UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effects of river ice on bank morphology and riparian vegetation : Peace River, Alberta Uunila, Lars Sakari 1999

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1999-0448.pdf [ 26.62MB ]
Metadata
JSON: 831-1.0089091.json
JSON-LD: 831-1.0089091-ld.json
RDF/XML (Pretty): 831-1.0089091-rdf.xml
RDF/JSON: 831-1.0089091-rdf.json
Turtle: 831-1.0089091-turtle.txt
N-Triples: 831-1.0089091-rdf-ntriples.txt
Original Record: 831-1.0089091-source.json
Full Text
831-1.0089091-fulltext.txt
Citation
831-1.0089091.ris

Full Text

E F F E C T S OF R I V E R ICE O N B A N K M O R P H O L O G Y A N D R I P A R I A N V E G E T A T I O N : P E A C E R I V E R , A L B E R T A by Lars Sakari Uuni la B . S c , The University of British Columbia, 1993 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF M A S T E R OF S C I E N C E in T H E F A C U L T Y OF G R A D U A T E STUDIES (Department of Geography) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y OF BRIT ISH C O L U M B I A August 1999 © Lars Sakari Uuni la, 1999 In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of Bri t ish Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Geography The University of Brit ish Columbia Vancouver, Canada Date /0-4uHi*tsr-/99? DE-6 (2/88) A B S T R A C T This study investigated the effects of river ice and related flooding on the bank morphology and riparian vegetation along 655 km of the Peace River from Clayhurst, Brit ish Columbia to Fort Vermil ion, Alberta. Regulation of this river since 1968 for the generation of hydroelectric power has caused a change in the flood regime and has resulted in variable channel adjustments and patterns of riparian succession. A major objective of this study was to determine how bank morphology and riparian vegetation might be influenced by disturbances created by ice jams and ice runs. Historical observations of ice jams have been focussed near the Town of Peace River, the largest population centre along the river. In order to examine frequency and magnitude of ice jam events for the remaining, sparsely populated reaches, a river-length reconnaissance was carried out. Both the direct physical effects of ice and the indirect effects of ice jam flooding on the channel margin were investigated. Bank morphology, sedimentary features, and vegetation patterns were surveyed on 270 transects of the riverbank at locations chosen to encompass as wide a range of hydrogeomorphic conditions as possible. During each transect, the extent o f vegetation damage in the form of broken stems and scars that could be attributed to ice jams, ice runs, or related flooding was measured. Approximately 300 wedges and cross-sectional samples of ice-damaged woody vegetation were collected and analyzed to determine the date of past ice disturbances. Analyses o f the geomorphic and botanical evidence collected in this study suggest that the frequency and magnitude of high stages due to ice jams peak near the central part of the study reach, between the Town of Peace River and Carcajou. This reach is highly confined and sinuous with a large number of mid-channel islands, confirming that channel morphology is a major factor determining ice jam location. Ice shove levels within the study reach peak at an elevation of 11 m above the mean summer stage. Based on the tree scar record along the river, ice jam flooding has occurred in 27% to 86% of the post-regulation years. Frequent events occur along a confined sinuous channel with a significant gradient reduction near Notikewin River. Ice jams occur less frequently in the upper Peace River, where the formation of an ice cover is intermittent. Ice scour and ice jam floods appear to maintain the active shelf, a characteristic bench found between the pre-regulation floodplain and the current, post-regulation channel bed. Although both erosional and depositional processes operate on this bench, evidence suggests the sedimentation during ice jam floods is most important. Flooding and sedimentation from ice jams also affect secondary channels that have rarely been inundated during open-water floods following regulation. Scour and push features are irifrequently scattered along the study reach and are short-lived due to the inherent erodibility of the sediments. The most conspicuous effects of ice jams and ice scour are found in the riparian vegetation. Numerous bent, broken and scarred stems caused by ice suggests that ice disturbance is frequent along the lower seres, particularly in high exposure areas such as island heads and cut banks. However, for the same elevations above the river, vegetation tends to be older and less disturbed in the lower reaches than the upper and middle reaches, due to differences in channel morphology that cause ice jams to halt along wide bar and bench surfaces. iv T A B L E O F C O N T E N T S A B S T R A C T i i T A B L E OF C O N T E N T S iv L IST OF T A B L E S v i LIST OF F I G U R E S v i i A C K N O W L E D G M E N T S x i i 1.0 I N T R O D U C T I O N . . . 1 1.1 BACKGROUND 1 1.2 FREEZE-UP 7 1.3 BREAK-UP 10 1.4 ICE JAMS 12 1.5 PREVIOUS STUDIES 14 1.5.1 Ice Research in Lake and Marine Environments 15 1.5.2 Ice Research on Rivers 16 1.6 RESEARCH OBJECTIVES 26 2.0 S T U D Y A R E A 29 2.1 LOCATION 29 2.2 PHYSIOGRAPHY 32 2.3 SURFICIAL AND BEDROCK GEOLOGY 34 2.3.1 Surficial Geology 34 2.3.2 Bedrock Geology 35 2.4 CLIMATE 36 2.5 VEGETATION 39 2.6 HYDROLOGY 40 2.6.1 Setting 40 2.6.2 Tributaries 41 2.6.3 F low Regulation.... 42 2.7 PEACE RIVER ICE 51 2.7.1 Freeze-up 51 2.7.2 Break-up and Ice Jams 56 2.8 CHANNEL MORPHOLOGY 62 2.8.1 Upper Peace River (Reaches l a and lb) 64 2.8.2 Middle Peace River (Reaches 2a and 2b) 68 2.8.3 Lower Peace River (Reach 3a) 69 3.0 M E T H O D S 71 3.1 RESEARCH DESIGN 71 3.2 L A B INVESTIGATION 72 3.3 FIELD OBSERVATIONS 74 3.3.1 Vegetation and Bank Surveys 78 3.4 ANALYSIS OF DATA 96 3.4.1 Ice Scar Data 96 3.4.2 Historical Data 97 3.4.3 Vegetation Communities 99 3.4.4 Bank Morphology 99 4.0 R E S U L T S 100 4.1 ICE JAM LOCATION, MAGNITUDE AND FREQUENCY FROM THE SCAR RECORD 100 4.1.1 Location and Magnitude of Ice Jams 100 4.1.2 Frequency of Ice Jams 110 4.2 EFFECTS OF RIVER ICE ON B A N K MORPHOLOGY 121 4.2.1 Erosional Features 122 4.2.2 Depositional Features 133 4.2.3 Bank Morphology 142 4.3 EFFECTS OF RIVER ICE ON RIPARIAN VEGETATION 146 5.0 C O N C L U S I O N S 156 5.1 LOCATION, MAGNITUDE AND FREQUENCY OF ICE JAMS 156 5.2 CHANNEL MORPHOLOGY 158 5.3 RIPARIAN VEGETATION 160 5.4 FUTURE STUDIES 161 6.0 R E F E R E N C E S 162 A P P E N D I X A . 1:50,000 S C A L E N.T.S. M A P S OF T H E P E A C E R I V E R 178 A P P E N D L X B. A D I S T A N C E D A T U M F O R T H E P E A C E RTVER 179 A P P E N D I X C. A H I S T O R Y OF ICE B R E A K - U P O N T H E P E A C E RTVER 181 A P P E N D I X D. P E A C E RTVER S T U D Y R E A C H M A P S 196 A P P E N D I X E. S E L E C T E D B A N K P R O F I L E S 204 A P P E N D I X F. A C C E S S POINTS A L O N G T H E P E A C E R I V E R 213 VI L I S T O F T A B L E S Table 2.1. Mean monthly temperature and precipitation at selected climate stations within the Peace River Basin for the period 1961 to 1990 (from Atmospheric Environment Service, 1993) 37 Table 2.2. Active and inactive Water Survey of Canada hydrometric stations on the Peace River (from Environment Canada, 1998) 43 Table 2.3. Summary streamflow statistics for selected Water Survey of Canada stations on the Peace River 45 Table 2.4. Annual maximum daily discharges for selected W S C stations on the Peace River '. 48 Table 2.5. Freeze-up, ice cover duration, and break-up statistics for hydrometric stations along the Peace River (from Prowse et al., 1996) 53 Table 2.6. Study sub-reaches in the Upper, Middle and Lower Peace River 65 Table 3.1. Study timeline 73 Table 4.1. Elevations of ice-related tree scars and trim lines on the Peace River between the BC/Alber ta border and Fort Vermil ion 102 Table 4.2. Dates and elevations of the highest annual tree scars for each sub-reach I l l Table 4.3. Dates and elevations of the highest annual tree scars for each principal reach ...113 vi i L I S T O F F I G U R E S Figure 1.1. General sequence of events during winter and spring break-up periods (from Wil l iams and MacKay, 1973) 6 Figure 1.2. V i e w of a mid-winter consolidated ice cover characteristic of the upper reaches of the Peace River 9 Figure 1.3. Close-up view of a consolidated ice cover near the head of an island on the upper Peace River 9 Figure 2.1. Location, relief, physiographic regions, and primary study reaches of the Peace River basin. Distances were measured in kilometres along channel centre-line from the head of Slave River 30 Figure 2.2. Mean monthly temperature and precipitation at selected climate stations within the Peace River Basin for the period 1961 to 1990 (from Atmospheric Environment Service, 1993) 37 Figure 2.3. Mean monthly flows at selected Water Survey of Canada stations along the Peace River pre- and post-regulation (from Environment Canada, 1998). See Table 2.3 for periods of record 46 Figure 2.4. Annual maximum daily discharges for selected W S C stations on the Peace River. Note that Wil l iston Reservoir was fi l led from December 1967 to May 1972, and that the spillway was operated at the W . A . C . Bennett Dam in 1972 and 1996 49 Figure 2.5. Estimates of 2-, 5-, 10-, 25-, and 50-year return period floods on the Peace River. Stations represented by the data in increasing drainage area order are: Hudson Hope, Taylor, Dunvegan, T P R , Fort Vermi l ion, and Peace Point. See Table 2.3 for periods of record 50 Figure 2.6. Peak freeze-up and break-up stages at the Town of Peace River (TPR) 55 Figure 2.7. Ice cover progression for 21 winters (from 1973/74 to 1993/94) on the Peace River (from Andres and Van Der Vinne, 1994). This plot of annual ice observations serves to indicate the high variability in timing, duration and extent of the ice cover from year to year since regulation 57 Figure 2.8. Annual peak water level versus discharge under break-up conditions at three locations on the Peace River. Break-up data are open circles, while open-water data are solid circles 61 Figure 2.9. General valley and channel morphology of the principal study reaches 63 Figure 2.10. Mean channel width and sinuosity of the study sub-reaches 66 Figure 2.11. Channel gradients along the Peace River based on 1:50,000 N T S maps and Andres (1996). The locations of the principal study reaches and points of interest are indicated 67 Figure 3.1. Mean daily stages at selected W S C gauging stations on the Peace River in 1995 and 1996 (from Environment Canada, 1997). The dates of field observations are indicated 76 Figure 3.2. Mean daily discharge at selected W S C gauging stations on the Peace River in 1995 and 1996 (from Environment Canada, 1997). The dates of field observations are indicated 77 Figure 3.3. V iew of the lower vegetation limit used as the field datum during the bank and vegetation surveys 81 Figure 3.4. V iew of the field datum and vegetation galleries on the middle Peace River 81 Figure 3.5. Example of the field datum and trim line on the upper Peace River 82 Figure 3.6. Example of the field datum along a straight reach on the upper Peace River 82 Figure 3.7. Mean daily water level records at selected W S C stations and elevations of field datum above the river level during the dates of the field investigation 83 Figure 3.8. Ice damaged balsam poplar and idealized sketch of ice block abrasion of a tree during an ice jam and ice drive (after Smith and Reynolds, 1983) 86 Figure 3.9. A ) Example of an ice damaged balsam poplar after a wedge sample was cut across the scar. B) Close-up view of a white spruce wedge sample clearly showing the year when the bark of the tree was abraded by ice 87 Figure 3.10 Backwater flooding of an island during a break-up ice jam near Sunny Valley....89 Figure 3.11 Damage to shrubs resulting from ice thrust during break-up. The photo was taken the morning after break-up took place near T P R 89 Figure 3.12. Example of a mature balsam poplar tree that sprouted four shoots after being shoved during an ice jam flood in 1974. A section was cut on the left for dating purposes 95 Figure 4.1. Extensive shrub damage and removal by the 1996 spring ice jam and ice run near Notikewin River. Parallel linear grooves carved by moving ice floes are clearly visible. Note a two metre scale in the foreground 101 Figure 4.2. Ice-damaged alder community near Fort Vermil ion. Spring 1996 break-up caused extensive damage along the bank and minor sedimentation on the floodplain 101 ix Figure 4.3. Locations of trees with ice scars greater than 6 m above field datum. Scar elevations above field datum are classified into four groups: 1) 6.0 - 6.9 m, 2) 7.0 - 7.9 m, 3) 8.0 - 8.9 m, and 4) greater than or equal to 9 m. 150 scars below 6 m are not shown because of map scale (see Appendix D). See Section 3.3.1 for discussion of field datum. Distances indicated are in kilometres from Slave River 104 Figure 4.4. Scar elevations along Peace River: a) ice scar envelope of all ice scars on mature trees; b) 20 km running means (calculated every 5 km) of all ice scars on mature trees; and, c) cumulative percentage deviations from the overall mean ice scar elevation (5.0 m) (see text for discussion). Elevations are referenced to field datum (see Section 3.3.1) 107 Figure 4.5. Tr im line elevations along Peace River: a) envelope of trim line elevations; b) 20 km running means (calculated every 5 km) of all trim line elevations; and, c) cumulative percentage deviations from the overall mean ice scar elevation (5.0 m) (see text for discussion). Elevations are referenced to field datum (see Section 3.3.1) 108 Figure 4.6. Highest tree scars dated for each year within each principal study reach. Reach 2a is not included because of small sample size. Data for Fort Vermi l ion were extracted from Gerard (1979). Note that elevations of Gerard's data are referenced to the 2-year summer flood level, and may not be directly comparable to the field datum used in this study (see text for discussion). Arrows indicate years of major ice jams at the T P R 114 Figure 4.7. Estimated ice jam stages for selected return periods for four of the five principal study reaches and the TPR. Reach 2a was not assessed due to the small number of tree scars dated in that reach. Return period estimates for ice jams at T P R are based in historical records. Study reach estimates were based on all events occurring within the post-regulation period (see Table 4.3 and Figure 4.6). Error bars represent the 5% and 95% confidence limits 115 Figure 4.8. Frequency distributions of break-up and open-water stages (pre- and post-regulation) at the TPR. Best-fit lines were fitted by eye. See text for discussion 117 Figure 4.9. Frequency distributions of pre-regulation ice jam and open-water stages at Fort Vermi l ion (from Gerard and Karpuk, 1979) 119 Figure 4.10. Locations of ice-pushed ridges, boulder/cobble pavements, ice-scoured bars or ice-scoured banks. Features indicated at a scale of 1 m or greater. Distances are in kilometres from the Slave River 123 Figure 4.11. Ice-scoured bar at the head of an island on the upper Peace River. Note the ice-pushed ridge in the background 124 Figure 4.12. Ice-pushed gravel ridge near the head of an island in the upper Peace River 124 Figure 4.13. Ice-pushed boulder and scour trough 126 Figure 4.14. Photos and contour map of large ice-pushed cobble-gravel ridge on the left bank between Dunvegan and Saddle (Burnt) River. Ice grounding is common at this location of a tight meander. Information from Chalmers (1995) suggests this ice-pushed ridge was likely formed during a freeze-up jam in 1993 127 Figure 4.15. a) A i r photo of a bend near 30-Mile Wel l , characteristic of the middle Peace River. Note that the island is attached to the shore and has a long and wide bar surface at its head; b) ice jam flooding of the secondary channel at the same location. Since regulation, ice jams have been the major process causing flooding of these environments. Note the toe of the jam was located approximately 20 km downstream 129 Figure 4.16. Ice-pushed alluvium and damaged alder stand near Carcajou. This photo was taken two months after break-up, when ice was shoved up this bank 130 Figure 4.17. V iew of the right bank at B ig Bend where 5 m of bank erosion occurred during the spring break-up in 1994. A t this site a remote hunting camp is located. Note the oi l drum, on the upper right, that was pressed into the bank under the force of ice 130 Figure 4.18. Close-up view of ice remaining on bank following break-up. Note the ice that has protected the bank from the shearing action of ice floes also acts to remove a thin layer of sediment as is slides down the bank 132 Figure 4.19. V iew of sediment frozen to ice being plucked from the bank after a stage reduction near the Smoky River confluence 132 Figure 4.20. Ice-pushed gravel ridge on a large bar near the Fairview water intake on sub-reach l b 134 Figure 4.21. Sequence of cobble ice-push deposits along the left bank near Hungry Bend on sub-reach 3a.4. Arrows indicate the locations of four deposits visible in the photo 134 Figure 4.22. Gravel ice-push deposit jutting into the channel on the lower Peace River 136 Figure 4.23. A n ice-pushed cobble hook rarely seen during the river reconnaissance 137 Figure 4.24. Typical boulder pavement of the lower Peace River 137 Figure 4.25. Isolated patch of cobble-gravel deposited during the melting of stranded ice blocks on a bar on the upper Peace River 139 Figure 4.26. Post break-up conditions at the mouth of the Whitemud River. Note the thick layer of mud within the wi l low shrubs 140 Figure 4.27 Example of sedimentation at the mouth of the Ksituan River following break-up. A t this location, break-up of the Ksituan River l ikely preceded that of the Peace River causing sediment from the Ksituan River to deposit on top of ice or within an area confined by ice 140 Figure 4.28. Idealized channel cross sections of the Peace River during mid-winter, dynamic break-up, ice jam flooding, and summer periods. Principal features include the active channel shelf that is inundated and over-run by water and ice, shoved and scarred mature trees and damaged riparian communities. The sketches are not to scale. The sketch for the dynamic break-up period was adapted from Thorson and Guthrie (1984) 145 Figure 4.29. Secondary channel vegetated with wi l lows and horsetails 147 Figure 4.30. Wide secondary channel approximately 1.5 m above the summer stage vegetated with primary successional wi l lows 147 Figure 4.31. 20 km running means (calculated every 5 km) of elevations of lower boundaries of vegetation communities along the Peace River: a) Class 5 - trees> 15 m tall; b) Class 4 - trees 5.0-14.9 m tal l ; and, c) Class 3 - shrubs 2.0-4.9 m tall and Class 2 shrubs - 1.0-1.9 m tall. Vegetation younger than Class 2 was highly discontinuous and therefore not plotted 149 Figure 4.32. Max imum vegetation ages versus elevation above field datum for selected transects within each principal study reach 153 Figure 4.33. Estimated ice shove frequency for selected transects on the Peace River 154 A C K N O W L E D G M E N T S This multi-year endeavor would not have been possible without the generous support of a number of persons. Foremost thanks go my supervisor, Dr. Michael Church, for his patience, support, guidance, and helpful advice. Sincere thanks go to the other members of my supervisory committee: Dr. Olav Slaymaker and Mrs. Margaret North. Thanks go to Dr. J . Ross Mackay for providing literature and sharing his knowledge of the North. Thanks go to Stephen Herold and Lesley Kalmakoff for their help in piloting the boat through some intimidating waters, bush clearing, surveying, and keeping spirits high despite the Peace River 's daily serving of rain and mud. Thanks to Dr. Jiongxin X u of the Chinese Academy of Sciences for his field assistance in the summer of 1996, and Anna and James Stokes of High Prairie for providing the field crew several home-cooked meals. Thanks go to Jack Chalmers, of Fairview, and Dennis Sivertson, formerly of Carcajou, for sharing their local knowledge of the Peace River. Many thanks go to Gordon Fonstad and Wi l l i Granson at Alberta Environmental Protection for providing historical information and the opportunity to fly over Peace River with Peace A i r Ltd. during break-up in 1996. Thanks go to Dr. Faye Hicks at the University of Alberta for loaning out an all but forgotten set of tree samples collected along the Peace River by the late Dr. R. Larry Gerard. Thanks also to David Andres at Tr i l l ium Engineering and Hydrographies Ltd. for providing photographs of the Peace River. Thanks go out to the Water Survey of Canada employees at Fort St. John, Peace River and Fort Smith for providing unpublished f low data. A t U B C , I thank Arnold Moy for his computer advice and support. Thanks also go to Darren Ham, Emmanuelle Arnauld, and Kel ley Hishon for their assistance. Finally, I would like to thank my wife, Pollyanna, and my parents for their love and support throughout the course of this project. 1.0 INTRODUCTION 1.1 B A C K G R O U N D The annual formation and break-up of river ice, along with ice jams and flooding that follow, has long been recognized as having the potential to affect bank morphology, channel substrate, and riparian vegetation along large northern rivers. In the Hudson Bay region, for example, Robson (1752) was cited by Newbury (1968) as follows: "When the snow melts, the Indians to the southward of York-fort who are near or within forty miles of the sea, keep their canoes always in readiness, that they may escape the torrent that pours down from the inland country, overflowing the adjacent plains, and bearing down on the trees...and it is easy in the summer to discern which rivers are subject to them, from the deep hollows which the ice constantly plows up on both sides." Similarly on the Mackenzie River, Kindle (1920) wrote: "There is scarcely a mile of river that does not show traces of the destructive work done by these great blocks as they travel seaward. In some places the spruce trees have been snapped off like straws; in others the clay banks have the appearance of having been ploughed by titanic plows." Modem studies in North America, Europe and Russia include work by Brooks (1993, 1996), Collinson (1970, 1971); Danilov (1972), Day and Anderson (1976), Dionne (1974, 1976), Gatto (1982), Gerard (1979), Hamelin (1972), Mackay and MacKay (1977), MacKay et al. (1974), Marusenko (1956), McLean and Anderson (1980), Newbury (1968), Parker and Josza (1973), Prowse (1996), Prowse and Gridley (1993), Scrimgeour and Prowse (1993), Scrimgeour et al. (1994), Smith (1979,1980), and Wentworth (1932a, 1932b). During periods of freeze-up and break-up on northern rivers, ice jams, especially during dynamic break-up events, impede river flows and can cause rapid increases in water levels upstream. Flooding may inundate side-channels and valley flats to levels substantially higher than those reached by open water floods (Bray and Kellerhals, 1979; Church, 1988; Gerard, 1990; Gerard and Karpuk, 1979; N R C C , 1989; Prowse, 1994; Walker and McC loy , 1969; Yaremko, 1968). Ice jam flood levels often reach a physical upper limit that is imposed by the morphology of the river ( N R C C , 1989). For example, on the Athabasca River in 1875, ice jam flooding was reported to have resulted in a 57 ft (17.4 m) stage increase in less than one hour (Moberly and Cameron, 1929). Upon the release of an ice jam, which may vary in rate from gradual to catastrophic, a wave of icy debris-laden floodwater progresses downstream as an ice drive or ice run. Progression continues downstream until it melts or is halted by another ice jam. Ice jam floods, from both the impounded water upstream and the ice run downstream, may cause damage to water intakes, bridges, pipeline crossings, telephone lines, f lood protection works, water level recorders, and other structures in the vicinity of the river (Andres, 1981; Doyle, 1982, 1988; Henoch, 1973; Nibler, 1980). Ice jam flooding poses a risk to economic development and loss of land, and it is a risk to persons l iving along floodplains. In addition, ice jams have historically interfered with transport and river navigation and have been a constraint in managing hydroelectric operations (cf. Beltaos, 1995). Such operations must balance flow releases required for power production with the potential of destabilizing the ice-cover and initiating damaging ice jams. Extreme water and ice levels combined with high velocities at break-up are responsible for periods of increased sediment transport. Rapid break-up fronts may increase suspended sediment concentrations by an order of magnitude above pre-break-up conditions (Prowse, 1992). Estimates of shear stresses near break-up fronts suggest bed-load up to 0.2 m diameter may be moved (Beltaos, 1993). In the wake of an ice jam flood and ice run, suspended, ice-rafted and shoved sediment may be deposited along channel margins, the bed and banks may be scoured, and riparian vegetation may be damaged or destroyed (Cameron and Lambert, 1971; Gerard, 1975, Koutaniemi, 1984; Lopatin, 1871; Martinson, 1980; McPherson, 1966; Outhet, 1974a; Smith, 1979). This process may lead to distinct erosional and depositional features along northern rivers (Smith, 1979, 1980; Marusenko, 1956). Furthermore, during ice jams, flows may be re-directed through side-channels and other parts of the main channel (Newbury, 1968). This, in turn, can cause increases in erosion rates by up to a reported 200% (Martinson, 1980). F low redirection to the inside of the meander may result in unusually high erosion rates in an otherwise depositional environment (Prowse and Gridley, 1993). Changes in the plan morphology of the river may result i f the channel shifts, avulses, or cuts off at meanders (Doyle, 1988, MacKay et al., 1974; Prowse, pers. comm., 1997; Wil l iams, 1973). Deep scour pools may form near ice jam locations (Baldrige and Trihey, 1996; Mercer and Cooper, 1977) often near locations of decreasing channel gradient (Mel in, 1954) or at rapids (Newbury, 1968). Although the effects of river ice are greatest on large alluvial rivers, bedrock lined channels may be affected by the scouring, grinding and polishing action of ice and sediment embedded in ice (Danilov, 1972, Dionne, 1974,1985; Leffingwell, 1919; Russell , 1890; Todd, 1892; Tolmachoff, 1928). These effects are generally superficial and rarely result in gross morphologic change. Ice runs and ice jams both destroy vegetation and promote future growth of vegetation, with the input of organic matter and nutrients. The removal of plant tissue, litter and nutritive particles with prolonged ice jam flooding, and the deposition of mud tend to reduce oxygen levels, temperatures, and solar radiation required for vegetation growth (Illichevsky, 1933; Wisheu and Keddy, 1996). Trees and shrubs scoured by ice may perish as ice-inflicted wounds activate dormant spores of decay fungi (Fil ip et al., 1989). 4 Ice jam sedimentation in northern rivers is a primary mechanism for supplying organic material and nutrients along channel margins (Prowse, 1996). In riparian environments, ice jam flooding has been cited as a major allogenic factor responsible for the maintenance of primary successional species such as Salix (Eggington, 1980; Prowse and Gridley, 1993; Prowse, 1996) which provides rich wildlife habitat (Baldrige and Trihey, 1996; Fi l ip et al., 1989; Strong and Leggat, 1992). The process of ice jam flooding may in fact be essential to the integrity of some lotic systems, such as inland deltas, where elevated lakes and side channels require periodic flooding to maintain certain forms of habitat (e.g., Peace-Athabasca Delta) (Prowse and Gridley, 1993; Prowse and Lalonde, 1996; Prowse et al., 1996). In summary, river environments that experience the annual formation o f ice tend to have riparian vegetation and morphology adapted to and maintained by the ice regime and particularly the periods of ice jamming during freeze-up and break-up (Prowse and Gridley, 1993; Prowse and Conly, 1996). The objective of the present study is to characterize the effects of river ice and especially ice jams on the bank morphology and riparian vegetation of the Peace River in northwestern Alberta. The Peace River is a northward flowing boreal river, and it is the largest tributary in the Mackenzie River Basin. In this environment, the ice-covered regime lasts about seven months of the year (Andres and Van Der Vinne, 1994). Furthermore, ice jams are often responsible for the major annual f lood (McLean and Anderson, 1980), particularly as the Peace River has been regulated since 1968. Since that time, flows have been managed for hydro-power generation, causing a significant reduction, but not a total elimination, of summer floods (Prowse, 1996; Prowse and Lalonde, 1996). The ice-covered regime on a river is defined here as the period from initial ice crystallization to final ice melt, and it is comprised of three main periods: 1) freeze-up, 2) mid-winter, and 3) break-up. Freeze-up and break-up are relatively short periods but usually represent the most significant periods with respect to geomorphic activity and vegetative damage, and wi l l be discussed in greater detail in Sections 1.2 and 1.3. During freeze-up and break-up, unsteady flows and fluctuating water levels result from rapidly changing hydraulic conditions as ice forms and disintegrates. The mid-winter period, on the other hand, tends to be a time of less sediment transport due to low and steady base-flows with lower bed shear stresses, velocities, and diffusivities caused by increases in hydraulic resistance imposed by ice (Prowse, 1996). However, on regulated rivers, fluctuating releases during mid-winter may destabilize the ice cover causing ice shoving, ice drives and mid-winter jams, unsteady flows and localized geomorphic activity. Knowledge of the fundamental processes operating during the ice regime is essential in understanding the effects of river ice and ice jams on morphology and riparian vegetation. The generalized processes of ice freeze-up and break-up are outlined below and are summarized in Figure 1.1. The main features of the ice regime along the study reach on the Peace River are described in Section 2.7. OPEN WATER I freezing conditions slow moving water INITIAL ICE SKIM FIRST ICE COVER thawing, winds freezing STABLE ICE UNSTABLE ICE I continuous freezing snowmelt, rising water levels Q O 5 CL r> 2 cc OQ CD 5 cu FIRST MOVEMENT OF ICE I restrictions rising water levels I FIRST ICE JAMS thawing, water pressure break-up of jam I MASSIVE MOVEMENT OF ICE I restrictions I MASSIVE ICE I repeated break-up and jamming I final melting turbulent water I FRAZIL & ANCHOR ICE ice run progression hanging dams restricted flow rising water levels I •ice r u n — ICE JAM I thawing, water pressure break-up, open water freezing I FRAZIL & ANCHOR ICE ice run stable ice • process continues until sufficient freezing to produce stable cover or sufficient melting to clear river of ice I melting Q O CL cc Hi RIVER COMPLETELY OPEN Figure 1.1. General sequence of events during winter and spring break-up periods (from Wil l iams and MacKay, 1973). 1.2 F R E E Z E - U P The process of freeze-up in late autumn and early winter begins with the cooling of water to a temperature of 0°C, which occurs by a transfer of heat from the water to the ground or air by conduction, evaporation or radiation. Once the heat of fusion from the water is removed, ice begins to form. Initially, border ice freezes at tranquil locations, including side-channels and shallow areas along banks. Meanwhile, near the center of the channel, frazil crystals begin to form. Frazi l crystals are hexagonal disk-shaped particles 0.1 to 5 mm in diameter (Prowse, 1994) and are commonly referred to as the building blocks of river ice (Ashton, 1978). Under turbulent, super-cooled conditions, with water temperatures of -0.01°C (Smith, 1980), frazil crystals grow and become mixed throughout the depth of water. A s frazil grows, its tends to stick to anything protruding into the f low including aquatic vegetation, boulders and large patches of gravel and coarse sand (Prowse, 1994). Frazil that accumulates on the bed of a river is known as anchor ice. Growth of anchor ice at downstream ends of steep turbulent reaches may eventually form ice dams. Under moderately turbulent conditions, frazil floats to the surface and forms slush ice. Slush ice agglomerates into hemispherical frazil pans. A s border ice grows towards the center of the channel and as frazil pans grow radially, a point is reached when ice pans lodge, jam, or bridge across the channel. Bridging of ice pans initiates the upstream progression of the ice cover. Meanwhile, downstream of the point of bridging, freeze-up is delayed, as new ice pans must develop from frazil produced in the open water. The characteristics of the initial ice cover depend on the wind velocity, air temperature, relative humidity, cloud cover, short-wave radiation, long-wave radiation, snowfall, river gradient, discharge, groundwater inflow, and reach geometry (Michel, 1971). Scenarios listed below provide examples of the factors controlling the formation of an ice cover. If the discharge and channel gradient are low, and the weather very cold, the ice cover wi l l progress rapidly by juxtaposition of incoming ice floes. On large rivers, ice cover progression by juxtaposition can proceed in excess of 10 km per day (Prowse, 1994). The ice cover w i l l be one pan thick and have a relatively smooth surface like ice on a calm lake. Juxtaposed ice covers are relatively stable and severe freeze-up jams under these conditions are uncommon on non-regulated rivers. However, on regulated rivers with large discharge fluctuations during freeze-up, a juxtaposed ice cover may be destabilized resulting in ice shoving, jamming and increases in water levels. If discharges are high during freeze-up, as is commonly the case on rivers regulated for hydro-power generation, the accumulating frazil upstream of the initial ice cover may be drawn below the ice surface and wi l l constrict the channel by forming a hanging dam of slush ice. A n ice cover that forms through progressive ice formation and collapse, or "telescoping", is described as consolidated (see Figures 1.2 and 1.3). Since ice covers reduce the conveyance of f low by reducing the cross-sectional area (i.e., ice floats with 90% of its thickness submerged) and increasing the resistance, the water levels increase (Gray and Prowse, 1993; Laszloffy, 1948) over open water conditions. The rise in stage continues until hydraulic conditions at the leading edge of the cover are modified enough to allow it to stabilize and proceed upstream. Assuming the ice cover and bed have equivalent roughness, water levels at freeze-up increase by about 30% above the mean depth of open water (Gray and Prowse, 1993; Prowse, 1994). The increase in water level, however, is substantially greater for highly consolidated (i.e., rough) ice covers and ice jams. A s water is taken into temporary storage with the stage increase upstream of an ice cover, low f low conditions may be Reach: 1b. 1 Site: near Fairview water intake Location: km 928.5 left bank Photo: 53.21 Date: Feb 6, 1996 Figure 1.2. V iew of a mid-winter consolidated ice cover characteristic of the upper reaches of the Peace River. Reach: 1b. 1 Site: near Fairview water intake Location: km 927.9 island Photo: 54.8 Date: Feb 6, 1996 Figure 1.3. Close-up view a of consolidated ice cover near the head of an island on the upper Peace River. 10 experienced downstream. The ice cover eventually freezes solid with a slightly translucent colour and fine-grained texture. If the ice cover is thin, freezing w i l l progress downwards as heat is lost to the atmosphere (Ashton, 1979), forming coarse-grained transparent ice. If snow accumulates on the ice cover, the ice may be submerged and downward growth wi l l be retarded or stopped. Ultimately, the weather conditions and thermal regime of the river throughout the winter control the thickening of the ice cover. On regulated rivers, such as the Peace River, warmer water temperatures from hypolimnetic reservoir releases significantly l imit the formation and development o f ice (Ashton, 1979). 1.3 B R E A K - U P Break-up is generally divided into three main phases: 1) pre-break-up, 2) drive, and 3) wash (Michel, 1971). Depending on the hydrometeorological conditions, break-up may be thermal or mechanical. During thermal break-ups, the pre-break-up phase dominates, whereas during a mechanical break-up, the drive is the primary phase responsible for the deterioration of the ice cover. The pre-break-up phase begins with fair weather and warm temperatures in spring. Immediately preceding this phase, discharges on non-regulated rivers would be the lowest of the year. On regulated rivers with uniform discharge, the stage of the river may be near the lowest of the winter, since steady flows would have decreased the hydraulic friction o f the ice cover and reduced the water level (Michel, 1971). The first stage in the pre-break-up phase is the thermal deterioration o f the ice cover as air and water temperatures increase the ice temperature to the 11 melting point. Initial melting occurs at ice crystal boundaries causing "candling", which dramatically reduces the ice strength but has little influence on the ice thickness. Candled ice is fine grained and opaque and resists melt by solar radiation due to its high albedo. The proportion of clear and white ice and the amount of water on the ice surface control the rate of melting by solar radiation. Meanwhile, snowmelt, rain and possibly reservoir releases increase the discharge and cause ice along channel margins to flood. This tends to promote the formation of shore leads and transverse leads, especially in shallow areas. Under increasing discharge and stage, the central portion of the ice wi l l float but not break. A t this point in a thermal break-up, the majority of the ice cover wi l l have deteriorated. The second, and most spectacular, phase of spring break-up is the drive - the main process operating during a mechanical or dynamic break-up. Farley Mowat (1967) describes it in the following passage: "Suddenly the murmur of the river flowing beneath its winter ice seemed to swell, to become resonant, then in an instant it was transformed into a heavy-throated roar. The cabin shuddered and the tin plates on the table slid and rattled as i f dancing to the erratic rhythm of an earthquake. I caught a terrifying glimpse of an immense cake of ice, at least ten feet thick, rearing out of the river not twenty paces from my door. It stood briefly on end, then toppled forward and as it fell a grey geyser of tormented water flew high above the shifting ice. The river, so long contained, was surging up between the shattered floes. The sound of the break-up moved downstream like the roll of a giant drum. A s it passed, a violent cacophony came into being as the great cakes shattered and moved ponderously down toward the still frozen lake. The air above the battleground was fi l led with a fine dust of ice crystals. Floes the size of buildings were ground out of existence in mere moments to be instantly replaced by others which drove the stubborn bay ice slowly backward." During the drive, ice remaining after the pre-break-up phase is mechanically destroyed by fracturing under increasing ice and hydraulic pressures. Abrupt warming and rapid snowmelt, a rain on snow event, or releases of water on a regulated river often initiate the ice drive or ice run. 12 Ice moves downstream in a surge until it encounters resistant ice or channel morphology (e.g., wider channel) that absorbs the surge. During an ice run, high velocities resulting from locally steep hydraulic gradients may entrain sediment larger than the competency of open water flows. The wash phase is the final stage o f spring break-up. During this period, ice concentrations are low resulting in stage reductions. Max imum discharge is commonly observed at this time (Michel, 1971) as headwater f low contributions peak. Fol lowing a mechanical break-up, a sudden and rapid reduction in stage wi l l often result in a heavy accumulation of ice along the banks. These ice accumulations define surfaces of shearing during break-up and tend to protect the bank from geomorphic activity where they are thick. Thermal break-ups on the other hand, tend to leave only occasional ice remnants on the shore (MacKay and Mackay, 1973). 1.4 I C E J A M S A stalling ice run becomes an ice jam which, by definition, is " a forced accumulation of ice fragments in a waterway" (Beltaos et al., 1990). A n ice jam may partially or completely block the cross-sectional area of a channel resulting in increased water levels upstream. Gray and Prowse (1993) classified ice jams into two categories: 1) hydraulically controlled, and 2) mechanically controlled. Hydraulically controlled jams, common during freeze-up, are characterized by a thickness of ice near the ice-water transition zone dependent upon the f low velocity (Michel , 1971). If velocities increase, ice jam thickness increases as ice floes submerge and deposit beneath the ice cover. Once the internal resistance of the ice cover can withstand increasing forces, the ice cover lengthens upstream. 13 Mechanically controlled jams, common during break-up, are composed of fragmented ice floes that are shoved and compacted into a thickness and sufficient internal strength to transmit the forces that are exerted on it to the banks or any obstruction (cf. Beltaos, 1995). A stable jam resists the drag force of the f low underneath and the down river component o f its own weight. Ice thickness is greatest at the downstream end (i.e., toe of jam) causing decreases in depth and abnormally high hydraulic gradients (Prowse, 1994). Under building hydraulic pressures, a sudden dynamic collapse of the jam wi l l occur. This releases a high velocity surge and initiates another ice run with increased water levels and velocities (Doyle, 1988). The process repeats sequentially until the whole river is free o f ice. If hydraulic pressures remain low, ice jams may melt in place. The severity of ice jamming is usually related to break-up severity; dynamic or mechanical break-ups produce the largest ice jam floods and thermal break-ups the most ineffective jams (Prowse et al. , 1996). In general, the severity of ice jams is greater when f low is in the direction of break-up progression (e.g., Liard, Mackenzie, Nelson, Peace, and Yukon Rivers). In Canada, this occurs in two circumstances: rivers that f low northward and rivers with headwaters that experience chinooks (Gerard, 1975; Smith, 1980). Although ice jams may form along any stretch of river based on the weather and streamflow, ice jams tend to recur from year to year at similar locations determined by channel morphology. Prime locations for ice jams tend to occur at locations of f low convergence. These locations include sharp bends (Mackay, 1958), islands, bars, shoals, rock ledges, bouldery reaches and near the mouths of tributary streams, especially where tributary fans have prograded into the main 14 channel. Ice jams also form where f low velocities decrease. Marked reductions in f low velocities occur at sites of decreasing channel gradient, increasing channel width, increasing channel depth or increasing sinuosity (MacKay and Mackay, 1973). 1.5 P R E V I O U S S T U D I E S Despite the wide range of effects of river ice on channel margins, there has been a tendency for researchers to avoid its study because of difficulties in observing and measuring the processes in the field. Dangerous conditions and the highly variable nature of river ice, particularly ice jams, have largely precluded quantitative measurements. Cold temperatures, remote locations, poor sampling and measurement methods (i.e., streamflow gauging) and formidable costs, in return for limited information, have slowed the progress of research (Church, 1988; Gatto, 1993; Gerard, 1975, Lawson et al. , 1986). Some have believed the ice season warranted less attention, since discharge and sediment transport are low during winter (Lawson et al., 1986). Likewise, few biological studies have been made along rivers that experience the intense activity associated with ice jams (Scrimgeour et al., 1994). This trend seems to be changing, however, as attention to northern river ecology is increasing due to expanding development to the north. In spite of the research difficulties, several workers have made significant contributions to the study of the geomorphic effects of river ice. A review of the salient contributions to drift or floating ice research from around the world is presented below. Since much of the early scientific research was focussed on the protective and destructive effects o f ice on shorelines of cold region lakes, oceans, or estuaries, a brief review is provided in Section 1.5.1. This is followed by a review of selected contributions to river ice research over the past 128 years. It should be noted that some authors use the term "glacie l " to describe the processes involving 15 seasonal drifting ice (i.e. lake, sea, and river ice) and the shoreline environment (Hamelin, 1972). 1.5.1 Ice Research in Lake and Marine Environments One and a half centuries' work, beginning in the early 19 t h century, concentrated on the notion that rock movements near shorelines, and lake ramparts or ridges of sediment onshore, were the consequence of ice expansion due to the volume change on freezing (Adams, 1825; Gilbert, 1908; Hobbs, 1911; Jennings, 1958; Petros, 1822), or ice shifting by winds during break-up (Wood, 1825). Peterson (1965) and Adams (1977) studied ice-push ramparts on lakes in Labrador and found that both thermal expansion and ice rafting were responsible for the movement of ice and sediment up to 18m inland and 6 m above lake levels. Rafting of sediment, however, is rarely seen because the partial open water season is very short. The formation of ramparts due to ice expansion requires: 1) sudden and large temperature changes, 2) fast transmission of temperature changes to the ice, 3) competent ice that can transmit stresses to the shore, and 4) repetition of the process (Hobbs, 1911). The second factor is inhibited when a snow cover insulates the ice. The third is regulated by the size of the lake. Smaller lakes have negligible expansion while larger ones may not have a fully rigid ice cover. Gatto (1982) summarized the effects of ice on reservoir shorelines and stated that ice could erode banks directly by loosening, removing and pushing soil or grinding along the face or toe of the bank. The main factors influencing erosion were: 1) water levels and their fluctuations, 2) ice strength and characteristics, 3) the mobility of the ice cover, 4) the degree of ice attachment to beach, nearshore and bank sediment, 5) the degree to which sediment is frozen, and 6) shore configuration. 16 Similar to lake ice, sea ice is associated with a variety of sediment transport processes. These include: 1) shoreline"scour by floating sea ice, 2) fluvial scour (i.e., eddy generation) triggered by discontinuous sea ice cover near shore, 3) suspended sediment freezing to sea ice, 4) waves washing sediment on top of sea ice, 5) wind blowing sediment on sea ice, and 6) ice freezing to bottom sediments and plucking these sediment as tides rise and ice refloats (Martini, 1981; Reineck, 1976). Boulder ramparts or ridges are the most conspicuous ice-related feature observed along northern coasts and can be up to 9 m high (Hume and Schalk, 1964). They form after large ice floes charge onto the shore under the influence of winds, currents or ice expansion (Ward, 1959). Boulder ridges tend to form on rocky coastlines with sufficient ice and water fluctuations and a distinct break in slope (Rosen, 1979). Ice thrusting also produces small-scale features such as gouges, furrows, striations and pits (Kovacs, 1983). The preservation of these features in the arctic may be high since the 6 to 8 week ice-free period is generally too short to allow wave and current destruction, especially in permafrost environments or i f stranded ice is present (Kovacs, 1983; Owens and McCann, 1970; Ward, 1959). However, along more temperate coasts, ice-push features may be short l ived since they become destroyed by waves and currents during storms (Greene, 1970; Owens, 1976). Shore-fast ice or an ice-foot may act in a protective manner to limit near shore wind, current erosion, and sedimentation (Dionne, 1974). 1.5.2 Ice Research on Rivers On the Yenisey River in Siberia, Lopatin (1871) provided one of the earliest accounts of the significance of river ice jams on northward-flowing subarctic rivers. Striated cobbles and outcrops in regions not subject to ice jams were interpreted as being due to the former wider 17 extension of the process of ice jamming. Russell (1890) provided one of the earliest and exhaustive descriptions of how river ice in Alaska was carried up banks resulting in the smoothing or polishing of rocky shores, especially on the upstream sides of promontories. He stated that ice can be responsible for the longitudinal and lateral transport of large boulders along a channel, as well as scarring and abrasion of trees and the furrowing of sand bars. Ice-related gravel deposits, boulder pavements, and abraded and striated stones were described as being similar to those created by glacial action, yet were less regular and less defined. Todd (1892) provided one of the earliest descriptions of striations on relatively soft bedrock (e.g., limestone) resulting from river ice movement. Although rather speculative, he did provide evidence from a number of locations in South Dakota, Missouri , and Illinois. A t the time, most rock striations had been ascribed to glacial action. A number of compelling reasons for crediting river ice rather than glacial ice for rock striations near rivers included: 1) striations are parallel to the river, 2) marks are relatively short, and 3) marks are in proximity to high energy areas along the river. Although Todd was not able to directly observe the process of break-up, he speculated that stronger ice (i.e., not thermally deteriorated) embedded with siliceous rocks would result in the greatest amount o f abrasion. Tolmachoff (1928) also described the intense abrasion of ice for rivers in arctic Eurasia. He believed river ice was responsible for the creation of rock masses with scratches and grooves similar to roche moutonnees produced by glaciers. In studies comparing the effect of river ice and glacial ice, Wentworth (1928, 1932a) claimed that striated erratics south of the southern limits of glaciation were likely transported by river ice. In previously glaciated areas, the differences between glacial and fluvial striations were held to be 18 mainly of degree (Wentworth, 1932a). On the Mackenzie River, Kindle (1918) observed evidence of the plowing and gouging action of river ice nearly everywhere, both on the bed and banks of the river. On bars, long broad grooves extended for considerable distances. River ice was responsible for plowing bed material up banks and forming piles o f mixed mud and gravel, which often results in elevated levels o f alluvium at the heads of islands. Repeated ice attack on island heads can form a boulder pavement on these landforms of mud. Boulder pavements, analogous to cobblestone roads, are smooth surfaces composed of sub-rounded cobbles and boulders, embedded in a matrix of fine sand, silt or clay, with their upper surfaces nearly even. Kindle (1918) observed these features along both banks for several kilometres, extending up concave banks from below water level to heights in excess of 7.5 m above summer water level. The genesis of pavements was attributed to the grinding and sliding action of ice during break-up, which presses and plucks boulders, locally supplied by glacial t i l l or glaciofluvial deposits, into and out of the river banks. Once formed, boulder pavements may be considered natural rip-rap which offers protection against lateral erosion and limits the degree of meandering (Kindle, 1918). The plucking action of ice was also believed responsible for a number of limestone bank concavities measuring up to 19 m long by 9 m wide by 1.5 m deep. One of the most comprehensive examinations of the erosion, abrasion, transport and deposition of sediment by ice jams was provided by Wentworth (1932b). Based on observations on the Yukon River, he described how ice "invades" banks and islands causing tree damage and bank modification. Principle bank features resulting from ice movements included: 1) fluted sand bars and striae aligned parallel or slightly towards the bank; 19 2) shifted bed material including "furrow lakes" and "bays" (described as small ice push features partially in the water with dimensions up to 4.5 m wide and 14 m long), "transverse bays" (described as ice-push gravel mounds that extend offshore and into the river on a sloping planed surface), and "excess-gravel ridges" (described as small deposits up to 1 m high that form as a result of the pushing of masses of stones in front of the ice); 3) shallow striations, bruises or rub marks on substrate and outcrops; and, 4) ice-planed gravel bars and boulder pavements. Boulder pavements, similar to those described by Kindle (1918), were the most characteristic feature of river ice on the Yukon River, stretching up to 500 m long and 30 m wide (Wentworth, 1932b). The landward margin of the pavement was commonly found at the level of the "ordinary" high water line but below " f lood" stage. Similar to the Mackenzie River, Yukon River pavements were situated on banks with concave-upward profiles with slope angles up to 15 degrees. The upper surface of the boulders dipped slightly upstream, suggesting the tendency to become firmly lodged only when they are originally flat or slightly imbricated (Wentworth, 1932b). Despite the evidence presented by Wentworth (1932b), he concluded that the overall geomorphic effects o f river ice were negligible and would unlikely survive a season of normal fluvial activity. Eardley (1938), also studying the Yukon River, noted that ice jams with recurrence intervals of 5 to 15 years caused extensive flooding and deposition of sediment between 0.01 and 1 m deep. Ice jam flooding was hypothesized as being responsible for the height to which some bar and inside-meander bend deposits were built. Overall, Eardley (1938) believed that river ice processes were negligible in comparison to normal fluvial processes. 20 Marusenko (1956) similarly described the effects of ice on northward flowing Siberian rivers. Ice-run erosion was understood to be dependent on a multitude of spatially and temporally varying factors including discharge, ice thickness, ice strength, water and air temperatures, geology, and bank configuration. Marusenko (1956) described "mostovoi", equivalent to boulder pavements, and "steps" carved out of unconsolidated banks by ice. Preservation of these features however, depends largely on subsequent ice and water stages. According to Mackay (1958), ice jams on the Anderson River, with stage increases of 3 to 5 m, resulted in damaged vegetation, ice erosion (including the formation and maintenance of boulder pavements) and floodplain deposition. Variations in the floodplain elevation (between 4.5 and 7.5 m) and gradient along the river were related to the effects of ice jamming. Similarly, Henoch (1960) recognized that break-up flooding on the Peel River was a major process responsible for sedimentation. High water marks were measured from 5.5 to 10 m above the summer stage. These levels decreased substantially in the delta where water is distributed more effectively among the many channels. Newbury's (1968) study of the Nelson River provided evidence to suggest the geomorphic effects of river ice are a function of lithology, hydrology, and configuration of a river channel. The build-up of ice at rapids was responsible for a downward shift o f the thread of maximum velocity and an increase in shear forces on the bed. A t several locations, this led to bed scour and transport of sediment up to 30 cm in diameter. However, the scour holes were rapidly infi l led during subsequent flows. The transport of bank sediments was found to depend on localized ice movements, especially in the vicinity of ice jams. Newbury (1968) noted that ice shove is responsible for a characteristic channel margin, with vegetation trim lines that fol low the river 21 surface profile, and a concave upward groove that is found between the trim line and open water level. Ice scour was particularly evident around low level, mid-channel islands. Walker (1969, 1975) and Walker and McC loy (1969) suggested that geomorphic work was dependent upon the nature of break-up and particularly the rate of discharge fluctuations. In the arctic deltas of the B low and Colvi l le Rivers, snow and ice act as a protective blanket for much of the year. However, during spring break-up, which lasts one month, most of the effects of river ice reported earlier were observed. Fol lowing ice shoving, ice may remain in place on bars and banks for weeks, causing a delay in vegetation growth. Stranded ice also may act to protect banks from further ice scour or fluvial attack. In the Mackenzie River Delta, G i l l (1971) and Outhet (1974a, 1974b) found that ice-related bank erosion was minimal during the break-up period since a large amount of ice in the channels keeps temperatures near freezing, thus minimizing thermally induced bank erosion common in permafrost environments (cf. Cooper and Hollingshead,1973; Lawson, 1983; Scott, 1978). They also noted that downstream moving ice is largely confined within shear lines to the central part of the channel, and that delta ice is in an advanced state of deterioration at break-up and thereby not effective in erosion. In addition, Lapointe (1984) found that ice-run erosion does not have a major effect on channel shifting in the Mackenzie Delta. Coll inson (1970, 1971) investigated small-scale bed relief generated by river ice on the sand bedded Tana River in Norway. He discovered that bed modifications, which encompassed areas of up to 1000 m 2 , included groove marks associated with the grounding of ice pans, ice-push crescents, ice-push ridges, and ice-pushed boulders. The degree of erosion by grounding ice was a function of the size of the ice block, the velocity on impact, and the rate of deceleration after impact, the last o f which is related to bed slope and grain size (Coll inson, 1971). He reported that scoured grooves were sometimes not rectilinear. Coll inson (1971) believed this was the result of ice blocks scouring the bed as they jostled behind ice jams or were moved by winds. Once grounded, ice may act to dissipate flows and protect the bed, but it also may induce scour by f low separation and eddy generation. Irregular bathymetry with relief up to 1 m, as well as large scour troughs up to 35 m long, 15 m wide, and 0.6 m deep were associated with this process. Although varied and extensive, Col l inson (1971) stated that the overall effects were minimal and localized. Day and Anderson (1976) arrived at a similar conclusion from observations made on the Thomsen River on Baff in Island. Based on the Aldan and Lena Rivers in U S S R , Hamelin (1972) reported observations similar to earlier workers in glaciel research. The main effects of river ice were: 1) the creation of microforms, 2) the protection of banks, 3) the accumulation of sediment, 4) the restructuring o f sediment, and 5) the creation of the "becevnik". "Becevnik" is a Russian term describing the river-side path along which boats are towed from shore. Hamelin (1972) believed the "becevnik" surface was in large part due to erosion and deposition o f sediments by river ice. It is the part of the riverbed, or lower bank that becomes uncovered after high water periods and ranges in sediment texture and slopes (from 4 to 26 degrees). The existence of the becevnik is governed by the channel configuration and on the Aldan and Lena Rivers can extend for several hundred kilometres. On the lower Yenisey River in Russia, the becevnik was found to be an average of 4 to 6 m above the summer stage (Danilov, 1972). In comparison, the floodplain elevation was 8-10 m 23 above the summer stage. Danilov also reported the characteristic erosional effects of ice described elsewhere. Ice was the inferred cause of mounds and ridges of alluvium 3 to 10 m above the becevnik level. Smith (1974, 1979, 1980) believed that the scouring action of ice jams and ice floes have resulted in shaping and maintaining a characteristic bench-like cross section similar to the becevnik described by Hamelin and Danilov. A low-level bench near the level of the 2-year summer flood was maintained largely by summer flows while a high-level bench, at the height of the 9-year summer flood, was maintained largely by ice scour. Smith contended that an enlargement of the cross sectional area of large Albertan rivers (by a factor of three) by ice action was a reason behind higher bankfull return periods for Alberta Rivers than non-icy rivers described in the scientific literature. Channel width was cited as a primary factor that determines whether ice moves as a drive and causes scour, or does not move and melts in place. Kellerhals and Church (1980) presented what they believed were more plausible explanations for the infrequent occurrence of bankfull conditions in Alberta. Since major rivers in Alberta have experienced relatively recent entrenchment, Smith (1979) may have mixed up the contemporary floodplain with low terraces in his measurement of bankfull levels. Smith (1979) also neglected ice jam floods in his frequency analysis, which would tend to decrease calculated bankfull return periods. Kellerhals and Church (1980) and Church and Mi les (1982) felt that most evidence does not support large-scale ice scour as a major channel forming process. To the contrary, stranded ice may act to protect banks and ice jam flooding may be responsible for aggradation on the floodplain, possibly to higher levels (Church, 1988). 24 In the early 1970's, considerable geomorphic research, including river ice studies, were conducted in the Mackenzie River Val ley as part of a series of environmental assessments for a proposed northern pipeline. Henoch (1973) and MacKay and Mackay (1973) studied the magnitude and frequency of ice shoving along the Mackenzie using ice jam and flood level evidence including driftwood, silt deposits, ice damaged trees, and ice shoved bank material. Ice jam frequency was reconstructed using dendrochronological techniques (i.e., counting tree rings from x-ray negatives of tree cores). The trim line elevation (marking the maximum extent of ice shove) along the Mackenzie River increased in the downstream direction, with increasing discharge and ice load, and was particularly high at bends, island heads, and confluences. A t the Ramparts, a narrow bedrock constriction, ice shoved debris was found up to 14 m above the summer stage. A t island heads, trim lines were up to 9.5 m above summer stage, whereas on the sides of islands, trim lines were only 2 m high. The dating of trees between ice pushed ridges suggested that the features were not formed contemporaneously. M in imum ages of the highest ridges varied from 6 to more than 60 years. This suggested that only a limited number of major jams occurred each year because the supply of ice required for a single large jam would constitute the total ice cover of a long stretch of river (MacKay and Mackay, 1973). On the Liard River, Parker and Josza (1973) found morphologic effects of ice break-up similar to those of the Mackenzie River. They also observed ice-pushed ridges of gravel, vegetation damage, and driftwood levels and found that on straight, wide reaches, ice levels are similar to water levels (i.e., shoving is insignificant). However, at channel constrictions, ice shove limits may reach 17 m above the normal summer stage. Parker and Josza (1973) also interpreted the fact that only young trees were found on banks as indicating episodic removal or destruction of the riparian zone cover by severe ice jams. 25 Mackay and MacKay (1977) assessed the stability of ice-push features common on the Mackenzie River. They observed that boulder pavements, typically found downstream of boulder sources (e.g., til l), were of two types: tight and loose. If a muddy matrix was present, boulder pavements were tightly pressed and were relatively stable, perhaps lasting for centuries. The boulders in the "tight" pavements were aligned parallel to the bank, were nearly all flush (representing minimum resistance to ice push), were in mutual contact, and tended to fine in the downstream direction. If boulders were deeply buried, they were extremely stable since the depth of thaw does not exceed the depth of burial (Mackay, pers. comm., 1994). "Loose" boulder pavements, the least stable of the ice-push features observed, lacked a muddy matrix and had a lower concentration of boulders. Mackay and MacKay (1977) also described boulder ridges along the Mackenzie River as rhythmically spaced successions of boulders formed near ice jam locations. These ridges often extended from the summer stage to the vegetation trim line. A t the head of alluvial islands, Mackay and MacKay (1977) observed that river ice was responsible for plastering riverbed muds up the banks, causing distorted stratigraphy. A t the heads of islands formed out of erosional remnants of t i l l or glaciofluvial gravels, ice-push tends to form stable bouldery island buttresses (Mackay and MacKay, 1977). On the Mackenzie River near the mouth of the Keele River, Brooks (1993) described an ice-scoured bank with distinctive morphology. Ice thrusting was assigned as the agent responsible for a zone of bank disturbance from 4 to 13 m above the late summer stage. The smooth bank surface was concave upward, sloped from 42 to 52 degrees, and consisted of massive matrix supported diamicton (i.e., pavement). Brooks (1993) believed the surface of the bank was unfrozen at the time of pavement formation by intense bank shearing. The cohesive matrix along the bank was a key factor in the preservation of its morphology. Brooks (1993) concluded that 26 ice scour may be responsible for the removal of material supplied by mass movement upslope, and therefore may destabilize the upper portion of the bank. On the Oulanka River in Finland, Koutaniemi (1984) found that the most conspicuous evidence of break-up was the damage to vegetation and small scale bar forms. Although at one bend, an ice jam caused 10 m of bank retreat, overall effects of ice were small since peak discharges followed break-up and obscured most effects. Similarly, Bergeron and Roy (1988) reported that, following ice jam bed modification, subsequent fluvial processes changed the bed morphology back to its pre-flood characteristics. This may be especially true for braided rivers with high bedloads (Brooks, 1996). In summary, it appears that the significance of river ice is extremely varied presumably in response to the severity of break-up and the local characteristics of the river in question. Although there is a lack of generalized theories on the geomorphic role of river ice, previous evidence from around the world suggests that ice may be an important modifier of the vegetation and substrate along the margins of some rivers, but the balance of opinion appears to be that it is relatively unimportant in establishing the overall morphology of river channels. 1 . 6 R E S E A R C H O B J E C T I V E S The variable significance of ice noted in prior studies is justification for further investigation into the role of ice on northern rivers. Furthermore, little information is available on the effects of ice on rivers f lowing in climates other than arctic or sub-arctic. The Peace River in Alberta, a northward flowing boreal river, provides an interesting setting for study because it experiences ice jams on a regular basis and it also is experiencing long-term morphologic change in response 27 to f low regulation (Church, 1995; Church et al., 1996). F low regulation has generally reduced sediment transport, especially in the proximal gravel reaches closest to the dams. A long the proximal reach, the former floodplain has become a terrace as open-water flooding of those surfaces is now rare. A long the distal sand-bed reach, reductions in summer "f lushing" have resulted in aggradation and channel narrowing, as well as abandonment of the former floodplain. The colonization of former floodplain surfaces by herbaceous and shrub communities has coincided with the post-regulation regime, and appear to be integral to the long-term adjustment of the channel. The general objective of the present study was to determine how ice processes, particularly those related to ice jams, modify the morphology and riparian vegetation of the channel margins of the Peace River. Such information on boreal rivers is limited, and on the Peace River may be key to understanding and predicting vegetation succession and ultimately the long-term evolution of the river under a regulated regime (Church and North, 1996). The specific objectives o f the study were to determine 1) the effects of ice jams on the spatial distribution of riparian vegetation, and 2) the effects of ice jams on bank morphology and substrate. A long some northern rivers, ice jams represent the greatest form of episodic disturbance. However, ice processes are complex, with each watercourse having its own characteristics (Wil l iams and MacKay , 1973). On the Peace River there is little doubt the highest stages are the result of ice jams (Gerard and Karpuk, 1979), especially since regulation began in 1968 (Prowse and Conly, 1996). It was hypothesized that locations of severe ice jams would be reflected by distinctive riparian vegetation structure (i.e., species and age) and spatial distribution. Additionally, it was hypothesized that ice-related morphologies would be present and most evident at ice jam locations. In order to test these hypotheses, it was necessary to compile a history of ice jams on the Peace River complete with 28 location, severity (i.e., stages), and dates. Although records exist for points along the river, there was a lack of information on ice jam severity in many remote reaches of the river. The study therefore involved a thorough reconnaissance of the Peace River not only to collect information on the characteristics and distribution of ice-related features and riparian vegetation, but also to obtain dendrochronologic information on past ice jam events. The study is a regional-scale geomorphic inventory o f the ice related effects on channel margins and w i l l be o f use in the preliminary stages of future engineering and environmental projects that need to consider river ice on the Peace River. 29 2.0 STUDY AREA 2.1 L O C A T I O N The study reach is the Peace River in northwestern Alberta, between the Brit ish Columbia/ Alberta border and the town of Fort Vermi l ion (Figure 2.1). The Peace River, which is regulated, flows 1,240 km from the Wil l iston Reservoir near Hudson Hope, B .C . to its confluence with the Riviere des Rochers and the Slave River. Near the mouth of the Peace River is the Peace-Athabasca Delta (PAD) which encompasses 3,820 k m 2 (The Peace-Athabasca Delta Project Group, 1972) and is formed by three river deltas: the Athabasca, Peace, and Birch. Approximately 80% of the Peace-Athabasca Delta is situated in Wood Buffalo National Park located at the western end of Lake Athabasca (Prowse and Lalonde, 1996). From the mouth of the Peace River, the Slave River flows northward to Great Slave Lake, which empties into the Arctic Ocean via the Mackenzie River. The Peace River is regulated by the W . A . C . Bennett and Peace Canyon dams in B .C . The W . A . C . Bennett Dam is an earth-fill structure, which was constructed between 1963 and 1968 (Coulson and Adamcyk, 1969) approximately 28 km upstream of the community of Hudson Hope and 166 km upstream of the B.C./Alberta border. A t 183 m high, 2,000 m long, and 9 to 850 m thick, it is one of the largest earth-fill dams in the world (B.C. Hydro, 1993). F low regulation of the Peace River began in December 1967, in advance of the completion of the dam and generating station (Coulson and Adamcyk, 1969). By May 1972, the Wil l iston Reservoir was fil led. The Peace Canyon Dam was constructed between 1975 and 1980 (B.C. Hydro, 1993) at a location 23 km downstream of the W . A . C . Bennett Dam. The Peace Canyon Dam is a concrete structure 50 m high and 534 m long (B.C. Hydro, 1993) operated as a run-of-river project since there is relatively little live storage upstream in Dinosaur Lake. i 31 Downstream of the two hydro-electric dams, the Peace River is generally undeveloped with a small population spread out among agricultural areas on the plateau above and terraces and rolling land within the Peace River Valley. The main centers along the Peace River, within the study area, are the Town of Peace River (TPR) and the town of Fort Vermil ion. Smaller settlements include Dunvegan, Carcajou, L a Crete and the Metis settlement of Paddle Prairie. Industry along the Peace River is dominantly agriculture, but there is increasing development in forestry and resource extraction sectors such as o i l , gas and coal. One of the largest industrial developments along the study reach is the Daishowa-Marubeni International Ltd. (Peace River Division) pulp mi l l approximately 20 km downstream of TPR. Native, non-native and Metis fishers, trappers and outfitters have also established themselves in the Peace River Val ley taking advantage of the rich wildl i fe resource. Since the river is deeply entrenched into the adjacent plateau, access to the river is limited. The valley sides are prone to slumping and the tributary creeks and rivers incise into the slopes so that road building and maintenance are costly. Consequently, between the B.C./Alberta border and the town of Fort Vermi l ion (a river distance 665 km) there are only five bridges (at Clayhurst, Dunvegan, T P R , Daishowa pulp mi l l , and Fort Vermil ion) and two ferry crossings (at Shaftesbury and Tompkins Landing). During the winter months, the ferries are replaced with ice bridges across the river. Although road access to the river is generally limited to the crossings, the river is entirely accessible by boat. Since the Peace River has a low gradient ranging from 0.0007 to 0.00007 (Kellerhals et al., 1972) and only two substantial rapids, most of the river is classified as Class 1 according to the International River Classification System and poses low risk to small boats. Boat launching and camping are available at several municipal and provincial parks, and various private campgrounds and unofficial sites (see Appendix F for details). 2.2 P H Y S I O G R A P H Y The Peace River Basin is located within two major physiographic divisions: the Western Cordillera and the Interior Plains (Jackson, 1975). Much of the basin within the Cordillera drains into Wil l iston Reservoir while the basin area within the Interior Plains empties entirely into the Peace River, downstream of the Peace Canyon Dam. From the headwaters to the confluence with the Slave River, the two major physiographic divisions are subdivided into six physiographic regions (Klassen, 1989). The Cordillera consists o f peaks exceeding 2,000 m elevation and is made up of the Omineca Mountains, Rocky Mountains, and Rocky Mountains Foothills. The Rocky Mountain Trench separates the Rocky Mountains from the Omineca Mountains. The Interior Plains are made up of the Alberta Plateau, Peace River Lowland and the Great Slave Plain (See Figure 2.1). Although hydrologically influenced by all regions, except for the Great Slave Plain, the study reach is wholly encompassed within the Peace River Lowland. The Alberta Plateau is located discontinuously along the Peace River watershed divide within the Interior Plains. This is a formerly glaciated environment, which includes forested rolling uplands and broad gently sloping prairies covered with unconsolidated glacial t i l l , aeolian, glaciolacustrine and recent deltaic deposits. Many lakes, ponds, and sloughs are scattered remnants o f larger proglacial lakes that covered the area (Jones, 1966). During post-glacial times, rivers have incised deeply into the plateau surface creating unstable valley walls. The Peace River Lowland borders the Peace River from near Hudson Hope to near the Peace-Athabasca Delta and gently slopes to the northeast. It is characterized by roll ing plateau locally 33 interspersed by steep cuesta slopes. The steep valley slopes exhibit extensive slumping, especially along stream courses in the upper Peace River Basin. Interestingly, there have been three landslides since 1900 that have blocked the Peace River in B.C. : in 1913 near Cache Creek, in 1930 downstream of Farrell Creek, and in 1973, (the Attachie Slide) opposite Halfway River (Thurber, 1974). The characteristics of the Peace River Val ley vary considerably in the downstream direction from Hudson Hope to the Peace-Athabasca Delta. The upper Peace River Val ley, between Hudson Hope and T P R , is a deep, stream-cut, post-glacial valley incised 180 to 355 m into Cretaceous strata (see Section 2.3). Val ley bottom widths vary from about 400 to 1,000 m while valley top widths are up to 10 km. The Upper Peace River Basin is dissected by a dendritic system of tributaries most of which are on the south side of Peace River. The largest tributaries include the Pine River in B .C . and the Smoky River in Alberta. Val ley flats along the Upper Peace River include narrow fragmentary terraces and discontinuous floodplain surfaces. Downstream of T P R , the middle Peace River Val ley widens and the depth of incision from plateau surface decreases from 200 m near T P R to only 30 m near Fort Vermil ion. A t the same time, valley bottom widths increase from 1 to 4 km. The landscape on the prairies above the middle Peace River is generally flat and gently dipping towards the Peace River. Some locations are poorly drained and peaty. There are few large tributaries between T P R and Fort Vermil ion. Most drainage from the plateau adjacent to the Peace River is from minor streams and coulees. Downstream of Fort Vermi l ion and beyond the study reach, the landscape becomes extremely flat with the Peace River Val ley not exceeding 25 m in depth. The Peace River flows across a plain 34 of glaciolacustrine and glaciofluvial silts and sands. A t locations where the river has cut into bedrock, rapids have formed. The main rapids on the lower Peace include the Vermi l ion Chutes and the Boyer Rapids. The main tributary is the Wabasca River near Fort Vermil ion. The Peace-Athabasca Delta (PAD) lies within the Great Slave Plain physiographic region. The P A D consists of swamps, bogs, and sloughs with topographic relief seldom above 1 m above the surface of the major P A D lakes, except for the levees and bedrock islands outlying from the Canadian Shield located primarily in the northeast. 2.3 SURFICIAL AND BEDROCK GEOLOGY A long history of downcutting is responsible for the steep valley sides flanking the Peace River. These steep sides, particularly along the upper and middle reaches, provides a clear view of the surficial and bedrock geology and therefore the nature of landscape evolution along the Peace River. 2.3.1 Surficial Geology Surficial materials are exposed near the top of the incised valleys along the study reach and are composed of a sequence of interglacial fluvial and glaciolacustrine deposits overlain by glacial t i l l from the last glacial advance. These deposits, in turn, are overlain by late-glaciolacustrine deposits from Lake Peace, postglacial f luvial deposits, alluvial fans, slide debris, and aeolian deposits (Thurber, 1974). A t the Wisconsin cl imax, three glacial systems coalesced near the study area: the Laurentide ice sheet from east, the Cordilleran Ice Sheet from distant west, and a local system of coalescent 35 valley glaciers from the northern Rocky Mountains directly west (Mathews, 1980). The final phase of these glaciers was characterised mainly by stagnation (Mathews, 1980). The first stage of recession was marked by a general lowering of the glacier surface to a point where large upland areas were exposed. Meltwater streams flowed to the lowlands to the south where ice was still present. A number of proglacial and supraglacial lakes formed, which eventually drained as ice melted and outlets to northeast became available (Kumar, 1977). During glacial recession, extensive aeolian dunes formed with the prevailing southeast to northwest winds, particularly along the lower reaches of river. The P A D developed following the recession of the Laurentide ice sheet, with melt-water draining into a much larger Lake Athabasca. Initially, the Peace Delta grew rapidly but, as levees attained sufficient height, most of the f low and sediment were carried directly to the Slave River. The Peace River Delta is now considered to be inactive (Prowse and Conly, 1996). 2.3.2 Bedrock Geology The underlying strata along the study reach of the Peace River consists of a series of marine and non-marine sandstones, shales and siltstones of Upper and Lower Cretaceous age (Jones, 1966). This series of rocks represents broad marine transgression - regression - transgression cycles. The rocks are soft, fissile, and have high concentrations of montmorillonite and bentonite, making them susceptible to slope failure following undercutting (Catto, 1991). The strata are generally undeformed, dip at low angles, and thicken to the southwest into the Alberta syncline (Jones, 1966). Generally, the rocks exposed along the Peace River become older in the downstream direction. Within the study area, four geological formations outcrop at river level: the Dunvegan, 36 Shaftesbury, Peace River, and Loon River Formations. The Dunvegan Formation consists of deltaic shales, siltstones and sandstones and outcrops from the BC/Alber ta border to near the Saddle River (Jackson, 1975). The Shaftesbury formation is composed of fissile marine shales and siltstones with minor sandstone and discontinuously outcrops along the river level from the BC/Alberta border to T P R (Jackson, 1975). The Peace River Formation is made up of sandstone, marine shale and siltstone and outcrops from T P R to near Manning (Jackson, 1975). The sandstone in this formation has a low erodibility and forms prominent cliffs north of TPR. The Loon River Formation is a poorly lithified, horizontally bedded marine shale with beds of siltstone and ironstone (Kumar, 1977) and outcrops from Manning to near the Vermi l ion Chutes (Jackson, 1975). 2.4 C L I M A T E Although the study reach covers a large area, meteorological conditions are relatively homogenous across the basin, with some latitudinal and longitudinal exceptions as shown by the climate normals in Table 2.1 and Figure 2.2. The Peace River Basin falls within three broad climatic regions: Prairie, Boreal and Cordilleran (Prowse and Conly, 1996). However, the study reach falls entirely within the Boreal climate region, characterized by cold dry winters, warm summers, and moderate annual precipitation. During winter there is a gradient of decreasing temperatures from west to east and south to north across the Peace River Basin. In the middle of winter, these differences may reach 9°C. On average, mid-winter mean monthly air temperatures vary from -12 to -22°C. During November, when river freeze-up typically begins, the mean monthly air temperature ranges from -6 to -10.5°C. There are on average 210 to 240 days in the year with freezing temperatures, and 60 to 37 Table 2.1. Mean monthly temperature and precipitation at selected climate stations ^rviL^l^Tr ^  ^ ^ Peri°d ^  * l"° ^  A t m o s P h e r i c Environment Month Ft. St. John TPR Ft. Vermilion Jan -13.4 -17.5 -22.5 Feb -10 -13 -17 Mar -5 -7 -10 Apr 3.2 3 2 May 9.5 10 10.5 Jun 13.5 14 15 Jul 15.5 16.25 17 Aug 14.5 14.6 15 Sep 10 9 8.75 Oct 4.75 3.5 2.5 Nov -6 -8.5 -10.5 Dec -12 -15.5 -20 Mean Monthly Precipitation (mm) Ft. St. John TPR Ft. Vermilion 33.5 22.75 22 20 37 72 67 65 47 22 28 30 22.75 19.5 15 15.5 31 63.5 61.5 51 40 24 24 20 20 17.9 20 18 35 47 64 55 35 28 21 19 Annual 466 383 372 • Mean Monthly Precip.-Ft. St. John S3 Mean Monthly Precip.-TPR • Mean Monthly Precip. - Ft. Vermilion Mean Monthy Temp.- Ft. St. John Mean Monthly Temp.-TPR Mean Monthly Temp.- Ft. Vermilion o o <u 3 re 0) a E c o c re a> Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 2.2. Mean monthly temperature and precipitation at selected climate stations within the Peace River Basin for the period 1961 to 1990 (from Atmospheric Environment Service, 1993). 38 90 days a year when the temperature is below -20°C (Andres and Van Der Vinne, 1994). It is important to note that these values were recorded at airports above the river valley. During winter months, the valleys may have colder temperatures than the adjacent plateau surfaces, due to cold air drainage. Radiative cooling is responsible for the thick fogs that blanket the Peace River Val ley in the fall. Such a topographic effect gradually decreases in the downstream direction (Carder and Siemens, 1971). During the summer, mean temperatures in the basin can reach 17°C in July, and daylight hours are long. A t Fort Vermi l ion, for example, up to 18.4 h of sunshine is received on June 21 (Carder, 1965). Additionally, topographic effects along the deeply incised valley create a thermal oasis with little spatial gradient. Temperatures within the valley regularly exceed uplands and prairies by 5°C (Catto, 1991). Not only does the Peace River experience large annual temperature fluctuations, but it also experiences a large diurnal variation in temperatures, particularly in late winter and early spring. In January, chinook winds from southwest can cause rapid thawing and temperature increases. Increases from -40°C to 15°C in one day are not uncommon (Carder, 1965). Across the Peace River Basin, there is a gradient in precipitation from west to east and south to north. Mean annual precipitation varies from 466 mm at Fort St. John, to 383 mm at T P R , to 372 mm at Fort Vermil ion. The heaviest precipitation both locally and regionally is associated with convective weather between mid-May and early September (Hudson, 1997). Local downpours, lightning, hail , strong winds, and even tornadoes may be observed during this period. A s a result, June and July are the wettest months of the year. Max imum snowfall occurs during 39 January and maximum month-end snow cover occurs in February. Snow covers typically last from early October to end of Apr i l . Although not within the study reach, discontinuous permafrost exists northward of 56° latitude and between 54° and 56° latitude in the Cariboo Mountains (Hudson, 1997; Energy, Mines and Resources, 1986). 2 . 5 V E G E T A T I O N The Peace River flows through the Low Boreal Mixedwood ecoregion (Strong and Leggat, 1992) of the Boreal Plains ecozone (Prowse and Conly, 1996). With the exception of cultivated areas, most of the ecoregion is forested with aspen poplar (Populus tremuloides) and balsam poplar {Populus balsamiferd) and to a lesser extent, white spruce (Picea glaucd) (the climax species), black spruce {Picea mariana), paper birch {Betula papyrifera), lodgepole pine {Pinus contorta), and jack pine {Pinus banksiana). The Low Boreal Mixedwood Ecoregion is associated with Gray Luvisol ic soils in well-drained areas and organic soils in poorly drained areas (Strong and Leggat, 1992). . . Aspect, edaphic conditions, and the topography influence the distribution of vegetation along the Peace River Val ley. In the upper Peace River, south-facing well-drained slopes with silty soils are typically covered with grasslands and with groves of aspen poplar in gullies. North-facing slopes are covered with mixed aspen and balsam poplar, white spruce, and paper birch. Riparian vegetation along alluvial flats consists of flood tolerant species including balsam poplar, white spruce, green alder {Alnus crispa), wi l low {Salix spp.), horsetail {Equisetum spp.), and 40 various shrubs, sedges, and grasses. Balsam poplar, the dominant pioneer species, is abundant along the channel margins and islands, where high f low events provide moisture and nourishment (Rood, 1996). The climax species, white spruce, often occurs on islands and floodplains where the river has provided protection from the forest fires for at least 100 years (Cameron, 1977). Some of the white spruce stands are harvested along the river, usually during winter when the river can be crossed by ice. 2.6 H Y D R O L O G Y 2.6.1 Setting The Peace River Basin extends through 7 degrees of latitude and 17 degrees of longitude, and has a total drainage area of 324,385 k m 2 (MacKay, 1967). It is the largest tributary in the 1,788,916 k m 2 Mackenzie River watershed (Energy, Mines and Resources, 1986), and it is the largest river by drainage area and runoff in Alberta (Harrington and McLean, 1985). The headwaters of the Peace River lie in the Rocky and Omineca Mountains. Glaciers and snowfields feed streams that drain into Wil l iston Lake, principally through two rivers in the Rocky Mountain Trench: the Finlay and Parsnip Rivers. A t the W . A . C . Bennett Dam, located at the eastern arm of Wil l iston Lake, the drainage area is 70,200 k m 2 (Harrington and McLean, 1985). Controlled flows from the W . A . C . Bennett Dam enter Dinosaur Lake, the impoundment of the Peace Canyon Dam (PCD). From the P C D , the Peace River flows due east through the Rocky Mountain Foothills at Hudson Hope and enters the Peace River Lowland near Fort St. John. A t T P R , the river turns northward to Fort Vermi l ion, at which it changes direction to the northeast, until it reaches the Peace-Athabasca Delta (PAD) near the Slave River. Although f low is normally routed through the Slave River to the north, it can reverse and flow south through 41 three major channels, the Riviere des Rochers, Revi l lon Coupe, and Chenal des Quatre Fourches, when the Peace River is higher than the level o f Lake Athabasca. The Peace River falls under the subarctic-nival hydrological regime, which is dominated by spring snowmelt in late May and early June. Secondary peaks are also produced by spring and summer rainstorms. The largest discharges are typically produced when regulated snowmelt runoff originating in the Cordillera combines with non-regulated runoff and rainfall from the foothills and prairies. Peak water levels are typically due to backwater effects of ice jams that accompany spring run-off (Church, 1974). 2.6.2 Tr ibutar ies There are several major tributaries to the Peace River. In B .C . these include the Halfway, Pine, Beatton, and Kiskatinaw Rivers. The Pine River is the largest and provides 10% of the annual runoff to the Peace River at Peace Point. In Alberta, the main tributaries include the Smoky and Wabasca Rivers. The Smoky River is the largest tributary, comprising 17% of the basin and 13% of the runoff at Peace Point. The Wabasca, in comparison, comprises 12% of the basin and provides 7% of the annual runoff. Smaller tributaries within the study reach include the Pouce Coupe, Clear, Saddle, Heart, Whitemud, Cadotte, and Keg Rivers. These generally drain relatively small areas of the adjacent plateau. The watersheds in the western portion of the basin draining the Rocky Mountains yield more than twice as much f low per unit area as the tributaries draining the prairies and foothills. The primary reason for this is the significantly greater precipitation, including snowfall in the mountains. In addition, the timing of runoff is quicker in the mountains since the slopes are 42 steeper and the soils of the mountainous watersheds are generally shallower providing less secondary storage. 2.6.3 F low Regulation Discussion of the hydrology of the Peace River necessarily involves the topic of f low regulation by the W . A . C . Bennett and Peace Canyon Dams in B .C. The Gordon M . Shrum generating station at the W . A . C . Bennett Dam went into full operation in June 1972 (Fonstad, 1992) producing 2730 M W with 10 turbines passing 2000 m 3/s (Ketchum, 1996). The W . A . C . Bennett Dam is used by B .C . Hydro as a swing reservoir to provide peaking capacity for the entire B .C. power system (Andres, 1994). A t the Peace Canyon generating station, 700 M W is generated with 4 turbines passing approximately 2000 m 3 /s (Lewis et al., 1996). Both the W . A . C . Bennett and Peace Canyon Dams have the capacity to pass in excess of 10,000 m 3 /s through spillways i f required. The spillways were in operation in 1972, during testing, and in 1996, during dam investigations and repairs. The Peace River has been gauged sporadically at a number of locations since 1915, but more consistently since the construction o f the W . A . C . Bennett dam in 1968. Water Survey o f Canada (WSC) hydrometric stations are distributed along the length of the Peace River (see Table 2.2) and provide a basis for comparing the pre- and post-regulation hydrology. The primary stations which are still operating, and which most clearly reflect the hydrology of the study reach, are at Taylor, Dunvegan, T P R and Peace Point. Table 2.2. Act ive and inactive Water Survey of Canada hydrometric stations on the Peace River (from Environment Canada, 1998). Station Name/ ID Drainage Area (km2) Distance from mouth (km) Latitude (N) Longitude (W) Period of record (asterisk denotes active station) Geodetic elevation of gauge zero (m) Channel gradient Hudson Hope 07EF001 69 900 1212 56°01'39" 121°53'56" 1917-1922, 1949-1996* 447.629 .00058 Pine River 07FA004 83 900 1120 56°11'58" 120°48'42" 1979-1996* 405.774 .00062 Taylor 07FD002 97 100 1119 56°08'09" 120°40'13" 1944-1996* 400.273 .00059 Alces River 07FD010 118 000 1079 56°07'32" 120°03'32" 1974-1984 (Stage), 1990-1991 (Stage), 1992-1996* NA .00032 Dunvegan Bridge 07FD003 130 000 943 55°55'09" 118°36'19" 1960-1969, 1974-1996* 336.664 .00025 (Town of) Peace River 07HA001 186 000 839 56°14'41" 117°18'46" 1915-1932, 1957-1996* 304.800 .00035 Carcajou 07HD001 210 000 587 57°44'30" 117°01'55" 1960-1967 NA .000094 Fort Vermilion 07HF001 223 000 410 58°23'15" 116°02'05" 1915-1922, 1961-1978, 1979-1993 (Stage only) 243.508 .000094 Fifth Meridian 07KA002 282 000 270 58°39'00" 114°01'20" 1960-1964, 1966-1967 NA .00012 Peace Point 07KC001 293 000 115 59°06'50" 112°25'35" 1959-1996* 207.121 .00010 Sweetgrass Landing 07KC004 NA NA 58°55'41" 111°55'00" May-Aug 1971 (Stage) NA NA Carlson Landing 07KC003 NA 37 58°58'40" 111°48'50" May-June 1971 (Stage) NA .00015 Below Chenal des Quatre Fourches 07KC005 NA 16 58°54'00" 1H035'00" 1972-1992 (Stage), 1994-1996* (Stage) NA .00015 The total runoff from the Peace River basin has remained unchanged since regulation. Approximately 46.3 mil l ion dam 3 originates in B .C . and 20.0 mil l ion dam 3 from Alberta. Since the fi l l ing of the Wil l iston Reservoir, mean annual flows are near pre-regulation values (see Table 2.3). However, seasonal demands for hydro-electric power have significantly changed the monthly flows along the Peace River. A s shown in Table 2.3 and Figure 2.3, mean monthly flows have generally decreased in the summer and increased in the winter as a result of f low regulation. A t Hudson Hope, immediately below the P C D , mean summer flows have been reduced by Vz and mean winter flows have increased up to four-fold. This has eliminated the natural fall flow recession, and at Hudson Hope, regulated summer flows are lower than the pre-regulated winter flows. The effect of regulation generally decreases with distance downstream as additional unregulated tributaries contribute flow. A t Peace Point, summer flows following regulation are 2/3 of the pre-regulation values, while winter f lows are 2.5 times greater. Unregulated tributary flows have become less important during winter and more important during summer since regulation. Prior to regulation, tributary inflows between Hudson Hope and Peace Point approximately equaled the f low at Hudson Hope, the point of regulation (Prowse and Conly, 1996). Fol lowing regulation, inflows between Hudson Hope and Peace Point are 20% of the f low at Hudson Hope during winter but in the summer the inflows may be as much as double the Hudson Hope f low (Prowse and Conly, 1996). The Smoky and Wabasca Rivers supply about 75% of the inflow between Dunvegan and Peace Point. F low contribution from the Smoky River, referenced at Peace Point, has increased following regulation during summer from 15% to 23%, and during the winter has decreased from 16% to 7%. 4 i U ) ( O — ¥ 3 * §••9 B 3 3 o I S T J & 3. 3 o S» & "o 3' § & I •O O. 3 B oo K. O M K » S n> -i « era s 2, <t "> o- g. O f t < SP f t o B I B "g. Q. oT S 1 O 3 D < ?o 3 O o 3 v o o ' T 3 » 2 2 o 3. " £ o » 2 > « S I g I 5' 5 ' £ -a 0 0 ^ ~ •"** C " *. o - < ? 3 ' . O v O 3 C z o » 7*. B. S: cr E. P. 3 3 0 0 o o o o - J ON ON N J \ © O J o o NO U i NO N J ON - J \o v O O ON o N J N J N J N J N> o o ON N J o OO 4 ^ U> 4 * u > U l 4 ^ O J NO U i O I O W » O •< •< etu etu etu 3 3 3 peri( peril peri( a. a. a. V) C/l (A 3 3 t o o o w u > to ~ = S — o 2? « « •- 2 3 2 2 O o ft 5 3 § O Z O w > c ft o o f! c & o < ~ 7 3 0 0 2 > 2 J ui ON oo -j ~~ N * ^ W t o O U l OO O J £ k • s i L A - 0 O NO o o —i O OO 0 0 V i Q \ U » N J N J N J v © N J O J ON 43* O OO ON — N J O J O J 4 ^ O O OO c © ON O J - o . - — <— N J U t —• O - 4 O OO 1 0 N J N J O J ON — W U i w ^ K ) M 4 ^ N| 4*> 4 * . ON N J O J O i U t N J NO O J — ON O J N J 4 * . ~ j \o O NO — N J VO O 4 * |— •—• • O J LA W - v j W OO O O J — - J 4 * -O OO O — 0 0 A K ) U i oo O J U l U> — 0 0 N J - - 4 O J v O 4 * NO NO U l N J N J - - S J J > O 4 ^ U i J > O O ^O - U hO 0 O A i T i OO O — ON r O U i O N O N O N U I - ^ J N J N O ^ J U W U i i k U s J t O W - O O U i A ^ ' O A -W s J A v O ^ O t VO — O 0 ON 45k N J 0 0 — — i - — to ^ - u i u i 4 ^ U i a s — h o J> NO — ON O — O OO N J N J O O 0 4 ^ OO O J N J O J O J NO O J N J O — N J N J O J O J U i H - ON O J O J O J v O N J NO O J N J O ON - J N J N J - - - t O U i - J N J ON O 0 ON -— 4 * . O U i O J O 4 ^ O 0 4 * . OO O O J O 0 •— — — N J N J O J O J N J — — — O N O N O O - 4 0 0 0 4 * O J 0 4 ^ U » U I O J O J I — O J 4 ^ N J O O U I U < U I 4 * V © O O O J 4 ^ 0 4 i . O O O J O J O O O J 2 i C L 0 0 D 5? I- § - 5 era _ o S 2 — *o „ — _ o\ VO 5 w ^ 5i " M y r- fJV L/> O 5 o - J o T o o o vo o t S 5 o — S ) o S = N §. S ~ v o v O ° f — U> ( O rn o o s •o T J fis O '1. o a 2 T S 13 OJ _ o S3 ?o -o I T l O O •v -1 ft f t "0 p o O f t fis O o 70 ~ = It to UJ p o P co O ft o CD p a> I-I on c o n 3 P a. P co &. o 3 co O 3 3" CD ^3 CD P O 3 < CD 46 Hudson Hope(07EF001) • Pre-regulated flows • Regulated flows Ch_JTTJ7"nl,r-l, Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Taylor (07FD002) a B ra o m a ~ >. « f "p o c ra a 8000 7000 6000 5000 4 0 0 0 3000 2 0 0 0 1000 0 J J J 1 t l | 1 r - i l Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec TPR (07HA001) a P ra £ o in i £ o S c ra tu S Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Peace Point (07KC001) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 2.3. Mean monthly flows at selected W S C stations along the Peace River pre- and post-regulation (from Environment Canada, 1998). See Table 2.3 for periods of record. Tributaries to the Peace River have been important in maintaining a muted form of the pre-regulation seasonal pattern (i.e., peak flows in summer and low flows in winter) at T P R and Peace Point. The range of mean monthly flows at T P R is only 29% of pre-regulation values while that at Peace Point is 31%. Immediately below the dams, upstream of the Pine River, average low flows have been shifted following regulation from March to June and average peak flows have been shifted from June to December. Regulation has reduced the magnitude and variability o f average annual peak flows (see Table 2.4 and Figure 2.4). Post-regulation average annual peak flows are 30%, 71% and 63% of pre-regulation values at Hudson Hope, TPR, and Peace Point respectively. Although average annual peak flows have been reduced fol lowing regulation, the peak f low of record within the study reach occurred in 1990, following regulation, suggesting a strong influence of tributary flows along the Peace River, particularly downstream of TPR. In Figure 2.5, f lood magnitude and frequency estimates are presented for the stations at Hudson Hope, Taylor, Dunvegan, T P R , Fort Vermil ion, and Peace Point. Although frequency analysis is not strictly val id on a regulated river such as the Peace River, the values serve to indicate the magnitude of regulation. For lower return period floods (<10 years), there is significant difference in flood magnitudes between pre- and post-regulation regimes. A t higher return periods, however, the difference decreases downstream to T P R due to the strong influence of the unregulated Smoky River. The attenuation of peak floods downstream of T P R is a result of water going into temporary storage along the many secondary channels in the lower Peace River. 48 Table 2.4. Annual maximum daily discharges at selected WSC stations on the Peace River. Year Hudson's Hope Taylor Dunvegan TPR Carcajou Fort Vermilion Peace Point 07EF001 07FD02 07FD003 07HA001 07HD001 07HF001 07KC001 - * , ~. Non-regulated flowregime , > •< 1915 9600 14-Jul 1916 7420 6-Jul 1917 4760 5-Jun 6260 6-Jun 8470 8-Jun 1918 6430 22-Jun 8010 20-Jun 11100 22-Jun 1919 6460 25-Jun 9030 25-Jun 9060 27-Jun 1920 5920 18-Jun 9540 19-Jun 9940 21-Jun 1921 7220 10-Jun 8750 10-Jun 8980 12-Jun 1922 7560 5-Jun 10600 6-Jun 9260 8-Jun 1923 8240 17-Jun 1924 7960 19-May 1925 7790 21-May 1926 8180 19-Jun 1927 8780 8-Jul 1928 6680 2-Jun 1929 6170 8-Jun 1930 9260 20-Jun No extreme discharge records are available for the period 1931-1944 1945 6460 31-May 1946 6710 29-May 1947 6400 1-Jun 1948 11500 30-May 1949 5380 5-Jun 1950 6800 16-Jun 6880 16-Jun 1951 5180 25-May 1952 5300 2-Jul 5890 12-Jun 1953 6140 22-May 8100 21-May 1954 6290 12-Jun 9430 28-May 1955 6120 27-Jun 7900 28-Jun 1956 5070 8-Jun 6540 7-Jun 1957 5830 25-May 7760 25-May 1958 6310 31-May 7960 30-May 11200 31-May 1959 5380 5-Jun 6990 5-Jun 8180 7-Jun 7650 3-Jul 1960 5920 30-Jun 7080 30-Jun 7280 1-Jul 11800 23-Jun 11000 24-Jun 10800 28-Jun 1961 7480 28-May 8330 28-May 8920 29-May 10700 29-May 9510 31-May 9910 31-May 9910 2-Jun 1962 5890 28-Jun 7360 28-Jun 7820 28-Jun 9150 29-Jun 8500 1-Jul 8890 E 1-Jul 10000 3-Jul 1963 5010 25-May 6460 25-May 6710 26-May 8890 26-May 7670 26-May 10000 E 1-Jun 1964 8810 14-Jun 10000 13-Jun 11100 14-Jun 12900 14-Jun 12100 18-Jun 12400 16-Jun 11900 19-Jun 1965 6060 4-Jun 7790 4-Jun 8830 30-Jun 13300 11-Jul 11200 12-Jul 11500 13-Jul 10500 15-Jul 1966 5470 13-Jun 6460 13-Jun 6510 13-Jun 8380 12-May 7480 15-May 7590 17-Jun 1967 6540 10 Jun 8160 10-Jun 9030 4-Jun 11300 5-Jun 10000 12-Jun 10400 6-Jun 10000 7-Jun Williston Reservoirfilling from December 1967 to May 1972 1968 360 E 22-Det 3280 14-Jun 4250 14-Jun 6850 15-Jun 6060 17-Jun 5830 19-Jun 1969 864 4-Nov 1790 5-Jun 2570 1-May 3620 2-May 2760 9-Jun 4810 6-May 1970 1050 30-Dec 2560 5-Jun 3710 6-Jun 3620 8-Jun 3910 11-Jun 1971 1140 24-Mar 4020 17-Jun 9660 13-Jul 7960 15-Jul 7820 17-Jul 1972 5130 12-Jul 5690 1' Jun 14100 14-Jun 10400 16-Jun 8810 18-Jun Regulated flow regime 1973 1650 16-Mar 2740 18-May 4530 19-May 4620 21-Jun 6340 22-Jun 1974 1680 27-Nov 2620 18-Jun 7280 B 26-Apr 8700 B 3-May 1975 1680 6-Feb 2250 29-Jun 2520 30-Jun 4130 1-Jul 3940 3-Jul 4870 6-Jul 1976 2090 12-Jul 4620 5-Aug 5410 6-Aug 7700 19-Aug 6710 E 21-Aug 7330 23-Aug 1977 1760 26-Jan 3030 10-Jun 5270 11-Jun 5970 11-Jun 6340 13-Jun 6800 16-Jun 1978 1770 4-Feb 2470 6-Jun 2860 1-Jun 3650 7-Jun 3820 E 6-Jun 4760 E 8-Jun 1979 1810 6-Feb 2930 5-Jun 3460 5-Jul 5180 16-Jun 6290 10-Jun 1980 1640 6-Jan 2410 19-Jun 2830 20-Jun 4710 6-Jun 3660 11-Jun 1981 1800 30-Dec 2900 28-May 3130 28-May 4420 29-May 4340 2-Jun 1982 1860 26-Nov 2760 16-Jul 2860 17-Jul 10800 17-Jul 6900 21-Jul 1983 3070 11-Jul 4770 9-Jul 5310 16-Jul 7140 17-Jul 6930 22-Jul 1984 2280 14-Oct 3480 8-Jun 4980 A 9-Jun 6800 10-Jun 6420 14-Jun 1985 1870 26-Nov 2480 24-Sep 2670 23-Sep 3470 23-May 4500 B 9-May 1986 1780 20-Jan 2510 30-May 2720 29-May 4590 30-May 4930 4-Jun 1987 1960 12-Apr 4010 2-Aug 5740 A 3-Aug 11400 3-Aug 9500 7-Aug 1988 2000 23-Jan 2630 13-Jun 3040 17-May 4270 13-Jun 5430 20-Jul 1989 1780 A 3-Feb 2200 3-Jun 2210 4-Jun 3650 25-Aug 4400 13-Jun 1990 1790 8-Nov 5190 13-Jun 7600 A 13-Jun 16500 14-Jun 12600 17-Jun 1991 1840 24-Oct 2310 12-Oct 2130 7-Jun 4380 11-May 5400 21-Jun 1992 1950 17-Dec 2800 3-Jun 2790 4-Jun 3830 4-Jun 4800 B, 27-Apr 1993 2000 16-Mar 2090 B 30-Nov 2820 25-Jun 3560 26-Jun 3600 6-Jul 1994 1870 8-Dec 2700 14-Jun 3600 15-Jun 4970 16-Jun 5310 B 28-Apr 1995 1760 5-Dec 1840 5-Dec 2370 6-Jul 3320 7-Jul 3050 B 8-May 1996 5190 31-Jul 6220 19-Jul 7300 20-Jul 7750 21-Jul 9760 25-Jul Notes 1 Data were extracted from Hydat CD-ROM 1996 (Environment Canada, 1998). 2 Discharges in nrVs; "A" denotes manual measurement, "B" denotes ice conditions, "E" denotes estimate. 3 WSC stations at Dunvegan, Carcajou and Fort Vermilion operate or have operated primarily in the summer months. 4 1996 peak flows were related to the operation of the W . A . C Bennett Dam spillway. 49 15000 T 10000 5000 HUDSON HOPE A. in o T - rg cn cn m CM o o co m co cn o cn in cn o in cn o co cn m CD cn 15000 10000 5000 0) UJ O < X o CD o >-_l < Q s 3 TAYLOR m o m •<- CM CM cn cn cn o CO cn m o io CO •<* T f CD CD CD in o IO CD cn cn m CD cn o in cn o co cn m o w co cn cn cj> cn cn 15000 10000 5000 DUNVEGAN CO o •<- CM o cn LO o CM co cn cn o cn cn o in cn o <o o m CO cn o co m co cn 3 s 15000 10000 5000 TPR 16500 V " A ~ V in o « - CM cn cn m o CM co cn cn m co cn o cn cn o in cn LO LO cn o CD cn in CD O cn m cn o co cn m co o 15000 10000 5000 PEACE POINT in o in o » - CM CM CO cn cn cn cn in co cn o Ti-en m •* cn o <o cn o cn in cn o co cn m cn cn Figure 2.4. Annual maximum daily discharges for selected W S C stations on the Peace River. Note that Wil l iston Reservoir was fi l led from December 1967 to May 1972, and that the spillway was operated at the W . A . C . Bennett Dam in 1972 and 1996. 2-YEAR RETURN PERIOD FLOOD 12000 10000 H 8000 6000 4000 2000 60 80 100 120 140 160 180 200 220 240 DRAINAGE AREA (1000 km1) 260 280 300 5-YEAR RETURN PERIOD FLOOD 14000 12000 10000 8000 6000 4000 2000 14000 12000 -| 10000 8000 6000 4000 2000 16000 14000 12000 10000 8000 -I 6000 4000 2000 80 100 120 140 160 180 200 220 240 260 280 300 DRAINAGE AREA (1000 km2) 10-YEAR RETURN PERIOD FLOOD 80 100 120 140 160 180 200 220 240 260 280 300 DRAINAGE AREA (1000 km2) 25-YEAR RETURN PERIOD FLOOD 60 80 100 120 140 160 180 200 220 240 260 280 300 DRAINAGE AREA (1000 km2) 16000 14000 12000 10000 8000 6000 4000 2000 50-YEAR RETURN PERIOD FLOOD - Non-regulated flows - Regulated flows 120 140 160 180 200 220 DRAINAGE AREA (1000 km2) 240 260 280 300 Figure 2.5. Estimates of 2-, 5-, 10-, 25-, and 50-year return period floods on the Peace River. Stations represented by the data in increasing drainage area order are: Hudson Hope, Taylor, Dunvegan, TPR, Fort Vermilion, and Peace Point. See Table 2.3 for periods of record. 51 2.7 P E A C E R I V E R I C E 2.7.1 Freeze-up Prior to regulation of the Peace River, flows during the early winter remained well below 1000 m 3/s and water temperatures declined in response to the ambient air temperatures along the river. During the fall f low recession, near-simultaneous cooling of the river resulted in the formation of ice floes that bridged at several locations (Acres, 1980). Ice began to form in early November, with the formation of a stable ice cover in the downstream reaches near Peace Point, and a series of intermittent covers upstream of Fort Vermil ion. Since ice bridging was common at several locations, complete freeze-up occasionally occurred in upper reaches prior to lower reaches. In general, the ice progressed upstream to T P R in late November to early December, and at Hudson Hope by early to mid-December. During freeze-up, stages increased by approximately 1 m at T P R as ice formed primarily by juxtaposition. Fol lowing regulation of the Peace River, the annual fall recession had been eliminated and winter discharge had increased on average two to three-fold. During the winter, flows were increased above 1640 m 3 /s (Ketchum, 1996), making tributary inflows, including the Pine, Smoky and Wabasca Rivers less important. Warm hypolimnetic water released from deep within the Wil l iston Reservoir, with temperatures typically between 0.5 and 4.0°C (Keenhan et al., 1982), increased temperatures significantly along the river, at least as far as T P R (Prowse et al., 1996) and possibly as far downstream as Vermi l ion Chutes (Andres and V a n Der Vinne, 1994). The timing of freeze-up has been delayed, and it now takes up to four months for the ice cover to progress, as a single front, upstream to its ultimate destination (Andres and Van Der Vinne, 1994). Between T P R and Fort Vermil ion, the freeze-up date has been delayed by as much as two months in some years. The duration of the ice-cover has also been shortened significantly, at 52 least as far dovvnstream as T P R (See Table 2.5). Upstream of the B.C./Alberta border, formation of a complete ice cover has become rare. The stage increase that is associated with freeze-up has increased dramatically since regulation. In the upper reaches, there is a 2 to 3 m increase in the backwater associated with the formation of the ice cover. This is largely the result of the significantly higher winter flows, which tend to result in thick and rough consolidated ice. Typical freeze-up events under regulated regime Below the Vermi l ion Chutes, a set of rapids and falls with a total drop of 9 m over 3 km, low gradients cause a stable ice cover to form by the extension of border ice from low velocity areas adjacent to the river bank. A long this reach, the timing of freeze-up and duration of the ice cover do not appear to have changed since regulation (Prowse and Conly, 1996). Ice progression above the Chutes depends on the staging of ice over the rapids and falls, which occurs by a combination of juxtaposition and shoving. Above the Chutes, gentle gradients allow ice to progress in an orderly fashion by juxtaposition, even under high discharges. Ice thickness averages about 0.5 m, unless local shoving occurs. Although records are short, regulation has delayed freeze-up at Fort Vermi l ion from late November or early December to late December or early January (Andres and Van Der Vinne, 1994). A s the gradient increases upstream of the Notikewin River, ice forms either by juxtaposition or consolidation, depending on the temperature and flow. Near T P R , ice typically forms by 53 Table 2.5. Freeze-up, ice-cover duration, and breakup statistics for hydrometric stations along the Peace River (from Prowse et al., 1996). Pre-regulation Post-regulation Shift in mean (alpha = 0.05) , . . Standard Mean date / _ . n ^ Deviation (days) ( y e a r s ) Standard Mean date / _ . .. n ^ Deviation , Days . (years) J (days) V J FREEZE-UP Hudson Hope Taylor Dunvegan TPR Fort Vermilion Peace Point Dec 13 7.8 5 Nov 15 7.2 5 Nov 16 7.9 6 no ice no ice Jan 01 16.9 21 Nov 21 9.3 21 yes no DURATION OF ICE-COVER Hudson Hope Taylor Dunvegan TPR Fort Vermilion Peace Point 141 58.4 24 158 19.3 17 124 18.1 5 168 11.2 5 169 10.9 6 no ice no ice 97 29.0 21 160 11.8 21 yes no BREAK-UP Hudson Hope Taylor Dunvegan TPR Fort Vermilion Peace Point Apr 27 10.6 3 Apr 16 9.1 5 Apr 29 5.0 5 May 02 5.0 6 no ice no ice Apr 10 13.2 21 Apr 24 7.6 12 Apr 28 6.1 21 no yes no Notes: 1. For stations with sufficient data, significance tests using Student's t-test were conducted to determine if a shift between the pre- and post-regulation mean dates occurred for freeze-up and break-up, and for the duration of the ice cover. 2. Due to the paucity of the pre-regulation data for Hudson Hope, the duration of the ice cover is based on the period of ice effect. Although the statistics for the other stations are based on continuous solid ice cover, using ice effect period for Hudson Hope illustrates the impact of regulation on the ice regime at this location. 3. Although the Taylor gauge has been affected by ice and temporary lodging of ice has occurred around the gauge, WSC personnel responsible for the gauge state that a permanent solid ice cover has not formed at this gauge since regulation. 54 juxtaposition with some thickening near the Highway 2 bridge due to local shoving. It is not uncommon for 1 m thick ice to become 2.5 m thick as shoving takes place. Upstream of the Notikewin River, the stage increase produced by a juxtaposed cover is less than 2 m. However, a consolidated ice cover may increase stages by as much as 5 m. This stage increase may be more significant than the increase during spring break-up (Andres, 1996). For example, at T P R freeze-up stages were consistently higher than the break-up stages in the period 1983-91 (Figure 2.6). Where consolidated ice forms, stages may remain relatively high throughout the winter period. Between Notikewin River and Dunvegan, high stages may last two to three months, a duration that far exceeds that produced by an open-water flood (Prowse and Conly, 1996). Since regulation, freeze-up at T P R has been delayed from 1 week to 2 months (Andres and Van Der Vinne, 1994). This has reduced the duration of a solid ice cover from 124 days prior to regulation, to about 97 days following regulation (Prowse et al., 1996). A t T P R following regulation, several temporary ice covers form and break up before a solid ice cover establishes. Further upstream, between Dunvegan and Hudson Hope, steep gradients prevent the formation of a juxtaposed ice cover under most combinations of temperature and discharge. Instead, consolidation occurs at freeze-up, resulting in ice thicknesses of up to 5 m and stage increases of up to 6 m (Andres, 1996). Smooth, thermal ice is rare in the upper reaches of the Peace River except in low-velocity areas such as in secondary channels, over shoals and where border ice can develop. The process of consolidation and ice shoving results in the formation of shear walls. These vertical faces of ice along the channel margin replace the natural bank as the contact point 55 Figure 2.6. Peak freeze-up and break-up stages at the Town of Peace River (TPR). 322 • Peak freeze-up stage (m) I Peak break-up stage (m) 320 f 318 . § . 3 1 6 2 _ i UJ 314 312 310 3 0 8 306 llll CO co cn CO cn in cn r-r-cn cn r-cn co co cn co cn co cn cn co cn co cn cn 56 between the floating ice cover and the bank. The rate of water level increase and the channel geometry are important in determining the amount of grounded ice and corresponding shear wall formation and type of ice-bank contact (Acres, 1984). Ice conditions on the upper Peace River have been extremely variable from year to year following regulation (see Figure 2.7). A t Dunvegan freeze-up occurs in early January, but occasionally the river is ice-free all year. The maximum extent of ice generally occurs in late January and early February (Andres, 1981) and on average reaches the provincial border. Ice has progressed upstream of the border in only 40% of the post-regulation years (Andres and Van Der Vinne, 1994). A t Taylor, significant ice conditions (with stage increases) have occurred in only 6 post-regulation years: 1972, 1974, 1979, 1983, 1985, and 1996 (Ketchum, 1996). 2.7.2 Break-up and Ice Jams The break-up partem along the river since regulation has changed mostly along the upper reaches of the Peace River. In general, freeze-up ice stages have been increased as a result o f increased winter flows. It is generally believed that the higher the freeze-up ice cover is stabilized, the lower the severity o f break-up wi l l be, since the ice cover w i l l convey greater flows without breaking (Prowse et al., 1996). Add to that the reduced significance of tributary flows at break-up since regulation, and it is expected that break-ups should be more thermal than mechanical. This has been observed in the upper reaches. Between Taylor and the Smoky River, break-up is determined by the regulated releases and to a lesser extent, flows from the Pine River and other smaller streams. Downstream recession of the ice cover begins in late February or early March. Gradual thermal erosion causes 57 1200 Nov Dec Jan Feb Mar Apr May Jun Date Figure 2.7. Ice-cover progression during 21 winters (from 1973/74 to 1993/94) on the Peace River (from Andres and Van Der Vinne, 1994). This plot of annual ice observations serves to indicate the high variability in timing, duration and extent of the ice-cover from year to year since regulation. 58 the formation of leads which lengthen, collapse and initiate short jams that advance at a rate of 5 to 15 km/day (Andres, 1981). Ice jams produce stage increases under these conditions rarely exceeding 1.5 m, which is generally less than the freeze-up stage increase. Between the Smoky River and TPR, break-up depends on the magnitude and timing of break-up on the Smoky River, and the location of the ice front on the Peace River ( B C H P A , 1977a). Because the Smoky River drains the foothills of the Rockies and the prairies south of TPR, peak flows and break-up on the Smoky River are typically earlier than the main stem of the Peace River near TPR. If Smoky River discharges are relatively high during winter, large hanging dams may form. These hanging dams are resistant to spring break-up and cause large surges when released ( B C H P A , 1979). If the Peace River ice has not receded past the Smoky River confluence at the time of Smoky River break-up, ice jamming and potentially serious flooding may occur near the mouth of the Smoky River and TPR. Several ice jam floods of this type have been documented for the T P R and are summarized in Appendix C. The T P R is the major center along the Peace River and therefore has the most complete record of ice jams on the Peace River. A t TPR, major ice jam floods that seriously jeopardized or damaged the town occurred in 1915, 1934,1948,1963, 1973,1974,1979, 1982, 1992, and 1997. The recorded peak stages of the floods at T P R are presented in Figure 2.6. Since the early 1970's, the regulation strategy adopted by B .C. Hydro is to release sufficiently high flows from the dams to promote thermal break-up and in-situ ice melt (Fonstad, 1992). The goal of the strategy is to minimize the risk of ice jams and flooding near TPR. In years with lower than average snowpack and/or an extended melt period, increased regulated flows may increase the rate of Peace River ice retreat such that it passes the Smoky River prior to break-up of the Smoky River. In those cases, regulation initiates an earlier break-up at T P R than would be expected under normal circumstances. However, during years with large snowpacks and a rapid increase in temperature, discharges from the dams are reduced i f the Smoky River break-up is imminent. This is done to provide sufficient freeboard for ice jam stage increases near TPR. As outlined in Appendix C , this has been a reasonably successful strategy. The mean date of break-up at T P R since regulation is earlier by about one week, but this may not be statistically significant (Prowse et al., 1996). However, break-up dates at T P R have become more variable since regulation (Prowse et al., 1996). It is important to note that since T P R is situated near the uppermost extent of the solid ice-cover, mid-winter break-ups are also possible, the most notable being in February, 1992. With an intense period of warming in the winter, such as that in 1992, a break-up event can be precipitated almost concurrently with the initial establishment of the freeze-up cover (Fonstad, 1992). Between T P R and Fort Vermi l ion, the channel gradient decreases and break-up occurs as a series of ice runs and jams. Serious flooding occurs during high runoff situations (e.g., 1963,1973 and 1974). It is unlikely that the effects of regulation have considerably altered the break-up patterns at the lower end of this reach (Andres, 1981), but determining the relative influence of regulation in the distal reach is complex because of: 1) climatic variability and its effect on f low regime, both upstream and downstream of the dam, and 2) the large contributing area downstream of the dam, much of which is ungauged (Prowse and Lalonde, 1996). A t Fort Vermi l ion, ice break-up occurs in late Apr i l or early May, which has changed little since regulation. 60 Break-up patterns at Peace Point, beyond the study reach, are likely unchanged as well . However, ice jams have become rare since 1974. Prowse and Conly (1996) reported that there have been at least 13 major ice jams near Peace Point since 1803. The most recent ones being in 1958, 1963, 1965, 1966, 1967, 1972, 1974, 1979, and 1992. Most of these events were associated with high runoff events from major tributaries, including the Smoky River. Prowse and Conly (1996) found no significant change in the f low during break-up following regulation. However, they discovered a significant change in the contribution to f low between the headwaters (regulated) and lower tributaries (non-regulated). While the headwater contributions to Peace Point break-up f low increased from 17 to 33% following regulation, the tributary contributions decreased. The reduction in lower tributary flows, perhaps due to climate change, was believed to be a major factor in the reduced severity of ice jams in the lower Peace River since the mid-1970' s. Although open-water floods, such as the one that occurred in June 1990, produce the highest discharges on the Peace River, the highest stages are often a result of ice jamming during spring or, in the upper Peace River, consolidation during freeze-up. Annual peak water levels versus discharge are plotted in Figure 2.8 for Dunvegan, T P R and Fort Vermil ion. Although, the confidence in discharge measurements under ice conditions is low, the plots serve to show that extreme stages can be produced under relatively low discharges i f ice jams occur. A t T P R , ice jam floods exceeded the open-water record flood on at least six occasions, which is probably similar to other locations downstream. DISCHARGE (m3/s) 2000 4000 6000 8000 10000 12000 14000 16000 o 8 7 348 -» 1988 freeze-up stage 346 Approx. level of June 1990 open water flood. • ^ DUNVEGAN 07FD003 344 1991 break-up ~~~~~~6~ stage 93 Measurements made between 1961 -1995 342 340 338 320 -g- 318 o 0) T3 O O) O 316 UJ UJ 314 312 310 258 256 254 252 250 248 246 9 2 -o-^ 1997 ice jam 7934 /ce jam TOWN OF PEACE RIVER 07HA001 Measurements made between 1961 -1995 1888 ice jam 1934 ice jam 1894 ice jam -1963 ice jam — 1876 ice jam Approx. level of June 1990 open water flood. 1965 ice jam n 1966,1974 break-up 1979 ice jam 1987 break-up FORT VERMILION 07HF001 Measurements made bewteen 1968-1977 Note: Elevations are referenced to the Highway 88 bridge. 244 I ' i ' ' i I 0 2000 4000 6000 8000 10000 12000 14000 16000 DISCHARGE (m3/s) Figure 2.8. Annual peak water levels versus discharge under break-up conditions at three locations on the Peace River. Break-up data are open circles, while open-water data are solid circles. 62 2 . 8 C H A N N E L M O R P H O L O G Y The Peace Paver has three primary reaches, each with its characteristic morphology (see Figure 2.9). The upper Peace includes a relatively steep, cobble-gravel bed reach between the P C D and the Smoky River. The middle Peace is a gravel-sand bed reach between Smoky River and Fort Vermil ion, and the lower Peace includes the mainly sand-bed reach downstream of Fort Vermil ion (Prowse and Conly, 1996). The middle reach, which is a transitional zone between a gravel and sand bed, has no clear lower boundary. Church (1995) set the boundary further upstream, near Carcajou, since there is a marked change in gradient, and since downstream of Carcajou the bed and banks are primarily sand (although gravel can be locally significant). In the present study, the lower limit of the middle reach was set at Tompkin's Landing slightly downstream of Carcajou, primarily to be consistent with the channel reach breaks outlined by Church et al. (1996). Note that the lower Peace investigated in this study, from Tompkin's Landing to Fort Vermi l ion, is considerably different than the Peace River below Fort Vermil ion. The reach upstream of Fort Vermil ion is confined within a low valley and has a considerable amount o f gravel. Downstream of Fort Vermi l ion, and particularly the Vermi l ion Chutes, the channel is generally unconfined and flows over fine-textured glaciolacustine and glaciofluvial deposits. The present study investigated the Peace River between the provincial border and Fort Vermil ion. The three principal reaches were divided into sub-reaches, each of which encompassed homogeneous morphology. Typical sub-reaches included individual island complexes or areas near tributaries (see Appendix D). Reach 1a Sub-reach: 1a.4 Site: 82 km: 1016.0 Photo: 4.24-25 Reach 1b Sub-reach: 1b.3 Site: 250 km: 912.55 Photo: 42.16-18 Reach 2a Sub-reach: 2a.5 Site: 360 km: 782.60 Photo: 45.14-16 Reach 2b Sub-reach: 2b. 8 Site: 536 km: 563.75 Photo: 29.15-17 111 O uj $ Reach 3a Sub-reach: 3a.4 Site: 649 km: 480.95 Photo: 17.19-21 Figure 2.9. General valley and channel morphology of the principal study reaches. 64 2.8.1 Upper Peace River (Reaches la and lb) The upper Peace River, as defined in this study, is from the provincial border to immediately upstream of the Smoky River. This reach consists of two principal study reaches, l a and l b 1 and 14 sub-reaches (see Table 2.6). It is characterized as a wandering channel split by several islands and occasional bars but always with an identifiable main thread. The channel of the upper Peace River is deeply incised below the adjacent forested plateau and has an average width of 500 m. Channel width, however, varies between 350 m at confined sections to 700 m near major island complexes (see Figure 2.10). These island complexes occur at two locations, Many Islands and Montagneuse Islands near river km 1000. Slumping of the valley walls is common along this reach, particularly where the river has impinged upon the toe of the slope. The valley flat is comprised of narrow and discontinuous terrace and floodplain surfaces. The mean sinuosity of the upper Peace is 1.33. This reflects the relatively straight reach near the border and between Dunvegan and the Smoky River, and a sinuous reach upstream of Dunvegan. The bed and banks are mainly gravel over soft and easily erodible shales. A long the banks, the gravel is overlain by a discontinuous accumulation of silt. The median bed material size (D50) is 53 mm while the D90 is 127 mm (Kellerhals et al., 1972). Channel gradients generally decrease from 0.00036 at the border to 0.00025 near the Smoky River (see Figure 2.11). The morphology of the Upper Peace River has been in a state of adjustment following regulation (cf. Church et al. , 1996). Reductions in peak flows have reduced the transport capacity of the river. Coarse-grained sediment supplied by tributaries is aggrading at confluences (Bray and 1. In Church et al. (1996), the Peace R. was divided into 3 discontinuous reaches 1,2, and 3 which fell within the gravel, transitional, and sand-bed reaches of the river: These 3 reaches correspond precisely with reaches la, 2a, and 3a used in the present study (see Figure 2.1). The unidentified channel between la and 2a, and between 2a and 3a was defined here as lb and 2b since they display characteristic morphologies very similar to la and 2a respectively. Table 2.6. Study sub-reaches in the Upper, Middle and Lower Peace River Sub-reach Sub-reach Name River Distance (km) Study Sites within Sub-reach1 Upper Peace River Reach la: B.C. /Alberta border to Dunvegan la.l Pouce Coupe River 1074-1054 1-30 la.2 Clear River 1054-1038 31-44 la.3 Campbell's Lease 1038-1018 45-78 la.4 Many Islands 1018-1005 79-109 la.5 Montagneuse Islands 1005-990 110-147 la.6 Highland Park 990-974 148-171 la.7 Pratt's Landing 974-957 172-191 la.8 Dunvegan 957-942 192-213 Reach lb: Dunvegan to Shaftesbury Ferry lb.l Green Island 942-927 214-239 lb.2 Erin Lodge 927-915 240-248 lb.3 Saddle (Burnt) River 915-906 249-254 lb.4 Long Island 906-891 255-263 lb.5 Peace Valley Ranch 891-878 264-270 lb.6 Mushikitee Island 878-865 271-288 Middle Peace River Reach 2a: Shaftesbury Ferry to Mannin 2a. 1 Shaftesbury Ferry 865-849 289-300 2a.2 Smoky River 849-840 301-313 2a.3 Town of Peace River 840-819 314-330 2a.4 Carmon Creek 819-798 331-345 2a.5 Whitemud River 798-774 346-363 2a.6 Cadotte River 774-749 364-374 2a.7 Sunny Valley 749-724 375-397 Reach 2b: Manning to Tompkin's Landing 2b. 1 Hotchkiss 724-694 398-420 2b.2 Keppler Creek 694-676 421-436 2b.3 Notikewin River 676-657 437-450 2b.4 Nina Lake 657-635 451-467 2b.5 Big Bend 635-618 468-487 2b.6 Scully Creek 618-591 488-509 2b.7 Carcajou 591-567 510-529 2b.8 Keg River 567-546 530-561 Lower Peace River Reach 3a: Tompkin's Landing to Fort Vermilion bridge 3a. 1 Tompkin's Landing 546-526 562-576 3a.2 Moose Island 526-500 577-610 3a.3 La Crete 500-485 611-638 3a.4 Prairie Point 485-441 639-693 3a.5 Bluemenort 441-419 694-735 Notes: 1. Study sites included all points where observations, measurements, and photographs were taken. They include survey sites where bank morphology was surveyed, vegetation transects were conducted, and scar elevations were measured. 66 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 a a a a a a a a b b b b b b a a a a a a a b b b b b b b b a a a a a 1 2 3 4 5 6 7 8 1 2 3 4 5 6 1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 1 2 3 4 5 STUDY SUB-REACH Figure 2.10. Mean channel width and sinuosity of the study sub-reaches. STUDY REACH MEAN CHANNEL GRADIENT 1a 0.00032 1b 0.00022 2a 0.00030 2b 0.00016 3a 0.00007 • Figure 2.11. Channel gradients along the Peace River based on 1:50,000 N T S maps and Andres (1996). The locations of the principal study reaches and points of interest are indicated. 68 Kellerhals, 1979) and is creating a stepped profile along the main stem (Church, 1995). In B.C. , Church et al. (1982) found that the channel has been static following regulation, hence, is no longer alluvial. The effects of regulation are proportionally greater near the dam with decreasing unregulated tributary influence. Since regulation, inundation of the floodplain by open-water flows on the Upper Peace River is rare. A s a result, the old floodplain has become a low terrace and the former bar surfaces have become the contemporary floodplain. A t the same time, vegetation has expanded onto the exposed alluvium and succession of semi-aquatic species to terrestrial species is occurring on the old bars. The channel has been narrowing as sediment accretion and bar extension take place in low velocity zones along the bank. In addition, secondary channels are slowly being abandoned and vegetated, which may affect channel sinuosity in the long-term. 2.8.2 Middle Peace River (Reaches 2a and 2b) The middle Peace River, in this study, is between the Smoky River and Tompkin's Landing. This reach consists of two principal study reaches 2a and 2b and 15 sub-reaches (see Table 2.6). The middle Peace River has an irregularly meandering habit, with frequent non-overlapping wooded islands and point bars. Narrow secondary channels separate the islands from the banks. The middle Peace River is partly entrenched and frequently confined within a deep stream-cut valley, with relief approximately 100 m lower than the upper Peace. The valley still exhibits slumping, but this decreases downstream as the valley walls decrease in height and increase in stability. The valley bottom varies between 1,000 and 2,000 m in width, and consists of discontinuous valley flats. Channel width gradually increases from 500 m near T P R to 750 m near Tompkin's Landing. Bed sediment textures range from fine gravel to sand, while the banks are primarily sands and silts over gravel, shale and sandstone. Mean gradients of the middle 69 Peace are 0.00030 for Reach 2a and 0.00016 for Reach 2b. The middle Peace River is relatively stable, however, entrenched loop development is common. The middle Peace River tends to have a high suspended sediment load. A t TPR, the mean annual suspended load is 33.7 M t (Carson, 1992). Ha l f of this comes from the Smoky River, which drains an area with friable bedrock and fine-textured glacial deposits. The headwaters of the Peace River, above the point of regulation, are not a major source of sediment, so the Bennett Dam has had virtually no impact on the sediment budget of the river (Church and North, 1996). Since regulation, the middle Peace River has experienced some growth of islands and bars but there has been little development of new bars (Prowse and Conly, 1996). Vegetation encroachment is widespread and is causing channel narrowing. Several side channels have been abandoned, but to a lesser extent than the upper Peace River. A t the mouth of the Smoky River, an alluvial fan has built out into the Peace River resulting in significant backwater for many kilometres upstream (Church, 1995). 2.8.3 Lower Peace River (Reach 3a) The lower Peace River, in this study, encompasses the reach between Tompkin's Landing and Fort Vermil ion. This reach is comprised of study reach 3 a and includes 5 sub-reaches (see Table 2.6). Sparsely forested lowlands and muskeg flank the river along this reach and there are no significant tributaries. The incised valley continues into the lower Peace, but becomes wider and shallower resulting in the river channel being less continuously confined and relatively isolated from hillslope and colluvial inputs. The channel has irregular meanders and frequently splits around island and bar complexes located at most bends. The bed of the lower Peace is primarily sand, although gravel can be locally significant. A t Fort Vermi l ion, the D50 is 0.31 mm while the D90 is 0.51 mm (Kellerhals et al., 1972). The banks are sand as wel l but have a higher silt content than upstream due to the decrease in f low velocities. Channel widths increase from 750 m at Tompkin's Landing to over 900 m near Fort Vermil ion. The mean channel gradient of Reach 3a is 0.00007. Since regulation, the lower reach has experienced bar and island accretion, mid-channel shoaling, transverse bar formation, and vegetation encroachment. This has resulted in side channel abandonment, which may shift the morphology of the river towards a single thread habit in the long-term. 71 3 . 0 M E T H O D S 3 . 1 R E S E A R C H D E S I G N The first objective of the investigation was to compile a record of major ice jams and ice drives, along the river. Historical data on the magnitude and frequency of these events was fundamental to the study. A considerable body of ice jam information was available for T P R , the main center along the river. Historical data (compiled in Appendix C) was primarily extracted from annual ice observation reports provided by the River Engineering Section of Alberta Environmental Protection and B.C. Hydro. Personnel from Alberta Environmental Protection provided annotated photographs of two major ice jam events at T P R and facts on the general ice regime. This information was supplemented by interviews of residents l iv ing along the river. The museum archives at the T P R and Fort Vermil ion were examined to gather anecdotes from past ice jam events. Unfortunately historical information was heavily biased to near the T P R , leaving large information gaps upstream and downstream. A s ice jams tend to be localized phenomena, it was necessary to gather environmental evidence of ice jams to f i l l the information gaps. Since the distribution and overall significance of ice jams on the Peace River was unknown prior to the investigation, the principal activity of the study was an extensive reconnaissance. The primary sources of data used to infer the frequency and magnitude of ice jams along undeveloped reaches were visible tree scars that lined the banks of the river. In addition, past disturbance events were dated by collecting ages of woody vegetation growing along the river. The nature of the disturbance was inferred by noting the type of damage to the stems (i.e., bent or snapped off) and the presence of other evidence such as silt or meltout deposits. Based on the author's knowledge, this reconnaissance represents the first comprehensive examination of the effects of river ice on 72 the banks of the Peace River. The second objective was to characterize the response of riparian vegetation to ice jams on the Peace River. In addition to tree damage along the edge of mature forest, riparian communities reflect the past history of flooding on the river (cf. Sigafoos, 1964). On the Peace River, riparian communities have been undergoing expansion and succession in response to reduced summer flooding since regulation (cf. Church et al., 1996). However, ice jams may tend to counteract the general successional trends, particularly where they are frequent and severe. A goal in this study was to determine how significant ice jams were in modifying the successional trends. The third objective of the research was to collect morphologic information along the channel margins in order to determine i f a relation existed between ice jam locations and bank morphology. This proved to be a complex issue to investigate since bank morphology depends on a multitude of factors. Additionally, f low regulation on the Peace River over the past 25 years has affected channel morphology. The abandonment of the pre-regulation floodplain and the creation of a new floodplain as much as 2 m lower is one form of channel adjustment that has been observed (Church, 1995). The present study included several tasks, which are outlined in Table 3.1. The primary fieldwork was conducted in the summer of 1995 with supplemental observations in 1994 and 1996. 3 . 2 L A B I N V E S T I G A T I O N Prior to the field investigation, 1:30,000 scale air photos were used to identify locations of probable ice jamming along the study reach. Potential locations were identified on the basis of Table 3.1. Study timeline Dates Tasks Sep 1 9 9 4 - N o v 1994 Develop thesis topic Dec 1994 Observe Peace River during winter at Dunvegan and T P R Jan 1 9 9 5 - M a y 1995 Collect background information; air photo interpretation; prepare for field work Jun 1 9 9 5 - A u g 1995 Conduct river reconnaissance from Many Islands to Dunvegan and from Sunny Valley to Fort Vermi l ion Aug 1995 Attend 8 t h Workshop on River Ice in Kamloops, B .C. Aug - Sep 1995 Conduct river reconnaissance from B.C./Alberta border to Many Islands, and from Dunvegan and Sunny Val ley Oct 1 9 9 5 - N o v 1995 Compile survey data, vegetation data, and prepare tree sections for analysis Jan 1 9 9 6 - M a r 1996 Analyze field data; Travel to Edmonton to gather information on Peace River ice conditions from Alberta Environmental Protection and Tri l l ium Engineering and Hydrographies Ltd. Feb 1996 Observe Peace River during winter conditions between Dunvegan and T P R Mar 1996 Present research findings at Geography Home Seminar Apr 1996 Observe Peace River during spring break-up between T P R and Notikewin River. Ground observations were supplemented with two flights over the river courtesy of Alberta Environmental Protection; present research findings at Geography Spring Research Symposium May 1996 Prepare for second major field trip to Peace River Jun 1996-Ju l 1996 Conduct ground-based observations during flood at several locations from Dunvegan to Fort Vermil ion. River travel was prohibited by high flows. Jul 1 9 9 6 - S e p 1997 Compile photo collection for the Northern River Basins Study; analyze field data Sep 1997 Present research findings at 9 t h Workshop on River Ice in Fredericton, N .B . Oct 1 9 9 7 - A p r 1999 Write thesis text May 1999 Submit thesis for review channel morphology, such as constrictions, sharp bends, tributary confluences, islands, shoals, and locations of abrupt channel widening. Most of the study reach was analyzed during the course of a mapping contract for the Northern River Basins Study (cf. Church et al., 1996). The mapping project entailed the delineation of morphologic and vegetative boundaries from air photos taken during five periods since 1950 along three discontinuous reaches of the Peace River. The mapped and intervening unmapped reaches were the focus of this investigation. During the mapping exercise, the author became familiar with the patterns of channel adjustment and vegetation growth along the river. Locations with suspected ice effects were noted for field examination. However, only a handful of ice-related features were identifiable at the photo scale and resolution. Field observations were therefore paramount in this investigation. 3.3 F I E L D O B S E R V A T I O N S Field observations were conducted during several trips to the Peace River. In December 1994, during a personal trip to northern Alberta, mid-winter conditions of the Peace River were observed and photographed at Dunvegan and TPR. M i l d temperatures in the early winter resulted in a late freeze-up. A s a consequence, river ice at the time was limited to the channel margins and side channels. Primary field observations were made between June and September 1995. A field crew of three persons traveled 665 km of the Peace River between Clayhurst and Fort Vermil ion. The trip was made in a 4.9 m (16 foot) aluminum boat powered by 15 and 20 horsepower outboard motors. Since it was logistically difficult to travel the entire reach with 3 persons in one continuous trip 75 down river, the study reach was covered in sections based on the locations of access points. These points, spaced 30 to 60 km apart, were used as base camps from which daily trips would be made in both the upstream and downstream directions. This method eliminated the requirement for an additional vehicle, but increased the requirement for outboard fuel, since all reaches of river were traveled at least twice. This meant that both banks could be inspected carefully with binoculars from the boat. Camps along the river without road access were established only twice during the field season. River stages and discharges were high in the early part of the field season (see Figure 3.1 and 3.2). This undoubtedly affected the number of minor features observed along lower elevations of the banks, but it is believed that all major features were visible under the circumstances. During the field investigation, an effort was made to gather local testimony on river ice jams. The methods used to record evidence of ice jams and the state of riparian vegetation and bank morphology is presented in Section 3.3.1. The Peace River was visited in February 1996. Mid-winter observations were made at locations between Dunvegan and T P R to observe and photograph consolidated ice conditions, which are common to the upper Peace River since regulation. Vegetation damage and protection by ice was observed. In Apr i l 1996, the author was fortunate to have first hand observations of break-up and ice jam flooding of the Peace River. Ice and bank conditions during break-up were documented and photographed from locations between Shaftesbury Ferry and Sunny Valley. The River Engineering Section of Alberta Environmental Protection provided two aerial reconnaissance DUNVEGAN (07FD003) T - T— CM CM CN CN CN c " I 3- S" i 5 < 5 -> 1995 1996 Figure 3.1. Mean daily stages at selected W S C gauging stations on the Peace River in 1995 and 1996 (from Environment Canada, 1997). The period of primary field observations is indicated. Figure 3.2. Mean daily discharge at selected W S C gauging stations on the Peace River in 1995 and 1996. The periods of field observations are indicated. 78 flights of the river, between Dunvegan and Notikewin River. These flights were part of their annual break-up observations. During the visit, ice jam flooding and bank conditions immediately following the flooding were observed between T P R and Sunny Valley. The second summer visit to the Peace River was made in June 1996, but field observations were curtailed due to high flows released through the spillway at the W . A . C . Bennett Dam. Since boating was deemed unsafe, observations were made from several locations along the river between the Hudson Hope and Fort Vermil ion. Although stages were high, clear evidence of spring break-up was observed and documented. 3.3.1 Vegetation and B a n k Surveys During the river reconnaissance in summer 1995, effects of ice on riparian vegetation, substrate and bank morphology were documented at 270 opportunistic survey sites with various hydrogeomorphic settings. Sites were chosen based on the evidence of scarred mature trees, widespread vegetation damage, well-defined vegetation galleries, unusual sedimentary features or highly scoured banks. The objectives of the level surveys were to: 1) date and measure the elevations of tree scars and other indicators of extraordinary ice jam stages; 2) determine the elevation, age, species, and degree of ice damage within distinct vegetation galleries; and, 3) describe, measure and map sedimentary and morphologic features related to river ice. Survey sites were generally located every 1 to 2 km along the river. If evidence of ice damage was limited, surveys were spaced approximately 10 km apart. A n effort was made to reduce bias 79 by selecting an equal number o f sites on both left and right banks of straight reaches, on cut banks, point bars, islands, and near tributary mouths. Bank profiles and elevations were measured using a Sokkisha C-3 automatic level (capable of tacheometric measurements of horizontal distance) and a 5 m stadia. Supplemental measurements were made using an abney level and a 50 m fibreglass tape. A t many sites, a strip of vegetation was cleared in order to provide a line-of-sight. Vegetation was cleared by hand using Sandvik bush axes and Swede-saws. This activity proved time consuming on the lower Peace River, because of wide and dense riparian communities. As a consequence, surveys along the lower Peace River were conducted further apart than on the upper reaches. Photographs were taken at all survey sites and other points of interest. A large selection of these photos has been compiled on C D - R O M for the Northern River Basins Study (Carson et al., 1997). Field Datum Since the river level varied during field observations, elevations were referenced to the lowermost edge of continuous vegetation taller than 0.25 m. This field datum was relatively stable and clearly defined along most of the river, not only by the vegetation boundary, but also by an interface between recently "washed" sediments (below) and deposited fine sediment (above). This level is the lower limit for sustained growth of terrestrial vegetation and is approximated by the river stage during the mean monthly f low in June (see Figure 3.3 to 3.6). The stability of the field datum is suggested by the moderately negative correlation between the height of the field datum above the water level and the water level records at two stations within the study area (rDunvegan = -0.60; rrpR= -0.55). In other words, a decrease in the water level on a given day corresponds to an equivalent increase in the height of the field datum above water level (see Figure 3.7). Using the field datum, elevation data are relative values, which are sufficient to indicate the general severity of ice jams and related flooding along the river. At the gauging stations of Dunvegan and T P R , the geodetic elevations of the field datum measured 341.2 m and 312.4 m respectively. Tree Scars The magnitude and frequency of disturbance events, such as ice jams, can be determined indirectly by using botanical evidence [see Alestalo (1971) and Sigafoos (1964) for comprehensive discussions on dendrochronology]. The term dendrogeomorphology was introduced by Alestalo (1971) to describe the application of plant ecology and tree-ring dating to research in geomorphology (Shroder and Butler, 1986). It focuses on site-specific geomorphic activities that are reflected in tree growth. Tree rings are an excellent, but often overlooked source of data, for the construction o f chronologies of geomorphic events. There are two main approaches to tree ring analysis in geomorphology (Shroder and Butler, 1986): 1) Examination from ring patterns of l iving trees that have been damaged by events and survived; and, 2) Determination of maximum age of trees and other plants growing on a surface previously denuded or newly produced by an event. Disturbances, such as flooding, may shove trees over, shear or cause corrasion, bury stemwood, 81 Figure 3 . 3 . View of the lower vegetation limit used as the field datum during the bank and vegetation surveys. Reach: 2b.8 Site: 551 Location: km 554.55 left bank Photo: 28.19a Date: Jul 23, 1995 Figure 3 .4 . View of the field datum and vegetation galleries on the middle Peace River. Figure 3.5. Example of the field datum and trim line on the upper Peace River. Reach: 1b.6 Site: 282 Location: km 870.85 right bank Photo: 14.7 Date: Jul 5, 1995 Figure 3.6. Example o f the field datum along a straight reach on the upper Peace River. Figure 3.7. Mean daily water levels at selected W S C stations and elevations of field datum above the river level during the dates of the field investigation. 84 and expose rootwood. Tree responses to these events include the formation of reaction wood, reduction or termination of growth, development of callus margins, germination of new shoots, and miscellaneous structural changes to the wood character. Disturbance is also reflected in wholesale differences in community structure. The analysis of plant successional patterns, and the ages of the oldest trees growing on each alluvial surface provides a minimum estimate of the time since disturbance. Corrasion scars on trees have widely been used on northern rivers to supplement hydrometric records of ice jam events, particularly in remote areas (see Gerard, 1981; Henoch, 1973; Parker and Josza, 1973; Smith and Reynolds, 1983). Studies conducted by Gerard (1981), Henoch (1973), Reynolds (1976), and Smith and Reynolds (1983) concluded that tree (corrasion) scars, visible along the margins of northern rivers, provide a reliable indication, and often the only evidence of past ice jam stages in remote areas (Gerard and Karpuk, 1979). This is particularly important since conventional stream gauging equipment may be rendered inoperable by large volumes of ice. The basic principle is that ice jams and ice floes wi l l scrape off bark and damage the underlying cambium o f trees adjacent to the river. Damage to the tree might also occur when the rough bark becomes frozen to ice floes that come in contact with the tree. When the ice floe moves, the bark peels off exposing the smooth cambium (Reynolds, 1976). The length of bark removed may be from a few centimeters to a few metres. If the tree survives, it produces scar-tissue or callus-margin growth over the wound and, in some cases, traumatic resin scars or canals (Henoch, 1973). Normal growth continues and, over a period o f several years, w i l l cover injured cambium. Since bark is not renewed, the stage of the event can be measured, and the event can 85 be precisely dated by counting the number of annual growth rings over the scar tissue on a wedge or cross-section of the tree, which cuts across the scar and new growth rings (Jakob, 1996; Smith and Reynolds, 1983) (See Figures 3.8 and 3.9). The date of the event is usually given as two possible calendar years because it is impossible to determine i f the event occurred during freeze-up in winter or break-up in spring, and trees are dormant during ice-covered (i.e., cold) season. Dates obtained this way are thought to be accurate unless the section is incomplete or decayed. For this study, corrasion scars on mature trees proved to be useful since they were ubiquitous along the Peace River, particularly on the upper and middle reaches. They are found on banks at sites of known ice jamming, or at sites a short distance up or downstream from the site of jamming (Sigafoos, 1964). On the lower Peace River, scars along the mature forest fringe can be difficult to identify from the river due to wide and dense shrubs and woodland galleries. In that case, identifying scarred trees must be done by walking along the banks. The primary species displaying ice scars included balsam poplar, aspen poplar, paper birch, green alder, wi l low, and white spruce. In general, the deciduous trees have rings that are more difficult to differentiate and interpret than the coniferous trees. The annual growth of a tree occurs in the form of a cone stacked upon cones of preceding years. Cross-sections of the tree intersect the cones, causing them to appear as rings. Annual tree rings vary in cell type depending on the phase of the growing season. "Ear lywood" cells in conifers are large and thin walled. A s the growing season progresses, physiological conditions within the tree change, such that cells added later in the season are smaller and thick walled. This "latewood" is denser and darker than earlywood and forms a distinct boundary with the following light-coloured earlywood (Shroder and Butler, 1986). Figure 3.8. Ice damaged balsam poplar (upper left), white spruce (upper right) and idealized sketch of ice block abrasion of a tree during an ice jam and ice drive (after Smith and Reynolds, 1983). Figure 3.9. A) Example of an ice damaged balsam poplar after a wedge sample was cut across the scar. In this case the tree was damaged during two events. B) Close-up view of a white spruce wedge sample clearly showing the year when the bark of the tree was abraded by ice. 88 Deciduous trees have less well defined rings because they have either diffuse-porous wood with vessels of uniform size, evenly distributed throughout each annual increment, or ring-porous wood with larger diameter vessels producing the inner or first formed part of the annual increment (Shroder and Butler, 1986). The belief that many scars are the result of ice impact was substantiated during mid-winter and break-up observations by the author (Figures 3.10 and 3.11). However, there was a possibility that some scars were not ice related. For example, it is possible that driftwood floating downstream during floods may cause tree abrasion. Although, damage could occur almost anywhere along the river, sites of impact by logs tend to be localized, often at the heads of islands or locations where the thread of maximum flow velocity impinges on the bank. In the field, locations of tree damage by floating wood are typically marked by piles or pieces of driftwood nearby. A long icy rivers, widespread tree damage is more l ikely a result of moving river ice (Reynolds, 1976). While driftwood can float into the head of islands or other locations, it tends to float with the thread of maximum velocity located off the riverbanks (Bolsenga, 1968). In contrast, during freeze-up or break-up, ice normally occupies the entire channel width. Since the concentration of downstream moving ice is often high, tree damage is common along the banks. In addition, woody debris incorporated in the ice during the rising stages of break-up may be a source of tree damage. Prior to measuring and dating scars, guidelines were followed in order to decrease the possibility that scars were related to processes other than ice disturbance. Firstly, the scars had to be located on the channel-side (i.e., in the path of the likely trajectory of ice movement) of the l iving tree. Secondly, i f scars were caused by animals (e.g., bears, beavers, or cows), disease, wind throw, Figure 3.11. Damage to shrubs resulting from ice thrust during breakup. The photo was taken the morning after breakup took place near TPR. 90 slope movement, driftwood, fire, lightening, or human activity they were not examined (see Jakob (1996) and Schweingruber (1988) for the characteristics of these scars). Thirdly, scars were examined only i f their origin was not in doubt and were located in an area exhibiting several scars or other evidence of ice disturbance. Since disturbance of riparian trees by humans is common near populated centers (i.e. TPR) , ice scars are difficult to positively identify. This precluded measurements near settlements such as TPR. A major problem interpreting tree scars is that their elevations reflect limits of ice shoving and not the peak water level during ice jams. Indeed, this can result in significant overestimation of future flood levels. Although the potential to include scar data from locations where ice shoving occurred cannot be eliminated, it can be minimized by accounting for the location of the tree scar with respect to the channel morphology. During the collection of scar data, a simple assessment of ice shove potential was adopted. Low potential areas included the lee of islands, point bars, and long straight reaches with few obstructions to flow. High ice shove potential existed at island heads, the outside of bends, and locations of f low constriction. When ice scars are plotted on a profile of the reach, anomalies are usually apparent and judgement is used to define a reasonable upper envelope of break-up levels (Gerard, 1981). Although generally reliable, scars can be misleading and should, i f possible, be validated by corroborating field evidence, local testimony, or historical photos and records. Another difficulty in interpreting the scar record is that it tends to be incomplete. Since the process of tree scarring by ice is inherently stochastic, ice stages are not uniformly recorded in space and time. Additionally, the preservation of scars depends on the long-term ice and flood regime of the river. A number of scarred trees are undoubtedly lost to bank failure, beaver 91 activity, or recent ice jams or ice runs. Old scars stand a poor chance of survival unless they were followed by less extreme events. A s a result, many scars identified on the Peace River had dates falling within the post-regulation period. A conservative estimate of the accuracy of ice scar elevations was found by Gerard (1981) to be on the order of 1 m for the Smoky River. On average, scars on the Smoky River were found to overestimate ice jam water levels by 0.5 m (Gerard, 1981). On the Red Deer River, Smith and Reynolds (1983) found that over a 41 year period, recorded peak stages of ice jam floods averaged 1.4 m (range 0.15 m to 3.41 m) below the elevations of tree scars. They noted that the larger the event, the closer the scar height was to the recorded stage. A t each site where a tree was believed to be scarred by ice, elevation measurements were made of the upper and lower edges of visible scars (on the highest trees). The term "vis ib le" is used since there are an unknown number o f trees with scars that have healed over. When this happens a vertical line remains where the growth of the two sides has come together (Smith and Reynolds, 1983). Although these are noticeable upon close inspection, they are difficult to identify from a boat on the river, even with binoculars. The trim line elevation was also measured along the river. The trim line is the upper limit of all vegetation damage, including trimmed branches and roughened bark. The trim line is a result of flood events and ice scour during ice drives (Smith, 1979) and serves as an index of the long-term severity of ice disturbances along the river. This line was well defined along the upper and middle Peace River (see Figure 3.5), but was difficult to define on the lower Peace River because it was often obscured by dense shrubs. 92 Approximately 300 scars found on trees and shrubs were dated by obtaining a wedge or section of wood cut through the scar using a Swede saw or small chain saw. Wedges of new wood and part of the scar (Figure 3.9) were cut from mature trees to avoid destroying them. Cross-sections of tree scars, which provide a more comprehensive means of event dating, were limited to shrubs and young trees. Samples with clearly defined rings and easily identifiable disturbance events were analyzed in the field. Samples with poorly defined rings were labeled and collected for laboratory analysis. Some samples were severely decayed and thus undatable. Wedges and sections provide a more complete picture of disturbance and the rings are easier to count than tree cores (cf. Schweingruber, 1988). Unfortunately, the cutting of wedges and sections is destructive and was kept to a minimum or not conducted in sensitive or restricted areas (e.g., provincial parks and private property). Most samples collected in this study were from moribund deciduous trees or rapidly growing shrub and immature forest species (i.e., Alnus spp.). A t a limited number of locations, wedges were sampled from damaged white spruce (Picea glauca). Unfortunately, the scar record was heavily biased towards lower scars formed during recent events. Wi th increasing time and number of events, there is a greater chance of evidence being destroyed. Trees scarred during pre-regulation events were difficult to identify because they were possibly removed by subsequent events or were healed over. This limited the period of analysis largely to the post-regulation period. Riparian Vegetation Communities In addition to tree scars, evidence of former ice disturbance was collected by dating and surveying riparian vegetation communities. Typically along the Peace River, vegetation is grouped into distinct linear galleries (see Figures 3.3 and 3.4) that occupy relict and 93 contemporary alluvial surfaces (at different elevations) (cf. Church et al., 1996). These galleries provide a record of the past and present environment along the river (Church and North, 1996). Since the vegetation patterns are a result of factors imposed by the river, the vegetation itself provides useful information as to the activity o f the river (Teversham, 1973), including past ice jams. Event frequencies may be established by dating post-disturbance communities, while magnitudes can often be measured by identifying the extent o f vegetation damage or vegetation removal. Eggington (1980) studied the riparian zone along the Mackenzie River and found that size and age of trees do not appear solely related to channel migration or degradation, but rather to ice activity. He stated that similar aged trees at different elevations at different sites were due to the variable extent of ice thrusting; an allogenic process that limits successional progression. In some cases, ice thrusting caused a shift in elevation of the vegetation galleries, and in others "vegetation islands" were left undisturbed. Using tree ages, Eggington was able to construct ice frequency curves for damaging ice thrusting events at specific locations. A similar method was adopted in this study. A n objective of the present study was to determine the extent to which the spatial variation in vegetation communities (e.g., age and species) can be attributed to allogenic processes imposed by river ice. A t approximately 270 sites, with various hydrogeomorphic conditions, the bank was surveyed to determine the bank profile (morphology) and to collect riparian vegetation data (see Appendix D 94 for locations). A t each site, transects were conducted from the water line to the mature forest, on a line normal to channel centre-line. Elevations of vegetation communities and stand heights were measured, and the presence/absence of riparian species were tallied within each community. To simplify data entry, heights were classified as follows: Class 0 <0.25 m (seedlings); Class 1 0.25 - 0.9 m (below waist high); Class 2 1.0 - 1.9 m (waist to head high); Class 3 2.0 - 4.9 m (above head but below stadia height); Class 4 5.0 - 15 m (above stadia); and, Class 5 >15 m (mature forest). Min imum ages of surveyed communities were estimated by counting tree rings from wi l low, alder and balsam poplar sections. The ecesis period for these species is often between 3 to 12 months (Scrimgeour et al., 1994) following a major ice jam flood. Wi l lows are hardy species and have extensive root networks that ensure future growth following extensive scour and removal by ice or deposition of sediment. In some cases, ice shoved trees react by sprouting new shoots which grow vertically. Figure 3.12 shows an example of four shoots growing from a balsam poplar that was shoved over in 1974; one was cut for dating purposes. Within each vegetation community, 5 to 10 sections were cut from the largest and presumably oldest shrubs and trees. Many sections from shrubs and small trees exhibited scars of recent events. For the most part, scar and germination dates were determined in the field. Bank profiles were surveyed at each site where ice scars or vegetation communities were investigated. A l l major slope breaks were surveyed to determine the elevations (and any downstream trends) of several alluvial surfaces along the Peace River. It was hypothesized that Reach: 2b. 7 Site: 525 Location: km 574.9 left bank Photo: 30.22 Date: Jul 26, 1995 Figure 3.12. Example of a mature balsam poplar tree that sprouted four shoots after being shoved during an ice jam in 1974. A section was cut on the left for dating purposes. 96 the elevation difference between relict (pre-regulation) and contemporary (post-regulation) floodplain surfaces is reduced at locations of frequent ice jam flooding. Additionally, ice-related sedimentary features were inventoried with descriptions and measurements for comparison with other rivers described in the literature. 3.4 A N A L Y S I S O F D A T A 3.4.1 Ice Scar Data Approximately 200 of the 300 tree sections collected in the field were brought back to the laboratory for dating. Sections were sanded using a belt-sander with 60 grit sandpaper. This was followed by hand sanding with progressively finer grit paper (up to 300 grit) until a smooth surface was obtained. Fol lowing the sanding, an oil-based wood penetrating stain was applied to the surface of each section to enhance ring visibility. Dates o f tree germination and ice scarring were determined by careful tree ring counting using a platform stereo microscope with up to 40 times magnification. Ring counts were repeated at least three times for each sample in order to reduce the chance for error. Error bars on individual events are uncommon in tree ring studies indicating that the method is founded on its replicable precision (Jakob, 1996). However, i f a section was decayed and the confidence in ring counts was reduced, an estimate of error was assigned. Ice scar elevations and dates representing ice jam events, and trim-line elevations representing long-term ice jam disturbance along the river, were compiled in a spreadsheet in order to determine i f river-length trends existed. Min imum and maximum scar elevations collected in the field were reduced to mean values to be consistent with Reynolds (1976) and Smith and 97 Reynolds (1983). Ice scars were grouped by sub-reach, and the highest scars found for each sub-reach for each year (i.e., ice jam event) were extracted. This space for time substitution was assumed valid since the ice conditions within each sub-reach were believed to be homogeneous. However, in order to have a sufficient record for frequency analysis, ice jam dates and elevations were also grouped on a reach basis. Such a grouping may not be strictly valid since the homogeneity of ice conditions within a reach is questionable. Only the higher scars found on mature trees were considered to mark extraordinary events. Numerous scars at lower elevations are formed each year during typical freeze-up stage increases and are mainly confined to shrub or woodland galleries consisting of young deciduous species. The locations of the higher scars were plotted on reach maps to determine their spatial distribution. Assuming the scar record adequately reflected the ice jam events within each study reach, four frequency distributions (Normal, Log-normal, Pearson III, Log-Pearson III) were fitted to the data using Weibul l plotting positions with the hydrologic analysis program Hydrotec (Greenland Engineering Group, 1997). Based on a visual assessment of fit (since statistical tests were not conclusive with such a small data set), the return period estimates were calculated by averaging the results of each distribution. The 5% and 95% confidence limits of each distribution were also determined. 3.4.2 Historical Data In order to reconcile the magnitude and frequency of ice jams, based on tree scar evidence with the ice jam history recorded at T P R (and to a lesser extent at Fort Vermil ion), historical ice jam 98 stage records were analyzed. A n accurate comparison of botanical and historical data is difficult because very few ice scars were actually measured and dated near the two towns during the study. A t T P R there was limited evidence of tree scars since bedrock cliffs line much of the river upstream and downstream of the town. The possibility of human damage to trees was high near T P R (e.g., during dyke building) and much land bordering the river was private property. Thus, dating was often not feasible or possible. Regardless, a cursory examination was made of the botanical evidence of ice jams and historical records at TPR. The comparison was made to find out i f the dates, magnitudes and frequencies of ice jams were similar along the river or not. For the comparison, flood records at T P R were converted to elevations above the field datum (field datum at T P R = 312.4 m). A t this elevation, the open-water discharge is approximately 2,500 m 3/s. Although the geodetic elevation of the field datum was also known at Dunvegan, the record of ice jam events was incomplete at that location. For the TPR, a frequency analysis was conducted for open-water (pre- and post-regulation) and break-up stages. Annual open-water peak daily flows from H Y D A T (Environment Canada, 1997) were converted to stages by using a series of rating curves provided by the Water Survey of Canada. Break-up stages were determined from ice observation reports and other sources (see Appendix C) . Weibul l plotting positions were used and data were plotted assuming a log-normal distribution. A distinct discontinuity was evident between thermal break-up events and mechanical break-up events. Subsequently, a frequency analysis was conducted exclusively on mechanical break-up stages (i.e., major ice jams at TPR) using the same method as for the ice scar record. 99 3.4.3 Vegetation Communities Elevations above field datum and ages of riparian vegetation communities were extracted from the field survey data and entered into a spreadsheet in order to facilitate sub-reach comparisons. Elevations of vegetation communities along the river were filtered by 20 km running means with a step length of 5 km. Plots were made to see i f spatial patterns of vegetation coincided with ice jam intensity as determined by the scar record. However, while the scar record was representative of the ice/water levels during major ice jams, the riparian community structure and ages reflect not only major ice jams, but may also reflect ice shoving and open-water floods. Based on vegetation transects of age and elevation, return periods of ice disturbance were estimated for each sub-reach by assuming tree age reflected the time since last disturbance (cf. Eggington, 1980). 3.4.4 Bank Morphology Survey data collected in the field were entered into a spreadsheet in order to plot bank profiles. These were used to determine the elevation of alluvial surfaces and the variation in height between the contemporary and relict floodplain along the river. Evidence of ice scour and ice-jam sedimentation was found at a number of levels. A goal of the bank surveys was to determine i f there was some relation between the observed ice effects and bank morphology. The locations of ice scoured morphologies identified in the field were plotted on base maps to determine whether they were widespread or localized. Sedimentary features associated with ice were described along the river with occasional measurements at prominent features. The review of these features is largely descriptive. 4.0 R E S U L T S 4.1 I C E J A M L O C A T I O N , M A G N I T U D E A N D F R E Q U E N C Y F R O M T H E S C A R R E C O R D Most previous studies using tree scars to infer ice jam events have investigated single sites or short reaches of river. A goal of this study was to take scar dating one step further in order to determine the magnitude and frequencies of ice jamming at the river-length scale. Observations made during and following break-up in 1996 confirmed that ice could shear banks and uproot and abrade trees (see Figures 4.1 and 4.2). However, it was also observed that moving ice could f low downstream within the confines of stranded ice along the banks, and thereby not damage trees. The occurrence of stranded ice and an offshore shear surface depends on whether break-up occurs on a fall ing or rising stage. In the upper Peace River, break-up is often on a fall ing stage, since f low releases early in the ice season result in freeze-up at elevated and often peak levels (see Section 2.7.1). Therefore, along the upper Peace River, stranded ice is common and ice damage tends to be limited. Since 1960, there have been six major ice jam floods at TPR, which is situated roughly in the centre o f the 655 km study reach. Little information on ice jams elsewhere along the river was available prior to the study. Below is a summary of the locations of major ice jams, their magnitude (i.e., stage), dates and frequency of occurrence as inferred from the tree scar record. 4.1.1 Locat ion and Magni tude of Ice Jams Table 4.1 presents the elevations of tree scars and trim lines at 208 sites along the Peace River. A l l elevations given are in metres above the field datum (see Section 3.3.1), which approximates the level o f the mean monthly f low in June. Data were compiled by sub-reaches, which 101 Reach: 2b. 3 Site: 440.5 Location: km 673.35 left bank downstream of Notikewin River Photo: 66.38 Date: Jun 27, 1996 Figure 4.1. Extensive shrub damage and removal by the 1996 spring ice jam and ice run near Notikewin River. Parallel linear grooves carved by moving ice floes are clearly visible. Note a two metre scale in the foreground. Reach: Site: Location: Photo: Date: 3a. 5 734.5 km 420.75 right bank near Fort Vermilion bridge 68.1 Jun 29, 1996 Figure 4.2. Ice-damaged alder community near Fort Vermilion. Spring 1996 break-up caused exensive damage along the bank and minor sedimentation on the floodplain. i r 2 J? I 6. o 3 3 a 3 D. O 3 1 a. 8-3 vs 8 e. o 3 1^ O N ON O N U l U t O 00 ON U ) - O S ) 00 vO v© N O vo N O \0 vO •O -O -O 00 00 00 vo O i , M OO - W si p ON O 4 » ' A N O — — o\ ui ui ui io j i j v b o ^ P° ^ * u ON ON O N vo io bo • J U l i i O I O - i O M U l M O M ' M " - VO !— to U J Ov - J ^ b\ ~ io a ui to A ^ bo io U J Ul io. bo ^ U J w Ui bo vo O N U I ' O N O N U I U I O N U I L T I U I O N U I - O HbobuiijLAwboL)b4>bob 8 99.5 NO J i UJ VO VO o — (O to UJ UJ iO io Ul io UJ iO s J l o -ON Ui U> to vo — O O O N ^ b W l / i W l / i U ) v D W W l o « - • J ; -">• p N O O ^ J U> Ul J ^ . Ul W ON Ul 0 0 > J i > » ON k ) 0 0 to UJ to UJ to Ul IO Ul o to ON o to ~J o UJ o UJ Ul Ul b bo 6 i-J io ~ ON ON Ul Ul ON 4 » to i ON b Ul W ON ; 0 Ul U I - 0 Ul •O Ul Ul Ul J > J i . N O LA V O U i ^ v O v O C N ^ i / i ^ L A w w W U l w 2 2 S O vo — w O N bo ON U\ W bo to W K> tO tO — — O 0 0 ON 4 ^ to O O O O O O Q O O ^ • v J M v O O v O W W W O W v O a b o M O N ON ON U I W ON U I ON ON vO ON vO LA ON A U J 0 0 A Ul LA to to N J U s i to in to C/3 S C T 3 M 3- D S 2 3 s 2 3 U J U J U ) U J N O N O 00 Ul U J Ul Ul -o )^ -o -o K l U 4 v & oo U J o N O b ki ij t O O U l Ul to io ~ 1 ON ON N O U I O ON U J U J A U J — IO 00 00 3. tn U J A cc oo to vo 00 vo Ul O N w w w W tO O Ln oo 00 00 00 to U J - U •— ON VO Ln O N b ^ vO O N 00 Ln to Ln U J U J U J to to S O O vo N O - J o oo 00 00 00 00 00 f A Ui A Ui 5 ON p N O — b ij b b M ON O N ~ J Ul to bo V O M O i * bv O N tO tO K) NO vO 0 0 ON to vo 0 0 0 0 Ln O N O N y i — ^ w o o Ln Ln ON — vo bo •o «o - « vO t o t o t o t o t o t o t o t o 0 0 0 0 0 0 0 0 0 0 0 0 - J - O 0 0 O N U l J > t O © W — 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 U > r> f > s i > j N | s j NO vo O O to ON L n t o b o w v o b L n b o v i ^ o o y i s i o v u i o v boLn^— w b » — t o b NO VO NO NO — to to to Ln w tn -o b vO W — ON -O. s J - O 0 0 oo -o U J O N ~ '.£». to to to to to to to w w w w to to " " OO tO — O v O W NO vO vO vO vO vO vO to w w w w w w s ) •— to W W 0 0 0 0 NO vo bo b bo. b to Ln ON s i Ln -O. LA 0 0 Ln ON tO b tO 0 0 tO to — O vo vo — w to vO vO vO Ln Ln Ln Ln Ln ON — Ln Ln 4 ^ r 4 oo vo t o O N Ln to bo oo -o -o -o • U vo oo ON Ln . U N O V O V O V O V O V O V O V O V O V O O N O N O N O N O N O N - ^ J - O - O ^ J p t o w u n O N v o p — — " w ^ O N t o t o b o w L n O N -o Ln A ON ON J > J>» ON W v© bo vo U -S 2 < s U 3 b o v o b o ^ L n O N O N O N — Ln Ln Ln Ln Cn Ln Ln Ln Ln Ln Ln • - ' oo Ln 4 > to ,- - - _ . - . - . u i U t L n u i L n U i L n n L n L n L n L n j > . ^ k j > ^ w w w w w N O ^ W * — v O O N A t O — Ul 4 * Ul Ul Ul Ul Ul Ul Ul Ul Ul Ul Ul Ul ui ui Ul O N Ul O N Ul ON Ul NO to 4 * O N 00 NO p UJ O N 4». ON iO UJ iO O N UJ 00 bo iO w w j > . O N f o O N L n w w p o t o j > i W t o J > w L n L n j > b b o b N L n — v o b w ^ - j W W Ln O N to O N ON W W VO W J> W to I f l U l l f l U l U i U i U U i U i U U ^ W - 0 0 0 - O L n W — W U U l W U i U i U l U l l / U i N I S I M O O O O M O O O O O O V O w — k ) w b o b o v o J > i t o O^ I O v^ vO ON W — Jt* p\ U VO 00 W to w Ln Ln Ln w -o o b\ ^ b c7\ v i l ^ j > . w w v o L n L n w o o O N W L n b b o ' « t o ^ j ! i > b o J > 4 ^ J > NO NO vO -J -O UJ o -o L n to ON o ON ON ON ON O to to to to ^ 0 J > w ON ON vO w bo NO ON w J ^ w ON bo Ln bo ON w L n w •o Ln ON NO b ON Ln ON OS • U 4 ^ . ON ON L n W — - O 2 2 2 to Ln vo vo vo b w *o Ln W oo Ln Ln — ON ON ON ON ON ON Ln ON ON ON ON -O N) - A Ui OO 0\ vO W W ON ^ - Ln 0 0 N) tO ON to ^-J b U J OO t o t O ON VO Ln v O ON — c o* J > . j > j > J > . j > . j > , j > i U J O N O N O N O N - O - O - O - V J 0 0 0 0 0 0 0 0 0 0 " - t O - U l s l O O - S O -- J ON J-^ ON Ji. W M 0 0 bo bo vo bo p Ln Ln Ln 1? R' .a. I W v O ON -O tO U I to W S 2 a. CL O < " J ^ J ^ J ' — J ^ J ^ J ^ I ^ J — l - v j —J — l - ^ J 0"N C~N U J t O M S J t O N J S J — — — O O O N & N O - N O O l s l U t o O ^ f t W O N U l O f c U w w ^ v o l ^ L n ^ w o N W O N L n b ^ b W W J > O N t O W W W - U JO W N -w w Ln ON w to W W Ul -0 w w Lo ON U J b — ^ J > J > j>. b to b to W « w O N •— b\ vo bo Ln A 4*. O N O N O N O N O N O N O N O N O N V O O O - O - V J O N O N O N U I J ^ to 0 0 J > O Ut w — t o v o O N O o b L n w b o b o w b w O N W w w w to W W W U l U i U i W W t O t o L i !u v i — w v o b o b o W W W W W W O N U I U I A O N ^ W J ^ U J O N O N O O U I — ON ON ON ON ON O N ON to to to SJ SJ ON Ul UJ O 0 0 ^ 0 *. J> u. •u 41. 4 0 NO NO NO NO NO NO NO O — S) UJ Ui b i o Ul ON 0 0 b IO SJ ~J UJ Ul — i c bo io UJ i o bo O N O N O N U » O O U \ U I U I Q O p vo OO OO * vO ON 4^ - vO • Ui — O O O O — — t O t O t O W W 0 0 LA ON — W N O N O - U I U I O -ON U J U J to w vo to bo 0 0 — ON ;_n w w vo — J s * ON U - 4^ , bo • W 4 ^ W ON - O b to bo to bo Ui Ul Ul Ul Ui Ul - O - J ON O N O N ON U I -C*. ON Ul J^. W to u ! UJ £ £ £ 0 0 O vO — W W to b Ln Uj ™ s l W to to Ui to bo vo Ln Ln b 0 0 A tO tO Ul K) — bo Uj L A ON hi 3 2 2 " S < SJ U 3 Si V> § 2 2 3. S 3 103 are expected to have similar f low and ice conditions due to similar channel morphology. The locations and elevations of all scarred trees identified are shown on 1:100,000 scale study reach maps in Appendix D. To illustrate the general distribution o f evidence collected along the Peace River, Figure 4.3 shows the locations where tree scars were found at elevations: 1) 6.0 - 6.9 m, 2) 7.0 - 7.9 m, 3) 8.0 - 8.9 m, and 4) greater than or equal to 9 m above field datum (afd). Scars above 6 m are expected to represent extraordinary ice jam events. The 6 m elevation above field datum at T P R has a geodetic elevation of 318.4 m. This elevation approximates the mean of the recorded stages of major post-regulation ice jam events at the T P R (see Figure 2.8). Figure 4.3 indicates that ice levels above 6 m have occurred along the entire length of the study reach. Considering that scar elevations may overestimate peak ice jam stages by about one metre (Gerard, 1981), an estimate of the water levels represented by these scars is probably closer to 5 m above field datum. Such overestimation l ikely applies to the other scars as wel l , but, as Smith and Reynolds found, lessens the higher the ice jam event. It is not surprising that scars are widespread above the 6 m elevation, especially in the upper Peace River. Consolidation of ice during freeze-up typically results in winter stage increases on the order o f 5 m (Andres, 1996). Within the upper Peace River there were three locations with scars in the 7.0 - 7.9 m range: upstream of Dunvegan Bridge, near Saddle River confluence, and directly upstream of the Smoky River confluence. These locations fit the classical criteria for ice jam locations. Upstream of Dunvegan Bridge, there is a 90 degree left bend in the river with 104 MEAN ELEVATION OF TREE SCARS ABOVE FIELD DATUM • > 9.0 m • 8.0 - 8.9 m ® 7.0 - 7.9 m : 6.0 - 6.9 m 20 Kilometres 40 B.C. / Alberta border \Dunvegan 9 Q Q 950 Ft. Vermilion bridge Town of Peace River Smoky River Figure 4.3. Locations of trees with ice scars greater than 6 m above field datum. Scar elevations above field datum are classified into four groups: 1) 6.0 - 6.9 m, 2) 7.0 - 7.9 m, 3) 8.0 - 8.9 m, and 4) greater than or equal to 9.0 m. 150 other scars below 6 m are not shown because of map scale (see Appendix D). See Section 3.3.1 for discussion of field datum. Distances indicated are in kilometres from Slave River. 105 a point bar extending into the channel. A t the Saddle River confluence, the river flows through a series of tight bends with prominent side channel and mid-channel bars. Near the Smoky River confluence, ice jams are initiated by the large ice volumes moving down the Smoky River. A secondary factor responsible for ice jams at this location may be the reduction in channel gradient upstream of the confluence that has resulted from the growth of an in-stream alluvial fan at the mouth of the Smoky River. Growth of this alluvial fan is partly a result of reduced summer flushing of the Peace River following f low regulation (Church et al., 1996). Ice jam levels, inferred from scar record, are the highest along the study reach in the middle Peace River, roughly between T P R and Carcajou. A long this northerly f lowing reach, scars above 8 m are clustered along three reaches, two of which had scars above 9 m. The first cluster falls within sub-reaches 2a.4, 2a. 5, and 2a. 6 located between the Daishowa pulp mi l l and Sunny Valley. The second includes sub-reaches 2b.2 and 2b.3 near Notikewin River, and the third cluster of high scars is located in sub-reaches 2b.7 and 2b.8 near Carcajou. The channel near the first scar cluster is generally straight with frequent side-channel bars and islands, but has the occasional sharp bend and wide point bar. The tight channel bend at 30-Mi le Wel l , near km 800, is confined and may be partially responsible for the three high scars identified upstream of km 800. Another scar in this cluster is located upstream of Tar Island near k m 790. Tar Island and a series o f extensive bars downstream represents a significant barrier to ice floes and may be a contributing factor to high ice levels in that area. Near Notikewin River, high scars are likely related to ice jamming in a reach with numerous overlapping islands and bars. In addition, near Notikewin River, there is a marked change in channel gradient of the Peace River (Figure 2.11). Much of the channel near Notikewin River is 106 confined as well. The cluster of high scars near Carcajou represents lower ice jam levels than the first two clusters discussed above. The channel near Carcajou is very sinuous (Figure 2.10) and is marked by large, mid-channel and point bars. However, the wide floodplain on the left bank provides relief for floods. The scars found on the lower Peace River were all below 8 m. Similar to Carcajou, the channel is highly sinuous, but has a wider floodplain, which acts to reduce peak ice jam flood levels. In order to display the longitudinal trends in scar elevation, all scars measured are plotted in Figure 4.4. The trim line data, which tends to mirror the scar record, but at a slightly higher level, is shown in Figure 4.5. On these figures, the ice scar and trim line envelopes for the river are plotted by eye. These envelopes indicate the upper limits of ice damage resulting in scarring (Figure 4.4a) or branch trimming (Figure 4.5a), regardless of date. This plot is not definitive and no frequency of occurrence can be attached to it. It does provide an indication of the variation of peak ice jam stages along the river. Most notably, the envelopes peak in the middle Peace River near Notikewin River at an elevation of approximately 10 m for scars and 11 m for the trim line. Both the upper and lower Peace River tend to have lower scar and trim line levels (by as much as 3 m). Since scar and trim line data exhibit considerable scatter, running means were calculated for 20 km lengths every 5 km (Figures 4.4b and 4.5b). This plot indicates a general increase in ice jam severity in the middle Peace River, although peak levels tend to be located further upstream than Figure 4.4. Scar elevations along Peace River: a) ice scar envelope of all ice scars on mature trees; b) 20 km running means (calculated every 5 km) of all ice scars on mature trees; and, c) cumulative percentage deviations from the overall mean ice scar elevation (5.0 m) (see text for discussion). Elevations are referenced to the field datum (see Section 3.3.1). Figure 4.5. Trim line elevations along Peace River: a) envelope of trim line elevations; b) 20 km running means (calculated every 5 km) of all trim line elevations; and, c) cumulative percentage deviations from the overall mean trim line elevation (5.8 m) (see text for discussion). Elevations are referenced to the field datum (see Section 3.3.1). 109 the peak for the ice scar and trim line envelopes. However, this curve tends to be biased by the small sample of high scars between km 760-820. Another index of overall ice jam levels is presented in Figures 4.4c and 4.5c. In these two figures, scar and trim line elevations were plotted as cumulative percentual deviations from the overall mean scar elevation (5.0 m) and overall mean trim line elevation (5.8 m) respectively. A flat curve, regardless of position on the plot, indicates that the elevations are similar to the mean. A descending curve indicates elevations consistently below the mean, while an ascending curve indicates elevations consistently above the mean. The slope of the curve is an indication of the degree to which elevations differ from the mean in the downstream direction. For example, a steepening of the curve indicates that scars were measured at elevations of greater and greater differences from the mean. Long-term ice jam severity, as represented by Figures 4.4c and 4.5c, is slightly above the study-length mean between the B.C./Alberta border and Many Islands (km 1000). This reach is characterized by several overlapping islands and large bars within a confined valley. Downstream of Many Islands to a point downstream of Montagneuse Islands (km 990), ice jam severity is below mean levels. Downstream of km 990, ice jam stages are above the overall mean level until km 650, approximately 30 km downstream of Notikewin River confluence. A notable "step" on Figures 4.4c and 4.5c at the Smoky River confluence, suggests that ice jam levels are substantially increased downstream of that tributary. Historically, ice jams have been triggered by the break-up of the Smoky River, so increased levels downstream of the Smoky are to be expected. From km 650 to km 580 near Carcajou, long-term ice jam intensity is about average for the study reach. Between Carcajou and Fort Vermi l ion, ice jam severity is below 110 mean levels. The river-length trend in severity suggests that the Smoky River confluence and channel morphology are major factors influencing ice jams along the Peace River. The ice scar trend also reveals a relatively low level of damage in the upper study reach, which is frequently covered by transient ice since regulation and the lower study reach, which has a wide valley flat. 4.1.2 Frequency of Ice Jams Ice Scar Record A total of 188 scars on mature trees were measured and dated within the study reach. Since no one location had a complete scar record, the information was compiled on both a sub-reach and a reach basis. Table 4.2 lists the highest scar identified in each sub-reach for all ice jam years identified in the scar record. To provide a more complete set o f ice jam information on the lower Peace River, data collected by Gerard (1979) were included; see Table 4.2 for a list of ice scars near Fort Vermi l ion. To confirm the accuracy of the data extracted, the author inspected the original scar samples, which are stored at the University of Alberta. Gerard's scar heights were referenced to the 2-year summer flood stage, which does not necessarily equal the field datum at Fort Vermi l ion. Note that the field datum was not precisely defined in the Fort Vermi l ion area (downstream of the study reach) so elevations presented in the table for Fort Vermi l ion should be read with caution. Since Gerard focussed his attention at one location, he was able to scrutinize the forest more closely for evidence of ice damage. Many of the ice scars he identified (i.e., healed over) would l ikely be missed during a river-length reconnaissance. Table 4.2. Dates and elevations of the highest annual tree scars for each sub-reach. River Elevation River Elevation River Elevation Sub- distance above field Sub- distance above field Sub- distance above field reach (km) Year datum (m) reach (km) Year datum (m) reach (km) Year datum (m) la.l 1074-1055 1948 5.2 2b.3 677-658 1963 8.7 3a.3 500-486 1979 5.8 1955 5.2 1982 8.7 1987 4.1 1967 4.4 1986 5.8 1990 3.9 1973 5.2 1990 5.8 1992 4.2 1982 5.5 1994 4.9 1994 4.2 1991 5.5 2b.4 658-636 1990 7.8 3a.4 486-442 1920 3.3 la.2 1055-1038 1955 3.9 1994 5.2 1973 5.1 1974 3.9 2b.5 636-618 1994 6.8 1980 5.7 1982 6.7 2b.6 618-592 1963 6.8 1986 3.2 1984 5.2 1974 6.3 1987 2.8 1992 6.7 1975 8.0 1990 3.9 1993 3.9 1980 4.8 1991 3.2 la.3 1038-1019 1968 6.4 1994 6.1 1992 1.2 1979 6.2 2b.7 592-568 1974 5.4 1994 5.3 1985 6.3 1979 8.2 3a.5 442-420 1965 6.3 1987 6.2 1980 6.8 1971 3.8 1990 4.5 1981 4.2 1979 6.3 1993 4.5 1982 7.7 1988 6.4 1994 6.2 1985 . 7.7 1990 3.4 la.4 1019-1006 1988 3.5 1987 4.0 1991 4.0 1992 4.5 1988 5.6 1992 3.4 la.5 1006-991 1985 4.5 1989 5.0 1994 3.8 1990 5.8 1990 5.4 Near Fort Vermilion' 1993 5.4 1992 3.6 1918 6.6 la.6 991-974 1987 5.0 1993 2.1 1934 5.8 1990 5.4 1994 3.1 1938 7.8 1993 5.4 1995 5.6 1939 5.0 la.7 974-957 No dates collected 2b.8 568-547 1965 4.5 1944 7.8 la.8 957-943 1992 5.8 1975 6.0 1945 3.7 lb.l 943-928 No dates collected 1976 8.1 1948 4.9 lb.2 928-915 1978 7.4 1980 5.3 1954 6.5 1982 7.3 1981 5.8 1955 4.8 lb.3 915-906 1978 6.3 1984 4.5 1960 6.4 lb.4 906-892 1987 4.9 1987 3.3 1962 2.2 lb.5 892-878 1979 6.7 1987 3.3 1963 7.0 1992 6.7 1988 3.6 1966 1.8 lb.6 878-866 1981 5.1 1989 4.0 1969 0.8 1982 4.4 1990 3.2 1970 5.5 1992 4.4 1991 4.0 1971 4.5 2a. 1 866-849 No dates collected 1992 3.4 1973 4.6 2a.2 849-840 No dates collected 1994 2.8 1974 1.6 2a.3 840-820 No dates collected 3a. 1 547-526 1978 6.9 1975 2.0 2a.4 820-799 No dates collected 1985 7.8 1976 1.9 2a. 5 799-774 No dates collected 1987 2.9 2a.6 774-750 No dates collected 1992 5.0 2a.7 750-725 No dates collected 1994 5.7 1982 7.2 3a.2 526-500 1953 7.1 1994 5.7 1974 3.6 2b. 1 725-694 1979 6.7 1983 6.9 1981 6.7 1987 4.1 1990 3.4 1988 3.6 1994 1.9 1990 4.7 2b.2 694-677 1990 4.8 1992 3.6 1991 3.8 1994 6.3 1994 2.9 Notes: 1) Scar elevation and dates for an unspecified area near Fort Vermilion were extracted from Gerard (1979), and were confirmed by inspecting the orginal tree sections collected in the field (stored at the University of Alberta). Analyses of flood return periods historically have relied on simple discharge records, which do not reflect the important backwater effects of river ice, hence are inadequate for estimation of i « jam stages (Prowse, 1994). Floods produced by ice jams are expected to have a highly variable frequency distribution (Church, 1988). In order to obtain a first approximation of the frequency of ice jamming along the Peace River, the scar record was analyzed on a reach basis, see Table 4.3 and Figure 4.6 for the data analyzed. Since the sample of scars in Reach 2a was limited, analysis was not conducted for that reach. Because the ice regime is known to be affected by regulation (the greater effect being noticed upstream), ice jam frequency analysis was conducted for events dated within post-regulation period (1973 to 1995). It should be noted that a number of statistical criteria were potentially violated in this exercise: 1) the events may not be entirely random due to flow regulation; 2) the events may not be homogeneous since the scar record does not distinguish between freeze-up and break-up jams; and 3) the ice jam events in the last 25 years may not have been stationary (i.e., there is the possibility of a climatically induced trend). The reliability of estimates is also affected by the serendipitous nature of the ice scar evidence and associated sampling errors. The elevation and return period of ice jamming for four of the principal reaches are plotted in Figure 4.7. Each column on this plot represents the average of estimates from four frequency distributions: Pearson III, Log-Pearson III, Normal and Log-normal. The error bars indicated are the 5% and 95% confidence limits. Reach 2b has the highest magnitude events. However, based on the small sample sizes, the error bars are wide and differences among reaches may not be statistically significant. 113 Table 4.3. Dates and elevations of highest annual tree scars for each principal reach. Reach Year Elevation above field datum (m) Reach Year Elevation above field datum (m) Reach Year Elevation above field datum (m) Upper Peace River Middle Peace River Lower Peace River la 1948 5.2 2a 1982 7.2 3a 1920 3.3 1955 5.2 1994 5.7 1953 7.1 1967 4.4 1965 6.3 1968 6.4 2b 1963 8.7 1971 3.8 1973 5.2 1965 4.5 1973 5.1 1974 3.9 1974 6.3 1974 3.6 1979 6.2 1975 8.0 1978 6.9 1982 6.7 1976 8.1 1979 6.3 1984 5.2 1979 8.2 1980 5.7 1985 6.3 1980 6.8 1983 6.9 1987 6.2 1981 6.7 1985 7.8 1988 3.5 1982 8.7 1986 3.2 1990 5.8 1984 4.5 1987 4.1 1991 5.5 1985 7.7 1988 6.4 1992 6.7 1986 5.8 1990 4.7 1993 5.4 1987 4.0 1991 4.0 1994 6.2 1988 5.6 1992 5.0 1989 5.0 1994 6.3 lb 1978 7.4 1990 7.8 1979 6.7 1991 4.0 Near Fort Vermilion1 1981 5.1 1992 3.6 1918 6.6 1982 7.3 1993 2.1 1934 5.8 1987 4.9 1994 6.8 1938 7.8 1992 6.7 1995 5.6 1939 5.0 1944 7.8 1945 3.7 1948 4.9 1954 6.5 1955 4.8 1960 6.4 1962 2.2 1963 7.0 1966 1.8 1969 0.8 1970 5.5 1971 4.5 1973 4.6 1974 1.6 1975 2.0 1976 1.9 Notes: 1) Scar elevation and dates for an unspecified area near Fort Vermilion were extracted from Gerard (1979), and were confirmed by inspecting the orginal tree sections collected in the field (stored at the University of Alberta). UJ UJ - i rr UJ REACH 1a 1910 1920 1930 1940 1950 1960 1970 UJ ^ 23 UJ UJ UJ 10 8 6 4 2 REACH 1b O ° O n u_00| O W 1980 1990 1910 1920 1930 1940 1950 1960 rW-1970 UJ „ O ~ 5 1 ° l p £ 3 UJ UJ -J i i tu 10 8 6 - 2 ll *• REACH 2b 1910 1920 1930 1940 1950 1960 1970 10 UJ 8 > O c AB UM 6 z l -o < 1 - a 4 < a > _ i UJ UJ _l UJ u. 2 • p REACH 3a (solid squares); FORT VERMILION (open squares)" 1980 1990 1980 1990 1910 1920 1930 1940 1950 1960 1970 1915 t t t f t t t 1934 1948 1980 1990 1963 1973 I 1979 1974 1982 1992 Figure 4.6. Highest tree scars dated for each year within each principal study reach. Reach 2a is not included due to small sample size. Data for Fort Vermilion were extracted from Gerard (1979). Note that elevations of Gerard's data is referenced to the 2-year summer flood level, and may not be directly comparable to the field datum used in this study (see text for discussion). Arrows indicate years of major ice jams at the Town of Peace River. Figure 4.7. Estimated ice jam stages for selected return periods for four of the five principal study reaches and the TPR. Reach 2a was not assessed due to the small number of tree scars dated in that reach. Return period estimates for ice jams at T P R are based on historical records. Study reach estimates were based on all events occurring within the post-regulation period (see Table 4.3 and Figure 4.6). Error bars represent the 5% and 95% confidence limits. 116 Historical Records A lengthy record of ice conditions for the Peace River is available at the T P R (see Appendix C). A frequency analysis of ice break-up stages was conducted on all available data. On Figure 4.8, these data were plotted assuming a log-normal distribution. Open-water peak stages recorded at the W S C station at T P R were also plotted for comparison. Best-fit lines were drawn through the data by eye. These results show that break-up stages are more severe than post-regulation open-water events above a return period of 1.5 years. A s expected, a noticeable discontinuity occurs between thermal and dynamic break-ups (i.e., major ice jams). The results o f a frequency analysis conducted on only post-regulation, dynamic break-up stages (i.e., ice jam stages) are plotted in Figure 4.7. In this plot, geodetic elevation was converted to field datum (field datum = 312.4 m). The results indicate that the frequency and magnitude of post-regulation ice jams in al l reaches are apparently similar to that at the T P R . Since the estimates are based on small sample sizes, there is a large uncertainty associated with each estimate. A t Fort Vermi l ion, Gerard and Karpuk (1979) analyzed the frequency of ice jams based on various sources of historical data, including resident interviews, archives, photographs, government records, W S C hydrometric records, and environmental evidence. A problem in analyzing historical data from various sources is to assign a rank and record length to each reported flood peak since each data source has a different level above which the event would be noted. Tree scars, for example, vary in the amount of information they can provide. A s you go back in time, the number o f scars in trees decreases while the stage o f the lowest scar increases, since there is a loss o f trees to bank erosion, ice jam activity or beavers. RETURN PERIOD (years) 1.01 1.1 1.5 2 3 4 5 10 20 50 100 thermal break-up .99 .95 .90 .80 .70 .60 .50 .40 .30 .20 .10 .05 .01 PROBABILITY OF EXCEEDENCE (%) Figure 4.8. Frequency distributions of break-up and open-water stages at the Town of Peace River. Best-fit lines were fitted by eye. See text for discussion. 118 The effect of a reduction in information content with time and various sources can be accounted for by using "perception stages". The perception stage is the stage above which the source would probably have provided information on the annual maximum ice related stage. The worth o f the perception stage follows from the fact that i f the source was in a position to notice and recall an event above the perception stage, but did not report it, it can be presumed the maximum water level was below the perception stage for that year (Gerard and Karpuk, 1979). Without perception stages, the frequency distribution tends to biased since lower events are omitted. The rank of the peak is determined by ranking all peaks in the group having a perception stage equal to or lower than peak level o f interest. To estimate the probability o f a single event, the record length (n) associated with that event is taken as the sum of the years in which the perception stage is lower than or equal to the stage of that event. The process may result in two maxima having the same rank, but never the same probability because they would have different record lengths. Fol lowing that, the exceedence probability of each peak is determined in the usual way. Figure 4.9 from Gerard and Karpuk (1979) shows the results o f an ice jam frequency analysis for Fort Vermi l ion using the concept o f perception stages. Note that this analysis was conducted exclusively on pre-regulation events. Although the elevations in this plot are difficult to compare with the data collected in this study, its overall trend is similar to the T P R record, with ice jam floods becoming more important flood mechanisms than open-water floods above a certain level. It should be noted that the concept of perception stage was not applied to the tree scar record used in this study, because the record is relatively short (25 years) and because no obvious perception stages could be identified in the record (Figure 4.6). 119 ?4 20 -"T r 1 1 1 1 1 1 1 1 1 r 1 r~-r—r™ a Ice jams Stages above this level more frequently due to ice jams — Open water _ i _ J L. JL J 1_ J 1 L 265.8 263.8 261.8 259.8 257.8 255.8 z 253.8 2 I UJ -i uu o 251.8 249.8 UJ o o UJ 247.8 99 8 98 95 90 80 60 40 20 10 5 2 1 0.5 PROBABILITY OF EXCEEDANCE (%) 0.1 0.01 Figure 4.9. Frequency distribution of pre-regulation ice jam and open-water stages at Fort Vermilion (from Gerard and Karpuk, 1979). 120 Figure 4.7 indicates that the scar data for the four study reaches do not fit the probability distribution very well (i.e., the confidence limits are far apart). Data points tend to fall below the distribution at low stages and level off at high stages. This is l ikely a result of the small sample sizes, especially for low and intermediate events. The lack of fit also may indicate the existence of a physical upper limit, above which ice jams are relieved by over bank f low (Gerard and Karpuk, 1979). Although differences in ice jam severity between reaches may not be significant, given the short record lengths, Figure 4.7 suggests that ice jams may be more severe in Reach 2b. Since regulation, there is evidence of 19 ice jam events in that reach, as opposed to 13 events in Reach l a , 6 events in Reach l b , and 14 events in Reach 3a. Unfortunately, the data in Figure 4.6 is insufficient to determine i f ice jam severity has significantly changed with f low regulation. With reference to Figure 2.6 and Appendix C , which indicate the ice jam levels at TPR, there appears to have been a greater number of recorded ice jams above 316 m (geodetic) in the post-regulation period. This tends to contradict the situation downstream near the Peace-Athabasca Delta, where break-up ice jamming has declined significantly since regulation (Prowse et al., 1996). Many ice jams at or downstream of T P R have been in large part triggered by break-up on the unregulated Smoky River. Thus, determining the effect of Peace River regulation on ice jam severity becomes a complex problem. One must consider not only post-regulation ice stages and f low releases prior to break-up, but also unregulated runoff from tributaries, which may have undergone a shift to lower values since the mid-1970's (Prowse et a l , 1996). The subject is beyond the scope of this thesis. 121 Reconciliation of scar record with historical record Comparison of botanical evidence with historical records is largely restricted to the TPR. Unfortunately, near the TPR, scar analysis is hampered by the level of development along the banks and the steep bank morphology, which limits tree growth, and thus the availability o f tree scars. A t the T P R , the highest ice jam stage recorded prior to the study was 319.9 m (geodetic) or 7.5 m above field datum in February 1992. The highest tree scars from this event that were identified and measured near the T P R were at 7.2 m above field datum. For this one example, the scar underestimates the actual peak stage by 0.3 m. Unfortunately, the accuracy of scar elevations could not be further tested in this study due to a lack of historical records at other locations. There always exists a high, but unknown, probability that ice shoving w i l l result in scar and trim levels above peak stages at other locations on the river. A long the upper study reaches, ice shoving into mature forest is expected along confined reaches with relatively steep banks. On the lower reaches ice shoving may be limited by ice grounding on extensive bars. 4 . 2 EFFECTS OF R I V E R ICE ON B A N K MORPHOLOGY The significance of ice effects on channel margins has historically been under-evaluated (Gatto, 1993). Since this investigation represents one of the first examinations of ice-effects on the banks of the Peace River, it is largely descriptive. The objective was to determine how the effects of ice on Peace River compare with other northern rivers (most of which are in severer climates). Ice effects are either a direct consequence of ice impact on the bank or indirect impact by the modification o f streamflows and stages. Although attention has been often paid to the reworking of sediments by ice, it is important to remember that ice also acts to protect sediments 122 when it is stable and grounded along banks. The following sections provide descriptions and locations of the principal ice-related bank features that were classified by morphology and process. Although the effects of ice do not exclusively fall under any single process, for discussion, they were grouped by erosional features, depositional features, and overall bank morphology. 4.2.1 Eros ional Features Bar and Bank Scour Bank erosion by ice is often associated with evidence of vegetation damage. Scour by ice jams and ice runs leaves a myriad of small-scale effects on the bank and substrate (see Wentworth, 1932b). Scour marks, keel marks or grooves are widespread on icy rivers and indicate locations where ice has grounded or has shoved. However, the size of these features depends on the severity of ice shoving and grounding. Scour marks on gravel bars are most noticeable when water fi l ls the bottom of the depressions. On the Peace River, small-scale effects were well distributed along the river with the exception to Reach 3 a where near vertical silt and sand banks do not favor the preservation of ice scour evidence. Scour features identified on the Peace River varied from a few centimetres to 2 m deep. Widths were on the order o f 1 m, and lengths varied from 1 m to 20 m. The locations o f ice scoured bars and banks are presented in Figure 4.10. The features indicated vary in scale, with most having a horizontal scale of 1 to 3 m. Although the distribution of scoured bars is widespread, the largest evidence, in both size and number, was found in Reaches l a and l b (Figures 4.11 and 4.12). In these two reaches, freeze-up consolidation and large stage fluctuations l ikely cause frequent grounding of ice along the channel margins and bars. o Ice scoured bank m Ice scoured bar • Boulder/cobble pavement • Ice-push ridge (f 450 550 600 20 40 Kilometres V > 550 mjl 650 X 7> 500 450 450 700 || 600 W 500 s^^ isib //' Ft. Vermilion bridge 650 T * 5 5 0 7050 ^ 7 1000 \ 900 1050fsf~^ """H J-^M**' 1 0 0 0 \ c 950 „ 7050 *1 J 1000 \ 1050 _ B .C. / Alberta border Smoky River 950 Figure 4.10. Locations of ice-push ridges, boulder/cobble pavements, ice-scoured bars, or ice-scoured banks. Features indicated are at a scale of 1 m or greater. Distances are in kilometres from the Slave River. 124 Reach: 1a. 8 Site: 196 Location: km 955.4 left bank Photo: 1.14a Date: Jun 21, 1995 Figure 4.12. Ice-pushed gravel ridge near the head of an island on the upper Peace 125 Additionally, with large discharges at freeze-up, bed scour by redirected flows may be an important process beneath the ice. Scour troughs are often located on gravel bars at the head of islands. Larger troughs were up to 20 m long, 10 m wide, and 2 m deep. Commonly, these features were associated with ice push deposits or sediments nearby (Figure 4.13). Although these features were oriented mostly parallel with the thalweg, in some cases, the linear scour marks deviated from the thalweg direction by up to 45 degrees. A relatively large deposit that formed near the time of freeze-up in 1993 (Chalmers, pers. comm., 1995) was discovered at km 924 on a tight meander bend, between Dunvegan and the Saddle (Burnt) River. Both ice scour and streamflow, presumably redirected by ice, resulted in the formation of a deposit nearly 3 m higher than the other bar surfaces in the area (Figure 4.14). The feature is on the left bank at the upstream end of a right bend. Downstream of the feature, an extensive bar is located along the outside of the bend, atypical of most rivers. It is likely that the channel in the vicinity of this bend is partially maintained by freeze-up ice jams. Ice that forms first along the shallow margin on the outside of the bend restricts flow to the opposite bank. As ice consolidation (i.e., jamming) occurs, greater volumes of ice are shoved to the outside of the bend and become grounded on bars, causing further redirection of flow to the inside of the bend. A similar process probably occurs at many bends (with islands) along the study reach. Following the formation of an ice jam, water building up under high pressures will "search" for the path of least resistance in order to bypass or "short circuit" the jam. The obvious path for flow is along the inside of the bend (i.e., the shortest and steepest route), through existing secondary channels 126 Reach: 2b.6 Site: 495 Location: km 605.2 right bank Photo: 32.4 Date: Jul 28, 1995 Figure 4.13. Ice-pushed boulder and scour trough. Reach: 1b.2 Site: 244 Location: km 923.8 left bank Photo: 48.3 Date: Sep 11, 1995 Contour interval: 0.25 m Figure 4.14. Photos and contour map of large ice-pushed cobble-gravel ridge on the left bank between Dunvegan and Saddle (Burnt) River. Ice grounding is common at this location of a tight meander. Information from Chalmers (1995) suggests this ice-pushed ridge was likely formed during a freeze-up jam in 1993. 128 or across the floodplain (Prowse, pers. comm., 1997). Streamflows are restricted from flowing around the outside of bends during ice jams because of channel confinement at most major bends on the Peace River (i.e., the channel is incised). F low diversion by ice jams at bends is likely responsible for maintaining, i f not forming, many of the secondary channels along the Peace River. Since regulation, flooding of secondary channels is almost always due to ice jams. During ice jam observations in 1996, the author observed ice jam flooding in many of these channels (Figure 4.15). However, in all cases where flooding took place, the secondary channel was conveying only a minor portion of the flow, and velocities were low. It is l ikely that i f secondary channels were to convey the majority of the flow, it would only be for a short period since the ice jam in the main channel would either collapse or progress downstream in a matter of hours. Major ice-scour features, such as the one shown in Figure 4.14, were not common. In total, six of these features were identified: four in Reaches l a and l b , and two in Reach 2b. Other scoured bar features were minor and would likely be reworked in subsequent years. Bank scour on the Peace River was widespread, but scour attributed to ice was difficult to determine. Besides the information provided by residents l iving along the river, ice scour was inferred i f evidence of ice-scarred trees or shoved sediment was found. Ice-scoured banks were confidently identified at only 11 sites (Figure 4.10). These features were most common along the lower end of Reach 2b, where the channel follows a confined, tortuous path with several mid-channel islands and bars. Figures 4.16 and 4.17 show ice-scoured banks in the middle Peace River. Both banks, located on the outside of meander bends, were nearly devoid o f shrubby 129 Figure 4.15. a) Air photo of a bend near 30-Mile Well, characteristic of the middle Peace River. Note that the island is attached to the shore and has a long and wide bar surface at its head; b) ice jam flooding of the secondary channel at the same location. Since regulation, ice jams have been the major process causing flooding of these environments. Note the toe of the ice jam was located approximately 20 km downstream. Figure 4.16. Ice-pushed alluvium and damaged alder stand near Carcajou. This photo was taken two months after break-up, when ice was shoved up this bank. Reach: 2b.5 Site: 475 Location: km 626.8 right bank Photo: 33.15 Date: Jul 29, 7995 Figure 4.17. V iew of the right bank at B ig Bend where 5 m of bank erosion occurred during the spring break-up in 1994. A t this site, a remote hunting camp is located. Note the oi l drum, on the upper right, that was pressed into the bank under the force of ice. 131 vegetation when observed. The photo in Figure 4.16 was taken 2 months after the break-up, so vegetation had not re-established yet. Fol lowing ice scour and the formation of an ice-pushed ridge of alluvium, ice jam flooding in 1996 had deposited a variable thickness of silt along the channel margins. Measurements of these thicknesses varied from 1 to 3 cm. According to the owner of a hunting camp at B ig Bend (Figure 4.17), an ice jam in the spring of 1994 eroded 5 m of the bank along a 100 m section (Sivertson, pers. comm., 1995). Assuming a prism of bank material 100 m by 5 m by 15 m (straight-line distance from water-line to top of bank) was eroded; this equates to 3,750 m 3 o f sediment scoured at one location during a single event. A s a consequence of the steep banks at B ig Bend, ice can impact a large range of elevations along the bank without being grounded. The degree of ice-induced erosion along the river depends on river levels and fluctuations. On a rising stage, ice wi l l tend to float. This would likely reduce the chance of erosion by direct impact. It is also possible that during stage increases, more ice is contributed from upstream as break-up occurs, causing increased shoving and scour at higher levels. On a falling stage, ice wi l l be grounded causing an amount of erosion dependant upon the rate at which the ice impacted the bank. The photo in Figure 4.18, taken near Dunvegan, shows that ice along the channel margin tends to protect the bank from the shearing action of downstream moving ice floes. However, a thin layer (approximately 1 cm) of sediment frozen to the grounded ice was removed following the stage reduction after break-up, when ice toppled down the bank. A similar process was observed in the middle Peace River, where stage fluctuations are lower (Figure 4.19). Attachment of ice to the bank sediment is common along most of the study reach, although direct observations on the lower Peace River were not made during winter. Ice toppling following a stage drop appears to be more significant in the upper Peace River. On the middle Reach: Site: Location: Photo: Date: 1b.1 213.5 km 942.6 right bank downstream of the Dunvegan Bridge 63.19 Apr 18, 1996 Figure 4.18. Close-up view of ice remaining on a bank following break-up. Note the ice that has protected the bank from the shearing action of ice floes also acts to remove a thin layer of sediment as it slides down the bank. Reach: Site: Location: Photo: Date: 2a.1 299 km 849.7 left bank 63.10 Apr 18, 1996 Figure 4.19. View of sediment frozen to ice being plucked from the bank after a stage reduction near the Smoky River confluence. 133 and lower Peace River, stages generally rise during the spring break-up causing ice along the channel margin to float off the bank rather than slide down the bank. Evidence of ice scour tends to be inconspicuous or not visible. Following bank scour by ice and the removal of sediment from the toe of the bank, it is common for the bank to slump or slide, since banks on the Peace River have a high silt and clay content. This possibly explains why scour features are not abundant. Similarly, on the Mackenzie River, Brooks (1996) concluded that ice scour may be a significant factor in reducing the stability of the upper banks. 4.2.2 Depositional Features Ice-pushed Ridges Sediment deposits shoved by ice were the most common features observed along the study reach; see Figure 4.10 for the location of ice-pushed ridges identified along the river. Although ridges of over 5 m height above normal summer river level have been reported on subarctic rivers (Dionne, 1976; Hamelin, 1972), those discovered on Peace River rarely exceeded 2 m (Figure 4.20), and most had a relief of less than one metre. Ice-pushed ridges are generally flooded during the mean annual open-water flood. They are composed of silty-gravel and tend to be armored by cobbles and boulders. Individual stones on some ridges have aligned striations and may be polished, presumably from the planing action of ice floes with entrained sediment. The degree of sorting varies, from loose sediment heaps to ridges armoured by boulder pavements (discussed below). Larger ice-pushed ridges are most abundant near cut banks in Reaches 2b and 3a. The orientations of ice-pushed ridges depend on the trajectory of moving ice floes. Bank configuration is the main factor in determining if ice floes will move parallel or at some angle to the bank. Ridges often occur where slopes consisting of cobbles and boulders 134 Reach: 1b.1 Site: 234 Location: km 930.9 left bank Photo: 2.20 Date: Jun 22, 1995 Figure 4.20. Ice-pushed gravel ridge on a large bar near the Fairview water intake on sub-reach lb. Reach: 3a.4 Site: 668 Location: km 469.8 left bank Photo: 20.7 Date: Jul 13, 1995 Figure 4.21. Sequence of cobble ice-push deposits along left bank near Hungry Bend. Arrows indicate the locations of four deposits visible in this photo. (i.e., glaciofluvial sediments) are being undermined by streamflows. Although summer flows are competent to transport silt and fine sand, only ice shoving moves large cobbles and boulders. The often jerky downstream progression of ice jams and ice runs over many years has resulted in discrete ridges of sediment, rhythmically spaced along the river (Figures 4.21 and 4.22). The spacing of the ridges ranges from 50 to 150 m in sequences no longer than 600 m. Based on the position of the ridges with respect to obvious sources of cobbles and boulders, ice-pushed ridges appear to migrate downstream. Ice-pushed deposits occasionally extend into the channel, forming a hook-like feature. In sub-reach 3a.3, a cobble hook was identified similar to those found on the Yukon River by Wentworth (1932b) (Figure 4.23). The hook feature, similar to most of the ice-pushed deposits identified, was only 1 m or less above the summer stage. The feature was located offshore but connected to the bank at low flow. Cobble hooks were rare and were only found in the lower Peace River. Pavements Cobble/boulder pavements are similar sedimentologically to ice-pushed ridges. Along the Peace River, pavements occupied a position well below the mean annual open-water flood, and thus were observed only during low flows. The surface of pavements resembles a cobblestone road consisting of sub-rounded cobbles and boulders embedded in a matrix of silt and fine sand (see Figure 4.24). Although not widespread, most are located in Reaches 2b and 3a (Figure 4.10), in proximity to sources of gravelly surficial deposits. The most extensive pavements were located in reach 3a at the heads of islands. During the reconnaissance, many pavement were submerged, but it is estimated that the larger pavements are up to 100 m long and 15 m wide (e.g., at the head Figure 4.22. Gravel ice-push deposit jutting into channel on the lower Peace River. Figure 4.24. Typical boulder pavement of the lower Peace River. 138 of Moose Island). Pavements on the Peace River are loosely packed. The removal of individual stones was usually possible by hand or by applying a small amount of force with a piece of wood. Sediment sizes within pavements often decreased in the downstream direction, with larger stones at the upstream end of the pavement usually the least pressed into the matrix. Meltout Deposits and Sedimentation Chaotic patches of clay to cobble sized sediment were found at several locations widely distributed along the river. These patches of sediment were attributed to the action of ice since they were discrete and were located near ice-damaged or matted shrubs, indicating the past presence of grounded ice floes. The deposits form as sediment melts out from ice floes on gently sloping bars. In several instances, deposits were found in tall shrubs over 100 m from the summer water line (Figure 4.25). Since regulation of the Peace River, ice jam floods are the primary means for supplying sediment to the upper banks and floodplain. The significance of ice jam floods, however, is difficult to determine especially when previous open water floods (especially the June 1990 flood on the Peace River) confound the sedimentary record. At a limited number of sites between Sunny Valley and Fort Vermilion, pre- and post-ice jam flood observations in 1996 revealed that deposits of silt averaged 1 cm (range 0.5-10 cm). The thickest deposits were found in relatively protected areas with dense shrub galleries. Willows and balsam poplars, the principal species in the shrub galleries, enhance sedimentation by checking currents. When ice jam flooding occurs for a considerable period, the build up of mud can bury all but the tallest shrubs (Figure 4.26). Silt deposits observed following the 1996 break-up generally thinned in the downstream direction, possibly in response to the widening of the channel and decrease in confinement. Figure 4.25. Isolated patch of cobble-gravel on bar surface deposited during the melting of stranded ice blocks on a bar on the upper Peace River. Reach: Site: Location: Photo: Date: 2a. 6 360.5 km 782.25 left bank at confluence of Whitemud River 63.1 Apr 17, 1996 Figure 4.26. Post break-up conditions at the confluence of the Whitemud River. Note the thick layer of mud within the willow shrubs. Reach: 1a.8 Site: 211 Location: km 947.10 right bank at confluence of Ksituan River Photo: 1.5a Date: Jun 18, 1995 Figure 4.27. Example of sedimentation at mouth of Ksituan River following break-up. At this location, break-up of the Ksituan River likely preceded that of the Peace River causing sediment from the Ksituan River to deposit on top of the ice or within an area confined by ice. 141 Unfortunately, only a limited number of measurements were made at points accessible by road, so that an overall assessment of volumes of sediment deposited was not possible. A t locations where the deposits were observed, silt was concentrated along a bench between the pre-regulation floodplain and the summer water level. This level, often cited in the literature as the active channel shelf, represents the post-regulation floodplain. Areas near confluences were found to have conspicuous silt deposits fol lowing break-up. Figure 4.27, for example, shows an area near the mouth of the Ksituan River. During spring break-up in 1995, it appears that water flowing from the ice-free Ksituan River pooled behind the still ice-covered Peace River causing sediments to be deposited along the ice margin. The sediments that form the banks of the Peace River are transported by mass wasting and erosion from the adjacent slopes and by water from its tributaries. Most o f the Peace River is lined by banks of fine-grained bedrock (i.e., shale) and surficial materials (i.e., glaciolacustrine sediments). Weathering, mass movement and erosion of these materials results in fine-grained sediments (chiefly sand and silt) dominating the surface of the floodplain. Coarser materials, as well as silt and sand, are delivered to the Peace River from its tributaries and adjacent slopes consisting of t i l l and glaciofluvial sediments (primarily in the lower reach). Due to the low competency of the Peace River fol lowing regulation, coarse sediments tend to deposit at, or slightly downstream, of their entry point with the river. Fine-grained sediments are responsible for a low degree of preservation of ice-related features along much of the river, particularly in the lower Peace River. A s indicated, ice-related features are preferentially located and preserved at sites with abundant coarse sediment. This bias means 142 that ice-push features cannot be reliably used on their own to indicate severity of ice disturbance along the river. Rather, ice-push evidence is useful only as corroborating evidence of ice activity. 4.2.3 Bank Morphology Several workers have proposed that icy rivers, especially those with severe ice jams and ice runs, have unique bank morphology in response to seasonal ice scour and ice jam related sedimentation (Hamelin, 1972; Marusenko, 1956; Smith, 1979). Bank characteristics of icy rivers tend to vary from being scoured clean to undisturbed. This was generally true for the Peace River. An interesting bank feature observed in the field was the active channel shelf, which is similar to the becevnik described by Hamelin (1972). Although its elevation varied with respect to the field datum, it consistently had the most abundant ice-related, morpho-sedimentological evidence. The active channel shelf, which has also been described on non-icy rivers (Hupp and Osterkamp, 1985), is a horizontal to gently sloping surface that is situated between the former (pre-regulation) floodplain and channel bed, above most open-water floods and below stages reached during most ice jam events. Hupp and Osterkamp (1985) report that this level is inundated between 5% and 25 % of the time on non-icy rivers. Marusenko (1956) concluded that similar levels in Russia were formed by a sequence of ice scour followed by bank slumping. Along the Peace River the channel shelf was restricted to banks with gentle slopes - typical of depositional environments. Eight profiles were chosen to represent the banks along the study reach in moderate to low exposure environments (i.e., typical of depositional environments); see Appendix E. These sites 143 were found on straight reaches, islands, and point bars. Banks in high exposure areas (i.e., erosional cut banks) were not chosen because they are similar along the river, regardless of the inferred process (e.g., ice scour or fluvial erosion). The profiles selected are accompanied by aerial and ground photos of the location, and a description of the woody vegetation found along the transect. In most of these profiles, the active channel shelf is located at an elevation of 2 to 4 m above the field datum and is vegetated by various shrubs and herbaceous species. Mature trees were located on the former floodplain surface at elevations of 4 to 6 m above the field datum. The elevations of the alluvial surfaces were surprisingly consistent along the river, despite the differences in sediment texture, hydrology, and ice jam intensity. N o trend in differences between the active channel shelf and the floodplain elevations was found at the river-length scale. Variations in bank exposure proved to be greater than any inter-reach differences. Differences between profiles were generally limited to the steepness of the bank, which is a function of the specific location of the profile as well as the sediment texture. In the lower Peace River, near-vertical steps of silts and sands separate each level on the banks. Banks on the lower Peace River have significantly wider shelf and floodplain surfaces than upstream. A s a result, shrubs growing on the shelves are generally older than upstream, since ice often damages only the edge of the surface of the shelf (see Figure 4.2). Botanical evidence on the active shelves in the upper Peace River suggests that they are affected by ice frequently, and in some cases on an annual basis (see Section 4.3). This is l ikely due to ice shoving during freeze-up consolidation and staging. 144 The genesis of channel shelves is believed to be entirely fluvial (cf. Hupp and Osterkamp, 1985). On the Peace River, they are relicts of the reduced open-water floods since regulation. However, these surfaces are not simply static. Ice jam scour and flooding of these surfaces has occurred from 27% to 86% of the post-regulation years (see Table 4.3), depending on the reach. From observations during and fol lowing ice jams on the Peace River, it appears that ice processes on Peace River are responsible for some form of maintenance of these shelves. Figure 4.28 presents an idealized series of cross sections of the Peace River at a location of a typical meander bend. Ice scour episodically destroys the shelf vegetation, and may produce local scour. However, widespread scour was not evident, suggesting that ice may be relatively immobile or shorefast along the channel shelf. Additionally, ice scour evidence, particularly in fine-grained sediments, has a low chance of preservation, especially i f followed by flooding. Ice floe velocities on the shelf are not l ikely high since grounding is expected to occur offshore. However, this is dependent on stage. Nevertheless, an ice-choked channel under extreme hydrostatic forces wi l l result in extensive shoving of ice into riparian vegetation. A t island heads shoving was observed to frequently topple mature and decadent balsam poplars. Ice jam flooding of the active channel shelf and secondary channels at similar elevations is responsible for widespread sedimentation. On average, each event supplies the channel shelf with a thickness of silt estimated from millimetres to a few centimeters. The thickness of silt deposited during each ice jam flood would likely be greater near tributaries and locations where elevated flood levels are sustained longest. Figure 4.28. Idealized channel cross sections of the Peace River during mid-winter, dynamic break-up, ice jam flooding, and summer periods. Principal features include the active channel shelf that is inundated and over-run by water and ice, shoved and scarred mature trees and damaged riparian communities. The sketches are not to scale. The sketch for the dynamic break-up period was adapted from Thorson and Guthrie (1984). 146 4.3 EFFECTS OF RIVER ICE ON RIPARIAN VEGETATION Riparian vegetation grows under unique edaphic conditions which are largely determined by the flood regime and associated sedimentation and scour (Teversham, 1973). These external environmental (allogenic) factors have often been cited as controlling the pattern, age and floristic composition of vegetation communities along rivers (Hupp and Osterkamp, 1996). Succession is the universal process of vegetation change (Cooper, 1926). In the boreal environment it remains largely uninvestigated (Church and North, 1996). On boreal rivers, such as the Peace River, succession is complex due to variations in species lifespans and differences in vegetation responses to changing environmental conditions. In addition, succession is complicated by factors such as variable seed sources, diseases and extreme events that are difficult or impossible to predict. Ice jams and ice runs represent discrete disturbances along channel margins. Ice scour and ice jam floods redistribute litter, nutrients and sediments and are expected to have strong effects on population, community, and ecosystem functioning (Scrimgeour et al., 1994). Vegetation galleries, which reflect unit successions or seres (Cooper, 1926), are commonly found at various levels along channel margins and islands of the Peace River. With increasing elevation above the river, vegetation becomes successively older and taller, reflecting a gradient in hydrogeomorphic conditions. On the Peace River, summer floods dominated hydrogeomorphic conditions prior to regulation. However, following regulation, reduced summer levels have resulted in extensive secondary succession within existing communities and primary succession (colonization) of bars and secondary channels (Figures 4.29 and 4.30) Reach: Site: Location: Photo: Date: 3a.1 570 km 534.0 right bank 27.20 Jul 21, 1995 Figure 4.29. Secondary channel vegetated with willows and horsetails. Reach: Site: Location: Photo: Date: 3a.4 678 km 463.0 right bank near Goose Neck Point 21.4a Jul 14, 1995 Figure 4.30. Wide secondary channel approximately 1.5 m above the summer stage vegetated with primary successional willows. 148 (cf. Church e t a l , 1996). Along most large rivers, there is a gradient of relative importance of physical and biological controls. A long the main channel, physical forces dominate, while in secondary channels biological interactions control community structure. In the post-regulation setting of the Peace River, ice jams are the major form of physical disturbance. Observations during and after break-up confirmed that partial or total destruction of vegetation communities occurs on the Peace River during ice jams and ice runs. However, the distribution of this disturbance is not uniform along the river. A t highly exposed locations, such as cut banks and island heads, ice tends to scour the surface and damage trees, shrubs and herbaceous species. In low exposure environments, such as secondary channels, the effects of ice are less conspicuous. Typically, ice jam flooding deposits only a thin veneer o f silt on the bed of secondary channels and vegetation damage is usually limited to matted shrubs caused by grounded ice floes. In Figure 4.31, the elevations of the lower boundaries of four of the riparian communities are plotted as 20 km running means calculated at 5 km intervals. This plot indicates that the elevation of each riparian community varies considerably along the study reach. The edge of mature forest (Class 5) ranges from 3 to 9 m above field datum, young trees (Class 4) vary from 2 to 7 m above field datum, and shrubs of height Class 2 (1.0-1.9 m) and Class 3 (2.0-4.9 m) occupy various positions between 0 and 6 m above the field datum. Communities less than 1 m tall were discontinuous and not plotted. The frequent removal and regrowth of the youngest shrub communities precludes their use an indicator of ice disturbances. However, mature trees and tall shrubs (Class 4 and 5) provide an index of the downstream severity of disturbances. 12 11 10 -9 -8 -7 -6 5 4 -3 -2 1 -n -1100 Trim line envelope Lower limit of Class 5 vegetation (trees > 15 m) 1000 900 800 700 600 500 400 12 11 10 9 8 7 6 5 4 3 2 1 -0 -1100 Lower limit of Class 4 vegetation (trees 5.0-14.9 m) r 1000 900 800 700 600 500 400 12 11 -10 -9 -8 -7 -I 6 5 4 • 3 -2 1 -n -1100 - Lower limit of Class 2 vegetation (shrubs 1.0-1.9 m) • Lower limit of Class 3 vegetation (shrubs 2.0-4.9 m) 1000 900 800 700 RIVER DISTANCE (km) 600 500 400 Figure 4.31. 20 km running means (calculated every 5 km) of elevations of lower boundaries of vegetation communities along Peace River: a) Class 5 - trees >15 m; b) Class 4 - trees 5.0-14.9 m; and, c) Class 3 - shrubs 2.0-4.9 m and Class 2 - shrubs 1.0-1.9 m. Vegetation younger than Class 2 was highly discontinuous and therefore not plotted. 150 Peak elevations of the forest edge (Class 5) are moderately correlated with peak elevations in the scar and trim line record (r2 = 0.55), suggesting that the elevation of the edge of the forest provides a similar index of the magnitude of ice disturbance along the study reach. However, the difference between scar/trim heights and the edge of the mature forest is not consistent along the river. Upstream of the Smoky River, the height of scars is on average 1 m above the base of trees. Near km 960, upstream of Dunvegan, scars are up to 3 m above the base of trees. This implies that in some locations, trees survive impacts by ice and associated scars. The presence of scars higher on trees above Dunvegan may indicate that: 1) ice floes may be moving or being shoved parallel with the river rather than laterally; 2) ice that stages at high levels during freeze-up is weaker and may cause abrasion and branch trimming but does not destroy trees; and, 3) during break-up, ice in the upper Peace River may be thermally deteriorated and thus weaker (i.e., ice thrusting up the banks may be crushed before trees are shoved over). Downstream of the Smoky River, vegetation elevations and scar/trim elevations correlate highly (r2 = 0.72). The elevation of the forest edge indicates that the most severe ice disturbances occur at the three reaches near km 780,680 and 525. A t km 780 and 680, the channel is relatively straight with numerous side and mid-channel bars and islands. However, the downstream ends of the reaches near km 789 and 680 are marked by highly sinuous meanders (P~2). Near km 525, ice jams are l ikely to form upstream of wide point bars that constrict ice passage. In general, along the lower Peace River, ice jam flooding is relieved by several secondary channels and by an extensive floodplain so mature trees are seldom impacted by ice (except at cut banks). In fact, the running average scar heights fall within the Class 4 vegetation zone on the lower Peace River. 151 Along the Peace River, significant differences in the elevations of similar aged (and height) vegetation communities have been observed, both at the local and river-length scales (see Appendix E). These differences are believed to be a result of allogenic processes such as ice scour and ice jam flooding. At some locations, the disturbances can occur with such frequency that certain shrub communities do not survive for more than a few years before being "mowed" down (Church and North, 1996). Evidence of scarred shrubs and meltout deposits suggests that this occurs at many locations. Each riparian species has a level of tolerance to stress and disturbance, which is a function of its physiology. Willows (Salix spp.), for example, are particularly well rooted and can withstand frequent over-riding by ice. Evidence of this was found along several transects where scar dates on mature trees were younger than adjacent shrub communities growing lower on the bank. Since many riparian vegetation communities recover rapidly after complete scour, measurements of the age and elevation of distinct vegetation communities provides estimates of the date and stage of former ice disturbance events. Open water floods can also disturb riparian vegetation and must be considered. These events, unlike ice jams, can be identified and accounted for in the hydrometric record. Also, open water floods on the regulated Peace River generally do not produce the stages that can be reached by ice jam events (Prowse et al., 1996). Only gentle to moderate sloping banks (i.e., depositional environments) contain obvious tree age trends. The maximum age of woody vegetation on a transect is assumed to approximate the return period for ice scouring events of a sufficient magnitude to uproot or crush vegetation at and below respective elevations. In Figure 4.32, maximum elevations along 19 transects, grouped by reach, are plotted. Eight of these transects are presented in Appendix E. Although data were selected from similar channel locations (point bars, sides of islands, and straight reaches), there is considerable scatter. This reflects differences in exposure, which is difficult to account for quantitatively. Although exposure can be assessed based on morphology, ice shoving tends to be stochastic and unpredictable. The data plotted in Figure 4.32 are representative of ice shoving events within the lower seres along the river. Thus maximum tree ages represent estimates of return periods of major ice scour events. These return periods were plotted in Figure 4.33 and best-fit logarithmic regression were determined. Data within each reach exhibit the widest scatter in the upper Peace River (reaches l a and lb) and lowest scatter in the middle (reaches 2a and 2b) and the lower (reach 3a) Peace River. However, since there are few old dates of mature vegetation along reach 3a transects, they should be interpreted with caution. The wide range of return periods for ice shoving at any given elevation in the upper Peace River was expected since ice cover there is transient and may not form in some years. So, the frequency of ice disturbance is highly variable, and often depends on the f low regulation strategy each year. When ice does form on the upper Peace River, disturbance is l ikely due to freeze-up staging rather than break-up. On the lower and middle Peace River, an ice cover forms consistently each year, and although stages are relatively high during this time due to f low regulation, freeze-up jams are not as extreme as on the upper Peace River, due to the lower channel gradient. Ice jams that occur on 153 110 100 90 80 ? 70 ra 1-60 g 50 < 40 1 30 20 10 REACH 1a 2 4 6 ELEVATION ABOVE FIELD DATUM (m) 110 100 90 80 ? 70 ra J.60 g 50 < 4 0 30 20 10 0 0 229 • 256 REACH 1b 2 4 6 ELEVATION ABOVE FIELD DATUM (m) 110 -100 -O 3 0 0 REACH 2a 90 - • 394 80 - • 395 I 70 " • |eo-S 5 0 : <4 0 ; o 30 - • 20 -10 -o n n - r P c ) 2 4 110 100 90 •) 80 ? 70 ra $.60 g 50 4 4 40 30 20 -I 10 0 0 429 • 446 • 493 • 517 A 551 A 490 REACH 2b a 4 A o» 1, <B,*f*V, ELEVATION ABOVE FIELD DATUM (m) 2 4 6 ELEVATION ABOVE FIELD DATUM (m) 110 100 90 80 ? 70 ra £ 6 0 S 50 < 4 0 30 20 10 0 REACH 3a 0 2 4 6 ELEVATION ABOVE FIELD DATUM (m) Figure 4.32. Max imum vegetation ages versus elevation above field datum for selected numbered transects within each principal study reach. See Appendix D for site locations. RETURN PERIOD (years) RETURN PERIOD (years) 1 10 1 0 0 RETURN PERIOD (years) Figure 4.33. Estimated ice shove frequency for selected transects on the Peace River. 155 the middle and lower Peace River seem to affect the lower seres more uniformly in time and space. Based on the transects presented in Figure 4.33, there conceivably is a single relation for reaches 2a, 2b, and 3a. This suggests that variations in site exposure or morphology may not be important in that part of the river. Based on a wide scatter in the dataset, no firm statements can be made regarding the differences in frequency and magnitude of ice shoving in the lower seres. Theoretically, one should expect transects with the greatest ice shove to be represented by steeper plots on Figure 4.33. 156 5.0 C O N C L U S I O N S The effects of ice jams and associated flooding on bank morphology and riparian vegetation have not been extensively studied on boreal rivers. Since ice jam floods generally create the highest flood levels on the Peace River, understanding where they occur and how they modify the riverine environment is important. A lso, in recent years, development has been increasing on this northern river. Increased infrastructure and habitation along the river can be affected by ice jam floods and ice floes. Thus, this study was undertaken to collect information on ice jam magnitude and frequency at the river-length scale. 5.1 LOCATION, MAGNITUDE AND FREQUENCY OF ICE JAMS Although ice jams may form along any stretch of the Peace River based on hydrometeorological conditions, ice jams tend to recur from year to year at similar locations determined by channel morphology. In the field, this was confirmed by identifying trees with multiple, ice-related scars. Most ice jam locations on the Peace River generally fit the classical criteria, including sharp bends, islands, bars, shoals, near mouths of tributaries, and locations of decreasing channel gradient (Mackay, 1958). The magnitude of ice jam events on the Peace River is known only in relative terms with respect to a field datum. However, the field datum approximates other researchers' reference to "normal summer stage." On the Peace River, ice shove is represented by the trim line, which reaches a maximum elevation of 11 m. This compares to 17 m reported on the Liard River (Parker and Josza, 1973) and 9.5 m on the Mackenzie River (Henoch, 1973; MacKay and Mackay, 1973). 157 In the upper Peace River, the reach most influenced by the effects o f regulation, freeze-up jams create higher stages than break-up jams. Ice jams, associated flooding, and downstream moving ice floes have the most severe impact on the middle Peace River. Stages from ice jam floods are relatively lower in the lower Peace River because the floodplain is relatively wide. Statistically, ice jam severity was not found to be significantly different among the study reaches. This was due to the large uncertainty associated with a restricted scar record for each reach. The frequency of ice jams along the river was found to vary among the study reaches. In the middle Peace River, the tree scar record suggets that 86% of the post-regulation years have had ice stages at or above the base of mature trees, and 45 % of the years ice jams stages exceeded 6 m above field datum. However in the upper Peace River, ice jams occurred in only 27 % of the post-regulation years. Several dates of major ice jams at the T P R were found in the scar record. However, a number of dates at T P R did not correspond with those determined by tree scar analysis, and vice versa. This suggests that transposing ice jam information from one location to another along a large river, such as the Peace River, may not be appropriate. Overall, tree scars proved to be less useful in dating ice jams than was originally anticipated. A s scars were collected over a substantial length of the river, time limited a complete sample of the tree scars at many locations. Scars that were sampled had to be visible from a boat traveling near shore. A n unknown number of old scars at high elevations were unrecorded as they were healed over. Additionally, where ice scour had caused tree removal, the scar record was incomplete. A t the reconnaissance level, however, tree scars provide a preliminary indication of the magnitude and frequency of ice disturbance along the Peace River, particularly where no information had previously existed. 158 5.2 CHANNEL MORPHOLOGY Erosional and depositional features related to ice jamming and shoving along the Peace River were, for the most part, inconspicuous. The alluvial sediments of the Peace River floodplain chiefly consist of fine sand and silt, which are largely derived from fine-grained bedrock (e.g., shales) and surficial materials (e.g., glaciolacustrine materials) adjacent to the channel. Ice-push and ice-scour features formed from these sediments are short-lived as they are rapidly eroded. Ice scour tends to be localized and is difficult to identify in the field unless the process is directly observed. Most sedimentary features described in the seminal paper by Wentworth (1932b) for the Yukon River were found on the Peace River. But those observed on the Peace River were mainly small-scale scour features that marked locations of shoving or grounding of ice. Only a small number of ice-pushed ridges were of sufficient size to be expected to survive for more than a few years. Evidence of secondary channel inundation during ice jams on the Peace River suggests that secondary channels are more active than previously assumed based on the open-water flood regime. Despite ice jam activity, serai vegetation appears to be thriving in many secondary channel environments. Sediment deposition is l ikely to occur within the back channels since f low velocities are low. Temporary ice jams in the main channel may create short-lived, high velocity flows through the secondary channels increasing the potential for ice scour. Persistent wi l low and balsam poplar communities growing within secondary channels appear to be only moderately affected by ice scour, and may in fact be benefiting from increased allochthonous inputs of nutrients. Areas of excessive ice jam sedimentation, which would hamper growth, are localized - the most common locations being near tributaries. 159 The active channel shelf on the Peace River is similar in appearance to "benches" or levels described in the literature on icy rivers (cf. Danilov, 1972; Hamelin, 1972; Smith 1979,1980). Although no firm statement can be made regarding its genesis, it appears that both the scouring action of ice and ice jam sedimentation are actively maintaining this bench. The importance of the two processes, however, appears to be in favour of deposition. A long Albertan rivers, Smith (1979, 1980) suggested that the 2-year summer flood maintains a low-level bench, while the upper bench, at the height of the 9-year summer flood, is maintained by ice scour. Based on vegetation transects in this study, ice shove on the active channel shelf occurs every 8 to 11 years. Lower return periods for ice shove occur within the middle Peace River, where ice jam severity (i.e., magnitude and frequency) tends to be highest. N o river-length trend was found for the difference in elevations between the active channel shelf and what is now considered the former floodplain surface along the Peace River. A long the upper Peace River, the greatest effects of ice on bank morphology were near Dunvegan and near the mouth of the Saddle River. On the middle Peace River, break-up on the Smoky River, the decrease in channel gradient near Notikewin River, and confined and sinuous morphology of the reach in general increase the potential for ice jams and geomorphic activity. Near the mouth of the Notikewin River, ice jams are particularly severe because of numerous tight bends and islands. On the lower Peace River, the hydrology and ice regime are influenced the least by regulation, and channel morphology is less confined than the upper and middle reaches. Val ley sides of the lower Peace River include coarse materials, such as t i l l and glaciofluvial sediments. The presence of cobbles and boulders along with fine-grained sands and silts allow for the formation of boulder pavements by the planing action of ice. These pavements are non-existent in the upper Peace River and are rare in the middle Peace River. 160 5.3 RIPARIAN VEGETATION Vegetation damage was the most visible effect of ice along the Peace River. The riparian zone is in a constant state of flux with respect to age and location of species over time. Ice-push is a major disturbance on the Peace River and causes partial or complete destruction of vegetation on and below the active channel shelf along most study reaches. Channel morphology and the degree of exposure determine the area affected by ice-push. A t gently sloping island heads, ice shove is common so that shrubby vegetation w i l l unlikely mature beyond 10 years. In low exposure areas, ice damage is usually limited to the occasional grounding of stray ice floes during ice jam floods. Depending on its size and location, ice may remain on bars and banks for weeks, causing a delay in vegetation growth. Prolonged ice jam flooding may result in thick mud deposits, which reduce levels of oxygen and solar radiation required for vegetation growth. Conversely, inputs of sediments, organic matter, and nutrients during ice jam flooding may promote future vegetation growth. On the Peace River, ice scour and ice jam flooding determine the location of the lower boundaries of vegetation communities. A l l community boundaries were found to be at relatively higher elevations in areas that are most severely affected by ice jams (i.e., reach 2b). Ice jams maintain vegetation galleries of different age, height, and species composition through events of various magnitudes and frequencies. Damage to vegetation was most severe in the upper and middle Peace River. In the upper Peace River, the channel is confined and banks are relatively steep. Within the upper Peace River, ice tends to impact riparian shrubs during annual freeze-up for those reaches that become ice covered. The greatest impact on vegetation was apparent in the middle Peace River. This is reflected by the relatively high elevation of the lower edge of the mature forest. A long the lower Peace River, vegetation galleries appear to be at an advanced state of riparian succession since direct ice scour is limited to a relatively narrow fringe. Due to the unconfined morphology, wide vegetation galleries have established. Ice jam flooding, however, has the potential to deposit thin layers of sediment across extensive areas along the lower Peace River. 5.4 F U T U R E STUDIES It is recommended that future studies investigate ice scour processes during both the freeze-up and break-up periods. In the present study, the type of evidence observed did not provide the necessary information to resolve the actual period that was responsible for the effects observed. A lso , detailed, site specific dendochronological investigations are recommended to refine the data presented in this study. Ideally, such an investigation would initially focus on the areas of greatest ice jam activitiy presented in this thesis and should utilize modem surveying techniques to accurately determine past flood levels. 162 6.0 REFERENCES Acres Consulting Services Ltd. (Acres). 1980. Dunvegan power project: Peace River ice study, Final Report for the Energy Resources Conservation Board, Calgary, Alberta. Acres Consulting Services Ltd. (Acres). 1984. Behavior of ice covers subjected to large daily f low and level fluctuations: Peace River ice regime study, V o l . 3, Canadian Electrical Association, Montreal, Quebec. Adams, J . 1825. On the movements of rocks by the expansive power of freezing water. The American Journal of Science and Arts, 9: 136-145. Adams, W. P. 1977. How spring ice break-ups alter our shorelines. Canadian Geographical Journal, 94 (2): 62-65. Alestalo, J . 1971. Dendrochronological interpretation of geomorphic processes. Fennia, 105 140 p. Andres, D. 1981. Break-up of the Athabasca and Peace Rivers: A generalized documentation. In Spring Break-up: Mackenzie River Basin Study Report Supplement 3, Mackenzie River Basins Committee, pp. 7:1-7:12. Andres, D. 1994. Peace River ice regime. Mackenzie Basin Impact Study Interim Report No . 2. In Proceedings of the Sixth Biennial A E S / D I A N D Meeting on Northern Climate & M i d Study Workshop of the Mackenzie Basin Impact Study, Apr i l 10-14, Yel lowknife, N.W.T., S.J. Cohen (Editor), pp. 237-245. Andres, D. 1995. Personal communication, Tr i l l ium Engineering and Hydrographies Inc., Edmonton, Alberta. Andres, D. 1996. The effects of f low regulation on freeze-up regime: Peace River, Taylor to Slave River. Northern River Basins Study Project Report 122, Edmonton, Alberta. Andres, D.D. and P .G. Van Der Vinne. 1994. Mackenzie Basin Impact Study: Effects of Climate Change on the Ice of the Peace River, Taylor to the Slave River, Draft Report. Prepared for theMackenzie Basin Impact Study Adaptation Branch, Canadian Climate Centre, Edmonton, Alberta, 57 p. Anonymous. 1987. Recreation sites in the upper Peace River Val ley, Pub. No . 1/182. Edmonton. Anonymous. 1991. Recreation sites in the lower Peace River Val ley, Pub. No . 1/303, Edmonton. Ashton, G.D. 1978. River ice. Annual Review of Fluid Mechanics, 10: 369-392. Ashton, G.D. 1979. River ice. American Scientist, 67: 38-45. Assaf, H. 1995. Analysis of events leading to the February 1992 flooding of the Town of Peace River, Alberta. In Proceedings of the 8 t h River Ice Workshop, August 16, Kamloops, Atmospheric Environment Service. 1993. Canadian Climate Normals 1961-90. Environment Canada. Baldrige, J .E. and E.W. Trihey. 1996. Maintaining biological integrity in instream flow studies in large rivers. In Northern River Basins Study Project Report No . 66, Proceedings of the Northern River Basins Study Instream Flow Needs Workshop, Walder, Gordon L. (Editor), October 14-15, 1993 and January 6-7, 1994, Northern River Basins Study, Edmonton, Alberta. Bart, B., S. Ward and M . Toews. 1988. Fort Vermi l ion, Alberta 1788-1988. Employment and Immigration Canada and Fort Vermi l ion and District Bicentennial Association. B .C. Hydro. 1982: Peace River ice observations 1981-82. Report No . HI566, Hydroelectric Generations Projects Divis ion, October. B .C . Hydro. 1983. Peace River ice observations 1982-83. Report No . H I641, Hydroelectric Generations Projects Divis ion, July. B .C. Hydro. 1986. Peace River ice observations 1983-84. Report No . H I 867, Hydroelectric Engineering Divis ion, February. B .C . Hydro. 1987. Peace River ice observations 1985-86. Report No . HI910, Hydroelectric Engineering Divis ion, March. B .C. Hydro. 1989a. Peace River ice observations 1986-87. Report No . H2014, Hydroelectric Engineering Div is ion, May. B .C . Hydro. 1989b. Peace River ice observations 1987-88. Report No . H2015, Hydroelectric Engineering Divis ion, June. B .C. Hydro. 1990a. Peace River ice observations 1988-89. Report No . H2261, Hydroelectric Engineering Div is ion, December. B .C . Hydro. 1990b. Peace River ice observations 1989-90. Report N o . H2407, Hydroelectric Engineering Div is ion, December. B .C. Hydro. 1993. Peace Canyon Dam and Generating Station: Peace River Power. Pamphlet. B .C. Hydro and Power Authority ( B C H P A ) . 1975. Peace River ice observations 1974-75. Report No . 768, Hydroelectric Design Division. B .C . Hydro and Power Authority ( B C H P A ) . 1977a. Peace River ice observations 1975-76. Report No . 860, Hydroelectric Design Divis ion, June. 164 B.C. Hydro and Power Authority ( B C H P A ) . 1977b. Peace River ice observations 1976-77. Report No . 870, Hydroelectric Design Divis ion, October. B .C. Hydro and Power Authority ( B C H P A ) . 1978. Peace River ice observations 1977-78. Report No . 961, Hydroelectric Design Division, December. B .C. Hydro and Power Authority ( B C H P A ) . 1979. Peace River Regulation Study. Report No . 800, Hydroelectric Design Divis ion. Beltaos, S. 1990. Break-up jams. In Cold Regions Hydrology and Hydraulics. American Society of C i v i l Engineers, N e w York. pp. 485-510. Beltaos, S. 1993. Transport and mixing process. In Environmental Aspects of River Ice, T.D. Prowse and N.C. Gridley (Editors), N H R I Science Report No . 5, National Hydrology Research Institute, Environment Canada, Saskatoon, Saskatchewan, pp. 31-42. Beltaos, S. (Editor). 1995. River Ice Jams. Water Resources Publications, Highlands Ranch, Colorado. Beltaos, S., R. Gerard, S. Petryk, and T.D. Prowse. 1990. Working group on river ice jams: Field studies and research needs. N H R I Science Report No . 2, Environment Canada, Saskatoon, Saskatchewan, 121 p. Bergeron, N . and A . G . Roy. 1988. Les effets d'un embacle sur la morphologie du l i d'un confluence de cours d'eau [The effects of an ice jam on the morphology of a river channel confluence]. Geographie physique et Quatermaire, 42 (2): 191-196. Bolsenga, S.J. 1968. River Ice Jams: A Literature Review. Great Lakes Research Center, U.S. Lake Survey. Bray, D.I. and R. Kellerhals. 1979. Some Canadian examples of the response of rivers to man-made changes. In Adjustments to the Fluvial System, D.D. Rhode and G.P. Wil l iams (Editors). Kendall/Hunt Publishing Company, Dubuque, Iowa, pp. 351-372. Brooks, G.R. 1993. Characteristics of an ice-scoured river bank near Keele River confluence, Mackenzie Valley, Northwest Territories. In Current Research, Part B, Geological Survey of Canada, Paper 93-1B, pp. 39-43. Brooks, G.R. 1996. The Fluvial Geomorphic Character of the Lower Reaches of Major Mackenzie River Tributaries Fort Simpson to Normal Wells, Northwest Territories. Geological Survey of Canada Bulletin 493. Cameron, T.F. 1977. Biophysical analysis and evaluation of capability: L a Crete Area. Alberta Energy and Natural Resources, 47 p. Cameron, J.L. and J .D.H. Lambert. 1971. The vegetation and its distribution on a Rideau River Island, Ottawa. Naturaliste Canadien, 98: 675-689. 165 Carder, A . C . 1965. Climate of the Upper Peace River Region. Publication No. 1224, Department of Agriculture, Ottawa, Canada, 19 p. Carder, A . C . and B. Siemens. 1971. Climate of the Lower Peace River Region. Publication No. 1434, Department o f Agriculture, Ottawa, Canada, 22 p. Carson, M.A. , 1992. Assessment o f sediment transport data and sediment budget analysis for the Peace-Slave river system, Alberta: Phase II (Sediment loads and regional budget). Report for Environment Canada, Calgary, Alberta. Carson, M.A . , J . Syrgiannis and F . M . Conly. 1997. Peace River photographic library on C D -R O M . Map and photographic companion documents prepared for the Northern River Basins Study, Edmonton, Alberta. Catto, N.R. 1991. Quaternary geology and landforms of the eastern Peace River region, Brit ish Columbia, N T S 94A/1,2,7,8. Open File 1991-11. Brit ish Columbia Ministry of Energy, Mines, and Petroleum Resources, Geological Survey Branch, Mineral Resources Divis ion, 19 p. Chalmers, J . 1995. Personal communication on September 10, 1995. Resident and property owner along Peace River near Fairview, Alberta. Church, M . 1974. Hydrology and permafrost with reference to northern North America. In Permafrost Hydrology, Proceedings of Workshop Seminar 1974, Canadian National Committee for the International Hydrological Decade, Ottawa, Ontario, pp. 7-20. Church, M . 1988. Floods in cold climates. In Flood Geomorphology, V . R . Baker, R .C. Kochel , and P.C. Patton (Editors), John Wiley & Sons Inc, pp. 205-228. Church, M . 1995. Geomorphic response to river f low regulation: Case studies and time-scales. Regulated Rivers: Research & Management, 11: 3-22. Church, M . and M.J . Mi les. 1982. Discussion on 'Processes and mechanisms of river bank erosion'. In Gravel-bed Rivers, R.D. Hey, J .C. Bathurst and C.R. Thorne (Editors), John Wiley & Sons Ltd., pp. 259-270. Church, M . and M . North. 1996. Post-regulation change and development of riparian vegetation along Peace River: predictions and initial observations. In Northern River Basins Study Project Report No . 66, Proceedings of the Northern River Basins Study Instream Flow Needs Workshop, Walder, G.L. (editor), October 14-15,1993 and January 6-7,1994. Northern River Basins Study, Edmonton, Alberta. Church, M . , K . M . Rood, M .E .A . North and J. Teversham. 1982. Peace River surveys 1979 and 1981. Department of Geography, The University of Brit ish Columbia, 54 p. Church, M . , J . X u , A . Moy and L. Uunila. 1996. Changes in morphology and riparian vegetation following f low regulation, Peace River, 1968 and 1993. Northern River Basins Study Project Report No . 102, Edmonton, Alberta, 91 p. Coll inson, J.D. 1970. Bedforms on the Tana River, Norway. Geograflska Annaler, 52A(1): 31-56. Coll inson, J.D. 1971. Some effects of ice on a river bed. Journal of Sedimentary Petrology, 41(2): 557-564. Conly, M . 1996. Preliminary break-up data collected from Water Survey of Canada chart recorders for Hudson Hope, Taylor, Dunvegan, Town of Peace River, Fort Vermil ion and Peace Point. National Hydrolological Research Institute, Environment Canada. Cooper, R .H. and A . B . Hollingshead. 1973. River bank erosion in regions of permafrost. In Fluvial Processes and Sedimentation, Proceedings of the Hydrology Symposium, University of Alberta, May 8-9,Department of Environment, pp. 187-196. Cooper, W.S. 1926. The fundamentals of vegetation change. Ecology, 7 (4): 391-413. Coulson, A . and R.J. Adamcyk. 1969. The effects of the W . A . C . Bennett Dam on downstream levels and flows. Technical Bulletin No . 18, Inland Waters Branch, Department of Energy, Mines, and Resources, Ottawa, Ontario, 9 p. Coulson, C . H . 1991. Manual of operational hydrology in Brit ish Columbia, second edition, Ministry of Environment, Water Management Divis ion, 234 p. Danilov, I.D. 1972. River ice as a factor of relief formation and sedimentation. Problemy Kriol i thologi i . Moscow University Press, 1972, No . 2, pp. 138-144. Davies, L .B. , R .K. Deeprose and H . M . Hunt. 1984. Observations and report on river ice conditions in the Peace River basin 1982 - 1983: Executive summary. Report to the Alberta Department of Environment and Brit ish Columbia Ministry of Environment. Joint Task Force, 6 p. Day, T.J. and J .C. Anderson. 1976. Observations on river ice, Thomsen River, Banks Island, District of Franklin. Paper 76- IB, Geological Survey of Canada, Terrain Sciences Divis ion, pp. 187-196. Dionne, J.C. 1974. How drift ice shapes the St. Lawrence. Canadian Geograpical Journal, 88 (2): 4-9. Dionne, J .C. 1976. L'action glacielle dans les schores du littoral oriental de la Baie de James. Carriers de Geographie de Quebec, 20(50): 303-326. 167 Dionne, J .C. 1985. Drift-ice abrasion marks along rocky shores. Journal of Glaciology, 31 (109): 237-241. Doyle, P.F. 1982. Significant river ice problems can occur even in southern Brit ish Columbia. In Proceedings of the workshop on hydraulics of ice-covered rivers, Edmonton, Alberta, June 1-2, D.D. Andres and R. Gerard (compilers), N R C C Subcommittee on Hydraulics of ice-covered rivers, pp. 296-302. Doyle, P.F. 1988. Damage resulting from a sudden river ice break-up. Canadian Journal of C i v i l Engineering, 15: 609-615. Eardley, A . J . 1938. Yukon channel shifting. Bulletin o f the Geological Society o f America, 49: 343-358. Eggington, P.A. 1980. Determining river ice frequency from the tree record. Current Research, Part A , Geological Survey of Canada, Paper 80-1 A , pp. 265-270. Energy, Mines and Resources Canada. 1986. The National Atlas of Canada, 5 t h Edit ion, Geographical Services Div is ion, Surveys and Mapping Branch. Environment Canada. 1997. H Y D A T C D - R O M , Version 95. Environment Canada. 1998. H Y D A T C D - R O M , Version 96-1.04. F i l ip , G . M . , L .D. Bryant, and C A . Parks. 1989. Mass movement o f river ice causes severe tree wounds along the Grande Ronde River in northeastern Oregon. Northwest Science, 63(5): 211-213. Flygare, H. 1983. Sir Alexander Mackenzie Historic Waterways in Alberta: A map/guide for river travel on the Clearwater, Athabasca, Slave & Peace Rivers. Banff Crag & Canyon, 88 p. Fonstad, G.D. 1992. Final report: Peace River ice jam flooding, February 28, 1992. Alberta Environmental Protection, River Engineering Branch, Appl ied River Engineering Studies, 18 p. Fonstad, G.D. 1995. Personal communications, River Engineering Section, Alberta Environmental Protection. Fonstad, G.D. 1996. Personal communications, River Engineering Section, Alberta Environmental Protection. Fonstad, G.D. and M . E . Quazi. 1994. A n analysis of ice jam stage increases, Peace River at the Town of Peace River. Alberta Environmental Protection River Engineering Branch, Technical Services and Monitoring Division, 23 p. 168 Garner, L .A . 1979. Ice break-up observations at the Town of Peace River. Alberta Environment, River Engineering Branch. Unpublished report. Garner, L .A. and M.E . Quazi. 1979. Ice break-up observations at the town of Peace River, 1979. Alberta Department of the Environment, Environmental Engineering Support Services, Technical Services Division, Edmonton, Alberta. Gatto, L .W. 1982. Reservoir bank erosion caused and influenced by ice cover. Special Report 82-31, U.S. Army Corps of Engineers, Cold Regions Research and Engineering Laboratory. Gatto, L.W. 1993. Effects of ice on shorelines. In Environmental Aspects of River Ice. T.D. Prowse and W . C . Gridley (Editors), National Hydrology Research Institute, Science Report No . 5, Environment Canada, Saskatoon, Saskatchewan, pp. 42-45. Gerard, R. 1975. Preliminary observations of spring ice jams in Alberta. In Proceedings of the Third International Symposium on Ice Problems, G.E. Frankenstein (Editor), August 18-21, Hanover, New Hampshire, pp. 261-273. Gerard, R. 1979. River ice in hydrotechnical engineering: A review o f selected topics. In Proceedings of the Canadian Hydrology Symposium: Cold Climate Hydrology, Vancouver, B .C. , N R C C , Associate Committee on Hydrology, pp. 1-29. Gerard, R. 1981. Ice scars: Are they reliable indicators of past ice break-up water levels? In Proceedings of the International Association for Hydraulic Research Ice Symposium, Quebec City, July 27-31, pp. 847-859. Gerard, R. 1990. Hydrology of floating ice. In Northern Hydrology: Canadian Perspectives. T.D. Prowse and C.S.L . Ommanney (Editors), National Hydrology Research Institute, Science Report No . 1, Environment Canada, Saskatoon, Saskatchewan, pp. 103-134. Gerard, R. and E. W. Karpuk. 1979. Probability analysis o f historical f lood data. American Society o f C i v i l Engineers Proceedings, Journal of the Hydraulics Div is ion, 105: 1153-1165. Gilbert, G .K . 1908. Lake ramparts. Sierra Club Bulletin, 6: 225-234. G i l l , D. 1971. Modif ication of levee morphology by erosion in the Mackenzie River delta, Northwest Territories, Canada. Polar Geomorphology, Institute o f Brit ish Geographers Special Publication 4, pp. 123-138. Gi l l is . 1997. Edmonton Journal Extra, Apr i l 23, 1997. Gray, D. M . and T.D. Prowse. 1993. Snow and floating ice. In Handbook of Hydrology, D.R. Maidment (Editor), McGraw-Hi l l Inc., New York, pp. 7.1-7.58. 169 Greene, H.G. 1970. Microrel ief on an Arctic beach. Journal of Sedimentary Petrology, 40 (1): 419-427. Greenland Engineering Group. 1997. Hydrotec 1995-96. Hydrologic analysis program for Windows 95, Concord, Ontario. Hamelin, L .E. 1972. Le glaciel de Iakoutie en Siberie Nordique. Biuletyn Peryglacjalny, 21: 75-86. Harrington, R.A. and D.G. McLean. 1985. Field observations of river bed scour on the Peace River near Fort Vermil ion. Canadian Journal of C iv i l Engineering, 11: 782-797. Henoch, W.E.S. 1960. Fluvio-morphological features of the Peel and Lower Mackenzie Rivers. Geographical Bulletin, 15: 31-45. Henoch, W.E .S . 1973. Data (1971) on height, frequency of floods, ice jamming, and tree-ring studies. In Hydrologic Aspects of Northern Pipeline Development, Report 73-3, Task Force on Northern O i l Development, Ottawa, pp. 153-190. Hicks, F. E., and M c K a y , K. 1996. Hydraulic Flood Routing Models of the Peace and Slave Rivers, Hudson Hope to Peace Point. Northern River Basins Study Project Report No. 77, Edmonton, Alberta. Hobbs, W . H . 1911. Requisite conditions for the formation of ice ramparts. Journal of Geology, 19: 157-160. Hudson, E. 1997. The Climate of the Peace, Athabasca and Slave River Basins. Northern River Basins Report 124, Edmonton, Alberta. Hume, J.D. and M . Schalk. 1964. The effects of ice push on Arct ic beaches. American Journal of Science, 262 (1): 267-273. Hupp, C.R. and W.R. Osterkamp. 1985. Bottomland vegetation distribution along Passage Creek, Virginia, in relation to fluvial landforms. Ecology, 66(3): 670 - 681. Hupp, C.R. and W.R. Osterkamp. 1996. Riparian vegetation and fluvial geomorphic processes. Geomorphology, 14: 277-295. Illichevsky, S. 1933. The river as a factor of plant distribution. The Journal of Ecology, 21: 436-441. Jackson, P.C. 1975. Geological highway map of Alberta. The Canadian Society of Petroleum Geologists, Calgary, Alberta. Jakob, M . 1996. Morphometric and geotechnical controls of debris f low frequency and magnitude in southwestern British Columbia. Unpublished Ph.D. thesis, Department of Geography, The University of Brit ish Columbia. 170 Jennings, J .N . 1958. Ice action on lakes. Journal of Glaciology, 3: 228-229. Jones, J.F. 1966. Geology and groundwater resources of the Peace River District, Northwestern Alberta. Bulletin 16, Research Counci l of Alberta, Edmonton, 143 p. Keenhan, T., U.S. Panu and V . C . Kartha. 1982. Analysis of freeze-up jams on the Peace River near Taylor, Brit ish Columbia. Canadian Journal of C iv i l Engineering, 9: 176-188. Kellerhals, R. and M . Church. 1980. Comment on "Effects of channel enlargement by river ice processes on bankfull discharge in Alberta, Canada" by D.G. Smith. Water Resources Research, 16(6): 1131-1134. Kellerhals, R., C R . Ne i l l , and D.I. Bray. 1972. Hydraulic and geomorphic characteristics of rivers in Alberta. River Engineering and Surface Hydrology, Report 72-1, Research Counci l of Alberta, Edmonton, 48 p. Ketchum, K. 1996. Review of peak river level frequencies for Peace River at Taylor. B .C. Hydro Power Supply Operations, 10 p. Kindle, E . M . 1918. Notes on sedimentation in the Mackenzie River Basin. Journal of Geology, 26: 341-360. Kindle, E . M . 1920. Arr ival and departure of winter conditions in the Mackenzie River Basin. The Geographical Review, 10: 388-399. Klassen, R.W. 1989. Quaternary geology of the southern Canadian interior plains. In Quaternary Geology of Canada and Greenland, Geology of Canada, R J . Fulton (editor), Geological Survey of Canada, 1: 138-174. Koutaniemi, L. 1984. The role of ground frost, snow cover, ice break-up and flooding in the fluvial processes of the Oulanka River, N E Finland. Fennia, 162 (2): 127-161. Kovacs, A . 1983. Shore ice ride-up and pile-up features, Part 1: Alaska's Beaufort Sea coast. C R R E L Report 83-9, 59 p. Kumar, P. 1977. Biophysical analysis and evaluation of capability: High Level. Alberta Energy and Natural Resources, 47 p. Lapointe, M.F. 1984. Patterns and processes on channel change, Mackenzie Delta, N.W.T. 1983-84 Progress Report. National Hydrological Research Institue, Surface Water Div is ion, Northern Hydrology Section, Ottawa, 52 p. Lawson, D.E. 1983. Erosion of perenially frozen streambanks. Report 83-29. U.S. Army Corps of Engineers, Co ld Regions Research and Engineering Laboratory. 171 Lawson, D.E., E.F. Chacho, Jr., B .E. Brockett, J.L. Wuebben, C M . Col l ins, S.A. Arcone, and A . J . Delaney. 1986. Morphology, hydrauilics, and sediment transport of an ice-covered river: Field techniques and initial data. Report 86-11. U.S. Army Corps of Engineers, Cold Regions Research and Engineering Laboratory. Laszloffy, W . 1948. Behavior of ice in rivers. Waterways Experimental Station, Translation 49-6, 47 p. Leffingwell, E.deK. 1919. The Canning River region. United States Geological Survey Professional Paper, 109: 174-177. Lewis, A . F . , G.J . Naito, S.E. Redden and B C Hydro. 1996. Fish F low Studies Project - Fish F low Overview Report. British Columbia Hydro and Power Authority Safety and Environment Report No . EA:95-06. Lopatin, I. A . 1871. On cobbles and rock outcrops striated and polished by ice on the shores of the Yenisei River. Memoirs of the Russian Geographical Society, St. Petersberg, 4: 293-328. MacKay, D.K. (1967). Discharge on the Liard and Mackenzie Rivers. Geographical Bulletin, 11:37. Mackay, J.R. 1958. The valley of the lower Anderson River, N.W.T. Geographical Bulletin, 11: 37-56. Mackay, J.R. 1994. Personal communication, Department of Geography, The University of Brit ish Columbia. MacKay, D.K. and J.R. Mackay. 1973. Locations of spring ice jamming on the Mackenzie River, N.W.T. In Hydrologic Aspects of Northern Pipeline Development, Report 73-3. Task Force on Northern O i l Development, Ottawa, pp. 233-257. MacKay, D.K., D.E. Sherstone, and K . C . Arnold. 1974. Channel ice effects and surface water velocities from aerial photography of Mackenzie River break-up. In Hydrologic Aspects of Northern Pipeline Development, Report 74-12, Task Force on Northern O i l Development, Ottawa, pp. 71 -107. Mackay, J.R. and D.K. MacKay. 1977. The stability of ice-push features, Mackenzie River, Canada. Canadian Journal of Earth Sciences, 14(10): 2213-2225. Marusenko, Y.I . 1956. Deyatel 'nost '1 'da naberegakh rek [Ice action on river banks]. Priroda, 12: 91-93. Martini, LP . 1981. Ice effect on erosion and sedimentation on the Ontario shores of James Bay, Canada. Zeitschrift fur Geomorphologie, 25 (1): 1-16. 172 Martinson, C. 1980. Sediment displacement in the Ottauquechee River: 1975-1978. C R R E L Special Report 80-20, U.S. Army Corps of Engineers C R R E L , 14 p. Mathews, W . H . 1980. Retreat o f the last ice sheets in Northeastern Brit ish Columbia and adjacent Alberta. Geological Survey Bulletin 331, Geological Survey of Canada, Ottawa, 21 p. McLean, D.G. and M.D. Anderson. 1980. Hydraulic and morphological processes on the Peace River near Fort Vermil ion. Report SWE80-2, Alberta Research Counci l , Transportation and Surface Water Engineering Department, 188 p. McPherson, H.H. 1966. Morphology and fluvial processes of the lower Red Deer River Val ley, Alberta. Ph.D. thesis, M c G i l l University. Mel in , R. 1954. Stream erosion and sedimentation at low water in winter. Union geodesique geophysique, I.A.H.S., Compt. rend., Assemb. gen., Rome 1954, 3: 257-260. Mercer, A . G . and R .H . Cooper. 1977. River bed scour related to the growth of a major ice jam. In Proceedings of the 3 r d Canadian Hydrotechnical Conference, Quebec, Canada, pp. 291-308. Miche l , B. 1971. Winter regimes of rivers and lakes. Co ld Regions Research and Engineering Laboratory, Monograph III-Bla, 131 p. Moberly, J.J. and W . B . Cameron. 1929. When Fur was K ing . J . M . Dent and Sons Ltd., Toronto. Mowat, F. 1967. Canada North, The Canadian Illustrated Library, McCle l land and Stewart Limited, Toronto, 128 p. Myles, E.L. 1965. The Emperor of the Peace River 1886-1952. Cooperative Press Ltd., Edmonton, Alberta. National Research Counci l of Canada (NRCC) . 1989. Ice jam floods. In Hydrology of Floods in Canada: A guide to planning and design, W.E . Watt (Editor), Ottawa, Ontario, pp. 169-183. National Research Counci l o f Canada ( N R C C ) . 1990. Optimum Operation of Hydro-electric Plants During the Ice Regime of Rivers: A Canadian Experience. Associate Cornmittee of Hydrology, Subcommittee on Hydraulics o f Ice Covered Rivers, 81 p. Ne i l l , C R . and D.D. Andres. 1984. Freeze-up flood stages associated with fluctuating reservoir releases. In Proceedings of the C S C E Cold Regions Engineering Specialty Conference, Edmonton, Alberta. Newbury, R. 1968. The Nelson River: A study of subarctic river processes. Ph.D. thesis. The Johns Hopkins University, Baltimore, Maryland. 318 p. 173 N ib le r ,G.J . 1980. The spring break-up. Weatherwise, 33(5): 207-209. Nuttall, J .B. 1974. Report on ice break-up at Peace River, Alberta. Alberta Environment, Technical Services Divis ion, Edmonton, Alberta. O'Connor. 1997. Peace River Record-Gazette, Apr i l 22, 1997. Outhet, D .N. 1974a. Bank Erosion in the Southern Mackenzie River Delta, Northwest Territories, Canada. Unpublished M.Sc. thesis, Department of Geography, University of Alberta, 89 p. Outhet, D .N. 1974b. Progress report on bank erosion studies in the Mackenzie River Delta, N.W.T. In Hydrologic Aspects of Northern Pipeline Development, Report 74-12, Task Force on Northern O i l Development, Ottawa, pp 297-345. Owens, E .H . 1976. The effects o f ice on the littoral zone at Richibucto Head eastern New Brunswick. L a Revue de Geographie de Montreal, 3 (1-2): 95-104. Owens, E .H . and S.B. McCann. 1970. The role o f ice in the Arct ic beach environment with special reference to Cape Ricketts, southwest Devon Island, N W T , Canada. American Journal of Science, 268 (5): 397-414. Parker, M.L . and L .A . Jozsa. 1973. Dendrochronlogical investigation along the Mackenzie, Liard and south Nahanni Rivers, N.W.T. Technical Report 10 to the Glaciological Div is ion, Water Resources Branch, Department of the Environment, 187 p. Parmley, L .J . 1987. Downstream ice problems encountered during winter operations of hydroelectric plants in the Peace River. In Proceedings of the CEA/T£PRI/Hydro Quebec Ice Problems Workshop, August 26-28,13 p. Peace-Athabasca Delta Project Group. 1972. The Peace-Athabasca Delta: A Canadian Resource, Summary Report, 144 p. Peace Country Advertising. 1994. Commemorative Edit ion: Peace River - 1919-1994. Peace River, Alberta. Peterson, J .A. 1965. Ice-push ramparts in the George River Basin, Labrador-Ungava. Arct ic, 18 (3): 189-193. Petros. 1822. On certain rocks supposed to move without any apparent cause. American Journal o f Science and Arts, 5(1): 34-37. The Province. 1982. Flood alert 'inconsiderate', Sunday Apr i l 18,1982. Prowse, T.D. 1992. Suspended sediment load during river ice break-up. Proceedings of the 49 t h Eastern Snow Conference, Oswego, New York, pp. 263-267. 174 Prowse, T.D. 1994. Environmental significance of ice to streamflow in cold regions. Freshwater Biology, 32: 241-259. Prowse, T.D. 1996. Significance of river ice to environmental processes and aquatic systems. In Proceedings of International Association of Hydraulic Research 13 t h International Symposium on Ice, Beij ing, China, August 27-30, 3: 865-872. Prowse, T.D. 1997. Personnel communication, National Hydrological Research Institute, Saskatoon, Saskatchewan. Prowse, T.D. and M . Conly. 1996. Impact of F low Regulation on the Aquatic Ecosystem of the Peace and Slave Rivers, Northern River Basins Study Synthesis Report No . 1, Edmonton, Alberta. Prowse, T.D., M . Conly and V . Lalonde. 1996. Hydrometeorological conditions controlling ice-jam floods, Peace River near the Peace-Athabasca Delta. Report prepared for the Northern Rivers Basins Study and Peace-Athabasca Delta Technical Studies, Report No. 103. Prowse, T.D. and N.C. Gridley (Editors). 1993. Environmental aspects of river ice jams. N H R I Science Report No . 5, National Hydrolgical Research Institute, Environment Canada, Saskatoon, Saskatchewan. Prowse, T.D. and V . Lalonde. 1996. Open-water and ice-jam flooding of a northern delta. Nordic Hydrology, 27: 85-100. Reineck, H. 1976. Drift ice action on tidal flats, North Sea. L a Revue de Geographie de Montreal, 30(1-2): 197-200. Reynolds, D . M . 1976. Determining frequency and magnitude of river ice jams from botanical evidence. Unpublished M.Sc. Thesis, Department of Geography, University of Calgary, Alberta. Robson, J . 1752. A n Account of Six Years' Residence in Hudson's Bay, from 1733 to 1736 and 1744 to 1747, London. Rood, S.B. 1996. Instream flows and riparian forests along Alberta's southern and northern rivers. In Northern River Basins Study Project Report No . 66, Proceedings of the Northern River Basins Study Instream Flow Needs Workshop, Walder, G.L. (Editor), October 14-15,1993 and January 6-7,1994, Northern River Basins Study, Edmonton, Alberta. Rosen, S. 1979. Boulder barricades in Central Labrador. Journal of Sedimentary Petrology, 49: 1113-1123. Russell, L C . 1890. Surface geology of Alaska. Bulletin of the Geological Society of America, 1 (116), 121 p. 175 Schweingruber, F .H. 1988. Tree Rings: Basics and Applications of Dendrochronology. D. Reidel Publishing Company, Dordrecht, Holland. Scott, K . M . 1978. Effects of permafrost on stream channel behaviour in Arct ic Alaska. Geological Survey Professional Paper 1068, U.S. Department of the Interior. Scrimgeour, G.J . and T.D. Prowse. 1993. Effects of ice on substrate. In State o f the Art Report on the Environmental Aspects of River Ice, T.D. Prowse and N . Gridley (Editors), National Hydrology Research Institute, Science Report No . 5, pp. 55-60. Scrimgeour, G.J . , T.D. Prowse, J . M . Culp, and P. A . Chambers. 1994. Ecological effects of river ice break-up: A review and perspective. Freshwater Biology, 32: 261-275. Shroder, J.F. Jr. and D.R. Butler. 1986. Tree-ring analysis in the earth sciences. Proceedings of the International Symposium an Ecological Aspects o f Tree-Ring Analysis, G .C. Jacoby, Jr. and J.W. Hornbeck (Compilers), August 17-21, 1986, Tarrytown, New York, pp. 186-212. Sigafoos, R.S. 1964. Botanical evidence of floods and flood-plain deposition. U.S. Geological Survey Professional Paper 485-A, U.S. Department of the Interior, 35 p. Sivertson, D. 1995. Personal communication, former resident at Carcajou and owner of hunting camp at B ig Bend. Smith, D.G. 1974. Remarks on: The effects of ice jams on Alberta rivers. Great Plains-Rocky Mountain Geographical Journal, 3: 122. Smith, D .G. 1979. Effects of channel enlargement by river ice processes on bankfull discharge in Alberta, Canada. Water Resources Research, 15(2): 469-475. Smith, D .G. 1980. River ice processes: Thresholds and geomorphic effects in northern and mountain rivers. In Thresholds in Geomorphology, D.R. Coates and J.D. Vitek (Editors), George A l len and Unwin, London, pp. 323-345. Smith, D .G. and D . M . Reynolds. 1983. Tree scars to determine the frequency and stage of high magnitude river ice drives and jams, Red Deer, Alberta. Canadian Water Resources Journal, 8(3): 77-94. Stong, W.L . and K .R. Leggat. 1992. Ecoregions of Alberta. Alberta Forestry, Lands, and Wildl i fe, Edmonton, Alberta. 59 p. Teversham, J . M . 1973. Vegetation response to fluvial activity in the Lil looet Floodplain. M .A . Thesis, U .B .C . Thomas. 1997. Edmonton Journal Extra, March 19, 1997 & Apr i l 23, 1997. 176 Thorson, R . M . and R.D. Guthrie. 1984. River ice as a taphonomic agent: A n alternative hypothesis for bone "artifacts". Quaternary Research, 22: 177-188. Thurber Consultants Ltd. 1974. Lower Peace River Shoreline Study. Report to B .C . Hydro and Power Authority. Todd, J .E. 1892. Striation of rocks by river ice. The American Geologist: A Monthly Journal of Geology and A l l ied Sciences, 9: 396-400. Tolmachoff, LP . 1928. The geology of arctic Eurasia; problems of polar research. American Geographic Society, Special Publication 7: 87. Vancouver Sun. 1982. Talks called on Peace River flooding. Saturday February 27, 1982. Walker, H.J. 1969. Some aspects of erosion and sedimentation in an arctic delta during break-up. In Symposium on the Hydrology of Deltas, International Association o f Scientific Hydrology, Publication 90, Bucharest, pp. 209-219. Walker, H.J. 1975. Intermittent arctic streams and their influence on landforms. Catena, 2: 181-192. Walker, H.J. and J . M . McCloy . 1969. Morphologic change in two arctic deltas. The Arctic Institute of North America, U.S. Army Research Office, 91 p. Ward, W . H . 1959. Ice action on shores. Journal o f Glaciology, 3 (25): 437. Wentworth, C .K . 1928. Straited cobbles in southern States. Geological Society of America Bulletin, 39. Wentworth, C .K . 1932a. A study of the abrasional work of river ice and of glaciers. Bulletin of the National Research Counci l , 89: 100-102. Wentworth, C .K . 1932b. The geological work of ice jams in subarctic rivers. Contributions on Geology and Geography, Papers 64-67, Washington University Studies-New Series, Science and Technology No. 7, St. Louis, pp 49-80. Wil l iams, G.P. 1973. Seminar on ice jams in Canada. University of Alberta, Edmonton, N .R .C . Technical Memorandum 107, 182 p. Wil l iams, G.P. and D.K. MacKay. 1973. The characteristics of ice jams. In Seminar on Ice Jams in Canada, University of Alberta, G.P. Wil l iams (Editor), Subcommittee on Snow and Ice, Associate Committee on Geotechnical Research, National Research Counci l o f Canada, Ottawa, Ontario, A C G R Technical Memorandum No. 107, pp. 17-35. 177 Wisheu, I.C.and P.A. Keddy. 1996. A model for managing emergent wetland vegetation: indicator species and maximization of diversity. In Northern River Basins Study Project Report No . 66, Proceedings of the Northern River Basins Study Instream F low Needs Workshop, Walder, G.L. (Editor), October 14-15, 1993 and January 6-7, 1994. Northern River Basins Study, Edmonton, Alberta. Wood, J . 1825. On the moving of rocks by ice. The American Journal o f Science and Arts, 9: 144-145. Yaremko, E.K. 1968. Flood and bank erosion study: Fort Vermi l ion, Alberta. Alberta Environment, Water Resources Divis ion, Hydrology Branch, 21 p. APPENDIX A. 1:50,000 S C A L E N.T.S. MAPS O F T H E P E A C E RIVER The following 1:50,000 scale map sheets cover the study reach on the Peace River. They listed in downstream order from the BC/Alberta border to Fort Vermi l ion: • 84 D/4, D/3, D/6, D/7, D/2, • 83 M/15 ,M/16 • 83 N/13 • 8 4 C / 4 , C / 3 , C / 6 , C / 1 1 , C / 1 4 • 8 4 F / 3 , F / 6 , F / 1 1 , F / 1 0 , F / 1 4 • 84 K /3 , K/2 , K/7, K 8 179 A P P E N D I X B . A D I S T A N C E D A T U M F O R T H E P E A C E R I V E R In order to directly refer to locations along the Peace River, distances along the channel center-line were measured manually from 1:50,000 National Topographic System map sheets (See Appendix A ) using dividers to mark 0.5 km intervals upstream from the mouth of the Peace River at the head of the Slave River (note the mouth of the Peace River is also the location where the Riviere des Rochers becomes the Slave River). The head of the Slave River was defined as the datum commencement point to be consistent with the distance referencing systems used by Alberta Research Counci l (Andres, pers. comm., 1995) and the River Engineering Department at Alberta Environmental Protection (Fonstad, pers. comm., 1995). A s the datum is located at the mouth o f the river, river distances increase in the upstream direction. Caution should be exercised when comparing distance values given below with those of other workers. The values depend on the scale o f the maps used and the interpretation o f the river center-line. The datum used by other workers has also been inconsistent. Some have defined the mouth of the Mackenzie River as km 0 while others such as Brit ish Columbia Hydro and contributors to the Northern River Basins Study have defined km 0 km as the W . A . C . Bennett Dam (Hicks and M c K a y , 1996). Below is a listing of the river distances at several well defined points along the Peace River. River distances at 10 km intervals are also plotted on maps A to G in Appendix D. Table B. 1. Locations of major features along Peace River. River Location River Location distance distance (km) (km) 0 Mouth of Peace River 847.50 Smoky River 17 Chenal des Quatres Fourches 853.25 Strong Creek 37 Carlson's Landing 857.60 Mackenzie's Cairn 115 Peace Point 865.70 Shaftesbury Ferry / Tangent Park 270 Fifth Meridian 896.00 Kieyho Park & Campground 338 Vermilion Rapids and Chutes 901.15 Elk Island Campground 356 Wabasca River 911.05 Saddle (Burnt) River 402 Boyer River 911.80 Leith (Little Burnt) River 409 Fort Vermilion (water intake) 928.50 Fairview water intake 419.95 Ft Vermilion bridge (Hwy 88) 939.13 Boucher Creek 432.00 Burnt Woods Landing 942.60 WSC Station, Dunvegan (07FD003) 438.85 Neustueter Flats 942.75 Dunvegan bridge (Hwy 2) 440.65 McDonald's Landing 943.80 Dunvegan Creek 444.60 Ward's Landing 944.00 Hines Creek 479.25 Aspin House 947.10 Ksituan River 491.60 Etna's Landing 962.70 Sawchuck Rapids 499.20 La Crete (Steephill) Landing 966.80 Hamelin Creek 500.05 Steephill Creek 971.30 Fourth Creek 546.55 Tompkins Landing / La Crete Ferry 995.72 Montagneuse River 555.60 Keg River 1011.45 Many Islands Campground 564.70 Buffalo River 1052.75 Sneddon Cr / Cotillion Campground 576.15 Carcajou Indian Settlement (abandoned) 1053.95 Clear River 585.00 Carcajou 1068.42 Pouce Coupe River 583.80 Wolverine River 1071.45 Moonlight Creek 584.25 Sivertson's boat launch at Carcajou 1074.35 BC/Alberta border 588.65 Howard's Landing 1079 Alces River (Clayhurst Bridge) 598.30 Scully Creek 1086 Kiskatinaw River 662.50 Hawk Hills access road 1099 Beatton River 673.35 Notikewin Provincial Park boat launch 1119 Taylor bridge (Hwy 97) 674.55 Notikewin River 1120 Pine River 688.25 Keppler Creek 1137 Moberly River 738.19 Buchanan Creek 1175 Halfway River 751.05 Sunny Valley boat launch 1212 Hudson's Hope 770.00 Cadotte River 1217 Peace Canyon Dam 782.25 Whitemud River 1240 W.A.C. Bennett Dam 785.90 Tar Island River Cruises Camp -795.00 30-Mile Well 810.00 Graham's Flats 811.15 Carmon Creek 822.78 Daishowa Bridge 836.60 Town of Peace River sewage plant 839.15 WSC Station, TPR (07HA001) 839.57 WSC x-section, Town of Peace River 839.85 Peace River bridge (Hwy 2) 840.82 Pat's Creek 841.15 Heart River 846.00 Town of Peace River water intake 181 APPENDIX C. A HISTORY O F ICE B R E A K - U P O N T H E P E A C E RIVER The following is a summary of the salient features of the annual ice conditions, particularly during break-up, on the Peace River upstream of the town of Fort Vermil ion. Information was extracted from historical records, including newspaper accounts, published and unpublished reports, and personal interviews. Sources prior to the 1970's provided largely anecdotal accounts of the general conditions at break-up, including the occurrence of major ice jam floods. By December 1972, information on Peace River ice conditions became more detailed and reliable as B.C. Hydro began ice observations from Hudson's Hope to Lake Athabasca. After two ice jam floods in 1973 and 1974, it was realized that winter f low releases from the W . A . C . Bennett and Peace Canyon Dams could significantly affect the ice regime on the Peace River. Therefore, on March 20, 1974, B .C . Hydro and water resource agencies from the B.C. and Alberta governments agreed to form the Joint Ice Task Force. In order to develop safe and economically feasible regulation strategies, the Task Force collected ice information on an annual basis. Originally, the Task Force studied the period of break-up, however, after the ice jam flood in January 1982, freeze-up was understood to be equally important (Fonstad, 1992). Since the Town of Peace River (TPR) had an established W S C gauging station, and was the largest settlement at risk from ice jams, observations were limited to its vicinity. When break-up was reported as "uneventful", no flooding occurred near or at TPR. However, flooding may have occurred at other locations. The date and the maximum instantaneous water level reached during break-up are presented i f available. For years when no direct observations were made, break-up data was extracted directly from W S C chart records (Conly, 1996). If available, the estimated mean daily discharge is provided. However, these should be interpreted with caution, as f low records during break-up and ice jam flooding are notoriously unreliable. If freeze-up stages were higher than break-up stages, the text in the annual description is underlined. 182 Year Location Date Peak Water Level at Breakup; Mean Daily Discharge on Date of Breakup 1826 Fort Vermilion (June 19) "The water has rose to such a height that the only spot above water is our dwelling house' [Anonymous quote cited by Batt et al. (1988)]. A n ice jam occurs near Fort Vermil ion according to Hudson's Bay Company archives (Gerard and Karpuk, 1979). A quote in Sheridan Lawrence's biography by Myles (1965) states that "Never before in l iving memory had the Peace hurled such havoc upon them; never before had a flood of such proportion occurred." Mrs. Rivard, a resident in Fort Vermil ion since 1917, stated that she had heard that the 1888 ice jam flood was just as bad as the 1934 flood (Gerard and Karpuk, 1979; Yaremko, 1968). 1894 Fort Vermilion 255.7 m A n ice jam occurs near Fort Vermi l ion according to Hudson's Bay Company archives (Gerard and Karpuk, 1979). 1910 TPR (March 26) Earliest recorded break-up since 1898. 1914 TPR The town is flooded to 6.1 m above an undefined "normal" level (Fonstad, pers. comm., 1996). 1934 TPR (Apri l 8) 319.43 m Break-up on the Smoky River initiated break-up on the Peace River to a point 9.6 km downstream of the TPR. A n ice jam at that point caused water levels near the town to rise to a level not exceeded since 1914. Water reached some Hudson's Bay Company boats in winter quarters. Peace River ice upstream of Smoky River was still intact (BC Hydro & Power Authority, 1975; Peace River Record, Apr 20, 1934). Fort Vermilion (Apri l 27) 258.4 m 1876 Fort Vermilion 254.8 m 1888 Fort Vermilion (May 7, 11:00) 258.7 m After an unusual winter with extreme and rapid variations in weather (Carder and Siemens, 1971) break-up was eventful with an ice jam forming 3.2 km downstream of Fort Vermi l ion near an island. The settlement was flooded with up to 1.8 m of water as far inland as 91 m. The open 183 water return period of this event was estimated to be > 10,000 years (Yaremko, 1968). Two photos in Yaremko (1968) show that ice jammed to at least the bankfull level. Ice shove was also shown several metres above bankfull. 1948 T P R Ice jam flood (Peace Country Advertising, 1994). 1950 Fort Vermilion 252.3 m Photographs examined by Gerard and Karpuk (1979) reveal the occurrence of relatively high break-up stage, possibly due to an ice jam. 1954 TPR (May 6) Latest break-up at T P R since 1898. 1960 TPR (Apri l 16) 313.20 m; 1274 m 3/s A n uneventful break-up year (Nuttall, 1974). 1961 T P R (Apr i l 20) 311.81 m; 1217 m 3 /s A n uneventful break-up year (Nuttall, 1974). 1962 TPR (Apri l 16) 314.17 m; 1500 m 3 /s A n uneventful break-up year (Nuttall, 1974). 1963 TPR (Apri l 20) 317.51m; 4275 m 3/s In this relatively high runoff year (Andres, 1981), water levels slowly rose on Apr i l 13. By Apr 19, the water level was at an elevation of 313.36 m (Fonstad, 1992). The W S C gauge at T P R was broken from Apr i l 16-19. On Apr i l 20, an ice jam caused an additional 4 m increase in stage (Fonstad, 1992; Nuttall, 1974). Fort Vermilion (Apri l 21,17:55) 255.4 m Break-up on Apr i l 21 increased the stage to 247.79 m (Conly, 1996). A n ice jam near the town raised levels to 255.4 m according to residents interviewed (Gerard and Karpuk, 1979). 184 1964 Hudson Hope (Apri l 29) The ferry was stopped by ice jams and shore ice (BC Hydro & Power Authority, 1975). TPR (Apri l 22) 313.71 m; 1100 m 3/s A n uneventful break-up year. The break-up stage was estimated from original W S C records (Conly, 1996). Fort Vermilion (May 2, 17:45) 248.634n; 1730 m 3/s The break-up stage was estimated from original W S C records (Conly, 1996). 1965 TPR (Apri l 14) 313.61 m; 2038 m 3/s A n uneventful break-up year (Nuttall, 1974). Fort Vermilion (Apri l 26, 16:00) 253.6 m Break-up was estimated from original W S C records with a stage of 248.41 m and a f low of 2550 m 3 /s (Conly, 1996). A n ice jam downstream of the settlement later flooded the airport road with 0.6 m of water and caused bank full flows at the settlement (Residents Mrs. Clarke & Mr . K idd quoted in Yaremko, 1968). 1966 TPR (Apr i l 5) 314.52 m; 1727 m 3/s A n uneventful break-up year (Nuttall, 1974). Fort Vermilion (May 4, 00:30) 251.1 m Break-up was uneventful. The break-up stage estimated from original W S C records was 249.57 m with a f low of 1730 m 3 /s (Conly, 1996). The stage peaked at 251.1 m (Gerard and Karpuk, 1979) possibly because of an ice jam. 1967 T P R (Apr i l 30) 313.09 m; 679 m 3/s A n uneventful break-up (Nuttall, 1974). Fort Vermilion (May 3,18:00) 248.98 m A n uneventful break-up. The break-up stage was estimated from original W S C records (Conly, 1996). 1968 TPR (Apri l 16) 311.96 m; 679 m 3/s A n uneventful break-up (Nuttall, 1974). 185 1969 TPR (Apri l 15) 314.88 m; 1415 m 3/s A n uneventful break-up (Nuttall, 1974). 1970 Taylor (February 2-6) 407.82 m A huge ice jam near Taylor caused water levels to rise 5.5 m resulting in overbank flooding of agricultural land on right bank and some damage to Peace Island Park (Province, Feb 4,1970). Clayhurst (February 4) A n ice jam downstream of Clayhurst caused backwater flooding for 35 km ( B C Hydro and Power Authority, 1975). TPR (Apri l 13) 313.37 m; 2000 m 3/s A n uneventful thermal break-up. The break-up stage was estimated from original W S C records (Conly, 1996). 1971 TPR (Apri l 19) 313.06 m; 1472 m 3/s A n uneventful thermal break-up year (Nuttall, 1974). 1972 TPR (Apri l 23) 314.85 m; 1812 m 3 /s A n uneventful break-up (Nuttall, 1974). Fort Vermilion 250.06 m The break-up stage was estimated from original W S C records and may be an underestimation of the actual value (Conly, 1996). 1973 TPR (Apri l 12, 11:30) 318.19 m; 2789 m 3 /s On Apr i l 10, water levels rose due to increasing snowmelt within the Peace River Basin. A review of gauge records indicated that Smoky River flows were increasing rapidly while the Peace River was slightly decreasing. On Apr i l 11, an initial 0.8 m stage increase levelled off until early on Apr i l 12, when break-up was initiated (Conly, 1996). Fonstad and Quazi (1994) believed the Smoky River was responsible for break-up, although no hard evidence was presented. A n ice jam downstream of T P R caused stages to rise over 4 m in an 11.5 h period (Fonstad, 1992) causing minor overtopping of dykes built in 1972 (BC Hydro & Power Authority, 1979). 186 1974 TPR (Apri l 20, 16:00) 317.52 m; 3624 m 3/s The events of 1974 are similar to those in 1973. The winter had an above normal snowpack and the spring had normal temperatures (Gerard, 1975). The maximum upstream extent of the ice cover was to a point 80 km downstream of the Peace Canyon Dam (Andres, 1981). Under the combined effects of rising temperatures and increasing tributary inflow, water levels increased slowly from Apr i l 17-19. A t 06:39, Apr i l 19, a dynamic break-up front progressed through the town (Nuttall, 1974). F low records indicated that break-up was initiated by the Smoky River as flows increased by an order of magnitude, during the previous 4 days. The Smoky River itself had a break-up associated with a jam failure (Gerard, 1975). A t least 3 jams on the Peace River were observed within a distance of 26 km downstream of the TPR. The first jam, 3 km long, was located 5 km downstream of the T P R at the end of a reach artificially weakened by dusting and scraping prior to break-up. This jam was blasted at 16:05 on Apr i l 20 (Nuttall, 1974), just after the peak stage was reached at the TPR. The blasting caused the ice to run and jam against solid ice between an island and the right bank. Authorities also blasted this jam. The last jam, 10 km long, some 26 km downstream of the T P R was probably the largest and caused a 4 m local rise in water levels (Gerard, 1975). Ice remained upstream of the Smoky River confluence until Apr 22. N o damage to the T P R was recorded. Daishowa-Marubeni Pulp Mill (Apri l 21) The area known as Six M i le Farm, near km 832, and the land where the Daishowa-Marubeni pulp mi l l is now located, near km 822, was partially flooded during break-up ( B C Hydro & Power Authority, 1979). Near Whitemud River (Apri l 24) A farm building on the left bank at km 781 was damaged by ice ( B C Hydro & Power Authority, 1979). Sunny Va l ley (Apr i l 23-25) Severe ice jam floods occurred as water was backed up behind successive positions of the break-up front. Farms were flooded (Andres, 1981; B C Hydro and Power Authority, 1979). Carca jou (Apr i l 22-23) Farms on the left bank near km 580 were damaged by ice and flood water. A telephone line crossing the river was brought down by ice floes. The Carcajou Indian settlement was covered by ice floes after the jam released (Andres, 1981; B C Hydro and Power Authority, 1979). Fort Vermilion 251.10 m The break-up stage was estimated from original W S C records. The elevation given may underestimate actual value (Conly, 1996). 1975 Near Clear River (January-February) Freeze-up stage increases of 3.0 - 4.5 m were observed near the BC/Alberta border. Two gravel operations, including one near Clear River were flooded by about 1 m of water (BC Hydro and Power Authority, 1979). A mild winter resulted in the relatively late arrival of freeze-up in mid-January. Two cold periods caused rapid ice formation and upstream advance and subsequent ice consolidation and shoving. Ice was observed near the tree line at some locations. The highest stage was reached during freeze-up: 314.94 m. Experimental dusting and blasting of the ice cover to induce localised flooding was carried out near Bewley Island by Alberta Environment in light of the previous year's events. The dusting had no significant effect, but blasting caused localised deterioration of the ice cover. Following a warm spell, the ice broke up between Dunvegan and the TPR. Low runoff, minimal variation in releases and a thin snowpack made this break-up relatively uneventful (BC Hydro and Power Authority, 1979). The Peace broke up 3 days prior to the Smoky River. Although no flooding occurred, ice was piled heavily against islands and banks (Acres, 1984). The break-up stage was estimated from original WSC records (Conly, 1996). The winter of 1975/76 had normal temperatures in the north and warm temperatures in the south. Freeze-up began in early December under cold temperatures. The stage increase at freeze-up was 3.0 - 4.5 m. During a warm period in late December, a 100 km length of the ice downstream of Dunvegan consolidated into 32 km just upstream of the TPR. The result was 3 m high shear walls and ice piled 2 m above islands near the Smoky River. Break-up was thermally induced with only minor ice jam flooding of the Smoky River at its mouth (BC Hydro & Power Authority, 1977a). Vermilion Chutes The falls were reported as being submerged by an ice jam (BC Hydro & Power Authority, 1979). 1977 T P R (March 12) 313.25 m; 847 m3/s A mild winter resulted in a late freeze-up and limited ice advancement (to Saddle River near km 910). The Smoky River broke on April 6. The highest stage of 314.94 m was recorded on January 8. Break-up occurred on the earliest date since 1910 and was by progressive thermal erosion of the ice cover (BC Hydro & Power Authority, 1977b). The uneventful break-up stage was estimated from original WSC records (Conly, pers comm., 1995). TPR (April 17) 314.53 m; 2250 m3/s 1976 TPR (April 12) 315.16m; 2270 m3/s 188 1978 TPR (April 14) 313.22 m; 1450 m3/s The winter of 1977/78 was near normal with the ice cover advancing to near Clayhurst (km 1078) (Parmley, 1987). Freeze-up resulted in stage increases of up to 7.6 m along a reach upstream of the TPR (km 880 to km 965). Stages increased also between TPR and Carcajou causing shear ridges and overbank flooding. At TPR, the stage peaked at 315.10 m on Dec 18 (Parmley, 1987). Break-up was similar to the previous 3 years in that it was thermally induced. Nevertheless, since stages were relatively high prior to break-up, minor flooding was witnessed along banks between Shaftesbury Ferry and the TPR (BC Hydro & Power Authority, 1978). The break-up stage was estimated from original WSC records (Conly, 1996). 1979 Taylor (February 18 - peak ice/water level near break-up) 407.2 m; 1340 m3/s Under extremely cold temperatures in February 1979, the ice cover advanced quickly upstream to a point 19 km upstream of Taylor on March 1. The ice cover progressed through the relatively steep reach in B.C. by successively forming, collapsing, and shoving (Keenhan et al., 1982). The freeze-up jams that formed were up to 6 m thick and caused stage increases up to 4 m. Jams were typically 500 m long with the jam toes frequently at the downstream end of islands (Keenhan et al., 1982). Prior to freeze-up, BC Hydro was releasing 1770 m3/s. This represented about 90% of full output (Ketchum, 1996). When freeze-up near Taylor was imminent, BC Hydro reduced flows to 1000 m3/s. Despite the flow reduction, flooding occurred. Farms and vehicles were damaged, cattle drowned, and moose were trapped on gravel bars and islands . Until that date, only the summer floods in 1948 and 1964 were higher (by 1.5 m and 0.8 m respectively)(Keenhan et al., 1982). Dunvegan During the January freeze-up, the water level increased up to 5.7 m under increasing discharges of 1580 to 1950 m3/s. Shoving was observed from Dunvegan to 70 km upstream of TPR (Acres, 1984). TPR (April 30,11:15) 318.75 m; 4100 m3/s The break-up of 1979 was unlike the previous 4 years. A large winter snowpack along with warm April weather provided the setting for an eventful spring. As pre-break-up flows were reduced substantially by B.C. Hydro, the events at the TPR were largely a result of the events that took place on the Smoky River. On April 25, Smoky River ice broke up at Watino (Garner and Quazi, 1979). On April 29, the Smoky River ice run reached the Peace River and caused a massive jam 3A of the way into the intact Peace River ice. On April 29, the stage at the TPR exceeded the 1974 peak stage by 0.76 m, causing the Heart River to overtop its banks at some locations. As the Smoky River flows increased from 225 - 1590 m3/s between April 25-30, the Peace River ice cover was slowly being lifted until it broke loose at 10:00 on April 30. This caused a 4.6 m high flood wave, which nearly overtopped the dykes (Garner, 1979). As the ice run passed town, ice jammed on the left side of Bewley Island causing floes to shift to the right side. A jam formed 1.6 km downstream of Bewley Island causing backwater flooding [a stage increase of 3.7 m according to Fonstad (1992)] of North Springfield and the Kinsmen Park. Sewers backed up near the Heart River and a power pole was destroyed by an ice floe. The West 189 Peace subdivision was evacuated. Over the next two days, the jam moved 32 km downstream. Although stages were high, no flooding was recorded downstream of the TPR (Garner and Quazi, 1979) . Ice upstream of the Smoky River confluence gradually melted and went out without incident. Fort Vermilion (May 10) 252.65 m The break-up stage estimated from original WSC records (Conly, 1996) was 251.18 m. An ice jam 13 km long was observed by McLean and Anderson (1980) 56 km upstream of Ft. Vermilion. It caused peak stages to reach an estimated 252.65 m at the Highway 88 bridge and left debris on the banks. The estimated velocity of the ice run was 2.5 m/s (McLean and Anderson, 1980). 1980 TPR (April 18,01:15) 312.17 m; 1040 m3/s A mild winter was the setting for a 1.0-2.4 m stage increase at TPR during freeze-up (Acres, 1980) . Although there was local shoving, no damage was reported (Acres, 1984). The maximum winter stage was 313.55 m on Jan 10. Break-up was uneventful and was estimated from original WSC records (Conly, 1996). 1981 TPR (April 5, 01:30) 314.06 m; 2590 m3/s The maximum winter stage was 314.65 m on Dec 19. Break-up was uneventful and was estimated from original WSC records (Conly, 1996). 1982 Clear River (March 23) An ice jam was observed at km 1045 near Clear River. Flooding occurred along shear walls (BC Hydro, 1982). TPR (Jan 8,01:30) 318.10 m; 2080 m3/s The flood in 1982 differed from most previous events since break-up occurred mid-winter. Under cold temperatures, as low as -38°C, the ice cover rapidly advanced in late December and early January. During that period, flows from the Peace Canyon Dam were fluctuating. On December 25-26, 1981, flows were 883 m3/s. By December 29-31, 1981, flows were 1811 m3/s. Flows were temporarily dropped to 998 m3/s on January 1-2,1982, but were back up to 1826 m3/s by January 6,1982. Freeze-up at the TPR occurred on January 2, just as the high flows from December 31 passed (the flow through time from the dam to TPR is 3 days). Initial freeze-up caused a large stage increase, but it soon dropped since flows were temporarily reduced at the dams (Jan 1-2). Continuing cold temperatures and low flows caused a thin ice cover to progress to Dunvegan by Jan 6. Under increasing flows required for electricity generation, the ice cover broke and consolidated above a jam 14 km downstream of Dunvegan. Ice levels were up to 1 m above the banks near Green Island (km 930)(BC Hydro, 1982). After several hours this jam broke at about 18:00 on January 7, releasing a surge estimated at 2500 m3/s (Neill and Andres, 1984). The surge dynamically broke 155 km of ice until jamming 20 km downstream of the TPR (Parmley, 1987). The 60 km of consolidated ice that jammed caused a 3.4 m stage increase over 190 a 3.5 hour period (Fonstad, 1992). Although the dykes were not overtopped (maximum stage was 1.5 m below top of dykes), residents were on flood alert. Basements of 20 homes were flooded with up to 28 cm of water in West Peace (Province, 1982; Vancouver Sun, 1982) and the Shaftesbury Ferry was almost lifted by ice (BC Hydro, 1982). A s the ice cover continued to form and progress upstream until March 4, stages fluctuated but remained relatively high (316.5 m) causing groundwater levels to remain high in West Peace River. Basements were flooded in some 60 residences ( N R C C , 1990). Spring break-up occurred on Apr i l 25 at 23:00 by thermal erosion and was uneventful. The stage at break-up only reached 315.94 m under a f low of 2020 m 3/s (Conly, 1996). Prior to break-up, snow was ploughed and the ice was sanded from T P R to near the Daishowa-Marubeni pulp mi l l in order to channelize overflow water and attempt to speed up the formation of a lead and melting (BC Hydro, 1982). Ice was also blasted at the mouth of the Heart River, but appeared to have little affect. Fort Vermilion (May 1) 250.71 m The break-up stage was estimated from original W S C records (Conly, 1996). 1983 Reach la (Apri l 13) B.C. Hydro (1983) aerial observations indicated that ice was being shoved 2.2 m above the waterline near Clear River (km 1054), and local flooding took place on both banks between Many Islands (km 1011) and Dunvegan (km 943). TPR (Apr i l 24) 312.85 m; 1360 m 3 /s Freeze-up occurred at the T P R on January 4 causing a stage increase of 3.32 m and a maximum stage of 315.27 m under a flow of about 1510 m 3/s (Fonstad, 1992). The flow from the Peace Canyon Dam prior to freeze-up was a relatively high 1350 m 3 /s. Even though the winter was relatively mi ld, causing slower ice progression, low flows later in the season permitted ice to progress to 3.2 km upstream of Site C near Fort St. John (km 1142). This was the furthest ice had formed since regulation (BC Hydro, 1983, Davies et al., 1984). By late Apr i l , 80 % of Peace River ice was thermally melted and the break-up of the Smoky River was uneventful. The break-up stage was estimated from original W S C records (Conly, 1996). 1984 TPR (Apr i l 13,08:30) 313.18 m; 1600 m 3/s A mi ld winter and low snowpack in Alberta made for a generally uneventful ice season. B .C. Hydro did not make any ice observations (B.C. Hydro, 1986). Freeze-up was on December 17, with the peak stage on Dec 25 at 314.68 m (Fonstad, 1992). Stages were well below the dykes (319.8 m) and basements (315.3 m). Break-up was thermally induced and uneventful. The break-up stage was estimated from original W S C records (Conly, 1996). Fort Vermilion (Apri l 16, 05:30) 248.31 m The break-up stage was estimated from original W S C records (Conly, 1996). 191 1985 TPR (Apri l 10, 16:00) 314.80 m; 2400 m 3/s The ice season of 1985 was uneventful. The maximum stage was 315.9 m on January 5. Break-up was thermally induced. The break-up stage was estimated from original W S C records (Conly, 1996). Fort Vermilion (May 2, 16:00) 250.60 m The break-up stage was estimated from original W S C records (Conly, 1996). 1986 Clayhurst Freeze-up stage increases of 4 m interrupted construction of the Clayhurst Bridge. Foundation work was stopped ( B C Hydro, 1987). TPR (Apri l 17, 01:30) 313.63 m; 2550 m 3 /s The winter was generally mi ld with cold weather in November and early December. Freeze-up was early on December 4, 1985, but the ice progressed slowly due to warm weather until mid-January. A t TPR, freeze-up caused a stage increase of 3.12 m and a peak stage of 315.09 m (Fonstad, 1992), still wel l below the dykes. Break-up was uneventful, and B.C. Hydro made no aerial observations. The break-up stage was estimated from original W S C records (Conly, 1996). Fort Vermilion (Apri l 26, 20:00) 250.47 m Break-up was thermally induced and the stage was estimated from original W S C records (Conly, 1996). 1987 Dunvegan (March 10) 349.10 m; 1700 m 3/s During freeze-up, ice consolidated and jammed due to high flows (B.C. Hydro was releasing 86% of full output). On March 3, the water level was 348.4 m. The thick ice caused stage increases that had not been exceeded in 30 years, according to Mr . Friesen, a resident of Dunvegan. He reported that 2 m of water flooded some fields and roads and ice damaged his water pump as it shoved up the banks ( B C Hydro, 1989a). TPR (Apri l 6; 04:40) 315.20 m; 2350 m 3 /s The winter o f 1986/87 was the second mildest between 1967 and 1987, causing freeze-up to occur late on January 23. Freeze-up increased the stage by 3.69 m to a maximum of 315.66 m on January 25 (Fonstad, 1992). Break-up was thermally induced and uneventful. The break-up stage was estimated from original W S C records (Conly, 1996). Fort Vermilion (Apri l 21,18:30) 251.66 m Break-up stage estimated from original W S C records (Conly, 1996). 192 1988 Dunvegan (February 5-6) 346-347 m Freeze-up stages were reported by a Dunvegan resident, Mr . Friesen, as being 2 m below 1987 levels (BC Hydro, 1989b). With the mildest winter on record (between 1967arid 1988) and high flows (1700 m 3 /s from Peace Canyon Dam), freeze-up occurred on February 1 - the latest on record. The freeze-up stage increase was 2.76 m to a peak stage of 314.76 m (Fonstad, 1992). The ice cover progressed to 29 km upstream of Dunvegan by Feb 19. Break-up of the Peace River on March 11 was thermally induced and uneventful. The break-up stage was estimated from original W S C records (Conly, 1996). On Apr i l 9, the Smoky River broke-up without incident (BC Hydro, 1989b). Fort Vermilion (Apri l 16, 12:30) 249.23 m Break-up was thermally induced and was estimated from original W S C records (Conly, 1996). Except for late December and mid-January, winter temperatures were about normal. Freeze-up took place on December 31 under low flows released over the Christmas period. The stage increased 3.57 m during freeze-up to an elevation of 315.10 m (Fonstad, 1992). The maximum stage of 315.70 m was reached on January 5 under a f low of 1275 m 3 /s. Stages remained above 315m over January and February since power demands required high releases by B .C . Hydro. Basements were within 0.1 m of being flooded (BC Hydro, 1990a). However, break-up was uneventful with the peak stage estimated from original W S C records (Conly, 1996). Fort Vermilion (May 4,13:55) 249.61 m Break-up was thermally induced and the break-up stage was estimated from original W S C records (Conly, 1996). 1990 Clayhurst (February 20) During freeze-up, ice consolidation caused a large stage increase near Clayhurst Bridge. 60 m of highway on the southern approach was flooded under 0.75 m of water as f low was diverted into an abandoned channel (BC Hydro, 1990b). Winter temperatures fluctuated from near freezing in November and early December to -33°C in late December to above freezing in late January to -25°C in mid-February. The freeze-up front fluctuated with the changes in temperature and flows. Ice consolidation was common. Freeze-up occurred at T P R on January 10. The stage increase was 3.44 m to a maximum stage of 314.98 m (Fonstad, 1992). The rate of further ice progression varied until reaching its maximum point upstream near Clayhurst on February 20. Spring break-up was thermally induced and uneventful TPR (March 11) 314.65 m; 1880 m 3 /s 1989 TPR (Apri l 23, 18:30) 314.10 m; 1630 m 3/s TPR (Apri l 9,15:00) 313.49 m; 1370 m 3/s 193 as the Peace River broke out prior to the Smoky River. The break-up stage was estimated from original WSC records (Conly, 1996). Fort Vermilion (April 22, 18:30) 250.25 m The break-up stage was estimated from original WSC records (Conly, 1996). 1991 Dunvegan (April 6, 20:30) 343.56 m The break-up stage was estimated from original WSC records (Conly, 1996). TPR (April 17, 03:00) 312.71 m; 1900 m3/s During freeze-up, the stage increased 2.49 m to a maximum of 314.25 m (Fonstad, 1992). Break-up was uneventful with the stage estimated from original WSC records (Conly, 1996). Fort Vermilion (April 24, 19:30) 250.46 m The break-up stage was estimated from original WSC records (Conly, 1996). 1992 TPR (February 28, 08:30) 319.90 m; 2820 m3/s The winter of 1991/92 was very mild resulting in freeze-up occurring on February 11-12 (Assaf, 1995). Freeze-up occurred during a period when B.C. Hydro had released the highest freeze-up flows since December 1967 (Peace Canyon Dam flow was 1830 m3/s and TPR flow was 2067 m3/s). An unusually large freeze-up stage increase of 4.7 m brought the stage to 316.78 m (Fonstad, 1992). Power commitments prevented BC Hydro from reducing flows in mid-February, even though they were made aware of the risk of basement flooding in West Peace. A thin weak ice cover progressed to near Saddle (Burnt) River (km 911) before breaking up. Both regulated flows and snowmelt from tributary basins were high. Break-up was unexpected since flow records showed no perceptible increase (Fonstad, 1992). Break-up began at midnight of February 26 (from gauge records and resident interviews) and released a surge, which started a downstream progression of temporary jams and surging ice runs. As the break-up front stalled behind a jam near Shaftesbury Settlement, at least one family was evacuated due to flooding. After that jam released, the break-up front progressed downstream before jamming again at 26 km downstream of the TPR (Assaf, 1995). A classical spring jam, nearly 50 km long, increased the stage 3.14 m above the pre break-up stage over a 5 hour period (Fonstad, 1992). The peak stage of 319.90 m overtopped the dykes in localised areas and caused 4000 people to evacuate under a state of emergency. The 1992 event was unusual in that the Smoky River did not initiate it. Fort Vermilion (April 15, 08:30) 250.17 m Break-up was thermally induced. The break-up stage estimated from original WSC records (Conly, 1996). 194 1993 Dunvegan (Apri l 15, 20:40) 343.22 m; 1700m 3 /s On January 4, freeze-up caused the peak stage of 345.60 m. The break-up stage was considerably less than recorded in original W S C records (Conly, 1996). TPR (March 29, 00:15) 313.20 m; 1650 m 3/s Break-up was thermally induced and uneventful. The break-up stage was estimated from original W S C records (Conly, 1996). Break-up was thermally induced and uneventful. The break-up stage was estimated from original W S C records (Conly, 1996). 1994 Carcajou (Apri l) A spring ice jam scoured the right bank near a hunting camp on B ig Bend (km 626). Several metres of bank were eroded and a part of the campsite was destroyed by ice. The event was described as the worst jam since the 1960's (Sivertson, pers comm., 1995). 1995 TPR The ice season was uneventful. 1996 TPR (Apri l 20-21) 315 m Break-up along the Peace River in 1996 occurred as a series of 3 break-up fronts progressing independently downstream for at least part of the break-up period. The upstream front thermally retreated from Taylor (Fonstad, pers. comm., 1995). From personal observations, the ice broke up late on Apr i l 20 initially along leads on the right bank downstream of the Smoky River. Water levels increased to a level approximately 1 m below the Heart River bridge but were well below the dykes. The only significant effect of break-up was the volume of ice shoved up banks. A long the left and right banks, ice was pushed up to 1 m above the bank top. Sunny Valley (Apri l 22) In response to a second break-up front which initially formed near the Daishowa-Marubeni Pulp M i l l , a jam formed in the Sunny Val ley area (km 750) causing at least 4 farms to be flooded between Whitemud River and Sunny Val ley (Fonstad, pers. comm., 1996). Tompkin's Landing (Apri l 22) A third break-up front, which formed, 16 km downstream of Sunny Val ley terminated in a jam near Tompkin's Landing. Downstream of the jam, ice was fractured from surging flows (Fonstad, pers. comm., 1996). Fort Vermilion (Apri l 21, 21:50) 250.21 m 195 Fort Vermilion (Apri l 24, 08:00) 252.8 m (est.) After a jam at Tompkin's Landing released, an ice run surged downstream before jamming near Fort Vermil ion. The jam was approximately 175 km long with its tail near Carcajou. Low-lying areas were flooded but the settlement was not flooded (Fonstad, pers. comm., 1996). The peak stage was estimated to be 1.08m the below peak stage on June 16,1990. 1997 TPR (Apri l 19,08:30) 319.92 m The winter of 1996/97 was colder than normal and had a snowpack 36% above normal (Thomas, 1997). The ice cover progressed as far as Taylor. Runoff increased substantially in Apr i l causing a rapid melting of the ice front. On the day prior to break-up, the ice had retreated 40 km. Early on Apr i l 19, the stage began its rise eventually peaking around 08:30 on Apr 19. The peak stage resulted from a jam 10 km downstream of the T P R (Fonstad quoted by Thomas, 1997). The high stage of the Peace River backed up the Heart River eventually causing water to surge through gaps in the dykes near the Heart River bridge (Thomas, 1997). About 65 shops, offices, and markets were flooded by up to 2 m of water (Gi l l is, 1997) causing over $50 M damage (Thomas, 1997). 4000 residents were evacuated in West Peace, North Peace, and downtown. A local state of emergency was issued 1 hour prior to the peak stage. To drain the water, dykes were intentionally breached near Pats Creek and the Heart River bridge (O'Connor, 1997). 196 APPENDIX D. PEACE RIVER STUDY REACH MAPS Legend for maps A-G Channel margin (lower limit of mature forest) Floodplain limit (approximate) Bars Sub-reach boundaries Boat launch Bridge Pipeline crossing I I Islands km 1000 River distance (km) from mouth of Peace River. Bank profile and riparian vegetation survey Max. elevation of tree damage above field datum (mY/ Mean elevation of tree scar above field datum (ml, REACH 3 a Location of sub-reaches REACH 1a 4 Key to map sheets A.2 Kilometres VO •VI vo 00 A P P E N D I X E . S E L E C T E D B A N K P R O F I L E S Site 729 Ft. Vermilion bridge Site 551 550/) Tompkin 's Landing Site 618 Carcajou ^ B.C./ Alberta border Site 116 Site 446 Site 300 Site 229 0 20 40 I I I Kilometres SITE 116 S U B - R E A C H 1a.5 km: 1002.6 Reach: Site: Location: Photo: Date: 1a.5 116 km 1002.6 left bank 7.17 Jun 28, 1995 206 SITE 229 SUB-REACH 1b.1 km: 933.8 Reach: 1b. 1 Site: 229 Location: km 933.8 island attached 50 40 30 20 10 0 DISTANCE FROM WATERLINE (m) 2 0 7 SITE 300 SUB-REACH 2a.2 km: 850.0 0 10 20 30 4 0 DISTANCE FROM WATERLINE (m) SITE 394 SUB-REACH 2a.7 km: 730.5 DISTANCE FROM WATERLINE (m) 209 SITE 446 SUB-REACH 2b.3 km: 665.6 Note: Photo was taken viewing upstream from site, not along transect 50 100 150 DISTANCE FROM WATERLINE (m) 210 SITE 551 SUB-REACH 2b.8 km: 554.6 2b. 8 551 km 554.6 island attai Jul 23, 19. DISTANCE FROM WATERLINE (m) 211 SITE 618 SUB-REACH 3a.3 km: 494.8 300 200 100 DISTANCE FROM WATERLINE (m) SITE 729 SUB-REACH 3a.5 km: 423.3 _ Reaeh-'—- 3a.5 Site: 729 Location: km423.3 island near Ft. Vermilion bridge Photo: 40.9 Date: . Jul 8, 1995 GREEN ALDER & BALSAM POPLAR (8 m tall) ELEVATION ABOVE FIELD DATUM (m) 6 4 2 0 -2 SILTY SAND ICE DAMAGED WILLOWS (1m tall) ICE DAMAGED WILLOWS (3 m tall) 100 viewing downstream 200 DISTANCE FROM WATERLINE (m) 213 APPENDIX F. A C C E S S POINTS A L O N G T H E P E A C E R I V E R Since the Peace River Val ley is deeply incised below the surrounding prairie landscape, access to the river is possible only at a relatively few locations along the river. Below is a list of the most reliable sites for boat launching and or camping along the river in Alberta upstream of Fort Vermil ion. A four-wheel drive vehicle is strongly recommended for towing and launching a boat along the Peace River because of the thick mud along the river and slippery roads during rainstorms. The best boat launches in terms of ease of access under most weather conditions are denoted with an asterisk. Useful information for anyone travelling the Peace River is available in "Recreation sites in the upper Peace River Va l ley" (Anonymous, 1987), "Recreation sites in the lower Peace River Va l ley" (Anonymous, 1991), and Flygare (1983). 1) Fort Vermilion area (km 418)* The Fort Vermi l ion boat launch is located at the foot of an old ferry crossing on the right bank approximately 100 m downstream of the Fort Vermil ion bridge. This gravel launch has a silty base which may be soft when wet. Four wheel drive vehicles are recommended. The boat launch may be accessed from Highway 88 by turning north onto a gravel road 500 m west of the turn off to Fort Vermil ion. Fol low the gravel road until it reaches the shoreline near overhead powerlines. There are no formal campsites in the Fort Vermil ion area. Informal campsites with no services include the land adjacent to the boat launch, and an abandoned campsite named "The Pines" on the left bank of the river directly opposite the boat launch. Alberta Transportation and Utilities previously operated the Pines. It was vandalized and infrequently used prior to our visit. A s an alternative, it is possibile to access the river directly in front of the Fort Vermi l ion settlement. 2) Bluemenort (km 430) There is primitive camping accessed by a dry weather dirt road which crosses private land. Boats must be carried to the river. 3) Neusteuter's Flat (km 440) There is primitive camping at the site at an old sawmill accessed by a dry weather road. Boats must be launched over cobbles and boulders. 214 4) Etna's Landing (Atlas Landing) (km 492) Etna's Landing is an informal boat launch and campsite frequented by residents of L a Crete. Two boat launches at the site are poor, and depending on the weather and stage of the river, may be unusable. Boats may need to be carried 20 metres over muddy shoreline. Etna's Landing can be accessed by driving west from the Country Corner Restaurant in L a Crete. The paved road becomes a gravel road and passes schools and a cemetery before being joined by a road from the left. Continue on the road, which becomes sand-based as it passes to the north of Linton Lake. A t the Y-intersection, stay right (left leads to a golf course). The road is generally in good condition during dry periods but may be impassable when wet due to the soft road bed. The travel time from Highway 697 in L a Crete to Etna's Landing is approximately 25 minutes. 5) La Crate Landing (sic.) (km 496) The river can be accessed by a steep dirt road across private land 3 km upstream of Steephill Creek. There are no provisions for camping at this location of an old ferry landing. 6) Steephill Creek (km 500) Poor boat launch used by locals. 7) Tompkins Landing/La Crete Ferry (km 546)* Boats may be launched at Tompkins Landing on either bank but it is preferable to launch on the right bank adjacent to the ferry ramp (lower gradient and wider turn around area). The launch has a good gravel base but is slippery when wet. Informal camping is possible immediately downstream of the ferry on the right bank. There are two clearings in the spruce forest suitable for camping. The location is particularly noisy as the ferry runs with diesel engines 24 hours a day during the summer. Tompkins Landing is reached by turning east onto Highway 697 from Highway 35 north of Paddle Prairie. A pay-phone and outhouse are located on left bank of river. 8) Carcajou (km 585.0) There are two locations with river access in the Carcajou area. Carcajou is located 38 km east of Highway 35 near K e g River. The road is well-graded but can be muddy during wet periods. There are no services and no formal camping sites. The main boat launch is on private land and is found by fol lowing the dirt road, which continues, straight from the end of the main gravel road to Carcajou. Permission should be obtained prior to using the launch which requires traveling through farm land. The condition of the launch is good, but is steep, and may be muddy and slippery during wet periods. During low flows, expect to carry boats to the waterline. A n area for camping is available with permission. The second location for river access is Howard's Landing a few kilometres upstream. To reach it, turn right at the main intersection on Carcajou flats. Fol low the gravel road past a home and take the main dirt trail through the forest (note: there are several trails of varying condition which reach the shoreline). Although the road reaches the river, boat launching is not recommended due to the rough nature of the road. 215 9) Hawk Hills (km 663) Local residents indicate that boat launching is possible approximately 10 km north o f Notikewin Provincial Park on the west side of the river. The road down the valley side is very steep and should only be used in dry weather. The author did not use the site since it is relatively close to the access point at Notikewin Provincial Park. 10) Notikewin Provincial Park (km 673) Notikewin Provincial Park is located northeast of Manning and is reached by turning onto Highway 692 from Highway 35 near Hawk Hi l ls. Road-side signs provide directions down the valley on a good gravel road. The condition of the unofficial boat launch depends on the stage of the river and can be extremely difficult during low stages. Under the poorest conditions, boats must be carried 25 m over thick mud that lines the banks. A lso, extensive bars located along the left bank must be avoided during low flows. A 20-site campground is located in the forest downstream of the boat launch and is supplied with firewood, outhouses, and a water pump. The cost of camping is $7-9 per night. 11) Sunny Valley (km 751)* Sunny Val ley is reached by turning east from Highway 35 onto an all weather road near North Star. The well graded gravel road winds down the valley-side onto a wide gravel-terrace covered by a forest o f jack pine. The road to the river passes gravel pits and farms before reaching the boat launch. The launch is generally in good condition with a concrete base. However, it may be covered with mud after floods. There are no formal campsites in the Sunny Val ley area. One informal campsite with a fire-pit is located adjacent to a gravel-pit approximately 2 km from the boat launch. Permission for extended camping should be obtained from the gravel-pit owner. Nearest services are located in Manning. 12) Whitemud River (km 782) The Peace River may be accessed near the mouth of the Whitemud River by driving down a dry weather road west of the gravel surfaced Highway 743, southeast o f Deadwood. The dry weather road crosses private land but the landowner has permitted recreational use of the site in the past. Boat launching is poor on the silty alluvial fan of the Whitemud River and should be used as a last resort. A small area for primitive camping is available. 13) Town of Peace River (km 840)* There are three boat launches located in the Town of Peace River. The Peace River River Rats boat dock and launch is located in West Peace River on the left bank upstream of the bridge crossing. It is well marked and is reasonably good during moderate to high flows. Launching is on a cobble-gravel bar and it is very shallow offshore. The second boat launch is located in downtown Peace River almost directly across from the River Rats boat launch. It is a concrete launch and is generally suitable under a variety of flows. The third launch is located north of the bridge in the Springfield area. The launch is over concrete blocks. H igh velocities during high flows can make launching at this site difficult. Camping in the Town of Peace River is limited to the L ion 's Club Park south of the bridge in West Peace River. This 115-site campground has showers and costs $12 per night. A s the only campground in town, it may be very crowded at times. 14) Strong Creek Park (km 851) This is a site for picnics and camping and may be used as emergency access. It has a poor launch. 15) Shaftesbury Ferry / Tangent Park (km 866)* Boat launching is possible on the north side of the river, adjacent to the ferry ramp. The gentle slope and gravel base provide a reliable boat launch. Camping is available at Tangent Park on the south side of the river. 16) Kieyho Park (km 896) There is a good boat launch and several camping sites on south side of river accessed from Eaglesham. 17) Elk Island Park (km 901) Campsites and a boat launch at E lk Island Park are accessed by a steep dry weather road. It may be difficult to launch at low flows due to shoals offshore. 18) Dunvegan (km 943)* The Dunvegan boat launch is located immediately west of the Dunvegan bridge on the south side of the river and is accessed from a gravel road 100 m south of the bridge. The concrete ramp is good, but may be covered with silt. 30 camp sites are available on the north side of the river at the Dunvegan Provincial Park. This is a very popular campground and may be especially crowded during holidays. 19) Pratt's Landing Campground (km 970) The Pratt's Landing campground and boat launch is located on a narrow floodplain in the upper Peace River valley. It is reached by travelling on secondary road #682 west of Fairview, and following the signposts. A steep dirt road winds down to the valley bottom and should not be traveled during wet conditions especially i f towing a boat. Four wheel drive is strongly recommended. A t the river, two poor launches are available. One is over a cobble bar and the other down a steep bank. Approximately 12 campsites are available along with a waterwell, shelter, firepits and outhouses. 20) Carter's Camp Campground (km 996) Carter's Camp Campground has a gravel boat launch and 3 formal sites for camping. The boat launch is generally good but may be muddy at times. The access road is reached by travelling 217 either west or north from Fairview on secondary roads 682 and 729 respectively, or north and west from Fairview on secondary roads 732 and 685 respectively. The latter route passes through Hines Creek. The gravel road down the valley-side is moderately steep and may be difficult to use during wet conditions. 21) Many Islands Campground (km 1011)* The Many Islands Campground and boat launch is located several kilometres south of Highway 964 at the end of an all weather access road. The boat launch has a firm gravel surface and is good under most conditions. 24 campsites are available. 22) Campbell's Lease (km 1030) This is an emergency access point on north side of river, used by locals for boat launching and picnics. Access is by a dry weather road across private land. Boat launching is difficult on the alluvial fan. 23) Cotillion Campground (km 1053) The Cot i l l ion Campground and boat launch is located on the right bank (south side) of Peace River and is reached by turning north from Highway 45 onto secondary highway 719 near Bonanza. The boat launch is on the fan of Sneddon Creek, and may be difficult during wet weather. Several formal campsites with showers and hot water are available. 24) Alces River (Clayhurst Bridge) (km 1079)* Boat launching is possible on cobbles on the left bank. Camping is possible downstream of the bridge on the right bank. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089091/manifest

Comment

Related Items