Meandering Processes and Channel Morphology in the Lower Mississippi River Prior to Major Human Disturbance

Home , Alluvial river, Channel pattern, Helicoidal flow, Mississippi River System, River morphology

Louisiana State University LSU Digital Commons

LSU Historical Dissertations and Theses Graduate School

1998 Meandering Processes and Channel Morphology in the Lower Mississippi River Prior to Major Human Disturbance. Paul Franklin Hudson Louisiana State University and Agricultural & Mechanical College

Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses

Recommended Citation Hudson, Paul Franklin, "Meandering Processes and Channel Morphology in the Lower Mississippi River Prior to Major Human Disturbance." (1998). LSU Historical Dissertations and Theses. 6741. https://digitalcommons.lsu.edu/gradschool_disstheses/6741

This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contact [email protected]. INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor MI 48106-1346 USA 313/761-4700 800/521-0600

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MEANDERING PROCESSES AND CHANNEL MORPHOLOGY IN THE LOWER MISSISSIPPI RIVER PRIOR TO MAJOR HUMAN DISTURBANCE

A Dissertation

Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of Doctor of Philosophy

The Department of Geography and Anthropology

by Paul Franklin Hudson B.S., Jacksonville University, 1991 M.S., University of Florida, 1993 August 1998

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 9902643

This microform edition is protected against unauthorized copying under Title 17, United States Code.

UMI 300 North Zeeb Road Ann Arbor, MI 48103

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DEDICATION

This work and effort is dedicated to the memory of my best friend, Willford I.

Soud, who died in March of 1996. Willie was one of those rare individuals who took

the road less traveled, and didn’t look back. He was a writer, an independent thinker,

a scholar, lover of life, and an all around good guy. The world is a little bleaker

without him.

iii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS

I would like to thank my doctoral advisor, Professor Richard H. Kesel, for his

assistance and effort in this project and during my graduate training at LSU. Thanks also

to my committee members for their support and insight, and for being good people, Dr.

Anthony Lewis in the Department of Geography and Anthropology, Dr. Gregory Stone

and Dr. Harry Roberts in the Coastal Studies Institute, and the Dean’s representative, Dr.

Jeffery Nunn in the Department of Geology.

I would like to thank the Robert C. West and the Richard Russell Field Research

Fund for financial support for my travels to Vicksburg, MS. I would also like to thank

DeWitt Braud and Farrell Jones in the Department of Geography and Anthropology for

research assistantships and sum m er funding, and for their knowledge of GIS and remote

sensing. I certainly wish to thank my friends and colleagues in Geography and

Anthropology, Michael Steinberg, Steven Rainey, Phillip Chaney, Michael Hawkins,

everyone of the CMSL crew, and many others far too numerous to mention 'who made

this experience more enjoyable, and even bearable. Also, a toast to the “draft-horse

scholars”, who find a way to get it done.

1 would also like to thank those who have helped me in the past, and continue to

encourage. In particular this includes Joann Mossa in the Department of Geography at

the University of Florida who first introduced me to fluvial geomorphology, and Ray

Oldakowski at Jacksonville University who encouraged me to pursue my interests in

graduate school. A warm and special thanks to Julianne Marie Walsh for helping me get

“there”. Lastly, I would like to thank my family, particularly my mom and dad, for I

would not have been here if it hadn’t been for the excellent example they set

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS

DEDICATION...... ii

ACKNOWLEDGEMENTS...... iv

LIST OF TABLES...... viii

LIST OF FIGURES...... x

ABSTRACT...... xiii

CHAPTER 1. INTRODUCTION...... I 1.1 OVERVIEW AND RESEARCH PROBLEM...... 1 1.2 PURPOSE...... 5 1.3 RESEARCH OBJECTIVES...... 5 1.4 APPROACHES TO MEANDERING PROCESSES...... 6 1.5 CHANNEL PATTERN AND ADJUSTMENT...... 8 1.6 RESEACH DESIGN...... 15

CHAPTER 2. PHYSICAL SETTING...... 17 2.1 DISCHARGE AND SEDIMENT REGIME...... 17 2.2 GEOLOGIC SETTING...... 24 2.3ALLUVIAL FLOODPLAIN...... 26 2.4 DELTAIC PLAIN...... 31

CHAPTER 3. DATA AND METHODS...... 32 3.1 HYDROGRAPHIC SURVEYS...... 32 3.2 GEOGRAPHIC INFORMATION SYSTEMS...... 33 3.2.1 Conversion to Vector Format, Channel Segments 33 3.2.2 Error...... 34 3.3 PRIMARY VARIABLES...... 35 3.3.1 Migration Rates...... 35 3.3.2 Channel Width ...... 37 3.3.3 Bankfull Channel Width ...... 37 3.3.4 Radius of curvature ...... 38 3.3.5 Sinuosity ...... 38 3.3.6 Thalweg, Pools and Riffles ...... 38 3.3.7 Bed Material ...... 39 3.3.8 Valley Slope...... 39 3.4 ANALYSIS OF DATA...... 41

CHAPTER 4. SPATIAL AND TEMPORAL TRENDS...... 43 4.1 OVERVIEW...... 43 4.2 LONGITUDINAL PROFILE...... 43

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.3 PLANFORM MORPHOLOGY...... 48 4.3.1 Channel Width ...... 48 4.3.2 Radius of Curvature ...... 53 4.3.3 Sinuosity ...... 56 4.4 CHANNEL MIGRATION...... 56

CHAPTER 5. PLANFORM ADJUSTMENT AND CHANNEL MIGRATION...... 63 5.1 OVERVIEW...... 63 5.2 PLANFORM INTERRELATIONSHIPS...... 63 5.2.1 Sinuosity and Radius of Curvature ...... 63 5.2.2 Sinuosity and Channel Width ...... 68 5.2.3 Radius of Curvature and Channel Width ...... 68 5.3 DOWNSTREAM TRENDS...... 69 5.4 VALLEY SLOPE AND CHANNEL ADJUSTMENT...... 72 5.5 RELATIONSHIPS BETWEEN BED MATERIAL AND CHANNEL PATTERN...... 78 5.6 CHANNEL CURVATURE AND CHANNEL MIGRATION... 83 5.7 CONTROLS ON MEANDER BEND MIGRATION ...... 85

CHAPTER 6. POOL AND RIFFLE MORPHOLOGY...... 90 6.1 INTRODUCTION...... 90 6.2 SPATIAL AND TEMPORAL TRENDS IN POOL - RIFFLE MORPHOLOGY...... 90 6.3 RELATIONSHIPS BETWEEN POOL - RIFFLE MORPHOLOGY AND CHANNEL WIDTH...... 94 6.4 RELATIONSHIPS BETWEEN POOL - RIFFLE MORPHOLOGY WITH VALLEY SLOPE...... I l l 6.5 RELATIONSHIP BETWEEN POOL - RIFFLE MORPHOLOGY AND BED MATERIAL...... 113 6.6 RELATIONSHIP BETWEEN POOL - RIFFLE MORPHOLOGY WITH PLANFORM MORPHOLOGY...... 117

CHAPTER 7 - DISCUSSION, CONCLUSIONS, AND FUTURE RESEARCH...... 120 7.1 DISCUSSION ...... 120 7.1.1 Mechanisms of Channel Change ...... 120 7.1.2 Influence of Sediment and Valley Slope ...... 122 7.1.3 Linking Pool and Riffle Morphology with Meandering Processes...... 124 7.1.4 Significance and Contribution ...... 127 7.2 CONCLUSIONS ...... 129 7.3 FUTURE RESEARCH ...... 131

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES...... 133

VITA...... 145

vii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES Table:

1.1 Significant variables influencing channel morphology of the Lower Mississippi River during modem time 101 -lCr ...... 2

1.2 Selected papers representative of the hydrodynamic and the morphologic approach to meandering processes in rivers ...... 7

4.1 Summary statistics for pool depth and pool to pool distance ...... 46

4.2 Results of Independent t-test used to compare pool spacing and depth in the alluvial valley and the deltaic plain ...... 47

4.3 Results of Paired Samples t-test between pool depth and pool to pool distance to detect if there was a significant temporal change between the 1880 and 1911 surveys ...... 49

4.4 Summary statistics for sinuosity, radius of curvature, and channel width ...... 51

4.5 Results of Paired Samples t-test to examine change over time in sinuosity, radius of curvature, and channel width ...... 52

4.6 Summary statistics for average migration rates (m/km/yr) and maximum migration rates of meander bends (m/yr) ...... 61

5.1 Correlation matrix of channel variables and controlling variables, Cairo to Head of Passes ...... 65

5.2 Correlation matrix of channel variables and controlling variables, alluvial valley...... 66

5.3 Correlation matrix of channel variables and controlling variables, deltaic plain ...... 67

5.4 Summary data for valley slope sections ...... 71

5.5 Results of simple and multiple regression models for estimating channel migration ...... 86

6.1 Summary statistics for pool spacing and channel widths ...... 95

6.2 Results of Independent t-test for channel widths of pools and riffles ...... 96

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.3 Paired t-test of pools and riffles...... 97

6.4 Independent t-tests of channel widths of pools and riffles ...... 103

6.5 Pearson correlation coefficients curvature and pool parameters ...... 116

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES

Figures

1.1 Theorized magnitude of influence of human modifications to the Lower Mississippi River system...... 4

1.2 Characteristics representation of planform, cross-sectional, and Longitudinal channel dimensions with their suggested form-relationships. 9

1.3 Form-process relationship between meander bend migration rates and channel curvature ...... 12

1.4 Classification of channel pattern based on the dominant mode of sediment transport ...... 14

2.1 Mississippi River drainage basin ...... 18

2.2 Average discharge (cms) for mean low and high stages, and for the 1927 flood, the largest flood on record ...... 20

2.3 Total annual suspended sediment load for the Lower Mississippi River below New Orleans ...... 21

2.4 Distribution and proportion of bed material for die Lower Mississippi River prior to the change in sediment regime ...... 23

2.5 Quaternary geology of the Lower Mississippi River Valley...... 25

2.6 Cross-section of the Lower Mississippi River Valley...... 27

2.7 Depositional environments of a Lower Mississippi River meander belt ridge ...... 29

3.1 Meander bend segment for 1880 and 1911 showing methods of determining migration rates...... 36

3.2 1932 bed material, da* (mm) ...... 40

4.1 1880 thalweg elevation (MASL) of the Lower Mississippi River ...... 44

4.2 Low Water Channel width for 1880,1911, and percent change ...... 50

4.3 Radius of curvature for 1880,1911, and percent change ...... 54

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.4 Sinuosity for 1880, 1911, and percent change ...... 55

4.5 1880 and 1911 surveys showing superimposed channels ...... 58

4.6 Average migration (m/km/yr) for channel segments...... 59

4.7 Maximum migration (m/yr) for meander bends ...... 60

5.1 Change in radius o f curvature between 1880 and 1911 ...... 64

5.2 Profile of elevation sampled at 2 mi intervals along the Lower Mississippi River meander belt ridge ...... 70

5.3 Percent change in sinuosity, radius of curvature, and width ...... 73

5.4 Relationship between valley slope and sinuosity ...... 75

5.5 Relationship between valley slope and radius of curvature ...... 76

5.6 Relationship between valley slope and width ...... 77

5.7 Relationship between bed material and sinuosity ...... 79

5.8 Relationship between bed material and radius of curvature ...... 80

5.9 Relationship between bed material and width ...... 81

5.10 Relationship between maximum migration of meander bends with channel curvature, Tc/W ...... 82

6.1 Location of pools and riffles along channel thalweg, 1-600 km below Cairo...... 91

6.2 Location of pools and riffles along channel thalweg, 600-1200 km below Cairo...... 92

6.3 Location of pools and riffles along channel thalweg, 1200 to Head of Passes ...... 93

6.4 Frequency of pool spacing by channel widths, Cairo to Head of Passes ...... 99

6.5 Frequency of pool spacing by channel widths, Cairo to Red River Landing ...... 100

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.6 Pool spacing by channel widths, Red River Landing to Head of Passes ...... 101

6.7 Relationship between distance between pools and channel widths, Cairo to Head of Passes ...... 105

6.8 Relationship between distance between pools and channel widths, Cairo to Red River Tending ...... 106

6.9 Relationship between distance between pools and channel widths, Red River Landing to Head of Passes ...... 107

6.10 Relationship between pool depth and channel widths, Cairo to Head of Passes ...... 108

6.11 Relationship between pool depth and channel widths, Cairo to Red River Landing ...... 109

6.12 Relationship between pool depth and channel widths, Red River Tending to Head of Passes ...... 110

6.13 Relationship between valley slope and distance between pools ...... 112

6.14 Relationship between valley slope and pool depth ...... 114

6.15 Relationship between the distance between pools and pool depth ...... 115

7.1 Control of channel activity by clay plugs ...... 123

7.2 Long-profile of bedrock and older sediments in the Lower Valley ...... 125

xii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT

Although the Lower Mississippi River has been intensively studied, few

studies have investigated meandering processes in the river when the channel was

considered to be relatively free of major human disturbance. This research is

significant in that it provides a detailed examination of meandering processes and

channel adjustment in a large fine-grained alluvial setting, offering a departure from

the base of knowledge on this topic that has evolved from studies on smaller coarse

grained river systems. The study spans the entire length of the Lower Mississippi

River (1,704 km), from Cairo, IL to Head of Passes, LA. Two sets of hydrographic

surveys (scale, 1:20,00) completed between 1877 and 1922 provide the primary source

of data. Each survey was segmented into individual meander bends and straight

reaches, digitized, and entered into a G1S for analysis of channel processes. Several

techniques were employed to examine spatial and temporal change, and

interrelationships between channel parameters and controlling factors, including;

simple and multiple linear regression analysis, Independent and Paired t-tests, and

Pearson correlation coefficients.

Two major meandering regimes are identified, which coincide with the

boundaries of the alluvial valley and deltaic plain. Channel morphology in the alluvial

valley is characterized by its planform morphology having a sinuous laterally

migrating channel. Although the cohesive sediments in the deltaic plain reduce lateral

migration, the river is able to scour the channel bed into a uniform pool and riffle

morphology. Migration rates for channel segments in the alluvial valley and deltaic

xiii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. plain averaged 26.7 and 3.3 m/km/yr, respectively. Analysis of meander bend

migration rates with the adjustment of channel curvature between the two surveys

suggests an equilibrium meander bend curvature, rc/W, between 3.0 and 4.0. Spatial

trends in sinuosity, channel width, and radius of curvature do not conform to common

downstream patterns due to the lack of a downstream trend in discharge. The

planform morphology in the alluvial valley increased in sinuosity and decreased in

radius of curvature, as meander bends became increasingly arcuate. However, channel

width and the pool and riffle morphology did not significantly change during the study

period.

xiv

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 1 INTRODUCTION

1.1 OVERVIEW AND RESEARCH PROBLEM

Alluvial rivers attempt to adjust their channel morphology towards an

equilibrium profile (Leopold et al. 1964). The actual mode of adjustment may vary in

response to a number of controlling variables (Table 1.1). The study of meandering

processes of alluvial river systems has long been an important topic of investigation

within fluvial geomorphology (Petts 1995). In general, two major approaches have

been employed; a hydrodynamic approach which seeks to relate channel form to their

causal mechanisms (e.g. Langbein and Leopold 1966; Yang 1971; Bathurst 1979;

Bridge 1984; Dietrich 1987), and a morphologic approach which often relies on

statistics to examine relationships between various indices of channel form and

controlling variables (e.g. Leopold and Wolman 1957; Brice 1973; Schumm 1963;

Ferguson 1975; Williams 1986). However, in spite of advances in recent decades

linking form and process in meandering rivers, many questions remain concerning the

adjustment of channel morphology in an actively migrating river (Petts 1995).

Although the planform morphology of meandering rivers has received a great

deal of attention, fewer studies have examined their interrelationships with the

adjustment of the cross-sectional and longitudinal channel morphology in the context

of a migrating channel. This may be due to the difficulty in acquiring adequate data

sets for large rivers owing to their longer history of disturbance, and the substantial

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1.1. Significant variables influencing channel morphology of the Lower Mississippi River during modem timeloMO2 (after Schumm and Lichty 1965).

Param eter Description Significance to Lower Mississippi River Processes Valley Dimensions Slope o f modem surface Major determinant of system energy, local variations may result in channel pattern change Depth to older surface Possible contact with bedrock or older sediments Width Determines width of zone of active migration and floodplain reworking, amplitude of meander bends Neotectonics Uplift domes, fault zones Alter energy gradient of valley profile, uplift bedrock or resistant sediments into channel, bank caving Older Valley Deposits Grain size, cohesiveness Influence on sediment regime and channel processes Floodplain Deposits Point-bar (coarse-grained) Erosive, active channel changes Natural levee (coarse-grained) Erosive, active channel changes Clay plugs (fine-grained, Resistant, retards lateral migration, cohesive) asymmetric meander bend patterns Backswamp (fine-grained, Resistant, retards lateral migration, cohesive) non-uniform channel patterns Deltaic Plain Deltaic deposits (fine-grained, Lack of erosion and development of cohesive) meander bends, straight channel Sediment Regime Fine-grained Channel narrowing, reduced lateral migration Coarse-grained Wide channel, increase in lateral migration Discharge Smooth hydrograph, seasonal Bank caving events during receding flooding hydrograph Human Agency Dams and reservoirs, Iand-use Change to discharge and sediment changes in upper basin regime

Engineering structures on Revetments restrict lateral migration, main-stem channel changes cut-offs shorten channel length and increase channel slope, levees contain flood waters and sediment within the channel

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. resources necessary to survey the channel for different time periods. The majority of

studies have examined coarse grained or gravel bed channels, thus providing a bias in

sedimentary setting. This is significant given the importance of sedimentary controls

on fluvial processes (Dietrich 1987). Additionally, large river systems have been

neglected due to the majority of studies having focused on small river systems and

laboratory flumes (Petts 1995). This represents a significant gap in the literature

(Dietrich 1991; Hooke 1995; Kesel et al. 1992), as the complexity of response between

causal mechanisms increases with spatial scale (Schumm 1991; Lawler 1993). Thus

the Lower Mississippi River offers an exception to many of the systems commonly

reported in the literature which form the base of knowledge on this topic, in that it is

much larger and has a relatively fine grained sediment regime.

The Lower Mississippi River is considered a classic example of a meandering

alluvial river (Fisk 1944), and represents an ideal setting in which to examine a

number of issues related to meandering processes and channel adjustment However,

the modem channel represents one of the most intensively regulated rivers in the

world. Since the early-1900s the system has undergone significant modification in the

form of main-stem channel changes and a reduction in sediment supply (Figure 1.1).

While the changes have been well documented, few studies have examined

meandering processes prior to the introduction of these changes to the system. The

majority of studies have examined the river following the major engineering works of

the mid-1900s (e.g. Anding 1970; Assifi 1966; Hannan 1969; Winkley 1977), and

have tended to focus on specific reaches of the river rather than considering the river

as a dynamic system. However, two series of hydrographic surveys (scale 1:20,000)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DEVELOPMENT

NEW MADRID E.O. NEOTECTONICS DELTA PROGRADATION LAND USE CHANGE LEVEES DIKES 5 4 TMeom&D HHATtVt MACMTUOa REVETMENTS (UPRIVER) 3 or oevttofMCNTs DAMS 2 DREDGING CONTINUING IMPACT GRAVEL MINING — >> (?) ouesTiONAuc toNcevnv CUTOfFS O f IMPACT OLD RIVER DISTRIBUTARY CONTROL STRUCTURE

1800

Figure 1.1. Theorized magnitude of influence of human modifications to the Lower Mississippi River system. Major developments effecting the fluvial system during the study period, indicated by the dotted vertical lines, were land use changes in the upper basin. The majority of large-scale engineering works occurred after 1930 (modified from Kesel 1988).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. published by the Mississippi River Commission during the late 1800s and early 1900s

document the morphologic characteristics of the natural channel regime prior to most

major human modifications. Soundings taken at approximately 250 m intervals along

the channel allow major bedforms and channel dimensions to be measured. By

employing a Geographic Information System (GIS), the two surveys may be

superimposed for analysis of channel adjustment during the period of this study.

1.2 PURPOSE

The purpose of this research is to contribute to the understanding of

meandering processes in the Lower Mississippi River by examining the

interrelationships between the adjustment of channel morphology with

sedimentological and valley controls, and to gain an understanding of the natural

channel regime prior to major human disturbance.

This research is significant in two major areas. It provides a detailed

examination of meandering processes and channel adjustment in a large fine-grained

alluvial setting, thus offering a departure from the base of knowledge on this topic that

has evolved from studies on smaller coarse-grained river systems. Secondly, it

provides a base-line understanding of the natural channel regime of the Lower

Mississippi River, which is essential for establishing the changes that have occurred

due to human modifications.

1.3 RESEARCH OBJECTIVES

The objectives of this study are directed at gaining a better understanding of

meandering processes in the natural channel regime of the Lower Mississippi River.

The primary objectives are to:

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1) Characterize the channel morphology of the Lower Mississippi River prior to the

major engineering projects of the early to mid-1900s;

2) Examine the interrelationships between planform pattern and migration rates with

the cross-sectional and longitudinal channel morphology;

3) Determine the influence of the sediment regime and valley controls on meandering

processes in the Lower Mississippi River;

4) Contribute to the body of literature on meandering processes in large alluvial

settings.

1.4 APPROACHES TO MEANDERING PROCESSES

The hydrodynamic and morphologic approaches to meandering processes differ

fundamentally in their assumptions and perspective (Table 1.2). The morphologic

approach may be advantageous in the study of historic channel change for several

reasons. The hydrodynamic approach to meandering processes is heavily grounded in

the principles of physics, and many of the advances in this frontier have occurred

within laboratory settings using hydraulic flumes. Often fluid and sediment transport

mechanics are employed in equations that seek to predict channel form and adjustment

in response to variations in energy expenditure and boundary conditions. From the

hydrodynamic perspective, the adjustment of channel morphology towards an

equilibrium profile occurs so that rivers minimize their variance of energy expenditure

(Langbein and Leopold 1966; Yang 1971). Each adjustment effects channel form

resistance, and uniquely controls the dissipation of energy within the channel (Leopold

and Wolman 1960; Maddock 1969; Richards 1982; Knighton 1998). Rigorous data

assumptions and the deductive nature of this approach make the issue of transcending

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1.2. Selected papers representative of the hydrodynamic and the morphologic approach to meandering processes in rivers.

HYDRODYNAMIC APPROACH Author/s) Setting Topic of Research Leopold and Range of settings Pattern of meandering rivers related to theory Wolman 1960 of flow mechanics Tanner 1960 Inclined glass plate Meandering related to helicoidal flow and turbulence Einstein and Shen Flume, University of California, Meandering pattern related to upper flow 1964 Berkeley, CA, U.S. regime Langbein and Numerous rivers in Wyoming, Meanders form to reduce variance of energy Leopold 1966 U.S. expenditure Yang 1971 Popo Agie River, Hudson, WY, Meander forms and pool riffle morphology secondary sources related to rate o f energy expenditure Bridge and Jarvis River South Esk, Glen Cova, Bed topography of a migrating meander bend 1976 Scotland related to flow and sediment transport mechanics Bathurst 1979 Upper River Severn, Wales; Distribution of shear stress within a meander Upper River Swale, Britain; bend Colorado irrigation canals Dietrich and Muddy Creek, Wyoming, U.S. Channel meandering and equilibrium related Smith 1984 to flow and sediment transport mechanics Thompson 1985 River Skirden Beck, northwest Secondary flow structures of pool and riffle Britain morphology related to development of meandering pattern S. Dceda 1989 Lower Wabash River, IL and IN, Lateral sorting of bed material in meander U.S. bends related to channel slope MORPHOLOGIC APPROACH Leopold and Rivers from different settings, Continuum o f channel patterns between Wolman 1957 flume experiments braided, meandering, and straight rivers related to discharge and channel slope Schumm 1960, Midwestern U.S. Sinuosity and W/D related to % silt-clay in 1963 bank material Chang and Wabash River, White River, Radius o f curvature and channel width related Toebes 1970a Indiana, U.S. to channel slope, discharge, and regional geology Brice 1974 Secondary data, U.S. Planform geometry analysis of meander bend evolution and classification Hickin and Beatton River, northeast British Migration rates related to channel curvature Nanson 1975 Columbia, Canada Hooke 1977 River Culm, River Axe, Britain Planform pattern and migration rates Williams 1986 Range of settings Meander geometry related to mean and bankfiill channel area and width and mean depth Rosgen 1994 450 rivers in a range of settings in River classification based on sinuosity, U.S., Canada, and New Zealand entrenchment, W/D, and slope van den Berg Rivers spanning a range of scale Channel planform pattern related to bed 1995 and settings material and valley slope Harbor 1998 Sevier River, Utah, U.S. Channel adjustment to changing valley slope

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. scale and setting problematic (Baker 1988; Richards 1982; Petts 1995). In contrast,

the morphologic approach emphasizes an empirical perspective, employing statistics to

analyze morphologic parameters. Findings often lead to morphologic relationships

between various channel parameters. Most of the significant advances on the topic of

channel adjustment in meandering systems have utilized this approach. Although,

during the past couple of decades with the advent of more sophisticated

instrumentation, studies combining both approaches have yielded promising results

(e.g. Richards 1976; Bridge 1977; Thompson 1986; Dietrich and Gallinatti 1991;

Clifford and Richards 1992).

1.5 CHANNEL PATTERN AND ADJUSTMENT

There is considered to be a strong relationship between the planform pattern of

rivers and the cross-sectional and longitudinal channel morphology (Richards 1982).

This is often represented in the literature by specific channel parameters commonly

being associated with meander form (Figure 1.2) (Leopold and Wolman 1957;

Ferguson 1975; Chang and Toebes 1970a; Keller and Melhom 1978; Williams 1986).

For example, the pool and riffle sequence is a fundamental component of the

longitudinal perspective (Richards 1976), and is often associated with specific features

of a river’s planform morphology, with pools located downstream of the apex of

meander bends and riffles found in reaches between meander bends (Leopold et al.

1964; Keller 1972; Keller and Melhom 1978; Harvey 1975; Thompson 1986). Closely

related to this are characteristic cross-sectional channel forms that are related to

meander bends and straight reaches. Cross-sectional channel morphology typically

displays asymmetry and lower width to depth ratios at meander apices, and channel

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Planform

Cross-sectional

.Channel slope

Longitudinal

Riffle Riffl. Pool Pool

Channel distance

Figure 1.2. Characteristic representation of planform, cross-sectional, and longitudinal channel dimensions with their suggested form-relationships.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. symmetry with higher width to depth ratios at riffles located in straight reaches

between meander bends (Dietrich and Smith 1983; Richards 1978; Richards 1982;

Konditerova and Popov 1966; Milne 1979; Knighton 1981, 1982). While the most

recognizable adjustment of a meandering river channel is in the planform dimension, a

meandering river also adjusts its channel in the cross-sectional and longitudinal

dimension (Langbein and Leopold 1966; Yang 1971). However, these dimensions of

channel morphology are usually not considered with studies of meandering processes

(Hooke and Harvey 1983; Hooke 1984; Howard 1987). In order to gain a

comprehensive understanding of channel processes in meandering river systems, it is

necessary to consider the adjustment of the longitudinal and cross-sectional channel

geometry within the context of a laterallymigrating river.

A number of studies have shown that there may be a continuum of channel

patterns which are related to a combination of controlling variables (Leopold and

Wolman 1957; Schumm 1977; Mosley 1987; Knighton and Nanson 1993; Rosgen

1994; van den Berg 1995; Alabyan and Chalov 1998). Variables used to discriminate

between channel patterns include; channel slope-discharge (Lane 1957; Leopold and

Wolman 1957), stream power (Schumm and Kahn 1972; McEwen 1994; van den Berg

1995; Alabyan and Chalov 1998), and the sedimentary characteristics of the channel

perimeter (Carson 1984; Schumm 1960, 1963; Jionxian 1991; Ebisemiju 1994). The

implication of this approach is that a change in the independent variable will result in

an adjustment in form towards a different channel pattern. Of issue is the magnitude

of change which may occur when a critical threshold is crossed (Schumm 1977). For

example, a slight increase in channel slope may result in a disproportionate change in

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. channel pattern if the system is close to a threshold (Schumm and Kahn 1972;

Schumm 1980; Winkley 1977). More recent studies have refuted this and suggest that

abrupt thresholds may not exist. Instead, rivers may experience a transition in form as

a series of smaller channel adjustments are made to accommodate the change in the

driving variables (Ferguson 1987; Rosgen 1994; van den Berg 1995). This may be

scale related, as the size and complexity of a system increases the sensitivity of a

system to change when a single variable is modified may be reduced (Schumm 1991).

In spite of the advances made in recent decades, there lacks a comprehensive model

which relates channel pattern and change to the independent variables of discharge,

valley slope, and sediment type (Knighton 1987; Schumm et al. 1994).

Channel patterns and rates of meandering may vary widely over the course of a

stream channel (Hooke 1977, 1980, 1995; Laczay 1977) and within individual

meander bends (Hooke and Harvey 1983; Ferguson 1977), so that rivers may vary

considerably in meander geometry (Brice 1974; Ferguson 1975). Channel curvature,

defined as the ratio of the radius of curvature (rc) to channel width (W), is considered

an important control on migration rates due to its influence on the distribution of

boundary shear stress (Callander 1978; Hickin and Nanson 1975; Bathurst 1979).

Rates of lateral migration have often been shown to increase where tJ w values range

from 2.0 to 3.0 (Figure 1.3). Keller (1972) developed a five-stage model for alluvial

rivers which relates meander migration with pool and riffle morphology. Pool and

riffle units developed as meander bends increased in length, which maintained pool

spacing at a constant 5 to 7 channel widths. Studies by Hooke and Harvey (1983) and

Thompson (1986) have substantially contributed to the comprehension of the causal

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a> A 2 a 0 100 s

_L_L_L 0 1 2 3 4 5 Meander bend curvature, r/W

Figure 1.3. Form-process relationship between meander bend migration rates and channel curvature, r /W . Maximum migration rates are associated with r/W of 2.0 to 3.0 (modified from Hickin 1974).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. mechanisms of this occurrence by emphasizing the role of flow dynamics as a linkage

between pool and riffle morphology and meander bend migration. Beyond a critical

path length of approximately 10 channel widths a new riffle unit is added, leading to

an increase in erosion and eventually to the development of a new meander bend.

While a river may vary considerably in meander pattern and rate of lateral

migration, a single meander bend can be seen as having an orderly pattern of erosion

and deposition. As .such, meander bends can be seen as dynamic sediment transfer

systems, involving erosion on the cut-bank and deposition on the inside of the channel

(Knighton 1998; H. Ikeda 1989). Sediments comprising the channel perimeter are of

importance because of their influence on channel form (W/D) and rate of lateral

migration (Friedkin 1945; Fisk 1952; Schumm I960, 1963; Rohrer 1984). Leopold

and Maddock (1953) suggest a wide channel (high W/D) is necessary where a river

transports a high proportion of its sediment as bed load. Schumm (1977) derived a

classification of alluvial channels based on the mode of sediment transport (suspended

load, mixed load, and bed load) and channel pattern. The shape (W/D), slope, and

stability of the stream channel are seen as related to stream power and fire amount and

size of sediment transport (Figure 1.4). The general consensus of the literature has

shown that as the percentage of silt and clay in the bed and bank material increases,

lateral migration decreases, W/D decreases, and channel stability increases (Alexander

and Nunnally 1972; Ebisimiju 1994; Fisk 1952; Kolb 1963; Hooke 1980; H. Ikeda

1989; Jiongxin 1991; Miller and Onesti 1979; Schumm 1960, 1963).

Although meandering processes involve interactions among various

dimensions of the channel morphology (Richards 1982), it is necessary to consider the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHANNEL TYPE Suspended Load Mixed Load Bed Load

O o X * C c c . c - coZ *o S < 2 B art £O 3 > m H < HIGH RELATIVE STABILITY LOW ( 3% >) Low-— Bed Load/Total Load Ratio -— High ( >11%) Sm all ■ 1 Sediment S ite ------► Large Smoll *•------Sediment Lood Large Low —------Flow Velocity — High Low Stream Power - High

Figure 1.4. Classification of channel pattern based on the dominant mode of sediment transport. The adjustment of channel morphology is considered with respect to a number of variables (modified from Schumm 1977).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. river system within the context of the geologic setting (Schumm et al. 1994). The

sediment and discharge regime of an alluvial river are often considered the dominant

controls on channel form (Leopold and Maddock 1953; Leopold and Wolman 1957).

However, valley controls, such as older sediments and bedrock forming the valley

perimeter, neotectonics, and valley slope also have significant control on channel form

and process over varying temporal and spatial scales (Adams 1980; Gregory and

Schumm 1987; Bumett and Schumm 1983; Fisk 1944; Winkley 1994; Saucier 1994;

Orlowski et al. 1995; Harbor 1998). This is important because variations in these

controls may account for downstream spatial irregularities in channel pattern. For

example, when an alluvial river encounters an increase in valley slope (increase in

energy) the river may adjust its channel in several ways. The mode of adjustment may

vary, but can include; lateral migration and an increase in sinuosity (Schumm and

Kahn 1972; Hooke and Harvey 1983), channel incision resulting in a decrease in

channel slope, an increase in width to depth ratio in the cross-sectional dimension, or

by increasing resistance in the longitudinal dimension by altering the pool and riffle

morphology (Knighton 1998).

1.6 RESEARCH DESIGN

This study adapts a morphologic approach to examine meandering processes

and channel morphology in the Lower Mississippi River over a 31 year span of time

during the late 1800s and early 1900s. Two series of hydrographic surveys (scale

1:20,000) competed by the Mississippi River Commission provides the primary source

of data. The surveys are superimposed in a GIS, which allow channel segments

consisting of both meander bends and straight reaches to be delineated. Sediment data

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and morphologic indices characterizing each channel segment are compiled in a

database for analysis of correlative and causal relationships between various indices of

channel form and change over time.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 2

PHYSICAL SETTING

2.1 DISCHARGE AND SEDIMENT REGIME

Draining a variety of climatic and geologic settings, the Mississippi River

drains 3.23xl06 km2, making it the largest drainage basin in North America (Figure

2.1). The Lower Mississippi River begins at the confluence of the Middle Mississippi

and Ohio Rivers, at Cairo, IL. Prior to the cut-offs of the 1930s, the river flowed over

1700 km to the Gulf of Mexico, where it discharged its sediment laden waters through

multiple bifurcating distributary channels at Head of Passes, LA. Five drainage basins

contribute streamflow to the Lower Mississippi River, including the Missouri

(1.37xl06 km2) and Arkansas (0.49xl06 km2) River basins draining the arid and semi-

arid Great Plains and Rocky Mountains, and the Ohio (0.53x106 km2), Upper

Mississippi (0.45xl06 km2), and Lower Mississippi (0.15xl06 km2) River basins,

draining the Appalachian Plateau and the humid central and south-central U.S. The

average annual discharge for the Lower Mississippi River is 12,800 cms. Stage

variations decrease towards Head of Passes, LA due to lower channel gradients and the

influence of the Gulf of Mexico (Mossa 1988), as well as being further downstream

from the Ohio River which supplies 70% of the discharge to the Lower Mississippi

River. However, discharge varies little in a downstream direction, as there are few

tributaries of considerable size contributing flow to the Mississippi River below Cario.

The mean high water discharge at Columbus KY, representing the total inflow at the

head of the Lower Valley, was 35,700 cms, while the mean high water discharge at

Vicksburg was 34,160 cms. Below Old River distributary, at Red River Landing, the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

OHIO BASIN .15X 10'km1 LOWER MISS. - BASIN .53x10' km8 S t. Louis \ \ New Orleans C airo BASIN UPPER MISS.' c 45x10'km* Gulf Gulf of Mexico

OFTHE THE SIX MAJOR MAJOR SIX THE BASIN -

DRAINAGE BASINS .2 4 x 1 0 'km*

RED-OUACHITA BASIN MISSOURI

1.37x10® km* MISSISSIPPI RIVER WATERSHED BASIN - .49 x 10' kmJ ARKANSAS-WHITE

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. mean high discharge was 29,120 cms (Figure 2.2). Discharge characteristics are

markedly seasonal, with a high flow period lasting several months during the late

winter and spring, and a low flow period during the summer and fall.

Several large floods occurred in the basin during the study period, from 1877 to

1924. The largest was the flood of 1882, which occurred before the extensive

construction of levees (Elliot 1932) that currently align almost the entire channel. At

Red River Landing the flood was above bankfull stage for eighty days, and crested on

March 31st with a discharge of44,460 cms. The flood resulted in extensive damage in

the Lower Valley, and caused 284 crevasses (Elliot 1932). The largest flood on record

in the Mississippi basin occurred in 1927, with a record discharge of 49,812 cms

(Ferguson 1940). Flood frequency during the study period appears consistent with the

larger trend for the historic Mississippi River discharge regime. From 1799 to 1864

there were 24 floods recorded, while from 1865 to 1931 there were 23 floods reported.

Thus, based on a 132 year record, the Lower Mississippi flood every 2.8 years (Elliot

1932).

The time period examined in this study is considered representative of the

natural condition of the system, since the major engineering works were initiated

dming the 1930s (Kesel 1988; Schumm et al. 1994). Several studies report that the

Lower Mississippi River during this period annually transported approximately 400 x

106 tons of suspended sediment (Figure 2.3) (Humphreys and Abbot 1861; Elliot 1932;

Keown et al. 1986; Kesel 1989; Meade and Parker 1985). The river also transported a

large percentage of its sediment as bedload. Typically an alluvial river will transport

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 80,000 - ^ 70,000 - avg low avghigh 1 60,000 - 1927 i 50,000 - J 40,000 - •| 30,000 -

20,000 -

10,000 - 0 - >- 04 co < < i4 < s o f eo 00 a S a £ e o x> 0 x>§ "v CO 0 3 as cs o CO U -J CJ d -o 1 04

Figure 2.2. Average discharge (cms) for mean low and mean high stages, and for the 1927 flood, the largest flood on record in the basin (data source: Elliot 1932).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 600

700

600

o u 500

■o o •ow 400 a

e 300

200

100

1830 1840 1850 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 Y ear

Figure 2.3. Total annual suspended sediment load for the Lower Mississippi River below New Orleans based on data from Humphreys and Abbot 1851; Quinn 1879 to 1893; and New Orleans Water and Sewage Board 1930 to 1982 (modified from Kesel 1988).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10% of its total sediment discharge as bedload (Richards 1982). However, a sediment

budget by Kesel et al. (1992) suggests that prior to the engineering works of the early

1930s, the Lower Mississippi River transported approximately 30% of its sediment as

bedload. Much of the bed material in the Lower Mississippi River was supplied from

channel bank failure events and lateral channel migration (Kesel et al. 1992). During

this period the channel was free to migrate laterally within its floodplain, with only

seven percent of the channel aligned by revetments (Elliot 1932). Currently, artificial

levees, channel training dikes, and concrete revetments align ninety percent of the

channel from Cairo, IL to Head of Passes, LA, effectively channelizing the entire

Lower Mississippi River.

A survey completed in 1932 shows the distribution and size of bed material

along the length of the Lower Mississippi River (Figure 2.4). Despite the overall

downstream decline in bed material size, the Lower Mississippi River does not exhibit

a smooth dowstream trend like many large alluvial systems (Leopold et al. 1964). The

increase in coarse bed material associated with the Arkansas River is particularly

distinct Also, the occurrence of several coarse sediment peaks coinciding with

smaller tributaries suggests the importance of small drainage systems in supplying

coarse sediment to the Lower Mississippi River. At Cairo, EL, 53% of the bed material

input into the system occurs as coarse sand or gravel. Downstream of Red River

Landing the proportion of fine bed material substantially increases, as there are no

tributaries of significant size in the deltaic plain.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Gravel

250m m tSd (opotH-^ fttSDd oa>

o 3 3E c_>

900 800 700 600 500 400 300 200 100 River Miles, Above Head of Passes

Figure 2.4. Distribution and proportion of bed material for the Lower Mississippi River prior to the change in sediment regime (data source: U.S. Army Corps of Engineers 1935).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.2 GEOLOGIC SETTING

The Lower Mississippi River Valley extends in a southerly direction more than

780 km from Cairo, IL to Head of Passes, LA, where the mouth of the delta bifurcates

into multiple distributary channels and discharges its sediment laden waters into the

Gulf of Mexico near the edge of the Continental Shelf (Figure 2.5). Valley widths

average 120 km, but range from 40 km near Natchez, MS to 200 km in width in the

central portion of the basin across the Arkansas Lowlands (Fisk 1944). Tectonic

controls have effected the Lower Mississippi River Valley development over a range

of temporal and spatial scales. Structurally, two major features effect the valley (Fisk

1944). The Mississippi Structural Trough, also known as the Mississippi Embayment,

is a southerly plunging syncline that orients North American interior drainage towards

the Gulf Coast The Gulf Coast Geosyncline trends east to west with an axis parallel

to the Gulf Coast, and is the dominant structural feature in the Central Gulf Coastal

Plain. During the Cretaceous the valley developed in igneous, sedimentary, and

metamorphic rocks as old as Precambrian, but has since been infilled with thousands

of meters of Gulf Coastal Plain sediments which dip towards the coast (Autin et al.

1991). In addition to large-scale structural controls, growth faults in South Louisiana

and uplift domes located within the valley represent neotectonic controls at regional

scales (Fisk 1944; Burnett and Schumm 1983; Autin et al. 1991; Saucier 1994).

Anticlinal domes, such as the Lake County, Monroe, and WigginsUplift in the Lower

Mississippi Valley may effect Quaternary landscapes by warping terraces and

floodplain surfaces, as well as modem channel processes by increasing valley slope

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Holocene deposits

Wisconsin stage deposits I Pre-wisconsm -ve^lCnjro 1 quaternary deposits

Coastal plain margin

'm m m

/Memphis

TCH/£A

/CZ~- f~miOBU9^

^pffVicksburg yJackson

Natchez

ed_Rivcr ______MS S&^Landing LA M obile s.

Baton ougc

EtTAlCr i i 60 km

GULF OF MEXICO Modified from Saucier. 1994

Figure 2.5. Quaternary geology of the Lower Mississippi River Valley (modified from Saucier 1994).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and emplacing resistant bedrock close to the bed of the channel (Fisk 1944; Gregory

and Schumm 1987).

The Quaternary valley formed in response to changes in sea level related to

episodes of continental glaciation during the Pleistocene (Fisk 1944). With lowering

sea level (base level) and increasing valley slope the river scoured and incised its

valley (Figure 2.6). Rising sea levels, due to melting of continental glaciers, resulted

in valley infilling due to a decrease in slope and sufficient energy for the river to

transport its large sediment load. Valley widening occurred during interglacial

periods, when the river developed a meandering system. This caused the deposition of

a thick clastic sedimentary wedge, increasing in thickness towards the Gulf of Mexico.

The depth of Holocene alluvium to the Pleistocene entrenched valley floor at Cairo, IL

and the southern extent of the valley is approximately 69 m and 180 m, respectively.

Holocene valley sediments have a classic fluvial fining upward sequence, grading

from a coarse gravel substratum representative of a braided stream system, to a

topstratum of fine sands, silts and clays representative of a meandering stream system

(Figure 2.6) (Fisk 1944; Saucier 1974; 1994). Recently this has been questioned by

Aslan (1996), who suggests that the early Holocene fluvial system may have been

more representative of a lacustrine setting similar to the Atchafalaya system, having

numerous lakes with interconnecting channels.

2.3 ALLUVIAL FLOODPLAIN

From Cairo, IL the river meanders for 1230 km through the alluvial floodplain,

where it intersects the deltaic plain at Red River Landing at the confluence with Old

River distributary (Russell et al. 1936; Fisk 1947; Saucier 1994). Old River, which

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MODERN Block Diagram of MEANDER BELT Lower Mississippi MISSISSIPPI RIDGE Alluvial Valley

BACKSWAMP f t S

EARLIER DEPOSITS* ’(substratum _ ^ o ZONE OF PROFILE OF RIVER MIGRATION LATE WISCONSIN WITHIN MEANDER BELT ENTRENCHED VALLEY SYSTEM

(Modified from fisk. 1951: Sw der. 1994)

Figure 2.6. Cross-section of the Lower Mississippi River Valley showing the incised Pleistocene valley floor overlain by Holocene alluvium. The Mississippi River flows within a meander belt ridge, which is elevated above the level of the marginal backswamp deposits (modified from Fisk 1951; Saucier 1994).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. functions as a Mississippi River distributary and a tributary to the Atchafalaya River,

flows within a former Mississippi River meander bend that was cut-off in 1831 by

Colonial Schreve (Elliot 1932.). There are two major sedimentary environments

within the Holocene alluvial floodplain, meander belt ridges and backswamps (Fisk

1947). Throughout its course the river flows within a meander belt elevated above

backswamp deposits (Figure 2.7). Formed due to lateral channel migration and

overbank deposition of flood deposits, the meander belt forms a ridge of coarse

sediments, and generally decreases in width and height towards the coast (Kolb 1963).

Within the meander belt there are three main fluvial sedimentary environments which

are significant to channel processes; natural levees, point bars, and clay plugs.

Natural levees are formed due to over bank flooding and the deposition of

coarse-grained sediments along the banks of the channel. This results in sediments

grading in thickness and texture, from silty-sands to silty-clays, away from the channel

towards the backswamp lowlands. Width and height dimensions of natural levees vary

along the length of the river, and are chiefly controlled by the sediments and height of

flood events. Although spatial variations occur, natural levee deposits tend to be finer

grained then point bar deposits, and decrease in grain size from sandy silts to silty

clays and clays in a downstream direction (Fisk 1947).

Point bar deposits are the most characteristic deposits of a meandering stream,

and are formed due to lateral migration of meander bends. Deposits have a fining

upward sequence, grading from basal coarse sands and gravel to silty sands. This type

of accretion topography is orientated parallel to the direction of channel migration, and

has an undulating pattern of topography termed ridge and swale, or, meander scroll.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Principal Depostitional Environments DISTRIBUTARY of a Meander Belt Ridge

b a c k s w a m p;

NATURAL LEVEE \

ABANDONED CHANNEL/ CLAY PLUG MEANDER/ .; ;;SCR0LL>’ : POINT BAR = EARLIER; CREVASSE DEPOSITS'; SPLAY

(Modified from Saucier. 1994)

Figure 2.7. Depositional environments of a Lower Mississippi River meander belt ridge. Major depositional units significant to channel processes include infilled abandoned channels (clay plugs), point-bar deposits, and natural levee deposits (modified from Saucier 1994).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Point bar deposits are larger and more extensive in the upper and central portions of

the valley, and occur in only a few bends below Red River Landing (Kolb 1963).

Clay plugs, or infilled sloughs, are ox-bow lakes or abandoned channels that

have infilled with fine-grained sediments during overbank flood events. Clay plugs

are important in that they greatly restrict lateral channel migration (Fisk 1947;

Krinitzsky 1965). Deposits are extensive in portions of the valley where channel

migration is most active, and do not appear below Donaldsonville. The thickest and

most resistant clay plugs are formed due to infilling of ox-bow lakes formed from

meander neck cutoffs. Chute cutoffs are located closer to the main channel and take

longer to occur then neck cutoffs, resulting in lower portions of the channel being

filled with coarser sediments which are buried by fine-grained silts and clay (Fisk

1944).

Backswamp are the second major depositional environment occurring within

the alluvial valley. Backswamp deposits are formed as overbank flooding transports

fine-grained sediments, primarily clay, to lowlands beyond the meander belt ridge.

They accumulate in thick deposits and form lowland swamps marginal to meander

belts. The thickness of backswamp deposits increases in a down valley direction. At

Mellwood, AR backswamp deposits occur to a depth of 9 m, but increases to 43 m in

the vicinity of White Castle, LA (Fisk 1947). Deposits are particularly extensive in the

Yazoo, Tensas, Beouf, and the southern extent of the St. Francis basin (Fisk 1947), but

do not occur north of Memphis.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.4 DELTAIC PLAIN

From Red River Landing the river flows 570 km through the deltaic plain to

the mouth of the delta at Head of Passes, LA. The deltaic plain formed as a result of

the Mississippi River shifting its center of deposition. Rising sea levels resulted in the

formation of four major abandoned delta complexes, the Maringouin, Teche, St.

Bernard, La Fourche, and two active deltas, the Plaquemine-Balize, or, modem "birds-

foot" delta, and the Atchafelaya (Fisk 1944; Frazier 1967). Deltaic deposits consist of

fine-grained clayey sediments which increase in thickness towards the Gulf of Mexico,

and are often buried beneath backswamps deposits between meander belt ridges and

interdistributary bays (Saucier 1994). Due to the cohesive nature of deltaic sediments,

rates of channel migration are particularly low (Kolb 1963).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 3

DATA AND METHODS

3.1 HYDROGRAPHIC SURVEYS

This study makes extensive use of historic data from the Mississippi River

Commission (MRC). Upon creation of the MRC by the U.S. Congress in 1879, the

MRC directed its energies to completing accurate topographic maps and surveys of the

river, which had been initiated several years earlier by the U.S. Lakes Commission and

the U.S. Coastal and Geodetic Survey. This resulted in a series of hydrographic

surveys that span the entire length of the Lower Mississippi River, from Cairo, IL to

Head of Passes, LA. The first set of hydrographic surveys, referred to as the 1880

survey, includes 84 maps (scale 1:20,000) and was surveyed between 1877 and 1895.

From Cairo, EL to Red River Landing the river was surveyed from 1877 to 1883. From

Red River Landing to Head of Passes, LA the river was surveyed between 1894 and

1895. The second set of hydrographic surveys, referred to as the 1911 survey, also

includes 84 maps (scale 1:20,000), and was surveyed between 1911 and 1924. From

Cairo, IL to Red River Tending the river was surveyed between 1911 and 1913, and

from Red River Tending to Head o f Passes between 1921 and 1924 (Elliot 1932).

There are on average 31 years between any two sections of the river channel.

Field surveying by the MRC was completed during low water stage, typically

from late fall to the rising stage of theannual spring flood. During the survey, stage

was measured three times a day and related to a local vertical datum. Topography

mapped included high water channel boundary, point bars, islands, towhead bars,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. vegetation, and all natural and cultural features adjacent to the channel. Hydrographic

work included soundings taken perpendicular to the channel and represents the low

water subaqueous channel profile. The 1880 survey was sounded at 250 to 400 m

intervals the length of the channel, while the 1911 survey was sounded at 250 m

intervals (Elliot 1932). Each sounding traverse typically recorded from 10 to 20

channel depths.

Although the hydrographic surveys were mapped during low water stage, due

to the size of the river completion required several field seasons. Thus mapping and

survey data reflect seasonal variations in stage. A slight change in river stage can

cause significant differences in channel width, making it difficult to differentiate actual

geomorphological change from changes inchannel width caused by fluctuations in

stage. To ensure uniform comparison of the channel over time, it was necessary to

reduce both the 1880 and 1911 hydrographic surveys to a common stage. This was

done by using a 40 year Average Low Water Plane (ALWP), from 1891 to 1931,

which is based on the elevation profile of the river at average low water stage. By

superimposing the elevation profile of the 40 year ALWP on to both surveys, the low

water channel width is delineated. This negates the issue of stage variation, and

ensures that similar comparisons are made of the river for the two surveys.

3.2 GEOGRAPHIC INFORMATION SYSTEMS (GIS)

3.2.1 Conversion to Vector Format, Channel Segments

Measurement of morphologic parameters and assessment of change was greatly

facilitated by use of an Intergraph MGE GIS. Each map survey (84 maps for each

survey) was scanned to a raster image and imported into M G E G eoV ec, where a raster

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to vector conversion was accomplished using heads-up digitizing. Features which

were digitized included; high water channel boundary, low water boundary adjusted to

the 40 year (ALWP), islands, major chutes, tributary confluence, channel cutoffs, and

control points. Each of the vectorized maps was registered using control points from

the hydrographic surveys. This resulted in 168 individual files, which were then

merged to superimpose the channel for the two time periods.

The Lower Mississippi River was then segmented into 167 individual meander

bends and reaches, averaging 10.2 km in length, which provided the primary

morphologic units for this study. A channel centerline was digitized down the length

of the channel, consisting of a series of 0.5 mi (805.0 m) line segments, which served

as the base line for delineating meander bends. To qualify as a meander bend, the

segment must have at least 60 degrees of arc, based on the curvature of the channel

centerline (Brice 1974). The upstream and downstream inflection points of meander

bends were identified based on the angle difference of successive channel centerline

segments. This allowed for a pseudo-objective identification of meander bends.

Reaches were located between meander bends, and generally were straight or irregular

in pattern. The division of the river into morphologic segments allows each to be

considered individually with respect to channel migration and the adjustment of a

range of variables through time.

3.2.2 Error

Error is inherent in the transformation of data from the river channel, to map

surveys, to digital format (Downward 1995). However, due to the source of data and

procedures that were employed in this study, the degree of error is not likely a serious

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. problem with regards to geomorphological analysis. There are several reasons,

including: 1.) Survey control by the MRC was established by a dense triangulation

system that resulted in a level of planimetric accuracy uncommon for historic channel

surveys in the U.S. (Lewin 1977; Hickin 1983). Previous surveys of the Lower

Mississippi River do not have a level of accuracy suitable for detailed planimetric

analysis. 2.) The ideal map scale for river channel analysis is recommended at

1:24,000 (Goudie et al. 1990), so the MRC surveys at 1:20,000 represent an

appropriate scale to examine planform channel change. 3.) H ea d s-u p digitizing of a

raster image is a more accurate technique then digitizing from hard copy, as it ensures

that the operator can precisely follow the line work being digitized. 4.) Finally, the

size of the river channel is of importance. The inherent error in a map survey for a

given scale represents a smaller proportion of the overall error for large river channels

in comparison to smaller river channels. The potential error of the digitized channel

was calculated at ± 20.0 m. This was determined based on the thickness of the

mapped channel boundaries (0.5 mm) and the map scale (1:20,000). Thus the

potential error from raster to digital conversion for a channel width of 750 m would be

± 2 .7% .

3.3 PRIMARY VARIABLES

3.3.1 Migration Rates

Superimposing the channel during the two time periods allows for migration

rates to be measured (Figure 3.1). Using the channel centerline as a base line, straight

lines were placed at a perpendicular angle from successive channel boundaries, at 1.0

mi intervals along the length of the channel. This represents the amount of change

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1*311

1880

1 KM

Figure 3.1. Meander bend segment for the 1880 and 1911 survey. The solid outer line represents the 1911 survey, whereas the 1880 survey is the dashed line. Radius of curvature, rc (m), was measured along the channel centerline. Sinuosity, P, was measured by dividing the length of the channel, measured along the channel centerline, by the straight-line distance between beginning and end inflection points. Average channel migration (m/km/yr) was calculated by taking the sum of the lines connecting the two outer channel bank boundaries, and dividing this by the length of channel and the number of years between surveys. Maximum migration rates (m/yr) were determined by measuring the distance between channel boundaries where the greatest amount of change occurred, represented by the bold dashed line, and dividing the length of the line.by the number of years between surveys.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. between the two surveys. For each channel segment, the total length of lines between

the channel boundaries were summed, and divided by the number of lines per channel

segment. This value was then divided by the number of years between successive

surveys, to give a value representative of the average rate of channel migration per

kilometer per year (m/yr/km). Maximum rates of migration (m/yr) were calculated

only for those channel segments which classified as a meander bend, and was

determined by measuring the greatest distance between die two surveys, and dividing

that length by the number of years.

3.3.2 Channel Width

Low water channel width (W) was determined by measuring the channel

between the ALWP channel boundaries at 1.0 mi intervals the length of the channel

for both the 1880 and 1911 survey. Channel width was then averaged for each channel

segment to obtain a channel width representative of the larger reach. Channel width is

often considered an independent variable due to its relationship with discharge, and

has been shown as being related to channel migration and bed material.

3.3.3 Bankfull Channel Width

Rankfull channel width (Wbf) was provided from cross-sectional data

calculated and tabulated by the Mississippi River Commission. The subaqueous

channel profile of each crossing along the length of the river were plotted on linen

sheets (available in the MRC archives in Vicksburg, MS) and scaled to the average

high water stage. The width was then measured directly from the linen sheets. Due to

the cross-section originally having been sounded at low water stage, this does not

represent a true bankfull channel depth, as it does not take into account channel scour

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. that occurs during bankfull discharge conditions. Additionally, the data has not been

adjusted to the 1927 vertical datum, and as such may contain vertical error.

3.3.4 Radius of Curvature

Radius of curvature (rc) was measured by use of GIS. The channel centerline

was used as a guide for fitting arcs with a minimum of 60 degrees of curvature for

each meander bend (Brice 1974). Radius of curvature is scale dependent, and provides

an index of the size of a meander bend. Large bends with a low curvature have high

radii, whereas smaller bends having a low curvature have lower radii (Figure 3.1).

3.3.5 Sinuosity

Sinuosity (P) was determined for each channel segment for both surveys.

Sinuosity was calculated by dividing the channel distance by the straight-line distance

between the beginning and end inflection points of each channel segment (Hooke

1977). Sinuosity is an index of channel meandering, and can be used in assessing the

adjustment of planform geometry (Figure 3.1).

3.3.6 Thalweg, Pools and Riffles

Data was obtained by averaging the channel thalweg of each crossing at 1.0 mi

intervals for the length of the channel for both time periods. The thalweg profile

reveals the undulating pattern of pools and riffles. The identification of pools and

riffles was by use of O’Neill and Abrahams (1984) bedform differencing technique.

The technique employs an index based on the standard deviation of successive

differences in thalweg depths as a filter to objectively identify pools and riffles.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.3.7 Bed Material

Bed material for the study was obtained from a survey completed by the U.S.

Army Corps of Engineers in 1932. Although not taken at even intervals the length of

the channel, the survey sampled bed material along the channel thalweg on average

every 2.8 km from Cairo, IL to Head of Passes, LA. For this study, the dg 4 (mm)

parameter is used, which is the particle size larger then 84% of the sampled bed

material. This is considered representative of the basal layer of channel sediments,

which a number of researchers have related to channel pattern and bank erosion (e.g.

Krinitzsky 1965; Thome and Lewin 1979; Nanson and Hickin 1986; Jiongxin 1991).

Additionally, preliminary research by the author has shown that in the Lower

Mississippi River the dg4 parameter explains a greater amount of variability in channel

form and migration rates then the di6 or dso parameter. Figure 3.2 shows that the dg 4

parameter does not display a smooth downstream decline in grain size, but is

characterized by a number of coarse sediment spikes within the alluvial valley.

3.3.8 Valley Slope

Valley slope represents an energy parameter (Knighton 1998) and is considered

an independent variable with respect to channel pattern (Schumm and Lichty 1965;

van den Berg 1995). Data for determining valley slope (Sy) was obtained from a

topographic map (scale 1:500,000,5.0 ft contour interval) of the Alluvial Valley of the

Lower Mississippi River (1930) prepared by the U.S. Army Corps of Engineers

Memphis and New Orleans District To determine valley gradient, a long profile of

valley elevation was plotted (NAVD 1927) from data taken at 2.0 mi (3.2 km)

intervals along a valley transect Of issue is the surface that is used to represent valley

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. r - OQ £ 20

o o o o o o © o o o o o o o o o o o o o o o o o o o o o o o o o o o o CM VO oo ON rj VO KM below Cairo

Figure 3.2. 1932 bed material, d $ 4 (mm), for the Lower Mississippi River (data source: Nordin and Queen 1992).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. slope. The surface of the alluvial valley of the Lower Mississippi River is complex,

with features differing in age, origin, and elevation. This makes it essential that

elevation data be obtained from a consistent suface. Although backswamp deposits

would be ideal, there are few deposits above Helena, AR. For this reason the modem

meander belt ridge was chosen, as it represents the surface which the river is migrating

within. Data was sampled perpendicular to the transect, with care taken not to sample

where there was pronounced meander scroll topography. When contour lines differing

in elevation were encountered, the average elevation was calculated. A plot of the

elevation data by valley kilometers below Cairo revealed a number of shorter slope

segments within the valley profile. Eleven shorter slope sections were identified on

the basis of breaks in slope along the valley profile. The slope (m/km) of each valley

section was determined by using least squares regression (Clark and Hosking 1986),

and converted to dimensionless units (m/m).

3.4 ANALYSIS OF DATA

The primary variables characterizing the 167 channel segments were input into

a database for statistical analysis. In order to gain greater insight into correlative and

causal relationships, secondary variables were generated from primary variables. To

illustrate, the ratio between rc and W was used to create an index of channel curvature

1974). Two of the major thrusts of this research were to examine spatial differences

between the alluvial valley and deltaic plain, and temporal change during the 31 years

between the two surveys. This requires indices of change, which were created from

the primary variables. For example, the percentage change in sinuosity (P%)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (secondary variable) was created from sinuosity (P) (primary variable) from the 1880

and 1911 surveys.

Descriptive statistics were used to characterize spatial and temporal trends, and

to provide diagnostic insight into parametric model building. The Independent and

Paired Samples t-tests were employed to examine whether there were significant

spatial and temporal differences in the planform and pool and riffle morphology.

Pearson’s correlation coefficient (SPSS 1998) and both simple and multiple linear

regression analysis was employed to examine relationships between variables, and to

isolate controlling variables unique to channel migration (Clark and Hosking 1986).

The variables were normalized using a log10 -log10 transformation, which increased the

amount of variance explained by the models (Nanson and Hickin 1986).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 4

SPATIAL AND TEMPORAL TRENDS

4.1 OVERVIEW

This chapter describes spatial and temporal trends of the planform and

longitudinal channel morphology, and estimates channel migration rates determined

from the two surveys. The Independent t-test is used to detect whether there are

significant differences in channel morphology between the alluvial valley and deltaic

plain. Significant temporal change is assessed by employing the Paired Samples t-test,

which compares differences in means between each channel segment for the 1880 and

1911 surveys. The degree of similarity between a single variable compared between

the two surveys is indicated by Pearson’s correlation coefficient. Higher correlation

coefficients are associated with little change over time, whereas low correlation

coefficients indicate that there has been considerable change between the two surveys.

4.2 LONGITUDINAL PROFILE

Figure 4.1 is a profile of thalweg elevations (MASL) at 1.6 km intervals

(converted from 1 mi intervals) along the entire length of the Lower Mississippi River,

from Cairo, IL to Head of Passes, LA. While data taken at this interval is too coarse to

delineate bedforms, the pool and riffle morphology is clearly delineated by the

rhythmic undulating thalweg pattern. The trend is fairly linear within the alluvial

valley, but the rate of decline in thalweg elevation decreases in the deltaic plain,

downstream of Red River Landing.

To further analyze the thalweg profile, pools and riffles were identified using

O’Neil and Abrahams (1984) bedform differencing technique. The technique uses an

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

o o in o vo t"- o o o - tj o o o o o o rs i—i

o o o

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. index based on the standard deviation of successive thalweg elevation differences. By

employing the technique separately above and below Red River Landing, the alluvial

valley and deltaic plain can be considered individually, as they are geomoiphologically

distinct

There are significant differences in the thalweg profile between the alluvial

valley and deltaic plain segments. In the deltaic segments, pools are deeper and spaced

closer together then in the alluvial valley segments (Table 4.1). Results of Independent t-

tests confirm significant spatial variations between the alluvial valley and deltaic plain

(Table 4.2). The average depth of pools for the entire river was 11.3 m in 1880. The

average depth of pools in the alluvial valley segments was 9.8 m, while in the deltaic

plain the average depth o f pools increased to 14.6 m.

The spacing of pools exhibits a similar trend. The average distance between

pools for the entire river was 9,628 m. In the alluvial valley segments the distance

between pools averaged 10,049 m, while in the deltaic plain segments the distance

between pools significantly decreased to 8,666 m. Additionally, in comparison to the

alluvial valley segments, the thalweg downstream of Red River Landing has a more

uniform pattern. Evidence of this is given by the coefficient of variation (CV), a

dimensionless unit defined as the ratio between the standard deviation and the average.

A lower CV reflects less variability, as the standard deviation is lower in proportion to

the average. The CV for the alluvial valley and deltaic plain segments are .54 and .45 for

pool depth, and .53 and .46 for the distance between pools, respectively.

During the period from 1880 to 1911, the average distance between pools for the

entire channel changed very little, as there was a slight increase in pool spacing in the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.1. Summary statistics for pool depth and pool to pool distance.

Pool Depth Pool to Pool (m) Distance (m)

Survey series 1880 1911 1880 1911 Cairo, IL to Red River Landing: N = 171 Average 11.3 112 9628 9666 Min. 2.7 2.6 3219 3219 Max. 36.3 29.2 45063 25750 CV .54 .48 .52 .47 Above Red River Landing N = 119 Average 9.8 10.2 10049 10366 Min. 2.7 2.6 3219 3219 Max. 31.5 23.7 45063 25750 CV .54 .47 .53 .47 Below Red!liver Landing: N = 52 Average 14.6 13.3 8666 8241 Min. 5.7 4.5 3218 3219 Max. 36.3 29.2 19312 20922 CV .45 .46 .45 .46

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.2. Results of Independent t-test used to compare pool spacing (m) and depth (m) in the alluvial valley and deltaic plain.

Independent t-test Comparing Differences in Pool Depth and Distance Between Alluvial and Deltaic Plain variable survey t Sig. (2-tailed) Pool Distance (m) 1880 1.667 0.097* 1911 2.872 0.005* Pool Depth (m) 1880 -5.080 0.000* 1911 -3.907 0.000* * = Significant difference of means between alluvial and deltaic plain at a = .10

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. alluvial valley segments and a slight decrease in pool spacing in the deltaic plain

segments. Similarly, the depth of pools increased slightly in the alluvial valley segments

and decreased in the deltaic plain segments. However, summary statistics alone are not

powerful enough to indicate whether the changes are significantly different Table 4.3

shows the results from the Paired t-tests, which compare pool depths and distances

between the two surveys. The results indicate that the pool and riffle morphology

remained stable over the 31 years between surveys, as the depth of pools and distances

between pools did not significantly change.

4.3 PLANFORM MORPHOLOGY

4.3.1 Channel Width

Channel width displays a concave curvilinear trend (Figure 4.2a). Width

decreases uniformly from Cairo to the Arkansas River, and changes only slightly

downstream from the Arkansas River to 900 km below Cairo. However, from 900 km

below Cairo to 1500 km below Cairo channel width substantially increases, before

decreasing in the lower sections of the river towards Head of Passes. The average

channel width in 1880 for the entire channel was 751 m. While in the alluvial valley it

averages 761 m, within the deltaic segment it declines by just 30 m, averaging 731 m

from Red River Landing to Head of Passes (Table 4.4).

While the downstream of trend of channel width is curvilinear, the percent

change in channel width is linear (Figure 4.2c). The Paired Samples t-test shows that

there was not a significant change in channel width between die 1880 and 1911 surveys

(Table 4.5). However, there are several sections where a change in channel width

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.3. Results of Paired t-test between pool depth (m) and pool to pool distance (m) to test for change in the pool and riffle morphology between the 1880 and 1911 surveys.

Paired t-test of Pools and Riffles 1880-1911 variable t Sig. (2-tailed) Cairo to Head of Passes Pool Distance (m) -0.115 0.908 Pool Depth (m) 0.146 0.883 Cairo to Red River Landing Pool Distance (m) -0.448 0.654 Pool Depth (m) -0.493 0.622 . 1 B 8

Red River 1 of Passes Pool Distance (m) 0.877 0.384 Pool Depth (m) 0.896 0.374 * = Significant did ference in means at a = .10 level between 1880 and 1911 surveys

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1500

r S 3 2 St 1000 8 u

500

0 o o o o o 8 8 8 8 8 a s s s O

b. 1500 s ■o £ 1000 a

500

0 o o o o o o o oo S9SOOO OO

c. 100 75- . e ■a so- 25- £ xz U 0- e -25- §> £ -50- o ev -75- -100 o

Kilometers BelowCairo, IL

Figure 4.2a,b,c. Low Water Channel Width (m): 1880,1911, percent change.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.4. Summary statistics for sinuosity, radius of curvature, and channel width. Statistics represent the average value of all channel segments within each river section.

Summary Statistics for Planform Variables River Sum Sinuosity Radius of Curvature Channel Width (m) section stat (m) 1880 1911 Avg. % 1880 1911 Avg. % 1880 1911 Avg. % change for change change all for all for all segments segments segments Above Max. 4.4 5.6 162 4671 4166 70 1397 1068 79 RRL Min. 1.0 1.0 -55 1099 921 -62 232 503 -45 Mean 1.6 1.8 10 2527 2281 -5 761 736 -2. Below Max. 3.2 3.0 64 5885 6282 26 1030 1042 24 RRL Min. 1.0 0.9 -37 830 841 -27 250 593 -27 Mean 1.3 1.4 1 1989 1954 1 731 735 0 Cairo- Max. 4.4 5.6 162 5885 6282 70 1397 1067 79 HOP Min. 1.0 0.9 -55 830 841 -62 232 503 -45 Mean 1.5 1.6 7 2347 2172 -3 751 736 -1

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.5. Results of Paired Samples t-test to examine change over time in sinuosity, radius of curvature, and channel width.

Paired Samples t-test 1880-1911 Section N Correlation t-stat Sig. level compared coefficient Sinuosity Cairo to 167 .791 -2.913 .004* HOP Cairo to 109 .760 -2.919 .004* RRL RRL to 58 .923 -.294 .770 HOP Radius o f Cairo to 125 .746 3.100 .002* Curvature HOP Cairo to 84 .585 3.335 .001* RRL RRL to 41 .980 -.758 .453 HOP Channel Cairo to 167 .627 1.719 .088 Width HOP Cairo to 109 .431 1.871 .065 RRL RRL to 58 .947 -.489 .628 HOP * = Significant difference of means between 1880 and 19 1 at the a = .10 level

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. does occur. Between the Arkansas River and approximately 900 km below Cairo

channel width decreased during the period between the two surveys.

4.3.2 Radius of Curvature

The radius of curvature of meander bends averages 2347 (m) for the entire river

1880 (Table 4.4). In contrast to channel width, the downstream trend of radius of

curvature displays a linear trend, decreasing from Cairo to Head of Passes (Figure 4.3a).

Radius of curvature averages 2527 (m) in the alluvial valley segments, and decreases to

1989 (m) in the deltaic plain segments. Although the downstream trend in radius of

curvature is linear for most of the channel, there are several large meander bends in the

deltaic plain, downstream of Red River Landing.

From 1880 to 1911 the average radius of curvature for the entire river

significantly declined, from 2347 m to 2172 m (Tables 4.4, 4.5). Most of this change

occurred within the alluvial valley segments, which recorded a 5% reduction in radius of

curvature. This is indicated in Figure 4.4b, which in comparison to the 1880 survey,

does not show as great of a downstream trend in radius of curvature. Downstream of

Red River Landing there was not a significant change in radius of curvature. While there

were a number of individual meander bends that underwent a large change in radius of

curvature, there are two sections of channel which show a decided trend in change

(Figure 4.3c). Between the confluence with the Arkansas River and 800 km below

Cairo, there was an increase in radius of curvature. This is followed by a marked

decrease in radius of curvature from 800 km below Cairo to 1000 km below Cairo.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a .

7000

6000

5000

4000

3000

a t 2000

1000

o oo o I 8 8 8 8 8 b.

7000

6000 5000

S 4000

3000

a t 2000

1000

o o o o I 8 8 8 8 S

100 g 75 a o§ 50 £o 25 CA -25 I

o o o o o 8 8 8 8 8

Kilometers Below Cairo, IL

Figure 4.3a,b,c. Radius of Curvature: 1880, 1911, and percent change.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 00

00 vo oo b.

* oeo a3 GO

ON I 150 o 00 oo 100 *s 3o .£ 4>00 3c _o e v0 Ps" 100 o o o o o o o o o o o o o o o © o o o CM V© oo CM vo oo

Kilometers Below Cairo, IL

Figure 4.4a,b,c. Sinuosity: 1880,1911, and percent change.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.3.3 Sinuosity

The pattern of sinuosity has a convex curvilinear trend, with increasing sinuosity

in the middle reaches of the channel (Figure 4.4a). Sinuosity averages 1.5 for the entire

river in 1880. However, there is considerable spatial variability between the alluvial

valley and deltaic plain segments, averaging 1.6 and 1.3, respectively (Table 4.4).

Sinuosity between Cairo and the mouth of the Arkansas River is 1.6, but increases to 1.7

between the Arkansas River and 800 km below Cairo (Figure 4.4a). Sinuosity decreases

below 800 km below Cairo, and is fairly uniform until several sinuous channel segments

at 1400 km below Cairo, downstream of Red River Landing in the deltaic plain. From

1400 km below Cairo to Head of Passes, approximately 300 km of channel, sinuosity

substantially declines, as the channel is nearly straight From 1880 to 1911 there was a

slight change in sinuosity, increasing from 1.S to 1.6 over the entire river (Figure 4.4b).

However, the degree of temporal change in sinuosity varies spatially along the length of

the channel (Figure 4.4c). The change in sinuosity was greater in the alluvial valley

segments, which increased on average 10%, from 1.6 to 1.8, while the channel segments

in the deltaic plain did not significantly change over time (Table 4.4,4.5). From Cairo to

900 km the rate of change was uniformly distributed, with no pronounced trends. The

major change occurred within a 300 km section of channel, between 900 to 1200 km

below Cairo.

4.4 CHANNEL MIGRATION

The annual rate of channel migration in each of the 167 channel segments was

determined by superimposing the channels for each map survey. The results indicate

that during the 31 yr period some segments remained stable, however, many exhibited

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. considerable change (Figure 4.5). Although the channel in the deltaic plain exhibits little

lateral migration, many channel segments within the alluvial valley had high rates of

lateral migration.

Figure 4.6 shows the distribution and range of average migration rates for each of

the 167 channel segments and their associated location along the Lower Mississippi

River. The average migration rate (m/km/yr) was determined by measuring die distance

between the 18S0 and 1911 channel boundaries at 1.0 mi (1.6 km) intervals along the

length of the channel. Thus values represent the average amount of lateral migration

along the entire channel distance. Migration rates were measured differently for channel

segments that were also identified as meander bends. Out of the 167 channel segments,

125 segments were classified as meander bends. Due to migration rates within meander

bends having considerable spatial variability (Hickin 1983), average migration rates

measured along the entire channel distance underestimate meander bend migration.

Thus for each of the channel segments which were also identified as meander bends, the

maximum rate of migration was determined for the point along the channel which

represented the greatest amount of lateral movement between the two surveys (Figure

4.7).

There is considerable variability in migration rates for the 167 channel segments

(Figure 4.6). Migration rates range from < 0.0 to > 83.6 m/km/yr and average 18.6

m/km/yr for the entire length of the Lower Mississippi River (Table 4.6). The average

migration rate in the alluvial valley segments is 26.7 m/km/yr, while in the deltaic plain

segments migration rates average just 3.3 m/km/yr. From Cairo to approximately 800

km below Cairo migration rates are fairly uniform, averaging 25.4 m/km/yr, although

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1880

KM RED RIVER LANDING .

Figure 4.5. 1880 and 1911 channel surveys superimposed, in the vicinity of Natchez and Red River Landing.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1600 . f 8 Channel Distance Below Cairo (km) *-H(Ncn't>nvor'000\o^(SroTt- oooooooooooooo ooooooooooooooo 5b Figure 4.6. Average migration rates (m/km/yr) for all channel segments (167) from Cairo to Head of Passes. to Head Cairo (167) segments from channel all for (m/km/yr) 4.6. rates Figure migration Average VO

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. New Orleans Channel Distance Below Cairo (km) ooooooooooooooooo oooooooooooooooooo I 100 Figure 4.7. Maximum migration rates (m/yr) for meander bends (125) from Cairo to Head of Passes. Head to Cairo (125) from bends meander for (m/yr) rates 4.7. Figure migration Maximum o\ o

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.6. Summary statistics for average (m/km/yr) and maximum (m/yr) migration rates.

Average migration Maximum migration of all channel of meander bends segments (m/yr) (m/yr/km) Above Red River Landing Max 83.6 123.4 Min 5.0 13.6 Mean 26.7 44.8 Below Red River Landing Max 17.5 28.8 Min 0.0 0.0 Mean 3.3 5.5 Cairo to Head of Passes Max 83.6 123.4 Min 0.0 0.0 Mean 18.6 31.6

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. there are four segments exceeding 50 m/km/yr. Migration rates increase between 800

km below Cairo and Red River Landing, averaging 40.8 m/km/yr, with the maximum

migration rate of 83.6 m/km/yr occurring just downstream of Vicksburg.

In comparison to all of the channel segments, the maximum migration rates of

meander bends show greater variability (Figure 4.7). Table 4.6 compares the average

and maximum rates of migration. The range in maximum migration rates of meander

bends ranges from < 0.0 to > 123.4 m/yr. In comparison to Figure 4.6, many of the

meander bends show considerable variation between the average and maximum values,

suggesting underestimation of the average values for meander bends. However, the

pattern is similar to the trend of average migration rates for all 167 channel segments. In

the alluvial valley meander bends migrated on average 44.8 m/yr and ranged from 13.6

to 123.4 m/yr, while in the deltaic plain migration rates of meander bends averaged 5.5

m/yr and ranged from 0.0 to 28.8 m/yr. From Cairo to the confluence with the Arkansas

River migration rates of meander bends average 29.1 m/yr. However, migration rates of

meander bends substantially increase between the Arkansas River and Red River

Landing, averaging 50.4 m/yr.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 5

PLANFORM ADJUSTMENT AND CHANNEL MIGRATION

5.1 OVERVIEW

While Chapter 4 assessed spatial and temporal trends of several parameters,

this chapter considers the interrelationships and adjustment of channel morphology.

Linear regression and Pearson’s Correlation Coefficient is used to examine the

direction of trend and strength of relationship between variables, providing insight into

controlling factors of channel migration. The data is examined from Cairo to Head of

Passes, and then compared between the alluvial valley and the deltaic plain.

5.2 PLANFORM INTERRELATIONSHIPS

5.2.1 Sinuosity and Radius of Curvature

Correlation analysis of hydrographic survey data for the entire river examined

in 1880 indicates that sinuosity is inversely related to radius of curvature (-.19). A

channel segment having a very large radius of curvature is usually straighter then a

meander bend with a lower radius of curvature. Often as sinuosity increases meander

bends become increasingly elongate, which reduces the radius of curvature (Figure

5.1). However, significant spatial differences exist when this relationship is examined

in the alluvial valley and deltaic plain (Tables 5.1-5.3). The strength of the correlation

coefficients increase to -.35 and -.27 for the alluvial valley and deltaic plain,

respectively. The lack of significant trends for the 1911 survey are likely explained by

the findings from Chapter 4, which indicated sinuosity significantly increased in the

alluvial valley, while radius of curvature declined between 1880 and 1911. This has

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28

2 7

1811

1880

Figure 5.1. Change in meander bend radius of curvature (rc) between 1880 and 1911.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.98 0.00 Avg Mig Avg 0.05 0.07 0.14 0.70 -0.14 W.9H w * 0.00 0.60 0.87 Wisgo 0.90 0.76 0.19 0.03 0.49 r9% 0.13 0.18 0.03 0.12 -0.04 0.12 0.49 -0.01 -0.12 0.03 0.12 0.75 0.05 -0.02 rcl*80 *cl9l 1 rcl*80 *cl9l 0.16 0.03 0.83 0.72 0.00 0.02 0.200.00 0.00 0.01 0.18 0.00 0.54 0.11 0.60 -0.26 0.17 P% 0.99 0.32 0.00 0.59 0.03 0.03 0.42 0.48 0.23 0.12 0.48 -0.06 -0.15 -0.20 P|9II 0.01 -0.15 -0.17 -0.16 0.18-0.13 0.14 -0.01 P|M0 0.37 0.85 0.72 0.69 0.87 0.73 -0.07 dg4 0.32 0.03 0.32 0.00 0.00 0.44 0.62 0.78 0.00 0.00 0.01 0.00 0.01 0.00 0.00 0.00 0.000.04 0.00 0.00 0.04 0.49 0.26 0.08 0.01 0.00 0.06 Sy Sig. Sig. 0.67 0.05 0.05 0.03 0.03 Sig. Sig. Pearson Correlation Pearson 0.40 0.34 -0.07 0.05 0.13 0.20 0.02 Pearson Correlation Pearson 0.03Correlation Pearson 0.15 -0.05 0.09 Sig. Sig. Correlation Pearson Sig. 0.49 Correlation Pearson Sig. 0.31 CorrelationPearson 0.21 Sig. 0.35 0.22 0.84 Sig. Correlation Pearson Sig. 0.18Correlation Pearson 0.31 -0.19 -0.05 0.09 0.03 0.20 -0.09 0.00 -0.09 Correlation Pearson 0.02 Sig.Sig. 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Pearson Correlation Pearson 880 1880 Wisgo w„„ rc% P P|9ll P%rcl Correlation Pearson 0.37rci9ii 0.26 0.38 0.50 w* MigAvg Correlation Pearson Mig Max 0.68 Correlation Pearson 0.53 0.29 0.71 0.38 0.55 0.54 0.30 0.21 Sv dg< Table 5.1. Correlation matrix ofparameters from Cairo to Head ofPasses.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.88 0.00 AvgMig 0.02 w * 0.02 0.22 -0.25 0.14 -0.26 0.22 W|9|| 0.47 W|880 0.15 0.66 0.09 0.16 -0.04 -0.16 0.85 r9% 0.07 0.11 0.03 0.01 0.08 0.15 0.34 0.02 0.69 0.17 0.75 0.00 0.12 0.15 0.03 -0.20 0.17 -0.21 -0.20 0.16 -0.11 -0.25 rcI91l 0.10 0.02 0.30 0.00 0.02 0.26 0.04 0.25 0.17 -0.02 0.82 -0.12 -0.18 -0.03 0.04 rci880 0.00 0.40 0.01 0.10 0.09 0.07 -0.18 -0.09 P% 0.23 0.53 0.06 0.00 0.00 0.02 0.93 0.70 -0.26 -0.01 P|91l 0.40 0.76 0.04 0.06 0.00 0.04 0.42 0.00 -0.18 -0.28 -0.25 Pisso 0.00 -0.04 0.21 0.09 0.06 0.00 0.99 0.74 0.01 0.42 0.00 0.16 -0.25 -0.26 0.71 0.67 0.55 0.36 0.16 0.07 -0.09 -0.03 -0.18 0.47 -0.02 0.80 -0.14 0.11 -0.09 0.32 dM 0.01 0.15 0.27 0.02 0.94 0.07 0.32 0.67 0.56 0.61 0.10 0.01 0.01 0.07 0.96 0.20 0.04 0.82 0.42 0.90 0.10 0.03 0.35 0.00 0.28 -0.01 -0.01 0.18 -0.35 -0.22 Sy ...... Sig. Sig. Sig. Sig. Pearson Correlation Pearson 0.20 Pearson Correlation Pearson 0.10 -0.04 Sig. Sig, Sig. Correlation Pearson 0.18 Sig. Correlation Pearson Sig. Correlation Pearson Sig. Correlation Pearson 0.20 Correlation Pearson -0.05 Pearson Correlation Pearson Sig. Sig. Pearson CorrelationPearson 0.24 Pearson CorrelationPearson 0.10 Correlation Pearson -0.05 Pearson Correlation Pearson Sig. ish 1880 Max Mig Max Wigjo W w* rc%Correlation Pearson 0.09 0.09 -0.39 P Sv P% rcl880 1*1911 Avg Mig Avg rcmi d«4 Table 5.2. Correlation matrix ofparameters in the alluvial valley.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Avg Mig Avg 0.00 0.00 w * 0.17 0.05 0.50 0.19 0.710.06 0.00 0.71 0.50 0.98 w ,w, Wiggo 0.12 0.37 0.39 0.41 0.04 r9% 0.02 0.15 0.34 0.44 0.66 0.01 0.81 •0.08 0.43 0.12 rci9ii 0.14 0.18 0.20 0.60 0.00 0.15 0.33 -0.01 Icl880 0.28 0.19 0.550.02 0.61 0.02 0.35 0.26 0.20 0.00 0.290.06 0.56 0.88 0.01 0.00 -0.21 0.09 0.08 P% 0.27 0.34 0.14 0.09 0.20 0.11 0.07 -0.30 0.110.29 0.65 0.06 0.97 0.00 0.62 -0.07 0.01 0.38 0.22 -0.16 -0.17 -0.20 -0.26 0.97 1911 OP 0.91 0.06 0.04 0.07 -0.31 0.17 0.24 188 P 0.190.23 0.08 0.62 0.12 0.45 0.13 0.31 0.75 0.61 0.22 0.62 0.58 0.180.19 0.28 0.20 0.04 -0.27 0.20 0.04 0.080.32 0.31 0.21 ds4 0.12 0.36 0.16 0.17 0.06 0.45 0.58 0.65 0.34 0.77 0.39 0.24 0.00 0.00 0.01 0.01 0.52 0.25 0.23 0.19 0.13 0.220.71 0.00 0.00 0.07 0.23 0.00 0.19 0.13 0.06 0.71 0.98 -0.14 0.16 -0.29 -0.10 Sy Sig. Sig. Sig. Correlation Pearson Sig. CorrelationPearson 0.20 0.16 Sig. Sig. Sig. Pearson CorrelationPearson Sig. -0.05 Correlation Pearson Sig. Correlation Pearson 0.15Correlation Pearson 0.16 0.07 Sig. Pearson Correlation Pearson Correlation Pearson Pearson Correlation Pearson Sig. Sig. Pearson Correlation Pearson 0.57 Sig. Pearson Correlation Pearson P l9 U P%Correlation Pearson 0.05 Wisso Sv dw P|880 lcl880 r cl911 rc% Correlation Pearson 0.29 W,91l w « Max Mig Max AvgMig Table 5.3 Correlation matrix ofparameters in the deltaic plain.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. important implications regarding the character of change associated with meander

bend morphology. Increasing sinuosity in association with a reduction in radius of

curvature suggests that the meander bends were becoming increasingly elongate as the

amplitude of the meander bends increased.

5.2.2 Sinuosity and Channel Width

Sinuosity and channel width are inversely related (-.15) from Cairo to Head of

Passes in 1880 (Table 5.1). This suggests that sinuous channel segments tend to have

narrower channels then straighter reaches, which are associated with wider channels.

The relationship exhibits significant differences when viewed within the alluvial

valley and deltaic plain (Tables 5.2, 5.3). The strength of the relationship increases to

-.25 in the alluvial valley, while in the deltaic plain it is not significant Between the

two surveys there was no appreciable change in the relationship for either the alluvial

valley or the deltaic plain.

5.2.3 Radius of Curvature and Channel Width

Radius of curvature and channel width is positively correlated in 1880 (.18),

which indicates that larger meander bends tend to have a wider channel. In

comparison to the entire Lower Mississippi River, the strength of the relationship

increases in the alluvial valley, from .18 to .25, but exhibits no correlation in the

deltaic plain below Red River Landing. Radius of curvature and channel width is not

correlated in 1911. The change in the relationship over time between the two variables

is primarily due to the significant decreasing trend in radius of curvature, discussed in

Chapter 4, while channel width did not significantly change over time.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.3 DOWNSTREAM TRENDS

The valley profile shown in Figure 5.2 represents the elevation of the active

meander belt ridge sampled at 2 mile intervals (converted to km) from Cairo to Head

of Passes. Eleven profile segments were delineated, and are shown in Figure 5.2. The

characteristics of the profile segments are noted in Table 5.4. Although the overall

trend declines towards the Gulf of Mexico, slope coefficients and the shape of the

profile show that the Lower Valley does not have a smooth concave curve, but consists

of a series of different slope segments. The major valley slope segments used in this

study are delineated where there is a significant change in the profile of the meander

belt Additionally, within the segments shown in Figure 5.2 there are two major

regions. The first region occurs in the middle valley, sections 5 and 6, where there is a

pronounced change in valley gradient due to the Monroe Uplift dome. Uplift domes

typically cause a flattening of the valley profile above the axis (section 5), while

causing gradient to increase down-valley from the axis (section 6). The second region

represents a major change in sediment regime and valley slope, and occurs at sections

9 and 10 at Red River Landing, the boundary between the alluvial valley and the

deltaic plain.

The valley sections delineated in Figure 5.2 coincide with neotectonic,

lithologic, or sedimentological controls, and appear to influence channel sinuosity.

Segment 2 has a low gradient and a sinuosity of 1.81. This segment occurs along the

New Madrid fault zone, and is followed down-valley by a high gradient segment (3)

having a channel sinuosity of 2.14. Section 5 is the low gradient segment up-valley of

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. © © © © o © 00 o © o o Valley Kilometers Valley Cairo, Below IL o o © © rs © o o 100 n u § 50- W VI Figure 5.2. Profile ofelevation (MASL) sampled at2 mi (3.2 km) intervals along the LowerMississippi Rivermeander belt interval) ofthe Alluvial Valley ofthe Lower Mississippi River (1930), U.S. Army Corps ofEngineers Memphis and New analyze the influence ofvalley slope on channel morphology. Data source: Topographic Map (scale 1:500,000,5 ft contour ridge, Cairo, IL to Head of Passes, LA. The profile was subdivided into 11 shorter slope sections, which were used to Orleans District.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 5.4. Summary data for valley slope sections.

Valley Sv (m/m) Valley km Valley D ist Channel km Channel P Section below Cairo (km) below Cairo Dist. (km) 1 0.000132 41.8 41.8 48.28 48.28 1.15 2 0.000006 67.6 25.75 86.90 46.70 1.81 3 0.000199 103.0 35.41 162.50 75.60 2.14 4 0.000135 370.2 267.16 569.70 407.20 1.52 5 0.000025 412.0 41.84 645.40 75.70 1.81 6 0.000131 569.7 157.72 949.50 304.10 1.93 7 0.000071 582.6 12.88 968.90 19.40 1.51 8 0.000113 708.1 125.53 1173.30 204.40 1.63 9 0.000019 772.5 64.38 1266.60 93.30 1.45 10 0.000143 817.6 45.06 1347.10 80.50 1.79 11 0.000017 1088.0 270.38 1705.00 357.90 1.32

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the Monroe uplift axis, and occurs downstream of the Arkansas River confluence.

Sinuosity values are moderate at 1.81. In section 6, on the steeper down-valley side of

the Monroe Uplift, sinuosity increases to 1.93. Section 7 is a relatively short segment

influenced by resistant Tertiary limestone bedrock at Vicksburg, MS, and has a

sinuosity of 1.51. Downstream of Vicksburg the valley slope steepens, and this

portion of the river has a sinuosity of 1.63. Section 9 is a low gradient segment that

coincides with the boundary between the alluvial valley and deltaic plain at Red River

Landing, and has a sinuosity of 1.45. Despite section 10 occurring within cohesive

backswamp and deltaic deposits, channel sinuosity increases to 1.79. Valley slope

decreases at the lowermost section, 11, and in combination with cohesive deltaic

sediments sinuosity declines to 1.32.

5.4 VALLEY SLOPE AND CHANNEL ADJUSTMENT

Temporal and spatial change in sinuosity, radius of curvature, and channel

width during the study period is shown in Figure 5.3. The percent change of each of

the individual channel segments occurring within a valley slope section was averaged.

This permits the change that occurred along each of the 11 major slope sections of the

valley profile to be considered. The spatial distribution of channel adjustment for

channel width, radius of curvature, and sinuosity does not occur in simultaneous

locations along the length of the river. Channel width declines from section 1 to

section 6, and then increases by 25 % at section 7 at Vicksburg. Down-valley of

Vicksburg, channel width shows little temporal change, except at section 10, below

Red River Landing, where it decreases by almost 10 %. Radius of curvature shows a

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. aly lp (m/m) Slope Valley P% rc% w% 0.0001 T 0.0002 - -- 0.00005 - 0.00015 © o © »n © vo o o o. cro o o O o o o o o o (N o o CO Tt- on Kilometers Below Cairo, IL •«fr •«fr

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. similar trend to channel width, except for section 1, where it increases by almost 25 %.

Except for section 5, sinuosity increases from section 1 to section 7 at Vicksburg,

which shows a decline. Sinuosity increases substantially at section 8, but shows little

change down-valley. The reduction in sinuosity at Vicksburg represents the largest

decline of sinuosity of any of the valley segments. This valley section consists of only

two channel segments. The sinuosity reduction occurs where the channel is increasing

its radius of curvature due to being impinged between resistant backswamp deposits

and the eastern valley wall.

The influence of valley gradient on sinuosity, radius of curvature, and channel

width was further assessed using simple linear regression analysis. When the

influence of an outlier is removed, valley slope explains 39% and 40% of the

variability in sinuosity for the 1880 and 1911 surveys, respectively (Figure 5.4). The

relationship between valley slope and sinuosity has been shown as being concave-

curvilinear (i.e. Schumm and Kahn 1972), with the highest sinuosity occurring at a

mid-slope value, and abruptly decreasing towards a straight channel at a higher

threshold slope. However, Figure 5.4 suggests little resemblance towards this pattern.

The percent change in sinuosity between the two surveys is also shown, with valley

gradient explaining approximately 27% of the temporal change in sinuosity.

While valley slope explained 39 % and 40 % of the variation in sinuosity, there

is no appreciable control of valley slope on radius of curvature or channel width for

either of the two surveys. This is shown in Figures 5.5 and 5.6, and suggested by the

correlation matrices of Tables 5.1-5.3. The pattern is consistent for both the 1880 and

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ecn Change Percent -30 -10 -20 -40 -50 0.00025 R R = 0.27 0.0002 R R = 0.39 R R = 0.40 Valley Slope (m/m)

(Pll) P80 Pll P% %) (P Linear (P80) Linear Linear 0.00005 0.0001 0.00015 2.5 I o 00 00 Figure 5.4. sinuosity. valley Figure (m/m) slope and between Relationship c/<

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cag r, Change % -r 50 -- 25 - 0 - -25 0.00025 50 X 0.0002 ■ ■ rc80 ♦ rcx ll rc% - Linear ■ (rcl 1) - —Linear (rc80) Linear (rc%) ______X X ♦ ♦ X m m m m m m m td M i Valley Slope (m/m) 0.00005 0.00005 0.0001 0.00015 y X m m m m m m m m m m m >fi ” X4— = —^ ■ ■ ■

- 1000 5000 5000 -i 3000 3000 - 4000 4000 -

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cag W Change % 10 -10 -20 20 - 0 - - j 30 - - 0.00025 30 (wll) a 0.0002 ♦ wll ■ w80 X w% “ “ (w%) - Linear — Linear (w80) Linear R2 0.0262 = R2 R2 = 0.0529 = R2 ■ X * - ■ t T Valley Slope (m/m) ■ R2 = 0.0361 R2 0.00005 0.00005 0.0001 0.00015 t* .

3000 3000 -i n £ O §§ 1000 H S 2000 Figure 5.6. Relationship between valley slope (m/m) and channel width (m).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1911 surveys, and the percent change over time. Although valley slope has been

shown to effect planform channel morphology (Adams 1980), there is little influence

of valley slope on radius of curvature or channel width. Thus in the Lower Mississippi

River, sinuosity represents the dominant planform parameter controlled by valley

slope.

5.5 RELATIONSHIPS BETWEEN BED MATERIAL AND CHANNEL PATTERN

In addition to the influence of valley slope on channel parameters, bed material

is also considered an important control on channel pattern (Schumm 1977, Thome and

Tovey 1981, van den Berg 1995). Although the three variables have a positive

correlation with bed material (dg 4 mm), the coefficients are relatively low (Tables 5.1-

5.3). Bivariate plots and regression analysis show considerable scatter (Figures 5.7-

5.9). The relationship between sinuosity and bed material (Figure 5.7) is particularly

surprising considering the range of bed material size and channel patterns that exist

within the Lower Mississippi. Given the sedimentological differences between the

alluvial valley and deltaic plain, it might be expected that relationships between the

two variables would be more distinct. However, there are a number of sinuous reaches

in the upper portions of the Lower Valley, which also have fine-grained sediments.

For example, channel segments #s 67 and 68 occur downstream of the Arkansas River

in the vicinity of Greenville, MS, and have a sinuosity of 4.22 and 3.65, respectively.

However, the bed material within these bends is particularly fine grained, having

average values of 0.78 (mm) and 0.53 (mm), which reflects the influence of cohesive

Tertiary sediments which the channel is incised within.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. B a

CO f C3 <1> ce 2 & > □ •av o. CQ

[ $ 3 3 g> , and sinuosity. 4

d 1161 -0881 Figure Figure 5.7. Relationship between bed material, ds

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. , and , and radius of curvature (m). 4 Figure Figure 5.8. Relationship between bed material, ds

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. , and channel width (m). width channel , and 4

(ui) ippiM. pmreqo Figure dg 5.9. material, Figure bed between Relationship

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

% o 43

o a 10 50 30 -20 20 0 -10 -40 -50 40 8 7 2% 6 4.0 reduced their curvature, rc/W, between between 1880 rc/W, curvature, their 4.0 reduced 4% rJW, > rJW, 4 23% 1880 r^W 3 5 33% 2 7% % ofN% . . , , . \ / - • « change % rc/w —♦—migration (m/yr) 0 50 - 60 -i 40 - 1-4 t G u »- u oo

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.6 CHANNEL CURVATURE AND CHANNEL MIGRATION

Insight into meandering processes may be gained from examining migration

rates in relation to the curvature of the meander bend. Due to the form of a meander

bend being determined by spatially irregular rates of migration, it is necessary to assess

the maximum rate of migration when considering meander bend processes. One of the

most useful indices employed to examine migration rates within meander bends is the

relationship between radius of curvature and channel width, r

considered an important control on meander bend migration because of its influence

on the distribution of shear stress within the channel (Bathurst 1979), and it has been

used in the consideration of migration rates (Hickin 1974). Generally it has been

shown that lateral migration rates are greatest where channel curvature ranges from 2.0

to 3.0, and decrease above and below these values (Knighton 1998).

Figure 5.10 compares migration rates of meander bends for classes of r J W in,

and considers the change of channel curvature, tJW, between the two surveys. Due to

the extremely low rates of meander bend migration in the deltaic plain, the analysis

considers only meander bends within the alluvial valley. Migration rates and the

percent change of r

respective classes of r

between meander bend curvature, r

River is bimodal. The highest rate of migration occurs between 5 and 6, although only

a few meander bends occurs within this class. A second pronounced peak, which

represents the greatest number of meander bends (33 %), occurs between 2 and 3.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. With the exception of 7 to 8, which consists of only 1 % of the meander bends, the

lowest rates of migration occur between 3 and 4, averaging approximately 37 m/yr.

This coincides with the class of meander bends having 30 % of the meander bends,

and suggests that this may be an equilibrium meander curvature, r

is adjusting towards.

Examining the adjustment of meander bend curvature, r

period provides further evidence of this. The least amount of adjustment occurred for

meander bends having a curvature, rc/W, between 3 and 4, which is the class of

meander bends having the lowest rate of migration and includes 30% of the meander

bends. Meander bends having a curvature, r

rc/W, towards the equilibrium curvature, r

bends having a curvature, rc/W, greater than 4 reduce their curvature, r

equilibrium curvature, r

a fundamental difference in interpretation of meander bend equilibrium and the

adjustment of meander bend curvature, r

equilibrium curvature, r

rates are highest, between 2 and 3. By examining the relationship between migration

rates and the adjustment of meander bend curvature, r

this study suggest that classes of meander bends having higher rates of migration are

adjusting their channel morphology towards an equilibrium meander form curvature,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.7 CONTROLS ON MEANDER BEND MIGRATION

The correlation matrices in Tables 5.1-5.3 reveal significant correlations

between several parameters, and provide insight towards further analysis. Several

parameters that correlate highly with meander bend migration include; Sv, dg 4 , P %, rc

%, and Wbf. The controls of these parameters on rates of meander bend migration are

examined by simple and multiple linear regression analysis of meander bend migration

rates from Cairo to Head of Passes. Table 5.5 shows the results of models containing

several combinations of these variables. The coefficient of determination, R2,

represents the amount of variation in the dependent variable (maximum migration

rates) explained by the independent variables. The unstandardized coefficients

represent the slope of the relationship, or, the amount of increase in the independent

variable required for a unit increase in the dependent variable. The standardized

coefficients are dimensionless, and allow the explanatory power of individual

variables to be compared. As the standardized coefficient increases, the greater the

explanatory power of the variable.

The parameter with the highest explanatory power was the 1911 bankfull

channel width (Wbf), explaining 61% of the variability in migration rates (Nanson and

Hickin 1986). Due to its relationship with discharge and energy, channel width is

often considered an independent variable with respect to channel processes (Williams

1986). Channel migration is scale related (Hooke 1980), and is often shown as a

function of drainage area, discharge, or channel width. Although discharge varies little

in the Lower Mississippi River, the higher explanatory power of bankfull channel

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 5.5. Simple and multiple linear regression models for meander bend migration. Cairo to Head of Passes.

Dependent variable = Maximum migration (m/yr) 1 R2.50 Unstandardized Standardized t Sig. Coefficients Coefficients (Constant) 5.051 14.562 .000 Sv .899 .706 10.972 .000 2 R2 .30 Unstandardized Standardized t Sig. Coefficients Coefficients (Constant) 1.316 30.943 .000 dg4 .565 .552 7.290 .000 3 R2.61 Unstandardized Standardized t Sig. Coefficients Coefficients (Constant) -6.818 -11.473 .000 w w 2.597 .781 13.642 .000 4 R2.61 Unstandardized Standardized t Sig. Coefficients Coefficients (Constant) -1.472 -7.288 .000 Wft/Dhf 1.385 .784 13.780 .000 5 R2 2 3 Unstandardized Standardized t Sig. Coefficients Coefficients (Constant) .924 12.534 .000 P % change .419 .482 5.995 .000 6 R2 2 3 Unstandardized Standardized t Sig. Coefficients Coefficients (Constant) .601 5.038 .000 rc % change .607 .484 6.052 .000 7 R2 .63 Unstandardized Standardized t Sig. Coefficients Coefficients (Constant) -6.377 -10.737 .000 W m 2.412 .725 12.409 .000 P 1911 .598 .175 2.992 .003 8 R2 .65 Unstandardized Standardized t Sig. Coefficients Coefficients (Constant) -5.176 -8.256 .000 Whf 1.929 .589 8.968 .000 P % Change .165 .191 3.186 .002 re % change .281 .235 3.898 .000 9 R2 .64 Unstandardized Standardized t Sig. Coefficients Coefficients

(Constant) -5.681 -8.058 .000 Ww 2.195 .660 9.643 .000 P 1911 .560 .164 2.810 .006 .123 .120 1.789 .076 Dg4 table continued

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10 R2.61 Unstandardized Standardized t Sig. Coefficients Coefficients (Constant) -6.048 -9.628 .000 Ww 2.310 .701 11.169 .000 P% Change .151 .172 2.740 .007 11 R2 .70 Unstandardized Standardized t Sig. Coefficients Coefficients (Constant) -2.202 -2.389 .019 Sv .345 2 1 8 4.179 .000 P % change .142 .164 2.911 .004 rc% change .227 .190 3.309 .001 Wbf 1.464 .447 6.374 .000 All models significant at oc = .10 level

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. width in comparison to low water channel width suggests the importance of the

bankfull discharge as a channel forming variable.

Several other variables also were statistically significant in explaining

migration rates. Valley slope (Sy) represents an energy parameter (van den Berg

1995), and explains 45% of the variability in migration rates. The high explanatory

power of valley slope on channel migration likely represents the channels attempt to

adjust its channel toward an equilibrium profile. Although bed material explains 30%

of the variability in migration rates, it is surprising that it becomes insignificant or

contributes very little when combined with combinations of other variables (model 9).

Channel migration in the Lower Mississippi River occurs mainly due to bank caving

events, which are often controlled by a coarse basal layer of sediments within a

vertically stratified channel bank (Krinitzsky 1965; Kesel et al. 1974). Since dg 4

represents the coarse component of the bed material, it might be expected that this

would be representative of the basal layer of bank material, and thus explain a larger

percentage of the variability in migration rates (Thome and Tovey 1981, Nanson and

Hickin 1986).

Sinuosity, Piggo and P1911, explain a low percentage of channel migration rates,

and were not significant at the .10 level. Sinuosity can also be considered

representative of the relationship of valley slope to channel slope in dimensionless

units. Therefore the adjustment of sinuosity between the two surveys, (P% change)

may be considered representative of the slope adjustment of meander bends (Chang

1988). When used as an independent variable it explains 23% of meander bend

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. migration (model 5). It remains significant when combined with other variables,

although its explanatory power decreases (models 8, 10, 11). Although radius of

curvature was not found to be significant, the adjustment of radius of curvature

between the two surveys has the same explanatory power as percent sinuosity change,

having an R2 o f .23 (model 6).

The model with the greatest explanatory power (model 11) combines bankfull

channel width, valley slope, and the percent change in sinuosity and radius of

curvature. Combined, these variables account for 70% of the variation of meander

bend migration. However, by examining the standardized coefficients, it can be seen

that the percent change in sinuosity and radius of curvature do not offer a great deal of

explanatory power, although being significant at the .05 level. The three variables

added to bankfull channel width only increase the explanatory power of the model by

9%, and have the effect of lowering the influence of bankfull channel width.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 6

POOL AND RIFFLE MORPHOLOGY

6.1 INTRODUCTION

This chapter ex am ines the pool and riffle morphology of the channel bed of the

Lower Mississippi River. Pools and riffles are macro bedforms (Richards 1982)

which give the thalweg an undulating profile. The oscillating pattern of pools and

riffles has been linked to meandering processes due to its relationship with the

planform morphology (Leopold et al. 1964), and is considered one of the four degrees

of freedom in channel adjustment (Knighton 1998). While Chapter 4 examined spatial

and temporal trends in the spacing and depth of pools, this chapter examines

interrelationships between several aspects of the pool and riffle morphology. This is

followed with an analysis of the relationships between pool and riffle morphology

with valley slope, bed material, and planform morphology.

6.2 SPATIAL AND TEMPORAL TRENDS IN POOL - RIFFLE MORPHOLOGY

Figures 6.1-6.3 provide a more detailed look at the thalweg profile, and

identifies the pools and riffles selected by employing O'Neil and Abrahams (1984)

bedform differencing technique. There are a number of irregularities within the

oscillating thalweg pattern in the alluvial plain, having pools and riffles that vary

considerably in spacing and depth. As is the case of the planform morphology,

deviations in the thalweg profile may be due to resistant sediments and the influence

of neotectonics within the valley (Fisk 1944). This is suggested by the reach of higher

thalweg elevation downstream of Memphis, between approximately 375 and 440 km

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 009 VI VI O O O VI VI O O O 3 2 n O VI r<1 O o CN O Kilometers Below Cairo Kilometers o (N O o o o in o ❖ Riffles ❖ □ Pools Pools □ 25 - 100 -i c o 5 8 o 00 oo & 75 • f i w 50 - Figure 6.1. Location of pools and riffles along the channel thalweg of the Lower Mississippi River, 1 -600 km below -600Cairo. km River, 1 Mississippi Lower of6.1. thalweg the channel Figure the riffles along ofpools and Location i—> vo

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. z ft % N < o* £ cr c ci3 Kilometers Below Cairo Kilometers <3 a g a □ Pools □ Riffles O Cairo. Figure 6.2. Location ofpools and riffles alongthe channel thalwegof the Lower Mississippi River, 600-1200 km below to vo

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. □ Pools 0 Riffles

Kilometers Below Cairo

Figure 6.3. Location of pools and riffles along the channel thalweg of the Lower Mississippi River, 1200 km below Cairo to Head of Passes. below Cairo (Figure 6.1). The distinct change in trend suggests possible lithologic and

fault control (Fisk 1944, Schumm et al. 1994). Although two pools are identified

within this reach, this section of the channel lacks an undulating trend that

characterizes much of the profile. The channel section coincides with channel

segments 33 to 37 (Figure 4.5), having a fairly low sinuosity, and is aligned by

numerous cutoffs and ox-bow Lakes. Although the thalweg from 600 to 1200

kilometers below Cairo also has an irregular pattern, the pools increase in depth

(Figure 6.2). The increased thalweg elevation in the vicinity of Greenville may be due

to the Monroe Uplift The high elevation of the thalweg slightly downstream of

Greenville coincides with the axis of the uplift (Fisk 1944). In comparison to the

upstream flank of the uplift the pools increase in depth on the downstream flank. This

coincides with an increased valley slope, and may reflect an adjustment of the thalweg

profile in response to change in valley slope. Additionally, there are two high points

in thalweg elevation that occur at Vicksburg and Natchez. This likely represents the

influence of resistant Tertiary sediments, as the river is impinged against the eastern

valley wall (Schumm et al. 1994). The channel within the cohesive deltaic plain

deposits has a lower W/D, with deeper pools that are more closely spaced (Figure 6.3).

6.3 RELATIONSHIPS BETWEEN POOL - RIFFLE MORPHOLOGY AND CHANNEL WIDTH

Although the above discussion has concerned pool and riffle morphology as

pertaining to fluctuations in channel bed elevation, the channel bed coinciding with

pools and riffles may also vary systematically in width (Richards 1976). Larger scale

differences in channel width between the alluvial and deltaic plain were discussed in

Chapter 5. However, channel width (Table 6.1) also has a rhythmic fluctuation, being

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6.1. Pool spacing (by channel widths) and width of channel (m) at pools and riffles.

Pool to pool Pool Channel Riffle Channel Spacing Width Width (channel widths) (m) (m)

Survey series 1880 1911 1880 1911 1880 1911 Cairo, EL to Red River Landing: N = 171 Average 11.7 11.6 682 626 918 866 Min. 3.4 2.4 280 320 320 400 Max. 98.1 33.9 1440 1080 2160 2080 C.V. .49 .50 30 23 .31 27 Above Red River Landing N = 119 Average 12.0 12.3 674 605 961 889 Min. 3.4 2.4 280 320 320 400 Max. 98.1 33.9 1,440 1080 2160 2080 C.V. .87 .50 .34 2 4 .33 29 Below Red River Landing: N = 52 Average 11.1 102 700 668 821 820 Min. 4.2 3.8 500 440 600 580 Max. 30.4 262 1100 1040 1320 1360 C.V. .53 .45 20 .18 22 .19

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6.2. Independent t-test of Differences in Pool and Riffle Morphology Between Alluvial and Deltaic Plain

variable survey t Sig. (2-tailed) Pool Spacing 1880 0.654 0.514 (channel widths) 1911 2.312 0.022* Pool Width 1880 -1.489 0.138 1911 -2.673 0.008* Riffle Width 1880 2.971 0.003* 1911 1.616 0.100* * = Significant difference of means between alluvial and deltaic plain at a = .10

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6.3. Paired t-test of Pools and Riffles: 1880-1911.

variable t Sig. (2-tailed) Cairo to Head of Passes Pool Spacing -0.921 0.358 (channel widths) Pool Width 3.182 0.002* Riffle Width 1.820 0.071* Cairo to Red River Landing Pool Spacing -1.314 0.191 (channel widths) Pool Width 2.565 0.012* Riffle Width 2.090 0.039* Rec River Landing to Head of Passes Pool Spacing 0.667 0.508 (channel widths) Pool Width 2.028 0.048* Riffle Width -0.345 0.731 * = Significant dL Terence in means at .10 level between 1880 and 1911 surveys

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. significantly wider at riffles then pools. From Cairo to Head of Passes the average

channel width of riffles is 35 % wider (236 m) then at pools. Richards (1976) found

that pools and riffles significantly varied in width by 15 %. This is in agreement with

theories of hydraulic geometry, which suggest that changes in depth are compensated

for by a change in width (Leopold and Maddock 1953). Although channel width at

pools and riffles were significantly different in both the alluvial valley and deltaic

plain, there is a greater difference between the channel widths of pools and riffles in

the alluvial valley (Table 6.1). The channel width of pools significantly decreased in

the alluvial valley and deltaic plain between the two surveys, but the width of riffles

significantly changed only in the alluvial valley (Table 6.3).

Pool spacing is often considered with respect to channel width, which has the

effect of standardizing pool to pool spacing where there is a change in channel form.,

and allows for comparison of rivers of different sizes (Keller and Melhom 1978).

Channel width is used because of its relationship to discharge, and can be considered

an independent variable with respect to channel form. To obtain the spacing of pools

in channel widths, the distance between successive pools was divided by the low flow

width of the upstream riffle. The data are shown in Figures 6.4-6.6, and summary

statistics reported in Table 6.1. Each figure shows the distribution for both 1880 and

1911, allowing comparisons to be made between the two surveys.

It is widely reported that the spacing of pools, regardless of geomorphic

setting, will average from 5 to 7 channel widths (Leopold and Wolman 1957, Keller

and Melhom 1978, Gregory et al. 1994). While there is considerable variation (Figure

6.2), the pools in the Lower Mississippi River are spaced further apart then what is

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. > 0s s 3 o 100 50 75 25 as Cairo to Head ofPasses to Head Cairo rn Pool Spacing (channel widths) (channel Pool Spacing In —A— 1911% —A— CZZ31880 CZZ31880 -e -1 8 8 0 % CZZ11911 CZZ11911 5 0 10 15 20 25 c o >» cr P-, Figure 6.4. Frequency ofpool to pool spacing by channel widths, Cairo to Head ofPasses. VO VO

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 ro Os Above Red River Landing AboveRiver Red Pool Spacing (channel widths) (channel Pool Spacing [ZZ51911 -e -1 8 8 0 % —A—1911% [5531880 5 0 10 15 20 25 o c S cr PL, Figure 6.5. Frequency ofpool to pool spacingby channel widths, Cairo to Head ofPasses. o o

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. U £ 3 - - 75 r 100 O s r j Below Red River BelowLanding River Red Pool Spacing (channel widths) (channel Pool Spacing —B—1880% —B—1880% - A - 1911% CZ31880 CZZI1911 - 25 T 20 Figure 6.6. Frequency ofpool to pool spacing by channel widths, Cairo to Head ofPasses.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. commonly reported in the literature. When the entire channel is considered, from

Cairo to Head of Passes, pools are spaced on average 11.7 channel widths apart (Table

6.1), and did not significantly change in the 1911 survey (Table 6.4). In general, for

both surveys it can be seen that the distribution has a slight negative skew, with the

majority of pools occurring in the lower classes, and tapering off towards the higher

ranges. However, the distribution of pool to pool spacing varies somewhat when

compared between the alluvial valley and deltaic plain (Figure 6.5, 6.6). In 1S80 the

average spacing of pools in the alluvial valley was 12.0 channel widths, compared to

11.1 channel widths in the deltaic plain. Both the alluvial valley and the deltaic plain

have several pools that are spaced further then 20 channel widths apart The

maximum pool to pool spacing was 98.1 channel widths, which occurred downstream

of Memphis between 375 and 400 kilometers below Cairo.

Although there is a significant difference in the depth and distance between

pools, results from Independent t-tests reveal that there was not a significant difference

in pool spacing between the valley and deltaic plain in 1880. This in agreement with

Keller and Melhom (1978), who found the spacing of pools to be constant across a

range of geomorphic settings. However, there was a significant difference between

the spacing of pools between the alluvial valley and deltaic plain in the 1911 survey.

Summary statistics in Table 6.1 shows that the average spacing in the alluvial valley

increased from 12.0 to 12.3 channel widths, while in the deltaic plain the spacing of

pools decreased from 11.1 to 10.2 channel widths. The difference in spacing between

the two years was great enough for the difference to become statistically significant.

Since the distance between pools did not significantly change (Table 6.2), the

102

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6.4. Independent t-test between channel widths o f pools and riffles.

Section of River t-stat Sig. Cairo to Head of Passes: 1880 -10.064 .000* Cairo to Head of Passes: 1911 -11.644 .000* Cairo to Red River Landing: 1880 -8.015 .000* Cairo to Red River Landing: 1911 -10.293 .000* Red River Landing to Head of Passes: 1880 -3.754 .000* Red River Landing to Head of Passes: 1911 -6.131 .000* * = Significantly different at the a = .001 leve

103

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. difference in 1911 must be due to the significant decrease in riffle width, which is

used as the denominator in determining the spacing of pools.

The depth of pools and distance between successive pools has often been

shown as being a function of channel width due to its relationship with discharge

(Harvey 1975; Knighton 1998). This suggests that pool depth and the distance

between pools will increase as channel width increases. Figures 6 .1 -6 .9 shows the

relationship between pool to pool spacing (m) with channel width. The Lower

Mississippi River does not show a significant trend between these variables, as there is

considerable scatter associated with both forms of channel width. Since the alluvial

valley and deltaic plain represent unique geomorphic settings, it may be expected that

the relationships would change when plotted separately. However, Figures 6.8 and 6.9

do not reveal any change in the relationship between these variables. What is

noticeable is the distinction between riffle width and pool width, which is revealed by

the differential plotting of the trend-lines. Since riffle width is on average greater then

pool width, it plots above pool width. The relationship between pool depth and

channel width is shown in Figures 6.10-6.12. However, as is the case with pool

spacing and channel width, there is considerable scatter between the variables. The

slight trend between pool width and depth suggests that as the width of pools

decreases, the depth of pools increases. The relationship is stronger above Red River

Landing, and is not significant in the deltaic plain. Given the number of studies that

have demonstrated a linkage between these variables, the lack of a significant

relationship is somewhat surprising. However, the majority of studies on this topic

involve rivers where there is an increase in discharge and channel width, either in a

104

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50000 1 40000 Riffle: =R2 0.000 Pool: =R2 0.001 S O RiffleO Width (m) □ Pool Width (m) Pool Spacing (m) Cairo to Head ofPasses 20000 30000 D 8 D 0 0 0 o o 10000 -\ - - 500 - 1000 2500 -I 2000 (L> £ U / " “ ■s & | 1500 Figure 6.7. Relationshipbetween distance between pools (m) and channel widthsPasses. (m) ofpools and riffles. Cairo to Head of o

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I 50000 40000 Pool: = R2 0.006 6 O RiffleO Width (m) □ Pool Width (m) Riffle: =R2 0.005 1880: 1880: Pool Spacing (m) Cairo to Red River Landing O O o 10000 10000 20000 30000

- 1000 1000 2000 2500 n

§ 500 i g 1500 "jo 1 u ?V Figure 6.8. Relationship between distance between pools (m) and channel widths (m) ofpools and riffles. Cairo to Red River Landing. os o

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 6.9. Relationship between distance between pools (m) and channel widths (m) of pools and riffles. Red River Red riffles. and Passes. pools of of (m) Head to widths Landing channel and (m) pools between distance between Relationship 6.9. Figure

1880: Channel Width (m) 2000 i 2500 1000 - 1500 0 - 500 - - 0 10000 Red River Landing to Head of Passes of Head to Landing River Red Pool: Pool: 1880: Pool Spacing (m) Spacing Pool 1880: 20000 R2 = 0.000 = 00 40000 30000 □ Pool Width (m) Width Pool □ O Riffle Width (m) (m) Width Riffle O 50000 40 —j 35 pool: R = 0.10 riffle: =Rz 0.01 riffle width (m) O □ pool width (m) o 1880: (m) Pool Depth 1880: Cairo to Head of Passes Head to Cairo 10 10 15 20 25 30 x . % o .

2500 2000 H P* P* 1500 - 13 1 -I 1000 0 1 500 & I Figure 6.10. Relationship betweenpools depth (m) and channel widths (m) ofpools and riffles. Cairo to Head ofPasses. o 00

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

40 Riffle Width (m) O □ Pool Width (m) 30 35 Pool: = R2 0.09 O a Riffle: = 0.000R2 O O 1880: (m) Pool Depth 1880: Cairo to Red River Landing River Red to Cairo 10 10 15 20 25 □ □ o 0 o

A A - 2000 2500 n (D o © 00 2 500 I £ 1500 - 1 iooo Figure 6.11. Relationship between pools depth (m) and channel widths (m) of pools and riffles. Cairo to Red River to Red Cairo riffles. of (m) pools and widths channel 6.11. and (m) Figure pools between depth Relationship Landing. o vo

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. p 35 40 Pool: R = 0.000 30 O RiffleO Width (m) □ Pool Width (m) a Riffle: = R2 0.005 TTT — 1 ------1 □□ □□ a ------1880: (m) Pool Depth 1880: 1 w0 w0 0 <» 00 Red River Landing to Head ofPasses Head to Landing River Red ------1 ------5 5 10 15 20 25 _i 500 - 2500 n 2000 o £ 1500 g 1000 o 00 00 Figure 6.12. Relationship between pools depth (m) and channel ofwidthsPasses. (m) ofpools and riffles. Red RiverLanding to Head

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. downstream direction due to the contribution of tributaries, or from exam ining several

rivers ranging in size and setting. This reduces the effect of the variability associated

with a river where there is little variability in downstream discharge and channel

width, which is the case of the Lower Mississippi River.

6.4 RELATIONSHIPS BETWEEN POOL - RIFFLE MORPHOLOGY AND VALLEY SLOPE

The poor relationship between channel width and pool and riffle morphology

suggests that other variables may have greater control on the characteristics of pool

and riffle morphology. Although the influence of valley slope on the planform

geometry, particularly sinuosity, has been extensively studied (e.g. Schumm and Kahn

1972), fewer studies have exam ined the influence of valley slope on the pool and riffle

morphology. This is somewhat surprising given the linkages between the pool and

riffle morphology with planform geometry and channel equilibrium. As in the case of

sinuosity, changes in valley slope may explain irregularities within die thalweg profile.

Figure 6.13 indicates the influence of valley slope on the distance between pools.

When the 11 valley segments are considered, valley slope explains only 6 % of the

variation in pool to pool spacing. However, owing to die distinct differences in the

channel characteristics between the alluvial valley and deltaic plain, the data is

considered individually for each geomorphic setting. In this case, the amount of

variation in pool spacing explained by valley slope in the alluvial valley increases to

11%. The amount of variation in the deltaic plain increases to 71 %, although there

are only three data points, and as such there should not be too much weight given to

these findings. However, the differing positions of the trend lines confirms the

differences in pool spacing explained in the earlier portions of this chapter.

I l l

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. iue61. Relationship(m/m)and slope between Figuredistancevalley (m)between pools.6.13.

1880: Pool Spacing (m) 00 -i 20000 10000 - 15000 00 - 5000 -0.00005 - 0 0.00005 Valley Slope (m/m) Slope Valley 0.0001 Below Red River Landing: R2 = 0.71 R2 = Landing: River Red Below Above Red River Landing: R2 = 0.11 R2 = Landing: River Red Above 0.00015 0.0002 0.00025 When pool depth is regressed against valley slope, there are again two distinct

groupings (Figure 6.14). Although only 9 % of the variation in pool depth is

explained by valley slope in the deltaic plain, the data plots substantially higher then

pool depth in the alluvial plain. The lack of a trend suggests the uniformity of pattern

of pool depth within the deltaic plain. There is an inverse relationship between pool

depth and valley slope above Red River Landing, with 24 % of the variation in pool

depth explained. Although there is some scatter within the data, the pattern suggests

that pool depth decrease with increasing valley slope.

The preceding analysis concerning pool spacing and depth revealed a poor

relationship between these variables (Figures 6.10-6.12). However, by standardizing

pool spacing with channel width and aggregating the data for each of the 11 valley

segments, the relationship between the variables can be assessed with the influence of

valley slope inherent within the data. Figure 6.15 shows the relationship between

these variables with the data subdivided between the alluvial valley and deltaic plain.

By utilizing this approach, it can be seen that 62 % of the variation in pool depth is

explained by pool to pool spacing in channel widths, suggesting the depth of pools

increases with increasing pool spacing. The data plotted for the deltaic plain does not

suggest a trend.

6.5 RELATIONSHIPS BETWEEN POOL - RIFFLE MORPHOLOGY AND BED MATERIAL

The association of bed material with pool depth is shown in the correlation

matrix in Table 6.5. There is a significant negative correlation between bed material

(dg4 mm) and pool depth for the 1880 and 1911 surveys from Cairo to Head of Passes

(Table 6.5). This indicates that pools become deeper as the size of the bed material

113

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.00025 0.0002 0.00015 Below Red River Landing: =R2 0.09 0.0001 Above Red River Landing: =R2 0.24 Valley Slope (m/m) 0.00005 0 - 1 0 -0.00005 20 ~ Figure 6.14. (m). Figure pool depth valley between and slope (m/m) Relationship

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20 15 Above Red River Landing: =R2 0.62 Below Red River Landing:= 0.002R2 10 1880: Pool Spacing (channel widths) (channel Pool Spacing 1880: 5 0 5 0 15 10 a u o Q ’o a, 00 ■s © 00 Figure 6.15. Relationship between the distance between pools (m) and pool depth (m) for data correspondingto each valley section.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

0.08 0.27 0.10 0.05 0.09 0.07 0.09 0.06 0.12 0.10 -0.22 -0.14 -0.15 -0.07 -0.06 0.29* rJW% r

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. decreases. The correlation is stronger when the entire length of the channel is

considered, from Cairo to Head of Passes, which reflects the distinct sedimentological

and morphological differences in the alluvial valley and deltaic plain. Since the bed

material was sampled in 1932, it is somewhat surprising that there was not a

significant correlation between the depth of pools above Red River Landing in the

1911 survey. Bed material is also significant when correlated with the percent change

in pool depth. The direction of the trend is positive, and is stronger when only the

channel above Red River Landing is considered. This indicates that between the 31

years of the two surveys, that pools have a tendency to increase in depth with

increasing size of bed material.

6.6 RELATIONSHIPS BETWEEN POOL - RIFFLE MORPHOLOGY AND PLANFORM MORPHOLOGY

There have been relatively few studies that have examined the pool and riffle

morphologywithin the context of the adjustment of planform channel geometry. In

Chapter 5 the influence of channel curvature on channel migration was shown.

However, channel curvature may also have an influence on the depth of pools. In this

section, the depth of pools is considered with respect to channel curvature, which is

expressed as radius of curvature, meander bend curvature (r

sinuosity. Correlation matrices are employed to explain the associations between

these variables and pool depth, the percent change in pool depth between the two

surveys, and the number of pools within a channel segment (Table 6.5).

Sinuosity does not show a significant correlation in 1880, but correlates with

several variables in the 1911 survey. Above Red River Landing, sinuosity correlates

negatively with the number of pools within a channel segment This indicates that

117

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. straighter channel segments tend to have a higher number of pools. There is a positive

correlation between sinuosity in 1911 with die percent change in pool depth. Since the

percent change in sinuosity also shows the same strength and direction of correlation

with percent change in pool depth, this indicates that channel segments that increased

in sinuosity between 1880 and 191 lalso increased the depth of pools associated with

these channel segments. Further proof of this is given from the negative correlation

between the percent change in sinuosity and depth of pools in 1880, which indicates

that sinuosity increased in those channel segments which did not have deep pools.

There are no significant correlations with sinuosity in the deltaic plain. Since the

strength of the correlations are nearly identical in the alluvial valley and when the

entire channel is considered (Cairo to Head of Passes), this supports the findings

reported from the t-tests, which found that sinuosity and pool and riffle morphology

did not significantly change in the deltaic plain.

There are several significant correlations between radius of curvature with pool

depth (Table 6.5). When the entire length of the channel is considered, there is a

negative correlation between radius of curvature and pool depth for both 1880 and

1911. The trend strengthens in the alluvial valley. Radius of curvature is a

measurement of the size of a meander bend, and thus this correlation indicates that

smaller meander bends tend to have deeper pools. However, radius of curvature alone

does not indicate the "sharpness" of a meander bend, as it is scale related. Channel

curvature, r

radius of curvature, and is more representative of the "sharpness" of a meander bend.

From Cairo to Head of Passes in 1880 and 1911 deeper pools tend to be associated

118

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with "sharper" meander bends, as r

bend curvature has been linked to the control of shear stress distribution within

meander bends (Bathurst 1979). Thus the correlation between meander bend

curvature and pool depth may suggest that smaller meander bends are able to more

effectively scour their channel bed, causing pools to increase in depth. The spatial

variation between the alluvial valley and deltaic plain is of interest The variables are

correlated in 1911 in the alluvial valley, but not for the 1880 survey. As Chapter 4 and

5 indicated, in the alluvial valley meander bends became increasingly elongate, as

sinuosity increased and radius of curvature decreased between the 1880 and 1911

survey. Thus the results of these correlations indicate that as the meander bends were

becoming increasingly elongate, the pools which were associated with the meander

bends increased in depth.

119

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 7

DISCUSSION, CONCLUSIONS, AND FUTURE RESEACH

7.1 DISCUSSION

7.1.1 Mechanisms of Channel Change

Although this study has mainly concerned the characterization and adjustment

of channel pattern, additional insight into meandering processes may be gained by

considering the findings within the context of their causal mechanisms. Controlling

mechanisms of bank retreat and the development of planform morphology in

meandering rivers has often been related to flooding and mass failure of heterogeneous

bank sediments (Kesel et al. 1974; Hooke 1980; Twidale 1966; Thome and Osman

1988; Lawler 1995). In the Lower Mississippi River the flood regime is markedly

seasonal, which means that planform change in this study has occurred over

approximately 31 flood events. However, the process of bank erosion occurs

differentially during the flood hydrograph. During the rising and peak stage of a flood

event, basal sediments erode due to scour and flow-failure of coarse-grained bank

sediments at title channel bed. This reduces the support of the upper bank sediments,

which become saturated and fail during the receding limb of the flood hydrograph

(Kesel et al. 1974). Krinitzsky (1965) found this to be the mechanism of bank failure

in the area between Lake Providence and Vicksburg, which in this study had the

highest rates of meander bend migration, 123.4 m/yr (Figure 4.7), and change in

sinuosity (Figure 4.4).

120

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. While the mechanism of bank retreat is related to the seasonal flow regime, the

geometry and helicoidal flow pattern of meander bends represent a self-regulatory

mechanism (Callander 1978). This enables the meander bend to maintain an

equilibrium form, which is often related to a ratio of radius of curvature to channel

width, r

has important implications regarding the control of meander bend processes in the

alluvial valley. While the most stable meander bend curvature, r

valley was found to be between 3.0 and 4.0 (Figure 5.10), the ratio is also likely

controlled by the distribution of shear stress within the meander bend. A number of

authors have shown that as the r

Lower Mississippi River), the flow path in the channel changes and causes increased

erosion upstream and downstream of the meander bend apex. This triggers a negative

feedback, as the meander bend increase its r

r,/W towards a more stable form. Conversely, a meander bend that increases its

curvature beyond a stable meander bend r J W causes excessive erosion at the apex,

resulting in a more arcuate meander bend (Bathurst 1979; Richards 1982).

The importance of helicoidal flow with regards to meander bend processes in

the alluvial valley is further suggested by the interrelationship between the planform

and pool and riffle morphology, as described in Chapters Five and Six. Meander

bends that became increasingly elongate were found to increase the depth of pools

associated with these bends. This is in agreement with a number of studies that have

found an increase in shear stress exerted at the channel bed on the concave portion of

121

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the meander bend. As the bend becomes more arcuate the flow is concentrated within

the meander bend, which increases superelevation of the flow. The increase in

boundary shear stress causes scour of the channel bed, which increases the depth of

pools (Callander 1978; Dietrich 1987).

7.1.2 Influence of Sediment and Valley Slope

Heterogeneous sediments within the channel perimeter (Fisk 1947, Winkley

1970) are likely responsible for the low explanatory power of many of the empirical

relations (Knighton 1987; Schumm et al. 1994). The complexity of sediments within

the channel perimeter makes it difficult to generalize downstream trends in grain size.

Tributaries entering the Lower Valley contribute coarse-grained sediments and are

responsible for the Lower Mississippi River lacking a smooth downstream decline in

bed material size (Leopold et al. 1964). The influence of sediment contribution by

tributaries is exacerbated by their low discharge contribution, which results in the

Lower Mississippi River having little downstream variation in discharge (Figure 2.2).

The issue of channel sediments is further complicated because sediments of different

sizes and cohesion are often vertically stratified within the channel bank (Krinitzsky

1965; Torrey et al. 1988). In addition to the sediment load within the channel, modem

sedimentary units of significance include backswamp deposits and clay plugs (Figure

7.1). Clay plugs are much less numerous below Vicksburg, as the river flows within a

younger meander belt and the channel is less confined (Fisk 1947). In combination

with the fewer number of clay plugs, the Grand Gulf Group has been deeply scoured

122

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Control of River Alignment by Clay Plugs in the Lake Providence Reach of the Mississippi River

3rocr

MayersvilJe

Sand deposits in which v :: l river can migrate freely ,/y//A Clay plugs of /////A variable thickness

Providence

-32*45*

Modified from Fisk. 1944

Figure 7.1. Control of channel activity by resistant clay plugs in the vicinity of Lake Providence, LA (modified from Fisk 1944).

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Figure 7.2) and does not restrict channel movement, as this portion of the river has the

highest rates of channel migration in the Lower Mississippi River (Figure 4.7).

In many reaches older sediments and bedrock are in contact with the channel

bed (Humphreys and Abbott 1861; Winkley 1994). In several instances., changes in

the profile of the meander belt ridge is due to the occurrence of resistant bedrock, and

therefore also represent a change in lithology. Thus it may be difficult to isolate the

effects of changes in slope from the change in lithology on channel processes. For

example, on the upstream flank of the Monroe Uplift, the valley slope is flattened and

channel sinuosity is lower then on the down-valley flank of the uplift dome. However,

up-valley of the uplift axis the thalweg is incised within cohesive Tertiary clay of the

Yazoo formation, restricting lateral movement of the channel. On the downstream

flank the channel is in contact with the erosive Cockfield sand and gravel (Winkley

1970), and both sinuosity and channel migration increase (Figure 7.2).

7.1.3 Linking Pool and Riffle Morphology with Meandering Process

The consideration of the adjustment of channel form with regards to energy

and resistance provides a conceptual framework in which to view meandering

processes between the alluvial valley and the deltaic plain. Channel morphology in the

alluvial valley is characterized by its planform morphology having a sinuous laterally

migratingchannel. The morphology of a meandering river adjusts to an equilibrium

profile that minimizes the rate of energy expenditure within the channel (Langbein and

Leopold 1966; Cherkauer 1973). A meandering pattern enables a river to regulate its

energy and resistance in several ways. As a river channel migrates laterally, the slope

124

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T S 'W - J33* U! UOI1SA3I3

O o !«■“ .5 e •E| s I s

c e» •u

■1 =•

^999999

Figure 7.2. Long-profile of bedrock and older sediments within the Lower Valley (modified from Fisk 1944).

125

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of the channel is reduced, which reduces kinetic energy within the channel (Friedkin

1945). Additionally, meanders directly increase energy expenditure by increasing

resistance due to channel curvature and increasing the length of the surface with which

the streamflow is in contact (Richards 1982; Knighton 1998). By development of a

series of meander bends, energy expenditure is more uniformly distributed along the

entire length of the channel, which reduces the variance of energy expenditure (Yang

1971).

Whereas in the deltaic plain there are fewer meander bends, the channel is

characterized by low sinuosity, low W/D, with deeper pools that are closely spaced.

While the river can not expend energy in the lateral dimension due to the resistance of

the cohesive bank sediments, the pattern of pools and riffles formed in the channel bed

function to increase the proportion of energy expenditure at the bed of the channel

(Langbein and Leopold 1966; Yang 1971), and has been referred to as vertical

meandering (Whol 1993). Additionally, while the cohesive nature of the deltaic plain

sediments can be seen as influencing the pattern of erosion on the channel bed, it also

may influence the pattern of sediment deposition. Since the river can not erode its

channel hanks, it inhibits the development of point-bar platforms necessary for

meandering (H. Ikeda 1989). The bed material that would be stored in point bar

platforms is confined to a smaller area within the channel bed. The sediment is stored

in riffles rather then point bars, which increases the relative depth of pools.

126

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7.1.4 Significance and Contribution

Findings from this research have regional significance in that they establish the

base-line geomorphic conditions of the river before major human disturbance, while

the river is considered representative of its natural channel regime. By analysis of

flow stage data, several researchers have suggested that the channel of the Lower

Mississippi River was infilling and becoming wider during this period. However this

conclusion may be faulty on several accounts. The data had not been corrected for the

error in the vertical datum associated with the 1880 and 1911 hydrographic surveys.

Additionally, the increase in stage coincides with increasing levee heights. By

adjusting the 1880 and 1911 surveys to a common ALWP it has been shown in this

study (Chapter Four) that channel width did not significantly increase between the two

surveys. Ockerson (1892) in a study of bank caving along the Lower Mississippi

River makes a similar observation.

The conclusion that the Lower Mississippi River was not significantly

degrading its channel during this period is further substantiated by findings from

analysis of the pool and riffle morphology. The pool and riffle morphology is one of

the primary means of channel adjustment during a degradational phase (Dury 1971).

Excessive erosion in the planimetric dimension would likely be responsible for a

decrease in pool depth due to channel infilling and an increase in channel width due to

bank failure. However, temporal analysis of the pool and riffle morphology between

the 1880 and 1911 surveys found that the depth and spacing of pools did not

significantly change between the two surveys.

127

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. From the larger perspective of meandering processes this study is significant

for several reasons. The majority of studies on this topic concern smaller coarse

grained systems, thus this study represents a departure from the literature in that the

Lower Mississippi River is a very large complex system. Of worthnoting is the

deviation from the expected trends of channel width and sinuosity with bed material,

which has been found by a number of researchers to have significant control on these

dimensions of channel morphology (e.g. Schumm 1960, 1963; Ebisemiju 1994; van

den Berg 1995). While the coarse component of the bed material is often considered

representative of the bank material in meandering rivers, the heterogeneously stratified

channel banks in the Lower Mississippi River indicate that this generalization does not

always hold true. Thus empirical relationships in large river systems such as the

Lower Mississippi that are influenced by older sediments and bedrock may be of little

use unless they are refined in a manner which addresses the unique conditions of the

system under consideration (Knighton 1987; Petts 1995).

One of the more significant findings of this study concerns the importance of

the pool and riffle morphology with respect to meandering processes. Much of the

literature suggests that this is mainly a characteristic of coarser-grained systems (e.g.

Knighton 1998), which it likely due to many studies on this topic having established

that pools and riffles differ with respect to their sedimentology (e.g. Hirsch and

Abrahams 1981; Milne 1982; Clifford 1993; Sear 1996). Findings from this study

suggest that the pool and riffle morphology is a fundamental component of the

meandering process in large fine grained fluvial systems, and can become the

128

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. predominant manner in which the river adjusts its channel morphology to reduce the

variance of energy expenditure within the channel.

7.2 CONCLUSIONS

1.) Downstream trends in thalweg elevation coincide with major changes to the

sediment regime and valley slope. The channel thalweg has a concave profile, with

the departure from linearity occurring in the lower alluvial valley, near the boundary

with the deltaic plain. The distance between pools decreases and the depth of pools

increases in a downstream direction, but did not significantly change over time.

2.) Downstream patterns of channel width, radius of curvature, and sinuosity depart

from common trends, having a concave, linear, and convex downstream profile,

respectively. The departure from linearity for channel width and sinuosity occurs in

the middle and lower portions of the alluvial valley where there is a significant change

in the sediment regime and valley slope. Sinuosity significantly increases within the

alluvial valley between the 1880 and 1911 surveys, with the greatest amount of change

occurring from 800 to 1200 km below Cairo.

3.) Channel migration varies considerably between the alluvial valley and the deltaic

plain, averaging 27 m/yr and 3 m/yr, respectively. The highest rates of channel

migration coincide with the greatest change in sinuosity, from 800 to 1200 km below

Cairo.

4.) Sinuosity increased and radius of curvature decreased between the two surveys

within the alluvial valley. Much of the reduction in radius of curvature occurred

129

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. where sinuosity increased, between 800 to 1200 km below Cairo, as meander bends in

this reach became increasingly elongate between 1880 and 1911.

5.) Analysis of meander bend migration with the adjustment of channel curvature,

rc/W, occurs between 3.0 and 4.0. Meander bends having a curvature, r

increase their curvature, r 4.0

reduce their curvature, r

6.) Channel migration is best explained by bankfull channel width. In contrast,

channel width for low water was not significant, which suggests the importance of the

hank-full discharge as a channel forming discharge.

7.) There is a significant difference between the widths of pools and riffles. Channel

width at riffles is 35 % greater then the channel width of pools.

8.) The spacing of pools by channel widths is greater then what is commonly reported

in the literature. In the alluvial valley pools are spaced on average 12.0 channel width

apart, whereas in the deltaic plain pools are spaced every 11.1 channel widths.

However, there is not a significant difference in pool spacing by channel widths

between the alluvial valley and deltaic plain in 1880.

9.) While in the alluvial valley straighter channel segments tended to have a higher

number of pools, sinuous channel segments had deeper pools that increased in depth as

the channel segments became more arcuate. Additionally, deeper pools were

associated with finer bed material. However, the depth of pools increased faster in

coarse-grained channel segments.

130

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10.) Although pool and riffle morphology is usually represented as being related to the

planform morphology, in the Lower Mississippi River it is better developed in the

deltaic plain where there are few meander bends.

7.3 FUTURE RESEARCH

While this study has addressed a number of issues concerning meandering

processes in the natural channel regime of the Lower Mississippi River, there is a need

for additional work on several key fronts. This study has focused on examining the

differences in the meandering processes between the alluvial valley and deltaic plain,

which may serve to. mask some of the significant relationships between variables.

More in depth research could be undertaken with the existing database considering

smaller channel reaches.

One approach that would likely enhance the current study would be to

subdivide the individual channel segments according to the type of sedimentary

environments they are migrating within. This would allow the different sedimentary

units to be examined individually with regards to their influence on meandering

processes. However, this approach could be further improved upon by employing a

detailed sedimentological database that takes into account the vertical stratification of

channel bank sediments. Such data could lead to the refinement of existing empirical

models concerning bank failure, channel migration, and channel pattern. Much of this

data exists in the form of reports and studies for different channel sections completed

by the U.S. Army Corps Engineers Waterways Experiment Station, but needs to be

assimilated into a single database.

131

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The hydraulic component of the system can be accounted for by using the

water surface slope for estimation of stream power and the boundary shear stress on

the channel bed. When combined with the sedimentological data, this may allow for a

greater comprehension of the mechanics of channel processes, which may elucidate

thresholds in channel patterns.

While this study has examined pool and riffle morphology, additional work on

this topic could be done which employs more sophisticated techniques that may aid in

identifying the relationship with other dimensions of the channel morphology. One

such approach that has not yet been utilized on pool and riffle morphology is fractal

analysis. While it has been employed in the analysis of meander bends (e.g.

Montgomery 1996) the oscillating pattern of the pool and riffle morphology suggests

an application in the Lower Mississippi River.

Finally, as this study has characterized geomorphic processes during the natural

channel regime of the Lower Mississippi River, additional research should examine

channel adjustment in response to human modifications. While there have been

several studies which have addressed this issue, the majority of these studies have not

considered change from the standpoint of the natural channel regime.

132

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES

Adams, J. 1980. Active tilting of the Unites States midcontinent: geodetic and geomorphic evidence. G eo lo g y, v. 8, pp. 442-446.

Alabayan, A.M. and Chalov, R.S. 1998. Types of river channel patterns and their natural controls. Earth Surface Processes and Landforms, v. 23, pp. 467-474.

Alexander, C.S. and Nunnally, N.R. 1972. Channel stability on the Lower Ohio River. Annals o f the Association o f American Geographers, v. 62, no. 3, pp. 411- 417.

Anding, M.G, 1970. Hydraulic Characteristics o f Mississippi River Channels. Potamology Investigations Report no. 19-3, U.S. Army Corps of Engineers Waterways Experiment Station, Vicksburg, MS. pp. 1-67.

Aslan, A. 1994. Holocene sedimentation, soil formation, and floodplain evolution of the Mississippi River floodplain, Ferriday, LA. Ph.D. dissertation, University of Colorado, Boulder, CO.

Assifi, A.T. 1966. Hydraulics and geometry of rivers. Masters thesis, Colorado State University, Ft. Collins, CO.

Autin, W.J., Bums, S.F., Miller, B.J., Saucier, R.T., and Snead, J.I. 1991. Quaternary geology of the Lower Mississippi Valley. In. R.B. Morrison (Ed.), Quaternary Nonglacial Geology, Conterminous GeologicalU.S., Society of America, Boulder, CO, pp. 547-582.

Baker, V.R. 1988. Geological fluvial geomorphology. Geological Society o f America Bulletin, v. 100, pp. 1157-1167.

Bathurst, J.C. 1979. Distribution of boundary shear stress in rivers. In D.D. Rhodes and G.P. Williams (Eds.), Adjustments o f the Fluvial System, Kendall Hunt, Dubuque, Iowa, 372 p.

Brice, J.C. 1973. Meandering pattern of the White River in Indiana - an analysis. In M. Morisawa (Ed.), Fluvial Geomorphology, Binghampton Publications in Geomorphology, pp. 179-200.

Brice, J.C. 1974. Evolution of meander loops. Geological Society o f America B u lle tin v. , 85, pp. 581-586.

Brice, J.C. 1984. Planform properties of meandering rivers. In C.M. Elliot (Ed.), River Meandering, New Orleans, American Society of Civil Engineers, pp. 1- 15.

133

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bridge, J.S. 1977. Flow, bed topography, grain size and sedimentary structure in open channel bends: a three-dimensional model. Earth Surface Processes and L a n d fo rm s, v. 2, pp. 401-416.

Bridge, J.S. 1984. Flow and sedimentary processes in river bends: comparison of field observations and theory. In C.M. Elliot (Ed.), River Meandering, New Orleans, American Society of Civil Engineers, pp. 857-872.

Bridge, J.S. and Jarvis, J. 1976. Flow and sedimentary processes in the meandering river South Esk, Glen Clova, Scotland. Earth Surface Processes, v. 1, pp. 303-336.

Burnett, A.W. and Schumm, S.A. 1983. Alluvial-river response to neotectonic deformation in Louisiana and Mississippi. S c ie n c e ,v. 222, pp. 49-50.

Callander, R.A. 1978. River meandering. American Review o f Fluid Mechanics, v. 10, pp. 129-158.

Carson, M.A. 1984. The meandering-braided river threshold, a reappraisal. Journal o fHydrology, v. 73, pp. 315-334.

Chang, H.H. 1988. On the cause of river meandering. In W.R. White (Ed.), International Conference on River Regime, Wallingford, U.K., Hydraulics Research, pp. 83-93.

Chang, H.H. and Toebes, G.H. 1970a. A statistical comparison of meander planforms in die Wabash Basin. Water Resources Research, v. 6, pp. 557-577.

Chang, H.H. and Toebes, G.H. 1970b. Geometric parameters for alluvial rivers related to regional geology.International Association fo r Hydraulic Research, pp. C24 — 1-C 24-9.

Cherkauer, D.S. 1973. Minimization of power expenditure in a riffle-pool alluvial channel. Water Resources Research, v. 9, no. 6, pp. 1613-1628.

Chitale, S.V. 1973. Theories and relationships of river channel patterns. Journal o fHydrology, v. 19, pp. 285-308.

Clark, W A V . and Hosking, P.L. 1986. Statistical Methods for Geographers. John Wiley & Sons, New York, NY, 518 p.

Clifford, N J. 1993. Formation of riffle-pool sequences: field evidence for an autogenetic process. In C.R. Fielding (Ed.), Current Research in Fluvial Sedimentology, Sedimentary Geology, v. 85, pp. 39-51.

134

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Clifford, NJ. and Richards, K.S. 1992. The reversal hypothesis and the maintenance of riffle-pool sequences: a review and field appraisal, hi P.A. Carling and G.E. Petts (Eds.), Lowland floodplain rivers: Geomorphological perspectives, John Wiley & Sons Ltd., pp. 43-70.

Dietrich, WJE., 1987. Mechanics of flow and sediment transport in river bends: In K.S. Richards (Ed.), River Channels: Environment and Process, Basil Blackwell, pp. 179-227.

Dietrich, W.E. and Gallinatti, J.D. 1991. Fluvial Geomorphology. In O. Slaymaker (Ed.), Field Experiments and Measurement Programs in Geomorphology, pp. 169-220.

Dietrich, W.E. and Smith, J.D. 1983. Influence of the point bar on flow through curved channels. Water Resources Research, v. 19, pp. 1173-92.

Dietrich, WJE. and Smith, J.D. 1984. Processes controlling the equilibrium bed morphology in river meanders.. In C.M. Elliot (Ed.), River Meandering, New Orleans, American Society of Civil Engineers, pp. 759-769.

Downward, S.R. 1995. Information from topographic survey. In A.Gumell and G. Petts (Eds.), Changing River Channels, John Wiley & Sons, New York, NY, pp. 304- 323.

Dury, G.H. 1971. A re-survey of part of the Hawkesbury River, New South Wales, after one hundred years. Australian Geographical Studies, pp. 121-131.

Ebisemiju, F.S. 1994. The sinuosity of alluvial river channels in the seasonally wet tropical environment case study of river Elemi, southwestern Nigeria. C a ten a , v. 21, pp. 13-25.

Einstein, H.A. and Shen, H.W. 1964. A study on meandering in straight alluvial channels. Journal o f Geophysical Research, v. 69, no. 24, pp. 5239-5247.

Elliot, D.O. 1932. The improvement o f the Lower M ississippi River fo r Flood Control and Navigation, US Army Corps of Engineers Waterways Experiment Station, Vicksburg, MS, v. 1-3.

Ferguson, H.B. 1940. History o f the Improvement o f the Lower M ississippi River for Flood Control and Navigation From 1932-1939. Mississippi River Commission Vicksburg, MS, 198 p.

Ferguson, R.1.1975. Meander irregularity and wavelength estimation J o u rn a l o f H yd ro lo g y, v. 26, pp.’315-333.

135

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ferguson, R.I. 1977. Meander migration: equilibrium and change. In K.J. Gregory (Ed.),River Channel Changes, John Wiley & Sons, New York, pp. 235-248.

Ferguson, R.I. 1987. Hydraulic and sedimentary controls of channel patterns. In K.S. Richards (Ed.), River Channels: Environment and Process, Basil Blackwell, pp. 129-158.

Fisk, H.N. 1944, Geological Investigation o f the Alluvial Valley o f the Lower M ississippi River. Mississippi River Commission, Vicksburg, MS, 44 p.

Fisk, H.N. 1947. Fine-Grained Alluvial Deposits and Their Effects on Mississippi River Activity, U.S. Army Corps of Engineers, Mississippi River Commission, Vicksburg, MS, 82 p.

Fisk, H.N. 1951. Mississippi River Valley geology in relation to river regime, American Society of Civil Engineers, no. 2511, pp. 349-364.

Fisk, H.N. 1952, Geological Investigation o f the Atchafalaya Basin and Problems o f M ississippi Diversion. U.S. Army Corps of Engineers, Mississippi River Commission, Vicksburg, MS, 145 p.

Frazier, D.E. 1967. Recent deltaic deposits of the Mississippi River, their development and chronology. G ulf Coast Association Geologists Society Transactions, v. 17, pp. 287-315.

Friedkin, JJF. 1945. A Laboratory Study of die Meandering of Alluvial Rivers, U.S. Waterways Experiment Station, Vickburg, MS, 40 p.

Gregory, D i. and Schumm, S.A. 1987. The effect of active tectonics on alluvial river morphology. In K.S. Richards (Ed.), River Channels: Environment and Process. Basil Blackwell, pp. 41-68.

Gregory, KJ., Gumell, A.M., Hill, C.T., and Tooth.S, 1994, Stability of the pool-rifQe sequence in changing river channels. Regulated Rivers: Research and Management, v. 9, pp. 35-43.

Hannan, A. 1969. Study of Mississippi River bends, PhJD. dissertation, Colorado Stale University, F t Collins, CO.

Harbor, D.J. 1998. Dynamic equilibrium between an active uplift and the Sevier River, Utah. Journal o f Geology, v. 106, pp. 181-194.

136

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Harvey, A.M. 1975. Some aspects o f the relations between channel characteristics and riffle spacing in meandering streams. American Journal o f Science, v. 275, pp. 470-478.

Hickin, E.J. 1974. The development of river meanders in natural river channels. American Journal o f Science, v. 274, pp. 414-42.

Hicltin, E J. 1983. River channel changes: retrospect and prospect In J.D. Collinson and J. Lewin (Eds.), Modem and Ancient Fluvial Systems, Special Publication no. 6, International Association of Sedimentologist, pp. 61-83.

Hickin, EJ. and Nanson, G.C. 1975. The character of channel migration on the Beatton River, Northeast British Columibia, Canada. Geological Society o f America B u lle tin v. , 86, pp. 487-494.

Hirsch, P.J. and Abrahams, A.D. 1981. The properties of bed sediments in pools and riffles. Journal o f Sedimentary Petrology, v. 51, no. 3, pp. 757-760.

Hooke, J.M. 1977. The distribution and nature of changes in river channel patterns: the example of Devon. In K.J. Gregory (Ed.), River Channel Changes, John Wiley & Sons: New York, pp. 265-280.

Hooke, J.M. 1979. An analysis of the processes of river bank erosion. Jo u rn a l o f H ydrology, v. 42, pp. 39-62.

Hooke, J.M. 1980. Magnitude and distribution of rates of river bank erosion. Earth Surface Processes, v. 5, pp. 143-157.

Hooke, J.M. 1984. Changes in river meanders: review of techniques and results of analyses. Progress in Physical Geography, v. 8, pp. 473-508.

Hooke, J.M. 1995. Processes of channel planform change on meandering channels in the UK. In A. Gumell and G. Petts (Eds.), Changing River Channels, John Wiley & Sons, New York, pp. 87-116.

Hooke, J.M. and Harvey, A M . 1983. Meander changes in relation to bend morphology and secondary flows. In J.D. Collinson and J. Lewin (Eds.), M o d em and Ancient Fluvial Systems, Special publication no. 6, International Association of Sedimentologist, pp. 121-132.

Howard, A.D. 1987. Modeling fluvial systems: rock-, gravel-, and sand-bed channels. In K.S. Richards (Ed.), River Channels: Environment and Process. Basil Blackwell, Oxford, U.K., pp. 69-94.

137

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Humphreys, A.A. and Abbot, H.L. 1861. Report Upon the Physics and Hydraulics o f the Mississippi River, Professional Papers of the Topographical Engineers, no. 4, J.B. Lippincott, Philadelphia, PA.

Ikeda, H. 1989. Sedimentary controls on channel migration and origin of point bars in sand-bedded meandering channels. In S. Ikeda and G. Parker (Eds.) R iv e r Meandering, American Geophysical Union Water Resources Monograph, no. 12, Washington D.C., pp. 51-67.

Ikeda, S. 1989. Sediment transport and sorting at bends. In S. Ikeda and G. Parker (Eds.) River Meandering, American Geophysical Union Water Resources Monograph, no. 12, Washington D.C., pp. 103-125.

Jiongxin, X. 1991. A study of river channel pattern information recorded by grain size parameters of fluvial sediment. Earth Surface Processes and Landforms, v. 16, pp. 129-142.

Keller, E.A. 1972. Development of alluvial stream channels, a five-stage model. Bulletin ofthe Geological Society o f America, v. 83, pp. 1531-1536.

Keller, E.A. 1983. Bed material sorting in pools and riffles. American Society o f Civil Engineers, Journal o f Hydraulics Division, v. 109, pp. 1243-1247.

Keller, E.A. and Mellhom, W.N. 1978. Rhythmic spacing and origin of pools and riffles. Geological Society o f America Bulletin, v. 89, pp. 723-730.

Keown, M.P., Dardeau, E.A., Jr., and Causey, EM. 1986. Historic trends in the sediment flow regime of the Mississippi River. Water Resources Research, v. 22, pp. 1555-1564.

Kesel, R.H. 1988. The decline in the suspended load of the lower Mississippi River and its influence on adjacent wetlands. Environmental Geology and Water S c ie n c e ,v. 11, pp. 271-291.

Kesel, R.H. 1989. The role of the Mississippi River in wetland loss in Southeastern Louisiana, U.SA.. Environmental Geology and Water Science, v. 13, pp. 183-193.

Kesel, R.H., Yodis, E.G., McCraw, D.J. 1992. An approximation of the sediment budget of the Lower Mississippi River prior to major human modification. Earth Surface Processes and Landforms, v. 17, pp. 711-722.

Kesel, R.H., Dunne, K.C., McDonald, R.C., Allison, K.R., and Spicer, B.E. 1974. Lateral erosion and overbank deposition on the Mississippi River in Louisiana caused by 1973 flooding. G eo lo g y, September, pp. 461-464.

138

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kolb, C.R. 1963. Sediments forming the bed and banks of the lower Mississippi River and their effect on river migration. Sedimentology, v. 2, pp. 227-234.

Knighton, A. D. 1981. Asymmetry of river channel cross-sections, Part I, quantitative indices. Earth Surface Processes and Landforms, v. 6, pp. 581-588.

Knighton, A. D. 1982. Asymmetry o f river channel cross-sections, Part D, mode of development and local variation. Earth Surface Processes and Landforms, v. 7, pp. 117-131.

Knighton, A.D, 1987. River channel adjustment - the downstream dimension. In K.S. Richards (Ed.), River Channels: Environment and Process. Oxford, U.K., Blackwell, pp. 95-128.

Knighton, A.D. 1998. Fluvial Forms and Processes. Edward Arnold, Baltimore, MY, 383 p.

Knighton, A.D. and Nanson, G.C. 1993. Anastomosis and the continuum of channel pattern. Earth Surface Processes and Landforms, v. 18, pp. 613-625.

Konditerova, EA. and Popov, I.V. 1966. Relation between changes in the horizontal and vertical characteristics of river channels. Soviet Hydrology, v. 5, pp. 515- 527.

Krinitzsky, E.L. 1965. Geological bfluences on Bank Erosion Along Meanders o f the Lower M ississippi River. U.S. Army Corpsof Engineers, Waterways Experiment Station, Reports 12-15,30 p.

Laczay, I.A. 1977. Channel pattern changes of Hungarian rivers: the example of the Hemad River. In K.J. Gregory (Ed.) River Channel Changes, John Wiley & Sons, New York, pp. 265-280.

Lane, E.W. 1957. A study o f the Shape o f Channels Formed by Natural Streams Flowing in Erodible Material. U.S. Army Corps of Engineers, Missouri River Division Sediment Services, v. 9,106 p.

Langbein, W.B. and Leopold, L.B. 1966. River Meanders - Theory o f Minimum V ariance. U.S. Geological Survey Professional Paper 262, pp. 782-794.

Langbein, W.B., and Leopold, L.B. 1968. River Channel Bars and Dunes- Theory o f Kinematic Waves. U.S. Geological Survey Professional Paper, 422-L, 20 p.

Lawler, D.M. 1993. The measurement of river bank erosion and lateral channel change: a review. Earth Surface Processes and Landforms, v. 18, p. 777-821.

139

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Lawler, D.M. 1995. The impact of scale on the processes of channel-side sediment supply: a conceptual model. Effects of Scale on Interpretation and Management of Sediment and Water Quality, International Association o f Hydrological Sciences Publication no. 226, pp. 175-184.

Leopold, L.B., and Maddock, TJr. 1953. The Hydraulic Geometry o f Stream Channels and Some Physiographic Implications. U.S. Geological Survey Professional Paper 252,57 p.

Leopold, L.B., and Wolman, M.G. 1957. River Channel Patterns: Braided, Meandering and Straight. U.S. Geological Survey Professional Paper, 282-B, 81 p.

Leopold, L.B., and Wolman, M.G. 1960. River meanders. Bulletin o f the Geological Society o f America, v. 71, pp. 769-794.

Leopold, L.B., Wolman, M.G., and Miller, J.P. 1964. Fluvial Processes in Geomorphology. W.H. Freeman, San Francisco, CA, 522 p.

Lewin, J. 1977. Channel pattern changes. In K.J. Gregory (Ed.), River Channel C hanges, John Wiley & Sons, New York, pp. 167-184.

Maddock, T. 1969. The Behavior o f Straight Open Channels With Movable B ed s. U.S. Geological Survey Professional Paper 622A.

McEwen, L.J. 1994. Channel planform adjustment and stream power variations on the middle River Coe, Western Grampian Highlands, Scotland. C atena, v. 21, pp. 357-374.

Meade, R.H. and Parker, R.S. 1985. Sediment in rivers o f the U.S. In National Water Summary 1984, U.S. Geological Survey Water-Supply Paper 2275, pp. 49-60.

Miller, T.A. and Onesti, L.J. 1979. The relationship between channel shape and sediment characteristics in the channel perimeter. Geological Society o f America B u lle tin v. , 90, pp. 301-304.

Milne, J. A. 1979. The morphological relationships off bends in confined stream channels Upland Britain. In A.F. Pitty (Ed.) Geographical Approaches to Fluvial P ro cesses, Norwich, Geobooks, pp. 241-260.

Milne, J.A. 1982. Bed-material size and the riffle-pool sequence. Sedimentology, v. 29, pp. 267-278.

Mississippi River Commission 1880. Hydrographic Surveys o f the Lower M ississippi River, Cairo, IL to Head of Passes, LA, Vicksburg, MS, 84 plates.

140

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Mississippi River Commission 1911. Hydrographic Surveys o f the Lower M ississippi River, Cairo, IL to Head of Passes, LA, Vicksburg, MS, 84 plates.

Mississippi River Commission 1921. Original Tabulation o f Bankfu.il Cross- Sections o f 1911 Hydrographic Surveys, Vicksburg, MS.

Mississippi River Commission 1930. Topographic Map of the Lower M ississippi Alluvial Valley, 5 ft contour interval, scale 1:500,000, U.S. Army Corps of Engineers, Vicksburg, MS.

Montgomery,. K. 1996. Sinuosity and fractal dimension of meandering rivers. A re a , v. 28.4, pp. 491-500.

Mosley, M.P. 1987. The classification and characterization of rivers. In K. Richards (Ed.) River Channels: Environment and Process. Basil Blackwell, Oxford, U.K., pp. 295-320.

Mossa, J. 1988. Discharge-sediment dynamics of the Lower Mississippi River. Transactions-Gulf Coast Association o f Geological Societies, v. 38, pp. 303-314.

Nanson, G.C. and Hickin, EJ. 1986. A statistical analysis of bank erosion and channel migration in western Canada, Geological Society o f America Bulletin, v. 97, p.497-504.

Nordin, C.F. and Queen, B.S. 1992. Particle Size Distributions o f Bed Sediments Along the Thalweg o f the M ississippi River, Cairo, IL to Head o f Passes, LA, September 1989. U.S. Army Corps of Engineers Waterways Experiment Station, Potamology Program, Report 7, Vicksburg, MS, 36 p.

Ockerson, J.A. 1892. Report on caving banks from Cairo to Donaldsonville with tabulated results and plats. House Documents, 50, 61st Congress, 1st Session, 3110-3111.

O’ Neil, M.P. and Abrahams, AD. 1984. Objective identification of pools and riffles. Water Resources Research, v. 20, pp. 291-296.

Orlowski, LA.., Schumm, SA., and Mielke, P.W. Jr. 1995. Reach classifications of the Lower Mississippi River. Geomorphology, v. 14, pp. 221-234.

Petts, G.E. 1995. Changing river channels: the geographical tradition. In A. Gumell and G. Petts (Eds.), Changing River Channels, John Wiley & Sons, New York, pp. 1-25.

Richards, K.S. 1976. Channel width and the riffle-pool sequence, G eo lo g ica l Society o f America Bulletin, v. 87, pp. 883-390.

141

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Richards, K.S. 1978. Simulation of flow geometry in the riffle-pool sequence. Geografiska Annaler, v. 60A , pp. 23-27.

Richards, K.S. 1982. Rivers: Form and Process in Alluvial Channels, Methuen NY, 361 p.

Roher, W.L.M. 1984. Effects of flow and bank material on meander migration in alluvial rivers. In C.M. Elliot (Ed.), River Meandering, New Orleans, American Society of Civil Engineers, pp. 770-782.

Rosgen, D.L. 1994. A classification of natural rivers. C a ten a , v. 22, pp. 169-99.

Russell, RJ., Howe, H.Y., McGuirt, J.H., Dohm, C.F., Hadley, W., Kniffen, FJB., and Brown, C A 1936. Lower M ississippi River Delta, Reports on the Geology o f Plaquemines and St. Bernard Parishes, Geological Bulletin, no. 8, Louisiana Geological Survey, New Orleans.

Saucier, R.T. 1974. Quaternary Geology o f the Lower Mississippi Valley, Research Series, no. 6, Arkansas Archeological Survey, Fayetteville, AR.

Saucier, R.T. 1994. Geomorphology and Quaternary Geologic History o f the Lower Mississippi Valley. U.S. Army Engineers Waterways Experiment Station, Vicksburg, MS, 364 p.

Schumm, SA . 1960. The Shape o f Alluvial Channels in Relation to Sediment T yp e. U.S. Geological Survey Professional Paper 352-B, pp. 17-70.

Schumm, SA . 1963. Sinuosity o f alluvial rivers on the Great Plains. G eo lo g ica l Society o f America Bulletin, v. 74, pp. 1089-1100.

Schumm, SA. 1977. Geomorphic thresholds and complex response of drainage systems. In S.A Schumm (Ed.), Drainage Basin Morphology, 353 p.

Schumm, SA. 1980. Some applications of the concept of geomorphic thresholds. In D.R. Coates and J.D. Vitex (Eds.), Thresholds in Geomorphology, George Allen & Unwin, London, p. 473-485.

Schumm, SA . 1991. To Interpret the Earth: Ten Ways to be Wrong. Cambridge University Press, 133 p.

Schumm, SA. and Kahn, H.R. 1972. Experimental study of channel patterns. Geological Society o fAmerica Bulletin, v. 83, pp. 1755-1770.

142

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Schumm, S.A. and Lichty, R.W. 1965. Time, space, and causality in geomoiphology.American Journal o f Science, v. 263, pp. 110-119.

Schumm, S.A. Rutherford, IX)., and Brooks, J. 1994. Pre-cutoff morphology of the Lower Mississippi River. In S A. Schumm and B.R. Winkley (Eds.), The Variability o f Large Alluvial Rivers, American Society of Engineers Press, New York, NY, pp. 13- 44.

Sear, D.A. 1996. Sediment transport processes in pool-riffle sequences. E a rth Surface Processes and Landforms, v. 21, pp. 241-262.

SPSS 1998. SPSS Base 8.0 Applications Guide. SPSS Inc., Chicago, IL, 372 p.

Tanner, W.F. 1960. Helicoidal flow, a possible cause of meandering. J o u rn a l o f Geophysical Research, v. 65, no. 3, pp. 993-995.

Thompson, A. 1986. Secondary flows and the pool-riffle unit: a case study of the processes of meander development Earth Surface Processes and landforms, v. 11, pp. 631-641.

Thome, C.R. and Osman, AM. 1988. The influence of bank stability on regime geometry of natural channels. In W.R. White (Ed.), International Conference on River R eg im e, Hydraulics Research Limited, John Wiley & Sons, pp. 135-147.

Thome, C.R. and Lewin, J. 1979. Bank processes, bed material movement and planform development in a meandering riven In D.D. Rhodes and G.P. Williams (Eds.), Adjustments o fthe Fluvial System, Kendall Hunt Dubuque, Iowa, pp. 117-137.

Thome, C.R. and Tovey, NJK. 1981. Stability of composite River banks. E a rth Surface Processes and Landforms, v. 6, pp. 469-484.

Toney, V.H. HI, Dunbar, J.B., and Peterson, R.W. 1988. Retrogressive Failures in Sand Deposits o f the Mississippi River, Report 1; Field Investigations, Laboratory Studies, and Analysis o f the Hypothesized Failure Mechanism. U.S. Army Corps of Engineers Waterways Experiment Station, Vicksburg, MS, 98 p.

Twidale, C.R. 1966. Erosion of an alluvial bank at Birdwood, South Australia. Zeitschriftfur Geomorphologje, v. 8, pp. 189-211.

U.S. Army Corps of Engineers 1935. Studies o f River Bed Materials and Their Movement With Special Reference to the Lower Mississippi River. Waterways Experiment Station Paper 17,161 p.

van den Berg, J.H. 1995. Prediction of alluvial channel pattern of perennial rivers. Geomorphology, v. 12, pp. 259-279.

143

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Walters, W.W. Jr. 1975. Regime Changes o f the Lower Mississippi River, Conception, Construction, and River Response. Potamology Investigations Report, 330- 2, U.S. Army Engineer District, Vicksburg, MS, 209 p.

Watson, C.C. 1982. An assessment of the Lower Mississippi River below Natchez, MS. PhJD. dissertation, Colorado State University, Fort Collins, Colorado.

Williams, G.P. 1986. River meanders and channel size. Journal o f Hydrology, v. 88, pp. 147-164.

Winkley, B.R. 1970. Influence of geology on regimen of a river. American Society of Civil Engineers National Water Resources Engineering Meeting, Memphis, TN, 35 p.

Winkley, B.R. 1977. Man-Made Cutoffs on the Lower Mississippi River, Conception, Construction, and River Response. U.S. Army Corps of Engineers, Vicksburg District, 209 p.

Winkley, B.R. 1982. Response of the Lower Mississippi River to river training and realignment In R.D. Hey, J.C. Bathhurst and C.R. Thome (Eds.), G ravel-B ed R iv e rs, Wiley, Chichester, U.K., pp. 659-681.

Winkley, B.R. 1994. Response of the Lower Mississippi River to flood control and navigation improvements. In S.A. Schumm and B.R. Winkley (Eds.), The Variability o f Large Alluvial Rivers, American Society of Engineers Press, New York, NY, pp. 45-74.

Wohl, E.E., Vincent, K.R., and Merritts, DJ. 1993. Pool and riffle characteristics in relation to channel gradient Geomorphology, v. 6, pp. 99-110.

Yang, T.Y. 1971. Formation of riffles and pools. Water Resources Research, v.7, no.6, pp. 1567-1574.

144

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. VITA

Although Paiil Franklin Hudson was bom in Eau Claire, Wisconsin on April 29,

1968, he considers himself a southerner. After living in Iowa until the age of ten, his

family moved to Jacksonville, Florida, the Sunshine State, in 1978. In Jacksonville

Paul attended Terry Parker High School, and graduated in 1986. The following fall he

entered college at Jacksonville University, where his father is a professor of music. His

decision to study geography late in his junior year would be a turning point in his life.

Upon graduation with honors from Jacksonville University in 1991, he attended

graduate school in geography at the University of Florida, and studied with Joann

Mossa (an L.S.U. alumna), where he developed an interest in fluvial geomorphology.

After graduating with a Master of Science degree in August of 1993, he moved to the

Bayou State to pursue a doctorate in geography at Louisiana State University, with an

emphasis in fluvial geomorphology. This spring he accepted an academic position at

the University of Texas at Austin, where he will move upon graduating from L.S.U. in

August of 1998 with the degree of Doctor of Philosophy in geography.

145

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DOCTORAL EXAMINATION AND DISSERTATION REPORT

Candidate: Paul Franklin Hudson

Major Field: Geography

Title of Dissertation: Meandering Processes and Channel Morphology in the Lower Mississippi River Prior to Major Human Disturbance

Approved: fLj~j / / • Major^vmm Professor

EXAMINING COMMITTEE: Jijri'i ft-

, % ■ c ^ kl <-

(Z lA ^bU A ,/ 9 ____ 7 /

Date of Kxamination:

July 2nd, 1998

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. IMAGE EVALUATION

1.0 s£ U2^ mm 2.2 3.6 £ 2.0 l.l 1.8 i m 1.25 II 1.4 1.6

150mm

IM/4GE. Inc 1653 East Main Street Rochester. NY 14609 USA Phone: 716/482-0300 Fax: 716/288-5989

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.