Front Cover
 Title Page
 Table of Contents
 Shear stresses acting in the...
 Development of cross-shore transport...
 Calibration of cross model with...
 Evaluation of cross with laboratory...
 Application of cross for November...
 Summary, conclusions and recom...
 Biographical sketch

Group Title: Technical report – University of Florida. Coastal and Oceanographic Engineering Program ; 112
Title: Improved cross-shore sediment transport relationships and models
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00075486/00001
 Material Information
Title: Improved cross-shore sediment transport relationships and models
Series Title: UFLCOEL-TR
Physical Description: vii, 213 p. : ill. ; 28 cm.
Language: English
Creator: Zheng, Jie, 1963-
University of Florida -- Coastal and Oceanographic Engineering Dept
Publisher: Coastal & Oceanographic Engineering Dept., University of Florida
Place of Publication: Gainesville Fla
Publication Date: 1996
Subject: Beach erosion -- Mathematical models   ( lcsh )
Sediment transport -- Mathematical models   ( lcsh )
Coast changes -- Mathematical models   ( lcsh )
Coastal and Oceanographic Engineering thesis, Ph. D
Dissertations, Academic -- Coastal and Oceanographic Engineering -- UF
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis (Ph. D.)--University of Florida, 1996.
Bibliography: Includes bibliographical references (p. 206-212).
Statement of Responsibility: by Jie Zheng.
Funding: This publication is being made available as part of the report series written by the faculty, staff, and students of the Coastal and Oceanographic Program of the Department of Civil and Coastal Engineering.
 Record Information
Bibliographic ID: UF00075486
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved, Board of Trustees of the University of Florida
Resource Identifier: oclc - 35916017

Table of Contents
    Front Cover
        Front Cover
    Title Page
        Title Page
    Table of Contents
        Table of Contents 1
        Table of Contents 2
        Table of Contents 3
        Abstract 1
        Abstract 2
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
    Shear stresses acting in the nearshore
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
    Development of cross-shore transport model - cross
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
    Calibration of cross model with laboratory experiments
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
    Evaluation of cross with laboratory experiments
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
    Application of cross for November 1991 and January 1992 storm erosion at Ocean City, Maryland
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
        Page 192
        Page 193
        Page 194
        Page 195
        Page 196
        Page 197
        Page 198
        Page 199
    Summary, conclusions and recommendations
        Page 200
        Page 201
        Page 202
        Page 203
        Page 204
        Page 205
        Page 206
        Page 207
        Page 208
        Page 209
        Page 210
        Page 211
        Page 212
    Biographical sketch
        Page 213
Full Text




Jie Zheng










I would like to express my deepest gratitude and appreciation to my advisor and the

chairman of my supervisory committee Professor Robert G. Dean, for his constructive

direction, enthusiasm, advice and unflagging support. His zeal and love for coastal

engineering have inspired me throughout this four-year study, which has been a challenging,

joyful and unforgettable experience in my life.

My thanks also go to committee members, Professors Daniel M. Hanes, Robert J.

Thieke, Hsiang Wang and Ulrich H. Kurzweg, for their valuable advice, suggestions,

discussions, comments and patience in reviewing this dissertation.

Appreciation is extended to all other faculty members and staff in the department for

supplying various components of knowledge and help during this study. Special thanks go to

Becky Hudson, Sandra Bivins, Lucy Hamm, Cynthia Vey, Helen Twedell, John Davis and Dr.

Li-Hwa Lin for their help and kindness in the completion of this study. Gratitude is due to my

friends in the department, Xu Wang, Yigong Li, Linda Charles, Emre Otay, Albert Browder,

Michael Dombrowski, Taerim Kim, Susan Harr, Michael Krecic, Michael Bootcheck, and

Paul Devine, for their friendship and assistance.

My final acknowledgment is reserved for those whom I owe the most, my husband

Jian Liu, for his love, support, encouragement and patience through these years, and my

parents, who instilled in me the values of life that have made me to go this far, and have

supported me all my life.



ACKNOWLEDGMENTS ............................................... ii

ABSTRACT .................................................... vi


1 INTRODUCTION ......................................... 1

1.1 M otivation ......................................... 1
1.2 Nearshore Sediment Transport .......................... 2
1.2.1 Transport Processes and Related Problems ........... 2
1.2.2 Bedload and Suspended Load ..................... 5
1.3 Review of Cross-Shore Sediment Transport Relationships and Models
................................................ 7
1.3.1 G general ...................................... 7
1.3.2 "Closed Loop" Model ........................... 8
1.3.3 "Open Loop" Model ........................... 11
1.4 Scope of Study ..................................... 15


2.1 Introduction ....................................... 20
2.2 Critical Shear Stresses ................................ 22
2.3 G ravity ........................................... 23
2.4 Shear Stress Due to Nonlinear Waves .................... 24
2.5 Boundary Rotation Flow Related Shear Stress ............. 25
2.6 Shear Stress Related to Undertow ....................... 33
2.6.1 Linear Shear Stress Relationship .................. 34
2.6.2 Quadratic Shear Stress Relationship ............... 37
2.7 Comparison of Magnitude of Shear Stress Terms ........... 48

........................ ............. ................. 53

3.1 Equilibrium Beach Profile ............................ 53
3.2 Scale Analysis ...................................... 57
3.3 Discussion of Transport Relationship ...................... 62
3.4 M odeling Process ................................... 66
3.4.1 Numerical Method ............................. 66
3.4.2 Wave Run-up and Set-up ....................... 69
3.4.3 Dune, Shoreline and Offshore Slopes ............... 70
3.4.4 Random Wave Generation ....................... 72

EXPERIM ENTS ........................................ 79

4.1 Introduction ....................................... 79
4.2 Saville's Experiments ................................ 80
4.2.1 General Description ............................ 80
4.2.2 Calibration ................................... 81
4.3 German "Large Wave Flume" ......................... 86
4.3.1 G general ..................................... 86
4.3.2 Experiments With Constant Waves ................ 90
4.3.3 Experiment With Irregular Waves ................. 95
4.4 SUPERTANK Experiments .......................... 99
4.5 Results of Laboratory Calibration and Comparison ......... 101


5.1 Application of CROSS to Laboratory Experiments ......... 106
5.1.1 Saville's Experiments ......................... 107
5.1.2 German "Large Wave Flume" ................... 112
5.1.3 SUPERTANK Experiments ..................... 118
5.2 Results of Evaluation ............................... 120


6.1 Background ...................................... 124
6.2 Brief Description of Three Commonly Used Models ........ 126
6.2.1 CCCLModel ............................... 127
6.2.2 EDUNE Model .............................. 127
6.2.3 SBEACH Model .............................. 128
6.3 Storm and Beach Profile Characteristics ................. 132

6.3.1 Storm Assessment ............................ 132
6.3.2 Beach Profile Characteristics .................... 136
6.4 Comparisons of CROSS Model With Three Commonly Used Models
................................. ...... ....... 140
6.4.1 Input Case Design ............................ 140 CROSS M odel ......................... 141 CCCL M odel .......................... 141 EDUNE Model ........................ 142 SBEACH Model ....................... 143
6.4.2 Numerical Results and Comparisons .............. 143
6.4.3 Average Profile Change at Different Elevation Contours 168
6.5 Sensitivity Study of CROSS Model ..................... 170
6.5.1 Transport Coefficient .......................... 170
6.5.2 Active W ater Depth ........................... 174
6.5.3 Storm Surge ................................ 184
6.5.4 W ave Height ................................ 185
6.6 Summ ary ........................................ 197


7.1 Summary ........................................ 200
7.2 Conclusions ...................................... 201
7.3 Discussion and Recommendations ...................... 204

REFERENCES ................................................... 206

BIOGRAPHICAL SKETCH .......................................... 213

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Jie Zheng

August 1996

Chairperson: Robert G. Dean
Major Department: Coastal and Oceanographic Engineering

A modified nonlinear cross-shore sediment transport relationship is developed based

on equilibrium beach profile concepts and dimensional scaling relationships. This nonlinear

relationship provides a reasonable explanation for the significantly different time scales of

beach evolution evident in various laboratory experiments. To predict a beach profile

response under wave action, a finite difference method is applied to solve the sediment

transport and continuity equations numerically. The proposed nonlinear model called

"CROSS" is calibrated and compared with the commonly employed linear transport

relationship using laboratory data. A total of seven large scale wave tank experiments from

three different facilities are examined. The results demonstrate that the nonlinear transport

model provides overall better predictions than the linear transport equations. The CROSS

model and the three other commonly used models are applied to predict beach erosion at

Ocean City, Maryland, during the November 11, 1991, and January 4, 1992, storms. Seven

survey lines are available for comparison with the numerical simulations. Two versions (2.0

and 3.0) of SBEACH are applied. Among the four models, the CCCL model is the only one

overpredicting average dune erosion and the other three models have different degrees of

underprediction with SBEACH version 2.0 underpredicting most. Overall, CROSS, EDUNE

and SBEACH version 3.0 present reasonable predictions for both dune erosion and the entire

profiles. The sensitivity of CROSS to the transport coefficient, active water depth, storm

surge levels and the storm wave heights are studied for the storm erosion at Ocean City. It

appears that CROSS is very insensitive to the transport coefficient. The subaqueous part of

a profile is quite sensitive to the wave height and the subaerial part is less affected. The

CROSS model provides better predictions with the ratio of active water depth to incoming

wave height of 1 than with the ratio of 1.28, and the 20% increased storm surge yields a

better simulation.


1.1 Motivation

The beaches are the transition zone between land and sea. They protect the shore from

damage by coastal storms and hurricanes and provide valuable recreational resources. Their

effectiveness as natural barriers depends on their size and shape, sediment characteristics and

the severity of storms. Under the seasonal action of waves, currents and winds, beaches often

erode during winters and accrete during summers. During storm surges the sea level rises

considerably above the normal high water level. As a result the storm driven waves reach the

front of the dunes and beach erosion occurs. The eroded sand is moved in an offshore

direction and a new beach profile is developed. In fact, the process of storm erosion can be

considered as the continuous adjustment of the beach profile to the changing hydrodynamic

and meteorological conditions during the storm.

In coastal engineering, accurate estimates of beach profile evolution in response to

storm waves and high water levels and the adjustment of beach fill to long term wave action

are particularly important for a variety of regulatory and design purposes. The time scale

associated with storm induced beach erosion is on the order of hours to days and depends on

the duration, water level and wave conditions of the storm, while the time scale of beach fill

adjustment is from months to years and depends on the fill material and the wave climate at


the site. Due to the complexities of surf zone hydrodynamics, sediment characteristics, and

wave and water level conditions, an analytical treatment of beach profile change is difficult

and requires recourse to numerical modeling, which requires an understanding of sediment

transport. The objective of this study is to provide a better understanding of cross-shore

sediment transport relationships and develop a numerical model to represent beach profile

changes associated with variable waves and tides.

1.2 Nearshore Sediment Transport

1.2.1 Transport Processes and Related Problems

Sediment transport at a point in the nearshore zone can be considered in terms of

cross-shore and longshore components. It appears that under a number of coastal engineering

scenarios of interest, the transport is dominated by either the cross-shore or longshore

component and this, in part, has led to a history of separate investigative programs for each

of these two components. Longshore sediment transport figures prominently in situations

involving loss of sediment supply, such as damming of rivers, and impoundment at structures

and inlets. In these cases, longshore sediment transport is the major process governing

nearshore topographic change. The cross-shore component determines profile evolution

primarily for beaches far away from structures, inlets and river mouths, and under the case

of increased water levels, storms and beach nourishment. In contrast to longshore sediment

transport modeling, which has been studied for about five decades, a focus on cross-shore

sediment transport modeling is relatively recent (about 20 years) and the uncertainty in

predicting effects of all variables thus may be considerably greater.


Cross-shore sediment transport encompasses both offshore transport such as occurs

during storms and onshore transport which dominates during mild wave conditions. Transport

in these two directions appears to occur in significantly distinct modes and with remarkably

disparate time scales. The offshore transport moves sand from the dune to the offshore bar

in hours to days during a storm, whereas the onshore transport may take months to years to

move sand from the offshore bar to the dry beach. The difficulties in predicting transport in

the two directions also differ substantially. Offshore transport is simpler and the more

predictable of the two with transport more or less in phase over the entire active profile. This

is fortunate for coastal engineers since there is considerably greater engineering concern and

interest in offshore transport due to the potential for damage to structures and loss of land.

Cross-shore sediment transport is relevant to a number of coastal engineering

problems including (a) beach and dune response to storms, (b) the equilibration of a beach

project that is placed at a slope steeper than equilibrium, (c) "profile nourishment" in which

the sand is placed in the nearshore with the expectation that it will move landward nourishing

the beach (this involves the more difficult problem of onshore transport), (d) shoreline

response to sea level rise, (e) seasonal changes of shoreline positions, (f) overwash, the

process of landward transport due to overtopping of the normal land mass due to high tides

and waves, (g) scour immediately seaward of shore parallel structures (e.g. seawalls), and (h)

cross-shore and longshore coupled transport around shore normal structures (e.g. jetties and

groins) in which the steeper and milder slopes on the updrift and downdrift sides induce

seaward and landward components, respectively. These problems are schematized in Figure


Pre-Slorm Profile

Storm Water Level \ Immediately
S-- ..Z.- .- -.. /Post-Nourished
,, _Normal Water Level ..

Post-Storm .' ,-Pre-Nourlshment '-
Profile Equllibrate' h'~ '---

a) Response to Storms b) Beach Nourishment Evolution

Shoreline Response

Increased Sea Level
S.... Initial Sea Level

"^g~ a~rnShoreward?

"Profile" Nourishment
c) Profile Nourishment Evolution d) Profile Response to Sea Level Rise

"Summer" Profile
Post-Storm Profile with
Washover Deposits
St By Norm Water L evel
,. Be Normal Water Level

"Winter" Profile .-- --
'.'..- Pre-Storm Profile *--.-

e) Seasonal Profile Changes f) Effects of Storm-Induced Overwash


Storm Water Lev.


Storm-induced Scour
Adjacent to Seawall

g) Storm-Induced Scour Adjacent to Seawalls h) Flow of Sand Around Structures

Figure 1.1 Problems and processes in which cross-shore sediment transport is relevant (after
Dean, draft of Chapter II of Coastal Engineering Manual).

1.2.2 Bedload and Suspended Load

A complete understanding of cross-shore sediment transport is complicated by the

contribution of both bed and suspended load transport components. Bagnold (1956) defined

bedload as the part of the total load which is supported by intergranular forces, and the

suspended load as the part supported by fluid drag. Usually particles entrained in the bedload

exhibit motions of rolling, sliding and sometimes jumping in the bed layer, whereas particles

entrained in the suspended load are suspended in the water column. Turbulence is the most

important factor in the suspension of sediment. Due to gravity, particles tend to settle to the

bottom; however, irregular motions of the fluid introduced by turbulence tend to lift particles

from the bed and support them in suspension. For the cross-shore sediment transport, bedload

is dominant outside the surf zone because of the relative mild fluid conditions. However,

owing to the highly turbulent flow field caused by wave breaking, suspended load prevails

inside the surf zone.

Generally bedload is related to the shear stress acting on the bottom. Several

researchers have proposed expressions for the transport rate in an oscillatory flow. Bailard

and Inman (1981) suggested that bedload transport rate was proportional to the 1.5 power

of bottom shear stress. Trowbridge and Young (1989) presented a linear relationship between

the transport rate and the bottom shear stress. There are several causes of shear stresses

acting on the bottom in the nearshore active zone: (1) shear stress induced by gravity, (2)

shear stress due to nonlinear waves, (3) shear stress caused by return flow, and (4) shear

stress related to boundary rotation flow. The magnitudes of these stresses can be remarkably


different inside and outside surf zone. A detail discussion of these shear terms will be

presented in Chapter 2.

Suspended sediment transport in the cross-shore direction under wave action primally

results from the contributions due to the oscillatory first-order velocity (intermittent

transport) and the mean cross-shore current. Dean (1973) developed a heuristic model for

cross-shore transport in the surf zone based on the intermittent suspension of sand grains by

wave breaking and the eventual settling of sand to the bottom. A wave breaking in the surf

zone lifts sand from the bottom up into the water column to a particular elevation. The time

required for the sand to fall back to the bottom is determined by the ratio of the elevation to

the sediment falling velocity. If the fall time is less than half the wave period, the sand grain

should move onshore. Alternatively if the fall time is greater, the sand particle would be

carried offshore. The elevation is dependent on the wave height, as it seems reasonable that

larger waves would lift sand to a higher elevation than smaller waves. Therefore Dean

assumed that the elevation is proportional to the breaking wave height. The direction of

intermittent transport was determined by the so-called "fall velocity parameter," which is a

non-dimensional parameter of breaking wave height divided by wave period and sediment fall

velocity. For most ocean wave cases, the mean cross-shore current is caused by the mean

mass return flow in the offshore direction. As a result the suspended transport contributed

from the mean flow is always in the offshore direction no matter how high the sediment

suspended. Because the cross-shore velocity profiles inside the surf zone are different from

those outside the surf zone, the contributions of the mean flow to suspended sediment

transport inside the surf zone are different from those outside the surf zone. Under unbroken


wave conditions since the upper part of water column has greater return flow velocity but

very little suspended sediment, the suspended transport due to the mean flow is almost

negligible. By contrast, inside the surf zone the lower part of water column has larger return

flow velocities and high sediment concentration, therefore the suspended sediment transport

caused by mean flow could be significant.

1.3 Review of Cross-Shore Sediment Transport Relationships and Models

1.3.1 General

Some recent interesting reviews of cross-shore sediment transport models have been

reported by Broker Hedegaard et al. (1992), Steetzel (1993), and Schoonees and Theron

(1995). Different methods and viewpoints have been applied. Hedegaard et al. made an inter-

comparison of six coastal profile models against measured profile evolutions from a large

wave flume (under dune erosion conditions). Steetzel classified transport formulae as

empirical transport, shear stress related transport, energy dissipation related transport, energy

dissipation related suspension and velocity related transport. Schoonees and Theron

summarized ten internationally developed transport models according to their theoretical

basis and morphological verification data.

In the present review, cross-shore sediment transport models are broadly classified

into two groups: "open loop" and "closed loop" models. An "open loop" model is not

constrained a priori to the final profile and the sediment transport is determined by sediment

concentration and fluid motions. "Closed loop" sediment transport models are based on

equilibrium beach profile concepts (Bruun 1954 and Dean 1977) and assume that a profile will

eventually achieve equilibrium if exposed to the same conditions for a long time. Cross-shore


transport is caused by deviations of a beach profile from equilibrium. Methodologically "open

loop" models are microscale approaches based on detailed physics of the process, while

"closed loop" models are more macroscale and utilize conservation laws and heuristic

arguments that provide reasonable solutions. Owing to the complexities of highly turbulent

flow inside the surf zone, there is considerable uncertainty in the detailed physics of wave

breaking processes and related sediment transport. Therefore the macroscale approach is a

powerful tool to predict beach profile change at the present time.

1.3.2 "Closed Loop" Models

The earliest relationship between increased water level and profile response was

developed by Bruun (1962), which does not require any specific form of the equilibrium

profile, but only requires that the form be known. The response is considered in terms of the

horizontal recession of the profile and sea level rise. Two requirements must be satisfied by

the new profile. First, the profile shape does not change with respect to the sea level.

Secondly, the sand volume in the profile must be conserved. One restriction of the Bruun

model is that the sea level rise should be much less than the berm height, since the berm height

relative to the rising sea level is considered fixed. Edelman (1972) removed the limitation of

relative small changes in water level by accounting for the progressive decrease in relative

dune elevation. Since the two above assume that a beach profile maintains its equilibrium form

immediately in response to the water level rise, no sediment transport relationship is required.

However, it is not suitable to apply such static models (no lag) to short-term profile changes.

Following the idea of Bakker (1968), Swart (1974 a, b) schematized the beach profile

into 4 zones referred to as backshore, onshore profile, offshore profile and transition area and


developed a time-dependent cross-shore model. With L, and L2 representing the schematized

lengths of onshore and offshore profiles, respectively, the transport equation is written as

Q = q [W-(L2-Li),] (1.1)

where Q is the time dependent cross-shore sediment transport rate, q is a constant for a

specific set of boundary conditions, and W and (L2-LI)t are equilibrium and time dependent

values of (L2-L,), respectively.

Chiu and Dean (1984,1986) developed the CCCL model, which is still used by the

State of Florida to establish the location of the State Coastal Construction Control Line

associated with the erosion limits of a 100 year return period storm event. Under erosive

water level and wave conditions, the time dependent beach recession, R(t), at each depth

contour is given by

R(t) = R(1 -e -Kt) (1.2)

in whichR_ is the equilibrium recession for the particular condition at that time and K is a

decay parameter. Strictly speaking, this is a beach profile change model since there is no

sediment transport relationship involved. Compared with Bruun and Edelman models, CCCL

is time dependent and can be applied to short-term profile change caused by storms.

Two wave energy dissipation related transport models are EDUNE (Kriebel 1982,

1990, Kriebel and Dean 1985) and SBEACH (Larson and Kraus 1989, Larson et al. 1989).

According to equilibrium beach profile concepts, offshore transport continues until the wave

energy dissipation per unit volume of water becomes uniform over the entire surf zone. Based


on this consideration, EDUNE expresses the local cross-shore sediment transport rate, Q, in

the surf zone as

Q = K(D D) (1.3)

in which K, is a dimensional transport coefficient, D and D. are the actual and equilibrium

wave energy dissipation per unit volume of water, respectively, and D. depends on the sand

size. Eq. (1.3) was first proposed by Moore (1982) and later modified by Kriebel (1982) and

Kriebel and Dean (1985). In SBEACH, the sediment transport rate across the nearshore

active zone is calculated by different relationships in four distinct zones identified as pre-

breaking zone (extending offshore from the breaking point), transition zone (from the break

point to the plunge point), broken wave zone (the plunge point to the seaward limit of the

swash zone), and swash zone (from the seaward swash zone to the run-up limit). A transport

relationship similar to Eq. (1.3) is applied in the broken wave zone with a term added to

account for the effect of local slope and written as

E dh E dh
Q = Ks(D -D, + -), D>D,

= O, D K, dx

where K, and e are dimensional coefficients and dh/dx is the local profile slope. The transport

rates in the other three ranges are established empirically in accordance with Saville's large

wave tank experiments (Kraus 1988).

1.3.3 "Open Loop" Models

"Open loop" models usually relate cross-shore sediment transport to the detailed

physics of the flow field such as sediment concentration, fluid velocity, and bottom shear

stress. Three different types of formulae will be briefly reviewed here.

(1) Sediment transport rate based on the product of flow velocity and sediment concentration.

The Dally model (Dally 1980, Dally and Dean 1984) presented the net time-averaged

flux of suspended sediment past a section in the nearshore zone as

Q = f u(z) c(z) dz (1.5)

where u(z) and c(z) are averaged horizontal velocity and sediment concentration at level z,

respectively. The flow regime is divided into an upper layer where only mean flow is

considered and a lower layer where both mean flow and orbital velocity are taken into

account. In the lower layer, the fall time of a sand particle is determined by its elevation and

fall velocity. According to linear wave theory, the particle will traverse a horizontal

displacement during the fall time. Thus, the average horizontal velocity of the particle during

one wave cycle due to the orbital motion is equal to the ratio of the horizontal displacement

to the fall time. The interface between the layers is determined by the elevation at which a

particle will fall back to the bottom in one wave period.

LITCROSS (Broker Hedegaard et al. 1992) is a coastal profile model developed by

the Danish Hydraulic Institute. The net sediment transport is calculated as bed and suspended

load. The instantaneous bed load and nearbed boundary condition for the vertical distribution


of suspended sediment are determined as functions of the instantaneous shear stress

(Engelund and Fredsoe 1976). The total suspended sediment transport over a wave period

is found from

Q= f fcudzdt+ f u,dz (1.6)
period depth depth

where the velocity, u, and the concentration, c, in the first integral are functions of time and

elevation, u, is Lagrangian drift, E denotes the time averaged concentration, and T is the

wave period. The time varying vertical distribution of suspended concentrations is calculated

by the vertical diffusion equation according to Deigaard et al. (1986).

Steetzel (1993) also presented his cross-shore transport model as the integral of the

product of concentration and velocity over time and depth. The sediment concentration is

solved from a one-dimensional, non-stationary convection-diffusion equation. The reference

concentration at the bed is assumed to be related to the kind of breaking and the dissipation

of the turbulent kinetic energy. The cross-shore transport rate is calculated in two parts, that

is, below and above the trough level. Above the trough level, concentration is a constant and

given by that at the mean water level.

(2) Sediment transport rate related to velocity powers

A representation of this type of model was proposed by Bailard (Bailard and Inman

1981, Bailard 1981, 1982) based on the concepts of stream transport developed by Bagnold

(1963, 1966). Bagnold assumed that the sediment was transported in two distinct modes,

each differing by way of the support of the sediment grains. Sediment transport as bedload


is supported by the bed via grain to grain interactions, while sediment transport as suspended

load is supported by the stream fluid via turbulent diffusion. In both modes, energy is

expended by the stream in transporting the sediment load. Bagnold, comparing the stream to

a machine, defined the sediment transport efficiency as the ratio of the rate of energy

expended by sediment load to the total rate of energy produced by the stream. In essence, he

assumed that the instantaneous sediment transport rate was directively proportional and

reacted immediately to the instantaneous energy dissipation rate per unit bed area.

Bailard (1981) refined Bagnold's work for the cross-shore sediment transport problem

and expressed the total load time averaged sediment transport equation as

Q = pcf n2 > t anf 3> pc u3u> tanp < I j 5> (1.7)
Stan4 tan I W W I.

where p is the density of water, cf is drag coefficient of the bed, EB and Es are bed and

suspended load efficiency factors, ( is the internal angle of friction for the sediment, tanp is

the bottom slope, iu is the instantaneous velocity vector, i is a unit vector in onshore

direction, W is the sediment fall velocity, and the < >means time average over a wave period.

This formula has been widely applied by others such as Stive (Stive 1986, Roelvink and Stive

1989) and Nairn (1990). Three coastal profile models have been developed based on Bailard

transport relationship. They are UNIBEST of Delft Hydraulics, NPM of Hydraulic Reseach,

and SEDITEL of Laboratoire National d'Hydraulique.

There are other velocity power related sediment transport relationships. One example

was given by Shibayama and Horikawa (1982). The transport rate accounts for bedload and


suspended load. Both bed and suspended loads are proportional to the velocity to the sixth


(3) Sediment transport determined by the product of velocity and bottom shear stress

Watanabe (1988) presented a sediment transport rate equation related to local wave-

current conditions and bottom shear stress. The total transport rate is the sum of contributions

from mean current, Qc, and waves, Q,, which are expressed, respectively, by

Qc = Ac(t-tc) p pg(1.8)
Q, = AwF ,(z-Zc)u pg

where Ac and A, are nondimensional coefficients for mean current and wave induced

transport, respectively, U is the current velocity vector, ib is the amplitude of the near-

bottom wave orbital velocity vector, T is the maximum value of the bottom shear stress in a

wave-current coexisting system, T; is the critical shear stress, F D is a direction function for

wave-induced net transport, and g is the gravitational acceleration. The coastal profile model

WATAN3 developed by University of Liverpool is based on the Watanabe transport


There are some relationships in addition to the above three types such as the one

developed by Nishimura and Sunamura (1986), in which the transport rate is related

empirically to the Ursell number representing the skewness of water particle velocity profile

and the Hallermeier parameter indicating the intensity of sediment movement.

Most "open loop" profile change models include modeling of hydrodynamic processes

to different degrees. The sediment transport calculation is related to the properties of


sediment and flow field. A interesting inter-comparison of five coastal profile change models

developed in Europe was presented by Broker Hedegaard et al. (1992). The five models are

LITCROSS of Danish Hydraulic Institute, UNIBEST of Delft Hydraulics, NPM of Hydraulic

Research, WATAN3 of University of Liverpool, and SEDITEL of Laboratory National

d'Hydraulique. The associated transport relationships have been described above. The models

were tested against measured profile evolution from a large wave flume. The initial profile is

a planar beach with a slope of 1:20. The experiment was carried out under regular waves with

a wave height of 1.5 m and period of 6 s. The velocities and the transport rates calculated

across the planar profile are shown in Figures 1.2 and 1.3. The measured and calculated

profiles after 4.3 hours are compared in Figure 1.4. It is noticed that the all five models

present quite good fits for the profile although the calculated velocity and transport rates are

markedly different. It seems that considerable uncertainty in the understanding of surf zone

hydrodynamics and sediment transport process still exists.

1.4 Scope of Study

A major objective of this dissertation is to improve "closed-loop" cross-shore

sediment transport relationships. In the following studies, a modified transport equation is

proposed according to dimensional arguments. A cross-shore profile change model called

CROSS is developed based on the modified transport equation and the conservation of sand.

Three sets of different laboratory experiments are applied to the calibration of CROSS.

Finally, the model is tested and compared with other three commonly used "closed loop"

models against data obtained from field storm erosion.

I U L ~

i I


o 1i 2 30 1 10 0S90 160Cr) X

DHI, Litcross
l, ------- .


i 44 .. 4.1 ` -- .12 as J. 1 -

LNH, Seditel

izil' Iz f~1I F



.- -- I4

a -U,.

10 46. 4 4 42 .0 0. 0.4 aL keI 1.

DH. Unibest
-- W* -- -- IY

. L1

I -I- -. .....I" /..l I I I

It -

S .1. 48 4* 4. 4.2 4.o 1 O. as as 1.*


S- Le $ad

1 Vertial Curenl
IO Componren at 52 Mi

Figure 1.2 Velocity components calculated by NPM, LITCROSS, UNIBEST and SEDITEL

and 2-D current field calculated by SEDITEL (after Broker Hedegaard et al., 1992).


0 4.1 -4. 4.4 41s 4. .I 1.4 0. as 1.





0- SWI.

0 . .
50 100 m
DHI, Ltcross DH, Unibest

s M 0 0 o I" -li e Is so a t s i i B oW I
LNH, Seditel UL, Woian3

4 -----------. .... '. -- -- ----

A io 4U 1 4U a i l ie 1 14 I is s U 40 i i o0 SO IN 1 10 I

Figure 1.3 Cross-shore sediment transport calculated by NPM, UNIBEST, LITCROSS,
SEDITEL and WATAN3 (after Broker Hedegaard et al., 1992).

To better understand the surf zone sediment transport process, Chapter 2 commences

with a qualitative description of the time-averaged bottom shear stresses acting in the

nearshore. The total shear stress is composed of four components. They are shear stress due

to gravity, shear stress induced by non-linear waves, shear stress related to the boundary

rotation flow, and shear stress caused by the mean mass return flow.

Chapter 3 is devoted to the development of CROSS based on the concepts of

equilibrium beach profiles. The relationship between the transport rate and the deviation of

the actual wave energy dissipation per unit volume from equilibrium is discussed according

*o -NPM

------------------ -----


": .... 5- -- .... -

o woo -.0 -0-00 7,S0.00 W.oot(m)

10- ----- ---- -0- EXPE00 -o-TAL (ES,
-10.00 t0.0 J.OO 5o0 7000 0.00 (m)

0.0* ---- ----- -----l----- ^^^lCY*LLE-I
-O100 00 00 OO 00 OCt




------- ----- -- ------,__ -------

-10.00 10.0 30.0 000 70.00 0.00 (m)

H 1.5m
T 6 s8
d k 0.33mm

Initial slopes: lower beach 1:20
upper beach 1:40

Figure 1.4 Comparison of measured and calculated coastal profiles after 4.3 hours of
exposure (after Broker Hedegaard et al. 1992).

to dimensional considerations. The profile change is determined by both the transport and the

continuity equations. The two equations are solved by the finite difference method with a

staggered grid. A detailed numerical process is presented.

The calibrations of CROSS are described in Chapter 4. Three sets of large wave tank

data are utilized, including three cases in the large German wave flume in Hanover (Dette and

Uliczka 1987), three cases from Saville's large wave tank experiments (Kraus 1988), and one

case from the SUPERTANK experiments at Oregon State University (Kraus and Smith

1994). Among the seven experiments, one of the German set was carried out with a Jonswap-

Spectrum of random waves, the Supertank case was run with the combination of random


and regular waves, and the others are conducted under regular waves. The best-fit transport

coefficient is determined as the value providing the least squares error for the fitting of eroded

volume in each experiment. Based on the calibration results, a transport coefficient is

recommended for the applications of CROSS according to the average of the best-fit values

in the seven experiments.

In Chapter 5, the CROSS model with the recommended transport coefficient is

evaluated by the same experiments used in Chapter 4 for the calibration of CROSS. The

predictions of CROSS with the recommended transport coefficient are compared with those

with the best-fit values for each experiment. The comparisons include the time history of

eroded volume and the profile evolutions.

The applications of CROSS for the November 1991 and January 1992 storm erosion

at Ocean City, Maryland, are presented in Chapter 6. The predictions of CROSS are

compared with three other models commonly used in the United States. The three models are

CCCL (Chiu and Dean 1984,1986), EDUNE (Kriebel and Dean 1985, Kriebel 1989,1990),

and SBEACH (Larson and Kraus 1989, Larson et al. 1989). A sensitivity study of CROSS

is also included in this chapter to test the sensitivity of the model to changes wave and surge

conditions during storms.

Finally, Chapter 7 presents a summary, documents conclusions of the study, and

proposes some recommendations for future study.


2.1 Introduction

A requirement for a proper understanding of coastal sediment transport is a

consideration of forces which act in the nearshore active zone. There are three kinds of

forces governing the behavior of sediment particles whether they are resting on the bed or

suspended in the water column. These are gravity forces, intergranular forces related to

continuous sediment contact, and the wave and current induced fluid forces. The magnitudes

of wave induced forces can by remarkably different inside and outside the surf zone. Due to

the strong and large scale turbulent flow caused by wave breaking, sediment is typically highly

mobilized inside the surf zone. Changes in beach profiles result from intermittent suspension,

turbulence, bottom shear stress and gravity. Outside the surf zone, the wave energy

dissipation and the turbulence occur mainly within the near-bed wave boundary layer. Within

this zone, the bottom shear stress related bedload sediment transport becomes the primary

factor in controlling profile change.

Under equilibrium conditions, the forces acting on the sediment particles are in

dynamic balance, and, although there is motion of individual particles under even low wave

activity, the profile remains more or less static. Cross-shore sediment transport occurs when

the nearshore hydrodynamic conditions change thereby modifying one or more of the forces


resulting in an imbalance and causing a sediment transport gradient and profile change.

Established terminology is that onshore and offshore directed forces are referred to as

"constructive" and "destructive," respectively. It is evident that some forces could behave as

constructive under certain conditions and destructive under others. As the result of the

combined constructive or destructive forces, sediment is transported onshore or offshore,

respectively. For suspended particles, the movement is determined by gravity, drag, lift, added

mass force, etc. acting on the particle. For particles resting on the bed, the bottom shear stress

is the dominant force. When the bottom shear stress is less than the critical shear stress

(friction), particles are under static conditions and no motion occurs. As the bottom shear

stress overcomes friction, particles will be moved in the shear stress direction and bedload

transport commences. Generally, the bottom shear stresses in the cross-shore direction

include contributions from gravity, waves, currents, bottom boundary layer rotation flow, and

undertow due to the mass and momentum flux. As waves and currents vary with time, the

shear stresses also change temporally. In principle, the sediment transport processes are

related directly to the instantaneous shear stresses. However, as a first step, only mean shear

stresses averaged over one wave period will be discussed here.

In this chapter, the critical shear stress is discussed first. When the total bottom shear

stress is less then the critical shear stress, the intergranular friction will keep sand particles

from moving. Under wave conditions, the total bottom shear stress is due to the superposition

of four components. These are shear stress induced by gravity, shear stress due to nonlinear

waves, shear stress related to the boundary layer rotation flow, and shear stress caused by the

mass return flow. Their magnitudes depend on the wave breaking, wind speed, sediment size


and bed slope. To provide a better understand of these forces, a detailed example will be

presented to quantify the directions and magnitudes of the four stress terms both inside and

outside the surf zone.

2.2 Critical Shear Stresses

The intergranular forces acting on a sediment particle are the normal and frictional

forces resulting from the surrounding sediment. For grains at rest, the static angle of repose,

(k, is determined by the friction coefficient, i.e. the ratio between the maximum sustainable

shear stress (i.e. critical shear stress), tr and the effective normal stress, oe (see Figure 2.1).

For mostly sandy materials, the static angle of repose is between 26 and 34 degrees (Nielsen,

1992). For shearing (moving) materials of near maximum concentration, the friction

coefficient is only slightly less than the static. Hence the dynamic angle of repose, (d is of

a similar magnitude to the static angle. Hance and Inman (1985) suggested a typical value of

31 degrees for beach sand.

A different parameter, the Shields parameter, describing the frictional shear stress was

suggested by Shields in a study of the incipient motion in steady flow as

= (2.1)

where 0, represents the Shields parameter, p is mass density of water, s is relative density of

sediment, and D is the sediment diameter. The critical Shields parameter, 0,cr, which

corresponds to the critical shear stress, tr,, is the effective Shields parameter at which

sediment movement starts. For sand grains quiescent under static water, the critical Shields

parameter is related to the angle of repose by the relationship, 0,cr = sin (4. Under waves and

/; Mg Figure 2.1. The friction
M Pi coefficient, (W/o),
equals tank,, where (,
f is the static angle of
a = Mg cosp Mg repose.

currents, this parameter becomes much less because the fluid lift force and near-bottom

turbulence decreases the effective normal stress. Although some laboratory studies (Natarajan

1969, and Hammond & Collins 1979) and field studies (Amos et al. 1988) have been

conducted, it is impossible to define a sharp criterion for the initiation of motion under

combined wave and current flows. Based on their observations, Amos et al. recommended

0.04 as the critical Shields parameter. Therefore, for the case of bottom shear stress less than

0.04 times the submerged weight of the sand, sediment will remain static on the bottom due

to friction effects.

2.3 Gravity

When the sand bed is horizontal, gravity acts normally to the bottom and has no

contribution in the shear direction. For a monotonic beach profile, gravity becomes an

obvious destructive force which acts downslope, i.e. seaward. However, for the case of a

barred profile, gravity can act in the shoreward direction over a portion of the profile. Overall,

gravity tends to "smooth" any irregularities that occur in the profile. If gravity were the only

force acting, the only possible equilibrium profile would be horizontal and sandy beaches as


we know them would not exist. For particles in the sand bed, the tangent component of

gravity force per unit bottom area at a layer of one grain thickness will be determined by the

component in the beach slope direction as

Tb = apg(s-1)dsinp (2.2)

Where a is the sediment concentration in the sand bed, d is the mean diameter of the

sediment, and p is the beach slope.

2.4 Shear Stress Due to Nonlinear Waves

As noted above, the constructive forces are those that tend to cause onshore sediment

transport. For classic nonlinear wave theories (Stokes, Cnoidal, Solitary, Stream Function

etc.), the wave crests are higher and of less duration than the troughs. This feature is most

pronounced just outside the breaking point. The oscillatory water particle velocities are

usually expressed as a sum of phase-locked sinusoids such as for the Stokes and Stream

Function wave theories. Although the time averaged water particle velocity is zero, the

average bottom shear stress, tb is directed onshore and expressed as

b=P IVbl (2.3)

Where fis Darcy-Weisbach friction coefficient and vb is the instantaneous wave induced water

particle velocity at the bottom. Dean (1987) developed the average bottom shear stress based

on the Stream Function nonlinear wave theory and presented the results in the non-

dimensional form shown in Figure 2.2.

2 MIH/H b =1.u--


j 37
,S, 2 -

5 (1) Hb =Breaking Wave Height

(2) La =2-
2 5 10'2 2 5 10 2 5 100 2

Figure 2.2 Isolines of non-dimensional average bottom shear stress,
Tb, versus relative depth and wave steepness (Dean 1987). Note the
bottom shear stresses are directed shoreward.

2.5 Boundary Rotation Flow Related Shear Stress

The second constructive force originates from the net mean onshore directed

streaming velocity induced by wave motion in the bottom boundary layer. This streaming

motion was first observed by Bagnold (1940) in the laboratory and has been studied

quantitatively by Longuet-Higgins (1953). The mechanism producing a mean bottom shear

stress is reviewed as follows"'. To satisfy the no-slip bottom boundary condition, the

[n The following development is a modification of the Longuet-Higgins approach due to
James Kirby (personal comminution).


particles of fluid in contact with the boundary must have the same velocity as the boundary

itself. But quite near the bottom the fluid is in motion with velocities comparable to that in

the interior potential flow of the fluid. This implies that there is strong vorticity in the

neighborhood of the bottom. This vorticity modifies the vertical velocity, which is correlated

with the streamwise velocity at the outer edge of the boundary layer, so that there is a change

in the mean flux of streamwise momentum across the top of boundary layer. This change in

the mean momentum flux must be balanced by a mean bottom shear stress.

The case considered here is that of periodic linear waves propagating shoreward

above a gently sloping sand bottom with straight and parallel depth contours. The equations

governing the water waves flow are the continuity equation and the Navier-Stokes equations

av 8w
-+_ = 0
ay az
av 8v2 vw 1 aP 2 v 2v
+ -- + -- = + e (-- (2.4)
t ay 9z p y y2 z2 (4)
8w + vw aw2 1 1P +(a2w 82w
at ay 8z p az ay2 az2

where the y axis is positive offshore and coincides with the mean water surface, the z axis is

perpendicular to the water surface and positive upward, (v,w) are the corresponding velocity

components in (y,z) directions, P is the water pressure, and e is the kinematic eddy viscosity.

It is convenient to divide the flow into two parts with potential and rotational components,

v = v+vr y+v
S r r
W vt + JZ + W


In which, the subscripts p and r represent the potential and rotational components

respectively, and D is the velocity potential for the interior wave flow. The no-slip boundary

condition requires the velocity (v,w) to be equal to zero at the bottom. It is assumed that the

conditions at the free surface are unaffected by the viscous flow since the boundary layer is

assumed thin. This requires

v- -- or v-0 as (h +z)-.o
ay (2.6)
w- -- or w,-0 as (h +z) -

where h is the mean water depth. The first order potential and viscous velocity components

can be solved from the linearized Euler and approximate boundary layer equations,

V2D =0
avr a2vr
at az2 (2.7)
wr 2Wr
8t z 2

where V is the Laplace operator. For waves propagating onshore, the solutions of Eqs. (2.7)

are of the forms

1l = 0.5Hei(kyOt)
S= (z) ei(ky+ot)
Vr= r(Z) e i(ky+ot) (2.8)
Wr = r() ei(ky+ot)


in which i is the imaginary number ( T/v ), rI is the instantaneous water elevation, H is the

wave height, k is the wave number, and a is the wave angular frequency. Solving Eqs. (2.7)

with the boundary condition (2.6) gives

(z) = A coshk(h+z)+B sinhk(h+z)
h +z
,(z) = Cexp[-(l+i) (2.9)

(z) = D exp[-(l+i) ]

where 6 = i/2e/ represents the boundary layer thickness and is very small compared with the

water depth, and A, B, C and D are constants to be determined. By substituting (2.9) into the

continuity equation and the no-slip bottom boundary condition, the constants B, C and D can

be replaced by the constant A. The solution becomes

0(z) = A [coshk(h+z) kb sinhk(h+z)]
=ikA exp[-( l +i)- -
V,(z) = ikAexp[-(l+i) ] (2.10)

*,(z) = -i k28A exp[-(1+i) h+z
2 8

in which, A is related to the wave height through the linearized free surface dynamic and

kinematic boundary conditions as

H = i A [coshkh -ik sinhkh]
2 g 2 (2.11)
H kA 1-i ( )
H = -k [sinhkh- k-bcoshkh]
2 ia 2


Eliminating H and A in Eqs. (2.11) results in the dispersion relationship

2 [1 --k tanhkh] = gktanhkh [1 --i k6cothkh] (2.12)
2 2

As viscous effects approach zero, (2.12) clearly approaches the classic wave dispersion

relationship, 02 = gk tanhkh, i.e., Dean and Dalrymple (1991). In most cases, k8 << 1, so that

the deviations to a and k determined by (2.12) are small.

Eq. (2.12) generally predicts complex o and k combinations, which correspond to the

wave energy decay temporally and spatially. The two simplest cases are as follows:

(1) o is real and k=k, +iki. Time periodic waves decay with propagation distance.

(2) k is real and a = or + io,. Spatial periodic waves decay in time.

The first case is more generally applicable to ocean waves, while the second case may be used

to describe cases such as wave decay in basins, etc. For the problems concerned, we assume

k1/k,l<< and klh < 1 and analyze the first alternative further in detail. Substituting k = kr + iki

into Eq. (2.12), retaining only the first order terms, and separating the real and imaginary

parts yield

kr6 k6
o2 [1 rtanhkrh] = gk [tanhkrh- ]
2 2
Skr8 gkrkh gkr2 (2.13)
o2 tanhkh = +- +g ki tanhkrh
2 cosh2krh 2

Assuming kr = k +Ak, where k is determined by 02 = gk tanhkh, we obtain

kr =k+
2kh + sinh2kh (2
k. = k2
2kh + sinh2kh

Thus waves of fixed frequency are shortened by friction, and their phase speed is reduced.

The constant A can be solved from (2.11) as

Ho k6 .n
A = -i Ho [1+-cothkhexp(-i-)] (2.15)
2ksinhkh 42 4

Combing Eqs. (2.8), (2.10) and (1.15) together, the real parts of potential and rotational

velocities are presented as

) H coshk(h +z) sin Hocosh kz sn
2ksinhkh 2/2 sinh2kh 4

Hocoshk(h+z) os Hok6coshkz cos(-)
v,(z) = -cos cos ( ) (2.16)
2 sinhkh 2r2 sinh2kh 4

SHasinhk(h+z) sin Haksinhkz sin(-
w,(z) sin sin(--)
2 sinhkh 22 sinh2kh 4


-r Ho H/o kicosh kh Q
r(z) = e s cos(-6-) + H -- coskh Cos -
2sinh kh 2v sinh2 kh 4(2.17)
w(z) = -e- Hk sin(O---)
2v2 sinhkh 4


where 6 = ky +ot and h+z Since 6/h is very small, the higher order terms are


We now consider the streamwise mean momentum flux in the bottom boundary layer

and the possibility of mean flows. Neglecting the effects of wave height damping and wave

length shortening caused by bottom friction, averaging the second equation of (2.4) over a

wave period yields a second order mean flow in the boundary layer

avw a2pr
e (2.18)
az az 2

where the over bar means the time average over a wave period. The mean shear stress caused

by the boundary viscous flow is obtained by integrating (2.18) from an arbitrary depth, z, to

the surface

t(z) = pe pvw|-pvwL. (2.19)

where the velocities v and w include both the potential and rotational parts. Applying Eqs.

(2.16) and (2.17) into the term vw yields

vw = vpwp +vpw,+v,Wp+Vrwr (2.20)

with the terms in the right hand side are expanded as

H2 2k6
VpWp z 16 sinh2kh
VpWr z = -_ H2 2k86 coshk(h+z) .
vw | = e sin-(-+-)
8 V/2sinh2kh 4

VrWp Iz = -e- H2 2 sinhk(h+z) sin (2.21)
8 sinh2kh
H2 2k6 k
-e [sinhk(h+z) coshkh sin ( +-) + sinh kz sin( -c)]
8 /2sinh3kh 4 4
VrWr Iz = e-2 H2a2k6
16 sinh2kh

Therefore the mean shear stress caused by the boundary layer rotation flow is

z(z) = pH2a2 inhk(h+z) sin +k6 [sinhk(h +z) sin +cosh k(h+z) cos -9 e -(2.22)
8 sinh2kh tanhkh 2

which decays exponentially outside the boundary layer. The corresponding bottom shear

stress is given by (2.22) with z=-h as

b = H2 (2.23)
16 sinh2kh

For shallow water conditions, kh <<1, substituting (2.22) into (2.19) and integrating from the

bottom to an arbitrary depth, z, with the no-slip bottom boundary condition, the time mean

flow in the boundary layer is given as

S=- H2ok [3 -2(+3+Z)cosh e- -2(h+z)sin e-+e-2+ ] (2.24)
16 sinh2kh h h


which is usually referred to as the "streaming velocity" and is directed in the onshore

direction. It is seen that just outside the boundary layer, this time averaged rotational related

velocity is independent of the value of the viscosity and is given by

VrL = (2.25)
16 sinh2kh

which is 1.5 times the average of the return flow due to the mass transport.

2.6 Shear Stress Related to Undertow

The seaward return flow of wave mass transport induces a seaward stress on the

bottom sediment particles. In a real beach, a net onshore or offshore mass flux may exist in

the presence of three dimensional circulations, therefore an additional shoreward or seaward

bottom shear stress can result from the non-zero net onshore or offshore mass flux. It is well

known that associated with the wave propagation toward shore is a shoreward linear

momentum flux (Longuet-Higgins & Stewart 1964) and wave energy damping. Outside the

surf zone, wave energy damps slowly with distance due to dissipation in the bottom boundary

layer. Inside the surf zone, the waves decay relatively rapidly due to turbulent dissipation

caused by wave breaking. As a result of wave energy damping, the momentum is transferred

to the water column causing a shoreward directed thrust and thus a wave induced set-up, the

gradient of which is proportional to the gradient of wave energy dissipation. This momentum

flux is distributed over depth as shown as Figure 2.3. In shallow water, linear wave theory

predicts that one third of the momentum flux originates between the trough and the crest

levels and has its centroid at the mean water level and that the remaining two thirds, which

Figure 2.3 Distribution of the onshore component of momentum flux over

originates between the bottom and the mean water level, is uniforrply distributed over this

dimension and thus has its centroid at the mid-depth of the water column. Because of the

momentum contribution within the free surface, wave energy dissipation by breaking induces

an equivalent shear force on the water surface which will be quantified later. This surface

stress causes a seaward bottom shear stress, which does not exist outside the breaking zone

and will become very significant within the breaking zone. The magnitude of the induced

bottom shear stress is dependent on the rate of wave energy dissipation and the relationship

of the velocity gradient and shear stress. Over the water column, this effect of surface shear

stress caused by the momentum transfer must be balanced by the bottom shear stress and the

pressure gradient due to the slope of the water surface.

2.6.1 Linear Shear Stress Relationship

The linear Boussineq shear stress relationship with eddy viscosity, e, is applied as

= pe- (2.26)


Considering linear wave theory, the momentum equation for cross-shore flow is given by the

second equation of (2.4). Assuming shallow water waves, kh<
period and integrating from the free surface, z = 0, to an arbitrary elevation, z, yield

( E E pgz (T (z)) = 0 (2.27)
ay 2 h ay y

where E = pgH2 is wave energy density per unit surface area, and i is mean water level.
The first two terms of Eq. (2.27) represent the so-called radiation stress components with the

first being the contribution between trough and crest and the second below the mean water

surface. The third term is the effect of the slope of the mean water level. The -r in the fourth

term is the wind shear stress on the free surface. Applying Eq. (2.26) into (2.27) and

integrating from the bottom, z= -h, to an arbitrary depth, z, with the no-slip bottom

boundary condition yield

p+(z) = (h+z) + g2- +h a (2.28)
T y 2 y h ay

Integrating the above equation over the entire water depth results in the volumetric flux,

which is equal to the sum of the net volumetric flux, qet, and the wave mass transport flux,

, where c is the wave celerity. The slope of the mean water surface is therefore
determined as

a8 -I -3 p E (2729)
ay pgh 2 4 Sy h pc qne


Substituting (2.29) into (2.28), we obtain the final expression for the velocity

h -aE 3 Z)2 z 3 (E +
()=- 2 -+ 1 -2
peC ay) 8 h 2h 8 2h pc J h


The corresponding bottom shear stress is given as

Pb 1 aE3pe E +
r P e- Z=-h + +
8 z 12 4 ay h2 pc


The first term on the above right hand side is the contribution of the surface wind. Onshore

winds cause a shoreward directed surface flow and a seaward directed bottom flow as shown

in Figure 2.4. Of course, offshore winds would cause a shoreward directed flow near bottom.

Thus landward and seaward directed winds result in destructive (offshore) and constructive

(onshore) forces, respectively. The second term is related to the part of momentum flux

originating between the wave trough and crest. Since the gradient of momentum flux is due

*- Wind Stress
-~ -- -

Shear Stress *-.
Acting on Bottom

Figure 2.4 Bottom stresses caused by surface winds.


to the rate of wave energy dissipation, this term is vary important for breaking waves but

almost negligible outside surf zone. The third term results from the mass flux. The seaward

return flow of the wave mass flux induces a seaward bottom shear stress. A net offshore or

onshore volumetric flux may exist when the return flow is enhanced or weakened due to

gradients in the longshore current. In storm events in which there is overtopping of the barrier

island, a portion or all of the potential return flow due to mass transport can be relieved. A

net onshore volumetric flux induced by the weakened return flow causes an onshore directed

bottom shear stress, while a net offshore volumetric flux acts oppositely and results in a

offshore directed bottom shear stress.

2.6.2 Quadratic Shear Stress Relationship

As waves propagate into shallow water, the wave related flow becomes more

turbulent. After wave breaking, the turbulence near the free surface is much stronger than that

near the bottom. Figures 2.5 and 2.6 present variations of mean turbulent kinetic energy, k,

from experiments of Ting and Kirby (1994), and Stive and Wind (1982) for spilling and

plunging breakers, respectively. In the figures, h/hb represents the ratio of local water depth

to the breaking water depth, and z-_ denotes the relative elevation. It appears that the
turbulent level decays in both types of wave breakers. The assumption of constant eddy

viscosity is not suitable anymore. Following Prandtl and Von Karman, the shear stress under

conditions of turbulent flow is related to the velocity shear as follows

av avy
az= pl2 I (2.32)
az aa

where I is the mixing length for an open channel flow. The form presented by Reid (1957)

S(a) 0

S-0.4 0 ,
:0 +0 *
0 X

No -
-0.8 *
0 I I ,.
0.02 0.03 0.04 0.05 0.08 0.07 0.08


-0.4 0


-0.8- o x +

0.03 0.04 0.05 0.08 0.07 0.08 0.09

(Ilgh) I

Figure 2.5 Variation of turbulent kinetic energy with depth in spilling breakers. (a) Ting
and Kirby (1994); h/hb = 0.879 (o), 0.809 (*), 0.744 (+), 0.668 (*), 0.563 (x). (b) Test
1 of Stive and Wind (1982); h/hb = 0.882 (o), 0.765 (+), 0.647 (*), 0.529 (x).

is applied here

S=K- (zo+h+z)(z -z) (2.33)

where zo and z, are roughness lengths at the bottom and the free surface, respectively. Given

a surface stress and nearshore beach profile, the momentum equation (2.27) is solved by

equating return flow to the sum of Stoke's drift, -, and net volumetric flux, qner

A comparison of cross-shore velocity variation with depth associated with the linear

and quadratic shear stress relationships is presented in Figures 2.7 and 2.8 for breaking and


S -0.5 +

a o a
0 0.01 0.02 0.03 0.04 0.05 0.08 0.07 0.08 0.09 0.

(b) 0

0.3 -0.5

1 -I
s0.01 0.02 0.03 0.04 0.05 0.06 0.07

Figure 2.6 Variation of turbulent kinetic energy with depth in plunging breakers. (a)
Ting and Kirby (1994); h/hb = 0.929 (+), 0.857 (.), 0.773 (o), 0.675 (x), 0.584 (.).
(b) Test 2 of Stive and Wind (1982); surface, 0.905 (+), 0.709 (), 0.544 (o), 0.381 (x),
0.143 (-).

non-breaking waves, respectively. Bottom streaming is not included in this model. In the

figures, the following wave conditions have been considered: water depth h = 1 m, beach

slope = 0.02, wave height H = 0.78 m, and wave period T = 6 s. In the quadratic relationship,

the bottom roughness is 1 cm for both breaking and non-breaking waves. Since breaking

waves cause a very rough water surface, the surface roughness is taken as half the wave

height for breaking waves and 1 cm for non-breaking waves, respectively. Generally, the

eddy viscosity in the linear stress relationship is difficult to determine and may vary with the

water and wave conditions. For the comparisons presented here, the eddy viscosities are given

S0.4 2...... 0.4 0. ...... .. ..

0 0 0L
-20 0 20 0 0.02 0.04 0 0.2 0.4
stress [N/m^2] eddy viscosity [mA2/s] velocity [m/s]

Figure 2.7 Comparisons of the linear (dashed line) and the quadratic (solid line)shear stress relationships for
breaking waves. Note: the net seaward transport is equal to the shoreward mass transport.

stress [N/m^2]

0 0.001 0.002 0 0.2
eddy viscosity [mA2/s] velocity [m/s]

Figure 2.8 Comparisons of the linear (dashed line) and the quadratic (solid line)shear stress relationships for
non-breaking waves. Note: the net seaward transport is equal to the shoreward mass transport.


by the average values of the quadratic shear stress relationship as 1.23x102 m2/s and

1.22x103 m2/s for the breaking and non-breaking waves, respectively. As shown in Figure

2.7, the magnitudes of velocity predicted by two methods are about the same order, but the

profiles are quite different. The position for the maximum velocity is at an elevation of 50%

and 25% above the bottom in the linear and quadratic relationships, respectively. In both

models, the shear stress varies linearly with the elevation and the surface value is equal to one

half the local wave energy gradient due to the wave breaking; but the bottom shear stress

predicted by the linear model is about three times large as that predicted by the quadratic

model. As shown in Eqs. (2.30) and (2.31) for the linear stress model, the velocity is inversely

proportional to the eddy viscosity and the bottom stress varies directly with eddy viscosity.

It is difficult to determine a proper value of eddy viscosity to satisfy both the velocity

distribution and the bottom shear stress under such conditions. For the non-breaking wave

case shown in Figure 2.8, the two methods yield much closer results. Due to the turbulence

caused by wave breaking, the eddy viscosity is much larger inside the surf zone. Since the

velocity profile is independent of eddy viscosity in the linear model for non-breaking waves,

a better fitting between the two methods for the shear stress can be achieved by decreasing

the eddy viscosity in the linear model. In the following, the predictions of the two methods

are compared with measurements from the literature.

Hansen and Svendsen (1984) measured the velocity distributions for undertow on a

planar bed with a slope of 1:34.3. The incident wave height was 12 cm in the uniform depth

portion of tank with a water depth of 36 cm, and a wave period of 2 seconds. Breaking

occurs at an approximate depth of 20 cm. The velocity distributions were measured in one


position outside the surf zone with a water depth of 22.5 cm and four different positions

inside the surf zone with water depths of 7.6 cm, 9.5 cm, 12 cm and 14.9 cm, respectively.

Figure 2.9 shows the comparison between the measured and the predicted results at

a water depth of 22.5 cm for non-breaking waves. The relative bottom and surface

roughnesses are 0.01 for the quadratic stress model. The eddy viscosity in the linear stress

model is taken as 0.609 cm2/s based on the average of the quadratic model. It appears that

the linear shear stress relationship presents a better velocity agreement than the quadratic.

However the difficulty in determine an approximate eddy viscosity in the linear stress model

should be noticed. Under non-breaking wave condition without winds, the velocity is

independent of eddy viscosity, but the shear stress is linearly proportional to it.

The velocity profiles of Hansen and Svendsen at the four positions inside the surf zone

are compared with the model predictions in Figure 2.10. The h/hb values shown in the figure

are the ratios of the local water depth to the depth at the initial breaking point. The relative

bottom roughnesses, zo/h, are 0.01 for all cases. Due to the wave breaking effects, the surface

roughness is given as half of the local wave height for each position. It appears that the linear

shear stress relationship is not suitable inside the surf zone anymore. The measured velocities

and these predicted by the quadratic shear stress relationship are in quite good agreement

except the case with 14.9 cm of water depth, which is a position very close to the breaking

point with a quite uniform velocity distribution.

Figures 2.11 and 2.12 compare the shear stresses, eddy viscosities and the velocities

computed from the quadratic stress model with those measured in Case 2 of Okayasu et al

(1988). The experiments were carried out with a bed slope of 1:20, incident wave height of

5 0 0.05 0 0.5 1
stress [N/m"2] eddy viscosity [cmA2/s]

0 5
velocity [cm/s]

Figure 2.9 Comparisons of the measurements (circles, Hansen and Svendsen 1984) with the predictions of
the linear (dashed line) and the quadratic (solid line)shear stress relationships for non-breaking waves at water
depth of 22.5 cm.

(b) h=12.0cm, h/hb=0.594

> 0.6

velocity [cm/s]

(c) h=9.5cm, h/hb=0.469

velocity [cm/s]

(d) h=7.6cm, h/hb=0.378

-5 0 5 10 15
velocity [cm/s]

0 5 10 15
velocity [cm/s]

Figure 2.10 Comparisons of the measured velocities (circles, Hansen and
Svendsen 1984) with the predictions of the linear shear stress model (dashed
lines) and the quadratic shear stress model (solid lines). Four different
positions are presented.

(a) h=14.9cm, h/hb=0.738

C a : 0

S0.4 .-- .. 0.4 0*. 0.4 .......... "I*:''-

02 0.2 ( 0.2

0L "- 01---.1 0
-2 0 2 0 5 10 -20 0 20
stress [N/mA2] eddy viscosity [cmA2/s] velocity [cm/s]

Figure 2.11 Comparisons of the measurements (circles, Okayasu et al 1988) with the predictions of the
quadratic (solid line)shear stress relationships for breaking waves at water depth of 6.5 cm and wave height
of 5.1 cm.

stress [N/mA2]

0 2 4 -10 0
eddy viscosity [cm"2/s] velocity [cm/s]

Figure 2.12 Comparisons of the measurements (circles, Okayasu et al 1988) with the predictions of the
quadratic (solid line)shear stress relationships for breaking waves at water depth of 3.0 cm and wave height
of 2.34 cm.


5.63 cm and period of 2 seconds at the uniform depth portion of the tank with a water depth

of 40 cm. The measurements inside the surf zone at two different locations are presented here.

In the experimental data, the Reynold stresses were calculated by Okayasu from the measured

turbulent fluctuating velocity; the mean shear stresses shown in the figures was obtained by

averaging the Reynold stresses over one wave period, and the mean eddy viscosity was

calculated from the mean shear stress and the gradient of the mean velocity. As expected, the

shear stresses varied quite linearly from the bottom to the surface. Since waves broke at the

surface, the eddy viscosity in the upper part of the water was greater than that in the lower

part. The turbulent level of flow decreased from the water surface to the bottom. In both

locations, the measurements and the predictions of the quadratic stress model agree well for

the shear stresses and the velocities. The predictions for the eddy viscosity are acceptable.

2.7 Comparison of Magnitude of Shear Stress Terms

In the field, all the bottom shear stress components mentioned in sections 2.3, 2.4, 2.5

and 2.6 exist however their relative magnitudes differed with some terms being much more

important than others. The momentum flux related shear stresses are remarkably different for

breaking and non-breaking waves. To provide an estimate of the relative magnitudes of these

shear stresses, the following conditions have been considered:

Water depth, h = 1 m, and density, p = 1025 kg/m3.

Beach slope of 0.02.

Wave height, H = 0.78 m, and period T = 6 s.

Mean sand diameter, D = 0.2 mm, and relative density, s = 2.65.

Wind speed at 10 m above the water level, uo = 20 m/s.


Because of the relative strong turbulence levels, the eddy viscosities are larger for breaking

waves than those for non-breaking waves. Under the water and wave conditions listed above,

the shear stresses will be calculated for both breaking and non-breaking wave cases.

(1) Critical shear stress

Generally, the critical shear stress inside the surf zone is much smaller than that

outside the surf zone due to the turbulence caused by wave breaking. Since very few studies

have been conducted under breaking wave conditions, the results from Amos et al. (1988) are

applied to determine the critical shear stress outside the surf zone as

T = 0.04p(s-l)gD = 0.13 N/m2 (2.34)

Inside the surf zone, the critical shear stress is considered as zero.

(2) Gravitation contribution

In Eq. (2.2), the sediment concentration, a, in the equation is assumed to be 0.66,

therefore the gravity induced shear stress is 0.04 N/m2.

(3) Shear stress due to non-linear waves

For the wave conditions given above, the wave length in the deep water, Lo, is 56.2

meters. The corresponding wave steepness, H/Lo, and the relative depth, h/L o, are 0.0139

and 0.0178, respectively. From Figure 2.2, the non-dimensional average bottom shear stress

is 5.4. Considering a Darcy-Weisbach friction coefficient of 0.08, the shear stress will be

given by Eq.(2.3) as 0.94 N/m2 directed shoreward.

(4) Shear stress due to streaming velocity

The shear stress due to the boundary layer streaming velocity is presented by Eq.

(2.23). The negative sign in the equation means the shear stress is acting in the onshore

direction. The bottom eddy viscosities computed by the quadratic shear stress relationship are

applied here for the calculations. For non-breaking wave conditions, the bottom eddy

viscosity is 1.66x104 rd/s, the boundary layer thickness, 6, is calculated as 1.86 cm, and

therefore the shear stress is 2.17 N/m2. Under breaking waves, the bottom eddy viscosity is

6.69x104 m2/s, the corresponding boundary layer thickness and shear stress are calculated as

3.74 cm and 4.36 N/m2, respectively.

(5) Shear stress due to the wave mass flux

The case considered here assumes no net mass flux in the cross-shore direction, which

means the seaward return flow balances the wave mass transport. The bottom shear stresses

are computed by the quadratic stress model as 0.64 N/m2 and 1.38 N/m2 for the non-breaking

and breaking wave conditions, respectively. The difference between the two conditions is

caused by the different eddy viscosities, which result from the different turbulence levels.

(6) Shear stress due to momentum flux

As seen from Eq.(2.31), the contribution of momentum flux is related to the gradient

of wave energy. Outside the surf zone, the change of wave energy can be neglected, therefore

the shear stress due to the momentum flux is almost zero. Inside the surf zone, there is a

substantial wave energy dissipation due to wave breaking. The shear stress caused by the

momentum flux becomes a very important term. For the wave and water conditions listed


above, the bottom stress computed by the quadratic stress model is 4.2 N/m2, which is the

total value presented in Figure 2.7 reduced by the contribution due to the mass flux.

(7) Shear stress due to winds

Winds blowing over water cause a wind shear stress and a slope of water surface. As

a result, a surface flow in the wind direction and a bottom flow in the opposite direction are

generated. The wind shear stress is expressed by a friction factor, f, (assumed as 0.005 here),

and the the wind speed at 10 m elevation

S= fu (2.35)

where Pa is the density of air and equal to 1.20 kg/m3 under the temperature of 20 oC. For the

given wind speed of 20 m/s, the surface stress is calculated as 1.20 N/m2. In the quadratic

stress model, the magnitude of bottom shear stress induced by this wind depends on the wind

direction and wave breaking. An onshore directed wind will cause an additional offshore

directed bottom stress of 0.549 and 0.255 N/m2 for the non-breaking and breaking wave

conditions, respectively. An offshore directed wind will induce an onshore directed bottom

stress of 0.31 and 0.26 N/m2 for the non-breaking and breaking wave conditions, respectively.

A summary of all shear stress terms discussed above is presented in Table 2.1. The

stresses due to the momentum flux transport and streaming velocity are the two dominant

terms. The critical shear stress is quite small compared with other terms. Because winds are

the only source of the surface shear stress outside the surf zone, they have the most significant

effects on bottom stress in this region.

Table 2.1 Constructive and destructive bottom shear stresses.
Magnitude N/m2
Type Description of Force Outside Inside
Surf Zone Surf Zone

Critical Shear Stress Friction 0.13 0
Nonlinear Wave 0.94 0.94

Constructive Streaming Velocity 2.17 4.36
Wind* in Offshore Direction 0.31 0.26

Gravity 0.04 0.04
Mass Flux 0.64 1.38
Momentum Flux 0 4.16

Wind* in Onshore Direction 0.55 0.26

* Wind speed at 10 m above the water level, u0o= 20 m/s.


The "closed loop" cross-shore sediment transport relationships, which are studied in

this chapter, are based on wave energy dissipation and implicitly "capture" the dominant

destructive shear stress mechanism (i.e. wave momentum flux) discussed in Chapter 2. Since

these models do not represent any constructive mechanisms, they would not be expected to

perform well for beach accretion. Based on equilibrium beach profile concepts (Dean 1977),

sediment transport is caused by deviations of a beach from its equilibrium form. According

to scale analysis, the linear transport relationship proposed by Moore (1982) should be

modified and a non-linear model is developed. In the non-linear model, the transport rate is

proportional to the cube of the difference of the local wave energy dissipation per unit volume

from the equilibrium value at each location across the surf zone. It appears that the non-

linearity of this transport relationship can reasonably explain the time scale difference in the

beach profile evolution in different experiments. A finite difference method is applied to solve

the transport and the continuity equations numerically. Beach profile response at different

times is determined by the numerical solutions.

3.1 Equilibrium Beach Profile

An equilibrium beach profile represents a dynamic balance of destructive and

constructive forces acting on the beach. A change of these two competing types of forces will


result in a disequilibrium. Considering wave energy dissipation per unit water volume to

represent the dominant destructive force, Dean (1977) has proposed that a sediment of a

given size will be stable in the presence of a particular level of wave energy dissipation per

unit volume, D.. This can be expressed in terms of wave energy conservation as

1 d(Eg)
D- (3.1)
h dy

in which y is the shore-normal coordinate directed offshore, E is the wave energy, and Cg is

the wave group velocity. As a first approximation, D. is assumed to be a function only of

sediment size (Moore, 1982), or equivalent sediment fall velocity (Dean, 1987). With linear

wave theory and shallow water assumption, Eq. (3.1) can be integrated to

24D 2
h = ypg23= Ay3 (3.2)
5pg /iK2)

where K is the ratio of breaking wave height to water depth, and A is defined as a profile scale

parameter. The relationship between A values and the median sediment size is shown in Table

3.1 (Dean and Dalrymple 1996). Since the profile scale parameter, A, is only a function of

sediment size, wave conditions are not included in the above relationship.

The two-thirds power law profile formula of Eq. (3.2) was first empirically identified

by Bruun (1954) as an appropriate representation of the profiles in a field study of beach

profiles at Monterey Bay, CA and along the Denmark coast. Dean (1977) examined 502

beach profiles from the east coast and Gulf shoreline of the United States by a least square


fit to each profile with a generalized power law profile, h = Ay m. It was found that the

average value of the exponent was 0.66, in very good accord with the result derived in Eq.

(3.2). Hughes and Chiu (1978) carried out a study of beach profiles and associated sediment

characteristics at different locations in the state of Florida and Lake Michigan and found that

Eq. (3.2) described the beach profiles reasonably well.

Table 3.1 Summary of Recommended A values, units of A parameter are min (after
Dean and Dalrymple 1996).

D 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
0.1 0.063 0.067 0.071 0.076 0.080 0.084 0.087 0.090 0.094 0.097
0.2 0.100 0.103 0.106 0.109 0.112 0.115 0.117 0.119 0.121 0.123
0.3 0.125 0.127 0.129 0.131 0.133 0.135 0.137 0.139 0.141 0.143
0.4 0.145 0.147 0.148 0.150 0.151 0.153 0.155 0.156 0.158 0.159
0.5 0.161 0.162 0.163 0.165 0.166 0.167 0.168 0.169 0.171 0.172
0.6 0.173 0.174 0.175 0.177 0.178 0.179 0.180 0.181 0.183 0.184
0.7 0.185 0.186 0.187 0.188 0.189 0.190 0.190 0.191 0.192 0.193
0.8 0.194 0.195 0.196 0.196 0.197 0.198 0.199 0.200 0.200 0.201
0.9 0.202 0.203 0.204 0.204 0.205 0.206 0.207 0.208 0.208 0.209
1.0 0.210 0.211 0.212 0.212 0.213 0.214 0.215 0.216 0.216 0.217

Two disadvantages of Eq. (3.2) are the infinite beach slope at the water line and the

monotonic form of the profile. The first shortcoming is overcome by including gravity as a

significant destructive force when a profile becomes steep. In this case, Eq. (3.2) is modified

to include the beach face slope, mo,

h h 3/2
y = O- + (3.3)
mo A

Figure 3.1 Definition sketch for Inman's curve fitting. Crosses denote the origin for bar-berm
(dotted) and shorerise (dashed) curves, with y coordinates, y,, Y2, and vertical coordinates,
hi, h2 (from Inman et al. 1993).

Larson (1988) modified Dean's approach and developed an equation of similar form by

replacing the spilling breaker assumption with the wave breaking model of Dally et al. (1985)

and retaining the requirement of uniform wave energy dissipation per unit volume.

Inman et al. (1993) developed an equilibrium beach profile that treats a profile as two

parts, the inner (bar-berm) and the outer (shorerise) portions. The two portions are matched

at the breakpoint-bar and fit by curves of the form

h = By m (3.4)

The coordinates of Inman's curve fitting are shown in Figure 3.1 with the subscripts 1 and

2 corresponding to bar-berm and shorerise curves respectively. In addition to B and m in Eq.

(3.4), the origins of two curves must be determined from the profile data. In total, seven


variables, yl, z1, B, m1, y2, B2,and m2 are required to fit a profile. This method is

diagnostic and generally useful for a beach with measured data available. Comparatively, the

method described by Eq. (3.2) is prognostic and needs only a description of sediment size.

3.2 Scale Analysis

A beach which is steeper than equilibrium has a smaller volume of water over which

the incident wave energy is dissipated. This causes the energy dissipation per unit volume to

be greater than the equilibrium value. As a result, the total destructive forces are greater than

the constructive forces. The profile will respond to the imbalance of forces through

redistribution of the sediments. Over time, sand will be carried from onshore to offshore and

deposited near the breakpoint, resulting in a milder profile which approaches the condition

of uniform wave energy dissipation per unit volume. Similar to this process, for a beach with

a milder slope than equilibrium, sediment will be moved from offshore to onshore.

Based on these concepts, Kriebel and Dean (1985) proposed that the cross-shore

sediment transport rate per unit beach width, Q, could be approximated according to the

deviation of actual wave energy dissipation per unit volume from the equilibrium at a

particular location across surf zone as

Q = K(D -D,) (3.5)

in which, D represents the actual total destructive force,

5 tg3/2
D= 5pg32 Kh /- (3.6)
24 ay


The transport parameter, K, was considered as a dimensional constant. The following scaling

relationship is established from Eq. (3.5)

[K(D -D,)]odel
Q [KD = = Kr(D-D,)r (3.7)
[ K(D -D ,) ]prototype

where the subscript r denotes scale ratios. For an undistorted model with same density fluid,

according to the definition of D, the disequilibrium scale (D-D.),can be expressed in terms

of the length scale, L,

(D-D.), = (3.8)

On the other hand, according to the Froude relationship the time scale, Tr, is expressed as

Tr = Vr (3.9)

Following this relationship, the cross-shore sediment transport per unit beach width should

be scaled as

Qr Lr2 (3.10)

This equation provides a basis for evaluation of transport models. If K is independent of the

length scale, Eq. (3.7) does not provide a valid scaling of transport.

Based on field observations, Dean (1973) hypothesized that sediment was suspended

during the wave crest phase position and if the fall time were less or greater than one half


wave period, the net transport would be landward or seaward, resulting in a bar formation for

the latter case. This mechanism would exist in the region of wave breaking. Considering the

height of suspended sediment to be proportional to the breaking wave height, Hb, resulted
in identification so-called fall velocity parameter, -. By examining small scale wave data
for which only deep water reference wave heights, Ho, were available, the following

relationship for net seaward sediment transport and bar formation was found

> 0.85 (3.11)

Later, Kriebel et al. (1986) examined only prototype and large scale laboratory data and found

a constant of 2.8 instead of 0.85 in Eq. (3.5). Kraus (1988) examined only large tank data and

proposed the following relationship for bar formation

H Ho
o 0.0007 (3.12)
LO [wT)

Several model studies have confirmed that -- is a valid modeling parameter such that if this
parameter is the same in model and prototype, they are scaled versions of each other and the

fall velocity is scaled by the length scale as w, = JL-. Two examples from Hughes et al. (1990)

and Kriebel et al. (1986) are presented in Figures 3.2 and 3.3, respectively.

It is of interest to develop and test a transport model which can ensure convergence

to the target (equilibrium) beach profile and also satisfy the scale relationship given by

Eq.(3.10). One approach is to consider the following form for the sediment transport


370 WAVES Proto oshed

S~-~S Ois ff a 0.484 a

.10 0 10 '20 30 40 50
L_.. i"* i i ioi c |odei Solid
S15so WAVES Prot Dshd

~.lS iFP R, 0.444 a1

-*1 0 10 20 30 40 SO

Figure 3.2 Comparison of beach profiles from medium and large scale wave tank, Scaling
according to (Ho/wT),=1 (from Hughes et al. 1990). Note concrete bore underlying initial
sloping sand beach.

Q = K(D-D,) ID-D,In-1 (3.13)

Which results in the scale relationship

Qr =K(D-D .)rD-D I (3.14)

Substituting Eqs. (3.8) and (3.10) to Eq. (3.14) yields

n 3
KL = L 2 (3.15)

If K is only a function of the fall velocity parameter, Kr equals unity and n=3 is determined

such that both the scaling relationship and convergence to the equilibrium profile can be

satisfied. Otherwise, if Kr is related to the length scale as Kr = Lm, the relationship, m+- = ,
2 2

Me- MdW 1G.T7a,.LTgo
Praftye H*L6a. TP5A*

Figure 3.3 Comparison of beach profile from medium and large scale wave
tank, Scaling according to (Hd/wT)r=1 (from Kriebel et al. 1986).

S500. -- Dune without foreshore
--. Case 300
S300. O

5. 10. 15. 20. 25. 30. 35. 40. 45. 50.
Time (hours)
(a) Eroded volumes versus time for the German "dune without foreshore" case and
Savill's Case 300.


~ L,-LO --;-,..-*, ,,,---.-o,, --.",*,,...
aSSss saw*
s0o 0ooo 505o 2Sooo 1 0oo
Time (hours)

(b) (LI L2) versus time for Swart's flume B Case with 86 = 0.15m and 6 2 = 0.10m

Figure 3.4 Time history of beach evolution in three different experiments.

must hold to satisfy Eq. (3.10). In this case, there is an infinite number of combinations of m

and n would satisfy the scaling requirement.

3.3 Discussion of Transport Relationship

Sediment transport in the nearshore region is a complicated process. Under different

conditions, some beaches reach equilibrium very fast, but others may vary slowly. The time

scale of beach evolution may vary from tens to thousands of hours for different experiments.

Figure 3.4 presents the results from the German "dune without foreshore" case (Dette and

Uliczka 1987), Saville's Case 300 (Kraus 1988), and Flume B case in Swart's experiments


(Swart 1974a). In these tests, the eroded volume at any time is determined as the cumulative

volume of material displaced between initial and current profiles, and L, and I2 are

schematized profile lengths at water depth 6, and (68 +86), respectively. Comparing these

three cases in Figure 3.4, a question arises as to the causes of the greatly different profile

response times.

In an attempt to understand the causes of the different time scale, we examine the

following equation

dx =K(x-x,) Ix-x, n-1 (3.16)

where x. is the equilibrium value of x. Eq. (3.16) is reminiscent of the transport relationship

discussed above. Non-dimensionalizing with x '= and t'= t yields
x. Kx-1

dx- (x -1) x'- 11 n-1 (3.17)

with the initial condition x'(t'=0) =x0, the normalized solution of Eq. (3.17) is given by

x/- 1 _t'
x e for n=
x'- 1 1 for n*(3.18)
for nr l

0 [(n-1)|xo-ll" 1|- +1 -1

Figure 3.5 presents the results of versus t' for n=l, 2 and 3, and x'=2 and 10
respectively. It appears that the time scale of the linear system (n=l) is independent of the

initial conditions and the two lines in Figure 3.5 (a) are coincident. However, for the nonlinear

(a) n = 1

(b) n = 2

2 4 6 8
(c) n = 3

2 4 6 8
normalized time, t'

Figure 3.5 The solution of Eq. (3.16).



- 0.6

a 0.4

a 0.2




i5i 0.4



^ 0.8


~S 0.4

" 0.2



systems, the initial conditions affect the time scale by the factor (n-1) |x -l|n-1. As n

increases, this factor becomes more and more significant.

As demonstrated in the solution of Eq. (3.16), a nonlinear transport equation provides

an explanation for the range of time scales observed in laboratory experiments of beach profile

evolution. In such nonlinear systems, the greater variation of initial condition from

equilibrium corresponds to smaller time scales of profile response. Returning to the question

raised about Figure 3.4, it appears the time scale differences in the three cases may be caused

by the different initial conditions and non-linearity of the transport relationship. Among three

cases, the "dune without foreshore" had a sand size of 0.33 mm and a initial slope of 1:4 with

a wave height of 1.5 m and period of 6 s, and its initial condition is farthest from its

equilibrium and the beach takes the shortest time to approach relative equilibrium. For Case

300, the sand size was 0.22 mm and the initial profile slope was 1:15 with a wave height of

1.68 m and period of 11.33 s. Swart's flume B Case had a sand size of 0.17 mm and an initial

slope of 1:10 with a wave height of 0.07 m and period of 1.04 s. Its initial condition is

hypothesized closest to the equilibrium and takes the longest time to arrive at relative

equilibrium. In the next chapter, the transport relationship, Eq. (3.13) with n=l and n=3, is

applied to three sets of large tank experiment data. Time dependent profile response is then

determined by the numerical solution of the transport equation and continuity equation, which


ay- -_ Q (3.19)
at ah



yi hn = A(yYmsi)23
---Y4.1 --

Figure 3.6 Grid with h and t as the independent variables and y dependent.

3.4 Modeling Process

3.4.1 Numerical Method

The sediment transport and continuity equations are solved by a finite difference

method. The computational physical domain with the staggered grid is presented in Figure

3.6. The cells are finite increments of the depth variable, h. Thus the depth is the independent

variable and the dependent variable is the distance, y, which varies with time. For each time

step, the transport and continuity equations are solved simultaneously. After applying Eq.

(3.6) as the formula of the actual wave energy dissipation per unit water column, the finite

difference form of the transport equation (3.13) at the (m+6)th time step, t=(m+6)-At,


P5 32k 2(hi3" -hi) 325 pg k 2(hn/-hi )
Qm+ = K 24 6 -D *24 +6 + -D (3.20)
e tm+6 m+6 m+6 m+6 *
Yi -Yi-1 Yi -Yi-1

where the superscript m+6 denotes the values at (m+8)th time step, and 6 is a value between


0 and 1. After rewriting the distance variable at the time, t=(m+l)-At, as yml =yim+Ayim+l

Eq. (3.20) becomes


-pg 32k 2(h3/2 -h.

m m
yi -yi-I



A m+ A m+1
For a small time increment At, the value e is much less than 1. Therefore
m m
Yi -yi-1
Eq. (3.21) is linearized as

5 p2-h) 5 2 (h 3/2_h3 n
QimK1 24 32 (l-) -D,.24 2 (1-e) -D (3.22)
Yi -yi-1 yi -Yi-1

When n is an odd number, the absolute sign in the equation can be dropped. Neglecting higher

order terms, the linearized finite difference transport equation is expressed as

M A-, i mAl
Ai*Ayi1m+ + Bi, Qm. + CiiAy = D

5P 3sk2(h3/ -hi n-I 5 P 32k2(h -hi3)
Ai*=nK 24 -D, ) *
Yi -Yi-1 Yi -Yi-1
B' = 1
C = -Ai
[5 3/22(h 3/2h 3/2 n
K 24
D* = i -- -D
yi -Yi-I



For the grid shown in Figure 3.6, the discrete form of the continuity equation (3.19) is

1 1
m+-- m+-
m+1 2 2
yi Qi+l Q (3.24)
At Ah

1+- 1
Recalling Qi 2 IQim + Qi,) Eq. (3.24) is written as

Aiimtl +BiAyimt + C-m+l = Di (3.25)
AQ,1 ++BAym+l+ C, =D,

A At
S 2Ah
B.= 1
Ci = -A.
D = At (Q m Qin)
2Ah il-I

Eqs (3.23) and (3.25) are tridiagonal equations and can be solved simultaneously by the

standard double sweep method. At each computation time, the active profile domain is

determined by the instantaneous incoming wave and water level conditions. Assuming this

active domain is the region of hib h< hhie, the boundary conditions, Qib = Qie = 0 can be

applied, which leaves 2x(ie-ib) variables as unknowns. The continuity equation (3.25) is

solved from i=ib to i=ie-l, and the transport equation (3.23) is solved from i=ib+l to

i=ie. They provide total 2x(ie-ib) equations. Starting with boundary condition Fi = Qie and

E i= 0, four new sets of variables are calculated as

A. Di-CiFi1 Ai Di*-Ci*Fi
E= --, F.= E. = Fi = D (3.26)
Bi+CiEi, Bi+CiEi* Bi* +Ci* Ei Bi +C* E
i i +1 +1Ci


These coefficients are determined sequentially from i = ie to i = ib. Therefore the current

transport rates and the profile changes are determined as

Ayi = EiQ +Fi ib (3.27)
Qi = Ei* Ay+Fi* ib+1li ie-1 (

3.4.2 Wave Run-up and Set-up

The active profile considered in the numerical model is from the wave run-up limit to

the wave breaking point. Outside this active region, the net sediment transport does not result

in profile changes. At each time step, the water level is determined by the sum of storm surge,

tide and wave set-up, and the elevation contours used in the transport equation are modified

according to the change of water level.

The wave induced set-up in the surf zone is calculated according to the balance of

pressure gradient and radiation stress. Inside the surf zone, based on the linear wave theory,

the set-up, l, is given (Dean and Dalrymple, 1991) as

(h) = + (hb-h) (3.28)

where K is the ratio of breaking wave height to the breaking water depth, and rb and hb are

the set-down and the water depth at the breaking point, respectively. For shallow water, the

value of tb is given by Longuet-Higgins and Stewart (1964) as

ib- -- (3.29)


In the numerical model, storm surge and tide are applied to the whole surf zone uniformly,

but the set-up is added according to Eq. (3.28) and varies with water depth. At the shoreline,

the set-up, %0 calculated from Eq. (3.28) is

e -0.0625 K2
-0 i- hb (3.30)

where e= According to linear wave theory, we have K = 0.78, and = 0.18 hb is

In the study of seawalls and breakwaters design, Hunt (1958) proposed wave run-up,

R, as a function of bed slope, wave height, and wave steepness, which can be expressed as

R =FHb (3.31)

where R is the run-up height measured vertically upward from storm water level, FR is a non-

dimensional coefficient and is approximately 1, m is the average bed slope from the run-up

limit to the breaking point, and Lo is the deep water wave length. Since the wave run-up

calculated by Eq.(3.31) is measured from the storm water level, after including wave set-up

in water level change, the wave run-up is established as the value calculated from Eq. (3.31)

by subtracting the value of wave set-up given in Eq. (3.30).

3.4.3 Dune. Shoreline and Offshore Slopes

In the numerical model, three characteristic slopes must be specified. As shown in

Figure 3.7, they are dune slope, md, shoreline slope, m and offshore slope, mff. The dune

Initial Profile

Dune Scarp
Run-up (m, / S.W.L.

Final Profile Offshore slope (m,,)

Figure 3.7 Definition sketch for dune, shoreline, and offshore slopes.

slope, which is defined as the averaged dune scarp slope after erosion, is the maximum slope

that the profile is allowed to achieve. If the profile reaches an angle steeper than the dune

slope at the end of any time step, avalanching occurs and the profile is smoothed to the dune


The shoreline slope, which is the anticipated profile slope between the shoreline and

the run-up limit, controls the profile evolution between the shoreline and the run-up limit. In

this region, the profile change is caused by the uprush and backrush, and the transport

equation discussed in section 3.2 cannot be applied. After each time step, the actual slope is

compared with the shoreline slope. If the actual profile is steeper than the given shoreline

slope at a point, the following adjustment is made for the distance variables of two adjacent


Ayi =-0.5 Yi-Yi-1 1h/ -e


Ayi- = 0.5 Yyi- hi-h 1 -e K)

where the subscript i and i-1 denote the indexes of cells, m, is the shoreline slope, and Ky is

a decay parameter and set somewhat arbitrarily to 10 hr- here, which represents a "folding

time" of 0.1 hours.

Sand conservation across the profile is ensured by the continuity equation, so that the

volume eroded from the beach face must be deposited offshore. Since the profile change ends

at the wave breaking point, there is an unrealistic sudden vertical change in the profile shape.

To overcome this problem, the offshore slope is introduced to control the slope at the end of

the deposited volume. Under most cases, the value of this slope is between 0.1 to 0.2.

3.4.4 Random Wave Generation

Under random wave conditions, the joint distribution of wave periods and heights

developed by Longuet-Higgins (1983) is applied to generate wave height and period series

in the time domain. It is convenient to non-dimensionalize wave height and period with the

quantities related to the spectral density, E(o). The average frequency is defined as

J = mlImo (3.33)

where mn denotes the nth moment of the spectral density,

m= on E(o)do (3.34)


A spectral width parameter v can then be defined in terms of the variance of E(o) about the

mean as

2 f (o-)2E(o) do mm2
C22 M 2
v_=_=__ _-_ _-- 1 ( 3 .3 5 )
o 0mo m

For a narrow-banded spectral random wave system, v is much less than 1. Under severe wave

conditions, v will increase. In practice, v values are between 0 to 0.6 for ocean waves. After

normalizing the wave height, H, and the wave period, T, by

H a
R 2 = (3.36)

the joint probability density function, p, is expressed as

2 -R2 +( 11 )212
p(R,-) = L(v) 2 -- e R (3.37)

where L(v) = 1+- is a normalization factor introduced to take account of the

fact that only positive values of I are considered.

The marginal probability density of wave height R is found by integrating p(R,') with

respect to 0 < T < c, that is

p(R) = L(v) [1 +erf(R/v)]Re -R2 (3.38)

where erf(Rv)= f e -t2dt is the error function. Eq. (3.38) states that the density of R

0.8 ---.. nu= 0.2
Snu =0.4

0.6 ........... nu = 0.6
o0 nuO\.



0 0.5 1 1.5 2 2.5 3
non-dimensional height

Figure 3.8 The probability density of wave height R when v = 0.2, 0.4 and 0.6
compared with the Rayleigh distribution (v = 0).

is almost Rayleigh distributed, but must be corrected by the factor L(v) [1 +erf(Rlv)]. For

large waves and small v values, the correction is exponentially small. Some examples are

presented in Figure 3.8. For the current application, the Rayleigh distribution is used as an

approximation of Eq. (3.38). Therefore, the cumulative probability of the wave height


P(R) = 1-e-R (3.39)

and the root mean square wave height, I-I, is related to the spectral density as Hrm, = 2 /2mo.

The distribution of wave period at fixed values of the wave height R is found by

dividing p(R,t) in Eq.(3.37) by p(R) in Eq. (38), that is

2 R 1 2R2
p('c/R) = ex 1-1 v (3.40)
rv[l+erf(R/v)] T2 T V2

Integrating p(r/R) with respect to t results in the cumulative probability

1 +tanh 1.2R (

To simulate an irregular wave condition, a random number P, uniformly distributed

from 0 to 1 is generated. A random non-dimensional wave height, R, is obtained from Eq.


R) = -ln(1-P) (3.43)

To obtain the corresponding wave period, a value of v is assumed according to the severity

of wave conditions. A second random number P2 with uniform density distribution over (0,1),
is generated. Substituting P2 and R(1 into Eq. (3.42) yields
is generated. Substituting P, and R i into Eq. (3.42) yields



0 0.5 1 1.5 2 2.5 3 3.5 4

Figure 3.9 Comparison of the error function erf(x) with tanh(1.2x).

P2 1+tanh
1 v In I v (3.44)
2.4R +tanh 1.2R,
2-P2 1 +tanh--1

Given a root mean square wave height, H-, and an average wave period, T, the dimensional

wave height and period are

H, =H R, T, = T (3.45)

The random wave heights and periods generated numerically are compared with Longuet-

Higgins's joint density distribution in Figure 3.10 and Figure 3.11 for narrow band (v = 0.1)

and wide band (v = 0.5) spectra, respectively. It appears that the numerical results agree with

Longuet-Higgins theory very well.

(a) Scatter diagram of numerical results

0 0.5 1 1.5 2 2.5
non-dimensional period

Contours of L-H Eq (dashed) and numerical

results (solid)

0 0.5 1 1.5 2 2.5
non-dimensional period

Figure 3.10 Comparisons of the numerical results with the density
function (Eq. (3.37)) of Longuet-Higgins for v = 0.1.


I 2
"5 1.5



(a) Scatter diagram of numerical results



, 1.5-

1 '
0.5 .

0 0.5
0 0.5

1 1.5 2 2.5 3
non-dimensional period

(b) Contours of L-H Eq (dashed) and numerical
3 1 1i

results (solid)

1 1.5 2
non-dimensional period

Figure 3.11 Comparisons of the numerical results with the density
function (Eq. (3.37)) of Longuet-Higgins for v = 0.5.


4.1 Introduction

The cross-shore sediment relationships discussed in Chapter 3 have two coefficients

requiring quantification. They are the transport coefficient, K and the exponent, n. Based on

the scaling analysis, n=3 was proposed in the last chapter. In this section, the proposed

nonlinear transport model with n=3 will be compared with the linear transport relationship,

n=l, for seven laboratory experiments. Calibrations of the two models are accomplished by

a series of simulations in which a trial K value is used to simulate profile development for each

experiment. The eroded volume at any time is defined as the cumulative volume of material

eroded between the initial and current profiles. The errors between the predicted and

observed eroded volume are obtained at the times for which profile information is available.

The best-fit K value is determined as the value yielding the overall least square error. The

calibrations utilize 3 sets of different laboratory data, involving a total of seven large scale

wave tank experiments. Among these, five cases were carried out with monochromatic wave

conditions, and the other two were conducted under random waves. According to the wave

tank facilities, the seven experiments are divided into three groups: Saville's experiments

which were operated by the Beach Erosion Board in the "Large Wave Tank" in 1956-1957

and 1962 (Saville 1957, Saville and Watts 1969, and Kraus 1988); the German "Large Wave


Flume" experiments, which were carried out in Hannover in 1987 (Dette and Uliczka 1987,

Dette and Rahlf 1992, and Dette et al. 1992); and SUPERTANK laboratory data collection

project, which was conducted at O.H. Hinsdale Wave Research Laboratory, Oregon State

University, in 1991 (Kraus and Smith 1994, and Smith and Kraus 1995).

4.2 Saville's Experiments

4.2.1 General Description

Saville's experiments (Saville and Watts 1969, and Kraus 1988) described here were

conducted during 1956-1957 and 1962 in the "Large Wave Tank" located at Dalecarlia

Reservation in Washington, DC. The concrete tank was 194 m long, 4.6 m wide, and 6.1 m

deep. Prior to the start of most new experiments, sand was emplaced to form a 1 on 15 slope.

A fine sand of median diameter 0.22 mm was used in the experiment performed during 1956-

1957, and a coarse sand of median diameter 0.40 mm was used in the experiment performed

in 1962. Since beach erosion is the process of interest in this study, only the set with a sand

size of 0.22 mm will be discussed here.

Seven experiments were conducted under constant wave conditions with sand size of

0.22 mm. They were referred as Case 100, 200, 300, 400, 500, 600, and 700. The

corresponding fall velocity for the sand at a temperature of 20 C is 3 cm/s and the profile

scale parameter, A, is 0.106 m1i3. Since the profile eroded back to the end of the tank during

the experiment in Case 100, this case will not be discussed. The wave and water level

conditions for the other six cases are listed in Table 4.1. The initial and the final profiles for

each of these experiments are presented in Figure 4.1 in the order of the fall velocity

parameter, H/wT. Since the wave heights were small in Case 600 and Case 200, there was


Table 4.1 Wave height, period, and water depth in horizontal section of the tank.

Case No. Wave Height (m) Wave Period (s) Water Depth (m) H/wT

200 0.55 11.33 4.57 1.62

300 1.68 11.33 4.27 4.94

400 1.62 5.60 4.42 9.64

500 1.52 3.75 4.57 13.51

600 0.61 16.00 4.57 1.27

700 1.62 16.00 4.11 3.38

almost no beach erosion. In the other four cases, beaches eroded and offshore bars formed

near the break points. It appears that as the fall velocity parameter increases the offshore bar

becomes more and more significant. In Case 500, the height of the offshore bar became about

1.5 m. According to observations recorded during the experiment, a second wave breaking

position occurred in this case. The time histories of observed eroded volume for the four

erosion cases are shown in Figure 4.2. It appears that the variation of eroded volume with

time is not consistent in Case 700. Although constant wave conditions were run during the

experiment, the eroded volume reached a maximum at 30 hours instead of increasing

monotonically. Therefore, for the calibration of CROSS model, only Case 300, Case 400 and

Case 500 will be used.

4.2.2 Calibration

The sediment relationship, Eq. (3.13) with both n=l and n=3, are applied. The active

water depth is determined as the breaking water depth, which is 1.28 times of the breaking

wave height according to McCowan (1894) theory. Based on the wave heights listed in Table

Case 600 H/wT = 1.27

20 .. ... .. ... .. ... . . . . .. . . .

-2. .. ... .. ... ...


-40 -20 0 20 40 60 80
distance [m]

Case 700 H/wT = 3.38

2 ... .... ... ....... ....... ....... ......

o .......... ............. ....... .......

- 4 ..... ....... ...... ... ."' ......


-40 -20 0 20 40 60 80
distance [m]

Case 400 H/wT = 9.64
: : : 1 :
.. . .. :. . ... .. ... . .. :. . . .. .

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

: : :
.... .
; .-

-40 -20 0 20 40 60 80
distance [m]

Case 200 H/wT = 1.62

2 ...... .... .. ... .. .. ..... .. ... ... .... .

0 ........ ...... '" ...... .... ............ .......



-40 -20 0 20 40 60 80
distance [m]

Case 300 H/wT = 4.94

0 ...... ... ...... ....... ........ ...... .......


-2 ... .


-40 -20 0 20 40 60 80
distance [m]

Case 500 H/wT = 13.51

2 ...... ..... ....... ....... ....... .. .......

0 ..... .. ... ....... .............. .......

-2 ...... ....... ......... .... .. ... ......

-4 .....

-40 -20 0 20 40 60 80
distance [m]

Figure 4.1 The initial (dashed) and the finial (solid) profiles of Cases 200, 300, 400, 500, 600,
and 700 in Saville's experiments.


Case 700 H/wT = 3.38

2 5 ........ ... .. 0 ......... ..... .......
200 0 0
2 0 ......... ....... ........ ......... ...........

15 0......

5 ..... ........ ........ ......... ....

0 20 40 60 80 100
time [hr]

Case 400 H/wT = 9.64

25 ........................ 0 ..
20 ........ .. ... .......... ............

10 0
10 5 -0 ....... ....................... ..........

5 0 .. ....................... ..........

10 20
time [hr]

30 40

Case 300 H/wT = 4.94
25 0.........--- ............



0 10 20 30 40 5(
time [hr]

Case 500 H/wT= 13.51
O n ________________________-- T-m -- -

0 20 40 60
time [hr]

80 100

Figure 4.2 The time histories of the observed eroded volume for Cases 700, 300, 400 and

4.1, the active water depths are 2.15 m, 2.08 m and 1.95 m for Case 300, Case 400 and Case

500, respectively. The eroded volume error is defined as

Err -1 [(Vol(t -VOlm(t]2
n j=


where the Vol, and Volm are the predicted and measured eroded volumes, respectively, and

ti denotes the time at which the measurements are available. In each case, the dune, shoreline

and offshore slopes used in the numerical model are determined according to the observed

......... ......... o .. .: ... ........ .......
0 0

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

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


, A



Table 4.2 The numerical model inputs and best-fit results for Saville's experiments.

Case Slope Best-fit K Error* [m2]

No. dune shoreline offshore n=3 [m8s2/N3] n=l [m4/N] n=3 n=l

300 0.50 0.15 0.20 8.55x10-o1 3.45x10-6 2.08 3.37

400 1.00 0.17 0.15 7.97x10-'0 3.71x10-6 1.78 2.40

500 0.50 0.13 0.30 5.77x10-'0 1.57x10-6 3.99 1.88

Error represents the root mean square between predicted and observed eroded volume
as shown in Eq. (4.1).



0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
K [m^8*s^2/N^3] 10-9

Figure 4.3 Root mean square error of eroded volume as a function of transport coefficient
K for the non-linear model (n=3) simulation of Case 300.

final profile. The values of these slopes are shown in Table 4.2. The calibration process for

the non-linear transport model simulation of Case 300 is shown in Figure 4.3. Based on the

results, a best-fit K value of 8.55x10-10 is adopted which gives the overall minimum error for

the eroded volume. The best-fit K values and the corresponding errors of eroded volume for

the three cases are presented in Table 4.2. The best-fit K values vary from 5.77x10'1 to

8.55x101-, a factor of 1.48 for n=3. The corresponding factor for the best-fit K values for

n=l is 2.36. Comparisons of predicted and measured eroded volumes are shown in Figure 4.4

4U i I i I I I I

01 -Case 300 ..--

E o
S20 0o -
20 0 measured
- 10 -- predicted by n=3
S- - .. predicted by n=1

0 5 10 15 20 25 30 35 40 45 50
time [hr]



5 10 15 20
time [hr]

25 30 35 40

10 20 30 40 50 60 70 80 90 100
time [hr]

Figure 4.4 Comparisons of prediction with best-fit K value to observed eroded volumes
for Cases 300, 400 and 500 in Saville's experiments.

Case 400 ,.---

0 measured
- predicted by n=3
- - predicted by n=1



1r V4 i I -- I

I i


for the predictions of both the linear (n=l) and the non-linear (n=3) sediment transport

relationships. It appears that the non-linear transport relationship with n=3 provides a better

fit for Case 300 and Case 400, while the linear transport relationship, n=l, predicts better

results for Case 500. As discussed before, Case 500 had a very significant offshore bar

(Figure 4.1) formed during beach erosion. Waves first broke at the offshore bar, then

reshaped in the offshore trough area and broke again in the nearshore region. It is not

appropriate to calculate wave energy dissipation per unit volume according to the offshore

incident wave height. The vortex caused by the first wave breaking also played an important

role in the sediment transport in the area shoreward of the bar.

The beach profile evolutions predicted by the non-linear (n=3) and the linear (n=l)

transport relationship are compared in Figures 4.5, 4.6 and 4.7 for Cases 300, 400, and 500,

respectively. In each case, the profiles at four different times are compared. Among three

cases, the proposed non-linear relationship presents overall better agreement for the profile

evolution in Case 300 and Case 400. Owing to the significant offshore bar in Case 500, the

numerical models which predict a smooth monotonic profile form have some difficulties in

representing the profile around the offshore trough and bar area.

4.3 German "Large Wave Flume"

4.3.1 General

The German "large wave flume" is 324 meters long, 7 meters deep and 5 meters wide.

Three series of experiments were carried out in the flume (Dette and Uliczka 1987, Dette and

Rahlf 1992, and Dette et al. 1992). Two of these experiments had the same constant wave

conditions and different initial profiles. Regular waves with a wave height of 1.5 meters and


a ,
aM -4





Z -4



20 40
offshore distance [m]

Figure 4.5 Case 300 from Saville's experiments. Comparisons of predicted to observed
profiles at different times for the best-fit K values for each transport relationships.

profiles at 5 hours

observed '
......... predicted by n=1
predicted by n=3

-20 0 20 40 60 8C

profiles at 10 hours

-20 0 20 40 60 8C

profiles at 30 hours

-20 0 20 40 60 8C

.profiles at 50 hours

| N.




z -4


2 profiles at 5 hours

0 -

a -2 -. initial profile. .
) -- observed
5 -4 ... .... predicted by n=1
- predicted by n=3
--6 I I I I
-20 0 20 40 60 80

2- profiles at 10 hours

0 -2-

-6 I I I I
-20 0 20 40 60 80

2- profiles at 30 hours

S. 0
-0 --


-6 II
-20 0 20 40 60 80

2 profiles at 40 hours

) -4
-6 I
-20 0 20 40 60 80
offshore distance [m]

Figure 4.6 Case 400 from Saville's experiments. Comparisons of predicted to observed
profiles at different times for the best-fit K values for each transport relationships.

2 profiles at 10 hours


-2 -- initial profile
-4 ......... predicted by n=1
- predicted by n=3
-6 I I I
-20 0 20 40 60 8C

2 profiles at 30 hours

0 -

-2 -


-6 i I I I
-20 0 20 40 60 8C

2- profiles at 60 hours

0 -



-6 I


-4- -


20 40
offshore distance [m]

Figure 4.7 Case 500 from Saville's experiments. Comparisons of predicted to observed
profiles at different times for the best-fit K values for each transport relationships.




profiles at 100 hours



a period of 6 seconds were generated in a water depth of 5 meters. The sand used for both

experiments had a mean diameter of 0.33 mm, which corresponding to a fall velocity of 5

cm/s at a temperature of 20 C and a profile scale parameter, A, of 0.131 mi"3. Two initial

profiles, termed "dune without foreshore" and "dune with foreshore", were used. The "dune

without foreshore" had a dune crest of 2 meters above still water level and a seaward slope

of 1:4 down to the channel floor. The "dune with foreshore" had a slope of 1:4 from the dune

crest of 2 meters above still water level to 1 meter water depth followed by a slope of 1:20

down to the channel floor. The third experiment, a "dune without foreshore" was conducted

with the same sediment as the other two but irregular waves. A JONSWAP-Spectrum was

applied to generate random waves with a significant wave height of 1.5 meters and peak

spectral wave period of 6 seconds in a water depth of 5 meters.

4.3.2 Experiments With Constant Waves

Two tests with the same constant wave conditions and different initial profiles were

carried out in the German "Large Wave Flume" in 1986. They are denoted by their initial

profiles as the "dune without foreshore" and the "dune with foreshore". During the

experiments, a test was stopped generally whenever the wave height in the horizontal bottom

portion (water depth of 5 meters) of the flume exceeded 20% of the originally generated

wave height so that the breaking point could stay in its original position. After motions in the

tank subsided, the tests were resumed.

Based on McCowan (1894) theory, waves break when the ratio of water depth to

wave height becomes 1.28. The active water depth, which is the water depth at the breaking

point, is determined as 1.92 meters for the numerical model according to the incident wave


Table 4.3 The numerical model inputs and best-fit results for the two constant wave cases
in German "large wave flume'.
Slope Best-fit K Error* [m2]
Case n=3 n=1
dune shoreline offshore [s/N3] [m4/N] n=3 n=l

Dune without 3.00 0.20 0.20 7.64x10'0 2.03x105- 0.74 8.08

Dune with 3.00 0.20 0.162 1.03x109 8.13x10-6 0.46 1.11

Error represents the root mean square between predicted and observed eroded volume
as shown in Eq. (4.1).

height of 1.5 meters. The input of the dune, shoreline and offshore slopes in the models are

adopted based on the finial profiles in the two tests and are presented in Table 4.3. The best-

fit K values and the corresponding errors of eroded volume resulting from the calibrations are

also shown in Table 4.3. For the two cases, the non-linear transport relationship (n=3)

produces much less error for the eroded volume than the linear transport relationship (n=l),

especially in the case "dune without foreshore". Of relevance is that the best-fit K values for

the non-linear transport equation (n=3) in the calibrations of the two cases here are close to

those in Saville's experiments with a difference of less than a factor of 2. However the

difference in the best-fit K value for the linear transport equation (n=l) is about a factor of

10 in Saville's experiments and German tests. The predicted eroded volumes as a function of

time are compared with measurements in Figure 4.8 for the two cases. The predictions from

the non-linear transport model (n=3) agree very well with the measured eroded volume over

the whole test time for both cases with the same wave and sediment conditions but different

initial profiles. This result supports the discussion in Chapter 3 that the non-linear model has

60 --
Dune without foreshore ..-
a 40
20 0, measured
--- predicted by n=3
/ -. predicted by n=1

0 1 2 3 4 5 6 7
time [hr]

15 -

S Dune with foreshore o
a 10 -

5 0 measured
S- predicted by n=3
2 - -. predicted by n=1
0 I I I I I
0 1 2 3 4 5 6 7
time [hr]

Figure 4.8 Comparisons of predictions with best-fit K values to observed eroded volumes
for the two tests with constant waves in German "large wave flume".

capabilities to handle different initial conditions, whereas the linear transport relationship

(n=l) cannot represent well the difference in time scale of profile evolution caused by the

different initial conditions. Figure 4.8 clearly shows that the linear transport relationship has

a difficulty in simulating the rapid response in the case of "dune without foreshore", although

it could provide an acceptable predictions for the case "dune with foreshore".

Comparisons of the profiles predicted by the non-linear and the linear transport

models and the measurements are presented in Figures 4.9 and 4.10 for the two cases,

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs