PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProj/Acc&Pres/ClRC/DateDue.Indd

TECTONIC IMPLICATIONS OF THE MAY 18, 1971 ARTYK EARTHQUAKE.
NORTHEAST RUSSIA

By

Melissa S. McLean

A THESIS

Submitted to
_ Michigan State University
In partial fulfillment of the requirements
For the degree of

MASTER OF SCIENCE
Geological Sciences

2009

ABSTRACT

TECTONIC IMPLICATIONS OF THE MAY 18, 1971 ARTYK EARTHQUAKE,
NORTHEAST RUSSIA

By

Melissa S. McLean

The Artyk earthquake of May 18, 1971, is the largest earthquake in
continental northeast Russia in the instrumental era. An extensive aftershock
sequence occurred during the three months following the mainshock. 286 of
these events were relocated using an expanded data set incorporating data
from both temporary and regional stations, and using a local travel-time curve.

The relocated aftershocks align very closely with the Kobdi fault as
mapped by Shilo (1961) and form two clusters. The northern cluster has the
same strike as the northwest-striking nodal plane of the focal mechanism,
indicating left-lateral strike-slip movement. The southern cluster shows a bend
to the east. The Kobdi fault appears to offset a diorite by about 8 km, yielding
an inferred rate of motion of 0.2 cm/yr on the fault.

The left-lateral strike-slip nature of the mainshock. the northwest-
striking aftershock distribution, and the lack of faults with a southwest or east
to northeast strike, and recent activity on the Ulakhan fault to the northeast
imply the earthquake cannot lie on the North America — Okhotsk or Okhotsk -
Eurasia plate boundaries or at a triple junction between them. Instead, the
Artyk earthquake was generated along a fault that represents internal

deformation in the Okhotsk plate.

Copyright by
Melissa S. McLean
2009

I dedicate this work to my son, Anthony.

ACKNOWLEDGEMENTS

Kaz F ujita, this wouldn’t have been possible without you. I sincerely
appreciate all of the hard work you put into teaching and guiding me through
graduate school, and the many steps you took to bring me back so that I could

properly finish. I can’t thank you enough.

Kevin Mackey, your optimism and unique way of looking at things were
instrumental in my understanding of northeast Russia. I truly enjoyed our trip
to Russia and I look forward to stories of your future adventures around the

world.

Bill Cambray and Dave Hyndman, thanks for being a part of my
committee. You have both helped me understand things in a different way, for

which I am thankful!

Jackie Bennett, your knowledge of the system really helped me get

through the bureaucracy so I could finish. I appreciate it.

Martha and Elizabeth F ujita, you guys were my family away from home.
I appreciate the tremendous hospitality and the love you showed me, as if I

was a member of your family.

Rob McCaleb, you were a great roommate and are a wonderful friend.
I like to think that our band practices were instrumental in understanding our

school work.

To my colleagues in Russia: Boris Koz’min for your experience and
knowledge. Larissa Gunbina and folks at the Magadan seismic station for

your hospitality.

To my friends, there are too many of you to thank here, but know that

you all made a big difference in getting me through some long years.

Mom and Dad, thanks for the financial backing and the support so that I
could go to graduate school in the first place. Shannon and Jenny, thanks for

reminding me that school is important, but family is even more important.

And last, I want to thank my husband, Larry, and my son, Anthony, for

putting up with me finishing school. I love yOu both.

This research was supported, in part, by the Department of Geological
Sciences of Michigan State University, National Science Foundation grant
OPP 98-06130, and the Department of Energy contract DE-FCO3-OZSF

22490, and is gratefully acknowledged.

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................... x
LIST OF FIGURES ................................................................................ xi
CHAPTER 1 Introduction ........................................................................ 1
1.1 Tectonic Setting ............................................................................ 4
1.2 Statement of the Problem ............................................................ 11

CHAPTER 2 The Artyk Earthquake and its Regional Tectonic Setting......... 16

2.1 The Artyk Earthquake of May 18, 1971 ............................................. 16
2.1.1 Hypocenter and Origin Time ......................................................... 16

2.1.2 Magnitude and Energy ................................................................... 20

2.1.3 Focal Mechanism ........................................................................... 23

2.1.4 Macroseismic Data ......................................................................... 26

2.1.5 Landslides and Surface Ruptures .................................................. 30

2.1.6 Aftershock Sequence and Temporary Deployment ........................ 33

2.2 Active Faults .............................................................................. 37

2.2.1 Ulakhan Fault System ......................................................... 38

2.2.2 Chai-Yureya Fault .............................................................. 41

2.2.3 Kobdi Fault ................................................................................... 43

2.2.4 Nera Fault ........................................................................ 52

2.2.5 Arkagala Fault ................................................................................ 53

2.2.6 Other Faults and Lineaments in the Epicentral Zone............... ...55

2.3 Cenozoic Basins along the Kobdi, Nera and Arkagala Faults ................ 55

CHAPTER 3 Seismological Database ...................................................... 57
3.1 Regional Networks ...................................................................... 57

3.1.1 Yakut Regional Network ...................................................... 57

3.1.2 Magadan Regional Network ................................................. 59

3.2 Artyk Sequence Data60
CHAPTER 4 Data Analysis .................................................................... 64

4.1 Crustal Velocities and Travel-Time Curves ....................................... 64
4.2 Location Procedure ..................................................................... 66

4.2.1 Removal of High-Residual Stations ........................................ 66

4.2.2 Impact of Individual Stations ................................................. 67

4.2.2.1 Regional Stations (General) .............................................. 68

4.2.2.2 Specific Stations ................................................................ 74

4.3 Location Quality .......................................................................... 77

4.3.1 Group1 Events .................................................................. 79

4.3.2 Group 2 Events .................................................................. 79

4.3.3 Group 3 Events .................................................................. 79

vii

4.4 Results ..................................................................................... 81

4.4.1 Mainshock ........................................................................ 81

4.4.2 Aftershocks ....................................................................... 83
CHAPTER 5 Discussion ........................................................................ 88
5.1 Correlation of the Aftershocks with Mapped Faults ................................... 88
5.1.1 Larina (1960) and Surrnilova et al. (1986) Variant ......................... 93

5.1.2 Shilo (1961) Variant ........................................................................ 94

5.1.3 Kurushin et al. (1976) Variant ......................................................... 96

5.1.4 lmaev et al. (2003) Variant ............................................................. 96

5.2 Correlation of Aftershocks with lsoseismal Data ....................................... 97
5.3 Temporal Distribution of Aftershocks ............................................... 98
5.4 Outlier Events .............................................................................. 99
5.5 Magnitude versus Time ............................................................... 100
5.6. Tectonic Implications of the Artyk Sequence .................................... 100
5.6.1 The Artyk Earthquake and Aftershocks .................................. 100

5.6.2 Plate Tectonic Setting of the Artyk Earthquake Sequence ......... 104

5.6.3 Displacement Rate on the Kobdi Fault .................................. 109

5.6.4 Continuations of the Kobdi Fault .................................................. 109
5.6.4.1 Northwest Continuation .......................................... 109

5.6.4.2 Southeast Continuation .......................................... 110

CHAPTER 6 Conclusions ..................................................................... 111
BIBLIOGRAPHY ................................................................................. 114

viii

Table I
Table II
Table III
Table IV

Table V

LIST OF TABLES

Epicentral Determinations for the Mainshock ................................. 19
Magnitude of the Mainshock ......................................................... 21

Focal Mechanism Determinations for the Mainshock ..................... 24
Temporary Seismic Stations .......................................................... 34
Quality Criteria Used in This Study ................................................ 77

LIST OF FIGURES

Images in this thesis are presented in color.

Figure 1-1. Seismicity map of eastern Russia based on MSU Eastern Russia

Figure

Figure

Figure

Seismicity Database. Events in the Aleutian Islands are not included in
the database and are not shown. Dots are proportional to event size
(inset). Older events (pre-1960) are shown as grey dots without outline.
The locations of the Chersky Seismic Belt (CSB) and OIekma-Stanovoi
Belt (OSB) are shown in yellow shading. The epicenter of the Artyk
earthquake of May 18, 1971 is labeled .............................................. 2

1-2. Generalized tectonic map of northeastern Russia. Solid green lines
show boundaries of major plates and blocks. Dashed lines show minor
boundaries and the light green shaded area represents the region of high
deformation in the northern Okhotsk plate. Arrows show directions of
relative plate motion. The Artyk earthquake is labeled. Plates: NA =
North American; EU = Eurasian; OK = Okhotsk, PA = Pacific. Blocks: LA
= Laptev; AM = Amur; BE = Bering. Representative focal mechanisms are
shown as lower hemisphere projections with the compressional quadrants
shaded. Approximate location of North America - Eurasia pole is shown
by shaded yellow dot. After Fujita et al. (in press) ............................... 5

1-3. Map of major faults and Iineaments in the central CSB. Dashed lines
represent faults whose location or existence is less certain. Thrust faults
are shown by toothed lines (teeth on upper plate). Blue dot shows
location of Artyk earthquake. ......................................................... 9

1-4. Schematic diagrams showing possible plate configurations relative to
the Artyk earthquake (red star). a. Artyk event is located on NA-EU plate
boundary, which may be changing location. b. Artyk event is on EU-OK
plate boundary. c. Artyk event is located at triple junction between NA,
EU, and OK. d. Artyk event represents an intraplate event in OK which is
breaking into blocks under compression between NA and EU. ............ 13

Figure 2-1. Index map of the local study area. Settlements are represented by

black dots. Mt. Khulamrin represented by black triangle ..................... 17

Figure 2-2. Comparison of epicentral determinations for the Artyk mainshock
(numbers refer to solutions listed in Table 1). Russian solutions shown as
red dots. Teleseismic solutions as blue dots. MSU solutions as orange
dots. Best solution is shown by larger orange dot. The diorite of Mt.
Khulamrin and an outlier exposure are shown by the red outlines .......... 18

Figure 2-3. Sections of representative vertical component WWSSN long-period
analog .seismograms showing surface waves digitized by Fujita et al.
(2002) to calculate the seismic moment of the Artyk earthquake. (top)
Corvallis, Oregon; (bottom) St. Johns, Newfoundland. Tick marks every
minute ...................................................................................... 22

Figure 2-4. Focal mechanism of the Artyk mainshock using synthetic
seismograms modeling of long-period WWSSN records (Riegel et al.,
1993). Upper trace for each station is the observed record, the bottom
trace is the synthetic. The parameters of the crustal model are given at
upper-right (a is P-wave velocity in kmlsec, [3 is S-wave velocity in
kmlsec, p is density in gm/cm3, and T is layer thickness in km).
Stereogram shows lower-hemisphere projection of first motions re-read
from WWSSN seismograms; open circles are dilitations, solid circles are
compressions. ........................................................................... 25

Figure 2-5. Intensity map of the main shock (Kurushin et al., 1976): 1 - main
shock epicenter; 2 — area of ground deformation; Intensity of shaking, in
points (MSK scale): 3 — greater than VIII; 4 - VII to VIII; 5 - VII; 6 - VI to
VII; 7-VI; 8-Vto Vl; 9-V; 10— IV; 11 - III to IV; 12- III; 13- II to III;
14 — not felt ............................................................................... 28

Figure 26 Area of mass movements, outlined in red, identified during field
surveys after Artyk earthquake (Kurushin et al., 1976; Smirnov and
Levashova,

1988) ....................................................................................... 31

Figure 2-7. Photographs of ground failures and landslides from the Kobdi
Stream valley referenced in figure 2-6 (photos courtesy of BM.
Koz’min) ................................................................................... 32

xi

Figure 2-8. Distriubution of aftershocks of the Artyk earthquake as determined
by the Yakut network. Inset shows cross sections of the aftershocks
along lines parallel (A-A1) and perpendicular (B-B1) to the strike of the
epicentral field (after Koz’min, 1984). 1 — Mainshock epicenter; 2 —
Aftershocks, scaled by size; 3 — Temporary station; 4 — Faults; 5 - Upper
Nera basin; 6 — Region of maximum ground failures ........................... 35

Figure 2-9. Russian “Meteor” satellite image showing the conspicuous Ulakhan
fault (arrows). Courtesy B. M. Koz’min ............................................. 39

Figure 2-10. Teleseismic events that have been mapped in the vicinity of the
Ulakhan and Kobdi faults. Relocated epicenters from MSU Database...42

Figure 2-11. Generalized faults and geographic features in the Artyk earthquake
area. Faults shown in black with black text. The Kobdi and Arkagala
faults form the Chai-Yureya Fault. Basins are highlighted red ............... 44

Figure 2-12. Satellite image of the Nera River basin showing proposed trace of
the Kobdi Fault (white arrows) and the offset diorite of Mt. Khuramlin.
Red arrows show possible left-lateral river offsets. Blue shaded area
shows region with greatest number of landslides. ............................. 46

Figure 2-13a. Detailed fault map of the epicentral region of the Artyk earthquake
showing the Larina (1960), and Surmilova (1976) and Kurushin et al.
(1976) interpretations for the Kobdi and Arkagala faults ...................... 48

Figure 2-13b. Detailed fault map of the epicentral region of the Artyk earthquake
showing the Shilo (1961) and lmaev et al. (2003) interpretations for the
Kobdi and Arkagala faults. ........................................................... 50

Figure 2-14. Detailed seismicity map of the epicentral region of the Artyk
earthquake . Teleseisms shown as large, blue dots, microseismicity
shown as small, green
dots ......................................................................................... 54

Figure 3-1. Permanent and temporary stations operating in northeast Russia
during the Artyk aftershock sequence. Network boundary shown as red
line; YAK refers to Yakutsk network; MAG refers to Magadan network.
Permanent stations are represented as blue triangles. Temporary
stations are represented by green triangles. ..................................... 58

xii

Figure 4-1: Overview showing locations of epicenter location experiments.
Inverted triangles represent temporary stations. Faults after lmaev et al.
(2003) shown as black lines. ......................................................... 69

Figure 4-2. Event of May 18, 1971, 23:09:09 UTC. Relocation is represented by
a large, red circle. Yellow triangles represent epicenters for which
regional and teleseismic arrivals used in the calculation; blue circles are
where only regional arrivals were used in the calculation. For both,
regional station arrivals were removed from the calculation on a station by
station basis (the station that was removed is noted when significant....71

Figure 4-3. Event of May 19, 1971 at 03:16:31 UTC. Relocated epicenter
calculation using all stations is represented by a large, red circle. Blue
circles are events for which station arrivals (regional, temporary) were
removed from the calculation on a station by station basis (the station that
was removed is noted when significant). .......................................... 72

Figure 4-4. Event of July 3, 1971 at 06:51:59 UTC. Relocated epicenter
calculation using all stations is represented by a large, red circle. Blue
circles are events for which station arrivals (regional, temporary) were
removed from the calculation on a station by station basis (the station that
was removed is noted when significant) ........................................... 73

Figure 4-5. Event of August 4, 1971 at 02:11:47 UTC. Relocated epicenter
calculation using all stations is represented by a large, red circle. Blue
circles are events for which station arrivals (regional, temporary) were
removed from the calculation on a station by station basis (the station that
was removed is noted when significant). .......................................... 75

Fig. 46: Plot of June and early July locations which had stations UN1S and
SUUS reporting. Inverted triangles represent temporary stations. Green
circles are the locations calculated with both stations; red triangles are
locations calculated with station SUUS removed; blue diamonds are
locations calculated with station UN1S removed. Faults after lmaev et al.
(2003) shown as black lines ........................................................... 76

Figure 4-7: Mainshock (blue), group 1 (yellow) and group 2 (red) quality
categories events on lmaev et al. (2003) mapped faults. Inverted triangles
represent temporary stations. ........................................................ 78

xiii

Figure 4-8: Mainshock (blue), group 3 (light blue) quality categories events on
lmaev et al. (2003) mapped faults. Inverted triangles represent temporary
stations .................................................................................... 80

Figure 4-9: Comparison of the Russian original (blue) and relocated (green)
epicenters. Inverted triangles represent temporary stations. Faults after
lmaev et al. (2003) shown as black lines. 82

Figure 4-10. Epicenters from the month of May. Main shock represented as
large, red circle. There were no temporary local stations (inverted
triangles) operating during this time period. Faults after lmaev et al.
(2003) shown as black lines ..................................................................... 84

Figure 4-11. Epicenters from the first of June through early July. Main shock
represented as large, red circle. Regional and local temporary stations
(inverted triangles) were operating during this time period. Faults after
lmaev et al. (2003) shown as black lines. ........................................ 86

Figure 4-12. Epicenters from late July through the end of August. Main shock
represented as large, red circle. Only regional stations were operating
during this time period. Temporary stations represented as inverted
triangles. Faults after lmaev et al. (2003) shown as black lines. .......... 87

Figure 5-1a. Relocated epicenters (blue dots) superimposed on top of faults as
mapped by Larina (1960) and Surmilova et al. (1986), and by Kurushin et
al. (1976). ................................................................................. 89

Figure 5-1 b. Relocated epicenters (blue dots) superimposed on top of faults as
mapped by Shilo (1961) and by lmaev et al. (2003). .......................... 91

Figure 5-2. Satellite image of Mt. Khuramlin and vicinity showing the Shilo
(1961) mapping of faults. ............................................................. 95

Figure 5-3. Distribution of aftershocks over time (after Kochetkov and Koz’min,
1976). The relative increase in June and early July reflects the time
when the temporary stations were operational ................................. 101

Figure 5-4. Relocated earthquakes plotted by magnitude. Red = main shock,
M6.4. Aftershock earthquake magnitudes identified by diameter of circle.
.............................................................................................................. 1 02

xiv

Figure 5-5. Seismicity and focal mechanisms of the northeast Russia study area.
Major faults shown in black, earthquakes in green (scaled by magnitude).
Focal mechanisms determined for earthquakes are shown in larger circles
as lower hemisphere equal-area projections with compressional
quadrants colored. Red events are well constrained, events in black are
poorly constrained or of unknown quality. Dates are given for events with
focal mechanisms and the Artyk earthquake is labeled. .................... 107

XV

Chapter 1: Introduction

The present-day plate boundary between North America and Eurasia is
well studied and understood in the North Atlantic and Arctic Ocean basins
where it is located along the Mid-Atlantic and Gakkel ridges. Details of how it
continues into the northeast Asian continent and its specific location between
the Arctic and Pacific oceans, however, have remained less well understood.
This segment of the plate boundary is one of the last frontiers in understanding

the framework of global plate movements.

The location of the plate boundary in northeast Russia lies within a
zone of seismicity known as the Chersky Seismic Belt (CSB; e.g., Koz’min,
1984, Parfenov et al., 1988; Fujita et al., in press). This is a wide zone that,
loosely defined, extends some 2000 km from the Laptev Sea of the Arctic
Ocean to northern Kamchatka and Sakhalin Island (Figure 1-1). It is an
intracontinental convergence zone and, like similar regions elsewhere, exhibits
considerable complexities with the possibility of numerous blocks or
microplates, and wide zones of diffuse deformation (Jackson and McKenzie,

1988; England and Jackson, 1989; Gordon, 1998).

Previous plate boundary studies within northeast Russia have focused
on the location and relative motions occurring between the North American
and Eurasian plates and the existence of an Okhotsk microplate (e.g.
Chapman and Solomon, 1976; Savostin and Karasik, 1981; Savostin et al.,

1982; Cook et al., 1986; Fujita et al., 1990a; Parfenov et al., 1988; Riegel et
1

Figure 1-1. Seismicity map of eastern Russia based on MSU Eastern Russia
Seismicity Database. Events in the Aleutian Islands are not included in the
database and are not shown. Dots are proportional to event size (inset).
Older events (pre-1960) are shown as grey dots without outline. The locations
of the Chersky Seismic Belt (CSB) and Olekma-Stanovoi Belt (OSB) are
shown in yellow shading. The epicenter of the Artyk earthquake of May 18,
1971 is labeled.

    

 

  

ON.
06 O
96 o
6.? .

:26
83.8.:

 

 

 

 

 

al., 1993). The region is understood to represent a zone of diffuse
compression (e.g., lmaev et al., 1990; Fujita et al., 1990a, in press) resulting
from the convergence of the North American and Eurasian plates rotating
around an Euler pole located near the southern end of the Laptev Sea (Figure
1-2; e.g., Cook et al., 1986; Calais et al., 2003; Steblov et al., 2003). Although
the general location and nature of the plate boundary has been relatively well
determined in the previously cited studies, its details and how the northern

Okhotsk plate is deforming are still open to interpretation.
1.1 Tectonic setting

The structure of the plate boundary region in northeast Russia is very
complex due to its geological history. Spatially, it lies within a zone of
accreted terranes wedged between the stable platforms of the North America
and Eurasia. Recent studies suggest that the terranes were rifted from the
eastern edge of the Siberian platform in the Devonian, amalgamated offshore
in the Early Jurassic and reaccreted to the edge of the platform in the Late
Jurassic (Nokleberg et al., 2000; Parfenov and Kuz’min, 2001). In the
Cenozoic the present-day plate boundary developed through northeastern
Russia that, with the opening of the Eurasia basin of the Arctic Ocean,
resulted in general compression and the southward extrusion of the Okhotsk

microplate (Cook et al., 1986).

Figure 1-2. Generalized tectonic map of northeastern Russia. Solid green
lines show boundaries of major plates and blocks. Dashed lines show minor
boundaries and the light green shaded area represents the region of high
deformation in the northern Okhotsk plate. Arrows show directions of relative
plate motion. The Artyk earthquake is labeled. Plates: NA = North American;
EU = Eurasian; OK = Okhotsk, PA = Pacific. Blocks: LA = Laptev; AM =
Amur; BE = Bering. Representative focal mechanisms are shown as lower
hemisphere projections with the compressional quadrants shaded.
Approximate location of North America - Eurasia pole is shown by shaded
yellow dot. After F ujita et al. (in press).

 

 

The present.day North America-Eurasia plate boundary extends along
the Mid-Atlantic Ridge through Iceland and then northward to the Arctic
Ocean. Within the Arctic Ocean basin the boundary forms a narrow line of
primarily nonnal-faulting earthquakes (Savostin and Karasik, 1981; Avetisov,
1996; Engen and Eldholm, 2003; Schlindwein et al., 2007) no wider than 50
km, following the Gakkel (Arctic) Mid-Ocean ridge (e.g., Wilson, 1963; Vogt et
al., 1979; Cochran et al., 2003), which extends to the Laptev Sea near the
Lena River delta. In the continental shelf of the Laptev Sea, the zone of active
seismicity is offset to the east and forms a band along the eastern edge Buor
Khaya Bay at the southern end of the Laptev Sea (Fujita et al., 1990b). A
second band of seismicity extends along the south coast of the Laptev Sea.
This, combined with the relative aseismicity of the central Laptev Sea has led
some workers to propose the existence or future development of a Laptev Sea
microplate between the two bands (Avetisov, 1993, 2000; F ranke et al.,2000;
lmaev et al., 2003). These two bands join near the Lena River Delta and

merge with the earthquakes of the CSB.

Global positioning system (GPS) based determinations for the North
America-Eurasia Euler pole have placed it near the southern end of Buor
Khaya Bay (Figure 1-2; Calais et al., 2003; Sella et al., 2002; Steblov et al.,
2003), very close to the location proposed by Cook et al. (1986) using
earthquake slip vectors. The present-day Euler pole location implies that there
is extension to the north in the Laptev Sea, and convergence to the south in

the CSB and the Sea of Okhotsk region, as first recognized by Wilson (1963).
7

South of Buor Khaya Bay, the earthquakes change to predominantly
thrust events with convergence between North America and Eurasia. How this
convergence south of the Euler pole is accommodated is not entirely
understood. Most studies recognize a separate Okhotsk plate (Figure 1-2)
being extruded to the southeast (Savostin and Karasik, 1981; Savostin et al.,
1983; Cook et al., 1986; Riegel et al., 1993; Fujita et al., 1990, 1997, in press;
lmaev et al., 1990, 2000; Parfenov et al., 1988). Large strike-slip faults
striking along and parallel to the proposed North America-Okhotsk plate
boundary (Figure 1-3), which Gusev (1979) has named the “Indigirka-Kolyma”
system of faults have been mapped through geological surveys and satellite
image interpretation. Features such as slickenlines, mylonites, zones of
crushed rocks and river offsets evidence of the strike-slip nature of many of
these faults. Linear magnetic anomalies and sharp gradients in the
gravitational field further verify the location of these faults (e.g. Vashchilov,

1963; Gusev, 1979; lmaev et al., 1990, 2000).

The seismicity becomes very diffuse south of the Euler pole, indicating
that the region is actively deforming. This deformation has been suggested to
be the result of distributed deformation and thrusting (e.g. Bobrovnikov and
Izmailov, 1989), rigid extrusion (e.g. Riegel et al., 1993) or a combination of

the two (Hindle et al., 2006, in press).

In the central CSB, earthquakes appear to branch off into two zones

(Figure 1-1), which strike NW-SE, parallel to the southern Chersky Mountains

8

Figure 1-3. Map of major faults and Iineaments in the central CSB. Dashed
lines represent faults whose location or existence is less certain. Thrust faults
are shown by toothed lines (teeth on upper plate). Blue dot shows location of

Artyk earthquake.

 

10

and the Ulakhan fault, and N-S, parallel to the Sette Daban Range and the
Ketanda fault system (Figure 1-3; lmaev et al, 1990, 2000; Fujita et al., in
press). These zones are believed to represent the boundaries of the Okhotsk
plate (Parfenov et al., 1988; Riegel et al., 1993). The NW-SE striking zone of
seismicity follows the Chersky Range (Figure 1-3) to the northeastern Sea of
Okhotsk and continues further across the neck of Kamchatka to join with
earthquakes delineating the edge of the Bering microplate (Figure 1-2; Mackey
et al., 1997) and possible convergence off of northeast Kamchatka (Pedoja et

al, 2006).
1.2 Statement of the Problem

The largest earthquake (Mw 6.4) in the continental part of the CSB in
the modern instrumental era (1962 to present) occurred on May 18, 1971, in
its central part (Parfenov et al., 1988; Fujita et al., in press), a significant
distance off of either of the two branches described above. It was located
within a broader zone of seismicity encompassing the northern part of the
Okhotsk plate which has been suggested to be a region of elevated
deformation (Figure 1-2; Fujita et al., in press). Because of its size and
location, its tectonic nature is important to understanding the plate
configuration and interactions between the North American and Eurasian

plates.

As noted above, numerous faults have been mapped in the CSB

(Figure 1-3; e.g., Larina, 1960; Shilo, 1961; Kurushin, 1976; Surmilova et al.,
11

1986; lmaev et al., 2000; Smirnov, 2000). Unfortunately, it is uncertain which
faults are active because previously located epicenters were less accurate,
due to methodologies used by Russian networks, a lack of data sharing
between networks, and a lack of western access to their local data. When
faults were mapped, no distinction was made between whether a fault was
active or ancient. This, coupled with the low-resolution of available mapped
geology, makes it impossible to develop meaningful conclusions regarding

exactly where and how the Okhotsk plate is deforming.

The Artyk earthquake of May 18, 1971, presents a unique opportunity
to further study this problem. The earthquake was accompanied by an
extensive aftershock sequence that lasted several months. A four station
temporary network was deployed in the area by the Yakutian Institute of
Geological Sciences (YIGS, now the Institute of Diamond and Precious
Mineral Geology) and the Yakutsk Experimental Methodological Seismological
Division (Yakutsk EMSD; now the Yakut Affiliate of the Siberian Division of the
Geophysical Survey of Russia; hereafter Yakut regional network) and the
Northeast Interdisciplinary Scientific Research Institute (NEISRI) and the
Magadan Experimental Methodological Seismological Division (Magadan
EMSD; now the Magadan Affiliate of the Geophysical Survey of Russia;
hereafter Magadan regional network) . The aftershock sequence provides a
data set which allows an opportunity to accurately map active faulting within a

part of this complicated region.

12

 

Figure 1-4. Schematic diagrams showing possible plate configurations relative
to the Artyk earthquake (red star). a. Artyk event is located on NA-EU plate
boundary, which may be changing location. b. Artyk event is on EU-OK plate
boundary. c. Artyk event is located at triple junction between NA, EU, and OK.
d. Artyk event represents an intraplate event in OK which is breaking into
blocks under compression between NA and EU.

13

The Artyk earthquake has been proposed to represent one of four
possibilities in the context of an extruding Okhotsk plate being compressed

and deformed between its larger neighbors:

1. The Artyk earthquake occurs on the North America - Okhotsk plate
boundary (Figure 1-4A). This proposal has not been seriously considered
except in the context of a possible change in the location of the plate boundary

(Fujita et al., 2002).

2. The Artyk earthquake occurs on the Okhotsk - Eurasia plate
boundary (Figure 1-4B). The Okhotsk - Eurasia plate boundary has been
drawn to pass through the Artyk sequence as it represented the largest events

in the area (e.g., Bird, 2003).

3. The Artyk earthquake cluster represents a triple junction between
the Okhotsk, North American and Eurasian plates (Figure 1-4C; lmaev et al.,

2003).

4. The Artyk earthquake occurs along a fault internal to the Okhotsk
plate that is accommodating intraplate deformation (Figure 1-40; Hindle et al.,

in press).

In this thesis I relocate aftershocks of the Artyk earthquake and use
satellite, topographic, and geologic data to determine which mapped local

fault(s) were ruptured by this earthquake and the implications this would have

14

for the resolution of the above scenarios within the context of deformation in

the North America — Eurasia plate boundary zone.

15

Chapter 2: The Artyk Earthquake and its Regional Tectonic Setting
2.1 The Artyk Earthquake of May 18, 1971
2.1.1 Hypocenter and Origin Time

The Artyk earthquake occurred at approximately 7:45 AM local time
(22:44:49 UTC) near the settlement of Artyk (Figure 2-1) in the Oimyakon
Uluus of the Sakha Republic (Yakutia) of the Russian Federation (formerly the
Yakut Autonomous Soviet Socialist Republic of the Soviet Union). The
earthquake was initially called the Oimyakon earthquake by Yakutian
seismologists. The epicentral region is located on the northeastern side of the
Upper Nera basin, along the Chai-Yureya Fault (lmaev et al., 2003; see further

below).

The epicenter was initially located by using arrivals at ten stations of the
Yakut regional network operated by the Yakutian Institute of Geological
Sciences (Y IGS) and the Magadan regional network operated by the
Northeast Interdisciplinary Scientific Research Institute (NEISRI; see section
3.1). These data were later supplemented by that of the Soviet national
network and subsequently from international global networks (Figure 2-2;

Table I).

We relocated the epicenter using a local crustal model developed by
Mackey (1999) (see section 4.1). The resulting relocated epicenter (Solution

12, Table I; see section 4.4.1) calculated using regional and teleseismic data
16

143' 144' 145' 146' 147'

   
    
    
    
 
  

 

   

     
 

  

 

 

  

 

65' E ) t 65'
Ust’ Nera ’
.9 Tu ngussky
Bu rustakh
Artyk R-
Artyk '
64' Mt. Khulamrin Q 54-
7%. a
‘P
Delyankir
Ust’
Kakchan
63' I 63‘
143' 144' 145‘ 146‘ 147'

Figure 2-1. Index map of the local study area. Settlements are represented
by black dots. Mt. Khulamrin represented by black triangle.

 

 

 

 

 

 

. l
u .
s 7 "‘T'I \ a
P. .3 I
‘- l i ; 0 '
K I b 5.:
1- 1 ’ .
. i “
64‘ i ‘-" 64'
J '
Ir
;
:53,“
5
C.
146'

Figure 2-2. Comparison of epicentral determinations for the Artyk mainshock
(numbers refer to solutions listed in Table 1). Russian solutions shown as red
dots. Teleseismic solutions as blue dots. MSU solutions as orange dots.
Best solution is shown by larger orange dot. The diorite of Mt. Khulamrin and
an outlier exposure are shown by the red outlines.

Table I

Epicentral Determinations for the Mainshock

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ID Reference Origin Time Latitude Longitude Depth
(UTC) (km)

1 Yakut regional network 22 44 43 64.03 145.98
LZemletryasenlya v SSSR 1971)

2 Yakut regional network (Byulletin’ 22 44 49.5 64.03 145.98
Zemletryasenii Sibiri, 1973)

3 Kondorskaya and Shebalin 22 44 43 64.0 146.1 16
(1977)

4 Obninsk (Moscow) 22 44 38 64.1 146.3

5 NEIS 22 44 43.8 63.95 146.112 33 fixed

6 ISO 22 44 39.25 63.9211 146.1013 0 fixed

7 Istanbul QSK) 22 44 43.8 63.9 146.1

8 Engdahl et al., 1998 22 44 42.22 63.9229 146.1149 8.5

9 Mackey, 1999 22 44 41.90 64.05 145.76 33.1

10 Mackey and F ujita, 2001 22 44 42.23 63.960 146.064 20.85

11 This study (regional orLIy'L 22 44 42.11 63.8530 145.8081 13“

12 This study 22 44 39.23 63.97 146.05 13"

 

* all regional stations from both the Yakut and Magadan networks.
** Depth constrained from waveform modeling by Riegel (1994) discussed below.

(see section 3.2) is very close to that of the teleseismic solutions of the
International Seismological Center (ISC), Engdahl et al. (1998), and Mackey
and Fujita (2001), and is located southeast of Mt. Khulamrin and lies near the
mapped trace of the Chai-Yureya Fault. It falls latitudinally between the
regional and teleseismic determinations and slightly to the west of other
solutions. The relationship of the mainshock to the aftershock sequence is

discussed in section 4.4.1.

19

All epicentral determinations (Table I) place the Artyk earthquake within
several kilometers of the trace of the Chai-Yureya as mapped by lmaev et al.

(2003; see also sections 2.2.3 and 5.1).

Focal depths (Table I) calculated using teleseismic stations are
generally unreliable; the ISC depth is fixed at 0 and the NEIC depth is fixed at
33 km. Engdahl et al. (1998), Mackey (1999), and Mackey and Fujita (2001)
calculated depths using teleseismic and regional P-wave arrivals; Mackey
(1999) and Mackey and Fujita (2001) also used 89 phases. Results from
these studies converged to crustal depths, but still differ greatly from each
other (8 - 30 km). Riegel et al. (1993) constrained the focal depth using long-
period waveform modeling at 13 km (see section 2.1.3); earlier modeling by
McMuIlen (1985) obtained a depth of 10 km. We adopted the Riegel et al.
(1993) solution of 13 km in relocating the mainshock. It is mid-way between
the high quality Engdahl et al. (1998) relocated depth (8 km) and the
Kondorskaya and Shebalin (1977) solution at 16 km. The Yakut regional

network did not calculate a focal depth for this event.
2.1.2 Magnitude and Energy

Estimates of the magnitude of the Artyk earthquake have varied

between 5.8 and 7.1 (Table II), depending on the author and method used.

In order to better quantify the magnitude of the event, Fujita et al.

(2002) used World-Wide Standardized Seismograph Network (WVVSSN) long
20

Table II

 

 

 

 

 

 

 

 

 

 

 

Magnitude of the Mainshock
Reference Type Value
Yakutsk regional network M 7.1
Zemletryasenia v SSSR mm 6.8
(Kondorskaya and Shebalin, 1977)
Zemletraseniya v SSSR mpv - SKM-3 (short 6.5

period)

Obninsk M, 6.4
NEIC M, 5.8
ISC M, 5.9
Obninsk ms 6.9
NEIC ms 6.6
Fujita et al. (2002); adopted in this Mw 6.4
study

 

 

 

 

period analog the seismograms (Figure 2-3) from eight stations to calculate a
moment for this event using the method of Okal and Talandier (1989). Vertical
component seismograms from VWVSSN stations COR, PMG, STU, GUA,
TRN, ESK, LEM, and STJ were digitized and moments were calculated for
various periods of 50 to 200 seconds at each station. The results were
corrected for the focal mechanism using the solution of Riegel et al. (1993; see
section 2.1.3) and averaged over the various periods. The results at the
individual stations were very robust except for LEM, which is located near a
nodal plane, thus small errors in the azimuth of the nodal plane would have
large effects on the focal mechanism correction; thus the station was omitted.
The final results for the seven remaining stations were averaged to obtain a

seismic moment, Mo, of 5.1 :t 0.1 x 1025 dyne-cm, and a corresponding Mw of

21

 

 

 

Figure 2-3. Sections of representative vertical component WWSSN long-
period analog seismograms showing surface waves digitized by Fujita et al.
(2002) to calculate the seismic moment of the Artyk earthquake. (top)
Corvallis, Oregon; (bottom) St. Johns, Newfoundland. Tick marks every
minute.

22

6.4. An earlier estimate by F ilson and Frasier (1972) was approximately 1025

dyne-cm.

The initial energy estimate by the YIGS was 3 x1016 Joules (Russian K-
class 16.5; see Rautian et al., 2007 for discussion of K-class). Using the
short-period record for Corvallis, Oregon (COR), F ujita et al. (2002) calculated
a value of 9.0 x 1015 Joules (Russian K-class 15.95) using the algorithm
described in Newman and Okal (1998). As these two estimates were

essentially identical, they did not calculate additional stations.

These analyses make the Artyk event the largest well-documented
event in the Chersky Seismic Belt in the modern instrumental period, although
significantly smaller than the magnitude of 7.1 often cited by Russian authors

(e.g., lmaev et al., 2000).
2.1.3 Focal Mechanism

The focal mechanism for this earthquake has been determined by a
number of authors using a variety of techniques. The solutions are all nearly

identical (Table III) and indicate that the solution is very robust.

Due to its comprehensive nature of the data used, the solution by
Riegel et al. (1993; Figure 2-4) is adopted as most reliable for this event. They

used long-period P-wave first motions re-read from WWSSN records

23

Table III

Focal Mechanism Determinations for the Mainshock

 

 

 

 

 

 

 

 

 

 

Reference Method Strike-1 Dip-1 Rake-1 Strike-2 Dip-2 Rake-2
McMuIlen, 1985 Synthetic 303 82 2 213 88 172
seismograms
and VVWSSN
P-wave first
motions
Chapman and P wave first 313 83 7 223 89 173
Solomon, 1976 motions
Filson and P-wave and 325 90 0 235 90 180
Fraser, 1972 surface wave
inversion
Koz’min, 1984 Regional and 308 86 9 218 81 176
bulletin P-
wave first
motions
Koz'min et al., Regional P- 327 90 , 0 237 90 180
1975 wave first
motions
Franke et al., P-waves 329 76 14 236 77 165
2000
Riegel et al., Long-period 305 82 6 214 84 172
1993; Riegel, synthetic
1994 seismograms
and P-wave
first motions
Balakina et al., Regional and 314 72 12 220 79 162
1993 bulletin P-
wave first
motions

 

 

 

 

 

 

 

 

24

 

I97l 05 IB

TOL LPZ HKC LPZ MECHANISM 305 82 6
DEPTH 13 km

SOURCE TIME 0.5 4.0 0.5
CRUST:

5.5 3.30 2.9 33

6.5 3.95 3.3

 

 

Figure 2-4. Focal mechanism of the Artyk mainshock using synthetic
seismograms modeling of long-period WWSSN records (Riegel et al., 1993).
Upper trace for each station is the observed record, the bottom trace is the
synthetic. The parameters of the crustal model are given at upper-right (a is
P-wave velocity in kmlsec, B is S-wave velocity in kmlsec, p is density in
gmlcm3, and T is layer thickness in km). Stereogram shows lower-hemisphere
projection of first motions re-read from WWSSN seismograms; open circles
are dilitations, solid circles are compressions.

25

(including those from McMuIlen, 1985), regional P-wave first motions from
Koz’min (1984), and synthetic seismogram modeling using the method of
Kroeger (1987). They calculated regional take-off-angles using the Jeffreys
and Bullen (1940) velocity model which Mackey (1999) showed was a very

good first approximation for the region.

The calculated strikes (305°-327°) of the northwest striking nodal plane
cluster about the strike of the Chai-Yureya Fault (314°) in the epicentral area
(see section 2.2.2) and therefore is presumed to be the fault plane. Thus the
focal mechanism is almost pure strike-slip and represents left-lateral faulting
on a northwest—southeast striking plane. The clip angle is very steep (72-90°),
with an average around 83°, close to the 70-75° dip on the fault cited by
Smirnov (2000).

Filson and Frasier (1972), using long-period WVVSSN waveforms,
determined that the rupture propagated to the northwest over a distance of 40

km at 4-5 km/sec.
2.1.4 Macroseismic Data

Immediately after the Artyk earthquake, scientists from the Institute of
the Earth’s Crust (IEC, Irkutsk), YIGS, and NEISRI jointly collected
macroseismic data using interviews and mail questionnaires (Koz’min et al.,
1975; Kurushin et al., 1976; Koz’min, 1984). These felt reports are

summarized below.

26

The total felt area of the earthquake was over 900,000 km2 (Figure 2-5).

The maximum intensity (MSK 1964 scale; similar to Modified Mercalli) of the
earthquake (VIII) was reported by a temporary geological field survey station,
30 km north-northwest of the epicenter. Here the earthquake was felt by all.
People ran out of buildings and the sound from the earthquake drowned out
that of a running tractor. In a wooden house, a table was overturned, objects
fell from shelves, and all the windows were broken. Intensity VII was reported
from the settlements of Artyk, Delyankir, Burustakh, Ozernoe, and parts of the
inactive Tungussky mine (Figure 2-1). Near the settlement of Ozernoe, heavy
drilling equipment was moved one meter. In Delyankir, two brick lined kilns
were destroyed. Heavy furniture moved in homes, plaster collapsed, and
people struggled to stay standing. Similar effects occurred at Burustakh. In
Artyk, northwest of the epicenter, the earthquake began with strong vibrations
accompanied by a muffled rumble and a column of dust rose into the sky. The
top of the chimney of the central boiler plant swayed with an amplitude of three
meters. In the repair shop of the Artyk truck base, a three meter high

ventilation chimney collapsed and in the main building, thin, cracks formed in
the walls and plaster broke off. Standing automobiles rocked as if driven
forwards and backwards. Dishes were broken and books fell off shelves at all
four settlements. At Tungussky, mine logs forming the ceiling of the workings
fell. On the Ust’ Nera to Magadan highway (Figure 2-5), numerous rockfalls

from steep hillsides were observed between Artyk and Ust’Nera.

27

 

 

 

o Epicenter
INTENSITIES
e e o 9
VIIIVII-VIII VII vr-vu
O 0 O 0
VI V-VI v IV

9900

III-IV III II-III Not
Felt -

 

 

 

 

 

 

 

Figure 2-5. Intensity map of the main shock (Kurushin et al., 1976): 1 - main
shock epicenter; 2 - area of ground deformation; Intensity of shaking, in points
(MSK scale): 3 — greater than VIII; 4 - VII to VIII; 5 - VII; 6 — VI to VII; 7 — VI; 8
-Vto VI; 9 - V; 10 — IV; 11 - III to IV; 12 — III; 13 — II to III; 14 - not felt.

28

Intensity VI was noted at Ust’ Nera, Tomtor, Susuman and elsewhere at
a distance range of 150-170 km. The earthquake was felt by all inhabitants
indoors and by many outdoors. People ran out of their homes and thin cracks
formed in plaster. Crockery and window panes rattled, objects on tables and
shelves fell, and furniture moved. A rumble from underground was heard at

distances up to 170 km (Koz’min et al., 1975).

The earthquake was felt with intensity V at Oimyakon, Predporozhnyi
and other localities at distances of 200—250 km. Windows, plates, and crockery
rattled and shelves and ceilings squeaked. Chandeliers vibrated and books
fell from shelves. In Predporozhnyi, paint was shaken off of walls and plaster

cracked.

In the intensity IV zone (Zyryanka, Khonuu and elsewhere) the
earthquake was felt on the first floor of structures but rarely by those outside.
Crockery, windows, and bookcases rattled. Hanging chandeliers oscillated and

some objects moved.

Intensities elevated by one unit were noted along the coast of the Sea
of Okhotsk (e.g., Magadan, Talon and Takhtoyamsk). Kurushin et al. (1976)
interpret this increase as a result of the presence of discontinuous, as
opposed to the regional continuous, permafrost. Alternatively, this may simply
be a result of increased unconsolidated sediment thickness near the coast or

of higher population density and the presence of larger buildings.

29

The isoseismals of the Artyk earthquake (Figure 2-5) are generally
ellipsoidal in form, with the major axis approximately aligned with the
orientation of both the fault plane and the strike of Mesozoic and Cenozoic
structures of the Verkhoyansk-Kolyma fold system (Kurushin et al., 1976;

Koz’min et al., 1975).

The Yakut network-calculated mainshock was located about 15 km
southeast of the region with highest intensities as determined by landslides

and surface ruptures (Kobdi Stream; Figure 2-2; see section 2.1.5).
2.1.5 Landslides and Surface Ruptures

Aerial surveys and field observations were conducted by the IEC near
the mainshock epicenter to search for landslides and surface ruptures
(Kurushin et al., 1976; Koz’min, 1984). No faulting associated surface
ruptures were observed but massive disruptions of the surface vegetation
layer were observed with mass movements into the Kobdi Stream valley over
an area of about 18 km2 (Figure 2-6 and Figure 2-7). Large amounts of
material consisting of soil, water, and snow, formed cone-shaped mud flows
with a thickness of up to 5—6 m, especially at the mouths of the tributaries of
the Kobdi Stream (Figure 2-7). The landslides and debris cones are variable
in shape and they ranged from a few square meters up to 20,000 m2 in size.
The total area of the landslides in the Kobdi Stream valley was about 250,000

m2 with a volume of displaced material reaching over 100,000 m3. Because

30

 

 

 

 

 

 

 

 

 

 

 

 

145' 146' 147'
e
BIaski/akh Str.. 0 O
o . ArtykR.
a. . o
, .Ko.bdr.Sfr ':I
. 0, C. “I, ’8” "’1’-
s.‘ s 0
o e e '. '
‘ . e 0"
. 05 no
. % .’ e o .
13 O . O ' r O
. o co .
o
e .0 e
e
e . .
e
e e
. a, -
3. .
.5
km 7’ e
0 10 20
145' 146' 147'

Figure 2-6. Area of mass movements, outlined in red, identified during field
surveys after Artyk earthquake (Kurushin et al., 1976; Smirnov and
Levashova, 1988).

31

 

Figure 2-7. Photographs of ground failures and landslides from the Kobdi
Stream valley referenced in figure 2-6 (photos courtesy of BM. Koz’min).

32

most of the landslides were concentrated on the southwest side of Kobdi
Stream, Kurushin et al. (1976) inferred that the rupture was located in the
middle of the valley, rather than on the mapped fault location several

kilometers to the northeast.

Subsequently, Smirnov and Levashova (1988) compared aerial
photographs taken in 1957 and 1976 and satellite images taken in 1981 and
identified additional landslides northeast of Kobdi stream and in the valley of
the Blaskilakh stream on the north side of the Artyk River (Figure 2-6). This

expands the area of landslides and other ground failures to about 90 kmz.

Kochetkov and Koz’min (1976) noted that the region of ground failures
was offset from the epicenter by 15 km. They ascribed this difference to the
dip of the fault plane and offset of the epicenter from the surface trace of the
fault. It should be noted, however, that most of the ground failures occur in a
topographically low region and that no failures are near the igneous intrusion
forming the higher elevations closer to the epicenter. Thus, the lack of ground
failures between the epicenter and the Kobdi Stream valley is likely simply due

to ground conditions.
2.1.6 Aftershock Sequence and Temporary Deployment

Immediately after the mainshock, a series of aftershocks began to be
recorded. Only one of the aftershocks (May 18, 1971 at 23:09:09 UTC) was
recorded teleseismically.

33

At the time of the mainshock, the nearest permanent seismic station
(see chapters 3 and 4) was Ust’ Nera (UNR), operated by the Yakut network,
approximately 150 km to the northwest of the epicenter. The second closest
station was Susuman (SUUS), operated by the Magadan network,
approximately 170 km southeast. Seimchan (SEY; Magadan network) was
located about 350 km to the east of the Artyk event. No other stations

operated within 500 km of the epicenter.

Following the mainshock and several significant aftershocks, the Yakut
and Magadan networks deployed four temporary seismic stations within 50 km
of the mainshock epicenter (Figure 2-8). The stations operated short-period
vertical VEGIK seismometers with GB-III galvanometers with a magnification
of about 15-20,000 (Kochetkov and Koz’min, 1976). Timing was by radio with
automatic clock correction. The open and close dates were determined by

examining the dates on which each station reported an arrival.

Table IV

Temporary Seismic Stations

 

 

 

 

 

 

 

 

 

 

 

 

Name MSU Latitude Longitude Elevation Open/Closed Operator
Code ('N) ('E) (m) (1971)

Artyk AYKS 64.18 145.13 700 6/3 — 7/31 YIGS

Kobdi AY1 S 64.20 145.51 800 6/8 - 7/13 YIGS

Ozernoe AYZS 63.75 146.11 875 6/1 - 9/28 NEISRI

Tungussky AY3$ 64.20 146.38 1080 6/1 - 7/24 YIGS

 

34

 

 

  

 

\

 

 

 

h . .
(kmloaizTquovB
km
20 ' o
h e
, (km)
—6"'o
AI

 

 

 

-: :-1 [OO O O 0 0'2

E3E‘I4E-j5-6

 

Figure 2-8. Distriubution of aftershocks of the Artyk earthquake as determined
by the Yakut network. Inset shows cross sections of the aftershocks along
lines parallel (A-A.) and perpendicular (B-B1) to the strike of the epicentral field
(after Koz’min, 1984). 1 - Mainshock epicenter; 2 - Aftershocks, scaled by
size; 3 - Temporary station; 4 - Faults; 5 - Upper Nera basin; 6 - Region of

maximum ground failures.

35

 

Some 1,200 aftershocks were recorded during the course of 1971, of
which about 500 were located (Figure 2-8) manually using intersecting arcs on
paper. Kochetkov and Koz’min (1976) contend that the location error was less
than 2 km for most events. They also calculated the focal depth for 212
events using xz-t2 relationships with a presumed error of :5 km. The focal
depths ranged from 2 to 29 km, with the majority being in the range of 10-18
km. The Yakut and Magadan networks worked cooperatively in the collection
and analysis of data during the period of the Artyk aftershock sequence. The
epicenters calculated by the Yakut network roughly paralleled the Chai-Yureya
Fault (although somewhat different in strike) and vaguely defined a near

vertical plane (Kochetkov and Koz’min, 1976; Koz’min, 1984).

K-class estimates, based on the methodology described in Rautian et
al. (2007), for events during the aftershock sequence are between 6.5 and 13
(M approximately 1.5 to 5). Eleven of the aftershocks were larger than K-class
11 (M about 4; Koz’min et al., 1975), although only one of them occurred while

the entire temporary network was operational (July 7, 1971).

The Russian-located events fall into two general clusters: one to the
northwest of Mount Khulamrin and one to the southeast. Their main shock
solution falls between the two clusters, with a fairly significant scatter of
aftershocks (approximately 50 kilometer radius). Kochetkov and Koz’min
(1976) contend that the aftershocks concentrated to the southeast in May and

migrated to the northwest through the course of the summer of 1971.
36

In this thesis, I use the phase data from both the temporary and
permanent stations to relocate the events using a better local travel-time curve
and automated techniques (see chapter 4), and study the spatial and temporal
distribution of the aftershock sequence and its relationship to the Chai-Yureya

fault system (see section 2.2 and chapter 5).
2.2 Active Faults

Significant mapping of faults was conducted by Russian geologists
through the 1980s under a pre-plate tectonics framework. Thrust faults, in V
particular, were not as commonly identified during this time frame. Following
the collaboration of Russian and western sCientists, a paradigm shift occurred
that led to a greater understanding of the significance of many of the features
that had been mapped (e.g., Parfenov, 1991; Parfenov and Kuz’min, 2001).
The advent of satellite imagery and computer modeling has further
dramatically changed our knowledge of large-scale tectonic features in

northeast Russia.

One of the biggest questions currently associated with the seismicity of
the CSB in general, and the Artyk sequence in particular, is the association of
large earthquakes with mapped faults and their tectonic significance. This
would be simple if faults were consistently mapped and published; however,

the “mapping” of faults in northeast Russia and the interpretation of the

37

geology is a problem in and of itself, as it varies depending on what sources

and references are utilized.

Russian scientists have mapped an extensive system of generally
northwest-trending strike-slip faults throughout the CSB (lmaev et al., 1990,
1994, 2000; Gusev, 1979; Smirnov, 2000), but often place them in different
locations or link them differently. These faults (Figure 1-3; e.g., Darpir,
Ulakhan and the Chai-Yureya fault system) have been interpreted as the main

seismogenic features in the CSB (Parfenov et al., 1988).

The Ulakhan fault system parallels the general trend of seismicity of the
CSB and is thought to represent the primary present-day locus of
displacement between North America and Eurasia, while the Chai-Yureya fault
system is about 20-30° discordant to it and has been interpreted in various

ways as noted in section 1.2.

2.2.1 The Ulakhan Fault System

The Ulakhan fault (Figure 1-3) is considered to be the largest fault in
the CSB and perhaps in northeast Russia. It is clearly visible throughout its
length in satellite imagery (Figure 2-9). It was initially identified by A. S.
Simakov in 1949, and has been studied by many geologists (e.g., lmaev et al,
1990; Mal’kov, 1971; Gusev, 1979; Smirnov, 2000). Its status as the largest
fault in the region is reflected in its name “Ulakhan,” which translates as “great”

in Yakutian.

38

 

Figure 2-9. Russian “Meteor" satellite image showing the conspicuous
Ulakhan fault (arrows). Courtesy B. M. Koz’min.

39

The Ulakhan fault has been traced by Russian workers from near the
Indigirka River to the Seimchan-Buyunda basin, a distance of about 600 km
(Figure 1-3; lmaev et al., 1990, 1994, 2000; Tret’yakov, 2003). Further
northwest, the fault has been continued towards the Selennyakh or Yana
rivers under various names, where it fades. To the southeast, the Ulakhan
fault has been extrapolated from its visible termination near the southeastern
edge of the Seimchan-Buyunda basin outward into the Sea of Okhotsk (lmaev
et al., 1990, 1994, 2000). With these extensions, the total extent of this fault is
over 1500 kilometers in length. The fault generally strikes northwest, with
several releasing bends forming pull-apart basins, throughout its length (lmaev
et al., 1990, 2000; Smirnov, 2000). Horizontal slickenlines are observed on
the smooth surface of the fault (lmaev et al., 1990, 1994, 2000) and, combined
with focal mechanisms of earthquakes, the fault is interpreted as being left-

lateral (e.g., Fujita et al., in press).

The total horizontal offset on the Ulakhan has been estimated by
Russian authors as 30-50 km, based on geology (Mal’kov, 1969, 1971; lmaev
et al., 1994), and as 24 km along the entire trace of the Ulakhan based on
offset rivers (McLean et al., 2000; Fujita et al., 2004, in press; Mackey et al.,
2008b). Given that the river network is believed to have been established
about 5 million years ago, this implies an average displacement of about 0.5
cm/yr (Fujita et al., in press), which is consistent with model calculations by
Hindle et al. (2006) for various extrusion scenarios for the Okhotsk plate.

40

The Ulakhan fault proper is active with about 10-20 teleseismic events
located within 10 km of it since the mid 19603. The largest event was the
October 19, 2006, event of magnitude Mw 5.2. The Ulakhan fault also marks
the northern boundary of the zone of high seismicity in northeast Russia and is
generally considered the northern edge of the Okhotsk plate (e.g., Parfenov et

al., 1988; Riegel et al., 1993; Fujita et al., in press).

Multiple faults splay off or intersect the Ulakhan at an angle of about
30° at various points along its length. These include the Darpir, Chai—Yureya,
In’yali-Debin (Umar), and other unnamed faults (Figure 2-10; Parfenov et al.,
1988, lmaev et al., 1990, Smirnov, 2000) that cut across the highly
seismoactive part of the CSB. Hindle et al. (in press) have suggested that

these faults dissect and sliver the northern Okhotsk Plate.

2.2.2 The Chai-Yureya Fault

The Artyk earthquake is generally regarded as occurring on the Chai-
Yureya fault (e.g., Koz’min, 1984). The Chai-Yureya fault, however, has been
defined somewhat differently by different authors. The fault was first defined
by B. l. Bronsky in 1936 and was initially mapped from the southern end of the
Upper Nera Basin (Figure 2-11; section 2.3) to the Kolyma River at Orotuk
(Figures 2-10, 2-11), continuing with a change in strike further southeast

(Vashchilov, 1963; also unnamed but mapped on Larina, 1960).

41

 

 

 

 

 

Orotuk

 

 

. 62'
143' 144’ 145' 146' 147' 148' 149'

 

 

Figure 2-10. Teleseismic events that have been mapped in the vicinity of the
Ulakhan and Kobdi faults. Relocated epicenters from MSU Database.

42

Subsequently, the Chai-Yureya has been extended by various authors
(lmaev et al., 1990, 2000; Gusev, 1979; Smirnov, 2000) to both the northwest
and southeast, resulting in a continuous proposed fault 600-700 km long
between Ust’ Nera on the northwest to the northeast of Magadan (Figure 1-3)
in the southeast. The northern extension, from Ozernoe to Ust’ Nera
incorporates the Kobdi fault of Kurushin et al. (1976; see section 2.2.3). The
southern extension is shown by a number of authors who extend the Chai-
Yureya Fault further to the southeast to Atka (Gorodinsky, 1982; Korol’kov,
1992; Gusev, 1979; lmaev et al., 1990, 2000) or to the Yama River basin
(Smirnov, 2000). Because the connections between the various fault sections
are uncertain, we use the term “Kobdi Fault” for the section between
Burustakh and the Delyankir River, “Arkagala Fault" refers to the well-mapped
section between Ust’ Kakchan and Orotuk (Figure 2-10), and “Chai-Yureya
Fault” refers to the entire fault, including extensions, as used by lmaev et al.

(2003; Figure 1-3).
2.2.3 Kobdi Fault

The Kobdi Fault (Figure 2-10; Kurushin et al., 1976), as mapped,
extends from south of Mt. Khulamrin northeast to Burustakh (Fig. 2-1) along
the northeastern edge of the Upper Nera basin (Fig. 2-11; Surmilova et al.,
1986). The Kobdi Fault may continue further northwest (q.v., Shilo, 1961).

lmaev et al.

43

 

143' 144' 145' 146' 147' 148‘ 149'

Figure 2—11. Generalized faults and geographic features in the Artyk
earthquake area. Faults shown in black with black text. The Kobdi and
Arkagala faults form the Chai-Yureya Fault. Basins are highlighted red.

(1990, 2000), Gusev (1979), and Koz’min (1984) consider the Kobdi Fault to
be a continuation of the Arkagala Fault and bound the Upper Nera basin on
the northeast (Figures 2-10, 2-11), with the Nera Fault (see section 2.2.5)
possibly continuing the displacement of the Chai-Yureya Fault to the northwest
towards Ust’ Nera (Figure 2-11). In this model, the Upper Nera basin is a pull-
apart basin due to the stepping of the Chai Yureya Fault (lmaev et al., 1990,
2000).

Investigations of the Kobdi Fault in the epicentral zone of the Artyk
earthquake show slickenlines on the fault surface that indicate left-lateral
motion. In exposures along the Artyk River and its tributaries on the edge of
the Upper Nera basin, transtension is also observed (lmaev et al., 1990,
2000). lmaev et al. (1990) also note that the strike-slip character the fault is
indicated by monotonic left-lateral “knee-shaped” bends in river channels in
tributaries of the upper Nera River and that the fault is marked in higher relief

by deep saddles and linear river valleys.

Analysis of satellite images and topographic maps suggests 1-4 km of
left-lateral offset on some of the tributaries of the Nera River northwest of
Mount Khulamrin (Figure 2-12). In addition, the Late Jurassic quartz-diorite
and diorite intrusion that forms Mount Khulamrin appears to be divided into

two parts offset by left-lateral motion of about 8 kilometers (Fujita et al., 2002).

45

Displaced portion of
Mt. Kuramlin diorite

 

Figure 2-12. Satellite image of the Nera River basin showing proposed trace
of the Kobdi Fault (white arrows) and the offset diorite of Mt. Khuramlin. Red
arrows show possible left-lateral river offsets. Blue shaded area shows region
with greatest number of landslides.

46

The Kobdi Fault between Mt. Khulamrin and Burustakh has been the
most seismically active transecting fault in the CSB (Figures 2-10, 2- 14). Slx
teleseisms, including the 1971 Artyk event and one of its aftershocks, have
been located in close proximity to this fault and an additional three epicenters
to the northwest may be associated with it further northwest (between the

Burustakh and Indigirka rivers).

The exact location of the Kobdi Fault varies between authors, especially
in the vicinity of Mt. Khulamrin (Figure 2-13). Surmilova et al. (1986) map the
fault as passing south of both Mt. Khulamrin and the offset part of the intrusion
noted above (Figure 2-13a). Shilo (1961), who does not explicitly name the
fault (Figure 2-13b), and Kurushin et al. (1976) map the fault between the two
parts (Figure 2-13a). In addition, Shilo (1961) also shows a slight curvature to
the east at the southern end of the Kobdi Fault. lmaev et al. (1990, 2000) and
Koz’min (1984) do not distinguish the Kobdi Fault and refer to it as the Chai-

Yureya Fault and consider them one and the same fault (Figures 2-10, 2-13b).

One significant discrepancy is the existence and nature of the
connection between the Kobdi Fault and the Arkagala Fault. If they are
continuous, a small restraining bend is required near Ozernoe; however a
topographically low region is mapped instead as part of the Upper Nera or
Khudzhakh basins. On face value, the curve in the trace of the Kobdi fault as
mapped by Shilo (1961) would extend the Kobdi fault either into the fault

segments in the west of the Khudzhakh basin (Figure 2-10) or curve it further
47

Figure 2-13a. Detailed fault map of the epicentral region of the Artyk
earthquake showing the Larina (1960), and Surmilova (1976) and Kurushin et
al. (1976) interpretations for the Kobdi and Arkagala faults.

48

 

L! .8... . hr x”: .33. b:

Figure 2-13b. Detailed fault map of the epicentral region of the Artyk
earthquake showing the Shilo (1961) and lmaev et al. (2003) interpretations
for the Kobdi and Arkagala faults.

50

 

L:

 

 

 

L:

 

 

51

towards the In’yaIi-Debin (Umar) Fault (Figure 2-11).
2.2.4 Nera Fault

Koz’min (1984) suggests the presence of a fault that trends along the
channel of the Nera River from Delyankir to Burustakh (Figures 2-1, 2-11, 2-
13b) and continues in an arc along the southern edge of the Nera River valley
to Ust’ Nera (lmaev et al., 1990, 2000; Gusev, 1979) . The fault then defines
the southeast edge of the Cenozoic deposits of the Upper Nera basin (see

section 2.3).

lmaev et al. (1990, 2000, 2003) suggest that the Nera Fault steps and
continues the displacement of the Chai-Yureya Fault to the northwest of the
Upper Nera basin (section 2.3), thus interpreting the Upper Nera basin as a
pull-apart basin stepping the displacement from the Nera Fault to the Kobdi
Fault. They then connect the Kobdi and Arkagala faults by an arc to the
southeast (Figure 2-13b). However, some mapped segments of the Arkagala
Fault are roughly found along an extrapolation of the Nera Fault to the
southeast and it is possible that the Nera fault is contiguous with the Arkagala

fault.

The Nera fault is not particularly seismic although a few teleseisms

have occurred along it during the instrumental period (Figure 2-10).

52

2.2.5 Arkagala Fault

The Arkagala Fault is clearly pronounced in the relief (lmaev et al.,
1990, 2000) southeast of the Upper Nera Basin (Figure 2-11), especially
where it has been exploited by the river network. The fault generally strikes
330°, and has a steep dip (Gusev et al., 1976; Gusev, 1979) of 70-75 degrees

to the northeast (Smirnov, 2000).

The fault is traced as a series of en-echelon segments which form
linear watershed divides and river valleys; there are a series of neotectonic
scarps 300-600 m high (Smirnov, 2000). Multiple strands of the fault are
mapped in a zone about 5 km wide (Larina, 1960; Gorodinsky, 1982; Gusev et
al., 1976). It is also coincident with a chain of linear magnetic anomalies and a
sharp step in the gravity field (Vashchilov, 1963). The amplitude of vertical
displacement on the Arkagala Fault has been estimated as greater than 600-
1000 meters (Gusev, 1979; Smirnov, 2000). There are no estimates of
horizontal displacement, however, its present-day motion is interpreted as left-
lateral (e.g., lmaev et al., 1990, 2000; Smirnov, 2000) based on offsets of

tributaries of the Kolyma River (lmaev et al., 1990).

The Arkagala fault has been significantly less seismically active than
the Kobdi fault and the level of activity is comparable to that of the Nera fault
(Figure 2-10, 2-14).

53

 

 

 

 

 

Figure 2-14. Detailed seismicity map of the epicentral region of the Artyk
earthquake . Teleseisms shown as large. blue dots, microseismicity shown as
small, green dots.

2.2.6 Other Faults and Lineaments in the Epicentral Zone

A number of other faults have been mapped or proposed in the
epicentral region of the Artyk earthquake. Shilo (1961) maps an unnamed,
southeast striking fault extending from Mt. Khulamrin for 15 km east of, and
parallel to, the Mel’yankir River. We will refer to this as the Mel’yankir Fault

(Figure 2-13b).

Shilo (1961) and Kurushin et al. (1976) also map a series of faults of
southeast strike between the Khudzhak basin and the lnyali-Debin Fault that
continue along the extrapolation of the Kobdi Fault, east of the Chai-Yureya
fault system (Figure 2-11). The Inyali-Debin fault of lmaev et al. (1990, 2000)

and Gusev (1970) is called the Umar Fault by Smirnov (2000).
2.3 Cenozoic Basins along the Kobdi, Nera, and Arkagala Faults

Numerous small Cenozoic basins are located along the Kobdi,
Arkagala, and Nera faults (Figure 2-11). The Upper Nera basin is composed
of Early Miocene and younger Iacustrine and coarse clastic sediments (Shilo,
1961; Grinenko et al., 1998b) up to 615 m thick (Shilo, 1961). It is possible
that some sediment deposits date as Paleogene in age (Shilo, 1961). The
Late Cretaceous basement of the basins dips to the southwest (Kurushin et

aL,1976)

Southeast of the Upper Nera basin is the Khudzhakh basin (Figure 2-

11) of unknown age (Pleistocene?) with a visible sedimentary thickness of
55

50 m (Shilo, 1961). This basin is located in the transition zone between the

Kobdi and Arkagala faults.

Farther southeast along Arkagala Fault are the Arkagala and other
coal-bearing basins (Figure 2-11). Figures in lmaev et al. (1990, 2000) appear
to interpret these basins as pull-apart basins, although they are older,
Cretaceous, in age with a thin veneer of younger sediment superimposed on

them (Cherepovsky, 1999).

56

Chapter 3: Artyk Seismological Database

The use of high-quality earthquake locations is important when
seismicity patterns are analyzed to characterize the tectonics of a region. The
MSU Northeast Siberian Seismicity database has significantly improved the
seismological database for eastern Russia by integrating and using all
available data - including international, regional and local data (see Mackey
and Fujita, 1999, for details and sources). This, combined with improved
velocity models and a standardized computer-based methodology for locating
earthquakes, should greatly improve upon previous Russian hand-determined

epicenters, especially when local stations are available.

3.1 Regional Networks

The Artyk region lies along the boundary between the spheres of
responsibility of the Yakut and Magadan regional seismic networks (Figure 3-
1). In the time period of the Artyk sequence, earthquakes were located by
hand in both networks and were quantified according to the Russian K-class

scale (q.v., Rautian et al., 2007).

3.1.1 Yakut Regional Network

During the Artyk aftershock sequence, the Yakutsk network operated 7

permanent stations (Figure 3-1).

57

 

Figure 3-1. Permanent and temporary stations operating in northeast Russia
during the Artyk aftershock sequence. Network boundary shown as red line;
YAK refers to Yakutsk network; MAG refers to Magadan network. Permanent
stations are represented as blue triangles. Temporary stations are
represented by green triangles.

58

Throughout the analog era, earthquake locations determined by the
Yakutsk network, presumably including 1971, were calculated by hand, by
drawing arcs on 115,000,000 scale maps. The network was broken into
northern (north of 60°N, including the Artyk region) and southern regions; only
stations from one region were used in the location process except for KHG
and YAK which were used in both regions. Epicentral distances were
determined by $9 - Pg time differences for each station. Many earthquakes
had phase arrival times from only three or four stations, and only one P arrival
(usually Pg). In this case, the one available 89 - Pg time was used to define
the origin time and other distances were determined using the SQ - origin time
difference for the remaining stations. Pg velocities of 6.0 or 6.1 kmlsec and $9
velocities between 3.5 and 3.7 kmlsec were used depending on location and

whichever “worked best” for a particular earthquake (Mackey, 1999).

3.1.2 Magadan Regional Network

During the time of the Artyk aftershock sequence, the Magadan district
operated 6 permanent seismic stations (Figure 3-1). All stations operated at

least one complete three-component set of instruments.

During the Artyk aftershock sequence, the Magadan network also used
the arc on map method of drawing by hand (Mackey, 1999). The travel-time
curve used was derived from the 1959 Magadan - Srednikan refraction profile
(Davydova et al., 1968) with a three layer crust and calibrated for a

hypocentral depth of 5 km. The uppermost layer is 6.0 km thick and has a Pg
59

velocity of 6.0 kmlsec. The second layer is 14 km thick with a P* velocity of
6.7 kmlsec. The third layer extends to 35.0 km depth and has a Pn velocity of

8.1 kmlsec.

Mackey (1999) also notes that date listed in global databases, such as
ISC, from both the Yakut and Magadan networks are preliminary time picks by
station operators and often yield high residuals. This can be a result of poor
time picks, Pg being reported as Pn, etc. In addition, historically, the
Magadan network only had access to preliminary arrival picks from the Yakut
network station UNR, thus UNR arrivals reported in the Magadan bulletin are
sometimes different from that in the Yakut bulletin. The MSU Northeast

Siberia database incorporates the final picks determined by network analysts.
3.2 Artyk Sequence Data

We collected phase data for all events of the Artyk sequence beginning
on May 18, 1971 and ending September 6, 1971. This later date was used as
the cut-off as earthquakes in the source region tapered off dramatically and fell
outside of the expected rupture region after that time. In addition, the most of
the temporary stations closed by the end of August, 1971. The data set was
also restricted to a geographic area covering roughly 63 to 65°N and 145 to
147°E (an area of approximately 22,000 kmz). This area was defined by
where the main cluster of aftershocks located, roughly equidistant around

Mount Khulamrin (see section 2.4.6).

60

The mainshock and the first several weeks of aftershocks were
recorded regionally only by the 13 operating permanent stations of the Yakut
and Magadan regional networks. Phase data for the main shock and larger
aftershocks were obtained from the MSU Northeast Siberia Seismicity data
base which incorporates data from the lntemational Seismological Center
(ISC) on-Iine bulletin and Byulleten’ Zemletryasenii Sibin' 1971 9. (Bulletin of
the Earthquakes of Siberia in 1971; Golenetskii, 1973; hereafter Byulleten’.
This publication is the predecessor to Malena/y po Seismichnost’ Slbin' - Data

on the Earthquakes of Siberia — usually cited for more recent data).

Subsequently, as described in the previous chapter (section 2.4.6), the
Yakut and Magadan regional networks deployed four temporary stations in the
epicentral region of the Artyk earthquake to record the aftershock sequence.

The phase arrivals for the temporary stations are not listed in Byulleten’.

The phase data in the MSU Northeast Siberia Seismicity database were
then supplemented by phase arrivals at the stations of the Magadan network
and the temporary station at Ozernoe (AY2S) obtained from the unpublished
B yulleten’ Aftershokov Artykskogo Zemletryaseniya 1 8 Maya 1971 9. (Bulletin
of the Aftershocks of the Earthquake of May 18, 1971) produced by the
Magadan network. This bulletin is composed of three parts covering different
time periods: May 18 - June 10, 1971; June 11 — July 31, 1971; and August 1
— December 31, 1971. Two versions of the bulletin for the May 18 - June 10

period exist. The phase arrival times are the same in the two versions, but

61

they differ in how amplitudes are calculated. One of the versions also includes
K-class determinations for the events while the other does not. These data
were typed into the MSU database format. Epicenters were taken from
Byulleten’, with a precision of 0.1 or 0.01 degree except for the events
recorded in July which were taken from a list provided by Dr. Koz’min with a

similar precision.

The resultant files were sent to Yakutsk where Dr. Boris Koz’min kindly
entered the arrival times from the Yakut network stations (both temporary and
permanent) through August 31, 1971 (only Ozernoe operated past this date).
The data from Dr. Koz’min for May and June, 1971, were the same as in
Byulleten’. For month of July, 119 events were new or had new phase data
and 18 events for August were new or had new phase data. Data for 86
events were received from Dr. Koz’min that were not associated with a

previously located earthquake and had no listed epicenters.

In general, there were no major discrepancies between the Yakut and
Magadan data; however, the combined file was examined for inconsistencies
and errors in October, 2002. Several specific problems were identified (see
below) and sent back to Dr. Koz’min for further verification. A revised set of
data was received in late November, 2002. These data were supplemented
by phase data for 18 events from Seimchan (SEY) which the author obtained

during a trip to northeast Russia in the summer of 2001.

62

The initial data set of 146 events (mainly from Magadan aftershock
bulletin and Byulleten’) increased to 416 events with additional data from Dr.

Koz’min, and the new picks from SEY.

The merging of the Yakut and Magadan data for their Artyk deployment
did encounter some difficulties. Dr. Koz’min indicated that there were some
interpretation or time problems with the data from June 1 through June 21. In
particular, data listed as AY2S (Ozernoe) in the data from Dr. Koz’min were
different than that reported in the Magadan aftershock bulletin and turned out
in some cases to actually be AY1S. In those instances, the Magadan network
data for AYZS were used, if available, as they were the operator of AYZS.
Otherwise, arrivals for that station were omitted during that time frame. SUUS
was also sometimes double listed with MAG listing a phase as SG and YAK
listing the same phase as S. These and similar conflicts were resolved as

best possible, however, they remain a possible source for some error.

63

Chapter 4: Data Analysis

This chapter discusses the methodology used in locating the
aftershocks, post-processing quality filtering, and additional steps taken to

understand the accuracy of earthquake locations used in this study.

Many multi-event relocation techniques (e.g., master event techniques,
joint hypocenter determination, etc.) have been developed to improve relative
locations within earthquake sequences. Such techniques depend on having at
least one well located event with stations common to other events or relatively
stable data sets with consistent and well distributed stations between the
events (e.g., Douglas, 1967; Spence, 1980; Pujol, 2000). Because of the
variability in station coverage and potential quality problems associated with
the Artyk data set, as outlined in Chapter 3, we used a single-event procedure
to determine the quality of the Artyk aftershock data set and to see if epicentral

locations could be improved.
4.1 Crustal Velocities and Travel-Time Curves

Northeast Russia is a very large area undergoing complicated
deformation, and was formed through the accretion of terranes of various
crustal affinities (see introduction; e.g., Parfenov, 1991; Nokleberg et al., 2000;
Parfenov and Kuz’min, 2001). Thus, a single or simplified velocity model for
the entire region, as was used in previous Russian studies, is unlikely to

represent the local crustal velocity within a specific part of the region.

64

In order to better determine a spatially varying travel-time curve for
northeast Russia, Mackey (1999) subdivided northeast Russia into cells 3-
degrees north-south by 5 degrees east-west. In each cell, events with 2 or
more Pn phase arrivals were selected, which includes only the larger events
which are recorded at more stations with a better azimuthal coverage. These
events were then relocated using Pg velocities between 5.875 kmlsec and
6.350 kmlsec in increments of 0.025 kmlsec and $9 velocities ranging from
3.47 to 3.65 km in 0.02 kmlsec increments. For each velocity combination, the
sum of the root-mean-squared (RMS) residuals for all events was calculated.
The Pg and $9 velocity combination yielding the lowest sum was selected as

the best-fitting velocity model for each cell.

Travel-time curves for both Pg and $9 phases were calculated using a
flat-earth, straight-ray assumption since both of these phases are confined to
the crust, thus the surface distance is roughly equivalent to the linear distance

between epicenter and station.

For the region of the Artyk aftershock sequence, the best fit velocity
based on 103 events was determined to be Pg = 6.025 kmlsec and $9 =
3.510 kmlsec (Mackey, 1999). Mackey (1999) also determined that the
Jeffreys and Bullen (1940) travel-time curves for Pn and P yielded better
locations than the IASPEI 1991 curve of Kennett (1991). Thus, these

velocities and the travel-time curves were used in the relocations in this thesis.

65

Steck et al. (2009) obtained Pg = 5.99 kmlsec and 89 = 3.487 kmlsec in a

recent tomographic study.
4.2 Location Procedure

In this study, we used a standard, least squares, best fit hypocenter
location program modified by K. G. Mackey to accommodate multiple travel-
time curves and phases, specifically Pg, Pn and $9. Sn phases were not used
because they were considered too noisy in the Northeast Siberian dataset
(Mackey, 1999). The program iteratively solves for improved hypocentral and
origin time estimates. Following an iteration of the program, the previous
solution becomes the new starting value for the location program. A total of
five iterations are completed during which the change in the RMS residual

decreases to some desired value acceptable to the operator of the program.

The aftershock sequence consists of 1,200 earthquakes, of which
approximately 500 were hand located by the Russian networks (Figure 2-8).
Of these, 416 were selected to be located using the computer algorithm. The
Russian determined epicenters obtained from the Byulleten’ and Dr. Koz’min
(section 3.2) were used as the initial estimate for origin time, location and focal

depth. The travel-time curves used were discussed in section 4.1.
4.2.1 Removal of High-Residual Stations

Following an initial relocation, each event RMS residual was examined.

In general, event residuals below $1.5 sec were initially considered

66

acceptable. If a first pass did not yield convergence of the routine to an
acceptable mean residual, individual outlying phase arrivals were removed
and the epicenter was recalculated. Following the initial relocation of the
earthquakes, stations with an individual arrival residual of greater than :I:3.2
seconds were eliminated. Some arrivals with lower, but significant, residuals
were also removed if the station was very close to the epicenter (for example,
the temporary network stations) and/or if a single residual had a value several
times larger than all others for a particular event. Final event residuals were
typically below :I:1.0 second. In general, higher event RMS residuals were
associated with events that had a higher number of reporting stations and

arrivals, or with stations that were a considerable distance from the epicenter.

After the initial relocation pass, 407 events were retained. The events
that were removed from the data set either relocated to a location significantly
outside of the study area (and therefore were not considered associated with
the aftershock sequence) or they did not converge to a stable value regardless
of the removal of outlying arrivals. Of the 407 events, 121 events retained
arrivals from three or fewer stations and were not used in further analysis. 286

events had four or more stations and were included in the final results .

4.2.2 Impact of Individual Stations

The Artyk aftershock data set is unique for the study area in that there
was considerable station coverage, both local and regional, which allowed the

opportunity to assess the impact of individual stations on earthquake locations.
67

It is expected that the greater the number of stations and defining phases, and
the smaller the azimuth gap between stations around an earthquake, will result
in a better quality location. Due to the extensive amount of faulting in the area,
location accuracy needs to be within a few kilometers to be able to associate
earthquakes to specific faults. The complicated geology combined with the
relatively low number of seismic stations operating in the area means that an

individual station could significantly bias an earthquake location.
4.2.2.1 Regional Stations (General)

An experiment was performed to determine the impact of removing
selected regional stations” arrivals used in event relocations. Several events

were evaluated with different combinations of regional coverage (Figure 4-1).

One aftershock (May 18, 1971, at 23:09:09 UTC) had both regional and
teleseismic stations reporting (7 regional, 12 teleseismic). The largest azimuth
gap between all stations was 90° (but between regional stations was between
90 and 180). For this event, the effect of removing one regional station out of
the location process was minor (an average of 1.57 km when the teleseismic
stations were also included; 1.67 km without). As expected, the effect of
removing a regional station was more significant if multiple phases were
reported and if the station was closer to the epicenter. OMS is the only
regional station for this event in which P, PG and SG phases are available.

Removing it moved the location 2.89 kilometers (1.7 km if the teleseismic

68

145' 146' 147'

 

 

 

 

 

 

 

 

R
4’
YKS Y1S AY3S
710703 09
[yank/7R
710519
“3 710804 710518
64' I 64
a.
:4.
6
Y
7
z ‘9
km s ‘e,
o 10 20 3’ ‘e
145' 146' 147‘

Figure 4-1: Overview showing locations of epicenter location experiments.
Inverted triangles represent temporary stations. Faults after lmaev et al.
(2003) shown as black lines.

69

stations were included). Another observation of this experiment is that
locations using teleseismic stations plotted more to the north; with just the

regional stations, they mapped more to the south and east (Figure 4-2).

For events where station coverage was less robust, with no teleseismic
coverage, but still fairly decent regional coverage and a moderate azimuthal
gap (greater than 90 and less than 180 degrees), the effect of removing one
regional station out of the location process actually moved the location less, on
average, than the event with teleseismic data. For example, for the event May
19, 1971 at 03:16:31, with 7 stations (all regional) the effect of removing one
regional station moved the location about 1.35 kilometers on average.
Removing station UN1S was most significant, With location difference of 2.43

kilometers (Figure 4-3).

Comparing the effect of regional stations with that of the temporary
network is also useful. An example of an event with decent regional and local
station coverage is the July 3, 1971 event at 06:51 :59. There were three
temporary stations reporting during this event AY1S, AYKS, AY3S. Removing
any of the temporary stations moved the location about 0.72 kilometers on
average. Removing any of the regional stations (UN1S, SUUS, SEY or USO)
moved the location about 1.20 km, on average; if the temporary stations were
also reporting, the location moved an average 1.0 km (Figure 44). Note that
SUUS has a significant effect, ahtough still only about 5.6 km (see section
4.2.2.2).

70

64°03'

64°00'

Figure 4-2. Event of May 18, 1971, 23:09:09 UTC. Relocation is represented
by a large, red circle. Yellow triangles represent epicenters for which regional
and teleseismic arrivals used in the calculation; blue circles are where only
regional arrivals were used in the calculation.
arrivals were removed from the calculation on a station by station basis (the
station that was removed is noted when significant).

146'06'

 

 

 

 

OMS

 

 

MGD
.ramoved

 

 

146°06'

71

 

 

 

 

64°03'

64'00'

For both, regional station

145°45' 145 '48'

 

 

 

 

 

 

 

 

I _1
SEY
removed USO
removed
a
e 3
e
e e
e
o o
o
o 0
64'03' e UN13 64.03.
OMS removed
removed
1 km
145'45' 145’48'

Figure 4-3. Event of May 19, 1971 at 03:16:31 UTC. Relocated epicenter
calculation using all stations is represented by a large, red circle. Blue circles
are events for which station arrivals (regional, temporary) were removed from
the calculation on a station by station basis (the station that was removed is

noted when significant).

72

145 '45' 145'48’

 

 

 

 

 

 

 

 

SUUS
removed
0
64'06' 64‘06'
AYKS
333...: ""'°"°d
' 8 I
e . ‘ a
Q j .e
o
e
USO
.removed
I—"——"“I
m .2
l45'45' l45‘48'

Figure 4-4. Event of July 3, 1971 at 06:51:59 UTC. Relocated epicenter
calculation using all stations is represented by a large, red circle. Blue circles
are events for which station arrivals (regional, temporary) were removed from
the calculation on a station by station basis (the station that was removed is

noted when significant).
73

Another example of an event with considerable regional station
coverage (10 stations) is the August 4, 1971 event at 02:1 1 :47. There was
one temporary station reporting during this event (AYZS), which had a minor
effect on the location (removing it only moved the location 0.98 km). The
effect of removing one regional station moved the location about 1.64
kilometers on average. The effect of removing SEY was most significant,

which moved the location approximately 5.2 kilometers (Figure 4-5).
4.2.2.2 Specific Stations

When assessing the distribution of the Artyk aftershocks, it became
apparent that certain individual stations (e.g. UN1S, SUUS) were more
significant in influencing the location of an earthquake than others. For events
which had both stations UN1S and SUUS reporting, locations were observed
to move significantly when one or the other station was not used. In general,
however, the locations moved less than 6 kilometers when one or the other

station was removed.

To assess how significant this might be, I plotted events from June
through early July for which both UN1S and SUUS reported arrivals. We
chose this time period because it was the time period of maximum coverage
from regional and local stations. Events were relocated with one of the two
stations removed and compared to the location with both stations reporting.
Events which had only UN1S reporting were an average of 3.5 km offset to the

northwest; those with only SUUS reporting were an average of 2.0 km offset to
74

145°33' 145’36' 145'39'

SUUS
.removed

 

 

64'00' O 64'00'

 

SEY
.removed

 

 

 

 

EA": _

145°33' 145'36' 145'39’

 

 

Figure 4-5. Event of August 4, 1971 at 02:11:47 UTC. Relocated epicenter
calculation using all stations is represented by a large, red circle. Blue circles
are events for which station arrivals (regional, temporary) were removed from
the calculation on a station by station basis (the station that was removed is

noted when significant).

75

146'

) -Y1S AYSS

/

Blaskilakh R.

 

 

 

 

 

 

 

Fig. 4-6: Plot of June and early July locations which had stations UN1S and
SUUS reporting. Inverted triangles represent temporary stations. Green
circles are the locations calculated with both stations; red triangles are
locations calculated with station SUUS removed; blue diamonds are locations

calculated with station UN1S removed. Faults after lmaev et al. (2003) shown
as black lines.

76

the southeast. The locations were offset from each other by an average of 3.8
km, the difference generally oriented northwest-southeast; the distribution and
general location of the events, however, does not substantially change (Figure

4-6).
4.3 Location Quality

Previous studies (e.g., Yang et al., 2000) have suggested that the
confidence in epicentral location accuracy is dependent on the number of
stations within a particular distance from the epicenter and the largest
azimuthal gap between recording stations. For this study, we developed
criteria to estimate the quality of the event locations that we use to associate
earthquakes to specific faults. Three quality groups were created, based on
the number of stations reporting and the largest azimuth gap between stations.

The criteria are summarized in Table V:

Table V

Quality Criteria Used in this Study

 

 

 

 

 

 

 

 

 

Quality Group Number of Stations Largest Azimuth Gap
(degrees)
1 6+ Less than 180
5 Less than 180
3 4 Less than 180
Any Greater than 180

 

 

Relocated events are plotted based on these quality categories.

77

145' 146' 147‘

 

 

 

 

 

145' 145' 147'

Figure 4-7: Mainshock (blue), group 1 (yellow) and group 2 (red) quality
categories events on lmaev et al. (2003) mapped faults. Inverted triangles
represent temporary stations.

78

4.3.1 Group 1 Events

A total of 114 events satisfy the Group 1 criteria (Table V); few events
in the Artyk sequence fit the Group 1 criteria without the temporary network

stations.

Group 1 events fall into two clusters: one to the southeast of Mt.
Khulamrin and one to the northwest of Mt. Khulamrin (yellow events in Figure
4-7). The nominal precision is about 1:5-10 km but based on their
correspondence and proximity to mapped faults (see section 5.1) for the most
part, we expect that locations are good to within approximately 13-5

kilometers.

4.3.2 Group 2 Events

A total of 67 events satisfy the Group 2 criteria (Table V) used in this
study. Epicenters in Group 2 also have a nominal precision of about :|:10 km,
however, are significantly more scattered than those in Group 1 (red events in
Figure 4-7). The plot indicates two clusters of events, as in Group 1, although

earthquakes fall a greater distance from the mapped faults (see section 5.1).
4.3.3 Group 3 Events

Group 3 events can either be recorded by at least 4 stations, with the
largest azimuthal gap between reporting stations of less than 180°, or

recorded by any number of stations but with the largest azimuth gap between

79

145' 146' 147‘

 

 

 

 

 

 

145' 145' 14')‘

Figure 4-8: Mainshock (blue). group 3 (light blue) quality categories events on
lmaev et al. (2003) mapped faults. Inverted triangles represent temporary
stations.

80

reporting stations being greater than 180°. A total of 103 events fell into the

Group 3 criteria.

Because of the generally small number of stations, the nominal
precision is about :I:7 km; however, the Group 3 events show a significant
amount of scatter, although two clusters are observed as in the first two cases
(Figure 4-8). These events were not considered in the fault analysis in chapter

5.

4.4 Results

Of the 286 aftershocks that were analyzed, 45 events were not located
by the Russian networks (station arrivals only). Figure 4-9 compares the 241
initial locations hand-calculated by Russian networks with their relocated
counterparts (all groups) from this study. Subjectively, the relocated epicenters
near the center of the epicentral field are more tightly clustered and form a

linear band striking about 320°.

4.4.1 Mainshock

The relocated Artyk mainshock is south of Mt. Khulamrin at the
southeast end of the aftershocks (Figure 4-8), consistent with the event
rupturing to the northwest as determined by directivity studies of F ilson and
Frasier (1972). The previous Russian locations place the main shock about 20
km to the northwest of the relocation, within the center of the aftershock

sequence (Figure 2-2; Table I; Kochetkov and Koz’min, 1976; Koz'min, 1984).

81

145' 146' 147'

 

 

 

 

 

 

A R
e
o . a
45% o o o e
I Y 0 ' Mr AY3S
YKS 0 ° ‘3' .
. ..O .8 ulagnrln Dad/ankI-rR
0 0" ° . ‘6 \ ' '
8 e
. I30 5’ f g
64' of} go .0. o 64
f o e
o
e o ‘5 o . o .
e 0 ' ‘Sv
o e 7,
° 0 e 0 ‘Ja
0 szs 1"“
o A e
o o 7

\ O 0

km .0
o 10 20 \\\\\\\\ 6 i

 

 

 

145' 145' 147'

Figure 4-9: Comparison of the Russian original (blue) and relocated (green)
epicenters. Inverted triangles represent temporary stations. Faults after
lmaev et al. (2003) shown as black lines.

82

Comparison of the relocated mainshock epicenter (Table I) to
previously determined solutions yields some systematic differences. The
Russian solutions (Yakut network solutions, Obninsk, Kondorskaya and
Shebalin, 1977) generally tend to fall to the north and northeast of my

relocated epicenter.

Western teleseismic solutions (NEIC, ISC, Engdahl et al., 1998) tend to
fall to the south and southeast, less than 10 kilometers from the relocation.
The mainshock relocation from this study differs insignificantly from that

reported in Mackey and Fujita (2001).
4.4.2 Aftershocks

The better quality (groups 1 and 2) relocated aftershocks, much like the
original Russian locations, plot in two clusters (section 2.4.6; Figure 4-7). One
is northwest of Mt. Khulamrin, the other is southeast. The relocations,
however, line up much more tightly and curve slightly more to the southeast
south of Mt. Khuramlin (see section 5.1). There is still some scatter in the
relocations and some events continue to locate in the Nera basin. Relocated
epicenters moved, on average, approximately 11.2 km. The largest distance
between original and relocated epicenters was about 70 km; the largest

changes appear to occur during the time the local network was not operating.

The scatter of aftershock relocations varies greatly between three

general time periods: May (when there were no temporary local stations),

83

145' 145' 147‘

 

 

 

 

 

145' 145- 147’

Figure 4-10. Epicenters from the month of May. Main shock represented as
large, red circle. There were no temporary local stations (inverted triangles)
operating during this time period. Faults after lmaev et al. (2003) shown as
black lines.

June through early July (when all four temporary stations were operating), and
late July through September (when only one temporary station was operating).
Because no stations were operating in May at distances of less than
150 km from the aftershock region, relocations from late May events are of
lower quality and, as a result, there is a significant amount of scatter (Figure 4-

10).

Most of the epicenters located during the month of June and the first
half of July are of high quality if there is data from 6 or more stations (section
4.3). The regional and all the temporary stations were operating at that time,
and there was very robust coverage. As a result, most of the epicenters from
that time frame are significantly less scattered than during other times of the
sequence (Figure 4-11). After the closure of the temporary stations, the

epicentral scatter increases again (Figure 4-12).

85

145' 146' 147‘

 

 

 

 

 

 

145' 143' 147'

Figure 4-11. Epicenters from the first of June through early July. Main shock
represented as large, red circle. Regional and local temporary stations
(inverted triangles) were operating during this time period. Faults after lmaev
et al. (2003) shown as black lines.

86

145' 146' 147‘

 

 

 

 

 

145‘ 146' 147'

Figure 4-12. Epicenters from late July through the end of August. Main shock
represented as large, red circle. Only regional stations were operating during
this time period. Temporary stations represented as inverted triangles. Faults
after lmaev et al. (2003) shown as black lines.

87

Chapter 5: Discussion

The relocated earthquakes from the Artyk aftershock sequence were
examined in terms of their temporal and spatial distribution, their relationship
with mapped faults and Iineaments visible on satellite imagery and maps, and
with macroseismic data. These observations are used to speculate about the

nature of deformation in the northern Okhotsk plate.

5.1 Correlation of the Aftershocks with Mapped Faults

The aftershock sequence relocates, in general, north and northwest of
the mainshock epicenter, consistent with the northwesterly rupture
propagation suggested by Filson and Frasier (1972). The aftershock
sequence extends over a distance of approximately 25-30 km, roughly that
expected for a surface rupture from the empirical scaling relationships for a

Mw 6.4 event (Wells and Coppersmith, 1994).

The relocated aftershocks generally form a linear trend in the cluster
northwest of Mt. Khulamrin with an approximate strike of 315-320° (Figure 5-
1). This strike is consistent with one of the nodal planes of the mainshock

focal mechanism (314°; section 2.1.3).

To determine which proposed local fault mapping is most consistent
with the aftershock data, I compared the relocated epicenters to the four

variant mapping of faults in the Mt. Khulamrin area that have been proposed:

88

Figure 5-1a. Relocated epicenters (blue dots) superimposed on top of faults
as mapped by Larina (1960) and Surmilova et al. (1986), and by Kurushin et
al. (1976).

89

.hvr
_

 

 

 

4... ..
be.

as: 2.535%.

 

   

 

 

 

k!

 

 

L:

J

63: mi.%m

.- 38: <2.

.
.0
«an. . .
I‘J'll

.t

      

 

         

90

Figure 5-1b. Relocated epicenters (blue dots) superimposed on top of faults
as mapped by Shilo (1961) and by lmaev et al. (2003).

91

 

. Q
s— a

L

 

.88. ..a s > .

   

 

 

.ht—

 

 

 

ah. I
:8: 0......»

 

. v

   

 

L:

.3—

 

Larina (1960) and Surmilova et al. (1986), Shilo (1961), Kurushin et al. (1976),

and lmaev et al. (2003).

5.1.1 Larina (1960) and Surmilova et al. (1986) Variant

This configuration is a composite with the northern half derived from the
geologic map of Surmilova et al. (1986), and the southern half derived from

the older map of Larina (1960).

The relocated epicenters do not fit well to this mapping (Figure 5-1a).
Epicenters northwest of Mt. Khulamrin are close to the Kobdi Fault, but not as
close as in other versions of the Kobdi Fault. South of Mt. Khulamrin, the
relocated epicenters take a sharp turn to the east, while the fault mapped in

this variant steps to the west.

The Surmilova et al. (1986) mapping of the Kobdi Fault does not place
it between Mt. Khulamrin and the detached diorite (Figures 2-2, 2-12), though

the relocated epicenters clearly do.

The Larina (1960) and Surmilova et al. (1986) variant is the only one
that suggests faults in the Upper Nera basin. Such faults could serve as a
connection between the Kobdi and Arkagala faults. The faults shown in the
variant (Figure 5-1a) are parallel and offset approximately 10 km to the
southwest of the Kobdi Fault but appears to splay off of the Kobdi. Some of
the earthquakes which plot in the Upper Nera basin fit close to this presumably

buried fault.

93

5.1.2 Shilo (1961) Variant

The relocated epicenters fit very well to this mapping (Figure 5-1b).
Epicenters that locate northwest of Mt. Khulamrin fall along or immediately
adjacent to the Kobdi Fault. South of Mt. Khulamrin, the epicenters continue
to follow the curved southern end of the Kobdi Fault, which has an

approximate strike of 285-300°, and may meet up with the Mel’yankir Fault.

F ujita et al. (2002) suggested that the southern cluster was associated
with the Mel’yankir Fault (section 2.2.6). Although some of the better located
events do plot directly on the Mel’yankir fault, the totality of the data would
seem to support their association with a curve to the east of the southern end
of the Kobdi Fault. Regretably, the basis for why Shilo (1961) curved the
south end of the Kobdi Fault is not stated, thus there is no way to evaluate the
reasonableness of this mapping; surface features visible in satellite imagery

are ambiguous (Figure 5-2).

The Shilo (1961) version of the Kobdi Fault and the epicenters aligning

along it, clearly separates Mt. Khulamrin and the detached diorite.

The geometry of the faults as mapped by Shilo (1961) suggests that the
Kobdi and Arkagala faults are distinct faults which are not connected to each

other.

94

 

Figure 5-2. Satellite image of Mt. Khuramlin and vicinity showing the Shilo
(1961) mapping of faults.

5.1.3 Kurushin et al. (1976) Variant

The relocated epicenters fit very well to this variant (Figure 5—1a), as the
Kobdi Fault is very similar to that in Shilo (1961). Epicenters northwest of Mt.
Khulamrin fall along or immediately adjacent to the Kobdi Fault, until reaching
the Kobdi Stream area where most of the landslide activity occurred (section
2.1.5). The main difference between the Kurushin et al. (1976) and Shilo
(1961) variants is that the Kobdi Fault continues with a strike of 315-3'20°
within and northwest of the Kobdi Stream area in the Kurushin et al. (1976)
map, whereas Shilo (1961) has the fault strike closer to 305-310°. There is
only one mapped epicenter past this region, which fits the Shilo (1961)
mapping, but there is not enough data to discriminate these variants. South of
Mt. Khulamrin, the epicenters continue to follow the curved southern end of

the Kobdi Fault, as in the Shilo (1961) map.

As with the Shilo (1961) map, the Kurushin et al. (1976) variant of the
Kobdi Fault clearly separates Mt. Khulamrin and the detached diorite.
Kurushin et al. (1976) maps the Arkagala Fault slightly different from Shilo

(1961 ), however, it is clear that the two faults are distinct in both variants.
5.1.4 lmaev et al. (2003) Variant

lmaev et al. (2003) is the only variant that combines the Kobdi and
Arkagala faults into a single Chai-Yureya Fault. The relocated epicenters fit

their mapping of the Chai-Yureya Fault in the area immediately west and

96

 

northwest of Mt. Khulamrin, as it is identical to the Shilo (1961) and Kurushin
et al. (1976) Kobdi Fault mappings. To the northwest, the Chai-Yuerya Fault
of lmaev et al. (2003) is the same as Kurushin et al. (1976) and does not fit the
data as well as Shilo (1961), although there are insufficient data to

discriminate the two.

South of Mt. Khulamrin, the lmaev et al. (2003) map does not fit the
aftershock data at all, as the epicenters take a sharp turn east, whereas their
mapping of the Chai-Yureya Fault continues to the south-southeast. lmaev et
al. (2003) map the Chai-Yureya Fault to the southwest of both Mt. Khulamrin
and the detached diorite suggesting that the fault does not offset them. This is

again discordant with the relocated epicenters.

The connection of the Kobdi and Arkagala faults as mapped by lmaev
et al. (2003) is not only discordant with the aftershock distribution but with the

apparent geology. This is further addressed in section 5.6.1.

Thus, the aftershock distribution is most consistent with the Kobdi fault
as mapped by Shilo (1961) and suggests that there is no direct connection

between the Kobdi and Arkagala faults in the Upper Nera basin.

5.2 Correlation of Aftershocks with lsoseismal Data

The intensity distribution of the Artyk main shock (see section 2.1.4) is
oriented roughly southeast to northwest, with the long axis oriented 310-315°,

coincident with the northwest cluster of the Artyk aftershocks in the vicinity of

97

 

—-"-I 1v.-.- u

Mt. Khulamrin. The region of maximum intensity as inferred from ground
disruptions (Kobdi Stream valley; section 2.1.5) is located north of the region
of maximal aftershock activity. Although Kochetkov and Koz’min (1976)
suggested that the fact that the epicenter is located at one end of the
aftershock zone and the maximum intensity is located at the other was
significant (see section 2.1.5), this is most likely due to differences in ground
conditions with widespread alluvial deposits in the Kobdi Stream valley which

would be absent on the higher slopes of Mt. Khuramlin.

5.3 Temporal Distribution of Aftershocks

In section 4.4.2 it was noted that the station distribution varied
considerably over time and that the highest quality events are concentrated in
June and early July when the entire local network was operating. Because of
this, assessing the temporal distribution of events over the entire aftershock
sequence is skewed. Comparison of the epicentral distributions for May, June
and July (Figure 4-8), shows that during the entire period, epicenters formed
two groups to the northwest and southeast of Mt. Khuramlin, contrary to
Kochetkov and Koz’min (1976), with variations in the amount of scatter about
the Kobdi Fault. There does seem to be some decrease in the relative

number of events in the southern cluster in July.

Detailed examination of subsets of the results from June and July show
possible clusters, linear trends and variations in strike. However, it is

impossible to tell if these are real based on the data available. The number of
98

events defining these clusters and trends is small (usually < 10) and could
easily be due to differences in the reporting stations, as discussed in section
4.2.2, or unknown local variations in seismic velocities. Application of more
sophisticated methods and careful, event by event, station by station analysis

of the data are required to determine if these small-scale variations are real.
5.4 Outlier events

In the second half of June, there are three outliers (Figure 4-11) within 5
kilometers of the presumed trace of the Nera Fault (section 2.2.4) of lmaev et
al. (2003) and Kurushin et al. (1976), near the channel of the Nera River and
in the Upper Nera basin. These are high quality (groups 1 and 2) events,
which suggests that their locations are relative accurate, they may have
occurred on the Nera fault. There is only one teleseism that could be
associated with this fault, which suggests that it is considerably less active at
present that the Kobdi. Altemately, it is possible that these events are simply

statistical outliers.

Additional outliers appear to form a trend between the Kobdi and
Arkagala faults as mapped by Larina (1960) and Surmilova et al. (1986) in the
southern Upper Nera basin (Figure 5-1a). Most of these events occurred in
May (Figure 4-10), when the local network was not operating, which implies
that the events locations may be less accurate. There are also no teleseisms
which fall in this region within the Upper Nera basin, which indicates that such

a fault within the basin, if it exists, is probably not as very active at present.
99

5.5 Magnitude versus Time

The number of aftershocks decreases with time (Figure 5-3; Kochetkov
and Koz’min, 1976), but with an increase during the period the temporary

network was operating reflecting a lower detection threshold.

The distribution of events by magnitude and location shows no biases;

both clusters have roughly equal numbers of larger events (Figure 5-4).
5.6 Tectonic Implications of the Artyk Sequence
5.6.1 The Artyk Earthquake and Aftershocks

Based on the relocated mainshock epicenter and the aftershock
sequence, the Artyk earthquake ruptured the southern part of the Kobdi Fault,
with most of the rupture near Mt. Khulamrin. Although there is some scatter in
the epicentral locations, within the estimated accuracy of the relocations
(section 4.3) there is no compelling reason not to believe that the rupture is
only on the Kobdi Fault; the aftershock sequence defines a relatively linear

trend and does not show any obvious splays or secondary trends.

The depth resolution of the relocations is poor (the majority had to be
fixed at 5 km as the calculated depth went above the surface or sub-crustal)
but, taking the Russian determined depths (Figure 2-8) at face-value and
combined with the focal mechanism (section 2.1.3), the Kobdi fault is a left-

lateral strike-slip, nearly vertically dipping, fault. The bend in the

100

 

 

 

 

 

 

 

 

40 Pi

3°" l

10 - U

19 May 1 June 1July . 1 August

Figure 5-3. Distribution of aftershocks over time (after Kochetkov and
Koz’min, 1976). The relative increase in June and early July reflects the time
when the temporary stations were operational.

101

 

 

 

 

 

 

Figure 5-4. Relocated earthquakes plotted by magnitude. Red = main shock,
M6.4. Aftershock earthquake magnitudes identified by diameter of circle.

102

aftershock sequence to the east south of Mt. Khulamrin into the valley of the

Mel’yankir River can be interpreted in several ways.

First, it is possible that there is some systematic location bias for events
in this area due to seismic velocity variations. This is unlikely in that the
geology to the east and west of the Kobdi fault is the same to the north and
south of Mt. Khulamrin (Surmilova et al., 1986), and the intrusion forming Mt.

Khulamrin is too small to cause a significant location bias.

Second, the rupture may have splayed off onto a secondary fault in a
manner similar to the later stages of the November 3, 2002, Denali earthquake
(Haeussler et al., 2004), where the rupture diverted off the Denali and on to
the Totschunda fault. This would allow a geometry in which the Kobdi Fault
does, in fact, bend to the south to link with the Arkagala Fault (the Chai-

Yureya fault system of lmaev et al., 2003).

As noted in section 2.2.3, however, this geometry which was suggested
by lmaev et al. (2003), would result in a weak restraining bend in the Chai-
Yureya Fault (Figure 2-13b). The region between the Kobdi Fault at Mt.
Khulamrin and the Arkagala Fault south of Ozernoe form part of the
Khudzhakh and Upper Nera basins (section 2.3.1) and is, in general, a
topographic low area with no indications of recent compression or uplift visible
in satellite imagery. If anything, the topography suggests a releasing bend.

Thus the Artyk earthquake and its aftershock sequence most likely

ruptured only the Kobdi fault which bends to the southeast at its southern end.
1 O3

5.6.2 Plate Tectonic Setting of the Artyk Earthquake Sequence

The lack of any visible faults or clear seismicity patterns in this area call
into question whether the Kobdi and Arkagala faults join in this area. It further
eliminates any possibility that a fault system exists that could curve to the
southwest as required in the plate configuration model of Bird (2003) or the
triple junction scenario of lmaev et al. (2003). All major faults either mapped
by Russian geologists or interpreted from satellite imagery strike southeast
(Figure 1-3; lmaev et al., 2000; Fujita et al., in press). This, plus the extremely
odd geometry of the promontory of the Okhotsk plate (Figure 1-4b) argues
strongly against the Artyk event having occurred on the Eurasia - Okhotsk
boundary. In addition, if the Artyk earthquake were located on the Eurasia —
Okhotsk plate boundary, the focal mechanism should be right-lateral along a
roughly north-south plane; this is inconsistent with the well constrained

solutions described in section 2.1.3.

The Kobdi Fault may curve slightly to the east into a zone of
displacement that may act as a transfer between the Kobdi Fault and the
Inyali-Debin Fault of lmaev et al. (1990, 1994, 2000) (Figure 1-3; Umar fault of
Smirnov, 2000) some 50 km to the east. Some east-southeast striking
lineations are visible in satellite imagery. The lnyaIi-Debin Fault is visible in
satellite imagery and is also considered a left-lateral strike-slip fault (lmaev et

al., 1990, 2000). Fujita et al. (2002) speculated that in such a scenario, the
104

InyaIi-Debin fault, in association with other Iineaments visible in satellite
images could represent a fault system that could extend out into the Sea of
Okhotsk and on to Kamchatka onto which the North America-Okhotsk plate
boundary may be stepping. Field studies by Mackey et al. (2007, 2008b),
however, conclusively demonstrate that the Ulakhan fault (section 2.2.2) is
currently active in the Seimchan-Buyunda basin (Figure 1-3), thus there is no
reason to postulate an alternative locus of displacement for the primary North

America — Okhotsk boundary (Figure 1-4a).

Finally, since the Artyk event is offset over 100 km from the Ulakhan
Fault, the event cannot be located on the present-day North America —
Okhotsk boundary. The strike-slip nature of the mainshock and the linearity of
the aftershock sequence, plus the lack of candidate faults that could link the
Artyk epicentral region and the Ulakhan fault or the Ketanda fault system also
precludes the suggestion of lmaev et al. (2003) that the Artyk sequence is
located at the triple junction between the Okhotsk, North American and

Eurasian plates (Figure 1-4c).

Previously, it was believed that the Okhotsk plate, while undergoing
deformation was, on the whole, semi-rigid and undergoing extrusion by slip
along highly conspicuous boundary faults such as the Ulakhan and the
Ketanda (Figure 1-3). The location of the Artyk earthquake and the Chai-
Yureya fault system, as mapped, however, are well-within the inside of the

presumed Okhotsk plate as defined by most authors (e.g.,Parfenov et al.,

105

1988; lmaev et al., 2000; Fujita et al., 1997, in press). Given the location of
the Artyk mainshock, if the Okhotsk plate was undergoing pure contraction,
one would expect a fairly large earthquake, but with more of a thrust
component. The pure strike-slip nature of the Kobdi Fault, however, suggests
it is a major locus of deformation within the Okhotsk plate, likely due to nested
extrusion or slivering of the northern Okhotsk plate in a manner modeled by
Hindle et al. (2006, in press). This deformation could be taken up along
several major strike-slip faults (Figure 1-4d), although the strike of the

Kobdi/Chai-Yureya is different than sketched by Hindle (in press).

An examination of the general seismicity, teleseisms, and focal
mechanisms of the continental part of the Okhotsk plate indicates that this
model of accommodated deformation makes sense. The Artyk earthquake is
a left-lateral strike-slip event which maps on the Kobdi fault, which may be a
part of a greater Chai-Yureya fault system. Other left-lateral teleseismic
events (e.g., June, 1985; June, 1970; January, 1972; December, 1986) have
occurred on parallel and subparallel strike-slip faults within the Okhotsk plate
(Figure 5-5). Many other left-lateral transpressional teleseismic events have
occurred on the Ulakhan fault to the northeast, which clearly marks the NA-OK
boundary, and to the southwest with right-lateral transpressional mechanisms
on the Ketanda Fault, which marks the EU-OK boundary (Figure 1-3 and

Figure 5-5; Riegel et al., 1993).

106

Figure 5-5. Seismicity and focal mechanisms of the northeast Russia study
area. Major faults shown in black, earthquakes in green (scaled by
magnitude). Focal mechanisms determined for earthquakes are shown in
larger circles as lower hemisphere equal-area projections with compressional
quadrants colored. Red events are well constrained, events in black are
poorly constrained or of unknown quality. Dates are given for events with
focal mechanisms and the Artyk earthquake is labeled.

107

v .
N87073:

   

e
e
egg—ac!
8&5:er .aoo...

108

The regional seismicity within the continental part of the Okhotsk plate
(Figure 5-5), suggests that there are regions with elevated seismic activity,
separated by relatively “quiet” zones which could indicate that there are some
rigid slivers or blocks embedded within the deforming zone of the northern

Okhotsk plate (Figure 1-2).
5.6.3 Displacement Rate on the Kobdi Fault

As discussed earlier, the aftershocks and the mapped trace (Shilo,
1961) of the Kobdi fault are located between the Mt. Khulamrin diorite and a
detached diorite exposure south of it with an offset of 8 km in a left-lateral
sense across the fault. If these represent parts of the same diorite and if the
Kobdi fault also initiated movement at about the same time as the Ulakhan (c.
5 Ma; Fujita et al., in press), the slip rate is somewhat less than 0.2 cm/yr, or

about one-third that of the Ulakhan.
5.6.4 Continuations of the Kobdi fault
5.6.4.1 Northwest Continuation

The relocated Artyk sequence clearly shows that the Chai-Yureya fault
system ruptured in the vicinity of Mt. Khulamrin and along the Kobdi Fault to
the northwest. Figure 2-10 shows that teleseismic seismicity occurs along an
extrapolation of the Kobdi fault to the northwest, possibly trending well to the
north of Ust’ Nera. If this represents a continuation of the Kobdi fault then the

Upper Nera basin is not now acting as a pull apart as suggested by lmaev et

109

al. (1990, 2000) and displacement is not transferred to the Nera Fault towards

Ust’ Nera (Figure 2-11).
5.6.4.2 Southeast Continuation

The seismicity, however, is not as clearly defined southeast of Mt.
Khulamrin, and the Artyk aftershocks definitely do not occur on the Chai-

Yureya fault as interpreted by lmaev et al. (2003).

The results of this study leaves open the question of whether the
displacement is transferred back to the Arkagala fault through presently
unknown faults, transferred to the InyaIi-Debin to the east, or to some other
fault. Also still unresolved is whether the fault, whether Arkagala or lnyali-
Debin, is locked south of the Artyk epicentral zone and will rupture at a later
date, or whether the strain is accommodated on a large number of small faults
which will not generate a large earthquake. The recurrence interval on these
faults is sufficiently long that it is likely we have not observed any significant

portion of a full cycle of seismic activity.

110

Chapter 6: Conclusions

This study has analyzed a previously unavailable data set to relocate the
aftershock sequence of a major earthquake in northeast Russia and to place the

event in a plate tectonic context.

The Artyk earthquake occurred at 22 44 39.23 UTC on May 18, 1971, with
an epicenter located at latitude 63.97°N and longitude 146.05°E, and a depth of
13 km. It had a moment magnitude, Mw, of 6.4 and released 1016 joules of
energy. The epicenter was located to the south of Mt. Khulamrin, at the southern
end of the aftershock sequence, supporting a propagation of the rupture to the
north. The focal mechanism was left-lateral strike-slip along a fault plane striking

315° with a dip of nearly 90°.

The best-located aftershocks cluster very closely to the Kobdi fault as
mapped by Shilo (1961), including a slight curve to the east to the south of Mt.
Khulamrin. At this time, there is not enough information as to if and how the
Kobdi Fault connects to the Arkagala fault and whether they form segments of a
greater Chai-Yureya fault system, although the geometry of the southern part of

the aftershock sequence suggests they may not be directly connected.

Almost all authors agree that the Ulakhan fault represents the primary
locus of deformation between the North American and Eurasian plates over at
least the past 5 million years (e.g., lmaev et al., 2000; Riegel et al., 1993; Fujita
et al., in press). The location of the Artyk sequence clearly places it more than

100 km from the Ulakhan and onto a completely different fault. The field studies

111

of Mackey et al. (2007, 2008b) indicate the Ulakhan is presently active and
mitigates against the proposal that the primary locus of displacement between
North America and Okhotsk is shifting from the Ulakhan fault, thus the Artyk
event is not associated with the North America — Okhotsk boundary (Figure 1-

4a).

The Artyk earthquake is also not likely to have occurred on the Okhotsk —
Eurasia plate boundary (Figure 1-4b). The earthquake clearly ruptures a fault
with a strike near 300°, which would require the boundary to do a significant
curve to realign with the known locus of displacement further south in the
Ketanda fault zone (Parfenov et al., 1988; Riegel et al., 1993; F ujita et al., in
press). Almost all significant mapped faults and Iineaments continue the strike of
the Kobdi (Chai-Yureya) fault in a southeast direction (lmaev et al., 1990, 2000;
Fujita et al., in press) and no Iineaments have been reported that strike
southwest of the Artyk epicentral region. The focal mechanism is also

inconsistent with this proposal.

The Artyk sequence ruptures along a known fault system in a linear,
strike-slip manner and teleseisms align along it. There is no indication of a fault
that would link the Kobdi (Chai-Yureya) to the Ulakhan and, as noted in the
previous paragraph, it would be very difficult to link the Kobdi (Chai-Yureya) to
the Ketanda system as well as having an inconsistent focal mechanism. Thus

even if the locus of displacement were shifting from the Ulakhan to the Kobdi and

112

other fault systems, the Artyk sequence could not serve as a triple junction

between the North American, Eurasian, and Okhotsk plates (Figure 1-4c).

Thus, of the possibilities presented in the introduction, the most likely is
that the Kobdi (Chai-Yureya) represents an intraplate fault that is accommodating
the convergence of North America and Eurasia by slivering and breaking up the
northern part of the Okhotsk plate (Figure 1-4d) as suggested by Hindle et al. (in
press). As the intraplate deformation is likely to be significantly slower than that
on the boundaries, seismic activity would be infrequent and the presently known
instrumental distribution represents a short snapshot of the activity; this is
supported by the estimate of a displacement rate on the Kobdi of ~0.2 cm/yr.
Other segments of this, or other faults, may rupture in the future. In any event,
the northern part of the Okhotsk plate appears to be accommodating

convergence by splintering into blocks separated by major strike-slip faults.

Further work with detailed field mapping, more refined relocation methods
(e.g., double-difference; Waldhauser and Ellsworth, 2000), global positioning
system surveys, will be necessary to resolve such questions as the nature and
location of the continuation of the Kobdi Fault and its relationship to the Arkagala
Fault. The 50 year instrumental period for seismology in northeast Russia
represents only a fraction of the expected recurrence interval given the low slip
rates, thus we have only recorded a snapshot of the deformation processes

occurring in this complex region.

113

BIBLIOGRAPHY

Avetisov, G. P., 1993: Some aspects of Iithospheric dynamics of Laptev Sea,
lzvestiya Physics of the Solid Earth, 29, 402-412.

Avetisov, G. P., 1996: Seismoactive Zones of the Arctic, VNIIOkeanologiya,
Sankt Peterburg. (in Russian)

Avetisov, G. P.: Yet again on the earthquakes of the Laptev Sea, Geo/ago-
Geoflzichiskie Kharakhten'stiki Litosfery Arkticheskogo Regiona, 3, 104-
113, 2000. (in Russian)

Bird, P., 2003: An updated digital model of plate boundaries, Geochemistry
Geophysics Geosystems, v. 4(3), 1027, doi: 10.1029] 2001GC000252

Balakina, L.M., Zakharova, A.l., Moskvina, AG, and Chepkunas, LS, 1993:
Investigation of focal mechanisms of strong crustal earthquakes of
northern Eurasia, 1927-1991, Seismichhnost’ i Seismicheskoe
Raionirovanie Sevemoi Evrazii, 1, p. 123-131. (in Russian)

Bobrovnikov, V. A., and Izmailov, L. I., 1989: Present-day structure and
geodynamics of the Earth’s crust of the southeastern part of the Yana-
Kolyma system, in Lin’kova, T. I., and Krasnyi, L. L., eds, Geophysical
In vestigations for the Solution of Geologic Problems, SVKNII DVO AN
SSSR, Magadan, 5-23. (in Russian).

Calais, E., DeMets, C., and Nocquet, J.-M., 2003: Evidence for a post-3.16-
Ma change in Nubia-Eurasia-North America plate motions? Earth and
Planetary Science Letters, 216, 81-92.

Chapman, ME, and Solomon, SC, 1976: North American — Eurasian plate
boundary in northeast Asia, Journal of Geophysical Research, 81, 921-
930.

Cherepovsky, V. F ., ed., 1999: Coal Basins and Deposits of the Far East,
Coal basis of Russia, v. 5, book 2, ZAO Geoinforrnmark, Moscow. (in
Russian)

Cochran, J. R., Kurras, G. J., Edwards, M. H., and Coakley, B. J., 2003: The
Gakkel Ridge: Bathymetry, gravity anomalies, and crustal accretion at

extremely slow spreading rates, Journal of Geophysical Research, 108,
doi:10.1029/2002JB001830, 2003.

Cook, D. B., Fujita, K., and McMuIlen, C. A., 1986: Present-day plate

interactions in northeast Asia: North American, Eurasian, and Okhotsk
plates, Journal of Geodynamics, 6, 33-51.

114

Davydova, N. I., Shvarts, Y. B., and Yaroshevskaya, G. A., 1968. Wave
patterns in deep seismic sounding along the Magadan-Kolyma profile,
in Deep Seismic Soundings of the Earth’s Crust in the USSR:
lntemational Geology Review— Book Section, 10, 93-102.

Douglas, A., 1967: Joint epicenter determination, Nature, 215, 47-48.

Engdahl, E. R., van der Hilst, R. D., and Buland, R., 1998: Global teleseismic
earthquake relocation with improved travel times and procedures:
Bulletin of the Seismological Society of America, 88, 722-743.

Engen, O, and Eldholm, O., 2003: The Arctic plate boundary, Joumal of
Geophysical Research, 1 08( B2), 2075, doi:10.1029/2002JB001809, 17

Pp-

England, P., and Jackson, J., 1989: Active deformation in the continents,
Annual Reviews of Earth and Planetary Science, 17, 197-226.

Filson, J., and Frasier, C. W., 1972: The source of a Siberian earthquake, Eos
Transactions of the American Geophysical Union, 53, 1041.

Franke, D., Kruger, F ., and Klinge, K., 2000: Tectonics of the Laptev Sea -
Moma ‘Rift’ region: investigation with seismologic broadband data,
Journal of Seismology, 4, 99-116, 2000.

Fujita, K., Cook, D. B., Hasegawa, H., Forsyth, D., and Wetmiller, R., 1990a:
Seismicity and focal mechanisms of the Arctic region and the North
American plate boundary in Asia, in Grantz, A., Johnson, G. L., and
Sweeney, J. F., eds., The Geology of North America, v. L., The Arctic
Region, Geological Society of America, Boulder, 79-100.

Fujita, K., Cambray, F. W., and Velbel, M. A., 1990b: Tectonics of the Laptev
Sea and Moma rift systems, northeastern USSR, Marine Geology, 93,
95-118.

Fujita, K., Stone, D. 3., Layer, P. W., Parfenov, L. M., and Koz'min, B. M.,
1997: Cooperative program helps decipher tectonics of northeast
Russia, Eos Transactions of the American Geophysical Union, 78, 245,
252-253.

Fujita, K., McLean, M. S., Mackey, K. G., and Kozmin, B. M., 2002b: The 1971
Artyk earthquake: Is the locus of motion changing in northeast Russia?
Eos Transactions of the American Geophysical Union, 83(47), suppl.,
F1247.

115

Fujita, K., Sella, G., Mackey, K. G., Stein, 8., Park, K.-D., and lmaev, V. S.,
2004: Relationships between seismicity and GPS determined velocities
in northeast Asia, Eos Transactions of the American Geophysical Union,
85(47), suppl., F667.

Fujita, K., Koz’min, B. M., Mackey, K. G., Riegel, S. A., McLean, M. S., and
lmaev, V. S., in press: Seismotectonics of the Chersky Seismic Belt,
eastern Sakha Republic (Yakutia) and Magadan District, Russia, in
Stone, D. B., editor in Chief, Northeastern Russia, Geology,
Geophysics and Tectonics: A tribute to Leonid Parfenov, Stefan Mueller
series, EGS.

Gorodinsky, M. E., ed., 1982: Geologic Map of the Northeast USSR. Vsesoyuz
Nauchnyi-Issledovatel’sky Geologichesky lnstitut, Leningrad, scale
1:5,000,000 (dated 1980). (in Russian)

Gordon, R. G., 1998: The plate tectonic approximation: Plate nonrigidity,
diffuse plate boundaries, and global plate reconstructions, Annual
Reviews of Earth and Planetary Science, 26, 615-642.

Grinenko, O. V., Sergeenko, A. l., and Belolyubsky, I. N., 1998b: Paleogene
and Neogene of Northeastern Russia, Part II, Izd-vo YNTs SO RAN,
Yakutsk, 57 pp. + 36 p. supplement volume. (in Russian)

Gusev, G. S., 1979: Folded Structures and Faults of the Verkhoyansk-Kolyma
System of Mesozoides, Nauka, Moscow, 206 pp. (in Russian)

Gusev, G. S., Mokshantsev, K. B., and Tret’yakov, F. F., 1976: Faults of the
Verkhoyansk-Chukotka folded region, in Mokshantsev, K. 3., ed., Fault
Tectonics of the Territory of the Yakut ASSR, Izdanie Yakutskogo
Filiala SO AN SSR, Yakutsk, 73-114. (in Russian)

Haeussler, P. J., Schwartz, D. P., Dawson, T. E., Steuner, H. P., Lienkaemper,
J. J., Sherrod, B., Cinti, F. R., Montone, P., Craw, P. A., Crone, A. J.,
and Personius, S. F., 2004. Surface rupture and slip distribution of the
Denali and Totschunda faults in the 3 November 2002 M 7.7
earthquake, Alaska, Bulletin of the Seismological Society of America, v.
94(6B) p. 823-852.

Hindle, D., Fujita, K., and Mackey, K.: Current deformation rates and extrusion
of the northwestern Okhotsk plate, northeast Russia, Geophysical
Research Letters, 33, doi:10.1 029/2005GL024814, 2006.

116

Hindle, D., Fujita, K., and Mackey, K., in press: Deformation of the
northwestern Okhotsk plate. How is it happening? in Stone, D. B., editor
in chief, Northeastern Russia, Geology, Geophysics and Tectonics: A
tribute to Leonid Parfenov, Stefan Mueller series, EGS.

lmaev, V. S., Imaeva, L. P., and Koz’min, B. M., 1990: Active Faults and
Seismotectonics of Northeast Yakutia, Yakutskii Nauchnyi Tsentr SO
AN SSSR, Yakutsk. (in Russian)

lmaev, V. S., Imaeva, L. P., Koz’min, B. M., and Fujita, K., 1994: Active faults
and recent geodynamics of Yakutian seismic belts, Geotectonics, 28,
146-158.

lmaev, V. S., Imaeva, L. P., and Koz’min, B. M., 2000: Seismotectonics of
Yakutia, GEOS, Moscow. (in Russian)

lmaev, V. S., Imaeva, L. P., and Koz’min, B. M., 2003: Dynamics of
seismotectonic processes in the zone of articulation between the
Eurasian, North American, and Okhotsk Sea plates (northeast Asia),
Otechestvennaya Geologiya, 2003(6), p. 69-74. (in Russian)

Jackson, J., and McKenzie, D., 1988: The relationship between plate motions
and seismic moment tensors, and the rates of active deformation in the
Mediterranean and Middle East, Geophysical Journal, 93, 45-74.

Jeffreys, H., and Bullen, K. E., 1940. Seismological Tables, British
Association for the Advancement of Science, London, 50 pp.

Kennett, B. L. N., 1991. IASPEI 1991 Seismological Tables: Research
School of Earth Sciences, Australian National University, Canberra, 167

pp-

Kochetkov, V. M., and Koz’min, B. M., 1976: The Oimyakon earthquake of
1971 and its aftershocks, in Solov’ev, S. L., editor in chief, Seismicity
and Deep Structure of Siberia and the Far East, DVNTs AN SSSR,
Vladivostok, p. 61-67. (in Russian)

Kondorskaya, N. V., and Shebalin, N. V., eds., 1977: New Catalog of Strong
Earthquakes in the Territory of the USSR From Ancient Times to 1977:
Nauka, Moscow, 534 pp.

Korol’kov, V. G., 1992: State Geologic Map, sheet O-56 (Magadan), P-56, 57
(Seimchan), VSEGEI, Sankt-Peterburg. scale 121,000,000. (in Russian)

Koz’min, B. M., 1984: Seismic Belts of Yakutia and the Focal Mechanisms of
Their Earthquakes, Nauka, Moscow. (in Russian)

117

Koz’min, B. M., Emel’yanov, N. P., Emel’yanova, A. A., Zhelinskaya, E. A.,
Larionov, A. G., and Kim, V. F ., 1975: Strong earthquakes of Yakutia,
in Gorbunova, I. V., Kondorskaya, N. V., and Shebalin, N. V., eds.,
Earthquakes of the USSR in 1971, Nauka, Moscow, p. 133-141. (in
Russian)

Kroeger, G. C., 1987: Synthesis and Analysis of Teleseismic Body Wave
Seismograms, Ph.D. Dissertation, Stanford University, Stanford.

Kurushin, R. A., Dem’yanovich, M. G., and Kochetkov, V. M., 1976:
Macroseismic manisfestations of the Oimyakon earthquake, in Solov’ev,
S. L., editor in chief, Seismicity and Deep Structure of Siberia and the
Far East, DVNTs AN SSSR, Vladivostok, p. 50-60. (in Russian)

Larina, N. I., 1960. State Geologic Map of the USSR, sheet P-55 (Kolyma
River), Ministerstvo Geologii i Okhrany Nedr SSSR, (Leningrad), scale
1:1,000,000. (in Russian)

Mackey, K. G., 1999: Seismological Studies in Northeast Russia, Ph.D.
Dissertation, Michigan State University, East Lansing.

Mackey, K. G., and Fujita, K., 1999: The northeast Russia seismicity database
and explosion contamination of the Russian earthquake catalog, in
Proceedings of the 213‘ Seismic Research Symposium: Technologies
for monitoring the Comprehensive Nuclear- Test-Ban Treaty, v. 1, US
Department of Defense, LA-U R-99-4700, 151 -1 61.

Mackey, K., and Fujita, K., 2001: Seismic calibration and discrimination in
northeast Russia, in Proceedings of the 23’" Seismic Research Review:
Worldwide Monitoring of Nuclear Explosions, v. 1, National Nuclear
Safety Administration, US Department of Energy, 80-89.

Mackey, K. G., Fujita, K., Gunbina, L. V., Kovalev, V. N., lmaev, V. S., Koz'min,
B. M., and Imaeva, L. P., 1997: Seismicity of the Bering Strait region:
Evidence for a Bering Sea block, Geology, 25, 979-982.

Mackey, K. G., Fujita, K., Gounbina, L., Koz’min, B., lmaev, V., Imaeva, L.,
and Sedov, B., 2003: Explosion contamination of the northeast Siberian
seismicity catalog: implications for natural earthquake distributions and
the location of the Tanlu fault in Russia, Bulletin of the Seismological
Society of America, 93, 737-746.

118

Mackey, K. 6., Hampton, B. A., Fujita, K., Kurtkin, S., and Gounbina, L. V.,
2007: Active faulting along the Ulakhan Fault, Seimchan-Buyunda basin,
northeast Russia, Eos Transactions of the American Geophysical Union,
88(52), suppl., T13D-1579.

Mackey, K., Fujita, K., Hartse, H. E., Steck, L. K., Gounbina, L., Leyshuk, N.,
Shibaev, S., Koz'min, B., lmaev, V., Gordeev, E., Masalski, O., Gileva,
N., Borrnotov, VA, and Voitenok, A.A., 2008a: Seismicity of Eastern
Russia 1960-2007, LAUR-04-1381, 1 sheet.

Mackey, K. G., Hampton, B., Fujita, K., Koz’min, B. M., Shibaev, S. V., and
Gounbina, L. V., 2008b: Field studies of active fault zones in eastern
Russia, in Malovichko, A. A., ed., Seismicity of Northern Eurasia:
Geophysical Survey RAS, Obninsk, p. 200-204.

Mal’kov, B. I., 1969: Structure and development of the Lyglykhtakh late
geosynclinal depression, in Mesozoic Tectonogenesis (Proceedings of
the VII Session Scientific Council on the Tectonics of Siberia and the
Far East), SVKNII, Magadan, p. 74. (in Russian)

Mal’kov, B. I., 1971: Scheme of development of the southeastern part of the
Yana-Kolyma geosynclinal system, in Mesozoic Tectonogenesis,
SVKNII DVO AN SSSR, Magadan, 43-50. (in Russian)

McLean., M. S., Fujita, K., Mackey, K. G., Kleber, E., Koz'min, B. M., and
lmaev, VS, 2000: The Ulakhan fault system, northeast Russia, Eos
Transactions of the American Geophysical Union, 81 (48), suppl.,

F 1 164.

McMuIlen, C. A., 1985: Seismicity and Tectonics of the Northeastern Sea of
Okhotsk, MS. Thesis, Michigan State University, East Lansing.

Newman, A. V., and Okal, E. A., 1998: Teleseismic estimates of radiated
seismic energy: The EIMo discriminant for tsunami earthquakes,
Joumal of Geophysical Research, v. 103, p. 20885-20898.

Nokleberg, W. J., Parfenov, L. M., Monger, J. W. H., Norton, I. O., Khanchuk,
A. I., Stone, D. B., Scotese, C. R., Scholl, D. W., and Fujita, K., 2000:
Phanerozoic tectonic evolution of the circum-north Pacific, U. 8.
Geological Survey Profesional Paper 1626.

Okal, E. A., and Talandier, J., 1989: Mm: A variable-period mantle magnitude,
Joumal of Geophysical Research, 94, 4169-41 93.

Parfenov, L. M., 1991: Tectonics of the Verkhoyansk-Kolyma Mesozoides in
the context of plate tectonics, Tectonophysics, 199, 319-342.

119

Parfenov, L. M, and Kuz’min, M. l., eds., 2001: Tectonics, Geodynamics, and
Metallognesis of the Territory of the Sakha Republic (Yakutia), MAIK
Nauka/Interperiodika, Moscow, 570 pp. (in Russian)

Parfenov, L. M., Koz’min, B. M., Grinenko, O. V., lmaev, V. S., and Imaeva, L.
P., 1988: Geodynamics of the Chersky seismic belt, Journal of
Geodynamics, 9, 15-37.

Pedoja, K., Bourgeois, J., Pinegina, T., and Higman, B., 2006: Does
Kamchatka belong to North America? An extruding Okhotsk block
suggested by coastal neotectonics of the Ozernoi Peninsula,
Kamchatka, Russia, Geology, 34, 353-356.

Pujol, J., 2000. Joint event location — the JHD technique and applications to
data from local seismic networks, in Thurber, C. H., and Rabinowicz, M,
eds., Advances in Seismic-Event Location, Kluwer Academic,

Dordrecht, 163-204.

Rautian, T. G., Khalturin, V. l., Fujita, K., Mackey, K. G., and Kendall, A. G.,
2007: Origins and methodology of the Russian energy K-class system
and its relationship to magnitude scales, Seismological Research
Letters, 78, 579-590.

Riegel, S. A., 1994: Seismotectonics of Northeast Russia and the Okhotsk
Plate, MS. Thesis, Michigan State University, East Lansing.

Riegel, S. A., Fujita, K., Koz’min, B. M., lmaev, V. S., and Cook, D. B., 1993:
Extrusion tectonics of the Okhotsk plate, northeast Asia, Geophysical
Research Letters, 20, 607-610.

Savostin, L. A., and Karasik, A. M., 1981: Recent plate tectonics of the Arctic
basin and of northeastern Asia, Tectonophysics, 74, 111-145.

Savostin, L. A., Verzhbitskaya, A. |., and Baranov, B. V.: Present-day tectonics
of the Sea of Okhotsk plate region, Doklady Akademii Nauk SSSR, 266,
961—965, 1982. (in Russian)

Savostin, L., Zonenshain, L., and Baranov, 8.: Geology and plate tectonics of
the Sea of Okhotsk, American Geophysical Union Geodynamics Series,
11, 189-221, 1983.

Schlindwein, V., Miller, C., and Jokat, W., 2007: Microseismicity of the

ultraslow-spreading Gakkel ridge, Arctic Ocean: a pilot study,
Geophysical Joumal lntemational, 169, 100-112.

120

Sella, G. F ., Dixon, T. H., and Mac, A., 2002: REVEL: A model for Recent
plate velocities from space geodesy, Journal of Geophysical Research,
107, doi:10.1029/2000JB000033.

Shilo, N. A., 1961. Quaternary deposits of the Yana-Kolyma auriferous belt,
conditions and stages of its formation, Trudy VNll-1, 66, 136 pp. (in
Russian)

Smirnov, V. N., 2000: Orogenic Regions, northeast Eurasia, in Grachev, A. F.,
ed ., Neotectonics, Geodynamics, and Seismicity of Northern Eurasia,
PROBEL, Moscow, 120-133. (in Russian)

Smirnov, V.N. and Levashova, S.V., 1988: Strong earthquakes of the
northeast USSR according to interpretation of space and air photos, in
Sidorov, A.A., Izmailov, L.I., Lin’kova, TI. and Silant’er, V.N., eds.,
Nature of Geophysical Fields of the Northeast USSR: SVKNII DVO AN
SSSR, Magadan, p39-53. (in Russian)

Spence, W., 1980: Relative epicenter determination usiong P-wave arrival-
time differences, Bulletin of the Seismological Society of America, 70, ‘
171-183. .

Steblov, G. M., Kogan, M. G., King, R. W., Scholz, C. H., Burgmann, R., and
Frolov, D. I., 2003: Imprint of the North American plate in Siberia
revealed by GPS, Geophysical Research Letters, 30,
doi:10.1029l2003GL017805.

Steck, L. K., Phillips, W. S., Mackey, K., Begnaud, M. L., Stead, R. J., and
Rowe, C. A., 2009. Seismic tomography of crustal P and S across
Eurasia: Geophysical Journal lntemational, 177, 81-92.

Surmilova, E. P., Maksimova, G. A., and Natapov, L. M., 1986. State Geologic
Map of the USSR, sheet Q-54, 55 (Khonuu), VSEGEI, Leningrad, scale
1:1,000,000 (dated 1985). (in Russian)

Tret’yakov, F.F., 2003. Darpir and Ulakhan Faults: Contemporary
Interpretation: Otechestrennaya Geologiya, 2003 (6), p. 78-80. (in
Russian)

Vashchilov, Y. Y., 1963: Deep faults of the southern Yana-Kolyma fold zone
and the Okhotsk-Chaun volcanic belt and their role in the formation of
granitic intrusions and the formation of structures (according to
geophysical data), Sovetskaya Geologiya, 1963(4), p. 54-72. (in
Russian)

121

Vogt, P. R., Taylor, P. T., Kovacs, L. C., and Johnson, G. L., 1979: Detailed
aeromagnetic investigation of the Arctic basin, Journal of Geophysical
Research, 84, 1071-1089.

Yang, X., Bondar, I., and Romney, C., 2000: PlDC Ground Truth Event (GT)
Database (Revision 1): CMR Technical Report CMR-00/15, 14 pp.

Waldhauser and Ellsworth, 2000: A Double-Difference Earthquake Location
Algorithm: Method and Application to the Northern Hayward Fault,
California, Bulletin of the Seismological Society of America, 90, 1353-
1 368.

Wells, D. L., and Coppersmith, K. J., 1994: New empirical relationships
among magnitude, rupture length, rupture width, rupture area, and
surface displacement, Bulletin of the Seismological Society of America,
84, 974—1002.

Wilson, J. T., 1963: Hypothesis of Earth’s behaviour, Nature, 198, 925-929.

122