from Prudhoe Bay on accommodated in the design, could cause defor- the Arctic coast to Valdez on Prince William Sound, will mation leading to failure in the pipeline, storage be subject to serious earthquake hazards over much of its length. To be acceptable from an environmental tanks, and appurtenant structures and equip- standpoint, the pipeline system is to be designed to mini- ment and ultimately to the leakage of oil. It mize the potential of oil leakage resulting from seismic might also induce effects such as seiching of shaking, faulting, and seismically induced ground de- liquids in storage tanks and liquefaction, land- formation. sliding, and differential compaction in founda- The design of the pipeline system must accommodate tion materials, all of which could result in defor- the effects of earthquakes with magnitudes ranging mation and potential failure. from 5.5 to 8.5 as specified in the "stipulations for Pro- To protect the environment, the pipeline sys- posed Trans-Alaskan Pipeline System." This report characterizes ground motions for the specified earth- tem is to be designed so as to minimize the po- quakes in terms of peak levels of ground acceleration, tential of oil leakage resulting from effects of velocity, and displacement and of duration of shaking. earthquakes. The magnitudes of the earth- Published strong motion data from the Western quakes which the design must accommodate are United States are critically reviewed to determine the given in "Stipulations for Proposed Trans-Alas- intensity and duration of shaking within several kilom- eters of the slipped fault. For magnitudes 5 and 6, for kan Pipeline System" ( [U.S.] Federal Task which sufficient near-fault records are available, the Force on Alaskan Oil Development, 1972, Ap adopted ground motion values are based on data. For Set. 3.4.1,. p.- 55), hereinafter referred larger earthquakes the values are based on extrapola- to "Stipulations*~ hi^ characterizes tions from the data for smaller shocks, guided by simpli- fied theoretical models of the faulting process. ground motions for the specified design earthquakes. The seismic design of the proposed pipeline INTRODUCTION involves a combination of problems not usually The route of the proposed trans-Alaska oil encountered. In the design of important struc- pipeline from Prudhoe Bay on the Arctic Ocean tures, detailed geologic and soil investigations to Valdez on Prince William Sound intersects of the site generally provide the background several seismically active zones. Sections of the data. Such detailed site investigations are not proposed pipeline will be subject to serious economically feasible for a linear structure earthquake hazards, including seismic shaking, nearly 800 miles long. In addition, a structure faulting, and seismically induced ground defor- more limited in extent can be located on compe- mation such as slope failure, differential com- tent foundation materials and away from known inelastic and nonlinear for large ground mo- There is substantial evidence that the durations. Finally, smoothed tripartite logarithmic tion of shaking strongly affects the extent of response spectra are constructed from the de- damage caused by an earthquake ; yet the prob- sign seismic motions by the general procedure lem of how duration is related to magnitude has of Newrnark and Hall (1969)) outlined in Ap- received little attention in the literature. In this pendix B. study, the measure of duration used corresponds The initial step in the design process discussed to the time interval between the first and last herein characterizes ground motion appropriate peaks of absolute acceleration equal to or larger to the design earthquakes. This step is based than 0.06 g. Operational definitions of the ac- solely on seismological data and principles and celeration and duration parameters are illus- does not incorporate factors dependent on soil- trated on an accelerogram in figure 2. structure interactions, deformational processes The values in table 2 are based on instrumen- within structures, or the importance of the tal data insofar as possible. Strong-motion data structures to be designed. It involves scientific have been obtained within 10 km of the causa- data and interpretation, whereas the subsequent tive fault for shocks as large as magnitude 6, steps involve engineering, economic, and social but no accelerograms are available from within judgments relating to the nature and value of 40 km of the fault for a magnitude 7 shock and the structures. from within more than 100 km for a magnitude The choice of parameters with which to spec- 8 shock. Estimates of intensity of near-fault ify ground motion was guided by the design ap- ground motion for shocks larger than magni- proach adopted for the pipeline project. A use- tude 6 are extrapolated from data obtained at ful set for the derivation of tripartite structural larger distances or from near-fault data from response spectra includes acceleration, velocity, smaller shocks. displacement, and duration of shaking. The ground motion values in table 2 are GROUND MOTION VALUES subject to several conditions as follows. They Table 2 characterizes near-fault horizontal are for a single horizontal component of motioh. ground motion for the design earthquakes. The The intensity of shaking in the vertical direc- intensity of shaking is described by maximum tion is typically less than two-thirds that in a values of ground acceleration, velocity, and dis- horizontal direction. They correspond to normal placement. In addition to the maximum acceler- or average geologic site conditions and are not ation, leve1,s of absolute acceleration exceeded intended to apply where ground motion is or attained two, five, and ten times are specified, strongly influenced by extreme contrasts in the because a single peak of intense motion may elastic properties within the local geologic sec- contribute less to the cumulative damage po- tion. They characterize free-field ground motion, tential than several cycles of less intense shak- that is, ground motion not affected by the pres- ing. Levels of absolute velocity exceeded or ence of structures. They contain no factor relat- attained two and three times are also given. ing to the nature or importance of the structure

Table 2.-Near-fault horizontal ground motion

Accoloration ig) Valoclty Icm/sacl Magnituda Peak absolute values Peak absolutm valuer Duration' IsecJ

lfime interval between flrst and last peaks of absolute acceleration equal to or greater than 0.05 q. Notms-I. Italic values are bared on instrumental data. 2. Tha values in this table are for a single horizontal componmnt of motion at a distance of a few 13-51 km of the causative fault: are for sltos at which ground motion Is not strongly altersd by extreme contrasts in the elastic properties wlthin the local geologic section or by tha presence of structures: and contain no factor ralating to the nature or lmpprtance of the structure baing daslqnod. 3. The values of actmlaration may bo oxcamdad if thara is apprsciable high-frequency (highor than 8 Hz1 anargy. 4. The values of dlrplacement are for dynamic ground displacements from which spoctral componantr with periods greater than 10 to 15 seconds are rmmoved. 0 I I I I I I I,O - -

- b w

Z - - 0- I- Q E W J * W 0 0 Q 0.01 - EARTHQUAKE MAGNITUDE - A 5.0-5.9 0 6.0-6.9 7.0-7.9

10 DISTANCE (KM) Figure 3.-Peak horizontal acceleration Venus distance to slipped fault as a function of magnitude. Except for 1949 Puget Sound shock (open squares), data shown are those for which distances to fault are most accurstely known (tabulated in Appendix C). Straight-line segments cennoct observations at differont statiom for an individual earthquake, for three magnitude 6 shocks and one magnitude 7 shock. From top to bottom, suites of magnitude 5 data are from 1970 Lytle Creek (m = 5.4), Parkfield (m = 5.5), and 1957 Duly City (m = 5.3) shocks. Closest Parkfield data point lies off plot to left at 0.08 km. For magnitude 6, most data within 100 km are from 1971 San Fernando earthquake (m = 6.6), and most data beyond 100 km are from 1968 Berrego Mountain earthquake (m = 6.5). Most magnitude 7 data are from 1952 Kern County shock (m = 7.7). Open squares are values from 1949 Puget Sound event (m = 7.1), for which distances are determined to hypocenter assuming minimum focal depth of 45 km. Arrows denote minimum values. San Fernando earthquake (m = 6.6). This shock more than 100 accelerograms at distances be- produced one accelerogram at a distance of yond 15 km. The peak acceleration from Paco- about 3 km from the inferred slip surface and ima at 3 km lies beneath a straight-line extrap- olation of the trend of the data beyond 10 km; this behavior is consistent with a zone of little attenuation near the fault as observed for the Parkfield data in figure 4. The maximum acceleration from the San Fer- nando earthquake was 1.25 g, nearly double the maximum acceleration recorded during any earthquake prior to 1971. The acceleration was recorded at a bedrock site adjacent to the Pacoima dam. Because the Pacoima accelerations are so much higher than those recorded in previous earthquakes, the question has arisen whether or not the record might be anomalous in the sense that the motion may have been significantly amplified by various site factors such as the rugged topographic relief, the presence of the dam, and the cracking and minor landsliding near the station. The authors are not aware of any investigations of poaaible site effects that conclusively demonstrate an anomalous amplification (greater than 25-50 percent) of recorded motion in the frequency range 1-10 01 I I 1 Hz. The Pacoima ground motion in the period 0 5 10 15 range 1 to 2 seconds is not inconsistent with DISTANCE FROM FAULT (KM) that predicated from a simple theoretical fault model for the earthquake (Trifunac, 1972). The near-fault acceleration values for magnitude 6.5, table 2, were derived from the Paco- ima accelerograms of the San Fernando earthquake. In the Newmark and Hall method for estimating velocity response spectra (Appen- dix B), the spectral amplitude in the approxi- mate frequency range 2-8 Hz is directly propor- tional to the peak ground acceleration. If the peak acceleration is dominated by higher frequency energy, the Newmark and Hall method overestimates the spectrum in this range. Fre- quencies higher than 8 Hz contributed significantly to the peak accelerations recorded at Pacoima (fig. 7) ; accordingly, the accelerograms were filtered to remove frequencies higher than about 9 Hz. Filtering reduced the accelerations by about 26 percent, as seen in table 3 and fig. 7. The near-fault acceleration values of table 2 for magnitude 6.5 were adopted from the filtered values. I I I 0 5 10 IS Near-fault accelerations for magnitudes larg- DISTANCE FROM FAULT (KM) er than 6.5 were extrapolated from strong mo- UNFILTERED

FILTERED (8~8.5 Hz)

Figure 7.-Unfiltered and filtered accelerograms of the 1971 San Fernando earthquake from the 5. 74" W. accelerograph component at Pacoima dam. Response of filter is 1.0 at frequencies less than 8 Hz and 0.0 at frequencies greater than 9 Hz with a half-wave cosine taper from 8 to 9 Hz.

VELOCITY to the fault is accurately known (large symbols) The response curve of the standard strong tend to separate according to magnitude; the motion seismograph operated in the United remaining data confirm this tendency, although States is flat to acceleration over the frequency their behavior is somewhat obscured by scatter range of the predominant ground motion. Ac- arising at least partially from errors in dis- cordingly, accelerations are measured directly tances. The plot reveals that beyond about 10 from the strong motion recordings, whereas km, peak velocity attenuates less rapidly with velocities are obtained by integration of the distance than peak acceleration. record. For this reason, there are few velocity The near-fault velocity values for magnitude data in the literature relative to acceleration 5.5 (table 2) are averages of the Parkfield data. values recorded at 0.08 and 5.5 km from the Peak velocity data in the magnitude range fault. The values for magnitude 6.5 are based 5-7 plotted as a function of distance from the on the San Fernando observations at the Paco- source (fig. 8) indicate that peak velocity in- ima site about 3 km from the fault surface. For creases with magnitude at all distance for which the larger magnitudes, the values were extrap- data exist. Those data points for which distance olated from those for 5.5 and 6.5 on the as- namic displacements excluding spectral compo- from shocks of similar magnitude suggest that nents with periods greater than about 10-15 the "intense" or "strong" phase of shaking men- seconds are available from double integration of tioned in felt reports corresponds to accelera- accelerograms or directly from displacement tions of about 0.05 g and greater. In comparison, meters. Both types of data are subject to uncer- the minimum perceptible level of acceleration is tainties. In the double integration of digitized 0.001 g (Richter, 1958, p. 26). accelerograms, errors may arise from low-fre- Durations obtained for several earthquakes quency noise in the digitization of the original in the magnitude range 5-7 indicate that for a accelerogram and from lack of knowledge of the given magnitude, duration decreases with in- true baseline of the accelerogram. On the other creasing distance from the source, and that at hand, there are instrumental difficulties associ- a given distance from the source, duration in- ated with displacement meters operating with a creases for larger magnitudes (fig. 10). The 0.05 free period of 10 seconds. The relative accuracy g duration for magnitude 5.5 (table 2) is the of the two types of data is not adequately un- mean of the maximum durations for the 1966 derstood (Hudson, 1970). Parkfield shock (m = 6.5) recorded at distances Peak displacement data obtained from double of 0.08 and 5.6 km from the fault surface (fig. integration of accelerograms and from 10-see- 5). The durations for magnitude 6.5 and 7.0 are ond displacement meters when plotted against based respectively on the measured 0.05 g dura- distance (fig. 9) show no apparent systematic tions of 13 seconds at Pacoima dam in the 1971 ditierence between the two types of data within San Fernando earthquake (rn = 6.6) and of 30 the scatter of the points. Peak displacement at seconds at El Centro in the 1940 Imperial Valley a given distance from the fault, like peak accel- earthquake, which was a multiple event charac- eration and velocity, increases with magnitude. terized by a surface-wave magnitude of 7.1. The near-fault value of peak displacement for These data were smoothed slightly to obtain a magnitude 5.5 (table 2) is the mean of the Park- regular increase of duration with magnitude in field values obtained at 0.08 and 5.5 km from table 2. The adopted near-fault durations of 17 the fault. For magnitude 6.5 the value is based and 25 seconds for magnitudes 6.5 and 7.0 are on the Pacoima record for the San Fernando consistent with the duration data in figure 10 earthquake. How peak dynamic displacement within the scatter of the points. (for periods less than 10-15 seconds) scales with In the absence of near-fault data for larger magnitude for larger shocks is uncertain. An magnitudes, durations can be estimated from upper limit to the increase of near-fault dynam- theoretical calculations in corroboration with ic displacement with magnitude is the rate at felt observations. Assume that a magnitude 8.6 which fault dislocation increases with magni- earthquake is a multiple event comprised of tude. The total fault slip in the 1964 Alaska several shocks as large as magnitude 7.5 distri- shock (rn = 8.5). Hence, an upper bound on the buted along a fault 500-1,000 km in length. Peak peak dynamic displacement for magnitude 8.6, acceIerations of 0.05 g or greater are expected after removal of low frequency energy, is about for a magnitude 7.5 earthquake at distances up 2 m. In this study, a value of 1 m is assumed to 100 km (fig. 3). For a rupture propagation for magnitude 8.5, and the values between mag- velocity of 2 to 3.5 km/sec, the 0.05 g duration nitude 6.5 and 8.5 are smoothly interpolated. at a near-fault station near the center of the DURATION fault would be 100 to 57 seconds, respectively. The measure of duration used in this study In comparison, felt reports of the duration of is the time interval between the first and last intense shaking in the aftershock zone of the acceleration peaks equal to or greater than 0.05 1964 Alaska earthquake (m= 8.6) ranged from g. Although crude, this measure is readily ap- 60-90 seconds at Whittier (Kachadoorian, 1966) plied to the existing accelerograms and approxi- to 150 seconds at Kodiak (Kachadoorian and mates the cumulative time over which the Plafker, 1967). The tabulated duration of 90 ground accelerations exceed a given level. Com- seconds for magnitude 8.6 (table 2) is consistent parison of felt reports for earthquakes of mag- with the calculated range of values and with nitude 6 and 6 with near-fault accelerograms felt data from the 1964 shock. DISTANCE (KM)

Figure 10.-Duration of shaking versus distance to slipped fault, if known, or epicentral distance for magnitudes 5, 6, and 7. Shown is 0.069 duration (see fig. 2 for definition; plotted data are tabulated in Appendix C). Distances represented by larger symbols are uncertain by less than 5 km; those indicsted by smaller symbols by 5 to possibly 25 km. Arrows denote minimum values.

The durations for magnitude 7.5 and 8.0 were 75, p. 4997-5009. interpolated between the values for magnitudes Gedney, L., and Berg, E., 1969, The Fairbanks earthquakes of 7.0 and 8.5 to obtain a smooth increase in dura- June 2 1, 1967: Aftershock distribution, focal mechanisms, and crustal parameters: Seismol. Soc. America Bull., v. 59, p. tion with magnitude. 73- 100. REFERENCES Hastie, L. M., and Savage, J. C., 1970, A dislocation model for Alyeska Pipeline Service Company, 1971, Alignment sheet 65 the 1964 Alaska earthquake: Seismol. Soc. American Bull., design, Appendix A-3.1023 of Project description of the v. 60, p. 1389-1392. trans-Alaska pipeline system: Houston, Texas, Alyeska Pipe- Hudson, D. E., 1970, Ground motion measurements, in Wiegel, line Service Co., 92 p. R. I.., (ed.) Earthquake Engineering: Englewood Cliffs, N.J., Arnbraseys. N. N., 1969, Maximum intensity of ground movements Prenties-Hall, p. 107-125. caused by faulting: Fourth World Conference on Earthquake Hudson, D. E., Brady, A. G., Trifunac, M. G., and Viiayaraghavan, Engineering, Santiago, Chile, 1967, Proc., v. I, p. 154-171. A., 1971, Strong-motion earthquake accelerograms, digitized Barrows, A. G., Kahle, J. E., Weber, F. H., and Saul, R. B., 1971, and plotted data, Volume 2, A: California Inst. Technology, Map of surface breaks resulting from the San Fernando, Cali- Earthquake Eng. Research Lab. Rept. 71-50, p, 321. fornia, earthquake of February 9, 1971: California Div. Mines Kachadoorian, R., 1966, Effects of the earthquake of March 27, and Geology Prelim. Rept. No. I I, plate I. 1964 at Whittier, Alaska: U.S. Ceol. Survey Prof. Paper Bruno, J. N., 1970, Tectonic stress and the spectra of seismic 542-8, 2 1 p. shear waves from earthquakes: Jour. Geophys. Research, v. proaching the design magnitude has occurred envelope of the actual response spectrum. New- on the pipeline route in this century, a magni- mark and Hall (1969) describe a graphical tude 7.3 shock in 1937. A recurrence interval of method for determining envelope response 50 years is assumed. spectra. First a tripartite logarithmic "ground In the magnitude 7.0 and 5.5 zones, there is motion spectrum" is constructed with three no historic record of shocks as large as the de- lines representing ground displacement, veloc- sign earthquakes. For the Willow Lake to Pax- ity, and acceleration. These lines are then son zone, the record of earthquakes equal to or shifted upward on tripartite log paper, by larger than magnitude 7.0 is probably complete amounts depending on damping, to reflect the for at least 50 years. From 67" N to Prudhoe dynamic amplification of the ground motion in Bay, the record for events as small as magni- the structure. The amounts by which the lines tude 5.5 is possibly complete since 1935, when are shifted are derived empirically from re- a seismic station was established at College, corded accelerograms and are subject to revision Recurrence intervals of 200 and 50 years are as new data become available. This procedure, assumed for the two zones. using the amplification factors given by New- mark and Hall (1969), is illustrated in figure 11, where the velocity response spectrum for 2 per- APPENDIX %PROCEDURE OF NEWMARK AND HALL FOR DETERMINATION OF RESPONSE SPECTRA A response spectrum for a given level of damping is defined by the maximum responses (usually expressed in terms of displacement, velocity, or acceleration) of linear, single-de- gree-of-freedom oscillators (with different free periods but identical values of damping) when subjected to a specified time history of ground motion. A single spectrum is a plot of the maximum responses as a function of oscillator period or frequency; there is a different response spectrum for each level of damping. The usefulness of the response spectrum comes from the ability to mode! engineering structures by equivalent simple damped oscillators and to estimate stresses induced by the particular ground mo- FREQUENCY (CPS) tion from knowledge of the equivalent period

and damping of the structure and of the appro- Figure I I.-Example of tripartite logarithmic ground (dashed) priate response spectrum. and response (solid) spectra (after Newrnark and Hall, 1969). The values of parameters describing the ac- Accelerations and displacements may be read from the plot in tual ground motion may be modified for non- addition to velocities. Response spectrum is for damping value linear energy-absorbing mechanisms before be- of 2 percent of critical. ing used in the construction of a response spec- cent damping is estimated for a ground motion trum. In the following example of the Newmark characterized by ground displacement of 12 and Hall method for constructing response spec- inches, velocity of 16 inches per second, and tra, the ground motion values are not modified. acceleration of 0.33 g. At high and low frequen- The example is illustrative only of the general cies, the response spectrum must theoretically method and not of an application to a specific equal the ground acceleration and displacement, problem. respectively; this accounts for the slope that Response spectra calculated from accelero- connects the 1.4 g and 0.33 g lines. The corres- grams often contain many peaks and troughs, ponding line at the low frequency side is off the hence prudent design requires the use of an graph to the left. Table 4.-Peak ground acceleration data for which distances to the causative fault are most accurately known--Continued

DATE EARTHQUAKE HAG STAT ION ACC OISf UNCER YR MCI DA G KM KH 70 09 12 LYTLE CK-rCACIFORNIA 594 WRIEHTWOOO 195 15- 2-5 CEDAR SPRlNGStRANCH 087 18- CEDAR SPRINGSwDAH -072 18- DEVILS CANYON 193 19- SAM BERNARDINO 125 28-

C OLTON 9 049 290 PUDDINGSTONE DAM 0022 32. LDMA LINDA .068 34. SANTA ANITA DAM 057 44.

HAGN ITUDE 6.0-4- 9

40 05 19 IMPERIAL VALLEYWCALIF. 6.4 EL CENTRO -36

68 04 09 BORREGO HTN.*CALIFr 6.5 EL CENTRO SAN DIEGO PERRIS RESERVOIR SAN ONOFRE COLTON SAN BERNARDINO DEVILS CANYON CEDAR SPRINGS SANTA ANA SAN DIMAS LONG BEACHIUTIL~BLDE. LONG BEACHvS- CAL. ED. SANTA ANITA RES. VERNON PASADENAIFAC.CLUB PASADENA.SEISMO.LAB. L.A.,SUBWAY TERM* L-A-tfDISON PEARBLOSSOM WESTWOOD GLENDALE HOLLYWOOD STOR- PE LOT PACOIHA DAM FAIRMONT RESERVOIR LAKE HUGHESCl DAVIS DAM CASTAIC GORHAN PORT HUENEHE SANfA BARBARA BAKERSFIELD TAFT

71 02 09 SAN FERNANDO, CALIF. 6.6 PACOIHA DAM 1-24 L-A-t GRIFFITH PARK 18 PASADENA* SEISMO. LAB. el9 SANTA AN~TADAM 24 LAKE HUGHES #12 037 LAKE HUGHES #9 -16 TEJON - 03 Figures 4 and 6 provide comparison of the derived by integration of accelerograms, were better acceleration data for magnitudes 5 and obtained primarily from three sources (Hudson 6, respectively, with acceleration data for which and others 1971; Wiggins, 1964; and Ambra- distances to the fault are less well known. The seys, 1969). The tabulated velocity is the larger figures include accelerations recorded within 100 of the two peak horizontal values. km of the fault or epicenter for shocks that pro- The displacement data in figure 9 are derived vided one or more accelerograms within 32 km. from displacement records obtained either di- Table 5 summarizes the data, which were ob- rectly from 10-second displacement meters or tained from several sources, including the an- analytically by double integration of accelero- nual issues of "United States Earthquakes." grams. Data from the 10-second displacement The tabulated acceleration is the larger of the meters are taken from the annual issues of two peak horizontal values. The tabulated dis- "United States Earthquakes." Displacements tance is the closest distance to the slipped fault, obtained from twice-integrated accelerograms if determinable, or epicentral distance. The un- are primarily from Hudson, Brady, Trifunac, certainty in distance is indicated by the letter and Vijayaraghavan (1971) and correspond to A, B or C, representing estimated uncertain- ground motion from which spectral components ties of less than 2 km, 2-5 krn, and 5-25 km, re- with periods longer than about 15 Hz are respectively. moved. The tabulated displacement is the larger Table 5 also summarizes the velocity, dis- of the two peak horizontal values. placement and duration data plotted in figures The 0.05 g durations plotted in figure 10 were 8,9 and 10, respectively. Figure 8 illustrates the measured from published accelerograms. The dependence of peak horizontal ground velocity larger of the two horizontal durations is tabu- upon magnitude and distance. The velocity data, lated. Table 5.-Strong motion data plotted on graphs showing peak horizontal acceleration, velocity, and dynamic displacement and duration of shaking as a function of distance to slipped fault (or epicentral distance) DATE EARTHQUAKE HAG STATION DISTANCE ACC VEL OZSP OUR YR HO DA KH G CM/SEC CH ** SEC MAGNITUDE 5.0-5.9

33 10 02 LONG BEACH 5.4 VERNON 14 C 9 115 200 LONG BEACH I5 .077 1.0 L A SUB TERM 19 082+ -8 0 O.O+ W~STWOOO 2 4 .009+ HOLLYYOOD STOR 27 -033+ PASADENA 30 .005+ 02 D

34 07 06 N CALIF COAST 5.7 EUREKA 149 C .9 0

35 01 02 C WENDOCINO 5.8 EUREKA 117 C • 1+0

35 11 28 HELENA, WONT 5.2 HELENA 8 C 0082 3-9

37 02 07 C MENOOCINO 5.8 FERNOALE 84 C 3.8 6+0

38 5 31 SANTA ANA MT 5.5 COLTON 47 C L A SUB TERM 78

38 09 12 C MENDOCINO 5.5 FERNDALE 51 C 6.1

40 05 19 IMPERIAL VAL 5.2 EL CENTRO 16 C -077 0.5

40 12 20 C MENOOCIYO 5.5 FERNDALE 91 C EUREKA 103

41 07 01 SANTA BARBARA 5.9 SANTA BARBARA 14 C .Its+ 20.3 L A SUB TERM 127 -2 0 Table 5.-Strong motion data plotted on showing peak horizontal acceleration, velocity, and dynamic displacement and duration of shaking as a fundion of distance to slipped fault (or epicentral distance)-Continued DATE EARTHQUAKE MAG STATION ~iSf4NCf ACC VEL OISP OUR VR HO DA KM r G CMISEC CM +* SEC 55 12 17 BRAWLEY 5.4 EL CENTRO 22 C -083+ 5-1 1. O+ 57 03 22 DALY CITV 5.3 SFGOLONGATE 8 B -129 4.9 203A 1.5 S F STATE 12 9105 5.1 1.1 A 100 S F STATE 12 1.1 D S F ALEXANDER 14 0056 2-9 lo3 A 0.0+ S F $0 PACIFIC 14 e049 5-0 1.4 A 0.0 OAKLAND 24 0048 2.0 1.5 A 0.0 SAN JOSE 58 007

57 04 25 CALIPATRIA 5.2 EL CENTRO 51 C -6 D

57 04 25 CALIPATRIA 5.1 EL CENTRO 51 C 04 D

60 01 19 HOLLISTER 5.0 HOLLISTER 8 C 064 3.0

61 01 09 HOLLISTER 5.6 HOLLISTER 20 C -193 17.1 3.8 A 5.5

62 08 30 LOGAN* UTAH 5.7 LOGAN 7 C • 12 2.5

63 02 28 FORT TEJON 5-0 WHEELER RIDGE 8 C .058

66 06 28 PARKFIELD 5.5 CHOLAME-SHAN 2 0.1~ .S2 72.2 22-3 A 12.0 CHOLAME-SHAM 5 5.5 e48 27.3 8.4 A 805 CHOLAHE-SHAN 8 9rb r28 1206 6.7 A 9.0 f EMBLOR 10.6 042 21.0 8.1 A 5.5 CHOLAHE-SHAN12 14.9 .O?4 6.5 7.1 A 5.0 SAN LUIS OIISP 59 -019

67 06 21 FAIRBANKS.AK 504 COLLEGE 15 0 006

67 12 10 N CALIF COAST 5.8 FERNDALE 32 C 010

70 09 12 LYTLE CREEK 5.4 WRfGHTWOOO CEDAR SPR RCH CEDAR SPR DAM DEVILS CANYON SAN BERNAROINO COLTON PUDDINGSTONE 0 LOWA LINDA SANTA ANITA D ~AGNI~UDE6.0-6.9 33 03 11 LONG BEACH 6.3 LONG BEACH VERNON L A SUB TERM

33 06 25 W NEVADA 6.1 S F SO PACIFIC 302 C 2+D

34 06 07 PARKFIELO 6. PASADENA 2 98 -2 D

34 12 30 MEXICALI* ME% 6.5 EL CENTRO 64 C 15.5 L A SUB TERM 328 03 D

35 10 31 HELENA 6.0 HELENA 7.5G 016 16.8 1.5

37 03 25 COAHUILA VAL 6 COLTON 94 C PASADENA 160 L A SUB TERM 167 Table 5.-Strong motion data plotted on graphs showing peak horizontal acceleration, velocity, and dynamic displacement and duration of shaking as a function of distance to slipped fault (or epicentral distance)-Continued D4TE EARTHQUAKE MAG STATION DISTANCEALL VEL DISP OUR YY #O 0A KH + G CM/SEC CM ** SEC 71 C2 09 SAN FERNAYDO 6.6 PACOXHA DAM 3 0 1.24 115. 439 A 13.0 L A GRIFFITH 16 -18 10.0 PASAOENAISEIS 17 .19 7. 0 SANTA ANITA 0 25 -24 11.5 LAKE HUGHES 12 26 - 37 14.0+ LAKE HUGHES 9 29 -16 4.5+ SANTA FELlCIA 29 .24 6.5+ LAKE HUGHES 4 30 19 4.0 CASTAIC 30 • 39 18-O+ LAKE HUGHES 1 32 -17 99 0 PALMDALE 35 13 FAIRMONT RE5 35 010 PEARBLOSSOM + 3 .15 PUDDINGSTONE U 48 09 PGLOS VERDES 52 04 OSO PUMP PLANT 54 05 LONG BEACH TRM 58 03 WRIGHTWOOD 61 05 TEJON 7 0 03 PORT HUENEME 71 03- GRAPEVINE 73 907 WHEELER RIDGE 88 03 CEDAR SPK RCH 94 002 CEDAR SPR DAM 94 03 COLT ON 97 .04 MAGNITUDE 7-0-7.9 40 05 19 IMPERIAL VAL 7.1 EL CENTRO 10.8

49 04 13 PUGET SNDtWASH 7.1 OLYMPIA 48+ C SEATTLE b9+

52 07 2~ KERN COUNTY 7.7 TAFT SANTA BARBARA HOLLYWOOD BSMT HOLLYWOOD LOT PASADENA PASADENA L A SUB TERM COLTON

54 12 16 FALLON* NEV 7.0 S F SO PACIFIC 404 L A SUB TERM 584

NOTES:

+ UNCERTAINTY IN DISTANCE: AXLESS 1HAN 2 KM 012 TO 5 KM C=5 TO POSSIBLY 25 KM

** SOURCE OF DISPLACEMENT DATA: A=DOUBLE INTEGRATION OF ACCELEROGRAH 0110-SEC DISPLACEMENT METER