by
Willem Gerald Zantvoort B.Sc., University of Victoria, 2000 A Thesis Submitted in Partial Fulfillment of the
Requirements for the Degree of Master of Science
in the Department of Geography
© Willem Gerald Zantvoort, 2008 University of Victoria
All rights reserved. This thesis may not be reproduced in whole or in part, by photocopy or other means, without the permission of the author.
Beach Profile Morphological Changes: Northeast Graham Island, Queen Charlotte Islands, British Columbia, Canada
by
Willem Gerald Zantvoort B.Sc., University of Victoria, 2000
Supervisory Committee Dr. Ian J. Walker, Supervisor (Department of Geography)
Dr. Dan Smith, Departmental Member (Department of Geography)
Dr. J. Vaughn Barrie, Outside Member (School of Earth and Ocean Sciences)
Supervisory Committee Dr. Ian J. Walker, Supervisor (Department of Geography)
Dr. Dan Smith, Departmental Member (Department of Geography)
Dr. J. Vaughn Barrie, Outside Member (School of Earth and Ocean Sciences) Abstract
The northeast coast of Graham Island is macrotidal, composed of unconsolidated sediments, and subject to extreme wind and wave conditions. Cape Fife coast is erosive, composed of sand to gravel, and is medium to low sloped with intertidal to subtidal bars. The north coast is mainly depositional, very low to steeply sloped, and composed of fine aeolian sands and cobbles. Rose Spit, trending north-northeast separates the two beaches. Cross-shore profiles documented seasonal morphologies, where active summer swash bar development is preceded by concave profile. This study identified that bars on the east coast are linked to erosive hotspots. There was a relationship between wavelength and amplitude of the bar and the erosive nature of the foreshore. It is proposed that bars protect against incident waves. Decadal and seasonal evolution of a portion of the northeast coast was compiled with the combination of aerial photography, bathymetric data and geomorphological mapping.
Table of Contents
Supervisory Committee ... ii
Abstract... iii
Table of Contents... iv
List of Tables ... vii
List of Figures... viii
Acknowledgements... 1
Chapter 1.0: Introduction... 2
1.1 Thesis Outline ... 2
1.2 Research Context ... 2
1.3 Purpose and objectives... 5
1.4 Literature review... 6
1.4.1 Background on morphological studies... 6
1.4.1.1 Surf scaling parameter (ε) and beach slope (tan β)... 8
1.4.2 Macrotidal beach morphologies... 12
1.4.3 Beach response to seasonal changes in wave and tide regime... 16
1.5 Study area... 19
1.5.1 Geomorphology ... 19
1.5.2 Previous research ... 22
1.5.3 Marine geology and oceanography... 22
1.5.4 Sea level history... 24
Chapter 2.0: Research methods... 26
2.2 Surveys... 27
2.2.1 Cross-shore profile site selection and rationale ... 27
2.2.2 Establishing a survey network ... 31
2.2.3 Topographic surveying methods... 36
2.2.4 Bathymetric surveys... 39
2.3 Sediment sampling... 39
2.4 Aerial photograph analyses... 41
2.5 Wave data... 44
Chapter 3.0: Results... 47
3.1 Introduction... 47
3.2 Cross-shore profiles ... 47
3.2.1 Elephant Cage Profile ... 47
3.2.2 White Creek Profile ... 51
3.2.3 Agate Beach Profile ... 54
3.2.4 North Beach 1 Profile ... 59
3.2.5 North Beach 2 Profile ... 67
3.2.6 North Beach 3 Profile ... 70
3.2.7 Rose Spit West Profile... 73
3.2.8 Rose Spit East Profile ... 77
3.2.9 Naikoon Profile... 81
3.2.10 Cape Fife Profile... 87
4.2 Intertidal and subtidal bar morphology... 101
4.3 Onshore aeolian sand transport and coastline stability... 106
4.4 Foredune erosion, rebuilding and progradation on North Beach... 110
Chapter 5.0: Conclusions... 113
5.1 Summary and conclusions ... 113
5.2 Recommendations for further research... 115
References... 117
List of Tables
Table 1. The 6 beach state classification scheme of Wright and Short (1984), from
dissipative to reflective beach with 4 intermediate states.……….………....10 Table 2. Macrotidal beach classification described by Short (1991), including 3 beach types that are described based on slope and morphological
features.………...………...13 Table 3. The conceptual beach model of Masselink and Short (1993). This model
attempts to integrate wave, tide and sediment properties to characterize beach
morphologies.………..………...15 Table 4. Profile site names and geographical benchmark location and elevation. ……...28 Table 5. Occupation dates of profiles on the north and east coast of Graham Island. Bathymetric survey dates are presented in bold italic………...29 Table 6. Survey pin location at Cape Fife with occupation dates and amount of coastal retreat measured. As of 15 February 2005 there were 4 pins remaining on the bluff with roughly 4 m between each pin………...38 Table 7. Sediment sample dates at each profile location. The letters h, l indicate high and low tide sample respectively………..40 Table 8. Year, series number and scale of aerial photographs used in study………42
List of Figures
Figure 1. Study area (black box on main figure) located off the northwest coast of mainland British Columbia, Canada. The detailed area of coastal monitoring is
highlighted in bold on the inset map. ………...…...4 Figure 2. LandSat image of Naikoon Peninsula indicating the approximate location and bearing (red arrows) of the ten cross-shore profiles measured during this study. Accurate profile bearings and co-ordinates for this study are provided in Table 4.………...20 Figure 3. Schematic showing relationships between ellipsoidal elevation as determined from GPS occupation at Tow Hill, Masset Mean Water Level as determined at Masset government dock and chart datum-mean water level separations……….32 Figure 4. Time series plot showing pressure sensor data obtained from Masset Inlet during 24 February through 27 May 2004……….34 Figure 5. GPS receiver positioned over brass survey marker in bedrock at the seaward edge base of Tow Hill, Naikoon Park, Queen Charlotte Islands. This occupation was used to determine ellipsoidal elevation for the study area……….35 Figure 6. Mean daily significant wave height data from Dixon Entrance wave buoy C46145. Dates of data presented are May 2003 through March 2005………..45 Figure 7. Mean daily significant wave height data from Hecate Strait wave buoy C46183. Dates of data presented are May 2003 through March 2005……….46 Figure 8. View to west-northwest of gently sloping, convex, upper beach face with broad low tide terrace at Elephant Cage Profile. Note exposed boulder lag at low tide with person for scale (circled).………...49
Figure 9. Figure showing topographic beach profile of Elephant Cage Profile from 1993 through 2005………..50 Figure 10. View of White Creek Profile location looking west-northwest at upper beach and vegetated log debris foredune complex. Note abundant aeolian deposition occurs within the backshore region………...52 Figure 11. Topographic profile results at White Creek Profile, 1993-2005……..………53 Figure 12. View is looking west across Agate Beach Profile at the reflective cobble boulder upper beach with well developed cusps and berms. At the right of the photograph note the abrupt transition from steep upper beach to low sloped, planar, low tide terrace. ………...56 Figure 13. Results of two years of topographic profile monitoring at Agate Beach Profile. The majority of sediment flux appears to occur above and below mean sea level with the middle beach being a point of no change………..57 Figure 14. 1997 aerial photograph showing embayed Agate Beach, with headlands Yakan Point to the southwest and Toe Hill to the northeast. Agate Beach is starved of sediment due to the groyne affect of Yakan point. Littoral drift is from southwest to northeast. Aerial photographs, BC97035-103J001 and 103J002, BC97036-103J012.…………..…58 Figure 15. Topographic profile results at North Beach 1 Profile 2003 through 2005. Note the repeated scarping at the base of he foredune………..60 Figure 16. View looking east along shore across North Beach 1 Profile. The photograph was taken from the crest of the vegetated foredune during February, 2005……….61
Figure 17. View is looking west across North Beach 1 Profile towards Tow Hill in the background. The Hiellen River mouth outflows into McIntyre Bay on the east side of Tow Hill and defines the western extent of North Beach……….62 Figure 18. View is looking east along the base of a scarped foredune at North Beach 1 Profile location. Note the extensive log debris matrix that is exposed………..63 Figure 19. Figure showing sediment transport to the base of the foredune during May through July 2004 at North Beach 1 Profile. Note, not all cross-shore profiles are shown as figure becomes difficult to interpret………..65 Figure 20. Detailed topographic surveys documenting onshore migration of bar feature at North Beach 1, May through July 2004. Note, not all cross-shore profiles are shown as figure becomes difficult to interpret. Truncated is data is shown from 100-450 m across shore to focus on bar migration……….66 Figure 21. View looking onshore, along North Beach 2 Profile, from sandy low tide terrace toward coarser grained and steeper upper beach face………68 Figure 22. Detailed topographic profile surveys of North Beach 2 Profile during July 2003 through February 2005...………...69 Figure 23. Topographic profile changes at North Beach 3 Profile during 2003 through 2005. Note the well developed persistent slope break at transition from low tide terrace to moderately sloped upper beach………..71 Figure 24. View of abrupt transition from moderately sloped upper beach face at left to low slope low tide terrace at right, looking west towards Tow Hill………..72
Figure 25. Slope transition line from intermediate sloped upper beach to low slope lower beach at Rose Spit West Profile. View looking west-southwest and Tow Hill is in the background at the right of the photograph. ………..……….74 Figure 26. Topographic profiles documenting changes observed at Rose Spit West Profile during 2004 through 2005……….….…75 Figure 27. Bathymetric profiles along the north coast of Graham Island from Elephant Cage to Rose Spit West. Note the steeping of the profiles from west to east. These
profiles were measured during July 2004.……….…76 Figure 28. Topographic profiles displaying onshore migration of intertidal bar and
summer and winter profile morphologies at Rose Spit East Profile. Note, not all cross-shore profiles are shown as figure becomes difficult to interpret.……….78 Figure 29. Figure showing swash bar development on the east coast of Rose Spit. Note well developed stoss and lee slope on current ripples in foreground and steep leading edge of gravel swash bar. View is looking north towards tip of Rose Spit………...79 Figure 30. Bar morphologies on the east coast from Rose Spit East to Cape Fife………80 Figure 31. Shown is the log debris zone at Naikoon Profile. Note the vegetated swale followed by a second log debris zone to the left (west) of photo. The view is looking north, toward tip of Rose Spit. Hecate Strait is to the right.………..82 Figure 32. Topographic profiles measured at Naikoon Profile exhibiting convex summer berm profile and concave winter profile. Note, not all cross-shore profiles are shown as figure becomes difficult to interpret………..83
Figure 33. Cross-shore profiles displaying vertical accretion at Naikoon Profile. Note, not all cross-shore profiles are shown as figure becomes difficult to interpret ………...…...85 Figure 34. Bar 1 and Bar 2 observed at Naikoon Profile. Bar 1 is subaerially exposed at low tide and the swale is intermittently traversable at spring low tide conditions. Ideally, all bars are surveyed with small vessel mounted bathymetry gear at high tide to overlap land based surveys. View is looking east- southeast to Hecate Strait………...86 Figure 35. View is looking north along base of Cape Fife bluff. Bluffs have retreated more than 7 m yr-1 (2003 through 2005).………..……….88 Figure 36. View of Bar 1 and Bar 2 exposed at Cape Fife Profile. Bar 1 is accessible by surveyor with rod, Bar 2 is surveyed with bathymetry capable vessel. View from top of bluff looking east-southeast to Hecate Strait. ………89 Figure 37. Cross-shore profiles showing measured bluff retreat at Cape Fife Profile during July 2003 through February 2005. More than 7 m yr-1 retreat was measured at this location. Note, not all cross-shore profiles are shown as figure becomes difficult to interpret.………..………...……90 Figure 38. Coastlines (approximate base of scarp) measured from 1937, 1966, and 1980 aerial photographs on base image of 1997 ortho photograph. Coastal retreat (1937-1997) at location of survey pins at south end of Cape Fife is roughly 200 m. Summary of retreat rates listed in Appendix 4………..91 Figure 39. Schematic shows the survey pin network at the south end of Cape Fife with the bluff edge location as of February 2005. Rebar survey pin location at bluff edge is noted on Figure 38.………....94
Figure 40. Topographic surveys documenting onshore bar migration of Bar 1 during May 2004 through February 2005 at Cape Fife Profile. Note, the asymmetrical summer bar and trough are symmetrical in the February profile………95 Figure 41. During May through July 2004 sediment was consistently removed from the upper beach save for 8 July through 31 July when minor accumulation is recorded……96 Figure 42. Schematic showing conceptual bar migration along a section of East
Beach……….………...108 Figure 43. Schematic showing simplified current (yellow arrows), wind (blue arrows) and possible sediment transport (yellow arrows) directions in the area of Naikoon Peninsula. Measured rates of erosion (red arrows) and progradation (green arrows) are also
indicated. Figure is intended to roughly display the cycling of sediments in a regional context. ……….……….………..112
I offer sincere thanks to the members of my supervisory committee, Ian Walker for his diligent editorial reviews, Vaughn Barrie and Dan Smith whose ongoing support, encouragement and patience were instrumental in the completion of this research. Valuable conversations on coastal processes were had with Chris Houser and my good friend Gavin Manson at GSC-Atlantic, who provided valuable reviews and a swift kick of encouragement at the 11th hour. Thanks to Brian Bornhold for taking time to externally examine this thesis.
Field work was assisted by Jeff Anderson, Teresa Conner, Brendan Mather, Kim Pearce, Matt Pope, and Kevin Tattrie. Last minute drafting assistance was provided by Ole Heggen from UVic Geography Department. My girlfriend Chandra Venables provided encouragement and displayed an amazing level of patience. The list of friends, mentors and colleagues at UVic, and Natural Resources Canada is too long to list and all the support was greatly appreciated. This research was made possible through funding from NRCan’s CCIAD. In kind support was provided by the Geological Survey of Canada, UVic, NSERC, CFI and BC Parks.
This thesis is dedicated to my family who has always supported my dreams, and my Uncle Bill, who we miss, and who would have greatly enjoyed seeing this day.
Chapter 1.0: Introduction 1.1 Thesis Outline
This thesis consists of five chapters. Chapter 1 states the research purpose and objectives of the thesis, and provides a thorough, wide-ranging literature review on the morphodynamics of macrotidal beaches. Chapter 1 also presents the geology, climate and oceanographic background for the study area. Chapter 2 describes the methods employed, including survey locations and dates, types of survey equipment, aerial photographs used and GIS analyses and techniques. Chapter 3 presents results on contemporary, event-based beach morphodynamics observed during the summer of 2004 that occurred in response to varying tidal cycles and storms. Chapter 3 also examines interannual to decadal-scale coastal changes documented in aerial photography and resurveyed topographic profiles. Chapter 4 presents a discussion of the research including key findings. Chapter 5 presents conclusions and includes directions for future research for the study area.
1.2 Research Context
The northeast coast of Graham Island, Haida Gwaii (Queen Charlotte Islands), British Columbia, is a highly dynamic geomorphic environment. The study region in this thesis extends north from Cape Fife, around Rose Spit, and west to Masset (Figure 1). The coast of the region is comprised of low-lying (5-20 m above sea level)
unconsolidated sediments, is subject to extreme wind and wave conditions, and has a mean spring tide range greater than 5 m.
coast, with monthly means in Hecate Strait and Dixon Entrance exceeding 2.8 m (Walker and Barrie, 2006; Walker et al., 2007). Given these conditions, the Geological Survey of Canada (GSC) has identified this region as highly sensitive to sea-level rise (Shaw et al., 1998).
During the late Quaternary, this region experienced sea level low stands indicated by submerged wave cut terraces in Hecate Strait 150 m below present mean sea level and sea level high stands approximately 16 - 18 m above present mean sea level (Barrie and Conway, 1999; Fedje and Josenhans, 2000; Barrie and Conway, 2002; Lacourse et al., 2003; Walker and Barrie, 2006; Wolfe et al., 2008). Potential impacts of climate change-induced sea-level rise include: increased coastal erosion; more extensive coastal
inundation; higher storm surge flooding; increased flood risk and potential loss of property (Walker and Barrie, 2006; Walker et al., 2007; Abeysirigunawardena and Walker, 2008).
The interaction of tides, winds, waves, sediment characteristics and pre-existing geology, controls coastal geomorphology, which is often described by the dominant force. For example, a coast may be tide-dominated with a macrotidal range (greater than 5 m) and little or no wave action, or at the other extreme in a microtidal environment (less than 2 m) coasts tend to be wave dominated, with little tidal influence.
Figure 1. Study area (black box on main figure) located off the northwest coast of mainland British Columbia, Canada. The detailed area of coastal monitoring is highlighted in bold on the inset map.
These are the two extreme ends of a spectrum and most coastlines around the world are some combination of wave- and tide-dominated (Anthony and Orford, 2002).
Most knowledge of sedimentary beach processes and morphology has been derived from studies conducted on temporally and spatially limited scales, largely in microtidal environments (i.e., tide range less than 2 m) (e.g., Sonu, 1973; Wright and Short, 1984; Lippman and Holman, 1990). However, ongoing research in macrotidal environments is adding to the knowledge base in coastal studies (e.g., Short, 1991; Masselink and Short, 1993; Amos et al., 1995; Kroon and Masselink, 2002; Masselink et al., 2006; van Houwelingen et al., 2006; Masselink et al., 2008). These studies have begun to demonstrate the importance and the effects of a macrotidal range on coastal morphodynamics.
1.3 Purpose and objectives
The purpose of this research was to document and characterize contemporary morphodynamics of a high-energy, unconsolidated, macrotidal coast on NE Graham Island, Haida Gwaii (Queen Charlotte Islands), British Columbia. This was done by comparative study of two beach types: (1) an intermediate multi-barred beach with SE exposure (East Beach) and (2) an ultra-dissipative beach with NW exposure (North Beach) (Figure 1).
Seven sites on the north coast and three sites on the east coast were studied using repeat topographic surveying, aerial photograph analysis using a Geographic Information System (GIS), and grain size analysis. The research objectives were:
1. To quantify rates of coastline change (i.e., erosion and/or progradation) via aerial photograph analysis and repeated topographic cross-shore profile surveys over
multiple years (1993 through 2005) on both seasonal (winter and summer) and weekly scales (summer).
2. To characterize and describe the coastal landforms at each profile location using aerial photography, grain size analysis, and repeat cross-shore profiling.
1.4 Literature review
1.4.1 Background on morphological studies
Beaches are often described based on cross-shore profile, sediment
characteristics, and bedform features such as bars, and ridge and runnel morphologies (e.g., Sonu, 1973; Hale and McCann, 1982; Wright and Short, 1984; Bruun, 1988; Dean, 1991; Dail et al., 2000; Dawson et al., 2002; Bernabau et al., 2003; Shand, 2003). The equilibrium beach profile (EBP) (Bruun, 1988; Dean, 1991) describes beach form and sediment motion in a two dimensional, onshore or offshore, direction and neglects the longshore component of sediment transport.
Equilibrium beach profiles, as described by Dean (1991) tend to be concave, and smaller and larger sand diameters are associated with gentler and steeper slopes
respectively. The beach face is approximately planar, and steep waves result in milder slopes and a tendency for bar formation (Dean, 1991). One implication of the EBP approach is the exclusion of sediment transported via longshore currents that play an important role in coastal morphology. In addition, the equilibrium profile, as described by Bruun (1962) and Dean (1991) did not initially incorporate the relationship between profile morphology, wave characteristics, sediment attributes, tidal extremes and onshore
predict the two dimensional response of the cross-shore profile under rising sea level conditions.
Two general classifications have focused on relating form with tidal range. In the first classification, on a beach with a microtidal range any incident wave energy will be focused over approximately 1-3 m tide range. The second framework considers a beach in a macrotidal environment which receives the same incident wave energy, though it is spread over 3-15 m tidal range (Wright and Short, 1984; Short, 1991; Masselink and Short, 1993).
According to Short (1991) there are three beach types: (1) microtidal with a tide range < 2 m; (2) mesotidal with a range between 2 - 4 m; and (3) macrotidal with tidal range > 4 m. Short (1991) notes that beaches with a mean spring tide range greater than 3 m usually do not exhibit the rhythmic and barred topography associated with microtidal beaches, and thus modifies the classification so that the microtidal range is < 3 m and macrotidal is 3-15 m. This is a weighted classification that does not separate tidal variations in the 3-15 m tide range, therefore neglecting a possible response in beach form that may result from a mesotidal environment.
Most studies of macrotidal environments are conducted on beaches with tidal ranges of 3 - 6 m (e.g., Short, 1991; Masselink and Anthony, 2001; Bernabau et al., 2003; Moore et al., 2003; Reichmuth and Anthony, 2007). With the exception of Wright and Short’s (1984) study of Cable Beach, Australia (tide range > 9.5 m), and ongoing work in Hecate Strait, northeast Pacific, where extreme high tide range exceeds 7 m (Walker and Barrie, 2006), there has been less work done in this high tide range environment, thus neglecting the range from 7 to 15 m.
1.4.1.1 Surf scaling parameter (ε) and beach slope (tan β)
Beaches have been distinguished by the surf scaling parameter ε and the beach slope (tan β) (Guza and Inman, 1975):
ε = αbω2/ (g tan2β) (1)
where αb is the breaker amplitude, ω is the incident wave radian frequency (2 π/T; T= period), g is the acceleration of gravity and β is the beach gradient. Three breaker types are identified in coastal studies: spilling, plunging and surging (Guza and Inman, 1975). Spilling breakers occur on relatively low slope beaches. This type of breaker spills over the leading edge of the wave over a long distance. Plunging breakers tend to curl over and break with a single crash and occur on steeper beach faces. Surging breakers occur on steep sloped beach faces and have a peak but do not ever break or spill. Surging breakers and complete reflection are expected when ε < 1.0. Strong reflection continues to occur at ε < 2.5, and at ε > 2.5 waves begin to plunge, thus dissipating energy until ε > 20, at which point spilling occurs in coincidence with an ultra-dissipative beach state (Guza and Inman, 1975).
The differences between the intermediate beach states (types 2-5) depend on the breaker type, beach slope (tan β = 0.03-0.1), and presence of rhythmic beach
morphologies such as bars, troughs, cusps, and ridge and runnel topography (Guza and Inman, 1975). This classification scheme effectively describes beach morphology
face morphodynamics. Most beaches included in Wright and Short’s (1984) study had a microtidal range of between 0.9 and 1.6 m. Cable Beach was the only macrotidal
environment with a spring tide range of 9.5 m.
Wright and Short (1984) defined six beach states by the beach face slope (tan β) and the surf scaling parameter (ε) (Table 1). At one end of the beach state classification is the dissipative beach with a very gently sloping face (tan β 0.01- 0.02) and high ε values (> 30). Dissipative beaches from Wright and Short (1984) on Seven Mile and Goolwa beaches in Australia showed a number of common characteristics: 1) a lack of consistent longshore rhythmic forms; 2) both beaches were sinks for fine-grained sand, had very low slope, and wide, multi-barred surf zones; and 3) the dominant waves breaking on both beaches were spilling followed by dissipation across the surf zone.
Wright and Short (1984) studied Bracken Beach south of Sydney, Australia to better understand reflective beach dynamics. This fully reflective beach is characterized by a very steep (tan β = 0.1 - 0.2), often coarse sand beach face dominated entirely by surging to collapsing breakers. Unlike a highly dissipative beach, fully reflective beaches are dominated by incident wave energy while infragravity energy plays a minor role (Butt and Russell, 1999). On most reflective to semi-reflective beaches, rhythmic forms consist of well developed beach cusps composed of coarser grained sediments (sand to cobble) (e.g., Masselink and Pattiaratchi, 1998; Nolan et al., 1999; Sanchez et al., 2003).
Table 1. The 6 beach state classification scheme of Wright and Short (1984), from dissipative to reflective beach with 4 intermediate states.
Beach Type Wave type ε Tan β Features
Dissipative Spilling High 0.01 Flat/concave beach
trough Intermediate
longshore bar-trough
Plunging to surging Moderate to low 0.05- 0.2 cusps
steep reflective beach deep trough and bar Intermediate
Rhythmic bar-beach Plunging to surging Moderate 0.01-0.2 mega cusps mega ripples bar and trough Intermediate
transverse bar-rip Plunging to surging Moderate 0.05 low berm bar and runnel Intermediate
ridge-runnel or low tide terrace
Plunging Moderate 0.01 low berm
terrace runnel flat bar
Reflective Surging Low 0.01-0.2 steep beach face
wide berm cusps
Between the highly dissipative and fully reflective beaches of the Wright and Short (1984) classification scheme exist four intermediate assemblages. All of these assemblages possess some aspect of a dissipative and a reflective beach with complex morphologies. Progressing from dissipative to reflective, there is a gradual development of rhythmic forms. The first intermediate beach, the longshore bar-trough state, has a slightly steeper beach face (tan β = 0.05 - 0.2) and a higher longshore bar. In this profile the breakers dissipate somewhat on the bar to reform in the 2 - 3 m deep trough thus providing high wave energy to a locally reflective beach face. The notable change in morphology from this type to the second intermediate type, the rhythmic bar and beach state, are the rhythmic forms of the crescentric bar and the subaerial beach, which have longshore feature spacing of between 100 and 300 m. Type (1) and (2) of the Wright and Short (1984) intermediate assemblage can both develop from a preceding dissipative accretionary profile.
The transverse-bar and rip-state is typically accretionary and develops when existing crescentric bars become welded to the beach face as a result of the interaction of two edge wave modes with similar frequencies (Holman and Bowen, 1982; Moore et al., 2003). This sequence of shoreline morphology leads to alternating, highly dissipative transverse bars and reflective rip-dominated embayments (Gelfenbaum and Brooks, 2003).
The final intermediate beach state is the ridge and runnel, or low tide, terrace state. This beach has a reflective high tide state, dissipative low tide state and commonly occurs in sandy, macrotidal environments exposed to fetch-limited waves (Reichmuth and Anthony, 2002). Ridge and runnel topography typically has 1 to 6 ridge complexes
over a 300 m cross-shore distance. In addition, they determined that a more subdued ridge and runnel topography results from greater protection from prevailing waves, and as wave energy increases so too do both ridge and runnel topography and mobility. An exception to this behavior may occur in response to a combination of high wave energy and large spring tides that appear to have a flattening effect on the ridges.
Beach states are also described by Wright and Short (1984) using the dimensionless parameter Ω:
Ω = Hb/wsT (2) where Hb is the breaker height (m), ws is the sediment fall velocity (m s-1) and T is the wave period (s). When Ω < 1 a steep, cusped, reflective beach state results. The four intermediate beach states have Ω values ranging from 1 - 6 and typically exhibit bar and rip morphology. At Ω > 6 a dissipative beach state occurs, possibly with subdued bar morphology.
Wright and Short (1984) offer little in terms of linking beach morphodynamics with macrotidal range. Although similarities in form exist between the microtidal and the macrotidal environment, the two are not analogous. It is necessary to study meso-macro tidal beaches to further our understanding of how a beach face responds to a greater vertical distribution of energy over larger spatial and temporal scales.
1.4.2 Macrotidal beach morphologies
Short (1991) reviews a number of studies on macrotidal beach environments and provides an overview of the related morphodynamics identifying three types of
Table 2. Macrotidal beach classification described by Short (1991), including 3 beach types that are described based on slope and morphological features.
Beach Type Slope (º) Features
Higher Wave, Planar, Uniform Slope 1.0-3.0 - planar surface - no ripples - no bedforms/bars
- steep, cuspate high tide beach Moderate Wave, Multi Bar 0.5-0.6 - flat, uniform intertidal
- 2-5 multiple bars
Low Wave Beach and Tidal Flat ~0.1 - reflective, coarse grained beach face - low slope, rippled tidal flat
Short (1991) also identifies problems with making accurate comparisons of these beach types including the variable tide ranges (3-15 m), grain sizes and levels of sorting, as well as high to low sea states. In addition, the majority of the morphodynamic analyses performed before 1991 provided limited linkage between tide range, wave height, grain size, and beach slope, and it is from this that Masselink and Short (1993) developed the relative tide range (RTR) relationship. The RTR relationship builds on the beach groups previously described by Short (1991) and Wright and Short (1984).
With an increased understanding of wave effects on beach morphology, Masselink and Short (1993) adapted the Ω value from Wright and Short (1984) and introduced the relationship RTR=TR/Hb, where RTR is the relative tide range and is a function of breaker height (Hb) and tide range (TR). From this relationship Masselink and Short (1993) developed a conceptual beach model that describes beach state as a function of RTR and Ω values in order to account for how a macrotidal environment affects beach state. This conceptual model has the ultra-dissipative and fully reflective end members that bound Wright and Shorts (1984) classification and has six intermediaries (1) low tide terrace + rip (2) low tide terrace (3) barred (4) low tide bar/rip (5) barred dissipative and (6) non-barred dissipative (Table 3).
Table 3. The conceptual beach model of Masselink and Short (1993). This model attempts to integrate wave, tide and sediment properties to characterize beach morphologies. Fall Velocity (Ω) 0-2 2-5 ≥ 5 0-3 reflective cusps step intermediate barred cusps/steep beach face subdued or pronounced bar
barred dissipative single-multiple bars 3-7 low tide terrace/rip
cusps
reflective beach face low tide terrace with rip
low tide bar/rip steeper beach face swash bar
low tide transverse bar and rip
non-barred dissipative flat/featureless Relative Tide Range (R TR)
7-15 low tide terrace cusps
reflective high tide beach dissipative low tide terrace
ultra-dissipative flat/featureless cusps ultra-dissipative flat/featureless cusps
1.4.3 Beach response to seasonal changes in wave and tide regime
Aubrey (1979) examined five years of beach profile data from southern California to study a long-recognized difference between summer and winter beach morphology. Two distinct seasonal pivotal points (point separating eroding and accreting regions) of sediment movement were identified by Aubrey (1979) at approximately 2-3 and 6 m below mean sea level. In the summer months, June through October, the beach foreshore accretes and the nearshore erodes. During the winter months of roughly November
through May, the foreshore erodes and the offshore accretes. This seasonal profile change was attributed to variability in wave dynamics. Small amplitude, long period waves are dominant during summer months while higher energy, high frequency waves dominate in winter. It is this seasonal wave climate that controls the local redistribution of sediments (Aubrey et al., 1980).
Aubrey (1979) found that early winter storms tended to erode sediment from the beach backshore and foreshore as well as from depths of 6 - 10 m. This sediment was subsequently deposited in depths of 2 - 6 m. The results of this study did not identify a sand movement direction following the winter storms, but suggested that some sand may move shoreward under wave influence and some may be transported seaward under gravity influence. Specific stages of beach rebuilding were documented during summer months when winter storm waves gave way to more gentle swell waves (e.g., Dubois, 1988; Kroon and Masselink, 2002; Quartel et al., 2008). Dubois (1988) documented the rebuilding of a segment of the Delaware coast via lower energy swell waves between March and August 1982. It was proposed that the post-storm accretion stage is caused by
propagating bores and swash (Houser and Greenwood, 2007). No net erosion or accretion occurred during the 5-year period of measurements.
In a study of inner and outer bar systems on the coast of the Outer Banks, North Carolina, beach change was observed to be a result of wave obliquity not intensity (Sonu, 1973). Sonu (1973) described two separate mechanisms of beach change, one associated with outer bars, a second associated with inner bars. During a five month field program from January through May 1964, local waves were dominated by a series of migratory atmospheric depressions (Dolan et al., 1969). The depression brought with it growing southerly wind waves which shifted to northerlies as the depression moved away. These were finally replaced by shore normal swells. In this scenario, the northerly waves were higher than the southerlies and the later stage shore normal swells were lowest.
Changes in nearshore bathymetry showed a direct relationship to wave incident angle. The oblique northerly and southerly waves effected a greater change on the outer bar system than did the shore perpendicular swells. This was seen even when the oblique wave height was less than that of the swells. Also described by Sonu (1973) was the accretion of the subaqueous profile during the southerly waves and the erosive nature during northerlies. Beach change was attributed to variable wave direction, as the difference in wave height and gradient was negligible (Sonu, 1973). Material from the inner bar was eroded during storm events, this material was dispersed on the surf zone bed and subsequently the subaerial beach formed a concave surface (Sonu, 1973). As with the work of Wright and Short (1984), Sonu’s (1973) work pertains to a microtidal environment and does not take into account the affect of tide range on beach morphology (Masselink and Short, 1993).
After the storm stage, during the swell-dominant stage, currents move shoreward over the shoal and seaward through rip channels between the shore-parallel shoals. These shoals that become stranded on the beach are commonly referred to as swash bars. This topography is often referred to as ridge and runnel where the swash bar is the ridge and the trough behind that drains through rip channels is termed runnel (Owens and Frobel, 1977). The shoreward progression of swash bars decreases as the bar reaches the upper limit of swash extent. The runnels decrease in size as the amount of wave overtopping diminishes. Ultimately the swash bar welds alongshore forming a continuous berm and a convex beach profile (e.g., Orford and Wright, 1978; Kroon and Masselink, 2002;
Masselink et al., 2006). This is a state of full beach accretion. This characteristic summer morphology with a berm in the upper intertidal and a ridge and runnel system in the lower intertidal has also been documented at the Truc-Vert Beach (French Atlantic Coast) (Apoluceno et al., 2002). Apoluceno et al. (2002) identified migration rates of the berm and ridge and runnel system to the upper and middle side of the intertidal zone at rates of 0.1 m day-1 and 0.8 m day-1 respectively.
These and other studies, conducted in the nearshore environment have identified many specific beach characteristics and their relationship to wave and tidal regimes. This thesis will attempt to address certain aspects of beach morphology in a macro-tidal multi-barred system.
1.5 Study area
1.5.1 Geomorphology
Graham Island is located approximately 100 km off the NW coast of mainland British Columbia and is the largest island in the Queen Charlotte Islands archipelago (Figure 1). The research sites consist of seven cross-shore profiles on the north coast and three profile sites on the east coast (Figure 2). All profiles include corresponding
bathymetric profile data obtained in 2004. The Naikoon Peninsula of NE Graham Island is comprised of glaciofluvial Quaternary sediments and is subject to rapid rates of coastal retreat, from 1-3 m yr-1 to as much as 12 m on East Beach, in response to individual storm events (Barrie and Conway, 2002). In addition, the east coast experiences localized growth of foredune complexes as sediment is transported onshore from the exposed bars at low tide (Pearce, 2005; Walker and Barrie, 2006). In contrast, North Beach is
prograding at rates of 0.3-0.7 m yr-1 (Harper, 1980; Clague et al., 1982; Walker and Barrie, 2006; Walker et al., 2007; Wolfe et al., 2008).
The local wind regime is bimodal as the region experiences predominantly NW winds during the summer months and SE winds during the winter months (Pearce, 2005). Given the north and west aspect of the north coast, it receives predominantly onshore winds in the summer and offshore winds in the winter. In contrast, the eastern coast is subject to predominantly offshore winds in the summer and strong onshore winds during the winter months. Competent (i.e., wind speeds above ~6 m s-1) NW winds during the summer months help maintain the foredune complex by transporting fine sands from the subaerial beach.
Figure 2. LandSat image of Naikoon Peninsula indicating the approximate location and bearing (red arrows) of the ten cross-shore profiles measured during this study. Accurate profile bearings and co-ordinates for this study are provided in Table 4.
The beaches on the north coast generally have low slopes (approximately 0.50). These beaches are comprised of fine sand, locally occurring boulder fields (intertidal at Elephant Cage Profile), and are backed by active foredunes. Locally, where the upper beach is steep and starved of fine sediment, and the high tide beach is comprised of cobbles and multiple cuspate berms, the foredune is less developed. This scenario occurs at Agate Beach Profile, the eastern extent of North Beach and the west side of Rose Spit.
Beaches on the east coast are typically more continuous that those on the north coast, they have a steeper slope from the foreshore to the nearshore, are comprised of predominantly sands and gravels, and are fronted by multiple intertidal to sub-tidal nearshore bars. Along the east coast of Graham Island from Tlell to Rose Spit is a
sequence of oblique nearshore bars that are shore attached at the southern end. These bars trend offshore in a NE direction. Three high-amplitude bars occur in the nearshore zone and at low tide from Cape Fife to the tip of Rose Spit; two of these bars are intertidally exposed. At low tide, the junction of the bar with the shoreline provides a large surface of sediment available to be delivered to the parabolic dunes via onshore winds (Walker and Barrie, 2006).
The coast surrounding the Naikoon Peninsula hosts multiple longshore bars and active parabolic dunes on the rapidly retreating east coast, contrasted by an extensive series of progradational beach ridges on the north coast. These two dynamic geomorphic environments meet at Rose Spit, a 15 km long, accretionary feature oriented N/NE separating Dixon Entrance from Hecate Strait.
On NE Graham Island, annual average wind speeds exceeding 8.0 m s-1 combine with a fetch length greater than 100 km to produce significant wave heights (Hs) from 1.8
– 2.1 m. It is the combination of high wind speeds, high wave energy and available sediments that produce the geomorphically dynamic shoreline of NE Graham Island (Barrie and Conway, 1996; Walker and Barrie, 2006; Walker et al., 2007).
1.5.2 Previous research
A study of coastal processes was conducted on the north coast of Graham Island by the Geological Survey of Canada (GSC) August 1978-July 1979 (Harper, 1980). Additional research was initiated in 1990 by the GSC to document coastal change on NE Graham Island (Conway and Barrie, 1994; Barrie and Conway, 1996; Barrie and
Conway, 2002). This program was then continued as part of a larger, interdisciplinary study on coastal vulnerability to climate change and sea-level rise on Northeast Graham Island (Walker and Barrie, 2006; Walker et al., 2007). This study was funded by Natural Resources Canada’s Climate Change Impacts and Adaptations Directorate (CCIAD) with contributions from the Natural Sciences and Engineering Research Council (NSERC), the Canada Foundation for Innovation (CFI), and the GSC. In-kind support was provided by BC Parks, GSC, Council of the Haida Nation (CHN), CHN-Forest Guardians, Old Massett Village Council, and the University of Victoria.
1.5.3 Marine geology and oceanography
The study region is bounded by Hecate Strait to the East and by McIntyre Bay and Dixon Entrance to the North (Figure 1). The north coast of Graham Island, from approximately Masset Inlet to Rose Spit, is a region of active sediment accumulation, as evidenced by a prograding beach-dune complex that receives sediment transported northward by longshore drift that subsequently passed through Rose Spit to be
the origin of the sediments brought onshore on the north coast is that deepwater sediments are brought onshore from the north (Harper, 1980). Wolfe et al. (2008) describe a continual progradation of the North Beach shoreline in response to ongoing and gradual rates of sea level regression coupled with a strong wind regime, resulting in a well developed series of foredunes. The implications of this, the authors describe, is that wind action is as important as wave swash in delivering sediment to the north coast.
To the east, the waters of Dogfish Bank are shallow (approx. 20 m deep) and the deepest portion of Hecate Strait is a NE-southwest trending trough that is 300 m deep at the south end, shoaling to less than 80 m at the north. This body of water has a fetch greater than 100 km from the southeast. Four prominent submerged terraces have been mapped in Hecate Strait displaying characteristic wave cut features at depths of 20-25, 30-35, 40-45, and approximately 110 m (Bornhold and Barrie, 1991). These terraces are interpreted as relict features from previous sea level low stands.
The seafloor characteristics of Dixon Entrance are less well known than those of Hecate Strait. Dixon Entrance to the west has a wide 300 m deep mouth that gradually narrows and shoals to the east (Bornhold and Barrie, 1991). McIntyre Bay is the southern extent of Dixon Entrance and is a shallow, less than 40 m deep embayment. Quaternary geology of the study area is comprised of subglacial to proglacial deposits that resulted from the coalesce of local piedmont glaciers and the Cordilleran ice sheet (Barrie and Conway, 2002).
Tides in the area range from 5-7 m, and are semidiurnal mixed, resulting in two high and two low tides each day, each differing in amplitude (Thomson, 1981). There are two branches to the tidal wave, one of which propagates eastward through Dixon
Entrance and the other northward through Hecate Strait, meeting at Rose Spit
approximately 1 hr into flood tide (Amos et al., 1995). Annual significant wave height is 1.8 m with significant wave heights > 3.5 m occurring 20-30% of the time (Thomson, 1981; Amos et al., 1995).
General current directions in the study area are from west to east in Dixon Entrance and McIntyre Bay and from south to north in Hecate Strait (Crawford and Thomson, 1991). These two transport paths converge at Rose Bank, located E-NE of Rose Spit. A tidal current study, conducted by the Canadian Hydrographic Service from 1990 through 1995 using drifter buoys deployed in Dixon Entrance and Hecate Strait (Crawford et al., 1998), identified a well-defined counter-clockwise gyre (the Rose Spit Eddy) in east Dixon Entrance that occurs within the general circulation pattern. However, during intense SE storms coupled with an ebb tide cycle this west to east transport can be reversed in the eastern portion of Dixon entrance inducing an east to west flow path in SE McIntyre Bay (Amos et al., 1995).
1.5.4 Sea level history
After the late Wisconsinan glaciation (70 ka BP – 15 ka BP) the Queen Charlotte Islands experienced a regional sea level regression. Around 14600-12400 14C yr BP isostatic rebound and the development of a glacioisostatic forebulge caused a sea level regression in southern Hecate Strait to as much as -150 m (Josenhans et al., 1995) and -100 m in northern Hecate Strait (Barrie and Conway, 2002). Eustatic sea-level rise in concert with a subsiding glacioisostatic forebulge, led to a rapid sea level transgression that resulted in a sea level highstand of approximately 16 to18 m above modern sea level
2008). Evidence of this regional transgression is stored in drowned shorelines, river channels, estuaries, deltas, spit platforms, barrier islands and wave-cut terraces.
Following this highstand, relative sea level has regressed since approximately 3800 yr at a rate of approximately 2.0 mm yr-1 (Fedje and Josenhans, 2000). Evidence of this regression is stored in coastal features including; relict beach ridges, prograding spit and foredune systems, and parabolic dune systems. Recent work however, suggests a reversal in mean sea level trends over the latter half of the twentieth century (+1.4±0.6 mm yr-1) (Abeysirigunawardena and Walker, 2008). Indications of how the modern coastline might respond to a rise in sea level can be inferred from high water storm events. For example, Barrie and Conway (2002) document 12 m of coastal retreat as the result of a 40 cm rise in sea level during the 1997-1998 ENSO event. They propose that the impact of ENSO events will provide insight into probable coastal response to sea-level rise.
Chapter 2.0: Research methods 2.1 Introduction
This chapter describes research methods that were employed to document coastal morphodynamics on NE Graham Island, Haida Gwaii (Queen Charlotte Islands), British Columbia over a two-year period from June 2003 through February 2005. Data collected include: cross-shore topographic profiles, high and low tide sediment samples on profiles, aerial and ground photography and nearshore bathymetric surveys. Data collected for cross-shore profiles, bathymetric profiles, sediment analysis, coastal retreat rates and aerial photograph retreat figures are presented in appendices 1-6. Cross-shore profiles were collected to characterize and classify beach and backshore morphology and to complement profiles collected during previous research (1993 through 1997) (Barrie and Conway, 1996; Barrie and Conway, 2002). Seasonal beach response has been studied on the east and west coasts of North America using similar profiling methods by a number of researchers (e.g., Winant et al., 1975; Aubrey, 1979; Dubois, 1988; Corbau et al., 1999; Masselink and Pattiaratchi, 2001; Ruggiero et al., 2005). The intention of
conducting weekly profiles during May through August 2004 was to document seasonal response on a macrotidal coast in the northeastern Pacific.
Nearshore bathymetric surveys were conducted to provide a detailed resolution survey of the network of nearshore bars on the east coast to compare to the morphology of the north coast. Predicted water level data were provided by Canadian Hydrographic Service (CHS).
2.2 Surveys
2.2.1 Cross-shore profile site selection and rationale
Profile locations were chosen to reoccupy four sites surveyed by Barrie and Conway (2002) from 1993 through 1997. Two on North Beach (Elephant Cage and White Creek profiles) and 2 on East Beach (Naikoon and Cape Fife profiles) (Barrie and Conway, 2002) (Tables 4 and 5). A third reoccupied site on East Beach, approximately 1 km south of Cape Fife, was not re-established because initial bench marks were eroded. Additional sites were added in July 2003 to be representative of beach characteristics and morphodynamics in the study area. Survey sites were chosen based on apparent beach slope, backshore and nearshore morphology, and sediment characteristics. The intention of this was to survey representative beach units. Seven sites on the north coast and 3 sites on the east coast were monitored between 2003 and 2005 (Figure 2, Table 4). Most sites are located on the north coast due to the difficult logistics in accessing the east coast, particularly the southern portions.
Elevation (m C D ) 9. 59 71 7. 98 42 10 .1 728 9. 08 41 9. 34 51 10 .6 9. 75 69 7. 49 41 27 .852 Elevation (m ellip so id al) -0 .802 81 -2 .515 702 -0 .627 135 -1 .7 159 -1 .5 549 -0 .3 -0 .8 331 -3 .7 059 16 .6 525 NTS Ma p 103K /1 10 3J/ 4 10 3J/ 4 10 3J/ 4 10 3J/ 4 10 3J/ 4 10 3J/ 4 10 3J/ 4 10 3J/ 4 Ori ent at ion (Tr ue) 7 348 331 330 309 303 RSE 110 RS W 2 90 66 106 Zone (UT M ) 9 9 9 9 9 9 9 9 9 East ing ( U T M ) 2971 76 .53 3105 15 .86 3165 66 .27 3181 71 .90 3231 72 .16 3240 82 .46 3263 87 .55 3260 54 .33 3257 53 .36 No rt hi ng ( U T M ) 5991 455 .3 3 5992 564 .6 2 5995 396 .3 5 5995 778 .8 6 6000 161 .0 7 6001 441 .3 2 6005 184 .4 1 6003 906 .8 1 6000 360 .0 2 Profile Elepha nt C age Wh ite Creek Agate Beac h North Beach 1 (NB-1) North Beach 2 (NB-2) North Beach 3 (NB-3) Rose S pit East (RSE) and Rose S pit West (RSW Nai koo n Cap e Fife
Table 4. Profile site names and geogra
El epha nt Cag e Mar 1 6, 93 Sep 2 9, 95 Oct 2 1, 97 Jul ,04 Feb 1 8, 05 Wh ite C k Mar 1 6, 93 Feb 4, 94 Sep 2 9, 95 Oct 2 2, 97 Jul ,04 Feb 1 9, 05
Agate Beach Jul
20 ,0 3 Ma y 16 ,0 4 Ju n 2,04 Ju n 8,04 Ju n 14 ,04 Ju n 29 ,04 Jul ,04 Jul 7, 04 Jul 14 ,0 4 Jul 20 ,0 4 Jul 29 ,0 4 Feb 1 4, 05 NB-1 July 17,03 Ma y 8, 04 Ju n 3,04 Ju n 8,04 Ju n 14 ,04 Ju n 25 ,04 Jul ,04 Jul 3, 04 Jul 7, 04 Jul 14 ,0 4 Jul 20 ,0 4 Jul 30 ,0 4 Feb 2 1, 05 NB-2 Ma y 17 ,0 4 Ju n 2,04 Jul ,04 Jul 30 ,0 4 Feb 1 8, 05 NB-3 Jul 15 ,0 3 Feb 2 6, 04 Ma y 16 ,0 4 Ma y 22 ,0 4 Jul ,04 Jul 30 ,0 4 Feb 1 9, 05 RSW Ma y 17 ,0 4 Ju n 1,04 Ju n 14 ,04 Jul ,04 Jul 4, 04 Jul 31 ,0 4 Feb 2 1, 05 RSE Feb 2 6, 04 Ma y 17 ,0 4 Ju n 1,04 Ju n 14 ,04 Jul ,04 Jul 4, _0 4 Jul 19 ,0 4 Jul 31 ,0 4 Feb 1 5, 05 Nai koo n Mar 1 6, 93 Feb 4, 94 Oct 3 0, 94 Ma y 10 ,0 4 Ju n 1,04 Jul ,04 Jul 19 ,0 4 Jul 30 ,0 4 Feb, 05 Cap e Fife Jul 18 ,0 3 Ma y 8, 04 Ju n 1,04 Ju n 30 ,04 Jul 04 Jul 8, 04 Jul 17 ,0 4 Jul 31 ,0 4 Feb 1 7, 05 Name Date
coast of Graham Island. B
at
Profiles were surveyed once in the summer 2003 field season to obtain a summer beach type representation and subsequently in February of 2004 to document a winter beach state. During the summer 2004 field season all beach profiles were measured once and select profiles, (Agate Beach, North Beach 1, Rose Spit East, Rose Spit West,
Naikoon and Cape Fife, Figure 2), were measured on approximately a weekly schedule to document seasonal beach recovery in response to the lower energy wind and wave
regime following the storm season. These summer measurements document the rebuilding of the beach profile. Survey timing was adjusted to meet optimal low tide conditions and other logistical constraints (e.g., equipment availability, weather).
All surveys were conducted as close as possible to the lowest daily tide to record profile morphology as far as possible in the seaward direction. These profiles were
extended farther seaward as the tidal cycle approached spring tides and therefore captured a more extensive nearshore data set. During neap tides, the tidal range is quite low and does not allow for a long cross-shore transect. The profile at Cape Fife captured the nearest bar when surveyed during a low spring tide and was frequently measured using a total station. It was difficult to survey the first bar at Naikoon Profile as the depth of the landward trough at low tide was greater that 1.5 m. The second and third bars are predominantly subaqueous and only the bar crest is exposed at spring low tide. These bars are only measurable using boat- mounted bathymetric equipment and consequently were only surveyed twice (Table 5).
The weekly monitored profiles(three on the north coast and two on the east coast) were selected on aspect and morphology to provide a representative sample set of local
on the north coast and are subject to a predominantly W-NW wind climate. These beaches range from low slope sandy beach backed by foredune (NB-1) to moderately sloped beach-face backed by steep cobble berm terrace (RSW). Rose Spit East (RSE), Naikoon Profile, and Cape Fife Profile, all located on the east coast, were surveyed approximately on a weekly basis to document the changing morphology of a multiple barred beach in response to summer wave regime, and tidal cycles.
Once locations for cross-shore profiles were chosen, a compass bearing was taken along the coastline and the cross-shore profile azimuth was set at right angle to the coast in a seaward direction. A series of benchmarks was then installed to mark the profile origin. The benchmarks consist of a 1-2 m piece of rebar driven into the sand with approximately 0.05 to 0.25 m left exposed. These benchmarks were surveyed by RTK-GPS in July 2003 relative to the Tow Hill benchmark. All surveys were conducted in UTM zone 9. Though the extreme west end of the study area lies in UTM zone 8, these surveys were derived from control in UTM zone 9 and were not reprojected. Benchmarks for this survey were installed in July 2003. A compass bearing was taken along the coastline and the cross-shore profile azimuth was set at right angle to the coast in a seaward direction.
2.2.2 Establishing a survey network
Coastal surveys using Global Positioning System (GPS) satellites are referenced to an idealized earth shape represented by the WGS84 ellipsoid. In Canadian coastal research, this is typically reduced to mean sea level (MSL, approximated by a datum known as Canadian Geodetic Vertical Datum 1928 CGVD28) or to Chart Datum (CD) which in Canada represents Lower Low Water Large Tides (Figure 3).
Figure 3. Schematic showing relationships between ellipsoidal elevation as determined from GPS occupation at Tow Hill, Masset Mean Water Level as determined at Masset government dock and chart datum-mean water level separations.
To perform these datum transformations, the local separation between the ellipsoid and CGVD28 and/or CD must be determined using a gravimetric geoid model. The Canadian Geodetic Survey has developed a model for Canada called HTv2.0 which consists of the CGG2000 geoid, locally constrained by a corrector surface, itself
determined using first order leveling survey techniques. A problem arises in the study area in that first order leveling has not been conducted and, therefore, HTv2.0 cannot adequately determine the separation between the ellipsoid and CGVD28 (aka mean sea level). Typically this problem could be resolved using local tide gauge data linked to ellipsoid elevations, but no tide gauge is operational within the study area, and the two closest (Prince Rupert and Queen Charlotte City) are considered too distant to adequately represent mean sea level on the two coasts of the study area. To overcome these vertical datum issues, in 2004 a temporary tide gauge consisting of a Vemco Minilog pressure transducer was installed at the government wharf in Masset (set to seafloor with weight). The gauge operated from 24 February through 27 May 2004, recording water depths over the instrument at 20 minute intervals (Figure 4). At time of deployment, the vertical distance between the instrument and the deck surface of the government wharf was recorded to allow later conversion to water level relative to the ellipsoidal elevation of the wharf deck.
Elevation control for the wharf surface and all other control in the study area was provided by RTK GPS surveys linking to a brass cap in bedrock at Tow Hill (Figure 5). The coordinates and ellipsoidal elevation of this benchmark were obtained by processing a 13-hour occupation of the monument in 2003 using the Precise Point Position software of the Canadian Geodetic Service.
Figure 4. Time series plot showing pressure sensor data obtained from Masset Inlet during 24 February through 27 May 2004.
1 April 15 April 29 April 13 May 27 May 18 March
Figure 5. GPS receiver positioned over brass survey marker in bedrock at the seaward edge base of Tow Hill, Naikoon Park, Queen Charlotte Islands. This occupation was used to determine ellipsoidal elevation for the study area.
Once an ellipsoidal elevation of the wharf deck was determined, the mean ellipsoidal elevation of water levels from the Minilog was calculated and set as the zero elevation of a new datum termed Masset Mean Water Level 2004b (MMWL2004b). The vertical separation between the ellipsoid and MMWL2004b shown in Figure 3 is 7.4 m. This separation was used to convert all GPS-derived ellipsoidal elevations to a datum assumed to accurately represent mean water level in the study area. A mean sea level chart datum separation of 3.8 m was applied on the east coast and a separation of 3.1 m was applied to the north coast (separation values provided by Canadian Hydrographic Service, Sidney, British Columbia).
In order to reduce land-based surveys to Chart Datum used for all boat-based echosounding, separations between Chart Datum and mean water level in the study area were obtained from the Canadian Hydrographic Service. These separations vary from 3.1 m at the north end of Masset Inlet to 3.8 m on the east coast; depending on proximity to the model stations, one of these two values was used at each profile site.
2.2.3 Topographic surveying methods
Beginning in 2003, cross-shore profile data were collected using either a Topcon GTS 229 Total Station or an Ashtech Z-Xtreme Real Time Kinematic GPS (RTK-GPS). The data from both instruments consist of northing, easting and elevation (along with other ancillary data) accurate horizontally and vertically to 3 cm, however the processing differs. Total station and GPS have sub-1 cm instrument accuracy, however accuracy is reduced when the surveyors rod penetrates the ground surface. The survey rod can often sink 1-3 cm depending on whether the surface is dense (beach sand) or porous (aeolian
For the Total Station survey datasets, a simple linear correction was applied to the profiles to position them geographically using UTM coordinates of the survey control for the profile site. Data were corrected to the top of the pin in order to observe accretion at the base of the pin. Cross-shore profile data were extracted using software provided by Cansel Systems. Following data extraction, point data was processed using Microsoft Excel, and final plots were produced using Golden Softwares Grapher. UTM eastings and northings were reduced to single point data using Pythagoras. These single point data were then summed to provide a cross-shore distance. The elevation data from each survey point was corrected to the known elevation of the survey pin to correct each profile to chart datum elevation.
For the RTK-GPS, custom processing software provided by the Geological Survey of Canada was used to correct raw data to the survey control point (top of rebar) and output UTM coordinates. As was done with the Total Station survey data, further processing and analyses used Microsoft Excel and Grapher.
In addition to Total Station and GPS surveys, a line of alternating rebar and brass rods approximately 5 m apart was installed shore normal in 1997, roughly 500 m south of the Cape Fife Profile (Table 6). These survey pins have been used to document rates of coastal retreat on the east coast and were monitored up to 15 February 2005.
Table 6. Survey pin location at Cape Fife with occupation dates and amount of coastal retreat measured. As of 15 February 2005 there were 4 pins remaining on the bluff with roughly 4 m between each pin. Location of pin nearest to the bluff edge when last surveyed was 325663 E, 5999951 N, UTM Zone 9.
Date Occupied Retreat (approximate) Estimated retreat rates
October 22, 1997 Installed n/a
June 8, 1998 12.10 m 19.4 m yr-1 July 2003 2.90 m 0.58 m yr-1 February 22, 2004 5.6 m 9.6 m yr-1 May 2004 0.3 m 1.2 m yr-1 July 31, 2004 1.3 m 5.2 m yr-1 February 15, 2005 6.9 m 12.8 m yr-1
2.2.4 Bathymetric surveys
The local bathymetry was surveyed for all profile locations during June and July 2004 using a Simrad EQ30 echosounder operating at 200 kHz with a 7.5 degree beam angle coupled with a Garmin handheld GPSMap 76 and Canadian Differential Global Positioning System (CDGPS) to provide horizontal positioning with differential corrections. The horizontal accuracy of the Garmin GPS coupled with the CDGPS is approximately 3 m. The intention of the bathymetric surveys was to extend the land-based profiles into the nearshore and record the major bathymetric features such as subaqueous bars. During the 2004 survey, shore-normal bathymetric transects were conducted on the same lines as the 10 subaerial cross-shore profiles (Table 4). On the north coast, bathymetric surveys extended to depths of approximately 6 m below chart datum, the typical depth beyond which no major bedforms were identifiable on the echosounder.
On the East coast, the vessel was too small to navigate safely in Hecate Strait to survey past the seaward extent of major bedforms. Profiles were typically surveyed to 1.5 km offshore to depths deeper than 8 m below CD.Using custom software provided by the Geological Survey of Canada, depths were merged with GPS positions and corrected to Chart Datum using predicted tides for McIntyre Bay and Dogfish Banks from June 2004, provided by CHS.
2.3 Sediment sampling
Over the course of the study, high and low tide sediment samples were obtained repeatedly at cross-shore profile locations to provide a documentation of sediment grain size characteristics (Table 7, Appendix 3).
El epha nt Cag e Ju n 6 h l Ju n 9 h l Jul 28 hl Wh ite Creek Jul 28 hl
Agate Beach Feb 1
9 hl Ma y 16 hl Ju n 2 h l Ju n 8 l Ju n 14 h l Ju n 29 l Ju l 7 l Ju l 1 4 l NB-1 Feb 1 7 hl Ju n 3 h l Ju n 8 h l Ju n 14 h l Ju n 25 h Jul 7 hl Jul 14 hl Jul 20 hl NB-2 Ma y 17 hl Ju n 2 l Ju n 8 h l Jul 30 hl NB-3 Ju l 3 0 l RSW Feb 2 2 hl Ma y 17 hl Ju n 14 h l Jul 31 hl RSE Feb 2 2 hl Ma y 17 hl Ju n 14 l Ju l 1 9 l Jul 31 hl Nai koo n Ma y 10 hl Ju n 3 h l Jul 31 hl Cap e Fife Feb 2 3 hl Jul 8 h Jul 17 h Jul 31 hl Name Date (200 4)
Table 7. Sediment sample dates at each
profile location. The lette
rs h, l indicate high and lo
Sampling was conducted coincident with a cross-shore survey at or near spring tides to obtain representative samples at the most recent high tide line as well at ocean water level (OWL). Sediment samples were taken to augment profile descriptions and to document changes in sediment characteristics in response to changes in wind and wave climate as well as tidal cycle. Samples of approximately 0.5 kg were collected by hand (approximately 5 cm deep) at each profile location at the most recent high tide mark as well as at the ocean water line. These samples were split in half, dried and sieved in ¼ phi intervals for finer grained sands and in ½ phi intervals for coarser sediments
(gravels). Coarser gravels (cobbles) were not sieved, rather, they were photo-documented with ruler for scale. The samples contained no silt or mud, so settling tube methods were not used. Sediment characteristics are a direct response to the formative energy in the environment, and grain size analyses are employed to enhance the characterization of the profile morphology.
2.4 Aerial photograph analyses
Aerial photograph analyses provides a proven method for assessing coastal change (e.g., Manson, 1999; Forbes et al., 2004; Walker and Barrie, 2006). Aerial photography was quantitatively analyzed by scanning the photography and then orthorectifying using a Digital Elevation Model (DEM) and Ground Control Points
(GCPs) collected in a Geographic Information System (GIS) or remote sensing software. In this study, aerial photography at various scales from 1937, 1966, 1974, 1980, 1984 and 1997 (Table 8) was analyzed.
Table 8. Year, series number and scale of aerial photographs used in study.
Aerial photograph year Series # Scale
1997 BC97035-103J001 1:40,000
BC97035-103J002 BC97036/037-103G092 BC97035/036-103G082 BC97036-103J012
1984 UAG II 3092-122 Not known
UAG II 3092-123 UAG II 3092-124 UAG II 3092-125 UAG II 3092-126 1980 BC80008-123 1:12,000 BC80008-222 BC80008-223 BC80008-225 BC80008-226 BC80008-228 BC80008-229 1974 BC5630-188 1:15,000 BC5630-189 BC5630-220 1966 BC4362-220 1:15,840 BC4362-221 BC4362-223 BC4362-224 1937 BC19:51 Not known BC19:63 BC19:62 BC19:61
These aerial photographs were scanned at 600 DPI and imported into ESRI’s ArcMap. Using the georeferencing extension in ArcMap, GCPs were first collected on the 1997 ortho photograph mosaic and all preceding years were georeferenced to this 1997 set of ortho photographs. The northeast coast of Graham Island has few
anthropogenic features (roads, houses, power lines, and bridges) that can be used as GCPs for georeferencing aerial photographs. Consequently, stream junctions, lakes, and distinct trees that have remained unchanged over the aerial photograph history were used as key landmark features with which to georeference photographs. Aerial photographs were georeferenced using affine (1st order) polynomial transformation with cubic convolution resampling. No fewer than 6 ground control points were used when
georeferencing each image and only the centre portions of each aerial photograph were used by clipping the distorted edges during scanning to eliminate distortion around the edges. Error of georeferencing is usually reported as a Root-Mean-Square (RMS) error - a measure of how entered coordinates of the GCPs match computed coordinates in the transformed raster. The distance between the correct and transformed coordinates is known as the residual error and the RMS is the root mean of the squared residuals. Using this technique the RMS error was kept below 3.5 m.
The majority of 1937 aerial photography record was georeferenced and the coastline and bar features were digitized. However, only a select few aerial photographs from this set are included in this analysis as the backshore-nearshore environment is recorded on the outer 2 cm of the aerial photograph and distortion and error on this portion of the photograph are too great to make accurate interpretations of coastal change. The aerial photograph numbers are included in Table 8 for reference. The aerial
