Long-term evolution of nearshore disposal berms

UFL/COEL-TR/105

LONG-TERM EVOLUTION OF NEARSHORE
DISPOSAL BERMS

by

Emre N. Otay

Dissertation

1994

LONG-TERM EVOLUTION OF NEARSHORE DISPOSAL BERMS

By

EMRE N. OTAY

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1994

ACKNOWLEDGMENTS

I sincerely do not believe that there is a single paragraph in this dissertation which has

not been influenced by my advisor Dr. Robert G. Dean. Even ideas we do not fully agree on

carry a spirit through his existence. And to Suna I owe this instant for carrying us through the

joys and hardships of our studies.

With this opportunity I would like to thank Dr. Ashish J. Mehta and Dr. Bent A.

Christensen for their caring support, Dr. Hsiang Wang and Dr. Daniel M. Hanes for their help at

important stages of my career, Dr. Donald M. Sheppard for advices at most needed times and Dr.

Peter Y. Sheng, Dr. Michel K. Ochi and Dr. Robert J. Thieke for their excellent lectures.

Specially I would like to thank Dr. Paul A. Work not only for establishing the original

monitoring program but also for his cooperation and friendship that I admire. I thank Dr.

Tae-Hwan Lee for his valuable discussions while sharing the same office. I also would like to

thank Lynda Charles for her sincere friendship, Becky Hudson for making my work place feel

like home, Sidney L. Schofield and Subara B. Malakar for their contributions in collecting and

analyzing the field data and Helen Twedell for her help in the archives.

TABLE OF CONTENTS

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

L IST O F FIG U R E S ............................................ .................................................... v

LIST O F TA B LES ........................................................... ................................... xi

A B ST R A C T ........................................................ ...... .. ............. ........................... xiii

CHAPTERS

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

2 LITERATURE REVIEW ................................................................................ 5

2.1 Previous Nearshore Disposal Projects ....................................................... 7
2.2 Physical M odel Studies ...................................................... .................... 13
2.3 General Guidelines and Classifications of Nearshore Disposal Projects ........... 17

3 FIELD MONITORING PROGRAM .............................................................. 23

3.1 Site Description ......................................... .............. ............................... 27
3.2 Wave, Current and Tide Measurements ........................................ .......... 29
3.3 Sediment Size Distribution .................................................................... 41
3.4 W weather Station .......................................................................................... 46
3.5 Photographic Documentation ..................................................... ............... 50

4 SURVEY ACCURACIES AND IMPROVEMENT TECHNIQUES .................... 51

4.1 State of the Art in Nearshore Surveying ........................................ .......... 52
4.2 Field Testing of Horizontal Distance Measurement Instruments ................... 54
4.3 Horizontal Positioning by Triangulation Method ....................................... 58
4.4 Post-Adjustment of Boat Survey Data ...................... .................... 64

5 SEDIMENT TRANSPORT PROCESSES AT THE NEARSHORE BERM ........... 72

5.1 Topographic and Hydrographic Surveys ..................................................... 72
5.2 Volumetric Changes and On/Offshore Migration of the Berm ...................... 82
5.3 Lateral Spreading and Diffusion ................................................................ 85
5.4 Sheltering Effects on Nearshore Sedimentation ........................................ ..... 90

iii

6 PHYSICAL MODEL STUDIES ................................................. ............... 101

6.1 Experim ental Setup ........................................ .. ....... ................ 101
6.2 Model Similitude Study ............................................. ...... ............... 103
6.3 Measurement Procedure.............................................................................. 104
6.4 Analysis and Interpretation of Laboratory Data ........................................ 105

7 SEDIMENT TRANSPORT MODELING ................................................... 111

7.1 Nearshore Forcing Mechanisms and Their Relative Importance ..................... 112
7.2 Effects of Berm Properties on Sediment Transport ........................................ 115
7.3 Deterministic and Probabilistic Parameterization of Sediment Transport ........ 117
7.4 Diffusion and Advection Processes in Bed Evolution ..................................... 128
7.5 Spectral Modeling of Bed Form Evolution .................................................... 136

8 SUMMARY AND CONCLUSIONS ................................................................. 144

R E FEREN C E S ............................................................................................................. 148

BIOGRAPHICAL SKETCH ......................................................................................... 155

LIST OF FIGURES

FIGURE PAGE
1.1 Beach and profile nourishment. ......................... ............. 2

2.1 Statistics of 193 disposal sites in the U.S.A. (data from Herbich,
1992). ..................................................... ......... ........................... 19

3.1 Site location chart. .................................................. ................... 24

3.2 Components of field monitoring program. ....................................... 28

3.3 Planview of disposal area. ............................................. ............... 29

3.4 Idealized profile cross-section at disposal area. ................................ 29

3.5 Time history of height, period and direction of waves at Ranger
Station. ...................................... .............. ................................. 32

3.6 Time history of height, period and direction of waves at Caucus
Shoal. .................................................................... ..... .......... 32

3.7 Magnitude and direction of mean current and tidal stage at Ranger
Station. ...................................... .............. ................................. 33

3.8 Magnitude and direction of mean current and tidal stage at Caucus
Shoal. ................................................. ................... ........ ........... 33

3.9 Histogram of significant wave heights at Ranger Station. ................... 34

3.10 Histogram of significant wave heights at Caucus Shoal ................. 34

3.11 Histogram of representative wave periods at Ranger Station .............. 35

3.12 Histogram of representative wave periods at Caucus Shoal .............. 35

FIGURE PAGE
3.13 Polar histogram of wave directions at Ranger Station ...................... 37

3.14 Polar histogram of wave directions at Caucus Shoal ......................... 37

3.15 Histogram of current velocities at Ranger Station. ............................ 38

3.16 Histogram of current velocities at Caucus Shoal. ............................................. 38

3.17 Polar histogram of current directions at Ranger Station .................... 39

3.18 Polar histogram of current directions at Caucus Shoal. ..................... 39

3.19 Directional distribution of current velocities at Ranger Station .......... 40

3.20 Directional distribution of current velocities at Caucus Shoal ............ 40

3.21 Longshore distribution of D50 for November, 1993 (solid line) with
envelope (dashed line) of sizes for 1989, 1990, 1991, 1992 and 1993. 43

3.22 Longshore averaged cross-shore distribution of D50. Temporal
variation from November, 1989 to November, 1993. ........................ 44

3.23 Percentage of fines for 5 m samples from November, 1989 to
N ovem ber, 1993. ............................................................................. 45

3.24 Percentage of fines for 8 m samples from November, 1989 to
November, 1993. ................................ ............................. 45

3.25 Time history of air temperature, wind velocity, wind direction and
rainfall. ................................... ....... ...... ............................... 47

3.26 Directional distribution of wind velocities. ..................................... 48

3.27 Histogram of wind velocities. ................................................... 49

3.28 Polar histogram of wind directions. ........................... ........ 49

4.1 Test results for survey tape. ............................................................. 55

4.2 Test results for range finders. ....................................................... 56

FIGURE PAGE
4.3 Test results for MiniRanger. ................................... 57

4.4 Schematic triangulation setup. ..................... .. ............ 59

4.5 Possible triangulation errors. ...................... ... ............. 60

4.6 Solution domain for the triangulation method illustrated for case (a)
in Figure 4.5. ...................... .... .......... ............... 62

4.7 Planview of Little Lagoon survey site ............................................ 63

4.8 Survey trajectory and relative positioning accuracy (thickness of
trajectory line is proportional to the error radius). ............................ 64

4.9 Overlap region in a beach profile survey. ........................................ 65

4.10 Figure 4.10: (a) Wading survey data (dotted) and the analytic
expression given in (4.13) as a curve-fit (solid). (b) Boat profile
Before (dashed) and after (solid) post-adjustment versus wading
profile (dotted) in overlap region. A=0.3 m, B=1.01, a=13.0 m,
sm,=0.02 m ........................................... ................... ... .......... 68

4.11 Three-Parameter post-calibration results for Perdido Key bathymetric
surveys. ..................... ............ ....... ............. 69

4.12 Reduction in vertical RMS-errors for different methods of
post-adjustment. .......................... ... ........ ............................ 71

5.1 Analysis procedure of line survey data. .......................................... 73

5.2 Berm cross-sections obtained from line surveys at R-54 ................... 74

5.3 Nearshore berm generated from box survey data of December, 1993. 75

5.4 Planview of survey lines in (a) state and (b) local coordinate systems. 76

5.5 Data intensity in 93/05 survey with 50x100 m search cells. ............... 79

5.6 Fraction of occupied cells. ....................... .. ... ... ......... 80

FIGURE PAGE
5.7 Number of original survey points per cell. ....................................... 80

5.8 Average vertical standard deviation. .......................... ... ........ 80

5.9 Evolution of the profile nourishment between October, 1992 and
Decem ber, 1993. ....................................... ........ ............. 81

5.10 Cross-sectional area and center of gravity characteristics. ................... 82

5.11 Volumetric changes relative to October, 1992. ................................. 83

5.12 On/offshore migration of center of gravity relative to October, 1992. 84

5.13 Changes in berm sections relative to October, 1992. (a) cross-
sectional areas (b) cross-shore position of center of gravity. ............. 86

5.14 Landward edge of nearshore berm at range R-50. ............................ 87

5.15 (a) Berm cross-section and (b) wave number spectrum at R-54 .......... 89

5.16 Evolution of bed shape. ......................................................................... ... 90

5.17 Average wave number spectrum. ................... ............................. 91

5.18 Average wave number response function. ......................................... 91

5.19 Evolution of dry beach width since completion of beach nourishment. 93

5.20 Average of profiles within protected beach nourishment. Averages
based on profiles at R-50, R-52, R-54, R-56, R-58 and R-60. ............ 95

5.21 Apparent cross-shore sediment transport within protected beach
nourishment, based on average profiles shown in Figure 5.20 from
September, 1990 to November, 1993. ............................................ 95

5.22 Average of profiles within protected beach nourishment. Averages
based on profiles at R-43, R-44, R-45, R-46, R-61, R-62 and R-63. .. 96

FIGURE PAGE
5.23 Apparent cross-shore sediment transport within unprotected beach
nourishment, based on average profiles shown in Figure 5.22 from
September, 1990 to November, 1993. ................................................ 96

6.1 Experimental setup in prototype dimensions. ................................... 102

6.2 Laboratory experiments on nearshore berm evolution. .................... 106

6.3 Erosion and accretion observations in the model studies .................. 107

6.4 Cross-shore sediment transport in the model studies ........................ 108

6.5 Migration of the model berm relative to its initial location .............. 110

6.6 Evolution of bed form in laboratory experiments ............................. 110

7.1 Geometric properties of nearshore berms. ........................................ 116

7.2 Berm classification using Hallermeier's depth limits ........................ 122

7.3 Predicted occurrence of incipient motion at Perdido Key nearshore
berm ............................................................................................ 123

7.4 Comparison of the on/offshore transport criteria and the measured
joint probability density distribution of wave heights and wave
periods at the Perdido Key nearshore berm (Numerical values
represent the joint probabilities in percent chance of occurrence).
Onshore movement lies below the line predicted by a particular 126
equation. ..........................................................................................

7.5 Predicted occurrence of on/offshore movement at Perdido Key
nearshore berm ................................................................................. 127

7.6 Forces acting on a sediment grain. .................................................. 131

7.7 Evolution of sinusoidal bed using tractive force model based on
Equation (7.18). ................................................................................ 135

7.8 Idealized wave number spectrum of river beds ................................ 137

7.9 Analogy between linear systems and spectral evolution of bed forms. 138

FIGURE PAGE
7.10 Solution domain for the spectral model. ..................................... 139

7.11 Comparison of Equation (7.34) with field observations .................. 143

LIST OF TABLES

TABLE PAGE
2.1 Previous nearshore disposal studies. ................................... 6

2.2 Laboratory tests on nourishment techniques (data from Vera-Cruz,
1972). .................... ....... .................... ........................... .. .15

2.3 Nearshore berms and associated depth limits (data from Hands and
A llison, 1991). ...................................................... ....................... 21

3.1 Chronology of Perdido Key field efforts. ....................................... 25

3.2 Wave gage statistics from 10/1/91 to 12/1/93. ................................... 41

4.1 General characteristics of EDM equipment (from ASCE Task
Committee on Hydrographic Investigations of the Committee on
Waterways of the Waterway, Port, Coastal and Ocean Division, 1983). 53

4.2 Comparison of vertical and horizontal accuracies (from Clausner et al.,
1986). ................................................................................... ... 54

4.3 Features of horizontal distance measurement instruments ................. 58

5.1 Horizontal and vertical errors in line and box surveys. ....................... 98

6.1 M odel scale factors ......................................................... ................. 104

7.1 Cross-shore forcing mechanisms effecting the evolution of nearshore 113
berm s. .............................................................. ............................

7.2 Sedimentologic, geometric and hydrodynamic properties of the berm. 117

7.3 Characteristic parameters related to sediment transport .................... 118

7.4 Criteria for the incipient motion ............................... ................... 120

TABLE PAGE
7.5 Criteria related to the on/offshore bar movement (from Larson and 124
Kraus, 1992). .........................................................

7.6 Representative variables at Perdido Key nearshore berm ................ 131

7.7 Sediment transport formula based on Shields parameter.. .................... 133

7.8 Spectral model parameters obtained from regression analysis ............ 142

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

LONG-TERM EVOLUTION OF NEARSHORE DISPOSAL BERMS

By

Emre N. Otay

December, 1994

Chairman: Dr. Robert G. Dean
Major Department: Coastal and Oceanographic Engineering

The long-term fate of underwater berms under the influence of hydrodynamic forces has

been studied using field experiments combined with physical and analytical models. Following

the placement of 3 million m3 of beach quality sand in water depths of 5 to 6.5 m, the evolution

of this underwater berm has been monitored with emphasis on sediment transport processes in

the nearshore zone. In connection with the field studies, four years of wind, wave, current,

sediment and bathymetric data have been collected. Results show that the changes in berm

centroids are within the accuracy limits of the surveys (3.5 m) during two years following

placement. Lateral spreading and diffusion of berm shape have been observed. The wave number

spectra of the bed form show a consistent decrease in total spectral energy within the two years of

monitoring. The higher rates of reduction occur at higher wave numbers consistent with the

preferential hydrodynamic smoothing of the shorter bed undulations. In addition to local changes

in berm geometry, both beach profile and shoreline change data showed evidence of sheltering

effects due to the presence of the berm. The average shoreline recession and the seaward flux of

sediment within the sheltered area has been substantially reduced compared to unprotected

portions of the beach. Small scale laboratory experiments supported findings on small net

transport compared to more dominant diffusive processes. Several dimensionless criteria related

to initiation of motion and direction of transport are evaluated against existing wave data.

Deterministic and probabilistic application of existing incipient motion criteria estimated the

berm to be active 63% of the time, although no net movement has been observed. Several criteria

related to the on/offshore movement of bars have been tested against wave data. On a preliminary

basis it is concluded that none of the criteria posed are suitable for predicting cross-shore motion

of an offshore berm. A spectral evolution model showed reasonable agreement with the field

observations of the bed form attenuation, although consistent differences were present in the

spectral form of this attenuation over the wave number range examined.

CHAPTER 1
INTRODUCTION

The underwater as an unknown frontier has fascinated mankind as early as the differences

between solids and fluids were recognized. Oceans, being the largest collection of water, also

represent one the most complex mediums. Even a simple task on land might challenge the

theoretical and practical capabilities of modem science and technology if conducted underwater.

The disposal of dredge material in the offshore is a fairly simple operation though it raises

many questions concerning the hydrodynamics and sediment transport in open water, enough to

make it the objective of this study to expand our understanding about the evolution ofnearshore

disposal berms. The simplicity of the process and the variety of applications make the nearshore

disposal and the construction of an underwater berm a favored tool in the coastal engineering

profession. In places where dredge material is available in quantities, nearshore disposal berms

provide a potential solution to coastal protection directly as a "feeder berm" or indirectly by

reducing the incoming wave energy.

With the improved knowledge about the problems associated with shore protection measures

using hard structures including seawalls, groin fields and offshore breakwaters, soft structures

such as beach and nearshore nourishment have become a popular means of erosion control. In

nourishment efforts, the relatively low cost of offshore placement of sediment compared to

stockpiling at the foreshore (Figure 1.1) is a critical factor, along with the reduced environmental

impact, for example, on benthic fauna in the swash zone, sea turtle nesting and the nutrition

transport to dune vegetation. Under certain conditions, offshore mounds may also serve as fish

habitats and depositories for unwanted or contaminated material.

2

beach nourishment

profile nourishment

Figure 1.1: Beach and profile nourishment.

Despite the advantages, the design of nearshore berms often involves uncertainties about its

performance. Prediction of the stability and the long-term fate of disposal material requires a

detailed understanding of the forcing mechanisms responsible for the underwater evolution.

Considering the complexity of the problem and the lack of well developed theories, pure

analytical studies are generally insufficient for a thorough investigation unless an experimental

study is accompanied with it.

Physical modeling has been accepted as a valuable means of qualitative and quantitative

observation although the latter is often questioned. Underlying dynamics of many known

processes experience substantial scale and other effects when simulated in physical models.

Especially seaward of the breaking zone where the changes may be small compared to any

synthetically created mechanism such as amplified wave reflection, model specific circulations

and boundary and scale effects, all of which reduce the validity of the data. A well designed field

monitoring project, on the other hand, is free of the above mentioned problems, but the challenge

still remains to carry out measurements under the uncontrolled conditions in nature.

The objectives of this thesis includes three components. (1) Monitor the evolution of a

nearshore disposal berm under the forcing of waves and currents, (2) evaluate the findings of the

field investigations to understand the underlying processes, and (3) examine the ability to predict

the response of the berm to the existing forcing.

Within the scope of this study, previous field investigations of other disposal projects are

reviewed. The core of the experimental part is based on an extensive field monitoring of the

Perdido Key nearshore disposal berm at Perdido Key, Florida. For more than four years large

amounts of field data have been collected including bathymetric changes at the berm area as well

as the measurement of hydrodynamic forces and sediment samples. Analysis methods are

developed to extract information on sediment size distributions, wind, wave and current climate

and nearshore bathymetry. Realizing the importance of measurement accuracy specifically for the

extraction of sediment transport quantities from bathymetric surveys, accuracy limits of different

surveying techniques are evaluated and improvement methods are developed and applied.

The key factor to predicting the long-term fate of underwater berms is to identify the physical

processes and the responsible mechanisms. Small scale laboratory tests and specifically planned

field measurements are very effective means to establish the relative importance of underlying

diffusion and advection processes. This information can later be used for theoretical models of

cross-shore sediment transport. The nearshore zone as the less active region in the beach system

is affected by complex mechanisms. The effect of the local bottom slope causing a gravity force

and the wave induced currents are major contributors to the transport of sediment. Wind, current

and tide driven circulations can gain considerable importance depending on the local conditions

at the disposal area. In many cases, these areas are close to inlets and dredged channels which

complicates local hydrodynamics. Irregularities in longshore transport, generation of rip currents

and surfbeat may cause local concentration of sedimentation or erosion patterns. At this stage a

4

realistic distinction between dominant forces and those that are relatively insignificant is

necessary. The prediction of nearshore berm evolution within an acceptable level of confidence

will require a further emphasis on focused field monitoring and subsequent correlation with

analytical and numerical approaches.

CHAPTER 2
LITERATURE REVIEW

Open-water disposal of dredged material has been practiced world-wide for over half a

century. The initial attempts have arisen from the search for a beneficial use of the large amounts

of dredged material obtained from channel maintenance operations. The removed material which

varies in size and quantity has been placed in nearshore disposal sites seaward of the surfzone.

Depending on the specific design purpose, the nearshore material may be expected to either

feed the eroding beach system or remain stable within designated limits without being activated

by the surrounding forces. In both cases, an extensive monitoring effort is required if the fate of

the disposal is of major concern. Most of the previous field studies reviewed in Section 2.1 dealt

not only with the question of how the underwater disposal evolves but also include

measurements of other quantities that are likely to affect the fate of the project. These are mainly

physical environmental forces including waves, currents, tides, and winds; sediment properties of

the native and placed material and finally the local geometry of the disposal site (Table 2.1). In

addition to field monitoring programs, a small number of laboratory experiments have been

conducted. Section 2.2 summarizes several physical model studies simulating nearshore disposal

sites. In the last two decades there have been attempts to interpret the physical processes at

nearshore disposals of dredged material. Experiences from previous field works were compiled to

establish case histories of disposal mounds and their evolution. A large set of information was

categorized according to relevant physical parameters. General guidelines were established for

the design of future nearshore disposal projects. This final group of literature was reviewed in

Section 2.3.

Table 2.1: Previous nearshore disposal projects.
Location Date Placed Water Mound Sand Wave Wave On/Off- Shore Reference
Volume Depth Relief Size Height Period shore Pro-
[m3] [m] [m] [mm] [m] [s] Motion tection

Santa Barbara, CA
Atlantic City, NJ
Long Branch, NJ
Durban, South Africa
Copacabana Beach, Brazil
Long Island Sound, CN
Lake Erie, OH
New River Inlet, NC
Limfjord Barriers, Denmark
Tauranga Bay, New Zealand
Dam Neck, VA
Sand Island, AL
Fire Island, NY
Jones Inlet, NY
Mobile Outer Mound, AL
Coos Bay, OR
Silver Strand, CA
Kira Beach, Australia
Mt.Maunganui, New Zealand
Port Canaveral, FL
Perdido Key, Florida

1935
1942
1948
1970
1970
1974
1975
1976
1976
1976
1982
1987
1987
1987
1988
1988
1988
1988
1990

1992
1992

154,000
2.7 mil
460,000
2.5 mil
2.0 mil
1.17 mil
18,000
26,750
22,000
2 mil
650,000
350,000
320,000
300,000
14.3 mil
4 mil
113,000
1.5 mil
80,000
120,000
3 mil

6.1
4.6-7.6
11.5
7-16
4-6
18.3
17
2-4
4-5
11-17
10-11
5.8
4.9
4.9
10.7-13.7
20-26
4.6-5.5
7-10
4-7
5.3-6.8
5-6

0.18
0.32
0.34
0.35
0.4-0.5
silt
silt
0.49
0.25-0.3

0.08
0.22

fine sand
0.25-0.3
0.2

0.3

7-9

10-14

7.3

3.4-4.6
11.5
13.1
8
5.9
6.3
5.7

stable
stable
stable
both
onshore
stable
stable
onshore
onshore
stable
stable
onshore

stable

loss
onshore
onshore

onshore
stable

none
none
none
indirect
direct
none
none
direct
direct

direct

indirect
none
direct
direct
direct
direct
indirect

Hall and Herron (1950)
Hall and Herron (1950)
Hall and Herron (1950)
Zwambom et al. (1970)
Vera-Cruz (1972)
Bokuniewicz et al.(1977)
Danek et al. (1978)
Schwartz and Musialowski(1977)
Mikkelsen (1977)

Healy et al. (1991)
Hands and DeLoach (1984)
Hands and Bradley (1990)
McLellan et al. (1988)
McLellan et al. (1988)
McLellan (1990)
Hartman et al. (1991)
Andrassy (1991)
Smith and Jackson (1990)
Foster et al. (1994)

Bodge(1994)
present study

L ______________ .1. ___________________ L ______________ .L _____________ ________________ ______________ I _______________________________________________________________

2.1
0-8.3

9.1
0.36
1-8
2.1
9
3.3
1.8-2.1
2
2
6.6
4.6-7.6
2.1
2
2
1.65
1.75

1-2
1-2
0.7
0.1
0.1
0.55

0.3-0.8
2.7
0.62
4
0.5-1.5
1.2
0.45

2.1 Previous Nearshore Disposal Projects

This review is limited to the field monitoring of the long-term physical evolution of offshore

disposal projects around the world's open waters. The earliest reports were found in U.S. Army

Corps of Engineers (USACE) dredging projects. Santa Barbara Harbor, California, was dredged

by hopper dredges in 1935. A total of 154,000 m3 of material was placed 300 m offshore, along

the 6 m contour. The resulting underwater mound was approximately 700 m long and 1.5 m high

(Hall and Herron, 1950). After two years of monitoring the berm was reported as "exceptionally

stable" although there was evidence that some of the new material had moved onshore to fill the

areas between the mound and the shoreward slope of the profile.

Another attempt by the USACE in 1942 was to nourish an eroding beach at Atlantic City,

New Jersey (Hall and Herron, 1950). Over a period of seven years, 2.7 million m3 of dredge spoil

was placed in 4.6 to 7.6 m of water with the expectation that the spoil would move ashore by

natural forces. Surveys showed no evidence that sand moved onto the beach.

In 1948, the USACE deposited 460,000 m3 of dredged material at Long Branch, New Jersey,

in 11.5 m of water (Hall and Herron, 1950). Local waves, sediments and offshore bathymetry

were measured for four years. The first report by Hall and Herron (1950) indicated continuing

shoreline recession and no net substantial sand movement in the mound area. A later report by

Harris (1954) confirmed the initial findings such that 94% of the originally deposited material

was found in place after four years of monitoring. It was concluded that future offshore

placements should be conducted in less than 6 m of water in order to benefit the beach.

These unsatisfactory results from early attempts of nearshore nourishment were followed by

two decades of no activity in the field. Between 1966 and 1970 a large nourishment project was

constructed at Durban Beaches, South Africa (Zwambom et al., 1970). A total of 2.5 million m3

of good quality sand was placed at depths between 7 m and 16 m, forming a 1.5 km long

nearshore berm to protect the leeward beaches. The field monitoring program included sediment

sampling, tracer tests and bathymetric surveys. In addition to the field efforts several physical

model tests were conducted to study the effect of crest width on wave attenuation and beach

protection. These laboratory findings will be summarized in Section 2.2. The nearshore berm

cross-sections during the field studies indicated random fluctuations in the vertical and horizontal

(on/offshore) dimensions of the berm. During two major storms with significant wave heights

ranging from 2.4 to 3.5 m, the berm temporarily lost 20% to 23% of its original volume which

was later returned in post-storm recovery periods. Four years of monitoring indicated that the

berm provided significant protection to the leeward beaches. Measurements along protected and

unprotected portions of the beach showed that the protected beaches had gained 20% whereas the

unprotected beaches had lost 20% of the sand volume above Mean Sea Level averaged over a

period of six months during construction of the berm. Results from the physical model study

supported the field observations of beach protection by the nearshore berm.

During the beach and offshore nourishment project at the 4.2 km long Copacabana Beach,

Brazil, 2 million m3 of medium size sand was dumped in water depths of 4 to 6 m (Vera-Cruz,

1972). A similar test was run in the laboratory parallel to the field monitoring. During and after

the nourishment project, sediment samples were collected and waves, tides and winds were

measured. The 14 profile lines surveyed every 15 days for two years indicated shoreline advance

even after the beach nourishment was completed. The additional volumetric gain of the beach

material and the increasing grain sizes in the beach sediment samples suggested a potential

feedback from the offshore site which was originally nourished with a coarser sand.

In 1974 a mixture of silt and sand was dumped over a naturally accreting mud bottom at the

New Haven disposal site in Long Island Sound, Connecticut (Bokuniewicz et al., 1977). A total

of 1.17 million m3 of dredge spoil was placed in 18 m depth, forming a 9 m high conical mound.

During 200 days of monitoring the mound crest was lowered by 2 m due to dewatering and self

compaction. No additional changes were observed even during Hurricane Belle in 1976.

In 1976 a relatively small amount of dredged material (18,000 m3) was placed in Lake Erie

near Ashtabula, Ohio (Danek et al., 1978). The average depth at the disposal site was 17 m.

Accurate measurements (2 cm) conducted with stationary survey rods and sediment traps led to

conclusions that the sediment pile was stable and there was no evidence of any compaction.

Generally it was estimated that the changes due to erosion were much more important than

compaction.

In 1976 a three-month monitoring program was conducted following a dredge spoil disposal

at New River Inlet, North Carolina (Schwartz and Musialowski, 1977). A total of 26,750 m3 of

coarse sand was placed between the 2 and 4 m contours, forming a berm relief of 1.8 m. After 13

weeks 75% of the initial berm volume was removed from the offshore zone. The location of this

material removal could not be identified. It was concluded that the lost sediment first moved

onshore. After entering the littoral zone it was transported away by strong longshore currents. In

a later report, Schwartz and Musialowski (1980) indicated more evidence to support onshore

transport of disposal sediment. The combination of erosion in the offshore and accretion in the

inshore and the fact that the steep sides of bed ripples were facing landward suggested that the

offshore deposited sediment was moved onshore to fill the trough of the surf zone bar and later

transported away by longshore currents.

In 1976 an offshore nourishment was conducted inside a groin field at Limfjord Barriers in

the Danish North Sea (Mikkelsen, 1977). A deposit of 30,000 m3 of medium sized sand was

placed in 4 to 5 m of water, forming a 2.1 m high artificial bar. After an initial loss during the

placement, the remaining material continually moved onshore. In the final survey, 10 months

after the nourishment, the bar adjusted itself to the existing profile by raising the profile and

steepening the foreshore. The fact that the borrow sand was coarser than the native sand

contributed to the development of a steeper inshore and to a small rate of material loss out of the

system. The groins have also played a protective role by preventing the material from dispersing.

Observed onshore transport was supported by offshore winds during the monitoring project.

Between 1977 and 1978 a 2 million m3 dredge spoil was placed in Tauranga Harbor near Bay

of Plenty, New Zealand (Healy et al., 1991). The disposal material formed an underwater mound

800 m diameter and 9 m vertical relief at 11 m to 17 m of water. During 10 years of observation

the mound height was reduced by 0.46 m. No significant migration was observed.

In 1982 a 3.3 m high test mound was constructed at the Dam Neck disposal site off Virginia

Beach, Virginia (Hands and DeLoach, 1984). A total of 650,000 m3 unpolluted mixture of fine

sand and silt were dredged from the Chesapeake Bay and deposited off Virginia Beach in 10 m to

11 m of water. A one-year field monitoring study indicated no substantial change in the

geometry. Several underwater measurement techniques were used during the field investigations.

These included hydrographic surveys, side-scan sonars, seismic sub-bottom surveys, referenced

rods, surface and bottom sediment samples, bottom sediment cores and diver observations. Their

relative advantages were evaluated and future potentials were discussed.

In 1987 the USACE constructed a submerged berm near Sand Island, Alabama (Hands and

Bradley, 1990). A total of 350,000 m3 of clean fine grain sand was placed creating a 1.8 km long

underwater berm. At the South comer of the berm, Hands and Allison (1991) and Hands (1991)

reported another preexisting man-made mound 250 m in diameter. Both features, the elongated

berm and the mound, were located in approximately 6 m of water with a 2 m relief. A three-year

field monitoring showed relatively small dispersion in the berm geometry and no evidence for

offshore loss of sediment. The landward portion of the berm migrated shoreward slightly. The

mound which had a conical shape was smaller than the berm and responded as a single unit by

I

moving shoreward at a rate of 30 m/year. The mound relief decreased by 0.6 m in 33 months.

The mound showed sustained onshore migration and slow dispersion.

The Mobile Outer Mound was built between 1988 and 1989 using dredged material from the

Mobile ship channel deepening operations (McLellan et al., 1990). A total of 14.3 million m3 of

sand and mud was placed approximately 8 km off Dauphin Island, Alabama, offshore of the 10

m contour. Maximum mound relief was 6.6 m. The initial efforts to design a stable mound for

wave energy reduction and protection of the shoreward beaches were successful. Two wave

gages were located seaward and landward of the mound. Wave records during normal and

extreme seas showed 29% to 75% reduction in wave heights. The latter was measured during

tropical storm Beryl.

From 1979 to 1988 a total of 4 million m3 of medium size sand was disposed at 20 to 26 m

water depth off Coos Bay, Oregon (Hartman et al., 1991). Mound relief varied from 4.6 to 7.6 m.

A nine-year wave buoy record indicated an average significant wave height of 2.7 m offshore of

the disposal site. Survey results indicated that each year 126,000 m3 of material were transported

away. After 10 years of monitoring 71% of the total volume remained in place.

In 1988 an underwater disposal berm was constructed at Silver Strand State Park near San

Diego Bay, California (Andrassy, 1991). The berm was composed of 113,000 m3 of medium

sized sand. With 2 m relief at 5 m water depth, the berm was expected to reduce a significant

portion of the incoming wave energy. Bathymetric surveys and wave records were used to

monitor the developments in the nearshore. Within approximately two years after construction

the berm had migrated 34 m onshore. The berm had also contributed indirectly to the nearshore

accretion by blocking erosive waves. The shoreline had advanced an average of 40 m. No

significant longshore movement was encountered in the nourished material.

Two projects involving offshore storm bars were reported on the Gold Coast of Australia

(Smith and Jackson, 1990). The second placement in 1988 was larger in size (1.5 million m3) and

was also better monitored. Material was placed between the 7 and 10 m contours elevating bar

crests up to 2 m. Two days after completion of the project, large swells were developed causing

two weeks of intense wave activity in the area. Waves up to 4 m height were observed breaking

directly on the artificial bar which reduced the waves to heights of 2 m. A post-storm survey

showed that approximately half of the original nourished material was transported onshore.

During a storm only 15 m of dune recession was measured which led to the conclusion that the

project protected the beach successfully.

In 1990, 80,000 m3 of dredged sediment was placed near Mt. Maunganui, New Zealand, in 4

to 7 m of water (Foster et al., 1994). Before the nourishment, the beach profiles showed

alternating characteristics of erosion and accretion depending on the survey season. Following

the placement, all sections of the profile landward of the dump ground experienced accretion

which was interpreted as an indication of continuous onshore transport of material from the dump

ground. This interpretation is rather inconclusive since the post-nourishment accretion rates in

the shoreward regions were not in balance with the amount of erosion at the dump ground. A

closer study of accreted and eroded volumes showed that these volumes were in the same order

of magnitude as the natural fluctuations in the sediment budget before the disposal. Similarly,

measured volumetric changes in regions other than the disposal site were equivalent to

approximately 10 cm vertical change of bed elevation which is within the limits of the survey

accuracy. Although the observed accretion in the foreshore could be due to natural fluctuations of

the profile and/or measurement errors and not necessarily due to the onshore transport of the

disposal material the erosion on the mound itself was real. Consequently 85% of the total

material was removed within a period of 3-4 months.

In 1992 Port Canaveral Harbor was dredged producing 120,000 m3 of beach quality material

(Bodge, 1994). The borrow material was placed in a nearshore disposal area offshore of Cocoa

Beach, Florida. The constructed berm reached an average vertical relief of 1.65 m at

approximately 6 m of pre-nourishment depth. Based on one-year monitoring including wave

measurements and bathymetric surveys it was concluded that the berm has moved onshore by 60

m under 1.2 m average significant wave heights.

2.2 Physical Model Studies

In this section previous laboratory experiments on artificial nearshore berms will be

reviewed. Studies of natural nearshore systems will be considered in Chapter 7, under sediment

transport modeling. Two of the previous model studies, Zwambom et al. (1970) and Vera-Cruz

(1972), were coupled with a nearshore nourishment project, such that the results from the

laboratory experiments were used in the design of the prototype disposal project. Other

investigations reviewed in this section include physical modeling of nearshore mounds of sand

by Gunyakti (1987) and the wave transformation characteristics at artificial submerged structures

by Sawaragi et al. (1988) and Vincent and Briggs (1989).

Several wave flume tests were conducted in connection with the nearshore nourishment

project at Durban Beaches, South Africa, (Zwambom et al., 1970). The mound was constructed

on a beach slope of 1:20 a water depth of 15 m. The mound crest was located 7.5 m below the

Mean Sea Level. The width of the mound crest was varied during the experiments to find the

optimum design value. All quantities were reported in prototype dimensions. The first set of

experiments was conducted on a moveable bed model to test the criterion for erosive and

non-erosive conditions based on previous work by Iwagaki and Noda (1962). Medium sized sand

and anthracite, with specific gravity of 1.35, were used to simulate the prototype beach profile.

Wave and tidal conditions were reproduced until equilibrium profiles were reached. Results

showed good agreement between measured beach deformations and the proposed criterion for

erosion. The beach deformation was given as a function of the deep water wave steepness, HJLo,

and the deep water Froude number, ws in which Ho is the deep water wave height, L. is

the deep water wave length, g is the gravitational acceleration and w, is the settling velocity of

the mean grain size. In another series of tests, different prototype mound crests were scaled in a

fixed bed model. For crest widths of 61 m, wave heights were reduced up to 30% while passing

over the mound. Further increases in width showed no significant changes in wave attenuation.

Additional moveable bed tests were conducted with and without a nearshore mound. The model

with mound showed considerable reduction in beach erosion compared to the model without

nearshore mound. Under the same extreme storm conditions the final positions of the beach

profiles measured in two separate models differed by 90 m in the horizontal and 2.5 m in the

vertical. The mound itself responded to the storm with 1 m reduction in crest elevation and 40 m

seaward migration. This final observation is consistent with the known fact of offshore migrating

natural bars during storm seasons.

The beach and nearshore nourishment of Copacabana Beach, Brazil was first tested in a

distorted movable bed model (Vera-Cruz, 1972). After successfully reproducing an equilibrium

shape of the beach similar to the prototype for normal wave conditions, four different

nourishment techniques were tested. The required sand volumes and run-times to reach 85 m

beach widening are presented in Table 2.2 for each placement technique. Stockpiling on the

foreshore and offshore placement were applied first separately and later as a mixed method to

nourish the beach. The mixed method was further distinguished as non-programmed when

stockpiling and offshore placement were applied simultaneously at the same range or as

programmed when the transects were alternated so that the two nourishment processes did not

interfere. During the non-programmed method, the falling of the stockpiled sand to deep levels

counteracted the climbing of the dumped sand to high levels. Except for this inconvenient

interference, the offshore dumping proved to be a cost effective way for nearshore nourishment

projects if designed carefully to guarantee that the offshore material feeds the beach.

Table 2.2: Laboratory tests on nourishment techniques (data from Vera-Cruz, 1972).
Placement Technique Required Volume to Reach 85 m Required Model Run-
Beach Widening [m3/m] Time to Reach 85 m
Beach Offshore Total Beach Widening [hrs]
Stockpiling Only 714 0 714 30
Offshore Placement Only 0 952 952 60
Mixed Method (non-programmed) 571 476 1047 53
Mixed Method (programmed) 357 476 833 ?

Gunyakti (1987) studied the behavior of offshore submerged mounds of dredged material in

two dimensional model tests. Using fine and medium sized sands, the offshore mound was

subjected to wave action. Test results indicated that coarser sediments were conserved between

the original mound and the shoreline suggesting onshore transport as the dominant mechanism.

Fine sediments got lost offshore with a fraction being transported first onshore and later

dispersed by longshore currents. A mixture of fine and coarse sand resulted with dispersion of

fine material to the offshore and onshore transport of the coarser sand. The amount of onshore

deposited material was found to be directly related to the relative height, hid (ratio of mound

relief to depth of mound base), the relative width, b/(d-h) (ratio of crest width to depth of mound
crest) and the depth of mound crest (d-h). For two conditions, 0.5
was found that 75% of the original material was deposited on the beach.

The experimental study by Sawaragi et al. (1988) on the effects of submerged breakwaters on

artificially nourished beaches produced some findings which were indirectly related to questions

of wave energy attenuation and cross-shore sediment transport at artificial mounds of dredged

material. By controlling the erosion and accretion processes at nourished beaches the artificial

reef had similar features to submerged nearshore mounds. In the first series of experiments, wave

breaking criteria and wave attenuation on an artificial reef were studied using a 2-D fixed bed

wave tank. The second series of tests which were more relevant to the present study were

conducted on a moveable bed model. Using 0.3 mm sand, artificial reefs were constructed where

the crest height to water depth ratio was varied between 0.5 and 0.85. Waves with 13.1 cm height

and 1.1 s period were generated to study erosion criteria on the reef. The ratio of the shear

velocity at the bottom to the settling velocity of the sediment (u./w,) was proposed as a critical

parameter to control the initiation of erosion. Critical values of (u./w,) were evaluated for 14

cases at regions where significant erosion was initiated. Results showed that erosion of the

mound occurs when u./w, > 0.5-0.6. It can be shown that these findings are consistent with the

Shield's criterion for incipient motion. A detailed discussion on initiation of motion and other

relevant dimensionless groups can be found in Chapter 7.

In a fixed bed wave basin, Vincent and Briggs (1989) studied the wave transformation

characteristics over a submerged mound. The spatial distribution of wave heights was measured

under various wave conditions, including monochromatic, spectral, unidirectional, directional,

breaking and nonbreaking waves and their combinations. All tests were performed using an

elliptical mound of 30 cm maximum height and a uniform water depth of 45 cm. The results

showed that the linear monochromatic wave ray theory was inadequate to estimate the actual

conditions with irregular waves and directional spreading. Only for small, unidirectional waves

did the tests for monochromatic waves produce good representations of the irregular waves. For

other cases monochromatic tests overestimated the amplification of irregular waves by 50 to over

100%.

2.3 General Guidelines and Classifications ofNearshore Disposal Projects

Except for a few early attempts (Basco et al., 1974), compilation of existing field experiences

into case histories is a fairly recent development which began in the 1990s. Learning from past

experiences has always provided a good approach to create new ideas in the field of engineering

and science. In this section, two groups of studies will be reviewed. The first group of papers by

Basco et al. (1974), Dortch (1990) and Fredette et al. (1990) consists of guidelines for

experimental and analytical methodologies to analyze the physical evolution of disposal berms.

The second group of studies by Pequegnat et al. (1990), Herbich (1992), McLellan (1990) and

McLellan and Kraus (1991) categorized existing nearshore berms according to local

characteristics and outlined design methodologies. Finally, Hands and Allison (1991) and Hands

(1991) took the berm concept one step further in an attempt to explain the berm behavior using

site specific berm and wave properties.

In a comprehensive report, Basco et al. (1974) described the physical aspects of dredged

material disposal in river, estuary, ocean and lakes. An extensive literature review was

subdivided into the above mentioned regions. Existing knowledge on physical properties of

disposal mounds, fundamental fluid mechanics and sediment transport mechanisms were

summarized. Available techniques for field and laboratory studies were described. Factors

controlling the long-term fate of disposal mounds were identified as the bottom-layer mud flow,

suspension by wind-wave action, transport by tidal currents and deposition affected by salinity

induced flocculation. It was concluded that no analytical method was available to estimate the

dispersion geometry of dredged material.

A large collection of information on the long-term fate of aquatic disposal sites was compiled

in a series of reports by the USACE's Dredging Operations Technical Support Program. The first

report in this series, dealing with the general problem of long-term evolution of open-water

disposal material, was published by Dortch (1990). The physical processes following the

placement were analyzed in two stages. These were the mound resuspension and dynamics, and

transport and redeposition. Steady state analytical methods, time and rate dependent analytical

methods, modeling (physical, numerical, hybrid) and experimental studies (field and laboratory)

were suggested as tools to analyze the long-term fate of the material. Physical processes such as

sediment transport by waves and currents, shear stresses, transport formula and armoring effects

of bed surface layer were explained. A list of recommendations for future efforts was presented

including improvement of measurement accuracies, better understanding of cohesive sediments

and long-term field monitoring.

Fredette et al. (1990) evaluated selected measurement techniques for physical and biological

monitoring of aquatic dredged material disposal sites. Instruments such as side-scan sonar,

seismic profiler, positioning and remote sensing devices, current meters, benthic and nekton

sampling devices were described using specifications and figures.

In a revised procedural guide for ocean dredged material disposal sites, Pequegnat et al.

(1990) reported 108 disposal projects in the U.S.A. These sites were categorized in USACE

districts according to project characteristics including water depth, sediment volume and offshore

distance.

A similar review of open-water disposal sites in the U.S.A. was given by Herbich (1992).

Results from a 1989 survey were presented in a statistic summary of project characteristics.

Figure 2.1 shows some of these statistics for 193 disposal sites in 20 USACE districts. According

to Figure 2.1 only 13% of the disposal sites were monitored out of which 58% was dispersive

and 43% was active. Related publications by the USACE's Dredged Material Research Program

were listed. Beneficial uses of dredged material were summarized.

unknown (43.52%) -

unknown (88.0W

unknown (74.0904

monitored (13.47%)

not-monitored (43.01%)

active (5.18%)
stable (6.74%)

dispersive (15.03%)
..', not-dispersive (10.88%)

Figure 2.1: Statistics of 193 disposal sites in the U.S.A. (data from Herbich, 1992).

McLellan (1990) presented a list of ten nearshore berm projects constructed with dredged

material and a list of split-hulled hopper dredges in the U.S.A. which can operate in shallow

areas. McLellan and Kraus (1991) gave definitions for nearshore berms as feeder or stable berm,

depending on the construction purpose. Design methodologies, in particular the significance of

location and timing of placement and the berm geometry, were explained. To determine the berm

stability a criterion was proposed, based on Dean Number, HJ(w, T) in which Ho is the deep

water wave height, w, is the sediment fall velocity and T is the wave period. The design

calculations were demonstrated for a proposed nearshore berm.

In a series of papers, Hands and Allison (1991) and Hands (1991) classified eleven nearshore

berms according to their stability characteristics. The measured response of the berms to the

wave forces was compared to estimates using Hallermeier's (1981) profile zonation limits. Most

of the required wave information was obtained from the Wave Information Study (WIS) hindcast

data. The inner (h, and outer (hot) depth limits for the "buffer zone" proposed by Hallermeier

(1981) were assumed as the seaward limit for the active zone and the landward limit for the

stable zone, respectively. These limit depths are shown in Equations (2.1) and (2.2).

hin=2 Hs + 11 oHs (2.1)

hout=(Hs-H- 3 Hs) (2.2)

where,

Hs = mean local significant wave height
oHs = standard deviation of local significant wave height
T = mean wave period
d = grain diameter

The offshore distance of the berm from the inner limit (seaward as positive) normalized by

the width of the "buffer zone" is defined as Xbe,. If the bottom slope between the berm and the

depth limits is assumed uniform, xber can be calculated by dividing the vertical difference of

measured water depth at the berm (hbnn) and the calculated inner depth limit (h,) by the vertical

drop between outer and inner depth limits (hou hj. as shown in Equation (2.3).

L

hberm-hin
Xberm- hout-hin

(2.3)

Since the original dimensions of the "buffer zone" varied with the site properties, by using

Xbe,, different nearshore berms can be compared on a common hypothetical profile. Table 2.3

shows the variation of berm depths with associated inner (hm) and outer (ho,) depth limits for

nine sites in the U.S.A. hm and ho vary from 3 to 10.3 m and from 7.4 to 57.3 m, respectively.

The corresponding "buffer zone" widths, depending on the bottom slope, would range from 0 to

470 m for a 1:10 slope, and from 0 to 4.7 km for a 1:100 slope. The same data set is illustrated

including the present berm under study later in Figure 7.2.

Table 2.3: Nearshore berms and associated depth limits (data from Hands and Allison, 1991).
SITE hbm hm hout x Active/Stable
[m] [m] [m] [-]
Dam Neck, NC 7.6 6.8 10.7 0.205 S
Atlantic City, NJ 5.8 5.9 8.1 -0.045 S
Santa Barbara, CA 6.7 3 10.2 0.514 S
Sand Island, AL 5.8 6.7 17.8 -0.081 A
Long Island, NY 4.9 5.9 7.4 -0.667 A
Brazos, TX 8.2 7.9 41.4 0.009 A
Silver Strand, CA 4.9 8.1 35.3 -0.118 A
Humboldt, CA 15.8 10.3 57.3 0.117 A
Humboldt, CA (SF3) 21.3 10.3 57.3 0.234 A

Independent of berm geometry and sediment size, the authors expressed the exceedance

statistics of wave heights and maximum near-bed orbital velocities. While the wave height

classification could not distinctly differentiate between active and stable berms, the near-bed

velocities, especially at the upper 75- to 95 percentile, correlated well with the observed berm

22

behavior. Calculations using WIS (Wave Information Study) data showed consistently that berms

were active if (um.x,75 > 0.4 m/s) or (um95% > 0.7 m/s). For listed sites, grain size, tide, wind and

oceanic currents were assumed not to have a critical effect on the stability of the berm.

CHAPTER
FIELD MONITORING PROGRAM

The field study portion of this dissertation is based on an extensive monitoring program for

the beach and nearshore nourishment project at Perdido Key, Florida (Figure 3.1). In 1989 the

U.S. Navy initiated a dredging project to deepen the navigation channel of Pensacola Pass by an

additional 4 m. Approximately 4.1 million m3 of dredge material were placed along the eastern 7

km of the Gulf of Mexico shoreline at Perdido Key between November 1989 and September

1990. The beach nourishment portion of the project was investigated by Work (1992), Work and

Dean (1992), Work et al. (1990a,b; 1991a,b) and by Otay and Dean (1993; 1994). The second

phase of the dredging operations provided approximately 3 million m3 of additional material for

the construction of the nearshore disposal berm. The offshore deposition took place between

September 1990 and October 1991 at pre-placement depths ranging from 5 m to 6.5 m.

The five-year monitoring program, sponsored by the National Park Service and the U.S.

Navy, incorporated physical, biological and environmental studies. The Coastal and

Oceanographic Engineering Department at the University of Florida was responsible for the

physical monitoring. In the following sections, the site and physical data collection methods will

be discussed focusing on the evolution of the nearshore berm from 1991 to 1994. The field

monitoring was divided into topographic and bathymetric surveys, wave, current and tide

measurements, sediment samples, meteorological data acquisition and oblique photography. The

hydrodynamic and meteorological data were collected continuously at fixed stations. The

complete history of field efforts is summarized in Table 3.1. With the exception of the

bathymetric surveys, the physical monitoring efforts are discussed in this chapter.

0 5 km
6 !

Figure 3.1: Site location chart.

Table 3.1: Chronology of Perdido Key field efforts.

Date

10/28-11/1/89

11/17/89
1/18/90

1/29/90
1/30/90
3/7-3/9/90

5/2/99
6/24/90
8/8/90
8/17/90
9/22-9/26/90

12/6/90

1/29-2/3/91

5/15-5/16/91
5/28-6/1/91

6/18-6/19/91

7/29-7/30/91
9/10/91

9/28-10/2/91

10/12-10/20/91

10/23-10/24/91

Task

Pre-nourishment survey:
Wading/swimming profiles (Gulf and Bay), offshore bathymetry
Sand samples, photos
Placement of nourishment material begins
Wave gage tripod and standalone gage installed
Tide gage installed at Ft. Pickens Pier, Santa Rosa Island
Mechanical (analog) weather station installed
Large stilling well installed for Ft. Pickens tide gage
56 sand samples collected, to replace those destroyed or not collected during
pre-nourishment survey
Standalone wave data collection package replaced with new package
Digital weather station installed
Standalone wave data collection package replaced with new package
Placement of nourishment material completed
First post-nourishment survey:
Wading/swimming profiles (Gulf side), offshore bathymetry
Sand samples, photos
Standalone wave data collection package replaced with new package
Ft. Pickens pier tide gage re-surveyed
Wading/swimming profile survey (Gulf side)
Sand samples
Wave gage cable re-buried
Wading/swimming profile survey (Gulf side)
Sand samples
Shore-connected wave gage removed; cable cut
Standalone wave gage installed
Wind vane and anemometer replaced
Standalone wave gage removed
Fresh standalone wave gage installed
Wading/swimming profile survey (Gulf side)
Sand samples, photos
Reattached Ft. Pickens pier tide gage
Yearly survey: Wading/swimming profiles (Gulf side), offshore bathymetry
Installed heavyweight data/power cable for wave gage
Replaced standalone wave gage near Ranger Station
Installed shore-connected wave gage near Ranger Station
Sand samples, photos
Replaced shore-connected wave gage near Ranger Station

Date ITask

1/16-1/22/92

4/15-4/16/92

7/8/92

10/17-10/20/92

10/27-10/29/92

1/22-1/25/93

5/14-5/18/93

6/23/93
8/3/93
9/8/93

10/15-10/17/93

11/13-11/16/93

12/8-12/10/93

Wading/swimming profiles (Gulf side)
Replaced wind vane/anemometer
Replaced shore-connected wave gage near Ranger Station
Replaced standalone wave gage near Ranger Station
Replaced shore-connected wave gage near Ranger Station
Removed standalone wave gage from Ranger Station
Installed new standalone wave gage near Caucus Shoal
Replaced shore-connected wave gage near Ranger Station
Replaced standalone wave gage near Caucus Shoal
Yearly survey: Wading/swimming profiles (Gulf side)
Replaced shore-connected wave gage near Ranger Station
Replaced standalone wave gage near Caucus Shoal
Sand samples, photos
Replaced weather station
Yearly survey: Offshore bathymetry (Gulf side)
Bathymetric survey of "Profile Nourishment"
Wading/swimming profiles (Gulf side)
Cleaned shore-connected wave gage near Ranger Station
Replaced standalone wave gage near Caucus Shoal
Reset weather station
Wading/swimming profiles (Gulf side)
Bathymetric survey of 8 lines along "Profile Nourishment"
Bathymetric survey of "Profile Nourishment"
Wading/swimming surveys of beach cusps
Cleaned shore-connected wave gage near Ranger Station
Replaced standalone wave gage near Caucus Shoal
Reset weather station
Replaced shore-connected wave gage near Ranger Station
Replaced shore-connected wave gage near Ranger Station
Replaced shore-connected wave gage near Ranger Station
Replaced standalone wave gage near Caucus Shoal
Yearly survey: Wading/swimming profiles (Bay side)
Reset weather station
Yearly survey: Wading/swimming profiles (Gulf side)
Sand samples, photos
Reset weather station
Yearly survey: Offshore bathymetry (Gulf side)
Bathymetric survey of "Profile Nourishment"

3.1 Site Description

The nearshore nourishment project is located within the Gulf Islands National Seashore at

Perdido Key, Florida. Perdido Key is a barrier island on the Gulf of Mexico coast near the

Florida-Alabama state border. The island is narrow in the North-South direction extending in

length from Pensacola Pass on the East to Perdido Pass on the West. Figure 3.2 shows the study

area and the data collection elements. The first portion of borrow material was used for beach fill

along the eastern 7 km of the island between DNR Monuments R-42 and R-64. The main focus

of this chapter was the second phase of the nourishment in which 3 million m3 of additional

dredged material was placed offshore of the eastern half of the nourished beach, between DNR

Monuments R-50 and R-60. The water depths at the dump site ranged from 6.5 m on the western

end of the construction site to 5 m towards the East where the disposal berm extends onto Caucus

Shoal. The deposit commenced at an approximate offshore distance of 500 m and extended up to

1300 m from the post-nourishment shoreline. In the horizontal plane, the disposal site had a

shore-parallel length decreasing from 4 km at the shore-facing side to 2 km at the seaward side

and a cross-shore width increasing from 300 m at the western end to 800 m towards the East. The

horizontal dimensions of the disposal area are shown in Figure 3.3 as reported by the dredging

contractor.

The nearshore bathymetry between DNR Monuments R-50 and R-60 is characterized by a

steep foreshore slope of 1:20 extending to 5 m depth followed by an abrupt change in the profile

slope to 1:1000 as idealized in Figure 3.4. The disposal berm was constructed on this rather flat

portion of the offshore zone. Towards the eastern project boundary, offshore profiles become

steeper and concave upward despite the approximately linear profiles elsewhere. Both

differences are related to the presence of the Caucus Shoal where the water depths are shallower

than the other areas within the disposal site.

METEOROLOGICAL

APPROXIMATE LIMITS Of
PROFILE NOURISHMENT

1 2 3 4 5 km

WAVE GAGE
NOTE:
R-40 is FLORIDA DEPARTMENT OF NATURAL
RESOURCES MONUENTED "RANGE 40"

Figure 3.2: Components of field monitoring program.

/

----------------------------------_^ ^^^^ ^^^ ^^^ ^^ 0

0

E -500

. -1000
C

a. -1500
0
-
S-2000

-2500
4

48

Disposal Area

. . . . ..
-F
'I '

50 52 54 56
Range No.

Figure 3.3: Planview of disposal area.

58 60

1:20
disposal berm

1:1000

Figure 3.4: Idealized profile cross-section at disposal area.

3.2 Wave. Current and Tide Measurements

Waves, currents and tides were measured at two stations, one located at 7 m water depth

approximately 750 m offshore from the Ranger Station near DNR Monument R-34, and the other

on Caucus Shoal at 5 m water depth approximately 1 km offshore from R-62. The gage at Ranger

Station has been operating since January, 1990. The second gage was installed in April, 1992 at

Caucus Shoal to focus on the local wave and current climate in the vicinity of the shoal and the

effects of the Entrance to Pensacola Bay. This gage was located near the Southeast corer of the

N

"""""'"'"""'"'

nearshore disposal area. Therefore it should identify wave and current effects acting directly on

the nearshore berm.

Both stations comprise P-U-V type gages mounted on tetrapod shaped steel frames placed on

the ocean floor. The P-U-V gages consist of two electronic sensors; a pressure transducer and an

electromagnetic current meter. Raw data include a pressure signal and two mutually

perpendicular velocity signals measured in a horizontal plane perpendicular to each other. All

signals are collected once every 6 hours at 1 Hz sampling frequency for a 17 minute duration.

The gage near the Ranger Station is connected by a cable to a shore station which serves as a

link for the remote control operation. The P-U-V data from this gage can be retrieved from the

University of Florida via telephone. The gage near Caucus Shoal carries a self contained storage

device which can store data until divers retrieve the package every 3-4 months. Additional

information about the wave packages can be found in Work (1992).

The P-U-V data are analyzed using directional spectrum methods to obtain wave height,

wave period, wave direction, tide and current velocity and current direction. Wave related

parameters such as wave period, significant wave height, modal wave direction and the spreading

parameters were obtained from a directional spectrum analysis of the P-U-V signals. Current and

tide variables such as the current velocity, mean current direction and the tidal elevation were

calculated from statistical values in the time domain. The complete set of wave, current and tide

parameters covered a period of January 1990 to May 1994. In this study, only the time frame

from the completion of the nearshore berm construction in October, 1991 through the last

bathymetric survey in December, 1993 will be evaluated. Results from this time period are

presented in Figures 3.5 through 3.20.

The time history of wave, current and tide parameters in Figures 3.5 to 3.8 indicated different

levels of coverage at the two gages. The shore-connected gage at Ranger Station (RS) operated

more consistently than the stand-alone gage at Caucus Shoal (CS) which included long periods of

no data between installation of successive wave packages (Figures 3.6 and 3.8). A comparative

study on continuity of pressure and current related data proved that the pressure transducer was

the most reliable element of the gage and due, in part, to biological fouling the current meter was

the least reliable. Therefore the availability of wave, current and tide parameters varied

depending on the particular sensor the data were originated from. The pressure sensor provided

the basis for determining wave height, wave period and tidal stage whereas the current meter

provided the basis for establishing wave direction and current related information. The vertical

inconsistencies in the tidal elevation shown in Figures 3.7 and 3.8 were caused by incompatible

calibration coefficients and offsets of successive pressure transducers.

The theoretical background of the spectral and statistical analysis performed on the time

series of pressure and current data can be found in Ochi (1990). In this section, only the results of

the data analysis will be discussed and not the method itself.

Significant wave height (H,) was calculated from the total spectral energy of the water

surface displacement as interpreted from the pressure measurements. Figures 3.9 and 3.10 show

histograms of H, at wave gages RS (Ranger Station) and CS (Caucus Shoal). The most probable,

maximum and mean values, 0.28 m, 2.91 m and 0.58 m respectively, at RS, were higher than

their counterparts at CS, 0.21 m, 1.39 m and 0.46 m respectively. Similarly the most probable,

maximum and mean wave periods at RS, 7.3 s, 12.8 s and 6.4 s respectively, were higher than

their counterparts at CS, 4.3 s, 10.6 s and 6.2 s respectively.

Information about the location and configuration of the two gages and the wave and current

statistics for the period from October, 1991 to December, 1993 were summarized in Table 3.2.

Ranger Station: Significant Wave Height

0 100 200 300 400 500 600 700 800 900 1000

Ranger Station: Peak Wave Period
15i -- ..------- i -----,--- i --
15

W 0 -- -- --- -- --- -- --- -- --- --
5-5

0
0 100 200 300 400 500 600 700 800 900 1000

Ranger Station: Wave Direction
400

0
0 100 200 300 400 500 600 700 800 900 1000
Days after Nearshore Nourishment

Figure 3.5: Time history of height, period and direction of waves at Ranger Station.

Caucus Shoal : Significant Wave Height

Caucus Shoal : Peak Wave Period

Caucus Shoal : Wave Direction

180 200 220 240 260 280 300 320 340
Days after Nearshore Nourishment

Figure 3.6: Time history of height, period and direction of waves at Caucus Shoal.

Ranger Station: Mean Current Velocity

100 200 300 400 500 600 700 800 900 1000

Ranger Station: Current Direction

100 200 300 400 500 600 700 800 900 1000

Ranger Station: Mean Water Level

0 100 200 300 400 500 600 700 800 900 1000
Days after Nearshore Nourishment

Figure 3.7: Magnitude and direction of mean current and tidal stage at Ranger Station.

Caucus Shoal : Mean Current Velocity

Caucus Shoal : Current Direction

Caucus Shoal : Mean Water Level

180 200 220 240 260 280 300 320 340
Days after Nearshore Nourishment

Figure 3.8: Magnitude and direction of mean current and tidal stage at Caucus Shoal.

Ranger Station: Histogram of Significant Wave Heights

0 0.5 1 1.5 2 2.5
Hs [m]

Figure 3.9: Histogram of significant wave heights at Ranger Station.

Caucus Shoal : Histogram of Significant Wave Heights
18

16
Peak = 0.21 m
Max = 1.39 m
Mean = 0.46 m

712-

T10

4-
a 6
C;

1 11_1_ 1__nI
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Hs[m]

Figure 3.10: Histogram of significant wave heights at Caucus Shoal.

215

g10
0a

Ranger Station: Histogram of Wave Periods

Tp [s]

Figure 3.11: Histogram of representative wave periods at Ranger Station.

Caucus Shoal : Histogram of Wave Periods
190.....

0 5 1611 -7 8- -U-
4 5 6 7 8

Peak= 4.3 s
Max = 10.6 s
Mean = 6.2 s

9 10 11

Tp [s]

Figure 3.12: Histogram of representative wave periods at Caucus Shoal.

12

o
10
0
8
.

The comparison of results at two locations indicated significant differences in the wave

directions (Figures 3.13 and 3.14). It should be noted that the times of the two records are not

exactly the same as shown in Figures 3.5 and 3.6. The predominant wave directions at RS and CS

were calculated as 345 N and 300 N respectively. Knowing that the shoreline between the two

stations makes an angle of 800 with the North, the predominant wave directions at RS and CS

were skewed 50 and 50 respectively westward from the shore-normal.

Figures 3.15 and 3.16 show histograms of current velocities. Contrary to the surface waves

the near-bottom current velocities at CS were higher than the currents at RS. The most probable,

maximum and mean current velocities at CS were 0.11 m/s, 0.34 m/s and 0.12 m/s respectively

whereas their counterparts at RS were 0.04 m/s, 0.28 m/s and 0.08 m/s. Although the maximum

current velocity at RS was 0.46 m/s the next highest value was 0.27 m/s. The comparison

indicated stronger current climate near Caucus Shoal which was generally controlled by the tidal

flows at the Entrance of Pensacola Bay whereas the weaker currents at Ranger Station were

dominated by the shoaling waves. The currents at RS were distributed over all directions with

some concentration in the landward facing half of the spectrum (Figure 3.17). This onshore

directed group of currents was associated with moderate current velocities (Figure 3.19). The

combination of frequent occurrences and moderate current velocities makes this group an

important component of the overall current climate at Ranger Station. A single peak at 1350N

contributed substantially to the total energy due to the very high current velocities associated

with that particular direction. By comparing Figure 3.17 to Figure 3.19 it can be concluded that

the most energetic current direction at RS was onshore with a narrow southeasterly component.

On the other hand, current directions at CS were confined in a narrow zone around a peak at 3000

N which is equivalent to 200 eastward from the shore-perpendicular.

37

Ranger Station: Percent Occurrence of Wave Directions
N20
Peak = 345.0 Deg.N ........20....
: 16

4

Caucus Shoal : Percent Occurrence of Wave Directions
40
Peak =300.0 Deg.N.......
o 30

20
w E
........ .. ..

S. .... ..... .... ... .

S

Figure 3.14: Polar histogram of wave directions at Caucus Shoal.
Fiue .4 Plr itgrmo'wv "2"iretosa acsSol

Ranger Station: Histogram of Current Velocities
I .. .

0.1 0.15 0.2
Um [m/s]

0.25 0.3 0.35

Figure 3.15: Histogram of current velocities at Ranger Station.

Caucus Shoal : Histogram of Current Velocities

0.05 0.1 0.15 0.2 0.25 0.3
Urn [m/s]
Figure 3.16: Histogram of current velocities at Caucus Shoal.

Peak = 0.04 m/s
Max = 0.28 m/s
Mean = 0.08 m/s

n nnnnnH -

39

Ranger Station: Percent Occurrence of Current Directions

Deg.N

S

Figure 3.17: Polar histogram of current directions at Ranger Station.

Caucus Shoal : Percent Occurrence of Current Directions
N
20 Peak = 30.0 Deg.N

W ...... .. ....... .. .... ........ ...... ...... E
W ~ ~ ~ ~ ~ ~~i .....dz .: ....:............

s

Figure 3.18: Polar histogram of current directions at Caucus Shoal.

Ranger Station: Magnitude and Direction of Current
0.5

0.4

0.3

S0.2

-0.1

00

z -0.2

-0.3

-0.4

-0.5 0 0.5
East-West Currents [m/s]

Figure 3.19: Directional distribution of current velocities at Ranger Station.

Caucus Shoal : Magnitude and Direction of Current

0.4

0.3

S0.2

0.1

00

0
o
y -0.1

0
z -0.2

-0.3

-0.4

S-

0 0.5

-0.5 0 0.5
East-West Currents [m/s]

Figure 3.20: Directional distribution of current velocities at Caucus Shoal.

Table 3.2: Wave gage statistics from 10/1/91 to 12/1/93.
Ranger Station Caucus Shoal
Operational since January, 1990 April, 1992
Type of Gage CDN SeaData
shore-connected stand-alone
Latitude N 300 17.60' N 300 18.60'
Longitude W 870 24.73' W 870 19.21'
Distance from Shore [m] 750 1,000
Water Depth [m] 7 5
Height of Pressure Transducer
above Bed [m] 0.53 1.3
Most
Significant Probable 0.28 0.21
Wave Height [m] Max. 2.91 1.39
Mean 0.58 0.46
Most
Representative Probable 7.3 4.3
Wave Period [s] Max. 12.8 10.6
Mean 6.4 6.2
Predominant Wave
Direction [0 North] 345 300
Height of Current Meter
above Bed [m] 1.66 1.37
Most
Current Velocity Probable 0.04 0.11
[m/s] Max. 0.28 0.34
Mean 0.08 0.12
Predominant Current
Direction [o North] 135 30

3.3 Sediment Size Distribution

Sand samples were collected at eight locations along the profiles: Dune, mid-beach, berm,

beachface, -1 m, -2 m, -5 m and -8 m. Details about the sampling locations and methodology

were given by Work (1992). Grain size distributions have been determined by mechanical sieve

analysis of each sample, using a series of twelve U.S. standard sieves with mesh numbers 10, 20,

30, 40, 50, 60, 70, 80, 100, 120, 140 and 160. Various characteristic parameters of the grain size

statistics have been analyzed, such as the median diameter (ds,), mean diameter, sorting index,

skewness and kurtosis. Only the d, results are presented here; other parameters are available on

disk.

Sediment characteristics (ds) are summarized in Figure 3.21 for the dune, mid-beach, berm,

beach face, 1 m, 2 m, 5 m and 8 m samples. The solid line in each of these figures represents the

median diameters from the sampling in November, 1993 and the dotted lines represent the

envelope of the measured size distributions including the native sand before the beach

nourishment. Knowing that the nearshore berm was located between Range 50 and 60 in the

longshore direction and the water depths from 5 m to 8 m, results from these locations should be

interpreted parallel to the evolution of the berm. In this region the wet samples of 1993 had

sediment sizes nearer to the lower limit of the envelope than the upper limit. An opposite trend

was observed for the landward samples collected between the dune up to the -1 m. Here the size

distributions of 1993 sediments fell between the upper and lower limits of the envelope. Figure

3.22 shows a similar decrease in sediment size at 5 m and 8 m water depths which is more

consistent than the decrease at other locations.

Fine sediment can influence significantly the performance of a nearshore nourishment

project. Nourishment material finer than the native sand is usually considered of lesser quality. In

addition to recreational advantages and biological concerns, fill material of equal or coarser grain

sizes as the native sand is more efficient in terms of additional dry beach length and the project

life of the beach and profile nourishment. Details of the beach nourishment concept and the

effects of sediment size selection is beyond the scope of this dissertation. For further information

readers may refer to Dean (1983) and Dean (1988).

Spatial Distribution of Median Diameter
Dune Mid Beach

0.5 0.5

E .... E

0 0.2 0 0.2
0.1 0.1

o0 40 50 60 s0 40 50 60
Range Number Range Number
Berm Beach Face

0. .5 .

E 0
S a 0 .3o i 4
0.2 0 0.2
0.1 0.1

30 40 50 60 30 40 50 60
Range Number Range Number

Spatial Distribution of Median Diameter
-1m -2m

0.5 .-- 0.5
00.0.4 .
-. '0.3
0 0.2 00.2
0.1 0.1

0 40 50 60 30 40 50 60
Range Number Range Number
-5 m -8 m

0.5 0.5
'0.4 .. :". 0.4
E 0.3 'EZ;0.3
00.2 '.. -. 0.2 V

0.1 0.1

o0 40 50 60 0 40 50 60
Range Number Range Number

Figure 3.21: Longshore distribution of D50 for November, 1993 (solid line) with envelope
(dashed lines) of sizes for 1989, 1990, 1991, 1992 and 1993.

--~--~-~--~-~I

Crosshore Distribution of D50
Longshore averaged for All Years
0.5

E 0.4
E 0.4 ---------- -- -------- --- -
a1)
.0.3 -

: 0.2

a)
U 0.1
dune berm -1 m -5 m
mid-beach beachface -2 m -8 m

D Nov.'89] Sep.'90 Oct.'91 Oct.'92 Nov.'93

Figure 3.22: Longshore averaged cross-shore distribution of D5. Temporal variation from
November, 1989 to November, 1993.

Fine sediments in the Perdido Key Project were located primarily between Ranges R-42 and

R-58. The distribution of fines, as shown in Figures 3.23 and 3.24 respectively, for the various

sampling events at 5 m and 8 m water depths, appears to be decreasing with time. The origin of

these fines is the Pleistocene mud deposits that were excavated in the dredging operation. The

decrease of fines with time is to be expected due to suspension during energetic wave events

which causes suspension of the fines and distribution over wide areas. However the

interpretation of the suspension and transport of fines must be tempered with the understanding

that the distribution can be somewhat "spotty" due to concentrating in local depressions and thus

the data must be interpreted in the "aggregate" sense rather than on the basis of individual

samples.

45

5 m Sand Samples

Percentage Finer than 0.0097 mm

100

80

" 60
.*--

o 40

20

0

S- --------- -------- ----- -------------i i
----------- ----- -- --- -------------

-- - - - - -

30 34 38 42 44 46 50 54 58 61 63 65 67
Range Number increasing towards East

E Nov.'89 = Sep.'90 I Oct.'91 Oct.'92 M Nov.'93

Figure 3.23: Percentage of fines for 5 m samples from November, 1989 to November, 1993.

8 m Sand Samples
Percentage Finer than 0.0097 mm

30 34 38 42 44 46 50 54 58 61 63
Range Number increasing towards East

65 67

Figure 3.24: Percentage of fines for 8 m samples from November, 1989 to November, 1993.

100

- 80
a,

IL 60
C
*

| 40
a)
S20

n.

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

I I Nov.'89 1- Sep.'90 M Oct.'91 i Oct.'92 i Nov.'93

'"~ "I"- ~-

'-i--I

'

I

u

-- I I i ----- ---
-- -- -- -- -- -- -- -

3.4 Weather Station

Throughout the 5 year monitoring project, three consecutive weather stations were used to

collect meteorological data at the Perdido Key Ranger Station. The first unit was a mechanical

weather station which recorded analog data for six months commencing in January 1990. In June

1990 it was replaced with a digital unit which was later replaced with a similar instrument. The

current weather station has been operating since October 1992 with a locally installed data

acquisition unit and a storage device. The station can be controlled remotely from the University

of Florida and the stored data can be retrieved via telephone. The data acquisition unit consists of

electronic sensors to measure wind velocity, wind direction, air temperature and rainfall. Data

are sampled at 1 Hz frequency and the minimum, maximum and mean values over a user defined

interval are saved in the storage device. The optimum interval was found to be two hours

considering the storage capacity and the retrieval time of stored data in case of system

malfunction. Approximately once a week, stored data are retrieved into a computer at the

University of Florida via telephone.

Wind speed and wind direction are measured approximately 3 m above the roof of the

Ranger Station at an elevation of 11.7 m-NGVD. Although the original sampling intervals ranged

from 15 min. to 2 hr the data are reduced to a uniform set with 2 hr interval for the statistical

analysis. Within the scope of this dissertation, only results from a 2 years time frame since the

completion of the nearshore berm construction in October, 1991 are presented. Figure 3.25

shows the time histories of air temperature, wind velocity, wind direction and rainfall for the

period of October, 1991 to December, 1993. The gaps between the data lines indicate periods of

no data due to either malfunction in the data acquisition system or interruption of data retrieval

during severe weather conditions accompanied by lightning.

Perdido Key Weather Station

E 10
?5-

II.... I, A

0-
400 500 600 700 800 900 1000
Days after Nearshore Nourishment

Figure 3.25: Time history of air temperature, wind velocity, wind direction and rainfall.

Figure 3.26 shows the distribution of wind velocities in 16 magnetic wind directions. The

length of arrows from the origin indicates the wind velocity whereas the arrow head points

towards the direction in which the wind blows. The number of arrow heads in each of the sixteen

zones indicate the density of occurrences in that particular direction. The strongest winds blew

~

4'.

~?

4

from the Northwest quadrant whereas the winds from the East South East were the weakest in

magnitude.

Magnitude and Direction of Wind

6-

4-

2

0-

0-2-/ ,
C1)

-6 -

-8 -

-5 0 5 10
East-West Velocities [m/s]

Figure 3.26: Directional distribution of wind velocities.

Histograms of wind velocity and wind direction are shown in Figures 3.27 and 3.29

respectively. The statistical distribution of wind velocities indicated a most probable velocity of

2.9 m/s with 18% probability of occurrence. Maximum and mean velocities were 10.7 m/s and

3.1 m/s respectively. Directional histogram indicated a clear predominant wind direction from

the North with 18% probability of occurrence. Winds from other directions were distributed

rather uniform with a slightly more chance for Easterly winds.

49

Histogram of Wind Velocities

2 4 6 8
Wind Velocity [m/s]

Figure 3.27: Histogram of wind velocities.

Percent Occurrence of Wind Directions
Peak= Deg. N N18
Peak= 0.0 Deg.N 18

. ..... E

Figure 3.28: Polar histogram of wind directions.

50

3.5 Photographic Documentation

Oblique color ground photography has been taken throughout the study to document changes

as the nourished beach evolved. Photography is conducted in conjunction with each survey.

Three photos are generally taken at each transect, viewing to the left along the beach,

perpendicular to the beach and to the right along the beach. The reader may contact the author

regarding availability of the photographs.

CHAPTER 4
SURVEY ACCURACIES AND IMPROVEMENT TECHNIQUES

In certain coastal engineering problems such as sediment transport or sediment budget

calculations, the accuracy of the measurement technique can be critical to the understanding of

the physical processes under investigation. In the extreme case, if the scale of change of the

measured quantity is in the same order of magnitude as the measurement accuracy, the reliability

of the findings becomes questionable. The importance of a detailed knowledge about the survey

accuracies has led to a series of field and data analysis investigations to test some of the more

commonly used positioning instruments.

In Sections 4.1 and 4.2, accuracy and reliability ranges are established and possible

approaches for improvement are discussed. Existing techniques are reviewed and new

post-survey techniques based on digital algorithms are introduced. One of these algorithms is

developed in Section 4.3 for surveying nearshore bathymetry with three fixed stations in a

triangular geometry and one mobile unit. This method can be applied at the post-processing stage

but also in real time during the data acquisition. The method applies for all geometrically

possible situations and provides the most probable position even for a case with an indefinite

solution domain. Another advantage is that at each position, an error radius is calculated and

provides a measure of uncertainty.

A numerical technique is introduced in Section 4.4 to correct the errors in the horizontal and

vertical positions measured by a boat/fathometer technique. This method operates as a

post-processing adjustment by applying independent calibration parameters to the horizontal and

vertical data. A portion of the profile must be overlapped by wading/swimming survey which

serves as a reference bathymetry to determine the optimum adjustment parameters using a least

squares procedure. The advantage of this method is its ease of use and minimum operating cost

compared to other hardware dependent techniques discussed in the text.

4.1 State of the Art in Nearshore Surveying

A comprehensive review of existing horizontal positioning techniques for hydrographic

surveys is given by Hart and Downing (1977). They considered systems based on various

technologies including microwave, laser, optics, acoustics and satellites and evaluated the

following different criteria: Cost, accuracy and ease of use. Their findings provide a useful

guideline for hydrographic surveyors.

Other studies have focused on improvement of vertical accuracy in hydrographic surveys.

Downing and Fagerburg (1987) have evaluated several techniques including a pendulum-

stabilized accelerometer based heave compensation system (HIPPY 120), Doppler equipment,

vertical displacement measurement equipment using automatic electro-optical tracking systems,

video-type optical tracking systems, laser leveling systems and satellites. Despite technological

problems, all of the techniques evaluated were found promising for future considerations.

ASCE Hydrographic Investigations Committee (1983) categorized Electronic Distance

Measurement (EDM) equipment in different operating frequencies and reported corresponding

range limitations and accuracies. As a general rule, the range of EDM systems which is equal to

the penetration distance of the electromagnetic wave, decreases with increasing frequency due to

higher energy losses at higher frequencies. On the other hand the accuracy increases with

frequency due to improved time resolution at shorter wave lengths. Table 4.1 summarizes the

range, resolution, accuracy and repeatability variations for different frequency bands. An

exception to the general trend is noted in the higher frequency bands; however, no explanation

was given in the ASCE report about this non-monotonic behavior.

Table 4.1: General characteristics of EDM equipment (from ASCE Hydrographic Investigations
Committee, 1983).
Frequency Range Resolution Accuracy Repeatability
10-20 kHz 2400 km 120 m 1.6 m 1.6 km
100 kHz 1900 km 15 m 0.4 m 0.4 km
1.7-3.3 MHz 500 km 0.5 m 10 m 3-4 m
2.8-3.2 gHz Line of sight 0.1 m 1 m 0.5 m
9.3-9.5 gHz Line of sight 1.0 m 3m 1.5 m

Accuracy and cost effectiveness are the primary considerations in choosing from several

state-of-the art techniques for hydrographic surveys. A combination of wading/swimming and

boat/fathometer survey is by far the most common technique due to the minimal costs and

general availability of required equipment. For this technique, surveys are conducted in two

stages. First, the nearshore profile is surveyed using standard rod and level procedures placing

the rod by walking/wading/swimming. The survey starts from a known benchmark on land and

extends into the sea up to a limiting depth of approximately 4-5 m. The main advantages of this

procedure are its simplicity in practice and its high accuracy. Secondly, the offshore portion is

surveyed by boat measuring the vessel's horizontal position and water depth simultaneously. For

this purpose two separate types of instruments operating independently from each other are used

to measure the horizontal and vertical positions. It is a common procedure to calibrate the

equipment before and perhaps during and after each survey but considering various factors

affecting the survey accuracy such as the water temperature, salinity, changing tide, draft of the

boat etc., an optimum setting of the instruments may not be established in the field. A

post-adjustment may be necessary to account for changes in the above mentioned conditions.

Other, more accurate methods exist for conducting the seaward portion of the survey,

including the sea sled and a self propelled wheeled vehicle, the CRAB, each of which supports a

survey prism. However, these are generally slower than boat surveys and require equipment not

commonly available. Clausner et al.(1986) have reviewed accuracies obtained with four different

types of survey equipment and have found that the boat/fathometer survey has an accuracy and a

repeatability error of 22.6 cm and 9.1 cm respectively (Table 4.2). The vertical accuracy of

fathometer measurements is quite dependent on wave conditions. The vertical accuracies of the

hydrostatic profiler, CRAB and Sled were approximately 2 cm. Clausner et al. (1986) have

measured similar on-line accuracies for different techniques on order of 0.5 to 3.5 m which

would indirectly affect the vertical accuracy in case of longshore irregularities in the bottom

topography.

Table 4.2: Comparison of vertical and horizontal accuracies (from Clausner et al., 1986).
Vertical [cm] Off-line [m]
Accuracy Repeatability Mean RMS
Sled used as reference 1.2 1.83 1.55
CRAB used as reference 1.8 0.49 0.40
Hydro. Profiler 1.8 2.7 2.38 3.57
Boat/Fathometer 22.6 9.1 1.80 1.31

4.2 Field Testing of Horizontal Distance Measurement Intruments

Field investigations have been carried out to establish the accuracy and repeatability of four

different horizontal distance measurement instruments. All four instruments: The survey tape,

optical range-finder, an infrared based Omni and microwave based MiniRanger units are standard

field equipment used in wading/swimming and boat surveys. Tests were conducted on a clear day

over a 1 km distance of flat land. The three transponder units of the MiniRanger system and the

reflector of the Omni were placed along an established baseline. The receiver unit of the

MiniRanger system, the Omni and the range finder were located at different points from the

origin. At each point, distances from the origin were measured using the tape, Omni, range-finder

and the MiniRanger. Omni readings were used as reference to calculate the accuracy for the other

instruments since the Omni was the most precise unit in the group with 5 mm manufacturer

claimed error. The following sections describe test results measured at 8 stations located along

the straight baseline. Each unit was tested separately for systematic and repeatability errors.

A standard 50 m floatable survey tape was used to measure the distance of each station from

the baseline. Deviations from the Omni readings are plotted in Figure 4.1. Repeatability was not

tested separately since it was already incorporated in the results due to the repeating nature of the

measurement technique every full tape length. Although the Omni is more accurate than the

survey tape the possibility may not be dismissed that part of the measured error originates from

the Omni readings.

Survey Tape
Accuracy and Repeatability
0.75.
I Tape- Omni

0.5.

^ 0.25-

-0.25
-0.25 I -II I I --- I
0 200 400 600 800 1000
Actual Distance [m]

Figure 4.1: Test results for survey tape.

At each station three independent readings were taken by different members of the survey

crew using two range-finders. Deviations from the Omni distances are plotted in Figure 4.2,

indicating the systematic error of the units at different distances. Similarly standard deviations of

readings taken by different crew members are shown in the same figure. These values represent a

measure of repeatability error. At distances smaller than 500 m both the accuracy and the

repeatability remain within 7 m. Beyond this range both values experience a sudden increase and

exceed the limit of usuability mainly due to optical limitations of range-finders. The maximum

distance measured with the range-finders in a wading survey is around 200-300 m.

Range Finder
Accuracy and Repeatability
40

30- -

0
I10--

-10- I I l I
0 200 400 600
Actual Distance [m]

F RF47577-Omni 7 RF47577:Std.Dev. -A.- RF47511- Omni I RF47511: Std.Dev.

Figure 4.2: Test results for range-finders.

MiniRanger distances were transmitted by three stationary transponders and recorded for

approx. 5 minutes at a sampling frequency of 1 Hz. The time average of each record was

calculated to find the deviation from the Omni reading. The standard deviation of the 5 minute

record is considered as the repeatability error. Figure 4.3 shows systematic and repeatability

errors at different distances. Systematic errors for all three units indicate a slight increase with the

measured distance. The maximum systematic error was around 3 m. However, the repeatability

error was not sensitive to the distance. It remained mostly unchanged (0.38 m) through the

experiment. For real-time data acquisition, where each point in space is measured by a single

burst of data point in time, the repeatability error is as important as the systematic error since the

effective accuracy will be the sum of both errors.

Mini Ranger
Accuracy and Repeatability
4
3
4-----------

2-
80

-2 I I I I I
0 200 400 600 800 1000
Actual Distance [ml

-....- Code Omnl -.m- Code 4-Omnl Code Omni Code 1:Std.Dev. C Code 4: Std.Dev. o Code 8: Std.Dev.

Figure 4.3: Test results for MiniRanger.

Experimental findings are summarized in Table 4.3. Accuracies were distinguished as

systematic and repeatability errors. Systematic error indicates the deviation from the Omni

readings whereas the repeatability error was found from the standard deviation of multiple

measurements of the same kind. Next to Omni the best performance within a range of 1 km was

achieved by the survey tape. MiniRanger units were second best with an average systematic error

of 1.36 m and an average repeatability error of 0.38 m. Within a range of 200 m range-finder

accuracies were comparable to MiniRanger accuracies. Beyond 700 m no range-finder readings

were taken due to difficulties in focusing on the target. A comparison with Table 4.1 is possible

for the MiniRanger which falls into the 'Line of Sight' category.

Table 4.3: Features of horizontal distance measurement instruments.
Horizontal Systematic Repeatability Factory Settings
Errors Mean RMS Mean RMS Frequency Range Resolution Accuracy

[m] [m] [m] [m] [Hz] [m] [m] [m]
Omni ref. ref. 10'2-5x104 0-5,000 0.001 0.005
Survey tape 0.18 0.25 0-50
MiniRanger 1.36 1.09 0.38 0.28 (5.4-5.6)x109 30-37,000 0.03 2.8-7.7
Range-finder 8.34 12.37 4.99 7.05 50-500

4.3 Horizontal Positioning by Triangulation Method

Triangulation is one of the most frequently used horizontal positioning techniques in

bathymetric surveys. The idea is based on simple geometry such that any point in the 2-D plane

can be described by its distance from two known points. In general this description does not

describe a unique point since the two arcs drawn from each known point intersect in two points.

In analytic geometry this problem leads to solution of a quadratic equation (equation for a circle).

Quadratic equations may have no, one or two real solutions depending on the discriminant. By

extending the problem with a third known distance the solution becomes unique since there is

only one point where all three arcs will intersect. We can apply this method to a horizontal

positioning problem in a bathymetric survey where a set of three stationary transponders are used

to obtain the horizontal location of the survey vessel. A 'master' unit in the vessel communicates

with three stationary transponders and records the distance to each transponder at a certain

sampling frequency. The transmission is made through microwaves so that the time lapse, At,

between the first reported distance and the third reported distance is in the order of 105 seconds.

At= distance 3103m 10-5 (41)
speed of light 3.108 m
s

The following section presents a method of geometric analysis to obtain the horizontal

coordinates of a point for which a set of three distances are reported from three stations with

known coordinates. At this stage we have to consider that the reported distances are not exact

due to various possible causes including impreciseness of the transponder locations, interference

of signal, etc. The additional difficulty introduced by the error in arc radii requires a special

technique to uniquely determine or estimate the best possible position of the unknown point.

Figure 4.4 shows a generalized case of three arcs which do not have a unique crossing point

indicating some uncertainty in the transmitted distances.

St-i

R-R

R-2
SSt-3

St-2-------------------------------St-3

Figure 4.4: Schematic triangulation setup.

Figure 4.5 shows all possible configurations in which the three arcs can occur. As can be

seen the region between the arcs is the area of uncertainty but also the solution domain since the

exact solution lies somewhere inside. A direct way to determine the most probable solution

would be the center of gravity of that area. But this area is not always finite as in Figures 4.5c,d,e

and therefore the solution cannot be defined uniquely.

(c)

Figure 4.5: Possible triangulation errors.

If (Xi,Yi) are the coordinates for station Sti and the distance to the survey vessel is Ri then the

three arcs can be described in Cartesian Coordinates as,

Arci: (x-Xi)2 + (y. y)2 = R2 i= 1,2,3 (4.2)

The following technique avoids the problem shown in Figure 4.5 by reducing the nonlinear

equations which are the source of the indefinite solution domain into linear equations with finite

solutions. Each of the linear equations defines geometrically a line (dashed lines in Figure 4.6)

which is perpendicular to the line connecting the two stations and crosses through the intersection

points of the two arcs drawn from those two stations. If the two arcs don't intersect the line will

pass through a point equally spaced from both arcs.

Line.2: A,2 x + B12 y = C12

(b)

(4.3a)

Linel.3

Line2.3

A13 x + B13 y = C13

A23 x + B23 y = C23

(4.3c)

After solving the three linear equations simultaneously we obtain three solutions each of

them indicating an intersection point of two of the lines such that S, for Line.2 and Line.3, S2 for

Line?.2 and Line2., S3 for Line1.3 and Line2.. Using the following substitutions,

Ai =Xi -Xj (4.4)

Bii = Yi Y, (4.5)

(4.6)

where Xi and Yi are the station coordinates. The intersection coordinates simplify to:

s B13C12-B12C13 B23C12-B2C23 B23C3-B3C23 (4.7ab,c)
s A12B13-A13B12 2 2B23-A23B2 3 = A13B23-A23B13

A12C13-Al3C12 A2C23-23C2 3C23-A23 3 (4.8a,b,c)
s1 A12B13-A13B12 2 A 12B23-A23B12 3 A13B23-A23B13

The intersection points S, S2 and S3 span a triangle whose center of gravity (Sc) is the most

probable solution of the original equations of circles as shown in Figure 4.6. Therefore the

coordinates of the unknown survey point is best estimated as (X, ,Y, ). The error radius, E is the

(4.3b)

Cij= I(X +, Ri ,- X + R)

Root-Mean-Square value of the three minimum distances between Sc and the three arcs. The s, can

be considered as a measure of accuracy for a particular solution.

The coordinates of the center of gravity (S ) for the triangle are given as,

Xcg = -(Xs1 +Xs2 +Xs3) (4.9)

Ycg =(Ys1 +Ys2 +Ys3) (4.10)

Ide 2-3

Line 1-2

Figure 4.6: Solution domain for the triangulation method illustrated for case (a) in Figure 4.5.

The error radius is given in Equation (4.11).

S[Ri- (Xcg-X)2+Ycg-Y,)2 2
sr = i=13 (4.11)
3

The technique described above was tested in a field study at Little Lagoon, near the city of

Gulf Shores, AL. With the survey setup shown in Figure 4.7, an area of approximately 250,000

m' was surveyed. Collected raw data were analyzed using the triangulation technique described

earlier in the text to compute the (xy) coordinates of the survey vessel. The resulting survey

trajectory and the error radii are plotted in Figure 4.8. The trajectory thickness indicates that the

error radius varies between 0 and 7.1 m with a mean value of 3 m. For this particular setup the

accuracy in horizontal positioning is most sensitive to the distance from Station 2. This

knowledge could provide a basis for refining the coordinates of the stations.

N

SULittle Lagoon

i Gulf of Mexico
Figure 4.7: Planview of Little Lagoon survey site.

64

Survey Trajectory
1000

9 0 0 ...... .......... .. ............. ........... ............ .. . ........ .....: S t.1 -
S* Max Err.Rad.= 7.1 m /

Ca
E 7 0 0 ..... ............ ..... .. .... .. ............ -............ ...... ..... .............
E 700 ...

600 ............
S500
o ::
5 400 .
Z
4 4 0 0 .. .. ............. ... ... 1 .- ..... ... ... .......... .... ...... .............

2 300. .............

100
100 i i i i
-600 -400 -200 0 200 400 600 800
Distance East of Benchmark [m]

Figure 4.8: Survey trajectory and relative positioning accuracy (thickness of trajectory line is
proportional to the error radius).

4.4 Post-Adjustment of Boat Survey Data

Measurement errors in wading/swimming/boat combination surveys may lead to profiles

which differ significantly from the real bathymetry. The accuracy of volumes, calculated from

such data, can be affected substantially. Post-processing is a cost-effective solution to minimize

systematic errors in boat surveys and improve data quality without replacing the existing survey

system. However, certain conditions noted below are required for the conditions to apply.

The method introduced and illustrated by application to a fairly large data set is a purely

mathematical technique based on a least squares analysis. The necessary condition to apply this

method is a profile surveyed by both wading/swimming and boat surveys such that these two

profile segments overlap (see Figure 4.9). The overlap region is used later to adjust the

calibration constants of horizontal and vertical positioning instruments in the offshore profile

segment.

-MWL
wading profile

boat profile

Figure 4.9: Overlap region in a beach profile survey.

The main assumption in this method is that the wading profile represents the true bathymetry

whereas the boat profile deviates horizontally and vertically from the exact profile. A measured

profile is a set of data points, where each point is a pair of horizontal and vertical coordinates

(X,Z)i. We can describe wading and boat profiles as (X,Z,)i and (Xb,Z)i respectively. To the first

order, we assume that the boat data can be corrected using an offset and a linear scale factor.

Since the horizontal and vertical positioning instruments function independently there should be

independent calibration parameters. The corrected boat profile (XbcZb )i can be written as:

Xb= a + bXb (4.12a)

,= A + B Zb (4.12b)

where for no errors in the original data, a=A=0, b=B 1.

~

Using a curve-fitting technique we can express the wading profile, (X,,Z,)i, as an analytic

function f(x) which represents the true bathymetry in a closed form. The choice of the closed

form expression used for the curve-fitting depends on the shape of the profile in the overlap

region. Generally for beach profiles the following function gives a good fit to the measured

depths. For other applications different analytic expressions may be more appropriate.

f(x) = Co+ c, X + c x x213 +cx+c4x2 (4.13)

The deviation of the corrected boat profile (Xb,,Zb,)i from the wading profile (X,,Zw)i can be

expressed as:
S= Zb,c [f()]x=xb,c (4.14)

Substituting our first order corrections, the error at each boat data point becomes a function

of the correction parameters A, B, a, and b only.

si(A,B, a, b)= [A +B. Zbi [f(x)] +b.xb (4.15)

By applying the least square error method we can find optimum values for A, B, a and b

which minimize the error between the wading profile and the corrected boat profile.

82= \A+B.Zb -[f(x)] +b-x ]2-min (4.16)

The least squares equations for the optimum calibration parameters are found from a set of

four nonlinear algebraic equations (4.18a-d). Because of our choice of a nonlinear fit function

(4.13) the least squares equations must be solved numerically through multiple iterations to

obtain the post-calibration parameters A, B, a and b.

aE.s2
-- = 0 -> i = 0 (4.17a)
OA i
D s2
-3= 0 -> E* i.Zbi = 0 (4.17b)

= 0 -+is f +b] =0 (4.17c)

0282
f=X0- e, i =0 (4.17d)
9b L -Jx _a+b.xb

A change in each post-calibration parameter indicates a change in the physical environment

for which the measurement instruments are originally calibrated. The vertical offset (A) is the

most likely to require adjustment. This parameter depends on the local tide, wave setup, vessel

squat etc. For acoustic depth finders the vertical calibration factor (B) is proportional to the speed

of sound in water which may vary with the physical and chemical properties of sea water such as

temperature, salinity, turbidity, density, etc. However, the value of B should be approximately

unity. The horizontal offset (a) may be caused by an error in distance measurements of

transponder positions but may also include a built-in error in the instrument settings. For

microwave based positioning systems the horizontal calibration factor (b) is the least sensitive

since it may only vary with a change in speed of electromagnetic waves in the atmosphere.

The post-adjustment method is applied to a set of field data, collected at Perdido Key, FL.

This area and the profile lines are shown in Figure 3.2. Figure 4.10a shows a wading profile at

Range-60 surveyed in May 1993 (dots) and the curve fit (solid line), using (4.13). The quality of

the curve fit is very important to the overall success of the analysis since the method assumes the

analytic expression calculated in (4.13) represents the true bathymetry.

For this field measurement only the first three parameters were allowed to vary. These are the

vertical offset and calibration factor (A and B) and the horizontal offset (a). As noted, the

horizontal calibration factor (b) was very unlikely to change and would require a variation in the

speed of electromagnetic waves in air. Results of the 3-Parameter post-calibration analysis to the

profile in Figure 4.10a are shown in Figure 4.10b. The Root-Mean-Square Error (s,) between

the wading and the boat profile is reduced by 80% of the original value before the

post-adjustment.

Figure 4.10: (a) Wading survey data (dotted) and the analytic expression given in (4.13) as a
curve-fit (solid). (b) Boat profile before (dashed) and after (solid) post-adjustment versus wading
profile (dotted) in overlap region. A=0.3 m, B=1.01, a=13.0 m, E~,=0.02 m.

The same analysis applied for four different surveys at 25 range lines at Perdido Key, FL, did

not reveal a consistency among different range lines for a particular survey. Instead considerable

correlation between the three post-calibration parameters were found as shown in Figure 4.11.

Vertical Offset

Range No.
Horizontal Offset
O

Vertical Factor

X X t
20 X + E O
x x x~ x o 10 X
x xx xx)IAMo + x
-------- oo -----------
-20 R +0
OX XX
o o
0
30 40 50 60 30 40 50 60
Range No. Range No.

Figure 4.11: Three-Parameter post-calibration results for Perdido Key bathymetric surveys.

From a physical point of view, there is no reason for the correlation between vertical and

horizontal parameters since they result from independent physical causes. Similarly calibration

offsets (A,a) should be independent of calibration factors (B,b). However, the correlation in the

results may be due to numerical effects artificially created from the data such as the shape of

curves or insufficient number of points in overlap region. An example for the shape effect is the

case of linear wading and boat lines that are separated from each other by a parallel shift. In this

case the offshore bathymetry could be worse than original. As a direct consequence, if the

k .. f

post-calibration analysis is applied to a set of profiles of similar shape, there will be a significant

correlation between the vertical (A) and horizontal (B) offset parameters. Correlation between

unknowns in a least squares solution are usually indicative of the non-orthogonal nature of the

parameters defining the error surface. As a result of this method, the vertical RMS-errors between

wading and corrected boat profiles in the overlap region were decreased down to the order of

several centimeters.

The RMS-errors of December 1993 survey are presented in Figure 4.12 for different methods

of adjustment. The original profiles of wading/swimming and boat surveys showed large

deviations in the overlap segment of the profile. The mean error averaged over 25 profiles was

16.6 cm for the original boat data. Significant improvements were observed in RMS-errors when

more sophisticated techniques were applied to the data. The mean error drops by 46% to 9 cm if

only a vertical shift is allowed (l-P) such that a=0, B=b=1 and only A is variable. With a

3-Parameter post-adjustment the mean error was reduced by 69% to 5.1 cm (3-P). An additional

correction method was applied as follows. The original boat profiles from November 1993 were

adjusted with the averaged values of (a) and (B) parameters over 25 range lines. Later the

adjusted data were corrected vertically at each range line. The resulting errors indicated

approximately the same RMS-error as the standard 1-P method.

The reported errors represent deviations of the boat/fathometer data from the

wading/swimming survey data. Although the latter method is generally more accurate, the

difference between these two survey accuracies may become very small or even reversed. Some

of the presented wading/swimming data were collected under difficult field conditions such as

strong longshore currents and large breaking waves. The overlap region which was used to

establish the RMS-errors is also the most difficult part of the nearshore zone to survey by

wading/swimming. On the other hand the presented boat/fathometer data are the average of two

71

or three measurements along each line, so that the effect of field conditions is filtered to some

extent.

RMS-Errors in Overlap Region
Perdido Key, Nov. 1993
0.4
I Original M 1-P 3-P

0.3-----------------------------------------------------

w
I0.2 ----- --- -- .---- -------

0.1 -- --

0 1 i I t
Range No.

Figure 4.12: Reduction in vertical RMS-errors for different methods of post-adjustment.

CHAPTER 5
SEDIMENT TRANSPORT PROCESSES AT THE NEARSHORE BERM

The physical evolution and the long-term fate of the nearshore nourishment project were

monitored by means of a series of bathymetric surveys. As shown in Figure 3.2 the nearshore

berm is located between DNR Monuments R-50 and R-60. The offshore disposal activities

commenced in September, 1990 and were completed in October, 1991. Starting with the October,

1991 survey, six of the survey lines encompassed the profile nourishment area. The measurement

efforts have included extended profile lines in the disposal area (line surveys) and additional

bathymetric surveys "blanketing" the area (box surveys). Except for the pre-disposal survey in

September, 1990 and the first post-disposal survey in October, 1991 line surveys reached the full

extent of the nearshore berm at all of the six ranges. The additional data from box surveys in

October, 1992, May, 1993 and December, 1993 were collected from a rectangular offshore area,

approximately 1 km by 5 km.

The results presented in the following sections of this chapter are obtained from the analysis

of both the line and the box surveys. Section 5.1 describes the surveying methods and important

features of data analysis. Sections 5.2 through 5.4 discuss, three aspects of the data analysis

results for the nearshore berm: (1) The volumetric changes and migration of the placed material,

(2) the lateral spreading and the evolution of surface features of the bed, and (3) the sheltering

effect of the underwater placed material on the leeward beaches.

5.1 Topographic and Hydrographic Surveys

One of the two methods used to measure the berm bathymetry was the line surveys which

were performed as a combination of wading/swimming and boat surveys. The landward portion

I

of the beach profiles were surveyed to approximate water depths of 4 to 5 m employing standard

rod-and-level techniques, by first wading and then swimming over the deeper portions of the

profiles. The offshore profile was surveyed by a boat equipped with a fathometer to measure

depths and a microwave rangefinder system to measure horizontal distances. The tide and other

long period fluctuations of the water surface were measured by a portable tide gage in

approximately 1 m of water at Range 54. Figure 5.1 summarizes the procedure which has been

used to analyze line survey data.

SRawv Boae Bad Point
S v a r Correction

Low-Pass Filter
for Wave Motion

ur Tidee Datua Tide Correction

Wading/ Vertical Adjustment
Swimmin Using Overlap Region
Data (Chapter 4)

Figure 5.1: Analysis procedure of line survey data.

During the annual surveys, twenty-five Gulf profiles were surveyed by boat, and an

additional eight have been surveyed to wading/swimming depth (generally 4-5 m) only, in order

to improve spatial resolution of the evolution of the beach nourishment. A total of twelve line

surveys have been conducted to date: The annual bathymetric surveys of 11/89, 9/90, 10/91,

10/92 and 11/93 and the additional wading/swimming surveys of 1/91, 5/91, 9/91, 1/92, 6/92,

1/93 and 5/93. Six of the profile lines, R-50, R-52, R-54, R-56, R-58 and R60, were located

within the disposal site. The nearshore berm could be detected only by boat surveys since the

wading/swimming surveys did not extend to disposal depths. The September 1990 survey was

the most recent pre-disposal survey whose results were used to establish the reference profiles

for the berm evolution. After the placement of berm material, a total of four long surveys was

conducted to cover the berm area. These are the 10/91, 10/92, 5/93 and 11/93 surveys. During the

first post-nourishment survey in 10/91 only two range lines, R-58 and R-60, extended completely

over the berm. At the other four ranges the survey lines documented only the shoreward half of

the berm sections. Figure 5.2 shows a berm section at R-54 for different surveys starting with the

pre-disposal survey in September, 1990.

Berm Cross-Section: R-54

4 '+ 90/09
..... 91/10
---92/10
S\ -. 93/05
-4.5 i .
5- .'\ -- 93/11

z
E -5-
I "\
i 0,:
ED~ ~ / .

700 750 800 850 900 950 1000 1050 1100 1150 1200
Distance Offshore [m]

Figure 5.2: Berm cross-sections obtained from line surveys at R-54.

The second method for surveying the disposal berm is the so called box survey where the

entire area is blanketed by the survey vessel resulting in a dense mesh. Figure 5.3 shows the

resulting nearshore berm area as a combination of surface and contour plots.

Perdido Key Nearshore Berm: Dec.93

1500

Shore-Parallel [m] 0 500 Shore-Perpendicular [m]

Figure 5.3: Nearshore berm generated from box survey data of December, 1993.

The horizontal vessel position during box surveys was obtained by triangulation technique

using MiniRanger units as described in Chapter 4 with an exception of two land-based units

instead of three. The third unit was not used because of interference with the other units. Without

the redundancy of a third signal, the horizontal positioning error could not be estimated as

described in Chapter 4. To avoid data inconsistency, the fixed MiniRanger units were always

placed at the same locations on DNR-Monuments R-46 and R-62. The box survey data were

collected using a local coordinate system whose origin is located at Monument R-46. The

positive longshore axis points towards Monument R-62 with an azimuth of 70.30 North. The

cross-shore axis is orthogonal positive extending seawards. Using monument coordinates and

azimuth angles, the complete set of line surveys were transformed to the local berm coordinate

system which provided a common basis to compare them with the box surveys. Figure 5.4 shows

the transformation of 91/10 line survey first into State Coordinates (Figure 5.4a) and than to local

berm coordinates (Figure 5.4b).

Perdido Key Profiles in State Coordinates: 91/10

490 6 8 (t

0 0
O 488-

\ S
0 486

484-
1092 1094 1096 1098 1100 1102 1104 1106 1108 1110 1112
Easting [1000-ft]
Perdido Key Profiles in Local Berm Coordinates: 91/10

0 1 4 6 4 f 6
0~- 6 8 4 6 08 j5OB1 52

S500-

1000

1500, -
-1000 0 1000 2000 3000 4000 5000 6000
Longshore [m]

Figure 5.4: Planview of survey lines in (a) State and (b) local coordinate systems.

The vertical position of the bed during the box surveys was measured with the same acoustic

sensor as in the line surveys accompanied by synchronous tide measurements. For each survey,

the vessel spent an average of 10 hours to complete approximately 150 km of survey lines. Since

the first box survey in October, 1992, the survey resolution increased with each successive

survey. Figure 5.5 shows the boat trajectory (solid lines) for the May, 1993 survey to provide an

example of the spatial coverage. The collected raw data from box surveys in October, 1992, May,

1993 and December, 1993 were analyzed using a similar procedure as in Figure 5.1 except that

the low-pass filtering and averaging was not applied and instead the original data in the form of

randomly distributed (x,y,z) triplets were interpolated into a 41 by 101 grid system which spans

an area of 1 km in the shore-perpendicular and 5 km in the shore-parallel directions, respectively.

During the bilinear interpolation, the unknown elevation of each grid point was calculated

from a set of original data points adjacent to that particular grid point. After selecting the set of

triplets for each grid point the final grid elevation was obtained by using a weighting function

proportional to the inverse square distance between the original survey points and the grid point.

There are several searching techniques available to identify the selection of triplets. Two of them

were applied to Perdido Key data. The first one is the "point search" technique in which to every

cell the closest N triplets are assigned where N is a pre-defined integer greater than zero. The

second one is the "area search" technique where all triplets within a certain search cell are

assigned to a corresponding grid point. Although the area of grid cells, here 25 m by 50 m, was

fixed by the size of the grid system and the number of grid points in each direction, the area of

the search cell is an independent variable. The search cell can be smaller or larger or the same

size as the grid cell.

Both methods have advantages and disadvantages. The "point search" method provides better

results in terms of data completeness since the method guaranties that to every grid point some

elevation will be assigned independent of how far the original survey points lie. On the other

hand the "area search" method is more reliable in terms of accuracy and gives more control over

the search process by changing the physical limits of data influence. Since for the particular

study, accuracy was more important than data completeness, the "area search" technique was

chosen for the analysis. To optimize the interpolation, existing box survey data were analyzed

using search cells in three sizes: 25x50 m, 50x100 m and 100x200 m. Figure 5.5 shows one result

of this analysis for May, 1993 data using 50x100 m search cells. Combined with the boat

trajectory (solid lines), the figure shows the number of data points within each cell. Darker cells

indicate larger number of original survey points collected in a particular cell. These are mostly

areas where the survey vessel slowed down for maneuvering which resulted in denser data

populations.

The findings of the bilinear interpolation tests using different sizes of search cells are

presented in Figures 5.6 through 5.8. Figure 5.6 shows the number of grid points which were

occupied with at least one original survey point divided by the total number of grid points, a

constant number equal to 4141. The shorter bars for 92/10 indicate that the survey trajectory was

not as dense as it was in the following surveys. Figure 5.7 shows the average number of original

survey points per grid cell. Since the same sampling frequency was used in all of the surveys

longer bars indicate slower boat speed. Figure 5.8 shows the vertical standard deviation of

original survey points within a cell, averaged over the number of occupied cells. This last

parameter can be considered as a measure of the vertical error which reflects a superposed value

of both the repeatability error of the survey and the round-off error due to the interpolation.

Depending on the search method, the vertical error for different box surveys varies from 7 cm to

21 cm. All three variables show consistent increase with increasing size of search cell which

indicates choosing large size of search cells would result in a more complete horizontal grid but a

larger standard deviation in the vertical; in other words, a tradeoff between the horizontal and the

vertical resolution. Although the vertical accuracy is generally more important than the spatial

resolution, Figure 5.8 indicates that the vertical accuracy is not as sensitive to changes in cell

dimensions as the other two variables (Figures 5.6 and 5.7). Therefore the optimum dimensions

for the search cell were chosen as 50 m by 100 m.

Intensity of Data Points Collected during 93/05 Survey

500050
4500 l ... ,, 70

4000

60

3000
A?

"40
c. 2500

2000 30

1E0010

500 600 700 800 900 1000 1100 1200 1300 1400 1500
Shore Perpendicular [m]

Figure 5.5: Data intensity in 93/05 survey with 50x100 m search cells.

80

Fraction of Occupied Cells
1
0.8 -- -- -

0.6 -- --- -
0.4 -

0.2 ---- --
0 .
25*50 50*100 100*200
Cell Dimensions [m]
o 92/10 E 93/05 E 93/12

Figure 5.6: Fraction of occupied cells.

Number of Points per Cell
80

60 - - - - -

40--------------------

0 ... .
25*50 50*100 100*200
Cell Dimensions [m]
E 92/10 U 93/05 E 93/12

Figure 5.7: Number of original survey points per cell.

Mean Vertical Standard Deviation

0.-- - - -- -
0.2---------B -

0.0
25*50 50*100 100*200
Cell Dimensions [m]
092/10 M93/05 93/12

Figure 5.8: Average vertical standard deviation.

81

Figure 5.9 displays a gray-scale map obtained using optimum cell dimensions. Darker areas

in the figure indicate deeper sections whereas the elevated bathymetry of the profile nourishment

are shown as lighter shading. These snapshots from 92/10, 93/05 and 93/12 show qualitatively,

the distributions of depths in the mound area. Note that each snapshot is a result of analyzed

survey data collected approximately seven months apart.

Offshore Mound: 92/10

500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Shore Parallel [m]

Offshore Mound: 93/05

500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Shore Parallel [m]

Offshore Mound: 93/12

500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Shore Parallel [m]

Figure 5.9: Evolution of the profile nourishment between October, 1992 and December, 1993.

-500

Cn

o
-1000
0-
0

-1500-
0

0 -1000

a-
E)

-1500
0

0

-1500
0

82

5.2 Volumetric Changes and On/Offshore Migration of the Berm

Six survey lines encompassing the offshore mound were analyzed to calculate volumetric

changes and the movements of the center of gravity of the placed material. Figure 5.10 presents a

general picture of the relative changes in the cross-sectional berm area and the position of center

of gravity from the completion of the placement in October, 1991 to the last survey in November,

1993. Each bar in Figure 5.10 represents the cross-sectional area of the mound for a particular

time and location. The height of the bar is a relative measure of the magnitude of the

cross-sectional area such that longer bars indicate larger amounts of material. The position of the

bars along the horizontal axis determines the cross-shore location of the center of gravity of the

corresponding section. As is evident from these results, there is no clear indication of significant

cross-shore movement of the center of gravity.

A
I
60
I
LU
w

58

6
Z 56
aD
r

54

S52

5
50

Offshore Mound: Changes in Center of Gravity and Area
I I I I

1 324

- t1: 91/10

2: 92/10

3: 93/05

4: 93/11

S i i I

920 940 960 980 1000 1020 1040
<---Onshore--- Distance from Monument [m]

1060 1080 1100
---Offshore--->

Figure 5.10: Cross-sectional area and center of gravity characteristics.

~

u-

Data presented in Figures 5.11, 5.12 and 5.13 were computed relative to the October, 1992

survey because that was the first complete set of profiles to fully cover the berm section in all of

the six range lines. Figure 5.11 shows volumetric changes of the underwater placed material

based on six profiles relative to October, 1992.

Offshore Mound: Volumetric Changes

S-20

S-40
0
0 -60

S-80

S-100
0
.S -120
E
S-140

-160

55
Range No.

Figure 5.11: Volumetric changes relative to October, 1992.

Measured volumetric changes, distributed across the entire section, are equivalent to an

average vertical change of 5 cm which is in the same order of magnitude as the vertical survey

accuracy determined in a study on nearshore measurement techniques and accuracies (Otay and

Dean, in press.). A similar result was found for the movement of centers of gravity in Figure

5.12. The horizontal displacements were found to range within approximately 10 m which is

again within the limits of the horizontal accuracy as found by Otay and Dean (in press.).

Offshore Mound: Center of Gravity
30 i ii

25- ..... 91/10

-92/10
20 93/05

-- 93/11
15

10

5-

I 0 -

ic _- I-- I I I- I I

50 51 52 53 54 55
Range No.

56 57 58 59 60

Figure 5.12: On/offshore migration of center of gravity relative to October, 1992.

In addition to the results presented in Figures 5.10 to 5.12 using line survey data, the 3-D

data from box surveys were analyzed to study the same features of berm evolution. The

nearshore bathymetry from successive box surveys was interpolated as 101 cross-sections. For

each cross-section, 2-D profiles of excess sand were computed and the areas and associated

centers of gravity of the profiles established. Resulting volumetric changes and centers of gravity

are shown in Figure 5.13 (a) and (b) respectively. Compared to the results from line surveys,

Figure 5.13 shows increased values in both variables. Specifically, erosion was observed in the

0)
E
Cl

a,
ca

0

0
t-
o
O-

western 1 km of the disposal site from October, 1992 to May, 1993 whereas the shallower

regions in the East indicated large fluctuations around zero mean. The rate of volumetric change

was considerably reduced during the period between May, 1993 and December, 1993 surveys

everywhere except in small regions near the eastern boundary between 4000 m and 4500 m. The

range of volumetric changes measured as 100 m3/m are equivalent to 10 cm in the vertical.

This value is very close to the mean vertical standard deviations (12-14 cm) presented in Figure

5.8. The change in centers of gravity between October, 1992 and May, 1993 varies from -20 m

shorewardd migration) to +70 m (seaward migration). These values were decreased to no change

in the East and -30 m in the West respectively from May to December, 1993. In general the

western 1 km of the nearshore berm experienced erosion rates of 50 to 100 m3/m between

October, 1992 and May, 1993 and stabilized afterwards. In the same region, a small but

consistent seaward migration of approximately 20 m was observed. Other sections have not

shown a significant horizontal movement.

Although it is theoretically possible to improve the measurement accuracy by adjustment of

the data or applying other high accuracy surveying techniques, it appears that the magnitudes of

the changes will remain small relative to those occurring in the beach nourishment portion of the

project because of the more energetic conditions and thus smaller time scales of the transport

processes in the vicinity of the beach nourishment and surf zone.

5.3 Lateral Spreading and Diffusion

Based upon analysis of the profile nourishment data, it appears that most of the berm

evolution is apparent as a "spreading out" of the placed material. To address the question of

lateral spreading, the landward edge of the placed material was examined and corresponding

volume changes and displacements of centers of gravity were calculated.

MILLISECOND	CLASS.METHOD	MESSAGE
0	sobekcm_page_globals.constructor
0	sobekcm_page_globals.constructor	Application State validated or built
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.constructor	Navigation Object created from URI query string
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.display_item	Retrieving item or group information
0	sobekcm_page_globals.get_entire_collection_hierarchy	Retrieving hierarchy information
0	sobekcm_assistant.get_entire_collection_hierarchy
0	cached_data_manager.retrieve_item_aggregation
0	cached_data_manager.retrieve_item_aggregation	Found item aggregation on local cache
0	item_aggregation_builder.get_item_aggregation	Found 'all' item aggregation in cache
0	system.web.ui.page.page_load (ufdc.page_load)
0	sobekcm_page_globals.constructor.on_page_load
0	html_echo_mainwriter.add_style_references	Adding style references to HTML
0	html_echo_mainwriter.add_text_to_page	Reading the text from the file and echoing back to the output stream
60	html_echo_mainwriter.add_text_to_page	Finished reading and writing the file