Dynamics of large and small scale bedforms on a macrotidal shoreface under shoaling and breaking waves

Marine Geology, 115 (1993) 207-226

207

Elsevier Science Publishers B.V., Amsterdam

Dynamics of large and small scale bedforms on a macrotidal shoreface under shoaling and breaking waves P h i l i p D . O s b o r n e ~ a n d C h r i s t o p h e r E. V i n c e n t

School of Environmental Sciences, University of East Anglia, Norwich, NR4 7TJ, UK (Received March 8, 1993; revision accepted July 27, 1993)

ABSTRACT Osborne, P.D. and Vincent, C.E., 1993. Dynamics of large and small scale bedforms on a macrotidal shoreface under shoaling and breaking waves. Mar. Geol., 1'15: 207-226. Bedform dimensions, bed position changes, near-bed velocities and suspended sand concentrations are analyzed from measurements at a single location on a macro tidal beach in the south west of England. The study was conducted in 0.5-2.25 m water depth under both swell and wind-generated storm waves with both weak and strong currents present. Bed positions and suspended sand concentrations were measured using the 3 transducers of a multi-frequency acoustic backscatter sensor with 5 mm vertical resolution. Two distinct bed types, based on wavelength (2), were observed, each with two or more subtypes possible: (I) small-scale bedforms (2< 20 cm) which include two-dimensional pre-vortex (1), vortex (2), post-vortex (3) and three-dimensional vortex (4) forms; (II) large-scale bedforms (2 > 20 cm) which include two-dimensional (5) forms and threedimensional vortex (6) forms. Small-scale forms were dominant under non-breaking conditions while the large-scale forms occurred under both non-breaking and breaking waves; the large-scale forms dominate under breaking conditions. Both types, but particularly low steepness forms, were highly mobile with maximum horizontal migration rates of 5 cm min- 1 Large (upto 15 cm) and rapid (upto 3.0 cm min -1) changes in vertical bed elevation were also observed in association with the development and migration of large-scale forms. Large-scale bedforms were also highly variable spatially, often being interspersed with smaller scale forms under decaying flow regimes and with areas of flat bed under increasing regimes. Observations suggest these forms are present even under high energy surf zone conditions (wave Shields /> 1). Models for predicting ripple dimensions did not perform well in this environment. This lack of agreement reflects the complex hydrodynamic regimes associated with random (grouped) shoaling and breaking waves together with the presence of currents, often at large angles to the waves, as well as the rapid rates of change in the wave forcing associated with tidal cycle oscillations in this macrotidal environment. Suspended sediment concentrations and transport rates are particularly sensitive to the bedforms present and also to bed position changes associated with ripple migration. Estimates of transport rates are subject to potentially large errors (upto 30%) without compensation for bed elevation changes relative to sensor position.

Introduction Recent m e a s u r e m e n t s o f s u s p e n d e d s a n d conc e n t r a t i o n s a b o v e the sea-bed (e.g. V i n c e n t a n d Green, 1990; G r e e n w o o d et al., 1990) have raised questions a b o u t the m e c h a n i s m s c o n t r o l l i n g the re-suspension o f s a n d a n d o f o u r ability to predict sand c o n c e n t r a t i o n s a n d sand t r a n s p o r t rates in the m a r i n e e n v i r o n m e n t . Bedforms generated by waves, currents, a n d sediment t r a n s p o r t a p p e a r to tPresent address: Department of Geography, Scarborough Campus, University of Toronto, Scarborough, Ont. M1C 1A4, Canada 0025-3227/93/$06.00

play crucial roles in these processes by m o d i f y i n g n e a r bed flows, by e n h a n c i n g the e n t r a i n m e n t a n d re-suspension o f sediment, a n d by influencing the timing a n d spatial d i s t r i b u t i o n o f sand resuspension. Models presently used assume that the b e d f o r m s are u n i q u e l y predictable from relatively simple flow a n d grain size parameters: o u r results show this is n o t necessarily the case a n d that timescales o f bed f o r m a t i o n are quite different from the time-scales of waves or wave groups. K n o w l e d g e o f ripple history, time-scales o f bed response a n d accurate p o s i t i o n i n g o f i n s t r u m e n t a tion m u s t be k n o w n before the kinematics o f the

(C') 1993 - - Elsevier Science Publishers B.V. All rights reserved.

SSDI OO25-3227 ( 93 ) EO112-.1

208

suspension process can be interpreted. Up to now, there have been few comprehensive field investigations of ripple formation, ripple migration and bed response under waves and currents (see review by Boyd et al., 1988) and fewer still have been made with associated sediment transport measurements using fast response sensors. The data described here were collected as part of the Bedforms and Suspension EXperiments (BASEX)field programme designed to investigate the processes controlling sand re-suspension under non-breaking waves and currents, particularly the control exerted by bedforms. Bedforms and local bed elevation changes, together with suspended sand concentrations, near-bed velocities and water surface elevations, were measured at a single location near the mid-tide position on a macro-tidal shoreface in the southwest of England. The measurements were taken over a 9 day period in August 1991 during spring tides (range upto 5 m) with mild swell and low energy wind-wave conditions and over a 12 day period in August 1992 with both swell and wind-generated storm waves present. Vincent and Osborne (in press) present a method for gaining information on bedform wavelength and migration using static arrays of high frequency acoustic backscatter sensors (ABS) and summarize such data obtained from the second BASEXexperiment. In this paper, we focus on the analysis of bed position measurements and direct bedform measurements of height, spacing, planform and profile which were obtained by SCUBA divers and high-frequency ABS. Measurements of ripple height and wavelength are compared with those obtained from various predictive formulae. Timeseries of bed positions and reconstructed bed profiles are used to examine the timescales of bed response and bedform development to changing hydrodynamic conditions. The implications with respect to sand re-suspension and sand transport rates are also considered.

Site description and field experimental design Longsands Beach situated on Whitsand Bay, Cornwall (Fig. 1) is a macrotidal beach (spring range up to 5 m) facing south-southwest and

P.D. OSBORNE AND C.E. VINCENT

exposed to storm waves and swell from the Atlantic Ocean. The shoreface is essentially concave upwards with a slope of 1:15 near the Mean HWL and decreasing rather abruptly to a fairly uniform slope of 1:70 at approximately 50 m offshore (Fig. 2). Sand samples from this latter slope were unimodal size distributions (modal size = 212 Ixm or 2.25 phi) when dry sieved at 0.25 phi intervals. The shoreface exhibits a distinct longshore periodicity with cuspate horns separated by shallow embayments on a wavelength of approximately 200 m or more. During the storm waves which occurred near the end of the August 1992 experiment, it was confirmed that the embayments or channels were occupied by well-developed rip currents. The instrumented station was situated near the crest of a cuspate horn in both experiments. Instrumentation, consisting of a 3 frequency acoustic backscatter sensor (ABS), a Valeport 8003.2 discuss head electromagnetic current meter (EMCM), and a Paro-scientific Digiquartz pressure transducer were mounted on an H-frame situated approximately 150 m offshore from the mean HWL (Figs. 2 and 3). The sensors were shore-connected and the data were recorded on a single PC-based data logging system. In addition, 3 optical backscatter sensors (OBS) and 3 syringe suction samplers were deployed in a vertical array close to the ABS during the 1991 experiment. The results from the suction samplers and the OBS are to be presented elsewhere (M.O. Green, pers. commun., 1992). Instrumentation was deployed in a similar configuration in both experiments but with an additional ABS sensor mounted in side scan mode (Vincent and Osborne, in press) in 1992. The OBS and additional current meters were not available for the second deployment. Local bed elevations were measured with a vertical resolution of 5 mm at 3 horizontal positions in a shore-normal transect using 3 ABS transducers. The transducers were mounted approximately 0.6 m above the sea bed and spaced approximately 0.1 m apart and 0.2 m apart in 1991 and 1992, respectively. Bedform geometry and bedform migration rates were measured using a combination of the ABS, and continuous SCUBA diver observation and measurement. Two types of diver observations were made:

BEDFORM DIMENSIONS ON A MACROTIDAL SHOREFACE

209

TORPOINT --

~.

= L o o ~ . . s .

.....

~ , ~

............

THE SOUND

0 I

"i: :i'): , /!/

fikm

I

..... :!:::

Fig. 1, Location of study.

i:;l \ -4.5-

-5. ~o

~o

~o

~o 1~o 1~o 1~o 1~o 1~o 26o 2~o 2~o 2~o 2so D~I'ANCE RELATIVETO BASEOF CLIFF(m)

--

August 5 - -

Aegust 8 - -

August 10 - -

1 August 13 ]

Fig. 2. Profiles across Longsands Beach, Whitsand Bay during the experimental period. (1) m e a s u r e m e n t s o f the b e d profile b e n e a t h the A B S t r a n s d u c e r s were m a d e b e t w e e n d a t a collection b u r s t s b y m e a n s o f a g r a d u a t e d m e a s u r e m e n t device (described below).

(2) detailed surveys o f the b e d f o r m g e o m e t r y in a 5 m 2 a r e a situated a p p r o x i m a t e l y 2 m s e a w a r d o f the H - f r a m e were m a d e d u r i n g d a t a collection runs.

P.D.OSBORNEANDC.E.VINCENT

210

3-5.5 CM SPHERICAL

EMCMs~

SYRINGE SAMPLERS 3 ABS TRANSDUCERS

CAMERA~j-~I

3.2 CM DISCUS EMCM F L

30BB /

/ Lk

BAMLER,?TAES /q l: 7//'/7-/-f/f/~,

~ / / / / / / / . / / / / / / / / / / / / ~

V/Z///

SECTION

ABS TRANSDUCERS

IoI,. ,MHz

PLAN

"

~' WEST ONSHORE 50cm L

SCALE

J

Fig. 3. Schematic of the instrumented H-frame.

The bedform measurement device consisted of 2 steel rulers approximately 0.5 m in length which were graduated in millimetres and intersecting at 90 ° to one another. One ruler was fixed parallel to the transducers for measuring horizontal scales in the on-offshore direction while the other ruler was capable of sliding both vertically and horizontally relative to the other for measuring vertical scales at various horizontal positions. In addition, an underwater camera was mounted adjacent to the ABS sensors. The E M C M was fixed to a vertical a r m of the H-frame at a nominal elevation of 0.1 m. The ABS used in this experiment operates at a basic rate of 50 Hz; each transducer transmitting

a short ( ~ 12 ItS) pulse sequentially then listening for the return echo. Sampling the backscattered return echo at 150 kHz and range gating allowed the sediment concentration profile and bed position to be estimated with a vertical resolution of 5 mm. In the basic data collection mode, profiles from 12 acoustic transmissions (pings) were averaged and a single profile recorded for each transducer every 0.24 s (4.22 Hz) with current speeds from the E M C M and other sensors collected for each profile. Burst duration was ~ 8 minutes. A second collection mode allowed longer burst duration (,-~ 17 minutes) by operating at 2.2 Hz and averag-

211

BEDFORM DIMENSIONS ON A MACROTIDAL SHOREFACE

ing 24 pings per profile. The first mode provided the maximum spatial and temporal resolution, while the second mode was useful for tracking temporal changes in bed position.

smooth the bed position series and remove random instantaneous variations above and below the average bed position.

Acoustic measurements of the sea bed position

Description of bedforms

High-frequency (>1 MHz) Acoustic Backscatter Sensors have been used occasionally for measuring small-scale (both spatial and temporal) changes in sea bed position for nearly two decades (see for example, Dingier, 1974; Dingier et al., 1977; Greenwood et al., 1985; Hanes et al., 1988; Vincent et al., 1991). ABS detect the presence of the bottom boundary due to the sharp change in acoustic impedance between the fluid and bed which causes a strong acoustic reflection (Hanes et al., 1988). In most acoustic profiles of the water column, the bottom location is clearly identified by a sharp peak in the backscattered pressure (see for example Hanes et al., 1988). Unfortunately, the presence of high concentrations of suspended sediment, especially in the near-bed region, and the presence of bubbles due to wave breaking can make the detection of the bed position unclear at times. In particular, high concentrations of suspended sediment near the bed result in significant attenuation of the acoustic beam and may temporarily shift the position of maximum acoustic backscatter away from the sea-bed (Vincent et al., 1991). Often the shape of the bottom echo is more complex than a single sharp peak and may be composed of a number of peaks which may be due to returns from substantial, but transient, features at or below the bed surface (Green and Boon, 1988) or to variable echoes from adjacent ripple crests and troughs within the finite width of the acoustic beam (Vincent and Downing, in press). Fortunately, these deviations from the "true" target are relatively short-lived phenomena thus making their detection and removal possible by a measure of both qualitative judgement as well as some de-spiking and temporal averaging. An algorithm was developed to detect the maximum acoustic return from each profile and produce time-series of bed positions. A simple binomial filter and a de-spiking algorithm were also used to

A total of 36 observation dives were made during the two experiments to record bedform height, wavelength and form. During these observation dives, measurements were made at regular intervals during and between data acquisition runs. In order to make the measurements more systematic and the description of forms less subjective, a set of four classification criteria was adopted:

A. Size (based on height and spacing) small-scale

(ripples):

height < 4 cm;

spacing

1-20 cm

large-scale (megaripples): height > 4 cm; spacing 20-100 cm

B. Type of vortex shedding~suspension observed." pre-vortex or rolling grain ripples--no vortex shedding and no suspension vortex--suspended sediment and vortices are shed from the crests of ripples at regular intervals under large waves post-vortex--sediment suspension and vortex shedding occurs as irregular bursts, ripples may or may not be detectable.

C. Number of crest dimensions: flat bed: no visible crests 2-dimensional: includes straight, sinuous, and lunate forms, long-crested and short-crested types with low bifurcation density ( > 1 0 c m between bifurcations) 3-dimensional: includes cross-ripples, complex irregular forms and ripples with high bifurcation density and numerous crest intersections (< 10 cm between bifurcations).

P.D. OSBORNE AND C.E. VINCENT

212

were slightly asymmetric and crescentic. The minor crests were oriented obliquely to, but in the same general direction as, the longshore current (Fig. 4d).

D. Profile." symmetric asymmetric indeterminate Six distinct bedform types, based on the above scheme, were observed during both experimental periods (Table 1; Fig. 4). Types 1, 2 and 3 have essentially the same planform but differ in terms of steepness, symmetry and suspension characteristics. Generally, the small-scale forms were dominant under non-breaking waves while large-scale bedforms were dominant under asymmetric and breaking waves. Crests of small-scale twodimensional forms (Types 1, 2 and 3) were generally oriented normal to the direction of wave propagation (essentially shore-normal) (Fig. 4a). However, crests of large-scale two-dimensional forms (Type 5) were generally oriented at approximately + or - 4 5 ° to the shore-normal with the orientation direction determined by the direction of the longshore current (Fig. 4c). These highly mobile forms were crescentic in planform, asymmetric landward and migrated onshore under breaking waves; they have been described as lunate-shaped megaripples (e.g. Clifton, 1976). Largescale three-dimensional forms (Type 6) were either well-developed cross-ripples with major and minor crests intersecting at 90 ° and oriented at 45 ° to shore-normal, or, transitional forms resembling the intersection points of the cross-ripples and appearing as large "hummocky" forms. In the case of well-developed cross-ripples, the major crests were symmetric and straight, while the minor crests

Bedform height and wavelength Calculation of the sediment transport by waves and currents over a sandy substrate requires a detailed knowledge of the frictional (form-drag) effects induced by bedforms. Therefore, a large effort has been directed at predicting wave-formed ripple geometry as a function of hydrodynamic and sediment parameters. Probably the most comprehensive physically based model of the bed roughness under unsteady oscillatory flows is that of Grant and Madsen (1982) which followed from their work on combined wave and current interaction with a rough boundary (Grant and Madsen, 1979). The model partitions the roughness into two distinct contributions which are due to the form drag around individual bedforms and to the near-bed sediment transport. The form drag is treated as a function of the ripple geometry and the shear stress. Using data from the laboratory experiments by Carstens et al. (1969) and other sources, Grant and Madsen (1982) present a set of empirical equations for the prediction of waveformed ripple geometry as a function of the boundary shear stress, i.e. skin friction. The ripple heights ~/and wavelengths 2 are given by: /~,\-o.16 ~-a-- 0.22 ~ )

(1)

TABLE 1 Summary of bedform observations and classification SIZE

#CRESTS

SHEDDING SUSPENSION PRORLE HEIGHT (m)

2-dlmermionaJ Pm-vo~x SMALL- SCALE Vortex

Post-vortex 3-dimensionaJ Vortex

TYPE

# OBSERVED

Us im/sI

h (m)

WAVE CONDITIONS

1 2 3

1 18 B

<0.60 >1.00 0 , 4 5 - 0 . 7 2 1.80-2.29 0.B0-1.20 0.45-2.03

symmetric shoaflng syrnmetdc/dnoaiing

0.15 inds~nninats

4

10

0.71-0.90 1.03-1.60

shoaling

5

16

0.62-1.14 0.30-1.54

breaking

6

12

0 . 5 1 - 0 . 9 7 0.44-2.02

shoaling/breaking

I

2-dimer~ional Vortex

,

0 symmetric 0.1 sym/Myrn/indi 0.05 indetsmdnate

<0.06- >0.2 asymmeffi¢

I

LARGE-SCALE a-dimensional Vor~x

>0.2

sym/~PFn/Ind

BEDFORMDIMENSIONSON A MACROTIDALSHOREFACE a)

213 b)

SMN.L SCALE TWO DIMENSIOkk~L (Types 1,2,3)

~.

SMALLSCALETHREEDIMENSIONAL(TYPE4)

~ .;'~c~;.

c) LARGE- SCALE TWO-DIMENSIONAL (TYPE 5)

d) LARGE SCALE THREE-DIMENSIONAL (TYPE 6)

i) CROSS-

,,~,'

,

~: .~:. :,:~. : .,,~ ~:~ j.~.:~,r.~:, • .,.::~,..,?~,~.~ ': :~-":

m~tns

) WASHED-OUT CROSS R PPLES OR HUMMOCKS •

~:

....

~.

...... : : ~ , . . : ' ~ .

-. - : ~ .

Fig. 4. Schematic representations of bedform types observed during the BASEXfield experiments: (a) small-scale two-dimensional (Types 1,2,3); (b) small-scale three-dimensional (Type 4); (c) large-scale two-dimensional (Type 5); (d) large-scale three-dimensional (Type 6).

2

_ _ _

6.25

\¢o1

(2)

below the break-off point where ripple steepness is increasing, and at/= 0.48SO.8 \~b¢/

(3)

'~ 3.6S, o.6 (~k¢') ~/

(4)

above the break-off point where ripple steepness begins to decrease, where a is the water orbital excursion amplitude, ~' is the maximum Shields parameter over a wave cycle, gcc is the critical Shields parameter, S. is a dimensionless sediment parameter. The model developed by Grant and Madsen (1979, 1982) and Glenn and Grant (1983), henceforth referred to as the Grant-Madsen-Glenn or GMG model, has been used here to predict the

214

P.D.OSBORNEANDC.E.VINCENT

equilibrium bedforms using the combined flow characteristics and boundary shear stresses estimated from the wave and current data collected by the EMCM during the field study; the onshore significant wave-induced current being used to represent the wave conditions in each run. The bedforms predicted by the G M G model rely to a large extent on laboratory data of Carstens et al. (1969). However, Nielsen (1981) concluded from an analysis of both field and laboratory data that the size and shape of field ripples are strongly influenced by the irregularity of the waves. Under irregular waves, as found in the field, the ripples are shorter and flatter than under regular laboratory waves. Nielsen argues that although ripple wavelength is a very complex function of flow and sediment parameters, the single most important factor in determining both ripple height and wavelength is a sediment mobility number (0) of the form:

0

field and laboratory measurements which have demonstrated the dependence of ripple wavelength on orbital diameter and wave period below a critical value which is a function of sand size (e.g. Inman, 1957; Mogridge and Kamphuis, 1972; Clifton, 1976; Lofquist, 1978; Miller and Komar, 1980; Clifton and Dingler, 1984), the relationship is misleading. The problem is that both the normalised ripple wavelength, 2* (=2/do), and wave Reynolds' number, Nr, are functions of orbital diameter, do, so: J,*= 2~Tz 2(Nr)°'5

where Tz is the wave period. Hence, a plot of log(2*) against log(Nr) as shown by Boyd et al. (1988) will have gradient of -0.5. Further, Tz is also a weak function of do and if the data given by Boyd et al. (p. 458) are used to regress do on Tz, then T~ oc do°'39. This leads to: (~1.61 Nr = \ 2 * ]

(s- 1)gD

where T is the wave period, s is the relative density, D is the grain diameter and g is the acceleration due to gravity. Nielsen suggests use of the following equations for irregular field waves: 2 { 693-0.37 lnT0"~ a = exp ~,10~--~--5 1-~-0)

(5)

If there is no variation in 2 with Nr, B0yd et al.'s normalization procedure would lead to a power law relationship of: 2*ocNf o.62

Including the ripple wavelength in their data, Boyd et al. found. ,~* ~ N r -0.67

-q =210 1.85 a

(6)

Boyd et al. (1988) collected field measurements of ripple wavelength and near-bed flow conditions at a mean depth of 10.5 m over a sandy shoreface deposit over a period of 18 days. They found that ripple wavelengths normalized with respect to wave orbital diameter are most highly correlated with a wave Reynold's number (Nr) of the form: Nr-

Vmdo 2v

where do is the near-bed orbital diameter, and Um=ndo/Tis the maximum near-bed orbital velocity. Although this is consistent with several other

Amos et al. (1988) derived wave-current ripple heights for their ripple spacing measurements, obtained at 21 m depth on the Scotian Shelf, using an empirical relationship suggested by Allen (1970, p. 70), where. r/= 0.0742 xA9

(7)

which is applicable to wavelengths between 0.04 and 0.6 m. In Fig. 5, observed ripple heights are plotted against those predicted by Eqs. 1 and 3 from the G M G model, by Eq. 5 from Nielsen (1981), and b y Eq. 7 from Allen (1970). The G M G model tends to overpredict the smaller ripple heights and underpredict the larger ripple heights while the

215

BEDFORM DIMENSIONS ON A MACROTIDAL SHOREFACE

WHITSANDBAY 1991 RIPPLE HEIGHTS

~71

m

|

n

st

• i

Xm

m

•

I ALLEN

"/

l

I

i y r . l :I -

: ..[]

/ : t .

T

m

aim

OBSERVEDP~PLE~IGm (CM) Fig. 5. Observed ripple heights against those predicted by the GMG model, Nielsen (1981) and Allen (1970).

Nielsen model tends to underpredict both large and small ripple heights. Allen's relatively simple equation provides a somewhat better overall prediction of ripple height but observations of ripple wavelength are required to use the equation. In Fig. 6, observed ripple wavelengths are plot-

ted against those predicted by Eqs. 2 and 4 from the GMG model and by Eq. 6 from Nielsen (1981). The patterns are similar to those for ripple height predictions. The GMG model overpredicts the spacing of small-scale ripples by a factor of 4-6, but seems to perform somewhat better for the

WHITSANDBAY 19 91 RIPPLE WAVELENGTHS(cm) CARSTENS et al. [] NIELSEN

80~ 70

•

.~. 504

m m

iN

•

/

.........

! t --... °

ml m

o

io

"

~o

•

3'o

/o

5'o

go

qo

go

~o

loo

OBSERVEDRIPPLEWAVEIk~IGTI4(CM) Fig. 6. Observed ripple wavelengths against those predicted by the GMG model (Carstens et al., 1969) and Nielsen (1981).

216

larger ripples. The Neilsen model seems better able to predict small-scale ripple wavelengths than large-scale ripple wavelengths. However, the analysis by Vincent and Osborne (in press) resulted in no statistically significant correlation between wave parameters and ripple wavelengths determined from the side-scan ABS for small-scale ripples and a weak but significant cors'elation between Reynold's number, Shields parameter and mobility number for the large-scale ripples. Sherman and Greenwood (1984) have found that for constant orbital velocity and wave asymmetry, bed roughness and bedform changes in the surf zone are influenced by the presence of mean currents, particularly the profile of the longshore current velocity. Amos et al. (1988) have further suggested that bedform types designated as wave ripples, wave-current ripples and current ripples are well separated by the wave Reynold's number and the dimensionless mean flow; even better separation is achieved using the wave Shields parameter (after GMG) and the current Shields parameter (after Sternberg, 1972). The maximum wave Shields parameter and shear velocity for the steady current were computed using the GMG model using observed ripple dimensions as input to the model. Although mean currents were small relative to wave-induced flows during the 1991 experiment, Fig. 7 shows that small scale ripples and large scale 3D forms are separated from about half of the large-scale 2D ripples (lunate megaripples); the latter being associated with higher values of U,c (> 8 cm s-a) and the former occurring at smaller values of U,c. However, as a number of the large lunate-shaped and three-dimensional ripples also occur at small values of U,c these criteria do not seem to be appropriate for separation of ripple types in this environment. The conclusion to be reached from the analysis up to this point is that conventional formulae for predicting ripple geometry and ripple existence do not perform well in this macro-tidal nearshore environment. This is perhaps not surprising when one considers the timescales of bedform change with respect to rates of change in hydrodynamic conditions in this environment and also the inherent difficulty of characterising largely threedimensional, spatially variable forms with simple

P.D. OSBORNE AND C.E. VINCENT

two-dimensional parameters such as height and wavelength.

Temporal changes in bedforms during tidal cycles Time series of bed positions were reduced to one representative value of bed position for each two minute segment of a data run by de-spiking, temporal averaging and, where necessary, a qualitative assessment of time-series. A series of reduced bed positions typical of those observed under low energy swell illustrates the changes in bed elevation during a falling portion of a tidal cycle (Fig. 8c). Zero-crossing periods of the cross-shore flow, mean water depths, mean cross-shore and shoreparallel velocities and the significant range of crossshore flow (hereafter significant wave orbital speed) associated with these bed positions are shown in Fig. 8a and b. The significant range of the crossshore flow is calculated as the average range (maximum velocity to minimum velocity) of the largest 1/3 of the velocity ranges in a given timeseries. In general, significant wave orbital speeds were between 0.45 m s- a and 1.2 m s- 1 for waves with zero crossing periods between 4 and 7 s (the corresponding near-bed orbital excursion was between 20 and 100 cm) and suggests that most swell was strong enough to entrain sediment during the experiments. Also evident in this particular time-series, is the inverse relationship between the wave orbital speeds and the water depth, due to the increase in wave height as the waves shoal in shallowing water. Mean cross-shore flows were usually small, less than 0.10 m s -a, and directed offshore (negative) during both rising and falling portions of tidal cycles. Although the mean shoreparallel flows were also relatively small (0.05-0.15 m s -~) during swell, the direction of the flow shows a distinct reversal during the falling tide on August 12/92 (Fig. 8b) from westward (negative) flow to eastward (positive) flow. Such reversals were not uncommon during the experimental periods and were most likely associated with the phase of the tide but also may have been associated with nearshore circulation patterns induced by periodicity in the beach topography. The two lines in each bed position series indicate the uncertainty in measured bed position due to

217

B E D F O R M DIMENSIONS ON A MACROTIDAL SHOREFACE

A

Small 2D +

•

X

A

+

A

+

A

Small 3D A Large 2D

xA

%-

X

X

Large 3D

I

r7

I

Sm 2D/L8 2D

+ +

> <

0.01

'~

6

8

1'0

12

U* CURRENT Fig. 7. Bed states plotted rclativeto m a x i m u m wavc Shields parameter and current shear velocity,both derived using G - M - G and observed ripple geometry.

10

(a)

90-

54-

0.15

(b)

0.1 0.05 0 - 0.05

1=

I> I:~

--0.1 0

i

i

-55

o~. _ 57_~

" ~ ~_~ ~-61

'

-

~.~

~

-0.15

i

t

~

~

--6,5

,.~

_.:..__:.._=

~

Time (min.)August 12

I

(C)

I

Fig. 8. Time-scales of: (a) burst-averaged zero-crossing wave period; (b) burst-averaged significant wave orbital speed, water depth, cross-shore and alongshore flow; (c) reduced bed positions representative of a falling spring tide (August 12) under low energy swell.

218

P.D. OSBORNE AND C.E. VINCENT

sampling resolution ( + 1 bin) of the ABS. In general, the periods about high tide (depths of 1.5-2.0 m and significant wave orbital speeds of less than 0.75 m s- 1) are characterised by relatively small and gradual changes in bed position and by the presence of bedform type 1, 2, and 4, while the rapidly rising and falling periods (depths 0.5-1.5 m and wave orbital speeds greater than 0.75 m s-1) are characterized by larger and more rapid bed elevation changes (Fig. 9) and by bedform types 3, 5 and 6. In Fig. 9c, the first 30 minutes of the series show variations in bed position of no more than 2cm. Over the next 30 minutes, there is an accretion of 2.5 cm followed by erosion of 6 cm. In the final 15-20 minutes of the series there is a rapid accretion of 8 cm. These observations indicate that large and rapid bed elevation changes of more than 0.5 cm min- ~ occur during both rising and falling portions of a tidal cycle under relatively calm swell with little or no mean currents present. The largest changes in bed elevation are associated with smaller depths near the end of a falling tide or the beginning of a rising tide when wave-induced cross-shore flows are large and waves are either breaking or near breaking. Diver observations confirmed that these bed position changes were associated with both

bedform migration bed forms.

and

the

development of

Bed profiles and bedform migration Although the ABS transducers were mounted in fixed positions it is possible to make some useful reconstructions of the bed profile and of bedform migration rates from the position of the echo from the sea bed, the known configuration of the 3 transducers and from the detailed diver observations and measurements of the bed beneath the transducers. In Fig. 10 a time sequence of bed profiles has been reconstructed from ABS bed positions and SCUBA diver measurements taken over a 25 minute period during a falling tide. Divers observed small-scale short-crested bedforms with bifurcations (Type 2); vortices were being shed from ripple crests during both strokes of the wave. The bedforms had a maximum height of 3.5cm and wavelength of 15-18cm. The time sequence of reconstructed bed profiles in Fig. 10 suggests an onshore migration of the form at a rate of 1 cm min- ~. In this case, there was relatively little change in the overall height/spacing of the migrating form. Bedform migration rates for both experiments

0.6

'~ 0.5\ cJ l m l

~0.4m

l

mm mm

.~ 0.3-

m m

m mm m

.~ 0.2-

~ o.1-

m m

mm

I

l

l

Immm

•

mlmmm m

m

mm 0

o

i

m

m

minim mm

~ Water Depth (m)

Fig. 9. Variations in the time-averaged rate of vertical bed position change as a function of water depth under low energy swell.

BEDFORM DIMENSIONS ON A MACROTIDALSHOREFACE

mR

ABOVE -59.5 -

-59.0 -59.0

-60.0-

-59.5

-60.5 -

-61.0 ~BELOW

219

-60.0

-60.5 -61.0

Fig. 10. Time sequence of reconstructed bed profiles indicating onshore bedform migration.

have been interpreted from coherent but spatially separated ABS bed position measurements, and also from sidescan ABS profiles. The sidescan ABS results presented by Vincent and Osborne (in press) indicate a positive and significant correlation between bedform migration rates and first order wave parameters (e.g. Wave Reynold's number, mobility number, Shields parameter) for the small scale bedforms; whereas, a positive relationship, but no significant correlation was found for the larger scale forms. Onshore migration was by far the most common type of migration observed during both experiments; only a single case of offshore migration was observed during the 1991 experiment. Maximum rates of observed migration associated with small scale forms were on the order of 4-5 cm min-1 with an average rate of 1.3 cm min -1. Large-scale lunate-shaped mega-ripples had maximum migration rates of 3 cm m i n - 1 and average rates of 1.1 cm rain- 1. SCUBA divers noted migrating ripples were asymmetric in the direction of migration. The rates measured in this study are considerably higher than the maximum rate reported (0.18 cm min -1) by Boyd et al. (1988) from their measurements at 10 m depth. Dingier (1974) on the other hand, measured rates ranging from 0.5cm min -1 to 4.2cm min -1 for smallscale wave ripples under shoaling waves which are more consistent with these observations.

Although it was not possible to obtained detailed diver observations of the bed during periods of storm wave activity, snorkellers were able to make occasional observations and the ABS provided continuous bed position series. These observations and measurements confirmed that significant bedforms were present even under a well-developed surf zone and that these bedforms were migrating at a considerable rate. The bed position series shown in Fig. 11 are from a 40 minute period during the storm peak on August 22, 1992 when wave heights in the vicinity were on the order of 1.25 m and the surf zone width was approximately 100 m; wave Reynold's numbers at this time were 6.85-7.57 x 105 and wave Shields parameter was 0.99-1.06. The time series clearly indicate the presence of a migrating bedform with a height of at least 15 cm. The bedform was apparently migrating obliquely with respect to the ABS transducers and in the direction of the longshore current. Such a large bedform is likely to have had a wavelength of at least 50 cm and perhaps as large as 150 cm. The measurements suggest that the form was displaced horizontally by a full wavelength in approximately 25 minutes. If this were indeed the case, the horizontal migration rate would be between 2 and 6 cm min-1. Once again, these measurements imply that the application of conventional ripple existence criteria developed for

220

P.D.

OSBORNE

AND

C.E.

VINCENT

WB822M22

WB822M01

55 60

~ 6s

f

i

!

i

i

._o 50

§ 8o m 651

,

,

,

,

,

55 60 65 0

F2

I 5

10

15 20 Time (minutes)

25

I 30

I

I

35

40

Fig. 11. Unsmoothed time-series of the position of maximum backscattered pressure from the three vertically oriented ABS transducers during the peak of storm wave activity on August 22, 1992. wave motion alone may be inappropriate for the surf zone where breaking waves and currents are present.

Bedform development and spatial variability In addition to information on bedform migration, bed profile information was also useful for observing the rate and pattern of sequential ripple development. Figure 12 shows a time sequence of reconstructed bed profiles for August 12 when the tide was dropping (see also Fig. 8). The bed profiles for the first 40 minutes indicate that while the height of the bedforms increased by a factor of 2-3, the spacing between crests increased by a factor of 3 or 4 such that the overall steepness of the forms decreased. Lofquist (1978) has also noted the increases in height and length during the growth of sand ripples. The profiles during the last 25-30 minutes indicate shoreward migration of a large-scale bedform at a rate of approximately 2.0 cm min-1. At the beginning of this sequence, divers observed small-scale symmetric, shortcrested, two-dimensional vortex ripples (Type 2) with the crests oriented normal to shore, and a height of 2.5 cm and a spacing of 10 cm. By the end of the sequence, the dominant bedforms were

large asymmetric landward lunate megaripples (Type 6) with a maximum height of 8-10 cm and a spacing of 40-50 cm. This type of sequential development of ripples occurred during all observed tidal cycles in response to changing water depths whenever near bed velocities were large enough to move sediment. Generally, the sequential changes followed the conceptual model outlined by Clifton (1976) in which he identified a hierarchical sequence of bed types associated with increasing asymmetry under shoaling waves: linear ripples, irregular ripples, cross ripples, lunate megaripples, flat beds. However, under the prevailing spring tides, water depths changed rapidly and as a result, local hydrodynamic conditions also varied considerably over short periods of time. Observations in the 5 m square just seaward of the instrumentation revealed significant spatial variability in both the bedform type and individual dimensions of a particular bedform type (Fig. 13). During the rising tide on August 17, large-scale lunate mega-ripples (Type 5) were generated under spilling surf bores. The heights of these megaripples varied from 3 to 10 cm while wavelengths ranged from 20 to 50 cm. At first, the mega-ripples were interspersed with areas of flat bed (Fig. 13a). As water depths increased, the larger scale ripples

221

BEDFORM DIMENSIONS ON A MACROTIDAL SHOREFACE

ABOVE ~--I

-58.0 -59.0 -60.0 -61.0

-

-62.0

-

~BELOW

Fig. 12. Time sequence of reconstructed bed profiles indicating sequential development o f bedforms and onshore bedform migration during a falling tide (August 12, 1991) under low energy swell.

a) 1745 h

b) 1815h

,,4-,,

,,'I",

c) 1830h I I I I

:"

f---.

-4--d)

I!II I

f) 1135h I I

II!IIII ===t

I

i----.,-..~

e) 110Oh

I

==~=====

"--'x ,_,x

I i....

la

/~-~7

--

J

-"x

/'-'x

I I I

"-'X I J

191 -"

Fig. 13. Spatial and temporal variability o f bedforms in the 5 m 2 observation area: (a) 08-17 1745 h: rising tide near low water (Nr=4.5 x 105; ~b'=0.63; 0=219); (b) 08-17 1815 h: rising tide (Nr=5.3 x 105; $'=0.62; 0=222); (c) 08-17 1830 h: rising tide (N r = 2.8 × 105; ~,'= 0.37; 0 = 120); (d) 08-18 0800 h: rising tide near high water (N, = 1.6 x 10s; ~ ' = 0.18; 0 = 57); (e) 08-18 1100 h falling tide (Nr = 6.9 x 10s; ~ ' = 0.64; 0 = 246); (f) 08-18 1135 h falling tide near low water (N, = 7.8 x 10s; ~ ' = 0.77; 0 = 303).

222

generated in the surf zone became less active and smaller ripples developed on the stoss slopes of the larger forms and in the areas of flat bed (Fig. 13b and c). These smaller forms first appeared as Type 3 ripples (post-vortex or transitional) on the longer duration seaward strokes of the waves and then were visible during both wave strokes. Gradually the height of these small-scale ripples increased under declining flow conditions until the area was dominated by sinuous-crescentic smallscale vortex and transitional post-vortex ripples. A similar situation was observed on August 18 near high tide (Fig. 13c) when the area was dominated by sinuous small-scale short-crested vortex ripples (Type 4; height= 1-3 cm; wavelength= 8-10cm) and long-crested post-vortex ripples (Type 3; height < 1 cm; wavelength= 10-12 cm). In the latter portion of the falling tide on August 18th, several large-scale cross-ripples had formed which gradually became individual lunate megaripples separated by areas of flat bed (Fig. 13d and e). Temporal variation in the cross-shore flow was induced by both the presence of wave groups and changing water depths associated with tidal variations. A set of time-series (Fig. 14) and burstaveraged concentration profiles (Fig. 15) taken over a one hour period (1121-1221 h on August 19) illustrate the nature of these hydrodynamic changes and their combined effects on bedforms and suspension. At the beginning of the first series (L21; t = 0-350 s; 1121-1127 h), steep, small-scale vortex ripples (Type 2/4; q = 3-5 cm; 2 = 12-15 cm) were present on the bed. Towards the end of the first series (t=375 s) an individual large wave flattened the bedforms. The bedforms re-appeared a few waves later, but with reduced steepness (q = 1-3 cm; 2= 12-15 cm); these forms were present during the second time-series (L33; 1133-1150 h) in which there is a gradual increase in both the magnitude of the cross-shore flow and the asymmetry of the flow. Also note the contrast in suspension events between the first 180 s of the second timeseries (L33) relative to those in the first time-series (L21); the suspension events in the second series appear to be much weaker for similar wave conditions reflecting the reduced ripple steepness at this time. Again towards the end of the second series

P.D. OSBORNE AND C.E. VINCENT

(t = 985 s), a single large wave flattened the bed. A few waves later, lunate megaripples (Type 5) had formed. These developed in height and wavelength during the third time series (L53; 1153-1201 h) and were well developed (q=5-7cm; 2= 30-35cm) throughout the fourth series (M04; 1204-1221 h) with low amplitude post-vortex ripples (Type 3) superimposed. The concentration profile (Fig. 15) associated with the steep vortex ripples (L21) is characterized by an exponential decrease in concentration from 180mg 1-~ at 0.01 m to 20mg 1-1 at 0.1 m above the bed. Profiles for the larger, but lower steepness lunate megaripples and post vortex ripples (L53, M04) exhibit a very steep exponential region near the bed with concentrations decreasing from 1 g 1-1 to 50 mg 1-x in the lowermost 0.02 m. The profile for burst L33 appears to be transitional between the steeply exponential and moderately exponential profiles. Although the magnitude of the cross-shore flows increased over time, suspension decreased due to the nature of the bedforms present. Discussion

Large and rapid changes in bed elevation and bed geometry were observed during both rising and falling segments of spring tidal cycles under low energy swell conditions. These changes were associated with bedform development resulting from changes in the local hydrodynamic forcing as well as with bedform migration. These processes accounted for local vertical bed elevation changes of upto 10 cm during a series of tidal cycles in which no significant local net erosion or accretion took place. Even larger and more rapid bed elevation changes were observed under storm waves associated with the migration of large scale forms. Bedform migration rates of upto 5 cm min 1 were associated with small-scale two-dimensional bedforms (height = 0.5-3.0 cm; wavelength = 8-20 cm) and rates of upto 3 cm min- 1 with welldefined lunate mega-ripples (height = 3-8 cm; wavelength= 30-80 cm) under spilling and shoaling waves. Assuming a simple triangular ripple section and excluding sediment by-passing this represents a bed load transport rate of approxi-

BEDFORM DIMENSIONS ON A MACROTIDALSHOREFACE (a)

223

d

~o. do. do. &.

~b~ ~ d. -g •

1'oo. 1%. 1'.o. 1'6o. ~'so. ~oo. 2'2o. ~o. J6o. 2'8o. ~oo. ~2o. 3'4o. ~6o. ~8o. ~'o~,.

Time i n seconds

Time i n seconds

[ LubB]g133

~ /.

F2 '

i

,

- ~ . O1MH= ..

,,#,

F2 - q.O]MHz .

.i ....

r..

,-'.

,.i

":,:

,

'

",:'.

"

~d

'.,

i

s-~

I

~ .d.

~o. do. do. do.

((:1) --o

l'oo. 1%. l',o. l'~o. {~o. ~oo. ;~o. ~o. ~o. ~'~o. ~oo. ~o. 3~o. ~'~o. ~'~o. do~,.

Time in seconds

~' ~ [ I

0,33-1.0g/l

Time i n seconds

~b819r'r, OL~

F2 - q. OlMHz

o"

~o.

do.

do.

do.

F2 - q,OlMHz

l'oo. ~'~o. l',o. l'~o. ]'~o. ~oo. ~o. Ao. &o. ~o. ~oo. ~o. ~,o. ~o. ~'~o. ,'o~,.

Time i n seconds

T~me i n seconds

Fig. 14. Time-series of suspended sediment concentration and cross-shore velocity, between 1120 h and 1200 h on August 19, 1992: (a) L21 1121-1129 h (N~= 3.37 x 105; ~k'=0.30; 0 = 106); (b) L33 1133-1150 h (N~=4.19 x 10s; ~b'=0.38; 0 = 139); (c) L53 1153-1201 h (N~= 3.71 x 105; ¢ ' = 0 . 3 5 ; 0 = 125); (d) 1204-1221 h (N~= 5.40 x 10s; ¢ ' = 0 . 5 1 ; 0 = 192)•

Rverage profile for each freq BRSEXII Rug 1992 L21,L33,L53,MOU, c~

__

wbBl912l wb819153 wb819139

t

wb81 gmotl

c~

E d

_

°

ld 4 ......l G ' .......1G' .......16-f ......1 ~ ....... 10' Conc

(giLl

Fig. 15. Burst-averaged suspended sediment concentration profiles for the series shown in Fig. 16.

mately 0.035-0.125 g cm -1 s -1 in the case of the small ripples and 0.075-0.2 g cm -1 s -1 for the large ripples. Evidently, form migration can represent a significant contribution to the net sediment transport. Results suggest transport associated with form migration may be particularly important for low steepness ripples which are less effective at suspending sediment high into the flow. These findings have significant implications for measurement and interpretation of sediment transport rates in wave and current environments. As an example, Fig. 16 shows time-series of de-spiked and smoothed bed position, suspended sediment concentrations at 2 cm above an arbitrary fixed bed position (indicated by the dashed line in Fig. 16a), suspended sediment concentrations at 2 cm above the measured time-varying bed position, and cross-shore velocity. The uncompensated

224

P.D. OSBORNE A N D C.E. VINCENT

-58 /-1

~-60 ~-62

0

100

O0

1(10

200

I

L/--

300

400

5~

600

7~

8~

900

1000

300

400

5(10

600

7~

8dlO

900

1()00 1100

~ 2 3 ~

~0

11~

(e)

0

100

2~

-0.81 0

. 100

.

3~

460

. 300

. 400

S~

600

7~

am

90o

1()00 1100

800

900

1000

0.8

200

.

.

. . 500 600 Time (s}

. 700

.

1100

Fig. 16. Time-series of: (a) smoothed and de-spiked bed position; (b) suspended sediment concentration at 2 cm above fixed bed (dashed line in a); suspended sediment concentration at 2 cm above true bed position; (c) cross-shore velocity.

concentration series (Fig. 16b) is representative of the type of measurements which would be obtained by a fixed point sensor such as an optical backscatterance sensor or transmissometer. There is clearly a non-stationary trend in this series which is coherent with the rising bed position. The trend is absent from the series compensated for bed position change (Fig. 16c). Interpretation of sediment resuspension processes and estimates of sediment transport rates computed from point measurements of sediment concentration and velocity will be subject to potentially large errors without compensation for bed position changes relative to instrument positions when large and rapid bed elevation changes occur. In the example shown, the time-integrated cross products between velocity and concentration differed by 30% for the compen-

sated as opposed to uncompensated series. Furthermore, the presence of migrating large-scale bedforms under relatively gentle conditions suggests that spatial variability in sediment concentrations, near-bed velocity structure and sediment transport are likely to be significant. Bedform dimensions under these conditions are not well predicted using the equations of Carstens et al. together with the Grant-Madsen-Glenn model or by using Nielsen's model. Ripple heights were best estimated using a simple empirical equation suggested by Allen (1970), although use of this equation requires apriori knowledge of the ripple wavelength. It is not surprising that conventional approaches to predicting ripple geometry and ripple existence do not work well in this macro-tidal nearshore environment when one con-

BEDFORM DIMENSIONS ON A MACROTIDAL SHOREFACE

siders the factors involved in controlling the bedform regime. These include: (1) groupiness in the incident waves, (2) the complex hydrodynamic conditions under shoaling and breaking waves; (3) the super-position of quasi-steady currents; Furthermore, the depth controlled or tidally controlled changes in local hydrodynamic conditions occur very rapidly in this environment. Bedforms are therefore likely to be out of equilibrium with flow conditions for most of the tidal cycle rendering direct computations of bedform geometry from local flow and sediment parameters difficult. Finally, the parameters of height and wavelength which are traditionally used to define bedform geometry exhibit a large amount of spatial variability even within a relatively small area (5 m2), such that even individual bedforms are often not easily characterised by a simple height and wavelength. These results indicate a need for measurements to examine the three-dimensional nature of these nearshore bedforms.

Acknowledgements The Bedforms and Suspension EXperiments is a combined project involving the University of East-Anglia, Bullard Laboratories, Polytechnic of the SouthWest and Cardiff University and is sponsored by the Natural Environment Research Council. PDO also gratefully acknowledges the support of a NSERC, Canada postdoctoral fellowship. The authors wish to thank Drs. Mal Green, Mark Davidson, Paul Russell and Mr. Gareth Lloyd for their enthusiastic assistance while in the field. Particular thanks to Prof. David Huntley for making the B-Band instrumentation available and for arranging access to the beach. Thanks also to Tony Westbrook (Toronto) and Philip Judge (East Anglia) for their assistance with the preparation of figures.

References Allen, J.R.L., 1970. Physical Processes and Sedimentation. Unwin, London, 248 pp. Amos, C.L., Bowen, A.J., Huntley, D.A. and Lewis, C.F.M., 1988. Ripple generation under the combined influences of

225 waves and currents on the Canadian continental shelf. Cont. Shelf Res., 8(10): 1129-1153. Boyd, R., Forbes, D.L. and Heftier, D.E., 1988. Time-sequence observations of wave-formed sand ripples on an ocean shoreface. Sedimentology, 35: 449-464. Carstens, M.R., Neilson, R.M. and Altinbilek, H.D., 1969. Bedforms generated in the laboratory under oscillatory flow: analytical and experimental study. U.S. Army Corps Eng., Coastal Eng. Res. Cent., Tech. Memo, 28. Clifton, H.E., 1976. Wave-formed sedimentary structures--a conceptual model. In: R.A. Davis, Jr. and R.L. Ethington (Editors), Beach and Nearshore Sedimentation. SEPM Spec. Publ., 24: 126-148. Clifton, H.E. and Dingier, J.R., 1984. Wave-formed structures and paleo-environmental reconstruction. Mar. Geol., 60: 165-190. Davidson-Arnott, R.G.D. and Greenwood, B., 1976. Facies relationships on a barred coast, Kouchibouguac Bay, New Brunswick, Canada. In: R.A. Davis, Jr. and R.L. Ethington (Editors), Beach and Nearshore Sedimentation. SEPM Spec. Publ., 24: 149-168. Dingier, J.R., 1974. Wave-formed ripples in nearshore sands. Ph.D. Dissert., Univ. California, San Diego, 136 p. Dingier, J.R., Boylls, J.C. and Lowe, R.L., 1977. A high frequency sonar for profiling small-scale subaqueous bedforms. Mar. Geol., 24: 279-288. Glenn, S.M. and Grant, W.D., 1983. A suspended sediment stratification correction for combined wave and current flow. J. Geophys. Res., 92: 8244-8264. Grant, W.D. and Madsen, O.S., 1979. Combined wave and current interaction with a rough bottom. J. Geophys. Res., 84: 1797-1808. Grant, W.D. and Madsen, O.S., 1982. Moveable bed roughness in unsteady oscillatory flow. J. Geophys. Res., 87:469-481. Green, M.O. and Boon, J.D., 1988. Response characteristics of a short-range high-resolution digital sonar altimeter. Mar. Geol., 91: 197-203. Greenwood, B., Dingier, J.R., Sherman, D.J., Anima, R.J. and Bauer, B., 1985. Monitoring bedforms underwater using high resolution remote tracking sonars (HRRTS). Proc. Can. Coastal Conf. (St. Johns, Nfld.) pp. 143-158. Greenwood, B., Osborne, P.D., Bowen, A.J., Hazen, D.G. and Hay, A.E., 1990. Nearshore sediment flux and bottom boundary dynamics: the Canadian Coastal Sediment Transport Programme (C-COAST). Proc. 22nd Int. Conf. Coastal Eng. (Delft, Netherlands.) ASCE, pp. 2227-2240. Hanes, D.M., Vincent, C.E., Huntley, D.A. and Clarke, T.L., 1988. Acoustic measurements of suspended sand concentration in a wave-dominated nearshore environment. Cont. Shelf Res., 6(4): 585-596. Inman, D.L., 1957. Wave generated ripples in nearshore sands. Beach Erosion Bd., U.S. Army Corps Eng., Tech. Memo, 100. Lofquist, K.E.B., 1978. Sand ripple growth in an oscillatoryflow water tunnel. C.E.R.C., Washington, D.C., Tech. Pap., 78-5. Miller, M.C. and Komar, P.D., 1980. A field investigation of the relationship between oscillation ripple spacing and the near-bottom water orbital motions. J. Sediment. Petrol., 50: 183-191.

226 Mogridge, G.R. and Kamphuis, J.W., 1972. Experiments on bed-form generation by wave action. Proc. 13th Coastal Eng. Conf., A.S.C.E., Vancouver, pp. 1123-1142. Nielsen, P., 1981. Dynamics and geometry of wave-generated ripples. J. Geophys. Res., 86(C7): 6467-6472. Sherman, D.J. and Greenwood, B., 1984. Boundary roughness and bedforms in the surf zone. Mar. Geol., 60: 199-218. Sternberg, R.W., 1972. Predicting initial motion and bedload transport of sediment particles in the shallow marine environment. In: D.J.P. Swift, D.B. Duane and O.H. Pilkey (Editors), Shelf Sediment Transport, Process and Pattern. Dowden, Hutchinson and Ross, Stroudsburg, Pa., pp. 61-83. Vincent, C.E. and Downing, A.J., in press. Variability of suspended sand concentrations, transport and eddy diffusiv-

P.D. OSBORNE A N D C.E. VINCENT

ity under non-breaking waves on the shoreface. Cont. Shelf Res. Vincent, C.E. and Green, M.O., 1990. Field measurements of the suspended sand concentration profiles and fluxes, and of the resuspension coefficient over a rippled bed. J. Geophys. Res., 95: 15591-15601. Vincent, C.E. and Osborne, in press. Bedform dimensions and migration rates under shoaling and breaking waves. Cont. Shelf Res. Vincent, C.E., Hanes, D.M. and Bowen, A.J., 1991. Acoustic measurements of suspended sand on the shoreface and the control of concentration by bed roughness. Mar. Geol., 96: 1-18.