Third International Conference on Port and Ocean ... - Poac.com

PROCEEDINGS OF THE
HELD IN FAIRBANKS, ALASKA
UNIVERSITY OF ALASKA
11-15 AUGUST 1975
THIRD INTERNATIONAL CONFERENCE ON
PORT AND OCEAN ENGINEERING
UNDER ARCTIC CONDITIONS
VOLUME I
INSTITUTE OF MARINE SCIENCE
D.C. BURRELL & D.W. HOOD
CONFERENCE CONVENERS

Copyright © 1976 by publisher: Institute of Marine Science
University of Alaska
Fairbanks, Alaska 99701
All rights reserved.
ISSN: 0376-6756:
Printed and bound by: Ken Wray Print Shop
Anchorage, Alaska
United States

SPONSORS
American Association for the Advancement of Science, Alaska Division
Arctic Institute of North America
Canadian Society for Civil Engineering
Engineering Committee on Oceanic Resources
Marine Technology Society
Norwegian Institute of Technology: Division of Port and Ocean Engineering
u.S. Office of Naval Research
u.S. National Science Foundation
University of Alaska: Institute of Marine Science; Alaska Sea Grant Program;
Geophysical Institute; Naval Arctic Research Laboratory
Financial assistance to help bring certain distinguished foreign participants
I
to this conference was provided by the Office of Polar Programs, National Science
Foundation under Grant No. OPP75-00l29, the Alaska Sea Grant Program with funds
provided by NOAA Office of Sea Grant under Grant No. 04-5-158-35, and by the
Institute of Marine Science with funds appropriated by the State of Alaska.
iv

Dr. D. W. Hood
Mr. J. E. Bowker
Dr. A. Brebner
Dr. C. Bretschneider
Dr. P. Bruun
Dr. T. Karlsson
Dr. D. C. Burrell
Dr. Vera Alexander
Mr. D. C. Crevensten
Mr. D. H. Rosenberg
Mrs. Sandra Rosenberg
Dr. W. M. Sackinger
Ms Patricia Dyer
INTERNATIONAL COMMITTEE
ORGANIZING COMMITTEE
v
University of Alaska
Fairbanks,
United States
Bowker, Associated,
Incorporated,
Boston, United States
Queens University
Kingston, Canada
University of Hawaii
Honolulu,
United States
Norwegian Institute of
Technology
Trondheim, Norway
University of Iceland
Reykjavik, Iceland
Chairman and Publications
Finances
Scientific Excursions
Local Arrangements
Spouses' Program
Technical Program
<strong>Conference</strong> Secretary

TABLE OF CONTENTS
VOLUME I
PREFACE. .iii
SPONSORS. iv
INTERNATIONAL COMMITTEE v
ORGANIZING COMMITTEE. • v
LISTING OF PROCEEDINGS PAPERS APPEARING IN ASSESSMENT OF THE ARCTIC
MARINE ENVIRONMENT: SELECTED TOPICS . ••.•.•.•..•...... xvi
SECTION 1 - PLENARY SESSION
The Future of Oil and Gas Exploration and Development in
the Canadian Arctic
C. R. Hetherington. • • . • • • • .
Alaska - The Search for a Unified Policy of Offshore Oil
and Gas Development
G. R. Martin. . . • • • • • . • • . • • . . .. . ...• 11
SECTION 2 - PROGRAMS ON NORTHERN RESEARCH
Status of Meteorological and Oceanographic Information
Relative to the Petroleum Industry in the Gulf of Alaska
W. R. McLeod •.•.••..••••••..•.••..••• 19
SECTION 3 - INSTRUMENTATION
Energy Supply in the Arctic (Abstract)
C. Bettignies • . .••••.••.... 41
Two Recently Developed Arctic Data Buoys
B. M. Buck, W. P. Brown, S. P. Burke and E. G. Kerut ..•.. 45
Technical and Logistics Problems Associated with Air Deployable
Instrumentation Systems in the Arctic Basin (Abstract)
R. W. Correll, T. McGuirk and W. Lenharth •
Effective Instrumentation Development for Arctic Ocean
Engineering Research
G. M. Gray ....
Development and Use of the Arctic Data Buoy
D. P. Haugen and E. G. Kerut. •
(Abstract)
Oceanographic Instruments for Arctic Use
E. L. Lewis . • • • . • • • •. • . • . • . . • . • . • . 71
vii
3
• 55
•.•.• 61
••••. 69

SECTION 6 - (continued)
The Significance of the Shear Rigidity and of the Poisson Ratio
for Sea Ice Plates
K. Hutter
.247
Measurement of Sea Ice Force by the Strain Rosette Method in
the North Water Area
H. Ito and F. MUller. • . • • • • • • .269
An Early Desalination and Ice Structures Project Using Natural
Freezing (Abstract; <strong>Conference</strong> Late News Paper)
P. R. Johnson . • • • • • . . • • • . • . . • .285
Ice Mechanics
H. R. Kivisild.
Glaciological Investigations for the Improvement
Ship Design Carried Out on the Sea Ice Near Pond
(Northern Baffin Island) in Spring, 1972
H. Kohnen •.••.•••••.•••••.
Islands of Grounded Sea Ice
A. Kovacs, A. Gow and W. F. Dehn ••.
of Ice-Going
Inlet, N.W.T.
• .287
· .315
.333
On the Flexural Strength of Brackish Water Ice by in situ Tests
M. Ml!1!ttltnen. . • • • • . • • • • • • • •. .349
Internal Stress Measurements in Ice Sheets Using Embedded Load
Cells
R. D. Nelson. • • . . . • .361
Interpretation of the Tensile Strength of Ice Under Triaxial
Stresses
D. E. Nevel and F. D. Haynes. • .375
A Narrow Beam Sonar to Measure the Submarine Profile of an Ice
Ridge
T. Pousi, M. Luukkala and E. Palosuo. • .389
Height to Draft Ratios of Icebergs
R. Q. Robe •..•
.407
A Preliminary Study of Ridging in Landfast Ice at Barrow, Alaska
Using Radar Data
L. H. Shapiro • .417
The Use of Flat jacks for the in situ Determination of the
Mechanical Properties of Sea Ice
L. H. Shapiro and E. R. Hoskins (<strong>Conference</strong> Late News Paper).427
On the Movement of Pack Ice
T. Tabata and T. Ishida
(Abstract)
ix
•.•••.••••••• 437

SECTION 6 - (continued)
An Elastic Structural Analysis of Floating Ice Sheets by the
Finite Element Method
K. D. Vaudrey and M. G. Katona. • . ... 439
Fast Ice Studies in Western Davis Strait
R. L. Weaver, R. G. Barry and J. D. Jacobs ••.•••.•.• 455
SECTION 7 - ICE DYNAMICS
The Simulation of Arctic Sea Ice Dynamics
R. Colony . . . . . • • . . . . . . • . . .469
AIDJEX Field Operations to August 1975
A. Heiberg. • . • . . . . • • • . . . .487
Water Stress Sub-Model for the AIDJEX Model
M. McPhee • . . . •. 495
A Mechanical Model for the Deformation of Arctic Pack Ice
(Abstract)
K. R. Maser • •••. 509
Applications of the AIDJEX Ice Model
R. S. Pritchard and R. T. Schwaeg1er. • •••••.• 513
Ice Motion in the Vicinity of a Grounded F10eberg
W. J. Stringer and S. A. Barrett •••••..•••.•.•. 527
SECTION 8 - ICE AND TRANSPORTATION
Ice Conditions Along the Alaskan Coast During Breakup (Abstract)
N. Borgert. . • . ..•• •••.•. 555
Some System Engineering Considerations for a Floating, Coal-Fired
100MW Power Plant
J. P. Craven, C. Gopa1akrishnan, I. Swatzburg, K. Hotta,
V. Ishibashi, C. K1oke, T. Nakajima, E. Oshiro, R. Watanabe,
D. Wilson and C. Wyse • • • • • . • • . • • • • . .. • .557
USCG Polar Class Icebreaker Design Parameters and Features in
Advancement of Icebreaker Design
Cdr. H. E. Fallon, Jr ••••• .571
Prestressed Concrete Floating Terminal for Arctic Ocean Service
B. C. Gerwick, Jr.. . • • • • • • •••••..•
Hydrographic Surveys and Charting in the Arctic - a Match of
Economics and Technology
A. J. Kerr ••••..•••••
x
.581
• • .597

SECTION 9 - (continued)
Experiences of Ice Forces Against a Steel Lighthouse Mounted
on the Seabed, and Proposed Constructional Refinements
M. MlI11ttlinen. • • • • . • • • . • • • • • . • • . . • • 857
Estimating Pile Icing Under Northern Climates and Tidal
Conditions (Abstract)
A. R. McKay •••••••.•• • .• 871
Possible Uses and Construction Methods for Arctic Coastal
Platforms Fabricated from Frozen Sea Water (Abstract)
A. R. McKay, R. C. Miller, R. H. Ragle and T. Sahin •••• 873
North Sea Offshore Structures Environment and Environmental Loads
B. Pedersen. • • • • • • • . • . • • • • • • • • • 875
Ice Force Response Spectrum Modal Analysis of Offshore Towers
D. V. Reddy, P. S. Cheema and A. S. J. Swamidas • • 887
Techniques for the Study of Ice/Structure Interaction
R. J. Robbins, P. H. Verity, T. P. Taylor and M. Metge ••• 911
Verification of Numerical Wave Computation Reconstruction
of Extreme Storms
E. Smiland. • •• 925
Reinforced Ice: Its Properties and Use in Constructing
Temporary Enclosures
R. G. Stanley and P. G. Glockner. •••• 935
Ice Forces Acting on Inclined Wedges
P. Tryde. . . • . ••••••••••• 957
Ice Forces Acting on Slender Structures
P. Tryde. . • • • • • • • • • • • • • • • • • • • • 961
SECTION 10 - SEABED AND SUBBOTTOM
Some Results of a Thermsl Analysis of Offshore Artificial Islands
G. R. Bafus, G. L. Guyman and R. F. Carlson ••••.••• 967
The Seaward Extension of Permafrost Off the Northern Alaska
Coast (Abstract)
M. C. Brewer •• •• 987
Determination of the Physical Characteristics of Marine Sediments
Related to Offshore Construction via Seismic Methods (Abstract)
W. R. Bryant ••••••.••••••••••••••••• 989
Seabed Scour by Currents Near Platforms
T. Carstens ••••••••.•.• 991
xii

SECTION 10 - (continued)
Coastal Erosion Processes With Special Reference to the
Function of Groins and Headlands in Coastal Protection
(Abstract; <strong>Conference</strong> Late News Paper)
F. Gerritsen. • • • • . • • • • • . • • • • • . • • .1007
Permafrost: From the Bottom Up (Abstract; <strong>Conference</strong> Late
News Paper)
J. O. Hakkila and J. C. Balch. • • • • • . • • • • .1009
Theoretical Models for Sub-Sea Permafrost
W. D. Harrison and T. E. Osterkamp. • •••• 1011
Geophysical Investigations of Sub-Sea Permafrost in the
Canadian Beaufort Sea
J. A. Hunter and A. S. Judge. • . • • • •. 1025
Determination of Wave-Induced Pressures in the Soil Media
Contiguous to a Buried Pipeline
N. W. Lai, R. F. Dominguez. • • • • • .1059
Near Shore Permafrost in the Vicinity of Pt. Barrow, Alaska
J. C. Rogers, L. H. Shapiro, L. D. Gedney and
D. Van Wormer. • • • • . • • • . • • • • . • • • .1071
SECTION 11 - NORTHERN OIL AND MINERAL
Polar Gas Project: Natural Gas Pipelines in the Arctic
Environment - Engineering Research
L. M. Etchegary and W. Hindle • . • • • • .1085
Pipelines and the North Sea
J. P. O'Donnell ••••
(Abstract)
• .•• 1105
Polar Gas Project: Natural Gas Pipelines in the Arctic
Environment - Environmental Research
H. E. Palmer and W. Hindle. . • . • • • • • • .1109
Unique Problems of Oil Development in the Arctic
O. G. Simpson. • ••••.• .1129
Mining in Antarctic: Survey of Mineral Resources and Possible
Exploitation Methods
J. F. Sp1ettstoesser •• .1137
Offshore Drilling from Artificial Ice Platforms
H. J. Strain •••••••••••••••.••.••••• 1157
xiii

SECTION 13 - (continued)
Research on Integrated Offshore Treatment System of Town
Wastes and Effective Utilization of its Byproduct
Misao Noguchi and Minoru Ishida . • • • . • • •. • ••. 1319
Governmental Administrations on Preservation of Ocean
Environment in Japan
Taisuke Samethima • • .
••.•. 1329
Application of Dielectric Radar Reflector (Lens ref)
John T. Takarabe. . • • • • • • • • . •• 1335
Construction, Towage and Installation of the Floating City
"Aqua polis"
Masayuki Tamehiro • • .1345
On the Mechanical Properties and Results of Soak Test in Sea
Water for Anti-Corrosive Wire Ropes
Isao Ueno • . . • • • • • • • . • • . .1375
AUTHOR INDEX •.•••••••..•.•••••••••••.••.. 1381
xv

LISTING OF PAPERS APPEARING IN
Assessment of the Apotio MaPine EnviPOnment: SeLeoted Topios
(Hood and Burrell, eds. IMS Occas. Publ. No.4)
PART I - ASSESSMENT CONCEPT AND PROJECT APPROACH
SECTION 1 - THE ARCTIC PREMISE
Chapter 1 D. W. Hood
Chapter 2 M. J. Dunbar
Chapter 3 P. Korringa
Chapter 4 B. E. Willard
SECTION 2 - REGIONAL PROGRAM PERSPECTIVES
Chapter 5
Chapter 6
Chapter 7
Chapter 8
Chapter 9
Chapter 10
H. C. Bugge
A. F. Treshnikov
H. E. Bruce
R. D. Muench
R. Elsner
C. P. McRoy
A. R. Milne
D. W. Hood
Introduction
PART II - CONTEMPORARY TOPICS
SECTION 3 - THE NORTHERN SEABED CONDITION
Chapter 11
Chapter 12
A. S. Naidu
C. J. Lee
T. C. Mowatt
C. H. Everts
xvi
Man in the Polar Marine Ecosystem
Ecological Assessment as Criterion in
the Rational Exploitation of Natural
Resources
Environmental Principles for Guiding
Engineering Design in Arctic Regions
The North Sea Problem
Recent Soviet Research in the Arctic
A Study Plan for the Alaskan Continental
Shelf
PROBES: A Study of the Bering Sea Shelf
Canadian Environmental Studies of the
Southern Beaufort Sea
Marine Terminus of the Trans-Alaskan
Pipeline
Chemistry of Deep-Sea Sediments in the
Canada Basin, West Arctic Ocean
Sedimentation in a "Half-Tide" Harbor

SECTION 3 - (continued)
Chapter 13
Chapter 14
J. C. LaBelle
C. M. Hoskin
S. M. Valencia
SECTION 4 - DYNAMIC PHYSICAL PROCESSES
Chapter 15
Chapter 16
Chapter 17
Chapter 18
Chapter 19
J. A. Dygas
D. C. Burrell
L. W. Gatto
W. J. Wiseman,
A. D. Short
O. G. Houmb
I. Vik
J. A. Dygas
D. C. Burrell
SECTION 5 - BIOLOGICAL FEATURES
Chapter 20 R. C. Clasby
V. Alexander
R. Horner
Chapter 21 S. A. Norrell
M. H. Johnston
Chapter 22 H. M. Feder
D. Schamel
Chapter 23 P. C. Craig
P. McCart
Chapter 24 C. Atkinson
Jr.
SECTION 6 - HYDROCARBONS IN THE ARCTIC ENVIRONMENT
Chapter 25
Chapter 26
C. D. McAuliffe
D. G. Shaw
L. M. Cheek
xvii
Fill Materials Between Barrow and the
Colville River, Northern Alaska
Sediment Transported by Ice-Rafting in
Southcentral Alaska
Dynamic Sedimentological Processes
Along the Beaufort Sea Coast of Alaska
Circulation and Sediment Distribution in
Cook Inlet, Alaska
Mesoscale Thermal Variations Along the
Arctic North Slope
Duration of Sea States in Northern Waters
Response of Waves and Currents to Wind
Patterns in an Alaskan Arctic Coast Lagoon
Primary Productivity of Sea-Ice Algae
Effects of Oil on Microbial Component of
an Intertidal Silt-Sediment Ecosystem
Shallow-Water Benthic Fauna of Prudhoe
Bay
Fish Use of Nearshore Coastal Waters in
the Western Arctic: Emphasis on Anadromous
Species
Development and Potential Yield of Arctic
Fisheries
Surveillance of the Marine Environment
for Petroleum Hydrocarbons
Hydrocarbon Studies in the Arctic Benthic
Environment

SECTION 6 - (continued)
Chapter 27 T.
Chapter 28 J.
S.
D.
A. Gosink
w. Short
D. Rice
L. Cheatham
xviii
Sample Size in Environmental Assessments
Comparison of Two Methods for Oil and
Grease Determination

THE FUTURE OF OIL AND GAS EXPLORATION AND DEVELOPMENT IN
THE CANADIAN ARCTIC
Charles R. Hetherington
Panarctic Oils Ltd
Calgary, Alberta
Canada
ABSTRACT
The Canadian Arctic regions north of the 69th paraLLeL have substantiaL proved reserves
of oiL and gas and very much Larger potentiaL reserves yet to be discovered. AccordingLy,
these Arctic areas, aLong with offshore Eastern Canada consti-tute the Last great frontier
for the suppLy of oiL and gas within Canada.
The eztent to which the Canadian Arctic is deveLoped, and the timing of deveLopment wiLL
dete1'l1line the eztent to which Canada wiLL remain seLf sufficient in energy and the eztent
to which Canada wiLL have surpLus oiL and gas avaiLabLe to share with the United States.
Thus, the subject of the future of oiL and gas expLoration and deveLopment in the Canadian
Arctic wiLL hopefuLLy be of interest to this <strong>Third</strong> InternationaL <strong>Conference</strong> on Port and
Ocean Engineering Under Arctic Conditions.
WhiLe Canada is the onLy industriaLized nation in the worLd that is seLf sufficient in oiL
and gas, this seLf sufficiency is short Lived. Shortages of oiL and gas are to reduce
ezports of oiL and gas to the United States to maintain suppLies to meet Canadian requirements
untiL frontier oiL and gas and other sources of energy can be deveLoped.
PeopLe skiLLed in oiL and gas expLoration through bJOrLd ezperience beLieve that the geo­
LogicaL prospectiveness of the Canadian Arctic coupLed with impressive discoveries to
date, indicate that there is the potentiaL in this area not onLy to suppLy seLf sufficiency
for Canada, but aLso to provide surpLus oiL and gas for many years into the future.
MY coLLeague Mr. H. J. Strain, is presenting a paper at this <strong>Conference</strong> deaLing with the
technoLogy and operating methods that have been and that are being deveLoped, which are
making it possibLe from a physicaL standpoint to carry on the operations in this remote
and inhospitabLe area necessary to evaLuate the potentiaL of the Arctic IsLands and bring
oiL and gas resources to market.
In this paper, I wiLL deaL with the other side of the coin, nameLy the economic, poLiticaL
and environmentaL factors which wiLL probabLy have more impact on the time frame within
which Arctic oiL and gas may be deveLoped than wiLL the physicaL abiLity to conduct Arctic
operations. In doing so, I wiU;
a. Review Canada's suppLy of and demand for oiL and gas and how this
affects oiL and gas avaiLabLe for export to the United States,
b. Review the sta-tus of oiL and gas deveLopments in the Canadian Arctic
and the potentiaL for future suppLy from this area, and
c. Review factors affecting the pace of Arctic deveLopment.
3

INTRODUCTION
The extent to which the Canadian Arctic is developed, and the timing of development will
determine the extent to which Canada will remain self sufficient in energy and the
extent to which Canada will have surplus oil and gas available to share with the United
States.
CANADIAN NEEDS AND THEIR EFFECTS
Canadians are the world's second largest per capita users of energy, with each individual
consuming on average the equivalent of 55 barrels of oil per year. This high
consumption is the direct result of Canada's climate, widespread population, high standard
of living, and historically low prices for energy.
About 65 percent of the energy requirements of Canada are supplied by oil and gas,
essentially all of which is produced in Alberta, British Columbia, Saskatchewan and
Manitoba, in that order of importance. Remaining reserves in these areas are about
9 billion barrels of oil and 53 trillion cubic feet of gas, which in total equate to
about a l5-year remaining supply at today's production rates.
Overall figures like this may be considered to be misleadingly alarming, because as shown
in the past, the discovery of new reserves has kept pace with increasing needs. However,
this trend is no longer true. For four years now since 1970, the production of oil in
Western Canada has exceeded discoveries so that remaining oil reserves are decreasing.
Similarly, the production of gas has exceeded reserve additions in two of the last
three years and remaining gas reserves are decreasing.
Energy consumption in Canada is growing at a rate which, if left unaltered, will
require four times as much energy in the year 2000 as was consumed in 1970. This
realization does become alarming when considered in the light of decreasing discoveries
of new reserves.
Concerned with the implications of this situation, the National Energy Board of Canada
in 1974 undertook a series of public hearings to determine what Canada's position
actually was with respect to supply and demand for oil. The Board reached the conclusion
that without frontier reserves Canada would become a net importer of oil in 1977
and that even with continued imports into Eastern Canada, a deficiency of supply from
indigenous oil to meet the demand of the rest of Canada was indicated in the early
1980's.
In view of this foreseeable deficiency, the Board reviewed the possibility of conserving
remaining reserves through reduction or elimination of exports, along with all of
the implications of such action. The Board observed that the best protection for
Canada in the long run, is obtained from an active and economically healthy industry
continually increasing available reserves, and it considered the adverse effect upon
the Canadian exploration industry of reducing production which would correspondingly
reduce funds available to the oil companies for exploration. The Board concluded that
at the point when forecasts show that oil surplusses will disappear in less than 10
years, exports should be phased out over the period of indicated surplusses. Using
forecasts available in 1974, and specifically excluding oil supplies in the frontier
areas, the Board concluded that oil exports to the United States should be progressively
phased out by 1983. The Government of Canada has instituted a policy generally
along these lines.
I think Canada's approach to protecting its own domestic oil supplies is generally
misunderstood, particularly by Americans who are feeling the impact of export restrictions.
Basically, we are only looking after our own oil requirements first, with
4 Hetherington 2

export restrictions phased over a period up to 10 years to minimize disruptions abroad,
and to provide time to arrange alternate supplies.
The National Energy Board formula for determining available exports contemplates that
exports can be increased if and when frontier oil, (or for that matter, any new oil
reserves) is developed to the extent that Canada once again has surplus oil for a
period of ten years into the future. Accordingly, when adequate oil reserves are
developed in the Canadian Arctic and can be transported to market, the mechanism has
already been set up in Canada to allow exports.
A somewhat similar situation is developing with respect to natural gas. The National
Energy Board of Canada held public hearings on gas supply and demand, and their report
confirms earlier concerns that natural gas supplies will not be adequate in the near
term to meet both the increasing domestic demand and existing export commitments.
The Board recommends efforts to increase short term deliverability, increased pricing
toward commodity value with oil, and conservation, and it stated its view that reasonably
foreseeable requirements for gas for use in Canada consistent with the pricing,
conservation and industrialization policies of Canadian governments must be given
priority over existing export commitments.
Here again, when Canadian Arctic gas reserves are developed and made available for
market, the mechanism will exist for allowing export of surplus gas just as in the case
of oil. But the Board cautions that new exports must be relatively short term and
conditional on Canadian needs having continuing priority.
ARCTIC DEVELOPMENT AND POTENTIAL
The search for oil and gas in Canada's Arctic is centered in two areas, the Mackenzie
River Delta and the adjacent offshore Beaufort Sea, and in the Canadian Arctic Islands
a thousand miles to the north and east of the Delta. Oil and gas discoveries have been
made in both areas.
Arctic Islands
In the Arctic Islands, 6 large gas fields have been discovered with reserves estimated
to be in excess of 13 trillion cubic feet. The largest reserves are located on the
Sabine Peninsula of Melville Island in the Hecla and Drake Point fields. Four other
gas fields are located on and around Ellef Ringnes Island, at Kristoffer Bay, King
Christian Island and Thor Island. The gas wells all have exceptionally high deliverability
ranging to over 400 million cubic feet per day Absolute Open Flow capacity.
In discussing Arctic technology, it is interesting to note the offshore drilling
method developed by Panarctic which has made it possible to drill wells today as far
as 8 miles offshore in wate? over 400 feet deep, using a modified conventional drilling
rig supported on artificially thickened ocean ice. The drilling method is both effective
and safe, and makes it possible to drill offshore wells where ice is landfast at a
fraction of the cost of any other available means.
All geological evidence in the Arctic Islands indicates favorable prospects for crude
oil as well as gas, and there have been several encouraging oil shows. Out of the
some 80 wildcats drilled to date, one well on Cameron Island has encountered potentially
commercial oil.
A total of 100 wells has been drilled in the Arctic Islands at a total cost including
seismic of some $350 million.
5 Hetherington 3

potential of the area so that energy policy can be established on a sound basis. If in
fact Canada's Arctic contains huge oil and gas reserves, Canadian energy policy is
certain to be different than if reserves in the Arctic prove to be minimal.
From an economic standpoint, it is also vitally important to discover and develop new
oil and gas reserves required by Canada in the next few years, because present Arctic
reserves are not sufficient to support the economic viability of transportation
facilities required to bring these remote resources to market. Both Canadian and
American markets are remote from the Arctic, and oil and gas must be transported great
distances over difficult terrain, where access and logistics are so costly that the
required threshold reserve to permit any marketing, is very large.
A threshold reserve of 25 to 30 trillion cubic feet of gas is needed for marketing of
Mackenzie Delta gas by pipeline. If Alaskan gas is combined with gas from the
Mackenzie Delta as proposed by the Canadian Arctic Gas Pipeline project for transportation
through Canada to both the west coast and eastern United States, there probably
is an adequate gas supply. But the Alaskan gas which constitutes the bulk of the
reserves may not be available for transportation through Canada.
A competing proposal for marketing Alaskan gas sponsored by El Paso Natural Gas Company
proposes to transport Alaskan North Slope gas across Alaska, paralleling the Alyeska
oil pipeline to an ice-free port where the gas would be liquefied and transported by
LNG tankers to markets in the Pacific northwest and California. The gas reserves in
Alaska would appear to be adequate to support this project. El Paso states that studies
establish that gas can be delivered via its project to United States markets through
facilities located entirely within the United States at about the same cost as a pipeline
system bringing this gas through Canada.
There is no question that Alaskan gas is needed to supplement declining energy supplies
in the United States, and El Paso maintains that keeping the gas in the United States
would favor the U.S. balance of payments; would eliminate uncertainties in foreign
energy sources and could be implemented far sooner than any other plan. El Paso maintains
that the transportation of natural gas through Alaska will stimulate the development
of Alaska's rich mineral deposits along the pipeline route and adjacent areas
with increased employment and tax base within Alaska. Public announcements would seem
to indicate that the State of Alaska supports keeping its gas within the State. However
one views the El Paso Alaska project, it must be considered a serious competitor
to the Canadian Arctic Gas Pipeline System.
If gas from Alaska is not available, there is not sufficient gas in the Mackenzie Delta­
Beaufort Sea area to support a pipeline project out of this area alone, either by the
Arctic Gas system or by the competing system proposed by Foothills Pipelines. Arising
out of the possibility that Alaska gas may not be available for transportation through
Canada, a competing project has been proposed by Foothills Pipelines Ltd. to take gas
only from the Mackenzie Delta-Beaufort Sea area to Canadian markets. Here again, this
project suffers from not having sufficient gas reserve in the Mackenzie Delta-Beaufort
Sea area to support the economic viability of the extensive pipeline facilities
involved.
In the Canadian Arctic Islands, the minimum threshold reserve required to support a gas
pipeline outlet to market, is about 20 trillion cubic feet. While over one-half of
this reserve has been established, more reserves must be discovered before a pipeline
project can be built based on gas from the Arctic Islands alone.
While the immediate objective is to establish enough threshold reserves to permit
marketing, the long term objective is to evaluate the oil and gas potential of the
7 Hetherington 5

When we have to start digging up $12 billion of foreign exchange to pay for this
imported oil, I rather suspect that oil and gas companies will be treated more generously,
providing they undertake to reinvest in exploration.
Other factors that are deterring Arctic exploration are more of a short term and annoying
nature, and are being corrected. The Government of Canada by a sort of forced
agreement with the Provinces, regulates the price of oil below world prices, and the
price of gas below its oil BTU equivalent. Fortunately, the Federal Government seems
to recognize that subsidizing eastern Canadian oil consumers and holding down the price
of natural gas contributes to waste rather than conservation in the market place, and
also acts to reduce funds available for exploration to find new oil and gas supplies.
It is announced Government policy to raise natural gas prices over a preiod of time to
their BTU equivalent with oil. The Government has also indicated its intention to allow
crude oil prices to rise toward present world values, also over a period of time, to
levels which would support adequate exploration and development of new reserves. But
in the meantime, we go on wasting gas and oil at low prices while exploration in the
high cost frontier areas is declining for lack of funds.
Land regulations and royalty rates on Canadian Federal lands which have been uncertain
for some period of time, are now being established. Recently, the Minister of Indian
and Northern Affairs announced the principles which will govern the leasing of Federal
lands, which he stated are intended to stimulate exploration.
Perhaps in the past, the oil and gas industry did not give proper consideration to
environmental and sociological matters, but this is certainly not true today. From my
own first hand knowledge, I believe every company operating in the Canadian Arctic is
conscious and respectful of the environment. In addition, these companies have progressive
programs for the employment and benefit of natives.
CONCLUSION
In conclusion, I would like to make this observation. When Panarctic first went into
the High Arctic, the big question was how to operate successfully in the harsh natural
climate. It was hard to foresee a political/economic climate that eight years later
would prove far more difficult to overcome. The fact is, that we cannot afford to slow
down exploration activity in the frontier areas if we are to discover and bring to
market required new energy supplies within an acceptable time frame.
Expedient short term decisions that retard the pace of development, though they may
look reasonable today, will certainly affect a broad aspect of Canadian life in the
long term. The answer can only be found in more mutual understanding and cooperation
between the public and private sector. Hopefully, this better understanding is
beginning to come about.
9 Hetherington 7

ALASKA - THE SEARCH FOR A UNIFIED POLICY OF OFFSHORE OIL AND GAS DEVELOPMENT
Guy R. Martin
Commissioner of Natural Resources
State of Alaska
United States
ABSTRACT
Thank you for the opportunity of being here. Although I am a newaomer to yoUP association
and not of the scientific corrmunity bJhich dominates yoUP program, I believe bJe have much
to shaPe. I am conscious of the breadth and quality of ezperience in Arctic engineering
bJhich is present here, and it gives me tremendous hope and encouragement to look at the
formidable program you face during this bJeek.
It bJOuZd be most useful, I bJOUZd guess, if I could focus my remarks on specific aspects
of offshore oil and gas development in the Arctic Ocean and the Beaufort Sea. For reasons
I bJiZZ e:x:plain, the perspective I must take is broader. At present, Alaska finds herself
poised on the brink of a second generation of offshore oil and gas development, the first
generation ocCUPPing in Cook Inlet over the past decade. The brink bJhich bJe approach is
precipitous, the size of the abyss is intimidating, and it is not yet certain bJhether bJe
are moving voluntarily tobJard the edge or being pushed.
AlZobJ me to describe at least some of the component parts of the situation the state faces
in this second generation. For those many of you from other nations, I ask that you forgive
me the domestic politics bJhich cannot be altogether separated from the report.
11

First, Alaska learned in October of last year that the Federal Government had firm plans
(denied then, but firm now) to lease, for oil and gas development, every area on Alaska's
outer continental shelf which had oil and gas potential. Nine regions were identified
covering virtually all of Alaska's coastline, from the Gulf of Alaska to the Bering,
Chukchi, and Beaufort Seas. Unbelievably, the entire leasing program--the selling of the
shelf--is to be completed in just three years.
Not without its own problems, the State of Alaska added a component in the form of a
revenue crisis brought about by trans-Alaska pipeline delays and the rapid dissipation of
the now famous $900 million from the Prudhoe Bay oil and gas lease sale of 1968. By a
logic which is curiously ironic, the first solution proposed for the crisis was another
oil and gas lease sale in the very same area. This is, of course, the Beaufort Sea, and
it represents to the State one of the principal means of raising substantial sums of revenue
at an early time, as well as being the first major step for Alaska into the Arctic waters
this conference knows so well.
For political and environmental flavor, there is a third ingredient in the form of a small
but volatile oil and gas lease sale held in the closing days of the prior State administration
in the mouth of beautiful Kachemak Bay at the south end of the Kenai Peninsula
adjacent to the Cook Inlet oil fields. This sale committed for oil and gas operations
that many believe to be one of Alaska's environmental and recreational jewels, as well as
an important shellfish habitat. In return, it produced less than $30 million in bonus
revenues.
There is more to say, but it is sufficient to conclude that the second offshore generation
will be huge, widespread, and immediate. Industry, government and others are prepared to
move into the water, (or for this conference, onto the ice) at an early time, in a big way.
Onto this stage came the new State administration of Governor Jay Hammond, having campaigned
on a platform suggesting the need for greater caution regarding major developmental
decisions, having secured a narrow victory, and having left three former Governors
of Alaska in his elective wake.
This administration, depending on your perspective and the papers you read, is either the
devil as manifested in a "no-growth" policy of economic strangulation or the last chance
to get a governmental handle on economic and developmental forces which threaten to
change forever what is best about Alaska. It may even be something in between. From day
number one a unified offshore oil and gas development policy was at the top of the priority
list for this administration.
Because of its immediacy and its scope, the Federal O.C.S. program became more a focus
for the policy process than did our own State issues in the Beaufort Sea or Kachemak Bay.
In retrospect, this may not have been bad, for it forced the State to broaden its view
to the entire coastal area, and to see if it could articulate a policy for the Federal
areas to which it would then subject itself.
That policy process is not complete, nor is it likely ever to be finalized and locked
away. Yet, there is a clear recognition by the administration that, in addition to the
need for policy consistency, there is a need for finality and certainty. The oil
industry is, to use its own cliche, a "high risk, high yield" venture which, while pursuing
its admitted self-interest, desires clear standards having continuity in excess
of the terms of political administrations. I believe the goal of this administration is
to address that legitimate desire and to provide a policy which will have continuity and
be internally unified and consistent.
12 Martin 2

The force of events, in the form of the Federal O.C.S. program, has both accelerated the
development of the policy and shaped it. Although there is not time here to summarize al
the emerging points, I believe it will be useful to discuss at least those cornerstones
of such a policy which relate to the interests of those here, and to indicate the way in
which these points could apply to a consistent State policy for the specific and diverse
offshore situations I earlier described.
One of the earliest lessons the State learned from the Federal O.C.S. program was that
its fixed structure and schedule was fully determined before it was ever made public.
Subsequent events have born this out here and elsewhere, in that the effect of a virtual
cavalcade of dissapproval for many aspects of the program have resulted in virtually no
change in it. Thus, while it is not possible to charge that the Federal program is devoi,
of public input, it may as well have been so, for there is no responsiveness to that whicl
is made. A similar history haunts the Kachemak Bay sale, where the files are full with
inconsistent and unfulfilled promises to the public.
As a result, the State is pledged to maximize the public process affecting offshore oil
and gas deCisions, and to maximize the breadth of the decision-making base in its own
Bureaucracy. This policy has been carried out thus far regarding a Beaufort Island sale,
even though unrequired by law, and it would be our intention to encourage public participation
continuing through the stipulation process at the time a lease sale is actually
decided. The end result, of course, should be an offshore lease sale which the public
not only participated in, but affected.
Similarly, we learned early in the Federal program that the pre-existing schedule and
structure preceded environmental studies in the lease area rather than the other way
around. Unbelievably, the first year baseline research effort in the northeast Gulf of
Alaska will just be winding up as the lease sale is held. The fine researchers from NOAA
who are coordinating this study do not believe, I think, that they have any relation to
the lease schedule at all. As engineers and scientists, I must believe that you would
want this information before proceeding, and the State will adopt a policy which attempts
to maximize environmental information prior to the making of a leasing decision. This is
the process we followed with the preparation of a Beaufort Islands environmental assessment,
again unrequired by law. No such state process existed for application to the
Kachemak Bay sale, nor was such a process evidenced in the furor following the sale and
existing now.
Going one more step with the Federal O.C.S. program, we discovered that there existed no
mechanism except administrative discretion for identifying areas where offshore oil and
gas development may be wholly inconsistent with other resource or use values, be they
fisheries, recreation, wilderness or simply aesthetic. Surely, even in our rush to "sell
the shelf" in Alaska, we can find and agree on those special areas where either we should
not proceed at all, or if necessary, do so only as a last priority.
If there are such areas, must we not ask for a way to define them and to identify them,
and a mechanism to exclude them? The State will adopt such a policy with regard to its
own sales, and is now defining the precise areas of concern. Included in the evaluation
phase are both Kachemak Bay and the Beaufort Sea areas. Thus far, there has been no indication
that there is either ability or willingness on the part of the Department of the
Interior to establish such a process to avoid extreme resource conflicts and protect special
areas.
Next, insofar as possible, it has appeared desirable to the State that new or "Frontier"
offshore operations occur first in areas where such operations are likely to cost the
least, create the fewest resource conflicts, and require the least amount of major changee
in the area. Regarding the Federal program, it is clear that it should be resisted in ite
13 Martin 3
,

intention to impose itself first on the northeast Gulf of Alaska, where the costs will be
high, the resource conflicts great and the change in the nearby area extraordinary. To
place the northeast Gulf so high on the schedule, the Department of the Interior had to
look blindly past east coast areas where, on every count, earlier leasing would have been
more appropriate.
In Kachemak Bay, resource conflicts are extremely high. Change in the immediate area will
be extreme, although prior Kenai Peninsula offshore operations provide some base against
overall major change and perhaps against excessive costs. In the Beaufort Island area,
particularly north of Prudhoe Bay, new offshore operations would build on existing North
Slope operations to moderate both change and costs. Resource conflicts, although present,
would appear to be far less severe than either of the other areas proposed.
Potential costs, those undefinable risks from a major spill to the existing resources of
an area such as fisheries, or scenic and recreational value, must be also a factor. Not
only the value of these resources, but also the difficulty and risks of oil and gas operations
must be balanced in comparing and prioritizing offshore areas. Such an evaluation
must result in potential costs far higher in the northeast Gulf and in Kachemak Bay than
in the Beaufort Island area of the Beaufort Sea.
A related component of a unified policy must be the place that any lease sale, and subsequent
offshore operations, will play in a rational scheme of oil and gas development.
Should it occur first in an area like the Beaufort Sea, where onshore support facilities
either exist or would be compatible, and where a pipeline transportation system will exist
with carrying capacity on a proper schedule?
Should it come first in an area like the east coast, where the energy demand is the greatest,
where the supply lines would be the shortest, and where processing facilities will be
available? Or should it come first and immediately where no such facilities exist and
transportation of gas or crude oil will be contingent on a new, yet to be designed and
constructed system and fed into a market which may well be in a surplus position when
delivery is made?
Perhaps the major emerging cornerstone of a unified State offshore policy is the concern
of the State for her communities. This is so because the communities, with their distinct
and special personalities, have so much to do with the quality of life in Alaska. Not
even the most zealous advocate of offshore development denies the toll that such activity
will take, in terms of change, on the coastal towns of Alaska. There is virtual unanimity
that "it will never be the same," although there is certainly less agreement whether this
result is good or bad.
In the beautiful towns of Homer and Seldovia on Kachemak Bay, the verdict is "bad", and
it is all the worse that their feelings were given so little public outlet prior to the
sale of oil and gas lease sales in their own backyard. In the northeast Gulf of Alaska,
the verdict respecting the merits of the change is different. There, the communities
plead for time to plan, and for assistance in doing so, before the explosion takes place.
In this area, towns like Yakutat, Cordova, Seward, and Kodiak, like the towns of Northern
Scotland before them, are becoming aware of what they can do, given time.
Let us give them the time they need and seek out those areas, whether they be in other
sections of the United States, or on the already developed North Slope of Alaska, where
the conflicts of community change will be less and the scale of the need, with fewer
community conflicts, will afford us all the capability to get ready.
14
Martin 4

It is my belief that the State will continue in its policy to seek to avoid the sacrifice
of its coastal communities, even to such an objective as energy supply, but rather will
seek the time, the assistance or the conditions necessary to protect the people. I hope
we never have to ask "how come nothing's like it was until its gone?" (Sammy Davis, Jr.
as Will Martin in "Yes, I Can").
Finally, let me touch on the policy themes related to technology, and its darker side,
inevitability. You know as engineers and scientists that many Alaskans are gripped by
the notion of inevitability--inevitability of growth, inevitability of offshore oil and
gas development, inevitability of the loss of "the Alaskan way of life"--hard to describe
but certainly real. Inevitability translates in engineering to " ••• show us the problem,
we'll solve it".
The economist Kenneth Boulding captured the feeling as follows:
"I have recently discovered the name of the devil, and that is something terribly
important to know. The real name of the devil is sub-optimization, finding out
the best way to do something which should not be done at alL •• "
Is this "sub-optimization" not the tool, the concept, which guarantees that even the marginally
inevitable can be accomplished with hard work, i.e., "The difficult we can do
today, the impossible will take until tomorrow".
This concern has translated into emerging State policy on offshore development not so
much as a request for statutory standards, but as a plea for the Alaska public and its
government to be able to say "no" to certain aspects and methods of offshore development.
This concept clearly motivates a policy which seeks a recognition that some coastal areas
are so valuable for other resource purposes that, even in the face of inevitability, and
even with the skill to "do the best job possible", we should be able to either preclude
oil and gas operations or undertake them only as a final priority.
Regarding technology, the nuts and bolts of offshore oil and gas exploration, the stipulations,
the design of equipment, the operating orders, this same concept has translated
into a request for consideration of a new technological standard. Again, it is more a
plea for a change of philosophy than for a statutory regime.
Under present law, offshore operations must meet a standard of the "best available technology".
That is, if we do it as well as we can, then we can proceed. This is the
standard of inevitability articulated in law.
The State has suggested that this standard be restated to require "environmentally safe
technology". This has been met, as I am sure some of you are thinking, with the charge
that "it can't be defined" and that "there will always be some risks". On the latter
count, I agree, but the risks should be evaluated, and taken where necessary, according
to a standard which measures the overall environmental safety of a thing or action
rather than merely whether it is "the best we can do".
On the latter--"it can't be defined"--let me agree that it is no easy task to define anything
which is subjective, for it consists of qualities which can't be added or subtracted.
Still, I would believe that there is agreement here that "environmentally safe technology"
is recognized not only as a concept of value which ought to be defined, but as a challenge
for engineers which far exceeds the lesser burden of simply locating those things and
processes which are at the top of the present "state of the art".
To apply a standard of "environmentally safe technology", requires not only that, as engineers,
you design and know what is the best possible, but also that it is safe. Should
we not seek to set minimum standards for offshore operations rather than being shaped
only by what is technologically available?
15 Martin 5

The task of definition not only on this policy but on the others I have described belongs
to us all, but falls most heavily on the State. To address the Federal O.C.S. program,
the ultimate fate of a prior lease sale in Kachemak Bay, and the sale of State leases in
the Beaufort Sea with integrity, credibility and consistency is not only the objective of
the State, but the absolute prerequisite for Alaska to deal fairly and successfully with
the oil industry over the coming years.
All of these points must come together for Alaska under the general commitment to make a
responsible and major contribution to the Nation's energy supply from the natural bounty
with which our State is blessed. They must come together in a policy which neither
acquiesces blindly to the preferred or even traditional perogatives of the oil industry,
nor arbitrarily denies it a stable, consistent and fair standard on which it can do the
work we agree is important.
State policy on the issues I have discussed, and many others, has been developing for
months, and there continues to be a press for practical decisions on the elements of
Alaska's second generation of offshore oil and gas development. Some of these decisions
must and will be forthcoming shortly.
The "second generation" offshore will be much bigger and far longer than our meager experience
in the Cook Inlet. The results of present decisions on the Beaufort Sea, on
Kachemak Bay and on the Federal O.C.S. program will affect this State long after they are
made, but they will set the pattern now.
Also setting the pattern will be the quality of the work the members of this conference
will do. As I look at your program, I see potential answers to many of the questions we
must ask as we move offshore in the Arctic. On behalf of Governor Hammond, I want to
extend gratitude for the work you are undertaking, to wish you the very best in your
deliberations this week, and to welcome your many distinguished visitors here to Alaska.
We are honored to have you.
16
Martin 6

SECTION 2
PROGRAMS ON NORTHERN RESEARCH
Denner, w. w.
A Proposed National Program in Aratia Researah for the United States
(Naval Arctic Research Laboratory, Alaska, United States)
[Paper, abstract unavailable)
McLeod, w. R.
Status of MeteoroZogiaal and Oaeanographia Information Relative to the
Petroleum Industry in the Gulf of Alaska
(Marathon Oil Company, Texas, United States)
17

STATUS OF METEOROLOGICAL AND OCEANOGRAPHIC INFORMATION
RELATIVE TO THE PETROLEUM INDUSTRY
IN THE GULF OF ALASKA
Wilfred R. McLeod
Marathon Oil Company
Houston, Texas
United States
ABSTRACT
Historical. devel.opment of physical. oceanographic and meteorol.ogical. data gathering efforts
in the Gul.f of Al.aska is reviewed briefl.y. Emphasis is given to the description of OUX'rent
programs sponsored by the petrol.eum industry. Basic objectiVes, instrumentation, and resul.ts
of expected data are incl.uded in the discussion of each study. Mention is made of
future pl.ans for both industry-sponsored and government-sponsored programs. NOAA and NASA
satel.l.ite data and its potential. benefits for marine operators are al.so discussed. Final.l.y,
instances of present and pl.anned cooperative efforts invol.ving government and the petrol.eum
industry in the Gul.f of Al.aska are given.
19

INTROOOcrION
Before commencing operations in a hostile environment like the Gulf of Alaska, or
indeed in any environment, a large body of meteorological and oceanographic data
is needed to aid in selecting equipnent appropriate for use in such an area. Knowledge
of average or frequently occurring weather conditions, or such parameters as
current speed and direction, wave height and wave forces, is an important factor in
planning for efficient and safe offshore operations. Evaluation of extreme or rarely
occurring conditions provides the basis for the design of structures and selection
of materials and equipnent. Finally, weather prediction is vital to safely conducting
all offshore operations.
The collection of meteorological and physical oceanographic data in the Gulf of
Alaska began as early as 1889 when the Coast and Geodetic Survey, now NOAA-National
Ocean Survey (NOAA-NOS), established a permanent tide-recording station at Kodiak.
At the same time, water temperature readings were also made, although this has not
been done continuously over the years. A few years later, this same agency began
measuring drift currents at harbor entrances, narrows and off headlands. The objective
of this data gathering effort was to provide information for publication in
the Tide Tables and Tidal/CUrrent Tables.
Some time prior to 1899 the u.S. Weather Bureau and the Navy oegan collecting ocean
and weather observations from ships-of-opportunity in the Gulf for use in the preparation
of synoptic weather maps and climatic charts and in compiling monthly pilot
charts. During the early 1900's, various other government agencies began gathering
data relating to the physical environment in the Gulf of Alaska. The u.S. Navy Oceanographic
Office, as well as various scientific institutions, began taking salinity
and water temperature readings in specified areas for the purpose of calculating geostrophic
currents and describing the thermal/density structure of the Gulf. At the
same time, onshore weather stations were being established in such places as Kodiak,
Yakutat, and Middleton Island. Observations at these stations were used to improve
synoptic weather maps and nearshore forecasts.
All of these programs were undertaken with certain objectives in mind, i.e., preparation
of tide prediction, pilot charts, and weather maps. Much of the data was
more than adequate to satisfy the intended mission, but is of very limited use in
planning offshore operations. Early weather data gathered by ships-of-opportunity
is generally inferior due to lack of appropriate instrumentation and properly trained
observers. Coverage was obtained infrequently and only along shipping lanes, leaving
large areas for which few or no reports are availaole. Onshore stations are frequentl
located too far inland or are excessively modified by onshore conditions to be of
use in determining conditions on the open Gulf. The geostrophic current values calculated
from data taken on oceanographic cruises are only very general averages; and
because the method of computation assumes deep water (usually 1,000 meters), it is
inapplicable to currents on the shallower continental shelf. Long-term, good data
during storms for offshore winds, waves, and currents are unavailable because of both
the desire of prudent operators to avoid storms and the problems of making measurements
in storms.
The petroleum industry, oowever, needs site-specific data on waves, wind, and other
related phenomena under both the best and worst possible weather conditions. Since
exact drilling sites are at present unknown, a large body of data describing the conditions
in the Gulf as a whole is desirable. This information will then be used to
design efficient structures and cost-effective operations. Fbr these reasons, the
industry has embarked on a number of meteorologic and oceanographic data gathering
20 McLeod 2

programs for its own use. These programs, as well as some closely related studies,
are described hereafter in this paper and listed in Table 1.
WAVE CLlMA'lOImY
In 1967, the petroleum industry initiated a Group Oceanographic Study, Gulf of
Alaska (GOSGA) to develop an estimate of the wave climatology of the Gulf of Alaska
oy using a hindcasting technique. The effort was large both in terms of the areas
to be investigated, Figure 1, and the 30-month length of the investigation.
Five years and thirty storms were selected as representing normal years and extreme
events through a process of reviewing and analyzing all past weather data for the
Northwest Pacific. Hindcast of these five years (along and with) one year of wave
measurement off Yakutat were made. These hindcast results gave the energy distribution
for the waves in ten period bands so that the specific height and periodsensitive
effects of waves on drilling vessels of all sizes could be evaluated.
The GOSGA hindcasts are probably the largest collection of such good quality data in
existence. However, as a direct verification, suppplement, or (if necessary) a replacement
of the GOSGA hindcasts, there existed a need for a set of high quality
measured wave data in the Gulf. These two studies tend to complement each other.
To satisfy this need, the industry coI!lllenced the Gulf of Alaska Wave and Wind
Measurement Program (GAWWMP) in March of 1974. This study will, at the same time,
provide a firm data base for reliably performing the following tasks with offshore
operations:
a) Evaluating the operational efficiencies of floating equipment and systems
(drilling rigs, pipe laybarges, derrick barges, supply boats, floating storage,
etc.).
b) Delineating optimal seasons for drilling and construction operations.
c) Describing the fatigue-producing wave climate for structural design.
d) Calibrating both hindcasting and forecasting models.
i) Long-term severe storm risk evaluation.
ii) Operational weather forecasting and planning of operations.
e) Developing an estimate of severe sea-state recurrence conditions and
resultant risk.
i) Using direct extrapolations of wave data.
ii) Possibly using newer methods of extrapolating wave and wind data.
The program has set wave rider buoys in clusters of three, at each of five locations
in the Gulf of Alaska, from Sitkinak Island on the west and Yakutat on the east.
All Waveriders carry onboard data loggers, schematically shown in Figure 2. In
addition to these internal recordings at all of the clusters, buoys at three of the
clusters are monitored by telemetry at recording stations established at Sitkinak,
Middleton, and Yakutat. At the other clusters near the Barren Islands and offshore
of Icy Bay, the data is only being recorded by the onboard data loggers. Weather
stations complement the buoy wave observations at the three shore recording stations.
These systems have been in operation since early SeptentJer 1974, and the present
21 McLeod 3

DATE
1. 1889-PRESENT
2. 1892-PRESENT
3. PRE-1899-PRESENT
N
N 4. PRE-1899-PRESENT
5. 1922-PRESENT
6. 1927-PRESENT
7. 1971-PRESENT
8. 1972-PRESENT
9. 1974-PRESENT
OPERATOR
COAST & GEODETIC SURVEY
(NOAA-NOS)
NOAA-NOS
NATIONAL WEATHER SERVICE
(SHIPS OF OPPORTUNITY)
NATIONAL WEATHER SERVICE
(SH I PS OF OPPORTUN ITY )
U.S. WEATHER BUREAU (NWS) ,
FAA, & COAST GUARD
OCEANOGRAPHIC INSTITUTIONS
& NAVY OCEANOGRAPHIC OFFICE
UNIVERSITY OF ALASKA
NOAA-SDS
NOAA-NOS
TABLE 1
SUIfo1ARY OF METEOROLOGICAL AND PHYSICAL OCEANOGRAPHIC
MEASUREMENT PROGRAMS IN THE GULF OF ALASKA
PART 1 - GOVERNMENT
PARAMETERS MEASURED
TIDE & SURFACE WATER
TEMPERATURE AT SHORE
TIDAL DRIFT CURRENTS
SURFACE SYNOPTIC
WEATHER
SHIP'S DRIFT
ONSHORE WEATHER
SALINITY & WATER TEMP­
ERATURE VS. DEPTH
TSUNAMIS
CLDUD COVER, WATER
SURFACE TEMPERATURE
CURRENTS
TYPE OF RECORDING INSTRUMENT
TIOE GAGE (FLOAT -WELL) &
BUCKET THERMOMETER
DRIFT POLES, VESSEL SET
METEOROLOGICAL INSTRUMENTS
AND VISUAL
VISUAL FIXES. LORAN LATER.
METEOROLOGICAL INSTRUMENTS
AND VISUAL
NANSEN BODLES WITH REVERSING
THERMOMETERS (PRIMARY TOOLS)
PRESSURE SENSOR
SATELLITE SENSORS - VISUAL,
IR & MICROWAVE
CURRENT BUOYS USING
AANDERAA METERS
OBJECTIVES
PROVIDE BASIS FOR TIDE
PREDICTIONS FOR PUB­
LISHING TIDE TABLES.
MEASURE TIDAL CURRENTS
AT BAY ENTRANCES,
NARROWS & OFF HEADLANDS
FOR PUBLISHING TlOE
CURRENT TABLES.--
FOR PREPARATION OF
SYNOPTIC WEATHER MAPS
& CLIMATE CHARTS.
COMPILE MONTHLY PILOT
CHARTS.
IMPROVE SYNOPTIC
WEATHER MAPS AND NEAR­
SHORE FORECASTS.
CALCULATE GEOSTROPHIC
CURRENT. DESCRIBE
THERMAL/DENS ITY STRUCTURE.
MEASURE TSUNAMI HEIGHT IN
INTERMEDIATE WATER DEPTH.
PROVIDE STORM LOCATION
& INTENSITY TO IMPROVE
WEATHER FORECASTS.
MEASURE 4 OR 5 LOCATIONS
IN NORTHEAST PART OF GULF.

N
'"
Transmitter
Figure 2
Radio Antenna
Internal
------
Logger
Circuit Board
Water Surface
E levat ion
Wove Propagation
DATAWELL WAVERIDER

N
\0
LEGEND
@ BBN/IRC Current meter buoy
-?- NOAA Current meter buoy
GULF OF ALASKA
Figure 4 Current Meter Measurements

W
I-'
• Synoptic Observations
o Aviation Observations
Figur. &
Alaskan Observing Stations
NOAA 1975. Weather Stations List for Alaska. NWS Alaskan Region.p.1I
o
950

Site
Yakutat
(airport)
Middleton Is.
(airstrip)
Kodiak
(City)
cape Spencer
(C.G. Lights)
Yakataga
(airstrip)
Cordova
(airport)
cape St. Elias
(C.G. Lights)
cape Hinchin Brook
(C.G. Lights)
Seward
(airport)
Homer
(Spit)
Sitkinak Is.
(C.G. Lights)
TABLE 2
AVAILABLE COASTAL METEOROLCX;ICAL DA'rA
GULF CF ALASKA AREA
Type of Information
wind Precip Bar.Pres.
No. Yrs No. Yrs No. Yrs
29 35 35
26 35 35
25 25 25
Unk. 17 Unk.
16 30 30
24 27 27
Unk. Unk. Unk.
Unk. Unk. Unk.
Unk. 49 49
14 34 34
Unk. 16 Unk.
32
Corrments
Approx. 143,824 wind
obs by '74
Approx. 173,824 wind
obs by '74
Approx. 59,824 wind
obs by '74
Approx. 46,216 wind
obs by '74
Approx. 26,304 wind
obs by '74
Approx. 60,824 wind
obs by '74
Approx. 52,056 wind
obs by '74
Approx. 14,130 wind
obs by '74
Approx. 33,912 wind
obs by '74
Approx. 66,824 wind
obs by '74
Approx, 21,608 wind
obs by '74
McLeod 14

light level images and thermal infrared (8-14 micron) images have been received
since 1973. These data are on 9-inch wide film which covers a 1600 nautical mile
wide swath. 'llie infrared images provide a 2 nautical mile wide resolution and
about 1.6 C thermal resolution, which is sufficient to define significant thermal
boundaries on ocean surface waters. 'the simultaneous visible (video) image has
1/3 nautical mile resolution and is useful for separating cloud-free regions over
the open ocean.
On Sunday, July 23, 1972, the Earth Resources Technology Satellite (ERTS-l or Land­
Sat) was launched into a sun synchronous polar orbit around the earth. The satellite
has the capability for producing coverage of most of the earth on an l8-day repetitive
cycle.
Visible and IR video are currently available from the NOAA-3, NOAA-4, and GOES
satellites. NOM-3 and 4 are identical polar-orbiting satellites similar in characteristics
to the DMSP described above, but with a superior thermal resolution
potential (1/4 C), and producing coverage only every 12 hours (10 a.m. and p.m.
Pacific Time). GOES is in synchronous orbit and has much lower resolution, but
it can provide continuous data in a motion picture time/lapse format. For study
of very large circulation cells, the NOAA data are readily available in near
real-time for forecasting use bY independent groups.
The availability of meteorological satellite data at intervals as frequent as 6
hours (DMSP) and 12 hours (NOM) now offers the possibility of synoptic time lapse
studies of surface waters over wide areas of the Gulf which may be cloud-free for
successive orbits. The basis for such studies is the recognition of surface visible
patterns due to suspended sediments in the nearshore region and thermal infrared
image temperature patterns in surface waters offshore, or nearshore as well. During
cloud-free periods, successive satellite images reveal changes in these patterns
from which the direction and magnitude of currents over a wide area can be determined.
Preliminary landSat studies have confirmed the generally accepted views of the circulation
of the Gulf.
Fundamental to the function of long-term weather forecasts are synoptic temperature
data over the Gulf. Satellite thermal infrared data offer the possibility of filling
in such temperature data over cloud-free regions, thereby supplementing surfacebased
measurements. The advent of mircowave radiometry in regions of the spectrum
free from the effects of water vapor absorption or sea state, should provide an allweather
capability for such remote temperature measurements.
Future plans include the launching of a satellite, termed SEASAT-A, in 1978 as part
of NASA's Earth and Ocean Physics Application Program, (EOPAP). 'llie SEASAT-A mission
is the first major step in developing and demonstrating a global ocean dynamics
monitoring system using space measurement techniques to provide information to the
users of the earth's oceans. It will provide inq;>roved data for the following
applications :
1) Predictions of wave heights, wave spectra and wind fields for ship routing,
ship design, storm-damage avoidance, coastal disaster warning, coastal protection
and development, and offshore power plant siting;
2) maps of current patterns and temperatures for ship routing, fishing, pollution
dispersion information, iceberg hazard avoidance;
3) charts of ice fields and leads for navigation and weather prediction; and
34 McLeod 16

e active at anyone time. To focus solely on one design factor at a time is a
mistake; a proper design methodology should take this concurrent multiplicity of
hazards into account.
Therefore, an analysis method which will treat both the effects and likelihoods
of all possible environmental load conditions on a platform is needed. The second
study sponsored will treat this problem. The probability distribution for the resistance
of the structure to any given combination of loads will be determined, so
that the likelihoods of overloads (i.e., failures either partial or total) can be
assessed. Naturally, inputs from the first study on seismic risk are important.
Possible interactions between loadings which result in the total loading not being
a linearly additive combination, will be considered. Additionally, the time-dependent
strength of the platform, a necessary consideration because of the structural
fatigue due to the severe wave climate and expected earthquake activity, will be
treated. The results will be a means for a rational treatment of the entire
spectrum of loads and resistance of an offshore platform.
Superstructure Icing
Drilling vessels, as well as crew boats, operating in certain northern areas face
the hazard of superstructure icing. Ice can often form on a ship at rates in excess
of one inch per oour. Excessive ice buildups create problems of stability in vessel
handling, which can eventually lead to catastrophe. Earlier this year, a program
undertaken to study the characteristics, extent, frequency of occurrence, and methods
of control of superstructure icing. Specifically, the extent of the Superstructure
Icing Program (SIP) study was as follows:
1) The SIP was a comprehensive literature review within the bounds of what was
found in the open literature. This constituted a logical first step, particularly
with regard to the reviews and analyses of the Japanese and
Russian activities in this area.
2) The search parameters included:
a) Past synoptic weather situations for severe superstructure icing in order
to relate the hazard potential to weather predictions, and then to index
such parameters producing the icing conditions;
b) probability of occurrence and maximum thickness by geographic area.
3) The final report covered superstructure icing due to wind-blown spray and
atmospheric precipitation. The participants received general design
principles by which the rate of icing accumulation and extent could be computed
for the Gulf of Alaska and other geographic areas.
By looking in to past synoptic weather situations where known superstructure icing
has occurred, we were able to relate, in a limited way, the hazard potential to
weather predictions, and then to index such parameters producing the icing conditions
- - in short, a definition of conditions necessary for the formation of
atmospheric icing, such as freezing fog, rate of accumulation and probability of
occurrence.
This information will be extremely useful in designing and scheduling of operations
of low-freeboard supply boats in the Gulf of Alaska. SOme guidance is provided for
the design and operation of semi-sul:mersibles, drillships and offshore platforms
36 McLeod 18

SECTION 3
INSTRUMENTATION
Bettignies, C.
Energy SuppZy in the Arctic
(University of Moncton, New Brunswick, Canada)
Buck, B. M., W. P. Brown, S. P. Burke, and E. G. Kerut
TWo RecentZy DeveZoped Arctic Data Buoy8
(Polar Research Laboratory, Incorporated, California, United States, and,
NOAA Data Buoy Office, Mississippi, United States)
Corell, R. W., T. McGuirk and W. Lenharth
TechnicaZ and Logi8tics ProbZems AS80ciated with Air DepZoyabZe Instrumentation
SY8tems in the Arctic Basin
(University of New Hampshire, United States)
Gray, G. M.
Effective In8trumentation DeveZopment for Arctic Ocean Engineering Re8earch
(Department of the Environment, University of Washington, United States)
Haugen, D. P. and E. G. Kerut
DeveZopment and U8e of the Arctic Data Buoy
(University of Washington, and NOAA Data Buoy Center, Bay St. Louis,
Mississippi, United States)
Lewis, E. L.
Oceanographic Instruments for Arctic U8e
(Department of Environment, British, Columbia, Canada)
39

ENERGY SUPPLY IN THE ARCTIC
Christian Bettignies
University of Moncton
Moncton, New Brunswick
Canada
EXTENDED ABSTRACT
This paper deaZs with unconventionaZ sources of energy and their possibZe utiZizations
in the Arctic environment. Power generation using wind and soZar energy wiZZ be reviewed
as weZZ as some other aZternate sources.
So Zar Energy
Summer utiZization of SoZar Energy is sometimes possibZe in the arctic due to the Zong
periods of sunshine in summer.
Production of Energy by soZar power is presentZy being utiZized and some actuaZ appZications
wiZZ be discussed; i.e •• to power geophysicaZ data recorders. to power Zighthouses.
transmitters. teZephones. etc •••• and to pump water.
SoZar Heating or air conditioning of buiZdings. which is feasibZe at some Zocations.
wouZd require some stand-by power in northern sites. StiZZ. it couZd save energy by
using this form of energy in summer. SoZar heating technoZogy components are stiZZ
quite expensive. however. their price shouZd decrease in the future and render soZar
energy more competitive.
wind Energy
In isoZated regions or in the Arctic. smaZZ and medium size aerogenerators can provide
a viabZe aZternative to some power needs. Indeed. in these regions. due to the remoteness.
the scattering of the demand of energy. reguZar power pZants are expensive to buiZd
and to run.
SeveraZ ezperiments reZating actuaZ tests done in the past wiZZ be presented. SimiZar
appZications are possibZe in the arctic where wind is pZentifuZ.
Indeed.
a)
with modern airscrew aerogenerators. it is sometimes possibZe
to power geophysicaZ data recorders and navigation aids:
i. e •• a Zighthouse was powered by a 2-bZade. 5.7 m diameter. 5 KW aerowatt
windmiZZ. In his study. Mr. Prunieras (1966). compared a conventionaZ system
consisting of 2 dieseZ motors of 13 HP coupZed to a 6.25 KVA aZternator with
cadmium-nickeZ batteries (110 V. 310 AH). to a wind power system using a 5 KW
aerogenerator. a stand-by dieseZ power pZant and Zead batteries (500 AH. 115 V).
41

In this aase, the windmiU is used as a fueL saving deviae and that can be ver>y eaonomiaaL
if you know that on a 5 year experiment, the aeoLian produaed eaah year, between 70
and 90% of the totaL energy required by the Ughthouse.
SimiLar set-ups aan be utiLized for powering geophysiaaL data reaorders. UsuaLLy, batteries
store enough energy to aope with no wind aonditions of 2 to :5 lJJeeks duration. Instruments
enaLosUPes aan aLso be kept within aertain thermaL Limits by these aerogenerators.
Data from ver>y isoLated sites aan be sent to sateLLites and then beamed baak to
earah at presaribed Loaations.
b) to purify or desaU lJJater:
SeveraL sahemes have been proposed. Mr. SoLomon (1968) proposes to use "Large
aerogenerators aoupLed to thermo-aompression distiLLators. The system is fuLLy
automatia.
a) To power radios, TVs, teLephones, miarowave reLay stations, eta •..
SeveraL exampLes of suah utiUzations wiU be presented.
d) To eLeatrify dzueUings and viUages: ..
An experiment wiU be desaribed in detaiL. It shows how an AUgaier 7.5 KW windmiU
lJJaS used to power a farm. This test (J. G. WaLker, 1960) used an automatia
Loading deviae, thanks to whiah most of the avaiLabLe energy in the wind lJJas empLoyed.
The eLeatrifiaation of viLLages is possibLe as, in the aratia, the instaLLed aapaaity
is sometimes quite smaU and, in the past, Larger aerogenerators of 25 to
250 KW have been buiLt and used for eLeatrifiaation purposes.
In the Soviet Union, many medium size windmiLLs have been used to produae shaft
power and eLeatriaity in their isoLated Siberian regions.
Some detaiLs wiLL be given as weLL as the resuLts of a teahniao-eaonomia study
showing the feasibiLity of using these modern airsarew aerogenerators in a
northern Quebea Loaation (C. Bettignies).
e) Cathodia proteation of pipeLines:
WindmiLLs have been used for this purpose in severaL aountries. Non-saarifiaiaL
anodes aan be used with an externaL sourae of E.M.F. produaed in this aase by an
aeoUan and aorrosion is then prevented.
f) To pump lJJater:
ExampLes are numerous and some wiU be briefiy disaU8sed.
g) To heat and refrigerate:
DweLLings or enaLosures aan be heated or refrigerated or aerogenerators aan be
aLso used to prevent the freezing of utiLities.
Storage of Energy
Different means of storing energy wiU be presented. They wiU be aompared to stand-by
power, whiah on a Larger saaLe, is more feasibLe.
Cost of Energy
Power generation is possibLe at priaes ranging between 5 to 25 a per KWHR depending on the
windmiLL used and on the avaiLabLe energy whiah can be produaed from the wind at the site.
42
Bettignies 2

TWO RECENTLY DEVELOPED ARCTIC DATA BUOYS
B. M. Buck, W. P. Brown, S. P. Burke
Polar Research Laboratory, Inc.
Santa Barbara, California
United States
and
E. G. Kerut
NOAA Data Buoy Office
Bay St. Louis, Mississippi
United States
ABSTRACT
TWo types of data buoys employing different means of data telemetry have been recently
developed by Polar Research Laboratory, Santa Barbara, California for general usage in
polar regions. Eighteen of these buoys were installed in the Beaufort Sea in April-May
1975 as part of the l4 month long AIDJEX.
The Arctic Environmental Buoy (AEB) is a remote unattended data acquisition and HF telemetry
system designed under contract with the NOAA Data Buoy Office for deployment on ice
covered seas. The total system as presently configured consists of 8 AEBs on a 400 km
radius circle and a Central Control Station (CCS). The Central Control Station under
computer control collects the data from the AEBs, processes the data and formats it on a
digital tape for future analysis. The CCS is also capable of controlling the majority
of the AEB functions via a command link. The AEB is configured to automatically sample
sensor and position data at three hour intervals. The present sensor configuration allows
six primary sensors with lO bit resolution and l6 auxiliary sensors with 5 bit resolution.
The auxiliary sensors are sampled only once per day. The sensor and position data are
stored in a digital memory which is transmitted via on H.F. link once per day to the
Central Control Station. A unique dual recireulating memory concept is utilized to prevent
data loss due to propagation vagaries and Polar Cap Absorption events. Position
measurements are accomplished at each AEB by an on-board NAVSAT receiver which provides
satellite identification and doppler measurements that are matched with NAVSAT data obtained
at the Central Control Station to provide a fix for the drifting AEB.
The Synoptic Random Access Measurement Station (SYNRAMS) was developed under an Office
of Naval Research Contract to automatically measure and record underwater ambient noise,
barometric pressure and air temperature data for relay to the NIMBUS-6 satellite. Data
is measured and stored in a solid state COS/MOS memory every three hours, the newest data
replacing the oldest data in memory. Station location is obtained through doppler
measurements by the satellite. Carbon zinc air batteries provide the energy necessary
for a two year system life. The eZeatronias and batteries are housed in a l6 foot aluminum
tube which extends through the paak ice into the relatively warm water below. SYNRAMS
can float as a spar buoy during the summer if it melts free. Ten SYNRAMS buoys were
suacessfully deployed this spring with seven adJacent to the AEBs.
45

These early experiments indicated that both the HF and satellite telemetry systems
provided cost effective means of Arctic data retrieval. The relative advantages
and disadvantages of the two techniques are given in Figure 1.
Both of these approaches were in 1974-1975 developed into practical Arctic data
buoys for use in the Arctic Ice Dynamics Joint Experiment (AIDJEX). This l4-month
experiment that started in March 1975 is designed to investigate the large scale
response of sea ice to changing environmental parameters. The AIDJEX program as
presently envisioned is the first of a series of studies that will subsequently be
incorporated under a larger Polar Experiment (POLEX). The objective of AIDJEX is
to reach, through coordinated field experiments and theoretical analyses, a fundamental
understanding of the dynamic and thermodynamic interaction between arctic
sea ice and its environment and to answer basic questions of the mechanisms which
cause large scale ice deformation and the effects of ice deformation and morphology
on the heat balance. Since underwater acoustic ambient noise is caused primarily
by ice dynamics, the simultaneous measurement of acoustic noise levels with wind
drag and ice movement will provide a better understanding of the causes and mechanisms
of noise, possibly leading to a prediction model. For this reason an acoustics
experiment was added to the AIDJEX project.
The NOAA Data Buoy Office (NDBO) engineering development activities include the
development of arctic data buoys in support of national and international scientific
experiments. As part of these activities a program was initiated to develop and
test three prototype high-capacity HF buoys called Arctic Environmental Buoy (AEB)
for first use on the AIDJEX. The AIDJEX Project Office set forth the requirements
for data handling and general performance specifications for these buoys. PRL
applied the technology learned in the previous work on LORAMS to this problem and
deSigned and constructed the three prototype buoys. Since the AIDJEX plan called
for an array of 8 to 10 of these buoys along with a manned Central Control Station
(CCS), the National Science Foundation, the major AIDJEX sponsoring activity, funded
the construction of seven more AEBs and the CCS.
Realizing the unique opportunity afforded by the large scale AIDJEX to learn more
of arctic environmental acoustics, the Office of Naval Research contracted PRL to
design and implement an ambient noise experiment coordinated with AIDJEX. For this
task it was decided to develop a smaller buoy utilizing the soon to be launched
NIMBUS F satellite. The reasons for a second type buoy were threefold:
the AEB was to employ a large HF antenna that would be very
difficult to noise-quieten and therefore could contaminate
underwater ambient noise measurements.
there was a need for a location and barometric sensor backup
for the relatively expensive AEBs.
there existed a need for the development of an economical,
easy to install arctic data buoy for future scientific uses
outside of AIDJEX.
This second buoy, called Synoptic Random Access Measurement System (SYNRAMS), was
to be installed adjacent to each AEB, preferably on the same floe, but far enough
removed to prevent both acoustic noise and radio frequency interference between
the two.
After completion all data buoys were shipped to the Arctic and installed by AIDJEX
and PRL personnel during the spring of 1975. Eight AEBs and 10 SYNRAMS were implanted
on a 400 km radius circle around the main AIDJEX manned camp as shown in Figure 2.
As of this time all 8 of the AEBs and 8 of the 10 SYNRAMS are in operation and
providing data on a continuous basis.
47 Buck et at. 3

periods of up to two years. An array of 10 of these ice stations was installed 250-
550 nautical miles north of the Alaskan coast during the spring of 1975. In each
station, 24 hours worth of the most recent data, made up of eight 32-bit words,
are retained in memory for burst transmission to the RAMS (Random Access Measurement
System) receiver in the polar orbiting NIMBUS F satellite. Surface platform location
to a CPE of about 5 km is obtained through doppler measurement of the transmitted
signal.
The SYNRAMS ice station is illustrated in Figure 4. The drawing shows a cross section
of 3.1 meters thick sea ice with about a half a meter of snow on top. A portable
gasoline-powered ice auger was used to drill a 23 cm diameter hole in the ice through
which the 20 cm diameter, 5.2 meter aluminum tube was installed.
The basic SYNRAMS ice station functionally consists of a solid state memory providing
data for the Random Access Measurement System (RAMS) satellite data relay platform.
The memory receives its information from a digital data formatter controlled by
crystal oscillator-based timing electronics. The formatter receives data from
sensor signal conditioners. The sensors used include a barometer, air temperature
sensor and a hydrophone. The primary battery bank supplies power to assorted power
conditioners which provide the necessary supply voltages required by the system.
The noise measurement hydrophone and preamplifier is located 30 meters below the
water level tethered to this aluminum tube by a length of passing link chain. A
6.1 kg teardrop lead weight is connected to the end of the chain. The hydrophone!
preamplifier cable is tied to the chain at 0.3 meter intervals, and "haired" fairing
attached to each loop to reduce the effects of current induced strumming. This
135 kg payload causes the 5.2 meter tube to float with its top about one meter above
the ice surface. At the top of the tube is a circularly polarized transmitting
antenna (especially designed for this application) protected by a plastic fairing.
A barometer and crystal oscillator for the RAMS transmitter is contained at the
extreme lower portion of the tube. These temperature sensitive components are
located here to take advantage of the very stable temperature of the water directly
below the ice.
The RAMS module is located above the barometer and oscillator unit. The lower
portion of this module contains an energy storage capacitor bank for the RAMS I-watt
transmitter. The other portion of this module contains a power oscillator, digital
encoder, power amplifier and power supply conditioners. The RAMS unit transmits a
one-second data burst, consisting of 64 bits, once every minute. These 64 bits
contain 32 bits of data, an identification word, and various system synchronization
words. Each 32-bit data group corresponds to one three-hour synoptic sample occurring
within the last 24 hours. Eight such transmissions cover a one-day memory
period, and then the data repeats. The reception of at least eight consecutive
transmissions within the l6-minute period of a pass is necessary to recover the
timing of the data samples.
The memory and control module is positioned directly above the RAMS. This module
contains all sensor signal conditioners, digital timing, data memory and control
circuitry. The module also contains a test panel used in checkout and system
setup. Connectors at the top of this module provide connections for primary power,
hydrophone preamplifier, RAMS RF output and a system test set.
Ten 1.2 volt carbon-air primary batteries are stacked above the memory and control
module. They are series connected to generate the +12 volt, 1000 amp-hour primary
power source to power the station for a two-year period.
52 Buck et at. 8

A special circular polarized helix antenna is mounted on top of the aluminum tube
covered with a ten inch diameter polyethelene fairing.
The air temperature sensor is mounted next to the antenna on a wood block. The
wood serves to thermally insulate the thermistor from the aluminum tube.
CONCLUSIONS
The complete array of 10 SYNRAMS was installed in May-June 1975. Preliminary tests
with the recently orbited NIMBUS F satellite indicate that eight of the ten are
operating satisfactorily and providing continuous data.
Eight AEBs are now operational on the Arctic ice pack, being used to gather data
necessary to finalize a mathematical model of pack ice dynamics.
Both data buoys have been designed with the flexibility to easily adapt to a wide
variety of other sensors and can be used wherever periodic data over a long term is
needed, and at a very small fraction of the cost of manned ice camps for similar
functions.
REFERENCES
1. Olenicoff, S. M. The Soviet DARMS.
AIDJEX Bulletin No.7, April 1971 and No. 22, August 1973.
2. Buck, B. M., McLennan, M. and Springer, M. January 1973. Underwater
Acoustic Measurements in the Arctic Ocean Using a Tropospheric Scatter
Radio Telemetering System (U), General Motors-DRL Report TR63-201, (Conf)
3. Buck, B. M. October 1972. LORAMS and SHRAMS - Two Unmanned Drifting Data
Collection Systems for Ice-Covered Seas. Presented at World Meteorological
Organization <strong>Conference</strong>, Tokyo, Japan.
4. Haugen, D. P. and Kerut, E. G. 1975. Development and Use of the Arctic
Data Buoy. POAC.
54 Buck et at. 10

TECHNICAL AND LOGISTICS PROBLEMS ASSOCIATED WITH AIR DEPLOYABLE INSTRUMENTATION
SYSTEMS IN THE ARCTIC BASIN
R. W. Corell, T. McGuirk and W. Lenharth
University of New Hampshire
Durham, New Hampshire
United States
EXTENDED ABSTRACT
As the initial phase of the research effort at the University of New Hampshire to develop
a Ughtweight coring system for operation in the Arctic, a study was conducted to select
the aircraft most suitable for supporting scientific research over large portions of the
Arctic Basin. The recommended aircraft would be required to transport the coring unit to
various sites throughout most of the Central Arctic Basin.
Early in the study the decision was made to review only those aircrafts which had previously
proven themselves in an arctic environment. Based on this, as well as several other
factors, (arctic availability, STOL capabilities, history of use in cold weather, ranges
beyond a few hundred miles and available manitenance support) the candidate aircraft were
chosen to be, USAF C-130A, UNS-LC-130F, R, commercial C-130, TWin Otter, Caribou, Buffalo,
and Helicopters, UH-1B, UH-1N, and The Lamma.
The eight candidate aircraft were compared using the foUowing criteria: 1) maximum
range, maximum distance to research site with fuel remaining for return flight, 2) fuel
consumption at cruise speed, 3) cost to user on a per hourly basis, 4) maximum payload,
5) cruising speed, 6) engine type, 7J operating cost per mite, and 8) maximum payload
with maximum fuel on board. This information is presented in Table 1. The sources from
which this information was obtained are also listed.
Aircraft operations in the Arctic depend very much upon the weather conditions at any
particular time and therefore due consideration was given to climatological factors duPing
this analysis. Low temperatures are capable of halting aircraft operation; however, in
the Arctic the temperatures do not appear low enough, nor do they last long enough, to
produce much more than a temporary and locaUzed effect. The wind speed wiU only be a
factor if it produces an adverse cross wind component duPing landing or take off and for
most of the larger aircraft, this seldom occurs in the Arctic. Since the smaller aircraft
require shorter runuxzys, they are capable of choosing alternate runuxzys to reduce this
cross wind component. Visibility becomes most important in arctic operations and clear
visibility for 3 miles with a ceiling of 1000 feet seems to be a reasonable minimum safe
visibility. Minimum operating temperatures and maximum cross wind components for the aircraft
under consideration appear in Table 2.
55

Aircraft Designation
ruin Otter
Caribou
Buffalo
(USAF-Comm) C-130A
LC-130R
UH-1B
UH-1N
Lamma
NOTES ON TABLE 2
TABLE 2. Arctic Operational Restrictions
Min. Operating Temperature Maximum Cross wind Component
About _45 0 F None (1)
18 KTS
25 KTS
35 KTS
35 KTS
40 KTS (2)
40 KTS (2)
40 KTS (2)
1. From operator at resolute. Because STOL Capability the runway can be abondoned and
any flat area used.
2. wind of this magnitude in any direction of gusts of 20 KTS of more.
A condition caUed "white out" which resuUs from ground fog or low level clouds, viewed
against the snow or ice surface giving the appearance of total whiteness can produce a
loss in depth preception and orientation making low level flying very dangerous.
A final factor which must be considered in planning operations at new remote ice stations
is the availability of daylight. Since the amount of light in a day changes with the time
of year and the location of the site, there are "windows of daylight" during which conditions
are favorable. However, since the warmer days generally occur during periods of extended
daylight, the surface conditions of the ice may not permit landings or take off
even though there is sufficient light.
Also included in this paper is a review of available air bases located in Canada, Greenland,
or the U.S.A. from which arctic research projects could be staged. The descriptions
of the bases include runway sizes, sophistication of vanigational aids, and availability
of adequate maintenance and crew facilities.
An overview, rating the five aircraft (helicopters having been eliminated due to range
and payload limitations) for six capabilities that appeared to bear on arctic operations
was prepared. The comparison matrix which was to be used as a guide in making the final
decision is presented in Table 3.
Based upon the data presented in Tables 1, 2, and 3 as well as additional information obtained
through communication with aircraft manufacturers and users, it appeared that the
C-130 was best suited for support of research projects in the Arctic. The Buffalo was
the only other aircraft which could possibly be used. Both the ruin Otter and the Caribou
have limited range and payload capabilities. It is also obvious that based upon cost the
VXE-6 military C-130 be recommended over the commercial C-130.
57 Corell et az' 3

As a step in gaining a better understanding of the penetration of a core barreZ into seafZoor
sediments, work has been started on a computer program describing this phenomenon.
PresentZy the program is in the earZy stages of deveZopment, but preZiminary resuZts have
reveaZed some interesting reZationships between the amount of penetration obtained and
severaZ parameters directZy reZated to penetration. In its finaZ form the program win
hopefuZZy aid in designing a better system for sampZing from the ocean fZoor.
ABSTRACT ONLY AVAILABLE
60
Corell et al. 6

EFFECTIVE INSTRUMENTATION DEVELOPMENT FOR
ARCTIC OCEAN ENGINEERING RESEARCH
Gordon M. Gray
Director, Ocean Engineering Research Laboratory
University of Washington
Seattle, Washington
United States
ABSTRACT
The philosophy of funding aratia researah and the utilization of teahnology to support
that researah are examined. Using the Naval Aratia Researah Laboratory (1974) program
as an example it is shown that the average ratio of field time to home time for all
projeats supported was 1:2.5. This is aonsidered to be unneaessarily high, and due
primarily to traditional funding praatiaes for aratia researah, whiah do not take into
aaaount effiaienaies available through long-life, autonomous, multi-sensor instrumentation.
Other examples are given wherein the ratios of field-to-home time are 1:15 and
1:7.7 respeatively. The suaaess of these latter programs depended, among other things,
upon the use of instrumentation designed to minimize the time spent by a saientifia
team in the field. This approaah is reaommended for future aratia researah, and a set
of poliay guidelines is provided whiah aan areate the management inaentives needed to
switah from present praatiaes. Several high performanae aratia instrument systems are
desaribed and given as examples of the type of equipment needed to reduae the field-tohome
time ratio for future aratia saientifia projeats.
61

The demand for tools to carry out petroleum resource exploration in the Arctic Ocean has
come on with such speed during the last few years that the small community of scientists
and engineers that traditionally carries on arctic research has found itself quite unprepared
to respond within the time frame of new commercial needs. For decades spartan
budgets have characterized the U.S. Government's investment in arctic research. During
that time, the explorer/investigator learned to made do with equipment that was often
garnered from the "excess" lists of his colleagues in oceanography, meteorology, and
geology; and whatever adaptations appeared necessary to insure that these instruments
would survive in the hostile arctic environment were usually made as afterthoughts. Consequently,
now that expeditious resource assessment has emerged as a prime objective in
arctic exploration, the arctic research community finds itself without adequate equipment
to meet the demand. What instruments, for example, are available to study detailed
variations in ice thickness of the polar pack? Or, how does one obtain bathymetric and
geophysical data relating to the bottom of the ice-covered Arctic Ocean without facing
the formidable task of boring countless holes through the ice? And how can an engineer
measure accurately and continuously the movements of the everchanging ice pack which his
exploratory structures must withstand?
These problems and a host of others must be solved, but the sOlution must involve more
than just a few pieces of new hardware. What is needed is a whole new philosophy of
arctic research instrument development and utilization. Our present equipment is not up
to the tasks at hand, generally for two reasons: first, its design usually does not
take into account fundamental requirements for operating in the Arctic, and, second, it
is usually built to satisfy one program and one investigator only. With few exceptions,
today's instrumentation is labor intensive to operate, logistically difficult to handle,
wasteful in its use of power, and marginal in terms of field performance.
What, then, must be done to rectify this situation? Frankly, I believe the arctic research
community needs an infusion of the systems-oriented philosophy and practice that
was used successfuly in space programs. First we must develop reliable, multi-function
instrument platforms, and sensors that can operate at least semi-autonomously for extended
periods. Then we must work together efficiently to maximize our return from the
new equipment. Unfortunately, complex systems take time to perfect. So, even if we are
to embark on this new approach immediately, more than a year would pass before instrument
systems could be assembled, checked out and operationally deployed. Significantly higher
development costs will also be involved, but I believe that these costs can be met by
restructuring rather than increasing the research budget. For example, the practice of
supporting sizable teams in the field for extended periods to perform investigations of
limited scope should be revised in favor of smaller groups that collaborate on diversified
experiments using multi-purpose instrumentation. In 1974, for example, the Naval
Arctic Research Laboratory provided support for more than 17,000 man days of scientific
field work, at a cost--not including scientific program funding--of some $5.7 million. l
If this is compared with the scientific funding for the same programs, one can show that
1 man day was spent in the field for approximately each 2.5 days of work at home. I submit
that this ratio of field-to-home time is too high for almost any research. At the
University of Washington's Applied Physics Laboratory (APL-UW), two recent arctic experiments
recorded field-to-home times of 1:15 and 1:7.7, respectively. The program that
resulted in the 1:15 ratio admittedly had too little field time associated with it. However,
there is general agreement at the Laboratory that a 1:7 ratio of field-to-home time
is a reasonable target for most arctic investigations. As a start for new programs, if
the field-to-home time could just be cut in half--that is, from 1:2.5 to say 1:5--the
$2.9 million saved annually would be enough to revolutionize arctic scientific instrumentation
and provide the incentive for its time-shared use by collaborating research
groups. Again, using a specific example at APL-UW, an under-ice submersible was recently
developed2 which is unquestionably one of the most sophisticated instrument platforms
ever used in Arctic Ocean research. It cost approximately $700,000, or less than one
62 Gray 2

quarter of the money that would be available in one year by changing the field-to-home
time requirements. That cost included the complete design, fabrication, testing and
operation at T-3 for one season. Data for three projects were taken simultaneously and
plenty of sensor and recording space was available for others. Thus, it would appear,
at least within the U.S. Government-supported arctic research programs, that internal
fund restructuring should be adequate to cover most of the costs associated with the development
of new high-capability instrumentation systems. It remains, then, to devise
a new policy for arctic research program support that can effectively change the funding
patterns of the past in a way that will provide incentive for greater efficiency in field
operations.
Such a policy must foster close program planning and a positive attitude among scientists,
engineers and instrumentation developers. On the one hand, the scientists must
be assured that the instrumentation upon which they are to stake the success of their
programs will, in fact, provide reliable, accurate, and sufficient data to satisfy their
needs; but on the other hand, the instrument system designers must be given complete
performance requirements, ample funding, and time to debug and test their equipment completely.
I believe this can be accomplished if the funding managers will adopt the following
three new policy guidelines as part of the review process for all future arctic
research proposals:
(1) establish a field-time to home-time ratio
(1:7 recommended) and use it as a prime
criterion for new program approval
(2) provide liberal instrumentation development
funding for all programs that include a clear
plan for time-shared use of instrumentation in
the field*
(3) establish a dependable arctic instrumentation
development resource by identifying laboratories
or manufacturers who have specific experience
and success in building and fielding arctic
instrumentation.
These technology centers would be placed on a selected source list and would carry responsibility
not only for instrument system development but also the proper conduct of
the complete field measurement program. Thus, if their equipment failed or their field
program was not properly executed, they would be in jeopardy of being removed from the
qualified list.
* For development of support instrumentation systems, the funding algorithm suggested
is F = K (EC/ED), where
F authorized funding to develop instrument system
EC sum of funding approved for all collaborating programs
ED total of all scientific man-days to be spent in field by collaborating programs
K constant, probably >1, which is a function of program complexity and bearing of
work on solution of major arctic problems.
63
Gray 3

None of the foregoing is meant to imply that stagnation has set in, and that no improvements
are being made in the capabilities of arctic instrumentation. Quite to the contrary.
Ingenuity continues to bloom in the fertile beds of technology. My principal
concern here is that our present rate of headway is insufficient. If one considers
just the time needed to made the standard, preliminary offshore geophysical survey, it
would take the United States over 10 years, using present techniques, to cover adequately
the 130,000 square miles of continental shelf between Cape Lisburne and Demarcation
Point. Our neighbors on either side with their much larger shelf areas could be at it
for more than a century! AIDJEX, which is culminating this year, has been a leader in
sponsoring collaborative efforts among scientists and the development of a broadly usable
data bank. Likewise, in Antarctica, development of the Unmanned Geophysical Observatory
has provided a good example of space program-type cooperation among scientists in the
use of a remote, fully-automated station to study upper-atmosphere physics. 3 Both of
these efforts, however, evolved over many years and could be said to have succeeded in
spite of the system, for AIDJEX is still a highly labor-intensive program4 and the UGO
has been beset by years of delay principally due to its small hardware budget.
To cite a few examples of what is being done to advance the state-of-the-art in Arctic
instrumentation, the Unmanned Arctic Research Submersible (UARS) , Figure 1,5,6 is a versatile,
free-swimming, multi-sensor instrument package built to satisfy multiple-user
needs, which, in the Arctic Ocean include:
(a) high field reliability
(b) adequate endurance and depth capability
(c) sensor and data handling capability to
accommodate several field investigations
simultaneously
(d) field portability and adaptability to new
missions
(e) low manpower requirements in the field.
In Figure 2 the submersible is being lowered on its launching rack preparatory to a run
under the pack ice near ice island T-3. UARS is a "sub-compact" minisubmarine that was
designed, built, and put into service in the Arctic in a period of just over a year and
a half. The vehicle carries enough battery power to operate all sensors and its propulsion/control
system for more than 10 hours. Since the device is untethered, it has
a high degree of spatial flexibility in its operational envelope. It can, for example,
be programmed to run over a straight course of more than 30 nautical miles or carry out
shorter range surveys using a variety of three-dimensional tracking configurations. It
was also designed around modular concepts and with ample reserve buoyancy so that a number
of instrument packages can be carried simultaneously, and the vehicle system can be
transported to any place on the ice pack by aircraft.
One type of multi-sensor system which has been given considerable attention--and achieved
remarkable success--in the Arctic during the past few years is the automatic data buoy.
Both the APL Arctic Data Buoy? and the Polar Research Laboratory (PRL) Arctic Environmental
Buoy were designed to meet multiple-user requirements for data collection and to
survive for extended periods (>1 yr) in the pack ice environment. The buoys also had to
be deployable by small crews and transportable by light aircraft that can operate from
unprepared ice surfaces in the Arctic Ocean. Early Arctic Data Buoys were designed to
relay air and water temperature, barometric pressure and buoy position, via NIMBUS satellite,
to a ground station for collection and analysis. Models that are scheduled for
64
Gray 4

Figure 1. Unmanned Arctic Research Submersible (UARS) developed by
the University of Washington's Applied Physics Laboratory
Figure 2. UARS being lowered on launching
rack prior to a data collection
run under the 20- ft thick arctic
ice pack
65 Gr ay 5

field deployment this year will also measure ocean current velocity and direction at two
depths. Between satellite passages, data samples will be collected at 3-hour intervals
and held for transmission. Figures 3 and 4 show an Arctic Data Buoy being installed in
the ice. This buoy was one of an array of six which were being deployed during 1972 in
the Beaufort and Chukchi Seas . PRL buoys are currently being used in AIDJEX. They utilize
NAV-SAT to obtain an order-of-magnitude greater position accuracy than is possible
with NIMBUS -F, and feature an HF radio telemetry link which is used to relay data from
the buoy to the AIDJEX ice camp.
Effective instrumentation does not necessarily equate with complexity and high cost. An
example, at the low end of the dollar spectrum, is the recently-fabricated Conductivity­
Temperature-Depth unit shown in Figure 5. This device was developed by the Applied Physics
Laboratory at a cost of less than $10,000 . The signal processing, data recording,
and power packages are all housed within the hub of the reel; and its weight of approximately
100 lb--which includes 600 ft of cable on the reel--gives this piece of equipment
a "go-anywhere" versatility. This year its successful use from ice pack stations, small
boats and light aircraft has cut the normal time-per-station in half.8
Still another arctic data system with mUlti-program flexibility is now on the drawing
board and should be available within the next few months. It is basically a bottommounted,
or tethered, instrument service platform which can accommodate a number of sensors
and carries power and data storage capacity for an extended period of operation.
A specia1 feature of this system is the method of data retrieval. A high-data-rate
acoustic transmission link will be used. This will obviate the need to retrieve the entire
instrument package in order to recover the data. Where the pack ice is moving rapidly
the acoustic link should prove particularly advantageous. To date the problems
associated with retrieving anything that is anchored to the bottom beneath a sheet of
moving pack ice have prevented the use of extended-time data collection systems in most
of the Arctic.
I would like to conclude by reiterating my concern over the way that arctic research has
been funded in the past, particularly with respect to the development of efficient instrumentation
systems. With the focus of activity, at least in the North American
Arctic, now centering around petroleum exploration, government agencies and industries
other than those with traditional interests in the high latitudes will be taking part in
the formulation of future arctic research policy. I believe, therefore, that it is both
important and timely that current policies and funding practices be reviewed and modified
so that our exploration program for the coming years can be responsive to the new needs.
I have tried to show that there is no shortage of good ideas on how to improve the efficiency
of arctic data collection--and some good systems are evolving--but I submit that
they will be too few and come too late to help in our search for new energy resources
unless we take steps now to insure their proper development and use.
REFERENCES
1. "Analysis of the Functions, Activities and Utilization of the Naval Arctic Research
Laboratory, Barrow, Alaska," Office of Naval Research, 16 September 1974.
2. "Unmanned Arctic Research Submersible (UARS) System Development and Test Report,"
APL-UW 7219, Applied Physics Laboratory, University of Washington, 11 September 1972.
3. "A Program for Use of Automatic Stations in Upper-Atmosphere Physics Research in Antarctica,"
National Academy of Sciences, 1974.
4. "Operations Manual for the AIDJEX Main Experiment," AIDJEX Office, University of
Washington, January 1975.
66 Gray 6

Figure 3. Arctic Data Buoy (ADB) main
body being inserted through
hole in ice pack
Figure 4. Fully assembled ADB, in place, receiving final checkout
67 Gray 7

5. "Unmanned Submersible as an Arctic Research Tool," G.M. Gray, Arctic Logistic Support
Technology Symposium, AINA, November 1971 .
6. "The Unmanned Arctic Research Submersible System," R.E . Francois, MrS Journal,
January-February 1973.
7. "Development and Use of the Arctic Data Buoy," APL-llW 7422, Applied Physics Laboratory,
University of Washington, January 1975.
8. "Light Aircraft Deployable CTD Systems, P. Becker, Plessey Environmental Systems STD
<strong>Conference</strong>, 11 February 1975.
Figure 5. Portable CTD instrument designed for use at
remote arctic sites. Signals are multiplexed
at the sensor head, at left, prior to being
transmitted up the cable to the recording and
display units in the drum hub.
68
Gray 8

uoy l.oaation wiU swwZy ahange as the iae moves. A study of the drift patterns indiaates
that buoys deployed jU8t north of the Alaskan-Canadian aoast generally migrate westerly
at an average rate of perhaps 2 to 3 degrees wngitude per month, with oaaasional
movement as high as 10 degrees per month.
An advanoed version of the aratia buoy, aonfigured with oaeanographia sensors as well as
meteorologiaal sensors, is aonaurrently under development by the NOAA Data Buoy Offiae.
The outer hull wiU aonsist of a single-pieae tube peZ'fllQ7lently seal.ed at the bottom end.
Telemetry and traaking will be aCJaomplished by the Random Aaaess Measuring System (RAMS)
aarried aboard the NimbU8-F satelUte that is saheduled for launah in the swmzer of 1975.
This faU, four of these buoys wiU be deployed north of the Alaskan-Canadian aoast to
support a study of iae dynamias in the shear zone.
ABSTRACT ONLY AVAILABLE
70
Haugen and Kerut 2

OCEANOGRAPHIC INSTRUMENTS FOR ARCTIC USE
E. L. Lewis
Frozen Sea Research Group
Ocean & Aquatic Sciences
Department of the Environment
Victoria. B.C.
Canada
ABSTRACT
Problems in making oaeanographia observations that are peauliar to polar regions are only
in part the result of the alimatia extreme. Equipment must fit through a small hole in
the iae sheet and must hold to a preaise position in the large density gradients
frequently found in aratia surfaae waters proteated from wind mixing by the iae sheet.
The growing iae sheet tends to move the position of instruments suspended from it within
these gradients as does the tidal rise and faU. If the horizontal aomponent of the
earth's magnetia field does not exaeed 6000y, other means must be found to give a
direational referenae.
A review is given of useful teahniques for position fixing and making depth and STD
measurements from the iae surfaae. Deployment systems and instruments for reaording
temperature profiles, tidal range, and aurrents are desaribed.
71

INTRODUCTION
There are both difficulties and advantages in using oceanographic instruments from the
sea ice in comparison to ship borne operations. The ice gives a stable platform for
instrument deployment superior to any ship and, if it is shorefast, is almost as good as
the sea bed as an attachment point for instrument strings; but it must be penetrated for
instrument insertion which greatly limits the acceptable package size. Problems due to
low temperature are less than they used to be. Modern electronic equipment will usually
store quite satisfactorily down to -50 0 C although most manufacturers' specifications of
components cite -40 0 C as the low temperature limit. Instruments may be subject to
thermal shock when placed into sea water but this can usually be avoided as far as
sensitive components are concerned by suitable design. When low temperature operation
is required batteries frequently are the limiting factor.
A sea ice cover 2 m thick presents a considerable barrier to instrument insertion and a
22.5 cm diam. hole is the largest that may be realistically drilled by hand held equipment.
As a requirement for significantly larger holes would preclude helicopter deployment,
instrument design must be based on a maximum diameter of say 20 cm to allow for
radial ice growth in the hole during instrument installation/recovery. The ice cover
eliminates the effects of wind mixing on surface waters so that extreme density gradients
caused by the annual melt may persist throughout the year. Currents at the ice/water
interface are of particular interest in studies of the movement of pollutants such as oil
and the relation of the instrument sensor to the density gradient is most important; some
mechanism for keeping the sensor-interface distance constant during ice growth must be
devised.
All commercial current meters depend for directional reference on the horizontal component
of the earth's magnetic field. When this becomes very small, the balance of the
compass card in the instrument becomes critical as any departure from horizontal will
give a turning moment on the compass needle in excess of that available from the true
horizontal field. A practical lower limit of useful horizontal field for the best
current meters is 6000y (Barfoot, 1972) which excludes most of the Canadian Arctic
Archipelago. Other techniques must be used for directional reference.
Frequently recording instrument packages for Arctic use must operate unattended for a
year so that they may be installed and recovered when the ice cover is particularly solid
or entirely absent. In addition logistic constraints often require an annual interval
between installation and piCk-Up. Many recorders offered as standard equipment with
commerical instruments do not have adequate capacity for this length of time and batteries
used to power the apparatus will run down. Where a magnetic field performs some function
in the measuring device, steel cased alkaline batteries are not suitable although their
high ampere/hr capacity and good low temperature performance makes them very attractive.
Each and every problem of this type has to be considered separately according to
individual circumstance.
The normal oceanographic instrument deployment system, a ship, is more or less taken for
granted and frequently supplied complete with crew to the oceanographer. The arctic
oceanographer is forced to develop his own "ship" and a description of the system used by
the author's group based on tracked vehicles as a prime mover was given at the time of
the First POAC <strong>Conference</strong> (Lewis, 1972). Some of the lightweight equipment developed for
use with helicopter transport has been described by Ganton (1968).
position and Depth
position information may be required for the location of an oceanographic station or, at
a much higher order of accuracy, to determine ice movement with respect to the sea bed.
In many coastal regions with mountains, a "fisherman" type radar system operated from a
72
Lewis 2

tracked vehicle is accurate enough to give station location and is the cheapest system
apart from a sextant. The radar also gives details of the surface roughness of the ice
sheet which is a great aid in route planning. Where the sea coast is low lying, and if
the range required is not too great, pyramids of empty fuel drums may be used to fix two
on-shore points on the radar screen. For accurate position fixing active radar beacons
systems are available which may be placed on hillsides or steel towers to give ranges up
to 75 krn. Such equipment is far more costly than a simple radar system and gives
accuracies of a few meters in range-range mode, transponders operational at -40 0 Care
available. Further from shore satellite positioning systems are probably the best available
in terms of cost, unless an existing low frequency navigation system may be utilized.
To record movement of "landfast" ice in shallow waters the lengths of wires used to
connect a point on the ice to fixed points on the sea bed have been measured (Croasdale
1974). The wires wound and unwound onto spring loaded reels and an analogue of the
motion was provided by potentiometers driven by the reel shafts. Limited ice movements
may also be determined from the time intervals between reception of signals from a single
sea bed pinger at hydrophones in an array below the ice (R. Goodman, Innovative Ventures,
Calgary, private communication). In deeper waters the ice position may be related to the
known positions of three acoustic transponders on the sea bed (Smith 1973) •
For many years hydrographic soundings have been made through the sea ice sheet using echo
sounders. Snow is cleared from the ice, a reasonably flat ice surface is produced by
scraping or cutting the ice with a special tool and the echo sounder transducer "stuck"
to the ice using motor oil or antifreeze/water mixture. Air within the ice will prevent
adequate reception of the reflected signal on occasion but statistically speaking this
is a minor problem. Recently this procedure has been automated so that the functions
may be controlled entirely from within a tracked vehicle, the transducer having been
placed within fluid filled rubber bag which is hydraulically loaded onto the cleared ice
surface (D. Caulfield, Bannister Technical Service, Edmonton, private communication).
Continuous bottom profiling from an ice sheet has been successfully attempted utilizing
a sparker, enclosed within a flexible bag of
antifreeze/water which was pulled behind a
tracked vehicle and loaded onto the partially
cleared ice. The receiving geophones were
pressed into contact with the ice and the
vehicle had to stop momentarily to receive the
signal as the noise produced by motion over the
ice surface was sufficient to mask the reflection
from the sea bed. With sophisticated
signal processing sub-bottom profiles as well as
depths of channels in the Canadian Arctic
Archipelago were obtained.
Determination of ice growth rates presents a
problem as, at a single site on the ice sheet,
successive drillings each make that location an
anomaly as far as the next drilling is concerned.
Drilling successive holes at nearby locations is
no better, it is quite easy to produce negative
growth rates using this technique due to local
variations in ice thickness. One solution to
this problem, the ice thickness wand, is shown
in Figure 1. Application of electric power for
a few seconds is sufficient to release the wand
from the ice when it may be pulled up against
the scale until the cross bar touches the ice/
water interface. The wand is then allowed to
73
Fig. 1 Wand for measuring ice thickness
(schematic). The diam.
of the wand is about 3 mm and
the power consumption about 60Ow.
Lewis 3

fall and remains in this lowered position until the next reading is taken. The breakable
link allows wand recovery at the end of a growth season.
Two pulsed radar systems of similar design have been developed for the continuous profiling
of sea ice thickness from tracked vehicles (OUtcalt 1974, Campbell and Orange 1974).
Electromagnetic energy at frequencies centred around 300 MHz is projected downwards into
the ice sheet from an antennae having dimensions in the order of a meter which is
suspended in front of the vehicle or pulled along behind it. Reflections from the ice
water interface and intermediate layers are received. These systems have been used
primarily to enhance the safety of over-ice operations.
Salinity/Temperature/Depth Measurements
The use of standard oceanographic bottle techniques in the Arctic winter and spring is
difficult unless the bottles may be brought up through the sea ice straight into a warm
environment. Until a given technique has been proven in practice, at least two samples
of sea water should be extracted from each oceanographic bottle to ensure that there has
been no freezing within the bottle during recovery. Even with a heated enclosure ice
crystal formation may occur in samples taken at- depth close to ice shelves as the bottle
is recovered, due to the heat sink provided by the variation in freezing point with depth
(Foldvik and Kvinge, 1974). Alternatively ice crystals may exist in the sea water immediately
below a growing sea ice sheet and in the vicinity of ice shelves, glacier tongues,
etc. which will melt within the sample giving erroneous salinity values. Once again
analysis of two samples taken from the same bottle at different times will help to clarify
matters.
These problems associated with ice crystals in the water still occur with in-situ
conductivity/temperature/depth equipment but may be recognized with experience. Sudden
changes in calibration, particularly of conductivity cells are often associated with ice
formation on the cell caused either by deposition of existing ice crystals or by growth
onto a cell at below freezing point temperature placed in the sea. As far as is known
only one CTD instrument is specifically adapted for Arctic use and a detailed study on
its deployment and calibration was made by Lewis and Sudar (1972). Others cequire a much
larger hole in the ice sheet for deployment. an instance of their use being that described
by Amos (1975). Most of the problems relating to accuracy and preciSion of CTD measurements
are not peculiar to the Arctic but one, the conversion of conductivity, temperature,
and pressure readings to salinity is particularly uncertain near freezing point. An
interim formula has been published by Perkin and Walker (1972), but further work is still
required.
It should be realized that effective use of CTD equipment from the sea ice demands a carefully
designed deployment system. To make measurements at levels of accuracy commensurate
with 0.01% 0 salinity on a routine basis requires a meticulous attention to detail and
it is unlikely that it can be approached without an integrated engineering design. This
is also true onboard ship but the comparative states of development of over-ice and overwater
deployment systems make the point particularly valid in an Arctic environment.
Tidal Measurements
Arctic tidal measurements have been made for many years using equipment designed for lower
latitudes. For example electrically heated stilling wells have been used to prevent ice
formation and to allow free movement of a float on the surface of the water in the well.
Such equipment requires protection from ice movement, a heating supply, and an attendant
so that its use is confined to the vicinity of permanent settlements. It is also expensive
to operate. Another technique is to measure and log the pressure at the sea bed
using an on-shore transducer and strip chart recorder. Available equipment will not
usually operate in an extremely cold environment and thus requires heated shelter. In
74
Lewis 4

For recording currents near the sea bed the equipment shown in Figure 4 has been developed
with telemetry to shore as a back-up to internal recording. A folding tripod is lowered
through a 10" diameter hole drilled in the ice sheet (Section A of fiqure) and the vehicle
is brought forward to allow attachment to a winch in the sledge (Section B of the figure).
The pressure case containing the gyro also contains two inclinometers. Together with a
touch-down switch these enable the user to be certain that the frame has deployed correctly
on the sea bed. The gyro is now unlocked remotely from the surface and allowed to
run free; a messenger releases the pressure case from the tripod frame. The case is
winched-in causing a thin line attached to the tripod to unwind from the spool shown.
The gyro in its case comes through the ice to the surface when its orientation is immediately
checked with respect to local land marks which give the frame orientation. OUt of
sight of land, the direction may be referred to that available from any existing navigation
system including a separate north seeking gyro. The line is cut, a float is pushed
down on it to a depth of about 7 meters where it locks. An acoustic beacon is tied to
the line end and thrown down the hole after the float to produce the situation shown in
the lower part of Figure 4C. The current meter hangs from the tripod apex by a short
length of hydraulic hose and the normal north seeking compass in the current meter follows
a magnet inserted in the vane used to give current direction with respect to the tripod.
Currents near the surface are measured by a sensor assembly consisting of rotor and vane
which are deployed separately to the recorder unit in order to enhance chances of survival
of the latter during break-up. The recorder with the telemetering transmitter and a small
radio beacon are contained within a heavy walled float made from 6" black polyethylene
pipe; the telemetering antenna is separate. The surface sensor package is moored on a
device which mechanically senses the position of the ice/water interface at daily intervals
and if necessary, moves the sensors to keep them at a constant distance from the
interface. The equipment is installed using a transponder type range-range navigation
system giving position accuracies of a few meters in tens of kilometers. This enables
the operator to return to the same location, say one year later, drill a hole in the ice
and with reasonable assurance listen for the acoustic beacon and determine its bearing.
He then drills further holes until the float can be seen either with a tv system or with
a periscope and hooks the line with a specially designed grab. The sea bed assembly
weighS about 40 lbs in water so that recovery does not present much difficulty. The
upper current sensor assembly mayor may not survive through break-up though the recorder
package will almost certainly be recoverable if either the acoustic or radio beacons can
be "heard". Telemetry ashore restricts the loss if the equipment is destroyed and enables
malfunctioning to be detected immediately. The system was first deployed in April 1974
with the exception of the sonic telemetry link between sea bed and float; this has still
to be tested.
Future Developments
The float described in the last section is a simple instrumented buoy suitable for use
in coastal waters. Much larger buoys including satellite navigation equipment has been
built by the staff of the AIDJEX Project and deployed in the Northern Beaufort Sea. They
have survived for over two years and the design has clearly proved itself satisfactorily.
A detailed description of the equipment is available (Haugen and Kerut 1973). The major
region for future development probably lies in sensors. The extreme stratification
existing in Arctic surface waters would make a conductivity chain to be allied with a
temperature chain most desirable. Although strictly outside oceanographic considerations,
wind speed sensors that are not susceptible to icing are highly desired. Reception of
data from remote areas in near real time is a possibility with modern satellites. Within
three years it may be possible to instrument the waters in the vicinity of a proposed
industrial development on the Arctic coast and to receive data in the south sufficient
to check equipment operation and to make decisions outside without having the usual long
term wait for an adequate information base.
78 Lewis 8

REFERENCES
Amos, A. F. 1975 Physical oceanography from the Arctic pack ice; project AIDJEX S/T/D
Programs. Proc. <strong>Third</strong> S/T/D Conf. & Workshop, Feb. 12-14, 1975, pp 125-142.
Plessey Environmental Systems, San Diego.
Barfoot, L. 1972 Current meter direction detection as a function of magnetic field
intensity. Canada Centre for Inland Waters, MSD Central Region, Burlington, Onto
Campbell K. J. and A. S. Orange 1974
ice thickness by impulse radar.
A continuous profile of sea ice and freshwater
Polar Record, Vol. 14, No. 106, pp 31-41.
Croasdale, K. R. 1974 The movement of Arctic landfast ice:
influence on offshore drilling. Proc. 2nd Int. Conf. on
Arctic Conditions, pp 617, Univ. of Iceland. Reykjavik.
its measurement and
Port & Ocean Eng. under
Foldvik, A. and T. Kvinge 1974 Conditional instability of sea water at the freezing
point. Deep Sea Research, Vol. 21, pp 169-174.
Ganton, J. H. 1968 Arctic field equipment. Arctic, Vol. 21, pp 92-97.
Haugen D. P. and E. G. Kerut 1973 The Arctic data buoy, a system for environmental
monitoring in the Arctic. AIDJEX Bulletin #22, pp 37-53. Arctic Ice Dynamics
Joint Experiment, Univ. of Washington, Seattle, Wash. Aug. 1973.
Lewis, E. L. 1972 The collection of oceanographic data from the sea ice surface in
winter. Proc. First Int. Conf. on Port & Ocean Eng. under Arctic Conditions, pp1219
Technical Univ. of Norway, Trondheim.
Lewis, E. L. 1973 A tide gauge for use in remote areas. In "Ocean 73", pp 504. Inst.
of Electrical & Electronic Engineers, New York.
Lewis, E. L. and R. B. Sudar 1972
sea for salinity determination.
Measurement of conductivity and temperature in the
J. of Geophy. Res., Vol. 77, No. 33, pp 6611-6617.
Outcalt, E. L. 1974 Radar peers through the Arctic ice. Canadian Petroleum, p 46.
Perkin R. G. and E. R. Walker 1972 Salinity calculations from in situ measurements. J.
Geophys. Res., Vol. 77, No. 33, pp 6618-6621.
Rapatz W. J. and F. Stephenson 1974 The effectiveness of the Aanderaa Water Level Gauge
in Canada's Arctic offshore tidal program. In "Ocean 74", pp 358, Inst. of Electrical
and Electronic Engineers, New York.
Smith R. D. 1973 Precision acoustic positioning with the AIDJEX acoustic bottom referencing
system. In "Ocean 73", IEEE Conf. on Engineering in the Ocean Environment,
IEEE, New York, pp 526-530.
80 Lewis 10

(the resolution of the satellite) to be detected. Besides, they did not necessarily have
to be open, in fact we found that most of them had a cover of thin, dark new ice.
OBSERVED FEATURES AND CAUSES
Some specific features can be distinguished on the analyzed map which represents a mean
of all observations for March 1973 (Fig. 2). The highest ice concentration, above 7 octas,
generally appears north of 75°N and SE of Wrangel Island in the western part of the
Chukchi Sea. In March 1973 the border of the 7 octa line coincided very closely with the
continental shelf as outlined by the 100 fathom contour in the East Siberian and Chukchi
Seas. Areas of low ice concentration, below 6 octas, occur NW of Wrangel Island, north
of the New Siberian Islands and W-SW of Banks Island. Concentrations below 5 octas are
found along the NW coast of Alaska. Northern Bering Sea is generally below 5 octas except
for the Gulf of Anadyr and a small area NE of St. Lawrence Island. The lowest ice
density, below 2 octas, was found in the eastern part of Norton Sound.
By studying the Canadian surface weather maps for the Arctic Ocean, the following general
weather situation could be derived for March 1973. This general pattern applied mainly
to the Chukchi and Beaufort Seas and was less reliable for the Siberian side where the
data were sparse. At the beginning of March the prevailing winds were generally northerly
to westerly with a small to medium intense pressure gradient. From the 16th of March a
strong pressure gradient prevailed and the wind direction started to shift towards NE.
On the 18th of March the wind shift to NE was fully completed. From then until the
end of the month, the winds were generally NE to E due to a strong high pressure system
centered over the northern Beaufort Sea. This general situation is also shown by Streten's
(1974) 5-day means of the 0000 GMT surface pressure differences between Barter Island
(70° 08'N; l43°38'W) and Point Barrow (7l 0 l8'N; l56°47'W) resulting in an on-shore
or northerly wind component during the spring and summer of 1973. His pressure differences
between Mould Bay (76°l4'N; l19°20'W) and Cape Parry (70 0 l0'N; l24°4l'W) give
a resulting easterly component of wind flow after the middle of March. Since the wind
is normally the main driving force of the ice, an easterly component would tend to open
the pack west of Banks Island. A northerly wind component would press the pack ice
against the northern shore of Alaska.
In March 1974 the ice concentration was an average of 10% higher than in 1973 (Fig. 3).
The reason can be found in the difference in wind pattern. Computing 5-day means of the
0000 GMT surface pressure difference between the same stations as in 1973, we find the
resulting wind components to be westerly for the entire month of March 1974. (Mould
Bay and Cape Parry, Fig. 4). In connection with a northerly wind component for the second
half of March (mean pressure difference between Barter Island and Point Barrow), the
pack ice would tend to get pressed against the Arctic coast of Alaska and Banks Island.
This results in a more southerly location of the 7 octa line for the ice concentration
in 1974 (Fig. 3). The 7 octa line reaches the coast of Siberia by the New Siberian Islands
and SE of Wrangel Island, eliminating the low ice concentration north of the New
Siberian Islands in 1973. The same holds true W-SW of Banks Island. The lowest concentration
in the eastern Norton Sound is slightly below 4 octas as compared to 2 octas in
1973. Areas of shore fast ice, which were not detected on the March 1973 imagery, occur
along the south coast of Norton Sound and in Kotzebue Sound. The shorefast ice, bordering
the low density ice along the NW coast of Alaska, was detected in both years. The
low ice concentration along the NW coast of Alaska and NW of Wrangel Island and the high
concentration SE of Wrangel Island seem to be rather typical. The pattern is probably
connected with the general circulation bringing warmer Bering Sea water north along the
coast of Alaska and cold Arctic Ocean water south along the coast of Siberia.
The ice concentration maps were planimetered and the mean amount of open water, or more
probably new ice, calculated. The observed difference in ice concentration is reflected
87 Ahlnas and Wendler 5

00
00
Figure 2. Ice concentration in octas for March 1973. Gradually darkening shading used for every whole octa
interval. Highest concentration 7-8 octas: large widely-spaced dots, lowest concentration, below 2 octas in
upper Norton Sound: darkest shading. Note coincidence between margin of continental shelf, as outlined by 100
fathom isobath and 7 octa line in East Siberian Sea.

Figure 3. Ice concentration in octas for March 1974. (explanation: see Figure 2)

Figure 5. Satellite imagery #6426 by NOAA-2 VHRR is the visible band on 12 March 1974
showing maximum advance of ice edge in Bering Sea . Note areas of low ice concentration
along the NW coast of Alaska, NW of Wrangel Island and St. Lawrence Island .
92 Ahlnas and Wendler 10

Figure 6. Enlargement of NOAA- 2 VHRR imagery #6463 VIS on 15 March 1974. Note decrease
in ice concentration from previous figure. Open water south of St. Lawrence Island and
in upper No r ton Sound .
93 Ahlnas and Wendler 11

partly be due to the large quantities of fresh water discharged by the rivers draining
into these oceans. The low density river water with its higher freezing point would stay
on top of the saline ocean water, creating a stable layer that would freeze early.
Another, if not main cause, would have to be sought in the ice dynamics itself. The total
amount of "open" water could be calculated in the same way as the ice concentration. For
each sub-section and each satellite pass the amount of "open" water was estimated and
the values averaged. For the Arctic Ocean north of the Bering Strait, where the subsections
were about the same size, a calculated amount of "open" water of 16% in March
1973 and 9% in March 1974 was found which is in good agreement with the planimetered
values from the analyzed maps which gave 15% and 11% respectively. The total values of
all sections given somewhat too high a value for the amount of "open" water because the
sub- divisions were smaller in the low ice density areas of Northern Bering Sea (23% in
1973 and 15% in 1974) than in the high ice density areas of the central Arctic Ocean.
The values found are in reasonable agreement with Wittman and Schule (1966) and Thorndike
et aL (1976).
When estimating the ice concentration f or each subsection the main direction of the ice
fractures, if any, was also recorded daily. The directions were estimated with an accuracy
of 22, degrees using 16 cardinal points or 8 main lead directions, since 180 degree
opposites would give the same fracture direction. In Figure 7 the mean or preferred lead
direction is shown for March 1973. The mean is calculated from the frequency of leads
running in each mentioned direction. The direction did at times change from day to day
as new leads opened and old ones disappeared. Generally, the leads run N-S in the central
Arctic Ocean and roughly parallel to the coast further south. The relationship between
the fracture pattern in the Beaufort Sea as seen on NOAA-2 IR and the atmospheric
pressure field for March 1973 was also studied by Ackley and Hibler (1974). The preferred
lead directions are roughly the same for March 1974 as for March 1973 in the central
Arctic Ocean (Fig. 1). In the far north and along the coast some 45° to 90° changes in
direction have occurred. The main directions are caused by shear, created as a combined
result of the prevailing mean wind direction and the underlying force of the ocean currents.
As a convenient working model, the general surface water circulation in the Arctic Ocean
was taken from the Oceanographic Atlas of the Polar Seas (1958). The large anticyclonic
gyre in the Beaufort Sea is most striking (Fig. 8) but may not always exist, as during
1974. The ice fracture maps showed the leads to be roughly parallel with the coastlines.
This is dee to air stress and the alongshore circulation shearing the ice parallel to the
shore. Away from the shore leads are opened and closed by synoptic pressure systems independent
of shoreline orientation.
Added to the surface circulation pattern are ice fractures from NOAA-3 VHRR VIS and IR for
13 April 1974. The orientation of the leads shows their relationship with the circulation.
Note the absence of fractures in the lower center of the gyre. Dynamically a current
gyre is comparable to an atmospheric pressure cell where the winds reach their minimum velocity
in the center. This explains the lower concentration of fractures in the center,
since the calmest conditions with the least stress exist there. In addition, the satellite
imagery, (Fig. 9) shows a low pressure cell centered over the gyre.
TEMPERATURE OF ARCTIC OCEAN LEADS
To get an idea of the actual temperature of the leads and to try to answer the question
whether they were open or refrozen, some IR enhancements were done for selected images
of March 1975. At that time higher accuracy could be obtained by reprocessing the digit-
94 Ahlnas & Wendler 12

Figure 7. Preferred lead direction for each subsection in March 1973.

Figure 8. General Surface water circulation in the Arctic Ocean (from Oceanographic Atlas of the Polar Seas 1958).
Note large anticyclonic gyre in t he Beaufort Sea. Thin lines are ice fractures from NOAA-3 VHRR for April 13, 1974.

Figure 9 . NOAA-3 VHRR #1961 IR imagery of the leads connected with the anticyclonic
circulation in the Beaufort Sea on 13 April 1974 .
97
Ahlnas and Hendler 15

-29·C UK.. -18·C lriim -3 -5 -10 -15 -20 -25 -30 -35 -40 -43"C
Figure 10. Enhanced enlargement of the NOAA- 4 VHRR imagery #1530 IR of Eas tern Beaufor t Sea on 17 March 1975 .
The gray scale from white t o black goes from -43°C to - 3°C. Black dots are -29 °C and white dots -18°C .

Weller, G. 1972. Radiation Flux investigation, AIDJEX Bull. No. 14, p. 28-30.
Wendler, G. 1973. Sea ice observation by means of satellite. J. Geophys. Res. 78: 1427-
1448.
Wittman, W.I. and J.J. Schule, Jr. 1966. Comments on the mass budget of Arctic pack ice.
Proceedings of the Symposium on Arctic Head Budget and Atmospheric Circulation -
pp. 215-246, ed. J.O. Fletcher. The RAND Corp. RM-5233-NSF.
104 Ahlnas and Wendler 22

MONITORING ARCTIC SEA ICE USING LANDSAT IMAGERY
James C. Barnes, Clinton J. Bowley, and Michael D. Smallwood
Environmental Research & Technology, Incorporated
Concord, Massachusetts
United States
ABSTRACT
High resolution, multispeatral satellite data-have now been available sinae late July
1972, when Landsat-l (formerly aalled ERTS-l, the Earth Resouraes Teahnology Satellite)
was launahed. Landsat-2 was launahed in January 1975. In this paper the appliaation of
Landsat imagery to monitoring aratia sea iae is disaussed. In partiaular, the results
of studies using data aolleated during the spring and summer of 1973 are presented.
These studies have shown that detailed information on iae aonditions aan be derived from
the Landsat imagery. Beaause of the high aontrast, linear iae features even smaller
than the aatual sensor resolution may be deteatable. Moreover, differenaes in refleatanae
between the shorter wavelength bands and the near-infrared band have been found
very useful for distinguishing iae types and for determining aertain aharaateristias of
the iae surfaae. The analysis teahniques that have been developed using Landsat data
represent powerful tools in fulfilling the sea iae statistiaal data requirements of both
short-term logistiaal planning and long-term program efforts for future off-shore operations.
105

Figure 1 . Landsat- l MSS-S imagery (ID Nos . 1226-22171 and 1226- 22174), 6 March 1973,
showing St . Lawrence Island area in the Bering Sea. Grey ice with numerous
shearing leads extends southward from the island; an area of open water with
developing stratus cloud streaks lies along the southern coast; first-year
and compacted floes cover the area east and north of the island .
108 Barnes et al . 4

As the flight approaches the western end of St. Lawrence Island the commentary indicates
that the amount of thick first-year ice is decreasing and that the ice ahead is mostly
grey ice; the location of the ice boundary separating these different ice types in the
Landsat imagery corresponds almost exactly with the location when the above comment was
made. After crossing the island, the observer reports all grey ice a few days old, some
of which has undergone deformation in the form of stretching to the southwest. This observation
confirms the ice type deduced from the satellite data; furthermore, leads
indicative of stretching of the ice can be seen in the imagery.
The ice conditions displayed in the imagery on 7 March (not shmm) are very similar to
those of the previous day . On the 7th, however, the stratus streaks have increased, obscuring
some of the grey ice south of St . Lawrence Island. Farther south some ice features
can be detected, but much cloudiness exists . One segment of the CV-990 flight on
this day follows the southern boundary of the grey ice, an area that is cloud free . The
observer reports a vast expanse of grey ice, and then reports running into a stratus
deck; the comment is made that the "stratus streaks are due to open water along the south
shore of St. Lawrence". Thus, this commentary verifies exactly the ice and cloud conditions
apparent in the satellite imagery.
Comparative analysis with aircraft photography
Vertical viewing photography from the CV-990 flight of 5 March was obtained from NASAl
Goddard Space Flight Center. The field of view of the camera system was ± 35°, providing
an areal coverage for an individual frame of about 12.5 by 12.5 km at the altitude flown .
Through a measurement of the scales of features detectable in the photographs as compared
with those detectable in the corresponding Landsat images, the spatial resolution of the
aircraft photographs was determined to be of the order of 5-10 m.
Ice conditions photographed on two segments of the flight of 5 March, are in excellent
agreement with the ice conditions mapped from the Landsat imagery a day later. An enlarged
portion of the satellite image north of St . Lawrence Island is given in Figure 2a,
and the corresponding aircraft photography for this flight segment is given in Figure 2b.
The photography shows numerous first-year ice floes (vast, big, medium and small)
embedded in grey and grey-white ice, as well as several fractures . Twenty-four hours
later these same vast, big and medium first- year floes embedded in grey and grey- white
ice can be identified in the Landsat enlargement. The small floes, however, are-not as
readily detected due to the overall brightness variations existing within the grey and
grey- white ice. The fractures (approximately 70 to 90 m wide) seen in the photography
are not visible in the satellite imagery, indicating that they may have either closed or
refrozen during the 24-hour interval. A fracture which did not appear in a vast floe in
the aerial photography is visible, however, in the satellite image; the fact that this
fracture is of the scale of those seen in the aircraft photography and is in a floe that
can be positively identified indicates that changes have actually taken place in the ice
(some fractures closing and others opening) during the 24-hour interval .
Another portion of the Landsat image is given in Figure 3a and the corresponding swath of
aircraft photographs in Figure 3b . This segment of aerial photographs shows young fast
ice (grey and grey- white) 200 m to 1.4 km in width along the west coast of St. Lawrence
ISland . Pack ice off the southwest coast comprised of young (grey) ice is also visible .
These same features are readily detectable in the Landsat image.
In summary, the sea ice features that can be detected in Landsat imagery, as seen in
these observations of the Bering Sea and in observations of other arctic areas, include
the following: the distinction between grey, grey-white, and older forms of ice, as well
as the distinction between ice floes and surrounding brash ice; the growth and deterioration
of leads; the formation of new grey ice within leads; the deterioration of the ice
surface as evidenced by the formation of puddled areas and flooded ice; linear dry areas
109 Barnes et al. ')

Figure 4 . Landsat-1 MSS-7 image (ID No. 1259-20130), 8 April
1973. showing developing lead in eastern Beaufort
Sea at M' C1ure Strait, south of Prince Patrick Island.
Figure 5. Landsat- 1 MSS- 7 image (ID No . 1261-20243), 10 April
1973, showing same area as in Figure 4.
113 Barnes et al. 9

Figure 6. Landsat-l MSS- 7 image (10 No. 1262-20302) , 11 April
1973, showing same area as in previous two figures.
Figure 7. Landsat-1 MSS-7 image (10 No. 1279-20242), 28 April
1973, showing same area as in previous three figures.
114 Barnes et al . 10

;
f
Figure 8. Landsat-1 MSS-7 imagery
ID Nos. 1283-20470, 1283-20472),
2 May 1973, showing same area as
in previous four figures .
Figure 9. Landsat-1 MSS-7 imagery
(ID Nos. 1298-20300, 1289-20302),
17 May 1973, showing same area
as in previous five figures .

Figure 10. Landsat-1 MSS-7 imagery (10 Nos.
1301-20465, 1301-20471), 20 May
1973, showing same area as in
previous six figures.
Figure 11. Landsat-1 MSS-7 imagery (10 Nos .
1318-20405, 1318-20411, and 1318-
20414), 6 June 1973, showing same
area as in previous seven figures.

Figure 13. Landsat-1 MSS- 7 mosiac, (ID Nos . 1245-13430, 1245-
13423 and 1245-13421) 25 March 1973, showing the
Dove Bay area along the east coast of Greenland .
The following features are indicated: lIe de France
(A), Germania Land (B), Dove Bay (C), Store Ko1deway
(D), Hochsetters Forland (E), and embedded icebergs (F) .
119
Barnes et al. 15

Figure 14 . Landsat- l MSS- 7 mosiac (ID Nos. 1337- 13532 and
1337-13530) 23 June 1973, showing the same
area observed in the previous figure. Flooded
fast ice (A) is evident in Dove Bay and southwest
of lIe de France.
120
Barnes et al. 16

Figure 15. Landsat-l MSS-7 mosiac (ID Nos. 1391-13523
and 1391-13521) 18 August 1973, showing the
same area as observed in Figures 13 and 14.
lIe de France (A), Dove Bay (B) and Store
Koldeway (C) are idicated.
121 Barnes et al. 17

Figure 18. Landsat-2 MSS-7 image (ID No. 2201-21445), 11 August 1975, showing
same area as viewed in the previous Figure . The ice edge can be
seen through a thin cloud cover .
125 Barnes et al. 21

areas east of Barrow appears darker, probably because of the existence of meltwater on
the ice surface; on 11 August some ice may also exist in these same areas, but it cannot
be detected in the near-IR image. Despite the existence of some c loud cover over the ice
in the August observation, the edge of the ice is sharply defined, and it is obvious that
the pack has not retreated during the interval from early July. The ice cover persisted,
of course, into September, causing the well-publicized difficulties in the passage of the
supply barges around Point Barrow to Prudhoe Bay. In contrast, a Landsat observation of
early September 1973 shows the same area with completely open water.
CONCLUSIONS
Continuing studies have further substantiated the application of Landsat for monitoring
sea ice. The Landsat configuration is sufficient for collecting substantial amounts of
sea ice data. Because of the overlapping of orbital passes at high latitudes, the potential
for obtaining sequential observations of the Arctic is considerably greater than for
lower latitudes. The sensor resolution and area viewed by Landsat are adequate for most
ice monitoring purposes, and ice monitoring is an application where the color data products
appear to offer little advantage over the more readily available black and white
imagery.
Despite the adequacy of Landsat for most ice monitoring purposes, certain changes in the
orbital configuration and sensor complement would enhance the potential for collecting
arctic data. The orbital coverage should be extended farther north than the current 79°
to BOoN limit, in order to permit data collection over additional arctic regions; for
example, the northern coast of Ellesmere Island, which cannot be observed in the current
Landsat configuration, is an important region to be monitored in future ice studies
because it is the origin of most of the arctic ice islands. Another limitation in the
use of Landsat data is that the existing sensors can provide ice observations only during
the period of adequate solar illumination. For heat balance studies, in particular, wintertime
ice monitoring is essential. The addition of a thermal-IR channel to the sensor
complement would enable ice observations to be made during the wintertime dark period.
A sufficient amount of Landsat data have now been accumulated to enable extensive studies
to be undertaken of ice deformations and movements and their year-to- year variations. It
is now possible to initiate the compilation of sea ice statistics related to parameters
such as floe size and the spatial and temporal frequencies of leads and polynyas. Also,
detailed studies of ice types and surface characteristics using Landsat digitized data
have yet to be accomplished. Clearly, the continued development of improved techniques
to interpret Landsat and other satellite data will lead eventually to the use of operational
satellite systems to collect ice data on a far more complete and economical basis
than has heretofore been possible.
ACKNOWLEDGMENTS
The studies reported in this paper were supported by NASA Goddard Space Flight Center.
The authors wish to acknowledge the assistance provided by Mr. James R. Greaves, Scientific
Monitor, and Mr. Edmund F. Szajna, Technical Monitor both of NASA Goddard. We also
wish to thank Commander William Dehn, retired Director of the Ice Forecast Office of the
U.S. Navy Fleet Weather Central, and Mr. William E. Markham, Director of the Canadian Ice
Forecast Central, for providing aerial ice survey charts and for their helpful suggestions
in the interpretation of various ice features detectable in Landsat imagery. The
BESEX ice data were obtained through the courtesy of Dr. Per Gloersen of NASA Goddard.
126
Barnes et al . 22

REFERENCES
Barnes, J.C. and C.J. Bowley. 1974. The application of ERTS imagery to monitoring
arctic sea ice. Final Report under Contract NAS 5-21802, Environmental Research &
Technology, Inc., Concord, MA, 93 pp.
Barnes, J.C., C.J. Bowley, D.T. Chang and J.H. Willand. 1974. Application of satellite
and visible infrared data to mapping sea ice. In Proceedings of Interdisciplinary
Symposium on Advanced Concepts and Techniques in the Study of Snow and Ice Resources,
National Academy of Sciences, Washington, D.C., pp. 467-476.
Barnes, J.C., C.J. Bowley and M.D. Smallwood. 1975. The application of ERTS imagery to
monitoring arctic sea ice: supplemental report. Report under Contract NAS 5-21802,
Environmental Research & Technology, Inc., Concord, MA, 47 pp.
Barnes, J.C., D.T. Chang and J.H. Willand. 1972. Image enhancement techniques for improving
sea ice depiction in satellite infrared data. Journal of Geophysical Research;
Oceans and Atmosphere Edition, 77 (3), pp. 453-462.
Baliles, M.S. and H. Neiss. 1963. <strong>Conference</strong> on satellite ice studies. Meteorological
Satellite Laboratory Report No. 20, U.S. Department of Commerce, Weather Bureau.
Campbell, W.J., et al. 1974. Results of the U.S. contribution to the joint U.S./U.S.S.R.
Bering Sea experiment. Preprint X-9l0-74-l4l, NASA Goddard Space Flight Center,
Greenbelt, MD, 193 pp.
Crowder, W.K., H.L. McKim, S.F. Ackley, W.O. Hibler, III and D.M. Anderson. 1973. Mesoscale
deformation of sea ice from satellite imagery. Paper presented at the Interdisciplinary
Symposium on Advanced Concepts and Techniques in the Study of Snow and
Ice Resources, Monterey, California.
Fletcher, J.O. 1966. Forward. In Proceedings of the Symposium on the Arctic Heat Budget
and Atmospheric Circulation, Memorandum RM-5233-NSF, The RAND Corporation.
McClain, E.P. 1973. Some new satellite measurements and their application to sea ice
analysis in the Arctic and Antarctic. In Proceedings of Interdisciplinary Symposium
on Advanced Concepts and Techniques in the Study of Snow and Ice Resources, National
Academy of Sciences, Washington, D.C., pp. 457-466.
McClain, E.P. and M.D. Baliles. 1971. Sea ice surveillance from earth satellites.
Mariners Weather Log, 15 (1), pp. 1-4.
Morse, R.M. 1964. Recommendations of panel on sea ice. Oceanography from Space, Woods
Hole Oceanographic Institution.
Wark, D. and R. Popham. 1962. Ice photography from the meteorological satellites TIROS
I and TIROS II. USWB, Meteorological Satellite Laboratory Report No.8.
127 Barnes et al. 23

Q)
c
(J If)
CI) g-
B
MAXIMUM RANGE
TEMPERATURE
VOLTAGE SIGNAL
Figure 2. The look up graph showing the gray tone for maximum
range of temperature
133 Jayaweera and Marvill 5

8
W
..J
c:::(
U
(f)
w
ENHANCED IMAGING
- CC °c
TEMPERATURE
Figure 3 (a). Look up graph for IR enhancement
134 Jayaweera and Marvill f

Figure 5 . Application of double gray scale to identify the ice
edge . The gray scale is chosen so that there is a
gray tone jump at the ice edge.
137 Jayaweera and Marvill 9

This way, open water areas can be easily distinguished. This technique, in addition to
showing at a glance open water, is also useful to compute the fraction of open water in
the Arctic at any time of the year.
The look up curve for this type of enhancement in Figure 6 indicates two gray tones for
each temperature. The sudden jump of gray scale from black to white is chosen for the
temperature at which a change in a physical property will occur, in the previous example
this corresponds to the melting point of sea ice. Then while one scale will give temperatures
of ice the other gives those of the sea water. In this way the information content
could be doubled from that of a single scale enhancement in addition to showing the boundary
of sea ice and water.
Ice Thickness Measurement
Up to a certain thickness of sea ice the radiative temperature from sea ice will be directly
related to its thickness provided there is no snow on the surface. The IR enhancements
could be used to map ice thickness by assigning a color, say black to the temperature
corresponding to a certain thickness; or if all areas where ice is thicker than a
certain amount or between two thicknesses is required, then a special single gray tone
could be assigned to that temperature range. This kind of gray tone could be superimposed
on the original image in the form of black dots or black areas.
CONCLUSION
These are a few of the examples of using IR enhancements for sea ice studies. These
special enhancements could be used to distinguish sea ice conditions in a near real time
basis or later research applications.
ACKNOWLEDGEMENT
This work is a result of research sponsored (in part) by the National Oceanic and Atmospheric
Administration, National Environmental Satellite Service Grant No. 5-35190 and
(in part) by the Alaska Sea Grant Prop,ram, supported by NOAA Office of Sea Grant, Department
of Commerce, under Grant No. 04-5-158-35.
139 Jayaweera and Marvi11 11

Introduction
ARCTIC OFFSHORE DESIGN CRITERIA FROM UNDER-ICE PROFILE DATA
In the early 1960's, several nuclear submarines of the U.S. Navy made cruises beneath
the Lce of the Arctic Ocean. Mounted on each of these submarines were several sonar
transducers to provide guidance for safe navigation (Fig. 1). As the submarine advanced,
sound pulses from the upward-looking transducer were reflected from the ice COver
above and recorded on strip-charts aboard the submarine. These strip charts reproduce
in analog form the profile of the underside of the Lce under which the submarine is
cruising (Fig. 2). The profiles are amenable to rapid analysis by computer (LeSchack,
et al., 1970) and are presently being analyzed by the Development and Resources Transportation
Company (D&RT Co.) to derive under-ice parameters for use by engineers concerned
with designing Arctic offshore structures to withstand the onslaught of the
moving pack ice. These data are providing, at relatively low cost, part of the ice design
information for which the oil industry has expressed a need (Hudson, 1973; Visser,
1973).
Analysis Of Submarine Under-Ice Profiles
An example of the raw submarine under-ice profile data is shown in Figure 2. Depth
measurements are made directly from these analog charts with a line-follower digitizer.
The base from which all depth measurements are made is the water surface; this is determined
by the characteristic reflection observed on the record when leads or polynyas
are present (as small ones frequently are). The digitizer records X and Y data values
along the profile by means of manually following the profile, from left to right, with
an instrumented arm at the end of which is a cursor with a magnifying lens. The data
points are digitized starting from an origin positioned at the lefthand side of the
record. The waterline is at the top of the record. Only the first arrivals of recorded
sound energy are digitized. The digitizer records one depth value, (Y), for every 0.01
inch (0.03 cm) of x-travel along the direction of the profile. Depth values are recorded
in inches in increments of 0.1 inch (0.3 cm) On the record. This translates to a
data point about every 15 ft (5 m) in the X-direction (for a typical submarine cruising
speed) and a depth increment of 3.2 ft (1 m) .
Once the analog strip chart data are converted into digital form, statistical analysis
of the profiles are conducted. The results of the analyses are expressed in tabular
form (Tables 1-10). Each profile sub-segment analysis contains: (a) the linear frequency
of occurrence of ice keels deeper than 6 ft (1.8 m), deeper than 15 ft (4.6 m),
deeper than 30 ft (9.1 m), and deeper than 45 ft (13.7 m), (b) the frequency of occurrence
of ice keels at each depth, (c) the percentage of the total ice profile found at
each depth, (d) the frequency of occurrence of ice masses below a depth of 10 ft (3.0 m)
having a given area (an ice mass is defined as the maximum area of an ice profile segment
deeper than 10 ft (3.0 m), i.e., an ice mass may be composed of one or more keels),
(e) the frequency of occurrence of ice masses having a given width at the base (taken
at a depth of 10 ft (3.0 m) as in (d) above, and (f) a tabulation of the width of a
given ice mass (e) vs. the area (d) of the same ice mass.
Geographical and Seasonal Variation of the Ice
It is clear to persons doing any extensive flying over Arctic ice that the ice cover
and the intensity of ridging (and hence, the effective thickness of the ice) vary from
place to place and from season to season. Since the submarine cruises made in the early
1960's covered large areas and spanned several seasons (Fig. 3), an opportunity is afforded
by these data to examine quantitatively the ice thickness variations. From the
point of view of the engineer designing offshore structures, however, analysis of
142 LeSchack 2

FIGURE 3: Crulse tracks of U. S. nuclear submarines in the Arctlc Ocean
(after Lyon. 1963).
146
LeSchack 6

FIGURE 4: Superimposed along the February 1960 cruise track of the USS SARGO are values of RMS ice depth (m).
The general boundaries of the three "ridging intensity" provinces described by Hibler et al. (1974) are indicated.

Examination of imagery and format of presentation
Prior to any numerical analysis, photographic presentations of the IR data were examined.
Using standard photo-interpretive techniques, an area of the Arctic pack in each image
was selected that was clear of cloud cover and reasonably close to an identifiable landmark
(Fig. 5). From the general areas selected on the imagery, it was then possible,
using the digitized satellite IR data, to generate for these areas a computer print-out
map, using different type font characters to represent different temperature ranges
(Fig. 6). This type of presentation, although physically unwieldy, permits locating
and referencing any area of interest on the map by a coordinate system of "lines and
spots". Once a geographic landmark can be identified on this map, the geographic
coordinates of the Arctic pack ice under study can be ascertained by cross referencing
the gridded image that corresponds to it. In addition to the print-out map, another
data presentation in the form of 32 x 32 matrices was used. This format gives greater
precision of radiant emittance values and also provides generalized contours of these
data values (Fig. 7). Once the pack ice areas of specific interest were localized, it
was then possible to conduct numerical analysis of the data associated with these areas.
Numerical analysis
Once data is in the matrix format shown in Figure 7, they are amenable to numerical analysis.
Each number in the 32 x 32 matrix represents an average radiant emittance
value (or because of the radiometer calibration, a radiant temperature value) for the
area of ice covered; in the case of Figure 7, each number represents an average radiant
temperature value over four adjacent 1 km resolution elements, thus the matrix actually
represents a 128 km x 128 km area over the ice. During the course of this research,
however, NOAA developed programs that permitted the formatting of 32 x 32 matrices of
the individual 1 km resolution elements. It is the latter 32 km x 32 km matrices that
have been used in the subsequent work.
As was discussed by LeSchack (1974), the data that make up the matrix can also be presented
as a histogram of frequency of occurrence of individual radiant temperature
values, and that the shape of the histogram is a function of the quantities of the ice
and clouds in the area covered. The variation of the shape of the histograms as a
function of ice types and cloud covers as interpreted from the imagery for 17 April
1974 is shown in Figure 8. Examination of these histograms shows that radiant temperature
values derived from mUlti-year ice have a normal distribution, multi-year and
first-year ice together have a positively skewed distribution, areas of ice with a cold
cloud cover have a negatively skewed distribution and areas of ice covered with warm
clouds have a normal distribution, but with a significantly higher mean radiant temperature.
The cause of this variation of histogram shapes is clear when the actual
temperatures of the ice and cloud types are considered.
Since visual examination of the histograms indicates that there are qualitative differences
as a function of ice and cloud types, it was determined to examine quantitatively
the IR data distributions, i.e., determine the mean, skewness and kurtosis of each 32 x
32 matrix. Accordingly, selected portions of passes 5725, 5989, 2010 and 2369 were
chosen, divided into arrays of 32 km x 32 km matrices, and then processed, using the
following statistical relationships:
M r
S
fl (Xl - m)r + f 2 (X 2 - m)r + •.•.••. fN(X N - m)r
fl + f2 + •••..••. fN
150
(1)
(2)
LeSchack 10

lated by the Maykut-Untersteiner Model as follows (Table 12):
TABLE 12
Radiant Temperature Surface Temperature
Satellite Date Measured Calculated
NOAA-l 15 Mar 1971 -33 -31.9
NOAA-l 21 Mar 1971 -26 -30.0
NOAA-l 17 Jul 1970 + 1 - 0.1
There are significant implications to any proposed automated ice mapping technique if it
can be shown that the Maykut-Untersteiner Model can be used to predict the mean temperature
of the multi-year ice for any day of the year. With such a predictive ability,
in conjunction with the measured M, Sand K values, it would then be feasible to determine
the possible range of ice temperatures that should be expected on a given day,
thereby eliminating from consideration by the algorithm areas of clouds or land masses
that fallout of this temperature range. Accordingly, one of the objects of the present
research was to investigate further the coincidence of calculated surface temperatures
with radiant temperatures recorded by the newer, VHRR-equipped satellites.
The imagery from the five passes used in the present research was examined by photointerpretive
techniques and multi-year ice areas within the regions for which digitized
data were available were selected. The criteria for selection were that (a) no
cloud cover be qbserved over the area; (b) that the ice appear unfractured, i.e., without
newly re-frozen areas; (c) that the area was far enough from land so that the ice
was not likely to be land-fast ice; and (d) that NIMBUS-5 low resolution microwave data
acquired independently on or about the same date showed brightness temperatures in the
same areas that are indicative of multi-year ice. The radiant IR temperatures of the
ice were then calculated from the numerical "count" provided bg the VHRR. As best as
could be determined, in the temperature range of interest, -40 C to DoC, the difference
between ascending integer values of the arbitrary count is + 1°C + 0.1°. An internal
blackbody reference in the radiometer is used to calibrate the arbitrary count, and
this in turn was checked (and corrected if necessary) against an external target consisting
of a body of known temperatures, viz. ice and water at a temperature of -1.2
to -l.SoC. Table 13 lists the comparison between satellite observed radiant temperature
values and surface temperatures for equilibrium ice calculated from the Maykut­
Untersteiner Model. An atmospheric correction (Smith, et al., 1970), which exceeds
O.loe only for temperatures higher than _22°C, has been-applied to the derived radiant
tempera tures •
Figure 9, a comparison of a section of Pass 2010 (17 April 1974) imagery
with the aSSOCiated skewness, mean, and kurtosis matrices, is shown
on the following four pages.
155
LeSchack 15

Line
SO 2952952922"7274283353315251271256262256245250241
279274281343284330217296269249254275267253249
62828127229028927611303282299278269275266252244240
650 )56292291318273278313279278212260268261248249
1012300113193613102942
1044 1434o_3183062933
l07325JlOlI63441073OJ334.Z89282
1102.90327334)08320319306360272298354362
K
250241253247230266311254
1161082772 as) 262 993822 712 9111 052 962 782 923142732873032832842 n3102 962 963 00
340316297292310307304.291306303311281294328281287274333315324321293296
295308_2912123092903422115312317299273275301283317297327363293306)20297
2116310331_321283113353.4'283333315324370285278285339280292283271284
2972953'01333314325 2912883102913133122832S9346310284281306313308263
3012902923012743043372,7630030929932S291307342316293291312310281295288325
3053112923721442843282911022743,26293J0733129728729S338312296311323285290
29730230S317298309.3033493262822853Jl32710S2832973302903012162913.02321
312326311284332.336282303275344303296298286339305304293275306331277279
321286298318 214291292287294267289298278307299311302315301348
FIGURE 9 (C): The kurtosis matrix corresponding to the imagery in 9 (a) is
shown. The shaded areas of the matrix have K values greater than 3.75
and are associated with the transition zone of multi-year ice and first-year
ice. The outlined area at upper right is associated with the Banks Island
land mass. This area has an average K value of 2.44, which is significantly
lower than the average K value for any similar sized area on the imagery.
158 LeSchack III

data distribution corresponding to the multi-year ice should then be compared with the
values calculated by the Maykut-Untersteiner Model.
Since the Maykut-Untersteiner Model was based on inputs associated with perennial Arctic
sea ice having only a small percentage of open water, it would, accordingly, be valuable
to modify this model, possibly in conjunction with the Schaus and Galt (1973)
open-lead model, and determine what effect increasing amounts of open water have on the
distributions of radiant temperatures assGciated with multi-year and first-year ice.
Additionally, it would be important to examine imagery that covers a range of areas
of open water. Since the amount of open water would be nearly impossible to obtain by
normal "ground truth" procedures, LANDSAT imagery, with a spatial resolution of 100 m,
could be particularly valuable for providing this complementary information during the
summer months.
Another key to an algorithm for objectively delineating ice types, clouds, and land
masses appears to be the use of the mean radiant temperature, skewness and kurtosis
of arrays of radiant temperature distributions. Although the analysis applied to pass
2010 appeared promiSing, this type of analysis should be applied to equally large
areas recorded at different times of year. Additionally, an objective way of delineating
the important surface features should be attempted. In the present work, subjective
judgment was used to interpret the M, Sand K values as predictors of features
on the imagery. In new research, it would be worthwhile to use a wholly objective
technique, such as cluster analysis, to delineate the features.
The above two recommendations appear to be the next steps to confirming the basic theory
outlined above. Upon such confirmation, an algorithm would have to be developed to implement
the various statistical techniques and present them in a rapid, useable format
to those organizations concerned with ice mapping and movement.
REFERENCES
Hibler, W.D. III, Mock, S.J. and Tucker, W.B., III, (1974), Classification and Variation
of Sea Ice Ridging in the Western Arctic Basin, J. Geophys Res. V. 79, n 18,
pp 2735-2743
LeSchack, L.A., Hibler, W.D. III and Morse, F.H., (1970), Automatic Processing of
Arctic Pack Ice Data Obtained by Means of Submarine Sonar and Other Remote Sensing
Techniques, Tech. Rept, under ONR Contract N00014-70-C-OIIO by the Development and
Resources Transportation Co., Silver Spring, Maryland 20902. (Also in AGARD Conf.
Proc. No. 90, NATO, "Propagation Limitations in RelOOte Sensing, October 1971).
LeSchack, L.A. (1974), Potential Use of Satellite IR Data for Ice Thickness Mapping,
Interim Report, NOAA Contract 3-35384, or published in, The Coast and Shelf of the
Beaufort Sea. Arlington, Va: Arctic Institute of N.A., 1974, pp 243-267.'
Lyon, Waldo K. (1963) The Submarine and the Arctic Ocean, Polar Record, 11:75
(699-705) •
Maykut, G.A. and untersteiner, N. (1971), Some Results from a Time-Dependent TherlOOdynamic
Model of Sea Ice, J. Geophs. Res. V 76, No.6, pp 1440-75
Schaus, R.H. and Galt, J.A. (1973), A Thermodynamic Model of an Arctic Lead, Arctic,
V 26, No.3, pp 208-221.
Smith, W.L, Rao, P.K., Koffler, R., and Curtis, W.R. (1970), The Determination of Sea
Surface Temperature from Satellite High Resolution Infrared Window Radiation Measurements,
Monthly Weather Review, V 98, No.8, pp 604-611.
162 LeSchack 22

SECTION 5
ICE AND ENVIRONMENT
Carlson, R. F.
A TheoPy of spring Rivep DisahaPge into· the Apatia Iaepack
(University of Alaska, United States)
Chen, E. C. and B. F. Scott
Aging ChaPaateristias of C!'ude Oil on Iae
(Inland Waters Directorate, Environment Canada, Ontario, Canada)
Getman, J. H., Lt., L. A. Schultz and P. C. Deslauriers
Tests of Oil ReaovePy Deviaes in Iae Coveped Wateps
(U.S. Coast Guard, Washington D.C., United States, and ARCTEC, Incorporated,
Maryland, United States)
Karlsson, T.
Wave Runup Due to Flash Floods Caused by Subglaaial Volaanism
(University of Iceland, Reykjavik, Iceland)
Lewis, E. L.
Oil in Sea Iae
(Department of Environment, British Columbia, Canada)
Tilsworth, T.
Alaska Coastal Enviponmental Engineering
(University of Alaska, United States)
Topham, D. R.
Simulation of an Oil WeU Blowout in Shallow Coastal Wateps
(Department of the Environment, British Columbia, Canada)
Weller, G.
PoZaP Meteopology - A Review of Some Reaent ReseaPah
(University of Alaska, United States)
163

A THEORY OF SPRING RIVER DISCHARGE INTO THE ARCTIC ICEPACK
Robert F. Carlson
Director and Professor of Hydrology
Institute of Water Resources
University of Alaska
Fairbanks, Alaska
United States
EXTENDED ABSTRACT
The spring breakup of arctic rivers presents one of the more significant and spectaauLar
events in the arctic coastal zone. The flow ocaurring from snowmelt in the mountains
to the south abruptly rises to a peak and falls through a strong recession. The resulting
flow provides the major freshwater input into the coastal system within a short
period of time along with sediment as bed and suspended load. As the implications of
the breakup phenomena are many, an understanding of the dynamics of arctic river breakup
is important to the nearshore coastal zone.
The breakup phenomena has been often observed on the Mackenzie and Colville Rivers and
somewhat less on other arctic rivers. Usually the flow is described as going over the
top of the ice creating a rather sizeable ponding of water over an extensive area. Many
have mentioned the sediment transported on top of the ice drastically affecting the
coastal regime of the arctic. In this paper a theory of an alternative flow occurrence
is offered which predicts two primary results: most of the [low must go under the ice;
and, large scour channels evolve as a result.
The following description can probably be extended to most of the arctic streams along
Alaska's coastal zone. However, the author is not familiar enough with other coastal regimes
to make an unqualified extension. The computations are based on the Kuparuk River
which is located about twenty miles to the west of the Prudhoe Bay oil region. The
rivers in the area enter the coastal zone where barrier islands are approximately three
miles from the coast.
The streamflow is characterized by a rather abrupt breakup which begins shortly after
June 1st and proceeds to rise to a peak within about 5 days. An abrupt recession begins
and the flow drops to a low level in about 15 days. The computations assume a peak of
100,000 cfs with a simplified triangular hydro graph shape. In this region bottom fast
ice ocaurs between the coastline and the barrier islands with an unknown amount and extent
of water between the ice and the bottom. The offshore ice is assumed to be anchored
to the bottom.
The approximate division of flow to the top of the offshore ice must first be calaulated.
Given a depth of flow on the ice of a truncated cone five by one feet by five miles in
radius, only 30 percent of the streamflow goes onto the ice surface. This volume in turn
reduces the triangular hydrograph to approximately 82,000 cfs. The simple argument follows
that since the 30 percent of the flow is available to the ice surface, the remainder
must go under the ice and the flow mechanisms must accommodate a flow of 82,000 cfs.
165

The under ice flow geometry assumptions are: ice remains grounded and is 6 feet thick;
the stream water flows through a rapidly enlarged conduit; the conduit is approximately
five miles long; and the conduit has an available head of 11 feet to be dissipated by
shear friction. These numbers are approximate and only illustrate the theory. The next
step is to assume a conduit of some dimension underneath the ice entirely within the gravel
bottom. Using the common hydraulic pipe flow equation, the several assumed shapes
require the following dimensions to accommodate the flow: circular shape, a 96 foot diameter
section is needed; rectangular shape, 6 feet deep, a width of 4400 feet is needed;
and, type B erodable channel section, a depth of 60 feet and a top width of 270 feet is
required.
The ramifications of these results are many. Time does not allow a complete discussion
except to mention that the phenomena appears to be of the size and significance to
warrant a rather careful engineering investigation and consideration in design of nearshore
and offshore structures.
Direct verification of the results of the theory are nearly impossible. One can appreciate
the difficulty of attempting to measure an under ice conduit at the time of a rather
large flow through the ice pack. A tentative verification has been made by estimating
apparent conduit widths from several U2 photos taken of the Kuparuk and Prudhoe Bay region
during the spring breakup season of 1974. These photographs indicate a single channel
width of 718 feet for the Kuparuk River and several multiple channels off the
Putuligayuk and Sagavaninktok Rivers. These indications of the channel existence provide
a tentative validation of the theory.
ABSTRACT ONLY AVAILABLE
166 Carlson 2

AGING CHARACTERISTICS OF CRUDE OIL ON ICE
E. C. Chen and B. F. Scott
Inland Waters Directorate, Environment Canada
Ottawa, Ontario
Canada
EXTENDED ABSTRACT
Knowledge on the aging characteristics of oil is useful for assessing the effect or planning
the cleanup of an oil spill. This work investigates the aging of crude oil on ice.
Laboratory experiments were carried out in an environmental chamber (127 x 60 x 46 cm)
with temperatures varying from -5 to -40°C and air flow rates from 0 to 18 l/min. TWo
different types of crude oil, Norman Wells and a Pembina blend were investigated. The
crude was poured on a flat ice surface and oil samples were withdrawn at selected intervals.
The changes in viscosity, density, surface tension and refractive index were determined
as a function of the elapsed time. All of these physical parameters were found
to increase as the oil aged; the higher the temperature or the higher the air flow rate,
the greater is the increase. Gas chromatography was also used to monitor the changes in
the n-alkane concentrations as a result of aging.
Field tests were conducted in man-made ponds (15.6 x 7.2 m) near a river during the winter.
Oil was poured on the natural ice surface and allowed to age there for several
weeks. Samples were taken regularly and their viscosity, density and water content analyzed.
The viscosity and density of the crude showed an increase with elapsed time while
the water content was found to fluctuate depending on the weather conditions. Composition
of the aged oils were investigated by gas chromatography. For Norman Wells crude
oil, an estimated 52% of the oil had evaporated before the normal spring break-up.
ABSTRACT ONLY AVAILABLE
167

TESTS OF OIL RECOVERY DEVICES IN ICE COVERED WATERS
Lt. James H. Getman
Office of Research and Development
U. S. Coast Guard
Washington, D. C.
United States
and
Lawrence A. Schultz and Paul C. Deslauriers
ARCTEC, Incorporated
Columbia, Maryland
United States
ABSTRACT
As part of its Arctic Pollution Response Program, the United States Coast Guard conducted
tests on two oil spill recovery devices operating in a simulated Arctic environment incorporating
below freezing temperatures and ice infested waters. The objective of the test
program was to determine the oil recovery capability of the two devices tested operating
in an oil/ice/water environment. Tests were conducted in broken fresh water ice and broken
salt water ice, with No. 2 diesel oil and crude oil selected to closely match the properties
of Prudhoe Bay crude oil, and at temperatures of +25°p and +15°P. These tests demonstrate
that with minor hardware modifications and the proper operating procedures, both
devices will successfully recover crude oil and No. 2 diesel oil spilled in a broken ice
field of moderate ice piece size.
169

INTRODUCTION
The present world-wide energy situation has focused attention on the Arctic as a
major source of oil. Petroleum exploration, production and transportation activities
in the Arctic will result in an increased potential for oil spills in ice infested
waters. The U. S. Coast Guard has the responsibility for promulgation and enforcement
of regulations concerning oil pollution of U. S. waters. A portion of the U. S. Coast
Guard's Arctic Pollution Response Program is being concentrated on the development
of a near-term Arctic oil pollution response capability. This effort is directed
toward the development of systems to detect, control, clean up and abate Arctic oil
pollution in the various oil/environment interactions that may exist. Examples of
these oil/environment interactions include oil spilled under ice, oil spilled on
top of ice, oil sandwiched within solid ice, oil contained between ice floes, and
oil along ice locked shorelines.
A major test of available oil spill recovery equipment was conducted by the Coast
Guard at Homer, Alaska in late 1973. The purpose of the test program was to evaluate
existing off the shelf oil containment and recovery devices in a freezing or near
freezing ice/water environment and to determine the effect of cold temperatures and
ice on their operation. Oil was not spilled for these tests for logistic as well
as environmental reasons. The tests showed that of the six recovery devices tested,
two devices, the Lockheed disc drum unit and the Marco oleophillic belt unit, could
be capable of recovering oil from an oil/ice/water environment. The.tests described
in this paper were directed toward an evaluation of the capability of these two
existing oil recovery systems in broken ice fields of moderate ice piece size.
The two devices tested were the Lockheed Clean Sweep Model R2003, a nominal 4 foot
diameter by 7 foot wide disc drum unit, and the Marco Pollution Control Class I Oil
Recovery System, an oleophillic belt device having a nominal belt width of 1 foot. A
sketch showing the method of operation of the Lockheed unit is presented as Figure 1.
The Lockheed Clean Sweep operates on the principle of oil adhering to an oil wetted
surface. The drum is immersed in the oily water and is rotated. The paddle wheel
effect of the cross vanes creates a current which causes the oily water to flow
into the drum between a series of parallel, vertically positioned metal discs as
shown in the sketch. The oil adheres to each rotating disc until it comes in contact
with stationary wiper blades which are mounted over the centerline of the unit and
wipe the oil from the disc. The oil then tends to run down the wiper due to
gravity forces and falls into a central trough which runs through the center of the
discs. A conveyor screw in the trough moves the recovered oil from the central
trough to a sump for disposal.
Figure 2 is a sketch of the Marco Class I Oil Recovery System. In operation, oil
and oil soaked debris are lifted from the surface of the water on a continuous
conveyor of Filterbelt material. The Filterbelt is a stranded open cell synthetic
foam material which removes oil from the surface of the water as water flows through
the open cells of the material. Oil entrained within the Filterbelt is then wrung
from the belt by a pneumatic tensioned squeeze roller. The oil drops into a reclaimed
oil sump and oil soaked debris is retained on a grid mounted over the sump. The
unit also incorporates an underwater impeller which draws the oil and water through
the belt to improve the unit's oil recovery efficiency primarily in stationary applications.
TEST SET-UP AND PROCEDURES
These tests were conducted for the U. S. Coast Guard in the Ice Model Basin of
ARCTEC, Incorporated, located in Columbia, Maryland during March and April of 1975.
170 Getman et aZ. 2

The Ice Model Basin consists of a 100 foot long, 12 foot wide, and 5 foot deep test
basin situated within a heavily insulated cold room which can be refrigerated to
+15°p with a mechanical refrigeration system or -150 o p by the controlled injection
of liquid nitrogen into the cold room. An aluminum carriage spans the width of the
basin supported through low friction ball bushings by cylindrical stainless steel
rails mounted to the basin walls. The carriage is propelled over a speed range
of 0.01 to 20.0 feet per second by an endless pretensioned stainless steel cable
driven by an electric motor through an eddy current clutch, both of which are located
outside the cold room. The test plan called for testing each of the two oil recovery
units in 95% coverage of both fresh water ice and salt water ice, in two thicknesses
of No.2 diesel oil and crude oil, at two temperatures of +25°p and +15°P. The ice
piece size ranged typically from 3.x 3 x 3 inches to 22 x 22 x 10 inches with
10 x 10 x 6 inches being about average. The nominal oil thicknesses selected for
the test program were 1/2 inch and 2 inches. Oil was deposited on a volumetric
basis with the volume deposited based upon the nominal thickness spread over the entire
1200 square foot surface area of the model basin without consideration given to the
ice cover. The actual thickness of the oil between ice pieces corresponding to the
1/2 inch nominal thickness was about 2 inches, while that corresponding to the 2 inch
nominal thickness was about 5 inches. While it was desired to conduct tests with
Prudhoe Bay crude oil, the cost of transporting Prudhoe Bay crude to the laboratory
was prohibitive. A substitute crude having physical properties very closely matching
Prudhoe Bay crude was located in south Texas and this crude was used for this
test program.
The test plan called for all tests to be conducted at a constant speed of advance for
the oil recovery devices. In order to determine the best speed of advance to use for
the test program, a preliminary test series was conducted before oil was added to
the ice/water mixture in the basin. In general, oil recovery operations in the field
are conducted in the 1 to 2 feet per second speed range. The preliminary tests conducted
at these speeds indicated that damage caused by impact of the oil recovery
devices by the ice chunks could be excessive at these speeds. Tests conducted at a
forward speed of 0.1 feet per second significantly reduced the ice impact problem,
however this speed was felt to be unrealistically low for field applications where
the vessel upon which the oil recovery device is mounted must maintain steerage. A
compromise value of 0.5 feet per second was then established as the standard forward
speed for the entire test program.
The model basin's main towing carriage was modified to provide the primary means of
support for both of the oil recovery devices tested. The power unit supplied with
the Marco device was used to supply power to both oil recovery devices. This
package consisted of a 6.4 horsepower air cooled diesel engine driving a hydraulic
pump and an air compressor. The test program was based upon testing each unit over
half the length of the model basin. After the oil/ice field was prepared for a test,
a divider was installed across the width of the basin to separate it into equal
sections. This division of the basin permitted a test to be performed in one half
of the basin with one device without disturbing the oil in the remaining half of the
basin. The test of the other device under the same conditons would subsequently be
conducted in this undisturbed half of the basin.
Data recorded during the test program included both physical property data and
operating data. Prior to a test, samples of oil were removed from the surface of the
basin and tested to determine the specific gravity, viscosity, surface tension,
temperature, and emulsification of the oil. Other data recorded preceeding a test
included the oil thickness, the ice thickness, the ice cake size, ice percent coverage,
ice salinity, ice temperature, water salinity, water temperature, and the
173 Getman et at. 5

immersion of the oil recovery device below the oil surface. Data measured and
calculated related to the operation of the devices included the belt or drum speed,
the speed of advance, the time of recovery, the total volume of oil, water and ice
recovered, the oil/water/ice recovery rate, the volume of oil recovered, the oil
recovery rate, the recovery efficiency, the volume of oil encountered, and the
throughput efficiency. Samples of the recovered oil were removed after each test
and tested to determine the viscosity, specific gravity, emulsification, and surface
tension of the recovered oil. Data was also recorded on the absorption of oil by
the ice.
The general testing procedure involved the preparation of the water/ice/oil conditions
on the afternoon preceeding the day of testing thereby allowing the test area to
soak overnight at the specified temperature. After the basin had been held at the
prescribed temperature overnight, the ice chunks that froze together were broken
apart. Oil samples were removed for the measurement of oil physical properties and
the oil recovery devices were prepared for testing. For the Lockheed unit, test
preparations consisted of hosing down all frozen components with hot water. The drum
was then rotated in air before being lowered into the oil/ice/water mixture to insure
that it was ice free and operating properly. In conducting the tests, once the drum
was lowered into place and secured, the drum was rotated and as soon as oil entered
the sump, the carriage drive system was engaged. For the Marco unit, test preparations
consisted of thawing out all air supply lines and ensuring that the belt, squeeze
rollers, tensioning device, impeller, and the belt drive were all operational in air.
The boom was then lowered to the specified depth for the test. In testing, the
Filterbelt was rotated and as soon as oil entered the sump, the forward drive system
was activated. Upon completion of a test, the boom was raised out of the oil/ice/water
mixture and the oil remaining in the belt was squeezed into the sump. Figure 3 is a
photograph of the Lockheed oil recovery device being prepared for testing in the basin.
TEST RESULTS
The results of this test program include both quantitative results concerned with the
oil recovery capability of the units tested and qualitative results concerned with
the operating characteristics of the units in the harsh simulated Arctic environment.
The recovery variables of primary concern are the oil recovery rate, the oil recovery
rate per unit width of the oil recovery device, the oil recovery efficiency and the
throughput efficiency. Figure 4 is a plot of the oil recovery rate data obtained
from the test program as a function of oil type and air temperature. The oil recovery
rate is defined as the net volume of oil recovered divided by the time of recovery.
For the tests conducted in diesel oil, the performance of the Marco unit is seen to
be essentially independent of the nominal oil thickness over the range tested and the
value of the air temperature over the range tested. The average oil recovery rate of
this unit operating in diesel oil was 5.5 gpm. For the Lockheed unit operating in
diesel oil, the average recovery rate was 30.4 gpm with a somewhat larger range of
variation. The recovery rates obtained during the crude oil tests varied over a much
wider range for both units. In crude oil the average recovery rate for the Lockheed unit
was 29.1 gpm while that obtained for the Marco unit was 11.7 gpm. Some of the
variations realized in the Marco test results can be attributed to the varying amount
of crude oil that was being conveyed up the belt in a layer on top of the belt surface
in addition to the oil which was entrapped within the fibers of the Filterbelt. Ice
interacted with the layer of heavy crude oil that was lying on the Filterbelt in
varying ways, sometimes scraping oil off the surface of the belt and in other instances,
not interfering with the conveying effect to any appreciable extent. In these tests,
the type of ice, whether salt water ice or fresh water ice, seemed to have little
effect on the test results.
174 Getman et aZ. 6

Figure 3. Photograph of the Lockheed Unit Being Prepared
for Testing in Diesel Oil and Fresh Water Ice
175 Getman et al. 7

Figure 8 is a plot of the oil recovery rate per foot of device width versus oil
viscosity. The two plots presented for the 0.5 inch nominal oil thickness show no
consistent variation or trend in the oil recovery rate as a function of oil
viscosity. However, in the case of the 2 inch nominal oil thickness tests, both
the Lockheed device and the Marco device show a trend in increasing unit oil recovery
rate with increase in the oil viscosity.
Several problems associated with operating each of these devices in an oil/ice/water
environment at low temperatures were revealed during the test program. A major
problem associated with the Lockheed unit was damage to the vanes due to interaction
with the ice. The ice had a tendency to bend the vanes at the ends as
shown in Figure 9 and generally loosen the vanes from their forward connection
point with the discs of the unit. For the test program, this problem was minimized
by strapping the vanes at each end with aircraft cable. Other problems associated
with the Lockheed unit include the small size of the sump which limited the storage
capacity and intensified the pumping problem associated with pumping the slurry of
oil, water, and ice chips. Ice chunks often wedged between the drum and the frame
of the unit. Freeze-up of moving parts overnight was often a problem especially
when operating in fresh water requiring a thawing of the unit by hosing with hot
water. In softer salt water ice, the rotating edges of the vanes often shaved the
ice into fine ice chips and slush which subsequently mixed with the oil and intensified
the slurry pumping problem. Problems associated with the Marco unit included
the overnight freezing of moving parts including the freezing of the Filterbelt to
the pan supporting it in fresh water tests, pumping of the collected mixture of oil,
ice slush, and ice chips, and the pile-up of ice pieces ahead of the boom as the
device progressed through the test run as shown in Figure 10. Some abrasions and
tearing of the Filterbelt did occur during the course of the test program, but the
damage was relatively limited. Although these problems were revealed when operating
the recovery devices in the test environment, it is important to note that both
units survived the harsh test program and, with a few adjustments, both were able
to recover oil throughout the entire testing period.
Aside from the ability of the devices to remain operational throughout the test
program, they each had certain design characteristics which proved advantageous
when encountering oil in a broken ice field. The rotating drum of the Lockheed unit
had a tendency to submerge the oil covered ice chunks beneath the water surface as
the ice chunks enountered the drum as shown in Figure 11. As a result of this, some
oil adhering to the surface of the ice had a tendency to float free to the oil/
water interface due to buoyancy, and a cleaner piece of ice would rise to the surface
on the downstream side of the unit. The Marco device had a tendency to recover
more oil coated ice pieces. While the recovery of slush ice and small ice chunks
presents greater pumping problems, the recovery of the small oil coated ice pieces
may be desireable in field operations.
CONCLUSIONS
The major conclusions which can be drawn from this program are as follows:
1. Both of the oil recovery units tested, the Lockheed R2003 Clean Sweep and
the Marco Class I Oil Recovery System, successfully recovered crude oil and No. 2
diesel oil spilled in a broken ice field of moderate ice piece size, and both
survived the rather harsh test program.
2. Both of the oil recovery units tested offer the potential for substantially
improved performance and reliability when operated in a broken ice field after
relatively minor adjustments and modifications are made to the units to eliminate
181 Getman et al. 13

Figure 9. Photograph of a Heavy Gauge Vane on the Lockheed Unit
After Being Bent by Interaction with Ice
183 Getman et at. 15

Figure 10. Photograph of the Build-Up of Ice
in Front of and on the Marco Unit
184 Getman et al. 16

Figure 11. Photograph of the Ice Processing
Action of the Lockheed Unit
185
Getman et al. 17

ACKNOWLEDGEMENT
The work described in this paper was sponsored by the Department of Transportation.
U. S. Coast Guard under Contract DOT-CG-S1487-A. The opinions or assertions contained
herein are the private ones of the writers and are not to be construed as
official or reflecting the views of the Commandant or the U. S. Coast Guard at
large. The citation of trade names and manufacturers does not constitute endorsement
or approval of such products.
187 Getman et at. 19

Figure 1. Map of Iceland and the surrounding areas.

The water flood wave breaking out from underneath the glacier will sDread
out over the sand flats east of the mountain range, but according to 1918
accounts, the greatest part of the total flow will find its way through the
pass between the mountains and Hafursey (Fig. 2}. From there it spreads
out over the sand and reaches the sea as a 12 km wide wave which runs on
both sides of Hjorleifshof6i.
An indication of the forces at work here may be found in Figure 3 which was
taken about a week after the 1918 eruption started at a location shown on
Figure 2. It shows two blocks of ice, presumably broken off the glacier by
the rushing water and deposited at this location. From the size of the two
men standing on top of one of the blocks, their size can be estimated. The
light colored block seems to be on the order of 10x25 meters and the other
block at least 12x20 meters. Visual evidence in the channel were reported
to indicate a water depth of 60 to 70 meters, but these observations are
probably somewhat distorted, although a water depth of 35 to 40 meters is
entirely possible. These ice blocks were found about 5 km from the water
outlet from the glacier as indicated in Figure 2.
Figure 4 gives another indication of the volume of the flood wave. The
photograph was taken in May 1975 down on the flat sand approximately 15 to
20 km from the origin of the flood. The picture shows a rock which was
carried by the flood water and deposited at the location shown in Figure 2.
Using the man in the picture as a scale gives a diameter of 8 m and a height
of some 5.6 m. A hole has been dug by the side of the rock down to a depth
of 2 m, where the digging was stopped due to water in the hole. It seems
therefore fairly safe to assume a rock height of about 8 m making the rock
approximately spherical with a volume of at least 270 cubic meters.
Samples taken of the rock show an average density of water saturated rock
of 2.12 metric tons/m 3 making the total rock weight of about 570 metric
tons.
The flood carried with it a tremendous amount of sand and volcanic ash as
indicated in Figure 2 where the approximate shoreline contour before and
after the eruption as well as the coastline at present are shown. The
flood moved the coastline out a distance of some 2000 m due to sand and
volcanic ash deposits, and reports from an earlier eruption (1721) relate
how one could walk on dry land after the eruption where before were the
deepest fishing banks of 70 to 80 fathoms depth.
Figure 2 shows also how the sandy beach below the village of V{k has been
moved out as a result of sand deposits which have been transported from the
1918 eruption flood deposits below Hjorleifsh6f6i.
FLOOD DANGER AT THE VtK VILLAGE
The village of V{k has a population of 400 to 500, most of whom live in
homes built in the foothills, but some buildings are located down on the
sand flats. The village master plan calls for extensions of the village
up into the foothills, but the villagers have expressed an interest in
moving down onto the sand. These wishes have up to now been resisted by
the planning authorities and by the Icelandic Civil Defense Council on
grounds of flood danger by a Katla eruption when and if one occurs. These
floods may come by either of two routes - on land or from sea. The land
route of the flood wave will follow the south side of the mountain range
to V{k (Figure 5) whereas the sea route will be formed by a flood wave
which moves to all directions at sea from the beach where the main flood
192 Karlsson 4

Figure 5 .
Figure 5 .
Katla flood area showing possible overland fleod
route to the Vik village . Route of main wave to
sea is indicated .
Map showing profile leveling sections on the
sand flats east from Vik .
196
Karls s on 8

.....
o
Figure 8 .
" ':.
Depth soundings of the waters off Vlk .
ViK - KOTLUTANGI
Soundi"91 '0( mean low watH SptlrMjlS
Conform conic Pl'OJKtion
Surveyed In Aug.· Sept 1913
SV-13

that the wave travels in each channel independent of the other channels
and by using the depth data (Fig. 8) the change in wave height as the wave
propagates out may be evaluated by the method described below.
ho
0
yo
/
{::,X
1 0
hI
YI EHl'
t
{::,X
11
j
0 1 0 1
Figure 10. Wave traveling in a channel of variable depth and width.
Consider a wave traveling over still water as shown in Figure 10. The wave
propagates at a velocity given by (Chow 1959, p. 555)
c = Igy(l + 3h/2y +(h/y)2/2)
and from continuity the following expression is obtained
c = (y + h)v/h (4 )
where v is the particle velocitv. Equation (3) shows that for small values
of h the wave velocity is reduced to
c = ;gy ( 5 )
which is the well known equation for waves traveling over shallow water.
If Y on the other hand is small compared with h, the wave velocity c as
given by (3) becomes very large. In that case friction on the channel bed
becomes a dominating factor and the flow velocity is given bv Manning-s
equation
where R is the hydraulic radius of the channel cross section and S is the
bottom slope. Equation (3) or equation (6) together with (4) are then used
depending upon which combination gives the lowest value for c.
An element of length {::,X between sections 0-0 and 1-1 (fig. 10) will be considered.
The mean wave propagation velocity over the element is (co+cl)/2
and the travel time of the wave from 0 to 1 is
From continuity it is found that the volume added to the body of water between
0 and 1 is equal to the inflow at 0:
1
( 3 )
(6 )
(7)
( 8 )
200 Karlsson 12

N
o
I-'
Sec.tlons -
Figure 11,
6
2 km
10 11 12 13
I I
Refracted wave rays approaching Vlk,
'.
I
I:;:;
J "
;!
f
--- i
/8
,.$2
;;
Ii c
. U
63'24-

205
'H
o
Karlsson 17

CasuaL observations made by oiL company empLoyees at existing bLowouts suggest that the
formation of stabLe emuLsions is improbabLe and that oiL particLes rising to the surface
may reform into a sLick. A series of studies on emuLsification is being carried out by
Drs. Topham and Cretney to provide confirmation. Incorporation of such a sLick into
growing sea ice has been studied in the Laboratory by WoLfe and BouLt (L9?4) who used a
comparativeLy smaLL tank so that the oiL compLeteLy covered the ice/water interface.
This prevented them from noting the typicaL sessiLe drop configuration of oil at the ice/
water interface. Photographs of these drops have been pubLished by KeeviL and Ramseier
(L9?5) and the matter has been investigated in detaiL by Rosenegger (L9?5) in continuing
contracted studies. Both Norman WeLLs and Swan BiLLs crude, taken as the extremes of the
range of oiLs, have sessiLe drop thicknesses of approximately 5 mm. Thus the minimum
thickness of an oiL sLick beneath ice has this vaLue and is controLLed by surface tension
effects; much deeper pooLs can occur due to ice/water interface topography. PLanned experiments
wiLL determine the vaLue of the "sticking friction" for sessae drop movement
and the forces required to sustain continual movement of such a sLick. Once such a sLick
or oiL Lens has ceased movement, it will become incorporated within the ice sheet by
naturaL growth processes at the appropriate seasons. Study G2, being conducted by NORCOR
of YeLLowknife at BaLaena Bay, Cape Parry, N.W.T., is concerned with the details of this
incorporation and is being reported on elsewhere at this meeting. It appears that after
incorporation, the oiL Loses LittLe if any of its higher fractions. The experiment
"Arctic IV" conducted by the McInnes Foundation of Toronto at ResoLute Bay, N.W.T., in
L9?4 indicated that once the snow had been removed by the intense radiation to be expected
in earLy summer, oil beneath the sea ice wouLd meLt its way to the surface rapidLy due to
the extreme contrast in aLbedos. As the oil moves upward in the ice sheet it enters regions
of higher radiation and progresses at an ever increasing rate; it finaLLy ends
fLoating on the water surface of its own meLt pooL. The NORCOR experiment is further
designed to study this effect, measure aLbedo change with time in the summer of L9?5, and
report on the eventuaL degradation of the fLoating oiL by radiation. The enhanced radiation
absorption per unit area wiLL thus be determined. The potentiaL of burning as a
method of cLean-up wiLl aLso be studied.
The containment of the oiL beneath the sea ice by interface topography is a probLem of
considerabLe difficuLty at the present LeveL of knowledge. The oiL wiLL be confined by
different scaLes of roughness, the Largest arising from mechanicaL faiLure of the ice
sheet to form pressure ridge keels. The statistics of the topography of the upper surface
of sea ice cover in the Beaufort Sea in both nearshore and offshore regions are
being investigated by the author's coLLeague, Dr. P. wadhams, by anaLysis of Laser profiLometer
data taken in September L9?4 and April 19?5. An attempt has been made to relate
along-ridge and along-keel profiles using data obtained by Dr. Francois of the Applied
Physics Laboratory, University of Washington, with the UARS unmanned submersible near T3
(private communication 19?5). This is the only data of its type available. The containment
of oil slicks by the keels is similar to that of containment by booms. Wilkinson
(l9?2) has studied this latter problem; Lau and Kirchhefer (l9?4) have produced a useful
review of literature. At the author's request, Moir and Lau (l9?5) extended their work
on boom containment and investigated the effects of simulated ice cover with a ridge keeL
in the flume tank at the Canada Center for Inland Waters, Burlington, Ontario. The
slope angle of the ridge had only a slight effect on containment. Under given initial
conditions shear flows can build up internal waves at the oil/water interface which can
then slop oil over the keel and allow escape.
In their original work, CampbeLL and Martin (Z9?3) postulated "lead-matrix pumping" as a
mechanism of oil distribution through the pack. We have not been able to devise a model
to test this hypothesis and if sufficient oil were released into the natural environment
to allow useful observations it would probably constitute a major disaster in its own
right. However, it is thought that all mechanical deformations of sea ice and ice growth
are best regarded as mechanisms for translating "free oa" resulting from the bLowout into
"fixed oa" within the sea ice. Consider an oa sUck of 5-mm thickness in a slowly
208 Lewis 2

alosing lead. The two masses of iae will first impaat at a aertain point, and later a
seaond point a distanae down the lead forming a long thin pool. If the motion progresses,
pressure ridge formation shortens this pool until the oil sliak has a thiakness equal to
the iae. Further movement will aause arude oil of speaifia gravity less than 0.9 to flow
over the surrounding snow and iae. The end result after aomplete lead alosure would be
aonsiderable surfaae aontamination, and large quantities of oil in the interstiaes between
the iae bloaks forming the pressure ridge where it might well produae a low strength
struature by exaluding the water required for bloak welding. It is probable that the oil
would remain inaorporated within the iae until its eventual melting.
Dr. E. R. Walker of this Group will use results from the above studies to prediat possible
alimatia effeats of the blowout based on the model of Maykut and untersteiner (l97l).
It is aonsidered that a good estimate of the maximum area aovered by a given quantity of
oil in iae aovered waters may be obtained by aonsidering that it is distributed in a uniform
layer of thiakness l mm above and 5 mm below the iae. Albedo ahanges remain to be
seen but the alimatia effeats of one blowout wiU be very minor. Movement of iae may distribute
this aalaulated maximum area widely so that alean-up operations would have to deal
with separated poakets. Suah operations should be aonaentrated in early summer as the
oil rises to the iae surfaae; burning from the surfaae of melt pools before weathering
appears attraative. If the oil is aUowed to esaape from these pools or the blowout oaaurs
in open water, the situation regarding spread is greatly worsened. Probably the major
hazard of a blowout in the Beaufort Sea is to the immense sea bird aolonies. These feed
on leads in early summer and aould be oiled and die by the million in unfortunate airaumstanaes.
REFERENCES
Ashton, G. D. 1974. Air bubbler system to suppress iae. Speaial report 2l0, u.S. Army
Cold Regions Researah and Engineering Laboratory, September 1974, pp l-36.
Ayers, R. C., H. o. Jahns and J. L. Glaeser. 1974. Oil spills in the Aratia Oaean:
extent of spreading and possibility of large saale thermal effeats. Letters in
Saienae 86:843-844.
Campbell, W. J. and S. Martin. 1973. Oil and iae in the Aratia Oaean: possible largesaale
interaations. Saienae l8l:56-58.
Glaeser, J. L. and G. P. Vanae. 1971. A study of the behavior of oil spills in the
Aratia. AD7l7l42 National Teahniaal Information Center, Springfield, Virginia,
U.S.A.
Keevil, B. E. and R. o. Ramseier. 1975. Behavior of oil spilled under floating iae.
Proa. Conf. on Prevention and Control of Oil Pollution. Ameriaan Petroleum Institute,
San Franaisao, Marah, pp 497-50l.
Lau, Y. L and S. A. Kirahhefer. 1974. A review of the dynamias of aontained oil sliaks
in flowing water. unpublished manusaript, Hydraulias Division, Canada Center for
Inland Waters, Burlington, Ontario, Marah, 29 p.
MaMinn, T. J. 1972. Crude oil behavior on Aratia winter iae. Final report projeat
734l08, Offiae of R. and D., U.S. Coast Guard, September.
Maykut, G. A. and N. Untersteiner. 1971. Some results of a time dependent thermodynamia
model of sea iae. J. Geophy. Researah 76(6):l550.
209 Lewis 3

ALASKA COASTAL ENVIRONMENTAL ENGINEERING
Timothy Tilsworth
Associate Professor and Head
Program of Environmental Quality Engineering
University of Alaska
Fairbanks, Alaska
United States
EXTENDED ABSTRACT
Alaska possesses some 38.4 percent of the U. S. tidal shoreline. This amounts to 33,904
miles, most of which is uninhabitated and undeveloped. Environmental impact and degradation
is and has been concentrated largely near communities and industrial operations.
Most of Alaska's communities are located near the coastline and most industrial operations
are headquartered in coastal areas. Pollution of Alaska's environment has and will continue
to be found where man's activities abound. Although the State has a relatively
small population (302,l73 - u.S. Bureau of Census, 1970) for its large land mass (586,400
square miles) the potential for environmental damage is tremendous because of exploration
and development of natural resources. Through a lack of engineering or a thorough understanding
of Alaska's environment the State will suffer much damage in the next decade.
There has been a notable lack of work in coastal zone management and engineering in
Alaska. Sound practice in this area will be necessary to prevent and/or abate pollution
of the air, land, and water for both the offshore and onshore coastal areas of Alaska.
This paper presents, examines, and evaluates environmental issues which should be considered
in coastal engineering efforts.
The practice of engineering is not hampered by Alaska's environment but must be modified
to reflect sound engineering judgment. Most engineering systems used in Alaska have been
"borrowed" from "lower 48" technology and modified to suit the northern environs. It is
these modifications, however, that one must critically evaluate and in many instances
failure of engineering systems or components has occurred because the engineer was not
thoroughly knowledgeable of Alaskan conditions. Alaska's environment is highly varied
and is sometimes erroneously described by the phrase "the fragile ecology of the Arctic".
Some of the parameters that must be confronted by coastal engineers include four climatic
zones (arctic, continental, transitional, and maritime) that have a wide temperature
range (_75° to +lOOo F) and other highly variable climatological conditions, continuous
and discontinuous permafrost to depths in excess of lOOO feet, severe earthquakes, tides
that approach 33 feet and negative tides nearing 6 feet, and sea ice (total coverage)
that extends as far south as the Alaska Peninsula.
Pollution control in arctic/sub-arctic Alaska is largely concentrated near industrial and
community complexes located near the coastline. Only a small portion of the population
is located in interior Alaska while numerous small communities and villages dot the coastline.
Additionally, most of Alaska's major industry is coastal oriented including lumber,
seafood and oil. Environmental engineering for these coastal complexes is not limited to
air, land and water pollution and should include evaluations of ecology, social and
211

eaonomia impaat. resourae depletion. and others. Conventional studies must address problems
of publia health. water supply and distribution. water quality. wastewater aolleation
and treatment. wastewater sludge handling. solid waste management. air pollution. and
others.
Reaent energy exploration and development in Alaska has areated both short and long range
proHems. The Trans Alaska OU PipeUne aan be used to show the need for long range
aoastal zone planning and engineering. Construation of the haul road to Prudhoe Bay LJiU
provide a new transportation mode and aaaess to the Brooks Range and subsequently a flurry
of resourae development. Additionally. plans are already underway LJith the "aaaess" road
to Nome whiah aould eventually allow development of massive aoal reserves loaated in the
Northwest aorner of Alaska. Long range engineering evaluation must be started for the
aoastal areas inaluding the gathering of a large amount of baseline data needed for engineering
design.
More reaent interest and attention has foaused on OCS (outer aontinental sheLf) development.
ocs amounts to some 550. 000 square miles near Alaska. Most of the OCS studies are
being direated at impaat identifiaation and evaluation for the aontinental sheLf itself.
Virtually no engineering evaluations are being aonduated involving potential onshore environmental
impaat. The long range impliaation of ignoring or negleating these areas may
be devastating. Conaern must be direated at projeats assoaiated LJith the ocs leases suah
as inland staging areas. airports. roads. housing. and aU environmental faailities and
faators.
ABSTRACT ONLY AVAILABLE
212 Tilsworth 2

0.5
VELOCITY, M/SEC
HEIGHT, M.
50
FIGURE 1. PLUME SHAPE, 22 M 3 /MIN OF FREE AIR
40
30
20
10
o
216
2 4
RADIUS, M.
6
Topham 4

2 4 6 RADIUS, M.
FIGURE 2. VELOCITY PROFILES, 22 M 3 /MIN OF FREE AIR. NUMBERS ON CURVES
INDICATE DEPTH IN METRES.
217 Topham 5

u
w
(/')
-......
M
:E
-3
0
..... --'
w
:E
=>
--'
0
>
60
50
40
30
20
10
27.6
22.0
20 40 60 PLUME HEIGHT, M.
FIGURE 3. ENTRAINED WATER VOLUME FLOWS. NUMBERS ON CURVES INDICATE AIR
FLOW, M3/MIN FREE AIR.
218
Topham 6

N
......
\0
5
10
15
Radial velocity, m/sec.
+0.2 o -0.2 +0.2 o -0.2 +0.4 +0.2 o +0.6 +0.4 +0.2 o
34 m
Depth, m
12 m
Radial Position, m.
Figure 4. Radial Velocity Profiles of Induced Currents, 26.6 m 3 /sec air at 60 m depth.
8 m

1
VELOCITY. M/SEC
HEIGHT. M.
20
15
10
5
o 2
FIGURE 5. PLUME SHAPE. 3.6 M 3 /MIN OF FREE AIR.
221
4
RADIUS. M.
Topham 9

1
VELOCITY, M/SEC
HEIGHT, M.
FIGURE 6. PLUME SHAPE, 29 M 3 /SEC OF FREE AIR.
20
15
10
5
o
222
2 4
RADIUS, M
Topham 10

Taylor G. I. 1955 The action of a surface current used as a breakwater. Proc. Royal
Soc. London. Series A, Vol. 231, pp 466-478.
Topham D. R. 1974 Hydrodynamic aspects of an oil well blowout under sea ice. Interim
Report, Beaufort Sea project, Dept. of the Environment, Victoria, B.C.
Turner J. S. Buoyancy effects in fluids. Cambridge Univ. Press, England, 1973, pp 195.
224 Topham 12

Introduction
POLAR METEOROLOGY - A REVIEW OF SOME RECENT RESEARCH
Gunter Weller
Geophysical Institute
University of Alaska
Fairbanks, Alaska
United States
EXTENDED ABSTRACT
The general large-scale circulation of the global atmosphere has its basic driving mechanism
in the equator-poleward temperature gradients in both hemispheres. It has become
increasingly obvious over the last few decades that to understand and predict the behavior
of the atmosphere at any point, it is essential to understand the behavior of the total
global fluid system. The Global Atmospheric Research FPoject (GARP) is an outcome of this
recognition. Studies of the heat sinks (the polar regions) are therefore just as important
as studies of the heat source (the equatorial regions) to understand the meteorology
of the planet. Perhaps the most variable feature of the polar heat sinks is the extent of
ice on the sea. Several studies during the past decades have shown that statistical relationships
exist between ice extent in the North Atlantic and certain regional features
of the atmospheric pressure field and that large variations of the ice boundary in the
North Atlantic are associated with fluctuations of the general circulation of the atmosphere.
More recently, it has been shown that temperature anomalies at the surface of the
subtropical Pacific Ocean are associated with anomalous conditions in the North Atlantic.
To understand and predict these complex interactive phenomena, the bahavior of the ice
itself must be sufficiently understood.
Polar Observational Systems
The network of existing stations making upper air observations, listed by CARP is reasonably
good for the northern hemisphere, except in the Arctic Basin where there are only
three manned drifting ice stations at present (two maintained by the USSR, one by the
USA). On the Antarctic continent the station network is very sparse. Satellite data
largely have to replace many of the observations normally made from manned stations. The
VHRR (Very High Resolution Radiometer) of the United States NOAA 2 and 3 satellites is
capable of taking data in both infrared and visible wavebands, to give nearly global
cloud cover once a day with a resolution on the order of l km. Both Soviet and U.S.
satellites have made passive radiometric observations in the microwave spectrum, in order
to demonstrate its use for temperature and humidity distribution. Satellite infrared
temperature soundings are the key to the fully three-dimensional global observations,
especially at high latitudes where the gradient wind approximation holds quite well.
Both NIMBUS-5 with the ITPR system and NOAA-2 with the VTPR system, have this capability
but several limitations exist. Conventional meteorological measurements in the polar
regions often must be made from automatic weather stations; some of these are discussed.
225

However. regionaL differences within the Arctic Basin and temporaL anomaLies are stiLl
poorly documented; the effect of the persistent summer stratus cLouds on the radiation is
aLso stiLl uncertain.
Atmospheric Chemistry and Aerosols
Arctic and Antarctic regions are important in the study of atmospheric chemistry because
of their remoteness from many sources of minor atmospheric constituents. This remoteness
has Led to the estabLishment of benchmark stations in these regions. Carbon dioxide. DDT.
and atmospheric concentrations and sources of trace metals in the poLar regions are discussed.
as are particLes in the polar troposphere and stratosphere.
ABSTRACT ONLY AVAILABLE
227 Weller 3

SECTION 6
ICE, GENERAL AND MECHANICAL
Assur, A.
Sea I ae Enginsel'ing
(USA Cold Regions Research and Engineering Laboratory, New Hampshire,
United States)
Burdick, J. L.
TensiLe creep-Rupture of PoLYCTystaLLins Iae
(University of Alaska, United States)
Hutter, K.
The Signifiaance of the Shear Rigidity and of the Poisson Ratio for Sea
Iae PZates
(Federal Institute of Technology, Zurich, Switzerland)
Ito, H. and F. MUller
Measurement of Sea Iae Porae by the Strain Rosette Method in the North
Water Area
(Swiss Federal Institute of Technology, Zurich, Switzerland and
McGill University, Quebec, Canada)
Johnson, P. R.
An EarLy DesaLination and Iae Struatures Projeat Using NaturaL Freezing
(CRREL, Fairbanks, Alaska, United States)
Kivisi1d, H. R.
Iae Meahanias
(FENCO Consultants, Ltd., Alberta, Canada)
Kohnen, H.
GLooioLogiooL Investigations for the Improvement of Iae-Going Ship Design
Carl'ied Out on the Sea Iae Near Pond InLet, N.W.T. (Northern Baffin
IsLand) In 8pl'ing, 1972
(University of MUnster, West Germany)
Kovacs, A., A. Gow and W. F. Dehn
IsLands of Grounded Sea Iae
(U.S. Army Cold Regions Research and Engineering Laboratory, New Hampshire,
and Maryland, United States)
229

MlUlttlinen, M.
On the FLe:curaL Stl'ength of B'I'ackish Wate'I' Iae by in situ Tests
(University of Oulu, Finland)
Nelson, R. D.
Inte'I'naL St'I'ess MeaSU'I'ements in Iae Sheets Using Embedded Load CeLLs
(University of Alaska, United States)
Nevel, D. E. and F. D. Haynes
Interp'I'etation of the TensiLe St'I'ength of Iae Unde'I' TriaxiaL St'I'esses
(u.S. Army Cold Regions Research and Engineering Laboratory, New Hampshire,
United States)
Pousi, T., M. Luukkala and E. Palosuo
A Na'I''I'OliJ Beam SOna'!' to Measu;roe the Submarine P1'OfUe of an Iae Ridge
(University of Helsinki, Finland)
Robe, R. Q.
Height to DTaft Ratios of Iaebe'I'gs
(U.S. Coast Guard Research and Development Center, Connecticut, United States)
Shapiro, L. H.
A PreLiminary Study of Ridging in Landfast Iae at Ba'I"I'OliJ. ALaska Using
Rada'I' Data
(University of Alaska, United States)
Shapiro, L. H., and E. R. Hoskins
The Use of FLatjaaks fo'I' the in situ Determination of the MeahaniaaL
Prope'I'ties of Sea Iae
(University of Alaska and South Dakota School of Mines & Technology,
United States)
Tabata, T. and T. Ishida
On the Movement of Paak Iae
(Hokkaido University, Sapporo, Japan)
Vaudrey, K. D. and M. G. Katona
An ELastia StruatU!'aL AnaLysis of FLoating Iae Sheets by the Finite
ELement Method
(u.S. Navy Civil Engineering Laboratory, California, United States)
Weaver, R. L., R. G. Barry and J. D. Jacobs
Fast Iae Studies in Weste'I'n Davis Strait
(University of Colorado, United States, and University of Windsor, Ontaria
Canada)
230

Not too many years ago onLy Russia had a capabiLity in ice modeLing. Now considerabLe
interest and capabiLities are deveLoping in the United States and Canada, FinLand, Germany,
Derunark, S!Jeden and Norway. A very promising start is aLso being made in Japan.
More than anything eLse recent deveLopments are dictated by the need for energy. The
arctic Basin has a staggering potentiaL for the discovery of oiL and gas resources. The
Prudhoe Bay area is onLy a smaLL beginning.
The oiL industry deveLoped its first experience with ice forces in the Cook InLet which
by now has a number of producing platforms under moving ice conditions and considerable
tides.
The ice conditions are not much worse in Bristol Bay but are becoming progressively more
difficult farther north. Nevertheless the development of offshore faciLities, terminaLs
and harbors is under active consideration. There are promising areas north of Prudhoe
Bay.
Activities requiring the application of methods in sea ice engineering are underway in
the Beaufort Sea, in particuLar in the Mackenzie Bay. A number of chaLLenging projects
invoLving sea ice are surfacing in the Canadian Archipelago. In many places we have to
worry about fragments of ice islands and that is not easy.
The presence of icebergs in the Baffin Bay, Davis Strait and Labrador Sea has not stopped
speculative thought and actions. The Newfoundland area with its hazards of icebergs has
been actively expLored.
One should not forget that developments motivated by the search for energy inevitably
wilL lead to further development of mineral resources. Harbors and shipping will demand
a better knowledge of sea ice engineering that is available now. No wonder that "outsiders"
to the Arctic as Germany and Japan are developing a very constructive attitude to
further the state of the art.
Mining development in Greenland already has to cope with sea ice difficulties, oil exploration
will follow.
The Norwegians are aggressively pursuing their potentials in Arctic waters. No wonder
that the first POAC conference was held in Norway.
It will be no surprise if gigantic energy resources will be discovered in the northern
offshore waters of the Soviet Union, especially off Siberia, but so far there is no need
for imminent development. However, Russian research activities on sea ice in Pacific
waters (Sakhalin) are of considerable professional interest. Sea ice difficulties are
aLso experienced in the Caspian Sea.
How can the scientist and research engineer support the forthcoming intense development?
What tools and what knowledge are available?
Scientific Aspects
The phase diagram of sea ice is crucial for the rational analysis of its properties
which are strongly determined by its brine content. Some refinements in the phase
diagram are now possible but a substantial experimental basis based upon modern methods
is still not available. Brine drainage problems including ionic selectivity are not sufficiently
resolved. The role of solid salts is still hypothetical.
Some headway has been made in the study of dielectric properties. Remote determination
of ice thicknesses may be within reach.
232 Assur 2

measured values differing at least by a factor of ten. It will take several years to
sieve the evidence and to obtain some consensus of opinion. However, agreement is developing
that the size of the object in relation to the ice thickness is of considerable
significance although the basic concepts and types of relationships are still controversial.
Friction, adfreezing and compliance with the object are of paramount importance.
So are little suspected factors such as the orientation of optical axis in ice crystals.
The importance of this was always maintained by scientists but was historically ignored
by engineers.
The use of inclined surfaces is important to cut down on ice forces. One of the favored
designs for offshore platforms and terminals are conical structures. We are well on the
way of developing a reliable design methodology. A possible controversy is the possible
adfreezing of random ice masses to the structures with a resulting multifold increase in
ice forces. Only a few ideas exist how to calculate such situations, how to avoid or how
to cope with such possibilities. Even the likelihood of such occurrences is not certain.
It seems that relatively thin mushy ice can adfreeze but thick ice will lead to a selfcleaning
process.
Artificial islands are well protected behind a collar of shore ridges. Expected forces
should be small. This and other requirements lead to the necessary development of the
mechanical principles for broken ice masses and pressure ridges. We are making headway
in this area but will not succeed unless we turn from classical to stochastic methods in
mechanics. A few approaches can be outlined.
There are several schools available for the calculation of thrust needed for icebreaking
ships. Many considerations are similar to the calculation of ice forces on structures.
The effect of friction and especially of side pressure on ships has been largely ignored
in the past, although powerful ships can be beset in surprisingly thin ice provided
that sUfficient horizontal pressure develops. Also there is no unanimity in the treatment
of various components in ice resistance which can lead to considerable misextrapolations
going from model tests to the full scale performance of ships.
Considerable, Interesting Work Being Conducted in This Area
Scoring and scouring by the keels of pressure ridges and the fragments of ice islands is
a potential problem in bottom technology. Advances in observational methods have been
made. There are interesting developments available how to cope with the potential hazards
of icebergs, including improved observational methods. The possible use of icebergs as a
water resource also created multinational interest.
All these developments, of course, require a deepening and intensification of basic research
in support of engineering needs. Interestingly these developments augmented by
needs in internal waterways have led to considerable interest for the establishment and
updating of ice engineering facilities in several countries.
ABSTRACT ONLY AVAILABLE
234
Assur 4

19·05
_t_
T
!
114. 30 38 . 10
1
T
19. 05
Y t
I
!--35.56-1
25 · 40
Figure 2 . Spec imen dimensions (mm)
238 Burdic k 4

For the specimens that sustained the load for over 100 minutes, the continuous variation
in geometry reduced the significance of the overall elongation measurements.
As no practical method of geometrically relating unit strain with acceptable precision or
attaching extensometers to the necked section of the ice was discovered, a photographic
approach was utilized. Markers were placed on the later samples. A 35 mm camera was
mounted on a rigid support and photographs of the specimens were taken at varying intervals
during the test. Relative displacement of the markers was scaled from enlargements of the
photographs.
Post Test Investigation
After a sample failed, a transverse slice was removed at the narrow point of the neck.
The two ends were then "glued" together with distilled, O·C water so that it could be sawed
longitudinally. A longitudinal, center line section was then removed. The transverse and
longitudinal sections that had been removed were mounted on glass plates and thin sections
were prepared.
The completed sections were photographed between crossed polaroids. The irregular area of
the minimum cross was measured with a polar planimeter.
RESULTS AND DISCUSSION
Because testing and recording techniques were being developed during the series of experiments,
the results are not as complete as might be desired. Elongations and deformations
of interest in the earlier tests were not recorded, or crudely measured, because the magnitude
of flow exhibited by some of the specimens was not anticipated and prOVision for recording
it was not included in the original planning. As creep tests are rather time consuming,
the data from all tests are included. Hopefully, deficiencies in tests or procedures
are noted.
The grain size distribution of the samples was not desirably uniform. It is felt that the
large grains, typically, at the bottom and outside of the specimen had some influence on
the results. It is now suspected that raising the temperature to allow flooding the specimens
promoted the formation of large crystals at the base and that the thermal inertia of
the mold allowed similar growth at the periphery of the samples. Similar large exterior
grains are evident in the research report of Haynes (1973). Further investigation of freezing
techniques would be of interest.
Creep-rupture data is plotted three ways: axial stress vs time; axial stress vs log time;
and maximum stress (axial plus bending stress) vs log time. As suspected, the plots using
log time exhibit greater linearity. The gap in failure time between 100 minutes and 2,000
minutes is of interest. It is suggested that early failure is caused by initial dislocations
concentrating at existing flaws and initiating fracture, but that if these early
stress concentrations are not high enough to cause failure, a mechanism of crystal realignment
will take place, reducing local stress and ultimately allowing the ice to carry
unit stresses two to three times as high (Table 1) as the maximum values reported by
Hawkes and Mellor (1972).
There is some relationship between initial stress and early failure. As shown in Table 2,
only one sample with an initial axial stress of over 1,400 MN/m 2 lasted over 100 minutes
and no specimens with an initial stress of less than 1,200 MN/m 2 failed early. Between
these limits, both failure modes occurred. The difference between the specimens in this
intermediate range was not discovered. Visual inspection did not reveal any significant
flaws in the early failure samples. Study of the thin sections, so far, has not led to
any conclusions as to the reason for early failure. Grain size may be significant, but
fine grained and coarse grained samples failed in ,both modes at similar stress levels.
241 Burdick 7

TABLE 2. Specimen data
Length
Sample Between Initial Final Initia 1 Time to
Number End Caps Area Area Stress Eccentricity Density Fail ure Remarks
(l1li1) (1lIl1 2 (Itt) (MN/m2) (1lIl1) (mg/m 3 ) (minutes)
456 .733 .281 .904 5 650 Many large bubbles
2 463 1.203 .572 .902 93
3 82.30 499 121 1.115 .070 .900 4 420
4 81. 99 503 115 1.184 .140 4 405 Strain Interrupted
5 504 same 1.392 .085 5.5
N
..".
w
6
7
83.06
84.20
499
499
same
same
1.264
1.408
.149
.088
.906
.893
57
12
8 83.95 497 161 1.266 .083 .893 2 907
9 82.80 502 72 1. 256 .107 .909 3 415
10 82.60 497 1.267 .048 .911 3 739 Strain Interrupted
11 81.92 494 same 1.450 .105 .900 21
12 82.17 502 159 1.427 .086 .910 2 238
13
Blank Specimen
14 Blank Specimen
15 82.55 498 same 1.563 .115 .906 0.9
16 82.30 497 same 1.567 .206 .905 0.6
17 82.15 501 same 1.482 .128 35
t:d 18 81.76 499 same 1.490 .133 .905 51
" 19 497 1.499 .117 3.3
'1 same
p.
..,.
n
?;"
""

N
.I'> """
.....
o
Table 2.
Sample
Number
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Continued
Length
Between
End Caps
(mm)
82.55
81. 99
83.19
83.19
83.01
82.91
81.66
82.85
82.83
82.88
82.25
82.70
Initial Final
Area Area
(mm 2 (mm 2 )
496 143
495
498 same
497 same
495 168
498 95
494 137
497
497
497 237
499 same
500
Initial Time to
Stress Eccentricity Density Fail ure Remarks
(MN/m2) (mm) (mg/m 3 ) (minutes)
1.353 .036 .908 3 134
1.353 .248 .906 39
1.348 .109 16
1.349 .165 32
1.205 .205 .908 4 520
.979 .152 .908 11 043
1.136 .281 .903 5 437
1.344 .081 .900 2 297
Blank Specimen
Failed in Throat
1.344 .165 .899 2 609
Blank Specimen
1.345 .152 .902 2 706
1.340 . 381 .905 90
1.336 .083 .902 2 466 Failed in Throat

a. Sample under load b. Samples after loading and
before loading
c. Longitudinal thin sections between polarized light
Figure 4.
sample #14
sample #24
245 Burdick 11

THE SIGNIFICANCE OF THE SHEAR RIGIDITY AND OF THE POISSON RATIO FOR SEA ICE PLATES
Kolumban Hutter
Laboratory of Hydraulics, Hydrology and Glaciology
Federal Institute of Technology, Zurich
Switzerland
ABSTRACT
A nono'lassioa'l p'late theo:ry is presented whioh inc'ludes shear deformation and acoounts
for the possibi'lity that Poisson's ratio may depend on brine oontent. The statio response
of such a pZate to strip Zike 'loade is presented and it is shown that miorostruotura'l
effeots (the dependence of Poisson's ratio on brine) and the inf'luenoe of
transverse shear are more pronounced the more the app Zied-'loade are ooncentrated. The
numerioa'l resuZts, based on sea ioe further indioate that the dependence of the Poisson
ratio on brine may in sea ioe safe'ly be disoarded but that the infZuence of the transverse
shear deformation may have to be taken into aooount. These oonc'lusions are a'lso
found for a dynamio prob'lem. The artio'le o'loses with a presentation of the numerioa'l
va'lues of the phenomeno'logioa'l ooeffioients neoessa:ry for this theo:ry.
247

1. INTRODUCTION
This paper is concerned with the significance of Fbisson's ratio in floating sea ice.
More precisely, sea ice is not a one-component material but rather a composite consisting
of pure ice and brine pockets, small liquid inclusions of spherical or cylindrical
shape. The size of these inclusions depends on the salinity, the phase diagram of saltwater
and through it also on the temperature. Under the assumption of elastic behavior
the material response will thus be a function of temperature. This implies that the
gross material behavior is not merely determined by the material itself but also by the
thermal state the latter is exposed to and therefore by climatic conditions.
If the behavior of sea ice is assumed to be elastic - and this will be the restriction
we are dealing with in this article - this elastic behavior necessarily depends on
salinity and through the phase diagram also on temperature. The behavior can be described
as follows: Consider a floating ice plate. If at a material point the temperature
is known, then through the phase diagram of sea water one may at that point determine
the proportion of ice to brine corresponding to that temperature. This then leads
to the determination of the elastic properties of the composite "pure ice-brine inclusions".
At the point in question the material may be considered to be composed of an
elastic matrix and liquid inclusions and if the latter are assumed to be spherical or
randomly oriented, it is clear that the composite material will be isotropic. Its properties
will depend on the properties of the matrix, those of the liquid and to first
order also on the porosity induced by the brine inclusions. For this isotropic material
a dependence of Young's modulus and Poisson's ratio on the above mentioned parameters
is therefore needed.
Usual calculations are based on the assumption that the elastic properties of the liqun
can be neglected and we follow this assumption. Much work has been done in the recent
years to attack the problem analytically, but the essential ideas go back to A.Einstein
(l906) who was looking at the viscosity of a Navier Stokes fluid carrying suspended
particles. In Einstein's situation the viscosity of the liquid was changed by the presence
of the particles (their number density). Our situation is entirely analogous.
In sea ice research the above mentioned situation has long been recognized. Much of
sea ice research is therefore connected with the determination of the material properties.
Its most important applications lie in the theory of plates and as far as elastic
plates are concerned one would think that all is known, once the plate constant is
determined, as a function of the surface temperature, say. Knowing the temperature
distribution across the thickness of the plate, it is then not hard to guess that the
plate constant for sea ice should come out as some "smearout" of Young's modulus and
Poisson's ratio over the thickness of the plate. That is exactly what one does and indeed
one uses a formula of the form
D (l)
This formula in which z denotes the distance from the neutral plane depends on both,
Young's modulus and the Poisson ratio.
On the other hand, all experimental techniques for the determination of elastic constants
as functions of brine inclusions are based on quasi-static ring tests, beam
tests and seismic methods, which either only give the Young modulus as a function of
brine inclusions or which assume from the outset that the Poisson ratio does not depend
on the volume porosity of brine. All in all, experimental evidence for its dependence
on brine content is not established yet but experts claim that there is a need for it.
248 Hutter 2

On the other hand, formula (1) shows that, as far as plate equations are concerned,
there is no way of detecting whether the value of D is determined by a variation of
E or V or both. Furthermore, most processes occuring in floating sea ice, either
natural or induced by man, have wave lengths long as compared to the thickness of the
plate so that a two dimensional approximation of the three dimensional theory of elasticity
is justified. On the basis of the classical theory which uses an expression like
formula (1) all possibilities to judge whether the dependence of the Fbisson ratio on
brine content is of any significance or not are eliminated. What is needed is a plate
theory, which goes beyond the classical one. It should contain a second phenomenological
constant and with two constants of the form (1) it is then possible to decide
whether a variation of Poisson's ratio with brine content is significant.
The above mentioned theory has been developed and it so happens that its structure is
such that shear deformations are naturally taken care of. This puts us into the position
not only to investigate the influence of the Fbisson ratio, but equally also to
account for shear deformations. In this connection it might interest the reader that
it is the latter rather than the former which we have found to be of moderate significance
in sea ice.
The experimentalist might wonder, why I put so much emphasis on plates despite the fact
that ring tests etc. do not have any direct connection with plates. MY point of view is
that plates are by far the most encountered structural elements in sea ice. Should it
therefore come out that certain features and subleties are too small to be of numerical
importance, it is no longer necessary to search for an explanation, no matter how sophisticated
ones experimental techniques may be and no matter how far the experimental
techniques may have digressed from the original situation encountered in nature. Experiments
are then no more than for scientific curiosity - a claim theoreticians usually
suffer under - and this is exactly the case for the dependence of Fbisson' s ratio on
brine content.
2. THE NONCIASSICAL PlATE THEORY
The usual derivation of plate theories goes along the following lines:
(i) The principle of d'Alembert is introduced and the equilibrium equations on a
plate element are formulated on the assumption that the stresses on the faces of
the plate are statically equivalent to membrane forces and bending and twisting
moments (assumption of a flat two dimensional Cosserat surface)
(11) The hypothesis of Bernoulli-Love is adopted. (Material directors perpendicular
to the middle surface of a plane remain perpendicular to that surface when deformation
occurs)
(iii) Material behavior is usually based upon Hook's law.
Of these three assumptions it is the second one, which destroys the possibility of a dependence
of Fbisson's ratio on brine content. Indeed, a closer look at the classical
derivation of plate equations shows that it is condition (ii) together with (iii) which
determines the plate constant as written down in formula (1).
There are ways of avoiding assumption (ii). They usually follow the lines of continuum
mechanics. There are several possibilities to derive such equations and the one I use
goes back to Mindlin. It is not the place here to derive these equations, but the ideas
may be sketched: Consider the equations of balance of linear momentum
* I use Cartesian tensor notation and the summation convention. Also,
f'i = Of/crx. i ·
249
Hutter 3

Lastly, I have taken the position that the results obtained in this article suggest to
stop searching for a dependence of Poisson's ratio on brine. This statement, of course,
is true only to within the provision, that salinities of sea water do not exceed 3.4%.
Moreover, it is probable that an experimental set-up for the determinations of certain
material constants may critically depend on the values of Poisson's ratio. In such a
case, of course, its dependence on brine is important, although ultimately it may be
neglected. I do not think that experimentalists ordinarily dream out such singular experiments,
but they should at least be aware of such a possibility.
ACKNOWLEDGMENT
This research was financially supported by the Federal Institute of Technology through
the Laboratory of Hydraulics, Hydrology and Glaciology.
Assur, A. 1958.
of Science,
REFERENCES
Composition of Sea Ice and its Tensile Strength.
p. 598·
National Academy
Einstein, A. 1906. Eine neue Bestimmung der Molekuldimensionen. Ann. Phys.
29:289-306.
Hutter, K. 1975. A General Theory for Floating Ice, Archives of Mechanics, 27:4.
Hutter, K. 1975. Floating Sea Ice Plates and the Significance of the Dependence of
the Poisson Ratio on Brine Content. Proc. Royal Soc., A343:85-103.
Hutter, K. 1974. On the Significance of Poisson's Ratio for Floating Sea Ice,
Mitteilung Nr. 11 der Versuchsanstalt fUr Wasserbau, Hydrologie und Glaziologie
an der ETH, Zurich.
Weeks, W.F. and Assur, A. 1966. The Mechanical Properties of Sea Ice. Proc. Conf.
Ice Pressures against Structures. Laval Univ., Quebeque, p. 25-78.
268 Hutter 22

MEASUREMENT OF SEA ICE FORCE BY THE STRAIN ROSETTE METHOD IN THE NORTH WATER AREA
Rajime Ito and Fritz MUller
Department of Geography
Swiss Federal Institute of Technology (ETH)
ZUrich, Switzerland
and
McGill University
Montreal, Canada
ABSTRACT
Pim Island (approx. l2 km by 6 km) in Kane Basin, between Ellesmere Island and Greenland,
and a hypothetiaal dam aaross the southern end of Kane Basin are treated as models of
large struatures sUffering sea iae forae. Field data for the study was aolleated as part
of the North Water projeat during spring 1974 and in 1975.
Strain rosettes of three different sizes were set in the sea iae 5 to l2 km north-east
of Pim Island: four of 50 m, two of 2 km and one of ZO km. The 50 m rosettes were measured
by steel tape whilst the others were measured by optiaal theodolite and an Eleatronia
Distanae Measurer (EDM). The strain at eaah point was aomputed and then the stress
aalaulated, assuming several aonstitutive laws: elastia, Newton's, and Glen's laws.
Foraes parallel and perpendiaular to the aoast of Pim Island were integrated along three
sides of an imaginary lO km square with one side resting on the aoast. The reaating
forae from the struature, Pim Island, was aomputed for equilibrium of foraes. The sea
iae forae on the hypothetiaal dam was aomputed in a similar way.
From the analysis it appew's that the 50 m rosette size is the most suitable one and that
Glen's law serves best for the aomputations required for this study. The greatest diffiaulties
arose from the unaertainties in the speaifiaations of the aonstitutive laws.
269

Lighting conditions for theodolite observations changed very rapidly with the advance
of the season at these high latitudes. The greatest distance at which the 30 cm by 40 cm
survey flag could be seen by theodolite in March was 3 km. Conditions improved rapidly
until May when the 4 m high and 2 m trianglular beacons were visible at a distance of
25 km. The sinking of the theodolite tripod was hardly noticeable, providing a single
series of measurements was made without delay. The estimated error was 10 seconds, i.e.
10 cm in 2 km or 50 cm in 10 km.
The Wild distomat proved highly suitable for the field work with respect to both low
temperature and salt. However the equipment is inconveniently bulky and heavy for transport
by skidoo and Nansen sledge. Meteorological data for the correction of the distance
measurements were obtained on site without difficulty except that dew point data measured
at Cape Herschel (about 30 km away) had to be used. Adverse weather conditions did
not hamper the distomat measurements. Distances of over 2 km were measured at several
occasions even when snow or fog reduced visibility to 200 m. The longest distance measured
was 11 km . The errors are estimated at 5 cm in 2 km or 13 cm in 10 km.
COMPUTATION
The sea ige is obviously neither under plane stress nor plane strain but in a threedimensional
stress-strain condition. Plane stress is, however, a rather better approximation
than plane strain if the three-dimensional treatment must be abandoned for the
sake of simplicity.
The assumption of plane stress is used throughout this study. It is emphasized that neither
the strain nor the stress deviator in the vertical direction is always zero and
that the principal stress deviators or strains on the horizontal plane are not necessarily
the maximum and minimum stress deviators or strains.
The one-dimensional engineering strain is the difference of the distance at two measurements
divided by the first distance. The mean strain rate is obtained by dividing by
the time interval between two measurements. The principal strains in the horizontal
plane are computed from the strains in three different directions. So far the strain
computation is general and the characteristics of the material are not considered. The
third principal strain is still unknown and cannot in fact be measured by current techniques
without disturbing the stress field in the sea ice. Therefore, no discussion about
the invariants is yet possible. The third principal strain and the invariants can only
be computed after a particular constitutive law is assumed.
Three constitutive laws were used to compute the stress from the strain (Table 1): the
elastic law, one of the instantaneous laws which is commonly applied to the study of
short period phenomena, e.g. seismic waves, and two time dependent laws, linear and nonlinear,
the former often applied in the study of pack ice and the latter for glacier
movement. The terms elastic law, Newton's law and Glen's law will be used in this paper.
Table 1 shows the constitutive equations for these laws and the constants applicable in
this study. The condition of continuity was assumed to apply to the force intensity
(stress multiplied by ice thickness) and only simpl€ fields, linear or quadratic, because
of the limited number of rosettes. The ice forces were obtained by integrating the force
intensity fields.
273
Ito and MUller 5

however, that the fast ice has no time independent characteristics but that the response
must be fast. This cannot be detected by the method described here. A different approach
would be necessary to obtain information on these short period characteristics (Goodman
et al. , 1975).
So far nO significant difference was found between the results obtained by Newton's law
and Glen's law. This is discussed later on in this paper.
The Models
Two models were designed (Fig. 2):
Model 1 - Pim niland was regarded as a large pier protruding into the sea from the land
(Ellesmere Island) ,ignoring Rice Strait. It was approximated by a rectangle of 16 km by
8 km. The major ice force was presumed to act on the eastern 10 km stretch of the
northern coast. The minor ice force on the other part of the coast was ignored as well
as the direct wind or sea current force against the island. General observation in the
area indicated that this simplification is reasonable. The structure was regarded as a
cantilever supported by Ellesmere Island. The purpose of the present study is to estimate
the force necessary to support this cantilever.
Model 2 - A dam across the entrance of Smith Sound, from Pim Island to Cairn Point, was
imagined. The dam is assumed to be floating and to offer no resistance to the subsurface
sea current. This hypothetical dam would, therefore, suffer only the ice force.
Two segments of the dam were assumed, each segment consisting of a simple beam of 5 km
length. The force necessary to support these beams was computed.
The Ice Forces
For Modell, the hypothetical pier, the area was divided into three sectors, each being
measured at three points. A linear field was assumed in each sector. The components of
force intensity, normal and parallel to the coast of the pier, were plotted. The results
for the period 6-14 April 1974 are shown in Fig. 5.
The field would have less accuracy in the vicinity of the coastline because of the irregular
local topography; the integration was, therefore, made along the other sides of
the imaginary square. The force acting on the coast was computed as the balance of the
forces, with those acting on the surface of the square being ignored. The results are
shown in Fig. 6. Special care was necessary when the ice boundary intersected the side
of the square.
As there were tide cracks along the coast it could be concluded that there was no tension
existing between the sea ice and the structure. Computations with Glen's law showed only
compression whilst Newton's law showed tension during the first period. This is the only
significant difference between the two laws. Glen's law appears more suitable for the
fast ice, although definite conclusions cannot be drawn because of the uncertainty of
the coefficients.
The following results were obtained with Glen's law. The normal force, pressure, increased
with the seasonal advance. This was obviously due to the northward movement of the
boundary (Fig. 7). The concentrated pressure was presumed to be at the base of the archshaped
fast ice boundary, which was to the right of the structure in the first period
and moved closer to the square in the second period. Finally it entered the square 'in
the last period, during which the actual length of the coastline in contact with the
sea ice was reduced to half. The force intensity, therefore, was twice as much as the
calculated value.
277 Ito and MUller 9

•
25 Km
I
\
I
\
I
21 JUNE
------- .... ,
" ...... - ... )
'" (11 JUNE ,/
POLYNYA
Figure 7 Change of the fast ice boundary location,
1974
280 Ito and MUller 12

AN EARLY DESALINATION AND ICE STRUCTURES PROJECT USING NATURAL FREEZING
Philip R. Johnson
Cold Regions Research and Engineering Laboratory
Research Engineer
Fairbanks, Alaska
United States
EXTENDED ABSTRACT
University of Alaska research engineers carried out an experimental ice building project
at Kotzebue, Alaska in early 1966. The project was based on earlier work by the Naval
Civil Engineering Laboratory and is in the line of development that may lead to fullscale
ice structures in the Arctic Ocean. It tested means of building ice structures.
It demonstrated the possibility of producing potable water from sea water in the coastal
Arctic by using natural freezing during the winter to freeze sea water and natural melting
and gravity drainage to freshen the ice in the spring. It also demonstrated the fact
that building ice structures from sea water is far from simple and the experience gained
will be of value in the future.
ABSTRACT ONLY AVAILABLE
285

ICE MECHANICS
lI. R. Kivisild
FENC

- ice sheets formed in situ
- icebergs and ice islands
- ice floes
- consolidated ridges.
In these cases, listed above, the ice masses act as one elasto-plastic body. An
integration of individual effects of solid mass contacts is required in an analysis
of the following ice formations:
unconsolidated ice masses
- pack ice
- unconsolidated ridges.
A field of ice floes can be analyzed by defining the interaction of various individual
elements. The failure process at the leading edge of an ice floe in collision can be
described by combined static and dynamic stresses at the edge. The interaction
between floes can be led back to tests on individual ice floes. The accumulation of
local effects can be combined to define the resultant action of an ice pack.
Even in a case of unconsolidated pack or ice ridge, the force mechanism depends
ultimately on the properties of solid ice.
Methods for the building of ice structures and quality control of natural or built-up
ice depends on the ability to observe and predict the mechanical properties of ice.
ENGINEERING PARAMETERS
Properties which are of interest from an engineering point of view use the mechanics of
consolidated ice as a basis.
As a first step there is a need to know the strength and rheological properties of
consolidated ice as listed on Figure 1:
- compressive and tensile strength
- genesis and crystal structure
- salinity, temperature, crystallography
- creep effects.
Compressive and tensile strengths have many direct engineering applications and
rheological properties are important.
Physical properties such as salinity, structure and temperature are useful in defining
the ranges of a set of mechanical properties. Creep effects will have a strong
interdependence between rheology and temperature which thus becomes prominent in the
analysis of the effects of long-term loads. Mode of formation and crystal structure
are of interest by demonstrating the range of applicability of measured properties.
SELECTION OF ICE WITH VALID PROPERTIES
As shown on Figure 2, ice properties vary substantially, but can be defined by using
various index properties:
- Salinity measurements will delineate major fields of application since salt
water ice is different from fresh water ice.
- Temperature measurements will define applicability of tests since cold ice
behaves as a brittle elastic material at rapid loadings while at warmer
temperatures, plastic effects dominate.
- A study of deformations can be important since past fracturing history can
have strong residual effects.
290
Kivisild 4

Thus mechanical properties depend on the structure of ice, temperature and restraints
and because of varying conditions, a set of strength and rheological data is valid
only for ice in similar regions with a similar structure and at the same temperature.
Although measurements of index properties would indicate the type of ice which must
be tested, changes of ambient factors during sampling and testing cause further scatter.
The means to reduce the number of unmanageable variables are indicated on Figure 2:
- Field tests would eliminate sampling and transportation effects as well as
change of conditions in a test tank or a laboratory.
- Tests have been devised which disturb ice as little as possible.
- In situ tests are indicated to eliminate sampling problems.
TESTS OF SIGNIFICANT PROPERTIES
Testing for design or to provide control of operations should be designed to reduce
extrapolations from test results to engineering parameters. The main objectives of the
particular application should be considered. A maximum ice strength would be required
to define the limit of ice thrust while the minimum strength would be required to
assure bearing capacity. Figure 2 indicates some basic lines of approach to reach
these objectives:
- field tests either in full size or on moderately reduced geometric scale
models will give required criteria for a particular design;
- geometric scale model tests in controlled environment in lakes or test tanks
are a good basis for engineering criteria for tested designs, define the
mode of ice action and indicate which ice properties are dominant in the
tested cases;
- testing of basic ice properties in laboratory conditions, carefully
establishing ambient conditions which would make tests applicable for design
through rigorous stress analysis;
- special tests which simulate ice action on parts of structure colliding with
ice and integrating the forces on various blocks of ice during the collision
process to obtain total ice thrust.
A clear definition of the main design parameters is available only for a few ice
engineering problems. A summary of various needed strength and strength-related
parameters for the assessment of ice forces and carrying capacity is given in Figure 1.
On the other hand, a great number of ice tests have been conducted. A comprehensive
review of past work on ice strength is given in Reference 2.
In order to avoid expensive and cumbersome large scale tests for each individual
design, the development of certain new tests was indicated. A system of these tests
are for ice thrust and are based on the testing of clearly distinct modes of failure
of blocks in collision. To devise the tests, a brittle elastic theory of ice failure
was derived to explain the action of ice in collisions.
Tests which have a wider range of applications such as tests which simulate the failure
of various blocks of ice in collision allow the application of results to a variety of
structures.
By measuring the properties of ice in the field and in situ, a much-needed link between
natural and artificial environments has been established. Since the station of
knowledge of ice mechanics is limited, these methods permit a reduction of the often
unmanageable variables.
The flaking strength test, described below, was devised to simulate the conditions in
ice thrust at the edges of a pile or a structure resisting ice although it was found
to give more than that. Stresses in various flaking failures are shown in Appendix 1.
292
Kivisild 6

The same type of test with modifications in the shape of flaking elements was devised
to determine ice thrust in the central portions of exposed faces in contact with ice.
For ice floes of limited size where lateral constraints would not apply, it was found
advisable to define changes caused by removal of lateral constraints, and tests with
this change were also conducted in the flaking test series. As described below, the
series has been reduced to largely one type of testing since it was found that results
in the alternative flaking conditions can be correlated and it was not warranted to
repeat the variations to define the properties of a different kind of ice.
A separate test series has been carried out as template on flat faces of ice and as
jacking tests in boreholes to determine confined compression strengths. Stresses under
strip loads are discussed in Appendix 2. Three dimensional confined compression
results are shown in Appendix 3. These tests have been applied to determine pressures
acting on structures deeper embedded in colliding ice masses. This would especially
apply for monopods and piles with relatively small aspect ratios.
Many collision processes depend on bending and tensile strengths. Field cantilever
tests, encastre beam tests over the full ice sheet thickness and small beam tests have
been used. For plate bending effects, it is necessary to consider the healing effect
of lateral pressure as occurs in nature and for small scale effects, solid sections
must be considered. As described below therefore vertical cantilevered beams were
supplemented to conduct these other tests.
In collision processes against very wide faces compared to ice thicknesses, a
simultaneous load would not be applied. The same question arises in collisions of ice
with very large structures where the total effect would be the time average force or a
rate of impulse rather than a peak thrust. In order to aid in the estimate of action
in these processes, energy to fail a unit volume of ice has been defined also by
measuring deflections and by calculating the work done up to the moment of ice failure
in the tests.
Borehole jack tests have also been found to be very practical to execute in all depths
of ice. This test can be used to determine the combined strength of ice in deeper
layers and these tests have been used also as index tests to define properties of ice
in general. Other properties can be found through the application of relationships
between the borehole jack tests and other tests as defined from other test series on
similar ice.
FLAKING STRENGTH
These tests are done in pits cut in the ice surface as shown in Figure 3. The
hypotenuse of a right triangular plate is placed parallel to the pit surface against
one wall and a hydraulic ram is used to apply pressure until failure. This particular
shape of plate was originally chosen since ice is observed to flake off in approximately
this shape when pushed against by a structure such as a ship or fixed monopod or pile
as shown in Figures 4 and 5. The major and minor principal stresses in such a case are
the applied ram pressure and a respectively (Figure 6). In addition to the biaxial
stress effects during the test, additional restraints arise from the three-dimensional
factors during flaking. Deviations from uniaxial compression condition are discussed
in Appendix 1, where ice is assumed to be a brittle elastic isotropic material.
Because of heterotropic effects and the existence of definite fluid zones, actual
flaking conditions approach even more the state of unconfined compression than shown
in this theoretical analysis.
Special tests have been conducted in an attempt to eliminate lateral constraints to
simulate collision by floes. One method has been to cut slots with the chain saw on
either side of the plate as shown in Figure 7a. Another test, tried during this study.
293 Kivisild 7

FAilS AS
IN CONFINED
COMPRESSION ---+-)
FRONT ELEVATION
D (STRUCTURE WIDTH)
EDGE flAKING
UNCONFINED
FIGURE 5. FLAKING FAILURE OF ICE SHEET IMPINGING
ON A WIDE STRUCTURE
SI DE ELEVATION

FIGURE 6. PRINCIPAL STRESSES DURING FLAKING TEST
297
Kivisild 1

(a) FLAKING TEST WITH SLOTS
PLAN
ELEVATION <
\'-_---.., ,---_J
y
SLOTS
TOP OF
ICE
.... --ADDITIONAL TRENCH TO ALLOW
SIDEWAYS FLAKING
(b) FLAKING TEST ROTATED 90° CLOCKWISE & SPECIAL PtT
FIGURE 7. FLAKING TESTS DESIGNED TO RELIEVE EFFECT OF
LATERAL FORCES
2913
Kivisild 12

was done with the plate rotated through 90 0 after a special type of pit had been cut
(Figure Tb). Strengths obtained from these tests did not differ from values obtained
from the regular tests.
Ice flaking strengths obtained from various locations in the arctic and sub-arctic have
been plotted in Figure 8 against the temperature of the ice at the time of testing.
Tests conducted by Hawkes and Mellor (3) on dumbbell-shaped specimens on ice with
random C axis orientation provide the best information available in the literature on
the uniaxial compressive strength of ice as obtained from laboratory small-scale tests.
Their specimens were carefully machined to eliminate end effects such as splitting and
all failures were by crushing of the ice in the central region of the specimen. Time
to failure was found to affect the strength in their tests so tests with the same range
of times to failure (average of 30 seconds) as encountered with the flaking tests are
reported. Their tests were conducted at -7°C and the range of strengths corresponding
to the times to failure are shown in Figure 6. The line of best fit shown is for the
flaking strength values. We see that it intersects the line defining the range of
Hawkes and Mellor tests at the lower part of the middle third. It should be emphasized
that each of the flaking test points represents the average of a minimum of 6 tests and
some represent an average of over 50 tests. The flaking tests are conducted at the ice
surface and the ice in this region usually has a random orientation of the C axis
(Reference 4). Thus there is a lot of relevant data to form the basis of our comparison
and we can only conclude that the flaking test can be used as a measure of unconfined
compressive strength in addition to the direct definition of ice stresses at particular
points in collisions with a' structure.
This is a very significant conclusion since it provides a low cost and easily performed
in situ measure of compressive strengths of ice.
If we examine the extreme left portion of the data, especially that along the vertical
axis (Temperature = OoC), we see that there is considerable scatter of data. Also,
the line of best fit indicates that the effect of ice temperature on compressive
strength is not large. This leads us to the conclusion, one which is substantiated by
a curve presented by Butkovich (Reference 1), that the strength of ice is moderately
affected by temperature down to the melting point and that once it reaches the melting
point, its strength can be anything between nearly the full ice strength of colder
temperatures and zero strength depending on the amount of heat of fusion which the ice
has lost. Thus it is not so important to know ice temperatures to estimate strength
but rather to know the degree of deterioration of the ice, especially in spring at
breakup time.
CONFINED COMPRESSIVE STRENGTH
Since the flaking tests can only be conveniently done near the ice surface, a different
in situ test must be conducted at greater depths. Borehole compression tests are
usually the quickest and most reliable type which can be performed in many naturally
occurring materials and this is what we have commonly used for ice.
For the purposes of the test, a 15 cm diameter hole is drilled in the ice. A borehole
compression testing jack developed by FENCO is then lowered into the hole and tests are
conducted on the side of the hole at required depth intervals. A diagram of the jack
is presented in Figure 9. Pressure is applied to the front and back plates
hydraulically by means of a piston located inside the jack body and which is activated
by a pump at the surface. Oil is transferred from the pump to the piston by means of
hydraulic high pressure hose supplied with quiCk-disconnect couplings. Displacement
is measured using a linear, variable resistor mounted on the jack body with the
plunger attached to the front plate. Resistance is related to displacement by a
calibration curve and is measured at the surface by means of a multi-meter or ohms meter.
299 Kivisild 13

After the jack is lowered into the hole and the test begun, the stress situation is as
shown in Figure 10. The major principal stress, aI' is the applied plate pressure.
Since the pressure is being applied to what is essentially the surface of a semiinfinite
solid, the stresses are three-dimensional. At the beginning of the test, a
bulb of crushed ice forms beneath the load plates because of surface unevenness and
high local stress concentrations. This bulb is confined by stresses ° 1 , o? and 03 as
shown in Figure 10. The bulb of crushed ice continues to grow as the app11ed pressure
is increased until at some point it pushes out around the load plate and failure is
reached. In many cases the crushed ice never escapes from the bulb but the size of the
bulb diameter grows larger and larger and the displacement increments increase
exponentially. Some of the ice near the load plate recrystallizes from the high
pressure. Large displacements are associated with this consolidation and plotting
pressure displacement curves enables one to determine the "ultimate" or confined
compressive strength of the ice since the curve approaches this value asymptotically.
These tests are normally done at one-foot intervals through the ice sheet and can be
correlated readily to other strength tests done in pits close to the surface.
Elastic modulus values for the ice are determined from the pressure-deflection curves
of the borehole tests. The analysis gives much useful information regarding the
relative brittleness or ductility of the ice throughout its depth.
TENSILE AND BENDING STRENGTH
Bending strength is a problem since tests, mostly one-dimensional, are greatly affected
by cracks. Crack effects in nature are not as pronounced because of plate effects and
because of partial or full healing of most cracks. Therefore, vertical bending tests
which avoid cutting across cracks have been devised.
LARGE SCALE TESTS
Although it is not always possible to carry out measurements which engage portions of
ice with a size of the same order of magnitude as involved in prototype action, tests
should aim for an involvement of as large portion of ice as possible.
Size effects have been considered and it is attempted in the tests to engage portions
of ice containing large numbers of grains. This is of course more easily done near the
surface than farther down in natural ice.
Various large scale tests have been conducted which reproduce the action of ice against
exposed members of a structure. A comprehensive series of nutcracker tests has
simulated the thrust of ice against columns and piles in ice. Such tests offer means
of checking the validity of the smaller scale tests.
CONCLUSION
Correlations are now available from tests of full scale structures resisting ice forces
and from platforms built of ice and carrying drill rigs. It is possible to state with
confidence that design criteria of ice effects can be predicted on a rational basis,
using results of in situ techniques of ice testing and relating them to ice mechanics
theories verified by observations on large scale test structures.
302 Kivisild 16

15c ...
CRUSHED ICE .. DIAMETER HOLE
A A
t t
SECTION A-A
• APPLIED STRESS
TOP OF ICE
PRINCIPAL STRESSES FOR CONFINED COMPRESSION TEST
FIGURE 10
303
Kivisild 17

10
9
8
7
4
3
2
o
-
6O/1k =8(0=51)
I
6 1 (45)
j
4
1 (37)
2 (19)
I
1 (0)
15 20 25 30 35 40 45 50 55
1
60 65 70
FIGURE 1.3 BRITTLE ELASTIC THEORY - FORCES
306
Kivisild 20

3
o
P v =0.4
o 0
o 10 20 30 40 50 60 70 80 90
0----.
VARIATION OF p/fc AND p WITH (2)
FIGURE NO. 3.1
312
300
200
100
[+]
Kivisild 26

4> p/fc
v = 0.3 v = 0.4
1 0 lo5 lo6
2
4
6
8
19
37
45
51
lo9
2.3
2.6
2.7
2.1
2.9
3.4
3.9
The value of 0c/Ot is usually in the range of 4 to 8 and v for ice is about 0.3 or a
little higher. A theoretical value of 2.5 is thus obtained for the ratio of p confined
to o. Tests carried out by FENCO to determine the confined compressive strength of ice
indi8ate a higher ratio with an average of test series normally in the range of 3.0 to
3.1 indicating that ice is not fully brittle, but has some plasticity. In the
interpretation of results a ratio of 3 is therefore used.
313 Kivisild 27

GLACIOLOGICAL INVESTIGATIONS FOR THE IMPROVEMENT OF ICE-GOING
SHIP DESIGN CARRIED OUT ON THE SEA ICE NEAR POND INLET, N.W.T.
(NORTHERN BAFFIN ISLAND) IN SPRING, 1972
Heinz Kohnen
Institut fUr Geophysik
University of MUnster
D-44 MUnster
West Germany
ABSTRACT
Transportation in aratia waters has been a steady growing probLem sinae the expLoration
of the riah mineraL resouraes in the North. Sea traveL, using big iae-going buLk and oiL
aarriers, is expeated to be a suitabLe and eaonomiaaL teahnique. For the aonstruation
of suah big vesseLs various gLaaioLogiaaL parameters of the sea ioe and the dependenae
on the environmentaL aonditions have to be k.nown. Among these, of major importanae are:
strength, eLastiaity, fPiation of metaL against iae, pressure, stress, iae movement and
their reLation to temperature, saLinity, wind fieLd and oaean aurrent; eLeatPiaaL and
eLeatromagnetiaaL properties of the iae for the navigation by the aid of remote sensing
teahniques.
The German shipyard A. G. Weser in Bremen initiated in L972 an expedition to Northern
Baffin IsLand to study aU reLevant parameters at one Loaation and the same time to aomprehend
the whoLe gLaoioLogiaaL regime. The expedition was a aooperative effort between
Broak University, St. CathaPines (Canada), and the University of MUnster, Germany.
TWenty-five saientists and teahniaians aarPied out gLaaioLogiaaL, meteoroLogiaaL and
oaeanographia investigations duPing May and June L972 on the sea iae between Pond InLet
and ByLot IsLand, N.W.T.
This artiaLe reviews the soientifia program and the researah aativities and summarizes
the important resuLts of the expedition. Furthermore, the author tPies to find the re­
Lations between the individuaL resuLts and to give a synthesis.
315

INTRODUCTION
From the earliest voyages to polar seas in small sailing vessels to more recent trips
over arctic seas in aircraft, travelers have reported discovering new islands which
have defied the attempts of all subsequent travelers to relocate. Mysterious disappearing
northern islands of the past include the famous Gunnbjorn Skerries found by a
Norwegian in 877 while sailing between Iceland and Greenland (Zukriegel, 1935; Nansen,
1911). While never seen again, Gunnbjorn's discovery remained on the maps and in the
minds of men for centuries. Perhaps the most famous island discovered was sighted
off the southern coast of east Greenland by Frobisher in 1578 (Best, 1578) and referred
to as the Isle of Buss, the Land of Buss and, finally, when it could not be
relocated, the Sunken Land of Buss. This land too remained on the maps for centuries
and as late as 1818 Edward Parry is known to have made a special search for it (Parry,
1821). Then there were the lands and islands known as Arctic Land, hypothesized by
Harris (1904) to be west of Prince Patrick Island; Keenan Land, searched for by
Mikkelson (1909) in the Beaufort Sea; Presidents Land, reported in 1876 to be in the
Lincoln Sea; Bradley Land, sighted by Cook (1911) at 85°N, north of Axel Heiberg
Island; Sannikov Land; Crocker Land, sighted by Peary (1907) northwest of Ellesmere
Island; the sandy island located in the Beaufort Sea at 76°N 150 0 W by Gracianski in
1937 while looking for the lost crew of a Russian polar flight (Brower, 1960); and
Takpuk's Island, found in the early 1930's at about 71 o N, 145°W in the Beaufort Sea.
This last island was reported by Takpuk to have small ponds, grass and moss on it and
a scattering of boulders up to the size of a man's head (Stefansson, 1934).
Several of these "islands" are now believed to have been large tabular icebergs, the
so-called ice islands, which have been seen in considerable numbers in the Arctic
Ocean during the last thirty years (Kovacs and Mellor, 1971 and 1974). These are
undoubtedly what Gracianski and Takpuk saw and reported as islands. New "islands"
have now been found and are discussed in this paper.
THE DISCOVERY
The first indication of an anomalous condition existing in the sea ice approximately
160 km northwest of Ft. Barrow, Alaska, occurred during a routine inspection of NOAA-l
satellite imagery in the spring of 1971. The site of the anomaly was indicated by
the existence of a large polynya which persisted in one general location. This area
of open water appeared as a dark spot in a field of gray sea ice on the NOAA imagery
(Fig. 1, position 1). Soon after this discovery, contact was made with the Naval
Arctic Research Laboratory (NARL) at Pt. Barrow and the director was informed of the
discovery. It was requested that a mission be flown to the site to determine if the
polynya was associated with the grounding of a large ice island or was due to an
extensive accumulation of grounded sea ice on a shoal shown to be 22 meters below sea
level at 71 0 54'N, 161 0 8'w on the 6th edition of the U.S.C. and G.S. Arctic Coast of
Alaska Map No. 9400. However, this mission was not flown.
Information of the find was also given to the U.S. Coast Guard and a request was made
for the icebreaker Burton Island, which was scheduled for summer operation in the
Chukchi Sea in 1972, to investigate ice conditions in the area of the polynya, which
continued to appear on NOAA satellite imagery through the winter and spring of 1972.
In August 1972, the Burton Island visited the area. A large ice formation was observed
with highly irregular relief that appeared to rise as much as 9 m above the
sea. Although the description given was of a large hummock field of pressured sea
ice, news quickly spread that the formation was a large piece of shelf ice, i.e., an
ice island, 8 by 20 km in size (Untersteiner, 1972a and b).
334 Kovacs/Gow/Dehn 2

Figure 1 . NOAA- l satellite image of 13 May 1971. Orbit No .
1922, shoving pol ynya associated vith the grounded
sea ice feature (1) and polynyas associated vith
grounded ice on Herald Shoal (2) and on a shoal
vest of Wrangel Island (3).
Photographs taken by Captain Robert G. Moore during the Burton Island visit did not
become available to us until the spring of 1975 (courtesy of Dr. A. Grantz). Three
of the tventy- one photos received are shovn as Figures 2, 3 , and 4 . Figure 2 shovs
approximately three quarters of the grounded island of ice as it appeared at a distance
from the air . The contrast betveen the sunlit pressure ridges incorporated in the
island and the rarefied summer pack is quite apparent . The extreme southvestern tip
of the island , here designated Dehn Point, is shovn in Figure 3 to consist of belts
of shear ridges , hummocked ice and floes of multi- year ice . A close- up of the summer
melt features on the island is shovn in Figure 4 .
None of the Coast Guard photos of the island shoved any fragments of shelf ice, i.e. ,
ice islands . In September 1972, W. Dehn flev over the site to assess and photograph
the surface structure of the grounded ice . Several photographs taken on this trip
revealed the existence of a cluster of tventy- seven ice island fragments vith abnormally
high freeboards (Fig. 5). This latter situation vas presumed to be the
result of uplift associated vith grounding . This discovery led to the speculation
that an ice island had drifted onto the shoal and vas then pushed higher up on the
shoal by the moving pack until the ice island had become firmly grounded. Stresses
associated vith grounding are believed to have resulted in fragmentation of the
island . It vas assumed that the grounded ice island fragments vere the anchor point
against vhich the moving pack failed, eventually to form the large area of grounded
ice shovn in Figure 2. Detection of the island of grounded ice on satellite imagery
335 Kovacs/Gov/Dehn 3

Figure 2. Oblique viev of grounded sea ice feature shoving approximately 75%
of its linear dimension . Dehn Point is at left end (southvestern
tip) of island.
Figure 3. Aerial viev shovi ng structural characteristics of ice incorporated
into the grounded sea ice feature at Dehn Point .
336 Kovacs/Gov/Dehn 4

Figure 4. Aerial view of summer melt features near the
edge of the island of grounded sea i ce.
Figure 5. Cluster of ice island fragments incorporated
into grounded sea ice feature.
Photogr aph was taken in September 1972.
was therefore made possible by the large polynya which forms on t he side of the
island o,posite the direction in which the pack ice is moving .
AGE OF THE ISLAND
To determine how long the island of grounded ice had been in existence , hundreds of
NOAA , ESSA and Nimbus satellite images of the area were inspected. The earliest
imagery available to us was that of 1966 from the Nimbus satellite . Although of very
low resolution , this imagery revealed the polynya associated with the island (14 June
1966 Nimbus orbit 408 , frame 30) . For the years 1967 through 1972 the polynya may be
seen on the following selected imagery : 16 May 67 ESSA- 3 orbit 2844 ; 7 May 68 ESSA- 3
orbit 7328 ; 11 May 69 ESSA- 9 orbit 932; 5 May 70 ESSA- 9 orbit 5416 ; 13 May 71 NOAA- l
orbit 1922 (see Fig. 1) ; and 3 May 72 NOAA- l orbit 4522 . Satellite imagery thus
establishes the fact that the island is not new but has indeed been in .existence for
many years .
337 Kovacs/Gow/Dehn 5

Figure 9 . View of landing site and surrounding features on
grounded island of sea ice .
formation was somewhat larger than Big Diomede Island which is approximately 9 . 3 km
in length and clearly visible in the NOAA- 3 image at position 2 . The 1975 ERTS
(Landsat) imagery also reveals that the smaller ice formation had not re- formed .
This finding indicates that islands of grounded ice are not necessarily permanent
features and may even be seasonal formations .
Neither of the two islands of grounded ice was located near the shoal shown at 71 0 54 ' N,
161 0 8 ' w on the 6th edition of U.S . C. and G.S. Arctic Coast of Alaska Map No. 9400.
The most recent edition of this map , published in 1973, shows not one but two shoals
near the area of the grounded ice formations , one 18 meters below the ocean surface
at 71 0 50 ' N, 161 0 10'W and the other approximately 22 meters below the surface at
71 0 50 ' N, 161 0 7 ' W. Kovacs et al. (1975) found that "these shoals are approximately 35
km east- southeast of the site of the two grounded ice format i ons found in the 1973
ERTS- l imagery" and speculated "that the two grounded ice formations are resting upon
these shoals." They reasoned that since corrected "ERTS - l imagery can be used to
position surface features to within 300 meters of their true geographic positions,
the location of the shoals in the bathymetric chart may be in error . "
ISLAND VISIT
A field trip was made to the island of grounded ice to obtain bathymetric measurements
around it, to investigate its surface morphology and compositional structure ,
and to ascertain if it could be safely used for a manned research base. On 13 April
1975 , a flight was made to the island . An extensive search was made for a place to
land, but without success. On 15 April another flight was made to the island and a
small area of first- year ice was found on its central western edge which was large
enough to land on. The landing area and the formidable hummock field surrounding it
are shown in Figure 9.
The search for a landing site revealed an ice surface of highly varying relief.
Incorporated in the island were small areas of first- year ice which had formed in
the numerous summer melt pools (see Fig. 4), many multi- year floes which either
devel oped on site or had become a part of the island during the yearly winter accretion
process, and endless belts of shear ridges and fields of brecciated ice
averaging 5 to 6 meters high with many peaks on the order of 15 meters high (Fig.
10) . A number of large floebergs were seen (Fig . 11) , but nothing that could be
identified as a fragment of shelf ice. This feature would thus appear to be a true
island of grounded sea ice.
340
Kovacs/Gow/Dehn 8

Figure 10. View of hummock field in grounded sea ice with
ridge peaks up to 15 m high .
Figure 11. View showing floebergs incorporated into the
grounded sea ice island.
Accompanying us to the island in a helicopter was Mr . Jack Lentfer of the U.S. Fish
and Wildlife Service. Mr . Lentfer, internationally known for his research on polar
bears , was at the time engaged in his annual spring polar bear tagging program. His
interest in the island in general , and in the polar bears which might be found there,
allowed us to prevail upon him to assist in the field deployment of our camp and to
make a low level inspection by helicopter.
The field camp was established on a thick multi- year floe incorporated into the
western side of Dehn Point. This site was selected because it provided access to a
long refrozen lead which meandered northwest toward the interior of the island. On
this lead we planned to walk toward the interior, taking water depth and ridge height
measurements en route. This route was denied to us the following morning when at
approximately 9 o ' clock the ice to the west began to pressure. Within two hours the
ice in the lead was transformed into a hummock field of first- year and multi- year
ice. The field began within 15 meters of our camp and extended as far as one could
see to the west from the top of a 7- meter- high pressure ridge. Ice conditions around
the island of grounded ice on 15 and 16 April 1975 are shown in Figures 12 and 13.
341 Kovacs/Gow/Dehn 9

Figure 12. Defense Meteorological Satellite Program (DMSP)
image of 15 April 1975 showing ice conditions
around the island of grounded sea ice (1) and the
existence of two islands of grounded ice on Herald
Shoal (2).
There was a large polynya on the southeast side of the island and a tongue of pressured
ice nearly 50 km long extended outward from the northwest side . It was the ice
movement and the resulting pressuring associated with the formation of this tongue
that caused the crushing of the lead next to our camp . In the Landsat image of 29
April (Fig. 14) this ice tongue is no longer apparent.
Our convenient route into the interior of the island now gone, we attempted to
move directly over the ice surface . This effort was soon terminated when it became
apparent that both we and the equipment would suffer considerably and there was not
sufficient time to make a long trek over the snow- slickened rubble extending toward
the interior. The relief which was before us is shown in Figure 15.
Our stay lasted two days. This short visit revealed that the surface relief was a
chaotic jumble of pressured sea ice covered by a surprisingly deep layer of snow .
Walking over this terrain was exceedingly difficult and treacherous when snow bridges
between the ice blocks gave way under foot .
Bathymetric measurements made at six sites around the periphery of the island varied
from 28 to 31 meters , averaging 30 meters. During the spring of 1976 we plan a
helicopter trip to the island for the purpose of making additional bathymetric
measurements throughout its interior. These measurements will allow us to determine
the depth of the shoal upon which the island of sea ice is grounded .
OTHER ISLANDS OF GROUNDED ICE
Besides the two islands of grounded sea ice discussed, we have found similar formations
on Herald Shoal (70 0 30 ' N, 171 0 30 ' W) , and on a shoal off the northwest coast of
Wrangel Island. Unlike the island of grounded sea ice northwest of Pt. Barrow , which
remains year round , these other islands do not survive the summer and in some locations
342 Kovacs/Gow/Dehn 10

Figure 13. NOAA- 4 satell ite image No . 1906 V2F1492 of 16
April 1975 showing ice conditions around the
island of grounded sea ice at 72°N 162°W (1),
two grounded ice features on Herald Shoal (2),
and a grounded ice feature west of Wrangel Island
(3) .
Figure 14. LANDSAT satellite image (ID No. 2097- 22081)
showing i ce condition around the island of
grounded sea ice on 29 April 1975 .
343 Kovacs /Gow/Dehn 11

Figure 15. Viev shoving characteristic surface rel i ef (snov- covered) in
vicinity of camp site located near edge of grounded sea ice
feature. The peak at the left center is 11 meters high .
islands do not form each year . An example of the latter is the grounding of ice on
Herald Shoal. The 1973- 74 NOAA-3 imagery did not reveal any grounded ice formation
on this shoal . Hovever , the lov resolution spring imagery from the ESSA- 3 satellite
in 1968, the ESSA- 9 satellite in 1969 , and the NOAA- l satellite in 1971 (see Fig . 1 ,
position 2) shovs a polynya associated vith an island of grounded ice on this shoal .
In the vinter of 1974- 75 t vo islands of grounded ice formed on this shoal . Both are
easily detected in the higher resolution NOAA- 4 and DMSP satellite imagery shovn in
Figures 12 and 13 at position 2 . Another area vhere an island of grounded sea ice
has been observed to form each year is approximately 50 km off the northvest coast
of Wrangel Island (see position 3 in Fig . 1 and 13) .
DISCUSSION
Satellite imagery has revealed the existence of islands of grounded ice on shoals in
the Arctic Ocean . One has been found to be a "permanent " feature , at another location
an islands re- forms each vinter, vhile at a thi rd location s uch formations do
not appear to occur each year. Recent calculations of pressure ridge keel depth
distribution in the Arctic Ocean indicate that several hundred ice keels 20 meter s or
more in depth can be expected to pass a given location during the course of a year
(Weeks et al. , 1971) . Therefore , the potential impact frequency of pressured ridge
keels and the occasion for them to become firmly grounded can be great in the shoal
area of the grounded ice formations. In addition , in recent years several hundred
ice island fragments have been found both adrift and gr ounded off the Beaufort Sea
coast of Alaska and Canada (Kovacs , 1972 , Kovacs and Mellor , 1974) . These fragments
typically have keels 15 to 30 meter s in depth. While their numbers are fev in relation
to the numerous ridge keels that can be expected to drift over a given location
in a year , the grounded ice formation at 72°N , 162°W has been shovn (Fig . 5) to have
had fragments of grounded shelf ice incorpor ated into it.
Ridge grounding on shoals has been found to create a barrier against vhich the floes
of the pack are pushed . Unable to r esist the stresses developed , the floes eventually
344 Kovacs/Gow/Dehn 12

moving pack while the ship was in winter quarters. A similar definition for floeberg
is given in the WMO Sea-Ice Nomenclature (Anon, 1970) under the heading "Forms of
Floating Ice." The two photos given in the WMO publication clearly show that a
floeberg is not a large area of hummocked ice but only a drifting fragment thereof.
Islands of grounded sea iae seems the appropriate term to describe the new island
discoveries reported in this paper. This term is technically correct because the
formations are islands in the sea which in this instance are composed of grounded sea
ice, though they may, on occasion, consist in part of shelf ice. The term grounded iae
island was not used because this would tend to imply that such features were pieces
of grounded shelf ice.
ACKNOWLEDGMENTS
This work was supported by the Office of Naval Research under MIPR No. N0001475MP50013.
The logistic support of the Naval Arctic Research Laboratory, Barrow, and the helicopter
support provided by the U.S. Fish and Wildlife Service, through Mr. Jack
Lentfer, are gratefully appreciated.
REFERENCES
Anon. 1970. WMO sea-ice nomenclature. World Meteorological Organizationreport
WMO/OMM/BMO-No. 259. TP. 145, Geneva, Switzerland.
Armstrong, Sir A. 1858. A personal narrative of the disaovery of the north-west
passage. Hurst and Blackett, London.
Best, G. 1578. True disaourse. London.
Brower, C.D. 1960. Fifty years beZow zero. Dodd, Mead and Company, N.J.
Cook, F.A. 1911. My attainment of the Pole, Mitchel Kennerly, New York.
Harris, R.A. 1904. Indications of land in the vicinity of the North Pole. Nat. Geog.
Mag. 15.
Kovacs, A. 1972. Ice scoring marks the floor of the Arctic Shelf. The Oil and Gas
Journal 70 (43).
Kovacs, A. and M. Mellor 1971. Investigation of the ice islands in Babbage Bight.
Creare Inc. Technical Note N-118, Hanover, N.H.
Kovacs, A. and M. Mellor 1974. Sea ice morphology and ice as a geological agent in
the southern Beaufort Sea. In The Coast and Shelf of the Beaufort Sea. Proceedings
of the Arctic Institute of North America Symposium on Beaufort Sea
Coast and Shelf Research.
Kovacs, A., H. L. McKim and C.L. Merry 1975. Islands of grounded ice. Arctic 28(3).
McClure, R. 1857. Disaovery of the northwest passage. Sherard Osborn, Editor,
London.
Mikkelson, E. 1909. Conquering the aratia iae. London and Philadelphia.
Nansen, F. 1911. In northern mists. William Heinemann, London, Vol. 1.
Nares, Sir G.S. 1878. Narrative of a voyage to the poZar sea during l875-76 in H.M.
ships Alert and Discovery. Second ed., New York.
347 Kovacs/Gow/Dehn 15

Parry, W.F. 1821. Journal of a voyage for the discovery of the northwest passage.
London.
Peary, R.E. 1907. Nearest the pole, a narrative of the expedition of the Peary
Arctic Club in the S.S. Roosevelt, 1905-l906. London.
Schindler, J.F. 1968. The impact of ice islands-the story of Arlis II and Fletcher
Ice Island, T-3, since 1962. In J. E. Sater, ed. Arctic Drifting Stations.
The Arctic Institute of North America.
Smith, C.L. 1971. A comparison of Soviet and American drifting ice stations.
The Polar Record 15(99).
Stefansson, V. 1921. The friendly Arctic. Macmillan, New York.
Stefansson, V. 1934. An Eskimo discovery of an island north of Alaska. Geographic
Review, January.
Stockton, C.H. 1890. Arctic cruise of the USS Thelis. Nat. Geog. Mag. 2.
Stringer, W.J., and Barrett, S.A. 1975. Ice motion in the vicinity of a grounded
floeberg. In Abstracts of papers presented at <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on
Port and Ocean Engineering Under Arctic Conditions, University of Alaska, August
11-15, 1975.
Sverdrup, o. 1904. New land. London, Vol. 1.
Underwood, D. 1974. This week at NARL. NARL Newsletter, 30 September to 6 October.
Untersteiner, N. 1972a. Letter of 14 August to LCDR W.F. Dehn, Fleet Weather Facility,
Washington, D.C.
Untersteiner, N. 1972b. Letter of 14 August to Dr. A.F. Treshnikov, Arctic and
Antarctic Research Institute, Leningrad, USSR.
Weeks, W.F., A. Kovacs and W.D. Hibler, III 1971. Pressure ridge characteristics in
the Arctic coastal environment. Proceedings of the First <strong>International</strong> <strong>Conference</strong>
on Port and Ocean Engineering Under Arctic Conditions, Vol. 1. The
Technical University of Norway, Trondheim, Norway.
Zukriegel, J. 1935. Cryologia maris (Study of sea ice). Geographic Institute
Charles IV, Praha.
348 Kovacs/Gow/Dehn 16

REFERENCES
1. E. Enkvist, 1971: Kalkkiranta -71, Jaanlujuuskokeet
Wartsila/Merenkulkuhallitus.
2. M. Maattanen, 1972: Murtovesijaan lujuustutkimus, Tvarminne 1972,
Oulun Yliopisto/Merenkulkuhallitus.
3. M. Maattanen, 1973: Murtovesijaan lujuustutkimus, Perameri-Oulun
edusta, Oulun Yliopisto/Merenkulkuhallitus.
4. v.v. Lavrov, 1969: Deformation and strength of sea ice, Leningrad.
5. T. Tabata, K. Fujino, A. Aota, 1966: The flexural strength of sea
ice in-situ, Proceedings I.L.T.S., Hokkaido University.
6. W. Week and A. Assur, 1963: The mechanical properties of sea ice,
CRREL.
1 Types of test beams
2 Test arrangement by hand lever
3 Lffect of loading rate
APPENDIX
355 MMMttMnen 7

INTERNAL STRESS MEASUREMENTS IN ICE SHEETS USING EMBEDDED LOAD CELLS
Richard D. Nelson
Geophysical Institute
University of Alaska
Fairbanks, Alaska
United States
ABSTRACT
'The interaction of an eLastic incLusion and a stressed visco-eLastic host is disauBsed
as a means for stress anaLysis in sea ice sheets. A simpLe technique is presented which
aLLows numericaL caLauLation of both the eLastic and the steady state visco-eLastic responses
of the incLusion. The behavior of a 4-node finite eLement modeL of host/transducer
interaction is used to estabLish criteria for transducer design.
361

Elastic bodies emhedded La elastic or viscoelastic continua have heen used in recent
years as a means for stress analysis in concrete (Rocha, 1965) frozen soil (Hawkes, 1969)
and ice (Nelson, 1974). The stress (or strain) induced in the elastic body is measured
by conventional means and used as an indicator of the stress in the host itself. The
theory of elasticity readily yields a calibration constant which may be used to relate
the transducer response to the stress in the surrounding continuum, provided the host
is elastic, the properties of both the host and transducer are known, and the geometry
is simple. If the host's elastic properties are uncertain but are known to lie within
a certain range, care must be taken to choose the elastic properties of the inclusion
so as to minimize error due to this uncertainty; usually this means the inclusion should
be much stiffer than the host. If the host is viscoelastic, then the theory of viscoelasticity
may be used to analyze the interaction and a time varying calibration factor
results. As with the case of uncertain elastic properties, the approach taken for
viscoelastic materials has been to minimize the range of variation in the calibration
constant by using a stiff transducer (Hawkes, 1969, Nelson et aI, 1972).
In previous work, Nelson et al (1972) used a one node, one dimensional spring and dash
pot model to elucidate the principals involved in the host/inclusion interaction when
an elastic body is embedded in a viscoelastic continuum. The model was used to establish
design criteria for transducers which were to be embedded in ice for the purpose of
stress analysis. This simple model did not attempt to analyze shear transfer from the
ice in the region near the transducer head nor the changes in local stress concentration
near the head as creep progressed. The purpose of the following work is to provide a
basis for numerical analysis of the host/inclusion interaction for viscoelastic hosts
and complex inclusion geometries which will allow optimization of "stress transducer"
systems. In addition, a four node element model is used to examine the roles of shear
transfer, elastic moduli, and viscous moduli in establishing the stress in such
inclusions.
The host material will be modeled as a network of linear springs and dash pots whose
force/deflection equations are:
F Ko for a spring
F C .L'"
dt
for a dash pot
The transducer will be modelpd as One or more springs surrounded by the continuum.
order to estimate the fidelity of the transducer's response to stress variations in
host, we will examine the case where a sudden change in stress occurs. That is, we
In
the
wish
to determine the response of the transducer/host pair to a step input. This would
require determination of the transfer function relating the transducer output (strain or
deflection of an elastic body) to the stress applied at the far boundaries of the
continuum, a task which can be formidable for even a modest network.
If the transfer function were available, three questions would be important:
(I) What is the initial response, at t = O?
(2) What is the final response at t = 00 ?
(3) What is the rise time associated with the approach to the final response?
Fortunately, the first two questions can be answered without using the full differential
expression of the transfer function. This is true because at t = 0, the dash-pots have
not had time to deflect and can be considered rigid, leaving only the elastic response
of the springs. Standard structural analysis programs can then be used to evaluate the
forces and deflections. When the continuum is modeled with spring and dash pots in
series, the spring deflections will not be changing at t = 00 since a steady state has
362 Nelson 2

The tensile strength of ice in a triaxial state of stress was investigated by Haynes
[1]. Although multiaxial stress states exist in many strength studies of ice, the
primary motivation for his investigation was to compare the results to the tensile
strength determined by previous Brazil tests. Mellor and Hawkes [2] and Butkovich
[3] found that Brazil tests produced tensile strengths much lower than uniaxial
tensile tests.
In the Brazil test the disc of ice fails in tension along the loaded diameter. The
maximum tensile stress occurs at the center where, according to elastic theory, the
ratio of compression to tension is 3 to 1. This ratio increases as the distance from
the center increases. Haynes [1] conducted tests with ratios up to 10 to 1.
The ice tested by Haynes was isotropic polycrystalline ice, similar to the porous ice
found in glaciers. A cylindrical dumbbell shape with a diameter of 2.54 cm was
chosen for the test specimens. The general method of preparing the ice specimens was
to pour snow grains into a vibrating mold, saturate them with water from one end, and
freeze them from one end. Thin sections of the specimens were examined under polarized
light and showed that the grains were randomly oriented with a size of about 0.7 mID.
The ice was bubbly with an average bubble size of about 0.2 mID.
During a test the cylindrical surface of the specimen was pressurized by hydraulic
fluid while the ends remained unpressurized. The hydraulic pressure produced axial
tension and equal radial and hoop compression in the necked-down part of the specimen.
The cross-sectional area of the end cap and the necked-down section determined the
ratio of compression to tension. This produced a constant compression to tension
ratio as the hydraulic pressure was increased. A thin rubber membrane was placed
over each specimen to prevent the hydraulic fluid from penetrating into the ice.
Additional information regarding the test equipment, test procedures, stress calculations,
and eccentricity of loading may be found in the report by Haynes [1].
The test results are shown in Figure 1. This graph shows the tensile strength of ice
as a function of the compression to tension stress ratio. The one test run at the
ratio of 10.14 to 1 is not shown in Figure 1. The tensile strength for this test was
1.71 kg/cm 2 . The uniaxial tension results in Figure 1 are the minimum, average, and
maximum values from 63 tests performed by Hawkes and Mellor [4] on similar ice. The
average strength found at a ratio of 3 to 1 is about one-third the average uniaxial
tensile strength. This indicates that the compressive stress present in the Brazil
test produces the low values obtained for the tensile strength.
In the previous paper by Haynes [1], the triaxial data for snow-ice were compared to
the brittle fracture theory of Griffith [5] and Babel [6,7]. Griffith considered the
two-dimensional state of stress around an elliptical hole. The ellipse was shrunk to
a flat crack. The failure theory was determined by equating the internal stress for
the uniaxial tension state to the internal stress of a biaxial stress state. Babel
did the same thing except he considered the range of shapes going from a circle,
through the ellipses, to a flat crack. Strictly speaking, the three-dimensional data
should not be compared to a two-dimensional theory. Since the general three-dimensional
theory is complicated, we will review and extend the two-dimensional theory in
order to gain an insight into the nature of the problem. Then we will compare the
test data to a special case of the three-dimensional theory.
Griffith [5] considered the two-dimensional state of stress around an elliptical hole
as shown in Figure 2. His solution was based on the work of Inglis [8]. The stress
R which is tangent to the elliptical surface was given by
376 Nevel and Haynes 2

y -2
crT
-4
-6
4 6
FIGURE 5. Biaxial fracture angles.
383 Nevel and Haynes 9

FIGURE 6. Principal stress space.
Loading Plane
02=0;
385 Nevel and Haynes 11

20
40
30
(0) Minimum
(6) Average
(D) MaXimum
FIGURE 7. Triaxial fracture results.
386 Nevel and Haynes 12

REFERENCES
1. F.D. Haynes, Tensile strength of ice under triaxial stresses, USACRREL RR 312,
Dec. 1973.
2. M. Mellor and I. Hawkes, Measurement of tensile strength by diametral compression
of discs and annuli, Engineering Geology, Vol. 5, p. 173-225, 1971.
3. T.R. Butkovich, Some physical properties of ice from the TUTO tunnel and ramp,
Thule, Greenland, USACRREL RR 47, 1959.
4. I. Hawkes and M. Mellor, Deformation and fracture of ice under uniaxial stress,
Journal of Glaciology, Vol. 11, No. 61, p. 103-131, 1972.
5. A.A. Griffith, The theory of rupture, Proceedings of the First <strong>International</strong>
Congress for Applied Mechanics, Delft, Netherlands, p. 55-63, 1924.
6. H.W. Babel, Biaxial-fracture strength of brittle materials, Air Force Materials
Laboratory, TR 66-51, Wright-Patterson Air Force Base, OhiO, March 1966.
7. H.W. Babel and G. Sines, A biaxial fracture criterion for porous brittle materials,
Journal of Basic Engineering, Transactions of the American Society of Mechanical
Engineers, Vol. 90, Series D, pp. 285-291, June 1968.
8. C.E. Inglis, Stresses in plate due to the presence of cracks and sharp corners,
Transactions of the Institute of Naval Architects, Vol. IV, p. 219-230,
London, 1913.
9. B. Paul, Macroscopic criteria for plastic flow and brittle fracture, p. 313-496,
In Fracture Vol. II, edited by H. Liebowitz, Academic Press, 1968.
10. S.A.F. Murrel and P.J. Digby, The theory of brittle fracture initiation under
triaxial stress conditions, Geophysical Journal of the Royal Astonqmical Society
Vol. 19, pt I, p. 309-334, pt II, p. 499-512, 1970.
11. H.M. Westergaard, Theory of elasticity and plasticity, Harvard University Press,
1952.
12. R. Frederking, Preliminary results of plane strain compression tests on columnargrained
ice, <strong>International</strong> Association of Hydraulic Research ice symposium,
Leningrad, 1972.
387 Nevel and Haynes 13

A NARROW BEAM SONAR TO MEASURE THE SUBMARINE
PROFILE OF AN ICE RIDGE
T. Pous i and M. Luukkala
University of Hel s inki
De partment of Physi cs
Helsinki, Finland
and
E. Pa los uo
University o f Helsinki
Department of Geophysi cs
Helsinki, Finland
ABSTRACT
A narrow beam verticaLLy scanning sonar has been buiLt to measure the submarine profiLe of
an ice ridge. The sonar consists of two main parts : the immersed portabLe scanning sonar
head and the centraL unit containing aLso the recorder and 220 V Honda generator. The centraL
unit is usuaLLy situated in a tent or in a heLicopter.
The principLe of use is as foLLows : a hoLe is driLLed in the ice and the scanning sonar
head is immersed at a suitabLe depth. The sonar transmits 0. 1 ms sound puLses five times
per second with 10 W power. underice features such as ridges, scatter sound back to the
transducer. The backscattering echoes are received and the echo distance and the corresponding
scan angLe are shown by two digitaL parreL meters and recorded by a xy- recorder.
The beam width of the sonar is 2. 4 0 and the distance resoLution is about ±15 cm . During
t he fieLd experiments the sonar performed according to expectations, and the profiLes of
severaL ridges were recorded. The distance range of the sonar was about 30 m but an expansion
is pLanned.
389

INTRODUCTION
This project was started when Professor E. Palosuo (Department of Geophysics, University
of Helsinki), in his research work concerning ice ridges, noticed the usefulness of a high
frequency narrow beam scanning sonar. Because such equipment was not available with a
graphic display, the sonar was to be built in the Department of Physics, University of
Helsinki.
As the project began the following demands were stated:
1. The sonar must have a graphic display from which the submarine profile of an ice
ridge can be seen or computed.
2. The 3 dB beamwidth of the sonar has to be less than 3 degrees.
3. The range of the sonar should be about 50 m.
4. It should be possible to operate the scanner head at a distance of about 100 m
from the central unit which was planned to be situated in a tent, helicopter or
on an ice breaker.
The final construction of the sonar fulfills all these demands except the distance range
which was about 30 m. Our purpose, however, is to expand the range by increasing the gain
of the receiver. The details of the sonar and some theoretical considerations are presented
in this paper. In actual measurements in the Gulf of Bothnia in March 1975 the sonar
performed well. The results of the measurements are also discussed and some measured profiles
are shown.
The electronic central unit of the sonar and the scanner head are shown in Figure 1.
Figure 1. Sonar electrical unit and scanner head.
390
Pousi et aZ . 2

W

Figure 7. Submarine unit
The larger profile has been "viewed" from three different dep t hs . These profiles are not
quite alike, but clearly t he parts pointed by arrows correspond with each other. One reason
for differences may be the change in the horizontal direction when changing the depth.
On the other hand it is obvious that the sound may reflect in different ways when the depth
of the transducers is changed . To get more precise information about these variations,
many systematic measurements have to be made and the results have to be checked, for example
by diving .
In Figure 14 the underside of an even ice field is shown . The reflection has been received
continuously till the distance of 7 m.
CHANGES AND IMPROVEMENTS
The most important improvement will be to increase the amplification of the receiver.
The gain of the receiver is planned to be raised up to 120 desibels and it may also be
necessary to use time variable gain. If the gain variation is made such that it compensates
for the signal level decrease as the distance increases it is also possible to use
an intensity modulated graphic recorder to study the back-scattering strength.
The display system will be changed so that we shall get the side
ectly on the recorder instead of the scan-angle/distance-output.
shall have to multiplicate the distance with the sine and cosine
398
profile of the ridge dir­
To accomplish this we
of the scan- angle to get
Pousi et aZ .lO

Figure 8. Construction of a submarine unit
ANTUHlN AKSEU
399 Pousi et at. 11

"'" o
.....
'"
2M
7M
HOLE 1
t
Figure 9 •
2.4M
10M
!
20M
!
30M
Typica 1 submarine pro file measu red with sonar
40M
!
50M 60M
210CM
..

HEIGHT TO DRAFT RATIOS OF ICEBERGS
R. Q. Robe
U. S. Coast Guard Research and Development Center
Groton, Connecticut
United States
ABSTRACT
A study of height to draft ratios of iaebergs near the Davis Strait reveats ratios whiah
raTIIJe from 1:1.28 to 1:10.56. The ratios of bergs dominated by their horiaontat dimension.
suah as tabutar or broken tabutar iaebergs. have average height to draft ratios of 1:4.46
and 1:4.26 respeativety. Bergs with a more vertiaat nature. pinnaate or drydoak bergs.
have ratios averagiTIIJ 1:2.31 and 1:2.41 respeativety. The smattest ratios are found in
domed bergs whioh average 1:6.30.
If we assume that the height to draft ratio of iaebergs is aharaateriaed by a aontinuous
distribution. then usiTIIJ a Kruskat-Wattis one-way anatysis of varianae teahnique we aan
test the hypothesis that the average ratio of iaebergs is not signifiaanHy different for
gross visuat shape atasses. The resutt is that for the sampted ioebergs there is no signifiaant
differenae. For summary purposes then the average of the averages (1:3.95) aan
be used as desariptive of the height to draft ratio of iaebergs regard tess of visuat shape
atass.
In this study iaebergs with the greatest height have the targest height to draft ratios.
That is to say the draft for taU ioebergs is proportionaUy tess than for tow bergs.
The reasons for this are aonjeatured to be as fottows:
a. The tattest bergs generatty have spires and pinnaates whiah add great height
with minimum mass white the 'Lowest bergs tend to be worn and smooth haviTIIJ
maximum mass for minimum height.
b. The towest bergs are worn and have onty the most dense ioe remainiTIIJ. aU unaonsoUdated
iae and snow haviTIIJ been washed away. and most voids haviTIIJ disappeared
aausiTIIJ them to fWat tower in the water.
Between the berg heights of 10 meters and 60 meters. whiah is the range of this sampte.
the height is retated to the height to draft ratio by the power aurve.
l/Ratio = 49.4 (Height)-·8
407

INTRODUCTION
The draft of icebergs is of interest for a variety of reasons. In areas where
pipelines or cables lie on the bottom, information on draft can be used to estimate
the probability of a break. For the <strong>International</strong> Ice Patrol the draft is of interest
because of the effect it may have on drift, groundings and deterioration. Approximately
seven eighths of the mass of an iceberg is submerged; however, this is not an
indication that the height to draft ratio is necessarily 1:7.
Estimates of height to draft ratios were made as far back as the late 19th century.
Steenstrup (1893) gives the ratio as 1:7.4 and 1:8.2; while KrUmmel (1907) gives a
ratio between the extremes of 1:8 and 1:4, with most falling in the range of 1:5
and 1:6. Grounded icebergs were used to obtain the earliest estimates of the ratio.
Dawson (1907) found a berg stranded in the Strait of Belle Isle in 1894 which had a
ratio 1:3. Again in the Strait of Belle Isle, Rodman (1890) found a 30-meter pinnacle
berg grounded in 29 meters of water for a ratio of nearly 1:1. To estimate draft,
Smith (1925) used a drag wire strung between two heavy weights and towed at known
depths by two small boats. The small boats, separated by about 137 meters, would
pass on opposite sides of the iceberg and lower the weights till the wire passed
freely under the iceberg. He found a ratio of 1:2. During the 1959 ice patrol,
Budinger (1960) examined the underside of an iceberg by diving under it. He found
that the berg had a height to depth ratio of 1:3.3. Budinger also observed another
berg 55 meters high grounded in 175 meters of water off Cape Race (ratio of 1:3.2).
Budinger erroneously states that the height to depth ratio cannot be smaller than
1:6. This was in conflict with earlier estimates by Steenstrup (1890) and Krummel
(1907) and also was not substantiated by data from the present study. Data collected
by the submarine USS SEA DRAGON, which studied nine bergs, found height to draft
ratios which ranged from 1:1.3 to 1:4.2 (Murray, 1960).
The height to draft ratio was highly dependent on the shape of the berg. The berg had
to float so that seven eighths of its mass was submerged and so that the berg was
stable. If, for instance, the iceberg was tabular (flat top and bottom and vertical
side) a ratio of 1:7 would be expected. If the above-water portion was rounded and
smooth, while the underwater part was pointed, then a ratio smaller than 1:7 could be
expected, even as small as 1:9 or 1:10. The other extreme was the case where the
underside of the berg was rounded and smooth and the above-water portion had towering
vertical walls. The most pronounced case of this type was the drydock berg, where an
embayment of great height and little mass. These could have a height to draft ratio
which approached 1:1.
The purpose of this study was to see if the above water shape of icebergs was
related in a significant way to the height to draft ratios for those bergs. Heightto-draft
ratios were obtained for a total of 30 icebergs.
METHODS
Measurements of iceberg draft were taken with a Kelvin-Hughes Transit Sonar during
a cruise aboard the CGC EDISTO, July 1974. The EDISTO was operating in the Davis
Straits area and along the west coast of Greenland. The Kelvin-Hughes Transit Sonar
was designed to conduct bottom surveys; however, we were interested in vertical
targets rather than in horizontal ones. The sonar transducer produced a fan-shaped
beam 1.5° wide in the horizontal and 52° wide in the vertical, both being to the
3db level. For our purposes the transducer was pointed down by 26°, so that the top
of the fan-shaped beam would just pass under the surface of the water and the bottom
of the beam would be depressed at 52°. The transit sonar was designed for use from
a small boat with only a few feet of freeboard. It was first mounted on the EDISTO's
408 Robe 2

arctic survey boat (ASB). This arrangement worked well, providing cover for the deck
gear and personnel, along with high maneuverability and good speed control. The first
five bergs were surveyed from the ASB with great success. Use of the ASB was then
discontinued because the single point bridle used to raise and lower it was hazardous
in any but the calmest weather. For the next two bergs the motor surf boat (MSB) was
used. It was inadequate because the equipment was exposed to the weather and because
the boat had such little stability that it was difficult to maintain the transducer
orientation with respect to the iceberg. The MSB was retired due to a failure of the
boat davit.
Finally, a method for using the transducer from the EDISTO itself was devised. The
freeboard of the EDISTO was approximately eighteen feet from the rail to the water
line aft of midship. A 21 foot pipe was fabricated that would support the transducer
three feet below the water line. The transducer was mounted on the bottom of the pipe,
and the pipe was manhandled from the deck to the outboard position for each run. Small
chunks of ice were a constant problem and once sheared the transducer off the supporting
pipe. A safety line attached to the transducer prevented loss of equipment.
With the SOnar on the EDISTO it was possible to have the deck gear in the oceanographic
laboratory and also to operate from a very stable platform.
When the ship was positioned near enough to the berg (Figure la), the beam of the
sonar was completely intercepted by the iceberg. As the ship circled the berg, it
increased the distance from the berg so that at some point part of the sonar beam
passed under the berg. The distance was increased till the ship was at maximum
range (550 meters slant distance from the bottom of the berg) or a good echo was no
longer received.
Five assumptions were made in interpreting the record, a sample of which is given in
Figure 2. First, that the first echo was returned from the near surface portion of
the berg; second, that the strong echos were reflected from vertical surfaces on the
underwater portion of the berg; third, that weak returns came from walls which slope
away from the observer along a radial of the sonar beam; fourth, that blank areas in
the return were the results of shadow areas caused by caves, holes or ridges in the
iceberg; and fifth, that if the transducer was far enough away from the berg the
last return from the berg comes from the deepest portion of the berg.
The entire record of the iceberg sonar trace was examined and points which were
representative of the deepest point on the berg were chosen. These points were
plotted on a radial grid so that the radial distances to the various portions of the
berg could be converted to vertical measurements of berg draft. These estimates of
draft were plotted versus distance to the berg. As the distance to the berg,
increased, the draft estimates approached an asymptote which was assumed to represent
the true draft of the iceberg.
DISCUSSION
The subaerial shapes of icebergs are extremely varied sometimes displaying fantastic
forms. Some bergs have "windows" in high vertical walls, while others are pockmarked
like a piece of Swiss cheese, and still others have huge grottos or voids. As a
means of organizing the shapes of the visible portion of icebergs into some system
certain prominate characteristics have been chosen and used for typing icebergs into
classes. These classes are based solely on visual identification.
This study examines whether or not the visual classification of icebergs is a meaningful
way to classify the height to draft ratios of these bergs.
409 Robe 3

(a)
(b)
Figure 1. (a) The beaM from the side looking sonar is completely intercepted
by the iceberg at very close range; (b) At a greater range a portion
of the sonar beam will pass under the iceberg and not return to the
transducer.
4W
Robe 4

!2
til
-t
• Z
n
-I: '"
-t '"
:D '"
!!
Figure 2. A sample of the side looking sonar record showing: (a) the
return from the bottom, (b) the shadow of the iceberg on the bottom glving
an approximate shape, (c) the return from the iceberg, (d) the return fror!
waves, (e) the zero line on the chart.
Based loosely on Murray (1968), the icebergs of this study were separated into five
general catagories based on gross visual shape characteristics.
1. Tabular bergs were horizontal, flat-topped bergs.
2. Broken tabular bergs were those that had a horizontal orientation, but whose
surface was highly fractured.
3. Pinnacled bergs had a large central spire or a pyramid of one or more spires
dominating the shape.
4. Drydocked bergs had an eroded U-shaped slot cut by wave action surrounded by high
vertical walls or pinnacles.
5. Domed bergs had a smooth, rounded top which had Once been either submerged or
highly \leathered.
The mean height to draft ratio for each of the five visual classes was computed and
compared statistically to the mean ratio for all other classes. The null hypothesis
411 Robe 5

is that there is no significant difference between the height to draft ratios for the
visual classes of icebergs.
The height to draft ratios for the icebergs studied ranged from 1:1.28 to 1:10.56
(Tables 1 through 5). The 1:1.28 value was in line with previous measurements, but the
1:10.56 value was smaller than any of the previously reported ratios. The 1:10.56
ratio was associated with a domed berg where the rounded above-the-water portion had
the maximum mass in the minimum height. To attain this value the underwater portion
probably had a taproot-like formation.
The tabular and broken tabular (Tables 1 and 2) had almost identical characteristics.
These were the most massive of the bergs, having lengths which were observed to reach
600 meters and masses in excess of nine million metric tons. The mean heights for
the tabular and broken tabular were both 28 meters. The mean drafts being 108 and 107
meters respectively. Of course, the height to draft ratios were quite similar also,
being 1:4.46 for the tabular and 1:4.26 for the broken tabular. The range of height
to draft ratios was 1:2.00 to 1:9.58 for the tabular bergs and 1:2.93 to 1:7.23 for the
broken tabular bergs.
Table l. Height to draft ratios for tabular bergs
Height Depth Ratio
(meters) (meters) (1:
35 122 3.48
40 80 2.00
30 137 4.57
21 97 4.62
32 84 2.62
12 115 9.58
28 121 4.32
Mean 28 108 4.46
Range 12-40 80-137 2.00-9.58
Table 2. Height to draft ratios for broken tabular bergs
Height Depth Ratio
(meters) (meters) (1: )
41 139 3.39
18 60 3.33
13 94 7.23
30 111 3.70
55 161 2.93
21 88 4.19
20 126 6.30
21 78 3.71
30 107 3.57
Mean 28 107 4.26
Range 13-55 60-161 2.93-7.23
Pinnacle bergs (Table 3) and drydock bergs (Table 4) also appear to have had average
height to draft ratios which were quite similar. Both berg types were generally
vertical in aspect and had the largest height to draft ratios of any of the visual
groupings. The range of the ratios was also more limited than for the other visual
groups. Perhaps the range was more limited because the physical characteristics of
these groups were less ambiguous than was the case for the tabular or domed bergs.
412 Robe 6

Table 3. Height to draft ratios for pinnacle bergs
Height Depth Ratio
(meters) (meters) (1:
16 37 2.31
59 111 1.88
32 84 2.62
34 83 2.44
Mean 35 79 2.31
Range 16-59 37-111 1.88-2.62
Table 4. Height to draft ratios for drydocked bergs
Height Depth Ratio
(meters) (meters) (1:
53 68 1.28
44 103 2.34
30 108 3.60
Mean 42 93 2.41
Range 30-53 68-108 1. 28-3. 60
Domed bergs (weathered, smoothed, deteriorated bergs) were the most deceptive
(Table 5). A few penetrated the water's depth as the pinnacle bergs penetrated the
air. The domed bergs had a range of height-draft ratios far greater than the other
classes (1:2.63-1:10.56) and also by far the smallest average ratio (1:6.30) of any of
the visual classes. Domed bergs were generally the smallest in size as a class.
Table 5. Height to draft ratios for domed bergs
Height Depth Ratio
(meters) (meters) (1:
30 79 2.63
16 52 3.25
12 65 5.42
21 157 7.48
13 92 7.07
9 95 10.56
12 92 7.67
Mean 16 90 6.30
Range 9-30 52-157 2.63-10.56
The assumption was made that the height to draft ratios of icebergs form a continuous
distribution. Using a Kruska}-Wallis one-way analysis of variance technique, Welsh
(1975), the hypothesis that the average ratio for icebergs was not significantly
different for the gross visual shspe classes was tested. This resulted in the conclusion
that, for the sampled icebergs, there was no significant difference between
classes. For summary purposes the average of the visual class averages (1:3.95)
can be used as descriptive of the height to draft ratio of icebergs regardless of
visual shape class.
Since one visual class was not significantly different from another with respect to
the height to draft ratio, all classes were combined and the ratios were plotted
against iceberg height. The distribution was by no means linear and was best
represented by the pbwer curve (Figure 3).
l/Ratio = 49.4 (Height)-·8
413 Robe 7

North Atlantic Ocean. Ice Seminar: A <strong>Conference</strong> sponsored by the Petroleum
Society of CIM, Calgary, Alberta, May 1968.
Rodman, H. 1890. Reports on Ice and Ice Movements in the North Atlantic Ocean.
U.S. Navy Hydrographic Office, No. 93, 26 pp.
Smith, E. H. 1925. <strong>International</strong> Ice Patrol. The Meteorological Magazine. The
Meteorological Committee, Air Ministry, Vol. 60, No. 718.
Steenstrup, K. J. V. 1893. Bidrag til Kjendskab til Braeerne og Brae-Isen i
Nord-Gronland, Meddelelser om Gronland, Vol. 4, pp. 71-112.
Welsh, J. P. 1975. Personal Communciation.
415 Robe 9

INTRODUCTION
In the open sea, ridging occurs most often by failure of relatively thin lead ice between
thicker ice floes. In the near shore area, that is, from the edge of the 1andfast ice
sheet to the shoreline, thin ice is usually absent, but when moving pack ice approaches
an ice sheet already in contact with the shore, failure and subsequent ridging or hummocking
must take place. In general, the line along which this occurs will be the boundary
between the ice sheets, but in some instances it can occur at the shore line, or it can
be initiated at points of stress concentration within the fast ice sheet itself. The
composition and morphology of the resulting ridges will depend upon a number of parameters.
Ice strength and thickness, and the distance through which the pack moves during
pressuring should obviously be important as they are in the case of ridging in the pack
ice away from shore. However, the role of other variables in the process of ridging in
shallow water still needs to be considered.
In their model of pressure ridging in the open sea, Parmerter and Coon (1972, 1973) examined
the partition of energy during the process of pressure ridge building. Three categories
were identified: (1) gravitational potential energy of both the sail and keel, (2)
friction, and (3) surface energy due to the creation of new surfaces by fracture, with
the first of these the most important. In shallow water this balance may be changed in
several ways. Grounding of a growing ridge will largely remove bouyancy as an aid in
achieving the ultimate height of the ridge so that beyond some sail height, which depends
upon the water depth and the average density of the keel, all the height of the ridge
must be produced by "bulldozing" of blocks up the pile. In addition, grounding will necessarily
alter the ratio of sail height to keel depth. Thus, it can reasonably be expected
that the forces required to build a ridge to some arbitrary height will differ
between shallow water and open sea, as will the partition of gravitational potential energy
between sail and keel. The "bulldozing" process can be expected to increase the
amount of grinding and crushing of the ice over that which occurs in ridging in the open
sea thus changing the contribution of the formation of new surface energy by fracture to
the energy balance of the ridging process. With regard to friction, Parmerter and Coon
(1972, 1973) consider only the friction at the top and bottom of the ice sheet as it is
,pushed into the rubble of the growing ridge. However, based upon limited field observations,
it appears that after grounding has occurred, the edge of the advancing ice sheet
tends to be driven up the offshore side of the ridge rather than ahead into the rubble
pile, so that the role of frictional forces also needs to be re-eva1uated. Gouging of
the sea floor during grounding is a distinct possibility and will also involve some energy
loss which must be considered. Further, note that in shallow water the ice on one
side of the line along which a ridge grows is usually stationary, so that it is likely
that the relative sense of motion across a growing ridge in shallow water will be more
divergent than is generally the case in moving pack ice. Thus, it appears that ridging
in shallow water is very different in many respects from ridging in the open sea.
Finally, there is the question of what causes a ridge to develop in a particular location.
Weeks, Kovacs and Hibler (1971) have pointed out that deep-keeled floes or pressure
ridges which drift into shallow water can serve this purpose. In addition, it is possible
that other variables such as coastal geometry (as modified by 1andfast ice) ice drift
patterns, and water depth may operate independently to localize ridges.
The purpose of this paper is to describe the use of radar data in the interpretation of
ridge building processes in the near shore zone. In particular, interpretation involving
knowledge of the movement vector of the pack ice and the duration of events is emphasized.
This provides a supplement to field data, and permits additional insights to
be made regarding these processes.
418 Shapiro 2

DESCRIPTION OF THE RADAR SYSTEM
The radar system operated by the University of Alaska at the Naval Arctic Research Laboratory,
Barrow, Alaska (Fig. 1), is a 3 cm X-band, standard ships radar which is mounted
-'
.J
.J --i\
.:.iI1Ii .. ...--..,."
Figure 1. View of the radar tower at Barrow.
on a tower 12 m above mean sea level. This gives the radar a maximum range of about 12
km. However, the effective range is less than this because of the way in which the data
are taken. No artificial reflectors are used. Instead, natural reflecting surfaces on
the ice supply the energy return, and the number of reflectors large enough to supply
sufficient return to be visible on the radar screen at a given distance decreases with
distance from the source. Experience has shown that the optimum scanning range is 3
nautical miles (5 . 6 km), although some data are taken at 1.5 nautical miles (2.8 km) and
6 nautical miles (ll.l km) scans. The data are recorded photographically by a 35 mm
camera placed in front of the screen which photographs the screen and a clock face every
3 minutes. The resulting pictures (Fig. 2) can then be displayed as time-lapse motion
pictures or as individual frames.
OBSERVATIONS AND DISCUSSION
To date, most of our work has been on a sequence of data acquired during the winter of
1973-74 when an intense storm drove the pack ice into the near shore zone creating a
variety of ridges and hummock fields. Combined with field data, this work has produced
some interesting observations on the processes involved in ridge building in shallow
water. The event and the resulting ice features have been described in detail elsewhere
(Shapiro, 1975).
419 Shapiro 3

BARROW
Jan 74
Figure 2. Sample frame of radar data. True north corresponds approximately to 90 0 on
the outer ring of the display. The coastline extends along a line from 130 0
to 310 0 (i .e., northeast-southwest) with land to the southeast. Black dots
in the offshore area are natural reflecting surfaces on the pack ice. The
prominent lines extending northwest-southeast across the frame are one nautical
mile (1.85 km) apart.
During this storm, which occurred between 30 December 1973 and 1 January 1974, pack ice
about . 5 - .6 m thick was driven essentially parallel to the shore near Barrow from
southwest to northeast (Fig. 3). The peak winds during the storm reached 90 km/hr at the
Barrow airport (Fig. 4), but there are unofficial reports of gusts up to twice that value.
The velocity of the pack ice during the period of high winds was about 8.0 km/hr.
The location map of the Barrow area (Fig. 3) indicates a shoal near the radar site, and
this feature ultimately marked the location of a linear hummock zone which will be described
below. In the early stages of the storm several reflectors, which represented
grounded floes with deep keels, stopped over this shoal, but this occurred only at low
drift velocities . No new floes grounded at higher velocities and, in fact, those already
grounded were dislodged by the impact of the moving pack ice when the ice velocity increased
towards it maximum value. Subsequently, no reflectors stopped over the shoal and
the pack ice simply crossed it as if it were not there. This continued until the movement
of the pack was blocked to the northeast, out of the field of view of the radar.
Then, reflectors were observed to stop in sequence from northeast to southwest, filling
the area shoreward of the shoal, while offshore the ice continued to move in a northeasterly
direction. Along the shoal a grounded hummock field formed as the result of the
interaction between the moving and stationary ice. This feature was 4-5 km long and 125-
420 Shapiro 4

he position of this hummock field can be interpreted as resulting from the drift pattern
)f the pack ice. Note that the orientation of the shoal is such that it tends to deflect
lortheast setting currents offshore around Pt . Barrow. It therefore seems possible that
1 similar flow geometry might be followed by the ice moving up the coast so that, when
the movement of the ice in the nearshore zone north of the radar field of view was blocked,
the drifting ice would be deflected offshore along the line of the shoal. Then, as
noted above, the hummock field reflects the deformation along the line between the moving
pack ice and the stationary ice inshore.
The hummock field itself was generally linear to sinuous in plan view, which implies a
shear origin, and this is to be expected from its orientation of 6° to the ice drift vector.
However, there is no evidence that the intense cataclastic deformation which is typical
of shear ridges, occurred while the hummock field was ,being formed although, as
discussed below, its outer boundary' was marked by a shear ridge. Instead, the resulting
feature consisted of a chaotic jumble of broken blocks simply piled as would be the case
in a pressure ridge. Three other ridges in the radar field of view showed the same
characteristics of rather straight aspect in plan combined with a block composition typical
of pressure ridges. These also formed at relatively low angles to the drift direction
of the ice and were not bounded by shear ridges. It thus appears that the smoothly
sinuous to straight form of a ridge indicates origin in shear even in the absence of extensive
brecciated ice.
As noted above, the dimensions of the hummock field were 125-150 m in width, with an
average elevation of about 3 meters. From the radar data it can be determined that this
width was attained by gradual growth in a seaward direction. This implies, in turn, that
the 3 m elevation may be regarded as a limiting height (possibly analogous to that determined
by Parmerter and Coon (1972, 1973) for ridging in the open sea) above which the ice
could not be piled under the conditions of ice properties and drift which existed at that
time. In contrast, ridges up to 9 m in elevation formed off Barrow during the winter of
1974-75. The thickness of the ice at the time these developed was .6-.8 m, similar to
that at the time of formation of the hummock field, but the drift direction associated
with these ridges was almost at a right angle to their strike, while the ice velocity was
much lower. The ridge crests were sharp, and thus it is likely that the limiting height
had not yet been reached when the movement stopped. Based upon this, it can be speculated
that the difference between the limiting heights in these two cases may largely be due to
the difference in the angle of approach of the pack ice to the stationary boundary, and
consideration of this variable should be a part of any effort to model ridging in shallow
water.
The idea of the limiting height of a ridge being a function of the angle of approach of
the pack ice to the landfast ice edge introduces the possibility of understanding one of
the more puzzling aspects of landfast ice di stribution . That is, the reason why the edge
of the landfast ice often tends to occur close to the 20 m depth contour along the Alaska
coast. The radar data has provided imagery of the landfast ice being emplaced at Barrow
on four occasions, and in each of these, the drift vector of the pack ice was nearly
parallel . to the shore at least over part of the boundary. Now, assume that the 3 m average
elevation of the hummock field described above represents a "l imiting height " for
ridges formed when young sea ice is driven against the coast at a low angle. Then, for
any reasonable value of the ratio of sail height to keel depth (say a range of 1:5 to
1 :8) it follows that the farthest from shore that a ridge could ground under these conditions
would be approximately at the 20 meter depth contour and this did indeed mark the
initial outer edge of the fast ice at Barrow for these occasions. Once grounded, as
Weeks and Kovac s (1970) point out, the ridge will tend to stabilize, and will thus serve
as a nucleus for further growth laterally along the edge of the landfast ice, as well as
seaward or in elevation as thicker, stronger ice is driven against the ridge at different
angles later in the year.
422 Shapiro 6

In dealing with a natural system as complex as drifting sea ice, there is a danger of
overgenera11zing results from studies in one area by extending them to other areas where
they may not apply. In the case of the hypothesis offered here, one obvious example of
this can be cited. Cooper (1973) described the distribution of 1andfast ice in Mackenzie
Bay, and noted that while the edge of the 1andfast ice sheet approximately coincides
with the 20 m depth contour, it is not marked by a continuous ridge. Thus, in this area
another explanation may be required. However, the 1andfast ice along most of the Alaska
coast is often bounded by a continous ridge system, of which shear ridges are a common
constituent (Stringer, 1974). This hypothesis may therefore have some validity in that
area, but more information is needed on the late fall drift pattern along the coast in
order to verify it.
As noted above, the offshore side of the linear hummock field described above was bounded
by a shear ridge and an accompanying zone of catac1astica11y deformed ice, such as described
by Weeks and Kovacs (1970). This ridge developed under the same drift conditions
as the linear hummock zone and it has not been possible to determine from the radar exactly
when the transition from hummocking to shear ridging occurred, although it appears
that the building of the hummock field was completed in less than one and one-half hours.
The reason for this transition in deformation mode is open to question. However, a possible
solution is suggested by the work of Pritchard and Schwaeg1er (this volume) in
which the deformation of an ice sheet fixed to shore along one edge and acted upon by a
constant, low velocity, longshore wind, was calculated. The results indicate that with
time, the ice in the nearshore area tends to diverge, which may indicate that the component
of internal stress normal to shore is decreasing. Now, if the assumption is made
that similar conclusions would result for winds of high velocity and slip along the boundary
of the ice sheet, then it follows that as the magnitude of the component of stress
to normal to the ridge is reduced, the ice will stop piling and instead will simply be
driven parallel to the offshore face of the hummock field resulting in the grinding and
crushing associated with the development of shear ridges. This conclusion is tenuous at
present, but seems worthy of further study.
Finally, it is of interest to consider the effect on the pack ice of ridge building at
the 1andfast ice edge. The radar system provides excellent data for this purpose. Figure
5 shows two plots of the component of the ice velocity vector parallel to shore versus
distance from shore. The plot in the left half of the figure represents movements
during the first 30 minutes after the linear hummock zone, described above, was formed.
It appears from the radar data that the zone was growing seaward throughout this time,
so that the interaction at the pack ice-1andfast ice boundary consisted of breaking of
the pack ice and "bulldozing" of blocks up the pile. The data in the right half of the
figure were acquired about 1-1/2 hours after the first set, during the building of the
shear ridge and adjacent zone of catac1astica11y deformed ice. The same number of points,
thirty-five, were used to prepare both figures. Note that the magnitudes of the components
of the velocity vectors normal to the shore at these times were small relative of
those of the components parallel to shore, and were not considered further.
It is apparent from both data sets that for distances greater than about 2.5 km from
shore there is only a slight increase in velocity with increasing distance from shore.
Further, the magnitude of the velocity at that distance is about the same for both sets.
However, inshore from 2.5 km there is a marked difference between the point distribution
of the two figures. For the time when the shear ridge was forming there is little change
in the velocity profile as the shore is approached, and it is only in the first kilometer
from the ridge that any deviation from a straight profile becomes apparent. In contrast,
the profile for the time when the hummock zone was being built is strongly curved in the
first 2.5 km from shore, and the minimum velocity, based upon data points only, for this
profile is above 2 km/hr less than that for the second profile. Extrapolation indicates
that this difference may be greater than 3 km/hr at the boundary of the stationary ice.
423
Shapiro 7

THE USE OF FLATJACKS FOR THE IN SITU DETERMINATION
OF THE MECHANICAL PROPERTIES OF SEA ICE
Lewis H. Shapiro
Geophysical Institute
University of Alaska
Fairbanks, Alaska
United States
and
Earl R. Hoskins
Department of Mining Engineering
South Dakota School of Mines and Technology
Rapid City, South Dakota
United States
ABSTRACT
A short series of tests was conducted at Barrow, Alaska, in April and May of 1975 to
evaluate the applicability of flat jacks for the in situ measurement of the mechanical properties
of sea ice. The results indicate that, with suitable sensing and timing instrumentation,
reliable values o{ the elastic and viscoelastic parameters can be obtained.
Approximate values of 2 x 10 1 dyne-min/cm 2 and 4 x 10 9 dyne/cm 2 were determined for the
coefficient of viscosity and Young ' s modulus, respectively.
427

INTROOUCTION
The objective of the project for which this investigation wa s done is to develop hardware
and methods for the in-situ determination of the viscoelastic properties of sea ice. The
general approach we are taking i s to use flat jacks to generate loads and then to sense
the effects of these loads using stress and strain transducers embedded in the ice.
The purpose of this short study was to examine the behavior of flat jacks in the ice environment
and determine what modifications were needed in the basic sys tem for this project.
While conducting these tests , data were incidentally obtained regarding mechanical
properties of the ice. For reason s described below, the numbers are too crude to report
as values of these properties, but they are close enough to those determined by other
investigators to illustrate the utility of flat jacks in tests of this type .
EXPERIMENTAL TECHNIQUES
Flat jacks (Fig. 1) are simple and inexpensive to build, consisting only of an envelope
of
Figure 1. Two 46 x 46 cm steel flat jacks.
thin, sheet steel, or other suitable material, welded together around the edges. The
load is applied by pumping a fluid between the plates and pressure in the system is
monitored by a pressure gauge installed in the fluid line. Note that the properties of
the fluid are not important except, of course, that for these experiments it must not
freeze. Further details regarding this technique can be found in Hoskins (1966) and
Jaeger and Cook (1969).
Flat jacks can be constructed to any dimensions and shape. For our purposes, square or
rectangular forms seem most appropriate, although circular and semi-circular flat jacks
have been used in rock mechanics experiments.
Stresses around an expanding flat jack are relatively constant over about 90% of the jack
face, with a tendency for somewhat higher stresses towards the margins. To get an indication
of the stresses in front of a flat jack face, we calculated the stresses along a
line normal to the center of the face using Muskhelishvili's (1963) plane-strain, elastic
solution for stresses around a slit under internal pressure. The results are shown
in Figure 2. Note that the normal load dominates, as would be expected, and reasonable
428 Shapiro and Hoskins 2

Figure 6. Single flat jack array in place prior to testing. The
top of the flat jack is the thin black line in the center
of the photograph. Note position of pegs.
while the flat jacks were installed with their upper edges at the i ce surface. Thus, the
loading was restricted to the upper part of the ice sheet.
Pressures in the flat jack were maintained using a hand pump, with JP-4 as the pressurizing
fluid, and were read to .07 kg/cm 2 (1 psi) from a pressure gauge. Displacements
were determined by measuring the distance between the pegs across the expanding flat jack
using a vernier caliper. Unfortunately, this device could only be read to an accuracy
of 2.5 x 10- 3 cm (.001 in), one significant figure (in inches) less than we believe is
necessary for good data. In addition, time was recorded only to the nearest minute.
However, based upon some simple calculations regarding the possible range of errors, the
results of the creep tests are considered to be within an order of magnitude of the correct
values.
RESULTS
A total of five creep curves and one relaxation curve were obtained in the series of
tests . A good example curve is shown in Figure 7. Approximate creep rates were established
for those parts of the curves which appeared to have reached steady state. Then,
plotting these against values of the stresses based upon the elastic solution shown in
Figure 2, a coefficient of viscosity of 2 x 101 1 dyne-min/cm 2 was obtained. This corresponds
to the linear viscous element of a Maxwell model when the ice is modeled using
either a Maxwell element alone or a Maxwell element in series with a Kelvin element to
form a 4-parameter model .
For purposes of comparison we made the same calculations using the data from the first
four creep tests in compression reported in Peyton (1966), and obtained a value for this
parameter of 1 x 10 12 dyne-min/cm 2 . The extensive series of in- situ tests of viscoelastic
properties reported in Tabata (1958) yielded values ranging from 1011 to 10 12 dynemin/cm
2 for horizontal beam deflection, while laboratory results on horizontally oriented
cores gave values of order 1011 . Based upon the similarity of these results to ours
we anticipate that, with certain necessary refinements in measurement procedures, flatjacks
can be used to obtain realistic creep curves.
433 Shapiro and Hoskins 7

INTRODUCTION
Existing floating sea ice sheets in the polar regions have been utilized to support surface
and air operations. Adequate airfields and ice roads can be economically Constructed
and maintained on annual sea ice; for example, the sea ice near McMurdo Station,
Antarctica, has been used each season since the <strong>International</strong> Geophysical Year (IGY).
Now, operations on ice-covered waters throughout the world have greatly increased due to
the search for and exploitation of new energy resources. Specifically, Naval polar operations
are dependent on sea-ice airfields and roads to provide heavy cargo and logistics
support for United States activities (Dykins, 1969).
This paper presents a summary of the analytical development of a finite element computer
code which provides an elastic solution to a laterally loaded ice sheet. Then, analytical
and experimental results are compared using laboratory saline-ice plate studies. Any
stress analysis prediction of ice sheet behavior requires prior knowledge of reasonable
sea-ice material properties. Past flexural strength and elastic modulus testing performed
at the Civil Engineering Laboratory (CEL) is reviewed to provide a foundation for
input material properties to the computer code. Flexural failure strengths are reduced
by an arbitrary factor of safety to give allowable stresses. Then, calculated stresses
determined from the finite element analysis are compared with allowable stress values.
Thus, a series of allowable-load-versus-ice-thickness curves for both C-130 and C-14l
aircraft can be formulated, along with a table of minimum allowable ice thicknesses for
different typical vehicles performing polar logistics operations.
ANALYTICAL METHOD
The main portion of the general axisymmetrical finite element computer code employed in
this investigation was written by Wilson (1965). The formulation is based on the direct
stiffness method with triangular or quadrilateral conical elements representing the axisymmetrical
ice sheet. Generally, the ice sheet may have a complex configuration and can
be considered as a layered medium. The constitutive relations can be orthotropic, nonlinear,
and temperature-dependent. External loading can be either structural or thermal
in nature with an option to consider gravitational forces.
A special mesh generator was added to the program to facilitate input for the plateshaped
solids considered in this study. In addition, subroutines are incorporated into
the main computer deck to simulate a fluid foundation and superposition of loads from any
combination of wheel or track configurations.
Fluid Foundation
Development of the simulated fluid foundation is achieved by calculating a vertical resistance
proportional to the displaced volume and adding it to the global stiffness
matrix of the solid for every node on the solid-fluid interface. In formulating the
vertical resistance, compatibility of displacements requires off-diagonal terms as well
as diagonal terms; all of these terms are determined by energy principles consistent with
the finite element formulation.
Superposition
The superposition development determines the combined state of stress in an ice sheet resulting
from an arbitrary configuration of circular loads. The principle of superposition
has inherent limitations; first, the stress-strain relations must be linear, and
second, the ice sheet must be free of variations of material property and layer geometry
with respect to any horizontal plane. Fortunately, typical ice sheets are composed of
horizontal layers that have material property variations and temperature changes solely
in the vertical direction. Moreover, the "infinite" extent of the ice sheet nullifies
440
Vaudrey and Katona 2

A dynamic method for determining Young's modulus involves vibrating an ice specimen with
a transmitter and measuring either the velocity of the longitudinal waves or the fundamental
transverse frequency from a receiver. The dynamic elastic modulus can be calculated
from simple equations, if either the velocity or frequency is known. Dynamic
modulus values are generally higher than those obtained by deflection measurement
methods; therefore, average elastic modulus values for different temperatures are used
to establish the modulus gradients for each thermal period of Table 1. For the first
three periods the modulus on the left corresponds to the ice sheet surface temperature,
while the one on the right corresponds to the temperature at the bottom of the ice
sheet. There is no elastic modulus gradient for the fourth period; it depicts an almost
isothermal condition.
In addition to the flexural strength and elastic modulus properties of annual sea ice
listed in Table 1, Poisson's ratio is required as input to the finite element computer
code to determine the bearing capacity of a floating ice sheet. Throughout this study
a constant value of 0.3 was assumed for Poisson's ratio.
ICE THICKNESS CALCULATIONS
After having performed experiments to determine sea-ice material properties and then developed
a finite element model to predict sea-ice behavior, it is possible to calculate
required ice thicknesses for aircraft and vehicular operations. First, a brief discussion
of the program input is presented.
Input Parameters
The input requirements for the finite element code are based on a circular, plate-shaped
solid supported by a fluid foundation. The ice sheet can be of arbitrary dimensions
with as many as 12 material layers and can accommodate circular loadings of uniform
pressure but abritrary radius and magnitude. These input parameters can be divided into
three basic categories: (1) program control and finite element mesh generation; (2) ice
sheet material properties; and (3) aircraft or vehicle loading conditions. The first
category sets the boundary conditions and defines the number of material layers and number
of element rows for each layer of the ice sheet. Also, the mesh generator establishes
the number of total elements, elements under the load, and elements in contact
with the supporting fluid. In the second category temperatures are assigned to the top
and bottom of different material layers within the ice sheet, along with a corresponding
elastic modulus and Poisson's ratio for each temperature. The thickness for each material
layer is given, as well as the density of the fluid foundation (in this case, seawater).
For the last category the input parameters are the number of circular loadings
(e.g., wheels or tracks) and the total load (weight of aircraft or vehicles), in addition
to such individual wheel or track circle characteristics as: (1) radius and pressure
of load; (2) x-y coordinates (to locate each loading); and (3) fraction of total
weight carried by individual loading.
Aircraft Load Curves
Calculating the required effective ice-sheet thicknesses for the load ranges of different
aircraft requires exercising the finite element computer code in two series of repeated
steps. Initially, the input parameters for a single thermal period are selected
from Table 1, and several ice sheet thicknesses are chosen for this period. By performing
consecutive computer iterations, maximum tensile stresses are found at the bottom
of the ice sheet for each specified thickness. This iterative process generates an aircraft-load
versus ice-thickness curve for a given thermal period, and the entire cycle
is repeated for the remaining periods.
449 Vaudrey and Katona 11

In Figure 10 curves for both C-130 and C-14l military aircraft are shown for all four
thermal periods defined in Table 1. Period one curves for both aircraft are determined
by comparison of the calculated stresses with an allowable stress that is found by reducing
beam failure strengths by 30%, an equivalent safety factor of 1.20. All other
curves reflect a reduction of 25% in the failure strength, representing a safety factor
of 1.15.
Vehicle-Ice Thickness Table
Minimum ice sheet thicknesses for typical polar vehicles are presented in Table 2.
Again, the iterative process is repeated for each thermal period. However, for relatively
constant loads (like vehicles), the iterations converge to a single, tabular
point rather than a curve in the case of aircraft, which nave a wide cargo/fuel range.
All vehicles are a sampling of typical logistics support equipment operating in polar
regions. Vehicle weights used in thickness calculations are also given in Table 2. Allowable
stresses represent a sufficient reduction in failure strength to produce a
safety factor of 1.50.
Operational Field Procedure
Once ice-thickness curves and tables are determined, proper field procedure is still
necessary to operate safely on any ice sheet. Of course, both ice tnicknesses and temperatures
must be monitored. At the start of the operating season, a detailed thickness
survey should be made down each edge of sea-ice runways, alongside ice roads, and around
all other operating areas. Maintenance of a complete up-to-date temperature record of
the ice sheet is necessary for a verification that the proper thermal period is being
used. Ice temperatures should be measured by thermocouple or thermistor stations at
five or more locations dispersed for general coverage of the entire operational area.
Routine inspection of the ice sheet surface should be made for existence of cracks.
These cracks should be recorded for length, location, and whether they are wet or refrozen.
CONCLUSIONS AND RECOMMENDATIONS
The elastic finite element computer code has demonstrated its capability as an accurate,
versatile, and powerful tool for analyzing and predicting sea-ice-sneet behavior subjected
to lateral loading. The finite element method can easily handle mUltiple
wheel/track loadings as well as the temperature gradient through the ice sheet. An increase
in the scope of the analysis to include dynamic response would require minimal
effort by considering inertial terms through a consistent mass matrix. Another expansion
of the finite element method to predict the linear viscoelastic response of sea-ice
sheets to surface loading has already been performed (Vaudrey and Katona, 1975). However,
as yet, the limit of permissible deformation of floating ice sheets under sustained
loadings has not been defined.
ACKNOWLEDGMENTS
This paper covered work performed under Naval Facilities Engineering Command project
number YF52.555.00l.0l.002, entitled "Structural Analysis of Ice for Naval Polar Applications."
The authors wish to thank Mr. J. E. Dykins for consulting on the material
properties of sea ice and reviewing this paper and Mrs. Barbara Hamilton for typing and
proofreading the manuscript.
450 Vaudrey and Katona 12

REFERENCES
Dykins, J. E. 1969. Technical Report R-641: Sea-Iae Bearing Strength in Antaratiaa
- Airaraft Load Curves for MaMUrdo Iae Runway. Naval Civil Engineering
Laboratory, Port Hueneme, 36 pp.
Katona, M. G. and K. D. Vaudrey. 1973. Technical Report R-797: Iae Engineering -
Summazy of EZastia Properties Researah and Introduation to VisaoeZastia and
NonZinear AnaZysis of SaZine Iae. Naval Civil Engineering Laboratory, Port
Hueneme, 71 pp.
Timoshenko, S. and S. Woinowsky-Krieger. 1959. Theory of PZates and SheZZs.
MrGraw-Hill, New York, pp. 63-67.
Vaudrey, K. D. 1975. Technical Note N-1417: Iae Engineering - EZastia Property
Studies on Compressive and FZexuraZ Sea Iae Speaimens. Civil Engineering
Laboratory, Port Hueneme, 16 pp.
Vaudrey, K. D. and M. G. Katona. 1975. Viscoelastic Finite Element Analysis of
Sea Ice Sheets. In G. E. Frankenstein, ed., Proceedings of the IAHR <strong>Third</strong>
<strong>International</strong> Symposium on Ice Problems. 18-21 August 1975, Hanover, NH,
pp 515-525.
Wilson, E. L. 1965. Structural Analysis of Axisymmetric Solids. AlAA Journal.
3:2269-2274.
453 Vaudrey and Katona 15

FAST ICE STUDIES IN WESTERN DAVIS STRAIT
R. L. Weaver and R. G. Barry
Institute of Arctic and Alpine Research
University of Colorado
Boulder, Colorado
United States
and
J. D. Jacobs
Department of Geography
University of Windsor
Windsor, Ontario
Canada
ABSTRACT
Shore fast ice is one of the predominant geographical features of the eastern coast of
Baffin Island. Since 1971 field parties from this Institute have studied various aspects
of the surface energy budget on the fast ice in the area of Broughton Island (67°N, 65°W)
during both the growth and decay phases of the annual ice cycle. The program has fooused
on relationships between synoptic weather events, observed changes in the surface energy
budget, and ice response. Most of the work has centered around the break-up phase (May
to mid August).
Research efforts during the growth phase have concentrated on (1) predictive modelling of
ice growth and (2) mapping inter-annual spatial variation of fast ice distribution. A
simple thermodynamic model has been employed to predict ice thickness using mean daily
air temperature and snowdepth data.
The ablation and break-up processes have been subjectively divided into five stages: (1)
pre-melt; (2) melting snow; (3) puddle formation; (4) thawhole formation and drainage; and
(5) break-up. The timing and length of these stages are shown to vary from year to year.
A subjective catalog of synoptic atmospheric pressure patterns, based on surface pressure
maps, has been used in an attempt to explain the large range in fast ice conditions observed
during the past five years. Examination of the variation in frequency of synoptic
types favorable or unfavorable to ice melt has proved inadequate by itself to explain the
variations in ice conditions. However, refinement of this work using an objective classification,
in conjunction with energy budget parameters, is underway.
455

INTRODUCTION
Shore fast ice is one of the predominant geographical features along the eastern coast
of Baffin Island, occurring with varying seaward extent from Bylot Island southward to
just north of Cape Dyer. Our study has concentrated on the region from Home Bay to
Padloping Island (see Figure I), with field studies since 1971 of various aspects of
the surface energy budgeL on the fast ice in the area of Broughton Island (67 0 N, 64°W)
during both growth and decay phases of the annual ice cycle. The program has focused
on relationships between synoptic weather events, observed changes in the surface
energy budget, and ice response.
This work began as an exploratory study of the synoptic conditions influencing the climate
of Baffin Island, with particular respect to its glacierization. First we examined
the seasonal climatic characteristics in terms of a classification of the mean
sea level (MSL) pressure field. Since then, the synoptic climatological catalog has
been used as a basis for interpreting present shore fast ice conditions.
The synoptic type gata!og wag dev 6 loped by subjectively dividing the MSL pressure field
over the sector 55 -80 N, 50 -100 W into 42 different "static" types or groups on a
daily basis over a five year period (Barry, 1974). The classification scheme was later
simplified into six cyclonic and six anticyclonic types so that sample sizes were large
enough for statistical testing of between and within group variances. Thus the rationale
of this study in part is to attempt explanation of observed changes in the fast
ice sheet in terms of the changes in frequency and distribution of these synoptic types.
This paper summarizes data from four distinct, but interconnected, phases of our work.
Results are presented on: (1) theoretical ice growth modelling with comparison to field
measurements in the Broughton Island region; (2) descriptive modelling of fast ice
ablation and break-up with application to the climatically diverse 1971-1975 ablation
seasons; (3) satellite mapping of summer fast ice morphology with special emphasis on
what appear to be persistent features from year to year; and (4) preliminary results
of an analysis of synoptic weather-pattern correlations to variations in seasonal fast
ice characteristics.
ICE GROWTH
Shore-fast ice is formed in situ from young ice during the yearly freeze-up or when
drifting first or multi-year ice is frozen into a sheet which is fixed to the shore
(Atmospheric Environment, Canada, 1965; WMO, 1970). The late winter fast ice in the
Home Bay-Broughton Island area extends eastward some 50-80 km in most years and generally
corresponds approximately with the 180 m isobath (Jacobs et al., 1974).
Our predictive modelling of ice growth is based on the one-dimensional heat flow equation
(Jacobs et al., 1974, p. 41).
h depth of snow
H ice thickness
t time
Ts
temperature of
Tw = temperature of
L latent heat of
p density of ice
snow surface
water
fusion
456
thermal conductivity of ice
thermal conductivity of snow
(1)
Weaver/Jacobs/Barry 2

Figure 1. Location Map of the eastern Canadian Arctic and
Study area (outlined)
457 Weaver/Jacobs/Barry 3

This equation accounts for the heat flow through two layers of differing conductivity,
snow and ice. A form of equation (1) was used to calculate fast ice thickness for the
deep water area some 5 km east of Broughton Island. The predicted ice thickness by the
end of May 1972 was 175 cm. Five holes drilled in the first week of June over a 20 km
transect east of the island showed an average thickness of 180 cm with a range of ± 10
cm. Model results in the 1973 and 1974 seasons were not as accurate (± 20%). In 1973,
this was probably due to the presence of up to 7/10 second year ice and to abnormally
heavy snow pack in the spring, which results in non-linear effects, not accounted for
by equation (1). In 1974, observed thicknesses were only available in the harbour area.
Ice thickness in estuarine zones cannot yet be predicted because the model does not
take account of the oceanic turbulent exchanges associated with tidal action.
BREAK-UP
The ablation and break-up processes may be divided into five stages (Figure 2): (1)
premelt, (2) melting snow, (3) puddle formation, (4) thawhole formation and drainage,
and (5) break-up (Jacobs et aZ., 1975). These stages generally apply to the fast ice
observed during the 1971-75 summers in the vicinity of Broughton Island. Similar
schemes have been reported by Zubov (1945) and WMO (1970). This sequence of events
however, probably does not occur in areas with heavily ridged or rafted fast ice.
In late May and early June, the cold, snow-covered fast ice sheet lFigure 2, stage 1)
starts to warm. Lieske (1964) has reported that thermal conduction from the water beneath
the ice was important during spring at Barrow, but we have no data for this early
stage. Daily totals of solar radiation are large in May-June and from 10 to 30% of the
incoming radiation is absorbed by the snowpack leading to melt with water percolating
to the relatively impermeable ice surface (Figure 2, stage 2) where it coalesces to
form incipient puddles. A snowcrust may persist for many days if air t"'peratures remain
below freezing and early summer storms deposit fresh snow. Eventually, however,
the snow is melted and the ice surface develops alternating hummocks and puddles
(Figure 2, stage 3). The meltwater puddles drain through fractures, leads, and seal
breathing holes to form a low density fresh water layer 0.5 to 1 m deep between the bottom
of the ice and the seawater below (Figure 2, stage 4) in areas where the tidal or
ocean currents are insufficient to overcome the strong density inversion. A similar
fresh water layer was also observed by Langleben (1966). Where runoff is restricted,
extensive sheets of melt water form on the surface to depths of some tens of centimeters.
With the average albedo of the composite hummock-puddle surface now considerably reduced
(from about 0.8 to 0.4), the absorption of solar radiation (with some 2100 joules
cm- 2 day-l potentially available at mid-summer at 70 0 N) leads to rapid ablation. Up to
5 cm of ice melt per day was recorded during stages (3) and (4) in 1973 and since ice
is a translucent material allowing solar radiation to penetrate deeply, the actual mass
change was probably greater. The puddles, with an albedo of 0.25 or lower (Langleben
1969, reported values of 0.19 at Tanquary fiord), absorb ca. 15-30% more radiation than
does the adjacent ice surface. The greater ice melt under the pools allows some of
them finally to melt through forming thawholes (Fig. 2, stage 4). The final break-up
(Fig. 2, stage 5) occurs when the structurally weakened fast ice sheet is broken apart
by wind and wave action. Rarely does the ice totally melt in situ. LANDSAT and NOAA-2
imagery show that at first giant and vast floes separate from the ice sheet at the
edges of polynyas or along flaw leads that have persisted throughout the winter, or appeared
early in the summer in regions of thin ice. Because there is little mixing of
the waters between these floes, solar heating of the surface waters is intense and the
final melting and break-up of the floes may occur in a matter of days.
The approximate melt stage transition dates for 1971-75 seasons are presented in Figure
3. These dates are only approximate because progression from one stage to the next is
458 Weaver/Jacobs/Barry 4

Figure 4. Map of generalized progression of break-up stages in the Broughton
Island area.
463 Weaver/Jacobs/Barry 9

fiord ice. Generally, the retreat starts to the south of Padloping and works northward
along the coast, controlled by local geographical features.
This seaward lag in break-up generally holds along the east Baffin coast. The northsouth
timing of break-up, on the other hand, does not appear to follow a set pattern
from year to year. Instead, the break-up has been observed to begin following rapid offshore
movement of the pack ice along the fast ice edge in mid- to late summer, suggesting
a link to strong off-shore winds.
SYNOPTIC EVENTS RELATED TO BREAK-UP
The characterization of synoptic patterns of atmospheric circulation which may be of
significance to the rate of fast ice decay has been carried out by cataloguing daily MSL
pressure patterns according to a classification by Barry (1974). The relative frequencies
of the major cyclonic and anticyclonic groups in individual summers show only
small departures from the average for 1961-70 and are clearly insufficient, by themselves,
to account for the year-to-year differences in ice conditions:
TABLE 1.
Cyclonic types
AnticYclonic types
1961-79(%)
49
51
Type frequency in July-August
1971
-2
+2
Departure
1972
+3
-3
from 1961-70(%)
1973 1974
-9 -6
+9 +6
The types can be categorized according to their associated weather characteristics,
based on Barry, Bradley and Jacobs (1975), as to those types more likely to advance or
retard ice wastage. Thus, warm, rainy cyclonic situations and warm, clear-sky anticyclonic
situations will both accelerate ablation through reduced albedo and increased
net radiation, whereas situations producing snowfall, cold and calm weather, or high
overcast skies will retard ablation. An analysis along these lines gives the results
shown in Table 2.
TABLE 2. Frequency and departures from average of synoptic patterns
favoring advanced/retarded ice wastage.
TZ:2e freguencz: and de2artures (%)
Types June Departures July-Aug. Departures
Favoring: 1969-74 1971 1972 1973 1974 1961-70 1971 1972 1973
Advanced
Wastage 32 +5 +4 -9 +12 43 +6 -6 -12
Retarded
Wastage 41 -25 +18 +2 -11 18 0 0 +6
Break-up E. 3rd 2nd
of Broughton Is. week none week
July Aug.
1974
+10
-1
2nd
week
July
The results are in line with the observed data, in at least a qualitative sense. The
1972 season is shown as being strongly "retarded," especially if June is considered;
464 Weaver/Jacobs/Barry 10

ACKNOWLEDGEMENTS
This work has been supported by a grant (GV282l8) from the National Science Foundation,
Office of Polar Programs. We are indebted to the residents of Broughton Island, N.W.T.
for their valuable assistance during the past five years. Numerous individuals have
helped our work but, in particular, we wish to mention Pauloosee and Mosesee Audlakiak
and Jaco Nukingiak who have greatly aided field operations; J. Pearson, Mrs. J. Deyell,
and T. B. Rose, cooperative observers at Broughton; the members of INSTAAR field parties;
and Mrs. M. Eccles for computer programming.
We thank the Canadian Defense Research Board and the Canadian Forces Maritime Proving
and Evaluation Unit for copies of SLAR imagery and Dr. Moira Dunbar and Major C. Chisholm
for arranging the imagery loans. We also acknowledge the continued assistance and cooperation
of the Atmospheric Environment Service and the Canada Center for Remote Sensing.
Atmospheric Environment, Canada. 1965.
practices for ice reconnaissance.
60 pp.
REFERENCES
MANICE. Manua l of sta:nda:l'd proaedures and
3rd Provisional Ed., Environment Canada, Toronto,
Barnes, J. C. and C. J. Bowley. 1973. Mapping sea ice from the Earth Resources Technology
Satellite. Aratia Bull. 1:6-13.
Barry, R. G. 1974. Further climatological studies of Baffin Island, Northwest Territories.
Inland Waters Directorate, Tech. Bull. No. 65. Environment Canada,
Ottawa, 54 pp.
Barry, R. G., R. S. Bradley and J. D. Jacobs. 1975. Synoptic climatological studies
of the Baffin Island area. In G. Weller and S. A. Bowling, eds., Climate of the
Arctic. University of Alaska, Fairbanka, pp. 82-90.
Bryan, M. L. 1975. A comparison of ERTS-l and SLAR data for the study of surface water
resources. NASA CR-ERIM 193300-59-F. Environmental Research Institute of Michigan,
Ann Arbor, 104 pp.
ERTS-l Data users handbook. Goodard Space Flight Center document. No. 71504249.
Jacobs, J. D., R. G. Barry, R. S. Bradley and R. L. Weaver. 1974. Studies of climate
and ice conditions in eastern Baffin Island, 1971-73. Inst. Arct. Alp. Res.
Occas. Pap. No.9. University of Colorado, Boulder, 78 pp.
Jacobs, J. D., R. G. Barry and R. L. Weaver. 1975. Fast ice characteristics with
special reference to the eastern Canadian Arctic. Polar Reaord 17:521-536.
Langleben, M. P. 1966. On the factors affecting the rate of ablation of sea ice. Can.
J. Earth Sai. 3:431-439.
Lieske, B. J. 1964.
Barrow, Alaska.
Net radiation over fast sea ice during spring break-up at Pt.
Proc. 15th Alaskan Science <strong>Conference</strong>. 52-60.
World Meteorological Organization. 1970. WMO sea-iae nomencLature. W.M.O. No. 259,
TP. 145, World Meteorological Organization, Geneva, 147 pp.
Zubov, N. N. 1945. L'ydy Arktiki. Izdat. Glavsevmorputi, Moscow (transl. Arctic Sea
Ice, U.S. Navy Oceanogr. Office, 1963, 491 pp.).
466 Weaver/Jacobs/Barry 12

SECTION 7
ICE DYNAMICS
Colony, R.
The SimuZation of Aratia Sea Iae Dynamias
(University of Washington, United States)
Heiberg, A.
Aidjex FieZd Operations to August 1975
(University of Washington, United States)
McPhee, M.
Water Stress SUb-Mode Z for the AIDJEX Mode Z
(University of Washington, United States)
Maser, K. R.
A MeahaniaaZ ModeZ for the Deformation of Aratia Paak Iae
(Foster-Miller Associates, Incorporated, Massachusetts, United States)
Pritchard, R. S. and R. T. Schwaegler
AppZiaations of the Aidjex Iae ModeZ
(University of Washington, United States, and Seattle University,
Washington, United States)
Stringer, W. J. and S. A. Barrett
Iae Motion in the Viainity of a Grounded FZoeberg
(University of Alaska, United States)
467

THE SIMULATION OF ARCTIC SEA ICE DYNAMICS
Roger Colony
AIDJEX
University of Washington
Seattle, Washington
United States
ABSTRACT
A mathematical model of large-scale sea ice dynamics has been formulated that provides a
theoretical background for the AIDJEX 1975-76 experiment in the Beaufort Sea. GOverning
differential equations are developed to predict ice motion, the distribution of ice
thicknesses, and internal ice stress.
The motion is induced by synoptic-scale air and ocean flow in the Beaufort Sea during the
1975-76 experiment. The data from a network of drifting data buoys and manned drifting
ice stations are used in the model formulation and in the verification of analysis. The
finite difference approximations of the differential equations and the computational
algorithm is outlined. Results of model calculations will be compared with data acquired
by airborne and satellite-based sensors.
469

Large-Scale Sea Ice Dynamics
INTRODUCTION
The objective of AIDJEX is to reach, through coordinated fieUi experiments and theoretical
analyses, a fundamental understanding of the dynamic and thermodynamic interaction
between arctic sea ice and its environment (Maykut, Thorndike and Untersteiner, 1972).
One of the principal scientific objectives of AIDJEX is to represent large-scale sea ice
deformation and the mechanics of this deformation. The temporal and spatial distribution
of large-scale properties of sea ice is little understood, yet its importance to global
climate models and shipping forecasts is manifest. The large-scale strain appears to be
the key to heat transfer and momentum exchange in the air/ocean/ice system and to the
internal stress state of the ice. It has been observed that the large-scale motion of
sea ice takes the characteristic lengths and times of synoptic weather patterns; that is,
about 100 kilometers and typical times of a day or less.
Both the small and mesoscale dynamics such as floe-to-floe interaction, buckling, cracking,
ridging, rafting, and lead formation are characterized by very discontinuous behavior
with a complicated pattern of strain rate, and stress distribution. The large-scale
properties arenhomogenization of the small and mesoscale behavior; and, consequently, a
continuum model is thought to govern velocity, stress, strain, ice thickness distribution,
yield strength, and heat transfer. Although the continuum model does not simulate such
features as individual pressure ridges, leads, and multiyear floes, it does reflect these
mechanisms in a spatially averaged sense. Indead, much of the "physics" in the largescale
ice model is strongly rooted in the physical processes thought to govern the
mesoscale dynamics.
Beaufort Sea Experiment
The AIDJEX 1975-76 field experiment is designed to take measurements necessary for both
the formulation of the modeling problem and for the verification of calculations from the
model (Untersteiner, 1974). The physical configuration of the experiment consists of
four manned drifting ice stations surrounded by an outer ring of automatic data buoys.
Figure 1 illustrates the station positions and primary data analysis area on May IS, 1975.
The relative position of the buoys provides a boundary for the primary modeling area. The
manned camps, data buoys and primary modeling area are in constant motion and deformation.
The position of the manned camps and data buoys is monitored using the Navy Navigational
Satellite System. The data buoys receive satellite data and record local atmospheric
conditions. The data is then relayed to a manned camp via a radio link. A time series
of position and local atmospheric conditions is then prepared. The absolute position has
a typical error of 100 meters; however, the position error relative to another receiver
is only about 10 meters. The manned stations receive about 35 position fixes per day.
The large-scale kinematics can be determined from the motion of the buoys and manned
camps. The motion and deformation of the primary modeling area is also measured by sate 1lite
photography. A sequence of LANDSAT photographs allows a number of distinguishable
features to be tracked. The displacements of the features provide another independent
measure of the kinematics. Aerial photography and other satellite imagery are used to
directly measure large-scale deformations.
Boundary Layer Studies
The principal motive force for sea ice is the synoptic-scale surface winds. The modeling
of sea ice dynamics will focus on the relation of the ice motion-to-the-surface traction
exerted by the winds. The surface atmospheric pressure from the data buoys, manned camps,
and other sources is combined with the northern hemisphere surface pressure map provided
470 Colony 2

y the U.S. National Weather Service. An enhanced surface pressure analysis local to the
primary modeling area is then constructed. Surface geostrophic winds are determined from
the gradient of the surface pressure.
The surface traction, hereafter referred to as the "air stress," is related to the surface
geostrophic winds through a theory of the planetary boundary layer. The field experiment
also measures the vertical structure of the lower portion of the atmospheric boundary
layer. Experimentally determined drag coeffiCients, together with the geostrophic winds,
then are used to calculate the air stress throughout the primary modeling area. In addition,
direct measurements of vertical momentum fluxes are obtained from aircraft flights
over a large portion of the modeling area. The accurate measurement and representation
of the distribution of air stress is perhaps the most difficult task of the field experiment.
o
CANADA
Figure 1. The area (gridded) of the Beaufort Sea in
the Arctic Ocean in which data was collected
during AIDJEX. On 15 May, 1975, data buoys
were located at the dots and manned camps
at the triangles.
471 Colony 3

The frictional coupling between the sea ice and the ocean is related to the differential
motion of the sea ice and the upper-level ocean currents. The ocean currents are obtained
from a long-term record of measurements. The motion of transient eddies of various sizes
has not been documented well enough to be used directly in the simulation. The field
experiment provides direct measurements of the complete oceanic boundary layer at the
manned camps. These vertical velocity and density profiles provide the drag coefficjent
needed to relate the surface traction, referred to as the "water stress," to the oceanic
flow. The vertical flux of heat and salt is also monitored and is related to the thermodynamic
growth of sea ice.
The Modeling of Sea Ice
The AIDJEX ice model is embodied in a set of partial differential equations. The continuum
model of sea ice is governed by a conservation law for momentum, a conservation
law for the distribution of thicknesses of ice in a given area, and a constitutive law
relating stress to strain and strain rate. The sea ice is modeled as a two-dimensional
platelike body of an isotropic elastic-plastic material. The important energetics associated
with the veritcal direction have been parameterized in terms of an independent
thickness coordinate. The modeling of sea ice will predict horizontal motions, strain
rates, the distribution of thicknesses of ice, stress state, and individual forces entering
into the momentum equation.
The basic mode of the simulation of sea ice dynamics is to observe the motion of the
drifting ice stations and choose some event, such as a storm or an unusual deformation,
to simulate. The time interval for the simulation is chosen and the initial conditions
are determined; the initial conditions are velocity, stress, and ice thickness distribution.
The ring of buoys defines the boundary of the region in which the simulation is
made (Colony, 1975). Boundary conditions are prescribed and the air stress field is
prepared from observed weather patterns. A digital computer is then used to calculate
the approximate solution to the governing differential equations (Pritchard and Colony,
1976). The simulation of the event is compared to actual observations of the event
(Coon et aZ., 1976). This modeling will allow us to isolate individual forces active on
the pack ice and will determine the stress state of the ice. It is noted that the model
does not "forecast" the ice motion and state because the driving forces are not forecasted.
DIFFERENTIAL EQUATIONS
On the large scale, sea ice is very thin, about 3 meters, compared to the horizontal
extent, about 100 kilometers; therefore, we consider a plane problem. Neither the plane
strain nor the plane stress problem is appropriate. Instead, the vertical velocity,
stress, and strain are somewhat ignored, similar to that of the shell theory.
Momentum Equation
The two-dimensional balance of momentum for a differential element of sea ice considers
the effect of internal ice stress (actually the in-plane stress resultant), air and water
stress, Coriolis force, and sea surface tilt.
To analyze sea ice motions within geometries that encompass only a portion of the Arctic
Basin, it is adequate to neglect the earth's curvature and describe the response in Cartesian
coordinates (x,y) In the plane of motion, momentum balance is expressed as
D
DT nt Q div Q" + J a + !w - mfc !s x Q + m§ grad H (1)
472 Colony 4

The particular yield curve now used by the AIDJEX ice model is
and is shown in Figure 2. Stresses are restricted to have no positive principal values
because the material is densely fractured and will support no stress in any opening mode.
The limiting strength p* is from Parmerter and Coon (1972) and involves the energetics
of ridge building.
plastic strain rate
PLASTIC J
ELASTIC
second principal
stress
first principal
stress
Figure 2. The yield curve now used in the AIDJEX ice model. The material
response is elastic if the stress state lies within the curve,
plastic for stresses on the curve. The dashed curve shows the
yield constraint for a slightly smaller yield strength p*.
The hardening parameter p* may increase or decrease (shown as dashed curve in Figure 2)
in time. The yield curve now used simply expands or contracts with the variation in p*.
The weakening, decrease in p*, may imply an instability of the material in the sense of
Drucker (1950); however, no computational difficulties have been found. This instability
may be suppressed by spatial and temporal variation of the air stress and the restraining
effect of the water stress.
The plastic flow rule is obtained by maximizing plastic work for a given deformation.
This normal flow rule [Coon et aZ. (1974), Pritchard (1975b)1 is
A > 0 (8)
where e P is the plastic stretching and A is a positive scalar. The plastic flow is further
decomposed into isotropic and deviatoric parts. Figure 2.shows.the plastic stretching
to be orthogonal to the loading function at the instantaneous stress state.
475 Colony 7

ecause p* typically varies much more slowly than Q. The average thickness, n, of a differential
element is
h= ('Gdh
The average mass per unit area of the differential element is then m = ph where p denotes
the ice density. The mass per unit area is associated with the stress node (or ice
thickness distribution node). In the later calculation of the momentum balance the mass
associated with a velocity node is needed. To satisfy this requirement the mass at the
velocity node is obtained from interpolation of the masses at the adjacent stress nodes.
Step 6 (stress divepgenae)
The average ice stress divergence in R is obtained from the values of Q on L, Figure 3.
The Green-Gauss theorem again gives
r!;n d Z (19)
The values of Q along L are given at the midpoints of the sides of the quadrilateral R
rather than at the vertices as in the case of the velocity gradient. However, the same
linearity assumption is used in evaluating the line integral. The divergence of ice
stress is used in the integration of the momentum equation.
Step 7 (momentum equation)
n_ll n+ll _
The difference equations for the momentum equation in the interval (t 2, t 2) X R can
be derived by formal integration
i
R
m r da dt
As R follows the material the spatial integration and the time differentiation may be
interchanged and the acceleration term becomes
m (21)
The average mass per unit area m in the region R is evaluated at time t n by interpolation
from neighboring stress nodes.
The integral of the divergence of ice stress becomes
482 Colony 14

A more complete description of the integration of the momentum equation is given by
Pritchard (1975) and Pritchard and Colony (1976).
Stability and Results
The system of difference equations satisfy consistency or convergence requirements and
have truncation errors which are generally first order. The mathematical stability of
the system of equations is conditional, however. The relation between the allowable time
step fit and a norm of the mesh size 6x
where C 2
This is the usual Courant condition characteristic of explicit difference equations arising
from wave propagation problems. The existence of elastic waves in large-scale sea
ice is doubtful; however, the elastic strain rate is a necessary part of the kinematic
relationship. In the main, the difference equations are robust provided the stability
criteria is not violated.
RESULTS AND VERIFICATIONS
Deployment of the drifting ice stations began in early March 1975. The complete scientific
experiment, including the data buoys, began about the first of May 1975. The
eleven-day period beginning on 15 May 0000 hours Greenwich Mean Time was chosen for the
first modeling exercise. Coon et al. (1976) has reported on a portion of the calculations
for that modeling period. It is the intent of the AIDJEX modeling group to simulate
the sea ice dynamics for several time periods from May 1975 to April 1976. Model calculation
reports containing detailed input and output quantities will appear in the AIDJEX
Bulletin (University of Washington, Seattle, Washington, USA).
No results of calculations will be included in this paper. Rather, a general description
of the available calculations and the means of their verification will be given here.
Calculated Quantities
The position, velocity, and strain rate fields are calculated. This description of the
large-scale kinematics is subject to direct verification from many sources. Each of the
manned drifting stations interior to the modeling region is equipped with a Navy Navigation
Satellite System receiver. The time series of the measured position and velocity
of each manned station is directly compared with the time series of position and velocity
from the calculations. The strain rate tensor is also measured from the relative motion
of the manned stations. This measured strain rate is compared to the calculations. The
processing of LANDSAT imagery and aerial photography is used in the measurement of the
large-scale displacement field over a major portion of the modeling area (Nye, 1975).
Spatial patterns of velocity and strain rate are compared to calculated results. The
pattern of strain is also revealed in the creation of leads which are recorded by remote
imagery. The calculations of the relative amount of area of these leads is an important
direct verification of calculated results. Coon et al. (1976) shows these direct comparisons
between the calculation of position, velocity and strain rate to measured values.
The ice thickness distribution model will be "tuned" to reflect experimental lIleaSJUements.
The thermodynamic growth of ice will be measured and correlated against surface air temperature,
snow cover, and the surface radiation balance. The mechanical redistribution
of the thicknesses of sea ice will be compared with remote imagery and ridge measurements.
An energetics argument by Rothrock and Hall (1975) also allows the shape of the yield
curve to be determined from LANDSAT imagery.
(26)
484 Colony 16

Lead orientation is not predicted by the ice model. It is anticipated, however, that the
lead orientation will be correlated with the principal stresses and the angle of the
principal stress. Note that the axis of principal stress and plastic strain rate are
aligned. The tractional forces due to air and water stress are the subject of a major
experimental effort. The forces and loads must be accurately represented before the value
of the ice model can be fully assessed. Measured force balances at the manned drifting
stations are compared to the various terms calculated from the momentum equation.
The large scale ice stress is not a readily defined property of sea ice and it almost
defies measurement. The primary use of large-scale ice stress in the sea ice model is
for the calculation of the forces due to divergence of ice stress. While the divergence
of ice stress was not large for the modeling period 15-25 May, there will be many cases
when it will dominate the response. This will be especially true for near-shore studies.
The relation between the large-scale continuum ice stress and the local stress experienced
on smaller spatial scales has been studied by Evans and Rothrock (1975) and Maser
(1976). The application of large-scale stress to the formulation of a structures problem
is not complete. It is certain that the local mesoscale ice stress can be much greater
than the large-scale ice stress.
The application of the calculations to ice forecasts or global climate models has not been
investigated. Perhaps more details such as anistropic behavior is needed, or perhaps less
is needed. However, one of the most valuable results of this sea ice dynamics simulation
will be a clarification of appropriate length and time scales.
ACKNOWLEDGMENTS
The author wishes to thank all of his colleagues in the AIDJEX modeling group. The formulation
of a large part of the sea ice model has been due to the efforts of Max Coon,
Gary Maykut, Robert Pritchard, Drew Rothrock and Alan Thorndike. The structure of the
computer program for the modeling of sea ice dynamics is due to Don Thomas. Particular
thanks are due Carolynn Padgett for the diligent typing. This work was supported by the
National Science Foundation Grant OPP7l-0903l, formerly GV 28807, to the University of
Washington.
REFERENCES
Brown, R. A. 1974. Analytiaal methodS in planetary boundary-layer modeling. John Wiley
and Sons, Inc., New York, p. 66.
Colony, R. 1975. The boundary determination problem for the AIDJEX Ocean. AIDJEX
Bulletin. 28:87-97.
Colony, R. and Pirtchard, R. S. 1975a. Integration of elastic-plastic constitutive
laws. AIDJEX Bulletin. 30:55-80.
Colony, R. and Pritchard. R. S. 1975b. Integration scheme for an elastic-plastic sea
ice model. In Proceedings of 12th Annual Meeting of the Society of Engineering
Science. The University of Texas at Austin, Texas, pp. 953-963.
Coon, M. D., Maykut, G. A., Pritchard, R. S. and Rothrock, D. A.
pack ice as an elastic-plastic material. AIDJEX Bulletin.
1974. Modeling the
24:1-105.
Coon, M. D., Colony, R., Pritchard, R. S. and Rothrock, D. A. 1976. Calculations to
test a pack ice model. In Proceedings of 2nd <strong>International</strong> <strong>Conference</strong> on Numerical
Methods in Geomechanics. Virginia Polytechnic Institute and State University at
Blacksburg, Virginia, in press.
485 Colony 17

Drucker, D. C. 1950. Some implications of work hardening and ideal plasticity. Quarterly
of Applied Mathematics. 7:411-418.
Evans, R. J. and Rothrock, D. A. 1975. Stress fields in pack ice. In Proceedings 3rd
<strong>International</strong> Symposium on Ice Problems. U.S. Army Cold Regions Research and
Engineering Laboratory at Hanover, New Hampshire, in press.
Maser, K. R. 1976. A mechanical model for the deformation of arctic pack ice. Published
elsewhere in these proceedings.
Maykut, G. A., Thorndike, A. S. and Untersteiner, N. 1972. AIDJEX scientific plan.
AIDJEX Bulletin. 15:1-67.
Maykut, G. A. and Untersteiner, N. 1971. Some results from a time dependent, thermodynamic
model of sea ice. Journal of Geophysical Research. 76,6:1550-1575.
McPhee, M. G. 1976. Water stress sub-model for the AIDJEX model. These proceedings.
Nye, J. F. 1975. Suggested procedure for observing ice displacement, strain, and thickness
distribution during the AIDJEX main experiment. AIDJEX Bulletin. 28:127-140.
Parmerter, R. R. and Coon, M. D. 1972. Model of pressure ridge formation in sea ice.
Journal of Geophysical Research. 77,33:6565-6575.
Pritchard, R. S. 1974. Elastic strain in the AIDJEX sea ice model. AIDJEX Bulletin.
27:45-62.
Pritchard, R. S. 1975a. A difference approximation to the momentum equation. AIDJEX
Bulletin. 30:81-93.
Pritchard, R. S. 1975b.
Applied Mechanics.
An elastic-plastic constitutive law for sea ice.
42E,2:379-384.
Journal of
Pritchard, R. S. and Colony, R. 1976. A difference scheme for the AIDJEX sea ice model.
In Proceedings of the 2nd <strong>International</strong> <strong>Conference</strong> on Numerical Methods in Geomechanics.
Virginia Polytechnic Institute and State University at Blacksburg,
Virginia, in press.
Rothrock, D. A. 1975. The energetics of the plastic deformation of pack ice by ridging.
Journal of Geophysical Research. 80,33:4514-4519.
Rothrock, D. A. and Hall, R. T. 1975. Testing the redistribution of sea ice thickness
from ERTS photographs. AIDJEX Bulletin. 29:1-19.
Thorne, B. J. and Herrmann, W. 1967. TOODY--A computer program for calculating problems
of motion in two dimensions. Sandia Laboratories, New Mexico. SC-RR-66-6l2.
Thorndike, A. S., Rothrock, D. A. Maykut, G. A. and Colony, R. 1975. The thickness distribution
of sea ice. Journal of Geophysical Research. 80,33:4501-4513.
Untersteiner, N. 1974. Arctic ice dynamics joint experiment. Arctic Bulletin.
1,4:145-159.
486 Colony 18

BACKGROUND
The objective of AIDJEX is to reach an understanding of the dynamic and thermodynamic
interaction between sea ice and its environment. The central purpose of AIDJEX has
been to develop a predictive numerical model of the sea ice which will describe its
behavior based on given inputs from the environment.
The concept of AIDJEX evolved over a number of years from research conducted by u.s.
and Canadian scientists interested in Arctic problems. Special scientific aspects of
the program will be well covered in three other AIDJEX papers this morning. The plan
as initially advanced in 1969 called for a series of pilot studies designed to resolve
scientific questions and to develop necessary technology. These studies were to be
followed by the main experiment, a year-long occupation of an array of manned stations
surrounded by a ring of automatic data buoys.
Four pilot studies preceded the main experiment. The first two, in March-April 1970
and 1971, were operated jointly with the Canadian Polar Continental Shelf Project
at their Camp 200 west of Banks Island. Both concentrated mainly on gathering oceanographic
data and testing the relative contribution of skin friction and form drag to
the total momentum exchange between ice and water.
From 25 February to 29 April 1972, the third and most ambitious pilot study took place
about 300 miles north of Barrow, at a triangle of three manned stations 75 km apart
surrounded by a ring of five data buoys on a radius of 400 km. Eighty scientists and
support personnel from the U.S., Canada, and Japan participated in this study, which
was designed to (a) measure the three-dimensional velocity and density structure in
the ocean, (b) test procedures for obtaining regional values of wind stress, and
(c) gather high resolution strain data on the same Bcales as would be used in the
main experiment.
In the spring of 1974, the fourth pilot study, a lead experiment on the sea ice north
of Barrow, acquired data on the modification of the atmospheric and oceanic boundary
layers near an open lead. This investigation concluded the preliminary field work
for the main experiment.
THE MAIN EXPERIMENT
After a funding cut for the main experiment was announced in 1973, AIDJEX devoted
much of its time to planning a "bare-bones" experiment, one that would yield the
information necessary to reach its objectives and yet remain within given budgetary
limits. What evolved was a four-station manned array surrounded by a ring of at
least eight buoys defining the boundary of the region to be studied. This array
is shown in Figure I, and the minimum observational program deemed necessary both to
test and to drive the theoretical model is summarized in Table 1. Then in March 1975,
with more than a year of planning behind us and a detailed operations plan in front
of us, we began to deploy the main experiment.
(The text that follows is the oral part of a slide presentation.)
The search for a site for the main camp began in late February with the PCSP Twin Otter
out of Mould Bay, joined a week later by the NARL R4D from Barrow. All searches were
restricted to an area north of 76°N and west of l40oW, far enough from the south and
east coasts that the anticipated drift of the buoys would not take them into shorefast
ice, yet close enough that the camp could be reached by supply aircraft. Within that
target area, we were looking for a large multiyear floe adjacent to a smooth frozen
lead sufficiently long and thick to support C-130 landings during the major airlifts.
488 Heiberg 2

..
o
o
q
o
o
Figure 7. AIDJEX array at completion,
in June 1975.
492
. ,
I den ti tiers
Squares: manned camps
Circles: NavSat buoys
Dots: RAMS buoys
Heiberg 6

Figure 8. Main camp in the summer . The photo, taken
during an airdrop flight, shows the fog and melt
ponds that are major obstacles to Arctic flying
in the warm season.
\ .' ,
•
•
Figure 9. Deformation of the array, shown by triangulation
of the three satellite camps.
493
Heiberg 7

WATER STRESS SUB-MODEL FOR THE AIDJEX MODEL
Miles McPhee
Arctic Ice Dynamics Joint Experiment
University of Washington
Seattle, Washington
United States
ABSTRACT
The AIDJEX numerical ice model expresses the stress between ice and ocean as a quadratic
drag law of the form Ti w = pcwG bijGj, where Ti w is the horizontal stress vector, Gj is
the relative velocity vector between ice and ocean, bi1 is a cosine matrix involving the
boundary layer turning angle, S, and C w is a dimensionless drag coefficient. In general,
S and C w may depend on other parameters in the model such as ice thickness, ice speed,
and season of the year. Values used in preliminary model runs are those determined from
the 1972 AIDJEX experiment, namely, C w = 0.0034 and S = 24°. Subsequent runs will make
use of oceanographic measurements from the 1975-76 field experiment. Various techniques
for estimating stress from mean current measurements arc discussed, and a method is introduced
that invokes turbulent boundary layer theory to estimate the shape of the average
profile. The method was tested on 1972 data and results are shown.
495

to ensure that all current meters were turning above threshold velocity. One-hour averages
were used for all profile calculations. For the fixed momentum integral method
(section 3.2), the triplet at 12 m was chosen for a reference since this was the deepest
triplet consistently in the well-mixed layer. For the generalized method (section 3.3),
a friction velocity coefficient, c 2 = u*IU(2) , was determined by considering u* as calculated
from turbulence measurements [MSj; the result was c 2 = 0.05.
Figure 2a shows the time series of frictional-influence depth H = 0.45 u*lf= 0.023 U(2)lf.
for the 36-hour period beginning at 1200 (AST) on all April 1972. In Figure 2b, the
reference speed calculated at H is shown with the measured speed at 32 m. Linear interpolation
was used between measurements at fixed levels of 20, 26, 32, and 38 m. The
relative angles between the streamlines at 2 m and 32 m and between 2 m and H are shown
in Figure 2c. The program calculates the magnitude of stress by (1) integrating Mx and
My by trapezoidal rule to 3Z m--this is referred to as the momentum integral method--and
by (2) performing the calculation suggested in section 3.3:
pyc 2 U(2) G sin B
with P = 1.0 gm cm- 3 and y = 0.2. The hourly stress estimates by both methods are plotted
in Figure 3a, with the corresponding drag coefficients C w = ITO I IpG 2 plotted in Figure 3b.
The larger value for the momentum integral method reflects t:,e "bulge" in the lateral profile
discussed in [MSj. The drag coefficient value of 3.4 is that determined from
composite turbulence measurements made during the last third of the period shown [MSj.
The comparative consistency of the generalized drag coefficient over a fairly large range
of ice speeds is encouraging.
N
I
E
o
"'Q
><
•
U
Figure
the
10
5
12 24 36 HOURS
0 0 12 24 36 HOURS
Momentum integral method ( 32 m reference)
Generalized method ITI = yu GSln/3
3. Water stress (a) and drag coefficient, c (b) for
36-hour period.
w
503 McPhee 9

Hunkins, K. 1974b. The oceanic boundary layer and ice-water stress during AIDJEX 1972.
AIDJEX Bulletin, No. 26 (September):109-l28.
Hunkins, K. 1975a. The oceanographic program for the AIDJEX main experiment. AIDJEX
Bulletin, No. 28 (March):48-60.
Hunkins, K. 1975b. Geostrophic drag coefficients for resistance between pack ice and
ocean. AIDJEX Bulletin, No. 28 (March):6l-68.
Leavitt, E. 1975. Determination of air stress from AIDJEX surface layer data. AIDJEX
Bulletin, No. 28 (March):11-20.
McPhee, M. G. 1975. An experimental investigation of the boundary layer under pack ice.
Technical Report Ref. M75-l4, Department of Oceanography, University of Washington,
Seattle, Washington, 164 pp.
McPhee, M. G., and J. D. Smith. 1975. Measurements of the turbulent boundary layer
under pack ice. AIDJEX Bulletin, No. 29 (July):49-92.
Newton, J. L.
features.
1973. The Canada Basin: mean circulation and intermediate-scale flow
Ph.D. dissertation, University of Washington, Seattle, Washington.
Rigby, F. 1974. Theoretical calculations of internal wave drag on sea ice. AIDJEX
Bulletin, No. 26 (September):126-l40.
Tennekes, H. 1973. The logarithmic wind profile. Journal of Atmospheric Sciences,
30:234-239.
508 McPhee 14

2. Opening of pre-existing cracks: The solid floes are assumed to be laced with
cracks which have arisen from out-of-plane sources, such as isostatic imbalance
in the ice sheet, or water waves.
3. Rigid body motion of solid floes: In the context of the scale of the Arctic ice
mass, these features are analogous to micromechanisms, much like dislocations
in metals. What AIDJEX has done is to develop a model for the macroscale in
which the overall deformation is a gross manifestation of these mechanisms.
Naturally, a number of assumptions have been made in order to achieve this end.
Some of these involve the way in which the ice thickness is redistributed as a
result of overall strain. Others involve the contributions to the overall work
done during deformation.
The purpose of the work reported herein was to evaluate the accuracy and consequences
of these assumptions by developing an independent technique. This technique
would compute the overall behavior of the aggregate material (pack ice)
by building from the basic deformation mechanisms of pressure ridging, crack
opening, and rigid body motion of floes.
The Computer Simulation
The technique selected for this purpose was the finite element technique. This is a
natural approaah, since the finite element technique itself involves a synthesis of overall
stress-strain behavior by assembling many small elements. It was also felt that the
particular characterizations for pressure ridging, cracking, and rigid body motions
would be available in the current state of the art in finite elements. This is, in
fact, the case.
The most relevant finite element developments for this application have come in the area
of rock mechanics. Rock masses tend to have joints and foliations as well as strata or
veins of material with different properties. A number of computer codes have been
written for the analysis of stresses and deformations around underground openings, taking
into account some of these peculiar features. Some of the relevant modelling capabilities
of the programs are:
l. Plasticity
Mohr-Coulomb
Von Mises with strain hardening
2. Joint Elements
3. Crack Development Analysis
No-Tension analysis
Craak propagation analysis
From amongst the above modelling capabilities we were able to find the basic features
needed to model pack ice. Either plasticity model is capable of simulating the constant
load level associated with lead deformation. Joint elements are capable of simulating
the opening, closing and sliding of predefined cracks. For cracks which are not specifically
predefined, crack development analyses have been developed. These techniques
assume that crack formation is isotropic. The simplest of these is the no-tension
approach. In this analysis, the material is assumed to be inaapable of supporting any
principal tensions. Principal tensions are eliminated by superimposing self equilibrating
stress fields. A more detailed procedure identifies the aatual crack propagation
development. Elements which have principal tensions above a threshold value are redefined
as orthotropic (cracked), and the analysis iterates until a final, stable crack pattern
is developed.
510 Maser 2

APPLICATIONS OF THE AIDJEX ICE MODEL
Robert S. Pritchard
AIDJEX
University of Washington
Seattle, Washington
United States
Richard T. Scbwaegler
Department of Civil Engineering
Seattle University
Seattle, Washington
United States
ABSTRACT
The AIDJEX model. is presented for the first time uri-th the el.astia-pl.astia l.ClIJ1, iae redistribution
funation and energetias argument simul.taneousl.y disaUBsed. Numeriaal. sol.utions
to the model. are obtained for two different probl.ems. The geometry ahosen is intended to
simul.ate aonditions near a aoastal. boundn.ry uri-th uri-nds bz.ounng either on-shore 01' al.ongshore.
Computed resul.ts indiaate spreading of a ridging region at the boundary CJaUsed by
aompression aaaompanying the on-shore uri-nds. HOIJJever, bJhen an al.ong-shore uri-nd oaaurs,
the vel.oeity differenaes aonaentrate at the shore beaause the shearing deformation induaes
softening of the material. model.. These resul.ts satisfy physiaal. intuition and proaessed
data from the LANDSAT sateUite. The resuz.ts suggest that the AIDJEX model. rruy be used
to simul.ate l.arge-sool.e deformations of Aratia sea iae in both the aentral. Beaufort Sea
(site of the AIDJEX main experiment) and in the near-shore marginal. iae zone.
513

INTRODUCTION
The AIDJEX model is designed to simulate the large-scale response of sea ice to driving
forces in the atmosphere and ocean. The focus of this effort is the simulation of
conditions observed during the AIDJEX main experiment during 1975-76 [Untersteiner, 1974,
Maykut, Thorndike and Untersteiner, 1972].
The mechanisms that affect the macro scale (tens of kilometers) response of the sea ice
are the opening of leads between ice floes, the subsequent freezing of open water and
thin ice in the leads and piling up of the thin ice into ridges as closing occurs. The
individual micro scale processes are understood [Maykut and Untersteiner, 1971, Parmerter
and Coon, 1972, 1973]. The effect of many of these events occurring simultaneously in a
large-scale element has been described in the AIDJEX ice model [Coon, et al.,1974].
The large space scales required in the modeling of sea ice do not permit us to use
laboratory tests to determine how sea ice deforms when it is loaded by in-plane forces.
Instead, the response must be inferred by understanding the microscale processes and
the effect of simultaneous interaction of many of them. Testing of the model then is
accomplished by simulating actual conditions observed during field experiments. However,
in field experiments the applied forces cannot be controlled and furthermore, cannot
be measured directly. Instead, it is necessary to also provide a model that relates the
forces exerted on the ice to variables that can be measured. In the development of a
viable sea ice model this means that the atmosphere and ocean must be modeled and
described just as well as the ice. These three systems are considered together in the
AIDJEX model and this consideration is reflected in the parameterizations that have
been chosen to describe the dynamic and thermodynamic response of sea ice. In this
paper we concentrate on understanding the sea ice response but must include an oceanic
boundary layer model to generate realistic test conditions (an atmospheric boundary
layer model is also required but it is not actively used for the present study).
The deformation of an ice element due to in-plane forces is described by an elasticplastic
material model [Coon, et al., 1974, Pritchard, 1974, 1975]. The elastic response
describes the small amount of strain induced by deforming ice floes and the plastic
response (larger and much more important) describes the relative motion of floes,
specifically the formation of leads and ridges. The strength at which flow occurs
depends on the amount and thickness of thin ice (and open water) present within an
element. To provide this information we parameterize the vertical spatial coordinate in
terms of an independent thickness coordinate. Then the ice thickness distribution is
used to measure the fraction of area covered by ice of each thickness [Coon, et al., 1974,
Thorndike, et al., 1975].
We have stated that the AIDJEX model was developed with the primary purpose of simulating
the ice response in the central Beaufort Sea (the location of the AIDJEX main experiment).
Therefore, we must justify the present applications which are intended to represent
conditions in the near-shore regions of the Arctic. We do this by noting that the same
physical mechanisms of lead and ridge formation account for deformation for the nearshore
ice when the length scales remain on the order of tens of kilometers. Finally,
the purpose of this paper is limited to understanding how ice responds to various
driving forces and so we are not concerned with simulating actual conditions. Therefore,
there is no need to account for local conditions that affect the atmospheric or oceanic
boundary layers and we ignore the effect of bottom topography and tides.
Perhaps the most important difference between the central Arctic and the near-shore
regions is the rate at which deformations occur. Strain rates on the order of 1% per
day have been reported by Thorndike [1974] during the 1972 AIDJEX pilot study. These
are average values over a 100 km length scale. Similarly, Hibler, et al. ,[1974a] have
514 Pritchard/Schwaegler 2

compression at point 3 in Figure 3. The original 40 km fetch has increased, and the
water drag stress has decreased. The combined effect of ridging in regions A and C is
sufficient to continue creating open water at the right boundary of region C in spite
of the thermodynamics involved. As a result, the stress state remains at zero at the
right boundary of this region while growing rapidly at the left edge. In summary, the
material response in region C is similar to that discussed previously in region A.
Because of the significant amounts of thin ice present in this region, thermodynamics
plays a much more important role here than in the other two regions.
Along-shore Wind
For this case we apply a constant wind stress of 0.5 dynes/cm 2 in the positive ydirection
(orthogonal to the cross-section viewed in Fig. 1). The material response is
strikingly different from that encountered with the on-shore wind. Immediately upon
loading, material at the left-hand edge fails in a state of plastic shear flow while
the remaining region remains elastic. Figure 3 shows the state of stress at point 4
in the failing cell shortly after loading begins. This state corresponds to pure shear
deformation where the compressive strength of the material weakens. Figure 8 shows
this compressive strength, p*, as a function of tioe.
With all the deformation concentrated in the first cell (called region A), the elastic
material (region B) moves essentially as a rigid body along the shore. The material
at the right-hand edge (region C made up initially of open water) undergoes the same
response as region A for along-shore motions. As the shore-ward cell continues to
weaken, the velocity of the elastic region increases. The components of this "rigid
body" velocity are plotted as a function of time in Figure 9. From this curve we
observe that the along-shore velocity attained during the first half day is very nearly
wind driven. The additional weakening at the left-hand edge after this time can therefore
have very little effect of the kinematic response of the basin. Hence, we conclude
that a quaSi-steady state has been reached after approximately one-half day of loading.
The region of plastic deformation does not expand as it did in the on-shore loading
case, but instead is concentrated to form a discontinuity at the shore. The velocity
profiles shown in Figure 10 reveal the geometry of this discontinuity. Several
mechanisms act simultaneously in region A to cause the complex response. These include
a combination of ridging and leading with simultaneous thermodynamic growth on open
water and thin ice.
To understand the model response in region A it is helpful to review some of the
characteristics of our tear-drop yield surface. When a stress-free element is subjected
to pure shear deformation two limiting behavior patterns are possible. Consider the
initial stress state to be located at point 1 in Figure 3. If the element is totally
unconfined, it will separate into a loosely knit group of fragments with a high
percentage of leads present. An increase in the area will occur with no change in the
internal stress state. The strain rate vector associated with this expansion will
combine with the shear component to produce a total strain rate vector which is
orthogonal to the yield curve at the origin. Note that this response is independent
of the initial ice compressive strength or thickness distribution. The other limiting
behavior is observed for an element which is highly confined to prevent expansion. When
pure shear deformation is applied to this element it is accompanied by a rapid increase
in both internal pressure and shear stress. Under this condition the stress state will
move rapidly to point 4 along the yield curve and the strain rate vector will retain
its orthogonality relationship to the yield surface. Unlike the previous condition,
this response is a function of the initial ice strength and thickness distribution.
In our test problem the material response falls somewhere between these two extremes.
At first the inertia associated with the large elastic zone provides sufficient confinement
to permit a rapid increase in the internal pressure. This build-up accelerates
the elastic zone off shore to a divergent velocity of 0.9 cm/sec within the first half
522 Pritchard/Schwaegler 10

The model response may be interpreted in terms of pressure ridge formation being
analogous to plastic flow in compression and also of shear ridging being analogous to
plastic flow in shearing. The former expands in space while the latter concentrates into
a narrow region (approximated by the model as a discontinuity in the velocity field)·
Additional calculations using winds oriented at different angles indicate that a
combination of effects occurs that is qualitatively like the on-shore and along-shore
cases. For all of them, we see a tangetial flow caused by the along-shore component,
but we also see the on-shore component produce hardening. The width of this plastic
region expands as in the on-shore case. Simultaneously, however, the lateral component
causes flow that reaches a quasi-steady state.
In all cases, we find a wide variety of strain rates appearing within the same problem
at different locations. We see that the AIDJEX model simulates these largely different
deformation rates in a proper manner. In regions where the ice remains elastic,
deformation is essentially zero with rigid body motion as the dominant mode; however,
in nearby cells, when plastic flow occurs, the deformation rate becomes significantly
larger and does produce the desired strain needed to simulate observed motions.
The ice thickness distribution fields have been graphed at various times for all of
the problems under consideration. From these graphs we are able to observe the categories
of ice that participate in the ridge building process. These graphs are also
quite useful in helping to interpret the model response and compare it with photographs
obtained by aircraft and the LANDSAT satellite. Data taken from these photographs
are the percentage area of a given range of thickness categories. These data provide
valuable checks for interpreting the response and checking the predictive capability
of the ice model.
We recommend that future work with the AIDJEX ice model include the effect of Coriolis
acceleration. We neglected it on the assumption that it unnecessarily complicated the
results without changing the fundamental response of the model. While this is true,
we now feel it is more important to include those effects that must always be present
in reality. Comparison calculations that include the Coriolis force have shown that
the turning induced by the Coriolis force cannot be reproduced simply by applying the
winds at a different angle. Material non-linearities, therefore, require that this
force be considered by realistic simulations.
ACKNOWLEDGMENTS
We are grateful to Max D. Coon, Research Coordinator of the AIDJEX Modeling Group, for
suggestions that helped us to better understand the results of the model calculations.
This work was supported under the National Science Foundation Grant OPP7l-0403l to
the Arctic Sea Ice Study at the University of Washington.
REFERENCES
Colony, R., and R. S. Pritchard. 1975a. Integration scheme for an elastic-plastic sea
ice model. In Proceedings of 12th Annual Meeting of the Society of Engineering
Science. The University of Texas at Austin, Texas, pp. 953-963.
Colony, R., and R. S. Pritchard. 1975b. Integration of elastic-plastic constitutive
laws. AIDJEX Bulletin. 30:55-80.
Coon, M. D., G. A. Maykut, R. S. Pritchard, D. A. Rothrock, and A. S. Thorndike. 1974.
Modeling the pack ice as an elastic-plastic material. AIDJEX Bulletin. 24:1-105.
525 Pritchard/Schwaegler 13

Hibler, W. D. III, W. F. Weeks, A. Kovacs, and S. F. Ackley. 1974a. Differential sea
ice drift I: spatial and temporal variations in sea ice deformation. J. of
Glaciology. 13,69:437-455.
Hibler, W. D. III, S. F. Ackley, W. K. Crowder, H. L. McKim, and D. M. Anderson. 1974b.
Analysis of shear zone ice deformation in the Beaufort Sea using satellite imagery.
In J. C. Reed and J. E. Sater, ed., The Coast and Shelf of the Beaufort Sea. Arctic
Institute of North America, pp. 285-296.
Maykut, G. A., A. S. Thorndike, and N. Untersteiner. 1972. AIDJEX scientific plan.
AIDJEX Bulletin. 15:1-67.
Maykut, G. A., and N. Untersteiner. 1971. Some results from a time-dependent, thermodynamic
model of sea ice. J. of Geophysical Research. 76:1550-1575.
McPhee, M. G.
Bulletin.
1975a. Ice-ocean momentum transfer for the AIDJEX ice model.
29:93-111.
AIDJEX
McPhee, M. G. 1975. Water stress sUb-model for the AIDJEX model. Published elsewhere
in these proceedings.
Parmerter, R. R., and M. D. Coon. 1972. A model of pressure ridge formation in sea
ice. J. of Geophysical Research. 77:6565-6575.
Parmerter, R. R., and M. D. Coon. Mechanical models of ridging in the arctic sea ice
cover. AIDJEX Bulletin. 19:59-112.
Pritchard, R. S., and R. Colony. 1974. One-dimensional difference scheme for an
elastic-plastic sea ice model. In Computational Methods in Nonlinear Mechanics.
The Texas Institute for Computational Mechanics, Austin, Texas, pp. 735-744.
Pritchard, R. S. 1974. Elastic strain in the AIDJEX sea ice model. AIDJEX Bulletin.
27:45-62.
Pritchard, R. S.
Mechanics.
1975. An elastic-plastic constitutive law for sea ice.
42,2:379-384. Series E.
J. of Applied
Rothrock, D. A. 1974a. Redistribution functions and their yield surfaces in a plastic
theory of pack ice deformation. AIDJEX Bulletin. 23:53-81.
Rothrock, D. A.
Bulletin.
1974b. The energetics of plastic deformation in pack ice.
27:63-83.
AIDJEX
Rothrock, D. A. 1975. Energetics of the plastic deformation of pack ice by ridging.
J. of Geophysical Research. 80,33:4514-4519.
Schwaegler, R. T. 1975. Effect of changing the yield surface and the kinematic
relationship in the AIDJEX sea ice model. AIDJEX Bulletin. 29:135-150.
Thorndike, A. S. 1974. Strain calculations using 1972 position data. AIDJEX Bulletin.
24:107-129.
Thorndike, A. S., D. A. Rothrock, G. A. Maykut, and R. Colony. 1975. The thickness
distribution of sea ice. J. of Geophysical Research. 80,33:4501-4513.
Untersteiner, N. 1974. Arctic ice dynamics joint experiment. Arctic Bulletin.
1,4:145-159.
526 Pritchard/Schwaegler 14

During spring, 1974, two pairs of images were obtained eighteen days apart. These
dates were March 20 and 21 and April 7 and 8. Much of the analysis presented here
is based on these data.
Meso-scale Ice Motions
Cursory examination of Landsat imagery reveals ice motions at this scale: That is,
the motions can easily be seen on sequential pairs of standard 1:1,000,000 scale
Landsat images. Observations of ice motions at local scale requires examination of
enlarged images while long range influences on ice motions can only be observed by means
of large field of view satellite data.
Ioe motions over 24-hour periode
i. March 20/21, 1974. Figures 1 and 2 show Landsat scenes containing Katie's floeberg
on March 20 and 21, 1974. The March 20 image shows the floeberg located to the north
of an adjacent polynya (at this latitude Landsat images are oriented top to bottom
in a nearly northeast-southwest direction). The March 21 image (whose center point is
located to the west of the center point of the March 20 image) shows the floeberg
located approximately eastward of an adjacent polynya. It is not obvious from examination
of these images just what relative displacements have taken place and whether
or not they are uniform across the image. It can be seen that several SE-NW oriented
leads have opened around the floeberg to the southwest.
Figure 3 shows the floeberg surrounded by the vector displacements of nearby ice
during the 24-hour period between the acquisition of these two images. The first
impression given by this figure is that displacements vary in a somewhat systematic
way, increasing from north to south. Actually, the displacements fall into two groups.
The first contains 35 vectors and extends from the northeast to the floebergs, and the
second, containing 25 vectors, extends from the floeberg to the bottom (southwest) part
of the figure. The velocity corresponding to the upper group is 6.5 em/sec with an average
deviation of 0.3 em/sec while the velocity of the lower group is 8.7 cm/sec with an
average deviation of 0.6 cm/sec. These larger velocities correspond to the ice broken off
by the NW-SE trending lead systems.
It is interesting to compare the ice motion during this time with isobars on simultaneous
weather charts. Figures 4 and 5 show reproduced portions of Canadian Enviromental
Service surface charts with arrows near the location of the floeberg (location
circled) indicating the direction of ice motion, which is nearly parallel to the isobars.
Actually, one would expect surface winds to deviate from the geostrophic approximation,
having an additional southward component resulting from surface friction and a
consequent diminished Coriolis acceleration. However, because of Coriolis acceleration
of the ice which has been set in motion by the wind, there will be a component of the
drift forces toward the right of the instantaneous direction of motion. Zubov (1944)
pointed out that these two factors generally compensate for one another with the result
that pack ice generally drifts along isobars as appears to be the case here.
ii. April 7/8, 1974. Figures 6 and 7 show Landsat images of the floeberg and vicinity
for April 7 and 8, 1974. Sky conditions were not entirely clear at this time: some
rather thin clouds cast a number of shadows across the image. However, not all ice
details are obscured. Examination of the April 7 image shows the floeberg on the lefthand
side of the image surrounded by fragmented pack ice and are-frozen polynya
adjacent to the southwest side. Apparently, major motion had ceased for a time
sufficient for the polynya to freeze up to the floeberg. However, adjacent to the
floeberg there is a fresh lead approximately one kilometer wide indicating that ice
motion toward the southwest had recently been initiated.
530
Stringer and Barrett 4

Figure 1. Landsat I image showing "Katie's floeberg" on 20 March 1974, at 1:1,000,000
scale
531 Stringer and Barrett 5

Figure 5. Canadian Environmental Services surface weather chart for 0000 hrs UT
on 22 March 1974, showing the location of Katie's Floeberg" and the
observed direction of ice movement.
535
Stringer and Barrett 9

It is interesting to note the large east-west oriented lead running across the image
just north of the floeberg. There is another lead, somewhat less conspicuous but
parallel to this lead passing just south of the floeberg.
The April 8 image shows more distinct lead systems than the April 7 image. Further,
it is clear that the ice adjacent to the floeberg has been in motion since the April 7
image: a large polynya has been formed to the west of the floeberg. It is still
possible to see the impression left by the floeberg in the ice which had formed adjacent
to the floeberg prior to this motion.
Figure 8 shows the floeberg surrounded by the vector displacements of nearby ice
during the 24-hour period between the acquisition of the April 7 and 8 images. The
displacements fall into five groups as shown in the following table:
Group
I
II
III
IV
V
Table of Velocities for April 7 - 8, 1974
Location/Description Average Velocity Average Deviation
Pack ice north of floeberg 17.8 cm/sec 0.7 cm/sec
Slower moving ice just downstream 8 cm/sec 0.2 cm/sec
Slower moving ice just upstream 9.6 cm/sec 0.5 cm/sec
of floeberg
Block of vectors in lower left 6 cm/sec 0.4 cm/sec
Pack ice south of floeberg 3.9 cm/sec 0.5 em/sec
The ice in Group I is located on the north side of the shear boundary just north
of the floeberg. Referring to Figures 6 and 7 it can be seen that they represent
a relatively continuous sheet of pack ice moving as one piece. Group V is the large
block of vectors lying on the south side of the floeberg. It can be seen from Figures
6 and 7 that these represent an area of relatively broken-up pack ice with several
freshly frozen-over leads.
Groups I and V are essentially part of the large scale picture and will be considered
in detail in that section. However, some mention of the large scale picture of ice
motions must be made at this point: Figure 9 shows an enlarged portion of a NOAA II
image obtained at nearly the same time that Figure 7 was imaged by Landsat. The
area covered by this image is approximately 1000 km east-west by 800 km north-south.
The floeberg can be seen to the west of Point Barrow. Examination of this figure should
verify that the ice in areas I and V is actually part of the large scale ice motion
pattern and that the significantly slower speeds measured for motion of ice located
between these two groups is essentially a result of the blockage of the large scale
motion by the floeberg.
Groups II, III and IV represent ice apparently influenced by the presence of the floeberg
and will be considered here in terms of meso-scale influences on ice motions.
The major group, Group III is roughly 15 km wide just upstream of the floeberg and
necks down to a width of 8 km at the floeberg. It appears that the lower speed of
this ice is a result of blockage of its motion by the floeberg. The mechanics of the
blockage will be considered in the section on local-scale ice motions.
Finally. we consider the ice in Group IV. These vectors represent motions of individual
pieces of ice which are part of a large block located "down stream" of the floeberg
During the 24-hour period observed here, this block broke loose from block V and
moved at a speed intermediate between blocks II and V.
539 Stri nger and Barrett 13

The conclusion drawn from this analysis is that during this period the floeberg was
responsible for altering the motion of a column of ice approximately 20 km wide upstream
and downstream of its position (which remains virtually fixed). However,
the previously cited example, March 20/21, 1974 did not exhibit this behavior. In
neither case did the presence of the floeberg create a zone of ice exhibiting velocities
varying with distance perpendicular to the motion vectors. In both cases, rather
than taking place within a zone, shear occurs along well-defined lines.
Ice motions over Z8-day period
Still considering meso-scale ice motion effects, it is worthwhile to examine the effect
of the floeberg on ice motions observed over the time of a Landsat cycle: 18 days.
Figure 10 shows displacement vectors for ice motions between the times of Figures 2
and 7. Hence these displacements include the ice displacements occuring between
March 21 and April 7 and also the displacements just considered: April 7 to April 8.
The following table lists the groups of displacemnt vectors and their average deviations:
Group Location/Description Average Velocity Average Deviation
I Pack ice vectors north of floeberg 3.7 cm/sec 0.1 cm/sec
II Column of slower vectors upstream of 1.8 em/sec 0.2 cm/sec
floeberg
III Pack ice vectors south of floeberg 2.5 cm/sec 0.1 em/sec
IV Slower moving ice downstream of 1.6 cm/sec 0.1 cm/sec
floeberg
V Block of vectors in lower left of 1.1 cm/sec 0.1 cm/sec
figure
This picture shows a column of slower-moving ice apparently resulting from the blockage
of pack ice motion by the floeberg. Note that the two zones of pack ice north and
south of the floeberg exhibit velocities with a 3:2 ratio over a two and a half
week average whereas the one-day ratio was nearly 2:1.
This two and a half week average shows the floeberg slowing a column of ice approximately
26 km wide by almost a factor of 1/2. Here again, shear takes place along lines and not
continuously within a zone.
Long Range (1000 km or more)
In the examples just given for influences on meso-scale ice motions by the floeberg it
should be noticed that pack ice velocities north and south of the zone of immediate
influence were not equal. Figure 9, the NOAA II image obtained at very nearly the same
time as Figure 7, shows that the line of shear on which the floeberg is located is
part of an extensive crack and lead system. Figure 11, drawn from NOAA II images
from April 7 to April 9, 1974, shows the full extent of this system. The vector field
shown in Figure 8 shows the Meso-scale ice motion during the 24-hour period ending
at the time of Figure 11.
Figures 12 and 13 show barometric data at the beginning and end of this 24-hour
period, respectively. On these figures the location of the floeberg is given by a
circle and the instantaneous direction of ice motion by arrows. There is a strong
suggestion that the large crack and lead system has been generated by an east-moving
high pressure area with ice motions nearly along isobars as described by Zubov (1944).
Seldom would one expect to find equal driving forces over a very large area of pack ice.
Crack and lead systems result from either unequal driving forces or unequal retarding
541
Stringer and Barrett 15

Figure 12. Canadian Environmental Services surface weather chart for 0000
hrs UT, 8 April 1974 showing the location of "Katie's Floeberg"
and the observed direction of ice movement.
544 Stringer and Rarrett lR

Figure 13. Canadian Environmental Services surface weather chart for 0000 hrs
UT, 9 April 1974 showing the location of "Katie's Floeberg" and the
observed direction of ice movement.
545 Stringer and Barrett lQ

forces - or both. Hence, given the wind systems suggested by the isobars in Figures 12
and 13 one might expect a long east-west lead separating pack ice to the north and ice
perhaps subjected to near-shore drag to the south. In the absence of the floeberg this
lead would probably occur somewhere within 50 km of its position shown on Figure 11.
The conclusion drawn here is that the floeberg acts as a stress concentrator defining
the location of a line of shear which may extend several hundred kilometers.
Local (within 20-30 km)
One might have expected to observe meso-scale ice motion vectors in the zone of immediate
influence to become shorter as the ice approached the floeberg, resulting from compaction
of the ice. However, this phenomena is not observed at the meso-scale and so the question
arises: does compaction occur at a more local scale?
In order to search for this phenomena Figure 14 was prepared from 1:250,000 scale
enlargements of Figures 6 and 7 (Figures 15 and 16) showing ice displacements in the
immediate vicinity of the floeberg between March 20 and 21, 1974. Wherever large floes
could be identified, they were drawn in so that the viewer could make the distinction
between vectors drawn for the motion of individual floes and vectors drawn for different
points on large floes . (Note that in the latter case two unequal or non-parallel vectors
indicate rotation as well as translation of the floe.)
Examining this figure it appears that the motion of floes on either side of the floeberg
(transverse to the direction of floe motion) is not significantly altered in its
immediate viCinity, except for a slight tendency on the south side to diverge upstream
and converge again downstream. The three equal and parallel vectors terminating on
the floe just north of the floeberg indicate that this floe is not rotating and is
therefore not in a region of shear. Shear is therefore confined to the narrow region
between this floe and the floeberg. Although the vectors on the south side of the floeberg
exhibit the divergence - convergence pattern mentioned earlier, there is no evidence
of shear taking place any closer than these vectors .
Upstream of the floeberg there is evidence of blockage of ice motion. On the north side
there are three short vectors which, by their pattern indicate that these floes have
been slowed considerably and are being deflected around the north side of the floeberg.
Just to the south of the vectors, a large floe is approaching the floeberg and is rotating
clockwise. However, vectors drawn on opposite sides of a large floe located further upstream
show no sign of its being rotated or slowed upon its approach to the floeberg.
Vectors located between these floes show evidence of smaller floes being deflected somewhat
around the rotating floe.
CONCLUSIONS
It would be useful to tabulate the conclusions regarding pack ice driven past an isolated
obstruction drawn from analyses at the three distance scales.
1. Large scale (regional). The obstruction may act as a "stress concentrator"
influencing the location of major components of large lead systems.
2. Meso-scale. Pack ice driven past the obstruction can be broken into two sheets of
ice moving with essentially unaltered velocities and a column of slower-moving ice
between the two sheets containing the obstruction. Under these circumstances, the
width of the column of ice directly affected by the obstruction is on the order
of 3 times the cross-section presented by the obstruction. Ice within this column
is subjected to considerable breakage. {Individual pieces of ice become difficult
546 Stringer and Barrett 20

Figure 16. Enlarged Landsat I image of 21 March 1974 UT, at approximately 1:250,000
scale.
549
St r i nger and Barrett 23

to distinguish from one Landsat scene to the next as they approach the obstruction.)
Differental motion takes place entirely along the boundaries of the breaking zone.
In this respect the concept of an instantaneous "shear zone" does not really apply -
rather it would be more appropriate to describe differential motion taking place
along "lines of shear".
3. Small (local) scale. As ice within the breaking zone approaches the obstruction
the velocities of smaller pieces of ice decrease. This effect only appeared to
propagate as far as twenty kilometers upstream of the obstruction.
The width of the breaking zone does not increase at the location of the obstruction.
Hence, since the breaking zone is approximately 3 times as wide as the obstruction,
under steady-state conditions, ice within the breaking zone column must be on the
average half again as thick when moving past the obstruction as when farther than
twenty kilometers upstream. This estimate assumes that ice moves past the obstruction
at the average velocity of the breaking zone column.
Implications to Offshore Petroleum Development
Considerable attention has been given to the possibility of construction of permanent
structures on the continental shelf for activities related to petroleum extraction.
Here we will consider the conclusions just discussed in terms of such structures.
The implications to be drawn from the conclusions regarding effects on large-scale
ice motions is that even one structure could determine the precise location of a major
lead, assuming that the force pattern was such that a lead was about to be created in
the vicinity. (Here we are discussing a lead, resulting from differential stress within
the ice pack, which may be 1000 km in length and not the actual track of the structure
in the ice pack.) If the structure is located far from shore the picture could be very
similar to that presented by the floeberg; the structure would be surrounded by moving
ice. However, if the structure were located near shore, it might determine the position
of the flaw lead and thereby alter the extent of ice stationary with respect to the
shore. The alteration would very likely be such as to increase the extent of this ice.
The conclusions drawn from meso-scale considerations show that under the proper conditions,
not only would the track of the structure extend "downstream" (as would be expected)
but also the "breaking zone" column would extend "upstream" from the structure.
Obviously the cross section of the structure would influence the width and length of this
zone in a direct way. Taken together, the track and breaking zone related to the
structure would be responsible for considerable amounts of open water and thin ice in
the vicinity of the structure. Of course, the magnitude of this effect would be dependent
on pack ice motion.
It would be interesting to consider a further implication of this phenomena. The large
amount of open water and thin ice in the vicinity of such a structure could act to
attract sea mammals (and their predators) and could possibly prove to be beneficial
to Sea mammal existence. On the other hand, petroleum spills originating at the structure
would almost certainly collect in these adjacent areas, and possibly prove harmful to the
animals congregated there.
The local scale conclusions imply that often the structure would almost certainly be
surrounded by moving pressured ice - except on the side of the downstream polynya -
and direct surface travel to and from the pack ice would often be impossible. Even
in the event that pack motion ceased, whenever the motion resumed, the structure would
very quickly become isolated from the pack ice.
550
Stringer and Barrett 24

These conclusions would suggest that serious and careful considerations should be given
to the location of an off-shore structure placed within the moving pack ice region,
and its possible effects on marine mammals.
Acknowledgements. The research reported here was supported in part by NAS5-20959 (NASA
LANDSAT follow-up investigation) and NOAA, Outer Continental Shelf Environmental Assessment
Program Office, contract number 03-5-022-55.
REFERENCES
Zubov, N. N., Arctic Ice, translation by U. S. Navy Oceanographic Office and the American
Meteorological Society under contract to the Air Force Cambridge Research Center,
1963, Published by Naval Electronics Laboratory (San Diego), 491 pp.
551
Stringer and Barrett 25

ICE CONDITIONS ALONG THE ALASKAN COAST DURING BREAKUP
Neil Borgert
Crowley Maritime Corporation
Seattle, Washington
United States
EXTENDED ABSTRACT
This paper draws on the author's experience in directing marine operations of the firm's
annuaZ seaZift to Prudhoe Bay, and wiU covel" the foUawing subjects:
1. Approximate dates conventionaZ (non-ice-strengthened) tugs and barges can first
expect to transit the area from Wainwright ViZZage to Prudhoe Bay.
2. Ice conditions that can be expected to be found near the west and north AZaskan
shores between Wainwright and Prudhoe Bay with a given wind direction and force.
J. The direction and veZocity of currents and rise 01" faZZ of water depths with a
given wind direction and force.
4. Areas where vesseZs have been safeZy anchored during periods of strong 01" shoreward
fZow of ice.
5. Conditions which wiU suggest when vesseZs shouZd not attempt to transit the
hazardous areas from Pt. BeZcher to Pt. Barrow.
6. A short swrunary of ice conditions experienced during sumner seasons 1969 through
1974.
ABSTRACT ONLY AVAILABLE
555

SOME SYSTEM ENGINEERING CONSIDERATIONS FOR A FLOATING, COAL-FIRED lOOMW POWER PLANT
John P. Craven, Chennat Gopalakrishnan, Irving Swatzburg, et a1.*
Marine Programs
University of Hawaii
Honolulu, Hawaii
United States
ABSTRACT
By 1985 energy consumption in the United States is expected to exceed 3.42 x lo1 2 KWff. To
meet this demand with the present technology base and domestic energy resource mix, President
Ford has established an energy program which calls for energy conservation, increased
petrolewn production, deVelopment of midJ,}estern and Alaskan coal fields, increased
nuclear power plant capacity, and the construction of 150 major coal-fired power
plants in the next decade. Within this framework, a University of Hawaii graduate seminar
in Ocean Engineering systems and Resource Economics undertook a stuay to develop the
concept of utilizing marine space for the deployment of a 100MW power plant.
Initial economics and technological trade-offs indicated that:
1. Coal should be the fuel used, if available from midJ,}est or Alaskan sources.
2. A semi-submersible platform configuration should be used.
This system, modular in nature and deployable within existing technology, consists of
three standard 50MW electric generating units modified with marine boilers, each contained
within a semi-submersible hull, with ballast and buoyancy control, transportation
system, underwater power cable, and environmental monitoring.
The plant consists of a platform 40 feet above the ocean surface, 390 feet long by 340
feet wide, and 78 feet high. The platform is supported by 12 vertical columns 90 feet
long and 24 feet in diameter, which are connected to three main semi-submersible hulls,
each of which is 90 feet in diameter and 750 feet long. The total dispZacement of each
module would be approximately 400,000 tons. The construction material would be an appropriate
mixture of concrete and prestressed concrete.
Economic feasibiZity studies, based on a 175MW capacity, indicate:
1. Expected capital cost of 3.6 mills/KWff;
2. Other fixed costs at 1.3 miUs/KWff;
3. Operating costs (excluding fuel) at 2 mills/KWH; and
4. Fuel cost between 4.7 and 7.6 mills/KWff.
*Kenji Hotta, Vance Ishibashi, Cort Kloke, Toshio Nakajima, Ernest Oshiro, Ronald Watanabe,
Douglas Wilson, and Charles Wyse.
557

INTRODUCTION
The drive to preserve and improve our environment and the need to deal with the energy
crisis are two of the major issues facing the United States in the seventies. The
demands by a public increasingly concerned about the deterioration of the environment and
the tightening constraints imposed by the energy shortage, particularly the oil shortage,
are felt in every area of economic activity, but perhaps no industry will be so dramatically
affected as the electrical power generation industry.
Energy consumption in the United States is expected to exceed 3.42 x 1012KWH l by 1985.
To meet this demand with the present technology base and domestic energy resource mix,
President Ford has established an energy program which calls for energy conservation,
increased petroleum production, development of midwestern and Alaskan coal fields,
increased nuclear power plant capacity, and the construction of 150 major coal-fired
power plants in the next decade. Within this framework and with the support of the
National Science Foundation, a graduate seminar in Ocean Engineering Systems and a seminar
in Resource EconOmics at the University of Hawaii undertook an exploratory study to
develop the concept of utilizing marine space for the deployment of an electric power
plant.
The systems design was predicated on methodology described in Reference 1. In such an
approach classical cost effective calculations are subordinated to subjective probability
estimates and engineering experience. The systems design required, therefore, a careful
statement of the subjective goal of the system. The subjective goal which was employed
for a power plant utilizing marine space is:
The design, construction, and deployment of an offshore energy conversion system
that is capable of operating at a location at sea near a point on shore
utilizing the energy derived via marine space and converting this energy into
100MW of electric power with pollutant emissions which insure maintenance of
established environmental standards on shore and having the capability to be
expanded at some future date to produce an additional 50MW of packaged energy,
and that this system will successfully perform 95% of the time with a conditional
probability of 95% that the average yearly outage will be less than or
equal to 8 hours and have a demonstrated reliability of 95%.
Inherent in this objective goal is the value judgment that mass-produced coal power
plant systems will be more saleable in 100MW units as prime power for small island
nations and communities and as supplemental power for major installations during growth
or transition periods.
To meet this goal, the system was subdivided into six major subsystems: hull and platform;
power plant; transportation; process material, supply and disposal; environmental
sensing; and economic feasibility. For the sake of brevity, only the salient aspects of
each subsystem will be presented in the following sections.
HULL AND STRUCTURE
The Floating Marine Power Station (Fig. 1 and 2) was designed to match existing operational
profiles for conventional power plants and to match, whenever possible, the standard
power plant arrangements. This requirement was set to eliminate the lead time and
costs of development of new marinized equipments. There was also a requirement that con-
IThe 1971 energy consumption was 1.6 x 1012KWH • This figure represents an annual consumption
increase of 7% from 1950 to 1970. If I assume that energy conservation efforts
are effectively able to reduce this rate to 5% annually, then the consumption in 1985
will be 3.42 x 1012KWH.
558 Craven et al. 2

struction would not require a shipyard and that construction could take place in a
sequence which would allow easy installation of the power plant. This requirement was
again established to eliminate the lead time and cost of development of construction
facilities. Cost and construction considerations dictated that the platform be modular
in nature and utilize reinforced concrete for the major hull and plastics where possible
to meet environmental constraints.
The basic trade-off of a barge hull versus semi-submersible hull design was decided in
favor of the semi-submersible hull design. Although initial platform construction cost
data indicated that a barge design (.740 mills/KWH) might be less costly than a semi-submersible
design (.985 mills/KWH), the advantages of a stable platform with its inherent
ability to utilize standard landbased systems and components rather than marine systems
and components suggested intangible cost savings that would outweigh the platform differential.
Indeed, the barge must absorb or reflect the ocean wave energy, whereas the
majority of ocean wave energy will pass through a semi-submersible hull. Techniques for
determining the cost of such energy absorption have not yet been developed, but the
stress cycles to which elements of the structure are exposed will be very large. If only
one extra period of major maintenance and overhaul is introduced in a ten-year period as
a result of this energy dissipation, the apparent cost advantage will disappear.
The Floating Marine Power Station at this design phase, is composed of three modular
units. The individual module consists of a single buoyancy hull and four vertical concrete
columns which support an upper composite structure of concrete and steel. The vertical
cylinders are 24 feet in diameter and extend 50 feet below the sea surface. At
this depth the columns penetrate the main hull which is 90 feet in diameter and 700 feet
in length. This hull will provide the spaces for the major subsystems of the Floating
Marine Power Plant. The vertical columns are to be 24-inch thick concrete -- reinforced
as required. Special high temperature concrete will be utilized in areas of high heat
loads. The horizontal ties and pratt trusses between vertical columns and modules are
from 18 feet in depth to 34 feet in depth with spans of 75 feet and 140 feet respectively.
These sections will be composed of steel and concrete integrated to create a
loading bearing structure which will produce functional spaces.
The horizontal ties required for structural integrity are at the 95-foot mark which is
the center line of the main hulls and are 20 feet in diameter and 60 feet long. These
structures are pre-stressed, post-tensioned cylinders of 24-inch thick concrete.
The interface 40 feet above and below the ocean surface, considered the "band of disturbance,"
is 24 feet in diameter and of a configuration for minimal stresses and water
drag forces. These 24-foot diameter and 24-inch thick continuous vertical columns provide
cores for the vertical movement of people, goods and circulating gases.
The three modules necessary for the creation of a 100MW power plant would be rigidly connected
at the upper platform and at the center line of the main hull. The overall dimensions
of this marine floating power station are: (1) a platform above the ocean interface
which is 390 feet long by 340 feet wide and (2) three main hulls which occupy a volume
of 390 feet wide by 750 feet long by 90 feet deep. The total displacement of one
module would be approximately 400,000 tons.
POWER PLANT
In designing the power plant, it was decided to utilize three identical 50MW units which
would be operated to produce lOOMW of electrical power at the desired reliability. Thus
during times of routine overhaul or unscheduled outages when one of the units becomes
inoperative, the remaining two units would be able to maintain the lOOMW output.
With the exception of the boiler, the power plant is a standard coal-fired 50MW compo-
561 Craven et aZ. 5

nent (Fig. 3). Due to overall size, weight, and motion constraints of the platform, it
was determined that the standard boiler would not be adequate. The critical specifications
developed for a marine boiler for this system are presented in Table 1. The functional
space allocation is indicated in Figure 1. The steam boiler, turbine generator,
condenser and feedwater components of the mechanical plant of the power plant are located
in the horizontal concrete cylinders below the water surface. The primary voltages of
the electric plant are also located in these cylinders while the control room and auxiliary
voltage transformers are located on the upper deck. The 2S00KW gas turbine emergency
power network and the other services of the auxiliary plant are located on the
upper deck.
TRANSPORTATION
The transportation subsystem is composed of four sections: fuel supply, power transmission,
ash products, and personnel transfer. After examining the characteristics, location,
projected availability and estimated costs of midwestern and Alaskan coal, it was
decided to transport the coal in a slurry form. A SO,OOODWT slurry tanker coupled to a
single buoy mooring looked both technically feasible and subject to the least sea state
operating constraints.
It was determined that a single buoy mooring would be viable within an envelope of 600
feet of water, 3S-knot winds and a 2.S-knot current. The buoy would have a maximum diameter
of 41 feet and a gross weight of 150 tons and a slurry pumping capability of 660
tons per hour for a solid concentration of 40-45% with a specific gravity of 1.4 and a
maximum particle size of number 8 mesh.
The power cable specifications developed are given in Table 2. To insure adequate reliability
two separate cables are recommended. The material speCifications for the cable
were beyond the scope of this paper, but it should be noted that solid, gas pressured,
and oil filled underwater power transmission cables are all currently in use in various
parts of the world.
To remove the ash products a standard tug-barge combination will be adequate. The personnel
transfer would normally be accomplished by marine surface craft with an emergency
helicopter back-up capability.
PROCESS MATERIAL SUPPLY AND DISPOSAL
This subsystem is comprised of eight functions: coal handling, slag handling, ash handling,
dynamic ballast control (i.e. the intake and discharge of water), dynamic positioning
(i.e. for the purpose of object delivery and retrieval), ventilation (i.e. the intake
and discharge of air), circulating water, and component delivery.
The coal handling subsystem receives the coal in slurry form with the main slurry pumps
located in the mooring buoy. The coal is stored, decanted, and dried, in the 3 submerged
hulls as indicated in Figure 1. A minimum of three months supply of coal, 120,000 tons,
would be available at all times. As the coal is consumed, sea water is pumped into the
coal storage area to maintain proper ballasting. In addition, if coal was used as ballast,
then an additional 180,000 tons would be available in the event of a shipping hiatus.
The dust handling subsystem stores and transfers smokestack ash precipitation. Dust
obtained from the exhaust gas by electrostatic precipitation located on the main deck is
discharged by gravity through four two-foot diameter holes into the vertical struts.
The area on the outside of the exhaust gas ducting and inside the vertical strut walls
acts as a temporary storage area for the precipitated dust. At the bottom of the strut
562 Craven et al. 6

Size
TABLE 1. 50MW power plant boiler specifications
Maximum Length
Maximum Height
Maximum Width
Output
Motion
Steam Pressure
Temperature
Maximum Acceleration (3 axes)
564
63 ft
55 ft
20 ft
1400 psi
10000F
.02 g's rms
Craven et al. 8

TABLE 2. Cable characteristics
Size S inches in diameter
Weight 88 1b/yd (in air)
63 1b/yd (in sea water)
Min. Bending Diameter
Max. Tensile Load
Max. Temperature
Avg. Temperature
Loss Per Cable
Capacity of Transmission
565
12.S feet
IS tons
SO.SKW per 1,000 yd
lS0MW
1,200A
200KV
Craven et al. 9

overland coal shipping costs would decrease, and a coal-fired power plant would have a
slight cost advantages over oil. Unquestionably the cost of both oil and coal will rise
in the future, but in view of the relative scarcities of the two fuels, competing demands
for their use and ownership of the reserves, it is reasonable to assume that the cost of
oil will generally rise more rapidly than the cost of coal. Another important consideration
in operating a power plant with an expected life of 20-30 years is that coal will
probably hold less surprises than oil in availability and price.
The critical figure sought for the analysis was the delivered cost of a kilowatt hour
(KWH) of electricity into the Oahu electric power grid so that a reasonable comparison
could be made with present and anticipated costs of power generated by conventional
onshore power plants.
Insofar as possible, recent industry expenses provided the basis of cost calculations,
supplemented by engineering estimates where appropriate and necessary. Many costs cannot
be traced to a single source, but are judgments based on many sources of industry
statistics and estimates (Table 3). The range of total cost of coal delivered was calculated
to range from 4.6 mills/KWH to 7.6 mills/KWH.
Fixed and variable costs of operation would differ between an offshore coal-fired plant
and an onshore coal- or oil-fired facility. Areas of probably significant cost differences
have been identified elsewhere. 4 Reducing these figures to per kilowatt costs and
eliminating those fixed-cost items having no particular advantages to either system
results in Table 4.
CONCLUSIONS
The foregoing analysis suggests that an offshore floating power plant fired by coal in a
semi-submersible hull is both technically and economically feasible. Technical considerations
indicate that the system should be composed of three 50MW electric generating
units modified with marine boilers, each contained within a semi-submersible hull, with
ballast and buoyancy control and dynamic positioning capabilities. Power transmission
to the land would be via an underwater cable. This system, modular in nature, would be
deployable within existing technology and have the advantage of a standardized design
and could provide prime power for small island nations and communities and supplemental
power for major installation during growth or transition periods. Economic analysis
indicates that an offshore plant is cost competitive with its landbased counterpart.
It also appears that coal could be a less expensive fuel than oil for power generation
in Hawaii, given the following conditions:
1. High productivity, low cost, large scale mining·of western or Alaskan U.S.
coal
2. Adequate, efficient overland coal transportation to a West Coast port by
unit or integral train
3. A cost efficient marine terminal capable of rapidly loading large scale
ships
4. Efficient marine transport of the coal in large scale self-discharging
colliers.
When the external social and environmental benefits of removing a power plant to the
offshore location.are included, the offshore plant appears markedly superior.
4Craven, J. P., C. Gopalakrishnan, et. al., "Some Economic and Engineering Considerations
for Floating Coal-Fired 100MW Power Plant," 11th Annual <strong>Conference</strong>, Marine Technology
Society, September 1975 (in press).
567 Craven et aZ. 11

Basic Plant Equipment
Land and Structure
Platform and Structure
Mooring and Connection
Undersea Cable
Delivery and Installation
S02 Pollution Control Equipment
Cooling System
Extra Engineering Costs
TOTAL COST PER KW
Total Plant Cost - 175MW say:
Annual Capital Cost @ 12% say:
Annual Capital Cost Mills/KWH
(80% operating factor)
Other Fixed Costs
Operating Costs, excluding fuel
Total Costs, except fuel
TABLE 4. Capital costs per KW
569
Land
$ 136
35
25
4
$ 200
$35,000,000
$ 4,200,000
3.4
1.3
2
6. 7 mills /KWH
Sea
$ 125
58
11
5
3
3
7
$ 212
$37,100,000
$ 4,460,000
3.6
1.3
2
6.9 mills/KWH
Craven et al. 13

REFERENCES
Craven, J. P. 1971. Ocean Engineering Systems. The MIT Press, Cambridge, Massachusetts
and London, England.
570 Craven et aL. 14

USCG POLAR CLASS ICEBREAKER DESIGN PARAMETERS
AND FEATURES IN ADVANCEMENT OF ICEBREAKER DESIGN
Cdr. H. E. Fallon, Jr.
U.S. Coast Guard
Baltimore, Maryland
United States
ABSTRACT
The design of the USCG Polar Class Iaebreaker began in the mid 1960's with the searah to
update the Coast Guard's Iaebreaker teahnology and define the short and long range mission
requirements for u.s. Iaebreakers. Iaebreaker teahnology at that time was aonfined
to a very small. group throughout the world and it has only been in reaent years, with the
onset of inareased aonaern for resourae development in iae aovered areas, that iae and
iae breaking teahnology has beaome of general aonaern. One has only to pursue the referenaes
listed in aul'l'ent papers to note the preponderanae of those listed and dated after
1968.
Design and aonstruation of any ship as large and aomplex as an iaebreaker is not a short
te:rm projeat. The Polar Class detail. design aould be said to have interrupted this
aforementioned evolving searah for teahnology in 1969-l970. As suah, the design improvements
01' advanaements, if you will., must be viewed in that time frame. The first of
the Class was saheduled for delivery and subsequent iae trials in February thru April
1976. The iae trial is the real and ultimate test of the appliaability 01' proof of
these advanaements.
571

INTRODUCTION
For the purpose of this discussion the areas of advancement are divided into three general
categories:
A. Hull and Structure
B. Propulsion and Control
C. Other Features.
HULL AND STRUCTURE
The hull design considered both the need to provide minimal ice resistance due to its
form and adequate strength to resist the loads imposed in breaking and displacing the
ice.
The hull areas of most significance in icebreaking are the bow and forebody shape, and
the stern configuration. The bow and forebody shape was optimized to provide both adequate
continuous and ramming mode icebreaking capability. Coast Guard design criteria
established 3 knots as the minimum speed considered continuous with a transition zone of
o to 3 knots. This involved optimum selection of the local waterplane and flare angles,
(Fig. 1) for continuous icebreaking and, in addition, the location of the point of maximum
beam forward of amid ships. The bow angles, i.e., stem and spread angles, (Fig. 2) were
selected to maximize the ramming performance while minimizing the extraction force
necessary after a ram.
Stern Configuration
The shape of the waterplane aft has a significant effect on the amount of broken ice
that comes in contact with the propeller. Adequate design provides sufficient space between
the side of the icebreaker and the broken channel to allow the submerged broken
ice to return toward the surface and hence clear the propellers (Fig. 3).
Hull Structure
The considered advancement in icebreaker hull structural design are the studies conducted
to define the loads experienced and apply these thru the use of established engineering
design methods. This establishes documentation in the design which can be reanalyzed
based on actual load experience to provide improved design information.
The hull structural form is different from earlier U.S. Icebreakers in that a grillage
frame system employing deep webs and frames throughout is used for the Polar Class as
opposed to the Truss Framing System used in the mid-ships six-tenths length of the Wind
and GLacier Class.
A grillage system has much better response to local overloads for resisting progressive
failure. However, it is a heavier weight system when considered on an equal-design
strength basis.
The icebelt configuration used in the Polar Class design is shown in Figure 4 along with
the design loads for elastic deformation in the areas shown.
For small areas, localized impact loads of 1200 psi, with a plastic deformation criteria
was used.
These ice loads were the subject of considerable early study by the Coast Guard and are
still an area of investigation for future Icebreaker design. Testing is planned during
the scheduled full scale trials to obtain more data to assist in determination of
accurate design load information.
572 Fallon 2

I'
-.'
WING rr10PEll«:n
____ 0; ...
Figure 3. Propeller design
(-.v,x.
____ ,u.E.AM 12
574
c
CHI'.!--!Nc l
WIDTH 12
Fallon 4

In preparation for full scale tests to determine these load levels and the points of
maximum stress a two dimensional photo elastic model was constructed and loaded to simulate
ice loadings.
Hull Material
The hull material (Table 1) selected is a Coast Guard derivative of ASTM A 537 quenched
and tempered steel. Designated as CG A537M this material has a reduced carbon and increased
manganese content, and exhibits excellent ductility and fracture toughness at
reduced temperatures down to -60°F.
The physical properties of this material allowed utilization of the design loads previously
discussed without the penalty of increased weight or reduced internal space
which would have been suffered with a lower strength material.
PROPULSION AND CONTROL
The most often discussed area of advancement is the Polar Class Propulsion and Control
System. The propulsion system consists of a Diesel Electric AC-Rectified DC plant and
a Geared Gas Turbine plant driving controllable reversible pitch propellers.
The Polar Class embodies a three-shaft system to give improved reliability and maneuverability
over the WindaUlSS. An even power distribution between shafts will be used.
This provides less thrust unloading of the wing screws due to propeller interaction and
gives better performance in the near bollard condition.
In a departure from conventional practice, we have specified a CODOG - or Combined
Diesel or Gas Turbine Plant consisting of an 18,000 shp diesel-electric plant and 60,000
shp in gas turbines. This ship will have a free running speed of about 17 knots and a
continuous icebreaking capability estimated in new, hard, sea ice of 4 feet in the dieselelectric
mode and 6 feet in the gas turbine mode. Normal operations, including icebreaking,
will be in the diesel-electric mode with the gas turbines being used for those
limited periods when 18,000 shp will not suffice. Incidentally, it is estimated that this
ship will be able to break 21 feet of ice by ramming.
The propulsion system for an icebreaker must have the ability to produce power with high
efficiency over a broad range and have the capability of high rates of change of power.
The system must have a low susceptability to damage in a hostile environment, low
maintenance and allow containment in a minimum geometry to reduce the beam required due
to this effect on hull form (i.e., ice resistance).
During the various design phases all available types of power plants were studied and
have application dependent on the mission, ship size and set operational requirements.
The new Coast Guard icebreaker is a replacement ship for the aging WindaZass. Early
preliminary design studies concluded that a rectified AC system driving DC motors was
the only practical electrical means of attaining high diesel-electric shaft horse powers.
Further study of this system revealed many desirable characteristics such as, smaller
size and weight, higher efficiency, lower maintenance, higher dependability and flexibility
of installation so that it was retained after the diesel power level was reduced
to a power level attainable with an all DC plant.
The gas turbine was selected for its well known weight-power ratio and size advantage, as
well as cost limitations to achieve the maximum horse power desired. In addition, the
gas turbine possesses torque characteristics which are highly advantageous to icebreaking.
Stall torque to full load torque ratios of 2.5 are not uncharacteristic and this is obviously
desirable when a propeller loads down in ice. To reduce the problems of reversing
in both the diesel and gas turbine mode a controllable pitch propeller is utilized.
The controllable pitch propeller has a great influence on the design aside from
576 Fallon 6

TABLE 1. Hull material - physical and chemical properties.
CHEMICAL REQUIREMENTS*
Element
Carbon, Max.
Manganese
Ladle
Check
Phosphorus, Max.
Ladle
Check
Sulfur, Max.
Ladle
Check
Silicon
Ladle
Check
*Small amounts of certain alloying elements will be present
but shall not exceed the following amounts:
Copper
Nickel
Chromium
Molybedenum
TENSILE REQUIREMENTS
Tensile strength
Yield point
Elongation in 2 inches
Elongation in S inches
BEND DIAMETERS
Thickness of material
To 1-1/4 in. (31.75 mm) thickness, incl.
Over 1-1/4 in. to 2 in. (50.S mm) thickness incl.
Charpy V-notch energy requirements
577
CG A537 M ASTM A537 B
Composition (per cent)
0.16 ladle
0.20 check
0.90 - 1.50
0.S5 - 1.55
0.035
0.045
0.040
0.050
0.15 - 0.35
0.13 - 0.40
0.35
0.25
0.25
O.OS
70,000 to
90,000 psi
50,000 psi
min.
22% (min.)
lS% (min.)
0.24
0.70 - 1.35
0.65 - 1.40
0.035
0.040
0.15 - 0.50
0.13 - 0.55
0.35
0.25
0.25
O.OS
SO,OOO to
100,000 psi
60,000 psi
min.
22%
Ratio of bend diameter
to thickness of specimen
2 2
2-1/2 2-1/2
20 ft. lbs.
Fallon 7

this ability to reverse large amounts of power. An analysis of existing icebreakers
fixed pitch propeller failures indicated a predominant number happened when the propeller
is stopped and ice encounter occurs.
Every reversal of stop of a controlled pitch system presents this opportunity for damage.
The CP system rotates continuously and thus experiences the far less strenuous condition
of "Milling" ice. The CP system also allows the efficient transmission of power over
this broad range. A typical cruising speed would require no more than 7000 shp while
the ice breaking mode could require up to full power of the gas turbines. ConSidering
these factors, and in light of the long cruising run before the vessel reaches her
operating area, you can see the favorable effect on required vessel endurance and, consequently,
size and cost. The rate of reversal also indicates significant advantages
of the CP system in the ramming mode since the speed of advance obtained before impact
is a function of the speed of reversal. It is expected that the CP system will provide
a 20 percent advantage over the fixed pitch propeller in the ramming mode.
The gas turbines are directly coupled to the shaft system through a clutch reduction
gear, the motor brushes are lifted from the commutator during this mode of operation.
A dental coupling aft of the reduction gear allows uncoupling of the reduction gear
during motor operation.
Control system operation in both modes is programmed to reduce propeller pitch first to
maintain RPM and resist stall.
Design of a propulsion system such as this introduced some new areas for concern and
more detailed consideration. The reduction gear, motor, propeller combination is a
three mass shaft system and the critical frequencies fall near, both above and below the
desired operating range. The introduction of the ice milling factor as a design input
or forCing function on shaft torsionals is critical in the case of a direct geared drive
propulsion system and the possibility of gear tooth separation during this condition
must be reviewed.
The Polar Icebreakers have been designed for unmanned machinery spaces, all operations
may be accomplished from the ECC (Engineering Control Center). This requires an extensive
control and monitor system.
In addition the ECC contains the master bench boards for the auxiliary power generators,
the set up switch boards for the main propulsion, the central hydraulic system control
panel, back-up alarm panels and other necessary auxiliary equipment controls.
The controllable pitch propeller material used for the Polar Class Icebreaker is a departure
from that preViously used for U.S. Icebreakers. These materials have previously
been MIL-S-16993A, Class 2, a chromium manganese stainless steel, or as an alternate a
NiAl Bronze material, the material being used for the Polar Class propellers is designated
CA6NM and has properties as listed in Table 2. This material has not previously
been used for icebreaker propellers but its chemistry, the mechanism by which it obtains
its hardness and its physical properties and the relation to previous materials and their
success indicate that it should be a superior material for Icebreaker propellers. Other
attributes of the material are its weldability which lends itself to repair of casting
defects.
Full scale trials are planned to investigate and attempt to document the results and
effects of these above stated features. Starting with establishing ice breaking capability
and hull loading (as indicated previously) complete monitoring of the propulsion
plant response will be accomplished. Table 3 is a listing of the scope of the trials as
established, at the present time. Conditions and time will obviously be a determining
faetor as to the capability of actual accomplishment.
578
Fallon 8

TABLE 2. Propeller material.
Carbon
Manganese
Silicon
Phosphorus
Chromium
Nickel
Molyadenum
Sulfur
Tensile
Yield
Elongation
Reduction area
MIL-S-16993A
Class 2
.15% Max
1.0 % Max
0.5 % Max
0.05% Max
11. 5-14%
0.65-1.0%
0.5-0.7%
0.05% Max
90 ksi
65 ksi
18%
30%
OTHER FEATURES
ASTM
CA15M
.15% Max
1.0 % Max
0.65% Max
0.040%Max
11. 5-14%
1.0 %
0.15-1.0%
0.040%Max
90 ksi
65 ksi
18%
30%
ASTM
CA6NM
.06% Max
1.0 % Max
1.0 % Max
0.040%Max
11.5-14%
3.5-4.5%
0.40-1.0%
0.040%Max
110 ksi
80 ksi
15%
35%
The Polar Icebreaker is a multi-mission vessel suited for operations of long duration in
ice covered waters. Hence a significant amount of the area above the machinery spaces
are outfitted to accomplish various scientific missions.
The oceanographic wet lab contains two winches, a hydraulic boom, covered controls, an
oceanographic dry lab, chemical lab, meteorological lab, ocean data center and quarters
and office area for 10 scientific personnel are also provided.
Other unique features are the central hydraulic system, provided to establish a central
power source for diverse equipment which is not used coincidently in the course of normal
operations, and a fixed helicopter hanger and flight deck for day-night flight operations.
The Polar Icebreakers are probably the most advanced ships in the Coast Guard from the
standpoint of habitability. A stateroom configuration is used throughout - four, two,
and one person staterooms are utilized.
The Polar Class deSign, while completed in 1971 and embodying advancements developed
thru that time, will have certain other features added during retrofit availability.
Coast Guard testing of underwater hull coatings for ice breaking vessels has continued
during this time and the polyurethane coating system appears to have considerable
promise in this application. As such it is planned to coat the underwater body with
this coating system.
Model test techniques have improved considerably since the time of the early Polar Class
studies, as such further model tests are planned to enhance the model, full scale
correlation data base and further evaluate the theoretical methods in existence at the
present time and those used in design at the Polar Class Icebreaker.
579 Fallon 9

Table 3. Test plan.
Icebreaking
Performance
1. Resistance
a. Continuous
icebreaking
resistance
b. Static
(starting)
resistance
c. Heeling system
during continuous
icebreaking
d. Clogged channel
resistance
2. Ramming/Extraction
a. Ramming & extraction
resistance
in level ice
b. Ramming & extraction
resistance
in pressure ridges
3. Maneuvering
a. Turning circle
4. Hull friction
a. Side friction
test
Ice Impact
Forces
1. Hull impact
forces
2. Rudder impact
forces
Machinery
Performance
Environmental
Data
1. Propulsion 1. Ice properties
sIstem con- measurement
troIs
2. Propeller ice 2. Meteorological
pact & ice Data
milling
3. Shaft vibration
tests
4. Engine cooling
& recirculation
5. Measured mile
senoor calibration
Photograph
IC Documen·
tation
In summary, the new Polar Class Icebreaker will be superior in all respects to the
GLacier and WindaLass icebreakers. It will be able to more successfully complete present
mission requirements as well as being capable of performing missions which are not impossible.
The larger physical size and total shaft horsepower available, will allow
the new class to be 25 per cent more productive in transit and perhaps equally as much in
the ice. This increased productivity will allow the assigning of new, heavier operating
requirements to the new class while decreasing the number of icebreakers in an area with
changing the level of support. The advancements then have surely accomplished their
desired purpose.
580 Fallon 10

PRESTRESSED CONCRETE FLOATING TERMINAL
FOR ARCTIC OCEAN SERVICE
Ben C. Gerwick, Jr.
Professor of Civil Engineering
University of California
Berkeley, California
United States
ABSTRACT
The development of the l'esoupces of the offshol'e APctic unU l'equil'e terminal. stZ'uctUI'es,
capabZe of staying "on station" in the deep watel'S undel' aU pl'obabZe soo ice conditions.
SUch terminaZs can be utiZized fol' pl'Ocessing, stol'age, and shipping opel'ations: The
Zattel' may not necessa1'iZy be by sUPface ship.
One soZution to this is a vePy Zal'ge pl'estl'essed concl'ete bal'ge, using its mass, shape,
and 'length in effective combination to continuousZy bl'eak the ice, and abZe to move
ZateroUy in the event of ice isZand encountel'.
This papel' e:r:pZol'es the fil'st component of such a terminal. stl'uatupe, the fZoating tel'minaZ,
in some detail., and suggests possibZe soZutions fol' the subsequent components of
mool'ing, risel' and off 'loading connections, needed to make the system viabZe.
581

INTRODUCTION
Papers presented at the First POAC <strong>Conference</strong> in Trondheim in 1971, on the Utilization of
Prestressed Concrete in Arctic Ocean Structures (Gerwick, 1971; Figs. I, 2, and 3), addressed
the potential advantages of high strength prestressed concrete for Arctic Ocean
structures and made specific reference to the proposed fixed offshore terminal module for
Humble Oil, to be located 40 km north of Prudhoe Bay. They featured the upward-breaking
cone as an effective means of causing the ice sheet and pressure ridges to fail in tension
Since that 1971 conference, the explosive development of fixed concrete platforms has
taken place in the North Sea: The Ekofisk Caisson and Condeeps Beryl A and Brent B have
been installed, and 11 other fixed concrete platforms are under construction.
A 70,000 ton displacement prestressed concrete floating terminal for LPG is under construc·
tion in Tacoma, Washington, destined for service for ARCO in the Java Sea (Fig. 4), and a
large offshore coal-shipping terminal off Queensland, Australia, founded on 10 prestressed
concrete caissons (Fig. S), is nearing completion.
NEW TECHNOLOGICAL DEVELOPMENTS
Since the 1971 paper, a number of new technological developments have occurred which have
a major impact in relation to Arctic Terminals:
1. High strength concretes, obtained by careful selection and control of ingredients
and the use of super-water-reducing admixtures enable practicable design compressive
strengths of 60 to 70 N/mm 2 to be realized.
Progress is being made in the development of practical means for using polymerimpregnated
concrete (P.I.C.) which gives very high compressive strengths (100
to 140 N/mm 2 ) , high tensile strengths (10 N/mm 2 ) and excellent qualities of
abrasion, resistance and durability.
Fiber concretes are also being brought to the stage of practical application,
using randomly-oriented steel fibers to give enhanced tensile strength.
For both of these (polymer and fiber concretes), the increased strengths are unfortunately
accompanied by a brittle mode of failure. However, the combination,
polymer-impregnated fiber reinforced concrete exhibits high ductility and toughness.
Ferro-cement, in which layers of closely-spaced wire mesh are embodied in fine
concrete or shotcrete gives the possibility of developing very high tensile
strengths (8 N/mm 2 and more) and can be effectively used in composite action,
especially for doubly-curved shell configurations.
High strength lightweight concrete is also now practicable, with compressive
strengths of S2 N/mm 2 and a unit weight of 1900 Kg/m 3 •
2. A better understanding of shear-resistant mechanisms enables us to use posttensioning
and conventional reinforcement much more effectively in resisting
local and over-all shears from ice.
3. New coatings, such as dense epoxies and particular, dense polyurethane, provide
proper "armor" for resisting abrasion and reduction of adhesion. The dense polyurethane
has been applied to the USCG "Polar class" icebreakers as well as to
Great Lakes icebreakers, and appears to be very effective. Coating thicknesses
up to 1.2 mm are practicable.
582 Gerwick 2

w
Figure 1. Proposed Humble Oil terminal, Prudhoe Bay

Figure 3. Proposed Arctic Ocean marine terminal of prestressed concrete (by Santa Fe-Pomeroy, Inc., USA)
Early concept

Figure 4. LPG terminal barge for ARCO, for service in Java Sea, now under construction in Tacoma, Washington
by Concrete Technology Corporation and Transgas, Inc.

---
Figure 5. Prestressed concrete terminal, Hay Point, Queensland, Australia

4. Better understanding of ice-structure interaction: The role of mass and the
modes of ice-breaking energy requirements.
5. Development of more practicable and efficient methods of construction of large
prestressed concrete floating structures, utilizing such construction concepts as
precast segmental construction, slip- forming, assembly afloat, and construction
afloat.
Practical methods have been developed for the efficient construction of complex,
double-curved shapes (Figs . 6, 7, 8, and 9).
RELATIVE MERITS OF FIXED AND FLOATING TERMINALS
Fixed terminals have the advantage of permanent connection to the seafloor and hence facilitate
connection of risers and subsea pipelines. Although subjected to dynamic loads, the
mass and rigidity of the structure permits design and construction resistant to fatigue
and wear.
On the negative side, fixed terminals must rigidly resist maxima pressure ridges, and the
unlikely but possible encounter of an ice island fragment . High loads are imposed on the
foundation, including bearing, rocking and shear .
Such terminal must be axi- symmetrical, designed to resist load from any direction. There
are obvious limits to water depth. In the event of seismic action, the structure must
resist the acceleration forces against the ice sheet.
A floating terminal can utilize its inherent dynamic response to aid in ice-breaking and
significantly reduce the forces . By proper design, these responses can be tuned to be
most effective. Such a terminal can "weather-vane", so as to present its most favorable
aspect to the on-coming ice.
A floating structure can be moored in such a manner as to permit it being moved laterally
through the ice to avoid an on-coming ice island fragment, assuming timely location and
determination of direction. In the event that overwhelming force breaks the structure
loose from its moorings, then it still rides safely in the ice pack. Finally, a floating
structure is effectively decoupled from seismic acceleration of the sea floor.
The principle disadvantages of the floating structure are the moorings and riser connections:
These will be discussed later, as will emplacement and offloading concepts.
A third type of terminal, the seafloor (submerged) terminal involves entirely different
considerations and is not considered further in this paper .
FLOATING STRUCTURE - ICE INTERACTION
Two papers in the 1975 SNAME conference, Berstad (1975) and Milano (1975), developed the
important parameters for ice-breaking in a floating structure and the various mechanisms
by which such a floating vessel breaks the ice. They emphasized the need for very large
size (mass), in order to smooth out the peak loads; conical shape, in order to force the
ice to fail in tension; the advantages inherent in downward-breaking, and the advantage of
length in obtaining a favorable pitch frequency.
Downward-breaking, as opposed to the upward-breaking cones previously designed for fixed
structures, reduces the work required. Instead of lifting the ice sheet out of the water,
the ice is pushed down into the water. Initiation of tension failur e may be facilitated
by multiple thermal cracks and discontinuities in the top surface of the ice .
588
Gerwick 8

ICE-BREAKING
Milano (1975) evaluates the relative energy expenditures of the several modes or phases of
ice-breaking. Total energy is expended as follows:
E 1 Movement through broken ice
E 2 Impact (crushing)
E 3 Rising up onto the ice
E 4 Downward pitching
E 5 Displacing ice along the sides of the vessel.
E 2 (impact and crushing) contributes about half the total resistance at all
speeds.
E 5 is important only at higher speeds.
Thus, the efficient design for a moored floating terminal will mobilize the dynamic energy
of the moving ice and translate it into a downward-breaking vertical force.
In the event of slow-moving ice, or excessive damping by the broken ice, pitch can be excited.
One such means is the Pneumatically Induced Pitching System (PIPS) developed by
Arctic Engineers and Constructors. Another means, proposed herein, is the uae of hydraulically-actuated
cable grip hoists, alternately pulling and releasing the primary mooring
lines in resonance with the natural pitch frequency of the terminal barge.
CONFIGURATION
The requirements of large mass and double-curvature considered along with the need for
economy, suggest the use of prestressed concrete for the hull structure.
In order to force downward breaking of the ice, an inverted cone shape is adopted forth
bow. Beam has a relatively small influence on the ice forces at low speeds, hence a substantial
bow width at waterline is acceptable. Aft of the conical bow, the below-water
hull tapers inward slightly to facilitate movement of the broken ice. Above water, the
cross-section flares out to give a large deck area.
The hull bottom has a downward-extending bulb, as a continuation of the cone, so as to
place the draft of the bulb below that of the deepest anticipated pressure ridge. This
permits mooring through a well, so that the lines do not encounter ice.
The vessel must be large: A displacement of 250,000 tons or larger, is contemplated.
Length is essential to reduce pitch frequency: An acceptable length might be 200 m.
Natural pitch frequency is planned to be about 60 sec.
CONSTRUCTION
Because of the complex double-curved shapes and large size, it becomes essential to develop
an efficient method of construction. Any system adopted must give adequate consideration
to the installation of piping, mechanical, and electrical systems. However, one cannot
be afraid to be bold, especially in view of the successful carrying out of critical operations
for North Sea platforms and offshore structures, (e.g., the Dubai Oil Storage)
which at the time represented extremely bold undertakings.
Obviously the lower hull of the terminal vessel could be built in a deep basin or graving
dock, launched when completed to a reasonable depth in relation to its length, (e.g., 15
m). If temporary buoyancy and support tanks have been installed around the perimeter, a
launching draft as low as 8 m may be practicable.
To facilitate concrete forming and placement, precast concrete "slab" segments would be
cast in the adjacent yard, in the most favorable position (nearly level). Then they would
593
Gerwick 13

e set on falsework supports or on the temporary support tanks. Joints between slabs
would be cast-in-place.
Post-tensioning would be extensively employed to overcome bending tension and to create
favorable normal forces to resist shear.
Once launched, the process would be continued afloat, using struts and tie-backs to hold
the precast slabs in place until jointed and tensioned. Slabs could be very large, since
they would be set by waterborne cranes.
While feasible, the above method does require temporary buoyancy tanks and extensive temporary
support. Further the basin required is very wide and long.
The lower hull could be built in two halves, thus requiring only half as long a basin.
Then the halves would be brought together while afloat and joined by post-tensioning, in
a manner similar to that used to join the sections of the Hood Canal Floating Bridge.
The problems and cost of temporary buoyancy tanks and overhanging supports would be eliminated
by building the lower hull segments upside-down, then rolling them over in the water.
Properly planned roll-over operations are well established in salvage work and in barge
repair operations, and would be practical for the lower half segments, either singly or
after joining. Temporary support of the upper hull overhangs does not present as serious
a problem for construction, since the interior structure would be built up ahead, giving
adequate structure against which to strut and tie the precast panels.
MOORING
Mooring of this proposed terminal barge in moderate offshore areas (3 to 4 m multi-year
ice, and 30 m deep pressure ridges), will develop lateral forces of several thousand kips.
These are significantly below those developed by a comparable fixed structure, e.g., only
20 percent. Peak forces will have been smoothed out by the mass of the vessel and the
hydraulic buffers built into the cable-actuated hoists. Present concepts envision the use
of a cluster of large diameter wire rope lines, brought up through the mooring well.
Catenaries will be laid out in respect to the water depth so that under normal ice conditions,
the departure angle of each line is 70' with the horizontal or steeper. The lines
will be pre-adjusted to approximately equal tension by the cable-grip hoists.
Where a riser must be maintained, the problem becomes similar to that of an SBM or SPAR
type structure. Risers will either be stopped off to the mooring line cluster, or a different
type of mooring; tension leg, rigid arm, or radial cluster will be employed, since
these would limit the excursion. The forces are large, but still within practicable
limits. A rotating swivel would be required in the mooring well, but fortunately would
be operating in a controlled and constant environment.
Anchors to develop the required resistance are also a separate but important part of the
total system, and must be specially adapted to the soil conditions and depth at a particular
site.
Proposed water depths for mooring are in the 30 to 60 m range, althoug? deeper installations
are possible.
EMPLACEMENT (DEPLOYMENT)
The movement of such a terminal to an offshore site may involve penetrating the ice during
summer months. The stern will, therefore, be fitted to accept a pusher-tug of about 30,000
S.H.P., since the vessel and bow already possess an adequate ice-breaker configuration.
594
Gerwick 14

The vessel is capable of setting and planting its own moorings, assisted by the pusher
tug. The tug may itself need ice-breaker assistance to return to open water.
OFFLOADING
Large tankers especially designed for ice-covered waters, (or conventional tankers operating
during ice free periods), would approach the terminal from astern and moor to it for
cargo transfer.
Submarines or semi-submarine tankers may utilize the terminal with under-ice transfer astern.
For example, a stinger pipe may be entered into a well contained within the stern
of the terminal vessel and connection made in the dry, all in a protected environment.
COST
While no detailed analysis of cost has yet been completed, it appears that the hull structure
cost would be about $35,000,000 to $40,000,000 at present day (1975) prices.
CONCLUSIONS
Progress is being made in the development of concepts for an Arctic offshore terminal.
One promiSing concept is the use of a large prestressed concrete floating terminal, embodying
a downward-breaking cone with induced pitch, moored so as to weathervane on heavy
moorings, and able to move laterally, if required, in order to avoid ice island fragments.
Obviously, a great deal of detailed development is required to complete the system design
and to prepare detailed designs for a specific location and function.
REFERENCES
Berstad, E. E. 1975. The importance of size in an Arctic ship. Proceedings, Society of
Naval Architects and Marine Engineers <strong>Conference</strong>.
Gerwick, B. C., Jr. 1971. Utilization of prestressed concrete in Arctic Ocean structures.
First <strong>International</strong> <strong>Conference</strong> on Port and Ocean Engineering under Arctic Conditions.
Milano, V. R. 1975.
motion in ice.
ference.
Variation of ship/ice parameters on ship resistance to continuous
Proceedings, Society of Naval Architects and Marine Engineers Con-
595 Gerwick 15

INTRODUCTION AND HISTORICAL DEVELOPMENT
The development of the Canadian Arctic is today in the exploratory stage similar to the
development of Canada itself two centuries ago. The rate at which exploration becomes
exploitation remains to be seen but there are clear indications that we are on the
verge of major development. Although the oil and gas has not yet begun to flow south,
nickel, asbestos and other minerals are already being shipped from Arctic ports.
Canadians can no longer afford to simply dream about the North, they must act.
A visitor from Europe flying over North America will note the work of the surveyor in the
neat rectangular pattern of agriculture and highways. This framework of basic environmental
information is an essential ingredient of planning. Therefore, in the Arctic the
basic description of the area both above and below the water, is essential to developing
the region. Part of this information is a description of the underwater topography.
In the initial stages it will only be necessary to have a general picture but as the
pattern of development becomes clear the information density must increase. It must be
emphasized that both the general and the specific information must be available before
the developments begin.
Apart from the work of the early British, American and later, Canadian explorers, the
present data base of hydrographic information in the Canadian Arctic may be considered
to have started with Canadian Hydrographic Service northern surveys about 1908. These
surveys continued intermittently up to the war years and included surveys of Tuktoyaktuk
and Churchill harbours. From 1945 onwards, surveys were directed towards the establishment
of the Joint Arctic Weather Stations and Distant Early Warning Station supply needs.
Both American and Canadian ships worked on these programs. In 1952 the Service began
to publish a uniform series of charts at 1:500,000, completing that program in 1960.
In 1949 ice strengthened charter ships began to be used to supplement the survey effort.
In 1957 BAFFIN, a major survey vessel was commissioned and during the next two years
was used in Frobisher Bay and Hudson Strait.
In 1960 through to 1962 BAFFIN worked in Lancaster Sound. This program was reactivated
in 1971 using the icebreaker LABRADOR and continued in 1973 and 1974 using BAFFIN again.
A total area of 40,165 square kilometres was covered.
In the Western Arctic a major program was instituted with three ships working in the
Beaufort Sea in 1970. In subsequent years PARIZEAU, a medium sized vessel continued
working in the area covering a total of 49,500 square kilometres.
Continuing on through all these operations, hydrographers have regularly sailed north
each year aboard the Coastguard icebreakers. Surveys of many settlements have been
conducted when opportunity permitted.
In spite of the considerable amount of effort that has been expended since 1908 only
about 147,377 square kilometres north of the Arctic circle have to date, been surveyed
by ship. Furthermore, as will be discussed later, only a portion of this area may be
considered adequately surveyed and charted to meet the needs of the navigators of
tomorrow.
NEW TYPES OF TECHNOLOGY
A significant departure from conventional methods of hydrographic surveying in the Arctic
was made as a result of the establishment of the Polar Continental Shelf Project in
1959. The initial thrust of that operation was to acquire environmental information on
the continental shelf off the Queen Elizabeth Islands. The area was completely inaccessible
to any of the icebreakers of that time. Consequently, methods of obtaining
depth measurements beneath ice covered waters had to be developed. Initial attempts
598
Kerr 2

and hovercraft systems, almost continuous operations over at least 12 hours a day should
be possible. However, like the other developments considerable logistic support will
add greatly to the cost.
A promising system for surveying the shallower parts of the Arctic such as Coronation
Gulf, is being developed at present, with field trials expected in 1976. The system
uses an inertial platform to provide precise control for the aircraft obtaining
coloured photography. Using photogrammetry it is hoped to be able to define depth
contours to the depth of photographic penetration of the water.
Adequacy of Information
Before deciding on the job to be done and the various alternatives available, some
comments must be made on the adequacy of the information. Methods available to date
to provide a complete bathymetric description of the sea floor are either very expensive
or cover the area very slowly. Therefore, for most medium and small scale surveying a
sampling technique must be employed. Although attempts have been made to quantify the
sampling system, no successful method has been developed.
The needs for bathymetric information are various. Pipeline engineers will require
sufficient information to ensure that the pipe is well supported over a critical length.
Engineers need detailed information for building wharves and other marine structures.
The Canadian Hydrographic Service's main objective is to provide charts that permit
ships to navigate in safety. Apart from ensuring that the land itself is correctly
positioned, safety means making sure that aZZ submerged obstacZes that have a depth
between peak and the water surface at extreme Zow tides Zess than the depth of the
keeZ of vesseZs expected in that area, are Zocated and defined.
A particularly interesting case that points to the difficulty of defining a sampling
interval has been reported by Shearer et aZ (1971). In 1969 when the MANHATTAN made
its first voyage across the Arctic in company with the icebreaker JOHN A. MACDONALD
it passed over an unusual rise in the sea bottom. Subsequently, a detailed survey of
the area was carried out and revealed 78 of these mounds, now thought to be pingos;
the depths above their summits ranging from 15.4 metres to 45 metres. The shallowest
of these would be less than the draft of MANHATTAN and considerably less than the draft
of modern supertankers. Of importance to hydrographers is that the diameters of the
bases of these mounds averaged 400 metres which is considerably less than the line
spacing of sounding profiles normally used for surveys of such an area.
The provision of adequate charts for surface vessels, even supertankers, may be
problematic but the provision of adequate charts for submarine tankers requires a completely
new look at the methods and technology. Clautice and Sheets (1974) in discussing
Submarine Tanker Navigation in the Arctic estimate that a submarine tanker of
300,000 DWT will have a sail to keel dimension of 120 feet. They state that ice keels
of 100 feet are not uncommon. Allowing for a minimal clearance of 20 feet it can be
realized that we must now be concerned with obstructions reaching to 40 fathoms
(73.2 metres) of the surface. The authors go on to examine the additional margin that
must be needed when the submarine is diving or ascending due to its great length. They
point out that Barrow Strait would be particularly critical as depths in the 40 fathom
range have been reported there.
Criteria for line spacing of sounding profiles used by the Canadian Hydrographic Service
is governed by the scale of the plot which in turn is controlled by the depth and
shape of the sea floor. In Lancaster Sound in 1973 and 1974, line spacing of 750 metres
in depths less than 200 metres and 1,250 metres in depths greater than 200 metres were
used. In the Beaufort Sea PARIZEAU used a line spacing between 500 metres and 1,000
metres. If we are to ensure that we detect all obstructions, such as the pingos with
600
Kerr 4

their 400 metres average diameter, it will be necessary to re-examine the adequacy of
present survey methods.
Economics and Productivity
Several different methods and vehicles have been examined with respect to their cost
effectiveness. Hydrographers for many years have been seeking criteria to describe
cost effectiveness, with limited success. Line miles, spot soundings and area surveyed
are used in these cases.
It has been reported,Q1inistry of Transport, Canadian Coastguard, personal communication),
that a major icebreaker of the Wind class, has operational costs of approximately
$7,500 to $8,000 a day, excluding capital depreciation. The cost of the major survey
vessel BAFFIN is reported,(At1antic Oceanographic Laboratory, personal communication),
to have risen from $3,400 a day in 1962-63 to $9,350 and $10,285 for 1973 and 1974
respectively, based on a 10% annual increase from a known 1972 cost. Once again no
capital depreciation is included. Table 1 shows productivity and rates and costs over
two years when BAFFIN was surveying in Lancaster Sound.
TABLE 1. Ship Surveys
Year Days on* Daily Total Square Line + Av. Cost Av. Cost Production Production
Operation Cost Cost Kms Kms per sq.km. per line rate sq. rate km/
$ $ $ km $ km/day day
1973 57 9,350 532,950 5,833 4,403 91. 4 121.0 102.33 77 .3
1974 48 10,285 493,680 3,381 3,127 146.01 157.9 70.4 65.1
* from southern port and return
+ includes some launch sounding
It is worth noting that the cost per line kilometre can be compared with $270 per
kilometre($500 a line mile) quoted for commercial seismic surveys in 1973 (Oil Week,
November 19, 1973, p. 32). However, a replacement for a ship of BAFFIN's size would
probably cost about $20 million. Amortizing this approximately over 25 years will
add over $2,000 to the daily costs. Consequently, the average cost for a line kilometre
in 1973 and 1974 respectively would become $146.9 and $188.6.
Two other ship operations were examined in 1969 (Summary of SRN-6 Hovercraft Operations,
Internal Report 1969). Both ships were working in the Beaufort Sea. In a very successful
year with considerable open water conditions, estimates were made on operating
PARIZEAU. This ship worked in conjunction with three launches and is reported to have
surveyed over 5,000 line miles (9,250 kilometres) during 70 days of operations at the
remarkably low cost of $28 per line mi1e($15.l per kilometre). A daily ship cost of
$4,000 was used in that study. RICHARDSON, a minor vessel, 66 feet in length, is
reported to have surveyed 990 line miles (1,832 kms) at a cost of $72 a line mile
($38.9 a kilometre). It must be noted that the smaller of these ships could not be used
in ice covered waters such as Viscount Melville Sound. Even the PARIZEAU, as an ice
strengthened ship appears to have considerable limitations for working in ice.
Two years of operations with an SRN 6 hovercraft have been examined (Summary of SRN-6
Hovercraft Operations, Internal Report 1969). The operational years were 1969 and 1970
and for comparison purposes these costs have been projected to 1974 using a 10% annual
increase. Table 2 outlines production and costs. Both year's operations were in the
601 Kerr 5

Beaufort Sea.
Year
1969
1970+
1974*
Days on Total
Operation Costs $
58 231,112
64 254,216
61 355,283
+ Projected cost
TABLE 2(a). Hovercraft Surveys (SRN 6)
Square
Kms
3,494
2,674
3,084
Line
Kms
3,114
2,825
2,970
Av. Cost
Per Sq.Km.
66.2
95.1
115.2
* Figures averaged and projected from 1970 to 1974
Av. Cost
Per Km.
74.2
90.0
119.6
Prod. Rate
Sq.Km/Day·
60.2
41.8
50.6
Prod. Rate
Km/Day
The above figures for the SRN 6 hovercraft include rental, fuel and transportation.
They do not include accommodation costs for personnel or for possible helicopter or
fixed wing support that may be required.
A more recent model of air cushion craft, the Viking 7501, has been examined using
figures provided by its manufacturers. The cost data have been looked at in the context
of the early SRN 6 Surveys in the Beaufort Sea.
TABLE 2(b). Hovercraft Surveys (Viking 7501) Estimated
Total Costs $ Sq. Kms Line Kms Av. Cost/Sq.Km Av. Cost/Km
386,567 3,084 2,970 125.4 130.2
In this case, the total hovercraft cost was estimated at $230,099 and $156,468 was added
for aircraft support. This represents a more realistic cost assessment than those
provided in Table 2(b). In discussions with a hydrographer who had worked on the SRN 6
program he was of the opinion that if the fixed strut had been perfected at the time a
very high rate of data collection could have been achieved.
For tracked vehicle costs for over ice operations we must use estimated costs. A report
made in 1973 (Oil Week, November 19, 1973, p. 32) on costs for over ice seismic
operations gives $3,000 to $3,500 a line mile (1,622 - 1,892 a kilometre). These
apparently high costs are of course for the more complex seismic operation rather than
for bathymetry. Note may also be made of a reported production rate of 10 miles per
day for over ice seismic operations (Oil Week, November 19, 1973, p. 32). The following
estimate has been developed recently for a survey utilizing two and four tracked
vehicles plus helicopter and fixed wing support for bathymetric surveys.
Year No. Days on
Vehicle Operation
1976 2 70*
1976 4 70*
TABLE 3. Over Ice Tracked Vehicle Surveys
Total Sq.
Costs Kms
$
3.5
Mill.
4.5
Mill.
Line Av.Cost
Y.ms Per Sq.Km
5,000+
10,000+
53.7
44.1
48.7
Av.Cost Prod.Rate Prod.Rate Cost Per
Per Km Sq.Km/Day Km/Day Sounding
700 70 70
450 140 45
* Operating season is approximately mid-March to beginning June.
+ At present spot soundings along profiles. Soundings spaced 100 metres apart.
602 Kerr 6

The estimates consider an operating crew of about 30 persons. Twelve hours of work per
day are guaranteed. Crews are exchanged every two weeks. All data reduction costs have
been included. Since mobilization costs are included in these estimates, it is probable
that second and future year costs will be less. All capital equipment costs are
amortized.
Through ice spot sounding methods using helicopter support has been examined for surveys
in Eureka Sound in 1974 and Nares Strait in 1975. These data are shown in Table 4. All
helicopter, fuel, transportation, accommodation and fixed wing aircraft support are
included. Instrumentation costs are not included.
TABLE 4. Over Ice Helicopter Supported Surveys
Year Days Total Number Cost $ Sq. Kms Av.Cost Prod.Rate Normal Range
Operation Cost $ Soundings Sndg. Surveyed Per Sq.Km Sq. Km/Day Grid Grid
Spacing Spacing
1974 73*(46) 355,305 4,189 84.82 12,925 27.49 177.1 2Km 1/2-3 Kms
1975 58(30) 458,005 2,988 153.28+ 18,225 25.13 314.2 3Km 1-4 Kms
* Total days of operations are included in calculations. Figures in parentheses are
actual sounding days.
+ The higher costs for 1975, apart from inflationary increase are due to the operation
being in a more remote area and also aircraft rates were re-negotiated and substantially
increased.
The surveys employed 3 Bell 206B helicopters for sounding and DC 3 and Twin Otter support.
Helicopter rates increased by 47% between the two years and 10% and 25% for fixed wing
DC 3 and Twin Otter support, respectively.
Discussion on Comparison of Systems
Summer operations in ice free or partly ice free waters should not be compared with
spring operations from the ice surface. They may in fact, be considered complementary
by permitting a longer overall operating season.
Since 1969 several of the large vessels of the Canadian Hydrographic Service have been
used to collect gravity, magnetic and other geophysical data as well as bathymetry.
Measurements with shipborne gravimeters require a quite stable platform, generally only
possible by using the larger ships. Due to the constant changes of course and changes
in velocity, gravity measurements from a ship working in ice are not practical. In
comparing a platform such as a hovercraft it should be realized that it is not capable
of carrying the variety of sensors as is a ship.
The method of spot sounding from the ice using helicopters for transportation may be
compared with the tracked vehicle. At present the two systems are seen as having different
applications. The helicopter systems provide a very general picture of the
bathymetry, whereas the tracked vehicles appear to be most suited to surveying limited
areas in great detail. It can be expected that if the helicopters were employed on
closely spaced soundings of perhaps 100 metres apart that the rate of sounding measurement
would increase considerably. At the present rate, if soundings were converted
into line kilometres 100 metres apart, the total line kilometres would have been 418.9
in 1973 and 298.8 in 1974.
603 Kerr 7

as it could be used as soon as any open water appears. The ship on the other hand may
often be barred by ice on the way to the survey area. Examples of this are the
constraints to shipping when passing Point Barrow, through Hudson Strait and in Baffin
Bay.
As it was remarked earlier, the ship can be employed for multi-parameter surveys whereas
the hovercraft is not so capable. On the other hand gravity measurements can often be
measured more effectively from the ice surface during the spring than from a ship in the
ice during the summer.
The tracked vehicle system also has a different operating season from the ship. The two
can be combined to extend the operating season for perhaps five months of the year.
Alternatively, the tracked vehicle and the hovercraft might extend the season to six
months. The very high costs of the tracked vehicle system at present are discouraging
and it is to be hoped that the present models can be improved both with respect to the
rate of data collection and a reduction in costs. It should be noted that the tracked
vehicle system as presently configured is capable of collecting very useful geological
information as well as bathymetry.
The writer has proposed in an earlier paper, (Proceedings of the Twelfth Annual
Canadian Hydrographic <strong>Conference</strong>, 1973, p. 307), that a total change in the planning
of surveys will provide the bathymetric information required in the most timely manner.
Traditionally, hydrographic surveyors have surveyed large blocks at a time, surveying
the waters from the coast out to the deep water. This is the most appropriate method
for southern waters where chart users are, many and various. The yachtsman is
interested in the shallow coastal waters, the deep sea ship is concerned about the deep
channels and the fisherman is interested in the shallow offshore banks. In the Arctic
the shipping pattern is different. Although the occasional ship may land an oil rig or
supplies at some remote beach, most of the traffic travels in the deep water from
settlement to settlement, from mine to the open sea, and in the future, from oil well to
the open sea. In order to provide these customers with detailed bathymetric information
before the heavy shipping starts the only method is to concentrate on the surveying and
charting of navigational corridors, superimposed over a much wider general bathymetric
coverage.
The corridor concept has been criticized by some of the hydrographers themselves. They
feel that the corridor is a second best and a compromise to the fine standards of the
nautical chart maker. The writer believes that such criticism is the result of a misunderstanding
of the system. The best analogy that can be used is that it is better to
build a two-lane highway linking two cities and extend it to four and then six lanes
as the traffic gets heavier, than to build a six-lane highway from one city that ends
in the middle of the fields and does not reach the other city.
Adding support to the corridor proposal is the fact that in many parts of the Arctic,
including the Beaufort Sea and Barrow Strait, nothing less than total bottom coverage
using sonar or very close echo sounding profiles will suffice. Such coverage over
large areas is today, and in the near future, impractical.
One argument raised against corridor charting is the fact that ships in Arctic waters
often must redirect their courses to avoid heavy ice. Against this it must be said that
the corridors would initially be located where the least ice is present each year. In
the Beaufort Sea this may be very difficult to predict, but amongst the Arctic Islands
there are well recognized ice patterns. For instance, in Viscount Melville Sound the
most likely open water areas exist close to Melville Island. In the Beaufort Sea it may
be necessary to either survey very wide corridors or perhaps alternative inshore and
offshore routes. It is proposed that the corridor width should normally be twice the
probable navigation error of ships navigating through them.
605 Kerr 9

In Table 5 a number of corridors linking known key points have been listed. (Table 5,
Proposed Arctic Corridors
It will be seen that a line spacing of 100 metres, or 10 lines to the kilometre has been
used throughout. This would provide far more thorough coverage than the surveys of
today which use line separation 7 to 12 times wider than this. The total number of line
kilometres can be seen to be 306,870. Using a season ship rate of 4,000 kilometres
this would entail 77 ship seasons at a cost of 38.5 million dollars, just under 60% of
the cost of the total survey method. If the ship method was combined with four units
and two ships were employed, the annual production rate could be raised to 18,000
kilometres per year. The task could then be completed in 17 years at a cost of 5.5
million dollars a year for a total of 93.5 million dollars. It can be readily seen
that the tracked vehicle system while adding greatly to the productivity is also adding
greatly to the cost. But ships suitable for the task are not now available and will
take several years to complete. The tracked vehicle could be in operation next season.
Hovercraft also can be acquired on much shorter notice than ships and appear to have
excellent potential.
CONCLUSION
It is hoped that this paper has spelled out the awesome task of providing adequate
bathymetry and consequently safe navigation, for the future shipping in the Arctic.
Unless changes are made in Canadian government policy it is evident that there will be
major resource development and consequently great increases in the size and numbers of
ships in Arctic waters. The task of providing these ships with the information they
need requires not only a major increase in financial commitment but also some rethinking
of existing methodology. Finally, it will be noted that no discussion has
taken place on untested methods. Although submarines, unmanned submersibles and other
exotic systems and platforms have been proposed over the years, only state of the art
systems have been considered in this discussion.
REFERENCES
Atlantic Oceanographic Laboratory. Personal Communication.
Clautice, W.G. and Sheets, H.E. 1974. Submarine Tanker Navigation in the Arctic.
Marine Technology Society Journal, Vol. 8 No. , pp. 29-37.
Eaton, R.M. 1963. Airborne Hydrographic Surveys in the Canadian Arctic. <strong>International</strong>
Hydrographic Review, Vol. XL No.2, July, pp. 45-51.
Facts from Canadian Maps. 1972. A Geographic Handbook, Energy, Mines and Resources,
Surveys and Mapping Branch, pp. 28.
Jollymore, P.G. 1971. A Portable Digital Sounding System for Arctic Use. <strong>International</strong>
Hydrographic Review, Vol. XLVII No.2, July, pp. 45-51.
Kerr, A.J. and O'Shea, J. 1973. Planning and Technology of Future Hydrographic
Surveys in the Arctic. Proceedings of the Twelfth Annual Canadian
Hydrographic <strong>Conference</strong>, pp. 307.
Knudsen, D., Pugh, D.
Sounding.
1975. Field Report, Rea Point Experiments in Through the Ice
Internal Report, Canadian Hydrographic Service, May.
Ministry of Transport, Canadian Coastguard. Personal Communication.
Oil Week, 1973. Ice Dictates Arctic Seismic Work, November 19, pp. 32.
606 Kerr 10

Pulkkinen, H.W. 1973. Hydrographic Surveying with air cushion vehicles, trials
report. Polar Continental Shelf Project, pp. 46.
Shearer, J.M., MacNab, R.F., Pelletier, B.R. and Smith, T.B. 1971. Submarine Pingos
in the Beaufort Sea. Science Vol. 174, November, pp. 816-818.
Summary of SRN-6 Hovercraft Operations on Hydrographic Survey Operations. 1969.
Internal Report, Polar Continental Shelf Project.
Yeaton, G. 1969. The Utilization of a Hovercraft and the Fixed Strut Sounding
Assembly in the Canadian Arctic. Internal Report, Polar Continental Shelf
Project.
607 Kerr 11

ESTIMATING POWERING REQUIREMENTS OF ICEBREAKING SHIPS IN THE DESIGN STAGE
Jack W. Lewis
President
ARCTEC, Incorporated
Columbia, Maryland
United States
EXTENDED ABSTRACT
The state-of-the-art for prediating the powering requirements of iaebreaking ships while
still in the design stage is reviewed and a proaedure whiah aan provide varying levels of
aonfidenae in these prediations is given.
A aritiaal review of reaent literature in this field is made from the point of view of a
designer who must take a set of owner's requirements and design the best ship he aan whiah
meets these requirements. Normally an owner will speaify suah items as trade routes. enduranae.
speed levels to be maintained. aargo aapaaity. mission. eta. The designer must
then design a ship whiah meets these requirements at the lowest possible aost to the owner
while satisfying all the neaessary safety rules and regulations and navigational restriations.
The basia starting point for the designer is the seleation of a hull form and its
proportions. Onae this has been aaaomplished. an estimate of the powering requirements
must be made in order to determine if the required propulsion maahinery will fit into the
hull and if the fuel aapaaity of the ship is suffiaient to satisfy the owner's enduranae
requirement. In the early design stages it is useful to have a mathematiaal proaedure
available whiah aan give approximate powering prediations so that the designer aan
quiakly determine whether or not it is feasible to satisfy the given set of owner's requirements.
If these requirements aan't be satisfied then they must be revised. If they
aan be satisfied. then the designer will normally turn to more aaaurate powering prediation
teahniques before aompleting his design.
Some mathematiaal proaedures are available whiah aan give approximate answers to the
iaebreaking powering problem. These proaedures are reviewed and a speaifia aomputational
proaedure is suggested. The aonfidenae level assoaiated with the mathematiaal proaedure
unfortunately is not as high as the designer might like and it is good praatiae to have
more aaaurate prediations made based on the results of physiaal model tests. Proaedures
for the aonduat of suah tests are also reviewed and a speaifia proaedure is suggested.
The paper aonaludes with a series of reaommendations for improving the state-of-the-art
in prediating the powering requirements of iaebreaking ships.
ABSTRACT ONLY, AVAILABLE
609

Fig. 2.J
Fig. 2.2 Young ridges in
the northern part of the
Baltic
Weathered ridges in the northern part of the Ba ltic
615
Makinen et aZ. 5

After the ship was extracted she was backed again 2 to 3 ship lengths and a new ram was
started. Several rams were needed to penetrate bigger ridges. After the ridge was penetrated
the next ship was surveyed and the ship was moved towards it.
Traversing a ridge usually requires several rams. Each ram consists of four phases: the
accelerating of the ship in a broken channel, the penetrating (decelerating) phase in the
ridge, the extracting of the ship from the ridge, and the backing of the ship for the
next ram. Detailed measurements were made only during the penetration phase whereas of
the other phases nothing else but the time needed for them was registered.
A test in a ridge took on the average a little less than one hour including the time
needed for measuring the ridge profile. In addition to that a long time was spent in
moving the ship close to the next ridge suitable for tests, because the whole ice field
in which the tests were performed was heavily ridged and the ramming mode of movement
had to be used here as well.
Figure 5.4 presents a ridge photographed from the ship's wheelhouse before the first ram.
As can be seen, the ship is accelerated for the ram in broken ice.
Tests in level ice present no problem as regards the synchronization of the data on ice
conditions with those on the ship's speed and propulsion machinery performance in case
the type of ice conditions is constant. In ridge tests, the synchronization is important
because the ridge thickness is not constant and because the length of the ridge cross
section is usually of the same order of magnitude as that of the ship. Special parameters
have been developed to facilitate the research on the influence of the ridge characteristics
on the ship's ice resistance. These parameters, i.e. the effective length of
the ridge cross section and the effective ridge thickness (the accurate term: average
ridge thickness for effective length of ridge cross section) are defined in Figures 7.1
and 7.2.
MEASUREMENTS DURING TESTS
The ship's ice resistance and speed are the desired output values of the penetration
(deceleration) phase. Because the ship's speed usually slows down very quickly when she
is ramming into a ridge, all the readings had to be registered on recorders so that the
values of any moment could be read later.
A modified traffic radar was used for measuring the ship's speed. It was equipped with
an UV-recorder, which gave the speed vs. time in graphic form. Because this was the
first time the radar was employed for this purpose the conventional speed measuring system
with a stop watch and sighting poles was also used. The traffic radar functioned
well. Figure 8.1 shows the radar antenna on the main deck, and the UV-recorder for speed
registration can be seen in Figure 8.2. An example of speed recording is given in Figure
8.3.
For determining the ice resistance the propeller revolutions of all the four screws had
to be known. The revolutions were registered by two UV-recorders, two shafts per one
recorder. In addition, the voltage and current in the electrical propeller motors were
registered for obtaining data for the calculation of the machinery power as a check of
the other measurements. Figure 8.4 shows and example of the propeller revolution recordings
during a ram.
A switch operated by hand was used for synchronizing the speed and revolution measurements
with the progress of the test.
Through a series of raMS the time used for each phase was recorded. Also this measurement
was synchronized with the others.
628 Makinen et al. 18

Fig. 8.1 The antenna of the traffic radar attached
to the deck
Fig. 8.2 The UV-recorder in the wheelhouse for reg-
istering the ship's speed
631 Makinen et al. 21

Thus the data collected during the trial consisted of recorder rolls of three recorders,
a detailed timetable of the tests with the different phases and of forms with the results
of the ridge profile measurements.
TREATMENT OF MEASUREMENTS
The parameters chosen for the treatment of the results, such as the speed at which the
ship enters the ridge, penetration (the distance penetrated), and the effective ridge
thickness are shown in Figures 7.1 and 7.2.
During the penetration phase recordings were read and tabulated at two-second intervals.
The speed and propeller revolutions varied so widely that by using a longer interval the
needed accuracy would not have been achieved.
In connection with the ship's open-water trial which was performed before the del ivery
in 1970, the bollard thrust and the open-water speed vs. propulsion power and propeller
revolutions were measured. Using the results of these measurements and those of the
ship's model tests in open water, the diagram of propeller thrust vs. propeller revolutions,
with ship speed as a parameter, was drawn for determinging the thrust. This diagram
is shown in Figure 9.1.
The total resistance can be expressed as fol lows:
where
R tot = T + ma
T total propeller thrust (the thrust of the four propellers added together)
m ship's mass
a deceleration
ma mass force
Usually the mass force was dominant.
The ice resistance in a ridge (i .e. the resistance due to the ice in a ridge) is obtained
by subtracting the open-water resistance from the total resistance. The open-water resistance
values measured in the ship's open-water trial are used.
The added water mass was not included in the mass force because its magnitude is not
known for a ship penetrating a ridge. It is probable that it is smaller for a ship
completely inside a ridge than for one in ice-free water because almost the whole hul I
is in contact with ice. On the other hand the significance of the added ice mass is
completely unknown . These both may create inaccuracy in the results.
After all the data of individual points (measurements at two seconds' intervals) had been
analysed, the results were collected into a convenient form, Figure 9.2. This form consists
of al I the information needed for further analysis, including the data on the characteristics
of the ridge. It includes the momentary values from the rams as well as the
time history of the whole series of rams in a ridge.
For investigating the influence of ridge characteristics and ship speed on ice resistance
the results obtained from the measurements of all the ridges were combined at three
levels:
1. Mrnnentary values of the deceleration rhase; for determining the ice resistance
in ridge vs. ship speed, penetration the distance penetrated) and ridge charac-
634 Makinen et aZ . 24

100
90
80
70
50
50
1.0
30
20
10
o
PROPE LLER THRUST OF
2 PROPELLERS
[10' N J
50 100 150
PROPELLER RPM
Fig . 9 .1 Thrust - propeller revolution - ship's speed
curves used in the analysis
635
Makinen et al. 25

Fie· 10.3
Ice resistance in
ridge vs.
thickness.
values
series
ICE RESISTANCE
• PENETRATION
110' Nm I
300
200
o
of
of
ridge
Average
each
rams
ICE • MEAN VALUE IN A RIDGE I E
RESISTANCE
MEAN VALUE OFA SERIES OF RAMS
110' N )
300
200
100
o 2
• • MEAN VALUE IN A II, 00£
IE "'UN VALUE OF" A SERIES OF R.tJ45
•
500 1000
AREA OF RIOGI: PI/OFILE 1m')
•
•
I.
•
•
•
•
34567
EFFECTIVE RIDGE THICKNESS 1m)
Fig. 10.4 Ice resistance
integrated over penetration
vs. ridge thickness inte­
grated over penetration.
Average values of each
series of rams
•
•
Makinen et aZ. 29

This paper is the third in a series delivered to the POAC <strong>Conference</strong>. The first dealt
with Marine Transportation in Alaska's Bering Sea and Aratia Oaean Areas; the second
with The Comparison of Off-shore and On-shore Terminal Faaility Loaations on Alaska's
Northwest Coast.
This third paper deals with the present marine transportation options in Northern Alaska.
It is intended to direct research and program efforts in the immediate future,
for it is in the immediate future that Alaska will have to face some rather extreme
pressures on its transportation systems that serve the Arctic areas if the federal and
private building toward the continued institution of massive energy-related programs
continues over the next decade.
The level of involvement of marine commerce in satisfying these transportation needs
will need a careful reappraisal now that Alyeska Pipeline Service Company has constructed
a secondary State highway from the Yukon to Prudhoe Bay. When the bridge across
the Yukon is ready for use in October of 1975 it will be possible to serve the central
areas of the North Slope by truck from Fairbanks or anywhere else in North America that
is served by roads.
Thus we are faced with several alternate means of serving the entire North Slope from
the traditional sealift by barge or airlift by Hercules and other cargo aircraft.
Before completion of the road, traffic analysts computed that between cargo airlifted
from Fairbanks having reached that city by sea, rail or highway truck and on cargo that
was barged to Prudhoe Bay and then airlifted south, the even cost line was about 75
miles south of Prudhoe Bay. (1) This, of course, varied east and west with distances
to the two Alaskan bases, Prudhoe Bay and Fairbanks.
The major cost increment in the sealift route through Prudhoe was in lightering and
moving goods to the various sites. The major cost for the Fairbanks route was in the
airlift and a large subsidiary cost in the first years was the transfer from the railroad
sidings to the airport.
The amounts of freight moved by these two systems since 1968 has not been tabulated exactly,
but reasonable figures are available. Until the road was completed for use in
October of 1974, about 375,000 tons was shipped by air through Fairbanks. A good part
of this was petroleum products and other life support necessities. During this same
period about 25,000 tons was shipped by ice road through Fairbanks, including the weight
of construction vehicles moved over the ice roads.
In contrast, about 650,000 tons came in through Prudhoe Bay. Much of this was outsized
equipment, large modules for the oil field operators, pipe for the line, and other
packages far too large to move by aircraft. The total cost of the sealift from Seattle,
Houston or other ports was about $75 million; that of the airlift about $100 million.
It must be remembered that these figures are estimates and only partially a result of
records search.
Thus, we have around one million tons moved into the Arctic areas at a cost of some
$175 million for direct transportation costs. Indirect costs are somewhat more difficult
to arrive at but come to about $105 million expended as follows:
(1) Studies conducted by the Federal Field Committee for Development Planning in
Alaska and the Federal Aviation Administration in file reports.
646
Parker 2

that cannot be hauled by air or truck. The cost savings involved in constructing these
modular units in lower cost areas would probably not be made up by any transportation
saving brought about by shipping them in smaller units that could be hauled by truck
and assembling on site. The question then becomes - how much should the marine transportation
system be improved to increase its efficiency vis-a-vis the road to the
Arctic?
The major drawback to increased reliability and more frequent shipping service is obviously
ice. The dislocations to other areas caused by accumulation of the annual
barge fleet for its once a year haul is a factor that should be weighed into any investment
in ice-breaking capability. This would make conceivable an icebreaker whose
amortization and operations costs equaled the savings in the reduction possible in the
barge fleet, reduction in insurance premiums and reduction in inventory costs. It is
doubtful if such tradeoffs are possible on cargo flows of less than 200,000 tons per
year in marine commerce. Thus, a $100 million icebreaker with yearly costs of $10
million would add $50 per ton to the costs, a figure which it would be difficult to
reach with savings just outlined, but which possibly might be reached.
It is obvious that thus far the discussion has been based on petroleum development in
the Central North Slope. The 35,000 tons a year to serve the private and military sectors
in the Arctic is negligible in these computations and will tend to remain a high
cost, inefficient system because of the scattered points to be served. As pointed out
in previous papers, an improved lightering service would help to reduce costs, but a
system such as a Sky Crane helicopter might add to present costs unless real utilization
efficiencies could be obtained.
When resource development is shifted west into the Naval Petroleum Reserve (Pet) 4 area,
a different equation might be used. As one gets over to the Chukchi Sea coast it becomes
obvious that some consideration must be given to tanker alternatives as against
pipelines down the main spine corridor from the Central North Slope to the Gulf of
Alaska.
In a study done in 1971 for the Corps of Engineers, the author estimated cargo flow to
the North Slope at some 350,000 tons per year if a moderate level of exploration continued
over the North Slope, including Pet 4. Cargo flows of 795,000 to 1,095,000 tons
were estimated for the Seward Peninsula, Kobuk Valley and the North Slope. The bulk of
the cargo was 565,000 tons of ore from the Lost River and Bornite areas. (3) Bornite
ores could be captured by the Arctic Road if the mine operators were forced to reach a
unilateral transportation decision.
If major oil finds similar to Prudhoe are found in the western area, they will be able
to utilize either central pipeline or tanker alternatives. The economic margin in an
oil field of this size can justify great costs and port investments of $2 to $3 billion
would be conceivable if tankers proved to be the best alternative.
The coal reserves of western Alaska, the smaller oil fields and the other minerals do
not possess this margin as yet. It becomes obvious that development of other resources
will ride the back of petroleums investments and therefore port systems should be designed,
if possible, with this in mind.
(3) Walter Parker, Dale Swanson, et aI, Northwest AU2ska Eaonomia and Transportation
Prospeats, for the Corps of Engineers, Alaska District, University of Alaska,
Fairbanks, March 1972, P. 129-130.
648 Parker 4

There will be 180 million tons of oil per year through the trans-Alaska line. The natural
resource cargo flows of western Alaska could, in the next 20 years, be anywhere
from zero, where they are now except for a few tons of fish, to 200 to 300 million tons
of oil, coal, LNG and other minerals. Obviously, resource flows of this magnitude make
possible many options.
In the eastern part of Alaska's Arctic entirely differing problems are at hand. Here,
surface transportation options must relate to the central spine, new pipelines and
roads through wilderness areas along the U.S.-Canada border, or going east through Canada.
There have been few alternative scenarios developed for this area, other than the
Alaska Gas Arctic proposal, because it is dominated by the Arctic Wildlife Range and
similar Canadian proposals to the east.
Whatever happens in this area will probably have little relationship to marine transportation
due to the rapidly developing road system to the south on both the U.S. and
Canadian sides. There is also the consideration that this area is even more adapted
towards pipelines to the Gulf Coast or through Canada than to the central part of the
North Slope between the Colville and the Canning Rivers where present development is
concentrated.
A certain grand pattern is beginning to become clear for Arctic development in Alaska.
Certain basic assumptions regarding markets are necessary in order to accept this pattern,
namely:
1. Oil and gas resources will be reserved for North American use.
2. Coal is available for export to all Pacific Rim users.
3. Other minerals are available on a limited basis to other than North
American markets.
Considering these assumptions as relatively valid, it becomes clear that pipelines to
ice-free ports on the Gulf of Alaska or overland through Canada will maintain their
pre-eminence as alternatives for oil and gas except for the western fringes of Alaska.
Certainly, any oil or gas located offshore will have a strong bias towards transportation
by tanker, especially in the Bering Sea where ice breaking costs will be minimally
less than north of Bering Straits.
The same conditions could prevail for oil or gas located immediately on shore.
Coal will probably seek a marine alternative rather than face an 800 to 1,000 mile haul
to a Gulf of Alaska port by rail. The same would hold with other hard minerals in
western Alaska.
It seems probable that an even cost line between shipments destined for Gulf Coast
ports and those destined for Bering Sea ports will develop rather rapidly. The role
of public financing is critical in the location of this even cost line. If roads are
treated as a free good to be provided by the government and rails or ports must seek
private finance, it is obvious that a true even cost line will not be obtained.
In any case, the development of substantial parts of western and northwestern Alaska
will rest upon the technology developed to provide port structures on these shallow
coasts that will accommodate large tankers and/or carriers. Environmental safety is a
prime concern that mitigates against marine oil shipments but is more neutral on coal
and ores.
649
Parker 5

The entire Bering and Chukchi ecosystems must have a superlatively high level of environmental
protection if the sadly reduced stocks of the Bering are to be replenished.
If a 200 mile fisheries jurisdiction includes pollution control, then both seas will be
largely divided between the Soviet Union and the United States. True protection will
only come if each has stipulations on construction and pollution control regulations
that are equally strict.
A maximum development regime would see 14 large crude oil carriers, 10 LNG tankers and
about 7 ore carriers present in the Bering and Chukchi Seas at all times. A similar
development on the Soviet side would double those figures.
There must be a tradeoff between environmental control and insurance costs in this area
that will result in substantially the same eventual costs since environmental protection
is largely a matter of ship safety insofar as major oil spills are concerned.
Hopefully, decision makers at that time will have sophistication enough in their planning
models to incorporate the positive cost saving aspects of maximum environmental
safety engineering rather than simply railing at the costs.
If this is not done and maximum production is used as an excuse for not incorporating
these protection measures in engineering programs from the first, there will be little
that can be accomplished by a strict regulatory regime imposed later other than a continuing
running battle between resource developers, either private or governmental,
and environmental regulators. Later imposition of the necessary engineering to make
satisfactory adjustments will be more expensive than doing it right the first time.
This is one of the major lessons learned from the State of Alaska's experience with
offshore platforms in Cook Inlet and tanker operations from that area thus far. There
are hopes that at the Valdez terminal and its associated marine transportation system,
we have made some incremental gains in developing a system where a reasonable level of
environmental balance is maintained, although a good deal of work remains to be done on
the marine operations of that system.
The major areas of concern for coastal development along the Beaufort and Chukchi Seas
have been outlined in the previous papers referred to at the beginning of this paper.
Some of the offshore problems are also detailed but should be restated here and some
additions made based on our experience of the past two years.
Treatment of ballast water in a totally offshore environment will be expensive but must
be more and more rigorous as development regimes increase in scope. A certain level of
development may permit 8 p.p.m. and the next level require reduction to 5 p.p.m. since
the important criteria is the total amount of oil released into the seas.
Totally separate ballast can certainly be a cheaper alternative than engineering treatment
plants with enough flexibility to meet ultimately very low separation standards.
The money must be spent on the tanker or on the facility in any case.
Navigation in multi-year ice will require tankers than can handle their own immediate
problems when icebreaker support is not immediately available. A review of the S.S.
Manhattan experience with multi-year ice will lead to the next level of experimentation
that is necessary in developing steering capabilities in heavy ice.
Incremental losses will be concentrated in the areas of loading, and loading areas
should probably offer reasonable protection against sudden ice changes. The alternative
of disconnecting in serious ice conditions is not attractive from the standpoint
of minimizing incidental losses of oil.
650
Parker 6

Personnel and crew capability will become more of a problem the farther north one goes.
One of the major problems with our present development regime is the inability of private
or government entities to secure large numbers of people with northern experience.
This is not the case with DEW line construction and previous efforts where a fairly
high level of Arctic expertise was constantly present.
More than any other factor, the presence in each onshore and offshore area of a project
oriented multi-disciplinary team to review stipulations and work programs will
raise the level of environmental engineering by bringing to new entrants the benefit of
past experience. Doing this can probably accomplish more than all the legislation in
the world. The legislation is necessary to provide the authority and the regulators to
impose penalties but without that critical third group the rest is somewhat of an empty
barrel which will resound with vigorous declamations but do little to protect the Arctic
as it enters into higher and higher levels of activity.
651 Parker 7

THE SEMI-SUBMARINE ICEBREAKING TANKER, AN ALTERNATIVE FOR TRANSPORTING
CRUDE OIL OR SIMILAR HEAVY BULK MATERIALS IN ARCTIC WATERS
Per C. Sandnles
Aker Group
Oslo, Norway
ABSTRACT
This paper presents the outcome of a preliminary design and feasibility study aimed at
finding the most suitable vesseL for transporting Arctic crude oil. The study was initiated
at the Aker Group in Norway shortly after the discovery of significant quantities of
petroleum on the North Slope of Alaska. Three principal conaepts were considered. The
assumptions and calauLations leading up to the seLection of the SSIT, a type of semisubmcwine,
as the most promixing alternative are outLined. Unit cost of transportation
is evaLuated based on estimated vessel data and assumed environmental conditions and ice
resistance coefficients. It is concluded that the SSIT, as described in this paper,
seems to provide the most suitabLe method for transporting crude oiZ from the North Slope
or the Canadian Arctic Islands to the prinaipaZ markets on either side of the North
AtLantic.
653

INTRODUCTION
During 1970/71, several methods of transporting crude oil in Arctic waters were studied
by the engineering staff of the Aker Group, the largest builder of ships and offshore
structures in Norway. The main objective of this study was to arrive at a preliminary
proposal for an icebreaking tanker with the capability of year round operation under
Arctic conditions. The most promising concept, referred to as the semisubmarine icebreaking
tanker, or the SSIT, is presented in this paper.
A fairly extensive technical and economical feasibility study was carried out. The main
tool for this study was a computer program specially developed for evaluating transportation
costs for various alternatives based on published ice resistance formulas and
coefficients. Method of analysis is outlined and principal assumptions and conclusions
are summarized.
The three principal design objectives were: 1) construction and operation as far as
possible within reach of available technology, 2) maximum safety against pollution of
the Arctic environment, and 3) minimum unit cost of transportation, for instance as
measured in $/bbl.
The first objective, and possibly the second one too, early seemed to rule out the
nuclear submarine alternative. Together with the third objective, they seemed to imply
a surface vessel with the capability of propelling itself through the ice with minimum
amount of resistance, engine power and fuel consumption.
5:7
CATAMARAN TYPE SEMISUBMARINE ICEBREAKING TANKER
SEMISUBMARINE ICEBREAKING TANKER (SSm
FIGURE 1. PRINCIPAL CONCEPTS
654 Sandn

GENERAL DESCRIPTION OF MAIN CONCEPTS
Initially three main concepts as shown on Figure 1 were considered. The first one, the
icebreaking tanker, hereinafter referred to as the IBT, is basically a conventional
surface vessel with exceptionally powerful propulsion machinery, extensive ice
reinforcements, and hull and bow shaped as to facilitate easy passage through ice.
The second one has several features in common with the main, final one, the SSIT. The
hull consists of a main, submersible cargocarrying section and two watertight superstructures
extending above maximum permissible waterline in order to provide reserve
buoyancy and longitudinal stability. Both the forward and aft superstructures are
connected to the main hull by two transition sections spaced transversely as far apart
as possible. The main reason for this and for making the entire vessel exceptionally
wide is to obtain sufficient transverse stability at the maximum draft waterline with
the aid of the waterplane inertia provided by the transition sections. On the other
hand, the transition sections should be as narrow as possible in order to reduce ice
resistance.
The SSIT at the bottom of Figure I has each superstructure connected to the main hull by
just one narrow transition section located on the centerline of the vessel. This way,
the waterplane inertia contribution to transverse stability is negligible at maximum
draft. Similar submarines, positive stability can only be attained by center of gravity
being located below center of buoyancy. This is obtained with the aid of permanent
ballast at the bottom of the vessel, for instance 40 000 tons of concrete or liquid mud,
and a dry compartment along the top of the lower hull. Piping, normally run on top of
the main deck, is located inside this dry compartment which also serves as an access
corridor between the two superstructures.
Despite the permanent ballast, the apparent advantages of the third concept were
considered significant enough to justify exclusion of the second one from any further
analysis. For instance, two ice cutting edges at the forward end instead of one give
considerably greater ice resistance. Furthermore, location of these off centerline
implies continuously changing inclining moments due to the ice cutting process. If the
high density permanent ballast is sought eliminated in favour of more payload, hardly
any space suitable for additional cargo oil is available. The superstructures cannot be
used due to stability considerations and the double shell compartments cannot be used
without substantially reduced pollution protection.
FIGURE 2. ARTIST' S IMPRESSION
655 Sandnaes 3

DESIGN WATER LINE 37.5 m
SECTION THROUGH MIDSHIP SECTION
ENGINE ROOM
FIGURE 4. TRANSVERSE SECTIONS
STRENGTH
BALLAST WATER
Steel weight as derived from preliminary strength calculations is an important parameter
for determining both construction cost and stability. These calculations are based on
the preliminary structural arrangement of Figure 4. (The implications of the additional
antipollution measure of extending the double skin all around the lower portions of the
main hull have not been assessed at this stage.)
Materials for most of the internal structural memb 2 rs will be normal shipbuilding steel
with a minimum required yield strenght of 240 N/mm (34 000 psi). Highly stressed structural
parts like the shell plating and stiffeners in waY20f ice zones are proposed made
of high strenght steel with a yield strength of 360 N/mm (51 000 psi.) Low temperature
steel qualities, mainly classification society grade E steel, is expected to be used
extensively, especially in the exposed portions of the outer shell.
657
Sandna!s 5

Considering the fact that the sloped sides should make it almost impossible for the ice
to exert full pressure perpendicular onto any portion of the main hull, it is felt that
proposed scantlings provide sufficient local strength margin. Certain exposed areas,
like the bow and the superstructure sides, need additional strengthening. This can be
provided for instance by using heavier steel plates, higher strength steel, additional
intermediate frames and stiffeners, heavy steel castings at sharp points and along
knuckle lines, concrete backing, etc.
These strength considerations have resulted in the weight data listed in Table 1.
A rough, preliminary steel weight have been arrived at for the IBT by assuming an
addition of 50 percent of ice reinforcement to the steel weight of a conventional
tanker. This implies that the SSIT is 30 percent heavier than the IBT and 90 percent
heavier than the conventional tanker.
STABILITY
This section deals with ability of the vessel to always remain afloat and in fairly
upright condition even when damaged or loaded by heavy ice.
Intact Stability
Stability calculation results are summarized in Table 2 and Figure 5. The indicated
ample stability is the result of an initial effort to maximize the metacentric height.
This effort includes minimizing free surface effects, minimizing the volume of, or
raising the center of gravity of necessary dry compartments in the lower hull (like
engine room and pump rooms), and providing buoyancy space along the top centerline of
the main hull.
TABLE 2. STABILITY SUMMARY
Minimum metacentrLc heigth, which occurs at the maximum draft of 37.5 meter, is shown
to be 1.80 meter. At this draft, waterplane inertia contribution to transverse metacentric
height is almost negligible. Taking the reserve buoyancy of the superstructures
into consideration, the vessel is expected to have an excellent transverse stability
curve (GZ curve or 'rightening lever curve) with a maximum value greater than 4 meter
occurring at an angle of inclimation close to 90 degrees. Similar most surface vessel,
the significant portion of the stability curve could be cut short by downflooding.
However, for the SSIT, this downflooding angle can easily be made greater than 45
degrees.
659 Sandna!s 7

SALTWATER DISPLACEMENT. METRiC TONS
FIGURE 5. MISCELLANEOUS HYDROSTATIC CURVES
500000
If less metacentric height is acceptable, some of the permanent ballast may be eliminated
in favour of more payload. However, due to the limited availability of space in
the main hull, cargo capacity may be increased 7 percent at the most and at the expense
of reducing the transverse metacentric height by approximately 50 percent.
Minimum required metacentric height could be governed by four factors: damage stability,
wind heeling moment, accumulation of ice on superstructures, and forces and moments
exerted on the vessel by floating ice. The vessel is so big and heavy that wind moments
should not have noticeable effect on it. The same argument is probably true for ice
accumulations on the superstructures, especially if proper precautions have been taken
to enable the crew to remove such accumulations. Even the icebreaking process should
not cause too much loss of stability. Any large vertical force component acting downward
on the bow is counterbalanced by a vertical force at approximately the same elevation.
This counterbalancing force is due to added buoyancy due to sinkage. Based on these
considerations and the favourable shape of the GZ-curve indicated earlier, it seems
reasonable to make the preliminary assumption that minimum required metacentric height
will not exceed one meter. This assumption should make room for extensive necessary or
desirable modifications of proposed vessel.
660 SandniEs 8

PROPULSION, RESISTANCE AND ICEBREAKING CAPABILITY
This section deals with the ability of the vessel to propel itself through ice-covered
waters. Coefficients are proposed for propulsive efficiency, hydrodynamic resistance
and ice resistance. Later on, these coefficients are used in the evaluation of the
total economy of the project.
Machinery
Detail investigation into which type of machinery would be the most suitable has not
been carried out at this stage. Both steam turbines and gas turbines seem applicable.
Enclosed general arrangement drawing shows the vessel equipped with three fixed blade
propellers directly connected to three direct current electric motors. This way, the
complications of geared transmissions and variable pitch are aVOided, and shock loads
can be absorbed more easily by the electric motors. As suggested in Reference 4, these
motors could receive power from constant frequency AC generators through rectifiers.
Three main gas turbine generator sets are located inside the aft superstructure. This
arrangement should lend itself to quick and easy replacement of anyone set whenever
overhaul is necessary. Plenty space should be available for auxiliary machinery
adjacent to the generator sets. Also, the lower engine room may be made fairly small,
thereby improving stability and payload capacity.
Preliminary proposed total engine power is based on the Comparative Cost AnalYSis which
for the chosen ice conditions, shows 125 000 SHP to be close to optimum for SSIT
versus 175 000 SHP for the IBT.
Propellers and Propulsive Efficiency
Preliminary proposed propellers are three fixed pitch propeller with a diameter of 9.5
meter. Number of revolutions will probably be somewhere in the range from 80 to 110 per
minute. Nozzles are omitted in order to prevent blade damage due to pieces of ice being
caught between nozzle and blade tip.
Let
'?p
propulsive efficiency
EHP effective horse power
SHP shaft horse power
R total resistance in metric tons
v vessel speed in m/s
By definition EHP R v/0.075 (2)
1'l.p EHP/SHP (3)
The general form of the relationship between propulsive efficiency and speed is assumed
to be
where Po' P l , P 2 and P 3 are coefficients depending on propeller and hull efficiencies.
662 SandruEs 10
(4)

3. Using full engine power, the SSIT should obtain 4 knots after approximately 2
minutes over a distance of 130 meter. Total kinetic energy of the vessel would be
3 times greater than the kinetic energy of the largest existing icebreakers with
an assumed ramming speed of 12 knots.
4. Final developed static vertical breaking force on the ice could be in the order of
magnitude of F = 10 000 tons, which should be sufficient for breaking ice as thick
as h = 10 meter (30 ft). According to a formula on page 12 in Reference 2.
V 10000'
= 1;25·160 12 m
5. After heavy ice has forced an icebreaker to a standstill, normal procedure would
be to back out and try another blow. The large amount of engine power available
for the SSIT, should make this backing out a fairly quick and reliable operation.
6. An alternative to backing out would be to discharge approximately 30 000 tons of
clean water ballast from the forward tanks while the vessel still has its nose
under the ice. This way, a total of 40 000 tons static vertical, uplifting,
icebreaking force can be mobilized, way beyond the capability of any existing
icebreakers. Thus, provided the SSIT is fully operational, it is hard to imagine
that icebreaker assistance would be necessary to prevent the vessel from getting
stuck.
Figure 8 is based on the following approximate equations.
Where
v ....
T
L
"In .6 v,l.
( PT)
to + tn g ";. vp O+V1+i") + *
Vr
ramming penetration length
peak ramming velocity
deceleration while breaking ice
virtual mass ratio (1.1)
total displacement (420 000 tons)2
acceleration of gravity (9.81 m/s
ice resistance for instance as estimated from
equations (10) or (11)
propulsive thrust (900 tons)
ahead to astern thrust ratio (1.2)
mean velocity
time needed to reverse engines
time needed to back out and pick up momentum
for another blow
deceleration time
667 Sandrues 15
(19)
(20)
(21)

FIGURE 9. ICE PROFILES
complete ATC output, could have the effect of giving too favourable results mainly
in the lower engine power range.
The main reason for the option of being able to specify an upper speed limit is to
see if the overall economy can be improved by running at reduced power and fuel
consumption in open water. No significant savings have been indicated so far.
b. Environmental data
Voyage subdivided into sections, one input card for each section length in nautical
miles and average ice thickness in feet for each season or month throughout the
year.
Proposed input environmental data can be read either from Figure 9 or Table 6. Ice
condition variations have been taken into consideration by subdividing each year
into three seasons: summer, spring-fall, and winter.
The computer program can handle as many as 12 periods. However, proposed subdivisions
is considered sufficient for supporting the main conclusions of this paper.
c) Vessel data like deadweight, steel weight, ballast weight, installed shaft horsepower,
waterline breadth and wetted surface as listed in Table 1, and propulsive
efficiency as specified earlier.
output Data
Sample output of the ATC program for a voyage from Prudhoe Bay to Davis Strait with an
SSIT equipped with 125 000 SHP is shown in Table 6. To a large extent, this printout
should be self-explanatory. Most input data are included. Special attention should be
focused on the following principal results at the top of the table:
- unit price of transportation u =
- price of one vessel TI
- number of round trips per year
- average speed
- number of vessel needed for a
production rate of 2.000 000 bbl/day
- total price of these vessels STI
- investment rate
$
$
$ 2
$
.96 x) (.70) per bb1
126.5 million
25 (34)
12.5 (18) knots
16.5 (12.2)
100 (l 550) million
0.62(0.46) bbl/day/n.mile
X)Numbers in parenthesis are based on Kashteljan instead of Edwards and Lewis.
671 Sandna!s 19

ESTIMATING OCEANOGRAPHIC RESEARCH SHIP RESISTANCE IN ICE
Richard P. Voelker
Vice President
ARCTEC, Incorporated
Columbia, Maryland
United States
and
J. Kwangse Kim
Consulting Engineer
ARCTEC, Incorporated
Columbia, Maryland
United States
ABSTRACT
In reaent years, investigators from universities and other institutions around our nation
have expressed interest and aonaern on the rapid developments oaaurring in the Aratia.
This aonaern is aaaentuated by the faat that there are no iae-worthy Ameriaan researah
ships operating in Alaskan-Aratia waters. The purpose of this paper is to present the
results from reaent iaebreaking model experiments whiah have aharaaterized the behavior
of an oaeanographia researah ship hull form in iae aovered waters. Results of the iaebreaking
model tests are presented in equation form and the full saale iaebreaking resistanae
in graphiaal form.
677

INTRODUCTION
Elsner (1974) has stated that as American research activities in polar seas steadily
increase, the need to resolve the problems of obtaining truly adequate ice-going
ships to serve as laboratory platforms for that work are even more urgent. More
specifically expanding laboratory studies in such disciplines as biological, chemical,
geological and physical oceanography, marine mammal ecology and physiology, fish biology,
ornithology, ice physics and other fields require marine vehicles. Present
investigations in the north, specifically in the region of Alaska, merit special
attention, inasmuch as no ice-worthy American research ships are operating there on a
regular basis.
The results of research in our Arctic marine frontiers can be of highly significant,
practical and economic importance as well as of theoretical and long-range value in
the understanding of our environment. Indeed, wise and sensible development and use
of resources in that environment will depend upon new knowledge obtainable only from
an ice-worthy oceanographic research ship. A lasting solution for accomodation of
present and anticipated new research programs would require a new construction of
ice-worthy laboratory ships suitable for multi-disciplinary research activities.
Much of the scientific interest, as reported by Elsner, is not in regions of heavy ice
pack, and as such, an icebreaker is not required for many of the scientific endeavors
in the seasonal ice zones. Urgently needed, therefore, are modern, powerful icestrengthened
laboratory ships capable of maneuvering in the loosely-packed ice which
makes up a large part of the seasonal Bering and Chukchi Sea ice and much of the
summer margin in the Arctic Basin. Such a ship would also be capable of operating in
vast areas of the Antarctic Seas.
Dinsmore (1974) has developed a preliminary set of criteria for the operation of such a
ship based upon geographic, environmental, and scientific requirements. These are
summarized as follows:
1. Geographic
Gulf of Alaska
Aleutian Chain
Bering Sea and Bering Strait
Chukchi Sea
Beaufort Sea (Shelf)
2. Environmental
Ice: Winter ice of 0.91 meters thickness with 75% ice
concentration
Fast consolidated ice of 0.45 meters thickness
Pressure - withstand besetment
Weather:
Air Temp: (-40)OC @ 5 knots
Sea Temp: (-2)OC
Sea State: operate in State 5
Withstand severe icing
3. Scientific
Multi-discipline onboard research
Geophysical and fisheries - limited
to reconnaissance surveys
Cruise duration - 60 days
Scientists - 10 to 20
678
Voelker and Kim 2

Figure 1. Length of Operating Seasons for a 3000 SHP Oceanographic Research Ship
680
Voelker and Kim 4

TEST RESULTS
Linear regression analysis techniques were employed on the twenty-four model test data
points shown in Table II to generate a predictor equation for full scale ship resistance.
TABLE II
Model Test Data
Date Run R h v of ..l v R
(kg) (am) (am/seed (kg/am 2 ) PuP h Igh puP Bh2
8-21-74 lA 3.45 2.40 13.41 0.501 208.8 0.276 9.43
lB 4.01 2.35 48.46 II 213.2 1.009 11.44
lC 6.10 2.50 74.98 200.4 1. 514 15.37
8-22-74 2A 1. 29 1.60 13.41 0.233 145.6 0.331 7.93
2B 1. 55 1.48 48.46 II 157.4 1.271 11.96
2C 3.15 1. 65 72.54 141.2 1.803 18.22
8-23-74 3A 1.84 1. 64 18.29 0.271 165.2 0.453 10.75
3B 2.76 1.75 46.94 II 154.9 1.087 14.17
3C 4.06 1. 70 74.07 159.4 1.84 22.14
8-26-74 4A 8.13 4.15 18.29 0.317 76.4 0.287 7.44
4B 9.46 4.10 47.55 II 77.3 0.750 8.86
4C 10.92 3.80 74.98 83.4 1.228 11 .91
8-27-74 5A 3.51 2.15 18.29 0.267 124.2 0.398 11.96
5B 5.44 2.83 46.94 94.3 0.891 10.70
5C 5.74 2.58 74.98 103.5 1.490 13.58
8-28-74 6A 4.36 2.70 18.29 0.534 197.8 0.355 9.42
6B 4.78 2.50 46.94 213.6 0.936 12.04
6C 6.92 2.60 76.20 205.4 1.508 16.12
8-29-74 8A 8.01 3.25 18.09 0.400 123.1 0.324 11.94
8B 7.42 3.20 46.94 II 125.0 0.838 11.41
8C 12.97 3.80 76.20 105.3 1.173 14.14
8-30-74 9A 7.69 4.10 18.29 0.198 48.3 0.288 7.21
9B 9.28 4.30 46.94 II 46.0 0.723 7.91
9C 11. 10 4.00 76.20 49.5 1.216 10.93
This process, as documented by Lewis and Edwards (1970), requires the development of
several different groups of dimensionless variables. The variable groups developed by
Vance (1974) resulted in the best fit of the data and take the form
682 Voelker and Kim 6

.. c
o
i
• c
... (J
Z
...
I­
Ul
iii
...
II:
.__"-----;>'"""--- B. B 5 kg / em 2
_____ =-----4.21 kg/em2
o 2 4 6 B 10 12 14 16
SPEED (km/hrl
Figure 5. Icebreaking Oceanographic Research Ship Full-Scale
Resistance for Various Flexural Strengths
(.6 meters ice thickness)
A few notes are in order on observations made during the performance of the model
test program. During tests at ice thicknesses equivalent to 0.915 meters (in excess of
the requirements for this program) and high ice strengths, the model motions became
very pronounced. More specifically, the bow had to be pushed up into the ice sheet a
considerable distance prior to breaking a cusp. Upon breaking the cusp, the model
swung to that side. This in turn makes directional stability difficult in heavy ice.
This phenomenon is explained by the fact that the oceanographic research ship has
about one-third the displacement of the WIND Class icebreaker and hence must ride up
onto the ice sheet a greater distance prior to breaking a cusp. This type of action,
however, must be expected for high horsepower to displacement type ships.
685 Voelker and Kim 9

REFERENCES
1. Elsner, R. "Arctic Research Ships, A National Requirement," University of Alaska,
1974.
2. Dismore, R. P. "Needs for an Alaskan Arctic Research Vessel," Report of Meeting by
UNOLS, Seattle, Washington, August 14-15, 1974.
3. Lewis, J. W. and R. Y. Edwards, Jr. "Methods of Predicting Icebreaking and Ice
Resistance Characteristics of Icebreakers," SNAME Transactions, November, 1970.
4. Vance, G. P. "A Modeling System for Vessels in Ice," Ph.D. Dissertation in Ocean
Engineering, University of Rhode Island, 1974.
5. Kim, J. K., R. P. Voelker, G. H. Levine, and J. G. Toeneboehn "Icebreaking
Resistance Tests of a Great Lakes Tub," for U. S. Coast Guard, September 19, 1974.
ACKNOWLEDGEMENT
The model testing program described in this paper was sponsored by the United States
Coast Guard under Contract No. DOT-CG-43070 to the University of Michigan for the
resistance determination of an icebreaking tug having a hull form typical to that required
for an ice transiting oceanographic research ship and is shown as Reference 5.
686
Voelker and Kim 10