Aspects of Pacific Seismicity

Aspectsof Pacific Seismicity

Editedby EmlleA.Okal

1991 Springer Basel AG

Reprint from Pure and Applied Geophysics (PAGEOPH), Volume 135 (1991), No. 2

Editor's address:

Emile A. Okal Northwestern University Department of Geological Sciences Evanston, IL 60208 USA

Library of Congress Cataloging-in-Publication Data

Aspects of Pacific seismicity / edited by Emile A. OkaI. p. cm.

Published also as v. 135, no. 2 of Pure and Applied Geophysics.

1. Seismology-Pacific Area. I. Okal, Emile A. QE537.2.P2A76 1991 551.2'2'091823-dc20

Deutsche Bibliothek Cataloging-in-Publication Data

Aspects ofpacific seismicity / ed. by Emile A. OkaI. - Basel Boston ; Berlin : Birkhäuser, 1991

NE: Okal, Emile A. [Hrsg.]

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically those of translation, reprinting, re-use of illustrations, broadcasting, reproduction by photocopying machine or similar means, and storage in data banks. Under § 54 of the German Copyright Law where copies are made for other than private use a fee is payable to »Verwertungsgesellschaft Wort«, Munich.

©1991 Springer Basel AG Originally published by Birkhäuser Verlag AG Basel in 1991.

ISBN 978-3-0348-5641-6 ISBN 978-3-0348-5639-3 (eBook)DOI 10.1007/978-3-0348-5639-3

Contents

167 Introduction, E. A. Okal

169 Circum-Pacific seismic potential: 1989-1999, S. P. Nishenko

261 Intraplate seismicity ofthe Pacific Basin, 1913-1988, M. E. Wysession, E. A. Okal and K. L. Miller

PAGEOPH, Vol. 135, No. 2 (1991)

Introduction

0033-4553/91/020167 -02$1.50 + 0.20/0 © 1991 Birkhäuser Verlag, Basel

This special issue of PAGEOPH regroups two extensive studies of the seismicity of the Pacific region. They are intended to provide a comprehensive update on our understanding of the occurrence of seismicity along the plate boundaries of the Pacific Basin, as wel1 as in its interior.

The first paper, "Circum-Pacific Seismic Potential, 1989-1999", by Stuart P. Nishenko, is an assessment of the present seismic potential in 96 circum-Pacific plate boundary zones, in the form of the specific conditional probabilities for the occurrence of large to great earthquakes along these segments, during a number of time windows extending from 1989 to 2009. Both in its goals, its approach, and its general philosophy, this study fol1ows in the steps of a previous study by Nishenko and three co-workers, published 12 years aga in PAGEOPH (McCann et al., 1979). The paper builds on the experience acquired in the past decade, notably by the detailed seismo10gical study of great historical earthquakes, but also through significant progress in the recognition and dating of the geological evidence for pre-instrumental events; in addition, we have obviously learned a great deal from the large subduction events of the past decade (Colombia, 1979; Mexico and Chile, 1985; Aleutian 1986, to name a few). The paper conc1udes by ranking the seismic gaps with the highest probability of activity in the next ten years, and the next twenty years, respectively.

The second paper, "Intraplate Seismicity of the Pacific Basin, 1913-1988", by Michael E. Wysession, Emile A. Okal, and Kristin L. Mil1er, presents a thorough compilation of al1 seismicity reported as intraplate in the Pacific Basin, since the inception of the regular listing of seismological observations in the International Seismological Summary. In addition to about 800 earthquakes be\onging to wel1-defined temporal and spatial swarms, approximately 900 events were critical1y analyzed and most of them relocated. In the end, only 45% proved to be genuinely intraplate, with the remainder a mixture of plate boundary events erroneously listed as intraplate, poorly constrained solutions for which an interplate location cannot be ruled out, and blatant errors resulting from typographicalor other systematic errors upon compilation. The paper also catalogs al1 available focal mechanisms and presents some statistics on the evolution with time of the detection of intraplate earthquakes, as wel1 as of the accuracy of the location process.

168 Introduction PAGEOPH,

Both papers in the issue are obviously intended and expected to serve as basic reference for future investigations of these problems.

Emile A. Okal

REFERENCE

MCCANN, W. R., NISHENKO, S. P., SYKES, L. R., and KRAUSE, J. (1979), Seismic Gaps and Plate Tectonics: Seismic Potential Jor Major Boundaries, Pure Appl. Geophys 1/7, 1082-1147.

PAGEOPH, Vol. 135, No. 2 (1991) 0033-4553/91/020169-91$1.50 + 0.20/0 © 1991 Birkhäuser Verlag, Basel

Circum-Pacific Seismic Potential: 1989-1999

STUART P. NISHENK01

Abstract - The seismic potential for 96 segments of simple plate boundaries around the circumPacific region is presented in terms of the conditional probability for the occurrence of either large or great interplate earthquakes during the next 5, 10, and 20 years (i.e., 1989-1994, 1989-1999 and 1989-2(09). This study represents the first probabilistic summary of seismic potential on this scale, and involves the comparison of plate boundary segments that exhibit varying recurrence times, magnitudes, and tectonic regimes. Presenting these data in a probabilistic framework provides a basis for the uniform comparison of seismic hazard between these differing fault segments, as weil as accounting for individual variations in recurrence time along a specific fault segment, and uncertainties in the determination of the average recurrence time.

The definition of specific segments along simple plate boundaries relies on the mapping of earthquake rupture zones as defined by the aftershock distributions of prior large and great earthquakes, and historic descriptions of feit intensities and damage areas. The 96 segments are chosen to represent areas likely to be ruptured by "characteristic" earthquakes of a specified size or magnitude. The term characteristic implies repeated breakage of a plate boundary segment by large or great earthquakes whose source dimensions are similar from cyc1e to cycle. This definition does not exc1ude the possibility that occasionally adjacent characteristic earthquake segments may break together in a single, larger event. Conversely, a segment mayaiso break in aseries of smaller ruptures.

Estimates of recurrence times and conditional probabilities for characteristic earthquakes along segments of simple plate boundaries are based on I) the historic and instrumental record of targe and great earthquake oc,currence; 2) paleoseismic evidence of recurrence from radiometric dating of Holocene features produced by earthquakes; 3) direct calculations of recurrence time from the size of the most recent characteristic event and the long-term rates of plate motion assuming the validity of the time-predictable model for earthquake recurrence; and 4) the application of a lognormal distribution for the recurrence times of large and great earthquakes. .

Time-dependent estimates of seismic potential are based on a physical model of earthquake occurrence which assumes that the probability for an earthquake is low immediately following the occurrence of a characteristic earthquake and increases with time as the stress on the fault segment recovers the stress drop of the event. This study updates earlier work on seismic gaps by explicitly inc1uding both recurrence time information and the temporal proximity to the next event as factors in describing earthquake hazards.

Currently, 11 out of 96 regions have a high (i.e., ~ 50%) probability ofrecurrence during the next 10 years and are characterized by either fairly short (i.e., less than 30-40 years) recurrence times or long elapsed times relative to the average recurrence time. The majority of these segments are located in the southwest Pacific (Vanuatu, New Guinea, and Tonga). When a longer time window is considered (e.g., 20 years or 1989-2009), 30 out of 96 regions have a high potential. Many of these regions are located near areas of high population density. These determinations do not preclude rupture of other fault segments, with less than a 50% chance in 10 or 20 years, or large and great earthquakes in areas we have

I National Earthquake Information Center, United States Geological Survey, Denver, CO 80225, U.S.A.

170 Stuart P. Nishenko PAGEOPH,

not studied in detail. While this study has summarized the seismie potential for a large number of regions aroul}d the eireum-Pacifie, there are still a number of geographie and seismotectonie regions that need to be eonsidered, inciuding Indonesia, the Philippines, New Zealand, and the eountries that surround the Caribbean basin.

Key words: Cireum-Pacifie, earthquake foreeasting, earthquake predietion, eharaeteristie earthquakes, probability, seismie hazards.

Introduction

In the past 30 years great strides have been made in the fields of seismology and geophysies towards understanding the occurrenee of large and great earthquakes along simple plate boundaries. These advanees have recently led to the development of long-term earthquake foreeasts for speeifie fault zones. The applieability of these teehniques and ideas to at least some areas of the cireum-Paeifie region was demonstrated by the suecessful foreeast ofthe great (Ms 7.8) 1985 Valparaiso, Chile earthquake (Nishenko, 1985). The loeations of other suecessful earthquake foreeasts and predietions sinee 1940 are shown in Figure I. At present, national earthquake predietion pro grams in the United States and Japan have identified speeifie areas for intensive study, based on regularities in the patterns of historie earthquake oeeurrenee and expeetations of similar sized events in the near future (i.e., Parkfield, California [BAKUN and LINDH, 1985] and the Tokai District, Japan [MOGI, 1981]). On a broader seale, The WORKING GROUP ON CALIFORNIA EARTHQUAKE PROBABILITY (1988, 1990) reports represent the first Federally sanetioned regional probabilistic forecasts for earthquake activity in the Uni ted States.

It is generally known which population centers and sites of eritieal faeilities around the circum-Pacific region have experieneed destruetive large (Ms 7.0-7.7) and great (M w 7.7-9.3) earthquakes in the historie past. These same loealities are also eandidates for the inevitable reeurrenee of similar earthquakes and tsunamis at so me future time. Henee, while it is of aeademic interest to know how long it has been sinee a prior destruetive earthquake occurred at a partieular loeation; it is more important, from a societal perspeetive, to know when the next damaging event will oeeur.

This report summarizes the known seismie history for 123 seismie gaps around

• Plate I

Cireum-Pacifie Seismie Potential 1989-1999. Colors portray the time-dependent eonditional probability for the reeurrence of either large (7.0< M, < 7.7) or great (M" M w > 7.7) shallow, plate boundary earthquakes during the time interval 1989-1999. See text for loeation of segment boundaries and expeeted magnitudes. Probabilities are eonditional on the event not having occurred prior to 1989 and are represented by dark blue (0-20%), green (20-40%), yellow (40-60%), and red (60-100%). Light blue areas are those regions with no historie record of great earthquakes. Specifie dates and magnitudes

refer to those areas with ineomplete historie records. See text for recent forecast updates.

Vol. 135, 1991 Circum-Pacific Seismic Potential 171

.. '

,"

11~U"

I I

1""

6",

IU{;,

I \\

t/!r

~<;v

::::

: I.

I~Ur

..

10°

/)

, f C~

J /

1/_

. .

b/

,

/ I

/ /

Fig

ure

I Su

cces

sful

ear

thqu

ake

fore

cast

s an

d pr

edic

tion

s. S

tars

sho

w t

he l

ocat

ion

of

succ

essf

ul e

arth

quak

e fo

reca

sts

and

pred

icti

ons

from

194

0 th

roug

h 19

86.

Whi

le e

vent

s in

bot

h in

tra-

and

inte

rpla

te e

nvir

onm

ents

are

sho

wn,

the

maj

orit

y o

f su

cces

sful

for

ecas

ts h

ave

been

ass

ocia

ted

with

int

erpl

ate,

or

sim

ple

plat

e bo

unda

ry,

eart

hqua

kes

(aft

er N

ISH

EN

KO

, 19

89).

.... N ~ S ~ z ~ =

i'I'" o ~ Cl 1:l -~


the circum-Pacific region and describes the potential for future large and great earthquakes in terms of conditional probability for the next 10 years (1989-1999) in 96 of those regions. The remaining 27 regions are those with either an incomplete or no historie record of large and great earthquakes. The ten-year interval (1989-1999) overlaps the International Decade of National Disaster Reduction and comes at a time when many nations are faced with the inevitability of natural disasters and the realities of restricted economic resources. The prudent development of disaster mitigation and reduction pro grams for specific locations, within socially beneficial time frames, requires an understanding of when and where natural disasters are to occur.

The portrayal of seismic hazards around the circum-Pacific region involves comparison of many fault segments, or portions of plate boundaries, that exhibit varying recurrence times, earthquake magnitudes, and tectonic regimes. The level of reliability associated with these forecasts va ries from region to region, and is influenced by the completeness of the historie earthquake record and our present understanding of the mode of earthquake rupture in these regions. Presenting these data in a probabilistic framework accounts for individual variations in recurrence time along a specific fault segment, as weIl as errors in our determination of the repeat time, and provides a basis for uniform comparison of seismic hazard between segments which have differing recurrence times. One advantage of studying such a broad and diverse area is that many seismic regions have relatively short recurrence times. The short observed recurrence times, 20-60 years, compared to 100-200-year intervals observed at many other plate boundaries, allow for the evaluation of earthquake forecasts in a relatively brief time. This rapid "turn around" time for scientific evaluations in turn can lead to a better understanding of longer-term processes occurring at other regions and the overall improvement of strategies for implementing disaster mitigation programs.

Table 1 lists the sites most likely to experience a large or great earthquake within the next 10 years, based on the assessment of the seismic potential of 96 gaps around the circum-Pacific region. Large or major earthquakes refer to those events with surface wave magnitudes (Ms ) between 7.0 and 7 3/4. Great earthquakes are those events with seismic moment-magnitudes (M w) larger than 7 3/4. The southwest Pacific region, including the islands of New Guinea, Vanuatu, and Tonga, presently contains the majority of high probability gaps. Some of these high potential gaps near population centers presently include those near Jama, Ecuador and southeastern Guatemala. This list does not account for possible risks of intraplate occurring away from the plate boundary proper. Nor does it preclude damaging earthquakes from occurring in areas that have not been studied in detail (e.g., southeast Asia, Indonesia, and the Caribbean). Many segments of Central America have gaps, that while presently assigned intermediate probabilities for the next 5 years, will become areas of high concern within the next 10 or 20 years. Table 2 lists all the high (i.e. ~ 50%) probability regions for a 20-year window


Table I

Top Seismic Gaps Gaps with ;;? 50% Conditional Probability for Recurrence During 1989-1999

Location Magnitude Last Event Probability

1. Parkfield, Califomia mb 6.0 1966 93% 2. Delarof Is., Aleutians M s 7.4 1957 (85%) 3. Vankolo Is., Vanuatu Ms 7.5 1980 83% 4. Jama, Ecuador M s 7.9 1942 (66%) 5. Nicoya, Costa Rica Ms 7.3 1978 64% 6. S. Santo Is., Vanuatu Is. M s 7.1 1971 60% 7. E. New Britain, New Guinea Ms 7.7 1971 59% 8. W. New Britain, New Guinea Ms 7.9 1945 (58%) 9. Central Tonga Ms 7.8 1948 58%

10. N. Bougainville, New Guinea M s 7.8 1971 53% 11. S. E. Guatemala mb 7.5 1915 51%

Probability values in parentheses reflect less reliable estimates.

Table 2

Top Seismic Gaps Gaps with ;;? 50% Conditional Probability for Recurrence During 1989-2009

Location Magnitude Last Event Probability

1. Parkfield, Califomia mb 6.0 1966 ~99% 2. Vankolo Is., Vanuatu M s 7.5 1980 99% 3. Delarof Is., Aleutians M s 7.4 1957 (98%) 4. Nicoya, Costa Rica M s 7.3 1978 98% 5. Jama, Ecuador M s 7.9 1942 (90%) 6. E. New Britain, New Guinea Ms 7.7 1971 92% 7. S. Santo Is., Vanuatu M s 7.1 1971 91% 8. N. Bougainville, New Guinea Ms 7.8 1971 90% 9. Central Tonga M s 7.8 1948 84%

10. W. New Britain, New Guinea Ms 7.9 1945 (84%) 11. Santa Cruz, Vanuatu M s 8.1 1966 82% 12. Loyalty Is., Vanuatu Ms 7.2 1980 80% 13. SJi Guatemala mb 7.5 1915 79% 14. Shumagin, Is., Alaska Ms 7.4 1917 75% 15. Ometepec, Mexico M s 7.3 1950 74% 16. C. Oaxaca, Mexico Ms 7.8 1928 (72%) 17. Guadalcanal, Solomons M s 7.5 1988 71% 18. San Cristobal, Solomons M s 8.0 1931 (71%) 19. E. Oaxaca, Mexico Ms 7.8 1965 70% 20. Unimax Is., Alaska M s 7.4 1946 (67%) 21. Fox Is., Aleutians Ms 7.4 1957 (67%) 22. Colima, Mexico Ms 7.5 1973 66% 23. West Oaxaca, Mexico Ms 7.4 1968 64% 24. Kamchatsky Pen., U.S.S.R. M s 7.5 1971 61% 25. S. Valparaiso, Chile Ms 7.5 1906 59% 26. Papagayo, Costa Rica M s 7.5 1916 (55%) 27. Tokai, Japan M s 8.0 1854 (53%) 28. Urup Is., Kuriles Ms 8.5 1963 (52%) 29. C. Guerrero, Mexico Ms 7.8 1899-1911 (52%) 30. C. Guatemala Ms 7.9 1942 50%

Probability values in parentheses reflect less reliable estimates.


(1989-2009) and is of use for long-term planning. The 30 highly ranked gaps in Table 2 represent roughly 1/3 of all the gaps studied. Many of these gaps are near urban centers and represent potential threats to those centers. The Appendix contains a list of all the gaps studied, their coordinates, the date and magnitude of the most recent large or great earthquake, and probability estimates for 5-, 10-, and 20-year windows (i.e. 1989-1994, 1989-1999, and 1989-2009).

The assessment of long-term seismic hazard for the simple plate boundaries of the circum-Pacific region is an active and rapidly developing field. The time-dependent nature of these forecasts necessitates that these results be regularly updated. In addition, new data for other seismogenic regions and improvements in the model on which these assessments are based will lead to revision and refinement of the seismic hazard forecasts presented here. Development of reliable intermediate-term forecasts, which cover time intervals of years to months (e.g., KEILIS-BOROK et al., 1988; V ARNES, 1989), would also refine these assessments. These intermediate-term data can narrow the earthquake forecast time window and provide additional motivation for increased awareness and action.

Methodology

This study attempts to quantitatively describe the seismic hazards associated with the future recurrence of large and great characteristic earthquakes along segments of the circum-Pacific seismic zone. These events are termed interplate earthquakes and reflect motion at the boundary of and between crustal plates. By characteristic earthquake, we mean an event wh ich repeatedly ruptures the same fault segment and whose dimensions define that segment (SCHWARTZ and COPPERSMITH, 1984).

Most of the cumulative seismic energy release, seismic moment, and cumulative seismic slip along major plate boundaries occurs in large or great earthquakes. The strain energy that is released in shallow large or great earthquakes is believed to build up slowly along simple plate boundaries for tens to hundreds of years. This strain comes from the movement of the plates, which va ries from 2 to 12 cm/yr for the major plate boundaries discussed here. Friction along plate boundaries prevents many major seismic zones from moving continuously on a scale shorter than tens to hundreds of years. Once stresses build to a criticallevel, the plate interface moves suddenly about 1-20 m during the rupture associated with a large or great earthquake. Aftershocks, which represent readjustments on the fault surface following the occurrence of a mainshock, are not included in the forecasts discussed by this report. With the exception of those earthquakes occurring within the trench outer rise region, intraplate earthquakes, which reflect deformation within plate interiors, are also not covered by this report.

Many analyses of earthquake hazards around the circum-Pacific region have


been based on the seismie gap hypothesis, i.e., the idea that segments of simple plate boundaries that have not ruptured in a large or great earthquake in many decades are the most likely sites of future large or great events (FEDOTOV, 1965; MOGI, 1968; SYKES, 1971; KELLEHER et al., 1973; MCCANN et al., 1979; NISHENKO and MCCANN, 1981). For a segment ofa plate boundary to be considered a seismic gap, it must have a history of prior large or great earthquakes and not have ruptured in a large or great event in at least 3 deeades (MCCANN et al., 1979). Given the present large number of seismic gaps around the cireum-Pacifie, additional information is neeessary to differentiate between those gaps whieh may be sites of large or great shoeks in the immediate future (i.e., the next 5 to 10 years) and those which may remain dormant for longer time intervals. Henee, in addition to knowing when the last event occurred at a partieular loeation, information on the repeat times of large and great earthquakes, the loeal rates of fault motion and strain accumulation, and the size of the expected earthquake are essential for eompletely deseribing the seismie hazard.

Deseriptions of interplate seismie potential are presented in terms of estimates of the expeeted repeat time for large and great earthquakes in this region, and the eorresponding time-dependent eonditional probability for time intervals of 5-, 10-and 20-years duration from the present (i.e., 1989-1994, 1989-1999 and 1989-2009). Estimates of earthquake repeat times are based on I) reeurrenee intervals from the historie and instrumental reeord, 2) radiometrie dating (i.e., 14C and 230Th) of uplifted marine terraces, eoral heads, and off set Holocene deposits, and 3) direet estimates based on the size of the most reeent earthquake and the loeal rate of plate motion. Presenting these data in a probabilistie framework accounts for variations in reeurrenee time along a speeifie fault segment, as weil as errors in our determination of the repeat time, and provides a basis for the eomparison of seismie hazard between segments whieh have differing reeurrence times. The flow ehart in Table 3 presents the proeedure used to estimate seismie potential or eonditional probability. The individual steps in this proeedure will be explained more fully in the following sections.

One approach to probabilistie hazards assessment has been to lump all known reeurrenee information for a speeifie magnitude class into one da ta set for analysis of regional recurrenee statisties. Pooling data from large regions is neeessary beeause weIl determined estimates of reeurrenee times and probabilities require more data than are usually available for individual fault segments. This approach, of pooling the available data, was attempted to RIKITAKE (1976) for a number of individual seismie zones using a Weibull funetion as the preferred statistical distribution. The amount of time elapsed since a prior large or great shock in each segment is an intrinsie faetor for the hazards estimate. The probability, or likelihood, for a large or great eharacteristic earthquake is low immediately following the oeeurrenee of a previous event, and inerease with time as the plate boundary segment reeovers the stress drop of the prior event (RIKITAKE, 1976; HAGIWARA,

Vol. 135, 1991 Cireum-Pacifie Seismie Potential

Table 3

Flowchart 0/ Earthquake Recurrence Time and Conditional Probability Calculations

HISTORIC Dates of historie

earthquakes

Observed Average Repeat Time

Tave

ij

Notes:

GEOLOGIC Dates of prehistorie earthquakes ± ut

Observed Average Repeat Time

Tave

ij Compute rand T;xp

ij Date of last event (10)

ij Tpred = (10 + Texp )

ij Year and window length for probability estimates

ij Longitudinal Distribution

ij Conditional Probability Estimates

DIRECT Estimates of eoseismic

displacement & rates of fault motion ± ur

ij Estimated

Repeat Time r ij

177

A. For prehistoric earthquakes, U i is a measure of 14C dating uncertainties and other errors if known. B. For direet estimates, U i is a measure of the uncertainties in coseismie displacement (u I) and rates of fault motion (u2 ), where Ui = J ui + u~ . C. In T = In(Tave + JlD' and Texp = Te <PD + UM /2), where JlD = 0.0099, UD = 0.215, and UM = Ju~ + U~. See text for additional details.

1974). This approach to hazards forecasting is termed time-dependent, and is fundamentally different [rom approaches that assume random or Poisson distributions of earthquake recurrence times (see Figure 2). In the latter case, the Poisson model leads to estimates of ~onditional probability that are independent of the amount of time elapsed since the last event. In both eases, averaging reeurrenee information over a geographie and seismieally diverse region usually results in a redueed ability to differentiate distinct segments, eaeh of whieh may have their own eharaeteristie reeurrenee behavior.

In an attempt to rigorously deseribe the time-dependent recurrenee behavior of individual fault segments, NISHENKO and BULAND (1987) used the ratio T/Tave for a number of segments along simple plate boundaries with weil known reeurrenee histories. For eaeh individual fault segment, Tave is the observed average reeurrenee interval and T is the observed individual reeurrenee interval. By utilizing apriori information about the underlying distribution of earthquake recurrence intervals, more accurate descriptions of recurrence behavior are possible than using a few available data alone.


,-..... Valparaiso, Chile Earthquake Forecasts rtl .... co Q; :>-. 1.0 Tave 86 yrs

0 1730 C\l , 1822

11 , 1906 ...., 0.8 /1 1985 "0 I I

Ms 8.3 Ms 7.8 ........, , , I I :>-. I I ~ 0.6 , I ......

I I .!:l co , I

.!:l 0.4 I I 0 I I .... I I 0..

0.2 co c 0 .....

0.0 ...., "0 1700 1800 1900 2000 c 0 u Year

Figure 2 Comparison of time-dependent and time-independent conditional probability estimates. Curves compare the lognormal (time-dependent) and Poisson (time-independent) conditional probability estimates for the Valparaiso, Chile seismic zone. Time-dependent models, which account for the regularity in characteristic earthquake occurrence, are low immediately following a large or great earthquake and grow as a function of the time elapsed since the last event. In contrast, Poisson models, which assume random earthquake recurrence behavior, are constant as a function of the time elapsed and are

time-independent.

Probability density functions for earthquake recurrence time, f(I), where 1 is the time elapsed since the last characteristic earthquake, form the basis of the timedependent probability approach. These density functions and their associated measures of variability-the standard deviation, G, and the coefficient of variation, G /mean, define the degree of temporal resolution and hence, the information content of the recurrence interval data. An important result of the TjTave analysis of NISHENKO and BULAND (1987) is that the standard deviation of the lognormal distribution appears to be a fixed fraction of the recurrence interval. The normalized coefficient of variation of the lognormal recurrence distribution, GD = 0.215 and is approximately constant over a wide range of recurrence times and seismic moments in a variety of tectonic environments. It is this property, when used in a normalized time frame, that permits a "generic" description of the underlying probability density function for earthquake recurrence studies and calculations of conditional probability. The marginal time-dependent probability density function, f(T), defined by NISHENKO and BULAND (1987), BULAND and NISHENKO (1988), WORKING GROUP ON CALIFORNIA EARTHQUAKE PROBABILITIES (1988, 1990),


and used in this study is

1 /(T) = e -(ln(T/ T)-/lD)2/2"it (1) T(JM fo

where In t = In(Tave + Jl.o). The mean and standard deviation of /(T) are Jl.D = - 0.0099 and (J M = J (J1 + (J~, respectively. The standard deviation, (J D, defines the intrinsic variability of recurrence intervals based on the global analysis of NISHENKO and BULAND (1987) discussed above and is equal to 0.215, the second standard deviation, (JB, describes the relative uncertainty in our estimates of the median recurrence interval. The complete joint probability density function, which describes the variability in the actual recurrence time, and our estimates of the median recurrence time is shown in Fig. 3.

For historie earthquakes, values of Tave , t and (JB are estimated from observations of T, the individual recurrence times. How weil the median recurrence time can be estimated is described by (JB, and depends on both the number and quality

2

o

Figure 3 Three-dimensional representation of the lognormal joint probability density function. Joint probability density amplitude is shown as a function of the variation in T/T (the ratio of individual recurrence time for a fault segment, T, to our best estimate of the median recurrence time, t) on the X-axis and the variation of 1fT (the ratio of a realization of ihe median recurrence interval, t, to our best estimate of the median recurrence interval, t) on the Y -axis. The coefficient of variation for the random variable In

(I/T) is 0.2 (from BULANO and NISHENKO, 1989).


of observations of T. In other words,

(2)

In equation (2), the u~ In term, where the subscript i denotes an individual recurrenee interval, accounts for the uneertainty in measuring a reeurrence time. For historieal data, there is generally little or no uneertainty as to when the events in question occurred and this term is set equal to O. Henee, UB is only a funetion of the number of historieal recurrence intervals observed and the intrinsie variability of those intervals. For geologieally based estimates, the U j ITj term is included in equation (2) to aeeount for the additional uneertainties in the geologie data (i.e., the ± lu values for the 14C dates). Henee the resolution in estimating recurrence times from geologie data is a funetion of both the errors in dating previous events and the intrinsie variability of the reeurrenee intervals, the latter being independent of age-dating uneertainties. For direct estimates of reeurrence time, the u;/Tj term in equation (2) is used to aeeount for the ± lu uneertainties in the rate of fault motion and eoseismie displaeement, and the direet recurrence time estimate itself is set equal to T.

The expeeted or mean reeurrenee time, Texp , is based on a number of observations of T for a specifie fault segment and is given by,

T = Te(I'D+<1~/2) exp oll • (3)

The predieted date of oeeurrenee of the next event (Tpred ) is simply the sum of the date of the last earthquake (to) with the estimated recurrenee time of the next earthquake (Texp ), (i.e., Tpred = 10 + Texp ). Tbe above recurrence time estimate ean also be presented in terms of a foreeast or predietion time window. Following NISHENKO and BULAND (1987) and BULAND and NISHENKO (1988), the predietion time window is defined by substituting f(T I Ic> for f(T) in equation (I), and ealeulating the 0.05 and 0.95 pereentage points for the new distribution, given that the event has not occurred by time Ic (i.e., the eurrent time).

In addition to estimating the reeurrence time for a specifie fault segment, we ean also estimate the eonditional probability or the likelihood for the event to occur in a given time interval 1 + M. This probability is eomputed from the ratio of areas under the lognormal distribution in equation (I) for two different time intervals and is of the form

p _ f2~,':f(TIT) dT

c - i: ,!(T/T) dT

(4)

and is eonditional on knowing both the expected recurrence time, and that the event has not happened by time Ic (see Figure 4). By implieitly ineorporating the

Vol. 135, 1991

I-

z o ;::: ü Z :::l LI..

~ Cf) z w Cl

~ :::; äi ~ CD o a: a..

Circum-Pacific Seismic Potential 181

t t..n T

TIME

Figure 4 Conditional probability for earthquake occurrence. The probability for an earthquake in the interval T, T + t1T, is given by the area of dark shading under the probability density curve. The probability, conditional on the earthquake not having occurred prior to T, is the ratio of the area of dark shading

to the sum of the areas with dark and light shading.

uneertamtles in t (i.e., O"B) into the probability distribution, the funetions in equations (l) and (4) beeome more broadly distributed with inereasing uncertainty in the data and dampen ehanges of probability as a function of time. For those eases where T ~ t and 0" M is greater than about 0.3, eonditional probabilities will begin to deerease as a funetion of time (COHEN, 1988; DAVIS et al., 1989). This is generally not a serious problem unless t is so poorly eonstrained that any forecast is highly suspect to begin with. The majority of forecasts in this study are based on the historie reeord, and have 0" M values that are less than 0.3. Overall, conditional probability values are dependent on the above uncertainties, and the width of the time window, M, chosen for an individual forecast. In other words, for a given seismic gap, the conditional probabilities for a 20-year window are always larger than those for a 10- or 5-year window. Hence the comparison of probability estimates among various gaps must be done using the same time interval.

The following sections deseribe the different types of basic data sets used in estimating the segmentation, recurrence times, and probabilities for the simple plate boundaries of the eireum-Pacifie region.

Historical / Instrumental Observations

Initial investigations of the seismicity and teetonie setting of the eireum-Paeifie region relied on the aftershoek distributions of prior large and great earthquakes and historieal descriptions of damage and feIt intensities to define rupture zones and segmentation along the strike of simple plate boundaries (FEDoTov, 1965; MOGI,

182 Stuart P. Nishenko PAGEOPH.

1968; KELLEHER et al., 1973; SYKES, 1971; MCCANN et al., 1979; NISHENKO and MCCANN, 1981). The above earthquake data provide the basic seismologieal foundation for defining the recurrenee behavior of individual fault segments in this analysis and will be diseussed in more detail in later sections. Overall, the limited number of available observations make it diffieult to define the recurrence eharaeteristies for many of the segments around the eireum-Pacifie. In fact, departures from idealized eharaeteristie earthquake behavior have been noted for some seismie zones in the eireum-Paeifie (THATCHER, 1990).

For individual fault segments, the penalty for few recurrenee observations is a large uneertainty in the estimated reeurrence time and associated eonditional probability, as indieated by the 90% eonfidenee limits of Texp and the eorresponding range of probabilities for a partieular exposure time. For one recurrence interval, the standard deviation, ii is taken to be 21 % of T. As diseussed in the previous section, and shown in equation (2), more recurrenee observations ean reduee this uneertainty. Overall, the uneertainties in using historie data are smaller than those in other data sets (i.e., geologie and direct estimates, see following seetions).

Geologie Investigations

Radiometrie dating of geologie features produced or altered by eoseismie displacements in previous great earthquakes have greatly extended the recurrence history along seetions of the cireum-Pacifie seismie zone. Unfortunately, these features have only been identified and studied in a few areas. In Alaska, studies of offset Holoeene moraines along the Fairweather fault (PLAFKER et al., 1978) and uplifted, wave-eut marine terraces on Middleton Island and the Cape YakatagaYakutat Bay region (PLAFKER, 1986) have helped eonstrain the rates of plate motion and the size and timing of prehistorie great earthquakes in this region. Radiometrie dating ofuplifted eoral heads on the Vanuatu Islands (TAYLOR et al., 1990) has extended the seismie his tory for this region of the southwest Pacifie. In the northeast Pacifie, studies of subsided eoastal terrains along the eoast of Washington and Oregon (ATWATER, 1987; GRANT et al., 1989) have begun to doeument the potential for a large or great earthquake in that region. The offset of sediments during large and great earthquakes on the San Andreas fault in southem Califomia (SIEH et al., 1989) and the Chixoy-Poloehie fault in Guatemala (SCHWARTZ, 1985) have been invaluable in doeumenting the behavior of major transform faults within the study area.

By taking into account the ± lu uneertainties in the radiometrie dates for individual earthquakes, as weil as the intrinsie variability of recurrence intervals, geologie data ean be included with and eompared to recurrence estimates based on historie data or direct ealeulations as shown in equation (2) (see also NISHENKO and BULAND [1987] for further details and examples).


Direct Estimates

For these fault segments laeking either a historie or geologie record of prior events, we have estimated reeurrenee times based on the size of the most reeent large or great earthquake along that fault segment. The prineipal assumption in this type of calculation is that the most recent large or great event refteets the typicalor characteristie size of earthquakes in a given seismic gap from eycle to eycle. Variations in the size of "gap-filling" earthquakes from eycle to cycle have been noted by various workers (see THATCHER, 1990) and are place limitations on the spatial resolution of the magnitude foreeasts presented here. Direct estimates of recurrence time are determined by dividing the best (median) estimate of eoseismie displaeement, or amount of fault slip, of the most reeent earthquake by the best (median) estimate of the rate of fault motion and assume that no aseismic slip is oeeurring. These direct estimates of reeurrenee time assume the time-predictable model (REID, 1910; SHIMAZAKI and NAKATA, 1980). For some events, only the instrumentally measured long-period seismic moment (Mo) is known. For older, pre-instrumental earthquakes, we ean estimate the seismic moment by using observations of tsunami wave heights (see Figure 5). In both cases, average eoseismie displacements are ealculated using estimates of the rupture area from aftershoek of feit area studies, where U, the displacement = Mo/p.A, Jl. is the shear modulus and A is the fault area. For strike-slip faults, observations of eoseismie surfaee offsets can be compared to estimates of displacement based on the seismic moment. In both eases, the degree of resolution in our foreeasts of reeurrence time and magnitude is dependent on knowing the distribution of coseismic displaeement along the fault surfaee. Within a single rupture zone, sub segments that have smaller amounts of displaeement may repeat sooner or more often than those segments with greater amounts of displacement. In so me cases, detailed investigations have outlined heterogeneous slip distributions, so reeurrenee estimates ean also be attempted for subsections of larger ruptures, as weil.

Uncertainties in the rate of plate motion and the amount of coseismie displaeement for a given fault segment ean reduee the reliability of direct recurrence time estimates. Overall, the uncertainty in the relative rates of plate motion are at the 3-8% level (MINSTER and JORDAN, 1978; CHASE, 1978). In this study we use a conservative estimate of 10%. What is more uncertain however, is the amount of aseismie slip that may be oeeurring along any one fault segment. All of the direet recurrence time calculations presented assume no aseismic slip, and hence represent minimum repeat time and maximum probability estimates. Where available, the historie reeord provides a eomparison for estimating the amount of aseismic slip. Estimates of coseismic displacement vary by a greater amount than the rates of plate motion, and reftect both the inherent uneertainty in the methods used to measure displaeement as weil as the variability in displaeements along a fault surface. For eaeh segment, we have used a variety of displaeement estimates (where available) to help bracket the reeurrenee time. The variation in displaeement (and

184

Vl ~

Q)

Qi E

II

10

~ 1.0 w I

0.1

0.01

TSUNAMI

Stuart P. Nishenko PAGEOPH,

OBSERVATIONS AT HILO " ~, ~

" ~" ,," 1960 S. CHtLE

,," 1837 b. 6 KAMCHATKA " S. CHILE LOG WH "~ 184 ( htd . 0 .72 LOG Mo' 2 1. 0

" ••• 1868 S. PERU 1906 A~ 0 1877 N. CHILE pO 92

COLOMBIA ,," , 952 ,,;'

1922 ", ",;" C. CHILE b..~ -+ fA----1 0 8 ,-'

1906 1\ -L"'-----l " C. CHILE ~ -r .. ----, ,,'

~ " ",," ~""G5 " ~ 1986 ANOREANOF IS. ~-' 0 ~" HILO " 11 "

1985 C. CHILE ~ 1974 0 10 ,," HAWAII t SOUTH ," 14 b. C. PERU" 7~" AMERICA

.,,' 10 b.!.966 ~ 0 9 19~.S MICHOACANO~ 15 ~,,~ 7;," 130 °lfr9q,s·

20-" COLOMBIA

,"

30 "", ~"

" ~"

CALIBRATION EVENTS

6. SOUTH AMERICA

o AL ASKA - ALEU TI ANS

20-' " -'

" " " -,'

~~

10

MOMENT

o JAPAN ' KURILES

& Mo Bosed On MT Estimote

I>. Mo Bosed On M. Est imote

~ Mo Estimot. (This Study)

100 1000

(X10 27 dyne-cm)

Figure 5 Tsunami observations at Hilo, Hawaii. Tsunami wave heights, recorded at Hilo, Hawaii are plotted as a function of seismic moment (Mo) for a set of calibration events (open, numbered symbols). Aleast squares fit for the calibration events (Log (Wave Height) =0.72 Log Mo -21.0) and±2t1 limits, is shown as solid and dashed lines, respectively. This calibration allows estimation of the seismic moment of older, nonseismographically recorded earthquakes and provides a rapid estimate of wave heights at

Hilo, Hawaii for recent earthquakes (after NISHENKO, 1985).

to a lesser extent relative plate motion) define the f1;/Tj values in equation (2). These estimates tend to have the largest ranges in probability, reftecting the greater uncertainty in the recurrence time estimates.

Discussion of Individual Regions

The following sections summarize the basic earthquake data for the simple plate boundary segments of the circum-Pacific region covered in this report. Individual

186

s

60°

PACIFIC OCEAN


50°

Conditional Probability 1989-1999

E:::::::::~ 60-100%

_40-60%

~ 20-40%

Figure 6

~0-20% f"'""'l No Historie Record I..!....!....!J of Great Earthquakes

"'i M, ,.~ Incomplet9 Historie Record

Seismic potential for the Chilean seismic zone: 1989- 1999. Patterns portray the level of conditional probability for occurrence of great (Ms 7.7 and larger) earthquakes during the next 10 years, 1989-1999, and range from haehured, 0-20%; eross-haehured, 20-40%, open dots, 40-60%; and fi1led dots, 60- 100%. Small eross pattern denotes those areas with no historie reeord of large or great earthquakes.

Specifie dates and magnitudes refer to areas with ineomplete historie reeords.

segment, the probability for the recurrence of a great earthquake in the next 10 years is at the 11 % level. In spite of these low estimates, however, more work is needed in this region to document the earthquake history for a better understanding of the seismic regime.

C-2. Chilean Archipelago, 46° -55°S. South of the intersection of the Chile


segments or source zones are identified by number, name and either the along strike latitudinal or longitudinal extent. Each section contains a description of the previous earthquake history, estimates of the average repeat time, and the conditional probability for recurrence in the next 10 years (1989-1999). Estimates for additional time windows (Le., 5 and 20 years, 1989-1994 and 1989-2009) can be found in the Appendix.

South America

Chile

The Chilean seismic zone marks the zone of interaction between the South American, Scotia, Antarctic (south of 46°S), and Nazca (north of 46°S) plates along the west coast of South America. The rates of plate convergence range from 2 cm/yr in the south (South America/Antarctic convergence) to 9 cm/yr farther north (South America/Nazca convergence). Variations in the tectonic regime, as evidenced by gaps in active Quarternary volcanism and changes in the attitude of the intermediate-depth seismic zone, are influenced by intersections of bathymetric features on the subducted Nazca plate and differences in the age of subducted sea floor (BARAZANGI and ISACKS, 1976). These various tectonic regimes also influence the size and recurrence times of large and great earthquakes that occur along the Chilean margin. The following section summarizes the basic earthquake data for 10 segments of the Chilean subduction zone, and is an update of NISHENKO (1985). Note that the probability estimates in NISHENKO (1985) were based on a Weibull distribution of recurrence times, while the current estimates are based on a lognormal distribution. Changes in probability distribution functions, as weil as recurrence time estimates for some areas, have resulted in forecast values that differ from earlier studies. These modifications will be discussed in the appropriate subseetions. Estimates of conditional probability for 1989-1999 are shown in Figure 6.

C-l. Tierra dei Fuego, 65° - now. Previous large or great earthquakes that are documented for this region inc1ude 2 February 1879 and 17 December 1949 (2 M s 7.7 events within 8 hours). Both the 1879 and 1949 events produced Modified Mercalli (MM) intensities of VII in Punta Arenas. While details are lacking for both events, it is assumed that both represent rupture along the Magellan fault system (the transform plate boundary between the South American and Scotian plates). Field studies of the 1949 events indicate rupture extended from the western end of the Brunswick Peninsula (near Punta Arenas) to Policarpo (Tierra deI Fuego), and suggest an overall fault length of about 450 km (WINSLOW, 1982). Based on these length estimates, the combined 1949 events rank as one of the largest strike-slip earthquakes of the 20th century.

If the 70-year recurrence indicated by the historie record is characteristic for this

188

NAZ - SAM

RELATIVE CONVERGENCE

9 CM/VR

Stuart P. Nishenko

1920

!....--

PAGEOPH,

CONSTITUCION

---r- -1 TALCA

1 1 1

1

1939*:

1 1 _CHLLAN

---C» W

1962 Oll •

1 1 1

_1-

I -=' (»

W

""

-0 0-0

400~ ____ ~ ________ ~~-& ______ ~ ______ ~ __ -& ____ ~

75 0

Figure 7 Reeent and historie earthquakes in the Coneepcion-Valdivia, Chile area. Rupture lengths for great earthquakes in this area (1835, 1837, and 1960) are solid where known and dashed where inferred. Cross-haehured areas of the eoast are areas of observed eoastal uplift associated with the 1835 Concepcion earthquake and include Mocha Island, Santa Maria Island, and the TalcahuanoConcepcion area. Larger haehuring is the zone of uplift associated with the 21 May 1960 earthquake (after PLAFKER and SAVAGE, 1970). Overlap ofuplifted areas suggests that the 21 May 1960 'foreshock' to the great 22 May 1960 earthquake may have been a repeat of the 1835 earthquake in this area (after NISHENKO, 1985). Coastal outline and bathymetry (contour interval 1000 m) after PRINCE er al. (1980).

Projection of the Mocha and Valdivia fracture zones from HERRON (1981).


Ridge with the South American plate at 46°S, the Antarctic plate is being subducted beneath the western coast of South America at about 2 cm/yr. This region is characterized by a sediment filled trench, a low level of seismicity, and no history of large or great earthquakes. In view of the low convergence rate, repeat times for large and great earthquakes may be a few hundred years. At present, no data are available to estimate a recurrence time or conditional probability for this segment.

C-3. Southern Chile, 400 -46°S. Previous great earthquakes that have ruptured this segment of the Chilean margin include: 16 December 1575; 24 December 1737; 7 November 1837 (M, 9.2); and 22 May 1960 (Mw 9.4). The 1960 event presently ranks as the largest earthquake in the instrumental record (KANAMORI, 1977a; CIFUENTES, 1989). All of the events in this segment are estimated to have rupture lengths between 700 and 1000 km long. Based on this historie record, the average repeat time for this region is 128 ± 16 years, and the probability for recurrence over the next 10 years is considered negligible (i.e., ~ 1%) due to the recency of the last event.

C-4. Concepcion, 35° -400 S. Previous great earthquakes located in this zone include: 8 February 1570; 15 March 1657; 25 May 1751; 20 February 1835; I December 1928 (Ms 7.9); 25 January 1939 (Ms 8.3); and 21 May 1960 (Ms 7.9). Comparison of 20th century earthquakes with those in preceding centuries suggests a variable mode of rupture for this region. In other words, previous events appear to have ruptured the entire segment at one time, while the most recent episode which began in either 1928 or 1939 took 20+ years to complete. The 21 May 1960 event is included in this zone rather than zone C-3, based on similarities with observations of the 1835 event. Both events produced coastal uplift on the Arauco Peninsula (see Figure 7). Mocha Island however, was uplifted during the 1835 event and the 22 May 1960 sequence and not the 21 May event (PLAFKER and SAVAGE, 1970). Previously, the 21 May event had been identified as a 'foreshock' ofthe great 22 May 1960 earthquake (zone C-3) (see CIFUENTES, 1989) and not included in the calculations of NISHENKO (1985) for this segment. Consideration of the closely coupled behavior of both zones (which historically have ruptured within less than a few years of one another), and the above noted spatial similarity in coseismic uplift suggests that the rupture of the Concepcion segment was completed on 21 May 1960. Accordingly, the average repeat time for this segment is about 95 ± 10 years, and the conditional probability for the recurrence of a great earthquake in the next 10 years appears to be at the 3% level. Inclusion ofthe 1960 event does not change the estimate of average repeat time substantially (92 vs 95 years); however, changing the date of the last event from 1928/1939 (NISHENKO, 1985) to 1960 does reduce the seismic potential estimate from greater than 70% (NISHENKO, 1985) to 12% (this study) for a 20-year window.

C-5 and 5a. Valparaiso and Pichilemu-Llico, 32°-35°S. Previous earthquakes in this segment occurred on: 13. May 1647; 8 July 1730; 19 November 1822; 16 August 1906 (Ms 8.3); and 3 March 1985 (Ms 7.8). Prior to 1985, NISHENKO


(1985) had identified the Valparaiso region as the most likely site along the Chilean margin for a great earthquake in the next 20 years (1984-2004). This forecast was confirmed by the 3 March 1985 event (CHOY and DEWEY, 1988). The average repeat time for great earthquakes along this segment of the margin is 85 ± 9 years. The small standard deviation of these repeat times (± 9 years or ± ll %) is shorter than the global compilation ( ± 21 %, NISHENKO and BULAND, 1987) and is surprising in light of new evidence that the rupture lengths of. these events have varied by a factor of 2 to 3 over the last 300 + years (Compte et al., 1986). A comparison of the 1906 and 1985 seismic moments indicates that the latter is only 17% of the former.

Prior to the 1985 earthquake, a compressional outer rise event occurred on 16 October 1981 (Ms 7.2) and has been interpreted as an intermediate-term precursor to the 1985 event by CHRISTENSEN and RUFF (1983, 1989). Both the 1730 and 1985 earthquakes were also preceded by aseries of pronounced foreshocks. The 1985 foreshock sequence is located offshore, near the postulated intersection to the Juan Fernandez Islands with the Chile trench and a sharp offset in the depth to basement in the Chile trench (see Figure 8). If the ocurrence of foreshocks in this particular area is a characteristic phenomena prior to great earthquakes in this segment, their future identification will be vital for short-term earthquake prediction.

While the probability for the Valparaiso region appears to be presently low, due to the recency of faulting (assuming that the 1985 is the gap filling event), the 1985 event did not fill in the same area as the 1906 rupture (see Figure 9). The region

Ju81'1 Fernandez Is_

4 ____________ ~_/\-------------~

~6 I Ia.. w o

8 _1985 ... 1973"":1971

-_-1906

Figure 8 Comparison of basement relief along the Chile trench and earthquake rupture zones in the Valparaiso, Chile area. Maximum depth to basement along the Chile trench is shown by solid line, stippled area represents depth of sediment fiJl. Dashed line at top of figure represents the regional depth of the Nazca plate at a distance of 300 km from the trench axis (from SCHWELLER et al., 1981). Stars are earthquake epicenters and horizonal bars depict length of rupture for large and great earthquakes during this century. Note that aJl epicenters appear to cluster near the trench axis high at 33°S which is inferred to be related to the intersection of the Juan Femandez Islands. Foreshocks to the 3 March 1985 earthquake were also located in this region. This spatial coincidence suggests that this tectonic feature, or its downdip extension, has an important infiuence on large and great earthquake occurrence in this area.


31 0r---------~~--~----_,--~~~~~~~------, S

"'''' ~ \ 0

RELATIVE CONVERGENCE O ( l'v * 0 \ 1

9CM'VR/o /

o 50 100

KM

Figure 9 Recent earthquakes in the Valparaiso, Chile region. Rupture zone of the 3 March 1985, M s 7.8 event shown as shaded region. Note the dose spatial proximity of the 1906, 1973 and 1985 epicenters. The 1981 earthquake is an outer rise event that indicated the existence of compressional stresses in the region prior to the 1985 rupture. Cross-hachured area along the coast is the area of uplift associated with the 1906 earthquake. At present, the region between 34.5" and 35°S labelIed GAP? has been unruptured

since 1906, and may be the site of a future large event in this region (COMPTE et al., 1986).


between 34S and 35°S (segment C-5a, Pichilemu-Llico) has remained unruptured since 1906 and may be the si te of a future large event (COMPTE et al., 1986). The corresponding probability for this subsegment is at the 33% level for the next 10 years, and is in general agreement with the 20-year Weibull estimates published by NISHENKO (1985).

C-6. Coquimbo-Los Vi/os, 300 -32°S. Previous great earthquakes that have occurred along this segment of the margin include: 8 July 1730; 15 August 1880; and 6 April 1943 (Mw 8.2). The 1730 event also ruptured zone C-5. There is no information concerning events between 1730 and 1880. Both the 1880 and 1943 events appear to be similar in size, however, and the 63-year recurrence time appears to be the most reasonable estimate for this zone at this time. Based on these data, the probability for recurrence in the next 10 years is at the 24% level.

C-7. Atacama 26°-300 S. While this area has experienced a number of large events in the 18th, 19th and 20th centuries, only the 30 March 1796 and 10 November 1922 (Ms 8.2-8.5), events appear to be great earthquakes that ruptured all or most of this segment. This region lies within the megetectonic element of the Chile an margin that is characterized by a shallowly dipping intermediate-depth seismic zone and an absence of Quaternary volcanism (BARAZANGI and ISAcKs, 1976). The epicenter of the 1922 event is located about 100 km inland but the earthquake produced a great tsunami (Mt 8.5). Most tsunamigenic earthquakes tend to be located at or near the coast. GUTENBERG (1939) suggests that the tsunami may have been triggered by submarine landslides (similar to the suggested mechanism for the 1946 Unimak Island, Aleutian tsunami [see Alaska-Aleutian section]). LOMNITZ (1970) indicates that the 1796 event may have been similar to the 1922 event, but no tsunami was genera ted by the former. If the 1922 tsunami was caused by landsliding, differences between the two events are diminished, and the time interval 1922-1796 or 126 years is a reasonable recurrence time estimate. For comparison, direct calculation of a recurrence time for the 1922 event, based on its magnitude and estimates of coseismic slip, average 104 years. Both estimates indicate low probabilities (4-11 %) for the recurrence of a great earthquake in the next 10 years.

C-8 Taltal-Copiapo, 25°-27°S. As well as being involved in the great earthquakes that originate in zone C-7, this segment has independently produced a number of large and great earthquakes during the last two centuries. Events in 3 April 1819; 25 May 1851; 5 October 1859; 4 December 1918 (Ms 7.6); 2 August 1946 (Ms 7.1); and 40ctober 1983 (Ms 7.4) all produced localized damage. This region lies within the rather diffuse tectonic boundary between 'normal' subduction to the north, with associated active Quarternary volcanism, and the shallowly dipping intermediate-depth Benioff zone to the south (zone C-7). The boundary between these two tectonic regimes may be coincident with the abrupt change in trench axis depth at 27°S (SCHWELLER et al., 1981). Given that the last large earthquake occurred in 1983, we estimate the probability for a future large event to be at the ~ 1 % level for the next 10 years.


C-9. Paposo 25°-24°S. This small segment, located between the 1966 Taltal and the great 1877 Arica - Antofagasta earthquake, has no prior history of large or great events. The 1877 earthquake (discussed below) may have ruptured into this segment, but no reports are available to confirm this suggestion. Additionally, estimates of tsunami source region (based on tsunami arrival times in MILNE, 1880) for the great 1877 earthquake terminate in the vicinity of Antofagasta. Hence the seismic hazard for this segment of the Chilean margin is not weil understood at this time, though we note that this region may have the potential for participating in a large or great earthquake that may initiate from either the north or the south.

C-IO. Arica-AntoJagasta 19°-24°S. This segment is distinguished by the great (MI 9.0) 9 May 1877 earthquake which, in addition to producing shaking damage along the Chilean coast, also produced a destructive tsunami throughout the entire circum-Pacific basin. We have no information about the previous great earthquakes in this region due to low population density and the la te development of this area at the end of the 19th century. Lacking any further information, the best recurrence estimate we can provide is based on tectonic analog with the Southern Chile segment (Zone C-3) and the adjacent Arica segment in Peru (see Peru discussion). Both comparison with zone C-3 (l28-year repeat time) or direct estimate based on the size of the 1877 event (111 years, NISHENKO, 1985) indicates that the probability for recurrence is at or near the 20% level for the next 10 years. If however, this zone is more analogous to the adjacent 1868 zone in Peru (264-year repeat time), the resultant probabilities are even lower over the next 10 years. The uncertainty in recurrence times for this segment, however, and the potential tsunami threat to the entire circum-Pacific community underscores the need for more research in this segment of the Chilean margin.

Peru

The Peruvian seismic zone marks the boundary between the South American and Nazca plates, and the continuation of the Peru-Chile trench system along the west coast of South America. In addition to having rates of plate convergence similar to those in Chile (9 cm/yr), Peru also exhibits a number of seismo-tectonic features that are similar to those in Chile. The absence of active Quarternary volcanism in Central Peru (2° -15°S) and a shallowly dipping intermediate-depth seismic zone is similar to the Atacama region in central Chile (27°-33°S, BARAZANGI and ISACKS, 1976). The Peruvian margin can be divided into three major segments, based on the patterns of historie seismicity and the presence of major tectonic elements along the convergent margin. These segments are defined by the intersection of the Carnegie Ridge (0° -2°S), the Mendana fracture zone (lOOS), and the Nazca Ridge (l5°S) with the Peru trench. Estimates of conditional probability for 1989-1999 are shown in Figure 10.

P-l. Arica, 19°-16.6°S. The only known prior great tsunamigenic earthquakes

Vol. 135, 1991 Cireum-Paeifie Seismie Potential

~~r-------------~~fA~~~----~r--------------+--~O°

PACIFIC OCEAN


~:::::::::~ 60-100% ~ 0·20% ..... ~

11140-60%

Figure 10

('ö"i'"il No Historie R-u ~ 01 GIRI EaI1hqu ....

193

Seismie potential of the Peru-Ecuador-Colombia seismie zone: 1989-1999. Patterns portray the level of eonditional probability for occurrence of great (Ms 7.7 and larger) earthquakes during the next 10 years, 1989-1999, and range from haehured, 0-20%; eross-haehured, 20-40%; open dots, 40-60%; and filled dots, 60-100%. Small eross pattern denotes those areas with no historie record of great earthquakes.

Specifie dates and magnitudes refer to area with ineomplete historie records.


in this segment occurred on 24 November 1604 and 14 August 1868 (Mt 9.0). The latter event is one of the most widely documented 19th century South American earthquakes. Comparison of intensities in Peru for both events indicates similar rupture zone lengths (approximately 400 km in Peru, see Figure 11). Unfortunately, comparable da ta does not exist in Chile for the 1604 event, and the southem boundary of the 1604 earthquake is uneonstrained. The long reeurrenee interval between these events, 264 years, is enigmatie when eompared to similar sized earthquakes in southem Chile (see previous diseussions for Chi1ean segments C-3 and C-20). Using the single, historie reeurrence time of 264 years, the probability for a Mt 9.0 earthquake appears to be negligible (i.e., 1 %) for the next 10 years. If we use the reeurrenee his tory from southem Chile as an estimate for this segment, the probabilities inerease to the 23% level for the next 10 years.

1868 Rupture Zone

-------- - - --XI

X >.

(/J IX 000 c: Cl>

c: VIII o

ca VII (.) "-Cl> VI ~

'0 V Cl> -'0 IV 0 ~

111

11 0 200 400 600

Distance trom

Figure 11

It Ica

o 0

800

Arica

o 1868

• 1604

, Lima

00

1000 1200

(km)

0

1400

Comparison of feit intensities for the 24 November 1604 and 14 August 1868 Africa, Peru earthquakes. Modified Mercalli intensities in Peru for the 1604 (solid squares) and 1868 (open circles) earthquakes are plotted as a function of distance from Arica, Peru (based on intensity data from SILGADO, 1985). Line at top of figure shows the estimated rupture zone for the 1868 earthquake (dashed where inferred). The similarity in the decay of intensity with distance suggests that these two events ruptured equivalent portions of the plate margin in Peru. Unfortunately, comparable data do not exist in Chile for the 1604

earthquake and the southem boundary of the 1604 earthquake is presently unconstrained.


P-2. Camana, 16.6°-15.8°S. Prior destructive earthquakes that have affected the Camana region of the Peru margin occurred in 1590; 27 March 1725; 10 July 1821; and 6 August 1913 (Ms 7.8). The average repeat time is 108 ± 13 years, and the probability for another great earthquake in this region appears to be at the 13% level for the next 10 years.

P-3. Nazca, 15.8°-14°S. At present, with the exception ofthe 12 May 1664 Ica earthquake, no predecessor is known for the 24 August 1942 Nazca (M w 8.2) earthquake. The 1942 event is located at the intersection of the Nazca ridge with the Peru trench. Preliminary analysis ofbody waves for the 1942 event (S. Beck, personal communication, 1988) indicates this to be a rather complex earthquake and about twice the size of the 1940 Lima event (see section P-5). The 1664 earthquake destroyed Ica and caused great damage in Pisco. Unfortunately, no reports are available southeast of Ica, and we could not directly compare the extent of damage in 1942 and 1664. As discussed in the next section, the 20 October 1687 double earthquake, appears to have ruptured the segment of the plate margin between Lima and PiscojIca. The 23-year interval between 1664 and 1687 appears to be too short an interval to accumulate the necessary strain for a great earthquake in the same segment of plate boundary. Hence we presently ass urne that the 1664 event ruptured the plate boundary east of Ica, ineluding the Nazca Ridge intersection. Accordingly, our best recurrence time estimate for this segment is 278 years (i.e., 1942-1664). In other locations around the circum-Pacific, the collision of bathymetric features at convergent zones is suggested to locally modify the subduction process, resulting in longer than average recurrence times (KELLEHER and MCCANN, 1967). Based on the historic 278-year interval, the probabilities for a great earthquake in the Nazca segment for the immediate future are at the :5; I % level.

P-4a, b. Lima-Pisco and Pisco-Ica, 12°-14°S. This and the adjoining Chimbote- Lima segment of the Peru margin are defined by the intersection of the Mendana fracture zone at lOGS and the Nazca Ridge at 15°S. Both segm\?nts have exhibited consideration variation in the mode of earthquake rupture during the last 250+ years (BECK and NISHENKO, 1990). On 20 October 1687, 2 earthquakes, approximate1y 2 hours apart, caused widespread shaking and tsunami damage from Lima to Ica. Reports and intensity data, however, are not detailed enough to determine which portion of the margin ruptured in which event. Comparison of these events with the more recent 3 October 1974 (M w 8.0) earthquake indicates that the combined effects of the. 1687 events covered a much larger area (see Figure 12). The 1974 event involved the rupture of two dominant asperities, in elose proximity, southwest of Lima (BECK and RUFF, 1989), with the maximum damage located near Lima (ESPINOSA et al., 1975, 1977). While the 1974 aftershock zone of DEWEY and SPENCE (1979) extends to the Pisco area, low intensities and low aftershock activity near Pisco and Ica indicate that this southern area apparently did not rupture in the same manner as in 1687. Based on the high intensities near Ica and Pisco in 1664 and 1687, we believe this area to be a presently unbroken

196

+

Stuart P. Nishenko

SOOw +

3 October 1974 Mw=8.0

+

+

3

9cmlyr

24 August 1942

Figure 12

76° +

20 October 1687

?

XI , \ ,-

\ \ VI VIII \

Pis~o -lca

I VII \

/ - I

- I I

": +

M-8~; 4

, 5 5

PAGEOPH,

SO S

12 °

Comparison of 1687 and 1974 Peru earthquakes. Modified Mercalli intensities for the 20 October 1687 (combined effects of 2 earthquakes, 2 hours apart) and 3 October 1974 earthquakes from EsPINOSA et al. (1975, 1977) and SILGAOO (1985). Aftershock zone and epicenter (star) of the 1974 earthquake from DEWEY and SPENCE (1979). Stippled region inside the aftershock zone represents the area of major seismic moment release in 1974 (from BECK and RUFF, 1989). In general, the combined feit area of both 1687 earthquakes is larger than that of the 1974 event, and suggests that the southem portion near Ica may represent a seismic gap that was not ruptured in 1974 (from BECK and NISHENKO, 1990).

Bathymetry from PRINCE er al. (1980), contour interval: 1000 m.

seismic gap. Previous events in 1687 and 1813 suggest that a 126-year repeat time may be appropriate for this region. Accordingly, probabilities for the southeastern portion of this segment, near Pisco (P-4b), may be as high as 28% für the next 10 years, and this region deserves further attention. The portion of this segment that ruptured in 1974 (segment P-4a) presently has a ~ 1% probability in the next 10 years.


P-5. Chimbote-Lima, 100 -12°S. The largest known earthquake along this segment of the Peru trench occurred on 28 October 1746 (M w 8.4). Since then, no earthquake along this segment of the coast has been as large or has produced the same pattern of damage. Analysis of the sequence of great underthrust events on 24 May 1940 (Mw 8.2) and 17 October 1966 (M w 8.1) have provided insight into the nature of fault rupture along this segment (BECK and RUFF, 1989; BECK and NISHENKO, 1990), and may explain the marked difference in the mode of rupture. Both the 1940 and 1966 events are characterized as relative1y simple events wh ich ruptured discrete asperities northwest of Lima. Comparison of the highest feit intensities for 1746, 1940 and 1966 earthquakes indicate that while the shaking and damage in 1746 is one to two Modified Mercalli values higher than 1940 or 1966, all occupy similar areas along the coast (see Figure 13). Both BECK and RUFF (1989) and BECK and NISHENKO (1990) suggest that the 1746 earthquake may have ruptured both the 1940 and 1966 asperities in a single event. Studies of multiple asperity earthquakes (e.g., see Ecuador-Colombia section) indicate that they are considerably larger, and more destructive, than the sum of single asperity earthquakes, even though they rupture the same segment of plate boundary. Based on the historie record of great earthquakes in 1746, 1828, and 1940-1966, we tentative1y estimate the repeat time to be about 91 years, and estimate the probability for future great earthquakes in this segment to be at the ~ 1-8% level for the next 10 years.

P-6. Chimbote-Guayaquil, 0.5-100 S. This segment of the Peru Trench is defined by the intersection of the Carnegie Ridge and the Mendana fracture zone. In contrast to the other segments of the Peru-Chile Trench, the area between Chimbote and Trujillo is characterized by in frequent large earthquakes and no record of great underthrust earthquakes for the past 300 years (KELLEHER, 1972). Subduction of the massive Carnegie Ridge complex could be responsible for greater than average repeat times in this region. At present we have no da ta with which to quantitatively evaluate the potential of this region.

Ecuador - Colombia

The Ecuador-Colombia seismic zone markes the boundary between the Nazca and South American plates in northwestern South America. The rate of convergence along this margin is about 8 cmjyr. Based on the patterns of seismicity over the last 100 years, the Ecuador-Colombia margin is comprised of 4 major segments. Ouring the last 80 years, 3 out of 4 segments have ruptured in aseries of large and great earthquakes. The mode of rupture, however, was not similar in all cases (KANAMORI and McNALLY, 1982; MENDOZA and OEWEY, 1984; BECK and RUFF, 1984). The great 31 January 1906 earthquake (M w 8.8) ruptured an estimated 500 km length of the Ecuador-Colombia coast. Following this great earthquake, events on 14 May 1942 (Ms 7.9), 19 January 1958 (Ms 7.8), and 12

198 Stuart P. Nishenko

~\ \ + " Trujillo

~ 17 October 1966

80·W +

~ •

Mw=8.1

" .' , \

\ \

~ ~i' VIII\. \,

~1~40 I .'

..() '( ~ Lima I

~"\ 2.1 m VII;

24 May 1940 M-8

4

+

3

VI

\.

\

I

•

Figure 13

•

76° +

29 October 1746

\

\ IX-X

~ima + ~ I

VIII I

24m /

Pisco

·/Ica V I •

+

~ 2

~; 5

5

PAGEOPH,

8 0 s

12 °

• /

VI

•

Comparison of 1746, 1940, and 1966 Peru earthquakes. Modified Mercalli intensities for the 29 October 1746, 24 May 1940, and 17 October 1966 earthquakes from ASKEW and ALGERMISSEN (1985) and SILGADO (1985). Aftershock zone and epicenter (star) of the 1966 earthquake from DEWEY and SPENCE (1979). Stippled regions represent the areas of major seismic moment release in 1966 (from BECK and RUFF, 1989). Hachured area for 1940 earthquake represents the estimated area of principal seismic moment release. MFZ is the Mendana fracture zone. Based on the spatial coincidence of areas of high shaking in 1746, 1940, and 1966, it is proposed that the 1746 earthquake simultaneously ruptured the same areas that subsequently ruptured individually in 1940 and 1966 (from BECK and NISHENKO, 1990).

Bathymetry from PRINCE el al. (1980), contour interval: 1000 m.

December 1979 (Ms 7.7) ruptured subsegments ofthe 1906 zone (see Figure 14). At this writing, the Buenaventura segment (EC-4) is distinguished as having no his tory of large or great coastal events.

Given the variable earthquake history, and the spaise settlement along the coast of Ecuador and Colombia in the 19th century, recurrence time estimates, based on


presently know the least about, and deserves more investigation in the immediate future.

EC-2. Esmeraldas, 1.2°-1. 7°N. As is the ease for the 1942 event, the time between the 1906 and 14 January 1958 event, 52 years, provides the primary reeurrenee da ta for this segment. For eomparison, the average displaeement in 1958,2.3 meters (KANAMORI and McNALLY, 1982), gives a direet reeurrenee time of 29 years; while the maximum displaeement inferred from body wave inversion, 3.75 meters (BECK and RUFF, 1984), gives a direet reeurrenee time of 47 years, whieh is in better agreement with the historie reeord. Using the historie interval of 52 years indieates probabilities over the next 10 years at the 19% level.

EC-3. Tumaco, 1.7°-4°N. With the exeeption ofthe 1906 and 12 Deeember 1979 events, there are known predeeessors for this segment of the Colombian subduetion zone. Direet estimates of reeurrenee time, using the average and peak displaeement in 1979,2.7 and 6 meters (KANAMORI and McNALLY, 1982; BECK and RUFF, 1984), range from 34 to 75 years. The latter estimate is in good agreement with the amount of time elapsed between the 1906 and 1979 events. Using 73 years as the reeurrenee time estimate, the probability for next ten years is at the ~ I % level.

EC-4. Buenaventura, 4° - 7SN. At this writing, this segment of the EeuadorColombia seismic zone is distinguished by not having any history of large or great coastal earthquakes. Hence estimates similar to others in this study cannot be made at this time.

Central America

The Central American seismic zone marks the boundary between the Cocos, Rivera, Caribbean, and North American plates along the Middle America and Mexican trenches. Rates of relative plate convergence between the Caribbean, North American, and Cocos plates range from 6 to 9 cmjyr, and are approximately 2 cmjyr between the Rivera and North American plates. The folIowing discussion for the Panama, Costa Rica, Guatemala, and Mexiean margins are based on work by MENOOZA and NISHENKO (1989), MONTERO (1986, 1989), WHITE and 0-FUENTES (1991), and NISHENKO and SINGH (1987a,b,c). Unfortunately, insufficient data exist presently for a quantitative analysis of seismic potential in EI Salvador and Nicaragua. Estimates of seismic potential for the zones studied are shown in Figure 15 for the time interval 1989-1999.

Panama

The Panama region has been interpreted as a microplate loeated at the junction of the Coeos, Nazca, Caribbean, and South American plates (BOWIN, 1976; PENNINGTON, 1981; MANN and BURKE, 1984; AOAMEK et al. , 1988). Along the

Vol. 135, 1991

82 4 N

2

o

Circum-Pacifie Seismic Potential 199

80 78 W

MS 7.7

", Colombla ........ r _._.

MS 7.9 Ecuador

Figure 14 Comparison of recent earthquakes along the Ecuador-Colombia seismic zone. Relocations of aftershocks occurring within three months of the 14 May 1942, 19 January 1958, and 12 December of 1979 earthquakes are shown by stippled zones (from MENDOZA and DEWEY, 1984). Epicenters are shown by stars. Darker shading highlights areas that generated the majority of seismic moment release in 1958 and 1979, and are suggested to represent the loeation of dominant asperities along the Ecuador - Colombia seismie zone (from BECK and RUFF, 1984). Line labelIed 1906 represents the estimated rupture length of the great 31 January 1906 earthquake (from KELLEHER, 1972). Note that the subsequent series of events in 1942, 1958, and 1979 all reruptured separate portions of the 1906 zone, pointing to a variable

mode of great earthquake occurrence in this region.

the historie reeord, are diffieult. The 4 February 1797 earthquake, suggested by HEATON and HARTZELL (1986) as a possible predeeessor to the 1906 event, is also suggested to have oeeurred in the Andes near Quito (see SILGADO, 1985). Laeking a eomplete historie reeord prior to 1906, ealculations of eoseismie displaeement for the 1942, 1958 and 1979 earthquakes provide the tentative estimates of reeurrenee time. These direet estimates ean be eompared to the time intervals between the 1906 and subsequent 20th eentury earthquakes, assuming that these times are proportional to the eharaeteristie repeat time for eaeh segment.

EC-l. lama, OSS-1.2°N. The interval between the 1906 and 14 May 1942 event, 36 years, provides the only historie repeat time estimate for this segment. Based on this single 36-year estimate, the probability over the next 10 years is at the 66% level. Pooling all of the Eeuador-Colombia repeats results in a longer average reeurrenee time (54 years), however, the 10-year probability is at the 48% level. See Appendix I for the range of single and pooled estimates for this zone. Unfortunately, the segment with the highest potential is the one we

200~

N

9 8

6 4

3 M

-1

1921

Ms ....:

....

I ~~~~

100

(() \~

V~ 2

Con

ditio

nal

Pro

babi

lity

1989

-199

9 ~:::

::;:::~

60

-10

0%

~ 0

-20

%

lIB88B

!!lI8

4" ~"%

,.....

......

., N

o H

isto

rie R

ec:o

rtI

_...-v

u l:..

..!...!.

.l 01

GrN

I E

artI

lqua

l<es

~ 2

Q..4

0%

,ni Mo "~

~tII H

lato

tlc

CA

RIB

BE

AN

S

EA

~.

CO

LO

MB

IA

0° I

1"[

--_

-I

-=:,

J),

Fig

ure

15

Seis

mie

pot

enti

al o

f th

e C

entr

al A

mer

ican

sei

smie

zon

e: 1

989

-19

99.

Pat

tern

s po

rtra

y th

e le

vel

of

eond

itio

nal

prob

abil

ity

for

occu

rren

ee o

f la

rge

and

grea

t (M

s 7.

0 an

d la

rger

) ea

rthq

uake

s du

ring

the

nex

t 10

yea

rs,

1989

-199

9, a

nd r

ange

fro

m h

aehu

red

, 0

-20

%,

eros

s-ha

ehur

ed,

20-

40%

; op

en d

ots,

40

-60

%,

and

fille

d do

ts,

60-

100%

. Sm

all

eros

s pa

Uer

n de

note

s th

ose

area

s w

ith

no h

isto

rie

reeo

rd o

f gr

eat

eart

hqua

kes.

Spe

cifi

e da

tes

and

mag

nitu

des

refe

r to

are

as w

ith i

neom

plet

e hi

stor

ie r

ecor

ds.

-< ~

w ~

~

(') §" 3 ~ n Si

n lr. 3 ;:;. ~ <> a E

N o


north or Caribbean eoast of Panama, geologie evidenee of eompression within the aeeretionary prism (SILVER et al., 1990) and the presenee of a shallow, southerly dipping seismic zone indieate the existenee of a eonvergent margin between Panama and the Caribbean. The laek of both aetive volcanism and seismieity deeper than 70 km, however, indieates that this is not a weil developed subduetion zone and may be related to oroclinal bending (SILVER et al., 1990). The triple-junetion between the Caribbean, Nazea, and South Ameriean plates exist in the eastern Panama-Colombia border region as a diffuse zone of faulting and seismicity. Along the southern eoast of Panama, diffuse left-Iateral faulting and evidenee of oblique eonvergenee between Panama and the Nazea plate has been diseussed by JORDAN (1975) and SILVER et al. (1990). To the west, the boundary between the Nazea and Coeos plates is the Panama fraeture zone; its assoeiated seismicity terminates near the Middle Ameriean Treneh at the Panama-Costa Riea border.

In eontrast to many of the adjaeent eountries in Central and South Ameriea, Panama is distinguished by relatively low levels of large and great earthquake aetivity (see Figure 16). During the 400-year historie reeord, only a few events stand out as eausing widespread damage. One event, 7 September 1882, appears to have originated along the north eoast of Panama (segment P-I). In addition to damaging the Panama-Colon railway line, this event produeed a tsunami that washed away a number of towns along the Ban Blas Arehipelago (see Figure 17 and MENDOZA and NISHENKO, 1989). No other similar events are known from the historie reeord for this region. The laek of similar earthquakes along the north eoast from 1600 to 1882 may indicate a reeurrenee time of approximately 300 years. Along the south eoast of Panama, a large event on 2 May 1621 eaused MM VII damage in Veraguas provinee and MM VII damage in Panama City. Patterns of damage appear to be similar to those produeed by events in 1845 and I Oetober 1913 (Ms 7.5), suggesting similar teetonie origins. Too !ittle is known, at present, to eonfidently assign reeurrenee time estimates for seismie zones in Panama; however, relative to the north eoast, both the southeast and southwest portions of the eoast have higher rates of seismie aetivity (MUNOZ, 1988a,b).

COSla Rica

The west eoast of Costa Riea ean be divided into 4 primary segments based on the seismie his tory from 1800 to the present (MORALES, 1985; MONTERO, 1986, 1989). These are the Papagayo, Nieoya, Quepos, and Osa segments. A number of signifieant teetonic features are present along the Costa Rican portion of the Middle Ameriean treneh, including ehanges in the dip of the Benioff zone and the subduetion of topographie features on the Coeos plate. Northwest of the Nieoya peninsula the Benioff zone is deeper and has a steeper dip, than to the south beneath the Nieoya and Osa peninsulas (GUENDEL and McNALLY, 1986). The ehange in dip of the Benioff zone southeast of the Nieoya peninsula has been

90

8

5

20

95

19

78

M

s 7

.0

15

CO

CO

S

10

Fig

ure

16

80

W

~~o~

0\ ~o

oo.·

1

98

3

M s

;.;

° \~

... .,

•• ~.

• 5N

... ,

NA

ZCA

Inst

rum

enta

l se

ism

icity

alo

ng t

he C

entr

al A

mer

ican

sei

smic

zon

e (m

odif

ied

from

RIN

EHA

RT

el

al.,

198

2).

Epi

cent

ers

of e

arth

quak

es w

ith M

s gr

eate

r th

an

3.9

from

190

0 th

roug

h 19

79 a

re s

how

n. S

tars

are

tho

se e

vent

s w

ith e

ithe

r M

s gr

eate

r th

an 7

.7 o

r as

soci

ated

with

mor

e th

an 1

0 re

cord

ed d

eath

s. A

ddit

iona

l ev

ents

dis

cuss

ed i

n th

e te

xt a

re a

lso

indi

cate

d. N

OA

M,

CO

CO

S,

NA

ZC

A,

and

eAR

IB r

efer

to

the

Nor

th A

mer

ican

, C

ocos

, N

azca

, an

d C

arib

bean

tec

toni

c pl

ates

. N

ote

the

stri

king

con

tras

t in

sei

smic

ity b

etw

een

Pan

ama

and

the

rest

of

the

Cen

tral

Am

eric

an s

eism

ic z

one.

-< ~ ..., .V

>

~

~

Ci i' ~ ~ W

~.

(i' ä' co a [ ~ ...,

204

8'N

11


eo'W

7 SEPTEMBER 1882 PANAMA EARTHQUAKE

78'

----......... ......... oe.... .......... .......... """"",.................... a.aga"",

cocos 11

Figure 17

, ......... (Na.acandl)

......... -.................. ......... _-

M ••• Hev, da :;{-de Her'll eo

Intensity map for the 7 September 1882 Panama earthquake. Contour lines (dashed where poorly constrained) separate zones of equal Modified Mercalli intensity associated with the 1882 event. Principal tectonic tl:atures presently known for the Panama region are shown in the in set map. Barbed lines represent zones of plate convergence. The intensity data indicate that the star, centered at lOoN, 78°W, is the most probable location of the 1882 event, and suggest that this event may be related to convergence between the Caribbean plate and the Panama micropiate (from MENDOZA and NISHENKO,

1989).

suggested as representing a contortion in the subducted Cocos plate. Farther to the southeast, the Osa peninsula is the site of a ridge-trench collision between the Cocos ridge and the Middle American Trench (ADAMEK et al., 1987).

CR-I. Papagayo, 87°-86°W. Only two earthquakes are associated with the Papagayo segment by MONTERO (1986, 1989), 18 February 1840 and 27 February 1916 (Ms 7.5). There are, however, uncertainties as to the exact location ofthe 1840 event. If the 1840 earthquake is the predecessor to the 1916 event, we estimate the probability for a similar event in the next 10 years to be at the 31 % level.

CR-2. Nicoya, 86°-85°W. The Nicoya region is one of the most active seismic zones along the Costa Rican margin and has experienced large events on 3 April 1827; 8 September 1853; 9 December 1863; 21 June 1900 (Ms 7.2); 24 April 1916


(Ms 7.4); 21 December 1939 (Ms 7.3); 10 May 1950 (Ms 7.7); and 23 August 1978 (two M s 7.0 events).

If the sequence of large (Ms 7.0-7.4) earthquakes is representative of the characteristic strain release in this segment, the average repeat time is 22 ± 2 years, and the probability for continued activity in the Nicoya region over the next 10 years is at the 64% level. The larger 1950 earthquake may represent a more complex event, or a c1ass of earthquakes with a longer recurrence cyc1e that has not been weIl sampled yet.

CR-3. Quepos, 85°-84°W. The only large earthquakes associated with this segment are the events of 3 March 1882 and 9 September 1952 (Ms 7.0). Based on this one recurrence time, 70 years, we estimate the probability for another event in the next 10 years to be at the 8% level.

CR-4. Osa, 84°-83°W. The Osa peninsula marks the site where the aseismic Cocos ridge is subducting beneath the Caribbean plate. Large events occurring in the Osa segment inc1ude 8 May 1822; 4 August 1854; 20 December 1904 (Ms 7.2); 5 December 1941 (Ms 7.5); and 3 April 1983 (Ms 7.3). The 1983 event occurred in a seismic gap identified by KELLEHER et al. (1973), MCCANN et al. (1979), and ASTIz and KANAMORI (1984). Both the 1941 and 1983 events were studied in detail by ADAMEK et al. (1987) who showed that while they were similar, the 1941 event exhibited more complexity in the source time functions and had a slower rupture initiation. Based on these data, the average repeat time for large events in the Osa peninsula segment is 40 ± 4 years, and we estimate the probability for recurrence to be at the :s; I % level.

Nicaragua and EI Salvador

The Pacific co ast of central EI Salvador and Nicaragua can be divided into three primary segments based on the occurrence of large earthquakes between 1898 and 1926. These segments are Central EI Salvador, ES-I (28 February 1926, M s 7.2); Western Nicaragua, N-l (28 March 1921, M s 7.4); and Eastern Nicaragua, N-2 (29 November 1898 M s 7.0 and 30 December 1907 M s 7.2). Unfortunately, few reports exist for 19th century and earlier earthquakes in this area. Those that do exist do not provide enough spatial coverage to es ti mate earthquake location or rupture extent. Hence recurrence estimates similar to others made in this study are difficult at this time.

We do note, however, that the time elapsed since the previous episode of large plate boundary earthquakes, 63-91 years, is equal to or longer than average recurrence times for adjacent segments of the Central American trench. All three segments have been identified as seismic gaps by KELLEHER et al. (1973), MCCANN et al. (1979), and ASTIZ and KANAMORI (1984). This region should be regarded as having a significant but unknown seismic potential. Clearly, more work needs to be


done to better eonstrain the hazard from large plate boundary earthquakes in this region.

Guatemala

The two principal seismie regions in Guatemala that are eovered by this study are I) the strike-slip Chixoy-Poloehie-Motagua fault system that defines the North Ameriean-Caribbean plate boundary in eentral and northern Guatemala, and 2) the eonvergent boundary between the Caribbean and Coeos plates along the west eoast of Guatemala and the Middle or Central Ameriean treneh. Henee, within Guatemala and neighboring Chiapas, Mexieo there exists a diffuse tripie junetion between the North American, Caribbean and Coeos plates. The following earthquake hazards estimates are based on analysis by SCHWARTZ (1985), SCHW ARTZ et al. (1979), WHITE (1984, 1985), and WHITE and CIFUENTES (1991).

Transform Faults: The Chixoy-Poloehie and Motagua faults represent the prineipal strike-slip struetures within a broad zone of deformation whieh defines the western portion of the Caribbean - North Ameriean plate boundary in Central Ameriea. While these faults have been the most aetive historieally, there are insuffieient data to evaluate the potential of other faults within this diffuse zone of deformation. During the time interval 1538 to 1983, WHITE (1984) has doeumented 2 or possibly 3 aetive periods for the eombined Chixoy-Poloehie-Motagua fault system. These aetive periods, whieh typically are terminated by large earthquakes on either the Chixoy-Poloehie or Motagua fault systems, have had durations of 50-100 years and are separated by quiet periods of about 120 years duration.

G-l. Motagua Fault, 88.3°-91 0 W. Geologie investigations, following the 4 February 1976 (Mw 7.5) Motagua earthquake, indieated that the previous large event occurred around 1280 A.D. (SCHWARTZ, 1985; SCHWARTZ et al., 1979). Given that the last-large event was in 1976, the probabilities for a similar event on the Motagua fault in the near future are negligible (Le., s; I %). In addition, WHITE (1984) suggests that the 1976 event has probably terminated the eurrent aetive period in this region which began in 1945.

G-2,3. Chixoy-Polochic Fault, 89°-91,50 W. Previous events that are associated with the eastern portion of the Chixoy-Poloehie fault, based on both geologie and historie investigation, occurred in 600 A.D., 938, 1538 (?), and 1785 (Ms 7.3?). The average recurrence time is about 250 years, and the estimated probability for the eastern Chixoy-Poloehie segment is at the 8% level. For the western segment of the Chixoy-Poloehic fault, only one event in 1816 is presently doeumented (WHITE, 1985). Given the temporal eoupling of earthquakes on the eastern and western segments in the last eycle (1785-1816), we qualitatively estimate that the ehanees of a future large event on the western portion of the Chixoy-Poloehie are similar to those for the eastern segment.


Convergent margin. The convergent margin of Guatemala can be subdivided into 3 principal segments, based on the historic analysis of WHITE and CIFUENTES (1991), which covers the time interval 1526 to 1987. Rupture zones for earlier events are determined from the location of the Modified Mercalli (MM) VII intensity contour. These 3 segments are termed the northwestern, central and southeastern portions.

G-4. Southeastern, 89° -90.5°W. This segment also includes the northwestern portion of EI Salvador and is located in the area between Guatemala City and San Salvador, EI Salvador. Additionally, this area contains the most consistent or regular history of great earthquakes in the catalog of WHITE and CIFUENTES (1991). Previous events occurred on 1575, 1658, 5 March 1719, 30 March and 30 May 1776, 19 December 1862, and 6 September 1915 (Ms 7.7). Based on these events, the average repeat time is 68 ± 6 years, and the probability for a future event is at the 51 % level for the next 10 years. Presently this segment stands out as having the highest hazard along the Guatemala portion of the Middle American Trench. Analysis of the 1915 earthquake indicates that it ruptured the lower portion of the plate interface, between 60 and 80 km, and was not accompanied by a tsunami. Presently, this gap is also associated with a zone of seismic quiescence for small and moderate sized earthquakes since 1963 (WHITE and CIFUENTES, 1991).

G-5. Central, 91S-90SW. Prior events in the central portion of the Guatemala seismic zone include 29 September 1717, 29 July 1773, 19 December 1862,3 September 1874 and 6 August 1942 (Ms 7.9). In contrast to the northwestern segment, there appears to be more regularity in the pattern of earthquake occurrence. The average repeat time is 66 years, and the probabilities for the next 10 years are at the 23 % level.

G-6. Northwestern, 92S-91SW. Prior great earthquakes along the northwestern portion of Guatemala include 24 October 1765, 18 April 1902 (Ms 7.9), and 6 August 1942 (Ms 7.9). The 1902 event was part of a sequence of 3 great earthquakes in western Guatemala and Chiapas, Mexico, from 1902 to 1903 (see also NISHENKO and SINGH, 1987c). The northwestern segment appears to have a much more variable recurrence history than the central or southeastern segments. WHITE and CIFUENTES (1991) have suggested that this marked variability may reflect complex interactions between the convergent and strike-slip fault systems in the vicinity of the Cocos-North American-Caribbean tripie junction. The longest interval between events (1765-1902) occurs during the time span in wh ich the western Chixoy-Polochic fault experienced its only known historical rupture ( 1816). While the average recurrence time is not weil known for this segment, and the nature of its interaction with transform faults is not weil understood at present, we tentatively estimate that the recurrence time is 89 years, and that the probabilities for future great earthquakes throughout this zone during the nex t 10 years are at the 13 % level.

208 Stuart P. Nishenko PAGEOPH.

Mexico

The following sections summarize the basic earthquake data for 14 segments along the Mexican subduction zone and are organized by state within the Republic of Mexico. A more complete analysis is presented in NISHENKO and SINGH (1987a). Place names and a space-time graph illustrating historie rupture zones and the locations of current seismic gaps are shown in Fig. 18.

Chiapas

M-l. Chiapas, 92S-94°W. Previous earthquakes known or thought to have occurred in this region include 1565 or 1591, 1743, 1902 (Ms 7.5, 7.8) and 1903 (Ms 7.7). The average repeat time is approximately 162 years, much longer than elsewhere along the Mexican or Guatemalan margin. At present, the probability for another great event in Chiapas appears to be at the 2% level for the next 10 years.

Oaxaca

M-2. Tehuantepec, 94°-95.2°W. The Tehuantepec gap is one of two segments along the Mexican margin that are presently distinguished as having no historie record of large or great earthquakes (KELLEHER et al., 1973; KELLEHER and MCCANN, 1976; SINGH et al., 1981). The location of this gap is coincident with the intersection of the Tehuantepec Ridge with the Middle American Trench, and is near the tripie junction of the North American, Cocos, and Caribbean plates.

At present, there are no data with which to base a seismic hazard estimate for this segment. One possible candidate for a prior event in this region is the 5 June 1897 M s 7.4 earthquake, which had previously been associated with the eastern Oaxaca segment (SINGH et al., 1981). Based on the dimensions of the Tehuantepec gap (125 km length, 80 km width) this region may be capable of producing an event of M w 8.0 (Mo 1 x 1028 dyne-cm). This region should be regarded as having a poorly known seismic potential, and should be the target of future research to better constrain both the earthquake history and the hazard level.

M-3. Eastern Oaxaca, 95.2°-96.4°W. Previous large earthquakes that are known to have occurred within this segment include: 22 March 1928 (Ms 7.7) and 23 August 1965 (Ms 7.8). The observation of similar body waveforms (CHAEL and STEWART, 1982; SINGH et al., 1984), and comparable magnitudes for both events indicate that the characteristic earthquake model may be appropriate in this region. Based on the one available recurrence time, 1965-1928 or 37 years, the corresponding estimates of conditional probability for the next 10 years are at the 35% level.

Vol. 135, 1991

2000

1932

,..., 1900 •

Cireum-Paeifie Seismie Potential 209

22' N

, j

" ....... -MJCHOACAN

, -,

106' W

16"

? ... --.-... 1 " ..

104' W

1985

lD1I1

1011 •

, , , " -, --

..!!!!....

_~!!-l.. _

20' N

NOAM

, ~ .......... __ otj,t

GUERRERO

,

' ... I ~ _~J ~

cocos

OAXACA

102' W l00' W 98' W

~

" " ~ 0

0

~2!!!. .!!!!. 0 JUS

~ 1928 (4 Events)

18' N 92' W

16' N

94' W

12' N

96' W

2000

... --.-----.-191 1

'90 • • •

,-• . '907

• ' ... 1197 1900 ,.80 '.0 . •

IU2' . '.'0 16~"

1845 • .---, 1._. ,,2'0

? _-. __ ~ 1 1800L-______ ~ __ ~~ __ L-__ L_ __ ~ __ ~ ____ L_ __ L_ __ ~ __ L__L __ _L __ __J 1800

104 ' 102 ' 100 ' 98' 96' 94'

Figure 18 (Top) Loeation of Mexiean states, eities, and major bathymetrie features. Bathymetry from CHASE el al. (1970), eontour interval: 1000 m. Solid arrows indicate the direetion and relative rates of eonvergence between the North Ameriean (NOAM), and Coeos (COCOS) or Rivera (RIV) plates [eonvergenee vectors after McNALL y and MINSTER (198 I) and EISSLER and McNALL Y (1984)]. (Bottom) Spacetime diagram of historie and recent earthquakes along the Mexiean subduetion zone, and loeations of eurrent seismie gaps. Large circles represent earthquakes with M s ~ 7.7. Small eircles are everits with M s ~ 7.0. Dashed horizontal Iines indieate the lateral extent of rupture zones sinee 1928, based on

aftershoek studies and are dashed where less well-determined (from NISHENKO and SINGH, 1987a).


M-4. and M-5. Central Oaxaca, 96.4°-97.7°W. Previous large and great earthquakes that have occurred along these segments of the margin include: 11 May 1870 (Ms 7.9); 27 March 1872 (Ms 7.4(?»; 17 June 1928 (Ms 8.0); 9 October 1928 (Ms 7.8); and 29 November 1978 (Ms 7.8). The Central Oaxaca region contains two segments or source zones, one between 96.4° and 97.3°W (Zone M-4) and a second between 97.3 and 97.7°W (Zone M-5).

In general there is good agreement between the earlier intensity data and the instrumental data that the 1870, 17 June 1928 and 1978 earthquakes ruptured the same segment of the plate boundary (Zone M-4). As in the case of the eastern Oaxaca segment (M-3), the observation of simple body waveforms, and comparable magnitudes indicates that the observed 54 ± 8-year average repeat time appears to represent a reliable recurrence estimate for this region. Based on these data, the conditional probability for the next 10 years appears negligible (i.e., ~ 1%).

In contrast, the adjacent segment (97.3°-97.7°W, zone M-5), wh ich last ruptured on 9 October 1928, has not had a repeat since that time and is presently considered to have a high probability for recurrence. The historie record is not c1ear as to what events ruptured this segment prior to 1928, or whether this segment is capable of rupturing independently. Hence, well constrained recurrence estimates, similar to others in this study, are not possible at this time. Based on the size of this gap (55 km length, 80 km width), the estimated seismic moment of an event that would fill this region is approximately 2 x 102 dyne-cm (equivalent to M s 7.7, or approximately the size of the 9 October 1928 earthquake). One possible candidate for a previous event is the 27 March 1872 earthquake. Using the interval 1928-1872, or 56 years, as a recurrence time estimate, the estimated conditional probability is at the 45% level for the next 10 years. For comparison, a recurrence time of 44 years is suggested, based on the sizes and times of events occurring in adjacent segments (M -4 and M -6). While the absolute values are not well-constrained, these estimates are significantly higher than the estimates for the adjacent regions in the state of Oaxaca.

M-6. Western Oaxaca, 97.7°-98.2°W. Previous large earthquakes that have occurred in this segment include 5 May 1854 (Ms 7.7); 2 November 1894 (Ms 7.4);

4 August 1928 (Ms 7.6); and 2 August 1968 (Ms 7.4). As in the ca se of the other Oaxaca segments (M-3, M-4), comparable body waveforms and magnitudes indicate that all events are essentially identical, and that the observed 38 ± 5-year average repeat time is a reliable recurrence time estimate. Accordingly, the probability estimates for the next 10 years are at the 21 % level.

Guerrero

M - 7. Ometepec, 98.2° -99.3°W. Previous large earthquakes that are known or suspected to have occurred in this segment include: 2 December 1890 (Ms 7.5); 15 April 1907 (Ms 7.9); 23 December 1937 (Ms 7.5); 14 December 1950 (Ms 7.3); and


7 June 1982 (Ms 6.9,7.0). This segment has been classified as a seismic gap by SINGH et al. (1981) on the basis of the amount of time elapsed since a prior large earthquake. During this century, the recurrence history along the Ometepec segment has been somewhat anomalous when compared to the adjoining segments in Oaxaca. Earthquakes in Ometepec exhibit both simple and complex modes of rupture, with large variations in both magnitude and interevent times. For example, the 1937 event was followed 13 years later, in an almost identical location, by the 1950 event. NISHENKO and SINGH (l987b) argue that this sequence may not represent a true recurrence, but rather a complex mode of rupture that took 13 years to complete. In contrast, GONZALEZ-RUIZ and McNALLY (1988) suggest that the Ometepec segment is characterized by randomly occurring large events. According to their interpretation, the best recurrence probabilities would be based on a Poisson type model, which gives estimates of 30-40% for the next 10 years. For comparison, the lognormal model gives slightly higher estimates of 47% for the same time interval.

M-8. Acapulco-San Marcos, 99.3°-100oW. Previous large and great earthquakes that are associated with this segment include: 4 May 1820(?) (Ms 7.6); 7 April 1845 (Ms 7.9); 15 April 1907 (Ms 7.9); and 28 July 1957 (Ms 7.7). Estimates of the seismic potential for this segment have been discussed by SINGH et al. (1982) based on the assumption that both the 1907 and 1957 events ruptured the same segment of plate interface. The seismic moment of the 1907 event is approximately twice that of the 1957 earthquake, and based on the above assumption and the time-predictable model, a possible recurrence time of 28 years was suggested. NISHENKO and SINGH (1987b), on the other hand, suggest that the 1907 event may have ruptured both the Acapulco and Ometepec segments (M-7 and M -8). Hence the difference in seismic moment may reftect the fact that the 1957 shock only ruptured half of the 1907 zone.

Based on historie observations of great shocks in the Acapulco-San Marcos region, there are two possible candidates for a predecessor to the 1907 event; 1820 and 1845. The average repeat times, depending on which event actually occurred in this segment, range from 56 to 68 years. In spite of this present uncertainty, the current probability estimates are low for the next 10 years (i.e., 13%).

M-9. Central Guerrero, 100o -101°W. Previous earthquakes of M s ~ 7.5 known or estimated to have occurred in this segment include: 7 April 1845(?) (Ms 7.9); 24 December 1899 (Ms 7.7); 26 March 1908 (Ms 7.8); 30 July 1909 (Ms 7.5); and 16 December 1911 (Ms 7.8). With the exception of the 1845 event, there are no presently well-documented earthquakes for this region prior to 1899, owing to 10w population densities during the 19th century and perhaps longer than average repeat times. Hence, recurrence estimates for this segment are speculative at best.

This segment may be capable of producing a single event of M w 8.0 (Mo 1 x 1028 dyne-cm) based on the physical dimensions of the gap (100 km length, 80 km width). The summation of moment release during the 1899-1911 sequence is


approximateJy 1.5 x 1028 dyne-cm, and suggests that while the exact locations of these earlier events are poorly known, most, probably all, contributed to filling in the entire gap. Hence future activity may involve either a great earthquake with a multiple source mechanism or aseries of large earthquakes, as were observed near the turn of the century (GONZALEZ-RUIZ and McNALLY, 1986).

While no historie recurrence times are available, we can estimate recurrence times by extrapolating the known recurrence behavior in Oaxaca. This comparison suggests recurrence times of approximateJy 60-70 years for M s 7.7 to 7.8 events, and 50 years for M s 7.5 earthquakes. These estimates are in relatively good agreement with a postulated 54-year recurrence time, provided the 1854 event is actually located in the Central Guerrero gap. These estimates are shorter than the amount of elapsed time since the previous series of large and great shocks at the turn of the century and we note that the Central Guerrero gap stands out as having a significantly higher potential (30-40%) than the surrounding regions.

M-IO. Petatlan, 10Io-I01.8°W. Previous shocks in this region include 22 February 1943 (Ms 7.7) and 14 March 1979 (Ms 7.6). In contrast to the earthquakes in Oaxaca, there appears to be considerably more variability in the mode of rupture for this region. Based on the assumption that the interval 1979-1943 represents a recurrence interval for this segment, the conditional probability for the next 10 years is negligible (3%).

Michoacan

M-I1. Michoacan, IOIS-103°W. This segment, like the Tehuantepec gap, is coincident with the intersection of a bathymetric feature, the Orozco fracture zone. KELLEHER and MCCANN (1976) and LEFEVRE and McNALL Y (1985) have suggested that collisions of this sort may locally modify the subduction process, resulting in longer than average repeat times. The occurrence of the great M s 8.1 19 September 1985 earthquake in this gap appears to support the model of infrequent great shocks with longer than average repeat times. An earlier event 7 June 1911 (Ms 7.9) has been relocated in this gap on the basis of locally recorded S-P times and similarities in intensity with the 1985 event (EISSLER et al., 1986). Hence, while still poorly known, a recurrence time of 74 years appears to be appropriate for this segment, and the hazard estimate for the recurrence of a great event within this gap in the next 20 years is small (~ 1%).

M-12. Colima 103°-103.7°W. Previous large earthquakes associated with this segment include: 15 April 1941 (Ms 7.9) and 30 January 1973 (Ms 7.5). Comparison of locally recorded S-P times and feit intensities indicate that the rupture zone of the 1941 event was larger than the 1973 event, and may have extended into the 1985 Michoacan rupture zone (UNAM SEISMOLOGY GROUP, 1986). These variable rupture patterns illustrate the current difficulties in exactly forecasting future earthquake activity. Nevertheless, if we base our forecast on the observed interval 1973-1941, the conditional probability is at the 25% level for the next 10 years.


Colima

M-13. Colima gap, 103.7-104,SDW. The Colima gap is defined on the basis of relocations of aftershocks of the great 3 June 1932 Jalisco earthquake (SINGH et al., 1985) and covers a zone approximately 60 km long between the 1932 and 1973 Colima rupture zones. The Colima gap is approximately coincident with the coastal extension of the Colima graben-the arm of the incipient tri pie junction in Jalisco (LUHR et al., 1985). The lack of any major shocks in this region during the last 80 years may indicate the existence of a modified stress regime in this area. Given this, any further discussion is beyond the scope of this study, with the exception of noting that the Jalisco graben system is also capable of producing large earthquakes (i.e., II February 1875, Guadalaraja (Ms 7.5».

Jalisco

M -14. Jalisco, 104.3° -105.7°W. Previous large and great earthquakes that have occurred in or near the state of Jalisco include: 25 March 1806 (Ms 7.5); 31 May 1818 (Ms 7.7); 20 January and 16 May 1900 (Ms 7.6 and 7.1); and 3 and 18 June 1932 (M s 8.1 and 7.8). The 3 June 1932 earthquake is one of the largest events to have occurred in Mexico during this century. Analysis of intermediate-period body waves, relocations of locally recorded aftershocks and compilation of intensity data (SINGH et al., 1984; SINGH et al., 1985) indicate that this was a complicated event that ruptured a 220 km portion of the Rivera plate. Low population densities along the coast of Jalisco, however, make positive identification of older rupture zones difficult. One or both of the 1806 and 1818 events may have ruptured either the Jalisco or Colima segments. SINGH et al. (1985) estimate a recurrence time of 77 years, based on the long-period seismic moment and rupture areas of the 3 and 18 June 1932 events. This direct estimate is shorter than the historically suggested intervals of 114-126 years, if either of the 1806 or 1818 events did actually rupture this segment of the margin. The corresponding conditional probabilities from both data sets range from 2 to 18% for the next 10 years.

Additionally, we have no information about the recurrence of large (i.e., 7< M s < 7.7) events, like the 20 January 1900 earthquake for this region. Hence, recurrence time and probability estimates for large shocks in the state of Jalisco are not possible at this time.

North America

California

The San Andreas fault system represents the primary locus of plate motion between the North American and Pacific plates in the western United States. The following discussion summarizes the results of the WORKING GROUP ON CALIFORNIA


EARTHQUAKE PROBABILITIES (1988, 1990) and is focused primarily on the San Andreas fault proper. The IO-year forecast (1989-1999) shown in Figure 19, and the 30-year (1988-2018) forecast of the Working Group is shown in Figure 20. Discussion of other faults that comprise the San Andreas fault system (i.e., the San Jacinto, Imperial, and Hayward faults) can be found in the 1988 and 1990 Working Group reports.

Primary segmentation of the San Andreas fault is based on the rupture zones of great through moderate-sized earthquakes that have occurred within the last 300 years and are thought to define characteristic segments along the fault trace. From north to south these primary segments include the 1906 San Francisco rupture, the Central Creeping Zone, the 1966 Parkfield-Cholame, the 1857 Fort Tejon, the 1812(?) San Bernardino, and the 1680 Coachella Valley earthquake ruptures. Within each of these rupture zones, subdivisions have been made, reflecting spatial variations in the amount of observed coseismic displacement associated with a particular rupture. These subdivisions will be discussed within the overall discussion of the primary segmentation.

SA 1-3. Northern San Andreas Fault, 36.8°-40.4°N. The northern San Andreas fault segment is defined by the rupture zone of the great 1906 San Francisco

50' N

40·1-----.-

30'

.~ D:o~~~aa.t 1ia2~1>". 1'~=-MIUM'ic

Figure 19 Seismic potential of the San Andreas fault and Washington-Oregon seismic zone: 1989-1999. Patterns portray the level of eonditional probability for occurrenee of large and great (M s 7.0 and larger) earthquakes during the next \0 years, 1989-1999, and range from haehured, 0-20%; eross-haunehed, 20 - 40%; open dots, 40-60%; and filled dots, 60-100%. Small cross pattern denotes those areas with no historie reeord of large or great earthquakes. Specifie dates and magnitudes refer to areas with

incomplete historie records.

Vol. 135, 1991

LEVEL OF RELIABILITY (with A being mOSI reliable)

• A .B • C

0 0

OE

Circum-Pacific Seismic Potential 215

Figure 20

CONDITIONAL PROBABIlITY OF MAJOR EARTHQUAKES ALONG SEGMENTS OF THE

SAN ANDREAS FAULT 1988-2018

;;!~ a~ E~ Oal ZO Oa: on.

Conditional probability for the occurrence of major earthquakes along the San Andreas fault, in the 30-year interval 1988- 2018) (from WORKING GROUP ON CALIFORNIA EARTHQUAKE PROBABILITIES, 1988). For individual fault segments, the box heights are equal to the 30-year (1988 - 2018) probability

and the pattern within the box refiects the reliability of da ta used to construct the forecast.

earthquake and extends from Punta Gorda south through the San Francisco Peninsula, terminating near San Juan Bautista. Over the 485 km long rupture zone the average amount of slip in 1906 reaches a maximum of 4-7 meters (from Punta Gorda to Olema) and steadily decreases in San Francisco Peninsula region from 3 to less than I meter, before terminating near San Juan Bautista.

The long-term rate of fault motion for the northern (Olema) and San Francisco Peninsula segments (1.6-1.9 cm/yr), and the amounts of displacement in 1906 indicate recurrence times of 228 - 303 and 136-169 years, respectively (see WORKING GROUP reports [1988, 1990] for a more detailed account of the models used to develop these estimates). Hence, probabilities for a great earthquake in the immediate future (i.e., the next 10 years) appear low (i.e., ~ 10%) for segment SA-1. For the southern terminus of the 1906 break, however, estimates are more equivocal, and are highly dependent on interpretation of available data (see SCHOLZ, 1985;


SYKES and NISHENKO, 1984; THATCHER and LISOWSKI, 1987). Vsing coseismic displacement values of 2.5 and 1.5 meters, and a slip rate of 1.6 cmJyr, the return time estimated in 1988 for the Peninsula segment (SA-2) range between 156 and 94 years respectively, and corresponding lO-year probabilities range between 5% and 15%. In 1988 this segment stood out as having the highest potential along the 1906 break for a large (i.e., M s ~ 7) earthquake. The southern Santa Cruz segment (SA-3) also bad a moderate potential (9% in 10 years) for the occurrence of a moderate sized event (Ms 6.5). On 17 October 1989, the M s 7.1 Loma Prieta earthquake overlapped portions of both the southern Santa Cruz Mountains and San Francisco Peninsula segments (V.S. GEOLOGICAL SURVEY STAFF, 1990). The complex nature of faulting during this event and the lack of surface offset on the San Andreas fault itself does not lend itself, at this time, to an unambiguous assertion that the 1989 event does, in fact, represent a rerupture of the 1906 break in this area and whether the probability has decreased along this segment.

SA-4. Central Creeping Zone, 36°-36.8°N. The 130-km long Central Creeping segment has exhibited continuous or quasi-continuous slippage, with a rate of creep that is similar to that of the long-term geologic rate of fault motion. Since all or the majority of fault motion is being accounted for by creep, this segment does not appear to be accumulating strain for release in a future large or great earthquake.

SA-5. Parkfield-Cholame, 36°-35.7°N. South of the central creeping zone is the Parkfield-Cbolame segment which historically has ruptured in moderate (Ms 6.0) earthquakes about every 22 ± 2 years. The last event occurred in 1966, and the next event is expected in a 5-year window centered on 1988. Currently this segment is the site of the Parkfield Earthquake Prediction Experiment (BAKUN and LINDH, 1985), and has a high probability (i.e., 93%) for recurrence within the next 10 years.

SA 6-8. Central San Andreas Fault, 35.7°-34.3°N. The rupture zone of the great 1857 Ft. Tejon earthquake defines the Central San Andreas fault, and like the 1906 break, can be divided into subsegments along its 300-km length. The three principal subdivisions are the Cholame (SA-6), Carrizo (SA-7), and Mojave (SA-8) segments. The Cholame segment lies at the northern end of the 1857 rupture and exhibited a northwards decrease in displacement from 9 meters to less than I meter over a distance of about 55 km in 1857. Application of existing methods for estimating recurrence times is difficult, due to the gradational nature of the displacement. A best estimate, using the median value for slip on segment SA-6, is 159 years and the corresponding probability for an event of M s ~ 7 during the next 10 years is at the 11 % level. The Carrizo segment sustained surface displacements that ranged from 6 to 10 meters in 1857, and at 3.4 cmJyr, indicate recurrence times of 296 years, or probabilities of 1% for the next IO years. Hence the possibility for a great earthquake in central California, involving the Carrizo segment of the San Andreas fault, appears to be negligible over the next IO years. The Mojave segment contains the southern terminus of the 1857 rupture, and


sustained surface displacements of about 3-5 meters. Pallet Creek, in the Mojave segment, has been extensively studied (SIEH et al., 1989), and offset sediments dated by 14C indicate recurrence intervals that range from 44 to 332 years. The lack of well-defined periodicity in the Pallett Creek da ta is puzzling at present, and may indicate overlap of adjacent rupture zones in this region (i.e., Carrizo and Coachella-San Bernardino) or temporal clustering. Based on interpretation of data at face value, the average recurrence time is 131 years, and probability for an event of M s ~ 7.5 over the next 10 years is at the 7% level. If we use the direct method to estimate a repeat time, with an average slip of 4.5 meters, and rate of fault motion of 3.0 cm/yr, the expected recurrence time is 162 years, and the probability is at the ll % level for the next 10 years.

SA 9,10. Southern San Andreas, 34.3°-33.00 N. The southern San Andreas consists of two primary segments, San Bernardino (SA-9) and Coachella Valley (SA-IO). Paleoseismic investigations of the Coachella Valley segment indicate a sequence of 4 events during the period A.D. 1000-1700 (SIEH, 1986). The average repeat time is 220 years, and the last event occurred in A.D. 1680 ± 20. Given these observations, the probability for a future M s ~ 7.5 event is at the 14% level for the next 10 years. While this estimate is relatively low, it still implies that this segment is the most likely site of the next great California earthquake. Much less is known about the San Bernardino segment. Dendrochronological studies indicate that the last event may have occurred in 1812 (JACOBY et al., 1988); however, the magnitude and extent of rupture at this writing are not weIl known. In addition, it is not weIl understood how the San Bernardino segment interacts with the adjacent Mojave or Coachella Valley segments. Tentative estimates of the seismic potential for San Bernardino are at the 8% level for the next 10 years.

Washington - Oregon

In recent years, scientific attention has focused on the possibility of great earthquakes occurring along the Washington-Oregon coast of the Uni ted States, in association with the subduction of the Juan de Fuca plate (see HEATON and KANAMORI, 1984; HEATON and HARTZELL, 1986). Most of the arguments for, and appraisals of, the seismic potential of the Cascade subduction zone have been made on the basis of tectonic analog with other convergent margins that are presently subducting young oceanic lithosphere, and have produced great earthquakes (see for example, discussions for Jalisco, Mexico, and Ecuador-Colombia). At present, the only firm evidence for recurring, large-scale tectonic events is based on observations of subsided coastal marsh deposits by ATWATER (1987), ATWATER et al. (1987), and GRANT et al. (1989). While these data indicate widespread subsidence or submergence events, it is not clear at this writing that these are unequivocally related to the occurrence of great earthquakes. Preliminary 14C dating of buried wetland soils indicates an average return period of 500-600 years. These


intervals, however, are not strictly periodic and individual interevent times range from 100 to 1000 years. The last event has been dated at A.D. 1618-87 by dendrochronological methods (YAMAGUCHI et al., 1989) and suggests that we are still in the middle of the recurrence cycle. Based on these 14C dates and the apparent aperiodic nature of the interevent times, a Poisson model for recurrence indicates a probability of 2% for the next 10 years and is approximately equal to the time-dependent estimate. Consequently, the hazard over the next 10 years appears to be small. One caveat, however, is that if these submergence events are related to earthquakes, they only describe the effects of great (i.e., M w 8.0 + ) events, and say little or nothing about the occurrence of large (i.e., M s 7.0-7.7) interplate or intraplate events in this region (WEAVER and SMITH, 1983; BUCKNAM and BARNHARD, 1939; SPENCE, 1989).

North Pacific

Queen Charlotte - Alaska - Aleutians

The Queen Charlotte-Alaska-Aleutian (QC-A-A) seismic zone extends north from the Juan de Fuca spreading center offshore of British Columbia through southern Alaska and west to the Aleutian and Kommandorski Islands in the northwest Pacific. Throughout the more than 5000-km long zone of interaction between the Pacific and North American plate, 5 distinct tectonic regimes are recognized to constitute the QC-A-A seismic zone. These include: 1) a predominately strike-slip regime along the Queen Charlotte-Fairweather fault zone, 2) a zone of transition between strike-slip and underthrust motion in south-east Alaska, 3) a continental-type subduction region in southern Alaska grading into, 4) island are type subduction in the Aleutian Islands and 5) a regime of oblique subductiontransform motion in the Kommandorski Islands. The rates of relative plate motion vary from about 5.5 cm/yr along the Queen Charlotte Islands to about 9 cm/yr near the Kommandorski Islands. The following discussion is based on work by NISHENKO and JACOB (1990), and Figure 21 summarizes the seismic forecast for this region for the time interval 1989-1999. Place names along the Queen Charlotte-Alaska-Aleutian seismic zone and the locations of recent earthquake rupture zones are sh9wn in Figure 22.

QCAA-l. Cape St. Jarnes, 51.7°-52.4°N. Relocation of earthquakes (M ~ 7 since 1900, M ~ 6 since 1917) on or near the Queen Charlotte fault by ROGERS (1986) indicates the existence of a 75-km long segment (south of the 1949 M w 8.1 Queen Charlotte Island shock, QCAA-2) which has no history of large earthquakes since 1898. Comparison with other areas along the strike of the Queen Charlotte fault indicates that this area may be capable of independently producing large earthquakes. The estimated size of a future event that would fill this gap is

I P

AC

IFIC

OC

EA

N

Con

ditio

nal

Pro

babi

lity

1989

-199

9 [i

ll6

0-1

00

% ~ 0

·20%

_ 4

(H;(

)%

~ 2

0-4

0%

IU

o-;

;;:m

1f'I

r:,or

np..

Hl&l

OI1

c

I /

I~

L..

..!I

_

Fig

ure

21

50'

~\140

'

Seis

mie

pot

enti

al o

f th

e Q

ueen

Cha

rlot

te-

Ala

ska-

Ale

utia

n se

ism

ie z

one:

198

9-19

99.

Pat

tern

s po

rtra

y th

e le

vel

of

eond

itio

nal

prob

abil

ity

for

oeeu

rren

ee o

f la

rge

and

grea

t (M

s 7.

0 an

d la

rger

) ea

rthq

uake

s du

ring

the

nex

t 10

yea

rs,

1989

-199

9, a

nd

rang

e fr

om

haeh

ured

, 0

-20

%;

eros

s-ha

ehur

ed,

20

-40

%;

open

do

ts,

40

-60

%;

and

fille

d do

ts,

60

-10

0%

. Sm

all

eros

s pa

tter

n de

note

s th

ose

area

s w

ith

no h

isto

rie

reeo

rd o

f gr

eat

eart

hqua

kes.

Spe

cifi

e da

tes

and

mag

nitu

des

refe

r to

are

as w

ith i

neom

plet

e hi

stor

ie r

eeor

ds.

<: ~

w

.V>

'l:>

'l:

>

(j

~.

c:: 3 ~ ~ ~

[3 ;:;. d' " a a N

:;;

50

0

(\'~ °a

\~

'u~

\l

17

00

E

...

" B

eri

ng

S

ea

\\

'-"\

"

Ale

uti

an

~

Pa

cif

ic

18

00 1

70

0 W

1

60

0

Fig

ure

22 (

Top

)

\:")

<'0

00

Oce

an

15

00

14

00

Ca

na

da

13

00

W

N ~

~

C- O>

::l

:-0

Z

,"' ir

::>

i'<"

o ;e Cl ~ '"C

I .::

t

60

0 N

~ 8

.6

50r~

: 1

70

0 E

\.

~ 1

96

4

<)

Mw

9.2

d

(V/

19

38

~

19

17

M

8.3

0'O-Q

0'O

-Q

0'O

-Q

19

46

\7

.5

19

29

"

Ms

7.4

cm

/yr

• M

w7

.8

8.3

_

_

11

I cm

/yr

__

I

18

00

17

00

W

16

00

Fig

ure

22 (

Bot

tom

)

I 1 I

c;:J

~~ -91

\~"W-

.. •

o'

~

.. ~

og,

0=

I

Q\

""

Pa

cif

ic

Pla

te

L-L

-%l

I ~

15

00

14

00

13

00

W

(Top

) B

athy

met

ry,

maj

or

tect

onic

fe

atur

es,

and

plac

e na

mes

al

ong

the

Que

en

Cha

rlot

te-A

lask

a-A

leut

ian

seis

mic

zo

ne.

F.F

. an

d Q

.C.F

. ar

e th

e F

airw

eath

er a

nd Q

ueen

Cha

rlot

te f

ault

zone

s, r

espe

ctiv

ely.

Con

tour

int

erva

l fo

r m

arin

e ba

thyr

netr

y is

100

0 frn

(af

ter

NA

VA

L O

cEA

NO

GR

APH

IC O

FFIC

E,

1978

). (B

otto

m)

Rup

ture

zon

es o

f la

rge

and

grea

t ea

rthq

uake

s: 1

917-

1986

(fr

orn

NIS

HEN

KO

and

JA

CO

B,

1990

). M

s an

d M

w v

alue

s re

fer

to t

he 2

0-se

c su

rfac

e w

ave

and

ultr

a-Io

ng p

erio

d m

agni

tude

sca

les.

Dir

ecti

ons

of

rela

tive

plat

e m

otio

n be

twee

n th

e Pa

cifi

c an

d N

orth

Arn

eric

an p

late

s in

dica

ted

by

arro

ws.

Rat

es o

f re

lativ

e pl

ate

mot

ion

(in

cmjy

r) a

fter

MIN

STER

and

JO

RD

AN

(19

78).

<: ?- ;:;:;

,CA

::0

'Ci - (j

~.

c:: 3 .;;, '" (") S;

(") cn

<>

0;' 3 (i'

"Cl

0 Ö a ~

IV

IV


approximately M s 7.5-7.6. The strain aeeumulation since 1900 is estimated to be about 4.8 meters and is equivalent to a M s 7.6 earthquake. Unfortunately, information about reeurrence intervals are laeking for this segment. Using the interval 1900-1987 as a minimum recurrence estimate, NISHENKO and JACOB (1990) suggested that the probability for the next 10 years eould be at the 22% level. In keeping with the eategorization used in this study, we have ranked this region as having no historie reeord. While poorly eonstrained, this segment stands out along the Queen Charlotte-Fairweather fault zone and deserves further study.

QCAA-2. Queen Char/olte /s/ands, 52.4°-56°N. The 22 August 1949 (Mw 8.1) Queen Charlotte lslands earthquake is estimated to have had a rupture length of approximately 490 km. Based on seismologieal studies, the majority of seismie moment release (and henee eoseismie displaeement) appears to have been released in a zone extending 265 km north of the epieenter (BEN-MENAHEM, 1978; BOSTWICK, 1984). This suggests that smaller amounts of displacement may have oeeurred in the adjaeent northern and southern segments (55S-56°N and 52S - 53SN, respeetively).

An estimate of the average eoseismie displaeement in 1949 ranges from 4 to 7.5 meters, depending on the length of faulting ehosen. Dividing these estimates by the rate of fault motion indieates a reeurrence time of 70-130 years. The historie and geologie record of prior great earthquakes along this segment of the plate boundary is poorly known. Henee independent estimates of recurrence time are not available at present. Using the average of the estimates, 108 years, and eonsidering the receney of faulting, the probabilities for a M w 8.0 event over the next 10 years are at the 4% level.

Note that this diseussion deals with great M w 8.0 events. The possibility that signifieant variations in displaeement may exist along strike would allow smaller fault segments that eould rupture independently in large earthquakes (i.e., M s 7-7.5) and have recurrenee times shorter than 100 years. At present there are no data with whieh to eonstrain a reeurrenee time for these smaller segments.

QCAA-3. Sitka, 56°-58°N. Previous earthquakes that are either known or suggested to have oeeurred along this segment of the plate boundary include: 26 November 1880; 24 October 1927 (Ms 7.1); and 30 July 1972 (Ms 7.6). The 1972 event was anticipated by SYKES (1971) on the basis of the seismie gap hypo thesis. If all three events ruptured the same segment of plate boundary, the observed average repeat time is 46 ± 7 years. However there is a large variation in the magnitude of the last two earthquakes (7.1 and 7.6). Given the reeeney of faulting, repeat times based on the historie record and the size of the 1972 event indieate negligible probabilities (2%) over the next 10 years.

QCAA-4. Lituya Bay, 58°-60SN. The Fairweather fault is the only historieally aetive transform fault along the Queen Charlotte-Alaska-Aleutian seismic zone that is exposed on land, and henee accessible for direct measurement of rate of fault motion and eoseismie displaeement. Observations following the 10 July


1958 (Ms 7.9) earthquake indieate a maximum eoseismie displaeement of6.5 meters at Crillion Lake and smaller offsets (2.5-3.5 meters) elsewhere (TOCHER, 1960; PLAFKER et al., 1978). Dividing the maximum offset by the rate of fault motion gives a reeurrenee time estimate of 125 years. The long-term geologie rate of fault motion gives estimates of about 110 years for 6.5-meter events or 60 years for 3.5-meter events (PLAFKER et al., 1978). For eomparison, dividing the average coseismie displacement (4.5 meters) by the rate of fault motion gives an estimate of 85 years. Given the range of recurrence time estimates for this segment of the QC-A-A seismic zone, 85-125 years, the probability for the repeat of an M w 8.2 earthquake is negligible (4%) over the next 10 years.

QCAA-5,6. Yakutat and Yakataga, 139°-145°W. Both the Yakutat and Yakataga segments define a complex transitional zone between predominately strike-slip motion along the Fairweather fault system and underthrusting along the Alaskan trench. Between these two simple fault systems, thrust faulting along the east-west trending Chugach-St. Elias fault zone, and evidence of active faulting offshore within the exotic Yakataga bloek (i.e., the Pamplona zone) indieate the existenee of a eomplex and diffuse plate boundary in this region. In other words, plate motion does not appear to be concentrated along a single throughgoing structure, and hazards estimates made for this segment may be less reliable.

Prior great earthquakes in this region occurred on 4 and 10 September 1899 (M w 8.2 and 8.1, respectively). Estimation of recurrence intervals for these two events, however, is highly dependent on the interpretation of the type of faulting involved. Estimates for the 10 September event range from 80 years for an interplate event to 380 years for an intraplate event (MCCANN et al., 1980; PLAFKER and THA TCHER, 1982). While there is doubt as to the type of faulting involved in the 10 September event, there is little doubt that the 4 September event was an interplate event that involved faulting along the Chugach-St. Elias fault zone and produced coastal uplift of about 1-3 m (PLAFKER and THATCHER, 1982; JACOBI and ULAN, 1983). For this event the estimated coseismic displacement ranges from 2.5 to 5 meters (MCCANN et al., 1980), and average repeat time is about 99 years. Strain accumulation measured in the Yakataga gap region from 1979 to 1984 is consistent with the rate of plate convergence in this area, and does not indieate that any appreciable aseismic slip is occurring (SAVAGE et al., 1986). While the amount of coseismic displacement is poorly constrained, the conditional probability based on the average of these estimates is at the 21 % level for the next 10 years.

QCAA-7. Prince William Sound-GulJ 0/ Alaska, 145°-156°W. The 24 March 1964 Prince William Sound earthquake (M w 9.2) is the largest earthquake to have occurred along the QC-A-A seismic zone in historic time, rupturing both the Gulf of Alaska (QCAA-7) and Kodiak Island (QCAA-8) segments. Estimates of the average coseismic displacement, based on seismologie and geodetie information, range from 14 to 30 meters (PLAFKER, 1969; MIYASHITA and MATSUÜRA, 1978),


and the estimated recurrence time ranges from 230 to 460 years. Hence the probability of another giant earthquake in this zone is negligible ( ~ I %) for the immediate future.

The recurrence estimates for M w 9.0 earthquakes indicate little about the possibility of smaller great earthquakes (i.e., M w 8.0) occurring in this area. Relocations of great events on 14 July 1899 and 9 October 1990 indicate that the region also has the potential for other destructive earthquakes in addition to the largest ones (BOYD and LERNER-LAM, 1988). Unfortunately, the hazard associated with these events is not weil understood, and the probability for these events to occur cannot be dismissed by virtue of the fact that the probability for M w 9.0 events may be low in the future.

QCAA-8. Kodiak /sland, 150o-155°W. The area in and around Kodiak Island is distinguished by having one of the longest, and presumably most complete earthquake histories along the entire QC-A-A seismic zone (DAVIES et al., 1981). Reports of large earthquakes feit on Kodiak Island during the last 200 years suggest a recurrence interval of about 60 years between periods of increased activity that may last as long as 10 years. The identification of specific fault segments is difficult, however, given the sparse distribution of feit reports. For this reason, we feel it appropriate to use a Poisson-based description of probability for the Kodiak Island region. In other words, we have sampled the strain release in a volume surrounding Kodiak Island, not just the behavior of a single fault or plate boundary segment. Using a Poisson model and the observed 58-year recurrence interval, the probability for the next 10 years is at the 16% level. Since this is a Poisson based probability estimate, it is static and will not change as a function of time. For comparison, if these earthquakes do in fact, represent the activity of a single source region centered near Kodiak Island with an approximate 60-year periodicity, the time-dependent conditional probability is at the 1-2% level for the next 10 years.

QCAA-9. Alaskan Peninsula, 156°-158SW. Previous large and great earthquakes that are known to have ruptured this segment include 22 July and 7 August, 1788; 16 April 1847; 20 September 1880; and 10 November 1938 (Mw 8.2). These events indicate recurrence times between 49 and 75 years for this segment of the QC-A-A seismic zone. The spread in the above repeat times represents uncertainty as to the size and location of the 1880 event. Whether the 1880 event represents an interplate or intraplate event is not clear (see descriptions in DA VI ES et al., 1981). The historie record does indicate that during the last 200 years, the mode of rupture in this region has changed from aseries of great shocks in 1788 and 1847 (with estimated rupture lengths of 400-600 km) to a sequence of relatively smaller events in 1880, 1917 (see next section) and 1938 which independently ruptured the Shumagin and Alaskan Peninsula segments (DAVIES et al., 1981; BOYD and LERNER-LAM, 1988). Hence, while we are able to estimate recurrence times for this region, we cannot quantitatively assess the likelihood that these two segments will


break independently or simultaneously. Based on the average reeurrenee time of 75 ± 11 years, the eonditional probability for the next 10 years is at the 18% level.

The seismie moment estimates for the 1938 earthquake range from 2.1 x 1021 Nm (based on the 100-see surfaee wave measurements of BRUNE and ENGEN, 1969) to 5 x 1021 N-m (based on the tsunami wave magnitude seale of ABE, 1979). Using the dimensions of the gap (250 km x 100 km) and a eonvergenee rate of 7.1 em/yr, the eorresponding reeurrenee time estimates range from 23 to 53 years. Note that these estimates are somewhat shorter than those indieated by the historie reeord and may imply aseismie slip in this area. Aeeordingly, we have based our foreeast on the historie reeord alone.

QCAA-I0. Shumagin [slands, 158S-161.7°W. Previous great earthquakes loeated in the Shumagin Island region include 22 July and 7 August 1788; 16 April 1847; and 31 May 1917 (Ms 7.4, ESTABROOK and BOYD, 1989). As in the ease of the Alaskan Peninsula gap, there is an indication that the region exhibits a variable mode of rupture. Based on the 1788, 1847 and 1917 events, the average repeat time is 65 ± 10 years, and the probability for reeurrenee over the next 10 years is at the 48% level.

Based on the size of the 1917 event (Ms 7.4 ± 0.3 or a maximum seismie moment of 4.5 x 1020 N-m) and the aftershoek zone (125-200 km length and 100 km width), the average eoseismie displaeement is on the order of 1 meter. Given the rate of plate motion in this area (7.5 em/yr), the estimated reeurrenee time is less than 20 years, signifieantly less than implied by the historie reeord (65 years) and the amount of time elapsed sinee the 1917 event (72 years). Possible explanations for this marked diserepaney are that the 1917 rupture area has been overestimated, the average eoseismie displaeement underestimates the peak slip, or that a signfieant amount of aseismie slip is oeeurring in this area. In any ease, it appears that the direet estimates are too unreliable at present and we use estimates based on the historie reeord for our foreeast.

Whether aseismie slip does, or does not occur in the Shumagin Islands region is eontroversial. LISOWSKI et al. (1988) reported that geodetie strain measurements on the Shumagin Islands, for the period 1980-1987, showed no appreciable horizontal strain aeeumulation in the direction of plate eonvergenee. BEA VAN (1988), while not excluding the possibility for aseismic slip, points to levelling data eovering virtually the same time span (1980-1988), whieh shows a pattern of eoherent tilting on several of the islands. To explain and unify these two data sets, BEAVAN (1988) developed a model of the main thrust zone that is loeked at depths from 27 to 55 km, shows a steepening of dip midway between these two depths, and has stable sliding or aseismie slip above and below the loeked depth range. The model is, within error bounds, eonsistent with both the strain and tilt data, but it also prediets larger sealevel ehanges than have been measured by tide gauges on the islands (J. Beavan, personal communieation, 1989). Thus arguments for appreeiable strain aeeumulation during the 1980's are questionable for the eastern half of the


Shumagin segment, to which strain and tilt measurements have been restricted. HUDNUT and T ABER (1987) point out that recent background seismic activity in the western half of the Shumagin gap is distinctly different, and that geodetic and tilt data from that region are not available. Thus it is not c1ear whether all or only portions of the Shumagins slip aseismically, and how much strain, if any, has been accumulating since the last large shallow earthquake in 1917. Given the occurrence of the M s 7.5 earthquake in 1948, which ruptured the deepest portion of the main thrust zone in the central section of the Shumagin Islands (BOYD et al., 1988), it is conceivable that the Shumagin gap may not be capable, by itself, of producing a great earthquake. It does, however, have ademonstrated historie capability to independently produce mid-7 magnitude earthquakes. Conversely, the Shumagin gap may participate in a great earthquake which either reruptures the 1938 zone to the east of the Unimak segment to the west. As noted previously, these uncertainties are not explicitly reflected in the conditional probability and error estimates for the Shumagin Islands segment.

QCAA-ll. Unimak Island, 161.7°-164°W. This segment is defined on the basis of the 1 April 1946 (Ms 7.4, MI 9.3) earthquake, which produced one of the largest tsunamis recorded during this century. The size of the tsunami, however, is not in agreement with the seismographically observed size of the earthquake, and it has been suggested that the tsunami was triggered by submarine landsliding at the time of the earthquake (H. Kanamori, personal communication, 1985; S. Lewis, personal communication, 1985). Based on the size of the earthquake, M s 7.4, we estimate a recurrence time of about 25 years. Locations of prior earthquakes in this region are poorly knowo. One event, 10 Oecember 1905, M s 6.9, is located on the eastern edge of the 1946 aftershock zone, and may indicate a longer 41-year recurrence interval. Oue to the amount of time elapsed, both the 25- and 41-year estimates indicate probabilities for the occurrence of a large (i.e., M s 7-7.5) event at the 43-55% level over the next 10 years. The unusual circumstances surrounding the 1946 tsunami, however, indicate that the next large Unimak Island earthquake does not necessarily have a high potential for producing a great tsunami.

QCAA-12-14. Central Aleutians, 165°-180oW. The Aleutian Islands portion of the QC-A-A seismic zone provides an excellent example of documented variations in the mode of subduction between seismic cyc1es. At the turn of the 20th century, aseries of large earthquakes ruptured this region within an interval of 10 years. Following this episode of subduction, the great (M w 8.7) 9 March 1957 Andreanof Islands earthquake and its aftershocks ruptured this entire plate boundary within a few days. The occurrence of another great earthquake on 7 May 1986 (Mw 8.0) appears to represent the return to a more segmented mode of earthquake recurrence in this region. Based on the analysis of the 1957 shock and the occurrence of the 1986 event, it is possible to divide the Aleutian Islands portion of the QC-A-A seismic zone into three segments: Fox Islands (l65°-173°W); Andreanof Islands (I73°-177°W); and the Oelarof Islands segments (177°-180°W). The locations of


these segments appear to be controlled by the presence of fracture zones of the subducted Pacific plate (Amlia FZ 173°W, Adak FZ 177°W) as weil as major structural discontinuities within the Aleutian Ridge and forearc regions: Adak Canyon, Bowers Ridge, Hawley Ridge.

QCAA-12. Fox Islands, 164°-173°W. This segment includes the Unalaska gap of HOUSE et al. (1981). Following the 1957 mainshock, the Fox Island segment exhibited a gradual expansion of the afters hock area, over aperiod of three days, in contrast to the more rapid 24-hour expansion in the Andreanof and Delarof segments (BOYD, 1986). All evidence suggests that the rupture characteristics changed dramatically in the region east of the Amlia fracture zone (173°W). In conjunction with the arrested rupture propagation of the 1957 event, it is noteworthy that a great normal faulting earthquake occurred seaward of this segment on 7 March 1929 (M w 7.8) (SYKES, 1971; KANAMORI, 1972). The occurrence of such an event is suggested by KANAMORI (1972) as evidence of partial decoupling of the plate interface in this area. A number of large earthquakes have been located in this segment during the 20th century, suggesting that the normal mode of strain release in this segment may characteristically occur as large, rather than great, earthquakes every 20-50 years. Based on the historical seismicity and estimates of displacement in this region in 1957 (approx. 240 cm, or a 30-year recurrence time) we estimate the probability for large (i.e., M s 7-7.5) events in this segment to be at the 44% level for the next 10 years.

QCAA-13. Andreanof Islands, 173°-177°W. This segment contains both the epicenter ofthe 1957 event and the rupture zone ofthe 1986 earthquake (EKSTRÖM and DZIEWONSKI, 1986; EKSTRÖM and ENGDAHL, 1989; ENGDAHL et al., 1989; HWANG and KANAMORI, 1986; BOYD and NABELEK, 1988). Both earthquakes nucleated in approximately the same area, just east of Hawley Ridge, suggesting that this forearc feature may playa significant role in the earthquake process there. Based on the short instrumental record for this region, the only great earthquakes known to have occurred are 1957 and 1986. A great event in 1906 is located near 180°, and may have ruptured either to the east or to the west (BOYD and LERNER-LAM, 1988). Hence the two return periods for this segment of the QC-A-A seismic zone are 29 and ~ 51 years. Given these two intervals, the average return time is at least 40 years, and the conditional probability for the recurrence of a great earthquake is at the :$; 1 % level.

QCAA -14. Delarof Islands, 177° -1800 W. The boundaries of the Delarof segment are delineated by Adak Canyon and Amchitka Pass (180°) on the Aleutian Ridge, and the presence of the Adak fracture zone on the subducted Pacific plate at 177°W. As in the case of the Fox and Andreanof Islands segments, the seismologic record for this century indicates a history of large earthquakes, with an apparent return period of 20-30 years, punctuated by the 1957 event. Large events located in this segment occurred on 14 February 1905 (Ms 7.3) and 5 and 7 July 1929 (Ms 7.0 and 7.3). Combined with the occurrence of the 1957 shock, this suite


of events suggests areturn period of 26 years for large earthquakes. Using the sparse instrumental record, in conjunction with the revised 30-year recurrence time from the tide gauge inversion of WAHR and WySS (1980, see NISHENKO and JACOB, 1990, for details), the conditional probability for the recurrence of a large earthquake in this region is at the 85% level for the next 10 years. This segment has been a gap for large earthquakes since the 1957 event. The only large event near this region was the 1971 Adak Ca ny on earthquake (Ms 7.1) whose source zone straddles the boundary between the Delarof and Andreanof Islands segments at 177°W but did not rupture into the Delarof segment (HABERMANN, 1981).

QCAA -15. Rat !slands, 180° -171 oE. Previous large and great earthquakes are only known from the instrumental record and include 29 June 1898 (M s 7.6), 17 August 1906 (Ms 7.8), 2 September 1907 (Ms 7.4),17 December 1929 (Ms 7.8) and 4 February 1965 (M s 8.7). While the exact locations of these earlier events are not weil constrained (BOYD and LERNER-LAM, 1988), these events suggest recurrence times of 21- 59 years.

Within this segment, the QC-A-A seismic zone is composed of 3 distinct tectonic crustal blocks: Rat (180° -176SE), Buldir (176S -174SE), and Near (174S -171 OE). The boundaries of these blocks are weil defined by submarine canyons that are of tectonic rather than erosional origin (GATEs and GIBSON, 1956), and reflect the block segmentation and rotation of the Aleutian arc (GEIST et al., 1988). The sequence of large events at the turn of the century indicates that these individual blocks ruptured separately but within a few years of one another. In 1965 all 3 blocks ruptured simuItaneously (SPENCE, 1977). Inversion of long-period body waves indicates that the majority of seismic moment release in 1965 is concentrated in the easternmost 2/3 of the aftershock area (i.e., from 180° to 173°E, RUFF and KANAMORI, 1983; BECK and CHRISTENSEN, 1991). The source zone of the 1965 tsunami terminates near 175°E and HATORI (1981) identified Attu Island (within the Near Islands block) as a tsunami gap.

Within the last 90 years there have been two modes of strain release in this segment. One involves the individual rupture of separate blocks and the second involves simultaneous rupture of all the blocks. Additional evidence for independent behavior of individual blocks mayaiso be indicated by the occurrence of the 1898 (M w 8.0) and 1929 (M w 7.9) events in the vicinity of Near block. If these earlier events are interplate earthquakes, then the Near block may have a recurrence cycle half as long as the Rat block.

Recurrence estimates based on the size of the 1965 event range from 58 to 83 years and indicate a low probability (4%) over the next 10 years. If the Near block is sufficiently decoupled from the adjacent blocks, the observed 33-year cycle indicates that the probability for a large event in this segment may be quite high over the next 10 and 20 years (i.e., 45 and 83% respectively, see NISHENKO and JACOB, 1990, for details).


QCAA -16. Kommandorski Islands, 171 ° -165°E. The motion of the Pacific plate is parallel to the Aleutian Arc along the Kommandorski Islands segment of the QC-A-A seismic zone, and terminates underthrusting at the Kurile Trench along the east co ast of Kamchatka. Within the Kommandorski segment motion is accommodated by a broad zone of strike-slip deformation which includes vertical strike-slip motion along the Bering Sea flank, and horizontal strike-slip motion along the southern or Pacific flank of the Aleutian Ridge (CORMIER, 1975).

Few large shocks have been located in this segment during this century. As discussed by MCCANN et al. (1979) there is evidence for large or great shocks along the southern flank in 1849 and 1858. Uncertain size and poorly known rupture dimensions for these shocks make estimation of repeat times difficult at this time.

Western Pacific

The western Pacific seismic zone extends south from Kamchatka through the Kurile, Japan, Izu-Bonin and Mariana trenches, and represents the zone of convergence between the Pacific and the Asian and Philippine Sea plates. Along this zone, the rates of northwesterly plate convergence vary from 8.5 cm/yr to 11 cm/yr, from north to south. Variations in the rate and size of shallow thrust earthquakes that occur along the western Pacific seismic zone reflect changes in the plate interface geometry and other geophysical parameters that influence the tectonic regime and earthquake potential of a particular segment of plate boundary. Seismic hazards estimates for the western Pacific seismic zone are shown in Figure 23.

Kamchatka and the Kurile Islands

In contrast to the lengthy Japanese earthquake history discussed in the next section, much less is known about the earthquake history of the Kurile and Kamchatka seismic zones. Along the Kurile Islands there are few sites for observation and the generally low population densities have resulted in a sparse earthquake record prior to the advent of instrumental recording. Hence, the majority of recurrence estimates for this portion of the circum-Pacific are based on only one or two observed repeat times, and have large uncertainties. In the southern Kurile Islands, the identification of segmentation and recurrence histories are further complicated by overlapping rupture zones and interacting asperities. The following discussion summarizes work by FEDOTOV (1965), FUKAO and FURUMOTO (1979), KONDORSKA YA and SHEBALIN (1982), FEDOTOV et al. (1982), SCHWARTZ and RUFF (1985, 1987), BECK and RUFF (1987), and SCHWARTZ et al. (1989). Figure 24 shows recent earthquake rupture zones along the Kurile-Kamchatka seismic zone.

KK-l. Kamchatsky Peninsula, 55°-57°N. Two large events have occurred in this segment during this century, 13 November 1936 (M s 7.1) and 15 December 1971 (Ms 7.5). The latter event is distinguished on the basis of having been



~:::::::::l 60-100% ..... _ 40-60%

m20-40%

~0-20% f"i""l No Historie Racortt l!....!....!..J 01 Greal Earthquakes

'9', .... 79~ Incomplele Historie Record

U-____________ ~ ______________ L_ ____________ ~ ______________ L_ __________ ~100

Figure 23 Seismic potential of the Western Pacific, including Kamchatka, the Kuriles, Japan, Izu-Bonin, and the Mariana ares: 1989-1999. Patterns portray the level of conditional probability for occurrence of great (Ms 7.7 and larger) earthquakes during the next 10 years, 1989-1999, and range from hachured, 0-20%; cross-hachured, 20-40%; open dots, 40-60%; and filled dots, 60-100%. Small cross pattern denotes those areas with no historie record of great earthquakes. Specific dates and magnitudes refer to

areas with incomplete historie records.

Figure 24 Major tectonic features, active volcanoes (solid triangles), and recent earthquake rupture zones (crosshatched areas) along the Kamchatka-Kurile-Japan-Izu-Bonin-Mariana trench system (after KELLEHER and MCCANN, 1976). Note the relatively smooth ocean ftoor, large earthquake rupture zones, and the nearly continuous line of active volcanoes to the north (Kamchatka-Kurile-Japan). Contrast this with the absence of great earthquakes, the irregular distribution of active volcanoes, and the rougher ocean ftoor bathymetry to the south along the Bonin and Marianas trenches. Bathymetry from CHASE el al. (1970).

Vol. 135, 1991 Circum-Pacific Seismic Potential

PACIFIC

OCEAN

\ .. . o

o • o (/

Figure 24

Ob.

231


anticipated by FEDOTOV (1965) using the basic concepts of seismic gaps. Based on this apparent 35-year recurrence, the probability within the next 10 years for another large event is at the 23 % level.

KK-2. Northeastern Kamchatka, 53°-55°N. Previous great earthquakes that have affected the Pacific coast of northeast Kamchatka occurred on 22 August 1792 (M w 8.4) and 3 February 1923 (M w 8.3, Mt 8.8). Based on this single 131-year repeat time, the probability for another great event in this northeastern zone is at the 3% level for the next 10 years.

KK-3. Southeastern Kamchatka and Paramushir Island, 49°-53°N. Prior great earthquakes that are located in this segment include 17 October 1737 (M 8.3), 6 May 1841 (Mt 9.0), 25-27 June 1904 (Ms 7.2-7.4), and 4 November 1952 (Mw 9.0). Descriptions of the 1737 event (KONDORSKAYA and SHEBALlN, 1982) indicate that this was a great earthquake with a rupture zone comparable to that of the 1952 event. Descriptions of the 1841 earthquake, however, do not provide as clear a comparison. The primary information concerning the size of the 1841 event comes from a comparison of tsunami waves recorded at Hilo, Hawaii (1841, 4.6 m and 1952, 3.7 m), and ABE (1979) has assigned a Mt of 9.0 to both events (see Figure 5). If the 1737, 1841 and 1952 events all ruptured similar portions of the seismic zone, the average repeat time would be 108 ± 16 years, and the probability for another great (i.e., M w 9.0) event in this region is less than 1 % for the next 10 years. If, in fact, only the 1737 and 1952 events are comparable, the recurrence time increases to 215 years. Both recurrence time estimates however, indicate negligible probabilities for the immediate future.

The 1904 events, as depicted by FEDOTOV (1965) and FEDOTOV et al. (1982) appear to have ruptured a central portion of the 1952 zone. These events point out the possibility of large earthquakes occurring within the rupture zones of great earthquakes (see also discussion for the 1964 Prince William Sound earthquake in the Alaska section). At present, we have no data to quantitatively evaluate the probability for these types of events to occur.

KK-4. Shiashkotan Island, 48°-49°N. Only one event, 1 May 1915 (Ms 8.0), is listed for this segment of the Kurile Islands by FEDOTOV et al. (1982). Extrapolation of recurrence times for similar sized events along the Kurile are suggests that intervals of 80-100 years may be appropriate if the 1915 earthquake is an underthrust event. In this case we tentatively estimate a probability at the 20% level for the next 10 years.

KK-5. Central Kurile Islands, 46°-48°N. No large or great shallow earthquakes are known for this segment of the Kurile Island are during this century. This segment is also coincident with a pronounced discontinuity in the structure of the island are, between the Bussol and Kruzenshtern Straits, where depths to the Vityaz Ridge increase by 2000 + meters. This unusual island are structure suggests that the lack of great earthquake activity may be symptomatic of the tectonic regime. Interestingly, a compressional outer rise event occurred opposite this segment in


1963 (mb 7.7) and may reftect the presence of compressive stresses at the plate interface zone (CHRISTENSEN and RUFF, 1989). At present, lacking both an historic record and knowledge of when the last event occurred, we do not have enough information to quantitatively evaluate the large or great earthquake potential of this zone or assess the intermediate term significance of the compressional outer rise events. Qualitatively, however, we note that this region should be considered one of high but unknown potential. If this gap ruptured at one time, the physical dimensions indicate that it could produce both a great earthquake and tsunami.

KK-6,7. North and South Urup Island, 149°-152SE. Prior great earthquakes along this section of the Kurile arc include 7 September 1918 (M, 8.7) and 13 October 1963 (M w 8.5). While an earlier event in 1780 is also associated with this segment, the completeness of the earthquake catalog during the 19th century is in question (see SYKES and QUITTMEYER, 1981). In contrast to many of the other seismic zones around the circum-Pacific, this segment of the Kurile are has exhibited a substantial degree of overlap between adjacent ruptures in 1918 and 1963. Analysis of the 1918 and 1963 events by BECK and RUFF (1987) show that the 1963 earthquake was a complex event that involved a rupture of 3 separate asperities. The southwestern asperity, near the epicenter offshore S. Urup Island (KK - 7), appears to have been unbroken since 1780. The remaining 2 asperities, offshore Northern Urup Island (KK-6), appear to have been ruptured in 1918 and subsequently reruptured in 1963.

Given the uncertain status the historic record in this area, recurrence times may range from 45 (1963-1918, KK-6) to 183 (1963-1780, KK-7) years. The 45-year recurrence estimate indicates probabilities at the 21 % level for the next 10 years, and the probabilities based on the 183-year estimate are significantly smaller.

KK - 8. Etorofu Island, 148° --:-1500 E. The only great earthquakes located in this segment occurred on 6 November 1958 (M w 8.3) and 24 March 1978 (M s 7.6). While separated by 20 years, both events ruptured different portions of the plate interface, and hence, constitute one complete episode of subduction for this segment of the Kurile arc. The great 1958 event was a high stress drop earthquake that exhibited 10w aftershock activity at shallow depths (i.e., ~ 20 km, FUKAO and FURUMOTO, 1979). The updip portion of this segment was subsequently ruptured 20 years later in 1978. No prior events are known for this segment, and the high stress drop of the 1958 event is suggested to reftect the long interval of strain buildup (FUKAO and FURUMOTO, 1979). Estimates of recurrence time (and future behavior), based on earthquakes in adjacent segments, indicate small probabilities (i.e., ~ 1 % for the next 10 years) but are tentative, at present.

KK-9. Shikotan Island, 146S-148SE. Earthquakes that have occurred along this portion of the arc include 4 June 1893 and 11 August 1969 (M w 8.2). According to SCHW ARTZ and RUFF (1987) the 1969 event can be classified as a simple earthquake that involved the rupture of a single asperity on the plate interface. The estimated coseismic slip in 1969, within this asperity region (7.2 meters),


equaled the strain buildup sinee the 1893 event (76 years x 9.3 emjyr). Reeonstruetion of the tsunami souree zones for the 1893 and 1969 events by HATORI (1979) indicates that the 1969 zone may have been larger than the 1893 zone (200 vs 150 km, respeetively). Henee, ifthe 1969 event, in addition to having a larger souree zone, also had eonsiderable eoseismie slip, then the reeurrence time may be slightly longer than the 76-year interval indieated by the historie record. Both estimates, however, indieate small probabilities for reeurrence in the immediate future (i.e., ~ 1 % for 1989-1999).

Eastern Japan - Eastern Honshu and Hokkaido

The earthquake history of the Japanese islands spans 1300+ years, and represents one of the longest historie reeords in the eireum-Paeifie region. The period of eomplete reporting, however, is eonfined to the period from the 17th eentury to the present. This interval includes a number of repeats for individual segments and henee provides us with reasonably weIl eonstrained recurrenee time and foreeast estimates. The hazard estimates for the Japanese seismie zone are based primarily on data summarized in WESNOUSKY et al. (1984). The ten-year foreeast is shown in Figure 23 and the 10eations of reeent rupture zones are shown in Figure 24.

Based on this 10ng historie reeord, a large variation in repeat times from Hokkaido to southem Honshu has been noticed (KANAMORI, 1977b). While the rates of plate eonvergenee vary by 10% from northem Hokkaido to southem Honshu, the observed reeurrenee times for great earthquakes vary by a faetor of 10. This large temporal variation is suggested to refleet a decrease in the plate coupling from north to south and an inerease in the amount of apparent aseismie slip.

J-1. Nemuro-Oki, 146.5°-147°E. Previous large events known to have occurred in this segment inctude 22 Mareh 1894 and 17 June 1973 (M w 7.8). The 1973 event was antieipated on the basis of the seismic gap hypothesis (FEDOTOV, 1965; SHIMAZAKI, 1974); however, the 1973 event appears to be only half the size of the 1894 earthquake (see THATCHER, 1990). Based on this single 79-year repeat, we estimate the probability during the next 10 years to be at the ~ 1 % level.

J-2. Tokachi-Oki, 144S-146SE. Previous great earthquakes include 1843 (?) and 4 Mareh 1952 (Mw 8.1). This single reeurrenee indicates a probability at the ~ 1 % level for the next 10 years.

J-3. Tokachi-Oki, 142°-144°E. This segment has a weIl known history of great earthquakes that include 13 April 1677 (Ms 8.1), 11 Mareh 1763 (Ms 7.7), 23 August 1856 (Ms 7.7), and 16 May 1968 (Mw 8.2). This sequenee of events indieates an average repeat time of 96 ± 12 years, and the probability for next great earthquake is at the ~ I % level for the next 10 years.

J-4. Sanriku-Oki, 37.7°-39°N. Previous events related to underthrusting along this portion of the Japan treneh include 2 Deeember 1611 (M s 8.1), 17 February 1793 (Ms 7.1), and 5 August 1897 (Ms 7.6). The average repeat time is 143 ± 21


years, and probability for an event in the next 10 years, based on these data, is at the 7% level. Reconstruction of tsunami source zones for the 1879 and 1793 events indicate that both were of similar size and ruptured the same portion of the plate boundary (HATORI, 1976); however, the coseismic displacements in both events were not comparable, and the 1897 event appears to be smaller than the 1793 event. Application of the time-predictable model by WESNOUSKY et al. (1984) indicates a recurrence time of 93 years. Based on this shorter recurrence time estimate, the probability in the next 10 years would increase to the 21 % level.

This section of the Honshu coast is also effected by great normal faulting earthquakes (e.g., the great 1933 Sanriku earthquake) that occur in the trench outer rise region and generate destructive tsunamis. These earthquakes appear to be rare events and are outside the scope of this study.

J-5. Miyagi-Oki, 37S-39°N, 142°E. The Miyagi-Oki region of Honshu has been the site of aseries of earthquakes over the last 350 years that exhibit remarkable regularity in recurrence time (see WESNOUSKY et al., 1984; and NISHENKO and BULAND, 1987). The last event occurred in 1978 (Ms 7.4), and based on the observed 40 ± 3-year periodicity, the probability for an event during the next 10 years is at the ~ 1 % level.

J-6. Shioya-Oki, 36°-38SN. Aseries of five large and great (Ms 7.1-7.7) earthquakes ruptured this portion of the plate boundary in 1938. These appear to be the only events to have occurred in this region within the last 1000 years (ASE, 1977, WESNOUSKY et al., 1984). Based on this observation, the probability for future events in the next 10 years appears to be negligible (i.e., ~ 1 %).

J-7. Japan-Izu Bonin Trench Junction, 35°-36°N. Infrequent large and great events are noted to occur near the junction and Izu-Bonin trenches, southeast of the Boso Peninsula. The 3 February 1605 (Ms 7.9) and 26 September 1953 (Ms 7.5) events occurred southeast of the 1703 rupture in segment J-8 (see below). Lacking further information, WESNOUSKY et al. (1984) used the interval between the 1953 and 1605 events as a recurrence time estimate. The 4 November 1677 (Ms 7.4) earthquake occurred along the Japan Trench in an area with no known predecessors. Given the length of the Japanese record in this region, recurrence times on the order of 1000 years seem appropriate, and the resulting probabilities are low for the immediate future.

J-8. Sagami Trough and Boso Peninsula, 35°-36°N. The last major earthquakes that ruptured the northern portion of the Sagami Trough and the Boso peninsula occurred on 31 December 1703 (Ms 8.2) and I September 1923 (Mw 7.9). The combined coseismic uplift of both events, wh ich comprises the Genroku terrace in this area, matches the height of the highest Holocene terrace (Numa terrace) in the area. This equivalence has led to recurrence time estimates for 1703 -1923 type events to be on the order of 800-1700 years (MA TSUDA et al., 1978; SHIMAZAKI and NA KAT A, 1980). More direct estimates, based on the size of the 1923 earthquake, indicate recurrence times of 200 years (SCHOLZ and KA TO, 1978). Based on


these recurrence time estimates, we calculate negligible ( ~ 1 %) probabilities for this segment of the Boso peninsula.

Southwestern Japan - Western Honshu

J9-11. Nankai Trough, 133°-139°E. The Nankai Trough marks the zone of interaction between the Philippine Sea and Eurasian plates in southwestern Japan. While the historie record of great earthquakes exist since A.D. 684, it is incomplete prior to A.D. 1707. ANDO (1975) and ISHIBASHI (1981) have divided the margin into a number of distinct segments or blocks which have ruptured either simultaneously or within a few days to years of one another. Previous great earthquakes (and the segments ruptured) incJude 28 October 1707 (M 8.4, ABCDE), 23 and 24 December 1854 (M 8.4, AB and CDE), 7 December 1944 (Mw 8.1, CD, segment J-IO), and 20 December 1946 (Mw 8.1, AB, segment J-11). See Figure 25 for the locations of these recent ruptures.

At present, the segment east of the 1944 Tonankai event (E, segment J-9) has been unruptured since 1854. This segment is called the Tokai gap, and is the present day focus of the Japanese earthquake prediction program (ISHIBASHI, 1981; MOGI, 1981). Prior to 1707, the rupture history of the Tokai gap was poorly known. During the 1707 Hoeii and 1854 Ansei earthquakes, the Tokai region appears to have ruptured in conjunction with blocks C and D. Hence it is not known if block E can rupture independently. Based on the one available recurrence interval that involved this block, 1707-1854 or 147 years, the probability for an event in the Tükai gap (J -9) is at the 16% level für the next 10 years. Für cümparisün,

Figure 25 Geographie place names, major tectonie features, and great earthquake source zones along southwest Japan (modified from NISHENKü and MCCANN, 1979). A, B, C, D, and E refer to the principal tectonie segments along the Nankai trough (after ISHIBASHI, 1981). Segments A, B, C, D, and E all ruptured within 32 hours of one another in 1854. During this century, segments A and B ruptured in 1946, C and D in 1944, and E (the Tokai gap) has not reruptured since 1854. The Tokai gap is eurrently the focus of the Japanese earthquake prediction program (see MOGI, 1981). N.T. is the Nankai trough and S.T.

is the Suruga trough.


applieation of the shorter recurrence histories for Kii and Nankai segments (blocks AB and CD, respectively), suggest a probability at the 31 % level for the Tokai gap. Probabilities for the Kii (J-IO) and Nankai (J-II) regions (which last ruptured in 1944 and 1946, respeetively) are at the ~ I % level for the next 10 years.

Izu-Bonin and Mariana Ares

IBM-I-Izu-Bonin and Mariana, 34°-lOoN. The Izu-Bonin and Mariana ares form the eastem edge of the Philippine Sea plate and the zone of interaction with the Paeifie plate. Neither GUTEN BERG and RICHTER (1954) nor ROTHE (1969) report any shallow earthquakes of magnitude greater than 7.3 along these ares between latitudes 10° and 35°S. The absence of great earthquakes and the smaller number of reported large shoeks along this margin have been diseussed by MCCANN et al. (1979) and KANAMORI (1977b).

Large shoeks do occasionally occur along this boundary as illustrated by the events on 22 September 1902 (originally listed as M = 8.1 by RICHTER (1958) and revised to M s 7.5 by ABE and NOGUCHI (1983» and 5 April 1990 (Ms 7.5). The 1990 event is, however, anormal faulting earthquake (PDE, 1990). With the possible exeeption of these events, both the instrumental and shorter historie record of the Marianas and Izu-Bonin ares give no indieation of the frequent occurrence of large thrust events that are typical of other subduetion zones with narrow plate interfaces. Based on these observations, we qualitatively assign this segment a low probability for the oeeurrenee of great earthquakes during the next 10 years.

Southwest Paeifie

The tectonies of the southwest Pacifie region are in response to the eonvergenee of the Pacifie and Australian plates. Subduetion of the Pacifie plate along a westerly dipping seismie zone extends from the southem Kermadec are to the northem end of the Tonga Islands. Transform and extensional tectonies eharaeterize the region between the northem end of the Tonga are and the southem portion of the New Hebrides subduetion zone. The Australian plate is subducted at the New Hebrides seismie zone whieh dips to the east. To the west, both Australian lithosphere and the smaller Solomon plate are subducted beneath the Solomon are. Along the northem eoast of New Guinea the plate motion is generally east-west, though a weil defined seismie zone is not present.

The rate of plate motion ranges from a low of 5-6 em/yr in the Kermadec region to near 10 em/yr in the northem portion of the Tonga are. Along the New Hebrides region, plate motion may be as high as ll em/yr, while plate motions near the New Guinea-Solomon region may be 10 em/yr. The following diseussion of the seismie potential of the southwest Pacifie is updated from that of MCCANN (1980) and is summarized in Figure 26 for the time interval 1989-1999.


14rrO~· ---------r--______ ~~~ ________ r-________ l~8~O.~ ________ r-______ ~~ 11 O.


!:::::::::j60.100'10 ~ 0.20%' _ 4()-6Q% ~ No ~JoIOtIc R_fll ~ I.:....!....:.J 0' GfUt EmhqulkH m 20-40% t'~,Mw l'J ~ HIIItWIc

PACIFIC OCEAN

~ _________ ~ __________ t-________ ~1~·

M,84

30

l-~~r::-t-----j1--140·

~_+----T--ISO·

Figure 26 Seismie potential of the southwest Paeifie, including New Britain, Solomon Islands, Vanuatu and Tonga-Kermadec: 1989-1999. Patterns portray the level of eonditional probability for occurrence of large and great (Ms 7.0 and larger) earthquakes during the next 10 years, 1989-1999, and range from haehured, 0-20%; eross-haehured, 20- 40%; open dots, 40- 60%; and filled dots, 60-100%. Small eross pattern denotes those areas with no historie reeord of great earthquakes. Specifie dates and magnitudes

refer to areas with ineomplete historie records.

New Britain and Solomon Islands

This can be divided into three distinct tectonic elements: east-northeast trending New Britain, northwest trending Bougainville, and the Guadalcanal-San Cristobal region. The junction of the New Britain and Bougainville arc segments is one of the most seismically active regions in the Pacific (see Figure 27), and may be a profitable site for earthquake prediction research during the next decade.

NB-I. Western New Britain, 148°-l5I oE. The section ofthe seismic zone near western New Britain and the Huon Peninsula has not been ruptured by earthquakes of magnitude 7.5 or greater since 28 December 1945 (Ms 7.9) and 6 May 1947


1400 E 1450

1400 E 1450

'71 (8.1) • -. ' 18 (7.6)

'26 (7.8) /D--"35 (8.1 ) '17 (7.9)

1500 1550 1600 165°E

PACIFIC ONTONG

. ..,.~ 0 CE A N l NEW NE~ELAND J:~:TEAU

BRITAIN r BOUGAINVILLE

~ r~~~RGIA ,.. -11~~\\'1 TR~."c. ~ ISLANDS SAN

,,~'II + \). ~

150"

'" r,.2 '" ~ ~,~ CRISTOBAL IS

~~~ ,DNLARK ~ 'J GUADALCANAL IS

RENNELL IS.

1550 160" 16S"E

'44 (7 .n • '46 (7.6, 7 .9)

'16 (8.0) cf) '13 (7.9)

/ '19 (8.1) '71 (7 .9, 8.0)

'38 (7.n ..J::~"'«.c~" '75 (7.8)

'47 (7 .7()J '55 (7.5) '06 (7.7) '43 (7 .5)

r---=::=:===::::::..._--~---,~ '75 (7.n ,,~ '39 (8.0

Ms >7.5 1900 - 1987

'36 (7.6)

<l&, "77 (7.5, 7 .8(2» ,

'. ~ '31 (8.0,8.1 )

'26 (7 .6)~ . ~ '39 (8.2) '00 (7 .8)

'26 (7.6) ' 10 (7 .5) ) '35 (7 .8)

• Relocated or HOF

o G & R (1953), R (1958) '66 (7.9) '84 (7.5)

Figure 27

239

50 S

10 0

5·S

100

Geographie place names and loeations of large and great earthquakes along the New Guinea-New Britain-Solomon seismie zone. (Top) Major geographie features of the New Guinea-Solomon Islands region . Bathymetry from CHASE er al. (1970), eontour interval: 1000 m. Shaded regions are deeper than three thousand fathoms. Solid triangles are aetive volcanoes (after KATSUI, 1971). (Bottom) Shallow earthquakes larger than M s 7.5 from 1900 to 1987. Eaeh event is labelIed with date and magnitude. Open eircles are loeation from GUTENBERG and RICHTER (1954) or RICHTER (1958). Solid eircles are either reloeated or taken from the National Earthquake Information Center Hypocentral Data File.

Haehured areas are aftershoek zones of more recent earthquakes. After MCCANN (1980).

(Ms 7.7). An event on 14 September 1906 (Ms 7.7) is located in this area, but could not be reliably relocated. If we assume that this earlier event was the predecessor to the events in the 1940's, the repeat time is approximately 40 years. Given this estimate, the tentative probability for recurrence within the next 10 years is at the 58% level.


Two compressional outer rise events occurred in this region in 1966 and 1973 and indicate that the trench outer rise region has been accumulating compressive stress over the last 25 years (CHRISTENSEN and RUFF, 1988). Presently, western New Britain stands out as the only segment along the New Britain-Solomon arc that has produced a compressional outer rise event.

NB-2. Eastern New Britain, 151 0 -153°E. In eastern New Britain earlier sequences of events (2 February 1920 (Ms 7.9); 29 September 1946 (Ms 7.7), 23 April 1953 (Ms 7.6); and 26 July 1971 (Ms 7.7» have repeatedly ruptured substantial portions of the plate boundary. Therefore recurrence times for this region appear to be as short as 25 ± 5 years, as noted previously by LA Y and KANAMORI

(1980). Given that the last event occurred in 1971, and the relatively short repeat time, the probability is at the. 59% level for the next 10 years.

S-I. Northern Bougainville, 1530 -154SE. The northern Bougainville segment has ruptured in aseries of great earthquakes that are closely linked in space and time with those in the eastern New Britain segment (1 January 1916 (Ms 8.0); 6 May 1919 (Ms 8.1); 29 September 1946 (Ms 7.9); and 14 July 1971 (Ms 7.8». There are uncertainties in the direction of a rupture of earlier events, and in fact, this event represents a gap filling earthquake. Both segments (NB-2 and S-I) appear to have average repeat times of about 25 years. Since both regions ruptured in 1971 and within a few years of one another in earlier episodes, the interchange of earlier earthquakes between these two regions, due to uncertainties in event 10cation and rupture direction, does not significantly change the recurrence times or probabilistic estimates for future great earthquakes. Like segment NB-2, the occurrence of events in 1971 and the relatively short 25-year recurrence time indicate a high conditional probability of 53% over the next 10 years.

S-2. Southern Bougainville, 154S -155 SE. Previous events in this segment occurred on 30 January 1939 (Ms 8.0) and 20 July 1975 (Ms 7.5, 7.6). While the epicenters of these earthquakes cluster near the SW part of Bougainville, the aftershock zone of the 1975 sequence clearly demonstrates that the rupture was to the northwest. The rupture direction of the 1939 event is not known, and we assume that it was also to the northwest. Based on this single 36-year recurrence, the probability over the next 10 years is estimated to be at the 10% level.

Tensional outer rise events occurred in regions NB-2, S-1 and S-2 following gap filling events in 1978, 1974 and 1975 and reftect the relaxation of compressive stress in the plate interface region (CHRISTENSEN and RUFF, 1989). Given the relatively short recurrence time for great earthquakes in the New BritainBougainville area, monitoring for changes in stress in the trench outer rise region will be of great use for narrowing the time windows of future earthquake forecasts. As of this writing, a moderate size event has recently occurred in the trench outer rise region between segments NB-2 and S-1 on 23 July 1988 (M s 6.7). Analysis of this event indicates it to be a tensional outer rise event (PDE, 1988)

Vol. 135, 1991 Circum-Pacific Seisrnic Potential 241

and indieates that stresses in this region have apparently not yet ehanged in preparation for the next episode of great underthrust earthquakes.

S- 3. New Georgia, 155S -157°E. A large event appears to have broken this segment of the plate boundary on 19 April 1936 (Ms 7.4). Its position between the seismieally aetive Bougainville and inaetive Woodlark segments of the Solomon are suggest that this is a transitional region. Comparison of magnitudes and repeat times for other regions in the Solomon are indieate that this region should have reruptured sinee 1936, if the 1936 event was an underthrust earthquake. At present insufficient information is available to explain this diserepaney or to eonfidently estimate a reeurrenee time.

S-4. Woodlark Basin, 157°-159SE. Within the Woodlark Basin lies the boundary between the Solomon Sea and Australian plates. The entire Woodlark Basin, a region of extensional tectonics, has not experieneed large events sinee the turn of the eentury, and is eonsidered to have a low potential for such events. The segment of the Solomon are beneath whieh this young seaftoor is subducted (region S-4) has also maintained its quieseent nature for large and great earthquakes sinee 1900. Regions of aetive seaftoor spreading or extensional tectonics generally do not produee large or great earthquakes as often as eonvergent or trans form boundaries (KELLEHER et al., 1973; MCCANN et al., 1979). Based on the lack of large or great earthquakes during this eentury, and the unique tectonic setting, region S-4 is qualitatively assigned a low probability for large earthquake occurrenee.

S- 5. Guadalcanal, 159° -161 oE. This segment of the Solomon plate boundary exhibits a clear cluster or souree for epieenters that have only been observed to propagate to the northwest. When taken as a whole, the Guadalcanal region has seen two episodes of high earthquake aetivity, 1926-1935 and 1977 to the present (1988). Therefore reeurrenee times appear to be about 50 years, or about twiee that of the New Britain - Bougainville segments. Given the estimated similarity in eonvergenee rates (about 10 emjyr) and in the sizes of earthquakes in both regions, this differenee may indieate that more aseismie slip is occurring along the southern portion of the Solomon are.

Because of our uneertainty in the loeation and meehanism of this earlier sequenee of earthquakes, it is not apparent that the present sequenee of earthquakes in this region has terminated. For eomparison, the eumulative seismie moment release in the interval 1926-1935 is approximately 2.3 x 1028 dyne-em, while the moment release in the eurrent sequenee is about 33% of the earlier (Le., 7.5 x 1027 dyne-em). Henee the end of the eurrent sequenee may still lie in the future. Continued aetivity in this region on 10 August 1988 (Ms 7.5) is eonsistent with this pattern. Aeeordingly, the probability for future aetivity is estimated to be at the 45% level.

S-6. San Cristobal, 161°-163°E. The San Cristoba1 region last ruptured in a great earthquake on 3 Oetober 1931 (Ms 8.0). No reeurrenee intervals have been observed for this segment during this eentury. Arecent 1arge event on 1978


(M s 7.1) appears to have ruptured the updip portion of the 1931 rupture zone; however, the major portion of this segment appears to be unbroken sinee 1931 and therefore eonsidered a likely eandidate for a future great earthquake. Using the reeurrenee time from the adjaeent Guada1canal region (S-5), 50 years, as an estimate for this region, the probability is estimated to be at the 45% level for the next 10 years.

No eompressional or tensional outer rise earthquakes are listed in the eatalog of CHRISTENSEN and RUFF (1989) for either the Guada1canal or San Cristobal segments. Given the uneertainty of reeurrenee time estimates for this region, monitoring the region for eompressional outer rise earthquakes would help to narrow the foreeast time windows.

S- 7. SE Solomons, 163° -165°E. The far southeastern portion of the Solomon island are defines the transition zone between the Solomon and Vanuatu plate boundaries, and has been nearly aseismie for large earthquakes during the last 30 years. One event has been assigned to this region (29 July 1900 (Ms 7.6», but the uneertainty in loeation is large. A more detailed understanding of the nature of this region will be neeessary before we ean reliably estimate the seismie potential.

Santa Cruz and Vanuatu Islands

VA-la,b. Santa Cruz and Vankolo Islands, l1°-13°S. Events on 18 July 1934 (Ms 8.3); 31 Deeember 1966 (Ms 8.1); and 17 July 1980 (Ms 7.9) have ruptured similar segments of the are (see Figure 28), but display different details in the rupture proeess. Events in 1934, 16 November 1944 (Ms 7.3) and 1966 have nearly identieal epieenters, suggesting similarities in the initiation of rupture at 11.9°S. The 1980 event ruptured the asperity that aeted as a barrier for the southern expansion of the 1966 rupture (T AJIMA et al., 1990) This region is named the Vankolo asperity and appears to spawn major earthquakes every 10-20 years. If the adjoining segments of plate boundary have stored enough strain energy, then rupture of the Vankolo asperity may spread into the adjoining region, resulting in a great rather than a large earthquake. The aftershoek zone of the 1934 earthquake overlaps both the 1966 and 1980 aftershoek zones (MCCANN, 1980). Based on the historie reeord, we would expeet the Vankolo asperity to rupture in a large event before the year 2000, and estimate the eonditional probability at 83%. The reeurrenee time for larger (i.e., great) events appears to be about 30 years, henee the next failure may initiate a rupture similar to that in 1966, and eonditional probability for a great event is at the 48 % level. If the 1980 earthquake reset the stress level in this region, the resulting probability would be lower still.

VA-2. Torres Islands, 13°-14°S. While no earthquake activity is known for the Torres Islands segments during this eentury, the occurrence of a eompressional outer rise event in 1985, near the northern end of the eollision zone between the East Rennell Island ridge and the are (CHRISTENSEN and RUFF, 1989), may


10·S

15·

20·

", ~sA~IAA~~~Z ~ ~

~VAN~OROISLANO RenneIl ... Island "l I \ < TORRES ISLANOS R,dge \~ /

r""2 \ '.

F I J I

p A ' ~ SANTO ISLAND

/ \J \~~ .;1 ~MALEKULAISLAND

Denlrecosloux \..-ol "'i\.. Frocture Zone r 0 ~

I iTANNA ISLANO

~ , ) ............ COrlOIl5 T,ough

L A T E A U

-:-ANEITYUM ISLAND

MATTHEW ISLAND

j __ ~ v

HJNTER I$lA"-'\)

10·S

15·

20·

175·

Figure 28

C

1961 (7,5)· n () .1943 (7 ,5)

1920 (8.1) •

ß~ 1901 (8,1)

,6:, ' 1928 (7.8)

~ 1943 (7.6)

170·

243

Geographie place names and loeations of large and great earthquakes along the New Hebrides-Vanuatu seismic zone. (Left) Major geographie features of the New Hebrides-Vanuatu Islands region. Bathymetry from CHASE et 01. (1970), eontour interval: 1000 m. Shaded regions are deeper than three thousand fathoms. Solid triangles are aetive volcanoes (after KATSUI, 1971). (Right) Shallow earthquakes larger than Ms 7.5 from 1900 to 1987. Eaeh event is labelIed with date and magnitude. Open eirc1es are loeation from GUTEN BERG and RICHTER (1954) or RICHTER (1958). Solid eirc1es are either relocated or taken from the National Earthquake Information Center Hypocentral Data File. Haehured areas are

aftershoek zones of more recent earthquakes. After McCANN (1980).

provide a basis for qualitatively upgrading the hazard in this segment. Reeonnaissanee studies of uplifted reef terraees on the Torres Islands by TA YLOR et a/. (1985)

in 1977, indieated that the ages .and altitudes of major terraees were eorrelated with Holoeene and interglacial sea level ehanges, and no evidenee of uplift was found within reeent deeades. Overall, the rate of uplift throughout the Torres Islands is about I mmjyr, and is a faetor of 2-3 slower than the rates observed on Santo and Malekula (see below).

VA-3. North Santo [sland, 14°-15.3°S. Emerged eorals on northwest Santo Island indieate that at least one event in 1865 ± 6 occurred prior to the 23 Deeember 1973 (Ms 7.5) earthquake (TAYLOR et al., 1990; EDWARDS et al. , 1987). The amount of uplift in this earlier event is twiee that of the 1973 event (0.6 vs 1.2 m), suggesting that the 1865 event was twiee the size, or that there is a missing event. Henee reeurrenee times for this segment of the Vanuatu are range from 54 to


108 years. Due to the receney of faulting, both estimates indicate reJatively low probabilities for this region over the next 10 years ( =:;; 1%).

VA-4. South Santo and Malekula Island, 15.3°-16.3°S. The boundary between the north and south Santo seismie zones eoincides with the region where the D'Entreeasteaux ridge intersects the New Hebrides treneh. As suggested by PASCAL et al. (1978) and TA YLOR et al. (1980), it is likely that the D'Entrecastaux fraeture zone represents a strong struetural boundary aeross whieh the rupture zones of large earthquakes generally eannot propagate.

Two large earthquakes on 5 January 1946 (7.3, 7.0) appear to have ruptured similar segments of south Santo Island that broke during events in II August 1965 (Ms 7.5) and 27 Oetober 1971 (Ms 7.1). The distribution and amounts of eoral emergence on south Santo suggest that the 1946 events are most similar to the eombined effeets of the 1965 and 1971 events. Henee the probability for the occurrenee of M s 7 events, based on the historie reeord, is at the 60% level. TAYLOR et al. (1990) note that this observed 19-25-year recurrenee is shorter, by a faetor of two, than the recurrenee inferred from dividing the average eoseismie uplift (0.28 m) by the long-term uplift rate (5.5 mm/yr). In the latter ease, the estimated reeurrenee time is 51 years and the eorresponding probability is at the 9% level.

In addition to the 1946-1965/71 events on South Santo Island, eoseismic uplift during the 1965 earthquake also occurred on Malekula Island. The next highest eoral terraee on Malekula has 230Th dates of 1718-1729 A.D., and is loeally uplifted by an amount similar to the 1965 event. The eoral data suggest long (240-year) reeurrence times for 1965 sized events along this portion of the Vanuatu are, and is in good agreement with the estimated reeurrence time based on rates of uplift. ConsequS!ntly, the probability for future events is relatively low over the next 10 years ( ~ 1%).

VA-5. Central Vanuatu, 17°-200 S. To the south there are fewer large shoeks along the plate boundary, and the plate eonvergence may be accommodated aseismieally. Little is known about the great 20 September 1920 (Ms 7.9) earthquake. It is loeated near the treneh axis and generated a tsunami observed on Samoa (ImA et al., 1967). The loeation of this event suggests that it may have been anormal faulting earthquake, signifying redueed eoupling in this area. No specifie forecasts for future behavior are possible at this time.

VA-6. Loyalty Islands, 200 -23°S. In eontrast to the central Vanuatu, the Loyalty Islands portion of the Vanuatu are exhibits frequent large earthquakes (VmALE and KANAMORI, 1983). Prior events on 16 March 1928 (Ms 7.6); 14 September 1943 (Ms 7.4); and 25 Oetober 1980 (Ms 7.2) indicate a short reeurrenee time of about 24 ± 4 years. In addition, on event on 9 August 1901 (Ms 7.9) is also loeated near this segment; however, this epieenter needs to be verified. Based on the pattern of oeeurrenee of post-1901 events, the probability for a large event in this segment over the next 10 years is at thc 22% level.


VA-7. Hunter Island, 171°-174°E. This segment is bounded by the Loyalty Islands segment to the west and the Fiji rift to the east. As shown in Figure 28, there is no history of great earthquakes in this segment, and hazards estimates similar to others in this study cannot be made at this time.

Tonga and Kermadec Islands

In contrast to many of the other subduction zones covered in this study, large areas of the Tonga and Kermadec arcs are distinguished by the lack of demonstrable repeats of shallow interplate earthquakes during this century. Both Tonga and northern Kermadec were sites of aseries of great earthquakes from 1907 to 1919. These events are poorly located, their focal mechanisms are not known, and if assumed to be interplate events, cannot cover the entire plate boundary. Given the present incomplete state of knowledge for this region, we will attempt to single out those areas where recurrence estimates can be made. Other areas will, for the present, have to be regarded as suspect. These suspect areas include Tafahi Is. (TK-l), Eua Is. (TK-2), Ata Is. (TK-4), and the junction of the Tonga and Kermadec trench systems (TK-6). Recent rupture zones are shown in Figure 29.

TK-3. Tongatapu, 20o -22°S. Of the seismic gaps in the Tonga arc, the region from 20° to ~2°S is clearly identified as one of the high seismic potential by the observations of a) one recurrence interval (2 January 1907 (Ms 7.6) or June 1913 (Ms 7.7) to 8 September 1948 (Ms 7.8),35-41 years) and the fact that 40 years has elapsed since the last event, b) three compressional outer rise events (1982-1986) (CHRISTENSEN and RUFF, 1988) and c) an area of high level background seismicity (WYSS et al., 1984). WYSS et al. (1984) identified 3 areas of high level background seismicity along the Tong- Kermadec arc. Two of these high background seismicity areas were spatially coincident with and preceded the occurrence of earthquakes in 1975 and 1976 in regions TK- land TK-8. On the basis of the historic seismicity and estimated 38-year repeat time, this area has a probability of 58% over the next 10 years, which appears to be consistent with the observations of other intermediate-term phenomena.

TK-5. Louisville Ridge Intersection, 23°-25°S. Collision with the Louisville Ridge has indented the arc geometry in this area. The great (Ms 7.7) 19 December 1982 interplate thrust earthquake was preceded by a compressional outer rise event in 1974 (Mw 7.4) and intermediate depth normal faulting events in 1977 (M w 8.0) and 1980 (Mw 7.6) (CHRISTENSEN and LAY, 1988). No great earthquakes are located in this region prior to the 1982 event, suggesting that repeat times in this region are on the order of 80 years or more. Using 82 years as a tentative repeat time estimate, the conditional probability, given the recency of faulting, is less than 1 % for the next 10 years.

TK-7,8. Northern and Southern Kermadec Islands, 27°-28S and 28S-30.2°S. This region experienced two great events on 1 May (Ms 8.1) and 16 November 1917

246 Stuart P. Nishenko

s.q: .... 15' S ~os

Q • 15"S :q /s

\s\..~ .& I <.q ~\~\ -:;:Z, TAFAHI IS . .--'· '1'0s

o LAU /i~NEIAFU IS. I ... BASIN _.:; .

?o· !(l 3 ?o· TONGATAPU IS. .& <ff

;1 ATA IS--

2" SOUIH 10"

FIJI

BASIN

RAOUl IS. .,

' '».f (Q ,0- 6'~ 3O' ~

& /d "'.<-~ ' e;, f ~ ~

~ ~ ." .Ii' "Cl

Q:- ~ j!lt ~ .'"

Figure 29

-:;:z, p ...

PAGEOPH,

1 1981 (7.7) "t;:t-ca,_

'.~7 (7,6;.'), '9'1 (8.")~'975 (7.7)

• •• 77 (7.6)

'9 •• (S,4) '. r/.. .

19'3 (7.9)~ '-'i. . , V '906 (7.5)

.907 (7.6) Q •

•• 48 (8m

I f:'982 (7.7)

.975 (7.9)

'917~S") .988(81)

,9'7 P.?) •• 955 (7.9)

2 ' • 1978 (8 1) ,~ '976 (19) . ,

1943 (7,8) ,

o '910 (7.5)

M. >7,~ 1900 - 1987

• Rdlocated or HOF

o G & R (1953), R (1958)

Geographie place names and loeations of large and great earthquakes along the Tonga-Kermadec seismie zone. (Left) Major geographie features of the Tonga-Kermadec Islands region. Bathymetry from CHASE er al. (1970), eontour interval: !OOO m. Shaded regions are deeper than three thousand fathoms. Solid triangles are aetive volcanoes (after KATSUI, 1971). (Right) Shallow earthquakes larger than Ms 7.5 from 1900 to 1987. Eaeh event is labelIed with date and magnitude. Open eirc1es are loeation from GUTENBERG and RICHTER (1954) or RICHTER (1958). Solid circ1es are either relocated or taken from the National Earthquake Information Center Hypoeentral Data File. Haehured areas are after-

shoek zones of more recent earthquakes. After MCCANN (1980).

(Ms 7.7) and again on 14 January 1976 (Ms 8.1, 7.9, within 50 min) and 20 October 1986 (Ms 8.2). The 1986 event resulted from the internal deformation of the Pacifie plate in response to interplate eollision (LUNDGREN et al., 1989). Preliminary investigations of the 1976 earthquakes indicates that the first event is of the underthrust type (S. Beck, personal eommunication, 1990) and may be an interplate earthquake. Given these observations, and the laek of reliable loeations and meehanisms for the earlier 1917 events, it is diffieult to eonfidently assign either a reeurrenee time or probability for these two segments of the Kermadee are. If the 1976-1917 sequenee does represent a reeurrenee, then the probability for the next 10 and 20 years is less than 1% and 3%, respectively.

TK -9. Southern Kermadec, 30.2° -38°S. The southern part of the Kermadee are is distinguished by the lack of large and great earthquakes, and low levels of background seismicity during this century. This segment coincides with the weIl


developed back-are spreading of the Harve Trough, and the section of the are with the lowest convergence rate. UVEDA and KANAMORI (1979) related the slow subduction of old seafloor to aseismic subduction, and this region may be a good example of that phenomena. Based on these observations, we have qualitatively assigned this section of the Kermadec arc a low probability for the occurrence of great earthquakes in the near future.

Summary

We have examined and characterized the seismic potential for 123 segments of simple plate boundaries around the circum-Pacific region. Ninety-six of these 123 segments have sufficient information to estimate recurrence times and seismic potential. This report represents the first probabilistie comparison of seismic hazards on this scale, and updates earlier work on seismic gaps by KELLEHER et al. (1973) and MCCANN et al. ( 1979) by including new information on earthquake repeat times, the proximity to the next expected earthquake, and a time-dependent probabilistic model to describe earthquake hazards along simple plate boundaries. Variations in observed earthquake magnitudes, recurrence times, and the completeness of the historie record for individual plate boundary segments have resulted in a diverse data set fOf-the circum-Pacific region. The probabilistic framework adopted in this study, in addition to accounting for uncertainties in the individual data, has provided a common base for comparing regional seismie hazards, and being able to rank gaps in terms of those most likely to rupture in the next 5, 10, and 20 years.

The estimates presented in this study are based on conditional probability and the primary results summarized in Plate 1, Figures 6, 10, 15, 19,21,23,26, and the Appendix cover the time interval 1989-1999, which overlaps the International Decade of NaturalDisaster Reduction. Table I is a ranked list of those gaps with greater than a 50% probability of recurrence within the next 10 years. This list includes 11 zones, or 11 % of the total number of gaps with hazard estimates. The majority of gaps on this list are characterized by relatively short (Le., < 50 years) recurrence times and the IO-year interval (1989-1999) covers a substantial fraction of possible recurrence times for those gaps. Table 2 presents a ranked list of gaps with a greater than 50% probability of recurrence in the next 20 years ( 1989-2009). This list is considerably larger, and includes 30 out of 96 gaps, or 31 % of our sampIe. While many of the gaps in Table 2 have Ion ger repeat times, the combination of time elapsed since the last event and a larger time window, has resulted in a larger list. Many more gaps in Table 2 than in Table I occur near urban centers. The 20-year list is useful for longer-term forecasting and may be a more reliable identification of areas with high or significant hazard levels.

One of the basic postulates used in this study holds that the variation in repeat times for characteristie, plate boundary earthquakes is proportional to the average


repeat time. The implication of varying degrees of temporal resolution impacts the planning of earthquake prediction programs or experiments, hazard mitigation work, and the funding and expected lifetimes of research pro grams. Given the inherent large uncertainties in many hazards estimates, which in some instances cover decades, research and hazard mitigation programs must be organized with these limitations in mind. Some programs, such as geodetic, strong motion, and seismological monitoring can be instituted and carried out for decades. Hence, the long-term estimates presented here can be considered a "blueprint" for future research. Other research pro grams, wh ich may focus on intermediate or short-term physical phenomena or have shorter life spans due to logistieal or instrumental reasons, should await deployment until the time window is sufficiently narrowed.

Earthquake forecasting is a rapidly developing field. Hence, in addition to updating these forecasts, due to the time-dependent nature of the problem, updates will be necessary as new information becomes available and improvements are made in the probabilistie models. Further improvements in the direct method of estimating recurrence times are necessary for areas with one or no observed repeat times. Twenty-seven areas are currently categorized as having either an incomplete or no historic record of large or great earthquakes. Additional recurrence time data would help to clarify hazard estimates for those segments. The observations of significant variations in the size of "gap-filling" earthquakes from cycle to cycle along many plate boundaries indicates that spatial as weil as temporal uncertainty needs to be addressed in the future. 80th spatial and temporal uncertainty ultimately affect seismic risk estimates and decisions conceming the type of mitigation effort needed for urban areas near hazardous seismic gaps.

While this study has summarized the seismic potential for a large number of seismic gaps around the circum-Pacific area, there remain a number of geographic and seismotectonic regions whieh need to be included. Additional geographie regions include Indonesia, the Philippines, New Zealand, and the cöuntries which surround the Caribbean basin and the Mediterranean. The occurrence of shallow, large earthquakes in voIcanic valleys (e.g., EI Salvador, Nicaragua, Colombia, and Ecuador) mayaiso be amenable to the probabilistic approach used in this study. Earthquakes in these regions pose an even greater risk to life and property due to high population densities.

Tsunamis associated with the occurrence of great, shallow plate boundary earthquakes pose a threat to the entire circum-Pacific community. Knowledge as to when and where great earthquakes can be expected is invaluable towards developing a tsunami forecast for the circum-Pacific. This type of forecast would provide a basis for the development of an active, rather than passive, tsunami mitigation program.

Finally, attention needs to be given to the development of areal-time earthquake warning and advisory system. The time windows for the forecasts presented here need to be narrowed, as additional da ta become available, to realize the uItimate goal of earthquake forecasting and prediction research-saving lives and


property. Advanees in the analysis of seismologie data, eoupled with satellite eommunieations, provide the opportunity for rapid data eolleetion, interpretation, and evaluation. The development of a global, real-time, seismie hazards network eould be an aehievement that would eontinue the goals of the International Deeade of Natural Disaster Reduetion.

Appendix

Summary 0/ Circum-Pacific Probability Estimates

Conditional Probabilill: Location Date of Magnitude Forecast 1989-1994 1989-1999 1989-2009 Method

Last Event Window

SOUTH AMERICA

Chile C-I Tierra dei Fuego 65°_72°W 1949 7.7 1995-2064 4% 11% 29% H

C-2 Chilean Archipelago 46°_55°5 No historie record of great earthquak ..

C-3 Southern Chile 40°_46°5 1960 9.4 2043-2148 :;1% <1% :;1% H

C-4 Concepcion 35°_40°5 1939-1960 7.9-8.3 2002-2078 1% 3% 12% H

C-5 Valparaiso 32°_34.5°5 1985 8.0 2041-2109 :;1% :;1% :;1% H

C-5a Pichilemu-Llico 34.5°_35°5 1906 (7.5-8.0) 1990-2044 17% 33% 59% H

C-6 COQuimbo- Los Vilos 30°_32°5 1943 8.2 1991-2049 11% 24% 49% H

C-7 Ata.cama 26°_30°5 1922 8.5 2002-2128 2% 4% 10% H

C-8 Taltal-Copiapo 25°_27°5 1978 7.3 2005-2038 :;1% :;1% 15% H

C-9 Paposo 25°_24°5 No historie record of greal earthquakes

C-IO Arica-Antoragasta 19°_24°5 1877 9.0 (1992-2074) (10%) (20%) (39%) H.D

Peru P-1 Arica 19°_16.6°5 1868 9.0 (1991-2300) (:; 1-12%) (:; 1-23%) (:; 1-43%) H P-2 Camana 16.6°_15.8°5 1913 7.8 1993-2073 6% 13% 29% H P-3 Nazca 15.8°_14°5 1942 8.2 (2109-2396) (:; 1%) (:; 1%) (:; 1%) H

P-4a Lima-Pisco 12°_14°5 1974 8.0 2028-2097 :;1% :;1% :;1% H P-4b Pisco-Ica 12°_14°5 1813 (1991-2079) (14%) (28%) (47%) H P-5 Chimbote-Lima 10°_12°5 1940/1966 8.2,8.1 1996-2089 :'01-3% :'01-8% 1-24% H P-6 Chimbote-Guayaquil 0.5°_10°5 No historie record of greal earthquakes

Ec:uador-Colombia EC-1 Jama 0.5°5_1.2°N 1942 7.9 (1990-2015) (25-41%) (48-66%) (78-90%) H EC-2 Esmeraldas 1.2°_1.7°N 1958 7.8 1993-2044 8% 19% 46% H EC-3 Tumaco 1.7°_4°N 1979 7.7 2023-2098 :'01% :'01% :;1% H EC-4 Buenaventura 4°_7.5°N No historie record of greal earthquakes

CENTRAL AMERICA Panama

P-I ~an Blas 77°_79.5°W 1882 7.5± Incomplete historie record

C'gsta Rica CR-1 Papagayo 87°_86°W 1916 7.5 (1991-2051) (16%) (31%) (55%) H CR-2 Nicoya 86°_85°W 1978 7.3 1993-2010 20% 64% 98% H CR-3 Quepos 85°_84°W 1952 7.0 1996-2067 3% 8% 25% H CR-4 Osa 84°_83°W 1983 7.3 2010-2041 :'01% :'01% 4% H

Nicaragua and EI Salvador E~-I C. EI Salvador 89°_88°W 1926 7.2 Incomplete historie record N-I West Nicaragua 88'-87°W 1921 7.4 Incomplete historie record N-2 East Nicaragua 87°_86°W 1898-1907 7.0-7.2 Incomplete historie record

Guatemala G-I Mot.gua 88.3°_91°W 1976 7.5 2392-3117 :'01% :;1% :;1% G C-2 E. Chixoy-Polochic 89.00 _90.5°W 1785 (7.3) (1996-2207) (4%) (8%) (15%) H,G G-3 W. Chixoy-Polochic 9O.5°_91.5°W 1812 (7.5) ( 1996-2207) (4%) (8%) (15%) H C-4 Southeast 89°_90.5°W 1915 7.4 1990-2025 29% 51% 79% H G-5 Central 91.5°_90.5°W 1942 7.9 1992-2041 10% 23% 50% H C-6 Northwest 92.5°-9I.5°W 1942 7.9 (1994-2056) (5%) (13%) (34%) H


Appendix con't. Summary of Circum-Pacific Probability EsUmates

Conditional Probabili~

Localion Date of Magnitude Forecast 1989-1994 1989-1999 1989-2009 Method Last Event Window

~ENTRAL AMERICA Mexi~Q

M-I Chiapas 92.5°_94°W 1902 7.8 (2008-2151) (1%) (2%) (5%) H M-2 Tehuantep('c 94°-95.2"W No historie record of great earthquakes M-3 East Oaxaca 95.2°-96.4 ° W 1965 7.8 1991-2026 15% 35% 70% H M-4 Central Oaxaca 96.4°_97.3°W 1978 7.8 2013-2060 ::;1% ::;1% 2% H M-5 Central Oaxaca 97.3°_97.7°W 1928. 7.8 (1990-2032) (25%) (45%) (72%) H M-6 W('~l Oaxaca 97.7°_982°W 1968 7.4 1994-2025 6% 21% 64% H M-i Ometf'pec 98.2°_99.3°W 1950 7.3 1990-2030 26% 47% 74% H M-8 Acapulco-San Marcos 99.3°_IOOo W 1957 7.7 1994-2042 5% 13% 40% H M-9 Central Cuerrero looo_lOl oW 1899-1911 7.8 ( 1990-2068) (16%) (30%) (52%) D

M-IO Petatlan IOl o_101.8°W 1979 7.6 2001-2038 ::;1% 3% 29% H M-II Micho&can IOIY _103° W 1985 8.1 2029-2106 ::;1% :0 1% :0 1% H M-12 Colima 103° _103.7° W 1973 7.5 1993-2025 8% 25% 66% H M-Il Colima Gap 103.7°_104.5°W No historie record of great earthquakes M-I·I Jalisco \04.3°_105.7°W 1932 8.2 1992-2129 1-9% 2-18% 7-37% D,H

NORTH AMERICA California

SA-I Olt'ma 40.4°_37.5°N 1906 8.0 2049-2452 :01% -:; 1% 1% D SA-2 Peninsula 37.5°-36.8°N 1906 (7.0) (1992-2218) (2-8%) (5-15%) (11-30%) D SA-3 Santa Cruz 37°-36.8°N 1906 (6.5) (1995-2165) (5%) (9':6) (19%) D SA-4 Central Crreping Zone 36.8°-35.7° N No historie record of grcal earthquake!'i

SA-5 Parkfi.ld 35.7°_37°N 1966 6.0 1989-2000 70% 93% >99% H SA-6 Cholame 35.7°_35.3°N 1857 (7.0) (1994-2234) (5%) (11%) (20%) [)

SA-7 Carrizo 35.3°_34.7°N 1857 8.0 2018-2370 1% 1% 3% 0 SA-8 Mojave 34.7°_34.3°N 1857 7.5 1993-2179 6% 11% 22% [)

SA-9 San Bernardino 34.3°_33.8°N 1812(') (7.5) (199.1-2:1.18) (4%) (8%) (15%) [)

SA-IO Coachena 33.8°-33.4°N 1680 7.5 1993-2167 7% 11% 26% G

Waehington.Oregon WO-I SW Washington 48.3' -45 .. 1 N 1680 (8.0) (2080-27 50) (s: 1%) (:0 1%) (::; 1%) G

NOR.'t."---~A...!:.U'lC Queen Cbarlotte-Alalka-Aleutianl

QCAA-I Cp. St. James 51.7°_52.4°N >1898(') (7.6) No historie record QCAA-2 Queen Charlotte 52.4°_56°N 1949 R.I 2000-2152 2% 4% 11% D QCAA-3 Sitka 56°_58°N 1972 7.3 2002-2042 ::;1% 2% 22% H QCAA-4 Lituya Bay 58°_60.5°N 1958 8.2 2002-2158 1% 4% \0% D.G QCAA-5 Yakutat 139°_142°W 1899 (8.0) (1992-2096) (\0%) (19%) (36%) D QCAA-6 Yakataga 142°_145°W 1899 (8.0) (1991-2094) (11%) (21%) (39%) D QCAA-7 Pr. William Sound 145°_156°W 1964 9.2 2270-2645 :01% :0 1% :0 1% D QCAA-8 Kodiak I •. 1500_155°W 1964 (8.0) (2002-2051) (8%) (16%) (29%) H QCAA-9 Alaskan Penn. 156°_158.5°W 1938 8.0 1993-2051 8% 18% 41% H

QCAA-IO Shumagin Is. 158.5°-161.7°W 1917 7.4 1990-2029 27% 48% 75% H QCAA-II Unimak Is. 161.7°_164°W 1946 (7.4) (1990-2047) (25-31%) (43-55%) (67-82%) H.D QCAA-12 Fox Is. 164°_173°W 1957 (7.4) (1990-2022) (22%) (44%) (78%) H QCAA-13 Andreanof Is. 173°·177°W 1986 8.0 2013-2049 ::;1% ::;1% 1% H QCAA-14 Del.rof Is. 177°W-180° 1957 (7.4) (1989-2004) (59%) (85%) (98%) H QCAA-15 Rat I •. 1800_171°E 1965 8.0 2001-2061 1% 4% 17% H QCAA-16 Kommandorski Is. 171°_165°E 1858 (8.0) Incomplete historie record

WESTERN PACIFIC

K.Hnchatk~ KK-I Kamchatsky Pen. 55°_57°N 1971 7.5 1993-2028 7% 23% 61% H KK-2 NE Kamchatka 53°_55°N 1923 8.3 2004-2137 1% 3% 8% H KK-3 SE Kamchatka 49°_53°N 1952 9.0 2021-2116 ::;1% ::;1% 1% H KK-4 Shiashkotan Is. 48°_49°N 1915 8.0 ( 1992-2075) (10%) (20%) (40%) D KK-.\ Central Kurile 46°_48°N No historie record of great earthquakes KK-6 N. Vrup ls. 150.5°-152.5'E 1963 8.5 (1993-2037) (8%) (21%) (52%) H KK-7 S. Urup Is. 149°-150.5°E 1963 8.5 (2073-2262) ::;1% ::;1% ::;1% H KK-R Etorofu I:<; 148°-I.\00E 1958/78 8.3/7.6 (2011-2080) (:0 1%) (:0 1%) (3%) D KK-9 Shikotan Is. 146.5°_148.5°E 1969 8.2 2015-2093 ::;1% ::;1% 2% H,D -------_.


Appendix <on't. Summary of Circum-Padfl< ProbabIlity Eotlmateo

Conditional ~IOHbilil! Location Date of Magnitude Forec .. t 1989-1994 1989-1999 1989-200II Method

Last Event Window

WESTERN PAClPl!;; flokkaldo " E. Honlhu

J-I Nemuro-Oki 146.5'-147'E 1973 7.8 202G-2102 :51" :51" 1" H J-2 Tokachi-Oki 144.5'-146.5'E 1952 8.1 2018-2130 :51" :51" 2" H J-3 Tokachi-Oki 142'-IH'E 1968 8.2 2031-2112 :51% :51% :51% H J-4 8anriku-Oki 37.7'-39'N 1897 7.6 1991-2109 3-11" 7-21% 17-39% H,D J-5 Miyagi-Oki 37.5'-39'N 1978 7.4 2004-2034 :51% :51% 17% H J-6 8hioYa-Oki 36'-38.5'N 1938 7.1-7.7 (2340-4360) (:51%) (:51%) (:51%) H J-7 Japan·lzu Bonin junction 35'-36'N 1953 7.5 (2162-2521) (:51%) (:51%) (:51%) H J-8 Sagami Troulh &: 8oso Pen. 35'-36'N 1703/1923 8.2/7.9 2060-2370 :51% SI% SI% G,D

SW Honshu J-9 Tokai 138'-139'E 1854 (8.4) (1990-2112) (8-17%) (18-31%) (30-53%) H

J-1O Kii 136'-138'E 1944 8.1 2018-2120 SI" SI% 2" H J-II Nankai 133'-I36'E 1946 8.1 2021-2124 :51% :51% 1% H

)zu Bon In " Mariana. IBM·1 [zu Bonin·Mariana 34'_10' N No historie record or II'ftI urlhquakes

SOUTHWEST PACIF]C N_ Brltlan-Solomon

NB-I W .. I New Brilain 148'-151'E 1945 7.9 (1990-2021) (34%) (58%) (84") H NB-2 E .. I New Britain 151'-153'E 1971 7.7 1990-2012 29% 59" 92% H 8-1 North Boulainville 153'-154.5'E 1971 7.8 1991-2013 23% 53% 90% H 8-2 80uth Boulainville 154.5'-155.5'E 1975 7.8 1997-2034 2% 10% 44% H 8-3 New Georgia 155.5'-157'E 1936 7.4 Incomplete historie record 8-4 Woodlark Basin 157'-159'E r No historie record of Ireat earthquakes 8-5 Guadalcanal 159'-161'E 1988 7.5 1990-2034 25% 45% 71% H 8-6 San Cristobal 161'-163'E 1931 8.0 (1990-2034 ) (25%) (45%) (71%) H S-7 SE Solomons 163'-165'E (1900?) Incomplete historie rec:ord

Vaguatu VA-la Santa erul 1I'-13'S 1966 8.1 1990-2020 23% 48% 82% H VA-Ib Vankolo I •. 1I.9'S 1980 7.5 1990-2003 41% 83% ~ 99% H VA-2 Ton .. 10. 13'-14'S No historie n~cord o! great earthquak .. VA-3 N. Santo 10. 14'-15.2'S 1973 7.5 2036-2152 SI% SI% SI% G VA-4 S. Santo and 15.3'_16'5 1971 7.1 1990-2013 30% 60% 91% H

Malekula 10. 15'_17'5 1965 7.5 21(l9-2359 SI% :51% SI% G

VA-5 Central Vanuatu 17'_21'5 Incomplete historie record VA-6 LoYalty Is. 21'_23'8 1980 7.2 1995-2016 3% 22% 80% H

VA-7 Hunter 15. 17I'-174'E No historie record of great earthquakes

ToD&@-Kermadec TK-I Tafahi Is. 15'_17.5'8 1917 8.4 Incomplete historie record TK-2 Eua I •. 17.5'-2O'S 1919 8.2 lncomplete historie record TK-3 Tongatapu 20'_22'8 1948 7.8 1990-2020 33% 58% 84% H TK-4 Ala Is. 22'_23'8 Ineomplete historie reeord TK-5 Louisville Ridse 23'_25'8 1982 7.7 (2024-2130) (S 1%) (:51%) (S 1%) H TK-6 C. Tonga-Kermadec 25'-27'S No historie record of great earthquakes TK-7 N. Kermadec I •. 27'_28.5'8 1917(?) (8.0) Ineomplete historie record TK-8 S. Kermadec I •. 28.5'-30.2'8 1917(?) (8.0) Ineomplete historie record TK-9 Southern Kermadec 30.2'_38'8 No historie reeord of great earthquakes

Notes:

Masnitudes are M s unless otherwise noted. Forecut window represents the 90% eonfidenee interval about the expected re(:urrenee time, and is eonditional upon the event .

not havinl occurred by 1989. All values in parentheses refted less reliable estimates. Method deseribes the principal data sets used ror the recurrenee time estimate: H, historie; G, geologie; D, direct.


Acknowledgements

A review and synthesis of this scope could not have been completed without the advice and support of numerous experts and friends in the circum-Pacific community. Thanks go to Brian Atwater, Sue Beck, Tom Boyd, Ray Buland, Doug Christensen, Jim Dewey, Bob Engdahl, Al Espinosa, Klaus Jacob, Paul Krumpe, Bob Masse, Bill McCann, Karen McNally, Carlos Mendoza, Dave Perkins, Enrique Silgado, Krishna Singh, Lynn Sykes, Rick Terman, Randy White, Mary Ellen Williams, and last, but not least, my fellow members of the Working Group on California Earthquake Probabilities. The additional task of reviewing this manuscript fell on Ray Buland, Jim Dewey, Al Espinosa, and Bill Spence. Their comments were invaluable in improving the overall presentation. This review was supported by the United States Geological Survey and the Agency for International Development, Office of U.S. Foreign Disaster Assistance under PASA BOF-OOOO-P-IC-4051-00.

REFERENCES

ADE, K. (1977), Tectonic Implications of the Large Shioya-Oki Earthquakes of /938, Tectonophysics 4/, 269-289.

ADE, K. (1979), Size of Great Earthquakes of /837- /974 lriferred from Tsunami Data, Geophys. Res. 84, 1561-1568.

ADE, K., and NOGUCHI, S. (1983), Revision of Magnitudes of Large Shallow Earthquakes, /897-1912, Phys. Earth Planet. Inter. 33, 1-11.

ADAMEK, S., TAJIMA F., and WEINS, D. A. (1987), Seismic Rupture Associated with Subduction of the Cocos Plate, Tectonics 6, 757-774.

ADAMEK, S., FROHLICH, c., and PENNINGTON, W. D. (1988), Seismicity of the Caribbean-Nazca Boundary: Constraints on Microp/ate Tectonics of the Panama Region, J. Geophys. Res. 93, 2053-2075.

ANDO, M. (1975), Source Mechanisms and Tectonic Significance of Historica/ Earthquakes a/ong the Nankai Trough, Japan, Tectonophysics 27, 119-140.

ASKEW, B. L., and ALGERMISSEN, S. T. Cata/og of Earthquakes for South America, Peru, Vol. 7 (CERESIS, Lima, Peru 1985), 1108 pp.

ASTIz, L., and KANAMORI, H. (1984), An Earthquake Doublet in Ometepec, Mexico, Phys. Earth Planet. Inter. 34, 24-45.

ATWATER, B. F. (1987), Evidence for Great Holocene Earthquakes Along the Outer Coast of Washington State, Science 236, 942-944.

ATWATER, B. F., HULL, A. G., and BEVIS, K. A. (1987), Aperiodic Holocene Recurrence of Widespread, Probably Coseismic Subsidence in Southwestern Washington (Abs), EOS, Trans. Am. Geophys. Union 68, 1468.

BARAZANGI, M., and ISAcKs, B. L. (1976), Spatial Distribution of Earthquakes and Subduction of the Nazca Plate Beneath South America, Geology 4, 686-692.

BAKUN, W. H., and LINDH, A. G. (1985), The Parkfield, California Earthquake Prediction Experiment, Science 229, 619-624.

BEAVAN, J. (1988), Crustal Deformation in the Shumagin Seismic Gap, Alaska, U.S. Geol. Sur. Open File Rep. 88-673, 183-188.

BECK, S. L., and RUFF, L. J. (1984), The Rupture Process of the Great /979 C%mbia Earthquake: Evidence for the Asperity Model, J. Geophys. Res. 89,9281-9291.

BECK, S. L., and RUFF, L. J. (1987), Rupture Process of the Great /963 Kurile Islands Earthquake Sequence: Asperity Interaction and Multiple Event Rupture, J. Geophys. Res. 92, 14123-14138.


BECK, S. L., and NISHENKO, S. P. (1990), Variations in the Mode o/Great Earthquake Rupture along the Central Peru Subduelion Zone, Geophys. Res. Leu. 17, 1969~1972.

BECK, S. L., and CHRISTENSEN, D. H. (1991), Rupture Proeess 0/ the Feb. 4, 1965 Rat Islands Earthquake, J. Geophys. Res., in press.

BECK, S. L., and RUFF, L. J. (1989), Great Earthquakes and Subduetion along the Peru Treneh, Phys. Earth Planet Inter. 57, 199~224.

BEN-MENAHEM, A. (1978), Souree Meehanism 0/ the 1906 San Franciseo Earthquake, Phys. Earth Planet. Int. 17, 163~181.

BOSTWICK, T. K. (1984), A Re-examination 0/ the August 22, 1949 Queen Charlotte Earthquake, M. S. Thesis, 112 pp., Univ. British Columbia, Vancouver.

BOWIN, C. L. (1976), Caribbean Gravity Field and Plate Teetonies, Geol. Soc. Am. Spec. Paper 169, 79 pp.

BOYD, T. M. (1986), The Andreano/Islands Earthquake: Previous Rupture History (abs.), EOS Trans. Am. Geophys. Union 67, 1082.

BOYD, T. M., and NABELEK, J. L. (1988), Rupture Proeess 0/ the Andreano/Islands Earthquake 0/ May 7, 1986, Bull. Seismol. Soc. Am. 78, 1653~ 1673.

BOYD, T. M., and LERNER-LAM, A. L. (1988), Spatial Distribution 0/ Turn-o/-the-eentury Seismicity along the Alaska~Aleutian Are, Bull. Seismol. Soc. Am. 78, 636~650.

BOYD, T. M., TABER, J. J., LERNER-LAM, A. L., and BEAVAN, J. (1988), Seismie Rupture and Are Segmentation within the Shumagin Islands Seismie Gap, Geophys. Res. Leu. 15, 201~204.

BRUNE, J. N., and ENGEN, G. R. (1969), Excitation 0/ Mantle Love Waves and Definition 0/ Mantle Wave Magnitude, Bull. Seismol. Soc. Am. 59, 923~933.

BUCKNAM, R. c., and BARNHARD, T. P. (1989), Evidenee 0/ Sudden Late Holocene Uplift in Central Puget Lowland (abs.), EOS, Trans. Am. Geophys. Union 70, 1332.

BULAND, R., and NISHENKO, S. P. (1988), Best Earthquake Forecast Probabilities and Conditional Earthquake Prediction (abs.), EOS, Trans. Am. Geophys. Union 69, 1299.

CHAEL, E. P., and STEWART, G. S. (1982), Recent Large Earthquakes along the Middle American Trench and their Implications /or the Subduction Process, J. Geophys. Res. 87, 329~338.

CHASE, C. G. (1978), Plate Kinematics: The Amerieas, East A/rica, and the Rest 0/ the World, Earth Planet. Sci. LeU. 37, 355~368.

CHASE, c., MENARD, H., and MAMMERICKX, J. (1970), Bathymetry 0/ the North Pacific, Charts 1 ~6, Scripps Inst. Oceanogr., Inst. Marine Res., La Jolla, CA.

CHOY, G. L., and DEWEY, J. W. (1988), Rupture Process 0/ an Extended Earthquake Sequence: Teleseismic Analysis 0/ the Chilean Earthquake 0/ March 3, 1985, J. Geophys. Res. 93, 1103~ 1118.

CHRISTENSEN, D. H., and LAY, T. (1988), Large Earthquakes in the Tonga Region Associated with the Subduction 0/ the Louisville Ridge, J. Geophys. Res. 93, 13367~13389.

CHRISTENSEN, D. H., and RUFF, L. J. (1983), Outer-rise Earthquakes and Seismic Coupling, Geophys. Res. Let!. 10, 697 ~ 700.

CHRISTENSEN, D. H., and RUFF, L. J. (1989), Seismic Coupling and Outer Rise Earthquakes, J. Geophys. Res. 93, 13421 ~ 13444.

CIFUENTES, I. L. (1989), The 1960 Chilean Earthquakes, J. Geophys. Res. 94, 665~680. COHEN, A. C., Three-parameter estimation, in Lognormal Distributions (eds. Crow, E. L., and Shimizu,

K.) (Marcel Dekker, Inc, New York 1988) pp. 113~138. COMPTE, D., EISENBERG, A., LORCA, E., PARDO, M., PONCE, L., SARAGONI, R., SINGH, S. K., and

SUAREZ, G. (1986), The 1985 Central Chile Earthquake: A Repeat 0/ Previous Great Earthquakes in the Region?, Science 233, 449~453.

CORMIER, V. F. (1975), Tectonics near the Junction 0/ the Aleutian and Kuril ~ Kamchatka Ares and a Mechanism /or Middle Tertiary Magmatism in the Kamchatka Basin, Bull. Geol. Soc. Am. 86, 443~453.

DAVIES, J., SYKES, L., HOUSE, L., and JACOB, K. (1981), Shumagin Seismic Gap, Alaskan Peninsula: History 0/ Great Earthquakes, Tectonie Setting, and Evidence /or High Seismic Potential, J. Geophys. Res. 86, 3821 ~ 3855.

DAVIS, P. M., JACKSON, D. D., and KAGAN, Y. Y. (1989), The Longer It Has Been Since the Last Earthquake, the Longer the Expeeted Time till the Next?, Bull. Seismol. Soc. Am. 79, 1439~ 1456.


DEWEY, J. W., and SPENCE, W. (1979), Seismic Gaps and Source Zones 0/ Recent Large Earthquakes in Coastal Peru, Pure Appl. Geophys. 117,1148-1171.

EDWARDS, R. L., TAYLOR, F. W., CHEN, J. H., and WASSERBURG, G. J. (1987), High-precision thorium-230 dating 0/ corals using thermal ionization mass spectrometry: Applications to paleoseismology, In Proceedings of Conference XXXIX, Directions in Paleoseismology (eds. Crone, A. J., and Omdahl, E. M.) (U.S. Geol. Survey Open File Rpt. 87-673), pp. 30-38.

EISSLER, H. K., and McNALLY, K. C. (1984), Seismicity and Tectonics 0/ the Rivera Plate and Implications /or the 1932 Jalisco, Mexico Earthquake, J. Geophys. Res. 89, 4520-4530.

EISSLER, H. K., ASTIz, L., and KANAMORI, H. (1986), Tectonic Selling and Source Parameters 0/ the September 19, 1985 Michoacan, Mexico Earthquake, Geophys. Res. LeU. 13, 569-572.

EKSTRÖM, G., and DZIEWONSKI, A. M. (1986), CMT Mechanisms/or the Andreano/ !slands Earthquake Sequence (abs.), EOS, Trans. Am. Geophys. Union. 67, 1082.

EKSTRÖM, G., and ENGDAHL, E. R. (1989), Earthquake Source Parameters and Stress Distribution in the Adak !sland Region 0/ the Central Aleutian Islands, Alaska, J. Geophys. Res. 94, 15499-15519.

ENGDAHL, E. R., BILLINGTON, S., and KISSLINGER, C. (1989), Teleseismically Recorded Seismicity be/ore and after [he May 7, 1986, Andreano/Islands, Alaska Earthquake, J. Geophys. Res. 94, 15481-15498.

ESPINOSA, A. F., HUSID, R., and ALGERMISSEN, S. T. (1975), Seismological and Engineering Features 0/ the October 3, 1974 Lima Earthquake, Proc. of the 5th European Conf. on Earthquake Eng., Chap. 7, Paper 141, 1-7, Istanbul, Turkey.

ESPINOSA, A. F., HUSID, R., ALGERMISSEN, S. T., and DE LAS CASAS, J. (1977), The Lima Earthquake o/October 3, 1974; Intensity Distribution, BuH. Seismol. Soc. Am. 67, 1429-1440.

ESTABROOK, C. H., and BOYD, T. M. (1989), Magnitude Estimate 0/ the Shumagin !slands Earthquake 0/ May 31, 1917 (abs.), Seismol. Res. LeU. 60, 33.

FEDOTOV, S. A. (1965), Regularities 0/ the Distribution 0/ Strong Earthquakes in Kamchatka, the Kuril Islands, and Northeastern Japan, Tr. Inst. Fiz. Zemli Akad. Nauk SSSR 36, 66-93.

FEDOTOV, S. A., CHERNYSHEV, S. D. and CHERNYSHEV, G. V. (1982), The Improved Determination 0/ the Source Boundaries /or Earthquakes 0/ M ~ 73/4, 0/ the Properties 0/ the Seismic Cyc/e, and 0/ Long-term Seismic Prediction/or the Kurile-Kamchatkan Arc, Earthq. Pred. Res. 1, 153-171.

FUKAO, Y., and FURUMOTO, M. (1979), Stress Drops, Wave Spectra, and Recurrence Intervals 0/ Great Earthquakes-Implications 0/ the Etoro/u Earthquake 0/ 1958 November 6, Geophys. J. R. Astron. Soc. 57, 23-40.

GATES, 0., and GIBSON, W. (1956), Interpretation o/the Configuration o/the Aleutian Ridge, BuH. Geol. Soc. Am. 67, 127-146.

GEIST, E. L., CHILDS, J. R~ and SCHOLL, D. W. (1988), The Origin 0/ Summit Basins on the Aleutian Ridge: Implications /or Block Rotation 0/ an Arc Massif, Tectonics 7, 327-341.

GONZALEZ-RuIZ, J. R. G., and McNALLY, K. C. (1986), Seismic Slip Gap and Potential/or a Large Multiple-shock Earthquake Sequence in Guerrero, Mexico (abs.), EOS, Trans. Am. Geophys. Union 67, 1115.

GONZALEZ-RuIZ, J. R. G., and McNALLY, K. C. (1988), Stress Accumulation and Release since 1882 in Ometepec, Guerrero, Mexico: Implications /or Failure Mechanisms and Risk Assessments 0/ a Seismic Gap, J. Geophys. Res. 93, 6297-6317.

GRANT, W. c., ATWATER, F., CARVER, G. A., DARIENZO, M. E., NELSON, A. R., PETERSON, C. D. and VICK, G. S. (1989), Radiocarbon Dating 0/ Late Holocene Coastal Subsidence above the Cascadia Subduction Zone-Compilation /or Washington, Oregon, and Northern California (abs.), EOS, Trans. Am. Geophys. Union 70, 1331.

GUENDEL, F., and McNALLY, K. C. (1986), High Resolution Evidence 0/ Smooth BeniojJ Zone Gradations Approaching the Southern Terminus 0/ the Middle America Trench (abs), EOS, Trans. Am. Geophys. Union 67, 1114.

GUTEN BERG, B. (1939), Tsunamis and Earthquakes, BuH. Seismol. Soc. Am. 29, 517-526. GUTENBERG, B., and RICHTER, C. F., Seismicity 0/ the Earth, 2nd Ed. (Princeton Univ. Press, New

Jersey 1954), pp. 161-234. HABERMANN, R. E. (1981), The Quantitative Recognition and Evaluation 0/ Seismic Quiescence:

Applications to Earthquake Prediction and Subduction Zone Tectonics. Ph.D. thesis, 253 pp., Univ. Colorado, Boulder, CO.


HAGIWARA, Y. (1974), Probability of Earthquake Occurrence as Obtained from a Weibull Distribution Analysis of Crustal Strain, Tectonophysics 23, 313-318.

HATORI, T. (1976), Propagation of Tsunamis from Source off the Pacijic Coast of Northeast Japan, BuH. Earthquake Res. Inst., Univ. Tokyo 5/, 197-207.

HATORI, T. (1979), Tsunami Sources in Hokkaido and Southern Kuril Regions, BuH. Earthq. Res. Inst., Tokyo Univ. 49, 63-75.

HATORI, T. (1981), Tsunami Magnitude and Source Area of Aieutian-Alsaka Tsunamis, BuH. Earthquake Res. Inst. 56,97-110.

HEATON, T. H., and KANAMORI, H. (1984), Seismie Potential Associated with Subduetion in the Northwestern United States, BuH. Seismol. Soc. Am. 74, 933-941.

HEATON, T. H., and HARTZELL, S. H. (1986), Source Charaeteristics of Hypothetieal Subduction Earthquakes in the Northwestern United States, BuH. Seismol. Soc. Am. 76, 675-708.

HERRON, E. M. (1981), Chile margin near 38°S: Evidenee for a genetie relationship between eontinental and marine geologieal features or a ease of eurious eoincidenee?, In Nazca Plate: Crustal Deformation and Andean Convergenee (eds. Kulm, L. D., Dymond, J., Dasch, E. J., and HUSSONG, D. M.) Mem. Geol. Soc. Am. /54, 683 - 702.

HOUSE, L. S., SYKES, L. R., DAVIES, J. N., and JACOB, K. H., Identijication of a possible seismie gap near Unalaska Island, eastern Aleutians, In Earthquake Prediction, An International Review, Maurice Ewing Ser., Vol. 4 (eds. Simpson, D. W., and Richards, P. G.) (Am. Geophys. Union, Washington, D.C. 1981), pp. 81-92.

HUDNUT, K. W., and TABER, J. J. (1987), Transition from Double to Single Wadati-Benioff Seismie Zone in the Shumagin Islands, Alaska, Geophys. Res. Lett. /4, 143-146.

HWANG, L. J., and KANAMORI, H. (1986), Souree Parameters of the May 7, 1986 Andreanof Islands Earthquake, Geophys. Res. Lett. /3, 1426-1429.

IIDA, K., Cox, D., and PARARAS-CARAYANNIS, G. (1967), Preliminary Catalog of Tsunamis Oceurring in the Paeijie Oeean, Hawaii Inst. Geophys. 67-10, Univ. Hawaii, 131 pp.

ISHIBASHI, K., Specijieation of a soon-to-occur seismic faulting in the Tokai district, eentral Japan, based on seismotectonies, In Ear/hquake Predietion, An In/erna/ional Review, Maurice Ewing Ser., Vol. 4 (eds. Simpson, D. W. and Richards, P. G.) (Am. Geophys. Union, Washington, D.C. 1981), pp. 297-332.

JACOBY, G. c., and ULAN, L. D. (1983), Tree Ring Indications of Uplijt at /cy Cape, Alaska Related to the /899 Earthquakes, J. Geophys. Res. 88,9305-9313.

JACOBY, G. c., SHEPPARD. P. R., and SIEH, K. E. (1988), Irregular Recurrence of Large Earthquakes along the San Andreas Fault: Evidenee from Trees, Science 24/, 196-199.

JORDAN, T. H. (1975), The Present Day Motions ofthe Caribbean Plate, J. Geophys. Res. 80,4433-4439. KANAMORI, H. (1972), Meehanism of Tsunami Earthquakes, Phys. Earth Planet. Int. 6, 246-259. KANAMORI, H. (1977a), The Energy Release in Great Earthquakes, J. Geophys. Res. 82,2981-2987. KANAMORI, H., Seismie and aseismie slip along subduetion zones and their tectonic implications, In Island

Ares, Deep Sea Trenehes, and Back-Are Basins, M. Ewing Sero I (eds. Talwani, M., and Pitman III, W. C.) (Am. Geophys. Union, Washington, D.C. 1977b), pp. 273-284.

KANAMORI, H., and McNALLY, K. C. (1982), Variable Rupture Mode of the Subduelion Zone along the Eeuador-Colombia Coast, BuH. Seismol. Soc. Am. 72, 1241-1253.

KATSUI, Y. ed., List of World Ac/ive Voleanoes (Volcanological Society of Japan, Sapporo, Japan 1971). KEILIS-BoROK, V. 1., KNOPOFF, L., ROTVAIN, I. M., and ALLEN, C. R. (1988), Intermediate-term

Predietion of Oeeurrence Times of Strong Earthquakes, Nature 335, 690-694. KELLEHER, J. A. (1972), Rupture Zones of Large South American Earthquakes and Some Predictions, J.

Geophys. Res. 77, 2087-2103. KELLEHER, J. A., SYKES, L. R., and OLIVER, J. (1973), Possible Criteria for Predicting Earthquake

Loeations and their Applieations to Major Plate Boundaries of the Pacijie and Caribbean Region, J. Geophys. Res. 78, 2547-2585.

KELLEHER, J. A., and MCCANN, W. R. (1976), Buoyant Zones, Great Earthquakes, and Unstable Boundaries of Subduetion, J. Geophys. Res. 8/, 4885-4908.

KONDORSKAYA, N. V., and SHEBALIN, N. V., New Catalog of Strong Earthquakes in the U.S.S.R.jrom Ancient Times through 1977 (U.S. Dept. of Commerce, Boulder, CO 1982), 608 pp.

LAY, T., and KANAMORI, H. (1980), Earthquake Doublets in the Solomon Islands, Phys. Earth Planet. Int. 21, 283-304.


LEFEVRE, L. V., and MCNALLY, K. C. (1985), Stress Distribution and Subduction of Aseismic Ridges in the Middle America Subduction Zone, J. Geophys. Res. 90, 4495-4510.

LiSOWSKI, M., SAVAGE, J. C., PRESCOTT, W. H., and GROSS, W. K. (1988), Absence of Strain Accumulation in the Shumagin Seismic Gap, Alaska, 1980-1987, J. Geophys. Res. 93, 7909-7922.

LOMNITZ, C. (1970), Major Earthquakes and Tsunamis in Chile during the Period 1535 to 1953, Geol. Rundseh. 59, 938-960.

LUHR, J. F., NELSON, S. A., ALLAN, J. F., and CARMICHAEL, I. S. E. (1985), Active Rifting in South-western Mexico: Manifestations of an Incipient Eastward Spreading Ridge Jump, Geology 13, 54-57.

LUNDGREN, P. R., OKAL, E. A., and WEINS, D. A. (1989), Rupture Characteristics of the 1982 and 1986 Kermadec Earthquakes, J. Geophys. Res. 94, 15521 -15539.

MANN, P., and BURKE, K. (1984), Neotectonics of the Caribbean, Rev. Geophys. 22, 309-362. MATSUDA, T., OTA, Y., ANDO, M., and YONEKURA, N. (1978), Fault Mechanism and Recurrence Time

of Major Earthquakes in the Southern Kanto District, Japan as Deducedfrom Coastal Terrace Data, Bull. Geol. Soc. Am. 89, 1610-1618.

MCCANN, W. R., Large- and Moderate-sized Earthquakes: Their Relationship to the Tectonics of Subduction, Ph.D. Thesis (Columbia Univ., New York 1980), 194 pp.

MCCANN, W. R., NISHENKO, S. P., SYKES, L. R., and KRAUSE, J. (1979), Seismic Gaps and Plate Tectonics: Seismic Potential for Major Boundaries, Pure Appl. Geophys. /17, 1082-1147.

MCCANN, W. R., PEREZ, O. J., and SYKES, L. R. (1980), Yakataga Gap, Alaska: Seismic History and Earthquake Potential, Science 207, 1309-1314.

McNALLY, K. c., and MINSTER, J. B. (1981), Nonuniform Seismic Slip Rates along the Middle America Trench, J. Geophys. Res. 86, 4949-4959.

MENDOZA, c., and DEWEY, J. W. (1984), Seismicity Associated with the Great Colombia-Ecuador Earthquakes of 1942, 1958, and 1979: Implications for Barrier Models of Earthquake Rupture, Bull. Seismol. Soc. Am. 74, 577-593.

MENDOZA, c., and NISHENKO, S. P. (1989), The North Panama Earthquake of 7 September 1882: Evidence for Active Underthrusting, Bull. Seismol. Soc. Am. 79, 1264- 1269.

MILNE, J. (1880), The Peruvian Earthquake of May 9th, 1877, Trans. Seismol. Soc. Japan 2, 50-96. MINSTER, J. B., and JORDAN, T. H. (1978), Present Day Plate Motions, J. Geophys. Res. 83,

5331 -5354. MIYASHITA, K., and MATSUURA, M. (1978), Inversion Analysis of Static Displacement Data Associated

with the Alaska Earthquake of 1964, J. Phys. Earth 26, 333-349. MOGI, K. Earthquake prediction program in Japan, In Earthquake Prediction, An International Review,

Maurice Ewing Ser., Vol. 4 (eds. Simpson, D. W., and Richards, P. G.) (Am. Geophys. Union, Washington, D.C. 1981), pp. 635-666.

MOGI, K. (1968), Some Features of Recent Seismic Activity in and near Japan, Bull. Earth. Res. Inst., Tokyo Univ. 46, 1225-1236.

MONTERO, W. (1986), Periodos de recurrencia y ti pos de sequencias sismicas de los temblores interplaca e intraplaca en la region de Costa Rica, Rev. Geol. Am. Central 5, 35-72.

MONTERO, W. (1989), Seismicidad historica de Costa Rica, 1638-1910, Geof. Int. 28, 531-559. MORALES, L. D. (1985), Las zonas sismicas de Costa Rica y alrededores, Rev. Geol. Am. Centr. 3,

69-101. MUNOZ, A. V. (1988a), Seismicity of the Panama Block, I. Magnitudes and Spatial Distribution of

Epicenters, Tectonophysics 145, 213-224. MUNOZ, A. V. (1988b), Tectonic Patterns of the Panama Block Deduced from Seismicity, Gravitational

Data, and Earthquake Mechanisms: Implications to the Seismic Hazard, Tectonophysics 154, 253-267. NAVAL OCEANOGRAPHIC OFFICE (1978), Bathymetric Atlas of the North Pacific Ocean, N.O. Pub.

1301-2-3 (S), 176 pp. NISHENKO, S. P. (1985), Seismic Potential for Large and Great Interplate Earthquakes along the Chilean

and Southern Peruvian Margins of South America: A Quantitative Reappraisal, J. Geophys. Res. 90, 3589-3615.

NISHENKO, S. P., Earthquakes: hazards and prediction, In Encyclopedia of Solid Earth Geophysics (ed. James, D. E.) (Van Nostrand Reinhold 1989), pp. 260-268.


NISHENKO, S. P., and MCCANN, W. R. (1979), Large Thrust Earthquakes and Tsunamis: Implications/or the Development 0/ Fore Arc Basins, J. Geophys. Res. 84, 573-584.

NISHENKO, S. P., and MCCANN, W. R., Seismic potential 0/ the world's major plate boundaries, In Earthquake Prediction, An International Review, Maurice Ewing Ser., Vol. 4 (eds. Simpson, D. W., and Richards, P. G.) (Am. Geophys. Union, Washington, D.C. 1981), pp. 20-28.

NISHENKO, S. P., and BULAND, R. (1987), A Generic Recurrence Interval Distribution /or Earthquake Forecasting, Bull. Seismol. Soc. Am. 77, 1382-1399.

NISHENKO, S. P., and SINGH, S. K. (1987a), Conditional Probabilities /or the Recurrence 0/ Large and Great Interplate Earthquakes along the Mexican Subduction Zone, Bull. Seismol. Soc. Am. 77, 2095-2114.

NISHENKO, S. P., and SINGH, S. K. (l987b), The Acapulco-Ometepec Earthquakes 0/ 1907-1982: Evidence /or a Variable Recurrence History, Bull. Seismol. Soc. Am. 77,1359-1367.

NISHENKO, S. P., and SINGH, S. K. (l987c), Relocation 0/ the Great Mexican Earthquake 0/14 January 1903, Bull. Seismol. Soc. Am. 77, 256-259.

NISHENKO, S. P., and JACOB, K. (1990), Seismic Potential 0/ the Queen Charlotte-Alaska-Aleutian Seismic Zone, J. Geophys. Res. 95, 2511-2532.

PASCAL, G., ISAcKS, 8., BARAZANGI, M., and DUBOlS, J. (1978), Precise Locations 0/ Earthquakes and Seismotectonics 0/ the New Hebrides Island Arc, J. Geophys. Res. 83, 4957-4974.

PENNINGTON, W. D. (1981), Subduction 0/ the Eastern Panama Basin and the Seismotectonics 0/ Northwestern South America, J. Geophys. Res. 86, 792-802.

PLAFKER, G. (1969), Tectonics 0/ the March 27, 1964 Alaska Earthquake, U.S. Geol. Survey Prof. Paper 543-1, 74 pp.

PLAFKER, G., (1986), Geologic studies related to earthquake potential and recurrence in the .. Yakataga seismic gap ", In Minutes 0/ the National Earthquake Prediction Council (ed. Shearer, C. F.), U.S. Geol. Survey Open File Rpl. 86-92, 135-143.

PLAFKER, G., and SAVAGE, J. C. (1970), Mechanism o/the Chilean Earthquakes 0/ May 21 and 22, 1960, Bull. Geol. Soc. Am. 81, 1001-1030.

PLAFKER, G., HUDSON, T., BRuNs, T., and RUBIN, M. (1978), Late Quaternary Offsets along the Fairweather Fault and Crustal Plate Interactions in Southern Alaska, Can. J. Earth Sei. 15, 805-816.

PLAFKER, G., and THATCHER, W. (1982), Geological and Geophysical Evaluation 0/ the Great 1899-1900 Yakutat Bay, Alaska Earthquakes (abs), Paper presented at AGU Chapman Conference on Fault Behavior and the Earthquake Generation Process, AGU, Snowbird, Utah, Ocl. 11-15.

PDE (1988), Preliminary Determination 0/ Epicenters, U.S. Geol. Survey. PRINCE, R. A., SCHWELLER, W. J., COULBOURN, W. T., SHEPARD, G. L., NESS, G. E., and MASIAS,

A. (1980), Bathymetry 0/ the Peru-Chile Continental Margin and Trench, 9 Maps, Map Chart Series MC-34, Geol. Soc. America, Boulder, CO.

RElD, H. F., 1910, The mechanics 0/ the earthquake, In The California Earthquake 0/ April 18, 1906 (Lawson, A. c., Chairman). Report of the State Earthquake Investigation Commission [reprinted 1969) (Carnegie Institution of Washington Publication 87, v. 2, 1910), 192 pp.

RElD, H. F. (1917), Note on the Earthquakes at Almirante, Republic 0/ Panama, in April 1916, Bull. Seismol. Soc. Am. 7, 27-30.

RICHTER, C. F., Elementary Seismology (W. H. Freeman and Co., San Francisco, CA 1958), 768 pp. RIKITAKE, T. (1976), Recurrence o/Great Earthquakes at Subduction Zones, Tectonophysics 35, 335-362. RINEHART, W., GANSE, R., TEIK, P., ARNoLD, E., STOVER, c., and SMITH, R. H. (1982), Seismicity

0/ Middle America, National Geophysical Data Center and National Earthquake Information Service, Boulder, CO.

ROGERS, G. C. (1986), Seismic Gaps along the Queen Charlotte Fault, Earthq. Pred. Res. 4, 1-11. ROTHE, J. (1969), The Seismicity 0/ the Earth, 1953-1965, (UNESCO, Paris), 336 pp. RUFF, L., and KANAMORI, H. (1983), The Rupture Process and Asperity Distribution 0/ Three Great

Earthquakes /rom Long-period Diffracled P-waves, Phys. Earth Planet Inter. 31, 202-230. SAVAGE, J. c., LISOWSKI, M., and PRESCOTT, W. H. (1986), Strain Accumulation in Ihe Shumagin and

Yakataga Seismic Gaps, Science 231, 585-587. SCHOLZ, C. H. (1985), The Black Mountain Asperily: Seismic Hazard 0/ the Southern San Francisco

Peninsula, California, Geophys. Res. LeU. 12,717-719.


SCHOLZ, C. H., and KATO, T. (1978), The Behavior of a Convergent Plate Boundary: Crustal Deformation in the South Kanto Distriet, Japan, J. Geophys. Res. 83, 783-797.

SCHWARTZ, D. P. (1985), The Caribbean-North Ameriean Plate Boundary in Central Ameriea: New Data on Quarternary Tectonics (abs.), Earthquake Notes 55, p. 28.

SCHWARTZ, D. P., and COPPERSMITH, K. J. (1984), Fault Behavior and Charaeteristie Earthquakes, Examplesfrom the Wasateh and San Andreas Fault Zones, J. Geophys. Res. 89,5681-5698.

SCHWARTZ, D. P., CLUFF, L. S., and DONNELLY, T. W. (1979), Quarternary Faulting along the Caribbean - North Ameriea Plate Boundary in Central Ameriea, Tectonophysics 52, 431-445.

SCHWARTZ, S. Y., and RUFF, L. J. (1985), The 1968 Tokachi-Oki and the 1969 Kurile Islands Earthquakes: Variability in the Rupture Process, J. Geophys. Res. 90, 8613-8626.

SCHWARTZ, S. Y., and RUFF, L. J. (1987), Asperity Distribution and Earthquake Occurrence in the Southern Kurile Islands Are, Phys. Earth Planet. Inter. 49, 54-77.

SCHWARTZ, S. Y., DEWEY, J. W., and LAY, T. (1989), Influenee of Fault Plane Heterogeneity on the Seismic Behavior in the Southern Kurile Islands Are, J. Geophys. Res. 94, 5637-5649.

SCHWELLER, W. J., KULM, L. D., and PRINCE R. A. (1981), Tectonics, strueture, and sedimentary framework ofthe Peru-Chile trench, In Nazca Plate: Crustal Deformation and Andean Convergence (eds. Kulm, L. D., Dymond, J., Dasch, E. J., and Hussong, D. M.) Mem. Geol. Soc. Am. 154, 323-350.

SHIMAZAKI, K. (1974), Nemuro-Oki Earthquake of June 171973: A Lithospheric Rebound atthe Upper Half of the Interface, Phys. Earth Planet. Inter. 9, 314-327.

SHIMAZAKI, K., and NAKATA, T. (1980), Time-predictable Recurrence Model for Large Earthquakes, Geophys. Res. Let!. 7, 279-282.

SIEH, K. (1986), Slip Rates across the San Andreas Fault and Prehistoric Earthquakes at Indio, California (abs.), EOS, Trans. Am. Geophys. Union 67, 1200.

SIEH, K., STUIVER, M., and BRILLINGER, D. (1989), A Very Precise Chronology of Earthquakes Produced by the San Andreas Fault in Southern California, J. Geophys. Res. 94, 603-623.

SILGADO, E. (1985), Destructive Earthquakes of South American 1530-1894, Earthquake Mitigation Program in the Andean Region, Project SISRA, vol. 10, Lima, Peru, 315 pp.

SILVER, E. A., REED, D. L., T AGUDlN, J. E., and HEIL, D. J. (1990), Implications of the North and South Panama Thrust Belts for the Origin of the Panama Orocline, Tectonics 9, 261-281.

SINGH, S. K., ASTIz, L., and HAvsKov, J. (1981), Seismic Gaps and Reeurrence Periods of Large Earthquakes along the Mexiean Subduetion Zone: A Reeexamination, Bull. Seismol. Soc. Am. 71, 827-843.

SINGH, S. K., ESPINDOLA, J. M., YAMAMOTO, J., and HAvsKov, J. (1982), Seismic Potential of the Acapulco-San Mareos Region along the Mexican Subduetion Zone, Geophys. Res. Let!. 9, 633-636.

SINGH, S. K., DOMINGUEZ, T., CASTRO, R., and RODRIGUEZ, M. (1984), P Waveforms of Large Shallow Earthquakes along the Mexiean Subduetion Zone, Bull. Seismol. Soc. Am. 74, 2135-2156.

SINGH, S. K., PONCE, L., and NISHENKO, S. P. (1985), The Great Jalisco, Mexieo Earthquakes of 1·932: Subduetion of the Rivera Plate, Bull. Seismol. Soc. Am. 75, 1301-1313.

SPENCE, W. (1977), The Aleutian Are: Tectonic Blocks, Episodie Subduction, Strain Diffusion, and Magma Generation, J. Geophys. Res. 82, 213-230.

SPENCE, W. (1989), Stress Origins and Earthquake Potentials in Caseadia, J. Geophys. Res. 94,3076-3088. SYKES, L. R. ( 1971), Aftershock Zones of Great Earthquakes, Seismicity Gaps, and Earthquake Predietion,

J. Geophys. Res. 76, 8021-8041. SYKES, L. R., and QUITTMEYER, R. c., Repeat times of great earthquakes along simple plate boundaries,

In Earthquake Predietion, An International Review, Maurice Ewing Sero 4 (eds. Simpson, D. W., and Richards, P. G.) (Am. Geophys. Union, Washington, D.C. 1981), pp. 217-247.

SYKES, L. R., and NISHENKO, S. P. (1984), Probabilities of Oeeurrenee of Large Plate Rupturing Earthquakes for the San Andreas, San Jacinto, and Imperial Faults, California: 1983-2003, J. Geophys. Res. 89, 5905-5928.

TAJIMA, F., RUFF, L. J., KANAMORI, H., ZHANG, J., and MOG!, K. (1990), Earthquake Source Processes and Subduction Regime in the Santa Cruz Islands Region, Physics Earth Planet. Int. 61, 269-290.

TAYLOR, F. W., ISAcKs, B. L., JOUANNIC, c., BLOOM, A. L., and DUBOIS, J. (1980), Coseismie and Quarternary Vertieal Tectonic Movements, Santo and Malekula Islands, New Hebrides Island Are, J. Geophys. Res. 85, 5367-5381.


TAYLOR, F. W., JOUANNIC, c., and BLOOM, A. L. (1985), Quarternary Uplift of the Torres Islands, Northern New Hebrides Frontal Are: Comparison with Santo and Malekula Islands, Central New Hebrides Frontal Are, J. Geo1ogy 93, 419-438.

TAYLOR, F. W., EDWARDS, R. L., WASSERBURG, G. J., and FROHLICH, C. (1990), Seismie Reeurrenee Intervals and Timing of Aseismie Subduetion Inferred from Emerged Corals and Reefs of the Central Vanuatu (New Hebrides) Frontal Are, J. Geophys. Res. 95, 393-408.

THATCHER, W. (1990), Order and Diversity in the Modes of Cireum-Paeifie Earthquake Reeurrenee, J. Geophys. Res. 95, 2609-2623.

THATCHER, W., and LlsowsKl, M. (1987), Long-term Seismie Potential of the San Andreas Fault Southeast of San Francisco, J. Geophys. Res. 92,4771-4784.

TOCHER, D. (1960), The Alaska Earthquake of July 10, 1958: Movement of the Fairweather Fault and Field Investigations of the Southern Epieentral Region, BuH. Seismol. Soc. Am. 50, 267-292.

UNAM SEISMOLOGY GROUP (1986), The September 1985 Miehoaean Earthquakes: Aftershock Distribution und History of Rupture, Geophys. Res. Lett. 13, 573-576.

U.S. GEOLOGICAL SURVEY STAFF (1990), The Loma Prieta, California Earthquake: An Anticipated Event, Science 247, 286-293.

UYEDA, S., and KANAMORI, H. (1979), Back-are Opening and the Mode of Subduetion, J. Geophys. Res. 84,1049-\061.

VARNES, D. J. (1989), Predieting Earthquakes by Analyzing Accelerating Preeursory Seismie Aetivity, Pure Appl. Geophys. 130, 661-686.

VIDALE, J., and KANAMORI, H. (1983), The Oetober 1980 Earthquake Sequenee near the New Hebrides, Geophys. Res. Lett. 10, 1137-1140.

WEAVER, C. S., and SMITH, S. W. (1983), Regional Tectonic and Earthquake Hazard Implieations of a Crustal Fault Zone in Southwestern Washington, J. Geophys. Res. 88, \0371-10383.

WESNOUSKY, S. G., SCHOLZ, C. H., SHIMAZAKI, K., and MATSUDA, T. (1984), Integration ofGeologie and Seismologie Data for the Analysis of Seismie Hazard: A Case Study of Japan, BuH. Seismol. Soc. Am. 74, 687 - 708.

WHITE, R. A. (1984), Catalog of Historie Seismicity in the Vieinity ofthe Chixoy-Poloehie and Motagua Faults, Guatemala, U.S. Geol. Survey Open File Rpt. 84-88, 40 pp.

WHITE, R. A. (1985), The Guatemala Earthquake of 1816 on the Chixoy-Poloehie Fault, BuH. Seismol. Soc. Am. 75, 455-473.

WHITE, R. A., and CIFUENTES, I. L. (1991), Seismie History ofthe Middle Ameriea Treneh Spanning the Guatemala Tripie Junetion and an Earthquake Forecast, BuH. Seismol. Soc. Am., in press.

WINSLOW, M. A. (I~82), The Struetural Evolution of the Magallanes Basin and the Neotectonics in the Southernmost Andes, Antarctic Geosci. 4, 143-154.

WORKING GROUP on CALIFORNIA EARTHQUAKE PROBABILlTlES (1990), Probabilities of Large Earthquakes in the San Franciseo Bay Region, California, U.S. Geol. Sur. Circu1ar 1053, 51 pp.

WORKING GROUP on CALIFORNIA EARTHQUAKE PROBABILITIES (1988), Probabilities of Large Earthquakes Oeeurring in California on the San Andreas Fault, U.S. Geol. Survey Open File Rpt. 88-398, 62 pp.

WYSS, M., HABERMANN, R. E., and GRIESSER, J. C. (1984), Seismie Quieseenee und Asperities in the Tonga-Kermadee Are, J. Geophys. Res. 89,9293-9304.

YAMAGUCHI, D. K., WOODHOUSE, C. A., and REID, M. S. (1989), Tree-ring Evideneefor Synehronous Rapid Submergenee of the Southwestern Washington Coast 300 Years B.P. (abs.), EOS, Trans. Amer. Geophys. Union 70, 1332.

(Received February 28, 1990, revised/accepted October 30, 1990)

PAGEOPH, Vol. 135, No. 2 (1991) 0033-4553/91/020261-99$1.50 + 0.20/0 © 1991 Birkhäuser Verlag, Basel

Intraplate Seismicity of the Pacific Basin, 1913 -1988

MICHAEL E. WYSESSION; EMILE A. OKAL1 and KRISTIN L. MILLER2

Abstract - We establish here a comprehensive database of intraplate seismicity in the Pacific Basin. Relocation and analysis of 894 earthquakes yield 403 reliable intraplate earthquakes during 1913-1988. These numbers do not inc1ude earthquake swarms, which account for another 838 events. Most of the remainder (304 events) are actually plate boundary earthquakes that have been erroneously located in intraplate regions. A significant number occur in recent years when location capabilities should have guarded against this situation. Relocations involve a careful linear inversion of P and S arrivals, accompanied by a Monte Carlo statistical analysis. We have also attentively removed the high number of c1erical errors and nuc1ear tests that exist in epicenter bulletins.

A geographical examina ti on of the relocated epicenters reveals several striking features. There are three NW-SE lineaments north of the Fiji Plateau and in Micronesia; diffuse seismicity and incompatible focal mechanisms argue against the southemmost, discussed by OKAL et al. (1986) and KROENKE and WALKER (1986), as the simple relocation of the Solomon trench to the North. Besides another striking lineament, along the \30oW meridian, there is also a strong correlation between seismicity and bathymetry in certain parts of the Basin. In the Eastcentral Pacific and Nazca plates there are many epicenters on fracture zones and fossil spreading ridges, and hot spot traces like the Louisville, Nazca and Cocos Ridges also display seismicity.

Key words: Intraplate seismicity, Pacific Basin, historical earthquakes.

1. Introduction

This paper presents an extensive evaluation of the intraplate seismicity of the

Pacific Ocean Basin over the 75 years since seismological data has been compiled

systematicaUy. We are motivated in this endeavor by continued in te rest in stresses

released during intraplate oceanic earthquakes. In particular, the nature, orienta

tion, and magnitude of these stresses provide important constraints on the deep

rheology of the plates, and on forces driving them (see STEIN and OKAL (1986) for

a review). Our investigation, involving the systematic relocation of more than 500 earthquakes, both historical (pre-1963) and recent, shows that less than one-half of

the catalog listings of intraplate events can be regarded as genuine (see Figure I). Most of the rest become plate boundary earthquakes mislocated to the interior of

I Department of Geological Sciences, Northwestem University, Evanston, Illinois 60208, U.S.A. 2 Department of Mathematics, Wheaton College, Wheaton, Illinois 60187, U.S.A.

262 Michael E. Wysession et al. PAGEOPH,

the plates for a variety of reasons. This result casts an ominous shadow on any attempt to reach quantitative conclusions regarding the magnitude of intraplate oceanic stress release based on existing, unrelocated catalogs. In addition, on a more local scale, the correlation of seismicity with existing bathymetric features, in the context of the possible existence of "weak" zones featuring preferential stress release, is even more dependent on an accurate knowledge of seismic epicenters. In both respects, then, the magnitude of the casualty rate among alleged intraplate events serves as an aposteriori justification of our present study.

Previous W ork

Interest in intraplate seismicity can be traced back to GUTENBERG and RICHTER'S (1941) monumental compilation of the earth's seismicity. Inside the Pacific Basin, which they recognized as one of their "stable masses", they identified

ORIGINAL BULLETIN 'INTRAPLATE' EPICENTERS

. . . . .' . . . . : . '. . ' ,

..... 20 . .

, ~ . . . .. '. , .... -. .' .. . .. . .•..

:.

-40

-60

120 140 160 180 2 40 260

Figure la Map of epicenters with Pacific Basin Intraplate locations, taken from original bulletins.

Vol. 135, 1991 Intrap1ate Seismicity 263

RELOCATED AND VERIFIED INTRAPLATE EPICENTERS

. . 40

, . ,

" '''b .. . , , 20 ,

" , "

" , , , ..

• • ~'\.It •• . :.," ..... : ' , , , , .

-60

120 140 160 180 200 220 240 260

Figure 1b Map of intraplate epicenters that we consider to be reliable, after either relocation or verification. Note that the large number of events near the Tonga, Japan, Aleutian and Central American trenches has

been grea tI y red uced.

22 earthquake listings by the ISS, and concluded that at most 2 could be correct. By the time of their last edition (GUTENBERG and RICHTER, 1954), they had come to the conclusion that none of the 22 could be trusted as intraplate. We will see that our relocations suggest that 5 out of 21 are indeed intraplate events, including the two these authors initially recognized (the 22nd event is listed in the Caroline plate, beyond our study area).

In more recent years, studies of intraplate seismicity can be classified in two categories. The first type, motivated by the investigation of the dynamic state of stress of the oceanic lithosphere, was concerned with retrieval of earthquake focal mechanisms and quantitative evaluation of the actual energy released. For these reasons, these studies were, in generallimited to the homogeneous dataset provided by the WWSSN, and did not involve historical events. Benchmark papers include


SYKES and SBAR (1974), RICHARDSON et al. (1979), BERGMAN and SOLOMON (1980, 1984), and WIENS and STEIN (1983, 1984).

The second type of study was generally of a more local nature, and was motivated either by the sudden occurrence of intense seismic swarms (e.g., FILSON et a/., 1973; LAY and OKAL, 1983; WIENS and OKAL, 1987), or even of a single event (STEIN, 1979; OKAL, 1980); or by the existence of a regional network providing improved detection capabilities over a limited spatial (and occasionally temporal) extent (TALANDIER and KUSTER, 1976; OKAL et a/., 1980; TALANDIER and OKAL, 1984a, 1987; WALKER and MCCREERY, 1985).

In addition, some investigations of a more global nature did address, at least partially, the question of intraplate Pacific seismicity, for instance in the framework of a general discussion of historical records (e.g., MIY AMURA, 1988).

At least three studies have attempted to compile the complete seismicity of all or part of the Pacific plate: OKAL (1981) listed 21 earthquakes belonging to the Antarctic plate, including 16 in the Pacific Basin; OKAL (1984) presented a compilation of the seismicity of the Pacific plate south of the Equator; more recently , WALKER (1989) has published a list of 406 reportedly intraplate earthquakes in the 4 main plates making up the Pacific Basin.

The present paper can be viewed as an extension of the work of OKAL (1981; 1984) in several directions: (i) full coverage of the entire Pacific Basin, including North of the Equator, and inside the Cocos and Nazca plates; (ii) more systematic relocation of all historical earthquakes, capitalizing on a more complete database; (iii) systematic examination of the statistical significance of the solution, inc/uding for recent earthquakes.

2. Data Base

As an initial working dataset we have tried to obtain for analysis all Pacific Basin intraplate earthquakes from 1913 through 1988. We have taken as intraplate events that are more than r from a plate boundary, and defined the Pacific Basin as being the full interiors of the Pacific, Nazca and Cocos plates, and the part of the Antarctic oceanic lithosphere that extends from 1600 E (the Pacific-AustralianAntarctic tripie junction) to 68°W (the southern tip of South America).

We have chosen a 2° buffer around plate boundaries for two reasons. There is a degree of uncertainty in the exact locations of earthquakes, both from errors and limitations in choosing travel times and from mantle heterogeneities between events and stations. A r buffer will certainly avoid the erroneous inclusion of most poorly located plate boundary earthquakes. Secondly, there are documented cases (e.g., CHEN and FORSYTH, 1978) where earthquakes, associated with plate bending, have occurred seaward of trenches, and while these are technically intraplate, they are associated with boundary tectonics and should not be part of this study.


It is certainly inevitable that in steadfastly choosing a 2° buffer around all plate boundaries we will omit certain verified intraplate earthquakes. Such an example is the January 21, 1970 event just south of the Siqueiros Fracture Zone, which had a M s = 6.8 and a thrust mechanism determined by BERGMAN and SOLOMON (1980). But hopefully, enforcing this stipulation will give us a hetter understanding of truly intraplate stresses, removed from the immediate inftuences of plate interactions.

Geographical Limits of Study

We have made certain exceptions to our intraplate regionalization, with the most notable being Hawaii. In the National Earthquake Information Center database of world seismicity (the "NEIC tape") there are 977 listings for Hawaii between 1868 and 1988. This is actually very small compared to the seismicity tabulated by the Hawaii Volcano Observatory, which recorded over 70,000 earthquak es during 1962-1983 (KLEIN et al., 1987). While this seismicity is certainly intraplate, its huge amount connected with hot spot volcanism makes it unreasonable as an inclusion in this study.

Approximately half of the total Pacific intraplate seismicity has occurred in the form of swarms located in less than a dozen distinct regions. A notable example is at the Gilbert Islands, where 224 earthquakes were detected teleseismically between 1981 and 1983 (OKAL et al., 1986). A discussion of these swarms appears in Appendix A, but will be brief because they have been studied elsewhere in more detail. The majority of our attention will focus on 894 independent events listed by previous agencies as having occurred in the intraplate regions of the Pacific Basin before 1989.

In most cases the geographical determination of a 2° buffer was straight forward due to the nume;ous well-defined plate boundaries in the Pacific region. On the basis of bathymetry, seismicity and numerous studies of plate motions, major subduction zones and mid-oceanic ridges and transforms can be clearly identified. There are areas, however, where sharply delineated plate boundaries do not exist, and we have excluded them from our study.

Perhaps the most seismically active of these is in the Caroline Islands region, whose scattered seismicity WEISSEL and ANDERSON (1978) descrihe as the diffuse plate boundary of a Caroline miniplate. A similar situation holds for the Fiji plateau, as described by CHASE (1971) and HAMBURGER et al. (1990).

We have also avoided the northernmost part of the Gulf of Alaska as it displays unusually high seismicity (including two recent M s = 7.6 intraplate earthquakes with aftershock distributions extending away from the coast) and seems to represent a fragmentation of the intraplate lithosphere and the establishment of a new plate boundary (LAHR et al., 1988).

Similarly, though the Shackleton Fracture Zone marks the proper boundary between the Antarctic and Scotia plates, we limit our study to events west of 68°W

266 Michael E. Wysession et al. PAGEOPH.

because the region between here and the Shackleton FZ is described by PELA YO

and WIENS (1989) as a diffuse seismic zone. The Juan de Fuca plate is also excluded, for though its boundaries are weIl

defined, its seismicity seems to suggest that it is undergoing internal deformation (STODDARD, 1987).

Sources

Our study, involving relocation and analysis of previously listed epicenters, is entirely dependent upon other supplying agencies for the original information. There are two requirements for these intraplate earthquakes which enable us to carry out relocations: they must have been located by a reporting agency (primary source) based on arrival times available at the time, and these epicenters must then be retrievable on a geographical basis from a presently accessible catalog (secondary source). In carefully reexamining these epicenters, our work represents the third stage.

Our main two sources of initial epicenters were the chronological NEIC tape and the database of WALKER (1989). Both sources, like the ISC, perform at two levels: they are responsible for some locations themselves, but also catalog earthquakes from other agencies. Taken as a whole there were several primary agencies from which we began our analyses, identified in Tables 1-13 by the following codes.

ISS = The International Seismological Summary (1913-1963). This was our preferred primary source for events through 1963 (when the ISS became the ISC), as it listed all arrival times received for a given earthquake as weIl as the location (though only through 1955, after which locations and arrival times were listed only for larger events). Their locations were often very unreliable as this was before the proliferation of computers and in early years they used the arrival times of surface waves to constrain locations. In many cases where they were unable to determine an epicenter but still listed enough body wave arrivals, this information was sufficient for us to determine a reliable location.

ISC = International Seismological Centre (1964-1988). This was the preferred primary source for events after 1963. They use a standardized location algorithm, many arrival phases are listed (though only P arrivals are used in locations), they supply magnitudes, standard deviations of the arrival time residuals and error estimates for the hypocenter parameters. Though there were many cases of ISC mislocations, for the most part their epicenters were reliable and within the standard error of our locations. The ISC performs both as a primary source (e.g., MUIRHEAD and ADAMS, 1986) and as a cataloging agency, publishing the listings of other agencies' solutions.

Vol. 135, 1991 Intraplate Seismicity 267

G = Gutenberg's personal anno ta ted copy of the ISS (SEISMOLOGICAL SOCIETY OF AMERICA, 1980), and

G-R = GUTENBERG and RICHTER (1941, 1954). The majority of these relocated as genuinely intraplate.

BCIS = Bureau Central International de Seismologie. This bulletin listed arrival times for its own locations as weil as those of other agencies, and was therefore very useful, especially during 1956-1963, the lean years of the ISS.

The following seven sources are all agencies of the Uni ted States government. Name changes often reftect no more than administrative reorganizations.

CGS = United States Coast and Geodetic Survey. NEIS = National Earthquake Information Service. Formerly known as the

CGS. NGDC = National Geophysical Data Center. PDE = Preliminary Determination of Epicenters. Published by NEIC. ERL = Environmental Research Laboratories. NOS = National Ocean Survey. Four earthquakes are referenced to the NOS,

all occurring on May 9, 1971 in the Northern Antarctic plate. GS = United States Geological Survey. USE = Uni ted States Earthquakes Bulletin. Of the 16 references to USE by the

NEIC tape, 15 were blatant typographical errors.

PPT = Laboratoire de Geophysique, Papeete, Tahiti. O}JS = Ocean Bottom Seismometer Array. From WALKER (1989). WIA = Wake Island Array. From WALKER (1989). W, E = Wake and Enewetak Arrays. From WALKER (1989). LAO = Large Aperture Seismic Array (LASA). The determinations by LASA

were the least reliable of the primary sources. Most of these locations are obtained from an interpretation of apparent slowness vectors across a small array. Modem work (e.g., CAPON, 1974; OKAL and KUSTER, 1975) has shown this procedure to be susceptible to errors from 10cal crustal structure. Though there are few actual references to LAO, many epicenters credited to the ISC were actually LASA determinations, and the relocations for these events often moved the epicenter great distances.

WEL = Wellington, Seismological Observatory. Of the 28 references to 10cal Southwest Pacific epicenters, 19 were genuinely intraplate, and our relocations usually would not change them.

T AC = Tacubaya, Instituto de Geofisica. The NEIC tape references T AC for five South American coast events, all of which relocate to the Central American trench.

The following three agencies were each referenced once by the NEIC tape, and in each case with a mistranscription.


JMA = Japan Meteorologieal Ageney. MOS = Moseow, Institute of Physies of the Earth. PAS = Pasadena, Califomia. Code reserved for loeal events in Southem Califomia.

The number of citations is by no means representative of an ageney's ability to deteet Paeifie intraplate events, but rather of our preferenees and those of NEIC and WALKER (1989) in seleeting primary sources for an event. The ageney listings are not independent, and most events are reeorded by at least two ageneies. In some eases the NEIC tape itself would have multiple listings (which we removed) for a single event beeause they differed enough in loeation and/or origin time (e.g., ISS, CGS and G + R).

The NEIC database that we used is the worldwide global set (mb ~ 3.0) in TAGGART et al. (1988). The WALKER (1989) dataset is a list of Paeifie intraplate earthquakes that he eonsiders to be reliable. Its strength lies in its access to loeal Mieronesian arrays and oeean bottom seismometers. The differenee between our study and W ALKER'S (1989) must be strongly stressed. Outside of the earthquakes he has loeated through loeal arrays, his study is only a eompilation of other events, however earefully they may be chosen. Our study eritieally analyzes eaeh individual earthquake, reloeating it if neeessary. This fundamental differenee is best illustrated by the result detailed in Seetion 5 that only 48% of Walker's events will be found genuinely intraplate by our standards.

The main two sourees, the NEIC tape and WALKER (1989), had surprisingly little overlap in their events: 24% be fore 1963, and 11 % post-62. While events listed by both tended to fare weIl in our reloeations and were most often genuinely intraplate, they were a small percentage of the total. Some of this lack in overlap is expeeted, but its magnitude is surprising. One would expeet the NEIC dataset to be included in the ISC, sinee it is prepared as mueh as a year earlier than, and is available to, the ISC. The dataset of WALKER (1989) is primarily from the ISC, but he uses a very eonservative and somewhat arbitrary parameterization of the intraplate region of the Paeifie, resulting in a signifieant lack of overlap between our two main sourees.

In eertain regions we also used the ISC regional eatalog as a souree for epieenter listings. While WALKER (1989) uses primarily ISC loeations, his parameterization of Paeifie plate boundaries uses latitude and longitude gridding on ascale ranging from 3 to 10°, eausing eertain large intraplate regions to be overlooked. As an example, 27 post-1963 intraplate epieenters had to be taken from the ISC regional eatalog east of the South Island of New Zealand, whieh is omitted from his regionalization. Most of these reloeated as genuinely intraplate.

The largest other eontribution to our dataset was for Polynesian events reeorded by the Polynesian Network (OKAL et al., 1980; updated by J. TALANDIER, pers. eommun., 1989) in an area mostly, but not totaIly, boxed out by WALKER (1989). Several events were also found in MIYAMURA (1988).


Our analysis begins with events occurring in 1913, because the International Seismological Summary (ISS) began publication then, printing original arrival times as weIl as hypocenters. We have also retained in our database two events from 1905 which, though without available arrival time information, were retained by both WALKER (1989) and MIY AMURA (1988). There were 10 other "intraplate" events referenced by MIYAMURA (1988) from before 1913 (mostly from the work of DUDA (1965)), but their proximity to the Tonga, Japan and Kuril subduction zones and their high magnitudes (7.0-8.0) suggest poorly located trench events.

3. Procedure

In relocating an event that has already been located, we seek to improve the reliability of the epicenter. Our ability to do so comes through the inclusion of more data, primarily by adding S arrivals to the relocation technique, through rejecting incompatible data more efficiently (notably by using data importances (MINSTER et al., 1974)), and through performing time-intensive computer-modeled statistical analyses. The large relocation vectors resulting from our analyses justify our efforts.

In determining an accurate database it was our hope at the onset that while events before 1963 would have to be relocated, events after this date would be found reliable. This was not the case. Many events, even those located in recent years, began as intraplate but ended as interplate after careful relocation, and of the 738 events considered (not associated with swarms or located by local networks) only 52 had hypocenters which we considered reliable enough not to warrant a relocation.

Our relocation was done through an interactive least-squares iterative regression of both P and S arrival times. This is different from the ISC inversion, which only uses P arrivals. The depth was usuaIly kept at 10 km, except in cases where there were enough arrivals, including those at nearby stations, to constrain it accurately. This is because, for Pacific intraplate events, most stations are at great teleseismic distances, resulting in the classic trade-off between depth and origin time. A depth of 10 km was chosen, consistent with the modem default value of the ISC and PDE for oceanic crustal faulting, and with recent studies of intraplate sources (BERGMAN and SOLOMON, 1980; WIENS and STEIN, 1984).

Our travel times are calculated using the Jeffreys-Bullen tables, both for P and

S. This model gives the travel times only for the Pn crustal phase at close distances, and being an averaged model, does not accurately represent local crustal structure. Thus there are cases, such as off-shore New Zealand events located by Wellington from local stations, where we feel that local agencies can use appropriate crustal models and additional crustal phases to better constrain the location. In such instances we have kept their locations, after checking their statistical significance.

The foIlowing sections detail the methods we used to significantly improve the reliability of our epicenters.


Inclusion of S Arrivals

Perhaps the greatest cause of large epicentral relocations was the inclusion of S arrivals in the inversions. Their addition often greatly helped to constrain the solution. S arrivals have been weighted by I/J3 to compensate for their slower velocities relative to P. Even so, errors in S wave arrival times were generally greater than for P, presumably due to the added complexity of choosing a later arrival.

An example can be seen for the May 18, 1974 earthquake listed by the ISC in the Caroline Islands region (Event 931). As shown in Figure 2, the ISC solution is

a. MAY 18, 1974

b. ISC LOCATION:

Ile 1Id 17h 3 .... 31'~'JI_ 1 :7~ :61NlC163:h:l7E "o~oooo .. 33k, R', 011111', 1 D, C-llne IlIMda Region

LAT lAe tAT L •• ESA E..AIa LMO lMnlnoton MTN M .. ton WRAW ........... ,.. aRS art-.

10'74 211 P S

".,. 1141P 11·13 201 P 21·81 23leP 21·72 220 P 28·.7 182 iP

174217·1 174309 174221'0 174225 174520 174630'1 1745311·5

+3'. -65

+1·1 -4,' +2-' -6,6 ."7

Figure 2

6

4

2

o

-2

c. RELOCATION:

S.lsoN IS0.1O"E d. 315 km O.T.- 17:41:11.4

Slaildard Deviation a 0.16 sec

Arrival DislallCe Residual Imponancc

LAT p 3.46 0.03 0.837 ESA P 4.63 -0.06 0.608 LMG P 4.22 0.19 0.416 MIN P 20.23 -0.06 0.531 WRA P 21.28 0.11 0.443 BRS P 22.32 -0.10 0.557 LAT S 3.46 -0.34 0.608

Example of relocation for Event 931 (May 18, 1974). (a) Map showing the ISC location in the Caroline Islands and the relocation under New Britain. (b) ISC bulletin listing. (c) Relocated hypocenter, with

greatly reduced residuals.


very impreeise, with a standard residual of 5.16 s for 6 arrivals, not including the S arrival at LA T whieh is off by over aminute. When the latter is included and the depth is floated, the solution eonverges very weil to the treneh. The hypoeenter has a depth of 315 km and is loeated beneath New Britain, a seismically active Benioff zone. Not only is the S arrival satisfied, but the P residuals are smaller as weil, with a standard residual for all 7 arrivals of 0.16 s, as shown in Figure 2c.

Floating the Depth

The ca se of Event 931 is also an example of how changing the depth will allow the hypocenter to locate in an aetive subduction zone. In cases like this the depth was varied either by the inversion algorithm itself, or if this proved unstable, by conducting several eonstrained reloeations at discretized depths and selecting the best one. In all cases we accepted an intermediate or deep hypocenter only if it coincided with a known Wadati-Benioff zone.

Removing Emergent Arrivals

Another source of error in bulletin locations is the inclusion of emergent arrivals, listed as eP or eS in bulletins. In some cases excellent solutions are obtained including all emergent arrivals, but in others these arrivals are clearly wrong and need to be removed from the inversion process. An example is Event 666 of February 7, 1974, which the ISC located 5° east of the Tonga Trench. The ISC solution fits ASP and WRA weil but otherwise does a poor job, especially at SPA, wh ich is inaccurate by aminute. This is questionable, because SPA is the only arrival marked impulsive, and even records a polarity. In addition we checked the original SPA record to guard against a clock error. When the two emergent arrivals at LA T and MTN are removed, the solution eonverges very weil to a loeation behind the subduetion zone with a standard residual of 1.13 s for five arrivals.

The process of rerunning the inversion many times after removing arrivals is philosophieally similar to a bootstrap technique, simply replacing the randomness of the choice of station to be removed with some better understanding of which arrivals may be suspeeL

Fixing Mislabelled Arrivals

An earthquake location can be in error due to the mislabelling of arrivals, either at the station, where a phase is incorrect1y identified, or during the location, with the inclusion of an arrival from a totally different earthquake. While the ISC does a very good job of screening out errors for the latter case, some still do slip through. An example is the May 23, 1967, earthquake (Event 664), located by the ISC (from a preliminary determination by LASA) in the Tokelau Island Region. Two entries from the ISC bulletin are shown in Figure 3, the second being a small

272 Michael E. Wysession el al.

3.

Preliminary determination given by lASA

837 May 23d 12h 3m 14<7.8'. Epieentre 8·,I.I·South by 170·,I.I·We,t. 9 ob, Oepth= O.OOOOR or H km. SO=2.7I' on 9 ob,. Mag =4 .4 on 5 ob,. IC (b25) Tokelau 1,Iands Region / (39) Paeifie Sa,in

GCA Gien Conyon 70.78 4b i 12 1430.2 +1.3 BMO Blut Mountains 71.00 37 i 121432.S +2.4 0.4 OUG Ougway 71 .43 43 i 12 14 34.b +1.Q 0.8 AL Q Albuquerque 73 .71 50 i 121444.S -1.9 o .b F G U Flaming Gorge 74.09 43 e 12 14 48 -0.4

CO L College OutPo,t 74.92 10 e 12 14 S3 +0.3 L F 3 lasa F Ring 77 .81 39 i 12 I S 07 .0 -1.8 0.7 WM 0 Wichita Mountains 79.71 S2 i 12 IS 13.8 -b .3 1.1

PAGEOPH,

L P S La Palma 83.21 75 i 12 IS 41 .IC +2.S ,p 12 IS S8 -597 +S

SSS May 23d 12h ISm 21s. 12.8·N 89.0 ·W. Oepth=SO km e7b) Off Coast of Central America / (b) Central America

838 The distances and residuals correspond to the position and time given by SSS

S S S San Salv.dor o .qO 348 i 12 15 H.7 -2.8

b.

15.67°S I 73.36°W d= 10.0 km O.T.= 12:02:30.3

Standard Deviation = 0.27 sec

Arrival Distance Residual Importance

GCA p 78.30 0.37 0.269 BMO P 78.91 -0.51 0.427 DUG P 79.12 0.35 0.224 ALQ P 81.02 -0.02 0.231 FGU P 81.74 -0.14 0.136 COL p 82.77 0.11 0.929 LF3 P 85.62 -0.04 0.163 WMO P 86.84 -0.11 0.621

Figure 3

12 15 45 -0

Example of epicenter contamination by arrivals from a different earthquake for Event 664 (May 23, 1967). (a) The S arrival at LPS (EI Salvador) for this event (top listing) is 597 s early, but LPS Sand P fit the following San Salvador event very weil. (b) Removal of LPS from the first event causes it to

relocate in the Tonga trench.

local earthquake detected only by San Salvador (SSS). The Tokelau Island event is largely constrained by the P arrival at LPS (La Pairna, EI Salvador), and the solution is only fair, with a standard residual of2.71 s for 9 P arrivals. The clue that something is amiss comes from the LPS S arrival, which if taken as an S is 597 s early, so the ISC interprets it as an sP. If, on the other hand, both LPS arrivals are from the event recorded subsequently by San Salvador, they agree excellently as they are, and the "Tokelau Islands" earthquake relocates to (15.67°S, 173.36°W), square in the trench, with a standard residual of 0.27 s for 8 P arrivals.

An example of a phase misidentification would be Event 1404, which occurred on July 2, 1980, and was located by the ISC in the northem part of the Cocos plate at (11.9°N, 96.9°W). The arrival time residuals for the ISC location were poor, with a = 6.28 s from 11 observations. We expected the relocation, therefore, to be to the


treneh near the northeast. Three arrivals from Colombia (BOG, FUQ and CHN) were ineompatible with the other stations, all in the United States, which aetually gave a superior loeation on a transform segment of the Galapagos Rise, to the southwest of the ISC loeation. When the BOG and FUQ arrivals were interpreted as S waves instead of P waves they agreed with the other P arrivals, and the final loeation (0" = 1.29 s for 10 observations) was at (3.09°W, 90.03°W), less than 1° north of the very seismieally aetive Galapagos Rise trans form fault. The CHN arrival is 20 s late when interpreted as an S wave, but comes at the appropriate time when interpreted as SS.

Monitoring Data Importances

We tried to guard against misloeations eaused by the heavy weighting of partieular arrivals in the inversion program. An arrival with a unique baek-azimuth will be given a higher data importanee and small errors in them ean greatly change the epieenter. In suspeet eases we both tested the inversion without them and examined the original seismograms, remeasuring the arrival times.

Determining Aftershock Relations

Common sources of misloeations were aftershoeks of large interplate earthquakes. Most of the supposed intraplate earthquakes off the co asts of Japan, the Kuril Treneh and the Aleutian Treneh reloeated into swarms of aftershoeks from large subduetion zone events. An example is shown in Figure 4, where the seismieity for June, 1975 is shown for the treneh off the eoast of Hokkaido, Japan. The seismieity is fairly quiet exeept for a large cluster of events eentered at (43°N, 145SE) eontaining over 180 aftershoeks of the June 10 "tsunami" earthquake (FUKAO, 1979). In Figure 4a the three intraplate listings (1034-1036) are shown with solid eircles, but their reloeations, aeeomplished by removing ineompatible emergent arrivals, are part of the aftershock sequenee shown in Figure 4b.

Monte Carlo Error Ellipses

While the inversion is a straight forward iterative proeess, using the partial derivatives of the hypocenter parameters with respeet to the arrival time residuals to determine the next loeation, the aetual process using real data is more like detective work, as there are many potential pitfalls. With perfect da ta there would be only one minimum weil in the N-dimensional residual spaee (determined by the number of arrival times) and the iterative inversion would quiekly find the hypocenter parameters that yielded travel time residuals eorresponding to this minimum. Using actual data, however, with errors from misset clocks, poody read arrivals, eontamination from reeord noise and arrivals from different events, the residual

274 Michael E. Wysession er al. PAGEOPH,

47

45

g 43

.., C> :> l::

3 41

39

37

LO~GITUOE (El

Figure 4 (a) Seismicity (open circles) for lune 1975 off the coast of Hokkaido, with the cluster centered around the lune 10 "tsunami" shock. The solid circles are three aftershocks of this event mislocated as intraplate. (b) The three events, when relocated, form part of the aftershock sequence. Many events in

the lapan-Kuril-Aleutian intraplate areas were aftershocks that relocated to trenches.

space can contain local minima into which the inversion can converge even though they are not the best solutions. A careful way of using the least-squares inversion, however, is to accompany it with a critical statistical analysis.

We use a Monte Carlo statistical simulation to give us an estimate of the reliability of our epicenters. As weil as possible pitfalls listed above, another danger with the least squares regression results from having a nonuniform distribution of event-to-station azimuths. A location can be done with a huge number of P and S arrivals, but if they are all at the same azimuth the epicenter will be very unconstrained in a direction perpendicular to the azimuth. Such situations are common in examining small Pacific events which often are only detected by a handful of stations at the nearest co ast. The Monte Carlo simulation, generating a confidence ellipse, gives us a quantitative estimate of how weil constrained an epicenter iso

The Monte Carlo test is extremely robust because it takes into account the nonlinearities of the inversion method by simply adding noise to the arrival times and rerunning the inversion process hundreds of times. The distribution of new epicenters gives an indication of the reliability of the best fit epicenter. We start each Monte Carlo simulation with the best epicenter and origin time found from our iterative inversion and then add random errors to all the arrival times and iterate the inversion to its best new location. Repeating this procedure a large number of times-we generally used 400-gives a significant distribution, whose


covariance matrix we use to obtain a 95% confidence ellipse. Having a confidence ellipse was useful though not always meaningful, as the Monte Carlo distributions were not necessarily Gaussian in latitude or longitude.

We used Gaussian deviates for our random errors, with standard deviations that varied according to the date and quality of the location. For recent events we used a standard deviation of 2 s for the Gaussian deviate errors to the arrival times, but for 1925 we might use 20 s. Because of the somewhat arbitrariness of this standard deviation, it was not so much the absolute size of the error ellipse that was important as the shape of the Monte Carlo distribution in relation to proximate plate boundaries. An example is shown in Figure 5 for the December 2, 1975 earthquake in the South Pacific (Event 443). Though the event locates west of the Pacific-Antarctic ridge, the 95% confidence ellipse, using only a 2 s standard deviation for travel time errors, makes this event unreliable as intraplate. The shape of the ellipse shows that while the epicenter is well constrained in the east-west

DECEMBER 2, 1975

-40

-45

-50

-55

-60

L...---I. ___ ..1..:-__ :....-J'-::---' -65 -130 -120

LONGITUDE (E)

Figure 5 Example of a Monte Carlo simulation for Event 443. The solid dots are the 500 simulation epicenters, done with a 2 s standard error, and the ellipse is the 95% confidence interval through them. The ISC and relocated epicenters are identical (buIl's eye) and are 4° from the Pacific-Antarctic ridge (dashed line), but the ellipse crosses two seismically active transform segments of the plate boundary, implying that the

event is most Iikely interplate.


direction, it is very poorly constrained north-south, and is most likely from a seismically active transform segment where the ellipse crosses the ridge. The uncertainty in location by the ISC (±2.2° in latitude) is elearly underestimated.

When the Monte Carlo distribution elearly crossed a plate boundary, the event was considered to be most probably interplate and was rejected.

Relocations to Islands

Earthquake relocations occasionally converge on epicenters in the immediate vicinity of populated islands. Under such conditions, an additional constraint should be the existence, or the documented absence, of feit reports. Several of our events were indeed reported feit (e.g., Event 214 on October 2, 1988, felt all over the Marquesas). In many instances, however, feit and instrumental reports were not correlated: In particular, earthquakes have been reported feit at intraplate locations (e.g., at Pohnpei on April ll, 1911) in the absence of any instrumental reports.

Proposing a relocation in the immediate vicinity of a populated island should not necessarily be contradicted by the absence of a feit report. The situation ean be iIlustrated best on the example of Event 177 (July 17, 1929), whose reloeated epieenter is just 25 km NE of Rarotonga in the Cook Islands. Body and surface waves from this earthquake were recorded (although imprecisely) in Australia and Ameriea. Thus the event probably has a magnitude of 5 to 5 1/2: Our present-day experienee with magnitude 4 1/2 events in the region is that they eseape detection outside Polynesia (despite improved instrumentation at teleseismic stations), while magnitude 6 events were better deteeted and located, even in those times. In the marine environment, and assuming similar source and propagation characteristics, our present-day experienee (for example in Hawaii) is that an earthquake of mb = 5-5.5 would be feit, at the most, 150 km away. This distance is less than the half-Iength of the Monte Carlo ellipse computed for the event, indieating that our relocation does not suggest, at a statistically signifieant level, that the event had to have been feit. At any rate, and assuming the earthquake would indeed have been feit, we eould not identify any newspaper published in the Cook Islands in 1929; a feit report eould therefore have simply been lost over the past 60 years.

Similarly, in the ease of Micronesia and the Marshall Islands, a number of events (886,887,896,898,899,900,901,902,907,908,913,914, 915) were identified as elose enough to populated islands that they eould have been feit. A systematic seareh of loeal newspaper eolleetions around the dates of the events (M. T. WOODS, pers. eommun., 1990) has failed to yield any report of felt earthquakes, even for Events 902 and 915, reported felt by the ISC.

As a result, we have chosen to list a few events in the immediate vicinity of populated islands, even in the absence of feit reports.


4. Results: Classification 0/ Events

The results of the analysis of the 894 non-swarm earthquakes are shown in Tables 1-13. For convenience we have divided the dataset into 13 parts, treating the Cocos, Nazca and Antarctic plates separately, and subdividing the Pacific into 10 separate regions. These subdivisions do not necessarily reflect distinct tectonic features, but rather facilitate an easier discussion of the Pacific intraplate seismicity with some geographical continuity. Some regions are obviously smaller than others, and feature fewer intraplate earthquakes; this situation is at least partly due to a higher casualty rate in these areas (e.g., Tonga to Samoa, and Japan-Kuriles). The locations of these regions are shown in Figure 6, and detailed descriptions of the regionalizations can be found in Appendix B.

Each table consists of a source code, date, initial hypocenter and magnitude, final hypocenter (where appropriate), processing code, site location (where given) and index number. The events are grouped within each table along geographical lines, in order to more easily recognize any spatial trends in the seismicity. Among casualties, events relocated to the plate boundaries (codes 7, 8, 9 and 10) have been

BOUNDRRY

'"'b

T POLYNESIR

SOUTHCENTRI=lL

I=lNTI=lRCTIC

Figure 6 Map of the regionalization of the Pacific Basin used in sorting the dataset into Tables 1-13. "J-K" stands for Off Japan and the Kuriles, "S-G" for Samoa to Gilbert, and "T _S" for East of Tonga and

Samoa.


TABLE 1. SORTED DATASET FOR POL YNESIA

Initial Localion ReJocation Code Site Index

See 0. .. Time Lot. Loal· Depdt Ma,. LaL Loa •. DepIh Tune NI Cf

DMY (GM1) "N oE (km) "N "E (km) (GMT) (s)

IlIlrap/ale Ew:nts

CGS 25 I 1962 10:03:06.8 .... 60 -152.60 33 4.6PPT -4.54 -152.59 10 10:03:05.5 38 1.38 U-O 101 !SC 13 41967 14:26:51.0 -6.80 -151.10 37 5.1 mb -6.80 -151.10 0 14:26:46.0 U-1 102 PPT 27 3 1966 04:26:43.0 -11.70 -151.40 0 4.2 PPT- TU-I 103 PPT 24 12 1977 11:08:11.0 -13.00 -151.30 0 2.9PPT TU-2 104 PPT 04 31977 22:00:15.0 -12.50 -150.20 0 3.9 PPT 5 TU-3 105 PPT 21 111972 16:28:33.0 -13.30 -149.50 0 3.7 PPT 5 TU-4 106 PPT 18 11974 08:01:56.0 -14.10 -148.40 0 2.0 PPT 5 TU-5 107 PPT 02 12 1974 17:23:30.0 -16.00 -145.10 0 3.2 PPT 5 TU-6 108 PPT 10 61975 19:35:12.0 -16.60 -145.10 0 2.5PPT 5 TU-7 109 PPT 24 91976 18:29:39.0 -17.30 -144.70 0 -3.1 PPT 5 TU-8 110 PPT 06101976 02:58:17.0 -17.30 -144.70 0 2.7 PPT 5 TU-8 111 PPT 05 10 1976 21:00:49.0 -17.30 -144.70 0 2.7PPT 5 TU-8 112 PPT 07 10 1976 02:06:24.0 -17.30 -144.70 0 2.1 PPT 5 TU-8 113 Pf\' 10101976 03:40:41.0 -17.30 -144.70 0 2.7PPT 5 TU-8 114 PPT 10101976 04:09:23.0 -17.30 -144.70 0 2.8PPT 5 TU-8 115 PPT 09 10 1976 21:35:52.0 -17.30 -144.70 0 2.4PPT 5 TU-8 116 PPT 10101976 13:10:13.0 -17.30 -144.70 0 3.3 PPT 5 TU-8 117 PPT 16101976 05:10:16.0 -17.30 -144.70 0 2.7PPT 5 TU-8 118 PPT 21 101976 00:57:26.0 -17.30 -144.70 0 2.7PPT 5 TU-8 119 PPT 23 10 1976 11:30:11.0 -17.30 -144.70 0 2.4PPT 5 TU-8 120 PPT 01 11 1976 12:06:00.0 -17.30 -144.70 0 3.1 PPT 5 TU-8 121 PPT 12121976 03:55:32.0 -17.30 -144.70 0 2.3 PPT 5 TU-8 122 PPT 10 11977 17:13:38.0 -17.30 -144.70 0 2.3 PPT 5 TU-8 123 PPT 26 41977 10:37:01.0 -17.30 -144.70 0 2.4 PPT 5 TU-8 124 PPT 13 21979 12:36:59.0 -14.20 -140.50 0 3.8 PPT 5 TU-9 125 GS 16 41982 02:52:52.1 -15.64 -137.93 0 4.8 mb 5 TU-la 126 PPT 25 21986 04:35:32.0 -12.47 -151.71 12 4.2 PPT 5 TU-li 127 PPT 10 71969 08:01:21.0 -14.90 -151.40 0 2.4PPT 5 SC-1 128 PPT 23 91969 16:28:52.0 -14.90 -151.40 0 2.5PPT 5 SC-1 129 PPT 31 31972 17:33:26.0 -14.90 -151.40 0 3.0 PPT 5 SC-1 130 PPT 18101972 04:23:48.0 .. 14.90 -151.40 0 2.1 PPT 5 SC-1 131 PPT 12 51973 05:53:02.0 -14.90 -151.40 0 2.6PPT 5 SC-1 132 PPT 12 51973 08:31:56.0 -14.90 -151.40 0 2.5 PPT 5 SC-1 133 PPT 18 71973 09:14:52.0 -14.90 -151.40 0 2.5 PPT 5 SC-1 134 PPT 05 51974 01:08:57.0 -14.90 -151.40 0 2.5PPT 5 SC-1 135 PPT 10 61975 18:28:25.0 -14.90 -151.40 0 2.1 PPT 5 SC-I 136 PPT 20 3 1976 07:06:57.0 -14.90 -151.40 0 3.6 PPT 5 SC-1 137 PPT 11 9 1977 00:06:01.0 -14.90 -151.40 0 2.3 PPT 5 SC-1 138 PPT 08 31978 02:43:46.0 -14.90 -151.40 0 2.4PPT 5 SC-1 139 PPT 14 81971 21:31:55.0 -14.70 -151.30 0 2.1 PPT 5 SC-2 140 PPT 1091971 11:13:19.0 -14.70 -151.30 0 3.2 PPT 5 SC-2 141 PPT 14 21974 17:07:55.0 -14.70 -151.30 0 2.5 PPT 5 SC-2 142 PPT 12 11 1985 21:16:05.9 -14.80 -151.30 12 3.9 PPT 5 SC-2 143 PPT 1411 1979 08:02:49.0 -16.70 -151.40 0 2.4PPT 5 SC-3 144 !SC 06 41979 22:49:53.1 -15.90 -150.50 0 5 SC-4 145 PPT 05 11 1988 14:02:54.8 -15.20 -150.78 12 3.2 PPT 5 SC-5 146 PPT 01 41968 14:56:34.0 -19.00 -156.50 0 3.7 PPT 5 NA-1 147 PPT 08 10 1968 17:17:12.0 -20.10 -152.00 0 3.3 PPT 5 NA-2 148 PPT 21 121973 10:00:00.0 -20.10 -152.00 0 2.3 PPT 5 NA-2 149 PPT 15 51974 16:56:07.0 -20.10 -152.00 0 2.5PPT 5 NA-2 150 PPT 07 91975 02:12:24.0 -20.10 -152.00 0 2.8PPT 5 NA-2 151 PPT 15 51976 09:58:43.0 -19.80 -150.10 0 2.6PPT 5 NA-3 152 PPT 18 81974 07:34:25.0 -21.10 -149.60 0 2.2PPT 5 NA-4 153 PPT 29 91974 05:04:18.0 -21.10 -149.60 0 2.8PPT 5 NA-4 154 PPT 20 I 1975 03:42:46.0 -21.10 -149.60 0 2.9PPT 5 NA-4 155 PPT 22 51975 05:50:45.0 -21.10 -149.60 0 lOPPT 5 NA-4 156 PPT 22 51975 05:50:45.0 -21.10 -149.60 0 3.0PPT 5 NA'" 156 PPT 06 61975 20:02:13.0 -21.10 -149.60 0 2.4PPT 5 NA'" 157 PPT 30 71975 02:37:46.0 -21.10 -149.60 0 3.3 PPT 5 NA'" 158 PPT 16101975 09:07:28.0 -21.10 -149.60 0 2.4PPT 5 NA-4 159 PPT 15 11 1975 20:34:47.0 -21.10 -149.60 0 3.4PPT 5 NA'" 160 PPT 16 11 1975 10:13:48.0 -21.10 -149.60 0 2.6PPT 5 NA'" 161 PPT 02 12 1975 12:11:43.0 -21.10 -149.60 0 2.3 PPT 5 NA-4 162 PP!' lS 3 1976 21:54:Z7.0 -21.10 -149.60 0 2.7 PP!' 5 NA-4 163 PP!' 12 51976 02:30:26.0 -21.10 -149.60 0 3.7 PPT 5 NA'" 164 PP!' 20 51977 13:43:05.0 -21.10 -149.60 0 2.8PPT 5 NA'" 165 PP!' 24 51977 06:08:54.0 -21.10 -149.60 0 2.8PPT 5 NA-4 166 PPT 29 71977 02:13:10.0 -21.10 -149.60 0 3.0PPT 5 NA'" 167 PPT 27 81977 01:08:36.0 -21.10 -149.60 0 1.7PPT 5 NA'" 168


Table 1 (continued) Sor/ed Da/ase/ for Polynesia

Initial Location Relocation Code Site Index

See Date Time Lot. Lan,. DepIh Ma,. l.aL Lan .. DepIh Tune NI " DMY (GMT) ON oE (km) oN "E (km) (GMT) (I)

PPT 00 91977 14:23:46.0 -21.10 -149.60 0 2.6PPT 5 NA-4 169 PPT 13 91977 03:11:49.0 -21.10 -149.60 0 3.4 PPT 5 NA-4 170 PPT 02 11 1977 03:38:55.0 -21.10 -149.60 0 2.8PPT 5 NA-4 171 PPT 22 12 1977 05:58:57.0 -21.10 -149.60 0 2.8PPT 5 NA-4 172 PPT 05 81978 09:49:45.0 -21.10 -149.60 0 2.8PPT NA-4 173 PPT 14 41979 09:19:48.0 -21.10 -149.60 0 2.8PPT NA-4 174 PPT 14101979 13:07:00.0 -21.10 -149.60 0 2.3 PPT NA-4 175 PPT 18101979 23:54:20.0 -21.10 -149.60 0 3.2 PPT NA-4 176 ISS 17 71929 19:52:12.0 -22.00 -160.50 0 -21.08 -159.76 10 19:52:18.3 2.55 NA-5 177 PPT 03 21970 14:47:51.0 -24.10 -151.80 0 3.5 PPT 5 AU-l 178 PPT 20 5 1970 16:31:53.0 -24.10 -151.80 0 3.5 PPT 5 AU-l 179 PPT 04 81970 11:02:23.0 -24.\0 -151.80 0 4.0PPT 5 AU-l 180 PPT 13 81970 12:04:31.0 -24.10 -151.80 0 4.4 PPT 5 AU-l 181 PPT 14 51970 09:35:52.0 -27.40 -148.60 0 4.5 PPT 5 AU-2 182 PPT 24 41977 16:31:03.0 -27.40 -148.60 0 3.4 PPT 5 AU-2 183 PPT 20 51977 20:30:53.0 -27.40 -148.60 0 3.3 PPT 5 AU-2 184 PPT 28 61977 03:37:01.0 -27.40 -148.60 0 4.1 PPT 5 AU-2 185 PPT 29 61977 05:27:41.0 -27.40 -148.60 0 3.6PPT 5 AU-2 186 PPT 16 91977 16:08:46.0 -27.40 -148.60 0 3.1 PPT 5 AU-2 187 PPT 20 91977 12:33:40.0 -27.40 -148.60 0 3.0PPT 5 AU-2 188 PPT 12101977 19:09:17.0 -27.40 -148.60 0 3.3 PPT 5 AU-2 189 PPT 20 10 1977 02:37:29.0 -27.40 -148.60 0 3.5 PPT 5 AU-2 190 PPT 1511 1977 18:57:16.0 -27.40 -148.60 0 4.6PPT 5 AU-2 191 PPT 31 I 1978 11:10:06.0 -27.40 -148.60 0 2.9PPT 5 AU-2 192 PPT 14 21978 0\:33:26.0 -27.40 -148.60 0 3.2 PPT 5 AU-2 193 PPT 08 3 1978 10:10:26.0 -27.40 -148.60 0 3.2 PPT 5 AU-2 194 PPT 25 3 1978 06:44:09.0 -27.40 -148.60 0 3.1 PPT 5 AU-2 195 PPT 31 3 1978 23:48:24.0 -27.40 -148.60 0 3.1 PPT 5 AU-2 196 PPT 05 5 1978 13:29:00.0 -27.40 -148.60 0 2.8 PPT 5 AU-2 197 PPT 21 51978 00:12:17.0 -27.40 -148.60 0 3.7 PPT 5 AU-2 198 PPT 08 61978 20:40:00.0 -27.40 -148.60 0 3.3 PPT 5 AU-2 199 PPT 09 61978 17:06:46.0 -27.40 -148.60 0 3.3 PPT 5 AU-2 200 PPT 09 61978 16:54:04.0 -27.40 -148.60 0 3.3 PPT 5 AU-2 201 PPT 10 61978 12:01:15.0 -27.40 -148.60 0 3.2 PPT 5 AU-2 202 PPT 10 61978 13:19:53.0 -27.40 -148.60 0 3.4 PPT 5 AU-2 203 PPT 12 61978 01:09:11.0 -27.40 -148.60 0 3.0 PPT 5 AU-2 204 PPT 12 61978 15:01:58.0 -27.40 -148.60 0 2.9PPT 5 AU-2 205 PPT 16 51979 02:36:56.0 -27.40 -148.60 0 3.6 PPT 5 AU-2 206 PPT 13 1 1972 17:48:27.0 -23.60 -146.90 0 4.0 PPT 5 AU-3 200 PPT 12 3 1977 17:31:00.0 -27.10 -144.40 0 3.7 PPT 5 AU-4 208 PPT 20 31977 01:02:48.0 -27.10 -144.40 0 4.2 PPT 5 AU-4 209 PPT 28 81977 01:41:31.0 -27.10 -144.40 0 4.5 PPT 5 AU-4 210 PPT 20 11 1979 06:45:04.4 -26.57 -138.84 0 5.3 PPT 5 AU-6 211 PPT 06 10 1982 08:29:44.0 -23.56 -140.11 0 4.6PPT 5 AU-7 212 PPT 1805 1987 18:33:17.4 -23.50 -149.03 12 4.0PPT 5 AU-8 213 PPT 02 10 1988 23:41:14.0 -9.47 -140.00 0 4.4 PPT 5 MQ-1 214 PPT 14 5 1975 11:49:42.3 -17.43 -136.05 0 5.0 mb 5 GB-1 215 PPT 01 31974 02:31:27.0 -17.00 -135.\0 0 3.3 PPT 5 GB-2 216 PPT 27 31970 22:25:25.0 -18.10 -133.70 0 3.\ PPT 5 GB-3 217 PPT 06 91971 22:00:56.0 -18.10 -133.70 0 3.5 PPT 5 GB-3 218

06 3 1965 11:10:53.1 -18.40 -132.90 0 5.5 mb -18.40 -132.85 10 11:10:52.2 3 GB-4 219 18 91966 06:40:35.5 -18.48 -132.81 0 5.0mb . -18.40 -132.86 10 06:40:35.8 3 GB-4 220

PPT 01 51968 08:50:03.0 -18.40 -132.80 0 4.5 PPT 5 GB-4 221 PPT 18 81968 05:41:28.0 -18.40 -132.80 0 4.7 PPT 5 GB-4 222 PPT 00 81971 14:47:59.0 -18.40 -132.80 0 4.2 PPT 5 GB-4 223 PPT 13 51972 11:24:49.0 -18.40 -132.80 0 3.3 PPT 5 GB-4 224 PPT 15 51975 02:09:35.0 -18.40 -132.90 0 3.1 PPT 5 GB-4 22S GS 25 5 1975 14:16:31.3 -18.46 -132.99 19 5.0 mb -18.40 -132.92 10 14:16:32.8 3 GB-4 226 PPT 13 9 1975 18:17:01.0 -18.40 -132.90 0 3.6 PPT 5 GB-4 227 PPT 15 5 1976 08:02:12.0 -18.40 -132.90 0 3.3 PPT 5 GB-4 228 PPT 22 91977 00:02:51.0 -18.40 -132.90 0 2.7PPT 5 GB-4 229 PPT 31 5 1979 0\:15:49.0 -18.40 -132.90 0 3.3 PPT 5 GB-4 230 PPT 30 I 1978 18:34:19.6 -16.23 -126.94 0 5.0 mb 5 GB-6 231 PPT 06 11 1978 09:05:54.0 -22.00 -132.00 0 4.5 PPT 5 GB-7 232 PPT 05 21986 04:04:36.0 -21.18 -135.10 12 3.9 PPT 5 GB-10 233 \SC 09 5 1974 09:52:14.7 -5.20 -129.20 33 -5.20 -129.20 33 09:52:14.7 2 SP-3 234 \SC 16 11 1983 12:31:21.8 -11.70 -126.10 10 4.8 mb 3 SP-4 235 ISC 08 11 \969 17:05:\3.0 -12.00 -140.30 33 4.3 mb -7.40 -148.00 10 17:05:17.0 6 U-2 236

280 Michael E. Wysession el al. PAGEOPH,

Table 1 (continued)

Sorted DatrJStlt tor Poly1lllsia

Initial Location Relocation Code Sile Index

Sec Da .. Time Lot. !.on,. DepIh Ma,. LaL !.on •. DepIh Tune N. (J

DMY (GMl) "N Da (km) ON Da (km) (GMl) (.)

Casua/ties WEl 27 91973 02:01:21.0 ·16.00 ·160.00 33 ·23.36 179.51 646 02:07:32.6 0.00 7 237 ISS 21 61918 03:59:05.0 ·22.00 ·141.00 0 1\ 231 ISS 17 41927 09:05:42.0 -6.50 ·126.00 160 ~5O 126.00 160 14 239 CGs 12 11942 16:07:11.0 .. .00 ·156.50 0 ·8.00 156.50 14 240 CGs 24 81961 18:29:58.9 ·22.21 ·138.81 0 5.0 mb 15 241 CGS 08 91961 18:59:59.3 ·21.78 ·139.16 0 4.7 mb 15 242 CGS 30 51970 17:59:58.5 ·22.24 ·138.79 0 4.7 mb 15 243 CGS 03 71970 18:29:59.1 ·21.80 ·139.19 0 4.8 mb 15 244 ERl 14 11971 18:59:59.2 ·21.88 ·138.95 0 4.7 mb 15 245 GS 11 71976 00:29:54.8 ·22.67 ·138.61 0 5.0 mb 15 246 GS 19 21977 23:29:51.9 ·22.10 ·138.76 0 5.3 mb 15 247 GS 19 3 1977 23:00:58.2 ·21.93 ·138.96 0 5.9 mb 15 248 GS 24 11 1977 16:59:58.5 ·21.89 ·138.95 0 6.0 mb 15 249 GS 22 3 1971 17:29:51.2 ·22.10 ·138.56 0 4.8 mb 15 150 GS 30 11 1978 17:31:58.4 ·21.90 ·138.97 0 5.9 mb 15 251 GS 19121971 16:57:00.1 ·21.73 -139.04 0 4.9 mb 15 152 GS 24 31979 16:27:58.8 -21.82 -139.06 0 4.9 mb 15 153 GS 18 61979 23:26:58.4 -22.04 -138.62 0 5.0 mb 15 154 GS 29 61979 18:55:51.9 -21.99 -138.88 0 5.3 mb 15 155 GS 15 71979 17:56:58.3 -21.89 -138.99 0 6.0 mb 15 256 GS 28 71979 19:55:59.2 -21.71 -138.90 0 4.4 mb 15 157 GS 23 3 1980 19:36:58.4 -21.87 -139.02 0 5.7 mb 15 151 GS 01 41980 19:30:58.6 -21.83 -138.84 0 5.1 mb 15 159 GS 16 61980 18:26:51.8 -22.02 -138.87 0 5.5 mb 15 260 GS 19 71980 23:46:51.2 -21.89 -139.02 0 5.9 mb 15 261 GS 03 121980 17:32:58.2 -21.94 -138.96 0 5.6 mb 15 262 GS 08 71981 22:22:51.9 -22.01 -138.90 0 5.3 mb 15 263 GS 03 81981 18:32:51.5 -22.07 -138.78 0 5.3 mb 15 264 GS 11 11 1911 17:06:58.3 -21.92 -138.99 0 4.5 mb 15 2M GS 05 12 1981 16:51:58.4 -22.03 -138.60 0 4.6 mb 15 266 GS 08 12 1981 16:46:58.4 -21.91 -138.89 0 5.2 mb 15 267 GS 20 3 1982 17:02:51.9 -21.99 -138.94 0 5.0 mb 15 261 GS 01 71912 17:01:58.8 -21.76 -139.05 0 5.2 mb 15 269 GS 15 71982 18:01:58.1 -21.86 -138.94 0 5.6 mb 15 270 GS 19 41983 18:52:58.4 -21.84 -138.90 0 5.6 mb 15 271 GS 15 5 1913 17:30:58.2 -21.89 -138.92 . 0 5.9 mb 15 272 GS 21 61913 17:45:58.6 -21.74 -138.92 0 5.4 mb 15 273 GS 04 81913 17:\3:58.2 -21.83 -138.92 0 5.2 mb 15 274 PDE 12 51984 17:30:58.3 -21.85 -138.96 0 5.7 mb 15 275 PDE 16 61914 17:43:51.9 -21.93 -138.99 0 5.3 mb 15 276 PDE 02 11 1984 20:44:58.5 -21.88 -138.99 0 5.7 mb 15 277 PDE 06 12 1984 17:21:58.6 -21.85 -138.91 0 5.6 mb 15 271 PDE 30 41985 17:28:51.9 -22.07 -138.90 0 4.5 mb 15 279 PDE 08 51985 20:27:58.8 -21.82 -139.05 0 5.7 mb 15 280 PDE 03 61985 17:29:58.0 -22.05 -138.86 0 5.2 mb 15 211 PDE 26 10 1985 16:34:58.3 -21.85 -138.97 0 5.4 mb 15 282 PDE 24 11 1985 16:00:58.5 -21.86 -\38.76 0 4.7 mb 15 283 PDE 26 1\ 1985 17:41:58.4 -21.86 -138.93 0 5.8 mb 15 284 GS 26 41986 17:01:56.7 -22.15 -139.11 0 4.8 mb 15 285 GS 30 51986 17:24:58.3 -21.90 -139.03 0 5.7 mb 15 286 GS 1211 1986 17:01:58.3 -21.91 -139.08 0 5.3 mb 15 217 GS 10121986 17:14:58.3 -21.92 -138.92 0 5.2 mb 15 288 GS 05 51987 16:57:51.7 -21.90 -139.10 0 4.9 mb 15 289 GS 20 5 1987 17:04:58.3 -21.89 -138.96 0 5.6 mb 15 290 GS 06 61987 17:59:59.3 -21.75 -138.89 0 4.7 mb 15 291 GS 21 61987 17:54:51.9 -22.05 -138.71 0 5.1 mb 15 292 GS 23 101987 16:49:58.8 -21.82 -139.03 0 5.5 mb 15 293 GS 05 11 1987 17:29:59.1 -21.79 -139.00 0 5.8 mb 15 294 GS 19 11 1987 16:30:58.3 -21.91 -139.D1 0 5.9 mb 15 295 GS 11 5 1988 16:59:58.4 -21.87 -139.07 0 5.5 mb 15 296 GS 25 5 1988 17:00:58.4 -21.90 -139.D1 0 5.6 mb 15 297 GS 23 61988 17:30:58.6 -21.91 -139.02 0 5.3 mb 15 298 GS 05 11 1988 16:29:58.4 -21.88 -139.04 0 5.4 mb 15 299 GS 23 11 1988 17:00:58.0 -21.99 -138.90 0 5.4 mb 15 300 GS 30 11 1988 17:54:58.0 -22.24 -138.83 0 5.5mh 15 301

Vol. 135, 1991 Intraplate Seismicity

Table 2

Sorted Dataset for Southcentral Paciftc (Rtcept Polynesia)

Initial Location

See Date DMY

CGS 05 I 1945 ISS 21 41960 ISC 22 71972 GS 30 91981 GS 04 10 1981 ISC 30 I 1980 GS 23 I 1988 CGS 22 11 1955 ISS 29-9 1951

CGS 03 81951 ISC 17 I 1976 CGS 10 11 1968 ERL 09 91971 CGS 14 91963 ISC 29 3 1975 GS 21 61988 ISS 22 31921 ERL 28 91972 BCI 14 71951 G 05 91938

G·R 15 12 1947 CGS 10 41950 ISS 17 10 1959 ISC 21 10 1976 GS 21 10 1988

Tune (GM1)

06:18:23.0 02:16:32.0 11:11:09.8 23:03:47.5 17:10:22.8 00:31:14.0 22:01:57.3 03:24:00.0 18:15:00.0 19:20:15.0 05:44:14.2 08:29:44.6 03:25:47.8 16:16:51.8 15:02:29.6 19:07:20.2 11:54:27.0 06:57:35.2 06:21:14.0 14:42:32.0 19:20:26.0 06:06:46.0 08:35:00.0 03:56:28.1 13:45:46.9

LaL ON

·2.00 ·2.60 ·2.40 4.80 4.67 ·5.30

·15.25 ·24.50 ·26.60 ·28.00 ·26.00 ·29.73 ·33.28 ·33.60 ·37.80 ·37.80 ·38.80 47.79 ·52.00 ·55.00 ·59.50 ·58.00 ·57.30 ·57.40 -46.52

ISS ISS ISS

CGS CGS CGS CGS BCI CGS BCI BCI BCI

02 21922 12121924 15 61930

02:51 :30.0 49.00 08:47:30.0 -40.00 21:08:11.0. -44.00

20 1 1940 09:58:06.0 1411 1944 00:30:24.0 17 9 1949 22:46:25.0 11 11 1949 10:57:36.0 16 61954 15:53:24.0 26 71958 08:35:\0.0 30 71958 15:10:12.0 17 81959 06:28:12.0 17 10 1959 01:23:00.0

ISS 03 41963 CGS 27 I 1964 CGS 09 10 1964 ISC 09 21968 ISC 26 71969 ISC 02 121975 GS 23 71976 ISC 31 71976 ISC 15 21975 ISS 26 3 1922 ISS 25 91924 ISS 04 11 1924 ISS 19 21920 ISS 20 71920

CGS 13 1 1938

14:47:58.0 10:32:13.4 00:14:22.3 08:36:24.0 05:21:40.0 05:23:09.7 17:44:26.8 05:11:20.7 22:00:58.5 13:25:32.0 04:0\:15.0 03:02:45.0 19:54:00.0 00:21:35.0 03:15:\8.0

·52.50 ·33.00 ·35.00

·9.00 ·37.50 -60.50 48.00 ·38.00 ·54.00 ·52.10 ·19.90 ·35.00 ·20.00 ·27.00 49.40 ·29.97 ·38.70 ·3.50

41.00 48.50 49.00 48.50 ·50.00 ·31.00

Lon,. DepIh oE (km)

·107.00 ·109.40 ·116.30 ·112.01 ·111.86 ·116.00 ·116.19 ·123.00 ·122.00 ·121.00 ·126.00 ·116.94 ·115.14 ·126.70 ·138.90 ·148.80 ·146.00 ·119.70 ·128.00 ·152.00 ·160.00 ·160.00 ·161.00 ·161.10 ·153.37

·132.00 ·120.00 ·118.00 ·134.50 ·115.00 ·154.00 ·119.00 ·114.00 ·168.50 ·120.00 ·115.00 ·165.00 ·131.50 ·115.40 ·115.00 ·114.00 ·120.00 ·120.10 ·114.43 ·123.00 ·108.80 ·135.00 ·160.50 ·132.00 ·160.50 ·127.00 ·155.00

o o o

10 10 10 10

o

33 33 33 33 o

10 o

33 o o

60 o

39 10

o o o o o o o o o o o o o

33 33 33 33 33 33 34 33 o o o

Ma,. Lal. "N

Relocation

Lon,. DepIh oE (km)

Time (GMT)

Intraplate Eve1l1s

4.8 mb 5.9 mb 5.1 mb 5.3 mb 5.1 mb 6.75PAS 5.9 mb

4.3 mb 4.8 mb 4.9 mb 5.0 mb 4.6 mb

5.1 mb 5.5 M. 6.0 PAS 7.2 PAS 6.4 M. 5.5 MI 5.4 mb 5.0 mb

·2.54 ·2.57 ·2.40

·24.27 ·26.55 ·26.92 ·26.00 ·29.73

·33.60

·37.80 ·40.67

·52.94 ·53.98 ·58.76 ·58.70 ·57.36 ·57.40 ·46.52

Casualties ·41.88 ·37.31 ·47.03

6.75PAS

4.2 mb 4.5 mb 4.1 mb 4.5 mb

5.1 mb 4.3 mb 4.4 mb

·54.23 ·30.64 ·64.44 ·8.61

·36.70 ·63.38 ·49.83 ·34.51 ·64.07 ·55.08 ·19.90 ·35.00 ·20.00 ·23.14 ·49.16 ·29.70 ·22.53

·3.43

·63.32

·26.65

·108.25 ·109.53 ·116.30

·122.77 ·121.95 ·121.42 ·126.00 ·116.94

·126.70

·148.80 ·142.99

·126.57 ·147.85 ·159.20 ·159.11 ·161.00 ·161.10 ·153.37

·134.19 ·120.85 ·116.24 ·135.78 ·113.20 ·164.30 ·109.03 ·112.51 ·163.75 ·114.35 ·109.38 ·169.40 ·128.28 ·115.40 ·113.80 ·114.00 ·117.25 ·120.11 ·114.20 ·113.45 ·108.77

·126.26

·170.22

10 10 o

10 10 10 33 33

33

10 10

10 10 10 10 10 39 10

10 10 10 10 10 10 \0 \0 10 10 10 10 10 33 33 33 10 10 33 10

10

10

10

06:18:21.6 02:16:35.2 11:11:09.8

03:24:05.0 18:15:00.0 19:19:59.8 05:44:14.2 08:29:44.6

16:16:51.8

19:07:20.2 11:54:18.4

06:21:19.7 14:42:36.6 19:20:26.3 06:06:43.7 08:35:02.5 03:56:28.1 13:45:46.9

02:51:31.8 08:47:39.8 21:08:08.8 09:58:13.0 00:30:39.6 22:46:25.5 10:57:40.2 15:53:34.7 08:35:44.2 15:10:24.1 06:28:57.8 01 :23:23.3 14:47:54.7

00:14:22.2

05:22:04.2 05:23:09.2 17:44:29.0 05:12:58.2 22:00:57.3

03:02:57.1

03:14:43.7

NI

14 60

19 12 13

16 14

13

11 6

\0 7

10

34

10

6 32

8

a (.)

1.66 1.87

1.48 0.50 0.70

4.44

1.32 1.78 1.79 0.97 1.28

1.95 1.41 2.86 1.10 0.56 2.60 1.75 1.18 0.99 1.38 0.52 0.34 3.15

1.81

0.93 1.28 2.58 0.43 1.66

0.00

0.53

281

Code Site Index

2 2

10 9

10

11 11 11 12 12 12

SP·1 SP·1 SP·2 !P·1 !P·1 !P·2 SP·5 !P·4 !P·5 !P-6 SP-6 SP·7 SP·8 !P·7 !P·9 SP·9 SP·IO !P·IO !P·11 !P·12 !P·13 !P·13 !P·15 !P·15 SP·11

!P·3

!P·16

!P·14

!P·8

!P·19

401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425

426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446

447 448 449 450 451 452

282

See Da'" DMY

!SC 21 41974 WEL 26 3 1982 !SC 26 61971 !SC 11 3 1984 !SC 10 21980 !SC 03 61983

WEL 01 81953 !SC 09 41965 !SC 26 7 1975

WEL 25 08 1964 WEL 04 41971 WEL 11 01 1974 WEL 15 21974 WEL 22 31974 WEL 26 4 1976 WEL 09 12 1976 WEL 25 12 1977 WEL 20 7 1978 WEL 28 21979 WEL 2 51979 WEL 25 41980 WEL 03 61982 WEL 02 21983 WEL 08 10 1985 CGS 24 91960 WEL 14 71964 !SC 27 7 1971 !SC 07 41972 !SC 24 41972 GS 01 11 1974 !SC 13 5 1982 !SC 02 41984 !SC 11 4 1984 !SC 13 41984 ISC 23 91984

WEL 31 51985 !SC 28 3 1976 ISS ISS

CGS ISS

ISS ISS ISS ISS ISS ISS CGS ISS BC! CGS !SC !SC

!SC LAO ISC !SC ISC !SC ISC PDE WEL PDE ISC ISC ISS

10 5 1924 10 8 1926 04 1 1940 01 8 1925

15 41921 26 12 1924 17 21925 04 1 1926 23 5 1936 09 7 1936 02 10 1940 01 21947 21 41959 24 8 1963 07 21964 01 10 1964

17 51966 10 1 1968 10 91974 27 7 1977 07 11 1979 27 5 1982 26 71982 30 3 1984 04 3 1985 22 5 1985 23 11 1985 28 2 1986 09 5 1920

Michael E. Wysession et af.

Table 3

Sor/ed Da/ase/ Jor Campbell - New aaland - Kermadec

Initial Location Relocation

Time (GMT)

06:54:27.9 19:45:30.5 22:02:18.2 16:42:38.0 22:59:55.1 05:44:37.6 13:36:28.0 13:18:57.0 11:36:55.5 02:37:12.0 08:56:11.2 03:07:22.1 07:15:01.9 22:16:20.5 16:06:08.2 03:54:40.0 04:35:54.8 14:56:42.2 23:36:34.0 08:32:03.8 09:20:12.1 08:32:46.9 15:38:59.8 06:23:06.1 11:06:39.2 01:11:02.0 14:03:05.7 00:53:07.9 09:29:17.1 20:44:31.2 18:04:13.1 18:45:08.6 11:10:14.3 09:33:54.6 03:57:53.2 08:16:26.5 06:48:26.2 02:48:40.0 21:16:20.0 01:10:18.0 02:25:40.0

21:06:10.0 23:32:43.0 14:13:18.0 04:01:36.0 15:32:07.0 10:20:45.0 10:24:00.0 13:28:56.0 15:24:42.0 12:50:13.0 14:45:05.5 20:38:06.0

04:27:24.0 00:37:45.0 15:44:23.0 15:26:45.7 01:00:01.0 10:12:17.2 10:20:18.0 16:38:38.2 18:44:04.0 06:55:58.2 02:44:16.0 14:05:54.0 08:00:04.0

Lat. "N

-55.28 -SO. 10 -51.35 -48.80 -49.00 -SO.20 -44.25 -43.60 -44.90 -42.10 -44.00 -42.13 -43.11 -43.28 -42.51 -43.42 -42.05 -42.74 -43.85 -44.56 -42.82 -41.99 -42.84 -42.25 -41.80 -40.40 -41.00 -'\0.40 -40.60 -39.06 -40.90 -40.10 -40.70 -40.10 -40.10 -40.10 -41.40 -36.30 -28.00 -34.00 -28.00

-33.30 -32.00 -51.70 -48.00 -41.00 -40.70 -30.50 -29.00 -62.50 -37.30 -32.10 -35.50

-40.66 -35.0 -38.70 -52.16 -38.1 -47.30 -41.76 -30.41 -40.50 -38.90 -41.90 -41.1 -51.70

lai,. Depch OE (km)

164.30 172.80 167.60 174.30 173.10 174.50

-179.00 -176.00 -175.00 178.40 177.97 178.10 176.90 177.16 177.82 179.04 178.27 177.40 176.62 177.10 176.87 178.03 177.12 178.10

-179.30 179.70

-179.30 -179.90 -179.40 -179.78 180.00

-179.80 180.00

-179.90 -179.30 -179.20 -171.40 -169.00 -163.50 -162.00 -163.50

-173.70 -173.00 173.80 170.00

-179.50 -179.40 -172.00 -162.00 -172.50 -178.20 -174.00 -170.50

179.70 -174.0 -176.50 165.60

-177.3 -178.70 179.05

-174.76 -178.70 -178.35 178.60 179.80 173.80

33 33 12 33 33 10

33 69

12 33 33 33 33 33 33 12 33 33 33 33 33 12 43 o

33 33 33 26 12 33 33 33 33 33 33 o o o o

o o o o

223 o o o o

36 33 o

33 33 33 33 33 33 33 .33 33 33 33 33 o

Ma,. Lot ON

lai,. Depch OE (km)

In/rapla/e Even/s

'5.1 mb 4.4 mb 5.6 MI 4.0WEL 4.1 mb 5.2 mb

4.5 WEL 4.6 mb 4.0WEL 3.8 WEL 4.1 WEL 4.0WEL 4.0WEL 3.9WEL 4.3 WEL 4.5WEL 4.1 WEL 4.2 WEL 4.1 WEL 4.0WEL 4.0WEL 4.3 WEL 3.9 WEL

4.5 WEL 4.5 mb 4.7 mb 4.4 mb 5.1 mb 3.7 mb

3.9WEL 3.7WEL 3.8 WEL 4.1 mb

5.9M.

-55.28 -51.19 -51.17 -49.00 -49.00

-44.24 -43.25 -44.90 -42.10 -44.00 -42.13 -43.11 -43.28 -42.51 -43.42 -42.05 -42.74 -43.85 -44.56 -42.82 -41.99 -42.84

-40.93 -40.40 -41.00 -40.40 -40.60 -39.06 -40.90

-40.08 -38.63 -35.21 -41.99 -33.84 -29.38

Casuallies

4.1 WEL 4.9 LAO 3.9 WEL 4.9 WEL 3.8 WEL 4.1 mb 4.2 WEL 4.7 mb 4.1 mb 4.7 mb 4.0WEL 3.7 WEL

-34.88 -29.66 -49.83 -45.83 -40.46 -39.75 -25.12 -19.90 -62.19 -37.47 -32.73 -32.92

-40.47 -18.99 -39.49 -51.69 -38.97 -49.77 -41.43 -30.57 -39.98 -37.95 -41.66 -40.42

164.30 168.30 167.69 172.63 173.10

-175.45 -177.60 -175.00 178.40 177.97 178.10 176.90 177.16 177.82 179.04 178.27 177.40 176.62 177.10 176.87 178.03 177.12

-178.72 179.70

-179.30 -179.90 -179.40 -179.78 180.00

-179.53 -170.67 -169.56 -161.25 -162.47 -164.34

179.96 -174.22 170.04 173.29 178.36 179.45 179.25

-174.23 164.48

-178.66 -179.86 -178.76

178.89 -167.31 178.02 164.25 179.30 164.70 178.25

-176.70 179.60

-179.80 178.14 178.59

33 10 11 10 33

10 10 69 o

12 33 33 33 33 33 33 12 33 33 33 33 33

10 o

33 33 33 26 12

10 10 10 10 10 10

10 10 10 10

249 231

10 10 10 10 10 10

33 10 33 33 33 60 33 33 10 33 33 33

Time (GMT)

06:54:28.3 19:45:29.5 22:02:20.0 16:42:47.6 22:59:55.1

13:36:28.1 13:19:08.2 11:36:55.5 02:37:12.0 08:56:11.2 03:07:22.1 07:15:01.9 22:16:20.5 16:06:08.2 03:54:40.0 04:35:54.8 14:56:42.2 23:36:34.0 08:32:03.8 09:20:12.1 08:32:46.9 15:38:59.8

11:06:33.9 01:11:02.0 14:03:05.7 00:53:07.9 09:29:17.1 20:44:31.2 18:04:13.1

08:16:27.2 06:48:25.5 02:48:47.8 21:14:57.2 01:10:21.4 02:26:03.8

21:07:20.3 23:32:34.1 14:14:29.3 04:01:45.0 15:32:14.1 10:20:42.8 10:23:56.3 13:28:56.4 15:25:02.3 12:SO:15.4 14:45:50.6 20:39:00.3

04:27:33.0 00:38:59.6 15:45:28.3 14:07:43.9 01:00:41.8 10:13:34.3 10:20:27.7 16:38:53.0 18:44:28.7 06:56:10.0 02:44:23.2 14:06:07.8

NI

15 5

40

11 9

6 25 15 13 15 23 15 20 16 25 21 18 20

16 12

4 5 9

10 10 14 4 6 6 6

11 4

11

29 26

4 26

11

(J

(I)

1.93 1.75 1.67 3.34

2.49 1.78

0.51 1.48 1.61 2.31 2.16 2.11 1.91 1.29 1.12 2.09 1.93 1.66 1.30 0.94

1.56 1.37

1.78 0.24 1.61 9.31 1.30 0.53

1.19 2.42

19.63 1.96 2.64 1.12 0.71 0.81 0.98 0.47 1.14 0.71

1.92 0.34 2.47

1.83 3.46 1.54 1.84 1.30 1.33 0.81 1.58

PAGEOPH,

Code Site Index

1 2 2

2 2 2 2 2

7 7

9 7 9

11

!P-17

!P-20

501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541

542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566


Table 3 (continued)

Sorled DalaseI tor Campbell - New Zealand - Kermadec

Initial Location Relocation Code Sitc Inclc:.

See Dale Time LaI. Looa· Depth Maa· Lot. Loo .. Depth Tune N. Cf DMY (GMT) "N OE (Ian) oN "E (Ian) (GMT) (s)

ISS 04 71923 22:54:55.0 ·28.00 ·163.50 0 ·30.79 ·160.55 10 22:55:01.9 0.00 11 567 ISS 25 21927 11:25:12.0 ·28.00 ·163.50 0 ~64 -163.46 10 11:24:45.8 0.00 11 568 ISS 01 61927 10:09:36.0 -45.00 180.00 0 -45.07 178.05 10 10:09:55.2 0.00 11 569

WEL 06 81964 08:34:00.0 -38.50 -176.00 33 29WEL 11 570 WEL 25 51984 01:04:42.0 -41.19 179.87 12 3.8WEL -41.19 179.80 10 01:04:42.1 0.18 11 571 ISS 06 81924 00:22:00.0 -28.00 -163.50 0 -26.SO -163.66 10 00:22:31.0 0.47 12 572 ISS 17121925 05:41:45.0 -48.00 170.00 0 -45.77 169.05 10 05:42:13.2 0.00 12 573 !SC 29 11 1974 22:14:43.6 -42.60 180.00 12 3.8 mb 12 574

WEL 30 61985 11:19:53.4 -49.10 174.60 33 4.3 mb 12 575 ISS 15101957 05:55:33.0 -30.24 -170.94 244 -30.24 -179.04 14 576

Table4

Sorled DalaseI tor Off Tonga and Samoa

Initial Location Relocation Code Site Inclc:.

See Dale Tone Lot. Lona· Depth Maa· LaL Looa· Depth Tune Ns Cf

DMY (GMT) "N OE (Ian) "N OE (Ian) (GMT) (s)

Inlraplale EVenlS

ISS 24121929 04:29:18.0 -3.00 -172.00 0 -2.00 -171.15 10 04:29:10.0 5 1.90 651 ISS 15 61930 07:33:03.0 -13.00 -162.80 0 -1268 -162.90 10 07:33:01.0 5 2.66 652 ISS 16 61932 23:13:47.0 -13.00 -167.00 0 -13.19 -166.69 10 23:13:46.4 6 0.88 653 ISS 20 21922 07:43:50.0 -17.00 -168.00 9 -1S.31 -166.43 10 07:43:45.9 5 1.36 1 654 ISC 04 7 1983 17:10:30.7 -2D.SO -166.80 33 -2O.SO -166.80 33 17:10:30.7 2 655 CGS 04 10 1937 07:40:30.0 -26.80 -163.10 0 6.2Ma -22.01 -166.99 10 07:40:34.8 1.20 1 !P-18 656

Casualties

ISS 20121914 14:08:27.0 -17.00 -168.00 0 -1S.72 -173.97 10 14:09:00.5 12 1.77 7 657 ISS 23 1 19%5 17:01:00.0 -11.00 -170.00 0 -8.60 -164.05 10 17:01:04.9 22.19 7 658 ISS 22 21928 12:57:20.0 -23.00 -163.00 0 -23.94 -174.20 10 12:59:28.2 4.13 659 ISS 14 21933 05:22:50.0 -24.60 -170.30 0 -2D.18 -174.88 571 05:23:53.3 0.44 7 660 CGS 29 61938 18:44:54.0 -16.00 -168.00 0 -21.00 -174.00 0 18:44:08.0 9 661 CGS 15 51939 06:09:00.0 -13.00 -168.00 0 -15.43 -174.55 235 06:09:35.6 9 0.78 7 662 CGS 31 31954 08:30:04.0 -1200 -171.50 0 -15.12 -174.79 10 08:29:44.7 11 0.41 7 663 ISC 23 5 1967 12:03:14.0 -8.00 -170.00 33 4.4 mb -1S.67 -173.36 10 12:02:30.3 8 0.27 7 664 !SC 12 11 1967 15:28:24.0 -25.60 -169.90 33 4.3 mb -26.76 178.96 600 15:28:53.0 6 2.30 7 66S !SC 07 21974 20:10:50.7 -19.10 -167.30 33 -23.70 -179.29 10 20:12:21.5 5 1.13 666 GS 28 51974 16:58:04.1 -24.48 -173.29 33 -26.12 -175.00 191 16:58:34.2 7 1.83 667 !SC 01 71974 06:21:49.7 -13.40 -167.40 33 -15.81 -169.56 10 06:22:09.8 8 0.71 668 !SC 15 8 1974 07:1S:10.4 -14.80 -167.30 0 3.9 mb -17.40 -174.52 199 07:15:13.0 7 1.14 669 ISC 17 3 1975 08:41:06.6 -17.00 -168.20 33 -17.06 -168.16 10 08:41:05.2 5 1.30 7 670 GS 05 3 1976 08:20:09.3 -21.78 -171.55 33 5.3 mb -21.56 -173.05 10 08:20:19.5 8 1.28 7 671 ISC 23 91976 09:39:04.7 -11.70 -172.00 33 -16.74 -175.48 362 09:40:19.5 8 0.30 672 ISC 23101976 21:31:05.4 -16.80 -168.90 33 -14.59 -170.20 10 21:30:55.0 11 0.53 673 !SC 14 41979 09:21:31.5 -l1.SO -166.10 33 -17.73 -175.25 5SO 09:23:37.5 7 0.93 674 !SC 09 11 1979 14:26:27.8 -11.00 -162.60 33 3.9 mb -16.35 -172.82 100 14:28:01.3 9 1.41 675 !SC 01 3 1980 07:1S:16.8 -21.90 -166.20 33 4.9 mb -20.26 -177.07 502 07:17:18.3 6 0.15 676 !SC 06 61980 08:57:58.4 -19.40 -167.60 33 4.2 mb -2240 -174.17 200 08:59:11.0 5 0.33 677 ISC 25101980 22:21:46.8 -13.60 -165.60 33 5.4 mb -17.60 -177.70 552 22:24:08.7 8 0.70 7 678 PDE 24 5 1984 01:22:10.5 -1S.83 -167.27 33 5.2 mb -19.80 -177.67 448 01:24:11.6 10 1.09 9 679 PDE 30 11 1984 06:42:40.1 . -26.27 -173.36 33 4.7 mb -26.33 -173.68 10 06:42:42.0 5 0.66 7 680 GS 26 91986 10:36:17.0 -21.84 -171.65 33 5.1 mb -2260 -176.20 162 10:37:04.0 11 0.72 9 681 ISS 18 51917 19:05:00.0 -11.00 -170.00 0 11 682 ISS 26 31927 02:26:40.0 -25.00 -167.00 0 -24.12 -167.59 10 02:26:58.0 0.00 12 683

WEL 11 11965 22:13:28.1 -18.10 -166.90 33 12 684 NGD 30 61905 17:07:00.0 -1.00 -168.00 60 7.6 13 68S BCl 17 71955 07:06:03.0 -17.SO -170.00 0 -17.35 169.21 10 07:06:04.1 2.33 14 686 BCl 28101958 19:07:48.0 -13.SO -167.00 0 -13.37 166.76 10 19:07:46.0 0.14 14 687 CGS 01 71961 03:48:36.9 -13.90 -166.10 33 -14.00 165.70 33 03:48:33.1 14 688

See

G CGS BC! ISS ISS

CGS ISC ISS W.E WIA WIA WIA NGD ISC ISC

ISS ISS ISS ISS ISS ISS ISS ISS ISS ISS ISS

CGS ISS ISS CGS CGS CGS CGS CGS BC! BC! BC! ISC ISC !SC ISC ISC ISC !SC !SC ISC ISC ISC ISC ISC ISC ISC PDE ISS ISS ISS ISS ISS ISS ISS ISS ISS

284

Da .. DMY

24 41937 26 61956 03 10 1957 30 41939 15 5 1931 05 3 1964 03 61981 10 21921 19 91967 28 5 1983 29 6 1983 19 91983 1911 1986 17 2 1982 02 51982

06 2 1918 01 8 1918 20 1 1922 01 121924 1911 1925 05 21928 08 61928 02 21932 23 1 1933 11 9 1933 12 9 1935 23 7 1936 02 4 1937 10 1 1939 19 1 1940 25 41940 12 91940 18 7 1949 10 12 1954 2S 1 1959 14 2 1959 04 1 1960 10 1 1968 19 71969 11 10 1969 29 10 1969 01 1 1970 02 3 1972 20 4 1972 08 4 1973 15 4 1973 30 7 1973 02 5 1974 10 61974 18 9 1974 10 11 1974 31 121974 27 2 1985

06 21918 20 9 1921 22 61926 23 3 1927 04 5 1921 01 61926 16121926 28 11 1927 21 10 1928

Michael E. Wysession et al.

Table 5

Sorted Da/aset for Samoa to Gi/ben


Time (GMT)

04:58:35.0 11:23:09.0 13:44:30.0 14:02:34.0 07:41:55.0 05:50:15.2 22:15:14.1 19:42:16.0 06:26:10.9 23:21:26.4 16:15:01.5 10:12:48.0 08:28:33.0 14:23:43.9 23:13:57.0

03:10:30.0 11:42:00.0 06:50:54.0 22:56:48.0 19:10:35.0 22:45:20.0 14:39:10.0 06:59:30.0 18:14:05.0 11:20:09.0 16:01:20.0 06:20:48.0 05:30:10.0 11:07:30.0 13:53:06.0 10:18:42.0 09:21:30.0 08:27:21.0 00:31:40.0 05:14:38.0 07:29:59.0 04:06:18.0 17:46:54.0 20:51:14.0 11:02:42.0 07:02:05.0 07:25:25.2 02:17:29.6 12:11:30.4 03:59:56.8 12:05:05.7 00:30:33.3 20:39:46.8 03:35:13.4 18:30:04.1 21:38:11.1 18:26:59.9 06:46:48.3

14:43:42.0 18:51:30.0 04:51:30.0 07:37:25.0 21:12:25.0 22:17:36.0 00:24:08.0 10:12:08.0 15:17:06.0

Lot "N

·12.00 ·10.00 -10.00 -5.60 -8.00 -6.10 -5.30 -3.00 -0.50 -1.00 0.00 0.50 0.30

-6.10 -1.10

-11.00 -11.00 -6.50

-12.00 -10_00 -10.00 -12.00 -11.00 -5.80

-10.80 -8.60

-10.00 -8.00

-12.00 -11.00

-8.50 -12.00 -11.50 -12.00 -3.75 -5.50 -8.00

-10.40 -8.30

-12.00 -9.00 -9.90 -7.50

-11.20 -11.10 -11.70 -11.70 -8.70

-11.70 -11.40 -11.40 -11.50 -6.79

-11.00 -11.00 -12.00 -12.00 -11.00 -10.00 -10.50 -3.00 -7.00

Lona· Deph OE (Ian)

-178.00 -173.50 -179.00 163.80 173.00

-172.50 -175.20 177.50 [email protected] 170.50 170.00 1@50 [email protected] 179.00 175.70

-176.00 -176.00 166.50

-177.00 176.00 176.00

-177.00 -176.00 175.00 178.50 179.80

-173.00 180.00

-176.00 -173.50 -176.50 -176.50 171.00

-172.50 158.75 162.50 170.00 179.80

-174_60 -178.50 -177.00 178.30 175.10

-177.60 -179.70 -177.70 -178.60 178.10

-177.70 -177.30 177.:0

-177.20 163.24

-176.00 -176.00 -177.00 ·177.00 -176.00 176.00

-17450 177.50 178.00

200 o o o o

33 33 o o o o o

33 33 33

o o o o o o

o 128

o o o o o

200 o o o o o

33 33 33 33 33 o o o o o o o

33 33. o

33 o o o o o o o o o

Moa· Lot "N

1.00,. Deph "E (Ian)

Intraplate Events 6.25PAS -10.14 -176.00

-10.26 -173.65 -10.58 -179.49

-6.36 163.22 -8.35 172.82

4.3 mb -5_77 -173.82 5.3 mb

5.3 mb 4.2 mb 4.4 mb

4.4 mb 4.3 mb 4.7 mb 3.8 mb 4.2 mb 4.3 mb 4.3 mb 4.1 mb 4.3 mb 4.2 mb 5.0 mb 4.7 mb

4.8 mb 4.4 mb 4.8 mb

-{).79 177.90

-3.61 177.81 -3.22 177.24

Casualties -12.48 -12.36

-5.54 -10.34 -15.61 -12.32 -14.82 -13.86 -12.12 -14.17 -16.99 -20.20 -21.22 -12.64 -16.49 -17.51 -16.63 -13_00 -14.56

-8.75 -10.70 -16.83 -15.41 -8.16

-15.92 -14.23 -16.64 -23.28 -15.92 -15.88 -15.11 -14.85 -17.@

-14.72 -16.01 -17.39 -16.28

-8.00

-8.13 -9.50 -6.76 -7.83

-174.80 -175.31 164.95

-174.72 -179.65 177.93

-175.60 -176.30 180.82

-179.54 180.14

-174.20 -178.40 -176.13 -173.00 -175.72 -176.24 171.00

-174.17 1S8.13 162.78

-178.60 -178.27 -174.63 -176.41 -174.01 -179.09 -176.43 -174.99 -177.27 -175.@ -176.73 -178.75 -175.97 -177.85 -178.75 -174.85 162.2

[email protected] -174.11 179.80 179.40

10 10 10 10 22 10

10

10 10

10 10 10 10 10 10 10

162 432 284 484

o 494

10 108 398 220

o 24 10 10 10

353 10

281 10

475 10

260 358 225 272 512 251 443 482 241 130

10 10 10 10

Tune (GMT)

04:58:30.0 11:23:14.2 13:44:32.8 14:02:35.9 07:41:57.2 05:50:22.0

19:42:14.5

14:23:49.7 23:13:51.7

03:10:57.5 11:42:21.4 06:51:19.4 22:56:42.4 19:11:49.2 22:45:55.0 14:39:10.6 06:59:40.0 18:14:58.3 11:20:40.0 16:01:31.1 06:20:13.0 05:30:20.7 11:07:35.7 13:53:06.9 10:18:57.3 09:21:32.8 08:28:20.0 00:31:32.7 05:14:43.9 07:30:02.1 04:05:45.8 17:47:25.1 20:51:13.7 11:03:04.1 07:01:52.2 07:26:03.5 02:17:04.4 12:12:01.3 04:00:31.6 12:05:33.1 00:31:02.7 20:40:33.7 03:35:42.0 18:30:35.2 21:38:57.4 18:27:27.4 06:47:13.0

22:18:31.6 00:24:005 10:13:09.2 1S:17:39.5

NI

21 5

10 11

4

4 7 7 9

10 18

13

16 7 6 8

6 5

10 8 6 5 7 9 5

11 5

" (I)

1.64 1.24 1.94 0.77 LS3 1.48

1.18

1.37 0.43

3.64 0.27 6.67 3.78

13.13 3.35 1.39 1.02 3.03 2.91 1.25

2.41 2.91 0.12 0_33 0.31

0.67 1.62 1.28 1.08 0.66 0.46 0.04 0.48 0.41 0.60 0.96 1.06 0.72 0.30 0.13 0.29 0.25 0.52 2.25 2.33

6.59 2_05 0.59 2.35

PAGEOPH,

Code Si.. Index

6 6

7 7 9 7 7

7

7 7 7 9

11 11 11 11 12 12 12 12 12

1P-21 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765

766 767 768 7@ 770 771 772 773 774 775 776 m 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 80S 806 807 808 809 810 811 812


Table 6 Sor/ed DflllJHt tor Microfll!sia anti Marshallislands

Initial Location Relocation Code Sire Index

See 0. ... lune l.aL Lona· Deph Mac· l..oL Lona· DepIh 1_ N. G DMY (GMT) "N DE (km) "N DE (km) (GMT) (.)

IlIIrap/ate Events ISS 29 91926 05:16:30.0 9.00 1S5.oo 0 5.84 -170.43 10 05:1S:01.7 4.71 1 876

NGD 03 51969 08:22:14.8 8.30 -175.60 33 5.2 mb 3 877 ISC 08 51976 23:11:35.4 13.50 -176.20 33 13.50 -176.20 33 23:11:35.4 2 878

NGD 27 1 1965 01:44:35.8 18.90 176.60 33 4.4 mb 18.90 176.60 33 01:44:35.8 2 879 WIA 26 81983 01:17:34.0 22.50 171.00 0 4 880 WIA 02 11 1984 17:17:205 26.00 162.00 0 4 881 !SC 16 11976 19:08:39.8 26.60 163.00 33 26.56 162.92 10 19:08:38.9 1.41 882 !SC 16 31976 08:05:50.4 28.90 157.00 33 28.90 157.00 33 08:05:50.4 2 883 WIA 08 7 1983 14:39:33.6 7.50 176.00 0 4 884 NGD 22 3 1982 06:04:10.7 6.60 175.10 33 5.6 mb 3 885 OBS 27 81981 04:04:08.0 9.00 167.00 0 4 886 OBS 13 81981 18:49:19.7 lQ.00 166.00 0 4 887 W,E 21 81965 18:44:34.5 15.00 167.50 0 888 W,E 23 51969 01:13:35.0 14.00 166.00 0 889 W,E 12121963 10:03:41.6 18.00 163.00 0 4 890 W,E 18121965 02:44:11.9 16.00 159.00 0 891 WIA 24 3 1983 00:24:08.2 19.00 154.00 0 4 892 WIA 04 71985 08:18:33.3 19.50 16050 0 4 893 W,E 17 7 1963 19:28:46.8 21.00 159.00 0 4 894 W,E 20 81963 23:38:25.2 21.50 157.00 0 4 895 W,E 15 81963 17:40:22.6 23.50 154.00 0 4 896 BCI 21 71960 09:51:36.0 25.00 147.00 0 25.30 148.34 10 09:51:43.4 6 1.25 897 WIA \0 1 1985 10:13:01.4 5.50 162.50 0 4 898 WIA 05 11 1982 16:26:07.9 5.50 162.50 0 4 899 !SC 05 11 1975 01:56:38.0 6.50 159.80 33 6.50 159.80 33 01:56:38.0 2 900 !SC 02 61977 17:49:16.3 6.50 160.50 33 6.50 160.50 33 17:49:16.3 2 901

NGD 18 41981 03:00:25.6 6.90 159.80 25 5.0 mb 3 902 E 26 21967 06:10:54.6 lQ.00 159.00 0 4 903

W,I! 27 3 1963 13:17:53.1 9.00 157.00 0 4 904 ISS 05 5 1924 15:56:45.0 12.50 158.00 0 11.30 156.8 10 15:57:15.8 0.16 1 90S W,E 28 1 1965 07:36:20.2 11.50 156.00 0 4 906

E 01 41964 19:56:57.9 7.00 157.00 0 4 907 NGD 16 51925 10:27:50.0 9.00 155.00 0 8.53 154.47 10 10:28:07.7 6 243 908 W,I! 21 71969 04:58:55.8 1200 151.00 0 4 909 WIA 21 9 1984 21:20:29.7 4.00 157.50 0 4 910 ISC 08 71969 04:09:03.1 5.50 154.70 33 5.50 154.70 33 04:09:03.1 2 911 E 07 12 1964 05:34:31.7 6.00 154.00 0 4 912

W,E 30 31964 08:47:34.4 7.00 151.00 0 4 913 NGD 12 31974 15:11:05.5 8.80 151.00 33 5.5 mb 8.80 151.00 33 15:11:05.5 2 914 !SC 14 31974 23:22:128 8.70 151.10 0 4.8 8.70 151.10 0 23:22:12.8 2 915 !SC 10 8 1985 17:02:11.9 8.77 148.20 33 5.0 mb 3 916 !SC 22 31966 07:52:51.0 10.13 147.90 68 4.7 mb 3 917 !SC 20 11 1964 00:56:21.6 0.60 156.30 0 4.2 mb 0.60 156.30 0 00:56:21.6 2 918 ISC 09 11 1970 02:40:06.0 280 150.40 33 2.80 150.40 33 02:40:06.0 2 919

Casualties NGD 30 11 1918 01:33:30.0 2200 151.00 0 22.08 151.47 10 01:33:37.1 5 13.90 7 920 ISS 19 91923 08:22:30.0 1250 168.00 0 22.42 143.81 10 08:26:42.6 4 3.33 7 921 ISS 18 11 1926 16:31:18.0 29.70 147.00 0 29.51 150.29 10 16:31:14.3 5 0.47 7 922 ISS 24 11 1926 17:40:54.0 29.70 147.00 0 28.91 144.05 10 17:41:10.3 4 2.10 7 923

NGD 18121926 09:38:45.0 2200 151.00 0 17.96 145.23 10 6 1.77 924 ISS 30 61928 22:21:45.0 25.00 150.00 0 48.48 147.38 10 22:24:00.3 10 204 925 ISS 06 8 1931 15:21:15.0 0.00 151.00 0 -1.62 151.65 10 15:21:04.7 5 1.54 926 ISS 09 21933 15:34:40.0 0.00 151.00 0 -3.55 148.70 10 15:34:21.2 7 1.41 7 927 ISS 21 5 1933 21:54:05.0 27.00 146.50 0 22.58 146.00 19 21:53:51.3 13 1.52 7 928

CGS 02 51939 06:15:120 0.00 150.00 0 54.03 -162.42 10 06:20:475 5 0.37 7 929 NGD 15 71951 06:05:220 21.00 15050 0 20.34 150.97 10 06:05:25.8 8 0.67 7 930 !SC 18 51974 17:39:38.9 1.70 153.70 33 -5.15 150.10 314 17:41:11.4 7 0.16 7 931 !SC 16 11976 15:34:05.5 27.60 158.10 33 45mb 44.03 149.60 10 15:35:34.3 12 0.78 7 932

!SC 20 21976 19:26:39.3 23.70 154.20 33 4.8 mb 43.07 147.35 10 19:28:31.1 9 0.34 7 933

ISS 14 3 1918 09:29:15.0 1.00 [4350 0 11 934 ISS 21 51918 11:15:10.0 11.70 176.00 0 11 935

ISS 29 91926 05:43:30.0 9,00 155.00 0 11 936 ISS 14 51923 07:02:320 16.00 15350 0 15.41 155.65 10 07:02:34.0 0.00 12 937 ISS 03 31928 17:15:10.0 9.00 155.00 0 927 157.22 10 17:15:06.9 4 3.44 12 938

ISS 24 81930 09:08:40.0 8.00 157.00 0 -8.00 157.00 14 939

ISS 28 41951 2t:19:41.0 5.60 150.50 0 -5.6 150.50 14 940 BCI 04 11 1953 08:47:19.0 1250 16550 60 -12.50 165.50 60 14 941 CGs 17101955 01:08:07.0 6.00 154.00 0 -553 154.17 10 01:08:12.8 16 1.22 14 942

Sce

ISS CGS SC! CGS CGS CGS CGS CGS CGS

See

ISC ISC ISC ISC ISC ISS GS

ISS ISS ISS ISS ISS ISS ISS ISS ISS ISS ISS ISS ISS ISS ISS

CGS ISC ISC ISC ISC !SC ISC ISC ISC ISC iSC ISC ISC ISC ISC ISC !SC !SC !SC ISS

NGD

286

Date DMY

14 7 1957 15101960 OB 71956 12 5 1958 16 5 1958 14 61958 28 6195C 12 7 1958 26 71958

Date DMY

14 2 1975 14 2 1975 05 5 1976 06 5 1976 21 8 1985 11 11 1936 07 3 1988

06 21916 31 10 1916 31 5 1917 06 91924 28 1 1925 04 10 1925 28 12 1925 04 3 1928 20 8 1928 16 21929 25 71929 24 21931 02 8 1931 26 41932 30 8 1932 28 61936 19 12 1966 18 1 1967 29 1 1968 12 81969 18 8 1969 27 81969 04 21972 17 12 1972 01 5 1973 27 91974 10 6 1975 10 61975 11 61975 21 1 1976 07 21976 24 3 1978 25 3 1978 21 41978 05 81931 17 21905

Initial Location

Time (GMT)

02:27:03.0 12:12:06.7 18:06:01.0 18:29:58.0 01:30:00.0 1C:29:59.0 19:29:58.0 03:29:58.0 20:29:59.0

LaL 'N

14.11 7.40

11.75 12.00 12.SO 12.00 12.00 12.00 12.00

Initial Location

Tune (GMT)

OB:33:01.6 OB:35:33.6 20:02:32.6 14:42:25.8 06:49:10.7 00:41:57.0 15:21:06.8

10:51:13.0 15:30:33.0 06:50:22.0 02:36:55.0 18:15:06.0 07:28:42.0 21:56:10.0 20:58:12.0 01:56:03.0 09:13:45.0 15:07:24.0 14:07:50.0 23:29:32.0 13:31 :32.0 16:18:50.0 OB:IO:24.0 07:56:43.0 11:17:10.0 14:58:43.0 01:17:55.0 12:04:40.0 03:26:36.0 00:16:38.3 09:38:35.2 04:00:36.1 06:08:03.8 15:38:10.5 20:40:59.9 19:11:10.8 23:49:04.9 06:50:20.8 21:49:26.7 23:20:44.0 19:07:00.3 07:26:11.0 11:39:26.0

LaL 'N

34.40 34.SO 35.20 36.80 37.70 41.00 41.67

41.00 46.SO 32.00 46.SO 39.00 41.00 39.00 41.SO 39.SO 41.SO 42.00 42.00 4O.SO 41.30 39.SO 32.SO 33.40 40.00 40.20 38.30 41.70 40.60 36.80 39.20 3S.SO 39.80 39.SO 39.30 41.00 41.00 41.60 41.SO 41.90 40.80 36.00 33.00

Michael E. Wysession el af.

Table 6 (continued)

Sorted DaJaset for Micronesia and MarshallIslands

Relocation

Lona· Depth Moa· oE (km)

151.39 157.20 163.00 162.00 161.00 161.50 162.00 165.00 161.50

80 416

o o o o o o o

LaL 'N

46.11 -7.40 11.65

Table 7

Lenc. Depdt oE (km)

151.39 157.20 162.10 10

Sorted Dataset for East of Japan and Kuri/es

Time (GMT)

18:06:02.0

Relocation

Lona. Depth Moa· LaL 'N

Lenl· Depdt OE (km)

147.90 147.80 156.80 158.00 149.50 155.00 152.22

156.50 160.00 147.50 160.00 155.00 156.50 155.00 153.00 157.00 153.00 158.00 158.00 159.00 157.60 157.00 145.50 148.10 153.00 149.80 148.00 151.80 148.20 148.60 148.50 147.30 147.70 147.00 148.70 147.80 151.60 1SO.50 151.50 1SO.9O 151.20 149.50 152.00

o o

33 39 20 o

10

o o o o o o o o o o o o o o o o

33 33 30 o o

45 o o o o o

33 o

33 109 33 33 40 o

60

'E (km)

Intraplate Events 3.7 mb 34.40 147.90 4.4 mb 34.50 147.80

35.74 156.09 5.1 mb 36.80 158.00 4.2 mb 37.70 149.SO

41.04 154.72 5.9 mb

Casualties 47.51 49.67 33.44 42.35 53.51 42.39 35.14 44.77 47.44 41.64 45.51 44.61 35.65 47.38 47.86 30.60

4.4 mb 4.3 mb

34.18 51.24 40.30 43.84 43.07 42.80

4.5 mb 3.8 mb 3.7 mb 4.2 mb 4.6 mb 4.3 mb 3.6 mb 4.5 mb 4.3 mb 5.4 mb 4.7 mb 4.1 mb

7.3

26.04 51.02 36.50 43.78 44.00 43.67 43.68 43.92 44.00 44.42 44.19 44.42 34.49

150.81 157.90 143.37 148.04 153.90 155.64 147.18 151.31 149.77 152.80 154.26 151.82 157.74 153.10 150.01 141.80 143.38 156.66 149.68 147.77 148.21 145.92 145.55 154.11 146.70 146.69 148.13 147.56 146.88 149.95 149.57 149.41 150.75 148.54 146.95

o o

10 39 20 10

10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 o

385 60 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10

Time (GMT)

08:33:01.6 08:35:33.6 20:02:30.8 14:42:25.8 06:49:10.7 00:41:54.7

10:52:10.1 15:31:02.3 06:06:00.4 02:37:01.5 18:16:36.4 07:29:00.3 21:56:23.2 20:58:35.6 01:57:07.7 09:13:SO.8 15:07:56.8 14:08:24.4 23:29:41.1 13:31:35.1 16:19:52.4 08:10:23.0 07:57:20.9 11:18:11.9 14:58:43.3 01:18:29.0 12:05:05.2 03:26:52.7 00:15:SO.1 09:39:40.9 04:00:45.5 06:08:30.9 15:38:39.2 20:41:22.9 19:11:30.4 23:49:23.8 06:50028.0 21:49:43.2 23:21:00.3 19:07:24.9 07:26:22.7

N.

10

N.

14 20

11 6

11 9 7 8 5

4

5 5 9 8 6 8 8

23

" (.)

1.11

" (.)

1.74

1.06

1.64 3.31 0.27 1.60 0.09 1.21 0.19 0.19 1.08 2.09 1.50 2.38 1.32 1.80 0.67

2.71 0.38 0.14 0.54 0.48 0.55 1.48 0.24 2.74 2.33 0.85 1.20 2.11 0.89 0.70 2.02 0.20 0.21 2.74

PAGEOPH,

Code Site Index

14 14 15 15 15 15 15 15 15

943 944 945 946 947 948 949 950 951

Code Sil. Index

2 2 1 3

7 7 7 7 7 7

7

7 7 7 7 7 7 7 7 7 7 7

7

12 13

1001 1002 1003 1004 1005 1006 1007

1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043


Table 8

Sorted DfJlase/ for NOT/her" BolUIdary

Initial Location Re1ocalion Code Sile lDdex

Sec Dole Time LoI. Loaa· Dopb Moa· LoI. Loaa· Dopb Tune N. CJ DMY (GMT) "N "I! (km) "N "I! (km) (GMT) (.)

11llrapla/e Eve,,/s !SC 28 41968 04:18:15.5 44.80 174.C50 36 5.5 mb 1101 !SC 28 41968 06:23:02.0 44.90 174.70 33 4.4 mb 1102 !SC 18 11976 05:17:15.8 43.20 -179.90 33 43.11 178.89 10 05:17:14.1 2.79 1103 !SC 05 31985 14:01:02.6 45.70 178.80 33 52mb 1104 COS 21 11 1938 21:34:42.0 45.00 -178.00 0 46.68 ·164.96 10 21:35:57.0 7 2.04 1 1105 !SC 14 81980 10:35:58.3 45.20 -161.60 0 4.4 mb 45.09 -1112.40 10 10:35:58.2 4 0.25 1 1106 !SC 11 11 1975 11:53:57.9 51.20 -153.80 33 51.20 ·153.80 33 11:53:57.9 2 1107 GS TI 31986 22:42:36.6 53.78 -148.48 33 4.5 mb 53.78 -148.48 33 22:42:36.6 2 1108 !SC 04 21979 21:25:41.9 42.40 ·14920 10 5.1 mb 3 1109 !SC 14 11975 21:56:33.6 45.00 -147.60 33 45.00 ·147.60 33 21:56:33.6 2 1110 GS 04 111978 18:22:56.7 53.96 ·139.84 10 4.4 mb 53.96 -139.84 10 18:22:56.7 2 1111 ISS 20 31940 00:35:40.0 46.90 -134.80 0, 46.93 ·134.79 10 00:35:44.6 9 1.84 1112 ISS 20 31940 02:45:TI.0 46.90 -134.80 0 47.12 -134.06 10 02:45:35.0 14 2.21 1 1113 ISS 05 11 1936 20:46:20.0 40.00 -133.00 0 40.04 -132.46 10 20:46:19.3 8 1.44 1 1114 !SC 21 11979 05:46:13.6 32.90 -131.80 33 3.5 mb 32.90 -131.80 33 05:46:13.6 2 1115 COS 20 51949 22:35:31.0 33.00 -126.00 0 4.8 PAS 33.03 -125.83 10 22:35:30.8 21 1.48 1 1116 GS 15101982 10:58:39.1 32.84 -125.82 5 5.1 mb 3 1117 GS 27101982 13:33:44.0 32.68 -126.02 10 4.0 mb 3 1118 GS 21 11981 20:49:04.7 31.15 -123.16 15 4.3 mb 1119 GS 05 61988 20:55:18.9 31.26 -126.04 10 3126 -126.04 10 20:55:18.9 2 1120

Casual/ies ISS 29 41913 23:28:59.0 50.00 171.00 0 55.35 171.09 10 23:29:34.7 13 1.65 1121 ISS 23 11 1913 21:17:33.0 49.50 165.50 0 51.88 1S9.55 10 21:18:06.5 8 2.94 7 1122 ISS 06 10 1921 15:59:36.0 43.00 170.00 0 49.27 149.74 10 15:00:45.2 13 2.65 1123 ISS 02 21923 01:06:15.0 50.50 164.00 0 53.40 1112.19 10 01:06:40.9 20 1.46 7 1124 ISS 02 21923 05:07:15.0 50.50 164.00 0 7.4 PAS 53.38 161.26 10 05:07:42.4 93 6.68 7 1125 ISS 05 21923 22:23:10.0 50.50 164.00 0 53.50 162.37 10 22:23:32.0 8 2.13 7 1126 ISS 03 21926 19:30:16.0 50.00 171.00 0 49.20 175.40 10 19:30:04.0 4 0.09 7 1127 ISS TI 81927 12:13:15.0 40.50 160.50 0 47.42 15427 10 12:14:08.3 5 4.58 7 1128 ISS 24 91927 17:41:00.0 47.50 -169.00 0 49.14 -168.81 10 17:41:18.0 10 2.32 7 1129 ISS 28 1 1928 23:59:00.0 47.40 176.70 0 49.18 179.44 10 23:59:12.9 7 2.04 7 1130 ISS 21 41930 21:56:57.0 40.50 160.50 0 44.87 151.96 10 21:57:40.1 2.17 7 1131 ISS 09 61931 12:14:06.0 50.50 164.00 0 52.37 160.48 10 12:14:19.9 24 2.29 7 1132 ISS 31 11932 09:22:09.0 40.00 -140.00 0 40.55 -130.01 10 09:23:37.0 9 1.54 1133 COS 11 8 1932 09:42:06.0 49.00 -166.00 0 51.98 -169.41 10 09:41:47.0 12 1.82 1134 ISS 21 71944 12:24:35.0 42.00 -130.50 0 42.51 -131.06 10 12:24:29.0 17 1.61 1\35 COS 14 61949 05:49:00.0 52.00 -160.00 0 48.34 ·17428 10 05:48:45.8 6 0.46 1\36 COS 15 71954 13:24:35.0 54.00 -138.00 0 53.63 -133.26 10 13:24:36.3 13 1.23 1137 BC1 12121960 11:24:42.0 50.00 166.00 0 55.93 160.54 10 11:24:42.5 8 1.28 7 1138 BRK 23 71964 08:46:26.6 41.20 -130.80 3 4.5 BRK 40.42 -126.99 10 08:47:15.9 12 5.64 8 1139 !SC 04 21965 06:42:58.1 48.70 171.50 33 5.2 mb 51.90 174.30 10 06:43:16.1 8 0.55 7 1140 !SC 04 21965 17:35:20.0 50.00 171.00 33 4.0 mb 51.34 17S.04 10 17:35:43.6 5 0.77 7 1141 !SC 08 21965 21:59:57.4 49.80 171.80 33 4.5 mb 49.96 172.07 10 21:59:58.5 9 1.22 7 1142 COS 22 21965 08:49:32.8 49.00 168.00 30 4.4 mb 52.83 -164.37 10 08:51:55.6 8 2.32 7 1143 !SC 30 31965 03:15:34.0 49.00 172.00 33 4.6 mb 51.82 173.26 10 03:15:44.9 9 0.36 '1144 !SC 28121967 07:30:26.0 44.80 -135.60 33 4.2 mb 44.32 -130.04 10 07:31:12.9 16 2.24 8 1145 !SC 06 91969 10:54:40.0 47.00 -169.00 33 5328 -168.54 10 10:54:55.8 4 0.07 7 1146 !SC 20 11 1969 07:47:43.0 48.40 173.00 33 4.2 mb 51.TI 179.05 10 07:48:21.4 6 0.59 7 1147 !SC 01 41972 13:02:33.1 45.40 170.30 0 3.8 mb 52.08 -171.70 10 13:04:22.2 6 0.79 7 1148 LAO 17 8 1974 09:25:25.1 47.40 172.80 0 4.1 mb 47.16 172.83 10 09:25:28.8 4 1.18 7 1149 !SC 02 21975 09:32:35.0 49.90 171.00 0 42mb 49.75 171.17 10 09:32:39.2 5 1.14 7 1150 !SC 08 41975 18:49:02.8 49.10 168.00 0 4.6 mb 52.04 160.64 10 18:49:34.9 4 0.17 7 1151 !SC 28101985 06:15:45.8 49.30 171.70 33 4.4 mb 49.34 171.83 10 06:15:46.2 5 0.24 7 1152 ISS 05 21923 22:58:20.0 50.50 164.00 0 11 1153 ISS 07 71923 06:09:54.0 52.00 -142.50 0 52.41 -144.00 10 06:09:52.0 4 4.01 12 1154 ISS 28 61927 17:19:25.0 40.00 168.00 0 6.43 138.87 10 17:19:27.3 1.00 12 1155

WBL 29 71958 22:21:21.0 40.50 176.30 0 5.0WBL -40.50 176.30 14 1156 MOS 07 61961 05:03:02.9 44.38 -14920 30 4.5 MOS 45.50 150.80 61 05:03:10.7 14 1157 BSS 11 5 1962 20:00:00.0 31.23 -12421 0 15 1158

288

Sec

ISS ISS ISS G·R ERL GS PDE GS ISS GS ISS ISS G·R CGS ISC ISC ISC ISS G

ISC ISC ISC G

CGS ISC

ISS ISS ISS ISS ISS ISS ISS ISS ISS ISS ISS

CGS ISS

CGS CGS CGS ISS

CGS TAC TAC TAC BQ CGS ISS ISS ISS ISS USE USE USE USE USE USE USE USE USE USE USE USE USE USE USE PAS ISS

D .. e DMY

14 1 1936 08 51930 22 91930 30 61945 (1/ 51973 Z1 81978 04 5 1985 24 3 1986 21 1 1936 22121977 24 61924 2511 1927 20 61938 23 7 1961 08 4 1977 11 3 1976 12 6 1983 17 61924 10121949 12 4 1972 25 6 1983 24 9 1966 04 1 1933 06 10 1963 23 5 1976

09 71919 26101920 02 41922 30 81922 09 91922 26 3 1924 29 61926 27111926 30 7 1928 17 1 1930 18 4 1931 21 11936 03 3 1937 1212 1938 11 1 1939

01 10 1940 25 4 1941 05 21943 17 5 1954

08121954 09 5 1955 06 61958 20 3 1962 1411 1919 20 61922 09 8 1927 22 91931 19 2 1933 06 61933 21 8 1933 03 12 1933 14 21934

27 11 1934 27 11 1934 08 2 1935 24 51935 02 81935 20 8 1937 05 21941 18 21941 29 91941 18 7 1957 22 61971 14 5 1955

Michael E. Wysession el al.

Table 9

Sorted Datasel for Wesl anti South of Baja California


Time (GMT)

00:04:08.0 22:45:02.0 05:02:Z1.0 05:31:18.0 16:27:12.1 01:10:28.4 12:47:09.9 19:43:23.9 04:08:02.0 02:57:50.8 13:48:50.0 19:52:30.0 14:02:28.0 14:37:56.9 11:06:51.6 00:45:19.9 18:30:06.0 20:51:15.0 19:15:42.0 21:42:46.3 02:46:40.1 08:57:17.0 21:10:46.0 08:48:12.2 17:36:32.1

19:19:25.0 19:04:20.0 17:00:45.0 22:40:42.0 00:15:47.0 20:22:50.0 18:55:40.0 14:40:00.0 02:41:10.0 16:54:30.0 13:05:02.0 04:54:18.0 09:21:12.0 22:02:42.0 21:13:54.0 20:43:12.0 12:11:39.0 03:52:18.0 22:42:(1/.0 20:44:45.0 11:06:49.0 22:46:18.0 01:00:25.7 06:38:35.0 09:43:06.0 01:19:00.0 01:25:46.0 20:36:00.0 18:29:00.0 05:46:00.0 00:45:00.0 20:00:00.0 04:10:00.0 23:48:00.0 09:30:00.0 21:37:00.0 06:(1/:00.0 19:59:00.0 06:05:00.0 10:06:00.0 01:58:00.0 15:24:20.0 09:53:59.2 20:00:01.0

Lat. "N

23.50 17.00 16.00 17.00 16.57 14.93 16.67 18.08 13.50 10.88 8.00 1.00 6.00 6.80 2.50 4.70 4.60 1.00 4.00 6.20

10.60 12.10 28.00 21.90 22.80

17.00 19.40 11.00 20.00 17.50 17.00 7.00

15.00 9.00 8.00 9.50 7.50

11.00 5.50

12.70 22.00 26.10 20.00 16.15 12.80 13.33 20.00 25.30 11.00 13.00 12.50 15.50 14.20 14.30 13.60 20.30 17.33 14.17 11.08 12.67 11.25 10.50 14.17 10.42 15.28 14.66 0.00 4.45

28.90

Lcna· DqMh OE (km)

·115.50 ·112.00 ·111.00 ·115.00 ·116.08 ·113.(1/ ·113.53 ·113.57 ·1(1/.50 ·1(1/.43 -118.00 -129.00 ·119.00 -123.50 ·115.70 -112.10 ·112.20 -129.00 ·129.00 ·130.30 -130.40 -130.80 ·127.50 ·127.40 ·125.10

-112.00 -122.20 ·108.00 -114.00 -116.50 -112.00 ·1(1/.00 -117.00 ·115.50 ·106.30 -1(1/.00 ·108.00 ·118.50 ·110.00 -119.50 ·129.00 ·116.60 -115.50 -112.92 ·117.50 -111.00 -11 1.00 -117.90 -108.00 -120.00 ·109.70 -122.50 -122.70 -121.60 -124.80 -121.90 -119.42 -120.12 -125.08 -121.83 -126.08 -126.42 -122.08 -126.55 -120.05 -119.66 -110.50 -118.33 -126.20

o o

o 33 33 10 10 o

33 o o o

33 33 33 10 o o o

33 92 o o

33

o o o o o o o o o o o o o o o o o o o o o o

33 o o o o o o o o o o o

60 o

. 0 o

200 75 o o I o

LaL oN

Lona· DqMh oE (km)

Inlraplale Evenls

6.75PAS 4.8 mb 5.0 mb 4.7 mb 5.1 mb

4.9 mb

5.6 PAS

4.6 mb 4.9 mb 4.5 mb

5.75PAS 4.4 mb 4.6 mb 4.9 mb 5.5 PAS 4.0 mb

23.18 15.73 15.08 16.60 16.57 14.93 16.67

13.01 10.88 8.38 5.36 5.77 6.81 2.50 4.70 4.60 0.78 4.12 6.20

12.11 28.42 21.90 23.46

Casualties

2.5ML

19.03 17.52 15.10 18.56 14.Z1 17.81 16.92 16.57 11.54 8.08 8.32

10.15 25.22 6.50

14.82 35.53 27.73 29.93 22.65 -4.67 19.Z1 8.00

29.47

36.39 11.06 15.50 14.20 14.30 13.60 20.30 17.33 14.17 11.08 12.67 11.25 10.50 14.20 10.42 15.28 14.66 40.00 34.45 28.87

·115.14 -114.46 -112.(1/ -115.80 ·116.08 -113.(1/ ·113.53

-1(1/.87 -1(1/.43 -119.09 ·114.18 -118.96 ·123.67 ·115.70 -112.10 -112.20 -129.32 ·128.85 -130.30

-130.87 -126.99 -1Z1.4O -126.37

-108.00 -116.83

-90.39 -1(1/.86 -116.04 -112.91

·93.26 -108.45

·87.62 -103.37 ·106.44 -104.40 -110.87

-78.00 -114.91 -115.55 -115.14 -114.97 -109.60 -1(1/.33 -1(1/.95

-84.30 -114.90

-100.70 -111.16 122.50 122.70 121.60 124.80 121.90 119.42 120.12 125.08 121.83 126.08 126.42 122.10 126.55 120.05 119.66

-110.50 -118.33 -126.02

10 10 10 10 33 33 10

10 33 10 10 10 10 33 33 10 10 10 o

31 10 o

10

10 10 10 10 10 10 10 10 10 10 10 10 10 o

10 10 10 10 10 10 10

10

10 10

1 10

Tune (GMT)

00:04:00.8 22:44:11.5 05:01:47.0 05:31:20.0 16:Z1:12.1 01:10:28.4 12:47:09.9

04:(1/:58.4 02:57:50.8 13:48:47.9 19:54:40.8 14:02:27.2 14:37:53.0 11:06:51.6 00:45:19.9 18:30:06.0 20:51:19.4 19:15:45.0 21:42:46.3

08:57:16.7 21:10:45.0 08:48:12.2 17:36:35.6

19:19:44.2 19:05:04.1 17:02:21.1 22:41:32.1 00:15:50.7 20:23:31.6 18:56:21.3 14:40:49.3 02:43:52.0 16:54:44.5 13:04:50.4 04:54:27.7 09:23:38.0 22:02:46.0 21:14:25.0 20:45:54.4 12:12:00.6 03:52:22.0 22:46:14.1 20:41:57.0 11:(1/:37.9 22:44:07.0 01:01:26.7

09:42:53.6 01:19:07.2

11:59:00.0

20:00:05.8

N.

11 4 6

55

9 20

17

40 17

4 8

10 6

4 20 7 5

10 14

6 5

52

o (.)

1.43 1.37 3.00 2.00

2.43

3.06 2.93 1.03 1.50

2.13 1.70

0.97 2.20

1.61

1.71 19.54 3.77 1.11 3.41 0.70 2.92 3.86 2.62 2.99 2.29 2.09 1.45

1.09 0.61 1.82 1.36 1.55 1.38 0.91

0.80

25.25 4.02

1.56

PAGEOPH,

Code Site Index

1 2

2

2 2

8 9 8 8 8 8 8 8 8 9 8

11 12 12 14 14 14 14 14 14 14 14 14 14 14 14 14 14 14 14 14 15

1201 1202 1203 1204 1205 1206 12(1/ 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225

1226 12Z1 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269

Vol. 135, 1991

Initial Location

See

ISC GS ISC !SC !SC

!SC ISS ISS USE ISS ISS ISC CGS CGS ISS lMA WEL BC! CGS ISC

Date DMY

Time (GMl)

05 5 1976 21:41:59.3 22 9 1988 22:28:44.6 30 6 1972 10:04:59.4 14 2 1984 07:10:03.0 02 5 1975 07:15:44.0

14 51976 \1 \0 1928 17 61917 17 9 1932

01 10 1918 \1 10 1928 17 31976 24 7 1938 21 7 1945 21 91951 01 9 1953 30 9 1958 04 12 1959 08 10 1960 09 4 1967

06:21:11.3 23:32:28.0 08:34:30.0 13:03:00.0 00:20:15.0 23:44:16.0 18:47:20.7 03:05:00.0 21:57:57.0 09:41:20.0 14:14:53.0 18:38:49.0 18:09:10.0 17:34:00.7 17:52:10.0

Ut. "N

37.50 23.81 29.50 30.70

8.10

3.80 26.80 34.00 26.25 30.00 26.80 12.70 32.00 29.00 39.20 34.00 38.40 36.00 36.10 39.00

Initial Location

Sce

G-R ISS

CGS CGS CGS ERL GS GS GS ISC CGS NE! NE! NEI ISC ISC ISC GS G

ISC PDE ISS

Date DMY

24 81937 03121928 18 1 1955 31 3 1968 17 12 1969 20 61973 2S 61975 09 8 1976 12 81976 \1 6 1977

03111968 03 1 1987 03 1 1987 05 1 1987 12 91974 06 61975 11101975 12 7 1978 30 8 1931 08 7 1978 22111984 15 11 1932

ISS 23 9 1916 ISS 2S 1 1918 ISS 05 11 1918 ISS 30 8 1928 G 26121928

ISS 07 4 1929 CGS 17 10 1933 ISS 24 7 1934

CGS 09 \1 1934 CGS 14 10 1935 CGS 29 12 1936

Time (GMl)

20:13:23.0 12:26:10.0 08:42:03.0 04:16:01.0 03:\1:46.5 03:53:18.1 06:53:24.2 12:55:39.7 03:53:50.9 00:16:19.8 19:53:22.9 18:14:41.3 18:19:35.9 15:25:33.8 15:38:14.3 14:43:30.5 07:32:02.4 01:33:35.1 07:34:34.0 07:57:42.3 07:48:35.4 10:30:19.0

05:41:54.0 01:20:30.0 22:39:00.0 21:58:51.0 21:32:52.0 19:31:45.0 13:33:30.0 02:47:24.0 16:\1:42.0 17:36:30.0 13:58:42.0

Lat. "N

5.00 4.20 5.00 5.20 4.34 4.03 5.94 4.32 4.14 5.00 7.91 5.30 5.60 5.80 8.00 8.30 8.00

10.34 6.00 8.60

10.17 11.90

10.00 12.00 12.00 12.00 6.00

12.00 1\.00 8.20 7.00 5.00 5.00

Intraplate Seismicity

Table 10 Sorted DalaseI for Norlhcentral Paciftc Basin

Relocation

Lala· DcpIh Maa· OE (km)

LaL "N

Long. DcpIh oE (km)

Inlraplale EvenlS 175.10

-167.20 -163.70 -157.10 -156.10

-167.60 172.00

-162.00 -148.60 -174.00 172.00

-171.80 -162.00 -176.00 171.50

-135.70 176.40

-179.50 176.90 163.00

33 20 o

10 33

33 o o o o o

33 o o o

70 160

o 176 33

5.5 mb 3.0mb 4.9 mb 4.5 mb

37.50

29.50

8.41

Casuallies 4.8 mb -24.78

47.54

43.35 12.70 32.45

-19.82 39.20 34.00

-38.40 -35.18 -36.10

-7.02

Table 11

175.10

-163.70

-155.72

rn.23 151.60

172.29 -171.80 -\15.86 179.32 71.50

135.70 176.40

-179.38 176.90 129.66

Sorted DalaseI for Cocos Plale

33

o

\0

475 \0

10 10 10 10

\0

\18

Tune (GMl)

21:41:59.3

10:04:59.4

07:15:46.6

06:19:16.3 23:35:14.4

23:45:46.1 18:47:27.0 03:05:06.8 21:58:1 \.2

18:09:08.7

17:41:55.0

Relocation

Long. DcpIh OE (km)

-89.00 -85.00 -87.50 -87.30 -88.04 -87.79 -87.37 -87.52 -87.48 -88.30 -85.92 -94.30 -93.70 -93.20 -91.90 -91.30 -90.90 -91.75 -99.00 -95.80 -95.34 -99.50

-97.00 -95.50 -95.50 -95.50 -99.00 -95.50 -95.00 -87.00 -89.00 -86.00 -90.00

o o o

33 33 33 33 33 33 33 25 10 \0 \0 o o

127 33 o

33 33 o

o o o o o

o o o

Lat. "N

Intraplale Evenls 6.0 PAS

4.2 mb 4.4 mb 4.3 mb 4.4 mb 4.9 mb 5.0 mb

4.7 mb 4.8 mb 4.5 mb 4.2 mb 4.3 mb 3.9 mb 4.211JL 3.9 mb 5.6 mb 4.0 mb 4.2 mb

4.98 3.47 5.04 5.04 4.46 3.91 5.48 4.32 4.14 4.74 7.79 5.36 5.75 s.s5 8.12 8.41 7.45

10.65 5.85 8.54

10.49 1\.49

Casuallies

6.0 PAS

12.12 14.77 13.27 12.56 5.83

14.37 10.57 9.00 7.21 4.74

13.50

-89.09 -87.62 -87.54 -87.33 -87.95 -87.72 -87.59 -87.52 -87.48 -87.60 -86.09 -94.30 -93.68 -93.64 -91.92 -91.31 -91.69 -91.59 -99.97 -95.62 -95.85 -99.26

-83.49 -9\.29 -88.12 -95.23

-101.44 -94.10 -93.90 -84.82 -79.46 -82.52 -92.50

\1 10 10 10 10 10 10 33 33 10 10 10 10 10 10 10 10 10 10 10 \0 10

10 10 10 10 10 10 10 10 10 10 10

Time (GMl)

20:13:27.3 12:26:20.8 08:42:05.8 04:15:59.0 03:\1:47.0 03:53:\8.2 06:53:20.7 12:55:39.7 03:53:50.9 00:16:19.8 19:53:23.2 18:14:43.1 18:19:38.7 15:25:35.0 15:38:\8.7 14:43:35.3 07:31:51.0 01:33:37.7 07:34:33.8 07:57:42.1 07:48:38.3 10:30:12.3

05:43:09.1 01:20:54.8 22:39:59.6 21:59:08.3

19:32:16.9 13:33:19.5 02:47:34.0 16:11:38.8 17:36:50.4 13:59:58.0

N. <I

(.)

10 6

N.

"t1 7

24 18 15 13 19

44 14 10 9 5 4 5

10 10 5

10 10

10

16 30 \1 \1 8

10

1.20

1.46 1.15

1.24

2.13 1.26

1.41

<I

(.)

\.72 2.11 1.72 1.45 1.64 2.62 2.26

1.15 2.63 1.59 0.82 1.04 0.95 2.78 1.59 1.82 1.83 1.37 1.70 1.00

\.63 1.54 2.31 2.58 \.94 3.24 \.33 4.05 1.34 0.73

289

Code Sile Index

2

\1 \1 12 12 12 14 14 14 14 14 14 14 14

1301 1302 1303 1304 1305

1306 1307 1308 1309 \310 1311 \312 1313 \314 \315 1316 1317 1318 1319 1320

Code Site Index

2 2 1 1 1

7 8 9

CO-I CO-2 CO-2 CO-2 CO-2 CO-2 CO-2 CO-2 CO-2 CO-2 CO-3 CO-4 CO-4 CO-4 CO-5 CO-5 CO-5 CO-6 CO-7 CO-8 CO-9 CO-IO

1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372

1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383


Table 11 (continued) Sorted DalaStl for Cocos Platt

Initial Location Relocation Code Site Inde.

See Olle Time Lal. Lona· Deph Mq. Lot. LooI· Deph Time NI " DMY (GMT) "N "E (km) "N oE (km) (GMT) (s)

BC! 12 41951 04:10:30.0 8.00 ·86.00 0 11.56 ·78.34 10 04:11:24.7 4.14 7 1384 CGS 10 3 19S5 04:SS:10.0 10.00 ·90.SO 0 11.29 ·90.11 10 04:SS:22.1 14 1.61 138S TAC 20 4195S 10:16:S9.0 7.50 -99.00 0 7.87 ·103.89 10 10:16:S3.8 11 0.49 8 1386 CGS 26 41961 12:18:32.0 UO ·91.10 33 IS.62 ·89.91 10 12:19:16.0 6 1.S7 7 1387 CGS 21 21962 23:46:49.4 3.00 ·86.SO ZS 3.79 ·83.47 10 23:46:44.4 6 0.84 8 1388 TAC 14 3 1966 IS:09:I1.5 11.42 -96.67 0 11.15 -9S.71 10 IS:09:12.8 4 0.69 7 1389 CGS 20 81967 16:43:29.8 7.20 -85.10 33 4.2 mb 7.41 -84.80 10 16:43:30.5 21 2.21 9 1390 ISC 01 S 1969 16:27:40.0 10.00 ·92.40 33 4.2 mb 18.08 -94.52 10 16:28:S8.8 10 2.13 7 1391 ISC 19 91969 22:21:SO.0 10.90 -96.90 33 13.78 -96.33 33 22:22:22.8 9 1392 !SC 09 1 1970 12:OS:23.0 11.60 -93.30 33 4.2 mb 15.71 -93.45 88 12:06:06.6 \1 1.26 7 1393 !SC 29 41970 16:01:17.0 11.10 -94.20 33 4.7 mb 12.30 -94.20 33 16:01:31.8 10 1.60 7 1394 !SC 22 51970 03:08:16.0 11.50 -94.SO 33 4.3 mb 17.S9 -93.10 45 03:09:05.2 8 2.80 7 139S ISC 09 S 1973 14:20:31.0 12.00 -97.00 0 4.3 mb 12.87 -97.56 10 14:20:39.0 4 O.SS 7 1396 !SC 18101973 21:S8:24.9 9.50 -90.60 0 4.3 mb 12.74 -91.47 10 21:S8:53.3 8 3.11 7 1397 ISC 11 12 1974 20:49:44.1 7.20 -91.30 0 3.4 mb 13.52 -87.39 10 2O:S0:33.1 6 0.29 7 1398 ISC 24 4197S 18:12:41.0 12.80 -99.30 SO 4.Smb 16.30 -98.70 33 18:13:07.9 16 0.80 9 1399 GS 26 21976 23:ZS:ZS.5 8.3S -87.46 33 4.6 mb 11.60 -87.68 188 23:26:05.7 19 0.99 9 1400 ISC 29 91977 OS:28:36.1 4.20 -95.00 0 2.41 -9S.61 10 OS:28:27.9 10 0.74 8 1401 ISC 08 3 1978 23:42:07.0 10.00 -93.90 33 4.4 mb 13.50 -91.10 194 23:42:36.8 9 1402 ISC 24 41978 OS:42:OO.0 7.30 -89.10 146 2.5 mb 10.68 -~S.12 146 OS:42:S4.0 9 1403 !SC 00 71980 23:56:40.0 11.90 -96.90 33 3.09 -90.03 10 2:l:55:O5.3 10 1.29 8 1404 ISS 04 1 1918 04:30:OS.0 10.50 -91.00 0 \1 1405 ISS 04 1 1918 04:32:ZS.0 10.50 -91.00 0 11 1406 ISS 1611 1918 OS:S6:30.0 12.00 -95.SO 0 15.80 -93.66 10 OS:57:06.7 0.00 11 1407 CGS 07 61937 04:03:12.0 5.00 -92.00 0 12 1408 !SC 05121966 00:10:00.0 7.00 -94.20 33 4.0 mb 2.92 -98.74 10 00:09:38.8 1.16 12 1409 ISC 13 81979 11:42:ZS.0 8.70 -88.80 33 5.0 mb 12 1410 BC! 28 31958 13:06:30.0 12.00 -95.SO 0 13 1411 CGS 26 I 1962 18:40:23.0 10.30 -90.60 45 13 1412

Table 12 Sor/ed Dalastl for Nazca Plalt

Initial Location Rclocation Code Site Inda

See Date Tune LII. Loo .. Deph Ma .. Lot. LooI· Deph Time N. " DMY (GMT) "N "E (km) "N OE (km) (GMT) (I)

In/rap/alt Evenls

!SC 29 71968 11:48:57.0 -2.00 -90.10 33 4.3 mb -1.96 -90.09 10 11:48:56.3 9 1.00 1 NZ-1 1501 ERL 14 81971 04:26:03.9 -2.62 -89.15 33 4.6 mb -2.50 -89.21 10 04:26:03.6 20 1.53 I NZ-I 1500 !SC 16 81975 09:43:48.3 -2.10 -90.33 33 4.8MB -1.95 -90.24 10 09:43:47.6 15 1.63 I NZ-I 1S03 CGS 07 81951 04:08:54.0 -6.50 -85.00 0 -6.54 -84.98 10 04:08:53.8 19 1.07 I NZ-2 IS04 ISS 20 819S9 07:18:34.0 -6.90 -85.20 0 -6.77 -85.10 10 07:18:37.9 47 1.26 I NZ-2 150S

CGS 24 12 1963 00:40:42.6 -6.60 -8S.10 33 4.3 mb -6.48 -8S.03 10 00:40:41.8 \1 0.78 I NZ-2 1506 !SC 18 61983 07:14:53.1 -6.10 -8S.40. 33 4.4 mb -6.22 -8S.4O 10 07:14:SO.8 7 2.44 I NZ-2 IS07 GS 00 10 1988 23:21:03.2 -7.07 -88.62 10 4.7 mb -7.07 -88.62 10 23:21:03.2 2 NZ-3 IS08 !SC 1211 1969 06:44:51.0 -3.80 -99.SO 33 -3.60 -99.SS 10 06:44:51.2 7 0.49 I NZ-4 IS09 ISS 23 81923 OS:12:4S.0 -S.OO -9S.oo 0 -S.39 -96.91 10 OS:12:48.8 4 0.87 1 NZ-4 ISIO !SC 11 11967 13:37:SI.0 -4.80 -96.60 33 4.8 mb -4.78 -96.64 10 13:37:48.9 5 0.66 I NZ-4 1S\l !SC 07 91986 07:SO:39.S -8.06 -97.60 10 4.8 mb 3 NZ-4 ISI2 !SC 17121972 16:00:33.9 -11.00 -90.20 6S 4.7 mb 3 NZ-4 IS13 G OS 81944 00:S7:17.0 -13.SO -92.SO 0 6.0 PAS -12.94 -92.44 10 00:S7:23.0 13 0.54 NZ-4 ISI4 G OS 81944 01:24:08.0 -13.50 -92.SO 0 6.ZSPAS -13.29 -92.62 10 01:24:12.4 16 0.7S NZ-4 IS15

!SC 12 31969 07:OO:ZS.6 -11.70 -9S.2O 33 4.3 mb -11.60 -9S.18 10 07:02:24.6 7 O'ZS NZ-4 ISI6 CGS 22 819S8 12:18:53.0 -11.50 -97.00 0 -12.72 -97.74 10 12:18:51.7 8 0.S7 NZ-4 1S17 !SC IS 41976 14:27:43.4 -14.00 -98.90 33 -13.98 -99.00 10 14:27:41.8 4 0.67 NZ-4 1S18 !SC ZS 11 1965 IO:SO:SO.8 -17.10 -100.20 143 S.3 mb -17.10 -100.20 13 10:SO:4O.0 NZ-4 IS19 !SC 08 21982 15:03:07.9 -17.80 -99.00 10 4.7 mb NZ-4 IS20 !SC 23 I 1983 06:00:52.9 -17.90 -96.80 10 4.2 mb NZ-4 IS21 !SC 08 101967 23:43:43.1 -17.20 -106.00 33 4.4 mb -17.17 -106.11 10 23:43:42.1 13 0.67 NZ-S IS22 OS 16 61975 OS:OS:24.3 -20.37 -109.12 33 4.8 mb NZ-6 1523 !SC 08 12 1964 09:14:31.6 -21.30 -82.00 48 5.0 mb -21.26 -82.06 18 09:14:30.9 30 1.18 NZ-7 1S24 !SC 19101972 20:31:54.3 -22.51 -82.94 33 4.3 mb -22.49 -82.99 10 20:31:53.0 26 1.56 NZ-7 ISZS ERL 20 41973 1S:16:tS.4 -28.76 -108.6S 33 4.9 mb -28.59 -108.64 10 15:16:14.7 8 0.33 NZ-8 1526 !SC 09 11986 09:OS:34.9 -29.00 -106.SO 10 4.8 mb -28.83 -106.14 10 09:05:36.8 24 1.79 NZ-8 1527


Table 12 (continued) Sorted Dataset for Nazca Plate

Initial Location Relocalion Code Site Index

See Date Tune LaL Lon,. Deprh Moa· LaI. Lon,. Deprh Time N. 0 DMY (GMT) "N oE (bn) "N OE (bn) (GMT) (I)

PDE tt 3 1984 13:42:56.4 -26.65 -108.51 10 5.7 mb NZ-8 1528 GS 06 71917 01:06:07.7 -27.00 -101.29 10 6.2 mb NZ-8 1529 GS 01 71917 tt:SO:14.8 -26.97 -101.16 10 6.1 mb NZ-8 1530 GS 02 61918 02:53:45.5 -26.82 -108.16 10 4.7 mb -26.82 -108.16 10 02:53:45.5 NZ·8 1531 JSC 01 11970 04:44:42.0 -29.40 -90.60 33 -28.73 -90.00 10 04:44:38.2 5 0.57 NZ-9 1532 JSC 21 61964 07:42:55.0 -30.00 -91.SO 33 4.5 mb -29.89 -91.75 10 07:42:54.2 10 0.71 NZ-9 1533 JSC 25 91965 16:27:26.0 -3O.SO -90.60 33 4.6 mb -30.37 -90.55 10 16:27:25.6 11 1.02 NZ-9 1534 G 25 51918 19:29:20.0 -30.50 -91.SO 60 7.0 PAS -30.15 -91.88 10 19:29:19.1 tt 1.75 NZ-9 1535

BCI 05 1 1960 02:27:36.0 -35.00 -92.00 0 -31.27 -90.34 10 02:27:56.2 8 1.20 NZ-9 1536 ERL 28 51972 01:38:51.5 -32.85 -92.11 33 5.0 mb NZ-9 1537 ISS 21 1 1959 10:04:09.0 -30.70 -78.90 0 -30.65 -79.01 10 10:04:12.9 52 1.25 NZ-10 1538 JSC 23 21968 12:28:04.0 -30.00 -76.00 33 -29.41 -75.96 10 12:27:59.3 4 0.47 NZ-11 1539 eGS 26 I 1962 14:34:45.7 -36.90 -88.90 60 -36.81 -89.21 10 14:34:43.3 14 1.24 NZ-12 1540 JSC 04 61981 07:39:41.5 -34.20 -78.70 37 5.1 mb -34.04 -78.86 10 07:39:39.7 60 1.57 NZ-13 1541 NE! 27 10 1986 01:02:52.3 -34.20 -78.90 10 5.0 mb NZ-lJ 1542 JSC 22 11981 02:23:58.7 -34.40 -79.60 33 4.1 mb -34.47 -79.74 10 02:23:56.9 0.91 NZ-13 1543 JSC 14 91974 17:20:17.0 -35.80 -79.60 33 -35.66 -79.65 10 17:20:16.6 0.18 NZ-13 1544 JSC 13 61973 15:10:38.8 -37.00 -79.30 0 4.1 mb -36.91 -79.39 10 15:10:42.5 0.56 NZ-13 1545

Casualties ISS 30 11921 18:58:38.0 -0.50 -93.00 0 -0.36 -93.84 10 18:58:49.9 2.19 1546 ISS 06 1 1922 19:20:38.0 -19.00 -76.00 0 -16.50 -73.0 1547 ISS 06 61925 20:45:25.0 -32.00 -95.SO 0 -47.77 -109.15 10 20:43:24.2 2269 1548 ISS 05 61926 01:20:15.0 -17.00 -78.SO 0 -15.20 -76.07 10 01:20:42.8 4 0.13 1549 ISS 05 81927 03:43:03.0 0.00 -85.00 0 -0.80 -83.23 10 03:43:35.6 7 2.19 15SO ISS 23 51928 20:24:44.0 -30.00 -77.00 0 -29.77 -79.45 10 20:24:32.1 4 3.36 7 1551 eGS 09 91937 05:29:00.0 -5.00 -84.00 0 12.01 -93.56 10 OS:32:35.4 9 1.16 1552 eGS 15 tt 1937 00:51:06.0 -14.60 -8240 0 -7.38 -77.33 10 00:51:22.5 5 0.27 1553 eGS 28 1 1940 07:27:48.0 -1200 -8200 0 10.74 -8.5.10 10 07:30:34.4 6 0.66 7 1554 eGS 31 81940 17:12:42.0 -17.00 -78.00 0 -15.39 -70.63 153 17:12:37.0 12 1.92 7 1SS5 eGS 10 81941 17:02:48.0 -5.00 -88.00 0 -32.00 -70.00 0 16:59:45.0 9 1556 ISS 15 21942 14:20:49.0 -9.70 -85.10 0 -1.41 -84.13 10 14:20:56.1 10 1.15 7 1557 ISS 13 2 1947 00:13:21.0 -19.00 -75.00 0 -11.27 -75.09 10 00:13:27.0 13 1.55 7 1551 eGS 07 91949 07:26:57.0 -2.00 -93.00 0 0.79 -90.26 10 07:27:09.5 I 1.25 I 1559 eGS 16 719SO tt:51:48.0 -2.00 -86.00 0 -1.25 -86.12 10 11:59:00.1 13 1.92 8 1560 Bel 29 12 1951 10:31:00.0 -31.00 -96.00 0 -]9.24 -95.17 10 10:37:19.9 12 1.72 I 1561 BC1 24 tt 1951 21:42:49.0 -10.00 -85.00 0 -10.19 -80.70 10 21:42';29 4 0.55 7 1562 eGS 23 51960 12:02:36.0 -38.00 -76.00 60 -39.04 -74.42 10 12:02:28.4 I 0.84 7 1563 eGS 19 61960 21:42:50.0 -36.50 -76.00 0 -36.84 -73.10 10 21:42:44.7 10 0.91 7 1564 eGS 22 10 1961 07:01:42.1 -0.20 -94.60 60 0.22 -98.39 10 07:01:53.8 tt 1.61 1565 eGS 27 10 1963 01:30:32.1 -29.50 -101.20 33 4.6 mb -36.07 -103.92 10 01:29:55.0 17 207 I 1566 eGS 05 11964 17:15:14.2 -29.80 -105.20 33 4.5 mb -30.57 -IOS.21 10 17:15:07.8 I 1.14 8 1567 CGS 09 1 1964 21:19:31.4 -1.20 -19.90 33 4.0 mb 0.11 -90.33 10 21:19:54.3 5 0.46 I 1568 eGS 11 11964 07:10:21.9 -32.50 -103.70 33 4.5 mb -31.88 -103.46 10 07:10:25.3 12 1.59 I 1569 eGS 29 11964 17:17:34.2 -3.20 -9220 33 4.2 mb 1.31 -92.24 10 17:18:01.8 10 1.72 8 1570 eGS 25 3 1965 20:36:32.0 -38.10 -ao.20 33 4.4 mb -42.00 -75.17 10 2O:36:OS.5 12 0.97 7 1571 eGS 30 51965 07:21:56.2 -37.70 -89.30 27 4.2 mb -40.65 -88.43 10 07:21:40.6 7 1.00 8 1572 eGS 23 71965 23:21:25.3 -39.10 -85.SO 33 4.9 mb -39.40 -15.64 10 23:21:22.9 12 0.86 1573 CGS 23 91965 iS:16:58.1 -36.80 -91.70 33 4.2 mb -36.57 -92.79 10 15:16:58.8 9 1.43 1574 JSC 03 51967 01:52:00.0 -29.00 -105.SO 33 4.1 mb -31.32 -IOS.99 10 01:51:43.3 8 0.95 1575 ISC 16 51967 08:25:07.0 -20.00 -100.60 33 4.4 mb -28.78 -106.07 10 01:24:15.3 10 1.83 8 1576 JSC 15101967 15:16:10.0 -30.00 -98.00 33 4.2 mb -30.49 -98.04 10 15:16:06.7 5 229 8 1577 ISC 07 21968 07:28:30.0 -28.00 -93.70 33 4.1 mb -32.63 -95.33 10 07:28:04.3 5 1.19 8 1578 JSC 19 31968 23:06:49.0 -32.00 -100.30 33 4.4 mb -30.64 -100.07 10 23:06:53.3 6 1.73 8 1579 ISC 05 61968 14:07:tt.0 -3.30 -98.00 33 4.3 mb 2.12 -97.04 10 14:07:48.5 9 0.49 8 1580 eGS 30 81969 05:43:33.2 -0.61 -89.62 33 4.2 mb -0.64 -89.54 10 OS:43:32.2 10 1.48 7 1511 ISC 20 51970 03:26:46.5 -30.70 -98.40 25 -34.09 -99.69 10 03:26:26.8 4 0.45 7 1582 JSC 10 91970 04:32:08.0 -2.00 -84.00 33 -2.25 -84.01 10 04:32:06.6 4 289 7 1583 ISC 12 61972 00:54:48.2 -18.50 -I04.SO 0 3.8 mb -28.12 -112.38 10 00:53:44.7 1.17 8 1584 JSC 15 4 1973 13:08:42.4 -30.50 -76.10 0 3.4 mb -3035 -75.40 10 13:08:49.3 10 1.91 7 15as GS 27 12 1975 02:44:00.2 -3.51 -88.62 33 4.5 mb 0.00 -90.97 10 02:44:31.9 14 0.65 I 1586 ISC 23 41977 14:27:tt.1 -27.70 -77.10 33 -30.15 -72.89 10 14:28:09.4 11 252 7 1517 JSC 15 51977 16:46:13.9 -35.60 -15.30 33 4.6 mb -40.03 -86.34 10 16:45:48.2 9 1.06 8 1518 JSC 01 101977 00:42:32.1 -37.90 -76.SO 33 -37.90 -76.56 10 00:42:31.5 tt 257 7 1519 NE! 19 71978 03:29:25.4 -3.10 -84.30 33 -3.44 -84.14 10 03:29:28.2 6 222 7 1590 JSC 04 41980 01:18:56.4 -10.60 -1220 33 -11.21 -80.96 10 08:19:13.9 6 1.8.5 7 1591 JSC 06 51911 09:30:55.0 -18.30 -75.20 33 -11.26 -74.86 14 09:30:53.1 9 1.21 7 1592 JSC 27 41982 21:51:36.9 -23.80 -75.30 33 -23.99 -74.67 10 21:51:43.4 7 2.16 7 1593 JSC 27 10 1982 14:37:07.7 -29.40 -75.20 33 -29.49 -74.45 10 14:37:13.5 22 2.45 7 1594 JSC 04 12 1983 00:14:12.0 -34.20 -71.10 33 -32.54 -71.84 10 00:15:19.9 16 0.71 7 1595

292

See

NEI ISS ISS ISS ISS ISS

CGS ISS ISS ISS ISS ISS ISS

CGS ISC ISS

See

G·R NOS NOS NOS NOS ISS

CGS ISC GS ISC ISS PDE CGS GS GS

ISS G G

G·R ISS ISC GS !SC ISS G·R WEL

Da .. DMY

30 71916 16 21917 21 21917 22 21917 28 91918 07 91919 11 111963 15 21917 03 21918 01 91920 06 61925 07 31926 28 10 1947 08 21962 27 71972 28 91918

DI" DMY

12 3 1927 09 51971 09 5 1971 09 5 1971 09 5 1971 04 12 1956 28 I 1961 11 1 1976 07 3 1988 29 4 1975 05 12 1927 15 6 1984 06 61970 05 2 1977 07111979

11 5 1922 27 12 1926 27121926 18 5 1944 18 1 1949 16 8 1967 20 6 1974 17 1 1967 01 91919 14 8 1929 13 9 1975

Michael E. Wysession er al.

Table 12 (continued)

Sorted Datasel Jor Nazca Plate


Time (GMD

06:09:16.6 02:12:10.0 09:41:22.0 09:12:20.0 10:19:30.0 20:21:16.0 16:21 :14.4 00:48:09.0 14:41:50.0 02:45:50.0 03:42:25.0 20:32:30.0 09:39:21.0 08:28:26.8 15:19:21.7 10:35:20.0

l..Il oN

·30.20 ·26.00 ·26.00 ·26.00 ·26.00 ·29.00 ·32.80 ·26.00 -3.00 -3.00

-30.00 -9.50

·14.80 ·1.70

-36.10 -26.00

Leng. DepIh oE (km)

-75.70 -80.00 -80.00 -80.00 -80.00 -98.00 -95.50 -80.00 -88.00 -88.00 -77.00 -84.00

·106.50 ·84.60 -91.70 -80.00

33 o o o o o

33 o o o o o o

45 o o

Mag.

4.2 mb

3.9 mb

Lat. 'N

-30.31

-24.73

-9.36 -30.32

·15.07

-29.71

Table \3

Leng. Depth oE (km)

-75.26

-75.80

-96.23 -77.69

·107.59

-78.15

10

10

10 10

10

10

Time (GMT)

06:09:21.2

00:49:02.7

02:45:07.7 03:42:23.9

09:39:22.5

10:17:38.1

Sorted Datasel Jor Alltarctic Plate

Initial Location

Time (GMD

18:44:32.0 08:25:01.7 08:53:25.9 18:00:59.9 18:35:09.8 10:07:54.0 14:06:12.6 23:22:41.2 08:50:44.1 08:33:27.3 17:49:30.0 04:18:57.4 06:14:11.9 03:29:18.9 11:31:49.6

00:44:32.0 08:42:55.0 09:20:30.0 19:55:12.0 04:43:18.0 10:03:08.0 09:03:20.9 21:37:08.0 19:12:25.0 02:16:50.0 08:40:10.0

Lat. 'N

-41.00 -39.78 -39.74 -39.84 -39.72 -45.50 -45.10 -46.40 -44.52 -45.80 -68.00 -68.54 -62.59 -66.45 -62.58

-48.80 -57.00 ·57.00 -44.00 -44.50 -55.20 43.91 -57.00 -69.00 -66.00 -51.00

Lenl· DepIh OE (km)

-106.00 -104.84 -104.93 -104.89 -104.98 -106.90 -106.40 -101.10 -93.26 ·87.20 -90.00

-111.73 -93.27 -82.58 -72.91

-79.00 -110.00 -110.00 -109.00

·90.50 -83.00 -88.58 -85.00

-108.00 175.00

-105.00

50 33 33 33 33 o

25 33 10 o o

10 33 33 10

o o o o o

33 33 33 o o

33

l..Il 'N

Relocation

Looa· Depth OE (km)

Tune (GMD

IlIlrap/ate Evellts

6.5 PAS 6.2 mb 5.2 mb 5.4 mb 5.4 mb

5.3 mb 4.9 mb 4.1 mb

5.2 mb 4.9 mb 6.2 mb 5.1 mb

-39.28

45.40 45.00 -46.40 -44.52 45.80 -66.81

-66.45 -62.58

CasUIJ/ties

6.0 PAS 6.25PAS 6.0 PAS

4.8 mb 5.3 mb

6.0 PAS

-50.43 -61.17 -61.88 43.84 42.40 -55.76 43.89

-65.89

·103.91

·107.30 -106.50 -101.10

-93.26 -87.20

-102.80

-82.58 -72.91

-77.54 -104.58 -103.61 -110.87 -90.80 -76.07 -88.91

-112.57

10

10 10 33 10 o

10

33 10

10 10 10 10 10 10 10

10

18:44:41.5

10:07:55.7 14:06:10.5 23:22:41.2 08:50:44.1 08:33:27.3 17:49:17.2

03:29:18.9 11:31:49.6

00:44:39.6 08:42:50.2 09:20:22.4 19:55:15.8 04:43:33.0 10:02:55.7 09:03:19.2

19:12:33.3

Ns

18

N.

13

4 15

" (s)

1.23

1.80

5.87 10.46

1.03

0.00

" (.)

2.33

0.90 1.30

2.03

1.97 1.19 2.08 0.92 1.80 1.81 1.30

2.21

PAGEOPH.

Code Site Index

11 11 11 11 11 11 12 12 12 12 12 12 12 12 14

1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611

Code Si.. Index

2

11 12 12 14

1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715

1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726


regrouped and sorted by date. The index number reftects this particular arrangement. The numbering sequence has gaps to allow for the possible updating of the catalog in the future.

A computerized version of the whole dataset is available from the senior authors. This package comprises both the sorted dataset, arranged geographically and by relocation code, and the chronological list of all events in Tables 1-13. In addition, it also includes separate files for the swarm events described in Appendix A.

The general processing code represents the nature of the analysis performed for each earthquake supposed to be intraplate, and is a factor of the type of data available and the results of the relocation. The events can be divided into two basic categories: those we deern to be true intraplate earthquakes, and the casualties we consider are not.

Intraplate:

• Code I (123 events): Relocation yields a statistically significant intraplate epicenter.

• Code 2 (72): Relocation confirms the bulletin listing as intraplate, and the bulletin listing has been retained.

• Code 3 (52): Epicenter very weIl located by the reporting agency, with no relocation deemed necessary.

• Code 4 (28): Modern event located by a local array of either Wake Island, Enewetak Island or Ocean Bottom Seismometers, retained here without relocation (WALKER, 1989).

• Code 5 (128): Modern event located by a local array of Polynesian stations, retained without relocation (OKAL et al., 1980; J. TALANDIER, pers. commun., 1989).

• Code 6 (3): Relocation makes them part of a swarm described independently in Appendix A.

Casualties:

• Code 7 (196): Earthquakes that were relocated to seismically active subduction zones.

• Code 8 (76): Events relocated to a ridge or transform fault system. • Code 9 (30): Original location is intraplate, but a second listing from a distinct

agency is interplate. • Code 10 (2): Listing appears in WALKER'S (1989) catalog, but epicenter locates

on a ridge system. • Code 11 (32): Two few arrival times available to perform a meaningful reloca

tion. • Code 12 (37): Relocation was attempted but failed to converge to a stable

solution.


• Code 13 (4): Events for whieh no arrival times eould be obtained. • Code 14 (41): Original loeation was blatantly wrong, as the probable result of a

clerical error in reporting. • Code 15 (70): Announeed or presumed to be a nuclear explosion.

Discussion

Code 1: For an event for whieh a reloeation was performed to be eonsidered truly interplate, several eriteria had to be satisfied. The new loeation had to not only be intraplate but also remain at least 2° from a plate boundary, a requirement diseussed above. In addition the event had to be statistieally signifieant, as determined through the Monte Carlo simulation, also diseussed above. Otherwise the epieenter was rejeeted as interplate and given a eode of either 7 or 8, depending on the type of plate boundary. All events initially deemed intraplate from a reloeation were given a Monte Carlo test.

Code 2: In these eases a reloeation and Monte Carlo test were performed but the new loeation did not differ signifieantly from that listed in the bulletin. Events with this eode did not reloeate more than OS and had depths whieh, when uneonstrained, did not beeome intermediate or deep. There are a few borderline eases, however, where the epieenter moved less than 0.5° but in sueh a direetion that it beeomes less than 2° from a plate boundary, thus giving it an interplate rating.

The depths listed by the various ageneies for these events (as weil as with eodes 3,4 and 5) ranged from 0 km to below 60 km, and were eommonly 33 km. We now know that oeeanie intraplate earthquakes are generally shallower than the latter, and therefore do not neeessarily endorse these depths. If an epicenter did not move mueh during the reloeation we kept its original depth unless it was substantially overestimated, in whieh ease we ehanged the depth and adjusted the origin time (e.g., Event 1367 loeated by the ISC as being 127 km deep in the interior of the Coeos plate. Reloeation verified it as intraplate, but a better standard residual was found with a depth of 10 km).

Code 3: If the loeation given by the reporting ageney was eonsidered to be very reliable, then no reloeation was performed. This required a large number of travel time arrivals, very small residuals (with a standard deviation on the order of 1.0 s) and a fairly eomplete azimuthai eoverage of stations. These events were in the middle of plates so there was little ehanee of reloeating to plate boundaries. It was also important that other arrivals listed but not used in the ageney's loeation, sueh as S, PKP, SeS, ete., were weil fit by the loeation. This eode is not given to any historical (pre-1963) events beeause all of these were reloeated (when enough data was available).

Codes 4 and 5: For the 152 events loeated by loeal arrays no reloeation was performed. A full deseription of the Polynesian array and its loeation algorithm ean


be found in OKAL et al. (1980). We refer to WALKER and MCCREERY (1985) regarding the Western Paeifie arrays.

Code 6: These three events had original loeations that were independent from any other seismieity, but their reloeations yielded epieenters and origin times that were eoneurrent with large swarms. The faet that this occurred only three times for swarms that included a total of over 800 events reassuringly implies a great degree of accuraey in the determination of these swarms.

Codes 7 and 8: These 272 events reloeated to plate boundaries-either trenehes and Wadati-Benioff zones or ridge and trans form systems. These earthquakes were labelled as being interplate for one of three reasons: either their hypoeenters were direet1y on, or within 2° from, or had 95% eonfidenee ellipses that interseeted, a seismieally aetive plate boundary. This was largely the ease for events listed by WALKER (1989) near the Tonga and Japan-Kuril trenehes, and events from the NEIC tape listed near the East Paeifie Rise and the Nazea-Antaretiea plate boundary. A large number of events reloeated to interplate regions using only the arrival times used in the original loeation. More often, however, the events reloeated to aetive plate boundaries after applying teehniques diseussed earlier: including S waves in the inversion, floating the depth and removing emergent arrivals.

Code 9: For these events the existenee of a well determined interplate epieenter by one ageney invalidated the intraplate loeation by another. The interplate epieenter was preferred either beeause it was a more thorough loeation with more data and/or smaller residuals, or beeause we reloeated the event and found it agreed with the plate boundary loeation.

Code 10: These two events, listed as intraplate in WALKER (1989), were within 2° of the East ~aeifie Rise and were therefore removed from the intraplate dataset.

Code 11: A loeation was given for whieh there were three or less P or S arrival tim es and a reloeation was not possible. Most of these events occurred before 1930, when ISS loeations were oeeasionally done on the basis of the arrival times of surfaee waves. Other exclusions were for loeations determined on the basis of only 3 arrivals. While an inversion for three parameters ean be earried out using three arrival times, the solution is uniquely determined and yields residuals of zero for the arrival times (unless the arrival times have differenees greater than a eertain threshold of eompatibility). We do not feel that this is a meaningful solution, and require a minimum of four arrivals to aehieve some measure of reliability.

Code 12: There were two possible results of the reloeation that were deemed as failing to eonverge to a stable solution. The first is astriet interpretation of this: The inversion, no matter how many times it was iterated did not reaeh a final loeation. Unable to find a minimum in the residual spaee, the least-squares regression would yield a very different loeation with eaeh sueeessive iteration. In the seeond ease the inversion pro gram would aetually iterate to a single epieenter for any given subset of arrival times, but the addition or exclusion of one arrival time


could significantly change its location (in some cases by over 10°), and there was no rationale for choosing one station over another. This was often accompanied by excessively large residuals for what was usually a small number of arrival times, indicating the inadequacy of the location. Both of these cases are representative of large errors in the available arrival times.

Code 14: There were a surprisingly large number of blatant typographical errors in the reporting bulletins, signified by Code 14 in Tables 1-13. These errors were usually in the form of a missing or improperly included negative sign or reversal of adjacent numerals either in the latitude, longitude or origin time. Most of these errors are not traceable to the original locating agency but rather seem to occur upon transcription by the cataloging agency.

A dramatic example consists of 14 USE epicenters (1253-1266) listed as occurring off the coast of Mexico during the years 1933 and 1941; these epicenters formed a very distinct curvilinear trend around (l5°N, 125°W), but it did not match with any known bathymetric features in this region, and no references to these events could be found in any other bulletins. However, when the negative signs of the longitudes were removed they located precisely on top of the known seismicity of the Philippines for the same time period. The events were found in the original USE catalogs, but as small local earthquakes feit on the Philippine Islands.

Some of the errors, when corrected, merely place the epicenter at the nearest trench, but some move them far from their true origin. Another USE epicenter, Event 1267, was listed at (0.00°, 11O.500W) but was actually a local Utah earthquake at (40.000N, 1I0.500W). A 1951 ISS event (1315) was put in the Northcentral Pacific at (39.200N, 171.500E) but was actually a Hindu Kush earthquake at (39.200N, 71.500E), and four other "Northcentral Pacific" earthquakes (Events 1I56, 1317, 1318, 1319) listed in the Northem Pacific near (38°N, I 76°E) were actually New Zealand events at (38°S, 176°E) which were missing minus signs on the latitudes.

In the case of Event 1726, the earthquake is credited to WEL but does not appear in the Wellington (or associated) bulletins (B. FERRIS, pers. commun., 1989); it is therefore assumed that there is an error in the listing of the date (and perhaps in the location as weil) and the event is for all practical purposes irretrievable.

There is one other segment of this category, though less populated, which consists of blatant errors in the original location procedure. An example is Event 1320 (April 9, 1967), which was given a preliminary determination by LASA, later located by the ISC (see Figure 7) and finally compiled into WALKER (1989). The LASA array received the arrival at 18:01:45.1 and located the event in the North Pacific at (36°N, 172°E) with an origin time of 17:51:39, giving it an epicentral distance of 59.72°. We surmise however that the earthquake never occurred because at 17:41:55 there was an event in the Banda Sea (7.02°S, 129.66°E; h = 1I8 km; mb = 5.3). The distance to LASA from this event was 1I8.12°, and taking into

Vol. 135, 1991 Intrap1ate Seismicity

a. LAD Apr 9d 17h 51m 39s. 3& 'N 172 'E. Mag=5.0

(&11) North Pacific Ocean I (39) Pacific Basin 318 The distances and residuals correspond to the position and time given by LAO

LAD Lasa Centre 59.72 52 i 180145.1 .2.4

319 Apr 9d 17h 52m 10±lls. Epicentre 39'.2 .7'North by 1&3'.I.O'East. Oepth= O.OOOOR or 33 km. 50=& .48s on 14 obs. 10 (&Ill North Pacific Ocean I (39) Pacific Basin

T F 0 Tonto Forest && .15 &4i 180240.3 WM 0 Wichita Mountains 74.54 58 i 1803 53.& FA Y Fayettevilie 7&.54 54 i 18 04 03.9 5 C H 5cheHervi Ile 7T .05 27 e 18 04 02 LN 0 London 79.50 42 e 18 04 18

5 I C 5even Islands 80.7& 30 e 18 04 17 5 F A 5even Falls 81.47 34 e 18 04 20 MRG Morgantown 82.5& 44 i 1804 28.30 C BM CJribou 82.78 33i 180425.4 B N H Berlln (N.H.) 83.23 3& i 1804 28.8

MIM Milo 83.71 34 i 1804 29.2 EMM East Machias 84.78 34 i 18 04 31 .& WES Weston 84.89 37i 18 04 J2.0 KER Kermanshah 8& .08 312 e 18 04 4&

b. Times from the Banda Sea Earthquake

7.02°5 129.66°E d = 118 km O.T.= 17:41:55.0

Arrival Distance Residual Station Type (0) (sec)

LAO pp 118.10 0.8 TFO ? 117.87 WMO SKP 127.81 0.1 SCH SKP 130.48 -1.3 FAY SKP 130.61 0.1 SIC SKP 134.79 -1.0 LND SKP 134.99 -0.6 SFA SKP 136.30 -2.7 MRG SKP 137.99 0.6 CBM SKP 138.12 -2.6 BNH SKP 138.32 0.2 MIM SKP 138.54 0.0 EMM SKP 139.50 -0.2 WES SKP 140.14 -1.5 KER S 87.84 -4.7

Figure 7

-12.1 .10.4 .9.3 .4.8 .7.2

-0.4 -I .2 .. 1 .4 -2.5 -I .4

-3.4 -& .5 -&.7 .1 .3

0.& 0.8

14 obs

297

The "ghost" event of April 9, 1967 (Event 1320). (a) The LASA and ISC locations for an event that most likely never existed. (b) The residuals for the stations used in (a) if the arrivals are interpreted as

PP, SKP and S waves from the earlier Banda Sea earthquake.

account the depth of the Banda Sea earthquake, the LASA arrival has the exact time and distance expected for a pp arrival from the Banda Sea event. However, since LASA located using the slowness vector across a small array, there was no way for them to distinguish between the two.

This ghost event was then located by the ISC at (39°N, 163°E) and encounters similar problems_ The ISC lists 13 P arrivals at North American stations, mostly in


New England and Northeastern Canada, and one 10 Iran. Of the 13 North American stations, all except TFO are fit excellently by SKP arrival times from the Banda Sea earthquake, at epicentral distances of 128° -140°, a range in wh ich P diff

is small but SKP can be a very significant arrival (CHOUDHURY, 1973). The Iranian station KER is nearly equidistant from the Banda Sea and ghost Pacific events, and the alleged P arrival time from the latter would fit the S time for the former. This case could be included in the category of events relocated to plate boundaries but unlike the cases where locations are contaminated by arrivals from other events or mislabelled phases, this earthquake never occurred.

Code 15: A large number of announced or presumed nuclear explosions took place in the Pacific Basin since 1946. Some are listed in computerized and other epicentral catalogues, but this dataset is far from homogeneous, with many events missing while officially announced as nuclear tests by the V.S. Government (ANONYMOUS, 1989). Actually, we failed to recognize a rational algorithm for the identification andJor elimination of nuclear tests (presumed or announced) on the NEIC tape. In particular, neither a threshold in yield or seismic magnitude, nor the nature of the event (atmospheric, underwater or underground) can explain it, even over periods of time as short as a few months. It is thus suggested that the policies of the various seismological agencies towards reporting nuclear tests varied over the years, or perhaps even were occasionally nonexistent or random. As a result, the dataset on the NEIC tape is confusing and could be misleading. We have chosen to compile in Tables I, 6, 8 and 9 events listed in various seismological catalogues and known or believed to be nuclear tests. These tables are intended as a warning to the seismological user of these catalogues (primarily the NEIC tape), and do not pretend to be a complete listing of nuclear explosions in the Pacific.

Seven Bikini and Enewetak events (945-951) listed on the NEIC tape were officially announced as nuclear explosions (ANONYMOUS, 1989).

Finally, Event 1269 (May 14, 1955; 28.9°N, 126.2°W), part OfWALKER'S (1989) dataset, is an announced underwater V.S. nuclear explosion (ANONYMOUS, 1989). We also include in Table 8 Event 1158 (May ll, 1962; 31.2°N, 124.2°W); because of its unusual location, this announced nuclear test could be a source of confusion.

In French Polynesia, the NEIC tape includes 61 events in the vicinity of Mururoa and Fangataufa atolls, 42 of which are flagged on the tape as presumed nuclear explosions. All 61 feature a computed origin time no more than 5.2 s away from an even minute, and all occurred during daylight hours (07:28 to 15:30 GMT-9). On the basis of these characteristics, and also for several of them of strong mb :Ms anomalies, we believe the remaining 19 are also explosions, and list all 61 as such in Table 1. Once again, the criteria for identification of events as explosions on the NEIC tape are unclear.

Conversely, the tape lists Event 126 at Polynesian si te TV-I0 on April 16, 1982 as an explosion, with a zero depth ("assigned by Geophysicist"). This is most unlikely, since the VSGS epicenter (equivalent to the ISC's) is more than 550 km from


the cluster of presumed explosions at Mururoa, and 200 km from Puka-Puka, the nearest island. Rather, this epicenter plots on the Eastern flank of the Northern Tuamotu chain, and complements the alignment (TU9-GBI-GB2-GB3-GB4) described by OKAL et al. (1980). We regard this event as a genuine earthquake.

5. Results: Statistics

Of the total of 894 events (not related to swarms) that we extracted from reporting agencies and examined, 406 were considered to be genuinely intraplate (including three that relocated to swarms), with the breakdown between categories given in the previous section. This represents an intraplate percentage of 45.4%. In other words, less than half of the events reported that were not associated with localized swarms were actually intraplate earthquakes. The locations of the 894 original bulletin listings are shown in Figure la and the 406 remaining intraplate epicenters are shown in Figure I b.

GENUINE INTRAPLATE

403

~ '"

PLATE BOUNDARY EVENTS

405 304 513

E-< r.Ll

5 36 Z U) Blatant Errors 41 <t: 0 E-< Z <t: 34 39

I

Q Too Little Info. 73 ~ ti3 ~ Nuc1ear Tests 70

69 g:j ....l <t: ~

~

SWARMS 838 SWARMS

(Non-Walker) 815

Figure 8 Chart of the repartition of the dataset between various sources (Jeft and right), and according to the

resu1ts of the processing (center).

300 Michael E. Wysession et 01. PAGEOPH,

The distribution of rate of success was fairly even amongst the reporting agencies. For example, of the 405 events available to us as part of the WALKER (1989) dataset, 194 were found to be intraplate (47.9%), though if we do not count the 24 earthquakes he includes that were only a small part of the Line !slands, Tahiti and Gilbert Islands swarms, then the WALKER (1989) set yields 170 intraplate out of 381, apercentage of 44.6% which is nearly identical to that of the total set. These results are shown schematically in Figure 8, where we show the final numbers and their sources. The sources have been divided into those events which were taken from WALKER (1989), and those that were from the NEIC tape and elsewhere. Those events that were found in both references are therefore included on the WALKER (1989) side, though as mentioned earlier, the overlap was small.

Expectedly, there was a noticeable difference between historical and modern events as to the rate of success for intraplate verification. Of the 340 historical events (before 1963), only 71 (18.5%) were actually intraplate. The majority, 105, relocated to trenches. Of the succeeding 554 modem events 60.5% were found to be intraplate.

Number 0/ Events with Time

The results as a function of time as shown in Figure 9. Figure 9a shows the total references to Pacific intraplate earthquakes between 1913 and 1988, with actual intraplate events shown in Figure 9b and plate boundary events shown in Figure 9c. The bimodal distribution seen in Figure 9a is an interesting feature, and is an artifact of the evolving location capabilities of recording agencies since the two major peaks also appear in Figure 9c. Through the late 1920s there was an increase in the number of reported intraplate events as the increase in stations world-wide allowed for an increase in the number of total events recorded. These do not represent an increl}se in actual intraplate activity, however, which as is shown in Figure 9b remains constant. For the next decade the number of reported events decreases and remains low during post-war depression. This actually mirrors very weH the reported world-wide seismicity, as contained in the NEIC tape, where we see seismicity increase until the mid-1930s and then decline, reaching a low in 1945 which is one-third the number from 1935. This pattern of an increase and then decrease in the number of reported intraplate events may therefore not only be a factor of the accuracies in locating Pacific events and the amount of Pacific activity but mayaiso be simply expressing the inftuence of world events (principally World War 11) on the activity of seismological observatories.

In the early years location accuracy significantly affected the number of reported events. In 1917 over 10% of global seismicity as listed by the NEIC, 7 out of 65 events, is placed in the Pacific Basin. The following year 17 of 431 are reported as Pacific intraplate events. Only one from both of these two years is reliably intraplate. By the 1950s only 0.5% of global seismicity was listed as being Pacific intraplate, and this continues until the 1980s when this drops to around 0.15%. It

Vol. 134, 1990

50

40

Vl I- 30 Z I.J.I > 20 I.J.I

"11:

10

0

30

Vl I-Z 20 I.J.I > I.J.I

"11: 10

0

15

Vl I- 10 Z I.J.I > I.J.I

~ 5

o

Intraplate Seismieity 301

a. TOTAL L1STINGS

1980

b. INTRAPLATE EQS

C. PLATE BOUNDARY EQS

Ih~ hj ~~n ~ ~ r ~ ~~ 1920 1940 1960 1980

TIME (YR)

Figure 9 Histograms, by year, of the data used and analyzed. (a) Total event listed as intraplate by reporting bulletins. (b) Reloeated and verified intraplate earthquakes. (e) Earthquakes that have relocated to plate

boundaries. Note: the vertieal scales are not the same.

IS mteresting to note, however, that until 1963 the number of reliably intraplate events remains nearly constant at -1.5 per year, implying that the magnitude detection threshold is not significantly lowered during this time. After 1963, we do see an increase in intraplate seismicity due to the emplacement of Pacific arrays. The large peak that appears in Figure 9b in the mid-1970s is the increase in Polynesian seismicity made detectable through the operation there of many local stations. Excluding this large peak, since the start of the 1970s the number of actual intraplate events has averaged around 10 per year.


Residuals

Figure lOa shows a plot for each event of the standard arrival time residuals from all listed P and S arrivals. Even though this was somewhat arbitrary, as arrivals that we considered blatantly wrong were removed and this value is also a function of the number of stations used, it still gives a good indication of the quality of the data over time. A continual improvement in either picking arrivals or clock maintenance is evident, especially from the late 1920s to the early I 940s.

,..... CI)

'-./

C :J

"C 'Vi Ql

oc "C ... c

"C c C

Vi

~ 'e

,..... Cl Ql

"C '-./

C .12 C u 0

Cii ... -0

Ql "C

.2 'e Cl C

::::!:

100

80

60

40

20

0

a. INITIAL BULLETIN STANDARD RESIDUALS

o

o

o

o

8 0 o

o

o 0° 0 0 0 0 0

o 00 0 0 0 o 0 0 0 0 o~ orSl 0

o

o'\, 8~ 0 toö> 0 0 o '!B 0 o~ 0 'b Q.. 0 0

00 0 qs cocJI 0 ~ 0 o. ~ 0 0 0 0

o 0 N. ooo<t\l_~oo~ 0 0 o'~ o~s.°Jii.Y~--.... o c9 000 0 0 'bBo ~ 0 q,o 0 0 :;tJ'!~~-oao'6

1910 1930 1950 1970 1990 Time (yr)

b. LENGTH OF RELOCATION VECTORS 40

o 0

30 0

0 0

20 o 0

0 o 0 0 <9 0

0 0 0 0 0 0

0 0

0 <90 0

10 q, 0 c!i\' 0 0

'" 0

cPoO°~J" 0

°o$.Ji 1A:i 0 0

0 oill (90'

1910 1970 1990

Figure 10 (a) Plot of the initial standard residuals for bulletin locations. The value is taken from all available P and S arrivals that are not blatantly in error. The quality of the data shows a steady improvement through time. (b) The length of the relocation vectors (in degrees). Open circles are individual values and the solid line is a 3-yr moving average through them. The trend of the average starts from 6° for early events and decreases to almost zero at the present. Note the significant deterioration of solutions

during WWII.


Relocation Vectors

The distance that the epicenters move as a result of our relocations is shown in Figure JOb, with the open circles marking individual values and the solid line representing the average change in distance using a three-year moving window. After 1963 the averaged curve is not as meaningful because in cases where the epicenter moves less than 0.5° we have kept the reporting agencies' epicenters. Nonetheless, there is an evident quasi-linearly decreasing trend in the average distance by wh ich the original epicenter was in error, starting at 6° be fore 1920 and decreasing to nearly zero by 1988. A plot of the median change in distance would be similar, starting at 6° (nearly 700 km) in 1915 but decreasing to less than OS earlier, by 1970.

The relocation vectors in Figure 11 show magnitude and direction of the relocations separately for the historical and modern epicenters. Only changes of more than 0.5° are shown. Figure Ila is visually dominated by the blatant errors that were due to typographical bulletin errors, such as the cluster of Philippine events with the longitudes reversed and the New Zealand earthquakes with the latitudes flipped. Figure 11 b shows an attractive force of "slab pull" in the Tonga and JapanjKuril regions, where nearly all events listed as intraplate actually relocate to the trenches. It is remarkable that on both figures, the plate boundaries

RELOCATION VECTORS 1913-1962 1963-1988

60,-________________________ -. 60,-________________________ -.

~. -~ T r .~

{ ... "'-

l/~ . > :;.1< ~

'< • <.1 " ~ \

-68~ __________________ ~~ __ ~ -68 115 -65 115 -65

Figure 11 Relocation vectors for all earthquakes relocating more than OS from their bulletin listing. The tail of each arrow is the original listing, and the head the relocated epicenter. The dataset is categorized into historical (pre-1963) and recent events. Note that the historical dataset is domina ted by c1erical errors affecting the sign of one coordinate. For c1arity, the 1951 event relocating into the Hindu Kush is simply shown as an arrow exiting the box. The recent dataset is dominated by small earthquakes relocating to

nearby subduction zones.

304 Michael E. Wysession el 01. PAGEOPH,

are weil delineated even though all events were initially considered to be intraplate. This is true not only for the trenches, but also for the ridge and transform systems that separate the Cocos, Nazca, Pacific and Antarctic plates.

Focal Mechanisms

Table 14 lists all focal mechanisms available in the literature for genuine intraplate events listed in our catalogues, with orientation conventions after AKI and RICHARDS (1980). Sources are various individual studies listed in Table 14, and for more recent events, the Harvard moment tensor solutions (DZIEWONSKI et al., 1983a,b,c; 1984a,b; 1987a,b,c; 1988a,b,c; 1989a,b,c).

While it is conceivable that some of the older mechanisms could be partially or fully constrained from relatively sparse datasets of old seismograms (STEIN et al., 1988; JIMENEZ et al., 1989), such studies are outside the scope of the present paper.

Tab1e 14 Focal Mechanisms AWlilable for Inlraplale Evenls

Date Epicenter Index Focal mechanism Moment Reference

DMY oN oE eHO) Ii (") Ä. (") (1(f5 dyn-<:m)

Regular Evenls

30 6 1945 16.60 -115.80 1204 80 40 270 11 a 15 12 1947 -58.76 -159.20 421 350 30 90 b 22 11 1955 -24.27 -122.77 408 270 70 210 b 14 9 1963 -33.60 -126.70 414 29 67 154 b 06 3 1965 -18.40 -132.85 218 84 69 204 25 11 1965 -17.10 -100.20 1519 22 46 113 0.18 d 18 9 1966 -18.40 -132.86 219 103 69 206 28 4 1968 44.80 174.60 1101 330 65 70 0.2 09 5 1971 -39.78 -104.84 1702 16 60 90 25 5 1975 -18.40 -132.92 225 115 63 187 OS 2 1977 -66.45 -82.58 1714 1 58 82 4.4 g 30 1 1978 -16.23 -126.94 230 187 85 169 c 07 11 1979 -62.58 -72.91 1715 353 48 121 0.036 h 2011 1979 -26.57 -138.84 211 139 47 109 0.097 h 03 6 1981 -S.30 -175.20 757 331 44 123 0.047 30 9 1981 -4.80 -112.oI 404 107 75 25 0.22 j 04 10 1981 -4.67 -111.86 405 302 19 71 0.05 i 22 3 1982 6.60 175.10 885 300 57 56 0.46 k 15 10 1982 32.84 -125.82 1117 282 79 17 0.09 k 03 6 1983 -50.20 -174.50 506 301 51 354 0.11 1 11 3 1984 -26.65 -108.51 1528 211 68 250 0.14 m 19 11 1986 0.30 169.70 763 118 79 146 0.10 n 06 7 1987 -27.00 -108.29 1529 38 50 249 15 0

08 7 1987 -26.97 -108.16 1530 51 42 260 5.5 0

07 3 1988 41.67 152.22 1007 350 54 72 0.22 p 22 9 1988 23.81 -167.20 1302 44 78 117 0.04 q

Swarm Events

29 7 1968 -7.47 -148.16 Reg. A 232 85 191 06 8 1969 -7.40 -148.19 Reg. A 241 75 199 19 1 1973 (A) -7.36 -148.24 Reg. A 253 75 195 19 I 1973 (B) -7.31 -148.27 Reg. A 265 80 211

05 1 1978 -20.80 -126.94 Reg. C 56 50 104 0.07 25 7 1978 -20.76 -126.97 Reg. C 152 90 0 31 7 1983 -20.13 -126.93 Reg. C 150 40 270 0.54

29 3 1976 3.929 -85.88 Cooos 208 77 183 11.3


Table 14 (continued)

Focal Mechanisms Available for Inlraplale Events

Date Epicenter Index Focal mechanism Moment Reference

DMY oN "E . (") li (") ).. (") (1<YS dyn<m)

12-17 61968 -O.2S ·91.5 Gaiap. 335 47 247 0.6-1.2 u 24 21988 -0.507 -91.653 Galap. 185 43 100 0.18 p 24 21988 -0.427 -91.627 Gaiap. 16 45 117 0.04 p 20 51988 -0.493 -91.668 Gaiap. 320 38 358 0.10 v 13121981 -3.45 177.55 Güllen 167 52 129 0.05

7 1 1982 -3.37 171.52 Güllen 12S 60 120 0.81 w 20 11982 -3.36 177.38 Güllen 351 81 172 0.22 k 15 21982 -3.43 177.38 Güllen 2S3 86 0 1.10 w 23 21982 -3.39 177.37 Güllen 78 56 26 0.20 k 16 31982 -3.20 177.38 Güllen 140 60 138 0.97 w 22 31982 -3.35 177.53 Güllen 12 2S 158 0.12 k 27 3 1982 -3.42 177.60 Güllen 150 37 126 0.09 k 14 41982 -3.40 177.52 Güllen 277 23 51 0.05 k 18 41982 -3.43 177.47 Güllen 121 16 95 0.12 k 22 41982 -3.46 177.61 Güllen 133 49 122 0.09 k 28 41982 -3.42 177.57 Güllen 281 38 (ß 0.15 k 17 5 1982 -3.46 177.54 Güllen 277 41 47 0.08 k 23 5 1982 -3.35 177.40 Güllen 120 54 120 0.72 w

7 71982 -3.2S 177.47 Güllen 281 50 26 0.08 k 11 7 1982 -3.22 177.55 Güllen 2 49 141 0.05 k 13 7 1982 -3.28 177.55 Güllen 300 32 95 0.10 k 15 71982 -3.37 177.62 Güllen 2S9 2S 36 0.10 k 15 71982 -3.39 177.54 Güllen 142 44 132 0.09 k 5 81982 -3.38 177.66 Güllen 316 33 100 0.04 k 9 91982 -3.27 177.57 Güllen 261 35 44 0.07 k

13 91982 -3.51 177.58 Güllen 279 48 34 0.17 k 14 91982 -3.47 177.70 Güllen 93 26 80 0.10 k 16 91982 -3.33 177.60 Güllen 72 53 354 0.11 k 26 91982 -3.44 177.66 Güllen 106 43 101 0.03 k 3101982 -3.48 177.66 Güllen 2S4 48 41 0.04 k 5101982 -3.49 177.69 Güllen 272 32 38 0.07 k 9101982 -3.44 177.76 Güllen 116 43 106 0.05 k 9 10 1982 -3.47 177.70 Güllen 60 15 56 0.06 k 8 12 1982 -3.37 177.57 Güllen 32S 41 54 0.03 k 9121982 -3.48 177.62 Güllen 124 42 114 0.16 k

31 12 1982 -3.56 177.73 Güllen 284 32 48 0.05 k 21 11983 -3.53 177.67 Güllen 163 43 144 0.12 x 31 11983 -3.39 177.62 Güllen 283 38 76 0.19 x

5 21983 -3.37 177.67 Güllen 130 36 95 0.19 x 8 31983 -3.48 177.63 Güllen 118 27 96 0.11 x

05101984 20.10 -116.ot So. Baja 60 45 290 0.14 a 02121984 20.36 -115.77 So. Baja 50 55 270 4.9 a 28 51986 19.96 -115.88 So. Baja 40 50 280 0.28 a

References: a: WIENS and ÜKAL (1987); b: ÜKAL (1984); c: ÜKAL el al. (1980); d: MemIOUREN (1971); e: STEIN (1979); f: FORSY11I (1973); g: ÜKAL (1980); h: DmwONSIO el al. (1987b); i: DmwONW el al. (1988b); j: Focal mechanism ÜKAL fI984), Moment DmwONSIO el al. (1988b); k: DmwONSIO el al. (1983a); 1: DmwONSIO el al. (1983c ; m: DzmwoNSIO el al. (1984b); n: DmwONSIO el al. (1987c); 0:

DmwONSIO el al. (1988c); p: DzmwoNSIO el al. (1989a); q: JIMENEZ el al. (1989); r: DmwONSIO el al. (1987a); s: WIENS (1985); t: average of WIENS and STEIN (1984) and BERGMAN and SOLOMON (1984); u: KAUFMAN and BUltDICK (1980) (21 events wilh similar mechanism); v: DmwONSIO el al. (1989b); w: LAY and ÜKAL (1983); x: DmwONSIO el al. (1983b).

Magnitudes

Magnitude values reported in Tables 1-13 were compiled from the reports of various primary agencies. With regard to several available magnitudes, the largest figure was kept. Available seismic moments, principally from the Harvard CMT solutions, but also from a number of individual studies, are listed as part of the focal mechanism data, in Table 14.


In our previous study, OKAL (1984) noted an absence of magnitude 6 or larger events in that portion of the Pacific plate genera ted at the Old Farallon, prior to the reorientation of spreading. In contrast, other oceanic plates, and those portions which were not involved in the reorientation, feature larger intraplate earthquakes, (commonly M = 6 and occasionally M = 7). OKAL (1984) interpreted this pattern as a consequence of greater vulnerability of the plate in a geometry where stresses are oriented at an angle to the lithospheric fabric. Our present results fully uphold this observation: the only earthquakes with M ~ 6 in our dataset are located either at the fringes of the Pacific plate (e.g., the 1984 swarm South of Baja California), in the Southern portion of the plate (e.g., Events 408, 420, 421), or in the Cocos, Nazca and Antarctic plates (e.g., Events 1351,1515,1535,1702,1714).

6. Geographical Discussion

It has long been suggested that the state of stress in oceanic intraplate regions is controlled largely by the gravitational sliding of the lithosphere as it ages ("ridge-push") and to a lesser extent by drag forces on the lithosphere-asthenosphere boundary (FORSYTH and UYEDA, 1975; SBAR and SYKES, 1977; RICHARDSON et al., 1979; OKAL, 1980). Where and how this stress is released, however, can be determined through the knowledge of the distribution of seismicity, especially if there are correlations with bathymetric features (OKAL, 1983).

SYKES (1978) discussed the possibility of intraplate seismicity occurring along pre-existing zones of weakness such as fracture zones and fossil spreading centers. This has been supported by studies of oceanic intraplate regions. BERGMAN and SOLOMON (1980) found that larger earthquakes were often associated with old fracture zones. OKAL and BERGEAL (1983) have suggested that there is increased seismicity on the boundary li ne of lithosphere genera ted at the old Farallon Ridge before it jumped and reoriented itself along the present East Pacific Ridge, and STEIN (1979) has also discussed ancient plate boundaries such as the Emperor Trough as preferential sites for seismicity. Ancient traces of hot spots have also been proposed as possible zones of weakness.

A major tectonic fabric in the Pacific is the presence and often high density of seamounts. While a correlation betwen seismicity and seamounts or islands is difficult-seamount dimensions can be smaller than the inherent uncertainties in the earthquake epicenters-its occurrence may signify the release of very local as proposed to plate-wide stress. Such seismicity may be the expression of isostatic compensation of oceanic islands or the injection of magma associated with volcanism. Indeed, seismic activity has been weIl documented in association with volcanism in the Polynesian Islands (TALANDIER and OKAL, I 984a, 1987).

What will foIlow here is a discussion, region by region, of the possible correlations between Pacific intraplate seismicity and the tectonic settings it occurred in.


The latter includes bathymetric features, such as fracture zones, fossil ridges, seamounts and hot spot traces, as weIl as seafloor age, determined through magnetic lineations.

In order to classify our dataset and arrange it in a manageable fashion, we proceeded with the regionalization described on Figure 6 and Tables 1-13. It must be realized, however, that while small-scale geological features, such as seamounts and short fracture zones, lend themselves weIl to such classifications, many largescale bathymetric expressions transcend any attempt to regionalize the Pacific Basin. As a resuit, our geographical discussion of the seismicity will not strictly adhere to the chosen regionalization, but rather proceed along a very broad clockwise spiral, starting with Polynesia.

We concentrate in this section on a description of the resuiting intraplate seismicity in the context of local geology. Technical details regarding the quality of the individual relocations, and a documenta ti on of the casuaity events, can be found in Appendix B. In areas where seismicity tends to cluster at identifiable sites, we have followed our earlier practice (OKAL et al., 1980; OKAL, 1984) of assigning location codes to these various sites. In some areas, the level of seismicity and its scattered character do not warrant this effort, and no codes have been assigned.

Polynesia

Most of the sites in Polynesia have been described in detail by OKAL et al. (1980). The present discussion will emphasize the new sites, and in particular their possible relationship with features recognized earlier. OKAL et al.'s dataset is compiled and updated to 1988 in Table I, with the exception of Regions A and C, whose swarms are described separate\y in Appendix A. Figure 12a shows the updated seismicity in the area. We keep these authors' coding scheme for the various seismic locations involved. A total of 136 genuine intraplate earthquakes were recognized, most of them located by the Polynesian network. In keeping with our exclusion of the seismicity in the immediate vicinity of Hawaii, we do not include in our da ta set epicenters located in the Tahiti-Mehetia area, defined as "TM" in OKAL et al. (1980), and outlined on Figure 12a. This results in the elimination of 4 events selected by WALKER (1989) out of the more than 70 listed in OKAL et al. (1980).

It is remarkable that activity in Polynesia has been extremely low during the 1980s: Event 127 defines a new site (TU-lI) in the Northeastern corner of the Tuamotu Islands. Renewed activity took place at a very low level Northwest of the Society Islands (Events 143, 145 and 146), two new epicenters were defined Northeast of the Austral Islands (AU-7 and -8, Events 212 and 213), and one North of the Gambier Islands, along the northern flank of the Marutea-Acton group (GB-IO, Event 233). Finally, one event (214, on October 2, 1988) was wide\y


feIt in the Marquesas Islands, and located by PPT in the center of the archipelago. We should recall, however, that volcanoseismic activity was extremely intense during 1981-1985 in the Tahiti-Mehetia area, and at Macdonald Seamount during the whole decade.

The two new sites SC-4 and SC-5 North of the Society Islands (Events 145 and 146) define, together with SC-l and SC-2, a cluster grossly parallel to the axis of the island chain. In the Tuamotu and North-of-Gambier regions, and as noted by OKAL (1984), the single 1982 event at TU-lO (number 126) is remarkably aligned with the lineation running from TU-9 (Event 125) to the cluster at GB-4 (Events 219-230), which can indeed be extrapolated to the swarm site GB-5, part of Region C. While the exact origin of this lineament is speculative, it also coincides with the 4000-m isobath (with the exception of Region C). It is remarkable that the two sites immediately North of the Gambier Islands (Events 232 and 233) are also in the immediate vicinity of the same isobath.

z

w o ::J I-..... Ier ....l

POLYNESIR

G

~ • 0

P'

-<?" • Cl .(J

""" 2

.234

• 411

0

-5

-10

-15

-20

-25

~~------~~----------~----______ ~~~-30 -160 -150 -140 -130

LONGITUDE (El

Figure 12a Map of intraplate epicenters in Polynesia. Numbers refer to individual indices in Table I. Triangles identify the swarm site at Regions A and C. Bathymetry is contoured at 1000 m intervals. The stippled quadrangle is the Tahiti-Mehetia area, as described by OKAL el al. (1980), and excluded from the present

study. Two-Ietter codes follow the scheme of these authors.


FOCRL SOLUTIONS FOR POLYNESIR

REGION R

29 JUL 1968 06 Aue /969 /9 JAN 1973 (A) 19 JAN 1973 (8)

~ ~ ~ ~ REGION C

05 JAN 1978 25 JUL /978 3/ JUL 1983 23/ -- 30 JAN /978

~ ~ 0 ~ OTHER SITES

2/9 -- 06 MAR 1965 220 -- 18 SE? /966 226 -- 25 MAY /975 2// -- 20 NOV 1979

Q ~ ~ ~ Figure 12b

Focal mechanisms available for Polynesia.

Southcentral Pacific

In the Southcentral Pacific and East of Kermadec (Figures 13 and 14) we update OKAL'S (1984) description of the active sites, by recognizing a few remarkable features:

• Four events (425,537,538,539) are located at most l.so from the Louisville Ridge. These earthquakes could represent tectonic processes linked to prolonged activity, reactivation, or differential subsidence along this feature, generally interpreted as the wake of a hotspot presently located in the vicinity of the Eltanin transform (HAYES and EWING, 1971). A precise interpretation remains however speculative in the absence of any knowledge of lower-magnitude seismicity. A remarkable North-South lineament between longitudes 130 and 125°W can be followed from 34°S (Event 414) all the way into the Northern Pacific (see discussion below). The maximum distance between sites is 7°. This "1300 W" seismic line, recognized by WALKER ( 1989), runs through the swarm site at Region C. Because of the numerous fracture zones in that part of the Pacific Basin, this North-South lineament is neither an isobath (ocean depths range from 2500 to 5500 m), nor an isochron (the various epicenters being spread out from Chron 6 (Region C) to Chron 18 (at SP-3)). Upon intersecting the Oeno-Ducie-Crough

310

z

UJ C ::l I--I-5


SOUTHCENTRRL PRCIFIC BRSIN

(Except POlynesial

421 -- 16 DEC ltU7

-170 -160 -150 -140 -130 -laO LONG ITUOE (E l

Figure 13

PAGEOPH,

404 -- 30 SEP 1981

405 -- IU OCT 1981

408 -- 22 NOV 1956

414 -- 14 SEP 1963

Map of intraplate epicenters for the Southcentral Pacific, with available focal mechanisms. Numbers refer to individual indices in Table 2. Bathymetry contoured at 1000 m intervals.

chain, the lineament branches out Eastward, the new epicenters SP-7 and SP-8 (Events 412 and 413) providing some continuity with Sites IP-4, IP-5 and IP-6 (Events 408-410), recognized by OKAL (1984).

In a previous study, OKAL and BERGEAL (1983) had suggested that seismicity might be preferentially released along a line of age discontinuity related to the Miocene jump of the Farallon ridge. This line does run between 125 and 1300 W at latitudes 1O-38°S, but is offset Eastwards further North. It should also be noted that three events used by OKAL and BERGEAL (1983) are proven casualties in the present study: Events 432, 445 and 1267.

Campbell Plateau to Samoa

The region east and south of New Zealand has a large amount of documented intraplate seismicity due to the complex geophysics of the region and the extensive network of local stations. Out of 76 intraplate events listed for this region, 41 were

Vol. 135, 1991

w o ::l I--I-a: ...I

Intrap1ate Seismicity

CRMPBELL PLRTERU TO KERMROEC

-50

-55

~~ __ ~~ __ ~ __ ~ ______ ~-60 -170

(El

Figure 14

-40

506 -- 03 JUN 1983

Same as Figure 13 for Campbell-to-Kermadec Region. See Table 3.

3\1

eonsidered genuine, and for many of these our reloeations did not differ signifieantly from the epieenters determined by Wellington (Figure 14; Table 3).

Mueh of the seismieity in this region is loeated on the Campbell Plateau and Chatham Rise, both of whieh are of eontinental origin and possess a eomplex history (FLEMING, 1970): They are thought to have been former extensions of a N-S trending Triassic geosyncline that were swung cloekwise about the South Island during the Early Cretaeeous, based on the eontinuation of geologie facies from the bottom of the South Island to the Chatham and Bounty Islands (FLEMING, 1970), and the eontinuation of the Stokes Magnetie Anomaly on the South Island out into the Campbell Magnetie Anomaly System (DAVEY and CHRISTOFFEL, 1978). The Campbell Plateau was then separated from and moved south relative to the Chatham Rise, with 330 km of right-Iateral displaeement determined by DA VEY and CHRISTOFFEL (1978) to have occurred along the Campbell Fault, whieh trends northeast. The seismieity eorrelates weIl with these lineations.

Two earthquakes, Events 502 and 503, oeeur at the same position just east of the Auekland Islands, at the junetion of the southern end of the proposed Campbell Fault and the western end of a large, positive magnetie anomaly. This magnetie anomaly terminates at the Pukaki Saddle, a loeation where seismie data has shown


a sediment-filled faulted graben (HOUTZ, 1975); this is the site of Event 506 (mb = 5.2), an earthquake whose CMT foeal solution is shown in Figure 14. Two other events (504-505) reloeate to the same loeation on the northernmost Campbell Magnetie Anomaly (the equivalent of the lunetion Magnetie Anomaly), whieh extends from northeast of the Auekland Islands to south of the Bounty Islands. There is also one earthquake (501) at the extreme southern end of the Campbell Plateau, along the Campbell Spur.

Seismicity is eonspieuously absent from the Bounty Trough, but is very aetive along the Chatham Rise, just north of it. The North Island of New Zealand represents a transition zone from the Kermadee Treneh to the South Island's Alpine Fault System, and though there has been Paeifie oeeanie lithosphere subducted under the North Island at the Hikurangi Treneh, the entire island is displaying internal deformation and the exaet loeation of the plate boundary is not evident (MOLNAR, 1988). We have therefore chosen our 2° buffer to extend from the eurrent eoastlines of New Zealand, and we find that the seismieity beyond this point is dominated by two large clusters.

Events 510-524, with loeal Wellington magnitudes of 3.8-4.5, occur at the western end of the Chatham Rise and are separated into two features. The first, eontaining most of the epieenters, lies just beyond our 2° buffer. Though this appears as a linear feature, extending northeast from the rise into oceanie erost, this is misleading as there is mueh seismieity direetly to the west of here. Particularly signifieant is a linear northeast-trending termination to these events which defines a zone of inaetivity between this cluster and the second feature, three events (511, 516 and 520) whieh also lie along a northeasterly trend. These three shoeks also define a previously reeognized fault shown on the Cireum-Paeifie Map of the AMERICAN ASSOCIATION OF PETROLEUM GEOLOG1STS (1981). Further out on the rise, centered around the Chatham Islands, are three more events (507-509).

The seeond cluster, separated by about 150 km, lies directly to the northeast of the first (Events 525-536). These are slightly more energetie than the other cluster, with 4 events of mb ~ 4.5. Unlike the first cluster, whieh is tempo rally eontinuous, these events oeeur in three separate groups: Events 526 (1964), 527-530 (1971-1974), and 531- 536 (1982-1985). They do not correlate with an obvious bathymetrie feature.

There are five more earthquakes, also listed in Table 3, whieh are further to the east. Three (537- 539) loeated on or just beyond the Louisville Ridge, and are discussed earlier. The remaining two are east of the Louisville Seamounts and are both old. Event 540 (lP-20 in OKAL, (1984)) is from 1940, and Event 541 is from 1925.

The Paeifie regions surrounding the Tonga Treneh and Fiji Plateau were found to have very low seismicity. This was surprising, eonsidering that most previous maps of Paeifie seismieity, such as WALKER'S (1989), have displayed a very high density of seismicity in this region. Of the 100 events listed in these regions,

Vol. 135, 1991

z

w 0 => I-

I-CI: ....J


OFF TONGA AND SAMOA

651

lBO -175 -170 -165 LONG ITUDE (E)

Figure 15

-12

-17

Same as Figure 13 for Region East of Tonga and Samoa. See Table 4.

313

delineated in Figures 15 and 16, only 21 were reliably intraplate. The majority (63) of reported events were reloeated to the Tonga Treneh.

The seismieity east of Samoa and the Tonga Treneh was found to be very minimal (Figure 15). Of the five earthquakes that are retained (Events 651-656), four oeeurred before 1930, and therefore have larger error ellipses. Nonetheless, the best found epieenters for four of these (653-656) form a lineation that extends north-south along the 167.5° meridian. We have no simple explanation for this feature, as the distanee from the treneh is too great to invoke a systematie buekling of the lithosphere in the manner of CHEN and FORSYTH (1978) and CHAPPLE and FORSYTH (1979).

Finally, a single historieal event (651) reloeates north of Canton Island, possibly at the eastern end of the Nova Cant on Trough.

Samoa 10 Micronesia

We eoneentrate here on a large region extending from Samoa to the Marianas Are as shown in Figures 16 and 17. This region exhibits substantial seattered seismicity whieh was dominated by the Gilbert Islands swarm of 1981-1984 (LAY and OKAL, 1983, summarized in Appendix A). On this basis it has been suggested

314

z

UJ o ::::l I-..... Ier ..J


S~MO~ - PHOENIX - GILBERT 759-763,.

@ eC!> e

165 170 175 LONGITUOE (El

07 JAN 1982 15 FEB 1982 16 JiAR 1982

~ ~ ~ Figure 16

PAGEOPH,

757 -- 03 JUN 1981

-3 ~

-8 763 -- 19 NOV 1986

23 MAY 1982

~ Same as Figure 13 for Samoa-Phoenix-Gilbert Region. The triangle identifies the Gilbert Island swarm site. See Table 5. The four bottom focal mechanisms are for those events in the swarm studied by LA Y

and OKAL (1983). Thirty-two additional CMT solutions are listed in Table 14.

that a process of incipient subduction is taking place. OKAL et al. (1986) have traced this line of activity from Samoa to the Ralik Fracture Zone, north of Ocean Island. KROENKE and WALKER (1986) have extended it all the way to the Caroline Ridge just east of the Yap Islands.

We will examine the seismicity of the region in the context of these speculative models, and in particular will show that a significant difference in the level of seismicity, as weil as inconsistencies in focal mechanisms, result in a picture both fuzzier and more complex than the simple relocation of the Solomon subduction zone to the North.

Immediately to the North of Samoa were three historical events (751-753) two of which relocated to the Robbie Bank, with the third one (753) 10 east of Nurakita. Further north, in the Phoenix Islands, there are two events (756 and 757), with the latter given a CMT solution that shares the NE-SW compressional axis common to all Gilbert Islands events (Figure 16).

In the SW part of the Tuvalu Basin a single event (755) is located on a chain of seamounts that extend west from Funafuti Atoll. Further west a single historical event (754) relocates to the interior of the Ontong Java plateau.

At the Gilbert Islands swarrn site two earthquakes in 1982 (764-765) became

Vol. 135, 1991

z

w o ::::> I-...... Icr ...J

146 150


MICRONESIR RND MRRSHRLL ISLRNDS

.~ß .~ ,,0 ·882 • . 881.·

885 -- 22 JlAR 1982

160 LONGITUDE (E)

Figure 17

26

21

16

1 -170

315

Same as Figure 13 for Micronesia and the Marshall Islands. On this particular figure, bathymetry is contoured at 2000 m intervals. See Table 6.

part of the swaIlI! when relocated, and there was a third event (758) that relocated north of the site but whose error ellipse incorporates this region.

To the northwest, the Ralik FZ site underwent a new strong event since the studies of KROENKE and WALKER (1986) and OKAL et a/. (1986). The CMT solution for this event (763; mb = 5.3), however, is irreconcilable with the stress orientation at the Gilbert Islands (Figure 16).

Further north, in Micronesia and the Marshall Islands, seismicity is arranged along several linear trends oriented about 307°; we have geographically ordered the 44 intraplate events along these lineaments.

While these trends do not represent a unique way of organizing the seismicity of this region, they form very striking features. Certainly the dominant trend, the subject of the previous studies, is the one that extends from Samoa to the Caroline Ridge. Largely due to the inclusion of the Gilbert Islands swarm this line of seismicity was estimated by OKAL et a/. (1986) to have accounted for 15-30% of the total Pacific intraplate seismic moment budget for the years 1937-1983. This line of seismicity does not indicate a localized active subduction zone as suggested by KROENKE and WALKER (1986), however, because there is no evidence for a trench (OKAL et a/., 1986) and because the trend is only moderately linear. As was discussed above, the seismicity at the Samoa and Ralik locations is very minimal,


certainly compared to the Gilbert Islands activity, and the extension of this trend into the Micronesian region (Events 898-919) is as a zone of diffuse seismicity.

In addition, there are two other distinct features (considerably more linear) that form ESE trends, pointed out by WALKER (1989). One extends from the Ogasawara Plateau (east of the Izu-Bonin Trench) down to the Marshall Islands, following the Jurassic Magnetic Quiet Zone (Events 884-897), and the other (seen across the top of Figure 17) extends from south of the Shatsky Rise through the Mid-Pacific Mountains into the Central Pacific Basin (Events 876-883). The orientation of these trends is the same as the direction of absolute plate motion, determined relative to hot spots. The top and middle lineaments, when taken SE to NW on a Mercator projection, have azimuths relative to north of 308° and 306°, and the absolute motions of the middle of the trends are 307° and 304°, respectively (GRIPP and GORDON, 1989). The earthquakes in these regions, with the exception of Events 877 (mb = 5.2) and 885 (mb = 5.6, with a CMT solution shown in Figure 17), are small and usually only detected by local arrays. The presence of these three distinct belts of seismicity supports the idea that Australian-Pacific subduction is being arrested, but suggests that it has not yet found a new horne.

Off the Coast 0/ Japan and Kuriles

Compared to the Micronesian and Polynesian regions, the northern part of the Pacific Basin displays very sparse seismicity. There are two sites of active seismicity-the Hawaiian Island and the swarm at (200 N, 117°W) studied by WIENS and OKAL (1987; see Appendix A) -and otherwise there is very little. There is one large tract of oceanic lithosphere, with latitudes 0-400 N and longitudes 152-132°W, that does not boast a single recorded epicenter.

This lack of seismicity is especially remarkable near the Japan, Kuril and Aleutian Trenches, wh ich like Tonga have many intraplate bulletin listings (WALKER, 1989). Off the coast of Japan and the Kurile Islands (Figure 18, Table 7) we retained 7 intraplate earthquakes of the 43 that were reported. The largest (Event 1007: mb = 5.9), on March 7, 1988, was given a CMT thrust fault mechanism, shown in Figure 18. Though this epicenter does not correlate with any bathymetric feature, it is interesting to note that a 1936 earthquake (1006) relocates near this site.

Of the remaining 5 events in this region, two occur on the same day adjacent to an unnamed seamount (Events 1001-1002) and two locate to the western flank of the Shatsky Rise. Of the latter, the larger earthquake (Event 1004 on May 6, 1976: mb = 5.1) has a reliable location on the 4000 m isobath. Occurring on the previous day, Event 1003 relocates 2S to the southwest of this, but it was a small event (no magnitude given) with a poor relocation, so it may very Iikely have occurred at the site of Event 1004.


OFF THE COAST OF JAPAN ANO KURILES

50

f 007 -- 07 JIAR f 988

z 45

(I) l.LJ 0 ::J I-.... I-

40 a ....I

35

U1~~ __ ~~ __ ~~ __ ~30

145 150 155 160 1.0NGITUDE (El

Figure 18 Same as Figure 13 for Region East of Japan and Kuriles. See Table 7.

Northern Boundary

The tectonics of the Northern Boundary (Figure 19; Table 8) are domina ted by the Aleutian Trench, and as with the Japan and Kuril Trenches, most of the reported seismicity relocates into the trench. There are 20 epicenters (out of 58) that are reliably intraplate, however, and they are distributed fairly evenly throughout the region, occasionally forming small clusters that corre1ate with bathymetric features.

The largest event in this geographical region was an mb = 5.5 shock in the Emperor Trough (Event 1101), which had a thrust mechanism (Figure 19) determined by STEIN (1979), and had one recorded aftershock (11 02). Event 11 08 is located on a seamount in the Seamount Province south of Kodiak Island, though the other earthquakes in the central part of the Northern Boundary (1105-1107, 1109-1110) are not associated with identifiable seaftoor structures.

Particularly recognizable in this and the Japan-Kuril regions, and eliminated as casualties of the relocation process, are aftershocks of the large interplate earthquakes of February 3, 1923 in Kamchatka, February 4, 1965 in the Aleutians, August 11, 1969 and March 23-24, 1978 in the Kuriles, and June 10, 1975 at the Hokkaido corner (with moments ranging from 8 x 1026 to 1.4 X 1029 dyn-ern). Our failure to document reliable intraplate seismicity among the aftershocks of these large intraplate events means that the oceanic lithosphere at a distance of 2° from

318

z

w o :::::l f-

fa --l

Michael E. Wysession et al. PAGEOPH,

NORTHERN BOUND~RY

55

5C

45

4C

"---=~..:=n...&.::>:i_ 3C 1101 -- 28 APR 1968 1117 -- 15 OCT 1982 -135 -125 -115

Figure 19 Same as Figure 13 for Northern Boundary Region. See Table 8.

the trench does not participate in the coseismic stress relaxation. In this respect, our study yields an interesting result, since the pattern of seismicity observed on published maps of unrelocated earthquakes would lead to the exact opposite conclusion.

Further east, one epicenter is retained from the Alaska Plain, Event 1111, which is in the far northeast corner of the Pacific, but still south of the incipient plate boundary seismicity described by LAHR et al. (1988).

Slightly southwest of the Eickelberg Ridge are the epicenters of two events (1112-1113), statistically identical, that occurred within hours of each other on March 20, 1940. Further to the south, there are two earthquakes that occurred on major fracture zones: Event 1114 relocated to the Mendocino, and Event 1115 occurred on the Murray.

West of Southern California there is a cluster of epicenters that occur near or on the Feiberling Tablemount. Event 1116, with a magnitude of 4.8 (PAS), occurred in 1949 at the northern edge of the Tablemount, the same 10cation as Events 1117-1118 wh ich happened in 1982. The CMT mechanism for the larger of these two was dominantly strike-slip (Figure 19). A 1988 earthquake (1120) was located directly on the Feiberling Tablemount, and one more occurred 3° to the east. These events may possibly be associated with volcanism, which was reported to have erupted in 1972 at a 10cation between Events 1119 and 1120. This report was qualified as "uncertain" by SIMKIN et al. (1981).


West and South of Baja California

The region west of Mexico is a very complicated one (Figure 20), and appropriately there is a significant deal of seismicity that correlates with the seaftoor structures. As listed in Table 9, there are 25 events of the 69 listed that are genuine intraplate earthquakes. The success rate of relocations was actually better than this would suggest, as 15 of the casualties were the mislisted USE epicenters that occurred in the Philippines.

The tectonics of the Pacific-North American-Cocos-Nazca-Rivera plate boundaries have been complex, and the fabric of the Eastcentral Pacific seaftoor displays this. There are many east-west trending fracture zones, several north-south fossil spreading ridges, and countless seamounts. It is interesting, however, that the most visually striking patterns in the seismicity are two north-south lineaments that do not seem to be associated with former tectonic activity: that along the 1300 W meridian, mentioned above, and along the 116°W meridian, extending into the South Pacific. The 1300 W lineament starts at 31 0 N the Northern Boundary Region (Figure 19) with Events 1116-1120 on the Feiberling Tablemount, passes through this region as Events 1218-1225 and extends into the Southern Pacific (Figures 12a

W D :::J I-.... Ia ...J

WEST RND SOUTH OF BRJR CRLIFORNIR 30

25

20

15

10

~ __ ~ __________ ~ __ ~ __ ~~ ____ ~O -130 -120

LONG ITUDE ( E) -110

Figure 20

1204 -- 30 JUN 1945

o 05 ocr 1984

o 02 DEC 1984

o 28 MAY 1986

o Same as Figure 13 for Region West and South of Baja Califomia. Triangle identifies the site of 1984

swarm studied by WIENS and OKAL (1987). See Table 9.

320 Michael E. Wysession et af. PAGEOPH.

and 13) as Events 230, 233, 234, Region C, all the way down to Event 414 at 34°S. This lineament is insensitive to magnitude thresholds, as the earthquakes contained within it cover a range of magnitudes, and does not seem to be an artifact of detection capabilities as similar patterns are not seen in the South Pacific areas where a gradient of detection threshold also exists. The 116°W lineament begins with Events 1212 and 1215 in this region and then passes into the Southern Pacific (Figure 13) as Events 403, 406 and 407.

Many of the earthquakes on these lineaments also correlate weil with small scale or east-west features, so the lineaments may not be statistically significant. Event 1220 is one offour small earthquakes, along with 1210, 1211 and 1214 (within error bounds) that occur on the Clipperton Fracture Zone. The epicenter of Event 1219 (and also 1213) relocates to an unnamed fracture zone south of the Clipperton, Event 1215 is located on the Galapagos Fracture Zone, and the relocation of a 1927 earthquake (Event 1212, if it can be trusted) puts it on a 10 Ma fossil spreading ridge (MAMMERICKX and KLITGORD, 1982). In addition, there are two earthquakes (1216-1217) that occur seven years apart at the same place on a fracture zone just north of the Galapagos Fracture Zone.

Much of the seismicity in this region also occurs in parts of the crust that are densely populated with seamounts. Nine events (1201-1209), as well as the swarm studied by WIENS and OKAL (1987; see Appendix A), occur in an area that includes the Suitcase Seamounts to the north, the Shimada Seamounts to the west and the Mathematicians Seamounts to the east. Three more events (1223-1225) are located in the Baja California Seamount Province. This region is very young, formed by a spreading center (now the Mathematicians Ridge) that created a microplate between the Rivera, Cocos and Pacific plates (KLITGORD and MAMMERICKX, 1982). Though there is still volcanism associated with the northern part of the Mathematicians Ridge at the Revilla Gigedo Islands (SIMKIN et al., 1981), none of the seismicity is associated with it. It is significant, however, that events 1202-1206 and one further to the east (1209) form a Iineation that locates very weil on the O'Gorman Fracture Zone, formerly a major transform in the development of this area.

Northcentral Pacific

With the exception of the Hawaiian hot spot, the region of the Pacific shown in Figure 21 is seismically very quiet, with only five earthquakes of a possible 25 located or relocated as intraplate (8 of those listed were c1erical or typographical bulletin errors). Though a large amount of Pacific intraplate seismicity has been found to correlate with seamounts or seaftoor ridges, it is somewhat surprising that the Hawaiian and Emperor Chains, which are very bathymetrically elevated, are mostly inactive (though their isolation would reduce detectability). The one notable exception is Event 1302 (September 22, 1988; mb = 5.5). This earthquake occurred

Vol. 135, 1991

z

w o ::l I-

ICI -l

6".

160 170


NORTHCENTRRL PRCIFIC BRSIN

-170 -160 LONGITUDE (El

Figure 21

321

40

35

~'-"~,:; ... :. 30 . ~- -25

20

15

10

5

0 -150 -140

Same as Figure 13 for Northcentra1 Pacific. The stippled quadrangle is the Hawaiian area, exc1uded from the present study. See Table 10.

on the southern flank of the Hawaiian chain, approximately 100 km west of Tern Island, where it was felt. In this region, the chain includes a broad structure, continuous at the 1000 m depth, comprising French Frigate Shoals, and the Brooks and St. Gregatien Banks. It is limited by the Necker Ridge to the East, and separated by a deep North-South channel from the Gardner Pinnacles group to the West. Event 1302 took place direcdy South of the Brooks Bank. The focal mechanism of the event was given by JIMENEZ et al. (1989) and DZIEWONSKI et al. (I 989c). It features nearly horizontal compressional stress release in a direction (293°) within 20° of the direction of absolute motion of the plate (311 0) (GRIPP and GORDON, 1989).

The deployment (even on a temporary basis) of a seismic station at Tern !sland could help shed light on the origin of Event 1302, by allowing the detection of any possible activity at lower magni~udes.

The other four shocks in this region were smalI, all with mb ~ 5.0. One of these, Event 1301, is only 2° from the Emperor Seamounts and is located on the 4000 m contour of the Hess Rise. Events 1303 and 1304 are north of Hawaii, with the former occurring in the Musician Seamounts and the latter farther to the east, just north of the Murray Fracture Zone, Event 1305 relocates directly south of Hawaii, east of the Christmas Ridge.

To the east of these events is the Northeast Pacific Basin, a region of smooth


bathymetrie eontours, mostly devoid of signifieant seamounts and disturbed only by the major fraeture zones: Murray, Molokai, Clarion and Clipperton. It is signifieant that until we reaeh the line of seismicity at 1300 W we do not have any intraplate earthquakes in this region, even though it is close enough to the eoast of North Ameriea for an mb = 5.0 earthquake to be weil detected.

Cocos Plate

There were 22 events of 62 that were listed in bulletins that reloeated as reliably intraplate in the Coeos plate (shown in Figure 22). Tbe loeation eapabilities for both Coeos and Nazea events are greater than for many other parts of the Pacifie Basin due to the presenee of loeal Central and South Ameriean stations, but the added eoverage also lowers the magnitude detection threshold, and it is the unreliability of the smaller events that eauses this low intraplate pereentage.

The Coeos plate is the produet of the East Pacifie, Farallon and Galapagos Rifts, and is therefore younger to the west and south, but the seismieity is seattered aeross the plate and does not bear any obvious eorrelation with age. There is a strong eorrelation, however, with the loeation of the Coeos Ridge, the northeastsouthwest trending fossil traee of the Galapagos hot spot (HEY, 1977). Eleven events (1351-1361) occur on the ridge, with ten reloeating to the middle and one

COCOS PUHE

15

z 10 W 1365,,1367" 0 • . ., ::J .1370 • CO':"'5

29 MAR 1976 I- () (7 cd . ..... • I- 1369

~ a ...J • 0

5

COCOS SWARM

0 -100 -95 -90 -85

LONGITUDE (El

Figure 22 Same as Figure 13 for Cocos plate. The triangle identifies the site of the 1976 East of Cocos swarm. See

Table 11.


to the north. In the middle there is a spatial cluster of ni ne events (Site CO-2) wh ich is tempo rally spread out between 1928 and 1976, with an isolated event slightly to the west (Site CO-I). They are weil separated from Event 1361 wh ich is located furt her up the ridge at Site CO-3. In addition, there occurred in 1976 a swarm of 28 earthquakes at a location on the east flank of the Cocos Ridge (see Appendix A). The age of the ridge at Site CO-2 is approximately 10 Ma, but there has been rejuvenescent volcanism along the Cocos Ridge (CASTILLO, 1987) so the seismicity in this region may be due either to attempts to isostatically compensate the ridge or to recent magma migrations.

There are two sites to the west of the Cocos Ridge that involve small clusters of events. Site CO-4 consists of three events (1362-1364) that occurred on January 3-5, 1987, and Si te CO-5 also contains three events (1365-1367) wh ich took place during 1974-1975. These earthquakes and the five others that are scattered to the west, however, do not suggest any spatial trends and do not correlate with any recognizable bathymetric features.

Nazca Plate

Within the Nazca plate seismlclty seems to correlate very weil with known bathymetric features, as can be seen in Figure 23. Outside of the very active Galapagos swarms, there are 45 earthquakes that we consider to be reliably intraplate (of the 111 originally listed here), of which fully 90% locate on or near seamounts, islands, or other bathymetric features. The especially strong correlation with extinct ridges is a striking feature of the Nazca seismicity.

The seismic activity at the northern part of the plate is dominated by the Galapagos hot spot. The Galapagos Islands are active volcanically, and there is much seismicity connected with this (see Appendix A). The Carnegie Ridge is the ancient trace of the hot spot, but unlike its symmetrical counterpart, the Cocos Ridge, wh ich is still seismically active, the Carnegie Ridge seems to be quiet. There are three events (1501-1503), however, which lie on its southern flank, and it is uncertain whether or not they are of volcanic origin. Two of these are 0.5° from the islands Floreana and Espafiola, which are certainly of volcanic origin but do not seem to be currently active. It was thought that eruptions occurred on these islands in 1813 and 1958, respectively, but these reports have since been discredited (SIMKIN et al., 1981).

There is one other site of seismicity in the Nazca plate, Site NZ-13 near the Juan Fernandez Islands, that may be associated with volcanic activity. Submarine volcanism occurred in 1835 just north of the main island at (33.62°S, 78.78°W), and possibly in 1839 at (33.62°S, 76.83°W) as weil (SIMKIN et al., 1981). There are five earthquakes (1541-1545) that occur at this site, three on the south flank of the islands, about 40 km from the eruption sites, and two further south.

South of the Carnegie Ridge lies the Grijalva Ridge and two bathymetric highs

Figure 23 Same as Figure 13 for Nazca plate. The triangie identifies the site of the Galapagos swann activity,

1968-1998. See Table 12.

that are south of and parallel to it, the southem being the Sarmiento Ridge. These features have complicated histories in which they began as fracture zones between the Pacilic and Farallon plates, but became involved with the early spreading between the Cocos and Nazca plates. The Grijalva Ridge seems to have been the southem scarp of the Cocos-Nazca spreading center when it initiated 25 Ma b.p. (HEY, 1977), though the scarps south of that mayaiso have been involved with the spreading. All live of the events occurring here locate on the sharp bathymetric scarp between the Grijalva and Sarmiento Ridges, with 4 (Events 1504-1507) clustering at 85°Wand one (1508) further west at 88.6°W.

There is also an association between seismicity and the bathymetric lineament that forms the Quebrada Fracture Zone (Site NZ-3, Events 1509-1511). Of the three events here, one locates just north of the fracture zone and two relocate to its eastem terminus. One other earthquake located on a fracture zone, that being Event 1540 on the Challenger Fracture Zone.

Lithosphere forming the present Nazca plate was involved in the reorientation and jump of the old Farallon Ridge and concurrent break up of the Farallon plate starting in Late Oligocene (HERRON, 1972; MAMMERICKX et al., 1980). As a result, a set of extinct spreading centers, in three distinct segments trending


north-south, divide the plate in half. These fossil remnants of the Farallon Ridge are the Galapagos Rise to the north, the Mendoza Rise in the center and the Roggeveen Rise in the southern end of the plate. Though these spreading centers have been extinct for some time, their active seismicity suggests that they constitute zones of tectonic weakness.

Ten events (1512-1521) locate in the region of the Galapagos Rise, as defined by the 400 m isobath. This spreading center was active until 6.5 Ma ago, as described by MAMMERICKX et al. (1980), when Pacific-Nazca motion was fully taken up by the East Pacific Rise. This area (Site NZ-4) has several ridge segments that extend from just south of the Quebrada Fracture Zone down to the Mendaiia Fracture Zone, and the 10 earthquakes are evenly distributed throughout. This includes Event 1519, quoted by the ISC as having a depth of 143 km but which was relocated to 13 km by MENDIGUREN (1971), who also showed that its mechanism (see Figure 23) is compatible with ridge-push stresses.

The Mendoza Rise is seismically quiet, but the Roggeveen Rise (Site NZ-9) is very active, with six events (1532-1537) clustering on the fossil ridge. The 3500 m isobath defines this feature as having two parallel north-south bathymetric highs, and the earthquakes occur along these prongs, three on a side. Event 1535 on the western prong is tentatively one of the largest events in our dataset (MpAS = 7.0).

There is also seismicity connected with the hot spot trace that extends across the plate, the Sala y Gomez Ridge and the Nazca Ridge. There are six events located at Site NZ-8, at the Western end of the Sala y Gomez. Four of these earthquakes are very recent and cluster 1 ° due east of Easter Island (1528 -1531). Though Easter Island is the most recent expression of the hot spot, there has not been any known Holocene volcanic activity there. There are two adjacent events that relocate to site NZ-7 along the 2500 m isobath of the Nazca Ridge, though the rest of the Sala y Gomez Ridge between here and Easter Island is seismically quiet.

In conclusion, the Nazca plate is remarkable in that only 4 events (1522, 1523, 1538, 1539) do not seem to be associated with prominent bathymetric features.

Antarctic Plate

There have been 15 intraplate earthquakes in the region of the Antarctic plate that is part of the Pacific Basin, shown in Figure 24. The seismicity is in two parts, a northern set that is in relatively young lithosphere and a southern segment that is older.

The seismicity of this region is sparse (there were no confirmed earthquakes west of 115°W) and does not match significantly with bathymetric features, though there are two locations of recurrent seismicity. One involves the earthquake of May 9, 1971 (mb = 6.2, studies by FORSYTH (1973» and its three aftershocks (1702-1705), as weil as an older event (170 I on March 12, 1927) at approximately the same site. The recurrence of seismicity at this 10cation is an interesting feature.

326

z

L1J CJ ::J I-..... ICI: ...J

Michael E. Wysession et al. PAGEOPH.

RNTRRCTIC PLRTE -34

1702 -- 09 MAY 1971

-39

-49

1714 -- 05 FEB 1977

-54

-59

-54 1715 -- 07 NOV 1979

~~ ____ L-__ ~ ____ ~~~-59

-110 -100 -90 -BO -70 LONG ITUOE (E J

Figure 24 Same as Figure 13 for the Antartic plate. See Table 13.

The second site is situated to the south of the first and eonsists of two shoeks (1706-1707) reloeated by OKAL (1981) that occur at nearly the same loeation in 1956 and 1961. Both of these sites are on very young lithosphere near the Paeific-Antarctic spreading center. The first site is between the 5 and 5a magnetie lineaments, implying an age of 10 Ma, and the seeond is even younger, with an age of no more than 6 Ma.

The two events (1709-1710) at the eastern side of the northern seismicity both oeeur on fraeture zones of the Chile Rift. Event 1710 relocates to the Guafo Fraeture Zone, and the other loeates on the unnamed fraeture zone directly north of it.

The five earthquakes ( 1711-1715) that are further to the south form a quasi linear trend and oeeur in mueh older lithosphere adjaeent to Antaretiea-the three western events are in Cretaeeous lithosphere. The two events to the east are both within 4° of the South Shetland treneh, whieh is an aetive subduetion zone south of the diffuse plate boundary with the Seotia plate (PELAYO and WIENS, 1989).

7. Conclusions

The prineipal aspeets of our study ean be summarized as folIows:

l. The measurable level of intraplate seismicity of the Paefie Basin is very low, eonsidering its size, and even less so than previously assumed. Our reloeations


have revealed 403 intraplate events (not counting localized swarms), which is only 48% of the amount reported from previously published bulletins.

2. The majority of earthquakes occur as parts of a small number of spatial and temporal swarms. Some of these, like the Galapagos Islands swarms and Hawaiian seismicity, are clearly connected with hot spot volcanism, while others, like the Line Islands swarms, do not have any recognizable tectonic or volcanic origins. Nothing is known at lower magnitudes and the exploration of swarm locations like the Gilbert Islands, Regions A and C, and South of Baja is long overdue.

3. A very large number of earthquakes, 304, relocated to seismically active plate boundaries, in particular the Tonga, Japan and Central American trenches. Errors in the original locations can be rectified through a more careful and critical use of a least-squares linear regression routine. Many events located in Pacific intraplate regions are smalI, poorly located aftershocks of larger plate boundary earthquakes.

4. The correlations between seismicity and bathymetry in the intraplate regions of the Pacific are inconclusive. In some regions, like the Nazca plate or Eastcentral Pacific, nearly all of the seismicity is highly correlated with identified bathymetric features. These include former plate boundaries (like fracture zones and fossil spreading ridges), seamounts and hot spot traces. In other regions, like Region A, the seismicity does not correlate weIl with bathymetric features, and often the release of large amounts of seismic energy is not understood in the context of local tectonics.

5. The seismicity also takes the form of distinct lineaments in several regions of the Pacific. Some, such as those along the northeastern flank of the Tuamotus and the three lineations through Micronesia and the Marshall Islands, follow the direction of plate motions. Others, like the spectacular 1300 W lineament, neither follow plate motions nor correlate with bathymetry.

6. The high amount of seismicity in the Western Pacific may represent the onset of a relocation of the choking Solomon trench further to the north. There is obviously considerable activity in the eastern part of the region, with the Gilbert Islands seismicity as the biggest contributor and so me level of consistent activity in the Phoenix Islands. But in the west the mechanism at the Ralik site is incompatible with the Gilbert Islands mechanisms, and seismicity further west is much more diffuse. A simple linear model for the creation of a new trench is therefore certainly an oversimplification, especially west of 1700 E.

7. Maps of global seismicity often show many earthquakes surrounding subduction zones and extending for hundreds of km into the oceanic plate. Most of these events, especially around the Tonga, Japan, Kurils and Aleutian trenches, are mislocated plate boundary events. This suggests that coseismic stresses from trenches are not transmitted into the approaching oceanic plate beyond 2° in a manner that can be released as teleseismic earthquakes.


8. Databases of intraplate events contain many erroneous listings. Most of those are blatant c1erical errors that have occurred either by the locating agency or in the transcription to a reporting bulletin, and many involved switched negative signs in the latitudes or longitudes. There are many nuclear tests that are listed, but as they are only a percentage of the full number, the criteria for their identification in published bulletins is unclear.

9. Pacific intraplate events tend to be smalI, with a paucity having reliable magnitudes of mb ~ 6.0. In regions where there is a good distribution of local stations, like Polynesia and New Zealand, the magnitude detection threshold is lowered and considerably more events are recorded. This detectability is not available for many isolated regions of the Pacific, and the numbers of reliable epicenters there is extremely smalI. The small magnitudes and large teleseismic distances for these events make the determination of focal mechanisms difficult, and there are surprisingly few available.

Appendix A

Swarms

In addition to events isolated in space and time, considerable seismic activity takes place in the Pacific Basin in the form of swarms. In order to avoid c1uttering our datasets, we have removed such events from Tables 1-13. In this appendix, we present a rapid review of the main swarms known in the Pacific Basin. Individual listings of all 838 events involved have been prepared and are available as part of our computerized datasets.

Gi/hert Islands, 1981-1984

Probably the most remarkable swarm to take place in the Pacific Basin, the Gilbert Islands activity las ted from December, 1981 to April, 1983, with two iso la ted events continuing into 1984. A total of 225 earthquakes were detected teleseismically, at a minimum magnitude threshold of mb = 3.9 (see Figure A-l). Two more (Events 764 and 765) were relocated into the swarm from a different bulletin location. In the absence of a local network, nothing is known of the probable activity at lower magnitudes. The major events in the swarm have been studied in detail by LA Y and ÜKAL (1983), and their conclusions have been confirmed by 36 available moment tensor solutions listed in Table 14. Using cluster relocation techniques, SVERDRUP and ÜKAL (1987) have estimated the spatial extent of the main events in the swarm to be 40 by 30 km. Despite its high b-value, LA Y and ÜKAL have indicated that the swarm is unlikely to have a volcanic origin;

Vol. 135, 1991 Intraplate Seismicity

GILBERT ISLANDS 1981 - 1984

.. 1&

~ ; 10

IQQQ

2.S .,..-,...--,,--r---r----,r-::.

Figure A-I

2.25

1.15

I.S

~ 1.25 !! I

0.15

O.S

0.25

4 4.5 5 5.6 6 a ..... -Wave Itagn i .udo

329

(b)

(d)

Gilbert Island swarm, 1981-1984. (a) History of individual events, with magnitude, as a function of time. (b) Histogram of number of earthquakes per 7-day window; (c) Histogram of energy release per 7-day window; (d) Frequency-magnitude analysis: dots are incremental values, squares cumulative ones.

The straight line shows a best-fit regression of the cumulative values.

interpretation of the orientation of released stress has been given by OKAL et al. (1986), in the general, although at this stage somewhat speculative, framework of the incipient development of a new "Micronesian" subduction zone (KROENKE and WALKER, 1986). The occurrence, in 1921, of an event about 200 km to the North, with a confidence ellipse encompassing the swarm site, suggests a pattern of recurrence of seismicity at the site. It is clear that this location should be earmarked for a small-scale marine geophysical study including OBS deployments.

South 0/ Baja California, 1984-86

As studied in more detail by WIENS and OKAL (1987), this swarm involved 69 teleseismically detected earthquakes (mb ~ 3.5). It started abruptly in September, 1984, culminated with the main shock on December 2 (Ms = 6.2), and died off very quickly with sporadic activity until July, 1985, followed by two isolated aftershocks in April and May, 1986. No new activity had been reported at the site as of June 6, 1990. Figure A-2 shows a b-value comparable to world-wide averages. The available focal solutions shown on Figure 23 point out to normal faulting; WIENS


SOUTH OF BAJA 1984 - 1986

ID 6 UI ... z 8 • ...

> ~ 5

... "-·e D .. .. 4 :! ... 5 2

D

1- JILIAN DAYS

2.5

2.25 b = 0.90

a (.16 ..

i 21 1.6

~ z a 1.25

>-2D !! I

(d)

u .. D.15 ~ lD ... D.5 D

D.25 i 11 u D 9

3.& 4 4.5 5 S.S 6 1G84 JILIAN DAYS Iady-Vav. I10gni 'ud.

Figure A-2 Same as Figure A-I for the 1984-86 swarm South of Baja Califomia, studied by WIENS and OKAL

(1987). Time windows in (b) and (c) are 5 days.

and OKAL (1987) have indicated that they are compatible with the release of thermoelastic stresses, as suggested by TURCOTTE (1974) and BRATT et al. (1985), but if so, the question of the spatial concentration of this activity, and of its magnitude, remains unclear.

Ga/apagos Islands

We discuss here seismic actlVlty in the immediate vicinity of the Galapagos Islands, despite its location within 2° of the Nazca-Cocos spreading center. Its characteristics, evaluated within the framework of documented volcanic activity, shed precious light on the nature of other swarms, notably underwater ones, for which no such evidence is accessible.

Activity in the Galapagos Islands is presented on Figure A-3. About 250 earthquakes (3.8 ~ mb ~ 5.4) were recorded teleseismicaly during the collapse of the Fernandina caldera, in June, 1968. We refer to FILSON et al. (1973) for a detailed analysis of this seismicity, and of a myriad of events of smaller magnitude detected either by the local station GIE, or by distant high-gain observatories (including Pacific Basin hydrophones). Focal mechanisms are available for the largest event

Vol. 135, 1991

• " , .: 5 c D 0 >:

'" ~

~ "- 2 o

~ § 1 z o ,

1000


GALAPAGOS Is. 1968 - 1988

(a)

7000

300

~ z ~ 200 '" "-0

'" ::g 100 >: :0 z

. D

~ 21

,. '-' 15 20 z '" o i 19

1000 3000 1968

5000 JULIRN ORYS

5 0 l..I-~:-!-....L..l..LII~:-L.L-'-:~L--_--:;:!:.:-'LlJ '-' o .J

1000 3000 1968 JULIRN ORYS

Figure A-3

331

(b)

7000

(d)

7000

Activity at the Galäpagos site, 1968-1988. (a) History of individual events; (b) Histogram of number of earthquakes per 30-day time window; note that the 1968 swarm dominates all activity; (c) Same as (b) plotted on a logarithmic scale; (d) Energy release per 30-day window. A comparison of (c) and (d) illustrates the difference in nature between the 1968 caldera collapse, and the relatively more energetic

swarms in the subsequent years.

(June 14, 1968, 22:27 GMT) and KAUFMAN and BURDICK (1980) have shown that at least 20 events of sm aller magnitude have exactiy similar geometries.

In the present study, we have limited ourselves to a compilation of Galapagos events reported on the NEIC tape (Figure A-4). A frequency-magnitude analysis of our dataset yields a high b-value (1.95) comparable to FILSON et al.'s (1973) result for the high-end of their magnitude range (b = 1.91).

Additional activity in the Galapagos Islands featured a short-lived swarm of 17 teleseismically detected events in the Spring of 1971, sporadic activity through the 1970s and 1980s, and a burst of 11 teleseismically detected earthquakes in 1988, eight of them on February 24. Of the total 51 events recorded posterior to the 1968 swarm, only the latest earthquake (September 14, 1988) is listed as associated with volcanism. Our b-value investigations of the 1971 and 1988 episodes, as weil as of the whole post-1968 activity, yield va lues typical of world-wide averages (0.97, 1.00 and 1.05, respectively), suggesting that these events were not direct1y controlled by magmatic processes (Figure A-5). Furthermore, CMT solutions available for two events on February 24, 1988 feature thrust faulting (see Figure A-6), and a larger earthquake on May 20 is predominantiy strike-slip on a dipping fault.

332

140 lU68 Jl.A..lAN QAYS

IB.5

UNi8 JlLIAN DAYS


GALAPAGOS Is. 1968 ONLY

(a)

40

(c)

200

Figure A-4

IQSB

" I J j

jl 140 ISO 180

1.5 z CI 1.25 o - I

0.75

0.5

0.25

lQ68 JUL.IAN ORYS

3.5 4 4.5 5 5.5 B~-Vay. I10gni lud.

PAGEOPH,

J

(b)

200

(d)

Same as Figure A-I for the 1968 swarm during the collapse of Fernandina. Time windows in (b) and (c) are I day.

East Flank 01 Cocos Ridge, 1976

The Eastern Flank of the Cocos Ridge, in the vicinity of 4°N, 87SW, was the site of a swarm of 28 teleseismically located events, during aperiod of 70 days in the first half of 1976. The ISC reprinted a rather long description of the main event (March 29, 1976; 05:39, mb = 5.9; M s = 6.5) by the Smithsonian Institution's Center for Short-lived Phenomena, as representing interplate motion along the Cocos-Nazca boundary. However, it is now c1ear that the plate boundary in this region runs along the Panama Fracture Zone, about 300 km to the East, and this swarm is actually intraplate.

The history of the activity of the swarm and a frequency-magnitude plot (b = 0.99) are shown on Figure A-7. A focal mechanism for the main event is given on Figure 22 (BERGMAN and SOLOMON, 1984; WIENS and STEIN, 1984; WIENS, 1987). It is characterized by strike-slip motion with one of the fault planes striking at an azimuth of 28°, identical to the direction of absolute motion of the plate, 28.6° (GRIPP and GORDON, 1989), which is of course also the trend of the Cocos chain.

The Cocos Ridge has been described as the wake of the Galapagos hots pot on the Cocos plate (HOLDEN and DIETZ, 1972; HEY, 1977). Significant volcanic edifices postdating the formation of the ridge itself have been recognized on its

CA

LAPA

CO

S I.

. , 9

68

-88

C

ALA

PAC

OS

I..

Alt

",.

, 968

C

ALA

PAC

OS

Is.

'91

' O

nl'll

2.5

2

.25

1

.75

b

= 1

.05

b

= 1.

41

b =

0.9

7

1.2

5

• 2

1.5

1.7

5

1.2

5

z 1

.5

z C

I 1

.25

C

I ;

0.7

5 [

.~

] ~

0.7

5

0.5

0.2

5

0

•

3.5

4

4.5

5

5.5

BO

ÖW

-Wav

. K

Dgn

itud

.

0

0.7

5

0.5

0.2

5

6 4

.5

5 5

.5

Bod

y-W

av.

Mag

nitu

d.

Fig

ure

A-5

•

-0

.5

0.2

5 f

-

o f-

-l -l

4.5

5

5.5

B

ody-

Wav

. H

agn

itu

d.

Fre

quen

cy-m

agni

tude

ana

lyse

s of

Gal

apag

os a

ctiv

ity f

or v

ario

us t

ime

win

dow

s: L

eft:

Ful

l da

tase

t; C

ente

r: S

ince

cal

dera

col

laps

e: R

ight

: 19

71

swar

m o

nly,

~ w

,!.A ~ a .... I>

l "0

~ " W <n' a ö:

Q w

w

w


FOC~L SOLUTIONS FOR G~L~P~GOS SW~RMS

1968 SWARM 24 FEE 1988 15:43

o 24 FEE 1988 20:22 20 MAY 1988

Figure A-6 Focal mechanisms available for Gahipagos swarm events. Sources are listed in Table 14. For the 1968

swarm, this mechanism is common to at least 21 events (KAUFMAN and BURDICK, 1980).

EAST OF COCOS RIDGE 1976

(a) i!! 4 (b) . ffi " ~ 3 .~ 5 I'; c co .. 2 l! w

~ I

40 &0 1&0. 40 &0 1&0

1.75 b = 0.99

1.5

.. 21 (e)

1.2S co z ~ I

.. 20 ! 0.75 u

(d)

.. ~ 0.6

IU

7 0.2S

u 0 .J 18

1&0 1871

Figure A-7 Same as Figure A-I for the 1976 swarm East of Cocos. Time windows in (a) and (b) are I day.


summit and its flanks, the most prominent being Cocos Island (CASTILLO, 1987). The site of the 1976 swarrn is on the East flank of Cocos Ridge, approximately aligned with Tortuga and West Cocos seamounts. The occurrence of a swarrn at this site is suggestive of volcanic activity. However, the strike-slip geometry of the main shock, together with the b-value of 0.99, suggest that the seismicity was not directly controlled by active magrnatism. It is clear that a more detailed investigation of the site is warranted.

Region A, Line /slands

OKAL et al. (1980) reported 90 events, 10 of which detected teleseismically, at a location ("Region A") East of the Line Islands, approximately 7SS and 148°W (see Figure 12a). A major swarrn took place during 1968-70, followed by sporadic energy release continuing until the end of 1976. The site was then quiet for 9 years, until a short burst of new activity took place in 1985 at an epicenter undistinguishable from Region A. Six events were detected teleseismically and 8 more recorded at the French Polynesia array, bringing the total to 104 events at the site since 1968. The history of activity at Region A is presented on Figure A-8.

JORDAN and SVERDRUP'S (1981) cluster relocations showed that the 4 main events were spread over approximately 25 km. A vailable focal mechanisms (OKAL et al., 1980) are all strike-slip, compatible with the release of ridge-push stresses.

111&8 JLLIAN DAYS

_20

~ - ID

~ ~ 18

9 I

9 11

REGION A 1968 - 1988

11100

15 UI

~ .. 10

Figure A-8

(c)

1000

Same as Figure A-I (a, b, c) ror activity at Region A, Line Islands. Time windows in (b) and (c) are 30 days.

RE

GIO

N A

--

19

68

-76

R

EG

ION

A --

19

68

-88

R

EG

ION

A --

19

85

On

I.y

2 •

b =

1.0

6

2 ~' . .

~

~ = 1.~

2~

1.7

5 F

b

= 0

.91

1

.75

1

.5

1.7

5

1.5

1

.5

.~

1.2

5

1.2

5

1.2

5

Z

Z

z C

I C

I g'

0.7

5

a a

0.7

5

0.7

5

0.5

0

.5

. •

0.5

0

.25

0

.25

0

.25

0 0

0 •

3 3

.5

4 4

.5

5 3

3.5

4

4.5

5

3.5

4

4.5

5

Bad

y-W

avo

Mag

nitu

do

Bad

y-W

avo

Mag

nitu

do

Bad

y-W

avo

Mag

nitu

do

Fig

ure

A-9

F

requ

ency

-mag

nitu

de a

naly

ses

for

thre

e pe

riod

s of

acti

vity

at

Reg

ion

A:

Lef

t: E

arly

sw

arm

s (1

968-

1976

); C

ente

r: F

ull

data

set;

Rig

ht:

1985

onl

y.

Sym

bols

as

on F

igur

e A

-Id.

w

w

0\ ::: (i'

::r'

'" !l. r"l ~

'< '" " '" '" 0' ::;

~

I:l ,..... ~ a ~ .~


The new events (mb ~ 4.8) are too small for reliable focal mechanism studies; however, our frequency-magnitude studies revealed no significant change in b-values between the older events and the new swarm in 1985 (Figure A-9). During a 1979 cruise on R/V Gyre, no sign of anomalous bathymetry could be recognized at the site (SVERDRUP, 1981), and thus the origin of clustering of activity at Region A remains unclear.

Region C, East of Tuamotu Islands

Another site of particular interest is Region C, 10cated approximately 21°S and 127°W. ÜKAL et al. (1980) documented 97 events (mb ~ 5.5) during a short-lived swarm lasting from August, 1976 to June, 1979. This precise epicenter (GB-5, which JORDAN and SVERDRUP (1981) estimated to be spread over 45 km) has been quiet ever since, but significant activity has taken place at nearby locations. A major event (mb = 5.9; M s = 5.3), followed by a same-day aftershock (mb = 5.4) occurred in 1983 at GB-8, about 75 km North of GB-5; three additional events took place in 1986 at GB-9, 200 km to the Northeast (see Figure A-lO).

z

w o ::J t-..... ta ...J

()

REGION C

GB-9

• ,. GB-5 1976-79

o

-17

-18

-19

-20

-21

-22

0 -23

-24

~~L-~~~ __ ~~~~~ __ ~ __ ~ __ ~ __ ~-25

-130 -129 -128 -127 -126 -125 -124 -123 -122 -121 -120 L.ONGITUDE (E)

Figure A-lO Map of seismicity in Region C. There were 97 events at G8-5 in 1976-79, 2 at G8-8 in 1983 and 3 at

G8-9 in 1986.


OKAL et al. (1980) demonstrated that the swarm had an extremely low b-value (b = 0.55), and obtained one strike-slip focal mechanism for the July 25, 1978 event at Region C; a CMT solution is now available for a second event (January 5, 1978) which they could only partially constrain (see Figure 12b). Both solutions are compatible with the release of ridge-push stresses. The 1983 mainshock at GB-8 features normal faulting (WIENS, 1985; DZIEWONSKI et al., 1 984a). The variety of mechanisms obtained in Region C is illustrative of the variation with depth of the stress regime of a young oceanic plate (BRA TT et al., 1985).

The origin of the activity at Region C is unclear. Little is known from in situ investigations, except for the presence of distorted topography on the ocean floor (J. FRANCHETEAU, pers. commun., 1980). Using satellite geodesy, OKAL and CAZENAVE (1985) have proposed a speculative model of the areas' evolution during the Miocene, in which Region C would have formed during an episode of ridge propagation of the Old Farallon spreading center, driven by the Easter Island hotspot. A more definite interpretation must await a systematic exploration of the site.

Other Polynesian Swarms

The existence of the high-gain Polynesian Seismic Network (OKAL et al., 1980) has allowed the detection of a number of local swarms in and around Polynesia, the most spectacular of which, involving up to 40,000 detected events in the TahitiMehetia area, resulted from underwater volcanic activity from the Society Island hotspot (TALANDIER and OKAL, 1984a, 1987). In addition, Macdonald Volcano has been the site ofintense volcanoseismic activity, whose Twaves are recorded routinely in Polynesia (T ALANDIER and OKAL, 1982, 1984b, 1987). In keeping with our exclusion ofHawaiian seismicity, we do not list such activity anywhere in our dataset.

Weaker swarms (involving up to 25 recorded events) have been detected at other Polynesian sites, including North of the Society Islands, and in the Austral Islands (OKAL et al., 1980). These events are listed as part ofthe Polynesian dataset in Table I, and their location is outlined on Figure 12d.

Appendix B

Description of Individual Regions and Sites

In this Appendix, we provide a detailed account of the results of our relocations, and in particular discuss some of the more difficult cases, in which the differences in original and relocated epicenters are most drastic. We refer to the main text for a discussion of the seismic locations notably in the context of bathymetry.


The discussion is arranged geographically, starting with the 10 Pacific plate regions, followed by the Cocos, Nazca and Antarctic plates. Figures 12-24 plot the genuine intraplate seismicity over the SYNBAPS bathymetry, contoured at 1000 m intervals (2000 m in Micronesia and the Marshall Islands).

Polynesia

We define Polynesia as extending from 300 S to the Equator, between longitudes 161°Wand 125°W (see Figure 12). This region is covered by the French Polynesia seismic network, wh ich has been described in various papers (T ALANDIER and KUSTER, 1976; OKAL et al., 1980). In addition to the swarms at Regions A and C, the total dataset includes 200 potentially intraplate events.

Intraplate Earthquakes

• In addition to new events described in the main text, only one historical earthquake could be relocated with confidence inside Polynesia: Event 177 (July 17, 1929).

Casualties

• Event 238 (June 21, 1918) could not be relocated. The ISS indicates that it probably occurred in the midst of a local shock off the coast of Southem Califomia. GUTENBERG and RICHTER (1941) similarly commented that the solution was defective.

• Event 237 (September 27, 1973) is listed by Wellington at 16°S and 160oW. Despite the short distance, it was not recorded in Polynesia (J. T ALANDIER, pers. commun., 1989). The WEL epicenter listed by the ISC provides a very poor fit to the 4 available times (er = 7.1 s). The only way to make the data consistent is to interpret the time at Mangaho (MNG) as S rather than P. This provides an excellent solution at the bottom of the Tonga Benioff plane.

• One event (236 on November 8, 1969; 17:05 GMT), allegedly South of the Marquesas relocates to the Region A swarm' as listed in OKAL et al. (1980).

• Two events (239 on April 17~ 1927; and 240 on January 12, 1942) are blatant mislocations due to clerical errors.

• Finally, and as discussed in Section 4, we identified 61 events as presumed nuclear explosions (241- 30 1).

Focal Mechanisms

Apart from mechanisms at Regions A and C (see Appendix A), focal solutions for three events at GB-4 were published by OKAL et al. (1980). The only new


solution is for Event 211 (Location AU-6; November 20, 1979). OKAL (1984) gave a partially constrained thrust faulting mechanism; the newly available CMT solution (DZIEWONSKI et al., 1987b) is in general agreement with it. Focal solutions are regrouped on Figure 12b.

Other Sites in the Southcentral Pacific

We regroup here all seismicity in the Pacific plate South of the Equator, bordered on the West by meridians 125°W to 300 S, 161°W to 43°S, and 1700 W further South (Figure 13). OKAL ( 1984) identified 22 potentially intraplate events in this area. Access to the ISC catalogue, and to W ALKER'S (1989) dataset resulted in the recognition of 27 additional, potentially intraplate earthquakes. In addition, three new events (all in 1988) occurred since the completion of OKAL'S (1984) study, for a grand total of 52 potentially intraplate events, all of which were re-examined in the present study. Events are described and numbered in a geographical pattern from Northeast to Southwest. Locations codes identified in OKAL (1984) are retained. New locations are given codes starting with SP- (for "Southern Pacific"). In general, our results are similar to OKAL'S (1984), but a number of improvements were possible:


• Apart from 6 events, described in detail below, all Southcentral Pacific intraplate earthquakes in OKAL (1984) are confirmed by the present study, although one of them (Event 420' on September 5, 1938) was moved 120 km to the Northeast while still remaining inside the plate.

• Of the 30 additional events considered, only 10 were confirmed as genuine intraplate: three historical earthquakes (40 I, 402, 417), four more recent ones (403,411,412,413), and the three 1988 earthquakes (407,416,425).

Casualties

• Access to the BCIS listings allowed relocations for the period 1957-1963, when the ISS listings are limited to large earthquakes. As a result, we relocated to the ridge Events 434 (July 26, 1958),435 (July 30, 1958), and 437 (October 17, 1959; 01:23) for which OKAL (l9S4) could not perform relocations. Consequently, his sites IP-14 and IP-16 are eliminated.

• Confidence ellipses are computed systematically for all events, including those post-1962 earthquakes supposedly well-located by the USGS or ISC. As a result, we eliminate from the intraplate listings Event 445 (July 31, 1976). Site IP-8 is eliminated.


o Regarding Event 452 (January 13, 1938), and as mentioned by OKAL (1984), arrival times are not listed in the ISS. Access to BCIS data and to a number of seismograms resulted in an inconsistent dataset, precluding convergence to an acceptable solution. Elimination of the times at WEL, where the instrument developed instability, yields an intraplate location (26.7°S, 170.3°W), but the confidence ellipse reaches the trench. Epicenter IP-19 is eliminated.

o Finally, regarding Event 431 (September 17, 1949), we were able to duplicate OKAL'S (1984) solution at IP-13, but a statistical analysis reveals that a ridge location at 63°S and 165°W cannot be excluded. We elect to consider this event as interplate.

o Among the newly identified events, 12 were relocated to the East Pacific or Pacific-Antarctic Ridges (428-430, 433, 436, 438-442, 444 and 446), and three had confidence ellipses intersecting the plate boundary (426, 427 and 443), the latter being described in the main text and on Figure 5.

o Finally, five older events (447-451, dating from 1920 to 1924) had insufficient data to allow stable, if any, relocation.

Focal Mechanisms

OKAL (1984) obtained 5 focal mechanisms in this region. One of them (Event 424 at IP-15) is unconstrained normal faulting. New solutions are available for only two earthquakes: Events 404 and 405 at IP-l. The Harvard solution for Event 404 (DZIEWONSKI et al., 1988b) is a thrust fault violating a strong, impulsive dilatation at NNA. On this basis, we prefer the strike-slip solution given in OKAL ( 1984).

Campbell Plateau to Kermadec

We define this area as the portion of the Pacific plate South of 27°S and West of l61°W; South of 43°S, the Eastern boundary is moved to 1700 W (see Figure 14). As documented for example on W ALKER'S (1989) map, this region is rieh in potentially intraplate earthquakes. However, because of his coarse gridding of intraplate areas, a very incomplete pieture is presented on his map. By combining Walker's dataset with the NEIC tape, and the original ISC regional eatalogues, we identified a total of 76 potentially intraplate earthquakes, of whieh 41 were retained as truly intraplate.

lntraplate Earthquakes

Intraplate events are sorted and diseussed by generally increasing latitude. See the main text for geographical diseussion; among them:


• Five events (502-506) regroup into three epicenters on the Campbell plateau in the vieinity of its magnetie anomaly. While these are not high-quality loeation (0" reaehing 3 s on 8 stations), their intraplate eharaeter is statistieally signifieant.

• OKAL'S (1984) epieenter IP-17 (Event 507 on August 1, 1953) was eonfirmed at 44.2°S and 175SW, in the immediate vieinity of the Chatham Islands. Two additional events (508 on April 9, 1965; and 509 on July 26, 1975) are listed by the ISC in the same general area, along the Chatham Rise, and were eonfirmed upon reloeation.

• In the vieinity of New Zealand, between approximately 45 and 400 S, lies an aetive seismic zone eomposed of two major clusters. The northern cluster is quite evident on W ALKER'S (1989) map, although approximately 7 of the earthquakes he lists in the region aetually qualify as interplate. However, because he limited his analysis to the quadrant East of 180°, and South and 400 S, he totally missed the existenee of the seeond cluster further South, and a signifieant number of events in the Northern cluster. As a whole, we identified 16 events in the Southern cluster, and 11 in the Northern cluster. An additional event (Number 537 on March 28, 1976) is isolated on the Northeastern flank of the Campbell plateau, about 130 km North of the Northern cluster. In general, our reloeations easily duplieated the WEL epieenters, but were often ineonsistent with the ISC. This is due primarily to our use of S times. The largest magnitude for all these events was mh ~ 5.1, too small to allow any foeal meehanism studies.

• In the northeastern quadrant of the area under study, three events (two historieal and one recent) relocate in the immediate vieinity of the Louisville Ridge (537 on March 28, 1976; 538 on May 10, 1924; and 539 on August 10, 1926), the latter being moved more than 1500 km from its Iisted loeation.

• Finally, we confirm Event 540 (January 4, 1940) at Site IP-20, as defined by OKAL (1984), and a single event (541 on August 1, 1925) reloeates to 29.4°S, 164.3°W; the intraplate eharaeter of this solution being statistically signifieant.

Casualties

A total of 23 events were reloeated to the Paeifie-Australia plate boundary, two of them at signifieantly deep foei (546 and 547 on May 23 and July 9, 1936), and three others as recently as 1984-85 (561-563). Included are 7 events listed as intraplate by the ISC, but whose reloeated epicenter (on the plate boundary, or less than 2° away) is the original WEL solution. Two events (561 and 563) had intraplate PDE listings, and reloeated to an ISC interplate solution.

• In the ease of Event 575 (June 30, 1985), the ISC solution on the Campbell plateau aehieves a residual of only 0" = 3.8 s on 3 P and 2 S times. A similarly medioere solution (0" = 4.2 s) is obtained by loeating the earthquake on the


Puysegur trench, at 49SS, 162.3°E. The residual could be further reduced (to 3.9 s) by deepening the focus to 60 km, but it is not clear that seismicity does occur at such depths in the Puysegur trench. We list the earthquake as failing to converge (code 12).

• Similarly, for Event 559 (May 27, 1982), the ISC solution on the Campbell plateau fails totally to fit S times, resulting in a residual (1 = 21 s for the full dataset. An inversion of P times only provides two equally excellent solutions, one at the ISC epicenter, the other on the plate boundary at 49.7°S, 164.0oE; this latter solution also fits 3 of the 4 S times, and the listed S at OMZ, if interpreted as P (this station reports only an alleged S time), with a global residual of only (1 = 3.4 s. We regard this solution as better quality than the ISC's, and classify the earthquake as interplate.

• In addition, one event (Number 550 on April 21, 1959) relocated to the PacificAntarctic ridge.

• The reported location of Event 576 (October 15, 1957) results from a clerical error; it actually belongs to the Kermadec Benioff zone.

• Finally, six historical events (566-569; 572 and 573, dating from 1923 to 1927), but also two recent ones (570 and 571) had too few data for stable, if any, relocation. The last event (574 on November 29, 1974) fails to converge to a satisfactory, stable epicenter, we believe that some of its readings may actually relate to a Japanese earthquake occurring only 9 minutes earlier.

Focal Mechanisms

Only one focal solution is available in this region: (Event 506, June 3, 1983; mb = 5.2), its CMT mechanism featuring strike-slip (DZIEWONSKI et al., 1983c).

Off Tonga and Samoa

We concentrate here on the area located East of the Tonga trench north of 27°S, and East of 172°W north of l3°S. The Equator and meridian 16 PW bound it to the North and East (see Figure 15). We identified 38 potentially intraplate events.


Only 6 earthquakes were kept as truly intraplate; they are arranged from North to South, to the East of the Samoa arc.

• Regarding Event 651 (December 24, 1929), we confirm an interplate location on the basis of five arrival times «(1 = 1.90 s), despite GUTENBERG and RICHTER'S ( 1941) opinion of insufficient data.


• It is worth discussing in detail Event 656 (October 4, 1937). As mentioned in OKAL (1984), a solution cannot be found fitting all arrivals reported in the ISS. While equally mediocre solutions can be found at various depths in the Tonga trench, they all violate the arrival time at Tucson (TUC). We obtained a copy of the TUC seismogram, including time correction infonnation, which weights this station very strongly in the available dataset. The only epicenter compatible with TUC is the retained intraplate solution (IP-18).

• Only one recent earthquake (655 on July 4, 1983) was located in the same general area, 150 km to the North.

No focal mechanisms are available for any of these events.

Casualties

• Twenty-one earthquakes were relocated to the Fiji-Tonga subduction zone, 10 of them at intennediate and deep foci. It is particularly significant that many of these events took place in the 1970s and 1980s. An additional four were listed on the NEC tape as intraplate, but had other listings (mostly ISC) in the subduction zone, three of them intennediate or deep.

• Three events (686 on July 17, 1955; 687 on October 28, 1958; and 688 on July I, 1961) were relocated to the Vanuatu trench, their alleged location obviously the result of clerical errors.

• Two historical earthquakes (682 and 683), but also Event 684 in 1965, had too few da ta to allow a significant relocation.

• We could find no da ta regarding Event 685 in 1905, quoted by WALKER (1989).

Samoa-Phoenix-Gilbert Area

We define this area as the part of the Pacific plate South of ION between Longitudes 155°E and 172°W (see Figure 16). We use ION rather than the Equator in order to avoid splitting an interesting cluster of seismicity in the vicinity of Ocean Island and the Ralik Fracture Zone. In addition to the 225 events detected teleseismically during the 1981-1984 Gilbert Islands swann, we identified 62 potentially intraplate earthquakes in this region, only 15 of them were detennined to be truly intraplate.

Intraplate Events

• Two 1982 events (764 and 765 on February 17 and May 2, 1982) relocate to the epicenter of the Gilbert Islands swann during its period of activity (code 6). Similarly, Event 758 (February 10, 1921) originally located at the 1982 swann


epicenter, relocates 250 km to the North, but its confidence ellipse intersects the swarm site. This event suggests that activity at the swarm site may be of a recurring nature. It is remarkable that GUTENBERG and RICHTER (1941) describe this event as "doubtful, but may be correct".

• In the case of Event 751 (April 24, 1937), we confirm the ISC's epicenter at 1O.2°S and 176.00 W. The NEIC tape uses Gutenberg's solution (SEISMOLOGICAL SOCIETY OF AMERICA, 1980), at IrS, 178°Wand 200 km depth. This solution is unacceptable since its location is clearly outside the Benioff plane. Indeed, all attempts to locate this earthquake in the Fiji-Tonga subduction zone failed, and moved the epicenter to the North, regardless of starting depth and of whether or not depth was constrained in the relocation process. We confirm OKAL'S (1984) evaluation of this event as intraplate. This represents a rare case when one of B. Gutenberg's personal confident locations (SEISMOLOGICAL SOCIETY OF AMERICA, 1980) is cast in doubt. Two events in the 1950s (752, on June 26, 1956; and 753 on October 3, 1957) were relocated respectively East and West of the 1937 epicenter.

• Relocation of Event 754 (April 30, 1939) does not move it substantially from its listed epicenter on the Ontong-Java plateau. All efforts to associate it with the nearby Solomon trench, which included floating its depth, failed.

• Finally, we confirm an older event (755 on May 15, 1931) in the Tuvalu Basin, and two events (756 on March 5, 1964; and 757 on March 6, 1981) in or around the Phoenix Islands.

• We also incorporate into the catalogue four earthquakes (759-762) detected by the Enewetak and Wake Island arrays in the vicinity of Ocean Island (WALKER, 1989), as weil as a teleseismically detected one at the same location (763 on November ll, 1986).

Casualties

• Of the remaining events, 35 were successfully relocated to the trenches, 21 of these relocations involving an intermediate or deep focus. In particular, we found ISC reports of epicenters determined by LASA in the late 1960s and early 1970s most unreliable. These solutions usually violate S arrivals, and move to the Fiji-Tonga Benioff zone once hypocentral depth is allowed to float.

• Among the earthquakes relocated deep in the Tonga trench, Event 781 (April 25, 1940) is the only casualty in this region listed (but not relocated) in OKAL (1984).

• Also, Event 776 (September 12, 1935), listed by GUTENBERG and RICHTER (1954) as the only case of erroneous ISS listing inside the Pacific Basin in the period 1931-1935 (a somewhat optimistic statement judging by this study), is relocated to the Tonga Benioff zone, but shallower than, and to the North of, Gutenberg and Richter's solution.

• Event 777 (July 23, 1936), listed on the NEIC tape as intraplate, has an ISC

346 Michael E. Wysession er al. PAGEOPH,

listing in the Tonga trench, and Event 783 (July 18, 1949) on the Fiji plateau. One recent event, (803 on February 27, 1985) given as intraplate by the USGS, is listed at an intermediate depth in the Solomon trench by the ISC.

• Finally, 9 older earthquakes (804-812, dating from 1918 to 1928) had too few reported times to yield a stable solution.

Focal Mechanisms

Outside of the Gilbert Island swarm, only two events have available focal solutions: Event 757 in the Phoenix Island (DZIEWONSKI et al., 1988b), featuring thrust faulting, and Event 763 (November 11, 1986) at the Ralik site (DZIEWONSKI et al., 1987c), featuring strike-slip. They are shown on Figure 16, together with the mechanisms obtained by LA Y and OKAL (1983) for the four main events in the swarm. An additional 32 swarm events have CMT solutions listed in Table 14 (DZIEWONSKI et al., 1983a,b).

Micronesia and Marshall Islands

We define this area as the part of the Pacific plate bounded by Latitude ION, Longitude 170oW, and Latitude 30oN, and to the Northeast by a great circle from 30oN, 1600 E to 15°N, 175°W. We identified in this area 76 potentially intraplate earthquakes. Epicenters are plotted on Figure 17. Because of the large number of islands and seamounts in this region, and in order to avoid cluttering of the figure, contouring of the bathymetry is only at 2000 m intervals. West of 160oE, the southem boundary is moved to the Equator.


We relocated 15 events and confirmed 5 more as genuine intraplate earthquakes. Among those, Event 914 (March 12, 1974) was feit on Moen (Truk), and Event 902 (April 18, 1981) on Mokil Atoll, halfway between Pohnpei and Kosrae.

• Also, Event 908 (May 16, 1925) is confirmed at a location not significantly different from that listed by the ISS. It is noteworthy that GUTENBERG and RICHTER ( 1941) list it as "cannot be rejected definitely; to bring it into the active belts would require an [unacceptable] error". In their later edition (GUTENBERG and RICHTER, 1954), they suggest that this earthquake could be in New Guinea, although that location could not match the quality of our solution.

• Finally, we regard Event 876 (September 29, 1926 at 05:16) as probably intraplate. Confidence ellipses with very large noise (0- = 20 s) do not intersect the trenches. It is probable that its aftershock 27 minutes later (Event 936) is at the


same location but data are insufficient and we could not relocate it. Both were listed as doubtful by Gutenberg and Richter.

• In addition, WALKER (1989) lists 24 events detected at the Enewetak and Wake arrays. Our index pattern follows the broad lineations described in the main text.

Casualties

We relocated 14 events as belonging to the various subductions zones around the Pacific, including New Britain, the Mariana Arc, Japan, the Kurile arc, and even the Aleutian trench near Unimak !sland (Event 929 on May 2, 1939).

• Among them, Event 921 (September 19, 1923) was relocated to the Mariana trench, rather than Tonga, as proposed by GUTENBERG and RICHTER (1941).

• Of particular interest are three cases of recent events grossly mislocated by the ISC:

• The ISC location for Event 931 (May 18, 1974) is poor (0' = 5.2 s on 6 P waves), and misfits the S time at Lae by more than one minute. See main text for discussion.

• Regarding Event 932 (January 16, 1976, 15:34), the ISC solution is unsatisfactory (0' = 6.2 s on 14 P waves). Removal of MAT and CLL leads to an excellent solution (0' = 0.8 s on 12 times) in the Kurile subduction zone.

• In the case of Event 933 (February 20, 1976), the ISC location, of poor quality (0' = 5.2 s on 11 stations), is totally controlled by two stations (Lae and Lamington) in Papua New Guinea. Removal of these stations yields an excellent solution (0' = 0.3 s on 9 stations) off Hokkaido. The New Guinea times are interpreted as due to local activity; indeed two other events are listed in the area less than 12 hours away from these reports.

• Five older events (934-938, dating from 1918 to 1928) had too few listings to provide a stable relocation. Among them, Event 935 (May 21, 1918) is listed by GUTENBERG and RICHTER (1941) as "may be anywhere in the Pacific area"; we also concur with these authors that Events 937 (May 14, 1923) and 938 (March 3, 1928) have insufficient data.

• Finally, six events were found to have listings clearly resulting from clerical errors (939-944), and seven are nuclear explosions at the Bikini and Enewetak test sites in the 1950s (945-951; see Section 4).

Focal Mechanisms

Only one focal solution is available in this region, a CMT mechanism for Event 885 (March 22, 1982), featuring thrust faulting (DZIEWONSKI et al., 1983a).


Off the Coast of Japan and the Kuriles

We define this region as the part of the Pacific plate North of 300 N and West of 1600 E (see Figure 18). It contains 43 potentially intraplate earthquakes.


• Intraplate earthquakes are indexed by order of decreasing latitude. We relocated or confirmed five recent events as intraplate (1001-1005), as weIl as a single older shock (1006, November 11, 1936).

Casualties

• Event 1043 (the oldest earthquake in our dataset) has no arrival time information, and Event 1042 (August 5, 1931) failed to converge to a stable solution.

• All 34 remaining earthquakes can be relocated to the subduction system. Among the latter, at least 15 are small aftershocks of major events in the Japan-Kurile subduction zone. In many instances, ISC epicenters (of poor quality in the first place) gave extremely poor fit to reported S arrivals. Their inclusion in the relocation algorithm, as weIl as the removal of a few (often a single) reported emergent P arrival was enough to bring the epicenter back into the cluster of aftershocks. We believe that the ISC mislocations are in such cases due to a single erroneous P arrival time, itself the result of amispick due to insufficient signal-to-noise ratio in the coda of larger aftershocks.

Focal Mechanisms

Only one event has a published solution: Event 1007 (March 7, 1988), featuring thrust faulting (DZIEWONSKI et al., 1989a).

Northern Boundary

We discuss here the region extending along the Northem border of the Pacific plate, East of 1600 E. This area is bounded to the South by Latitudes 400 N West of 135°W, and 300 N East of that line (see Figure 19). We identified 58 potentially intraplate events in this region.


• We relocated five historical earthquakes (1105, 1112, 1113, 1114, and 1116) and two recent ones (1103, January 18, 1976; and 1106, August 14, 1980) as truly

Vol. 135, 1991 Intraplate Seisrnicity 349

intraplate. We further confinned 13 more from recent catalogues, including Events 1101 and 1102 (April 28, 1968), the Emperor Trough earthquakes studied by STEIN (1979). Earthquakes in this region are arranged by increasing longitude to the Alaskan corner, then by decreasing latitude.

Casualties

• We relocated 29 events to the circum-Pacific subduction zones, among which were two foreshocks ( 1124 and 1125) and an aftershock (1126) of the great Kamchatka earthquake of February 3, 1923. Our relocation of the main foreshock (Event 1125, February 2, 1923 at 05:07) is consistent with Gutenberg's location (SEISMOLOGICAL SOCIETY OF AMERICA, 1980). It is probable that Event 1152 (February 5, 1923), another aftershock in this sequence, for which too few data were available, is also in the subduction zone.

• Similarly, we relocate to the subduction zone four aftershocks of the great 1965 Rat Island earthquake (1140-1142 and II 44).

• Events 1123 (October 6, 1921), 1128 (August 27, 1927) and 1131 (April 21, 1930) were described by GUTENBERG and RICHTER (1941) as very doubtful. We relocate all three to the Kurile trench.

• We relocate one event (1137, July 15, 1954) to the Queen Charlotte Transfonn Fault, two to the Blanco Transfonn Fault (1135 on July 21, 1944, and 1145 on December 28, 1967) and one to the Mendocino Transfonn Fault (1139 on July 23, 1964).

• In addition, two old events (1154 and 1155, from 1923 and 1927 respectively) failed to converge, two were blatant clerical errors (1156 on July 29, 1958; and 1157 on June 7, 1961), and one event (1158, May 11, 1962) is the underwater nuclear test "SWORDFISH" (see Section 4).

Focal Mechanisms

In addition to STEIN'S (1979) dip-slip focal solution for Event 1101, a CMT mechanism is available for Event 1117 (DZIEWONSKI et al., 1983a), featuring strike-slip.

West and South of Baja California

This region extends from 300 N to the Equator, East of 135°W (see Figure 20). In addition to the 69 earthquakes detected during the 1984 swann South of Baja California (see Appendix A), we identified another 69 potentially intraplate events in this region.



• We retained 25 confirmed intraplate earthquakes, in addition to the 1984 swarm, in this part of the Pacific Basin. For five of those (1203, 1204, 1214, 1219, and 1223), we use the results of the WIENS and OKAL (1987) study, carried on with similar techniques. Events 1212 and 1218 were described by GUTENBERG and RICHTER (1941) as doubtful, but we confirm their intraplate character with residuals (J < 3 s.

Casualties

• Most remarkable in this region are 17 blatant mislocations due to clerical errors: 15 belong to the Philippine trench (1252-1266) between 1931 and 1941, one (1267) is in Central Utah in 1957, and another one a small Southern California shock detected only at Pasadena (1268, June 22, 1971).

• We relocated to the East Pacific Rise or the Middle America trench 21 earthquakes; two more (1239, December 12, 1938; and 1247, June 6, 1958) were listed by the NEIC tape as intraplate but had alternate listings in the Central American trench.

• Three older events (1249-1251, dating from 1919 to 1927) had too few reported times to allow a stable relocation.

• Event 1269 (May 14, 1955) is the underwater nuclear test "WIGWAM" (see Section 4).

Focal Mechanisms

Apart from three events during the 1984 swarm, WIENS and OKAL (1987) obtained a solution for Event 1204 (June 30, 1945), featuring normal faulting. No other mechanism is available in this region.

Northcentral Pacific Basin

This region comprises the rest of the Pacific plate, extending roughly from 1600 E to 135°W at latitude 40oN, and from 175°W to 135°W at the Equator (see Figure 21). It excludes Hawaii, defined as extending from l7°N to 23°N and 152°W to 162°W. This area has an extremely low level of teleseismically recorded activity, with only 20 potentially intraplate events.


• We relocated or confirmed only five events as intraplate, which we have sorted by increasing longitude. By far the largest (mb = 5.5) and most interesting event in this region is the Tern Island earthquake of September 22, 1988 (Event 1302). Its


focal mechanism was given by JIMENEZ et al. (1989) and DZIEWONSKI et al (1989c). This is the only event known to occur on the Hawaiian ridge away from the active hotspot.

Casualties

• By letting hypocentral depth be unconstrained, we relocated deep into the Tonga Beniotf plane Event 1306 (May 14, 1976). In addition, we relocate an older event (1307, October 11, 1928 at 23:32) to the Kurile subduction zone, our solution fitting all but 1 P times. It is probable that a second event (1311) 12 minutes later (and for which too few data were available for stable relocation) also belongs to this system. GUTENBERG and RICHTER (1941) had suggested that these events belong to the Aleutian trench, but no solution of comparable quality could be achieved in that region.

• Six additional events have blatantly wrong locations, presumably due to clerical errors (including Event 1315 on September 21, 1951, actually in the Hindu Kush).

• Similarly, Event 1313 (July 24, 1938) relocates to the Imperial Valley. The correct latitude in the originallocation, and the simultaneous occurrence of a large swarm in the Imperial Valley at the time, argue that this mislocation was the result of a clerical error.

• We refer to the main text for a detailed discussion of the "ghost" event, Number 1320 (April 9, 1967); its listing is the result of a gross misinterpretation of observations.

• Regarding Event 1312 (March 17, 1976), the ISC solution is of poor quality (0- = 6 s on 6 stations). Despite the recent date of this event, we were unable to identify more arrivals at high-quality WWSSN stations located in the vicinity of stations having reported arrivals (COL, LPS). We regard the available data as insufficient to provide astahle epicenter.

• In the case of Event 1309 (September 17, 1932), the only report is that it was feit by the SS Mericos S. Whittier at location 26.3°N, 148.6°W (NEUMANN, 1934). The isolated character of this location away from plate boundaries would suggest that this shock would have to be intraplate. We report this event as (code 11) "not enough information to allow relocation".

• Two older events (1308 on June 17, 1917 and 1310 on October 1, 1918) had too few data for stable, if any, relocation. GUTENBERG and RICHTER (1941) had particularly harsh comments on the quality of their ISS listings ("No evidence for the indicated epicenter" for Event 1308 and "A badly forced solution" for Event 1310).

Cocos Plate

In addition to the 28 events of the 1976 swarm, we identified a total of 62 potentially intraplate events inside the Cocos plate (see Figure 22).


lntraplate Earthquakes

• We relocated or confirmed 22 events as genuinely intraplate inside the Cocos plate. As discussed in the main text, some earthquakes have a tendency to form distinct clusters; consequently, we have identified sites with a location code starting with CO- for "Cocos".

• Regarding Event 1367 (October 11, 1975), we regard the ISC depth (127 km) as inconsistent with the location of the event removed from a Benioff zone. This depth was probably suggested by the Huancayo report of a later arrival, identified as pP by the ISC. While we have recognized a second arrival at Tueson, it is only 13.5 s after P, suggesting a depth of about 50 km, if interpreted as pP, or 30 km as sP. The latter is on the order of the maximum depth of reported intraplate seismicity for oceanic lithosphere of 40-45 Ma age (WIENS and STEIN, 1983). It is worth noting that a number of seamounts are present in the immediate vicinity of CO-5; if the event was associated with magmatic activity, higher strain rates could allow brittle fracture at greater depths. The discrepancy between the two time lags at HUA and TUC argues against a multiple event. Constraining the depth at 30 km, we obtain an excellent solution, with u < 1 s. We can only speculate on the origin of the second arrival at HUA, which was not available for inspection.

No focal mechanisms are available for any events in the Cocos plate, excepting the 1976 swarm mainshock (see Appendix A).

Casualties

Thirty-three events had alternate bulletin listings or were relocated to the adjacent plate boundaries.

• The available dataset for January 4, 1918 is extremely confused due to the apparent occurrence of two events within 150 s of each other (1405 and 1406). As a result, no relocation of either event could be performed.

• Similarly, Event 1407 (November 16, 1918) has only three body wave arrivals listed; we list it as "unsufficient data", but it is worth noting that the 3 times converge on the Middle America trench in Southern Mexico).

• An additional historical event (1408, June 7, 1937) could not be relocated in a stable way.

• Regarding event 1409 (December 5, 1966), the ISC solution is poor (u = 6.7 s on 7 stations). It is remarkable that the event is followed by a sequence of 4 weil located shocks on the Galapagos Ridge. Such a location would fit 5 of the 7 stations with u = 1.2 s. We list the event as having nonconverging data, but regard it as most probably interplate.

• In the case of two events (1411, March 28, 1958; and 1412, January 26, 1962), we could not obtain arrival time information.


Nazca Plate

In addition to the events recognized as part of the Galapagos swarms (see Appendix A), a total of 111 potentially intraplate events were identified in the Nazca plate (see Figure 23).

Intraplale Earthquakes

We relocated or confirmed 45 earthquakes inside the Nazca plate, the majority having a tendency to cluster at locations correlating with the main bathymetric features of the plate; we have ordered and numbered the events following decreasing latitudes of these clusters, wh ich have been given codes starting with NZ- (for "Nazca").

• For Event 1505 on Aug 20, 1959, the interesting case arose where the Sand P

waves individually yielded similar epicenters but times mutually incompatible. When combined, 14 S waves averaged 8.8 s slower than predicted for the epicenter based on P arrivals (the median lag was + 12.2 s). This unusual result could be the result of poorly selected S waves, selection of sS instead of S arrivals or perhaps even a deviation of the Poisson ratio for the upper mantle from that used in the Jeffreys-Bullen tables. This was the only event from Si te NZ-2 containing a significant number of S arrivals.

Casualties

• We isolated 14 events for which no sufficient data were available to perform a significant relocation. Among those are six events in February 1917 and September, 1918, listed in the ISS at 26°S and 800 W (1597-1600, 1603 and 1611). It is probable, as repeatedly stated in the ISS, that these epicenters are common. If an intraplate epicenter could be confirmed, it would represent a rather unique spacio-temporal accumulation of intraplate stress release, in view of the numerous reports of Love and Rayleigh waves in Europe which would suggest magnitudes of at least 5 1/2. The 1918 events (1600 and 1611) are probably one and only, this clerical error resulting from the use of a nonstandard time at two Argentinian stations.

• Finally, we relocated 27 events to the South American or Middle American trenches, and 24 to the nearby ridge systems.

Focal Mechanisms

In addition to MENDIGUREN'S (1971) thrust solution for Event 1519, normal faulting CMT solutions are available for the Easter Island events of 1984 and 1987


(1528,1529,1530) (DZIEWONSKI et al., 1987c; 1988c) (see Figure 23). Focal Solutions for Galapagos swarrn events are given separatelyon Figure A-6.

Antarctic Plate

We identified 26 potentially intraplate events in the Pacific Ocean part of the Antarctic plate and relocated or confirrned 15 as genuine intraplate earthquakes (see Figure 24).


• As discussed in the main text, we confirm 11 events listed by OKAL (1981) as truly intraplate.

• In addition, we identified two events (1711 on December 5, 1927; and 1710 on April 29, 1975) listed in the ISSjISC, but not on the NEIC tape, and two new events (1709 and 1712), since the completion of OKAL'S (1981) study in 1980.

Casualties

• Of the events listed by OKAL (1981), we did not consider the August 13, 1937 earthquake since its epicenter (57.4°S, 129.9°W) is less than 2° from the plate boundary. In addition, we confirm that Event 1720 (January 18, 1949) is actually interplate on the Nazca-Antarctica boundary.

• Relocation of Events 1717 and 1718 on December 27, 1926 moved them further inside the Antarctic plate than proposed by OKAL (1981). However, as already noted in that study, these locations are very poorly constrained and confidence ellipses run with l-u noise of 10 s (a conservative value in the 1920s) intersect the Menard Transforrn Fault at 500 S and 115°W. We regard these events as presumably interplate.

• Event 1723 (January 17, 1967) is listed in the ISC catalogue through a single report from LASA; we inspected records from regional high-gain WWSSN stations (SPA, LPB), and could find no trace of the event. In view of the large errors often associated with LASA slowness solutions, we regard this event as unsupported by data.

• As discussed in the main text, the listing for Event 1726 (September 13, 1975) is probably the result of a clerical error.

Focal Mechanisms

In addition to FORSYTH'S (1973) ridge-push mechanism for Event 1702, CMT Solutions are now available for two earthquakes in the Antarctic plate: Event 1714


was studied by OKAL (1980). The new CMT solution is basically identical to his mechanism and confinns the release of horizontal ridge-push (DZIEWONSKI et al., 1988a). A thrust mechanism is available for Event 1715 (DZIEWONSKI et al., 1987b).

Acknowledgments

This work would not have been feasible without the assistance of many scientists and library Staff who provided us with bulletins and seismograms. We are especially grateful to Daniel A. Walker, who gave us a computerized version of WALKER (1989) in advance of publication. Our interest in Pacific seismicity owes a great deal to many years of collaboration and friendship between one of us (EAO) and Jacques Talandier. We especially thank hirn for the update of the Polynesian dataset. The use of the data facilities at Lamont-Doherty Geological Observatory, Caltech's Seismological Laboratory, the Berkeley Seismographic Station, and Saint Louis University is gratefully acknowledged. Most figures were plotted using software developed by Paul Lundgren and Paul Stoddard. We thank Mark Woods for sifting through the Micronesian newspaper collection at the Hamilton Library of the University of Hawaii. Alice Gripp and Richard Gordon made available their absolute plate motions model before publication. We thank Eric Bergman and Doug Wiens for thoughtful reviews of this voluminous work. This research was supported by JOI, Inc., under Contract Number OCE-84-09478. KLM was the recipient of a Pew Foundation Fellowship at Northwestern University. Additional and Partial funding came from the Office of Naval Research, under contract NOOOI4-89-J-1663, and a Northwestern University Research Grant.

Note added in prooj (September 16, 1990)

Since this paper was accepted, we have received confinnation that events 1541 and 1542 were feit at the Juan Fernandez Islands (E. W. Bravo M., pers. comm., 1990). It is also worth mentioning a burst of activity (including a mb = 6.1 event) at the eastern end of Region NZ-8, around 27°S and 104°W (August 21, 22; September 2, 1990).

REFERENCES

AKI, K., and RICHARDS, Quantitative Seismofogy (San Francisco, W. H. Freeman and Co. 1980),932 pp.

AMERICAN ASSOCIATION OF PETROLEUM GEOLOGISTS, Pfate Tectonic Map o[ the Circum Pacific Region (Map, Tulsa 1981).


ANONYMOUS, Announced United States Nuclear Tests, July 1945 through December 1988 (U.S. Dept. of Energy, Office of External Affairs, Publication NVO-209, Washington, D.C. 1989), 64 pp.

BERGMAN. E. A., and SOLOMON, S. C. (1980), Oceanic Intraplate Earthquakes: Implications Jor Local and Regional Intraplate Stress, J. Geophys. Res. 85, 5389-5410.

BERG MAN, E. A., and SOLOMON, S. C. (1984), Source Mechanisms oJ Earthquakes Near Mid-ocean Ridges Jrom Body WaveJorm Inversion: Implications Jor the Early Evolution oJ the Oceanic Lithosphere, J. Geophys. Res. 89, 11415- II 44 I.

BRATI, S. R., BERGMAN, E. A., and SOLOMON, S. C. (1985), Thermoelastic Stress: How Important as a Cause oJ Earthquakes in the Young Lithosphere?, J. Geophys. Res. 90, 10249- 10260.

CAPON, J. (1974), Characterization oJ Crust and Upper Mantle Structure under LASA as a Random Medium, BuB. Seismol. Soc. Am. 64, 235-266.

CASTILLO, P. R., Geology and Geochemistry oJ Cocos Island, Costa Rica: Implications Jor the Evolution oJ the Aseismic Cocos Ridge, Ph.D. Thesis (Washington University, St. Louis 1987), 223 pp.

CHAPPLE, W. M., and FORSYTH, D. W. (1979), Earthquakes and Bending oJ Plates at Trenches, J. Geophys. Res. 84, 6729-6749.

CHASE, C. G. (1971), Tectonic History oJthe Fiji Plateau, GeoI. Soc. Am. BuB. 82,3087-3110. CHEN, T., and FORSYTH, D. W. (1978), A Detailed Study oJ Two Earthquakes Seaward oJ the Tonga

Trench: Implications Jor Mechanical Behavior oJ the Lithosphere, J. Geophys. Res. 83, 4995-5003. CHOUDHURY, M. A., Observation des ondes de volume ayant traverse le manteau et le noyau, In Traite

de Geophysique Interne (eds. Coulomb, J., and Jobert, J.) (Masson, Paris 1973), pp. 283-320. DAVEY, F. J., and CHRISTOFFEL, D. A. (1978), Magnetic Anomalies across Campbell Plateau, New

Zealand, Earth Planet. Sci. Letts. 41, 14-20. DUDA, S. J. (1965), Secular Seismic Energy Release in the Circum-Pacific Belt, Tectonophysics 2, 409-452. DZIEWONSKI, A. M., FRIEDMAN, A., GIARDINI, D., and WOODHOUSE, J. H. (1983a), Global Seismicity

oJ 1982: Centroid Moment-tensor Solutions Jor 308 Earthquakes, Phys. Earth Planet. Inter. 33, 76-90. DZIEWONSKI, A. M., FRIEDMAN, A., and WOODHOUSE, J. H. (I 983b), Centroid Moment-tensor

Solutions Jor January-March 1983, Phys. Earth Planet. Inter. 33, 71 -75. DZIEWONSKI, A. M., FRANZEN, J. E., and WOODHOUSE, J. H. (1983c), Centroid Moment-tensor

Solutions Jor April-June 1983, Phys. Earth Planet. Inter. 33, 243-249. DZIEWONSKI, A. M., FRANZEN, J. E., and WOODHOUSE, J. H. (I 984a), Centroid Moment-tensor

Solutions Jor July-September 1983, Phys. Earth Planet. Inter. 34, 1-8. DZIEWONSKI, A. M., FRANZEN, J. E., and WOODHOUSE, J. H. (I 984b), Centroid Moment-tensor

Solutions Jor January-March 1984, Phys. Earth Planet. Inter. 34, 209-219. DZIEWONSKI, A. M., EKSTRÖM, G., FRANZEN, J. E., and WOODHOUSE, J. H. (l987a), Global Seismicity

oJ 1978; Centroid Moment-tensor SolutionsJor 512 Earthquakes, Phys. Earth Planet. Inter. 46,316-342. DZIEWONSKI, A. M., EKSTRÖM, G., FRANZEN, J. E., and WOODHOUSE, J. H. (1987b), Global

Seismicity oJ 1979; Centroid Moment-tensor Solutions Jor 524 Earthquakes, Phys. Earth Planet. Inter. 48, 18-46.

DZIEWONSKI, A. M., EKSTRÖM, G., WOODHOUSE, J. H., and ZWART, G. (1987c), Centroid Momenttensor Solutions Jor October-December 1986, Phys. Earth Planet. Inter. 48,5-17.

DZIEWONSKI, A. M., EKSTRÖM, G., FRANZEN, J. E., and WOODHOUSE, J. H. (I 988a), Global Seismicity oJ 1980; Centroid Moment-tensor Solutions Jor 515 earthquakes, Phys. Earth Planet. Inter. 50, 127-154.

DZIEWONSKI, A. M., EKSTRÖM, G., FRANZEN, J. E., and WOODHOUSE, J. H. (I 988b), Global Seismicity oJ 1981; Centroid Moment-tensor Solutions Jor 542 Earthquakes, Phys. Earth Planet. Inter. 50,155-182.

DZIEWONSKI, A. M., EKSTRÖM, G., WOODHOUSE, J. H., and ZWART, G. (1988c), Centroid Momenttensor Solutions Jor July-September 1987, Phys. Earth Planet. Inter. 53, I - I I.

DZIEWONSKI, A. M., EKSTRÖM, G., WOODHOUSE, J. H., and ZWART, G. (l989a), Centroid Momenttensor Solutions Jor January-March 1988, Phys. Earth Planet. Inter. 54, 22-32.

DZIEWONSKI, A. M., EKSTRÖM, G., WOODHOUSE, J. H., and ZWART, G. (1989b), Centroid Momenttensor Solutions Jor April-June 1988, Phys. Earth Planet. Inter. 54, 199-209.

DZIEWONSKI, A. M., EKSTRÖM, G., WOODHOUSE, J. H., and ZWART, G. (I 989c), Centroid Momenttensor Solutions Jor July-September 1988, Phys. Earth Planet. Inter. 56, 165- 180.


FILSON, J., SIMKIN, T., and LEU, L.-K. (1973), Seismicity of a Caldera Collapse: Gatapagos lslands, 1968, J. Geophys. Res. 78, 8591-8622.

FLEMING, C. A. (1970), The Mesozoic of New Zealand: Chapters in the History of the Circum-Pacific Mobile Belt, Quat. J. Geol. Soc. London 125, 125-170.

FORSYTH, D. W. (1973), Compressive Stress Between two Mid-oceanic Ridges, Nature 243, 78-79. FORSYTH, D. W., and UYEDA, S. (1975), On the Relative lmportance of the Driving Forces of Plate

Motion, Geophys. J. Roy. astr. Soc., 43, 163-200. FUKAO, Y. (1979), Tsunami Earthquakes and Subduction Processes near Deep-sea Trenches, J. Geophys.

Res. 84, 2303-2314. GRIPP, A. E., and GOROON, R. G. (1989), Current Plate Motions Relative to the Hotspots, EOS, Trans.

Am. Geophys. Uno 70, 1351 [abstract]. GUTENBERG, B., and RICHTER, C. F. (1941), Seismicity of the Earth, Geol. Soc. Am. Spec. Pap. 34,

1-3l. GUTENBERG, B., and RICHTER, C. F., Seismicity of the Earth and Associated Phenomena (Princeton

University Press 1954), 2nd ed., 310 pp. HAMBURGER, M. W., EVERINGHAM, I. B., ISACKs, B. L., and BA RAZANGI , M. (1990), Seismicityand

Crustal Structure of the Fiji Plat/orm, Southwest Pacific, J. Geophys. Res. 95, 2553-2573. HA YES, D. E., and EWING, M., The Louisville Ridge: a possible extension of the Eltanin fracture zone, In

Antarctic Oceanology I, (ed. Reid, J.L.) Antarctic Res. Sero 15 (Amer. Geophys. Un., Washington, D.C. 1971), pp. 223-228.

HERRON, E. M. (1972), Sea-floor Spreading and the Cenozoic History of the Eastcentral Pacific, Geol. Soc. Am. Bull. 83, 1671-1692.

HEY, R. N. (1977), Tectonic Evolution of the Cocos-Nazca Spreading Center, Geol. Soc. Am. Bull. 88, 1404-1420.

HOLDEN, J. c., and DIETZ, R. S. (1972), Galapagos Gore, NazCoPac Tripie Junction, and CarnegiejCocos Ridges, Nature 235, 266-269.

HouTz, R. E. (1975), South Tasman Basin and Borderlands: A Geophysical Summary, Init. Repts. Deep Sea Drilling Proj. 29, 1135-1146.

JIMENEZ, E., CARA, M., ROULAND, D., and OKAL, E. A. (1989), Fault Plane Solutions from Single-station Surface-wave Data, EOS, Trans. Am. Geophys. Uno 70, 397 [abstract].

JOROON, T. H., and SVERDRUP, K. A. (1981), Teleseismic Location Techniques and their Application to Earthquake Clusters in the Southcentral Pacific, Bull. Seismol. Soc. Amer. 71,1105-1130.

KAUFMAN, K., and BURDICK, L. J. (1980), The Reproducible Earthquakes of the Gatapagos lslandv, Bull. Seismol. SOC. Am. 70, 1759-1770.

KLEIN, F. W., KOYANAGI, R. Y., NAKATA, J. S., and TANIGAWA, W. R. (1987), The Seismicity of Kilauea's Magma System, U.S. Geological Survey Prof. Paper 1350, 1010-1186.

KLITGORD, K. D., and MAMMERICKX, J. (1982), Northern East Pacific Rise: Magnetic Anomaly and Bathymetric Framework, J. Geophys. Res. 87, 6725-6750.

KROENKE, L. W., and WALKER, D. A. (1986), Evidence for the Formation of a New Trench in the Western Pacific, EOS, Trans. Am. Geophys. Uno 67, 145-146.

LAHR, J. c., PAGE, R. A., STEPHENS, C. D., and CHRISTENSEN, D. H. (1988), Unusual Earthquakes in the Gulf of Alaska and Fragmentation of the Pacific Plate, Geophys. Res. leUs. 15, 1483-1486.

LAY, T., and OKAL, E. A. (1983), The Gilbert lslands (Republic of Kiribati) Earthquake Swarm of 1981-83, Phys. Earth Planet. Inter. 33, 284-303.

MAMMERICKX, J., and KLITGORD, K. D. (1982), Northern East Pacific Rise: Evolutionfrom 25m.y.b.p. to the Present, J. Geophys. Res. 87, 6751-6759.

MAMMERICKX, J., HERRON, E. M., and DORMAN, L. M. (1980), Evidence for two Fossil Ridges in the Southeast Pacific, Geol. Soc. Am. Bull. 91, 263-27l.

MENDIGUREN, J. A. (1971), Foral Mechanism of a Shock in the Middle of the Nazca Plate, J. Geophys. Res. 76, 3861-3879.

MINSTER, J.-B., JORDAN, T. H., MOLNAR, P., and HAINES, E. (1974), Numerical Modelling of lnstantaneous Plate Tectonics, Geophys. J. Roy. astr. Soc. 36, 541-576.

MIYAMURA, S., Some remarks on historical seismogram.v and the microfilming project, In Historical Seismograms and Earthquakes of the World (eds. Lee, W. H. K., Meyers, H., and Shimazaki, K.) (San Diego, Academic Press 1988), pp. 40 1-419.


MOLNAR, P. (1988), Continental Tectonics in the Aftermath of Plate Tectonics, Nature 335, 131-137. MUIRHEAD, K., and ADAMS, R. D. (1986), Earthquakes in the "Aseismic" Regions of the Western

Pacific, Geophys. Res. LeUs. 13, 169-172. NEUMAN, F., United States Earthquakes, 1932 (U.S. Dept. Commerce, Washington, D.C 1934),21 pp. OKAL, E. A. (1980), The Bellingshausen Sea Earthquake of February 5, 1977: Evidence for Ridge-gener

ated Compression in the Antarctic Plate, Earth Planet. Sci. LeU. 46, 306-310. OKAL, E. A. (1981), lntraplate Seismicity of Antarctica and Tectonic lmplications, Earth Planet Sci. Leu.

52, 397-409. OKAL, E. A. (1983), Oceanic lntraplate Seismicity, Annual Rev. Earth Planet. Sei. 11, 195-214. OKAL, E. A. (1984), lntraplate Seismicity ofthe Southern Part ofthe Pacific Plate, J. Geophys. Res. 89,

10053-10071. OKAL, E. A., and BERGEAL, J.-M. (1983), A Modelfor the Plate Tectonics Evolution ofthe Eastcentral

Pacific Based on SEASAT lnvestigations, Earth Planet. Sei. Lets. 72,99-116. OKAL, E. A., and KUSTER, G. (1975), A Teleseismic Array Study in French Polynesia: lmplications for

Local and Distant Structures, Geophys. Res. Lett. 2, 5-8. OKAL, E. A., TALANDlER, J., SVERDRUP, K. A., and JORDAN, T. H. (1980), Seismicity and Tectonic

Stress in the Southcentral Pacific, J. Geophys. Res. 85, 6479-6495. OKAL, E. A., WOODS, D. F., and LAY, T (1986), lntraplate Deformation in the Samoa-Gilbert-Ralik

Area: APrelude 10 a Change of Plate Boundaries in the Southwest Pacific?, Tectonophysics 132,69-78. PELAYO, A. M., and WIENS, D. A. (1989), Seismotectonics and Relative Plate Motions in the Scotia Sea

Region, J. Geophys. Res. 94, 7293-7320. RICHARDSON, R. M., SOLOMON, S. C, and SLEEP, N. H. (1979), Tectonic Stress in the Plates, Rev.

Geophys. Space Phys. 17, 981-1008. SBAR, M. L., and SYKES, L. R. (1977), Seismicity and Lithospheric Stress in New York and Adjoining

Areas, J. Geophys. Res. 82,5771-5786. SEISMOLOGICAL SOCIETY OF AMERICA, Seismology Microfiche Publications from the Caltech Archives,

Sero II and III (eds. Goodstein, J. R., Kanamori, H., and Lee, W. H. K.) (Berkeley, 1980). SIMKIN, T., SIEBERT, L., MCCLELLAND, L., BRIDGE, D., NEWHALL, C, and LATTER, J. H., Volcanoes

of the World (Stroudsburg, Pennsylvania, H. Ross Publishing Co. 1981), 233 pp. STEIN, S., (1979), lntraplate Seismicity on Bathymetric Features: The 1968 Emperor Trough Earthquake,

J. Geophys. Res. 84, 4763-4768. STEIN, S., and OKAL, E. A., Seismological studies of the deformation of oceanic lithosphere, In Space

Geodesy and Geodynamics (eds. Anderson, A. J., and Cazenave, A.) (London, Academic Press 1986), pp. 407 -450.

STEIN, S., OKAL., E. A., and WIENS, D. A., Application of modern techniques to analysis of historical earthquakes, In Historical Seismograms and Earthquakes of the World (eds. Lee, W. H. K., Meyers, H., and Shimazaki, H.) (London, Academic Press 1988), pp. 85-104.

STEWART, L. M., and OKAL, E. A. (1983), Seismicity and Aseismic Slip along the Eltanin Fracture Zone, J. Geophys. Res. 88, 10495-10507.

STODDARD, P. R. (1987), A Kinematic Model for the Evolution of the Gorda Plate, J. Geophys. Res. 92, 11524-11532.

SVERDRUP, K. A., Seismotectonic Studies in the Pacific Ocean Basin, Ph.D. Thesis (University of California, San Diego 1981),436 pp.

SVERDRUP, K. A., and OKAL, E. A. (1987), Cluster Relocations of the Gilbert lslands Swarm, EOS, Trans. Am. Geophys. Uno 68, 363 [abstract).

SYKES, L. R. (1978), lntraplate Seismicity, Reactivation of Preexisting Zones of Weakness, Alkaline Magmatism and Other Forms of Tectonism Postdating Continental Fragmentation, Rev. Geophys. Space Phys. 16, 621-688.

SYKES, L. R., and SBAR, M. L., Focal meehanism solutions of intraplate earthquakes and stresses in the lithosphere, In Geodynamies of lceland and the North Atlantie Area (ed. Kristjansson, L.) (Dordrecht, D. Reidel 1974), pp. 207-224.

TAGGART, J. N., LEE, W. H. K., and MEAGHER, K. L., The U.S. Geologieal Survey's databases of seismie events, In Historieal Seismograms and Earthquakes of the World (eds. Lee, W. H. K., Meyers, H., and Shimazaki, K.) (London, Academic Press 1988), pp. 142-149.


TALANDIER, J., and KUSTER, G. T. (1976), Seismicity and Submarine Volcanic Activity in French Polynesia, J. Geophys. Res. 81, 936-948.

TALANDIER, J., and OKAL, E. A. (1982), Crises sismiques au volcan Macdonald (Ocean Pacifique Sud), C.R. Aead. Sei. Paris, Ser. 11. 295, 195-200.

TALANDIER, J., and OKAL, E. A. (1984a), The Volcanoseismic Swarms of 1981-1983 in the TahitiMehetia Area, French Polynesia, J. Geophys. Res. 89, 11216-11234.

TALANDIER, J., and OKAL, E. A. (1984b), New Surveys of Macdonald Seamount, Southcentral Pacific, Following Volcanoseismic Activity, 1977-1983, Geophys. Res. Lett. 11, 813-816.

TALANDIER, J., and OKAL, E. A. (1987), Seismic Detection of Underwater Volcanism: The Example of French Polynesia, Pure Appl. Geophys. 125, 919-950.

Tu RCOTTE, D. L. (1974), Are Transform Faults Thermal Contraction Cracks?, J. Geophys. Res. 79, 2573-2577.

WALKER, D. A., and MCCREERY, C. S. (1985), Significant Unreported Earthquakes in "Aseismic" Regions of the Western Pacific, Geophys. Res. Letts. 12, 433-436.

WALKER, D. A. (1989), Seismicity of the lnteriors of Plates and in the Pacific Basin, EOS, Trans. Am. Geophys. Uno 70, 1543-1544.

WEISSEL, J. K., and ANDERSON, R. N. (1978), ls There a Carotine Plate?, Earth Planet. Sei. Letts. 41, 143-158.

WIENS, D. A., Oceanic lntraplate Seismicity: lmptications for the Rheology and Tectonics of the Oceanic Lithosphere, Ph.D. Thesis (Northwestern University, Evanston, Illinois), 220 pp.

WIENS, D. A. (1987), Effects of Near-source Bathymetry on Teleseismic P Waveforms, Geophys. Res. Letts. 14, 761-764.

WIENS, D. A., and OKAL, E. A. (1987), Tensionallntraplate Seismicity in the Eastcentral Pacific, Phys. Earth Planet. Inter. 49, 264-282.

WIENS, D. A., and STEIN, S. (1983), Age Dependence of Oceanic lntraplate Seismicity and lmptications for Lithospheric Evolution, J. Geophys. Res. 88, 6455-6458.

WIENS, D. A., and STEIN, S. (1984), lntraplate Seismicity and Stresses in Young Oceanic Lithosphere, J. Geophys. Res. 89, 11442-11464.

(Received February 28, 1990, revisedjaecepted June 6, 1990)

• • • Ecology Hydrology Geology

New titles

Edited by

K. Roth H. Flühler W. A. Jury J. C. Parker

Field-Scale Water and Solute Flux in So Is In autumn 1989 the worId's leading experts in soil science gathered at the Centro Stefano Franscini to hold a symposium on the movement and reactions of water and chemicals in material field soils. Experimenters as weil as theoreticians and mathematicians discussed various aspects of transport problems which arise from the daiIy impact on the soiI by fertilizers, pesticides, acid rain or oiI and chemical hazards. The best contributions have been carefully reviewed and prepared for publication here. These are introduced by a preface providing a survey of the field and giving testimony to the state of affairs in soil science today. A special feature of the book is the «think tank» section, capturing the tenor of group session discussions and summarizing the concIusions reached by all participants coIIaboratively. This stands out as a lively account of present research in fieldscale soiI science and will be the starting point for further investigations in civil engineering and environmental protection as weil as the special areas of chemistry, physics and mathematics involved.

1990. 304 pages. Hardcover sFr. 78.-IDM 88.-ISBN 3-7643-2510-0

W. A. Jury K. Roth

Transfer Functions and Solute Movement Through Soil Theory and Applications

This book develops adescription of solute transport trough porous media in terms or transfer functions, using the travei time probability density function as the fundamental propertyor the system through which transport occurs. Existing transport process models are expressed as transfer functions and are compared with alternative approaches based on different hypotheses. Over 60 worked examples are contained in the text. and an additional 50 solved problems in an appendix. The book should be suitable for an advanced undergraduate or graduate level course on salute transport through porous media, or as a reference book for hydrologists, soiI scientists civiI or environmental engineers.

1990.228 pages. Softcover sFr. 54.-/DM 64.-ISBN 3-7643-2509-7

Birkhäuser Birkhäuser Verlag AG Basel· Boston· Berlin

Aspects of Pacific Seismicity

Documents

Transcript of Aspects of Pacific Seismicity