Airflow and sand transport variations within a backshore–parabolic dune plain complex: NE Graham Island, British Columbia, Canada

(2006) 17–34www.elsevier.com/locate/geomorph

Geomorphology 77

Airflow and sand transport variations within a backshore–parabolicdune plain complex: NE Graham Island, British Columbia, Canada

Jeffrey L. Anderson, Ian J. Walker ⁎

Boundary Layer Airflow and Sediment Transport (BLAST) Laboratory, Department of Geography, University of Victoria, PO Box 3050,Station CSC, Victoria, British Columbia, Canada, V8W 3P5

Received 17 December 2004; received in revised form 12 December 2005; accepted 22 December 2005Available online 27 January 2006

Abstract

Onshore aeolian sand transport beyond the beach and foredune is often overlooked in the morphodynamics and sedimentbudgets of sandy coastal systems. This study provides detailed measurements of airflow, sand transport (via saltation andmodified suspension), vegetation density, and surface elevation changes over an extensive (325×30 m) “swath” of a backshoreforedune–parabolic dune plain complex. Near-surface (30 cm) wind speeds on the backshore ranged from 4.3 to 7.3 m s−1,gusting to 14.0 m s−1. Oblique onshore flow is steered alongshore near the incipient foredune then landward into a troughblowout where streamline compression, flow acceleration to 1.8 times the incident speed, and increasing steadiness occur.Highest saltation rates occur in steady, topographically accelerated flow within the blowout. As such, the blowout acts as aconduit to channel flow and sand through the foredune into the foredune plain. Beyond the blowout, flow expands, vegetationroughness increases, and flow decelerates. Over the foredune plain, localized flow steering and acceleration to 1.6 times theincident speed occurs followed by a drop to 40% of incident flow speed in a densely vegetated zone upwind of an activeparabolic dune at 250 m from the foredune.

Sediment properties reflect variations in near-surface flow and transport processes. Well-sorted, fine skewed backshore sandsbecome more poorly sorted and coarse skewed in the blowout due to winnowing of fines. Sorting improves and sands become fineskewed over the foredune plain toward the parabolic dune due to grainfall of finer sands winnowed from the beach and foredune.During the fall–winter season, significant amounts of sand (up to 110 kg m−2) are transported via modified suspension anddeposited as grainfall up to 300 m landward of the foredune. No distinct trend in grainfall was found, although most fell on thedepositional lobe of the blowout and at 200 m near an isolated, active parabolic dune. Grainfall amounts may reflect severaltransporting events over the measurement period and the transport process is likely via localized, modified suspension from thecrest of the foredune and other compound dune features in the foredune plain. This evidence suggests that the process of grainfalldelivery, though often overlooked in coastal research, may be a key process in maintaining active dunes hundreds of metres fromthe shoreline in a densely vegetated foredune plain. The effectiveness of this process is controlled by seasonal changes invegetation cover and wind strength as well as shorter term (e.g., tidally controlled) variations in sand availability from the beach.© 2006 Elsevier B.V. All rights reserved.

Keywords: Aeolian; Dune; Grainfall; Saltation; Foredune; Parabolic dune; Coastal swath; Driftwood

⁎ Corresponding author. Tel.: +1 250 721 7347; fax: +1 250 721 6216.E-mail address: [email protected] (I.J. Walker).

0169-555X/$ - see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.geomorph.2005.12.008

18 J.L. Anderson, I.J. Walker / Geomorphology 77 (2006) 17–34

1. Introduction

Aeolian processes play a key role in the geomor-phology of most sandy coastal systems by transportingsand delivered to the beach via littoral processes into thebackshore. This sand is then stored within incipient andestablished dune systems and occasionally cycled backto the littoral system via coastal erosion and stormsurges. Research on coastal aeolian dynamics to date hasfocused largely on several key areas including: (i) theinfluence of vegetation type and density on sedimententrainment and deposition (e.g., Hesp, 1981, 1983,1984, 1989, 2002; Buckley, 1987; Sarre, 1989; Arens,1996a; Arens et al., 2001a; Davidson-Arnott et al.,2003); (ii) incident wind angle and resultant beach fetcheffect and sediment supply to coastal dunes (e.g.,Jungerius et al., 1981; Nordstrom and Jackson, 1993;Arens et al., 1995; Davidson-Arnott, 1996; Davidson-Arnott and Law, 1996; van der Wal, 1998; Jackson andCooper, 1999; Bauer and Davidson-Arnott, 2003); (iii)topographic influences on near-surface wind speed anddirection (e.g., Svasek and Terwindt, 1974; Rasmussen,1989; Hesp and Hyde, 1996; Hesp and Pringle, 2001;Hesp et al., 2005; Walker et al., in press); and (iv) theeffects of moisture content on rates of sand transport onbeaches (e.g., Belly, 1964; Sarre, 1989; Kocurek et al.,1992; Namikas and Sherman, 1995; Arens, 1996b; vanDijk et al., 1996; Jackson and Nordstrom, 1997;Sherman et al., 1998; Wiggs et al., 2004). Theinfluences of these factors on beach–dune sedimenttransport have been well documented, although largelyindependent of one another. However, with theexception of remotely sensed, GPR or topographicsurvey-based characterizations of morphologicalchanges in coastal dune systems (e.g., Brown andArbogast, 1999; Bristow et al., 2000; Andrews et al.,2002) and empirical and/or conceptual models of barrierisland and beach–dune evolution and sediment balance(e.g., Armon and McCann, 1979; Hesp, 2002),relatively little research has been conducted on theprocess-response dynamics and controls of beach–dunesystems at the “meso” scale (i.e., morphologicalresponses at the landform assemblage scale over periodsof hours to years) (Psuty, 2004; Sherman, 1995).Furthermore, sediment delivery well into the backshore(e.g., 10s to 100s of metres beyond the foredune) viasuspended grainfall has received little attention incoastal research.

To address this, the purpose of this study is to examinethe influence of variations in vegetation and topographyon airflow and sediment transport over a spatiallyextensive “swath” of a backshore foredune–parabolic

dune plain complex. The study site includes severaldistinct geomorphic units of driftwood jammed back-shore, foredune, trough blowout and parabolic dune.Albeit limited in temporal scope, this study describes atypical onshore SE wind event representative of theformative winds in the study area. In addition, seasonalmeasurements of sediment deposition via suspendedgrainfall are presented, and implications for net sedimenttransport and dune maintenance are discussed.

2. Physical setting

The study site is located 15 km south of Rose Spit onEast Beach in Naikoon Provincial Park, NE GrahamIsland, Queen Charlotte Islands (Haida Gwaii) ∼80 kmoffshore of the central coast of British Columbia,Canada (54°N, 131°W, Fig. 1). The Naikoon Peninsulaconsists of a low plain of unconsolidated Quaternaryglaciofluvial sediments (Clague et al., 1982) that, duringa marine regression over the late Holocene, has beenreworked by energetic littoral and aeolian processes(Barrie and Conway, 2002; Walker and Barrie, in press),leaving a series of relict shorelines and progradingforedune ridges (Fig. 2). Over the twentieth century,relative sea level has risen at a rate of +1.6 mm a−1

(Abeysirigunawardena and Walker, unpublished data)and the coastline of East Beach has retreated at 1 to 3 ma−1 (Barrie and Conway, 2002) while the shores ofNorth Beach have prograded at 0.3 to 0.6 m a−1 (Harper,1980).

Littoral sediments are moved onshore by a highlycompetent wind regime dominated by strong SE windsin fall through winter and W–NWwinds in summer (seewind rose in Fig. 1). Annual average wind speed is 8.5m s−1 with b1% calm conditions. For the period 1995–1999, winds above the accepted sand transport thresholdof 6 m s−1 (Fryberger, 1979) occurred 67% of the time(Walker and Barrie, in press). Potential aeolian activityis high in the region with a total sand drift potential (DP)(per Fryberger, 1979) of 4566 vector units (VU) (Pearce,2005). This is well above those documented for desertregions (80–489; Fryberger, 1979) and for parabolicdunes in the Canadian prairies (300–1600; Wolfe andLemmen, 1999). The resultant drift potential vector(RDP) is 2967 VU aligned NW (316°), reflecting thedominant SE winds in the study area (Pearce, 2005).Despite a moist maritime (Cfb) climate and densevegetation and forest cover, the high onshore sandsupply in this wind regime maintains active parabolicdunes and foredunes (Figs. 2 and 3).

East Beach is subject to semi-diurnal mixed tidesranging 5–7 m with HHWMT exceeding 7 m. Annual

Fig. 1. Study site location on East Beach, Naikoon Provincial Park, NE Graham Island, Queen Charlotte Islands (Haida Gwaii), British Columbia.Wind rose (inset, upper right) derived from Environment Canada data from Rose Spit station (1995–1999) shows strong bimodal wind regime.Annual drift rose (inset, lower right) derived from the same data shows directional potential drift (DP) vectors and large resultant drift potential (RDP)vector toward the NW resulting from dominant, strong SE winds.

19J.L. Anderson, I.J. Walker / Geomorphology 77 (2006) 17–34

significant wave height (Hs) is 1.8 m and the peak periodis 10 s. Higher values of Hs to 3.5 m occur in theshallower waters of Dogfish Banks along East Beachand prevail for 20–30% of the time during wintermonths (Eid et al., 1993; Thomson, 1981). Beaches inthe study region are of the intermediate class (Masselinkand Short, 1993) and are wave-tide-dominated(Anthony and Orford, 2002). Multiple, transversenearshore bars occasionally weld to the shoreline(Figs. 2 and 3) and provide enhanced localized sandsupply to the backshore (c.f., Anthony, 2000; Aagard etal., 2004). At low tide, as much as 250 m of exposedbeach fetch that increases significantly to 500 m underoblique onshore SE winds (Fig. 3A).

The study site is a 325 m deep×30 m wide “swath”of the coastal landscape of East Beach. The site beginsas a driftwood jammed backshore that extends ∼50 to70 m landward from a 0.5-m high storm-cut scarp on the

beach to a low 0.3-m sparsely vegetated incipientforedune backed by a 5-m established foredune ridge(Figs. 3B and 5A). A 2.5-m deep trough blowoutextends through the foredune for ∼30 m and adjoins adepositional lobe that continues 40 m onto a hummockybackshore foredune plain. An isolated, partly vegetatedparabolic dune exists 175 m landward of the foredunewith an approximate surface area of 0.2 ha. Vegetationcover in the study site ranges from 0% to 95% densityand is dominated by three species of vegetation: Large-headed sedge (Carex macrocephala), dune grass(Elymus mollis), and Pacific alkali grass (Puccinellianutkaensis). Density of cover for these species isseasonally variable during the growth season from lateApril to late November. Tree species present includeSitka spruce (Picea sitchensis) and red alder (Alnusrubra). To date, few studies have considered the mesoscale suite of geomorphic features and surface

Fig. 2. Airphoto showing location of the study site on East Beach in the Naikoon Peninsula region of NE Graham Island, Haida Gwaii (Source: 1980National Airphoto Library, photo #A25613-39).


characteristics encountered by onshore airflow andsediment in transport well into the backshore. Thisstudy examines flow and sand transport responses over abroader landform assemblage (i.e., a coastal swath) ofdune form, surface roughness, and vegetation at bothevent-based and seasonal temporal scales.

3. Methods

3.1. Airflow properties

Airflow properties were measured throughout thestudy site using two methods: (i) high frequencymeasurements from ultrasonic anemometers, and (ii)time-averaged flow vectors from precise handheldanemometers. High frequency wind speed and directionwas measured from a transect of four Gill Windsonicanemometers at 30 cm above the surface (u0.3) sampledat 1 Hz. Co-located with each was a SAFIRE-typesaltation probe (Baas, 2003), a Guelph-Trent wedgetotal flux trap (Nickling and McKenna Neuman, 1997),a surface elevation pin, and a suspended sediment

grainfall trap (Fig. 4). Stations were installed in thebeach backshore, foredune trough, depositional lobe,and foredune plain locations (Fig. 5B). Incident flowconditions are referenced to the backshore instrumentstation.

Windspeed data from each ultrasonic station werenormalized by measurements at the backshore stationusing

U0:3 ¼ u0:3 station x=u0:3 station 1 ð1Þwhere x is station location 1 to 4. As such, normalizedU0.3 wind speeds provide a relative measure of flowacceleration and deceleration relative to incident flowconditions in the backshore. In addition, flow steadinessfactor, Fs was derived for each station using thecoefficient of variation

Fs ¼ u0:3r=u0:3 mean ð2Þ

where u0.3σ is the standard deviation of the wind speeddataset. As such, lower Fs values represent steadierflows.

Fig. 3. Oblique airphotos of the study site showing the extent of the backshore and driftwood jam (A) and a closer view showing an outline of thestudy swath within the backshore–foredune–parabolic dune plain complex (B).


Time-averaged velocity vectors were measured at50 cm (u0.5) at 39 locations over the study site (Fig.5B) using Kestrel 1000 handheld anemometers.Average and maximum wind speeds (u0.5 mean andu0.5 max, respectively) were recorded from the instru-ment over a 20-s interval. Normalized U0.5 values forall locations were produced using Eq. (1) and u0.5values. In that additional wind speed statistics werenot available from the handheld instruments (e.g., σ),a flow gust factor (Fg) was calculated for eachlocation using

Fg ¼ u0:5 mean=u0:5 max ð3Þ

As such, lower values of Fg represent gustier flowconditions. Flow vector direction was measured over thesame interval from the alignment of flow streamers(flagging tape) attached to surface elevation pins at 50cm using a Silva 2° surveying compass. Incident flowdirection remained essentially constant (varying by only5°) during this period. Given the relatively shortmeasurement interval, not all frequencies of gusts maybe captured in the gust factor. However, longer testintervals to 2 min at select locations revealed littledifference in u0.5 values. Measurements were collectedtwice at all locations over the 5 h experiment andcompared for representativeness.

Fig. 4. Instrument stations showing “SAFIRE” style saltation sensors, grainfall trap, surface elevation pin, “Guelph-Trent Wedge” style total flux trap,and Gill Windsonic™ ultrasonic anemometer.


3.2. Sediment transport

Two modes of sediment transport were measured inthis study. First, quasi-instantaneous saltation intensitywas measured using four SAFIRE-type saltationimpact sensors (Baas, 2003), each inset 5 cm abovethe surface and co-located with an ultrasonic ane-mometer and total flux trap (Fig. 4). Saltation intensitywas measured at the same frequency as wind speed(1 Hz). Second, grainfall of sediments in suspensionwas measured on staggered transects every 5 minland from the foredune at 39 locations. Grainfalltraps consisted of 10-cm sections of 10-cm I.D. PVCtubing attached to steel rods that also served as surfaceelevation monitoring pins (Figs. 4 and 5B). Traps weresituated 1 m above the surface to exclude significantsaltation inputs and a plastic sample bag was attachedto the bottom of each trap to collect sediment. Nograinfall was observed during the transport experimentin July, so traps were left in the field and were checkedat 2 months (18 September 2003) and 7 months (15February 2004). Only a very small amount of sediment(1 to 5 g) was observed in September at somelocations, and traps were emptied in February.Amounts of sand captured in the 0.008 m2 samplingarea were proportionally extrapolated to, and assumedto be representative of, the surrounding 1 m2 of surface

surrounding each trap over this period. Grainfallsamples were dried at 130 °C for 72 h then weighed.Total dry weights were then converted to quantities ofkilograms per square metre.

3.3. Vegetation density

Vegetation density was measured at 61 plots withinthe study site, each 50 m2 (5×10 m) (Fig. 5B). For eachplot, vegetation height, species and cover density wereestimated. The cover of all species in each plot was todetermine total density (i.e., vegetation cover by planttype was not distinguished).

3.4. Surface elevation change

A network of 61 surface elevation monitoring pinswas installed within the study area, in three separate andstaggered transects, each ∼5 m apart (Fig. 5B). Pinsconsisted of a 1.5-m piece of 1/2″ steel rod with a zeromark set flush with the initial surface. Surface changeafter the 5-h experiment was measured from this linewith a tape measure to a precision of 1 mm. A map ofsurface elevation change was produced initially usingSurfer's® default inverse distance interpolation algo-rithm and then was modified manually in graphicssoftware using point data to refine the map.

Fig. 5. Digital elevation model of study site (A) showing distinct geomorphic regions within the study swath as well as sampling and instrumentdeployment layout (B).


3.5. Grain size variations

Eight surface sediment grab samples were collectedalong the central axis of the study site in each

geomorphic region from the backshore to the parabolicdune (Fig. 5B). Samples were dried at 130 °C for72 h then sieved at 1/4 ϕ intervals from −1.0 to 4ϕ. Grain size statistics (mean grain size, sorting, and


skewness) were calculated from the frequencydistribution of each sample in a spreadsheet programusing the method of moments.

4. Results

This study presents 5 h of airflow and sedimenttransport data measured during a SE storm on 19 July2003. During this event, regional wind speed recordedat 5 m from the nearby BLAST02 met station (∼1 kmnorth of the study site) was 5 to 10 m s−1 (18 to36 km h−1) from the SE (125° to 130°). Near-surfacewind speed measured at 30 cm on the beachbackshore ranged from 4.3 to 7.3 m s−1, gusting upto 14.0 m s−1. Rain fell consistently during theexperiment and measured amounts at the met station(11.8 mm total) are likely an underestimate because ofunder sampling of the rain gauge during high winds.Despite this, aeolian sediment transport was observedat some locations in the study site.

4.1. Near-surface flow vectors

Normalized flow vectors (U0.3 and U0.5) throughoutthe study site are shown in Fig. 6A. Incident flowconditions averaged 5 to 10 m s−1 from 125° to 130°during the study period. Backshore flow vectors showslight topographic acceleration (U0.5 approaching 1.1)and steering in the alongshore direction (N) in thevicinity of the incipient foredune. A minor decelerationoccurs upwind of the established foredune to 0.9. Afterentering the trough blowout, flow accelerates from 1.3to 1.8 times that of the incident flow at station 1, andflow vectors steer slightly up the north wall of theblowout. Beyond the blowout, flow decelerates to 0.6due to flow expansion in the lee of the depositional lobethen accelerates gradually down the lobe to 1.1.Throughout the foredune plain, localized positiveslope effects on a NE slope at 160 to 250 m causeslight northward topographic steering and accelerationfrom U0.5=1.1 to 1.6. As flow encounters the denselyvegetated region upwind of the parabolic dune (40% to95%, Fig. 6C) from ∼250 to 275 m, U0.5 values dropfrom 1.0 to 0.4. On the parabolic dune, U0.3 increasesslightly to 0.6 on the stoss slope.

4.2. Flow gustiness (Fg)

During the study period, incident wind speed on thebackshore (u0.3 station 1) averaged 4.3 to 7.3 m s−1 withgusts up to 14.0 m s−1. An interpolated contour map ofFg derived from u0.5 values using Eq. (3) is presented in

Fig. 6B. Flow is slightly gusty at the dune toe (Fg=0.78)and becomes steadier over the incipient foredune (0.88)and through the blowout (from 0.84 to 0.89). Flowbecomes slightly gustier (0.80) at the break in slopefrom the blowout trough to the depositional lobeleeward of the foredune. Flow is gustier above themore densely vegetated surface of the foredune plain(0.80 to 0.72) while positive slope effects on the NEslope serve to increase steadiness (0.83 to 0.91). On theparabolic dune, flow is moderately gusty with Fg valuesfrom 0.79 to 0.82.

4.3. Vegetation density

Vegetation density in the backshore ranges from2% to 10% and increases to 20% on the incipientforedune (Fig. 6C). Vegetation cover on the estab-lished foredune increases from 20% to 30% at the toeto 50% at the crest. In contrast, vegetation density waslow (∼2%) in the trough blowout. Both the deposi-tional lobe and lower regions of the foredune plainwithin 50 m of the foredune crest have low vegetationcover (5% to 10%). Vegetation on the foredune plaingenerally increases in density from 25% downwind ofthe depositional lobe to a noticeable transition toN40% at 175 m. Higher hummocks and the NE slopein the foredune plain show higher vegetation densitiesfrom 30% to 60%. From 225 to 255 m, anotherdistinct increase in vegetation cover occurs from 70 to95, followed by a distinct drop in density toward thebare surface of the parabolic dune.

The windward edge of the parabolic dune is borderedby several sparsely spaced Sitka spruce ranging 2 to 4 mhigh, while the lee side of the dune head is flanked by adense (i.e., N85%) stand of Red alder (Figs. 3B and 7A).Spruce trees show significant abrasion damage andcrown flagging by dominant transporting winds. Alderstands are sculpted on the parabolic dune head into astreamlined profile by abrading near-surface winds (Fig.7B). Although not explored in this study, a distinctinteraction exists between near-surface airflow and treestands that exert some control on the morphodynamicsof this parabolic dune.

4.4. Sediment transport and grainfall deposition

Sand transport recorded from saltation sensors isplotted with normalized wind speed (U0.3) and flowsteadiness (Fs) in Fig. 8. Over the period of theexperiment, no saltation was recorded at stations 1(backshore) or 4 (vegetated foredune plain). Station 2 atthe mouth of the blowout recorded over 13,000 counts,

Fig. 6. Normalized flow vectors throughout the study area (A), flow gust factor (Fg) derived from 50-cm wind speed measurements (B), andvegetation density map (C).


and station 3 had 2896 counts. Total flux traps capturedinsignificant amounts of transport, perhaps because ofrainfall and the relatively small width of the trapopening.

Topographic forcing effects are evident in the near-surface wind speed, flow steadiness, and sedimenttransport data. In general, an inverse relationship existsbetween wind speed and flow steadiness, and more

Fig. 7. Oblique airphoto of vegetation associations with the parabolic dune including stands of Sitka spruce (Picea sitchensis) on the windward flanksand red alder (Alnus rubra) leeward of the head (A). Red alder stand on the head of the parabolic dune moves landward with the dune as saltationabrades the windward side of the stand (B), whilst avalanching and grainfall within the stand provide favourable conditions for growth on the leewardside.


sediment is moved in faster, steadier flows. For instance,as flow is forced toward and accelerates into the blowoutat station 2, wind speed increases from U0.3 of 1.0(station 1) to 1.6, flow becomes steadier (from Fs 0.3 to0.22), and moves the most sediment. Downwind of thedepositional lobe, flow at station 3 is slower andmoderately gusty (Fs 0.27). Flow at station 4 is slower,but similar in gustiness to that on the backshore (Fs 0.29).

No grainfall was observed during the experiment inJuly and only a very small amount (1 to 5 g) wasobserved in September at some locations. This suggeststhat the majority of grainfall at the site occurs during thefall–winter storm season. Between September 2003 andFebruary 2004, at least two major storms occurred inHaida Gwaii — one in November 2003 and the second

on 24 December 2003 with SE winds approaching110 km h−1. In this wind regime, winds approaching100 km h−1 are relatively frequent and occur in mostmonths on record (Walker and Barrie, in press) andare capable of transporting sand in suspension.

Given the long duration of sampling and inability toidentify the occurrence of transporting events, grainfalldata are not normalized over a shorter time interval.Grainfall data are shown as an interpolated contour mapin kilograms per square metre (Fig. 9B). Grainfalloccurred only landward of the foredune, varying by twoorders of magnitude from 1 to 110 kg m−2 with anaverage of 12.69 kg m−2. In general, no observabletrend was found in grainfall deposition with distancefrom the foredune, though most sediment fell in two

Fig. 8. Flow steadiness (Fs), normalized wind speed (U0.3), and saltation intensity (total impact counts) plotted against distance from the shoreline atstations 1 to 4.


locations. First, 108 kg m−2 sediment fell in theimmediate lee of the trough blowout (at the head ofthe depositional lobe) with quantities dropping rapidlyto 11 kg m−2 within 50 m of the foredune ridge. Onlysmall amounts (3 to 5 kg m−2) fell in the foredune plainregion from ∼170 to 250 m. Second, nearing theparabolic dune (i.e., beyond 250 m) grainfall increasesrapidly from 15 to 110 kg m−2 between the arms of thedune just upwind of a small stand of Sitka spruce. On theexposed parabolic dune itself, grainfall was low (1 to2 kg m−2) and increased considerably to 84 kg m−2

on the lee side slip-face within the Red alder stand.

4.5. Surface elevation change

Changes in surface elevation were measured after theexperiment at 61 monitoring pin locations that wereinitially set flush to the surface. An interpolated,spatially averaged contour plot of resulting areas oferosion (−values) and deposition (+values) is shown inFig. 9A. Quantities ranged from +9 to −11 mm over thestudy area. In the backshore region, an average of 8 mmof deposition occurred, with slight erosion (−2 mm) onthe windward slope of the incipient foredune. In the leeof the incipient foredune, 8 mm of deposition occurrednear the toe of the established foredune. Though littlechange in surface elevation occurred at the entry of the

blowout, most of the trough surface was erosional to amaximum of −11 mm toward the head. On thedepositional lobe, +4 to +6 mm of deposition occurredto a distance of ∼50 m downwind of the blowout head.No measurable deposition was noted in the seawardregion of the foredune plain from 160 to 200 m. In thedensely vegetated region upwind of the parabolic dune(from 210 to 275 m), the surface was generallydepositional although localized in amount. The southern(left) side of this zone experienced little change (±1mm), while the sloping north side showed +9 mm at 210m to −8 mm on the stoss slope of a small dune featurewith +4mmof deposition in the lee.Much of this surfacechange and deposition in this region is manifested insmall (0.15 to 0.30 m high) shadow dune features(Fig. 10). At the toe of the parabolic dune slighterosion (−2 mm) then deposition occurred on theupper stoss slope (+7 mm), erosion of −5 mm at thecrest, and deposition of +8 mm in the immediate lee.

4.6. Sediment properties

Mean grain size (ϕ), sorting (σϕ), and skewness (sk)values from eight surface samples along the central axisof the study site are shown in Fig. 11. The upper beachsample is a poorly sorted (σφ=1.367), fine skewed(sk=0.212) medium sand (0.102 ϕ or 0.932 mm). All

Fig. 9. Surface elevation change map derived from surface elevation pin data following the experiment (A) and spatially averaged grainfall derivedfrom 7 months of collection during the fall–winter season (B).


aeolian sands, from the backshore to the parabolic dune,are medium in size (1.427 to 1.684 ϕ or 0.372 to 0.311mm) and moderately well to well sorted (σϕ from 0.451

to 0.653). In general, as mean grain size increases,sorting (in ϕ units) becomes poorer. Sorting declinesfrom the backshore sample 2 (σϕ=0.484) into the

Fig. 10. Small shadow dune features (0.15 to 0.30 m high) found in the densely vegetated backshore ∼225 m from the shoreline.


blowout throat (σϕ=0.517) then improves slightlytoward the parabolic dune, then declines to moderatelywell-sorted on the dune head. Skewness values progressfrom fine (positively) skewed on the beach to coarse(negatively) skewed in the foredune–trough blowoutregion perhaps from progressive winnowing of finersands in this region. On the foredune plain, sediments

Fig. 11. Variation in sediment properties with distance from the beach to t(symmetrical).

change from coarse skewed (σϕ=−0.163) to symmet-rical (σϕ=0.052) then fine skewed on the parabolicdune (σϕ=0.177). This shift to symmetrical and fineskewed with distance from the foredune may reflectgrain fall deposits of finer sediments selectivelywinnowed and transported in modified suspensionfrom the beach and foredune region.

he parabolic dune. A dashed line is provided to indicate 0 skewness


5. Discussion

5.1. Backshore

The driftwood-laden backshore of East Beach posesan abrupt roughness on boundary layer airflow thatextracts momentum and sediment during dominantoblique onshore SE winds. As such, driftwood jamsact as a sink for coastal sediments in the backshore(Komar, 1976; Hesp, 1983, 1989; Walker and Barrie, inpress). On the beach, wind action on typically coarser,poorly sorted sands winnows the finer (medium sand)fraction, which produces well-sorted backshore depos-its. Aside from common supply- and transport-limitingfactors that control the aeolian transport process (e.g.,moisture content, vegetation cover), the rate of infillingof the driftwood matrix is also dependent on (i) thepresence of shore-attached intertidal bars (Figs. 2 and3A) that provide enhanced sediment supply (Anthony,2000; Aagard et al., 2004) and (ii) tide stage duringtransporting events that controls effective fetch andpotential sand transport (Wal and McManus, 1993;Jackson and Nordstrom, 1998; Hesp, 2002; Bauer andDavidson-Arnott, 2003). Although the rate of infilling isunknown, a considerable amount of sediment wastransported into the backshore over the 7 months ofthe grainfall observations, completely filling some areasof the driftwood matrix.

During this 7-month period, the storm cut scarp alsoretreated by ∼1 m in response to one of the higheststorm surges on record (0.7 m) on 24 December 2003(Abeysirigunawardena and Walker, unpublished data).This event occurred just after low tide, and the extent ofwave runup at the site is unknown. Remnant scarps inthe established foredune indicate that complete erosionof backshore driftwood jams by wave attack occurs on alonger timescale than our observations. Subsequently,driftwood and flotsam return, the roughness matrixrebuilds (in some areas nearing 2 to 3 m deep) andpromotes the trapping of aeolian sediments. The erosionand rebuilding of sediment-laden driftwood jams issimilar to Hesp's (1983, 1999) observations of incipientforedune rebuilding, although on a much larger scale;and in this case, vegetation is not required for dunegrowth within the matrix. This process is believed to beimportant for incipient dune formation, sedimentcycling and storage on similar beaches in the NEPacific (Walker and Barrie, in press). Such sedimentstores may also serve as an important buffer againsterosive winter storms and gradual sea-level rise.

Airflow over the incipient foredune shows slighttopographic forcing and acceleration. Nearing the

established foredune, flow is steered alongshore anddecelerates in response to increasing vegetation cover(from 2% to 20%) and possible flow stagnation upwindof the established foredune. This promotes sedimentdeposition and growth of the incipient foredune asshown by net positive surface elevation changes andconfirms other accounts from settings with little to nodriftwood (e.g., Hesp, 1983, 1989; Rasmussen, 1989;Sarre, 1989; Arens, 1996b; Arens et al., 1995, 2001a).

5.2. Foredune–trough blowout

As flow enters the blowout, it accelerates by as muchas 1.8 times that of incident flow on the beach and issteered up to the wall of the blowout. Once flow entersthe blowout, it becomes steadier and faster and promotesincreasing saltation. This confirms similar observationsby Hesp and Hyde (1996), Hesp and Pringle (2001) andsuggests that blowouts act as transport “conduits” thatchannel flow and sediment from a variety of incidentflow angles through the foredune into the foreduneplain. In addition, flow speed in this region is inverselyrelated to flow steadiness, which confirms Walker andNickling's (2003) wind tunnel observations of acceler-ated flow toward the crest of an artificial dune.Furthermore, in the narrowest reach of the blowoutwhere streamlines are most constricted, flow steadiness,speed, and surface deflation are greatest. These effectsare reflected in the more poorly sorted, coarser(winnowed) sands in this region. Under drier conditions,enhanced sand transport toward the head of the blowoutwould occur, promoting continued erosion of the troughand increased sand delivery into the foredune plain.

5.3. Vegetated foredune plain

Based on observations of as much as a 1000-folddecrease in near-surface wind speed, Arens (1996a)concluded that negligible amounts of sediment movebeyond the crest of vegetated foredunes via saltation.However, the influence of vegetation on sand transportis clearly contingent upon plant density, distribution,morphology, and height as well as timing during thegrowth season (Hesp, 1989, 2002). In essence, thehigher and denser the vegetation canopy, the greater thereduction in sediment transport. This study suggestshowever, that during the fall–winter season whenvegetation density is lowest and storm winds are morefrequent, significant amounts of sand are transported asfar as 300 m beyond the foredune. It is unlikely that thisresults from full suspension of grains from the foredunecrest over this distance. Rather, it is likely that grains are


transported via localized, modified suspension from thecrest of the foredune and other compound dune featuresfurther landward in the foredune plain. This evidencesuggests that modified suspension at the landscape scalemay play a key role in maintaining active dune forms onthe foredune plain distant from the shoreline.

Sediment channeled through the blowout is depositedinitially on a sparsely vegetated depositional lobedownwind of the trough as flow expands and decelerates(Hesp and Hyde, 1996) to ∼60% of the incident flow.With distance down the depositional lobe, flow thengradually accelerates into the foredune plain. Localpositive slope effects (e.g., the NE slope at 160 to 250 m)promote slight (northward) topographic steering andfurther acceleration of flow along the slope. Despitemoderate to dense vegetation cover, sediment erosionand deposition occur causing surface elevation changesin some areas of the foredune plain in response to thesetopographically induced flow patterns. Localized jettingand erosion around vegetated dunes also occurs (see Fig.3B) (c.f., Hesp, 2002) in the hummocky area betweenstations 3 and 4. Otherwise, as vegetation densityincreases from 40% to 95% deep into the foreduneplain (beyond 200 m from the foredune), surfaceroughness increases thereby reducing near-surface wind-speeds and sheltering the surface from wind action.

5.4. Parabolic dune

In coastal environments, parabolic dunes oftendevelop from the migration of depositional lobes ofblowouts (Hesp, 1999, 2002). However, no evidenceexists to suggest that this is the case for the parabolicdune at the study site. This dune appears to bemaintained by sediment delivered via grainfall 150 to200 m downwind of the foredune during high-magnitude events of moderate (seasonal) frequency. Itsshape is partly controlled by a bio-geomorphic interac-tion between woody vegetation on the flanks and headof the dune that alters near-surface airflow and trapssediment on the dune (Fig. 7). In turn, the dune providesa distinct micro-environment with a different distur-bance regime (i.e., wind and sand abrasion) andhydrology from that of the surrounding coastal plain.

5.5. Grainfall delivery

Sediment transport via saltation has been researchedextensively in beach and backshore settings over the pasttwo decades (e.g., Hesp, 1983, 2002, 2003; Hesp andHyde, 1996; Hesp and Pringle, 2001; Davidson-Arnottand Law, 1990, 1996; Arens, 1996a,b, 1997; Arens et

al., 2001a,b, 2002; Jackson and Nordstrom, 1998;Davidson-Arnott et al., 2003). In contrast, sand deliveryvia grainfall in dune systems has received comparativelyless attention, particularly in coastal research.

Over transverse desert dunes, Hunter (1985) ob-served that the rate of decay in grainfall leeward of thecrest could be described by a power function. Furtherwork by Anderson (1988) found an exponential decaywith deposition concentrated at 0.2 and 0.4 m from thebrink on the lee slope. More recently, Nickling et al.(2002) also observed an exponential decline in grainfalldeposition where up to 99% of total grainfall wasdeposited within 2 m of the crest for a variety of dunesizes and aspect ratios. Nickling et al. (2002) found,however, that Anderson's (1988) model, based onsaltation trajectories, under-predicted grainfall rates bymore than an order of magnitude. This was attributed tovertical lift and modified turbulent suspension of grainsby secondary flows in the wake region that cause longertransport paths than those of true saltation, as observedby McDonald and Anderson (1995). Detailed windtunnel measurements by Walker and Nickling (2002)confirm the presence of vertical lift and balancedvertical mixing in the immediate lee of transversedunes. These recent observations indicate that saltationis not the sole mechanism for sediment delivery overand beyond dunes and that secondary lee-side airflowpatterns (e.g., flow separation and reversal cells) have asignificant effect on dune sedimentary dynamics. Todate, very little research exists to document such effectsover coastal dunes.

No trend in grainfall deposition was found inlandfrom the foredune, although these data reflect theinfluence of several transporting events over the periodof study. For instance, large quantities farther inland maybe the product of a few, less frequent storms, whilst highquantities in the lee of the foredune and blowout mayreflect more frequent grainfall during lower magnitudeevents. McKenna Neuman et al. (2000) concluded thatdune morphology is a product of the frequency anddistribution of wind speeds above threshold as well asthe nature of the regional wind regime. In this study,grainfall driven by higher magnitude (and relativelyfrequent) SE storm wind events contributes a great dealto sediment delivery and, thus, to dune maintenancethroughout the foredune plain. During these events, near-surface flow may separate from the foredune crest andfrom other compound dune features in the foredune plainto transport grains in modified suspension for hundredsof metres beyond the beach — an order of magnitudefarther than Arens' (1996a,b) observations along theDutch coast. In this environment, the amount of


sediment delivered by this mechanism is appreciable andin the order of 10 to 100 kgm−2 over a fewmonths of thetypically stormy fall–winter season. As the parabolicdune is located near a zone of high grainfall, it seems thatmodified suspension, coupled with flow–vegetationinteractions (e.g., bluff body stagnation effects andsand trapping on the flanks of the dune), may beresponsible for maintaining active dunes hundreds ofmetres from the shoreline in an otherwise vegetatedforedune plain.

6. Conclusions

This study presents extensive measurements ofairflow vectors, vegetation density, sand transport, andseasonal grainfall delivery and surface elevationchanges over a 325×30-m swath of a topographicallycomplex backshore–foredune plain complex. Theresults indicate a clear need to consider onshore aeoliansediment transport well beyond the beach and foreduneas potentially significant in the morphodynamics ofsandy coastal systems. Key findings include:

(i) Flow vectors are topographically steered andforced. Oblique onshore flow is steered along-shore over an incipient foredune, landward at theestablished foredune, and into a trough blowout.In the blowout, flow acceleration to 1.8 times theincident flow and increasing flow steadinessoccur. Beyond the depositional lobe of theblowout, flow expands, vegetation roughnessincreases, and flow decelerates to 0.6 times theincident flow. Over the foredune plain, flowsteering and acceleration to 1.6 times slightoccur (due to positive slope effects) followed bya drop to 0.4 in a densely vegetated zone upwindof an active parabolic dune.

(ii) Topographic forcing influences the relationsbetween near-surface wind speed (U0.3), flowsteadiness (Fs), and sand transport. High-fre-quency wind speed and transport intensity fromsaltation probes reveal an inverse relationshipbetween U0.3 and Fs. More sediment is moved infaster, steadier flows within the trough blowoutand, via topographic steering and flow accelera-tion, the blowout channels flow and sedimentthrough the foredune into, and beyond, theforedune plain.

(iii) Sediment properties (sorting and skewness) areinfluenced by topographic forcing and themechanisms of sand transport. Well-sorted medi-um sands in the backshore become more poorly

sorted in the blowout throat, better sorted over theforedune plain, then more poorly sorted at theparabolic dune. Sands are fine (positively) skewedon the beach to coarse (negatively) skewed in theforedune–blowout region due to progressivewinnowing of fines. On the foredune plain,sands are coarse skewed to symmetrical andprogress to fine skewed toward the parabolicdune, perhaps because of increasing deposition offines transported in modified suspension.

(iv) Significant amounts of sand are transported inmodified suspension and deposited (up to 110 kgm−2) as far as 300 m beyond the foredune duringthe fall–winter season when vegetation density islow and storm winds are frequent. Although noobservable trend in grainfall was found, mostwas deposited in two locations: (a) in theimmediate lee of the foredune blowout and (b)upwind of an isolated, active parabolic dune∼200 m landward of the foredune. Thisdistribution may reflect the influence of severaltransporting events over the period of study and/or localized separation and modified suspensionfrom compound dune forms on the foreduneplain. Thus, sand delivery via modified suspesionand grainfall coupled may be significant inmaintaining active dunes hundreds of metreslandward of the shoreline.

Acknowledgements

Thanks are extended to Kim Pearce for fieldassistance and to Dr. S.A. Wolfe for helpful contribu-tions on research design and interpretation. Gratitude isalso extended to the Council of the Haida Nation and toNaikoon Provincial Park staff Dan Bates and LucyStefanyk for access and logistical support. Supportfunding was provided by an NSERC operating grant anda Canadian Foundation for Innovation New Opportuni-ties grant to IJW. Thorough reviews by Drs. BernardBauer and Karl Nordstrom also substantially improvedthis manuscript.

References

Aagard, T., Davidson-Arnott, R.G., Greenwood, B., Nielsen, J., 2004.Sediment supply from shoreface to dunes: linking sedimenttransport measurements and long-term morphological evolution.Geomorphology 60, 204–224.

Abeysirigunawardena, D.S. and Walker, I.J., unpublished data. Sealevel responses to climate variability and change in northernBritish Columbia, Canada. Manuscript in preparation.

Anderson, R.S., 1988. The pattern of grainfall deposition in the lee ofaeolian dunes. Sedimentology 35 (2), 175–188.


Andrews, B.D., Gares, P.A., Colby, J.D., 2002. Techniques for GISmodeling of coastal dunes. Geomorphology 48 (1–3), 289–308.

Anthony, E.J., 2000. Marine sand supply and Holocene coastalsedimentation in northern France between the Seine estuaryand Belgium. In: Pye, K., Allen, J.R.L. (Eds.), Coastal andEstuarine Environments — Sedimentology, Geomorphologyand Geoarchaeology. Special Publications of the GeologicalSociety of London. Geological Society of London, London,pp. 87–97.

Anthony, E.J., Orford, J.D., 2002. Between wave- and tide-dominatedcoasts: the middle ground revisited. Journal of Coastal ResearchSpecial Issue 36, 8–15 (ICS 2002 Proceedings).

Arens, S.M., 1996a. Patterns of sand transport on vegetated foredunes.Geomorphology 17, 339–350.

Arens, S.M., 1996b. Rates of aeolian transport on a beach in atemperate humid climate. Geomorphology 17, 3–18.

Arens, S.M., 1997. Transport rates and volume changes in a coastalforedune on a Dutch Wadden island. Journal of CoastalConservation 3, 49–56.

Arens, S.M., van Kaam-Peters, H.M.E., van Boxel, J.H., 1995.Airflow over foredunes and implications for sand transport. EarthSurface Processes and Landforms 20 (4), 315–332.

Arens, S.M., Baas, A.C.W., van Boxel, J.H., Kalkman, C., 2001a.Influence of reed stem density on foredune development. EarthSurface Processes and Landforms 26, 1161–1176.

Arens, S.M., Jungerius, P.D., van Der Meulen, F., 2001b. Coastaldunes. In: Warren, A., French, J.R. (Eds.), Habitat Conservation:Managing the Physical Environment. John Wiley and Sons Ltd.,Toronto, ON, pp. 229–272.

Arens, S.M., van Boxel, J.H., Abuodha, J.O.Z., 2002. Changes in grainsize of sand in transport over a foredune. Earth Surface Processesand Landforms 27, 1163–1175.

Armon, J.W., McCann, S.B., 1979. Morphology and landwardsediment transfer in a transgressive barrier island system, southernGulf of St. Lawrence, Canada. Marine Geology 31, 333–344.

Baas, A.C.W., 2003. Evaluation of saltation flux impact responders(Safires) for measuring instantaneous aeolian sand transportintensity. Geomorphology 59 (1–4), 99–118.

Barrie, J.V., Conway, K., 2002. Rapid sea level change and coastalevolution on the Pacific margin of Canada. Sedimentary Geology150, 171–183.

Bauer, B.O., Davidson-Arnott, R.G.D., 2003. A general framework formodeling sediment supply to coastal dunes including wind angle,beach geometry, and fetch effects. Geomorphology 49 (1–2),89–108.

Belly, P.Y., 1964. Sand movement by wind. Technical memo 1, UnitedStates Army Corps of Engineers, Coastal Engineering ResearchCenter, Washington, DC, 80 pp.

Bristow, C.S., Chroston, P.N., Bailey, S.D., 2000. The structure anddevelopment of foredunes on a locally prograding coast: insightsfrom ground-penetrating radar surveys, Norfolk, UK. Sedimentol-ogy 47 (5), 923–944.

Brown, D.G., Arbogast, A.F., 1999. Digital photogrammetric changeanalysis as applied to active coastal dunes in Michigan.Photogrammetric Engineering and Remote Sensing 65 (4),467–474.

Buckley, R., 1987. The effect of sparse vegetation cover on thetransport of dune sand by wind. Nature 325 (6103), 426–428.

Clague, J.J., Mathewes, R.W., Warner, B.G., 1982. Late Quater-nary geology of eastern Graham Island, Queen CharlotteIslands, British Columbia. Canadian Journal of Earth Sciences19, 1786–1795.

Davidson-Arnott, R.G.D., 1996. Measurement and prediction of long-term sediment supply to coastal foredunes. Journal of CoastalResearch 13 (3), 654–663.

Davidson-Arnott, R.G., Law, M.N., 1990. Seasonal patterns andcontrols on sediment supply to coastal foredunes, Long Point, LakeErie. In: Nordstrom, K.F., Psuty, N.P., Carter, R.W.G. (Eds.),Coastal Dunes: Form and Process. John Wiley and Sons, Toronto,ON, pp. 177–200.

Davidson-Arnott, R.G., Law, M.N., 1996. Measurement and predic-tion of long-term sediment supply to coastal foredunes. Journal ofCoastal Research 13 (3), 654–663.

Davidson-Arnott, R.G., Ollerhead, J., Walker, I.J. and Hesp, P.A.,2003. Spatial and temporal variability in intensity of aeoliantransport on a beach and foredune. In: R.A. Davis and P. Howd(Editors), Coastal Sediments '03. The Proceedings of The FifthInternational Symposium on Coastal Engineering and Science ofCoastal Sediment Processes, 2003, Sheraton Sand Key Resort,Clearwater Beach Florida. World Scientific Publishing Corp. andEast Meets West Productions, Corpus Christi, Texas, USA.CDROM-ISBN 981-238-422-7.

Eid, B., Calnan, C., Henschel, M., McGrath, B., 1993. Wind and waveclimate atlas volume IV: the west coast of CanadaTransportCanada (Report no. TP 10820E), Halifax, NS.

Fryberger, S.G., 1979. Dune forms and wind regime. In: McKee, E.D. (Ed.), A Study of Global Sand Seas. USGS ProfessionalPaper, vol. 1052. United States Geological Survey, Washington,DC, pp. 137–169.

Harper, J.R., 1980. Coastal Processes on Graham Island, QueenCharlotte Islands, British Columbia, Current Research, PartA. Paper 80-1A. Geological Survey of Canada, Ottawa, ON,pp. 13–18.

Hesp, P.A., 1981. The formation of shadow dunes. Journal ofSedimentary Petrology 51 (1), 101–112.

Hesp, P.A., 1983. Morphodynamics of the incipient foredunes in NewSouth Wales, Australia. In: Brookfield, M.E., Ahlbrandt, T.S.(Eds.), Eolian Sediments and Processes. Developments in Sedi-mentology. Elsevier, Amsterdam, The Netherlands, pp. 325–342.

Hesp, P.A., 1984. The formation of sand “beach ridges” and foredunes.Search 15 (9–10), 289–291.

Hesp, P.A., 1989. A review of biological and geomorphologicalprocesses involved in the initiation and development of incipientforedunes. Proceedings of the Royal Society of Edinburgh 96B,181–201.

Hesp, P.A., 1999. The beach backshore and beyond. In: Short, A.D.(Ed.), Handbook of Beach Shoreface Morphodynamics. JohnWiley and Sons Ltd., Toronto, ON, pp. 145–270.

Hesp, P.A., 2002. Foredunes and blowouts: initiation, geomorphologyand dynamics. Geomorphology 48, 245–268.

Hesp, P.A., 2003. ENSO and parabolic dune dynamics in the roaringforties, Manawatu coast, New Zealand. In: W. Kamphuis (Ed.),Proceedings of the Canadian Coastal Conference 2003, Queen'sUniversity, Kingston, ON, on CDROM.

Hesp, P.A., Hyde, R., 1996. Flow dynamics and geomorphology of atrough blowout. Sedimentology 43, 505–525.

Hesp, P.A., Pringle, A., 2001. Wind flow and topographic steeringwithin a tough blowout. Journal of Coastal Research, Special Issue34, 597–601.

Hesp, P.A., Walker, I.J., Davidson-Arnott, R.G., Ollerhead, J., 2005.Flow dynamics over a vegetated foredune at Prince Edward Island,Canada. Geomorphology 65, 71–84.

Hunter, R.E., 1985. A kinematic model for the structure of lee-sidedeposits. Sedimentology 32, 409–422.


Jackson, D.W.T., Cooper, J.A.G., 1999. Beach fetch distance andaeolian sediment transport. Sedimentology 46 (3), 517–522.

Jackson, N.L., Nordstrom, K.F., 1997. Effects of time-dependentmoisture content of surface sediments on aeolian transport ratesacross a beach, Wildwood, New Jersey, U.S.A. Earth SurfaceProcesses and Landforms 22, 611–621.

Jackson, N.J., Nordstrom, K.F., 1998. Aeolian transport of sedimenton a beach during and after rainfall, Wildwood, NJ, USA.Geomorphology 22, 151–157.

Jungerius, P.D., Verheggen, A.J.T., Wiggers, A.J., 1981. Thedevelopment of blowouts in “De Blink”, a coastal dune area nearNoordwijkerhout, the Netherlands. Earth Surface Processes andLandforms 6, 375–396.

Kocurek, G., Townsley, M., Yeh, E., Havholm, K., Sweet, M.L., 1992.Dune and dunefield development on Padre Island, Texas, withimplications for interdune deposition and water-table-controlledaccumulation. Journal of Sedimentary Petrology 62 (4), 622–635.

Komar, P.D., 1976. Beach Processes and Sedimentation. Prentice HallInc., Englewood Cliffs New Jersey. 429 pp.

Masselink, G., Short, A.D., 1993. The effect of tide range on beachmorphodynamics and morphology: a conceptual model. Journal ofCoastal Research 9, 785–800.

McDonald, R.R., Anderson, R.S., 1995. Experimental verification ofaeolian saltation and lee side deposition models. Sedimentology 42(1), 39–55.

McKenna Neuman, C., Lancaster, N., Nickling,W.G., 2000. The effectof unsteady winds on sediment transport on the stoss slope of atransverse dune, Silver Peak, NV, USA. Sedimentology 47,211–226.

Namikas, S.L., Sherman, D.J., 1995. A review of the effects of surfacemoisture content on aeolian sand transport. In: Tchakerian, V.P.(Ed.), Desert Aeolian Processes. Chapman and Hall, London, pp.269–293.

Nickling, W.G., McKenna Neuman, C., 1997. Wind tunnel evaluationof a wedge-shaped aeolian sediment trap. Geomorphology 18,333–345.

Nickling, W.G., Neuman, C.M., Lancaster, N., 2002. Grainfallprocesses in the lee of transverse dunes, Silver Peak, Nevada.Sedimentology 49, 191–209.

Nordstrom, K.F., Jackson, N.L., 1993. The role of wind direction ineolian transport on a narrow sandy beach. Earth Surface Processesand Landforms 18 (8), 675–686.

Pearce, K.I., 2005. Aeolian geomorphology of northeast GrahamIsland, Haida Gwaii (Queen Charlotte Islands), British Columbia.Unpublished MSc Thesis. Department of Geography, University ofVictoria, British Columbia, Canada. 184 p.

Psuty, N.P., 2004. The coastal foredune: a morphological basis forregional coastal dune development. Ecological Studies 171,11–27.

Rasmussen, K.R., 1989. Some aspects of flow over coastal dunes.Proceedings of the Royal Society of Edinburgh 96B, 129–147.

Sarre, R.D., 1989. Aeolian sand drift from the intertidal zone on atemperate beach: potential and actual rates. Earth SurfaceProcesses and Landforms 14, 247–258.

Sherman, D.J., 1995. Problems of scale in the modeling andinterpretation of coastal dunes. Marine Geology 124, 339–349.

Sherman, D.J., Jackson, D.W.T., Namikas, S.L., Wang, J., 1998. Wind-blown sand on beaches: an evaluation of models. Geomorphology22, 113–133.

Svasek, J.N., Terwindt, J.H.J., 1974. Measurements of sand transportby wind on a natural beach. Sedimentology 21, 311–322.

Thomson, R.E., 1981. An analysis of wind and current observationscollected in the Queen Charlotte Sound-Hecate Strait-dixonentrance region during 1954 and 1955. Institute of Ocean SciencesReport 81-10. Government of Canada, Marine Sciences Director-ate, Pacific Region, Victoria, BC. 84 pp.

van der Wal, D., 1998. Effects of fetch and surface texture on aeoliansand transport on two nourished beaches. Journal of AridEnvironments 29 (3), 533–547.

van Dijk, P.M., Stroosnijder, K., de Lima, J.L., 1996. The influence ofrainfall on transport of beach sand by wind. Earth SurfaceProcesses and Landforms 21, 341–352.

Wal, A., McManus, J., 1993. Wind regime and sand transport on acoastal beach–dune complex, Tentsmuir, eastern Scotland. In: Pye,K. (Ed.), The Dynamics and Environmental Context of AeolianSedimentary Systems. The Geological Society, London, UK, pp.159–171.

Walker, I.J., Barrie, J.V., in press. Geomorphology and sea-level riseon one of Canada's most ‘sensitive’ coasts: Northeast GrahamIsland, British Columbia. Journal of Coastal Research, SpecialIssue 39.

Walker, I.J., Nickling, W.G., 2002. Dynamics of secondary airflow andsediment transport over and in the lee of transverse dunes. Progressin Physical Geography 26 (1), 47–75.

Walker, I.J., Nickling, W.G., 2003. Simulation and measurement ofsurface shear stress over isolated and closely spaced transversedunes. Earth Surface Processes and Landforms 28, 1111–1124.

Walker, I.J., Hesp, P.A., Davidson-Arnott, R.G.D., Ollerhead, J., inpress. Topographic steering of offshore airflow over a vegetatedforedune: Greenwich Dunes, Prince Edward Island, Canada.Journal of Coastal Research, J.R. Allen Memorial Special Issue.

Wiggs, G.F.S., Atherton, R.J., Baird, A.J., 2004. The dynamic effectsof moisture on the entrainment and transport of sand by wind.Geomorphology 59, 13–30.

Wolfe, S.A., Lemmen, D.S., 1999. Monitoring dune activity in theGreat Sand Hills region, Saskatchewan. Geological Survey ofCanada Bulletin 534, 199–210.

Airflow and sand transport variations within a backshore–parabolic dune plain complex: NE Graham Island, British Columbia, Canada

Documents