ISSN 00014370, Oceanology, 2014, Vol. 54, No. 2, pp. 205211. Pleiades Publishing, Inc., 2014.

Original Russian Text I.O. Leontyev, 2014, published in Okeanologiya, 2014, Vol. 54, No. 2, pp. 226232.

MARINE
GEOLOGY

Calculation of Longshore Sediment Transport

I. O. Leontyev
Shirshov Institute of Oceanology, Russian Academy of Sciences, Moscow, Russia
email: igor.leontiev@gmail.com
Received March 27, 2012; in final form, February 21, 2013

AbstractCalculation approaches to longshore transport of sandy sediments are discussed. The estimation
of the total sediment transport rate is shown to be possibly based on the socalled CERC formula, where the
proportionality factor K should be calculated from relationships of Bayram et al. [8] or Leontyev [4]. In both
cases, the results are very close to each other if the authors determination of the wave breaking depth is used.
Under the condition of contrasting variations in the sediment grain size over the coastal profile or in the case
of fragmentary sand distribution on the surface of the bed, the local approach implying processbased mod
eling is more effective. A model is suggested to compute the local longshore sediment transport rates.
DOI: 10.1134/S0001437014020131

INTRODUCTION portionality, and the subscript B belongs to the wave

Longshore sediment transport created by waves
and currents is a major factor of shore development,
and the problem of its calculation has long attracted
the attention of researchers and coastal engineers. The
purpose of the calculation is to assess the longshore
transport rate of the beachforming material in various
wave situations. Traditionally, there exist two
approaches to this problem: integral and local. The
former largely uses empirical regularities, and its result
is the assessment of the total (integral) transport in the
coastal zone. The latter approach is based on modeling
the processes on which the longshore transport
depends, leading to the determination of the local
transport rate on the shoreface profile, which allows us
to take into account the bottom relief features and sed
iment distribution. However, each approach has its
own advantages and limitations and is chosen to
match the problem under study.
This paper discusses the calculation methods of
both integral and local characteristics of sand trans
port. The purpose of this discussion is the choice of the
most acceptable calculation relationships.
INTEGRAL APPROACH
Key formula. The backbone of this approach is the
idea that the total longshore sediment transport rate
Q y is related to the longshore component of the energy
flux Fy, which is generated by waves that approach the
shore at an angle:
Q y = KFyB ,
where is the coefficient that correlates the dimen
sions Q y and Fy, K is the dimensionless factor of pro
break point. The FyB value is determined as follows:
FyB = ( EC g cos sin ) B , E B = 1 gH B ,
2
8 (2)
C gB = ghB , H B = B hB ,
where E is the energy, C g is the group velocity, is the
angle between the wave direction and the normal to
the shore, is the water density, g is the acceleration of
gravity, H is the wave height, h is depth, and B = 0.8 is
the breaking index. The H value may imply the mean
( H ), rootmeansquare ( H rms ), or significant ( H s )
wave heights, which, according to the Rayleigh distri
bution, are interrelated by the ratios
H s = 2H rms = 2 2 / H . (3)
If Qy expressed in m3/h, then =
is
3600 [ g ( g )(1 )], where g and are the density
and porosity of the sediments.
Relationship (1) was initially established empiri
cally (for example, [2]), but then it was theoretically
substantiated [13]. In recent decades, it has been
known as the Coastal Engineering Research Center
(CERC) formula.
Proportionality factor. For practical applications
(1), it is necessary to determine the factor K. The ini
tial version treated K as constant [18] and equal to 0.77
or 0.39 depending on the wave height used, H rms or
H s . According to the assessment in [17], the above val
ues are overestimated by about a factor of two. In a
(1) later version of formula [10], K depends on the size of
the sediment particles d g , which increases as d g .
decreases. At the same time, there is an opinion [12]
that the particle size should not noticeably influence

205
206 LEONTYEV
the transport value since, as d g decreases, the rough
ness of the bed decreases together with the drag force,
which causes particle transport.
Relatively recently, Bayram et al. [8] proposed their
model for the calculation of the longshore transport
rate, which has ultimately transformed into the CERC
formula. Having calibrated the formula on the basis of
a new data array, the researchers concluded that, first,
the previous K value was largely overestimated and,
second, the K value depends on the Dean parameter,
S D = H sB ( w gT p ) , where w g is the particle sedimenta
tion rate (fall velocity), and T p is the wave spectrum
peak period. The relationship derived can be written as
follows:
H A few iterations are enough to obtain stable values of
K s = 0.0256 0.9 + 0.4 sB , (4)
w gT p
where the s subscript in K means that the calculation
by formula (1) is based on a significant wave height. An
increased Dean parameter means an increased con
centration of suspended sediments. Bayram et al. have
shown that, if we use (4), the transport rate calculated
according to the CERC formula falls into an interval
of double deflections against the measured values in
62% of all cases and that this is the best result com
pared to the other formulas tested.
Leontyev [3, 4], applying his model for local sedi
ment transport rates, obtained a relationship for total
transport, similar to (1), where the K factor with the
values of included constants can be represented as
ghB
K rms = 0.04 0.8 + 0.02 . (5)
wg
The rms subscript in K means that calculations
according to formula (1) use the rootmeansquare
wave heights. Although relationship (5) does not con
tain a wave period, it is somewhat similar to (4). Of
interest is the possibility to compare transport calcula
tions by either relationship.
Wavebreaking depth. First, we should consider the
definition of the height of breaking hB . Unfortunately,
many investigations do not show the way this impor
tant parameter is assessed, which hinders the compar
ison of the results. For the purposes of unification, we
may propose to assess hB depending on the wave
parameters at deep water, which usually serve as input
parameters for practical calculations. Assuming that
the shoreward energy flux from the open sea is con
stant until the beginning of the wave breaking,
(EC g cos ) = (EC g cos ) B ,
(6)
C g = (g 4)T p
(the subscript refers to deep water) we have, taking
into account relationships (2), the following:
25 25
1 5 cos
hB = 1 2 H 4/5 ( )
gT p2
cos B
, (7)
4 B
where the wave approach angle B is determined from
Snells refraction law,
sin B C B = sin C ,
(8)
C B = ghB , C = ( g 2) T p,
where C is the wave phase speed. Initially, the hB value
is calculated without regard to changes in the
angles. At the next step, we correlate hB with regard to
the B value, after which we also correct the B angle.
hB and B .
Note that, in the model of Bayram et al., H = H s
and H sB = B hsB , while, in the model of Leontyev, hB
corresponds to breaking of waves of 1% exceedance;
i.e., H = H1%, and, taking into account the Rayleigh
distribution, H 1% = 2.15H rms and H rmsB =
( B 2.15)h1%B . According to (7), at low angles, we
have h1%B hsB = (H 1% H s )4 5 = 1.40.
Calculations vs. measurements. Since our purpose
is not to verify formula (1), which has already been
done by Bayram et al., but to compare the calculations
according to (1) with the inclusion of relationships (4)
and (5), we may limit ourselves to a relatively small
amount of data, which, however, cover a fairly broad
range of Q y . values.
In this case, we use three arrays of field data given
in Table 1. The measurements of Voitsekhovich [1] and
Leontyev [3] were made in the Black Sea (respec
tively, on the Ukrainian and Bulgarian shores), and the
data of Miller [15] relate to the Atlantic coast of the
United States (Duck, NC). The initial wave parame
ters were converted to the deepwater mean parame
ters. As we see from Table 1, the range of transport
changes includes three orders of values.
The results of the comparison are shown in Fig. 1.
The inclined lines in the graphs correspond to the
coincidence of the calculated and measures values. It
is obvious that the model calculations of both Bayram
and Leontev agree well with the measured data. In
addition, graphs (a) and (b) almost repeat each other;
i.e., both models yield very close results. The differ
ence in the calculated values in the majority of cases
does not exceed 10% (Table 1). This result seems
amazing since the model precursors and the data
arrays used for their calibration are not at all related to
each other.
The inclusion of the wave period into relationship (4)
does not appear to have a decisive effect on the results.
The growth of the period, according to (4), should
OCEANOLOGY Vol. 54 No. 2 2014
 CALCULATION OF LONGSHORE SEDIMENT TRANSPORT 207

cause a decrease in K. However, the breaking depth hB Table 1. Measured data and the longshore sediment trans
increases simultaneously, and this tendency appears to port rates calculated using the models of Bayram et al. and
be dominant, causing an increase in transport. Leont'yev
Taking into account the consistency of the calcu Sediment transport rate, m3/h
lated values, as well as the fact that the model of H, T, , dg, mea
Bayram et al. has already undergone overall verifica m s deg mm Bayram's Leont'yev
sure
tion, we may recommend the practical application of estimate estimate
ments
formula (1) with the K factor calculated according to
relationship (4) or (5) and with the determination of Voitsekhovich [1]
the height of breaking hB according to relationship (7). 0.4 4.0 10 0.25 6.7 3.0 3.1
0.4 3.8 15 0.25 6.7 4.4 4.4
0.5 5.7 10 0.32 5.9 4.5 5.1
LOCAL APPROACH
0.4 3.8 10 0.36 3.0 2.4 2.3
Local sediment transport modeling implies the cal 1.0 6.0 30 0.43 101 66.8 66.1
culation of the wave parameters at each point of the 1.0 5.4 18 0.34 85.1 55.8 52.2
shore profile, the determination of the longshore cur 0.6 4.8 10 0.25 22.2 9.6 9.7
rent profile, and the assessment of the amount of solid
particles raised from the bed by waves and currents. To 1.1 5.9 25 0.36 76.6 89.0 85.8
calculate the transport rate, various relationships are 0.6 5.8 25 0.36 16.6 15.4 16.8
used, the most popular among which are the formulas 0.5 5.9 18 0.32 15.2 7.6 8.8
by Bijker, EngelundHansen, Watanabe, Ackers 0.5 4.2 20 0.28 21.1 9.7 9.5
White, Van Rijn, and Bailard. They are considered and 0.6 3.8 10 0.30 17.0 8.6 7.6
compared in [7] on the basis of the data measured from 0.8 5.0 5 0.28 8.9 10.3 9.8
the Duck Research Pier in NC. It turned out that some 0.6 4.2 5 0.40 3.7 3.6 3.3
models behave better at weak wave activity but operate
badly in other conditions. The Bijker and Watanabe 0.8 6.2 5 0.37 4.4 8.0 8.4
formulas systematically make overassessments, and 1.0 5.8 20 0.28 121 69.7 68.7
the least deflections in general are typical of the Van 0.8 5.3 30 0.28 67.3 47.8 47.9
Rijn model. 0.5 4.1 5 0.29 4.4 2.6 2.5
Of interest is a similar comparison for Leontyevs 0.6 5.4 10 0.32 10.4 8.5 8.9
model [4], whose updated version includes a number 0.9 4.5 30 0.25 176 78.6 69.6
of additional mechanisms. 0.5 4.3 35 0.70 11.8 8.5 8.1
Key calculated relationships. The model is based on 0.6 5.9 15 0.48 7.0 9.7 10.3
the energetics concept [5, 6, 9], and the local longshore 0.5 3.6 15 0.50 4.1 5.5 4.9
sediment transport rate, expressed in m3 m1 h1, is 0.4 4.3 10 0.47 1.5 2.0 2.0
determined as
0.5 3.8 10 0.28 4.4 5.3 4.9
0.9 4.7 10 0.32 42.9 26.0 23.1
q y = 9 b D f + s m ( 4D f + B )
u
1.2 5.7 20 0.23 186 147.3 137.2
8 tan wg (9) 0.4 3.3 20 0.24 5.2 8.0 7.1
V + Vc 0.6 4.2 30 0.22 34.8 26.9 25.8
w + q BS sin ,
um 0.9 4.7 25 0.23 95.8 76.3 69.0
where is the angle of the natural slope (for sand, Leont'yev[3]
tan 0.6). The value D f is the energy dissipation rate 0.5 4.2 30 0.4 7.8 10.0 9.7
owing to the bed friction: 0.7 4.8 30 0.3 48.8 31.2 30.5
0.6 4.2 30 0.4 13.7 16.9 15.8

3
( )
D f = 2 f wum3 , f w = exp 5.5(r am )0.2 6.3 , (10) 1.1 4.8 30 0.4 68.4 96.6 83.3
1.4 6.1 10 0.4 45.1 77.5 70.3
where f w is the wave friction coefficient; um is the 1.2 6.9 10 0.4 48.8 48.0 48.4
amplitude of the orbital velocity at the bed; 1.4 6.0 10 0.3 129 96.8 87.6
H Miller [15]
um = 1 rms , (11) 1.9 6.3 20 0.165 530 906 778
2 sinh kh
2.8 5.8 13 '' 1780 2180 1518
am = um is the amplitude of the water particle oscil 1.8 5.6 33 '' 560 1080 886
lations along the bed; = 2 T p is the frequency; 1.7 5.8 35 '' 670 906 782

OCEANOLOGY Vol. 54 No. 2 2014

208 LEONTYEV

(a) (b)
1000 1000
Calculated transport rate, m3/h

Calculated transport rate, m3/h

100 100

10 10 1
2
3

1 10 100 1000 1 10 100 1000

Measured transport rate, m3/h Measured transport rate, m3/h

Fig. 1. Comparison of the calculations made using (a) the Bayram et al. relationships and (b) Leontyevs relationships with the
measured data: (1) Voitsekhovich [1], (2) Leontyev [3], and (3) Miller [15].

k = 2 L is the wave number; L is the wavelength; r is The dissipation rate D is determined by the follow
the roughness parameter; ing relationships:

r = 85 2.5 0.05d g + 4 ,
2
  H 4  H rms
D = 2 P E , P = rms , = 0.5 + . (15)
f w2.5um
2 (12) Tp h L
2.5 = ,
2 ( g ) gd g
2
The coefficient equals 0.1 or 0.5 for depths h 2hB
2.5 is the Shields parameter based on the conven or h < 2hB .
tional roughness, r2.5 = 2.5d g , and are the height The local wave heights H rms are calculated with a rela
and length of the bed ripples. The values f w , r, , and tionship derived from the energy balance equation [4]:
are calculated on the basis of the Nielsen relation
ships [16]. H rms =
1 2
The transport efficiency parameters b and s take x

12

the form [4] = C g cos 44

dx + 1 , (16)
T p h 4 ( C g cos ) 3 F02
0
b = 0.1 2.5 0.05, s = 0.02 2.5 0.05,
F0 = H rms0C g 0 cos 0,
2
(13)
0.05 < 2.5 < 1, b = 0.1, s = 0.02, 2.5 1.
where the 0 subscript refers to the selected initial point
The initial sediment movement meets the condition on coastal profile, and C g is the group velocity:
2.5 > 0.05.
The B value takes into account the additional
energy in the bottom layer by the penetration of turbu
(
2 sinh 2kh
L
Tp )
C g = C 1 + 2kh , C = tanh(kh), (17)

lence from the surface layer during wave breaking: C is the phase speed calculated by Eckarts approxima
tion [11].
B = D exp 3 (1 P 1) , =
(hB l B ) The additional transport in the breaking zone q BS is
, (14)
H s L due to discharge of suspended solid material to the
upper part of the water column in the segment
where D is the energy dissipation rate during wave
x B x x B + l BS :
breaking, P is the fraction of breaking waves in the
spectrum, l B is the length of the surf zone, and s x xB
q BS = c1(x) Dum, (x) = , l BS = c2hB , (18)
L = ( g 2)T p2 is the wave length at deep water. wg l BS

OCEANOLOGY Vol. 54 No. 2 2014

 CALCULATION OF LONGSHORE SEDIMENT TRANSPORT
where l BS is the length of the effective area of q BS , the
point x B corresponds to the breaking depth hB , c1 =
0.05, c2 = 8. The OX axis is normally directed to the
shore.
The mass transport flow velocities in the bottom
layer, V w and Vc , are due to waves and current respec
tively. The velocity V w is expressed as
2 3
V w = d m sin , d = D* D , D* =
u EC g
, (19)
C D* X 5
where D* is the mean gradient of the energy flux in the
distance X between a given point and the shoreline. In
the wave shoaling zone, D 0, d 1, V w > 0 (shore
ward velocity), and expression (19) passes to the
known LonguetHiggins formula [14]. In the surf
zone, D increases, d 0, and V w 0.
The depthaverage velocity of the longshore current V
is determined from the equation of longshore momen
tum balance, which, in a stationary onedimensional
case ( t = 0, y = 0 ), has the following expression:
1 d S + R T
( yx yx ) wy
dx
(20)
+ y d l
dx (d(Vh)
dx )
= 0,
where S yx is the waveinduced radiation stress, R yx is
the stress predetermined by rollers on the crests of
breaking waves, Twy is the drag force of the wind, y is
the bed friction stress, and l is the coefficient of the
eddy viscosity [4]. The transfer to the nearbed current
velocity Vc is made with the following relationship:
Vc = V ( ln(h z a ) 1) , z a = e 1w,
(21)
w = f w 2am,
where z a is the apparent bed roughness for the current
superimposed on a wave oscillatory flow and w is the
thickness of the wave boundary layer.
Calculations vs. measurements. For comparison,
we use the data of three series of field measurements
near Duck, NC. One series was made in 1996 and was
characterized by Miller [15], as was mentioned above.
The other two series of measurements were conducted
during the DUCK85 and SANDYDUCK experi
ments in 1985 and 19971998; their results were dis
cussed in [17]. Unfortunately, the wave approach
angles were not given there. Therefore, we had to cal
culate the values and H by selecting the known
wave heights and periods at reference depths (about
2 m under weak wave conditions and 78 m during sea
storms). The size of the sediments and wave parame
ters at deep water for the respective measurement
series are given in Table 2. As we can see, weak waves,
moderate storm, and heavy storm situations are repre
sented.
OCEANOLOGY Vol. 54 No. 2 2014
209
Table 2. Calculated mean wave parameters in deep water
during field experiments near Duck, NC (according to the
data in [7, 15])
Measurement d g, H, T, ,
No.
date mm m s deg
1 06.09.1985 0.18 0.35 10.7 30
2 06.09.1985 0.30 10.9 30
27.03.1996 0.165 1.8 5.6 33
4 02.04.1996 1.7 5.8 35
20.10.1997 0.18 2.8 10.7 60
6 04.02.1998 3.6 10.7 60
The comparison of the calculated and observed
profiles of the local sediment transports is shown in
Fig. 2. As the wave heights grow, the measured trans
port rates increase from 101 to 101 m3 m1 h1. In low
wave conditions (Fig. 2a), transport is observed mainly
in the wave breaking zone. During a moderate storm
(Fig. 2b), the sediments drift within the whole range,
but the maximum is observed near the shore, while,
during a heavy storm (Fig. 2c), the main flow of the
material concentrates in the outer part of the range.
The calculations yield values of the same orders; in
addition, the predicted transport distributions repro
duce mainly the observed tendencies.
The deviations of the calculated values from the
measured ones depend on a number of reasons. Thus,
the calculations were based on relatively conventional
bed profiles and were conducted without taking into
account the tidal fluctuations of the sea level (with
amplitude up to 1 m), while even small changes in
depth significantly affect the values of the sediment
transports. Another cause is probably the nonunifor
mity of the bed relief along the shore, which causes the
development of gradient currents that increase the
transport at some segments and decrease it at other
segments of the shoreface. Finally, the third cause of
deviations is the heterogeneity of the size of the mate
rial on the bed.
Variable sizes of sediments. The size of the sand par
ticles in the above model can be set constant (as in the
cases considered above) or variable along the profile.
In the latter case, the estimated transports can be
updated significantly, which is illustrated in Fig. 3. A
shore section in the eastern Gulf of Finland serves as
the prototype here, where the sand material (d g =
0.250.5 mm) is deposited as separate bars between
which the bed is covered with boulders (about 10 cm in
size). The calculations reflect the conditions of a typi
cal WSW storm ( H = 1 m, T = 5 s, = 45). For
comparison, Fig. 3 also shows the results obtained
under the assumption of uninterrupted sand layer
(d g = 0.3 mm). As we can see, if we take into account
the real situation, the total transport rate turns out to
be several times smaller. This is also partially related to
the decrease in the current velocities due to the
210 LEONTYEV

(a) 1.0
0.3
3 1 1

0.8

V, m/s
Elevation, m qy, m m h qy, m m h

1 0.6
0.2 0.4
0.2
0.1 0

3 1 1
0 1.0

Elevation, m dg, mm qy, m m h

3 1 1

0.3 0.8
2 0.6
0.2 0.4
0.2
0.1 0
0.6
0 0.4
0.5 0.2
1.5
0
2.5 4
120 140 160 180 2
Distance, m 0
2
6 (b) 4
3 1 1
Elevation, m qy, m m h qy, m m h

3 0 40 80 120 160 200 240 280 320 360 400

4 1 Distance, m
2
2
Fig. 3. Calculated profiles of the longshore sediment trans
0 port rates and current velocities for (1) the actual distribu
3 1 1

8 tion of the material on the bed and (2) for the continuous
6 4 sand layer. The wave parameters are as follows: H = 1 m,
4 T = 5 s, = 45.
2
0 increased bed friction at the section with boulder
0 armoring.
4
8
0 50 100 150 200 250 300 350 CONCLUSIONS
Distance, m
The longshore sediment transport is described in
(c) terms of the local and integral characteristics. From
3 1 1

8
Elevation, m qy, m m h qy, m m h

the practical point of view, of major interest is the total

6 5 sand transport along a given shore (Q y ). It has long
4
been known that the Q y value is directly proportionate
2
to the longitudinal component of the energy flow Fy ,
0 which is easily calculated using preset wave parame
3 1 1

16
ters. The lack of data has long hindered the unambig
12 6 uous determination of the proportionality factor K;
8 however, to date, acceptable solutions to this problem
4 have been suggested. This allows us to recommend
0 formula (1) for calculations, where K is calculated
0 using relationships (4) or (5) and the depth of the wave
breaking is determined using ratio (7).
4
8 The recommended method of assessing Q y is appli
100 200 300 400 500 cable in the presence of a continuous sand layer on the
Distance, m shoreface. In the conditions where the size of the sed
1 iments changes significantly along the profile or where
2 sand occurs as separate spots intermingled with a
Fig. 2. Comparison of the calculated profiles of the long boulder and rock bed, the local approach is more effi
shore sediment transport rates (2) with the measured data
(1) near Duck, NC [7, 15] in conditions of (a) low waves, cient, implying the simulation of elementary processes
(b) a moderate storm, and (c) a heavy storm. The graph that participate in the sediment transport, including
numbers correspond to the numbers in Table 2. wave transformation and longshore current develop

OCEANOLOGY Vol. 54 No. 2 2014

 CALCULATION OF LONGSHORE SEDIMENT TRANSPORT
ment. On the basis of the findings of the conducted
verification, we can recommend the updated version
of the authors model to assess local longshore sedi
ment transport rates.
ACKNOWLEDGMENTS
This work was supported by the Russian Ministry of
Education and Science (The Development of Auto
mated Information System Technology for Monitor
ing Marine Coastal Zones and Inner Water Bodies of
the Russian Federation, project no. 16.515.11.5075)
and by the European Commission through program
FP7.20091 (project no. 244104, THESEUS (Innova
tive Technologies for Safer European Coasts in a
Changing Climate)).
REFERENCES
1. O. V. Voitsekhovich, Transportation of drifts along the
shore: general principles and real data, Vodn. Resur.,
No. 5, 108116 (1986).
2. C. A. M. King, Beaches and Coasts (Edward Arnold,
London, 1961).
3. I. O. Leontyev, Dynamics of Surf Zone (Shirshov Insti
tute of oceanology, USSR Science Academy, Moscow,
1989) [in Russian].
4. I. O. Leontyev, Coastal Dynamics: Waves, Currents, and
Sediment Transport (GEOS, Moscow, 2001) [in Rus
sian].
5. R. A. Bagnold, Mechanics of marine sedimentation,
in The Sea (Wiley, New York, 1963), Vol. 3, pp. 507
528.
6. J. A. Bailard, An energetics total load sediment trans
port model for a plane sloping beach, J. Geophys.
Res., C: Oceans Atmos. 86 (11), 1093810954 (1981).
OCEANOLOGY Vol. 54 No. 2 2014
211
7. A. Bayram, M. Larson, H. C. Miller, and N. C. Kraus,
Crossshore distribution of the sediment transport:
comparison between predictive formulas and field mea
surements, Coastal Eng. 44, 7999 (2001).
8. A. Bayram, M. Larson, and H. Hanson, A new for
mula for the total longshore sediment transport rate,
Coastal Eng. 54, 700710 (2007).
9. A. J. Bowen, Simple models of nearshore sedimentation;
beach profiles and longshore bars, in Coastline of Canada
(Geol. Surv., Halifax, Canada, 1980), pp. 111.
10. Coastal Engineering Manual (CEM), EM 111021100,
Part 3, Ch. 2: Longshore Sediment Transport, 2002,
pp. III21III4111.
11. C. Eckart, The propagation of gravity waves from deep
to shallow water, in National Bureau of Standards, Cir
cular 521 (Washington, DC, 1952), pp. 165173.
12. K. A. Haas and D. M. Hanes, Processbased modeling
of total longshore sediment transport, J. Coastal Res.
20 (3), 853861 (2004).
13. P. D. Komar and D. L. Inman, Longshore sand trans
port on beaches, J. Geophys. Res. 75 (30), 59145927
(1970).
14. M. S. LonguetHiggins, Mass transport in water
waves, Phil. Trans. R. Soc., A 245 (903), 535581
(1953).
15. H. C. Miller, Field measurements of longshore sedi
ment transport during storms, Coastal Eng. 36, 301
321 (1999).
16. P. Nielsen, Coastal bottom boundary layers and sedi
ment transport, in Advanced Series on Ocean Engi
neering (World Scientific, 1992), Vol. 4.
17. J. S. Schoonees and A. K. Theron, Review of the field
data base for longshore sediment transport, Coastal
Eng. 36, 301321 (1993).
18. Shore Protection Manual (SPM), Coastal Engineering
Research Center, US Army Corps of Engineers (US Gov.
Print. Off., Washington DC, 1984).
Translated by B. Alekseev