Citation
Sediment Management in Low-Energy Estuaries: Loxahatchee, Florida

Material Information

Title:
Sediment Management in Low-Energy Estuaries: Loxahatchee, Florida
Creator:
PATRA, RASHMI RANJAN
Copyright Date:
2008

Subjects

Subjects / Keywords:
Calibration ( jstor )
Canals ( jstor )
Estuaries ( jstor )
Sand ( jstor )
Sediment deposition ( jstor )
Sediments ( jstor )
Time series ( jstor )
Tributaries ( jstor )
Velocity ( jstor )
Water tables ( jstor )
City of Tallahassee ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Rashmi Ranjan Patra. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
12/20/2003
Resource Identifier:
54502651 ( OCLC )

Downloads

This item has the following downloads:

patra_r ( .pdf )

patra_r_Page_048.txt

patra_r_Page_005.txt

patra_r_Page_017.txt

patra_r_Page_115.txt

patra_r_Page_029.txt

patra_r_Page_010.txt

patra_r_Page_041.txt

patra_r_Page_051.txt

patra_r_Page_124.txt

patra_r_Page_019.txt

patra_r_pdf.txt

patra_r_Page_054.txt

patra_r_Page_079.txt

patra_r_Page_047.txt

patra_r_Page_024.txt

patra_r_Page_122.txt

patra_r_Page_035.txt

patra_r_Page_083.txt

patra_r_Page_121.txt

patra_r_Page_118.txt

patra_r_Page_016.txt

patra_r_Page_130.txt

patra_r_Page_052.txt

patra_r_Page_065.txt

patra_r_Page_043.txt

patra_r_Page_072.txt

patra_r_Page_073.txt

patra_r_Page_012.txt

patra_r_Page_066.txt

patra_r_Page_082.txt

patra_r_Page_132.txt

patra_r_Page_013.txt

patra_r_Page_021.txt

patra_r_Page_050.txt

patra_r_Page_018.txt

patra_r_Page_074.txt

patra_r_Page_101.txt

patra_r_Page_058.txt

patra_r_Page_110.txt

patra_r_Page_062.txt

patra_r_Page_120.txt

patra_r_Page_104.txt

patra_r_Page_060.txt

patra_r_Page_056.txt

patra_r_Page_119.txt

patra_r_Page_112.txt

patra_r_Page_057.txt

patra_r_Page_117.txt

patra_r_Page_103.txt

patra_r_Page_081.txt

patra_r_Page_127.txt

patra_r_Page_076.txt

patra_r_Page_040.txt

patra_r_Page_128.txt

patra_r_Page_061.txt

patra_r_Page_063.txt

patra_r_Page_107.txt

patra_r_Page_025.txt

patra_r_Page_089.txt

patra_r_Page_045.txt

patra_r_Page_123.txt

patra_r_Page_126.txt

patra_r_Page_131.txt

patra_r_Page_116.txt

patra_r_Page_113.txt

patra_r_Page_070.txt

patra_r_Page_023.txt

patra_r_Page_067.txt

patra_r_Page_015.txt

patra_r_Page_032.txt

patra_r_Page_006.txt

patra_r_Page_002.txt

patra_r_Page_108.txt

patra_r_Page_092.txt

patra_r_Page_097.txt

patra_r_Page_078.txt

patra_r_Page_085.txt

patra_r_Page_114.txt

patra_r_Page_011.txt

patra_r_Page_088.txt

patra_r_Page_084.txt

patra_r_Page_109.txt

patra_r_Page_125.txt

patra_r_Page_069.txt

patra_r_Page_096.txt

patra_r_Page_049.txt

patra_r_Page_129.txt

patra_r_Page_037.txt

patra_r_Page_030.txt

patra_r_Page_100.txt

patra_r_Page_111.txt

patra_r_Page_093.txt

patra_r_Page_095.txt

patra_r_Page_014.txt

patra_r_Page_075.txt

patra_r_Page_034.txt

patra_r_Page_090.txt

patra_r_Page_046.txt

patra_r_Page_044.txt

patra_r_Page_091.txt

patra_r_Page_071.txt

patra_r_Page_064.txt

patra_r_Page_042.txt

patra_r_Page_038.txt

patra_r_Page_077.txt

patra_r_Page_080.txt

patra_r_Page_003.txt

patra_r_Page_001.txt

patra_r_Page_068.txt

patra_r_Page_059.txt

patra_r_Page_033.txt

patra_r_Page_094.txt

patra_r_Page_087.txt

patra_r_Page_039.txt

patra_r_Page_022.txt

patra_r_Page_102.txt

patra_r_Page_086.txt

patra_r_Page_055.txt

patra_r_Page_008.txt

patra_r_Page_098.txt

patra_r_Page_106.txt

patra_r_Page_027.txt

patra_r_Page_007.txt

patra_r_Page_036.txt

patra_r_Page_028.txt

patra_r_Page_026.txt

patra_r_Page_053.txt

patra_r_Page_009.txt

patra_r_Page_004.txt

patra_r_Page_099.txt

patra_r_Page_105.txt

patra_r_Page_031.txt

patra_r_Page_020.txt


Full Text












SEDIMENT MANAGEMENT IN LOW ENERGY ESTUARIES: THE


LOXAHATCHEE, FLORIDA















By

RASHMI RANJAN PATRA


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2003
















ACKNOWLEDGMENTS

My sincere gratitude is reserved for Dr. Ashish J. Mehta for his guidance in my

education and research, which made my studies very precise and rewarding, as well as

the entire Coastal and Oceanographic Engineering Program faculty. Also deserving my

gratitude for their guidance and assistance are Dr. John Jaeger, Dr. William McDougal,

and Kim Hunt. Most of the analysis in the study was made possible by the valuable

assistance provided by Mr. Sidney Schofield, who taught me the basics of analysis.

Special thanks are due to Dr. Earl Hayter for setting up, supporting and guiding me

through the numerical model in its entirety.


Thanks are also due to Dr. Zal S. Tarapore, for his encouragement and guidance,

which marked my initial years as a coastal engineer and my studies here possible. My

wife Sumitra and my friend Anjana also deserve special kudos for their emotional and

editorial support. Finally, my mother and father merit unlimited praise for providing me

with mind, body and soul, as do my other friends and families for developing it.
















TABLE OF CONTENTS
page

A CKN O W LED G M EN TS ........................................... .................................................. ii

LIST O F TA BLES .................................................... ............................ .................... vi

LIST O F FIG U RES ............................................ ............................... .................... viii

LIST O F SY M BO LS ............................................ .............................. .................... xii

A BSTRA CT ....................................................................................... .. .................. xv

CHAPTER

1 IN TRO DU CTION ............................................ .............................. .........................

1.1 Problem Statem ent....................................................................... .................... 1
1.2 Study Tasks ................... ...... ... ....... ........... ....................3
1.3 Outline of Chapters .................................................................... ...................... 3

2 SEDIMENT MANAGEMENT ALTERNATIVES ................................................... 4

2.1 Present Condition of the Loxahatchee Estuary.......................... ...................... 4
2.2 C-18 Canal .................................................................................................. 12
2.2.1 Present Condition ............................................................... .................... 12
2.2.2 M anagem ent Options.......................................................... .................... 17
2.3 Central Em baym ent ................................................................. ..................... 21
2.3.2 M anagem ent Option: ...................................................... ..................... 27
2.4 N orthw est Fork: ..................................................................... 28
2.4.1 Present Condition:........................................................... ..................... 28
2.4.2 M anagem ent Options...................................................... ..................... 32
2.5 N north Fork...................................................... .......................... .................... 33
2.5.1 Present Condition............................................................ ..................... 33
2.5.2 M anagem ent Options...................................................... ..................... 34

3 D A TA CO LLECTION ....................................................................... ..................... 35

3.1 Field Setup in the Southw est Fork........................................... ..................... 35
3.2 Instrum ents D employed .............................................................. ..................... 37
3.2.1 Current ............................................... ......... ............... ...... ............... 37
3 .2 .2 T id e ...................................................................................................... ... 3 8










3.2.3 Salinity/Tem perature.................................... ......................................... 38
3.2.4 Sediment Concentration...............................................39
3.3 Field Data Results in Southwest Fork............................ ...................41
3 .3 .1 C u rren t .......................................................................... .... ... ... ..... 4 1
3.3.2 Tidal Level ............... ............................................. ..................... 44
3.3.3 Total Suspended Solids ................................................... .................... 47
3.3.5 O their D ata B locks............................................................................. 55
3.4 Field Data Results in Northwest Fork............................ ...................56
3 .4 .1 F ield S etu p .............. ....... ................................................... .................56
3.4.2 Tidal Level ............... ............................................. .................... 56
3.4.3 Total Suspended Solids ................................................... .................... 58
3.4.5 A additional D ata B locks ................................................... .................... 59
3.4.5.1 Tidal Level ...................................... .................59
3.4.5.2 Total Suspended Solids ........................................ ....... ........ 60

4 MODEL CALIBRATION AND VALIDATION ....................... ...............63

4.1 M odel D description ........................................................................ 63
4.3 Grid G generation ........................................................................69
4.4 Boundary Conditions .................................... .................................... 72
4.5 Model Calibration and Validation ............................ .........................78
4.5.1 Calibration............................ ....................78
4.5.2 M odel V alidation .......................................................... .................... ... 80
4.5.3 Simulation of trap scheme of Ganju, 2001 ..........................................86

5 EVALUATION OF SEDIMENTATION CONTROL ALTERNATIVES .............87

5 .1 D design B asis ................ .... ...................................................... ..... ............ ... 87
5.1.1 G general Principle .................................... ....................................87
5.1.1.1 Sediment Entrapment ...........................................................87
5.1.1.2 Self-cleaning Channel ........................................ .................... 88
5.1.2 Design Alternatives.......................... ..........................89
5.1.2.1 Alternative No. 2: C-18 Canal Trap...............................................90
5.1.2.2 Alternative No. 3: Bay Channel .....................................................91
5.1.2.3 Alternative No. 4: Bay Y-channel............................................ 93
5.1.2.4 Alternative No. 5: Northwest Fork Channel ...................................94
5.1.3 Efficiency A analysis .................................... ....................................94
5.1.3.1 Velocity Vector Calculation............................................94
5.1.3.2 Sediment Deposition Calculation................... ..... ...............95
5.1.3.3 Trap Efficiency.......................... ...... ..... ....................97
5.1.3.4 Channel Efficiency ................................. ........................................ 98
5.2 D design Sim ulations ................................................................. .................... 98
5.2.1 Design Flows ........................................................... 98
5.2.2 Alternative 1.........................................................98
5.2.3 A alternatives 2, 3, 4 and 5 ........................................ .................... 99
5.3 Deposition Equation Calibration.............................................102
5.3.1 C alibration for Sand ................................................... .................... 102


iv










5.3.2 F ine Sedim ent .............................. ..... .. ..... .............. ...... ................ 102
5.4 Sand Deposition due to Alternatives.............................................103
5.4.1 Bay Channel...................................................103
5.4.2 C -18 C anal ................. .......................... ... ....... ........................ 103
5.4.3 B ay Y -channel ........................ ................... .................... ... 104
5.5 Fine Sediment Deposition due to Alternatives ...........................................104
5.6 Sedim ent R em oval ................................................................ .................... 105
5.6.1 Calculation of Deposition ................................................................... 105
5.6.2 Calculation of Channel Efficiency ...................................................... 106
5.6.3 Rem oval of Bay Sedim ent ................................................................ 107
5.7 A ssessm ent of A lternatives............................................................................ 107

6 C O N C LU SIO N S ............................................ ................................................ 109

6.1 Sum m ary ......................................................... ...... ................... ... 109
6.2 C conclusions ........................ ...................... .................... .................. 110
6.3 Recomm endations for Future W ork..............................................................112

LIST O F REFEREN CES ........................................................... .......... .................... 113

BIOGRAPHICAL SKETCH .............................................. ....................116
















LIST OF TABLES


Tableage

2.1 Basin area distributions in the Loxahatchee River estuary watershed ...................7

2.2 Statistical tributary flow (based on Figures 2.6 a-c) ..............................................14

2.3 Median and high flow concentration data and coefficients for equation 2.1 .........15

2.4 Spring/neap tidal ranges and phase lags for three gauges........................................27

3.1 Instrumentation for data collection and data blocks...............................................36

3.2 Discharge data for the period 04/14/2002 to 04/21/200........................................41

3.3 Typical mean current magnitude values for data blocks........................................44

3.4 Characteristic values of the tidal data ................................... .................... 47

3.5 TSS concentrations for the representative data blocks...........................................51

3.6 Characteristic salinity values......................... .......... .............. ....... ............. 51

3.7 Characteristic temperature values ......................................................54

3.8 Summary of parametric value (Days 37-59 in year 2003).......................................55

3.9 Summary of parametric value (Days 90-101 in year 2003).................................... 55

3.10 Summary of parametric value (Days 101-135 in year 2003)...................................56

3.11 Characteristic values of the tidal data ................................... .................... 58

3.12 TSS concentrations for the representative data blocks...........................................58

3.13 Characteristic values of the tidal data ................................... .................... 60

3.14 TSS concentrations for the representative data blocks...........................................62

4.1 Definition of cell type used in the model input ..................................................69

4.2 Amplitude and phase correction factor for the tides ..................... ....................77









5.1 Alternative schemes for evaluation ..................... ...... ................................ 89

5.2 C critical velocities for sand .................................................................................. 96

5.3 Design flows in tributaries .................. ..................................... 98

5.4 Maximum currents at alternatives: calibration discharges...................................00

5.5 Maximum currents at alternatives: Different discharges .....................................100

5.6 Calibration for sedim ent fluxes .................................... .......... ............ ........... 102

5.7 Rate of sand deposition in bay channel ............................................................103

5.8 Rate of sand deposition in C-18 canal.................... ............................................. 103

5.9 Rate of sand deposition in Y-channel ....................................... 104

5.10 Rate of fine sedim ent deposition in alternatives ....................................................105

5.11 Annual sand budget: Calibration discharge .....................................105

5.12 Annual sand budget: Peak discharge.................... ............................................... 105

5.13 Annual fine sediment budget: Calibration discharge ............................................106

5.14 Annual fine sediment budget: Peak discharge ........................ ....................106

5.15 Annual sedim ent loading............................ ..... ....... .................... 106

5.16 Assessment of impacts of proposed alternatives....................................................108
















LIST OF FIGURES


Figure

2.1 Location m ap of the study area ........................................ ............................. 5

2.2 Loxahatchee River estuary and tributaries .......................... ...... ................. 5

2.3 Hydrographic survey of the estuary (November 2001)....................... ........... 7

2.4 Loxahatchee River estuary watershed basins, estuarine limits (dashed arcs) and
central em baym ent ................. .... .......................................................................... 8

2.5 Ariel photograph showing development of new shoal...........................................11

2.6 Cumulative discharge plot. a) Northwest Fork. b) North Fork.
c) Southw est Fork ................... .............. ... ................ .......... ........ ...14

2.7 Dredging Plans for C-18 canal, 1956 ..................................... ..... ............... 16

2.8 Current variation under the effect of released discharge from the S-46
stru ctu re ...................................... ............................ ............. ...... 19

2.9 Effect of S-46 discharge on the suspended sediment concentration........................19

2.10 Arial Photograph showing the Central Embayment, the Inlet and the
T rib u tries ..................................................... ..................... 2 3

2.11 Location of tide gauges marked UFG1, UFG2 and UFG3.....................................26

2.12 Sample records of tidal measurements at three locations (09/14/00-09/15/00)-
Datum NAVD 88. ....................................................... 27

2.13 Location of stream-gauging stations and sampling site for suspended
sedim ents, .................................................................................. .................... 30

2.14 Location indicating fresh mud depositions and the Shoals the estuary ..................33

3.1 Location of instrument tower in the Southwest and Northwest Forks ...................35

3.2 Calibration plots used for calibration of OBS sensors ..........................................40

3.3 Record of current magnitude: Days 94-114 (year 2002) ...................................... 42









3.5 Record of current magnitude: Days 332-356 (year 2002)........................................43

3.6 Record of current direction: Days 332- 356 (year 2002)...................................... 43

3.7 Water level time-series: All levels relative to NAVD 88. Days 94-114 (2002). .....44

3.8 Water level time series: Upper plot shows original time series with mean trend
and the lower plot is without the mean oscillations. All levels relative to
NAVD 88. Days 94-114 (2002). .......................... ....................................45

3.9 Water level time-series: All levels relative to NAVD88. Days 332- 365 (2002)
and D ays 01-35 (2003). ................................................ ................................. 45

3.10 Water level time series: Upper plot shows original time series with mean trend
and the lower plot is without this trend. All level relative to NAVD 88. Days
332-365 (2002) and Days 01 -35 (2003). ........................................ ... ........46

3.11 TSS time-series at four elevations: Days 94-114 (year 2002). ................................48

3.12 TSS time-series at three elevations: Days 352- 365 (year 2002) and 01-35
(y ear 2003). ............................................................................... .................... 48

3.13 Depth-mean TSS concentration time series: Days 94-114 (year 2002)...................49

3.14 Depth mean TSS concentration time series: Days 352- 365 (year 2002) and
Days 01 -35 (year 2003). ................. ...................................... 49

3.15 Depth mean TSS concentration time series and tidal trend indicating their
dependence: Days 352- 365 (year 2002) and Days 01 35 (year 2003). ...............50

3.16 Salinity time series: Days 94-114 (year 2002). ............................. ......................52

3.17 Salinity and Current magnitude time series: Days 94-114 (year 2002). .................52

3.18 Temperature time series: Days 94-114 (year 2002)................................................53

3.19 Salinity time series: Days 352- 365 (year 2002) and 01-35 (year 2003). ..............54

3.20 Temperature time series: 352- 365 (year 2002) and 01-35 (year 2003)...................54

3.21 Record of water level variation. Days 245 -255 (year 2003). ..............................57

3.22 Water level time series: Upper plot shows original time series with mean trend
and the lower plot is without the mean oscillations. All levels relative to
NAVD 88. Days 245 -255 (year 2003). ........................................................57

3.23 Depth-mean TSS concentration time-series: Days 245-255 (year 2003) ...............58

3.24 Record of water level variation. Days 310.5 -313.5 (year 2003). ..........................59









3.25 Water level time series: Upper plot shows original time series with mean trend
and the lower plot is without the mean oscillations. All levels relative to
NAVD 88. Days 310.5 313.5 (Year 2003).................................. ..................... 60

3.26 TSS time-series at two elevations: Days 310.5 313.5 (year 2003). ....................61

3.27 Depth mean TSS concentration time series: Days 310.5 -313.5 (year 2003).........61

3.28 TSS time-series at three elevations: Days 315.5 318.5 (year 2003)...................62

3.29 Depth-mean TSS concentration time series: Days 315.5 -318.5 (year 2003). .......62

4.1 Model domain showing input bathymetry and shoreline.........................................71

4.2 Computational grid showing the flow boundaries ..............................................72

4.3 Tidal time series from UFG1, 09/14/00-10/13/00, a) Raw data, b) Tidal plot
after the m id-tide trend is rem oved. ................................................ .................... 74

4.4 Tidal time series from UFG3, 09/14/00-10/13/00, a) Raw data, b) Tidal plot
after the m id-tide trend is rem oved. ................................................ .................... 76

4.5 Flow tim e series applied at S-46 boundary .................................... ..................... 77

4.6 Flow time series applied at Northwest Fork boundary..........................................77

4.7 Model calibration measured vs. predicted current, a) Cold Start, b) Hot Start........81

4.8 Model calibration measured vs. predicted current direction.................................82

4.9 Model calibration measured vs. predicted water surface elevation (UFG2) Year
2000, a) C old Start, b) H ot start. .................................................... ..................... 83

4.10 Model calibration measured vs. predicted water surface elevation (UFG3) Year
2000, a) C old Start b) H ot start. .................................................... ..................... 84

4.11 Model calibration measured vs. predicted water surface elevation
(Northwest Fork) Year 2003, a) Cold Start, b) Hot start..........................................85

4.12 Validation results using trap used by Ganju, 2001..............................................86

5.1 Design concepts for sedim ent manager ent .................................. ..................... 88

5.2 Alternatives considered, with existing bathymetry..................... .....................90

5.3 Dredged bathymetry of the C-18 canal with plan form view of the proposed
trap. Trap considered by Ganju (2001) is also shown............................................91

5.4 Planform view of the proposed self-cleaning channel in the bay. .........................92









5.5 Location of the sea grasses indicated in model with increased roughness...............92

5.6 Planform view of the proposed self-cleaning Y-channel in bay...........................93

5.7 Planform view of the proposed self-cleaning channel in the Northwest Fork ........94

5.8 Current comparisons for a model cell at the upstream end of the Northwest
Fork channel: calibration discharges .............................................. ..................... 99

5.9 Current velocity vectors over the modeled domain; maximum flood velocities
at spring tides. .............................. ....................... .................... 101

5.10 Current velocity vectors over the modeled domain; maximum ebb velocities at
sp rin g tid e .......................................................................................................... ..... 10 1















LIST OF SYMBOLS

area

vertical turbulent viscosity

width of the basin

wave celerity

uniformly distributed initial sediment concentration (kg/mi)

sediment concentration (kg/m 3)

constant multiplier for u-velocity conversion to true east

constant multiplier for u-velocity conversion to true north

constant multiplier for v-velocity conversion to true east

constant multiplier for v-velocity conversion to true north

sediment deposition under reduced flow

deposition at the entrance to the channel

deposition at the exit of the channel

dimensionless projected vegetation area

total water column depth

length of the channel/trap

coefficient of conductance

volume source or sink

Richardson number











steady mean flow velocity

velocity vector

width of the channel in equation 5.8

settling velocity

buoyancy

vegetation resistance

Darcy-Weishbach friction factor

coriolis acceleration

acceleration due to gravity

water depth

scale factor along x-axis

scale factor along y-axis

turbulent intensity

amount of sediment in influent

amount of sediment in effluent

removal ratio

velocity along the channel (x-axis)

friction velocity

velocity amplitude under current

current at the entrance


current at the exit








critical velocity for erosion


curvilinear-orthogonal horizontal velocity

velocity in true east direction

velocity across the length of the channel(y-axis)

curvilinear-orthogonal horizontal velocity

velocity in true north direction

variable water depth

water density

reference water density

bed erosion shear stress

x-component shear stress

y-component shear stress

bed bottom shear stress

mid-tide elevation

high-tide elevation

low tide elevation

vertical diffusivity

Karman constant

free surface potential

velocity angles
















Abstract of Thesis to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science


SEDIMENT MANAGEMENT IN LOW ENERGY ESTUARIES: THE
LOXAHATCHEE, FLORIDA

By

Rashmi Ranjan Patra

December 2003


Chairman: Ashish J. Mehta
Major Department: Civil and Coastal Engineering

Implementation of schemes for sediment entrapment and self-cleaning channels

was examined in the micro-tidal estuarine environment containing both sand and fine

sediment. The central embayment of the micro-tidal Loxahatchee River estuary on the

Atlantic Coast of Florida was chosen as the candidate location due to its unique

characteristics with respect to the influx of sand and fine sediment in its central

embayment, and concerns regarding the potential for long term impacts of this flux on the

embayment. An ideal sediment trap captures all of incoming sediment, i.e., the removal

efficiency is 100%. A self-cleaning channel allows no net deposition of incoming

sediment, which passes through, so that its removal efficiency is nil.

Hydrodynamic model simulations were carried out for selected trap/channel

alternatives, and their efficiency was calculated by relating sediment deposition to change

in the flow regime due to implementation of these alternatives. Calculations indicated









that the concepts of sediment entrapment and of self-cleaning can operate only

imperfectly in the study area due to the low prevailing forcing by tide and the episodic

nature of freshwater discharges in the tributaries.

Fine sediment accumulation in the central embayment can be reduced by dredging

the C-18 canal, as the trapped sediment would account for more than half of the total fine

sediment entering the bay. A channel close to the southern bank of the embayment could

improve bay flushing by ebb flow, reduce bay-wide sedimentation and serve as a

navigation route. Careful design with regard to channel alignment would be required to

avoid sea grass beds in the area. Long term simulations of flow and sediment transport

are required to assess sediment circulation patterns and the formation of shoals in the

central embayment and the Northwest Fork.















CHAPTER 1
INTRODUCTION

1.1 Problem Statement

Sedimentation due to the influx of fine and coarse particles is an issue affecting

numerous estuaries and coastal waterways. Often enough, these particles originate far

inland, and are transported into the coastal zone by runoff and stream flow. In the

estuarine regime, inorganic sediment almost never occurs in isolation, as it is

complemented by measurable organic fraction produced by either indigenous sources

(e.g., native phytoplankton, swamp vegetation, wind blown material), or allochthonous

sources (e.g., river-bome phytoplanktons, swamp vegetation, windblown material)

(Damell, 1967). In turn, such organic-rich sediments can degrade water quality by

oxygen uptake and a reduction in light penetration. In this study, the question of

preemptive dredging of sediment prior to its deposition in an area of concern or, as an

alternative, preventing its deposition in the area of concern by channelizing flow, was

studied. The candidate water body was the estuarine segment of the Loxahatchee River

on the east coast of Florida.

Loxahatchee River, which discharges mainly through its Northwest Fork, supplies

mainly quartz sand and organic detritus. Clay mineral makes up less than 5% of the mud

in the estuary, but because this mud is rich in organic matter, its accumulation has

become a matter of concern in the central embayment of the estuary. This flow, in

addition to controlled discharges from the S-46 structure in the C-18 Canal at the head of









the Southwest Fork, brings in much of the sediment (mean concentration 0.014 kg/m3;

Sonnetag and Mcpherson, 1984) in the central embayment.

A commonly employed solution to reduce sedimentation is the implementation of a

trap scheme by trenching the submerged bottom. Such a trench-trap is a means to

increase the depth at the chosen location by dredging. Increased depth results in a

decreased flow velocity (and associated bed shear stress), thereby allowing incoming

sediment to settle in the trap, instead of being carried further downstream. The removal of

sediments becomes much easier as it can be then be removed from the trap, rather than

dredging the otherwise distributed deposits from a considerably broader area. As an

alternative to sediment entrapment, creating a self-cleaning channel in the area of concern

for sedimentation would mean that sediment would pass through the system, without

deposition. The degree to which both approaches can function depends on the flow

conditions, type of sediment and the morphology of the estuary.

Given the above background, the objectives of this study were: 1) to determine the

efficiency of traps installed at selected locations in the estuary, and 2) to evaluate the

efficiency of channels as a means to pursue the goal of a self-cleaning sedimentary

environment.

Shoaling has occurred the Loxahatchee in many areas, especially near the

confluences of the major tributaries (Northwest Fork and Southwest Fork) in the central

embayment where the velocities are typically low (Sonntag and McPherson, 1984).

Recent studies (Jaeger et al., 2002) suggest internal recirculation of sediments as an

important factor governing sediment transport within the estuarine portion of the river.

Accordingly, in order to manage sedimentation in the central embayment, it may be









desirable to test trap/channel deployments at multiple locations. The performance of these

schemes was evaluated with regard to efficiency of sediment removal.

1.2 Study Tasks

The tasks undertaken included:

1. Data collection from the site and scrutiny of data from the existing literature to
characterize the nature of flow, sediment transport and sedimentation. This
included measuring tidal elevations, current velocities, sediment concentrations and
bed sediment distribution (Jaeger et al., 2002) in the estuary, and obtaining stream
flow data for the tributaries from the literature.

2. Simulating the flow field using a hydrodynamic model, in order to determine the
velocities, water surface elevations and bed shear stress distributions.

3. Introduction of trap schemes in the calibrated flow model to determine flow
velocities with and without the trap, and development of relationships for
calculating trap efficiency.

4. Introduction of self-cleaning channels and an assessment of their viability.

5. A qualitative assessment of the usefulness of the approaches based on selected
criteria.

1.3 Outline of Chapters

Chapter 2 describes the sediment management alternatives including existing

conditions and the proposals for implementation. Chapter 3 deals with the field data

collection for this study including data analysis and interpretation. Flow model

calibration and validation is included in Chapter 4 and evaluation of management

alternatives is described in Chapter 5. Summary of the results and conclusions are made


in Chapter 6, followed by a bibliography of studies cited.















CHAPTER 2
SEDIMENT MANAGEMENT ALTERNATIVES

2.1 Present Condition of the Loxahatchee Estuary

Loxahatchee River empties to the Atlantic Ocean through the Jupiter Inlet located

in northern Palm Beach County on the south coast of Florida, about 28 km south of St.

Lucie Inlet and 20 km north of Lake Worth Inlet. The three main tributaries, which feed

the estuary, are the Northwest Fork, the North Fork, and the Southwest Fork. In addition,

the Jones Creek and Sims Creek, which are far lesser tributaries than the others, also feed

the estuary through the Southwest Fork. Figures 2.1 shows the general location map of

the study area.

The major surface flow in to the estuary historically was through the Northwest

Fork draining the Loxahatchee Marsh and Hungry land slough (refer Fig 2.1). The

upstream reach of the Southwest Fork, referred to as the C-18 canal, was created in

1957/58 in the natural drainage path in order to lengthen the area of influence of the

Southwest Fork and facilitate drainage of the westward swampland (Refer Figure 2.1 and

Figure 2.2). The flow in the canal is regulated by the S-46 automated sluice gate

structure. Whereas, the Southwest and the Northwest fork converge on the estuary

approximately 4 km west of the inlet, the North fork joins the central bay about 3 km

west of the inlet. Down stream of the Florida East coast Railroad (FECRR) Bridge the

Intracoastal Waterway (ICWW) intersects the estuary in a dogleg fashion. Five

navigation/access channels exist on the south shore of the central embayment















8021 D i 1 r wr a

L N I jE G L\







JiV. itl MILESS QLflL







Figure 2 1 Location map of the study area (Source US Geological Survey report no 84-
4157, 1984)





















A detailed hydrographic survey of the central embayment (Figure 2 3) and the
Northwest and Southwest Forks carried out November' 2001 (Ldberg Land


Surveying, Inc) indicates the depths in the estuary, which range between 0 m (reference


to North Atlantic Vertical Datum 1988, (NAVD 88)) near the sandy shoals to almost 6 m
to North Atlantic Vertical Datum 1988, (NAVD 88)) near the sandy shoals to almost 6 m









in the entrance channel near the FECRR Bridge. The average depth over the embayment

is just over 1 m. The navigation channel (maintained by the Jupiter Inlet District) runs

westward from the Inlet, under the FECRR bridge, and through the central embayment

approximately 14 km upstream from the Inlet. The navigation Channel has a bottom

width of about 30.5m (100 feet) and is maintained at 1.75m(5.74feet) (reference to

National Geodetic Vertical Datum 1929, (NGVD 29) and 2.21m (7.24feet) with

reference to NAVD 88) with a side slope of 1:3. Flood shoals, which approximately

bisects the central embayment exists mainly due to the sand influx from the ocean, and

smaller shoals exist at the termini of the three main tributaries. Small shoal islands are

located west of the FECRR bridge, on both sides of the channel.

The Northwest Fork and North Fork are natural tributaries draining in to the central

embayment. However, as mentioned the Southwest Fork was lengthened westward by

construction of C- 18 canal with a control structure (S-46), in order to divert flow from the

Northwest Fork to the Southwest Fork. A channel was then constructed allowing the

diversion of flow from the Northwest Fork to the Southwest Fork. For easy reference

from this point on, the C-18 canal will be indicated as the narrow channel section and the

broader section at the root will be called Southwest Fork (Figure 2.2).






























Figure 2 3 Hydrographic survey of the estuary (November 2001)

The Loxahatchee River estuary draus over 1000 km2 of land through the three

main tributaries, the ICWW, and several minor tributaries The individual watershed
basins are shown in Figure 2 4 and listed in Table 2 1 The watershed constitutes

residential areas, agricultural lands, and unimhabited maush and slough areas

Table 2 1 Basin area distributions in the Loxahatchee River estuary watershed
Basin Area (km )
Intracoastal Water way 545

Jonathan Dickinson 155
South Indian River 65
Loxahatchee Rive 6
main~~~~ ~ ~ trbtais th JCW dsvrlmio rbtre h idvda aese
basnsareshwn n igue 4 i lite ini- Tal 1Tewaesedcntiue

reienil ra, giclualluis ui nihbte mashaidsouhara
















Loxahatchee Rive_____ _____6____









-



- '.1... -.
I.


-I .i ..x''or -


i-_ -, '







i tLo {.







Unlike more northerly estuaries, upland drainage in to the Loxahatchee provides

only quartz sand and organic detritus Clay mineral makes up less than 5% of the mud in

the estuary (McPherson, 1984) Earlier studies indicate that the estuary was periodically

open and closed to the sea due to various reasons Originally, flow from the Loxahatchee

River along with that from Lake Worth Creek and Jupiter Sound kept the inlet clean

With the construction of the ICWW and the Lake Worth inlet and the modifications of

the St Lucie Inlet in 1970, some flow was diverted Subsequently, Jupiter Inlet generally

remained closed until 1947, except when it is dredged periodically After 1947, it was









maintained open by dredging by the Jupiter Inlet District and the U.S. Army Corps of

Engineers.

Dredge and fill operations have also been carried out in the estuary embayment and

forks. In the early 1900's, there was significant amount of filling at the present FECRR

Bridge, which narrowed the estuary from 370 m to 310 m. The areas east and west of the

bridge (and also under the bridge) were dredged in mid-1930's, and also in 1942. The

material was high is shell content and was used in construction of roads. In 1976-77,

additional estimated 23,000 m3 materials were removed from the estuary at the bridge

and from an area extending 180 m from the west. Some dredging was also carried out in

the Southwest Fork near the C-18 canal in the early 1970's (Wanless, Rossinsky and

McPherson, 1984). In 1980, three channels were dug in the embayment, and an estimated

23,000 m3 of sediment were removed.

After 1900, the estuary was greatly influenced by the dredging and alteration of the

drainage to the basin. With gradual lowering of the water table and resultant effect on the

water quantity, the direction and pattern of inflow (McPherson and Sabanskas, 1980)

were considerably affected. Historically, the major surface flow to the estuary was in to

the Northwest Fork from the Loxahatchee Marsh and the Hungry-land Slough (Figure

2.1), both of which drained north. A small agricultural canal was dug before 1928 to

divert a small amount of water from the Loxahatchee Marsh to the Southwest Fork. As

noted, in 1957-58, C-18 canal was constructed along the natural drainage way to divert

flow from the Northwest Fork to the Southwest Fork of the estuary.

Jaeger et al., (2001) carried out extensive studies in the estuary to reevaluate the

nature of environmental sedimentology in the lower Loxahatchee River Estuary and as a









companion study to Ganju et al. (2001). Specifically, new samples were collected in

order to 1) examine changes in surficial sediment types between 1990 and 2000, 2)

attempt to determine the sources of fine-grained muddy sediments accumulating within

the estuary; and 3) examine rates of sedimentation within the central embayment and

three forks (North, Northwest, and Southwest/C-18 canal) by collecting a suite of -l-m

long pushcores and -3 m long vibracores within the estuary. Grab samples were collected

in all regions of the estuary and were analyzed. One of the main findings of the study was

the internal movement of the sediments in the estuary system. With the growth of the

population on the shoreline and associated human activities the mangroves dotting the

shoreline started vanishing. The removal of these Mangrove cover from the shoreline

released a large quantity of sediments, which was otherwise trapped in their roots.

Essentially fine grained, these sediments moved with the flow and started getting

deposited in the estuarine bounds. According, to this study new shoals were

developed/grown by this process, especially the submerged one in the Northwest Fork,

down stream of the shoal identifiable from a satellite map and Figure 4 of the Report

(Jaeger et al., 2001). The aerial photograph reproduced in Figure 2.5 also indicates an

additional shoal developing from the root of the existing shoal, suggesting that the

general nature sediments being fed by the Northwest Fork is coarse grained with the fine

grained ones carried downstream with the current before deposition.

Tidal flow into and out of the estuary is much larger than freshwater inflow from

all the major tributaries. Fresh water flow is reported to be about 2 percent of the total

tidal inflow (Sonnetag and McPherson, 1984). Tides are mixed semidiurnal (twice daily

with varying amplitude) with a tidal range of about 0.6 to 0.9 m. Tidal waves advances









up the estuary at a rate of 2.23 m/s to 4.46 m/s (McPherson and Sonnetag, 1984) and

shows little change in the tidal amplitudes over to about 16 km river km. Winds have a

significant effect on the tidal ranges especially the strong northeast winds which prevails

during autumn and winter for example can push in additional water into the estuary

affecting the tidal ranges.


Figure 2.5 Ariel photograph showing development of new shoal

Estuarine conditions extends in the estuary from the inlet for about 8 river km into

Southwest Fork, 9.6 river km in to the North fork and 16 river km into the Northwest

Fork.

Of late, the environmental condition of the Loxahatchee River and the estuary has

become a matter of great concern. The major factor affecting the environmental health is

the sediment transported in to the estuary. Large amount of the sediments settling in the









basin might affect the bottom life, alter circulation patterns, and accumulate shoals,

thereby impeding boat traffic (McPherson, Wanless and Rossinsky, 1984).

2.2 C-18 Canal

2.2.1 Present Condition

The C-18 canal drains the Loxahatchee Slough, a shallow swamp-like feature

containing diverse flora and fauna. However, estuarine conditions persist for 8 km up the

Southwest Fork/C-18 canal measured from the inlet.

Flow data obtained from USGS stream flow gage data, for all available years

(1971-2002 N.W. Fork, 1980-1982 N. Fork, and 1959-2002 S.W. Fork) indicate that C-

18 canal/Southwest Fork carries a maximum discharge of 61.54 m3/sec. Cumulative

frequency distribution curves were constructed to designate (Figure 2.6 a-c) median and

extreme flow events (Table 2.2) for all the tributaries. The C-18 canal is regulated at S-46

structure, which is basically a gated sluice. The criterion for controlling the flow at the S-

46 structure is based on water level behind the structure. When the level exceeds a

predetermined mark, the sluice gates are opened until the level recedes by 30 cm (Russell

and McPherson, 1984), at which point the gates are closed. This regulation has resulted in

a discontinuous flow record; with weeks of no flow passing the structure, and days when

storm flows have been released. During normal wet season, the level behind the S-46

structure is not always sufficiently high for releasing flow, while the other tributaries are

freely discharging to the estuary.














































--------- -- -- -- -- -- ---
|_ _ | 4 I I_ 4 _ |


0 5 10 15 20 25 30 35 40 45 50


Flow rate (m3/s)


_______


55 60 65 70 75 80


0 02 04 06 08 1 12 14 16 18


Flow rate (m3/s)


02j


I 2 I I I 1 2
I- I I I I -

ftI +I I I I+- H











I 1 2 I I I 1 1 2
_ - - -- - 4- 1 1 -
I I_ _ _


_ _ _ _ _


II_ _ _ _






II_ _ L J _
I I

I I I II


2 22


-------- -- -- -- -- _:_ ::: ::


I I H1 H1 -



I 1 -


I_ I 2_ 2 -



I_ I H1 1 -






I_ I I2 I2 -



I- I I I1 A -






I_ I 2_ 2_ -











c

09
08 - --- ------ -- ----- -----+ -- -

07 I I I I

06 -- ------ --- -- ----------
0 5 _ _


0 3 II ----

02 -

01 -- --- --- -------- -- -


0 5 10 15 20 25 30 35 40 45 50 55 60 65
Flow rate (m3/s)
Figure 2.6.Cumulative discharge plot. a) Northwest Fork. b) North Fork. c) Southwest
Fork.

Table 2.2 Statistical tributary flow (based on Figures 2.6 a-c)
Tributary Median Flow High Flow Maximum Flow
(50%) (90%) (98%)
(m3/s) (m3/s) (m3/s)
Northwest Fork 0.7 4.1 76
North Fork 0.1 0.21 1.9
Southwest Fork 1.3 7.8 61

Sonnetag and McPherson (1984) reported two values of suspended solid sediment

concentration (0.059 kg/m3, 0.017 kg/m3) with corresponding flow data for the C-18

canal (31 m3/s, 28 m3/s, respectively) and a mean concentration value for duration (1980-

82) of their study (0.014 kg/m3). The median flow for the C-18 canal (1.3 m3/s) from the

Figure 2.6c was correlated to this mean value of concentration in the present study. A fit

in the form of (Mfiller and F6stner, 1968)


C = aQ









was used (Ganju et al., 2001), where a, and bk are site specific coefficients, with a is

indicative of the erodibility of the upstream banks/bed and exponent b, is indicative of the

intensity of the erosional forces in the river.

Table 2.3 Median and high flow concentration data and coefficients for equation 2.1
Median flow High flow a
Tributary Concentration Concentration
(kg/m3) (kg/m3) Coefficient Coefficient

Northwest Fork 0.011 0.023 0.012 0.27

North Fork 0.01 0.018 0.018 0.02

Southwest Fork 0.014 0.059 0.012 0.49

Fieldwork, consisting of bottom profiling and sampling, was carried out during July

2001 (Jaeger et al.,) by collecting a suite of -l-m long push cores and -3 m long

vibracores within the estuary. A total of 110 samples were collected from sampling

locations covering the entire estuary and river (Figure 1, Final Report on Sedimentary

processes in the Loxahatchee River Estuary, 5000 Years ago to the Present, Jaeger et al.,

2001) including from outcrops of regional surficial geological unit (undifferentiated 1.8

million year-old Pleistocene sediments) in order to examine the potential sediment

sources (Loxahatchee River, C-18 canal, Inlet, and Pleistocene-Age (last 1.8 million

years) sediments exposed along the banks of the C-18 canal. 56 of the samples were

reoccupations of sites sampled in 1990 and were reported (Mehta et al.,1992). These new

samples were collected to examine the changes in sediment characteristics pattern over a

10-year period. Positions of all sampling sites were determined by differential GPS

providing a position accuracy of ~1 m. Each grab sample recovered approximately

1,000-2,000 cm3 of sediment, removing approximately the uppermost 1-5 cm of the

sediment surface. Sediment distribution maps produced from these grab samples indicate









particle sizes reveal that the majority of the estuary is dominated (by weight) of fine,

well-sorted sand in the -150 micron (3 phi, 0 15 mm) size range (Jaeger et al, 2001)

In the same study conducted by Jaeger et al, (2001), poling depths (obtained by

pushing a graduated pole into bottom until a hard substrate is reached) in the C-18 canal

were determined to estimate sedimentation rates along the length of the canal Since the

bottom was dredged at the time of construction of the canal in 1957/58, the bed thickness

can be considered to represent the subsequent accumulation This is because, the

dredging of the canal in 1958 would have most likely left behind a hard, sand rich

horizon that could not be easily penetrated with the solid rod Figure 2 7 shows these

thicknesses along the canal length Sediment thickness increases with the distance from

the S-46 structure, possibly due to the large erosional forces near the structure (when

flow is released), and reduction of these forces as the flow moves along the canal,

allowing more deposition of sediment



E Natural bottom (pre-1957)


I-


.... I- D'D edged canal bottom
S(post-1957)


0 500 1000 1500 2000
Distance from S-46 structure (nm)
Figure 2 7 Dredging Plans for C-18 canal, 1956 (Source: Ganju et al., 2001)

This coarse sand layer was sampled at the base of push cores (see Figures 21 and

22 from Jaeger et al, 2001) There appears to be a trend of increasing thickness away









from the S-46 control structure (Figure 19). Modeling of sediment transport in the canal

(Ganju et al., 2001) also supports such a trend. The overall sedimentation rates (10-50

mm/yr) in the canal are very high for most coastal areas, where sedimentation has kept

pace with the rise in sea level (3-5 mm/yr) (Davis, 1994). However, this sampling

technique of poling only provides mean sedimentation rates over this 42-year (1958-

2001) time period. Analyses of push cores collected in the canal document alternating

layers of clean sand and muddy sand/sandy mud (see Figures 21-22, Jaeger et al., 2001).

This inter-layering of sediment types is characteristic of time-varying deposition

rates/erosion rates. When the sluice gates are opened, fast currents can erode the

sediment surface followed by rapid deposition of sand and mud. The best way to evaluate

time-varying sedimentation rates is with either accurate annual bathymetric profiles or by

measuring naturally occurring radioisotopes in the sediment cores (Jaeger et al., 2001).

2.2.2 Management Options

Dredging plans for the C-18 canal from 1956 is shown in Figure 2.7 (U.S. Army

Corps of Engineers, 1956). The existing bottom was deepened to 3 m at some locations to

facilitate drainage. The depths refer to the National Geodetic Vertical datum of 1929

(NGVD). The present mean depth of the canal as measured along the length is 1.2 m.

Hence there has been substantial sedimentation in the canal, which in turn means that it

no longer serves as a sediment trap and allows sediment to be transported to the central

embayment. One way of maintaining the depth in the canal is to devise a suitable

dredging option coupled with a designed flow regime in order to maintain the canal in the

self-cleaning mode. However, one of the main difficulties in this is the lack of continuous

supply of water. As described earlier, the flow in the canal is erratic and controlled by the

S-46 control structure. Accordingly, although the median flow in this canal is higher than









the other tributaries, the flow is episodic and therefore not enough to overcome the bed

shear resistance of the deposited sediments. This situation can be illustrated by data

collected during between April 4th and April 24t, 2002.

Figure 2.8 indicates the dependence of the current velocity on the released

discharge. The sudden jump in the over all current magnitude recorded downstream of

the structure therefore exhibit strong erosional trend as can be seen from the Figure 2.9.

In addition, it indicates that, sediment concentrations in the bottom layers are much more

pronounced due to the obvious reason of erosion of the bed. It can therefore be concluded

that, a sustained and regular flow regime would help keeping the canal sediment free.

An option is to increase the depth in the canal by dredging part or all of it, thereby

recreating the sediment trap. As an alternative, a detailed study of the flow pattern can be

undertaken and a suitable flow regime worked out. This would involve redesigning of the

control structure and a better regulation of the flow. However, the following points

should be noted:

1. The capacity of flow from the structure appears to be insufficient to flush out
sediment beyond 1.2 km (Ganju et al., 2001) from the structure even under "high"
discharges when the gates are open.

2. A potential option is to change the gate configuration but not the flow regulation
schedule. If changing the gate configuration from sluice to weir is successful, it
would create a sediment trap upstream of the gate, which would "buy time" for
the downstream reach of the canal, but this upstream trap would eventually have
to be dredged to maintain it effectiveness. The volume of material trapped will be
restricted the weir height. Over-depth dredging upstream is a viable option.

3. However, because sediment transported across the gate is believed to be quite
heterogeneous (ranging from fine sand to clay and organic matter) and the organic
material is presently not found in the bed there, predictive modeling the transport
of sediment across the gate will be an uncertain exercise without extensive data
collection on both sides of the gate. An option would be to carry out gate
conversion and work with the new system based on a rough estimation of the new
flow/sediment regime. It is likely that some modification of gate opening schedule
may also have to be carried out to improve the efficiency of the upstream trap.







19






1,4














0" 2
I

O 95 100 106 110 115
Days of the Year (2002)

Figure 2 8 Current variation under the effect of released discharge from the S-46
structure


OBS 4




40




=3 0OS 2 los 110 r15
90 95 100 OBS 3 105 110 115









90 95 100 105 110 115
90 95 100 OBS 1 105 110 415


300~~~~~~~~~~ ------ -- ---- ] ------ --- ------i-------------


Days of the Year (2002)

Figure 2 9 Effect of S-46 discharge on the suspended sediment concentration

The present study envisages examining the option dredging the downstream canal


Ganju (2001) warned out such an exercise by testing the effectiveness of a comparatively









short sediment trap. The trap design and results of the investigation are summarized

below.In order to quantify the sedimentation rate as a function of discharge in the C-18

canal investigations were carried out using calibrated sediment transport models. The

boundary conditions were designed to simulate the episodic unsynchronized (with

Northwest and North Fork discharges) discharges from the S-46 structure. The results

indicated that as discharge increases the change in the rate of sedimentation rate

decreases. However, they do not share a direct straight-line relationship. For instance,

doubling of flow from 2.5 m3/s to 5 m3/s results an increase of 71% in the sedimentation

rate and similar increase from 10 to 20 m3/s changes the rate only by 25% indicating that,

the sedimentation rate is more sensitive to lower discharges. This is evidently due to

increasing discharge is associated with increased concentration. The regulation of the C-

18 canal by the S-46 structure is manifested in the high frequency of zero-discharge

periods (54% of the days) and the spikes. The deposition rates were found to be 0.15 m

for a period of 10 years, which compared well with the poling results.

The study also compares the sedimentation in a regulated C-18 canal to that of

hypothetically unregulated canal by applying flow record for the Northwest fork for the

same period pro-rated so that the discharges over the 10 year flow period remains

identical. Resulting in a 10-year deposition thickness of 0.22m (0.022m/yr), implying that

the episodic discharges in actuality reduced the rate of sedimentation. This is a direct

consequence of near constant high discharge attenuating the increasing trend of

sedimentation.

The study incorporates a trap near the area of greatest post dredging thickness, with

a poling depth of approximately 1.2 m. A dredging depth of 3 m (from the original bed









level) width of 60 m, and a length of 180 m were chosen for the trap, which was

considered sufficient to reduce the velocity in the canal, and allow a measurable amount

of sediments to settle. This trap configuration reduced the current magnitude by 67% over

the trap. As a consequence a number of factors were evaluated by the study namely,

* Simulations showed that the removal ratio, i.e., the ratio of sediment influx (into
the trap) minus out flux divided by influx), was maximum at an S-46 discharge of
approximately 1.7 m3/s. At higher discharges sediment was transported beyond the
trap, while at lower discharges sediment settled before the trap.

* The second simulation involved testing the trap efficiency as a function of sediment
concentration. It was observed that increase in sediment concentrations in the free
settling range in general increases the settlement. The increase in trapped load
followed a linear trend up to concentrations of 0.25 kg/m3 (free settling zone),
which is explained by the increase of deposition flux with concentration (with
constant settling velocity). Above this concentration, and below 7 kg/m3
flocculationn range), the increase in settling velocity yields a similarly increasing
trend for trapped load. In the hindered settling zone, however, (which lies above
this concentration) trapped load decreases as the settling velocity deceases. It was
therefore be inferred that trapped load is a function of concentration because
settling velocity (and hence the deposition flux) is also a function of concentration
at values greater than 0.25 kg/m3.

* The simulations on varying organic content indicated that, increase in organic
content led to decease in settling velocity, which resulted in lower removal ratio.
Sedimentation rate in the trap increased with increased organic content, due to
corresponding decease in dry density. In addition, the increase in influent load with
increasing organic content as less sediment was deposited upstream of the trap at
higher organic content.

2.3 Central Embayment

2.3.1 Present Condition

Jupiter Inlet, which is about 112 m wide and 3.9 m deep at the jetties, allows the

tidal flow in and out of the estuary. The channel starting at the jetties leading up to the

Florida East Coast Railroad Bridge is fairly uniform, with width varying from 206 m to

247 m and the mean depth varying between 3.92 m at the inlet and 2.6 m near the

FECRR Bridge. The ICWW meets the channel down stream of the FECRR bridge.









Upstream of the FECRR Bridge the embayment widens and the channel is divided in to

two parts by shoals often exposed under low water conditions. These shoals, presumably

created by the sands introduced in to the system through the inlet and the tributaries, and

carried by the flood tide, occur where the sediment carrying capacity of the flow reduces

with the reduction of current at wider sections. In addition, east of these sandy shoals

there occurs a small mangrove island. Similar Islands occur near the north bank close to

the FECCR Bridge. The deepest portion of the embayment lies to the north of the sandy

shoal, easily identifiable even from a real photograph (Figure 2.9) is currently used for

navigation. The shoreline is basically sandy with little or no clay present. The percentage

of clay and silt is barely 5%. The average depth in the central embayment is 1.2 m. The

depth in the deeper portions along the flood channel however exceeds 3 m in patches. A

similar deep channel can be found along the south bank, which has been presumably

created by the ebb circulation. A clear ebb channel can also be seen from the satellite

photographs to the south of the sandy shoal. Boats returning to their docks use this

channel at high water. There are many private wooden docks along the entire coastline.

At the turn of the century, the Loxahatchee River estuaries along with its

immediate environ was a pristine ecosystem consisting of mangroves, salt marshes, and

scrubland. Prior to Word War II, agricultural interests transformed the area in to a rural

landscape with citrus groves and vegetable farms. As a result, a significant increase in

residential population occurred around this time. These developments ultimately

prompted the declaration of the estuary an aquatic preserve in 1984. Nonetheless, the

construction activities, especially of the residential homes, still continue along the

shoreline and the entire estuarial shoreline of the central embayment as well as a









significant portion of the tributary shorelines is residentially occupied. Recreational

boating is widely practiced in the estuary by the local residents. Access is necessary to

the upstream areas for recreational activities, and also to the open sea and the ICWW.

Many of the natural and artificial access routes have shoaled in recent years (Antonini et

al., 1998), leading to hazardous boating practices such as high-speed entry/ exist to

prevent grounding of vessels. The channels adjacent to the south shore of the central

embayment are more susceptible to shoaling (Sonntag and McPherson, 1984), directly

affecting the boaters who rely on these channels for access.














Figure 2.10 Arial Photograph showing the Central Embayment, the Inlet and the
Tributaries

Estimates with regard to grain size, composition and age of bottom sediments are

given by McPherson et al. (1984) for the entire estuary. The samples collected by vibro-

core boring were analyzed in the laboratory for micro-faunal and macro-faunal

assemblages, grain size distributions, constituent composition and radiocarbon age. With

regard to the grain size it was seen that, the characteristics of the bed material were

identical to those of the underlying sediments in the core. Fine-grained sediments

dominate the central bar at the lower reaches of the estuary; whereas medium to coarse-

grained sand dominates upper reaches of the bar. Patches of fine to medium sand draping









the muddy sediment surface can be seen in the main body of the estuary. The shell

content in the bed material varies from 0 to 5% at the eastern end to 20 to 30% at the

western end.

Grain-size analysis reveals that there are two distinct different populations. The

first, well-sorted sediment with a mode between 62.5 to 125 microns, and the second,

poorly sorted sediment commonly showing bimodality. The bimodal distributions

generally have one mode at about 300 microns and the other at 100 microns.

Jaeger et al, (2001) measured the particle sizes in the estuary, which, reveal that the

majority of the estuary is dominated (by weight) of fine, well-sorted sand in the -150

micron (3 phi; 0.15 mm) size range. This size sand is ubiquitous in the estuary and is

observed in Pleistocene-age coastal deposits exposed in outcrops within the study area.

The ultimate source of the sand accumulating within the upper estuary is from erosion of

these older deposits. The amount of mud-sized sediment (<63 microns) is minimal with

the exception of the upper reach of the Northwest Fork, the North Fork, and near the

junction of the C-18 canal and the Southwest Fork. Clay mineral analyses on the mud

fraction accumulating throughout the estuary reveals that the ultimate source of the clay-

sized sediment is from erosion of the Pleistocene-age deposits.

Comparison of the sediment characteristics (median particle-size, sorting) between

1990 and 2000 within the Central Embayment reveal that this region has not changed

significantly over the past decade. However, the navigation channels have become

coarser apparently due to the removal of fine sediment. Portions of the lower Northwest

Fork and the Southwest Fork have gotten finer.









Based on the analyses of 20 push cores, there does not appear to be a widespread

organic-rich flocculent "muck" layer within the three major forks of the estuary.

Although mud is a common component of the sediments in these locations, by weight it

usually represents less than 20% of the total core mass.

In addition, study by Jaeger et al., (2001) indicate that in the main navigation

channels, the sediments have become coarser and more poorly sorted over the last ten

years. The study attributes this to the likely inclusion of shelly material in the 2000

samples that was not sampled in 1990. It is possible that maintenance dredging during

this time period resulted in the exposure of older shelly material or that changes in the

shape of the navigation channel has led to stronger currents that have removed the finer

sands. Although the western portion of the Central Embayment has seen no change in the

median particle diameter, it has gotten marginally better sorted, and could reflect a

decrease in fine sediments accumulating.

Freshwater runoff enters the Loxahatchee River estuary by river and canal

discharges, by storm drains, and by overland subsurface inflow. Most of the freshwater

from the tributaries is discharged from the Northwest Fork of the estuary. These flows, as

expected, vary seasonally, occurring chiefly in the wet season. The median, high and

maximum flow discharges are given in Table 2.2.

Tidal flow into and out of the estuary is much larger than the freshwater inflow

from all the major tributaries. The combined freshwater flow into the estuary is found to

be about 2% or less of the average tidal inflow at the Jupiter inlet (McPherson, Sonnetag,

1984). However, during tropical storm Dennis, freshwater inflow per tidal cycle

increased to 18% of average tidal inflow (McPherson, Sonnetag, 1983). Tides are mixed









senm-diurnal with varying amplitudes, with a tidal range of approximately 0 6 to 1 m The

tidal wave advances to the estuary at a rate of about 2 3 m/s to 4 5 m/s Higher than usual

tides can be noted dunng the autumn and winter when strong northeast winds pushes

additional water m to the estuary causing higher than average tides

Ultrasonic water level gauges (Model 220, Infinties USA, Daytona Beach, FL)

with stillng walls were installed to measure tidal elevations between September 14t and

October 18ft, at three locations m the estuary one each m the Central embayment (tied to

the FECRR bridge pier), Northwest Fork, and Southwest Fork The gauge locations are

shown m Figure 2 10 Tidal elevations were recorded with respect to North Atlantic

Vertical Datum 1988 (NAVD 88) and are reproduced m Table 2 3 Tidal ranges indicate

the total change m water surface elevation between low and high tides and phase lag

refers to the difference in time between high/low tide at UFG1 gauge and the other

gauges In Figure 2 12 sample records from three tidal gauge locations are shown


Figure 2 11 Location of tide gauges marked UFG1, UFG2 and UFG3










045



03

02


E 0 0
00 0






04

05
Time (hour)
Figure 2.12 Sample records of tidal measurements at three locations
Datum NAVD 88.

Table 2.4 Spring/neap tidal ranges and phase lags for three gauges


0


-- --UFG2
- al- *UFG3


(09/14/00-09/15/00)-


Gauge ID Spring range Neap range Phase lag from UFG1
(m) (m) (min)
UFG1 0.90 0.66 0 0
UFG2 0.85 0.65 21 60
UFG3 0.86 0.64 28 60


Ganju et al., (2001) compared the data obtained from these gauges to a station on

the Northeast Florida coast and inferred that trends in water surface elevation followed

similar increases and decreases in mid-tide elevations and the increased elevations in side

the estuary is a direct result of onshore winds. The wind records from two offshore

stations were averaged and correlated with the mid-tide elevation, resulting in a positive

correlation. Accordingly, mid-tide elevation was subtracted from the measured elevations

(filtering) in order to obtain tidal data without any variation.

2.3.2 Management Option:

Presently, the dredged spoil from the embayment is disposed on land. Land

disposal of marine sediment is often times not optimal for the environment, especially for









the ground water. According to earlier studies by Sonnetag and McPherson (1984) the

central embayment receives sediment from two main sources, the inlet and upland

discharge. Regular maintenance of the navigation channel is a clear indication of this

supply. Ideally, a large enough central shoal (if developed to correct contours) could

serve the process of self-cleansing of the bay. The shoal when developed would decrease

the water flow area and thereby, increasing the velocity of flow. The increased current in

the limiting case would develop erosional stresses equal to the critical bed shear of the

sediment and therefore would be able to prevent the further sedimentation of the bay.

However, numerical modeling for such an examination is outside the scope of this study.

The present study will however deal with the development of an additional

navigation/flow channels for improvement of ebb flow.

2.4 Northwest Fork:

2.4.1 Present Condition:

The Northwest Fork meanders through typical South Florida swampland within the

Jonathan Dickinson State Park (JDSP). The extensive swampland and scrubland east of

JDSP is drained by the North Fork. It is therefore evident that the watershed is

biologically productive, and the sediment carried by the runoff is rich in organic content

eventually finds its way in to the estuary (Sonnetag and McPherson, 1984).

Most of the freshwater from is discharged through this fork. From February 1",

1980, to the September 30th, 1981, for example, 77.3 percent of the freshwater was

discharged into the Northwest Fork, 20.5 percent in to the Southwest Fork (C-18 Canal),

and 2.2 percent into the North Fork (Sonnetag and McPherson, 1984). The Loxahatchee

River (i.e., Northwest Fork) at SR-706, site 23 as shown in Figure 2.14 (Figure 2, U.S.









Geological report no 83-4244, 1984), contributed the greatest percentage of flow to the

estuary (37.4 percent) of all the tributaries.

Vertical variation of the sediments in the Northwest fork is Found at site 5 and 5E

(Figure 2, U.S. Geological report no 84-4157, 1984) during both incoming and outgoing

tides. Presumably, greater water velocities, particularly at 0.6 m above the bottom at the

mid depth, associated with higher tide stages contributed to the greater vertical variation

of suspended sediments (Sonnetag and McPherson, 1984). Concentration of the

suspended sediments and the percentage of sediments of organic origin were variable

with season and weather conditions as indicated by the data collected and listed in U.S

Geological Survey report 84-4157 (Sonnetag and McPherson, 1984). The greatest

increases were observed in Cypress Creek, lying upstream of the Northwest Fork.

Concentration of the suspended sediment in the tributaries also changed as a result of

man's upstream activities. During September 1981, suspended sediment concentration in

the Cypress Creek and Hobe Grove Ditch increased as much as 21 times over

concentrations in early September (Sonnetag and McPherson, 1984). Cleaning and

dredging operations on the irrigation canal connected to the Cypress Creek and Hobe

Grove Ditch were presumably responsible.

Suspended sediment load from the tributaries are highly seasonal and storm related.

The 5 major tributaries to the Loxahatchee estuary Loxahatchee River at SR-706,

Cypress Creek, Kitching Creek, Hobe Grove Ditch, and C-18 at S-46 discharged 1,904

tons of suspended sediments to the estuary during the 20-month period (February 1, 1980

to September 30, 1981) (Table 2.3). During the 61 days period of the above-average

rainfall (August 1 to September 30, 1981) that included tropical storm Dennis, the major









tributares discharged 926 tons of suspended sediment to the estuary This accounted for

49 percent of the suspended sediment discharged to the estuary dunng the 20-month

penod and about 74 percent of the suspended sediment discharged dunng 1981 water

year (Sonnetag and McPherson, 1984) Sediment loads from C-18, Loxahatchee River at

SR-706, and Cypress Creek accounted for more than 94 percent of the total tributary

input of the sediment load


F.lure --L. t o ,1 4ltrear.'l dg1Yg Rfali cr a~, for *pd L....hatch.e

Figure 2 13 Location of stream-gaugmg stations and sampling site for suspended
sediments, (Source US Geological report no 83-4244 and 84-4157)

Unlike the central embayment concentration of mud was quite high (-50%) in the

Northwest Fork (Jaeger et al, 2001) The study by Jaeger et al, (2001) also analyses

vibracores takes which, reveal that there has been roughly 0 5-1 cm/yr of sedimentation

within a part of the Northwest Fork when compared to data from a USGS-sponsored









study completed, in 1984 (Sonnetag and McPherson). The study further concludes that,

these accumulation rates are close to those averaged over the past 50 years, assuming that

an observed change in the cores from layered sediment not mixed by organisms to those

that are well mixed by organisms occurred in 1947 when the inlet was stabilized. Inlet

stabilization would have led to increased tidal flushing that allowed for better

oxygenation of bottom waters and sediments permitting occupation of sediments by

organisms. However, this datum has not been substantiated as pre 1947 and the

accumulation rates are bulk averages. A comparison of the collected data and studies by

Ganju et al., (2001) showed that accumulation rates within the upper reaches of the three

Forks are about 2-3 times higher than the modeled fine-sediment budget prepared by

Ganju et al. (2001). Accordingly, the study concludes that, this discrepancy could be due

to poor age constraints of the core layers or to the substantial presence of sand in the core

sections, which was measured in this stratigraphic (i.e., core layering) approach but not in

the fine-sediment budget.

Upstream of the outfall point of the Northwest Fork is marked by a horseshoe-

shaped shoal (Figure 2.14). Presumably this shoal is formed due to the reduction in

current velocity of the sediment-laden flow by the ebb tide. In addition, the ebb flow

velocity gets reduced upon meeting a large body of water (central embayment). Upstream

of this shoal there occur a series of sand shoals also formed by the same processes.

Downstream of the shoal however, the depths are uniform gradually increasing as moves

in to the central embayment area. Formation of deposits presumably from the erosion of

old deposits in side the estuary was also reported by Jaeger et al., (2001). Figure no 4 of









the report are reproduced here for reference with regard to the deposition and material

composition.

In Figure 2.15 the mass percent of the mud (particles smaller than 63 microns) in

the upper -5 cm of the sediment surface is shown. Location of the sampling sites are

shown as dot symbols.

2.4.2 Management Options

Discounting the sedimentation from the internal sources of erosion, the Northwest

Fork contributes the maximum discharge as well as the maximum sediment into the

central embayment (Sonnetag and McPherson, 1984). However, Jaeger et al, (2001)

indicate that fresh deposits are found in the Fork (Figure 2.14), suggesting that the source

of such deposits may be mostly internal to the estuary, and most likely due to the erosion

of old deposits. Sediment from external sources entering the estuary with fresh water

discharge as reported by McPherson (1984) would have deposited in the proximity of the

horseshoe shoal.

In order to minimize the deposition of fine sediment in the area of high mud

percentage in the Northwest Fork (Figure 2.14), a self-cleaning channel will be

examined. According to Jaeger et al., (2001), the origin of deposits (Figure 2.14) is due to

the erosion of old deposits. Therefore the channel is proposed to be located downstream

of these deposits. Design aspects of the channel are considered in Chapter 4.






33




75










Clay deposits 3
70
65



50













Figure 2 14 Location ndcating fresh mud depositions and the Shoals the estuary Source
Sedimentaiy Processes in the Loxahatchee River Estuary 5000 Years Ago to
the Present-FINAL REPORT, Jaeger et al (2001)

2.5 North Fork

2.5.1 Present Condition

The North Fork is a natural tnbutary draining the eastern part of the Jonathan

Dickson State Park Discharge as given i Table 2 2 is the least of the three main

tnbutanes (2 2% of total), and water depth is fairly uniform at around m, with virtually

no shoals McPherson and Sonnetag (1983) reported that i the 1981 water year the

tnbutanes of the North Fork were dry at the gauging stations (Figure 2 13) from March
3

trough mid-August Dunng the rest of the year the average flow was 0 12 m /s, a very

small value Discharge following Topical Storm Dennis was also small for the amount of

rainfall associated with the storm Daily discharges for the last 10 days of August 1981

averaged 0 31 m /s but increased to 0 71m/s in September Jaeger et al (2001) found






34


some mud deposits in the upper reaches. Depths in the fork appear to be adequate for the

recreational boating.

2.5.2 Management Options

The North Fork as indicated above has the least river inflow as well as the least

sediment contribution to the estuary. In addition, the depths are fairly uniform and good

for the types of boats presently using it. Hence no additional facility is believed to be

required for this area. Therefore no dredging is planned for this tributary nor appears to

be required.















CHAPTER 3
DATA COLLECTION

3.1 Field Setup in the Southwest Fork

Field data were collected at two sites, one in the Southwest Fork and the other m

the Northwest Fork Section 3 3 collection effort and results in the Southwest Fork, and

Section 3 4 m the Northwest Fork

The field data collection set up in the Southwest Fork of the estuary had

geographical coordinates of latitude 260 56' 36 78" N and longitude 800 07' 17 34" W In

the Northwest Fork the corresponding coordinates were 260 59' 16 78" N and longitude

80 07' 56 34" W These two locations are shown in Figure 3 1 The locations of the tidal

gages installed in the year 2000 were shown in Figure 2 11 The depth (below North

Atlantic Vertical Datum, 1988, (NAVD88)) at the sites ware 2 1m and 2 18, respectively


Figure 3 1 Location of instrument tower in the Southwest and Northwest Forks









Data in the Southwest Fork were collected in two phases. The first phase of the

data collection was carried out between 4h and 24th April 2002, and the second phase was

between 6h of February and 2nd of June 2003. The instrumentation deployed is given in

Table 3.1.

Table 3.1 Instrumentation for data collection and data blocks
Instrument Data Date
Data logger (*) Current (mag.) -u Apr 04 to Apr 24 Nov 27 to Jun 2
Data logger (*) Current (dir.) -u Apr 04 to Apr 24 Nov 27 to Jun 2
Data logger (*) Current (mag.) v No data Nov 27 to Jun 2
Data logger (*) Current (dir.) v No data Nov 27 to Jun 2
Data logger (*) Tide levels Apr 04 to Apr 24 Nov 27 to Jun 2
Data logger (*) OBS 1 Apr 04 to Apr 24 Nov 27 to Jun 2
(Poor quality)
Data logger (*)) OBS 2 Apr 04 to Apr 24 Nov 27 to Jun 2
Data logger (*) OBS 3 Apr 04 to Apr 24 Nov 27 to Jun 2
Data logger (*) OBS 4 Apr 04 to Apr 24 No data collected
Data logger (*) Temperature Apr 04 to Apr 24 Nov 27 to Jun 2
Data logger (*) Salinity Apr 04 to Apr 24 Nov 27 to Jun 2
*With ultrasonic current meter in April 2002 replaced with an electromagnetic current meter

Instruments were attached to a tower erected for this purpose and was powered by

rechargeable batteries. The instrument assembly consisted of a Marsh-McBirney

electromagnetic current meter, a Transmetrics pressure transducer for the measurement of

water surface elevation, a Vitel VEC-200 conductivity/temperature sensor for

measurement of salinity and temperature, and three Sea point Optical Backscatter Sensor

(OBS) turbidity meters for measuring the sediment concentration at 3 different levels. In

the first phase instrument setup, however, turbidity sensors were deployed at 4 different

levels. In addition Lidberg Land Surveys, Inc. carried out a hydrographic survey and

collected data with regard to the bottom bathymetry of the central embayment and the

tributaries.









In the Northwest Fork the data collection started on 14t of August 2003, with 3

level OBS sensors, one Conductivity Temperature sensor and one Pressure gauge

3.2 Instruments Deployed

3.2.1 Current

Current data were collected using Marsh-McBirey electromagnetic current meter

(Model 585 OEM). This meter consists of a 10 cm diameter spherical sensor, OEM

motherboard, and signal processing electronics (Figure 3.5). The instrument senses water

flow in a plane normal to the longitudinal axis of the electromagnetic sensor. Flow

information is output as analog voltage corresponding to the water velocity components

along the y-axis and x-axis of the electromagnetic sensor. The velocity sensor works on

the Faraday principle of electromagnetic induction. The conductor (water) moving in the

magnetic field (generated from within the flow probe) produces a voltage that is

proportional to the velocity of water. The Marsh-McBirey requires periodic cleaning of

the probe with mild soap and water to keep the electrodes free of non-conductive

material.

Since the instrument has essentially a cosine response in the horizontal plane, the

flow magnitude and the direction information are retained. In addition, the spherical

electromagnetic sensor has an excellent vertical cosine response. This unique

characteristic allows the sensor to successfully reject vertical current components that

may be caused by mooring line motions. As the flow changes direction, the polarities of

the output signal also change. So the u (velocity along axis of the channel flow) and the v

(velocity across the channel) velocities are stored and can be combined to give the

resultant magnitude and direction. It must however be noted that, the v velocity

component was largely insignificant due to the width of the channel at the tower location.









3.2.2 Tide

Water surface elevation was measured using a Transmetrics pressure transducer

installed at the instrument tower. The instrument incorporates three major design

elements that allow it to measure pressure accurately and reliably; bonded foil strain

gages configured in a Wheatstone bridge (for temperature stability), high precision

integral electronics for signal amplification, and stainless steel construction for durability

and corrosion resistance. The instrument was calibrated and temperature-compensated

against standards applicable for the region.

3.2.3 Salinity/Temperature

Conductivity is the measurement of the ability of a solution to carry an electric

current. It is defined as the inverse of the resistance (ohms) per unit square, and is

measured in the units of Siemens/meter or micro-Siemens/centimeter. The measurement

of conductivity is necessary for the determination of the salinity of a solution. Salinity is

proportional to the conductivity and is expressed in terms of concentration of salt per unit

volume (mg/1, or ppt). The field measurement of salinity was carried out following

similar procedures using a Greenspan Electrical Conductivity (EC) sensor substantially

eliminating a basic source of error arising out of the inaccuracies due to temperature and

electrode effects. In this instrument the electrical conductivity is a function of the number

of ions present and their mobility. The electrical conductivity of a liquid changes at a rate

of approximately 2% per degree Centigrade for neutral salt and is due to the ionic

mobility being temperature dependent. The temperature coefficient of the conductance

(or K factor) varies for salts and can be in the range 0.5 to 3.0. As electrical conductivity

is a function of both salt concentration and temperature, it is preferable to normalize the









conductivity measurement to a specific reference temperature (250C) so as to separate

conductivity changes due to salt concentration from those due to temperature changes.

* The instrument deployed consisted of the following primary elements:
* Toroidal sensing head (conductivity sensor)
* Temperature sensor
* Microprocessor controlled signal conditioning and output device

The conductivity sensor uses an electromagnetic field for measuring conductivity.

The plastic head contains two ferrite cores configured as transformers within an

encapsulated open-ended tube. One ferrite core is excited with a sinusoidal voltage and

the corresponding secondary core senses an energized voltage when a conductive path is

coupled with primary voltage. An increase in charged ion mobility or concentration

causes a decrease in the resistivity and a corresponding increase in the output of the

sensor.

A separate PT100 temperature sensor independently monitors the temperature of

the sample solution. This sensor provides both a temperature output and a signal to

normalize the conductivity output.

3.2.4 Sediment Concentration

The instrument deployed was a Sea Point turbidity meter. This instrument measures

turbidity by scattered light from suspended particles in water. The turbidity meter senses

scattered light from a small volume within 5 centimeters of the sensor window. The light

sources are side-by-side 880 nm Light Emitting Diodes (LED). Light from the LED

shines through the clear epoxy emitter window into the sensing volume, where it gets

scattered by particles. Scattered light between angles 15 and 150 degrees can pass

through the detector window and reach the detector. The amount of scattered light that







40


reaches the detector is proportional to the turbidity or particle concentration in the water

over a very large range.

The sensors were calibrated using a sample from the measurement site. Periodic

calibrations were conducted in order to evaluate the conditions of the windows and the

sensitivity to scattering. In addition, only black containers were used in calibration so as

to prevent any probable scattering events due to reflection off the container wall. The

calibration was carried out using known volume of sediments in known volume of water

and the voltage output of the instrument recorded. A linear fit curve was generated in

order to determine the accuracy of the calibration. The calibration plots are given below,


20 30 40
Concentration (mg/L)

25

2


0
"15
a
5
O
1 /


10 20 30 40
Concentration (mg/L)


10 20 30 40 50
Concentration (mg/L)


Figure 3.2 Calibration plots used for calibration of OBS sensors









3.3 Field Data Results in Southwest Fork

3.3.1 Current

The electromagnetic current meter was located at a height of 96.5 cm from the

bed level. The velocity data in two directions, one parallel to the flow and the other

perpendicular to it, were combined vectorially to find the resultant magnitude and

direction. The ultrasonic current meter deployed in April 2002 collected the current

magnitude and direction directly. Based on these data the depth-mean magnitude time

series for Julian days 94-114 is shown in Figure 3.3 and the corresponding direction plot

is given as Figure 3.4. A sudden increase in the current magnitude in the plot is

attributable to the opening of control structure S-46. The directional plot indicates a uni-

directional flow driven by the discharge from the structure. The discharge record for the

period is given in Table 3.2 for ready reference.

Table 3.2 Discharge data for the period 04/14/2002 to 04/21/200
Date Julian Days of 2002 Discharge
(m3/s)
04.14.2002 104 0.03171
04.15.2002 105 0.00821
04.16.2002 106 0.01416
04.17.2002 107 0.01501
04.18.2002 108 0.03483
04.19.2002 109 0.05777
04.20.2002 110 0.03568
04.21.2002 111 0.00934

























S-46 Gate Opned


JULIAN DAYS IN YEAR 2002

Figure 3 3 Record of current magnitude Days 94-114 (year 2002)


200 r

150





^ 0 FLOOD
z




-15(





SO 95 100 105 110 115
JULIAN DAYS IN YEAR 2002

Figure 3 4 Record of current direction Days 94-114 (year 2002)















Maximum 0 17 ms
Mean 0 06 m/s


335 340 345 350
JULIAN DAYS IN YEAR 2002

Figure 3 5 Record of current magnitude Days 332-356 (year 2002)


335 340 345 350 355
JULIAN DAYS IN YEAR 2002

Figure 3 6 Record of current direction Days 332- 356 (year 2002)

Figure 3 5 is a representative plot of the current magnitude for the second data

block This plot indicates a more uniform velocity pattern dnven by the tidal flow in the







44


estuary The current magnitudes reach a maximum value of 0 17 m/s with the mean

value at 0 06 m/s In addition it is seen that the flow is predominantly along the estuary

with very low values observed for transverse current (v) In Table 3 3 typical mean

current values are summarized

Table 3 3 Typical mean current magnitude values for data blocks
Current magmtude (m/s)
Julian days in Curent ma ltude (m/s) Velocity u Velocity v
2002 With S-46 Only tidal flow (is) (i/s)
2002 (mls) (mls)
discharge
94-114 025 004 a- a
332-356 j_ 0 06 0057 0018
a No data

3.3.2 Tidal Level



Maximum Level 1 7 m
Mean Level 1,2 m
18 Minimum Level 07m











s / I i I I l



9 95 100 105 110 115
JULIAN DAYS IN YEAR 2002

Figure 3 7 Water level time-senes All levels relative to NAVD 88 Days 94-114 (2002)






























0 95 100 105 110 115
JULIAN DAYS IN YEAR 2002

Figure 3 8 Water level time senes Upper plot shows onginal time series with mean trend
and the lower plot is without the mean oscillations All levels relative to
NAVD 88 Days 94-114 (2002)


,1h1,11 I nl ,lIlll ll1' I
II



el 292m
1 49 m
221 rm


340 350 360 370 38
DAYS OF THE YEAR (2002)


Figure 3 9 Water level time-senes All levels relative to NAVD88 Days 332- 365 (2002)
and Days 01-35 (2003)


l i I ,


I I | 1 | I I I .'

Spnng tde Range 08 m
Neap tide Range 0 5 r



i'1 S i ( I 1 1 1 1'i'1 1 1


i i*' n | l n n' i 'i*>'i















15-








W 0




Spring range 1 00 m

Neap range 0 50 m

-1 5


-2 IIIII
340 350 360 370 380 390 400
JULIAN DAYS IN YEAR 2002/03

Figure 3.10 Water level time series. Upper plot shows original time series with mean
trend and the lower plot is without this trend. All level relative to NAVD 88.
Days 332-365 (2002) and Days 01 35 (2003).

In Figure 3.7 the raw tidal time-series is shown for the period April 4 to April 24th,

2002. In Figure 3.8 the upper plot shows the original time series with the tidal trend and

the lower plot is with the tidal trend removed.

The tidal plots indicated in the Figure 3.7 to Figure 3.10 are representative plots

from the phase II and I. The characteristic values of the tidal data are given in Table 3.4.

In addition it can be noted that the tidal ranges compares well in both the phases with the

spring range equal to 1.0m and the neap range around 0.5m. As will be explained later,

the tidal fluctuations (as could be noted from Figure 3.9) between Julian days 360 to 365

in Year 2002, 01 to 5 and 17 to 25 in Year 2003, is likely to affect the sediment

concentration in the estuary.









Table 3.4 Characteristic values of the tidal data
J n ds Mean Water level/Tidal range
Julian days
in 2002/03 water depth Water level (m) Spring/neap range (m)
in 2002/03
(m) Maximum Minimum Spring Neap
94-114 1.20 1.70 0.30 0.90 0.50
332-365
332-365 1.20 1.90 0.30 1.00 0.50
01 -35


3.3.3 Total Suspended Solids

Total Suspended Solid (TSS) was recorded at four elevations in the first phase and

three elevations in the second phase. The elevation of the OBSs relative to the bed level

was OBS-4 = 1.17 m, OBS-3 = 0.80 m, OBS-2 = 0.48 m and OBS-1 = 0.22 m. The

corresponding total suspended solid time series are reported in Figures 3.11 and 3.12 for

days 94-114 (Phase I) and 352- 365 in 2002 and 01 to 35 in year 2003 (Phase II),

respectively.

Table 3.5 provides the maximum, mean and minimum values of sediment

concentrations at different levels for each data block. Depth-mean concentration averaged

every 12 hours is presented in Figures 3.13 and 3.14. The mean concentration figures

(Figure 3.13 and 3.14) indicate the average variations in the concentration over time with

out the instantaneous variations (spikes).

















3 95 100 OBS2 105 110 11s


100 105 110
JULIAN DAYS IN YEAR 2002


Figure 3 11 TSS time-senes at four elevations Days 94-114 (year 2002)


OBS 3


Maximum O mIgL




350 355 360 365 370 O0 1 380 385 390 395


Maximum 2500 mgL
Mean 1600 g/L
Minimum 270 mrnL


355 360 385 370 375 380 385
JULIAN DAYS IN YEAR 2002/03


Figure 3 12 TSS time-senes at three elevations Days 352- 365 (year 2002) and 01-35
(year 2003)


90 395












OBS1





0 95 100 OBS 2 105 110 1





S95 100 OBS 3 105 110





S95 100 OBS 4 105 110 1





S95 100 105 110 1
JULIAN DAYS IN YEAR 2002


Figure 3 13 Depth-meanTSS concentration time senes Days 94-114 (year 2002)


OBS 3


100 Maximum 150 mgL
Mean 17 mg/L
E M'irMmm 7 mgL


350 355 360 365
3000

000- Maximum 2900 mg
SMean 1900 mg
E 1000 Minimun 400 ma


370 O 1 380 385 390 395 400


JULIAN DAYS IN YEAR 2002/03


Figure 3 14 Depth mean TSS concentration time senes Days 352- 365 (year 2002) and
Days 01 35 (year 2003)







50




80


70 Tide Trend x 25


S60

-J
W 50
F-
4 OBS 3 Trend
40


E30
-- Driven by 46
Z Discharge a
L 20 December 2 th
O

S10 Drven by tidal
fluctuations

0
355 360 365 370 375 380 385 390 395 400
JULIAN DAYS IN YEAR 2002/03

Figure 3.15 Depth mean TSS concentration time series and tidal trend indicating their
dependence: Days 352- 365 (year 2002) and Days 01 -35 (year 2003).

It can be noted from Figures 3.11 and 3.13 that there is a sudden increase in

sediment concentration with the discharge from the S-46 structure on 14h of April 2002

(Refer Table 2.2 for discharge details). This clearly indicates that sediment concentration

is discharge driven. Results of Figures 3.12 and 3.14 indicate that the lowest OBS1

sensor was too close to the bed and recorded almost saturated sediment content. There

was no discharge from the structure between December 14th and February 20th, except for

0.01 m3/s discharge on the December 20h, 2002, which explains the increase in sediment

concentration recorded around Julian day 355 (December 201h). However, the increase in

TSS reported between days 17 and 27 (Year 2003) without any discharge from S-46,

could be attributed to spring tidal effects (Refer to Figure 3.15). In general, it appears that









TSS concentration is dependent on the local tidal current and flow discharges down the

S-46 structure. The TSS concentrations with regard to other data blocks are given in

Table 3.7 to 3.9.

Table 3.5 TSS concentrations for the representative data blocks
Julian Days Maximum TSS Mean TSS Minimum TSS
in 2002/03 concentration concentration concentration
(mg/L) (mg/L) (mg/L)
94-114 165 50 10
352-365 158 17 7
01-35

3.3.4 Salinity and Temperature

The conductivity and temperature measurements carried out for the location is

presented in Figures 3.16 (days 94 114 of 2002). The salinity curve indicates the effect

of the fresh water discharge. Due to this flow fresh water from the S-46 structure the

salinity values dropped to 11 mg/L from a mean value of about 28 mg/L. In order to

examine this hypothecation the current magnitude and the salinity was plotted together in

Figure 3.17, which, indicated a decrease in salinity with an increase in the current

magnitude. Accordingly, it can be concluded that the fresh water discharge reduces the

salinity in the estuary.

Table 3.6 Characteristic salinity values
Julian days Maximum Salinity Mean Salinity Minimum Salinity
in 2002/03 (mg/L) (mg/L) (mg/L)
94-114 34.2 24.9 11.1
352-365 39.5 36.5 26.7
01-35




















Maximum 34 mg/L
Minimum 11 mg/L
Mean 24 mg/L


SO 95 100 105
JULIAN DAYS IN YEAR 2002

Figure 3 16 Salinity time series Days 94-114 (year 2002)


JULIAN DAYS IN YEAR 2002

Figure 3 17 Salinity and Current magnitude time series Days 94-114 (year 2002)







53


32 T


30 Maximum 310C
Minimum 210 C

o 28 Mean 26 C


12


W 24


22


204
90 95 100 105 110 115
JULIAN DAYS IN YEAR 2002

Figure 3 18 Temperature time senes Days 94-114 (year 2002)

Similarly the temperature time-series shows a positive correlation with the

discharge, with temperature increasing with the discharge from S-46 structure However

any defimte conclusion could not be deduced fiom this the absence of adequate data on

temperature of the freshwater discharged

For the second data block between days 352 and 365 (of year 2002) and days 01

and 35 (of year 2003) the Figure 3 19 indicates an apparent malfunctioning of the sensor

that seems to have contaminated the conductivity time senes that calculates the salimty

by measunng its conductivity of the solution at a given temperature Although the

temperature time senes for the same penod appears to give correct readig consistent

with the environment, the incorrect conductivity data have made the salimty

determination inaccurate Therefore salinity values reported in this penod appear to be

rather high Tables 3 5 and 3 6 summarize the charactenstic values of salinity and







54


temperature for both the data blocks The results from the other data blocks are furnished


in Table 3 7 to 3 9




Maximum 39 50 mgL
Mean 35 60 mg/L
SMinimum 26 70 mg/L
4J







20

10



355 360 365 370 375 3 0 385 390 395 400
DAYS OF THE YEAR 2002/03
Figure 3 19 Salinity time series Days 352- 365 (year 2002) and 01-35 (year 2003)


Maamum 2 C
o Mean 11C
2' MItnmml 8 C



0T15







350 355 360 365 370 375 380 385
JULIAN DAYS IN YEAR 200203
Figure 3 20 Temperature time senes 352- 365 (year 2002)

Table 3 7 Characteristic temperature values
Julian days Maximum Temperature Mean Ten
in 2002 (co C) ('o C)
94-114 314 263
352-400 267 109


390 395


and 01-35 (year 2003)










3.3.5 Other Data Blocks

The foregoing discussions included the various aspects of data collection their

analysis and results for two representative data blocks (Julian Days 94 to 114, 330-365 in

year 2002 and 01 to 35 in year 2003). However since the second phase data collection

lasted from November 26th, 2002 to May 15th, 2003, it was considered necessary to

include the characteristic values obtained from the other data blocks, which would offer a

better insight in to the overall site conditions.

Table 3.8 Summary of parametric value (Days 37-59 in year 2003)
Parameter Maximum Mean Minimum
Depth (m) 1.9 1.2 0.5

OBS 1 (mg/L) *
OBS 2 (mg/L) 240 30 0.7
OBS 3 (mg/L) 110 20 1.0
Salinity (mg/L) 40 35 23
Temperature ("C) 27 16 11
Current Magnitude (m/s) *


* Poor quahty data

In Table 3.7 a summary of parametric

February 6h and February 28th is presented. The

are presented in Tables 3.8 and 3.9.

b1 T, r x* t,


values of the data collected between

Sdata obtained for the other two blocks


le j3.9 summary of paramedic value (Days 9u-101 in year 20u3)
Parameter Maximum Mean Minimum
Depth (m) 1.6 1.1 0.6

OBS 1 (mg/L) 2670 1710 1470
OBS 2 (mg/L) 80 50 1.0
OBS 3 (mg/L) 96 51 20
Salinity (mg/L) *
Temperature (TC) 20 11 3
Current Magnitude (m/s) *
*-Bad Data


Tla


'^' ^^^"`









Table 3.10 Summary of parametric value (Days 101-135 in year 2003)
Parameter Maximum Mean Minimum
Depth (m) 1.9 1.3 0.7

OBS 1 (mg/L) 2090 1670 1180
OBS 2 (mg/L) 170 46 2
OBS 3 (mg/L) 210 56 10
Salinity (mg/L) 26 19* 17*
Temperature ("C) 22 12 4
Current Magnitude (m/s) 1.40 0.60 0.10
Bad Data

3.4 Field Data Results in Northwest Fork

3.4.1 Field Setup

In the third phase of data collection, in the Northwest Fork, the instrument tower

included three optical backscatter sensors (OBS), a pressure transducer (for water level)

and a conductivity/temperature sensor. Data collection began on 08/14/2003. Data on

water level and TSS are presented. The conductivity/temperature sensor malfunctions

during this phase and yielded values of questionable accuracy. Hence these data are not

reported.

3.4.2 Tidal Level

The pressure transducer was located 0.45 m from the bed. Figure 3.21 shows the

original time series of the water level.














Maximum Level 1 75 m
Mean Level 12 m
Minimum Level 070 m


244 246 24 50 22 254 256
JULIAN DAYS IN YEAR 2003

Figure 3 21 Record of water level vacation Days 245 -255 (year 2003)


w Sprm. Range 0,0 rm
-0$ Neap Range 05 m







244 246 248 250 252 24 256
JULIAN DAYS IN YEAR 2003

Figure 3 22 Water level time series Upper plot shows original time senes with mean
trend and the lower plot is without the mean oscillations All levels relative to
NAVD 88 Days 245 -255 (year 2003)

In Figure 3 22 the upper plot shows 12-hourly mean trend with the original time


scenes, and in the lower plot this trend is removed As can be seen from the latter plot, the






58


nsmg mean trend indicates the effect of fresh water discharge The tidal range was 0 80

m Characteristic values are given in Table 3 10

Table 3 11 Charactenstic values of the tidal data
J n Mean Water Level/Tidal range
in 2003 Water depth Water level (m)Spng/nea range (m)
(m) Maximum Minmum Spnng Neap
245-255 120 175 0 70 0 80 0 50

3.4.3 Total Suspended Solids

Total suspended sohds (TSS) concentration was recorded at three elevations The

elevations of the OBS sensors relative to the bed were OBS-1 = 1 04 m, OBS-2 = 0 66 m

and OBS-3 = 0 30 m The corresponding depth-mean concentration time senes is

reported in Figure 3 23 Characteristic values are given in Table 3 11


Maximum 230
Mean 100
Minimum 50


244 246 248 250 252 254
JULIAN DAYS IN YEAR 2003
Figure 3 23 Depth-meanTSS concentration time-senes Days 245-255

Table 3 12 TSS concentrations for the representative data blocks


256


(year 2003)


Julian Days Maximum TSS MeanTSS Mimnmum TSS
in 2003 concentration concentration concentration
(mg/L) (mg/L) (mg/L)
245-255 230 100 50










3.4.5 Additional Data Blocks

3.4.5.1 Tidal Level

Two additional data blocks were collected between November 6h, 2003 and

November 24 2003. Tide data for Julian days 310 and 313 are presented here. The

remainder was found to be of poor quality.


311 311 5 312 3125 313
JULIAN DAYS IN YEAR 2003


3135


Figure 3.24 Record of water level variation. Days 310.5


313.5 (year 2003).




























311 3115 312 3125
JULIAN DAYS IN YEAR 2003


Figure 3 25 Water level time series Upper plot shows onginal time
trend and the lower plot is without the mean oscillations
NAVD 88 Days 310 5 313 5 (Year 2003)


scenes with mean
All levels relative to


Table 3 13 Characteristic values of the tidal data
Mean Water Level/Tidal range
Jul days in Water depth Water level (m) Spnng/neap range (m)
(m) Maximum Miimum Spring Neap
310 5-3135 140 1 90 1 00 090 050


3.4.5.2 Total Suspended Solids

Two data blocks for the TSS concentration was collected and are presented below


Charactenstic values are presented in Table 3 14














Maxmum 834 mg/L
Mean 578 mgL
Min 397 mg/


312 3125 313


OBS1


311 3115 312 3125
JULIAN DAYS IN YEAR 2003


Figure 3 26 TSS time-senes at two elevations Days 310 5


OBS 2
45

40
EJ


311 31 2 3125
JULIAN DAYS IN YEAR 2003


313 5 (year 2003)


Figure 3 27 Depth meanTSS concentration time senes Days 310 5


311 3115


313 5 (year 2003)










OBS 3


60-Mean 190 m
Mmi 156 mg/L
140
3155 31 36165 317 3175
JULIAN DAYS IN YEAR 2003
Figure 3 28 TSS time-senes at three elevations Days 315 5


OBS3


318 3185

318 5 (year 2003)








318 3185


EO
76


318 3185


Figure 3 29


220
10 --




3155 316 3165 317 3175 318 3185
JULIAN DAYS IN YEAR 2003
Depth-mean TSS concentration time senes Days 315 5 318 5 (year 2003)


Table 3 14 TSS concentrations for the representative data blocks
Julian Days Maximum TSS MeanTSS MimmumTSS
in 2003 concentration concentration concentration
(mg/L) (mg/L) (mg/L)
3105 3135 834 304 19
3155-3185 219 140 81


316 3165 0& 1















CHAPTER 4
MODEL CALIBRATION AND VALIDATION

The analyzed data presented in Chapter 3 give a qualitative insight into the

prevailing environmental conditions. However, in order to have a quantitative

understanding of the flow regime in the estuary, it is necessary to apply a numerical

simulation technique. This chapter includes a brief description of the numerical model,

generation of the computational grid, initial and boundary conditions and the model

operational scheme. Model calibration and validation are then carried out.

Certain aspects of the estuary have been idealized in the formulation of the model

in order to reduce the computational time and avoidance of potential errors. These

idealizations are as follows:

1. The central embayment domain is terminated at the FECRR bridge excluding the
ICWW (Intracoastal Waterway). This enables use of tide data from UFG1 gage
installed at the bridge.

2. The traps and the navigation channels have rectangular cross-sections.

4.1 Model Description

Flow simulations were carried out using Environmental Fluid Dynamics Code

(EFDC) maintained by the Environmental Protection Agency, and developed by

Hamrick, 1992. This code works through a Microsoft Windows-based EDFC-Explorer

pre- and post-processor. Developed on a Fortran platform, the physics of EFDC and

many aspects of the computational scheme are equivalent to the widely used Blumberg-

Mellor model (Blumberg and Mellor, 1987) and the U.S. Army Corps of Engineers'

Chesapeake Bay model (Johnson, et al, 1993). EFDC solves the three-dimensional









hydrostatic, free surface, turbulent averaged equations of motion of a variable density

fluid. The model uses a stretched or sigma vertical coordinate and Cartesian or

curvilinear, orthogonal horizontal coordinates. Dynamically coupled transport equations

for turbulent kinetic energy, turbulent length scale, salinity and temperature are also

solved. Externally specified bottom friction can be incorporated in the turbulence closure

model as a source term. For the simulation of flow in vegetated environments, EFDC

incorporates both two and three-dimensional vegetation resistance formulations

(Moustafa, and Hamrick 1995).

The numerical scheme employed in EDFC to solve the equations of motion uses

second-order-accurate spatial finite difference on a staggered- or a C-grid. The model's

time integration employs a second-order-accurate, three-time-level, finite-difference

scheme with an internal-external mode splitting procedure to separate the internal shear

or baroclinic mode from the external free surface gravity wave or barotropic mode. The

external mode solution is semi-implicit, and simultaneously computes the two-

dimensional surface elevation field by the preconditioned conjugate gradient procedure.

The external solution is completed by the calculation of the depth averaged barotropic

velocities using the new surface elevation field. The models' semi-implicit external

solution allows large time steps that are constrained by the stability criteria of the explicit

central difference or upwind advection scheme used for the nonlinear accelerations.

Horizontal boundary conditions for the external mode solution include the option for

simultaneously specifying the surface elevations, the characteristic of an incoming wave,

free radiation of an outgoing wave or the volumetric flux on arbitrary portions of the

boundary. The model's internal momentum equation solution, at the same time step as








the external, is implicit with respect to vertical diffusion. The internal solution of the

momentum equations in terms of the vertical profile of shear stress and velocity shear,

which results in the simplest and most accurate form of baroclinic pressure gradients, and

eliminates the over-determined character of alternate internal mode formulations.

The model implements a second order accurate in space and time, mass

conservation fractional step solution scheme for the Eulerian transport equation at the

same time step or twice the time step of the momentum equation solution. The advective

portion of the transport solution uses either the central difference scheme used in the

Blumberg-Mellor model or hierarchy of positive definite upwind difference schemes. The

highest accuracy up-wind scheme, second order accurate in space and time, is based on a

flux corrected transport version of Smolarkiewicz's multidimensional positive definite

advection transport algorithm, which is monotonic and minimizes numerical diffusion.

The EFDC model's hydrodynamic component is based on the three-dimensional

hydrostatic equations formulated in curvilinear-orthogonal horizontal coordinates and a

sigma or stretched vertical coordinate. The momentum equations are:

4 (mnm,Hu)+ ed(mHuu ) ey (mnHvu)+ (n(mxmwu)- f ,mm Hv

=-myHdj+P + + )+ 0my (* + zrH p+9 mxm, A u

+4 HAu +d 4 HA,u m my cDp(U2 + v2 /2 U
m \m ) (4.1)

4 m(nm Hv)+ d, (m Huv)+ (mnHvv)+ mx (m mwv)+ fm, mHu

= -mH4,(P+Pom + 0)+ nk (r': + H p + mmym v (4.2)

+d HAX +4 HA v mD(2 +V2 2
\m+ ( LH I A V) MY y) i/









m m f = mXm f U mx + v my (4.3)


(rxz, yz)= AH- lz (, v) (4.4)

where u and v are the horizontal velocity components in the dimensionless

curvilinear-orthogonal horizontal coordinates x and y, respectively. The scale factors of

the horizontal coordinates are mx and my. The vertical velocity in the stretched vertical

coordinate z is w. The physical vertical coordinates of the free surface and bottom bed

are z, and z* respectively. The total water column depth is H, and 0 is the free surface

potential which is equal to gz, The effective Coriolis acceleration f incorporates the

curvature acceleration terms, with the Coriolis parameter, f according to (4.3). The Q

terms in (4.1) and (4.2) represent optional horizontal momentum diffusion terms. The

vertical turbulent viscosity A, relates the shear stresses to the vertical shear of the

horizontal velocity components by (4.4). The kinematic atmospheric pressure, referenced

to water density, is p,, while the excess hydrostatic pressure in the water column is given

by:

,p = -gHb = -gH(p- po )po' (4.5)

where p and po are the actual and reference water densities and b is the buoyancy.

The horizontal turbulent stress on the last lines of (4.1) and (4.2), with A, being the

horizontal turbulent viscosity, are typically retained when the advective acceleration are

represented by central differences. The last terms in (4. 1) and (4.2) represent vegetation

resistance where c, is a resistance coefficient and D, is the dimensionless projected

vegetation area normal to the flow per unit horizontal area.

The three-dimensional continuity equation in the stretched vertical and

curvilinear-orthogonal horizontal coordinate system is:









4 (m,yH)+ (mHu)+ d (mHv) + e(m, ( w)= Q, (4.6)

with QH representing volume sources and sinks including rainfall, evaporation, infiltration
and lateral inflows and outflows having negligible momentum fluxes.
The solution of the momentum equations, (4.1) and (4.2) requires the specification

of the vertical turbulent viscosity, A,, and diffusivity, K,. To provide the vertical turbulent

viscosity and diffusivity, the second moment turbulence closure model developed by

Mellor and Yamada (1982) (MY model) and modified by Galperin et al (1988) and

Blumberg et al. (1988) is used. The MY model relates the vertical turbulent viscosity

and diffusivity to the turbulent intensity, q, a turbulent length scale, 1, and a turbulent

intensity and length scaled based Richardson number, R,, by:

A, = ql

A Ao(I+R1R )


A= A(1 3C,1 = B--


(B2-3 3C (B2 +6A1)
S= 3A
6A
1- 3C -B

1 = 9A1A2
R31 =3A (6A, + B2) (4.7)

K, = ql
01 KK


K=4( (4.8)


R gHdb 12 (4.9)
S q2 H2









where the so-called stability functions, 0, and 0K, account for reduced and enhanced

vertical mixing or transport in stable and unstable vertically density stratified

environments, respectively. Mellor and Yamada (1982) specify the constants A,, B,, C,,

A2, and B as 0.92, 16.6, 0.08, 0.74, and 10.1, respectively.

For stable stratification, Galperin et al. (1988) suggest limiting the length scale

such that the square root of R, is less than 0.52. When horizontal turbulent viscosity and

diffusivity are included in the momentum and transport equations, they are determined

independently using Smagorinsky's (1963) sub-grid scale closure formulation.

At the bed, the stress components are presumed to be related to the near bed or

bottom layer velocity components by the quadratic resistance formulation

( "r ) '"ywr2 1 2 ( u1, (4.10)
(rxzT yz), = ( rb by) = c u+ l (l) (410)

where the 1 subscript denotes bottom layer values. Under the assumption that the near

bottom velocity profile is logarithmic at any instant of time, the bottom stress coefficient

is given by


Cb K)2
( In(A1/2z,)) (4.11)


where c is the von Karman constant, A, is the dimensionless thickness of the bottom

layer, and zz, /H is the dimensionless roughness height. Vertical boundary conditions

for the turbulent kinetic energy and length scale equations are:

q2 = B2/3 :z =1 (4.12)

q2 = B23 : z = 1 (4.13)

1=0 : z=0,1 (4.14)









where the absolute values indicate the magnitude of the enclosed vector quantity which

are wind stress and bottom stress, respectively.

4.3 Grid Generation

The first step in the setup of the modeling system is to define the horizontal plane

domain of the region being modeled. The horizontal plane domain is approximated by a

set of discrete quadrilateral and triangular cells. Developed on a digitized shoreline, the

grid defines the precise locations of the faces of the quadrilateral cells in the horizontal as

well as in the vertical plane. However, all the computations are carried out at the center of

the cells. Since the model solves the hydrodynamic equations in a horizontal coordinate

system that is curvilinear and orthogonal, grid lines also correspond to lines having a

constant value of one of the horizontal coordinates. The shoreline as well as the cell

reference is provided by a local set of Coordinates in MKS unit, as the code uses MKS

system internally. Seven identification numbers were used to define the cell types. The

cell identification details are given in Table 4.1.

Table 4.1 Definition of cell type used in the model input
Cell ID Definition of cell type
0 Dry land cell not bordering a water cell on a side or corer of the model
1 Triangular cell with land to the northeast of the model
2 Triangular cell with land to the southeast of the model
3 Triangular cell with land to the southwest of the model
4 Triangular cell with land to the northwest of the model
5 Quadrilateral water cells of the model
9 Dry land cell bordering a water cell on a side or on a corer of the model


The type 9 dry land or fictitious dry land cell type is used in the specification of

no flow boundary conditions. The horizontal geometric and topographic (bottom

bathymetry) and other related characteristics of the region, files dxdy.znp and lxly.znp are

used. The program then directly reads these quantities expressed in meters. The lxly.znp









provides cell center coordinates and components of a rotation matrix. Cell center

coordinates are used only in graphics output and can be specified in the most convenient

units for graphical display such as decimal degrees, feet, miles, meters or kilometers. The

rotation matrix is used to convert pseudo east and north (curvilinear x andy ) horizontal

velocities (u and v respectively) to true east and north for graphics vector plotting,

according to;



S tn) cue cvn e co J

where the subscripts te and tn denote true east and true north, while the subscripts

co denotes the curvilinear-orthogonal horizontal velocity components. The coefficient C

is the multiplier term for conversion to true east and true north.

The width of the C-18 canal, which varies between 75 m at the Southwest Fork

junction to less than 40m at the S-46 structure, dictated the dimensions of the cells. It was

decided that a 25 x 25m cell would be accurate enough for representing the width of the

C-18 canal resulting in desired level of accuracy. The same cell size was then

conveniently extended to the rest of the model domain. The bottom bathymetry was

based on the Hydrographic survey carried out in November '2001 by Lidberg Land

Surveying, Inc. However additional data for areas not covered under this survey were

obtained from other available surveys. The roughness coefficient of the bottom

bathymetry in the model is composed of two components. A fixed component viscosity

(for the present model fixed at 0.020m) and a variable component, which is varied

uniformly on the entire model domain during calibration process, both the component

together constitutes the factor z0, defined in equation 4.11. The dimensionless thickness






71

of the bottom layer AI, defined in the same equation, equals to 0 25, since four vertical

layers are used The fixed component of the roughness factor, how ever can be

increased/decreased in the areas of vegetation or other special features The details of sea

grass locations in the central embayment can be referred fiom Drawing no LOX-001

(Cuthcher & Associates, Inc Coastal Engineer, 2002) provided by the Jupiter Inlet

District The sea grass was input in the model as an overlay file In this way the cells

having the sea grasses are enclosed by a polyline so that, the roughness coefficient can be

easily edited The sea grass was represented as cells having more roughness (fixed

component = 0 040m) than that of the surroundings In Figure 4 1 the input bathymetry

and the shoreline as generated by the model are shown


5





*11


Figure 4 1 Model domain showing input bathymetry and shoreline










BottomElev
6S36 Time27500 306
_* _____ _


Figure 4.2 Computational grid showing the flow boundaries

In the computational grid (Figure 4.2), each land cell was assigned number zero or

nine as the case may be and each water cell was assigned five. There were no triangular

cells used for this grid. Figure 4.2, in addition, indicates the locations of the tide gages

and the Instrument tower in the Southwest fork. The S-46 structure in the C-18 canal is a

flow boundary (black cells), as are the two main tributaries, and the FECRR bridge on the

East. The eastern boundary was restricted to the FECRR bridge. The flow boundaries

were kept straight; so as to allow flows perpendicular to the cell faces, as the model does

not allow non-orthogonal flows.

4.4 Boundary Conditions

In the beginning of the simulation, velocities throughout the model domain are

considered to be zero. It was observed that a full tidal cycle was required before the water

surface elevation reached a quasi-steady state. This was verified by recording water









surface elevations at the location of the two tide gauges (UFG2 and UFG3) over multiple

tidal periods.

Tidal forcing at the FECRR bridge (eastern boundary) is perhaps the most

important boundary condition in this system, because it is this mechanism by which the

majority of the water flows through the estuary. The data obtained from the UFG1 gage

(Figure 2.10) were used to simulate this forcing. The raw data were examined for the

mean trends in the water surface elevation (Figure 4.3). The raw data contains a sub-tidal

frequency trend, which was also noticed in the water surface elevation data of the Miami

Harbor. The trends were of a similar in nature and therefore it was hypothesized that

onshore winds may have created increased elevation in side the estuary. The wind records

from two offshore sites (37 and 221 kilometer east of Cape Carnival, Florida) were

correlated with the mid-tide elevation, which indicated a positive correlation (Ganju et

al., 2001). In order to overcome the effects of these variations imposed on the

astronomical tide, the mid-tide elevation was subtracted from each measured elevation in

the same tidal cycle. The mid tide elevation 7,m is given by Equation 4. 1, where, 77H and

77r are the water surface elevation at high and low tides respectively.


c =HT + L (4.16)
2








74




08
a

06 -











-4










-08
06














S02
0














0.4


-06
0 50 100 150 200 250 300 350 400 450
TIME (h)

















Figure 4 3 Tidal time sees from UFGI, 09/14/00-10/13/00, a) Raw data, b) Tidal plot
after the mid-tie trend is removed Time o 12 00 am
0204 i 1 j i M I n









In Figure 4.3a the raw tidal time series is shown along with the tidal trend and in

Figure 4.3b the tidal time series is shown after subtracting the mean-tide trend. The

eastern boundary accordingly used this water surface elevation boundary condition.

For the boundary in the C-18 canal, two sets of boundary condition data were

available. The daily average flow time series of the S-46 structure and the water surface

elevation time series. The elevation time series was obtained from the tide gauge UFG 3

(same period as at UFG 1) installed in the Southwest Fork (Figure 2.10). In order to make

these data usable at the flow boundary (S-46 Structure) amplitude corrections were

carried out by trial and error till both predicted and measured time series matched. In

order to calculate the phase correction (lag) following calculations were carried out

assuming shallow water conditions. The tidal wave celerity C is given by,

C = h (4.23)

where, g is the acceleration due to gravity and h is the water depth. Then the phase

shift AT is given by,


AT' AT = (4.24)
C


where, AL is the distance for which the water depth is considered uniform, accordingly

the phase lag for the distance between the UFG 3 gage station and the S-46 structure was

calculated and verified (0.13 hour). Figure 4.4 gives the plot of the raw data collected at

UFG 3, including the mean trend and the amplitude with trends removed. It was

hypothesized that these data, corrected for the phase and amplitude could be applied as

boundary condition to simulate actual flow conditions.
































TIME (h)


O
02















TIME (h)

Figure 4 4 Tidal time sees from UFG3, 09/14/00-10/13/00, a) Raw data, b) Tidal plot
after the mid-tide trend is removed Time ongm 12 00 am

Note that the flow discharge time series (Flgure 4 5) from the S-46 structure was

selected, as the model is known to be giving better simulation results under discharge
g:F- ~ ~i IH || II
I-
lii 6
-0 8 --.-- '-------------------
0 C(O 15 0 50 30 30 0 5
TIEjh
Fiue4 ia tm eesfo UG,0/1/01/1/0 a a atOb idlpo
afe2h i-ietedi rmvdTm rgn1 0a
Noeta0h lwdshretm ee (iue45 rmteS4 tutr a

selected, O as 4emdli nw ob igbttrsmlto eut ne icag


boundary condition










70 00


6000)


75o00
E
w 40i0
0
< 3000
M M
L)
U)m


1000


S -l rF -r v


I I I J I i
V

b Iit i


000 50000 10000 15000 20000 25000 30000 35000 40000 45000
0 0 0 0 0 0 0 0
DAYS
Figure 4.5 Flow time series applied at S-46 boundary


0 1000 2000 3000 4000


5000


Figure 4.6 Flow time series applied at


DAYS
Northwest Fork boundary


In the Northwest Fork boundary as well, two sets of boundary conditions, namely,

the water surface elevation boundary condition (obtained from transferring the collected

data of the tidal station UFG 2) and flow discharge boundary condition were evaluated.

The flow time series used is shown in Figure 4.6.

Table 4.2 Amplitude and phase correction factor for the tides
Boundary Amplitude factor Phase correction
C-18 1.14 0.13 hour
Northwest Fork 1.18 0.042 hour


I


I











Per U.S. Geological Survey Report 84-4157 (Russell and McPherson, 1984) the

majority (77.3 %) of the fresh water flow in to the estuary enters through the Northwest

Fork. Therefore, the flow discharge boundary condition for this tributary was considered

as most appropriate as opposed to the water surface elevation. The corrected water

surface elevation data from UFG 2 tide gage was used for calibration.

The North Fork carries the least discharge (2.2%) of the total freshwater flow in

to the estuary in the mean, (Russell and McPherson, 1984)) hence at this boundary also

flow discharge boundary condition is applied. The flow discharge was worked out from

2.2
the Northwest boundary data applying a constant multiplier ( = 0.0285).
77.3

4.5 Model Calibration and Validation

4.5.1 Calibration

In general calibration of the model aims at simulating conditions identical or close

to that in the prototype so that prototype conditions can be accurately replicated and

reproduced. Calibration involves matching multiple parameters, which is often times, is

practically impossible. However, depending on the nature experiments and the results

desired, the type of calibration differs. Since the present model simulation aims to relate

the velocity and the associated stress field to the erosion/accretion of the sediments in the

estuary, it would be highly desirable to calibrate the model with comparison of the flow

velocities. But current data for the model simulation period, between September 14t

2000 and 18t October 2000 was not available and therefore, it was decided to calibrate

the model using the data collected at the instrument station located in the Southwest Fork

(Figure 3.1) between November 26t and May 15t 2003 for which current as well as









water surface elevation data was available. The amplitude multiplier and phase lag

factors are given in Table 4.2.

Accordingly a simulation for this period was carried out using flow discharge

boundary conditions for the Southwest, the Northwest and North tributary boundaries and

water surface elevation boundary condition for the East boundary. For the eastern

boundary the tidal data from Miami Harbor were "transferred" to the FECRR bridge

boundary by applying suitable correction factors for the amplitude and the phase lag. This

procedure was carried out in two steps. In the first step, the Miami harbor data for the

period 14t September 2000 to 18th October 2000 were transferred to the boundary with

application of recommended coefficients (for method of calculation refer to NOS Tide

Tables for year 2000). The calculated tidal elevations were compared with the UFG 1

data and the final multiplication correction factor was obtained as 1.023. For the model

simulation period in year 2002 the same correction factor was used to transfer tidal

elevations of Miami harbor to the flow boundary.

Model calibration began with an initial run for 48 hours (referred to as 'cold start')

in order to make the tide and discharge mutually compatible throughout. In addition, the

flow attains stability in this period. The results of the cold start period were compared

with the current velocities as well as the water surface elevations obtained from the

instrument tower. The process was continued by changing the variable component of the

bottom friction coefficient (one component of z0) (applicable uniformly throughout the

model domain), until an approximate match of the current magnitude and phase was

obtained. In the second step RESTART.OUT and RSWT.OUT, the two output files of the

cold start were used as input, and model run was performed for a longer period (15days)









in order to obtain simulation for final calibration. The predicted and measured currents

were then compared and is given in Figure 4.7a (Cold start) and 4.7b (Hot start) for a

variable bottom friction factor of 0.027. It can be seen that the agreement is very good for

the current, with a maximum error of 1.8% of the total current amplitude. The water

surface elevation however differs by about 2.8 cm, which is about 3% of the tidal

amplitude. Since current is in better agreement with the measured data the calibration was

considered accurate enough for simulation. In addition, comparison of the predicted and

measured current direction exhibited good agreement as indicated in 4.8.

4.5.2 Model Validation

Model validation was carried out using the same calibrated parameters and

simulating the flow conditions of year 2000 (between September 14t and October 18t).

The measured as well as the model results at both the tidal gage stations after cold start as

well as hot start periods are compared and reproduced as Figures 4.9 and 4.10. As

indicated in the figure 4.10, the agreement is fairly accurate with a maximum variation of

2.7 cm, which is about 3.4% of the maximum tidal amplitude reported in the estuary.

Similar validation was also carried out using the Northwest Fork data collected between

September 3rd and September 12b which also showed equally good agreement as shown

in Figure 4.11.






























JULIAN DAYS IN 2002


----- Model Run Measured


JULIAN DAYS IN YEAR 2002

----- Model run Measured


Figure 4.7 Model calibration measured vs. predicted current, a) Cold Start, b) Hot Start.












200 00

150 00-

10000-------

5000 ----------- -- ------ -----------
z
S 0*0
w 33"00 3320 33340 33, 0 33380 3340 20 33410 33460 3480 3500
F -50 00
I-
z
0 -100 00 -

-15200 ---- --- ---------------

-200 00-

-250 00
JULIAN DAYS IN YEAR 2002

.... Model -- Measured



Figure 4.8 Model calibration measured vs. predicted current direction.
































JULIAN DAYS IN 2000


------ Model Measured


0 6000


JULIAN DAYS IN 2000


....... Model Measured


Figure 4.9 Model calibration measured vs. predicted water surface elevation (UFG2)
Year 2000, a) Cold Start, b) Hot start.





























JULIAN DAYS IN 2000


----- Model Measured


0 6000


JULIAN DAYS IN 2000


----- Model Measured

Figure 4. 10 Model calibration measured vs. predicted water surface elevation (UFG3)
Year 2000, a) Cold Start b) Hot start.




Full Text

PAGE 1

SEDIMENT MANAGEMENT IN LOW ENERGY ESTUARIES: THE LOXAHATCHEE, FLORIDA By RASHMI RANJAN PATRA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2003

PAGE 2

ii ACKNOWLEDGMENTS My sincere gratitude is reserved for Dr. Ashish J. Mehta for his guidance in my education and research, which made my studi es very precise and rewarding, as well as the entire Coastal and Oceanographic Engineer ing Program faculty. Also deserving my gratitude for their guidance and assistance are Dr. John Jaeger, Dr. William McDougal, and Kim Hunt. Most of the analysis in the study was made possible by the valuable assistance provided by Mr. Sidney Schofield, w ho taught me the basics of analysis. Special thanks are due to Dr. Earl Ha yter for setting up, supporting and guiding me through the numerical model in its entirety. Thanks are also due to Dr. Zal S. Tara pore, for his encouragement and guidance, which marked my initial years as a coastal e ngineer and my studies here possible. My wife Sumitra and my friend Anjana also de serve special kudos for their emotional and editorial support. Finally, my mother and fa ther merit unlimited praise for providing me with mind, body and soul, as do my other friends and families for developing it.

PAGE 3

iii TABLE OF CONTENTS page ACKNOWLEDGMENTS..................................................................................................ii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES.........................................................................................................viii LIST OF SYMBOLS........................................................................................................xii ABSTRACT....................................................................................................................... xv CHAPTER 1 INTRODUCTION........................................................................................................1 1.1 Problem Statement.............................................................................................1 1.2 Study Tasks........................................................................................................3 1.3 Outline of Chapters............................................................................................3 2 SEDIMENT MANAGEMENT ALTERNATIVES.....................................................4 2.1 Present Condition of the Loxahatchee Estuary..................................................4 2.2 C-18 Canal.......................................................................................................12 2.2.1 Present Condition.......................................................................................12 2.2.2 Management Options..................................................................................17 2.3 Central Embayment.........................................................................................21 2.3.2 Management Option:..............................................................................27 2.4 Northwest Fork:...............................................................................................28 2.4.1 Present Condition:...................................................................................28 2.4.2 Management Options..............................................................................32 2.5 North Fork........................................................................................................33 2.5.1 Present Condition....................................................................................33 2.5.2 Management Options..............................................................................34 3 DATA COLLECTION...............................................................................................35 3.1 Field Setup in the Southwest Fork...................................................................35 3.2 Instruments Deployed......................................................................................37 3.2.1 Current....................................................................................................37 3.2.2 Tide.........................................................................................................38

PAGE 4

iv 3.2.3 Salinity/Temperature...............................................................................38 3.2.4 Sediment Concentration..........................................................................39 3.3 Field Data Results in Southwest Fork..............................................................41 3.3.1 Current....................................................................................................41 3.3.2 Tidal Level..............................................................................................44 3.3.3 Total Suspended Solids...........................................................................47 3.3.5 Other Data Blocks...................................................................................55 3.4 Field Data Results in Northwest Fork..............................................................56 3.4.1 Field Setup..............................................................................................56 3.4.2 Tidal Level..............................................................................................56 3.4.3 Total Suspended Solids...........................................................................58 3.4.5 Additional Data Blocks...........................................................................59 3.4.5.1 Tidal Level....................................................................................59 3.4.5.2 Total Suspended Solids.................................................................60 4 MODEL CALIBRATION AND VALIDATION......................................................63 4.1 Model Description...........................................................................................63 4.3 Grid Generation...............................................................................................69 4.4 Boundary Conditions.......................................................................................72 4.5 Model Calibration and Validation...................................................................78 4.5.1 Calibration...............................................................................................78 4.5.2 Model Validation....................................................................................80 4.5.3 Simulation of trap scheme of Ganju, 2001.............................................86 5 EVALUATION OF SEDIMENTAT ION CONTROL ALTERNATIVES................87 5.1 Design Basis.........................................................................................................87 5.1.1 General Principle....................................................................................87 5.1.1.1 Sediment Entrapment.......................................................................87 5.1.1.2 Self-cleaning Channel......................................................................88 5.1.2 Design Alternatives.................................................................................89 5.1.2.1 Alternative No. 2: C-18 Canal Trap.................................................90 5.1.2.2 Alternative No. 3: Bay Channel.......................................................91 5.1.2.3 Alternative No. 4: Bay Y-channel....................................................93 5.1.2.4 Alternative No. 5: Northwest Fork Channel....................................94 5.1.3 Efficiency Analysis....................................................................................94 5.1.3.1 Velocity Vector Calculation.............................................................94 5.1.3.2 Sediment Deposition Calculation.....................................................95 5.1.3.3 Trap Efficiency.................................................................................97 5.1.3.4 Channel Efficiency...........................................................................98 5.2 Design Simulations...............................................................................................98 5.2.1 Design Flows..........................................................................................98 5.2.2 Alternative 1............................................................................................98 5.2.3 Alternatives 2, 3, 4 and 5........................................................................99 5.3 Deposition Equation Calibration....................................................................102 5.3.1 Calibration for Sand.................................................................................102

PAGE 5

v 5.3.2 Fine Sediment...........................................................................................102 5.4 Sand Deposition due to Alternatives..............................................................103 5.4.1 Bay Channel..........................................................................................103 5.4.2 C-18 Canal............................................................................................103 5.4.3 Bay Y-channel......................................................................................104 5.5 Fine Sediment Deposition due to Alternatives.............................................104 5.6 Sediment Removal.........................................................................................105 5.6.1 Calculation of Deposition.....................................................................105 5.6.2 Calculation of Channel Efficiency........................................................106 5.6.3 Removal of Bay Sediment....................................................................107 5.7 Assessment of Alternatives.................................................................................107 6 CONCLUSIONS......................................................................................................109 6.1 Summary........................................................................................................109 6.2 Conclusions....................................................................................................110 6.3 Recommendations for Future Work...............................................................112 LIST OF REFERENCES.................................................................................................113 BIOGRAPHICAL SKETCH...........................................................................................116

PAGE 6

vi LIST OF TABLES Table page 2.1 Basin area distributions in the L oxahatchee River estuary watershed.......................7 2.2 Statistical tributary flow (based on Figures 2.6 a-c)................................................14 2.3 Median and high flow concentration da ta and coefficients for equation 2.1...........15 2.4 Spring/neap tidal ranges and phase lags for three gauges........................................27 3.1 Instrumentation for data collection and data blocks.................................................36 3.2 Discharge data for the period 04/14/2002 to 04/21/200...........................................41 3.3 Typical mean current magnitude values for data blocks..........................................44 3.4 Characteristic values of the tidal data......................................................................47 3.5 TSS concentrations for the representative data blocks.............................................51 3.6 Characteristic salinity values....................................................................................51 3.7 Characteristic temperature values............................................................................54 3.8 Summary of parametric va lue (Days 37-59 in year 2003).......................................55 3.9 Summary of parametric va lue (Days 90-101 in year 2003).....................................55 3.10 Summary of parametric va lue (Days 101-135 in year 2003)...................................56 3.11 Characteristic values of the tidal data......................................................................58 3.12 TSS concentrations for the representative data blocks.............................................58 3.13 Characteristic values of the tidal data......................................................................60 3.14 TSS concentrations for the representative data blocks.............................................62 4.1 Definition of cell type used in the model input........................................................69 4.2 Amplitude and phase correction factor for the tides................................................77

PAGE 7

vii 5.1 Alternative schemes for evaluation..........................................................................89 5.2 Critical velocities for sand........................................................................................96 5.3 Design flows in tributaries.......................................................................................98 5.4 Maximum currents at alternatives: calibration discharges.....................................100 5.5 Maximum currents at alternat ives: Different discharges.......................................100 5.6 Calibration for sediment fluxes..............................................................................102 5.7 Rate of sand deposition in bay channel..................................................................103 5.8 Rate of sand deposition in C-18 canal....................................................................103 5.9 Rate of sand deposition in Y-channel....................................................................104 5.10 Rate of fine sediment deposition in alternatives....................................................105 5.11 Annual sand budget: Calibration discharge...........................................................105 5.12 Annual sand budget: Peak discharge......................................................................105 5.13 Annual fine sediment budget: Calibration discharge.............................................106 5.14 Annual fine sediment budget: Peak discharge.......................................................106 5.15 Annual sediment loading........................................................................................106 5.16 Assessment of impacts of proposed alternatives....................................................108

PAGE 8

viii LIST OF FIGURES Figure page 2.1 Location map of the study area .................................................................................5 2.2 Loxahatchee River estuary and tributaries.................................................................5 2.3 Hydrographic survey of the estuary (November 2001)..............................................7 2.4 Loxahatchee River estuary watershed basins, estuarine limits (dashed arcs) and central embayment.....................................................................................................8 2.5 Ariel photograph showing development of new shoal.............................................11 2.6 Cumulative discharge plot. a) Northwest Fork. b) North Fork. c) Southwest Fork.....................................................................................................14 2.7 Dredging Plans for C-18 canal, 1956 ......................................................................16 2.8 Current variation under the effect of released discharge from the S-46 structure....................................................................................................................19 2.9 Effect of S-46 discharge on the su spended sediment concentration........................19 2.10 Arial Photograph showing the Central Embayment, the Inlet and the Tributaries................................................................................................................23 2.11 Location of tide gauges ma rked UFG1, UFG2 and UFG3.......................................26 2.12 Sample records of tidal measurements at three locations (09/14/00-09/15/00)Datum NAVD 88.....................................................................................................27 2.13 Location of stream-gauging stations and sampling site for suspended sediments, ................................................................................................................30 2.14 Location indicating fresh mud depositi ons and the Shoals the estuary....................33 3.1 Location of instrument tower in the Southwest and Northwest Forks.....................35 3.2 Calibration plots used for calibration of OBS sensors.............................................40 3.3 Record of current magnitude: Days 94-114 (year 2002)..........................................42

PAGE 9

ix 3.5 Record of current magnitude: Days 332-356 (year 2002)........................................43 3.6 Record of current direction: Days 332356 (year 2002).........................................43 3.7 Water level time-series: All levels relative to NAVD 88. Days 94-114 (2002)......44 3.8 Water level time series: Upper plot shows original time series with mean trend and the lower plot is without the mean oscillations. All levels relative to NAVD 88. Days 94-114 (2002)...............................................................................45 3.9 Water level time-series: All levels relative to NAVD88. Days 332365 (2002) and Days 01-35 (2003).............................................................................................45 3.10 Water level time series: Upper plot shows original time series with mean trend and the lower plot is without this trend. All level relative to NAVD 88. Days 332-365 (2002) and Days 01 – 35 (2003)................................................................46 3.11 TSS time-series at four elevations: Days 94-114 (year 2002).................................48 3.12 TSS time-series at three elevations: Days 352365 (year 2002) and 01–35 (year 2003)...............................................................................................................48 3.13 Depth-mean TSS concentration ti me series: Days 94-114 (year 2002)...................49 3.14 Depth mean TSS concentration time series: Days 352365 (year 2002) and Days 01 – 35 (year 2003).........................................................................................49 3.15 Depth mean TSS concentration time series and tidal trend indicating their dependence: Days 352365 (year 2002) and Days 01 – 35 (year 2003).................50 3.16 Salinity time series: Days 94-114 (year 2002).........................................................52 3.17 Salinity and Current magnitude tim e series: Days 94-114 (year 2002)...................52 3.18 Temperature time series: Days 94-114 (year 2002).................................................53 3.19 Salinity time series: Days 352365 (year 2002) and 01-35 (year 2003).................54 3.20 Temperature time series: 352365 (year 2002) and 01-35 (year 2003)...................54 3.21 Record of water level vari ation. Days 245 – 255 (year 2003).................................57 3.22 Water level time series: Upper plot shows original time series with mean trend and the lower plot is without the mean oscillations. All levels relative to NAVD 88. Days 245 -255 (year 2003)....................................................................57 3.23 Depth-mean TSS concentration ti me-series: Days 245-255 (year 2003).................58 3.24 Record of water level varia tion. Days 310.5 – 313.5 (year 2003)...........................59

PAGE 10

x 3.25 Water level time series: Upper plot shows original time series with mean trend and the lower plot is without the mean oscillations. All levels relative to NAVD 88. Days 310.5 – 313.5 (Year 2003)............................................................60 3.26 TSS time-series at two elevat ions: Days 310.5 – 313.5 (year 2003).......................61 3.27 Depth mean TSS concentration tim e series: Days 310.5 – 313.5 (year 2003).........61 3.28 TSS time-series at three elev ations: Days 315.5 – 318.5 (year 2003)......................62 3.29 Depth-mean TSS concentration time series: Days 315.5 – 318.5 (year 2003)........62 4.1 Model domain showing input bathymetry and shoreline.........................................71 4.2 Computational grid showing the flow boundaries...................................................72 4.3 Tidal time series from UFG1, 09/14/0010/13/00, a) Raw data, b) Tidal plot after the mid-tide trend is removed..........................................................................74 4.4 Tidal time series from UFG3, 09/14/0010/13/00, a) Raw data, b) Tidal plot after the mid-tide trend is removed..........................................................................76 4.5 Flow time series applied at S-46 boundary..............................................................77 4.6 Flow time series applied at Northwest Fork boundary.............................................77 4.7 Model calibration measured vs. predicted cu rrent, a) Cold Start, b) Hot Start........81 4.8 Model calibration measured vs. predicted current direction....................................82 4.9 Model calibration measured vs. predicted water surface elevation (UFG2) Year 2000, a) Cold Start, b) Hot start...............................................................................83 4.10 Model calibration measured vs. predicted water surface elevation (UFG3) Year 2000, a) Cold Start. b) Hot start...............................................................................84 4.11 Model calibration measured vs. predicted water surface elevation (Northwest Fork) Year 2003, a) Cold Start, b) Hot start..........................................85 4.12 Validation results using trap used by Ganju, 2001...................................................86 5.1 Design concepts for sediment management.............................................................88 5.2 Alternatives considered, with existing bathymetry..................................................90 5.3 Dredged bathymetry of the C-18 canal with plan form view of the proposed trap. Trap considered by Ganju (2001) is also shown..............................................91 5.4 Planform view of the proposed self-cleaning channel in the bay............................92

PAGE 11

xi 5.5 Location of the sea grasses indicated in model with increased roughness...............92 5.6 Planform view of the proposed self-cleaning Y-channel in bay..............................93 5.7 Planform view of the proposed self-cl eaning channel in the Northwest Fork.........94 5.8 Current comparisons for a model cell at the upstream end of the Northwest Fork channel: calibra tion discharges........................................................................99 5.9 Current velocity vectors over the modeled domain; maximum flood velocities at spring tides.........................................................................................................101 5.10 Current velocity vectors over the modeled domain; maximum ebb velocities at spring tide...............................................................................................................101

PAGE 12

xii LIST OF SYMBOLS A area vA vertical turbulent viscosity B width of the basin C wave celerity 0C uniformly distributed initial sediment concentration (kg/m3) s C sediment concentration (kg/m3) cueC constant multiplier for u-velocity conversion to true east cunC constant multiplier for u-velocity conversion to true north cveC constant multiplier for v-velocity conversion to true east cvnC constant multiplier for v-velocity conversion to true north D sediment deposition under reduced flow enD deposition at the entrance to the channel exD deposition at the exit of the channel pD dimensionless projected vegetation area H total water column depth L length of the channel/trap K coefficient of conductance H Q volume source or sink q R Richardson number

PAGE 13

xiiiT time U steady mean flow velocity U velocity vector W width of the channel in equation 5.8 s W settling velocity b buoyancy pc vegetation resistance f Darcy-Weishbach friction factor e f coriolis acceleration g acceleration due to gravity h water depth xm scale factor along x-axis ym scale factor along y-axis q turbulent intensity () s iq amount of sediment in influent () s eq amount of sediment in effluent r removal ratio u velocity along the channel (x-axis) *u friction velocity cu velocity amplitude under current cenu current at the entrance cexu current at the exit

PAGE 14

xivcru critical velocity for erosion cou curvilinear-orthogonal horizontal velocity teu velocity in true east direction v velocity across the length of the channel(y-axis) cov curvilinear-orthogonal horizontal velocity tnv velocity in true north direction z variable water depth water density 0 reference water density s bed erosion shear stress bx x-component shear stress by y-component shear stress b bed bottom shear stress c mid-tide elevation H T high-tide elevation LT low tide elevation vertical diffusivity Karman constant free surface potential velocity angles

PAGE 15

xv Abstract of Thesis to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science SEDIMENT MANAGEMENT IN LOW ENERGY ESTUARIES: THE LOXAHATCHEE, FLORIDA By Rashmi Ranjan Patra December 2003 Chairman: Ashish J. Mehta Major Department: Civil and Coastal Engineering Implementation of schemes for sediment entrapment and self-cleaning channels was examined in the micro-tidal estuarin e environment containing both sand and fine sediment. The central embayment of the micro-tidal Loxahatchee River estuary on the Atlantic Coast of Florida was chosen as the candidate location due to its unique characteristics with respect to the influx of sand and fine sediment in its central embayment, and concerns regarding the potentia l for long term impacts of this flux on the embayment. An ideal sediment trap captures all of incoming sediment, i.e., the removal efficiency is 100%. A self-cleaning cha nnel allows no net deposition of incoming sediment, which passes through, so that its removal efficiency is nil. Hydrodynamic model simulations were ca rried out for selected trap/channel alternatives, and their efficiency was calcula ted by relating sediment deposition to change in the flow regime due to implementation of these alternatives. Calculations indicated

PAGE 16

xvi that the concepts of sediment entrapme nt and of self-cleaning can operate only imperfectly in the study area due to the lo w prevailing forcing by tide and the episodic nature of freshwater discharges in the tributaries. Fine sediment accumulation in the central embayment can be reduced by dredging the C-18 canal, as the trapped sediment would acc ount for more than half of the total fine sediment entering the bay. A channel close to the southern bank of the embayment could improve bay flushing by ebb flow, reduce ba y-wide sedimentation and serve as a navigation route. Careful design with regard to channel alignment would be required to avoid sea grass beds in the area. Long term simulations of flow and sediment transport are required to assess sediment circulation pa tterns and the formation of shoals in the central embayment and the Northwest Fork.

PAGE 17

1 CHAPTER 1 INTRODUCTION 1.1 Problem Statement Sedimentation due to the influx of fine a nd coarse particles is an issue affecting numerous estuaries and coastal waterways. Often enough, these particles originate far inland, and are transported into the coastal zone by runoff and stream flow. In the estuarine regime, inorganic sediment almo st never occurs in isolation, as it is complemented by measurable organic frac tion produced by either indigenous sources (e.g., native phytoplankton, swamp vegetation, wi nd blown material), or allochthonous sources (e.g., river-borne phytoplanktons, swamp vegetation, windblown material) (Darnell, 1967). In turn, such organic-ri ch sediments can degrade water quality by oxygen uptake and a reduction in light pene tration. In this study, the question of preemptive dredging of sediment prior to its deposition in an area of concern or, as an alternative, preventing its deposition in the area of concern by channelizing flow, was studied. The candidate water body was the estu arine segment of the Loxahatchee River on the east coast of Florida. Loxahatchee River, which discharges ma inly through its Northwest Fork, supplies mainly quartz sand and organic detritus. Clay mineral makes up less than 5% of the mud in the estuary, but because this mud is ri ch in organic matter, its accumulation has become a matter of concern in the central embayment of the estuary. This flow, in addition to controlled discharges from the S-46 structure in the C-18 Canal at the head of

PAGE 18

2 the Southwest Fork, brings in much of the sediment (mean concentration 0.014 kg/m3; Sonnetag and Mcpherson, 1984) in the central embayment. A commonly employed solution to reduce sedi mentation is the implementation of a trap scheme by trenching the submerged bo ttom. Such a trench-trap is a means to increase the depth at the chosen location by dredging. Increased depth results in a decreased flow velocity (and associated bed shear stress), thereby allowing incoming sediment to settle in the trap, instead of be ing carried further downstream. The removal of sediments becomes much easier as it can be then be removed from the trap, rather than dredging the otherwise distributed deposits fr om a considerably broader area. As an alternative to sediment entrapment, creating a self-cleaning channel in the area of concern for sedimentation would mean that sedime nt would pass through the system, without deposition. The degree to which both appro aches can function depends on the flow conditions, type of sediment and the morphology of the estuary. Given the above background, the objectives of this study were: 1) to determine the efficiency of traps installed at selected loca tions in the estuary, and 2) to evaluate the efficiency of channels as a means to pur sue the goal of a self-cleaning sedimentary environment. Shoaling has occurred the Loxahatchee in many areas, especially near the confluences of the major tributaries (Northwe st Fork and Southwest Fork) in the central embayment where the velocities are typi cally low (Sonntag and McPherson, 1984). Recent studies (Jaeger et al., 2002) suggest in ternal recirculation of sediments as an important factor governing sediment transport within the estuarine portion of the river. Accordingly, in order to manage sedimenta tion in the central embayment, it may be

PAGE 19

3 desirable to test trap/channel deployments at multiple locations. The performance of these schemes was evaluated with regard to efficiency of sediment removal. 1.2 Study Tasks The tasks undertaken included: 1. Data collection from the site and scruti ny of data from the existing literature to characterize the nature of flow, sediment transport and sedimentation. This included measuring tidal elevations, current velocities, sediment concentrations and bed sediment distribution (Jaeger et al., 2002) in the estuary, and obtaining stream flow data for the tributaries from the literature. 2. Simulating the flow field using a hydrodyna mic model, in order to determine the velocities, water surface elevations and bed shear stress distributions. 3. Introduction of trap schemes in the ca librated flow model to determine flow velocities with and without the trap, and development of relationships for calculating trap efficiency. 4. Introduction of self-cleaning channels and an assessment of their viability. 5. A qualitative assessment of the usefulness of the approaches based on selected criteria. 1.3 Outline of Chapters Chapter 2 describes the sediment management alternatives including existing conditions and the proposals for implementati on. Chapter 3 deals with the field data collection for this study including data an alysis and interpretation. Flow model calibration and validation is included in Chapter 4 and evaluation of management alternatives is described in Chapter 5. Summa ry of the results and conclusions are made in Chapter 6, followed by a bibliography of studies cited.

PAGE 20

4 CHAPTER 2 SEDIMENT MANAGEMENT ALTERNATIVES 2.1 Present Condition of the Loxahatchee Estuary Loxahatchee River empties to the Atlantic Ocean through the Jupiter Inlet located in northern Palm Beach County on the south co ast of Florida, about 28 km south of St. Lucie Inlet and 20 km north of Lake Worth Inle t. The three main tributaries, which feed the estuary, are the Northwest Fork, the Nort h Fork, and the Southwest Fork. In addition, the Jones Creek and Sims Creek, which are far le sser tributaries than the others, also feed the estuary through the Southwest Fork. Figur es 2.1 shows the general location map of the study area. The major surface flow in to the estuar y historically was through the Northwest Fork draining the Loxahatchee Marsh and Hungry land slough (refer Fig 2.1). The upstream reach of the Southwest Fork, refe rred to as the C-18 canal, was created in 1957/58 in the natural drainage path in order to lengthen the area of influence of the Southwest Fork and facilitate drainage of the westward swampland (Refer Figure 2.1 and Figure 2.2). The flow in the canal is regulated by the S-46 automated sluice gate structure. Whereas, the Southwest and th e Northwest fork converge on the estuary approximately 4 km west of the inlet, the North fork joins the central bay about 3 km west of the inlet. Down stream of the Fl orida East coast Railroad (FECRR) Bridge the Intracoastal Waterway (ICWW) intersects the estuary in a dogleg fashion. Five navigation/access channels exist on the south shore of the central embayment

PAGE 21

5 Figure 2.1 Location map of the study area (Source: U.S. Geological Survey report no.844157, 1984) Figure 2.2 Loxahatchee River estuary and tributaries A detailed hydrographic survey of the cen tral embayment (Figure 2.3) and the Northwest and Southwest Forks carried out in NovemberÂ’ 2001 (Lidberg Land Surveying, Inc) indicates the depths in th e estuary, which range between 0 m (reference to North Atlantic Vertical Datum 1988, (NAVD 88)) near the sandy shoals to almost 6 m

PAGE 22

6 in the entrance channel near the FECRR Bridge. The average depth over the embayment is just over 1 m. The navigation channel (maintained by the Jupiter Inlet District) runs westward from the Inlet, under the FECRR bridge, and through the central embayment approximately 14 km upstream from the In let. The navigation Channel has a bottom width of about 30.5m (100 feet) and is ma intained at 1.75m(5.74feet) (reference to National Geodetic Vertical Datum 1929, (NGVD 29) and – 2.21m (7.24feet) with reference to NAVD 88) with a side slope of 1:3. Flood shoals, which approximately bisects the central embayment exists mainly due to the sand influx from the ocean, and smaller shoals exist at the termini of the th ree main tributaries. Small shoal islands are located west of the FECRR bridge, on both sides of the channel. The Northwest Fork and North Fork are natural tributaries draining in to the central embayment. However, as mentioned the Southwest Fork was lengthened westward by construction of C-18 canal with a control structur e (S-46), in order to divert flow from the Northwest Fork to the Southwest Fork. A channel was then constructed allowing the diversion of flow from the Northwest Fork to the Southwest Fork. For easy reference from this point on, the C-18 can al will be indicated as the narrow channel section and the broader section at the root will be called Southwest Fork (Figure 2.2).

PAGE 23

7 Figure 2.3 Hydrographic survey of the estuary (November 2001) The Loxahatchee River estuary drains over 1000 km2 of land through the three main tributaries, the ICWW, and several mi nor tributaries. The individual watershed basins are shown in Figure 2.4 and liste d in Table 2.1. The watershed constitutes residential areas, agricultural lands, and uninhabited marsh and slough areas. Table 2.1 Basin area distributions in the Loxahatchee River estuary watershed Basin Area (km2) Intracoastal Water way 545 C – 18 Canal 278 Jonathan Dickinson 155 South Indian River 65 Loxahatchee Rive 6

PAGE 24

8 Figure 2.4 Loxahatchee River estuary watershed basins, estuarine limits (dashed arcs) and central embayment Unlike more northerly estuaries, upland dr ainage in to the Loxahatchee provides only quartz sand and organic detritus. Clay mi neral makes up less than 5% of the mud in the estuary (McPherson, 1984). Earlier studies indicate that the estuary was periodically open and closed to the sea due to various r easons. Originally, flow from the Loxahatchee River along with that from Lake Worth Cr eek and Jupiter Sound kept the inlet clean. With the construction of the ICWW and the Lake Worth inlet and the modifications of the St.Lucie Inlet in 1970, some flow was dive rted. Subsequently, Jupiter Inlet generally remained closed until 1947, except when it is dredged periodically. After 1947, it was

PAGE 25

9 maintained open by dredging by the Jupiter Inlet District and the U.S. Army Corps of Engineers. Dredge and fill operations have also been carried out in the estuary embayment and forks. In the early 1900Â’s, there was signi ficant amount of filling at the present FECRR Bridge, which narrowed the estuary from 370 m to 310 m. The areas east and west of the bridge (and also under the bridge) were dr edged in mid-1930Â’s, and also in 1942. The material was high is shell content and was used in construction of roads. In 1976-77, additional estimated 23,000 m3 materials were removed from the estuary at the bridge and from an area extending 180 m from the west Some dredging was also carried out in the Southwest Fork near the C-18 canal in the early 1970Â’s (Wanless, Rossinsky and McPherson, 1984). In 1980, three channels were dug in the embayment, and an estimated 23,000 m3 of sediment were removed. After 1900, the estuary was greatly influen ced by the dredging and alteration of the drainage to the basin. With gradual lowering of the water table and resultant effect on the water quantity, the direction and pattern of inflow (McPherson and Sabanskas, 1980) were considerably affected. Historically, the major surface flow to the estuary was in to the Northwest Fork from the Loxahatch ee Marsh and the Hungry-land Slough (Figure 2.1), both of which drained north. A small agricultural canal was dug before 1928 to divert a small amount of water from the L oxahatchee Marsh to the Southwest Fork. As noted, in 1957-58, C-18 canal was constructed al ong the natural drainage way to divert flow from the Northwest Fork to the Southwest Fork of the estuary. Jaeger et al., (2001) carried out extensive studies in the estuary to reevaluate the nature of environmental sedimentology in th e lower Loxahatchee River Estuary and as a

PAGE 26

10 companion study to Ganju et al. (2001). Speci fically, new samples were collected in order to 1) examine changes in surfic ial sediment types between 1990 and 2000, 2) attempt to determine the sources of fine-grained muddy sediments accumulating within the estuary; and 3) examine rates of sedimentation within the central embayment and three forks (North, Northwest, and Southwes t/C-18 canal) by collecting a suite of ~1-m long pushcores and ~3 m long vibracores within the estuary. Grab samples were collected in all regions of the estuary and were analy zed. One of the main findings of the study was the internal movement of the sediments in the estuary system. With the growth of the population on the shoreline and associated human activities the mangroves dotting the shoreline started vanishing. The removal of these Mangrove cover from the shoreline released a large quantity of sediments, which was otherwise trapped in their roots. Essentially fine grained, these sediments moved with the flow and started getting deposited in the estuarine bounds. According, to this study new shoals were developed/grown by this process, especially the submerged one in the Northwest Fork, down stream of the shoal identifiable from a satellite map and Figure 4 of the Report (Jaeger et al., 2001). The aerial photograph re produced in Figure 2.5 also indicates an additional shoal developing from the root of the existing shoal, suggesting that the general nature sediments being fed by the Northw est Fork is coarse grained with the fine grained ones carried downstream with the current before deposition. Tidal flow into and out of the estuary is much larger than freshwater inflow from all the major tributaries. Fresh water flow is reported to be about 2 percent of the total tidal inflow (Sonnetag and McPherson, 1984). Ti des are mixed semidiurnal (twice daily with varying amplitude) with a tidal range of about 0.6 to 0.9 m. Tidal waves advances

PAGE 27

11 up the estuary at a rate of 2.23 m/s to 4.46 m/s (McPherson and Sonnetag, 1984) and shows little change in the tidal amplitudes ove r to about 16 km river km. Winds have a significant effect on the tidal ranges especia lly the strong northeast winds which prevails during autumn and winter for example can push in additional water into the estuary affecting the tidal ranges. N ew Shoal Figure 2.5 Ariel photograph showing development of new shoal Estuarine conditions extends in the estuary from the inlet for about 8 river km into Southwest Fork, 9.6 river km in to the North fork and 16 river km into the Northwest Fork. Of late, the environmental condition of th e Loxahatchee River and the estuary has become a matter of great concern. The major f actor affecting the environmental health is the sediment transported in to the estuary. Large amount of the sediments settling in the

PAGE 28

12 basin might affect the bottom life, alter circulation patterns, and accumulate shoals, thereby impeding boat traffic (McPherson, Wanless and Rossinsky, 1984). 2.2 C-18 Canal 2.2.1 Present Condition The C-18 canal drains the Loxahatchee Slough, a shallow swamp-like feature containing diverse flora and fauna. However, estuarine conditions persist for 8 km up the Southwest Fork/C-18 canal measured from the inlet. Flow data obtained from USGS stream fl ow gage data, for all available years (1971-2002 N.W. Fork, 1980-1982 N. Fork, a nd 1959-2002 S.W. Fork) indicate that C18 canal/Southwest Fork carries a maximum discharge of 61.54 m3/sec. Cumulative frequency distribution curves were constructe d to designate (Figure 2.6 a-c) median and extreme flow events (Table 2.2) for all the tr ibutaries. The C-18 canal is regulated at S-46 structure, which is basically a gated sluice. The criterion for controlling the flow at the S46 structure is based on water level behind the structure. When the level exceeds a predetermined mark, the sluice gates are ope ned until the level recedes by 30 cm (Russell and McPherson, 1984), at which point the gates are closed. This regulation has resulted in a discontinuous flow record; w ith weeks of no flow passing the structure, and days when storm flows have been released. During normal wet season, the level behind the S-46 structure is not always sufficiently high for re leasing flow, while the other tributaries are freely discharging to the estuary.

PAGE 29

13 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.105101520253035404550556065707580 Flow rate (m3/s)Cummulative frequency distribution 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 00.20.40.60.811.21.41.61.822.2 Flow rate ( m3/s ) Cumulative frequency distribution a b

PAGE 30

14 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 05101520253035404550556065 Flow rate ( m3 / s ) Cumulative frequency distributionc Figure 2.6.Cumulative discharge plot. a) Nort hwest Fork. b) North Fork. c) Southwest Fork. Table 2.2 Statistical tributary flow (based on Figures 2.6 a-c) Tributary Median Flow (50%) (m3/s) High Flow (90%) (m3/s) Maximum Flow (98%) (m3/s) Northwest Fork 0.7 4.1 76 North Fork 0.1 0.21 1.9 Southwest Fork 1.3 7.8 61 Sonnetag and McPherson (1984) reported two values of suspended solid sediment concentration (0.059 kg/m3, 0.017 kg/m3) with corresponding flow data for the C-18 canal (31 m3/s, 28 m3/s, respectively) and a mean concentration value for duration (198082) of their study (0.014 kg/m3). The median flow for the C-18 canal (1.3 m3/s) from the Figure 2.6c was correlated to this mean value of concentration in the present study. A fit in the form of (Mller and Fstner, 1968) s b sCaQ (2.1)

PAGE 31

15 was used (Ganju et al., 2001), where s a and s b are site specific coefficients, with s ais indicative of the erodibility of the upstream banks/bed and exponent s bis indicative of the intensity of the erosional forces in the river. Table 2.3 Median and high flow concentrati on data and coefficients for equation 2.1 Tributary Median flow Concentration (kg/m3) High flow Concentration (kg/m3) s a Coefficient s b Coefficient Northwest Fork 0.011 0.023 0.012 0.27 North Fork 0.01 0.018 0.018 0.02 Southwest Fork 0.014 0.059 0.012 0.49 Fieldwork, consisting of bottom profiling and sampling was carried out during July 2001 (Jaeger et al.,) by collecting a suite of ~1-m long push cores and ~3 m long vibracores within the estuary. A total of 110 samples were collected from sampling locations covering the entire estuary and ri ver (Figure 1, Final Report on Sedimentary processes in the Loxahatchee River Estuary, 5000 Years ago to the Present, Jaeger et al., 2001) including from outcrops of regional su rficial geological unit (undifferentiated 1.8 million year-old Pleistocene sediments) in order to examine the potential sediment sources (Loxahatchee River, C-18 canal, In let, and Pleistocene-Age (last 1.8 million years) sediments exposed along the banks of the C-18 canal. 56 of the samples were reoccupations of sites sampled in 1990 and were reported (Mehta et al.,1992). These new samples were collected to examine the change s in sediment characteristics pattern over a 10-year period. Positions of all sampling sites were determined by differential GPS providing a position accuracy of ~1 m. Each grab sample recovered approximately 1 000-2 000 cm3 of sediment, removing approximate ly the uppermost 1-5 cm of the sediment surface. Sediment distribution maps produced from these grab samples indicate

PAGE 32

16 particle sizes reveal that the majority of the estuary is dominated (by weight) of fine, well-sorted sand in the ~150 micron (3 phi; 0.15 mm) size range (Jaeger et al., 2001) In the same study conducted by Jaeger et al., (2001), poling depths (obtained by pushing a graduated pole into bottom until a hard substrate is reached) in the C-18 canal were determined to estimate sedimentation rates along the length of the canal. Since the bottom was dredged at the time of construc tion of the canal in 1957/58, the bed thickness can be considered to represent the subs equent accumulation. This is because, the dredging of the canal in 1958 would have mo st likely left behind a hard, sand rich horizon that could not be easily penetrated with the solid rod. Figure 2.7 shows these thicknesses along the canal length. Sediment th ickness increases with the distance from the S-46 structure, possibly due to the large erosional forces near the structure (when flow is released), and reduction of these forces as the flow moves along the canal, allowing more deposition of sediment. Figure 2.7 Dredging Plans for C-18 canal, 1956 (Source: Ganju et al., 2001) This coarse sand layer was sampled at th e base of push cores (see Figures 21 and 22 from Jaeger et al., 2001). There appears to be a trend of increasing thickness away

PAGE 33

17 from the S-46 control structur e (Figure 19). Modeling of sediment transport in the canal (Ganju et al., 2001) also supports such a trend. The overall sedimentation rates (10-50 mm/yr) in the canal are very high for most coastal areas, where sedimentation has kept pace with the rise in sea level (3-5 mm/ yr) (Davis, 1994). However, this sampling technique of poling only provides mean se dimentation rates over this 42-year (19582001) time period. Analyses of push cores coll ected in the canal document alternating layers of clean sand and muddy sand/sandy m ud (see Figures 21-22, Jaeger et al., 2001). This inter-layering of sediment types is characteristic of time-varying deposition rates/erosion rates. When the sluice ga tes are opened, fast currents can erode the sediment surface followed by rapid deposition of sand and mud. The best way to evaluate time-varying sedimentation rates is with eith er accurate annual bathymetric profiles or by measuring naturally occurring radioisotopes in the sediment cores (Jaeger et al., 2001). 2.2.2 Management Options Dredging plans for the C-18 canal from 1956 is shown in Figure 2.7 (U.S. Army Corps of Engineers, 1956). The existing bottom wa s deepened to 3 m at some locations to facilitate drainage. The depths refer to the National Geodetic Vertical datum of 1929 (NGVD). The present mean depth of the canal as measured along the length is 1.2 m. Hence there has been substantial sedimentation in the canal, which in turn means that it no longer serves as a sediment trap and allows sediment to be transported to the central embayment. One way of maintaining the dept h in the canal is to devise a suitable dredging option coupled with a designed flow regi me in order to maintain the canal in the self-cleaning mode. However one of the main difficulties in this is the lack of continuous supply of water. As described earlier, the flow in the canal is erratic and controlled by the S-46 control structure. Accordingly, although the median flow in this canal is higher than

PAGE 34

18 the other tributaries, the flow is episodi c and therefore not enough to overcome the bed shear resistance of the deposited sediments. This situation can be illustrated by data collected during between April 4th and April 24th, 2002. Figure 2.8 indicates the dependence of th e current velocity on the released discharge. The sudden jump in the over all current magnitude recorded downstream of the structure therefore exhibit strong erosiona l trend as can be seen from the Figure 2.9. In addition, it indicates that, sediment concentrations in the bottom layers are much more pronounced due to the obvious reason of erosion of the bed. It can therefore be concluded that, a sustained and regular flow regime w ould help keeping the canal sediment free. An option is to increase the depth in the can al by dredging part or all of it, thereby recreating the sediment trap. As an alternativ e, a detailed study of the flow pattern can be undertaken and a suitable flow regime worked out. This would involve redesigning of the control structure and a better regulation of the flow. However, the following points should be noted: 1. The capacity of flow from the structure appears to be insufficient to flush out sediment beyond 1.2 km (Ganju et al., 2001) from the structure even under “high” discharges when the gates are open. 2. A potential option is to change the gate configuration but not the flow regulation schedule. If changing the gate configura tion from sluice to weir is successful, it would create a sediment trap upstream of the gate, which would “buy time” for the downstream reach of the canal, but this upstream trap would eventually have to be dredged to maintain it effectiveness. The volume of material trapped will be restricted the weir height. Over-depth dredging upstream is a viable option. 3. However, because sediment transported across the gate is believed to be quite heterogeneous (ranging from fine sand to clay and organic matter) and the organic material is presently not found in the bed there, predictive modeling the transport of sediment across the gate will be an un certain exercise without extensive data collection on both sides of the gate. An option would be to carry out gate conversion and work with the new system based on a rough estimation of the new flow/sediment regime. It is likely that some modification of gate opening schedule may also have to be carried out to improve the efficiency of the upstream trap.

PAGE 35

19 Figure 2.8 Current variation under the effect of released discharge from the S-46 structure. Figure 2.9 Effect of S-46 discharge on the suspended sediment concentration The present study envisages examining the option dredging the downstream canal. Ganju (2001) carried out such an exercise by testing the effectiveness of a comparatively

PAGE 36

20 short sediment trap. The trap design and re sults of the investigation are summarized below.In order to quantify the sedimentation rate as a function of discharge in the C-18 canal investigations were carried out usi ng calibrated sediment transport models. The boundary conditions were designed to simulate the episodic unsynchronized (with Northwest and North Fork discharges) discha rges from the S-46 structure. The results indicated that as discharge increases the ch ange in the rate of sedimentation rate decreases. However, they do not share a dir ect straight-line relationship. For instance, doubling of flow from 2.5 m3/s to 5 m3/s results an increase of 71% in the sedimentation rate and similar increase from 10 to 20 m3/s changes the rate only by 25% indicating that, the sedimentation rate is more sensitive to lower discharges. This is evidently due to increasing discharge is associated with increas ed concentration. The regulation of the C18 canal by the S-46 structure is manifested in the high frequency of zero-discharge periods (54% of the days) and the spikes The deposition rates were found to be 0.15 m for a period of 10 years, which compared well with the poling results. The study also compares the sedimentation in a regulated C-18 canal to that of hypothetically unregulated canal by applying flow record for the Northwest fork for the same period pro-rated so that the discha rges over the 10 year flow period remains identical. Resulting in a 10-year deposition thickness of 0.22m (0.022m/yr), implying that the episodic discharges in actuality reduced the rate of sedimentation. This is a direct consequence of near constant high disc harge attenuating the increasing trend of sedimentation. The study incorporates a trap near the ar ea of greatest post dredging thickness, with a poling depth of approximately 1.2 m. A dre dging depth of 3 m (from the original bed

PAGE 37

21 level) width of 60 m, and a length of 180 m were chosen for the trap, which was considered sufficient to reduce the velocity in the canal, and allow a measurable amount of sediments to settle. This trap configura tion reduced the current magnitude by 67% over the trap. As a consequence a number of factors were evaluated by the study namely, Simulations showed that the removal ratio, i.e., the ratio of sediment influx (into the trap) minus out flux divided by influx) was maximum at an S-46 discharge of approximately 1.7 m3/s. At higher discharges sediment was transported beyond the trap, while at lower discharges sediment settled before the trap. The second simulation involved testing the trap efficiency as a function of sediment concentration. It was observed that increase in sediment concentrations in the free settling range in general increases the settlement. The increase in trapped load followed a linear trend up to concentrations of 0.25 kg/m3 (free settling zone), which is explained by the increase of deposition flux with concentration (with constant settling velocity). Above th is concentration, and below 7 kg/m3 (flocculation range), the increase in settling velocity yields a similarly increasing trend for trapped load. In the hindered settling zone, however, (which lies above this concentration) trapped load decreases as the settling velocity deceases. It was therefore be inferred that trapped load is a function of concentration because settling velocity (and hence the deposition flux) is also a function of concentration at values greater than 0.25 kg/m3. The simulations on varying organic conten t indicated that, increase in organic content led to decease in settling velocity, which resulted in lower removal ratio. Sedimentation rate in the trap increased with increased organic content, due to corresponding decease in dry density. In additi on, the increase in influent load with increasing organic content as less sediment was deposited upstream of the trap at higher organic content. 2.3 Central Embayment 2.3.1 Present Condition Jupiter Inlet, which is about 112 m wide and 3.9 m deep at the jetties, allows the tidal flow in and out of the estuary. The channel starting at the jetties leading up to the Florida East Coast Railroad Bridge is fairly uniform, with width varying from 206 m to 247 m and the mean depth varying between 3.92 m at the inlet and 2.6 m near the FECRR Bridge. The ICWW meets the channe l down stream of the FECRR bridge.

PAGE 38

22 Upstream of the FECRR Bridge the embayment widens and the channel is divided in to two parts by shoals often exposed under low water conditions. These shoals presumably created by the sands introduced in to the syst em through the inlet and the tributaries, and carried by the flood tide, occur where the sedi ment carrying capacity of the flow reduces with the reduction of current at wider sections. In addition, east of these sandy shoals there occurs a small mangrove island. Similar Islands occur near the north bank close to the FECCR Bridge. The deepest portion of the embayment lies to the north of the sandy shoal, easily identifiable even from a areal photograph (Figure 2.9) is currently used for navigation. The shoreline is basically sandy with little or no clay present. The percentage of clay and silt is barely 5%. The average depth in the central embayment is 1.2 m. The depth in the deeper portions along the flood channel however exceeds 3 m in patches. A similar deep channel can be found along the south bank, which has been presumably created by the ebb circulation. A clear ebb channel can also be seen from the satellite photographs to the south of the sandy shoal. Boats returning to their docks use this channel at high water. There are many private wooden docks along the entire coastline. At the turn of the century, the Loxa hatchee River estuaries along with its immediate environ was a pristine ecosystem consisting of mangroves, salt marshes, and scrubland. Prior to Word War II agricultural interests transformed the area in to a rural landscape with citrus groves and vegetable farm s. As a result, a significant increase in residential population occurred around this time. These developments ultimately prompted the declaration of the estuary an aquatic preserve in 1984. Nonetheless, the construction activities, especially of the residential homes still continue along the shoreline and the entire estuarial shoreline of the central embayment as well as a

PAGE 39

23 significant portion of the tributary shorelines is residentially occupied. Recreational boating is widely practiced in the estuary by the local residents. Access is necessary to the upstream areas for recreational activities, and also to the open sea and the ICWW. Many of the natural and artificial access routes have shoaled in recent years (Antonini et al., 1998), leading to hazardous boating practices such as high-speed entry/ exist to prevent grounding of vessels. The channels ad jacent to the south shore of the central embayment are more susceptible to shoa ling (Sonntag and McPherson, 1984), directly affecting the boaters who rely on these channels for access Figure 2.10 Arial Photograph showing the Central Embayment, the Inlet and the Tributaries Estimates with regard to grain size, co mposition and age of bottom sediments are given by McPherson et al. (1984) for the en tire estuary. The samples collected by vibrocore boring were analyzed in the labor atory for micro-faunal and macro-faunal assemblages, grain size distributions, constituent composition and radiocarbon age. With regard to the grain size it was seen that, th e characteristics of the bed material were identical to those of the underlying sediments in the core. Fine-grained sediments dominate the central bar at the lower reaches of the estuary; whereas medium to coarsegrained sand dominates upper reaches of the bar. Patches of fine to medium sand draping

PAGE 40

24 the muddy sediment surface can be seen in the main body of the estuary. The shell content in the bed material varies from 0 to 5% at the eastern end to 20 to 30% at the western end. Grain-size analysis reveals that there ar e two distinct different populations. The first, well-sorted sediment with a mode between 62.5 to 125 microns, and the second, poorly sorted sediment commonly showing bimodality. The bimodal distributions generally have one mode at about 300 microns and the other at 100 microns. Jaeger et al, (2001) measured the particle sizes in the estuary, which, reveal that the majority of the estuary is dominated (by we ight) of fine, well-sorted sand in the ~150 micron (3 phi; 0.15 mm) size range. This size sand is ubiquitous in the estuary and is observed in Pleistocene-age coastal deposits exposed in outcrops within the study area. The ultimate source of the sand accumulating with in the upper estuary is from erosion of these older deposits. The amount of mud-sized sediment (<63 microns) is minimal with the exception of the upper reach of the Nort hwest Fork, the North Fork, and near the junction of the C-18 canal and the Southwest Fork. Clay mineral analyses on the mud fraction accumulating throughout the estuary reveal s that the ultimate source of the claysized sediment is from erosion of the Pleistocene-age deposits. Comparison of the sediment characteristics (median particle-size, sorting) between 1990 and 2000 within the Central Embayment re veal that this region has not changed significantly over the past decade. Howeve r, the navigation channels have become coarser apparently due to the removal of fine sediment. Portions of the lower Northwest Fork and the Southwest Fork have gotten finer.

PAGE 41

25 Based on the analyses of 20 push cores, th ere does not appear to be a widespread organic-rich flocculent “muck” layer with in the three major forks of the estuary. Although mud is a common component of the se diments in these locations, by weight it usually represents less than 20% of the total core mass. In addition, study by Jaeger et al., (2001) indicate that in the main navigation channels, the sediments have become coarser and more poorly sorted over the last ten years. The study attributes this to the likel y inclusion of shelly material in the 2000 samples that was not sampled in 1990. It is possible that maintenance dredging during this time period resulted in the exposure of olde r shelly material or that changes in the shape of the navigation channel has led to st ronger currents that have removed the finer sands. Although the western portion of the Cent ral Embayment has seen no change in the median particle diameter, it has gotten marg inally better sorted, and could reflect a decrease in fine sediments accumulating. Freshwater runoff enters the Loxahatchee River estuary by river and canal discharges, by storm drains, and by overland s ubsurface inflow. Most of the freshwater from the tributaries is discharged from the No rthwest Fork of the estuary. These flows, as expected, vary seasonally, occurring chiefl y in the wet season. The median, high and maximum flow discharges are given in Table 2.2. Tidal flow into and out of the estuary is much larger than the freshwater inflow from all the major tributaries. The combined freshwater flow into the estuary is found to be about 2% or less of the average tidal in flow at the Jupiter inlet (McPherson, Sonnetag, 1984). However, during tropical storm Dennis, freshwater inflow per tidal cycle increased to 18% of average tidal inflow (McPherson, Sonnetag, 1983). Tides are mixed

PAGE 42

26 semi-diurnal with varying amp litudes, with a tidal range of approximately 0.6 to 1 m. The tidal wave advances to the estuary at a rate of about 2.3 m/s to 4.5 m/s. Higher than usual tides can be noted during the autumn and winter when strong northeast winds pushes additional water in to the estuary causing higher than average tides. Ultrasonic water level gauges (Model 220, Infinities USA, Daytona Beach, FL) with stilling walls were installed to meas ure tidal elevations between September 14th and October 18th, at three locations in the estuary one each in the Central embayment (tied to the FECRR bridge pier), Northwest Fork, a nd Southwest Fork. The gauge locations are shown in Figure 2.10. Tidal elevations were recorded with respect to North Atlantic Vertical Datum 1988 (NAVD 88) and are reprodu ced in Table 2.3. Tidal ranges indicate the total change in water surface elevati on between low and high tides and phase lag refers to the difference in time between high/low tide at UFG1 gauge and the other gauges. In Figure 2.12 sample records from three tidal gauge locations are shown. Figure 2.11 Location of tide gauges marked UFG1, UFG2 and UFG3

PAGE 43

27 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 05101520253035404550 Time (hour)Elevation from MSL (m) UFG1 UFG2 UFG3 Figure 2.12 Sample records of tidal measurem ents at three locations (09/14/00-09/15/00)Datum NAVD 88. Table 2.4 Spring/neap tidal ranges and phase lags for three gauges Gauge ID Spring range (m) Neap range (m) Phase lag from UFG1 (min) UFG1 0.90 0.66 0 0 UFG2 0.85 0.65 21 60 UFG3 0.86 0.64 28 60 Ganju et al., (2001) compared the data obt ained from these gauges to a station on the Northeast Florida coast and inferred that trends in water surface elevation followed similar increases and decreases in mid-tide elev ations and the increased elevations in side the estuary is a direct result of onshore winds. The wind records from two offshore stations were averaged and correlated with the mid-tide elevation, resulting in a positive correlation. Accordingly, mid-tide elevation was subtracted from the measured elevations (filtering) in order to obtain tidal data without any variation. 2.3.2 Management Option: Presently, the dredged spoil from th e embayment is disposed on land. Land disposal of marine sediment is often times not optimal for the environment, especially for

PAGE 44

28 the ground water. According to earlier studi es by Sonnetag and McPherson (1984) the central embayment receives sediment from two main sources, the inlet and upland discharge. Regular maintenance of the naviga tion channel is a clear indication of this supply. Ideally, a large enough central shoal (i f developed to correct contours) could serve the process of self-cleansing of the bay. The shoal when developed would decrease the water flow area and thereby, increasing the velocity of flow. The increased current in the limiting case would develop erosional stresse s equal to the critical bed shear of the sediment and therefore would be able to pr event the further sedimentation of the bay. However, numerical modeling for such an exam ination is outside the scope of this study. The present study will however deal with the development of an additional navigation/flow channels for improvement of ebb flow. 2.4 Northwest Fork: 2.4.1 Present Condition: The Northwest Fork meanders through typi cal South Florida swampland within the Jonathan Dickinson State Park (JDSP). The extensive swampland and scrubland east of JDSP is drained by the North Fork. It is therefore evident that the watershed is biologically productive, and the sediment carried by the runoff is rich in organic content eventually finds its way in to the estuary (Sonnetag and McPherson, 1984). Most of the freshwater from is discha rged through this fork. From February 1st, 1980, to the September 30th, 1981, for example, 77.3 percent of the freshwater was discharged into the Northwest Fork, 20.5 percen t in to the Southwest Fork (C-18 Canal), and 2.2 percent into the North Fork (Sonnetag and McPherson, 1984). The Loxahatchee River (i.e., Northwest Fork) at SR-706, site 23 as shown in Figure 2.14 (Figure 2, U.S.

PAGE 45

29 Geological report no 83-4244, 1984), contributed the greatest percentage of flow to the estuary (37.4 percent) of all the tributaries. Vertical variation of the sediments in the Northwest fork is Found at site 5 and 5E (Figure 2, U.S. Geological report no 84-4157, 1984) during both incoming and outgoing tides. Presumably, greater water velocities, pa rticularly at 0.6 m above the bottom at the mid depth, associated with higher tide stages c ontributed to the greater vertical variation of suspended sediments (Sonnetag and McPherson, 1984). Concentration of the suspended sediments and the percentage of se diments of organic origin were variable with season and weather conditions as indicat ed by the data collected and listed in U.S Geological Survey report 84-4157 (Sonneta g and McPherson, 1984). The greatest increases were observed in Cypress Cree k, lying upstream of the Northwest Fork. Concentration of the suspended sediment in the tributaries also changed as a result of manÂ’s upstream activities. During September 1981, suspended sediment concentration in the Cypress Creek and Hobe Grove Ditc h increased as much as 21 times over concentrations in early September (S onnetag and McPherson, 1984). Cleaning and dredging operations on the irrigation canal c onnected to the Cypress Creek and Hobe Grove Ditch were presumably responsible Suspended sediment load from the tribut aries are highly seasonal and storm related. The 5 major tributaries to the Loxahatchee estuary Loxahatchee River at SR-706, Cypress Creek, Kitching Creek, Hobe Grove Ditch, and C-18 at S-46 discharged 1,904 tons of suspended sediments to the estuary during the 20-month period (February 1, 1980 to September 30, 1981) (Table 2.3). During th e 61 days period of the above-average rainfall (August 1 to September 30, 1981) that included tropical storm Dennis, the major

PAGE 46

30 tributaries discharged 926 tons of suspended sediment to the estuary. This accounted for 49 percent of the suspended sediment disc harged to the estuary during the 20-month period and about 74 percent of the suspe nded sediment discharged during 1981 water year (Sonnetag and McPherson, 1984). Sediment loads from C-18, Loxahatchee River at SR-706, and Cypress Creek accounted for more than 94 percent of the total tributary input of the sediment load. Figure 2.13 Location of stream-gauging stations and sampling site for suspended sediments, (Source U.S. Geological report no 83-4244 and 84-4157) Unlike the central embayment concentrati on of mud was quite high (~50%) in the Northwest Fork (Jaeger et al., 2001). The st udy by Jaeger et al., (2001) also analyses vibracores takes which, reveal that there ha s been roughly 0.5-1 cm/yr of sedimentation within a part of the Northwest Fork when compared to data from a USGS-sponsored

PAGE 47

31 study completed, in 1984 (Sonnetag and McPher son). The study further concludes that, these accumulation rates are close to those aver aged over the past 50 years, assuming that an observed change in the cores from layere d sediment not mixed by organisms to those that are well mixed by organisms occurred in 1947 when the inlet was stabilized. Inlet stabilization would have led to increased tidal flushing that allowed for better oxygenation of bottom waters and sediment s permitting occupation of sediments by organisms. However, this datum has not been substantiated as pre 1947 and the accumulation rates are bulk averages. A compar ison of the collected data and studies by Ganju et al., (2001) showed that accumulation rates within the upper reaches of the three Forks are about 2-3 times higher than the modeled fine-sediment budget prepared by Ganju et al. (2001). Accordingly, the study conc ludes that, this discrepancy could be due to poor age constraints of the core layers or to the substantial presence of sand in the core sections, which was measured in this stratigra phic (i.e., core layering) approach but not in the fine-sediment budget. Upstream of the outfall point of the No rthwest Fork is marked by a horseshoeshaped shoal (Figure 2.14). Presumably this shoal is formed due to the reduction in current velocity of the sediment-laden flow by the ebb tide. In addition, the ebb flow velocity gets reduced upon meeting a large body of water (central embayment). Upstream of this shoal there occur a series of sand shoals also formed by the same processes. Downstream of the shoal however, the depths are uniform gradually increasing as moves in to the central embayment area. Formation of deposits presumably from the erosion of old deposits in side the estuary was also re ported by Jaeger et al., (2001). Figure no 4 of

PAGE 48

32 the report are reproduced here for reference with regard to the deposition and material composition. In Figure 2.15 the mass percent of the mud (particles smaller than 63 microns) in the upper ~5 cm of the sediment surface is shown. Location of the sampling sites are shown as dot symbols. 2.4.2 Management Options Discounting the sedimentation from the inte rnal sources of erosion, the Northwest Fork contributes the maximum discharge as well as the maximum sediment into the central embayment (Sonnetag and McPhers on, 1984). However, Jaeger et al, (2001) indicate that fresh deposits are found in the Fo rk (Figure 2.14), suggesting that the source of such deposits may be mostly internal to the estuary, and most likely due to the erosion of old deposits. Sediment from external s ources entering the estuary with fresh water discharge as reported by McPherson (1984) woul d have deposited in the proximity of the horseshoe shoal. In order to minimize the deposition of fine sediment in the area of high mud percentage in the Northwest Fork (Fi gure 2.14), a self-cleaning channel will be examined. According to Jaeger et al., (2001), th e origin of deposits (Figure 2.14) is due to the erosion of old deposits. Therefore the ch annel is proposed to be located downstream of these deposits. Design aspects of the channel are considered in Chapter 4.

PAGE 49

33 Figure 2.14 Location indicating fresh mud depositi ons and the Shoals the estuary. Source: Sedimentary Processes in the Loxahatchee River Estuary: 5000 Years Ago to the Present-FINAL REPORT, Jaeger et al., (2001) 2.5 North Fork 2.5.1 Present Condition The North Fork is a natural tributary dr aining the eastern part of the Jonathan Dickson State Park. Discharge as given in Table 2.2 is the least of the three main tributaries (2.2% of total), and water depth is fairly uniform at around 2 m, with virtually no shoals. McPherson and Sonnetag (1983) re ported that in the 1981 water year the tributaries of the North Fork were dry at the gauging stations (Figure 2.13) from March trough mid-August. During the rest of the year the average flow was 0.12 m3/s, a very small value. Discharge following Tropical Stor m Dennis was also small for the amount of rainfall associated with the storm. Daily di scharges for the last 10 days of August 1981 averaged 0.31 m3/s but increased to 0.71m3/s in September. Jaeger et al. (2001) found Clay deposits

PAGE 50

34 some mud deposits in the upper reaches. Depths in the fork appear to be adequate for the recreational boating. 2.5.2 Management Options The North Fork as indicated above has the least river inflow as well as the least sediment contribution to the estuary. In addition, the depths are fairly uniform and good for the types of boats presently using it. Hence no additional facility is believed to be required for this area. Therefore no dredging is planned for this tributary nor appears to be required.

PAGE 51

35 CHAPTER 3 DATA COLLECTION 3.1 Field Setup in the Southwest Fork Field data were collected at two sites, one in the Southwest Fork and the other in the Northwest Fork. Section 3.3 collection effo rt and results in the Southwest Fork, and Section 3.4 in the Northwest Fork. The field data collection set up in th e Southwest Fork of the estuary had geographical coordinates of latitude 26o 56' 36.78" N and longitude 80o 07' 17.34" W. In the Northwest Fork the corresponding coordinates were 26o 59' 16.78" N and longitude 80o 07' 56.34" W. These two locations are show n in Figure 3.1. The locations of the tidal gages installed in the year 2000 were shown in Figure 2.11. The depth (below North Atlantic Vertical Datum, 1988, (NAVD88)) at the sites ware 2.1m and 2.18, respectively. Instrument Stations Figure 3.1 Location of instrument tower in the Southwest and Northwest Forks

PAGE 52

36 Data in the Southwest Fork were collected in two phases. The first phase of the data collection was carried out between 4th and 24th April 2002, and the second phase was between 6th of February and 2nd of June 2003. The instrumentation deployed is given in Table 3.1. Table 3.1 Instrumentation for data collection and data blocks Instrument Data Date Data logger (*) Current (mag.) – u Apr 04 to Apr 24 Nov 27 to Jun 2 Data logger (*) Current (dir.) – u Apr 04 to Apr 24 Nov 27 to Jun 2 Data logger (*) Current (mag.) v No data Nov 27 to Jun 2 Data logger (*) Current (dir.) v No data Nov 27 to Jun 2 Data logger (*) Tide levels Apr 04 to Apr 24 Nov 27 to Jun 2 Data logger (*) OBS 1 Apr 04 to Apr 24 Nov 27 to Jun 2 (Poor quality) Data logger (*)) OBS 2 Apr 04 to Apr 24 Nov 27 to Jun 2 Data logger (*) OBS 3 Apr 04 to Apr 24 Nov 27 to Jun 2 Data logger (*) OBS 4 Apr 04 to Apr 24 No data collected Data logger (*) Temperature Apr 04 to Apr 24 Nov 27 to Jun 2 Data logger (*) Salinity Apr 04 to Apr 24 Nov 27 to Jun 2 *With ultrasonic current meter in April 2002 replaced with an electromagnetic current meter Instruments were attached to a tower erected for this purpose and was powered by rechargeable batteries. The instrument assembly consisted of a Marsh-McBirney electromagnetic current meter, a Transmetrics pressure transducer for the measurement of water surface elevation, a Vitel VEC-200 conductivity/temperature sensor for measurement of salinity and temperature, and three Sea point Optical Backscatter Sensor (OBS) turbidity meters for measuring the sedime nt concentration at 3 different levels. In the first phase instrument setup, however, turb idity sensors were deployed at 4 different levels. In addition Lidberg Land Surveys, Inc. carried out a hydrographic survey and collected data with regard to the bottom ba thymetry of the central embayment and the tributaries.

PAGE 53

37 In the Northwest Fork the data collection started on 14th of August 2003, with 3 level OBS sensors, one Conductivity Temperature sensor and one Pressure gauge 3.2 Instruments Deployed 3.2.1 Current Current data were collected using Marsh-McBirney electromagnetic current meter (Model 585 OEM). This meter consists of a 10 cm diameter spherical sensor, OEM motherboard, and signal processing electronics (Figure 3.5). The instrument senses water flow in a plane normal to the longitudinal axis of the electromagnetic sensor. Flow information is output as analog voltage corresponding to the water velocity components along the y-axis and x-axis of the electromagnetic sensor. The velocity sensor works on the Faraday principle of electromagnetic i nduction. The conductor (water) moving in the magnetic field (generated from within th e flow probe) produces a voltage that is proportional to the velocity of water. The Marsh-McBirney requires periodic cleaning of the probe with mild soap and water to keep the electrodes free of non-conductive material. Since the instrument has essentially a cosine response in the horizontal plane, the flow magnitude and the direction informati on are retained. In addition, the spherical electromagnetic sensor has an excellent vertical cosine response. This unique characteristic allows the sensor to successfu lly reject vertical current components that may be caused by mooring line motions. As th e flow changes direction, the polarities of the output signal also change. So the u (veloc ity along axis of the channel flow) and the v (velocity across the channel) velocities are stored and can be combined to give the resultant magnitude and direction. It must however be noted that, the v velocity component was largely insignificant due to the width of the channel at the tower location.

PAGE 54

38 3.2.2 Tide Water surface elevation was measured usi ng a Transmetrics pressure transducer installed at the instrument tower. The instrument incorporates three major design elements that allow it to measure pressure accurately and reliably; bonded foil strain gages configured in a Wheatstone bridge (for temperature stability), high precision integral electronics for signal amplification, and stainless steel construction for durability and corrosion resistance. The instrument was calibrated and temperature-compensated against standards applicable for the region. 3.2.3 Salinity/Temperature Conductivity is the measurement of the ab ility of a solution to carry an electric current. It is defined as the inverse of th e resistance (ohms) per unit square, and is measured in the units of Siemens/meter or micro-Siemens/centimeter. The measurement of conductivity is necessary for the determina tion of the salinity of a solution. Salinity is proportional to the conductivity and is expressed in terms of concentr ation of salt per unit volume (mg/l, or ppt). The field measurement of salinity was carried out following similar procedures using a Greenspan Electrical Conductivity (EC) sensor substantially eliminating a basic source of error arising out of the inaccuracies due to temperature and electrode effects. In this instrument the el ectrical conductivity is a function of the number of ions present and their mobility. The electri cal conductivity of a liquid changes at a rate of approximately 2% per degree Centigrade for neutral salt and is due to the ionic mobility being temperature dependent. The te mperature coefficient of the conductance (or K factor) varies for salts and can be in the range 0.5 to 3.0. As electrical conductivity is a function of both salt concentration and te mperature, it is preferable to normalize the

PAGE 55

39 conductivity measurement to a specific reference temperature (250C) so as to separate conductivity changes due to salt concentration from those due to temperature changes. The instrument deployed consisted of the following primary elements: Toroidal sensing head (conductivity sensor) Temperature sensor Microprocessor controlled signal conditioning and output device The conductivity sensor uses an electro magnetic field for measuring conductivity. The plastic head contains two ferrite core s configured as transformers within an encapsulated open-ended tube. One ferrite core is excited with a sinusoidal voltage and the corresponding secondary core senses an en ergized voltage when a conductive path is coupled with primary voltage. An increase in charged ion mobility or concentration causes a decrease in the resistivity and a corresponding increase in the output of the sensor. A separate PT100 temperature sensor inde pendently monitors the temperature of the sample solution. This sensor provides both a temperature output and a signal to normalize the conductivity output. 3.2.4 Sediment Concentration The instrument deployed was a Sea Point turb idity meter. This instrument measures turbidity by scattered light from suspended particles in water. The turbidity meter senses scattered light from a small volume within 5 centimeters of the sensor window. The light sources are side-by-side 880 nm Light Em itting Diodes (LED). Light from the LED shines through the clear epoxy emitter window into the sensing volume, where it gets scattered by particles. Scattered light between angles 15 and 150 degrees can pass through the detector window and reach the det ector. The amount of scattered light that

PAGE 56

40 reaches the detector is proportional to the turb idity or particle concentration in the water over a very large range. The sensors were calibrated using a sample from the measurement site. Periodic calibrations were conducted in order to eval uate the conditions of the windows and the sensitivity to scattering. In addition, only black containers were used in calibration so as to prevent any probable scattering events due to reflection off the container wall. The calibration was carried out using known volum e of sediments in known volume of water and the voltage output of the instrument r ecorded. A linear fit curve was generated in order to determine the accuracy of the calibration. The calibration plots are given below, Figure 3.2 Calibration plots used for calibration of OBS sensors

PAGE 57

41 3.3 Field Data Results in Southwest Fork 3.3.1 Current The electromagnetic current meter was lo cated at a height of 96.5 cm from the bed level. The velocity data in two direc tions, one parallel to the flow and the other perpendicular to it, were combined vectorially to find the resultant magnitude and direction. The ultrasonic current meter de ployed in April 2002 collected the current magnitude and direction directly. Based on these data the depth-mean magnitude time series for Julian days 94-114 is shown in Figure 3.3 and the corresponding direction plot is given as Figure 3.4. A sudden increase in the current magnitude in the plot is attributable to the opening of control struct ure S-46. The directional plot indicates a unidirectional flow driven by the discharge from the structure. The discharge record for the period is given in Table 3.2 for ready reference. Table 3.2 Discharge data for the period 04/14/2002 to 04/21/200 Date Julian Days of 2002 Discharge (m3/s) 04.14.2002 104 0.03171 04.15.2002 105 0.00821 04.16.2002 106 0.01416 04.17.2002 107 0.01501 04.18.2002 108 0.03483 04.19.2002 109 0.05777 04.20.2002 110 0.03568 04.21.2002 111 0.00934

PAGE 58

42 Figure 3.3 Record of current magnitude: Days 94-114 (year 2002). Figure 3.4 Record of current direction: Days 94-114 (year 2002).

PAGE 59

43 Figure 3.5 Record of current magnitude: Days 332-356 (year 2002). Figure 3.6 Record of current direction: Days 332356 (year 2002). Figure 3.5 is a representative plot of th e current magnitude for the second data block. This plot indicates a more uniform velo city pattern driven by the tidal flow in the

PAGE 60

44 estuary. The current magnitudes reach a maximum value of 0.17 m/s with the mean value at 0.06 m/s. In addition it is seen that the flow is predominantly along the estuary with very low values observed for transver se current (v). In Table 3.3 typical mean current values are summarized. Table 3.3 Typical mean current magnitude values for data blocks Current magnitude (m/s) Julian days in 2002 With S-46 discharge Only tidal flow Velocity u (m/s) Velocity v (m/s) 94 114 0.25 0.04 aa332 356 a0.06 0.057 0.018 a No data 3.3.2 Tidal Level Figure 3.7 Water level time-series: All levels relative to NAVD 88. Days 94-114 (2002).

PAGE 61

45 Figure 3.8 Water level time series. Upper plot shows original time series with mean trend and the lower plot is without the mean oscillations. All levels relative to NAVD 88. Days 94-114 (2002). Figure 3.9 Water level time-series. All levels relative to NAVD88. Days 332365 (2002) and Days 01-35 (2003).

PAGE 62

46 Figure 3.10 Water level time series. Upper plot shows original time series with mean trend and the lower plot is without this trend. All level relative to NAVD 88. Days 332-365 (2002) and Days 01 – 35 (2003). In Figure 3.7 the raw tidal time-series is shown for the period April 4th to April 24th, 2002. In Figure 3.8 the upper plot shows the origin al time series with the tidal trend and the lower plot is with the tidal trend removed. The tidal plots indicated in the Figur e 3.7 to Figure 3.10 are representative plots from the phase II and I. The characteristic valu es of the tidal data are given in Table 3.4. In addition it can be noted that the tidal ra nges compares well in both the phases with the spring range equal to 1.0m and the neap ra nge around 0.5m. As will be explained later, the tidal fluctuations (as could be noted from Figure 3.9) between Julian days 360 to 365 in Year 2002, 01 to 5 and 17 to 25 in Year 2003, is likely to affect the sediment concentration in the estuary.

PAGE 63

47 Table 3.4 Characteristic values of the tidal data Water level/Tidal range Water level (m) Spring/neap range (m) Julian days in 2002/03 Mean water depth (m) MaximumMinimum Spring Neap 94-114 1.20 1.70 0.30 0.90 0.50 332-365 01 35 1.20 1.90 0.30 1.00 0.50 3.3.3 Total Suspended Solids Total Suspended Solid (TSS) was recorded at four elevations in the first phase and three elevations in the second phase. The elev ation of the OBSs relative to the bed level was OBS-4 = 1.17 m, OBS-3 = 0.80 m, OB S-2 = 0.48 m and OBS-1 = 0.22 m. The corresponding total suspended solid time series are reported in Figures 3.11 and 3.12 for days 94-114 (Phase I) and 352365 in 2002 and 01 to 35 in year 2003 (Phase II), respectively. Table 3.5 provides the maximum, mean and minimum values of sediment concentrations at different levels for each da ta block. Depth-mean concentration averaged every 12 hours is presented in Figures 3.13 and 3.14. The mean concentration figures (Figure 3.13 and 3.14) indicate the average variations in the concentration over time with out the instantaneous variations (spikes).

PAGE 64

48 Figure 3.11 TSS time-series at four elevations: Days 94-114 (year 2002). Figure 3.12 TSS time-series at three elevations: Days 352365 (year 2002) and 01–35 (year 2003).

PAGE 65

49 Figure 3.13 Depth-mean TSS concentration time series: Days 94-114 (year 2002) Figure 3.14 Depth mean TSS concentration time series: Days 352365 (year 2002) and Days 01 – 35 (year 2003).

PAGE 66

50 Figure 3.15 Depth mean TSS concentration time series and tidal trend indicating their dependence: Days 352365 (year 2002) and Days 01 – 35 (year 2003). It can be noted from Figures 3.11 and 3.13 that there is a sudden increase in sediment concentration with the di scharge from the S-46 structure on 14th of April 2002 (Refer Table 2.2 for discharge details). This cl early indicates that sediment concentration is discharge driven. Results of Figures 3.12 and 3.14 indicate that the lowest OBS1 sensor was too close to the bed and recorded almost saturated sediment content. There was no discharge from the structure between December 14th and February 20th, except for 0.01 m3/s discharge on the December 20th, 2002, which explains the increase in sediment concentration recorded around Julian day 355 (December 20th). However, the increase in TSS reported between days 17 and 27 (Year 2003) without any discharge from S-46, could be attributed to spring tidal effects (Ref er to Figure 3.15). In general, it appears that

PAGE 67

51 TSS concentration is dependent on the local tidal current and flow discharges down the S-46 structure. The TSS concentrations with regard to other data blocks are given in Table 3.7 to 3.9. Table 3.5 TSS concentrations for the representative data blocks Julian Days in 2002/03 Maximum TSS concentration (mg/L) Mean TSS concentration (mg/L) Minimum TSS concentration (mg/L) 94-114 165 50 10 352-365 01-35 158 17 7 3.3.4 Salinity and Temperature The conductivity and temperature measurements carried out for the location is presented in Figures 3.16 (days 94 –114 of 2002). The salinity curve indicates the effect of the fresh water discharge. Due to this flow fresh water from the S-46 structure the salinity values dropped to 11 mg/L from a m ean value of about 28 mg/L. In order to examine this hypothecation the current magnitude and the salinity was plotted together in Figure 3.17, which, indicated a decrease in sa linity with an increase in the current magnitude. Accordingly, it can be concluded th at the fresh water discharge reduces the salinity in the estuary. Table 3.6 Characteristic salinity values Julian days in 2002/03 Maximum Salinity (mg/L) Mean Salinity (mg/L) Minimum Salinity (mg/L) 94-114 34.2 24.9 11.1 352-365 01-35 39.5 36.5 26.7

PAGE 68

52 Figure 3.16 Salinity time series: Days 94-114 (year 2002). Figure 3.17 Salinity and Current magnitude time series: Days 94-114 (year 2002).

PAGE 69

53 Figure 3.18 Temperature time series: Days 94-114 (year 2002). Similarly the temperature time-series shows a positive correlation with the discharge, with temperature increasing with the discharge from S-46 structure. However any definite conclusion could not be deduced from this the absence of adequate data on temperature of the freshwater discharged. For the second data block between days 352 and 365 (of year 2002) and days 01 and 35 (of year 2003) the Figure 3.19 indicates an apparent malfunctioning of the sensor that seems to have contaminated the conductiv ity time series that calculates the salinity by measuring its conductivity of the soluti on at a given temperature. Although the temperature time series for the same period appears to give correct reading consistent with the environment, the incorrect c onductivity data have made the salinity determination inaccurate. Therefore salinity valu es reported in this period appear to be rather high. Tables 3.5 and 3.6 summarize the characteristic values of salinity and

PAGE 70

54 temperature for both the data blocks. The results from the other data blocks are furnished in Table 3.7 to 3.9. Figure 3.19 Salinity time series: Days 352365 (year 2002) and 01-35 (year 2003). Figure 3.20 Temperature time series: 352365 (year 2002) and 01-35 (year 2003). Table 3.7 Characteristic temperature values Julian days in 2002 Maximum Temperature (‘0’ C) Mean Temp (‘0’ C) Minimum Temperature (‘0’ C) 94-114 31.4 26.3 21.6 352-400 26.7 10.9 7.8

PAGE 71

55 3.3.5 Other Data Blocks The foregoing discussions included the vari ous aspects of data collection their analysis and results for two representative data blocks (Julian Days 94 to 114, 330-365 in year 2002 and 01 to 35 in year 2003). However since the second phase data collection lasted from November 26th, 2002 to May 15th, 2003, it was considered necessary to include the characteristic values obtained from the other data blocks, which would offer a better insight in to the overall site conditions. Table 3.8 Summary of parametric value (Days 37-59 in year 2003) Parameter Maximum Mean Minimum Depth (m) 1.9 1.2 0.5 OBS 1 (mg/L) * OBS 2 (mg/L) 240 30 0.7 OBS 3 (mg/L) 110 20 1.0 Salinity (mg/L) 40 35 23 Temperature (0C) 27 16 11 Current Magnitude (m/s) * * Poor quality data In Table 3.7 a summary of parametric va lues of the data collected between February 6th and February 28th is presented. The data obtained for the other two blocks are presented in Tables 3.8 and 3.9. Table 3.9 Summary of parametric value (Days 90-101 in year 2003) Parameter Maximum Mean Minimum Depth (m) 1.6 1.1 0.6 OBS 1 (mg/L) 2670 1710 1470 OBS 2 (mg/L) 80 50 1.0 OBS 3 (mg/L) 96 51 20 Salinity (mg/L) * Temperature (0C) 20 11 3 Current Magnitude (m/s) * *-Bad Data

PAGE 72

56 Table 3.10 Summary of parametric value (Days 101-135 in year 2003) Parameter Maximum Mean Minimum Depth (m) 1.9 1.3 0.7 OBS 1 (mg/L) 2090 1670 1180 OBS 2 (mg/L) 170 46 2 OBS 3 (mg/L) 210 56 10 Salinity (mg/L) 26 19* 17* Temperature (0C) 22 12 4 Current Magnitude (m/s) 1.40 0.60 0.10 Bad Data 3.4 Field Data Results in Northwest Fork 3.4.1 Field Setup In the third phase of data collection, in the Northwest Fork, the instrument tower included three optical backscatter sensors (OBS), a pressure transducer (for water level) and a conductivity/temperature sensor. Data collection began on 08/14/2003. Data on water level and TSS are presented. The c onductivity/temperature sensor malfunctions during this phase and yielded values of ques tionable accuracy. Hence these data are not reported. 3.4.2 Tidal Level The pressure transducer was located 0.45 m from the bed. Figure 3.21 shows the original time series of the water level.

PAGE 73

57 Figure 3.21 Record of water level variation. Days 245 – 255 (year 2003). Figure 3.22 Water level time series. Upper plot shows original time series with mean trend and the lower plot is without the mean oscillations. All levels relative to NAVD 88. Days 245 -255 (year 2003). In Figure 3.22 the upper plot shows 12-hourly mean trend with the original time series, and in the lower plot this trend is removed. As can be seen from the latter plot, the

PAGE 74

58 rising mean trend indicates the effect of fr esh water discharge. The tidal range was 0.80 m. Characteristic values are given in Table 3.10. Table 3.11 Characteristic values of the tidal data Water Level/Tidal range Water level (m) Spring/neap range (m) Julian days in 2003 Mean Water depth (m) MaximumMinimum Spring Neap 245-255 1.20 1.75 0.70 0.80 0.50 3.4.3 Total Suspended Solids Total suspended solids (TSS) concentration was recorded at three elevations. The elevations of the OBS sensors relative to the bed were OBS-1 = 1.04 m, OBS-2 = 0.66 m and OBS-3 = 0.30 m. The corresponding depthmean concentration time series is reported in Figure 3.23. Characteristic values are given in Table 3.11. Figure 3.23 Depth-mean TSS concentration time-series: Days 245-255 (year 2003). Table 3.12 TSS concentrations for the representative data blocks Julian Days in 2003 Maximum TSS concentration (mg/L) Mean TSS concentration (mg/L) Minimum TSS concentration (mg/L) 245-255 230 100 50

PAGE 75

59 3.4.5 Additional Data Blocks 3.4.5.1 Tidal Level Two additional data blocks were collected between November 6th, 2003 and November 24th, 2003. Tide data for Julian days 310 and 313 are presented here. The remainder was found to be of poor quality. Figure 3.24 Record of water level variation. Days 310.5 – 313.5 (year 2003).

PAGE 76

60 Figure 3.25 Water level time series. Upper plot shows original time series with mean trend and the lower plot is without the mean oscillations. All levels relative to NAVD 88. Days 310.5 – 313.5 (Year 2003). Table 3.13 Characteristic values of the tidal data Water Level/Tidal range Water level (m) Spring/neap range (m) Julian days in 2003 Mean Water depth (m) Maximum MinimumSpring Neap 310.5-313.5 1.40 1.90 1.00 0.90 0.50 3.4.5.2 Total Suspended Solids Two data blocks for the TSS concentration was collected and are presented below. Characteristic values are presented in Table 3.14.

PAGE 77

61 Figure 3.26 TSS time-series at two elevations: Days 310.5 – 313.5 (year 2003). Figure 3.27 Depth mean TSS concentration time series: Days 310.5 – 313.5 (year 2003).

PAGE 78

62 Figure 3.28 TSS time-series at three elevations: Days 315.5 – 318.5 (year 2003). Figure 3.29 Depth-mean TSS concentration time series: Days 315.5 – 318.5 (year 2003). Table 3.14 TSS concentrations for the representative data blocks Julian Days in 2003 Maximum TSS concentration (mg/L) Mean TSS concentration (mg/L) Minimum TSS concentration (mg/L) 310.5 – 313.5 834 304 19 315.5 – 318.5 219 140 81

PAGE 79

63 CHAPTER 4 MODEL CALIBRATION AND VALIDATION The analyzed data presented in Chapter 3 give a qualitative insight into the prevailing environmental conditions. However, in order to have a quantitative understanding of the flow regime in the estu ary, it is necessary to apply a numerical simulation technique. This chapter includes a brief description of the numerical model, generation of the computational grid, initial and boundary conditions and the model operational scheme. Model calibration and validation are then carried out. Certain aspects of the estuary have been idealized in the formulation of the model in order to reduce the computational time and avoidance of potential errors. These idealizations are as follows: 1. The central embayment domain is termin ated at the FECRR bridge excluding the ICWW (Intracoastal Waterway). This enable s use of tide data from UFG1 gage installed at the bridge. 2. The traps and the navigation channels have rectangular cross-sections. 4.1 Model Description Flow simulations were carried out usi ng Environmental Fluid Dynamics Code (EFDC) maintained by the Environmental Protection Agency, and developed by Hamrick, 1992. This code works through a Microsoft Windows-based EDFC-Explorer preand post-processor. Developed on a Fo rtran platform, the physics of EFDC and many aspects of the computational scheme are equivalent to the widely used BlumbergMellor model (Blumberg and Mellor, 1987) a nd the U.S. Army Corps of EngineersÂ’ Chesapeake Bay model (Johnson, et al, 1993). EFDC solves the three-dimensional

PAGE 80

64 hydrostatic, free surface, turbulent averaged e quations of motion of a variable density fluid. The model uses a stretched or sigm a vertical coordinate and Cartesian or curvilinear, orthogonal horizontal coordinates. Dynamically coupled transport equations for turbulent kinetic energy, turbulent length scale, salinity and temperature are also solved. Externally specified bottom friction can be incorporated in the turbulence closure model as a source term. For the simulation of flow in vegetated environments, EFDC incorporates both two and three-dimensi onal vegetation resistance formulations (Moustafa, and Hamrick 1995). The numerical scheme employed in EDFC to solve the equations of motion uses second-order-accurate spatial finite differen ce on a staggeredor a C-grid. The modelÂ’s time integration employs a second-order-accurate, three-time-level, finite-difference scheme with an internal-external mode splitti ng procedure to separate the internal shear or baroclinic mode from the external free surface gravity wave or barotropic mode. The external mode solution is semi-implicit, and simultaneously computes the twodimensional surface elevation field by the pr econditioned conjugate gradient procedure. The external solution is completed by the cal culation of the depth averaged barotropic velocities using the new surface elevation field. The modelsÂ’ semi-implicit external solution allows large time steps that are cons trained by the stability criteria of the explicit central difference or upwind advection sc heme used for the nonlinear accelerations. Horizontal boundary conditions for the extern al mode solution include the option for simultaneously specifying the surface elevations the characteristic of an incoming wave, free radiation of an outgoing wave or the volumetric flux on arbitrary portions of the boundary. The modelÂ’s internal momentum equation solution, at the same time step as

PAGE 81

65 the external, is implicit with respect to ve rtical diffusion. The internal solution of the momentum equations in terms of the vertical profile of shear stress and velocity shear, which results in the simplest and most accurate form of baroclinic pressure gradients, and eliminates the over-determined character of alternate internal mode formulations. The model implements a second order accurate in space and time, mass conservation fractional step solution scheme for the Eulerian transport equation at the same time step or twice the time step of the momentum equation solution. The advective portion of the transport solution uses either the central difference scheme used in the Blumberg-Mellor model or hierarchy of positiv e definite upwind difference schemes. The highest accuracy up-wind scheme, second order accurate in space and time, is based on a flux corrected transport version of SmolarkiewiczÂ’s multidimensional positive definite advection transport algorithm, which is m onotonic and minimizes numerical diffusion. The EFDC model's hydrodynamic component is based on the three-dimensional hydrostatic equations formulated in curv ilinear-orthogonal horizontal coordinates and a sigma or stretched vertical coordinate. The momentum equations are: tmxmyHu xmyHuu ymxHvu zmxmywu femxmyHv myHxp patm myxzb zxHzp zmxmyAvH zu xmymx HAHxu ymxmy HAHyu mxmycpDpu2 v21/2u (4.1) tmxmyHv xmyHuv ymxHvv zmxmywv femxmyHu mxHyp patm mxyzb zyHzp zmxmyAvH zv xmymx HAHxv ymxmy HAHyv mxmycpDpu2 v21/2v (4.2)

PAGE 82

66 mxmy f e mxmy f u ymx v xmy (4.3)1(,)(,)xzyzvzAHuv (4.4) where u and v are the horizontal velocity components in the dimensionless curvilinear-orthogonal horizontal coordinates x and y, respectively. The scale factors of the horizontal coordinates are mx and my. The vertical velocity in the stretched vertical coordinate z is w. The physical vertical coordinate s of the free surface and bottom bed are zs* and zb* respectively. The total water column depth is H, and is the free surface potential which is equal to gzs*. The effective Coriolis acceleration fe incorporates the curvature acceleration terms, with the Coriolis parameter, f, according to (4.3). The Q terms in (4.1) and (4.2) represent optional horizontal momentum diffusion terms. The vertical turbulent viscosity Av relates the shear stresses to the vertical shear of the horizontal velocity components by (4.4). The kinematic atmospheric pressure, referenced to water density, is patm, while the excess hydrostatic pressure in the water column is given by: zp gHb gH o o 1 (4.5) where and o are the actual and reference water densities and b is the buoyancy. The horizontal turbulent stress on the last lines of (4.1) and (4.2), with AH being the horizontal turbulent viscosity, are typically retained when the advective acceleration are represented by central differences. The last te rms in (4.1) and (4.2) represent vegetation resistance where cp is a resistance coefficient and Dp is the dimensionless projected vegetation area normal to the flow per unit horizontal area. The three-dimensional continuity e quation in the stretched vertical and curvilinear-orthogonal horizontal coordinate system is:

PAGE 83

67 tmxmyHxmyHuymxHv zmxmyw QH (4.6) with QH representing volume sources and sinks in cluding rainfall, evaporation, infiltration and lateral inflows and outflows having negligible momentum fluxes. The solution of the momentum equations, (4.1) and (4.2) requires the specification of the vertical turbulent viscosity, Av, and diffusivity, Kv. To provide the vertical turbulent viscosity and diffusivity, the second mome nt turbulence closure model developed by Mellor and Yamada (1982) (MY mode l) and modified by Galperin et al (1988) and Blumberg et al. (1988) is used. The MY model relates the vertical turbulent viscosity and diffusivity to the turbulent intensity, q, a turbulent length scale, l, and a turbulent intensity and length scaled based Richardson number, Rq, by: Av Aq l A Ao1 R1 1Rq1 R2 1Rq1 R3 1Rq Ao A11 3 C1 6 A1B1 1 B1 1/3 R1 1 3 A2B2 3 A21 6 A1B1 3 C1B2 6 A11 3 C1 6 A1B1 R2 1 9 A1A2R3 1 3 A26 A1 B2 (4.7) K v Kq l K Ko1 R3 1Rq Ko A216A1B1 (4.8)Rq gH zb q2 l2H2 (4.9)

PAGE 84

68 where the so-called stability functions, A and K, account for reduced and enhanced vertical mixing or transport in stable and unstable vertically density stratified environments, respectively. Mellor and Yamada (1982) specify the constants A1, B1, C1, A2, and B2 as 0.92, 16.6, 0.08, 0.74, and 10.1, respectively. For stable stratification, Galperin et al. (1988) suggest limiting the length scale such that the square root of Rq is less than 0.52. When horizontal turbulent viscosity and diffusivity are included in the momentum and transport equations, they are determined independently using Smagorinsky's (1963) sub-grid scale closure formulation. At the bed, the stress components are presum ed to be related to the near bed or bottom layer velocity components by the quadratic resistance formulation 22 1111(,)(,),xzyzbxbybcuvuv (4.10) where the 1 subscript denotes bottom layer values. Under the assumption that the near bottom velocity profile is logarithmic at any instant of time, the bottom stress coefficient is given by cbln( 1/2 zo) 2 (4.11) where is the von Karman constant, 1 is the dimensionless thickness of the bottom layer, and zo=zo*/H is the dimensionless roughness height. Vertical boundary conditions for the turbulent kinetic energy and length scale equations are: 23 2 1:1sqBz (4.12) 23 2 1:1bqBz (4.13) l 0:z 0,1 (4.14)

PAGE 85

69 where the absolute values indicate the magnitude of the enclosed vector quantity which are wind stress and bottom stress, respectively. 4.3 Grid Generation The first step in the setup of the modeling system is to define the horizontal plane domain of the region being modeled. The hor izontal plane domain is approximated by a set of discrete quadrilateral and triangular ce lls. Developed on a digitized shoreline, the grid defines the precise locations of the faces of the quadrilateral cells in the horizontal as well as in the vertical plane. However, all th e computations are carried out at the center of the cells. Since the model solves the hydrodyna mic equations in a horizontal coordinate system that is curvilinear and orthogonal, gr id lines also correspond to lines having a constant value of one of the horizontal coor dinates. The shoreline as well as the cell reference is provided by a local set of Coordi nates in MKS unit, as the code uses MKS system internally. Seven identification numbers were used to define the cell types. The cell identification details are given in Table 4.1. Table 4.1 Definition of cell type used in the model input Cell ID Definition of cell type 0 Dry land cell not bordering a water cell on a side or corner of the model 1 Triangular cell with land to the northeast of the model 2 Triangular cell with land to the southeast of the model 3 Triangular cell with land to the southwest of the model 4 Triangular cell with land to the northwest of the model 5 Quadrilateral water cells of the model 9 Dry land cell bordering a water cell on a side or on a corner of the model The type 9 dry land or fictitious dry land cell type is used in the specification of no flow boundary conditions. The horizontal geometric and topographic (bottom bathymetry) and other related characteristics of the region, files dxdy.inp and lxly.inp are used. The program then directly reads these quantities expressed in meters. The lxly.inp

PAGE 86

70 provides cell center coordinates and components of a rotation matrix. Cell center coordinates are used only in graphics output and can be specified in the most convenient units for graphical display such as decimal de grees, feet, miles, meters or kilometers. The rotation matrix is used to convert pseudo east and north (curvilinear x and y ) horizontal velocities (uand v respectively) to true east and north for graphics vector plotting, according to; tecuecveco tncuncvncouCCu vCCv 4.15 where the subscripts teand tndenote true east and true north, while the subscripts codenotes the curvilinear-orthogonal horizontal velocity components. The coefficient C is the multiplier term for conversion to true east and true north. The width of the C-18 canal, which vari es between 75 m at the Southwest Fork junction to less than 40m at the S-46 structure, dictated the dimensions of the cells. It was decided that a 25 x 25m cell would be accurate enough for representing the width of the C-18 canal resulting in desired level of accuracy. The same cell size was then conveniently extended to the rest of the model domain. The bottom bathymetry was based on the Hydrographic survey carried out in November ‘2001 by Lidberg Land Surveying, Inc. However additional data for areas not covered under this survey were obtained from other available surveys. The roughness coefficient of the bottom bathymetry in the model is composed of two components. A fixed component viscosity (for the present model fixed at 0.020m) and a variable component, which is varied uniformly on the entire model domain during calibration process, both the component together constitutes the factor 0z, defined in equation 4.11. The dimensionless thickness

PAGE 87

71 of the bottom layer 1, defined in the same equation, e quals to 0.25, since four vertical layers are used. The fixed component of the roughness factor, how ever can be increased/decreased in the areas of vegetation or other special features. The details of sea grass locations in the central embayment can be referred from Drawing no LOX-001 (Cuthcher & Associates, Inc. Coastal E ngineer, 2002) provided by the Jupiter Inlet District. The sea grass was input in the model as an overlay file. In this way the cells having the sea grasses are enclosed by a poly line so that, the roughness coefficient can be easily edited. The sea grass was represented as cells having more roughness (fixed component = 0.040m) than that of the surr oundings. In Figure 4.1 the input bathymetry and the shoreline as generated by the model are shown. Figure 4.1 Model domain showing input bathymetry and shoreline

PAGE 88

72 -6.836-.305Bottom ElevTime: 275.00 Figure 4.2 Computational grid showing the flow boundaries In the computational grid (Figure 4.2), each land cell was assigned number zero or nine as the case may be and each water cell was assigned five. There were no triangular cells used for this grid. Figure 4.2, in additi on, indicates the locations of the tide gages and the Instrument tower in the Southwest fo rk. The S-46 structure in the C-18 canal is a flow boundary (black cells), as are the two ma in tributaries, and the FECRR bridge on the East. The eastern boundary was restricted to the FECRR bridge. The flow boundaries were kept straight; so as to allow flows pe rpendicular to the cell faces, as the model does not allow non-orthogonal flows. 4.4 Boundary Conditions In the beginning of the simulation, ve locities throughout the model domain are considered to be zero. It was observed that a full tidal cycle was required before the water surface elevation reached a quasi-steady stat e. This was verified by recording water InstrumentStation

PAGE 89

73 surface elevations at the location of the two tide gauges (UFG2 and UFG3) over multiple tidal periods. Tidal forcing at the FECRR bridge ( eastern boundary) is perhaps the most important boundary condition in this system, because it is this mechanism by which the majority of the water flows through the estu ary. The data obtained from the UFG1 gage (Figure 2.10) were used to simulate this fo rcing. The raw data were examined for the mean trends in the water surface elevation (F igure 4.3). The raw data contains a sub-tidal frequency trend, which was also noticed in th e water surface elevation data of the Miami Harbor. The trends were of a similar in nature and therefore it was hypothesized that onshore winds may have created increased elev ation in side the estuary. The wind records from two offshore sites (37 and 221 kilomete r east of Cape Carnival, Florida) were correlated with the mid-tide elevation, whic h indicated a positive correlation (Ganju et al., 2001). In order to overcome the effect s of these variations imposed on the astronomical tide, the mid-tide elevation was s ubtracted from each measured elevation in the same tidal cycle. The mid tide elevation c is given by Equation 4.1, where, H T and LT are the water surface elevation at high and low tides respectively. 2 H TLT c (4.16)

PAGE 90

74 a b Figure 4.3 Tidal time series from UFG1, 09/1 4/00-10/13/00, a) Raw data, b) Tidal plot after the mid-tide trend is removed. Time origin 12:00 am.

PAGE 91

75 In Figure 4.3a the raw tidal time series is shown along with the tidal trend and in Figure 4.3b the tidal time series is shown af ter subtracting the mean-tide trend. The eastern boundary accordingly used this water surface elevation boundary condition. For the boundary in the C-18 canal, tw o sets of boundary condition data were available. The daily average flow time series of the S-46 structure and the water surface elevation time series. The elevation time seri es was obtained from the tide gauge UFG 3 (same period as at UFG 1) installed in the S outhwest Fork (Figure 2.10). In order to make these data usable at the flow boundary (S -46 Structure) amplitude corrections were carried out by trial and error till both predic ted and measured time series matched. In order to calculate the phase correction (lag) following calculations were carried out assuming shallow water conditions. The tidal wave celerity C is given by, Cgh (4.23) where, gis the acceleration due to gravity and h is the water depth. Then the phase shift T is given by, ; L Cgh T L T C (4.24) where, L is the distance for which the water de pth is considered uniform, accordingly the phase lag for the distance between the UF G 3 gage station and the S-46 structure was calculated and verified (0.13 hour). Figure 4.4 gi ves the plot of the raw data collected at UFG 3, including the mean trend and the amplitude with trends removed. It was hypothesized that these data, corrected for th e phase and amplitude could be applied as boundary condition to simulate actual flow conditions.

PAGE 92

76 b a Figure 4.4 Tidal time series from UFG3, 09/ 14/00-10/13/00, a) Raw data, b) Tidal plot after the mid-tide trend is removed. Time origin 12:00 am Note that the flow discharge time series (Figure 4.5) from the S-46 structure was selected, as the model is known to be givi ng better simulation results under discharge boundary condition.

PAGE 93

77 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 0.00500.001000.0 0 1500.0 0 2000.0 0 2500.0 0 3000.0 0 3500.0 0 4000.0 0 4500.0 0 DAYS DISCHARGE (m3/s) Figure 4.5 Flow time series applied at S-46 boundary 0 10 20 30 40 50 60 70 010002000300040005000DAY S DISCHARGE (m3/s) Figure 4.6 Flow time series applied at Northwest Fork boundary In the Northwest Fork boundary as well, two sets of boundary conditions, namely, the water surface elevation boundary condition ( obtained from transferring the collected data of the tidal station UFG 2) and flow discharge boundary condition were evaluated. The flow time series used is shown in Figure 4.6. Table 4.2 Amplitude and phase correction factor for the tides Boundary Amplitude factor Phase correction C-18 1.14 0.13 hour Northwest Fork 1.18 0.042 hour

PAGE 94

78 Per U.S. Geological Survey Report 84-4157 (Russell and McPherson, 1984) the majority (77.3 %) of the fresh water flow in to the estuary enters through the Northwest Fork. Therefore, the flow discharge boundary condition for this tributary was considered as most appropriate as opposed to the water surface elevation. The corrected water surface elevation data from UFG 2 tide gage was used for calibration. The North Fork carries the least discharge (2.2%) of the total freshwater flow in to the estuary in the mean, (Russell and Mc Pherson, 1984)) hence at this boundary also flow discharge boundary condition is applied. The flow discharge was worked out from the Northwest boundary data applying a constant multiplier ( 2.2 0.0285 77.3 ). 4.5 Model Calibration and Validation 4.5.1 Calibration In general calibration of the model aims at simulating conditions identical or close to that in the prototype so that prototype conditions can be accurately replicated and reproduced. Calibration involves matching multiple parameters, which is often times, is practically impossible. However, depending on the nature experiments and the results desired, the type of calibration differs. Since the present model simulation aims to relate the velocity and the associated stress field to the erosion/accretion of the sediments in the estuary, it would be highly desirable to calib rate the model with comparison of the flow velocities. But current data for the model simulation period, between September 14th 2000 and 18th October 2000 was not available and ther efore, it was decided to calibrate the model using the data collected at the inst rument station located in the Southwest Fork (Figure 3.1) between November 26th and May 15th 2003 for which current as well as

PAGE 95

79 water surface elevation data was available. The amplitude multiplier and phase lag factors are given in Table 4.2. Accordingly a simulation for this period was carried out using flow discharge boundary conditions for the Southwest, the No rthwest and North tributary boundaries and water surface elevation boundary condition for the East boundary. For the e astern boundary the tidal data from Miami Harbor were “transferred” to the FECRR bridge boundary by applying suitable correction factors fo r the amplitude and the phase lag. This procedure was carried out in two steps. In th e first step, the Miami harbor data for the period 14th September 2000 to 18th October 2000 were transfe rred to the boundary with application of recommended coefficients (for method of calculation refer to NOS Tide Tables for year 2000). The calculated tidal elevations were compared with the UFG 1 data and the final multiplication correction factor was obtained as 1.023. For the model simulation period in year 2002 the same correc tion factor was used to transfer tidal elevations of Miami harbor to the flow boundary. Model calibration began with an initial run for 48 hours (referred to as ‘cold start’) in order to make the tide and discharge mutu ally compatible throughout. In addition, the flow attains stability in this period. The resu lts of the cold start period were compared with the current velocities as well as the water surface elevations obtained from the i nstrument tower. The process was continued by changing the variable component of the bottom friction coefficient (one component of 0z) (applicable uniformly throughout the model domain), until an approximate match of the current magnitude and phase was obtained. In the second step RESTART.OUT and RSWT.OUT, the two output files of the c old start were used as input, and model r un was performed for a longer period (15days)

PAGE 96

80 in order to obtain simulation for final calibration. The predicted and measured currents were then compared and is given in Figure 4. 7a (Cold start) and 4.7b (Hot start) for a variable bottom friction factor of 0.027. It can be seen that the agreement is very good for the current, with a maximum error of 1.8% of the total current amplitude. The water surface elevation however differs by about 2. 8 cm, which is about 3% of the tidal amplitude. Since current is in better agreemen t with the measured data the calibration was considered accurate enough for simulation. In addition, comparison of the predicted and measured current direction exhibited good agreement as indicated in 4.8. 4.5.2 Model Validation Model validation was carried out using the same calibrated parameters and simulating the flow conditions of year 2000 (between September 14th and October 18th). The measured as well as the model results at bot h the tidal gage stations after cold start as well as hot start periods are compared and reproduced as Figures 4.9 and 4.10. As indicated in the figure 4.10, the agreement is fairly accurate with a maximum variation of 2.7 cm, which is about 3.4% of the maximu m tidal amplitude reported in the estuary. Similar validation was also carried out using the Northwest Fork data collected between September 3rd and September 12th which also showed equally good agreement as shown in Figure 4.11.

PAGE 97

81 Figure 4.7 Model calibration measured vs. predicted current, a) Cold Start, b) Hot Start. -0.10 -0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 331.00331.20331.40331.60331.80332.00332.20332.40332.60332.80333.00JULIAN DAYS IN 2002CURRENT (m/s) Model Run Measured -0.06 -0.05 -0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 0.05 346.00346.20346.40346.60346.80347.00347.20347.40347.60347.80348.00JULIAN DAYS IN YEAR 2002CURRENT (m/s) Model run Measured

PAGE 98

82 -250.00 -200.00 -150.00 -100.00 -50.00 0.00 50.00 100.00 150.00 200.00 333.00333.20333.40333.60333.80334.00334.20334.40334.60334.80335.00 JULIAN DAYS IN YEAR 2002CURRENT DIRECTION (degree Model Measured Figure 4.8 Model calibration measured vs. predicted current direction.

PAGE 99

83 -0.4000 -0.3000 -0.2000 -0.1000 0.0000 0.1000 0.2000 0.3000 0.4000 0.5000 257.5000258.0000258.5000259.0000259.5000260.0000260.5000 JULIAN DAYS IN 2000WATER LEVEL (m ) Model Measured -0.4000 -0.3000 -0.2000 -0.1000 0.0000 0.1000 0.2000 0.3000 0.4000 0.5000 0.6000 273.400 0 273.600 0 273.800 0 274.000 0 274.200 0 274.400 0 274.600 0 274.800 0 275.000 0 275.200 0JULIAN DAYS IN 2000WATER LEVEL (m ) Model Measured a b Figure 4.9 Model calibration measured vs. predicted water surface elevation (UFG2) Year 2000, a) Cold Start, b) Hot start.

PAGE 100

84 -0.4000 -0.3000 -0.2000 -0.1000 0.0000 0.1000 0.2000 0.3000 0.4000 0.5000 257.5000258.0000258.5000259.0000259.5000260.0000260.5000JULIAN DAYS IN 2000WATER LEVEL (m) Model Measured -0.4000 -0.3000 -0.2000 -0.1000 0.0000 0.1000 0.2000 0.3000 0.4000 0.5000 0.6000 273.400 0 273.600 0 273.800 0 274.000 0 274.200 0 274.400 0 274.600 0 274.800 0 275.000 0 275.200 0 JULIAN DAYS IN 2000WATER LEVEL (m) Model Measured a b Figure 4.10 Model calibration measured vs. predicted water surface elevation (UFG3) Year 2000, a) Cold Start. b) Hot start.

PAGE 101

85 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 245.6245.8246246.2246.4246.6246.8247247.2247.4247.6JULIAN DAYS IN 2003WATER LEVEL (m) Measured Model -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 253.00253.50254.00254.50255.00255.50JULIAN DAYS IN YEAR 2003WATER DEPTH (m) Measured Model a b Figure 4.11 Model calibration measured vs. predicted water surface elevation (Northwest Fork) Year 2003, a) Cold Start, b) Hot start.

PAGE 102

86 4.5.3 Simulation of trap scheme of Ganju, 2001 As noted, Ganju et al., (2001) carried out the model simulations by installing a trap in the C-18 canal. The trap was located 480m down stream of the S-46 structure and was 180 m long, 60 m wide and dredged 3 m from the existing bed level. The same scenario was simulated in the present model in order to revalidate data and have a check on the model results. The simulations indicated a 60% reduction in the current magnitude against 67% reported by Ganju, 2001. The results of the simulation (current) are given as Figure 4.12. With this final validation, it was considered that the model is calibrated for the estuary and therefore ready for the actual simulations. -0.01 0.00 0.01 0.01 0.02 0.02 0.03 258.00258.50259.00259.50260.00260.50JULIAN DAYS IN YEAR 2002CURRENT (m/s) Gtrap Existing Figure 4.12 Validation results using trap used by Ganju, 2001.

PAGE 103

87 CHAPTER 5 EVALUATION OF SEDIMENTATION CONTROL ALTERNATIVES 5.1 Design Basis 5.1.1 General Principle Two concepts for management of sediment are adopted with respect to the central embayment, as shown schematically in Figure 5.1. These include sediment entrapment (Figure 5.1a,c) and sediment self-cleaning by channelization (Figure 5.1b,d). 5.1.1.1 Sediment Entrapment For sediment entrapment an area of the submerged bottom is deepened to a depth greater than the surrounding bottom (Figure 5.1a ,c). This works on the simple concept of decreasing the velocity by increasing the flow area. The carrying capacity of the flow being proportional to its velocity, a reduction in the velocity would reduce the carrying capacity and thereby result in sedimentati on. This in turn would allow maintenance dredging to be performed at a specific loca tion (the trap), rather than over a broad submerged area. There is an existing sand trap of this nature at Jupiter Inlet, as shown in Figure 2.9. Trap efficiency is defined as the per cent by which the effluent suspended solid load is reduced with respect to the influent lo ad. In a tidal situation, the seaward edge of the trap will be the influent side during flood flow and the effluent side during ebb, and vice versa for the landward edge.

PAGE 104

88 5.1.1.2 Self-cleaning Channel The concept of a channel designed in such a way that the flow through it is in nonsilting non-scouring (or self-cleaning) equili brium can be employed where it is desired that the sediment flows through without net de position (Figure 5.1b,d). In an ideal setup the trap efficiency of such a channel would be zero. Figure 5.1 Design concepts for sediment management. Uniform channel Dredged trap Shoal Dredged channel Higher flow lesserdeposition Non-uniform channel Bay Bay Low flow deposition Low flow deposition Higher flow lesserdeposition (a) (b) (c) (d)

PAGE 105

89 Trap and the channel are incorporated in the model by changing the bottom profile file (dxdy.inp). Accordingly, the depths and elevati ons of the grid cells in the designated area are changed per design. 5.1.2 Design Alternatives Alternative schemes indicated in Table 5. 1 were formulated and studied for their efficiency and function. Note that Alternatives 6 and 7 were introduced to determine if there was an interactive effect of multiple alternatives implemented simultaneously. However, simulations showed that this was not the case, i.e., there was no measurable impact of an alternative on others. As a re sult, only the first five were investigated further. General locations of Alternatives 2 through 5 are shown in Figure 5.2. Detailed drawings are shown later in the chapter. The basis of selection and design for each alternative (considering the existing “no-action” condition as Alternative no. 1) is described next. Table 5.1 Alternative schemes for evaluation Alternative no. C-18 Canal Trap Bay Channel Bay Y-channel Channel Northwest Fork Channel 1 Existing (“no-action”) condition 2 X 3 X 4 X 5 X 6 X X X 7 X X X

PAGE 106

90 Figure 5.2 Alternatives considered, with existing bathymetry. 5.1.2.1 Alternative No. 2: C-18 Canal Trap The design sediment trap includes the entire stretch of the C-18 canal downstream of the S-46 structure, in order to take care of the drawback of a much shorter trap of Ganju (2001), which was shown to trap only 60% of the sediment under all conditions of S-46 discharge. In addition, the sensitivity to sediment concentration, flow velocity and location were also found to be pronounced in th at analysis. Accordingly, it was decided to dredge the entire length of the canal to -3.5 m (with respect to NAVD 88). The trap was considered to have a width equal to that of the canal. The dredged section is shown in Figure 5.3. Bay Channel C-18 Canal tra p Bay Y-channel N W Fork Channel

PAGE 107

91 Figure 5.3 Dredged bathymetry of the C-18 canal with plan form view of the proposed trap. Trap considered by Ganju (2001) is also shown 5.1.2.2 Alternative No. 3: Bay Channel A self-cleaning channel close to the south bank of the bay (Figure 5.4) was adopted in order to improve the flow in that portion of the bay, which is weaker than in the existing channel. It was considered that such a channel along the ebb flow path would concentrate the flow and thereby flush out incoming fine sediment arriving from the Southwest Fork. In addition, this channel would improve small-craft navigation, provided it was designed appropriately. The design depth in the channel was based on the California Department of Boating and Safety (1980), which stipulates a minimum depth of 0.9 m below the deepest draft of the vessel or 1.5 m, which ever is greater. Considering 1.5 m as the minimum depth for navigation plus 0.63 m for sedimentation a nd allowance for over-depth dredging, the final dredged depth of the channel works out to 2.13 m. The California Department of Boating Safe ty stipulates a minimum width of 23 m at the design depth. The existing navigation cha nnel (Figure 5.5) is maintained at 33 m at the design depth of 2.2 m. However, with a side slope of 1:3m, the top width C-18 canal dredged to –3.5m Ganju (2000) trap

PAGE 108

92 works out to 23 m + 13.2 m = 36.2 m. Consider ing that cell dimensions are fixed at 25 m x 25 m, a self-cleaning channel width of 50 m was adopted in the model. Figure 5.4 Planform view of the proposed self-cleaning channel in the bay. Figure 5.5 Location of the sea grasses indicated in model with increased roughness. Existing channel dredged to -2.2 m Bay channel dredged to -2.13 m

PAGE 109

93 The channel was designed with three bends Per design stipulations (Bruun, 1989), the bend angle should not be more than 30o and radius of curvature not less than 1,500 m. Also, at bends widening is generally carried out. For the model the width at the bends were widened by 25 m (one extra cell). As s een from Figure 5.4 and 5.5 the channel is largely outside the zones where sea grass beds occur. Figure 5.6 Planform view of the proposed self-cleaning Y-channel in bay. 5.1.2.3 Alternative No. 4: Bay Y-channel The proposed Bay Y-channel originates fr om the existing navigation channel, and bifurcates into the Southwest and Northwest Forks (Figure 5.6). The width of the channel must be equal to that of the exiting channel, i.e., 30.5 m at the design depth of 2.2 m. The same depth is adopted for the extension. With a side slope of 1:3 and a depth of 2.2 m, the width at the top works out to 30.5 + 2x3x2. 2 = 43.2 m. Hence in the model a two-cell wide (50 m) channel was simulated. At th e bifurcation a total channel width of 100 m (four cell widths) was chosen for navigation purposes. Y-channel dredged to -2.2 m

PAGE 110

94 5.1.2.4 Alternative No. 5: Northwest Fork Channel This alternative is located in fairly shal low depths and is devised to channelize the flow into a deeper channel so that mud would be prevented from depositing under increased velocity. Accordingly, the channel would be self-cleaning and thereby reduce the cost of maintenance. The location of th e channel, shown in Figure 5.7, goes around a shoal. The width of the channel is maintained at 50 m with a (navig able) depth of 2.13 m and side slopes of 1:3. The widths at the bends are increased by one cell width for navigational purpose. -6.836-.305Bottom ElevTime: 0.00 Figure 5.7 Planform view of the proposed self-cleaning channel in the Northwest Fork. 5.1.3 Efficiency Analysis 5.1.3.1 Velocity Vector Calculation The direction of the resultant flow velocity vector is calculated from 1180 tan u v (5.1) N W Fork Channel dredged t o -2 1 3 m

PAGE 111

95 where u and v are the two measured velocity components. The corresponding magnitude, u, is obtained from 22u uuv u (5.2) 5.1.3.2 Sediment Deposition Calculation The rate of mass deposition rate under flow is given by 1;b s bc cDWCLW (5.3) where b is the bed shear stress, c is the critical shear stress for erosion, s W is the settling velocity, C is the suspended sediment concentration, and L and Ware the length and width of the trap or channel, respectively and deposition D is given in kg/sec. The bed shear stress in the model is calculated from 22 1111(,),bxbybcuvuv (5.4) where 2 1ln(/2)b oc z (5.5) and where 1= 0.25, 00.047mz and the Karman constant 0.4 Substituting these values in Equation (5.5) gives bc = 0.167. The critical bed shear stress of erosion is determined based on the type of sediment, i.e., sand or fine-grained. These values are determined next. Fine sediment: For the fine sediment, a c value of 0.1 Pa is selected for calculation of the critical velocity for erosion, uc (Mehta and Parchure, 2000). From Equation (5.4) c is given as

PAGE 112

96 2cbccu (5.6) Therefore, with bc= 0.167, cu= 0.247 m/s is obtained. Sand: For sand, uc can be calculated from the ShieldsÂ’ relationship, under fully rough turbulent flow, which is given by (Buckingham and Mehta, 1985) 0.5 500.0133[()]csud (5.7) where s is the unit weight of sand (= 2,650 kg/m3), is the unit weight of estuary water (= 1,020 kg/m3) and 50d is the median grain diameter, which ranges between 0.1 and 0.4 mm (Jaeger et al., 2001). For calculations th ree median diameters were considered, namely, 0.1, 0.2 and 0.3 mm. The correspondi ng critical velocities from Equation (5.7) are given in Table 5.2. In connection with these values, we note that the shear stress being proportional to the square of flow ve locity under turbulent flows, Equation (5.3) may be expressed as 2 21s cu DWCLW u (5.8) where u is the current velocity. Table 5.2 Critical velocities for sand Median Diameter, d50 (mm) Critical Velocity, uc (m/s) 0.1 0.17 0.2 0.24 0.3 0.29 Deposit thickness: The calculation for sand deposition is carried out using Equation (5.8). For that purpose, the mass of sand m was converted into the corresponding volume using relation (1)sm V n (5.9)

PAGE 113

97 where, Vis the sand volume, s is the sand granular density, and n is the porosity. For the present calculations, s =2,650 kg/m3, and n=0.4. Conversion of volume of sand deposited into sedimentation thickness Dr can be accomplished from (1)r sVm D AnA (5.10) where A is the area of deposit. With regard to fine sediment the same calculations were performed and the total sedimentation rate was calculated. The deposit volume was calculated from dm V (5.11) where d is the sediment dry density, which wa s calculated with 15% organic content (OC) from the relation 1900exp(0.156)80dOC (5.12) with 15% OC (Ganju, 2000) the density works out to 263 kg/m3. 5.1.3.3 Trap Efficiency Trap efficiency is defined in terms of the sediment removal ratio r given by ()() () s ise siqq r q (5.13) where () s iq is the influent sediment load, and () s eq is the effluent load. Efficiencies for different trap scenarios were examined by Ganju et al. (2001). Accordingly, sediment flux calibration was carried out using those results, so that the r values could be determined on a consistent basis. Separate calculations were carried out for sand and for fine sediment.

PAGE 114

98 5.1.3.4 Channel Efficiency The method of calculation of the efficiency of the self-cleaning channel is similar to the one given above. In the present analysis an ideal self-cleaning channel is defined as one in which the removal ration r = 1 (or 100%). Erosion of channel, as might occur under very high flows, was not explored separately because the environment in question is largely a depositional one. 5.2 Design Simulations 5.2.1 Design Flows Model simulations were carried out under three tributary flow scenarios – flows used for model calibration, median (50 per centile) flows and peak (98 percentile) flows, following Ganju (2000). The discharges are listed in Table 5.3. Table 5.3 Design flows in tributaries Tributary Southwest Fork Northwest Fork North Fork Calibration discharge (m3/s) 14.5 32 0.3 Median discharge (m3/s) 1.3 0.7 0.1 Peak discharge (m3/s) 32 61 1.9 5.2.2 Alternative 1 Water surface elevation time series at th e three tidal stations (UFG1, UFG2, and UFG3) were available for a period of 34 da ys (09/14/2000 to 10/18/ 2000). As noted, data from UFG1 were used at the eastern boundary and data from other two stations were used for verification of model accuracy. Accordi ngly, model runs were restricted to these 34 days. Simulation with the existing bathymet ry, i.e., Alternative 1, was intended to observe the present flow regime and comp are the same with results from changed bathymetry under other alternatives. This woul d enable the calculation of the change in

PAGE 115

99 velocity and bottom shear stress due to depth change at a trap or a channel (see, for example, Figure 5.8). 0.0000 0.0050 0.0100 0.0150 0.0200 0.0250 0.0300 0.0350 0.0400 0.0450 277.5000278.0000278.5000279.0000279.5000280.0000280.5000JULIAN DAYS IN YEAR 2002CURRENT SPEED (m/s) Existing Dredged Figure 5.8 Current comparisons for a model cell at the upstream end of the Northwest Fork channel: calibration discharges. 5.2.3 Alternatives 2, 3, 4 and 5 Under calibration discharges in the tri butaries, the maximum peak flood and ebb velocities are compared in Table 5.4 for the four alternatives. For each alternative, the three values (upstream, mid-point and downstr eam) have been averaged and given for all three tributary discharge scenarios (calibration, median and peak) in Table 5.5. Figure 5.9 shows the current velocity vectors for th e maximum flood flow condition at spring tide over the model domain. Figure 5.10 shows the corresponding ebb flow vectors.

PAGE 116

100 Table 5.4 Maximum currents at alternatives: calibration discharges Upstream Mid-point Downstream Alternative Condition Ebb (m/s) Flood (m/s) Ebb (m/s) Flood (m/s) Ebb (m/s) Flood (m/s) Non-dredged 0.03 0.025 0.05 0.028 0.05 0.03 C-18 canal Dredged 0.01 0.008 0.035 0.021 0.04 0.03 Non-dredged 0.30 0.23 0.26 0.12 0.12 0.08 Bay channel Dredged 0.21 0.17 0.14 0.12 0.10 0.07 Non-dredged 0.30 0.26 0.23 0.20 0.08 0.08 Y-channel Dredged 0.21 0.16 0.10 0.09 0.04 0.04 Non-dredged 0.06 0.04 0.04 0.04 0.05 0.04 NWF channel Dredged 0.03 0.027 0.02 0.02 0.025 0.02 Table 5.5 Maximum currents at alternatives: Different discharges Calibration discharge Median Discharge Peak discharge Alternative Condition Flood (m/s) Ebb (m/s) Flood (m/s) Ebb (m/s) Flood (m/s) Ebb (m/s) Non-dredged 0.07 0.05 0.03 0.01 0.16 0.11 C-18 trap Dredged 0.06 0.04 0.04 0.03 0.11 0.10 Non-dredged 0.30 0.23 0.21 0.18 0.29 0.26 Bay channel Dredged 0.21 0.17 0.18 0.14 0.35 0.30 Non-dredged 0.30 0.28 0.22 0.20 0.21 0.20 Y-channel Dredged 0.21 0.20 0.17 0.16 0.32 0.30 Non-dredged 0.06 0.04 0.04 0.03 0.10 0.06 NWF channel Dredged 0.03 0.027 0.02 0.02 0.09 0.06

PAGE 117

101 .25 (m/s)VelocitiesTime: 258.08 Figure 5.9 Current velocity vectors over the modeled domain; maximum flood velocities at spring tides. .25 (m/s)VelocitiesTime: 259.92 Figure 5.10 Current velocity vectors over the modeled domain; maximum ebb velocities at spring tide.

PAGE 118

102 5.3 Deposition Equation Calibration In order to apply Equation (5.3), it must be calibrated against existing deposition rates within the modeled domain. This is described next. 5.3.1 Calibration for Sand Taking the bay channel as an example, we note that Ganju et al. (2001) reported entry of 9.2 Mkg (5,786 m3) of sand into the central embayment from the inlet. Accordingly, the report calculates a uniform deposition of 3.25 mm/year. Since the bay area is 1,780,308 m2, using Equation (5.3) and a uni form (non-dredged condition; calibration discharge) current of 0.12 m/s (ave rage of all cells), the sand settling flux can be worked out as follows: 3. Rate of sedimentation per unit bed area = 9.2 Mkg/ 1,780,310 m2 = 5.16 kg. 4. For 0.1 mm diameter sand, critical velocity = 0.17 m/s (Eq.5.7). 5. Using Equation (5.8), sand settling flux = 2.81x10-7 kg/m2 s. Table 5.6 provides all results based on such calculations. 5.3.2 Fine Sediment Table 5.6 Calibration for sediment fluxes Alternative Total supply (Mkg) Area of deposit (m2) Unit deposition (kg/m2) Deposition flux (kg/m2 s) Sand = 0.0 577,560 0.00 0.0 C-18 canal Fines = 1.27 577,560 2.20 8.72x10-8 Sand = 9.20 1,780,310 5.16 2.81x10-7 Bay channel Fines = 0.46 1,780,310 0.26 9.36x10-9 Sand = 9.20 1,780,310 5.16 2.81x10-7 Bay Ychannel Fines = 0.46 1,780,310 0.26 9.36x10-9 Northwest Fork 2.00 1,949,890 0.87 3.42 x 10-8 Similar calculation for fine sediment was carried out. As an example, for the bay channel, the total inflow of fine sediments is 0.46 Mkg. So the total volume works out to

PAGE 119

103 1,749 m3. Accordingly, the sediment flux works out to 9.36 x 10-9 kg/m2 s. Table 5.6 provides all results. 5.4 Sand Deposition due to Alternatives 5.4.1 Bay Channel Table 5.7 gives the sand deposition flux in the bay channel averaged over the entire length for the three discharge scenarios – calibration, median and peak. In these calculations the channel length is taken as 1,430 m, and width 50 m. The respective mean sediment concentration values are 0.04, 0.02 and 0.10 kg/m3. Table 5.7 Rate of sand deposition in bay channel Deposition flux (kg/m2 s) Median diameter (mm) Calibration discharge Median discharge Peak discharge 0.1 8.80x10-6 9.67x10-6 5.64x10-6 0.2 1.32x10-5 2.76x10-5 1.14x10-5 0.3 3.52x10-5 4.06x10-5 2.74x10-5 5.4.2 C-18 Canal Table 5.8 Rate of sand deposition in C-18 cana Deposition flux (kg/m2 s) Median diameter (mm) Calibration discharge Median discharge Peak discharge 0.1 1.38x10-5 2.62x10-5 0.89x10-5 0.2 2.07x10-5 3.68x10-5 1.24x10-5 0.3 3.64x10-5 5.06x10-5 2.26x10-5

PAGE 120

104 In this case, for the calibration discharge case the mean sediment concentration in the C-18 canal was taken as 0.048 kg/m3, as recorded at the Southwest Fork tower. For the median and peak discharges the respective values are 0.020 and 0.10 kg/m3. The percent of fines is taken as 15. The 1,890 m long canal is considered having a uniform width of 50 m. Results are given in Table 5.8. 5.4.3 Bay Y-channel Deposition fluxes for the 850 m (stem plus one arm) long and 50 m wide channel are given in Table 5.9. Table 5.9 Rate of sand deposition in Y-channel Deposition flux (kg/m2 s) Median diameter (mm) Calibration discharge Median discharge Peak discharge 0.1 4.41x10-6 5.76x10-6 3.08x10-6 0.2 8.80x10-6 3.16x10-6 6.24x10-6 0.3 1.38x10-5 3.06x10-5 0.74x10-5 5.5 Fine Sediment Deposition due to Alternatives Table 5.10 gives the fine sediment budget based on the concentration reported in the data collection in Southwest and Northw est Forks. The results compare well with those of the values predicted in Ganju et al. (2001).

PAGE 121

105 Table 5.10 Rate of fine sediment deposition in alternatives Deposition flux Alternative Calibration discharge Median discharge Peak discharge C-18 canal 2.25x10-5 2.75x10-5 1.45x10-5 Bay channel 2.26x10-6 2.87x10-6 1.38x10-6 Bay Y-channel 2.26x10-6 2.87x10 -6 1.38x10 -6 NW Fork channel 4.06x10-6 5.16x10-6 2.76x10-6 5.6 Sediment Removal 5.6.1 Calculation of Deposition Deposition in the trap and the channels was calculated using the above results. The removal ratio from Equation (5.13) was calculated by adopting the sediment load in the tributaries reported by Ganju et al. (2001). Tables 5.11 and 5.12 give the annual sand budget and the Tables 5.13 and 5.14 the fi ne sediment budget in the channel for calibration and peak discharge cases (leaving out median discharges for illustration. Table 5.11 Annual sand budget: Calibration discharge Alternative Rate of deposition (mm/s) Total deposition (mm) Total deposition (kg) Bay channel 8.81x10-6 8.30 943,590 Bay Y-channel 4.40x10-6 8.83 989,780 Table 5.12 Annual sand budget: Peak discharge Alternative Rate of deposition (mm/s) Total deposition (mm) Total deposition (kg) Bay channel 5.64x10-6 5.32 604,800 Bay Y-channel 3.08x10-6 5.81 585,000

PAGE 122

106 Table 5.13 Annual fine sediment budget: Calibration discharge Alternative Rate of deposition (mm/s) Total deposition (mm) Total deposition (kg) C-18 canal 2.25x10-5 17.5 286,400 Bay channel 2.26x10-6 1.93 38,730 Bay Y-channel 2.26x10-6 1.93 20,400 NW Fork channel 4.06x10-6 7.50 62,220 Table 5.14 Annual fine sediment budget: Peak discharge Alternative Rate of deposition (mm/s) Total deposition (mm) Total deposition (kg) C-18 canal 1.45x10-5 11.30 184,570 Bay channel 1.38x10-6 1.17 23,630 Bay Y channel 1.38x10 -6 1.17 12,440 NW Fork channel 2.76x10-6 5.10 42,310 5.6.2 Calculation of Channel Efficiency For an assessment of the efficiencies of the trap/channels, we will assume that the two channels in the bay accumulate sand onl y, whereas C-18 canal and the Northwest Fork accumulate fine sediment only. On that basis, Table 5.15 summarizes the annual load and shoaling (rounded to nearest mm) in the canal and channels. From these results that while the C-18 canal trap acts as such, the three channels are unlikely to be selfcleaning. On the other hand we note that th e performance of the channels improves at peak discharges, which, in general, highli ghts the sediment-flushing role of high river discharges, as at all estuaries in their natural state. Table 5.15 Annual sediment loading Calibration discharge Peak discharge Alternative Load (metric tons) Shoaling (mm) Load (metric tons) Shoaling (mm) C-18 canal 286 18 185 11 Bay channel 944 8 605 5 Bay Y-channel 990 9 585 6 NW Fork channel 62 8 42 5

PAGE 123

107 5.6.3 Removal of Bay Sediment The role of proposed alternatives as sediment traps, especially under typical prevailing flow conditions in the estuary lends itself to an assessment of these alternatives as a means to reduce sedimentation in the bay. This is noted next. As mentioned, Ganju et al. (2001) reported entry of 9,200 (metric) tons of sand into the central embayment, corresponding to a uniform shoaling thickness of 3.3 mm/year (accurate to tenth of mm). Since the bay ch annel would remove 940 tons (at calibration discharge), the net shoaling would reduce to 2.9 mm/year. Similar calculation for fine sediment re moval by the C-18 canal trap plus the Northwest Fork channel can be shown to re duce the deposition of 0.78 mm/year of fine sediment in the bay to 0.57 mm/year. 5.7 Assessment of Alternatives It is instructive to make a qualitative asse ssment of the proposed alternatives based on their roles in improving water quality (by trapping fine sediment) and navigation (by trapping sand). Based on assignment of num bers: +1 (good), 0 (no effect) and -1 (negative impact), Table 5.16 provide the assessment.

PAGE 124

108 Table 5.16 Assessment of impacts of proposed alternatives Alternative Sedimentation control Navigation Overall Comment 1 No-action -1 -1 -2 -2, however does not mean that the present condition is severe 2 C-18 canal +1 0 +1 If fine organicrich sediment continues to accumulate in to the central bay, this option should be considered 3 Bay channel +1 +1 +2 Should be considered for implementation; careful design will be required so that sea grass beds are not disturbed 4 Y-channel -1 +1 0 Despite its effectiveness as a trap, shoaling may be rapid because it would cut existing and active shoal 5 NWF channel +1 0 +1 If fine sediment accumulation continues, this option should be considered

PAGE 125

109 CHAPTER 6 CONCLUSIONS 6.1 Summary The objective of this study was to asse ss the implementation of schemes for sediment entrapment and self-cleaning channels in the micro-tidal (< 2 m range) estuarine environment containing both sand and fine sediment. The central embayment of the Loxahatchee River estuary on the east (Atlan tic) coast of Florida was chosen as the candidate location due to its unique character istics with respect to the influx of fine sediment and sand in the central embayment, a nd the concerns that have arisen in recent years due to the potential for long term impact s on the system due to this influx. The spring tidal range in the central embayment is 0.8 m, and three main tributaries feed the bay; Southwest Fork/C-18 canal, Northwest Fork and North Fork. Bay sediment is a mixture of sand, silt and clay along with organic detritus. An ideal sediment trap captures all of incoming sediment, i.e., the removal efficiency is 100%. A self-cleaning channe l allows no net deposition or erosion of incoming sediment, which passes through, so that its removal efficiency is 0%. In addition to the “no-action” present condition in the study area, four alternatives were examined: a sediment trap in the C-18, a self-cleaning channel in the central bay itself, a (self-cleaning) Y configured extension of th e existing navigation channel in the bay and a self-cleaning channel in the Northwest Fork All the self-cleaning channels were designed to meet the minimum depth and wi dth required for the shallow draft vessels plying the area.

PAGE 126

110 The basis for the introduction of the sediment trap was to capture sediment arriving from the S-46 sluice-gate structure at the h ead of the C-18 canal, prevent the material from settling in the Southwest Fork and the cen tral embayment. The overall rationale for the introduction of the self-cleaning channe ls was to ease the passage of sediment, especially the fine-grained component from th e tributaries, so as to prevent its deposition in the central embayment, and thereby enable it to exit to the Atlantic from Jupiter Inlet. Hydrodynamic flow modeling was carried out using the Environmental and Fluid Dynamic Code (EFDC) to calculate water el evations, velocities, and bed bottom shear stresses in the estuarine domain. The model was calibrated using data on water levels, currents and suspended sediment concentrati on collected in years 2002/2003 at a site in the Southwest Fork. Validation was then carri ed out using data collected in 2000 in the central embayment and in the Northwest Fork in 2003. Following model validation, model runs were carried out for the selected alternatives, and their efficiencies were calculated by relating the sediment settling flux with change in flow momentum and hen ce bed shear stress. The results of these simulations and conclusions are discussed in the following section. 6.2 Conclusions 1. Calculations indicate that the concepts of sediment entrapment and of selfcleaning can operate only partially in th e study area due to the weak prevailing forcing by tide and the episodic nature of the freshwater discharges in the tributaries. 2. The simulations indicated that under c onditions of typical discharges from the tributaries, with the introduction of Bay channel and Bay Y-channel, annually 1,930 (metric) tons of sand out of the total 9,200 tons entering the bay could be captured. The Bay channel would reduce the sand load by 940 tons, thus lowering the present bay-mean 3.3 mm/year shoaling thickness by 0.4 mm/year. The addition of the Y-channel would reduce shoaling by 0.9 mm/year. Also, these two channels could entrap about 68 tons of fi ne sediments, thus reducing the present 0.78 mm/year fine sediment shoaling thickness by 0.11 mm/year. Simulations

PAGE 127

111 under peak discharges showed reduced en trapment by the two channels, because under high discharges their self-cleaning performance improved. For sand, the Bay channel reduced the shoaling thickne ss by 0.4 mm/year, and the two channels together by 0.8 mm/year. Similarly, the tw o channels would reduce fine sediment shoaling by 0.11 mm/year. Note however that such flows do not occur frequently in the study area. 3. On account of its length (same as that of the canal), the C-18 trap was found function better than the short (60 m l ong) trap of Ganju (2001), with annual entrapment increasing from 159 tons to 290 tons under typical discharge sequence from the S-46 structure, and reduced to 190 tons under peak discharge. 4. The Northwest fork channel did not function per expectation. Although the percent-wise flow velocity reduction wa s found to be the least (40%) in this channel of all the alternatives, as a resu lt of velocity reduction self-cleaning was not achieved. The annual entrapment of fine sediment was about 62 tons. Under peak discharge this value would reduce to 42 tons. 5. From the simulations it can be concluded that while the C-18 canal trap acts as such, the three channels are unlikely to be self-cleaning. On the other hand, the performance of these channels improves at peak discharges, which highlights the sediment flushing role of the high river di scharges, as at all estuaries in their natural state. 6. The implementation of C-18 trap, th e Bay channel and the Northwest Fork channel could collectively reduce bay-mean sedimentation from as much as 3.3 mm/year to 2.9 mm/year from sand and from 0.78 mm/year to 0.47 mm/year for fine sediment. 7. A qualitative assessment of the proposed alternatives based on their roles in improving water quality (by trapping fine sediment) and navigation (by trapping sand) was carried out based on the a ssignment numbers, +1 for good, 0 for no effect and –1 for negative impact. Th is assessment leads to the following observations. 8. Although the present condition in the study area with respect to sedimentation does not appear to be severe (in compar ison with numerous estuaries elsewhere), implementation of the one or more of the above alternatives may be considered at a future date, if necessary. 9. Fine sediment accumulation in the central bay due to ingress of sediment from upland discharge and erosion of existing and old deposits could be reduced with the installation of C-18 trap, as it accounts for more than half of the total fine sediment entering the bay. Periodic dredgi ng of the canal would then remove the trapped sediment. 10. Bay channel (close to the southern bank of the central embayment) appears to be a good option. Since it would act as an effici ent ebb flow channel, it could improve

PAGE 128

112 bay flushing as well as navigation and se rve to reduce bay-wide sedimentation. Under peak river discharges the channe l appears to possess a degree of selfcleaning capability for fine sediment and, accordingly, it may require low maintenance dredging. However, careful design considerations with regard to its alignment will be required to stay clear of sea grass beds in the area. 11. The Y-channel, despite having good sand trapping capability, is likely to shoal up with sand rapidly, as it cuts through active shoals. Also, this channel is not effective for trapping of fine sediment par tly due to its cross-flow orientation (70110) with respect to the direction of the prevailing flows. 12. It appears that while the Northwest Fork channel will not be able to draw additional flow into it, it would assist in trapping fine sediment, and may be considered for implementation if the presen t rate of accumulation of fine sediment in that area continues. 6.3 Recommendations for Future Work Trap efficiency modeling would be more accurate if based on a sediment model linked to EFDC. Simulations can then be ex tended to calculate sediment transport by accounting for the role of sediment compos ition in greater detail and with greater accuracy. Sources of sediment internal to the modeled domain may have to be simulated in order to identify the internal movement of sediment and formation of shoals in the Northwest Fork and the central embayment. Long-term simulation of flows and sediment transport on the order of years (at least one year) is required to evaluate the net sediment movement necessary for a more effective ex amination of the efficiency of traps and channels.

PAGE 129

113 LIST OF REFERENCES Antonini, G.A., Box, P.W., Fann, D.A., and Grella, M.J., 1998. Waterway evaluation and management scheme for the south shore a nd central embayment of the Loxahatchee River Florida. Technical Paper TP-92, Florida Sea Grant College Program, Gainesville, Florida,. Brunn, P., Port Engineering, Fourth Edition, 1989, Gulf Publishing Company, Houston, Texas Buckingham, W.T., Mehta A.J., 1985. Physi cal Modeling of Tidal entrances: a case study. Proceeding of the First National Conference on Dock and Harbor Engineering, Vol. 2, Indian Institute of technology, Mumbai, E.39-E.47 Department of Army, U.S. Army Corps of Engineers, 1984, Shore Protection Manual, U.S. Government Printing Office, Washington, D.C. Ganju, N.K., 2001. Trapping of organic-rich fi ne sediment in an estuary. M.S. Thesis, University of Florida, Gainesville, Florida Ganju, N.K., Mehta, A.J., Parshukov, L.N., and Krone, R.B., 2001. Loxahatchee River Florida Central Embayment: Sedime nt budget and trap evaluation. Coastal and Oceanographic Engineering Progr am, Report no UFL/COEL-2001/008 University of Florida Goodwin, C.R., and Russell, M.G., 1984. Simu lation of tidal flow and circulation patterns in the Loxahatchee River estuary, southeastern Florida. Water-Resource Investigation Report 87-4201, U.S. Geological Survey, Tallahassee, Florida. Hamrick, M.J., 1992. Three dimensional flui d dynamics computer code: Theoretical and computational aspects, Special Report No 317 in Applied Marine Science and Ocean Engineering, Virginia Institute of Marine Science, Gloucester Point, Virginia Ippen, A.T. and Harleman, D.R.F., 1966. Tidal dynamics in estuaries. In: Estuary and Coastline Hydrodynamics, A.T. Ippen, ed., Engineering Societies Monographs, McGraw Hill, New York, Jaeger, J. M., and Hart, M., 2001. Sediment ary processes in the Loxahatchee River Estuary: 5000 years ago to the present. Final Report, submitted to the Jupiter Inlet District Commission, Jupiter, Florida, Department of Geology and Geophysics, University of Florida, Gainesville, Florida.

PAGE 130

114 Jupiter Inlet District, 1999. Environmental re storation program for the Loxahatchee River central embayment. JID Report, Jupiter Inlet District Commission, Jupiter, Florida. Lambe, T.W., and Whitman R.V., 1969., Soil Mechanics, Wiley, New York. Marsh-McBirney, 2001 Electromagnetic Current meter (Model 585 OEM), Operation Manual., USA McPherson, F.B., and Sonnetag, H.W., 1984. Sediment concentrations and loads in the Loxahatchee River estuary, Florida, 1980-82. Water-Resource Investigation Report 84-4157, U.S. Geological Survey, Tallahassee, Florida. Mehta, A.J., and Parchure, T.M., 2000. Su rface erosion of fine-grained sediment revisited. In: Muddy Coast Dynamics and Resource Management, B.W. Flemming et al. eds., Elsevier, Amsterdam. Mehta, A.J. and Li., 2003., Coastal Cohe sive Sediment Transport Class notes, Principles and Process-modeling of C ohesive Sediment Transport, University of Florida, Florida Moustafa., J. and Hamrick M.J., 1995. A three-dimensional environmental fluid dynamics computer code: Theoretical and computational aspects. Special Report 317. The College of William and Mary, Virginia Institute of Marine Science, Virginia, USA. Nakamura, S., 2001. Applied Numerical Methods with Software. Prentice Hall, Englewood Cliffs, New Jersey Ochi, M.K., 1990. Applied Probability and Stochastic Processes. John Wiley and Sons, New York. Russell, M.G., and McPherson, F.B., 1982. Fres hwater runoff and salinity distribution in Loxahatchee River estuary, Southeastern Florida, 1980-82. Water-Resource Investigation Report 83-4244, U.S. Geological Survey, Tallahassee, Florida. Simpson, M.R., and Oltmann, R.N., 1993. Discharge-measurement system using an acoustic Doppler current profiler with appli cations to large rivers and estuaries. Water Supply Paper 2395, U.S. Geological Survey, Sacramento, California. Sonnetag, H.W., and McPherson, F.B., 1984. Nutrient input from the Loxahatchee River environmental control district sewage-tr eatment plant to the Loxahatchee River estuary, southeastern Florida. Water-Resource Investigation Report 84-4020, U.S. Geological Survey, Tallahassee, Florida. Vito, A.V., Editor, 1975. Sedimentation Engineering. ASCE Manuals and Reports on Engineering Practice – No 54, PWS-KENT Publishing Company, Boston, USA

PAGE 131

115 Wanless, H., Rossinsky, V., Jr, and McPhers on, F.B., 1984. Sediment ologic history of the Loxahatchee River estuary, Florida. Water-Resource Investigation Report 84-4120, U.S. Geological Survey, Tallahassee, Florida.

PAGE 132

116 BIOGRAPHICAL SKETCH Rashmi Ranjan Patra was born in Bhubaneswar in the state of Orissa, India, to Mrs. Sabitri and Dr. Gouranga Ch. Patra. After schooling in M.K.C. High School and M.P.C. College in Baripada, Orissa, the author we nt to the Indian Institute of Technology, Kharagpur, for a bachelorÂ’s degree in civ il engineering. During initial years after graduation, he worked as a design engineer for the development of the first dry-dock project in India for Keppel Shipyards, Singapore, and then for Water and Power Consultancy Services (India) Limited before joining the coastal engineering program of the Department of Civil and Coastal Engineer ing at the University of Florida for the masterÂ’s degree. Upon obtaining the degree the author plans to practice as a professional and continue the work he has been doing, and thereby contribute to the field, which is so fascinating.