The URL can be used to link to this page

Home My WebLink AboutSUS466SUSI TNA RIVER ICE PROCESSES: NATURAL CONDITIONS AND PROJ ECTED EFFECTS OF HYDROELECTRIC DEVELOPMENT VOLUME I 4 SUSITNA RIVER ICE PROCESSES: NATURAL CONDITIONS AND PROJECTED EFFECTS OF HYDROELECTRIC DEVELOPMEN T DRAFT REPORT Prepared by: R&M Consultants, Inc. Harza-Ebasco Susitna Joint Venture Arctic Environmental Information and Data Center, University of Alaska LGL Alaska Research Associates, Inc. Agriculture and Forestry Experiment Station, University of Alaska Submitted by: Arctic Environmental Information and Data Center University of Alaska-Fairbanks To: Harza-Ebasco Susitna Joint Venture 711 H. Street Anchorage, Alaska 99501 For: Alaska Power Authority 324 W. 5th Avenue, Second Floor Anchorage, Alaska 99501 April 5. 1985 This report was prepared by: R&M Consultants, Inc. C. s. Schoch Harza-Ebasc o Susitna Joint Venture D. W. Beaver H. W. Coleman E. J. Gemperline K. Jawed Arctic Environmental Information and Data Center University of Alaska S . V. Cuccarese R. J. Hensel M. D. Kelly J. C. LaBelle LGL Alaska Re s earch Associates, Inc. D. R. Herter R. H. Pollard D. G. Roseneau R. G. B. Senner W. D. Steigers, Jr. J . C. Truett .l. D. Woolington Agriculture and Forestry Experiment Station University of Alaska D. Helm TABLE OF CONTENTS i LIST OF TABLES ii LIST OF FIGURES VOLUME I I. EXECUTIVE SUMMARY II. INTRODUCTION •••••••••••••••••••••••••••••••••••••••••••••••••• A. B. c. PURPOSE ••••••••••••••••••••••••••••••••••••••••••••• SCOPE ••••••••••••••••••.••••••••••••••••••••••••••••• BACKGROUND •••••••••••••••••••••••••••••••••••••••••• 1 1 2 5 III . SUSITNA RIVER MORPHOLOGY AND CLIMATE.......................... 8 IV. A. B. c. UPPER RIVER ••••••••••••••••••••••••••••••••••••••••• MIDDLE RIVER •••••••••••••••••••••••••••••••••••••••• LOWER RIVER ••••••••••••••••••••••••••••••••••••••••• 1. River Mile 98.5 to RM 78 ••••••••••••••••••••••• 2. River Mile 78 to RM 51 .........•.•..•.•••...... 3. River Mile 51 to RM 42. 5 ••••••••••••••••••••••• 4. River Mile 42.5 to RM 2 7 ••••••••••••••••••••••• 5. River Mile 27 to RM 0 ...••.......•.••...•.•...• RIVER ICE PROCESSES ••••.•••.•••••••••••••••••••••••••••••••••• A. B. GENERAL FREEZEUP I CE PROCESSES ••••••••••••••••••••••••••• 1. Prazil ..................•••.....•................... 2. 3. 4. 5. Shore Ice .......................................... . lee Bridges ........................................ . lee Cover Progression .............................. . Anchor Ice ..•••........................ e •••••••••••• SUSITNA RIVER ICE COVER DEVELOPMENT •••••••••••••••••••••• 1. Cook Inlet to the Chulitna River Confluence ••••••••• 2. Chulitna River Confluence to Gold Creek ••••••••••••• 3. 4. 5. 6. 7. Gold Creek to Devil Canyon •••••••••••••••••••••••••• Devil Canyon (to Devil Creek) ••••••••••••••••••••••• Devil Canyon to the Oshetna River ••••••••••••••••••• The Freez eup of the Lower and Middle River Side Channels, Sloughs and Tributaries •••••••••••••• Summarized Historical Freezeup Chrono l ogies ••••••••• a. b. c. d. 1980 Freezeup 1981 Freezeup Chronology ..•........•.••..•...•• Chronology ••••..•••••.•..•.•••.•• 1982 Freezeup Chronology ••••••••••••••••• ~ ••••• 1983 Freezeup Chronology ••••••••••••••••••••••• 33RD2-007x 11 12 17 18 21 22 23 24 28 28 28 30 33 36 43 44 45 58 63 65 67 69 73 73 77 80 82 c. D. E. TABLE OF CONTENTS (Continued) e. 1984 Freezeup Chronology •••••••••••••••••••••• 86 GENERAL RIVER ICE BREAKUP PROCESSES •••••••••••••••••••••• 96 1. Channe 1 Leads. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 96 2. Candling............................................ 99 3. Ice Jams •••••••••••••••••••••••••••••••••••••••••••• 100 SUSITNA RIVER ICE COVER DISINTEGRATION ••••••••••••••••••• 100 102 103 105 105 108 112 123 125 1. 2. 3. 4. ICE 1. 2. 3. 4. Talkeetna to Cook Inlet ••••••••••••••••••••••••••••• Devil Canyon to Talkeetna •••.••••••••••••••••••••••• Sum.arized Historical Breakup Chronology •••••••••••• 1981 Breakup Chronology •••••••••••••••••••••••• 1982 Breakup Chronology •••••••••••••••••••••••• 1983 Breakup Chronology •••••••••••••••••••••••• a. b. c. d. 1984 Breakup Alternate Sources Chronology •••••••••••••••••••••••• of River Ice Information •••••••••• EFFECTS ON THE ENVIRONMENT ••••••••••••••••••••••••••• 127 Morphology and Vegetation ••••••••••••••••••••••••••• 128 Sediment Transport ••••••••••••••••••••••••••••• , •••• 130 Slough Overtopping •••••••••••••••••••••••••••••••••• 133 Groundwater. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 136 V. WITH-PROJECT STUDIES.. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 138 A. B. MEmODOL()(;Y AND SCOPE •••••••••••••••••••••••••••••••••••• 1. Reservoir Ice Modeling ••••.••••••••••••••••••••••••• 2. Instream Temperature Modeling ••••••••••••••••••••••• 3. Instream Ice Modeling ••••••••••••••••••••••••••••••• SIMULATION RESULTS ••••••••••••••••••••••••••••••••••••••• 1. Reservoir Ice •••••••••.••••••••.•••••••••••••••••••• a. Watana Operations •••••.•••••••••••••••••••••••• b. Watana Filling •••••••••••.••••••••••••••••••••• c. Devil Canyon Operation ••••••••••••••••••••••••• 2. Instream Temperature Simulations •••••••••••••••••••• a. The 0 C Isotherm Position •••••••••••••••••••••• b. Effects on the Ice Cover ••••••••••••••••••••••• 3. River lee ••••••••••••••••••••••••••••••••••••••••••• a. Preezeup ••••••••••••••••••••••••••••••••••••••• (1) Natural Conditions •••••••••••••••••••••••• b. c. (2) With-Project Conditions ••••••••••••••••••• (a) Project Operation •••••••••••••••••••• (b) Watana Filling ••••••••••••••••••••••• (3) Alternative Intake Designs and Operating Policy •••••••••••••••••••••••••••••••••••• (4) Case E-VI Flow Constraints •••••••••••••••• Breakup ••••••••••••••••••••••••••••••••••• · •• • • • Effects of Power Flow Variation on Ice 138 138 140 141 145 145 146 149 150 151 151 152 153 153 153 154 155 158 159 161 162 Cover. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 164 33RD2-007x 4. 5. 6. TABLE OF CONTENTS (Continued) ( 1) Experience Survey ••.•••••••••••••••••••••• (2) Susitna Operations ..•••••••••••••••••••••• Project Effects on River Cross-Section Characteristics .................................... . Effects on Slough and Side Channel Flow ••••••••••••• a. Q\lant 1 ty of Flow ..•.•...•••..........•.....••.. (1) Period immediately Prior to Ice Cover Formation .....•...••••..........•... (2) Within Ice Cov~red Areas •••••••••••••••••• (3) Upstream of the Ice Covered ••••••••••••••• b. Water Temperature ••••••..•..•..•••..••.•..•.•.• Ice Effects on Sedim·ent Transport and Channel Stability .................•......................... a. Mainstream Habitat .......•......•...•.••.••••.• (1) Suspended Sf!diment Concentration •••••••••• (2) Channel Stability •••••.••••••••••••••••••• b. Side Channe~.s •••••••••••••••••••••••••••••••••• (1) Suspended Sediment Concentration •••••••••• (2) Channel Stability ••••••••••••••••••••••••• c. Side Sloughs ................................... . (1) Suspended Sediment Concentration •••••••••• (2) Channel Stability •••••••••••••••••••.••••• d. Upland Sloughs ................................ . e. Tributary Mouths ................••........•.... (1) Suspended Sediment Concentration •••••••••• (2) Channel Stability ••••••••••••••••••••••.•• f. Tributaries ................................... . 164 167 167 170 171 172 173 174 175 179 180 180 181 183 183 184 184 184 185 186 186 186 186 187 VI. ENVIRONMENTAL EFFECTS • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 188 A. OVERVIEW OF ICE RELATED ISSUES 1. Sources of Issues ••••••••••••••••••••••••••••••••••• a. Fish and Aquatic Habitat ••••••••••••••••••••.•• b. Riparian V~getation •••••••••••••••••••••••••••• c. Wildlife and Terrestrial Habitat ••••••••••••••• d. Public Use .................................... . B. MECHANISMS OF EFFECT 1 . Impoundment Zone .....••••••••....•.••.••.........•.. a. Wa tana Filling ................................ . b. Watana Only On-Line ....•............•.•.......• ( 1) Freezeup. . . . . . . . . . . . . . . . . . . . ............ . (2) Meltout/Breakup ••••••••••••••••••••••••••• c. Watana and Devil Canyon On-Line •••••••••••••••• ( 1) Freezeup ................................. . (2) Meltout/Breakup ••••••••••••••••••••••••••• 2. Middle River Zone ..•••.....•..•..................... a. Freezeup ....••.....•........•....•.......•.•... b. Meltout/Break.up ...............•.•..........•... 3. Lower River .....•.......................•........... a. Fr•ezeup ...................................... . 33RD2-007x 188 189 189 190 190 191 192 192 192 193 194 194 195 195 195 197 197 197 c. D. E. TABLE OF CONTENTS (Continued) b. Mel tout /Breakup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 EFFECTS ON f ' ISHERIES ••••••••••••.••••••••••.•••••••.•••.• 1 . Introduction ........................................ . 2. Methodology ........................................ . 3. Impoundment Zone Effects Analysis •••••.••••••••.•••• a. Fish Resource ..........•....................... b . The With-Project Environment ••••••••••••••••••• c. Anticipated With-Project Effects ••••••••••••••• (1) Watana Reservoir ...........•.............. (2) Devil Canyon Reservoir •••••••••••••••••••• 4. Middle River Effects Analysis ••••••••••••••••••••••• a. Fish Resource ................................. . b. The With-Project Environment ••••••••.•••••••••• c. Anticipated With-Project Effects ••••••••••••••• 5. Lower River Effects Analysis •••••••••••••••••••••••• a . Fish Resources ••••••••••••••••••••••••••••••••• b. The With-Project Environment .•••••••••••••••••• c. Anticipated With-Project Effects ••••••••••••••• 6 • Su111D8. ry ......................•...................... EFFECTS ON RIPARIAN VEGETATION ••••••••••••••••••••••••••• 1. General ............................................ . a. Impoundment Zone ..•••••••.••••.••.••.•.•..•..•. b. Middle River Zone •...•.••...••..•............•. c. Lower River .......••.••••.•..•................. 2. Effects of Altered River Ice •••••••••••••••••••••.•• a . Impoundment Zone ............•.......•...•.••.•. b. Middle River Zone •.•...•.........•••••.......•. c. Lower River ..........•......................... EFFECTS ON WILDLIFE ..........•...•......•...•.••••..••.•. 1. General .............•........•...................... a. Impoundment Zone ...•••.••.••.....•..•........•. b. Middle River ...•....•....•....•.............•.. 2. Effects of Alteration of River Ice •••••••••••••••••• a. Impoundment Zone ..•..•••.....•..•.............. (1) Watana Dam On-Line Only ••••••••••••••••••• (a) Non uniform ice formation during 199 199 205 206 206 208 210 210 211 212 212 224 225 236 236 238 239 242 245 245 245 246 251 251 251 253 255 256 256 257 262 263 263 263 freezeup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264 b. (b) Ice deposition along reservoir margin during winter drawdown •••••••• 264 (c) Determination of reservoir ice cover in spring •.••••••••••••.••••••• 266 (d) Accumulation of windblow snow along impoundment shoreline •••••••••••••••• 267 (e) Increased extent of open water during winter ........••.........•.... (2) Watana and Devil Canyon Dams On-Line •••••• Middle River .•.....•.•..•.•..•......•..•...•••. ( 1) Wa tana Dam •••••••••••••••••••••••••••••••• 269 270 270 271 33RD2-v"7x F. TABLE OF CONTENTS (Continued) (a) Longer open-water period and larger open areas resulting from higher temperatures of regulated flows •••••• 271 (b) Higher staging resulting from increased winter flows ••••••••••••••• 273 (c) Early in situ melting of ice during spring breakup ••••••••••••••••••••••• 275 (2) Watana and Devil Canyon Dams On-line •••••• 277 (a) Longer open-water periods and larger open-water areas resulting from higher temperatures of regulated flows................................ 277 EFFECTS ON PUBLIC USE •••••••••••••••••••••••••••••••••••• 1. Natural Ice Cycle Chronology •••••••••••••••••••••••• 2. Public Use Perspective ........•...•.••.••••......... 3. Present User Group Activity ••••••.•••••••••••••••••• 4. Effects of Altered Ice Processes on Public Use •••••• 279 279 280 280 282 VII. REFERENCES (Chapter I-V, VI, A B C D E F) (Additional Reading) VIII. GLOSSARY OF TEIU>IS AND DEFINITIONS VOLtJlofE II FIGURES I-91.f- Appendix A Lower River Photography Appendix B Middle River Photography 33RD2-007x LIST OF TABLES Table No. Page No. 1. Slough and side channel study areas in the lower and middle Susitna River ...................................... . 2. Relative contribution of flows at Chulitna-Susi.tna- Talkeetna confluence....................................... LC) 3. Susitna River available weather data ••••••••••••••••••••••• ~7 4. Major annually recurring open leads between Sunshine RM 83 and Devil Canyon •••••••••••••••••••••••••••• 5. Susitna River Ice Thickness Summary 1981-84 •••••••••••••••• 6 . 1980-81 River Ice Summary •....••••••.•.•.••••..•••.•••••... 7. 1981-82 River Ice Su1111D8ry •••••••••••••••••••••••••••••••••• 8. 1982-83 River Ice Summary ..•.•.................•.......•.•• 9. 1983-84 River Ice Summary .••••••..•...••.••.•..•.•••••..••• 10. 1984 Freezeup .......................•............•..•...... 11. Common and scientific names of fish species recorded in th Susitna River Basin •••••• , •••••••••••••••••.••••••••• -;too 12. Susitna River salmon escapement estimates, 1981-84 ••••••••• 13. Simulated Susitna middle river ice front progression and winter overtopping of sloughs •••••••••••••••••••••••••• 14. Susitna hydroelectric project ice issues list for fish ••••• 15. Susitna river salmon escapement for the middle Susitna River, 1981-84 .........•.........•.....•........... 16. Susitna river salmon phenology ••••••••••••••••••••••••••••• 17. Peak salmon survey counts above Talkeetna for Susitna River tributary streams •••••••••••••••••••••••••••• 18. Peak slough escapement counts above Talkeetna •••••••••••••• 19. Chum salmon escapement for the ten most productive sloughs above RM 98.6 1981-83 •••••••••••••••••••••••••••••• 33RD2-007w LIST OF FIGURES Figure No. Vo \v~~ It Page No . 1. Basin Ma.p •••••••••••••••••••••••••••••••••••••••••••••••••• 2. Basin 'Ma.p Upper River ............•..•....................... 3. Basin Hap Middle River..................................... 3 4. Basin 'Map Lower River...................................... Jf 5. Average Historical Accumulated Freezing Degree (; Days 1980-1983 ••.•.••••••••.•••••••...•••.•• ., ••.•.••••.•.•• 6. Upper Susitna River Basin Air Temperature Variation •••••••• 7a. Middle River Mean Daily Air Temperature Record December 1984 ••••••••••••••••••••••••••••••••••.•••• /C 10. 11. 12. 13. 14. Photo of shore ice on the left bank at the Chulitna River Confluence ........................•.................. 15. Photo of Susitna River mouth ••.•••••••••••••••••••••••••••• 16. Photo of Yent~ River Confluence •••••••••••••••••••••••••••• 17. Photo of Kashwitna River Confluence •••••••••••••••••••••••• 18. Photo of Goose Creek Confluence •••••••••••••••••••••••••••• 19. Typical ice cover development on lower and middle river .............................................. . 20. Relative stage levels at selected sites during 1983 Susitna River freezeup ..••••••..•.....•••.•..•.••••.••••••• 21. Photo of Susitna-Chulitna-Talkeetna Confluence ••••••••••••• 33RD2-007s LIST OF FIGURES (Continued) Figure No. " IJ \ u .,...... (. :r:r Page No. 22. 23. 24. 25. 26a. 26b. 27a. 27b. 28. 29. 30. 31. 32. 33. 34. 35a. Photo of Middle River ice bridge and Leading Edge •••••••••• Photo of ice bridge at RM 105; November 4, 1984 •••••••••••• Photo of RM 105 November 17, 1982 •••••••••••••••••••••••••• Photo of Wlliskers Creek. ...........................••....... Photo of RM 103.3 December 28, 1983 •••••••••••••••••••••••• \) ~ (.~ I 9 ~ ~ # ., ~ 5 ) 'fC ~~ok ~~~ 6 RM I O ~ Photo of ~ ~.u ... .-\-.::.. 0 ... /\_..Q...~ 1\..o1..:- 1 I q ~ (_ ~ , , Y, ~-/ Photo of ~ ~{. CL'\..Lc...__, t) .oi.<..J lq &l 'L ~ 0., tf ~ ..) Photo of Shore Ice at Devil Canyon ••••••••••••••••••••••••• £"o Photo looking upstream from RM 134 December 1983 ••••••••••• Photo of anchor ice dam near Slough 21 ••••••••••••••••••••• Photo looking downstream at anchor ice dam at RM 144 • 5 ••••••••••••••••••••••••••••••••••••••.••••• Photo Gold Creek November 17, 1982 and December 23, 1982 •••••.•••••.•••••••••••••••••••••••••••••• Photo of Anchor Ice Dam ......•••.••.•••••.•.••....••.••.... Photo of Slough 8A November 17, 1982 and December 23 • 1982 •••••••••••••.•••••••••••••••••••••••••••• Photo of Slough 9 November 17, 1982 •••••••••••••••••••••••• 35b. Photo of Slough 9 December 23, 1982 •••••••••••••••••••••••• 61 33RD2-007s Figure No. 36. 37. 38. 39. 40. LIST OF FIG URES (Continued) \J 'j \ 'J ~ "t TI Page No. Annual variation of freezing degree days from 1380-1984 ............................................. . Annual variation of stream flow during freezeup ••••••••••.• lee front progression relative to cumulative freezing -ro degree days ............................................... . Water temperatures at Gold Creek ••••••••••••••••••••••••••• Photo of Candling .......................................... . 41. Photo of simple ice jam, RM 134............................ 7 ~ 42. Photo of an in situ ice cover melting during breakup ••••••• 1 ~ 43 . Photo of a grounded o r dry jam at RM 131 ••••••••••••••••••• 8 0 44 . Photo of ice jam producing island overtopping near RM. 131, May 7, 1983 .............•.......•..•.•........ 45. Photo of overtopped island and damage to mature vegetation .................................•.•............. 46. Photo of showing in an ice jam ••••••••••••••••••••••••••••• g ~ 47. 48. 49. 50. 51. Photo taken after release of ice jam at RM 101 ••••••••••••• Photo of a shear zone between flooding and grounded ice at RM 105..................................... '1 0 Photo of an ice shear wall, RM 137 ••••••••••••••••••••••••• q L Photo of rafting of large cobbles near Slough 21 ........... qy Photo of sediment deposited by ice ••••••••••••••••••••••••• ~' 52. Photo of formation of Slough 11 by ice scouring •••••••••••• 9 S 53. Susitna -Talkeetna air temperatures November -March •........................................... 54. Average monthly air temperatures at Talkeetna •••••••••••••• l O L 55 . Susitna River natural streamflow ••••••••••••••••••••••••••• t C3 56. Release constraints discharge for Susitna at Cold Creek. •••••••••••••••••••••••••••••••••••••••••••••• 33RD2-007s Figure No. 57. 58. 59. 60. 61. 62. LIST OF FIGURES (Continued) v c \ \J .,_., ~ Jl Page No. Watana Reservoir ice growth 1996 Case c ................... . Devil Canyon Reservoir ice growth 2002 Case C ••••••••.••••• Watana Reservoir water levels •••••••••••••••••••••••••••••• Watana Reservoir water levels Watana and Devil Canyon operating 2002 .........••....•.•..........•..........•..... Watana Reservoir Water Levels; Watana and Devil Canyon Operating 2020 Simulation Case E-VI •••••••••••••••••••••••• Devil Canyon Water Levels 2002 Simulation Case E-VI •••••••• ,o5 /0 ~ 10 7 1 ocr 1/C 63. Devil Canyon Water Levels.................................. II I 64 . Photo reservoir bank erosion caused and influenced by ice cover ..................................... • • • · · ... • · } f "2... 65. Middle Reach Ice Front Maximum Progression................. {I 3 66. Middle Reach Ice Cover Duration............................ II J.J 67. Su Hydro Maximum Simulated Winter River Stages •••••••••••.• 68. Susitno Hydro -Total Ice Thickness Maximum Simulated Values .•......................................... II ftJ 69. Winter 1971-72 maximum slough stages....................... Jl7 70. Winter 1976-77 maximum slough stages....................... 1/ ~ 71. Winter 1981-82 maximum slough stages....................... /I q 72 . Winter 1982-83 maximum slough stages ••••••••••••••••••••••• {~u 73 . Winter 1971-72 Total Ice Thickness ••••••••••.•••••••••••••• p .. J 74. Winter 1976-77 Total lee Thickness •••••••••••••••••••••••• 75. Winter 1981-82 Total Ice Thickness ••••••••••••••••••••••••• 76. Winter 1982-83 Total Ice Thickness ••••••••••••••••••••••••• 77. Watana Filling Maximum Slough Stages ••••••••••••••••••••••• 78. Winter 1971-72 Filling Maximum Slough Stag ~s •••••••••••.••• 79 . Winter 1981-82 Maximum Slough Stages Case C vs. 33RD2-007s Figure No. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. LIST OF FIGURES (Continued) v 6 ' \) VV> o( .:rr Page No. Case E-V I 200 1 .•.•........................•.•.............. Winter 1981-82 Maximum Slough Stages Case C Vs. Case E-Vl 2002 ., • ,. <"' • wo Winter 1981-82 Middle Reach Freeze-up Simulated Pre-Project ................................................ . Winter 1981-82 Middle Reach Freezeup With-Project-1996 ••••• Middle Reach Freeze-up With-Project Winter 1981-82 ••••••••• Middle Reach Freeze-up Simulated Preproject Winter 1982-83 ..............•.....................•........ Middle Reach Freezeup Simulated Preproject Winter 1982-83 ••••••••••••••••••••••••••••••••••••.•••••••• Middle Reach freezeup with-project 1996 winter 1982-f.3 ••••• Plan view showing flow components typical slough ••••••••••. Detail of Intragravel Flow From Exhibit 34 ••••••••••••••••• Susitna River Drainage Basin Species by Study Zone ••••••••• Month river crossing initiated ••••••••••••••••.•••••••••••• Nelchina caribout head estimates ••••••••••••••••••••••••••• Historical range of the Nelchina caribout herd ••••••••••••• Dall sheep study area ..................................... . The Jay Creek mineral lick complex ••••••••••••••••••••••••• ;7-7 ,-;;z.q 3o l 3 I I 3 2 t4 0 I c.f I I'/ 1. 33RD~-007s I. SlJMlofARY (To be written) II. INTRODUCTION A. PURPOSE The Susitna ~iver system is dominated by river ice for more than hal f of each year. The processes of ice formation in the late fall, ice development and evolution thoughout the winter, and decay of ice in the early spring have large effects on the river's morphology and fish and wildlife habitats, as well as patt~rns of human use . The purpose of this report is to summarize all that has been learned to date regarding natural river ice processes in the Susitna River, the predicted alteration of those ice processes by the proposed Susitna Hydrolelectric Project, and the effects of altered ice processes on aquatic and terrestrial habitats, fish, wildlife, and vegetation, and man's use of the Susitna basin. This assessment is an integral component of an overall environmental investigation of the effects of the Susitna Hydrolelectric Project on fish , wildlife, and the human environment. This report is part of a series, called the lnstream Flow Relationships Report Series (!FRS), which summarize aquatic environmental studies. This report incorvorates the work of many groups over the past five years, all of whom contributed dir ectly or indirectly to this effort . Especially notable are the efforts of the following organizations: I 33tu>2-007v R&M Consultants, Inc, who performed river ice observa tions and analyses every winter; Harza-Ebasco Susitna Joint Venture (H-E), who carried out computer simulation modeling and analyses of natural and with-project river ice processes; The Alaska Department of Fish and Game (ADF&G), who performed winter field work in terrestrial and aquatic studies; The Arctic Environmental Information and Data Center (AEIDC), Univ. of Alaska, who carried out instream temperature simulations by computer model and performed fish and aquatic habitat analyses; LGL Alaska \ssociates, Inc., who performed studies of wildlife and terrestrial habitat in the Susitna Basin; The Agriculture and Forestry Experiment Station (AFES), Univ. of Alaska, who carried out riparian vegetation studies along the river system. The report is organized into two volumes. The first volume includes text and tables; the second volume contains all figures and appendixes. Figures are not collated into the text because many of them will be in 11 by 17-inch format in the final report. For the draft version of the report, all figures and maps have been reduced to 8 1/2 by 11-inch size. B. SCOPE 33RD2-007v This report analyses river ice processes under both natural and with-project conditions, and describes the expected effects of projec t-altered ice conditions on the biological environment, for the entire Susitna River system including the proposed impoundment areas. Specifically addressed are the following subjects : 1. Natural river ice processes in the Susitna River as observed for the past five years, including freezeup, mid-winter ice development, and breakup processes; 2. Predicted alterations of river ice processes by the proposed Susitna Hydroelectric Project, produced by computer simulation modeling of with-project instream temperatures and ice processes (SNTEMP and !CECAL). These include predicted ice conditions in the proposed Watana and Devil Canyon reservoirs, and river ice regimes under various operational scenarios including Watana filling, Watana alone on line, and both Watana and Devil Canyon on h .ne. Each of these scenarios is tested against varying climatic conditions, for wh i ch the climatic years 1971-72 (cold), 1976-77 (very warm), 1981-82 (average), and 1982-83 (warm) are used; 3. Predicted effects of altered river ice processes on terrestrial and aquatic habitats, fish, wildlife, and vegetation, and public use. Many of the conclusions reached in Chapter VI of thi$ report are necessarily speculative, as few studies have been carried out regarding the effects of natural ice processes on fish. wildlife, or vegetation. Also, little information exists on the effects of river ice processes altered by 33RD2-007v J other northern hydroele c tric facilit i es on fish and wildlife habitats. The statements of fact or conclusions contained in this report have been derived fro• several sources. These ar ~: 1. Field observations and interpretations by investigators contractually involved with determining the environmental effects of the Susitna Hydroelectric Project; 2. Information on the Susitna River basin anj its resources and environs contained in both accredited journals and in gray literature; 3. Information on the physical and biological effects of river ice processes learned from studies performed in other river systems,prin~ipally reported in accredited journals. ~~·~ever possible, statements in the report purporting to be factual are documented as • so~rce. Conclusions or expressions of professional opinion are similarly documented unless they are the opinions of the authors. Additional relevant winter studies in the Susitna Basin are now in progress by ADF&G, but these results were not available in time for inclusion in this report. Conclusions regarding the effects of altered ice regimes on fish and wildlife and their habitats rest entirely upon the with-project river ice scenarios predicted by computer simulation models. These simulation models, while state-of-the-art, have limited predictive capabilities. For example, the !CECAL model is one-dimensional in scope, limiting its capabilities to predict ice processes in peripheral parts of the river where important fish 33RD2-007v habitat often occurs . Als~ • the model is incapable of modeling spring ice bre akup processes; conclusions about spring meltout are based on professional judgement using !CECAL-predictea ice regimes. This report, therefore, serv,s to give project planners the best view of with-project river ice conditic:ms and their effects on the biological environment that is possible wi th presenl predictive technol ogy. Tne authors believe that the scenarios discussed herein ~re the most likely to result from t'tte project. C. BACKGROUND The Susitna River drains an area of 19,600 .•quare miles, the sixth largest river basin in Alaska. The Susitna flows 320 m.'les from its origin at Susitna Glacier to the Cook Inlet estuary. Its basin is b,rdered by the Alaska Range on the north, the Chulitna and Talkeetna mountain: on the west and south, and the northern Talkeetna plateau and Gulkana upla1 ds to the east. This area is largely within the coastal trough of southcentral Alaska, a belt of lowlands extending the length of the Pacific mountaiL system and interrupted in Alaska by the Talkeetna, Clearwater, and ~rangell m1untains. Major Susitna tributaries include the Talkeetna, Chulitna, ~nd Yentna rivers (figure 1). The Yentna River enters the Susitna at river mil~ (RM) 28 (28 miles upstream from the mouth ?t Cook Inlet). The Chulitna River ·ises in the glaciers on the south slope of Mount McKinley and flows south, e1tering the Susitna River at RH 99 near Talkeetna. The Talkeetna River originat ·s in the Talkeetna Mountains, flows west, and joins the Susitna at RH 97 1 ear Talkeetna. Tributaries in the no~thern portions of the Susitna basin originate 1 . the glaciers of the eastern Alaska Range. The east and west forks of the 33RD2-007v s Susitna and Maclaren rivers join the mainstem Susitna River above RM 260. Below the glaciers the braided channel traverses a high plateau and continues south to the Oshetna River confluence near RM 233. There it takes a sharp turn west and flows through a steeply cut canyon which includes the Watana (RM 184.4) and Devil Canyon (RM 151.6) damsites. In this predominantly single-channel reach the gradient is quite steep, averaging approximately 10 feet per mile (Acres American 1983). Below Gold Creek (RM 137) the river alternates between single and multiple channels until the confluence with the Chulitna and Talkeetna rivers (RM 97), below which the Susitna broadens into widely braided channels for 97 miles to Cook Inlet. The proposed project consists of two dams to be constructed over a period of about 15 years. The Watana dam would be completed in 1994 at a site 3 miles upstream from Tsusena Creek (RM 184.4). This development would include an underground powerhouse and an 885-foot high ~arthfill dam, which would impound a reservoir 48 miles long with a surface area of 38,000 acres and a usable storage capacity of 3 .7 million acre feet (maf). The dam would house multiple level intakes and cone valves. Installed generating capacity would be 1020 megawatts (Mw), with an estimated average annual energy output of 3460 gigawatt hours (gwh) (Acres American, Inc. 1983). The concrete arch Devil Canyon dam would be completed by 2002 at a site 32 miles downstream of the Watana damsite. It would be 645 feet high and would impound a 26 mile-long reservoir witt 7,800 surface acres and a storage capacity of • 36 maf. Installed generating capacity would be about 600 Hw, with an average annual energy output of 3450 gwh (Acres American, Inc. 1983). Construction and subsequent operation of the Susitna dams are expected to alter the norma.. flow and thermal regimes of the river . Mainstem flows downstream of the project would be higher in the winter than they are 33RD2-007v naturally. Mainstem water temperatures downstream from the projec: would be cooler in the summer and warmer in the winter than under natural conditions. A change in the river ice regime downstream from the project is expected due to these al t ered flows and tecperatures . 33RD2-007v 7 II I ~ SUSITNA RI VE R MOR PHOLOGY AND C LI~~TE The Susitna River drainage basin, sixth largest in Alaska, is located in the Cook Inlet subregion of southcentral Alaska. The drainage basin covers 19.600 square miles. It is bordered on the west and north by the Alaskan Range, on the east by the Talkeetna Mountains and the Copper River lowl ands, and on the south by Cook Inlet. The river is 320 miles long from the mouth at Cook Inlet to the headwaters at Susi tna Glacier . Major tributaries include the Chulitna, Talkeetna, and Yentna _vers, all located downstr1.!am of the proposed project. Extensive glaciers in the headwaters contribute substantial suspended sediment loads during summer months. Streamflow is characterized by high flows between May and September and low flows from December to April. The headwaters of the Susitna River and the major upper basin tributaries are characterize d by broad, braided, gravel floodplains emanating from glaciers on the south flank of the Alaskan Range. Below the West Fork tributary confluence, the river develops a split-channel configuration with numerous gravel bars, flowing south between narrow bluffs for about 55 miles. Below the confluence with the Oshetna River, the Susitna River flows west for 96 miles through steep-walled canyons before reaching the mouth of Devil Canyon. This reach contains the Watana and Devil Canyon damsites at River Hiles {RM) 184.4 and 151.6, respectively, as measured from Cook Inlet. River gradients are quite high, averaging nearly 14 feet/mile in the 54 miles above Watana damsite, 10.4 feet/mile from Watana downstream to Devil Creek, and 31 feet/mile in the 12-mile stretch between Devil Creek and Devil Canyon. Below Devil Cany on, the gradient decreases from about 14 feet/mile to 8 feet/mile above Talkeetna . The river in this reach is generally characterized b y a split-channel coufiguration , with numerous side-channels and sloughs. About 4 33RD1-007m - 8 - miles above tri~ confluence with the Chulitna River, the Susitna River begins to braid, and remains braidec:l the remainder of its length to Cook In]et. Numerous islands and side channels appear. The gradient continues to decrease, ranging from 5.5 feet/mile for the 34-mile r each below Talkeetna to 1.6 feet/mile for the last 42 miles. In order to facilitate morphological descriptions, this report refers to three distinct and easily identifiable river reaches labelled the upper, middle and lower reaches. These river reaches have been referred to in other reports concerning a variety of specific studies. This report deals with ice, the formation of which is primarily controlled by air temperature but to a great extent is affected directly by solar radiation as well. For the following discussions regarding river ice, the "upper river" will refer to the initial reach which is subjected to colder air temperatures due to the higher elevation and latitude of th~ headwaters, but also receives a substantial amount of solar radiation during the freezeup period because of the north-south orientation and lack of major topographic features (figure 2). The "middle reach" is the section of river that flows generally east to west, from tlae vicinity of the Oshetna tributary confluence, through the Watana and Devil Canyon impoundment areas, then southwest past Gold Creek and ending at the Chulitna River confluence. This reach flows through a mountainous area where steep canyon walls shade the turbulent water surface for much of the year (figure 3). Downstream of the Chulitna River confluence the river mvrphology changes suddenly from essentially a narrow confined channel with a steep gradient to a broad channel containing a braided flow pattern. This configuration is retained through this final "lower reach" to Cook Inlet (figure 4). 33RDl-007m - 9 - Identification of i pecific sites on the r i ver is best don e with place-names . However, in some areas the lesser features have no names or the cartographers have failed t o a ttach one, so a substitute system of river mile (RM) numbers was developed with a common refecence point (Cook Inlet, &~ 0) s n that all chose concerned with the river study could mutually orient themselves. Photomo saic river maps snowing the Susitna from Cook Inlet to the proposed Devil Can y on dam site are included in Appendixes A and B. No maps are available of the river upstream of Devil Canyon except for the standard U.S. Geological Survey (USGS) topographic sheets at 1:63,360 scale. The entire USGS map set showing the Susitna River from Cook Inlet to the Susitna Glacier has been compiled and reproduced with river mile numbers in the Susitna River Mile Index (R&H, 1981a). The Susitna River originates in the continental climatic zone, flowing south into the transitional climatic zone. Due to the maritime influence and the lower elevations, temperatures are more moderate in the lower basin than in the upper basin. Freezing temperatures occur in the upper basin by mid-September, with frazil ice generated in the reach from Denali through Vee Canyon by early October. Several meteorological stations have been installed along the river since 1980. Records from these stations, located at Susitna Glacier, Denali, Kosina Creek (between Vee Canyon and Watana), Watana, tevil Canyon and Shen:~an, together with records from the National Weather Service (NWS) at Talkeetna, illustrate the 1harp difference in freezing degree-days along the length of the river (figure S). In general, the meteorology within the Susitna River basin is highly variable between weather station s ites. This is due, in part, to the movement of storm systems, the topographic variance, and the change in latitude, but the major reason for the temperature variance between Denali and 33RD1-007m -10 - Talkeetr.a is the 2,400-foot elevation dif f erence. Of the seven weather stations currently in oper~tion, only three will be considered in this report as prc-viding repre.,e:ntative data for describing ice processes in the upper, middle, and lower reaches. These stations are located at Susitna Lodge on the Denali Highway, at Watana Camp, and at Talkeetna. The following section~ discuss the specific morphological and climatological characteristic:; unique to each section and how they relate to river ice formation. A. UPPER RIVER The waters flowing through this reach originate primarily from four major glaciers and to a lesser extent as runoff via nu~erous tributaries. Meltwater and runoff drain from the West Fork, Susitna and East Fork Glaciers, flow through broad gravel floodplains and merge into a single channel, to flow through a narrow pass between the Clearwater Mountains on the east and an unnamed range to the west. The Denali Highway bridges the river at this point (RM 291). The u.s. Geological Survey records daily river stages at the bridge. This gage provides information for computations of daily flow, which averages 2, 759 cubic feet/second (cfs ) from a drainage area of 950 square miles. During freezeup, or between October and December of any given year, tne flow drops rapidly from about 2,000 cfs to under 400 cfs. The lowest flow occurs in March, and is usually estimated at less than 300 cfs. Just to the south and eas t of the highway bridge, at Susitna Lodge, R&H Consultants operates a weather recorder. This station monitors air temperature, wind speed and direction, solar radiation, humidity and precipitation. Along the Susitna River, freezing air temperatures are generally first recorded at the Denali weather station. 33RD1-007m -11 - Downstreac of the Oena l i a rea, the river develops a split-channel config- uration •ith man y gravel oars but few vegetated islands. The route meanders through a broad plain, wit n the channel generally confined by low bluffs . The Maclaren River ent ~rs from the northeast about 31 miles downstream of the Denali Highway Bridge. Average annual discharge of the Maclaren River is 979 cfs. During freezeup the flow drops from about 700 cfs in October to 200 cfs by the end of December. This river drains the Maclaren Glacier and a large portion of the C1.earwater Mountains. Fourteen miles further downstream the non-glacial Tyone ~lver enters the Susitna from the southeast, draining the lakes Louise, Susitna and Tyone, as well as hundreds of square miles of muskeg and black-spruce bogs. The last major tributary entering this reach is the Oshetna River (RM 233). This river flows north into the Susitna, draining the north flank of the Talkeetna Mountains. Several tributaries to the Oshetna are glacial, the largest being the Black River which emerges from a sizeable unnamed glacier. The climate in this upper river reach is characterized by being colder, drier and sunnier than in the lower reaches. Figure 6 shows a comparison plot of air temperatures recorded at Talkeetna, Watana, and Denali during the 1984 freezeup. Storm systems usually pass to the west of this river reach, funneled northward through the lower Susitna Valley and over Chulitna Pass. The Talkeetna Mountains rise to over 8,000 feet and cause much of the water in the warm maritime air masses to precipitate out, so that the region to the north and east of this range is in a rain shadow. B. MIDDLE RIVER This reach description focuses on the general east-west course of the Susitna River beginning at approximately RM 233. 33RD1-007m -12 - The unique characteristics of this reach are the steep gradient of the channel and the steep-walled canyon throu,~h which the river flows. The elevation drops from about 2,150 feet down to 350 feet in roughly 140 miles, for an average gradient of 13.4 feet /mile. This contrasts considerably with the upper river gradient of 5.9 feet/mile and the lower river gradient of 3.6 feet/mile. Downstream of the Oshetna River confluence the Susitna enters Vee Canyon, the site of a U.S.G.S. streamgage. This gage measures flow draining an area of 4,140 square miles. The average annual flow is 6,404 cfs. From October to December, the flow drops from roughly 5000 cfs to under 1400 cfs. The middle reach is primarily either a single channel or a split channel with intermittent vegetated islands and gravel bars. The water is very turbulent through the entire reach. The river valley or canyon is generally quite deep, averaging about 1000 feet at the proposed Watana damsite. and the mountainous terrain along the south bank shields the river from direct sunlight for much of the year. This causes an air temperature gradient between the cold canyon bottom and the warmer plateau adjacent to the river. This gradient is especially evident during the winter, when sun angles are lowest and dense cold air settles in the canyon. In December, 1984, temperature deviations of over 10 C were measured between the Watana weather station and a thermograph located near the water surface . The average monthly deviation, however, measured between 2-3 C (figure 7). The weather recorder at Watana Camp is similar to the recorder at Denali. Winter precipitation is not generally measured at the weather stations except for what snow may accumulate on the ground. However, at Watana a Wyoming gage has been operating since October 1981, giving daily precipitation readings 33RD1-007m -13 - when a heated tipping bucket was operating. Since October 1983 monthly totals have been measured froc an accumulating snowgage charged with an ant ifreeze mixture. These data allow a comparison between winter precipitation at Talkeetna anrl Watana. Figure 8 summarizes precipitation data over the winters of 1982-1983 and 1983-1984. The effects of storm patterns is illustrated by the large volumes of snowfall at Talkeetna compared to Watana. The combination of turbulent water, cold air temperatures and little solar radiation creates conditions where massive volumes of ice can form. This reach of river is therefore a major source of ice during freezeup compared to the upper and lower reaches. The upper reach has cold air temperatures but lacks the turbulence necessary to generate large volumes of ice. In October the sun shines directly on the water surface for much of the day, raising the effective water temperature, if not the air temperature, sufficiently to prevent further ice from forming. In the shaded middle reach, freezing air temperatures are sustained, allowing ice to form over a longer period of til!le. In contrast, the lower river has neither the cold air temperatures in October nor the turbulence to generate much ice. The reach between Devil Canyon and the Chulitna River confluence has received considerable attention during the project environmental studies. Project impacts would be most evident in this reach because no major tributaries enter the Susitna River to offset the effects of flow regulation. Smaller tributaries include Portage Creek, Indian River, Gold Creek and Fourth of July Creek. The U.S.G.S. maintains a streamflow recorder and conducts monthly measurements of discharge at the Gold Creek Bridge, where the Alaska Railroad crosses the Susitna River (RM 136.6). This gage measures the flow from a drainage area of 6,160 square miles. The average annual discharge is 9,724 cfs. Freezeup flows range from about 10,000 cfs in October to under 33RD1-007m -14 - TABLE 1 SLOUGH AND SIDE CHANNEL STUDY AREAS IN LOWER AND MID DLE SUSITNA RIVER Observed Freezeup Year of Threshold Name Location lee Effects Observation Elevation (River Mile) (Feet) Hooligan Side Channel 35.2H None 1984 Unknown Eagles Nest Side Channel 36.2H Some flooded snow 84 Unknown Kroto Slough Head 36.3H None 84 Unknown Rolly Creek Mouth 39.0M None 84 Unknown Bear Bait Side Channel 43.0H None 84 Unknown Last Chance Side Channel 45.4H None 84 Unknown Rustic Wilderness Side Ch 59.5H Overtopped 83,84 Unknown Caswell Creek -Mouth 63.0M None 84 Unknown Island Side Channel 63.2M Flooded snow 84 Unknown Mai n stem West Ba n k 74.4M Some flooded snow 84 Unknown Circular Side Channel 75 .6H None 8 4 Unknown Goose 2 Side Channel 75.8H Overtopped 83,84 Unknown Sauna Side Channel 79.8H Non e 84 Unknown Sucker Side Channel 84 .5M 11· me 84 Unknown Beaver Dam Slough 86.3M None 84 Unknown Sunset Side Channel 86.9 None 84 Unknown Sunrise Side Channel 87.0 None 84 Unknown Birch Creek Slough 88.4M Ncne 83,84 Unkn own Trapper Creek Side Channel 91.6 None 83 ,84 Unknown Whiskers Slough 101 .5H Overtopped 80-84 367 Side Channel at Head of Gash Creek 112.0 Overtopped 82,83,84 Unknown Slough 6A 112.3M Backwater 80-84 u Slough 8 114 .lH None 83 476 Side Channel MSII 115.5 Overtopped 82,83,84 482 Side Channel MSII 115. 9H None 8 2 ,83,84 4S7 Curry Slough 120.0H None 84 Unknown Moose Slough 123.5H None 84 Unknown Slough SA-West 126.1H Overtopped 8l,S2 .83,S4 573 Slough SA-East 127.1H Overtopped S1,S2 ,83,84 582 Slough 9 129.3H Some flooded snow 81,S2,S3,84 604 Side Channel Upstream of Slough 9 130.6 None S2,S3,84 Unknown Side Channel Upstream of 4th of July Creek 131.S None 82,S3,84 Unknown Slough 9A 133 . 7H None S3,84 651 Side Channel Upstream of Slough 10 134.3 None 82,83,S4 65 7 Side Channel Downstream of Slough 11 135.3 None 8 2 ,S3,84 Unknown Slough 11 136. SH None 82,83 ,84 687 33RD1-007m Name Slough 17 Slough 20 Slough 21-Entrance A6 Slough 21 Slough 22 TABLE 1 (c o..,f',.,· •• Jj SLOUGH AND SIDE CHANNEL STUDY AREAS IN LOWER AND MIDDLE SUSITNA RIVER Location (River Mile) 139.3H 140.5H 141. 8H 142.2H 144.8H Observed Freezeup Ice Effects None None None None None Year of Observation 82.83 82,83 82.83 82,83 82.83 H -Indicated location represents the head of the slough or channel. M -Indicated location represents the mouth of the slough or channel. U .. "Upland" slough wtth no upstream head or berm. 33RD1-007m Threshold Elevation (Feet) Unknown 730 747 755 788 3.000 cfs by the end of December. Downstream of the Gold Creek Bridge. the river gradually resumes a more southe r ly f l ow direction. retaining the steep gradient and mostly split channel configura tion. R&H Consultants operates a weather station at Sherman. which together with the Watana Station provides representative data for the middle river reach. Downstream of Devil Canyon there are a series of sloughs and side channel habitats that are particularly sensitive to changes in the mainstem flo·w regime. The project-name given to these sloughs for identification. and the river mile location of the upstream entrances are included in table 1. The majority of these habitats have both an upstream entrance and a downstream exit. During high water events in the mainstem. flood flows overtop the entrance and spill into these overflow channels. The upstream entrances arc often protected to some extent by a low gravel berm. These berms have been observed to form when ice floes are pushed laterally from the mainstem by forces usually generated in ice jams. The floes contact the .channel bottom and shove gravel 1 cobbles or soil before them. ultimately forming these berms. A critical mainstem stage must be achieved before overtopping of the berms occurs. These critical elevations are also listed in table 1. At low mainstem flows the berms are not overtopped and the sloughs often convey clear water from small tributaries and upwelling groundwater. C. LOWER RIVER This final reach begins at the confluence of the Susitna and Chulitna Rivers at RM 98. 5. During the spring. summer and fall the Chulitna River contributes most of the sedjment to the lower river. The large material is deposited as the river gradient decreases. The deposition of sediment eventually causes the unconfined river channel to shift. This on-going 33RD1-007m -x- 11 proc ess results in numerou s interlac ed channels. Upstream of ots confluence with the Susitna, the Chulit na River currently has two major channels. The larger one flows along the northeast (left) bank, while the other flows along the extreme right bank. During freezeup the r.ight bank channel usually de-waters at flows under 4,000 cfs, and the left bank channe l contains all the Chulitna flow. The exact confluence of this channel and the Susitna varies from year to year. At the Chulitna Canyon, the USGS maintains a gage measurin g streamflow from a drainage area of about 2,570 square miles. The average annual discharge is 8, 798 cfs. This is about 90% of the average annual flow measured at Gold Creek , although the drainage area is less than half (approximately 40 %) the size. To some extent this is due to the higher perc entage of the basin that is glaciated, but otherwise indicates the high volume of precipitation this region receives compared to the Watana and Denali area. The Chulitna River flow decreases rapidly during freezeup from about 10,000 cfs in early October to about 2,000 cfs by the end of December. The lower river has been subdivided into five reaches, each with distinct characteristics: Segment 1, RH 98.5 at the Chulitna confluence to RH 78 near the confluence with Montana Creek; Segment 2, RH 78 to RH 51, which is approximately the upstream enri of the Delta Islands ; Segment 3, RM 51 to RM 42.5 through the Delta Islands ; Segment 4, RH 42.5 to the Yentna River confluence at RH 27; and Segment 5, the remaining reach from RM 27 to Cook Inlet . The following discussion presents brief descriptions of each river segment including pertinent data based on photo-interpretation and field observations. 1. SEGMENT 1 : RIVER MILE 98.5 TO RIVER MILE 78 33RD1-007m -~- IY The Talkeetna River f lows into the Susitna from the northeas t, upstream and adjacen t to the town of Talkeetna. This r i ver is also gaged by the US GS at a site about 3 miles up s tream of the Susitna confluence. The strea~flow is measu red from a drainage area of about 2 ,006 squar e miles. The average dis charge is 4,055 cfs. The t y p i cal range of flow during fre:ezeup is fr om about 6,000 cfs in early October to about 1,000 cfs by the end of December. The relative flow contributions from the Susitna, Chulitna and Talkeetna Rivers have been summarized in table 2. Downstream from the three-river confluence area, tht. Susitna continues t h rough a broad, low floodplain with multiple, interlaced channels . This network genera l l y consists o f the main channel and a s eries of secondary cha nnels. The main channel meanders irregul arly a cross the wide floodplain, occasionally contacting the steep bluffs of the su~rounding terrace . Secondary channels are usually flooded during the spring a nd summer high w':er period only, since their thalweg elevation is higher than the main channel. They are generally located near or against veget a ted islands or directly along either bank, and usually on the opposite side o f the floodplain from the main channel. The main channel flow continues all year while most of the secondary channels normally de-water at some time during the winter, not necessarily prior to freezeup . The floodpla i n consis t s mostly of gravel bars and some partially vegetated islands. Sev eral complex side channel systems e x i s t but these are generally flooded only at flows exceeding 13,000 cfs at Sunshine. Some side channels have a separate source of wa ter, either from z tributary or groundwate r seeps, and are cons i dered side s !oughs at lower d ischarges. These s i de channels are separate d from the mainstem by l arge heavi l y vegetated i slands, and ma y occur along either the left or right bank. Birc h, Sunshine , 33RD1-007m -~ (1 R20/16b TABLE 2 RE LATIVE CONTRIBUTION OF FLOWS AT CHULIT NA-SU SITNA-TALKEETNA CO NFLUEN CE ( PRE-PROJ £C T) Tota 1 Flow Contribution b~ (cfs Fl o w D/S(cfs) Pe rce n t~ b~ Chu I i tna I 1 l Ta I keetna( ll Susitna (ll Talkeetna Chu I i J.na Talkee!,na Susi!<na October 4859 2537 5 639 13U35 r t % 20% 4 3% Nov ember 1994 1187 2 467 56 48 35% 2 1% 411% De cember 14 57 838 1773 4068 36% 2 1% 11 3% January 1276 67 1 14 5 4 31t 0 1 37% 20% 43% February 1095 565 1236 2896 38% 19% ld~ March 976 4:'12 11 111 2582 38 % 19% 113% N Apri I 1 158 557 1 368 3083 38% 18 % 1111% 0 May 851 1 4176 13317 26004 33% 16% 5 1% J une 225 40 1191 0 27928 6 2378 36% 1\1% 4~% July 26330 10390 23 8 53 60573 411% 1 7% 3\1% August 22 190 9 '149 2 14 79 53418 42% 18 % 4 u% Sep t em be r 11 740 5853 13171 30764 38% 19% 4 3% Annua l 8 ·fl18 11 0 86 9567 221101 39% 18% 4 3% ( 1 ) 0 i sc ha ry e data from U.S.G .S. records up to September, 1981 . Source: Bre dthaue r and Drage , 1982 . Rabideux and \o."hitefish sloughs are the most extensive :md significay;-,t side channel systems along this reach . Six tributaries enter this reach, including the Chulitna and Talkeetna rivers, Lesser contributions are added by Trapper, Birch, Sunshine, and Rabideux creeks. The Susitna River downstream of Talkeetna is confined to only one channel at few places, most notably at the Parks Highway Bridge area called Sunshine, and immediately below the Yentna River confluence at Susitna Statton. The USGS monitors streamflow at both sites. At Sunshine the gage measures the cummulative flow from the Chulitna, Susitna an<.! l'alkeetna rivers, a drainage area of about 11,100 square miles. The mean annual discharge at this site is about 24,000 cfs (unofficial estimate) with the flow usually diminishing from about 25,000 cfs in early October to about 5500 cfs by the end of December. 2. SEGMENT 2 : RIVER MILE 78 TO RIVER MILE 51 This reach is characterized by extensive side channel complexes along the entire reach. These consist of a network of interconnecting channels which are normally flooded only at high flows or during the elevated stages induced by an ice cover. Many of the outermost channels in the complexes are fed by one or more tributaries which keep water flowing in a small portion of the side channel regardless of the mainstem flow. Six significant tributaries enter this reach, although only Montana Creek enters the Susitna mainstem directly. Goose Creek, Sheep Creek, Kashwitna River, 197 Mile Creek and Caswell Creek enter side channels which are isolated from the mainstem except at high water stages. The gradient through this segment starts out at 6 feet/mile and decreases near the Delta Islands for an average of 5.6 feet/mile. This segment has the 33RD1-007m -)(- 2..{ steepest s lope on the lower river and subsequently has the highest velocities. Due to mechanical thickening (shoving), this reach also has the thickest ice cover. The mainstem (excluding the side channel complexP.s) appears simil~r to the main channel in Segment 1, with a broad expansE: of gravel and sand bars exposed at low flows when the mainstem is generally confined to one or two channels. The maximum width of the flood plain is 6,000 feet and tre minimum is 1,000 feet. The majority of the gravel bars are devoid of veg~tation. High summer flows generally inundate the gravel bars, with debris carried along by the flow often piling up on the islands as log jams. At high flows, the water breaches the entrances to side channels and spills into these systems. The side channels seem to function primarily as overflow channels, diverting water away from the mainstem during floods. 3. SEG!-!ENT 3: RIVER !-!ILE 51 TO RIV ER HILE 42.5 (Delta Islands) This reach runs through an intricate system of islands. The mainstem at some high flows becomes diffused and is difficult to differentiate from side channels. Only at the low flows prior to freeze-up can the thalweg be defined. Even then it is split into two channels flowing along both the extreme left anri right banks. The majority of the side channels are dewatered at these low flows. The maximum channe l width is 4,500 feet at RM 51, with the narrowest portion of 700 f~et at RM 4 2 .5. RM 42.5 also marks the joining or convergence of the two main channels emErging from the Delta Islands anJ the end of this segment. Field investigations documented ground water seeps entering several of the side channels, providing these with a separate source of water isolated from the mainstem. The groundwater seeps are probably related to the mainstem stage since the contrihutto n of flow by groundwater in the side channel seems to diminish with lower water levels in the mainstem. 33RD1-007m -)(- 22.... Two tributaries enter this rear.h along the east bank. Little Willow Creek and Willow Creek initia lly flow into rt side channel, which th en en~ers the east mainstem at RM 52 about 1,000 feet downstream of the Willow Creek confluence. The river gradient reduces substantinlly from 5.6 feet/mile in Se gment 2 to 2.9 feet/mile in Segment 3. This may provide an explanation for the complex morphology of this reach. The lower gradient results in reduced water velocities which could result in less degradation and perhaps some aggradation, causing the channels to meander and intertwine. The east channel conveyed the majority of the flow in 1984. However, this could shift to the west channel if the contr0lling gravel deposits at the upstream end of the Del ta Islands are eroded. The multiple channels of the Delta Islands are fo rced together by terraces just upstream of the Deshka River at RM 42. 4. SEGMENT 4: RIVER MILE 42.5 TO RIVER MILE 27 This reach is similar to Segment 2, with a well defined mainstem and numerou s side channels along both the left and right banks. The Deshka River, at RM 40.6, is th~ only major tributary entering this segment. Kroto Slough represents one of the major side channel complexes in this segment. The upstrean: entrance is located about one-half mile below the confluence of the Deshka River. Although Lhis side channel has several branches which connect with the Susitna m.-dnstem, one channP.l continues on separately to the Yentna River. This side chan.1el sys tem dewaters at flows less than 13,000 cfs (USGS at Sunshine). However, when the mainstem is ice covered , the stage increases enough tc.. flood the channel, so for a major 33RD1-007m -.-M- ~3 portion of the year this side channel flows with Susitna and Deshka River •aters. The gradient through this reach continuc!s to decrease with respect to p r eceding segments. The gradient average of 2 .6 feet /mile is also reflected in the lower surface water velocities. Velocities from 3 to 4 fps have been measured when Sunshine flow is 10,000 cfs. Channel widths range from a maximum of 5,500 feet at RM 32.2 to the narrow section of 800 feet at RM 38.5. The side channels through this reach are strictly overflow channels at high water, are generally dewatered at flows below 13,000 c:s, (USGS at Sunshine), or usually between October and April. At RM 28 the Yentna River joins the Susitna. This is a maj o r tributary draining an area over 6,200 square miles. The Yentna River contributes approximately 40 percent of the annua 1 flow measured at Susitna Station (RM 25.9) by the USGS. However, this is not cons istent at all flow ranges. The proportion may vary greatly depending on storm system movement and the glacier mass wasting characteristics of each system. The Yentna discharge approximates the flow on the Susitna measured at Sunshine during low flow periods but often does not respond sicultaneousl y to the same hydrograph peaks. The average annual flow in 1983 was 18 ,214 cf.. During freezeup the discharge typically drops from about 20,000 cfs in October to under 4,000 cfs by the end of December. 5. SEGMENT 5: RIVER MILE 27 TO RIVER MILE 0 Just downstream of the Yentna River confluence the USGS maintains the last streamflow gage on the Susitna at RM 26 . This gage, located at Susitna Station, measures essentially the total flow of the entire Susitna River waters hed, an area of about 19,400 square miles. The average discharge is 33RD1-007m 49,940 cfs, but typically during freezeup the flows dro p from about 60,000 cfs to 9,000 cfs during freezeup. The river reach downstream of Susitna Station represents an area of transition from a river system to an estuary. A dominating feature of this sep;"'lent is Alexander Slough, al so called the Susitna west channel. This represents a major side channel at most o pen water flows but dewaters just prior to freezeup. When mainstec water enters this side channel the flow essentially becomes isolated and does not re-enter the mainstem except at flood stages. Then an interconnecting channel at RN 9. 7 floods. At low flows, such as just prior to freezeup, the side channels are generally dewatered and the mainstem is confined to one channel, although encompassing many exposed sand bars. The slope through this reach was determined from USGS topographic contours to be about 1.5 feet/mile. Surface velocities average about 2 to 3 fps. Tributaries entering this reach include Alexander Creek and Fish Creek. Alexander Creek enters Alexander Slough and continues out to Cook Inlet without joining the mainstem. Fish Creek drains the swamplands adjacent to, and east of, the Susitna east channel and enters a side-channel at RM 8. As can be expected, the gradient is so low here that flow froc this tributary is greatly restricted by backwater created by mainstem stages. The National Weather Ser•Tice has o perated a weather station in Talkeetna since 1941. The data from this site are fairly representative of the lower river area and provide good bas~line climatic trends for the entire basin. Air temperatures are known to vary considerably between Talkeetna and Cook Inlet, with extremely low winter air temperatures observed between the Delta Islands and Susitna Statio.1. In 1984 R&M Consultants placed a thermograph at 33RD1-007m -k- ~$' RM 48 in the Delta Isla nds in order to quantify the air temperature deviation from Talkeetna . These data are plotted in figure 7. Statistical information presented in th is section have been summarized and are listed in table 3. 33RD1-007m R21/1a TABLE 3 SUS I TNA RIVER "" .... ; \ ... \, \ ~ \v ~ '\-\-. •. 'f"' \:)-\." Basin Annual Arua Ave ra4e Ava i lnl.l le f\uu .... Ga g ed Gradient D 1 scha rge Majo r wea titer Locat,ion b£!!gJj} ~il ( ft,Lmi l (cf s ) T r i bu J.a r i e s __ * Dnta Upper River West ro r k (R&M) Dena I 1 ( RM 29 1) 2 1 9')0 9.3 2,7':;9 [II S l f ork DOnA I I Mn c laren 1960-19811 Mac l aren I< i vur 280 979 Tyone OshP.tna RM 233 ~ 5.9 M iddl e River Goose Jay (R&M) Vee Ca n yon (RM 22 3) 10 4, 140 25.0 6, 4 011 Ko si na Wat iHta Watana 1980 -19811 GOII.I Cruc k (HM 136) I)( 6,160 1~.0 9, 7<!11 Duallman 0CVI I Ca nyon l su s en a 1 98()-1 '.1811 RM 98 ill 13.4 Duv1 I SIIHnnan ~ Portage 1CJ!l 2-1<;84 I nll ian -J Gold Cree k fourth of July Lower River Chu I 1 tna River 2,570 8, 798 .. (NW S) I • , I 2,006 4,05 ':; l~tlkeetna Talkee tn• _!:•v~r \-Chulitna 19111-1984 Sunshine ( RM 81~) lit 1 1,100 5.7 24,000 Talkeetna r rappc r (R&:M) Wi I l o w Cruek 11 6 417 Birch De Ita 1 s 1 and Rabi lluaux 1 ho rmoy roph Des h k<l R I vor ')92 9 41 Mnnt<t na 19811 Goose Yen trw H 1 vor 6, 180 18,2 14 Sheep Ka s llw 1 t na Susitna S t a t1on (RM 26) 72 19,1t00 II. 3 lt<J, 91 10 L 1 t t 1 e W 1 I 1 ow W i I I ow Cook Inlet (RM 0) 2..!! 3.6 Doshka Yl~n tna Alexander Tcta 1 318 IS.3 * Other wc<~tller l>tatio n s opera tell 1n t h e Susitna River b asin by R&M Consu l tants are G 1 ac i cr 1980 -84 , and Kos1na 1980 -81 1, IV. RIVER ICE PROCESSES A. GENERAL FREEZEUP ICE PROCESSES Previous stud i es of ice formation on northern rivers have ranged from qualitative descriptions of events to analytical studies of ice ccver stability (Newbury, 196). Through these studies a consistent sequence of events has emerged by which an ice cover forms on northern rivers. These events have also been documented on the Susitna River and can be summarized as follows. Generally in October, a continuous trend of cold air temperatures causes the river water temperatures to drop to the freezing point, and frazil ice forms (figure 9). Slush floes appear on the water surface and collect in quiet water, eddies and along the shore, freezing into ice sheets. Anchor ice accumulates on the channel bottom in turbulent shallow reaches. The river current carries the majority of the slush downriver where, on entering the backwater of Cook Inlet, it jams at a cons triction to form an ice bridge. From tnis point the slush packs into an ice cover which accumulates upstream at a rate dependent on the intensity of cold weather, river morphology and the hydrodynamics. Growth of the ice cover continues until the upper and middle river is frozen over and frazil ice is no longer generated. Remaining open water is gradually closed as i c e grows laterally from the banks. 1. FRAZIL ICE The main process of ice development on turbulent northern rivers is the formation and accumulation of fra?.i l ice. Fra?.il ice fot'll's on open water surfaces when air temperatures are below freezing. Frazil ice is microscopic ice crystals which fo r ms wh en the river water becomes supercooled. Supercooling occurs when water loses heat to the atmosphere, producing a 33RD1-007m -..)h-- 2¥ temperature below the free.zi.ng point (i.e. le ss than 0 C). Supercooling of river water normally does not go below -0.05 C (Michel, 1971 ). The supercooling of water stops when the heat of fusion of ice formation produces an increase in water temperature. In supercooled water, newly formed frazil ice c rystals are very "active" and either agglomerate to form larger grains or adhere to underwater objects. This stage of development may last only a few minutes. Frazil ice takes the shape of flat, circular plates. The frazil ice becomes inac tive once the river water returns to 0 C from the supercooled state. The inactive frazil ice grains float to the surface and continue to grow from atmospheric heat e xchange. Frazil ice crystals are continuously changing form. From the initial miniscule crystaline discs t hey grow rapidly into larger grains (figure 10). The nucleation of frazil ~s known to be associated with foreign particles such as fine sedioents (Osterkamp, 19 78). The crystals form around these nuclei, and while the water remains supercooled they grow rapidly in size or stick to other crystals to form ice floes. The ice floes have densities nearl y equal to water and remain entrained even at low water velocities. If supercooling continues the frazil floes grow and eventually gain sufficient mass to counteract the turbulence and float to the surface. The frazil slush floes drifting on the surface are agglomera tions of highly porous, poorly bonded ice grains that can easily break apart and become re-entrained (figures 11 and 12). The slush is therefore constan tly broken up and submerged by turbulence enroute downstream. On the water surface and in close contact to cold air, the slush ice grains grow with th e water in the slush pores crystalizing onto the grain surface . Ice partic lt:! growth tends towards a spherical shape, suggesting that the flat surfaces grow more ice than at edges or ends. The p oros ity of the slush decreases as the grain size increases. Measurements of 33RD1-007m porosity on the Susitna River indicate decreasing values in the downs tream direction, which correlates well with the increasing age or residence time of the slush in the downstream direction. (New frazil ice is formed in the colder and more turbulent upstream reaches, then floats downstream.) When flow velocity decreases so that frazil slush remains on the surface, a continuous layer of solid ice forms on the top of the slush floe. Under these conditions the ice grains are no longer being broken apart and the water held in the interstices simply freezes solid, binding the ice particles together in a solid sheet. The majority of frazil ice is generated at night when air temperatures are usually coldest, and generation is reduced or stopped with sunlight when heat added by solar radiation stops the cooling of the water (Hichel, 1971). The variable climatic conditions throughout the Susitna River basin significantly effect the net volume of fra z il slush present in the river and, consequently, the rate of ice cover development. The dominating meteorological parameters governing ice formation are air temperature and solar radiation. 2. SHORE ICE Shore, also called border ice, ice is the first type of ic~ cover to form over non-turbulent water along a river bank. On many northern rivers an area of laminar flow exists along the banks . Michel (1971) states: because in laminar flow there is no intermixing of the top layer with the botto~ layers, ••• considerable cooling will occur in the top laye r while the average water temperatu re i s still far above the freezing point. Ice 33RD1-007m ~- Jo is nucleated on the ·water surface starting in contact with the cold material along the river bank .•. forming a clear and solid ice sheet. Michel also recognizes this as a primary ice cover process, second only to frazil ice accumulation. Shore ice growth on the Susitna River has not been documented to form exactly the way Michel describes. This is probably due to the high velocity and turbulence of this river. An absolute laminar flow area along the banks is rare on the Susitna and the rapid cooling and subsequent clear ice formation has only occasionally been observed. Shore ice does develop, however, often to extensive dimensions . Flow margins, while not laminar, are relatively slow-moving and water temperatures are close t o those at mid-channel. Because the turbulence is less, cooling of the surface does occur, but even a slight amount of mixing prevents the formation of ice on the surface. Sho re ice on the Susitna begins forming soon after frazil ice is fir s t formed, and continues growing t owards the channel center until the rate of growth equals the rate of erosion by water veloci ty (figure 13). Shore ice development begins when frazil slush drifts into the low velocity flow margins along the river banks. Friction against the channel bottom stops the slush, which freezes to the bank. Slush continues accumulating along the flo~ margins, extending the shore ice out into the channel until flow velocity is high enough to keep the slush moving, preventing it from freezing in place, At this point a shear zone develops where moving slush , carried by the flow, slides along the edge of the fixed shore ice. This usually occurs when the velocity at the shore ice edge exceeds 1. 0 fps. The face of the shore ice edge is very rough, consisting of ice grains frozen in place, and the friction 33RD1-007m -~- 11 on the moving slush ice is enough to slow the v elocity of the floes. The slush ice flo es are then be ing forced downstream by the water velocit y along their out side edge while also being f o rced t o a s tandstill along the inside edge b y friction against the shore ice. The slush floes then tend to break apart, with th e ice grains either being fo rced underneath the shore ice o r deposit e d along the top edge in a n a rrow layer, Successive layers are added in this manner, so that the shore ice grows laterally out into the channel and vertically down t o the channel bottom by trapping drifting slush ice. This process is known as "buttering" (Lavender, 1981). Low velocity areas subject to this t ype of frazil slush accumulation include: the inside bank of river meanders; shallow flow margins; eddies downstream of any flow obstruction such as rocks, logs and rapid s; backwater areas above flow cons trictions; and eddies immediately downstream of a tributary confluence (figure 14 ). In some reaches the conditions are favorable for extensive shore ice growth from adjacent banks, so that the surface of the flowing water is constricted. In these cases floating slush ice often bridges the narrow gap and free z es in place (see Ice Bridges). Heavy snowfalls during cold weather often initiate shore ice growth by rapidly covering low-veloc ity flow margins . If the turbulence is low enough to not entrain the snow laye r, then water infiltrated into the s now qui c kly freezes into a rigid snow-ice cover. Once in place this ice cover continues growing vertica lly by building up add itional layers of black ice. This process has been documented on some shallow areas of the Susitna where velocities are nea r zero, including backwater zones, sloughs, beaver ponds and lakes. The growth of s hore i ce in c reases the we tted perimeter of the channel while decreasing the c ross-sectional area of the channel. 33RD1-007m -Jtr- Jz, This causes an increase in water level. ·The rising water level often fractures the shore ice, sometimes simply hinging the ice shelf and separating it from the bank. This creates a narrow lead of open water between the bank and the shore ice shelf. The rate of shore ice growth seems to depend on the channel depth. Shore ice grows quickly over shallow flow margins and slowly in deeper water with high velocities. The gradual decrease in river discharge, therefore, plays an important role in controlling the extent of shore icP growth. If discharge remains high or constant (i.e., the same water lev el) tht. shore ice growth is less. 3. ICE BRIDGES Ice bridges usually initiate an upstream ice cover progression by frazil slush accumulation. This eventually leads to a continuous ice cover on most northern rivers. On the Susitna River, several formation processes have been documented, all of which may occur during the course of one freezeup period. The most important ice bridges to form on the Susitna are those on the lower river, which usually form in the final 10-12 miles upstream of Cook Inlet. In this reach, an ice bridge forms primarily due t o low flow velocities induced by a high tide event, allowing t!-!e slush floes to accumulate and form a continuous cover from bank to bank that freezes in place. Ice bridges may also form at shore ice constrictions or at shallow riffles where slush ice floes become grounded while having sufficient cohesion to resist breaking apart. Of less significance are ice bridges which form late during the freezeup when water levels rise due to anchor ice (see Ancho r l ee) or shore ice growth. The rising water lifts the shore ice away from the river banks. The shore ice fractures into large blocks which are c a ught by the current and 33RD1-007m -y- ]3 drift downstream until they wedge firmly between a channel constriction or become grounded along the banks. Any type 0f ice bridge acts as a surface obstruction to slush floes drifting downs tream . These floes either contact the ice bridges and submerge, become entrained in the flow and re-emerge further dowr.~tream, or accumulate ag~i~st the upstream z dge of the ob - struction. The tide fluctuations in Cook Inlet create a backwater that influe nces water velocities in the final 15-mile reach of the Susitna ~iver. The tides often fluctuate over 30 feet above the Anchorage reference datum. This datum is 16.4 feet below the :!.ocal mean sea level. \.fuen a high tide of 34 feet occurs at Anchorage the high water line is approximately 17.6 feet ar ove me an sea level. Water velocities a re visibly reduced iP the 10-mile reach of the Susitna above Cook Inlet . During the latter half of October, when slush ice floes are drifting downriver and a high tide occurs in Cook Inlet, the floe s tend to concentrate in the backwater zone, This accumula tion occ~rs rapidly since the floes are not conveyed through the reach at the same rate as t hey en ter. Tte accumulations often attain extensive proportions resembling a continuous ice cov er but moving at a slow v eloc ity of ab o ut 0.5-l.O fps. When the tide begins to recede the water level drops and flow v elocity increases . However, as the surface area of the river decreases it can no longer transpo rt the massive volumes of accumulated ice, resulting in a jam . The i c e jam gains stability as the water level c ontinues dropping a nd more ice floes become grounded. This ice jam prevents incoming ice fl oes from pas s ing out t o s ea. At low c oncentrations of ice f loes, a bridge does not develop and the ice accumulation is flushed out t o sea . Ice bridges hav e been observed to f orm at RM 5 and RM 9 during the 4 years (1981 -19 84) of ice study. The factors 33RD1-007m -~ J<t which appear to coincide be"fore on ice bridge forms in this rea c.h are cold air temperatures, large volumes of slush ice and a high tide event. Subsequent to the formation of an ice bridge below RM 10 and the upstream progres sion of an ice cover, other ice bridges may form above the advancing ice front. The flow velocity is generally reduced in the backwater a rea upstream of the leading edge. When the slope gradient is low, the flow velocity can be reduced to less than 1 fps for a distance of at least o ~e mile abov e the ice front. During a cold weather event and with high slush ice dis- charges, the slush may jam at a channel constriction or river me~nder upstream of the leading edg~ but within the backwater zone, leaving an open water reach downstream betwe en the new ice bridge and the old leading edge. Ice cover progres s ion resumes at the new ice bridge. Some slush may break free from the underside of the upstream ice cover, emerge in the open water below the ice bridge, and accumulate along the up s tream edge of the old ice front in a thin sheet. This is shown in Figure 14, along with an example of the pr ~viously described ice bridges . From 1981 to 1984 river ice bridges have been observed to form at the following river miles: 5 9 10 12 14 16.5 25.9 30 46.1 (West Channel, Delta Islands) 52.1 At low river discharges (i.e ., les s than 5000 cfs at Gold Creek), several reaches of the middle river between Talkeetna and Gold Creek have a channel configuration that allows wide shore ice development. 33RD1-007m -~ J:r' The shore ice constricts the open water surface area to a narrow lead which becomes plugged by ice floes during a high slush ice discharge. This ice jam freezes in place, and additional ice floes drifting downstream begin accumulating along the upstream edge of the ice bridge. lee bridges of this type have been documented to form at the following river miles : 97.4 (lower river) 120.5 98.8 128.5 105.1 135.5 Ice bridge formation is of paramount importance to further development of an ice cover. How or where the bridge forms is not critical, but an obstruction must develop in order to stop the flow of slush out to sea. Once an ice bridge forms the frazil slush rapidly accumulates along the upstre~m edge by various processes which are dependent on the physical characteristics of the ice and the hydrodynamics of the river. Trese processes are discussed in the following sections. 4. ICE COVER PROGRESSION After the formation of an ice bridge, the upstream progression of an ice cover is controlled by air temperature at the leading edge, the volume of incoming ice discharge, the hydrodynamics of the river flow and the physical properties of the incoming slush ice. Published literature recognizes four conditions for the progression of the leading edge (Calkins, 1983). The following diagrams show the four basic processes of ice cover progression, Vp is ice cover progression rate and V is water velocity, t is ice thickness, and H is water depth. 33RD1-007m 1. Progression by simple juxtaposition of the arriving floes with no subsequent thickening of the ice cover. Ice cover thickness equals initial slush floe thickness. H ---v t H < 0.33 / 2. Slush floes arriving at the leading edge are compressed to a greater thickness than the original ice floe thickness. This is termed hydraulic thickening and can occur in various degrees when combined with the mechanical thickening process of shoving. 33RD1-007m l < 'll H' _o.3.., H -)6- J7 ---V 3. Arriving slush ice floes are compressed and added to the cover but some also submerge and break apart, eventually being deposited underneath the ice cover further downstream if lower velocities occur. v, .. c , -- -v 4. Arriving slush floes do not accumulate at the ice front but are subducted beneath the cover and may be deposited some distance downstream. - 7YY' The upstream progression of the ice front by juxtaposition leads to a rapid ice cover development. On the Susitna River this is the predominant process of progression downstream of RM 25. The slush ice floes that drift 33RD1-007m -)6- lP through this reach have been on the water surface long enough to have formed a solid surface layer. This significantly strengthens the floes so that they resist crushing or breaking apart. Virtually all the floes therefore remain on the surface and c c me to a stop against the leading edge. The floes pack together and spread across the channel from bank to bank. They accumulate so rapidly during a high ice discharge that voids of open water may remain (figure 15). The shearing stresses of flowing water under the ice cover are no! strong enough to compact the ice floes into a tight configuration. Hydraulic thickening is the primary process of ice cover advance from near RM 25 upstream to the reach near Sherman (RM 130). Even as far downstream as RM 25, the ice floes have not remained continuously on the surface long enough to form a solid laye-:-. The floes therefore lack coi.~sive strength and deform readily when contacting the leading edge. Hydraulic thickening occurs in various forms, depending primarily upon the water velocity and the cohesive properties of the slush. Generally, with colder air temperatures the cohesion increases. Less ice cover thickening occurs in areas of slower water velocities. Arriving ice floes are plastered along the leading edge to a thickness roughly one-third the water depth in areas where flow velocity is low. Where the river is steeper, the velocity higher, and the cohesion of the slush relatively low, thickening continues by a shoving or compression of the cover. The compression of the cover occurs repeatedly, creating higher upstream water levels and lower velocities until progression can resume. This process is known as staging. Water levels can rise dramatically in a short period of time to many feet above the initial level. The remaining two processes of undercover deposition are difficult to document, but probably occur to some extent on the Susitna River. However, juxtaposition and hydraulic thickening seem to be the predominant processes. 33RD1-007m -~- 31 Ai r temperatures influence the st<1bili t y of an accumulating ice cov er, thus affecting the freq u ency and extent of ice shov e s and mechanical thicken ing. Rapid upstream progress i ons during cold weather periods are due to !arg e volumes of ice being generated ups tream and t o the fast stabilization of th€ ice cover as the slush rapidly free z es solid, resulting in fewer and less e x tensive shoves. Colder air temperatures result in faster sta bilization of the ice cover. The equation for estimating solidification of the slush cover i s : 1 h (t) ~ k z~ p where: h(t) = ice thickness at time t in inches k empirical coefficient based on snow cover z = accumula ted freezing-degree days (F) since slush floes stopped down s tream drift p = porosity of slush For example, if the mean daily temperature is 20 F, then in one day the freezing degree-days accumulated equal 12. If no snow lies on the ice then k is about 0.6. With a slush porosity of 0 .35, the ice cover on the main s tem would have a solid layer about 1.2 inches thick after 1 hour, about 3.0 inches thick after 6 hours, and 6 inches thick in 24 hours. At 0 F the rate of stabilization would increase so that a solid layer 2.0 inches thick would form i n 1 hour and about 5 .0 inches thick would form in 6 hours . Slush ice is relatively weak when compared to black ice, due to the number of bonding surfaces between the small ice grains of slush ice. Black 33RD1-007m ice, which grows at a slower rate, develops large crystals with fewer boundary surfaces resulting in a substantially stronger material. During warmer temperatures, a solid layer may not form in the slush floes. The compressions then occur concurrently with the arrival of ice floes at the leading edge. This is more difficult to distinguish, as several processes occur simultaneously. The slush ice accumulates in the backwater and rapidly fills the wetted surface area of the open water. The slush floes become grounded along the flow margins, with a shear zone developing where moving slush slides past grounded or fixed shore ice. Slush sliding along the shear zone deposits layers of ice grains along the shore ice, increasing the width of the latter. Additional slush is pushed under the shore ice and fills in any space between the channel bottom and the underside of the shore ice. The slush continues building rapidly onto the shore ice until it has extended out into the channel to some distance where the velocity of the flowing slush prevents furt-her deposition. The shore ice is now in equilibrium for the prevailing flow and temperature conditions, neither building nor eroding. If either the water velocity or air temperature decreases, then the shore ice would begin building out again. The velocity of slush ice is decreased by friction when contacting the shore ice. This decrease in velocity is transferred to other ice floes further out in the channel. There is generally an increase in velocity from the flow margin to the channel center. Slush therefore tends to accumulate first at the channel perimeter and proceeds towards the channel center, creating a V-shaped leading edge . The fastest flow velocities at center channel are in the notch of the V. Along the sides of the V, the slush floes are compressed and are no longer distinguishable in the resulting ice cover. If air temperatures remain warm enough to prevent the rapid freezing of the 33RDl-007m -.)t- tl surface layer, then the ·accumulating, unconsolidated ice mass constantly compresses due to the mass of upstream ice and flow friction acting on the ice cover. The momentum of the moving ice is sufficient to compress the cover to thicknes.1es greater than that necessary 1.or upstream progression. For this reason, ice thickness measurements often detect 10 to 15 feet of slush beneath a solid ice layer of several feet, although the channel gradient or water velocity are not high enough to justify that much resultant stagin~. Compression or shoving occurs only in reaches of high water velocity (generally greater than 4 fps), but this also depends on the degree of solidification of the slush. Unsolidified slush compresses at lower veloc ities. The massive bank-to-bank compressions of the entire ice cover usually only occur where water velocities exceed 4 fps. On the lower river and in some reaches of the middle river, the com- pressions occur sporadically. The final ice cover shows evidence of compression zones followed by a reach where the floes juxtaposed. At the end of the juxtaposed section, another compression occurs (figure 16). This pattern extends from Cook Inlet to Talkeetna. In constrast, much of the middle river is entirely compressed, with the ice cover shoved laterally until it contacts the surrounding terrace. During a compression the forces within the cover are transmit ted to the banks. If the bank slope is low, ice is pushed laterally up the bank well beyond the water level. If the banks are steep, the ice is contained without the lateral spreading. The latter condition is more prevalent on the middle river, with the former more typical of the lower river. The rapid development of open water leads in the ice cover is evidence of instability between the stable ice and flowing water. The widespread occurrence of the leads suggests that an ice cover, particularly on the middle 33RD1-007m river, is either eroding -or building, but is rarely stable. The rate of erosion is dependent on the air temperature and the water velocity. Open leads can form within 1-3 miles below :he ice front, with exceptions noted where leads have developed within a few hundred feet below the leading edge. During e x tremely cold weather (i.e ., below-15 C) the channel leads begin to form a secondary ice cover by a gradual accumulation of slush at the downstream end. This secondary ice cover progresses upstream until completely covering the open water. If the cold weather ends before a complete ice cover reforms, open water may remain all winter. Erosion of the ice cover is evident by a decre3 se in water level and a sagging of the ice cover by many feet. This is due partially to a steady but gradual decrease in discharge. However, the primary cause is the erosion or removal of unsolidified slush ice under the solid layer. The entire ice cover between banks settles until the slush contacts the channel bottom along the low flow margins. 5. ANCH'JR ICE In supercooled water, frazil ice crystals hav e a propensity for adhering to any object in contact with the river flow. When frazil ice adheres to rocks on the channel bottom, it is commonly referred to as anchor ice. Anchor ice generally forms in shallows immediately downstream of a turbulent reach. The turbulence is required to supercool the entire water column down to the channel bottom so that frazil ice forms and is maintained in the active state until it adheres to the substrate. Deeper reaches are generally free of anchor ice because they are less turbulent and the supercooled condition does not effect the entire water column. Anchor ice can accumulate in thick layers in and below shallow rapids where depths do not exceed 4 feet. The greatest accumulations therefore occur in the rapids themselves, often altering the 33RD1-007m -~ YJ flow regime by either effectively raising the bed surface and increasing the water level, or by restricting the flow and creating a backwater area. Anchor ice "dams" have been documented between Sherman (RM 130 ) and Portage Creek (RM 149). The thickness of the anchor ice dams can exceed 2 feet. They tend to locally lncreas e water velocity by restricting the cross sectional area, creating turbulence which helps to maintain the supercooled condition of the flow. When supercooling ceases, due to increased air temperature or the formation of an ice cover, anchor ice tends to release from the bottom and float, often taking along material from the channel bottom. Anchor ice is usually readily visible, as it takes on a dark tint (generally appears brown but this varies with depth) from accumulating small sediment particles . The sediRent probably was saltating along the bottom until contacting the rough and highly porous surface of the anchor ice. The sediment is trapped between the crystals and is eventually covered by more frazil ice layers. On clear days, solar radiation or warm air temperatures can raise the substrate temperature sufficiently so that anchor ice breaks away and floats to the surface, oftea carrying with it accumulations of sediment. The anchor ice floes drift downstream, eventually becoming incorporated in a downstream ice cover. Studies of anchor ice have shown it to be an important ecological factor relative to lentic macrofauna and fishes (Needham and Jones, 1959). Anchor ice forms only in areas which have no upwelling of relatively warm groundwater through the substrate. When anchor ice releases from the channel bottom it generally dislodges bed material, causing bottom macrofauna to become entrained and made available as food to resident fish species. B. SUSITNA RIVER ICE COVER DEVELOPMENT 33RD1-007m -K- tf This section discusses jce cover formation on the Susitna River from the mouth at Cook Inlet to the middle river reach near Vee Canyon . 1. COOK INLET TO THE CHULITNA RIVER CONFLUENCE The initiation of ice cover developcent on the Susitna River usually occurs when large volumes of slush ice fail to pass through a channel constriction near the river mouth at Cook Inlet. The meander at RM 9 forces the ice floes to contact the outside (west) bank. At a high tide the resulting backwater furthP.r reduces the water velocity. With high ice concen- trations and cold air temperatures, bridging is likely to occur. Cold air temperatures are necessary to quickly freeze the ice in place. Upstream ice cover progression by accumulating ice floes can begin as soon as the slush ice velocity slows. The higher upstream velocity of incoming slush causes a greater volumt! of slush to accumulate against the upstream edge than can be expelled from the downstream end. Therefore, with a low channel gradient and slow water velocity the ice cover "advances" upstream by juxtaposing (Figure 15). A fixed ice cover imparts a frictional resistance to flowing water, causing an increase in water level. The increase in water level, called staging, is required to slow water velocities to such a point that ice floes are not swept beneath the leading edge of the ice cover. The maximum staging level observed below RM 26 is about 2-3 feet, with ice thickness averaging 3 feet. This ice thickness refers to the total of solid surface ice (frozen slush) and the underlying loose slush. Air temperature controls the thickness of the solid ice fraction simply by continually freezing additional slush ice. If the underlying slush ice is removed by erosion, then growth of the solid surface ice layer slows significantly. 33RD1-007m Prior to the initi~~ion of freezeup, Alexander Slough is usually dewatered when decreasing mainstem flows drop below th,, critica l level to overtop the entrance. The developing ice cover stages suft ~ciently to flood the channel. Water usually inundates the snow cover in the side channel and along the flow margins. This quickly freezes solid, producing a shorefast ice cover. The flowing open water in Alexander Slough often requires more than four additional weeks to freeze, primarily because the stage does not increase enough to allow the passage of slush ice. The slush ice rafts are usually 2-3 feet thick. Unless the stage increases by that value above the threshold elevation of the channel entrance, the floes can not drift into the side channel. The depth of water over the channel entrance at Alexander Slough has been observed at 1 foot, so the ice rafts become grounded a short distance from the main chailnel. No slush ice cover progression has therefore been observed in Alexander Slough. Closure is achieved by border ice growth. Higher water velocities near RM 26 prevent ice cover progression by simple juxtapositiorl . and mechanical thickening of the cover begins. This process of thickening occurs after the slush ice cover is in place. The frictiona l shear between high velocity water and the fixed ice creates an unstable condition, which can cause a portion of the ice cover to shift. This sudden movement upsets the stability of adjacent ice and in seconds the entire local cover is moving downstream and consolidating. A chain reaction of this type has been observed to affect over 2,000 feet of ice cover. Compression of unconsolidated slush ice during this move causes the total thickness to increase. The ice c vv er may also be shoved laterally ~nto the ·Janks often above the water line. Several ice compression phases have been timed to last more than 8 minutes, which brought the leading edge downstream about one-half mile and increased the s tage about two feet. 33RD1-007m Aerial observations noted that the Yentna River often contributes about 50-60% of the total estimated icc volume below the Susitna confluence. When the ice front reaches the confluence it separates and continues up the Susitna River, while another leading edge goes up the Yentna River (figure 16). The progression rate on the Yentna River is faster due to its morphological characteristics and high ice discharge. The Yentna is generally narrower, shallower and has fewer channels compared to the Susitna. The s l ower water velocities also permits less ice thickening, therefore requiring less ice volume to develop a stable cover. A stage increase of 3.9 feet has been measured at the entrance to Kroto Slough (RM 40.1). This is sufficient to overtop the slough with a flow depth of 1. 5 feet at the entrance, but few ice floes can enter due to a t y pical thicknesses of about 2 feet. The elevated mainstem stag·· also effects the Deshka River by creating a backwater zone which extends about 2 miles upstream. The surface water velocities on the Deshka are reduced enough to allow Susitna ice floes to be pushed up the Deshka for about 100 feet. Slush ice drifting down the Deshka River encounters this barrier to flow and an upstream advance by accumulating ice is initiated. The Deshka has low water velocities and the slush ice advances by juxtaposition, eventually freezing into a solid ice cover. When the ice cover progression enters the Delta Islands the leading edge splits. Ice fronts advance separately up the east and west channels. The east channel ice cover progresses more slowly. The advancing ice cover generally causes stage increases high enough to inundate the snow cover over the Willow Creek gravel fan. This saturated snow eventually freezes into an ice cover. However, the water course from Willow Creek is not u s ually altered. The stage through this area increases about 3 feet during the ice 33RD1-007m -}( - Y7 front advance. Slush ice . from the Susitna does not encroach on the creek confluence. In 1983, the stage increase measured on the west channel of the Delta Islands, near RM 48, was about 2.5 feet. This channel was flooded but no slush ice entered. The Susitna ice cover progresses through the Delta Islands and converges near RM 51, then continues to proceed upstream. The reach above the Delta Islands contains more secondary channels within a broad gravel and sand floodplain. The primary or main channel is relatively shallow at freezeup and when the water level rises a wide area is generally flooded. The ice floes remain contained within the main channel, since water depth is not sufficient t" float them laterally out of the thalweg. As the ice cover proceed e d through this reach in 1983, a large portion of the flood plain was inundat ed. The saturated snow eventually froze solid, creating an ice cover but without the hummocked appearance of the main channel slush ice cover. Many of the lower river side channel complexes are flooded during ice cover progression. vfuen the staged mainstem overtops the channel entrances, existing ice over isolated pools in side channels is immediately broken up and washed downstream. Mainstem slush ice often accompanies the :>urge through the side channel at RM 60. The slush ice and ice debris occasionally accumulates in small jams a short distance below the side cha nnel entrance, but is usually carried out to the mainstem. A mainstem stage increase of about 3 feet occurs near the mouth of Kashwitna River (RM 60) (figure 17). The effects of mainstem staging are not evident to a significant degree at the mouth of Sheep Creek. Sheep Creek enters a side channel that exte nc~ from RM 62 to RM 67. Through this reach the mainstcm is along the west bank. Since the side channel complex is on the east bank, it is not usually affected by backwater or overtopping during ice cover progression. 33RD1-007m Goose Creek enters a -side channel that runs from RM 69 to RM 72. This side channel was flooded in 1984 but not in 1983 (figure 18). The mouth of Montana Creek is significantly influenced by the staging process. The existing channel mouth steadily degrades when the mainstem water level recedes. T~e absence of an extensive backwater area results in higher tribut:ary velocities at the mouth and subsequently more downcutting at low mainstem flows. Montana Creek can therefore become entrenched in the alluvial fan. Heavy anchor ice deposits usually accumulate on the substrate, and a large ice dam has been observed to develop about 200 feet above the confluence. When the ice front approaches RM 73, 2 miles downstream of the tributary confluence, the mainstem stage adjc.cent to Z.tontana Creek increases by about 1 foot and creates a backwater zone that floods the tributary channel and ice dam. A maximum stage increase of 7.1 feet was measured on November 18, 1983, and most of the confluence area was inundated. The snow cover over much of the alluvial fan was flooded and subsequently formed ice. An additional 2 feet of staging would have been required to completely overtop the alluvial fan. Ice thicknesses measured adjacent to the Hontana Creek confluence (RM 77) in late January, 1984 averaged 6.8 feet, with a minimum of 1.3 feet and a maximum of 7.0 feet. The channel gradient is relatively steep in this a~ea, with the ice cover usually remaining unstable. After the initial progression through this reach, an open lead appears from about RM 71 to RM 85. This lead eventually freezes over again when entrained frazil ice floats and accumulates at the lower end and along the sides. This secondary progression may not completely close the lead. In March 1984, open water remained from RM 81 to RM 85. 33RD1-007m Sunshine Slough and side channels are usually ov ertopped and flooded. Slush ice has not been observLd entering this system due to an insuf f icient overtopping depth at the entrance. These channels subsequently require an additional 8-12 weeks to freeze over, with many leads existing all winter. The side chann ... ls leading to the entrance of Birch Creek Slough are flooded but the stage does not increase enough to overtop the slough entrance. The maximl'.m observed increase was 3. 1 feet in 1983, near the upstream entrance to Birch Creek Slough. An additional foot would have been necessary for over- topping. Trapper Creek is not affected by Susitna mainstem freezeup. At prefreezeup stages Trapper Creek does not merge with the Susitna until RM 90. No slush ice floes drifted up into the creek mouth, and flow remains unrestricted by ice. With the except ion of some backwater, Birch Creek and Sunshine Creek are also unaffected by the ice advance. The flow in Rabideux Creek is low during freezeup (discharge estimated less than 10 cfs) and the staging has been measured to reached 7 feet over the open water level. Most of the side channels below Talkeetna are flooded to sone extent, often only saturating the snow cover. Several side channels, such as Sunshine Slough and Kroto Slough, remain flooded all winter. The maximum staging levels seem to be temporary, and water levels along the entire lower river recede once the leading edge has moved upstream several miles. This may be due to ice cover erosion and the development of leads, or seepage of water into the adjacent banks. A reduction in mainstem stage may cause the ice cover to sag and eventually collapse. A thinning of the ice cover by erosion has also been measured over high velocity cells along a cross section. Ice thickness measurements along the banks usually reveal thicknesses representative of the 33RD1-007m original ice cover at the time of progression. Thin covers have been located over fast flowing water, either at mid-channel or along either bank. The thin ice covers are indicative of areas where water velocity (friction) is high enough to erode the underside of the ice cover. Table 4 lists the major open leads documented by aerial photography in 1983 on the lower and middle river. Figure 19 illustrates the general fr~ezeup sequence for the lower Susitna River. Water level fluctuations due to staging at sev eral lower river sites have been plotted and are shown in Ligure 20. thicknesses. See table 5 for ice The following sequence suD'Illlarizes the highlights and general freezeup characteristics of the lower river from Cook Inlet to the Chulitna River confluence. 1. An ice bridge forms at a channel constriction near the mouth of the Susitna during a high tide and high slush ice discharge. 2. A rapid upstream advance of an ice cover by slush accumulation. 3. Thin, unconsolidated initial ice cover forms. 4. Minimal staging (2-4 feet) occurs up to Sunshine, with over 4 feet occurring near Talkeetna. 5. Little telescoping or spreading out of the ice cover occurs due to shoving. Ice cover generally is confined to the thalweg channel. 6. Tributaries generally continue flowing through the winter. 33RD1-007m -H- J{ Table 4 Major Annually Recurring Open Leads Between Sunshine RM 83 and Devil Canyon RM 151 Locations and Dim*sions on Harch 2, 1983 Location of !I£!. Continuous UEsteam End Channel of AEErox. Widest or River Mile If T~Ee Lead (1) Len~th (Ft) Point ~Ft) Discontinuous 85.0 Mainstem Velocity 550 80 Continuous 87.1 Slough Velocity 4,500 50 Discontinuous 87.6 Mainstem Velocity 700 100 Continuous 89.0 Hainstem Velocity 1,200 100 Continuous Side Channel Velocity 2,500 40 Continuous 89.5 Mainstem Velocity 1,400 60 Discontinuous 91.0 Mainstem Velocity 1,700 80 Discontinuous 92.3 Mcdnstem Velocity 1,300 110 Discontinuous 93.7 Mainstem Velocity 3,500 110 Continuous 94.0 Mainstem Velocity 3,500 20 Discontinuous 95.2 Side Channel Velocity 2,400 100 Continuous 96.9 Side Channel Velocity 5,600 lJO Discontinuous 97.0 Mains tern Velocity 1,100 30 Continuous 102.0 Mainstem Velocity 2,400 100 Discontinuous 102.9 Mains tern Velocity 600 100 Continuous 103.5 Mains tern Velocity 1,850 100 Discontinuous 104.1 Mainstem Velocity 280 70 Continuous 104.5 Mainst.em Velocity 1,700 110 Continuous 104.9 Mainstem Velocity 900 150 Continuous 105.9 Mainstem Velocity 1,050 100 Continuous 106.1 Mainstem Velocity 200 60 Continuous 106.4 Mainstem Velocity 370 50 Continuous 106.6 Mains tern Velocity 350 50 Discontinuous 107.4 Mainstem Velocity 200 50 Continuous 109.1 Mainstem Velocity 550 100 Discontinuous 110.3 Mainstem Velocity 150 100 )iscontinuous 110.5 Mainstem Velocity 290 50 Continuous 110.9 Mainstem Velocity 450 50 Discontinuous 111.5 Mainstem Velocity 1,600 100 Continuous 111.7 Mainstem Velocity 500 90 Continuous 111.9 Mainstem Velocity 900 150 Continuous 112.5 Hainstem Velocity 700 100 Discontinuous 112.9 Mains tern Velocity 500 110 Continuous 113.8 Mainstem Velocity 600 110 Continuous 117.4 Mainstem Thermal 780 60 Continuous 117.9 Side Channel Thermal 1,260 120 Discontinuous 119.6 Side Channel Thermal 550 50 Continuous 119.7 Mainstem Velocity 350 50 Continuous 33RD1-007m TABLE 4 Location of ~ Ufsteam End Channel of River Mile # T;a~e Le~(l) 120.3 Mainstem Velocity 121.1 Mainstem Velocity 121.8 Side Channel Thermal 122.4 Slough (7) Thermal 122.5 Slough (7) Thermal 122.9 Slough (7) Thermal 123.1 Hainstem Velocity 123.9 Side Channel Thermal 124.4 Side Channel Velocity 124.9 Mainstem Thermal 125.3 Slough (8) Thermal 125.5 Hainstem Velocity 125 .5 Slough (8) Thermal 125.6 Hainstem Velocity 125.9 Slough (8) Thermal 126.1 Slough (8) Thermal 126.3 Slough (8) Thermal 126.8 Slough (8) Thermal 127.2 Side Channel Thermal 127.5 Hainstem Velocity 128.9 Slough (9) Thermal 128 .5 Side Channel Thermal 128.8 Side Channel Thermal 129.2 Slough Thermal 130.0 Hainstem Velocity 130.8 Side Channel Thermal 130 .7 Hainstem Velocity 131.1 Mainstem Velocity 131.3 Mainstem Velocity 131.5 Side Channel Thermal 131.3 Side Channel Thermal 132.0 Mainstem Velocity 132.1 Mainstem Velocity 132.3 Hainstem Velocity 132.6 Mainstem Velocity 133.7 Slough Thermal 133 .7 Mainstem Velocity 134.3 Slough (10) Thermal 134.0 Side Channel Thermal 134.5 Side Channel Thermal 135.2 Hainstem Velocity 33RD1-007m (Continued) ApfrO_~. Lensth (Ft) 800 550 1,450 1,850 380 1,950 1,000 200 270 600 3,50() 2,140 800 350 580 500 250 1,500 2,450 700 5,060 1,210 380 4,000 600 5,000 150 490 800 5,000 900 150 500 400 1,350 6,000 1,110 4,500 -~ .s-] 1,200 850 1,580 Continuous Wi d est or Point (Ft) Discontinuous 100 Continuous 100 Continuous 30 Discontinuous 60 Discontinuous 50 Continuous 80 Discontinuous 80 Continuous 50 Continuous 40 Continuous 90 Continuous 50 Discontinuous 100 Continuous 500 Continuous 60 Continuous 50 Continuous 30 Continuous 50 Continuous 80 Discontinuous 50 Continuous 80 Continuous 100 Continuous 30 Discontinuous 20 Continuous 30 Discontinuous 90 Continuous 50 Discontinuous 50 Continuous 90 Continuous 100 Continuous 80 Discontinuous 90 Discontinuous 20 Continuous 20 Discontinuous 80 Continuous 80 Continuous 60 Continuous 100 Continuous 40 Continuous 50 Continuous 100 Continuous 90 Discontinuous TABLE 4 Location of ~ U~steam End Channel of River Mile I Tl:~e Lead(1 ) 135.7 Slough (11) Thermal 136.0 Mainstem Velocity 136.3 Side Channel Thermal 136.7 Mainstem Thermal 137.1 Mainstem Velocity 137.4 Side Channel Thermal 137.8 Slough (16) Thermal 138.2 Mainstem Velocity 138.9 Mainstem Thermal 139.0 Mainstem Velocity 139.1 Mainstem Velocity 138.4 Mainstem Velocity 140.6 Side Channel Thermal Slough (20) Thermal 142.0 Slough (21) Thermal 141.5 Mainstem Velocity 142.0 Mainstem Velocity 142.6 Mainstem Velocity 142.8 Mainstem Velocity 143.6 Mainstem Velocity Main stem Velocity 143.8 Mainstem Velocity 143.9 Mainstem Ve.1.ocity 144.5 Mainstem Velocity Slough (22) Thermal 144.6 Slough (22) Thermal 145.5 Mainstem Velocity 146.9 Mainstem Velocity 147.1 Mainstem Velocity 147.7 Mainstem Velocity 14 8 .1 Mainstem Velocity 148.5 Mainstem Velocity 149.0 Mainstem Velocity 149.5 Mainstem Velocity 150.0 Hainstem Velocity 150.2 Mainstem Velocity 151.? Mainstem Velocity (1) Velocity indicates lead kept open by kept open b y groundwater seepage. 33RD1-007m (Continued) A~~rox. Widest Lensth (Ft) Point (Ft) 5,500 230 2,050 1,620 750 2,500 1,400 2,000 2,100 780 500 600 1,900 1,100 3,850 850 950 1,600 850 550 280 780 500 900 250 300 1,150 700 850 150 420 680 400 500 350 750 2,800 high-velocity -]I(- s-r 80 80 40 80 60 20 30 150 150 20 30 30 100 20 40 40 50 150 150 20 20 100 30 100 20 20 100 100 80 40 so 140 60 80 20 100 100 flows. Thermal Continuous or Discontinuous Continuous Continuous Continuous Continuous Continuous Discontinuous Discontinuous Continuous Continuous Continuous Continuous Continuous Discontinuous Continuous Discontinuous Continuous Continuous Discontinuous Continuous Discontinuous Continuous Continuous Continuous Discontinuous Continuous Continuous Continuous Continuous Discontinuous Continuous Discontinuous Continuous Continuous Continuous Discontinuous Continuous Discontinaous indicates lead M15/65 TABLE 5 SUS ITNA RIVER ICE THICKNESS SUMMARY 19 81-1984 1981 1982 1983 Average Th ic kness Ave r age Thi ckness Average Thickness Location Date -so l i d -S lu s h ·• Location Date So I id S lu s h • Location Dat.e So I id Slush · Vee Canyon 0 1 /13/8 1 3. 1 30 Deadman Cr . 2.9 11.4 (RM 223) 'I"_(RM 187 ) Deadmiln <.r 02/27/8 1 2.6 4.0 • wa t ana 03/11/82 3 .0 1.5 Wattt na 02/04/83 2 .II 0 ( RM 187) 011/04/8 1 3.0 0 ( RM 18 2 .3 ) ~(RM 182.3) 04/12/83 2.8 0 UR X-1 02 03/04/81 2.8 1 . 3 •fr-Portage Cr 03/13/82 2 .5 1 . 7 .Portage Cr 02/0 4/83 2.5 (RM 186.6 ) (.( RM 148.8) ~· -{ UR X-1 03 03/0 4 /8 1 2.5 1 . 5 1 "' (RM 146 .8 ) 011/12/83 3. 0 1 . 1 ( RM 165 . 7) ' URX-104 03/05/6 1 2.0 3.6 LR X-6 1 04/16/62 3 .0 3.0 ·I Go ld Cr 02/04/83 1 .6 0 (RM 18 5.3) ' I;, UR X-1 05 03/05/61 2.5 2 . 1 "t RM 146.6) ~ (RM 1 36.6 ) 04/12/83 2.3 0 (UM 184.6 1 UR X-1 06 03/06/6 1 2.4 3.4 LRX-53 04/13/62 2 .0 2 .5 (RM 184.5 1 I UR X-107 03/06/81 2.4 2.4 ~( RM 140.2 ) I (RM 184 .2 1 L""Y UR X-1 06 03/07/81 2.2 3.3 ~ ( RM 18 3.6 ) UR X-1 0 9 03/07/6 1 2 .5 2.3 Curry 03/13/82 3.2 1 . 5 02/04/63 1.9 (RM 181,.9) \ RM 120.6 ) ·'. UR X-11 0 03/06/8 1 2.0 2.0 I ( RM 120 .6) 04/12/63 2.2 ( RM 162.6 ) ! wa tana 0 4 /0 1 /6 1 2. 1 2. 1 (RM1 82.3 ) URX-11 2 03/08/61 3.0 1 •• chu l itna Co n~ 0?/04/63 2 .9 2 ( RM 18 1 .6 ) \ URX-113 03/09/81 3 .3 1.9 (RM 96.6) 04/12/63 2.8 2 ( RM 180 . 1) UR X-1 14 03/09/61 2.5 2.2 (RM 179 .7 ) URX-11 5 03/10/81 2.5 2.2 ( RM 178 .8 ) URX-116 03/11 /81 1.3 5 .5 ( RM 1'(6. 6) URX-11 7 03/11/82 2 .6 1. 5 (RM 176.2) URX-118 03/11/81 2.0 3 .0 ( RM 17 3.9 ) UR X-11 9 03/1 2/81 2 .0 3.0 (RM 170.0) UR X-1 20 03/12/61 2 .0 2 .3 ( RM 167. 1) UR X-1 2 1 03/13/61 3 .9 (RM 162.5 ) Portage Cr 03/15/8 1 2 .4 1.3 ( RM 1'18 . 8) Gold Creek 0 1/1 4 /81 2. 1 1 .0 (RM 136.6 ) 02/27/8 1 2.3 0.9 Sherman 03/05/81 2 .1 0.6 ( RM 12 8 .5 ) Cur r y ( RM 120.6) Chase (RM 103.3) 02/27/8 1 03 /05/81 1 . 8 2.5 1. 9 2.0 7. The following major side channel complexes are subject to overtopping during freezeup due to mainstem staging that exceeds the thres~old elevation. a. Alexander Slough b. Delta Islands Side Channels c. Rustic Wilderness Side Channel d. Goose Creek Side Channel e. Sunshine Side Channel 8. Flooded snow forming snow ice along channel margins, with variable widths. 9. High initial freezeup discharges near 16,000 cfs at Sunshine are common, with low final discharges of about 5,000 cfs (based on USGS daily computed values). 10. Gravel bars and islands are seldom overtopped. 11. Flow is often diverted into connecting side channels. 12. Ice cover sags due to a gradual decrease in discharge, bank storage, and erosion of the ice cover. 33RD1-007m -')fl- .s-7 U. Open leaas persist !n side channels and high velocity zones through March. i4. Surface area of open water decreases due to steady ice accumulations and decline of water surface. 15. Clear ice builds up under slush ice c o ver. 16. Minimal shore ice develops, due to relatively warm air temperatures before ice cover advances. 2. CHULITNA RIVER CONFLUENCE TO GOLD CREEK When an ice bridge forms at the Chulitna confluence, ice progression moves upstream from the confluence to the vicinity of Gold Creek. Depending on cliruatic conditions, this bridge may form either when ice cover progression in the lower river reaches the confluence, or is wel l s hort of it. Depending on the flow rate, ice concentrations and channel morphology, an ice bridge may form in November or December just upstreac of the confluence of the Susitna -· and Chulitna Rivers (figure 21). The flow discharge a t Gold Creek during this period is typically ab0ut 4,900 cfs. In some year ~ with severe cold periods occurring during ice front progression, one or more secondary bridges may form upstream of the confluence bridge, forming secondary leading edges. Th~ processes of ice cover telescoping, sagging, open lead development and secondary ice cover progression are the predominant characteristics through this reach (figure 2 2, 23 and 24). Telescoping, or shoving, occurs during consolidation of the ~ce cover. When the velocity at the leading edge is low, ice floes drifting downstream contact the edge, remain on the surface, 33RD1-nn7:: -~- S¥ and accumulate upstream b y· jux taposition at a rate proportional to the con- centration of slush ice and to the channel width. This accumulation z o ne can be extremely long, generally being governed by the local channel grad ient , amount of staging and extent of the resulting backwater. This buildup continues until a critical velocity is encountered, causing the leading edge to become unstable with ice floes submerging under the ice cover. The pressure on a thin ice cover increases as ice mass builds up and higher velocities are reached in conjunction with upstream advance . At an undetermined crit i cal pressure, the ice cover becomes unstable and fails . This sets off a chain reaction , and within seconds the entire ice sheet is mo ving downstream. Several miles of ice c over b elow the leading edge can be affected by this consolida tion. This process r e sults in icc cover stabi- lization due to a shortening of the ice cover, substantial thickening as the ice is compressed, a stage increase, and telescoping. The shoving occurs only during each consolidation. As the ice compress es downstream, tremendous pressures are exerted on the ice cover below the accumulation zone and on the river banks. The ice mass s hifts to relieve the stresses exerted on it by the upstream cover, often becoming thicker in the process. This tends to further constrict the flow, resulting in an inc rease in stage. As the stage increa ses, t!"te entire ice cover lifts. Any additional pressures within the ice cover can then be reliev ed by latera l expansion of the ice across the river channel . This process of latera~ telescoping can continue until the ice cover has either expanded bank to bank or else has encountered some other obstruction (such as gravel islands) on which the ice becomes strande d. The ict. cover over water-filled channels continues t 0 float during ice cover progression. However, because of constant contact with high-velocity water, the ice cover erodes rapidly in areas, sagging and eventually 33RD1-007m collapsing. In some reaches these open leads can extend for several hundred yards. A secondary ice cover generally accumulates in the open leads, often completely closing the open water by the end of March. The process is similar to the initial progression except on a smaller scale. Slush ice begins accumulating against the downstream end of the leads and progresses upstream. Generally it takes several weeks to effect a c omplete closure (figure 25). Ice cover saggirtg, collapse, and open lead development usually occurs within days after a slush ice cover s tabilizes (figure 26 and 27). A steady decrease in flow discharge gradually lowers the water surface elevation along the entire river. Also, the staging process which had raised the water surface within the thalweg channel tends to se ~k an equilibrium level with the lower water table by percolating through the gravels of the surrounding terraces. The ice cover continues to move up the Susitna River, although at a steadily decreasing rate as the channel gradient increases. Since the gradient and the river velocities are increasing, staging levels must increase in order to create sufficient backwater to slow velocities to allow ice juxtaposition. Although flows are only in the range of 3,000-5,000 cfs at this time, the water rises to levels equivalent to open water flows of up to 45,000 cfs. This often causes breaching of upst':'eam berms on many of the sloughs and side channels. Significant quantities of slush ice are swept into these channels, entering the backwater area caused by the downstream staging. The slush ice then consolidates and freezes in the side channel, resulting in ice thi cknesses of up to 5-6 feet. This process occurs at different levels in different years and at different locations on the river. Many of the sloughs have groundwater seeps which persist through the winter. This groundwater is relatively warm, with winter temperatures of 1-3 33RD1-007m C(R&M, 1982). This is sufficiently warm to prevent a stable ice cover from fo nn ing in those areas not filled with slush ice. This relatively warm flow develops ice along the margins, constricting the surface area to a narrow lead. The leads rarely freeze over, often extending for thousands of feet downstream. Open water was observed all winter in the following s~oughs in this reach: Slough 7 Slough SA Slough 9 Slough 10 Slough 11 The ice front progrEssion rate decreases as the ice front moves upriver. In 1982, the progression rate slowed to 0.05 miles per day by the time it reached Gold Creek. The slush ice ~ront progression from the Susitna/Chulitna confluence generally terminates in the vicinity of Gold Creek, about 35 to 40 miles upstream from the confluence, by December or early J<~nuary. This is probably due to the increase in gradient, and to the reduction in frazil ice generation in the upper river as it develops a continuous ice cover. The upper river freezes over by border ice growth and bridging before the advancing leading edge has an opportunity to reach there. See table 5 for ice thicknesses. Figure 19 shows a generalized schematic of middle river freezeup. The freezeup characteristics on the Susitna River between Talkeetna and Gold Creek are summarized as follows: 1. Frazil ice plumes appear as early as September, but more commonly in early October. 33RD1-007m 2. Velocities are generally be~een fps. 3. Discharges at Gold Creek range from 4,900 cfs on November i to 1,500 cfs by the end of March. (USGS estimates). 4. lee bridge sometimes initiate an independent ice cover progression from the Susitna/Chulitna confluence. 5. The rate of ice advance gradually decreases from 3.5 miles per day near the confluence to 0.05 miles per day at Gold Creek. 6. Flow diversion~ into side channels and some sloughs occur. 7. Surface ice constrictions are formed by border ice growth. 8. Staging levels of 4-6 feet occur. 9. lee pack consolidates through telescoping of ice cover late~ally across channel. 10. lee cover sags. 11. Open leads and secondary ice covers are common. 12. Berm breaches at Slough SA. 13. Staging affects the local water table. 33RD1-007m 14. Thermal influx by groundwater seepage prevents ice cover formation in sloughs that are not breached and inundated with slush. 3. GOLD CREEK TO DEVIL CANYON The reach from Gold Creek to Devil Canyon freezes over gradually, with a complete ice cover occurring mu c h later than on the river further downstream. The delay can be explained by the relatively high velocities induced by the steep gradient and by the absence of a continuous ice pack progression past Gold Creek. The river upstream of this reach usually freezes over by late December (figure 28), resulting in an insufficient length of o p en water remaining to generate the large volumes of frazil necessary to cause the ice cover to progress past Gold Creek (figure 29). The most significant features of freezeup between Gold Creek and Devil Canyon are wide border ice layers, ice build-up on rocks and formation of ice covers over eddies. Ice dams have been identified at several locations below Portage Creek (figures 30 to 33). Generally, these dams form when the rocks to which the frazil ice adheres are located near the water surface. When air temperatures are cold (less than -10 C), the ice-covered rocks continue accumulating additional layers of anchor ice until they break the water surface. The ice-covered rocks effectively increase the water turbulence, stimulating frazil production and accelerating ice formation. The ice dams are often at sites constricted by border ice. The dams dOd constrictions cr2ate a backwater area by restricting the streamflow, subsequently causing extensive overflow onto the border ice. The overflow bypasses the ice sills and re-enters the channel at a point further downstream. Within the backwater 33RD1-007m area, slush ice accumulates in a thin layer from bank to bank and eventually freezes. Since the ice formation process in this reach is primarily due to border ice growth, the processes described for the Talkeetna to Gold Creek r each do not occur. There is only minimal staging. Sloughs and side channels are not breached at the upper end, and remain open all winter due to groundwater inflow. Open leads exist in the main channel, but are primarily in high- velocity areas between ice bridges. To summarize, the following are the significant freezeup characteristics of the river reac h between Gold Creek and Devil Canyon. 1. The reach has a steep gradient, high velocities, and a single channel. 2. A discontinuous ice cover occurs, usually by formation of local ice covers s eparated by open leads. This results in a late freeze over, generally in March. 3. There is extensive border ice growth, with very wide layers of shore-fast ice constrictins the channel. 4. Anchor ice dams create local backwater areas, which form ice covers and cause overflow. 5. Ice covers exist over ed d ies which form behind large boulders in the channel. 33RD l -007m -):2 -,, 6. Minima l staging oc c urs. No sloughs are breached, and no flow is diverted into side channels. 7. Few leads open after the initial ice cover. Minimal ice sagging occurs. 8. Thermal influx by groundwater seeps keeps sloughs open all winter. 9. Extensive development of snow ice occurs. 4. DEVIL CANYON (to De v il Creek) Ice proces ses in Devil Canyon (RH 150 to RH 151.5) create the thickest ice along the Susitna River, with mee ~ured thic knes ses of up to 23 feet (R&M, 19818). The remote, inaccessible canyon has a narrow, confined channel with high flow velocities and extreme turbulence, making direct observations difficult. Consequently, in 1982 a time-lapse camera, on loan from the Geophysical Institute, University of Alaska, was mounted on the south rim of the canyon to document the proc esses causing th~se great ice thicknesses. Large volumes of slush ice enter t h e canyon from upstream, generated either by upstream rapids or by heavy snowfall. Additional frazil ice is generated in the extremely turbulent flow through the canyon. The slush i c e jams up repeatedly in a plunge pool below the canyon and the ice cover progresses upstream, staging the water level over 25 feet above the normal open water level. However, the slush ice has little strength, and the center of the ice c over rapidly co l lapses after the downstream jam disappears and the water drains from beneath the ice. Some slush ice freezes t o the canyon 33RD1-007m -~- ,J,. walls, increasing in thickness each time the staging process is repeated. The ice cover forms an ~ erodes several times during the winter. Two reaches in Devil Canyon were noticed on the Geophysi~al Institite film. There were a total of 6 ice cover advances observed on the lower reach and 3 on the upper. This difference is due primarily to a steeper gradient, higher velocities and turbulence in the upper section. Only during extreme ice discharges did the upper reach form a n ice cover. The initial ice cover developed in October over both reaches, but rapidly eroded away, leaving only remnant shore ice. The second major ice cover event occurred in December. with the final ice cover forming in January. All of the major ice advances seemed to be related to heavy snowfalls. A storm in January left an ice c over on the lower reach which appeared to be stable. The low discharges in January could explain the longevity of this ice cover. Devil Canyon and the reach between Devil Creek (RM 161) and the Devil Canyon damsite (RM1S1) are the first areas on the Susitna to form ice bridges and develop an extensive ice cover. Ice covers of one mile in length were observed to form about two miles below the Devil Creek confluence as early as October 12, 1982, despite relatively warm air temperatures. The ice formation process at this point is believed to be similar to that in Devil Canyon. To summarize the freezeup in Devil Canyon: 1. The narrow. confined channel has high flow velocities and turbulence. 2. There is early formation of ice bridges and loosely packed slush ice covers. 33RD1-007m -,..- '" 3. Ice covers form and erode several times during the winter. 4. The ice covers are inherently unstable, eventual collapsing long before breakup. 5. There are extreme staging levels and ice thicknesses, with ice accumulations up to 23 ft. 5. DEVIL CANYON TO THE OSHETNA RIVER Lateral accumulation of border ice layers is the predominant process of ice cover development through this reach. The border ice often constricts the open water channel width to less than 10 feet. The slush ice then jams between the shorefast ice and freezes, forming an unbroken, uniform ice cover acros s the river channel. However, since this process does not occur simulta- neously over the entire reach, a very discontinuous ice cover results. Numerous open leads generally exist until early March . Upstream of the Devil Creek confluence the Susitna River has a steep gradient, single channel with gradually sloping banks to the vegetation trim line. At this point on a typical cross section the bank slope increases • dramatically and rises to nearly 1000 feet above the channel bottom. The gradually sloping banks below water line result in shallow water along the flow margins that abound in large boulders. These boulders provide anchors for slush ice that drifts into and stops along the river banks. Shere ice rapidly develops laterally out into the channel until encountering water velocities greater than 2 fps . Water velocities prevent the slush from adhering to the shore ice fringe or from anchoring to flow obstructions. Howeve r , eddies downstream of boulders do fill with frazil ice, resulting in 33RD1-007m small patches of surface ice with a fluted shape, the upstream tip being the boulder. An ice bridge usually forms at a flow constriction dow"llstream of the Tsusena Creek confluence at RM 181. Flow is funneled between a vertical rc·ck cliff and a gravel bar. Shore ice grows in the shallow water over the gra,el bar and the water depth does not reach the limiting 2 feet until all but 10 feet of the channel is ice covered. The rem~ining 10 feet of open water plJgs up with slush ice, which subsequently freezes solid. The water velocities are too high for an upstream progression, and all the slush from upstream is swept under the ice bridge. An ice bridge also formed near the mouth of Watana Creek, usually by mid-November. A continuous ice cover advances upstream and in some years reaches the mouth of Vee Canyon at RM 223. Water levels stage to over 10 feet above the initial open water, and slush ice is shoved laterally onto gravel bars and up to the vegetation trim line on both banks between ~~ 210 and 22 0. Anchor ice accumulates on the channel bottom to thicknesses exceeding 2 feet in some areas, raising the water level correspondingly. Th i s usua.l ly occurs when shore ice is present, and the rising water either fractures t he solid ice or overflows on top. In the latter case, snow laying on the shcre ice is flooded and eventually freezes, increasing the overall ice thickness significantly. The lateral growth of shore ice also increases the wat ~r level, but this is less noticeable since the discharge is simultaneous :~y decreasing. By the end of December the reach from R}f 170 to RM 210 has a discontinuous ice cover resulting f rom numerous bridgings between wide shore ice formations. The remaining open leads eventually f~eeze over if ait temperatures remain below -10 C for a long period, but otherwise water 33RDl-007m -~ '~ velocity and turbulence Keeps the leads open. se~ table 5 for ice thicknesses. Characteristics of Susitna River freezeup between Devil Creek and the Oshetna River are summarized as follows: 1. Extremely wide accumulations of border ice layers occur, resulting in gradual filling of the narrow open channel with slush ice which freezes and forms a continuous ice cover. 2. There is extensive overflow and flooded snow. 3. Few reaches exist where staging or telescoping occur. 4. Low discharges through the winter result in shallow water and moderate velocities. S. Minimal ice sagging occurs , with few leads opening after initial freezeup. 6. Extensive anchor ice forms, with high sediment concentrations. 6. THE FREEZEUP OF LOI.JER AND MIDDLE RIVER SID£ CHANNELS, SLOUGHS AND TRIBUTARIES To review processes of ice cover development on bodies of open water, black ice develops and grows on the surface of non-tu-bulent water. Snow can 33RD1-007m stiruulate this by covering and floating on the water surface, effectively decreasing the volume of water that needs to freeze before solidification occurs. An ice cover finally develops by frazil slush accumulation. The side channels, sloughs and tributa ries of the Susitna are subjected to one or all of these processes. :>i.!e channels generally have bed elevations higher than the adjacent mainstem, and therefore normally convey water only at relatively high stages such as those ass ociated with spring runoff and summer rain s torms . Susitna River side channels vary considerably in length, width and complexity from the relatively short systems on the middle river to the extensive multi-channel complexes on the lower river. Most lower river side channels are de-watered prior to freezeup when mainstem flows drop below 10,000 cfs. Some of these may have separate sources of water such as a minor tributary or groundwater seeps. These contribute enough water to maintain shallow pools. With the advent of cold air temperatures, these pools often develop ice covers by either black ice or snow ice formation. Snow ice obviously forms only after a heavy snow storm. This cools the water surface rapidly and often freezes into a solid but thin ice sheet. The snow can initiate an ice cover man y days before it would have frozen without the snow. This ice appear':.' white or opaque and grainy in contrast to the clear and smooth black ice which begins to grow underneath the snow ice. Staging, or the rising of mainstem water levels due to the accumulation of frazil slusn, often increases the water level sufficiently for t he side channel entrance to be overtopped. The sudden increase in water volume washes away the snow cover and flushes out the pools, fracturing the ice cover . This often results in small ice jams. The flooding continues until the mainstem water level recedes below the side channel threshold elevation. In some cases 33RD1-007m -~- ?a overtopping continues all winter long. Depending on the depth of water over the channel entrance, an ice cover then forms by either shore ice growth (shallow overtopping depth) or frazil slush accumulation (deep overtopping depth). In the case of a shallow overtopping depth, mainstem water enters the side channel, but due to the slush floe thicknesses which usually exceed 1 foot, the floes become stranded at the channel entrance and cannot enter. When overtopping depths are greater than 1 foot, then slush ice flows into the side channel and e i ther continues through to the mainstem again, or becomes lodged and initiates an ice progression that develops into a solid ice cover. If a progressio n does not start, then border ice forms along the banks and continue growing laterally into the channel, eventually closing over the open water. Side channels with separa te sources of water generally develop open leads through an ice cover. This is due to thermal erosion by relatively warm water emanating from the ground or tributary. ~· Side channels on the lower river which are usually overtopped during freezeup are: Alexander Slough Delta Island Side Channels Rustic Wilderness Side Channel Goose Creek Side Channel Sunshine Side Channel Side sloughs by definition are side channels with a source of water separate from the mainstem. They generally have an upstream entrance, bifurcating from the mainstem, and an exit which rejoins the mainstem. Upland sloughs have no upstream entrance, only a mouth, but the water level in the mouth is controlled by backwater regulated by mainstem stage. At low mainstem 33RD1-007m stages the mouth areas of tnese sloughs are dry except for the contribution o f its tributary or groundwater source. Side sloughs that are overtopped during mainstem freezeup behave similarly to the previously described side channels . An ice cover of either snow ice or black ice usually forms over the ponded water by mid-November. As the mainstem ice cover advances upstream, a large volume of water percolates into the substrate and surrounding terraces as bank storage. This has been doc umented by an increase in the water table adjacent to a developing ice cover. This is also noticeab le by an increase in the surface area of the ponds and the subsequent flooding of the surrounding snow (figure 34). The water level rise is not sudden enough to fracture the ice cover. Several sloughs between RM 120 and RM 130 are overtopped by the mainstem during freezeup. The overtopping usually consists only of water and no slush ice enters. An exception to this occurred in 1981 when Slough 8A was breached by an estimated 140 cfs for over 1 week. Any ice cover that forms over a slough is usually eroded away soon after mainstem ice progression because of the increased heat flux from groundwater flow into the sloughs (figure 35). This source of heat is significant enough to erode through an existing ice cover and keep the open water ice free thro ugh the winter. Often the thermal erosion continues out into the mainstcm ice cover, which develops an open lead wit h its source in the slough. Middle river sloughs and side channels that are c ommonly overtopped during freezeup inclutie: 33RD1-007m Lane Slough Mains tem II Side Channel Curry Side Channel Slo ugh 7 -~- 72- Slough SA Slough 9 Upland sloughs, such as Sloughs 6 and 10, generally do not develop a continuous ice cover but shore ice will form along the channel fringes. The majo~ tributaries of the Susitna form an ice cover by surface accumulatio:-.s of fr.:~zil slush. These include the Yentna, Deshk.a, Talkeetna and Chulitna Rivers. Most of the minor tributaries develop an ice cover by shore ice and anchor ice accumulations. They are generally too shallow and turbulent to fonn an ice cover by frazil progression. A majority of these tributaries drain basins large enough to contain a sufficient volume of stored groundwater to maintain a flow through the winter. This groundwater flow retains enough heat to erode through an existing ice cover when air temperatures begin moderating in early April. Before breakup these tributaries usually show a discontinuous ice cover and a lead eroded through the mainstem ice cover near the confluence. 7. SUMMARIZED HISTORICAL FREEZEUP CHRONOLOGIES This section is included for the purpose of consolidating the river ice observations from 1980 to 1984. This will facilititate evaluating the significance of annual variations in the freezeup sequence. Included in this section are tables 6 to 10 and figures 36 t.J 39, which summarize pertinent information from each year of the study to show how the varying climate conditions control ice cover formation. a. 1980 Freezeup Chronology 33RD1-007m -)It- 71 Climate conditions in"the Susitna Basin varied significantly from normal during the study period, influencing the processes of ice cover formation and breakup on the river. In early December air temperatures were well below normal. This was followed by unusually warm air temperatures in January after the ice cover had formed over the length of the river. During these early winter months, precipitation was low. Snow survey data showed that the snowpack in the Susitna Basin was 30-50% below normal through January. The combination of these factors resulted in an average ice thickness of 2 .5 feet on the Susitna River at Gold Creek in January, close to the historical average at that site. On October 11, 1980 in the vicinity of Gold Creek, areal coverage of frazil ice in the main channel was estimated to be 40%. It appeared that frazil was being generated primarily through Devil Canyon. At this time, there were no signs of frazil or shore ice developing in the Chulitna or Talkeetna Rivers. By late afternoon on October 12th, the leading front of floating frazil ice was approximately 5 miles above the Kashwitna River confluence (approximately RM 66). Frazil ice was flowing in the Yentna River, but no ice was observed in the Deshka (Krotc Creek). Frazil ice coverage in the main channel of the Susitna averaged 30% in the river above Talkeetna. Floes were beginning to accumulate at natural constrictions and in low velocity areas. Shore ice was also beginning to form in the quiet-water areas, but there was no significant constriction of the main channel due to shore ice growth. The following day, October 13th, first frazil ice was observed in the Talkeetna River, but there was still no sign of frazil ice in the Chulitna River. Ice floes in the Susitna River above the Chulitna/Susitna conf:~enc~ were more concentrated, with coverage in the main channel estimated at 80%. 33RD1-007m -)1- 7'1 The floes varied from 2-S feet in diameter through more turbulent reaches up to 50-10 0 feet long in the constrictions below Curry and Portage Creek confluence. A thin ice cover had formed on some quiet-water s loughs and side channels. Tributaries upstream from the Susitna/Chulitna confluence showed no signs of flowing frazil ice. On October 31st, anchor ice was first observed in the river near Sherman. The ice accumulated in masses 3-4 inches thick over SO % of the cobble bed in the near-shore area. Anchor ice was still present in water depth s of 4 feet up to 30 feet from shore in the main channel. Several ice bridges were ob- served through Devil Canyon and upstream to Devil Creek. By mid-November, anchor ice could be clearly seen along the length of the river from Talkeetna to Portage Creek. In the main channel, most reaches of shallow, high velocity water had anchor ice over 50-70% of the bed. Spring-fed side channels showed no signs of anchor ice formation. No ice bridges existed below Portage Creek. The most noticeable channel constrictions occurred just upstream of Curry near RM 120.8, at RM 126.1, at the bedrock outcrop near R}f 128.5, just upstream of Sherman at RM 131 and at the rock point near RM 135.8. By November 13, most of the tributaries below Talkeetna had formed ice covers near their confluences with the Susitna. In the lower river, the leading edge of the ice cover wa s observed at RM 75.5 approximately 8.4 miles below the Parks Highway Bridge. Upstream from this bridge to Talkeetna, flow was confined to the main channel, and side channels were either ice-covered or dry. On November 29, frazil ice coverage in the Talkeetna River was 40-50%, with most flow through the north channel. There was no sign of an ice cover forming in the Chulitna River near Talkeetna, with approximately 40% frazil 33RD1-007m ice coverage. The Susitna River at the confluence with the Chulitna showed 80-90% coverage of frazil slush ice, but the channel was still open. On December 1, an ice bridge was observed across the Susitna River at the Susitna/Chulitna confluence. The Chulitna River was still ice free. Evidence of a rise in water level of 3 to 4 feet occurred between November 29 and the morning of December 1 upstream of the ice bridge. Over the next two weeks the progression of the ice cover between the confluences and Gold Creek was monitored to determine the rate of ice cover growth upstream. The average rate of ice cover growth was 2.7 miles per day. Overall, there was little observed variation from this rate. It is important to note here that during ice cover formation air temperatures at Talkeetna were far below normal, which would tend to accelerate the rate of ice cover growth. On December 2nd, the leading edge of the ice cover was at RM 108 .5 . At RM 110.4, the width of open water was 100 to 125 feet and the edge of shore ice was approximately 80 feet from the toe of the right bank. Upstream from the leading edge of the ice cover, there appeared to be little change in the ice conditions along the river through Devil Cany ~n. However, from Tsusena Creek upstream, the channel was severely constricted by snore and anchor ice growth. At Watana Creek, an ice cover had formed which extended upstream to approximately 3 miles above the Kosina Creek confluence by the afternoon of December 3. At a few sites there was water spilling into side channels, indicating a rise in water level. The next re c onnaissance trip for ice observ ations was carried out on December 8. By this time, the i c e c ov er in the river below Talkeetna had progressed as far as RM 93.5 . Above the Susitna/Chulitna confluence, the leading edge of the ice cover was observ e d at RM 126.4. 33RD1-007m The final reconnaissance trip f o r freezeup observations was conducted on December 12. The ice co·1er t:xtended a s far upstream as Gold Creek. Upstream of Gold Creek, there were n o h .e bridges in the ch'lnnel until just below Portage Creek, where a small bridge had formed on the upstream side of a constricted bend in the channel. On December 15th, the ice cover extended upstrea m past Portage Creek and in t o Devil Canyon. On December 30, the i c e cover extended intermittently through Devil Canyon upstream to 4 miles above Devil Creek. Open water persisted in several turbulent flow reaches. Further upstrea m there was a discontinuous ice cover with several open leads. b. 1981 Freezeup Chronology The 1981 freezeup process was prolonged by the delay of cold weather. This contrasts with the 1980 freezeup, when a November cold snap caus ed rapid ice formation. Fluctuating air temperatures and relatively heavy precipitation through October precluded the formation of a stable ice cover. By the second week of December the lea ding edge of ice on the lower river was near Talkeetna, about two weeks later than in 1980. Long before an ice cover formed adjacent to Talkeetna, an ice bridge formed at the Susitna /Chulitna confluence. The i c e bridge at the Susitna /Chulitna confluenc e presented a barrier to ice floes, greatly re ducing the ·;olume of ice f eeding the down- stream ice pack. Co n sequently, it took over 6 ~eeks for the confluenc e area between Talkeetna and the ice bridge on the Susitna to develop an i c e cover. Fr azil i c e was first o bserved on October 2 , 198 1 at RM 11 0 during a morning flight up the Susitna River on October 2, 1981. No frazil ice was observ ed in the confluenc e area. 33RD1-007m --,1- 77 With air temperatures fluctuating above and below 0 C all through October, no permanent ice formations developed. Between October 12-1), temperatures increased sufficiently to melt much of the remaining border ice. On O~tober 29, border ice was again building along both sides of the river, and most side channels were de-watered. Ice ?ans and rafts from the Susitna formed 70% of the t o tal floating ice below the confluence with the Chulitna River. By November 2, mean daily air temperatures had remained consistently low (about -11.0 C) for several days. Above the mouth of Deadman Creek, the border ice had extended into and closed the channel to form an ice bridge. Ice pans were accumulating against this obstruction, causing an upstream advance of the ice cover. Another channel closure was forming just down s tream of Bear Creek confluence, about 1 mile below Tsusena Creek. An extensive ice bridge had developed below Fog Creek confluence, but the ice cover was not progressing further than the rapid s immediately below the Fog Creek confluence. A continuous ice cover had formed over the two mile long rapids section below the Devil Creek confluence. Many ice bridges were building between RM 155 and RM 160. Devil Canyon had a co ntinuou c ice cover from the proposed damsite down to RM 150. The discharge at Gold Creek at the time of these observations was 4,100 cfs. remained open. Below Gold Creek, the river channel Ice rafts were periodically broken up and reformed by local variations in flow. At RM 115 c hannel constrictions concentrated the ice rafts, and bridging seemed imminent. Cold air temperature s continued, and on November 6 the following aerial observations were recorded. Below Talkeetna, the Susitna wa s ice-covered from Cook Inlet to approximately the Kas hwitna River. The channe 1 at the Parks 33RD1-007m -~- -----~ Highway Hridge was choked· with slush ice rafts. The confluence area showed some frazil ice being contributed by the Chulitna and Talkeetna Rivers, but most of the ice was drifting down from the Susitna. In the Chase area 50-60% of the river channel was covered by border ice. An apparently stable ice bridge had formed at RM 105.5. Slush ice rafts were accumulating against it, creating an upstream progression of ice coverage. Channel constrictions were ohserved at RM 123, at RM 131 at Sherman, at RM 136 below Gold Creek, at RM 145 and at RM 149 just above the Portage Creek confluence. The ice cover and bridges through Devil Canyon remained stable with no significant growth observed. Little further ice formation was reported in the reach from Devil Canyon to Watana. On November 18, the mean daily air temperatures ranged from -13 C at Talkeetna tc -16.0 C at Watana. The lead ing edge of the ice cover had progressed upstream to within 4 miles of the Parks Highway Bridge. The Chulitna River showed increasing i ce formation activity, with moderate concentrations of frazil ice and ~idening border ice . The Talkeetna Riv~.r was completely ice-covered and was no longer contributing ice to the Susitna. Slush ice rafts on the Susitna River had consolidated and j ammed at a bo rder ice constriction near the confluence. The rea ch b e tween Cur r y and Sherman was characterized by extensive anchor ice, giving the water a milky appearance. The ice bridge below Gold Creek remained stable, with ~10 i ce progression. No significant ice formation had occured above the Devil Canyon area. By December 14, the i ce c uvet had progressed to RM 95 below Talkeetna. From there to the Susitna/Chul i.tna confluence, the river maintained an open channel. At the confluence, an ice cover 1·esumcd on the Susitna Riv~r and continued to RM 12 7 with the exception of narrow open leads of varying lengths, usually less than one-half mile long. The open channel above RM 12 7 33RD1-007m -x- 71 was 40-50 feet wide, and contained 70% frazil ice. Extensive patches of anchor ice were also observed. At Gold Creek the channel was 60-70 feet wide with no visible frazil ice. The ice-covered reaches in Devil Canyon and below Devil Creek confluence had developed narrow open leads about ~ and 1 mile long, respectively. Above Devil Creek, the river remained open with extensive herder ice formations constricting the remaining open water. On January 4, the Talkeetna, Chulitna and Susitna rivers were frozen at the confluences with the exception of open leads through the ice cover resulting from high water velocities. The Susitna above the confluence was generally ice-covered, with many reaches of narrow open l ead s . Near Sherman at RM 127 , an open channel about 1 mile long persisted. Above Sherman, the open leads became more numerous and generally longer. Above Go ld Creek, the river was essentially open but had many anchor ice dams. Little had changed through the Devil Canyon reach and further upstream. c. 1982 Freezeup Chronlolgy Between October 22 and Oc tober 26, 19 82, slush ice jammed near RM 10 and accumulated upstream 57 miles to the Sheep Creek confluence. Assumi.tg that the ice cover began progre ss ing upstream on October 22, then the progression rate was 11.5 miles per day . On Oc tober 26, the i ce cover had progressed past Rl-1 25 but was not continuous. There was no ice cove r on th e Susitna near the confluence of the Yentna River . The Yentna was also completely f ree of drifting ice and s hore ice. At Rl-1 3 2 , a loosely packed ice cover re sume d and continued upstream to RM 67. From RM 67 to RM 97 near Talkeetna, the river remained free of shore ice, even though a large volume of slush ice was con tinually drifting 33RD1-007m downstream. All of the major tributaries to the Su~itna below Talkeetna were still flowing and remained ice-free. The leading edge of the ice pack o n October 29 was near RM 87 , just upstream f r om the Parks Highway Bridge and adjacent to Sunshine Slough. However, the ice cover remained discon tinuous, with long o pen leads at the Yentna River confluence near Susitna Station, the Deshka River confluence, Kashwitna River, and Montana Creek. These tributaries were still flowin g but showed signs of an ice cover beginning to develop. From RM 76 up s tream to RM 87 the ice cover was thin and discontinuous, with long open leads adjacent to Rabideux Slough and in a side channel that extended from ~ mile below the confluence of Rabideux Creek down s tream for about 1 mile. By November 2, the leading edge had advanced to RM 95 at a rate o f 2 .1 miles per day during the previous 4 days. Many side channels had filled with water and the surface of the ice pack was near the vegetaLion line along the le f t (east) bank. The c hannel along the west bank remained dry and snow covered. An ice bridge formed at the Susi tna and lhulitna confluence on November 2, greatly reducing the volume of s.:.ush ice flowing in to the lower river and s lowing the rat e of ice cove r advance s u bs tantially . The formation of this lee bridge was dependent on decreased velocities b r o ught on by the proximit y of the leading edge . A snow storm immediately preceded the formation of the ice bridge at the Susitna /Chulitna confluence. This storm ma y have caused a s ubstantial local inc rease in ice di scha rge which could not pass through the cons tri ct~d channe l at one time. I ce disc harge estimates were substantially l ower a t Talkeetna after Nove mber 2. Several ice cons trictions were located near Curry (RM 120.6). Sl o ug h 9 (RM 128 .5) and Gold Creek (RM 135.9). Slush ice had been 33RD1-0 07 m -~ ~~ passing easily through these narrows since October 26 , but was now being compressed into long narrow rafts which broke up within several hundred f eet downstream. Unlike the co nfluence area, these constrictions were formed by successive layers of frozen slush ice along the shore. The r ate of ice advance averaged 1. 6 miles per day for thirteen days after passing Whiskers Creek. On November 22 the leading edge was situated adjacent to Slough ~A . The ice cove r had staged approximately 4 feet and was overtopping the berm at the head of Slough 8A. The estimated discharge through the s l o ugh was 138 cfs. The ice cover ~as very slow in advancing thro ugh the shallow section of river between Sloughs 8A and 9 . By December 2 , the ice cover had advanced at a rate of only 0 .3 miles per day for the previous 10 days, even though high frazjl slush discharges were observed at Gold Creek. On December 9 the leading edge had reached RM 136, just downstream of the Gold Creek Bridge. The ice cover advance stalled here for over 30 days , as the ice needed to accumulate in thickness before it could stage past this high-velocity channel constriction. On January 14, 1983, the leading e t.!:;€! finally crept past the Gold Creek Bridge at a ra te of 0.05 miles per day. d. 1983 Free zeup Chronology On October 17, 1983, slush ice was flowing through the middle and lower river, depositing along flow margins where it quickly froze into border ice. From Oct ober 23 until October 26 s lush ice f loes were estimated to cover 60% of the open wat er surface area on the Yentna River and about 40% o n the Susitna. On the morning of Oc tober 26 , 1983 an ice bridge formed at IU1 9 . On November 1 , 1983 the leading edge was a t RM 31.5, having progressed more than 16 miles in five days. By November 4 the ice front had passed the 33RD1-007m -~- ?'2- confluence of the Deshka River at RM 40.5. The maximum stage inc~ease l!'easured at the entrance to Kroto Slough (RM 40.1) was 3. 9 feet. This was sufficient to overtop the slough with a flow depth of 1.5 feet at the entrance, but no slush ice floes could enter due to their thicknesses of abou t 2 feet. The elevated ma j nstem stage also effected the Deshka River by creating a backwater zone which extended about 2 miles upstream. On November 5 the ice c0ver progression entered the Delta Islands. The leading edge split, and ice fronts advanced separately up the east and west channels. The east cha nnel ice cover progressed more slowly. The advancing ice cover caused stage increases high enough to inundate the snow cover over the Willow Creek alluvial fan. This saturated snow then froze into an i~.;e cover. However, th'2 water course from Willow Creek was not altered. The measured stage incr1!ased about 3 feet during the ice front advance. Slush ice from the Susitna d jd not encroach on the creek confluence, The stage increase measured at the entrance of a side channel near RM 48 on the west channel was about 2.5 feet. This channel was flooded but no slush ice entered. The Susitna ice covel progressed through the Delta Islands and converged near RN 51. Most of the major sid : channel complexes on the lower river were flooded during ice cover progre ss ion. Nainstem slush ice was observe d to accompany the surge thr o u ~h the Rustic Wilderness side channel, The slush ice and ice debris occasionally accu~ulated in small jams a short dista nce below the side channel entrances but usual ly were carried back out to the mainstem.. A maximum mainstem stage increase of 3 feet was measured near the mouth o f Kashwitna River (RM 60) on November 11. On November Y the leading edge was at RM 66, but the new ice cover remained un stable due to warm air t e mperatures that preve nted the s lush from 33RD1-007m ~ YJ freezing. This was apparent by the quickly deteriorating ice cover below the leading edge. An open lead had formed froo RM 62 to RM 65. The leading edge continued to advance at an average rate of 2 miles per day, even though the channel gradient gradually increases beyond RM 66 and more ice was required to produce a sufficiently stable cover. The effects of mainstem staging were not evident to a significant degree at the mouths of either Sheep Creek or Goos e Creek. Sheep Creek drains into a side channel that extends from RM 62 to ~~ 67. Through this reach the mainstem is along the west bank and since the side channel complex is on the east bank, it was there fo re not affected by back- water or cn:e rtopping. Goose Creek enters a side channel that runs from RM 69 to RM 72. This side channel was also not flooded or affec ted by backwa ter when the mainstem water level staged. The stage at the se tributary mouths did increase slightly due to a rise in the local water table. On November 19 the stage was rising at entrances to Suns hine Slough. Th e slough and side channels were eventually overtopped and fl oo ded, although no slush ice entered. These channels s ubsequently required an additional 8 -12 weeks to freeze ove r and many leads remained open all winter. The side channels leading to the entrance of Birch Creek Slough were flocded but the stage did not increase enough to overtop the slough entrance. The maximum increase was 3 .1 feet near the entrance to Birch Creek Slough. An additional foot would have been ~ecessary for overtopping. The tempor a~y arrival of the leading edge at Rl-1 95.5 initiated a separate ice progression up the Talkeetn n River. This progression o n the Talkeetna wa s so late, however, that the ruajority of the river had already frozen over with anchor ice and bo rder ice, significan tl y reducing the volume of frazil being genera ted. By mid-December the ice cove r had r eached a position about 300 ya rds upstream 0f the railroad bridge and e ssentially remained there for the rest of the winter. 33RD1-007m By December 9 the Susitna River ice front had advanced upstream into the middle reach above Talkeetna, No intermediate ice bridges formed. The ice cover on the lower river remained unstable and was marked by many extensive open water areas, either in mainstem leads or in flooded side channels. The Chulitna River, like the Talkeetna, had frozen over by lateral ice growth at the headwaters and was b y this time generating so little ice that no upstream accumulation occurred. The coni luence area of the Chulitna did not freeze over until late March, 1984. This was entirely due to anchor ice and lateral growth of surface ice. On December 22, a second leading edge was observed progressing from an i c e bridge at RM 120.7, jus t upstream of Curry. The river downstream bet\olee n RM 120 and RM 118 wa s still open -=1nd an ice fr on t was not longer advancing there. Heavy anchor ice deposits were observed within the open lead. This anchor ice had noticeably raised the water level and flooded the surrounding shore ice and s now. A new leading edge was subsequently started and moved past Curry (LRX-24) probably on December 21, 1983. The open water below Cur ry on the mainstem eventually froze over by border ice growth. Two anchor ice dams were observed in t he lead, at RH 120 and RM 119.6. Thi s created some backwater pending which facilitated faster lateral growth of border ice. By January 5, 19 84 the second leading edge was located at Sherman near RM 130. Since very little slush ice was flowing in the open water, the ice cover progression was relatively s l ow . By this time the river above Devil Canyo n had essentially frozen over and s t opped generating 3ubstantial volumes of fr a z il. An ice bridge formed a t RM 135.6 and a third leading edge began but 33RD1-007m p ~o gressed only about ( mile before becomir~ an indefinite zone of accumulation. The open water area between Rl-1 130 and RM 135.6 eventually covered by border ice closure. e. 1984 Freezeup Chronology Unusually mild weather during September and early October delayed the formation of significant volumes of frazil ice until the fourth week in October. The lack of late summer rainfall resulted in lmv freezeup stages compared to previous years. Slush ice was first observed flowing down the mainstem at Go ld Creek on October 16, 1984, although it had probably started flowing during the previous night. Variable concentrations of ice were o bserved until the af ternoon of October 22, when the air temperatur ~s warmed to 3 C and all ice disappeare~ A full 6 fee t of accumulated border ice disintegrated at Gold Creek during the follow ing two days . Slush ice concentrations began to increase again on <l::tober 25. On October 26, at river mile (RM) 9, near the mouth of the Susitna River, a dense con centration of ice f loes had ac cum ulated during the high tide of 32.4 feet (Anchorage referenc e station) at about 7:30 a.m. At R}l 9 the tidal fluctu a tio n was measured to range over 6 feet during this particular cycle. Near the river mou th t he top 1-2 inches of the ice pans had solidified, forming a rigid shee t on the surface. Pan size was variable with average diamete r s ranging fro:n 2 feet to over 6 feet. The water velocity during the high tide was le ss than foot/sec. and at low tide about 2.5 feet/sec:.: at center channel. The ice floes which drifted into the flow margin a long th·~ ea s t bank were barely moving, and became grounded when the t i d e receded. 33RD1-007m -)il!"- Pt The following day, October 27, the ice concentration in the area below r~ 9 again increased during high tide. With substdntially higher volumes of i c e floes coming into this area from upstream due to the co ld air temp e ratu·.es, an ice bridge developed near RM 5. On October 29, a complex picture unfold~d of ice cover develoyment on the lower reach. From R}f 5 a somewhat continuous cover extenJed to RM 19, adjacent to the entrance of Alexander Slough. The predr,minant process of aelv.:'1ce was by juxtaposition. Large areas of ope:. water were present througho ut ..:1-)e cover, indicating that little pre !'"_,u re was acting on the ice and no comprcssio.: had occurred. By 10:30 a.m. on Oc tober 29, the leading edge was located at Rt-1 1':1. ~-:-·;::·.o::r , due to insufficient ice from upstream, the leading edge was no longer advancing, and this ice fron t was essentially stalled at RM 19. The leading edge consisted of a thin layer of fine slush that was building in diagonal layers across the channel from the area of high water velocity on the outside of the river bend to low velocity on the inside of the bend. Open water with no slush was noted f r om Rl'l 19 to RM 25 .9 at Susitna Station (USGS gage site), where a second ice bridge had formed. A continuous ice cove r had devel oped upstream f r om ~~ 25 .9 to RM 43 of the east channel through the Del ta Islands. The ice cove r had also progressed up the Yentna River abou t 12 mil~s. The west channel through the Delta Isiands was entirely open from ~~ 42.5 to RM 46 . At RM 46 on the west channel a third i ce bridge had formed. This obstruction had prevented slush ice from drifting downstream to advance the ice cover above ru-: 42.5. From the ice bridge at ~~ 46, the ice cover had progressed up the west c hannel to RM 51. At this point the ma in c hanne l bifurcates , creating the west and east channels. The ice cover progression had sto pped here, and there was open ~ater up to P~ 52. The east channel was open from RM 43 t o RM 52. A fourth ice b ridge had fo rmed 33RD1-007m -~ Y7 at RM 52. Very little slush ice emerged from under the downstream edge of the bridge, indicat ing that mos t of the ice floes were retained by the advancing ice cover near the leading edge. This ice cover had progressed up to RM 5 5 . Visual estimates of slush concentrations at Gold Creek during the 4 days following the initial ice bridge formation were never less than SO% of the total open water surface area. By November 3, the leading edge of the ice cover, which now originated from the ice bridge at lUI 52, had progressed to R~! 71.5 at an average rate of 4.1 miles per day . At the three rivers confluence, the Chulitna l:<.iver and Talkeetna Rivers appeared to be contributing most of the slush ice to the lower Susitna River. The Susi tna above this con fluen c e area contained v ery little slush. At RN lOS, s lu s h ice had bridged the river at a shallow reach. Thi s bridge had remained s t ab le long enough to initiate an upstream progressio n of an ice cover on the middle reach of the Susitna. The consequence of this new progression was a de creased supply of s lush ice to the lower river ice front, ultimately delaying ice cover formation below Talkeetna. The leading edge progression rate slowed to under 2 miles per day on the l ower rive r, being e ntirely dependent on slush from the Ch ulitna and Talkeetna Rive rs and o n f razil ice gene rated below the ice bridge at RH lOS . A warm weather period began or. November 5 and lasted until the lOth. Ice conce ntra tions s harply decreased during this period to less tha n 10% at Gold Creek on November 8. Thi s subsequently decrea sed the rate of leading edge progres sion to 0.5 miles per day on the middle river and 0 .2 mile s per day o n the lower river. At thi s time, a n estima te d 75% of the slush forming the lower r iver ice cover above t he Yentnd River conf luence o riginated from tr~ Chulitna and Talkeetna Rivers. 33RD1-007m On November 10, cold air temperatures once again increased the ice concentrations, and on Novembe~ 13 the surface coverage was estimated at 80% by the Gold Creek observer. The middle river ice front advan c ed 6 miles (up to RM 121) and the lower river front moved upstream about 2 miles (up to RM 86). The middle river ice front progressed more rapidly due to a larger volume of slush ice generated in the available open water reach from Gold Creek to Watana. On November 14, the Chulitna and Talkeetna Rivers had forced ice bridges several miles upstreac of the Susitna confluence. These ice bridges prevented slush from entering the Susitna, and the ic~ cover progression on the Susitna stopped at ~~ 88. An insufficient supply of slush prevented further upstream progression at the rates previous ly observed. In Slough 8A, ponds with black ice about 4-6 inches thick began overflowing and flooding the surrounding snow :over on November 16, ~hen the leading edge was located at "J-1 12 7 . This indicated that groundwater levels were rising. The entrance berm at Rl·! 127 had not ye t been overtopped. HO\•ev- er, the berm at RM 126.1 had been f~ooded . The upper entrance to Slough 8A began overtopping on November 19 when the leading edge was at RM 128. This event was not nearly as dramatic as the previous overtopping in 19 82. From the air it was difficult t o tell that overtopping had occurred. The snow cover was about fo o t thick at the time, and the mainstem water seeped through the snow pack. On November 21, the leading edge of the ~iddle river ice fro nt reached RM 129, near the entrance t o Slough 9. No ove rtopping o f the entrance berm occurred. By this time the river upstream of Devil Canyo n had become ice cov~red, ~ever e ly limiting the volume of frazil capable of being generated, The rate 33RD1-007m of leading edge advance ·subsequently slowed t o about 0. 2 miles per da}. Anchor i ce had accumula ted on the bottom in massive proportions. Thi ck layers often broke free from the bottom and floated downstream to also become part of the downstream ice cover. On Decemb er 15, backwater from an ice dam at RM 135 caused a fracturing of upstream border ice. A large solid fragment drifted downstream, but instead of floating down to the leading edge at RM 131, it be came lodged on the anchor ice dam at IU-1 135, creating a new ice bridge. This ice bridge accumulated slush ice at the upstream cage. The new ice front prevented slush from continuing d ownstream and advancing the previous leading edge. By December 20, the river under the Gold Creek bridge had fro zen over and the leading edge was approaching RM 137 . The o p e n water belcw the ice dam at RM 135 remained as it appeared a week earlier. By the final observation flight on December 20, 1984, the leading edge on the Susitna River below the thr e e rivers confluence had reached IU-1 92, at a rate of 0. 1 2 miles per day . The Talkee tna Rive r was fr ozen ove r above the railroad bridge. The Chulitna River was frozen over f r om about three miles a bove the Susitna conf lue n c e. Exte n sive o pen l eads exis ted in the Susitna River ice cove r below Talkeetna . Open wa t er s till persis ted on the east channel o f the Delta Islands , al t hougil the flow velocity had dimin ished in man y places and b0~de r ice wa s beginning to c lose the open c hannel i n several areas. lo 33RD1-007m -,x--- M18/5 1 Date Nov. 29 Dec. 1 Dec. 3 Dec. 5 Dec. 8 Dec. 12 De. 15 Ha:y 3 Ma:y 4 ~ Ma:y6 May 9 Ice Bridge Loca t ion ( RM l 97 98.5 147.6 TABLE 6 1980 • 1981 RIVER ICE SUMMARY Leading Edge Location I RM l 75 .5 104 .5 112.8 118.8 126 .4 136.9 Stage Increase lftl 4.0 3.3 4.0 Slough or Side Channels Overt.oppud Ice Jams I RM) 126. 1 135.8 138.8 142 .3 138.8 129.7 119.3 112.8 138 .8 101.8 Ice Free M18/5 2 Date Nov . 2 Nov. 6 Nov. 18 Dec. 14 Jan. 4 May 10 ~ May 12 tv May 14 Ice Bridge Location !RMl 186.6 181.0 176.5 150.0 105.5 98.5 TABLE 7 1981 -1982 RIVER ICE SUMMARY Leading Edge Location (RMl 60 82 95 127 137 Stage Increase ( ft l Sloughs or Side Channels Overtopped Sunshine Ice Jams !RMl 1 5 3 111 2 139 130 1 0 7 Ice Free L\ ~ M1o /5 3 Date Oct. Oct. Oct. Nov. Nov. Nov. Dec . Dec. Jan. Apri I Ma~ 4 Ma~ 6 Ma~ 7 Ma~ 8 May 9 22 26 28 2 9 22 2 9 14 27 Ice Bridge Location !RM) 10 98.5 TABLE 8 1982 -1983 RIVER ICE SUMMARY Le ading Edge 1f i. ion !RMl 57 87 95 106 126 129 136 137 Stage Increase ( ft) 3 5 4 4 Sloughs or Side Channels Overtopped Alexander Rustic Wilderness Sunshine Slough 8A Slough 21 Slough 9 Slough 8A Slough 7 Ice Jams ( RMl 148.8 1115.5 1112.0 1 35.9 131 .0 120 .0 113.5 141.8 1 34.5 131.4 129 .0 122.0 119.5 113 .2 89.0 85.5 183.0 131 .5 129 .0 124 .5 122.0 120 .5 113 .0 183.0 1 3 1 .5 122 .0 11 2.5 99.5 Ice Free M18/5 4 TABLE 9 1983 -1984 RIV ER ICE SUMMA RY S l oug hs o r I ce Br i dg e Leading Ed ge St age Side Channe l s Ice Date Loca 11i o n (RMl LOCa !iiOn !RMl 1 nc r ease (ft l Ove r t opped ,J am s (R Ml Oc t . 26 9 2-3 Oct . 2 7 15 .0 No v. 1 27 31.5 Al exande r Nov . 4 40 .5 4 Nov . 5 50 .0 3 Nov. 9 6 6.0 3 Ru s t ic Wilde rne s s Nov . 10 73 .0 7 No v . 18 82 .5 Nov . 19 8 4 .5 6 s un s h i ne Nov. 26 95.5 Dec. 9 98.6 4 Dec. 22 120 .7 1 18 .0 124 .0 10 ~ Jan. 5 135 .6 1 30 .0 4 ~ 1 36.5 2 May 5 79 May 6 I ce f r ee M18/5 5 Date Oct. 27 Oct . 29 Nov. 3 Nov. 5 Nov. 10 Nov. 14 Nov . 19 Nov. 21 Dec. 15 Dec. 20 -'\ \ Ice Bridge Location !RMl 5 26 46 52 105 135 TABLE 10 1984 (FREEZEUP ONLY) Leading Edge Location IRMl 1';1 43 55 71.5 109 121 86 127 88 128 129 137 92 Stage Increase I ftl 3 4 9 Sloughs or Side Channels Overtopped Alexander Slough 8A Sunshine Ice Jams I RM) C. G~ERAL RIVER ICE BREAKUP PROCESSES Breakup is a gradual process of ice cover d!sintegration that begins with the formation of open leads through the ice cover, the melting of snow in the basin, and subsequent rising water levels that lift and fragment the ice cover. The multitude of resultant ice floes often accumulate in areas where the flow or channel configuration cannot convey the ice volume, and jams may develop. The stability of the jams is dependent on their configuration (floating or grounded) and on the presence of secondary channels that can divert water and relieve pressure from the jam. Ice jams are not common on channels with broad flood plains since a rapid rise in water level, necessary to lift and shatter the ice cover, is prevented by a greater increase in open water surface area relative to depth with the rising discharge. Jams may still develop upstream of an ice cover on this type of channel, but last only until the cover weakens and collapses. On confined channels, the river water eventually reaches a level where the ice jam is freed from its anchor and the mass of ice debris is swept downstream with a great enough force to destroy other ice blockages. This breakup drive leaves the channel essentially ice-free, with the exception of ice blocks left stranded above the high water line. 1. CHANNEL LEADS Open water leads are common in the ice cover on the Susitna. They are either formed shortly after the progression of a frazil cover, exist as a remnant in channels that were flooded but never ice-covered, or indicate reaches of mainstem that were either bypassed by the progression or by the progression ending further downstream. Examples of each case follow. 33RD1-007m -,k- tiL The upstream progression of a frazil slush ice cover is dependent on a slush accumulation of sufficient thickness to slow the water velocity upstrea m of the leading edge. Massive thicknesses can thereby be attained in reaches where high water velocities prevail, since the cover continues thickening until the rapids section is essentially drowned out. Slush ice tends to accumulate in thicker deposits near the banks, since water velocities are generally slower in shallow water. The thinnest ice cover can usually be found over the faster flowing water. The ice cover thickness necessary for continued progression only needs to be sustained long enough for the leading edge to move upstream. If progression occurs rapidly, the downstream slush ice cover may have little time to freeze solid, and leads may consequently form through the cover. After the passing of the leading edge the water stage drops as flow is lost to bank storage and slush ice is eroded from under the cover, thereby increasing the cross sectional area. The elastic properties of the ice cover cause it to sag, conforming to the configuration of the channel cross section in shall•)W water and floating over the area conveying deeper flow. If the bank slopes are low the ice cover forms an undulating surface with bulges where the slush is grounded on the bottom, and depressions corresponding to areas where the cover still floats. If the bank slopes are steeper, the cover deforms only to a critical angle before failure occil rs. In this case the ice cracks and falls into the water, creating an open lead. Leads also form in areas of high water velocity soon after an ice cover forms. On some reaches of the middle and lower river, leads have formed within hours after the slush stopped moving. This suggests an instability between the ice thickness and the water velocity which causes the unsolidified 33RD1-007m -~- 17 slush to rapidly erode away, usually only ov er a narrow portion of the channel where a critical velocity is exceeded. Leads are also prevalent in a reas that were bypassed by the slush cover progression or in areas such as upstream of Gold Creek where a continuous progression usually does not ~each. Many side channels on the lower river are dewatered prior to freezeup, only to be flooded when the mainstem water level stages upstream of the advancing ice front. The side channels carry water diverted from the mainstem, but often are too shallow to allow passage of the 2-3 feet thick ice floes . Without the entry of slush, the ~ide channels can only free z e over by direct heat loss to the atmos phere. This process requires significantly more time than a slush accumulation, and many side channels hav e never entirely frozen over in the years o f ob s ervation. Leads on the mainstem may grow an ice cover by slush accumulation if a supply can be generated in the open water area and if the resultant frazil can be captured at the downstream end of the lead. Leads often develope downstream of tributary mouths, side sloughs, or groundwater seeps. Groundwate r upwelling in sloughs results in relatively warm water being exposed to the surface. This rapidly erodes any ice cover on the sloughs (if it ever developed) and the thermal erosion conti nues downstream until the water has lost most of its heat . Thermal leads have been documented on the middle river, with their source emanating from sloughs and extending hundreds of feet into the main channel ice cover. The ice cover often perpetuates the flow of groundwater by raising the local water table so as to increase the rate of upwelling (see Section E, part 4 : Groundwater). 33RD1-007m -~- 1f 2. CANDLING Ice grows in hexagonal crystals. The crystal structure is described by .5c.c Jio.jnrtt '-•low four intrusive axes , three A-axes and one C-ax e in the basal plan of the crystal and the C-axis is perpendicular to the basal plane. An ice crystal resembles a pencil with the C-axis as the lead. When ice crystals nucleate on a body of calm water the axes are oriented randomly. The crystals grow more easily along the A-axes and crystals with horizontal C-axes tend to grow faster. Growth along the C-axis a l so occurs, but at a slower rate. Horizontal growth is usually restricted by a crowding out effect after about 1 em. During the process of crystal growth, impurities within the water are rejected, and concentrate along the crystal boundaries where they cause a weak bond between adjacent ice crystals. Because of the trapped impurities concentrated at crystal interfaces , solar radiation weakens this bond first when melting begins, causing the crystalline bonds to fail and the individual crystals to separate into "candles". Eventually the innumerable single crystals comp osing the ice cover one no longer frozen together and easily collapse (Figure 40). ., ·.· .· . -: , .... •:.· .... : .: .·. :·::·.; ;' ~ ...... :..·:..::· . 1,·•· ,.:.. ; ~·· ,. ,. .. .-..• <· ,Crys1al~raphlc axes ... : ... ..: ~" ..• Example of a-axes growth ... : •• • ..... ~ .... • r •• ;._ ;:: ~~.··; ;;:_'.:;"r~ ··.', . :.. ... ·: Example ol c·axis growth -Cr)·stallogr.tphic :~xes of snow :~nd ice cryst:~ls. If the growth m te a lo ng the ,,.,,1.:~ 1!\Co:o:J, that o~long th.: f'·a\1~. l.he cryS_I:Jis tend tow:~rd ~ pl:~tel i k~ struCiure. If C·3Jl is growth dominates. the cryst:~ls :~>SU me a columnhke :tppe:~r:~nce. The mecb:~msms t h3t c:~use dtlferences m growth r.ttes are not fully understood: air kmper:~ture pl 3 ys 3n impon:~nt role. 33RD1-007m -~- Qe; Reference: Perla & Martinelli (1976) 3. ICE JAMS The ice cover breakup on northern rivers is characterized by a weakening of the ice crystal structure by solar radiation, warm air temperatures, and an increasing river discharge initiated by snowmelt and augmented by rainfall. These factors combine to fracture the ice cover. The ice fragments drift downstream until they encounter a barrier such as a solid ice cover, an ice jam, shoals or constrictions. Ice jams are also caused by islands, sharp river bends, bridge piers and abutments. Two types of ice jams are common during breakup. The simple jam (figure 41) consists of an accumulation of ice fragments floating on the water surface and prevented from continuing downstream by a barrier, usually a stable ice cover. The dry jam (figure 42) is also formed by the accumulation of ice floes at an obstacle, but in this case the jam completely blocks the channel down to the river bottom. The water is forced to flow by percolating through the ice plug, and water levels upstream of the jam rapidly rise. This type of ice jam is responsible for the major flooding commonly associated with breakup. Water generally continues to rise upstream of the ice jam until the flow is either diverted via a side channel, flows overbank, or lifts and destroys the ice jam. The environmental effects of breakup are discussed more thoroughly in Section 7, part 1: Morphology. D. SUSITNA RIVER ICE COVER DISINTEGRATION Destruction of a river ice cover progresses from a gradual deterioration of the ice to a dramatic disintegration which is often accompanied by ice jams, flooding, and erosion. The duration of breakup is primarily dependent 33RD1-007m -~- (00 on the intensity of solar radiation, air temperature, and the amount of rain- fall. An ice cover rapidly breaks apart at high flows. Ice debris accumulates at flow constrictions and can become grounded. The final phases of breakup are characterized by long open reaches separated by massive ice jams. A large jam releasing upstream usually carries away the remaining downstream debris, leaving the river channel virtually ice free. A pre-breakup period occurs as snowmelt begins in the area, usually by early April. Snowmelt begins first at the lower elevations near the Susitnn River mouth and slowly works northward up the river. By late April, snow has usually disappeared on the river south of Talkeetna and snowmelt is beginning on the reach above the Chulitna confluence. Tributaries to the lower river have usually broken out in their lower elevations, and open water exists at their confluences with the Susitna Rivet. Increased flows from the tributaries usually erode the Susitna ice cover for considerable distances downstream from their confluence. As water levels in the river begin to rise and fluctuate with spring snowmelt and precipitation, overflow often occurs onto the ice since the rigid and impermeable ice cover fails to respond quickly enough to these changes. Standing wacer appears in sags and depressions on the ice cover. This standing water reduces the albedo, or reflectivity, of the ice surface, and open leads quickly appear in these depressions. As the water level rises and erodes the ice cover, ice becomes undercut and collapses into the leads, drifting to their downstream ends and accumulating in small ice jams. In this way, leads become steadily wider and longer. This process is especially noticeable in the reach from Talkeetna to Devil Canyon. In the wide, low-gradient river below Talkeetna, open leads occur less frequently, and 33RD1-007m -x- /61 extensive overflow of mainstem water onto the ice cover is the first indicator of rising water levels. The disintegration of an ice cover into individual fragments or floes and the drift of these floes downstream and out of the river is called the breakup drive. The natural spring breakup drive is largely associated with rapid flow increases, due to pr~cipitation and s nowmelt, that lift and fracture the ice surface. When the river discharge becomes high enough to break and move the ice sheet, the breakup drive begins. Its intensity is dependent upon meterological conditions during the pre-breakup period. For example, in 1981 a minimal snowpack and light precipitation during spring caused an insufficient increase in the flow to develop strong forces on the ice cover, and the ice tended to slowly disintegrate in place, producing few significant ice jalllllling events. Conversely, in 1982 a heavy snowpack with cool early spring temperatures prevented the ice cover from deteriorating significantly during the pre-breakup period. The ice remained strong into the later period of normal spring temperatures and rising flows, and the cover broke dramatically, producing several large ice jams. 1. TALKEETNA TO COOK INLET Solid and continuous ice covers can fragment en masse when the pressure created by the rising water level can no longer be contained. This is especially true on the lower river downstream of Talkeetna. The shattered ice cover, however, may remain in place for several days if the ice downstream remains intact. Increasing daily duration of exposure to solar radiation begins to have a marked effect in April. Existing leads lengthen as the floating ice cover melts from underneath, and once the snow has melted, solar radiation bearing 33RD1-007m -x- to 1- directly on the ice surt"ac..: causes the familiar candling process. This gradual melting seems to characterize "breakup" on the lower river (figure 43). The broad flood plain (relative to the area occupied by channels on the lower river) prevents a rapid increase in stage with rising discharges. When ice jams do occur, such as when ice debris from the middle river accumulates against a solid cover on the lower river, water spills over onto the flood plain and bypasses the congested main channel. Although erosion and damage to vegetation have been observed during breakup, these are isolated incidents and are considered insignificant when compared to damage incurred during summer floods. 2. DEVIL CANYON TO TALKEETNA By the end of April the middle Susitna River is usually laced with long, narrow open leads. Floes that have fragmented from the ice cover accumulate into small ice jams. The configuration of these small ice jams often resembles a U-or V-shaped wedge. the apex of the wedge corresponding to the highest velocities in the flow distribution. The constant pressure exerted by these wedge-shaped ice jams effectively lengthens and widens many open leads. reducing the potential for major ice jams in these areas. The actual breakup of the ice cover occurs when the discharge is high enough to break and move the ice sheet. lee jam sites generally have similar channel configurations. consisting of a broad channel with gravel islands or bars. and a narrow, deep thalweg confined along one of the banks. Sharp bends in the river are also good jam sites. The prese~ce of sloughs on a river reach may indicate the locations of frequently recurring ice jams. Many of the sloughs on the Susitna River between Curry and Devil Canyon were carved through terrace plains by some 33RD1-007m -x- lt>J extreme flood. Summer floods, although frequentl y flowing t h r ough sloughs, do not generally result in wate r levels high enough to ov ertop the river bank . During breakup, h owever , ice jams commonly cause rapid, local stage increases that continue rising until either the jam releases or the sloughs are flooded. While the Jam holds, channel capacity is greatly reduced, and flow and large amounts of ice are diverted into side channels and overbank (figure 44). The ice has tremendous erosive force, and can rapidly remove large sections of bank (figures 45 and 46). 0ld ice scars on tree trunks up to 10 feet above the bank level have been noted along side-channels. It appears that these sloughs are an indicator of frequent ic.e jams on the adj acent mainstem, influencing the stability and longevity of these jams by relieving the stage increases and subsequent water press ures acting against the ice. The following channels between Devil Canyon and Talkeetna are regularly influenced by ice-induced flooding ~uring breakup: Slough 22 Slough 21 from RM 142.2 t o RN 141 Slough 11 from RM 136.5 to RM 134.5 Side channels from RM 133.5 to 131.1 Side channels from RM 130.7 to 129.5 Slough 9 Slough 8A and 8 Slough 7 In general, the final destruction of the ice cover is accomp l ished when a series of ice jams break in s uc c ession. The resulting on-rush of water and 33RD1-007m ice debris clears the channel of any remaining obstructions and ice (figures 47 to 49). 3. SUMMARIZED HISTORICAL BREAKUP CHRONOLOGIES a. 1981 Breakup Chronology There was no significant precipitation during early spring to increase runoff in the watershed. Therefore, river discharge did not increase sufficiently to create strong forces on the ice cover and initiate breakup. Instead, the ice began to slowly disintegrate in place with long open leads developing through the length of the river. Pre-breakup conditions were observed during a reconnaissance trip on April 23. At that time, open leads were growing by ice calving off the lead perimeters. Ice floes accumulated at the downstream end. t.o floes were observed being carried underneath the ice cover. There was also little evidence of rising water level increasing pressure on the ice cover. For the next few days changes in the character of ice accumulations and water levels along the river were monitored, especially at Gold Creek. Increased overflow on top of the ice and fracturing of the ice cover indicated that the water level was steadily rising during the first week of May. Open leads continued to grow and connect. By May 3, the rise in water level and ice movement created ice jams upstream of the Parks Highway Bridge, above Curry at RM 120.5 where the channel bends sharply and begins to constrict, at RM 126.2, at RM 131.3, downstream from the Gold Creek bridge at RM 135.8, above Indian River in the vicinity of RM 139, and upstream at a constriction in the channel at RM 142 .3. 33RD1-007m On the morning of May 4th, it was o~served that most of the prev ious day's ice jams had released and new jams reformed at several different sites. The jam at RM 142 .3 had released sometime overnight, adding more ice and increasing pressure on the ice jam upstream from Indian River. A sharp bedrock outcrop along the left valley wall at RN 139 appeared to be the principal factor holding the ice. The far right channel was acting as an overflow channel, conveying flow around the ice and relieving pressure on the jam. Flow in this channel increased noticeably with the addition of ice from upstream. It also appeared that the center of the ice jam had sagged due to a change in water level. Parallel shear lines could be traced through the ice jam along the boundaries of the main channel on May 4th. This apparent drop in water level may have been related to increased flow spilling into the far right channel or possibly to release of the ice jam below Gold Creek. Duration and maximum water surface elevations resulting from the jam which keyed at the rock point near RM 135.6 could be read from the streamgage chart at Gold Creek (USGS). On tl, .. morning of May 4th remnant ice was stacked up to 6 feet high along both shores upstream and downstream of the bridge. Average thickness of the ice blocks was three feet, but much of it was candled and easily broken apart. From Gold Creek downstream, the main channel was free of ice accumulations until just below Sherman. Sometime during the night of May 3, the ice jam above Sherman released. Icc from that jam combined with upstream ice packed into the main channel throu.gh the reach just below Sherman. The ice jam key was located above a reach of shallow, turbulent flow near RM 129. 7, where the channel bed was extremely irregular. In this reach of divided flow, the left channel provided overflow relief, carrying flow around the ice so there was little effect on water levels upstream. This jam held in 33RD1-007m -,>l- Ied- place until sometime during the night of May 7th, as the channel was clear of ice on the morning of May 8th. The i~e jam downstream of Curry released during the early morning hours of May 4th. The ice sheet that previously existed at Curry broke up and accumulated in the reach at RM 119.2. Another ice jam keyed near RM 112.7 and extended upstream to the confluence with Lane Creek. On May 4th, there was a noticeable increase in overflow on the upstream ice indicating a rise in water level. Flow had also spilled into the right channel below IUt 112.7. The ice jam held until the early morning of May 6th, when the jam released. Ice floes packed into the channel extending from approximately RM 98.5 up to river Mile 101. 8. On the morning of May 8th the jam was still in place. Examination of streamgaging charts from Sunshine indicate the jam released sometime later on the 8th or early on the 9th causing the peak recorded on the Sunshine gage chart. New ice floes adding to the upstream edge of the jam at the confluence and the flood wave associated with release of the jam at Gold Creek on May 7 aggravated conditions at the confluence. Water levels were already high through this reach, with water and ice rising well up into the vegetation on both sides of the floodplain. The accumulating ice floes and rising water level created an unstable situation and the jam released on the morning of May 9. Ice cover in the lower river had broken up and been washed out several days before the ice moved down from above Talkeetna. First movement of the ice cover on the Deshka River and the lower Susitna River at the confluence was reported on the morning of ~~y 2. Sporadic movement continued throughout the day in this area. iiy early evening ice movement was also reported downstream at Susitna Station. 33RD1-007m For the next few days observers reported continued ice movement in the Susitna, rising wat:er levels, and breakup of the ice cover. (n May 3, the Deshka was 95 % ice-free, but a jam had developed at the confluence with the Susitna. The Yentna River was also ice-free except for a jam at the con- fluence with the Susitna River. By mid-day on May 5. the river at Susitna Station was reported free of ice and the jams at the Deshka/Susitna and Yentna/Susitna confluences ha d released. Through the length of the river channel, remnant ice was stranded on shore or packed into side channels with little or no flow. Over the following weeks rising water levels flushed out the rest of the ice or it melted in place. Overall, breakup during 1981 on the Susitna River was mild. Ice scarring of trees from the release of ice jams was noticed in a few locations, most dramatically in the vicinity of Whiskers Slough (lUI 101. 5). on the vegetated islands in the channel. However, no major changes in channel configuration or significant scouring of river banks due to ice were observed during the breakup process. b. 1982 Breakup Chronology Breakup was more dramatic in 1982 than in previous years, as demonstrated by extensive erosion and by damage to the Alaska Railroad tracks. Air temperatures increased during the second hal f of April, but nighttime lows still dipped below 0 C. By May 7 minimum daily temperatures averaged 4 C and ice moveruent began. Jams occurred in most of the areas described for 1981 but with greater consequences, rang i ng from scarring and denuding of vegetation to flooding and wa s hing away of railroad tics from under the tracks . In several 33RD1-007m areas below Talkeetna, massive amount s of soil were eroded from cutbanks, jeopardizing at least one residence. In the vicinity of the proposed Watana damsite, breakup effects were not as dramatic, with more ice melting in p!ace and less erosion. The jam just downstream of the mouth of Watana Creek caused total channel blockage and ice accumulations for 1 mile upstream. The only other significant jamming observed in the upper river took place near the mouth of Jay Creek. This jam backed up ice floes and impounded water for several miles. However, since the channel here is confined, no significant flooding took place. On April 26, 1982, the river below Talkeetna remained ice-covered, with many areas showing overflow. South of Bell Island, however, the ice had gone out, and the river was open. The Talkeetna River was still frozen, with open leads beginning to extend and connect. Heavy overflows were observed near Chase, indicating some localized runoff. Open leads dominated side channels and sloughs which were influenced by seeping groundwater. With the exception of high-velocity reaches, the ice cover remained stable and continuous from Sherman to Gold Creek. In high-velocity reaches, usually marked by open water leads, ice rafts were breaking away from the ice cover and drifting downstream. From Gold Creek to Indian River, the Susitna had a narrow open channel, probably a direct result of flows from Indian River, which was beginning to breakup. The ice bridges between Devil Canyon damsite and Devil Creek were beginning to show accumulations of ice floes and some jamming activity. No significant water level increases were reported. The areas of overflow previously observed above Devil Creek were showing open water. The quantity and extent of open leads were less upstream of tl-.e Fog Creek confluence, with no change in river ice above the Watana damsite. 33RD1-007m Between May 10 and May 15 the river showed little change upstream of Devil Creek except for the open leads getting wider and more numerous. The ice cover seemed to be melting in place rather than "breaking up". Ice movement began on many reaches of the Susitna River below Devil Creek. All ice bridges had disappeared except at RM 153, where an accumulation of ice floes had jammed and extended several hundred yards upstream. The river was open from Portage Creek to Gold Creek except for ice jams at RM 142 and RM 139. The ice cover remained stable about one-half mile below the Gold Creek Sridge. Ice had jammed below Sherman at RM 129 and 130, but appeared unstable and reportedly did not last long. The main channel between RM 118 and RM 120 retained its ice cover and appeared stable. Several jams of lesser consequence appeared at RM 115 to 117. At RM 107 (LRX-11), the river remained entirely frozen over. A continuous open lead had formed from Chase upstream to t:.~ mouth of Lane Creek. The confluence area was characterized by opening leads on the Susitna, the Chulitna was in final stages of breakup with no ice remaining over the channel. Hany ice blocks were stranded on sand bars and bank areas adjacent to the Chulitna. From May 12-15 a jam occurred at RM 107, flooding the railroad tracks and scouring the east bank. Although jams have occurred in generally the same location in previous years the 1982 breakup caused unusually severe erosion. The section of railroad track adjacent to the Susitna Rive1 at RM 108.5 was undermined when impounded water rose about 15 feet. The ice cover at breakup was shorefast and extended far out into the river channel, constricting the flow to a narrow deep channel against the right (west) bank. This cover was very resistant to lifting. Drifting ice blocks were up-ended upon striking this barrier, causing water inpoundment and subsequent increases in stage upstream of the jams. Witnesses claim the impounded water rose high enough to 33RD1-007m -~- //0 erode the railroad grade and wash away several ties and damage the support structure on a bridge crossing a tributary at RM 110. The jam persisted for three days and backed up ice floes for approximately 1 mile before releasing on May 15. While the j a m held, some water flowed over the ice. An extensive area on the right overbank was also flooded. This was by far the most significant damage in recent years according to railroad personnel. After the final ice drive, a river reconnaissance was made by boat on May 27 to observe the damage caused during breakup. The river reach just below Talkeetna was characterized by significant erosion of river banks on the outside of natural bends. A signi.ficant erosion problem exists just downstream of Talkeetna where a cabin situated on a 10-15 foot bank is potentially threatened by future breakup scouring of equal severity as that in 1982. At the confluence, the Susitna left bank at LRX-3 had eroded 3-4 feet, with many mature cottonwood trees now overhanging the river. At RM 99 and 100, ice blocks measuring 20-30 feet diameter had been pushed up onto the banks and sand bars. The upstream ends of vegetated islands had been scoured by ice, some being completely denuded of any vegetation for 100 feet or more inland from the bank. The left river bank had eroded 4-5 feet at RM 102. Areas most notably damaged by ice were characterized by mature (15-20 inch) cottonwoods and birch trees being knocked down and piled up against the upstream ends of islands . The Alaska Railroad had to heavily reinforce the grade by depositing large rip-rap on che river bank from RM i04 to 105 and from RM 108 to 116. particularly evident. The effects of breakup at Slough 9 (IU-1 129) were The berm at the head of the slough consisted of unconsolidated cobbles and sand, suggesting recent deposition. The ground on the islands was covered by 3-4 inches of freshly deposited silt, and ice 33RD1-007m -~ Ill blocks were observed wich.in the forest, all evidence of a major flooding event. The jam which caused this flooding was not observed. In addition to the ice jam at Chase, the Alaska Railroad reported damage to tracks at several locations along the river up to Gold Creek. The most extensively damaged section of railroad track lies between Curry and Chase where recurring ice jams are formed between RM 126.1 and RM 127.5. Additional jamming and damage was reported at Railroad Mile 260 (RM 132) following an ice jam near LRX-37. Upstream of Gold Greek between RM 141 and 142 is another overflow channel (Slough 21) which receives flood waters during breakup and high summer flows. Extensive damage to the channel and overbank vegetation ~as reported after the 1982 breakup. Scarring of 30 inch cottonwoods to heights o.= 5 feet above ground level was estimated. These trees were previously undamaged and are situated well above and away from the normal channel. c. 1983 Breakup Chronology The major streams flowing directly into the lower Susitna River were contributing substantial discharges by April 27, 1983. The ice was in varying stages of decay on these tributaries, with Kashwitna River retaining a virtually intact ice cover, and Montana Creek, Sheep Creek, and Willow Creek breaking up rapidly. By April 28, there was an open channel for most of the reach between Talkeetna and the Parks Highway Brid~:e. Observation during an aerial reconnaissance on April 29 documented a rapidly disinteg,:ating mainstem ice cover from Talkeetna down to the Montana Creek confluence. ~urther downstream, the mainstem ice cover was extensively flooded but remained intact. Above the Parks Highway Bridge the ice cover had shattered into large ice sheets in several areas. The large size of these fragments prevented the 33RD1-007m -~- ltz..., ice from flowing out. At Sunshine, an ice covered reach was flooded by about 0.5 feet of overflow, but remained intact. No ice jams had occurred. Observers at Susitna Station reported ice beginning to move downstream on May 2, with flowing ice continuing to pass for several days. Deshka River residents observed the first ice moving on May 4, with the steady ice flows ending on May 10. No significant jams were noted. This pattern indicates an upstream progression of ice breakup, which confirmed the aerial observations on the river belc~ Montana Creek. The largest ice jam observed on the lower river occurred on May 3 near the confluence with Montana Creek at RM 77. Here an extensive accumulation of drifting ice debris had failed to pass around a river bend and jammed. The Montana Creek confluence was flooded but no damage or significant impact by ice or water was noted. On April 27, 1983, daily observations and data acquisition began upstream of Talkeetna. By this time, the river had opened in some areas by the downstream progression of small ice jams. These minor ice floe accumulations remained on the water surface, often breaking down any intact ice cover obstructing their passage. As described earlier, this process is initiated in open leads which gradually become longer and wider until extensive reaches of the channel are essentially ice free. These small ice jams may be important in preventing the occurrence of larger, grounded ice jams. This was evident in 1983 when large ice jams released, sending tremendous volumes of floating ice downstream. The small jams had provided wide passages for the flowing ice, which may have jammed again i f the channel had remained constricted. On April 27, extensive channel enlargements and small ice jams were steadily pro- gressing downstream near the following locations: 33RD1-007m -~ /13 Portage Creek, RM 148.8 Jack Long Creek, RM 145.5 Slough 21, RM 142.0 Gold Creek, RM 135.9 Sherman Creek, RH 131 Curry Creek, RH 120 A large jam had also developed near Lane Creek at RH 113.5 and was apparently grounded. Flooded shore ice surrounding the jam indicated that some water had backed up. A noticeable increase in turbidity occurred on this day. On May 1, the ice jam key at Lane Creek had shifted down to RM 113.3 and was still accumulating ice floes at the upstream end. The source of the floes was limited to fragmenting shore ice. No significant accumulation would occur here until ice jams further upstream released. The ice jam near Slough 21 had increased in size and was raising the water level along the upstream edge. This backwater extended approximately 300 feet upstream. Figure 52 shows a relative stage increase at this measurement site of over 3 feet in 24 hour~. illustrating the water profile before and after this ice jam occurred. By May 2, 1983, several large ice jams had developed. The small ice jam at Gold Creek had broken through the retaining solid ice sheet, forming a continuous open channel from RM 139 near Indian River to a large ice jam at RM 134.5. The small ice jam that had been fragmenting the solid ice at the downstream end c.f an open lec;d adjact=nt to Slough 21 had progres sed down to Rl-1 141. A large jam had developed at RM 141.5, leaving an open lOater area between the two jams. The upstream ice jam was apparently created when a 33RD1-007m -~- II Y massive ice sheet s napped ·loose from shore-fast ice and slowly pivoted out into the mainstem flow, maintaining contact with the channel bottom at the downstream left bank corner. The ice sheet was approximately 300 feet in diameter and probably between 3 and 4 feet thick. The upstream end pivoted around until it contacted the right bank of the mainstem. The ice sheet was then in a very stable position, jammed against the steep t'ight bank and grounded in shallow water along a gravel island on the left bank. Several small ice jams upstream had released and were accumulating against this ice sheet, extending the jam for about one-half mile. The water level rose, with an esrimated 2,000 cfs flowing around the upstream end of the gravel island at RM 142 tnto a side channel. The entrance berm to Slough 21 at cross section H9 was also overtopped. Although the estimated discharge at Gold Creek was less than 6,000 cfs ba~ed on a staff gage reading, the normal summer flows required to breach this berm exceed 20,000 cfs. The entrance channel at cross section AS was breached, with about 150 cfs being diverted into the lower portion of Slough 21. Many ice floes also drifted through this narrow access channel and were grounded in the slough as the flow was distributed over a wider area. This illustrates the extreme water level changes caused by ice jams. By May 4, 1983, stable ice jams had developed and were gradually growing in size at the following locations between Talkeetna and Devil Canyon: 33RD1-007m Lane Creek at RM 113.2 Curry at RM 120.5 and RM 119.5 Slough 7 at RM 122 Slough 9 at RM 129 Sherman Creek at RM 131.4 -~ lt.f" Slough 11 at RM 134.5 Slough 21 at lUI 141.8 Downstream from the ice jam at Lane Creek, the ice cover was still intact, although extensively flooded. Between Lane Creek and Curry, the channel was open and ice free with the exception of some remnant shore ice. From Curry upstream to the ice jam adjacent to Slough 7 some portio~3 of the ice cover remained, but were severely decayed and disintegration seemed imminent. An intact ice cover remained from Slough 8 past Slough 9 to the ice jac at Sherman. This ice cover had many open leads and large areas of flooded snow. Between the remaining ice jams at Sherman, Slough 11 and Slough 21, the mainstem was essentially open. The jam at slough 21 was still receiving ice floes from the disintegrating ice cover above Devil Canyon. As ~ce floes accumulated against the upstream edge of the jam, the floating layer became increasingly unstable. At some critical pressure within this cover, the shear resistance betl'een floes was exceeded, resulting in a chain reaction of collisions that rapidly caused the entire cover to fail. At this point, several hundred feet of ice cover consolidated simultaneously. These consolidation phases occurred frequently during a 4 hour observation period at Slough 21 on Hay 4. The frequency was dependent on the volume of incoming ice floes. With each consolidation, a surge wave resulted. During one particular consolidation of the entire half-mile ice jam, a surge wave broke loose all the shorefast ice along the left bank and pushed it onto an adjacent gravel island. These blocks of shore ice were up to 4 feet thick and 30 feet wide. The zone affected was almost 100 feet long, with the event lasting only a few seconds. This process is essentially the same as telescoping during freezeup except 33RD1-007m -~- 1/(, that the ice is in massive rigid blocks instead of fine frazil sl\:,sh, and is thus capable of eroding substantial volumes of material in a very short time. The ease with which these ice blocks were shoved over the river bank indicates the tremendous pressures that build within major ice jams. During all of the observed consolidations at Slough 21, the large ice sheet forming the key of the jam never appeared to move or shift. The surge waves would occasionally overtop the ice sheet, sending smaller ice fragments rushing over the surface of the sheet. Towards the end of the day, the ice sheet b e~~n to deform. Solar radiation, erosion and shear stresses were rapidly det eriora: ·~~ this massive ice block. Final observations showed it to have bucklec in an undulal.._ -~ wave and fractured in places. Observers at the Gold Creek Bridge reported treme ndous volumes of ice flowing downst r eam at 6 p.m. on May 4. Taking into account the travel time, this indicates that the jam had probably released about 1 hour earlier. The ice released at Slough 21 continued downstream unobstructed until contacting the jam adjacent to Slough 11 at river mile 134. 5. The sudden influx of ice displaced the mainstem water and caused a rapid rise in water levels. The stage increased sufficiently to breach berms and flood the side channel belo~ Slough 11 adjacent to mainstem river mile 135. The jam key at this site consisted of shorefast ice constricting the mainstem flow to a narrow channel of no more than 50 feet. Large ice floes. mostly from the original jam at Gold Creek, had lodged tightly in this bottleneck. Pressures appeared to be exerted laterally against the shorefast ice which inherently is resistant to movement due to the high friction coefficient of the contacting river bed substrata. 33RD1-007m -~- 117 On May 5, few significant changes were observed in the ice jams despite warm, sunny weather and constantly increasing discharges from the tributaries to the mainstem. It was at first thought that when the ice broke at Slough 11 on May 6, it would carry away the ice jam at Sherman and start a sequence that could destroy the river ice cover potentially as far dow~river as Lane Creek. This was prevented by an event that actually increased the stability of the jam at Sherman so that it held for several more days. When the ice jam released near Slough 11 and the debris approached the jam at Sherman, it created a momentary surge of the water level. This surge broke loose huge sheets of shore ice which slowly spun out into the mainstem. One triangular ice sheet about 100 feet wide wedged tightly between two extended sheets of shore-fast ice. Ice floes continuing to accumulate against the upstream edge of this wedge exerted tremendous pressures on the obstruction. A pressure ridge rising at least 10 feet above the ice formed along the contact surfaces of the wedge. This ridge consisted of angular fragments and ice candles. The water level continued to rise as the mainstem channel fill~d with ice, which eventually extended upstream to RM 132.5. The ice jam had lengthened to over 1.5 miles. Flooding quickly occurred on the side channels adjacent to the mainstem, and some ice drifted away from the main channel. The volume of water flowing through the side channel was estimated at approximately 2,000 cfs. As the ice jam consolidated and thl! water level rose, P-ven more water was diverted through the bypass channels. This volume of diverted flow was critical to the stability and duration of the ice jam. Even though the jam increased in size, any additional hydrostatic pressure was relieved by diverting water into Lhe side channels. The entire sequence of events lasted only about 10 to 15 minutes. The water level rose over 1 foot 33RD1-007m -~ I;Y during this time span. Consolidations occurred periodically for the rest of the day but the jam key was never observed to shift. Other major ice jams key s on May 6 were located at: Watana Damsite Sherman Creek at RM 131 .5 Slough 9 at RM 129 Slough 8 near Skull Creek at RM 124.5 Slough 7 at RM 122 Curry at RM 120.5 (Photo 5.12) Lane Creek at RM 113 A small and unstable ice jam at RM 126 near Slough 8 had consolidated and the resulting surge started a rapid disintegration of the remaining ice cover down to the mouth of Slough 8 near Skull Creek. This same surge appeared to have breached the entrance berm to Slough 8. Slough 9 was flooded by a jam at RM 129 -near the upstream channel entrance. The Slough 7 ice jam received some additional floes wh ~ the jam at Slough 8 released. This resulted in a rise in water level and flooding at RM 123. At 6:30 p.m. on May 6. a moving mass of ice debris that stretched continuously from RM 136 to RM 138. with l esser concentrations extending for many more miles upstream. wa s observed approaching the Sherman ice jam. However. the consequences of this on the Sherman jam were not immediately ob- served. The condition of the floes indicated that this ice originated from above Devil Canyon. The well-rounded floes appeared to be no larger than 1 foot in diameter and were presumably shaped by the high number of collisions 33RD1-007m experienced in the turbulent rapids through Devil Canyon. Reconnais s ance of the river above Devil Canyon on Ma y 6 revealed a mainstem entirely clear of an ice cover for many miles. Stranded ice floes and fragments littered the river banks up to the confluence of Fog Creek. In several short reaches from Fog Creek upstream to Watana, the ice cover remained intact. A large jam had deve loped near the proposed Watana damsite and extended approximately 1 mile. On May 7, the following ice jams persisted: Key Location Watana Damsite, RM 184 Sherman, RM 131.5 Slough 7, RM 122 Slough 6A, RM 112.5 (formerly Lane Creek jam) Length 1 mile 3.5 miles 1 mile 2 miles Downstream from the jam at Slough 6A , the river retained an intermittent ice cover that was severely decayed and flooded. Below the Chulitna confluence, the mainstem was ice-free and no ice jams were observed. The reaches between the remaining ice jams were generally wide open. The Curry jam had released overnight and traveled all the wdy to the Lane Creek jam. Here, the sudden increase in i.ce mass shoved the entire ice jam downstream about 1 mile where it again encountered a solid but decayed ice c over . At about 10:30 p.m. on Hay 8, the ice jam at Sherman released, sending the total 3.5 miles ~f accumulated ice drifting downstream en masse at approximately 4-5 feet per second. This accumulation of ice easily removed 33RD1-007m -,li:C- /2() the remaining ice jams at· Slo-.6n 7 and Slough 6A. The l ust solid ice cover between Slough 6A at RM 11 2 and the Susitna/Chulitna c onfluence at RM 98.5 were also destroyed and replaced by one long, massive ice jam. This jam extended continuously from RM 99.5 to RM 104, and then was interrupted by an open water section up to RM 107. From this point a second ice jam extended upstream to RM 109.5. This blockage was later measured to be over 16 feet thick ia some sections, but more commonly was about 13 feet thick. These ice jams released on the night of May 9. Further observations were conducted on May 10 between lUI 109 and lUI 110. Along this rea ch, the final ice release had left accumulations of ice and debris stranded on the river banks, leaving ice floe s deep in the forest. When the ice jams released, the ice floes piled up along the margins did not move, probably due to strong frictional forces against the boulder strewn shoreline. This created a fracture line parallel to the flow vector where shear stresses were relieved. The main body of the ice jam flowed downstream, leaving stranded ice deposits with smooth vertical walls at the edge of water. These shear walls at RM 108.5 were 16 feet high. The extreme height of the water surface within the ice jam was demarcated by a differe nce in color. A dark brown layer represented the area through which water had flowed and deposited sediment in the ice pack. A white layer near the surface was free of sediment and probably was not inundated by flowing water. On May 10, the only remaining ice in the mainstem was on the upper river above Watana. Here an i c e jam about 1.5 miles long had dev eloped near Jay Creek. Ice floes continued to drift down s tream for several week s after the final ice jam at Chase released. As increasing discharges gradually raised the water level, ice floes that had been left strandP.d by ice jam surge waves were 33RDl-007m -)=t(- 12.1 carried away by the current. On May 21. the massive deposits of ice floes. fragments. slush. and debris were still intact near Whiskers Creek and probably would not be washed away until a high summer flow. The ice breakup of 1983 occurred over a longer time span than in previous years. according to historica.l. information and local residents. This was primarily due to the lack of precipitation during the critical period when the ice cover had decayed and could have been easily and quickly destroyed by a sudden. area-wide stage increase. During a year with more precipitation in late April. ice jams of greater magnitude may form and cause substantially more flooding and subsequent damage by erosion and ice scouring. Several important aspects related to 1 ce jams were observed this year and are summarized here: 1. Ice jams generally occur in areas of similar channel configuration. that is. shallow reaches with a narrow confined thalweg channel along one bank. 2. Ice jams commonly occur adjacent to side channels or s lcughs. 3. Sloughs act as bypass channels during extreme mainstem stages. often relieving the hydrostatic pressure from ice jams and controlling the water level in the main channel. Ice jam flooding probably formed the majority of the s loughs between Curry and Gold Creek. 4. Ice jams commonly create surge waves during consolidation which heave ice laterally onto the overbank. 33RD1-007m 12.7_ 5. Large ice sheets can break loose from shore-fast ice and wedge across the mainstem channel, creating extremely stable jams that generally only release when the ice decays. d. 1984 Breakup Chror.ology The 1984 lower river breakup was not marked by any unusual or dramatic events. The processes observed in the spring of 1983 were essentially repeated. As previously described, open water leads developed immediately in some a=eas where water velocities were high enough to erode the underside of the ice. The following river reaches seem to be particularly susceptible to open lead development, where an ice cover cannot remain stable for any period . ~ time unless cold air temperatures override all other influences: Below RM 9 (tidal influence) RM 62 to RM 66 RM 70.5 to RM 74 RM 78 to RM 86 RM 93 to RM 95 RM 96.5 to RM 98.5 The reach from RM 96.5 to RM 98.5 opened within 24 hours after ice cover progression from November 27 to December 8, 1983 to a width of about 100 feet at LRX-3 (RM 98.5). The open water surface area gradually diminished through the winter but wa s not observed to close in 1984 . Reach 5 also opened shortly after the initial cover developed in mid-November 1983. A secondary accumulation progressed upstream though the lead but never achieved a complete 33RD1-007m -_)>it:.- 123 closure. The remaining reaches eventually froze over by late January 1984. An ice cover that forms over open leads, by nature is less thick that the initial cover. For this reason these areas are the first to open up again with warmer air temperatures. This is the pattern observed on the lower river over the past three years. By early April 1984 the reaches listed above were again ice free over a portion of the cross section. The 1983 freezeup initiated with flows at the Sunshine gage of about 13,000 cfs. The leading edge of the ice cover arrived at Talkeetna with the discharge at Sunshine approximately 5,000 cfs. The majority of the ice cover in the downstream reaches of the lower river, formed at higher stages, is subsequently no long floating prior to breakup. Discharge generally begins to increase in late March from th~ Sunshine base flow of about 3,000 cfs. The corresponding stage increase consequently breaks up the ice cover over the upper reaches of the lower river first, since this ice developed at lower freezeup flows. If the ice is still structurally competent during the discharge increase then large ice sheets break free from the shorefast ice. These remain intact and drift downstream until they contact solid ice or become lodged across the channel. In the latter case a new barrier is created, whic h may cause i ce debris to accumulate into an ice jam. This was observed at RM 79 in 1984. This ice jam remained on the surface and no significant backwater occurred. The ice floes causing the blockage weakened after three days and dislodged. All the accumulated ice debris rushed downstream about l mile before contacting a solid ice cover. Here a new ice jam formed, which also remained on the surface with no substantial increase in stage. Historically, ice jams have been documented between RM 77 and R.M 96, but rarely do they cause much flooding since the broad flo o d plain adjacent to the ice choked channel has a large flow capacity. 33RD1-007m -~- 1~7 The lower river is us.ually ice-free by May 6. At this time the middle river usually has several very large ice jams and the upper river ice may still be intact. When the upper river ice finally disintegrates and moves downstream, it takes out the remaining middle river jams and the ice moves unrestricted through the lower river. 4. ALTERNATE SOURCES OF RIVER ICE INFORMATIOi.~ Additional inf ormation on the nature and timing of breakup of the ice cover on the Susitna River can be obtained through the National Weather Service River Forecast Center and the Alaska Railroad. Data from the Alaska Railroad The table below lists breakup dates on the Susitna River from 1975 to 1980 based on observations by personnel from the Alaska Railroad. It also d~scribes the nature of breakup a nd identifies specific problem sites. 33RD1-007m -~ ~ Year Dates 1975 May 12-15 1976 May 5-17 1977 May 16th 1978 May 8-9 1979 May 8 1980 May 12-13 Description Ice out by the 15th. no damage to track. Some minor flooding, Washouts on the Sth on tracks in the vicinity of Curry from river miles 119.8 to 122. Washouts related t o large jam extend- ing from river mile 118 .4 to 123 during the same time. Short stretch of track also lost downstream of LRX-30 at river miles 127.0 to 127.2. Heavy flooding of tracks in vicin- ity of LRX-18 and cant bank scouring tracks from LRX-13 just and (R.M. upstream. ice pushed 110. 4) to (R.M. 113.0). Ice out on the 17th. Signifi- up on LRX-18 Ice out , some bank scouring, but no signif- icant damage. Some jams and flooding, minor damage. Ice on tracks at curve approximately river mile 109.6, below LRX-13. Gentle breakup, no flooding or damage to ·racks. No flooding, ice and rockc pushed up on tracks at a few spots, no serious damage. Overall, the Railroad has never had ice problems with the track from Sherman upstream to Gold Creek. The track is farther from the main channel of 33RD1-007m -~- /)/ the Susitna and is higher above the river through that reach. However. flooding and damage to the tracks occur consistently in come reaches below Sherman. The track in the vicinity of LRX-30, where the river channel bends to the west has been damaged often. Rock rip-rap has been dumped Lo retard active bank erosion during breakup along the far left bank. Another section that appears vulnerable during breakup is that area below Curry from LRX-23 to below LRX-21. Ice jams of varying magnitude form through this reach nearly every year, causing flooding of the tracks or other damage. Farther downstream, active bank erosion is threatening the tracks in the vicinity of LRX-20. Rip-rap has been dumped to prevent further erosion. Rip-rap has also been dumped through the entire reach from LRX-18 to below LRX-13 along the left bank. This reach suffers nearly every year from flooding, ice on the tracks and scouring of the banks. The sharp bend in the riv er channel between LRX-09 and LRX-10 has also been the site of ice jams several times in the past. Water flooded the tr~cks and ice was pushed up on top of the banks, with some scouring occuring. E. ICE EFFECTS ON THE ENVIRONMENT River bed and bank material may be displaced directly by moving ice or by flow conditions altered by ice phenomena (Newbury. 1968). The rising ice cover and accompanying shifts of the solid ice fragments during breakup removes vegetation from the banks to a level not attained by high summer flows. The middle river vegetation trim line corresponds to the late winter ice profile rather than the peak annual flood. However, the lower river is not affected severely by breakup, and erosion on this reach is primarily associated with summer floods. 33RD1-007m -~- 12.7 Sloughs and side channels are usually overtopped during the open water season, but severe erosion generally takes place only when flows are accompanied by solid ice fragments during breakup or during extreme floods. Some sloughs may also be overtopped at freezeup. This usually leads to development of snow ice and in some cases short-term deposition of anchor ice. A rise in groundwater levels corresponding to the staging maj .nstem flow increases the seepage rate in the sloughs. This results in a storage heat flux to the slough water, causing anchor ice to melt and opening leads open through an existing ice cover. 1. MORPHOLOGY AND VEGETATION Breakup ice processes historically have been a major environmental force on the Susitna River, affecting channel morphology, vegetation, and aquatic and terrestrial habitats. The impacts vary along the length of the river. Ice processes appear to be a major factor controlling morphology of the river between the Chulitna confluence and Portage Creek. Areas with frequent jams have numerous side channels and sloughs. The size and configuration of existing sloughs appear to be dependent on the frequency of ice jamming in the adjacent mainstem. Major breakup ice events probably formed sloughs. The size and configuration of existing sloughs is dependent on the frequency of ice jamming in the adjacent mainstem, Ice floes can easily move the bed material, substantially modifying the elevation of entrance berms to the sloughs. In May, 1983, a surge wave overtopped a shallow gravel bar that isolated a side channel near Gold Creek. The surge also created enough lifting force to shift large ice floes. These floes barely floated but were carried into the side channel by the onrush of water, dragging against the bot toe for several 33RD1-007m hundred feet and scouring troughs in the bed material. This same process also enlarges the sloughs. When extreme staging occurs in the mainstem and a large volume of water spills over the berms, then ice floes drift into the side channel. These ice floes scour the banks and move bed material, expanding the slough perimeter. This scouring action by ice can therefore drastically alter the aquatic habitat. rhe erosive force of ice affects vegetation along the river. The frequency of major ice jam events is often indicated by the age or condition of vegetation on the upstream end of islands in the mainstem. Islands that are annually subjected to large jams usually show a stand of ice-scarred ma- ture trees ending abruptly at a steep and often undercut bank. A stand of young trees occupying t he upstream end of islands probably represents second generation growth after a major ice jam event destroyed the original vegetation. Vegetation is often prevented from Te-establishing by ice jams that repeatedly override some islands. Ice processes have several impacts on aquatic habitat. The sloughs may have snow ice up to 5-6 feet thick. Diversion of flow and ice into the sloughs may cause large changes in channel morphology. Large amounts cf silt may be deposited in the system at breakup. The s ilt deposits move downstream during summer high flows, covering good spawnin g habitat. Ice processes do not appear to play as important a role in the morphology of the Susitna River below the Chulitna confluence. This river reach regularly e xperiences extensive flooding during summer storms. These summer floods seem to have significantly more effect on the riverine environment than do ice processes (R&M Consultants, Inc. 1982a, 1982b). This reach is characterized by a broad, multichannel configuration with distances between vegetated banks of ten exceeding 1 mile. The thalwe g is represented by a 33RD1-007m ~- 12( relatively deep meandering-channel that usually occupies less than 20 percent of the total bank-to-bank width. At low winter flows the thalweg is bordered by an expanse of sand and gravel (R&M Consultants, Inc. 1982b ). Although ice cover prog ression frequently increases the stage about 2-4 feet above normal Octob er water levels, no significant overbank flooding takes place, although some sloughs a~d the mouths of some tributaries do receive some overflow. The ice cover below Talkeetna is usually confined to the thalweg, and surface profiles rarely approach the vegetation trim line along the banks. 2. SEDUIENT TRANSPORT The transportation of sediments decreases substantially between freezeup and breakup, primarily because of the elimination of glacial sediment input and the reduction in flows. The glaciers contribute the maj or ity of the suspended sediment by volume to the Susitna. Other factors that significantly influence the sediment regime are turb u lence, velocity, and discharge, all of which are greatly reduced during the winter. However, t he advent of frazil ice in October provides a variety of processes by which particles, both in suspension and saltation, c an be moved. Ice nucleation, suspended sediment fi4tration, an~ entrainment of larger particles in anchor ice are some of the processes described in t~lis section. The dramatic nature of breakup often introduces sediment to the flow by re-entraining particles that had settled to the bottom. This ice event is characteristically accompanied by ice scouring and erosion during extreme stages. Ice jam induced flooding commonly flushes sediments from side channels and sloughs. Ice blocks are heaved on t o riv er banks or scraped against unconsolidated depositional sediments, remov ing soils which may become entrained in the turbulent flow and carried downstream. 33RL1-007m Laboratory investigations have determined that ice readily nucleates around supercoo l e d particles. These particles may be in the form of organic detritus, soils, or even water droplets (Osterkamp, 1978). Prior to freezeup, the Susitna River abounds in clay-size sediment particles wldch may form the nucleus of frazil ice crystals. The first occurrence of frazil is generally also marked by a reduction in turbidity. Visual observations seem to indicate that the decrease in turbidity is proportional to the increase in frazil ice discharge. It is not certain whether this occurs because of the nucleation process or by filtration. As described in previous sections, frazil ice crystals t ~ to flocculate into clusters and adhere together as well as to other objects. When frazil floccules agglomerate they form loosely packed slush (Newbury, 1978). Water is able to pass through this slush but suspended sediments are filtered out. Sediment particles are therefore entrained in the accumulating ice pack. Ice shavings from bore holes drilled through the ice often contain silt-size particles of sediment. Early flows of slush ice accumulate on the lower river below Susitna Station and progressively advance upstream. These early slush floes possibly filter high sediment concentrations in October and retain them in suspension all winter. When frazil ice collects on rocks lying on the channel bottom, it is re- ferred to as anchor ice (Hichel, 1971). Anchor ice is usually a temporary featl.re, commonly forming at night when air temperatures are coldest, and releasing during the day. Like slush ice, anchor ice is porous and often has a dark brown color from high sediment concentrations. These sediment particles were either once suspended and subsequently filtered out of the water, or else were transported by saltation until they adhered 0:1 contc.•:t with the frazil. When anchor ice breaks loose from the bottom, it generally 33RD1-007m -~- /J/ lacks the structural competence to float any particles larger than gravel. Frazil slush is therefore an effective medium for sediment transport during freezeup, whether the process is nucleation, filtration or entrapment. An ice cover advancing upstream can cause a l ocal rise in water levels, often flooding previously dry side channels and sloughs. Substantial volumes of slush ice may accompany this flooding. On December 15, 1982, Sloughs 8 and 8A were flooded when the ice pack increased in thickness on the mainstem immediately adjacent to the slough entrance. These sloughs r eceived a disproportionate volume of slush ice relative to water volume since the water breaching the berm constituted only the very top layer of mainstem flow. The majority of slush ice floats near the water surface despite only minimal buoyancy. The flow spilling over the slough berms therefore carried a high concentration of ice. This slush ice and entrained sediment rapidly accumulated into an ice cover that progressed up the entire length of Slough 8A. Side channels and sloughs that were breached during freezeup and filled with slush ice are not necessarily flooded during breakup. If these sloughs are not inundated then the ice cover begins to deteriorate in place. The entrained sediment consolidates in a layec on the ice surface and effectively reduces the albedo, further increasing the melt rate. What finally remains is a layer of fine silt up to ~-inch thick covering the channel bottom and shoreline. If berms are breached during breakup, t hen ice fragments from the main channel may be washed into the slough and become strai].ded in the shallow reach. These ice floes then simply melt in place, depositing their sediment load in the side channel. This occurred in Hay 1983 when the "AS" access channel to Slough 21 flooded during a major mainstem ice jam. 33RD1-007m -~ /)~ Shore-fast ice along the perimeter of an ice jam is usually not floating. When debris accumulating behind a jam consolidates, the resulting surge wave may provide the critical lifting force to suddenly shift the border ice. This occurred near Slough 21 on Hay 4, 1983. Tons of ice were shoved onto a gravel island, entraining particles up to boulder-size and producing ridges of cobbles, gravels and organics. By this process of laterally shoving substrate material, ice can build up or destroy considerable berms and change the size of gravel bars near ice jam locations. When the lateral pressure exerted by ice is comp o unded by simultaneous downstream movement such as during an ice jam release, the effec ts on the riv er banks can be devastating. Many cubic feet of bank material were scoured away in minutes when massive jams released near Slough 21, Sherman, and Chase in t-fay 1983 (figure SO). An interesting phenomeno n observed during breakup was the e f fective filtering capability of ice j ams and individual ice blocks. Sediment-laden water flows through the man y channels and interstices between the fragment s in an ice jam. The s e interstic e s are usually filled with porous slush which removes suspended s e diments from the water. Ice jams can concentrate sediment in this manner and often become v ery dark in color . As discus s ed, Susitna Riv er ice generally consists of alternating layers of rigid, impermeable clear ice and porous, loosely packed, rounded crystals of metamorphosed fra z il ice. Water can percolate through the permeable lay ers, which strain out suspended sediment particles. This sediment becomes concentrated when the ice melts and is ei t her re-entrained into suspension or deposited on the river bank if the i ce floes were stranded (figure 51). 33RDl-007m -~ !Jj 3. SLOT JGH OVERTOPPING The sloughs and side channels of the middle and lower Susitna River convey mainstem flow for most of the open water season and generally dewater prior to freezeup. Overtopping occurs when mainstem water levels exceed the threshold elevation at the slough entrances. This typically occurs during floods initiated by snowmelt and augmented by rainfall and glacial melt, and often during freezeup when mainstem water is staged to a higher level before the advancing ice front. Breakup ice jams dramatically incre;:..s e the water level in a relatively short reach immediately upstream of the obstruction. Water levels can easily exceed those associated with even extreme summer flooding. When this occurs, the river banks can b e c vertopped as the \'later seeks an alternate route to bypass the jam. The sloughs on the middle river are overtopped at different river discharges . The bed material is often composed of large cobbles and gravels which may be tra nsported during the peak annual flood but not by the typi~Gl overtopping flow which occurs at the usual mainstem discharges of between 25,000 cfs and 40,000 cfs. This is evident also by the numerous beaver dams and ponds located in some sloughs indicating that flows do not reach extrem2 velocities. During breakup, the activity of solid ice floes moving bank material has been well documented and little doubt remains that i c e can move great volumes of material, even up to the largest boulders (Newbury, 1968). Entranc e berms can be bu1.lt up by ice floes shoved laterally from the main channel, literally bulldozing the substrate in front of them. Conversely these same berms can be removed if ice floes override the channel entrances. The typical sequence of breakup overtopping occurs as follows. An ice jam develops on the mainstem. This starts as a simple floating jam again s t a solid ice cover obstruction but evo lves into a dry dam when ice accumulations 33RD1-007m ~ IJ7 cause a compression of the debris, increasing the thickne ss and ground i n g the ice on the channel bottom. The jam n o w complete l y b locks the riv er chann el. Water levels rapidl y rise, spilling into s ide channels o r fl owi ng overb ank. If the flow can be conveyed through side channels, then the mainstem wa ·:er level stops rising. If the diverted flow is shallow, then ice f l o es can ·1ot leave the ice-cho ked mainstem bef o re beco ming grounded o n the chan nel bad. Subsequent surges due to ice d e bris compressions may shove the ice blocks laterally, pushing bank material il,to low mounds or berms . Compres-;ions also re s ult in increased water lev els, als o c a u sin g trore water to f low o ut of the main c hannel. If c om pressio ns cont inue, then a water level may be reache d that is deep eno ugh to drive the ic ·~ floes out of the main channel and into t he diversio n c hannel. The solid floes , which may have large dimensions and we i gh thou sands of pounds, are carried through the side channel, bumping and scrapin g over the entrance berm, impacting the ban<s, scouring the bed until finally b e coming stran ded . Severe modific atio n s o f :he side channel morpho logy o ccur i n th is manner . F igure 52 s h ow s an e x tn!me example of this process. A W4 SSive ice j am on the ma ins tem cau s ed water to flow into a side channel, wl1ich also became b l o cked and c ause J flo o ding of the terrac e along the left bank. I c e blo cks and wa ter ero d e d o ut what i s now calle d Slough 11. Overtopping may occur at some sloughs during freezeup as well. Thet:e event s a Le considerably less dramatic and usually do not affe c t the c hannel morpho logy since s o lid ice is ab s ent. Main s t e m staging c auses the wa ter lev el to rise and eventually s eep through the snow cover lying over the entranct! berm. During some years this results only in the formation o f s now i c e. If the water c ontinues t o r ise, t he s now rapidly ero des away a nd open wa ter flo ws thro ugh the side channel. If o vertopping continues , the entire s lough ma y b e 33RD1-007m -~- /]J- flooded. Water cont inues flowing until the mainstec recedes. The side channel flow may be accompanied by slush ice. This, however, has no detrimental influence on the channel since the ice lacks cohesion and usually flows around obstacles . The presence of shallow open water in the slough may resul t in the formation of anchor ice on the channel bed if air temperatures are cold enough (i.e. less than -10 C). Anchor ice was reported in Slough SA during the 198 2 freezeup when an estimated 140 cfs flowed through that channel. 4. GROUNDWATER Most sloughs on the middle river have channel bed elevations low enough to intercept the local groundwater table . This water seeps into the channel and prevents any permanent ice development . Groundwater temperatures are generally warm (2-3 C), melting existing ic~ und causing leads to fore in the sloughs . The se leads usually remain all winter, often extending beyond the slough mouth and into the main channel ice cover. Observations and recordings of groundwater levels in wells ad jacent to the mainstem indicate that water goes into or out of ground storage depending on mainst em s tage. At high stages, when the mainstem water level is higher than the l ocal water table (i .e. during flood s), water percolates thro ugh the substrate and groundwater levels gradually rise . Conversely, when mainstem stages are low, groundwater seeps from the banks until it reaches a level near the mains tem water level. Groundwater observation wells located at Slough 9 indicated a rising water table as the main channel ice cover a pproached in both 1982 and 1983 . This coincided with a noticeable increase in the s urface a re a of isolated pools in the slough system. Snow surrounding these pool s was inundated and 33RD1-007m -~- /]' subsequently froze into snow ice. Slough banks began t o seep ground water and the flow in the sloughs inc r eased slightly. The inc rease in groundwater flow rapid ly mel ted through ice covers over the pools and prevented further ice from developing. The thermal influence of the groundwater is overwhelmed if the slough is overtopped , but since overtopping during freezeup i s a rela tiv ely short term e vent, the groundwater provides heat when overtopping ceases. This heat influx continues all winter. The sloughs e ffec ted by groundwater seeps during the winter include: Slough 21 Slough SA Slough 11 Slough 7 Slough 10 Slough 6 Slough 9 Lane Sl ough t.lhi sk ers Slough 33 RD1 -007m -ya~ 13? V. WITH PR I)JEC T STUDIES A. METHOD OLOGY AND SCO PE Wint e r with-project studies included simulations of reservoir and river temperatures and ice processes, groundwater, sediment, and channel s tabilit y studies. The re se rvoir and river tempera ture and ice simula tions provided the necessary information on flows, water levels, and temperatures t o make the other studies. The seque n ce of these simulations is given be low. 1. Reservoir(s) o peration simulatio n s , in whi ch the power and f l ood release s are determined, on a weekly bas i s, thro ugho ut the y e ar. 2. Reservoir(s) temp e rature /ice s imulations utilize meteorologic and hydrologic data to determine outflow temperature s and winter reservoir ice cover on a daily basis. 3. In s tream temperature simulations pro ceed f r om the release r a t e and temperature at the dam(s ) a nd produce t empe rature profiles in the o pen- water o n a weekly basis. 4. I nstream ice s i mula tions begin a t th e 0 C isotherm dete rmined from the ins tream t e mp e r a tu re mod e l a nd predic t the ice r egimr d own s tream. Additionally , a mail su rvey of experience in o perating hyd r o electric project s in cold regions wa s undertaken. Letters were sen t t o operators of h y droelect ric proj ects and concerned environ~ental agencies to asce rtain 33RD2-007p /JcY' measures employed at the projects to minimize potential problems. Concerns addressed by the survey were: 1. project operation to control ice formation in the river downstream, 2. project operation to control ice cover cracking in the reservoir, 3 . effects on t errestrial animals of exposure to ice covered reservoir banks, 4. problems of bank erosion associa ted with ice in the reservoir and river downstream. 1. RESERVOIR ICE MODELLING With-project reservoir temperature and ice cover characteristics are simulated with the Dynamic Reservoir Simulation Model (DYRESM), developed by Imberger, Patterson, and others (imberger and Pa tt e rson, 1981). The model has b P.en calibra ted by Harza-Ebasco u sing Eklutna Lake data from 1982-1 984 (Harza-Ebasco Susitna Joint Venture, 198 4a). Acres used the same model for the License Application studies, with a limi ted calibration study on Eklutna. In order to include effects of an ice cover, an ice subroutine developed by Patterson and Hamblin for Canadian lakes has been incorporated in the model. :ce wc~ld begin t o f o rm on th e reservoi r when the surface temperature has cooled to 0 C and dail y average air temperatures are below 0 C. At Watana and De vil Ca nyon thi s normally occ ur s in November. On ce the river temperature at the upper end of the reservoir ha s dropped to 0 C, an estimate is made for fr azil ice input also . The frazil input is esti~a ted to be 5% of the ~~ater fl ow , and is assumed to b e presen t beginning No v ember l and decreasing t<, 0% 33RD2-007p /Ji on December 31. Once an ice cover has formed on the reservoir, snow is also allowed to accumulate based on actual precipitation records. The snow acts as insulation, slowing the growth of ice. The ice cover generally grows throughout the winter in accordance with daily air temperature and precipitation data. In the spring, solar radiation begins to melt the cover before air temperatures exceed 0 C. Later, the warm air temperature in co~bination with the increasing solar radiation tends to melt the cover rapidly, usually in less than 1 month. 2. INSTREAM TEMPERATURE MODELLHIG With-project stream te~peratures are modeled by SNTEMP, the Stream Network Temperature Simulation Model, developed by the U. S. Fish and Wildlife Service. SNTEMP predicts water temperatures at selected points in a river network. The model uses discharge and temperature output from the reservoir simulation, DYRESM, and routes the flow downstream utilizing meteorologic, hydrologic, and stream geometry data to compute heat flux relationships and water temperature along the river. SNTEMP has been documented by the Arctic Environmental Infornation and Data Center (AEIDC, 1983). SNTEMP operates between Watana o~ Devil Canyon downstrean to Sunshine Station at the Parks Highway bridge. Flow and thermal input froo tributaries between the dam and Sunshine Station are included. Topographic shading is also an important feature of the Susitna River which is included in the mode l . Generally, the SNTE~W output which is utilized by !CECAL is the location of the 0 C river isotherm as a function of time during the winter. From this point, !CECAL computes ice producti~n in the river downstream to the ice front. However, in the spring, the 0 C isothenn as cot:lputed tj SNTEHP is 3JRD2-007p downstream of the ice front, which indicates meltout is in progress. In this case, the water temperature at the front is interpolated from the SNTEMP results in order to estimate the temperature at the ice front. With this value , the melting of th ~ front and under-ice water temperature decay are computed by ICECAL. 3. INSTREAM ICE MODELLING Preliminary river ice simulations with the ICESIM model were undertaken by Acres American, Inc. (Acres American, Inc., 1983) in preparation of the FERC License Application. Harza-Ebasco Susitna Joint Venture (1984b) documented the river ice model ICECAL and its calibration to the Middle Susitna River for use in the present study IC EC AL was used to generate the river ice simulationR presented in this report . The model provides a d a ily summary of hydraulic, temperature and ice conditions throughout the study reach for the period November through April. The particular hydraulic and ice operations performed by the ICECAL model include the following: a. Hydraulic profiles are computed daily for the study reach. b. The 0 C isotherm is located based on SNTEMP results. c. Frazil ice produc tion is computed between the 0 C isotherm and the ice front. If the 0 C iso therm is determined to be downstream of the i ce front, meltout of the ice f r o nt a nd und er ice tempe r acu re decay are computed. 33RD2-007p d. Shore ice (border ice) growth proceeding from shore is computed in the reach between the 0 C isotherm and the ice front. e. As frazil ice coalesces into loosely-consolidated slush floes, hydraulic conditions at the ice front are analyzed to determine whether the floes accumulate at the upstream (leading) edge of the ice cover. If not, the ice is swept under the ice cover and deposited on the underside of the ice cover downstream, in accordance with the under-ice velocity profile. f. Computations are made of the slush and solid ice component thickness of the river ice cover. g. Meltout of the ice cover is simulated by computing the melting of the c ove r and retreat of the ice front ~hen warm water, above OC, reaches the ice front. Input data utilized by IC ECAL include the following: a. River cross-sectional geometry and bed roughne ss for the s tudy reac h. b. Weather co nditions (daily air t em perature and wind ve l oc ity) for the stud y reach. c. Water inflow hydrograph at up s tre am boundary of s tudy reach. 33RD2-007p d. Daily frazil ice discharges a t upstream boundary of study reac h, or e. Location of the 0 C isotherm, or water temperature at the ice f r ont. Cal ibration of !CECAL was carried ou t using observa tions of natural ice processes during 1982-83 and 19 83-84 (R&M Consultants , rnc., 19 84 a, b). Computer simulations of natural a nd with-project ice processes were made for the winters of 1971-7 2 , 1976-77, 1981-82, and 1983-83. The winter of 1971-72 is considered to be a repre sentative co ld winter, whereas, the winter of 1981-82 is average in te~perature. The winter of 1982-83 represents a warm winter and 1976-77 r ep resents a very warm winter. Climatic data fo r these yea r s are summa rized on figures 53 and 54 . Natural streamflows at Gold Creek for these winter s are shown in figure 55. The basic sioulations were run using the Case C release cons traint, shown on figure 56. Other simula tions have been made using the Case E-VI release constraint, al3o sho~~ on figure 56. The mul ti -level intakes at Wa tana and Devil Canyon were initially oper a ted in an attemp t to match the natural flow temperature . Other simulations were made as s uming release of warmest water year-round, end additional low power intakes at Watana. These variations were made to test the sensitivity of release temperatures and ice regime to operating policy and intake design. The simula tions are as follows: a. Basic Runs Case C Flow Coustraints 33RD2-007p /'( 3 Inflow-Matching Release Policy Intake Geometry per License Application Watana + Year Watana Only Devil Canyon Watana Simulated Pre-Project 1996 2001 2002 2020 Filling (Cold) 1971-72 1 5 9. 12. 16. (Very Warm) 1976-77 2 6 13. (Average) 1981-82 3 7 10. 14. 19. (2nd winter) (Warm) 1982-83 4 8 11. 15. 17. 18. (1st winter) b. Attempt to Provide Warmer Releases Throughout the Winter. (i) Repeat run 5, assuming 4 C releases throughout the w~nter. (ii) Repeat run 9, using warmest water available, year-round. (iii) Repeat run 9, using warmest water available year-round, with additional low intake at El.l800, (approach channel El.l770). (iv) Repeat run 9, using warmest water available year-round, with additional low intake at El . 1800, (approach channel El. 1500). (v) Repeat run 9, with additional low intake at El. 1636 (approach channel El. 1470). c. Effects of Revised Flow Constrain ts (Case E-VI) 33RD2-007p j'tY (i) Repeat run 10, with Case E-VI flow constraint. (ii) Repeat run 14, with Case E-VI flow constraint. B. SIMULATION RESULTS 1. RESERVOIR ICE Figures 57 and 58 show typical results of the ice cover simulations on Watana and Devil Canyon reservoirs, respectively, based on DYRESH results. On Watana, the ice cover begins to form in November and gradually thickens until March or April. The cover begins to melt in April or May and the meltout is generally more rapid than the fo~ation. The ice cover is generally melted sometime in May. In a cold winter , such as 1971-72, the Watana cover is not completely melted until early June. The timing for the cover on Devil Canyon Reservoir is very similar to Watana Reservoir. However, the ~o ver at Devil Canyon is generally thinner than at Watana. For ins tance, in a cold winter s uch as 1971 -72, the maximum computed thickness at lJatana was about 60 inches, compared to about 48 inches at Devil Canyon. The thinner cover at Devil Canyo n is because of the above 0 C releas es from Watana into the Devil Canyon reservoir during the winter, compared to the 0 C inflow to Watana reservo ir in the winter. The reservoir cover thickn esses shown on figures 57 and 58 are ba sed on inflow matching" temperature control. That is, the intake port which mo s t closely matches the reservoir inflow temperature is operated. Also , the releases are constrained by Case C. Other o p e rating rule curves have been studied including: 33RD2-007p a) Warmest available water year-round, b) Additional low intakes at Watana, and c) Lase E-VI release constraint. All of the above operations result in slight differences in the ice thickness, but the timing of formation and meltout are not substantially different. Figures 59, 60, and 61 show the minimum, mean and maximum Watana Reservoir water levels for operations in 2001, 2002 and 2020. Reservoir levels in 1996 would be similar to 2001. Figures 62 and 63 show the same information for Devil Canyon Reservoir in 2002 and 2020. Watana Reservoir is normally be drawn down between 40 and 90 feet between October and May. Between mid-April and mid-May the water level is fairly stable. Beginning in mid-to late Hay, the reservoir water surface begins to rise by about !0 to 30 feet a month. During the same periods the Devil Canyon water levels remains relatively constant and is not drawn down. a. Watana Operation At Watana, the reservoir would generally freeze over while the water level is near its maximum for the year. Ice would first begin to form in shallow areas and along the edges of the reservoir. This would consist of a very thin sheet. The ice cover would progress toward the reservoir center until it covered the entire surface. It is not uncommon for an overnight cold spell to result in a thin sheet of ice over large portions of the reservoir surface. If solar radiation and winds do not melt or break this sheet up, it is not unlikely that much of the reservoir surface could become ice covered almost simultaneously. Once an ice cover forms, it would thicken to a few 33RD2-007p I~ inches very rapidly. The thickening rate would be reduced with time because of the insulation effect of the ice and any snow cover. The ice thickness would be relatively uniform over the entire reservoir surface but a little thicker at the edges where the cover first forms. The reservoir ice surface would be relatively smooth. Strong winds may result in delays to ice cover formation by mixing reservoir water and keeping the surface water warm. This wind effect is accounted for in the DYRESM simulations. As the water level drops, ice sheets and blocks would be deposited on the banks. These would generally appear as in figure 64 and would conform to the topographf of the banks. Local discontinuities in the ice on the banks would occur due to rocks, stumps, changes in slope and other morphological and topographical features on the banks. Shelving of ice which would result in large discontinuities in the bank ice is not expected. The reservoir water level would begin to draw down in October, prior to freezeup of the reservoir surface. The draw down would be continuou ~, at a rate of 0 . 25 to 0. 5 feet per day. This would preclude the possibility that the ice might freeze into the banks if the water level were stable during freezing and form an ice shelf when the water level draws down further. The thickness of the ice would gradually increase as the water level dropped. Cracks would separate the ice sheets on the banks. The ice would generally be covered with snow. The ice cover would begin to melt in early April and be completely melted between early May and early June . In general, the melting of the ice cover would coincide with the lowest extent of the reservoir drawdown and reservoir levels during this period would be relatively stable . By the time reservoir water levels begin to rise, the reservoir ice cover would normally be nearly completely melted. 33RD2-007p lee on the reservoir banks would melt from solar and atmospheric radiation as the ice on the reservoir. Heat from water just below the reservoir surface and warm inflows would melt reservoir ice slightly faster than ice which was deposited on the banks. lee deposited in shaded areas on the banks and along the south shoreline would remain longer. Ice deposited on the northern shore and exposed to direct solar radiation would melt faster t:1an along the southern s.1ore . As the water level rises, ice which remains on the banks would be refloated into the reservoir where it would melt rapidly due to the warm reservoir water. lee on the reservoir surface in contact with the underlying soil and along the banks may melt sooner than the ice in the center of the reservoir. This would free the main body of the reservoir ice cover to move with winds. The ice on the reservoir banks near the maximum water level would be thinner than the ice on the banks near the minimum water level, and would thus melt sooner. For this reason, much of the ice on the reservoir banks may,in fact, melt sooner then the reservoir ice cover. Soon after the reservoir water surface is exposed to direct radiation and winds the upper layers of the water would overturn. The reservoir surface would be at 4 C or higher, and any ice refloated from the banks would be melted rapidly. During the meltout period, winds would generally be from the north and west and would tend to blow refloated ice blocks and broken ice on the reservoir toward the south and east shores. Some localized accumulation of ice may occur along the banks. However, the continuing rise of the reservoir- during this period should keep most ice afloat. Some erosion of reservoir shoreline material may be expected to occur in the spring as the mobi l e reservoir ice cover or floating ice blocks comes in 33RD2-007p contact with the banks. This is expected to result in localized increases in suspended sediment concentrations near the shorelines, but is not expected t o cause slope instability. Travel across the reservoir by animals may be impeded by the ice on the banks during the winter . However, the e:~perience survey indicates this is not a problem. Travel is more likely to be difficult during the period of initial ice formation when the ice thickness is not sufficient to hold an animal and during the melt out period when the ice is deteriorating. Thus there would be periods in November and May of every year when travel across the reservoir would not be assured. At other times, animals which could swim or walk across the reservoir would not be impeded. Ice blocks or sheets may be deposited in tributary mouth areas along the reservoir shore. Where the tributary walls are steep, this may shield the ice from direct solar radiation and melting of the ice before it is refloated by the rising water levels, may be delayed. However, it is not believed that this would prevent passage by fish through these areas since normal tributary flows during the melt out would be above free~ing, would tend to melt this ice and provide adequate passage conditions. b. Watana Filling Watana reservoir would be filled over a three year period as the dam is raised from elevation (El.) 1700 to El. 2205. The winter releases during the two winters of impounding would be the same as natural. That is, the water levels would be held constant in the reservoir. During the first winter of filling the dam crest wouid be at approximately El. 1950 and the water level would be stable at a level near El. 1880. An ice cover would form on the reservoir in much the s a me manner as during operation. The thickness would be 33RD2-007p similar and the melt out would occur a t about the same date. During the second winter of filling the dam crest would be at approximately El 2130 and the water level would be stable at a level between El. 20 50 and El. 2100 . Ice cover formation, thickness and melt out would all be similar to the first winter of filling. Since there would be no drawdown during the win ters of filling there would be no ice blocks on the banks. Passage over the reservoir would not be assured during the initial freeze-up and the melt out periods. c. Devil Canyon Operation The Devil Canyon water level would be relatively constant at its maximum level throughout the winter . In dry years, only, there would be period during freezeup, when the water level would be rising from its minimum level to El. 1455. During this period, ice along the reservoir shoreline would be somewhat thinner than ice in the center of the reservoir. This could make travel across the ice cover hazardous. In most years the water lev el would be at its maximum level by early December. In dry years, the water level would be at its maximum level in January. So, generally, between December and melt out in May, travel across the reservoir would be safe. During initial freezeup and meltout, passage would not be assured. Because the water level would not be drawn down there would be no ice blocks on the banks. Since the Devil Canyon winter water level would be relatively stable there is a possibility of cover expansion and ice push induced erosio n of bank materials and damage to vegetation r oo ted in the banks (Gatto, 1982). This could lead to some localized increases in suspended sediment concentrations near the eroded areas. The potential for this is probably greater upstream of the canyon, between river mile 162 and the Watana Da~ (RM 184). 33RD2-007p 2. INSTREAM TEMPERATURE SIMULATIONS a. The 0 C Isotherm Position The position of the ice cover leading edge, or ice front, is ultimately related to the position of the 0 C isotherm in the river. The position of the 0 C isotherm, for both natural and with-project conditions, has been simulated by the SNTEMP model, utilizing data from variouti climatic years. Under natural conditions the 0 C isotherm in the river progress es downriver in early winter. Climatic conditions in early winter normally cause the upper river basin to freeze earliest, and freezing temperatures then progress southward and downriver . With-project, however, warm waters above 0 C would be released from the dams year-round , and the water would slowly cool as it is exposed to below-freezing air temperatures while it moves downriver . Therefore, as air temperatures cool in early winter, the 0 C isotherm i n the river would progress upriver. The position of the 0 C isotherm is dependent upon climatic conditions, flow rate, and reservoir temperature releas es. Given a steady flow and temperature release scenario, the 0 C isotherm would progress upstream during the freezeup season at a rate determined by climatic conditions. If the flow rate were decreased or the release tempera ture decreased, the progression upstream of the 0 C isotherm would be quicker.ed . If the flow rate were increased or the release temperature increased, the 0 C isotherm progression would be slowed. Generally, with the project in place, the 0 C isotherm would advance upriver during freez e up, and during meltout/breakup it would retreat downriver. However, with-project simulations of the progress of the 0 C isotherm show that the progression rate fluctuates over a relatively wide 33RD2-007p I~ range from week-to-week, including reversals in direction from advancing upriver to retreating downriver and back again. This fluctuation is caused by short-term air temperature variations in the river basin. Comparisons of with-project conditions for the same climatic year under one-dam and two-dam scenarios, however, show a greater degree of fluctuation of 0 C isotherm progression for the one-dam scenario than for the two-dam. This can be explained by the fact that there are moderately different flow releases and slightly different release temperatures under e ach operational scenario . Changes in flow and temperature releases could be used, in fact, to gain some degree of control over the position of the 0 C isotherm and, hence, the ice front. A multi-level intake structure is included in the design of both dams to draw off waters from selected thermal strata for this purpose. b. Effects on the lee Cover SNTEMP assumes an open water sc~nario and cannot accurately predict the position of the 0 C isotherm in an ice covered reach. Therefore, whenever the 0 C isotherm is downstream of the ice front (i.e.: beneath the ice cover), the ICECAL model supercedes SNTEMP and calculates a new position for the 0 C isotherm. Whenever the 0 C isotherm is again upstream of the ice front, SNTEMP results again take over. Simply stated, it is important to note that whenever the 0 C isotherm is upstream of the ice front, SNTEMP results apply; whenever the ice front is upstream of the 0 C isotherm, ICECAL results apply. When the 0 C isotherm is upstream of the advancing ice front, cooling air temperatures prevail and the ice front advances upriver. Under these circumstances a reach of open water exists upstream of the ice front with water temperatures at 0 C. In this reach, border ice and anchor ice could form. Simulations predict that a maximum of about 30 miles of border ice 33RD2-007p could form with a one-dam scenario, while only as much as 10 miles of border ice could form under a two-dam scenario. Anchor ice could form in this reach in a manner similar to that which forms upstrean of the ice front under natural conditions. Whenever the 0 C isotherm is coincident with the ice front, stable air temperatures prevail and the position of the ice front would be stable, neither advancing nor retreating. In this circumstance, all open water upstream of the ice front would be above 0 C, and no significant border ice or anchor ice could form there. When the 0 C isotherm decays to a position downstream of the ice front, warming air temperatures prevail and the 0 C isotherm lies beneath the ice cover. Under these conditions, water above 0 C underlies the ice front. This causes the ice iront to disintegrate and the ice front retreats downriver. In this circumstances, no border ice or anchor ice could form upstream of the ice front. 3. RIVER ICE a. Freeze-up (1) Natural Conditions. Observations of natural conditions in the middle river are available from the winter of 1980-81 through 1983-84 by R&M Consultants (R&M Consultants, Inc., 1982a, 1983a, b). !CECAL model calibrations were made by Harza-Ebasco Susitna Joint Venture (1984a) using the 1982-83 and 1983-84 observations. Additional !CECAL simulations of natural conditions were made for the winters of 1971-72, 1976-77, and 1981-82, for comparison with the with-project runs. 33RD2-007p ISJ Simulations of natural conditions in the middle river are started when the ice bridges at the Chulitna confluence. This bridge has generally not occurred until the lower river front has reached the confluence area. In cold years with low flows, the middle river progrP-ssion has begun prior to the lower river being completely covered. For simulation purposes, the 1981-82 and 1982-83 middle reach progression was begun on the observed date, and the 1971-72 and 1976-77 progressions were assumed to begin at the earliest and latest observed bridging dates, res pectively. For natural conditions, the upstream simulation boundary was just upstream of Gold Creek, with ice inflow based on observations for 1981-1982 and 1982-1983, and ice inflow based on estimates correlated to Talkeetna temperatures for 1971-72 and 1976-77. (2) With-Project Conditions . For with-project simulations, lower river ice front progression is assumed to have reached the Yentna River on November 1. This is similar to natural timing, and is expected to be conservative for with-project ice conditions in the middle reach. In fact, the ice bridge near Cook Inlet with-project may form 1 or 2 weeks later than natural becaus e of trapping of the early Upper River frazil by the r eservoir(s). Following the November 1 start of the lower river front, ICECAL computes daily ice generation in the Middle Reach, estimates the ice contribution from the Chulitna and Talkeetna rivers, and estimates the contribution of frazil generated on the lower river. The daily ice vo lume is accumulated until the estimated volume of ice required to fill the lower river is produced. At this time, the progression is permitted to begin in the middle reach. 33RD2-007p The upstream boundary of ICECAL for with-project conditions is the 0 C isotherm location, as determined b y the i1 .stream temperature model (SNTEMP). ice Droduction is then computed in the reach from the 0 C isotherm downstream to the ice front. At the front, the ice either contributes to the front progression or is drawn down under the ice cover and deposited downstream, depending on hydraulic conditions at the front. (a) Project Operation. The annual maximum sin ulated progressions of the ice front in the middle river are summarized in figure 65 . The following conclusions can be made: 1. In warm winters, the ice front is not expected to progress past RH 1 26 (Slough 8A-East head). 2. In a ve ry cold winter, the ice front is expected to progress to the vicinity of Gold Creek, similar to natural progressions. 3. In an average winter, the front is expected to progress to very near Gold Creek with Watana only, and 5-10 miles downstream of Gold Creek with Devil Canyon on line. The simulated duration of ice cover in the middle reach is summarized figure 66 . This summary indicates the following: 1. In the coldest winter, 1971-72, the natural duration is about 6 months. The with-project duration is 4. 5 to 5. 5 months. The duration with Devil Canyon is slightly less than with Watana only . The progression 33RD2-007p typically begins 3-4 weeks later than natural, but meltout is ge~erally early to mid-May, similar to natural . 2. In the warmest winter simulated, 1976-77, the natural coverage exists for only 5 months, compared to 3-4 months for with-project conditions. The with-project ice progression starts 2-4 weeks later than natural and meltout is 2-3 weeks earlier. 3. In an average or warm winter, the natural coverage is about 6 months, and the with-project covera ge is valy abo ut 3 months. With-project progression typically begins in mid-to-late December and melt-out is complete by mid-to-late March. Maximum stages and ice thicknesses for basic simulations are tabulated in figures 67 and 68. Maximum stages are plotted relative to threshold elevation for sloughs in the middle reach in figures 69 to 72. Maximum ice thicknesses are plotted in figures 73 to 76 . These results indicate the following: 1. In a cold winter such as 1971 -72, many sloughs are overtopped in the natural condition as well as with-project. However , with-project, most of the sloughs are overtopped by a larger amount and for longer durations. 2. In a warm or very warm winter such as 1982-83 or 1976-77, there is little or no overtopping naturally, and overtopping is minor, with-project. 33RD2-007p 3. In the average winter, 1981-82, there is minor overtopping naturally, and moderate overtopping with Watana only. 4. thicknesses, with-project, are generally about the same as natural in the downstream half of the reach and less in the upstream half of the reach. This is particularly true after 2002 since the ice front rarely reaches Gold Creek with Devil Canyon on-line. Border lee, (or shore ice, or lateral ice) can be expected along the river banks where the water has cooled to 0 C upstream of the ice front. ICECAL estimates the development of border ice based on the water velocity and air temperature. The greatest rate of border ice development occurs in low velocity zones with cold air temperatures. The development of border ice is important because the water surface covered by border ice is not available for producing frazil ice. Border ice is simulated based on general experience data, and observations on the Susitna. Figure 65, which shows the maximum progression of ice in the Middle Reach does not indicate the border ice which would exist upstream of the front. Simulations indicate that, with Watana only, border ice would exist, as times, as far as 30 miles upstream of the maximum progression and may cover up to 25% of the water surface in that reach. With Devil Canyon, the border ice is expected to be restricted to a reach of only about 10 miles upstream of the maximum progression, again covering as much as 25% of the surface area. Anchor ice is generally found in the same general locations as border ice, that is, between the 0 C isotherm and the ice front. Anchor ice is not included in the simulation because little is known about the mechanism of its 33RD2-007p IS? formation and it is generally not modelled. Natural observations indicate it is quite common in the middle reach, presently, and would also be there with project. It is expec ted to have a relatively minor effect on with-pr<•ject water levels upstream of the ice front, which are generally lower than natural levels. Anchor ice generally produces less staging than an ice cover ·•ith the same volume of ice, since an ice cover results in a displac ement of the water surface as well as an additional roughness surface (shear face). The anchor ice produces a similar displacement but with no additional roughne ss surface. (b) Watana Filling. Results of simulations for the first and r.econd winters of filling compared to natura l are shown on figure 77 . A wa1~ winter was us e i to simulate the first winter of filling since relatively wa l ~ releases would be made through the low level cutlet for that condit~~. For the second winter, colder winter tenperatures were used since ~older releases would be made from near the reservoir surface with the cone vQ ~ves. The rates of flow during the winters of filling would be equal to natural flow rates. Thus, these two winters represent a wide range of conditions during reservoir filling. Figure 77 indicates that middle reach winter stages during filli1g are similar to or slightly lower than natural stages. Simulated thicknes:;es are likewise similar to or less than natural conditions. The ice front during filling is simulated to progress to the v icinity of RH 156-162, which is in the Devil Canyon area. The simulated ice proc~sses during filling upstream of Gold Creek are questionable, howev er, becaut~e observations of natural conditions in this zone have indicated that intermittent closure by border ice is the predominant process, rather t1an 33RD2-007p ISY front progression . No significant ice is expected upstream of Devil Canyon during filling under average or warm winter conditions. Simulated freezeup of the middle reach during filling begins 5-7 weeks later than natural, and meltout is expected in early May, similar to natural conditions. 3. ALTERNATIVE INTAKE DESIGNS AND OPERATING POLICY The basic simulations discussed above have been made with the "Inflow- matching" release criteria. This generally results in the coldest water available being released in the winter in an attefup t to match the 0 C winter inflow temperature. However, there is some interest in warmer releases in the winter in order to reduce the extent of ice coverage and staging with- project. The following studies were made to determine the sensitivity of ice related stages to intake operation for Watana in 1996 and 2001 using hydrologic and meteorologic data for the winter of 1971-72, the coldest winter: 1. A theoretical 4 C release for the entire winter period. This is not a DYRESM result, but assumes that the warmest water in the reservoir can be selected for release throughout the winter. 2. Warmest water available for the present intake design. This generally means using the bottom opening of the multi-level intake throughout the winter. 3. Warmest water available with an additional lower intake at El. 1800. This intake would be 240 ft. below the lowest present intake opening. 33RD2-01)7p I 51 4 . Warmest water available with an additional lower intake at El. 1636, instead of El. 1800. Figure 78 is a comparison of the runs in relation to their performances in preventing overtopping of significant sloughs and side channels. Conclusions are as follows: 1. The theoretical 4 C water is not effective for preventing overtopping significantly in the downstream third of the reach between Talkeetna and Gold Creek. 2. In the middle third of the reach, 4 C water would be effective. However, the lowest intake tested, El. 1636, is no more effective than the present design, and 4 C water cannot be withdrawn from the reservoir in this winter. 3. In the upstream third of the reach, 4C wate r would be very effective. However, only the intake at El. 1636 begins to approach the effectiveness of the 4C water . The present intake design, using warmest water, is a substantial improvement over "inflow-matching ." However, an intake at El. 1800 is no more effective than the present intake design. In summary, an additional low intake at El. 1636 would have some limited benefits in reducing ice related stages . However, the additional cost and other water quality questions related to releases from such depths make the alternative questionable. 33RD2-007p 4. Case E-Vl Flow Constraints. The bulk of the river ice studies have been made using the Case C flow constraints. However, Case E-VI has now replaced Case C as the recommended flow guidelines. The two release constraints are shown in figure 56. Case E-VI provides a more uniform summer flow . Additional !CECAL simulations have been made for Watana only in 2001, and Devil Canyon in 2002, with Case E-Vl, for comparison with Case C. The comparison of slough stages is shown on figures 79 and 80. Conclusions are as follows : 1. For 2001, average stage is about the same for both cases. Case E-VI results in one less slough overtopped (Slough 8), but the upstream sloughs (SA, 9 and 9A) are overtopped with a slightly larger head with Case E-VI. 2. For the 2002 comparison, Case E-VI results in slightly higher stages in the downstream reaches and slightly lower stages near Gold Creek. However, most sloughs are .. ~t overtopped for either case. Therefore, with Devil Canyon on-line, slough over topping is not considered important in either case except in a cold winter such as 1971-72. 3. In summary, Case E-VI and Case C produce similar winter ice conditions for the inflow-matching release policy for the years 200 1 and 2002, even though the winter release temperatures are slightly colder with E-VI. 33RD2-007p b. Breakup Breakup processes are expected to be different from natural in the Susitna River below the project, especially in the Talkeetna to Devil Canyon reach. Since the maximum upstream extent of the ice cover below the dams would be between RM 124 and RM 142, there would be no continuous ice cover between the damsite and the ice cover and, consequently, no breakup or meltout in that reach. Any border ice attached to the shore would probably slowly melt away in place. Occasional pieces of border ice might break away from shore and float downstream. Ice in the river reach above the project reservoirs would break up normally, but would not drift into this area as it normally does because it would be trapped in the reservoirs. Ice drifting into the headwaters of the Watana Reservoir may jam against the reservoir ice cover. This could result in elevated water levels in the Susitna River immediately upstream of the reservoir. As the reservoir ice cover melts, the ice from upstream would move further into the reservoir and the elevated water levels would decrease. At the time of the potential headwater ice jamming, the water level in the reservoir would be at its lowest leve l and would be between 40 and 120 feet below the normal maximum pool level. Water levels resulting from any ice jam would probably not exceed the normal maximum pool and would therefore not result in any flooding above the normal maximum pool level. The natural spring breakup drive is usually brought on by rapid flow increases that lift and fracture the ice cover. The proposed project reservoirs would regulate such seasonal flows, yielding a more steady flow regime and resulting in a slow meltout of the ice cover in place, from upstreac to downstream. 33RD2-007p The warmer than natural water temperatures released from the project would cause the upstream end of the ice cover to begin to decay earlier in the season than natural. Gradual spring meltout with Watana operating alone is predicted to be as much as 4 to 6 weeks earlier than natural, and 7 to 8 weeks earlier than natural with both dams operating. By May, flow levels in the river would be significantly reduced compared to natural as the project begin~ to store incoming flows from upstream. The result is expected to be that breakup drive processes that now normally occur in the middle river area would be effectively eliminated. Instead, a slow and steady meltout of river ice in this reach would probably occur. Since there would not be an extensive volume of broken ice floating downstream and accumulating against the unbroken ice cover, ice jamming in the middle river would probably not occur or would be substantially reduced in severity. This would eliminate or substantially reduce river staging and floo~lng normally associated with ice jams. thereby eliminating or greatly reducing the overtopping of berms. flooding of side sloughs and scour of the bank and channel. Observations of natural conditions indicate that breakup ice jam flooding in the lower river is not as severe as in the Middle Reach. Breakup ice jams have been noted near Montana Creek (RM 77), RM 85. 5 and RM 89 in 1983. However, the wide flood plain with many channels tends to prevent significant staging. With-project, spring flood flows from the Talkeetna. Chulitna and Yentna Rivers would make breakup in the lower river more similar to natural than break-up in the Middle Reach. The earlier meltout, warmer than natural temperatures and reduced flows from the Middle Reach with-project, are expected to cause the lower river meltout or breakup to begin earlier and be milder than natural . 33RD2-007p c. Effects of Power Flow Variation on lee Cover Natural ice cover formation is generally a very stable process, with minor stage fluctuations in the zone of front progression . In fact, natural flows are generally low and decreasing during the freezeup process which contributes to a well-behaved condition. Natural breakup, however, can be a much more unstable process. Flows are generally increasing in the spring and the ice cover is weakened by solar radiation and warm air. This combination can result in collapse of covers over large reaches, causing- ice jams and flooding. With hydroelectric power developments on a river, the natural flow and temperature conditions are altered. Power releases in the winter are generally larger than natural and warmer. In addition, hydroelectric generation can respond quickly to system load changes and is well suited to load following or peaking. Flow fluctuations resulting from load following or peaking can result in fracture of an advancing ice cover in early winter, or a ~~treating cover in spring. Such a fracture of cover can lead to flow surges, ice jamming, and flooding of aquatic, terrestrial, or human habitat. (1) EXPERIENCE SURVEY. A mailed survey has been conducted in order to ascertain the effects of winter power operation on river and reservoir ice. Results indicate the following: 1. There are few or no problems within impoundment zones. There are isolated instances of animal casualty, but no specific operation constraints for this. 33RD2-007p 2. There are a few cases reported in Canada where ice jam tlooding of towns has been attributed to upstream hydroelectric power operations. 3. Where operational procedures are in effect, they are generally directed toward protecting human life and property rather than aquatic or terrestrial habitat. 4. Operational restrictions such as those for British Columbia Hydro 's Peace River project include the following : a. During freezeup ice cover progression through sensitive areas on the river, flows are controlled at a high level until the cover develops sufficient strength to withstand flow fluctuations. b. After freeze-up, the plant can be operated freely without endangering the cover except in the frontal zone. c. During break-up and melt-out, flows are again maintained at a high level in the sensitive areas to erode the cover as quickly as possible . If tributary break-up appears imminent, B. C. Hydro releases are decreased in order to minimize the effect of the tributary ice surge. 5. In other cases operational constraints are employed to prevent the formation of hanging dams downstream or to reduce water levels upstream after a dam forms. These dams may result in high water levels which can reduce the plant generating capacity, endanger ~he powerhouse, or result 33RD2-007p in flooding of areas adjacent to the river. These types of constraints include: a. Inducing an early ice cover on the river upstream of known sites of hanging dams, by artificial means such as ice booms or other obstructions. When an ice cover forms, frazil ice production stops and the hanging dams, which result from frazil accumulation, are minimized. b. Inducing an early ice cover on the river by keeping powerhouse discharges low while the ice cover forms. This may result in more rapid ice cover advance, preventing further frazil production. formed, powerhouse discharges can be increased. After the ice cover is c. Preventing ice cover formation in sensitive areas by fluctuating discharges, continually breaking up the ice cover and keeping it downstream. This may result in higher water levels further downstream, but lo'l-rer water levels in sensitive areas. d. Reducing discharges after a hanging dam forms in order to reduce water levels upstream of the hanging dam. 6. The Canadian Electrical Association and many plant operators indicated that powerhouse operations during the winter to maintain a stable cover would be site specific and require operating experience over a number of years (ACRES, 1980). Climatic conditions, ice characteristics of years (Acres Consulting Services, Ltd. 1980). Climatic conditions, i~e 33RD2-007p characteristics, channel morphology, and water temperature are all variables which must be considered. (2) Susitna Operations. The License Application indicates that with Watana only, the Susitna project would be operated to generate base load electrical energy. In addition, following the construction of Devil Canyon, Watana would load-follow and Devil Canyon would be base-loaded. It is presently proposed to operate Watana before Devil Canyon is built, and later Devil Canyon, as follows (Alaska Power Authority 1985): 1. Allow maximum of t 10% flow variation from the weekly average flow within a given week. 2. Allow maximum t 10% flow variation from the current flow within a given hour. A flow variation of 10% (for instance, 1000 cfs with a base flow of 10,000 cfs) results in an open-water steady-state stage change of on 1 y about 0.3 feet. With an ice-cover, the steady-state stage change would be about 0.5 feet. This is considered to be safe from the standpoint of preventing breaking of the cover during freezeup, mid-winter, or breakup conditions. It is not believed that other constraints on winter powerhouse operation would be necessary to keep the ice cover stable. 4. PROJECT EFFECTS ON RIVER CROSS-SECTION CHARACTERISTICS 33RD2-007p Project operation in the winter is expected to change the flow cross-section characteristics in the middle reach compared to natural conditions. Parameters such as velocity, cross-section area and wetted bed perimeter would be somewhat different with-project. For e x ample, figures 81 and 82 show the middle reach velocity profile before and after freezeup for the winter of 1981-82, for natural and with-project 1996 conditions. These plots show the following: 1. Before freezeup, velocities are highly variable from section to section, for natural or with-project conditions. Velocities for natural conditions range from about 1.5 to 7 fps. With-project the range is from about 2 to 9 fps . Average velocity is about 3 fps, natural and 4 fps, with-project . 2. After freezeup, velocities range from 1 to 3 fps for natural conditions, and about 2 to 3 fps with project conditions. Average velocity is about 2 fps for natural and 2.5 fps with-project. 3. After freezeup, velocity is controlled by the critical velocity for erosion-deposition of slush beneath the ice cover, for pre-project and with-project conditions. This velocity is about 3 fps and is the practical limit for velocity beneath the cover. This mechanism controls the distribution of ice deposited beneath the cover. ~igures 83 and 84 show conditions at a typical cross-section in the middle reach, LRX 22, RM 119.3, for natural and with-project conditions. These plots show the following : 33RD2-007p 1. For natural conditions, the under ice water surface after freezeup is almost the same as the open-water stage . Figure 78 shows that at this section, the velocity drops from 3.6 fps to 2.3 fps during the freezeup, corresponding to the flow change from 3000 cfs to 1750 cfs during the freezeup. This also means that, in this case, the velocity decrease just compensates for the friction perimeter increase to produce a similar friction slope before and after freezeup. 2. For natural conditions at this section, there is little change in flow area or bed perimeter as a result of freezeup. 3. For with-project conditions, the under-ice water surface increases d~ring freezeup resulting in a slight increase in exposed bed perioeter. Figure 84 shows that the velocity drops from 5.2 fps to 2.2 fps during freezeup. This is due to the flow reduction from 12,000 cfs to 10,000 cfs and the increase in flow area during freezeup required to balance the effect of additional friction surface with the ice cover. 4. For with-project conditions at this section, the flow area doubles during freezeup and the bed perimeter increases slightly. The under-ice flow velocity after freezeup is almost the same as natural. Upstream of the with-project ice cover there would be changes from natural conditions as well. Figures 85 and 86 show natural and with-project conditions in 1996 at LRX 33, RM 130.1 for the winter of 1982-83. These plots show the following : 33RD2-007p 1. The natural condition velocity after freezeup is 1.6 fps and the with-project velocity is 3.3 fps. 2. The flow area after freezeup changes from 1800 square feet for na t~ral condi tions to 3000 square feet for with-project conditions . 3. The water surface elevation after freezeup changes from 615 .5 feet m ~J. under natural conditions to 613.2 feet msl with-project. 5. EFFECTS ON SLOUGH AND SIDE CHANNEL FLOW Flows in sloughs and side channels are affected by winter river water levels which can increase infiltration from the mainstem and overtop berms at upstream ends of the c hann ~ls. Figure 87 j~ ~ ~i~gramatic presentation of t~e sources of slough flow. At a particular slough (or side channel), the g~neral effect of increasing river water levels without overtopping the upstream berm would be to increase the amount of groundwater flow from the mainstem without signific~ntly altering its temperature. At the same location, if the upstrean berms were overtopped, both the quantity and temperature of intragravel flow would generally be affected: the quantity of flow could incre ase significantly, while the temperature would be dominated by that of the diverted mainstem water. Intragravel flow refers to the flow in the bed material of the habitat area and has two sources (see figure 88): 1. Groundwater flow which has three components (see discussion below under Quantity of Flow). 33RD2-007p /?0 2. Longitudinal flow in the strea~bed material caused by surface flow in the habitat area. In the winter, when a berm at the upstream end of a habitat area is not overtopped, intragravel flow is generally caused by groundwater flow. When the habitat berm is overtopped, the intragravel flow many be dominated by flow induced by surface flows. Winter water levels and thus slough and side channel flows would be affected differently in areas within and upstream of ice covered reaches of the mainstem. The extent of the ice cover would vary from year to year as shown on figures 65 to 67, depending primarily on climatic conditions. The upstream limit of the ice cover would be located further downstream with Devil t Canyon in operation than with Watana operating by itself. In addition, the ice covered reaches would be dynamic and thus areas near the upstream end of the ice cover may experience changing water levels throughout the winter. Figure 67 can be used to determine whether important habitat areas would be affected by overtopping of upstream berms, as well as the mainstem water levels influencing groundwater flow toward each habitat area. This exhibit assumes existing slough berms are not protected from overtopping. The following discussion concentrates on effects on quantity and temperature of slough and side channel flow in areas within and upstream of the ice cover. a. quantity of Flow The rate of flow in sloughs and side channels depends on local runoff, groundwater discharge, and flow resulting from overtopped berms. Under natural conditions, local runoff and overtopping of berms are relatively insignificant from early September until an ice cover forms on the river in 33RD2-007p /7{ November or December. During this period, the main source of flow in these habitat areas is groundwater which may originate f-om any of three sources: 1. Shallow localized infiltration from the mainstem, 2 . More regional groundwater transport in the downstream directio n ~ithin alluv ial materials comprising the Susitna River Valley, and 3. Transport towar d the river from upland glacial till and sedimentary r o cks comprising the Susitna River Valley walls . Between November and May, the middle Susitna River is normally ice covered. Water levels are elevated and overtop some side channels and sloughs. In non-overtopped areas, the higher water levels increase groundwater flow from the mainstem toward the sloughs and side channels . With-project mainstem flows would be higher than natural between September and April. Forma tion of ice cover i n the middle Susitna Riv er would be delayed relat i ve to natural conditions by approximately two to six weeks, until sometime in December. The maximum extent of ice-affected wa ter levels would be reduced relative to natural conditions. Water levels would be increased within of the ice-covered are as, and dec reased upstream of ice-covered areas. (1). Period Immediately Prior to Ice Co ver Formation. Under with-project conditions, mains tem water levels ~ould be higher than natural between September and November when an ice cover normally occurs . When Watana is operating alone, flows in this period would vary between 6,000 33RD2-007p /7~ cfs and 13,000 cfs and would average approximately 9,000 cfs. This is about 6,000 cfs above average natural flows for the period. When Devil Canyon first comes on line, flows would be in a much narrower range and would average approximately 7,000 cfs. As system energy demands increase, flows would also incr~ase to near Watana-only levels. Resulting water levels would be about two feet above natural levels, and would be similar to minimum with project water levels for the summer period (May through August). The groundwater flow resulting from shallow infiltration from the mainstem would thus be increased about natural conditions and would be similar to that component for the previous summer of operation. The component of groundwater discharge to sloughs resulting from regional downstream transport within alluv ial valley materials would be largely unchanged from natural conditions o r the previous summer period because the downstream gradients would remain about the same. Elevated groundwater levels resulting from elevated mainstem stage, with ice cover, could result in somewhat larger areas of upw e lling, thus increasing slough discharge somewhat. The component of groundwater discharge originating from upland areas would be unchanged from natural conditions. However, this component normally declines throughout the winter as recharge to these aquifers is reduced . Consequently during this period, groundwater flow in side channel~ and sloughs would be higher than under natural conditions, and similar to, although perhaps slightly smaller than, groundwater flows corresponding to minimum mainstem flows during the previous summer. (2). Within Ice-Covered Areas. An ice cover would form on the middle Susitna River during December and January. With-project discharges during this period are approximately 9,000 cfs above natural conditions and water levels within the ice-covered areas are 33RD2-007p 173 generally a few feet higher than natural. For this reason, the groundwater component due to local mainstem infiltration would be increased above that under natural conditions and would also be increased above that during the with-project summer open water period.. The other two components of groundwater flow would be largely unchanged from natural conditions, although the component due to regional down-valley groundwater flow may increase slightly if the area open to upwelling increases because of generally elevated groundwater levels in the valley. Groundwater discharge from upland areas would be the same as under natural conditions, and thus lower during this period than during the summer open-water period. There would be an increased frequency of berm overtopping in these areas relative to natural conditions if they are not protected, and the total flow in the sloughs and side channels would increase considerably from natural flows where this occurs. (3). Upstream of the Ice-Covered Area. The ice-covered reach of the middle Susitna River would extend upstream to between river mile 125 and river mile 142, depending on climatic conditions and whether or not Devil Canyon is operating. Upstream of the ice front, open water conditions would prevail. Winter time discharges would normally result in river water levels which are less than or equal to natural ice-affected water levels in this reach. The differences are generally on the order of zero to two feet. The maximum allowable winter discharge for Case E-VI is 16,000 cfs, which is not sufficient to overtop berms at the upstream ends of any of the sloughs upstream of the ice front. Some side channels, however, could still be affected by ~ainstem flows . The groundwater component due to shallow mainstem infiltration would be lower than natural in this area. 33RD2-007p However, winter discharges are higher than minimum with-project summer discharges so that the mainstem-affected component of groundwater flow would be higher in the winter than the minimum summer component. Other components of groundwater flow would be largely unaffected. Therefore, upstream of the ice-covered area, side channel and slough groundwater flows would be higher than minimum with-project summer groundwater flows but lower than natural winter groundwater flows . 6 . Water Temperature Surface water temperatures in sloughs and side channels are affected by thP-temperature of upwelling groundwater, climatic conditions, the temperature of mainstem water if the berm is overtopped, and the temperature of local surface runoff. In the winter, runoff can be neglected. Intragravel temperatures are primarily influenced by the temperatures of the components of groundwater flow. In habitat areas, intragravel temperatures appear to remain more stable than river temperatures. The degree to which intragravel temperatures at a habitat site change in relation to changes in mainstem temperature appears to be a function of 1. the distance between the site and the mainstem, which affects the travel time for groundwater from the mainstem: 2. overtopping of berms by cold mainstem water, which may depress intragravel temperatures; and 3. the relative contributions of the three components of groundwater flow. 33RD2-007p II~ Measurements of intragravel temperatures by the Alaska Department of Fish and Game (ADF&G 1983) at a site near the mouth of Slough 8A showed that, following the progression of an ice front past the slough, the intragravel temperature dropped to near 0 C within approximately a mouth. The ice cover raised mainstem water levels near the slough by 3.4 feet (R&Z.l Consultants, Inc., 1984b) and caused overtopping of the upstream berm. ADF&G hypothesized that the cold surface water depressed intragravel temperatures. It may also be possible that intragravel temperatures were responding to mainstem temperature transmitted with groundwater flow. Intragravel temperatures at this site seem to reflect mainstem temperature influence more than at another site upstream in Slough SA. lntragravel temperatures at the upper site are constant near 3 C all winter. The differences in the temperatures at the two sites may be related to the fact that the downstream site is within 1,000 feet of the river while the upstream site is approxi~ately 4,000 feet from the river. Observations on intragravel temperatures and mainstem staging for the winter of 1983-84 were similar to 1982-83, although overtopping of the berm was not reported (ADF&G 1985, R&M Consultants, Inc., 1984a). Measurements by ADF&G (1983) at a site in Slough 9 indicate that slough surface water temperatures drop to near 0 C during the winter, while suggesting that intragravel temperatures remain fairly constant at about 3 C. However, this latter inference cannot be confirmed because of the lack of intragravel temperature data for the period from mid-November 1982 through mid-March 1983. Intragravel temperatures recorced in 1983-84 (ADF&G 1985) were constant at about 3.5 C throughout the winter while surface water temperatures varied to as low as p.5 C. As a result of overtopping during ice breakup in May of 1983, intragravel temperatures fell rapidly although only minimally (less than 0.5 C) in response to a rapid 5 C decline in surface 33RD 2-007p /7~ water temperature. This suggests a response of intragravel temperatures to overtopping flows of very cold water, but still a dominance by warmer upwelling water. At slough 11, both surface water and intragravel temperatures remained approximately constant at about 3.5 C from September 1982 through April 1983. However, it was discovered on April 29, 1983, that the surface water probe was cov ered with about one inch of silt (ADF&G 1983). When the probe was uncovered, the recorded temperature quickly rose by 1 C to the inferred true surface water temperature . Thus it is unclear whether the recorded surface water temperature reflects a predominance by warm upwelling water, or whether it is in effect a measurement of intragravel rather than surface water temperature. Intragravel temperatures measured in 1983-84 (ADF&G 1985) show similar values to those measured in 1982-83. Surface water tempetatures are more sensitive to climate conditions, are less varied than at Slough 9 and generally between 1 C and 2 C between November and February. At Slough 21, intragravel temperatures near the mouth of the slough remained fairly constant at about 3.5 C during the winter of 1982-83, while surface water temperatur~s were somewhat colder, fluctuating between 0 C and 2 C (ADF&G ~983). At a site further upstream in the slough, both the surface water and intragravel temperatures varied in apparent direct correlation with each other, with the surface water t e mpe ratures about 1.5 C-2 C colder than the intragravel temperatures. During overtopping events in September (warm water) and May (cold water), both the surface water and intragravel temperatures showed approximately equivalent responses of several degrees. This suggests that surface water temperatures in the upper portions of the slough may be dominated by upwelling groundwater, except during overtopping of the upstream berm, when diverted mainstem water would dominate. Intragravel 33RD2-007p /?7 temperatures measured in 1983-84 (ADF&G 1985) showed similar trends, although overtopping was not reported. Intragravel temperatures measured in side channels (ADF&G 1985) show the same characteristics as in sloughs. The measured temperatures are much more stable than mainstem and side channel surface temperatures, but show more variability over the year and more short term fluctuations than slough intragravel temperatures. This may be explained by their closer proximity to the mainstem. Side channels may be subject to more frequent overtopping than sloughs. In general, it appears that the intragravel tP.mperature in many habitat sloughs and side channels would be dominated by warm groundwater flows during the winter, except for periods of overtopping of upstream berms as a result of ice staging or ice breakup if the berms were not protected from overtopping . the temperature of groundwater flow in slough areas appears to remain fairly stable, at a temperature approximately equal to that of the mean annual mainstem water temperature (Alaska Power Authority 1984). Side channel intragravel temperatures show about 1 C to 2 C variability about the mean annual river temperature. The mean annual mainstem temperature is not expected to change significantly with-project, and the temperature of that component of groundwater flow which is directly related to mainstem flows should not be changed . For those slough areas and side channels which appear to be more directly influenced by mainstem temperature variations, such as near the mouth Slough 8A and near the upstream end of Slough 21, the increased winter f lows relative to natural conditions and increased mainstem water level~ as a result of ice formation could produce colder than natural upwelling groundwater and thus colder intragravel te~~eratures. However, the intragravel temperatures in 13RD2-007p these sites are often found to be near 0 C especially at Slough 8A, for natural con~itions. 6. ICE EFFECTS ON SEDIMENT TRANSPORT AND CHANNEL STABILITY Ice processes affect sediment transport and channel stability in the following manners: 1. The formation of an ice cover on a river may double the flow wetted perimeter, reduce the flow velocity and reduce the tractive force of the flow. This will reduce the sediment transport capacity of the flow . (Sayre and Song 1979) 2. Elevated water levels associated with ice cover formation and breakup jamming may result in overtopping of sloughs and side channels along the perimeter of the main channel. Flow velocities associated with these events may remove material in the slough streambed. Velocities associated with breakup jamming and overtopping may be sufficient to erode large amounts of material and change the character of the slough . A breakup ice jam in 1976 resulted in changing Slough 11 from an upland slough to a side slough. (LaBelle 1984) 3. High flow velocities associated with surges res~lting from the failure of an ice jam may result in the scour of river banks and may remove vegetation and streambed material. (Gerard 1983, R&M Consultants, Inc. 1984b) 33RD2-007p !?( 4. Anchor ice may attach to a streambed and when refloa ted by rising temperatures and solar radiation, may carry some fine sediments from the streambed with it. (R&M Consultants, Inc. 1984b) 5. The flow of water throug~ a porous slush ice cover may result in the filtering of some fine material carried by the water. (R&M Consultants, Inc. 1984b) 6. The melting of ice deoosited on riverbanks, floodplain areas, or sloughs may result in the deposition of sediments carried in the ice on the underlying material . (R&M Consultants, Inc. 1984b) The discussion of effects of project operation is organized by habitat type . a. Mainstem Habitat (1). Suspended Sediment Concentration. With-project suspended sediment concentrations in the mainste m of the middle river would be increased over natural winter concentrations , which are near zero mg/1 . Fine materials influent to the reservoir are expected to remain in suspension and be discharged throughout the year. It has been estimated that the suspended sediment concentratio n in the powerhouse flows would be between 0 and 100 mgl (Peratrovich, Nottingham, and Drage 1982) with an average of about 60 mgl. This concentration would remain relatively stable between the dams and the upstream end of the ice cover . There would be a r~duction in the amount of fra7~1 and anchor ic£ occurring in the middle reach 33RD2-007p due to the warmer releases from the dams. Some sediment may be picked up by the frazil and anchor ice. However, any reduction in s uspended sediment concentration caused by frazil and anc hor ice may be compensated by sand particles picked up from the river bed as a result of higher winter fl ows (Harza-Ebasco Susitna Joint Venture 1984c). Downstream of the ice cover, some sediment may be trapped by the fra zil ice. The ice cov er would reduce the capacity of the river to pick up additional sand particles. Suspended sedime nt c oncentrations are e x pected to remain relatively constant at level s s imilar to levels upstream of the ice cover. In the reach downstream of the c onfluenc e with the Chulitna and Talkeetna Rivers, the suspended sediment concentration is also expected to be controlled by the concentration upstream of the confluence. Flow from the Chulitna and Talkeetna Rivers would dilute the suspended sediment concentration slightly. Winter discharges in this reach would be higher with project than natural conditions. This would cause additional sediment to be picked up. However, some sediment may also be trapped by the ice. Suspended sediment concentrations would be similar to the mi ddle river. (2 ). Channel Stability. Winter flows throughout the midd l e and lower rivers would be higher than natural. However, they would not be high enough to affect the ~tability of the streambed. The water levels d o wn s tream of the ice front would be higher than under natural conditions but the velocities wo uld be similar to natural conditions . Streambed subs trate would b e stable during this period, altho ugh some sand ma y be picked up by the increased flows. 1his would not alter the channel geometry. 33RD2-007p Breakup ice jams in the middle river would be reduced in frequency or eliminated Jue to project operation . Stream flows would be regulated by the project and spring floods, which normally lift and fracture the ice cover, would not occur. Warmer than natural releases from the reservoir would m«!lt the ice cover from the upstream end. Meltout would occur up to 8 weeks earlier than natural breakup. Therefo re, the main channel streambed and ·>anks would not be subject to scour resulting from ice jams. Some sediment suspended in the flow may be trapped in the porous frazil ice cover. The presence of layers of sediment in frazil ice covers has teen observed in the Tanana River. The mecha nism for this is not known . The concentration of sediment in the i c e appears to be highest at the bottom of the ice cover (Chacho 1985). Since the amount of material suspended in the flow may increase witt project, the amount of sediment trapped in the ice may also i ncrease. Dtring the spring meltout, the water level would reduce gradually as the ice co\er melts. The ice cover would melt primarily from the bottom due to the wa rm water in the river. Air temperatures would not be high e nough to cause significant surficial melting of the ice. Most sediment trapped in the i :e would be near the bottom of the ice and should be transported downstream as it is freed from the melting ice, rather than deposited on the river banks at1d underlying material. There would not be breakup ice jams which normally transport ice blocks onto overbank areas. Because of the higher than natural water levels within the ice covere•l area, there would be some stream-deposited ice in unprotected sloughs, s idt· channels and on the banks. Under natural conditions, this does not occur, except at breakup. deposited on banks. 33RD2-007p However, some border ice commonly breaks up and is In shallow areas the ice would not be exposed to much flow. The potential for accumulation of sediment in this ice would be lower than in the main channel . During the melt out of the main channel ice cover, the ice in these areas would not be as exposed to warm water as the mainstem ice and may remain in place after the main channel ice is melted out. Deposition of sediment from the melting of this ice is not expected to be significant and may be less than has been noted to be deposited on streambanks from ice remaining after breakup ice jams. In other areas along the stream margin the ice may be exposed to more flow throughout the winter. Potential sediment accumulation in this ice would be higher than in shallow areas, but less than in the mainstem. These areas would be ~xposed to melting from warm water in the same manner as the main channel. Sediments accumulated in the ice would be washed downstream as the ice melts and the sediment is exposed to flows. The ice would melt out in the same manner and time as the ice in the main channel. Deposition of sediments on underlying material might not exceed lev els resulting from ice deposited during spring breakup. b. Side Cham•els (1). Suspended Sediment Concentration. With-project winter flows and water levels would be sufficient to prevent most middle reach side channels from being dewatered . Downstream of the ice front, elevated water levels would keep side channels watered. Suspended sediment concentrations would be similar to the main ch~nnel with-project conditions in these areas. Upstream of the ice front, water levels would not be affected by ice. Side channels which would not be dewatered at flows of 33RD2-007p I? J 12,000 cfs to 13,000 cfs would have suspended sediment concentr J tions similar to conditions in the main channel. Dewatered side channels would be free of the main channel influence and would have clear upwelling flows. Anchor ice may form in watered side channels prior to the formation of an ice cover . Warm groundwater upwelling would inhibit this in some areas. This anchor ice may attache to and, when floated, transport some fine sediment out of the side channel. However, this is not expected to be significant . (2). Channel Stability. Within the ice covered area, high winter water levels would result in greater depths of flow in side channels than under natural conditions. Velocities, however, would not exceed approximately 2 feet per second. Additionally, the presence of an ice cover would effectively reduce the tractive force in the side channel. Substrate materials are expected to remain stable; however, materials such as sands and silts may be flushed from the substrate. The meltout of the ice cover would eliminate breakup ice jams as a source of scour in th side channels in the spring . Thus their streambeds would be more stable at that time also. Substrate materials in side channels upstream of the ice covered area would remain stable. Since ice jams would be less frequent or eliminated there would be less frequent stranding of ice which may contain sediment that would deposit in the area. Some ice from the ice cover may remain o n the side channel banks, and could result in deposition of sediment. c . Side Sloughs (1). Suspended Sediment Concentration. 33RD2-007p With-project winter flows and water levels would be sufficient to overtop unprotected berQ s at the head of slvughs within the ice covered reaches (figure 67). Suspe11ded sediment concentrations in these sloughs would have the characteristics of main channel flow. Upstream of the ice front water levels would not be sufficient to overtop berQs. Water in the se areas would be clear, resulting from groundwater f low. An ice cover may form in some side sloughs prior to ice caused staging in the mainstem. Warm groundwater flow would inhibit thi s in some areas of these slough. This ice may transport some fine sediments downstream when it is bro ken up and floated downstream, but this is not expected to be significant. Ice which may form in sloughs not late overtopped may cause some increase in suspended sediment concentration when broken up and float ed , but thi s also would occur under natura l conditions. Shnrefast ice would also form in sloughs not subject to overtopping. This also occurs under natural conditions. In general, sloughs which are not overtopped by elevated water levels would not have any change from natural conditions with respect to suspended sedimen t concentrations. (2). Channel Stability. Flow depths in side sloughs withi n ice covere d reaches would be greater than under natural conditions. However, the velocities wou]j not exceed approximately 2 feet per secori and the ice cov er would reduce the tractive force of the flow. Sediment transp o rt s~oulc be limited to fine materials. Side slough substrate would be sta~le under with project conditions. Substrate in side sloughs which are not overtopped would remain stable. As with side channels, the melt-out of the ice cover would elimina te breakup ice jams as a source of scour in side sloughs. This would improve the 33RD2-007p I~ stability of their substrate material. Since ice jams would be less frequent or eliminated, there would be less frequent stranding of ice which may contain sediment that would deposit in the area . In overtopped sloughs, some ice from the ice cover may remain stranded on banks. This could result in sowe deposition along the banks. In non-overtopped sloughs, shore ice that forms would melt out in much the same manner as for natural conditions. This ice would remain for longer periods than main channel ice since it is not subjected to warm water fr~m the main channel . Warming air temperatures, increased solar radiation and warm groundwater upwelling would melt t he shore ice. d. Upland Sloughs Project effects on winter time suspended sediment and channel stability in upland sloughs would be similar to side sloughs. Fewer upland sloughs would be affected by overtopping than side sloughs. e. Tributary Mouths (1). Suspended Sediment Concentrations. Suspended sedimen t concentration in the tributary couth habitat areas would be affected in the same manner a3 the main channel habitat areas. (2). Channel Stability. Channel stability in tributary mouth habit a t areas would be affected in the same manner as main channel habitat areas. The manner of ice deterioration would eliminate the potential for main channel ice to remain in tributary mouth areas and potentially block fish passage in these areas. 33RD2-007p I~ Tributary mouth habitat areas are subject to deposition of sediments brought down from the tributaries (Harza-Ebasco Susitna Joint Venture 1984d). This deposited material would normally be removed periodically by high summer flows. It can be speculated, although it hasn't been observed, that failure of breakup ice jams may cause surges which could also carry this deposited material downstream. With project, the manner of ice cover deterio r ation would preclude the removal of deposited sediments in these areas. Thus, aggradation affecting access may be encouraged by with project ice conditions. However, normal summer flows in the mainstem and tributary would cause downcutting of aggraded sediments and removal of sediments at most tributary mouths. f. Tributaries With project ice conditions would not affect sediment concentrations or channel stability in tributaries. 33RD2-007p !T7 Vl. ENV I RONMENTAL EFFECTS A. OVE RVIEW OF ICE-RELATED ISSUES This section describes the enviromrental effects of ice processes on major resources of the Susitna River specific to the upper, middle and lower river zones referenced by proposed location of hydroelectric facilities. Natural ice processes would perceivably undergo dramatic changes due primarily to different flow and temperature regimes anticipated during the construction and operational phase of the project. 1bis transformation raises issues and problems central to the welfare of fish, animal and habitat resources, and the people dependent upon them for recreational, commercial and subsistence purposes. Natural ice processes have a dynamic effect on the physic;1l characteristics of a subarctic river s y stem by constantly altering channels and surrounding flood plain. Ice and related flooding phenomena have negative and positive effects on a riverine environment. Fish and animals are most notably effected by the destruction, creation and alteration of habit:c.ts. Human dependence on fish and animal life is incontrovertibly linked to habitat stability and availability. 1. SOURCES OF ISSUES Because altered ice processes apparently evoke less public controversy, virtually all ice-related environmental issues originated with investigative agencies and consultants during the enviro ncental analysis phase of the Susitna River study. Based on previous environmental asses s ments, the Alaska Power Authority and Alaska Department of Fish and Game identified the majority of issues associated with altered ice processes. LGC Alaska Resource 33RD1-007n Associates, Inc. and E. ~oody Trihey and Associates identified pertinent issues and problems in conjunction with ongoing environmental studies respective to their contract responsibilities. Except for commentary from a local citizen (letter dated March 3, 1984 from Leon B. Dick) ice-related issues from the private sector have been remarkably nil. Issues and problems can be categoriztd as those pertaini~g to fish and aquatic habitat, riparian vegetation, wildlife and terrestrial habitat, and public use activities. a. Fish and Aquatic Habitat Fish related issues focus o· .. the effect of ice proc:esses on incubation and fry production. More perplexing is the effect of altered ice processes on habitat. This aspect is considerably more complex since aquatic habitats could be destroyed, created or altered during different freezeup and breakup scenarios as a result of project construction and operations. l-laj or issues relate to changes in channel morphology and substrate composition resulting from project-related ice and flooding processes. b. Kiparian Vegetation Changes in riverine morphology obviously affect the shore line and adjacent substrate and subsequently the successional stage of vegetation. Plant composition and growth patterns would also be influenced by sediment deposition and different ambient temperature caused by higher water temperature and rate of heat transfer. 33RD1-007n ;r( c. Wildlife and Terrestrial Habitat The most significant issues concern the relationship of different ice conditions to the mobility and survival of large mammals. The number of potential ice-related mishaps that normally occur as animals move and forage in bottoc land habitats may change as a result of more varied and less stable ice associated with higher winter flows and warmer temperatures. Of equal concern is the effect of ice and flooding forces on food and cover components of riverine habitats, especially those of low relief. During construction and operational phase, feeding niches subjected to flooding and overtopping would be significantly altered. d. Public Use The question of ice stability and presence of open water and its effect on winter human travel and a c cess is addressed in this section. Untenable conditions prevail as a result of warmer water temperatures, open water stretches and weaker ice formations. 33RD1-007n I io B. Mechanisms of Effects The physical effects on river ice processes that would ensue from the construction and operation of the Susitna River Hydro-electric Project are detailed in Chapter V. The altered physical processes that are expected to have important effects on the biological and human use environments are briefly summarized here in order to help orient the reader to the discussions that follow . 1. IMPOUNDMENT ZONE In this zone, where only the river occurs under natural conditions, Watana reservoir would first occupy the area from approx imately RM 184 to 239 when full, a distance of about 55 miles. It would vary in width along its length from 500 feet to 22,000 feet (4 .2 miles), not including runup into tributary mouths . Later, Devil Canyon reservoir would occupy the area between about RM 152 to 184 when full, a distance of acout 32 miles. Devil Canyon reservoir would vary in width along its length from about 500 feet to 3,800 feet, not including runup into tributary mouths. While Watana reservoir is on line alone, the river would extend continuously b ~low the dam . Once Devil Canyon reservoir comes on line, there would be no river reach between the two reservoirs when Devil Canyon reservoir is full; Devil Canyon reservoir would back up all the way to Watana Dam. Little drawdown of Devil Canyon reservoir would occur during the winter season, so there would never be much river reach between the two reserv oirs in winter (although, in the summer, drawdown is sufficient to produce about 6 miles of river reach between the reservoirs). Below Devil Canyon dam, there 33RD2-007t 111 would be a river reach in Devil Canyon of about 3 miles to the end of the zone at the canyon 's mouth. a. Watana filling During filling, winter water levels in Watana reservoir would be held constant. During the firs t winter of filling, the water level would be stable at an elevation of 1880 feet, while during th second winter of filling it would remain at about an elevation of 2100 ft. Ice cover formation , thickness , and meltout would be similar in character to those of normal reservoir operations (see below), except that no drawdown would occur, preventing the formation of draped ice ramps along tn reservoir shores. Below the reservoir, no significant ice would occur in the river upstream of Devil Canyon under average or warm climatic condi tion s . b. Watana only on line (1). Fr ~ezeup. Watana reservoir would generally begin t o freeze ove r sometime during mid November. Initially, very thin border ice would form along the shores of the reservoir, and then progress toward the center. At maximum ice cover development , the ice thickness on the reservoir would average about 3 to 5 feet. The ice cover would be relatively smooth. Drawdown of the reservoir during the winter would average ab o ut 90 feet , and would cause the ice near shore to fracture and become draped along the banks of the reservoir, creating an ice ramp surrounding the entire impoundm en t. The steepness and width of the ramp would vary depending upon local topog raphy . Narrow cracks would appear at topographic breaks in the ice 33RD2-007t I t!f L ramp. Where the ice cover comes in c ontact with the shore some bank erosion might occur, resulting in loca l ized increases in suspended sediments near shore. Below Watana reservo i r for the 36 miles to the end of the zone, river water temperatures would remain above freezing and the river would not freeze over all winter. Open water would prevail throughout the river in this zone. (2). Meltout/breakup. In the spring, meltout on the Watana reservoir would occur generally from early May to early June, depending on the climate . The ice would begin melting along the shores first. The melting ice cover would then fracture into numerous large and small pans o f ice. These pans would float around the reservoir until they melted away, probably being concentrated from time to time along the south and west shores due to winds. Draped ice above the water level would melt away more slowly. Some of it would be refloated and quickly melted away as the water level ri ses to store summer flows. Ice drifting from the upper river into the headwaters of Watana reservoir might jam against the ice cover there. This could cause river staging to occur just upstream of the reservoir until its ice cover begins to melt and retreat. However, since the reservoir would the~ b e al i ts minimum level due to drawdown, no flooding would probably occur outside the reservoir limits. Where the draped ice ramp caused by drawdown overlies tributary mouths entering the reservoir, the ice would probably become grounded and block the stream mouths. However, these tributaries begin flowing early in the spring, often sometime in ~rch, and the their warm flows would probably melt o r erode out the draped ice in the s tream mouth long befo re the reservoir itself began to melt out in April or May. 33RD2-007t c. Watana and Devil Canyon on line (l). Freezeup Watana reservoir would freeze over in a similar fashion and timing to tlte Watana-only scenario, again starting with very thin border ice along shore and then freezing toward the center. Ice thicknes s would again average about 3 feet at maximum ice cover development . During the winter, drawdown on Watana reservoir would average about 40 feet, fracturing the ice cover and causing draping of ice along shore. However, because the drawdown in this case would be less, the severity of draping and ice ramp formation would be somew~a t reduced. compared to th~ Watana-only scenario . Devil Ca nyon reservoir would begin freezeup during November. Durj ng early freezeup, border ice would probably form along the shoreline befc re freezing toward the center. During dry years, when water levels may be risjng during freezeup. ice near shore may be somewhat thinner than in the center. Ice thickness at the max imum development of the ice cover would average about 2-l/2 to 4 feet. Since no significant drawdown of the reservoir during t he winter NOuld occur, no nearshore ice fracturing and draping would occur. Dow~stream from Devil Canyon reservoir , a river reach of about 3 miles •>f open water would occu r to the end of the zone. This reach would all be with:.n th e confines of Devil Canyon . It is unlikely that any significant amount cf border ice would form with the r e latively warm water temperatures released from the dam within the turbulent canyon . /1'1 33RD2-007t (2). Meltout/breakup. In the spring, Watana reservoir would melt out in a similar fashion and timing to the Watana-only scenario. Devil Canyon reservoir would slowly melt out throughout May. There would oe little rise in water level and therefore little fracturing of the ice cover until it had decayed and thinned considerably by melting, at which time winds might fracture the remaining cover and drive the debris to downwind shores. 2. MIDDLE RIVER ZONE a. !_reezeup When Watana reservoir is filling, freezeup of the middle river would be delayed compared to natural conditions. Delay during the first year of filling would amount to about 7 weeks, while during the second year of filling it would be delayed about 5 weeks. The maximum upstream position of the ice front is predicted to be between RM 156 and RH 162. Ice thicknesses should be about one foot thinner than natural. Maximum river stages are expected to be 0 to 5 feet lower than normal during the first year of filling, and 0 to 3 feet lower than normal during the second year of filling. When the project comes on line, freezeup in the middle river is predicted to be delayed due to higher winter flows, warmer water temperatures, and reduced frazil ice input into the middle river compared to natural conditions. The advancing ice front would be delayed in reaching the Susitna/Chulitna confluence and freezeup in the middle river would be delayed by 17 to 44 days with Watana dam alone on line, and by 27 to 47 days with Devil Canyon dam also on line. 33RD2-007t Ice cover duration normally lasts 5 to 6 months under natura l conditions. Duration with-project would range from 3 to 5.5 months, depending on climate. With only Watana d am on line, the advanc ing ice front is pred icted to reach only to somewhere between RM 124 and 14 2 , while with Devil Canyon dam also on line, the ice front would reach somewhere between RM 123 and 137, depending on the climate and the operational scenario. No ice cover would mantle the river upstream from there; open water with temperatur0.s above 0 C would prevail all winter . The position of the ice fr on t would probably fluctuate significantly throughout the winter, a s weather conditions c hanged . When the ic ~ f ront is downstream f rom the 0 C isotherm, border ice would form between the ice front and the 0 C isotherm, cove ring approximately 25 percent of the water surface. Anchor ice would also form at times in this reach, but would have little effect on water leve ls . Upstream from the ice front no freez e up staging wo uld occur; ther e for ~ no sloughs would be overtopped or flooded in that area. Downstream from the ice front freez e up staging would occur, and the higher winter flows wo uld produce overtopping and flooding in more sloughs, more f requently, and for longer durations, than under natural conditions. ~lith Watana darn alone, freezeup staging is expected to be 2 t o 7 fee t higher than under natural conditions. When Devil Canyon comes on line , fre ezeup staging is predicted to be 1 to 6 feet higher than natural. The higher stage levels may be sufficient to flood som~ vegeta ted islands as ~ell as sloughs. Upstream from Lhe ice front, mainstern water temperatures would depend on the weather and the pro ject ope rating scenario. Just below the dams, water temperatures would range froM 0.4 C to 5.6 C f rom early Novem ber to late April. From there, temp era tures would decline downstream until r eaching the 0 C isotherm. With Watana dam operating alor.e, ice thicknesses below the ice 33RD2-007t front arc expected to be s imila r to natural conditions. With Devil Canyon dam on line, ice thicknesses are expecteo to be natural conditions. to 2 feet less than under The altere ~ flow regime would probably cause a change in t he lengths of open leads throughout the winter. As under natural conditions, anchor ice which h a d formed in the river before the passage of the ice fr o nt would sometimes be exposed in these open leads. b. Meltout/breakup In the spring, the river ice is predicted to slowly melt awa ~· iu place 4 t o 6 weeks earlier than natural breaku p with Watana alone, and 7 to 8 week s earlier than natural with Devil Canyon also on line. An y border ice would slowly melt away in place . There would probably be little or no breakup drive, with its associated ice jamming. Therefore, there would be no overtopping and flooding of sloughs during this period, and n o associated i c e scouring and fl ow erosion in the s loughs . 3 . LOWER RI VER a . Freezeup The ice cover is expected to begin forming in the l owe r part of the lower rive r early in Novembe r, ab o ut th e same as under natural conditions. Howev e r, the inc reased flows and redu ced frazil ice produ c tion from the middle river are predicted t o de lay the movemen t up s tream of the ice cove r s uch that it reaches th~ Susi tna/Chulitna rive r confluen ce 17 t o 44 days later than natural with Watana dam alone on line , and 27 t o 47 day s later than natural with Devil Canyon dam also on line. 33RD2-007t /i} Due to the increased flows from the middle Susitna River throughout the winter. it is speculated that there 1.10uld be somewhat more ice in the lower river than under natural conditions. This. however. has not been quantified. b. Meltout/breakup The lower river is expected to melt out sooner than it does naturally because of the earlier meltout of the middle river and the resulting earlier influx of warm water into the lower river. The middle river is predicted to melt out earlier than in natural conditions by 4 to 6 weeks with Watana alone, and 7 to 8 weeks with Devil Canyon also on line. Therefore, the lower river c ould be expected to melt out at similar times. The meltout should be gentle compared to natural conditions because there would be no breakup drive of ice into the lower river from the middle river. The ice i10 expected to slowly melt away in place, with no ice jamming and its associated scour and flooding. 33RD2-007t C. EFFECTS ON FISHERIES 1. INTRODUCTI ON This section describes the potential effects of ~hanges in Susitna River ice processes on fish habitats, resulting from various project operational scenarios. The Susitna River and its tributaries provide habitat for at least 19 species of fish (table 11); seven are anadromous and 12 are year-round res- idents. (Figure 89 shows the distribution of these species by study area), The Susitna River basin provides reproductive and rearing habitat for millions of salmon (recorded escapements vary from ove r 500,000 to more than 5 million, table 12). More than 99% spawn in its tribut a r y systems. Available informa- tion indicates that the mainstem Susitna R ~v er may provide essential rearing habitat for two salmon species--chum and chinook. Like salmon, resident species such as rainbow trout and Arctic grayling mostly u s e the mainstem for mjgration. Some resident spec ies also depend on the mainstem for overwintering habitats; others like burbot, whitefish, and longnose sucker reside year-round in the mainstem. This analys is examines the likely with-project ice effects of Watana Dam alone and of ~a tana and Devil Canyon dams together on fish habitat in three stretches of the Susitna River . These are kn own as impoundment zone, middle river zone, and lower river (figure 89). 1ne results of 21 ICECAL simulations are e xamined (table 13), four of which are natural (i.e. ·.Jithout project) and 15 of which are with-project and two are filling scen a rios. These simulations were run under various meteorologic and hydro logic conditions. Thirteen of the ~oJith-pr ojec t ICECAL simulations used Case C flow re gimes while two used Case E VI Flows. 33RD1-007 Table 11. Common and Scientific Names of Fish Species Recorded in the Susitna River Basin. Arctic lamprey Lampetra japonica (Martens) Eulachon (hooligan) Thaleichthys pacificus (Ricltards 'J n) Arctic grayling Thymallus arcticus (Pallas) Bering cisco Coregonus laurettae Sean Round whitefish Prosopium cylindraceum (Pallas) Humpback whitefish Coregonus pidschian (Gmelin) Rainbow trout Salmo gaird~eri Richardson Lake trout Salvelinus namaycush (Walbaum) Dolly Varden Salvelinus malma (Walbaum) Pink (humpback) salmon Oncorhynchus gorbus c ha (Walbaum) Sockeye (red) salmon Oncorhynchus nerka (Walbaum) Chinook (king) salmon Oncorhynchus tshawyt scha (Walbaum) Coho (silver ) salmon Oncorhynchus kisutch (Walbaum) Chum (dog) sa lmon Oncorhynchus keta (Walbaum) Northern pike Esox lucius Linnaeus Longnose sucker Catostomus catostomus (Forster) Threespine s tickleback Gas teros teus aculeatus Linnaeus Bur bot Lota lota (Linnaeus) Slimy sculpin Cottus cogna tus Richardson 33RC1/007g Table 12. Susitna River Salmon Escapement Estimates, 1981-1984 Year Chinook Socke:z:e 1 Pink Chum Coho Total 2 1981 272,500 85,600 282,700 36,800 677,600 1982 265,200 890,500 458,200 79,800 1,693,700 1983 176,200 101 ,300 276,800 24' 100 578,400 1984 250,000 605,800 3,629,900 812,700 190,100 5,488,500 1 Second run sockeye only. 2 Total 19 84 drainage escapement estimate. Esc~pement counts for 1981 through 1983 do not include chinooks or any escapemen ts into tributaries downstream of RM 77, with the exception of those into the Yentna River. Source: ADF&G 1983a; Barrett, Thompson & tHck 1984 and 1985. 33RD1-007i-1 Scenario Start o f freeze up TABLE 13 <;IMULATED SUSITNA HlDDU: RIVER ICE fRONT PROGRF.SSI ON 1 AND Heltout WINT ER OVER TOP P INC Of SLOUGHS Ha xioum Upstre a"' Ex tent of l e e Front Slough Overtopping Opi>ll l.a ter Dovns tr cam of Devil Canyon (RH 1 52) Date r•ate Ri ver Hile 8 SA 9 11 2 1 (RH 113 .7) (R.'f 125.1) (RH 128.3) (RH 1)5 .3) (RH 141.1) Natu ral 2 15 2A 1971-72 Nov 05 0 197 6-77 Dec 08 15 2 0 19111-82 Nov 18 Hay 10-15 152 0 1982-83 Nov 05 May 10 152 c 0 Watana Only 1996 Case C 1971-72 Nov 28 May 15 140 D D D 12 1976-77 Dec 26 Apr 18 126 D D 2n 1981-82 Dec 28 Apr 0) 137 D D D 15 1982-83 Dec 12 Mar 20 126 D 26 2001 Case c 1971-72 Nov 28 Hay 15 14 2 D D D 10 1981 -82 Dec 30 Apr 03 134 D D D 18 19 82 -83 Dec l'J Har 16 124 D 28 2001 Case E-VJ 1981-82 Dec 28 Har 23 134 D D 18 Watana and Dev il Canyon 2002 Case C l'lll-72 Dec 02 Ha y \l3 13 7 D f) 15 19 76 -77 Jan 08 Apr 14 124 28 1981-82 Dec 30 Har 12 124 28 1982-83 Dec 22 Mar 20 123 29 2002 Case E-Vl 1981-823 2020 Case c 1971-iZ Dec 03 Apr 15 13 3 D D 19 1982-83 Dec 14 l-f.a t 12 127 25 Wata na filling 1911 2-83 (YR I ) Dec 23 Hay 02 156 0 198 1-82 (YR2) Oec 23 ~ray 30 16 2 0 Legend: A -lee cover for natural condi tions extend s upst r ea~ of River HUe 137 to Devil Can yo n (lUI 152) by mean s of lateral ice bridging. 8 -Number o f miles of upen water from the i ce front upstream to Devil Canyon at River Hile 152. Sollie open lead s can be found in the ice cover. c -Observed natural overtopping. D -Slough is overtopped with project, but 110t under 5 ic1Ul a ted natural conditions. Notes: I. !CECAL Hodel Simul at i ons 2. Weather Conditions: 1971-72, co ld winter; 1976-77 , very warm winter; 19 81-82, average ..,inter; 1982 -8), wa re win ter . 3. Results una~ailable to date. Source: Harza/Ebasco Susitna Joi11 t Vent u re, 1985. 33RC1 /007e l. 2. 3. 4. 5. 6. 7. Table 14. SUSITNA HYDROELECTRIC PROJECT ICE ISSUES LIST FOR FISH. ISSl:E The effects of altered ice processes on salmon and resident fish habitats and populations down- stream of the dams, including fi s ~ access and changes due to staging. T~pacts on egg incubation, rearing and rearing cover, ~specially disruption of incubation or emergence timing from lower water temp e rutures and dewatering. Attention to middle river sloughs and side chann~ls are of primary impor- tance. The effect of ice staging on upwelling in slough and mainstem habitats. The level of ice cover. Could it effect primary productivity and rate of food production? Would it increase available overwin t er habitat? Importance of altered breakup on the forma - tion of ice jams and corresponding flushing of slough hab:!.tats. Ice process effects on staging overtopping. Changes in anchor ice formation. SOURCEl APA ADF&G/Su Hydro EWT EWT EWT AEIDC AEIDC 33RCl/007f -1 - UIPOUNDMENT ZONE MIDDLE RIVER ZONE X X X X X X X LOWER RIVER REACH X X X X ISSUE Table 14 (cont'd). SUSITNA HYDROELECTRIC PROJEC T ICE ISSUES LIST FOR FISH. SOURCE 1 IMPOUNDMENT ZONE MIDDLE RIVER ZONE LOWER RIVER REACH 8. Effect of ice shelf formation in the reservoir drawdown zone. AEIDC X 9. Increased amount of ice in the inundation area near tributary mouths impeding fi s h passage. AElDC X 1 Source: APA -Alaska Power Authority Issues List March 6, 1984 . 33RC1/007f ADF&C/Su Hydro -Alaska Department of Fish & Game/Su Hydro Aquatic Studies memorandum to AEIDC October 5, 1984. EWT -E. Woody Trihey & Associates memorandum to AEIDC October 2, 1984. AElDC -Arcti c Environmen tal Information and Data Cen t er plu s information from literature reviews o f hydroelec tric projec ts in northern environments. -2 - This analysis addresses nine ice-related issues of concern, which were identified by the Susitna Hydroelectric Study Group (table 14). The majority focus on middle river incubation and rearing habitats, which for the most part, are limited to sloughs and side channels. Concerns here rest principal- ly with the potential for increased frequency of overtopping of slough berms by ice-induced staging and with the potential for changes in the thermal regime of upwelling waters on natal habitats. Expressed concerns directed at the lower river are fewer in number. The project staff feels that this reach has less significant salmon spawning and rearing habitats and !CECAL modlers believe that with-project ice processes would change less dramatically here than in the middle river. Impoundment zone ice-related concerns focus on the potential for blockage of tributary stream mouths by ice and on the influence of ice on the littoral environment. 2. METHODOLOGY To conduct this analysis, AEIDC first reviewed the literature on high latitude water bodies (both regulated and unregulated) for information describing various ice condition effects on fisr. habitats. Relevant informa- tion was then synthesized to provide an overview of the type and scope of po- tential ice-related with-project effects which co1:ld occur in the Susitna Riv- er basin. Next, pertinent Susitna River-specific biological and physical in- formation was assembled. These two steps provided the basis for the ice ef- fects analysis, which was performed by comraring knowledge of fish overwinter life history stages to !CECAL simulations. This analysis was constrained by a number of factors. Chief among these was that ICECAL was designed prima rily to address river ice physical proc~sses .£!!_~(rates of formation, timin~ of freezeup, etc), r a ther than the effects 33RD1-007 of icing on the environment. Being one dimensional, ICECAL simply lacks the power to describe site-spe~ific ice processes in areas possessing multiple habitat types (e.g. side sloughs, mainstem, etc). Second, extant Susitna River basin data on fish distribution, abundance, and habitat uses focus on salmon and are temporally and spatially limited. Third, knowledge of the effects of various winter conditions on fish mortality is particularly scant. The fourth problem relates to the fact that issues were identified for evaluation primarily by a approach. Representatives of E. W. Trihey & Associates, Harza-Ebasco Susitna Joint Venture, the Alaska Department of Fish and Game, R&M Consultants, Inc. and AEIDC met on September 9, 1984 in AEIDC's offices to plan the ice assessment approach . It was agree d that participants would nominate ice related issues of concern to be addressed in the report by October 1, 1984. Subsequently, E. W. Trihey & Associates, Alaska Department of Fish and Game, and AEIDC responded. Offered issues of concern are listed in Table 14. Since no similar project exists in Alaska and time was of the essence, the project team relied on iciug information from other Northern areas to nominate issues. Because of this, issues raised are not necessarily wholly pertinent to the basin nor are they necessarily all encompassing. 3. IMPOUNDMENT ZONE EFFECTS ANALYSIS a. Fish Resource The principal source of information on fish distribution, abundance, habitat use, and life histories in the impoundment zone is ADF&G 1983b. The natural environment between Devil Canyon and the upstream end of the proposed Watana Reservoir provides habitats for nine fish species (ADF&G 1983b); eight are year-round residents and one (chinook salmon) is anadromous. Within Devil 33RD1-007 Canyon, Cheechako Creek (RM 152.5) and Chinook C~eek (RM 156.8) mark the up- stream limit of salmofl in the mainstem Susitna River. Devil Canyon's con- stricted river channel apparently creates a velocity barrier to upstream mi- grants. In total, fewer than 100 salmon utilize these two tributary habitats for reproductive purposes (Barrett, Thompson & Wick 1985). Arctic grayling are the most widely distributed and abundant species uti- lizing habitats above the c.anyon. The total 1982 Arctic grayling population in the impoundment zone was estimated to be over 1~,000 (ADF&G 1983b). Mainstem impoundment areas above the canyon provide essential overwintering habitat for Arctic grayling, which move into its tributaries to spawn following breakup in late May or early June (ADF&G 1983b). Arctic grayling migrate out of natal tributaries in September as water levels begin to drop. They overwinter in mainstem environments which become less turbid following freezeup (ADF&G 1983b). Except for documentation of their presence, little is known of the life histories or relative abundance of other species resident in the impoundment zone. Based on limited capture data, it seems that both burbot and longnose sucker are relatively abundant in the proposed impoundment areas (ADF&G 1983b). Elsewhere in the Susitna River, burbot spawn under the ice from January to February over gravel near tributary stream mouths (such as the Deshka River) (R. Sundet, pers. cotmn.). During the rest of the year, they apparently distribute themselves throughout the deeper portions of aquatic environments. Susitna River longnose sucker are spring spawners which move from overwinter habitats in the mainstem to tributary natal areas from late May to early June (ADF&G 1983b). Small numbers of round and humpback whitefish have been captured (at two locations) within the impoundment areas, but there are no estimates of their relative abundances. If they behave 33RD1-007 similarly to lower river reach and middle river whitefish, they overwinter in mainstem environments. Although available information is scant, it appears that this species spawns in early October in clearwater tributary streams. Although not present in mainstem impoundment areas, some lake trout and rainbow trout might gain access to them as a result of the project. Sally Lake, which supports a lake trout population of undetermined size, would be inundated by the Watana Reservoir (ADF&G 1984b). Lake trout generally spawn from August through December and require stable lake shore gravel substrates for reproduction. High Lake (located immediately north of Devil Canyon) is a tributary system to Devil Creek which has a resident population of rainbow trout. Nothing is known of the lake's trout population size. Should the project be completed, we believe that some rainbows might outmigrate down Devil Creek to the Devil Canyon Reservoir. Elsewhere in the basin, rainbow trout typically overwinter in lakes and mainstem habitats, returning in the spring following breakup to spawn in tributary streams. Most rainbow trout spawn in clearwater streams, which are paved with relatively small cobbles and have relatively moderate velocities (ADF&G 1983c). b. The With-Project Environment The following is a synopsis of selected aspects of the with-project envi- ronment that are rele•,ant to this ice effects analysis. (A detailed descrip- tion of with-project ice processes is found in Chapter IV.) The Watana Reser- voir would inundate roughly 55 linear miles of the mainstem Susitna River and about 30 miles of tributary stream environments, converting them to a lentic system. The Watana Reservoir would generally begin to freeze sometime in mid-November, with probable maximum ice thicknesses ranging from 3 to 5 ft. Winter reservoir drawdown would cause nearshore ice to fracture and drape over 33RD1-007 exposed banks. While on line alone, the Wat a na dam reservoir's winter draw- down would avera ge about 90 ft. With both dams drawdown would av~rage about 40 ft. In mid-winter, grounded ice would probably form barricade s at tribu- tary mouths. Based on observations of natural ice processes withi n the upper basin, project team members g ~~erally believ e that tributary flows would down- cut grounded ice before reservoir ice meltout. This would generally occur be- tween May and early June. The Devil Canyon Reservoir would inundate a maximum of 32 linear miles of mainstem Susitna Riv er environments. Freezeup times would be similar to those of the Watana Reservoir, but probable maximum ice thickness would not exceed four ft . Yearly winter reservoir drawdown would be slight if it occurred at all. Consequently , less ice draping would occur than in the Watana Reservoir. A few miles of open water may occur in the upper part of Devil Canyon Reser- voir due to the warm water released from the Watana Reservoir. 33RD1-007 c. Anticipated With-Project Effects (1). Watana Reservoir. Ice processes attendant to winter reservoir drawdown would affect reservoir salmonid spawning and rearing habitat quality. The littoral zone environment would experience periodic dehydration, substrate freezing, ice gouging, and erosion. Lake drawdown, coupled with ice draping, would prec lude evolution of a stable littoral zone conducive to lake trout (from Sally Lake) reproductive and rearing success. Lake trout reaching the i~~ounrlment would likely live a normal life span. The effects on other sal~onids would be less severe, because they spawn in tributary stream habitats. Thus, only their rearing life stages would be affected. Impoundment rearing habitats for Arctic grayling and whitefish would probably be less than ideal. Lake drawdown, ice draping, ice gouging, erosion and associated effects would likely reduce cover and food availability, because together they would preclude establishment of riparian vegetation and limit invertebrate produc- tivity. The effects of Watana Reservoir operation on burbot are more difficult to predict, because they have more generalized habitat requirements. They often inhabit deep, cold, and turbid environments. Although burbot can utilize lake shore gravels for spawning, most spawn in tributary stream environments. These would be unaffected with-project. Thus, the impoundment probably would not exert discernable negative effects on them. The impoundments littoral zone would not afford them viable reproductive habitat because of its unstable nature (see above). Ice blockage of tributary stream mouths should be a problem for fish only in extremely cold years, when spring ice mel tout is retarded. If climatic 33RD1-007 2./C . conditions match long term averages, the tributary mouths should be ice free before late May when Arctic grayling and longnose sucker migrate to instream spawning habitats . If spring meltout does not occur until after early June, both grayling and longnose sucker could experience reproductive failure that year. From a fish population biology standpoint, loss of a single-year class is not particularly troubling unless the population is being simultaneously stre .... sed by other factors such as epldemics or sport fishing. In Alaska, it 1a common for some local fish populations to have certain year classes predom- inate while others are absent or nearly so. Once the Devil Canyon Dam was on-line, the Watana Reservoir operations schedule could exert somewhat less influence on fish habitats because the ex- pected dr~wdown would be less. However, this point is moot, since its pre- dieted drawdown still exceeds 40 ft. For reasons discussed above, a drawdown of this magnitude would severely limit littoral zone productivity. (2). Devil Canvon Reservoir. Because of its smaller scale, winter drawdown of the Devil Canyon Reser- voir would be iess influential on impoundment littoral zone habitats than that predicted for the Watana Reservoir. Ice draping would be minimal (if it oc- curred at all) and ice gouging negligible given the bedrock substrate and lack of ice fracturing from extensive drawdown (Chapter V). Perhaps, importantly, impoundment area geomorphology and geology are such that they naturally con- strain the amount of potential lentic spawning habitat available. The can- yon's steep side walls and bedrock substrate severely limits potential use by spawning fish. For this reason it is unlikely that this reservoir would be a productive environment for fish. 33RD1-007 2.1/ Arctic grayling, buibot, longnose sucker, and (possibly) rainbow trout could gain access to the Devil Can y on Reservoir and become residents. None depend on lentic littoral zones for reproductive purposes. Arct~c gray ling, burbot, and longnose sucker naturally occur within the area to be iffipounded. Rainbow trout might gain access through outmigration from High Lake. Lake trout are not resident within the Devil Canyon impoundment area. They would have to gain access from the Watana Reservoir either by passing through the turbines, over the spillway , or through the gate valves. With-project ice blockage of tributary strean mouths should not be a pro- blem in this reservoir. The two main tributaries capable of providing repro- ductive habitats for the subject species (Fog Creek (RM 177) and Tsusena Creek (RM 181)] are located in the upper end of the reservoir where little, if any, ice accumulation is expected to occur (G. Gemperline, pers . comm.). Normal spring tributary meltout in this area would then easily wash out any remaining ice allowing timely access to spawning and rearing habitats for all reservoir residents. 4. MIDDLE RIVER ZONE FISH RESOURCE EFFECTS ANALYSIS a . Fish Resource A complete summary of available information on middle river fish resources is available in a report by Woodward Clyde Consultants and Entrix 1985 . Sixteen fish species are known to inhabit middle river zone waters (figure 89). All are ultimately dependent on mainstem environments for some aspects of their life histories. salmon) are anadromous. 33RD1-007 Five of the fish species present (all Salmon utilize mainstem river environments for migration, rearing, over- wintering, and to a lesser extent spawning (Woodward Clyde Consultants & Entrix 1985). An indication of the importance of middle river mainstem habitats as a travel corridor to returning salmon adults is found in escapement counts made at the ADF&G fish wheel stations at Carry and Talkeetna (table 15). Population estimates (based on Talkeetna station data) for 1984 indicate that approximc:ttely 6% of all coho, 12% of all chum, 2% . of all sockeye, 10% of all chinooks, and 5% of all pink salmon spawning in the entire Susitna drainage basin travel through the mainstem middle r~ver to reach ·their natal grounds (table 5). Adult migration timing varies by species, but in general the peak inmigration time above Talkeetna is from late June through the end of September (table 16). At least eighteen tributary streams in the middle river provide salmon 1'7 spawning habitats (table 17). Viewed as a whole, most salmon spawn in tribu- tary streams (which would be unaffected by the project). Based on escapement counts for 1984, 34 middle river sloughs collectively provided habitat for ap- proximately 5.5% of all salmon migrating above Talkeetna station (Barrett, Thompson & Wick 1985). Coho and chinook in this river reach apparently spawn only in tributary stream environments, pink salmon primarily in tributary streams (with a small number utilizing slough habitats) chum salmon in both tributary and slough environments, and sockeye spawn almost exclusively in sloughs (Barrett, Thompson & Wick 1985). Despite their relative importance to both chum and sockeye salmon in the middle river, slough spawning habitats are not central to the maintenance of the total Susitna River stocks of either species. Only about 2% of all chum and less than 0.5% of all sockeye spawning in the drainage in 1984 utilized these sloughs for this purpose. 33RD1-007 213 tABLE lS. SUSITHA RIVER SAL~ ESCAPEKEHT FOR tHE HIDDLE SUSITHA RIVER, 1981·84.1 Supllna River Chinook2 Pink Chum Sockev~ Coho ~catlon Hlle 19b2 1983 1984 1981 1982 lq83 1984 l981 1 982 1983 1984 1~81 1982 1983 1984 1981 19112 198) 1984 talkeetna 10) 10,900 11t,SOO 24,S91 2,300 73,000 9,SOO 177,881 20,800 49,100 S0,400 98,236 4,800 3,100 4,.,00 U,OSO 3,300 S ,lOO 2,400 11,847 Station Percent of total Susltna Eecapellent 9.8 2. 7 8.2 9.3 4.9 7.4 10.7 18.2 12.0 1.8 1 .2 2 .4 2.2 8 .9 6.4 9.9 6.2 CUrry 1 20 11,300 10,000 17,lS1 1,000 S8,800 s,soo 116 ,8S8 13 ,100 29,400 21,100 49,278 2,800 1 ,300 1,900 3 ,S93 1,100 2,400 800 2 ,162 N Station ' Percent of 3 ~ t otal Su tltna Elcapement 6.9 1.1 6.6 S.4 3 .2 4 .6 6.4 7.6 6.1 1.0 o.s 1.1 0.6 ).0 3.0 '· ~ 1.1 ~otal baein eecapeMent data for chinook are only available for 1984. 3 Eecape~~ent nuabeu for 1981·83 du not Include any count• Into tr1butar1u dovnotre'"" of RH77, vith the exception of the Yentna River. Source • ADFt.C 1983&1 Sarrett , th011peon and Wick 1984, 198S. 33RC 1/007d Table 16. Susitna River Salmon Phenology. DATE HABITAT RANGE PEAK CHINOOK (KING) SALMON Adult Inmigration Cook Inlet -Te lkeetna Hay 25 -Aug 18 Jun 18 -Jun 30 Talkeetna -D.C. Jun 07 -Aug 20 J un 24 -Jul 04 Middle River Tributaries Jul 01 -Aug 06 Juvenile Migration Middle River Hay 18 -Oct 031&3 Spawning Middle River Tributaries Jul 01 -Aug 26 Jul 20 -Jul '27 Lower River Tributaries Jul 07 -Aug 20 Jul 20 -Jul 27 N "' COHO (SILVER) SALMON ~ Adult Inmigration Cook Inlet -Talkeetna Jul 07 -Sep 28 Jul 27 -Aug 20 Talkeetna -D.C. Jul 18 -Sep 19 Aug 12 -Aug 26 Middle River Tributaries Aug 08 -Sep 27 Juvenile Migration Middle River May 18 -Oct 121&3 May 28 -Aug 21 Spawning Middle River Tributaries Sep 01 -Oct 08 Sep 05 -Sep 24 Lower River Tributaries Aug 08 -Oct 01 CHUM (DOG) SALMON Adult Inmigration Cook Inlet -Talkeetna Jun 24 -Sep 28 Jul 27 -Aug 0 2 Talkeetna -D.C. Jul 10 -Sep 15 Aug 01 -Aug 17 Middle River Tributaries Jul 27 -Sep 06 Middle River Sloughs Aug 06 -Sep 05 Juvenile Migration Middle River May 3 18 -Aug 20 May 28 -Jul 17 33RC1/007h - 1 - N ' t' Table 16 (cont'd). Susitna River Salmon Phenology. DATE HABITAT RANGE PEAK Spawning Middle River Tributaries Jul 27 -Oct 01 Aug 05 -Sep 10 Middle River Sloughs Aug 05 -Oct 11 Aug 20 -Sep 25 Middle River Mainstem Sep 0 2 -Sep 19 Lower River Tributaries Jul 27 - Se p 09 Aug 06 -Aug 14 SOCKEYE (RED) SALMON 2 Adult Inmigration Cook Inlet -Talkeetna Jul 04 -Aug 08 Jul 18 -Jul 2 7 Talkeetna -D.C. Jul 16 -Sep 18 Jul 3 1 -Aug 05 Juvenile Migration Middle River May 18 -Oct 111&3 Jun 22 -J ul 17 Spawning Middle River Sloughs Aug 05 -Oct 11 Aug 25 -Sep 25 PINK (HUMPBACK) SALMON Adult Tnmigration Cook Inlet -Talkeetna Jun 28 -Sep 10 Jul 26 -Aug 0 3 Talkeetna -D.C. Jul 10 -Aug 30 Aug 01 -Aug 08 Middle River Tributaries Jul 2 7 -Aug 23 Middle River Sloughs Aug 04 -.\ug 17 Juvenile Migration !>Iiddle River May 18 3-Jul 24 May 29 -Jun 08 Spawning Middle River Tributaries Jul 27 -Aug 30 Aug 10 -Aug 25 Middle River Sloughs Aug 04 -Aug 30 Aug 15 -Aug 30 Lower River Tributaries Jul 27 -Sep 09 Aug 06 -Aug 09 1 2 All migration includes migration to and between habitat. n o t just outmigration. 3 Second run sockeye only. No data available for pre-ice movement; earlier date of range refers to initiation of outmigrant trap operation. Source: Barrett. Thompson and Wick 1984. 1985; Schmidt et al. 1984; ADF&G 1983a.c. 33RC1/007h -2 - r.uu n Ptalr. S.l~ Sur'"#ty Coun t a Abooft Ta lk.Htn. for Sualtu lhn Tr l tM.atu y St r t •a Sli~VEY S'IIIUI< ~ Coho CblnooiL 1914 1976 1981 1981 198) 1954 19lS 1916 19 71 1918 1919 1981 1982 l9•l ·-..tlisu u 0 .2S 21 10 116 u s l01 22 •' Cru ll (lUI 101.4 ) ....... O.lS 40 80 )6 u 2)9 IS Crt ~ll (ltll 10...9) .ln ... o.n ,, .... CIUI 11 1.2) """' 1 .0 141 ,. 19 2 :14 Creek (R."t l ll.•) l Ane o.s 24 40 •1 12 2l Cre ek Cl.~ 111.6) L-r 1 .5 S6 lll 18 24 McKenzie U ll 116.2) "'.CI(enzle 0 .2S Cnco Cl.~ U6.1) l l t t le o.n Po rta .. (lUI 111.1) rt(CI\ 0.2S 11 of Ju.ly (lUI 121.1 ) Sku ll 0.25 Cn c o Ull 124.1 ) ........... 0.2S CrHO (lUI llO.I) J."nur rh 0,2 5 26 11 4 14 S6 ., o f July U ll U 1.0) C..>ld 0 .2S 21 2l 21 Cnck !M 1l6.1l lncU .n n .o .. lO as 101 SJ 46S lD sn )9) 114 liS 4 22 1 ,0Sl 1,19) 1,4 S6 ., .. , (lUI 118.6) Jac k 0.25 6 ...... !lUI 144 .5) ;·on: a,. 1S.O uo 100 22 81 n 128 29 102 ,. 140 140 6S9 1 ,2Sl l,l.O 5,444 c ..... (lll 1•1.9) o. .. ct..lr.o 3.0 16 25 2t Creek (lll U2,S) .:hlnooiL 2,0 • I lS CrH O (lll U6.1) MAL )01 141 •sa 6 )) 240 1,434 6 2 1 ,261 761 n• •2s 1.1 ~1 2,47 ) 4 ,416 1 ,171 • t.C:l/0071'"1 217 TAaLI l '1 (cont 'd) ~ak Sa..._ s..arv., Counu Abcwe t a lkeetna for SulltN lt"r Tributary Stn•• s..vn <1'1I1AII ~ a-. Sod nn 1•n lt16 19}1 1 .. 1 1912 191) 1-197. l 9n 1976 nn 1 .. 1 19ft2 1"'81 ~-- Y\lU.ra o.u c.--• <.., 101.•1 O\ON 0.2S CAd. CAPt 10.. 9) s••~h o.n c~ck (11:1 111.2) r~.,. 1.0 Cre-ek Ca. lll.U ...... o.~ 76 11 ]1 Cnc k (Ill 111.61 '--< 1.~ 14 u """'"''" (lUI 116. 2) '<Yule 0.2S Cnok 1111 116.7) L.tttle o .u ll 11 Port .. • ( ... 111.71 Fl ttfil 0.2~ of JuiJ ( ... 121.7) Sloull o.u 10 Cre~k (Ill U". 7) 51" .. .,..,. 0.2~ Cn•ll ( ... llO.I) rou uh o.u ~ .. 71 11 90 191 ... ltl ot J11ly ( ... 1)1.0) ...... o .u Crttk (M ll6. 7) ln4h" n .o HI 70 114 776 40 1,)1.6-Ill 2.24'1 Rhu (a.~ 111.6) ~·~· o.u l.ona (IL~ .... ~) rorup u.o 216 )0() IU ~26 l 1 21S ., , .... ( ....... , U.Cechako ).0 C<Hk (M Ul.~) O.l.....,. ... 2.0 Cnok (M I SO.I) MTAL J ,.01 1) ~ll 719 2•1 1,716 1 .... l,llt. .. 1: l*=l /0071•2 2tY tAIIZ 17 (cont'd) Pult S.lMn Survey Cowlu Above Ta l lleetu for ~ltn. llver Tributa ry Stre.-a SL'aVIT S'!UNI !!!!!!!S Pink 1974 1975 1976 1977 1981 1901 1901 .,.. l.ttlallert 0.25 a 1:18 193 Creek (R.'t lOl.«t ) n~n Cret'll. CRM 10..9) 0.25 so ll 107 4)1 s t .. :n Cr e e k CaM U l. 2) 0.75 C.:.th Crt<!k Cltll 111.61 1.0 lAne Crull UK U 3.6) 0.5 12 106 1,101 291 21 1 ,184 ~ •• ,"tt c .... (ltll 115.6) 0.2S 107 Lowtr 1.5 Zl 17 S6~ NciCtonJle (liM U6.2) ~len. tot 0.2S 17 11 t:ruk CR.'I 116.7) Uu.l• 0 .25 140 101 Porta ae (all 117.7) f'lrfo.nthorM 0 .25 ))7 Cr «lr. (lUI 120.1) fLtth 0 .25 111 411 o f July (all Ill. 7) S&ull Crc~ll Call 124 . 7) 0 .25 12 111 'l.rraan Cru ll (ltll UO.I) 0.25 24 ·~ t~n h 0.25 159 4 ,000 612 29 70 2 71 1 ,14:' o f July (aM 1)1.0) Cold r. .... (all 116. 7) 0.25 32 11 ~2 India n ll .. r CM 111.6) u.o 577 321 5,000 1 ,6 11 711 ... 9 ,0.6 J a t.ll Lor., (R.'I 1 .... 5) 0.25 •• rortaa• lS.O 211 3 ,000 169 215 2 ,101 , .... (ltll 148.9) O..ectt.ko 3.0 u Creek laM 1S2.5) 011-.k 2 .0 c .... (ltll 1S6.1) totAL 1 ,036 5 75 12,157 3 ,326 311 2,15S 1,)29 1 7,417 Soutc•: brrett 191t.; l.arrett, ~ton and Wl clt 1961., 191S; Rill 1977; ADh.C 1976, 1971, 1911, l983e. llRC1/007a • 3 2t1 Spawning habitat quaiity apparently varies greatly between sloughs as. in the last four years. the majority of chum salmon spawners counted were in 10 of the 34 (table 18 & 19). Three of the 10 most used sloughs have added sig- nificance in that they also provided over 90% of all sockeye spawning habitat in the middle river (table 18). Relatively few salmon spawn in mainstem non-slough habitats. of those which do chum salmon predominate. Generally. spawning habitats within the mainstem proper are small areally and widely distributed. In 1984. ADF&G made a concerted effort to identify mainstem middle river spawning habitats; they identified 36 spawning sites . Numbers of spawning fish counted at each of these sites varied from one to 131 with an average of 35 (Barrett. Thompson. & Wick 1985). Four of the five salmon species use middle river waters for rearing pur- pc ·ses (Dugan et al. 1984). At this time insufficient information exists to characterize the relative importance of individual mainstem rearing habitats. From May to September juvenile chinook rear in tributary and side channel environments. coho mostly rear in tributary and upland sloughs. and sockeye are evenly distributed between upland and side sloughs. From May to July rearing chum juveniles are distributed throughout side slough and tributary stream environments (Dugan et al. 1984). Of the five salmon species present. only two have been captured in winter in the middle river (ADF&G 1983c). Preliminary studies indicate that significant numbers (perhaps 25 % to 50%) of chinook and c oho juveniles reared in this zone overwinter in side slough and tributary stream environments (ADF&G 1985a). Perhaps significantly. preliminary evidence indicates that few juvenile salmon utilize the mainstem proper for overwintering purposes (ADF&G 33RD1-007 220 TASU: 11 Oft.lt soat£YE >I "" uvu ~ !lli!-.!!lli ~ !lli .!.!!.! !!!! !.!!1 ill! ~ !.ill .!!!! !!!! !!.!.! !!!! !.!!! !!!! ~ !lli !!!! .!.!!! .1.!!!2 !.!!.! ~ 99.6 12 10 100.4 21 ., 129 •• 101.4 50 I n u 20 ,. )A 101.9 17 11 .. :'•lkHtr...l S t. l O).n 105.2 101.1 108.2 -· 112.) 11 I' ll.J.l Jll. 1 )02 aS 2S l'u•t'lro4 117.1 90 10 Curry St. 120.0 .u 121.1 2) 49 ~( 121.9 41 4 121 1& 122.2 ao 104 400 .. .._ .. 12). 5 107 21 .. 16 2 2 n .. 114.6 140 77 1U , . 114.7 )4 2 2 ,. 125.1 S1 no ))6 17 917 10 177 .. 66 121 21 1)<. 1U.I sa 108 • 2 ' ll • 121.1 su 111 )6 lf>O )00 169 ISO 10 s 1J u 129.1 90 7) 11 •• 1)).1 Ul n1 105 )01 10 111.1 )6 l1 115. I )) " 116 411 459 211 l,S86 7t 84 75 n• 191 456 I loa 564 Ill 1 ~1 II 115.4 l l 115.7 11 , .. 115.9 15 117.2 100 Ill 500 ,. 117. I 12 u 11 " 111.9 24 II 21 90 66 16 :d 119,1 11 ,,, 119.7 • 45 )2 II 11 I 10 lloO.O 107 II 14 )0 u uo 10 2 .. as II lU.l 661 250 )0 )04 ll• 7)6 II' 2 ,lS4 II 75 21 )8 S) 197 112 .. ~u 145.5 10 ll 1 ... 5 11.4 U1 roru 1,152 49S " 541 1,596 2,2&.4 1,4» 1 ,S41 10) 194 114 )00 l,l~tl 607 S55 tn II 71 507 ' l,O.'t Soll.ru; t.rren 1974; 141nttt, n..p1on •M Vl ck 1964, 191S; lll1, U7l; ADF&.C 1976, 1978 . 1911, 1911o. llAC1/00lb-1 22.1 River Sloush Mile 8 113.7 8B 122.2 Moose 123.5 A' 124.6 8A 125.1 9 128.3 9A 133.8 11 135.3 17 138.9 21 141.1 TABLE 19 Chum Salmon Escapement for the Ten Most Productive Sloughs Above RM 98.6, 1981-83 . 3-Year 1981 1982 1983 Averas~ 695 0 0 232 0 99 261 120 222 59 86 122 200 0 155 118 480 1,062 112 551 368 603 430 467 140 86 231 152 1,119 1,078 674 957 135 23 166 108 657 1,737 481 958 Source: Barrett, Thompson and Wick 1 98 4. 33RC1/007c 2'22... Percent of Total Esca2ement 5.6 2.9 2.9 2.8 13.2 11.2 3.6 23.0 2.6 23.0 1985a). This helps somewhat to demonstrate the significance of sloughs to these species. Of the 11 resident middle river fish species (figure 89), capture data indicate that only rainbow trout, Arctic grayling, burbot, round whitefish, longnose sucker, and slimy sculpin are common (ADF&G 1983c). Dolly Varden, humpback whitefish, threespine stickleback, and Arctic lamprey also occur, but all appear to be more abundant in the lower river (Sundet and Wenger 1984). Lake trout are found only in surrounding area lakes, none of which would be influenced by the project. Little is known of either the numbers or the life histories (especially during the winter) of any fish species residing year-round in the middle river. Given the naturally reduced winter flow regimes of tributary streams in winter, it is probable that the majority of the~e resident fish (with the exception of lake trout) overwinter somewhere in the mainstem. It is general- ly believed, however, that most resident fish overwinter further down stream in the lower river (ADF&G 1983c). Of the most common resident species, three (burbot, longnose sucker, and slimy sculpin) :---ur year-round in the mainstem. Rainbow trout, Arctic gray- ling, and round whitefish spend most of the open wa~er season in tributary en- vironments which provide spawning and rearing habitat. Aspects of the winter life histories of these species (with the e x ception of slimy sculpin) pertain- ing to this analysis have been discussed previously. Too little is known of Susitna River slimy sculpins to adequately describe their habitat needs. How- ever, it is known with some certainty that they are distributed y ear-round throughout lentic and lotic environments within the basin, and that no large- scale movements or migrations have been noted . Spawning p~obably occurs around mid-June (ADF&G 1983c). 33RD1-007 b. The With-Project Environment The with-project Middle River Zone ice environment would differ dramati- cally from natural conditions. Formation timing of a contiguous river ice cover would be delayed, there would be an extensive reach of ice-free water below Devil Canyon, river flows would be four to five times more voluminous, and ice meltout would be earlier. These changes would occur as a consequence of dam interception of mainstem frazil ice input, increased wint~r flows due to the reservoir's operating schedule(s), and warmer than normal instream win- ter temperatures (the reservoirs would function as heat sinks). Middle river freezeup is predicted to be delayed between 17 to 44 days with Watana Dam and 27 to 47 day s with both dams in place. Depending on the year and with only Watana Dam on-line, the ice front is predicted to range somewhere between RM 124 and 142 and ice thickness below the front would be similar to natural. With both dams operating, the ice front is expected to range between P~ 123 to 137 and ice thickness is expected to be less than for natural conditions. It would likely be dynamic, changing location signifi- cantly throughout the winter in response to changes in weather conditions and with-project flows. Upstream of the ice front, no ice cover would mantle the river; open water with temperatures above 0 C would prevail throughout the winter. Between the 0 C isotherm and the upstream edge of the ice front, a zone of anchor ice form~tion would occur; no anchor ice would form upstream of this zone . Portions of the river near the ice front would be subject to free z eup staging, a phenomena which occurs as flowing water encounters the rough bottom surface of the ice mantle. When this happens, water velocity slows and the water sta ge rises (Chapter IV). Staging generally lasts one to two weeks un- der natural cond itions and could last a month or more with-project. 33RD1-007 Regardless of the final reservoir operation regime adopted, winter flow volumes would increase significantly with-project. Consequently, the aquatic instream environment would be substantially greater in extent, i.e. the wetted area would be increased. Because no ice staging would occur in the open water reaches immediately below Devil Canyon, no winter flooding is anticipated in this area. However, localized flooding would occur within ice-mantled river reaches, since higher than natural flows would be coupled with ice-induced staging . Higher flow volumes might also increase the length of open leads in ice-mantled areas downstream of the ice front. From early November to late April with-project water temperature is pre- dicted to range between 0.4 C to 5.6 Cat the dam outlet (AEIDC 1984). Water temperatures are expected to decline relatively uniformly downstream, reaching the 0 C isotherm near the ice front. The with-project springtime environment would differ from natural condi- tions chiefly in breakup phenology . Predicted higher than normal stream flows and instream temperatures would cause a gradual in-place melting of the ice mantle; there would be no breakup drive. Ice mel tout is expected to occur four to six weeks earlier than natural with the Watana Dam and seven to eight weeks earlier with both dams operating. c. Anticipated With-Project Effects Seven of the nine ice-related iss ues of concern identified for the pro- posed project relate to the middle river (table 14). Most are interrelated. Chief among these are concerns with slough incubation and rearing habitat quality. One deals with the potential introduction of near freezing water to slough incubation and rearing environments through ice-induced overtopping . Other issues concern the r otential of with-project flows altering the Z3RD1-007 character of upwelling waters and the lack of with-project ice cover upstream of the ice front. Potential effects in the ice-free reach include increased primary production (as more light penetrates the ice-free water surface) and increased overwinter habitat (as a result of higher than normal with-project winter flows). Of lesser concern, are issues pertaining to the with-project end of the natural cycle of breakup-induced flooding of slough habitats, and the a!llount of with-project anchor ice. Natural breakup-induced floods are thought necessary by some project team members to flush fines from spawning grounds. When anchor ice breaks up, melts, or otherwise disperses, it dis- lodges considerable amounts of substrate which can be life threatening to de- veloping embryos. Overtopping of slough b e rms occurs naturally during freezeup as a result of ice-induced staging and during breakup as a consequence of ice dam forma- tion. It is believed to influence overwinter embryo mortality in the middle river (ADF&G 1983d). Overtopping from freezeup-induced staging is the most troublesome to salmon, because it would introduce colder ~han ambient mainstem waters to developing embryos, for relatively long periods of times. During the period of incubation, survival of developing embryos naturally varies greatly and is dependent on several factors. The principal natural phenomena inducing embryo mortality are freezing of the spawning habitat, redd desiccation from dropping water levels, changes in the thermal and chemical characteristics of groundwater, and silting of redds (Buklis and Barton 1984, Canada Department of Fisheries & Oceans 1984). Dewatering and freezing of salmon redds have been identified as the principal factor inducing chum salmon embryo mortality in the middle Susitna River (ADF&G 1985b). Natural mortality is generally high during incubation; reported survival rates from North Amer- ica and Asia range between 1.5% to 30% (Buklis and Barton 1984; McNeil 1980). 33RD1-007 Temperature ranges that cause increased mortality to embryos are much narrowe r than those for adults (Alabaster and Lloyd 1982). Generally, the lower and upper temperature limits for successful initial incubation of Pacific salmon eggs fall between 4.5 and 14.5 C (Reiser and Bjornn 1979 ). Salmon embryos are most vulnerable to temperature stress in their early devel- opment stages, before closure of the blastopore . This occurs at about 140 ac- cumulated Celsius temperature units (Combs 1965; Bams 1967). (A t empe rature unit is one degree above freezing experienced by developing fish embryos per day). Merrell (1962) suggested that pink salmon embryo survival in Sashin Creek, southeastern Alaska, may be related to water temperature during spawn- ing. Embr yos exposed to cooler spawning environmental temperatures have been shown t o experience greater incubation mortality than those which began incu- bation at warmer temperatures (McNeil 1969). Bailey and Evans (1971) reported an increase in pink salmon mortality when initial incubation water tempera- tures were held below 2 C during the initial incubation period. Laboratory experiments with developing Sus:l.tna chum and sockeye salmon embr yos resulted in increased mortality and alevin abnormality when average temperatures were maintained at a level less than 3.4 C (Wangaard and Burger 1983). However, these increases were relatively slight. Following the period of initial sen- sitivity to low te-mperatures, i.e ., after the blastopore has closed (approxi- mately 30 days at 4.5 C), embryos and alevins can survive temperatures near 0 C (McNeil and Bailey 1975), but their development is slowed . During the incu- bation period, mean intragravel water temperatures in the primary middle river spawning sloughs range from 2 . 0 to 4. 3 C (ADF&G 1983d). Since peak chum s a lmon spawning in sloughs occurs between late August and September (table 16), it follows that blastopore closure occurs by October. 33RD1-007 Slough 8A was naturally overtopped in late November 1982 by cold mainstem water (near 0 C), providing some insight into potential effects of with- project overtopping events. Slough 8A intragravel water temperature was de- pressed during this event. Subsequently, embryo dev~lopment and emergence was delayed, and large numbers of dead embryos were seen (ADF&G 1983d). This sug- gests that increased mortality may have occurred. The significance of with-project overtopping to developing salmon varies between sloughs, being more p~oblematic in those downstream of the predicted ice front. As noted above, the predicted ice front location with the Watana Reservoir occurs between RM 124 to 142 (table 13). When it is at RM 124 (the farthest downstream ice front location predicted with the Watana Reservoir), none of the sloughs upstream of this point would be overtopped (table 13). Of the five most productive chum salmon sloughs in the middle river, only slough 8 is located downstream of RM 124 and would be overtopped. An average of 232 chum salmon spawned in slough 8 between 1981 and 1983 (table 19). T\,is represents approximately 5. 6% of the total chum salmon escapement to middle river sloughs for those three years (table 19). At the other extreme, when the predicted ice front is RM 142, three of the top five chum salmon producing sloughs (8, SA and 11) would be overtopped (table 13). From 1981 to 1983, these three sloughs supported an aggregate average of 1, 740 spawning chum salmon, approximately 42% of those spawning in middle river sloughs (table 19). Predicted river freezeup dates with the Watana Reservoir range from November 28 to December 30 (Harza-Ebasco Susitna Joint Venture 1984). Ice formation in all simulations is assumed to begiu at the Chulitna River confluence and progresses upstream from there. Of the eight !CECAL simulations run, six predict overtopping of sloughs 8 and 8A. three pr~dict 33RD1-007 overtopping of sloughs 9 and 11, and none predict overtopping of slough 21 (table 13). The expected rate of ice front progression upstream from the Chulitna River confluence varies annually due to climatic influence and temperature of the outflow. With the Watana Reservoir, ice front advance is predicted to take between one to six weeks (Harza-Ebasco Susitna Joint Venture 1984). Given the predicted start of river freezeup (late November) and the predicted rate of ice front advance, the earliest an overtopping event could occur is early December, which is generally post-blastopore closure. Most model runs indicate that freezeup start dates would be later, occurring in mid to late December (table 13). Therefore, the majority of predicted overtopping events from ice staging could not occur before late December and perhaps not until sometime in January . Based on !CECAL simulations of river freezeup timing, subsequent ice front advance, and what is known of the relationship of temperature to chum salmon embryo development, with-project ice-induced overtopping events could lead to widespread embryo mortality in affected sloughs . While the likelihoud of any direct embryo mortality from thermal stress diminishes after October following blastopore closure, some I CECAL simulations predict that staging overtopping even~s could last until spring meltout. Indirect mortality could be significant given that cold t~mperatures of this severity (near 0 C) and duration should delay embryo development and fry emergence to such an extent that they would be unable to complete their life cycle. The env ironmental consequences of ice-staging overtopping events appear to be less with both dams on-line. This is because initial freezeup dates ar~ predicted to be later, meltout dates are expecred earlier, and ice thickness would be less (see Chapter V). Further, the predicted duration of overtopping events is shorter, and they wot .ld occur later in winter. 33RD1-007 According to the two-dam ICECAL simulations, only sloughs 8 and 8A would be overtopped due to ice staging (table 13). Together, these two sloughs ac- counted for about 19 % of all chum salmon spawning in middle riv er sloughs from 1981 to 1983 (table 19). Importantly, only the "cold wint e r" simulations, which represent environmental extremes, predicted overtopping. Preliminary evidence indicates that (at least in sloughs proximal to the mainstem, e.g. 8, 9, and 11) intragravel water temperature is somewhat influenced by mainstem water temperature (Beaver 1984). overtopped slough environments . If true, it would serve to further cool Overtopping of slough berms by colder mainstem waters could also affect overwintering fish as water temperature affects fish metabolism, growth, food capture, swimming, and disease resistance. Juvenile salmonids can tolerate a wider range of water temperatures than embryos and can survive short exposures to temperatures which could ultimately be lethal. They can live for long periods at relatively l ow temperatures at which time they a bstain from feed- ing, are less active, and spend more time resting in secluded habitats (Ala- baster and Lloyd 198 2 ; Chapman and Bj ornn 1969). For example, in Carnation Creek, British Columbia, fish stopped feeding and moved into deeper water or clo ser to objects providing cover at temperatures below 7 C (Bustard and Narver 1975). Similarly, in Grant Creek near Seward, Alaska, juvenile salmon- ids were inactive at water temperatures between 1.0 to 4.5 C and inhabited cover afforded by streambed cobbles (AEIDC 1982). Regardless of whether one or two dam s are on-line, some slough overwintering fish would be exposed to colder overflow waters. As mentioned above, the chief difference between the one and two-dam options in this regard lies in the frequency and duration of overtopping events . 33iU>l-007 2 Jo Overwintering salmonids exposed to cold overflow waters (near 0 C) could respond in one of two ways, given that a critical thermal minimum has not been demonstrated in them short of actual freezing (AEIDC 1984). They conceivably might simply seek cover within the slough, becoming relatively inactive until temperatures once again rise following the end of the overtopping event. Al- ternately, they might elect to leave since they are mobi!.e. However, given that overflow water temperature would be identical to mainstem temperature, it is arguable whether they would do so. If they did emigrate, their survival would ultimately depend on their finding suitable replacement habitat which appears limited in this reach. Overtopping of slough berms from breakup-driven ice jams is not a with- project issue, given !CECAL predictions . According to model simulations. river ice would melt in place rather than breakup. Thus, no ice jams are pre- dicted to form at this time and no flooding of slough environments would oc- cur. The next concern to be discuss ed is the effect of with-project ice- staging on upwelling water in middle river spawning sloughs (table 14. Maximum winter river stages upstream of the with-project ice front are predicted to be lower than corresponding natural conditions, because there would be no freezeup staging (Harza-Eb asco Susitna Joint Venture 1984). Hence, there is concern that this low ~r stage could reduce the amount of slough upwelling. This should be of minimal concern since with-project winter flows upstream of the ice front with either dam scenario are predicted to be similar to those occurring naturally in September. As upwelling is presently sufficient for incubation purposes during natural September flows, one could assume that with-project upwelling would also be sufficient. Downstream of the ice-front, with-project river stages with both dams on-line are predicted 33RD1-007 2..31 to be higher than natural. Consequently, concern over project effects on upwelling rates in this zone are apparently moo t. The third issue raised deals with the potential effects of the with- project open wat er zone below Devil Canyon on fish habitat quality (table 14 . Regardless of whether one or two dams are built, an ice-free zone of open water would occur each winter below Devil Canyon. With Watana Reservoir, this (predicted b y I CECAL) would stretch between 10 to 28 miles; with both dams operational the zone would stretch between 15 to 29 miles (table 13). Conceivably, primary productivity could be enhanced in this area because there would be less snow and ice cover. Taken by itself, ice removal would allow more light to penetrate the water column, stimulating primary production. However, the question is complicated by the fact that released reservoir waters would be turbid, whereas natural winter flows are relatively clear (Acres American 1983). question. Estimates An ongoing study seeks to answer of released water turbidities are the productivity ~urrently being reforecasted. At present, there is no reliable information to use to describe the probable influences of the with-project open water area on winter productivity. Another aspect of the open water reach lies in its potential to become overwintering habitat. Present juvenile salmon overwintering areas are char- acterized by the presence of ice cover and upwelling warmer than ambient water (Mike Stratton pers •omm). Little is known about resident species overwinter- ing habitats, uut generally it seems that upwelling is uot as critical a com- ponent for them . Resident species are thought to overwinter in deeper main- stem pools and at tributary mouths (ADF&G 1983c). The open water reach could conceivably provide some overwinter habitat for juvenile salmon, since released reservoir waters (0.5 to 5.6 C) would be 33RD1-007 within the normal range of upwelling temperatures (0. 8 to 4. 2 C) and cover could be afforded by the turbid conditions. Since it is presently impossible to accurately predict turbidities, it is prema ture to speculate on the effec- t:~veness of this type of cover. The open water area should provide more over- winter habitat for resident species than now exists, chiefly because of the combined effects of higher with-project flows (which could create favored deep pool environments) and the relat ively warmer temperatures. The open water area could also provide additional salmon spawning and in- cubation habitat. Chum s almon have been observed spawning in other mainstem areas influenced by upwelling groundwater (ADF&G 198Sb). Although undocument- ed, it is possible that upwelled mainstem water temperatures at these sites are similar to those seen in sloughs. Given that released water temperatures are predicted to be in the range of upwelled slough water temperatures, and given the proclivity of chum salmon for spawning in ma i nstem environments, it is conceivable that the middle river could function as reproductive habitat provided that suitable substrate exists there. A more detailed analysis of water temperature effects on incubation is found in AEIDC 1984. Another expressed ice-related concern in the middle river pertains to the natural flushing of fines from slough spawning habitats by breakup-induced flooding (table 14). Regardless of whether one or two dams are built, !CECAL simulations predict that breakup e v ents would no longer occur; the river ice mantle would gradually melt in place. The issue rests in some visual estimates made o f the appearance of slough spawning substrates following breakup. Fo r example, an ADF&G biologist (Drew Crawford pers. comm) reported that following a breakup flood, s lough SA substrates appeared cleaner . Unfortunately, no sediment samples have been taken before and after breakup floods, so the issue remah.s founded on subjective appraisal of 33RD1-007 2.33 environmental conditions . While it is conceivable that breakup flooding is important for the maintenance of slough spawning substrates (at least in some locations), it is equally possible that hydraulic upwelling pressure (coupled with the actions of redd building adults) is sufficient for this purpose . Given the lack of information on the amount and size of intragravel fines be- fore and after floods, no defensible conclusions can be drawn. The last question analyzed is that of the effect of with-project anchor ice on fish and their habitats (table 14). Mechanisms of anchor ice formation are poorly understood, but is is known to occur most often in supercooled reaches over gravel substrates (Michel 1971; Mason 1958). Anchor ice is relatively couunon in the middle river, but none has been found to date in either mainst~m or slough salmon incubation areas. Presumably, this is due to the influence of warm upwelling at these sites. Little is known of the influence of anchor ice on Susitna River fish hab- itats. Benson (1955) studied the effect of anchor ice on trout stream ecology in Michigan. There, anchor ice was not found to affect trout eggs buried in the gravel. However, trout fry were apparently vulnerable to mortality if they were emerging at the same time that anchor ice formed. In California, Needham and Jones (1959) noticed that when anchor ice was dispersing and breaking up, it dislodged substrates and considerable numbers of invertebrates were carried away. In the middle river, anchor ice can carry gravel sub- strates away in a similar manner (R&M Consultants Inc. 1984). However, no invertebrate sampling was done at these times, so its influence in this regard is unknown. !CECAL does not simulate anchor ice formation; therefore, no with-project predictions of its rates or timing of formation, distribution, or thicknesses have been made. Project team ice modelers believe that there would be less 33RD1-007 2JY anchor ice with-proj ect in the middle river. Upstream of the 0 C isotherm in the open water lead below Devil Canyon, no anchc r ice formati on is likely due to the influence of the warmer than natural released water. This could be seen as a potentially stabilizing effect on instream invertebrate habitats there. Anchor ice would form with-project between the upstream edge of the ice-front and the 0 C isotherm in a manner similar to that seen naturally . However, its annual aggregate areal extent could increase over natural conditions, given the predicted dynamic nature of the ice front. More anchor ice would form with the Watana Reservoir than with both dams on-line because of the greater variability of anticipated flow r~leases. It is probable that no anchor ice could form in areas influenced by relatively warm upwelled waL ~--Thus, with-project anchor ice should not influence salmon reproductive habita::s. Given the imprecise nature of current knowledge of anchor ice formation processes and its influence on other fish habitat co~ponents, it is impossible to ~peculate further on how changes in the present ice regime would affect other species. 33RD1-007 5. LOWER RIVER EFFECTS ANALYSIS a. Fish Resource Nineteen species of fish are known to inhabit lower river waters (figure 89). All are dependent to some extent on mainstem environments to fulfill as- pects of theic life histories. Seven of these species are anadromous; they include five species of Pacific salmon and eulachon and Bering cisco. Avail- able information on fish species presence, distribution, habitats, and behav - ior in the lower riv er is not as complete as in the middle river . Based on extant data, it seems that fish use of lower river environments largely paral- lels that described for the middle river and impoundment zone, with a few ma- jor exceptions . At least 17 tributary streams and six sloug ~s provide salmon reproductive habitats in this reach. To date, no chinook, sockeye or pink salmon have been observed spawning in project-affected lower river mainstem waters; all appar- ently use tributary streams exclusively for this purpose (Barrett, Thompson & Wick 1985). Small numbers of chum and coho salmon have been seen spawning in 13 separate mainstem sites and six side sloughs; most members of these two species also spawn in tributary environments. ADF&G estimates that, in aggre- gate, the number of chum salmon spawning within main ~tem environments there represents roughly 0.3% of 1984 escapement to the basin. The estimated number of spawning coho in the mainstem represents roughly 0.2% of the 1984 escapement (Barrett, Thompson & Wick 1985). Chum salmon were the principal users of side slough spawning environments, being present in fiv e of the six sloughs used. Their estimated numbers rep resent roughly 0.1% of the total 1984 escapement. Only six coho were seen spawning in sloughs in 1984; all were in one of the six sloughs ( Barrett, Thompson & Wick 1985). This 33RD1-007 indicates that, unlike the middle river, lower river sloughs are less important for spawning purposes. Less is known of salmon rearing and overwintering habitats in lower river mainstem environments than in the middl~ river. Given their respective life history requirements and the natural hydrologic conditions occurring in win- ter, it is possible that some chinook, coho, and sockeye salmon overwinter in the mainstem and that some chum, chinook, and coho rear there. A few coho and chinook have been captured during winter in mainstem environments (ADF&G 1983c). Several million eulachon spawn in late May to early June in the lower 50 miles of the mainstem Susitna River. Most of these fish spawn below RM 29 in main channel habitats near cut banks over loose sand and gravel substrates (Barrett, Thompson & Wick 1984). Bering cisco return to the Susitna River in late August and spawning takes place in September through October. In 1981 and 1982, spawning activity peaked the second week of October. Bering cisco are known to spawn only in main channel environments; the majority of spawning a~parently takes place between RM 75 and RM 85 (Barrett, Thompson & Wick). Little is known about resident fish life histories in the lower river. A forthcoming ADF&G report (due to be released in late April), reportedly will contain a synopsis of available information on resident species. The 12 resi- dent fish ·pedes found in the lower river, with the exception of lake trout and northern pike, are generally believed to be common (Sundet & Wenger 1984; ADF&G 1983c). Rainbow trout, Arctic grayling, and Dolly Varden probably spend most of the open-water season in tributaries, using the mainstem principally for mi- gration and overwintering (ADF&G 1983c). Burbot, whitefish, longnose sucker, sculpin, stickleback, and Arctic lamprey are found in both the mainstem and 33RD1-007 ..23/ tributaries during the open-water season . All of these species are believed to overwinter in the mainstem. but only rainbow trout. burbot • and slimy sculpin were captured there during 1982 winter sampling (ADF&G l983c). Based on ongoing radio-telemetry studies. it appears that favored main- stem overwinter habitats for adult rainbow trout and burbot differ principally by depth and location. 7agged rainbows are most frequently relocated in main- stem side channels near tributaries in waters generally less than five feet (Rich Sundet. pers. comm.). They are often found close to open leads. Tagged burbot are most frequently located in winter in pools greater than six feet deep along river bends (Rich Sundet. pers. comm.). Both species seem to favor low velocity environments . Only one Arctic grayling has been successfully radio-tagged; it was frequently relocated in close association with rainbow trout (Rich Sundet. pers. comm.), No other resident species have been radio tagged. It may be that other resident salmonids with habits like rainbow trout also frequent relatively shallow low velocity environments in winter; the same type of relationship may exist between burbot and other bottom feeders such as longnose sucker. THE WITH-PROJECT ENVIRONMENT !CECAL simulations have not been run for the lower river. and the follow- in~ discussion is based wholly on subjective input provided by the Project Team. Ice would probably begin forming with-project in early November about the same time it does naturally. Increased with-project flows could delay upstream movement of the ice front. This delay is thought likely to be simi- lar to that modeled for the middle river (i.e. 17 to 44 days with Watana Res- ervoir and 27 to 47 days with both dams on-line). Increased winter flows might produce somewhat more ice than now occurs; however. this is uncertain. 33RDl-007 23?' Lower river ice meltout could be adv anced over natural conditions due to the expected earlier tha n normal meltout of the middle river . Meltout timing (if the abr ·1e is true) would be c losely coincidental to that predicted for the middle river. Since ice is e x pected to melt gradually , there would be no breakup ev ent as such. b. Anticipated With-Project Effects Four o f the nine ice-related issues of concern identified in the middle river cou ld also perta ~_n to lower river aquatic resources (table l4). Three issues (Nos . 1, 3, and 6) r~late to staging and one (No. 4) to the amount of ice cov er . As indicated above, no ice modeling has been done for the lower riv er; thus, conclusions presented here are tentative. With regard to staging, freezeup is thought likely to occur later than norma l with either one o r two dams operating. The view held by project team ice modelers is that freezeup staging would not lead to overtopping of slough berms in the lower river. Consequently , there would be no with-project ice effects on slough incubating salmon embryos. It is important to reiterate, however, that no ice modeling has yet been done for this reach . Should the prediction of no overtopping of slough berms prove false, the consequence to the salmon resource as a whole would be minimal. Lower river slough reproductive habitats are severely limited areally and a re utilized by only a sma l l number of chum salmon. Consequently, their collective contribu- tion to maintenance of Susitna River salmon stocks is very low. As in the middle river, the question of ice-related effects on upwelling pertains to salmon reproductive habitat quality. In ess ence, the question rests with two points: the rate of upstream migration of the ice front and the assumption that mainstem upwelling has a controlling influence on embryo 33RDI-007 2.31 survival. Salmon spawning naturally occurs in the mainstem at a time when river flow is decreasing . Successful mainstem reproduction is partly depen- dent on freezeup staging, which raises the water level and assures that up- welling is not diminished. This c oncern is more acute near the Chulitna River confluence than further downstream for two reasons; it would take longer for the ice front to arrive and more fish spawn in this area. With the project, ice front advance would be slower than natural, but flows would be greater than those now occurring. These two factors seem to offset each other. If so, effects to incubating embryos would be minimal , be- cause flows should be sufficient to maintain upwelling. However, it is i.mpor- tant to point out that to date there is no direct evidence that mainstem upwelling in the lover river exerts a controlling influence on incubation en- vironments there. The last lower river ice-related issue raised pertains to the question of how the with-project ice cover would affect primary productivity and the amount of overwinter fish habitat (table 14). Project ice modelers believe that regardless of wheth~r one or two dams is built, there would be more ice in the lower river with-project than naturally. However, the e.xact morphology of the ice cover is unknown. Provided that extensive lead systems did not develop, instream primary production with-project should be reduced in rough proportion to the increase in ice cover seen. If an extensive system of open water leads does develop, then the· converse would be true. It is possible that winter habitat availability could increase with-project, given the combined effects of ice-induced staging and greater flows. However, overwinter habitat is comprised of more components than just water volume. Numerous other variables, such as bed morphology, water depth, water velocity, temperature, and cover are at play. So, the belief that overwinter habitat 33RD1-007 might increase with-project is provisional and pending on acquisition of information describing how all habitat variables would be affected. Regardless, there would be no decrease in the amount of current overwinter habitat available for fish. 33RD1-007 2 'II 6. sm~Y In conclusion, winter drawdown of the Watana Reservoir would be a desta- bilizing influence on its littoral zone, makin~ it unproductive for salmonids. Some species would be more affected than others. In all likelihood, winter drawdown would preclude successful fall reproduction by lake trout; if spawn- ing took place at all, eggs would desiccate or freeze. Ice draping, gouging, and associated erosion would probably limit invertebrate productivity and cover availability, which in turn would diminish rearing habitat quality for Arctic grayling and whitefish. Burbot and longnose sucker should not be nega- tively influenced by ice-related processes. Both are bottom dwellers which do not depend on stable littoral zones for any of their life requirements. In s ome extremely cold years, ice blockage of tributary stream mouths could delay Arctic grayling and longnose sucker natal migrations. At such times, it is likely that reproductive failure could occur. This is not considered a major problem, since loss of a single year class is not overly threatening to rela- tively long-lived and fecund organisms like f ·~h. and given that the necessary cold climatic conditions seldom happen consecutively . The environment of the Devil Canyon impoundment would be much more sta- ble, given its winter drawdown schedule. However, the canyon's geomorphology and substrate geology limit establishment of a productive littoral zone. Fish reproductiv e habitats near the mouths of Fog and Tsusena creeks may not be in- fluenced by with-project icing events. Both are located in the upper end of the reservoir where little ice accumulation is expected. The chief with-project middle river ice concern lies in potential alter- ing of slough incubation habitat qua lity. Ice staging downstream of the ice front would cause overtopping of slough berms by colder than ambient mainstem water. This would have conseque nce to natal habitats. 33RD1-007 !CECAL simulations predict that all with-project ice-induc ed overto ppin g event s would occur after emb ryonic blastopore closure. Thus, there is little likelihood that direct mortality of embry os would ensue . However, indirect mortality would be significant given the predicted duration of most over- t ~~p i ng events (~one month). This would delay embryonic development to such a degree that it is unlikely that any could complete their life cycles. Over- topping waters could also affect overwintering juv~~~:? fish. Effects would be more severe the longer cold exposure lasted . Overtopping events would be more frequent and severe with the Watana Reservoir alo ne than with both dams on-line. Concern h a s been raised that the absence of with-pro ject ice staging in the area upstream of the ice front would alter slough upwelling rates. This does not seem likely as with-project winter flows are forecast to be between 8,000 and 12,000 cfs. This is similar to flows occurring naturally in Septem- ber . Since September upwelling rates are apparently sufficient to maintain salmon natal habitat qua lity, it seems likely that with-project winter flows should also be adequate. The with-project 10 to 29 mile long open water zone in winter below De v il Canyon could enhance primary productivity in the main- stem. The o retica lly, more light would be able to penetrate the open water column thereby stimulating photo synthesis. Howe v er, winter flows would be somewhat turbid c o nfounding the issue. A more likely effect of this open water zone could be the creation of ad- ditional overwinter habitat due to the combined influence of higher flows and warmer than natural water temperatures. Higher flow volumes could create deep pool overwinter habitats for res ident species. Since released reservoir wa- ters are predicted to be abo ut the same temperature as that of upwelled sloug h groundwater , this area might also prov ide some salmon overwinter and spawning 33ID1 -007 habitat. The with-project flow regime would eliminate breakup-induced flood- ing of slough habitats. Although unsubstantiated, this process may be neces- sary for maintenance of slough natal habitats (through flushing of fines from interstitial gravel spaces). Given present knowledge, it is impossible to predict the long term consequences of elimination of breakup-induced flooding on these habitats. Anchor ice has been shown to be a destabilizing influence on invertebrate and fish embryo habitats by dislodging substrates during melt- ing or breakup . No anchor ice is expected to form with-project in the open water lead upstream of the 0 C isotherm; however, it would form between the ice front and the 0 C isotherm in a manner analogous to that seen naturally. Cessation of anchor ice formation in the open water zone could stabilize incu- bation habitats. Less physical and biological information exists on the lower river than on the other two zones. No temperature or ice modeling has been attempted for this reach, making evaluation of with-project effects completely subjective. The general belief held by project ice modelers is that ice-induced staging would not lead to overtopping of lower river sloughs. With-project winter icing probably would not negatively influence upwelling rates, given that the effects nf the predicted slower than normal ice front advance and the higher than n atur al flows would likely offset each other. Higher with-project winter flows coupled with ice-induced staging could increase the amount of overwinter fish habitat (since wetted area would b 2 increased); however, this is uncer- tain given the present level of knowledge. Overwinter habitat is comprised of more than just water volume. Regardless, it seems likely that no existing overwinter habitats would be lost with-project. 33RD1-007 E. EFFECTS OF ICE ON RIPARIAN VEGETATION 1. GENERAL a. Impoundment Zone The early successional area in the potential impoundment zone is usually a band approximately 15 feet or less wide under lain by a cobbly or sandy substrate. It is dominated by a variety of forbs, graminoids (grass and grass-like species), and shrubs, including low willows and tall alders. Vegetation advancement is inhibited by ice and, to a lesser extent, s · ... mmer floods. Existing effects of ice on vegetation in the proposed impoundment zone include ice scars on trees, bending and scraping of low and tall shrubs, local sediment deposition from melting ice blocks, and silt deposition in the waters backed up behind ice jams. These events occur primarily during spring breakup. No ice effects on vegetation during freezeup have been reported along any reach of the river. As ice jams form, water levels may rise rapidly upstream from the jam and inundate vegetation for a short period of tine, probably less than 5 days. This inundation probably has little effect on plant species or vegetation succession because of its relatively short duration. However, sediment may be deposited at that time or as the ice melts in place . Ice blocks in a jam may be pushed laterally against and over the bank where they may scar, spp.) and several break, species and scrape woody species such as alder (Alnus of willow (nost commonly Salix alaxensis). Herbaceous species such as sedges (Carex spp .). fireweed latifolium), h e dysarum (Hedysarum alpinum), and horsetail (Epilobium (Equisetum 33RD2-007o variegatum) may also be scoured, but undamaged graminoids (grass or grass-like plants) have been observed under ice blocks piled several feet high near Clarence Creek (Helm and May er 1985). Ice has the greatest effect when the jam breaks, releasing huge quantities of ice and water moving very rapidly downstream. The ice blocks may scrape against trees, remov ing b~rk and living cambium, and possibly some outer layers of wood. Cambium is a thin layer of living tissue between the bark and wood, and is essential for wood production. When the cambium is removed by scraping, wood cannot form in that area in following yea rs, re sulting in a effects on an scarred white sca r. Most scarring does not have any important individual plant's growth or vegetation succession . lasting Freshly spruce trees (Picea glauca) were observed in spring 1982 downstream from Goose Creek. Some shrubs, usually willows and some alders, have been bent at about 45 degrees by bloc~s of ice pushing laterally along the shore. b . midd le river Vegetation succession seque~ces along the Middle and Lower reaches are similar to each other except for the more important r o le o f ice in the Middle River . Since the channel is still incised in the middle river, early successional sites occur in a relat ively narrow band compared with similar areas on consists the of lower river. Early horsetail (Equisetum successional vegetation variegatum), balsam usually poplar (Populus balsamifera), and feltleaf willow \.:lalix alaxensis), alone or in various combinations. Willow and poplar between 0.4 and 2.0 meter tall and larger stems that can be bent over are important as moose browse. Dryas (Dryas dru~mondii) occurs on more cobbly areas, while occasional 33RD2-007o forbs and graminoids are found in some areas. Approximately 10 to 20 years after stabilization, alder (Alnus tenuifolia) becomes dominant (McKendrick et al. 1982). Balsam poplar then overtops it after another 25 years. These alder and immature balsam poplar stages are considered intermediate and have few shrubs of either the species or size classes needed for browsing. When poplars are 70 to 100 years old, the overstory becomes patchy and rose (Rosa acicularis) and highbush cranberry (Viburnum edule) dominate the shrub understory. Both of these shrub species are browsed. Mature paper birch (Betula papyrifera) white spruce (Picea glauca) forests occur approximately 200 years after stabilization. These sites also have patchy canopies and rose and highbush cranberry in the understory. More details have been reported in McKendrick et al. (1982) and will be reported in Helm (1985). (Ages and time spans may be changed as data are analyzed for Helm (1985).) Natural effects of ice on vegetation in the middle river also include ice scars on trees, bending and scraping of low and tall shrubs, local sediment deposition from melting ice blocks, anc! silt deposition during staging behind ice jams. Old scars have b ~en found on balsam poplar (Populus balsamifera) approximately 7 feet above the ground near the mouth of Whiskers Creek (R&M Consultants, Inc. 1981, Helm 1985). These individuals were probably 30 feet or more from the present water line under normal summer flows. Freshly scarred trees have also been found along a cutbank near the confluence of the Susitna and Chulitna Rivers (R&M Consultants, Inc. 1982). Woody species, especially low willows and tall alder shrubs may also be bent or even have stems partly broken, but are capable of growing new shoots from the existing stem or rootstock (McKendrick et al. 1982). The 33RD2-007o degree of bending varies from slight (less than 10 degrees) to major (approximately 90 degrees). Most bent alder are flattened against the ground while most bent willow are only bent at approximately a 45 degree angle. Stems are usually partly broken only in intermediate stages of vegetation succession, such as 15 to 20-year ~ld alder, where the stems have become more rigid , but may still bend. Some cracked stems can still grow new shoots if enough cambium u:mains. Younger plants in earlier successional sites are usually only bent, and stems are rarely broken. Older, more rigid trunks, especially of balsam poplar, may be broken near the base and carried away by floods. This appears to happen to individuals near the edge of the vegetation type which get in the way, but not to the whole stand, large flood . Noberly and Cameron although this would be possible in a 1929, in Gerard, (no date) report e d an ice jam flood that knocked down trees 3 feet in diameter near Fort McMurra y on the Athabasca River in 1875. This bending and resprouting proce ss does not normally change the successional stage of vegetation in terms of type of vegetation, but it does change the age structure. Where 20-year old alder stems once occurred, one-year old sprouts would occur the next year. This scraping may increase the shoot production of willows to create more browse. If a particular portion of an island is subjected to ice scour repeatedly (say, every 20 years for alder) then vegetation succession would not a dvance beyond this stage, and individual stems would not grow more than 20 years above ground on the average. Rootstocks, however, could be 50 or more years old. A longer period of time without a jam becaus e of weather may allow vegetation succession to advance beyond that stage. Large ice blocks containing unsort e d sediment ~ay become grounded and stay in one place until they melt. The sediment load is then deposited 33RD2-007o where the ice melts (R&M Consultants, Inc. 1984), possibly forming a hummocky microtopography which could affect moisture availability for plants. Individuals growing between mound s would have more moisture and more litter deposited in those depressions. This could slow evaporation in the summer time and increase soil nutrient status if the material decomposes sufficiently. However, decomposition in northern latitudes is fre~uently slow. drier Individuals growing on top the m.:>unds would be subjected to a moisture regime, but would probably experience wanner temperatures. The unsorted nature of the sediment deposited by ice creates a diverse substrate for plant growth. Silt ma y also be deposited when water backs up b ehind ice jams and the water velocities slow (R&M Consultants, Inc. 1984). On some rive rs, such as the Yukon River with a broad floodplain, layers of s ilt 1/2 t o 2 inches thick may be deposited over large areas (Eardley 1938). Local deposition could not occur this high any other way on some bars and in side meander bends because summer floods did not reach thi s high (Eardley 19 38). Because of the rela tively incised channel on the middle Susitna River, deposition would not cover as large an area. Studies of aerial photographs from 1949 to 19 80 from Talkeetna to Devil Canyon have indicated that the size and shape of islands have changed little over this time period although they have become more vege tated (LaBelle 1984 ). Hence, the borders of young vegetation (less than 15 years) around most of these islands must be regrow th after sone riparian event either prevented suc cession from advancing or denuded a site to set vegetation succession backward to an earlier stage , and do not repre sent colonization of newly depos ited substrate (R&M Consultants, I nc. 1982, Helm 1985). Ice is believed to keep the younger vegetation (5 to 25 years old) 33RD2-007o from advancing while sum~er flows a r ~ believed v egetation becomes established on bare s urfaces. to control where new Bare surfaces probably result from ice scour where soil has been eroded or shoved by ice blocks (R&M Consultants, Inc. 1982 ). However, these surfaces may be colonized between ice events which would not occur every year . These scouring events are relatively localized, but occur repeatedly in the same places although not every year (R&M Consultants, Inc. 1984). The line between young vegetation, dominated by graminoids, that has not been allowed t o advance, and the riparian woodland is sometimes considered the trimline (Gill 1973). Mackenzie (Gill 1973) and This is probably controlled by ice on the Susitna Rivers (Shoch, pers. comm.). Rapidly developing woody species may be able to colonize the are.a between years when ice jams occur. A classic example of ice effects on vegetation occurs near the mouth of Whiskers Creek near LRX-7. The downstream end of the island contains an immature (approximately 50 to 70 years old) balsam poplar stand with some alder in the understory (Helm 1985, subject to revisi on pending further ana l ysis). Old ice scars occur approxinately 7 feet high on balsam poplar. Just upstream is a younger site. Although we do not have information on those ages, stems appear to be 20 30 year old alder. These do not appear to have been affected by ice recently (last 10 or 15 year s) although many alder stems have been bent and then r esp routed . On one side of the island, willows have been beaten down (approximately 45 degree angle), but do not appear to have had any danage o ther than bending and minor ~craping . At the upstream end of the i sland, the s urface was almost bare except for a few young willow stems that had survived the scraping (Helm 33RD2-007o 1985). Visual evidence of ice shoving soil and of silt being deposited from melting ice blocks was present. Many of these islands have gradually sloping banks above water during normal summer flo~s. In some areas, however, particularly along the main bank, cutbanks are lined with mature vegetation. These areas may be eroded by ice shoving against them (R&M Cor.sultants 1982). Mature balsam poplars were observed overhanging the left bank near LRX-3 after a large ice jam flood (R&M Consultants, Inc. 1982). Tree roots have been observed sliced by ice along a cutbank (Shoch, pers. comm .). Summer floods may also contribute to this erosion started by ice. soil, as well as the vegetation, is removed . In this case, howe v er, the c. lower river Ice effects on the lower river are less dramatic because of the width of the river and its braided morphology; however, ice jams have been reported near Montana Creek in 1983 (R&M Consultants, Inc. 1984). Occasionally minor jams may occur, but because there are so many other places for the water to flow, no major effects have been observed. Vegetation succession is similar to the middle riv er extensive areas are usually available for colonization. 2. EEFFECTS OF ALTERED RIVER ICE a. Impoundment Zone except that more The majority of areas that are now directly affected by the river, as well as many mature forests, in th e propos ed icpoundment zone would be inundated by the proje c t. Relatively mature forests and shrub lands 33RD2-007o would border a band of relatively unvegetated area around the impoundment. Water levels would increase through the summer, but would be drawn down for power pro duccion during the winter. Hence the summer wat ~.:r levels would probably determine where vegetation grows. species that can vithstand prolonged flooding, Except the area for annuals and between the winter ice and summer high water levels would probably be unvegetated . Some of the area near the highest fill level could be vegetated by species tha~ tolerate wet rooting zones. Some esLablished species could be affected by higher water tables, but this area is expected to be narrow. Species that prefer well-drained soils generally occur on steeper slopes; hence, an extensive area of existing vegetation would not be affected. Nilsson (1981) reports that since ice along lake reservoirs in Sweden melts in place along the shore, substrate is exposed for colonization for a few weeks before the reservoir fill s again. If the reservoir does not fill up each year, then more time is available for coloni z ation. The Susitna reservoirs are expected to fil! to maximum level each year so this should not occur. Ice effects should be mi nor, at most, with the project . Bears use the potential impoundment zone in the s pring (Miller and HcAllister 1982), when they are believed to be fora ging for hedysarum roots and possibly horsetail. Hedysarum generally grows in sandy or gravelly areas and could be found in forests or tret:less sites above the impoundment . It has not been established what horsetail species bears eat. One spec i es, Equisetum variegatum, is usually found in floodplain sites, while E . silvatic um and E. arven se are found in wooded areas. Hen c e, it is not clear if there would be a n impact on bears. Similarly , many plant species eaten by moo s e alo ng the floodplain are found at higher elevations. 33RD2-007o b. Middle River Areas of the middle river will not have continuous ice cover above RH 124 to 14 2 (Harza-Ebasco Susitna Joint Venture 1984). Existing vegetation in these sites is expected to advance successionally without setbacks attribut.lble to the river except where areas are being eroded away by cha :mel movement, which is relatively minor because of the incised channel. Areas that are currently unvegetated may become vegetated because of the lower summer flows if the substrate has sufficient fines (sand-sized particles and finer) to support plant growth and if sufficient moisture is available. With enough time (possibly 5 years or decades, depending on the environment), enough fines may be deposited to support plant grcwth. Much of the newly exposed substrate probably would consist of cobbles . Areas of the midd le river that would have ice cover would not be subject to the destructive breaku p floods that currently occur because the ice W)uld melt in place (LaBelle 19 8 4a). However, because of the winter fl oo din~ of areas that could be colonized in the summer, winter inundation and forma ;:ion of ice may retard vegetation development . Species that currently l .row along the river are adapted to summer flooding, sometimes for long periods of time, as well as overflow flooding and refreezing in the winter. Host of the natural winter flooding is short-lived and probably does not seal :he ground. Gas exchange may still occur among the roots, soil air, and the above-ground atmosphere. Winter flooding as a result of the project would remain all winter and would freeze in place (downstream from the icefront l, possil-ly inhibit;'_ng gas exchange between soil and atmosphere . Plant roots respire in winter although at a slower rate than in waroe r temperatures of sumoer. Winter ice-sealing is harsher than summer flooding becaust! 33RD2-007o suru::~er flooding is usuall y of sh o rter dur a tion and i nvolves moving wa ter, wh ich allows better gas e xchange (Kozlowski 1984). This seali ng of the ground by ice i s deva s tating t o some forage crops , such as alfalfa and grasses grown for hay (Smith 1981). This is a more severe treatment than the riparian plants currently receive , but they would proba bly t olerate it, although growth ma y be hampered. Some ar~a s with establi s he d vegetation (LRX 3, 7 , and 17) ma y become cov ered with winter ice under certain cond itions (Harza-Ebasco Susitna Joint Ventu re 1984), but presumably th ese plants would be able to survive . Established vege t a tion would be expe c ted i ndividuals. t o t o lera t e this treatme nt better than new The fa c tors wh ich influence vegetatio n establishme n t mo s t are probably prec ipita tion and water tab les during the s hort per iod in May and June wh e n viable wil low and bal sam poplar seeds are di spe rs e d. Seeds of these early successio nal s pe cies are s h or t-lived (2-6 we e k s), are dispersed in late May and J une, and r equ ire moi s ture for ear ly growth (Sch reiner 1974, Zasada et al. 1 98~,). Even under natural condi tions , new areas may not be immedia t ely colonized depend ing on moisture (Helm 1985). Colonizati o n of new a re as with the projec t in place migh t be slowed from normal rate s , but as a r~s ult of flows in ~~y a nd J une rat he r than the i c e eff ec t s (Helm 1985). Th e e a rl ier meltout (4 to 6 week s with Watana on l y and 7 to 8 wee ks with both Wat a n a and Devil Canyon) a nd reduc ed water l eve l s b y Ma y (Harza-Ebasco Su s itna Jo int Venture 1984) c o uld make it difficult for vege t a tion t o become est ablishe d from seed unle ss s ufficient precipitation occur s at th is time . Mosa i cs of vegeta t ion types and their resultant edge effec t which a r c con side red imp o rt a nt for wildlif e would be l os t witho ut ice effects, which 33RD2-007o control much of the vegetation dynamics in the middle river. However, some of the mature paper birch -white spruce sites may have patches of trees alternating with shrubby areas where old trees have fallen and left a gap in the canopy. Shrubs become more abundant and larger in these openings. Once vegetation becomes established, it is expected to advance unhindered by river effects. Artificial means of mitigation, such as logging, could be used to setback vegetation successio11 if that is c0nsidered desirable (Helm 1985). Early sites would probably advance to alder sites in 10 to 20 yeilrs, while alder stands would have advanced to immature balsam poplar stands in 20 years. Another 30 years would be required for maturation of the stand. It may take 5 years for a new site to colonize under the new flow regimes unless favorable rainfall occ rs during Hay and June. t-lore details will be found in Helm (1985) when it is completed. c. Lower River Ice effects on the lower river would probably be similar in nature to those of the middle river, but would be less because the projec t would have less effect on the lower reaches of the river. Other rivers, such as the Chulitna, Talkeetna, and Yentna, reduc e the effects of the middle Susitna below Talkeetna. Areas that currently have ice jams, such as near Montana Creek, would be less affected by ice as a result of the project . The lower summer water levels would control vegetation establishment. Colonization and vegetation succession would be similar to the middle riv e r. 33RD2-007c E. Effects on Wildlife 1. GENERAL A large and diverse group of wildlife species uses the Susitna project area year-round or seasonally (Alaska Power Authority 1983, LGL 1985). Moose, caribou, Dall sheep, black bear, brown bear, wolf, red fox, beaver, muskrat, and some members of the weasel family river otter, marten, mink, short-tailed and least weasel --are all locally abundant there. Lynx and wolverine are present in low densities, ranging widely through large territories, and evidence of coyotes is limited (Gardner and Ballard 1982, Whitman and Ballard 1984, Gipson et al. 1982, 19134). Many hird species have been do~umented in the Susitna Basin, including 15 species of raptors (eagles, hawks, falcons, harriers, and owls), up to 60 species of waterbirds (swans, geese, ducks, gulls, and shorebirds), and many more species of terrestrial birds (songbirds, ravens, woodpe ckers , gro use, and ptarmigan) (Kessel et al. 1982a,b). However, only a few of these bird species are permanent residents in the project area. Most use the region primarily as summer breeding habitat, migra ting in thP fall to temperat e or trop i ca l latitudes up to thousands of miles distant from the Susitna Bas in, and returning to northern latitudes in the spring. The comparatively few bird species that remain as ' winter residents include sp ruce grouse, ptarmigan, ravens, magpies, gray jays, chickadees, and several other species (Herter 1985). The following dis cu ss ions, based o n the re s ults of studies sponsored since 1980 by the Alaska Power Authority, emphasize those mammal and bird species most likely to be affec t ed , either ben eficially or adversely, by altered ice conditions assoc iate d with the Susi tna Project. 33RD2-007u a. Impoundmen t Zo ne Moose (Alces alces) occur throughout the impoundment z one and surrounding region at all times of the year, but are present in the Watana impoundment area in grea ter numbers than in th.:o Devil Canyor. impoundment area. For example, an estimate of 19 3 to 278 moose was made from an aerial census of the Watana impoundment area on March 28 , 1983; a similar census o n March 3 1, 1983 of the Devil Canyon impoundment area out to ~ mile beyond its maximum fill level estimated only 14 moo se (Ballard et al. 1984a). Studies of radio-c o llared mo ose have shown tha t some a nimals consistently confine their mo v ements to relative ly sma ll home range s in and n ea r the imp o undment zone, with considerable overlap between summer and winter range areas , whereas others tr~vel far from the impoundment zone, s pending on ly a portion of the year there (Ballard et al. 1984a). The impoundme nt zone ma y be esp ec ia 1ly i mpo rtan t to moose during the winter (November through April), because th e c anyo n of the Susitna River off e r s lower elev a tions wher e f o ra gt ma y be more ac c essible due t o shallower snow than on surruunding higher lands. Availabili t y of winter fo r age i s an important limiting fac t o r to moose . Accum~ca ting snow covers browse vegetation and, if deep enough, r es trict s th e ability of moose to rea ch their f ood either by digging or by moving t o areas of shallowe r snow . Studies of r adio-collared moose have s hown that use of lower eleva tior,s (1 ,800-3,000 feet) within th e impounciment zone increases d uring l a t e winter and early spring, when s now i s d eepes t . Many moose r em<'i_n at these low e r e l eva tio n s during the s u mme r, but move to higher elevat i on s in Oc t obe r and r emain there until accumula t ing snow ma y again influence their return tc lower eleva tions (Ballard e t al. 1984a). 33RD2-007u Studies of radi o -collared moose have a l so revealed that mo ose cross t he Susitna River at all times of the year ; 79 c rossings were documented from 1980 thrnugh 1982 (figure 90), and a March 1981 survey counted 14 set s of noose tracks crossing th e rive r between Watana and Kosina creeks (Ballard et a l. 198 2 , 1983a). Ca ribou (Rang ifer ta~a ndus) in the vicinity of the impoundment zone belong to the Nelchina herd, whi ch cu rrentl y co ntains about 24,0 00 anima ls (Pitcher 19 85) (figure 91). This is the herd most accessible t o the ma jorit y of sport hunt ers in Alaska, because of it s proximity to roads and to human population centers. Size, ca l ving areas, migratory movements, and o ther herd characteristi cs have been documented annually since 19 48 (see Skoog 1968, Hemming 1971, and Pitcher 1983, 1985). The known range of the Nelchina he rd i s generally bounded as follows : to the we s t by the Chulitna River and Parks Highway , t o the north by th e Ala ska Range, t o the east and southeast by the Mentasta and Wrangell mo untain s, and to t he s ou th by the Gl enn Highway (figure 92) (LGL 1985). The r ange of th e Nelchina herd thu s includes the en tj re imp oundment z one a nd its sur;ounding a r ea . Caribou of the Nelchina herd migrat:.! across the Susitna Ri ver several times each year . Th e Wa t ana im poundment area includes th e reach of the river wh e r e most crossings occur, be tween Deadman a nd Jay creeks, and i t is likely that members of the Nelc hina herd wo uld continu e t o cross the Watana reservoir annually in the futu re. As re cently as 19 82 , approximate ly 50% of t he female segment of t he Nelchina herd mig rated t hrough the upper reaches of the Watana im po un dmen t area en r ou t e to s pring calving g r ounrls (Pi t cher 198 3). Th e Wa t ana imp oun dme nt a~ea wo u ld probabl y se rv e as a crossing r ou te in fu ture yea rs fo r l arge numb e r s of migra t ing ca r :bou (Pitcher 1984). 33RD2-007u Dall sheep (Ovis dall~) occur at high elevations (above about 3,000 feet) in the Watana Creek Hills, Mount 1-Jatana-Grebe Mountain, and Portage-Tsusena creek areas, each of which supports an identifiable sheep population (figure 93) (Ballard et al. 1982, Tankersley 1984). However, in early summer (mid-May through mid-~uly) sheep of the Watana Creek Hills population descend overland 5 miles or more to use a mineral lick complex along Jay Creek, just inside the Watana impoundment high-pool margin (2 ,185 feet) (figure 94) (Tobey 1981, Ballard et al. 1982, Tankersley 1983, 1984). A minimum of 46 sheep used the lick area in 1983, about 31 percent of the observed Watana Creek Hills population in 1983 (Tankersley 1984). Because all of the major lick sites (including the intensively used Bluff and East Ridge; see figure 94) are on the banks of Jay Creek canyon above 2,185 feet, they would not be flooded by the Watana reservoir. However, most sheep arrive at the lick complex fr otr. the northwest, and sheep have been observed crossing the creek to reach lick sites on the southeast side (Tankersley 1984). The o bserved cro~sing point (just downstream from the Bluff and East Ridge) would bt inundated at the maximum ·reservoir fill level in late sumrner, but would be exposed when sheep cross during May and June (figure 94). Brown bear (Ursus arctos) and black b ear (Ursus americanus) are both abundant within the impoundment zone an~ would ~e affecte d in important ways by the Watana and Devil Canyo~ reservoirs (LGL 1985). However, bLcause these species hibernate , effect s related direc tly to ice would be limited to the early spring, when b ears emerge from dens while the Sus itna River is stiE f r o zen. Because bears are powerful s wimmer s and :limber s , th e y would not be affected greatly by re s ervoir conditions. 33RD2-007u Wolv e s (Cani s lupus) range widely throughout the impoundment zone, with heavies t concentrntions usually in areas that support their major prey, moose and caribou (Ballard et al. 1984b). Wolves commonly o ccur in social units called "packs'' whi c h maintain exclusive territories. Ballard et al. (l983b) found pack territories occupying all available habitat around the impoundment zone and along the upper river. Areas inside the impoundment zone were occupied b y at least six packs. For nine intensively studied packs, territory size averaged 452 squ a re miles (range 124-803 square miles). Although observations concerning wolf numbers in the Susitna Basin have been rec orded since the 1950's, compariso n of those estimates is difficult because of different methods used and different are as included in the estimates (Van Ballenbe rghe 1975 ). Pack sizes of the wolves studied by Ballard et al. (l983b, l984b) ranged from abou t 2 t o 6 individuals per pack and fluctuated seasonally. In contrast to wol ves , evidenc e of coyotes (Canis latrans) in the impoundment zone is limited, a nd trappers jn t he area report ca tching f ew (Gipson et al. 1984 ). Sma ller forbearers are al so abundant in and a r ound the impoundment zone. Red fo x (Vulpes fulva) frequent the rolling upl ands and foothills (between ab ou t 825 and 1,90 0 feet) of the middle Susitna Basin (Gipson et al. 1982, Hobgood 1984). Fox abund ance ge ~e rall y increases with distance upstrea m from Devil Canyon t o the mou th o f the Tyone River. Hobgood (1984) found s um:ner home ranges of adult red fo x to ave rage about 14 square miles, with o ne fox famil y per 3 2 t o 48 squ a re mil es . Tra ppers commonly cat c h foxes alor.g Susitna Ri ve r tributaries throughou t t he impoundmen t zone (Gipson e t al. 19 84). River otter (Lutra canadens is ) and mink (l-1ustela vison) are abundant along the Susitna mainstem throughou t the imp o undment z one and along most tributa ries vr to 2,000 feet in e le·1ation (Gipson et al. 1982). Mart e n 33RD2-007u (Martes ameri cana) are locally abundant but res '.ricted to mature spruce and spruce-birch forest ~elow 1,700 feet . Gipson ct al. (1982) f o und tha t marten occur along the Susitna mainstem in highest d~nsities between Devil Creek and Vee Canyon; Buskirk (1983) noted that indiv·.dual marten in the area tend to use the highe r elevations of their home ranges in the spring and lower elevations in the autumn. Short-tailed and least weasels (Mustela erminea and ~ rixosa , respectively) are locally abundant th roughout the impoundment zone (Gipson et al. 1982, 1984). Beaver (Cas t or canadensis) and muskrat (Ondatra zibethica) are present in the slow-flowing sec tions of mos t of the larger tributaries of the Susitna River and in lakes and porJs associated '-'ith those tributaries. However, these two species appear t o be absent from the Su s itna mainstem in the impoundment zone; the current is pr o bably t oo swift to suppo rt them (Gipson et al. 1982, 1984). Host birds, because of their great mobility and seasonal •·se of the impoundment zone, would not be aff ected appreciably by changes in i c e c onditions resulting from the Susit ~a Project . However, the bald eagle (Haliaeetus leucocephalus) is generally restricted in central Alaska t o rive r valleys , including the Susitn a , where large, mature white spruce and balsam poplar provide suitable nest trees along banks and on isla nd s (LGL 1985). Ice c onditions affecting active or potential nest trees would influence the d ·_!::L ribut ion of this species in and around the impoundment zone. In June 1984, three bald ec.gle nests. two currently active and one inactive, were present in the imp o undment zone (Rose11eau 1 98 4). Two waterbird species, the American dipper (Cinclus mexi canus) ar:d occasionally the common merganser (Mergus merganser), r emain as far north as the impou nd~ent zone during the winter if open water is present (Kessel 1982a,b). The extent and l oc ations o f 33RD2-007u open water and available food resources would determine with-project winter distributions of these species. b. Middle River Zone Moose use the middle river throughout the year but are present in greatest numbers during the winter (late October through late April) (Modafferi 1982). Census data from the four winters of 1981-82 through 1984-85 indicate that movements of moose onto the floodplain tend to correspond with the timing of snowfall, and that numbers of moose on the floodplain relate closely to snow accumulation. During the moderate winters of early 1982, 1983, and 1984, census results showed highs of 36, 84, and 88 moose present in the middle river (Modafferi 1984). In contrast, a winter census high of 132 moose --50 percent greater than any previous census of the middle river --was recorded in January 1985 (r·lodafferi 1985 pers. comm.) during the hei:iviest snow accumulation in ten years (SCS 1985). Hoose cross the river at all times of the year and freely walk on stable river ice. In May and June, some females calve in riparian habitats along banks and on islands, then move with their calves to south-and southeast-facing slopes away from the floodplain during the rest of the summer and fall (Hodafferi 1982). Males also move away from the river to higher elevations during spring and summer, returning with winter to the floodplain. Black bear popula tion s are substantial in the middle river zone and lower river and this species relies heavily on riparian habitats throughout its active period from approxima tely early Ha y through early October (Hiller and McAlli s ter 1982; Hiller 1983 , 1984). Wolves are probably not abundant throughout the middle river. Although Ballard e t al. ( 1983b) found most available habitat in the upstream portion 33RD2-007u (near Devil Canyon) to be occupied by wolves, Modafferi (1984 pers. comm.) has observed little evidence of wolves in the majority of the middle river, despite high numbers of wintering moose. Coyotes are apparently common between Portage Creek and Talkeetna (Gipson et al. 1984), but population size, distribution, and habitat use have not been documented. Red fox occur along the :loodplain but are more abundant at higher elevations away from the river. Beaver and muskrat occur throughout the midd)e river. October 1984 surveys conducted by Woolington et al. (1984) documented at least 45 active beaver colonies preparing to overwinter. Fourteen colonies were found in sloughs, 14 in upland sloughs, 13 in the mainstem, and four in side channels. Evidence of muskrats was infrequently observed, but they probably occur in side channels, sloughs, and upland sloughs. River otter, mink, marten, short- tailed weasel, and least weasel are known to be present throughout the middle river, but distribution and abundance have not been documented for these species. Only a few bird species are directly dependent on riparian habitat in the middle river for nesting and feeding. These include the semipalmated plover (Charadrius semipalmatus), spotted sandpiper (Actitus macularia), harlequin duck (Histrionicus histrionicus), and common merganser. With the potential exception of bald eagle nest trees, terrestrial birds and their habitat would not be appreciably affe~ted by altered ice conditions within the floodplain. 2. EFFE CTS OF /,LTERATION OF RIVER ICE a. Impoundment Zone (1) Watana Dam vn Line Alone. ~3RD2-J07u (a) Non-uniform ice formation during freeze-up. form along shorelines while the reservoir center In November, ice would remains open. Moose or caribou venturing tc the outer edge of the border ice may fall through and be unable to regain a solid footing. Because only a few animals are likel~ to only occasionally die in this way each year, the effect would not be important. (b) Ice deposition along reservoir margin during winter drawdown. As winter reservoir drawdown proceeds, shorefast ice would fracture and become draped along the banks; on steeper slopes, ice-shelving may occur (Hanscom and Osterkamp 1980, Alaska Power Authority 1983). Cracks would form as the ice d~apes and settles over irregular shoreline topography. Cracking would also occur between shorefast ice a nd the receding reservoir surface. These effects may impede or injure moose and caribou. The potential for ice-related hazards would be greater with Watana dam alone, as compared to the Watana and Devil Canyon facilities together. The deeper drawdown (about 90-foot) with Watana alone would expose more of the lower, steeper portions of the Susitna River canyon, as well as a greater overall depth of shoreline. Exposure of steeper slopes would produce higher gradients of shorefast ice and tend to facilitate ice-shelving. Moose crossing ice-draped reservoir slopes may be injured by slipping. However, the sloping surfaces would not necessarily be smooth; overflow onto the frozen impoundment would partially melt and freeze snow to the surface ice (Nilsson 198 la ,b), creating a coa rse texture and reducing the s ubsequent hazard to moving animals. Moose ma y be injured als0 by stepping intc snow-covered cracks. Mos t inju~ed moose would probably s tarve from an inability to move and forage efficiently in deep s now, or succumb to wolf 33RD2-007u predation. Although individual mo ose would occasionally die from ice-relatec. injuries of this nature, the resulting level of mortality is not likely t o !>•! important. In contrast, ice deposition in the drawdown zone would create a p o tentia lly i mportant hazard to ca ribou. Segment s of the Nelchina herd would cross the Watana impoundment southward in group migra tio11s from late April 1.0 mid-May, en r ou te to their calving grounds. The se c r ossings would occur whom the ice-covered drawdown zone is exposed to its maximum extent and, at t ·l e same time, unstable breakup cond itions exist (Pitche r 1984 , LC.L 1985). [n Norway and Sweden , groups of reincieer have been killed when attemp ting to negotiate similar ice conditions (K lein 1971, Villmo 1975). Along ste ~p banks, caribou unable to gain purchase o n ice-covered slopes may be forced to swim sufficiently t o deplete energy reserve s , which are at their lowest ebb ln late winter. At thi s time, pregnant fema les are in their poorest condit~on )f the year (Skoog 1968). An extended or unusually difficult migr a tion prior :o calving could result in higher-than-normal adult mor t ality rates and decreas~d viab ility of newborn calves, thereb y affecting herd p r oduc tivi ty (Pitch•!r 1984). If a la rge proportion of the Nelchina herd attempts to cross the Wa tara impoundment during hreakup, a substantial mortality could result from ice haza rd s in the water and a l ong the re servo ir banks. I t is not feasible t ) predict with accuracy the behavioral res ponses of caribou encountering spring ice hazards i n the impoundment , n o r to gauge the probable ex tent of mortalit) if caribou are trapped in the wa t er or injured on ice-covered slo pes . Ice-related effects would be impor t ant only if they produced mortaliLies con sistently f r om year to year, o r a single event large enough to suppre s s population levels t o a point at which annual calf rec r uitment wo u ld not of fs et 33RD2-007u losses to predation and other causes. Given the known resourcefulness of migrating caribou, the probability of s uch major mortalities is considered to be low, but sufficient to warrant routine annual monitoring of herd movements during breakup (LGL 1985). Dall sheep use of the Jay Creek minE!ral lick area may be affected slightly by shorefast ice. The drawdown zone would include the Jay Creek streambed immediately below the heavily -used Bluff ai.ld East Ridge lick sites. Sheep attempting to cross Jay Creek at this location may encounter some residual ice during May but would probably not be deterred from crossing. If ice-related hazards occur, wolves may benefit from a greater availability of weakened, injured, or dead moose and caribou, their major prey in the Susitna Basin (Bal l ard et al. 1984b). Smaller carnivores (e.g., coyote, red fox, wolverin e , marten) that feed on wolf-killed or winter-killed moose and caribou might also benefit; the extent of any benefit to these species would depend on the distribution and number of injured or killed animals. If mortalities occur over a sufficient area of the impoundment margin to include many different carniv ore territories or home ranges, and occur consistently from year to year, the beneficial effect of thi.s enhanced winter food availability on carnivore populations could be important. However, as explained above, moose and caribou mortalities of this magnitude are unlikely . (c) Deterioration of reservoir ice cover in spring. The slow deterioration of the melting reservoir ice cover would present a barrier to movements across t he Watana impoundment from early May through early June. The unstable ice conditions would be hazardous to moose and caribou throughout this period. As meltout progresses, there would be an increasing probability 33RD2-007u that moose or caribou will fall through the ice cover and be unable to regain solid footing. Swimming animals would encounter numero us ice floes, produced as the melting reservoir surface fractures. These may impede crossings and, where prevailing winds cause pile-ups, delay or prevent anicals from leav ing the impoundment. In the spring, some female moose cross the Watana impoundment area in either direction and calve on the opposite side. The majority of females probably do not cross the river prior to calving, as vegetative cover used for calving exists on both sides, and crossings appear to be infrequent. Parturition generally occurs in the middle Susitna Basin from May 15 though June 15, peaking between May 25 and June 2 (Ballard et al. 1982). Individual moose attempting to cross the Watana reservoir during this period would encounter unstable ice conditions. However, suitable calving habitat would remain on both sides of the Watana impoundment after filling, and the existing pattern of calving would probably continue. Therefore, although moose may be lost while attempting spring crossings on unstable ice, this loss is not likely to be important because relatively few individuals would be affected. Migrating caribou normally encounter hazardous breakup conditions during spring crossings of rivers and lakes, and the melting reservoir ice cover would probably not have an important adverse effect on the Nelchina herd . As discussed above, caribou are more likely to be affected by ice-related hazards as they reach the drawdown zone. (d) Accumulation of windblown snow along i mpoundment shoreline. Prevailing northeast winds would tend to sweep snow from the frozen reservoir surface, producing drifts along the southwest shore. Because of winter drawdown, it is likely that much of the windblown snow accumulation would be 33R!Y~-007u 2(,7 confined to the immediate reservoir area. However, snow would also accumulate in the vegetation growing above the edge of the high-water level (Nilsson 198la,b). The magnitude of effects of snow drifting on vegetation and wildlife would depend on such fartors as prevailing wind direction, fetch, wind velocities, cumulative snow depth, presence or absence of crusted layers in the snow profile, proportion of reservoir surface snow melted and/or frozen by overflow, slope of exposed impoundment shorelines, local varia tiona in shoreline topography, and vegetation types on the windward reservoir margin. Snow accumulation on the reservoir surface would not occur until after the surface freezes during November. Direct observations of snow accumulation along the downwind shorelines of lakes in the middle Susitna Basin (e.g., the Fog Lakes) indicate that snow tends to be removed from exposed areas and redeposited in downwind drifts behind trees or topographic irregularities (Steigers 1985 pers. comm.). However, snow drifting is a dynamic phenomenon, and drifts would tend to shift or be remov ed by changes in wind speed or direction. Although snow accumulations would occur along downwind shorelines of the Watana impoundment, especially in wider areas of the reservoir where fetch is increased (e.g., opposite Watana Creek, where the impoundment will be 4.2 miles wide), snow drifting along shorelines would be partially offset by shifting winds, sublimation of snow, and water overflow onto the reservoir ice cover, whi.ch would not melt and freeze it in place (Nilsson 198la,b). The effects of windblown snow accumulation on wildlife are not expected to be important. Only moose and their primary winter predators, wolves, would potentially be affected. Snow drifting along the reservoir shoreline is not expected to cover sufficient browse to produce a population-level food shortage. Although deep snow may hinder the mobility and browsing efficiency of moose, increasing their vulnerability to wclf predation, these effects are 33RD2-007u more likely to be important during a severe winter with deep snowfall, rather than as a result of local snow drifting. Snow accumulation would fill cracks and cover irregular ice formations along windward s horelines, reducing the hazard potential for moose and caribou; these effects have been observed at the Williston Lake reservoir in northern ~ritish Columbia (Thomas 1982 pers. comm .). (e) Increased extent of open water during winter. The approximately 36 miles of open water between Watana dam and the mouth of Devil Canyon, coupled with much higher winter flows than under pre-project conditions, would inhibit river crossings by mo vse in this reach. However, since winter moose crossings are infrequent in the De v il Canyon impoundment area (Ballard et al. 1984a), the effect of open water would not be important . The ~xtensive reach of open water would provide locally production foraging habitat for mink and river otter. Augmented winter flows would increase water depth a nd backwater volume at tributary mouths, providing inc reased overwintering h~bitat for stream-dwelling fish that would congregate in high densities at these locations (Alaska Power Authority 1983). Access to this improved winter food supply may have an important beneficial effect on mink and river otter populations . Open water availability in late fall and early spring, when other waterbodies are froze n, 'lo•ould attract and potentially benefit migrant waterbirds by affording safe resting areas . Although fish and invertebrate prey bases are expected to be generally low i n the impoundment zone (Alaska Power Authority 1983), fish overwintering areas near tributary mouths would provide food appropriate for some migrants, as well as for overwintering 33RD2-007u mergansers and bald eagles. The number of individual birds affected would probably be too small to be important. (2) Watana and Devil Canyon Dams on Line. With both the Watana and Devil Canyon dams in operation, ice-related effects on wildlife would be similar to those discussed above for the Watana dam alone. Because of its narrower width and steeper sides , the De v il Canyon impoundment area currently provides much less habitat value for overwintering wildlife than the Watana impoundment area (Alaska Power Authority 1983). Although many of the ice-related effects described above for the Watana impoundment would also occur in the Devil Canyon impoundment, they would be minimal in comparison. Caribou do not presently, and hav e not historically, crossed the Devil Canyon impoundment area in large numbers during seasonal migrations (Hemming 1971; Pitcher 1983, 1984), and few winter crossings by moose have been recorded (Ballard et al. 1984a). With both dams operating, drawdo•."D in the Watana impoundment would be reduced from ab out 90 feet to about 40 feet, min in. "'_zing the expos ure of lower, steeper canyon sides and correspondingly reducing haz ards associated with ice-covered slopes in the drawdown zone . This would have a beneficial effect on moose and caribou in comparison with the Watana dam alone. Devil Canyon reservoir would experience little drawdown in winter. In terms of ice-related conditions only, the Watana and Devil Canyon dams together would produce fewer adverse eff ects on wildlife than the Watana dam alone, because the benefit of the reduced Watana drawdown zone would offs et most ice -related adverse effects of the Devil Canyon impoundment. b. Middle River 33RD2-007u (1) Watana Dam. (a) Longer open-water period and larger open-water areas resulting from higher temperatures of regulated flows . With Watana only, freezeup in the middle river would be delayed 17 to 42 days and the ice would melt 28 to 48 days earlier than under pre-project conditions. Moreover, the ice front would reach only to the Curry-Gold Creek vicinity (RM 124 to 142), with its final location depending on ambient temperature and the temperature and volume of released flews. This will produce an open-water reach up to 6 0 miles l o ng downstream from the Watana dam. Under pre-project winter conditions, mo o se frequently cross the frozen middle river and use it extensively as a corrido r for upstrea~ and downstream travel, avoiding the deeper snow on surro unding land (Modafferi 1983). A prolonged open-water period would restrict these movements in the early winter and early spring. Persistent thin ice and open leads may cause increased mortality; even under pre-project conditions, moose attempting to cross the river someti mes fall through the ice and are killed (Modafferi 1983, Schock 1985 pers . comm.). Browse v egetation on islands rendered inaccessible by open water would not be available t o wintering moose, and in a few cases, female mo o se would be restricted in the early spring from reaching islands where they might otherwise have calved. None of these potentially adverse conditions is expected to affect a sufficiently large number of moose to be considered important, and wintering success of moose populations using the middle river would not appreciably change as a direct result of regulated flows. The lack of ice cover between Watana dam and the ice front, along with open leads downstream from the ice front, would provide foraging habitat for aquatic furbearers (river otter, mink, beaver, muskrat). As noted above, mink 33RD2-007u 27( and river otter would have increased access to overwintering fish. The extent to which beaver and muskrat would benefit from delayed freezeup and open leads during the winter is not clear. Beaver cache surveys in October 1984 indicated that there were no active colonies upstream of RM 140; therefore open water above the projected maximum ice front would not affect beaver unless the area is colonized in the future. Sixteen colonies, including an estimated 80 beaver, were identified between the projected maximum and minimum ice front locations (i.e., between RM 140 and 124). Colonies in this reach and farther downstream would be affected by delayed freezeup and open leads . These conditions may benefit beaver by extending the fall cache construction period, increasing the probability of overwinter survival. However, where open water allows beaver to reach the shore, border ice and deep snow would probably prevent them from foraging on land, and beaver that do forage on land during the winter would be highly vulnerable to wolf and coyote predation. In late winter or spring, 2-year-old beaver disperse from the parent colony (Leege 1968). Early spring melting of the ice cover, along with the absence of dynamic pre-project breakup conditions, may facilitate the dispersal of young beaver and extend the active season for adults. Prolonged open water in the middle river may increase available overwintering habitat for the common merganser, American dipper, and bald eagle. Mergansers and dippers would probably not benefit in sufficient numbers for this effect to be considered important. However, the availability of relatively ice-free side channels and sloughs may increase access of bald eagles to living fish and to salmon carrion . congregate near winter food sources, there Because bald eagles tend to is a potential for numerous individuals to benefit. However, because the project area is near the edge of bald eagle range (Alaska Power Authority 1983), population density is 33RD2-007u relatively low, and the increased winter feeding habitat may not attract a sufficient number of eagles to produce an important effect. Moreover, high turbidity may offset the value of open-water areas as eagle feeding habitats. (b) Higher staging resulting from increased winter flows. Higher staging would probably cover (with flowing water or ice) early-successional willow and balsam poplar otherwise available to moose as winter browse, and open water coupled with higher staging may prevent ~oose from reacl.ing islands where browse is available. Although access to browse plants would thus be reduced along the active floodplain, these riparian shrubs are relatively tolerant to flooding and would not be permanently affected. However, the higher staging, increased flows, and daily variations in flow would cause overflow of water onto the existing ice cover, forming a progressively thicker ice sheet (aufeis) that would spread to higher islands and terraces supportin~ mature shrubs and trees (Nilsson 198lb). The effect of yearly winter flooding and ice cover on mature floodplain vegetation cannot be predicted with certainty, and the probable locations and cumulative areal coverage of aufeis are not known. It is likely, however, that the annual spring development of trees, shrubs, and herbaceous plants overlapped with thick ice layers would be measurably delayed (Nilsson 198lb). The interrelated effects of higher winter flows and staging on moose mo "dlity, winter browse availability, and spring plant development are considered important to the long-term productivity of moose populations wintering in the middle river. Ice cover that persists into the late spring on banks and terraces will locally cover or delay the development of herbaceous floodplain vegetation foraged by black bears during the immediate post-denning period (Hiller and 33RD2-007u 273 McAllister 1982). However, the reduction in available forage will probably not be extensive enough to produce population-level effects on this species. Higer winter staging would hav e more important effects on furbearers . As desctlbed above, overtopping of floodplain islands, sloughs, and low terraces would flood habitat and in places form a lasting ice cover, removing hunting areas otherwise available to wolves, wolverine, mink, marten, and weasels. Beaver, however, would be the furbearer most sev erely affected by higher winter staging, because of the potential for flooding of lodges in the mainstem, side channels, and sloughs. Because summer and fall river stages would govern the siting and construction of beaver lodges and caches, higher winter staging would flood many of these sites, and the beaver in the colonies would be lost to drowning, starvation , and predation (Hakala 1952, Bo y ce 1974). Lodges within or nearest to the mainstem would be most sus ceptible to this effect, but side channels and sloughs would also be exposed to higher staging, and even upland sloughs may receive increased water from upwelling. October 1984 food cache surveys in the middl e river ide1 ified 45 beav er colonies preparing to overwinter between De v il Canyon and Talkeetna. Of these, 13 were in mainstem habitat, four in side channels, 14 in sloughs, and 14 in upland sloughs (Woolington et al. 1984). Based on these results, with an estimated five beaver per colony, the long-term capacity of the mi ddle river to support about 155 beaver in mainstetn, side-channel, and s l ough habitats could be permanently lost, with as many as 70 additional beaver potentially affected by higher water levels in upland sloughs. This would be an important adverse effect. Persistent s pring ice c over reduce available foraging and on flo odplain islands and terraces would nesting habi ~ats for shorebirds (e.g., semipalmated plover, spotted sandpiper), gulls (e.g., herring and mew gulls), 33RD2-007u 27 'I arctic terns, and potentially other species. However, few e x t~~s ive areas of early-successional riparian habitats selectively used by these species are present in the middle river (Kess~l et al. 1982a), and the effect on birds is therefore not expected to be important. (c) Early in situ melting of ice during spring breakup Warmer temperatures and reduced volumes of rel~ased flows in the spring would cause river ice to melt in place several weeks earlier than under pre-project conditions, avoiding the dynamic spring breakup drive that normally characterizes the Susitna and other northern rivers. This change would reduce the probability of moo s e injuries or mortality resulting from floating ice and debris (Modafferi 1983), but the small number of moose likely to benefit in this way would not appreciably affect population levels. A more important effect of changed breakup characteristics would result from habitat alteration . Ice-scouring associated with normal dynamic bre akup would be greatly reduced in the middle riv er zone under with-project conditions. As riparian vegetation encroaches into river channels towards the mean s uliUDer high-water line, the availability of early-successional browse would increase during approximately the first ten years of project operation. However, the spring ice-scouring that normally helps to maintain early-successional stages in the riparian zone would be largely absent. Over time, the quality of riparian habitat for wintering moose would decline as the early successional stages encouraged by reduced suliUDer flows mature and grow out of reach. Because it is likely to occur throughout the middle river, this reduction in moose winter habitat quality would become important during the life of the project. 33RD2-007u Reduced spring flows during the meltout period would allow stranded ice cover to remain in slou~hs and on overtopped islands and terraces. Because this ice would not be in contact with active river flows. it would tend to melt slowly and persist into the late spring in sheltered places. If this persistent ice cover is extensive in area, the resulting habit~t reduction and delay in spring plant development may be important to forbearers and birds. (2) Watana and Devil Canyon Dams on Line. With both dams in operation, ice-related effects on habitat and wildlife would be qualitatively similar to those described for the Watana dam alone. The annual open-water period would be slightly longer. as warmer released flows would delay freezeup and advance the spring melting of river ice. In addition, the ice front would extend only to between RM 123 and 133, maintaining an open-water reach up to about 30 miles long downstream from the Devil Canyon dam. Staging levels at freeze-up and short-term fluctuations in released flows would decrease slightly. These conditions would moderate, but not substantially reduce, the ice-related adverse effects discussed above. c. Lower River (1) Watana Dam on Line Alone. (a) Longer open-water period and larger open-water areas resulting from higher temperatures of regulated flows. With the Watana dam alone. ice would begin forming in the lower river at about the same time as under pre-project conditions (November), but the ice front would not reach the Susitna/Chulitna confluence until 17 to 44 days later than under pre-project conditions. Delay 33RD2-007u 27~ in the formation of a stable ice cover would correspondingly delay the period during which moose can safely cross the lower river zone. Under pre-proj(·ct conditions, moose freely cross this reach during annual migrations to win::er range, and extensively use the frozen lower river zone as a movement corridor (Modafferi 1982, 1984). Increased numbers and areas of open leads, coupled with delayed fre ~ze /up, may impede migratory movements, reduce access to browse vegetation, and contribute to direct mortality from falling thrcugh thin ice. However, variations in the timing of freezeup and extent of cpen water exist under pre-proj ect conditions, and moderate project-relcted increases in these factors are not expected to produce important advErse effects on moose. B~cause of the major contributions of flows from the Chulitna and Talkeetna rivers, along wi th lesser tributaries, ice-relc ted effects on ':egetation would not differ appreciably from pre-proj ect conditions. An increase in the annual ice -free period may benefit mink al"'d river otter by prolonging the availability of aquatic hunting habitat. This eff~ct would also allow beaver to remain active and store food later into the fall or winter, as cache construction continues until free z eup . In upstream reaches of tb~ lower riv er zone where these effects are Jreatest, they may be important to furbearer populations. lee-related effects in the lower river will not differ sufficently fr~m pre-project conditions to influence bird po~ulatio ns in importa nt ways. (2 ) Watana and Devil Cany on Dams on Line. (a) Longer open-water period and larger open-water a reas resulting fron higher temperatures of regulated flows . With both dams in operation, ice 33RD2-007u 27/ formation would be delayed slightly longer into the early winter, and spring melting would occur earlier. Effects on moose and furbearers would be similar to those described for the Watana dam alone. 33RD2-007u F. EFFECTS ON PUBLIC USE This section describes preproject ice conditions relative to public use of the Susitna floodplain. The effect of altered ice processes on winter-oriented public use activities are then discussed based on commentary from local residents and casual observations incidental to regular investigative field work. 1. NATURAL ICE CYCLE CHRONOLOGY Frazil ice appears in the upper river by late September and by late October, frazil slush accumulates into an ice cover that begins to form near Cook Inlet and extends upriver toward Talkeetna. Later in November the ice cover progresses into the middle river. Ice cover progression above Gold Creek is at a reduced rate due to a steeper gradient and less frazil ice generation. The reach from Gold Creek to Devil Canyon takes longer to freeze and differs further by this area's subjection to shore ice development and anchor ice daming. The upper river develops wide shore ice by forming successive layers of frazil and snow s l ush which causes the channel to narrow and eventually freeze into continuous ice cover with the entrapment of flowing slush. Open leads develop over turbulent water and may close as a result of fine slush accumulating against the downriver edge of the lead. Many open leads persist at intermittent locations along the river throughout the winter . Increased solar radiation early in April signals the pre-breakup period. This process begins in the lower river and gradually extends upriver. Snow disappears along the river south of Talkeetna by late April. With aG increase in snow melt discharge, the ice begins to lift and fracture as leads widen and small jams form downriver of leads. Breakup accelerates with increased 33RD1-007h 21 '7 discharge and more ice jamming. The ice cover continues to fragment. deteriorate and flow until the river in ice free usually in early May. 2. PUBLIC USE PERSPECTIVE Major public use activities may be categorized as outdoor recreation. commercial. and subsistence oriented functions. Major recreational activities during winter include dog mushing. ice fishing. snow mobi ling. small game hunting and cross country skiing. to trapping fur animals and Commercial activities are generally limited operating recreational lodging facilities. Subsistence oriented activities refer to gathering plant and animal material for personal consumptive purposes. Prevalent activities i:l this category include tree cutting for fire wood. house logs and saw timber as well as fishing and hunting for sustenance. 3. PRESENT USER GROUP ACTIVITY The level of public use varies by river zone and its relationship to population density. The river area between Devil Canyon and Watana Creek receives little if any public use during the winter period. There are no permanent or seasonal residents in this zone. The lack of public use is attributed to the area's inaccessibility • rugged topographic features and, perhaps. unusable ice sub3trate due to the river's gradient and treacherous shoreline. These conditions would not preclude occasional aircraft landings in select places since relatively straight and even stretches of river surface occur between Devil Canyon and Watana. Ice shelving provides a sufficient platform to land and temporarily moor air craft. however, such events were not evidenced during this study. Aircraft operators are deterred from using this 33RD1-007h river zone during winter because of the area's remoteness and unsafe landing conditions. Moderate public use occurs in the middle river. About 12 trappers reside and travel between the Chulitna confluence and Gold Creek area during winter. With some exception, these people wisely avoid using and crossing river ice because of unsafe conditions. The ice surface continually changes with new leads forming and freezing over in which case drifting snow may conceal an unsafe ice surface. In the area between Talkeetna and the Chulitna confluence only two river crossings considerable risk. Within (noted in separate years) were noted and a few days of each observation. river at ice snowmachine tracks gave way. Two trappers from the Chase area operate along either side of the river by crossing at a point where a freezeup ice jam forms a suitable thickness for safe crossings. non-existent in the middle river. Non-resident use appears to be The proximity to railbelt settlements and homesteaders, as well as the increased accessibility afforded by the Parks Highway. account for greater public use of the lower river zone. People in the Talkeetna. Sunshine, Montana. Caswell. Kashwitna and Willow areas use the river corridor for recreational. commercial and subsistence purposes. The preponderance of use occurs along the east side of the river that has highway access. Evidence of snowmachining is commonly observed along the river's edge and on frozen side channels. Variable ice conditions and accident potential deters the casual traveler from attempting to cross the river. River crossings do .ccur during the period of mid-winter ice coverage and appear more portions of the river between the Caswell area and prevalent in those the Deshka River confluence, and confluences of Alexander Creek and the Yentna River. Public use activities of major importance include trapping. ice fishing. dog mushing. 33RD1-007h 2. f( snowmachining and gathering firewood. Several commercial lodge owners accommodate a few recreationists from Anchorage and outlying urban areas. Nonresident use of the lower river zone appears to be negligible. 4. EFFECTS OF ALTERED ICE PROCESSES ON PUBLIC USE The effect of altered ice processes on public use wou "i.d vary by river zone and level of human activity known to occur within each zone. In the impoundment zone, the Watana reservoir would begin to freeze over around mid-November with the formation of border ice. Solid ice would be expected later in December and gradually increase in thickness to an estimated 4 feet as freezing temperatures persisted. Reservoir drawdown would ~ause near-shore ice to fracture and drape, thus creating an ice ramp. \tJa ter temperatures below the reservoir would remain above freezing and moving water would prevail throughout this portion of the zone. A similar scenario is expected to occur in the Devil Canyon reservoir except that ice fracture and drape along the periphery would be minimal or none because of a greatly reduced drawdown. Downriver from Devil Canyon, an open water reach would extend for several miles. Since public use of this zone has been minimal during winter these physical changes would have no appreciable effect on human activities. Future public use may be slightly enhanced since solid ice cover in the reservoir areas may provide suitable landing places for ski-equipped aircraft during the mid-winter period. Draping and the formation of an ice ramp after reservoir drawdown may deter or impede foot access to shoreline areas, should aircraft landings be attempted. Freezeup of the middle river would be delayed because of altered flow, warmer water temperatures and reduce inflow o f frazil ice. Under these conditions, the firmness of river ice would be untenable during mid-winter 33RD1-007h from the vicinity of Curry upriver to about Gold Creek. The altered flow regime would probably change the size and duration of open leads occurring dow:1river from the ice edge. Flooding and overtopping would occur in the downriver portion of the middle river freezeup. The limited public use this area currently receives would probably be significantly reduced. Mechanized and foot travel on an unstable ice substrate, expecially in the zone of instability between Curry and Gold Creek, would be extremely hazardous. The reach between the Chulitna confluence and Curry probably would not be adversely effected except that the duration of solid ice cover would be confined to a shorter period during January and February when leads and overflow reaches form enough solid ice cover to permit safe travel. The few individuals from the Talkeetna and Chase areas who trap fur animals on the opposite side of the river would probably cont ~~ue to operate at essentially the same lev el of effort and risk. Ice formation in the lower river is e x pected to be near natural cond~tions. Above Talkeetna, however, ice c over in the upriver portion is not expected to be solid enough for surface transportation until later in the winter, probably 17 to 47 days later than normal. Because of increased water flow from the middle river zone, more ice may form in the lower zone than under natural conditions. Public use activ ities are expected to oc cur at a near-no rmal level except that the outset of winter time activity could be delayed by a month or so to produce safe ice cover. However, surface transportation in the river corridor under natural conditions does not occur until users feel confident that safe operating condi tions exist. The size and number of open leads may increase as a result of more water flow in the lower river and could disrupt human activities to an undetermined degree. 33RD1-007h VI I. REFERENCES REFERENCE CITED -CHAPTER II: INTRODUCTION Acres American, Inc. 1983. Application for license for major project, Susitna Hydroelectric Project, before the Federal Energy Regulatory Commission. Vol. SA. Exhibit E, Chap. 2. Alaska Power Authority. Susitna Hydroelectric Project. 1 Vol . 33RD2-007y REFERENCES CITE -CHAPTER III: DESCRIPTION OF RIVER AND CHAPTER IV: RIVER ICE PROCESSES Calkins, Darryl J. 1983. Hydraulics, Mechanics and Heat Transfer for Winter Freezeup River Conditions. Class notes for: Ice Engineering for Rivers and Lakes, University of Wisconsin, Madison, Wisconsin. Michel, Bernard. 1971. Winter Regime of Rivers and Lakes. U.S. Army Corps of Engineers, Cold Regions Research and Engineering Laboratory, Hanover, New Hampshire. 130 pp. Needham, Paul and Jones, Albert. 1959. Flow, Temperature, Solar Radiation, and Ice in Relation Ecology, Vo. 40, No. 3. Newbury, Robert W. 1968. The Nelson River: A Study of Subarctic River Processes. University Microfilms, Inc., Ann Arbor, Michigan. 319 pp. Osterkamp, Tom E. 1978. Frazil Ice Foroation: A Review. Journal of the Hydraulics Division. Proceedings of the American Society of Civil Engineers . September, pp. 1239-1255. Perla, R. and Martinelli, M. 1976. Avalanche Handb r,ok. U.S. Department of Agriculture, Forest Service Handbook No. 489 . July. 198la. Susitna River Mile Index. Anchorage, Aalska. Alaska Power Authority. Susitna Hydroelectric Project. Report for Acres American, Inc. 1 Vol. R&M Consultants, Inc. 1981b. Preliminary Channel Geometry, Velocity and Water Level Data for the Susitna River at Devil Canyon. Anchorage, Alaska. Alaska Power Authority Susitna Hydroelectric Project. Report for Acres American, Inc. 1 vol. 1982a. Field Data Collection and Processing. Anchorage, Alaska. Alaska Power Authority. Susitna Hydroelectric Project. Report for Acres American, Inc. 3 vol. 1982b. Hydrographic Surveys Report. Anchorage, Alaska. Alaska Power Authority, Susitna Hydroelectric Project. Report for Acres American, Inc. 1 vol. U.s. Army Corps of Engineers. 1982. Ice Engineering, Engineer Manual No. 1110-2-1612. Washington, D.C. ADDITIONAL READING Alaska Department of Fish & Game. 1982. Susitna Hydro Aquatic Studies Phase II Basic Data Report. Anch0rage, Alaska. 5 vol . Ashton, George D. 1978. River Ice. Annual Reviews on Fluid Mechar.ics. Vol. 10. pp. 369-392. 33RD2-007y Benson, Carl S. 1973. Fairbanks, Alaska. A Study of the Freezing Cycle in an Alaskan Stream. Institute of Water Resources. 25 pp. Bilello , Michael A. 1980. A Winter Environmental Basin of the Upper Susitna River, Alaska. Army Corps of Engineers, Cold Regions Laboratory, Hanover, New Hampshire. 1 vol. Data Survey of the Drainage Special Report 80-19, U.S. Research and Engineering Edinger, J.E., et. al. 1974. Heat Exchange and Transport in the Environment. Baltimore Maryland. John Hopkins University. 124 pp. Gerard, Robert. 1984. Hydrological Effects of Ice. Cold Regions Hydrology and Hydraulics Session, California . October. Paper presented at the ASCE, San Francisco, Pariset, f ., Hausser, R. and Gagnon, A. 1966. Formation of Ice Covers and Ice Jams in Rivers. Journal of the Hydraulics Division, Proceed ins of the American Society of Civil Engineers. November. pp. 1-23. R&M Consultants, Inc. 1981a. Hydrographic Surveys Closeout Report. Anchorage, Alaska. Alaska Power Authority. Susitna Hydroelectric Project. Report for Acres American, Inc. 1 vol. 1981b. Ice Observations 1980-1981, Anchorage, Alaska . Alaska Power Authority. Susitna Hydroelectric Project. Report for Acres American, Inc. 1 vol. 1982b. Hydraulic and Ice Studies. Anchorage, Alaska. Power Authority. Susitna Hydroelectric Project. Report for American, Inc. 1 vol. 1982d. Ice Observations 1981-82 . Power Authority. Susitna Hydroelectric American, Inc. 1 vol. Anchorage, Alaska. Project. Report for Alaska Acres Alaska Acres 1982e. Processed Climatic Data, October 1981 to September 1982. Anchorage, Alas ka. Alaska Power Authority. Susitna Hydroelectric Project. Report for Acres American, Inc. 8 vol. 1982f . River Morphology. Anchorage, Alaska. Alaska Power Authority. Susitna Hydroelectric Project. Report for Acres American, Inc. 1 vol. 1982g. Slough Hydrology. Anchorage, Alaska. Alaska Power Authority. Susitna Hydroelect!'i ~ Project. Report for Ac;:-es American, Inc. 1 vol. 1983. Susitna River Ic e Study 1982-1983. Anchorage, Alaska . Alaska Power Authority . Susitna Hydroelectric Pro j ect. Report for Harza-Ebasco Susitna Joint Venture. 1 Vol. 33RD2-007y 1984. Susitna River Ice Study 1983-1984. Anchorage, Alaska. Alaska Power Authority. Susitna Hydroelectric Project. Report for Harza-Ebasco Susitna Joint Venture. 1 Vol. 1985. Susitna River Ice Study, Freezeup 1984. Anchorage, Alaska. Alaska Power Authority. Susitna Hydroelectric Project. Report for Harza-Ebasco Susitna Joint Venture. 1 Vol. Smith, D.G. 1979 . Effects of Channel Enlargement by River Ice Processes on Bankfull Discharge in Alberta, Canada. Water Resources Research, Vol. 15, No. 2 (April). pp . 469-475. U.S. Geological Survey. 1982. Water Resources Data, Alaska, Water Year 1981. Anchorage, Alaska. Water Resources Division, U.S. Geological Survey. United States Department of the Interior. PERSONAL COMMUNICATION Lavender, Thomas. 1981. Personal Communication. 33RD2-007y REFERENCES CITED -CHAPTER V: WITH-PROJECT STUDIES Acres Consulting Services, Ltd . 1980, Behavior of Ice Covers Subject to Large Daily Flow and Level Fluctuations, For the Canadian Electric Association. Acres American, Inc. 1983, Susitna Hydroelectric Project, Application for FERC License, Volume SA, Exhibit E, CHapter 2 for the Alaska Power Authority . Alaska Dept. of Fish and Game. 1983 . Susitna Hydro Aquatic Studies, Phase II Data Report, Winter Aquatic Studies (October 1982-May 1983). Alaska Dept . of Fish and Game, Susitna Hydro Aquatic Studies. 1985. An Evaluation of the Incubation Life Phase of Chum Salmon in the Middle Susitna River. Alaska Power Authority . 1984. Comments on the Federal Energy Regulatory Commission Draft Environmental Impact Statement of May 1984, Volume 9, Appendix VII -Slough Geohydrology Studies . Alaska Power Authority. 1985. Susitna Hydroelectric Project CASE E-VI Alternative Flow Regime, Submittal to FERC. Arctic Environmental Information and Data Center (AEIDC). 1983. Stream Flow and Temperature Modelling in the Susitna Basin, Alaska, for Hc:<rza-Ebasco Susitna Joint Venture for the Al aska Power Authority. Gatto, L.W. 1982. Reservoir Bank Erosion Caused and Influenced by Ice Cover, Special Report 82-31 for the U.S. Army Corps of Engineers, Cold Regions Research and Engineering Laboratory. Gerard R. 1983. Notes on Ice Jams for Ice Engineering for Rivers and Lakes, University of Wisconsin Extension. Harza-Ebasco Susitna Joint Venture. 1984a. Eklutna Lake Temperature and Ice Study, for the Alaska Power Authority. Harza-Ebasco Susitna Joint Venture . 1984b . Instream Ice, Calibr.ation of Computer Model, for the Al1ska Power Authority. Harza-Ebasco Susitna Joint Venture. 1984c. Lower Susitna River Sedimentation Study, Project Effects on Su s pended Sediment Concentration, (draft) for the Alaska Power Authority. Harza-Ebasco Susitna Joint Venture. 1984d. Susitna Hydroelectric Project Reservoir and River Sedimentation, for the Alaska Power Authority. Imberger, J ., and Patterson J.C. 1981. A Dynamic Reservoir Simulation Model-DYRESM:S, Transport Models for Inland and Coastal Waters, Chapter 9, Academic Press. L~Belle, Joseph C. 1984. Geomorphic Change in the Devil Canyon to Talkeetna Reach of the Susitna River since 1949 (preliminary draft) for Harza-Ebasco Susitna Joint Venture for the Alaska Power Authority. 33RD2-007y Peratrovich, Nottingham and Drage, Inc. 1982. Susitna Rese~oir Sedimentation and Water Clarity Study, prepared for Acres American, Ire., for the Alaska Power Authority. R&H Consultants. Inc. l982a. Winter 1981-82. Ice Observations Report. for Harza-Ebasco Susitna Joint Venture for the Alaska Power Authority. R&H Consultants, Inc. 1982b. Susitna Hydroelectric Project. River Morphology. for Acres American. Inc •• forth~ Alaska Power Authority. R&H Consultants. Inc. 1984a. Susitna River Ice Study. 1983-84. Draft Report. for Harza-Ebasco Susitna Joint Venture for the Alaska Power Authority. R&H Consultants, Inc. 1984b. Susitna Hydroelectric Project, Susitna Fiver Ice Study. 1982-83 for Harza-Ebasco Susitna Joint Venture. for the Alaska Power Authority. Sayre, W.W. and Song, G.B. 1979, Effects of Ice Covers on Alluvial Chc:nnel Flow and Sediment Transport Processes. prepared for U.S. Geolo@iC~l Survey, IIHR Report No. 218, Iowa Institute of Hydrc.ulic Research, The University of Iowa, Iowa City, Iowa. PERSONAL COMMUNICATION Chaco, E. 1985. Personal Communication. 33RD2-007y REFERENCES CITED -CHAPTER VI, PART C: EFFECTS ON FISHERIES Alabaster, J. S., and R. Lloyd . 1982. Water quality criteria for freshwater fish. 2nd ed. Butterworth Scientific, Boston, MA. 361 pp. Alaska Dept. of Fish & Game. 1976. Fish and wildlife studies related to the Corps of Engineers Devil Canyon, Watana Reservoir Hydroelectric Project. Anchorage, AK. Report for U.S. Fish & Wildlife Service. 1 vol. Alaska Dept. of Fish & Game. 1978. Preliminary environmental assessment of hydroelectric development on the Susitna River. Anchorage, AK. Report for U.S. Fish & Wildlife Service. 1 vol. Alaska Dept. of Fish & Game. 1983a. Susitna hydro aquatic studies, phase 2 final data repoTt. Vol. 2. Adult anadromous fish studies, 1982. Final Report. Anchorage, AK . Alaska Power Authority. Susitna Hydro Aquatic Studies. Report for Acres American, Inc. 2 vols. Alaska Dept. of Fish & Game. 1983b. Susitna hydro aquatic studies, phase 2 basic data report. Vol. 5. Upper Susitna River impoundment studies, 1982. Final Report. Anchorage, AK . Alaska Power Authority. Susitna Hydro Aquatic Studies. Report for Acres American, Inc. 150 pp. Alaska Dept. of Fish & Game. 1983c. Susitna hydro aquatic studies, phase 2 basic data report. Vol. 3. Resident and juvenile anadromous fish studies on the Susitna River below Devil Canyon, 1982. Final Report. Anchorage, AK. Alaska Power Authority. Susitna Fydro Aquatic Studies. Report for Acres American, Inc. 2 vols. Alaska Dept . of Fish & Game. 1983d . Susitna hydro aquatic studies, phase 2 data report. Winter aquatic studies (October 1982 -May 1983). Final report. Anchorage, AK. Alaska Power Authority . Susitna Hydro Aquatic Studies. Report for Harza-Ebasco Susitna Joint Venture. 137 pp. Alaska Dept . of Fish & Game. 1985a. Proposed Task Statement for FY86 Su-Hydro Scope of Work. Alaska Dept . of Fish & Game. 1985b . An evaluation of the incubation life phase of chum salmon in the middle Susitna River . Draft. Anchorage, AK. Susitna Hydro Aquatic Studies Winter Aquatic Investigations, September 1983 -~~y 1984 . 1985 Report 5. Vol. 1. Report for Alaska Power Authority. 157 pp . Alaska Dept. of Fish & Game. phase of chum salmon in the Susitna Hydro Aquatic September 1983 -May 1984 . 1 vol. 1985c. An evaluation of the middle Susitna River. Draft. Studies Winter Aquatic 1985 Report 5. Vol. 2. incubation life Anchorage, AK . Investigat:!.ons, Appendices A-F. Alaska Power Authority . 1980. A report on the first series of community meetings on the feasibility studies for the Susitna Hydroelectric Projec t and other power a lternatives, April 1980, Fairbanks, Talkeetna, Wasilla, Anchorage. Anchora ge, AK . 61 pp . 33RD2-007y Alaska, Univ., Arctic Environmental Information & Data Center. 1982. Summary of environmental knowledge of the proposed Grant Lake hydroelectric project area. Alaska Power Authority , Anchorage, AK. Report for Ebasco Services . 212 pp. Alaska, Univ., Arctic Environmental Information & Data Center. 1983. Methodological approach to quantitative impact assessment for the proposed Susitna hydroelectric project. Alaska Power Authority . Susitna Hydro Aquatic Studies. Anchorage, AK. Report for Harza-Ebasco Susitna Joint Venture. 71 pp. Alaska, Univ., Arctic Environmental Information & Data Center . 1984. Assessment of the effects of the proposed Susitna Hydroelectric Project on instream temperature and fishery resources in the Watana to Talkeetna reach. Final report. Anchorage, AK. Alaska Power Authority. Susitna Hydroelectric Project. Report for Harza-Ebasco Susitna J oint Venture. 2 vol. in 1 (looseleaf). Bailey, J.E., and D.R. Evans. 1971. The low-temperature threshold for pink salmon eggs in relation to a proposed hydroelectric installation. Fisheries Bulletin. 69{3):587-593. Bams, R.A. 1967 . A review of the literature on the effects of changes in temperature regime of developi ng sockeye salmon eggs and alevins. Journal of Fisheries Research Board of Canada. Manuscript 949:14-22 . Barrett, B.M. 1974. An assessment of the anadromous fish populations in the upper Susitna River watershed between Devil Canyon and the Chulitna River. Alaska Div. of Commercial Fisheries, Anchorage, AK. 56 pp. Barrett, B.M., F.M. Thompson, and S.N. Wick. 1984. Adult anadromous fish investigations: Hay -October 1983. Alaska Dept. of Fish & Game, Anchorage, AK. Susitna Hydro Aquat i c Studies . Report 1. Report for Alaska Power Authority. 1 vol . Barrett, B.M., F.M. Thompson, and S.N. Wick. 1985. Adult anadromous fish investigations (Hay-October 1984). Draft report. Alaska Dept . of Fish & Game, Anchorage, AK . Susitna Hydro Aqua t ic Studies. Report 1 . Report for Alaska Power Authority. 1 vol . Beaver, D.W. 1984. Slough discharge re ~r ession relations. E.J. Gemperline, Harza-Ebasco Susitna Joil't Venture, October 12. Benson, N. Ecology. 1955. Observations of anchor ic~ in a Michigan trout 36(1) :529-530. Memo to 4 pp. stream. Buklis, L.S., and L.H . Barton. 1984 . Yukon River fall chum salmon biology and stock status. Alaska Dept. of Fish & Game, Anchorage, AK . 67 pp. Bustard , D.R., and D.W. Narver. 1975. Aspects of water ecology of juvenile coho salmon (Oncorhynchus kisutcn) and steelhead trout (~ gairdneri). Journal of the Fisheries Research Board of Canada. 32(5):667-680. 33RD2-007y Canada Dept. of Fisheries & Oceans. 1984. Toward a fish habitat decision on the Kemano Completion Project: a discussion paper. Vancouver, British Columbia, Canada. 78 pp. Chapman, D.W., and T.C. Bjornn. 1969. Distribution of salmonids in streams with special reference to food and feeding. Pages 153-176 in T.G. Northcote, ed . Symposium on Salmon and Trout in Streams. University of British Columbia, Vancouver, B.C. H.R. MacMillan Lectures in Fisheries. Combs, B.D. 1965. Effects of temperature on the development of salmon eggs. Progressive Fish-Culturist. 27:134-37. Dugan, L., D. Sterritt, and M. Stratton. 1984. The distribution and relative abundance of juvenile salmon in the Susitna River drainage above the Chulitna River confluence. Draft report. Part 2 of D.C. Schmidt, S.S. Hale, and D.L. Crawford, eds. Resident and juvenile anadromous fish investigations (May-October 1983). Alaska Dept. of Fish & Came, Anchorage, AK. Susitna Hydro Aquatic Studies. Report 2 . 1 vol. Harza-Ebasco Susitna Joint Venture. 1984. Final report. Alaska Power Authority. 1 vol. Instream ice simulation study. Susitna Hydroelectric Project. McNeil, W.J. 1969. Survival of pink and chum salmon eggs and alevins. Pages 101-117 in T.G. Northcote, ed. Symposium on Salmon and Trout in Streams. University of British Columbia, Vancouver, B.C. H.R. MacMillan Lectures in Fisheries. McNeil, W.J. 1980. Vulnerability of pink salmon populations to natural and fishery mortality. Pages 147-131 in W.J. McNeil and D.C. Himsworth, eds. Salmonid ecosystems of the North Pacific. Oregon State University Press, Corvallis, OR. Mason, B.J. Physics. 1958. Supercooling and nucl~~~~on of water. 7:2-49. Advances in Merrell, T. R. 1962. Freshwater survival of pink salmon at Sash in Creek. Pages 59-72 in N .. J. Wilimovsky, ed. Symposium on Pink Salmon. University of British Columbia, Vancouver, B.C., 1960. H.R. MacMillan Lectures in Fisheries. Michel, B. 1971. Winter regime of lakes and rivers. Research and Engineering Laboratory. Monograph 3. U.S. Army Cold Regions 131 pp. Needham, P.R., and A.C. Jones. 1959. Flow, temperature, solar radiation, and ice in relAtion to activities of fishes in Sagehen Creek, California. Ecology. 40:465-474. R&M Consultants, Inc. 1984. Susitna River ice study 1983 -1984. Draft report. Alaska Power Autho rity . Susitna Hydroelectric Project. Report for Harza-Ebasco Susitna Joint Venture. 1 vol. 33RD2-007y Reiser, D.W., and T.C. Bjornn. 1979. Habitat requirements of anadromous salmonids. No. l in Influence of forest and rangeland management on anadroruous fish habitat in the western United States and Canada. U.S. Forest Service, Portland, OR. General Technical Report PNW-96. 54 pp. Riis, J.C. 1977. Pre-authorization assessment of the proposed Susitna River Hydroelectric Projects: preliminary investigations of water quality and aquatic species composition. Alaska Div. of Sport Fish, Anchorage, AK. 91 pp. Schmidt, D.C., et al., eds. 1984. Resident and juvenile anadromous fish investigations (May -October 1983). Parts 1-7. Alaska Dept. of Fish & Game, Anchorage, AK. Susitna Hydro Aquatic Studies. Report 2. Report ior Alaska Power Authority . 1 vol. Sundet, R., and M. Wenger. 1984, Resident fish distribution and population dynamics in the Susitna River below Devil Canyon. Draft report. Part 5 of D.C. Schmidt, S.S. Hale, and D.L. Crawford, eds . Resident and juvenile anadromous fish investigations (May-October 1983). Alaska Dept. of Fish & Game, Anchorage, AK. Susitna Hydro Aquatic Studies. Report 2 . 1 vol. Wangaard, D.B., and C.V. Burger. 1983. Effects of various water temperature regimes on the egg and alevin incubation of Susitna River chum and sockeye salmon. Final report. National Fishery Research Center, U.S. Fish & Wildlife Service, Anchorage, AK. 43 pp. Woodward-Clyde Consultants, and Entrix, Inc. 1985. Fish resources a,d habitats in the middle Susitna River . Draft final report. Anchorage, AK. Alaska Power Authority. Susitna Hydroelectric Project. Instream Flow Relationships Report Series. Technical Report 1. Report for Harza-Ebasco Susitna Joint Venture. 159 pp. PERSONAL COMMUNICATIONS Crawford, D. 1985. Telephone conversation, March 4. Alaska Dept. of Fish & Game, Anchorage, AK. ~tratton, M. 1985. Interview, February 14. Su-Hydro Gold Creek Camp, AK. Sundet, R. 1985. Telephone conversation, March 6 . Alaska Dept. of Fish & Game, Anchorage, AK. 33RD2-007y REFERENCES CITED -CHAPTER VI, PART E: EFFECTS ON WILDLIFE Alaska Power Authority. 1983. Susi tna Hydroelectric Project, Application, Exhibit E, Chapter 3. Submitted to Federal Regulatory Commission by Alaska Power Authority. License Energy Ballard, W.B., C.L. Gardner, J.H. Westlund and J.R. Dau . 1982. Moose- Upstream. Vol. III. In : Susitna Hydroelectric Project big game studies, Phase I Final Rep . Report by Alaska Dept. Fish and Game to Alaska Power Authority. Anchorage, 119 pp. plus appendices. Ballard, W.B., J.S. Whitman, N.G. Tankersley, L.D . Aumiller and P. Hess~~~. 1983a. Moose-Upstream. Vol. III. In: Susitna Hydroelectric Project. big game studies, Phase II Prog . Rep . Report by Alaska Dept. Fish and Game to Alaska Power Authority. Anchorage. 141 pp. Ballard, W.B., J.S. Whitman, L.D . Aumiller and P. Hessing. 1983b. Susitna Hydroelectric Project Phase II Progress Report. Big Game Studies. Vol. 5 -Wolf . Alaska Dept. Fish and Game. Prepared for Alaska Power Authority . 40 pp . Ball.:1rd, W.B., J .S . Whitman, N.G. Tankersley, L.D. Aumiller and P. Hessing. 1984a. Moos t.:-Upstream. Vol. III. In: Susitna Hydroelectric Project big game studies, 1983 Ann. Rep. Report by Alaska Dept. Fish and Game to Alaska Powr Authority . Anchorage. 147 pp. Ballard, W.B ., J.S . Whitman, L.D. Aumiller and Hydroelectric Project 1983 Annual Report. Wolf. Alaska Dept . of Fish and Game . Authority. 41 pp . P. Hessing. 1984b . Susitna Big Game Studies. Vol. 5 - Prepared for Alaska Power Boyce, M. 1974 . Beaver population ecology in interior Alaska. M.S. Thesis. Univ. of Alaska , Fairbanks. 161 pp. Buskirk, S.W . 1983. The ecology of marten in southcentral Alaska. ?h .D. Thesis. Univ. of Alaska, Fairbanks. 131 pp. Gardner, C.L. and W.B. Ballard . 1982. Susitna Hydroelectric Project Phase I Final Report. Big Game Studies. Vol. 7 -Wolverine. Alaska Dept . of Fis h and Game. Prepared for Alaska P ow~r Authority . Gipson, P .S., S.W . Buskirk and T.W. Hobgood. 1982. Susitna Hydroelectric Project Phase I Final Report . Furbearer Studies. Alaska t::oop. Wild!. Res. Unit , Univ . of Alaska, Fairbanks. 81 pp. Gipson, P.S ., S .W. Buskirk, T.W. Hobgood and J.D. Woolington . 1984. Hydroele c tric Project Phase I Report Update. Furbearer Studies. Coop. Wild!. Res . Unit , Univ. of Alaska, Fairba nks . 100 pp. Susitna Alaska Hakala , J.B. 1952. The life history and general ecology of (Ca s tor canndensis K.) in interior Alaska . M.S. Thesis. Alaska, Fairbanks. 181 pp. the beaver Univ. of 33RD2-007y Hanscom, J. T. and T. E. Osterkamp. 19 8 0. Potential caribou-ice problems in the Watana reservoir, Susitna Hydroelectric Project. The Northern Engineer 12:4-8. Hemming, J.E . 1971. The distribution and movement patterns of caribou in Alaska. Alaska Dept. of Fish and Game, Wildl. Tech. Bull. No. 1, 1971. Herter, D.R. 1985. Winter bird surveys of the middle Susitna basin . Report prepared by LGL Alaska Research Associates, Inc. for Harza-Ebasco Susitna Joint Venture. Anchorage. [In preparation). Hobgood, T.W. 1984. Red fox ecology in the upper Susitna Basin, Alaska. M.S. Thesis. Univ. of Alaska, Fairbanks. 151 pp. Kessel, B., S.O. MacDonald, D.D. 1982a. Susitna Hydroelectric non-game mammals. Report university of Alaska Museum. Gibson, Project. prepared Fairbanks. B.A. Cooper and B.A. Phase 1 Final Report. for Alaska Power Anderson. Birds and Authority. Kessel, B., D.D. Gibson, S.O. MacDonald, B.A. Cooper and K.C. Cooper. 1982b. Avifauna of the lower Susitna River floodplain. Report prepared under contract to LGL Alaska Research AssociaLes, Inc., for Alaska Power Authority. University of Alaska Museum. Fairbanks. Klein, D R. 1971. Reaction of reindeer to obstructions and disturbances. Science 173 :343-398. LGL Alaska Research Associates, Inc. 1985. Mitigation plan for wildlife and botanical resources. Susitna Hydroelectric Project. Prepared for Alaska Power Authority. Anchorage. [In preparation.] Leege, T.A. 1968. Natural movements of beavers of southeastern Idaho. J. Wildl. Manage. 32:973-976. Lieb, J.W., R.W. Tobey, and S.H. Eide. 1985. Analysis of Nelchina Caribou Range. Draft Report. Alaska Dept. of Fish and Game. Anchorage. Miller, S.D. and D.C. McAllister. 1982. Susitna Hydroelectric Project. Phase I final report. Big game studies. bear. Alaska Department of Fish and Game. Vol. VI. Black bear and brown Anchorage. 233 pp. Miller, S.D. 1983. Susitna Hydroelectric Project. Phase II progress report. Big game studies. Vol. VI. Black bear and brown bear. Alaska Department of Fish and Game. Anchorage. 99 pp. Miller, S.D . 1984. Susitna Hydroelectric Project. Phase II second annual progress report. Big game studies. Vol. VI. Black bear and brown bear. Alaska Department of Fish and Game. Anchorage . 174 pp. Modafferi, R.D. 1982. Hydroelectric Project Alaska Dept. Fish and pp. 33RD2-007y Moose-Downstream. Vol. II. In: Susitna big game studies, Phase I Final Rep. Report by Game to Alaska Power Authority. Anchorage . 114 Modafferi, R.D. 1983. Moose-Downstrec.m. Vol. II. Hydroelectric Project big game studies, Phase II Prog. Alaska Dept. Fish and Game to Alaska Power Authority. pp. In: Susitna Rep. Report ty Anchorage. 116 Modafferi, R.D. 19&4. Moose-Downstream. Vol. II. In: Susitna Hydroelectric Project big game studies, 1983 Ann. Rep. Report by Alaska Dept. Fish and Game to Alaska Power Authority. Anchorage. 116 pp. Nilsson, C. 1981a. Dynamics of the shore vegeta!:ion of a North Swedish hydroelectric reservoir during a 5-year period. ~eta Phytogeogr. Suec. 69. Uppsala. 96 pp. Nilsson, C. 1981b. Riparian vegetation of northern Swedish rivers . Wahlen- bergia 7:113-124. Ume~. Sweden. Pitcher, K.W . 1983. Susitna Hydroelectric Project, Phase II Progress Report. Big Game Studies, Vol. IV -Cari" 10u. Alaska Dept. of Fish and Game. Prepared for Alaska Power Authority. Pitcher, K.W. 1984. Susitna Hydroelectric Project, 1983 Ann. Rep . Studies, Vol. IV. -Caribou. Alaska Dept. of Fish and Game. for Alaska Power Authority. Big Game Prepared Pitcher, K.W. 1985. Susitna Hydroeletric Project, Draft 1984 Ann. Rep. Big Game Studies, Vol. IV. -Caribou. Alaska Dept. of Fish and Game. Prepared for Alaska Power Authority. Roseneau, D.G. 1984. Summary Statement on Nest Losses and Conflicts for Bald and Golden Eagles in the Susitna Hydroelectric Project Area. LGL Alaska Research Associates, Inc. Susitnd Hydroelect~ic Project. Prepared for Alaska Power Authority. 10 pp. Skoog, R.O. Alaska. 1968. Ecology of the caribou (Rangifer tarandus granti) in Ph.D. Thesis. Univ. California, Berkeley. 699 pp. Soil Conservation Service. 1985 . Snow surveys and water supply outlook for Alaska. Soil Conservation Serv ., U.S. Dept. of Agric. March 1985. Tankersley, N.G. 1983. Susitna Hydroelectric Project. Phase II Progress Report. Big Game Studies, Vol. 8 -nall Sheep. Alaska Dept . of Fish and Game. Prepared for Alaska Power ~uthority. Tankersley, N.G. 1984. Susitna Hydroelectric Project. Game Studies, Vol. 8 -Dall Sheep. Alaska Dept. Prepared for Alaska Power Authority. Final Report. Big of Fish and Game. Tobe y , R.W. 1981. Susitna Hydroelecuic Project. Annual Progress Report. Big Game Studies , Vol . 8 -Dall Sheep . Alaska Dept. of Fish and Game. Prepared for Alaska Power Authority. Van Ballenberghe, A.W. Erickson and D. B}'lJlan. 1975. Ecology of the timber wolf in northeastern Minnesota. ·Wildlife Monograph No. 43. The Wildlife Society. 43 pp. 33RD2-007y Villmo, L. 1975. The Scandinavian viewpoint . Pp. 4-9. In: J.R. Luik et al., ed., Proc. of First Int'l. Reindeer and Caribou Symp. Biol. Papers of the Univ. of Alaska Special Report No. l. Whitman, J.S. and W.B. Ballard. 1984. Susitna Hydroelectric Project. 1983 AnLual Report. Big Game Studies. Vol . 7 -Wolverine. Alaska Dept. of Fish and Game. Prepared for Alaska Power Authority . 25 pp. Woolington, J.D., P.S. Gipson and D. Volsen. 1984. Susitna Hydroelectric Project. Furbearer Studies. Fall 1984: Beaver. LGL Alaska Res. Assoc./ Alaska Coop . Wildl. Res. Unit, Univ. of Alaska, Fairbanks. Prepared for Alaska Power Authority. 30 pp. PERSONAL COMMUNICATIONS Modafferi, R.D. 1984. Game Biologist, Alaska Dept. of Fish and Game, Anchorage. Modafferi, R.D. 1985 . Alaska Dept. of Fish and Game. Schock, C. 1985. Hydrologist, R&M Consulta nts, Inc., Anchorage. Steigers, W.D. 1985. Wildlife Biologist, LGL Alaska Research Assoc., Inc ., Anchorage. Thomas, D. 1982. British Columbia Wildlife Branch. 33RD2-007y REFERENCES CITED -CHAPTER VI, PART D: EFFECTS ON RIPARIAN VEGETATION Eardley, A.J. 1938. Yukon channel shifting. Bulletin of the Geological Society of America 49:343-356. Gerard, R. (no date). Notes on ice jams. Department of Civil Engineering, University of Alberta, Edmonton, Alberta, Canada. T6G 2G7 17pp. Gill, D. 1973. development. Modification of northern alluvial Canadian Geographer 17:138-153. habitats by river Harza-Ebasco Susitna Joint Venture. 1984. Instream ice simulation study. Final Report. October 1984. Prepared for Alaska Power Authority. Helm, D. 1985. ~iparian succession report. (in preparation). Prepared for Harza-Ebasco Susitna Joint Venture for Alaska Power Authority. Helm, D. and P.V. Mayer. 1985. Plant phenology report 1983 (in preparation). Prepared for Harza-Ebasco Susitna Joint Venture for Alaska Power Authority. Kozlowski, T. T. 34:162-167. 1984. Plant responses to flooding of soil. Bioscience LaBelle, J. 1984. Geomorphic change in the Devil Canyon to Talkeetna reach of the Susitna River since 1949. Preliminary Report. Prepared for Alaska Power Authority. McKendrick, J. ; H. Plant ecology subtask 7.12. Collins; D. Helm; ..J. McMillen; and J. Koranda. 1982. studies, phase 1, final report. Environmental studies Prepared for the Alaska Power Authority. Miller, S.D., and D.C. McAllister. 1982. Big Game Studie'l Volume VI Black Bear and Brown Bear. Alaska Department of Fish and Game. Prepared for the Alaska Power Authority. Moberly, H.J., and W.B. Cameron. Sons, Ltd., Toronto. 151 pp. 1929. When fur was king. (cited in Gerard, R.). J.M. Dent adn Nilsson, C. 1981. Riparian vegetation of northern Swedish rivers. Wahlenbergia 7:113-124. R&M Consultants Inc. 1981. Ice observations 1980-1981. Task 3. Hydrology. Prepared for Alaska Power Authority. R&M Consultants. 1982. Winter 1981-1982. Ice Observations Report. Prepared for Alaska Power Authority. R&M Consultants. 1984 1982-1983 Susitna River lee Study. Final Report. Prepared for Alaska Power Authority . 33RD2-007y Schoch, C. 1985. Conversations throughout February 1985 regarding effects of Susitna River ice Processing on riparian vegetation. Schreiner, E.J. 1974. Populus L. Poplar. pp. 645-655 in : Schopmeyer, C.S. (coord.) Seeds of woody plants of the United States. Smith, D. 1981. Winter survival of forage stands. pp. 37-38 in: Forage Management in the North . Kendall/Hunt Publishing Company. Dubuque, Iowa. 258 pp. Zasada, J.D., and R.A. Norum, R.M. VanVeldhuizen, and Artificial regeneration of trees and tall shrubs in upland black spruce/feathermoss stands in Alaska. Forest Research 13:903-913. 33RD2-007y ::; 9 9 C. E. Tuetsch. 1983. experimentally burned Canadian Journal of VIII. GLOSSARY OF TEJU.IS AND DEFINITIONS* Agglomerate Anchor Ice Anchor Ice Dam Beginning of Breakup (Date) Beginning of Freezeup (Date) Black Ice Border Ice Breakup 33RD2-007r An ice floe formed by the bonding or freezing together of various forms of ice. Submerged ice attached or anchored to the bottom, irrespective of the nature of its formation. An accumulation o f anchor ice which acts as a dam and raises the water level . Date of definite breaking or movement of ice due to melting, currents or rise of water level. Date on which frazil ice, forming stable winter ice cover, is first observed on the water surface. Transparent ice formed in rivers and lakes. An ice sheet in the form of a long border attac hed to the shore and growing laterally out over the channel; same as shore ice, or lateral ice. Disintegration of ice cover. Breakup Date BS/sn Candle lee Channel Lead Degree-Day Dry Crack Floc 33RD2-007r The date on which a body of water is first observed t ,o be entirely clear of ice and remains clear thereafter. Rotten columnar-grained ice. Elongated opening in the ice cover caused by a water current or thermal erosion by warn gro undwater seeps. A measure of the departure o f the mean daily temperature from a given standard, usually 0 C. For example, a day with an average temperature cf -5 C represents S freezing degree-days . Cumula- tive degree-days are simply the sum of any number of degree-days. For example, the cumulative freezing degree-da7 s of a week with mean dail 1 temperature of -5 C, 0 C, +S C, -2 C, -5 C, -8 C, and -5 C are 25 freezing degree-days. Crack visible at the surface but which does not extend through the ice cover, and therefore remains dry . A cluster of fraz i l particles. Flooded lee Fracture Frazil Frazil Slush Freezeup Date Freezeup Period Frozen Frazil Slush Grounded lee Hinge Crack 33RD2-007r lee which has been flooded by melt water or river water and is heavily loaded by water and wet snow. Any break or rupture formed in an ice cover or floe due to deformation. Fine spicules. plates or disco ids of ice sus- pended in water. In rivers and lakes it is formed in supercooled. turbulent waters. An agglomerate of loosely packed frazil which floats or accumulates under the ice cover. The date on which the water body was first observed to be completely frozen over. Period of initial formation of an ice cover. Accumulation of slush that has completely frozen solid . Ice which has run aground. Crack caused by significant changes in water level . Hummocked lee lee Bridge lee Cover Ice Floe Ice Free Ice Jam Ice Ledge Ice Push Ice Run lee Sheet lee piled haphazardly, one piec~ over another to form an uneven surface. A continuous ice cover of limited size extending from shore to shore like a bridge. A £ignificant expanse of ice of any form on the surface of a body of water. Free floating piece of ice. No floating ice present. A stationary accumulation of fragmented ice or frazil. Narrow fringe of ice that remains along the shores of a river after breakup. Compression of an ice cover, particularly at the front of a moving section of ice cover. Flow of ice in a river. An ice run may be light or heavy, and may consist of frazil, anchor, slush or sheet ice. A smooth, continuous ice cover. Ice Shove In situ Breakup Lateral Ice Lead Pressure Ridge Shear Crack Shearing Shore Ice Shore Lead Shore Depression 33RD2-007r On-shore ice push caused by a compression in an unconsolidated ice cover that transmits internal forces laterally towards the banks . Melting in place. See Border Ice. Long, narrow opening in the ice. A line or wall of broken ice forced up by pressure. Crack formed by movement parallel to the surface of the crack. Motion of an ice cover due to horizontal shear stresses. See Border Ice. A water opening along the shore. Depression in the ice cover along the shore often caused by a change in wate r level . Snow Ice Snow Slush Stranded Ice Unconsolidated (Ice Cover) Ice that forms when snow slush freezes on an ice cover. It appears white due to the presence of air bubbles. Snow which is saturated with water on ice surfaces, or as a viscous mass floating in water after a heavy snowfall. Ice that has been floating and has been deposited on the sho re by a lowering of the water level. Loose mass o f float ing ice. * Source: U.S . Army Corps of Engineers, 19 82 l3RD2-nn7r