HomeMy WebLinkAboutAPA1255Alaska Power Authority
Susitna Hydroelectric Project
ARLIS
IK
\'i 2. '?
Alaska Resources
Library & Information Services
Ancl ·b .• "'ud.Ska
Subtasks 4.09 through 4.15
FINAL REPORT ON
SEISMIC STUDIES FOR
SUSITNA HYDROELECTRIC PROJECT
February 1982
Prepared by
Woodward·Ciyde Consultants
for
Acres American Incorporated
1000 Liberty Bank Building
Main at Court
Buffalo, New York 14202
Telephone: (716) 853-7525
PREFACE
This report presents the results of the seismic studies conducted
during 1980 and 1981 for the Feasibility Study of the proposed Susitna
Hydroelectric Project site. These studies included geologic evaluation
of faults and lineaments, a historical and microearthquake seismicity
study, an assessment of the potential for reservoir-induced seismicity,
and the estimation of ground-motion parameters.
The report includes ten sections \<Jhich summarize the results of the
studies to date. The three appendices present support data for the
interpretations and conclusions presented in Sections 1 through 10.
Tables and figures appear at the end of each section and appendix.
In most cases, measurements reported in this volume were made in the
metric system and then converted to U.S. Customary Units. For these
conversions, the measurements reported in the U.S. system are rounded
off to the nearest single unit (e. g., 70 km converts to 43 miles) even
when in the context of the sentence the conversion should be rounded off
to the nearest ten units (e. g., 70 km converted to 40 miles). This was
done to retain the original number used to make t~e conversion. Con-
versely, some measurements were made using the U.S. system; in these
cases, the conversion to the metric system also have been rounded off to
the nearest single unit. Both sets of numbers have been presented for
the convenience of the reader.
The results and conclusions presented in this report refine those
presented in the Interim Report (HoodvJard-Clyde Consultants, 1980b).
Woodward-Clyde Consultants
Figures 3-S(A and B) and 3-6(A) incorrectly depict the distribution of
ice disintegration deposits. On Figure 3-S(A), the area northwest of
the Black River shown to be ice disintegration deposits of Early Wiscon-
sin age is, in fact, an area of till. On Figure 3-5(8), ice disinte-
gration deposits of Late Wisconsin age should be shown only on the
valley floor; other areas indicated as ice disintegration deposits are
areas covered by till. On Figure 3-6(A), the area of ice disintegration
deposits of Early Wisconsin age east of Butte Lake should have been
shown as till.
The above corrections address errors in graphical presentation and do
not affect the conclusions of the report.
TABLE OF CONTENTS
PREFACE
TABLE OF CONTENTS
LIST OF TABLES
LIST OF FIGURES
DEFINITION OF KEY TERMS
ACKNOWLEDGMENTS
Page
1 -SUMMARY --------------------------------------------1-1
1.1 Project Description ---------------------------1-1
1.2 Fault Study Rationale-------------------------1-3
1.3 Approach --------------------------------------1-4
1.4 Tectonic Model --------------------------------1-6
1.5 Quaternary Geology-----------------------------1-7
1.6 Faults with Recent Displacement ---------------1-8
1.7 Seismicity ------------------------------------1-10
1.8 f~aximum Credible Earthquakes (MCEs) -----------1-13
1.9 Effect of Reservoir-Induced Seismicity (RIS) --1-15
1.10 Ground Motions --------------------------------1-17
1.11 Conclusions -----------------------------------1-19
2 -INTRODUCTION ---------------------------------------2-1
2.1 Project Description and Location --------------2-1
2.2 Objectives ------------------------------------2-2
2.3 Scope-----------------------------------------2-4
2.4 Fault Study Rationale -------------------------2-7
2.4.1 Conceptual Approach --------------------2-7
2.4.2 Guidelines for the Identification
of Faults with Recent Displacement ----2-10
2.5 Methodology -----------------------------------2-13
3 -QUATERNARY GEOLOGY ---------------------------------3-1
3.1 Introduction ----------------------------------3-1
3.2 Regional Pleistocene Geology Setting ----------3-4
3.3 Age and Extent of Quaternary Surfaces in
the Quaternary Study Region ------------------3-7
3.4 Key Quaternary Study Areas --------------------3-9
3.4.1 The Black River Area -------------------3-10
3.4.2 The Clear Valley Area------------------3-11
3.4.3 The Butte Lake Area--------------------3-13
3.4.4 The Deadman Creek Area -----------------3-14
3.5 Glacial History and Distribution of
Quaternary Surfaces --------------------------3-15
3.6 Quaternary Geology and Significant
Features -------------------------------------3-19
TABLE OF CONTENTS (CONTINUED)
Page
4 -SIGNIFICANT FEATURES -------------------------------4-1
4.1 Introduction ----------------------------------4-1
4.2 Detectability of Faults with
Recent Displacement ---------------------------4-2
4.2.1 Selected Worldwide Earthquakes and
Faults with Recent Displacement -------4-3
4.2.2 Occurrence of Surface Faulting
in California-------------------------4-6
4.2.3 Preservation of Recent Displacement ----4-7
4.2.4 Detection Level Earthquake-------------4-9
4.3 Talkeetna Terrain Boundary Faults -------------4-11
4.3.1 Castle Mountain Fault (AD5-1) ----------4-11
4.3.2 Denali Fault (HB4-1) -------------------4-14
4.3.3 Benioff Zone---------------------------4-17
4.4 Features Within the Talkeetna Terrain ---------4-19
4.4.1 Watana Site ----------------------------4-19
4.4.2 Devil Canyon Site----------------------4-39
4.5 Assessment of Recent Displacement -------------4-65
5 -SEISMICITY AND STRESS REGIME -----------------------5-1
5.1 Regional Seismicity---------------------------5-2
5.1.1 Tectonic Setting-----------------------5-2
5.1.2 Historical Seismicity Record -----------5-4
5.1.3 Analysis of Large Historical
Earthquakes ----------------------------5-6
5.2 Benioff Zone Seismicity-----------------------5-16
5.2.1 Benioff Zone Zonation ------------------5-16
5.2.2 Magnitude Estimates of
Benioff Zone Earthquakes ---------------5-20
5.2.3 Maximum Earthquakes from Significant
Benioff Zone Sources -------------------5-23
5.3 Local Microearthquake Activity----------------5-26
5.3.1 Network Operation ----------------------5-26
5.3.2 Recorded Earthquakes -------------------5-27
5.3.3 Talkeetna Terrain Stress Regime --------5-28
5.4 Recurrence ------------------------------------5-29
5.4.1 The Interplate Region ------------------5-30
5.4.2 The Intraplate Region ------------------5-31
5.4.3 The Talkeetna Terrain ------------------5-31
5.4.4 Summary--------------------------------5-33
Woodw~rdl·Ciyde Consultants
TABLE OF CONTENTS (CONTINUED)
Page
6 -RESERVOIR-INDUCED SEISMICITY (RIS) -----------------6-1
6.1 Evaluation of Potential Occurrence ------------6-3
6.1.1 Likelihood of Occurrence---------------6-3
6.1.2 Location and Maximum Magnitude---------6-4
6.2 Effect of RISon Earthquake Occurrence
Likelihood-----------------------------------6-5
6.2.1 Description of the Model ---------------6-6
6.2.2 Implementation of the Model for
the Susitna Project -------------------6-10
6.3 RIS and Method of Reservoir Filling -----------6-12
6.4 Potential for Landslides in the Devil Canyon-
Watana Reservoir Area Resulting from RIS -----6-13
7 -MAXIMUM CREDIBLE EARTHQUAKES (MCEs) ----------------7-1
7.1 Sources Outside the Talkeetna Terrain---------7-3
7.2 Talkeetna Terrain Boundary Sources ------------7-3
7.2.1 The Castle Mountain Fault--------------7-3
7 .2.2 The Denali Fault -----------------------7-4
7.2.3 The Benioff Zone-----------------------7-5
7.3 Talkeetna Terrain Sources ---------------------7-7
7.4 Effect of Reservoir-Induced Seismicity--------7-8
8 -GROUND MOTIONS -------------------------------------8-1
8.1 Introduction ----------------------------------8-1
8.2 Seismicity Environment ------------------------8-2
8.2.1 Potential Sources of Earthquakes -------8-2
8.2.2 t~aximum Credible Earthquakes (MCEs) ----8-3
8.2.3 Earthquake Recurrence------------------8-3
8.3 Deterministic Estimates of Earthquake
Ground Motions -------------------------------8-6
8.3.1 Attenuation of Earthquake
Ground Motion -------------------------8-6
8.3.2 Estimates of Peak Ground
Acceleration and Response
Spectra at the Dam Sites --------------8-8
8.3.3 Estimates of the Duration of Strong
Ground Shaking at the Dam Sites -------8-9
8.4 Assessment of Seismic Exposure ----------------8-9
8.4.1 Methodology ----------------------------8-9
8.4.2 Assessment of Inputs for Analysis ------8-10
8.4.3 Results --------------------------------8-12
8.5 Use of Results of Ground Motion Studies in
Selecting Design Ground-Motion Criteria------8-13
TABLE OF CONTENTS (CONTINUED)
9 -TRANSMISSION LINE AND ACCESS ROUTE
SUSCEPTIBILITY TO SEISMICALLY INDUCED FAILURE -----9-1
9.1 Introduction ----------------------------------9-1
9.2 Areas of Potential Susceptibility-------------9-3
9.2.1 Alternate Route 1 ----------------------9-3
9.2.2 Alternate Route 2 ----------------------9-4
9.2.3 Alternate Route 3 --------------------~-9-6
9.3 Summary---------------------------------------9-7
10 -CONCLUSIONS ----------------------------------------10-1
10.1 Feasibility Conclusions----------------------10-1
10.2 Technical Conclusions ------------------------10-2
APPENDIX A -METHODS OF STUDY ---------------------------A-1
A.1 Quaternary Geology ----------------------------A-1
A.1.1 Scope of Studies-----------------------A-1
A.1.2 Methods--------------------------------A-2
A.1.3 Age Dating-----------------------------A-5
A.l.4 Interpretation of Six Subregions
of the Quaternary Study Region --------A-9
A.2 Radiometric Age Dating ------------------------A-25
A.2.1 Radiocarbon Age Dating -----------------A-25
A.2.2 Potassium-Argon Age Dating-------------A-25
A.3 Field Mapping ---------------------------------A-28
A.4 Trench Logging Methods ------------------------A-29
A.5 Geophysical Studies ---------------------------A-32
A.5.1 Field Operations -----------------------A-32
A.5 .2 Data Reduction and
Interpretive Procedures ---------------A-34
A.6 Low-Sun-Angle Aerial Photography--------------A-35
A.7 Earthquake Recurrence Calculations ------------A-36
APPENDIX B -GLOSSARY
APPENDIX C -REFERENCES CITED
LIST OF TABLES
2-1 Project Subtasks and Objectives
3-1 Summary of Quaternary Glaciogenic Feature Characteristics
3-2 Radiocarbon Age Dates and Sample Descriptions
3-3 Summary of Relative Age Data for Early and
Late Wisconsin Moraines
4-1 Summary of Selected Worldwide Earthquakes
4-2 Summary of Surface Faulting in California
4-3 Summary of Boundary Faults and Significant Features
5-1 Parameters of Selected Historical Earthquakes
5-2 Microearthquakes Analyzed in the Stress Regime Study
5-3 Summary of Selected Benioff Zone Earthquakes
6-1 Reported Cases of Reservoir-Induced Seismicity (RIS)
6-2 Reservoir-Induced Seismic Events with Maximum Magnitude
of 5 or Greater
7-1 Maximum Credible Earthquake (MCE) Summary and
Seismic Source Data
8-1 Summary of Earthquake Sources Considered in
Ground-Motion Studies
8-2 Summary of Earthquake Recurrence Assessments
A-1 Relative Age Data in the Talkeetna Mountains and
A 1 ask a Range
A-2 Summary of Potassium-Argon Whole Rock Age Dates
LIST OF FIGURES
1-1 Summary of Seismic Sources
2-1 Project Location Map
2-2 Project Flow Diagram
3-1 Quaternary Time Scale
3-2 Quaternary Geology Map
3-3 Generalized Cross-Section of Quaternary Deposits and Surf aces
3-4 Significant Features and Quaternary Surf ace Map
3-5 Quaternary Geology of the Black River and Clear Valley Areas
3-6 Quaternary Geology of the Butte Lake and Deadman Creek Areas
4-1 Talkeetna Terrain Model and Section
4-2 Geologic Time Scale
4-3 Regional Tectonic Terrane Map
4-4 Talkeetna Thrust Fault and Susitna Feature Location Map
4-5 Diagrammatic Cross-Section of the Broxson Gulch Thrust Fault
at Windy Creek (Location W1)
4-6 Diagrammatic Cross-Section of the Talkeetna Thrust Fault
near Butte Creek (Location W2)
4-7 Geologic Map of Locations W8, W9, and WlO near Talkeetna Hill
4-8 Diagrammatic Cross-Section of the Talkeetna Thrust Fault at
Talkeetna Hill (Location W9)
4-9 Diagrammatic Cross-Section of the Talkeetna Thrust Fault at
Trench T-2 (Location W10)
4-10 Orientation of Folded Strata in Watana Creek (Location W3)
4-11 Location of Trench T-1 at Location W7
4-12 Trenches T-1 and T-2, Trench Logs
LIST OF FIGURES (CONTINUED)
4-13 Photographs of Trench T-1
4-14 Photograph of Trench T-2
4-15 Geologic Map of Location W11 near Butte Lake
4-16 Butte Lake Magnetic Profiles
4-17 Location of Trench S-1 at Location W12
4-18 Trench S-1, Trench Log
4-19 Photographs of Trench S-1
4-20 Watana Lineament (KD3-7) Location Map
4-21 Fins Feature (KD4-27) Location Map
4-22 Devil Canyon Features Location Map
5-1 Historical Earthquakes of Focal Depth Less than
30 km in the Site Region from 1904 through 1978
5-2 Historical Earthquakes of Focal Depth Greater than
35 km in the Site Region from 1904 through 1978
5-3 Focal Mechanisms for 29 June 1964 and 1 January 1975 Earthquakes
5-4 Observed and Synthetic Seismograms for P and SH Waves at
PAS for the 3 November 1943 Earthquake
5-5 Observed Polarities and Theoretical Radiation
Pattern for the 3 November 1943 Earthquake
5-6 1943 Earthquake Geology Map
5-7 Tectonic Interpretation and Cross-Section of Crustal and
Benioff Zone Microearthquakes Located Within the 1980 Network
5-8 Focal Mechanisms for 30 August 1980 and 31 August 1980 Earthquakes
5-9 Frequency-Magnitude Relationships
6-1 Plot of Water Depth and Volume for Worldwide Reservoirs and
Reported Cases of RIS
Woodward·CDyde Consultants
LIST OF FIGURES (CONTINUED)
8-1 Mean Attenuation Relationships for Deep Focus
(Benioff Zone) Earthquakes
8-2 Mean Attenuation Relationships for Shallow Focus Earthquakes
8-3 Comparison of Selected Attentuation Relationships With
Data From Alaska
8-4 Mean Response Spectra for Maximum Credible Earthquakes on the
Benioff Zone and Denali Fault -l~atana Site
8-5 Mean Response Spectra for Maximum Credible Earthquakes on the
Benioff Zone and Denali Fault-Devil Canyon Site
8-6 Mean Response Spectrum for Maximum Credible Detection
Level Earthquake
8-7 Schematic Representation of Seismic Exposure Analysis Approach
8-8 Relationships Between Magnitude and Fault Rupture Dimensions
8-9 Probability of Exceedance Versus Peak Ground Acceleration
at the Watana Site
8-10 Mean and 84th Percentile Response Spectra for a Maximum
Credible Earthquake on the Benioff Zone -Watana Site
8-11 Acceleration Time History for a Maximum Earthquake
on the Benioff Zone
8-12 Mean and 84th Percentile Response Spectra for a Maximum
Credible Earthquake on the Benioff Zone-Devil Canyon Site
9-1 Transmission Line and Access Routes Potential Hazards Map
A-1 Quaternary Geology Location Map
DEFINITION OF KEY TERMS
Site Region:
Project Area:
Devil Canyon Area:
Devil Canyon Site:
Devil Canyon Reservoir:
Watana Area:
Watana Site:
Watana Reservoir:
Woodward· Clyde Consultants
The area within a 62-m il e ( 100-km)
radius about either site.
This generally includes the Devil
Canyon and Watana areas and the
region in between.
The area within a 6-mile (10-km)
radius about the Devil Canyon site.
The presently proposed location of
the Devil Canyon Dam and related
f ac i l it i e s.
The area of the Susitna River
upstream from the proposed Devil
Canyon site which will be inundated
by impoundment by the dam.
The area within a 6-mile ( 10-km)
radius about the Watana site.
The presently proposed location
of the Watana Dam and related
facilities.
The area of the Susitna River up-
stream from the proposed Watana
site which will be inundated by
impoundment by the dam.
DEFINITION OF KEY TERMS (CONTINUED)
Microearthquake Study Area: The area in which microearthquake
monitoring was conducted in 1980.
The boundaries are 62.3° to 63.3°
north latitude and 147.5° to 150.4°
west longitude.
ACKNOWLEDGMENTS
The l~oodward-Clyde Consultants 1 personnel who participated in the
geological portion of the investigation during the two-year study
included: Phillip Birkhahn, George Brogan, Dan Collins, Tom Freeman,
Bob Goodwin, Paul Guptill, Jon Lovegreen, Ron Mees, Hector Reyes, Ed
Sabins, Ray Sugiura, John Waggoner, Dennis Welsch, and Jerry Williams.
Dr. Norm Ten Brink of Grand Valley State College, Michigan, assisted
with the Quaternary studies. Dr. Kerry Sieh of the California Institute
of Techno logy assisted with trench logging and Quaternary studies.
The Woodward-Clyde Consultants 1 personnel who participated in the
seismological portion of the seismic geology investigation during
the two-year stydy were: Jim Agnew, Jean Briggs, Jim Cullen, Don
Helmberger, John Hobgood, Barbara Leitner, Jose Rial, Bette Shepard,
William Savage, and Paul Somerville. Jason McBride and Richard
Thompson, both of Woodward-Clyde Consultants, and Milton Mayr, Dianne
Marshall, Cole Jonafrauk, and Rodney Viereck, all of the University of
Alaska Geophysical Institute, assisted with operation of the micro-
earthquake network in 1980. Jose Rial analyzed and interpreted syn-
thetic seismograms and first-motion parameters to estimate focal-
mechanism parameters. These results were reviewed by Don He lmberger.
Reservoir-induced seismicity studies were conducted by Thalia Anagnos,
Ram Kulkarni, Peter Knuepfer, and Duane Packer. The earthquake
engineering portion of the investigation was conducted by Ezio Alviti,
Barbara Bogaert, John Egan, Maurice Power, and Jim Rafidi.
Discussions were held with members of the Alaska Geological Survey,
the University of Alaska Geophysical Institute, the Lamont-Doherty
Geological Observatory, the University of Alaska, and the U.S. Geologi-
cal Survey.
Dr. Ulrich Luscher was principal-in-charge of the investigation in
1980. In 1981, George Brogan assumed Dr. Luscher 1 s position. Jon
Lovegreen was the project manager and directed the geology study along
Woodward-Clyde Consultants
-2-
with Phillip Birkhahn. ur. William Savage directed the seismology
study; Dr. Packer directed the reservoir-induced seismicity study; and
Maurice Power directed the earthquake engineering study.
Project peer review was provided by: Dr. Robert Forbes, Professor
Emeritus of Geology of the University of Alaska (geology); and Dr. I. M.
Idriss (earthquake engineering), Dr. George Linkletter (Quaternary
geology), Dr. Duane Packer (geology), Dr. William Savage (reservoir-
induced seismicity), and Dr. Tom Turcotte (seismology), all of Woodward-
Clyde Consultants.
Tha Alaska Power Authority authorized the project and provided funding
for this investigation. Among those involved were Eric Yould (Direc-
tor), Nancy Blunck, Robert Mohn, and Dave Wozniak.
Acres American Inc. (Acres) provided logistical support during the
field work and participated in discussions which ultimately led to the
technical conclusions presented in this report. Among those directly
involved were John Lawrence (Project Manager), Mike Bruen, Lance Duncan,
Jim Gill, Robert Henschel, Don MacDonald, Virendra Singh, Stuart
Thompson, and Leib Wolofsky.
Radiometric (K-Ar) age dates along with thin sections and analyses were
provided by Dr. Ken Foland and his colleagues at Ohio State University.
Krueger Enterprises, Inc., of Massachusetts provided radiocarbon (c14)
age dates. Air Photo Tech, Inc., of Anchorage, Alaska, provided low-
sun-angle color near-infrared aerial photography.
Aircraft support was provided by Akland Helicopter, Inc., Air Logistics,
Inc., ERA Helicopters, Inc., and Kenai Air, Inc., and was arranged by
Acres American Inc. Lodging and logistics in the field were also
arranged by Acres.
-3-
This report and the Interim Report were prepared by Woodward-Clyde
Consultants under subcontract to Acres. Technical support and typing
were provided by Jill Pelayo and Carole ~~ilson, assisted by Barbara
Belton, Esther Martin, and Hazel Boyd. Illustrations were prepared by
Arlene Padamada, assisted by Dennis Fischer, Al Herron, Robyn Sherrill,
Chairoa.ch Siripatanapaibul, and Mark Winters. Technical editing was
done by Karen L undegaard. Printing and binding of the Interim Report
vJas done by Continental Graphics, of Los Angeles. This report was
printed and bound by Woodward-Clyde Consultants and Hendricks Printing
Company, Inc., of Irvine, California, under the direction of Marcel
Arboleda, assisted by Belinda Spicer.
Woodward-Clyde Consultants
1 -SUMMARY
1.1 -Project Description
The Susitna Hydroelectic Project as currently proposed involves two
dams and reservoirs on the Susitna River in the Talkeetna Mountains of
south central Alaska. The Project is approximately 50 miles (80 km)
northeast of Talkeetna, Alaska, and 118 miles (190 km) north-northeast
of Anchorage, Alaska (Figures 1-1 and 2-1). The downstream dam at Devil
Canyon (62.8° north latitude, 149.3° west longitude) is currently being
considered as an arch dam to be approximately 645 feet (197m) high. It
would impound a 26-mile-(42-km-) long reservoir with a capacity of
approximately 1,092,000 acre-feet (1,348 x 106m3). The upstream
dam at Watana (62.8° north latitude, 148.6° west longitude) is currently
being considered as an earthfill or rockfill dam to be approximately 885
feet (270m) high. It would impound a 48-mile-(77-km-) long reservoir
with a capacity of approximately 9,515,000 acre-feet (11,741 x 106m3).
These dimensions are approximate and subject to revision during design
of the project.
Collectively, the proposed dams and related struc.tures are referred
to as the Project. This report is part of a feasibility study for the
Project being managed and conducted by Acres American Inc. for the
Alaska Power Authority. The purpose of this report is to summarize the
results of the seismic geology, seismology, and earthquake ground-motion
investigation conducted during 1980 and 1981.
The primary objectives of this two-year investigation were to pro vi de
values of earthquake ground-motion parameters that would be used for dam
design and assessment of feasibility of the Project and to identify
faults that have the potential for surface rupture through the area of
the Project.
1 - 1
Consu~tants
The 1980 study included: review of available geologic and seismologic
literature and data; monitoring of mi croearthquake activity for three
months within approximately 30 miles (48 km) of both proposed dam
sites using a 10-station microearthquake network; a preliminary review
of the potential for reservoir-induced seismicity; interpretation
of existing remotely sensed data; a 10 person-month geologic field
reconnaissance of mapped faults and lineaments within 62 miles (100 km)
of the Project; analysis and interpretation of these data for selection
of 13 features (faults and lineaments) for detailed study in 1981; and
an estimate of potential earthquake ground motions for the Project. The
results of the 1980 study were reported in an Interim Report (Woodward-
Clyde Consultants, 1980b).
The 1981 study included: the acquisition and interpretation of low-
sun-angle co lor near-infrared aerial photography; a 10 person-month
field program of geologic field mapping, aerial reconnaissance, and
trench excavation and logging of 13 features; a three person-month field
program of Quaternary geologic studies; geophysical surveys; potassium-
argon and radiocarbon age dating; analysis of seismograms for moderate
to large earthquakes that have occurred within or adjacent to the
Talkeetna Terrain; review of the Benioff zone seismicity and refinement
of the location and size of the maximum credible earthquake for the
zone; analysis of focal mechanisms of selected earthquakes to refine
understanding of the regional stress regime; and development of a model
for reservoir-induced seismicity that incorporates the relationship of
natural seismicity with reservoir-induced earthquakes. From this work,
seismic sources were identified, and the maximum credible earthquake for
each source and its recurrence interval were estimated. These data were
used to estimate the potential ground motions for the Project. A manual
was also prepared for the operation of a long-term seismic monitoring
network.
1 - 2
Woodward· Clyde Consultants
The results presented in this report were developed for two reasons:
1) for the purpose of evaluating Project feasibility; and 2) to support
submittal of a license application for the Project to the Federal
Energy Regulatory Commission. The results should be reviewed when final
dam design is considered.
This summary abstracts many important details that should be considered
in any application of the results to seismic design. Consequently, the
concepts, interpretations, and conclusions presented in this summary
should be used only within the context of corresponding sections in the
text.
1.2-Fault Study Rationale
According to the present understanding of plate tectonics, the earth's
lithosphere contains 12 to 22 miles (20 to 35 km) or so of relatively
light, brittle crust that overlies the mantle, which is denser and less
brittle than the crust. Major horizontal movements of the crustal
plates are considered to be related to, or caused by, thermal convective
processes within the mantle.
Within the plate-tectonic framework, faults that have the potential for
generating earthquakes have had recent displacement and may be subject
to repeated displacements as long as they are in the same tectonic
stress regime. In regions of plate collision such as Alaska, the
tectonic stress regime is the result of one plate being subducted, or
underthrust, beneath the adjacent plate.
Faults with recent displacement-both in the downgoing plate and in the
upper plate can generate earthquakes, which cause ground motions at the
surface that need to be considered for seismic design purposes. However,
1 -3
Woodward-Clyde Consultants
faults in the downgoing plate are not considered when evaluating poten-
tial for ground-surface rupture, because these faults do not extend into
the upper plate.
A guideline for defining 11 recent displacement 11 was prepared by Acres
American Inc. and is discussed in Section 3 of the Interim Report
(Woodward-Clyde Consultants, 198Gb). According to that guideline, faults
that have caused rupture of the ground surface within approximately the
past lGG,GGG years are classified as being faults with recent displace-
ment and should be considered in seismic design. Conversely, faults
that have not caused rupture of the ground surface during the past
lGG,GGG years are classified as faults without recent displacement.
These faults are considered to be of no additional importance to Project
feasibility and dam design because faults without recent displacement
are not known to be sources of large earthquakes or surface rupture.
To be identified as a fault with recent displacement, earthquakes that
occurred on the fault during the past lGG,GGG years need to have been
large enough to produce geological evidence of surface rupture that
could be detected by our investigation. Consequently, an estimate was
made of the magnitude of the largest earthquake that might have occurred
without leaving any detectable geologic evidence. This earthquake was
designated the "detection level earthquake 11 and is considered to be a
potential seismic source.
1.3 -Approach
The 198G study led to the i dent i fi cation of features considered to be
potentially important to seismic design. The rationale and methods for
identifying these features is discussed in Section 8 of the Interim
Report (Woodward-Clyde Consultants, 198Gb).
1 -4
Woodwarrd·CDyde ConsuUiants
The purpose of the 1981 study was to evaluate the feasibility and
design significance of three potentially important earthquake sources
with known recent displacement; these included the Castle Mountain and
Denali faults and the Benioff zone. The 1980 study also identified
13 features (four faults and nine lineaments) in the vicinity of
the proposed dam sites that required additional study to evaluate
whether they had been subject to surf ace displacement during the past
100,000 years and, therefore, might need to be considered as seismic
sources for purposes of dam design. These 13 features were examined in
detail during the 1981 field study. The study involved the following
objectives:
a) Assessing the likelihood that each of the 13 features is a fault.
b) Assessing the age of the sediments overlying each of the 13 fea-
tures.
c) Selecting and excavating trenches across topographic features that
resembled topographic expression of faults in the young geologic
deposits.
d) Evaluating the likelihood that each of the 13 features is a fault
with recent displacement using the guideline established for the
project, i.e., rupture of the ground surf ace during the past
100,000 years.
e) Assessing the detectabi lity of faults that may have ruptured the
ground surface during moderate to large earthquakes in the past
100,000 years and estimating a detection level earthquake that
could theoretically occur on a fault that might be below the
detection level of geologic investigation.
1 - 5
Woodwa~rd·CByde Consultants
f) Evaluating seismological records of moderate to large historical
earthquakes in the project region to estimate focal mechanism
parameters and assess the relation of the earthquakes to recognized
faults with recent displacement.
g) Applying judgment and experience gained from the study of other
faults with recent displacement in Alaska and in similar tectonic
environments (e.g., Japan and South America).
h) Estimating the maximum credible earthquake and recurrence interval
1) for each fault that is considered to be a seismic source, 2) for
the Benioff zone, and 3) for a detection level earthquake.
i) Estimating the potential for surface rupture on any faults with
recent displacement within 6 miles (10 km) of the dam sites.
j) Estimating the values of ground-motion parameters for the seismic
sources i dent ifi ed in objective (h) above that are appropriate for
seismic design.
1.4 -Tectonic Model
A tectonic model for the area encompassing the Project region was
developed in 1980 to provide a framework in which to: assess fault
activity; estimate maximum credible earthquakes; evaluate the potential
for surface fault rupture; and evaluate the potential for reservoir-
induced seismicity.
On the basis of the tectonic model, a relatively stable tectonic unit
was identified in which the Project is located. This tectonic unit was
named the Talkeetna Terrain (Woodward-Clyde Consultants, 1980b). The
Terrain boundaries are the Denali and Totschunda faults to the north and
1 - 6
Woodward-Clyde Consultants
east, the Castle Mountain fault to the south, and a zone of deformation
with volcanoes to the west. The thickness of the Talkeetna Terrain is
limited by the Benioff zone or base of the crust at depth (Figure 1-1).
With the exception of the western boundary, which is primarily a broad
zone of uplift marked by Cenozoic age volcanoes, all of the boundaries
are (or contain) recognized faults with recent displacement. The
Terrain is part of the North American plate.
Because the Talkeetna Terrain is a relatively stable tectonic unit,
major strain release occurs along its boundaries rather than within the
Terrain. The basis for this conclusion is: the clear evidence for
recent displacement along the Castle Mountain, Denali, and Totschunda
faults and the subducting plate defined by the Benioff zone; the general
absence of large historical earthquakes within the Terrain; and the
absence of faults within the Terrain that display evidence of recent
displacement. Some compression-related crustal adjustment within the
Terrain is probably occurring as a result of the plate movement and the
stresses related to the subduction process. This crustal adjustment is
expressed by small earthquakes such as those recorded during the 1980
microearthquake study.
1.5 -Quaternary Geology
Surfaces and sediments of late Quaternary v1ere studied in detail to
provide information on the recency of fault displacement. These studies
were designed to: a) prepare a map showing the geographic extent and
age of surficial glacial sediments and surfaces; and b) evaluate the
likelihood that each of the 13 significant features is a fault with
recent displacement. In the site region, the Quaternary surfaces
and sediments are primarily glacial in origin. This origin reflects
the wide-spread glacial activity in south central Alaska during late
Quaternary time.
1 -7
Woodward-Clyde Consultants
Four distinct Quaternary glacial episodes are evident: pre-Wiscon-
sin, >100,000 years before present (y.b.p.); Early Wisconsin, 75,000 to
40,000 y.b.p.; Late Wisconsin, 25,000 to 9,000 y.b.p.; and Holocene,
<9,000 y.b.p. (Figure 3-2). Each glacial episode was less extensive
than the preceding one. The limits of each are defined by the elevation
and geographic distribution of glacial erosional or depositional
features. Glaciers advanced repeatedly from three main source areas:
the Alaska Range to the north; the southern and southeastern Talkeetna
Mountains; and the Talkeetna Mountains north and northwest of the
Susitna River. Glacial flow was dominantly south and southwest,
following the regional slope and structural grain. Multi-directional
and convergent flow, differing glacial magnitudes, topographic influ-
ences, and other parameters make the glacial chronology of the Project
region complex.
The four features near the Watana site are located in areas where Late
Wisconsin (25,000 to 9,000 y.b.p.) surfaces predominate except in the
vicinity of the Talkeetna River where pre-Wisconsin (>100,000 y.b.p.)
surf aces are present (Figure 3-4). The nine features near the Devil
Canyon site are in areas of Early Wisconsin and older (>40,000 y.b.p.)
erosional surfaces, except in the vicinity of the Susitna River and
major creeks where Late Wisconsin (25,000 to 9,000 y.b.p.) surfaces
predominate (Figure 3-4).
1.6 -Faults with Recent Displacement
Faults for which evidence of recent displacement was found were con-
sidered to be potential seismic sources. Each potential seismic source
was evaluated during this study to estimate its potential seismic ground
motions at the Watana and Devil Canyon sites and its potential for
surface rupture within 6 miles (10 km) of the sites.
1 -8
Consultants
On the basis of the 1980 study, the Talkeetna Terrain boundary faults
were identified as seismic sources that need to be considered as
potential sources of seismic ground motion at the sites. These include:
the Castle Mountain fault, the Denali fault, the Benioff zone interplate
region, and the Benioff zone i nt rap late region (Figure 1-1). These
sources are considered to be or to contain faults with recent displace-
ment that could cause seismic ground motions at the Watana and Devil
Canyon sites; however, because of their distance from the sites, these
faults do not have the potential for rupture through the sites. The
1980 study also identified 13 features near the sites that required
detailed evaluation during the 1981 study to assess their importance for
seimsic design.
On the bas is of the 1981 study, no evidence for faults with recent
displacement other than the Talkeetna Terrain boundary faults has
been observed within 62 miles (100 km) of either site and none of
the 13 features near the sites are judged to be faults with recent
displacement. Therefore, when applying the guideline defining faults
with recent displacement to the results of our investigation, the 13
features are considered not to be potential seismic sources that could
cause seismic ground motions at the sites or surface rupture through the
sites.
Our interpretation that none of the 13 features are faults with recent
displacement is based on data collected during our investigation.
The data are limited in the sense that a cant i nuous 100,000 year-a l d
stratum or surface was not found along the entire length of each of
the features. For this reason, the available data were analyzed and
professional judgment was applied to reach conclusions concerning the
recency of displacement on each of the 13 features.
As discussed previously, earthquakes up to a given magnitude could occur
on faults with recent displacement that might not be detectable by our
1 - 9
geologic investigation. The size of such an earthquake, designated the
detection level earthquake, varies according to the degree of natural
preservation of fault-related geomorphic features and from one tectonic
en vi ron men t to another . The detect i on 1 eve 1 earth q u ak e h as been
estimated by: 1) evaluating the dimensions of surface faulting asso-
ciated with worldwide historical earthquakes in tectonic environments
similar to the Talkeetna Terrain; 2) identifying the threshold of
surface faulting using a group of thoroughly studied earthquakes in
California; and 3) evaluating the degree of preservation of fault-
related geomorphic features in the Talkeetna Terrain. For this project,
we have judged that the detection level earthquake is magnitude (Ms) 6.
1.7 -Seismicity
Historical earthquake activity within 200 miles (322 km) of the Project
is associated with displacement along crustal faults in the upper plate
and with the subducting (downgoi ng) plate. The largest earthquake
within 200 miles (322 km) of the Project is the 1964 Prince William
Sound earthquake of magnitude (Ms) 8.4. This earthquake occurred
outside the Talkeetna Terrain on the interface between the North
American plate and the Pacific plate (Figure 1-1); the associated
rupture and deformation extended to within appro xi mate ly 88 miles (142
km) of the Project.
Within the site region (62 miles [100 km] from the Project), the leve 1
of historical seismicity on the Benioff zone is at least several
times greater than that of the crustal region. The largest reported
earthquake in the site region (magnitude [MsJ 6-1/4) occurred on
3 July 1929. The focal depth appears to be below the crust, possibly in
the depth range of 25 to 31 miles ( 40 to 50 km). This depth suggests
that this earthquake may have occurred on the Benioff zone.
1 -10
Woodward-Clyde ConsLdtants
During three months of mi d-1980, a ten-station mi croearthquake array was
operated to study the area within 30 miles (48 km) of the Project. More
than 260 earthquakes in the magnitude (ML) range 0.0 to 3.7 were
analyzed.
Earthquake activity recorded by the microearthquake array clearly
delineates two seismic zones. The upper zone of crustal activity occurs
predominantly in the depth range of 5 to 12 miles (8 to 20 km). The
lower zone of activity defines a northwestward dipping zone (the Benioff
zone) at a depth of 25 miles (40 km) in the southeast to 50 miles
(80 km) in the northwest portion of the microearthquake study area
(Woodward-Clyde Consultants, 1980b). The Benioff zone is approximately
6 to 9 miles (10 to 15 km) thick and is characterized by widely distri-
buted seismicity.
During the three-month period of monitoring, 13 earthquakes of magnitude
(ML) 3.0 and larger were located in the Benioff zone. This level of
activity is about ten times greater than that recorded for the shallow
(crustal) zone. The slope of the magnitude-frequency relationship for
the Benioff zone microearthquakes is 0.68, similar to that for many
areas worldwide. The magnitude-frequency relationship suggests a
relatively low number of larger earthquakes compared to smaller earth-
quakes. These results are consistent with the historical seismicity
record.
The crustal earthquake activity was found to contain relatively few
events at depths shallower than 5 miles (8 km) or deeper than 12 miles
(20 km). No seismic activity that appeared to be associated with
the crust was deeper than 19 miles (30 km). The level of seismicity
within the crustal zone within 30 miles (48 km) of the Project is very
low, about one-tenth of the Benioff zone activity. The slope of the
associated magnitude-frequency relationship is 1.48.
1 -11
Woodward· Clyde Consul~anis
The 1981 seismicity study included: an evaluation of the tectonic
associ at ion of moderate to large earthquakes within or adjacent to the
Talkeetna Terrain; review of small earthquakes within the Talkeetna
Terrain to assess the nature of the stress regime; review of Benioff
zone seismicity to refine the assessment of the magnitude and location
of the largest earthquake that could be expected to occur on the Benioff
zone; and development of a manual for a long-term seismic monitoring
network. The first three topics are discussed in Section 5 and are
summarized below. The network manual has been prepared as a separate
document from this report.
The evaluation of moderate to large historical earthquakes shows that
all of these events larger than magnitude (Ms) 5.6 in the Talkeetna
Terrain occurred in the Benioff zone, adjacent to recognized faults with
recent displacement (such as the Castle Mountain fault), or in the crust
adjacent to the western boundary of the Talkeetna Terrain. The event
near the western boundary of the Terrain is the 1943 earthquake of
magnitude (Ms) 7.3, which had a focal depth of 11 miles (17 km) and
was located approximately 90 miles (145 km) southwest of the Project. A
review of available small-scale satellite imagery and aerial photography
showed several lineaments that could have been sources for this earth-
quake. However, no obvious source was observed during an aerial recon-
naissance of the epicentral region.
Evaluation of the regional stress regime, using records obtained from
the mi croearthquake network operated during the 1980 study and records
from the University of Alaska Geophysical Institute (UAGI), supports
a northwest-direction of compression. This orientation is consistent
with regional plate tectonic motion and with the tectonic setting of the
Talkeetna Terrain.
Review of worldwide and Alaskan Benioff zone seismicity resulted in
a refined configuration of the Benioff zone. The Benioff zone in
1 -12
Woodward-Clyde Consultants
south central Alaska is comprised of two regions. In the interplate
region, earthquakes occur along the interface between the subducting
Pacific plate and the overlying North American plate (Figure 1-1).
Relatively large earthquakes, such as the 1964 magnitude (Ms) 8.4
Prince William Sound earthquake, occur along this region. In the
intraplate region, earthquakes occur within the subducting Pacific plate
where it is decoupled from and dips beneath the North American plate.
The maximum earthquakes in this region of the Benioff zone are of
moderate to large size and are smaller than the maximum earthquakes in
the interplate region.
1.8-Maximum Credible Earthquakes (MCEs)
Maximum credible earthquakes (MCEs) were estimated for the boundary
faults (in the crust and in the Benioff zone) and for the detection
level earthquake (discussed in Section 1.6). The MCEs for the crustal
faults (the Castle Mountain and Denali faults) were estimated using the
magnitude-rupture-length relationships developed by Slemmons (1977b) and
relationships based on Slerrrnons' data base (U.S. Nuclear Regulatory
Commission, 1981). The rupture area relationship of Wyss (1979) was
also considered. Application of these relationships represents the
state-of-the-practice in the estimation of maximum credible earthquakes.
(Appendix E in the Interim Report [Woodward-Clyde Consultants, 198Gb]
describes the details of these relationships.) These relationships
provide essentially the same results for the crustal boundary faults.
The estimated magnitude of the MCE for the Denali fault was revised
from the estimate that was presented in the Interim Report (Woodward-
Clyde Consultants, 1980b). The revision resulted from the evaluation
of recent data, such as that of Slemmons (U.S. Nuclear Regulatory
Commission, 1981).
1 -13
The refined configuration of the Benioff zone (discussed in Sections 1.7
and 5.2) distinguishes between the shallow interplate region (where
earthquakes accommodate relative plate motion) and a deeper intraplate
region (where earthquakes accommodate internal deformation within
the Pacific plate) (Figures 1-1 and 5-7). Reviev1 of the historical
seismicity in analogous interplate and intraplate regions of the world,
including Japan and South America, was used to estimate the MCE for both
regions of the Benioff zone.
Sources of moderate earthquakes appear to exist within the Talkeetna
Terrain, although no faults with recent displacement were detected
by our investigation. Therefore, an MCE was estimated for the de-
tection level earthquake that would be associated with a fault along
which no surface rupture was observed. In summary, the MCEs for the
crustal and Benioff zone seismic sources are estimated as follows:
Closest Approach to
Proposed Dam Sites
MCE 0 evi 1 Canyon Wat ana
Source (Ms) miles/(km) miles/(km)
Castle Mountain fault 7-1/2 71 ( 115) 65 (105)
Denali fault 8 40 ( 64) 43 (70)
Benioff zone ( i nt erp 1 ate ) 8-1/2 57 ( 91) 40 (64)
Benioff zone (intraplate) 7-1/2 38 ( 61) 31 (50)
Detect on level earthquake 6 <6 (<10) <6 (<10)
Note:
1. This ~~CE accommodates the 1964 Pri nee Wi 11 i am Sound earthquake of
magnitude (Ms) 8.4. As discussed by Kanamori (1977), the M5
magnitude scale appears to saturate for this size event, and the
M~v scale more accurately describes the size of the energy release.
The magnitude of the 1964 earthquake is 9.2; this Mw magnitude
is represented in the table and in the report by Ms 8-1/2 for
comparison to the other MCEs.
1 -14
1.9 -Effect of Reservoir-Induced Seismicity (RIS)
The reservoir that would be impounded behind the proposed Devil Canyon
dam would be 551 feet (168m) deep, and that behind the Watana dam would
be 725 feet (221 m) deep. The volume of the Devil Canyon reservoir
would be 1.09 x 106 acre-feet (1,348 x 106m3). The volume of the
Watana reservoir would be 9.56 x 106 acre-feet (11,741 x 106m3). Using
classifications cited by Packer and others (1977), both reservoirs would
be very deep; the Devil Canyon reservoir would be large, and the Watana
reservoir wou 1 d be very 1 arge. Because of the proximity of the two
reservoirs to each other, they would constitute a combined hydrologic
unit that would be very deep and very large.
Given that the combined hydrologic unit described above will be very
deep and very large, the potential for reservoir-induced seismicity
(RIS) was estimated by evaluating reservoir-induced seismicity at
other deep, very deep, and very large reservoirs. Our interpretation of
the results of this comparison indicates that the expected likelihood of
a reservoir-induced earthquake (including microearthquakes) at the
proposed reservoir is 0.46 (on a scale of 0 to 1).
Since the likelihood of a reservoir-induced event is high, it is
important to understand what the maximum reservoir-induced earthquake
is likely to be. Previous studies (Packer and others, 1977; Packer
and others, 1979) have presented data that support the concept that
reservoirs can trigger earthquakes by means of increases in pore
pressure or by incremental increases in stress. Because reservoirs act
as triggering mechanisms, they are not expected to cause an earthquake
larger than that which could occur in a given region "naturally."
Rather, the reservoirs are expected to have a potential effect on the
length of time between events and possibly on the location of the
event. Thus, if the tectonic and seismologic setting of a region is
known, and if the maximum earthquake for the region has been adequately
1 -15
Woodward· Clyde Consuifcants
defined, the maximum size of a reservoir-induced event is limited by
the maximum earthquake that would occur in the region independently of
RI S.
Review of historical RIS data (Packer and others, 1977; Packer and
others, 1979) strongly suggests that reservoir-induced earthquakes
of magnitude (Ms) larger than 5 occur where faults with recent dis-
placement lie within the hydrologic regime of the reservoir. Since
no faults with recent displacement were found within the hydrologic
regime of the proposed reservoirs (discussed in Sections 1.6 and 4.4),
the likelihood of an RIS event of magnitude (Ms) greater than 5 is
considered to be very low. However, the detection limits for faults
with recent displacement in this region suggest the theoretical pos-
sibility of the presence of a fault within the hydrologic regime of
the reservoirs that could generate a magnitude (Ms) 6 earthquake
(Section 1.6). Consequently, there is some likelihood that a reservoir-
induced earthquake of up to magnitude (Ms) 6 could occur.
A model was developed to estimate the likelihood that earthquakes of
magnitude (Ms) ~4 could occur within the hydrologic regime of the
reservoir during the design life of the Project (Section 6). Applica-
tion of the model shows that a moderate to large RIS earthquake is
unlikely and that RIS has little effect on values for ground-motion
parameters. The limited effect is due in large part to there being no
faults with recent displacement within the hydrologic regime of the
reservoir and the remote possibility of the presence of an undetected
fault that could be the source of the detection level earthquake. The
results of using this model were incorporated into design analysis
(Section 8. 2) .
1 -16
Woodward-Clyde Consultants
1.10-Ground Motions
Both deterministic and probabilistic assessments were made of earth-
quake ground motions at the sites. The parameters of ground motions
addressed in these studies included peak horizontal ground acceleration,
response spectra, and duration of strong shaking. Estimated mean peak
horizontal ground accelerations and duration of strong ground shaking
(significant duration) at the sites due to maximum credible earthquakes
are the following:
Earthquake Maximum
Source Magnitude
Benioff zone 8-1/2
(interplate)
Den a 1 i f au 1 t 8
Detection 1 eve 1 6
earthquake
Mean Peak Acceleration
(g IS)
Watana Devi 1 Canyon
Site Site
0.35 0.3
0.2 0.2
0.5 0.5
Significant
Duration
(sec)
45
35
6
Response spectra of site ground motions for these maximum credible
earthquakes are presented in Section 8.
Probabilities of exceedance were estimated for various levels of peak
ground acceleration at the Watana site. For probability of exceedance
1 eve l s of 50 percent, 30 percent, 10 percent, 5 percent, and 1 percent
in 100 years, the corresponding peak ground accelerations are the
following:
Probability of Exceedence
in 100 years
50 percent
30 percent
10 percent
5 percent
1 percent
Peak Ground Acceleration
at Watana Site (g)
1 -17
0.28
0.32
0.41
0.48
0.64
Woodward-Clyde Consultants
The i nterp 1 ate region of the Benioff zone was found to dominate the
cant ri but ions to the probabi 1 it i es of exceedance. Other sources of
earthquakes, including the Denali fault and the detection level earth-
quake contributed only slightly to the probabilities of exceedance.
At the Devil Canyon site, peak ground accelerations for given probabili-
ties of exceedance would be slightly lower than those at the Watana site
because the Devil Canyon site is somewhat farther from the Benioff
zone.
Possible design ground-motion criteria were formulated for a maximum
credible earthquake occurring on the interplate region of the Benioff
zone. The criteria are consistent with those typically used for
critical facilities. These criteria, which are presented in Section 8,
included 84th percentile smooth response spectra and an acceleration
time history for use in the seismic analysis of Watana Dam. The
acceleration time history has an appropriately long duration and a
response spectrum that closely fits the 84th percentile response spectra
at the anticipated fundamental natural period of Watana Dam.
Design ground-motion criteria for a detection level earthquake can be
formulated using the same approach as that used for a maximum Benioff
zone earthquake. However, it is also appropriate to consider the
relatively low probability of detection level earthquakes in comparison
to Benioff zone earthquakes in selecting design criteria for a detection
level earthquake.
For non-critical facilities, such as a powerhouse or transmission
towers, the results of the probabilistic studies can be used to aid in
selecting design criteria. Selection of the design criteria may include
consideration of acceptable levels of probabilities of exceedence,
economics, and acceptable risks of damage to these facilities.
1 -18
Woodward-Clyde Consultants
1.11-Conclusions
Two sets of conclusions were drawn from the results of the invest i ga-
t ion. The feasibility conclusions are those considered important in
evaluating the feasibility of the Project. The technical conclusions
are those related to the scientific data collected. Both sets are
presented in Section 10. The feasibility conclusions, in summary, are
the following:
1) The faults with known recent displacement closest to the Project
sites are the Castle Mountain and Denali faults. These faults,
and the Benioff zone associated with the subducting Pacific plate,
are considered to be seismic sources. Maximum credible earthquakes
(MCEs) for the Castle Mountain and Denali faults, and the inter-
plate and intraplate regions of the Benioff zone, have been
estimated as: a magnitude (Ms) 7-1/2 earthquake on the Castle
Mountain fault, 71 miles (115 km) from the Devil Canyon site and
65 miles (105 km) from the Watana site; a magnitude (Ms) 8
earthquake on the Denali fault, 40 miles (64 km) from the Devil
Canyon site and 43 miles (70 km) from the Watana site; a magnitude
(Ms) 8-1/2 earthquake on the interplate region of the Benioff
zone, 57 miles (91 km) from the Devil Canyon site and 40 miles
(64 km) from the Watana site; and a magnitude (Ms) 7-1/2 earthquake
on the intraplate region of the Benioff zone, 38 miles (61 km) from
the Devil Canyon site and 31 miles (50 km) from the Watana site.
2) Of the 13 significant features, nine were found to be lineaments
and four were found to be faults. No evidence of faults with
recent displacement (displacement in the past 100,000 years) was
found on features that pass through or adjacent to the Project
sites; therefore, none of the 13 significant features near the
sites are judged to be faults with recent displacement for purposes
of seismic design.
1 -19
3) The detection level earthquake (an earthquake that theoretically
could occur on an undetected fault with recent displacement) was
judged to be a magnitude (Ms) 6 earthquake that could occur
within 6 miles (10 km) of either site.
4) Estimates of peak acceleration, response spectra, and duration of
strong shaking at the sites were made for the Denali fault, the
interplate region of the Benioff zone, and the detection level
earthquake. The results of the probabilistic ground-motion
(seismic exposure) study indicate that the source most likely to
cause ground shaking at the site is the interplate region of the
Benioff zone. Possible design criteria have been formulated for
the Benioff zone earthquake.
1 -20
I A G \U
0 50
A. MAP VIEW OF SEISMIC SOURCES ??4d
0 50
Location of
1964 Ea rthquake
Plate Motion Relative to North American Pl ate
0 50
B. THREE DIMENSIONAL VIEW OF SEISMIC SOURCES
0 50
WOODWARD-CL YD E CONSULTANTS 41410A February 1982
A
I
100 K i lometers
Benioff Zone
Seismicity
100 Miles
100 Kilometers
LEGEND
WI
D I
NOTES
Active crustal fault conside red
as a seismic source. Arrows
show sense of horizontal dis-
placement: U is up; D i s down.
Active crustal fault segment not
considered as ·a seismic source
Watana Site
Devil Canyon Site
1. Fault locations after Beikman (1974a ; 1974b)
2 . A-A' is the section orientation shown below in the
three dimensional view of seismic sources.
SUMMARY OF SEISMIC SOURCES
FIGURE 1-1
2 -INTRODUCTION
2.1 -Project Description and Location
Present conceptual plans for the Susitna Hydroelectric Project (referred
to hereafter as the Project) include two dams and reservoirs in the
Talkeetna Mountains of south central Alaska (Figure 2-1). This study to
evaluate the feasibility of the Project was authorized by the Board of
Directors of the Alaska Power Authority (APA) on 2 November 1979. Acres
American Inc. (Acres) was selected by the Alaska Power Authority (APA)
to conduct the feasibility study. A Plan of Study (POS) was developed
by Acres that identified the scope of services to be conducted for the
feasibility study (Acres American Inc., 1980).
The overall objectives of the feasibility study were to:
1) Establish the technical, economic, and financial feasibility of
the Project to meet the future power needs of the Railbelt Region
of the State of Alaska;
2) Evaluate the environmental consequences of designing and con-
structing the Susitna Project; and
3) File a complete license application with the Federal Energy
Regulatory Commission.
Woodward-Clyde Consultants was one of a six-member team of consultants
assembled by Acres to meet the objectives of the study. The objectives
and scope of participation in the feasibility study by Woodward-Clyde
Consultants are described in Sections 2.2 and 2.3.
2 - 1
Woodward-Clyde Consultants
The Project is located on the Susitna River, 50 miles (80 km) north-
east of Talkeetna, Alaska, in the Talkeetna Mountains (Figures 1-1 and
2-1). The Devil Canyon site is located at river mile 133 (62.8° north
latitude, 149.3° west longitude); the Watana site is located at river
mile 165 (62.8° north latitude, 148.6° west longitude). This report
encompasses the region within 62 miles (100 km) of either site. Thus,
the Project site region includes the Talkeetna Mountains, the north-
central portion of the Alaska Range, and portions of the Susitna
and Copper River lowlands.
The Project, as presently planned, would involve two dams on the Susitna
River (Figure 2-1). Plans for the downstream location--the Devi 1
Canyon site--call for a concrete arch dam having a structural height of
approximately 645 feet (197 meters) and an estimated maximum water
depth of 545 feet (166 meters). The impounded reservoir would be
approximately 26 miles (42 km) long and have a storage capacity of
approximately 1,092,000 acre-feet (1,348 x 106m3). Plans for the
upstream location--the Watana site--include an earthfill or rockfill dam
having a structural height of approximately 885 feet (270 meters) and an
estimated maximum water depth of 725 feet (449 m). Its impounded
reservoir would be approximately 48 miles (77 km) long and have a
storage capacity of 9,515,000 acre-feet (11,741 x 106 m3) (Acres
American Inc., in press). A transmission line, approximately 365 miles
(588 km) long, is planned to connect the power plants at the dam sites
with existing transmission lines.
2.2-Objectives
The responsibility of Woodward-Clyde Consultants for the Project
feasibility study was defined in Task 4 of the POS prepared by Acres and
issued by the APA in February 1980. The objectives of the POS assigned
to ~Joodward-Clyde Consultants were to:
2 -2
a) Identify faults that have the 'potential for surface rupture through
the Project or the potential for significant ground motions at the
Project;
b) Estimate the dimensions of surface rupture and estimate the ground
motions at the Project;
c) Undertake preliminary evaluations of the seismic stability of
proposed earth-rockfill and concrete dams;
d) Assess the potential for reservoir-induced seismicity and seis-
mically induced landslides in the reservoir area; and
e) Identify soils that would be susceptible to seismically induced
failure along the proposed transmission line and access routes.
Task 4 of the POS is subdivided into subtasks 4.01 through 4.15. These
subtasks were identified by Acres to meet the overall objectives
of Task 4 of the POS. The subtasks were established to provide the
geologic, seismologic, and earthquake engineering data needed to assess
the feasibility of the Project. The correspondence of the subtasks
identified by Acres to the objectives of Task 4 of the POS are listed in
Table 2-1.
The objectives (a through e) described above are addressed by the
following subtasks and sections of the report:
0
0
Objective (a) is addressed by Subtask 4.11, and the results are
presented in Section 4;
Objective (b) is addressed by Subtasks 4.11 and 4.13, and the
results are presented in Sections 4 and 8;
2 - 3
0
0
Objective (c) is addressed by Subtask 4.14 as a consulting service,
as described below;
Objective (d) is addressed by Subtask 4.10, and the results are
presented in Section 6; and
Objective (e) is addressed by Subtask 4.15, and the results are
presented in Section 9.
Subtasks 4.01 through 4.08, reported in the Interim Report (Woodward-
Clyde Consultants, 1980b), were conducted in 1980 to provide data for a
preliminary assessment of project feasibility. Subtasks 4.09 through
4.15 were conducted in 1981. This Final Report presents the results
from a 11 s ubt asks conducted during the two-year investigation. except
Subtasks 4.09 and 4.14. Subtask 4.09 is a manual for a long-term
earthquake monitoring system that is presented in a separate document.
Subtask 4.14 consisted of consultation without a report requirement.
The results presented in this report were developed for the purpose of
evaluating Project feasibility and should be reviewed when final dam
design is considered. The results are provided in support of the
license application to be submitted to the Federal Energy Regulatory
Commission (FERC) by Acres on behalf of the APA.
2.3 -Scope
This report is the product of a two-year investigation intended to
provide data for seismic design feasibility considerations. In this
report, the work conducted during the first year will be referred
to by the term 11 1980 study"; work conducted during the second year
will be referred to by the term "1981 study." The term "investigation"
will be used for the two-year program.
2 - 4
Woodward· Clyde Consultants
The multidisciplinary approach used during this investigation involved
interaction among a team of structural geologists, Quaternary geolo-
gists, seismologists, and earthquake engineers. Their task was the
analysis of potential seismic sources, recency of fault displacement,
surf ace rupture potential, and ground-motion parameters. The scope of
work for Subtasks 4.01 through 4.08 is discussed in Section 2 of
the Interim Report (Woodward-Clyde Consultants, 198Gb). For subtasks
4.09 through 4.15, a scope of work was developed which included the
following:
a) a detailed compilation and review of information describing the
Quaternary geology of the site region; this information included
aerial photography at scales of 1:24,000, 1:44,000, and 1:125,000
and LANDSAT imagery at a scale of 1:250,000;
b) acquisition and interpretation of low-sun-angle aerial photography
of the area within 6 miles (10 km) of each dam site and along
selected segments of the Talkeetna thrust fault and the Susitna
feature;
c) geological field studies of the Quaternary geology and of 13
significant features. These studies included aerial reconnais-
sance, field mapping, interpretation of aerial photographs,
geophysical surveys, age dating, trench excavation and logging, and
review of pertinent mine data;
d) development of a plan for a long-term seismic monitoring network
and preparation of a manual for use in implementing the network;
e) review and analysis of records of selected historical moderate
to large earthquakes within or adjacent to the Talkeetna Terrain;
2 - 5
Woodward·CDyde Consultarats
f) review and analysis of records of selected earthquakes in the
Talkeetna Terrain to provide additional insights into the stress
regime of the Terrain;
g) review and analysis of worldwide Benioff zone earthquakes to
refine preliminary estimates of the maximum credible earthquakes
(MCE s) ;
h) review and analysis of worldwide moderate to large earthquakes to
assess the magnitude of the detection level earthquake;
i) development of a statistical model to incorporate the effect
of reservoir-induced seismicity on seismic design parameters;
j) assessment of the potential for reservoir-induced seismicity;
k) estimation of maximum credible earthquakes (MCEs) for Talkeetna
Terrain boundary faults and estimation for each MCE recurrence
interval and rupture length;
l) evaluation of attenuation relationships for crustal and Benioff
zone seismicity in Alaska and selection of appropriate relation-
ships for establishing seismic design parameters;
m) selection of seismic design parameters using both deterministic and
probabilistic approaches;
n) interpretation of aerial photography along the rights-of-way of the
proposed access road and transmission line to identify potential
areas of seismically induced failure, such as landslides and
liquefaction; and
o) preparation of this report to summarize the results of the two-year
investigation.
2 -6
Woodward· Clyde Consultants
Completion of the scope of the two-year investigation involved approxi-
mately a 130-person-month leve 1 of effort. This inc 1 uded: appro xi-
mate ly 30 person-months for the data compilation, 70 person-months for
the field studies, and 30 person-months for data analysis and report
preparation.
2.4-Fault Study Rationale
2.4.1-Conceptual Approach
The conceptual approach to studies of faults, including faults
with recent displacement, was reviewed in detail in the Interim
Report (Woodward-Clyde Consultants, 1980b). The following is a
summary of the key aspects of the approach used to guide the
overall investigation.
a) The earth's crust is comprised of a series of plates that are
moving in relation to one another. Plate movement can result
in collisions with resultant subduction (underthrusting of one
plate beneath another).
b) When two crustal plates collide, the plate with the heavier
crust is usually subducted (underthrust) beneath the other.
Eventually the subducted plate falls or is thrust downward
into the upper mantle and becomes detached from the overriding
plate. Since plates from the oceanic crust are heavier
than continental plates, the oceanic plates are generally
underthrust.
c) Where subduction is occurring, the subduction process gene-
rates tectonic stress: within the downgoing plate, within the
overriding crustal plate, and along the interface between the
two plates where they are in contact with one another.
2 - 7
Woodward-Clyde Consultants
d) The subduction process leads to a complex pattern of deforma-
tion, faulting, and seismic activity. An understanding of the
subduction process provides the seismotectonic framework
within which to evaluate the significance of faults and
earthquakes.
e) Tectonic stress can lead to displacement along fault planes.
The resulting instantaneous release of energy (an earthquake)
produces seismic waves; these waves are propagated through the
earth's crust and mantle and result in ground motion that is
commonly referred to as earthquake shaking.
f) Faults that are sources of earthquakes are typically subject
to repeated displacements as long as the tectonic stress
environment remains unchanged. Therefore, faults that show
evidence of recent displacement are assumed to have the
potential for future displacement.
g) Rupture of the ground surface occurs during an earthquake if
the energy is released at a sufficiently shallow depth along a
fault that intersects the ground surface. When the energy
release occurs deeper in the crust or beneath the crust, or
when the energy release is sma 11 re 1 at i ve to the depth of
release, the fault does not rupture the ground surface.
h) The direction and rate of movement between plates has changed
during geologic time, resulting in a changed tectonic stress
environment. After such a change, displacement may occur
on some pre-existing faults and cease on other faults.
Therefore, faults that 1>1ere sources of earthquakes in a
previous tectonic environment, and are found to be faults
without recent displacement in the current tectonic environ-
ment, are not likely to be sources of earthquakes in the
current tectonic environment.
2 - 8
i) The frequency of earthquakes is related to the cyclic elastic
strain buildup and release by fault rupture; therefore, the
frequency varies greatly from one part of the earth's crust to
another. The interval between earthquakes on the same fault
or fault system may be longer than the period for which
historical records of earthquakes are available. Therefore,
the most reliable approach to evaluating the frequency of
earthquakes is one which utilizes an understanding of both the
geologic record and the historic seismicity record.
j) Earthquakes and the related surface rupture potential at a
given location in the earth's crust or 1 i thosphere can be
evaluated by using the history of surface fault rupture (or
displacement) that is expressed in the geologic record of the
past. As most commonly applied, if displacement has occurred
on a fault within a specified past time period, the fault is
classified as having recent displacement. Faults with recent
displacement (as defined for a particular project) are then
inferred to have a potential for surface rupture and earth-
quakes. This potential is considered in the design of that
project. Guidelines defining what is considered "recent
displacement" for this project are described in Section 2.4.2.
k) A fault that has been subject to frequently occurring and
large recent displacement appreciably affects the surface
geology and topography. For these faults, the record of past
earthquakes in the surface geology and topography is clearly
recognizable. A fault that has been subject to relatively
infrequent and small displacement may not greatly affect
the landscape. The evidence of these displacements may
be difficult to detect and to evaluate; however, it is
improbable that all evidence of young faulting would be
2 - 9
completely obliterated by weathering, erosion, and deposition.
Experience during the past decade or so indicates that it is
very unlikely for faults with recent displacement to have no
effect on the landscape. Geologists that are experienced in
assessment of recent fault displacement can detect these
faults (Sherard and others, 1974).
2.4.2 Guidelines for the Identification of Faults with Recent
Displacement
Regulatory definitions of a fault with recent displacement, such as
those discussed in the Interim Report (Woodward-Clyde Consultants,
1980b), can lead to a simplistic and possibly misleading concept of
the significance of a particular fault. If a fault has been
subject to displacement within a specified past period of time,
whether it is 11,000 years, 35,000 years, or 100,000 years,
it is important to understand how much displacement has occurred,
how often it has occurred, and the sense of displacement. For
example, consider a fault that has been subject to 0.2 inches (5
mm) of displacement every 75,000 years and a second fault that has
been the source of 3.3 feet (1 m) of displacement every 10,000
years. Both faults can be considered to have recent displacement
(if displacement within the past 100,000 years is used as the
definition of a fault with recent displacement); however, the first
fault is a much less important source of earthquakes than the
second fault when measured in terms of the size of the displacement
and the frequency of occurrence. For purposes of dam design, the
effect of displacement on these two faults can be significantly
different. In addition, the sense of relative displacement is also
important. As discussed by Sherard and others (1974), the effect
on dam design of displacements on thrust faults, normal faults, and
strike-slip faults is different for each type of fault.
2 -10
Woodward-Clyde Consultants
Dams have been designed to accommodate ground motions from rela-
tively large earthquakes that may occur relatively close to the
dam. For example, the San Pablo Dam in California is designed
to accommodate the ground motions of a magnitude (Ms) 8-1/2 event
on the San Andreas fault and a magnitude (Ms) 7-1/2 event on the
Hayward fault, approximately 12 miles (20 km) and 10 miles (16 km)
from the dam, respectively. Dams have also been designed to
accommodate surface rupture. For example, the Coyote Springs Dam,
built in California in 1936, was designed as an earth dam to
accommodate 20 feet (6 meters) of horizontal displacement and
3.3 feet (1 meter) of vertical displacement in the foundation.
Consequently, any consideration of faults with recent displacement
ultimately needs to address not only how recently fault displace-
ment has occurred, but also how much displacement has occurred, how
often it has occurred, and what the sense of displacement has
been. From these data, an assessment can be made of the likelihood
that the fault will have these characteristics in the future. From
this assessment, the seismic source potential and potential for
surface rupture for a particular fault can be considered in an
appropriate manner by the designer of the dam. Explicit geologic
evidence for the recency of displacement along a fault may not
exist; for these faults, the avai 1 able data and experience with
other faults are used in the final evaluation.
The guidelines for this investigation were conservatively selected
by Acres in the absence of criteria established by FERC, the
review agency to whom the license application will be submitted.
The guidelines were selected after available regulatory and
dam-building agency guidelines were reviewed (these agency guide-
lines are summarized in the Interim Report [Woodward-Clyde Consul-
tants, 198Gb]) and after discussions were held with project team
members. The guidelines were based on state-of-the-practice
2 -11
knowledge for identifying faults with recent displacement that
should be considered in dam design; further refinements of this
knowledge should be incorporated into future studies and into these
guidelines. The guidelines are as follows:
1) All lineaments or faults that have been identified by the
geology and seismology community as having been subject to
recent displacement should be included in assessing the
seismic design criteria for the Project.
2) All features identified as faults that have been subject to
displacement in approximately the past 100,000 years should be
considered to have had recent displacement. All faults having
recent displacement should be considered when assigning design
criteria for ground motions or for surface displacement at the
structure sites.
3) If a lineament or a suspected branch of a lineament is within
6 miles (10 km) of either site, then a detailed investigation
should be made to establish whether or not the lineament or
branch can be considered to have recent displacement and
whether the potential exists for displacement in the dam
foundation.
4) Lineaments more distant than 6 miles (10 km) from a dam site
that could cause significant ground motion at the site if they
were faults with recent displacement should be investigated in
detai 1. An assessment should be made of whether or not a
lineament is a fault and if it has been subject to recent
displacement.
2 -12
2.5 -Methodology
The purpose of this investigation was to estimate, for consideration
{if Project feasibility, the values of seismic design parameters for
9round motion and rupture of the ground surf ace during earthquakes. The
'1nterrelationship of each step of our interdisciplinary program is shown
in Figure 2-2.
The methode 1 ogy used to identify seismic source parameters and sources
of potentia 1 fault rupture through the sites inc 1 uded the fo 11 owing key
iitems:
J!.) Identification of faults with recent displacement. This included
studies of crustal faults and lineaments using the methods dis-
cussed in Section 4 and Appendix A.
.. "J) ~~~ Identification of seismic sources, including crustal faults with
recent displacement and Benioff zone sources. This included
studies of historical earthquakes using methods discussed in
Section 5 and the studies of faults with recent displacement
discussed in item (1) above.
3) Evaluation of and selection of a detection level earthquake,
which represents an earthquake whose source (a fault with recent
displacement) would not be detected at the surface. The detailed
methology for this work is discussed in Section 4.2.
·~) Estimation of maximum credible earthquakes for each of the seismic
sources. The methodology for this work is discussed in Section 7.
'.-) ;) Development of attenuation relationships for the seismic sources
using methodologies described in Section 8.
2 -13
Woodward-Clyde
Estimation of ground-motion parameters for the seismic sources.
A deterministic approach has been used; that is, the maximum
earthquakes is assumed to occur during the 1 ifet ime of the Project
at the closest approach of the seismic source to the Project.
2 -14
TABLE 2-1
PROJECT SUBTASKS AND OBJECTIVES
~btask Number and Description
4.01 Review of Available Data
4.02 -Short-term Seismologic
Monitoring Program
4.03 -Preliminary Reservoir-
Induced Seismicity
4.04 -Remote Sensing Image
Analysis
4.05-Seismic Geology
Reconn ai ss ance
4.06 -Evaluation and
Reporting
4.07 -Preliminary Ground-
Motion Studies
4.08 -Preliminary Analysis of
Dam Stability
4.09 -Long-term Seismologic
Monitoring Program
4.10 -Reservoir-Induced
Seismicity
4.11 -Seismic Geology
Field Studies
4.12 -Evaluation and Reporting
4.13 -Ground-Motion Studies
4.14-Dam Stability
Consulting Services
4.15-Soil Susceptibility to
Seismically-Induced
Failure
Objective
Acquire, compile, and review existing data, and identify
the earthquake setting of the Susitna River area.
Establish initial monitoring system, obtain and analyze
basic seismologic data on potential earthquake sources
within the Susitna River area and supply information
required to implement a more thorough long-term monitoring
program (Subtask 4.09).
Evaluate the potential for the possible future occurrence
of reservoir-induced seismicity (RIS) in the Susitna
Project area.
Select and interpret available remote sensing imagery to
identify topographic features that may be associated with
active faulting.
Perform a reconnaissance investigation of known faults in
the Susitna River area, and of lineaments that may be
faults, identify active faults, and establish priorities
for more detailed field investigations.
Complete a preliminary evaluation of the seismic environ-
ment of the project, define the earthquake source para-
meters required for earthquake engineering input in
design, and document the studies in reports suitable for
use in design studies.
Undertake a preliminary estimate of the ground motions
(ground shaking) to which proposed project facilities may
be subjected during earthquakes.
Provide input for preliminary evaluations of the seismic
stability of proposed earthfill, rockfill, and/or concrete
dams during maximum credible earthquakes.
Develop a long-term seismologic monitoring program to
provide a continuing source of seismological data for
refinement of the seismic design aspects of the project
during the detailed design phase.
Refine the estimate for the potential for reservoir-
induced seismicity made in Subtask 4.03.
Perform seismic geology field studies to identify faults
that may be active and in the vicinity of the selected dam
sites.
Refine the evaluation of the seismic environment and the
earthquake source parameters derived in Subtask 4.06,
complete the reporting of all the fieldwork and studies
undertaken in Subtasks 4.01, 4.05, and 4.09 to 4.11,
and provide coordination and management to Subtasks 4.09
to 4.11.
Refine the estimate of ground-motion characteristics made
in Subtask 4.07.
To provide consulting assistance to the Acres design group
engaged in the feasibility design of the dams.
Provide input on behavior of those areas along the
transmission line and major access road routes that appear
to be underlain by soils particularly susceptible to
seismically-induced ground failure such as liquefaction or
landsliding.
Stephar.
WOODWARD-CLYDE CONSULTANTS 41410A February 1982
Monahan
Flat
r?
Butte Lake
Copper River
Basin
Brooks Range \
~~._.,. \
fl;,O ~~ \
ALASKA Fairbanks\ CANADA . ' I
Gulf of Alaska
NOTE
1. Proposed reservoir configuration is after
Acres American Inc. (In Press).
PROJECT LOCATION MAP
0 5 10 15 20 Miles ~~~~
0 10 20 30 Kilometers
FIGURE 2-1
1980
Yes
No
lt.mottly tented dlta
lnttfl)fetltion and geologK:
dJ\1 review
0...-eiop tectonec model
M.ke preltmilllry esumate
~---------lof ground motton parameters+------!
No
for feasJbil•tv IUe'SSmtnt L-----y----------------'
1981
Conduct detailed analysis
of 11lecttd hinorlc
earthquake records
As.sen the ltv@f o ' dettct~bllitv for fauhs wnh recent
dtspl.cement tncludmg m•n•mum eanhauake 1nd mntmum
recurreo~e interval (t.e the detectton level urthquakel
AIR'A stress regtme wtthtn
tiClonk model
Select "'~••mum crecf•ble
t:lr1hquakn and wurctl"\ 10t
8~·off 2om
O«Nelop a longterm micro-
a,hquake operations manual
Provide muimum credible
ground motion parameters to
dam and nam.mis.sion tower
designer
No lddttional SU.Jdy
Pto¥tdt: rn~xtmum amount,
1YP« and ltkf'hhood of
tJC)IICemcnl to dam destg~
ptrlmtters to d1m designer
WOODWA RD-CLYDE CON SULTANTS 41410A February 1982
Assess potential for
teismically induced
landslides wlthtn reseNoir
AtiHS potential for
telsmiatly induced failure
•long uans,mission line and
ecceu room
PROJECT FLOW DIAGRAM
FIGURE 2-2
WoodwaJrd·Ciydle Consultants
3 -QUATERNARY GEOLOGY
3.1 -Introduction
Quaternary geology involves the study of geological processes in
recent geologic time (i.e., the past 1.8 million years). In the site
region these processes are primarily of glacial origin. As shown in
Figure 3-1, the Quaternary Period includes the Holocene and Pleistocene
epochs.
An understanding of the extent and age of Quaternary sediments and
surfaces (collectively referred to here as surfaces) is important to
seismic geology studies. This importance is related to the assessment
of recency of displacement along a fault. If activity on a fault
has produced surface displacement in recent geologic time, the young
(Quaternary) surfaces would likely include evidence of deformation.
When the age of the deformed surfaces is known, then the recency of the
fault displacement and the number of displacements can be assessed.
Conversely, if a fault has not produced surface displacement in recent
geologic time, then the overlying young (Quaternary) surfaces would not
be deformed. When the age of the overlying surfaces is 'known, then the
length of time during which there has been no displacement can be
assessed.
Within the site region, the late Quaternary surfaces are of greatest
interest to the seismic geology study. These surfaces include those of
Holocene and Pleistocene age (including the Wisconsinian and Ininoian
stages). This interest is due in large part to their extent and age.
These surfaces are present throughout the site region, and their age
ranges from a few years to approximately 120,000 years before present
(y.b.p.) (Figure 3-1). One of the project objectives was to identify
3 - 1
Woodward·CHyde
faults with recent displacement. A fault with recent displacement was
defined for this Project as a fault that has had surface rupture during
the past 1GG,GGG years (Section 2.4 .2). Therefore, the ability to
assess the presence or absence of fault displacement in the late
Quaternary surfaces is pivotal to the identification of faults with
recent displacement.
The approach to Quaternary geology studies used during this invest i ga-
t ion was to develop an understanding of 1) the Pleistocene glacial
geology of southern Alaska, and 2) the late Pleistocene (Wisconsini an
and Illinoian) glacial geology within a region (designated as the
Quaternary study region) encompassing the project sites and major
segments of the 13 features identified in the Interim Report (Woodward-
Clyde Consultants, 198Gb). Figures 3-2 and A-1 show the Quaternary
study region.
In 198G the Quaternary geology study involved a preliminary literature
review, limited aerial photography and imagery interpretation, and
aerial reconnaissance of selected areas. The results of this work are
summarized in Sections 6 and 7 of the Interim Report (Woodward-Clyde
Consultants, 198Gb). During 1981, the second year of the investigation,
the Quaternary geology study involved:
a) a complete review of pertinent literature;
b) detailed interpretation of large-and small-scale aerial photo-
graphy;
c) preparation of a preliminary map based on (a) and (b), prior to
the field season, showing the extent and estimated age of late
Quaternary surfaces;
3 - 2
d) field programs, including field mapping of key areas, excavation
of test pits, and collection of relative age (weathering) data, and
collection of material for radiocarbon age dating;
e) radiocarbon age dating of 11 samples; and
f) data synthesis and revision of the preliminary map of late Quater-
nary surfaces, listed in (c) above; this revised map is shown in
Figure 3-2.
This section of the report presents a summary of the results of these
Quaternary studies. The details of the work, i ncl udi ng the methodology
used, are presented in Appendix A. Two components of the methodo 1 ogy
are particularly important to understanding how age determinations and
correlations were made in the Quaternary study region. These are the
relative age (weathering) dating technique and the correlation technique
used to correlate glaciogenic features according to their relative
elevation and age.
The relative age (weathering) dating technique involved the use of
procedures and relationships similar to those used by Ten Brink in the
Alaska Range (Ten Brink and Waythomas, in press; Werner, in press)
(Appendix A.l.3.2). The technique is referred to hereafter as the
relative age dating technique. It involves the measuring of weathering
characteristics (such as the proportion or ratio of fresh, partially
weathered, and weathered granitic boulders, or the depth of oxidation)
of moraines. Differences in these characteristics, combined with
radiocarbon age dates, moraine elevation, and moraine morphology were
used to distinguish between the different aged surfaces (e.g., Early and
Late Wisconsin surfaces) in the Quaternary study region. In addition,
the relative age dating results from this study were compared with the
radiometrically dated scale of Ten Brink and Waythomas (in press)
developed in the Alaska Range (Table A-1).
3 - 3
The correlation of maximum elevations among glaciogenic features with
similar morphologic characteristics was used to assess the vertical and
horizontal extent of the glaciations in the Quaternary study region.
{Similar morphologic characteristics are suggestive of similar age, as
discussed in Appendix A.l.3.2.) To make these correlations, radiocarbon
and relative age dating of glaciogenic features was conducted in key
areas (such as those described in Section 3.4 below) to establish
the maximum elevation of a glacier in those areas during specific
glaciations. Subsequent identification of glaciogenic features with
similar characteristics in other portions of the study area were used to
correlate the glacial maximum throughout the Quaternary study region for
each given glacial period. In this manner, the maximum elevation of ice
for the various glaciations was estimated and the extent of these
glaciations was assessed for the Quaternary study region (Figure 3-2).
Using the above two components of the methodology, as well as the
others discussed in Appendix A.l.3, the late Quaternary geology for the
Quaternary study region was accomp 1 i shed.
The following discussion in this section: a) briefly reviews the
regional late Quaternary setting of southern Alaska; b) surrmarizes the
age and extent of Quaternary surfaces that were encountered in the
Quaternary study region; c) briefly discusses the key areas for which
Quaternary geologic re 1 at i onshi ps were used to develop the Quaternary
geology map (Figure 3-2); d) summarizes the glacial geology of the
Quaternary study region; and e) summarizes the relationship of Quater-
nary surfaces to the 13 features selected for field work during the 1981
field study (Section 4.4).
3.2 -Regional Pleistocene Geology Setting
Previous investigations of the Quaternary geology of south centra 1
Alaska have been either generalized regional studies or detailed studies
3 -4
of specific areas of limited extent. Representative regional studies
include those of Karl strom and others (1964), Coulter and others (1965),
Pewe and others (1965), and Pewe (1975).
Among the numerous detailed studies that have been conducted in areas
near the Talkeetna Mountains are those by: Miller and Dobrovolny
(1959), Karlstrom (1964), Trainer and Waller (1965), Schmoll and others
(1972), for the Cook Inlet area; Wahrhaftig, (1958), Thorson and
Hamilton (1977), Ten Brink and Waythomas (in press), for the central
,1\laska Range; and Chapin (1918) and Ferri ans and Schmoll (1957), for the
Copper River Basin.
little information was available prior to this investigation regarding
the Quaternary surfaces and history of the Talkeetna Mountains, although
a limited amount of data pertinent to Quaternary surfaces in the
Talkeetna Mountains has been presented by Bowers (1979) and by Terres-
trial Environmental Specialists (1981).
The reginnal and detailed studies cited above provide information
regarding the Quaternary geologic history of southcentral Alaska.
These studies suggest that the Talkeetna Mountain region existed as
an extensive mountainous to rolling upland at the beginning of the
Quaternary Epoch, approximately 1.8 m.y.b.p. Subsequent to that time a
series of climatic fluctuations, with conditions ranging from temperate
to polar, apparently began to affect the region. The fluctuating
climate, which characterized the region throughout the Quaternary
Epoch, lead to several periods of extensive glaciation during polar
conditions. At its Quaternary maximum, glacier ice formed an ice
cap over the Talkeetna Mountains. These periods of glaciation were
separated by interglacial periods with relatively temperate climatic
conditions generally similar to those now found in the region.
3 - 5
Woodward-Clyde Consultants
The general regional picture of alternating glacial and interglacial
periods along with specific evidence within the Talkeetna Mountains
suggests that, following the Quaternary glacial maximum for the region,
subsequent glacial advances were not extensive enough to produce an ice
cap over the mountains. In fact, the available evidence indicates a
series of glaciations of sequentially decreasing extent. It was these
more recent glaciations that produced the glacial, glaciofluvial, and
glaciolacustrine landforms and sediments that now dominate the Talkeetna
Terrain. There is relatively little evidence, however, on which to base
an interpretation of interglacial conditions.
Although glaciers covered only about 50 percent of the present area
of Alaska during the Quaternary Epoch, south central Alaska, south of
the crest of the Alaska Range, was nearly completely glaciated (Pew{,
1975).
Among the more recent glaciations that occurred in the Talkeetna
Mountains, four were recognized during this investigation in the
Quaternary study region (Figure 3-2). The glaciations and their
respective ages are: pre-Wisconsin, >100,000 y.b.p.; Early Wisconsin,
75,000 to 40,000 y.b.p.; Late Wisconsin, 25,000 to 9,000 y.b.p.,
which included four stades, each of which was less extensive than
the preceding one; and Holocene, <9,000 y.b.p. Figure 3-2 shows the
surfaces of the pre-Wisconsin stage, the Early Wisconsin stage, and the
first and last stade (Stades I and IV) of the Late Wisconsin stage.
These are shown because of their extent in the Quaternary study region.
Holocene surfaces are not shown in this figure because of their limited
extent and their distance from the 13 features being studied for recency
of fault displacement.
The ages for the four glaciations were assigned on the basis of radio-
carbon age dates obtained for this study and generally accepted age
assignments for similar glacial sequences elsewhere in Alaska (Hopkins,
1967; Flint, 1971; Pewe', 1975; Weber and others, 1980; Ten Brink and
Waythomas, in press).
3 - 6
During the four glaciations, ice advanced from three source areas: the
Alaska Range to the north of the Quaternary study region; the southern
and southeastern Talkeetna Mountains; and the Talkeetna Mountains north
and northwest of the Susitna River. In the Quaternary study region, the
glaciers from these sources coalesced to form a piedmont glacier that
flowed through the intermountain basin (shown in Figure A-1) that
includes the Susitna River (near the Watana site), Watana Creek, and
Stephan Lake areas. Glacial flow was dominantly to the south and
southwest and left a variety of landforms, surfaces, and sediments.
The following sections surrmarize the results of our investigation of
these landforms and sediments. They are intended to document the basis
for the application of the results to the assessment of the recency of
fault displacement on features near the sites.
3.3 -Age and Extent of Quaternary Surfaces in the Quaternary
Study Region
The 1981 Quaternary geology study led to the identification of ten types
of Quaternary glaciogenic features that were used in part to interpret
the age and extent of Quaternary surfaces in the Quaternary study
region. These features include: till, lacustrine deposits, outwash
deposits, ice disintegration deposits, kame terrace deposits, fluvial
deposits, fluting, trimlines, side glacial channels, and an assortment
of glacially sculptured bedrock forms including whalebacks, stoss
and lee, and grooves. The characteristics used to distinguish these
glaciogenic features are summarized in Table 3-1.
The morphologic characteristics of the Quaternary glaciogenic features
along with their elevation, radiocarbon age dates obtained from car-
bonaceous material, and relative age characteristics were used to
develop an understanding of the age and extent of late Quaternary
3 - 7
surfaces in the Quaternary study region. The results of this inter-
pretation are shown in Figures 3-2 and 3-3. The relationship of these
surfaces to the 13 features (whose seismic source potential was studied
in 1981) is summarized in Figure 3-4.
The age and extent of late Quaternary surfaces in the Quaternary study
region, as summarized in Figure 3-2, are interpreted to be as follows:
a) Pre-Wisconsin surfaces are preserved at higher elevations than are
Wisconsinian surfaces. Elevations range from above 4,200 feet
(1,280 m) in the northern section of the Quaternary study region to
3,100 feet (945 m) in the southern portion of the region.
b) Early Wisconsin surfaces typically are preserved along the margins
of topographically elevated areas (such as those on either side
of Butte Lake [Figure 3-6A] and the broad upland south of the
Susitna River between Kosina Creek and Oshetna River [Figure 3-2]).
c) Surfaces associated with early stades of Late Wisconsin glaciation
are present on the basin and valley floors and mid to low valley
walls of the Quaternary study region within the Talkeetna Mountains
(Figure 3-2). The Portage Creek, Deadman Creek, and Watana Creek
valleys and the Stephan Lake-Fog Lakes area are typical of the
areas with Late Wisconsin glacial surfaces.
d) Surfaces associated with the last stade of Late Wisconsin glaci a-
t ion (i.e., those resulting from the last advance of glacial ice
in Wisconsinian time) are found in valleys leading down from the
high elevations in the Talkeetna Mountains and in the basin floor
areas at the mouths of some of these valleys. In addition, the
Monahan Flat area (which lies between the Alaska Range and the
north end of the Quaternary study region) and the Butte Creek
3 - 8
Woodward-Clyde Consultants
area have extensive surfaces from the last stade of the Late
Wisconsin stage (Figure 3-2). These last stade deposits are. the
result of mountain glaciers which emanated from the high elevations
of the Talkeetna Mountains (in the northwest corner and the
southern portion of the Quaternary study region) and from glacial
ice that moved southward from the Alaska Range across Monahan Flat
and locally into the lower elevations of the Butte Creek area.
3.4 -Key Quaternary Study Areas
Four key areas within the Quaternary study region were studied in detail
to provide a basis for interpreting the age and extent of Quaternary
surf aces. These four areas have been designated as the Black River
area, the Clear Valley area, the Butte Lake area, and the Deadman
Creek area. The location of these areas is shown in Figure 3-2.
Figures 3-5 and 3-6 show the results of the studies conducted in these
areas and include the morphostratigraphic units that are present, the
age of these units, the location of test pits, and radiocarbon sample
1 ocat ions.
In order to develop the Quaternary geology map shown in Figure 3-2 and
the cross-section shown in Figure 3-3, the results of the studies in
these four key areas were used along with: additional radiocarbon
age dates from the Quaternary study region (the locations of the
dated material are shown in Figure 3-2, and the ages are summarized in
Table 3-2); the interpretation of aerial photos; and aerial reconnais-
sance and ground reconnaissance mapping. The following discussion
summarizes the key data that were obtained from these four areas.
3 - 9
3.4.1-The Black River Area
The Black River area is located south of the Susitna River near the
Copper River basin in the southeastern part of the Quaternary study
region, as shown in Figures 3-2 and 3-5A. The area is part of a
broad undulating plain in the eastern Talkeetna Mountains that
merges with the adjacent Copper River basin, as discussed in
Section A.l.4.6. Three morphologically distinct, glacially scoured
topographic surfaces of pre-Wisconsin, Early Wisconsin, and Late
Wisconsin age have been beveled into this plain by succeeding
less extensive glaciations. Two of these surfaces, those of the
Early and Late Wisconsin stages, were observed in the Black River
area. In addition, the last stade of the Late Wisconsin stage is
represented by glacial sediments, as shown in Figures 3-2 and 3-5A.
Studies were conducted of the Late Wisconsin surfaces to assess the
extent of glacial ice in the eastern Talkeetna Mountains and the
adjacent Copper River basin during Early and Late Wisconsin
time. At the junction of the Susitna and Oshetna Rivers, north of
the Black River area (Figure 3-2), till was observed to interfinger
with highly deformed lacustrine and deltaic deposits. A sample
(S47-4) of wood obtained from the lacustrine sediments gave a
radiocarbon age date of >37,000 y.b.p. (Table 3-3). These rela-
tionships strongly suggest that the t i 11 and 1 acustri ne deposits
are of Early Wisconsin age. Stratigraphic evidence suggests that
Late Wisconsin ice did not advance into the lower reaches of the
Black and Oshetna Rivers nor into the Susitna River in the eastern
Talkeetna Mountains (Figure 3-5A).
Mapping, based on aerial photo interpretation, of the extent of
moraines and relative age dating results corroborate the limited
extent of Late Wisconsin ice in this area. As shown in Table 3-3
3 -10
and Figures 3-2 and 3-5A, moraines BR-1, BR-2, and BR-3 (Late
Wisconsin age) and deposits of similar age are interpreted to
terminate 6 to 7 miles (10 to 11 km) south of the Susitna River.
From these data, Early Wisconsin glacial ice is inferred to have
moved northward into the Black River area from southerly sources
in the high elevations of the Talkeetna Mountains. This ice is
believed to have coalesced with ice moving from the Copper River
basin westward along the Susitna River valley. Late Wisconsin
glacial ice is inferred to have moved northward, down the Black
River area, from southerly sources at high elevations in the
Talkeetna Mountains. The glaciers were of limited extent, and ice
did not move out of the Black River area.
3.4.2-The Clear Valley Area
The Clear Valley area lies approximately 7 miles (11 km) south of
the Watana site (Figure 3-2). It is a glaciated valley that opens
into the lowland area associated with Fog Lakes and Stephan Lake,
as discussed in Section A.l.4.1. Twelve closely nested moraines,
glacial trimlines, and ice marginal channels have been used
to develop the glacial chronology in this area, as shown in
Figures 3-2 and 3-5B.
Pre-Wisconsin periglacial effects are present above elevation
3,100 feet (945 m). These effects are primarily those of well
developed frost-shattered boulder fields.
Studies of the 12 nested moraines were conducted to distinguish
the extent and elevation of Early and Late Wisconsin glacial
ice in proximity to the Watana site. Moraine morphology, relative
age dating, and cross-cutting relationships were interpreted to
show a distinct difference in age between the lower seven moraines
3 -11
(CL-1 through CL-7 in Figure 3-68) and the upper five moraines
(CL-8 through CL-12) (Table A-1). In addition, the distal end
of mora i n e s C L-6 and C L-7 bend s out h westward and 1 o s e the i r
topographic identity in the area of ground moraines at elevation
2,500 to 2,700 feet (762 to 823 m). This deflection and loss of
topographic expression suggests that Late Wisconsin ice from the
Clear Valley area merged with a southwestward flowing piedmont
ice sheet that flowed through the intermountain basin that is
delineated in Figure A-1. It further suggests that the 2,500 to
2,700 feet (762 to 823 m) elevation represents the upper limit of
Late Wisconsin ice in the basin.
The last stade (Stade IV) of Late Wisconsin ice is prominently
represented in the Clear Valley area by ice disintegration features
including eskers, kame deltas, and kettles. The relationship of
these features to the earlier Late Wisconsin moraines suggests that
glacial ice moved northward from the high areas south of the Clear
Valley area. The advance, however, did not extend into the basin
floor area at the mouth of Clear Valley and adjacent valleys.
A small Holocene lake was dammed by a till and bedrock ridge in the
basin floor area north of the Clear Valley area (Figure 3-58)
probably beginning in the waning stages of the Late Wisconsin
time and continuing into Holocene time. A radiocarbon age date of
approximately 3,500 years before present was obtained from the
lacustrine deposits (Sample 54-1).
The glacial chronology in the Clear Valley area (along with the
results of studies conducted in the Stephen Lake and Fog Lakes
areas) shows that the age of surfaces overlying the Talkeetna
Thrust fault and Susitna feature south of the Susitna River are
predominantly of Late Wisconsin age (Figure 3-4). This chronology
3 -12
provided the basis for interpreting the age of the higher surfaces
along the southwestern section of the Talkeetna thrust fault to be
Early Wisconsin and pre-Wisconsin in age (Figure 3-4).
3.4.3-The Butte Lake Area
This area lies at the north end of the Talkeetna Mountains and is
separated from the Alaska Range to the north by a broad lowland
called Monahan Flat (Figures 3-2 and A-1). The area includes a
northeast-southwest trending linear valley, containing Butte Lake,
that opens into the Monahan Flat lowland (as discussed in Section
A.1.4.5). Within the broad, U-shaped valley, drainage patterns and
directions have been altered by glacial erosion and deposition.
The valley bottom is predominantly mantled by till while the upper
valley walls are mantled by frost-shattered boulder fields.
Late Wisconsin glaciation resulted in glacial ice moving through
the Butte Lake valley up to a maximum elevation of 3,900 feet
(1,189 m). The elevations of Late Wisconsin end moraines suggest
that as many as nine individual moraines may be present. The
clustering of these moraines, the breaks in slope, and the surface
morphologic contrasts led to the identification of four Late
Wisconsin stades (ice pulses of glacial advance within the Late
Wisconsin glacial stage). These stades appear to be similar in age
and duration with those observed by Ten Brink and Waythomas (in
press) in the Alaska Range.
The maximum elevation of these four stades are: Stade I is 3,900
feet (1,189 m); Stade II is 3,600 to 3,800 feet (1,098 to 1,159 m);
Stade III is 3,200 to 3,300 feet (976 to 1,006 m); and Stade IV
is 3,000 to 3,100 feet (915 to 945 m). Stades I and IV are shown
in Figures 3-2 and 3-6A; Stades II and III are not delineated
because of the limited control on their extent.
3 -13
The Late Wisconsin ice in the Butte Lake valley is of pre-last
stade, i.e., older than 11,000 years before present. The Stade IV
glacial ice from the Alaska Range was of insufficient thickness
to move into the Butte Lake valley from Monahan Flat, and the
glaciers from the area to the west did not move into the valley.
These studies show that the age of the surfaces overlying the
Susitna feature in the northern Talkeetna Mountains is Late
Wisconsin in age.
3.4.4-The Deadman Creek Area
This area lies north-northeast of the Watana site as shown in
Figure 3-2. Deadman Creek flows southwestward in the intermountain
basin at the base of the northwest section of the Talkeetna
Mountains, as discussed in Section A.1.4.4. This part of the basin
floor is almost entirely mantled by hummocky ice disintegration
deposits and lacustrine plains (Figure 3-68). Intervening areas
are ground moraine or beveled bedrock outcrops.
Frost-shattered boulder fields above 4,100 feet (1,250 m) suggest a
maximum elevation for Wisconsinian glaciations. The maximum
elevation for Early Wisconsin moraines is 4,100 feet (1,250 m).
The age of these moraines is based on relative age dating results
(Tables 3-3 and A-1) as well as on their elevation. These deposits
are interpreted to be the product of coalescing valley glaciers
which merged with glacial ice emanating from the Alaska Range to
produce a piedmont glacier in the intermountain basin.
Late Wisconsin glaciation resulted in a sequence of closely spaced
end moraines along the base of the mountains to the west of
Deadman Creek (Figure 3-6B). In early Late Wisconsin time, the
glacial ice responsible for these deposits was probably a piedmont
3 -14
glacier similar in nature to that of Early Wisconsin glaciers.
During later stades of the Late Wisconsin stage, local valley
glaciers moved from the high region in the northwestern part of the
Talkeetna Mountains down into the Deadman Creek area and flowed
to the northeast. The evidence for this is the northward slopes on
moraines DC-1 and DC-2 (Figure 3-6B) and an arcuate moraine damming
Big Lak (immediately east of the Deadman Creek area) that is
concave southward.
During the last stade of the Late Wisconsin stage, a valley glacier
moved out of Tsuena Creek (Figure 3-2) and into the southwestern
portion of the Deadman Creek area. Subsequent stagnation of the
ice produced hummocky ice disintegration deposits which locally
dammed a 1 a k e, rep res en t e d by the 1 a c u s t r i n e de p o s its s how n
in Figure 3-6B. Radiocarbon age dating of a sample (545-1) from
these lake deposits gives an age of 3,450 ~170 y.b.p. (Table 3-2).
This age date tends to confirm relative youthfulness of the lake
and the existence of the dam until late Holocene time.
The glacial chronology of this area, as well as of the Butte Lake
area, shows that the age of the surf aces overlying the Susitna
feature north of the Susitna River are Late Wisconsin in age
(Figure 3-4). In addition, the chronology of this area and the
Clear Valley area, along with radiocarbon age dates and morphologic
characteristics in the Watana Creek area, show that the surfaces
overlying the Talkeetna thrust fault are also predominantly Late
Wisconsin in age (Figure 3-4).
3.5 -Glacial History and Distribution of Quaternary Surfaces
The data obtained from the key areas described above, along with
observations in the intervening sections of the Quaternary study region,
3 -15
radiocarbon age dates (shown in Table 3-2 and Figure 3-2) were used
develop an understanding of the glacial history and the distribution
the resulting late Quaternary glacial surfaces. The following
summarizes this understanding.
E'1!"!idence was found for four distinct Quaternary glacial episodes: pre-
::;consin, >100,000 y.b.p.; Early Wisconsin, 75,000 to 40,000 y.b.p.;
LJfte Wisconsin, 25,000 to 9,000 y.b.p.; and Holocene, <9,000 y.b.p.
iE.~i!:;:h glacial episode was less extensive than the preceding one. The
'ilt!i~Jits of each are defined by the elevation and geographic distribution
~-;~f glacial erosional or depositional features. Glaciers advanced
repeatedly from three main source areas: the Alaska Range to the north;
'!i:.h•i:~ southern and southeastern Talkeetna Mountains; and the Talkeetna
t-·~~~):umtains north and northwest of the Susitna River. Glacial flow was
iJi.r,;:;minantly to the south and southwest, following the regional slope and
~;.tructural grain. Multi-directional and convergent flow, differing
~!.!'l·~Jcial magnitudes, topographic influences, and other parameters make
th~ interpretation of the glacial chronology of the Quaternary study
region difficult.
Pt1;:~-Wisconsin glaciated surfaces are present at elevations above the
:::.:u~j!gested upper limit of Early Wisconsin glaciation. Bedrock scour
.:~:~"!d ice-sculptured forms (such as stoss and lee, and whalebacks)
d•r,::rni nate the topography. The geographic extent and e lev at ion of these
:~.u:rfaces suggest that ice cap conditions probably existed throughout
ti'lif::• Talkeetna Mountains. Broad areas of the plateau south of Devi 1
{;.~1if"!:yon were overridden by glacial ice. Periglacial processes during
lf:lter glaciations produced extensive veneers of colluvium and frost-
shattered boulder.fields .
. l:i, tentative age was assigned to the glaciated surf aces above the 1 imit
the Early Wisconsin glaciation on the basis of elevation and degree
•!:J.·f weathering. The data co 11 ected within the Quaternary study region
3 -16
only that these surfaces are older than the Early Wisconsin stage
the prior interglacial stage (120,000 to 75,000 y.b.p.), although
e surfaces are probably Illinoian in age (>120,000 y.b.p. ). Worldwide
::.d::udies of the Illinoian glacial stage show that its youngest age
variable (Flint, 1971). However, it is generally accepted to
>120,000 y.b.p. For purposes of this study we have accepted an age
>100,000 y.b.p. to be appropriate.
E.c~tly Wisconsin glaciation was less extensive than glaciation during the
pr1:~·ceding glacial period, and ice was present in existing valleys at
l!r,:i\!~~er elevations than the pre-Wisconsin ice was. Regional evidence for
ttl!!:~ maximum upper extent is limited. Prominent ice-marginal features,
~J!o!.tticularly trimlines and weathering contrasts, indicate the upper
l1nilit to be: 3,750 feet (1,143 m) in the Black River area; 3,100 feet
U~4~5 m) in the Clear Valley area; 4,200 feet (1,280 m) in the Butte
area; 4,100 feet (1,250 m) in the Deadman Creek area; and 3,100
(945 m) in the Devils Canyon area.
{;;lr.lciers from the Alaska Range and Talkeetna Mountain ice sources,
r'Ji•!:)•:>cribed above, coalesced to form a piedmont glacier in the inter-
~]l.ri,ij.mtain basin of the Susitna River (near the Watana site), Fog Lakes,
m'i!d Stephan Lake; but large areas of the upland plateaus were ice-free.
;)uccessive Early Wisconsin moraines in the Clear Valley area indicate
U:1.at several stades or recessional stillstands took place in Early
'i,i1ii::>consin time. Ice flow that emanated from the Alaska Range passed
~;tnJthwestward to the Talkeetna River and westward and southwestward
through Devils Canyon to merge with glaciers in the Chulitna Valley.
TIH~ Late Wisconsin glaciation was composed of four distinct glacial
.:~.des. The maximum limit of Late Wisconsin ice was 2,700 feet (823 m)
iin the Clear Valley area; 3,900 feet (1,189 m) in the Butte Lake area;
,J:md 3,900 feet (1,189 m) in the Deadman Creek area.
3 -17
Woodward-Clyde ConsuDtants
The ice during each stade was less extensive than that during the
preceding one. Glaciers in the first stade of Late Wisconsin time had a
geographic extent similar to prior glaciations. During later stades of
the Late Wisconsin stage, ice from the Alaska Range was not thick enough
to advance into valley passes in Brushkana and Deadman Creeks, but it
was thick enough to flow southward through the pass between Butte Creek
and Watana Creek. Valley glaciers in Tsusena Creek and adjacent valley
glaciers advanced into ice-free areas of the intermountain basin.
During Stades III and IV, individual valley glaciers were generally
confined to valleys and the piedmont glacier had retreated north of the
Susitna River. Ice from the Alaska Range continued to retreat northward
up Watana Creek and supplied abundant sediment for a lacustrine environ-
ment in lower Watana Creek. Rapid regional deglaciation produced
extensive hummocky ice disintegration deposits in topographic low
areas. Widespread glaciation came to an end approximately 9,000 y.b.p.
The Susitna and Talkeetna Rivers served as outlet channels for meltwater
from the retreating glaciers, producing extensive terraced outwash
gravels in the lower reaches of the rivers.
Glaciation of Holocene age is limited to cirque glaciers in the upper
reaches of valleys and to the formation of glaciolacustrine lakes
in lowland areas. The small cirque glaciers formed moraines within
a few miles of the cirques. Rock glaciers in the upper reaches of
mountain valleys are the most visible remnant of the Holocene glacia-
tion. Glaciolacustrine lakes, such as those that formed in the Clear
Valley and Deadman Creek areas (Figures 3-5B and 3-68) were dammed by
glacial units deposited by the Late Wisconsin Stade IV glacial ice.
These lakes have breached their dams and are now represented by plains
of lacustrine deposits.
3 -18
-Quaternary Geology and Significant Features
F\;tur features (faults and lineaments) were studied near the Watana site
during this investigation to assess their significance to seismic
dit;::•sign, as discussed in Section 4.4.1. The late Quaternary surfaces
that overlie these four features are predominantly Late Wisconsin
(25,000 to 9,000 y.b.p.) in age, except in the vicinity of the Talkeetna
River where pre-Wisconsin (>100,000 y.b.p.) surfaces are present
;J:Figure 3-4). For example, the Talkeetna Thrust fault from Denali to
tile Talkeetna River is overlain by surfaces of Late Wisconsin age along
t:!:!5· percent of its length. The remaining 15 percent is Early Wisconsin
and pre-Wisconsin in age, as shown in Figure 3-4.
r,,!;ine features were studied near the Devil Canyon site, as discussed
i:n!; Section 4.4.2. The Quaternary surfaces that overlie these nine
f~~atures are generally older than those encountered in the vicinity
of' the Watana site. These surfaces are of Early Wisconsin age or older
,[::1·40,000 y.b.p. ), except in the vicinity of the Susitna River and its
tributaries where Late Wisconsin surfaces (25,000 to 9,000 y.b.p.)
present (Figure 3-4).
'ii'he age of the Quaternary surfaces, the size of morphologic features
that have been preserved on each of these surf aces, and the extent of
tii"!e surfaces were evaluated and incorporated in the analysis of recent
I'·!Wlt displacement (as discussed in Section 4.2) for the four features
near the Watana site and the nine features near the Devil Canyon site.
The Quaternary field study and the subsequent data analysis showed no
E;:vidence of displacement in the Quaternary surfaces overlying the the
features (as discussed in Section 4.4).
Hie evaluation of Quaternary surfaces also focused on the resolution
p·r·ovided by these surfaces. This resolution was used to analyze the
rn::;1ximum earthquake (designated the detection level earthquake) and
3 -19
!'"•i:~sultant fault displacement that could occur in the site region and
still maintain the Quaternary surfaces in their observed undeformed
~~tate. This analysis of the detection level earthquake is discussed in
s.,;:~ct ion 4. 2 .
3 -20
SUMMARY OF QUATERNARY GLACIOGENIC FEATURE CHARACTERISTICS
s la9es
narrow
and
to subparallel,
:,,..,ales . ..,ith
low
MORPHOLOGIC CHARACTERISTICS
1. Broad undulating plains, generally confined
to U-shaped valleys
2. Forms blanket deposits
3. Fluting co!lll10n
4. Elongated~ narrow ridges built along margins
of glaciers, comnonly paired
1. Nearly flat featureless plains comnonly in
topographic lows
2. Slumping conmon along mar-gins of dissected
areas
3. No fluting
4. Drainage pattern corrrnonly contrasts with
patterns on adjacent sur-faces
5. Uniform vegetation type and density
l. Valley trains of fluvial sedim€nts
deposited on braided floodplains headed
at the terminus of glaciers
1. Random assemblage of huffiTKlcks, ridges~
basins, and small plateaus
2. Slopes vary, but many are steep
3. Generally confined to distal end of
glacier deposits
4. Slumping corrrnon
5. Generally limited to topographic lows
6. Kame and kettle, and knob and kettle
topography
Narrow~ flat surfaced, constructional
terrace along hillsides
2. Slopes down valley
3. May be discontinuous along valley sides
4. Surface may be pitted by kettles
5. Carrmonly no apparent source area for
sediments
6. Parallels direction of glacial flow
l. Confined to floodplains of active major
streams and rivers
2. Broad to narrow fiat floodplains
3. Corrrnon ly terraced
4. Possible meander scars
5. Stream bars corrroon
l. 'arge scale forms developed corrrnonly on
broad till plains
2. Contrast of vegetation type and density
3. Linear parallel to subparallel broad
topographic swales with intervening
well-rounded low ridges ..
2.
3.
4.
L
2.
3.
4.
5.
6.
7.
Abraided rock or trirrmed-off vegetation line
at former ice contact, separates surface
materials of differing reflectivity
Corrrnon ly paired on opposite va 11 ey walls
Slopes downvalley with same gradient as
va 11ey floor
Slight breaks in slope to nearly vertical
planar cliffs
Channel formed between ;nargin of glacier and
adjacent valley wall
Oiscont inuous elongated, narrow stream
channel
Typically oblique to hillside
Open at both ends
Parallels direction of glacial flow
Channel that leaves ice margin cutting notch
across ridge 1 i ne ca 11 ed overflow ch anne 1
Channels that have no apparent source area
for format ion but may now be accentuated
and oodified by hills ide runoff
1. Polished, srrooth bedrock outcrops
2. Strong preferred orientation of long
axis of forms and grooves
3. Rock structural features may be
differentially eroded
4. Forms with lengths greater than widths
1. Unsorted gravel and sand in matrix of silt
and clay, unconsolidated, nonstratified
l. Rhythmically beddea silt and/or clay with
occasional ice rafted sand, pebbles, and
grave 1
2. Forms vertical cliff faces, but slumping
is corrmon
3. Grain size variations preclude loess origin
4. Sorting and stratification are corrmon in
areas of deltaic sedimentation
1. Well-sorted and stratified rounded sands
and gravels
2. Cut and fi l 1 channels are corrmon in
cross sect ion
3. Surface relief is less than approximately
10 ft (3m), unless dissected or terraced
by subsequent fluvial processes
1. Chaotic and irregular nature of morphology
2. Kettles corrmon
3. Gravel, sand, silt, and clay; grain size,
degree sorting and stratification (structure)
of sediments varies with the aroount of
water reworking
4. Highly deformed structure caused by slumping,
settlement, and compaction are comnon
1. Stratified and sorted rounded sands and gravels
2. Hillside terrace may be discontinuous and
surface may contain kettles
3. Very narrow compared to fluvial terr·aces
4. Generally many hundreds of feet above outwash
and fluvial terraces, but if long enough, can
merge oJ~ith outwash terraces near terminus of
glaciers
1. Well-sorted and stratified rounded sands and
gravels from re't't'Orked glacial deposits and
eroded bedrock
2. Cut and fi 11 channels are corrmon in cross-
section
3. Surface relief is minimal unless terraced
l. DifficUlt to identify on ground because of scale
2. Relief of only a few lO's of feet (3m) over
large horizontal distances
3. Drainages differences reflected by vegetation
contrasts
1. Vegetation contrast or truncation and/or
contact between bedrock and unconsolidated
talus
2. Change in slope grade
l. Channels contain little or no sediment
2. Downhill side is lower than uphill side
3. Discontinuous
4. Locally open, hanging channel ends
5. No source areas for format ion
6. Cut into rock with corrrnon1y no outside wall
7. Rock channel walls have steep slopes
l. Polished curved bedrock outcrops
2. Gentle slopes upstream {stoss) and steep slopes
downstreom (lee)
3. Soooth rock outcrops may record small-scale
features of glacial abrasion, stri{l.tions,
po 1 ish, grooves
1. Geographic (spatial) position in relation to
other glacial deposits
2. Topographic (elevational) relationship to
adjacent clepos its
3. Elevation of end rroraines
4. Degree of dissection and surface rrodifi-
cat ion
5. Orientation and cross cutting relationships
of end roorai nes
1. Geographic (spatial} position in relation to
other glacial deposits
2. Topographic {e1evationa1) relationship to
adjacent deposits
3. Degree of stream dissection
4. Degree of surface rrodificat ion
1. Geographic (spatial) position in relation to
other glacial deposits
2. Topographic (elevational) relationship to
adjacent deposits
3. Abandonment or elevation of outwash surface
above present base leve 1
4. Degree of rrodification to surface rrorphology
5. Degree of local stream dissect ion
1. Geographic (spatial) position in relation to
other glacial deposits
2. Freshness of roorpho logy
3. Degree of slurrping, infilling, and rounding
4. Degree of stream dissection
1. Elevation above valley floor
2. Degree of erosional segmentation
3. Degree of surface modification and
dissect ion
4. Geographic position in relation to other
glacial deposits
1. Deposits confined to recent age but terraces
adjacent to active floodplain may be
assigned age based on elevation above
active floodplain
2. Degree of modification to terrace surface
and scarp
3. Degree of vegetation growth
l. Age of fluting scrre as age of till on which
it is developed
2. Orientation and cross-cutting relationship
of 1 i near forms
l. Elevation
2. Sharpness of 1 i ne
l. Elevation of channel
l. Orientation and cross-cutting relationships
of longitudinal axis
2. ~weather texture of rock surf ace
3. Degree of form rrodlfication
1. Various quantitative soil
morphologic rreasure;rents.
(weathering) data Appendix
Tables 3-3 and A-1
l. Soi 1 profile development
l. Soi 1 profile development
2. Degree of cementation
3. Degree of oxidation
4. Degree of vegetation cover
1. Soil profile development
2. Local relief
3. Degree of vegetation cover
age
4. Degree of roundness, infilling, and sluf'!'\)ing
infilling, and slu~ing
l. Soil profile development
2. Degree of infilling of kettles
3. Oxidation depth
4. Modification of form by erosion (dissection
or burial)
l. Soil profile development
2. Surface elevation
J. Depth of oxidation
l. See criteria for till
1. Elevation
2. Degree of rock surface weathering
1. Degree of weathering of channel rock
2. Degree of channel infilling by post-
erosional sediments
l. Degree of rock surface weathering
2. Degree of form rrodification by weathering
3. Preservation of glacial polish and small-
scale abrasion features
RADIOCARBON AGE DATES AND SAMPLE DESCRIPTIONS
Age date
ci4 years
before
_present
2245 ~ 140
2290 ~ 130
3450 + 170
3540 + 160
9395 + 200
9920 + 265
> 27 ,000
> 37 ,000
> 37,000
> 37 ,000
> 37,000
Note:
Field
sample
number
S49-l
S47-1
545-1
S4-1
S42-l
S54-1
S34-1
S29-l
Sl2-2
547-4
S62-1
Sample Stratum
Frozen silty sand
Frozen lacustrine
silt with striated
ice raft cobbles
Frozen lacustrine
silt and clay
Frozen lacustrine
silt and clay
Frozen lacustrine
s i lt and c l ay
Fine-grained ice
disintegration
deposits
Sand matrix
surrounding angular
bedrock blocks,
colluvial (?)deposits
Interbedded lacustrine
sand and silt from
ice marginal lake,
12 ft ( 3. 7 m) bel ow
outwash deposit
Lacustrine/deltaic fine
sand 51 ft (15.5 m)
above till
Contact of lacustrine
fine sand with inter-
fingered till
Ice marginal lacustrine
silt and sand
1. Site locations are shown in Figure 3-2
r~aterial
dated
Wood
chips
Wood
\~ood
Charcoal
Wood
Peat
Charcoal
Wood
1-Jood
chips
1-Jood
Wood
Site
locationl
Measured section
along Watana
Creek
Measured section
at Oshetna
River mouth
Slump exposure
along Deadman
Creek
Slump exposure
at mouth of
Clear Valley
Slump exposure
in upper Watana
Creek valley
Slump exposure
along upper
Deadman Creek
Measured section
along drainage
north of Stephan
Lake
Measured section
along Moraine
Creek
Measured section
along Brushkana
Creek
Measured section
at Oshetna
River mouth
Measured section
west of Danek a
Lake
Quadrangle
Talk. t~tn.
( D-3)
Talk. Mtn.
(C-1)
Talk. Mtn.
(D-3)
Talk. Mtn.
(C-4)
Talk. Mtn.
( D-2)
Healy
(A-3)
Talk. Mtn.
(C-4)
Talk. Mtn.
( D-4)
Healy
(A-3)
Talk. Mtn.
( C-1)
Tc.lk. Mtn.
( C-5)
Section
township
& range
NEl/4; NWl/4;
Sec. 9;
T32N; R7 E
NEl/4; NEl/4;
Sec. 4;
T29 N; Rll E
SWl/4; SEl/4;
Sec. 32;
T225; R4E
NI>Jl/4; NWl/4;
Sec. 34;
T30N; RSE
S~~l/ 4; NWl/ 4;
Sec. 21;
T22S; R2W
SWl/4; SEl/4;
Sec. 15;
T215; R4W
NEl/4; SEl/4;
Sec. 29;
T31N; R4E
St>Jl/4; NWl/4;
Sec. 24;
T32N; R4E
NEl/4; SWl/4;
Sec. 8;
T20S; R3W
NEl/4; NEl/4;
Sec. 4;
T29N; RllE
NEl/4; NWl/4;
Sec. 22;
T30N; R2E
Lab Sample
Number
GX-8056
GX-8055
GX-8059
GX-8054
GX-8035
GX-8062
GX-8060
GX-8034
GX-8057
GX-8058
GX-8124
Significance
of date
Dates young deposits which slumped into
older deposits and were subsequently
frozen
Anomalous date, may represent the age of
wood that was incorporated into lacustrine
deposits by cryoturbation
Minimum date of last retreat of
ice from valley of Deadman Creek
Maximum age of permafrost formation (i.e.,
the ground froze after deposition of
the charcoal); also represents time of
lacustrine deposition near the north of
Clear Valley
Dates the time of last retreat of ice from
Watana Creek valley
Dates the time of last retreat of ice from
valley of Deadman Creek
Maximum date on oxidized outwash overlying
the sampled stratum
Maximum date on oxidized outwash and till
blanket overlying the sampled stratum
Minimum date on till buried by the
sampled stratum
Dates advance of ice into i ce-margi na l
lake
Dates ice marginal deposits of the Early
Wisconsin stage; limits the maximum elevation
of Late Wisconsin glacial ice in the
intermountain basin
TABLE 3-3
SUt~MARY OF RELATIVE AGE DATA FOR EARLY AND LATE WISCONSIN MORAINES
Area 1 Subsurface
Surf ace
Moraine Morphology Degree of Side Average Granite
Form Mod if i cat i on2 Segmentation S l OQ es Weathering Ratio4
(%F/%PW/%W)
Clear Valley Fresh, prominent, Slight Slight Moderate
moderate crest width to steep
Butte Lake Fresh to weathered, None to Slight Gentle 56/34/10
prominent to sub-moderate to steep
dued, narrow to
broad crest width
Deadman Creek Fresh, prominent, None to Very Steep 50/39/11
moderate crest slight slight
width
Black River Weathered, subdued, Moderate High Gentle to 52/40/8
broad crest width to high moderate
Clear Valley Weathered, subdued, Moderate High Gentle to
broad to moderate to high moderate
crest vJidth
Butte Lake \fJeathered, prom-Moderate Moderate Moderate 30/46/24
i nent, moderate
crest width
Deadman Creek Fresh, prominent, Slight slight Moderate 40/40/20
moderate crest width to steep
Black River Fresh, prominent, None to Very Steep 42/32/26
narrow crest width s l i gh t slight
Notes:
1. Areas are shown in Figures 3-5 and 3-6.
2. Surface morphology modification reflects the effect of slumping and frost heaving.
3. The subsurface data summarized here are presented for each moraine in which measurements were made and
cited in Table A-1.
4. Granite weathering ratios are discussed in Appendix A.l.2.2. The numbers (e.g., 56/40/10) are the percentage
of granite boulders that are fresh (%F), partially weathered (%PW), and weathered (%W), respectively.
Data3
Oxidation Depth
inches (em)
11 to 24 (3 to 7)
8 to 12 (2 to 4)
7 to 12 (2 to 4)
11 to 17 (3 to 5)
26 to 40 (8 to 12)
14 to 29 (4 to 9)
17 to 18 (5 to 6)
21 to 22 (6 to 7)
TABLE 3-3
.
BSOLUTE AGE ~years Before Present l
ERA PERIOD EPOCH STAGE STADES -HOLOCENE -9,000-
IV
-11 ,000---
Ill
-13.500--f-
>-w
0: z La te II (..) <( w Wiscon sin -(..) -15,000-0 z -f-N a: 0
0 w I-
C/) z I--Wisconsin ian w <( w Glaciation -17,000-(..) :::> ..J -f-
0 a..
I
-25,000 --f-
I nterglaci'al
-40,000--f-+ -f.-
Early Wi sco nsin
-75,000--f-t -f.-
Sangamon
lnterglaciation --f--
-120,000---f-+ Pre -Wisconsin -f-
j Illinoian
Glaciation
!
z z z UJ
V5 -z <t Vl UJ -z z u 0 0 0 0 z u u _J -Vl Vl _J 0
_J 3 3 ::r -
170 1i0 715 u.i 4p _j 1 ·J
BAR SCALE
(In Thousands of Years Before Present)
NOTES
1. Era through Epoch terminology and absolute ages are after Van Eysing a (1978).
2. Stage terminology and ages ar e after Pewe' (1975).
3. Stade ages are modified aft!lr Ten Brink and 'Waythomas (in press).
QUATERNARY STUDY REGION
TIME SCALE
'--.
WOODWARD-CLYDE CONSULTANTS 41410A February 1982 FIGURE 3-1
LEGEND
CJ
D
D
S4 -1e
WI
Dl
NOTES
L ate Wiscons i n surfaces 11,000 to 9,0 0 0 y .b .p.
L at e Wi sc on sin su rf ac es 25,000 t o
1 1,000 y .b .p.
Early Wis co nsi n surfaces 75,000 t o
40,000 y .b.p.
Pr e-Wiscon sin sur f aces >100,000 y .b .p .
Radiocarbon sa mple locality and number
Glacial Age Boundary
Detailed study ar eas are shown in
Figures 3 -5 and 3-6 .
Generalized projected cross section is
shown in Figu re 3 -3 .
Watana Site
D ev il Canyon Site
1 . Figure A-1 shows the locati on of the Quate r na r y
Study Region .
2 . Areas with no color are bedrock and/or surfaces of
undifferentiated glacial age .
3 . y.b .p . is the abbrev i ation for years before present
4. Glacial age boundaries ar e interpreted from morpho-
stratigraph i c relat i onships and age dates .
~
-N-
~
QUATERNARY GEOLOGY MAP
0 10 20 Miles E:::::=EQk3~=1~~1====:3
0 10 20 Kilometers
F I GURE 3 -2
LEGEND
-Till
o -Glaciofluvial Deposits
S34
-Deltaic Deposits
-Ice Contact Deposits
-Undifferentiated Bedrock
Field Location Designation
Unconformity, queried
where inferred
Contact, queried
where inferred
Section location and
designation
NOTE:
1. Section location is shown on Figure 3-2.
0
E:-:3
Vertical Exaggeration 26x
SCALE
5
lf¢t": Hfj¢1
10 A .A .ol 44
50
• • • Till, gray
20 • "
o A 4
10 Miles
I
0 5 10 Kilometers
WOODW ARD-CLYDE CONSULTANTS 41410A February 1982
Maximum elevation of Early Wisconsin Ice
Tsusena
Creek
Maximum elevation of Late Wisconsin Ice
Deadman
Creek
Exposures sec 23 I
& 26, T32N, R5E
Delusion
Creek
Exposures I
sec 20 & 30,
T32N, R5E
-~u('(l
1'}\a')(.'"'
elevation of last stade of Late
Thickness
(ft) (m)
0 0"
10 .0. A A 4
6
Del ta i c, tan
sand, terraced
50 • • •• • Till, gray
100
4 •• ..
20" ".
" .
30
• 0 -. .
. -.. ..
40 . . ... ..
Glaciofluvial,
reddi sh, highly
oxidized,
moderately to
highly
consolidated
Lacustrine,
olive gray,
si lt and clay,
Sample S29-1
50
75+
Till, gray,
slumping
common
Measured section S14
N EY. SWY. sec 7,
T32N, R7E
dated >37,000 y.b.p.
Measured section S29
SWY. NWY. sec 24,
T32N, R4E
Thickness
(ft) (m) D
0 0
50
100
150
199
30:: ~-:=::: Lacustrine, tan
:__:-:__-:.:: to gray, silt with
:__:-~::-.:: fine sand beds,
:_::....:..: varves common
40 :..:.---~
Glaciofluvial,
oxidized
Measured section S44
SWY. NWY. sec 9,
T32N, R7E
D
Tributary
to
Watana Creek Lacustrine deposits
Samp l e S42-1 dated
9395+200 y.b.p.
4400
f
---------------4200
~~~\
Maximum elevation
last stade of Late
..;
/
I
-/ /'
I
• I,','"'
4000
bA: \'~\
"•I 3800
., I
• I
• 3600 :1 I
• •f I
• ··"· 3400 • I I ·"I 32oo
I I j·
I
"/ 2800
I
2600
2400
.... ....
c
0 2200 .... ro >
Q) w 2000 ,_..~--?.A-..___...__?----.A-
Thickness E
(ft) (m)
0 0
10 :__~·::.::.
. o----
50
30 :-_:o_:.-;:_ 100 :-..:.~:.~ Lacustrine, tan,
:-..:.:-_:_o~ random ice
~-:::::;;:rafted cobbles
40~-~~
150
60:=·.o_:_-:-200 -·0·-·
gray
245
Measured section S48
NWY. NWY. sec 9,
T32N, R7E
Pre Late Wisconsin
glacial deposits
Thickness F
(ft) (m)
0 0 •••••
50
100
;'1)::
10 :?.'::: Deltaic silt to :; '!); coarse ~and, ; :!. ~ graded and
••••• cross bedded,
20 ~i~· ripples common
==~ 30 • • '.j/ Lacustrine, silt
~~:-..: and clay, varves
....... ~~,AJ:J. common ...
40 ·<4: 4
·.·:Till, gray
150 •••
175
' .
5Q a. ... ,.·
River level
Measured section S15
SEY. SEY. sec 27,
T33N, R3W
Measured section S16
SWY. SEY. sec 15,
T33N, R3W
1800
1600
1400
GENERALIZED CROSS SECTION OF
QUATERNARY DEPOSITS AND SURFACES
FIGURE 3-3
LEGEND
---
KD5-12 •. / ........
WI
D I
NOTES
Late Wisconsin surfaces 11,000 to 9,000 y.b.p
Late Wisconsin surfaces 25,000 to
11,000 y .b .p.
Early Wisconsin surfaces 75,000 to
40,000 y .b .p .
Pre-Wisconsin surfaces >100,000 y .b.p.
Glacial age boundary
Thrust fault, dotted where concealed,
sawteeth on upper plate
Fault and code number
Concealed shear zone or fault
Lineament and code number
Watana Site
Devil Canyon Site
1. Figure A-1 shows the location of the Quaternary
Study Region .
2 . Areas with no color are bedrock and/or surfaces of
undifferentiated glacial age .
3. y.b .p . is the abbreviation for years before present.
4. Glacial age boundaries are interpreted from morpho-
stratigraphic relationsh ips and age dates.
~
-N-
~
SIGNIFICANT FEATURES AND
QUATERNARY SURFACE MAP
0 10 20 Miles ~=EWi3~~~~~1===:=3
0 10 2 0 Kilometers
FIGURE 3 -4
LEGEND
UNITS •
[]
AGE
Lacust r ine sediments D Holocene surfaces
<9 ,000 y.b.p.
Ice d i ~integration D Late Wiscon sin surfaces
deposits
Till
Bedrock
S4·1·
BR·1·
11,000 to 9 ,000 y .b.p.
D Late Wisconsin surfaces
25,000 to 11,000 y .b.p.
D Early Wisconsin surfaces
75,000 to 40,000 y .b .p.
D Pre-Wisconsin surfaces
>100,000 y.b.p.
Moraines
Direction of meltwater flow and length of
side glacial or overflow channels
Radiocarbon sample locality and number
Relat i ve dati ng locality and number
A . BLACK RIVER AREA
't'iOODWARD-CLYDE CONSULTANTS 41410A Febru ary 1982
NOTES
1. Regional location is shown on Figure 3-2.
2. y.b.p. is the abbreviation for years before present.
B. CLEAR VALLEY AREA
-N -
~
QUATERNARY GEOLOGY OF THE
BLACK RIVER AND CLEAR VALLEY AREAS
OE===E~~==o3.~5~~~~3===:31 M i le
0 0.5 1 Kilometer
Fl URE 3-5
A. BUTTE L A K E AR EA
LE G END
U NITS
II -. . -
Err~
CJ
Lacustrine sediments
Ice disintegration
deposits
Till
Glaciofluvial
sediments
Bedrock
I'IO ODWARD-CLYDE CONSULTANTS 41410A
AGE
D
D
D
D
D
February 1982
Holocene surfaces
<9,000 y .b.p .
Late Wisconsin surfaces
11,000 to 9,000 y .b .p.
Late Wisconsin surfaces
25,000 to 11,000 y.b.p.
Early Wisconsin surfaces
75,000 to 40,000 y.b.p.
·Pre-Wisconsin surfaces
>100,000 y .b.p.
S4 5·1 ·
DC·5 ·
B. DEADMAN CREE K AREA
Moraines
Direction of meltwater flow and length of
side glacial or o verflow channels
Radiocarbon sample l ocality and number
Re l ative dating locality and number
NOTES
1. Regional location is shown on Figure 3-2.
2 . y .b.p. is the abbreviation for years before present .
--N ·
~
QUATERNARY GEOLOGY OF THE
BUTTE LAKE AND DEADMAN CREEK AREAS
0
0 0 .5 1 Kilometer
FIGURE 3-6
Woodward-Clyde Consultants
4 -SIGNIFICANT FEATURES
4.1 -Introduction
Prior to evaluation of the faults and lineaments in the project region,
a tectonic model of the region was developed as discussed in Section 1.4
and the Interim Report (Woodward-Clyde Consultants, 1980b). This model
provided a conceptual framework in which the likelihood of recent fault
displacement and seismic activity could be evaluated. The tectonic
model applies to a region of the earth's crust that we have called the
Talkeetna Terrain. This terrain is bounded by the Denali and Totschunda
faults on the north and east, the Castle Mountain fault on the south, a
broad zone of deformation and volcanoes to the west, and the Benioff
zone and base of the crust at depth (Figure 4-1).
During the 1980 study, 216 features were studied by reconnaissance from
helicopters and fixed-wing aircraft and by ground mapping at selected
locations. At the conclusion of the 1980 study, the Talkeetna Terrain
boundary faults were identified as being faults with recent displacement
that should be considered in design. In addition, 13 significant
features closer to the dam sites were selected for additional study on
the basis of their potential affect on ground motion and surface
rupture considerations (Woodward-Clyde Consultants, 1980b).
During the 1981 study, the boundary faults were reviewed to refine
estimates of the maximum credible earthquakes (MCEs). Two of them. the
Castle Mountain and Denali faults, are discussed in detail in Section
4.3; the two regions of the Benioff zone are discussed in detail in
Section 5.2 and are summarized in Section 4.3.
The 13 features selected for additional study were the subject of
detailed field studies in 1981 using the methodology described in
4 - 1
Woodward-Clyde Consultants
dix A. The approach used to guide the field studies was twofold:
the bedrock along each feature to assess whether or not
e: feature was a fault; and 2) to examine the surficial units along the
e:.!'!!ture to evaluate the geologic evidence for recency of displacement.
·!!:f.; section on significant features presents our evaluation of faults
ilifl,i:J lineaments within and bounding the Talkeetna Terrain. Section 4.2
:i·'Ul1'!'!i!inarizes the detectability of faults with recent displacement.
;::;.,r;~c;:tions 4.3 and 4.4 discuss our interpretation of fault activity for
t~'i11;:: boundary faults and the features in the vicinity of the Watana and
:Ufndl Canyon sites. Our evaluation is summarized in Section 4.5.
·~L~: -Detectabi 1 ity of Faults With Recent Displacement
Th!f:!' detectabi lity of faults with recent displacement depends primarily
,m·:1 the following factors: 1) the age of the sediments overlying the
1fiHJU; 2) the amount of displacement at the surface during earthquakes;
;~) the recurrence interval of these earthquakes (i.e., how often do the
~;~;:trthquakes and displacements occur); 4) the type of displacement that
~~~c:r::,urs; 5) the length of the fault along which displacement occurs; and
c~\) the length of time that displaced features are preserved (i.e.,
~·\c~[Yidity of erosion and despositional processes relative to the rate of
~"'''' ·~:fi'i 1 a cement) ,,., ! '·''ii'' •
F,:,1.iL;!lts that generate earthquakes that are too small to rupture the
'f:!.!,,i!'!'''W'ace or to cause fault scarps large enough to be preserved may not
h'·~!t''i!B been detected by our geologic investigation. Consequently, an
~:::~:;.t'imate was made of the size of earthquake that might have occurred
!,\!'1!t:hout leaving any detectable geologic evidence. This earthquake was
!:;!!f::":;.·>ignated the "detection level earthquake."
·r\) .address this question, we conducted a three-step evaluation during
ti·n:: 1981 study. These steps included: 1) a review of a select group
4 - 2
Woodward-Clyde ConsuDtants
,~,f worldwide moderate to large earthquakes (Table 4-1) to evaluate
the degree of association between earthquakes and faults with recent
d'1splacement; 2) a review of moderate to large historical earthquakes
ii~1: California, which has a relatively complete data base, to evaluate
the threshold earthquake magnitude that results in recognizable surface
di'!isplacement; and 3) a field evaluation of the Talkeetna Terrain for
preservation of geological evidence of past earthquakes.
4.2.1-Selected Worldwide Earthquakes and Faults with Recent
Displacement
The primary purpose of evaluating the worldwide earthquake data
base was to assess the degree of associ at ion between earthquakes
and faults with recent displacement. The data analysis concen-
trated on the following topics: 1) assessing the threshold
magnitude of earthquakes for which surface faulting was recogniz-
able; 2) reviewing reports in the literature about large magnitude,
shallow earthquakes that resulted in no surface faulting; and
3) evaluating reports of faults that had been dormant for tens
of thousands of years and then were reactivated during a large
earthquake. From these data and analyses, an attempt was made to
estimate the threshold earthquake magnitude that would be expected
to produce surface rupture.
Data were collected and analyzed for worldwide earthquakes that
occurred in regions having geologic conditions similar to those of
the Talkeetna Terrain. The earthquakes were selected according to
the following screening criteria: tectonic setting, depth of
hypocenter, geologic setting, geomorphic terrain, and style of
faulting. If an earthquake occurred within the Talkeetna Terrain,
it would be classified as a shallow (less than 12 miles [20 km])
intraplate earthquake associated with crustal deformation within
the North American plate. However, the Talkeetna Terrain is
4 -3
Woodward·Ciydle Cons!Ulltants
close to the plate boundary on a regional scale, and the Terrain
may be under the influence of the North Ameri can-Pacific p 1 ate
co ll i s i on . Thus , an earth q u ak e t h at occurred co u l d a l so be
classified as an interplate earthquake related to plate edge
deformation on a regional scale. Both tectonic settings were
considered in the selection of worldwide data because the classi-
fication of worldwide earthquakes is not standardized.
Approximately 30 earthquakes of magnitude (Ms) 5.4 to 7.8 that
occurred in environments that are representative of the Talkeetna
Terrain were selected from three sources: Kanamori and Anderson
( 19 7 5 ) , S l e mm o n s ( 19 77 b ) , an d W y s s ( 19 7 9 ) . T h e d at a r e p or t e d
by these three sources were reconciled and combined with data from
other literature. For the earthquakes selected, available data
were compiled for the following variables: date, location,
magnitude, and association with faults having recent displacement.
These data are presented in Table 4-1.
The most prevalent limitations in the data base are caused by
differences in reporting of data. This is especially true for
field reports of earthquake effects such as surface faulting. The
quality of the reported data most often depends on the location and
size of the earthquakes (or loss of property and lives). These
limitations on quality of the worldwide earthquake data base make
rigorous comparisons difficult, but conclusions can be reached and
trends in the data can be described that are informative regarding
the concept of the detection level earthquake.
Evaluation of the worldwide data set shows that all of the reported
earthquakes occurred near a mapped fault with recent displacement
or within terrains where faults with recent displacement exist
(Table 4-1). Occasional reports of faults rupturing new ground or
being reactivated after tens of thousands of years do exist but are
4 - 4
unverified. An example is the magnitude (Ms) 7.0 Inangahua, New
Zealand, earthquake of 1968. During the earthquake, rupture
occurred both along a pre-existing scarp and a subsidiary fault;
the latter had not been previously recognized. The rupture
along the subsidiary fault was called 11 new.11 However, the earth-
quake occurred on a recognized fault with recent displacement
and the displacement along the subsidiary fault should not have
been considered an unexpected event.
The results of the evaluation of this data set show that, in
general, surface faulting is reported for earthquakes with magni-
tudes larger than magnitude (Ms) 6.5. The exceptions are mainly
earthquakes in C a 1 iforni a and Japan where earthquakes are intensely
studied and surf ace faulting has been reported for earthquakes as
small as magnitude (Ms) 5. Conversely, some shallow earthquakes
as large as magnitude (Ms) 7-1/4 have not had observed surface
rupture; however, these earthquakes have occurred in areas where
faults with recent displacement are present.
Although the quality of the worldwide earthquake data base is
variable, the primary conclusions that can be drawn from the
worldwide data base are: 1) moderate to large earthquakes in areas
similar to the Talkeetna Terrain are consistently associated with
recognizable faults with recent displacement or terrains where
recognizable faults with recent displacement are present; 2)
earthquakes less than magnitude (Ms) 5.5 have been observed to
result in surface faulting; and 3) shallow earthquakes as large as
magnitude (Ms) 7.0 or even greater have been documented with no
surface faulting.
Because the worldwide data vary in quality, and trends in the data
are lacking vJith regard to size and depth of earthquake versus
occurrence of surface rupture, a more uniform data base was sought
4 - 5
to examine the threshold of surface faulting. The most uniform and
detailed data base available is from California and is reviewed in
the following paragraphs.
4.2.2 -Occurrence of Surface Faulting in California
The magnitude threshold at which shallow earthquakes result
in recognizable surface faulting was estimated by evaluating
earthquakes that have occurred in California. Although the
historical record in California is short, a sufficiently large
number of moderate to large earthquakes have been intensively
investigated to estimate the threshold magnitude of surface
faulting.
Data were evaluated for earthquakes that occurred between 1900 and
1980 by reviewing compilations by Slemmons (1977b), Real and others
(1978), Toppozada and others (1979), and various reports of
earthquakes by the California Division of Mines and Geology,
U.S. Geological Survey, California Institute of Technology,
and Earthquake Engineering Research Institute. These data are
summarized in Table 4-2 according to periods of time from 1900
through 1969, and from 1970 through 1980. The data are for
earthquakes that occurred onshore within the borders of California.
Three magnitude ranges are shown to illustrate the sensitivity of
recognizing surface rupture according to variation in earthquake
magnitude.
Table 4-2 shows that during the first 69 years of this century
surface rupture was recognized for only five of the 34 earthquakes
over magnitude (Ms) 6. In contrast, from 1970 through 1980,
surface rupture was documented for all three earthquakes over
magnitude (Ms) 6 and for 75 percent of earthquakes over magnitude
(Ms) 5.5. The reason for the difference in recognition of
4 - 6
surface rupture during the past eleven years is the greater
interest that scientists now have in recognizing and understanding
surface faulting as a phenomenon. Thus, it can be inferred that as
studies of moderate to large earthquakes have become more thorough,
the detection of surface faulting has increased.
From these data, it can be concluded that the threshold of surface
faulting within California is on the order of magnitude (Ms) 5.5
and that for earthquakes over magnitude (Ms) 6.0 a high certainty
of surface faulting is expected.
4.2.3-Preservation of Rec~nt Displacement
The geologic detection of faults with recent displacement depends
on preservation of geologic evidence of past earthquakes. Earth-
quakes over the threshold magnitude of (Ms) 5.5 to 6.0 rupture
the sediments overlying the fault and rupture the ground surface,
thus creating a fault scarp. The ruptured sediments may be
detected if they are not buried by younger sediments. Detection
of the fault scarp depends on: 1) the height and length of the
original fault scarp, and 2) the degree of degradation of the scarp
between the time it was created and the present.
Faults with recent displacement may be detected if the sediments
and geomorphic surfaces present along the trace of the fault were
in existence prior to the last earthquake that produced surf ace
rupture. For example, the Talkeetna thrust fault has sediments
from the earlier part of the Late Wisconsin epoch (25,000 to 15,000
y.b.p.) along 85 percent of its length (Figure 3-4). Consequently,
single fault displacements from earlier than 25,000 y.b.p. cannot
be detected by surface geologic studies along 85 percent of the
fault. Erosion of a new geomorphic surface may not preclude
4 - 7
Woodward· Clyde
detection of the fault. Geologic evidence for recency of displace-
ment may be destroyed, but the fault should still be detectable as
a lineament.
The initial height and length of a fault scarp depends on the
magnitude and depth of the earthquake. The degree of degradation
depends on the type of geomorphic surfaces and Quaternary sediments
present in the Talkeetna Terrain and the geologic processes, such
as annual freeze and thaw, slumping, and solifluction, that reduce
the sharpness of small topographic features such as fault scarps.
Glacial scarps of varying heights and lengths that were similar
in form to fault scarps were examined in the project region to
establish the detection limit for preserved scarps in the Talkeetna
Terrain. On the basis of this examination, the detection limit for
initial surface rupture is a rupture length of approximately
9 miles (15 km) and a scarp height of approximately 1.7 to 3.3 feet
(1/2 to 1 m). Geologic evidence of this initial surface rupture
would be recognizable for thousands of years. Surface ruptures of
these dimensions are associated with earthquakes of magnitude
(Ms) 6 to 6-1/2 (Slemmons, 1977b).
Since most faults with recent displacement have many magnitude
(Ms) 6 to 6-1/2 or greater earthquakes during any 25,000-year
period, detectable evidence usually remains somewhere along the
length of the fault. However, it is theoretically possible for a
fault to have a single displacement associated with a magnitude
(Ms) 6 to 6-1/2 earthquake that occurred between 25,000 and
100,000 years ago. Such cases of long recurrence intervals of
displacement with surface rupture dimensions near the limits of
detection present the most difficulty in identifying faults with
recent displacement.
4 - 8
Again using the example of the Talkeetna thrust fault, 62 miles
(100 km) of the total fault length of 78 miles (126 km) are covered
by sediments that are only 15 ,000 to 25 ,000 years o 1 d (as shown in
Figure 3-4). It is hypothetically possible, though geologically
unlikely, that eight single ruptures with lengths that are smaller
than the detection limit of our investigation (rupture length of
9 miles [15 km]) could be fit within the 78-mi le (1Z6-km) fault
length. However, even if earthquakes occurred on each of the eight
sections of the fault only once every 100,000 years, it is highly
unlikely that all eight earthquakes would have occurred before
25,000 y.b.p. Our investigation should have detected evidence of
one of the ruptures somewhere along the fault. Therefore, faults
with recent displacement having surface ruptures associated with
earthquakes smaller than magnitude 6 to 6-1/2 are likely to have
been detected by our field studies.
4.2.4-Detection Level Earthquake
Evaluation of worldwide data for moderate to large earthquakes in
areas similar to the Talkeetna Terrain shows that recognizable
faults with recent displacement occurred in close proximity to or
were directly associated with the earthquakes.
On the basis of the review of the worldwide and California data on
surface faulting associated with earthquakes, it is reasonable to
assume that the threshold of surface faulting within the Talkeetna
Terrain is on the same order of magnitude, that is, magnitude
(Ms) 5.5 to 6.0. On the basis of the study of scarp preserva-
tion, we consider it likely that scarps that were originally 1.7 to
3.3 ft (0.5 to 1.0 m) high would be detectable today. Scarps of
these dimensions wou 1 d result from earthquakes of magnitude (Ms)
6.0 to 6.5 (Slemmons, 1977b). Thus, faults with recent displace-
ment associated with magnitude (Ms) 6.5 or greater earthquakes
may be directly detected by observing the individual fault scarps.
4 - 9
Faults with recent di sp 1 acement that are associated with earth-
quakes of magnitude as low as (Ms) 6.0 are also considered likely
to have been detected by our geologic investigation for four
reasons: 1) there are no confirmed cases of faults that have been
dormant for tens-of-thousands of years being the source of a
moderate to 1 arge earthquake; 2) a fault characterized by many
short ruptures along its length would probably have been detected;
3) lineaments on geomorphic surfaces or in sediments often make a
fault detectable even if the fault scarp is not preserved; and
4) terrains with moderate to large shallow earthquakes also have
clearly recognizable faults with recent displacement, and, within
the Talkeetna Terrain, no faults with recent displacement are
apparent.
Because no fault-related scarps were detected in the Talkeetna
Terrain, we concluded that the magnitude of the detection level
earthquake is (Ms) 6.0. This is the largest earthquake that
could theoretically occur on a fault that might not have been
detected by our geologic investigation. Such an earthquake could
occur on a source anywhere within the Talkeetna Terrain to the
depth of the crust, which is approximately 12. miles (20 km) as
discussed in Section 5. The recurrence interval for such an
earthquake is low, as shown in Figure 5-9. Consequently, the
likelihood of this earthquake occurring within 6 miles (10 km) of
either of the sites is considered to be low. This low likelihood
incorporates the size of the rupture plane that can be estimated
for a magnitude (Ms) 6 earthquake and the variety of ori en tat ions
(from horizontal to vertical) that the rupture plane could have
within the crust. As shown in Figure 8-8, the rupture plane
associated with a magnitude (Ms) 6 earthquake is 27 square miles
(71 km2) with a length and width of 5.2 miles by 5.2 miles (8.4 km
by 8.4 km).
4 -10
Woodward· Clyde Consultants
-Talkeetna Terrain Boundary Faults
4.3.1-Castle Mountain Fault
The Castle Mountain fault is predominantly a strike-slip fault that
dips steeply to the north. The fault is approximately 295 miles
(475 km) long and trends east-northeast/west-southwest about
71 miles (115 km) south of the Devil Canyon site and 65 miles
(105 km) south of the Watana site (Figure 4-1). It is nearly
vertical or steeply dipping to the north (Detterman and others,
1974' 1976).
The fault is present as a single trace along its mapped western
section. Along the eastern section of the fault, in the Matanuska
Valley, the fault consists of the main trace and a major splay,
which is known as the Caribou fault (Grantz, 1966; Detterman and
others, 1974, 1976). Detterman and others (1976) propose that the
main trace represents the older and more fundamental break of the
two traces, while the Caribou fault is the trace along which late
Cenozoic displacement has occurred. As is reported for the Denali
fault, the Castle Mountain fault is generally regarded as a major
suture zone within the earth's crust.
Displacement along the fault has been occurring since about
the end of Mesozoic time (Grantz, 1966), approximately 60 to
70 m.y.b.p. (Figure 4-2). The fault incorporates a combination of
right-lateral and reverse motions with the north side up relative
to the south side (Grantz, 1966; Detterman and others, 1974,
1976). The maximum amount of vertical displacement is approxi-
mately 1.9 miles (3 km) or more (Kelley, 1963; Grantz, 1966); the
maximum amount of strike-slip displacement is estimated by Grantz
(1966) to have been several tens of kilometers, although Detterman
and others (1976) cite 10 miles (16 km) as the total displacement
that has occurred along the eastern traces of the fault.
4 -11
Woodward-Clyde Consultants
During aerial reconnaissance for this study, the fault was observed
to be expressed as a series of linear scarps and prominent vegeta-
tion alignments in the Susitna Lowland. Along its eastern portion
in the Talkeetna Mountains, a lithologic contrast and possible
offset of the Little Susitna River and other streams provide
evidence for the location and recent age of the fault.
Evidence of Holocene displacement is observed only in the western
segment of the fault in the Susitna Lowland (Detterman and others,
1974, 1976). To date, no evidence of Holocene displacement has
been reported in the Matanuska Valley, although Barnes and Payne
(1956) propose that up to 0.8 mile (1.2 km) of vertical displace-
ment has occurred in the Matanuska Valley in Cenozoic time.
Slip on the Castle Mountain fault during Holocene time (<9,000
y.b.p.) continues to be predominantly strike-slip with a component
of dip slip as indicated by displacement of undated Holocene
features. In the Susitna Lowland horizontal displacement of a sand
ridge has involved 23 feet (7 m) of right-lateral displacement;
near-surface sediments have been displaced vertically 7.5 feet
(2.3 m) (Detterman and others, 1974). Bruhn (1979) excavated two
additional trenches across the fault. A river terrace near one of
two trenches has been right-laterally displaced approximately
7.9 feet (2.4 m), and one of the trenches across the fault revealed
3.0 to 3.6 feet (90 to 110 em) of dip-slip displacement of sedi-
ments. At this location, the north side is up relative to the
south side along predominantly steeply south-dipping fault traces.
This reversal of dip in faulted youthful sediments is commonly
caused by curving of the fault plane near the surface. The
dominance of right-lateral over vertical displacement appears to be
continuing in a similar proportion to that which occurred earlier
during Cenozoic time.
4 -12
Oetterman and others (1974) found evidence suggesting that the
7.5 feet (2.3 m) of dip-slip movement they observed has occurred
within the past 225 to 1,700 years, and Bruhn (1979) estimates
that the displacements he observed are similar in age, 222 to
1,860 y.b.p. Oetterman•s age estimate was interpreted from
Carbon-14 age dates obtained from the soil horizons displaced by
the fault and tree-ring counts from trees growing on the scarp.
From the available data, the rate of strike-slip displacement
cannot be calculated directly; however, the data imply a strike-
slip rate of displacement of 0.05 to 0.4 inches/year (0.13 to
1 em/year). A value of 0.2 inches/year (0.5 em/year) is used to
estimate the average recurrence of the MCE. Considering this
slip rate and the earthquake recurrence relationship for the
Castle Mountain fault (Figure 5-9), and applying the model for
the proportion of slip caused by the maximum credible earthquake
(Appendix A.7), we have estimated that a magnitude (Ms) 7.5
earthquake can occur on the Castle Mountain fault with an average
recurrence of 235 years.
There is no documented displacement along the Castle Mountain fault
in historic time. Plafker (1969) reports no observed displacement
during the 1964 Prince William Sound earthquake. A magnitude
(Ms) 7.0 earthquake occurred in the vicinity of the Castle Mountain
fault west of Anchorage in 1933. It is not known if the earthquake
was related to the Castle Mountain fault, and no investigations to
look for surface displacements have been reported (Page and Lahr,
1971).
Detterman and others ( 1976) have reviewed hi stori cal seismicity
in the vicinity of the fault for the time period 1934 through
October 1974. Most of the events in the vicinity of the fault have
reported focal depths of more than 19 miles (30 km) with the
precision in hypocenter depths estimated by the authors to be up to
4 -13
Woodward= Clyde
_:12 miles (20 km). The depth of these earthquakes suggests that
the events may be occurring at depth below the crust. In summary,
there has been seismic activity in the vicinity of the fault but no
reported correlation of earthquakes with the fault.
The Castle Mountain fault was classified during this investigation
as being a fault with recent displacement. The effect of potential
seismic ground motions from the Castle Mountain fault is considered
to be significantly less than that of the Denali fault because the
Denali fault has the potential for a larger earthquake and is
closer to the sites (as discussed in Sections 7 and 8). The Castle
Mountain fault is too far from the sites to affect potential
surface rupture considerations.
4.3 .2 -Dena 1 i Fault
The Denali fault is predominantly a right-lateral strike-slip fault
that is approximately 1,358 miles (2,190 km) long (Richter and
Matson, 1971). The fault consists of a number of segments and has
an arcuate east-west trend in the site region as shown by Grantz
( 1966) and Richter and Matson ( 1971), among others. North of the
site, the fault divides into two traces or strands. The northerly
segment is the Hines Creek strand, as shown in Figure 8-2 of the
Interim Report (Woodward-Clyde Consultants, 1980b). The southerly
strand is the part of the McKinley strand that passes within
43 miles (70 km) north of the Watana site and 40 miles (64 km)
north of the Devil Canyon site.
The fault has been the subject of numerous studies and is generally
agreed to represent a major sut~re zone within the earth's crust as
discussed by St. Amand (1957), Grantz (1966), Cady and others
(1955), Richter and Matson (1971), Page and Lahr (1971), Stout and
others (1973), Forbes and others (1973), Wahrhaftig and others
4 -14
(1975), Hickman and others (1978), and Stout and Chase (1980),
among others. The total amount of displacement along the fault
is the subject of continuing discussion. Some investigators
suggest the amount of strike-slip displacement is relatively small
(Csejtey, 1980), while others cite evidence supporting total
displacements of up to 155 miles (250 km) (St. Amand, 1957).
The Hines Creek strand of the Denali fault is believed to be
the older of the two strands with strike-slip displacement ceasing
by approximately 95 million years before present (m.y.b.p.)
(Wahrhaftig and others, 1975; Hickman and others, 1976). Subse-
quent strike-slip displacement has principally occurred along the
McKinley strand of the Denali fault (Wahrhaftig, 1958; Grantz,
1966; Hickman and Craddock, 1973; Stout and others, 1973). Because
the McKinley strand is the closer of the two strands to the sites,
and because most of the major strike-slip displacement is thought
to be occurring along this strand (rather than along the Hines
Creek strand), the Denali fault (in the site region) was considered
for the purposes of this investigation to consist of the McKinley
strand along with the Togiak-Tikchik fault segment, the Holitna
fault segment, the Farewell fault segment, and the Shakwak valley
fault segment (west of the Totschunda fault), as described by
Grantz (1966). These segments and the McKinley strand comprise the
Denali fault as cited in this report. The Denali fault is shown in
Figures 1-1 and 4-1.
Aerial reconnaissance of the fault in the vicinity of Cantwell
during this study revealed strong morphologic expressions of recent
displacement, such as fault scarps, offset ridges, linear valleys,
and sag ponds in bedrock or surficial sediments of undefined age.
The linearity of these features across the topography suggests that
the fault plane is close to vertical in this area.
4 -15
Woodward·CDyde Consultants
Holocene age displacements along the southern segment have been
studied by several investigators. In the Nenana River area,
Hickman and Craddock (1973) found evidence for as much as 443 feet
(135m) of right-lateral displacement and 10 to 13 feet (3 to 4 m)
of dip-slip offset, with the south side up relative to the north
side, in Holocene time. These data suggest a displacement rate of
' approximately 0.53 inches/year (1.3 em/year). Stout and others
(1973) measured right-lateral offsets as great as 197 feet (60 m)
and as much as 33 feet (10 m) of dip-slip displacement, with the
north side up relative to the south side, in Holocene units east of
the Black Rapids Glacier (northeast of the site region); on the
basis of these data, they estimated the displacement rate to be
between 0.20 and 0.24 inches/year (0.5 and 0.6 em/year) of right-
lateral motion and less than 0.06 inches/year (0.15 em/year) of
dip-slip motion during Holocene time. Other studies, including
Plafker and others (1977), Hickman and others (1977; 1978), and
Richter and Matson (1971), found evidence supporting a displacement
rate between 0.4 to 1.4 inches/year (1.0 to 3.5 em/year) on the
southern segment in Holocene time.
In summary, displacement rates in Holocene time along the Denali
fault locally range from approximately 0.2 to 1.4 inches/year
(0.5 to 3.5 em/year). Considering this slip rate and the earth-
quake recurrence relationship for the Denali fault (Figure 5-9),
and applying the model for proportion of slip caused by the
maximum credible earthquake (Appendix A.7), we have estimated that
a magnitude (Ms) 8 earthquake can occur on the Dena 1 i fault with
an average recurrence of 290 years. There is no documentation of
displacement on the fault in historic time. Hickman and others
(1978) suggest that the latest movement was several hundred to
several thousand years ago.
4 -16
Review of historic seismicity during this investigation, including
review of other published historical seismicity studies (e.g.,
Tobin and Sykes, 1966; Boucher and Fitch, 1969; Page and Lahr,
1971), suggests that seismic activity has occurred in the vicinity
of the Denali fault. This seismicity includes microseismicity
reported by Boucher and Fitch (1969) and macroseismicity, i.e.,
events of up to magnitude (Ms) 5 to 6 (Tobin and Sykes, 1966).
As discussed in Section 5.1, two large earthquakes (magnitude
greater than 7) have occurred in the general vicinity of the Denali
fault. However, uncertainties in the location and focal depth of
these events preclude their correlation with the Denali fault.
The Denali fault was classified during this investigation as being
a fault with recent displacement. The fault affects consideration
of the potential for seismic ground motions at both sites.
However, the fault does not affect consideration of surface rupture
potential through either site because of the distance of the fault
from the sites.
4.3.3 -Benioff Zone
The Pacific plate is moving northwestward at a relatively faster
rate than the North American plate. Along the Aleutian Trench
in the Gulf of Alaska, the differential rate of movement is
accommodated by subduction or underthrusting of the Pacific
plate beneath the North American plate, as shown in Figure 4-1.
The subducting Pacific plate dips beneath Alaska to a depth of
approximately 93 miles (150 km), as discussed by Packer and others
(1975), Davies and House (1979), Agnew (1980), and Lahr and Plafker
( 1980).
Evidence for the subducting Pacific plate is the zone of seismicity
associated with the p 1 ate. This zone of seismicity, the Benioff
4 -17
zone, has been observed in the site region by Davies (1975) and
Agnew (1980) and is discussed in Section 5 and shown in Figure 5-7.
Southeast of the site (beneath the Matanuska Valley region),
the Benioff zone becomes de coup led from the North American p 1 ate
and increases in dip. Hypocentral data obtained during this
investigation show the Benioff zone to be at depths of 31 mi1es
(50 km) and 38 miles (61 km) at its closest distance to the Watana
and Devil Canyon sites, respectively (Figure 5-7).
The Benioff zone is considered to consist of two subzones or
regions separated by a transition zone, as discussed in Section
5.2.1 and shown in Figure 5-7. The two regions are referred to as
the interplate region and the intraplate region. The interplate
region includes earthquakes along the interface between the
crustal plate and the subducting plate. The intraplate region is
that portion of the Benioff zone that is detached from and dips
beneath the crustal plate. This region includes earthquakes that
occur within the subducting plate.
Both regions of the Benioff zone are considered to be a source
of seismicity for the sites. This judgment is based on the
association of earthquakes with the downgoing slab and the proxim-
ity of the slab to the sites. The zone is not considered to affect
consideration of surface rupture potential through the sites
because of the depth of the zone at the sites and the decoupling
from the crust.
4 -18
-Features Within the Talkeetna Terrain
4.4 .1 -Watana Site
Talkeetna Thrust Fault
The Talkeetna thrust fault is a reverse or thrust fault that was
active during Cretaceous and early Cenozoic time, more than 50
million years ago (Figure 4-2). Although it is apparently very
old, it is the largest identifiable fault that passes near the
dam sites; therefore, field studies were conducted to detect any
evidence for recent fault displacement.
The Talkeetna thrust fault generally trends N50°E and passes
4 miles (6.5 km) southeast of the Watana site and 16 miles
(25 km) southeast of the Devil Canyon site. The fault is
approximately 78 miles (126 km) long and is part of a longer zone
of deformation that includes the Broxson Gulch fault to the
northeast (Beikman, 1974a) and an inferred extension beneath the
Susitna lowland to the southwest (Csejtey and others, 1980).
The Talkeetna thrust fault is recognized primarily as a litho-
logic contrast of regional extent between early Cretaceous
or Jurassic and older volcanic and sedimentary rocks of the
Wrangellia terrane to the southeast, and Lower Cretaceous
sedimentary rocks of the Maclaren terrane to the northwest
(Figure 4-3) (Csejtey and St. Aubin, 1981). Adjacent to the
fault within the site region, the Wrangellia terrane includes
interfingered subaerial and submarine basalt, argillite, sand-
stone, and limestone (Silberling and others, 1981). The Maclaren
terrane includes argillite and greywacke sandstone (Csejtey and
St. Aubin, 1981). Both terranes have been metamorphosed.
4 -19
Woodward-Clyde Consultants
Csejtey and others (1978) suggest that the Talkeetna thrust
fault is associated with large-scale thrust faulting along the
continental margin that juxtaposed the Maclaren and Wrangellia
terranes in late Mesozoic to Tertiary time, 50 to 100 m.y.b.p.
Faults of this type have been found throughout southcentral
Alaska; they are the result of accretion of parts of the Pacific
plate onto the North American plate during Cretaceous and early
Tertiary time (Jones and Silberling, 1979).
This accretionary process initially started in the Fairbanks area
in Jurassic time approximately 141 to 195 m.y.b.p. As accretion
continued through geologic time, the faulting associated with
it shifted southward from the Tintina fault near Fairbanks,
through the area of the Talkeetna thrust fault, and is currently
occurring in the Gulf of Alaska along the Aleutian trench. Thus,
the tectonic environment in which the Talkeetna thrust fault
is believed to have been active is no longer present in the
Talkeetna Terrain. Faulting is now occurring along the Aleutian
trench.
The location of the fault is imprecisely known along major
sections of its length because it is covered by relatively young
(Cenozoic) deposits except in the headwaters of Windy Creek
(east of Location W1), and at Locations W9 and vJlO near the
Talkeetna River (Figure 4-4). Between these locations, the fault
is inferred to be located between outcrops of the Maclaren and
Wrange 11 i a terranes that are widely separated by areas covered by
younger Cenozoic sediments. The location of the fault can be
inferred to be within a three-mile-(4.8-km-) wide band along
most of its length and within a one-mile-(1.6-km-) wide band at
the Susitna River (near location W5 in Figure 4-4).
4 -20
Woodward-Clyde Consultants
The Talkeetna thrust fault dips steeply to the southeast accord-
ing to: geologic mapping conducted during this study; the work
of Turner and Smith (1974} and Csejtey and others (1978}; and
aeromagnetic surveys of the Talkeetna Mountains quadrangle
(Csejtey and Griscom, 1978}. Southwest of Denali to the town ,of
Talkeetna, the aeromagnetic surveys and geologic mapping suggest
a southeastward to near vertical dip of the fault. The south-
eastward dip is the original dip from early Cenozoic formation of
the fault (Csejtey and St. Aubin, 1981).
Northeast of Denali, the lithologic contrast between the Wran-
gellia and Maclaren terranes can be recognized across northwest-
dipping thrust faults. Outside the project region to the
northeast, Stout and Chase (1980) and Chase (1980} have observed
Oligocene dikes and sediments that are offset by the northwest-
dipping Broxson Gulch thrust fault. Nokleberg {1981) has
extended the Broxson Gulch fault into the project region on
the basis of detailed mapping of the Mount Hayes quadrangle.
Detailed geologic mapping of the adjacent Healy quadrangle by
Smith (1981} shows the same lithologic contrast and northwestward
fault dip east of Denali. Because of these similarities, we
applied the name Broxson Gulch to the zone of deformation east of
Denali.
West of the Susitna River at Talkeetna, the aeromagnetic data
across the lithologic contrast were interpreted by Griscom {1979}
to suggest an unconformity that dips to the northwest rather than
a fault. The unconformity is also broken by later northwest-
trending faults (Griscom, 1979). Displacement by later faulting
and loss of character as a fault begin at the Talkeetna River
(near location W10 in Figure 4-4}.
4 -21
Woodward-Clyde Consultants
Although they may be part of the same long zone of deformation,
each part of the zone may be a different tectonic boundary. The
different segments may have developed as separate faults having
an opposite sense of underthrusting in Cretaceous and early
Tertiary time. Alternatively, the Talkeetna thrust fault may
represent the original dip of the zone of deformation to the
southeast, with the Broxson Gulch thrust fault and inferred
southwestward ext ens ion having been rotated to a northwestward
dip by later deformation.
Field studies were conducted along the Talkeetna and Broxson
Gulch thrust faults during 1981 to: 1) confirm the basis for
differentiati-ng the two faults, 2) refine knowledge of the nature
of the Talkeetna thrust fault, and 3) assess whether the fault
has been subject to recent displacement (i.e., displacement
within the past 100,000 years). During these studies, a number
of key locations were mapped or logged that contributed signifi-
cantly to our understanding about the nature and age of the
Talkeetna thrust fault. The key locations are shown in Figure
4-4.
The Windy Creek cross-section at location W1 is representative of
the Broxson Gulch trust fault. The fault dips to the north, and
the lithologic contrast is between metasedimentary rocks of the
Maclaren terrane and volcanic rocks of the Wrangellia Terrane
(Figure 4-5). The northward dip is inferrred from the north-
westward dip of overturned drag folds in Windy Creek, which is
consistent with the· dip inferred from mapping by Smith (1981) and
Nokleberg (1981). Smith (1981) suggests that the Maclaren
terrane was thrust over the Wrangellia terrane in the same
direction as on the Broxson Gulch thrust fault i.n the Mount Hayes
quadrangle.
4 -22
Woodward-Clyde Consultants
The cross-section of the fault near Butte Creek at location W2 is
believed to be typical of the Talkeetna thrust fault in the
vicinity of the dam sites (Figure 4-6). However, the Talkeetna
thrust fault itself is covered by Quaternary glacial sediments
in Butte Creek valley and in the vicinity of the sites. Its
location in Butte Creek valley is inferred from the presence of
Maclaren terrane rocks within and to the northwest of the valley
and Wrangellia terrane rocks to the southeast of the valley.
Good exposures in the Clearwater Mountains indicate that the two
subsidiary thrust faults shown in Figure 4-6 dip southeastward
and have thrust progressively older rocks northwestward over
younger rocks. The dip of the Talkeetna thrust fault is inferred
to be southeastward, simi 1 ar to the better exposed subsidiary
thrust faults. Aeromagnetic data for the area of Location W2
show a low gradient that was interpreted to suggest a northwest-
ward dip (Csejtey and Griscom, 1978); however, the observed low
gradient may well be caused by the intervening fault slices of
nonmagnetic sedimentary rock (KJs and TrPs shown in Figure 4-6).
The steep gradients of narrow, elongate magnetic highs (southeast
of the trace of the fault) appear to be 100re suggestive of a
southeastward dip.
The Talkeetna thrust fault is exposed in the bedrock near the
Talkeetna River at Location W9, on Talkeetna Hill (Figure 4-7).
Here the fault traverses an area of low relief and~is marked by a
sharp differential erosion contrast between the argillite on the
northwest and basalt on the southeast. The fault is inferred to
extend southwest from this area of low relief, down a steep
canyon. The slightly sinuous trace and steeply northwestward
dipping axial planes of folds indicate that the fault dips
steeply to the northwest or is vertical (Figure 4-8).
·The detailed variation in lithology within the fault zone at
Location W9 is shown in Figure 4-9. The zone is approximately
4 -23
Woodward-Clyde Consultants
180 feet (55 m) wide between competent argillite on the northwest
and competent basalt on the southeast. The rock units in between
are slightly to extremely sheared.
Tertiary age rocks near the fault at Watana Creeks Fog Creek, and
Talkeetna Hill were evaluated for the presence of deformation
that would be related to displacement on the Talkeetna thrust
fault during or after the Tertiary period.
At Watana Creeks the inferred trace of the Talkeetna thrust
fault is less than one mile (1.6 km) from 15 good exposures of
Oligocene deltaic and marine sediments distributed along a
7-1/2-mile-(12-km-) long stretch of Watana Creek (Figure 4-10).
The sediments have been folded (with minor faulting) into north-
west trending anticlines and synclines. The orientation of these
folds strongly suggests that they are related to a northeast-
southwest compressional stress regime that has existed since
deposition of the sediments approximately 22.5 to 38 m.y.b.p.
The northeast-southwest compression is inconsistent with s i gni-
ficant thrust or lateral faultin_g along the Talkeetna thrust
fault.
At Fog Creek (Location W6 in Figure 4-4}, undated Tertiary
volcanic rocks overlie approximately one-third of a three-mile-
(4.8-km-) wide area across the inferred trace of the fault.
-
Although undated, these volcanic rocks are probably similar
in age (50 m.y.b.p.) to the volcanic rocks at Talkeetna Hill
'f
(Location W8, Figure 4-4). The Tertiary volcanic rocks are
flat-lying to gently tilted. Deformation of the volcanic rocks
would be expected to be more extensive if the Talkeetna thrust
fault had been active at a low level during the past 50 million
years.
4 -24
Woodward-Clyde Consultants
The absence of recent displacement on the Talkeetna thrust fault
is supported by overlying volcanic rocks and later faulting that
displaces the Susitna segment. At Location W8, the Talkeetna
thrust fault projects beneath Tertiary andesite and basalt
(Figures 4-7 and 4-8). The volcanic rocks are not displaced by
the fault; thus, this part of the Talkeetna thrust fault has not
been active since the volcanic rocks were deposited. The age of
these rocks has been estimated from a potassium-argon age date
(Csejtey and others, 1978) as 50 m.y. b. p. Because this part of
the fault has not been subject to displacement for at least
50 million years, it provides strong circumstantial evidence that
the Talkeetna thrust fault has not had recent displacement near
the dam sites 20 miles {32 km) to the northeast. The Talkeetna
thrust fault is also displaced by later faulting at the Talkeetna
River (Figure 4-7). Several miles (several kilometers) of
displacement have occurred on this and similar faults since the
Talkeetna thrust fault had displacement along it.
Low-sun-angle aerial photographs were interpreted for the
Talkeetna thrust fault between Location W2 and the intersection
with the Susitna feature near the Talkeetna River {Figure 4-4).
Seven linear features were observed that were considered to have
a low to moderate likelihood of being fault related. These seven
features were then field checked to evaluate their origin. Of
the seven features, six were found to be clearly of glacial
origin. One of the seven features, at location W7 (Figure 4-4)
was considered to have a moderate likelihood of being fault
related.
The feature at location W7 is a N68°E trending linear scarp
(Figure 4-11) that is approximately 1,700 feet (518 m) long
and 7 feet (2.1 m) high and has a northwest facing slope of 26
degrees {Figure 4-11). The feature is located within 0.5 miles
4 -25
Woodward-Clyde Consultants
{0.8 km) of the inferred location of the Talkeetna thrust fault.
Tertiary and Quaternary cover in this area precludes precise
location of the fault trace. The scarp at location W7 is similar
in form to a fault-related scarp.
Trench T-1 was excavated across the scarp in sediments that are
glacial in origin and of probable Late Wisconsin {25,000 to
15,000 y.b.p.) age; this age estimate was based on the elevation
and surface morphology (Figure 3-4). Examination of the expo-
sures in the trench showed that the sediments are not faulted at
the base of the scarp or northwest of the scarp within the trench
(Figures 4-12 and 4-13).
On the basis of the logging of the trench and detailed mapping of
nearby glacial deposits, the scarp is interpreted to mark the
edge of the Late Wisconsin ice sheet. At the margin of this ice
sheet a large ice-marginal river flowed parallel to the scarp and
nearly perpendicular to the trench. The direction of flow is
indicated by the orientation of channel banks across the trench
at three locations. The river flowed across lodgement till (Unit
6 in Figure 4-12) and the ice sheet to the northwest of the
scarp. The river carried a heavy bedload of gravel (Unit 5) .
. When the ice sheet melted, the grave 1 co 11 apsed onto the lodge-
ment till (Unit 6) northwest of the scarp. As evidence of this
collapse, bedding and clast orientations appear to have been
slightly disrupted within the trench in the interval 95 to 105
feet (29 to 32m).
Trench T-2 was excavated at Talkeetna Hill, Location W10 (Figure
4-4). The trench site was selected because it is on a clear
trace of the Talkeetna thrust fault, and because the low swale
along the fault may have preserved Early Wisconsin sediments. At
this location, the fault is a zone of deformation approximately
4 -26
Woodward-Clyde Consultants
180 feet (55 m) wide. Logging of the trench showed that the
Wisconsin glacial sediments had been flushed from the swale
and replaced by colluvium and loess of probable Holocene age
(Figure 4-12 and 4-14). No fault displacement was observed in
these Holocene sediments. However, the youth of these sediments
makes this observation relatively insignificant. Bedrock within
the trench was observed to be locally highly sheared (Figures 4-9
and 4-12).
The Talkeetna thrust fault is judged to be a fault without recent
displacement for the following reasons:
1) Tertiary volcanic rocks with a potassium-argon age date
as being 50 m.y.b.p. (Csejtey and others, 1978) in age
overlie the fault near the southwest end of the fault
(Location W8). These volcanic rocks have not been displaced.
2) There is no evidence of faulting in surficial units whose age
is 15,000 to 25,000 y.b.p. (Location W7).
3) Folding of Oligocene strata in Watana Creek (Location W3)
suggest that deformation is being accommodated in a fashion
that is inconsistent with fault displacement along the
Talkeetna thrust fault.
4) The Talkeetna thrust fault is offset at least six miles
by younger faulting at the Talkeetna River south of Location
W10.
5) The Talkeetna thrust fault is the result of continental
accretion during late Cretaceous and early Tertiary time.
The present site of continental acc~etion is 350 miles (564
km) to the south-southeast.
4 -27
Woodward-Clyde Consultants
Therefore, the Talkeetna thrust fault does not affec~ considera-
tion of seismic ground motion or surface rupture potential at
either the Devi 1 Canyon or Watana Dam sites.
Susitna Feature (KD3-3)
The Susitna feature is a northeast-southwest trending 1 i neament
that is 95 miles (153 km) long and approaches to within 2 miles
(3.2 km) of the Watana site (Figure 4-4). The feature was
first described by Gedney and Shapiro (1975) as a prominent
topographic 1 i neament, which they observed on LANDSAT imagery.
These authors postulated that the lineament was a fault, in part
on the basis of data assembled by Turner and Smith (1974), which
is described below, and also on the basis of their interpretation
of seismic activity in the vicinity of the southern end of the
feature.
Evidence that the feature is a fault has been inferred by Turner
and Smith (1974) in the Susitna Glacier area of the south flank
of the Alaska Range. The inference is based on K-Ar dates on
plutonic bodies and interpreted cool-down rates associated with
these plutons ~Smith, 1980). According to this hypothesis, the
plutonic units on the east side of the Susitna feature cooled
down more rapidly than those on the west side of the feature,
suggesting that the west side was at greater depth than the east
and, subsequently, was faulted up into contact with the units
that cooled down more rapidly.
Gedney and Shapiro (1975) report that the Susitna feature'"
corresponds to the eastern boundary of the metasedimentary
units in the project area; however, mapping by Turner and Smith
(1974) and Csejtey and others (1978) shows metasedimentary rocks
extending east of the Susi tna feature sever a 1 miles (severa 1
4 -28
Woodward-Clyde Consultants
kilometers) to the Talkeetna thrust fault. Csejtey and others
(1978) report finding no evidence for the postulated Susitna
feature.
Gedney and Shapiro (1975) also suggest that there is seismic
activity associated with the Susitna feature. In particular,
they cite a magnitude (mb) 4.7 event that occurred on 1 October
1972 and a magnitude (mb) 5.0 event that occurred on 5 February
1974. The location given by Gedney and Shapiro (1975) shows the
earthquakes to be spatially close to the surf ace trace of the
Susitna feature. They hypothesize that the focal mechanisms
suggest a right-lateral strike-slip sense of displacement.
Review of the 1972 and 1974 earthquakes during this investigation
showed that, with the error in location reported by Gedney and
Shapiro (1975), the two epicenters could be more than 8 miles
(13 km) from the feature and that the focal depths put the
events at depths of 46 to 47 miles (75 to 76 km). Even with the
imprecision associated with focal depth determinations, it is
reasonable to associ ate these two earthquakes with the Benioff
zone. The correlation of these events with the Susitna feature
appears to be unwarranted because the Susitna feature lies in
the crust and the maximum thickness of crust is approximately
12 miles (20 km) as shown in Figure 5-7.
Field studies were conducted during. 1981 to: 1) locate any
evidence that the Susitna feature is a fault; and 2) evaluate the
likelihood that the feature is a fault with recent displacement.
Figure 4-4 shows field locations where interpretation of rock
exposures and surficial deposits contributed important knowledge
about the nature or age of the Susitna feature.
4 -29
Woodward-Clyde Consultants
Bedrock was mapped at five key locations to assess whether or not
the Susitna feature is a bedrock fault. These locations, shown
in Figure 4-4, are the Butte Lake area (Location W11), outcrops
between Butte Lake and Deadman Lake (Location Wl3), an outcrop
near Deadman Lake (Location W14), Tsusena Creek (Location W15),
and the Talkeetna River south of the confluence with Prairie
Creek (Location W16). Each of these locations is discussed
below.
The Butte Lake area has been mapped by Smith {1973). Some of the
results of this mapping are summarized on a map by Turner and
Smith {1974). As a result of their mapping, Turner and Smith
{1974) inferred that the Susitna feature may be a fault because
of relationships of crystalline units on either side of the
valley in which the Susitna feature is located. They observed no
evidence of a f au 1 t in the 1 ow 1 and where bedrock is covered by
glacial deposits.
During the 1981 study, mapping was conducted by Woodward-Clyde
Consultants in the Butte Lake area (Figure 4-15, Location Wll).
The lithologic units shown by Turner and Smith {1974) were
confirmed; however, the relationship of these units can be
explained by several reasonable alternatives (including the
presence of a fault). No direct or indirect evidence of a fault
was observed, and there is circumstantial evidence of no faulting
on the basis of the geophysical data described below. Smith
{1980) has re-examined his mapping of the Butte Lake area and did
not find evidence of a fault. In addition, he has not observed
evidence of the Susitna feature being a fault at any location
other than the Susitna Glacier area.
Three magnetometer traverses were conducted at Location W11
(Figure 4-16). One of the traverses was run across the lowland
4 -30
Woodward-Clyde Consultants
in which the Susitna feature is located. Two of the traverses
were run across mapped contacts between lithologic units; these
traverses served to calibrate the magnetic signature between
units where no fault was inferred. As shown in Figure 4-16, the
traverse across the Susitna feature (profile 2A) shows no change
in magnetic signature across the Susitna feature (shown by Turner
and Smith [1974] as a mapped contact between the p aragnei ss and
the migmatitic intrusive rocks). The magnetic signature is
nearly identical with that observed in profile 1 which crosses
the contact between the same two rock units where no fault is
inferred. Although the absence of a magnetic anomaly is not
conclusive evidence of no faulting along the Susitna feature, it
does provide circumstantial evidence that a fault is not present.
No evidence of a bedrock fault was observed at Location W14
between Deadman and Butte Lakes. At Location W13 near Deadman
Lake, minor faulting and a bedrock shear zone was observed within
one mile (1.6 km) of the trace of the Susitna feature. The
orientation of the shear zone is subparallel to that of the
Susitna feature. This faulting and shear zone are similar to
that observed elsewhere in the site region and are not considered
to be related to a throughgoing fault zone.
Tsusena Creek (Location W15) is the one location in the site
region that has exposed bedrock where the Susitna feature is
mapped. Turner and Smith (1974} show the Susitna feature as a
fault that cuts across meanders in the creek. Mapping in 1980
and 1981 showed no evidence of a fault in the canyon walls of
Tsusena Creek. A prominent oxidized zone with shearing and
faulting is present in the creek. However, mapping at this
exposure showed the oxidation to be related to a secondary
intrusion within the host granitic rocks. The faulting and
shearing within the oxidized zone is oriented northwest-south-
4 -31
Woodward-Clyde Consultants
east, at right angles to the Susitna feature. This structural
orientation appears to be related to the Fins feature (as
discussed below in this section), not to the Susitna feature.
Therefore, we have concluded that there is no evidence of a major
fault passing through Tsusena Creek with an orientation parallel
to that of the Susitna feature.
At Location W16 on the Talkeetna River, joint orientations were
measured .to make an assessment as to whether the Susitna feature
could be related to joint control. No evidence of a fault was
observed at this location and the joint orientations are at right
angles to the trend of the Susitna feature.
Low-sun-angle aerial photographs were interpreted for the Susitna
feature between Location W16 and the Denali Highway near Location
Wll (Figure 4-4). Twenty-eight lineaments were marked for
consideration of their possible relation to faulting. On the
basis of interpretation of the aerial photographs, all but 10 of
the lineaments were explained by glacial processes or were
not part of a pattern of lineaments that might be related to
faulting. These ten remaining lineaments were field checked and,
on this basis, assigned to a glacial origin or very low likeli-
hood of relation to surface rupture by faulting.
One lineament at Location W12 was trenched in order to confirm
our interpretation of glacial origin for the lineaments. For
this purpose, a prominent scarp was selected that is oriented
parallel to and within 500 to 2,500 ft (152 to 762 m) of the
Susitna feature (Figure 4-17). The scarp is approximately
18 feet (5.5 m) high and 8,000 ft (2.4 km) long.
The slope of the scarp is approximately 22° and faces eastward.
Examination of exposures in the trench (Figures 4-18 and 4-19)
4 -32
Woodward-Clyde Consultants
showed that the scarp is the product of glacial processes,
probably an ice contact process. No faults were observed in the
trench and the glacial sediments are continuous across the base
of the scarp and 90 feet (27 m) to the southeast. On this basis,
we have concluded that the scarp is not related to faulting and
that the other morphologically similar scarps are also not
related to faulting.
The Susitna feature is judged to be a lineament that is not
related to faulting, except for possible consequent erosion along
some bedrock faults at Location Wl3. This judgment is based on
the following reasons:
1) Bedrock exposures at three locations within the topographic
valley forming the Susitna lineament show no evidence of
faulting or subparallel joints.
2) The outcrop pattern of late Mesozoic and early Tertiary
bedrock at Butte Lake does not support the existence of a
fault. Although it is possible to hypothesize the existence
of a fault, a fault is not required to explain the outcrop
pattern.
3) No evidence of a fault was observed in Tsusena Creek where
the Susitna feature crosses exposed rock in canyon walls.
4) Glacial sediments that overlie the fault along most of its
length do not have any surface geomorphic features related to
fault rupture of the ground surf ace. The age of these
glacial sediments is primarily 25,000 to 9,000 y.b.p.
5) A trench excavated across a prominent scarp located along the
Susitna lineament revealed that the scarp is not related to
faulting and is of glacial origin.
4 -33
Woodward-Clyde Consultants
We classified the Susitna feature as being a lineament; there-
fore, it does not affect consideration of seismic ground motion
or surface rupture potential at either site.
Watana Lineament (KD3-7)
The Watana 1 i neament trends approximately e·ast-west a long the
Susitna River for a distance of 31 miles {50 km). At its western
end, the lineament passes through the Watana site {Figure 4-20).
The lineament was identified by Gedney and Shapiro (1975) on
LANDSAT and SLAR imagery; they reported no fault cont ro 1 was
reported. At the seale of the imagery, the 1 i neament approxi-
mately corresponds to a series of somewhat linear sections of the
Susitna River between approximately the confluences of Tsusena
Creek on the west and Jay Creek on the east.
Field studies were conducted in 1981 to: 1) search for bedrock
outcrops across or near the plotted location of the lineament,
and 2) search for any evidence of Quaternary faulting.
Locations along the Watana lineament where exposures of Mesozoic
bedrock occur near the lineament were checked for the presence of
direct or indirect evidence of faulting (Figure 4-20). Exposures
of bedrock were observed by aerial reconnaissance or ground
investigation at Locations W17, W19, W20, and W22. Color aerial
photographs (at a seale of 1:24 ,000) of bedrock outcrops at
Locations W24 and W25 were interpreted for linear features that
might be related to faulting.
Approximately six miles {10 km) upstream of the Watana site,
the lineament cuts across the south bank of the Susitna River
and trend~ across the low plateau northwest of Mount Watana
(Figure 4-20). No evidence of faulting was observed at Location
4 -34
Woodward-Clyde Consultants
Wl9 or at a small outcrop in a canyon at Location W22. A large
hill of argillite that is separated from the canyon wall of the
Susitna River is located at Location W20. A zone of sheared rock
in the east side of the hill trends approximately N65°E. Joints
at Location W20 trend approximately N45°E. The zone of sheared
rock and joints are representative of the small-scale faults and
joints found throughout the Talkeetna Terrain and are not
considered indicative of fault control for the Watana River
lineament.
Bedrock exposures along the east and west projection of the
lineament (Locations W17, W24, and W25) show no evidence of
faulting. Angle borings DH-21 and BH-6 drilled at the Watana
site by Acres American Inc. and the U.S. Army Corps of Engineers,
respectively, show no evidence of a throughgoing structural zone,
although there is a remote possibility that such a zone could be
present that would not have been encountered by the borings
across the Susitna River at the Watana site.
Low-sun-angle co lor infrared and color aeri a 1 photographs were
interpreted for the Watana lineament. The color aerial photo-
graphs cover the entire lineament as far east as Location W24,
and the co lor infrared photographs cover the 1 i neament as far
east as Location W22. No lineaments or patterns of lineaments
suggestive of youthful faulting were observed on the photographs;
however, sever a 1 short, isolated 1 i neaments were field checked.
All of these lineaments were interpreted to have glacial origin
or to be related to surface processes such as slumping. The
triangular areas on the west end (Location Wl8, Figure 4-20) and
the east end (Location W23, Figure 4-20) were intensively checked
for subparallel or splay lineaments. These two areas, like
Location W21, are on plateau areas where some surface morphologic
4 -35
Woodward-Clyde Consultants
expression should reveal the presence of a fault. No surface
morphology was observed that might be related to a fault along
the Watana lineament.
The Watana lineament is judged to be a series of disconnected
short lineaments that are not related to youthful faulting.
Therefore, it does not affect consideration of seismic ground
motion or surface rupture potential at either site.
Fins Feature (KD4-27)
The Fins feature is a shear zone or fault that trends northwest-
southeast between the Susitna River and Tsusena Creek and is
nearly vertical (Figure 4-21). The feature is 2 miles (3.2 km)
long and is shown as a shear zone or fault dipping 70° to 75° to
the northeast on an undated U.S. Army Corps of Engineers A 1 ask a
District map (Plate 05 entitled 11 Watana Reservoir Surficial
Geology 11
). The Fins feature is prominently exposed on the north
side of the Susitna River (Location W27) as a series of vertical
shear zones, which has a total width of approximately 400 feet
(122 m). The shear zone is approximately 2,500 feet (762 m)
upstream from the proposed Watana dam axis and is in a dioritic
unit mapped as being Paleocene in age by Csejtey and others
(1978).
Evidence of the feature has not been observed on the south side
of the Susitna River (Location W28). However, the south bank
does not have the prominent bedrock exposures that are present on
the north bank in this area.
The Fins feature observed on the north bank of the Susitna River
appears to correlate with a moderately to highly weathered,
oxidized shear zone present on the northeast bank of Tsusena
4 -36
Woodward-Clyde Consultants
Creek approximately 2 miles (3.2 km) upstream from the confluence
with the Susitna River at Location W15.
Joint measurements were obtained during the 1980 field season by
Acres American Inc. and Woodward-Clyde Consultants in Tsusena
Creek (Location W15 and other locations). These measurements
show a prominent northwest-southeast trending set of joints that
dip steeply northeast to southwest.
Observations during this investigation at Tsusena Creek included
that of a 6.5-foot-(2-m-) wide fault zone (within the oxidized
zone) that is oriented N30°W and dips 72°NE. The fault zone is
in granitic units of reported Paleocene age and contains mylonite
and possibly pseudotachylite. Elsewhere in the oxidized zone,
small-scale faults oriented northwest-southeast with a northeast
dip and slickensides were observed. The faulting and shearing at
Location W15 appears to be related to shearing associated with
the intrusive events rather than due to displacement along a
fault zone.
No evidence of the feature was observed northwest of the Tsusena
Creek exposure. However, prominent exposures similar to that at
Tsusena Creek are lacking.
The Fins feature appears to underlie a morphologic depression in
surficial units between the Susitna River and Tsusena Creek. It
is also coincident, in part, with a buried paleochannel, which is
filled with glacial deposits. Evidence for the paleochannel is
based on seismic refraction studies conducted by Dames and Moore
(1975) and Woodward-Clyde Consultants (1980a; 1982).
The approximately triangular areas at W26 and W29 (Figure 4-21)
were investigated for surface morphology that might be related to
4 -37
Woodward-Clyde Consultants
an extension of the Fins feature to the northwest or southeast.
Low-sun-angle color infrared and color aerial photographs at a
scale of 1:24,000 were interpreted for subparallel and splay
lineaments. No lineaments were found. Several lineaments
trending at a high angle to the Fins feature were checked to
confirm their relation to glacial processes.
The Fins feature is judged to be a fault at the Susitna River,
but its relation to faulting at Tsusena Creek is questionable.
No Quaternary expression of the feature was found between the
Susitna River and Tsusena Creek or along the two projections of
the feature.
The Fins feature was classified as a fault without recent
displacement; therefore, it does not affect consideration of
seismic ground motion or surface rupture potential at either
sit e.
This judgment is supported by the extremely short length of the
feature and the tectonic setting of the region. Experience in
other parts of A 1 ask a and other areas of the world, including
California, Japan, and South America, suggests that short faults
with recent displacement exist in association with other longer
faults whose recency of displacement is clearly recognizable.
That is, if a region is being subjected to stresses that are of.
sufficient magnitude to cause surface faulting, it is extremely
unlikely that strain release in the region would produce no
surface rupture longer than along one fault a few miles (a few
kilometers) in length. Rather, it is logical to expect that
the strain release would cause extensive rupture during some
earthquakes and smaller amounts of rupture on both the main
faults and shorter subsidiary faults.
4 -38
Woodward· Clyde Consultants
4.4.2 -Devil Canyon Site
Feature KC5-5
Feature KC5-5 trends N15°W for a distance of 12 miles (19 km) and
approaches within 4.5 miles (7 km) east of the Devil Canyon site
(Figure 4-22). The feature was originally identified as a
lineament, in part, by Gedney and Shapiro (1975} on LANDSAT
imagery. Subsequent examination of U-2 photography and aerial
reconnaissance during the 1980 field study resulted in the
extension of the 1 i neament at its northern and southern ends.
North of the river the lineament is a linear stream drainage.
South of the river the morphologic expression is that of a
prominent linear canyon that becomes a shallow linear depression
{Figure 4-22).
Review of the feature in 1980 and 1981 showed clear evidence
of fault control in the canyon south of the Susitna River
(Segment 2 in Figure 4-22} as well as possible fault control
a long a scarp at the southern end of the feature (Segment 4 in ·
Figure 4-22). For the purposes of this investigation, the
feature was considered to be a fault along its entire length; it
is referred to in this report as Fault KC5-5.
Four segments of the fault are discussed below; each segment has
characteristics which bear on the consideration of whether or not
the original lineament is a fault and/or on the recency of
displacement. Segment 1 is the section north of the Susitna
River; Segment 2 is the 6-mi le-(10-km-) long section south
of. the Susitna River which is marked by a prominent canyon;
Segment 3 is the upland plateau south of the canyon section; and
Segment 4 is the low, curvilinear scarp at the southern end of
the fault, which was shown in Figure 8-12 of the Interim Report
{Woodward-Clyde Consultants, 1980b}.
4 -39
Woodward-Clyde Consultants
Segment 1 is a linear stream drainage north of the Susitna River
which has no observed outcrops and no evidence which conclusively
confirms or precludes a fault origin. Joint measurements taken
on the north side of the Susitna River, approximately 656 feet
(200 m) up river from Segment 1, have a subparallel orientation
suggesting possible joint control of this segment of Fault KC5-5.
Segment 2 is a prominent canyon which is approximately 6 miles
(10 km) long and has a maximum relief of 1,000 feet (305 m).
Rugged terrain and 1 i mi ted access at the bottom of the canyon
precluded ground observations. However, detailed aerial review
of the canyon by helicopter showed clear evidence of faulting
in the canyon at three locations (summarized as Location 01 in
Figure 4-22}. The observed fault zones are parallel to the axis
of the canyon and are observable in three ridges which jut into
the canyon.
The zones are aligned, have orientations of N15•w to N2o•w, and
are near vertical. The zones form sharp, distinct boundaries
between what appear to be intrusive rocks. They may also locally
separate intrusive rocks from metamorphic rocks. The width of the
fault zone varies from a few inches (few centimeters) in width to
a few feet (few meters). Locally there is discoloration of the
fault zone to a light gray color. No direct evidence of the
sense of displacement was observed.
The linear trace of the fault and the canyon suggest that the
fault may be predominantly a strike-slip fault; however, a
substantial oblique component to the sense of displacement
cannot be precluded. If the orientation of the fault zone
(approximately N15.W} relative to the generally northwest
orientation of the regional maximum compressive stress direction
is considered, a strike-slip sense of displacement is reasonable
4 -40
Woodward-Clyde Consultants
(assuming that the Mesozoic stress regime associated with
accretion of this land mass to the North American craton was
responsible for origin of the bedrock fault}.
On the basis of observations to date, the amount of displacement
along the fault zone cannot be estimated. The prominence
of the zone suggests that substantial displacement has taken
place, but it isn't clear if this involves several tens of feet
(tens of meters) or hundreds of feet (hundreds of meters) of
displacement.
This segment of the fault lies within Tertiary intrusive rock
units (Figure 4-22} whose age is inferred to be 50 to 60 m.y.b.p.
(Csejtey and others, 1978). There are no recent geologic units
in this segment of Fault KCS-5 which provide direct evidence
about the recency of displacement. However, the faulting is in
bedrock, and morphologic relationships in Segment 3, discussed
below, strongly suggest that the fault has not been subject to
displacement in the last 100,000 years, and probably not in the
last several million to several tens of millions of years. The
development of the canyon appears to be related to differential
erosion a long the fault zone. There may be other factors which
also contributed to development of the canyon, such as additional
faults and/or joints along which differential erosion occurred.
The entire canyon shows evidence of both brittle and ductile
deformation. Taking into consideration this deformation and the
prominent bedrock fault observed in the bottom of the canyon, we
have concluded that this segment of Fault KCS-5 is clearly fault
controlled.
Segment 3 is a shallow, broad curvilinear depression on the
upland plateau which lies south of Segment 2 and the Susitna
River (Figure 4-22). Regional mapping by Csejtey and others
4 -41
Woodward-Clyde Consultants
(1978) originally suggested that this segment of the feature
could be related to a lithologic contact between the Cretaceous
argillite and greywacke metasedimentary sequence on the east and
the Tertiary intrusive sequence on the west. However, mapping
conducted during this investigation shows clearly that the
metasediment-intrusive contact is irregular, and it clearly
does not coincide with Segment 3 of the fault. The fault lies
entirely within the Tertiary intrusive rocks in this segment.
A shear zone is exposed in close proximity to the fault at
Location D2 and may be related to the fault. The shear zone
consists of highly altered, locally decomposed rock, which may be
the result ·of fault displacement, hydrothermal alteration,
or a combination of both processes. The shear zone appears to
be related to intrusion of a Tertiary pluton, but conclusive
evidence to support this conclusion was not obtained.
Southeast of Location D2 is a broad depression in which meander-
ing streams and marshlike conditions exist. No bedrock exposures
are present in this area. The sediments which are present are
interpreted to be of Early Wisconsin age (approximately 40,000 to
75,000 years in age) as discussed in Section 3.4. Detailed
aerial reconnaissance of this area showed no morphologic features
suggestive of fault displacement in the Early Wisconsin sedi-
ments. It is, therefore, concluded that there clearly has not
been displacement along this fault in the last 40,000 years.
Segment 4 consists of an a 1 i gnment of 1 i near scarps in bedrock
which face northeast on the northeast side of a series of low
rises. These scarps can be followed for a distance of 3.2 miles
(2 km); the· longest scarp is approximately 3,280 feet (1,000 m).
The maximum height of the scarp is approximately 3 ft (1 m),
and it has a rounded, subdued expression. The terrain is one of
4 -42
Woodward-Clyde Consultants
subdued relief with a thaw in the mantle of glacial sediments and
in situ soil. Bedrock is exposed intermittently in the area and
is rarely exposed at the scarp or near it.
Reconnaissance mapping conducted in 1980 suggested that the scarp
could be related to joint control or to slumping. Additional
mapping in 1981 suggested that slumping is not a likely cause of
the scarp, as there is no morphologic evidence for slumping
besides the presence of the scarp itself. The scarp may be
joint controlled. Joint orientations near Location 03 (Figure
4-22) are Nl1°W to N39°W, parallel to subparallel to the N15°W
trend that Fault KC5-5 has in this segment. The scarp could,
therefore, be related to differential erosion controlled by
jointing.
During the 1981 field studies, we examined the possibility
that the scarp was controlled by the lithologic contact between
the Cretaceous metasedimentary strata to the southwest and the
Tertiary intrusive rocks to the northeast. Mapping in and near
Location 03 showed that the contact between the two units is
irregular, as is typical of such contacts, and that the contact,
in general, lies up to a half mile (800 m) northeast of the scarp
(Figure 4-22). The scarp, therefore, does not appear to be
controlled by a lithologic contact.
A number of springs were observed to emanate from the base of the
scarp in or near Location 03. Hand excavation of two of these
springs suggested that they emanated from bedrock, but the
controlling mechanism within the bedrock was not directly
observable. The springs may simply be related to 11 daylighting 11
of the groundwater table at the scarp or a decrease in permea-
bility between weathered rock and fresh rock. The springs could
also be related to a groundwater barrier imposed by a fault
zone.
4 -43
Woodward-Clyde Consultants
Our assessment of this segment of the feature is that it could
be fault controlled. No direct evidence of faulting was observed;
however, circumstantial evidence of faulting (e.g., the springs
and the scarp itself) was present, and no compelling alternate
explanation for its presence has been found.
Feature KC5-5 was classified during this investigation as being a
fault without recent displacement; therefore, it does not affect
consideration of seismic ground motion or surface rupture
potential at either site. The judgment about fault inactivity is
based on the absence of morphologic features in 40,000-to
75,000-year-old sediments which overlie part of the fault,
and the absence of any compelling evidence of recent fault
displacement (e.g., systematic stream drainage offsets, scarps in
recent sediments, or offset of youthful geomorphic units). As
discussed in Section 4.2, we believe the level of detectability
of surface rupture in this region is approximately 2 to 3 feet
(0.5 to 1.0 m) of vertical displacement over a distance of
9 miles (15 km). That is, rupture greater than this would be
detectable. We feel that such a displacement has not occurred
along this feature during the last 100,000 years.
Feature KD5-2
Feature KDS-2 trends N55° E for a distance of 0.8 miles (1.3 km)
and approaches within 3.5 miles (5 .6 km) northwest of the Devi 1
Canyon site (Figure 4-22). The feature is a fault mapped by
Richter ( 1967}. The f au 1 t is described by Richter ( 1967} as
having a strike of N70°E and a dip of 30° to the northwest,
although his map shows a strike of N20°E to N60°E for various
segments of the fault. Richter (1967) mapped the fault as having
normal displacement which downdropped Cretaceous metasedimentary
rocks (argillite) on the northwest relative to Tertiary intrusive
4 -44
Woodward-Clyde Consultants
rocks (quartz monzonite) on the southeast. The amount of
displacement is not reported. The fault zone is exposed in and
near Treasure Creek (Figure 4-22) and is described by Richter
(1967) as having clay gouge, slickensides, and limonite staining.
Richter (1967) describes the structural setting of the fault to
be generally northeast-southwest trending, e.g., fold axes have a
strike parallel to the fault strike. From these relationships he
concludes that the fault and the other structural features were
formed during intrusion of the quartz monzonite. The quartz
monzonite was intruded approximately 50 to 60 m.y.b.p., assuming
the intrusive is of an age similar to that reported by Csejtey
and others (1978) for nearby intrusives.
During the 1980 study, an indistinct linear depression was
observed on U-2 photography, which appeared to represent a
possible extension of this fault to the northwest. Feature KDS-2
was, therefore, considered to include both the fault and the
lineament with a total length of 3.5 miles (5.6 km) as reported
in the Interim Report (Woodward-Clyde Consultants, 1980b).
During the 1981 study, large scale (1:24,000 scale) low-sun-
angle photography and co lor photography were reviewed for
Feature KDS-2. The indistinct linear depression showed no
evidence of fault morphology and no evidence suggesting that it
should be considered as part of the fault mapped by Richter
(1967). Detailed aerial review during the 1981 study supported
this con c 1 us i on •
We classified Feature KDS-2 as being a lineament with a short
bedrock fault along part of its length; therefore, it does not
affect consideration of seismic ground motion or surface rupture
potential at either site. Based on the work conducted on this
4 -45
Woodward-Clyde Consultants
fault, our conclusion is that a short bedrock fault is present.
There is no evidence to suggest that the fault has had recent
di sp 1 acement. The 1 i neament does not now represent a pass i b le
extension of the fault (i.n our judgment). These conclusions are
based on Richter • s (1967) suggestion that the fault is Mesozoic
in age, the absence of any evidence suggestive of a fault with
recent dis p 1 a cement, and the extremely short 1 ength of the
fault.
Feature KDS-3
Feature KDS-3 trends N45• E to N55• E for a distance of 51 miles
(82 km) and approaches within 3.6 miles (5.8 km) northwest of the
Devil Canyon site (Figure 4-22). Part of the lineament is
identified as a fault by Kachadoorian and Moore (1979) .on the
basis of mapping by Csejtey and others (1978). The remainder of
the lineament was identified by Gedney and Shapiro (1975) on SLAR
and LANDSAT imagery. Subsequent examination of U-2 photography
(Woodward-Clyde Consultants, l980b) showed the lineament to be
expressed morphologically as a prominent linear segment of
Portage Creek and as a prominent 1 i near bench a long the south
bank of the Susitna River southwest of Portage Creek.
Ground and aerial reconnaissance studies conducted in 1980 along
Portage Creek showed the lineament to consist of an elevated
prominent linear depression along the northwest bank of Portage
Creek (Woodward-Clyde Consultants, l980b). At the northeast end
of the 1 i neament, mi nera 1 i zed zones were observed in Port age
Creek. Further to the south, along the northwest side of the
creek, an apparent shear zone was observed which could not
be reached on the ground. It was concluded in 1980 that the
shear zone could be related to the lineament (Woodward-Clyde
4 -46
Woodward-Clyde Consultants
Consultants9 1980b). Elsewhere along this linear depression 9 the
depression appeared to be underlain by bedrock and to represent a
glacial meltwater side channel.
In 1981 9 evidence strongly indicative of no fault control of the
lineament was observed where the lineament trends across a low
plateau in the southwest quadrant of the confluence of the
Portage Creek and the Susitna River. This area is designated
Segment 1 in Figure 4-22 and is referred to here as the Portage
Creek p 1 ate au.
The Mint Mine is located at the northeast margin of the Portage
Creek plateau (Location 04; Figure 4-22). Review of State of
Alaska claims records and discussions with miners show that the
claims were staked in 1922 and some silver has been mined from
the property intermittently. Bedrock in the mine is Cretaceous
slate. These slates have been intruded by andesite dikes along a
shear zone. The dikes are accompanied by extensive brecciation
and silicification. The mineralization appears then to be
related to the intrusive and/or shearing episodes. These
episodes can be inferred to have occurred in Tertiary time 9
approximately 50 to 60 m.y.b.p.9 if they are assumed to be
related to the same tectonic activity as the Tertiary volcanic
rocks and/or intrusive rocks in the region 9 as dated by Csejtey
and others (1978).
Immediately northeast of the mine is a northeast-southwest
trending tributary to Portage Creek. On the southwest wall of
this tributary is an exposure of bedded metasedimentary rocks
(Location 05; Figure 4-22). Feature KD5-3 projects directly
through this exposure. Detailed aerial examination of this
exposure was made from a helicopter (limited access and the
4 -47
Woodward-Clyde Consultants
steepness of the tributary walls precluded ground examination).
The beds were observed to dip southward into the tributary wall.
No displacement of these beds was observed (the resolution was
judged to be 12 to 20 inches (5 to 8 em).
Southeast of the Mint Mine, on the Portage Creek Plateau, two
magnetic profiles were conducted across the feature during
the 1981 study at Location 06 (Figure 4-22). The traverses
were conducted using the procedures described in Appendix A,
Section A.5. The magnetic signature across the zone where the
fault is projected was absolutely flat. No deviation in magnetic
signature was observed. From these data, it was concluded that
no magnetic anomalies are associated with this segment of the
feature. The absence of an anomaly is circumstantial evidence
that no fault is present.
Northeast of the Mint Mine, Csejtey and others (1978) shows a
14-mile (23-km) long inferred thrust fault segment near Thorough-
fare Creek (Figure 4-22). This fault segment coincides with the
Feature at this location. This fault segment is related to late
Mesozoic-early Tertiary tectonism (at least 60 to 70 m.y.b.p.).
No evidence was observed during this field investigation to
suggest that the fault has been subject to displacement since
that time.
Southeast of the Portage Creek plateau, the feature coincides
with either a glacial meltwater side channel along the south bank
of the Susitna River or with the margin of the Susitna River
floodplain. (The uncertainty results from the width of the trace
when it is transferred from the scale of Gedney and Shapiro
[1975] to a scale of 1:24,000.)
4 -48
Woodward-Clyde Consultants
On the basis of work conducted in 1981, we believe the floodplain
margin of the Susitna River is the best location of the lineament
drawn by Gedney and Shapiro (1975). No evidence was observed
along this floodplain to suggest that a fault is present or
that it has been subject to recent displacement. There are no
prominent scarps, there is no consistent offset of tributary
drainages to the Susitna River, nor is there any other morpho-
logic evidence of a fault.
The pronounced change in lithologic texture and color and
possible structural fabric in the vicinity of Curry reported in
the Interim Report (Woodward-Clyde Consultants, 1980b) was not
re-examined during the 1981 study. The clear evidence of no
fault control elsewhere along the feature from the Portage
Creek plateau renders this relationship inconsequential to the
evaluation of active fault potential.
A possible zone of sheared rock was reported in the Interim
Report (Woodward-Clyde Consultants, 1980b) to be present north-
east of the Portage Creek plateau. The exposure which contained
this zone of sheared rock was not observed in 1981. It app ar-
ently had been covered by slumped material or sediment deposited
by the tributary.
Feature KD5-3 was classified during this investigation as being a
lineament; therefore, it does not affect consideration of seismic
ground motion or surface rupture potential at either dam site.
The origin of the lineament is related to an alignment of
the Portage Creek and Susitna River drainages which locally have
linear glacial features. The lineament was judged not to be a
fault and is, therefore, not a fault with recent displacement.
This judgment concurs with that of Kachadoorian and Moore (1979)
who concluded that there was no evidence of active faulting along
Portage Creek.
4 -49
Woodward-Clyde Consultants
Feature KDS-9
Feature KDS-9 trends N55°W to N70°W for a distance of 2.5 miles
(4 km) and approaches within one mile (1.6 km) south of the
Devil Canyon site (Figure 4-22). The lineament initially was
identified on SLAR imagery by Gedney and Shapiro (1975).
Subsequent examination of U-2 photography during the 1980 study
showed the lineament to be expressed morphologically as a
linear alignment of a stream drainage, several small lakes, and
marsh 1 and. This feature consists of three segments, which are
discussed below. Due to the short length, the segments are
not labelled in Figure 4-22.
The eastern segment of the feature is a ravine with approximately
200 ft (60 m) of relief. Where the ravine intersects the west
bank of Cheechako Creek, exposures of Tertiary intrusive rocks
were examined during the 1981 field study. No evidence of a
fault was observed; however, the exposure is not continuous
across the projection of the lineament. Therefore, there is some
possibility of a fault being present but concealed beneath
vegetation and/or colluvium.
Joint measurements were taken in the eastern segment. Stereonet
plots of the joint orientation show dominant orientations of
N25°W, N55°W, and N65°W. The latter two orientations and the
advance of glacial ice through this region strongly suggest that
this segment of the lineament is the result of glacial fluting,
the orientation of the fluting being controlled by the ice flow
direction and the joint orientations.
Woodward-Clyde Consultants (1980b) reported a knickpoint in
Cheechako Creek which aligned with the feature. Detailed review
of the knickpoint (with waterfalls) during the 1981 study showed
4 -50
Woodward-Clyde Consultants
that the knickpoint is not aligned with feature KD5-9. It is
concluded, therefore, that the knickpoint is not related to the
feature.
The central segment of the feature is the northeastern shoreline
of two small lakes which lie in a depression whose northwest
margin is 1 i near. The area is generally marshy with standing
water. No evidence of fault control was observed during either
the 1980 or 1981 studies.
West of the intersection of feature KD5-9 with feature KD5-45
(Figure 4-22), a small outcrop is present which is overlain by
glacial deposits. The orientation of the schistosity within the
outcrop is parallel to the alignment of the feature. No evidence
of a fault was observed in the outcrop nor in the glacial
sediments overlying this area.
The western end of the lineament is a broad, shallow depression
on top of an elongate, broad rise. The depression is marked by
standing water and marsh. The margins of the depression are
bordered by spruce forest on the slightly higher ground. No
evidence of fault control was observed. The depression is
considered to be the result of glacial fluting.
Feature KD5-9 was classified during this investigation to
be an unrelated series of linear features whose origin is related
to glacial processes and to control by dominant joint and
foliation orientation. The feature is considered to be a
lineament; therefore, it does not affect consideration of seismic
ground motion or surface rupture potential at either site.
4 -51
Woodward-Clyde Consultants
Feature KD5-12
Feature KD5-12 trends N50°E for a distance of 14.5 miles (24 km)
and approaches within 1.5 miles (2.4 km) upstream of the Devi 1
Canyon site (Figure 4-22). The lineament initially was identi-
fied, in part, on SLAR imagery by Gedney and Shapiro (1975) as a
linear stretch of Cheechako Creek south of the Susitna River.
The 1 i neament was extended northward across the Susitna River;
this judgment was based on morphologic relationships observed
on U-2 photography during the 1980 study by Woodward-Clyde
Consultants (1980b). North of the Susitna River, the lineament·
is expressed in part as a linear depression in which lie several
small lakes, and in part as a linear stream drainage. This
depression cuts across the predominant structural grain of
this area.
Five locations (Figure 4-22) examined during the 1981 field
season clearly demonstrate that feature KD5-12 is not a fault,
and these five locations locally provide alternate explanations
for its origin. The segment of the feature northeast of the
Susitna River corresponds to the contact between the Tertiary
intrusive unit to the southeast and the Cretaceous metasediments
to the northwest (Figure 4-22). At location D7, Cretaceous
argillite has been intruded by the Tertiary pluton. There is no
evidence of faulting along this contact.
No morphologic evidence of a fault or structural control was
observed where the feature crosses the Susitna River (Figure
4-22, Location D8). Vegetation and soil cover the river banks at
the projection of the feature, so observations in bedrock could
not be made at the river crossing.
4 -52
Woodward-Clyde Consultants
The northeast wall of Cheechako Creek, at the projection of the
feature (Figure 4-22, Location 09), was examined on the ground
from a distance of 1,000 feet (305 m). No evidence of fault
control was observed in the Tertiary intrusive rocks; however,
the resolution of this observation is limited by the distance of
the observation and access limitations imposed by the canyon
walls.
At location 010 (Figure 4-22) in Cheechako Creek, the east bank
juts westward into the creek and has deflected the creek into a
horseshoe-shaped bend. Rock is continuously exposed on the
promontory and the west wall here where the feature is projected
to lie within Cheechako Creek. Examination of the east promon-
tory on the ground and the east wall, using binoculars, provided
evidence of no faulting through the exposure. The resolution of
this observation was judged to be 4 to 9 inches (10 to 15 em).
During the 1980 field study, a prominent shear zone was observed
from the air near the southwestern end of the feature at Location
011 (Figure 4-22). Mapping was conducted at the shear zone
during the 1981 field study. The results of the mapping confirm
that a 23-foot-(7-m-) wide shear zone is present in the Tertiary
intrusive rocks. The zone consists of highly altered and
decomposed rock which has been sheared and faulted along a
N35°W to N40°W trend. This trend is perpendicular to the N50°E
orientation of Feature KDS-12. From this orientation of the
shear zone, we concluded th~ it is not related to the feature,
and that there is no evidence of structural control at this
location.
The observations made at the five locations described above
provide the basis for our judgment that Feature KDS-12 is not a
fault. It is, rather, a series of unrelated linear features
4 -53
Woodward-Clyde Consultants
whose origin is related to a lithologic contact, or to the
alignment of discontinuous stream drainages. No evidence of a
bedrock fault was observed, nor was any morphologic evidence of
recent displacement observed. We classified Feature KOS-12
as being a lineament; therefore, it does not affect consideration
of seismic ground motion or surface rupture potential at either
dam site.
Feature KOS-42
Feature KOS-42 trends N50°W to N6s•w for a distance of 3 miles
(5 km) and approaches with in 0.5 miles (0 .8 km) south of the
Oevi 1 Canyon site (Figure 4-22). The feature was originally
identified as a lineament on U-2 photography by Woodward-Clyde
Consultants during the 1980 field study. The lineament is
comprised of two aligned stream drainages. The eastern drainage
trends N65°W and the western drainage trends N50°W.
Information from three locations was used to evaluate this
feature. The eastern location, 012 (Figure 4-22), is an exposure
of argillite within the Cretaceous metasedimentary sequence. The
outcrop showed no evidence of faulting with the resolution
described below. A foliation or indistinct bedding trend was
observed in the argillite. The orientation was uniformly N20o E
to N60°E. The presence of this foliation or bedding trend
provided a high degree of resolution in assessing the absence of
a fault. Tempering this resolution was an intermittent cover of
vegetation and surficial materials which provided an overall
resolution at the outcrop of at least 6 feet (2 m).
Location 013 (Figure 4-22) is a linear trough several hundred
feet (200 m) long and 200 feet (60 m) wide, which has been cut
into the Tertiary intrusive rock. The feature lies in the middle
4 -54
Woodward-Clyde Consultants
of this trough. A dike has been intruded into the Tertiary
intrusive in this trough. The dike is oriented approximately
perpendicular to the axis of the trough, but it is not exposed on
the bottom of the trough. Measurements taken on the dike show it
to have a uniform strike of N50°E. This strike can be projected
across the trough and shows evidence that it has not been
displaced. The accuracy of the alignment was judged to be 2°;
so the resolution in the absence of displacement is approximately
6 feet (2 m).
The feature projects across Cheechako Creek, and there are
limited exposures in the vicinity of that projection. Steep
canyon walls and limited access in the creek bottom precluded
gro;und examination of the exposures. From the air, however,
there appeared to be a contact between the Cretaceous metasedi-
mentary sequence on the north and the Tertiary intrusive unit to
the south. The orientation of the contact was not visible.
Consequently, the contact may coincide with the feature, but this
interpretation remains to be confirmed.
No evidence of a fault was observed in the walls of Cheechako
Creek. However, the limited extent of exposures and the inabil-
ity to review the available exposures on the ground limit the
confidence in this observation.
The feature projects into the south wall of Devils Canyon at
location 015 (Figure 4-22). These walls have been examined from
\,. the air by geologists for both Woodward-Clyde Consultants and
Acres. In addition, Acres• geologists mapped the walls using
rock-climbing methods. No evidence of a fault was observed in
the canyon walls. A near vertical joint was observed in the
wall, which is infilled with what appeared to be a dike that
is 1 foot (30 em) wide.
4 -55
. ~ ..
Woodward-Clyde Consultants
Feature KD5-42 was classified during this investigation as
being a series of short lineaments which originated from glacial
enhancement of surface morphology and stream drainages. The
feature was judged to be a lineament; therefore, it does not
affect consideration of seismic ground motion or surface rupture
potential at either site.
Feature KD5-43
Feature KD5-43 trends N80°E for a distance of 1.5 miles (2.4 km)
and passes through the left abutment of the Devi 1 Canyon site
(Figure 4-22). The lineament is expressed morphologically as a
short, prominent depression, approximately 300 feet (91 m) wide,
which is oriented p·arallel to the Susitna River. Within the
depression are two small lakes with a low saddle of glacial
material between them.
The depression associated with the lineament was considered to be
a potential spillway during initial feasibility studies conducted
by the U.S. Bureau of Reclamation (USBR) in 1957 and 1958 (U.S.
Bureau of Reclamation, 1960). During the USBR study, five
borings were drilled across the depression on the saddle between
the two lakes. An additional boring was drilled on the southwest
shore of the eastern 1 ake, and a test pit was excavated in the
saddle near the northwest shore of the eastern lake during this
study.
During the 1980 feasibility study, Acres drilled an angle boring
(BH-4) southward from the north shore of the eastern lake. The
boring was dri 11 ed beneath the 1 ake for a distance of 501 feet
(153 m) across the axis of the depression. In 1981, Acres
drilled an angle boring (BH-7) northward from the south shore of
the eastern lake. The boring was drilled beneath the lake for a
distance of 498.3 feet (151.9 m).
4 -56
Woodward-Clyde Consultants
The results of this drilling program were that no significant
fault zone or shear zone was encountered in boring BH-4. In
Boring BH-7, a zone of brecciated rock with zones of clay
(gouge?) and slickensided surfaces was encountered beneath the
eastern lake. One of the borings drilled in the center of the
buried valley during the USBR study (DR-6) encountered 11 Sheared
rock 11 for the 20-foot (6-m) distance that the boring was drilled
in rock.
In 1978, Shannon and Wilson conducted a seismic refraction
traverse (along the saddle between the two lakes) across the
feature for the U.S. Army Corps of Engineers (1979). As part
of this feasibility study, Woodward-Clyde Consultants (1980a)
conducted two north-south seismic refraction traverses across
the eastern lake and a northwest-southeast traverse at an
oblique angle to the north-south traverses and the axis of the
depression.
The data obtained from these studies show that a buried bedrock
channel is present beneath the eastern part of the depression.
The channel has a maximum depth of approximately 90 feet (27 m)
and is filled with 80 feet (24 m) of sand and gravel (glacial
outwash) which is overlain by approximately 10 feet (3 m) of
silt, sand, gravel, and cobbles (glacial till).
The faulted or sheared rock encountered in Boring BH-7 (and to a
lesser extent in DR-6) suggests that a fault or shear zone may
underlie the feature. However, review of the core obtained
from BH-7 suggests that it does not necessarily represent a
major throughgoi ng structure. The shearing and f ault-re 1 ated
features are not dissimilar from those which would be expected in
rocks which have been subjected to sever a 1 periods of tectonic
activity. They could represent a throughgoing fault; although it
is not clear that they do.
4 -57
Woodward-Clyde Consultants
If the sheared zone encountered in boring BH-7 does represent a
fault which controls feature KD5-43, the fault dips southward.
The absence of the shear zone in boring BH-4 precludes its
dipping northward toward the river.
Ground reconnaissance studies conducted along the feature during
this investigation in 1980 included fracture analyses in bedrock
on both sides of the depression and ground traverses of the
saddle between the two lakes. The fracture analyses showed
that fractures on both sides of the depression have similar
orientations. The dominant orientation is N35°W with a steep
northeast to southwest dip. Thus, there is no major disruption
of joint attitudes from one side of the feature to the other.
The canyon wall of Cheechako Creek at the east end of the
lineament was examined from the air. No evidence of faulting was
observed. No evidence of di sp 1 acement was observed from the
air on the Susitna River canyon wall at the west end of the
1 i neament. However, access 1 imitations and vegetation cover
limit the confidence in the interpretations on both canyon
walls.
Considering the above information and data, the depression
associ a ted with lineament KD5-43 appears to be a meltwater
side-channel that may be structurally controlled. According to
this interpretation, the depression may have developed from
differential erosion along a prominent structure such as a
fracture zone or bedrock fault. Subsequent glacial and/or
meltwater processes served to enhance and probably deepen the
depression, and it was later filled with sediments during a late
glacial episode (perhaps in Late Wisconsin time).
4 -58
Woodward-Clyde Consultants
If it is conservatively assumed that the feature is a fault~
then the question about the recency of di sp 1 acement needs to
be addressed. Given the short length of the feature and the
tectonic setting of the region (the only observed faults with
recent displacement in the site region were the Talkeetna Terrain
boundary faults~ as discussed in Section 4.5), we consider it
unlikely that Feature KD5-43 is the only fault with recent
displacement in the region. We~ therefore~ consider it to be
extremely unlikely that Feature KD5-43 is a fault with recent
displacement.
This judgment is based on the extremely short length of the
feature and the tectonic setting of the region. Experience in
other parts of Alaska and other areas of the world~ including
California~ Japan~ and South America~ suggests that short faults
with recent displacement exist in association with other longer
faults for which recency of displacement is clearly recognizable.
That is~ if a region is being subjected to stresses that are of
sufficient magnitude to cause surface faulting~ it is extremely
unlikely that strain release in the region would produce no
surface rupture longer than along one fault a few miles (a few
kilometers) in length. Rather, it is logical to expect that
the strain release would cause extensive rupture during some
earthquakes and smaller amounts of rupture on both the main
faults and shorter subsidiary faults.
Feature KD5-43 has been classified as being a possible fault.
There is circumstantial evidence of a fault zone in the sub-
surface; however~ there is no evidence of a fault on the canyon
walls along the projection of the feature. We have concluded
that the feature is a fault without recent displacement; there-
fore, Feature KD5-43 does not affect consideration of seismic
ground motion or surface rupture potential at either site.
4 -59
Woodward-Clyde Consultants
Feature KD5-44
Feature KD5-44 trends Nl5•w for a distance of 21 miles (34 km)
and approaches within 0.3 miles (0 .5 km) upstream of the Devi 1
Canyon site (Figure 4-22). The feature initially was identified
south of the Susitna River as two discontinuous lineaments on
SLAR imagery by Gedney and Shapiro (1975). One of the lineaments
followed, in part, the northern end of Cheechako Creek, whose
confluence with the Susitna River is immediately upstream from
the Devil Canyon site. Air photo interpretation conducted during
the 1980 investigation identified a lineament with a similar
alignment along a stream drainage whose confluence with the
Susitna River is opposite that of Cheechako Creek.
During the 1980 field study, it was the opinion of the Woodward-
Clyde Consultants geologists that the two lineaments identified
by Gedney and Shapiro (1975) and the lineament identified by
Woodward-Clyde Consultants should be considered as a single
feature (Woodward-Clyde Consultants, 1980b). The 1981 field
study and the subsequent analysis of the feature considered the
feature to be a single lineament, 21 miles (34 km) long.
The feature is expressed morphologically as a linear stream
drainage north of the Susitna River. On the south side of
the Susitna River, the feature is the northern segment of
Cheechako Creek and a tributary to Cheechako Creek. South
of this tributary, the feature is a shallow, broad, linear
depression on the upland plateau which lies south of the Susitna
River (Figure 4-22).
Three locations were studied in detail in 1981 to determine if
Feature KDS-44 is a fault and whether it is a fault with recent
displacement. Location D15 (Figure 4-22) is on the north side of
4 -60
Woodward-Clyde Consultants
the Susitna River opposite the confluence of Cheechako Creek.
Here an oxidized mafic dike is exposed in the canyon wall.
Aerial measurements made during the 1981 field investigation show
that the dike has a northwest strike and an apparent dip of 70°
to 80°NE. The dike can be traced for a distance of approximately
300 feet (91 m) and has an apparent maxi mum width of approxi-
mately 20 feet (6 m). The dike was inaccessible on the ground
during the 1981 field study. Observations made from the heli-
copter strongly suggest that the dike dies out at its eastern
end. At its western end, it may die out, or it may cant inue on
strike across the drainage associated with Feature KD5-44. The
reason for the apparent ambiguity is due to the orientation of
the exposure relative to the dike and the inaccessibility of the
exposure. The dike visually appears to wedge out, or die out
on the east bank of the drain age. However, during aeri a 1
examination of the west bank, a dike was observed on strike with
the main dike. It is not clear if the two dikes are segments of
a single dike or if they are completely separate.
Whether or not the dike dies out on the east bank of the tribu-
tary or continues across the tributary is not significant to
the evaluation of Feature KD5-44. Either relationship provides
circumstantial evidence that the dike is not truncated by a fault
associated with Feature KDS-44.
Location 016 is the large point bar which juts into the Susitna
River upstream from Devil Canyon (Figure 4-22). Seismic refrac-
tion studies were conducted across the point bar by Shannon and
Wilson in 1978 for the U. S. Army Corps of Engineers (1979). The
results of the study suggest that a buried step or scarp in
bedrock steps from a depth of approximately 100 feet (30 m) below
the point bar (on the downstream side) to a depth of 300 to
4 -61
Woodward-Clyde Consultants
330 feet (91 to 100 m) on the upstream side. On the basis of
these two seismic refraction lines, the buried step can be
inferred to have a buried relief of approximately 200 to 230 feet
(61 to 70 m). Its base is oriented N10°W, subparallel to the
trend of Feature K05-44. The slope dip is approximately 20°NE.
The southwest side of the step is up relative to the northeast
side.
The presence of the buried bedrock step beneath the point bar is
an anomaly. Its origin is open to question at this point. It
could represent a fault scarp which was modified by glacial and
fluvial processes prior to burial. However, the absence of a
fault northwest and southeast of Location 015 strongly suggests
that it is not part of a throughgoing fault. It may be the
product of differential erosion along a joint system. Evaluation
of the glacial history of the region (Section 3) has provided no
compelling explanation to date.
In summary, the origin of the buried bedrock step at Location 015
remains an enigma. It does not, however, represent a through-
going fault scarp, in our opinion, because there is no evidence
observed of faulting northwest and southeast of this location.
In the Interim Report (Woodward-Clyde Consultants, 1980b), zones
of light-colored, fractured, and highly weathered rock in
Cheech ako Creek were described. The origin of the zones was
suggested to be possibly related to faulting. The walls of
Cheechako Creek were mapped under the direction of Acres during
the winter of 1980-1981, and no evidence of a throughgoing fault
system was observed (Bruen, 1981). We conclude then that these
zones of rock are not related to a throughgoing fault.
4 -62
Woodward-Clyde Consultants
The location with the most compelling evidence that Feature
KD5-44 is not fault-controlled is Location D17 (Figure 4-22).
Here, Tertiary intrusive rock is exposed across the entire zone
in which the feature is projected. The rock is exposed on the
banks and in the channel of a tributary to Cheechako Creek.
There is clearly evidence of no fault dis-placement· in the
exposure with the resolution being approximately one inch
(2 to 3 em).
In the Interim Report (Woodward-Clyde Consultants, 1980b) an
apparent morphologic anomaly was reported a long Feature KD5-44.
A linear shallow depression, approximately 500 feet (152 m) long
was observed in a terrace. During the 1981 field study, the
origin of the terrace was reviewed as a part of the Quaternary
geology studies. It has been judged to be a glaciofluvial
deposit associated with a glacial stade following early Late
Wisconsin time. The linear depression is a drainage channel
which was subsequently cut into the terrace.
Feature KD5-44 was classified during this investigation as
being a series of unrelated lineaments whose origin is related to
the alignment of stream drainages. The feature is a lineament;
therefore, it does not affect consideration of seismic ground
motion or surface rupture potential at either site.
Feature KD5-45
Feature KD5-45 trends N80°E to N85°E for a distance of 19.5 miles
(31 km) and approaches within 0.8 miles (1.3 km) of the left
abutment of the Devil Canyon site (Figure 4-22). The lineament
was identified during the 1980 field investigation as a prominent
north-facing linear bluff along the south bank of the Susitna
River. Aligned with this bluff is a small, linear stream
4 -63
Woodward-Clyde Consultants
drainage at the west end of the lineament, a linear topographic
depression along the eastern portion of the lineament, and
several small lakes along the lineament.
Ground and aerial reconnaissance conducted during the 1980 field
study showed that the lineament corresponds primarily to the
front of the hills (i.e., range-front) along the south bank of
the Susitna River and locally is expressed as a linear trough
approximately 150 feet (46 m) wide and 10 feet (3 m) deep. The
linear trough is underlain by argillite and glacial till. Water
was observed flowing at a rate of approximately 3 to 5 gallons
per minute ( 11 to 19 liters per minute) out of the till at the
base of the trough. Reexamination of the till during the 1981
field study showed that the flowing water was emanating from
the contact between the till and the underlying argillite.
Thus, the water is flowing locally along the top of the bedrock
surf ace.
Detailed aerial reconnaissance during the 1981 field study
disclosed no anomalous morphologic relationships that required
ground-checking. There are no exposures a long the base of the
range front which could be examined for information on the origin
of the feature.
The origin of Feature KD5-45 is that of a range front which has
been modified by glacial processes. There is a conspicuous
absence of faceted spurs, scarps, and other morphologic features
representative of a fault origin. We have concluded that the
feature is a lineament; therefore, it does not affect considera-
tion of ground motion or surface rupture potential at either
sit e.
4 -64
Woodward-Clyde Consultants
4.5 -Assessment of Recent Displacement
Faults for which evidence of recent displacement was found were con-
sidered to be potential seismic sources. Each potential seismic source
was evaluated during this study to estimate its potential seismic ground
motions at the Watana and Devil Canyon sites and its potential for
surface rupture within 6 miles (10 km) of the sites.
On the basis of the 1980 study, the Talkeetna Terrain boundary faults
were identified as seismic sources that need to be considered as
potential sources of seismic ground motion at the dam sites. These
include: the Castle Mountain fault, the Denali fault, the Benioff
interplate region, and the Benioff intraplate region (Figure 1-1).
These faults are considered to be or to contain faults with recent
displacement that could cause seismic ground motions at the dam sites;
however, because of their distance from the sites, these faults do not
have the potential for rupture through the sites (Table 4-3). The 1980
study also identified 13 features nearer the dam sites that required
detailed evaluation during the 1981 study to assess their importance for
seismic design.
On the basis of the 1981 study, no evidence for faults with recent
displacement other than the Talkeetna Terrain boundary faults has been
observed to date within 62 miles (100 km) of either dam site and none of
the 13 features near the dam sites are faults with recent displacement.
Therefore, the 13 features are not considered to be potential seismic
sources that could cause seismic ground motions at the sites or surface
rupture through the sites (Table 4-3).
Our interpretation that none of the 13 features are faults with recent
displacement is based on data collected during our investigation. The
data are limited in the sense that a continuous 100,000 year-old
4 -65
Woodward-Clyde Consultants
or surface was not found along the entire length of each of
features. For this reason, the available data were analyzed and
rofess ion a 1 judgment was app 1 i ed to reach conclusions concerning the
ency of displacement on each of the 13 features.
4 -66
Woodward-Clyde Consultants
OF SELECTED WORLDWIDE EAR
Location
1tude1
)
w. Nelson, New Zealand 7.6
Izu, Japan 7.0
Hawkes Bey, New Zealand 7.8
N1sh1-Sa1tama, Japan 6.75
7.0
Wairoa, New Zealand 6.8
Taiwan 7.1
Ituango, Coluntlla 6.25
Tot tori, Japan 7.4
San Juan, Argent ina 7.6
Mikawa, Japan 7.1
British Coluntlia, 7.2
Vancouver Island
Kern, California 7.7
~ 7.2
Japan 5J
Central Alaska Mb 7.3
Kit a Mino, Japan 7.0
Buyln-Zara, I ran 7 .25
Truckee, California ~ 5 J5
Saltama, Japan 5 .a
Meckering, Australia 7.0
R~art, Alaska 6.5
Parlahuanca, Peru 5.7
Mb 5.9
Pariahuanca, Peru 6.4
6.2
Sylmar, California 6.6
~ 6.4
Point Mugu, California ~ 5.75
6.0
Friuli, Italy 6.4
San Juan, Argent ina
Izu-Oshima, Japan
Coyote Lake, California
Livermore, California
El Asnam, Algeria
7.4
6.8
5.7
~ 5.9
~ 5.5
7.3
UAKES
Act1ve2
Tectonic
Terrain
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes(?)
Yes
Yes
Yes
Yes
Yes
(1)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Distance to
Fault with
Late Quaternary
0
0
0
0
0
0
(1)
0
0
0
0(1)
0
0
(1)
0
0
0
(1)
0
0(1)
0
0
0
0
0
0-5(1)
0
0,
0
0
Yes
Yes
Yes
No
Yes
Yes
(?)
Yes
Yes
Yes
(?)
Yes
No
No
No
Yes
Yes
No(?)
Yes
No
Yes
Yes
Yes
No
Yes
No
Yes
Yes
Yes
Yes
1s expressed as M5 unless otherwise Indicated. Some earthquakes are assigned more than
Ms value In the literature by different authors.
lve tectonic terrain Is discussed In Section 4.2.1.
References
Slemmons (1977b)
Slemmons (1977b); Kuno (1936);
Matsuda (1977)
Adams and others (1933); Wyss (1979);
Slemmons (1977b)
Kanamor1 and Anderson (1975)
Matsuda (1977); Abe (1974)
!ida (1965); Richter (1958)
Slemmons (1977b)
Woodward-Clyde Consultants (1979)
Kanamor1 (1972); !Ida (1965)
Slemmons (1977b); !Ida (1965)
A.ndo (1974); I ida (1965)
Rogers and Hasegawa (1978);
Slauson and Savage (1979)
Gutenberg (1955); Dlbblee (1955);
Dunbar and others (1980)
Okada and Ando (1979)
Davis (1960); Sykes (1971)
Bonilla (1979); Kawasaki (1975)
Antlraseys (1965); Slemmons (1977b)
Ryall and others (1968); Tsai and
Aki (1970)
Kanamor1 and Anderson (1975);
Abe (1975)
Everingham and others (1969)
Lander (ed.) (1969); Gednet and
others (1969); Huang (1981)
Lander (ed.) (1970)
Lander (ed.) (1970)
S 1 ngh and others ( 1980) ;
Langston (1978); Yerkes (1973)
Castle and others (1977);
Boore and Stierman (1976)
Cagnett i and Pasquale (1979);
Cipar 1980); Clpar (1981);
Flnett i and others (1979)
Rojan and others (1977)
Shlmazak i and Somerville (1978)
Lee and others (1979);
Urhammer (1980)
Bolt and others (1981)
Burford and others (1981)
Woodward-Clyde Consultants
Y OF SURFACE FAULTING IN CALIFORNIA
ercent of
Number of Events Events with
Total Number With Surf ace Surface Faulting
Range of Events Faulting (%)
A. 1900 through 1969
> 5.0 287 8 3
> 6.0 34 5 15
B. 1970 through 1980
> 5.0 24 8 33
> 5.5 8 6 75
> 6.0 3 3 100
Woodward-Clyde Consultants
TABLE 4-3
SUMMARY OF BOUNDARY FAULTS AND SIGNIFICANT FEATURES
Distance 5
mi 1 es (km) from
Feature1 •2 Fault (F)
Length 4
Feature Name 3 or Linea-Devil No. ment (Ll miles(km) Can:r:on Watana
BOUNDARY FAULTS
ADS-1 C-astle Mountain F 295 (475) 71 (115) 65 ( 105) Fault
HB4-l Denali Fault F 1,358 ( 2 ,190) 40 (64) 43 (70)
Benioff Zone F 434 (700) 56 (91) 40 (64) (Interplate)
Benioff Zone F 46 {75) 38 (61) 31 (50) (Intraplate)
WATANA SIGNIFICANT ·FEATURES
KC4-l Talkeetna Thrust F 78 ( 126) 16 (25) 4 (6.5)
KD3-3 Susitna Feature L 95 ( 153) 16 (25) 2 (3.2)
KD3-7 Wat ana River L 31 (50) 22 (35) 0 (0)
KD4-27 Fins Feature F 2 (3.2) 23 (37) 0 (0)
DEVIL CANYON SIGNIFICANT FEATURES
KCS-5 L 12 (19) 4.5 (7) 19 (31)
KDS-2 F 0.8 (1.3) 3.5 (5.6) 24 (38)
KDS-3 L 51 (82) 3.6 (5.8) 14 (23)
KDS-9 L 2.5 (4) 1 ( 1.6) 24 (39)
KDS-12 L 14.5 (24) 1.5 (2.4) 17 (28)
KDS-42 L 3 (5) 0.5 (0.8) 22 (35)
KDS-43 L 1.5 (2.4) 0 (0) 24 (38)
KDS-44 L 21 (34) 0.3 (0.5) 23 (37)
KDS-45 L 19.5 (31) 0.8 (1.3) 35 ( 41)
~
1. Alpha-numeric code number is based on: a) a first letter designation for the
1:250,000 quadrangle where the feature is located (A = Anchorage, H = Healy,
K = Talkeetna Mountains); b) the letter and number of the 1:63,380 quadrangle
at the midpoint of the feature; and c) a number designating the order of the
feature's recognition.
2. Feature locations are shown in Figures 4-3, and 4-18 through 4-20.
3. Feature name is given where assigne9.
4. Length is from the text. ·
5. Distance is the closest approach of the surface trace of the fault or
lineament as measured on the base maps referred to in Note 2.
Fault with
Recent
Dis~lacement
Yes
Yes
Yes
Yes
No
No
No
No
No
No
No
No
No
No
No
No
No
so•
LE GEND
_,u
--D
----
..........
TERRAIN
• Gl ennallen
---®
--....;,;; .................
A. TALKEETNA TERRAIN MODEL
Mapped str ike-slip fault, arrows show
sense of horizontal displacement
Mapped strike-slip fault with dip slip
compon ent, letters show sen se of
ver~ical displ ar:ement: U i s up ;
D is down.
Mapped fault , sense of horizontal
displacement not defined
Inferred strike-slip fault
Mapped thrust fault, sawteeth on
upper plate
41410A Febru ary 198 2
NOTES
CD 0.9 -2.0 cm /yr Hickman and Campbell ( 1973); and Page (! £'12 ).
@ 0.5 -0.6 cm/yr Stout and others (1973).
@ 3. 5 cm /yr Richter and Matson ( 1971 ).
@ 1.1 cm /yr, no Holocene activity farther east, Richter and Matson ( 1971 ).
® 0.9 -3.3 cm /yr Richter and Matson (1971 ).
® Inferred connection with Dalton fault; Plafker and others ( 1978).
(j) Inferred conn ection with Fairweather fault ; Lahr and Plafker (1980).
@ Connection inferred for this report .
® 0 .1 - 1.1 cm /yr Detterman and others (1974); Bruhn (1979).
@ 5.8 cm /yr Lahr and Plafker (1980).
Q]) Aleutian Trench and Postulated Shelf Edge Structure afte!' Guptill
and others (1981 ).
@ Slip rates cited in notes CD through @ are Holocene slip rates .
@ All fau It locations and sense of movement obtained from Beikman ( 1978; 1980).
Ploue Mot on Relative to North Amer1can Plate
B. SCHEMATIC TALKEETNA TERRAIN SECTION
TALKEETNA TERRAIN MODEL
AND SECTION
0 100 200 Miles
~=EF£003~31~~~~~
0 100 200 Kilometers
FIGURE 4-1
z
Ul c
r
--i
)> z
--i
Ul
,
"' CT
2
!?;
<
ABSOLUTE AGE ERA PERIOD EPOCH
(In Million s of Years Before P rus en t
HOLOCEI\!f
--0.01---QUATERNARY (RECENT
--1.8 --PLEISTOCE!'JE -u PLIOCENE
0 NEOGENE N MIOCENE 0 --22.5 ---z
LU TERTIARY OLIGOCENE u
PALEOGENE EOCENE
PALEOCENE --65 --u CRETACEOUS
--141 ---0
N
0 JURASSIC --195 ---(/)
LU TRIASSIC :2 --230 --
PERMIAN --280 ---C: ., PENNSYLVANIAN 0 ':J .c 0 u .... ....
"' Q) MISSISSIPPIAN 0 (.)!'= --345 ---N
0 DEVONIAN --395 ---LU
..J
4: SILURIAN --435 ---0..
ORDOVICIAN --502 --__,
CAMBRIAN
--572 --
' HADRYNIAN
--1000 ---a::u
z LU_ r-o HELIKIAN 4: ON --1800 ---a: a:: a
CD 0.. APHEBIAN --2600 ---:2
4: u
LU u a: ARCHEAN u u
0.. 0
N 0 0
0 N N
4 --' LU 0 0 ----500 ..J (/) z
4: LU LU
0.. :2 u PROTEROZOIC
2600 572 230 65 0
BAR SCALE
(In Millions of Y ears Before Present)
STRATIGRAPHIC UNITS \ GEOLOG\C EVENTS
Colluvium, volcanic ash. fluvial sediments
Till, lakebeds, outwash sediments Glaciation, uplift of Alaska Range
Watana Creek sediments folded
Faulting oriented NE and NW Watana Creek shale, coal, conglomerate, fo rmation of sedimentary basins sandstone
Fog Creek and Talkeetna Hill andesite
-crranitic plutonism, subaerial vulca
Talkeetna thrust fault formed,
or basalt, granitic intrusive rocks accretion of island arc, or collisio
n i
n of
_QQ[ltinental and oceanic plates
Maclaren terrane argillite and graywacke Flysch deposition
Wrangellia terrane argillite and
metaconglomerate
Wrang'ell ia terrane meta-basalt, marble Subaerial vulcanism
Wrangellia terrane meta-andesite Island arc vulcanism .
NOTES
1. Time scale is after Van Eysinga (1978).
2. Stratigraphic units and geologic events are from sources identified in
Section 4 .
GEOLOGIC TIME SCALE
oenali Fault
.,;·
• Cantwell
MAC LAREN TERRANE ·
630
Devil Canyon
Site
. . :~ .
.·· ~, .....
<(~ .~ .. . ~:~~ ~ ~ ~
A .~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ <:'>., .................. .
~~~( ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Watana Site .~··:::::::::::::::::::: • . So. ·: . : : : : : : : : : : : . : : : : : : : "I ·.. 'S/fr,. . . . . . . . . . . . . ...
... ~ s~·J~~~···Il u i~'Je, n
. ~( ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ : : ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ; ; ........... · .......................................................... .
~2°30~····· ··········•····••••·• ···••···~~~~ri~tL~~ J A . ::~ ~~~ H [~ T [:. [ [H ~ [ [ [ ~ [i H ~ H ~ [ [ [ [ [ [[ [:::::::::::::::: :::::::::::::::::::::::::
·::::::::::::::::: :::·:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::.
Talkeetna : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : ::: : : : : : : : :: :: : : : : : : : : : : : : : : : : : : : : : : : : : I .... ::::::::::::: ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Sir ;;~~ ;; ~;; ~ ~ ~ ~ ~ ~ ; :; ; : ·; 4'9b ; ; ~ ~ ~ ~ ~ ~ ~ ~; ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~; ~ ~ ~ ~ '1'48°" ~ ~ ~ ~ ~; ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ............... ........ ··················-················ .................... . ............... ........ ................................... .................... .
LEGEND
NOTE
Thrust fault, dashed where inferred,
dotted where concealed , sawteeth on
upper plate
Inferred terrane boundary
1. Ge ology is from Csejtey and others (1978),
Nok leber g and others ( 1981 ). and Turner
and Smith (1974).
2. Terranes are desc ribed 10 Section 4.4.1
································· ..... ·······-························· ·····
································· .... . . .................................... . ...................................... ································· .....
TERRANE ........ ........
............................. ... .... . ............................. ... .... . ............................. ... .... . . ................................... .
····························· ....... . ····························· ....... . ............................. ... ..... .....................................
~ ~ ~ ~ ~ ~: i"4 j O ~; ~::; ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ········-··················· ....... . . ........ ··················· ....... .
··········· ···········
··········-···········
REGIONAL TECTONIC TERRANE MAP
0 10 20 30 40 Miles
?43 §I
0 10 20 30 40 Kilometers
OQowAR D·CL Y DE CONSULT ANTS 4141 OA February 1982 FIGURE 4 -3
LEGEND
W4.
, .. ~·
-···-
NO TES
Location and age designation of Quaternary
units (0 to 1.8 m.y.b.p.)
Location and age designation of Tertiary
units (1.8 to 65 m.y.b.p.)
Location and age designation of Mesozoic and
Paleozoic units (65 to 570 m.y.b.p.)
Thrust fault, dotted where concealed ,
sawteeth on upper plate
Lineament
1· m.y.b.p . is the abbreviation for million years
before present.
2· Geologic re lations at locations W2, W3, etc. are
discussed in Section 4.4 .1.
3· Lineam ent and fault locations are modified from Csejtey
and others (1978), and Turner and Smith (1974).
TALKEETNA THRUST FAULT AND
SUSITNA FEATURE LOCATION MAP
0 10 20 40 Miles ~~Gs=31~~~ ~~-
o 20 40 Kilometers
~DW~A~~------------------------------------------------------------------------~ RD·CLYDE CONSULTANTS 41410A February 1982 FIGURE 4-4
NNW
WINDY
CREEK
Os
I
1. ..
. /.. ... . ···t I
X ":.i• .. • . ::::::: ... ,z ~ ~ ~ ~ ~ ~ ~ ~
,,, .. :_::{j jjT jjjj
.!#~~~~~~~~~~~~~
·· 1/J.<HTTHH
/::::::::::::::::::
::::::::::::::::::
·/:::::::::::::::::::~
/ ...................... .
. · :::::::::::::::::::::: . .........................
LEGEND
Os
KJag
llv
\
~~
\
NOTES
Quaternary glacial sediments
Cretaceous and Jurassic meta-
sedim entary rocks
T ri assic vo lcan ic rocks
Th ru st faul t , infe rr ed from
lithologic contrast, arrows
show sense o f dip slip di s -
pl ace me nt
1. Section location is shown in
Figure 4-4 .
2. Geology is from Woo dward-Clyde
Consultants' 1981 f ie ld study and
Smith (1981).
SSE
-6,000
TALKEETNA THRUST FAULT
..
., -2,800
DIAGRAMMATIC CROSS -SECTION OF
THE BROXSON GULCH THRUST FAULT
AT WINDY CREEK (LOCATION W1}
0 2 Miles
E=1 ~~~~~3l -~ I
0 2 Kilom et e rs
.:= -c
0
·~
(1:)
> Q.)
w
w~o~w~~--------------------------------------------------------------~~~~~ ARD-CLYDE CONSULTANTS 41410A F ebruary 1982 FIGURE 4-5
NW
TALKEETNA
THRUST FAULT
BUTTE CREEK
LEGEND
Os
Kag
KJs
F!v
"A Ps
NOTES
Quaternary glacial sediments
Cretaceous metasedimentary
rocks
Cretaceous and Jurassic meta-
sedimentary rocks
Triassi c volcanic rocks
Triassic and Permia n sedimentary
rocks
Thrust fault, inferred from
lithologic contrast, arrows show
sense of dip slip displacement
1. Section location is shown in Figure 4-4.
2. Geology is from Woodward-Clyde Con-
sultants' 1981 field study and Csejtey
.(1981).
SUBSIDIARY
THRUST
FAULTS
SE
6,000
-2,800
DIAGRAMMATIC CROSS-SECTION OF
THE TALKEETNA THRUST FAULT
NEAR BUTTE CREEK (LOCATION W2l
OEI ===J~~~==!~~~~~===31 Miles G I
0 2 Kilometers
c::
.2 ...
ctJ
~
w
~ow~~~~~~~~;-~~~~--~~------------------------------------------~ ARD-CLYDE CONSULTANTS 41410A February 1982 FIGURE 4-6
---
D __.::::..-. --u~
Quaternary glacial sediments
Quaternary colluvium and talus
Tertiary granod i orite
Tertiary andesite or basalt
Cretaceous argi II ite
Paleozoic basalt, greenstone
Paleozoic li mestone
Lithologic contact, dashed where
approximate l y located
Fault, dashed where approximately
located, dotted where concealed;
arrows show sense of horizontal
disp l acement, letters show sense of
vertical displacement: U is up;
D i s down .
\' Trench located between arrows
'ooowAR D-CLYDE CONSULTANTS 41410A February 1982 --------
NOTES
1. Geology is from Woodward-Clyde Consultants ,
1981 field studies and photogeology, and
Csejtey and others ( 1978).
2. Geologic relations shown in this figure are
discussed in Section 4.4.1.
GEOLOGIC MAP OF LOCATIONS
W8, W9, AND W10 NEAR TALKEETNA HILL
0 2 Miles
I
0 2 K i lometers
FIGURE 4-7
NW
Kag
... ~ ...
'
........ . .........
f ~~~~~~~ ......... 1 ........ .
f:::::::
f~~~~~~~ , ........ . . . . . . . . .
Jl-.r·::::::: ......
I ...... .
·. 1: Hj)
c~~~~~
f~~~~~~ !:······· to"'''''
t ..... .. t ...... .
···-··· f ...... .
LEGEND
Tv
t<ag
Pzv
NOTES
Tertiary volcanic rocks
Cretaceous metasedimentary rocks
Paleozoic volcanic rocks
Thrust fault, inferred from lithologic
contrast, arrows show sense of vertical
displacement
1. Section location is shown in Figures 4-4
and 4-7.
2. Area labeled "Tv" represents a hill of
Tertiary volcanic rock into which the
Talkeetna thrust fault projects. The
hill (location W8 in F1gure 4-4) is.
separated from Talkeetna Hill by a
narrow valley represented 10 the figure
as a space.
3. Geology is from Woodward-Clyde
Consultants' 1981 field study and
Csejtey (1981 ).
""'oowi\RD-CLYDE CO NSU LTANTS 4 1410A February 1982
SE
-3,600
TALKEETNA THRUST
FAULT
:Pzv
rTALKEETNA
RIVER
-800
-~
t:
0 ..... ca > Q.)
LU
DIAGRAMMATIC CROSS-SECTION OF
THE TALKEETNA THRUST FAULT
AT TALKEETNA HILL (LOCATION W9)
0 2 Miles
I
0 2 Kilometers
FIGURE 4-8
NW t----Limits of deformation within the Talkeetna
thrust fault as shown in Figures 4-4 and 4-7.
SE
---,-2700
1 Trench T2 1
~H~HHHU ............
::::::::::::l ........... 1 ............ ············ ··········· ............
u::::::::::::
············ ............
Kag zvp .:.:Pzvm :· ::::::::::::'
•.
........... ........... ........... ...........
;;;;;;;;;! .......... . ·········
LEGEND
Os Quaternary sediments
Kag Cretaceous argillite
Pzvp PaleoZOIC phyllitic greenstone,
ongmally basalt
Pzvm PaleoZOI C massive greenstone,
ongmally basalt
Pzl Paleozoic limestone with
argillite inclusions
ss 5 Sheared rock
~~ Fault wtth gouge, dashed where
inferred, arrows show sense of
' vert1cal displacement
" Lithologic contact, dashed
' wh ere mferred ' 0 Sense of vertical displacem ent
from regional geologic relation·
ships, Csejtey (19811.
Pzvp
NOTES
{0
Stream_l
..
Pz I V. ~ ~:: : :~.2;~1!1. . .
It
A .. L
~····· u·····
............... ............. ..
............... ············· .. ............... ...............
............... ············· .. ............... ............. .. ............. .. ············· .. ............. .. ············· .. . ............. .
1. Section location is shown m
Figures 4 -4 and 4-7.
2. Geology is from Woodward-Clyde
Consultants 1981 field study.
-2650
c:
0 .... co > Q.)
w
DIAGRAMMATIC CROSS -SECTION OF
THE TALKEETNA THRUST FAULT AT
TRENCH T-2 (LOCATION W10)
~0~~~5~~5~0=:5~~~1~0~0 Feet
0 10 20 30 Meters
FIGURE 4-9
LEGEND
~ ....
Strike and dip of bedding
Plunging syncline, dashed
where inferred
Plunging anticline, dashed
where inferred
Thrust fault, dotted where
concealed, sawteeth on
upper plate
NOTES
1. Orientations were measured in Tertiary strata.
Fold axes are inferred from the bedding orientations.
2. Geology is from Woodward-Clyde Consultants 1981
field studies.
3. Geologic relations shown in this figure are
discussed in Section 4.4.1.
ORIENTATION OF FOLDED STRATA
IN WATANA CREEK (LOCATION W3)
0 2 Miles
E=EA~~~,3
0 2 Kilometers
-~0 ~~--------------------------------------------------------------------------~ WARD -CLYDE CONSULTANTS 41410A February 1982 FIGURE 4-10
LEGEND
[QU
[Tgd]
lliJ
lliYJ
.---
..........
\
Quaternary glacial deposits
Tertiary granodiorite
Tertiary volcanic rocks
Paleozoic volcanic rocks
Lithologic contact, approximately
located
Linear scarp, line at top of slope,
hachures on slope
Concealed thrust fault, sawteeth
on upper plate
Trench located between arrows
NOTES
1. Geology is from Csejt ey and others (1978).
2 . Geologic re l ations shown in this
t;guco "' d;.ored ;n Seot;on 4.4 .1.
-N-
~
LOCATION OF TRENCH T-1
AT LOCATION W7
0 2 Miles
I
0 2 Kilometers
~DW~A~~~----------------------------------------------------------------------~ AD-CLYDE CONSULTANTS 41410A February 1982 FIGURE 4-11
LEGEND DESCRIPTION OF UNITS
Lithologic Contacts, Description, Width
Abrupt, <0.1 ft (<3 em)
Gradational, 0.1 to 0.5 ft (3 to 15 em)
----/ '
I \, ____ .,.,.// Very gradational, >0.5 ft (> 15 em)
Soil Boundarie~ Description, Width
.---....
/ \ I
·"·-·"/
··---...
\.
Other Symbols
~or
s
...
NOTES
1. Trench data :
Length :
Abrupt, <0.1 ft (<3 em)
Gradational, 0 .1 to 0.5 ft (3 to 15 em)
Very gradational, >0.5 f t (> 15 em)
Laterally gradational or
i nterfi ngeri ng contact
Water table at time of excavation
Boulder
Gravel layer
Sand layer
Original surface disturbed by
excavation
Fracture due to shear
Sample location
T -1
205 ft (62.5 em)
T-2
78 ft (24 m)
Maximum
Depth: 10.5 ft (3 .2 m) 6 ft ( 1.8 m)
Excavated: 4 to 13 August 1981 26 August 1981
TRENCH T-1
1. Or@nic soil horizon. Fibrous, peaty mat of muskeg with silt
and pebbles, alive at top, very dark brown ( 1 OY R 2/2, moist).
A. Buried organic soil horizon.
2. Mixed loess and co lluvium . Si lt: Pebbly and sandy, poorly
to well sorted, subrounded, por ous, yellowish brown (10YR
5 /4, moist), mottled w i th very dusky red (2.5YR 2.5/2,
moist), massive except fer fa i nt, very pale brown (10YR
7/4, moist) and dark brown (7.5YR 3.5/3, mo ist) hori-
zontal bands that may r epresen t bedding.
3. Loess. Silt: Local patches of 5% very fine sand, well
sorted, porous, yellowish brown (10YR 5/4, moist) w i th
f aint horiz o nta l strong brown ( 1 OY R 5/6, mo ist) bands,
massive, moist.
4. Volcanic ash. Grayish brown ( 10 YR 6 /2, moist) to light
brownish gray (10YR 5/2).
5. Glaciof luvial gravel. 35% pebbles and cobbles, sandy matrix
except locally clast supported, some boulders up to 2 ft
(61 em) in diameter, poorly sorted, subangular to sub-
rounded, very dark brown ( 1 OY R 2/2, moist) with patchy
stains of very dusky red (2.5YR 2.5/2, moist) or dark
reddish brown (5YR 2.5/2, moist), poorly bedded, clasts
of greenstone with less than 1% granitoid rocks.
A. Less than 5% cobbles, pebbles weathered to clayey
silt, loose, clast supported.
B. Pebbles with no cobbles or boulders, very loose, clast
supported.
C. Sandy pebble gravel, <5% cobbles and boulders near
base, locally bedded.
6. Lodgement till. Clay: Sandy and gravelly with boulders up
to 3 ft (91 em) in diameter, very poorly sorted, subrounded
to subangular, dark grayish brown (2,5Y 4/2, moist), massive,
very hard greenstone clasts.
Geologists: Phillip C. Birkhahn Phillip C. Birkhahn
Edward H. Sabins Robert G. Goodwin
Kerry E. Sieh Edward H. Sabins
Logging
Scale: 1 in. = 5 ft 1 in. = 5 ft
2. Description of units is according to Compton (1962), and Pettijohn (1949).
3. Trench T -1 and T -2 locations are shown in Figures 4-11 and 4-7, respectively.
HEET 1 OF 2)
TRENCH T -2
1. Organic soil ho r izo n. Fibro us, pe aty m at of m uske g with
silt and pebbles, ali ve at top, very da rk brow n (10Y R 2/2,
moist).
2. Colluvium . Si lt w ith sand, clay, pebbles, an d cobbles,
poorl y so rted, an gular, p oro us, olive bro w n (2.5Y 4/4,
moist) to ol ive (5Y 4/4, mo ist), m assive, argil lite and
greenstone clasts, gradational contact over weathered bedrock
and abrupt contact over fresh bedrock.
3. L o ess. ·ilt w ith very fin e sa nd, light yellowish brow n
(10YR 6 /4, moist), locally weathe r ed to da rk brow n (7.5Y R
3 /4, moist), massive, pa r tly mi x ed w ith colluvium, 1-in.-
(2 .5-cm-) thick, discontinuous paleosol at base, very dark
brow n (10YR 2/2, mo i st).
4. Weathered bed r oc k . Sand, silt, and clay, decomposed ar-
gill ite and gree nsto ne.
5. A r gil lite (K ag). Medium light gra y (N6, m o ist), weathe r ed
light brow n ish gr ay (1 0 Y R 6/2, m o ist) t o b r o wn i sh ye ll ow
(1 0 YR 6/6, m o ist), thi n bedded, bedding p l ane cle avage.
A. Hard.
B. Hard, m oderatel y sheared.
C. I nten sely shea r ed.
D. C l osely she ar ed, gr ayish black (N2, moist).
E. C lose l y sheared, red ( 10R 4/6, m o ist).
F. G o uge of crushed argi llite and argi llite f la k es.
G. Closely shea r ed .
6. G ree n sto ne (Pzv). Basalt metamo r ph osed to pumpe ll yite-
p r ehnite gr ade of metamo r ph ism.
A. D ar k gra y (N3, m o ist) t o dark gr eenish gr ay (5GY 4/1,
moist), suboph it ic an d fin ely granul ar, traces of pyrite
and q u ar tz.
B. Phyllitic with interbeds of Unit 6A, dusky yellow green
(5GY 5/2, moist) and grayish olive green (5GY 4 /2, moist).
7. Fault gou ge. C l ay, r ed (2.5YR 5/8, moist).
Relative Horizontal Location --
It 10 15 20 25 30 35 40 45 50
10 12 14
10
Trench Bottom
15
Relative Horizontal Location --
TRENCH T -1,
TRENCH LOG
55
16
SOUTHWEST WALL
N22"W--
18
so 65 70
20
75 00 85 90 95
22 24 26 28
It 11Lo~,-----1~15~---r--~120L_ ______ ,12L5 ________ 1J3_o-, ______ 1 3L5----,---1~4o ______ -r1J4_5 ________ 150L,-------1~55----r---~150L_ ____ ~_16L5 ________ 1~70,-------1_L75 __ ,-__ L_100L_ ____ -._18L5 ________ 1J90,_------~19L5 __ ,-____ 2LOO-----,~'~
58 60 62 34 J6 38 40 42 44 46 48 50 52 54 56
It
10 Trench beyond 205 ft (62.5 m) was used for drainage
sump and not logged
15
~
.3 ..
;;;
20
25
TRENCH T-2,
TRENCH LOG
Relative Horizootal Location --
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
10 12 14 16 18 20 22 24
SOUTHWEST WA LL It
N52°W-
Ground Surface
·-··-··----··-··-·-··-··-··-··-··-
Trench Bottom
10
10
DE CONSULTANTS 41410A February 1982
100
30
10
15
~
.3 ..
;;;
20
25
105 110
32
m
10
15
TRENCHES T --1 AND T --2,
TRENCH LOGS
FIGURE 4-12 (SHEET 2 OF 2)
A. Low sun-angle aerial photograph of the Trench T-1 location.
B. Southeast view of Trench T-1 at the scarp. Units are described in Figure 4-12.
PHOTOGRAPHS OF TRENCH T-i
FIGU RE 4-13
Southwest wall of Trench T-2 between 63 and 67 feet (20 .7 and 22.0ml. Units are described in Figure 4-12.
PHOTOGRAPH OF TRENCH T-2
0 6 12 Inches
g¢¢1 I
0 15 30 Centimeters
FIGU RE 4-14
AD-CLYDE CONSULTANTS 414 1 OA February 1982
LEGEND
Turner and Smith (1974)
---
.--.. ·--
Quaternary surficial deposits
Tertiary quartz monzonite
Tertiary or c t . rocks re aceous mtgmatitic intrusive
Tertiary or Cret quartz diorite aceous granodiorite and
Cretaceous paragneiss
Cretaceous pelitic schist
Lithologic contact
Lineament
Woodward-Clyde Consultants [GJ Quartz monzo .t1981d Field Studi es
nt e an equtvalent gneiss
w~~\§~\\1 Biotite schist and gneiss
-Pelitic schist
Location and n · b urn er of magnetic profile
NOTES
1. Geologic relations shown in h. . . tn Section 4.4.1. t ts ftgu re are dtscussed
2. Magnetic profiles are shown in Figure 4-16.
GEOLOGIC MAP OF
LOCATION W11 NEAR BUTTE LAKE
0 E=lES~3:=i~~~r=31 Miles
0 2 Kilometers
FIGURE 4-15
sw NE sw NE
6900
6500
6100
Tkm Kgn Ksp Tqm
A. MAGNETIC PROFILE 1 B. MAGNETIC PROFILE 1A
NW SE
7000
SUSITNA FEATURE
6600
6200
Kgn TKm Tqm
C. MAGNETIC PROFILE 2A
Station reading
Tertiary quartz monzonite
Tertiary or Cretaceous
migmatitic intrusive rocks
Cretaceous paragneiss
Cretaceous pelitic schist
Susitna feature location and bedrock geology is
from Turner and Smith (1974).
Magnetic profile 1 is a calibration line across
rock units similar to those in magnetic pro-
file 2A across the Susitna feature.
Magnetic profile locations are shown in Figure 4-14.
Magnetic data acquisition methodology is described
in Section A.5.
LYDE CONSULTANTS 41410A February 1982
BUTTE LAKE
MAGNETIC PROFILES
0 2000 4000 Feet EFf±i4~o~~~s3b~o~~~1 o~'a~~ Meters
FIGURE 4-16
LEGEND
[QQ
[Tgd I
~
~Krli]
~
[§] m ---
TT''"''rrn
Qu aternary glacial sediments
Tertiary granodiorite
Tertiary felsic intrusive rock
Tertiary and Cretaceous
migmatitic instrusive rocks
Cretaceous paragneiss
Cretaceous pelitic schist
Jurassic flysch
Inferred lithologic contact
Linear scarp, line at top of slope,
hachures on slope
-• • ·-Lineament
\
\
Trench located between arrows
NOTES
1. Geology is from Turner and Smith { 1974).
2. Geo logic relations shown in this
figure are discussed in Section 4.4.1.
LOCATION OF TRENCH S-1
AT LOCATION W12
0 2 Miles
lb3 I
I
0 2 Kilometers
~DWuA~~----------------------------------------------------------------------~ RD·CLYDE CONSULTANTS 41410A Febru ary 1982 FIGURE 4-17 ----
LEGEND
Lithologic Con t acts, Descrip ti on, Width
Abrupt, <0 .1 f t (<3 em) ---...... r \
\. / Gradational, 0 .1 to 0.5 ft (3 to 15 em) -----.,.----.... ,
f I ' / -_,
Very gradational, >0.5 ft (> 15 em)
Soil Boundaries, Description, Width
/--.._ \
\
/ ............
I
-----
/
\. ___ . /
Other Symbols
oo••~~~oo•G••••••••••
. . . . . . . . . . .
5
NOTES
1. Trench Dzta:
Length:
Abrupt, <0.1 ft (<3 em)
Gradat i ona l, 0.1 to 0.5 ft (3 to 15 em)
Very gradational, >0.5 ft (> 15 em)
Laterally gradational or
interfinger i ng contact
Water table at time of excavation
Boulder
Gravel layer
Sand layer
Original surface disturbed by
excavation
Fracture due to shear
Sample location
S-1
228 ft (69 .5 m)
Maximum
Depth: 9 ft (2.7 m)
Excavated: 23 to 24 August 1981
Geolog ists: Phillip C. Birk hahn
Robert G. Goodwin
Kerry E. Sieh
Logging
Scale: in. = 5 ft
DESCRIPTI ON OF U N ITS
TRE NC H S-1
1. Organic soi l horizon. Fibrous, peaty mat of muskeg with
silt and pebbles, alive at top, very da rk b r o wn (1 0 YR 2/2,
moist), massive.
6 . A blat io n till. Sand and gr ave l wi th boul d ers .
2. Colluvium. Silt, sand , pebbles, and bou l de r s, p oorl y sorted,
angular, strong brown (7.5YR 4 /6, moist) wi t h strea k s of yel -
lowish brown ( 1 OY R 3 /4, moist), massive.
3. Clay. Silty with some pebbles, sticky and p la stic, sti ff, olive
gray (5Y 4/2, moist), massive, weak horizontal platey partings,
wet.
4. Sandy loam to sandy c lay. 5% angular pebbles with clay
coatings, slightly sticky and sl ightly plastic, crumbles, very
dark brown ( 1 OY R 2/2, moist) with patches of brown ( 1 OY R 4 /3,
4/3, moist) silt, massive.
5. Lodgement till. Olive gray (5Y 4/2, moist).
A _. Sand: Trace of clay, massive, weak horizontal platey
partings.
B. Clay: Sandy and silty, sticky and plastic, olive gray (5Y
4/2, moist), massive, strong horizontal platy f abr ic.
C. Clay: Silty, sticky and slightly plastic to plastic, dark
olive gray (5Y 3/2, moist), massive.
A. San d : f i n e t o medi um w ith <5 % peb b les an d c obb les ,
m o d erately so rted, subangu l ar to subroun d ed, m ass ive to
poorl y bedde d; w ith grave l l enses : 50% pe bbl es an d
cobbles, poo rl y sorted, subangu l ar, 4° to 6° i n-to-slope
d i p .
B. G r ave l : Coarse sand and pebbles with cobbles and boulders,
mode r at ely sorted, subangu lar, dark grayish brown ( 1 OYR
3.5 /2, moist) to very dark grayish brown (2.5Y 3.5 /2, moist)
weak f abric parallel to slope, clast supported.
C. Silt: Sandy and pebbly, poorly sorted, subrounded, very
dark grayish brown (2 .5Y 3 /2, moist), massive, friab l e.
D. Sand : Very fine with < 1 0 % coarse sand and pebbles with
cobbles and boulders, moderat~ly sorted, subrounded , olive
gray (5Y 4 /2, moist), massive, friable, faint platy fabr ic
subparallel to slope.
E. Gravel: Sandy and silty, moderately sorted, angular to
subangular, dark olive gray (5Y 3 /2, moist), moderately
bedded, quartz and lithic clasts, silt and sand interbeds.
F. Sand : Very fine, well sorted, subangular, dark grayish
brown (10YR 3.5/2, moist) to very dark grayish bronw
(2.5Y 3.5/2, moist), massive.
G . Sand: Medium to coarse, gravelly, moderately sorted,
angular to subangular, olive gray (5Y 4 /2, moist), mas-
sive quartz , granitoid, and argillite clasts , weak fabric
subparallel to slope.
2 . Descr i ption of units is accord i ng to Comp t on (1962), and Pe tt ijohn (1949).
FIGURE 4 -18 (SHEET 1 OF 2) WOODWARD-C L YDE CONS ULTAN TS 41410A FebruarY
1982
h
15}5
z
~
'"}:
"' z
;;;
25
20+6
I L,
h
Relative Horizontal locatwn --
Trench to southeast wa::o used lor
dn1nage sump and was not logged
Relative Hanzontal LocJtiDn --
D-CLYDE CONSULTANTS 41410A February 1982
TRENCH S-1,
TRENCH LOG
SOUTHWEST WALL
N39°W
Ground Surface
,f'
i ~
6+20 "'
~
I ,,
I
LL5
Trench 173 It {52 7 ml
caved wa~ not logged
To
.~
I '
2-f
I + 41
~ 15
'I
6 i-20
I ,J
TRENCH S-1,
TRENCH LOG
FIGURE 4-18 (SHEET 2 OF 2)
A. Low sun ~angle aeri al photograph of the Trench S·1 location.
B. Southwes t wall of Trench S-1 between 20.5 and 24.7 feet (6.3 and 7.5 m).
Units are d escr ibed in Figure 4 -18. PHOTOGRAPHS OF TRENCH S-t
------~--~~--D~W;A~R~~~C-L Y __ D_E_C_O_N_S_U_L_T_A_N_T_S __ 4_1_4_1_0_A __ F_eb_r_u-ar-y--19_8_2--------------------------------------~F~I G~U~R~E~4~-~1~9
LEGEND
__.,. ... ---
•
•
NOTES
Lineament
Location and age designation of Quaternary
un its (0 to 1.8 m.y.b.p.)
Location and age designat i on of Mesozoic
and Paleozoic units (65 to 570 m .y.b.p .)
Area ch eck ed for features related to
the Watana lineament.
1. m.y.b.p. is the abbreviation for million years before
present.
2 . Geologic relations at locations W17 through W25
are discussed in Section 4.4 .1.
-N-
WATANA LINEAMENT (KD3-7)
LOCA Tl ON MAP
0 5 10 Miles ~F;d~~t=f--~Fd~5Fj~-------~--~
0 5 10 Kilometers
FIGURE 4-20
LEGEND
~
G
I Tbgd I
I Tsmg I
~
--·
W26.
W27.
NOTES
Quaternary glacial sediments
Tertiary volcanic rocks
Tertiary granodiorite
Tertiary schist, migmatite, granite
Cretaceous argillite and graywacke
Inferred lithologic contact
Shear zone or fault, dashed where inferred,
dotted where concealed
Location and age designation of
Quaternary units (0 to 1.8 m.y .b .p.)
Locati on and age designation of Mesozoic
and Paleozoic units (65 to 570 m.y.b.p.)
1. m.y.b.p. is the abbreviation for million years before
present.
2 . Geologic relations at these locations are discussed in
Section 4.4.1
3 . The location of the Fins Feature is from U.S. Army
Corps of Eng i neers, Alaska District (undated).
4 . Geology is from Csejtey and others ( 1978).
0
~
-N-
~
FINS FEATURE (KD4-27)
LOCATION MAP
2 M il es
I
0 2 Kilometers
L EGE N D
0
0
~
~
EJ
----
KD~ ... -
KC~
~-··~·
3 ~
D I
NOTES
Tert iary volcanic rocks
Tertiary intrusive rocks
Cretaceous argillite and graywacke
Triassic volcanic rocks
Paleozoic volcanic rocks
Inferred I ithologic contact
Lineament and Code No.
Fault and Code No.
Concealed thrust fau l t ,
sawteeth on upper plate
3 is the segment number d iscussed in
Section 4.4.2
D-3 is the location number discussed in
Section 4.4.2.
Devil Canyon Site
1 . Geology is modified afte r Csejtey and others ( 1978).
2. Geologic relations shown in this figure are discussed
in Section 4.4.2 .
3. The Talkeetna thrust fault i s concealed by
Quaternary sediments.
0
-N-
DEVIL CANYON FEATURES
LOCATION MAP
5
E3
10 Miles
0 5 10 Kilometers
FIGUR 4-22
5 -SEISMICITY AND STRESS REGIME
This section su11111arizes the regional historical seismicity study and
the results of the microearthquake network operation that were conducted
in 1980 (Woodward-Clyde Consultants, 198Gb). The results of the work
conducted in 1981 are also presented. The overall objective of these
studies was to increase the understanding of earthquake potential within
the Benioff zone of the subducted Pacific plate and within the Talkeetna
The results of these studies have were used primarily to
refine our understanding of the location, size, and focal mechanism of
possible maximum credible earthquakes associated with these regions.
The 1981 seismicity study included: an evaluation of the tectonic
associations of moderate to large earthquakes in or adjacent to the
Talkeetna Terrain; a review of the Benioff zone seismicity (to refine
the assessment of the size and location of the largest earthquake
that could be expected to occur on the Benioff zone); a review of small
earthquakes i·n the Talkeetna Terrain to refine the assessment of
the nature of the stress regime; development of recurrence relationships
the crustal zone seismicity and seismic sources within the Benioff
zone; and preparation of a manual for a long-term seismic monitoring
network. The first four topics are discussed below. The manual is
presented as a separate document from this report.
To assess the maximum credible earthquake in the Talkeetna Terrain and
the Benioff zone, historical earthquakes of about magnitude (Ms) 6 or
larger (that occurred within or beneath the Talkeetna Terrain) were
re-examined in terms of location, focal depth, focal mechanism, and
tectonic association. To augment this work, an additional literature
review and an analysis of worldwide historical earthquake data were
conducted to refine the earthquake potential of the Benioff zone. The
stress regime within the Talkeetna Mountains portion of the Talkeetna
5 - 1
Woodward-Clyde Consultants
ain was also studied further by analyzing records from the 1980
twork (Woodward-Clyde Consult ants, 1980b) and from the University
Alaska Geophysical Institute (UAGI).
ITt~ following discussion presents: a) the regional historical seis-
including the analysis of moderate to large historical earth-
b) the results of the Benioff zone seismicity review; c) a
unrmary of the results of the microearthquake network operation in 1980;
\ a discussion of the regional stress regime in the Talkeetna Terrain,
!'
nr:luding the analysis of focal mechanisms for selected events; and
:1' recurrence relationships developed for the crustal zone seismicity
d the Benioff zone.
-Regional Seismicity
5.1.1 -Tectonic Setting
Recent concepts of plate tectonics have had a major influence on
interpretations of the contemporary tectonics of Alaska. According
to these concepts, the underlying cause of the geologic and seismic
activity in central and southern Alaska is the subduction of
the Pacific plate at the Aleutian Trench as the plate spreads
northward from the east Pacific Rise (!sacks and others, 1968;
Tobin and Sykes, 1968). This northward movement occurs at a rate
of approximately 2 .4 inch es/yr (6 cm/yr) relative to the North
American plate and is illustrated in Figure 4-1. As the Pacific
plate reaches the Aleutian Trench, it is thrust under the portion
of the North American plate that includes Alaska and the Aleutian
Islands.
In the Gulf of Alaska area, the interplate movement is expressed as
three styles of deformation: right-lateral slip along the Queen
5 - 2
Woodward-Clyde Consultants
Charlotte and Fairweather faults; underthrusting of the oceanic
Pacific plate beneath the continental block of Alaska; and a
complex transition zone of oblique thrust faulting near the eastern
end of the Aleutian Trench (Figure 4-1). The Trench represents the
ground surface expression of the initial bending of the oceanic
plate as it moves downward beneath the North American plate.
The regional earthquake activity is closely related to the plate
tectonics of Alaska. Figure 4-1 shows an oblique schematic view
of the major geologic and tectonic features of the regional plate
tectonics. The subducting plate is shown moving to the northwest
away from the Aleutian Trench and dipping gently underneath the
upper Susitna River region. The subducted material is located at
depth on the basis of the hypocenter distribution of instrumentally
located earthquake activity. This kind of subcrustal seismic zone
is called a Benioff zone. In some areas, such as to the southwest
of the site region a long the Alaska Peninsula, the presence of
subducted oceanic crust is inferred from the presence at the ground
surface of andesitic volcanic rocks.
The Benioff zone in the site region is characterized by earthquake
activity extending to a depth of about 93 miles (150 km) (Agnew,
1980). No autochthonous andesitic volcanic rocks or volcanoes
currently are known to be present at the ground surface in the site
region above the Benioff zone.
Beneath the Prince Willi am Sound area, which is on the North
American plate, the subducted plate moves nearly horizontally. The
two plates appear to be closely coupled in this region and have the
capacity to accumulate and release very large amounts of elastic
strain energy. The most recent example of this process was the
28 March 1964 earthquake of magnitude (Ms) 8.4 (Mw 9 .2). Part
of the rupture zone of this earthquake near the sites, as evidenced
by aftershocks, is shown in Figures 4-1 and 5-7.
5 - 3
The overlying North American plate is also disrupted by intraplate
deformation that is related to the interplate motion. Evidence
for tectonic deformation is found in the Alaska Range more than
279 miles ( 450 km) northwest of the surf ace i nterp late boundary
at the Aleutian Trench in the Gulf of Alaska. Much of this
deformation is the composite expression of the plate interaction
process over millions of years and of the seaward migration of the
subducting zone, which has periodically accreted additional crust
to the continental land mass.
The historical seismicity within 200 miles (322 km) of the Project
is associated with three general source areas: the crustal seismic
zone within the North American plate; the intraplate region of the
Benioff zone; and the interplate region of the Benioff zone. The
seismicity of these three source areas is reviewed in this section
following a discussion of the record of historical seismicity in
the Talkeetna Terrain and adjacent areas.
5.1.2-Historical Seismicity Record
Prior to the installation in 1935 of a seismograph near Fairbanks
at College, Alaska (CMO), only local felt reports or seismograph
recordings made at distant stations were available to determine
epicenters and focal depths of earthquakes in south central Alaska.
Among these distant stations vJere: one at Sitka, Alaska, installed
in April 1904, consisting of two Bosch-Omori horizontal seismo-
meters; one each at Berkeley and at Lick Observatory in California,
installed in 1887 (published readings began in 1910 and 1911,
respectively); and some Japanese stations developed in 1879. Davis
and Echols (1962), Davis (1964), and Meyers (1976) have published
lists of felt earthquakes for Alaska dating from the 18th century,
although the very low-population density in Alaska prior to
1900 has precluded historical felt reports of earthquakes in the
interior of Alaska earlier than the large event of 1904.
5 - 4
Woodward-Clyde Consultants
Davis (1964) and Meyers (1976) have published lists of felt earth-
quakes for Alaska dating from the 18th century, although the very
low-population density in Alaska prior to 1900 precluded historical
felt reports of earthquakes in the interior of Alaska earlier than
the large event of 1904.
During the early and middle portion of the twentieth century, prior
to 1964, epicenters and focal depths of earthquakes in Alaska were
computed primarily from teleseismic data. Location uncertainty
varied greatly and depended on the specific combination of earth-
quake size and source region depth. For example, larger earth-
quakes (magnitude [Ms] greater than 6) that occurred within the
shallow Benioff zone may have been well-recorded worldwide, but may
not have had clear pP phases to constrain depth estimates and may
have been located using travel time curves that did not account for
local tectonic structure. Uncertainties in location and depth
could be as large as 62 miles (100 km) or more. For earthquakes of
uncertain focal depth, the focal depth is often constrained to a
depth of 20 miles (33 km) to compute the epicentral location. In
addition, recomputations of some earlier earthquake locations, such
as those published by Sykes (1971), have reduced some of the
original catalog errors.
The accuracy of epicentral locations improved slightly with the
installation of the seismograph at College, Alaska (CiviO), (near
Fairbanks) in 1935, but it was not until the mid 1960s, after the
devastating Prince Willi am Sound earthquake of 28 March 1964, that
earthquake monitoring was significantly improved in central and
southern Alaska. After the 1964 earthquake, epicentra1 and focal
depth accuracy improved with the installation of seismic networks
operated by the University of Alaska Geophysical Institute (UAGI),
the National Oceanographic and Atmospheric Administration (NOAA),
5 - 5
Woodward· Clyde Consultants
and the U.S. Geological Survey during the period 1964 to 1967, and
with the preparation of a velocity model for the area by Biswas and
Battacharya (1974).
Since 1974, the focal depths of earthquakes recorded and located
by the UAGI are accurate to approximately plus or minus 9 miles
(15 km) with the ep i central accuracy generally being better than
the depth accuracy. Location accuracy and magnitude detection
levels have varied due to the number of stations in operation at a
given time and changes in data handling procedures and priorities,
so the above values may be too small for some poorly recorded
events. From 1967 to 1974, the focal depth error estimates are
approximately plus or minus 12 to 19 miles (20 to 30 km), with
epicentral uncertainty of plus or minus 12 to 16 miles (20 to
25 km). The accuracy of focal depth estimation within the U.S.
Geological Survey seismograph network is very good, probably plus
or minus 6 miles (10 km) or less. However, this network is south
of the Project and generally outside of the site region.
5.1.3-Analysis of Large Historical Earthquakes
Within the tectonic setting of the site region (Section 5.1.1)
the Benioff zone has a substantially higher level of seismicity
than the crust, which includes the Talkeetna Terrain (as shown
in Figures 5-1 and 5-2 and discussed in Section 4.3 of the Interim
Report [Woodward-Clyde Consultants, 1980b]). This observation
implies that most large historical earthquakes of magnitude
(Ms) 6 or larger have occurred in the Benioff zone rather than in
the crust. To obtain a better understanding of this concept and to
ascertain its validity, we reviewed available records for these
events using the computational seismology techniques of Helmberger
(1968) and others cited below. The earthquakes that were studied
5 - 6
Woodward· Clyde ConsuH:annts
are listed in Table 5-l and shown in Figures 5-1 and 5-2, which are
revised from Figure 4-6 of the Interim Report (Woodward-Clyde
Consultants, 198Gb).
In carrying out this study, seismograms of the selected events were
acquired from data sources in the United States and Europe. The
seismograms were used to evaluate focal parameters such as location
and depth of the earthquake, orientation of the fault plane, and
sense of displacement. The focal parameters were then determined
or estimated by using waveform analysis based on synthetic seismo-
grams (e.g., Helmberger, 1968; Langston and Helmberger, 1975; Rial,
1978) and first-motion polarities of body waves. For events prior
to 1960, only two or three seismograms of the events were usually
available, so some of the determinations are necessarily more
uncertain and a greater 1 eve 1 of judgment was used. In order to
calibrate the judgments of waveform interpretation, two more recent
earthquakes were analyzed, the 1964 and 1975 earthquakes listed in
Table 5-l.
1 January 1975
The focal depth of this mb 5.9 earthquake, as listed in Appendix C
of the Interim Report (Woodward-Clyde Consultants, 198Gb), is
41 miles (66 km), suggesting its association with the subducting
plate. Examination of the available World-Wide Standardized
Seismographic Network (WWSSN) records shm.;s clear depth phases
(pP) that allowed the focal depth estimate to be revised to
34 miles (55 km). The first-motion directions of P-waves read
from nine of these records do not provide much constraint on the
focal mechanism, but examination of the waveforms suggest the
occurrence of normal faulting. The first-motion observations and
the assumed normal fault mechanism are shown in Figure 5-3A.
This mechanism and focal depth are consistent with earthquake
sources occurring within the subducted plate.
5 -7
Woodward-Clyde Consultants
29 June 1964
The first-motion plot for this magnitude (mb) 5.6 event is
shown in Figure 5-38. Data from stations ApatHy, U.S.S.R.
(APA), Nurmijarvi, Finland (NUR), and Kongsberg, Norway (KON),
were taken from the International Seismological Centre (ISC)
bulletin; other readings were taken from the WWSSN seismograms.
Examination of the available WWSSN records shows waveform
characteristics typical of a shallow crustal event (depth=
9 to 12 miles [15 to 20 km]). This earthquake was clearly not
associated with the Benioff zone which 1 i es at a depth of about
80 miles (125 km) beneath the epicenter. It is concluded that
this event occurred in the crust and is assumed to have a pure
thrust solution (as shown in Figure 5-38); however, a substantial
amount of oblique slip is possible. The strike of the nodal
planes is N55°E with dips of 10°NW and 80°SE.
19 August 1948
This magnitude (Ms) 6-1/4 earthquake was examined to verify
its focal depth, which was estimated as 62 miles (100 km) in
Appendix C of the Interim Report (Woodward-Clyde Consultants,
1980b). The records at Pasadena, California (PAS), and several
other southern California stations show clear depth phases
indicating a focal depth of 55 miles (90 km). Thus, this event
is associated with the Benioff zone. The wave form ch aracteri s-
ties suggest a thrust mechanism.
3 November 1943
This earthquake of magnitude Ms 7.3 is the largest his tori cal
earthquake that has been located within the Talkeetna Terrain.
It has, therefore, been studied in greater detail during this
study than the other events.
5 - 8
Good seismograms showing clear body and surface waves were
obtained from Kew, England (KEW), Ottawa, Canada (OTT), College,
Alaska (C~10), Honolulu, HavJaii (HON), Puerto Rico (SJG), Pasa-
dena, California (PAS), and additional stations in the southern
California network. Synthetic seismograms were constructed to
match the observed records, as shown for the long-period record
at PAS in Figure 5-4. The shape and complexity of the P-wave
form indicates the presence of tvJO sources separated in time by
1.7 seconds and occurring at nearly the same location. A focal
depth of 11 miles (17 km) gives the best fit of the synthetic
record to the observed records for the phases pP and sP (with an
accuracy of +1 mile [2 km]). The source model for both sources
is a high-angle, nearly pure reverse mechanism (dip and rake of
80° to 90°) with strike of N45°E. The second source is slightly
larger, with relative seismic moment of 5.5 x 1Q26 dyne-em
compared to 5.0 x 1Q26 dyne-em for the first source. The
source time-functions for both sources indicate a relatively
high stress drop; this interpretation is consistent with the
interpretation that this double event is an intraplate earth-
quake (Kanamori and Anderson, 1975).
The observed polarities of the body waves are consistent with the
theoretical radiation pattern of the NE-striking reverse-slip
model. Figure 5-5 shows the theoretical radiation patterns for
P, SV, and SH for epicentral distance greater than 30°. Stations
and polarities are plotted at their azimuths from the epicenter.
P-wave polarities at OTT, KB~. PAS, and CMO allow for up to 15"'
uncertainty in the fault orientation and dip.
With the focal depth accurately estimated at 11 miles (17 km).
the e p i cent e r of the earth q u ak e (as l i s ted i n Append i x C of
1-Joodward-Clyde Consu1tants, 1980b) can also be refined. To
analyze the epicenter, the residuals at stations reported in the
5 - 9
Woodward·Cmyde
ISC Bulletin were minimized using a trial and error approach that
took into account the 11 mile (17 km) focal depth. The revised
location of 61.9oN latitude and 151.3°W longitude is 19 miles
(30 km) northwest of the NOAA location. The uncertainty of the
revised location is 6 to 12 miles (10 to 20 km). This location
is approximately 90 miles (145 km) southwest of the Project
(Figure 5-1) and lies along the transitional western boundary of
the Talkeetna Terrain.
Geological Association
The seismological analysis of the 1943 earthquake places it in
the crust along the western boundary of the Talkeetna Terrain
(Figure 5-6). For an earthquake of this size and focal depth, we
would expect it to have occurred on an identifiable source (e.g.,
a fault with recent displacement) as discussed in Section 4.2.
No faults with recent displacement had been reported in this
region, so a limited literature review and remotely sensed data
interpretation program was conducted during this study. The
literature review involved the review of geophysical studies and
field mapping, primarily by Griscom (1979), Csejtey and Griscom
(1978), and Csejtey and others (1978) and evaluation of the
tectonic model (including the Talkeetna Terrain); the review led
to the following observations:
a) The earthquake occurred along the western boundary of
the Talkeetna Terrain. As discussed in Section 4.1, this
boundary is a broad zone of deformation and volcanoes.
b) A prominent northeast-southwest trending aeromagnetic
anomaly (separating aeromagnetically dissimilar terrain) was
identified within 15 miles (25 km) north of the epicenter by
5 -10
Griscom (1979), as shown in Figure 5-6. This anomaly
corresponds in part (such as near Mount Yenlo [Figure 5-6])
to what Griscom (1979) refers to as an unconformable contact
between the Mesozoic metasedimentary sequence of the
Maclaren Terrane to the northwest and the Paleozoic and
Mesozoic volcanic sequence of the Wrangellia Terrane to the
southeast (these terranes are discussion in Section 4.4.1).
In the project site region, this contact was mapped as the
Talkeetna thrust fault (Csejtey and others, 1978).
c) The Talkeetna thrust fault is offset near the town of
Talkeetna by a series of strike-slip faults, as shown in
Figure 5-6. This offset presents convincing evidence that
this segment of the Talkeetna thrust fault has not been
subject to recent displacement (i.e., if the Talkeetna
thrust fault had had youthful displacement, it would have
displaced these strike-slip faults rather than having
been displaced by them). It also shows that the contact
described in (b) above is not part of the same fault
system.
d) The Mount Yenlo contact dips to the northwest as shown in
Figure 5-6. This dip is in the opposite direction from that
of the Talkeetna thrust fault described in Section 4.4.1 and
shown in Figure 5-6.
e) The southwestern end of the Talkeetna thrust fault is
capped by Tertiary volcanic units that have no observed
displacement.
5 -11
Woodwa~rd-Ciyde ConsuHant:s
f) The aeromagnetic anomaly described in (b) above does not
correlate with the mapped trace of the Talkeetna thrust
fault southwest of Stephan Lake. A number of causes could
have produced this deviation, for example, the presence of
intrusive bodies with a high percentage of minerals with
magnetic properties.
In addition to the above geophysical and geologic data review,
interpretations were made of LANDSAT imagery for a 705-square-
mi le (1,960-km2) area encompassing the epicentral region.
Available U-2 photography at a scale of 1:60,000, which covered
approximatelty 60% of the region, was also interpreted, and a
winter overflight was made of the epicenter region. The purpose
of the overflight made on 16 December 1981 was to review the
lineaments identified on the remotely sensed data and to assess
if any clearly identifiable faults with recent displacement
were present in the epicentral region. Lighting and weather
conditions during the overflight were poor, the cloud ceiling was
at 2,500 feet to 3,000 feet (762 m to 915 m), and light was
diffuse, with approximately three hours of usable daylight.
The results of the remotely sensed data analysis and overflight
led to the following observations:
a) A number of lineaments are possible candidate sources for
this earthquake. These lineaments are shown in Figure 5-6.
b) The observed lineaments predominantly trend northwest-
southeast perpendicular to the northeast-southwest trend
obtained from the focal mechanism studies.
5 -12
c) The observed lineaments are parallel to the range front (and
the Talkeetna Terrain boundary) and the direction of glacial
ice.
d) Many of the lineaments are of glacial origin. Some of the
lineaments may be of tectonic origin.
From the geophysical, geological, and remote sensing observations
described above, it was concluded that:
a) Potential sources for the 1943 earthquake are present in the
epicentral region (e.g., the source of the aeromagnetic
anomaly; or one of the lineaments observed in the epicentral
are a).
b) Among the potential sources for the 1943 earthquake, no
obvious faults with recent displacement were observed.
c) The 1943 earthquake did not occur on a fault that is part of
the Talkeetna thrust fault.
27 April 1933
This magnitude (Ms) 7 earthquake was reported by NOAA (Appendix C
in Woodward-Clyde Consult ants, 1980b) with an assigned focal
depth of 0 miles (0 km). Available records from Tucson, Arizona
(TUC), Berkeley, California (BKS), and Pasadena, California
(PAS), and additional southern California stations show clear
evidence that a multiple event occurred in the crust at an
appro xi mate depth of 9 miles (15 km). Comparisons with waveforms
for earthquakes with thrust mechanisms, presented by Langston and
Helmberger (1975), indicate that waveforms for P and S arrivals
5 -13
are typical of a thrust mechanism. This earthquake occurred
outside of the Talkeetna Terrain. It may have been associated
with the Castle Mountain fault.
4 July 1929
This magnitude (Ms) 6-1/2 earthquake was reported by NOAA
(Appendix C in Woodward-Clyde Consultants, 1980b) to have the
same epicenter as the 21 January 1929 event of Ms 6-1/4. No
focal depth was reported for either event. A record from TUC,
was the only one obtained for this event during this investiga-
tion. A qualitative interpretation of this record suggested that
the earthquake occurred in the crust. Its epicentral location
north of the the Denali fault (Figure 5-1) appears to be justi-
fied, on the basis of work by Sykes (1981); therefore, it
occurred outside of the Talkeetna Terrain.
3 July 1929
This earthquake had a magnitude of (Ms) 6-1/4 and no reported
focal depth, according to NOAA (Appendix C in vJoodward-Clyde
Consultants, 1980b). The only record available for this earth-
quake is TUC. The focal depth appears to have been below the
crust, possibly in the depth range 25 to 31 miles (40 to 50 km).
This estimate is based on a subjective assessment of the wave-
forms and the general aspect of the record. This event appears
to have occurred in the Benioff zone below the Talkeetna Terrain.
Sykes (1981) has relocated the event to 62.3°N and 149.3°W
(Figure 5-2), about 12 miles (20 km) south of the NOAA location.
21 January 1929
The 21 January 1929 earthquake had a reported magnitude (Ms)
of 6-1/4 and was located north of the Denali fault (Appendix C in
5 -14
Woodward-Clyde Consultaurnts
Woodward-Clyde Consult ants, 1980b). No reported focal depth
was given. Review of the TUC record suggested that this event
occurred in the crust, as shown in Figure 5-l, and that its
epicentral location north of the Denali fault appears to be
justified, considering the work of Sykes (1981). Therefore, it
occurred outside the Talkeetna Terrain.
7 July 1912
This earthquake had a reported magnitude of Ms 7.4 and an
original location north of the Denali fault (Appendix C in
Woodward-Clyde Consultants, 198Gb). Review of felt reports
during the 1980 study were used to relocate the event to a
location near the Hines Creek strand of the Denali fault (Figures
4-6 and 4-8 in Woodward-Clyde Consultants, 1980b, and Figure 5-l
in this report). Evaluation of the available Honolulu (HON)
record for this event suggests that it is crustal. The location
along or north of the Denali fault lies outside of the Talkeetna
Terrain.
27 August 1904
This earthquake of magnitude (Ms) 7-3/4 was reported by NOAA
to have a location north of the Denali fault (Appendix C in
Woodward-Clyde Consultants, 1980b). Review of felt reports
during the 1980 study, and data obtai ned from Sykes (1981) have
resulted in the relocation of the earthquake 53 miles (85 km)
northwest of the NOAA location. Review of the available HON
record for this event suggested that it is crustal. The crustal
location appears to lie substantially north of the Talkeetna
Terrain.
5 -15
Woodward-Clyde Consultants
Summary
This additional analysis of historical earthquakes provides
significant confirmation of the low level of crustal deformation
occurring within the Talkeetna Terrain. As shown in Figure 5-l,
the largest earthquakes with reported locations within the
Talkeetna Terrain are the Ms 5.6 earthquakes of 17 July 1923,
29 May 1931, 17 October 1931, and 26 July 1933. These are too
sma 11 to be analyzed by the procedures used above, but they may
in fact have occurred in the Benioff zone, as the 3 July 1929
event apparently did. The 3 November 1943 earthquake occurred in
the crust along the western margin of the Talkeetna Terrain.
-Benioff Zone Seismicity
the Interim Report (Woodward-Clyde Consultants, 198Gb), the Benioff
beneath the Talkeetna Terrain was clearly identified and exhibited
al detailed features of Benioff zones that have been defined in
In order to assess more fully the impact of potential future
.hquakes occurring in the subcrustal region, additional comparative
.· £j.'1,;vses were used to characterize the subduction process in central
These characteristics were used to distinguish the intraplate
tnterplate regions more atcurately than was done for the 1980 report
to refine further the estimates of maximum earthquakes for these
5.2.1-Benioff Zone Zonation
The Benioff zone shown in profile in Figure 5-7 was interpreted in
the Interim Report (Woodward-Clyde Consultants, 198Gb) on the
basis of Yoshii 's "aseismic front 11 model (Yoshii, 1975, 1979) and
the model of Davies and House (1979). The principal element of
5 -16
Woodward· Clyde Consultants
Yoshii 's model is the aseismic front, which is a vertical plane
separating seismic (trenchward) from aseismic (landward) portions
of the mantle above the subducting plate. The front also separates
the trenchward portion of the plate interface, which is subject to
interplate earthquakes, from the landward portion of the plate
interface, which is aseismic but below which intraplate earthquakes
occur within the subducting plate.
The physical interpretation of the aseismic front is that it marks
the lower boundary of coupling between the two plates on the
shallow portion of this interface. This coupling is reflected in
interplate earthquakes on the interface and in the earthquakes in
the mantle of the overriding plate above the plate interface. The
latter earthquakes are believed to be caused by stress transmitted
across the interface from below by the subducting slab. Landward
of the aseismic front, there is no coupling across the plate
interface and thus interplate earthquakes on the interface and
earthquakes above the interface are absent.
The aseismic front shown in Figure 5-7 was located on the basis of
mantle earthquake locations. It intersects the plate interface at
a depth of 29 miles (47 km). The apparent absence of mantle
seismicity in the adjacent region to the southwest (Lahr, 1975)
prevents the identification of the aseismic front using this method
in that locality.
According to Yoshii 's model, it should be possible to locate the
aseismic front by finding the point of transition from interplate
to intraplate earthquake mechanisms. On re-examination of the 1980
microearthquake data combined with the UAGI data, two additional
focal mechanisms in the region to the southeast of the area were
covered by the network. These two mechanisms are plotted in
Figures 5-8A and 5-88, and their locations in cross-section are
shown in Figure 5-7. Event A is the only apparent thrust mechanism
5 -17
Woodward-Clyde Consultants
found in the interplate region of the Benioff zone, and this
mechanism is somewhat uncertain. The horizontal axis of maximum
compression (P) is not well-aligned with the northwest plate motion
vector, and the inferred fault planes are too steep to represent
interplate slip. This event (A) lies southeast of the aseismic
front and the transition zone (discussed shortly) and is located
approximately along the plate interface.
Events within the intraplate region of the Benioff zone (Figure
5-7) generally show down-dip tensional axes, indicating intraplate
mechanisms, and their depths range from 23 to 54 miles (37 to 89
km). This suggests that the aseismic front may lie farther east
than shown in Figure 5-7, because the aseismic front occurs at a
depth of 29 miles (47 km) on the plate interface.
Earthquake mechanisms in an adjacent region southwest of the study
area were studied by Lahr (1975). He found a composite solution
for events in the depth range 26 to 34 miles (42-55 km), with an
average depth of 30 miles (48 km), that indicated interplate
faulting. Intraplate earthquake mechanisms were generally at
depths greater than 26 miles (42 km), although a composite intra-
plate solution was found for a group of earthquakes in the depth
range 18 to 26 miles (29 to 43 km), with an average depth of
23 miles (38 km). These results suggest that the transition occurs
in the depth range 21 to 31 miles (35 to 50 km) and may not
necessarily be sharply defined. Since Yoshii 's study, Seno and
Pongsawat (1981) have found that the interplate and intraplate
regions in a 1 imited segment of the northern Honshu Benioff zone
may overlap over an interval of approximately 31 miles (50 km)
trenchward of the aseismic front,
just below the interplate ones.
location identified by Lahr (1975)
with intraplate events occurring
This area of overlap is in the
as the transition zone.
5 -18
Woodward-Clyde Consultants
There is a distinct change in the level of seismicity in the
Benioff zone at a depth of about 21 miles (35 km). It is possible
that the interval from this depth to the aseismic front represents
overlapping of interplate and intraplate zones as found by Seno and
Pongsawat (1981). This possibility is compatible with the apparent
existence of a transition zone in focal mechanisms as discussed
previously.
Davies and House (1979) distinguish between the "main thrust"
zone (which is termed the "interplate region of the Benioff zone"
in this report) and the 11 Benioff" zone (which is termed the
"intraplate region of the Benioff zone" in this report). Davies
and House (1979) provide three lines of evidence that argue for a
distinction between the "main thrust" (shallower, interplate) zone
and the "Benioff" (deeper, intraplate) zone. These are the change
in dip of the subducting plate, the transition from interplate to
intraplate earthquake mechanisms in the vicinity of this bend, and
the occurrence in the 11 main thrust zone (shallow plate interface)
of great thrust-type earthquakes followed by periods of quiescence
(Mogi, 1969; Kelleher and others, 1974; Kelleher and Savino, 1975)
as compared to the relatively uniform seismicity in the "Benioff
zone 11 (deeper intraplate zone).
Davies and House (1979) find a change in dip at a depth of approx-
imately 24 miles (40 km) in a seismicity profile across the
subduction zone in the vicinity of Skwentna, south of the site
region. Although the seismicity in Figure 5-7 does not extend very
far trenchward, it is clear that a change in dip must occur at
about a depth of 21 miles (35 km) in order that the projected
plate interface intersect the surface at the Alaska trench. This
projected interface has a dip of approximately 8° and has a depth
of 15 miles (25 km) at the northern limit of the aftershock zone of
the 1964 Prince William Sound, Alaska, earthquake (Figure 5-7).
5 -19
Woodward-Clyde Consultants
The distinct change in level of seismicity in the subduction zone
that occurs at a depth of 21 miles (35 km) has already been noted.
This change is interpreted to be the boundary of the "Benioff" and
the 11 main thrust 11 zones of Davies and House (1979), since it
satisfies all three of their criteria. There is a change in dip,
the beginning of a transition in earthquake mechanisms, and a
change in seismicity level.
No synthesis of Yoshii 's (1975; 1979) "aseismic front 11 model and
the 11 Benioff-main thrust" model of Davies and House (1979) has
been published. However, it would appear that these two boundaries
define lower and upper depth limits of a transition zone from
seismic to aseismic behavior of the plate interface.
It is concluded from the 11 Benioff -main thrust" model for Alaska
that great shallow interplate earthquakes do not rupture down-dip
beyond the transition at a depth of 21 miles (35 km). The after-
shock zone of the 1964 Prince William Sound earthquake suggests
that rupture terminated at a shallower depth, approximately
15 miles (25 km) deep, at a distance of 43 miles (70 km) trench-
ward of the transition. It is not clear that this 43-mile (70-km)
interval ever ruptures seismically in great interplate earthquakes.
Instead, it is likely that this interval undergoes plastic deforma-
tion after great i nterp late earthquakes or undergoes steady creep
during the interseismic period of the earthquake cycle (Thatcher
and Rundle, 1979).
5.2.2-Magnitude Estimates of Benioff Zone Earthquakes
The Benioff zone earthquakes fall into two categories: interplate
earthquakes (which represent the relative motion of plates) and
intraplate earthquakes (which represent the internal deformation of
the subducting plate). The maximum magnitudes of interplate
earthquakes \'Jere evaluated from estimated rupture dimensions and
5 -20
from a comparative study of worldwide seismicity. Intraplate
earthquakes have much smaller dimensions than large interplate
earthquakes, but these dimensions are usually not precisely
determined because aftershocks are scarce. For this reason, the
maximum magnitude of intraplate earthquakes was estimated from a
comparative study of worldwide seismicity.
Prior to assessing the maximum magnitude values, a relationship
between magnitude scales and rupture dimension is needed. Kanamori
(1977) has suggested that magnitude scales for large earthquakes
should be based on seismic moment. This is because surface wave
magnitude Ms begins to saturate above Ms = 7.5. Seismic moment
M0 is defined by the relationship:
M0 = l1AD . . . . . . . . . . . . . . . . . . . . Equation 5-1
were l1 =the rigidity,
D = the average displacement on a fault, and
A = the area of the fault surface.
Kanamori (1977) defines a magnitude scale Mw based on seismic
moment which does not saturate at the upper end and is equivalent
to surface-wave magnitude in the range 6.0 and 8.0. Mw can
therefore be considered as a continuation of the Ms scale for
large earthquakes. Relationships between seismic moment and
magnitude have been derived from worldwide earthquake data. For
Mw greater than 7-1/2 and Ms greater than 5 and less than
7-1/2, Hanks and Kanamori (1979) define a moment magnitude
scale M, which is related to seismic moment by the relation:
M = 2/3 log M0 -10.7 ........ Equation 5-2
Direct determination of M0 is made by using long-period body
waves, surface waves, free oscillations, and geodetic data.
5 -21
Indirect estimates of M0 can be made from measured surface
displacements, rupture lengths, and estimated fault width. For
faults that have not been subject to historical rupture, geologic
and seismic studies may provide information on these parameters;
seismic moment can then be calculated, and a moment magnitude can
be derived from the equation given above.
An assumption made in the derivation of the moment magnitude
relationship is constant stress drop for large earthquakes
(Kanamori, 1977). Some error may be introduced into moment
magnitude calculations because of temporal on regional variations
in stress drop. In addition, uncertainties in the estimation of
displacement, rupture length, and fault width may lead to
errors in the estimation of seismic moment. However, we shall
now see that the uncertainties involved in the estimation
of future large earthquake moments or magnitudes may be quite
small if the fault area can be accurately estimated.
Estimating Magnitude of Benioff Zone Earthquakes from Fault Area
Abe (1975) showed that a remarkably linear relationship exists
between fault area and seismic moment of large earthquakes. This
relationship is given by:
M0 = 1.23 x 1022 s3/2 dyne/em . . . . Equation 5-3
where S is the fault area in km2. The measured moment values
for this data set deviate from this equation within a factor of
1.3 for M0 larger than 1o27 dyne/em, i.e .• for Ms larger
than 7 .3.
We may convert from moment M0 to moment magnitude M using the
relation of Hanks and Kanamori (1979):
5 -22
M = 2/3 log M0 -10.7 . Equation 5-4
This moment magnitude is equivalent to Ms in the range 5 < Ms
< 7-1/2. Then Abe's equation becomes:
M = log A+ 4.03 .. . ...... Equation 5-5
This equation is very similar to an equation obtained by Wyss
(1979):
M = log A + 4.15 . . ........ Equation 5-6
Abe's equation (Equation 5-5) is preferred for our purposes
because it is based on a data set for which all earthquakes
larger than 7.5 are subduction zone earthquakes, whereas Wyss'
equation (Equation 5-6) is based on a global set of data.
5.2.3-Maximum Earthquakes from Significant Benioff Zone Sources
There are two significant subduction zone sources to be considered
at the sites:
0
0
Source 1 is an interplate earthquake on the shallow portion of
the plate interface southeast of the sites.
Source 2 is an intraplate earthquake vdthin the subducting
plate beneath the sites.
Maximum Magnitude of Interplate Earthquakes (Source 1)
The source 1 interplate earthquakes can be very large because of
the large dimensions of plate interfaces. The length of the zone
that ruptured in the 1964 Pri nee Wi 11 i am Sound earthquake is
5 -23
Woodward-Clyde Consultants
approximately 525 miles (750 km) (Sykes and Quittmeyer, 1981).
The width that ruptured is variously estimated, for example from
112 miles (180 km) (Sykes and Quittmeyer, 1981) to 223 miles
(360 km) (Davies and House, 1979). The width of the aftershock
zone (Page, 1968) appears to increase from approximately 124
miles (200 km) in the southwest to approximately 217 miles
(350 km) in the northwest. Using the dimensions assumed by Sykes
and Quittmeyer, we obtain a magnitude (moment magnitude) of 9.2,
which agrees with the value measured by Kanamori (Mw = 9.2).
As discussed previously. the p 1 ate interface appears to have
ruptured to a depth of approximately 15 miles (25 km) during the
1964 Prince William Sound earthquake. If we assume that it is
possible for rupture to extend to a depth of 23 miles (35 km), an
additional strip 43 miles (70 km) wide is added at the lower edge
of the 1964 fault plane, giving a magnitude (moment magnitude) of
9.35.
Maximum Magnitude of Intraplate Earthquakes (Source 2)
A search of the NOAA worldwide earthquake catalog in the depth
range 22 to 43 miles (35 to 70 km) was conducted in order to find
the largest historic intraplate earthquakes in the Benioff zone
in that depth interval. All of the earthquakes having catalog
magnitudes greater than (Ms) 7.5 are listed in Table 5-3. Of
these earthquakes, all but six were excluded from the prescribed
c at ego r y for r e as on s g i v en i n t h e 11 c o mm en t 11 c o 1 u m n . T he s e
reasons include more precise magnitude or depth determinations.
the focal mechanism determinations, and the hypocentral locations
that indicate interplate faulting or other dissimilarities in
tectonic setting. Four of the remaining six earthquakes have
unknown mechanisms; thus, it is not known whether they are
5 -24
intraplate events. These earthquakes, for which the magnitudes
(Ms) are all 7.6, are the 1933 Peru, 1935 Kuril, 1943 Kermadec,
and 1953 Chile events.
The two remaining earthquakes are known to be intraplate events.
They are the 1970 Peru earthquake (Ms = 7.6, Mw = 7 .9) and
the 1959 Kamchatka earthquake (Ms = 7.7, Mw = 8.2). These
are the only two earthquakes having down-dip tension axes within
the Benioff zone that are believed to have ruptured a large
fraction of the descending lithosphere at depths shall ower than
100 km (Abe, 1972; Seno, 1981). The estimates of Mw (Abe and
Kanamori, 1980) are based on the dimensions of the aftershock
zones.
I sacks and Barazangi (1977) showed that although the 1970 Peru
mainshock had a down-dip tensional mechanism, its aftershocks
included events having both down-dip tensional and down-dip
compressional mechanisms, which were spatially separated into
lower and upper regions within the subducting plate. By com-
paring this distribution of mechanisms with that found in
Kamchatka by Vieth (1974), they inferred the existence of upper
and lower seismic planes within the descending lithosphere. This
phenomenon is now widely recognized, although Fujita and Kanamori
(1981) do not include Peru in their worldwide list of double
seismic zones.
The presence of a double-planed seismic zone beneath the Shumagin
Islands, Alaska, has recently been reported by Reyners and Coles
(1982). However, no evidence for the lower plane is found
beneath south-central Alaska (Figure 5-7).
Both of the large intraplate earthquakes occurred in double-
planed seismic zones and appear to have ruptured the interval
5 -25
Woodward-Clyde Consu§tants
between the two planes. Beneath the Susitna site, the absence of
a lower plane implies the absence of seismogenic stress in the
lower region of the plate and, thus, the absence of large (Ms
> 7-1/2) intraplate earthquakes. Therefore, it is concluded that
the maximum magnitude intraplate earthquake that can occur in
the intraplate region of the Benioff zone beneath the site is
magnitude (Ms) 7-1/2.
Microearthguake Activity
l}i,.Jri ng the three-month period 28 June to 28 September 1980, Woodward-
Clyde Consultants conducted microearthquake recording and analysis
tt:::1 study sei smi city in the vicinity of the proposed Devil Canyon and
The objective of the study was to collect mi croearth-
q!\J:ake data of value in assessing earthquake sources within approxi-
l'!:::;:jtely 30 miles (48 km) of the sites. The data were used to calculate
E~:a.rthquake locations, focal depths, local Richter magnitudes (ML), and
first-motion plots that could be interpreted with respect to regional
a.nd local geologic features, tectonic models, and historical seismicity.
"ii'hese results have been combined with seismic geology results to assess
the seismic design bases for the Project. These results have also been
to plan a program of long-term seismic monitoring.
'T'J,·:d s section summarizes the monitored activity and presents the results
Df additional focal mechanism studies that were conducted to refine the
1980 focal mechanism analysis.
5.3.1-Network Operation
During the period 25 June to 4 July 1980, ten seismograph stations
were installed around the Watana and Devil Canyon sites at the
locations shown in Figures 9-1 and B-1 of the Interim Report
5 -26
Woodward-Clyde Consultants
(Woodward-Clyde Consultants, 1980b). Data from eight of the
ten stations were telemetered into the Watana Base Camp, where
seismographs continuously recorded data on smoked drum recorders.
Two of the ten stations recorded data at their respective field
sites and required servicing every other day by helicopter. This
station configuration and instrumentation provided a reliable field
operation and produced a high-quality data set. The seismic
records were read at the field camp, and local earthquakes were
located with a portable microcomputer. The field data analyses
provided the latitude, longitude, depth of the focus, and local
Richter magnitude (ML) of each processed earthquake.
After t h e f i e l d season , the earth q u ak e s were repro c e s sed by
Woodward-Clyde Consultants using data analysis procedures described
in Appendix B of the Interim Report; final locations were cataloged
in Appendix D of the Interim Report (Woodward-Clyde Consultants,
1980b).
5.3.2-Recorded Earthquakes
Between 28 June and 28 September 1980, a total of 268 earthquakes
were located within an area bounded by 62.3° to 63.3° north lati-
tude, 147.5° to 150.4° west longitude, designated the microearth-
quake study area. Of these 268 earthquakes, 98 occurred be low a
depth of 19 miles (30 km) in the dipping Benioff zone, and 170
occurred in the crust above 19 miles (30 km), primarily in the
depth range of 5 to 12 miles (8 to 20 km) as shown in Figure 5-7.
These earthquakes are shown in Figures 9-1 and 9-2 of the Interim
Report (Woodward-Clyde Consult ants, 1980b). The accuracy of the
earthquake locations is considered to be very good (within a few
kilometers) for those events that occurred within the network, but
the accuracy of the location of earthquakes outside the network
decreases as the distance from the network increases.
5 -27
Among the crustal events, the largest earthquake was of magnitude
(ML) 2.8 and occurred approximately 7 miles (11 km) northwest of
the Watana site on 2 July 1980. Five smaller events have also been
located within 6 miles (10 km) of the Watana site. A magnitude
(ML) 1.66 earthquake occurred within 3 miles (5 km) of the Devil
Canyon site on 12 September 1980 at 0428 Universal Coordinated Time
(UCT). In addition, six smaller events occurred in the Devil
Canyon area.
In addition to the crustal zone of seismicity, the existence of a
subcrustal zone of seismicity is clearly demonstrated in Figure
5-7. The deeper zone dips in the direction of approximately N45°W
at an angle of 20°. The depth of 19 miles (30 km) separates the
crustal zone from the deeper seismicity.
It is clear by inspection of Figure 5-7 that the Benioff zone is
characterized by having more frequent larger earthquakes than
does the shallow crustal zone. Thirteen Benioff zone earthquakes
were assigned a magnitude (ML) of 3.0 or larger, the largest of
which had a magnitude (ML) of 3.68 and occurred on 13 July 1980.
The contrast in level of seismicity in the crustal and Benioff
zones is consistent with the Benioff zone being about an order of
magnitude more active than the crustal zone.
5.3.3-Talkeetna Terrain Stress Regime
In the Interim Report (Woodward-Clyde Consultants, 1980b), compo-
site focal mechanisms were prepared using data obtained during the
1980 mi croearthquake monitoring study. In order to refine and
extend this analysis, additional data were obtained from the
University of Alaska Geophysical Institute (UAGI) by reading the
original film records for the three-month time period of the
microearthquake study. Eleven earthquakes were large enough to
5 -28
record on both networks during July, August, and September 1980.
These earthquakes include the events listed in Table 5-2.
Of the 11 events, three occurred in a cluster, designated as
Cluster #1 in the southern portion of the network (shown in
Figures 9-6 and 9-7 of the Interim Report [Woodward-Clyde Consul-
tants, 1980b]). Four of the events (including catalog numbers
49, 116, and 197) are applicable to the focal mechanism shown in
Figure 9-8 of Woodward-Clyde Consultants (198Gb). The remaining
four events did not pro vi de we 11-constrai ned focal mechanisms and
have not been considered further.
The additional first motion data obtained from the UAGI stations
did not alter or improve the mechanism shown in Figure 9-7 of
the Interim Report (Woodward-Clyde Consultants, l980b). That
mechanism, using the first motion data and the geologic structural
trends, was judged to be a N23oE plane dipping 50°NW, with maximum
compression oriented northwest-southeast.
The additional four events associated with the mechanism shown in
Figure 9-8 of Woodward-Clyde Consultants (1980b) did not change the
reported composite mechanism, which indicates northwest-southeast
compression.
!;i ,•il -Recurrence
EJn't.hquake magnitude-frequency re lat i onsh ips (recurrence models) were
developed for this investigation to be used as part of the input for the
setsmic exposure analysis discussed in Section 8. The recurrence
~;!.!J;'I''•i~meters a and bin the relation Log N =a-bM for the Talkeetna
Terrain, the intraplate Benioff zone, and the interplate zone are listed
in Table 8-2. The recurrence relations are plotted in Figure 5-9 for a
5 -29
Woodward-Clyde Consultants
·11Tea of 386 square miles (1,000 km2) per 100 years along with the
,r,~ points that are discussed below. The data were taken from the
area of the 1980 microearthquake network, an area of 6,564 square
(17,000 km2), as shown in Figure 5-l.
recurrence relationships shown in Figure 5-9 are intended to
pr·~~~~)ent the seismicity of the three zones shown. However, the
used to develop or constrain these relations are subject to several
of uncertainty. A primary uncertainty is due to the limited
;periods of observation at various magnitude completeness levels .
. · uakes of magnitude (ML) 2 and 1 arger were observed for three
(Woodward-Clyde Cons u 1t ants, 1980b); earthquakes of magnitude
and larger were observed for about 15 years (Agnew, 1980);
earthquakes of magnitude 5-1/2 and larger, for about 55 years
·ard-Clyde Consultants, 1980b). Secondly, the magnitude scales
by the various data sources are not necessarily compatible and
varied over time. Finally, the form of the frequency-magnitude
lationship for the larger events in the zones is not constrained to be
·Ji~'l•f;),:J.r by the few data points, but is only assumed to behave so. Some
,. such as interplate regions, may rupture in "characteristic" large
quakes (the 1964 earthquake, for example) that recur in a periodic
f';:j:~ii"don and are not necessarily accompanied by a suite of smaller
~~;.~~~~"·thquakes that fit a linear relation as shown in Figure 5-9. Addi-
1 OI'H:t.'l aspects of uncertainty and constraints of the i ndi vi du a 1 recur-
rence relations are discussed in the following sections.
5.4.1 The Interplate Region
The slope of the interplate relationship is assumed to be 0.85
!:Woodward-Clyde Consultants, 1978). The position of the curve
(a-value) is determined so as to allow the recurrence of a magni-
tude 8-1/2 earthquake in 160 years, consistent with the repeat time
5 -30
Woodward· Clyde Consultants
of the 1964 earthquake as estimated by Davies and others (1981).
The area assumed for such an event is 69,498 square miles (180,000
km2). Because this source is at a large distance from the sites,
only the larger earthquakes would generate ground motions of
possible engineering significance. Thus, the distribution of
smaller earthquakes is not particularly important here.
5.4.2-The Intraplate Region
The recurrence relationship for the intraplate region was taken
from Woodward-Clyde Consultants (1978). In that study, recurrence
was estimated for a large section of the Benioff zone and was based
on earthquakes in the magnitude range of 5 to 6 and larger. To
test the applicability of this relation to the 6,564 square-mile
(17,000-km2) area considered here, two additional data points
were used, as shown in Figure 5-9. The first data point was for
magnitude ~4 (using earthquakes with focal depths greater than 19
miles [30 km]), and the second was for magnitude (Ms) ~6. These
two data points were derived from Agnew (1980). Recurrence data
obtained from the 1980 microearthquake study (Woodward-Clyde
Consultants, 198Gb) were extrapolated to magnitude ~4 and plotted
in Figure 5-9. These points suggest that the b-value should
be higher (perhaps as high as near 1.0) than that obtained from
Woodward-Clyde Consultants (1978). The Woodward-Clyde Consulta.nts'
(1978) relationship, therefore, may somewhat overpredict the events
of magnitude ~7.
5.4.3-The Talkeetna Terrain
The frequency-magnitude relationship for the Talkeetna Terrain
is based upon the three data points shown in Figure 5-9. The
point for earthquakes of magnitude ML 2_2 is taken from the
seismicity data recorded during the summer of 1980 (Woodward-Clyde
5 -31
Consultants, 1980b). The b-value observed for this data set, 1.48,
is very high as a representation of large-earthquake crustal
seismicity in other worldwide areas. Thus, only the number of
events of magnitude (ML) 2_2 was used here, not the b-value. As
was noted in the Interim Report (Woodward-Clyde Consultants,
1980b), this data set was strongly influenced by several clusters
of earthquakes, which may have resulted in the large b-value.
The data point for events of magnitude 2_4 represents the number
of events reported by Agnew (1980) for the 6,564 square-mile
(17 ,000 km2) study area with focal depths of <19 miles (<30 km).
This sample is assumed to be complete for the 15-1/2 year period
analyzed by Agnew (1980), but it may in fact be missing some events
of approximately magnitude 4 because of inadequate detection.
Also, the focal depths for these events are not necessarily
accurate, and events may be improperly included or excluded from
this data set.
One earthquake (of magnitude [Ms] 5-1/2 on 24 May 1931) has been
reported in the study area, as shown in Figure 5-1. This earthquake
was used to obtain the data point for events of magnitude 2_5-1/2
(Figure 5-9). The 3 July 1929 earthquake (magnitude [Ms] 6-1/4)
was assigned to the Benioff zone (Section 5-1) and is not included
in this frequency-magnitude relationship.
Five events are shown in Figure 5-1 to lie immediately east of the
study area. Three of these (the events of 1923, 1931, and 1933)
have magnitude (Ms) 5.6 and the same location, while the two
events of 1958 and 1963, reported by Tobin and Sykes (1966), are
assumed in this study to have magnitudes less than 5-1/2. As was
discussed in the Interim Report (Woodward-Clyde Consultants,
1980b) and as can be seen by comparing the seismicity patterns of
Figures 5-1 and 5-2, the likelihood is high that several if not
5 -32
Woodward·CQyde
most of these events should be associated with the Benioff zone at
a depth of about 25 miles (40 km) rather than with the crust at a
depth of 12 miles (20 km). Focal depths that would be sufficiently
accurate to resolve the structural association of these earthquakes
have not been determined or reported. However, the higher rate of
earthquake occurrence in the Benioff zone (relative to the crust)
favors their occurrence at depth.
If the earthquakes to the east of the study area were included in
an analysis and if the area were expanded to include much more of
the Talkeetna Terrain, the recurrence of magnitude (Ms) 5-1/2 and
larger earthquakes could increase by up to a factor of about two.
This would not significantly affect the results of the seismic
exposure analysis presented in Section 8.4.3. Such an increase (by
a factor of approximately two) in recurrence of larger events for
the Talkeetna Terrain would suggest that the rate of seismicity
in the Terrain is close to that of the Benioff zone within the
subducted Pacific plate. This suggested similarity is clearly not
correct. The 3 November 1943 earthquake (magnitude [Ms] 7 .3)
·was not included in recurrence considerations for the Talkeetna
Terrain. As discussed in Section 5.1.3, it occurred along the
western boundary of the Terrain and is not considered to represent
the earthquake potential within the interior of the Terrain, which
includes the Project sites.
5 .4 .4 -Summary
The frequency-magnitude relationships shown in Figure 5-9, and for
which parameters are listed in Table 8-2, were developed as inputs
to the seismic exposure analysis discussed in Section 8.4. These
relationships are considered to represent the average seismicity of
the regions discussed and do not reflect conservative or upper
5 -33
Woodward-Clyde Consultants
bound recurrence estimates, except in that the relationships are
limited by the maximum earthquake magnitude values selected for
each region. Further data analysis and a more detailed evaluation
of uncertainties in these relationships would be needed if these
relationships were to be used for purposes other than the seismic
exposure anlaysis described in Section 8.
5 -34
AAR~ETERS OF SELECTED HISTORICAL EARTHQUAKES
Date --
1 January 1975
29 June 1964
19 August 1948
3 November 1943
27 April 1933
4 July 19293
3 July 19293
21 January 19293
7 July 1912
27 August 19043
Notes • ~
Location Magnitudel
(Ms)
61.9°N, 149. r w 5.9 mb
62 .r N, 152.0°W 5.6 mb
63 .0° N, 105.5°W 6-1/4
61.9° N, 151.3oW 7.3
61.3°N, 150.8°W 7
64.2°N, 147 .9°W 6-1/2
62.3°N, 149.3°W 6-1/4
64.2°N, 148.0°W 6-1/4
63.8°N, 147 .5°W 7 .4
64.8°N, 151.5°W 7-3/4
Magnitude is Ms except where otherwise noted.
2. Focal Mechanism is not well constrained.
3. Location is from Sykes (1981).
Woodward-Clyde Consultants
De~th Mechanism
(miles)(km)
34 55 Normal
9-12 15-20 Thrust-oblique
55 90 Thrust2
10 17 Thrust
9 15 Thrust
crustal Stri ke-Sl i p2
25-31 40-50 Not Known
crustal Stri ke-Sl ip2
crustal Stri ke-Sl i p2
crustal Stri ke-Sl i p2
Woodward· Clyde Consultants
~
CR OEARTHQUAKES ANALYZED IN THE STRESS REGIME STUDY
d=---
Time (GMT) Latitude Longitude Magnitude Hypocenter
atalog No. 1 Date Hr Min Sec ON ow (ML) Dep t h (km)
14 8 Jul 1980 01:22:08 63.1 149.2 1.40 15
25 13 Jul 1980 10:17:45 63.2 148.8 2.53 9
49 19 Jul 1980 06:12:04 62.9 149 .1 1.79 6
103 6 Aug 1980 16:15:14 62.6 148.9 1.12 14
105 6 Aug 1980 11:36:51 62.6 148.9 1.07 15
116 9 Aug 1980 01:27:12 61.9 149.0 1.46 16
171 25 Aug 1980 16:17:09 63.1 149.3 1.31 17
172 25 Aug 1980 20:10:07 63.1 149.2 1.40 8
187 30 Aug 1980 06:33:17 62.6 148.9 1.70 16
197 1 Sep 1980 01:49:30 62.9 149.0 0.85 12
199 2 Sep 1980 05 :18:11 62 .5 149 .0 1.23 15 ,.
/
...--" -
l«l te • ---..::..:.
Catalog number is li sted in Appendix C of the Interim Report (Woodward-Clyde
Cons ul tants, 1980b). ·
s~MARY OF SELECTED BENIOFF ZONE EARTHQUAKES Woodward-Clyde Consultants
Date! ~
!9 NoV !906
22 ~~ar 1925
7 Mar 1929
23 Feb 1933
11 sep 1935
5 NOV 1938
5 NOV 1938
6 Nov !938
!4 sep 1943
z Aug 1946
1 Mar 1948
z Hov 1950
6 May 1953
4 May 1959
4 Feb 1965
12 Mar 1966
31 May 1970
14 Jul 1971
26 Jul 1971
17 Jun 1973
20 Jul 1975
20 Jul 1975
31 Oct 1975
14 Jan 1976
12 Jun 1978
14 Mar 1979
!!!.tes:
Lat1tude2,3 c· Nl
-22.0
-18.5
51.0
-20.0
43.0
36.8
37.2
37.2
-30.0
-26.5
-3.0
-6.5
-36.5
52.5
51.3
24.2
-9.2
-5.5
-4.9
43.2
-6 .6
-7.1
12.6
-29.2
38.2
17 .8
Longi tude3 ,4 c· E)
109.0
168.5
-170.0
-71.0
146.5
141.8
141.8
142.2
-177.0
-70.5
127.5
129.5
-73 .0
159.5
178 .6
122.6
-78.8
153.9
153.2
145.8
155.1
155.2
126.0
177.9
142.0
-101.3
MagnitudeS
(Hs)
7.75 (7 .5)
7.6
8.6 (7 .5)
7.6
7.6
7.7 (7 .7)
7.7 (7 .8)
7.6 (7.7)
7.6
60 .
50
50
40
60
60 (30)
60 (45)
60 (17)
60
Region
Java
Nt!'<l Hebrides
Aleutians
Peru
Kurils
Japan
Japan
Japan
Kermadecs
7 .9 (7-1/2) 50 Chile
7.9 (7-1/2) 50 Banda Sea
8.1 50 (220) Banda Sea
7.6 60 Chile
8 .0 (7 .7) 60 (74) K.sochatka
7.75 40 Aleutians
7.6 48 Taiwan
7.8 (7.6) 43 Peru
7.9 (7.8)
7.9 (7 .7)
7.7
7.9
7.7
47 (65) Solomons
48 (53) Solomons
48 Japan
49 Solomons
44 Solomons
Coornent 7
Hs=7.5, not 7.75
Inferred to be
i nterp 1 ate from
aftershock zone
Ms=7.5, not 8.6
Focal mechanism
unknown
Focal mechanism
unknown
Interplate
Interplate
Norma 1 fault
beneath trench
Focal mechanism
unknown
Ms=7-1/2, not
7.9; inferred
to be interplate
Hs=7-1/2, not
7.9; geometrically
c~lex arc
Focal Depth =
220 km
Foca 1 mechanism
unknown
Ms=7.7; M...=8.2
Interplate
Strike slip
Ms=7 .6; M...=7 .9
Interplate
I nterp 1 ate
Interplate
Inferred to be
interplate from
aftershock zone
Inferred to be
i nterp 1 ate from
aftershock zone
7.2 (7 .6) 50 Philippines Normal fault
beneath trench
7.7
7.7
7 .6
69
44
49
Kermadecs
Japan
Mexico
Inferred to be
interplate from
aftershock zone
Interp 1 ate
Interplate
l. These earthquakes are of Ms >7 .5 with focal depths between 35 and 70 km.
Reference
Geller and Kanamori (1977)
McCann (1981)
Geller and others (1978)
NOAA8
Abe (1977)
Abe (1977)
Abe (1977)
Gutenberg and Richter
(1954); Kelleher (1972)
Cardwell and !sacks
(1978); Gutenberg and
Richter (1954)
Abe and Kanamori (1979)
NOAA8
Abe and Kanamori (1980);
Seno (1981)
Stauder (1968)
Sudo (1972)
Abe and Kanamori (1980);
Abe (1972); Stauder (1975)
Magnitudes: Abe and
Kanamori (1980)
Focal Depth: Pascal (1979)
Tectonic Interpretation :
McCann (1981)
Magnitudes: Abe and
Kanamori (1980)
Focal Depth: Pascal (1979)
Tectonic Interpretation:
McCann (1981)
Shimazak i (1975)
McCann (1981)
I
McCann (1981)
Cardwell and others (1980)
McCann (1981)
Seno and others (1980)
Gettrust and others (1981)
2. Latitude is ·s when given 1n negative degrees (e.g., -9.2 is 9.2•s latitude).
3, Latitude and longitude are from the National Oceanographic and Atmospheric Administration (NOAA) locations. 4 · Longltude is ·w when given in negative degrees (e.g., -78.8 is 78.8•w longitude).
5. Magnitude is from NOAA. Alternate magnitude values are cited as (7.6). The source of the alternate magnitude
value is cited as a reference. 6 · Fo cal depth is from NOAA. Alternate focal depths are cited as (65). The source of the alternate focal depth is
cited as a reference. 7 · The comments summarize Woodward-Clyde Consultants' conclusions after reviewing the NOAA data and the sources cited
as references. 8 · This reference is the NOAA worldwide earthquake catalog available from NOAA in Boulder, Colorado .
63 .oo
27 AUG 1904
CRUSTAL
+
I
I
I
I
I
I
Microearthquake 1
Study Area~1
I
I
I
{
21 JAN 1929
CRUSTAL 4 JULY 1929
HINES CREEK
C) "TRAND
MCKINLEY ~
@
CANTWELL ________
1 --------I
A
+
29 MAY 1931
/
WATANA
SITE
I . sA.. .~0 ·~0
DENALI e, .~ .. ,.
I
I
I
I I
17 JULY 1923
17 OCT 1931
26 JULY 1933 I v7)
29 JUNE 1964
d = 15-20 km
I
I
I
I .·/
27 OCT 1958
19 JUNE 1963
I
I
(')
TALKEETNA'
I • ~-----------------
+
100 km
(') "' C)-3 NOV~1943
d = 17 km
1933
(')
• /1933
CJ ANCHORAGE
®
I
I
I
I
I
/
/
/
/
/
/ (')
/
/ -
I
I
A'
--
+
........ ..... .....
BROXSON GULCH
THRUST FAULT
62 .oo
~ I C)
I ;;;;----_.Q_---+~:::i1~93ij3 __ --f--------±-;::::--j_l_~-4itt------~r.v-~ J.OO Sl .oo ;'!:-C) 1941 I -00 -14 .o -\4 .so
-15 .QQ -150.00 -149.00
ARD-CLYDE CONSULTANTS 14658A December 1980
LIMIT OF 1964 EARTHQUAKE
AFTERSHOCK ZONE
LEGEND
REPDRTEO MAGNITUDE
C) a.o
C) 7.0
C) s.o
C) s.o
C) •• Q
!NTEN5!TY VX!l
~''
<:)x v IX
0 V!!!
0 VJ!
0 Y!
C>
1936 or
3 NOV 1943 Earthquake date of occurrence
d = 17 km Hypocenter depth
d == 17 km Approximate hypocenter depth
CRUSTAL Crustal -hypocenter depth not known
Faults with recent displacement
Faults with no observed evidence of recent
displacement, dotted where concealed
NOTES
1. Earthquakes of magnitude greater than 4 or intensity
greater than MM V are shown.
2. Magnitude symbol sizes are shown on a continuous
nonlinear scale.
3. Earthquakes are listed in Appendix C of Woodward-
Clyde Consultants (1980b).
4. Epicenter locations and hypocenter depths, where
revised from Woodward-Clyde Consultants (1980b),
are summarized in Table 5-1.
5. Section A-A'-B is shown in Figure 5-7.
~
-N-
~
HISTORICAL EARTHQUAKES OF FOCAL
DEPTH LESS THAN 30 km IN THE SITE
REGION FROM 1904 THROUGH 1978
0 10 20 30 40 50 Miles
F@ Ek 143 E'"""3
0 10 20 30 40 50 Kilometers
FIGURE 5-1
C)
C)
C)
C)
C)
C)
A
C) C)
C)
CANTWELL
@
C) (') DENALI
@
C) C)
63 .oo C) C) + ~ +
::"\DEVIL
(f) m.~~ ~
C) C) 19 AUG 1948
C) C)
CCI
CCI C)
~
CCI
C)
C)
C)
52 .oo C) C) C) +
C)
C) C)C":\:9 C)
C)
C) 8
C)
C)
C) C) C)
CCI C)
C) C)
61 .QQ
-152.00
DE CONSULTANTS 14658A Dacember
C)
C)
CCI
C) d = 90 km
C) (')
CCI C)
TALKEETNA
CANYON C)WATANA
~ ~
SITE C'J SITE C)
l{lJI C)
C)
CCI C)
C) 3 JULY 1929
/~= 40-50 km
C) C)
CCI
C) C) C) + C)
100 km radius C)
C)
C) CJ'--~C)~~~~~~~-----
1 JAN 197y
C) d = 55 km
• C) C)
C) C)
C) C) C)
C) C)C)C) CCI
ANCHORAGE (') C) ® C)
C)
C) I
I
I
I
C)
(')
I C)
LIMIT OF 1004 EARTHQUAKE
AFTERSHOCK ZONE
C)
+ 63 .oo
C)
+ 52.00
CCI CJ C)
C)
(')
LEGEND
REPBRTEO MAGNITUDE
C) s.o
u 7.0
C) 6-0
C) s.o
C) 4 .o
INTENSITY <(>XII
<Y X!
<Jx
<:> !X
~ Vll!
<:t> Vff
~ Y!
0
__.-Depth to Benioff zone in kilometers
---75 after Agnew ( 1980).
19 AUG 1948 Earthquake date of occurrence
d = 55 km Hypocenter depth
NOTES
1. Earthquakes of magnitude greater than 4 or intensity
greater than MM V are shown.
2. Magnitude symbol sizes are shown on a continuous
nonlinear scale.
3. Earthquakes are listed in Appendix C of Woodward-
Clyde Consultants ( 1980b).
4. Epicenter locations and hypocenter depths, where
revised from Woodward-Clyde Consultants ( 1980b),
are summarized in Table 5-1.
5. Section A-A'-B is shown in Figure 5·7.
HISTORICAL EARTHQUAKES OF FOCAL DEPTH
GREATER THAN 35 km IN THE SITE
REGION FROM 1904 THROUGH 1978
0 10 20 30 40 50 Miles
0 10 20 30 40 50 Kilometers
FIGURE 5-2
f
AlA IC.ON ~~0 . e,;OH
.,.OL
GE O HKCll
... ~'-M CAfte ~-.ADS
DAle 6~G wa~eoz
GOL
tKSe ~~;r
A. 29 JUNE 1964 EARTHQUAKE
Origin Time :
Location :
Magnitude (mb):
Depth :
LEGEND
• Compression
o Dilatation
x Nodal
0721 hr.
62. 7°N latitude
152.0°W longitude
5 .6
5 to 20 km
./ Fault or Nodal Plane ,
/ long dashes indicate s
/ low reliability,
, short dashes indicate
very low reliability
DE CONSULTANTS 41410A February 1982
I
I
I
I
I
I
I
I
I
I
I
B. 1 JANUARY 1975 EARTHQUAKE
Origin Time :
Location:
Magnitude (mb):
Depth :
NOTES
0355 hr.
61 .9°N latitude
149.7°W longitude
5.9
55 km
1. Lower hemisphere plots.
2 . Equal area projection.
FOCAL MECHANISMS FOR 29 JUNE 1964 AND
1 JANUARY 1975 EARTHQUAKES
FIGURE 5-3
p Wave
SH-Wave
---0 bserved (PAS)
~--Synthetic (PAS)
Observed (PAS)
----Synthetic (PAS)
OBSERVED AND SYNTHETIC SEISMOGRAMS FOR
P AND SH WAVES AT PAS FOR THE
3 NOVEMBER 1943 EARTHQUAKE
41410A February 1982 FIGURE 5-4
SH Wave
Radiation Pattern
LEGEND
(D) Dilatation
(C) Compression
(CC) Clockwise from above
(CW) Counter clockwise from above
KEW Azimuth to Station ---..
WARD-CLYDE CONSULTANTS 41410A February 1982
s
(D) ~It .
"'()_;
~
......... -<~.s "'~
SJG (CW\
~
$~,..
P and SV Wave /Jo
Radiation Pattern ~~ Cc_~
OBSERVED POLARITIES AND
THEORETICAL RADIATION PATTERN FOR
THE 3 NOVEMBER 1943 EARTHQUAKE
FIGURE 5-5
/
/GENERALIZED ~WESTERN BOUNDARY
/ OF THE TALKEETNA TERRAIN
1\lit Yenlo
D-CLYDE CONSULTANTS 41410A February 1982
B
D
Quaternary
Kag
Tz,lkeetna Thrust
0
Pzv
...
LEGEND
Tv
Kag
KJs
Vc
Pzv
---
u .... p
D •• . • ' u
Tertiary volcanic rocks
Cretaceous argillite and graywacke
Cretaceous -Jurassic marine sedimentary rocks,
undivided
Paleozoic and Triassic rocks (inferred from
aeromagnetic data)
Paleozoic volcanic rocks
Inferred lithologic contact
Strike-slip fault with recent displacement,
arrows show sense of horizontal displacement,
letters show sense of vertical displacement:
U is up; D is down, dotted where concealed
Strike-slip fault without recent displacement,
dashed where inferred, dotted where concealed.
Thrust fault without recent displacement,
dashed where inferred, dotted where concealed,
sawteeth on upper plate
Line separating aeromagnetically dissimilar
terrain, dashed where indistinct
U-2 lineament
LANDSAT lineament
Inferred fault from aero!':agnetic data,
letters show sense of veti ;Gal displacement,
U is up; D is down
WB, W9 Locations studied during this investigation
W D Watana Site
D U Devil Canyon Site
NOTES
1. Line separating aeromagnetically dissimilar terrain and
Mt Yenlo geology are shown by, or interpreted from
Csejtey and others ( 1978) and Griscom ( 1979).
2. The Talkeetna thrust fault and adjacent geology
are from Csejtey and others ( 1978\.
3. Castle Mountain fault location is from Magoon
and others (1976).
4. The 1943 epicenter location is from analysis con-
ducted during this investigation (Section 5.1.3) and
Tobin and Sykes (1966). h is the focal depth of
this earthquake.
5. Locations WB and W9 are shown in Figure 4-7
and discussed in Sections 4.4.1 and 5.1.3.
1943 EARTHQUAKE GEOLOGY MAP
FIGURE 5-6
0 0 0
0
0
NOTES
0
Approximate Upper Edge
of Benioff Zone \
0
0
0
1. Location of Section A-A'-B is shown in Figures 5-1 and 5·2.
2. The ear<:hquakes projected to the plane of this cross-section
are showr. in Figures 9-1 and 9·2 of Woodward-Clyde Con-
sultants (198Gb).
3. Seismic and Aseismic? delineation of the interplate region is
discussed in Section 5.
4. Focal mechanisms for earthquakes A and B are shown in
Figures 5-SA and 5-88, respectively.
E CONSULTANTS 41410A February 1982
0
I
I
I
I
B
SE
0
9J 00 I --
IAF 0
I
Po
I
-----------\;Limit of 1964 0 _---\ Earthquake
_-_.---Aftershock
---\Zone
~ -<!1--Seism ic--ll""
LEGEND
DEV£ Devil Canyon Site and Microearthquake Station Location
WAT A Watana Site and Microearthquake Station Location
Hypocenter Locations: 0 M=4.0
0 M 3.0
Q M = 2.0
0 M = 1.0
AF Aseismic Front after Yoshii (1975).
T Transition from the Aleutian Megathrust "Main Thrust" to
\ the Benioff zone of Davies and House ( 1979).
6'..; Distance in kilometers from proposed dam site to the closest
\approach of the respective regions of the Benioff zone.
Hypocenter of 30 August 1980 earthquake of
magnitude (mb) 3.6. Arrows show direction of
maximum stress.
B ~{\ ,__Hypocenter of 31 August 1981 earthquake of ~magnitude (mb) 3.4. Arrows show direction of
maximum stress.
TECTONIC INTERPRETATION AND
CROSS-SECTION OF CRUSTAL AND
BENIOFF ZONE MICROEARTHQUAKES
LOCATED WITHIN THE 1980 NETWORK
20 Miles
0 10 20 30 40 Kilometers
FIGURE 5-7
oo
;r
30 AUGUST 1980 EARTHQUAKE
Magnitude (mbl:
Depth:
LEGEND
1732 hr.
61.8°N latitude
148.9°W longitude
3.6
35 km
•T Minimum Compressive Stress Axis
• P Maximum Compressive Stress Axis
• Compression-high confidence in
interpretation
o Compression-low to moderate
confidence in
interpretation
0 Dilatation-high confidence in
interpretation
0 Dilatation-low to moderate
confidence in
interpretation
Fault or Nodal Plane
DE CONSULTANTS 4141 OA February 1982
..
B. 31 AUGUST 1980 EARTHQUAKE
Origin Time:
Location:
Magnitude (mb):
Depth:
NOTES
1552 hr.
62.2°N latitude
147.5°W longitude
3.4
35 km
1. Lower hemisphere plots.
2. Equal area projection.
FOCAL MECHANISMS FOR 30 AUGUST 1980
AND 31 AUGUST 1980 EARTHQUAKES
FIGURE 5-8
1000~---.---.---.r---~---r---.----,
QJ .6. "'0 ~-::::J .t: c
Ol BENIOFF ZONE co \ A: ~ (INTERPLATE) 1 ~N 10
"' E \ <lJ.,::.!
..:>!
coo ~ ::::Jo
rro ..c ~
t .... co ·~ ~ LlJ QJ BENIOFF ZONE Q.
.... "' (INTRAPLATE) 1 0 .... ·'\ co
.... QJ \.\ 15 >
EO
:::;0
z~ \ & ....
QJ QJ
> Q. 0.1 ·;:;
co TALKEETNA~\\
::::J TERRAIN .1 \ E
::::J u
0.01
0.001
2 3 4 5 6
Magnitude
LEGEND
Benioff Zone Intraplate Region
A Calculated from Agnew ( 1980) data
6. Calculated from Woodward-Clyde
Consultants' ( 1 980b) data.
Talkeetna Terrain
e Calculated from Woodward-Clyde
Consultants' ( 1 980b) data
0 Calculated from data reviewed
during this 1981 study 2
7
I
I
I
8
o Calculated from Agnew (1980) data
NOTES
FREQUENCY-MAGNITUDE
RELATIONSHIPS
1. Benioff zone interplate and intraplate
frequency-magnitude curves are
from Woodward-Clyde Consultants ( 1978).
2. Earthquakes reviewed for the Talkeetna Terrain
during this 1981 study are shown in Figure 5-1.
CONSULTANTS 41410A February 1982 FIGURE 5-9
Woodward-Clyde Consultants
-RESERVOIR-INDUCED SEISMICITY
e objective of this part of the investigation \'Jas to evaluate the
otential for the possible future occurrence of reservoir-induced
5 eismicity (RIS) in the vicinity of the proposed reservoirs. Reservoir-
induced seismicity is defined here as: the phenomenon of earth movement
and resultant seismicity that has a spatial and temporal relationship to
a reservoir and is triggered by nontectonic stress.
In the early 1940s in a study of Hoover Dam in the United States
(Carder, 1945), a relationship was first recognized between the level
impounded by a dam and the rate of occurrence of local earth-
quakes. Since that time, similar relationships have been reported for
99 other reservoirs around the world. A review of these reported cases
(Packer and others, 1977; Packer and others, 1979; Perman and others,
1981) resulted in 55 cases being classified as either accepted or
questionable cases of RIS (Table 6-1, Figure 6-1). A later reanalysis
of these cases resulted in 68 cases being classified as accepted cases
of RIS (Perman and others, 1981).
Several reservoir-induced seismic events (at Kremasta, Greece; Koyna,
India; Kariba, Zambia-Rhodesia; and Xinfengjiang, China) have ex-
ceeded magnitude (Ms) 6. Damage occurred to the dams at Koyna and
Xinfengji ang, and additional property damage occurred at Koyna and
Kremasta.
Studies of the occurrence of RIS (Simpson, 1976; Packer and others,
1977; Withers, 1977; Packer and others, 1979) have shown that RIS is
influenced by the depth and volume of the reservoir, the filling
history of the reservoir, the state of tectonic stress in the shallow
crust beneath the reservoir, and the existing pore pressures and perme-
ability of the rock under the reservoir. Although direct measurements
6 - 1
Woodward-Clyde Consultants
difficult to obtain for some of these factors, indirect geologic and
smologic data, together with observations about the occurrence of RIS
other reservoirs, can be used to assess the potential for and the
sible effects of the occurrence of RIS at the proposed Project
irs.
r.: scope of this study included: a) a comparison of the depth,
lume, regional stress, geologic setting, and faulting at the Devil
and Watana sites with the same parameters at comparable reser-
worldwide (discussed in Woodward-Clyde Consultants, 1980b);
assessment of the likelihood of RIS at the sites based on the
comparison; c) a review of the relationship between reservoir
filling and the length of time to the onset of induced events and
length of time to the maximum earthquake; d) an evaluation of the
stgnificance of these time periods for the sites; e) the development of
,:!. model to assess the impact of RIS on ground-motion parameters; f) a
review of the relationship between RIS and method of reservoir filling;
.:J..~ld g) an assessment of the potential for landslides resulting from
Items (a) through (d) were discussed in Woodward-Clyde Consultants
(],980b) and are updated in Section 6.1 of this report to reflect a
r1:::Gent addition to the data base by Perman and others (1981). Items (e)
t~H·ough (g) are discussed in Sections 6.2 through 6.4, respectively.
For this study, the two proposed reservoirs were considered to be
~;;'ni!f;! hydrologic entity (designated the proposed Devil Canyon-Watana
reservoir) because the hydrologic influence of the two proposed reser-
'u';:;r1rs is expected to overlap in the area between the Watana site and the
lii~i::;tream end of the De vi 1 Canyon reservoir. The proposed Devi 1 Canyon-
~;ljjj,f!;ana reservoir will be approximately 87 miles (140 km) long. Based on
/!peres American Inc. (in press) data, the predicted parameters for the
reservoir will be the following:
6 - 2
Devil Cani:on Watana Combined
Water Depth 551 ft (168m) 725 feet (221m) 725 ft (221m)
Water Volume 1.09x106 acre-feet 9.52x106 acre-feet 10.61x106 acre-feet
( 1 ,348x1o6m3) ( 11 ,7 41x1o6m3) ( 13 ,089x106m3)
s Regime Compression a 1 Compressional Compression a 1
Metamorphic Igneous Igneous
combined hydrologic body of water, as proposed, would constitute
;1!'~ry deep, very large reservoir within a primarily igneous bedrock
n that is undergoing compressional tectonic stress.
-Evaluation of Potential Occurrence
6.1.1-Likelihood of Occurrence
For comparative purposes, a deep reservoir has a maximum water
depth of 300 feet (92 m) or deeper; a very deep reservoir is 492
feet (150m) deep or deeper; a large reservoir has a maximum water
volume greater than 1x106 acre feet (1,234x106m3); and a very
large reservoir has a volume greater than 8.1x106 acre feet
(10,000x 106m3) as discussed in Packer and others (1977). Twenty-
one percent of all deep, very deep, or very large reservoirs
have been subject to RIS. Thus, the likelihood that any deep,
very deep, or very large reservoir will experience RIS is 0.21.
However, the tectonic and geologic conditions at any specific
reservoir may be more or less conducive to RIS occurrence.
Models have been developed by Baecher and Keeney in Packer and
others (1979) to estimate the likelihood of RIS at a reservoir,
characterized by its depth, volume, faulting, geology, and stress
regime. The models from which the likelihoods are calculated are
6 - 3
Woodward-Clyde Consultants
sensitive to changes in data classification for the geologic and
stress regime (Packer and others, 1979; Perman and others, 1981).
The calculations from models, however, do not significantly
influence the basic relatively high likelihood of RIS at the Devil
Canyon-Watana reservoir considering its depth and volume. The
calculation of the likelihood of occurrence of RIS at the Devil
Canyon-Watana reservoir is presented in Section 6.2.2.
6.1.2-Location and Maximum Magnitude
Packer and others (1979) and Woodward-Clyde Consultants (1980b),
among others, have discussed the concept, based on theoretical
considerations and existing cases of RIS, that an RIS event is
a naturally occurring event triggered by the impoundment of
a reservoir. That is, reservoirs are believed to provide an
incremental increase in stress that is large enough to trigger
strain release in the form of an earthquake. In this manner,
reservoirs are considered capable of triggering an earlier occur-
rence of an earthquake (i.e., of decreasing the recurrence interval
of the event) than would have occurred if the reservoir had not
been filled. In this regard, reservoirs are not considered capable
of triggering an earthquake larger than that which would have
occurred 11 natura lly. 11
The portion of crust that a reservoir may influence is limited to
the area affected by its mass and pore pressure influences. This
area of influence is often referred to as a reservoir's hydrologic
regime. Defining the precise extent of the hydrologic regime of a
reservoir is a complicated task and is the subject of numerous
studies (e.g., Withers, 1977; Withers and Nyland, 1978). However,
documented cases of RIS (Packer and others, 1979) indicate that the
RIS epicenters occur within an area that is related to the surface
area that the reservoir covers. For the purposes of this study,
6 - 4
Woodward-Clyde Consultants
the hydrologic regime of the proposed reservoir has been described
as an envelope with a 19-mi le (30-km) radius that encompases the
reservoir area, as discussed in Woodward-Clyde Consultants (1980b).
Previous studies (e.g., Packer and others, 1979) present evidence
that strongly suggests that moderate to large RIS events are
expected only to occur along faults with recent displacement.
Among the reported cases of RIS, at least 10 have had magnitudes
of (Ms) .2_5 (Table 6-2). Field reconnaissance and information
available in the literature indicate that Quaternary or late
Cenozoic surface fault rupture (i.e., rupture on faults with recent
di sp 1 a cement) occurred within the hydrologic regime of eight of
these ten reservoirs (Packer and others, 1979).
On the basis of this investigation, it has been concluded that
there are no faults with recent displacement within the hydrologic
regime of the proposed reservoir (Section 4.5). Therefore, the
maximum earthquake which could be triggered by the reservoir
is an earthquake with a magnitude below the detection level
of currently available techniques (i.e., the detection level
earthquake discussed in Section 4.2.4). Thus, the magnitude of
the largest earthquake that could be triggered by the proposed
reservoir is judged to be (Ms) 6, which is the maximum magnitude
of the detection level earthquake.
-Effect of RISon Earthquake Occurrence Likelihood
.!~, wobabilistic seismic exposure analysis was performed during this
·in'!/i;::•stigation to evaluate the likelihood of exceeding design levels of
!:rr1:::,u:nd motion at the Devil Canyon and Watana sites (Section 8). An
'1iil1~:H:wtant input to such an analysis is the rate (or recurrence) of
!f:::,:~rt.hquake activity on potential seismic sources in the vicinity of the
6 - 5
Woodward-Clyde Conso.dtants
As discussed in Section 6.1, the impoundment of a very deep and
large reservoir may trigger the occurrence of earthquakes that
1d not have occurred otherwise within the design life of a dam. The
jective of the model described below is to characterize the likelihood
occurrence of RIS events given the specific seismologic and tectonic
ting of the Devil Canyon and Watana sites. The information developed
this model has been included in the seismic exposure analysis in
A description of the overall model and the implementation
various components of the model are discussed in Sections 6.2.1 and
6.2.1 -Description of the Model
The basic approach consists of:
1) calculating the likelihood of occurrence of earthquakes within
an area and within a time p~riod affected by RIS;
2) converting the likelihood of occurrence of earthquakes into
the mean number of earthquakes within the specified area and
time;
3) distributing the mean number of earthquakes among faults with
recent displacement (if present) within the specified area or
distributing the earthquakes randomly throughout the specified
area if faults with recent displacement are not present; and
4) using the expected number of earthquakes on each fault or
within the specified area in the seismic exposure analysis to
obtain the contribution of that fault or the specified area
(when faults with recent displacement are not present) to the
probability of exceeding a given level of ground motion.
6 - 6
Woodward-Clyde Consultants
A brief description of each step in the approach follows.
1. Calculation of Likelihood of Occurrence of RIS Events
The likelihood of occurrence of earthquakes triggered by
the impoundment of reservoirs can be calculated from a
prediction model developed by Baecher and Keeney in Packer and
others (1979). The prediction model is based on statistical
discriminant analysis and is calibrated by the data to provide
an approximate estimate of the likelihood of RIS events for
specific sites. The model incorporates the worldwide RI S
data base and the historical seismicity of the site region
(Section 5); it is designed to predict the likelihood of
occurrence of an RI S event with magnitude (Ms) 2_4. Events
of lesser magnitude were deleted because they are considered
to be too small to have a significant contribution to seismic
design.
2. Calculation of Mean Number of RIS Events
From the likelihood of occurrence of RIS events, one can
calculate the mean number of RIS events by assuming a Poisson
model. The following equation can be used:
= -1n (1-p) ........ Equation 6-1
where ;.. = mean number of RIS events within an area and
time period affected by reservoir impoundment.
and p = probability of occurrence of RIS events within
the assumed area and time period.
Estimates of the time period and area within which RIS events
would generally occur can be obtained from the analysis of a
6 -7
Woodward-Clyde Consultants
large data base of characteristics of the major reservoirs
of the world (Packer and others, 1979; Perman and others,
1981). The time between reservoir impoundment and the largest
suspected RIS event ranges from 0 to 25 years, with most of
such events occurring within 5 to 10 years.
For long, thin reservoirs (such as the proposed Devi 1 Canyon-
Watana reservoir) most of the RI S events (80 to 90 percent)
are located within a three-dimensional space that has the
configuration of a half-pipe (the bottom half). This half-
pipe space encompasses the hydrologic influence of the
reservoir and typically has a radius equal to three times the
width of the reservoir (Withers, 1977).
For this study, the maximum width of the proposed reservoir
was defined as 6 miles (10 km) at Watana Creek. The radius of
the half-pipe space then is 19 miles (30 km).
For the purposes of model calculations, the half-pipe space
was converted to a rectangular three-dimensional space. The
length and width of the rectangular space is 37 miles by
37 miles (60 km by 60 km) vJhich is twice the 19-mi le (30-km)
distance cited above. The depth of the rectangular space is
19 miles (30 km) corresponding to the half-pipe radius of
19 miles (30 km). This rectangular space was centered about
each site, such that the distance from the site to the edge of
the space in all three dimensions \vas 19 miles (30 km). This
con f i g u rat i on was used to f a c i 1 i tate mode 1 c a 1 c u 1 at i on s
because a rectangular space is easier to model than a cylind-
rical space and because the effect of ground motions from a
RIS event that might occur more than 19 miles (30 km) from
either site would be negligible.
6 - 8
Woodward-Clyde ConsuHanri:s
3. Distribution of Mean Number of RIS Events
It is generally believed that stresses resulting from reser-
voir impoundment can affect the timing of earthquakes. An
earthquake that would have occurred in the vicinity of
a reservoir because of natural seismicity may be triggered
sooner because of RIS (i.e., its recurrence interval may
be reduced). Because there are no faults with recent dis-
placement within the hydrologic regime of the reservoir, it
would seem appropriate to assign the mean number of RIS events
to discrete units of volume within the hydrologic regime of
the reservoir. For the present study, the extent of the
hydrologic regime in which RIS events are expected to occur
is defined by the rectangular space described in (2) above.
An earthquake with a magnitude up to that of the detection
level earthquake is assumed to be able to occur on a source
anywhere within this three-dimensional space.
Given the above guidelines, the volume units within which
RIS events are to be distributed are defined. Then the
seismicity may actually be distributed by proportioning the
number of events according to the mean number of events that
would have occurred naturally. Since RIS would be expected to
decrease the recurrence i nterva 1 between earthquakes, the mean
number of RIS events for a given area would generally be
greater than the mean number of naturally occurring events for
the s arne area.
4. Use of RIS Events in the Seismic Exposure Analysis
In order to perform the seismic exposure analysis, it is
necessary to know not only the mean number of RIS events
greater than some minimum magnitude of interest but also
6 - 9
the distribution of these events over the appropriate magni-
tude range (defined by the b-slope), and the size of the
maximum credible earthquake (MCE) at which the earthquake
recurrence curve would be truncated. The values of both of
these parameters for RIS events (b-slope and the size of
the MCE) are assumed to be equal to the values of the same
parameters for naturally occurring earthquakes. This is
consistent with the hypothesis that RIS only shifts the
timing of earthquakes and does not have a significant effect
on the magnitude distribution or the magnitude of a maximum
credible earthquake.
6.2.2-Implementation of the Model for the Susitna Project
The implementation of the four steps of the model for the Susitna
project is discussed in this section. Since the potential earth-
quake sources were at different distances from the two sites (Devil
Canyon and Watana), the analysis using these sources was performed
separately for each site.
1. Calculation of Likelihood of Occurrence of RIS Events
Baecher and Keeney (in Packer and others. 1979) have discussed
two mode l s for cal c u l at i n g the l i k e l i h o o d of occurrence
of RIS events: in one model, reservoir characteristics
(depth, volume, stress state, and geology) are assumed to
be independent; in the other, dependence between reservoir
depth and volume is assumed. For the Devil Canyon-Watana
reservoir, the first model produced an expected likelihood of
0.37 for a RIS event (of any magnitude) with a standard
deviation of 0.13, while the second produced an expected
likelihood of 0.46 with a standard deviation of 0.22. Since
some dependence between reservoir depth and volume would be
6 -10
expected and since the assumption of dependence produces a
more conservative result (i.e., a higher likelihood for a RIS
event to occur), the results of the second model were used for
this study.
The worldwide RIS data base on which the Baecher and Keeney
in Packer and others (1979) model was developed is limited.
To accommodate the uncert ai nt i es associated with the 1 imi ted
data base, the likelihood of occurrence of RIS events was
assumed to be the mean plus one standard deviation value
(i.e., 0.46 + 0.22 = 0.68). In order to examine the sensi-
tivity of results to this assumption, a mean plus two standard
deviation value (0.46 + 2 x 0.22 = 0.90) was also analyzed.
2. Calculation of Mean Number of RIS Events
From equation 6-1, the mean plus one standard deviation number
of RIS events with magnitude ~4 was calculated to be 1.14;
for M ~5, it was calculated to be 0.93. It was assumed that
these events would be expected to occur within 10 years after
the reservoir was impounded and that they would occur within
the three-dimension a 1 rectangular space described in Section
6.2.1, Item 2. After the first 10 years, only the naturally
occurring seismicity was assumed to occur during the remaining
design life of the dam.
3. Distribution of Mean Number of RIS Events
As there are no known faults with recent displacement within
an area of 37 miles by 37 miles (60 km by 60 km) around each
site (Section 4.5), the mean number of RIS events for use in
the seismic exposure analysis was distributed as a random
source over a rectangular space of 37 miles by 37 miles by
19 miles (60 km by 60 km by 30 km).
6 -11
4. Use of RIS Events in Seismic Exposure Analysis
The mean number of events was calculated for ground motions
that exceeded a given level of peak ground acceleration in the
first 10 years. The rate of naturally occurring sei smi city
was used to ca leu 1 ate the mean number of events (for which
ground motions exceeded the given level of peak ground
acceleration) during the design life of the proposed dams
after the first 10 years. The sum of the mean number of
incremental events resulting from RIS during the first
10 years and the mean number of events due to natural seismi-
city during the remainder of the proposed dam design life
yields the total mean number of events for which ground
motions are expected to exceed a given level during the
design life of the Project. The results of these calculations
are included in the analysis presented in Section 8.
Method of Reservoir Filling
The occurrence of RIS events has often been correlated with rapid
initial filling of a reservoir, especial-ly with irregular filling
histories or rapid reservoir refill following major drawdowns (Packer
Ltnd others, 1979). The precise relationship between irregularities
in the filling cycle and the occurrence of RIS events is not well-
dn!:;umented in most cases. Furthermore, no controlled experiments have
b'!i~:!~n performed at reservoirs to vary fi 11 i ng rates and examine the
,;:~i'fect on seismicity. However, detailed information is available on the
correlation between seismicity and filling rates for at least one
reservo1r--Nurek, U.S.S.R.
i\'lthough impoundment at Nurek began in 1968, the first significant
iin::!poundment (328 feet [100m]) took place between late August and early
6 -12
ovember 1972. A step was made in the filling curve late in September;
llowing this step, seismicity increased. Upon completion of the first
1i: .. !iige filling cycle, seismicity reached a peak with maximum magnitudes
~~1;!~1,) of 4 .6 and 4.3. Sei smi city between November 1972 and June 1976
adly paralleled changes in water level (Simpson and Negnatullaiv,
"18) ,! •
the basis of this experience, it was recommended that second-stage
lling of the Nurek reservoir, resulting in a water depth of 656 feet
be accomplished by a smooth filling cycle with no abrupt
towdowns in filling rate. Seismicity remained low during this filling
Unnitil a minor but rapid fluctuation in filling rate occurred in August
Following this fluctuation, there was a pronounced increase in
smicity, along with the occurrence of the largest earthquake reported
that time, a magnitude (Ms) 4.1 earthquake. It has been implied
the increase in seismicity during this second filling cycle may
f,·::.(~'\!'e been directly related to the sudden change in rate of filling
Negnatullaiv, 1978; Keith and others, 1979).
Fr1;1m this experience at Nurek, and from consideration of the correla-
thms between filling curves and seismicity for other cases of RIS, it
ap~~~1ears that sudden changes in water level and sudden deviations in
f'.!:!:'!i:,e of water level change can be triggers of induced seismicity. A
~r::r.mtrolled, smooth filling curve, with no sudden changes in filling
:f',!::t'i!;e, should be less likely to be accompanied by induced seismicity than
rc:tpid, highly fluctuating filling rates.
6,4 -Potential for Landslides in the Devil Canyon-Watana Reservoir
Area Resulting from RIS
[1J::y assessment of the potential landslides in the Devil Canyon-Watana
~"·!f.!::l1ervoir (this area is considered to include the banks of the present
6 -13
e of the Susitna River) resulting from RIS should be considered
Ulin the context of the overall potential for landslides and rockfalls
'!i;he reservoir area. That is, the potential for landslides which can
triggered by impoundment of the reservoir by natural processes (such
freeze-thaw conditions) as well as by RIS should be considered.
thin this context, we have considered the potential for landslides
ggered by RIS by making a preliminary assessment of whether in situ
tions suitable for landslides exist in the proposed reservoir area,
whether earthquakes are likely to release enough energy to trigger
Detailed studies of potential landslide-prone areas
outside the scope of this investigation and therefore, were not
cted. Consequently, the judgment presented below represents a very
assessment of the potential for RIS induced landslides.
investigation, a very preliminary assessment of landslide
j:){Jitlj:mtial was made from remotely-sensed data interpretation, review
previous studies conducted for the project, and limited aerial
!'"•ii~{:{mnaissance. On the basis of this assessment, it is concluded that
1;1'j!~; potential exists for landslides to occur in the reservoir area.
,f\ iHS event occurring within the hydrologic regime of the reservoir
could trigger landslides if the earthquake occurred close enough
to a potential slide area and if it released sufficient energy to
tr·!if:~!ger a slide. Within the scope of this investigation, the location
td:' a RIS event within the hydrologic regime of the combined reservoir
•r.;:.!:~nnot be estimated with sufficient precision to provide a meaningful
i:::::~:~!:~:Ssment of where in the reservoir area a landslide could occur, how
li~~qj:e the landslide would be, and how large an earthquake would be
!iii!::~·~:~;,:!.ssary to trigger a landslide. Given these constraints and the
(;:(:·~'':figuration of the Susitna River valley, the likelihood of a large
'i!,;;:r!d:s 1 ide in the proposed reservoir during a reservoir-induced earth-
qu;!,~ke appears to be low. This judgment should be reviewed when final
da~ design is considered.
6 -14
TABLE 6-1
REPORTED CASES OF RESERVOIR-INDUCED SEISMICITY (RIS)
~2 Dam Name 1 Reservoir Name3 Country Classification-of RIS
Magnitudo of Largest4
RIS Event
Akosootlo Main~ Lake Volta Ghana Accepted, macro
Almendra, Tonnes Reservoir Spain Accepted, micro
Bajina Basta Yugoslavia Accepted, micro
4 Benroore New Zealand Accepted, macro and mi era
5 Blowering Australia Accepted, macro and mi era
6 Cabin Creek. USA Not RlS
Cajuru Brazil Questionable
Cll11ari llas Spain Accepted, macro
Cane lles Spain Accepted, macro
10 Clark Hill USA Accepted, micro (macro!)
11 Contra, Lake Vogorno Switzerland Accepted, micro
12 Coyote Valley, lilke Mendocino USA Accepted, macro
13 El Grado Spain Not RIS
14 Etrosson Switzerland Accepted, micro
15 Euculibene Australia Accepted, macro
16 Fairfield, Lake Monticello USA Accepted, mi era
17 Ghirni India Questionable
18 Grancarevo Yugos hvi a Accepted, micro
19 Grandva l France Accepted, macro and mi era
20 Hendrik Verwoerd South Africa Accepted, micro
21 Hoover, lake Mead USA Accepted, macro and mi era
22 ltezhitezhi Z.OOia Accepted, macro
23 Joe as see USA Accepted, macro and mi era
24 Kanafusa Japl!l1 Accepted, micro
25 Kari ba Zmi a/Rhodesia Accepted, macro and m1 era
26 Kas trak i Greece Accepted, macro
27 Keban Turkey Accepted, mi era
28 Kerr, Flathead Lake USA Accepted, macro
Kinarsani lndi a. Question ab 1 e
29 Koyna, Shivaji Sagar Lake India Accepted, macro and m1 era
30 Kremasta Greece Accepted, macro and mi era
31 Kurobe Japan Accepted, macro and mi era
32 LaCohilla Spain Questionable
33 La Fuensanta Spain Questionable
34 Mangalam India Questionable
35 Mangl a Pakistan Not RlS
36 Mani cougan Canada Accepted, macro and mi era
37 Marathon Greece Accepted, macro
38 Mica Canada Not RIS
39 Monteynard France Accepted, macro
40 Mula India Accepted, micro
41 Nurek. USSR Accepted, macro and micro
42 Oro vi 11 e USA Accepted, macro
43 Oued Fodda Algeria Accepted, micro
44 Palisades USA Accepted, micro
45 Paralibikulam lndi a Questionable
46 Pi astra Italy Accepted, macro and mi era
47 Pieve di Cadore Italy Accepted+ macro and micro
48 Porto Colombia Brazil Accepted, macro
49 Rocky Reach USA Not RIS
50 San Luis USA Not RIS
51 SanfOTd USA Not RIS
52 Schlegels Austria Acce~ted, mi era
53 Sefid Rud Iran Que•tiooab1e
Sharavathi India Questionable
54 Shasta USA Accepted, micro
55 Sholayar India Questionab~e
56 Talbingo Australia Accepted, macro and mi era
57 Uk.ai India Questionable
58 V.ajcnt Italy Accepted, micro
59 Vo1ta Grande Bra.zi1 Accepted, macro
60 Vouglans France Accepted, macro
61 Warragamba~ Lak.e Burragorang Australia Questionable
62 Xinfengji ang Chjna Accepted, macro and mi era
~
1. Data source: Packer and others {197g).
2. film'Oers corresj)onel to nurTbers in Figure 6-1; Kinarsani and Sharavathi are unplatted because
of insufficient data.
3. Where on 1y one narre is given. either the reservoir name is the s anlE.' as the dam name or only
the darn name is k.nOW'n.
4. A dash indicates the magnitude was not obtained. Intensities are given in the Modified
Mercalli Scale of Wood and Neumann (1931) for cases in whid1 this value was given in lieu
of the Richter magnitude.
Intensity v
less than 2
less than 3
5 (1)
3.5
Approx.
4.1
4.7
4.3 (1)
less than
5.2
Less than 3
5 (1)
2.8
Less than 3
Intensity V
Less than 2
5.0
4 or
3.2
Less
6.25
4.6
Less
4.9
6.5
6.3
4.9
4.1
5.75
less (1)
than 3
than 3
Intensity Vll
less than 1
4.5
5.7
Less than 3
3.7 (1)
4.4
Intensity V
Intensity Y! to Vl!
Less than 2
4.7
less than 3
3.5
Less than 3
Less than 4
4.4
5.4
TABLE 6-2
RESERVOIR-INDUCED SEISMIC EVENTS WITH MAXIMUM MAGNITUDE OF 5 OR GREATERl
Active Fault2
Dam Reservoir Magnitude Present
Koyna Shivaji Sagar Lake 6.5 Yes3
Kari ba Lake Kari ba 6.25 Not obtai ned4
Kremasta Lake Kremasta 6.3 Yes3
Xinfengji ang Xinfengji ang 6.0 Yes
Marathon Lake Marathon 5.75 Not obtai ned4
Oroville Oroville Reservoir 5.7 Yes
Coyote V a 11 ey Lake Mendocino 5.3 Yes
Benmore Lake Benmore 5.0 Yes3
Eucembene Lake Eucembene 5.0 Yes3
Hoover Lake Mead 5.0 Yes3
Notes:
1. Data Source: Packer and others (1979).
2. Active faults are those defined as having displacement
in the present tectonic stress regime.
3. Determination is based on field reconnaissance studies.
4. The presence of an active fault has not been obtained but is
considered probable because of the tectonic setting.
240
220 i
I
200
I
18o 1
160
140
120
100 ,,32
80 r
60
40
20
LEGEND
.sa
11
~~56
: ~-,60
..... . "'::::,139_,-
: .i-~~ '
:.~2 ..... . ·.
r-::;12.
8
Watana
42
Combined
. 27
25
·• 47 -••
0 .;~··J-•
~8
'8 24 lS!£
·. ~ ....
{55
~ 44
51~
49 40128 1 ~ 820
__ J6 ° G 1~,48 ...... ~
-·~· ,62 ,~,
.·
Approx·imately 11,000 reservoirs
without reported RIS not plotted
r;:::c 12
'L:~7
100 1,000
~-ss
S:::.J
10,000 100,000 500,000
Reservoir Capacity in 106 m3 (1ogarithmic scale)
Noto: The following r0§arvoirs ware not plotted because of
insufficient data: KinorUJnl, Sharavstr•i.
•41 • Nurnk (USSR) d0pth is in excess of 285m.
Deep and/or very large reservoir
Accepted e<~se of R IS, maximum magnitude 2: 5
Accepted e<~se of RIS, maximum magnitude 3-5
Accepted case of R IS, maxi mum magnitude S 3
Questionable case of R IS
Not RIS
CONSULTANTS 41410A February 1982
PLOT OF WATER DEPTH AND VOLUME
FOR WORLDWIDE RESERVOIRS AND
REPORTED CASES OF RIS
FIGURE 6-1
.i
I
!
l
I
Woodward-Clyde Consultants
J -MAXIMUM CREDIBLE EARTHQUAKES (MCEs)
The approach to estimating the maximum credible earthquakes (MCEs) in a
region, and thereby to establishing a basis for estimating the ground-
motion parameters at a specific site, is based on the premise that
significant earthquake activity is associated with faults with recent
displacement. The eva 1 u at ion of the MCE that may be associated with
a given fault is closely related to the geologic and seismologic
setting in the site region. Therefore, it is necessary to identify the
characteristics of the faults with recent displacement in order to
assess their seismic source potential. For this study, the only faults
considered to have been subject to recent displacement within or
adjacent to the site region are the Castle Mountain fault and the
Denali fault. The Benioff zone passes at depth beneath the site
and is also considered to be a potential seismic source. These three
potential sources are the Talkeetna Terrain boundary faults discussed
in Section 4. 3.
In addition to assessing the MCE for seismic sources in or adjacent to
the site region, the size of the maximum earthquake that could occur
on a fault with recent displacement that might not have been detected by
our geologic investigation was evaluated. This earthquake has been
designated the detection level earthquake as discussed in Section 4.2.4.
The selection of ground-motion parameters for the project was based on
both deterministic and probabilistic approaches (Section 8). For the
deterministic approach, MCEs were estimated for each of the seismic
sources and for the detection level earthquake. The closest distance of
these sources from the Watana and Devil Canyon sites were estimated and
used in the seismic design analysis. The MCEs and their distance from
the sites are summarized in Table 7-1. The 13 significant features near
7 - 1
sites were judged either not to be faults or to be faults without
,1;mt displacement; therefore, they were not included in this analysis.
deterministic approach can be relatively conservative as it includes
assumptions that the MCE will occur during the lifetime of the
i:lity and will occur at the closest approach of the respective
rces to the sites. The approach does not take into account whether
e assumptions are geologically or seismologically reasonable.
probabi listie approach models the occurrence of earthquakes using
geological and seismological characteristics of the site region.
approach, estimates were made of the ground motions that
during the life of the Project. The MCEs estimated for the
~;mic sources are the same as those estimated for the deterministic
In addition, the recurrence interval, maximum rupture length,
displacement, fault geometry, and slip rate were estimated for
the MCEs and their sources along with their b-slope. The MCE
in the probabilistic ground-motion analysis are summarized in
and discussed in Section 8.
~~~probabilistic analysis specifies the contribution of each of
U"H~: seismic sources to the overall ground motions. This approach is
d,r;r~:~iigned to provide a more realistic model of seismic ground motions to
'ii'i'h'lth the site may actually be subjected than does the deterministic
,~:pp:r-oach. Using the probabilistic model, a systematic evaluation is
!:fi'l;;:1.d!~: of the ground motions which could result from the MCE, not only
,£!rt the closest distance to the site, but also at different locations
~j'iin;~'lg a particular seismic source. The approach also incorporates the
recurrence interval of earthquakes on the sources and uses these data to
,~::;::;~~":~ss the likelihood of occurrence during the lifetime of the project.
7 - 2
Outside the Talkeetna Terrain
sources outside the Talkeetna Terrain, such as the mega-
thrust zone at the Aleutian Trench or the Fairweather fault, are not of
significance to the Project because of the distance of these sources
from the Project and because of the presence of seismic sources such as
the Denali fault and the Benioff zone that are closer to the Project.
Even if it is assumed that a magnitude (Ms) 8.5 event could occur on a
known seismic source outside the Talkeetna Terrain, the resultant ground
motions would be significantly less than those for the Denali fault or
the Benioff zone. Consequently, MCEs associated with seismic sources
outside the Talkeetna Terrain have not been considered further for this
investigation.
7.2-Talkeetna Terrain Boundary Sources
The MCE s were estimated for three of the boundaries of the Talkeetna
Terrain. These boundary sources are the Castle Mountain fault to the
south, the Denali fault system to the north and east, and the Benioff
zone at depth.
7.2.1-The Castle Mountain Fault
The MCE for the Castle Mountain fault is estimated to be a magni-
tude (Ms) 7-1/2 event. This estimate is based on the following
assumptions:
1) The length of the Castle Mountain fault, 295 miles (475 km),
is considered to be a discrete zone of crustal weakness that
would be subject to an earthquake during a period of strain
release;
7 - 3
2) The Castle Mountain fault is a strike-slip fault according to
the definition of Bonilla and Buchanan (1970);
3) Using Slemmons' data (U.S. Nuclear Regulatory Corrmission,
1981) for strike-slip faults, the maximum rupture length that
could occur during a single earthquake is estimated to be
55 miles (89 km); and
4) The rupture length cited in (3) would be expected to be
associated with an MCE of magnitude (Ms) 7-1/2 using Slemmons•
(1977b) relationship for magnitude vs. rupture length for
strike-slip faults.
7 .2.2-The Denali Fault
The MCE for the Denali fault is estimated to be a magnitude
(Ms) 8 event. This estimate is based on the following assumptions:
1) The part of the Denali fault closest to the Project sites
includes the Togiak-Tikchik, Holitna, Farewell, and Shakwak
Valley (west of the Totschunda fault) fault segments and the
McKinley strand described by Grantz (1966) and discussed in
Section 4.3. The total length of the fault considered for
this analysis is 670 miles (1,080 km);
2) The fault length described in (1) above is the longest section
of the Denali fault near the Project sites that appears to be
a discrete zone of crustal weakness that could rupture in an
earthquake;
3) The Denali fault is a strike-slip fault;
7 -4
Woodward·CByde ConJsuHants
4) An estimated maximum rupture length of 178 miles (287 km) may
be postulated to occur during a single earthquake using
Slemmons' data (U.S. Nuclear Regulatory Commission, 1981) for
strike-s lip faults; and
5) The rupture length cited in (4) would be expected to be
associated with an MCE of magnitude (Ms) 8, using Slemmons'
(1977b) relationship for magnitude vs. fault rupture length
for strike-slip faults.
7 .2.3-The Benioff Zone
Interplate Region
The MCE for the interplate region of the Benioff zone is esti-
mated to be the equivalent of the 1964 Prince William Sound,
Alaska, earthquake of Mw 9.2. (For consistency of presenta-
tion, elsewhere in the report we have used the Ms 8-1/2 magnitude
designation for this same event. It is recognized that the
Ms scale has substantial shortcomings as an adequate indicator
of energy release and size,when an earthquake reaches a magnitude
of about (Ms) 8. However, the Ms 8-1/2 designation is used
here to represent the 1964 Prince William Sound earthquake size
and to correspond with the Mw 9.2 designation.) The closest
approach of the interplate region to the Watana and Devil Canyon
sites is 40 miles (64 km) and 57 miles (91 km), respectively.
These distances are measured to the 22-mile (35-km) depth contour
on the interface between the subducting Pacific plate and the
North American plate (Figure 5-7). As discussed in Section 5,
this depth is assumed to mark the down-dip limit of great shallow
interplate earthquakes and the closest distance to the sites at
which the MCE could occur.
7 - 5
The assumptions used to derive the i nterp 1 ate region MCE are
summarized in the following paragraphs.
The MCE that could occur on the interface of any of the worldwide
Benioff zone interplate regions is estimated to be magnitude
(Mw, moment magnitude) 9.5. The fault rupture area and magni-
tude of this earthquake exceed the corresponding values for the
1964 Prince William Sound earthquake of magnitude (Mw) 9.2. In
addition, the assumed magnitude (Mw) of 9.5 is equal to the
magnitude of the largest earthquake that has occurred this
century, namely the 1960 earthquake near Chile.
It is considered unlikely that an earthquake of magnitude (Mw)
9.5 would occur on the interplate region interface. Rather, it
appears much more probable that future great earthquakes would
rupture approximately the same plate boundary segment that
ruptured in the 1964 earthquake in Alaska. As was the case in
1964, the magnitude (Mw, moment magnitude) would be expected
to be 9.2 and the closest distances to the Watana and Devil
Canyon sites would be 88 miles (142 km) and 108 miles (174 km),
respectively. These distances to the sites are the closest
approach of the area that ruptured in the 1964 earthquake and
are considerably greater than those used for seismic design
(i.e., 40 miles [64 km] and 57 miles [91 km], respectively).
One of the more conservative assumptions that could be made about
the interplate region earthquake is that it could rupture from
the surface down dip to the aseismic front (the latter is shown
in Figure 5-7). The magnitude (Mw, moment magnitude) would
be 9.5 and the closest distances to the Watana and Devi 1 Canyon
sites would be 40 miles (64 km) and 57 miles (91 km), respec-
tively. However, such an assumption would be contrary to the
expectation (based on the considerations outlined in Section 5)
7 - 6
Woodward·CHyde Consui'!l:ants
that great interplate thrust events should not rupture farther
down dip than 22 miles (35 km). If rupture of the deeper segment
adjacent to the aseismic front were to occur, it would most
likely occur independently of rupture on the shallow segment.
and it would be unlikely to have a magnitude greater than (Ms)
7.5, as occurred during the 1978 Miyagi-oki, Japan, earthquake
(Seno and others, 1980).
Intraplate Region
The MCE for the intraplate region of the Benioff zone is esti-
mated to be magnitude (Ms) 7-1/2. This magnitude is based on
an estimated maximum possible fault rupture area within the
intraplate region and on historical seismicity (Section 5). The
estimated maximum fault rupture dimensions of 9 miles (15 km) by
62 miles (100 km) yield a magnitude (Mw, Moment magnitude) of
7.2. A survey of the magnitudes of past intraplate earthquakes
worldwide in the depth range 25 to 45 miles (40 to 70 km) reveals
no events greater than magnitude (Ms) 7.6 (Section 5). The
closest distance of the intraplate region of the Benioff zone
to the Watana and Devi 1 Canyon sites is 31 miles (50 km) and
38 miles (61 km), respectively.
7.3 -Talkeetna Terrain Sources
,i\::; discussed in Section 4.4 and at the beginning of Section 7, the 13
·f1;~atures in the Talkeetna Terrain near the Watana and Devil Canyon sites
have been judged either not to be faults or to be faults without recent
None of these features are considered to be seismic
sources; thus, it is inappropriate to assign MCEs to them.
7 - 7
Woodward-Clyde Consultants
The absence of recognizable faults with recent displacement in the
Talkeetna Terrain sets an upper 1 imit on the magnitude of earthquakes
that could occur on a fault (near one of the sites) that might not
have been detected by our geological investigation. Consequently, an
estimate was made of the size of earthquake that might have occurred
without leaving detectable geologic evidence. This earthquake was
designated the "detection level earthquake." The detection level
earthquake for the Talkeetna Terrain is estimated to be magnitude
(Ms) 6 (Section 4.2.4). The closest distance of this earthquake to
either of the sites is estimated to be <6 miles (<10 km), as shown in
Tab 1 e 7-1.
7.4 -Effect of Reservoir-Induced Seismicity
The hydrologic effects of the proposed reservoirs are postulated to
influence an elliptically shaped area that extends up to 19 miles
(30 km) about the center (i.e., with a diameter of 37 miles [60 km]) of
the proposed Devi 1 Canyon-Watana reservoir, as discussed in Section 6.
However, the reservoir and RIS will not affect consideration of MCEs
along the faults outside the hydrologic regime of the reservoir,
including the Castle Mountain and Denali faults and the Benioff zone.
Within the hydrologic regime of the reservoir, the influence of a
reservoir is believed to be limited to that of a triggering mechanism
(as discussed in Section 6). Thus, the reservoirs are not expected
to cause an earthquake larger than that which could occur "naturally."
Moderate to large RIS events tend to have occurred where faults with
recent displacement lie within the hydrologic regime of the reservoir.
No faults with recent displacement have been observed within the
7 - 8
regime of the proposed reservoir. Consequently, the effect
expected to be limited to the detection level earthquake,
'!ich is discussed in Section 4.2.4. This effect is expected to be
latively small, as discussed in Section 8.2.3.
7 - 9
AXIMUM CREDIBLE EART SUMMARY AND SEISMIC SOURCE DATA
Source MCE Closest Ap~roach to
De vi 1 C anyon2 Watana2
(Ms) 1 miles(km) miles(km) Type Strike
Castle Mountain 7-1/2 71 ( ll5) 65 ( 105) Strike-N60' to 62" E
Fault s 1 i p
Fault 8 40 (64) 43 (70) Strike-N59" to 63' E
s 1 i p
Benioff Zone 8-1/2 57 ( 91)
(Interplate)
40 (64) N/A N45' E
Benioff Zone 7-1/2 38 ( 61) 31 (50) N/A N/ A
( Intrap 1 ate)
6 <6 (<10) <6 ( <10) N/A N/ A
1. Analysis or data are discussed in Sect ion 7.
2. Analysis or data are discussed in Sect ion 4.
3. Analysis or data are discussed i.n Section 5.
4. Displacement is from relationships in Slemmons (1977b).
Estimated average return period of the MCE that may occur on the fault.
Calculated by dividing fault length by MCE rupture length.
Slope of earthquake recurrence curves that are shown in Figure 5-9.
Length Width
Dip miles(km) miles (km)
90' 295 (475) 12 (20) 9
90' 670 ( 1080) 12 ( 20) 9
7' NW 434 (700) 161 ( 260) 3
N/ A N/ A14 N/ A14
N/A N/A N/A
Calculated by multiplying the fault recurrence by the number of MCE rupture lengths (e.g., for the Denali
fault: 290 years x 3.8 = 1,100 years).
Slip rate
in. (cm)/yr
0.2 (0.5)
0.4 (1.0)
N/ A
N/A
N/A
The fault, and fault rupture during an MCE, are assumed to extend to the base of the crust, i.e., 12 miles (20 km).
Estimate calculated using the procedures described in Appendix A.7.
These value~ are consistent with the b-slope that is used for the Talkeetna Terrain and are compatible with
the b-slope for other major strike-slip faults.
Davies and others (1981).
These b-slopes are discussed in Section 5.4 and shown in Figure 5-9.
The intraplate earthquake was assumed to occur on a fault anywhere within a 10,425 square-mile (27,000-km2) section
of the intraplate region of the Benioff zone.
Woodward-Clyde Consultants, 1978.
Obtained in the b-slope shown in Figure 5-9.
MCE Ru~ture Fault Number of Fault Segment
Length1 Width Displacement 4 RecurrenceS MCE Rupture b-RecurrenceS
miles(km) miles (km) feet (m) years Lengths6 Slope3 ,7 years
55 (89) 12 (20) 9 7.5 ( 2 .6) 235 10 5.3 0.85ll 1,245
178 ( 287) 12 (20) 9 21.6 (6.6) 290 10 3.8 0.85 11 1,100
434 (700) 124 (200)3 N/A 16o 12 N/A 0.85 13 N/A_
62 (100) 12 (20)3 N/A 275 15 N/A 0.68 13 N/A
2 2 N/ A 2 700 16 N/A 0.9 13 N/ A 5.2 (8.4) 5.2 (8.4) ,
TABLE 7-1
8 -GROUND MOTIONS
8.1 -Introduction
The objective of this study is to develop estimates of the parameters of
ground shaking at the Watana and Devil Canyon sites that may result from
earthquakes in the site region. The ground motion parameters addressed
in this report include peak acceleration, response spectra, and duration
of strong shaking.
Deterministic estimates of ground motions are presented in this section
for maximum credible earthquakes on significant faults with recent
displacement in or adjacent to the site region. These faults are
the Denali fault to the north of the sites and the interplate and
intraplate regions of the Benioff zone beneath the sites. Deterministic
estimates are also presented for an earthquake in the Talkeetna Terrain
(designated the detection level earthquake, as discussed in Section
4.2.4), which, for purposes of these estimates, is assumed to occur
close (within 6 miles [10 km]) to the sites.
A probabilistic analysis of ground motions has also been made for
this study. The purpose of this analysis, referred to in this section
as a seismic exposure analysis, is to assess the probabilities that
values of peak ground acceleration may be exceeded at the sites and to
identify the seismic sources that have a dominant contribution to
the probability of exceedance. Also, the results of this analysis
may be used to select design ground motion levels for appurtenant,
less critical project facilities such as intake towers, powerhouse
structures, and transmission towers.
This section is organized as follows. In Section 8.2, the seismicity
environment of the sites is summarized, including the potential seismic
8 - 1
Woodward· Clyde Consultants
sources, maximum credible earthquakes on these sources, and recurrence
of earthquakes on the sources. In Section 8.3, deterministic estimates
of ground motions at the sites are presented. Section 8.4 describes the
seismic exposure analysis and results. The use of the deterministic and
probabilistic results in formulating criteria for ground-motion design
for project facilities is discussed in Section 8.5.
8. 2 -Se i sm i c i Environment
8.2.1-Potential Sources of Eart akes
As described in Section 4, the Watana and Devi 1 Canyon sites
are located within a tectonic unit designated the Talkeetna
Terrain. Known faults with recent displacement are present along
the boundaries of the Talkeetna Terrain, as shown in Figures 1-1
and 4-1. These faults are considered to be potential seismic
sources for the sites and include: the Castle Mountain fault 9
approximately 71 miles (115 km) south of the Watana site; the
Denali fault, approximately 40 miles (64 km) north of the Watana
site; and the interplate and intraplate regions of the Benioff
zone, which underlie the Talkeetna Terrain. The interplate region
is at a depth of 40 miles (64 km) south of and beneath the Watana
site; the intraplate region dips downward to the northwest and
is about 31 miles (50 km) below the ~~atana site at its closest
approach. In addition to these seismic sources, a detection level
earthquake is considered in this analysis; it is assumed that this
earthquake would occur close to either site, that is, within
approximately 6 miles (10 km) of either site.
For deterministic estimates of ground motions, the Denali fault,
the interplate and intraplate regions of the Benioff zone, and the
detection level earthquake are all considered to be potential
8 - 2
Woodward·Cmyde
sources of earthquakes and ground motions. The Castle Mountain
fault is not specifically addressed, because ground shaking at the
sites due to a MCE on the Castle Mountain fault would be less
intense than ground shaking from MCEs on the other boundary
faults. For the probabilistic studies of ground motions, all of
the potential earthquake sources mentioned above, including the
Castle Mountain fault, were included. Table 8-1 summarizes the
potential earthquake sources addressed in the deterministic and
probabilistic studies and includes the closest distance of each
source to the sites.
8.2.2 -Maximum Credible Earthquakes (MCEs)
As described in Section 7, estimates of MCEs have been made for the
various potential seismic sources. The MCEs assigned to these
potential sources are magnitude (Ms) 7-1/2 for the Castle Mountain
fault, magnitude (Ms) 8 for the Denali fault, magnitude (Ms)
8-1/2 for the interplate region of the Benioff zone, and magnitude
(Ms) 7-1/2 for the intraplate region of the Benioff zone. A
maximum magnitude (Ms) of 6 has been assigned for the detection
level earthquake. MCE magnitudes for each seismic source are
summarized in Table 8-1.
8.2.3-Earthquake Recurrence
The frequency of occurrence of different magnitude earthquakes
is characterized by a Gutenberg-Richter relationship:
log N(M) = a -bM
8 - 3
where N(M) = average annual number of earthquakes greater than
or equal to magnitude M;
a, b = empirical constants.
The above relationship is considered valid for earthquakes up to
the estimated maximum credible magnitude for a source. Table 8-2
summarizes the estimated values of a and b for each source, and the
following paragraphs summarize the basis for these estimates.
Recurrence relationships were determined for the earthquake sources
using historical seismicity data, geologic data, and information
derived from the Susitna microearthquake network operated in 1980.
These data and information are discussed in Section 4.3 (for the
Denali and Castle Mountain faults) and in Section 5.4 (for the
interplate and intraplate regions of the Benioff zone and the
detection level earthquake).
The i nterp late region of the Benioff zone is the portion of the
Benioff zone that ruptured in the 1964 Prince William Sound
earthquake (magnitude [Ms] 8.4). Using the rate of convergence
and the down-slip length of interaction between the Pacific and
North American plates, Davies and others (1981) calculated a repeat
time of about 160 years for a great earthquake at shallow depth
along this portion of the Benioff zone. This recurrence interval
and the size of the interplate region considered (approximately
69,498 square miles [180,000 km2]) established the recurrence
relationship for great earthquakes (Section 5.4). The b-value of
0.85 was selected on the basis of regional historical seismicity.
The recurrence relationship for the intraplate region of the
Benioff zone was derived from the historical data compiled by the
University of Alaska (Agnew, 1980) and from data collected during
8 - 4
operation of the 1981 microearthquake network (Woodward-Clyde
Consultants, 1980b). The data collected by the network showed a
b-slope of 0.68 (Section 5.4). During a 40-year period, Agnew
(1980) observed one magnitude (Ms) 6 earthquake which he attributed
to the deeper Benioff zone in a 6,564 square-mile (17 ,000 km2)
area beneath the sites. The relationship given in Table 8-2
appears to best fit the historical data and data obtained from the
seismic network.
Geologic field studies conducted during the two-year investigation
and reviews of pertinent literature were used to estimate earth-
quake recurrence for the Castle Mountain and Denali faults. A
recurrence interval of 235 years was estimated for a magnitude
(Ms) 7-1/2 earthquake on the Castle Mountain fault, which is
about 295 miles (475 km) long; this estimate is based on observed
slip rates (Section 4.3). A recurrence interval of 290 years
was estimated for a magnitude (Ms) 8 earthquake on the 670-mi le
(1,080-km) length of the Denali fault west of the Totschunda fault
(Figure 4-1). A b-value of 0.85 was chosen for both sources on the
basis of historical seismicity in the region.
The recurrence relationship for the detection level earthquake
source was determined using data from the 1980 microearthquake
network and historical seismicity data. Few historical earthquakes
have occurred in the area considered for the recurrence of the
detection level earthquake, and the locations of the earthquakes
are subject to uncertainty as a result of the sparse station
coverage. Determining a recurrence relationship for the detection
level earthquake is therefore subject to some uncertainty because
of the limited quantity and quality of the historical seismicity
data. Agnew (1980) observed five earthquakes of magnitude (Ms) 2_4
in 16 years in an area of 4,633 sqaure miles (12,000 km2). The
results of the 1980 mi croearthquake network indicate a recurrence
8 -5
rate for magnitude (Ms) 2_4 earthquakes of approximately 0.40
events per year in the 6,564 square mile (17 ,000 km2) mi ere-
earthquake study area. Agnew (1980) did not observe any shallow
earthquakes df magnitude (Ms) 6 or greater in this area. One
event of magnitude (Ms) 6 in the area was located at great depth
and is not considered to contribute to the shallow seismicity. A
b-value of 0.9 was chosen for the detection level earthquake
because it is the value considered to be most consistent with the
historical seismicity data.
The influence of reservoir-induced seismicity (RIS) on the earth-
quake recurrence interval for the detection level earthquake
was also evaluated and incorporated into the seismic exposure
analysis. Section 6 describes the influence of RIS on earthquake
recurrence. The added effect of reservoir-induced seismicity on
the estimate of seismicity was characterized by a Gutenberg-Richter
relationship as summarized in Table 8-2.
8.3 -Deterministic Estimates of Earthquake Ground Motions
Information on maximum credible earthquake magnitudes and closest
distances of the faults from the sites (summarized in Table 8-1) was
used to estimate levels of earthquake ground motions at the sites. The
relationships employed for these estimates and the resulting ground-
motion characteristics are described in the following paragraphs.
8.3.1 -Attenuation of Earth uake Ground Motion
Attenuation relationships were selected to describe the variation
of peak ground acce 1 erat ion and response spectra 1 acce 1 erat ions
at the ground surface at the sites in relation to earthquake
magnitude and distance of earthquakes from the site. These
8 - 6
Woodward-Clyde Consultants
attenuation relationships were selected on the basis of analyses
of ground motions recorded during previous earthquakes, using
recordings selected to be appropriate for conditions at the sites,
i.e., rock at or very near the ground surface. The published work
of Schnabel and Seed (1973), Seed, Muraka, and others (1976), Seed,
Ugas, and Lysmar (1976), Woodward-Clyde Consultants (1978), Idriss
(1978), Crouse and Turner (1980), Seed (1980) and the results of
ground-motion studies at Woodward-Clyde Consultants were considered
in the selection of the attenuation relationships.
Woodward-Clyde Consultants (1978), Idriss (1978), and Crouse and
Turner (1980) indicate that ground motions from Benioff zone
(subduction zone) earthquakes may attenuate differently than ground
motions from shallow focus crustal earthquakes. To account for the
possible differences, two sets of attenuation relationships were
selected:
0
0
A relationship for earthquakes occurring on the Benioff zone
beneath the sites. This attenuation relationship was based
primarily on analysis of recordings from South America and
Japan for subduction zone earthquakes.
A relationship for crustal earthquakes occurring on the
Castle Mountain fault and Denali fault, and the detection
level earthquake. The primary basis for this attenuation
relationship was recordings from locations in California and
other parts of the western United States.
The mean (average) attenuation relationships used in this study
for peak ground acceleration are illustrated in Figure 8-1 for
Benioff zone earthquakes and in Figure 8-2 for shallow focus
crustal earthquakes.
8 -7
The selected attenuation relationships were compared with limited
data available from Alaska. This comparison is presented in
Figure 8-3 and indicates reasonably good agreement between the
Alaskan data and the attenuation curves used in this study.
8.3.2 -Estimates of Peak Ground Acceleration and Response
ectra at the Dam Sites
Using the attenuation relationships (Benioff zone or shallow focus)
discussed above, mean peak ground accelerations at each site were
estimated to be:
Earthquake
Source
Denali fault
Benioff zone
Mean Peak Ground Acceleration (g)
Watana Site Devil Canyon Site
0.2 0.2
0.35 0.3
(Interplate region)
Detection Level 0.5 0.5
Earthquake
It was found that the site ground motions from the Benioff zone are
governed by the interplate region of the zone; therefore, the
estimates for the intraplate region are not presented.
The response spectra of ground motions at the sites were also
estimated for MCEs on each of these seismic sources. The resulting
mean acceleration response spectra (damping ratios of 0.05 and
0.10) for the Denali fault and the interplate region of the Benioff
zone earthquakes are illustrated in Figure 8-4 for the Watana site
and in Figure 8-5 for the Devil Canyon site. The mean response
spectra at either site for a detection level earthquake is shown in
Figure 8-6.
8 - 8
8.3.3 -Estimates of the Duration of Strong Ground Shaking
at the Dam Sites
The duration of strong ground shaking (significant duration)
was estimated primarily on the basis of results presented by Dobry
and others (1978). In that study, significant duration is defined
as the time during which from 5 to 95 percent of the energy of
an accelerogram is developed. The significant duration may be
estimated from the following table:
Earthquake Magnitude
6
7
8
8-1/2
Significant Duration
(seconds)
6
15
35
45
8.4 -Assessment of Seismic Exposure
8.4.1 -Methodology
Estimates of the probability of exceeding various levels of peak
ground acceleration at the sites were made using the approach
illustrated in Figure 8-7. As indicated in that figure, the
probability analysis requires the characterization of certain input
parameters. Specifically, these include:
identification and geometry of seismicity sources;
seismic activity (recurrence and maximum magnitude) of each
source;
relationship between rupture area and earthquake magnitude;
8 - 9
0 ground motion attenuation relationships; and
0 anticipated design time period of interest.
With these inputs, the exposure analysis was conducted to calculate
the mean number of occurrences by which a given level of ground
motion would be exceeded at each of the sites during the time
period of interest, by combining the contributions of different
magnitude earthquakes occurring on the various sources at different
distances from the sites. The calculations were made using the
computer program called PROGRAM SEISMIC EXPOSURE developed by
Woodward-Clyde Consultants (in press).
The resulting mean number of occurrences that would exceed a given
level of ground motion at a site within the time period of interest
may then be used to estimate the probability of that level being
exceeded at least once during that time period of interest. The
analysis also provides an indication of the relative importance
of an individual source based on its contribution to the total
exposure of each site.
8.4.2 -Assessment of Inputs for Analysis
The earthquake sources considered in the analysis and the charac-
terization of their seismic activity are sunmarized in Tables 8-1
and 8-2. The Castle Mountain and Denali faults were modeled as
vertical planes extending in depth from 0 miles (0 km) (ground
surface) to 12 miles (20 km). The Benioff zone was modeled as
dipping planes at depth, representing the interplate and intraplate
regions. The fault on which the detection level earthquake might
occur was modeled as a series of vertical planes extending to
a depth of 12 miles (20 km), simulating the possible location of a
detection level earthquake at any location within the area of the
Talkeetna Terrain.
8 -10
In the analysis, it is assumed that an earthquake can occur with
equal likelihood at any location on the planar surfaces defined for
a given earthquake source. Because the attenuation relationships
are defined in terms of closest distance from the fault rupture
surface to the site, it is necessary to characterize the dimensions
of fault rupture for given earthquake magnitudes. The relation-
ships shown in Figure 8-8 between magnitude and rupture area,
length, and width were used to characterize the dimensions of fault
rutpure. A further constraint is necessary for the cases in which
the fault width is limited by the fault geometry (e.g., to a width
[depth] of 12 miles [20 km] for shallow crustal faults). For these
cases, the rupture width was limited to the fault width, and the
rupture length was selected to provide the total rupture area given
in Figure 8-8.
The peak acceleration attenuation relationships (mean values)
used in the analysis are shown in Figures 8-1 and 8-2. For a
probabilistic evaluation, it is important to include the uncer-
tainties in the predicted acceleration values for any given
earthquake magnitude and distance. A random error term was used
in the analysis to represent that uncertainty as a statistical
distribution about the median values. A log normal distribution
was assumed and the standard error term taken to be s = 0.40
for the shallow focus relationship and s = 0.60 for the Benioff
zone relationship. The median (50th percentile), mean, and
median-plus-standard-deviation (84th percentile) values of peak
acceleration are related as follows:
s2/2 amean = amedian · e
as4th percentile= amedian · es
8 -11
The methodology also provides for constraining the probability
distribution of peak acceleration so that unrealistically high
values of peak acceleration are not included in calculating
probabilities of exceedance. An upper bound on peak acceleration
was specified to be three standard deviations on the basis of the
trends and bounds suggested by empirical data.
For this analysis, the design time period of interest was assumed
to be 100 years.
8.4 .3 -Results
Probabilities of exceedance were calculated for various levels of
peak ground acceleration at the Watana site. The calculations
were not repeated for the Devi 1 Canyon site; however, the results
would not be significantly different between the two sites because
of their similar relative proximity to the seismic sources.
(The probabi 1 it i es of exceedance wou 1 d be s 1 i ght ly lower at the
De vi 1 Canyon site because this site is somewhat further from the
interplate region of the Benioff zone than is the Watana site).
A plot of the probability of exceedance versus peak acceleration
of the Watana site is shown in Figure 8-9. For probability of
exceedance leve 1 s of 50%, 30%, 10%, 5%, and 1% in 100 years, the
corresponding peak ground accelerations are the following:
Probability of
Exceedance
50%
30%
10%
5%
1%
Peak Ground Acceleration
at Watana Site (g)
8 -12
0.28
0.32
0.41
0.48
0.64
The re 1 at i ve cont ri but ions of each seismic source to the proba-
bilities of exceedance were also examined. The interplate region
of the Benioff zone was found to contribute about 80 percent of the
probabilities of exceedance. The reason for the dominance of this
source on the results is primarily the projected high level of
activity on the source, which is reflected in the recurrence
relationships (Table 8-2) used in the analysis. The high maximum
magnitude (Ms 8-1/2) for the interplate region of the Benioff
zone and the higher acceleration attenuation curves used for
the Benioff zone as compared to those used for shallow sources
(Figures 8-1 and 8-2) are also reasons for the dominant influence
of the interplate region of the Benioff zone on the results of the
seismic exposure analysis.
Most of the rest of the contribution to the probabilities of
exceed a nee comes from the i nt rap 1 ate region of the Benioff zone.
The Cast 1 e Mountain fault, the Dena 1 i fault, and the detection
level earthquake contribute only slightly to the prob~bilities of
exceedance. The contributions of the Castle Mountain fault and the
Denali fault are small primarily because of the relatively large
distances of these faults from the site. The contributions of the
detection level earthquake are small primarily because of the
projected low level of activity for the Talkeetna Terrain.
8.5 -Use of Results of Ground Motion Studies in Selecting Design
Ground-Motion Criteria
"!!'he results of the deterministic and probabi listie estimates of earth-
quake ground motions may be used in selecting ground-motion levels to be
used for design of the dams and other project facilities.
8 -13
critical facilities, such as dams, has often been based on
estimates of ground motion for MCEs. However, it is also
irable to take into account the probability of exceeding ground-
in selecting design levels. The results of this study
the most likely source of strong ground shaking at the
·!ite is the interplate region of the Benioff zone. The other sources
much less likely to cause the same levels of ground shaking at the
P'~:~~>sible design ground-motion criteria for the proposed dams have been
f~Jtmulated for a MCE occurring on the interplate region of the Benioff
For the Watana site, the estimated mean response spectrum for
th·js earthquake is plotted in Figure 8-4. For a critical facility, the
dif~:::;ign is often made for ground motion at an 84th percentile level
r.:rther than at a mean level. In Figure 8-10, both the mean and the 84th
· pi.~~~·centi le response spectrum are shown for the Watana site for a MCE on
the interplate region of the Benioff zone.
EH~:,cause the seismic analysis of Watana dam will be made using an
~~~~:c;eleration time history rather than a response spectrum, a possible
di;:~·~~ign acceleration time history was developed. A plot of this accel-
eration time history is shown in Figure 8-11, and the response spectrum
'i:If the time history is shown in Figure 8-10 superimposed on the smooth
::"·,;;:·:~ponse spectra. As can be seen in Figure 8-10, the response spectrum
the time history lies between the mean to somewhat above the 84th
p,;;;;tcentile smooth spectra. It is anticipated that the fundamental
]:H~~i!"i od of Watana dam will be in the range of about 1 to 2 seconds.
fii•r.~ure 8-10 shows that the response spectrum of the time history is
e'I,Dse to the 84th percentile smooth spectrum in this period range.
Also, the significant duration of the time history is about 45 seconds,
!1-,!ii·i,iich is consistent with the duration expected for a magnitude 8-1/2
i:~:a.rt hq u ak e.
8 -14
For the Devil Canyon site, mean and 84th percentile response spectra for
,i:li MCE on the interplate region of the Benioff zone are presented in
Figure 8-12. Because the seismic analysis of the proposed Devil Canyon
d,c!!l'!l may be made using a rupture spectrum rather than an acceleration
time history, a time history has not been developed for the Devil Canyon
1te. If a time history should be needed, it could readily be developed
b'Y modifying the time history shown in Figure 8-ll. The modifications
'i;~!·i:J!iJld involve scaling the time history shown in Figure 8-ll and modify-
'!lng its frequency ch aracteri st i cs to pro vi de response spectral values
to the 84th percentile level at the fundamental period of the
n~~$ign ground-motion criteria for a maximum credible detection level
·~~~H"thquake could be formulated in a manner similar to that described
,:~!lmve for a Benioff zone MCE. However, it is also appropriate to
~r::m!:sider the relatively low likelihood of detection level earthquakes in
~:::tHmparison to Benioff zone earthquakes in developing design criteria for
,!j: ~:::letection level earthquake.
F{H' non-critical facilities, such as a powerhouse or transmission
t~;~!~~~er, the results of the probabilistic studies can be used to aid
'iilf~l selecting design ground-motion criteria. Selection of the design
teria may include consideration of the acceptable levels of proba-
!;;;'1'Jities of exceedance, economics, and acceptable risks of damage
t~:! these facilities.
8 -15
TABLE 8-1
SUMMARY OF EARTHQUAKE SOURCES CONSIDERED IN GROUND-MOTION STUDIES
Closest Approach to
Proposed Dam Sites {km)
MCE
Earthquake Source (Ms) Devi 1 Canyon Watana
miles/(km) miles/ (km)
Castle Mountain fault 7-1/2 71 ( 115) 65 ( 105)
Denali fault 8 40 (64) 43 (70)
Benioff zone (Interplate) 8-1/2 57 ( 91) 40 (64)
Benioff zone (Intraplate) 7-1/2 38 ( 61) 31 (50)
Detection Level 6 <6 (<10) <6 (<10)
Earthquake
TABLE 8-2
SUMMARY OF EARTHQUAKE RECURRENCE ASSESSMENTS
Constants in
Gutenberg-Richter Relationship1
Earthquake Source
Castle Mountain fault
Den a 1 i f au 1 t
Benioff zone (Interplate)
Benioff zone (Intraplate)
Detection Level Earthquake
Natura 1 s ei smi city
only
Incremental seismicity
resulting from RIS
Notes:
a b
5.30 0.85
5.33 0.85
4.75 0.85
3.23 0.68
3.97 0.9
0.9
7-1/2
8
8-1/2
7-1/2
6
6
1. Gutenberg-Richter Relationship: log N(M) = a -bM. For the
constants summarized herein, N(M) is the number of earthquakes
per 100 years per 62 miles (100 km) of fault length (in the case
of the Castle Mountain and Denali faults) and per 386 square miles
(1,000 km2) (in the case of the Benioff zone sources and the
detection level earthquake source).
2. Reservoir induced seismicity (RIS) is considered for 10 years per
386 square miles (1,000 km2).
Cl
0.. co
c
0 ·;:;
co ...
Cl)
(l)
u u
<(
co .....,
c
0
N ...
0
I
.::.!. co
Cl)
0..
1.0.-------------------------------------------------~
0.8
0.4
0.2
0~--------~----~--~--~~~~~~--------~----~
10 30 100 300
Distance from Rupture (km)
MEAN ATTENUATION RELATIONSHIPS FOR DEEP
FOCUS (BENIOFF ZONE) EARTHQUAKES
LYDE CONSUL FIGURE 8-1
10 30 100 300
Distance from Rupture (km)
NOTE
1. Curves are applicable only within the distance
range shown; at distances less than 3 km, peak
accelerations are constant and equal to the
values at 3 km.
0-CLYDE CONSULTANTS 41410A February 1982
MEAN ATTENUATION RELATIONSHIPS FOR
SHALLOW FOCUS EARTHQUAKES
FIGURE 8-2
0.3~-----------------------------------------------------,
-
0.1 \
ca
.EJ -\ c -
0 ·;:; ~
~
Ill ~ a:;
(.)
(.) -<(
""iO ..
+-' -c
0
N ·;;:
0 ,...._
I
..:,(.
"' Ill
CL 0.03 --,-
'~ ' ' ' ....
,~}.
....
'
0
'
\
0
' ' ....
' &. ' ....
&.
I I I I I I I I I I 0.01 ~--------~----~~--J---~-L~--~~----------~----_j
10 30 100
Distance from Rupture (km)
LEGEND
Attenuation Relationships for:
Benioff zone earthquakes (shown in Figure 8-1)
--=-Crustal earthquakes (shown in Figure 8-2)
Alaskan Earthquakes
Date Magnitude Focal Depth (km)
() 21 June 1967 5.4 14
0 11 March 1970 6.4 29
6 1 May 1971 7.1 43
() 3D July 1972 6.5 25
& 10 Nov 1974 5.2 68
21 Feb 1976 4.0 58
300
COMPARISON OF SELECTED
ATTENUATION RELATIONSHIPS WITH
DATA FROM ALASKA
D-CLYDE CONSULTANTS41410A February 1982 FIGURE 8-3
0.03 0.1
LEGEND
Damping ratio= 0.05
----Damping ratio= 0.10
Denali fault
0.3
Period (sec)
3 10
MEAN RESPONSE SPECTRA FOR MAXIMUM CREDIBLE
EARTHQUAKES ON THE BENIOFF ZONE AND
DENALI FAULT-WATANA SITE
RD-CLYDE CONSULTANTS 41410A February 1982 FIGURE 8-4
'/
~
/
/
/
I
----..... -·-/ .....
/ '
/ ' / '
0 ~----~--~--~~~~~~----~--~~~~~~~~------~--~~~~~~~
O.Q1 0.03 0.1
LEGEND
Damping ratio= 0.05
----Dampingratio=0.10
0.3
Period (sec)
3
MEAN RESPONSE SPECTRA FOR MAXIMUM CREDIBLE
EARTHQUAKES ON THE BENIOFF ZONE AND
DENALI FAULT-DEVIL CANYON SITE
10
L YDE CONSULTANTS 41410A February 1982 FIGURE 8-5
2.0r---------------------------------.
0
0.01
LEGEND
NOTE
/,----...... ,
/ '
/ '
/ ' v ' ~ '
Damping ratio= 0.05
Damping ratio = 0.10
Period (sec)
\
' ' ' ', ' ............ -.... ----
1. Spectrum applicable to both sites.
MEAN RESPONSE SPECTRA FOR MAXIMUM
CREDIBLE DETECTION LEVEL EARTHQUAKE
10
D·CLYDE CONSULTANTS 41410A February 1982 FIGURE 8-6
INPUTS ANALYSIS
Obtain probability distribution of
ground motion parameter by
combining probabilities of
exceeding different levels of the
parameter due to occurence of
different magnitude earthquakes
on various sources.
RESULTS
Calculate probabilities
of exceeding specified
levels of ground motion
parameter
SCHEMATIC REPRESENTATION OF SEISMIC
EXPOSURE ANALYSIS APPROACH
D-CLYDE CONSULTANTS 41410A February 1982 FIGURE 8-7
100 200 300 400
Rupture Length (km)
Magnitude Area 1 Length 1 Down-<iip 1
Width
(Ms) (sq. km) (km) (km)
5.0 7 2.7 2.7
5.5 22 4.7 4.7
6.0 71 8.4 8.4
6.5 224 15 15
7.0 708 27 27
7.5 2240 47 47
8.0 7080 90 80
8.5 22,400 180 125
NOTE
1. Fault rupture dimensions are from Wyss (1979) and
Woodward-Clyde Consultants ( 1 978).
500 600 700 800
RELATIONSHIPS BETWEEN MAGNITUDE AND
FAULT RUPTURE DIMENSIONS
DE CONSULTANTS 41410A February 1982 FIGURE 8-8
Q)
u c ro
'"0
Q)
Q)
u
X
UJ ,.._
0.3
0.1
0 0.03
>-.....
.0 ro
.0
0 .....
0....
0.01
0.003
0.001
0 0.2 0.3 0.4 0.5 0.6 0.7
Peak Acceleration (g)
PROBABILITY OF EXCEEDANCE VERSUS PEAK GROUND
ACCELERATION AT THE WATANA SITE
0-CLYDE CONSULTANTS 41410A February 1982 FIGURE 8-9
Damping Ratio= 0.05
II &
I\ tv Response spectrum of
I ~I acceleration time
1
1 history shown in
1 Figure 8-1 0
I
I
I
I
0.4
0.2
0~--~--~~~~~~----~--~~~~-WU-----U-~~~~~~
0.01 0.03 0.1 0.3
Period (sec)
1 3 10
MEAN AND 84th PERCENT! LE RESPONSE SPECTRA
FOR A MAXIMUM CREDIBLE EARTHQUAKE
ON THE BENIOFF ZONE-WATANA SITE
D-CLYDE CONSULTANTS 41410A February 1982 FIGURE 8-10
10 20 30 40 50
Time (sec)
LYDE CONSULTANTS 41410A February 1982
60 70 80 90 100
ACCELERATION TIME HISTORY FOR A MAXIMUM
CREDIBLE EARTHQUAKE ON THE BENIOFF ZONE
FIGURE 8-U
0.03 0.1
LEGEND
Damping ratio = 0.05
---Damping ratio = 0.10
D·CLYDE CONSULTANTS 41410A February 1982
0.3
Period (sec)
3
MEAN AND 84th PERCENTILE RESPONSE SPECTRA
FOR A MAXIMUM CREDIBLE EARTHQUAKE ON
THE BENIOFF ZONE-DEVIL CANYON SITE
FIGURE 8-12
9 -TRANSMISSION LINE AND ACCESS ROUTE SUSCEPTIBILITY TO SEISMICALLY
INDUCED FAILURE
9.1 -Introduction
The objective of this part of the investigation was to provide input
regarding the behavior of those areas along the transmission line and
major access road routes that appeared to be underlain by soils that are
potentially susceptible to seismically induced ground failure such as
liquefaction or landsliding. The approach used to meet this objective
was to interpret large scale and small scale aerial photographs along
the rights-of -way and to review surfi cia 1 geology mapping conducted by
Woodward-Clyde Consultants during the 1981 field study (Section 3) and
by R & M Associates, Inc. (1981).
The scope of this part of the investigation involved the interpretation
of aerial photographs within 5 miles (8 km) of the three proposed
routes shown in Figure 9-1. The photography used for the investigation
was: U-2 false color near-infrared photography flown by the National
Aeronautics and Space Administration (NASA) in 1977 at a scale of
1:125,000; color photography flown for Acres in 1980 at a scale of
1:24,000; low-sun-angle color near-infrared photography flown for
Woodward-Clyde Consultants in 1981 at a scale of 1 :24,000; and black and
white photography flown by the U.S. Army in 1949, primarily at a
scale of 1:40,000.
Criteria were established to guide the identification of two types
of potentlally hazardous areas. The two types of hazards are seis-
mically triggered landslides and seismically induced liquefaction. The
hazards were assumed to have the potential to occur in the site region
during moderate to large earthquakes. No attempt was made to refine the
analysis for different magnitude events occurring at various distances
from the potential hazard areas.
9 - 1
Woodward-CWyde
Areas identified as having the potential for seismically induced land-
slides included:
a) areas with previous landslides and slumps;
b) river and stream valleys with steep slopes and overhanging promon-
tories; and
c) areas with ground fracturing on slopes above river and stream
valleys.
Criteria for areas of potential liquefaction ~Jere derived by reviewing
reports of the areas that liquefied during the 1964 Prince William Sound
earthquake (magnitude [Ms] 8.4). This review included reports by
Foster and Karl strom (1967), McCulloch and Bonilla, (1970), Plafker
(1969), and Tysdal (1976). Liquefaction during the 1964 earthquake was
concentrated in areas of unconsolidated deposits, particularly where
the ground was saturated with water. Typically, these were areas
underlain by glacial till, glaciolacustrine and glaciofluvial deposits,
pro-glacial lake sediments, marine sediments, alluvium, loess, and
clay.
Areas identified as having the potential for seismically induced lique-
faction in and along the transmission and access routes included:
a) floodplain deposits along the margins of rivers and streams,
and
b) areas underlain by glacial deposits, particularly kettles and
deposits with standing or near-surface water.
Section 9.2 discusses each of the areas identified as potentially being
susceptible to seismically induced landslides and/or liquefaction. It
9 - 2
Woodward·C~yde
should be emphasized that these locations have not been field checked.
These locations and the selected right-of-way should be field checked
prior to final design and construction.
9.2 -Areas of Potential Susceptibililty
This section presents the results of the study for each of the three
transmission line and access route alternates. Each of the areas of
potential susceptibility have been given a location number (Tl-1, Tl-2,
etc.) and are shown in Figure 9-1.
9.2.1-Alternate Route 1
Location T1-1 is located on a ridge 1-1/4 miles (2 km) northeast of
Gold Creek (Figure 9-1). It consists of several depressions which
may be kettles. The kettles appear to contain unconsolidated
sediments and excessive water. This location may be an area of
potential liquefaction under seismic loading conditions. The
location of this area is indicated by the ground pattern and the
brownish coloration of the vegetation within the depressions. At
the point the Susitna bends eastward, upstream from Gold Creek,
the floodplain (outwash material) contains kettles and pits
(Figure 9-1; Location Tl-2); this area may also contain areas of
potential liquefaction.
Where Alternate Route 1 enters the stream valley southeast of
Location T1-1, a site of possible landsliding is encountered
(Location T1-3). At the top of the ridge east of the stream
valley, Alternate Route 1 follows a colluvial deposit before
reaching the Devil Canyon site (Location Tl-4). The colluvi urn
appears to contain numerous depressions where unconsolidated
material is present and the water content is higher, as indicated
9 - 3
by the vegetation. These locations, T1-1, T2-2, and T1-4, have the
potential for liquefaction. On the south side of the Susitna
River, adjacent to the Devil Canyon site, an old river channel
(associated with Feature KD5-43) on the hillside contains several
similar depressions which may be susceptible to liquefaction
(Location Tl-5).
North of the Devil Canyon site, Alternate Route 1 trends northeast
between several lakes in the High Lake area. The drainages
from these lakes and the depressions in several low-lying areas
may be areas of potential liquefaction and should be evaluated
(Location Tl-6). A bench of glacial till a long Devil Creek at
Location T1-7 is another area of potential liquefaction.
East of Devil Creek, numerous kettles and depressions at Locations
Tl-8 and Tl-9 in till appear to contain saturated sediments that
may present liquefaction problems. The route then crosses areas of
extensive outwash that contain numerous depressions (Locations
Tl-10, 12, 13) that may be prone to liquefaction. North of the
Watana site, the route crosses several stream valleys which have
landslide scars at Locations Tl-ll and Tl-14. Near Tsusena Creek
(Location Tl-15) several landslides are present. South of Tsusena
Creek, an area of extensive small lakes and drainage channels is
present at Location Tl-16. This area also has the potential for
liquefaction. It is recommended that the route be carefully
located to avoid the numerous depressions at Location Tl-17 (a
potential area of liquefaction) and the slide areas evident along
Tsusena Creek.
9.2.2-Alternate Route 2
Alternate Route 2 from Devil Canyon to the Watana site starts out
at the intersection of Alternate Route 1, adjacent to a small lake
9 - 4
0.5 mile (0.8 km) south of the Devil Canyon site (Figure 9-1). It
trends eastward on bedrock, but several stream valleys are crossed
at Location T2-1. At each crossing, slide areas are present.
Where the route crosses the bedrock exposures, several sites of
glacial drift deposits are present (Location T2-2). Depressions
containing what appear to be unconsolidated sediments that would
be potentially susceptible to liquefaction are present at this
location.
Southward of Location T2-2, the route is located on bedrock
before turning eastward 6 miles (10 km) west of Stephan Lake.
Near this bend, the route goes from bedrock to glacial deposits
(Location T2-3). These deposits contain numerous depressions and
hummocky areas and are underlain by relatively thick deposits
of unconsolidated sediments. The vegetation indicates that the
sediments are saturated in many of the sites. Near the west side
of Stephan Lake, an outwash deposit is crossed and several boggy
depressions are present (T2-4). Both of these locations (T2-3 and
T2-4) are areas of potential liquefaction.
At the northeast end of Stephan Lake, the route passes several
small lakes with intersecting drainages (Location T2-5). East of
Location T2-5, the route enters an area of morainal sediments
(Location T2-6). Several streams with evidence of slides are
present near Location T2-6. During the 1981 field season, ground
thawing and resultant slumping of surficial units occurred near
Location T2-6. Locations T2-5 and T2-6 are areas that are poten-
tially susceptible to liquefaction.
There are many slide areas where the route crosses Fog Creek; some
are in ablation till (Location T2-7). This creek valley appears to
be potentially unstable at numerous locations. Near Fog Lakes,
there are numerous kettles in the glacial till, with possible
9 - 5
permafrost underlying the area (Location T2-8). Because of the
potential susceptibility to liquefaction, the route location
should be carefully planned through this area. Near the Susitna
River and the Watana site, several slide areas are also present
(Location T2-9).
9.2.3-Alternate Route 3
Alternate Route 3 trends northward from the Watana site to the
Denali Highway (Figure 9-1). North of the Watana site, a group of
small lakes are present in the glacial deposits at Location T3-l.
This area is potentially susceptible to liquefaction. These lakes
and deposits continue northward to Deadman Creek. Several areas of
outwash are present, and the vegetation pattern indicates sluggish
drainage and potential sites of liquefaction at Location T3-2.
On the high slopes above and west of Deadman Creek, bedrock is
present. The only areas that are considered to be potentially
susceptible to slide hazards are at stream crossings (Location
T3-3). Northwest of Deadman Mountain, the route enters the
Brushkana Creek area (Location T3-4). The entire area is pitted by
kettles and depressions (particularly in Locations T3-5 through
T3-8), is saturated, and has extensive areas of standing water. It
is expected that the area is underlain by unconsolidated sediments
and is potentially susceptible to liquefaction. The slide hazards
in this region appear to be limited, except for small areas at
stream crossings. Several scarps in a ground moraine are crossed
four miles (seven km) south of the Denali Highway at the Seattle
Creek Valley (Location T3-9). The route should be placed to avoid
the areas containing these potential slide sites.
9 - 6
9.3 -Summary
In general, the areas which are expected to be potentially susceptible
to landslides are those where steep slopes are encountered (particularly
where they are underlain by weak or unstable rock and young, unconsoli-
dated sediments), along river and stream drainages where previous
landslides have occurred. An example of this type of area is the
Susitna River banks upstream from the Watana site. It is recommended
that these potential landslide areas, and areas with characteristics
similar to those discussed previously in this section, either be avoided
or examined carefully prior to final route selection for both the
transmission line and the access route.
Areas generally expected to be susceptible to liquefaction are those
underlain by saturated, cohesionless, unconsolidated sediments. These
areas include river and stream floodplains, glacial depressions (such as
kettles), and unconsolidated glacial sediments such as glaciolacustrine,
glaciofluvial, outwash, and ice disintegration deposits. Prominent
examples of these types of areas are the Susitna River floodplain, the
Stephan Lake and Fog Lakes areas, the area between the Watana site and
Deadman Lake, and the Brushkana Creek area. It is recommended that
these potential liquefaction areas, and areas with characteristics
similar to those discussed previously in this section, either be avoided
or examined carefully prior to final route selection for both the
transmission line and the access route.
9 -7
" 1'/ ,•
LEGEND
~ ·>-
' -~
' '
~-
~
'·'\..
... -., /
'
/ /
~ 7
~--. \ "y"
' • 1' }'rrl i .~, -
/ --
l.
'(~-.... , '
Alternate access and transmission line
route and number
Potential slide area and location number
.,,
'..l '
&/ .(1/"'?...t~.c.: .,.. ,. .
.. tJ
!
Potential area of I iquefaction and location
number
WOODWARD-CL YDE CONSULTANTS 41410A February 1982
/
/
/ ·"'\
( ' . . •..
NOTES
'\,
'·1
1. Potential slide and I iquefaction areas have
been interpreted from aerial photography .
Their location and extent should be field
checked prior to design and construction
along the selected right-of-way.
2 . Route locations are from Acres American Inc .
Drawing Sk-5700-C2-1 01 dated August, 1981.
. ·, : --:~ .. /
I
,~ 0
l ,,oe,
V T3 -8,' ,I
T3-7,-
·' ('
. ,.---"' /~-v-c-· -..../ ' '/ . -,..-/
-· ··;
TRANSMISSION LINE AND
ACCESS ROUTES
POTENTIAL HAZARDS MAP
5
/
10 Miles
0 5 10 Kilometers
FIGURE 9-1
Woodward-Clyde ConsuStants
10 -CONCLUSIONS
For the purpose of evaluating Project feasibility, two sets of con-
clusions were drawn from the results of the investigation: 1) the
feasibility conclusions, i.e., those considered important in evaluating
the feasibility of the Project; and 2) the technical conclusions
related to the scientific data collected. The results upon which
these conclusions were based should be reviewed when final design is
considered.
10.1-Feasibility Conclusions
1) The faults with known recent displacement closest to the Project
sites are the Denali and Castle Mountain faults. These faults,
and the Benioff zone associated with the subducting Pacific plate,
are considered to be seismic sources. Maximum credible earthquakes
(MCEs) for the Castle Mountain and Denali faults, and the inter-
plate and intraplate regions of the Benioff zone, have been
estimated as: a magnitude (Ms) 7-1/2 earthquake on the Castle
Mountain fault, 71 miles (115 km) from the Devil Canyon site
and 65 miles (105 km) from the Watana site; a magnitude (Ms) 8
earthquake on the Denali fault, 40 miles (64 km) from the Devil
Canyon site and 43 miles (70 km) from the Watana site; a magnitude
(Ms) 8-1/2 earthquake on the interplate region of the Benioff
zone, 57 miles (91 km) from the Devil Canyon site and 40 miles
(64 km) from the Watana site; a magnitude (Ms) 7-1/2 earthquake
on the intraplate region of the Benioff zone, 38 miles (61 km~ from
the Devil Canyon site and 31 miles (50 km) from the Watana site.
2) Of the 13 significant features, nine were found to be lineaments
and four were found to be faults. No evidence of faults with
10 -1
Woodward-C§yde Consul'itanis
recent displacement (displacement in the past 100,000 years) was
found on features that pass through or adjacent to the Project
sites; therefore, none of the 13 significant features near the
sites are judged to be faults with recent displacement for purposes
of seismic design.
3) The detection level earthquake (an earthquake that theoretically
could occur on an undetected fault with recent displacement) was
judged to be a magnitude (Ms) 6 earthquake that could occur
within 6 miles (10 km) of either site.
4) Estimates of peak acceleration response spectra and duraction of
strong shaking at the sites were made for the Denali fault, the
interplate region of the Benioff zone, and the detection level
earthquake. The results of the probabilistic ground-motion
(seismic exposure) study indicate that the source most likely to
cause ground shaking at the site is the interplate region of the
Benioff zone. Possible design criteria have been formulated for
the Benioff zone earthquake.
10.2-Technical Conclusions
1) The site is located within the Talkeetna Terrain. This tectonic
unit has the following boundaries: the Denali fault to the
north and northeast; the Totschunda fault to the east; the Castle
Mountain fault to the south; a broad zone of deformation and
volcanoes to the west; and the Benioff zone at depth.
2) The northern, eastern, and southern boundaries of the Talkeetna
Terrain are major fault systems along which displacement has
occurred in Quaternary time. The Benioff zone beneath the Talkeet-
na Terrain represents the upper margin of the Pacific plate which
10 - 2
is being subducted beneath the North American plate. The western
boundary does not appear to have brittle deformation occurring
along a major fault.
3) The Talkeetna Terrain appears to be a relatively stable tectonic
unit within the present stress regime. Major strain release occurs
along the fault systems bounding the Terrain. Within the Terrain,
strain release appears to be randomly occurring at depth within the
crust. This strain release is possibly the result of crustal
adjustments resulting from stress within the Terrain caused by the
subduction of the Pacific plate and/or by stress imposed by fault
displacement along the Terrain margin.
4) The only fault system within the site region (within 62 miles
[100 km] of either Project site) which is known to have been
subject to displacement in late Quaternary time (the past 100,000
years) is the Denali fault. This fault is approximately 40 miles
(64 km) north of the sites at its closest approach. The only
other fault near the site region that has been subject to recent
displacement is the Castle Mountain fault which is immediately
south of the site region. This fault has been subject to displace-
ment in late Quaternary time.
5) Thirteen significant features were identified by the 1980 studies
as needing additional investigation. These 13 features were
selected on the basis of their seismic source potential and
potential for surface rupture through either Project site. Four of
these features are in the vicinity of the Watana site and include
the Talkeetna thrust fault, the Susitna feature, the Fins feature,
and the Watana lineament. Nine of the features are in the vicinity
of the Devil Canyon site and include Faults KD5-2 and KC5-5, and
lineaments KD5-3, KD5-9, KD5-12, KD5-42, KD5-43, KD5-44, and
KD5-45.
10 - 3
6) No evidence that the Susitna feature is a through-going fault was
recognized during this study. Aerial reconnaissance, ground
checking, low-sun-angle aerial photograph interpretation, and
trenching produced no evidence of a through-going fault in bedrock
along the lineament, and no evidence of deformation in overlying
surficial units.
7) The Talkeetna thrust fault is a northeast-southwest trending
fault that extends from the town of Denali southwest to the
Talkeetna River. The fault dips to the southeast except near
Talkeetna Hill where the fault is near vertical. On the basis of
aerial reconnaissance, ground checking, low-sun-angle aerial
photograph interpretation, and trenching, the fault is classified
as a fault without recent displacement.
8) The Broxson Gulch thrust fault is aligned with the Talkeetna
thrust fault at Denali and dips in the opposite direction (to the
northwest). Because of this dip reversal, the Broxson Gulch has
been judged to be unrelated to the Talkeetna thrust fault within
the present seismotectonic environment.
9) Seismicity within the Talkeetna Terrain can be clearly delineated
as crustal events occurring at depths to approximately 5 to
12 miles (8 to 20 km) and as Benioff zone events occurring at
greater depths. The depth to the Benioff zone increases from
approximately 25 miles (40 km) in the southeastern part of the site
region to more than 50 miles (80 km) in the northwestern part of
the mi croearthquake study area, and to more than 78 miles (125 km)
in the northwestern site region.
10) The Benioff zone can be divided into two regions, the interplate
region and the intraplate region. The former includes the area
where the Pacific plate is passing beneath, and in contact with,
10 - 4
the North American plate. The largest earthquakes associated with
the Benioff zone occur on or adjacent to the interface between the
plates in this region. The intraplate region dips to the northwest
and is decoupled from the North American plate. This region has
smaller maximum earthquakes than the interplate region.
11) The largest reported historical earthquake within the Talkeetna
Terrain is the magnitude (Ms) 7.3 event of 1943 which occurred
along the western margin of the Terrain approximately 90 and
107 miles (144 and 171 km) west of the Devil Canyon and Watana
sites, respectively. Several lineaments in the epicentral area may
represent faults that could have been the source on which this
earthquake occurred.
12) The largest crustal event recorded within the microearthquake
study area during three months of monitoring in 1980 was magnitude
(ML) 2.8. It occurred 7 miles (11 km) northeast of the Watana
site at a depth of 9 miles (15 km) on 2 July 1980.
13) No association of microearthquake activity with any of the 13
significant features is apparent.
14) The two reservoirs are considered to be one hydrologic entity.
This combined Devil Canyon-Watana reservoir would be among the
deepest and largest in the world. After comparing this reservoir
to similar reservoirs, our interpretation suggests that the mean
likelihood of a reservoir-induced earthquake within the hydrologic
regime of the proposed reservoir is 0.46 (on a scale of 0 to 1)
with a standard deviation of 0.22.
15) Since no faults with recent displacement were found within the
hydrologic regime of the proposed reservoir, the likelihood of a
R IS event of magnitude (Ms) 2_4 is considered to be low. However,
10 - 5
the detection limits for faults with recent displacement in this
region suggest that there is some likelihood that an RIS event up
to magnitude (Ms) 6 could occur. This is the magnitude of the
detection level earthquake.
10 -6
APPENDIX A -METHODS OF STUDY
This appendix describes the methods used for conducting the 1981 field
studies, including Quaternary geology, age dating, field mapping,
trenching, geophysi ca 1 surveys, and 1 ow-sun-angle photography acqui s i-
tion. It also describes the methods used for estimating the recurrence
intervals of
di sp 1 acement.
Appendix A of
maximum credible earthquakes for faults with recent
Methods used for the 1980 studies are summarized in
the Interim Report (Woodward-Clyde Consultants, 1980b).
A.1 -Quaternary Geology
A.1.1-Scope of Studies
This section presents the methodology, procedures, and measuring
techniques used to perform the Quaternary geology tasks during the
1981 study (described in Section 3). Section A.1.2.1 describes the
pre-field subtasks used to identify the type, areal distribution,
and relative age of the Quaternary surfaces. These pre-field
studies included: literature acquisition and review; photogeologic
interpretation; data transfer to topographic base maps; and pre-
paration of a preliminary map of Quaternary surfaces. Field
studies used to supplement, refine, and support the interpreta-
tions derived from the pre-field studies are described in Section
A.1.2.2. The field studies included: aerial reconnaissance mapping
and excavation of test pits; collection of samples for radiocarbon
dating; and collection of relative (weathering) age data. Section
A.l.3 and A.2.1 discuss the field techniques used for radiocarbon
and relative (weathering) age dating of the surficial deposits.
Subsequent to the field studies, radiocarbon age dates were
A - 1
obtained for 11 samples. These dates along with a synthesis of the
field data were used to refine the map of Quaternary surfaces
(Figure 3-2).
Detailed evidence from which the maximum lateral and vertical
limits of four chronologically distinct glaciations were derived
are presented in Section A.l.4. This evidence is discussed for
each of six subregions into which the Quaternary study region
was divided. These six subregions are shown in Figure A-1 and
discussed in Section A.1.4.
A .1.2 -Methods
A.1.2.1-Pre-Field Studies
Published and unpublished literature was reviewed to provide
a regional perspective of the Quaternary glacial history of
south central Alaska. Applicable information was documented and
briefly summarized from both regional and site specific studies
that would help in the identification of glacial deposits in the
Quaternary study region and in understanding their distribution and
age. The results of this review are discussed in Section 3.2. A
copy of the literature and recorded summaries are included in the
project file.
Detailed photogeologic interpretation was used as the primary
method for evaluating the glacial chronology of the Quaternary
study region. Emphasis was placed on this method because the
glaciogenic features that can be of diagnostic value in deciphering
the geographic and elevational limits of glaciations lend them-
selves to identification on stereographic aerial photographs.
Three stereographic aerial photo sets, each providing different
levels of detail or resolution, were used to map the glaciogenic
A - 2
features. The three sets of aerial photographs were: 1)
1:120,000-scale high altitude color near-infrared; 2) U.S. Army
Corps of Engineers 1:40,000-scale black and white; and 3) Acres
1:24,000-scale color. Although they were not available during the
pre-field photo analysis, low-sun-angle 1:24,000-scale color
near-infrared photographs were used during aerial reconnaissance,
field mapping, and field verification of glacial surfaces.
For each set of photographs, photogeologi c mapping was done on
registered mylar overlays. The different mappable units were color
coded or represented by symbolization.
In order to ensure as much objectivity as possible in identifying
features and assessing their elevations, interpretation of aerial
photographs was completed for all photographic sets before any data
were transferred to the topographic base maps. In addition, each
flight line was interpreted independently; data from adjacent
flight lines were compared only during the transfer of data to the
topographic base. Evaluation and analysis of the data began only
after all of the data was transferred to the topographic base. The
detailed maps are not presented here, but they are filed in the
project file.
The photogeologic data were transferred from the photographs to the
U.S. Geological Survey topographic quadrangle maps (1:63,360
scale) using a Bausch and Lomb zoom transfer scope. The transfer
of these data to the maps introduced limitations in the accuracy of
locations and elevations. The sources of accuracy limitations
were: radial distortion at the photograph margins; lack of
distinguishing features for precise registration of photograph and
map; large variation in topographic elevation over small horizontal
distances; and the small inherent distortions introduced by
the transfer scope. The elevations of features, on which the
conclusions and results of this study are based, are accurate to
+200 feet (61 m).
A - 3
Woodward-Clyde Consultants
Evaluation and analysis of the data on the topographic base maps
allowed for construction of a preliminary morphostratigraphic
map showing the maximum extent of each major glacial episode.
The morphostratigraphic units show the elevational limits of
geographical extents of ice-marginal glaciogenic features having
similar morphologic, stratigraphic, or geographic expressions or
relationships. These relationships were interpreted to provide
a preliminary definition of the relative ages of each glacial
episode. Characteristics used to assign a relative age to similar
glaciogenic features or deposits are summarized in Tables 3-1 and
3-3.
A.l.2.2-Field Studies
The field studies were designed to supplement and refine the
interpretations and conclusions derived from the analysis of aerial
photographs. The studies involved helicopter-supported low-level
aerial reconnaissance of morphostrat i graphic unit contacts and
visual verification of glaciomorphologic evidence used to delineate
the morphostratigraphic units. Lithologic characteristics of
surficial deposits in natural exposures were identified during the
aerial reconnaissance. This identification was supplemented, where
needed and possible, by ground checking. Exposures with more
than one lithologic unit and of significance in stratigraphic
interpretation were measured and described to help define the
glacial chronology and history. Descriptions of the lithologic
units included lithology, grain size, color, structure, texture,
other distinguishing characteristics, and unit thickness. Apparent
unit thickness was measured by tape and the slope angle was
measured by Brunton compass. The true thickness of the unit was
calculated from these data.
A -4
During detailed ground study of the exposures, material suitable
for radiocarbon age dating was collected where possible. Relative
age (weathering) dating (referred to here as relative age dating)
studies were also conducted.
A.1.3-Age Dating
A.1.3.1-Radiocarbon Age Dating
To increase the confidence in the ages assigned to various units in
the Quaternary study region, selected samples collected from
surfi cia 1 deposits were radiocarbon age dated. During the aerial
reconnaissance mapping and stratigraphic studies, exposures
were identified that had a high likelihood of containing dateable
material, i.e., lacustrine, deltaic, or glaciofluvial deposits.
From these exposures, 28 samples of carbonaceous wood and plant
material were collected. Of these 28 samples, 11 (at the 11
locations shown in Figure 3-2) were submitted to the Geochron
Laboratory of Cambridge, Massachusetts. to obtain radiocarbon age
dates. The results of the dating are surrmarized in Table 3-2.
The samples were collected in accordance with the guidelines and
theory described in Woodward-Clyde Consultants (1975). To prevent
contamination, samples were collected without hand contact and
were wrapped in aluminum foil. Samples were marked with a locality
designation and sample number (e.g., S34-1, as shown in Table 3-2
and Figure 3-2). Samples were dried and then wrapped in new foil
and placed in marked, self-sealing plastic bags.
A.1.3.2 -Relative Age Dating
The discontinuous nature of radiocarbon age dated glaciogenic
units and limited geologic exposures precluded continuous tracing
A - 5
Woodward-Clyde Consultants
of these units to ice-marginal end moraines. Dating of these
end moraines is important for understanding the age and extent of
glaciation and Quaternary surfaces because of the spatial limits
that end moraines place on glacial episodes. Quantitative and
qualitative relative age dating techniques were applied to lateral
moraine sequences in four key areas in the Quaternary study region
(Figures 3-2, 3-5, and 3-6). The relative age dating results
from these moraines were then compared to those from moraines that
were relatively and absolutely age dated in the Alaska Range by
Ten Brink and Waythomas (in press) and Werner (in press).
Relative age dating techniques measure variations in morphologic
and soil weathering characteristics in similar glaciogenic deposits
through time. The basic premise in all the relative age dating
techniques is that, with age, surface weathering and erosion
processes will alter and modify both the form of the moraine and
the till deposits within the moraine. The degree of alteration or
modification is a function of age. Therefore, if other weathering
parameters such as slope, drainage, or climate are similar for the
various moraines, then differences in the physical characteristics
between moraines can be used as an indicator of relative age.
Relative dating techniques and measurement procedures used in
this study "'ere generally similar to those used by Ten Brink and
Waythomas (in press) in glacial geology studies of the Alaska
Range. Some modifications were made in these techniques and
procedures to provide the information in a manner consistent with
the scope of this investigation.
These modified techniques of Ten Brink and Waythomas (in press)
were used because of: the close proximity of the Quaternary study
region to the Alaska Range; the applicability of the technique to
the scope of this study; and the successful application of these
A - 6
Woodward-Clyde Consultants
techniques in the Alaska Range (including the south flank of
t h e A l a s k a R an g e ) . T h e f o 1 1 ow i n g d i s c u s s i on s u mm a r i z e s : t h e
procedures that were used for field measurements at each field
location; criteria that were followed for data collection; and the
modified techniques that were used.
Rock \'leathering characteristics were measured on coarse-grained
granitic rocks at the surface and near surface, and soil weathering
characteristics were measured in the near surface. At each field
location, these characteristics were measured at intervals along
the crest of lateral moraines for a distance of up to approximately
1,000 feet (312 m).
The following method was established for data collection: 1)
procedures were standardized for each re 1 at i ve-dat i ng technique;
2) techniques were quantified as much as possible to minimize
operator-induced bias; 3) the sample size was selected to cover the
range of small-scale inhomogeneity of the rock and soil; 4) each
technique was performed by the same operator at all localities; and
5) compilation and comparison of data from separate areas was not
undertaken until all the data were collected.
Morphologic characteristics of lateral and end moraines that were
measured included: inner and outer slope angles; crest width;
local relief; and post-depositional modification. Comparison of
the res u lt s o b t a i ned from f i e l d 1 o cat i on s d uri n g t h i s study
showed that none of these characteristics were diagnostic in
differentiating age. Therefore, the data obtai ned on morphologic
characteristics are not presented here, but they are kept in the
project file.
Surface weathering characteristics that were measured included:
1) boulder frequency per 1,000 feet (312m); 2) percent of boulder
A -7
Woodward-Clyde Consultants
surface roughened by weathering; 3) ratio of the number of weath-
ered boulders to fresh boulders; 4) ratio of boulders with pits to
boulders without pits; 5) relief of individual grains above the
surrounding rock surface of boulders; 6) ease of removal of
individual grains from the surface of boulders; 7) size (depth and
diameter) of weathering pits on boulder surfaces; and 8) total
relief of boulder surfaces caused by direct weathering. The
results for all of the surface weathering techniques have a large
amount of scatter and overlap in three of the four key areas
(Figure A-1) where they were applied. At the fourth area, the
Butte Lake area, moraines were differentiated by two of the
techniques: the size of the weathering pits and the total boulder
surface relief. These techniques indicate a marked increase in
degree of weathering between moraines BL6 and BL7. This break
corresponds to a similar weathering-age contrast indicated by the
subsurface weathering data (Table A-1). Although the data are not
included here, they are kept in the project file.
Two techniques for measuring subsurface rock and soi 1 weathering
characteristics--the granite weathering ratio and depth of oxi da-
tion--were consistently found to be of diagnostic value in dif-
ferentiating age in all four of the key areas shown in Figures 3-2,
3-5, and and 3-6. The granite weathering ratio is the percentage
of 50 boulders taken from the B horizon that, when hit with a
hammer, have the characteristic of being fresh, partly weathered,
or completely weathered (rotten). It is a measure of the degree
and ease with which the boulder fractures. There is an increase of
approximately 50 to 100 percent between the values obtained for
Early moraines and those obtained for Late Wisconsin moraines
(Table A-1). To collect these data, five subsurface test pits were
excavated by a backhoe at intervals along a 1,000 foot (312 m)
length of crest on each moraine. Pit depth ranged from 2 to 4 feet
(0.6 to 1.2 m). The depth of oxidation is the thickness of
A - 8
•
Woodward-Clyde Consu§ttants
the weathered zone above the unaltered parent material. The
weathered zone contains iron minerals that are oxidized and stained
orange. This depth generally corresponds to a gradational or sharp
color break in the soil profile. Depths were measured by tape.
Table A-1 shows the range of depth of oxidation in the soil profile
exposed in five pits on each moraine for which measurements were
obtai ned.
A.1.4-Interpretation of Six Subregions of the Quaternary Study
Region
The glacial chronology of the Quaternary study region is complex.
Unlike the systematic sequence of alpine glacial events on the
north side of the Alaska Range (Wahrhaftig, 1958; Ten Brink
and Ritter, 1980; Ten Brink and Waythomas, in press), repeated
convergant and multi-directional glacial flow occurred throughout
much of the Talkeetna Mountains. Glaciers from the south side of
the Alaska Range pushed southward through the areas of Butte Lake,
Deadman Creek, Brushkana Creek, and Watana Creek (Figure A-1)
to merge and coalesce in an intermountain basin with glaciers
flowing from ice centers in the higher elevations of the Talkeetna
Mountains. The intermountain basin includes the Watana site and
extends approximately 25 miles (40 km) to the southwest, 10 miles
(16 km) to the northeast, 3 miles (5 km) to the northwest; and
8 miles (13 km) to the southeast from the site (Figure A-1). It
includes the Susitna River area (near the Watana site), the Fog
Lakes area, the Stephan Lake area, the lower reaches of the Watana
Creek area, and the Deadman Creek area. The Susitna River bisects
the basin from east to west.
The Quaternary study region was divided into six subregions
(Figure A-1) according to similarities in the physiography and
character of the glacial morphology within each subregion. The six
A - 9
Woodward-Clyde Consultants
subregions have been designated as: 1) Clear Valley-Fog Lakes;
2) Chunilna Plateau; 3) Portage Creek-Devils Canyon; 4) Tsusena
Creek-Deadman Creek; 5) Butte Lake-Brushkana Creek; and 6) Kosina
Creek-Black River. The four key areas discussed in Section 3.4
each lie within one of these subregions (Figure A-1).
The elevation, nature, and geographic extent of the glacial
features in the six subregions (including the four key areas) form
the basis for the delineation of Quaternary surfaces shown in
Figure 3-2. Within each subregion, the major glacial features
were identified. However, the scope of the study precluded
identification and mapping of many small glaciogenic features that
could alter or improve knowledge of the glacial geology.
Glacial episodes recognized in the subregions were assigned to the
following ages: pre-Wisconsin, Early Wisconsin, Late Wisconsin,
and Holocene. Assignment was made on the basis of relative and
radiocarbon age determinations, similarities in geographic and
elevational extent, comparison with similar chronologie sequences
in adjacent areas of southcentral Alaska, and professional judg-
ment. The following discussion describes the pre-Wisconsin,
Early Wisconsin, and Late Wisconsin glacial activity and surfaces
encountered in the Quaternary study region. Glacial activity
and surfaces of Holocene age are discussed only in a cursory
manner because the activity is generally restricted to higher
intermountain valleys and cirque areas. The limited lacustrine and
fluvial (deposits) occur in the intermountain basin. Because of
their limited extent, these deposits, which are associated with
Holocene glaciation, were not considered a significant part of the
seismic geology evaluation.
A -10
A.1.4.1-Clear Valley-Fog Lakes Subregion
This subregion inc 1 udes the southern portion of the i nterrnountai n
basin from the Susitna River on the north, southward to the Tal-
keetna River area (Figure A-1). The Clear Valley area, described
in Section 3.4.2, lies in this subregion (Figures A-1, 3-6B).
Much of the subregion is mantled by fluted ground moraine deposits
of inferred Late Wisconsin age. These deposits locally overlie
highly oxidized glaciofluvial outwash deposits and angular collu-
vial deposits that were radiocarbon age dated at greater than
27,000 y.b.p. (Sample S34-l shown in Table 3-2 and Figure 3-2).
Oxidization of the outwash probably represents a major unconformity
during an interglacial period prior to the Late Wisconsin glacia-
tion.
Grooving on local bedrock nobs indicates that glaciers flowed to
the southwest and west. These glaciers were probably piedmont
glaciers that flowed through the intermountain bas in in Early and
Late Wisconsin time.
Data from the Clear Valley area (Section 3.4.2) were interpreted to
show that the elevational limit of Early Wisconsin ice was at
3,100 to 3,300 feet (945 to 1,006 m) and the elevation limit of
Late Wisconsin ice was at 2,500 to 2,700 feet (762 to 823 m). The
upper elevational limit for Early Wisconsin ice appears to decrease
from 3,100 to 3,300 feet (945 to 1,006 m) at Clear Valley to 2,900
to 3,100 feet (884 to 945 m) near the junction of the Talkeetna
River and Prairie Creek southwest of the C 1 ear Valley area. The
evidence on which this conclusion is based includes trimlines 9
side glacial channels, and end moraines. The elevational decrease
in the upper limit suggests that the ice surface for the Early
Wisconsin piedmont glacier, which flowed through the intermountain
A -11
Woodwaurd·Ciyde Consultants
basin, sloped approximately 13 to 14ft/mile (2.5 to 2.7 m/km) to
the southwest. The maximum elevation of Early Wisconsin glacial
features establishes that this glaciation was more extensive than
the Late vJisconsin glaciation.
The elevational limit of Late Wisconsin ice in the subregion is
indicated by prominent trimlines and morphologic contrasts between
2,500 and 2,700 feet (762 and 823 m) throughout the subregion. Ice
marginal lacustrine deposits that have not been overriden by
glacial ice are present at an elevation of 2,600 feet (793 m). The
maximum elevation of Late Wisconsin ice at this locality is,
therefore, limited to an altitude of 2,600 feet (793 m) or less.
Numerous ice-marginal features suggest that the piedmont glacier
continued to decrease in size throughout successive stades of the
Late Wisconsin glaciation in response to fluctuations in size
of individual valley glaciers that coalesced with the piedmont
glacier. Stephan Lake is presently dammed by a small moraine
representing the last stages of the Late Wisconsin piedmont glacier
as it retreated northward. The configuration and extents of
individual moraines that are estimated to be of the last stade
(Stade IV) of the Late Wisconsin stage, such as those in the Clear
Valley area, indicate that the valley glaciers did not coalesce but
were confined to individual valleys during the last stade of the
Late Wisconsin glaciation. Extensive ice-disintegration deposits
in the upper reaches of valley bottoms indicate rapid deglaciation
at the end of the Late Wisconsin.
A.1.4.2-Chunilna Plateau Subregion
This subregion is predominantly a glaciated bedrock plateau above
3,000 feet (915 m) elevation located south of the Susitna River and
extending from the intermountain basin of the Clear Valley-Fog Lake
A -12
Woodward-Clyde Consultants
subregion westward to the Gold Creek area on the Susitna River
(Figure A-1). Glacial grooving and other landform features
produced by ice scour are common. These ice-scoured features
suggest that ice flow was primarily to the west and southwest.
Major drainages in plateau margins have sharp V-shaped valleys
deeply incised into bedrock. This morphology suggests that
a fluvial environment has existed without interruption since
the glacial ice, which scoured the features, retreated from the
subregion.
Prominant trimlines, grooving orientation contrasts, surface
morphology differences, and side glacial channels on the margin of
the plateau all suggest an upper elevational limit to the Early
Wisconsin glaciation of 3,000 to 3,200 feet (915 to 976 m). The
Late Wisconsin ice limit is not expressed on the plateau but is
defined by ice-marginal features at lower elevations in adjacent
subregions.
Scoured glacial features above 3,200 feet (976 m) indicate that the
plateau was overridden in an earlier glaciation. The age of the
earlier glaciation was not dated during this study but is estimated
to be pre-Wisconsin. This old glacial surface was overridden by a
younger glacial event in the Chunilna and Disappointment Creek
areas. Valley profiles are U-shaped and floored by glaciofluvial
sediments in the lower reaches. Grooving is oriented parallel to
the valley, rather than obliquely, as it is higher on the older
surfaces. The regional snowline as determined for the Wisconsinian
shown by Pewe (1975) would suggest that the plateau elevations may
have been just high enough for small valley glaciers to develop
during the Early Wisconsin stage and possibly the first stade of
the Late Wisconsin stage.
A -13
A.1.4.3-Portage Creek-Devils Canyon Subregion
This subregion is a topographical continuation of the Chunilna
Plateau subregion. The east-west trending Susitna River valley
separates the two subregions. The plateau north of the river
rises to merge with a major mountainous area in the northwestern
Talkeetna Mountains (Figures A-1 and 3-2). This mountainous area
was a major area of ice accumulation. Drainage from the mountains
is controlled by the regional northeast-southwest trending struc-
tura 1 grain. The plateau is dissected by these southwest-trending
glaciated tributary valleys that drain to the Susitna River. These
drainages contrast with those of the Chunilna Plateau that have all
been cut by fluvial processes.
The elevational limit of pre-Wisconsin ice in this subregion
appears to be approximately 4,000 feet (1,219 m). Grooved and
beveled bedrock occurs up to at least 4,000 feet (1,219 m),
suggesting that the plateau was almost completely overridden by ice
during a pre-Wisconsin glacial episode. Later, less extensive,
glaciations left substantially larger areas of bedrock (at lower
elevations) above the glacier ice.
Markedly different surf ace weathering, triml i nes, and the presence
of end moraines between 3,000 and 3,300 feet (915 and 1,006 m)
indicate that this elevation is the upper limit of Early Wisconsin
ice. Relative age dating results suggest that Early Wisconsin
ice formed prominent but very weathered-looking moraines at
3,000 to 3,200 feet (915 to 976 m) on the west side of the mountain
ridge separating the Susitna River and Chulitna River valleys. The
oxidation depth of 21 inches (53 em) on the moraines is within
the range of pre-Late Wisconsin data obtained in the Alaska Range
and elsewhere in the Talkeetna Mountains (Table A-1).
A -14
The limits of Late Wisconsin ice are better defined than those of
earlier gl aci ati ons. A long the southern portion of the Susitna
River valley from Sherman Creek to Indian River (Figure 3-2),
pronounced grooving orientation contrasts, deflected drainages,
prominent trimlines, and kame terraces mark a sharp upper limit
to the Late Wisconsin glaciation at an elevation of 2,100 feet
(640 m). To the east through Devils Canyon, the upper limit
rises slowly in elevation. It lies between 2,100 and 2,200 feet
(640 and 671 m) at Portage Creek, as indicated by slope breaks,
fluting truncations on older surfaces, surface weathering con-
trasts, trimlines, and end moraines. Topographic contrasts and
side-glacial channels indicate an increase to 2,200 to 2,400 feet
(671 to 732 m) in the High Lake area; in the Devil Creek area,
trimelines, side-glacial channels, and surface weathering contrasts
suggest an increase to 2,400 to 2,600 feet (732 to 793 m).
The maximum Late Wisconsin limit at Devil Creek of 2,600 feet
(793 m) is similar to that found in the adjacent Clear Valley-
Fog Lakes subregion (Section A.1.4.1). This suggests that the
piedmont ice mass from the intermountain basin advanced through the
topographic constriction at the western bend of the Susitna River
(near Devil Creek) into Devils Canyon and merged with local valley
glaciers from the northeast. Ice margi na 1 features suggest that
the ice was of sufficient extent for the piedmont and valley
glaciers to coalesce and override the Devils Canyon area only
during the first and possibly second stades of the Late Wisconsin
stage.
The piedmont glacier and valley glaciers did not merge during
later stades of the Late Wisconsin stage. In Portage Creek, the
extent of subtle, heavily vegetated moraines and corresponding
drainage profile changes indicate that the southwestern limits of
glacial ice during Stade III and IV of the Late Wisconsin were
A -15
Woodward· Clyde Consultants
4 and 6 miles (6 and 10 km), respectively, upstream from the
confluence of Portage Creek with the Susitna River. The valley
profile of Devil Creek indicates that the extent of Late Wisconsin
glacial stades in that creek was similar to the extent of stades in
Port age Creek.
The age of the Susitna River canyon~ which cuts across the Quater-
nary study region, is difficult to estimate. Extensive river
terraces located in the Indian River area indicate that the Susitna
River and associated tributaries served as major outlets for
glacial meltwater and copious amounts of glacially derived sedi-
ments carried by the meltwater. The Susitna River at Devils Canyon
and many adjacent tributary canyons are bedrock gorges that are
500 to 600 feet (152 to 183 m) deep in some places. They do not
exhibit glacial modification in some areas that were clearly
overridden by Early Wisconsin and early Late Wisconsin ice (e.g.,
the lower Portage Creek area). This suggests that the Devils
Canyon is younger than Stade II of the Late Wisconsin stage
(15,000 y.b.p). However, glaciofluvial outwash deposits located
near river level approximately 1.2 miles (2 km) south of Fog Creek
along the Susitna River were radiocarbon age dated by Terrestrial
Environmental Specialists (1981) at 37,000 y.b.p. The location of
these deposits suggests that the river canyon is at least 30,700
years old (i.e., the outwash sediments were deposited after the
canyon was cut).
The apparent discrepancy in canyon age may be caused by deposition
of older organic material from another (upstream) source into the
outwash deposits. These outwash deposits are in the area that
morphologically appears to have been glaciated until near the end
of the Late Wisconsin stage (Figure 3-2). Thus, older organic
material could have been reworked at this time.
A -16
Elsewhere along the Susitna River valley, hanging tributary
drainages suggest rapid down-cutting by the Susitna River after
retreat of the Late Wisconsin Stade II ice, 15,000 y.b.p. Alter-
natively, Devils Canyon and other bedrock canyons along tributary
drainages could have been repeatedly filled and flushed during
glacial advances and retreats and, therefore, not greatly modified
by ice erosion. Following the second stade, individual valley
glaciers were confined between valley walls, and the piedmont ice
mass in the Stephan Lake area trough may have retreated north of
the Susitna River. Retreat and deglaciation near the end of
the Late Wisconsin stage would have produced large quantities
of sediment-loaded meltwater. Isotatic readjustments during
deglaciation and minor regional tectonic uplift may also have
assisted down-cutting by effectively lowering the base level.
River terraces adjacent to the Susitna River floodplain indicate
that at least 50 to 100 feet (15 to 30 m) of do~tm-cutting occurred
during Holocene time.
A.1.4.4-Tsusena Creek-Deadman Creek Subregion
This subregion forms the northern half of the intermountain
basin which is divided by the westward-flowing Susitna River
(Figure A-1). The Deadman Creek area, described in Section 3.4.4,
lies in this subregion (Figures A-1, 3-68).
The elevation of the basin floor rises gradually to the north and
merges with major drainages and valleys that carried ice from the
Alaska Range in the Butte Lake-Brushkana Creek subregion. Ice from
the Alaska Range and the Talkeetna Mountains coalesced in the
basin. The extent to which the glaciers coalesced during each
glacial episode was different and was a function of the magnitude
of each glacial episode and the restrictions to glacial flow caused
by local topographic relief.
A -17
Surficial glaciogenic units in this subregion are predominantly
hummocky ice-disintegration deposits and extensive lacustrine
deposits which contrast sharply with the surface morphology of
the south side of the Susitna River in the Stephan Lake-Fog Lakes
subregion (Figure A-1). Fluted ground moraine or beveled bedrock
outcrops are also present in the Tsusena Creek-Deadman Creek
subregion.
The geomorphology of the subregion and relative and radiocarbon age
dates from the deposits indicate that the sediments mantling the
subregion are Late Wisconsin in age. The generalized cross-section
(Figure 3-3) shows the stratigraphic relationships that were
interpreted from numerous measured exposures. Highly oxidized
outwash deposits overlie lacustrine deposits that are older than
37,000 y.b.p. (Sample S29-1, Table 3-2). The oxidized outwash
deposits are overlain by till v1hich in turn is overlain by ice-
disintegration deposits. In the northern part of Watana Creek
valley, the till (which stratigraphically correlates with the till
described above) is overlain by younger lacustrine deposits and by
glaciofluvial, deltaic, and ice disintegration deposits. These
latter deposits are in turn overlain by lacustrine deposits that
have a radiocarbon age date of 9,395 ~200 y.b.p. (Sample S42-1,
Table 3-2). These younger sediments (shown in measured Sections C,
E, F, and G in Figure 3-3) date the last retreat of ice from the
basin and are interpreted to represent a sequence of northward
shrinking, retreating, and stagnating glacial ice that was debouch-
ing large amounts of meltwater and sediments toward the south.
Interpretation of relative age data for closely spaced lateral
moraines west of Deadman Creek (Figure 3-68) indicates that the
elevational limit of Early Wisconsin ice was at least 4,200 feet
(1,280 m) and that the limit of Late Wisconsin ice was 3,900 feet
(1,189 m).
A -18
Woodward-Clyde Consultants
Local valley glaciers merged with glaciers from the Alaska Range
that advanced southward down Deadman, Watana, Brushkana, and Butte
Creeks during Early Wisconsin time and probably during early Late
Wisconsin time. The glaciers merged to produce a piedmont glacier
that filled the intermountain basin. During later stades of the
Late Wisconsin stage, Alaska Range glaciers were of insufficient
thickness to advance southward through the relatively high passes
at Deadman Creek, Brushkana Creek, and Butte Lake, but still
advanced southwestward through the northwest end of the Watana
Creek valley as discussed in Section A.l.4.5. Northward-sloping
slopes of morainal crests DC-1 and DC-2 at Deadman Creek and a
concave southward arcuate terminal moraine (Figure 3-6B) at Big
Lake suggest that local valley glaciers from Tsusena Creek flowed
northward during later s tades of the Late Wisconsin stage to fill
areas that were left free of ice by the limited southward advance
of Alaska Range glaciers.
The piedmont ice mass appears to have retreated north of the
Susitna River by the last stades of the Late Wisconsin stage.
Fluted ground moraine on the south side of the Susitna River
contrasts with extensive younger lacustrine and ice-disintegration
deposits on the north side. This contrast indicates ice was
present north of the river and absent to the south during this
time. The lower reaches of Watana Creek were dammed by ice
during the later stades of the Late Wisconsin stage, and extensive
lacustrine sediments were deposited in the Watana Creek area.
Ice stagnation and ablation during the last stades of the Late
Wisconsin stage caused large areas to be mantled with ice disin-
tegration deposits, particularly in topographic low areas.
Near the end of the Late Wisconsin stage, the Tsusena Creek-Clark
Creek valley glaciers extended southward into the Susitna River
A -19
valley south of the mouth of Tsusena Creek and modified the valley
profile. Retreat and deglaciation produced hummocky ice-disinte-
gration deposits north of Watana Camp near the Watana site.
Ice-disintegration deposits in lower Deadman Creek formed a low dam
that lasted well into the Holocene Epoch. The radiometric age
of the lacustrine sediments is 3,450 +170 y.b.p. (Sample S45-1,
Table 3-2). This date suggests that the lake existed at least
until this time, after which the stream profile was re-established
by down-cutting through the ice disintegration deposits.
A.1.4.5-Butte Lake-Brushkana Creek Subregion
This subregion includes the northern ends of the major valleys that
drain southward and merge with the intermountain basin described
in the Clear Valley-Fog Lakes and Tsusena Creek-Deadman Creek
subregions (Section A.l.4.1 and A.l.4.4, respectively). The Butte
Lake area, described in Section 3.4.3, lies in this subregion
(Figures A-1, 3-6A).
Broad U-shaped valleys opening to the north and rounded topography
are characteristic of this subregion. This topography was caused
by extensive glacial scour and overridding by glaciers from
the Alaska Range. Drainage patterns and directions have been
altered by glacial erosion and deposition. Valley bottoms are
predominantly mantled by till, and upper valley walls are exten-
sively mantled by frost-shattered boulder fields and a type of
patterned ground called nonsorted stripes. The boulder fields and
nonsorted stripes formed in an ice-free periglacial environment.
Their lower limit marks the upper limit of younger ice.
Numerous moraines in the Butte Lake area (Figures 3-2 and 3-6A) are
interpreted to represent repeated southward glacial advances by
Alaska Range glaciers. On the basis of morphologic contrasts and
A -20
relative age data (Table A-1), moraines at elevation 4,100 feet
(1,250 m) and above are considered to be probably Early Wisconsin
in age. The maximum elevation of Early Wisconsin ice may be marked
by prominent trimlines at elevation 4,200 to 4,300 feet (1,280 to
1,311 m) ~hat correspond to the lower limit of the extensive
frost-shattered boulder fields. The surface morphology above
elevation 4,300 feet (1,311 m) includes rounded topography,
erosional features suggestive of glacial scour, and glacial
erratics. This surface morphology indicates a pre-Wisconsin
glaciation above elevation 4,300 feet (1,311 m).
The configuration and elevation of the moraines in the Butte
Lake area suggest that only ice of the Early Wisconsin stage and
Stade 1 of the Late Wisconsin stage was thick enough to advance
southwestward through the passes of Brushkana Creek (at elevation
3,250 feet [991 m]) and Deadman Creek (at elevation 3,350 feet
[1,021 m]) to merge in the intermountain basin with valley glaciers
from the Talkeetna Mountains. Ice thickness during later stades
was insufficient to allow ice to flow through the passes.
Elevations of the Late Wisconsin end moraines suggest that as many
as nine individual end moraines of Late Wisconsin age are present
in the Butte Lake area. The moraines have been tentatively grouped
into four distinct Late Wisconsin stades according to clustering
of end moraines, breaks in slope, kettle-frequency contrast, and
surface-morphology contrasts (Figure 3-6A). These stades and their
maximum elevation are: Stade 1, 3,900 feet (1,189 m); Stade 2,
3 ,600 to 3,800 feet (1 ,098 to 1,159 m); Stade 3, 3 ,200 to 3,300
feet (976 to 1,006 m); Stade 4, 3,000 to 3,100 feet (915 to 945 m).
Most evidence at the Butte Lake-Brushkana Creek subregion supports
a maximum elevation of 3,800 to 3,900 feet (1,159 to 1,189 m) for
Late Wisconsin ice. Prominent trimlines at 3,800 to 3,900 feet
A -21
Woodward·CByde Consultants
(1,159 to 1,189 m) at Butte Lake are consistent in elevation
with the upper limit of Late ~~isconsin moraines in lower Deadman
Creek (Section A.1.4.4). Some apparently contradictory evidence
was found in Brushkana Creek where surficial lacustrine/deltaic
deposits at elevation 3,000 feet (915 m) are older than 37,000
y.b.p. (Sample S12-2, Table 3-2). The age of the sediments suggest
that they were deposited before Late Wisconsin time. Because the
deposits have not been overridden, the maximum elevation of Early
Wisconsin ice would have been at 3,000 feet (915 m) or less rather
than 3,900 feet (1,189 m). Instead, the deposits are probably from
the late stade of the Late Wisconsin substage deposited as Alaska
Range glaciers retreated northward. The organic material may have
been reworked from an older deposit into a Late Wisconsin ice
marginal lake.
In contrast to Brushkana and Deadman Valleys, where the passes are
higher than elevation 3,250 feet (991 m), the upper Watana Creek
valley drainage divide is at elevation 2,850 feet (869 m). The
lower elevation of this divide allowed Late Wisconsin ice of
Stades II, III, and IV to advance through the Watana Creek valley
into the intermountain basin where it coalesced with ice from local
Talkeetna Mountain sources. This was the only route in the
subregion along which glacial ice from the Alaska Range was able to
advance into the intermountain basin following Stade I of the Late
Wisconsin stage. Lacustrine deposits dated at 9,395 2_200 y.b.p.
(Sample S42-1, Table 3-2) just south of ice disintegration deposits
in the northeast part of the Watana Creek valley (Figure 3-2) date
the last retreat of Alaska Range ice from the Watana Creek valley
area. During retreat northward, meltwater and sediment formed
extensive deltaic, outwash, and lacustrine sediments in lower
Watana Creek.
A -22
Woodward· Clyde ConsuStants
A.1.4.6-Kosina Creek-Black River Subregion
This broad undulating bedrock platform south of the Susitna River
in the eastern Talkeetna Mountains merges indistinguishably with
the adjacent Copper River basin to the east. Three morphologically
distinctive, glacially scoured topographic surfaces have been
beveled into the platform by succeedingly less extensive glacia-
tions (Figure 3-2). The Black River area, as described in Section
3.4.1, lies in this subregion (Figures A-1 and 3-5A).
An extensive glaciation of suggested Early Wisconsin age spread out
over the platform. Side glacial channels, moraines, and discordant
topography suggest an upper limit to this glaciation between 3,550
to 3,750 feet (1,082 to 1,143 m). A more extensive pre-Wisconsin
glaciation above 3,750 feet (1,143 m) is indicated by rounded
topography, glacial debris, and erosional features.
The distinction between Early and Late Wisconsin glaciations has
been made on the basis of morphology and elevation of moraines,
by relative age data from the Black River moraines (Tables A-1
and 3-3), and on the basis of valley morphology. The youngest or
Late Wisconsin glaciation deeply incised and modified preexisting
valleys and deposited four Late lrJisconsin end moraines in the upper
valleys of Tsisi Creek, Kosina Creek, and Black River. The maximum
northward extent of the Late Wisconsin ice in the Black River
area is identified by the arcuate termination of the uppermost
1 ate r a l mo r a i n e ( B R-3 o n T ab l e A-1 ) , ap pro x i mat ely 6 t o 7 m i l e s
(10 to 11 km) south of the Susitna River. Downstream from the
front of the maximum ice extent, the head of a sharp V-shaped
fluvial-cut canyon also suggests the northernmost advance of ice in
Black River. Similar morphology and configuration of end moraines
in the Tsisi and Kosina Creek area suggest a similar order of
magnitude for the extent of Late Wisconsin glacial events in
this subregion.
A -23
Woodward·CDyde Consultants
Near the confluence of the Oshetna and Black Rivers (Figure A-1).
conflicting radiocarbon age dates were collected during this
study and by Terrestrial Environmental Specialists (1981). These
differing dates have led to contrasting interpretations of the
glacial history in this subregion. Within the exposure at this
location, till interfingers with highly deformed lacustrine
sediments. A wood sample (S47-4) obtained from the lacustrine
deposits during this study provided a radiocarbon age date of
>37,000 y.b.p. (Table 3-2). The relationship of the till to the
1 acustri ne deposits and the radiocarbon age date suggest that the
till is Early Wisconsin in age. In addition to these data, the
extent of moraines and the relative age dating studies of these
moraines in the Black River valley suggest that Late Wisconsin ice
did not advance into the Susitna River west of the Copper River
basin in the subregion (Figure 3-2).
Terrestrial Environmental Specialists (1981) obtained a radiocarbon
age date for the same till unit, but somewhat higher in the
stratigraphic section than the sample described above. This date
was 24,900 ..:!:_3,125 y.b.p. Terrestrial Environmental Specialists
(1981) interpreted the sediments to be recessional ice-contact
drift that was deposited by Late Wisconsin glacial ice.
This difference in the age of glacial surfaces in the eastern
Talkeetna Mountains does not affect the interpreted age and extent
of glacial surfaces in the vicinity of the Project sites. It does,
however, point out the difficulty in interpreting the age and
extent of gl aci a 1 deposits where glaciers from different sources
and different ages converge.
A -24
Woodward-Clyde Consultants
A.2 -Radiometric Age Dating
A.2.1-Radiocarbon Age Dating
Radiocarbon age dates were obtai ned from 11 samples collected in
the Quaternary study region (Figure 3-2). The dates measured by
Geochron Laboratory are presented in Table 3-2. The procedures
used to collect, store, and transport the samples are described in
Section A.1.3.1. These procedures were in accordance with those
described in Woodward-Clyde Consultants (1975).
A.2.2-Potassium-Argon Age Dating
Whole rock potassium-argon age dates were obtained for three
samples of andesite collected in the vicinity of the Talkeetna
thrust fault. The purpose of obtaining the dates was to limit
the area in which the fault is present upriver from the Watana
site.
The locations from which the samples were collected are shown in
Figure A-1. The procedures used to collect the samples were in
accordance with those described in Woodward-Clyde Consultants
(1975).
Analyses of the samples were conducted by Dr. Kenneth A. Foland at
the Potassium-Argon Laboratory at Ohio State University. In
addition to the whole rock age dates, thin sections were prepared
of the samples, and petrographic descriptions of the thin sections
were made . T h e res u lt s of t h e age d at i n g are s u mm a r i z e d i n
Table A-2. The procedures used to conduct the age dating and the
thin section analysis are discussed in the following paragraphs.
A -25
Woodward·CUyde Consultants
A.2.2.1 -Sample Preparation
Samples for whole-rock dates were prepared from the hand-samples
provided. Each sample was first trimmed of weathered rind using a
rock saw with water as a cutting fluid. After drying, the trimmed
sample was crushed with a jaw crusher and then was ground with a
disc grinder. The ground rock was separated into size fractions by
sieving with nested sieves on a Ro-Tap shaker. The size fraction
falling between #40 and #60 U.S. standard sieve sizes was used for
analyses. Prior to analysis, the #40 and #60 fraction was vJashed
with water to remove adhering dust and with dilute hydrochloric
acid (HCl) to remove undesirable carbonates. The procedure was:
10% HCl wash; reagent acetone wash; reagent ethanol wash; and,
triplicate wash with doubly-distilled water. Each sample was then
dried in an oven at 80°C,
A.2.2.2-Argon Analysis
The concentrations of argon were determined by isotope dilution.
The sample was fused by induction heating in a molybdenum crucible
in a high vacuum system at pressures less than 5xlo-7 torr. An
38Ar tracer was added from a reservoir-gas pipette metering
system. Argon was purified using hot CuO and Ti gettering. The
purified argon was isotopically analyzed using a 6-inch gas mass
spectrometer, Nuclide Corp. model SGA-6-60, operated in the static
mode on line with a Hewlett Packard minicomputer.
A.2.2.3-Potassium Analysis
Potassium concentrations were analyzed by flame photometry using an
Instrument Laboratory Model 443 Flame Photometer with a lithium
internal standard. Rock samples were fused as 850°C with lithium
metaborate which serves as a flux and internal standard. The fused
A -26
samples were then dissolved in a 4% HN03 solution. The resulting
solutions were then diluted and analyzed for potassium on the flame
photometer which was calibrated with gravimetric standards.
A.2.2.4-Age Calculation and Uncertainties
The averages of repetitive analyses were used to calculate K-Ar
whole-rock dates using the following constants:
40K = 1.167xlo-2 atom % of K
t..e = 0.581x1o-10 y-1
t..B = 4.962x1o-10 y-1
>..tot a 1 = 5.543x1o-lO y-1
(40Ar/36Ar) in air = 295.5
The quoted age uncertainties reflect only analytical uncertainties
and make no provision for those in the accepted and recommended
above constants. Also, the quoted uncertainties do not reflect any
11 geological error,11 such as failure of the rock systems to behave
as ideal systems. The K and Ar calibrations were checked by
analyzing interlaboratory mineral standards.
A.2.2.5-Petrographic Analyses
Thin sections were prepared for each sample along the marked
orientation (if indicated). Two sections were prepared for each
sample to ensure a representative petrographic description. The
sections were studied with standard petrographic microscopes.
Microphotographs were taken of the thin sections.
A -27
A.3 -Field Mapping
Geologic data were collected by aerial reconnaissance from helicopters
and fixed-wing aircraft, interpretation of aerial photographs, and study
of approximately 300 locations on the ground. Photographs were taken
during these activities to document the aeri a 1 appearance of features.
The observed geomorphology, lithology, and geologic structure were
described using methods described in geologic field texts such as
Compton (1962). The orientation of features were measured with Brunton
compasses. Samples were collected for later reference and comparison of
rock types at different locations.
Ground locations were assigned outcrop location designations based on
the first letter of the name of the geologist making the observation and
on the consecutive number of the observations. For example, outcrop
locations studied by Phillip Birkhahn have designations such as B-32,
B-33, etc. Closely spaced observations were given the designation
B-32A, B-32B, etc., as needed. Samples were labeled with the outcrop
location designation and a number (for example, B-32-1, B-32-2, etc.).
Photography was done with 35 mm single-lens reflex, 35 mm rangefinder,
and color and black-and-white Polaroid cameras. The rolls for the 1981
study were numbered consecutively beginning with S-200. The types of
film used were: Kodachrome 64, Ektachrome 200, Plus-x pan, Polaroid
Type 108, and Ektachrome near-infrared.
The geologic data collected were recorded in field books (J.L. Carling
Corporation Field Book No. 350). The field locations of these data were
plotted on U.S. Geological Survey maps at a scale of 1:63,380 and
1:24,000 on aerial photographs. Aerial photograph interpretations were
marked on clear mylar overlays. Where appropriate, geologic maps were
produced of selected areas using the data obtained.
A -28
Woodward=C!yde Consultants
The geologic data and photographs were also recorded on the following
forms: Remote sensing lineament worksheet (SHP-4), Fault and lineament
data summary sheet (SHP-3), Field observation documentation sheet
(SHP-6), Photo log (SHP-7), Fault and lineament photo log (SHP-8),
and Fault and lineament index sheet (SHP-5). The use of these forms
is described in Appendix A of the Interim Report (Woodward-Clyde
Consultants, 1980b), and copies of the forms are provided.
The data are on file in Woodward-Clyde Consultants office at Santa Ana,
California, as follm>~s: field locations and outcrops described in field
notebooks; maps of the field locations plotted on a single set of
1:63,380-scale maps; photographs, slides, and negatives; field maps; and
air-photo overlays.
A.4 -Trench Logging Methods
Trench logging and photography were conducted by Woodward-Clyde Consul-
tants in three trenches excavated within 26 miles (41 km) of either
site. Two of the trenches, Treches T-1 and T-2, were excavated across
or near the Talkeetna thrust fault (Figures 4-11 and 4-7, respectively).
One trench was excavated near the Susitna feature (Figure 4-17). The
purpose of the logging and photograpy of these two trenches was to
document the presence or absence of faults with recent di sp 1 acement.
The trench logs which document the evidence of no fault displacement
in the trenches are presented as Figures 4-12 and 4-18. Selected
photographs of the trenches are presented as Figures 4-13, 4-14,
and 4-19.
Interpretation of the stratigrahic relationships shown in the trench
logs is presented in Section 4. The original trench logs and photo-
graphs of the trenches are in the Project file at Woodward-Clyde
Consultants, Santa Ana, California.
A -29
Woodward·CByde Consuitanfts
The three trenches were excavated by R & M Consultants of Anchorage,
Alaska. A track mounted John Deer 350 backhoe with a 0.07 cubic yard
(0.05 m3) bucket was then used to excavate the trench; subsequent to
excavation, the trenches were shored using hydraulic aluminum shoring.
Also subsequent to excavation, one of the trench walls was selected for
logging; the southwest wall was selected primarily because of lighting
restrictions imposed by the spoils pile on the northeast wall. Hori-
zontal and vertical control lines were then established. The horizontal
control line (level line) was established using three strand nylon
string and a K&E hand level or Stanley string level. After nailing the
level line to the trench wall, horizontal distances were measured using
100-foot Keson plastic coated cloth tapes. The stationing system was
established and recorded in feet southeast or northwest of the point
where trench logging commenced. Vertical distances relative to the
level line were measured using a standard six-foot wood carpenter's
rule.
Following placement of the level line, the trench wall was cleared using
a variety of techniques. A mason's hammer or a pick-axe was used to
remove smear and slough material. A stiff-bristled brush or paint
brush was used to clean and enhance stratigraphic relationships in sand
and gravel deposits. In some instances the trench wall could not be
prepared adequately for logging due to deep water and trench wall
failures. These parts of the trenches were well away from the scarps
and thus do not affect the conclusions regarding scarp origin.
After the trench wall was prepared for logging, stratigraphic relation-
ships were documented by recording the relationships on trench logs and
photographing the trench walls. Selected stratigrpahic units were
logged to document the absence of fault displacement in the trench and
to document stratigraphic relationships encountered in the trench.
A -30
Woodward-Clyde ConsuStl:ants
Trench logging procedures and results were reviewed upon completion of
the trench logging by Woodward-Clyde Consult ants' project review team.
Representatives of Acres and the Alaska Power Authority also examined
the trenches.
In order to effectively log and display observed stratigraphic relation-
ships, trench logs were drawn at a scale of 1 inch equal to five feet
vertically and horizontally. Previous experience in logging trenches
has demonstrated that this scale best portrays the observed strati-
graphic relationships and adequately displays potentially significant
features. For publication in this report, the trench logs were reduced
to a scale of one inch equal to approximately ten feet vertically
and horizontally (Figures 4-12 and 4-18).
The stratigraphic relationships were documented photographically with 35
millimeter SLR cameras equipped both with 35 millimeter 1.4 lenses and
haze filters and with 50 millimeter 1.8 lenses and haze filters.
Ektachrome film (ASA rating of 200) was used for color photography and
Tri-X Panchromatic film (ASA rating of 400) was used for black-and-white
photography.
An electronic flash was often used as a fill-in to illuminate the trench
walls while photographing the exposed stratigrahic units. Photographs
were taken at various spacings depending on trench width. After some
experimentation, six-foot centers were selected to provide adequate
photographic coverage without sacrificing necessary detail. This
spacing also permitted relatively rapid photographic documentation of
the trench wall. Locally, where the trench walls were narrow, or where
the trench was deepened to expose significant features, photographs were
taken on smaller centers. Generally, three-foot centers provided
adequate coverage in these areas.
A -31
A.S -Geophysical Studies
Geophysical studies in support of the geological portion of the investi-
gation were undertaken during the summer field season of 1981. The
objectives of these studies were to locate and delineate the subsurface
trace of the Talkeetna thrust fault and possible sources of the Susitna
feature and Feature KDS-3 (discussed in Section 4.4). Geophysical lines
were conducted in areas believed to have faults or possible faults in
order to confirm their extent, direction, and sense of movement.
The techniques selected for use were seismic refraction and magnetic
surveys. The scope of work consisted of the following tasks:
a) An evaluation of potential geophysical line sites to determine
which method or methods would be most effective at each site.
b) A survey of the sites using the selected technique(s).
c) Interpretation and evaluation of the data to locate the desired
faults and features.
A.S.l-Field Operations
Thirteen geophysical profiles were obtained in the project region.
The following sections cover the methods and equipment used.
A.S.l.l-Seismic Refraction Measurements
Seismic refraction measurements were obtained on one line across
the buried trace of the Talkeetna thrust fault near Fog Creek
using a GeoMetries model ES-1210F twelve-channel signal enhancement
seismograph. The signal enhancement feature allows the signal-
to-noise ratio of a low-energy source to be improved by adding
A -32
together 5 to 15 source inputs whose waveforms are consistent and
therefore add, while noise being random tends to cancel out.
The source used for this survey consisted of 1 to 6 pound charges
of ANFO. Triggering was obtained directly from the blaster. The
seismic refraction lines were recorded using a 100 foot (30 m)
geophone spacing except for the outside geophones which were
moved inward 50 feet (15 m). The source locations were 50 feet
(15 m) off each end of the line and midway in the line. Terrain
constraints often caused some minor variations from the line of the
survey and the geophone spacing.
Geophone and source locations were cleared of vegetation and
soft surface material. Geophones were firmly planted away from
disturbing influences such as tree roots, moving water, and wind
disturbance.
A.5.1.2-Magnetometer Measurements
The magnetometer readings were obtained along 12 profiles across
the Talkeetna thrust fault, the Susitna feature, and Feature KD5-3
with a GeoMetries Model G-816 portable proton magnetometer. This
instrument measures the total magnetic field of the earth with an
accuracy of 1 gamma under optimum conditions.
A base station was established at an accessible and magnetically
stable point near each profile. Repeated readings were made at the
base station in order to establish the diurnal variation in the
earth's magnetic field during the survey period. Additional base
stations were established on the longer profiles in order to
minimize travel time. Stations were repeated when high diurnal
drift rates occurred. Each station value was obtained by averaging
the readings taken at four sites in a 10-foot (3-m) radius circle
A -33
Woodward-Clyde Consultants
around the station. Measurements were made every 200 feet (61 m)
on all profiles except those at Talkeetna Hill where a 50 foot
(15 m) spacing was used.
A.5.2-Data Reduction and Interpretive Procedures
This section discusses the reduction procedures required to
prepare the data for interpretation and describes the interpre-
tative techniques utilized. Selected profiles are presented in
Section 4. Originals of all profiles are in the Project file of
Woodward-Clyde Consultants, Santa Ana, California.
A.5.2.1-Seismic Refraction Measurements
The arrival times of the seismic waves were plotted for each shot
to produce a continuous time-distance profile for the seismic
line. The profile was interpreted using a two-layer model with
lateral velocity variations. The velocities were obtained from
the slopes of the plotted data. The depth to the second layer and
lateral variations in the model were determined by using ray
theory, as discussed in any basic geophysics text.
A.5.2.2 -Magnetometer Measurements
The plotted magnetic field values represent the average values
at each station after correcting for diurnal variation. The
unaveraged values at each station usually fall within a ten-gamma
range. The diurnal variation was obtained from the plot of base
station values as a function of time.
No attempt was made to correct for topographic effects. Where
such effects do occur, they are noted in the discussions of the
individual profiles.
A -34
Qualitative interpretations of the profiles are provided. Quanti-
tative interpretation is not appropriate for the data obtained.
Two features of the data are used for interpretation: the spatial
frequency and the amplitude of the magnetic variations. A change
in either characteristic implies a change in subsurface character
which may be fault related.
A.6 -Low-Sun-Angle Aerial Photography
An aerial photographic survey of a 36 square-mile (100 km2) area
centered on each site and of a three-mile-(5-km-) wide strip along the
Talkeetna thrust fault and Susitna feature was flown by Air Photo Tech
of Anchorage, Alaska. The survey was made during May 1982. The survey
was specially designed to emphasize the presence of faults with recent
displacement.
The photographs were taken from a Piper Aztec aircraft at an elevation
of 12,000 feet (3,630 m) above the terrain. The camera used was a Zeiss
Jena MRB 15/2323 having a focal length of 6 inches (15.3 em). The film
was exposed using a Kodak Wratten number 12 filter. The film used was
a 9.4 inch (24 em) Eastman Kodak Aeroch rome Infrared 2443 having an
effective aerial film speed (AFS) of 40 daylight. Approximately 220
images were obtained at a scale of 1:24,000 and were printed on a 9 inch
by 9 inch (23 em by 23 em) format as one set of positive transparencies
and one set of Cibachrome prints. These are in the Project file at
Woodward-Clyde Consultants, Santa Ana, California.
The photographs were taken when the angle of the sun above the horizon
was 10° to 25•. The grazing angle of the sunlight enhanced the appear-
ance of small topographic features such as fault scarps. Infrared
film was used to enhance the appearance of any vegetation 1 i neaments
that might mark the location of a fault. Infrared film provides an
exceptionally strong contrast between different types of vegetation.
A -35
A.7 -Earthquake Recurrence Calculations
The average recurrence i nterva 1 of an earthquake is the average time
between earthquakes of a given magnitude. Recurrence i nterva 1 s are
often obtained from field evidence of individual earthquakes that caused
surface faulting and that left some permanent record to be interpreted
by geologists. When direct geologic evidence cannot be obtained as
evidence of recurrence interval, slip rates can be useful in estimating
recurrence intervals based on theoretical seismology.
The recurrence calculations are primarily based on the magnitude-
frequency relationship presented by Richter (1958):
Log N = a - b M . . Equation A-1
where N = the number of earthquakes of magnitude M and 1 arger
occurring within a defined time interval for a group of
faults or area of the earth's surface containing a
group of faults;
a and b = empirical constants;
and M = magnitude.
Use of this relationship leads to numerical estimates for return
period (recurrence) as a function of magnitude for that area of earth's
surf ace.
Because faults are the sources of earthquakes, it is considered reason-
able to apply the magnitude-frequency relationship to individual faults,
although significant uncertainites exist in how the relationship is
related to single faults. The uncertainty stems largely from a lack of
historical data to verify the theoretical model. For example, from
A -36
Woodward-Clyde Co111su6iants
the data available, it appears that the frequency of occurrence of
earthquakes of a specific magnitude on a fault is highly variable and
may be related to cyclic periods of activity and inactivity lasting many
tens to hundreds of years. Thus, the geologic and historical data
available in California for the past 50 to 180 years may provide
evidence of consistency with a recurrence model, but they do not provide
a basis for constructing the model.
An alternative to estimating recurrence intervals based on the magni-
tude-frequency relationship alone is to refine that estimate by incor-
porating geologic data into the calculation. If it is true that the
total energy released by a fault in the form of earthquakes is related
to the net offset occurring along the fault for the same time period,
then the magnitude-frequency relationship can be adjusted to a single
fault by using the slip rate of the fault. The measure of earthquake
energy per event is often expressed in terms of moment, M0 :
M = d 0 A . J..l ..... Equation A-2
where d = the average slip per event occurring within a fault rupture
area (A);
A = the fault rupture area;
and J..l = the shear modulus.
Dividing this equation by a time period gives the rate of energy release
(moment rate) for that period as a function of the average slip rate.
The assumptions made for the analysis of recurrence intervals are:
1) The total moment rate on a fault is given by the product of the
area of the fault, the slip rate, and shear modulus.
A -37 •
Woodward·Ciyde Consultants
2) All of the displacement is considered to occur seismically.
3) The magnitude-frequency relationship is considered to be linear
with an assumed slope (b) of -0.85 up to the maximum magnitude
assigned to the fault. This slope is selected to be appropriate
for the southern Alaska tectonic setting and seismicity.
Using these assumptions, we have calculated the recurrence intervals for
possible earthquakes on the Castle Mountain and Denali faults with a
slip rate of 0.5 and 1.0 cm/yr, respectively, for a possible maximum
magnitude of 7.5 and 8.0, respectively. For each fault, the area of
the fault is the product of MCE (maximum credible earthquake) rupture
length and width from Table 7-1. The shear modulus was assumed to be
3x1011dyne/cm2.
Anderson's method (1979) was used to calculate the "a" values for the
maximum magnitude selected. Using the calculated "a" value and the
assumed "b" value, we calculated the numbers of earthquakes expected
annually within one-half magnitude ranges. The inverse of each of these
numbers is the recurrence interval of earthquakes within the respec-
tive magnitude ranges. It is important to note that the recurrence
intervals presented here represent a mathematical distribution within a
half-earthquake magnitude range.
Using the above method, a magnitude (Ms) 7-1/4 to 7-3/4 event may be
expected to recur on the Castle Mountain fault at an average rate of
235 years. A magnitude (Ms) 7-3/4 to 8-1/4 event may be expected to
recur on the Denali fault at an average rate of 290 years.
A -38
TABLE A-1
Woodward-Clyde Consultants
RELATIVE AGE DATA IN THE TAL~EETNA MOUNTAINS AND ALASKA RANGE
Granite Weathering
Study Station Ratio Depth of Interpreted
Area Number1 %Partially Oxidation Age2
%Fresh Weathered %Weathered {inches[
A. TALKEETNA MOUNTAINS3
Black BR-1 54 38 8 11 to 14 Late
River BR-2 52 40 8 11 to 13 Wisconsin
BR-3 50 42 8 15 to 17
BR-4 42 32 26 21 to 22 Early
Wisconsin
Clear CL-2 ND4 ND ND 15 to 24 Late
Va 11 ey CL-4 ND ND ND 11 to 12 Wi scans in
CL-6 ND ND ND 13 to 14
CL-8 ND ND ND 26 to 30 Early
CL-10 ND ND ND 29 Wisconsin
CL-12 64 30 6 35 to 40
Butte BL-25 50 32 18 11 to 16 Late
Lake BL-3 62 30 8 8 to 10 Wisconsin
BL-4 58 32 10 10 to 12
BL-5 50 38 12 10 to 12
BL-6 53 35 12 10 to 12
BL-7 39 37 24 14 to 17 Early
BL-8 20 56 24 29 Wi scans in
Deadman DC-1 46 40 14 10 Late
Creek DC-2 52 38 10 7 to 9 Wisconsin
DC-3 50 40 10 8.5 to 12
DC-4 52 38 10 11 to 12
DC-5 40 40 20 17 .5 to 18 Early
Wi scans in
B. ALASKA RANGE6
Stade 38 60 2 6 to 13 10,500 to
MP-IV 9,500 y.b.p.
Stade 58 42 0 15 .5 to 17 12,800 to
McKinley MP-I I I 11,800 y.b.p.
Park Stade 50 46 4 10 to 17 .5 15,000 to
Glaciation MP-II 13,500 y.b.p.
Stade 28 66 6 11 to 19 25,000 to
MP-I 17 ,000 y. b. p.
Pre-Stade 16 66
MP-I
18 17 .5 to 35 >25,000 y.b.p.
Notes:
1. The locations of moraines are shown in Figures 3-5 and 3-6.
2. Interpreted age as estimated from photogeologic interpretation (Section A.1.2.1).
3. Data obtained during this investigation.
4. ND indicates that no data were obtained at this location because the rock types
were primarily metamorphic and volcanic and because the few observed granitic
rocks were transported to this location after deposition of the moraine.
5. The surface of the moraine was slightly disturbed by man-made activities.
6. Age determinations are from Ten Brink and Waythomas
are from Werner (in press).
(in press); weathering data
Woodward-Clyde Consultants
TABLE A-2
SUMMARY OF POTASSIUM-ARGON WHOLE ROCK AGE DATES
Sample No.1 ( wt. %)
W-1 0.207
0.206
0.205
0.205
Average: 0.206
W-4 0.199
0.193
0.194
0.190
Average: 0.194
Notes:
40Ar d ra
(x1ollmo1/g)
40Ar d ra
3.333 0.386
3.293 0.422
Calculated age3
(millions of years
before present)
3.313 90.2 (+2.3)
4.535 0.622
4.538 0.654
4.536 130 (~3)
1. Sample locations are shown in Figure A-1.
2. K is potassium, Ar is argon.
3. Uncertainties are given at the 68% confidence level and reflect
analytical uncertainties of +2% in K and ~1.5% in radiogenic 40Ar.
4. Analyses were conducted by K. A. Foland, Ohio State University.
QUATERNARY I
STUDY REGION
I
PORTAGE CREEK/;-
DEVILS CANYON J
SUBREGION I
,..r-/
CLEAR VALLEY -
FOG LAKES SUBREGION
WOODWARD-CLYDE CONSULTANTS 41410A February 1982
BUTTE LAKE -
BRUSKANA
CREEK
SUBREGION
KOSINA CREEK -
BLACK RIVER
SUBREGION
C<:;pper
Brooks Range \
~ ......
·c,o'-'-\
<,;,~6' \
ALASKA Fairbanks\ CANADA .. ~
I
Gulf of Alaska
NOTE
1. Proposed reservoir configuration is after
Acres American Inc. (In Press).
2. W-1 and W-4 are potassium-argon (K-Ar)
sample localities.
QUATERNARY GEOLOGY
LOCATION MAP
0~~~53:=~1~0:JE1~5~~20 Miles
0 1 0 20 30 Kilometers
FIGURE A-1
APPENDIX B -GLOSSARY
Ab 1 at ion Till
Aleutian Megathrust
Allochthonous
Alluvium
Amygdaloidal
Andesite
Anelasticity
Aseismic
Loosely consolidated rock debris, formerly
contained by a glacier, that accumulated in
place as the surface ice was removed by
ablation.
The major collision boundary between the
Pacific and North American plates where the
Pacific plate is descending into the earth•s
mantle.
Formed or occurring elsewhere than in place;
of foreign origin or introduced.
A general term for all sediment deposited in
land environments by streams.
Gas cavities in igneous rocks that have been
filled with secondary minerals such as
quartz, calcite, chalcedony, or zeolite.
A type of volcanic rock that is believed to
be associated with plate subduction.
The effect of attenuation of a seismic wave;
it is symbolized by Q.
An area of generally 1 ow sei smi city that can
have tectonic deformation which is not
accompanied by earthquakes.
B - 1
Autochthonous
Bathe 1 ith
Benioff zone
Candidate Feature
Candidate Significant
Feature
Woodward·C&ydle Co~rnsult:canrts
Formed or occurring in the place where found;
used for a rock or sediment derived in the
place where it is now found.
A large, generally discordant mass of
igneous rock which was intruded originally at
depth and now has more than 40 square miles
(104 km2) in surface exposure. It is
composed predominantly of medi urn to coarse
grained rocks, often of granodiorite com-
position.
Seismicity associated with plates of the
earth's crust which are sinking into the
upper mantle. In Alaska, the Benioff zone is
associated with underthrusting of the Pacific
plate beneath the North American plate.
A term used in this study to identify faults
and lineaments that may affect Project design
considerations. Candidate features were
selected by applying length-distance screen-
ing criteria prior to field reconnaissance
studies.
A term used in this study to identify faults
and lineaments that may affect Project design
considerations based on length-distance
screening criteria and a preliminary assess-
ment of seismic source potentia 1 and poten-
tial surface rupture through either site
using the results of the field reconnaissance
studies.
B -2
Catacl asti c
Cirque
Cons an gu i neou s
Crag and Tail
Cryoturbat ion
Dextral Fault
Drift
Woodward-Clyde Corilsu!tants
The granular fragmental texture induced in
rocks by mechanical crushing.
A steep-walled, flat-or gently-floored,
half-bowl-like recess or hollow, situated
high on the side of a mountain and produced
by erosive activity of mountain glaciers; it
is commonly found at the head of a glacial
valley.
The relationship that exists between igneous
rocks that are presumably derived from the
same parent magma.
An elongate hill or ridge resulting from
glaciation. The crag is a steep face or knob
of ice-smoothed, resistant bedrock at the end
of the ridge from which glacial ice came.
The tail is a tapering, streamlined, gentle
slope of intact weaker rock and/or till that
was protected in part from the glacial ice by
the crag.
A term to describe the stirring, churning,
modification. and all other disturbances of
soil resulting from frost action.
A strike-slip fault along which, in plan
view, the side opposite the observer appears
to have moved to the right.
All rock material transported by a glacier
and deposited directly by the ice or by
meltwater from the glacier.
B -3
Ductile
Dynamometamorphism
End Moraine
Fault
Fault with Recent
Displacement
Flysch
Geosyncline
Woodward-Clyde Corilsultants
A rock that is able to sustain, under a given
set of conditions, 5 to 10 percent deforma-
tion before fracturing or faulting.
The alteration of rock characteristics
primarily by mech ani cal energy (pressure and
movement).
A ridge of glacial sediments deposited at the
margins of an actively flowing glacier.
A surface or zone of closely spaced fractures
along which materials on one side have been
displaced with respect to those on the other
side.
As defined for this study, a fault that has
been subject to displacement within approxi-
mately the last 100,000 years.
A thick and extensive deposit largely of
sandstone that is formed in a marine environ-
ment (geosyncline) adjacent to a rising
mountain belt.
A mobile downwarping of the crust of the
earth, either elongate or basin-like,
that is subsiding as sedimentary and volcanic
rocks accumulate to thicknesses of thousands
of meters. Geosynclines are usually measured
in scores of kilometers.
B - 4
G 1 aci a 1 Err at i c
Glacial Scour
Glaciogenic
Gouge
Hypocenter
Ice Pulses
I nterc a 1 at ed
Kame
Kettle
A rock fragment carried by glacier ice or
floating ice, and deposited when the ice
melted at some distance from the outcrop from
which the fragment was derived.
The eroding action of a glacier.
Oepos its or topographic forms derived from
a glacial origin.
Soft clayey material often present between a
vein and a wall or along a fault.
That point within the earth that is the
center of an earthquake and the origin of its
elastic waves.
Small advances of ice associated with a minor
decrease in temperature during interglacial
periods.
A material that exists as a layer or layers
between layers or beds of other rock; inter-
stratified.
A short ridge, hi 11 , or mound of poor 1 y
stratified sediments deposited by glacial
meltwater.
A steep-sided, usually basin-or bowl-shaped
hole or depression without surface drainage
in glacial deposits.
B -5
K 1 i ppe
Lee
Lineament
Lit-par-lit
Lodgment Till
Magnitude
Woodward· Clyde Consultants
An outlying isolated remnant of an overthrust
rock mass.
The side of a hill, knob, or prominent rock
facing away from the direction from which an
advancing glacier or ice sheet moved; facing
the downstream side of a glacier.
A linear trend with implied structural
control (including but not limited to frac-
tures and faults) typically identified on
remotely sensed data.
Having the characteristic of a layered rock,
the layers of which have been penetrated by
numerous thin, roughly parallel sheets of
igneous material.
A basal till commonly characterized by
compact fissile structure and containing
stones oriented with their long axes gen-
erally parallel to the direction of ice
movement.
Magnitude is used to measure the size of
instrumentally recorded earthquakes.
Several magnitude scales are in common usage
(Richter, 1958). The differences in these
magnitudes are caused by the way in which
they are each calculated, specifically, by
the periods (frequency) of the waves which
are used in each measurement. ML is the
B - 6
Metabasalt
Mi croearthquake
Mi gmat ite
Miogeosyncline
Modified Mercalli Scale
Woodward-Clyde Consultants
original Richter magnitude which was devel-
oped for Southern California earthquakes
recorded on Wood-Anderson seismometers (free
period 0.8 second) at distances of 372 miles
(600 km) or less. Ms and mb use signals
recorded at teleseismic distances of 1,240
miles (2,000 km) or greater. Ms measures
the amp 1 itude of surf ace waves with periods
of 20 seconds and the mb is a measure of
the 1 second body waves. Mw depends on the
seismic moment according to the relation log
W0 = 1.5 Mw + 11.8. The variations in
the magnitude calculations result in part
because different size earthquakes generate
relatively different amounts of energy in
these frequency bands.
Volcanic rock (basalt) altered by temperature
and pressure to a metamorphic rock.
An earthquake having a magnitude (~1L} of
three or less on the Richter scale; it is
generally not f e l,t.
A rock (gneiss) produced by the injection of
igneous material between the laminae of a
schistose formation.
A geosyncline in which volcanism is not
associated with sedimentation.
An earthquake intensity scale, having twelve
divisions ranging from I (not felt by people)
to XII (damage nearly total).
B - 7
Morphostratigraphic
Noncomf ormi ty
Nonsorted Stripes
Normal Fault
Periglacial
Pluton
Woodward· Clyde Consultants
A distinct stratigraphic unit comprising a
body of rock that is identified primarily
from the surface form it displays.
A substantial hiatus in the geologic record
that typically implies uplift and erosion.
The gap occurs between older igneous or
metamorphic rocks and younger sedimentary
rocks.
Alternating bands comprising a form of
patterned ground characterized by a striped
pattern and nonsorted appearance due to
parallel lines of frost-shattered rubble and
intervening strips of relatively bare or
vegetated ground.
A fault along which the upper (hanging) wall
has moved down relative to the lower wall
(footwall).
Describing processes, conditions, areas,
climate, and topographic features at the
i mm e d i at e mar g i n s of former or ex i s t i n g
glaciers; influenced by the cold temperature
of the ice.
An igneous intrusion formed at great depth.
8 -8
Pot ass i urn-Argon
Age Dating
Pyroclastic
Rejuvenation
Reservoir-Induced
Seismicity
Reverse Fault
Seismic Moment
Significant Feature
Woodward·CI}!de Consultants
A type of age dating that is based on the
decay rate of potassium-40 (an isotope of
potassium) to argon-40. It is useful for
dating rocks rich in potassium and is
applicable for dating materials that are
generally older than 100,000 year.
Farmed by fragmentation as a result of a
volcanic explosion or aerial expulsion from a
vo 1 can i c vent.
Renewed downcutting by a stream caused by
regional uplift or a drop in sea level.
The phenomenon of earth movement and resul-
tant seismicity that has a temporal and
spatial relationship to a reservoir and is
triggered by nontectonic stress.
A fault in which the upper (hanging) wall
appears to have moved up relative to the
lower wall (footwall).
Measurement of the 11 S i ze 11 of an earthquake
equal to the rupture area times the average
displacement times the shear modulus.
A term used in this study to identify the
faults and 1 i neaments that are considered to
have a potential affect on Project design
considerations pending additional studies.
Selection of these features was made on the
B - 9
S 1 i ckens ides
Solifluction
Stade
Stoss
Stoss and Lee
Topography
Woodward· Clyde Consultants
basis of length-distance screening criteria
and final assessment of their seismic source
potential and potential for surface rupture
through either site using the results of the
field reconnaissance studies.
A polished and smoothly striated surface that
results from friction during movement along a
fault plane.
The slow (0.2 to 2 inches/yr [0.5 to 5
cm/yr]) creeping of wet soi 1 and other
saturated fragmental material down a slope,
especially the flow initiated by frost action
and augmented by meltwater from alternate
freezing and thawing of snow and ground
ice.
A substage of a glacial stage; time repre-
sented by glacial deposits.
The side or slope of a hi 11, knob, or
prominent rock facing the direction from
which an advancing glacier or ice sheet
moved; facing the upstream side of a glacier.
An arrangement, in a strongly glaciated area,
of small hills or prominent rocks having
gentle slopes on the stoss side, and somewhat
steeper, plucked slopes on the lee side.
B -10
Stratovolcano
Time stratigraphic
Whaleback
Woodward· Clyde Consultants
A volcano composed of explosively erupted
cinders and ash interbedded with occasional
lava flows.
A subdivision of rocks considered solely
as the record of a specific interval of
geologic time.
A small hill or hillock with a curved surface
that is the result of abrasion by glacial
ice. The resulting hillock has a smooth form
resembling a whale•s back.
B -11
APPENDIX C -REFERENCES
Abe, K., 1972, Mechanisms and tectonic implications of the 1966 and 1970
Peru earthquakes: Physics of the Earth and Planetary Interiors 5,
367-379.
1974, Seismic displacement and ground motion near a fault; The
S aitama earthquake of September 21, 1931: Journ a 1 of Geophys i ca 1
Research, v. 79, p. 4393-4399.
1975, Static and dynamic fault parameters of the Saitama earthquake
of July 1, 1968: Tectonophysics, v. 27, p. 223-228.
1977, Tectonic implications of the large Shioya-Oki earthquakes of
1938: Tectonophysics, v. 41, no. 4, p. 269-291.
Abe, K., and Kanamori, H., 1979, Temporal variation of the activity
of intermediate and deep focus earthquakes: Journal of Geophysical
Research, v. 84, p. 3589-3595.
1980, Magnitudes of great shallow earthquakes from 1953 to 1977:
Tectonophysics, v. 62, p. 191-203.
Acres American Inc., 1980, Design transmittal--Initial version--Prelimi-
nary licensing documentation: Alaska Power Authority, Anchorage,
Alaska Task, 10.2, 60 p., 3 appendices.
1982, Susitna Hydroelectric Project Feasibility Report: Prepared
by Acres American Inc. for the Alaska Power Authority [in press].
Adams, C. E., Barnett, M. A. F., and Hayes, R. C., 1933, Seismological
report of the Hawke•s Bay earthquake of 3rd February, 1931: The
New Zealand Journal of Science and Technology, v. XV, no. 1,
p. 93-107.
Agnew, J. D., 1980, Seismicity of the Central Alaska Range, Alaska,
1904-1978: Master•s Thesis, University of Alaska, Fairbanks,
95 p.
Ambraseys, N. N., 1965, An earthquake engineering study of the Buyin-
Zahra earthquake of September 1st, 1962: Proceedings of the
3rd World Conference on Earthquake Engineering, New Zealand,
v. I II.
Ando, M., 1974, Faulting in the Mikawa earthquake of 1945: Tectono-
physics, v. 22, p. 173-186.
c -1
Barnes, F. F., and Payne~ T. G., 1956, The Wishbone Hill District,
Matanuska Coal Field. Alaska: U.S. Geological Survey Bulletin
1016' 85 p.
Beikman, H. M., 1974a, Preliminary geologic map of the southwest
quadrant of Alaska: U.S. Geological Survey Miscellaneous Field
Studies Map MF-611, scale 1:2,000,000, 2 sheets.
compiler, 1974b, Preliminary geologic map of the southeast quadrant
of Alaska: U.S. Geological Survey Miscellaneous Field Studies Map
MF-612, scale 1:1,000,000, 1 sheet.
1978, Preliminary geologic map of Alaska: U.S. Geological Survey,
scale 1:2,500,000, 2 sheets.
1980, Geologic Map of Alaska: U.S. Geological Survey, 2 sheets,
scale 1:2,500,000.
Biswas, N. N., and Bhattacharya, B., 1974, Travel-time relations for the
upper mantle P-wave phases from central Alaskan data: Bulletin of
the Seismological Society of America, v. 64, p. 1953-1965.
Bolt, B. A., McEvilly, T. V., and Uhrammer, R. A., 1981, The Livermore
Valley, California, sequence of January 1980: Bulletin of the
Seismological Society of America, v. 71, p. 451-463.
Bonilla, M. G., 1979, Historic surface faulting--map patterns, relation
to subsurface faulting, and relation to preexisting faults, in
Evernden, J. F., convener, Proceedings of Conference VIII--AnalysTS
of actual fault zones in bedrock: U.S. Geological Survey Open-File
Report 79-1239, p. 36-65.
Bonilla, M. G., and Buchanan, J. M., 1970, Interim report on worldwide
historic surface faulting: U.S. Geological Survey Open-File
Report, 31 p.
Boore, D. M., and Stierman, D. J., 1976, Source parameters of the
Pt. Mugu, California, earthquake of February 21, 1973: Bulletin
of the Seismological Society of America, v. 66, p. 385-404.
Boucher, G. C., and Fitch, T. J., 1969, Microearthquake seismicity
of the Denali fault: Journal of Geophysical Research, v. 74,
p. 6638-6648.
Bowers, P. M., 1979, The Cantwell ash bed, a Holocene tephua in the
central Alaska Range: Alaska Division of Geological and Geophy-
sical Surveys, Geologic Report 61, p. 19-24.
c -2
Bruen, M., 1981, Personal
Susitna Hydroelectric
Alaska.
corm1unication,
Project, Acres
Woodward·CDyde Consll.dtants
Project geologist for the
American Inc., Anchorage,
Bruhn, R. L., 1979, Holocene displacement measured by trenching the
Castle Mountain fault near Houston, Alaska: Alaska Division of
Geological and Geophysical Surveys, Geology Department 61, p. 1-4.
Burford, R. 0., and Harsh, P. W., 1981, 7.3 earthquake in Algeria
reviewed: Geotimes, v. 26, no. 5, p. 18-20.
Cady, W. M., Wallace, R. E, Hoare, T. M., and Webber, E. J., 1955,
The central Kuskakwim region, Alaska: U.S. Geological Survey
Professional Paper 268, 132 p.
Cagnetti, V., and Pasquale, V., 1979, The earthquake sequence in
Friuli, Italy, 1976: Bulletin of the Seismological Society of
America, v. 69, p. 1797-1818.
Carder, D. S., 1945, Seismic investigations of the Boulder Dam area,
1940-1944, and the influence of reservoir loading on earthquake
activity: Bulletin of the Seismological Society of America, v. 35,
p. 175-192.
Cardwell, R. K., and !sacks, B. L., 1978, Geometry of the subducted
lithosphere beneath the Banda Sea in eastern Indonesia from
seismicity and fault plane solutions: Journal of Geophysical
Research, v. 83~ p. 2825-2838.
Cardwell, R. K., !sacks, B. L., and Karig, D. E., 1980, The spacial
distribution of earthquakes, focal mechanism solutions, and sub-
ducted lithosphere in the Philippine and Northeastern Indonesia
Islands, in The tectonic and geologic evolution of southeast
asian seas-and island: American Geophysical Union, Geophysical
Monograph 23 (Hayes ed.), Washington, D.C., p. 1-35.
Castle, R., Church, J. P., Elliott, M. R., and Savage, J. C., 1977,
Preseismic and coseismic elevation changes in the epicentral
region of the Point Mugu earthquake of February 21, 1973:
Bulletin of the Seismological Society of America, v. 67, p. 219-
231.
Chapin, T., 1918, The Nelchina-Susitna region, Alaska: U.S. Geological
Survey Bulletin 668, 67 p.
Chase, C. G., 1980, University of Minnesota, Minneapolis, personal
corm1unication.
c -3
Cipar, J., 1980, Teleseismic observations of the 1976 Friuli, Italy
earthquake sequence: Bulletin of the Seismological Society of
America, v. 70, p. 963-983.
1981, Broadband time domain modeling of earthquakes from Friuli,
Italy: Bulletin of the Seismological Society of America, v.
71, p. 1215-1231.
Coulter, H. W., Hopkins, D. M., Karlstrom, T. N. V., Pewe, T. L.,
Wahrhaftig, C., and Williams, J. R., 1965, Map showing extent of
glaciations in Alaska: U.S. Geological Survey Miscellaneous
Geologic Investigations Map I-415, scale 1:2,500,000.
Compton, R. R., 1962, Manual of field geology: New York, John Wiley and
Sons, 378 p.
Crouse, C. B., and Turner, B. E., 1980, Processing and analysis of
Japanese accelerograms and comparisons with U.S. strong motion
data: Seventh World Conference on Earthquake Engineering, Istan-
bul, Turkey, v. 2, p. 419-426.
Csejtey, B., Jr., 1980, U.S. Geological Survey, Menlo Park, California,
personal communication.
1981, U.S. Geological Survey, Menlo Park, California, personal
communication.
Csejtey, B., and Griscom, A., 1978, Preliminary aeromagnetic inter-
pretative map of the Talkeetna Mountains quadrangle, Alaska:
U.S. Geological Survey Open File Report 78-558-C, scale 1:250,000,
14 p., 2 plates.
Csejtey, B., Jr., Nelson, W. H., Jones, D. L., Silberling, N.J.,
Dean, R. M., Morris, M. S, Lanphere, M.A., Smith, J. G., and
Silbermen, M. L., 1978, Reconnaissance geologic map and geochron-
ology, Talkeetna ~~ountain Quadrangle, northern part of Anchorage
Quadrangle, and southwest corner of Healy Quadrangle, Alaska: U.S.
Geological Survey Open-File Report 78-558-A, 62 p.
Csejtey, B., and St. Aubin, D. R., 1981, Evidence for northwestward
thrusting of the Talkeetna superterrane and its regional signifi-
cance: U.S. Geological Survey Circular 823-B, p. B49-B51.
Csejtey, B., Stricker, G. D., Mullen, M. W., and Cox, D.P., 1980, The
Denali fault of southern Alaska: The case for minor rather than
major displacement (abs.): American Geophysical Institute (EOS)
Transactions, v. 61, p. 1114.
c -4
Dames and Moore, 1975, Subsurface geophysical exploration--proposed
Watana damsite on the Susitna River: U.S. Army Corps of Engi-
neers, Alaska District, unpublished report, 12 p., 2 appendices.
Davies, J. N., 1975, Seismological investigations and plate tectonics in
southcentral Alaska: Ph.D. dissertation, University of Alaska,
Fairbanks, 192 p.
Davies, J. N., and House, L., 1979, Aleutian subduction zone seismicity,
volcano-trench separation, and their relation to the great thrust-
type earthquakes: Journal of Geophysical Research, v. 84, p.
4583-4591.
Davies, J., Sykes, L., House, L., and Jacob, K., 1981, Shumagin seismic
gap, Alaska peninsula: history of great earthquakes, tectonic
setting and evidence for high seismic potential: Journal of
Geophysical Research, v. 86, p. 3821-3855.
Davis, T. N., 1960, A field report on the Alaska earthquakes of
April 7, 1958: Bulletin of the Seismological Society of America,
v. 50, p. 489-510.
1964, Seismic history of Alaska and the Aleutian Islands: Biblio-
graphical Bulletin of American Geophysics and Oceanography, v. 3,
p. 1-16.
Davis, T. N., and Echols, C., 1962, A table of Alaskan earthquakes
1788-1961: University of Alaska, Geophysical Institute Research
Report No. 8, 2 p.
Detterman, R. L., Plafker, G., Hudson, T., Tysdal, R. G., and Pavoni,
N., 1974, Surface geology and Holocene breaks along the Susitna
segment of the Castle Mountain fault, Alaska: U.S. Geological
Survey Miscellaneous Field Studies Map MF-618, scale 1:24,000,
1 sheet.
Detterman, R. L., Plafker, G., Tysdal, R. G., and Hudson, T., 1976,
Geology of surface features along part of the Talkeetna segment of
the Castle Mountain-Caribou fault system, Alaska: U.S. Geological
Survey Miscellaneous Field Studies Map MF-738, scale 1:63,360,
1 sheet.
Dibblee, Jr., T. W., 1955, Geology of the southeastern margin of the
San Joaquin Valley, California, in Earthquake in Kern County
during 1952: California Division Of Mines and Geology Bulletin
17 1 ' p. 2 3-34 .
c -5
Woodward-Clyde Consultants
Dobry, R., Idriss, I. M., and Ng, E., 1978, Duration characteristics
of horizontal components of strong motion earthquake records:
Bulletin of the Seismological Society of America, v. 68, p. 1487-
1520.
Dunbar, W. S., Boore, D. M., and Thatcher, W., 1980, Pre, co and post-
seismic strain changes associated with the 1952 ML=7.2 Kern
County, California earthquake: Bulletin of the Seismological
Society of America, v. 70, p. 1893-1905.
Everingham, I. B., Gregson, P. J., and Doyle, H. A., 1969, Thrust
fault scarp in the western Australian shield: Nature, v. 223,
p. 701-703.
Ferrians, 0. J., and Schmoll, H. R., 1957, Extensive proglacial lake of
Wisconsin age in the Copper River basin, Alaska: Geological
Society of America Bulletin, v. 68, p. 1726.
Finetti, I., Pursi, M., and Slejko, 1979, The Friuli earthquake (1976-
1977): Tectonophysics, v. 53, p. 261-272.
Flint, R. F., 1971, Glacial and Quaternary geology: New York, John
Wiley and Sons, 892 p.
Forbes, R. B., Turner, D. L., Stout, J., and Smith, T. E., 1973,
Cenozoic offset along the Denali fault, Alaska (abs.): American
Geophysical Union (EOS) Transactions, v. 54, p. 495.
Foster, H. M., and Karlstrom, T. N. V., 1967, Ground breakage and
associated effects in the Cook Inlet area, Alaska, resulting
from the March 27, 1964 earthquakes: U.S. Geological Survey
Professional Paper 543-F, 28 p.
Fujita, K., and Kanamori, H., 1981, Double seismic zones and stresses
of intermediate depth earthquakes: Geophysical Journal of the
Royal Astronomical Society, v. 66, p. 131-156.
Gedney, L., Berg, E., Pulpan, H., Davies, J., and Feetham, W., 1969,
A field report on the Rampart, Alaska earthquake of October 29,
1968: Bulletin of the Seismological Society of America, v. 59,
p. 1421-1423.
Gedney, L., and Shapiro, L., 1975, Structural lineaments, seismicity and
geology of the Talkeetna Mountains area, Alaska: Unpublished
report prepared for the U.S. Army Corps of Engineers, Alaska
Division, 18 p.
c -6
Geller, R. J., and Kanamori, H., 1977, Magnitudes of great shallow
earthquakes from 1904 to 1952: Bulletin of the Seismological
Society of America, v. 67, p. 587-598.
Geller, R. J., Kanamori, rl., and Abe, K., 1978, Addenda and correc-
tions to "Magnitude of great shallow earthquakes from 1904 to
1952 11
: Bulletin of Seismological Society of America, v. 68,
p. 17 6 3-17 6 4 .
Gettrust, J. F., Hsu, V., Helsley, C. E., Herrero, E., and Jordan, T.,
1981, Pat terns of loca 1 sei smi city preceding the Pet at 1 an earth-
quake of 14 March 1979: Bulletin of the Seismological Society
of /\merica, v. 71, p. 761-769.
Grantz, A., 1966, Strike-slip faults in Alaska: U.S. Geological Survey
Open-File Report, 82 p.
Griscom, A., 1978, Aeromagnetic map and interpretation, Talkeetna
Quadrangle, Alaska: U.S. Geological Survey ~1iscellaneous Field
Studies Map MF-870B, Scale 1:250,000, 2 sheets.
1979, Aeromagnetic interpretation of the Big Delta Quadrangle,
Alaska: U.S. Geological Survey Open-File Report 78-529-B, 10 p.
Guptill, P. D., Brogan, G. E., and Turcotte, T., 1981, Tectonic setting
of the Gulf of Alaska: Marine Geotechnology, v. 4., no. 3,
p. 223-241.
Gutenberg, B., 1955, The first motion in longitudinal and transverse
waves of the main shock and the direction of slip, in Earthquakes
in Kern County during 1952: California Division-of Mines and
Geology Bulletin 171, p. 165-170.
Gutenberg, B., and Richter, C. F., 1954, Seismicity of the earth and
associated phenomena: Princeton, New Jersey, Princeton University
Press, 310 p.
Hanks, T. C., and Kanamori, H., 1979, A moment magnitude scale: Journal
of Geophysical Research, v. 84, p. 2348-2350.
Hasegawa, A., Umino, N., Takagi, A., and Suzuki, Z., 1979, Double
planed deep seismic zone and anomalous structure in the upper
mantle beneath northeastern Honshu (Japan): Tectonophysics,
v. 57, p. 1-6.
Helmberger, D. V., 1968, The crust-mantle transition in the Bering
Sea: Bulletin of the Seismological Society of America, v. 58,
p. 179-214.
c -7
Wood ward · Clyde Consultants
Hickman, R. G., and Craddock, C., 1973, Lateral offsets along the Denali
fault, central Alaska Range, Alaska (abs.): Geological Society of
America Abstracts with Programs, v. 5, p. 322.
Hickman, R. G., Craddock, C., and Sherwood, K. W., 1976, The Denali
fault system and the tectonic development of southern Alaska
(abs.): International Geological Congress Abstracts, v. 3,
p. 683.
1977, Structural geology of the Denali Fault system, central Alaska
Range: Geological Society of America Bulletin, v. 88, p. 1217-
1230.
197 8 , The Denali f au lt system and the tectonic development of
southern Al aska: Tectonophysics, v. 47, p. 247-273.
Hopkins, D. M., 1967, The Bering Land Bridge: Palo Alto, California,
Stanford University Press, 495 p.
Huang, P., 1981, Focal parameters of October 29, 1968 Rampart, Alaska
earthquake and its tectonic imp 1 i cations: Massachusetts Institute
of Technology [in press].
Idriss, I. M., 1978, Characteristics of earthquake ground motions,
state-of-the-art p ap er: Specialty Conference on Earthquake
Engineering and Soil ~yna mi cs, Geotechnical Engineering Division of
the American Society of Civi 1 Engineers, Pasadena, California,
Proceedings, v. 3, p. 1151-1263.
!ida, K., 1965, Earthquake magnitude, earthquake fault and source
dimensions: Journal of Earth Science, Nagoya University, v. 13,
p. 115-132.
Isacks, B. L., and Barazangi, M., 1977, Geometry of Benioff zones:
Latera 1 segmentation and downwards bending of the s ubducted 1 itho-
sphere, in Island areas, deep sea trenches and back-arc basins,
Ewing series, American Geophysical Union, Washington, D.C., v. 1,
p. 99-114.
Isacks, B., Oliver, J., Sykes, L. R., 1968, Seismology and the new
global tectonics: Journal of Geophysical Research, v. 73, p.
5855-5896.
Jones, D. L., and Silberling, N. J., 1979, Mesozoic stratigraphy, the
key to tectonic analysis of southern and central Alaska: U.S.
Geological Su r vey Open File Report 79-1200, 37 p.
c -8
Kachadoorian, R., and Moore, G. W., 1979, Reconnaissance of the recent
geology of the proposed Devils Canyon and Watana dam sites, Susitna
River, Alaska, in U.S. Army Corps of Engineers, Southcentral
railbelt area, Alaska-upper Susitna river basin: U.S. Army Corps
of Engineers, Alaska District, Appendix Part 1, Exhibit, D-2,
42 p.
Kanamori, H., 1977, The energy release in great earthquakes: Journal of
Geophysical Research, v. 82, p. 2981-2987.
Kanamori, H., and Anderson, D. L., 1975, Theoretical basis of some
empirical relations in seismology: Bulletin of the Seismological
Society of America, v. 65, p. 1073-1095.
Karlstrom, T. N. V., 1964, Quaternary geology of the Kenai lowland and
glacial Cook Inlet region, Alaska: U.S. Geological Survey,
Professional Paper 443, 69 p.
Karlstrom, T. N. V., Coulter, H. W., Jernald, A. T., Williams, J. R.,
Hopkins, D. M., Pewe, T. L., Drewes, H., Huller, E. H., and Candon,
W. H., 1964, Surficial Geology of Alaska: U.S. Geological Survey
Miscellaneous Geologic Investigations Map I-557, scale 1:1,584,00,
2 sheets.
Kawasaki, I., 1975, The focal process of the Kita-Mino earthquake
of August 19, 1961, and its relationship to a Quaternary fault,
the Hatogaya-koike fault: Journal of Physics of the Earth,
v. 24, p. 227-250.
Keith, C. M., Simpson, D. W., and Soboleva, 0. V., 1979, Induced
seismicity and deformation at Nurek reservoir, Tadjik, USSR:
Unpublished manuscript, 28 p.
Kelleher, J. A., 1972, Rupture zones of 1 arge South American earth-
quakes and some predictions: Journal of Geophysical Research,
v. 77, p. 2087-2103.
Kelleher, J. A., and Savino, J., 1975, Distribution of seismicity
before large strike-slip and thrust-type earthquakes: Journal
of Geophysical Research, v. 80, p. 260-271.
Kelleher, J. A., Savino, J., Rowlett, H., and McCann, W., 1974, Why
and where great thrust earthquakes occur along island arcs:
Journal of Geophysical Research, v. 79, p. 4889-4899.
Kelley, T. E., 1963, Geology and hydrocarbons in Cook Inlet Basin,
Alaska, in Childs, D. E., and Beebe, B. W., eds., Backbone of the
Americas-s-ymposium: American Association of Petroleum Geologists
Memoir 2, p. 278-296.
c -9
Kuno, H., 1936, On the displacement of the Tanna fault since Pleisto-
cene: Bulletin of the Earthquake Research Institute, University
of Tokyo, v. 14, p. 619-631.
Lahr, J., 1975, Detailed seismic investigation of Pacific-North America
plate interaction in southern Alaska: Ph.D. thesis, Columbia
University, New York, 88 p.
Lahr, J. C., and Plafker, G., 1980, Holocene Pacific-North American
plate interaction in southern Alaska: Implications for the
Yakataga seismic gap: Geology, v. 8, p. 483-486.
Lander, J., ed. 1969, Seismological notes: Bulletin of the Seismo-
logical Society of America, v. 59, p. 1023-1030.
----, ed., 1970, Seismological notes: Bulletin of the Seismological
Society of America, v. 60, p. 685-693.
Langston, C. A., 1978, The February 9, 1971, San Fernando earthquake:
a study of source finiteness in teleseismic body waves: Bulletin
of the Seismological Society of America, v. 68, p. 1-30.
Langston, C. A., and Helmberger, D. V., 1975, A procedure for modeling
shallow dislocation sources: Geophysical Journal of the Royal
Astrological Society, v. 42, p. 117-130.
Lee, W., Herd, D. G., Cagnetti, V., Bakun, W., and Rapport, A., 1979,
A preliminary study of the Coyote Lake earthquake of August 6,
1979 and its major aftershocks: U.S. Geological Survey Open
File Report 79-1621, 43 p.
Matsuda, T., 1977, Estimation of future destructive earthquakes from
active faults on land in Japan: Journal of Physics of the Earth,
v. 25, Supplement, p. S251-S260.
McCann, W., 1981, Research Associate, Lamont-Doherty Geological Obser-
vatory, Palisades, New York, personal communication.
McCulloch, D. S., and Bonilla, M.G., 1970, Effects of the earthquake
of March 27, 1967, on the Alaska Railroad: U.S. Geological Survey
Professional Paper 545-D, 161 p.
Meyers, H., 1976, A historical summary of earthquake epicenters in and
near Alaska: National Geophysical and Solar-Terrestrial Data
Center, Boulder, Colorado, NOAA Technical Memorandum EDS NGSDC-1,
57 p., 7 appendices.
c -10
Miller, R. D., and Dobrovolry, E., 1959, Surficial geology of Anchorage
and vicinity, Alaska, U.S. Geological Survey Bulletin 1093,
128 p.
Mogi, K., 1969, Some features of recent seismic activity in and near
Japan, activity before and after great earthquakes: Bulletin
Earthquake Research Institute, Tokyo University, v. 47, p. 395-417.
National Oceanic and Atmospheric Administration, 1981, Hypocenter data
file (computer tape and cards): Boulder, Colorado, Environmental
Data Services.
Nokleberg, W., 1981, U.S. Geological Survey, Menlo Park, California,
personal communication.
Okada, A., and Ando, M., 1979, Active faults and earthquakes in Japan:
Science (Japanese publication), v. 49, p. 158-169 (in Japanese).
Packer, D. R., Brogan, G. E., and Stone, D. B., 1975, New data on plate
tectonics of Alaska: Tectonophysics, v. 29, p. 87-102.
Packer D. R., Cluff, L. S., Knuepfer, P. L., and Withers, R. J., 1979,
Study of reservoir induced seismicity: Final Technical Report to
the U.S. Geological Survey, Contract 14-08-0001-16809, 222 p.
Packer, D. R., Lovegreen, J. R., and Born, J. L., 1977, Earthquake eval-
uation studies of the Auburn Dam area--Volume 6--Reservoir induced
seismicity: U.S. Bureau of Reclamation, Denver, Colorado, Contract
6/07/DS/72090, 124 p., 7 appendices.
Page, R. A., 1968, Aftershocks and microaftershocks of the great Alaska
earthquake of 1964: Bulletin of the Seismological Society of
America, v. 58, p. 1131-1168.
1972, Crustal deformation on the Denali fault, Alaska, 1942-1970:
Journal of Geophysical Research, v. 77, p. 1528-1533.
Page, R. A., and Lahr, J., 1971, Measurements for fault slip on the
Denali, Fairweather, and Castle r~ountain faults, Alaska: Journal
of Geophysical Research, v. 76, p. 8534-8543.
Pascal, G., 1979, Seismotectonics of the Papua New Guinea-Solomon
Islands region: Tectonophysics, v. 57, p. 7-34.
Perman, R. C., Packer, D. R., Coppersmith, K. J., and Knuepfer, P. L.,
1981, Collection of data for data bank on reservoir induced
seismicity: Report to the U.S. Geological Survey, Contract No.
14-08-0001-19132, Final Technical Report, 45 p.
c -11
Woodward-Clyde Consultants
Pettijohn, F. J., 1949, Sedimentary rocks: New York, Harper and
Brothers, 513 p.
/' Pewe, T. L., 1975, Quaternary geology of Alaska: U.S. Geological
Survey Professional Paper 835, 145 p.
Pewe, T. L., Hopkins, D. M., and Giddings, J. L., 1965, Quaternary
geology and archeology of Alaska, in Wright, H. E., Jr., and Frey,
D . G . , e d s . , T h e Q u at ern a r y of file U n it e d S t at e s : P r i n c e t on ,
New Jersey, Princeton University Press, p. 355-374.
Plafker, G., 1969, Tectonics of the March 27, 1964, Alaska earthquake:
U.S. Geological Survey Professional Paper 543-I, 74 p.
Plafker, G., Hudson, T., Bruns, T. R., Rubin, M., 1978, Late Quaternary
offsets along the Fairv.;eather fault and crustal plate interactions
in south Alaska: Canadian Journal of Earth Science, v. 15,
p. 805-816.
Plafker, G., Hudson, T., and Richter, D. H., 1977, Preliminary observa-
tions on late Cenozoic displacements along the Totschunda and
Denali fault systems, in Blean, K. ~1., ed., The United States
Geological Survey in Alaska: Organization and status of programs
in 1977: U.S. Geological Survey Circular C 751-A, p. 867-869.
R & M Associates Inc., 1981, Photointerpretation report and terrain unit
maps: Susitna Hydroelectric Project Report for Subtask 5.02
submitted to Acres American Inc., 5 sheets, 18 maps at a scale of
1:24,000.
Real, C. R., Toppozado, T. R., and Parke, D. L., 1978, Earthquake
epicenter map of California 1900-1974: California Division of
Mines and Geology t~ap, Sheet 39, scale 1:1,000,000.
Reyners, t~ .• and Coles, K. S., 1982, Fine structure of the dipping
seismic zone and subduction mechanics in the Shumagin Islands:
Journal of Geophysical Research, v. 87, p. 356-366.
Rial, J. A., 1978, The Caracas-Venezuela earthquake of July 1967: A
multiple source event: Journal of Geophysical Research, v. 83,
p. 5405-5414.
Richter, C. F., 1958, Elementary Seismology: San Francisco, California,
W. H. Freeman and Company, 768 p.
Richter, D. H., 1967, Geology of the Portage Creek -Susitna River
area: Alaska Division of Mines and Minerals Geological Report No.
3, scale 1:24,000, 2 sheets.
c -12
Woodward-C&yde ConsuBtants
Richter, D. H., and Matson, N. A., Jr., 1971, Quaternary faulting in the
eastern Alaska Range: Geological Society of America Bulletin,
v. 82, p. 1529-1539.
Rogers, G. C., and Hasegawa, H. S., 1978, A second look at the British
Columbia earthquake of June 23, 1946: Bulletin of the Seismologi-
cal Society of America, v. 58, p. 653-675.
Rojan, C., Brogan, G. E., and Slemmons, D. B., 1977, Preliminary report
on the San Juan, Argentina Earthquake of November 23, 1977:
Conference on Earthquake Engineering, San Salvador, 1978, reprint.
Ryall, A., Van vJormer, J. D., and Jones, A. E., 1968, Triggering of
microearthquakes by earth tides, and other features of the Truckee,
California, earthquake sequence of September 1966: Bulletin of
the Seismological Society of America, v. 58, p. 215-248.
St. Amand, P., 1957, Geological and geophysical synthesis of the tecton-
ics of portions of British Columbia, the Yukon Territory and
Alaska: Geological Society of America Bulletin, v. 68, p. 1343-
1370.
Schmoll, H. R., Szabo, B. J., Rubin, M., and Dobrovolny, E., 1972,
Radiometric dating of marine shells from the Bootlegger Cove Clay,
Anchorage area, Alaska: Geological Society of America, v. 83,
p. 1107-1114.
Schnabel, P. B., and Seed, H. B., 1973, Accelerations in rock for earth-
quakes in the western United States: Bulletin of the Seismological
Society of America, v. 63, p. 501-516.
Seed, H. B., 1980, University of California at Berkeley, Revisions to
Schnabel and Seed (1973), personal communication.
Seed, H. B., Murarka, R., Lysmer, J., and Idriss, I. M., 1976, Relation-
ships of maximum accelerations, maximum velocity, distance from
source and local site conditions for moderately strong earth-
quakes: Bulletin of the Seismological Society of America, v. 66,
p. 1323-1342.
Seed, H. B., Ugas, C., and Lysmer, J., 1976, Site-dependent spectra for
earthquake resistant design: Bulletin of the Seismological Society
of America, v. 66, p. 221-243.
Seno, T., 1981, International Institute of Seismologic and Earthquake
Engineering, Ministry of Construction, Tsukuba, Japan, personal
communication.
c -13
Seno, T., and Pongsawat, B., 1981, A triple-planed structure of seis-
micity and earthquake lnechanisms at the subduction zone off Miyagi
Prefecture, Northern Honshu, Japan: Earth and Planetary Science
Letters, v. 55, p. 25-36.
Seno, T., Shimazaki, K., Somerville, P., Sudo, K., and Equchi, T.,
1980, Rupture process of the Miyagi-Oki, Japan earthquake of
June 12, 1978: Physics of the Earth and Planetary Interiors,
v. 23, p. 39-61.
Shannon and Wilson, 1978, Seismic refraction survey, Susitna hydro-
electric project--Watana and Devils Canyon dam sites: Report
to U.S. Army Corps of Engineers, Alaska District, 27 p.
Sherard, J. L., Cluff, L. S., and Allen, C. R., 1974, Potentially active
faults in dam foundations: Geotechnique, v. 24, p. 367-428.
Shimazaki, K., 1975, Nemuro-oki earthquake of June 17, 1973: A litho-
spheric rebound at the upper half of the interface: Physics of the
Earth and Planetary Interiors, v. 9, 314-327.
Shimazaki, K., and Somerville, P., 1979, Static and dynamic parameters
of the Izu-Oshima, Japan earthquake of January 14, 1978: Bulletin
of the Seismological Society of America, v. 69, p. 1343-1378.
Silberling, N. J., Richter, D. H., and Jones, D. L., 1981, Recognition
of the Wrangellia and vicinity, south-central Alaska: U.S.
Geological Survey Circular 8238, p. 851-855.
Simpson, D. W., 1976, Seismicity changes associated with reservoir
loading: Engineering Geology, v. 10, p. 123-150.
Simpson, D. W., and Negnatullaiv, S. K. H., 1978, Induced seismicity
studies in south central Asia: Earthquake Information Bulletin,
v. 10, p. 209-213.
Singh, S. K., Bazan, E., and Esteva, L., 1980, Expected earthquake
magnitude from a fault: Bulletin of the Seismological Society
of America, v. 70, p. 903-914.
Slauson, W. F., and Savage, J. C., 1979, Geodetic deformation associated
with the 1946 Vancouver Island, Canada earthquake: Bulletin
of the Seismological Society of America, v. 69, p. 1487-1496.
Slemmons, D. B., 1977a, Faults and earthquake magnitude: U.S. Army
Corps of Engineers, Waterways Experiment Station, Vicksburg,
Mississippi, Miscellaneous Paper S-73-1, Report 6, 129 p.
c -14
1977b, State-of-the-art for assessing earthquake hazards in the
United States; Part 6: Faults and earthquake magnitude with
appendix on geomorphic features of active fault zones: U.S. Army
Engineer Waterways Experiment Station, Vicksburg, Contract No. DACW
39-76-C-0009, 12 p., plus 37 p. appendix.
Smith, T. E., 1973, Alaska Department of Geology and Geophysical
Surveys, Fairbanks, Alaska, unpublished geologic maps, scale
1:24,000.
1980, University of Alaska, Fairbanks, Alaska, personal communi-
cation.
----1981, Alaska Department of Geology and Geophysical Surveys,
Fairbanks, Alaska, personal communication.
Snow, D. T., 1972, Geodynamics of seismic reservoirs: International
Association of Engineering Geologists, Symposium on Percola-
tion through Fissured Rocks, Stuttgart, Germany, Proceedings,
p. TJ1-TJ19.
Stauder, W., 1968, Mechanism of the Rat Island earthquake sequence
of February 4, 1965, with relation to island arcs and sea-floor
spreading: Journal of Geophysical Research, v. 73, p. 3847-3858.
1975, Subduction of the Nazca plate under Peru as evidenced by
focal mechanisms and by seismicity: Journal of Geophysical
Research, v. 80, p. 1053-1064.
Stout, J. H., Brady, J. B., Weber, F., and Page, R. A., 1973, Evidence
for Quaternary movement on the McKinley strand of the Denali fault
in the Delta River area, Alaska: Geological Society of America
Bulletin, v. 84, p. 939-947.
Stout, J. H., and Chase, C. G., 1980, Plate kinematics of the Denali
fault system: Canadian Journal of Earth Sciences, v. 17, p. 1527-
1537.
Sudo, K., 1972, The focal process of the Taiwan-Oki earthquake of March
12, 1966: Journal of Physics of the Earth, v. 20, p. 147-164.
Sykes, L. R., 1971, Aftershock zones of great earthquakes, seismicity
gaps, and earthquake prediction for Alaska and the Aleutians:
Journal of Geophysical Research, v. 76, p. 8021-8041.
1981, Lamont-Doherty Geological Observatory, Palisades, New York,
written communication to Acres included as part of the Review of
Woodward-Clyde Consultants Interim Report on Seismic Studies--Final
Draft dated 23 February 1981.
c -15
Sykes, L. R., and Quittmeyer, R. C., 1981, Repeat times of great
earthquakes along simple plate boundaries: in Simpson, D. W., and
Richards, P. G., eds., 1981, Maurice Ewing Srnes No.4, Earthquake
prediction and international review: Washington, D. C., American
Geophysical Union, p. 217-246.
Ten Brink, N. W., and Ritter, D. F., 1980, Glacial chronology of the
north-central Alaska Range and implications for discovery of
early-man sites (abs.): Geological Society of America, Abstracts
with Programs, v. 12, p. 534.
Ten Brink, N. W., and 1-Jaythomas, C. F., 1982, Late Wisconsin glacial
chronology of the north-central Alaska Range--A regional synthesis
and its implications for early human settlements: Accepted for
publication in the Geological Society of America Bulletin, 27 p.
Lin press].
Terrestrial Environmental Specialists, Inc., 1981, Environmental studies
annual report 1980 Subtask 7 .06, Cultural resources investigation:
Report prepared for Acres American Inc., Buffalo, New York and
the Alaska Power Authority, Anchorage, Alaska, by Terrestrial
Environmental Specialists, Inc., Phoenix, New York, 257 p.
Thatcher, W., and Rundle, J. B., 1979, A model for the earthquake
cycle in underthrust zones: Journal of Geophysical Research,
v. 84, p. 5540-5556.
Thorson, R. M., and Hamilton, T. D., 1977, Geology of the Dry Creek
site: A stratified early man site in interior Alaska: Quaternary
Research, v. 7, p. 149-176.
Tobin, D. G., and Sykes, L. R., 1966, Relationship of hypocenters of
earthquakes to the geology of Alaska: Journal of Geophysical
Research, v. 71, p. 1659-1667.
Toppozada, T. R., Real, C. R., and Pierzinski, D. C., 1979, Seismicity
of California January 1975 through March 1979: California Geology,
v. 7. p. 139-142.
Trainer, F. W., and Waller, R. M., 1965, Subsurface stratigraphy of
glacial drift at Anchorage, Alaska: U.S. Geological Survey
Professional Paper 525-0, p. D167-D174.
Tsai, Y., and Aki, K., 1970, Source mechanism of the Truckee, Cali-
fornia, earthquake of September 12, 1966: Bulletin of the Seis-
mological Society of America: v. 60, p. 1199-1208.
c -16
Woodward· Clyde Consultants
Turner, D. L., and Smith, T. E., 1974, Geochronology and generalized
geology of the central Alaska Range, Clearwater Mountains, and
northern Talkeetna Mountains: Alaska Division of Geological and
Geophysical Surveys Open-File Report 72, 11 p.
Turner, D. L., Smith, T. E .• and Forbes, R. B., 1974, Geochronology of
offset along the Denali fault system in Alaska (abs.): Geological
Society of America Abstracts with Programs, v. 6, p. 268-269.
Tysdel, R. G., 1976, A preliminary evaluation of selected earthquake-
related geologic hazards in the Kenai Lowland, Alaska: U.S.
Geological Survey Open File Report 76-270, 30 p.
U. S. Army Corps of Engineers, undated, Upper Susitna River basin-Watana
reservoir, surficial geology: U.S. Army Corps of Engineers,
Alaska District, scale 1:63,360, Plate 5.
1979, Southcentral railbelt area, Alaska, Upper Susitna River
basin, Hydroelectric power supplemental feasibility report:
U.S. Army Corps of Engineers, Alaska District, main report and
appendices 1 and 2.
U.S. Bureau of Reclamation, 1960, Devil Canyon project, Alaska,
feasibility report: Bureau of Reclamation, Alaska District,
99 p.
U. S. Nuclear Regulatory Commission, 1981, Safety evaluation report
related to the operation of San Onofre Nuclear Generating Station,
Units 2 and 3: U.S. Nuclear Regulatory Commission, vJashington,
D.C., NUREG-0712, Supplement No. 1, p. E-16 and E-17.
Urhammer, R. A., 1980, Observations of the Coyote Lake, California
earthquake sequence of August 6, 1979: Bulletin of the Seismo-
logical Society of America, v. 70, p. 559-570.
Van Eysinga, F. W. D., 1978, Geological timetabie: New York, Elsevier
Scientific Publishing Company, 1 sheet.
Vieth, K. F., 1974, The relationship of island arc seismicity to plate
tectonics: Ph.D. thesis, Southern Methodist University, Dallas,
Texas.
Wahrhaftig, C. A., 1958, Quaternary geology of the Nenana River valley
and adjacent parts of the Alaska Range: U.S. Geological Survey
Professional Paper 293, p. 1-70.
c -17
Woodward-Clyde Consultants
Wahrhaftig, C. A., Turner, D. L., Weber, F. R., and Smith, T. E., 1975,
Nature and timing of movement on Hines Creek strand of Denali fault
system, Alaska: Geology, v. 3, p. 463-466.
Weber, F. R., Hamilton, T. D., Hopkin, D. M., Repenning, C. A., and
Haas, H., 1980, Canyon Creek: A Late Pleistocene vertebrate
locality in interior Alaska: Quaternary Research, v. 16, no. 2,
p. 167-180.
Werner, A., 1982, Glacial Geology of the McKinley River area, Alaska:
with an evaluation of various relative age dating techniques:
Master's thesis, Southern Illinois University at Carbondale,
Illinois, 147 p. [in press].
Withers, R. J., 1977, Seismicity and stress determination at man-
made lakes: Ph.D. Dissertation, University of Alberta, Canada,
241 p.
vJithers, R. J., and Nyland, E., 1978, Time evo 1 uti on of stress under an
artificial lake and its implication for induced seismicity:
Canadian Journal of Earth Sciences, v. 15, p. 1526-1534.
Wood, H. 0., and Neumann, F., 1931, Modified Mercalli Intensity Scale of
1931: Bulletin of the Seismological Society of America, v. 21,
p. 277-283.
Woodward-Clyde Consultants, 1975, Age dating of geologic materials--A
survey of techniques: San Francisco, Professional Development
Committee, Woodward-Clyde Consultants, 183 p.
----1978, Offshore Alaska Seismic Exposure Study, Volumes I-VI:
Unpublished report prepared for the Alaska Subarctic Offshore
Committee, Exxon Corporation, Houston, Texas, Project Admini-
strator, by Woodward-Clyde Consultants, Orange, California.
----1980a, Final Report--Susitna Hydroelectric Project--Seismic
Refraction Survey--Summer, 1980: Unpub 1 i shed report prepared for
R & M Associates, Anchorage, Alaska, 17 p., 1 appendix.
1980b, Interim Report on Seismic Studies for Susitna Hydroelectric
Project: Report prepared for Acres American Inc., Buffalo,
New York, by Woodward-Clyde Consultants, Orange, California, 225 p,
8 Appendices.
1980c, Preliminary Seismic Hazards Study for the Ituango Project;
Colombia--Phase I: Report submitted to Integral, LTDA, Medellin,
Colombia, 7 Sections, 2 Appendices, 15 plates.
c -18
1982, Susitna Hydroelectric Project Seismic Refraction Surveys
1981: Unpublished report prepared for R & M Consultants, Anchor-
age, Alaska, by Woodward-Clyde Consultants, Santa Ana, California,
59 p .• 3 Appendices.
Wyss, M., 1979, Estimating maximum expectable magnitude of earthquake
from fault dimensions: Geology, v. 7, p. 336-340.
Yerkes, R. F., 1973, Effects of the San Fernando earthquake as related
to geology~ San Fernando, California earthquake of February
9, 1971: National Oceanographic and Atmospheric Administration,
Washington, D.C.
Yoshii, T., 1975, Proposal of the 11 aseismic front": Zisin 2, v. 28,
p. 365-367 (in Japanese).
1979, A detailed cross-section of the deep, seismic zone beneath
northeastern Honshu, Japan: Tectonophysics, v. 55, p. 349-360.
c -19