The petrofabrics of aufeis in a turbulent Alaskan stream

Item Type Thesis

Authors Kreitner, Jerry D.

Download date 25/09/2021 22:40:49

Link to Item http://hdl.handle.net/11122/7952 THE PETROFABRICS OF AUFEIS IN A TURBULENT ALASKAN STREAM

APPROVED:

Department Head

APPROVED _DATE_ Dean of the College of Earth Science and Mineral Industry

Vice President for Research and Advanced Study THE PETROFABRICS OF AUFEIS IN A TURBULENT ALASKAN STREAM

A

THESIS

Presented to the Faculty of the

University of Alaska in Partial Fulfillment

of the Requirements

for the Degree of

MASTER OF SCIENCE

By

Jerry D. Kreitner, B. S.

College, Alaska

May 1969

-UNIV. Of ALASKA LIBRARY ABSTRACT

The growth, form, and decay of in turbulent Goldstream Creek, Alaska, has been observed each year since 1964. Overflows which occur throughout die early part of the winter deposit layers of ice (aufeis) upon the pre-existing ice surface. Vertical and horizontal thin sections of the stream ice from the years 1965-66, 1966-67, and 1967-68 were examined and photographed under ordinary and polarized light. The c-axes of the crystals were oriented with a Rigsby four- axis stage and plotted on Schmidt equal-area nets. Examination of photographs and stereograms revealed five basic types of ice in Goldstream Creek: (1) clear, massive, original stream ice composed of elongate, tapered crystals in which the c-axes are primarily horizontal and randomly oriented; (2) bubbly overflow ice layers (aufeis) with horizontal c-axes which are sometimes aligned parallel to the stream flow; (3) skim ice layers with vertical to horizontal c-axis crystals; (4) fine-grained equigranular snow ice; and (5) underwater ice masses of slightly coherent, rounded plates with the c-axes normal to die plates and randomly oriented. During break-up the melt-water flows on top of the stream ice and slowly erodes the ice layers in the stream by a combination of melting and mechanical fragmentation. The layers are eroded away in descending order from the top to the bottom of the stream. TABLE OF CONTENTS

Page ABSTRACT ...... iii TABLE OF CONTENTS ...... iv LIST OF FIGURES ...... v LIST OF ILLUSTRATIONS...... vi INTRODUCTION...... 1 General Statement...... 1 Environment...... 3 Previous Work ...... 3 Present Work...... 5 Acknowledgments...... 5 PHYSICAL MEASUREMENTS OF THE 1967-68 ICE CYCLE ...... 6 Freeze-up Period ...... 6 Mid-winter Period ...... 14 Break-up Period ...... 14 STRATIGRAPHY AND PETROFABRICS OF THE ICE ...... 21 1967-68 Ice ...... 21 1966-67 Ice ...... 31 1965-66 Ice ...... 31 COMPARISON WITH OTHER TYPES OF IC E ...... 43 Lake I c e ...... 43 Sea I c e ...... 45 Ice ...... 46 CONCLUSION ...... 47 BIBLIOGRAPHY...... 49 APPENDIX A ...... 51 APPENDIX B ...... 57

iv LIST OF FIGURES

No. • Page

1. Plane table map of Goldstream Creek observation site ...... 4 2. Cross-section of Goldstream Creek showing various ice surfaces and thermocouple placements...... 10 3. Measurements of the electrical conductivity of the water; air and water temperatures; and water surface elevation during freeze-up in October 1967 ...... 11 4. Isotherms on 24 January 1968...... 12 5. Maximum and minimum air temperatures during the 1967-68 sea son...... 13 6 . Ice elevations during the 1967-68 season ...... 13 7. Plot of the electrical conductance for each of the ten different layers of the 1967-68 season ice ...... 15 8. Ice surface topography at Goldstream Creek on 20 April 1968...... 19 9. Cross-section of Goldstream Creek during break-up ...... 20

v LIST OF ILLUSTRATIONS

No. Page

1. Aerial view of Goldstream Creek...... 2 2. Surface ice forms at Goldstream Creek during freeze-up in October 1967 ...... 8 3. Underwater ice from Goldstream Creek during freeze-up in October 1967 ...... 9 4. Dr. Carl Benson photographing west end of trench which is near center of stream ...... 16 5. Ice layering in west end of trench...... 17 6. Horizontal thin sections from the 1966-67 icecore under polarized light...... 22 7. 1967-68 vertical thin section from bottom part of the stream section...... : . 25 8. 1967-68 horizontal thin sections from the bottom part of stream section under polarized light...... 26 9. 1967-68 vertical thin section from the middle part of stream section...... 27 10. 1967-68 horizontal thin sections under polarized light from the middle part of the stream section...... 28 11. 1967-68 vertical thin section from the top part of the stream section...... 29 12. 1967-68 horizontal thin sections from the top part of the stream section under polarized light...... 30 13. 1966-67 vertical thin section from the bottom part of the ...... 32 14. 1966-67 horizontal thin sections under polarized light from the bottom part of the ice core ...... 33 15. 1966-67 vertical thin section from the top part of the ice core ...... 34 16. 1966-67 horizontal thin sections under polarized light from the top of the ice core ...... 35 17. 1965-66 vertical thin section from the bottom of the stream section ...... 37 18. 1965-66 horizontal thin sections under polarized light from the bottom of the stream section ...... 38

vi No. Page

19. 1965-66 vertical thin section from the middle part of the stream section...... 39 20. 1965-66 horizontal thin sections under polarized light from the middle part of the stream section...... 40 21. 1965-66 vertical thin section from the top part of the stream section ...... 41 22. 1965-66 horizontal thin sections from the top part of the stream section under polarized light...... 42

vii INTRODUCTION

Genera] Statement

Water freezing in a turbulent stream behaves differently than water freezing in a relatively calm environment under die same temperature regime. The growth of ice in bodies of still water begins as a surface skim which grows downward. In a turbulent stream, ice often forms on the bottom and within the stream before ice develops on the surface. The ice that forms on the bottom of the stream is termed "anchor ice" and that which forms within die stream itself is called "frazil ice". These two terms have been collectively called "under­ water ice" by Altberg (1923, 1936), who pointed out diat diey are formed by the same mechanism. After the surface ice has formed a complete cover on the stream, pres­ sure bulges begin to form. Benson (1966, p. 103; 1967, p. 19) has shown that the freezing of water builds up pressure by constricting the channels. The pressure causes the bulges and, eventually, the overflows. Gold and Williams (1963) described a "hump" some 10 meters high, 100 meters wide, and 1200 meters long on the Ottawa , , that was the direct result of frazil ice deposition in a trench which resulted in a constriction of the flow and a corresponding rise in the water elevation. After the bulge or hump has been formed, expansion and contraction of the ice surface results in the formation of cracks. Following the formation of the cracks, the bulged ice surface gradually sags down in the center. As freez­ ing continues, the space available for running water is constricted. The pressure which develops is released when water escapes through cracks in the ice, or along the margins of the stream. This water flows out over the existing ice where it freezes to form overflow ice layers or "aufeis". Overflows occur throughout the early part of the winter and can build up an extensive thickness of aufeis that may be an order of magnitude greater than the original depth of the water in the stream (Benson, 1968, p. 19). The process continues until late January or early February. After the last overflow, the stream remains relatively stable, except for cracking, until the break-up begins in the late spring (F ig. 2). During break-up, when the ice choked channels cannot contain the runoff, Plate 1. Aerial view of Goldstream Creek. 3

the stream water flows on top of the ice surface. With increasing air temper­ atures and increasing water depth on the ice surface, die ice layers begin to erode away. The layers on top erode away first, followed by the lower layers. The last ice left in the stream is that frozen to rocks and branches along the banks and stream bottom.

Environment

The study area is located six kilometers north of the University of Alaska, where Goldstream Road crosses Goldstream Creek (Plate 1). Most observations and measurements were made between the bridge and the first upstream bend which is about 110 meters north of the bridge (Fig. 1). Goldstream Creek, which meanders through a broad, generally north-east trending , has a gradient of about five meters per kilometer. The stream bed consists of material rang­ ing in size from very fine-grained silt to pebbles, with the majority in the sand ra n g e. The study area was chosen because it was easily accessable throughout the winter and during the break-up period.

Previous Work

'The 1967-68 season was the fifth consecutive season that data were re­ corded at Goldstream. The first measurements were taken in 1963 by Dr. Carl Benson and graduate students from the University of Alaska. As the observers gained knowledge of the stream, they designed more elaborate observations and measurements. The measurements included air, water, and ice temper­ atures; water and ice elevations; and cross-sections of the stream. Summaries of the work accomplished each year were published in the Geophysical Institute Annual Reports for the years 1963 to 1968, and in the 1966, 1967, and 1968 Annual Reports of the Institute of Water Resources (IWR), University of Alaska. Reference-*-^ o -Snow tables

Thermocouple pit Undisturbed snow area ' Meteorological. Buildinq S helter y iV i X \ \ 1 . !---Bridge 7?—1 -Thermocouples

/^/-West bolt 2 A 4 6t 8 o o 10 3s? o i i C-,-Center bolt /-Bar i W ater Nv A i ______-1— i \ 12 o V \ ° HBast bolt . o 5 7 ° 0 3 7 9 1

Scale 1 : 2500 O i i x - — l 0 10 20 30 40 50 11 Meters

Figure 1. Plane table map of Goldstream Creek observation site on 23 September 1967. 5

Present Studies

During the 1967-68 season an extensive program for the recording of data was initiated at the Goldstream site. In addition to the study of aufeis char­ acteristics of turbulent streams, two students conducted research on the diagene­ sis of the snow cover and on the moisture content of the soil during a freeze-thaw cy cle . This thesis was derived primarily from the data and ice samples obtained from the 1967-68 season. However, ice cores and blocks from the two proceed­ ing years were also available for petrofabric analysis. Some data are being re­ corded during the 1968-69 season. Ice samples were collected at various stages of the freeze-up. Before break-up, a small trench will again be excavated to check the ice layering for any anomalous features.

Acknowledgments

The research for this thesis was accomplished under the direction of Dr. Carl Benson of the Geophysical Institute, University of Alaska. Dr. H.D. Pilkington served as acting committee chairman while Dr. Benson was on sab­ batical leave in 1968-69, and also provided technical information concerning petrofabric s. Financial support for the project was arranged by Dr. Benson in the form of a research fellowship through the Institute of Water Resources at the Univ­ ersity of Alaska. Others involved with the Goldstream project include: Dr. A. H. Thompson, visiting professor of Meteorology from Texas A and M University; Dr. T. D. Hamilton and Dr. D. B. Hawkins, both of the Geology Department at the University of Alaska; and Ed Wood and Dennis Trabant, graduate students from the University of Alaska. All have provided valuable assistance, both in the field and the laboratory. The use of a coldroom for studying the ice samples was made available by the Environmental Engineering Laboratory (A. E. E. L .) at the Univ­ ersity of Alaska. The efforts of the A. E. E. L. staff to keep the coldroom oper­ ational were greatly appreciated. \

PHYSICAL MEASUREMENTS OF THE 1967-68 ICE CYCLE

Freeze-up Period '

Prior to the freeze-up of Goldstream Creek, the stream banks were cleared of trees and shrubs to facilitate a plane table survey of the area. Iron pipes were installed on the stream banks at the end points of six different cross­ sections (Fig. 1). Hose clamps were installed on each pipe at a fixed elevation arbitrarily defined by the center reference bolt on the bridge. Profiles of the stream bed were surveyed at each cross-section; once before the freeze-up, each week during the winter, and once after break-up. The freeze-up of Goldstream Creek started on 30 September 1967, when the first ice of the season formed on puddles of water near the road. The first three thermocouple strings were installed in three, one meter deep pits near the building (Fig. 1) in order to determine the temperature gradient in the soil. A fourth string of thermocouples was placed on the stream bed along the cross­ section between stakes 5 and 6 (F ig. 1). Tvvo days after the thermocouples were installed, ice began forming along the banks and on a bar in the stream where the water was relatively shallow. This ice consisted mainly of c-axis vertical crystals and gradually grew out from the stream banks toward the center in the form of large discs and plates (Plate 2). Some dendritic blades (Plate 2), similar to those observed by Schaefer (1950), Prupacher (1967, Fig. 1), and Williams (1959, p. 56) were observed forming on the discs and plates in the middle of the stream. On 5 October the first "underwater ice" (frazil and anchor ice) was ob­ served on the bridge pilings, along the banks, and in the shallow part of the stream (Plate 3). Underwater ice development increased in the stream every day until freeze-up occurred on 21 October. The daily variation in temperature (Fig. 3) throughout the freeze-up period resulted in maximum formation of under­ water ice late at night and early in the morning. The growth of ice on the bottom increases the resistance to stream flow; this decreases the velocity. Since the discharge is essentially constant, the decrease in velocity demands an increased cross sectional area. This has been verified by the increase in water depth every

6 7

night. When the stream was freed of ice during the following day the stream height returned to its previous value (Fig. 3). Late in the morning the current of the stream began to peel the underwater ice off the bottom of the stream, carrying large amounts of debris with it. The underwater ice reached maximum development just before the stream completely froze over. During the freeze-up period water samples were taken. Electrical con­ ductivities of the samples were measured and plotted with the corresponding values of water elevation and temperature (Fig. 3). The conductance values in­ creased gradually until just before the stream had a solid ice cover on it when the values began to increase rapidly. Nine days after the first solid layer of ice had formed on the stream, a fifth string of thermocouples was installed on it. Two days later, large cracks formed in the ice on the west side of the stream parallel to the bank. As a result, the elevation of the ice surface at the center of the stream decreased (Fig. 2, surfaces 3 and 4). . Pressure bulges began to develop on the ice surface prior to the occur­ rence of the first major overflow. The water pressure, which caused the pres­ sure bulges (Benson, 1968) was released when the stream water broke through the cracks in the ice and at the stream margins. This water which overflowed froze to form a layer of aufeis on top of the first ice surface. This process was repeated throughout the freeze-up period and deposited several more layers of aufeis for a total thickness of about 100 cm upon the original ice surface. Three additional thermocouple strings were installed at various levels in the aufeis (Fig. 2). Temperatures in the ice and surrounding area were measured once a week until break-up. The isotherms (Fig. 4) indicate that water flowed along the stream bottom until the middle of January. A comparison of the maximum and minimum air temperatures (Fig. 5) with the plot of the ice height (Fig. 6) reveals that all the overflows, except one, occurred during relatively warm periods. A. Skim ice growing from sides of creek and also on bar near the right hand side.

B. Dendritic bladed ice. Bottom C. Initial ice skim of needles, side up. plates, a n ddendrites.

D. Platelet structure of skim ice. E. Initial ice skim in polarized light. Dark area contains c-axis vertical ice. Plate 2. Surface ice forms at Goldstream Creek during freeze-up in October 196 C. Close up showing dirt contained D. Close up showing platelet in the ice. structure of the ice.

E. Underwater ice frozen into the F. Underwater ice under bottom layer of stream ice. polarized light.

Plate 3. Underwater ice from Goldstream Creek during the freeze-up period in October 1967.

•j »*«** c y ? '*! 6 5 p

m /p*----- f 7r; 4 0 0 -I v A v / v y XXXX ///^ X ,,, r» r\ o -Q. n /f=\

------■ — f t - ® m m Water level on 3 0 0 - //A\ ' / ^ r / / / / 5 23 September 1967 i i ! ^ „ „ (J) 30 September 1967 % 2//%: m m . ( 2) 23 September 1967 Q 28 October 1967 S 20°- @ 1 November 1967 CD 18 N ovem ber 1967 C) 29 November 1967 1 0 0 - Q) 22 December 1967 0 ) 6 January 1968 C) 31 January 1968 1 > « . o l 200 400 600 800 1 0 0 0 1 2 0 0

C m . ------«—

Figure 2. Cross-section of Goldstream Creek showing various ice surfaces and thermocouple placement. Each small circle represents one thermocouple. 4 F 250- u '

in 200- O s~ E 150- Electrical conductivity 13 i i i . » 1 » t i 1-...... 6 7 8 9 10 11. 12 13 14 15 16 17 18 19 O c t o b e r -

+ A

o - o 3 V v f l O M A A j \ l ° -1 (5 Ambient air temperature

6 7 8 9 10 11 12 13 14 15 16 17 18 19 O c t o b e r

+.2°

U +.f- o sj\ f! \ f \ f ^ \ / ^./•\A 5 £. oiioo/wWater temperature * . 6 7 8 9 10 11 12 I • » 1 1 l 1 i 113 14 15 16 19 17 18

300- (§290- 77\ v4 firfW ^AAf 280- " Water elevation I ..... 1 - » 1 » » i 6 7 8 9 10 11 12 13 14 15 16 17 18 19 O c t o b e r

Figure 3. Measurements of the electrical conductivity of the water; air and water temperature; and water surface elevation during freeze-up at Coldstream Creek in October 1967. 5 ...... -...... - -JU 500-r n _ _ Snow surface— v 1 i } 1 i—- ! 4-24'-1 O x ■' 4l\\ \ \ \ \ W/T \VAA \ \ i H H ^* 5/ / / O' ' / y ' \ i? # > ■ n \ % /t V 4 5 0 - 1° 0~~ Oo l ( J o / . '■ ,'//// 1 i .. _1 6 — 4 m i s_ j . pi j a. A rrrrrr 1 o 1 1 it I J *///.X H—1 1 1 o o

' \ \ 4 t t o li O •- > ^ 4 / 2 2 ------o o oji o __ \i---- 400- \ w W ------' \ W 2: ^ \ y y v / / - V' 4 W AP, _. , ^ 'y jy / x i p

7 x y r\ U1 11 111 iy w utci ———...... | 350- 800 600 400 200 0 200 4 0 0 6 0 0 800 1000 1200 C e n t i m e t e r s

h-* Figure 4. Isotherms in Goldstream Creek on 24 January 1968. Temperatures are in degrees Centigrade. to Ambient air temperature at time of measurements was -24. 5 C at a height of one meter above snow surface. Elevation in centemeters uig h 16-8 season. 1967-68 the during Figure 5. Maxim um and m inim um air tem peratures at G oldstream C reek reek C oldstream G at peratures tem air um inim m and um Maxim 5. Figure Time 13 14

■ Mid-winter Period

" During the mid-winter period the ice surface elevation remained rela­ tively stable. The crack: that had formed after the last overflow continued to enlarge. Snow depth inci ;ased on the ice surface to a maximum of about 50 cm. Excavation of a tr nch (Plates 4 and 5) for studying the ice structures was started on 9 March, bout five meters south of the line between stakes 3 and 4 (Fig. 1). The trenc h was 1 m wide, 7 m long and 110 cm deep. Pebbles and debris were observed frozen into the ice as much as 30 cm above the stream bed (Plate 3E), and are pi obably the result of anchor ice being peeled off the bottom of the stream, ov« rturned, and then being incorporated into the over­ lying ice surface (Devik, 1949, p. 309). Examination of the debris laden ice under polarized light rev« aled that it was made up of very fine-grained to medium-grained equigrar ilar ice with some elongate tabular crystals (Plate 3F). Apparently, most of the t aded or platelet structure of the underwater ice was destroyed during its incoi poration into the overlying ice surface. Pieces of ice were cut out of the trench and thin sections prepared for petrofabric study. In add tion, small pieces from each layer were melted so that electrical conductivit es could be measured (Fig. 7). The layers that were relatively free of snow ic generally had lower conductivities. All conductivities were at least 100 umho le ;s than those obtained from the stream water during the freeze-up period (Fig 3).

Break-up Period

The first indicatic i of break-up was the appearance of yellow water and ice near the snow tables ( 7ig. 1). The amount of water gradually increased until 17 April when slight melt ng occurred at stake 6. On 18 April, water flowed from the cracks in the st: earn and increased in depth on the ice surface until 3 May, when the water le el began to recede. After the snow on the surface of the ice had melted, the ic ; surface topography was measured with an alidade and plane table. Profiles also were measured at each of the six cross-sections and a topographic map of he ice surface was prepared (Fig. 8). ; 5

L a y e r 10

J i M M t f i 9

■Mil j U i r o M 87

^ ■ y y - T T r - —

HlhMiHMillMMH 5 I

2

) 4 0 ' 8 0 120 1 9 6 8 I c e L im h o s c m

Figure 7. Plot of the electrical conductance for each of the ten different layers of the 1967-68 season ice. The clear, snow free layers tend to have lower values. 10

Plate 4. Dr. Carl Benson photographing west end of trench which is near the center of the stream. Plate 5. Ice layering in west end of trench at Goldstream Creek. 18

Erosion of the first surface layers of ice, by a combination of melting and mechanical fragmentation, began on 27 April. This process continued to remove succeeding layers of ice until 11 May, when the only ice remaining was that along the banks of the stream. Cross-sections of the ice surface measured at the line between stakes 5 and 6 and at the bridge during the break-up period are shown in Figure 9. a •

Bui Idimg- ^ I ' " " 1 /-Bridge

f .— 4 3 5 - ^ ---TTT" ‘ 8 ------425---o —

hf -- ---__1 415- '-'■’ X s ~ 4 4 ■--- 435 ... —V \ \ s l O --- o 5 7 ' ; T re n ch— *• 3

Scale 1 : 2500 1 1 I i o o 2 0 30 4 0 5 0 M e t e r s

F igure 8. Ice surface topography at Goldstream Creek on 20 April 1968. Contour interval is five centimeters. Note large crack in center of the stream and at extreme right hand side of drawing. 6 5 A A \ 20 ) April 1968

■—.... ■■'-'■■■... - • .°---0

---^--- -...... ' ^ , f / / / 0 //V^C) 0 <1

_n n r> o , ~fi—(^)—^ ///

6 w M A m m m 5 A ' A

1 M a y 1968

- n ni j! fp\, .X>//A O n n r -It o - J ^ V /v _Q_ o r» r> O J^X^W^Cvy/ I ■

1 % m .

3 M a y 1968

o'^bc.. o '!

%W/3o = trs =___C_ .. c /TN „ cAbb

6 ' / / / f t . 5

8 M a y 1968

Figure 9. Cross-sections of Coldstream Creek during break-up. Stipled area represents the water. See Figure 2 for identification of numbered layers. STRATIGRAPHY AND PETROFABRICS OF THE ICE

1967-68 Ice

In March 1968, ice cores from the preceeding year's work at Goldstream Creek were used for practicing the techniques of petrofabric analysis as described by Bader (1951), Haff (1938), Langway (1958), and Rigsby (1960). Vertical sec­ tions of die ice cores were cut and mounted in the A. E. E. L. cold room. After examining the vertical sections under polarized light, a number of horizontal sections were cut for petrofabric study. Examination of the horizontal sections revealed diat a few layers consisted of elongate, tabular crystals with the c-axis perpendicular to the long dimension of the crystal (Plate 6). Since orientation of the core relative to the stream direction was unknown, a large block of oriented ice was cut from the trench located five meters south of the line between stakes 3 and 4 (Fig. 1, Plates 4 and 5) so that orientation with respect to the stream direction could be determined. Two thin sections about 10 cm wide and totaling 110 cm in length were cut and mounted on large glass plates. The sections were then smoothed with a warm electric iron'*' until they were approximately 0. 4 mm thick. Ice of this thickness displays first order gray interference colors and gives sharp extinction positions. The 1967-68 ice (Plates 7 to 12) consists of ten definite layers with many sublayers. The zero elevation reference for the ice section is defined at the top of the ice formed when the original stream surface froze. Layer 1 (Plate 7) represents the original stream ice and consists of clear, dense ice with long tapering crystals. In the vertical section these crystals range in width from one mm near the top of the layer to two cm at the bottom.

In order to assure that the warm iron did not alter the crystalline struc­ ture of the ice, two pieces of ice were cut from the same section. One was mounted using the iron for smoothing and the other piece was mounted after init­ ially smoothing with a knife blade and finally with the palm of the hand. No differ­ ences could be detected in the two pieces.

21 Plate 6. Horizontal thin sections from the 1966-67 ice core under polarized light. The c-axis is perpendicular to the long dimension of the crystal. 23

Examination of the horizontal sections (Plate 8) reveals subhedral, equigranular crystals. The crystals range in size from less than one mm diameter at the top of the layer to as much as two cm in diameter near the bottom. The stereo­ grams for Layer 1 (Appendix A, -6 cm and -11 cm) show that, in general, the crystals lie with the c-axis nearly horizontal and randomly oriented. Near the top of the layer, some crystals are almost vertical but within a few centimeters depth, these crystals become wedged out by the horizontal crystals. Scattered throughout Layer 1 are occasional patches of very fine-grained bubbly ice containing rocks and plant material (Plate 3, E and F). Such ice is thought to have resulted from accumulation of underwater ice upon the overlying ice cover on the stream (Benson, 1968, personal communication). Layer 2 consists t>f fine-grained equigranular, stratified snow ice, with one minor overflow layer at +10 cm. The aufeis layer consists of medium-grained, tapered crystals. Many small air bubbles, less than one mm in diameter, are dispersed throughout the section, with a layer of larger bubbles located two cm below the top of the layer. Three distinct layers of snow were saturated with overflow water and then solidified to form the snow ice layers. The boundary between Layers 1 and 2 is marked by a thin zone of black stream sediment which is concentrated along the grain boundaries in a polygonal pattern. The black stream sediment disappears within a few cm below the top of Layer 1. The sediment indicates that the first surface water that froze con­ tained more debris than the water which froze later in the season. The very bottom of Layer 2 contains some elongated air bubbles less than one cm in length. The boundary between Layers 2 and 3 is marked by a layer of very fine, reddish-brown, decayed plant material (Plate 7, ordinary light). Minute air bubbles up to 0. 5 cm in length and very thin elongate, tubular air bubbles are also present. The base of Layer 3 is aufeis and consists of clear, dense ice in which the crystals are long, thin, and tapered. The top portion (Layer 3-B) is made up of very fine-grained, equigranular snow ice with some large bubbles. The boundary between 3-A and 3-B is quite irregular and indicates that snow fell upon the overflow before the surface had frozen. Layer 4 is identical to Layer 3-B except for a lack of the air bubbles. Layer 5 is aufeis consisting of medium-grained tapering crystals with many elongate tubular air bubbles. The horizontal section for this layer (Plate 10, +29 cm) shows an interlocking arrangement of medium-grained subhedral 24

crystals. The stereogram for the Layer (Appendix A, +29 cm) reveals that some of the crystals are nearly vertical and some are nearly horizontal, with no pre­ ferred orientation with respect to stream flow. L ayer 6 is made up of fine-grained equigranular snow ice with zones of large bubbles, up to 1/2 cm in diameter, at the top and bottom of the layer. Layer 7 consists of interbedded, bubbly aufeis and bubbly snow ice. Bubbles are more numerous in this layer than in any other layer in the section, possibly because it froze at a faster rate, since this overflow occurred at tem­ peratures of -30° C. and below. L ayer 8 is subdivided into two secondary' units. The basal unit ( 8-A ) consists of clear, bubble-free aufeis that contains some large pieces of yellow- brown, decayed, plant material and some green algae (Dr. V. L. Harms, 1968, personal communication). In vertical section the crystals of this unit are coarse­ grained, long, and taper from very narrow near the top to as much as two cm wide near the bottom. Examination of the horizontal sections (+63 cm, + 66 cm , +68 cm, Plate 12) shows medium to coarse-grained subhedral crystals. Some very fine-grained crystals containing organic material are also present. The stereograms of two of these sections (Appendix A, +66 cm , +68 cm) clearly show a maximum in the northeast quadrant. Unit 8-B is another aufeis layer of thin, long, tapered crystals that contains very finely disseminated organic material. This thin tapered ice gradually grades into very fine-grained snow ice at the top of the layer. This layer shows a broad distribution of points on the stereogram with most of the c-axes being nearly horizontal. Layer 9 consists of clear, coarse-grained aufeis. In the vertical section this ice has a type of "blocky" structure under polarized light. The horizontal sections from the Layer (Plate 12, +83 cm and +89 cm) contain elongate, tabular crystals with the c-axis perpendicular to the long dimension of the crystal. The stereograms for these sections (Appendix A, +83 cm and +89 cm) show a definite parallel alignment of the c-axis with the stream flow. Layer 10, the final layer of the season, was made up of fine-grained, bubbly, stratified snow ice. Layer 3B Fine-grained snow ice +20' / * *t .■ ; f 1

•If, \ Layer 3A 1 4 II , fi | , » | Tabular overflow ice

Thin tubular air bubbles

■ 1 • 1,1 ' Layer 2 Fine-grained bubbly snow ice +10-^ interbedded with clear, bubbly

overflow Ice i i M

m

+ ^ l m m m .

Reference level

Layer 1 Original stream ice

ery fine-grained black sediment i T i t * 1 U 1 r*HO 1 1 LB 4 ■ 1

Ordinary light Polarized light

Plate 7. 1967-68 vertical thin section from the bottom part of the stream section. Bottom of stream is at the bottom of page. Each square on the grid equals one square centimeter. 26 -i-19 cm •1 cm Medium-grained, Medium-grained, equigranular over­ equigranular ice. flow ice.

+11 cm_ Fine to medium- Coarse-grained, grained equigranular subhedral crystals snow ice.

+9 cm ■11 cm Medium - grained, Coarse-grained, equigranular over­ subhedral crystals flow ice.

Plate 8. 1967-68 horizontal thin sections from the bottom part of stream section under polarized light. Each square on grid equals one square centi­ meter. 460 Layer 7 Fine-grained, bubbly snow ice interbedded with fine-grained bubbly snow ice. 455

+50

Bubbly layer

445

.*/«■* \ V , 'CJ Layer 6 Jm V* *+■ * • * '* '*V t Fine-grained snow ice +40 % 8 m m

r +•'.'.1 ,-.w ,v ift. +35 Bubble layers ■-•-• i‘‘.'r v r .V j, 1 -4 •» . • • {?*•*££. 'tl*,'•« •'•* , y '- i •£js*w *1 • * Layer 5 Coarse-grained tabular overflow+30■ +30 ice ~ Tubular air bubble Layer 4 Very fine-grained snow ice +25

Layer 3B

Ordinary Light Polarized Light

Plate 9. 1967-68 vertical thin section from the middle part of stream section. Each square on the grid equals one square centimeter. 28 +56 cm +29 cm

Fine-grained snow Interlocking ice. medium-grained, elongate crystals and medium to fine­ grained subhedral crystals.

Fine to medium- Fine-grained, grained overflow squigranular snow ice. ice.

+34 cm Fine-grained snow ice with equigranular crystals.

Plate 10. 1967-68 horizontal thin sections under polarized light from the middle part of the stream section. Each square on the grid equals one square centi­ meter. Layer 10 -A Fine-grained, bubbly, stratified:d -• - '-.i »■«+•.' ' ‘ •' snow ice. 4-100 ;*■

+95

+90

Layer 9 Clear overflow ice +85

S z j a "Blocky" crystals MHI +80

‘ Vl

+75 Layer 8B a J H K ' i i'-1 Overflow ice I , ■ — Very fine-grained organic » . ' V ' k i material +70

Layer 8A

_ Decomposed plant remains +65 H

- Small (cm) tubular air bubbles

Ordinary light Polarized light

Plate 11. 1967-68 vertical thin section from the top part of the stream section. Top of stream is at the top of the page. Each square on the grid equals one square centimeter. .30 +78 cm Fine-grained snow Very fine-grained ice. ice containing very fine decomposed plant material.

+83 cm +66 cm Coarse-grained Coarse-grained tabular, elongate subhedral crystals crystals. of overflow ice.

i-79 cm +63 cm Coarse-grained Coarse-grained subhedral crystals subhedral crystals with fine-grained of overflow ice. equigranular ice.

Plate 12. 1967-68 horizontal thin sections from the top part of the stream section under polarized light. Each square on the grid equals one square centimeter. 31

1966-67 Ice

Several ice cores were obtained from Goldstream Creek in the spring of 1967 and stored in a freezer until a petrofabric analysis could be performed on them. Vertical and horizontal thin sections, made from one of the 1967 cores, were examined under polarized light. These sections (Plate 6) revealed that c-axis orientation with respect to the stream flow should be investigated further and this was done on the 1967-68 ice. In addition to the preferred direction of the c-axis in some of the sec­ tions, another important difference in the 1966-67 ice was the appearance of a clear layer composed almost entirely of c-axis vertical crystals (Layer 3, Plates 13 and 14). An ice layer similar to this type was observed forming at Goldstream Creek in October 1968. This was the first ice skim of the season and grew out from the banks in a shallow quiet area. The vertically oriented ice was not noted in the 1967-68 ice.

1965-66 Ice

During the spring of 1966, ice blocks were obtained from a trench located just south of the line between stakes 1 and 2 (Fig. 1). The pieces were marked with a felt pen to indicate orientation with respect to the stream, then were stored in a freezer. Most of the ice was not wrapped and as the ice evaporated, the markings gradually disappeared. Proper stream orientation is therefore unknown for these samples. The ice from this season closely resembles the 1967-68 ice. The bottom of the section consists of clear, original stream ice which is overlain by a succession of layers of snow ice and aufeis. Two layers in this section have crystals oriented with the c-axes primarily vertical. Layer 2 (Plates 17 and 18) consists of a section of bubbly ice containing large vertical crystals. The hori­ zontal section labled +0. 5 cm (Plate 18) consists mostly of crystals which are either vertical or nearly vertical. Several grains which are more nearly hori­ zontal are also present. The horizontal section (+1. 5 cm) which lies one cm above the vertically oriented ice is made up of an interlocking arrangement of +15 Layer 4 ♦ v . Fine-grained equigranular snow ice

+10 Layer 3 Overflow ice

Vertical crystals - +5

Layer 2 Fine-grained equigranular snow ice Reference level

Layer 1A Clear overflow ice i mil■ a UW M I All

- 5

-1 0 « n 1

-1 5

Layer IB -20 Fine-grained underwater ice t l M i

^ i ±l . „ r M

Ordinary light Polarized light

Plate 13. 1966-67 vertical thin section from the bottom part of the ice core. Each square on the grid equals one square centimeter. 33

+3 cm Very coarse­ grained, bladed vertical crystals.

Plate 14. 1966-67 horizontal thin sections under polarized light from the bottom part of the ice core. Each square on the grid equals one square centimeter. Layer 6 +55 H R Very fine-grained, bubbly, stratifiec snow ice

Layer 5B +50 M m Bubbly overflow ice

-Bubbles greater than 1/4 cm diameter

445

440 -9 H H

- Bubbles less than 1 mm diameter

_±35 ^ ^ B

Layer 5A Bubbly, stratified overflow and snow iceice

4-30

Layer 4 Fine-grained equigranular snow ice 4-25 J H H K |

+2° t M n B l O f X *

Ordinary light Polarized light

Plate 15. 1966-67 vertical thin section from the top part of the ice core. Top of stream is at the top of the page. Each square on the grid equals one square centimeter. 35

Fine-grained equigranular to elongate crystals of snow ice, and overflow ice.

4-*5 1 r* m Fine to medium-grained equigranular overflow ice.

Plate 16. 1966-67 horizontal thin sections under polarized light from the top part of the ice core. Each square on the grid equals one square centimeter. 36

elongate, tabular, horizontal crystals with several vertical crystals. Layer 5 (Plates 19 and 20) is made up of clear, dense aufeis and con­ tains many crystals that are vertically oriented. The horizontal section at +27 cm (Plate 20) contains coarse-grained, anhedral, elongate, tabular crystals, with orientations ranging from horizontal to vertical. The irregular grain boundaries of the crystal in this section are similar to the grain boundaries described by Knight (1962, p. 323) and could be the result of dendritic growth. Horizontal section +28 cm (Plate 20) is dominated by a single vertically oriented 2 crystal with a surface area of more than 35 cm . Layer 3 Bubbly overflow ice 1 +15 - P f l f t T

+10 -

Layer 2 bubbly snow ice ^ | | | | +5 _

i MBB 1 E u m '-fiv Reference Level ■ u -i E -$'v Layer 1 u ■Sr Original stream ice l b Very fine-grained black sediment 1I \W\ l i | M j l ! *11 - 5 - -Plant remains

1 ‘ 1 K . I* *. Long tapered crystalss ,,

-10 .

Clear ice W m -15 - ■ I i m

— 20 s

Ordinary light Polarized light

Plate 17. 1965-66 vertical thin section from the bottom part of the stream section. Bottom of stream is at bottom of page. Each square on grid equal-, one square centimeter. 38

lO cm Fine to medium - Coarse to very grained overflow coarse grained ice with interloci: subhedral vertical ing structure. crystals.

1-11 cm — 5 cm Fine-grained, Medium-grained, equigranular equigranular ice. overflow ice.

3-5 cm — 14 cm Fine-grained, Coarse-grained equigranular subhedral crystals snow ice. of original stream ice.

-ri. 5 cm — 19 cm Interlocking, Coarse-grained elongate, tabular subhedral crystals horizontal crystals of original stream with coarse-grained ice. vertical crystals.

Plate 18. 1965-66 horizontal thin sections under polarized light from the bottom part of the stream section. Each square on the grid equals one square centimeter. Layer 8 Clear, bubbly overflow ice 1 mm diameter bubbles •K50 . / /

Long, tapered crystals

+45 _ I _1 cm diameter bubbles

Layer 7 Bubbly overflow ice + 4 0 ­ , > 7 , \ * : , ' . /, ■ * 1 S WY , \ 4 r 4 - .

+35­ . .'-Y f : Ci Layer 6 Bubbly snow ice

— Brown organic material +30­

1m u - . Layer 5 • Clear overflow ice m w m

Vertical crystals —

Layer 4 ^4 * i *, * — * M * » Stratified bubbly snow ice * V c ; v • .v r j - k. i - . J. Brown organic material

Ordinary light Polarized light

Plate 19. 1965-66 vertical thin section from the middle pari of the stream section. Lacli square on the grid equals one square centimeter. 40 4-44 cm '-27 cm Medium-grained Coarse-grained overflow ice. anhedral elongate, tabular crystals with horizontal to vertical c-axes.

+35 cm +26 cm Fine to medium- Fine-grained grained equi­ equigranular granular snow overflow ice. ice.

+28 cm Very coarse­ Interlocking grained vertical crystals of fine crystal. to medium- grained over­ flow ice.

Plate 20. 1965-66 horizontal thin sections under polarized light from the middle part of the stream section. Each square on the grid equals one square centi­ meter. +85 Layer 9 Fine to very fine-grained, bubbly, equigranular, strat­ ified snow ice +8° B |

Bubbles up to one cm in diameter

+75 B |

H i ■fiM KSgjl

+65

+60 ■ H H g

Layer 8 Clear, massive, overflow ice containing few bubbles +55

Ordinary light Polarized light

Plate 21. 1965-66 vertical thin section from the top part of the stream section. Top of the stream is at the top of the page. Each square on the grid equals one square centimeter. 4

Plate 22. 1965-66 horizontal thin sections from the top part of the stream section under polarized light. Each square on the grid equals one square centimeter. COMPARISON WITH OTHER TYPES OF ICE

Lake Ice

In many respects the growth and form of ice in lakes or melt ponds is quite similar to that of turbulent streams. In lakes, the formation of the ice cover begins with the initiation of a thin ice skim made up of needles, feathers, and discs or stellar forms which have the c-axis normal to the plane of the growth and normal to the water surface (Arakawa, 1955). Maguruma and Kikuchi (1964, Fig. 1) state that if the supercooled layer is thick enough, barb-like vanes with c-axes normal to the vanes begin to form and extend down into the water, forming an interlocking arrangement of dendritic blades. Similar dendritic blades were observed at Goldstream Creek (Plate 2-B) during the 1967-68 season. The first growth of ice in lakes may be random, or consistently oriented with either horizontal or vertical c-axes. Lyons and Stoiber (1962, p. 368) found that if the winds over the lake surface have a velocity greater than 2. 7 m /sec, mechanical fragmentation of the ice surface will take place. When this happens, the vertical barb-like vanes or dendrites break off and float to the surface with the c-axis of the dendrites being normal to the water surface. There­ fore, depending on the wind and the amount of supercooling, any c-axis orientation from vertical to horizontal may prevail in lake ice. Except for the virtual absence of wind during the winter in Goldstream Valley, and the existence of turbulent water and underwater ice in the stream, the formation of the first layer of ice at Goldstream is almost identical to that formed in lakes. Lake ice crystals in the surface layer which have their c-axes horizontal are always fine-grained (1 mm) at the top and randomly oriented in the hori­ zontal plane. The c-axis vertical crystals are generally coarse-grained (cm) (Lyons and Stoiber, 1962). After the initial ice skim has formed, the ice continues to grow downward. Knight (1962, p. 322) states, "Below the surface all the ice was c-axis horizontal. Evidently phenomena occurring during growth of ice at the bottom of an can be more important to the orientation of the bulk of the ice than the initial ice

43 44

formation. " This change in orientation from the random mixture of horizontal and vertical crystals of the surface layer to the c-axis horizontal orientation of the deeper ice has been attributed to a wedging out of the vertical grains by the horizontal grains and to an alignment of the basal plane, i. e. , the plane of the most rapid growth, with the vector of the heat flux (Knight, 1962, p. 324 and 2 Shumskii, 1964, pp. 160-67). Lake ice crystals grow to large sizes (cm ) with depth and the c-axes tend to become more nearly horizontal. The orientation of the c-axis in the horizontal plane is random. If the pictures of the vertical and horizontal ice sections for the original ice and the overflow ice of Goldstream Creek (Plates 7 to 22) are compared to those of Knight (1962, p. 323); Magruma and Kikuchi (1963, pp. 694-95); and Ragle, Blair, and Persson (1960, pp. 44, 51), it is found that there is little difference at all between the two types of ice. When dendritic growth occurs in melt ponds, i.e., lakes formed from the melting and freezing of ice, the grain boundaries become irregular in a rather periodic fashion, doubtless reflecting the platelet structure. Every grain having a tilted c-axis develops one side as a straight face (Knight, 1962, pp. 323, 329). This face is the one parallel to the trace of the basal plane toward which the horizontal sections of Layer 9 of the 1967-68 ice (Plate 12, sections +83 cm and +89 cm). The dendritic growth irregular pattern mentioned by Knight (1962, p. 323) also appears in the Goldstream ice (Plate 20, section +27 cm). Crystals in the melt ponds on the Ward Hunt (Ragle and others, 1960, p. 45) were found to consist of medium to very coarse-grained, long, columnar crystals with straight grain boundaries. Also noted here were many long, linear, well-oriented tubular air bubbles which terminated at a boundary defined by one to two centimeters of crystal clear ice. The clear ice is thought to have crystalized last, i. e. , the layer frozen from the bottom up. Many bubbles similar to these were found in the aufeis from Goldstream Creek. In comparing lake ice to the ice formed in turbulent streams, a close similarity is noted from the initiation of the first ice crystals to the final form that the ice assumes. 45

Sea Ice

Although sea ice and stream ice are both classified as congelation ice by Shumskii (1964, pp. 83-84), their structure and fabric are for the most part quite different. In the horizontal section, sea ice typically displays a "platelet structure", as shown by Weeks and Assur (1967, Fig. 10), Bennington (1963, Figs. 11, 12, 15, and 16), and Knight (1967, Plates 7 and 8). This typical plate­ let structure is the result of ice that freezes from a solution rather than from a

liquid (Dorsey, 1940, pp. 423-24). I ' Weeks and Assur (1967, pp. 6-7) state that the first crystals to form in sea ice are in the form of minute spheres of pure ice. Growth changes these spheres into thin circular disks, which then change to hexagonal stars. These stars then overlap and form a continuous thin ice skim with a slight concentration of crystals with their c-axis vertical and a few in almost every other direction. Comparison of Figure 2 of Weeks and Assur (1967) with Plate 2-C of the Gold- stream Creek ice skim shows a close resemblance of the two ice covers. As in the formation of lake ice and stream ice, the amount of turbulence during the freezing process and the amount of supercooling greatly affect the crystal form of the initial cover. During conditions of significant turbulence more crystals form per unit volume of sea water, therefore increasing the abra­ sive action between crystals. This results in a slush that congeals to give fine­ grained, equigranular ice with a random c-axis orientation, similar to the snow ice layers of the Goldstream Creek ice. After the initial ice skim has formed on the surface, the growth rate of the crystals is determined by the heat flux in the sheet. A short distance below the upper ice surface, sea ice is mainly composed of c-axis horizontal crystals. Apparently the same mechanism of geometric selection of horizontal c-axis crystals, previously discussed by Knight (1962, pp. 322-28), and Shumskii (1964, pp. 160-67) can be applied to sea ice as well as lake ice and stream ice. Perey and Pounder (1958, p. 500) also found this mechanism of selection of horizontal grains when conducting experiments on fresh water and various solutions in tanks. Sea ice differs from lake ice and river ice in that it has the platelet structure characteristic of freezing in solutions. 46

Glacier Ice

According to Shumskii (1964) there are basically three types of ice: (1) congelation ice, (2) metamorphic ice, and (3) sedimentary ice. Congela­ tion ice, the equivalent of igneous ice, is the direct result of freezing water, i.e ice formed in lakes, ponds, streams, and in the sea. Sedimentary ice in­ cludes ice forms such as hail, sleet, hoarfrost, and rime. Metamorphic ice is ice that has undergone a change or transformation of form after its original deposition. Glacier ice is an example of a truly metamorphic form of ice. In comparing one form of ice with another, the basic type and mode of growth should be considered. As previously pointed out, glacier ice is con­ sidered to be metamorphic in nature and should therefore have a characteristic fabric that is different from stream ice. A comparison of stream ice to glacier ice reveals that not only do the two types of ice have different modes of growth, but that they each possess distinguishing fabrics and structures. However, the fine-grained snow ice layers of Goldstream Creek are similar in appearance to fine-grained glacier ice. CONCLUSION

' The five basic forms of ice produced by turbulent streams are: (1) skim ice, (2) original bottom stream ice, (3) overflow ice (aufeis), (4) snow ice, and (5) underwater forms of ice. There are two types of the skim ice: (1) almost entirely c-axis vertical crystals in the form of large plates and (2) dendritic blades and plates with c-axes normal to the plates and randomly oriented. The turbulent nature of the stream and the existing meteorological conditions at the time of freezing have a large influence on whether the original skim ice will be preserved. This skim ice is similar to that which forms on puddles, lakes, and ponds. The original bottom stream ice is generally clear and massive and con­ tains elongate, tapered crystals. In the upper part of the layer, which usually contains a large amount of silt, the crystals are relatively small and randomly oriented from horizontal to vertical. The silt is apparently brought to this posi­ tion by anchor ice floating up from the stream bottom. Some of it becomes incorporated in the ice forming at the surface. Within a few centimeters depth from the top of the layer, the horizontal crystals wedge out the vertical crystals. Scattered throughout this bottom layer of ice are occasional large (30 cm long) tubular air bubbles and patches of fine-grained ice containing large amounts of rock and debris. Except for the underwater forms of ice which become preserved in the bottom layer, the original stream ice is quite similar to lake ice. The aufeis layers are generally clear, sometimes bubbly layers with long tapered crystals. In certain cases a definite parallel alignment of the c-axis with stream flow occurs. Dendritic blades and plates were observed forming at various inclinations on the underside of the skim ice at Goldstream Creek and extending dow n into the stream. The snow ice layers consist of bubbly, sometimes stratified ice, in which the grains are very fine to fine-grained, euhedral, and equigranular in both the horizontal and vertical sections. This ice clcsely resembles fine-grained glacier ice. In their initial state, the underwater forms of ice appear as spongy, slightly coherent masses of plates and dendrites with the c-axes normal to the

47 48

plates and randomly oriented. Much of this ice becomes anchored to the bottom of the stream. As soon as it becomes bouyant enough, lumps of it are peeled off the bottom of the stream by the current, carrying large amounts of rock and debris with it. These ice masses overturn and bounce along the bottom of the overlying ice layer, often depositing large amounts of debris upon it. Apparently, a large portion of the plates are destroyed during this process because the under­ water ice found frozen into the bottom layer of the stream generally consisted of fine-grained equigranular crystals with only a small amount of bladed crystals. BIBLIOGRAPHY

Altberg, W. J. , 1923, On the cause of the formation of ice at the bottom of and lakes: Quar. Jour. Roy. Met. Soc. , v. 49, no. 205, p. 54-60. , 1936, Twenty years of work in the domain of underwater ice forma­ tion (1915-35): Ass. Int. D'Hydrologie Sci. Bull. , no. 23, p. 337-407. Arakawa, Kiyoshi, 1955, The growth of ice crystals in water: Jour, of , v. 2, no. 17, p. 463-467. Bader, Henri, 1951, Introduction to ice petrofabrics: Jour, of Geology, v. 59, no. 6, p. 519-536. Bennington, Kenneth O. , 1936, Some crystal growth features of sea ice: Jour, of Glaciology, v. 4, no. 36, p. 669-688. Benson, Carl, 1964, Freezing in turbulent streams: Annual Report, 1963-64, Geophysical'Inst. , U. of Alaska, p. 55. , 1965, Freezing in turbulent streams: Annual Report, 1964-65, Geophysical Inst. , U. of Alaska, p. 59-60. * , 1966, Freezing in turbulent streams: Annual Report, 1965-66, Geophysical Inst. , U. of Alaska, p. 59-60. , 1967, Freezing in turbulent streams: Annual Report, 1966-67, Geophysical Inst. , U. of Alaska, p. 70-71. , 1968, Freezing in turbulent streams: Annual Report, 1967-68, Geophysical Inst. , U. of Alaska, p. 62-63. , 1966, A study of the freezing cycle in an Alaskan stream: 1966 Annual Report of the Inst, of Water Res. , U. of Alaska, p. 100-117. , 1967, A study of the freezing cycle in an Alaskan stream: 1967 Annual Report of the Inst, of Water Res. , U. of Alaska, p. 19-22A. , 1968, A study of the freezing cycle in an Alaskan stream: 1968 Annual Report of the Inst, of Water Res. , U. of Alaska, p. 13-21. Dorsey, N. E. , 1940, Properties of Ordinary Water Substances: Am. Chem. Soc. Monograph Series, no. 81, 673 p. Devik, Olav, 1949, Freezing water and supercooling: Jour, of Glaciology, v. 1, no. 6, p. 307-309. Gold, L. W. and Williams, G.P. , 1963, An unusual ice formation on the Ottawa River: Jour, of Glaciology, v. 4, no. 35, p. 569-573. Haff, John C. , 1938, Preparation of petrofabric diagrams: The Am. Mineralogist, v. 23, no. 9, p. 543-574. Knight, Charles A. , 1962, Studies of Arctic lake ice: Jour, of Glaciology, v. 4, no. 33, p. 319-335. , 1967, The Freezing of Supercooled Liquids: D. Van Nostrand Company, Inc. , Princeton, New Jersey, 145 p.

49 V 50

Langway, Chester C. Jr., 1958, Ice fabrics and the universal stage: U.S. A rm y SIPRE Technical Report No. 62, August, 1958. Lyons, J. B. and Stoiber, R. E. , 1962, Orientation fabrics in lake ice: Jour, of Glaciology, v. 4, no. 33, p. 367-370. Muguruma, Jiro and Kikuchi, Katsuhiro, 1963, Lake ice investigation at Peters Lake, Alaska: Jour, of Glaciology, v. 4, no. 36, p. 689-708. , 1964, The origin of vertical c-axis ice on Peters Lake, Alaska: Jour, of Glaciology, v. 5, no. 39, p. 372-375. Perey, F.G.J. and Pounder, E.R. , 1958, Crystal orientation in ice sheets: Canadian Jour, of Physics, v. 36, no. 4, p. 494-502. Prupacher, H. R. , 1967, Growth modes of ice crystals in supercooled water and aqueous solutions: Jour, of Glaciology, v. 6, no. 47, p. 651-662. Ragle, R. H. , Blair, L.E., and Persson, L. E. , 1964, Ice core studies of Ward Hunt Ice Shelf: Jour, of Glaciology, v. 5, no. 37, p. 39-59. Rigsby, George P. , 1960, Crystal orientation in glacier and in experimentally deformed ice: Jour, of Glaciology, v. 3, no. 27, p. 589-606. Schaefer, Vincent J. , 1950, The formation of frazil and anchor ice in cold water: Trans. Am. Geophysical Union, v. 31, no. 6 , p. 885-893. Shumskii, P. A. , 1964, Principles of Structural Glaciology: Dover Publications, Inc. , New York, 497 p. Weeks, W. and Assur, A. , 1967, The mechanical properties of sea ice: U. S. Army CRREL Monograph No. 11-C3, September, 1967, 80 p. Williams, G. P. , 1959, Frazil ice: The Engineering Jour. , v. 42, no. 11, p. 55-60. APPENDIX A

' 1967-68 STEREOGRAMS

The following stereograms were derived from the c-axes of crystals plotted on Schmidt equal-area nets. Contour lines were drawn through points correspond­ ing to densities of 1, 2, 3, 4, 5, 7, and 10 percent in one percent of the area. Each stereogram represents 25 to 35 c-axes.

Percent of c-axes in a one percent area of the stereogram

51 \ -6 cm 52

N +29 cm 53

+63 cm 54 55 56

1967-68 Ice

+89 cm APPENDIX B

1965-66

STEREOGRAMS

57 -1 9 cm 58

+18. 5 cm 59

1965-66 Ice

+55 cm