d
i 3' fi"' 2‘ ‘P. 1,?
i 5“; ,1 m? w ‘1'. ' ':. ‘5
i n ,- . 3‘ u
. '4- ar r . ,2,
j ..‘ ( 1! r} ‘51:.- 'q‘
i o A ( i ' h."- np§v$9 Vb ‘3:
.1 V
'4‘ x ' : \44-5
:1: ‘6' I“ ' ' . V r‘ «I
l‘ U
i
'x . \

This is to certify that the

thesis entitled

SEISMICITY AND TECTONICS OF
THE FAR WESTERN ALEUTIAN ISLANDS

presented by

James Tyler Newberry

has been accepted towards fulfillment
of the requirements for
Geological Sciences

M' S degree in

 

Date 11 November, 1983

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

MSU .

LIBRARIES

“

 

 

BEIURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

’4

 

 

 

 

 

 

 

SEISMICITY AND TECTONICS
OF THE FAR WESTERN ALEUTIAN ISLANDS

BY

James Tyler Newberry

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the Degree of

MASTER OF SCIENCE

Department of Geological Sciences

1983

ABSTRACT

SEISMICITY AND TECTONICS
OF THE FAR WESTERN ALEUTIAN ISLANDS

BY

James Tyler Newberry

The present day tectonics of the Near and Komandorsky
Islands (far western Aleutians) are studied using
seismicity, focal mechanisms, and depth phase analysis. The
Near IslandS‘ are characterized by normal faulting south of
the Aleutian Ridge and shallow thrusting near the ridge.
Results indicate underthrusting as far west as Ostrov Mednyy
with strike-slip faulting restricted to the south of Ostrov
Beringa; this disagrees with prior models which suggested
strike-slip motion through the entire region. A zone of
north-east striking left-lateral faulting, which may
separate the Aleutian Ridge from Kamchatka, is proposed near
164.5 degrees east. Normal faulting north of the
Komandorskys may be related to bouyancy of a remnant slab.
Depth phase modelling indicates bulletin reported depths are
too deep. Moment release calculations imply that there may

not be a true seismic gap in the Komandorsky Islands.

ACKNOWLEDGEMENTS

I would like to express my appreciation to Kaz Fujita
for all of his help, insight, and patience during the past
several years. Thanks to him also for helping me overcome
my great fear of raw fish, chopsticks, computers, Amtrak,
public speaking, and professors. Thanks as well Hugh
Bennett and Tom Vogel for helpful comments as members of my
committee. Thanks to F. W. Cambray who unknowingly sparked
my interest in matters of geology some 8 years ago via
public television. Special thanks go to Seth Stein, Doug
wiens and Joe Engeln at N.U. for their exceptional
hospitality and generosity and use of their seismological
data collection and surface wave analysis package. Thanks
to Bill Rogers for whiskey and shelter in Evanston. Thanks
to Glenn Kroeger and Bob Gelleeror use of their synthetic
seismogram program. Thanks to Norm Sleep for helpful
comments and fish stories. Thanks to Klaus Jacob and Bob
Engdahl for their comments and criticisms. Appreciation
goes to Bill Monaghan for sharing the macroheadaches derived
from microcomputing. Thanks to Ann Goulette for
cartographic criticisms and to the Geography Department for
the use of their digitizing facility. Special thanks go to
David LaClair, whose meticulous work and keen insight laid
' the groundwork for this study. Thanks the the Harp Brewing
Company, Elderly Instruments, and the Peanut Barrel for
assisting me in efforts to keep my head on my shoulders.
Most of all, thanks to the fellow students: Mike, Ben,
Bruce, Dave, Bill, Gary, Jody, Cindy, Bill, Don, Soo-Meen,
Dave, Mongo, Randy, and many others, past and present, whose
friendship I have enjoyed.

Support was provided by Chevron Oil Company, and

support and research were provided by NSF grant EAR
80-25267, and ONR contract no. N00014-8-3K-0693.

ii

TABLE OF CONTENTS

INTRODUCTION
GEOLOGIC SETTING

Regional Setting . ...

Geology of the Western Aleutians
SEISMICITY

Magnitude and Frequency

Event Depths

Fault Plane Solutions

Hypocentral Relocations . . . . . . . . .
DISCUSSION OF RESULTS
CONCLUSIONS . . . . . . .
APPENDIX A: Geology of the Komandorsky Islands
APPENDIX B: Data Analysis Techniques
APPENDIX C: Focal Mechanisms

APPENDIX D: Surface Wave Mechanism

iii

C‘J—‘D

ll
11
20
33
40
44
51
53
58
68
86

LIST OF TABLES

l Earthquakes for which mechanisms are presented . . . 34

iv

LIST OF FIGURES

l Seismicity of the Aleutian Arc . . . . . . . . . . . 2
2 Basins of the Bering Sea . . . . . . . . . . . . . . 5

3 Seismic Gaps of the Aleutians, from Sykes et al.,

(1981) . . 12

4 Annual moment release . . . . . . . . . . . . . . . . 14
S Seismicity of the Far Western Aleutians . . . . . . . 16
6 Seismicity of the Komandorsky Islands . . . . . . . . l9

7 69-Jan-20 Synthetics . . . . . . . . . . . . . . . . 21
8a 75-Aug-15 Synthetics . . . . . . . . . . . . . . . . 24
8b 75-Aug-15 Synthetic traces at several depths: KIP . 25
8c 75-Aug-15 Synthetic traces at several depths: SHI . 26
9 Synthetic seismograms for 77-Feb-19
a:Mechanism and stations
b—f:Synthetic traces (arranged by

increasing azimuthal distance) . . . . . . . . 27-32
10 Near Islands mechanisms (solid: this study) . . . . 35
11 Aleutian tectonic blocks (from Spence, 1977) . . . . 37
12 Komandorsky mechanisms (light shaded: Cormier,

(1975a) . . 39

13 Results of relocation of 1978 sequence . . . . . . . Al
14 Results of relocation of 1981 sequence . . . . . . . 42

15 Zones of like focal mechanisms . . . . . . . . . . . 46

16 Models for faulting in the Komandorsky Islands . . . 50

I NTRODUCTI ON

The Aleutian island arc represents a zone of
convergence between the Pacific plate to the south and the
North American plate to the north. This convergence is
characterized by a subduction zone at which the North
American plate is being underthrust by the Pacific plate
which is moving in a northwesterly direction. Associated
with this subduction zone is the characteristic magmatic
arc, consisting of basaltic and andesitic volcanism and
their associated sediments, which make up the Aleutian
Island chain. Also associated with this convergence is a
zone of seismicity which has been shown to be discontinuous
along the length of the arc (Figure l). The far western
Aleutians constitute a zone where the relative convergence
vector of the Pacific plate with respect to the North
American plate changes from obliquely convergent to
essentially parallel with the plate margin, as one moves

westward along the arc (Figure 1).

Upon examination of the geometry of the tectonic

setting, a zone of northwest striking right lateral

ou< cmaunma< onu mo huwowamwmmua ouswwm

 

 

'nlhahl'.hl.ln if!!! Ill-Int!
in: lilio'. gziiiiu Oil-III!!!
:8 on: guard-hung..." Sin 3 3.3.33 .43 ......Iuflakll. ail-ah?!”
W- . o ...4— , l ..m— o’— .3.— .... Ills—l: lug... -.-. .. -. o p- o.— . .02 l lib - - i-
a , / a . . l; y . x . Q: Aw

 

 

 

  

 

 

 

(2344. no 33.335

strike-slip faulting would be expected in the far western
Aleutians. Previous work has shown that although strike
slip faulting is indicated in this region, a significant
number of events indicating other types of faulting occur
(Cormier, 1975a). Cormier noted that the horizontal
component of the slip vectors of all of the solutions
maintained a uniformity toward the northwest, consistant

with the direction of the regional convergence vector.

The zone of transition from oblique convergence to
essentially strike-slip motion is the subject of this
thesis. The objective will be to try to gain an
understanding of the tectonics in the area by studying the
seismicity, the nature of faulting as indicated by
fault-plane solutions, and the waveforms produced by these
earthquakes. This study is intended to supplement the work
of Cormier(1975a) and that of Stauder(1968) in the light of

several years of additional data.

GEOLOGIC SETTING

Regional Setting

The area of interest lies at the farthest western
reaches of the Aleutian Island chain, between 164 degrees
east and 178 degrees east, and includes the Komandorsky
Islands (Beringa and Mednyy), the Near Islands (Attu,
Agattu, and others), and the western Rat Islands (including
Kiska). At the western boundary of the study area is the
Aleutian arc - Kurile-Kamchatka arc junction which is also
the site of visable termination of the Emperor Seamount
chain. The Komandorsky Basin section of the Bering Sea to
the north has been the subject of some attention due to its
anomalously high heat flow and apparently young age
(Cormier, 1975a; Rabinowitz and Cooper, 1977). Tholeiitic
basalts found at DSDP site 191 (Figure 2) have been dated at
32.3 ma. (mid-Oligocene, using the time scale of Harland et
al., 1982) and are postulated to be of regional extent in
the Komandorsky Basin (Stewart et al., 1973). The
Komandorsky Basin is bounded on the east by the Shirshov
Ridge, an enigmatic bathymetric rise which has been
postulated to be several things, including a remnant island
arc, an ancient spreading center and a result of hot spot
activity near a transform fault (Ben-Avraham and Cooper,
1981). North of the Near Islands is the Bowers Basin which

is delineated by the extent of the Bowers Ridge, an arcuate

mom magnum mnu mo mcammmuu ousmwm

 

 

 

 

 

81:3u44

 

 

 

 

 

 

 

bathymetric feature which has been interpreted as an
inactive island arc with a sediment-filled trench on its
north flank (Cooper et al., 1981). It has been suggested
that underthrusting activity at the Bowers Ridge occurred
mainly in Mesozoic and early Tertiary time, with minor

underthrusting continuing throughout Cenozoic time.

The formation of the Bering Sea-Aleutian Ridge system
is currently a subject receiving considerable attention in
the light of the aforementioned discoveries, along with some
new and surprising paleomagnetic data in the eastern reaches
of the Aleutians and Alaska. This evidence implies that
much of southern Alaska and parts of the Bering Sea may have
developed at more southerly latitudes and moved northward to
their current position; much of this movement to have taken
place continuously throughout Cenozoic time (Stone et al.,
1982). The contradiction of the paleomagnetic data with the
geologic data, which infers that the Peninsular terrane of
southern Alaska and the Beringian margin had accreted to
Alaska by the end of the Mesozoic, has yet to be resolved

(Marlow and Cooper, 1983).

Geology of the Western Aleutians

Considering the enigmatic nature of the Komandorsky
Basin, careful inspection of the geologic record of the

western Aleutians is necessary to determine if this section

of the arc is a simple continuation of the eastern section.
If not, it may be taken to imply that the Komandorsky Basin

is itself of allocthonous character.

The geology of the Komandorsky Islands has been quite
thoroughly studied by Soviet geologists in recent years
(Schmidt, 1978; Tsvetkov, 1980; Borsuk and Tsvetkov,
1980). The features noted by these geologists have been
interpreted in accordance with standard Soviet geological
principles, dismissing the western view of global plate
tectonics (Schmidt, 1978). The descriptive elements of the
Soviet literature have been extracted and are summarized in

Appendix A and commented upon below.

The rocks of the Komandorsky Islands have been
described in terms of six stratigraphic suites, the lower
four of which have been grouped as the Komandorsky Series.
The rock units found contain mainly volcanic and
volcanogenic sedimentary rocks. The lowermost unit consists
of basaltic pillow lavas, massive lavas, rhyolites, and
tuffs. This unit has been paleontologically dated as
Paleocene. The rhyolites of this unit have been suggested
to be younger material (middle Oligocene, by K/Ar dating),
intruding the Paleocene basalts (Tsvetkov, 1982). The
rest of the Komandorsky Series is characterized by increased
volumes of tuffs and volcanogenic sediments and the

disappearence of pillow lavas. Plant detritus, coal, and

sedimentary structures in the uppermost suite of the
Komandorsky Series indicate a proximity to land and a

shallow water environment.

Unconformably overlying the Komandorsky Series is the
Nikol'skoye Suite, which is marked by the appearence of
alkaline andesites and basalts, interpreted to be the
products of shield volcanism. The unit overlying this, the
Vodopad Suite, contains subaerial lava flows of basalts,
andesites, dacites, and ignimbrite-like formations. These
two youngest units found on the Komandorsky Islands have

been faunally dated as late Oligocene to Miocene.

The pillow basalt-andesite-rhyolite assemblage, like
that described for the Komandorsky Islands, has been
suggested to be produced by the maturation of a young
subduction zone (Ringwood, 1974). The content and form of
the rocks of the Komandorsky Series appear to reflect the
continuous building of a volcanic arc, from early deep-water
pillow lavas to volcaniclastic rocks deposited in a
shallower. water environment to extrusion of andesites and
dacites in subaerial conditions. The geological description
implies that the arc had been built above sea level around

Miocene time.

The rocks forming the Near Islands seem to reflect a
similar history of early oceanic volcanism (late Mesozoic to

early Tertiary), and later uplift, subaerial erosion and

extrusion of andesitic and dacitic lavas, probably in middle
to late Tertiary (Gates et al., 1971). Kiska Island,
located on the far eastern edge of the study area, is
characterized by three volcanic rock formations. The
lowermost formation contains pillow basalts and pyroclastics
and has been estimated to be middle Tertiary age; the
determination being made by analogy to paleontologically
dated rocks found on some of the Rat Islands and Amchitka
(Coats et al., 1961). Resting unconformably on this
formation is a unit composed of subaerial lava flows,
autoclastic breccias, pyroclastic rocks, and volcanogenic
sediments. This formation has been dated in a similar
fashion as being of late Tertiary to early Pleistocene in
age. The third geologic entity on Kiska Island is Kiska
Volcano, an andesite volcano which last erupted in

September, 1969 (Simkin et al., 1981)

The uniformity of geologic events across the Aleutian
Ridge has been noted by DeLong et a1. (1978), who correlated
events as far west as the Near Islands. Examination of the
data assembled by the Soviets appears to justify extention
of this uniform character west to include the Komandorsky
Islands. DeLong et al. (1978) describe a major event
occuring around 30 million years ago (mid-Oligocene) which
is expressed as an unconformity (reflecting uplift)
ubiquitous to the arc and a clustering of radiometric age

dates, which they ascribe to the signature of a metamorphic

event. They suggest that these findings reflect the
subduction of the Kula Ridge at the Aleutian Trench. (A
major unconformity has been described as existing between
rocks of the dominantly marine volcanic rocks of the
Komandorsky Series (analogous to the "Early Series" rocks
described by DeLong et al.(1978) as spanning the rest of the
arc) and the rocks of the Nikol'skoye and Vodopad suites
(which may be analogous to the lower portion of the "Late
Series” rocks). The paleontologically determined ages of
the units on either side of the unconformity are consistant
with the 30 million year date assigned by DeLong et al.
(1978). The main notable difference between the rocks of
the Komandorsky Islands and those of the central and eastern
sections of the arc is the lack of Pliocene and younger
volcanism in the west. The ability to correlate the rock
units of the Komandorsky Islands with the rest of the
Aleutian Ridge lends support to the notion that these far
western islands have been part of the Aleutian are system

since its beginnings in early Tertiary time.

10

SEISMICITY

Magnitude and Frequency

As can be clearly seen on the seismicity map of the
Aleutian are as a whole (Figure l) the density of epicenters
is not uniform along the arc. This segmentation of
seismicity has lead to the identification of several seismic
"gaps” in the Aleutian arc . As many as five of these gaps
have been proposed for the Aleutian arc (Figure 3, Sykes et
al., 1981), one of the most "mature" of these located in the
Komandorsky Islands region, from 170 degrees east to near
the coast of the Kamchatka Peninsula (Jacob, 1983). This
section of the far western Aleutians has not been the site
of a very large earthquake (magnitude > 7.4) since 1858,
although it is not clear if the record for events is

complete over the last 125 years (Sykes et al., 1981).

In this study, attention was focused on the events
which occurred in the far western Aleutians from 178 degrees
east to 164 degrees east, with the Komandorsky Islands
region receiving the most rigorous treatment. From 170
degrees east to 164 degrees east, only 211 events were
reported by the International Seismological Centre (ISC)
and/or the United States Geological Survey (USGS) for the
time period 1964-1982; only 46 of these being of magnitude
> 5.0 (mb) and only 3 of these of magnitude > 6.0 (mb).

11

 

 

      
           

 

   

 

 

 

 

 

  

            
  

   

 

  

 
    

     

 

 

 

 

Aammav ..Hm um moxmm Eoum .mcmfiusma< mzu mo mamu oHEmHmmnm muswflm
0°“- 00'. 000- 9. 30°F. 08— U005—
— — 4 4 fl‘ J...—PCO.—.W’.O.;\4. saw: _ saws.
, ...... .......\
. o 9 5.3635 3\_&wm xx ..m‘. twig .
22 Z. on . .
a \v
93283 ..
, eat I I . .....n
, f/ $64
332.2”... 6%,. if new a :23
3.934 6
are ..
s I x .,
O \ /\ 2.8
¢xm<J< \. 3
q l— H . .
.x o o o. .oo
...: . ... ... a I 5.2.5.! 5.5:... .8
. \ 'F.’ ;_._.\; EofiN o . \v \.
.... \‘ «w.~ ...u ‘ s . 0". Mb Eu. . . ......ku
: c . .\ ......w: a, co..:o_< . no.” fl
.“ a; . ....AQ . POGO; aw...\/
3:15 a. .. .. a. .
.v ....551/
. . {.6
a . r . ... I
. o . ... .
o» a 2.9 No .7, K . x \
s. «.5 1 .
m/ s. 2.0.. e. 33a cou:oE< 532
Cu: .32 at. O 2.3
3 53(44 1
h b C
$.02 .02 .00. 3.3.: .3: .3... ...:2

12

It should be noted that the smaller magnitude portion
of the data set is likely to be temporally biased due to
detection-level changes related to closing and opening of
several seismological arrays in the continental United

States, Alaska and the Aleutians (Habermann, 1983).

The moment release corresponding to these events was
estimated (Figure 4 ) using the assumption that in the lower
portion of the magnitude spectrum, there is roughly a 1:1
ratio of Ms:mb (Evernden, 1975). Considering the difficulty
of obtaining precise magnitudes for small earthquakes using
teleseismic data, as well as the error introduced by the
above assumption, these moment estimates may be viewed as
more qualitative than quantitative. The relationship

between moment (Mo) and magnitude (Ms) is given below:
Log Mo = % MS + 15.5 (Farmer, 1982)

The average annual moment release in the Komandorsky
Islands over the time period studied is about 1.6 x 10E24
dyn-cm with the cumulative total estimated at around 3 x
10825 dyn-cm. To emphasize the small absolute value of this
number, it should be noted that the moment release in the
Great Chilean earthquake of May 22, 1960 (Ms = 8.3) was
determined to be 2.7 x 10830 dyn-cm (Kanamori and Cipar,
1974), five orders of magnitude greater than the cumulative

total of nineteen years of seismic activity in the

Komandorsky Islands region.

13

 

  

 

 

 

 

 
    

 

 

. ...n...n.u.n.. . Wax....3..3.....u.n.u.u...u..4..u.u . ..q...a........ . .. ... - . ... . . “mummyfiuwwxmEvan“?”Na...n....w.v..x.wave...4.3.x;.U....”.H.Ummmu.o
I I I‘ I II I IIIII I I III I I I I I I I III III-I II I I ..Iu. I ..I-nunvn II IAII. IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IV ....................
film... .. P. . . .....m flux”.3"..n.n.u.".u”.u.n.u....n.. HEW...“ ..u..."....“u.su...“....m .. ......”.3.x.”wanna.”.HMAAAUVHAA.”+2.“.n...rn.n¢a.¢u.u.u.u.u.n.3.3+“ERMA.”
q 411‘ II. I I ... n 1 r .. . .I IIIIIII-Iw lllll ... .. .nv ...
. .. ..v. .... .....x... ..."...3" ..v .x v. . "...... . .u... .. . .. u rip.” ............. a
I I I I I I I I III I I rI IIInI .I II ”HI”- I-I can-XI
g. ANDRE“... “5...”, . . -. . . v. .. an". . .. . .unmufsfivamw .............................................
“.Mnunx.fl.a.u..u..n....uaunfiauvmwmfifm
_u.fix”xwvfimxhkaaax.”.72.”.”.HA»£.?<V¢. ..

IIIIIIIIIIIIIIII

wwwinumfi. .mnuumuvzmn..un..un 5&3“. “mummfiwuwnfifi ...)..." ....v. . ......w.“.u.ux..u. .
luxuwlnmmw“a.».nmvmnflaw.”.u¢w..@mw............. ...... a...» MWWWoflwufifiwukufiufiva a»

IIIIIII

 

mwnmmfi ..u. , Sufi". ......“..uwnwnnmmfiwfiwwm 5
.3».....»” wwwffiq fiuw”....wmuvkmuwwfi 7

I.
. I:
Y I O I.‘ u ‘- .-

... .
”.94?

%%.§m

  

I l
“.wxaaxlaxaiox.3.x
threw“?.HAAAAEAAAA.
.....fi ................ a .............

70

. . . . . ..x.<...u.w..nvn "n”
I

.. Inflammaifl . «fix.
I I III” II I I III.
ghefian. . 4.x

”r.

I.”

.. Anna}. .3... n H.
. .. .fiwfwn ...wwnfiufi.».§..m%.

 

 

 

 

KOMANDORSKY

Figure A-Annual moment release
14

The overall seismicity of the Near Islands is much
greater, as can be observed in Figure 5. To keep the scale
of the study in within managable bounds, only events which
were reported by 50 or more stations were included in the
base data set. This corresponds to a cutoff at magnitude
4.7 (mb), although some smaller events were included which
fit the necessary criterion. The resulting set of events
consists of 393 events reported by the ISC and/or the USGS
ocurring in the section of the Aleutian arc between 178
degrees east to 170 degrees east during the 19 year study
period. The average annual moment release (Figure 4) was
estimated to be between 4.69 x 10824 and 8.59 x 10824
dyn-cm, depending whether or not the year 1965 is included
in the averaging, since it represents the abnormally high
value of moment release corresponding to the aftershocks of
the Rat Island event of February 4 (Ms = 8.1). The lesser
of the two values therefore more accurately represents the
background seismicity of the area. The cumulative total
moment released, including the year 1965, was estimated at
1.63 x 10826 dyn-cm, a value 5.5 times that calculated for

the Komandorsky Islands region to the west.

Qualitatively comparing the shapes of the histograms of
Figure 4, it is clearly seen that with the exception of
three years out of a total of nineteen, the year to year
correspondence is very similar. The moment release of two

of the anomalous years, 1969 and 1977, is dominated by a

15

mamaunma< cumummz mom can mo huwowfimwomam muswfim

3.: .020.—

 

 

 

 

 

 

 

 

8.’_ 8.0}. '. — '.2- 00.2.. 00.2.. 8. 2— s. :_ '.2— ...-a qu-v—lllg.~hv.. 8....il.oo.a.-Lv_i .00....“
I . W

I. I“ I 0 V“

l - , w.

”.‘I .II MSW/H. ....

I Q 9 O [HQ/j n

. ,v .... Smog .
v\l\ln/ .- . P v z...
Si/ll/Jif I \K\ LU ‘Q .u

 

 

 

 

 

 

A4, <M/

‘ l I .
xx
\. R
rI-II. v‘.‘ t .
v.

\ .
r\a\ll\l.|r liaIaIA all}

.WEEQQ

 

:xj
\v/l
\

 

 

 

 

 

/
K
K.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a‘ t,

NmmTvmm. >._._0:2w_ww .

l6

single large event in the Komandorsky Islands in each of the
years, which may account for the discrepency in the
histograms. This correspondence is remarkable, considering
that the annual average moment release is 3 to 5 times

greater in the Near Islands region.

The spatial pattern of seismicity varies considerably
in the study area (Figure 5). East of 170 degrees, there
exist three easily distinguishable zones. A small number of
events occur to the southwest of the bathymetric trench in a
zone extending to about 172 degrees. Another, less easily
recognized group occurs on the northeast side of the
Aleutian Ridge, and is limited to the area from 172 degrees
to about 174.5 degrees. The majority of epicenters in the
Near Islands region occur in a broad, tapering zone which
encompasses the trench and ridge area to around 169 degrees.
Within this zone, an aseismic lineation occurs between 172
and 175 degrees which divides the zone into two groups each
trending northwest. As will be noted in a later section,
representative fault plane solutions from either side of
this lineation appear to be very similar. The average ISC
reported focal depth for the events within the central zone
to the northeast of the lineation for the study period is
30.0 km, compared to 27.6 km for events south of this
aseismic zone. In the zone to the southwest, 52% of the
earthquakes considered in this study occured during

February, 1965 as aftershocks of the Rat Island event of

17

February 4. Only 9% of the activity of the area northeast
of the aseismic lineation occured during February, 1965;
the bulk of the seismicity occuring later in 1965 and in
subsequent years, so the quiet zone appears to seperate two

groups of epicenters whose main difference is temporal.

In the Komandorsky Islands region, several zones of
earthquake activity can be recognized (Figure 6). On the
northeast side of the Aleutian Ridge, from 166 to 170
degrees, there occurs a sparce, northwest trending zone of
epicenters. On the southern side of the ridge is a group of
earthquakes which appear to be an extension of the main zone
of epicenters in the Near Islands and extends to at least
168.5 degrees and may include the group of events located
south of Ostrov Mednyy at 167.5 degrees. The densest
cluster of earthquakes in the region occurs to the southwest
of Ostrov Beringa, closely following the trend of the
bathymetric trench. Between 164 and 165 degrees occurs a
zone of epicenters that may be viewed as being somewhat
aligned in a north-south fashion, particularly the events of
larger magnitude. Another cluster of events occurs near 164
degrees at around 55.7 degrees latitude. These trends of
seismic activity will be discussed in relation to
representative fault plane solutions in the region in a

later section.

18

SEISMICITY I964- I982

j.

 

hi

9/
55

IS.

 

 

 

 

 

 

 
     
 

. [EDN‘

3. I978.
Master Event

  

I98I
Master Event

69/

 

c
/

PAC/x376.
or A

 

 

 

 

 

: WK”

 

 

 

 

(I
“554.00 "4.50 ”6.00 103.:

Figure 6-Seismicity of the Komandorsky Islands

19

Event Depths

The overwhelming majority of the events occuring in the
Komandorsky Islands region have ISC reported depths
shallower than 40 kilometers. Only 23 of the 211
earthquakes reported are 40 kilometers or deeper, only 6 of
which are reported at depths of 80 kilometers or greater.
Of the 393 events considered in the Near Islands region, 89

are reported to be at least 40 kilometers deep.

Since there are so few intermediate and deep events
reported in the Komandorsky Islands, those that are may be
considered suspicious. In order to more accurately
constrain event depths in the region, synthetic seismograms
were generated for three of the largest events in the
region, two of which were reported to have occurred at

greater than 40 km depth.

Since waveforms generated by earthquakes are sensitive
to the relatively complex structure in and around island
arcs, realistic layered models were assembled for each
event. The compressional wave velocities, material
densities and estimates of layer thicknesses were inferred
from results of gravity and refraction work done on the
Komandorsky Islands (Gaynanov et al., 1968) and other parts
of the Aleutian arc (Stone, 1968; Shor, 1964; Murdock,
1969). Since the amplitude {of the 5P phase is small

relative to the water surface reflection pwP for many types

20

monumsucsm om-cmn-mo a magmas

00a 0.
uonm. .
Trill I l [.11

 

 

mom
I

 

 

:00
.1

5.. «n: o.wumc mu_.—u:.—z>m Nun owizafiumw

 

 

 

Elli: ‘ - ~52
I

. I

21

of submarine earthquake mechanisms (Hong and Fujita, 1981)
rough estimates of the shear wave velocities were considered
sufficient for the crustal models. Considering the
uncertainty of crustal structure and that of comparing
waveforms, the depths determined by generation of synthetic
traces should be considered to be accurate to within 5 to 7
kilometers for shallow focus events (Hong and Fujita, 1981;
Wiens and Stein, 1983), an accuracy much better than that

achieved from consideration of teleseismic arrival times.

The event of January 20, 1969 has an ISC reported focal
depth of 17 km and a magnitude of 6.0 (mb). This magnitude
determination is questionable, since very little energy was
noted on long period records. This may simply reflect a
fast rupture, as indicated by the short triangular
source-time function used (.25,0.,.25 sec) or may be the
product of constructive interference between the P and pP
phases in the first few seconds of the wavetrain. The event
was modelled on the basis of six short period records
(Figure 7) using the focal mechanism determined by Cormier
(1975a) which resulted in a focal depth of 1 km. The
apparent water depth in the vicinity of the epicenter is
deeper for stations to the south, consistant with the local
bathymetry (Figure 12). Even though only six short period
stations were used, the azimuthal coverage is good and the

results are consistant.

22

The earthquake of August 15, 1975 was modelled using a
trapezoidal source-time function (1.,1.,1. sec) for nine
long period records (Figure 8a-c), resulting in a
recalculated depth around 13 km, as opposed to 41 km
determined by the ISC using P wave data. The water layer
was found to be 1.75 km, one to two km shallower than the
bathymetry directly over the reported epicenter. Moving the
epicenter a few kilometers southwest toward the Komandorsky

Islands would satisfy this observation.

The largest event to occur in the area in the past 20
years (February 19,1977; mb=6.1) was modelled on the basis
of 11 long period records using a simple velocity structure
from Gaynanov et a1. (1968) and a symmetric trapazoidal
source-time function (2.,2.,2. sec). The depth this
analysis yielded was 20 to 23 km (Figure 9a-f) with an
estimated water depth of 2.2 km. The determination of the
water depth was not as clear cut as was hoped, since it
appears that the event may have been a complex rupture,
complicating the later parts of the waveform. The original
focal depth of 44 km was reportedly determined by the ISC on
the basis of pP-P times. Since the water surface reflecton
pwP is often very prominent. in oceanic thrust mechanism
earthquakes (Hong and Fujita, 1981), it can be misidentified
as pP resulting in a significant overestimation of focal

depth, which is likely the case with this event.

23

mowumnucmm mHuwD<umn mm mudwwm

SET

,
W."
w<<éQ<§é (l- a I x
a: l. u m o

 

 

 

 

 

_
_
_
_
.
_
_
r- 2: ml”: m.mnmc mu_cuzzz»m Nan m2-osa-ms g

. it is.
3:;115 u §>2<<>71ooo 32%;}? :8

 

 

 

Hrw

 

24

KIPAPA

Obs.

5km

 

 

25

 

35

 

45

 

ééééaré

l I

SOsec '

Figure 8b:75-Aug-15 Synthetic traces at several depths: RIP

25

SHIRAZ

 

 

 

 

 

Obs. WW
fSknm ———]flvvh~
* I5 ‘NT
25 Wm
45 ‘W

 

l
I

r 60 sec

Figure 8c:75~Aug-15 Synthetics at several depths: SHI

26

 

 

77-FEB-19 LPZ SYNTHETICS “8:6.1 H:23 KM

 

 

Figure 9:Synthetic seismograms for 77-Feb-19
a:Mechanism and stations
b-f:Synthetic traces (arranged by

increasing azimuthal distance)

27

 

 

 

 

 

 

 

 

 

 

S

The earthquakes in the Komandorsky Islands with focal
depths reported at 80 kilometers or greater are so few and
small in magnitude (these deepest events all have magnitudes
mb < 4.4), that using them to draw any conclusions about
brittle rupture at these depths would be unfounded. The
epicenters of several of these earthquakes are located
seaward of the trench, which also tends to contradict the
conclusion that these events are a result of rupture of
subducted oceanic lithosphere at depth. Since these events _
are so small, no depth phases could be seen clearly on the
available short period records, so verification of the

depths was not possible.

Fault Plane Solutions

Fault plane solutions were determined for earthquakes
in the region utilizing P-wave first motion data from the
Bulletins of the ISC and USGS, and reading key stations
whenever possible. Short period records provided the only
useful information for all but the very largest events.
Confidence on the restraint of focal planes is quite varied.
Events for which mechanisms are presented are listed in
Table l and focal mechanisms and data are given in Appendix

C.

The mechanisms characterizing the eastern part of the

study area, from 170 to 178 degrees east, are shown on

33

Table l-Earthquakes for which mechanisms are presented

 

Mach 1 Event 1 Time 1 Lat 1 Long 1 Depth : Mag . Source
1 65-Oct-16 20 01 56.07 164.68 4 5.2 N
2 69-Apr-04 22 57 54.46 169.46 1 5.5 N
3 75-Jan-28 11 53 56.06 164.66 7 5.1 N
4 77-Apr-12 3 54 55.64 164.59 42 5.0 N
5 78-Mar-03 10 53 55.60 164.70 45 5.5 N
6 79-Sep-14 7 28 53.64 169.74 28 5.8 N
7 79-Nov-09 13 45 55.61 164.08 26 5.7 N
8 65-Jul-21 17 52 53.31 170.38 0 5.7 N
9 70-Apr-26 14 20 52.93 171.45 12 5.8 N

10 71-Jul-25 15 41 52.13 173.03 36 5.8 N
11 71-Nov-22 0 46 52.33 174.23 38 5.7 N
12 73-Mar-19 11 41 52.78 173.85 81 5.7 N
13 73-Mar-23 6 55 51.27 174.16 21 5.7 N
14 74-Aug-20 20 44 52.17 174.95 42 5.7 N
15 75-Feb-02 8 43 53.08 173.58 10 6.0 N
16 75-Feb-02 7 24 53.00 173.47 2 5.9 N
17 80-May-03 9 30 51.21 173.61 38 5.8 N
A 59-May-12 4 57 54.95 168.17 33 6.5 C
8 63-Aug-08 2 12 54.28 168.23 38 6.2 C
C 65-Feb-08 15 46 55.12 165.60 35 5.7 C
0 65-Sep-13 13 07 55.30 165.99 21 5.4 C
E 66-Jul-19 1 40 56.24 164.83 20 5.3 C
F 68-Jul-28 21 12 55.39 166.69 14 5.4 C
8 69-Jan-20 14 20 54.84 166.00 1' 6.0 C
H 69-Apr-17 12 48 55.27 167.00 9 4.9 2
I 73-Nov-17 5 45 54.17 169.27 16 4.7 L
d 73-Dec-29 8 20 54.66 168.63 32 5.4 L
K 74-Feb-08 14 21 54.32 167.61 23 5.4 L
L 75-Aug-15 7 28 54.92 167.87 13‘ 5.8 L
M 75-Nov-04 12 05 54.34 167.54 15 5.4 L
N 77-Oct-09 4 06 54.74 166.01 42 4.8 L
0 78-Nov-17 5 33 54.09 169.27 9 5.1 L
P 82-May-31 10 21 55.14 165.40 33 6.0 U
0 65-Feb-04 12 06 52.74 172.05 30 5.8 S
R 65-Feb-04 14 18 53.03 171.08 16 5.7 5
S 65—Feb-05 9 32 52.37 174.33 16 5.9 S
T 65-Feb-05 20 47 51.83 174.41 30 5.7 S
U 65-Feb-06 4 02 52.05 175.60 35 5.9 S
V 76-Feb-07 2 17 51.34 173.44 45 5.8 S
U 65-Mar-30 2 27 50.32 177.93 20 6.5 6
X 65-May-23 23 46 52.17 175.17 31 5.9 S
Y 66-Jun-02 3 27 51.01 175.98 48 5.7 S
2 77-Feb-19 22 34 53.54 169.96 25' 6.1 L

34

NEAR

ISLANDS

x--/¢

 

 

/
\Jx

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

\&

 

 

 

 

55

 

171 I72 I73 I74 I75 176 l77 I78

170

study)

Islands mechanisms (solid: this

Figure 10 Near

35

Figure 10. As was described by Stauder (1968), these
solutions fall into two major groups: a zone of normal
faulting mechanisms to the south of the ridge, in and around
the bathymetric trench, and a zone of low angle thrust
faulting located beneath the Aleutian ridge. Of the 19
mechanisms shown, all but three of the ten mechanisms
determined in this study are concordant with these
groupings. Note that representative mechanisms from either
side of the northwest trending aseismic lineation in the
central zone of seismicity are essentially identical. One
of the solutions that fits neither catagory (Solution 12)
was reported to occur at 81 kilometers depth and indicates a
large component of normal faulting. The two remaining
mechanisms constitute a strike slip doublet of events
occurring on February 2, 1975. These mechanisms were
interpreted as representing right lateral strike slip motion
along a northwesterly fault plane (Cormier, 1975b), as an
extension of strike slip faulting described by Cormier
(1975a) in the Komandorsky Islands to the west. However,
the choice of the northeast trending nodal plane as the
fault plane can be justified by noting that Agattu Canyon to
the south of Attu Island (Figure 11) and major faulting on
Attu Island itself (Gates et al., 1971) are aligned quite
well with the trend of this focal plane. Agattu Canyon has
been interpreted as delineating the boundary between two

tectonic blocks of the Aleutian Arc (Spence, 1977) at least

36

I ouswwh
am Eoumv mxooHn ofiaOuomu amausoa< HH
Annma .moam

o: .

 

£12. 8 o. o o.

5.000-4<>mmcz_xt<m0m
205mg 298: we}. 39a (I
3.55% 438- O
mutzwuau- .
rzozqzqdaxm

Ln

   

:1; 3 gum

r
a

37

as far north as the 100 meter bathymetric contour south of
Attu Island (Figure 11). Choice of the northeast nodal
plane as fault plane for these events suggests that this
boundary may extend at least to a point north of Attu. To
the east, an event with a similar mechanism which occurred
on July 4, 1966 has been suggested to be representative of
"ridge rebound" at the boundary of the Rat and Delarof
blocks in response to the Rat Island event of February 4,

1965 (Spence, 1977).

The fault plane solutions are more varied and in
general less well constrained west of 170 degrees, but
spatial consistancy is observed. Mechanisms determined for
the area south of Ostrov Mednyy from 167 to 170 degrees
east imply thrust faulting on a nearly horizontal plane
(Figure 12). The largest event to occur in the area (Feb.
19, 1977, mb=6.1) is of this type. On the north side of the
ridge between 167 and 170 degrees is a zone of earthquakes
which appear to have mechanisms opposite to those just to
the south. If the nearly vertical focal planes are taken to
be the fault planes, the mechanisms imply very high angle
reverse (or normal) faulting, with the southeastern side of
the fault downdropped with respect to the northwest. These
events are generally reported to be comparatively deep, but
the depth phase analysis of one such event (August 15, 1975,
mb=5.8) introduces significant doubt in to this

interpretation. The area surrounding Ostrov Beringa is

38

KOMANDORSKY ISLANDS

 

 

 

 

 

 

g
1352/6 V5

 

 

 

 

 

 

 

xv \

,\\

\

\Qfi
“ \
\ \

53-0 \ .
I640 (650 (660 l6"'0 I680 (690 (700

Figure lZ-Komandorsky Mechanisms (light shaded: Cormier,

1975a)

b)
\D

.characterized by both strike slip and reverse faulting
mechanisms. To the west of Ostrov Beringa and extending
north parallel to the coast of Kamchatka peninsula at least
as far north as 56.5 degrees lies a zone of mechanisms of
dominantly strike slip character. The northwest trending
focal plane has been interpreted to be the fault plane for
these strike slip mechanisms and those to the east (Cormier,
1975a). As will be summarized later, there may be evidence
to support choice of the northeast trending nodal plane as

the fault plane for these events.
Hypocentral Relocations

Master event hypocentral relocations were performed on
two earthquake sequences which occurred in the Komandorsky
Islands region in 1978 and 1981 (Figures 13 and 14). Each
sequence consists of a main shock with associated foreshocks
and aftershocks. Relocations were performed using residuals
from location of the main shocks as station corrections to
be applied to the arrival time data of the associated
shocks, resulting therefore in improved locations relative
to the main shocks. In this way, it is hoped that
constraints can be placed on the direction of the fault

planes on which these events occurred.

40

 

 

 

 

 

 

 

 

RESULTS OF RELOCATION OF I978 SEQUENCE

Figure 13

41

56-0
CDBdVeRwumhn
0 After Relocation
@Moster Event
555 "- 1:—
55 "o M
O
‘ 54-5
I64-O l65-O I66°O

|98| RELOCATIONS

BEFORE
T

AFTER

..55J

 

 

FocoI
doth
(km)

 

30-

oo-

50-

60- o

70-.

 

 

 

DEPTH

Figure 14-Results of relocation of 1981 sequence

42

The sequence of events occurring on February 9-10, 1981
consists of 4 foreshocks, the master event, and 11
relocatable aftershocks. The mechanism for the master event
has been determined using both first motion data and surface
wave analysis (D. L. LaClair, pers. comm.). The relocated
hypocenters appear to delineate a nearly vertical, northwest
striking fault plane (Figure 13) which does not appear to
closely resemble either nodal plane of the mechanism of the
master event. However, since the hypocenter of the master
event falls to the northeast of the projection of the main
cluster of events, the possibility exists that this
earthquake stimulated rupture of a seperate fault slightly
to the south. Only one of the relocated earthquakes was
considered to be unreliable, and therefore was not included

in the plot of the relocations.

The sequence of March through December of 1978
consisted of four foreshocks, the master event, and five
aftershocks that could be confidently relocated. As can be
seen in Figure 14, the events which originally defined a
somewhat northwesterly trend tend to cluster on a more
northerly plane as a result of relocation. This may be
taken to indicate this sequence ruptured a roughly northerly

trending plane.

43

DISCUSSION OF RESULTS

The study of the seismic activity of the far western
Aleutians during the past 19 years has lead to several
interesting results. Although the level of seismic activity
diminishes abruptly between the Near Islands and the
Komandorsky Islands, it has been shown that the temporal
moment release characteristics are quite consistant. The
geologic evidence implies no major structural boundary
between the Near and Komandorsky Islands regions. These
observations lead to questions regarding the maturity of the
Komandorsky Islands seismic gap. The consistancy noted in
Figure 4 indicates that either stress is accumulating in the
region at a rate linearly related to the annual moment
release, or more likely is being accomodated aseismically.
If the latter case is correct, then the Komandorsky Islands
region is no more likely to experience a great earthquake

than any other section of the arc.

The study of two event depths using body wave synthetic
seismograms yielded results that differed significantly from
the focal depths determined by consideration of p-wave
arrival time data by the ISC. Since the synthetics
indicated consistantly shallower depths, it may be concluded
that some sort of seismic velocity anomaly is indicated in

the source region. This anomaly may reflect the effects of

44

a remnant lithospheric slab at the source, or may simply be
a result of bias in the travel time tables used in the
location procedure resulting in an apparent velocity
anomaly. The focal depth of the third earthquake modelled
with synthetic seismogram analysis was likely overestimated

as a result of the misidentification of depth phases.

The zones of similar tectonic activity as inferred by
the study of fault plane solutions in the far western
Aleutians are presented in Figure 15. In the eastern
section of the area, a zone of normal faulting is indicated
which likely is a result of tensional stresses arising from
bending of the lithosphere seaward of the site of subduction
at the Aleutian trench. This zone, and likely this stress

regime, ends near 172 degrees.

To the north of Attu Island in the Near Islands, the
occurance of a~ doublet of strike slip earthquakes may be
taken as indication that the boundary between the Stalemate
and Near tectonic blocks of Spence (1977) extends north of
where it had been previously described. The other mechanism
in this area, which is dominantly of normal faulting
character, is reported to occur at 8L kilometers depth (it
may be shallower, as has been shown for other events in this

study), and may be a result of tensional stresses in the

subducted lithospheric slab beneath the Near Islands.

45

 

STRIKE SLIP

STRIKE SLIP w/ REVERSE
LOW ANGLE REVERSE
HIGH ANGLE REVERSE
NORMAL

 

OOG®®

Figure lS-Zones of like focal mechanisms

46

The zone of low angle reverse faulting which
encompasses the Aleutian trench and ridge extends from the
Near Islands to 169. degrees and possibly to around 167
degrees. The mechanisms of this zone indicate that
thrusting on nearly horizontal planes takes place south of
the Aleutian Ridge at least as far west as Ostrov Mednyy.
This may be taken to indicate that either active
underthrusting of the Pacific plate is occurring in this
section of the Aleutian trench or that the strike slip
relative motion between the two plates is occurring on a
horizontal fault plane. Since normal faulting mechanisms
occur only as far west as 172 degrees and active volcanism
is absent west of Kiska Island, vigorous underthrusting

seems not to be occuring in the Komandorsky Islands.

The area around Ostrov Mednyy is dominated by the
occurance of dip slip faulting on vertical planes, although
one strike-slip mechanism was found for this area. Cormier
(1975a) noted one of these high angle reverse faulting
mechanisms in his study and suggested that it may have been
mislocated toward the Aleutian Ridge and actually reflected
tectonism in the Komandorsky Basin to the north, not
directly related to ridge processes. This study has
identified several additional events with similar focal
mechanisms, one of which (August 15, 1975) was modelled by
synthetic seismogram analysis, yielding a water depth more

characteristic of the ridge area than the oceanic basin to

47

the north. These focal mechanisms are not aligned along the
strike a single vertical focal plane, and therefore are
interpreted as representing at least three seperate
occurances of dip slip faulting, each indicating a
downthrown southeastern block. Gravity modelling of the
Komandorsky Islands and the western Bering Sea by Gaynanov
et al. (1968) suggests a long narrow block (extending to 70
km depth) of material of reduced density under the western
Aleutian Ridge. If this interpretation is accurate, it may
be indicative of a remnant of subducted oceanic crust. The
contrast of densities between this slab and the surrounding
material could produce bouyant forces, causing the vertical

movements suggested by the fault plane solutions.

West of 167 degrees, mechanisms of dominantly
strike-slip character appear. In the zone surrounding
Ostrov Beringa, reverse faulting occurs with strike-slip
faulting but the mechanisms determined do not allow the
identification of specific fault trends. Cormier (1975a)
chose the northwest trending focal planes as the fault
planes for the strike-slip mechanisms in the Komandorsky
Islands, the plane and sense of motion being most consistant
with the relative motion vector in this area. This choice
appears to be appropriate for the events occuring in the
bathymetric trench south of Ostrov Beringa, considering the
results of relocation of the 1981 sequence and the general

local alignment of epicenters. For the mechanisms between

48

164 and 165 degrees, there are several indications that
choice of the northeast trending plane may be justified.
The seismicity west of Ostrov Beringa tends to be aligned
roughly north-south, especially the larger events, and
extends at least to 56.5 degrees north. The bathymetry of
the area (Figure 6) indicates a northward trending trough in
the area of some of the mechanisms northwest of Ostrov
Beringa. If the northeast trending focal planes are chosen
as representing the plane of strike slip faulting in the
area, then a northerly component of the stress field in the
area is causing some (localized?) northward movement of the
western Aleutian Ridge with respect to Kamchatka. The
results of the relocation of the 1978 earthquake sequence
indicate clustering of the epicenters along a more north
striking plane, implying that the thrusting noted southwest
of Ostrov Beringa is a continuation of the strike slip zone
to the north (Model A, Figure 16). If the northwest
striking planes are chosen to represent the fault planes in
the area (Model B, Figure 16), then a series of enechelon
faults is envisioned, which extends to at least 56 degrees
north, some 100 kilometers north of the apparent plate
boundary at the trench. The simplicity of the first scheme,
requiring only one fault plane instead of a wide diffuse
zone of faulting, lends support to the choice of Model A as

the most plausible.

49

mpamHmH zxmuopsmon mnu cw wawuasmm pom maovozuoa mudwwm

 

 

 

 

 

 

 

 

 

 

In
NI
~

7

 

.1
I
1
ii
’
I
,gk
A
i ,
1
I-
%§’
4I

[/11

7%;

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I a V. a
a s a

m .5905. < .6002

 

 

50

 

 

CONCLUSIONS

This study has shown that the far western Aleutian
Islands have likely been part of a continuous Aleutian arc
since its formation in early Tertiary and that no major
structural boundary seems to exist to account for
differences in the seismicity of the Near Islands and the
adjacent Komandorsky Islands region, therefore introducing

doubt as to the existance of a true seismic gap in the area.

It has also been shown that the tectonics of the far
western Aleutian arc can not be simply described by large
scale right lateral strike slip faulting. Northwest
trending strike slip faulting may only occur in a limited
zone to the south of Ostrov Beringa. Thrusting and vertical
movements have been found to be significant manifestations
of strain in the region and left Ilateral relative motion
between the western Aleutian Ridge and Kamchatka Peninsula
has been postulated. The most recent model of modern plate
motions (RMZ of Minster and Jordan, 1978, shown on Figure
12) indicates a relative motion vector between the North
America plate and the Pacific plate which contains a small
component of convergence between the two plates even as far
west as the Komandorsky Islands. This component may be
sufficient to account for the thrusting and lack of

extensive strike slip faulting east of 166 degrees, and may

51

give rise to some of the northward movement of the Aleutian
Ridge. Another possible explanation for north-south strike
slip faulting in the west involves the consideration of
possible southward movement, or clockwiSe rotation of the
proposed Okhotsk plate which may contain the Okhotsk Sea and
Kamchatka Peninsula (Chapman and Solomon, 1976). Recent
work of Stone et a1. (1982) in southern Alaska and the
eastern Aleutians has indicated that the Bering Sea and the
Aleutian Ridge have been moving northward throughout the
Cenozoic. The movement postulated by this study in the west
may possibly be an indication that the whole of the Aleutian

arc is moving northward today.

52

APPENDI CBS

APPENDI X A

Geology of the Komandorsky Islands

The Komandorsky Islands consist mainly of volcanic
rocks and volcanogenic sediments. The Soviet field
geologists have described the assemblege in terms of six
suites of rocks, the lower four of which constitute the
Komandorsky Series (Schmidt, 1978). The entire
stratigraphic sequence reflects a transition from basaltic
submarine volcanism to subaerial extrusion of andesitic and
dacitic magmas. Each component of the stratigraphic

sequence is commented upon below.

The Komandorsky Series

 

The Preobrazhenskoye Suite

This lowermost suite of the Komandorsky Series contains
the oldest rocks found to outcrop on the Komandorsky
Islands. The suite consists of contrasting effusives
(define), basaltic pillow lavas, and pelitic and psammitic
tuffs. These rocks grade from small pillows of porphyritic
basalt (containing augite, hypersthene, and basic
plagioclase) to larger basaltic pillows with coarse
autoclastic breccias and basic tuffs. Massive lavas with

crude columner jointing are found and the upper parts of the

53

suite grade in to lighter colored basalts and tuff sandstone
and tuff spilites. The suite is estimated to be at least

600 meters thick.

The Gavan' Suite

This suite of rocks contains lava breccias,
tuff-spilites, pelitic and psammitic tuffs and tuff
sandstones and conglomerates. The I Gavan' Suite is
characterized by the absence of pillow lava flows, although
some fragments of pillows are found. Pyroclastics and some
concretions appear in the upper parts of the suite which is

reported to be about 900 meters thick.

The Gavrilov Suite

This suite of rocks is described as containing pelitic
and psammitic tuffs, light gray concretions of tuff-calcite,
calcite horizons, tuff-conglomerates and tuff-sandstones,
and lava breccias. All but the very top of this suite is
composed of the pelitic and psammitic tuffs. The upper part
is marked by the appearence of bands of tuff-breccias,
tuff-sandstones and tuff-conglomerates which are suggested
to indicate a transition to the overlying Poludennaya suite.

The thickness of this suite is around 600 meters.

54

The Poludennaya Suite

This uppermost unit of the Komandorsky Series is
composed of rhythmically interbedded tuff-conglomerates,
tuff-gravellites, and tuff-sandstones. Large amounts of
plant detritus and some coal appear anong with signs of wave
action which are taken to indicate a shallow water
environment with proximity to land. This suite is thought

to be 500 to 700 meters thick.

The units of the Komandorsky Series are distinguished
on lithologic bases, since each appears to lie conformably
upon the preceding unit (except at the base of the lowermost
Preobrazhenskoye Suite, where contact is unknown). The age
of this lowermost suite has been determined to be Paleocene
based on the occurence of tricamerate globigerinas, reported

to be similar to the species Globigerina nana Chalilov, and

 

Globigerina cf. 9, pseudobulloides Plummer (Schmidt, 1978).

 

 

The fauna in the middle and upper parts of the series are
poorly preserved so the age has noit been well-determined,
but by analogy to other suites in Japan and California, it
is estimated that these upper units may be lower Paleogene,

probably Eocene (Schmidt, 1978).

55

The Nikol'skoye Suite

The Nikol'skoye Suite lies unconformably over the rocks
of the Komandorsky Series and is composed of alkaline
andesites and basalts, stratified tuffites, diatomaceous
tuffites, and tuff-conglomerates and sandstones. This
suite is characterized by the appearance of shield
volcanics. The extrusive rocks are found to grade from
alkali basalts to andesites. Coalified detritus and
fragments of tree trunks are common. The occurance of the
minerals titanomagnetite, pyroxene, pigeonite, alkali
hornblende (barkevikite), and dolomite is an assemblege of
minerals unique to this suite. The titanomagnetite content
in some of the lavas may reach 10-15 percent (Schmidt,
1978). The Nikol-skoye Suite has been dated at late

Oligocene to early Miocene.

The Vodopad Suite

The Vodopad Suite consists of tuff-conglomerates and
breccias, dacites, tholeitic basalts, and subaerial
andesites. This suite is characterized by discrete, deeply
eroded relicts of subaerial stratovolcanos, which appear as
terrestrial lava flows of basalts, andesites, and dacites,
and mud flows. Also noted are ignimbrite-like formations,

and the floral remains of forests.

56

The Vodopad Suite lies unconformbly on the highly
disected paleorelief of the Nikol'skoye Suite and
Komandorsky Series below, and reaches a thickness of around
500 meters. Paleontological data are scarce within this
suite, but it has been suggested that an age of Pliocene

may be assigned (Schmidt, 1978).

57

APPENDIX B

Data Analysis Techniques

This appendix outlines the techniques used in the
analyses of this study. Earthquake relocation, the
determination of focal mechanisms and body wave analysis by
the generation of synthetic seismograms will be discussed
with regard to the assumptions and approximations involved
in each technique and the validity of the conclusions drawn

from their use.

Earthquake Relocation

Single event and master event relocations were
performed using an iterative, least squares algorithm
implimented by Hiroo Kanamori in a computer program known
locally as ”HYPZDT”. Due to the execution time, precision,
and core memory requirements of the FORTRAN coded program,
the MSU Cyber 750 mainframe computer was used for all

calculations.

The program accepts as input data the station code,
P-wave arrival times for all stations (with epicentral
distance < 100 degrees) reporting the earthquake to be
located, and hypocentral estimates from which it calculates

a theoretical P-wave arrival time to compare with the

58

observed time for each station. The difference between the
theoretical and observed times (station residuals) are then
minimized by a least squares process by allowing the
hypocenter and origin time to vary. The latitude and
longitude are the major factors in determining this minimum
with the event focal depth and origin time being of
subordinate influence. In this regard, the formal precision
with which the origin time and focal depth are often
presented should not be confused with absolute accuracy.
The travel time tables used to generate the theoretical
arrival times (Jeffreys and Bullen, 1940) are empirically
determined from enormous amounts of data collected at points
all over the globe, they result in theoretical calculations
reflecting a symmetrical earth and "average" crustal models.
Since island arcs are by their very nature abnormal in
structure, biases can be introduced in the location process.
This bias may in part account for the apparent greater focal
depths determined by this type of algorithm for the events

upon which body wave analysis was performed.

Since the hypocentral estimates given by the ISC are
generated by the above procedure, "relocation” of a single
event must involve differences in assumptions to yield any
further information. There are a couple of assumptions
which can prove useful in this regard. One is to assume
knowledge of one or more of the source parameters and

disallow the program to change it. The event parameter most

59

often chosen for this is the focal depth, often obtainable
from analysis of body waves. Since focal depth and origin
time are not highly influential parameters in the
minimization, fixing one likely acts to vary the other much
more than it affects the epicentral estimates. Another
method of obtaining different results involves differences
in the consideration of high residual stations. The ISC
locations include high residual stations in their data set,
although their effects are diminished by the use of
weighting factors. Since modern recording, timing, and
measuring equipment allow the determination of arrival times
on short period records to within .3 seconds, residuals of
greater than a few seconds may be considered suspicious.
Inhomogenieties in the earth and slab effects can cause such
high residuals, but can often be identified on the basis of
consistancy of stations in a given direction or area. The
assumption that is helpful in the relocation of events is
that isolated high residuals are a result of
misidentification of incoming phases, and are therefore
esentially removed from the data set, hopefully resulting in

more accurate hypocentral locations.

The program also allows for relative relocation of a
master event and foreshock and aftershock (or master event
and neighboring events) by assuming that the residuals
calculated for the master event (usually the largest

magnitude event in the area) are a result of inhomogeneites

60

in the earth between the earthquake and its recievers and
therefore may be applied as station corrections for other
events located near the master event. The validity of this
assumption may only be relied upon within a few hundred
kilometers of the master event (Fujita, Engdahl, and Sleep,
1981). The result is that when these station corrections
are applied to the subordinate events, precise locations,
relative to the master event, will be obtained. This is
commonly used to determine the direction of the fault plane
involved in an earthquake using the pattern of its
foreshocks and aftershocks thereby resolving the inherent
ambiguity of a double couple focal mechanism. Since the
largest shock in an area is usually considered to be the
most accurately located by the standard technique, the
relocations may not only be precise in a relative sense, but

also more accurate in an absolute sense.

Fault Plane Solutions

Fault plane solutions were determined for earthquakes
in the study area by using two standard techniques. The
representation most heavily relied upon is that of
presenting compressional wave polarizations on an azimuth
versus focal angle of incidence plot on an equal area

sterographic projection. The angle of incidence at the

61

focus (or take off angle) has been calculated for different
values of epicentral distance (delta) and for different
focal depths by Pho and Behe (1972). Each point plotted
represents the point of emergence of a ray from the a sphere
around the focus of the earthquake in lower hemisphere
Schmidt net projection. It is then attempted to delineate
zones of compressional- and dilatational first motions,
(p-wave polarizations) by the projections of two orthogonal
planes, representating the particle motion at the source due
to a double couple faulting mechanism. The symmetry
involved in the determination makes the distinction between
the fault plane and the auxillary plane impossible to make.
One must call upon other evidence to make this choice.
Three primary sources of error are involved with plotting
first motion data. The take off angle is highly dependant
on the focal depth of the earthquake, which is often not
well constrained. Errors of focal depth are often
manifested in the inablilty to make the nodal planes meet
the condition of orthogonality. Another source of error is
the lack of complete azimuthal coverage and lack of close
stations reporting first motions. In the case of western
Aleutian earthquakes, the southeast quadrant of the focal
sphere (representing the open Pacific Ocean) is nearly void
of data, and reports of stations close to the source region
are very scarce. The third major source of error results

from misidentification or misreporting of the first motion

62

by station operators and the ISC. To minimize this problem,
many stations were read first hand to verify and augment the

data set.

The other common method of determining fault plane
solutions involves analysis of amplitude radiation patterns
of surface waves (Rayleigh waves in this study) as a
function of azimuth. The theoretical basis of modelling
surface wave radiation patterns is quite mathmatically
rigorous and will not be presented here. The procedure
basically involves collecting Rayleigh wave amplitude data
(in digital form) and plotting it as a function of azimuth,
taking in to consideration propagation effects such as
attenuation and dispersion. Fourier transformations are
used to isolate amplitude data corresponding to desired
ranges of frequencies. The resulting data is plotted and
compared with theoretical radiation patterns. The observed
patterns are usually azimuthally incomplete and contain
variations due to non-uniform attenuation characteristics
and are therefore quite jagged, but the lobe maxima and
nodal characteristics are usually discernable. The
radiation patterns are quite sensitive to the source fault
geometry, so finding the pattern which best matches the
observed data is often not a difficult task. Used in
conjunction with mechanisms produced from P-wave first
motion data, a well-constrained solution is often

achievable.

63

One surface wave mechanism was attempted in this study,
but the lack of available data and time constraints
precluded the confident determination of a fault plane
solution. The event occured on March 3, 1978 and was
considered to be the master event of the relocated sequence
of earthquakes occuring around 165 degrees at 55 degrees
north latitude. The first motion data indicated a thrust
event with one plane fairly well constrained, but the
surface wave data in hand were not sufficient to
substantiate or improve this determination. The waveform
produced by the event appears to indicate a complex rupture
history, so it is not certian that a solution even
obtainable. The results of the attempt and the stations

used are presented in Appendix D.

Body Wave Analysis

Body waves were analysed in this study by the
(generation of synthetic seismograms in order to place
constraints on some of the focal parameters of three of the
largest earthquakes in the Komandorsky Islands. Refinement
of the focal depth, near source structure, and focal
mechanism can be achieved through successful matching of
observed seismograms with traces generated artificially.

The procedure involves theoretically generating seismic rays

64

taking into account effects of the focal mechanism, near
source velocity structure, and source-reciever seperation.
The resultant seismogram consists of theoretical pulses
generated at the source and at the interfaces of the
velocity structure, scaled for relative amplitudes and time
delays. The method is especially useful for shallow focus
earthquakes, where the effects of the near source structure

are seen in the earliest parts of the seismogram.

The synthetic seismogram for teleseismic events is
expressed by Kroeger and Geller (in press) as a convolution

of the form:
u(t)=S(t)*NSS(t)*E(t)*RS(t)*I(t)

where u(t) is the seismogram, S(t) is the source time
function of the earthquake, NSS(t) represents the effects of
the focal mechanism, propagation and near source structure,
E(t) is the term representing the effects of propagation
through the source to the reciever, RS(t) represents the
near reciever structure, and I(t) is the response of the
recording instrument. The term E(t) includes the effects of
geometrical spreading, analastic attenuation and travel time
from the source region to the reciever region. The near
reciever structure term is neglected since the near source
structure has the dominant effect on the waveform produced

by submarine earthquakes recorded by land-based recievers.

65

The main function of the synthetic seismogram routine
is to calculate the near source structure term NSS(t). This
term, as considered in the Kroeger-Geller algorithm, is a
time series of scaled impulses, one for each ray which
enters a homogeneous halfspace as a result of interaction
with a velocity structure of horizontal layers in the
vicinity of the source. This term is convolved with the
propagation term to produce a discrete spike representation
of the signal, and this in turn convolved with the source
time function and appropriate instrument response to produce
the synthetic seismogram. The character of each pulse
(duration, symmetry) is determined by the source time
function, which is altered to match the observed pulse
shape. The overall waveform can be smoothed (by the removal
of higher frequencies) by altering the attenuation
characteristics corresponding to the raypath to a particular

station.

Synthetic seismogram analysis is particularly useful
for obtaining constraints on the focal depth and the depth
of the water over the epicenter af a submarine earthquake
event. Long period waveforms are quite sensitive to the
effects of the pP and pwP phases produced by earthquakes of
body magnitude of roughly 5.6 and above. Of the three
events analysed in this study, the event of Mb=5.8 was best
suited for long period body wave analysis. The high

amplitude of the waveforms produced by the 6.1 event made

66

matching the long period records more difficult. The ISC
reported 6.0 (mb) event produced very little long period
body wave energy and thus was modelled on the basis of
relatively few short period records. The results of this
event were quite consistant however, and are presented with

a high level of confidence.

The most significant error introduced in to this type
of analysis is due to representation of the velocity
structure in the source region. The structure of island
arcs is generally complicated and variable, making velocity
constraints particularly difficult to determine.
Consideration of this imprecision, coupled‘ with the
uncertainty involved in matching waveforms implies that
depth determinations will be accurate to only a few
kilometers (5-7 km as suggested by Wiens and Stein, 1983 and

Hong and Fujita, 1981).

67

APPENDIX C

Focal Mechanisms

The following figures reprisent fault plane solutions
determined in this study by means of plotting compressional
wave first motions in a lower hemisphere stereographic
projection of the focal sphere. Compressional first motions
are represented by crosses and dilitational first motions
by diamonds. The larger symbols indicate data read first
hand, the smaller symbols are picks from the ISC and USGS

bulletins. Planes are dashed where uncertain.

68

 

KOI’I BS-UCT-IS 56.07N 164-68E H24 I1:5.2

 

 

 

 

69

 

 

 

1 “:5.5

KG” 59-RPR-04 54.46N 169.46E H

 

 

 

7O

 

 

KOM 75—JRN-28 56.08N 164.86E H:7 ”25.1

 

 

71

 

 

 

KOH 77-RPR-12 55.64N 184.59E H242 M:S.O

 

 

72

 

 

 

KOI‘I 78-HRR-03 55.06N 164.70E H:45 H:S.5

 

 

73

 

 

KOI‘I 79-SEP-l4 53.64N 169.74E Hz28 ”25.8

 

 

 

74

 

 

 

KOH 79-NOV-09 SS-BIN 164.08E H:26 H:5.7

 

 

75

 

 

 

NERR 65-JUL-21 53.31N 170.38E H:0 H:S.7

 

 

76

 

 

 

NERR 70-RPR-26 52.93N 171.45E H:12 “25.8

 

 

77

 

 

 

NERR 71-JUL-25 52.13N 173.03E H236 (1:5.8

 

 

IO

78

 

 

 

NERR 71-NOV-22 52.33N 174.23E H238 ”25.7

 

 

79

 

 

 

NERR 73-HRR-19 52.78N 173.85E H:81 (‘1:5.7

 

 

 

l2

80

 

 

 

NEFIR 73-MRR-23 51.27N 174.18E H:21 H:S.7

 

 

I3

81

 

 

 

NERR 74-RUG-20 52.17N 174.95E H:42 Fla-5.7

 

 

I4

82

 

 

 

NERR 75-FE8-02 53.08N 173.58E H:10 ”26.0

 

 

I5

83

 

 

 

NEFIR 75-FE8-02 53.00N 173.47E H22 ”25.9

 

 

I6

84

 

 

 

:38 ":5-8

”RY-03 51.21N 173.815 H

NERR 80

 

 

 

I7

85

APPENDIX D

Surface Wave Mechanism

An attempt was made to determine a fault plane solution
using Rayleigh wave radiation patterns for the earthquake
of March 3, 1978. Time considerations and lack of sufficient
data resulted in a poorly constrained solution. The results
of analysis using 14 long period vertical station records

appear on the following pages.

86

 

 

 

 

 

STRIKE: O DIP: O SLIP: 0
T: 23 SEC
RRYLEIOH HRVEU

87

 

 

 

 

 

 

 

 

 

 

 

 

 

STRIKE: 0 DIP: 0 SLIP: O
T: 33 SEC
RflYLEIOH NEVIS

88

 

 

 

 

 

 

 

 

 

 

 

STRIKE: D DIP: D SLIP: D
T: 4‘ SEC
RRYLEIDH HHVES

89

LIST OF REFERENCES

LIST OF REFERENCES

Ben-Avraham, 2., and A. K. Cooper, Early evolution of the
Bering Sea by collision of oceanic rises and north
Pacific subduction zones, Geol. Soc. Amer. Bull., 92,
485-495, 1981.

Borsuk, A. M., and A. A. Tsvetkov, Magmatic associations of
‘ the western part of the Aleutian Island Arc, Internat.
Geol. Rev., Engl. Transl., 24, 317-329, 1982.

Chapman, M. E., and S. C. Solomon, North American-Eurasian
plate boundary in northeast Asia, J. Geophys. Res., 81,
921-930, 1976.

Coats, R. R., W. H. Nelson, R. Q. Lewis, and H. A. Powers,
Geologic reconnaissance of Kiska Island, Aleutian
Islands, Alaska, U. S. Geol. Surv. Bull., 1028-R,
563-581, 1961.

Cooper, A. K., M. S. Marlow, and z. Ben-Avraham,
Multichannel seismic evidence bearing on the origin of
Bowers Ridge, Bering Sea, Geol. Soc. Amer. Bull., 92,
474-484, 1981.

Cormier, V. F., Tectonics near the junction of the Aleutian
and Kurile-Kamchatka arcs and a mechanism for Middle
Tertiary magmatism in the Kamchatka Basin, Geol. Soc.
Amer. Bull., 86, 443-453, 1975a.

Cormier, V. F., Composite focal mechanism solution of the
Near Islands (Aleutian Islands) earthquakes of February
2, 1975 (abstract), Eos Trans. AGU, 56, 393, 1975b.

DeLong, S. E., P. J. Fox, and F. W. McDowell, Subduction of
the Kula Ridge at the Aleutian Trench, Geol. Soc. Amer.
Bull., 89, 83-95, 1978.

Evernden, J., Further studies on seismic discrimination,
Bull. Seimol. Soc. Amer., 65, 359-391, 1975.

Farmer, R. A., Seismicity, tectonics and focal mechanisms in
the Scotia Sea area, M. S. Thesis, Mich. State Univ.,
1982.

Gates, 0., H. A. Powers, and R. E. Wilcox, Geology of the

Near Islands, Alaska, U. S. Geol. Surv. Bull. 1028-U,
709-821, 1971.

90

Gaynanov, A. G., I. P. Kosminskaya, and P. A. Stroyev,
Geophysical studies of the deep structure of the Bering
Sea, Acad. Sci. USSR, Izv., Earth Phy. Ser., nr. 8,
461-465, 1968.

Habermann, R. E., Teleseismic detection in the Aleutian
Island Arc, J. Geophys. Res., 88, 5056-5064, 1983.

Harland, W. B., A. V. Cox, P. G. Llewellyn, C. A. G.
Pickton, A. G. Smith, and R. Walters, A Geolggic Time
Scale, 131 pp., Cambridge Univ. Press, Great Britian,

 

Hong, T.-L., and K. Fujita, Modelling of depth phases and
source processes of some Central Aleutian earthquakes,
Earth Planet. Sci. Lett., 53, 333-342, 1981.

Jacob, K. H., Aleutian seismic gaps quantified: High
probability for great Shumagin earthquake in next 10
years (abstract), Eos Trans. AGU, 64, 258, 1983.

Jeffries, H., and K. E. Bullen, Seismolo ical Tables, 50
pp., Neill & Co. Ltd., Edinburgh, 19 .

 

Kanamori, H., and J. Cipar, Focal process of the Great
Chilean Earthquake of May 22, 1960, Phys. Earth Planet.
Inter., 9, 128-136, 1974.

Kroeger, G. C., and R. J. Geller, An efficient method for
synthesizing teleseismic body waves in a vertically
stratified medium, (in press).

Marlow, M. S., and A. K. Cooper, Wandering terranes in
southern Alaska: The Aleutia microplate and
implications for the Bering Sea, J. Geophys. Res., 88,
3439-3446, 1983.

Minster, J. B., and T. H. Jordan, Present-day plate motions,
J. Geophys. Res., 83, 5331-5354, 1978.

Murdock, J. N., Crust-mantle system in the central Aleutian
region: ~ A hypothesis, Bull. Seismol. Soc. Amer., 59,
1543-1558, 1969.

Pho, H. T., and L. Behe, Extended distances and angles of
incidence of P waves, Bull. Seismol. Soc. Amer., 62,
885-902, 1972.

Rabinowitz, P. D., and A. K. Cooper, Structure and sediment
distribution in the western Bering Sea Basin, Marine
Geology, 24, 309-320, 1977.

91

Ringwood, A. E., The petrological evolution of island arc
systems, J. Geol. Soc. Lond., 130, 183-204, 1974.

Schmidt, 0. A., Tektonika Komandorskikh ostrovov i struktura
Aleutskoi griady, Akad. Nauk SSSR Geol. Institut,
Trudy, vyp. 320., 100 pp., 1978.

Shor, G. G. Jr., Structure of the Berings Sea and the
Aleutian Ridge, Marine Geology, 1, 213-219, 1964.

Simkin, T., L. Siebert, L. McClelland, D. Bridge, C.
Newhall, and J. H. Latter, Volcanoes of the World, 232
pp., Smithsonian Institution, Hutchinson Ross Publ.
Co., Stroudsburg, PA, 1981.

Spence, W., The Aleutian Arc: Tectonic blocks, episodic
subduction, strain diffusion, and magma generation, J.
Geophys. Res., 82, 213-230, 1977.

Stauder, W., Mechanism of the Rat Island earthquake sequence
of February 4, 1965, with relation to island arcs and
sea-floor spreading, J. Geophys. Res., 73, 3847-3858,
1968.

Stewart, R. J., J. H. Natland, and W. R. Glassley, Petrology
of volcanic rocks recovered on DSDP leg 19 from the
north Pacific Ocean and the Bering Sea, in Creager, J.
S., D. W. Scholl, et al., Initial Reports of the Deep
Sea Drilling Project, Volume 19, pp. 615-627,
Washington(U. S. Government Printing Office), 1973.

 

Stone, D. B., Geophysics in the Bering Sea and surrounding
areas: A review, Tectonophysics, 6, 433-460, 1968.

Stone, D. B., B. C. Panuska, and D. R. Packer,
Paleolatitudes versus time for southern Alaska, J.
Geophys. Res., 87, 3697-3707, 1982.

Sykes, L. R., J. B. Kisslinger, L. S. House, J. N. Davies,
and K. H. Jacob, rupture zones and repeat times of
great earthquakes along the Alaska-Aleutian Arc, in
Earthquake Prediction, edited by D. W. Simpson and P.
G. RiEhards, pp. 73-80, American Geophysical Union,
Washington, D. C., 1981.

Tsvetkov, A. A., First data on rare-earth elements in
igneous rocks of the Komandor Islands, Acad. Sci. USSR,
Dokl., earth Sci. Ser., Engl. Transl., 251, 184-187,
1980.

Tsvetkov, A. A., Pseudodikes in rhyolite of the Komandorskiy
Islands, Int. Geol. Rev., Engl. Transl., 24, 313-316,
1982.

92

Wiens, D. A., and S. Stein, Age dependence of oceanic
intraplate seismicity and implications for lithos-
pheric evolution, J. Geophys. Res., 88, 6455-6468,
1983.

93

1293 03145 8239

II
H
l
"I
V. I”
II.
n
"I
H
l
I
III
"I
"II
n
H

3