i MORPHOLOGICAL CHANGES IN CHANNELS DUE TO CYCLONE GENERATED HYDRODYNAMIC SHOCK BY MD. WASIF-E-ELAHI MASTER OF SCIENCE IN WATER RESOURCES DEVELOPMENT INSTITUTE OF WATER AND FLOOD MANAGEMENT BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY JUNE 2016
105
Embed
MORPHOLOGICAL CHANGES IN CHANNELS DUE TO CYCLONE …
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
i
MORPHOLOGICAL CHANGES IN CHANNELS DUE TO CYCLONE
GENERATED HYDRODYNAMIC SHOCK
BY
MD. WASIF-E-ELAHI
MASTER OF SCIENCE IN WATER RESOURCES DEVELOPMENT
INSTITUTE OF WATER AND FLOOD MANAGEMENT
BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY
JUNE 2016
ii
Morphological Changes in Channels due to Cyclone
Generated Hydrodynamic Shock
A thesis by
MD. WASIF-E-ELAHI
Submitted in partial fulfillment of the requirements for the degree of
MASTER OF SCIENCE IN WATER RESOURCES DEVELOPMENT
Institute of Water and Flood Management
BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY
June 2016
iii
iv
CANDIDATE’S DECLARATION
v
Dedicated to My parents
vi
ACKNOWLEDGEMENT
In the first place, I would like to thank the Almighty Allah for giving me the ability to complete
this research work. I would like to express my sincere and heartfelt gratitude to my supervisor
Dr. Mohammad Anisul Haque, Professor, IWFM, BUET, Dhaka for his constant guidance,
valuable advice, generous help and constructive discussion to carry out this research. I feel
proud and lucky to work with him. His keen interest in the topic and enthusiastic support for
my effort was a source of inspiration to carry out the study. I also express gratefulness to the
present Director Dr. Mashfiqus Salehin and former Director Dr. Md. Munsur Rahman, IWFM,
BUET, Dhaka, Dr. Rezaur Rahman, Professor, IWFM, BUET, and Malik Fida Abdullah Khan,
M.Sc., Deputy Executive Director (Operation), CEGIS, Dhaka for their comments on thesis
work that helped me a lot to improve the quality of the thesis.
I want to express my sincere gratitude to Dr. Md. Munsur Rahman, Team leader,
DEltas, Vulnerability and Climate Change: Migration and Adaptation (DECCMA) Project and
the modelling team of the project who have helped by providing valuable information in
different stages of my research work.
I would like to extend my thanks to Sumaiya, Tamanna Kabir, Mohiuddin Sakib, Fatin
Nihal and Tariq Omarr, M.Sc. students and PhD candidate at IWFM, BUET, who have helped
in different stages of my research work.
I would like to give thanks to all staff members of IWFM and all my friends for helping
and inspiring me in different ways.
I am very much indebtedness to my heavenly father whose encouragement and support
was a continuous source of inspiration for my higher study.
Finally, I wish to extend my gratitude to my family members for their moral support during my
research work.
vii
Abstract
River stability and response to changing environmental conditions are highly dependent on local watershed features and exposed environmental condition. During hydrodynamic shock like severe cyclones, stream power increases above normal levels, resulting in dramatic changes in the riverine landscape. In the natural landscape, cyclone generated storm surges can play a major geomorphic role, especially for sediment transport in channels and deposition on floodplains. The geographical location and climatic condition make Bangladesh one of the most cyclone prone countries in the world. For Bangladesh coast, not much information is available on changes in river morphology due to this phenomenon. To fill this research gap - present study aims to develop a semi-analytical model to assess the morphological changes in channels due to hydrodynamic shocks and apply the model for a cyclonic event in Bangladesh coast.
Changes in channel morphology depend on the channel characteristics. In this study channels are classified based on channel conveyance. During cyclone, the morphological changes in channel are mainly controlled by bed shear stress. To develop the semi-analytical model, width depth ratio is used as an indicator of channel morphology. From the Manning’s equation, the relation between channel geometrical shape factor (which is the ratio of channel conveyance and roughness) and channel width depth ratio is developed. Afterward width depth ratio is expressed as a function of channel geometrical shape factor. It is found that conveyance of a channel is directly proportional to the bed shear force. By introducing proportionality constant and assuming that change of channel conveyance due to bed shear force varies linearly, the non-dimensional relation between channel conveyance and bed shear force is established. Later the channel conveyance is replaced by channel geometrical shape factor in the non-dimensional relation. By substituting the width depth ratio with the function of geometrical shape factor in the non-dimensional relation and re-arranging, the semi-analytical model is developed.
Magnitudes of variables in the semi-analytical model are determined using the Delft 3D model simulations for different cyclone scenarios which are termed as hydrodynamic shocks. To determine the coefficients of the semi-analytical model, these variables are used. Delft Dashboard is applied to incorporate the cyclonic condition in the Delft 3D model simulations. Performance of the semi-analytical model is evaluated by comparing the model results with the Delft 3D model simulation. Performance of the model is found to be within ±0.5% to ±5% of error margin.
viii
Table of Content
Page No.
Certificate of Approval iii
Declaration iv
Acknowledgment vi
Abstract vii
Table of Contents viii
List of Figures xi
List of Tables xiv
Abbreviations xv
List of Symbols xvi
Chapter 1. Introduction
1.1 Brief Background of the Study
1.2 Rationale of the Study
1.3 Objective of the Study
1.4 Outline of the thesis
1
3
4
4
Chapter 2. Literature Review 6
2.1 River Classification and Morphology
2.2 Factors Governing the Morphological Changes of Channel
2.2.1 Response of channel
2.2.2 Change of channel in large scale
2.2.3 Thresholds for morphological change
2.2.4 Change of morphological characteristics during
cyclonic condition
2.3 Cyclone in Bangladesh
2.3.1 General
2.3.2 Reasons of severity of storm surges
2.3.3 Impacts of major cyclone on river morphology along
Bangladesh coast
2.4 Modelling of river morphology and application
2.4.1 Coupling of storm surge and morphology models
2.4.2 Model selection for the study: Delft 3D and Delft
6
9
9
12
13
15
15
17
17
18
19
26
26
29
ix
Dashboard
Chapter 3. Development of Semi-analytical Model 31
3.1 Channel Conveyance
3.2 Bed Shear Force and Change in Channel Morphology
3.3 Development of Semi-Analytical Model
31
33
37
Chapter 4. Computation of Coefficients for the Semi-Analytical Model 39
4.1 Application of Numerical Model
4.1.1 Model parameters
4.1.2 Model grids
4.1.3 Model bathymetry
4.1.4 Model boundary conditions
4.1.5 Model validation
4.1.6 Numerical model result of cyclone SIDR
4.2 Channel Classification
4.3 Generating the Scenarios
4.4 Incorporating the Cyclone in Delft 3D
4.5 Computing the Variables for the Semi-Analytical Model
4.5.1 Calculating the cross sectional area and wetted perimeter
4.5.2 Calculation of bed shear force
4.5.3 Generating relation between geometric shape factor and
bed shear force
4.5.4 Computation of coefficients
39
39
40
41
41
41
43
46
47
48
49
49
50
52
54
Chapter 5. Application of Semi-Analytical Model 57
5.1 Application of Semi-Analytical Model for Different Scenarios
5.2 Application of Semi-Analytical Model for cyclone SIDR
5.3 Sensitivity of Model Results
5.4 Application of the Semi-Analytical Model without the Application
of Delft 3D Model
5.5 Comparison of the Model Results with Measured Cross Section
57
60
62
67
73
Chapter 6. Conclusions and Recommendation 76
6.1 Conclusions
6.2 Limitations of the Model
76
76
x
6.3 Recommendations for Future Study 77
References 78
Appendix 85
xi
List of Figures
Figure No. Page No.
Figure 2.1 Channel pattern as a function of channel slope and bankfull
discharge
7
Figure 2.2 Independent and dependent variables impact on channel
morphology
10
Figure 2.3 Spatial and temporal scales of channel response variables in
alluvial rivers
11
Figure 2.4 Work and scale of channel change accomplished for flood
events of different magnitude and duration
12
Figure 2.5 Channel change depends on the overlap between frequency
distributions of driving and resisting forces for different
scales of morphologic response
14
Figure 2.6 Cyclone tracks along Bangladesh 18
Figure 2.7 Locations of embankment breaching in Polder-32 area 20
Figure 2.8 Comparison of satellite images of Nalian River between year
2008 and 2014
21
Figure 2.9 Picture of new narrow channel formation at weak point 23
Figure 2.10 Picture of new narrow channel formation at embankment
cutting point
24
Figure 2.11 Breaching of Embankment at Kamarkhola canal 24
Figure 2.12 Newly constructed sluice gate at Kamarkhola canal 25
Figure 4.1 Domain of the large model (a) and locations of the model
validation (b)
43
Figure 4.2 Magnitude of bed shear stress at the time of landfall for
cyclone SIDR
44
Figure 4.3 Magnitude of flow velocity at the time of landfall for
cyclone SIDR
45
Figure 4.4 Resultant erosion/sedimentation at the time of landfall for
cyclone SIDR
45
Figure 4.5 Relation between geometric shape factor G and WDR for
different dimensions of channels
46
xii
Figure 4.6 Hypothetical tracks of different strengths of cyclones 49
Figure 4.7 A typical channel cross-section 50
Figure 4.8 Variation of bed shear force for cyclones with variable
strengths
51
Figure 4.9 Relation between geometrical shape factor and bed shear
force for channel with low conveyance
52
Figure 4.10 Relation between geometrical shape factor and bed shear
force for channel with medium conveyance
53
Figure 4.11 Relation between geometrical shape factor and bed shear
force for channel with high conveyance
53
Figure 4.12 Relation between geometrical shape factor and bed shear
force for channel with very high conveyance
54
Figure 4.13 Schematic representation of ranges of coefficient i 55
Figure 4.14 Schematic representation of ranges of coefficient j 55
Figure 4.15 Schematic representation of ranges of coefficient a 56
Figure 4.16 Schematic representation of ranges of coefficient b 56
Figure 5.1 Variation of WDR with time for low conveyance channel
when cyclone condition is HSC
58
Figure 5.2 Variation of WDR with time for medium conveyance
channel when cyclone condition is VHSC
58
Figure 5.3 Variation of WDR with time for high conveyance channel
when cyclone condition is VHSC
59
Figure 5.4 Variation of WDR variations with time for very high
conveyance channel when cyclone condition is VHSC
59
Figure 5.5 Temporal variation of measured bed shear force for the
channel where the semi-analytical model is applied
60
Figure 5.6 Variation of bed shear stress and flow velocity as a function
of non-dimensional bed shear force
61
Figure 5.7 Comparison of model WDR with measurements for VHSC 62
Figure 5.8 Variation of WDR with different values of i for very high
conveyance channel during VHSC
63
Figure 5.9 Variation of WDR with different values of j for very high
conveyance channel during VHSC
63
Figure 5.10 Variation of WDR with different values of a for very high 64
xiii
conveyance channel during VHSC
Figure 5.11 Variation of WDR with different values of b for very high
conveyance channel during VHSC
64
Figure 5.12 Variation of WDR with time after applying modified
coefficients
65
Figure 5.13 Time series of measured non-dimensional bed shear force at
the mouth of Baleswar estuary during cyclone SIDR
66
Figure 5.14 Variation of WDR with non-dimensional bed shear force at
the mouth of Baleswar estuary.
66
Figure 5.15 Comparison of time series of WDR between the model and
the measurements at the Baleswar mouth
67
Figure 5.16 Model application when water depth and depth average
velocity are known
68
Figure 5.17 Variation of bed shear stress as a function of depth average
velocity
69
Figure 5.18 Model application when only velocity of flow is known 70
Figure 5.19 Steps of the model application when no data is available 72
Figure 5.20 Temporal variation of WDR during cyclone SIDR at
Bishkhali mouth
72
Figure 5.21 Variation of WDR with bed shear force during cyclone SIDR
at Bishkhali mouth.
73
Figure 5.22 Change of planform in Bishkhali estuary due to cyclone
SIDR
74
Figure 5.23 Comparison of the semi-analytical model result with the
observed planform
75
xiv
List of Tables
Table No. Page No.
Table 4.1 Delft 3D morphology model parameters 39
Table 4.2 Reliability of Delft 3D morphology model 42
Table 4.3 Channel classification based on conveyance 46
Table 4.4 Channel properties that are used to generate the scenario of
model runs
47
Table 4.5 Classification of different intensities of cyclone 48
Table 4.6 Computed coefficients values for different channel types 55
Table 4.7 Ranges of coefficients variations for different channel types 55
Table 5.1 Coefficient values of the semi-analytical model 57
xv
ABBREVIATIONS
IWFM Institute of Water and Flood Management
BWDB Bangladesh Water Development Board
MoE Ministry of Environment
CEGIS Center for Environment and Geographic Information Service
IWM Institute of Water Modelling
WB World Bank
BUET Bangladesh University of Engineering and Technology
BMD Bangladesh Meteorological Department
xvi
LIST OF SYMBOLS
V = Mean velocity of flow (m/s)
R = Hydraulic Radius (m)
S = Slope of channel
Q = Flow discharge (m3/s)
A = Cross-sectional area of channel (m2)
n = Manning's roughness coefficient
K = Conveyance of channel (m3/s)
G = Geometrical shape factor of conveyance
B = Width of the channel (m)
D = Depth of the channel (m)
휏 = Bed shear stress (N/m2)
훾 = Specific weight of water (1000 kg/m3)
P = Wetted perimeter (m)
F = Bed shear force (N)
WDR = Width depth ratio
a,b,i,j = Coefficients
휌 = Density of water (kg/m3)
푓 = Friction factor
H = Water depth (m)
LSC = Low strength cyclone
MSC = Medium strength cyclone
HSC = High strength cyclone
VHSC = Very high strength cyclone
1
CHAPTER ONE
INTRODUCTION
1.1 Brief Background of the Study
River morphology deals with the interaction between flowing water in rivers
and their environment. River stability and response to changing environmental
conditions are highly dependent on local watershed features and exposed environmental
condition (Buffington, 2012). At larger spatial and temporal scales, altered
environmental conditions may cause changes in channel properties (Montgomery and
Buffington, 1997) and planform morphology. Possible change in channel type can be
presented in terms of slope, confinement, discharge and sediment supply. In some
cases, this change can be related to the specific process like coastal hydrodynamic
process (Montgomery, 1999). By Leopold and Wolman (1957), the shape and pattern of
a natural channel is governed by the combined effects of climate, rocks and
physiography of the region where the channel is located. Channel width is largely
determined by effective discharge (Biedenharn et al., 2000) which is bankfull discharge.
Changes of channel are controlled by the discharge and sediment load provided by the
drainage basin (Leopold and Wolman, 1957). This change of channel is directly linked
to bed mobility and can be evaluated by comparing the bankfull shield stress to the
critical value for incipient motion of the median grain size (Buffington, 2012).
Tropical storm surge, large floods, landslides and earthquake can produce
significant geomorphic features. In contrast, smaller floods in the Valley and Ridge can
produce such erosional and depositional features that probably require thousands of
years to overcome the effects (Jacobson et al., 1989). During hydrodynamic shock like
severe cyclones, stream power increases above normal levels, resulting in dramatic
changes in the riverine landscape (Gupta, 2000; Terry et al., 2008). In the natural
landscape, cyclone-induced storm surges also play a major geomorphic role, especially
for sediment transport in channels and deposition on floodplains (Gupta, 1988; Terry et
al., 2002; Kostaschuk et al., 2003). Tropical cyclones are one of the most devastating
natural disasters (Mohanty et al., 2014; Pattanayak et al., 2014) globally. Tropical
cyclones (also known as hurricanes or typhoons) can generate extreme floods in parts of
the tropics and subtropics between about 10° and 30° of latitude. Accounts of the
resulting high shear stress and unit stream power, enhanced stream conveyance,
2
sediment transport and storage, and channel forms are now available in nature for a
limited number of streams in Australia, South Asia, and the Caribbean. These extreme
floods generated from tropical cyclones tend to occur in a regular interval of decades
rather than return periods of 50 years or 100 years flood. Such tropical cyclones
generated extreme floods, which occur on the decadal scale, can determine the size and
coarse sediment of many river channels in the hurricane-affected areas. Other high
flows, which are relatively smaller and occur between the storm surge induced large
floods, build spatial morphological features in such channels. The final form of affected
channel, therefore, is a function of three different parameters- sizes and frequencies of
floods, and sediments (Gupta, 2000). Modified channels of this type have been reported
from many areas: northeastern Australia (Wohl, 1992 as cited in Gupta, 2000), India
(Gupta, 1995 as cited in Gupta 2000), and the Greater Antilles group of islands in the
Caribbean (Ahmad et al, 1993 as cited in Gupta, 2000). Channel formation and
preservation depend on the magnitude of flood in arid region like arid central Australia
(Bourke and Pickup, 1999).
The geographical setting of Bangladesh makes the country vulnerable to natural
disasters (MoEF, 2005). This vulnerability is increased more due to its dense population
(Choudhury, 2007). Geographic setting and river morphology contribute to regular
disasters in Bangladesh. Almost three-fourths of Bangladesh border is surrounded by
mountains and hills, along with the funnel-shaped Bay of Bengal in the south. Such
geographic settings have caused of life-giving monsoon rains, but also, make it more
vulnerable to natural disasters. The major disasters those are concerned in Bangladesh
are the occurrences of flood, cyclone and storm surge, flash flood, drought, tornado,
riverbank erosion, and landslide (Hossain, 2008). Among these disasters, the cyclone is
considered as the major and most devastating to the human habitation of this country.
Of the 508 cyclones that have originated in the Bay of Bengal in the last 100 years, 17
percent have hit Bangladesh, amounting to a severe cyclone almost once every three
years. Of these, nearly 53 percent have taken more than five thousand lives
(Khalequzzaman, 1976 as cited in Sharbari, 2012). In the early 1960s for optimizing the
crop production, 37 polders, and 282 sluice gates were constructed by Bangladesh
Water Development Board (BWDB) under the Coastal Embankment Project. Poldered
areas in southwest Bangladesh have lost 1.0–1.5m of elevation, whereas the
neighboring Sundarban mangrove forest has remained comparatively stable (Auerbach
et al., 2015). One major consequence of this elevation loss occurred in 2009 when
3
several polders (e.g. polder 5, 32) were overtopped due to storm surge and breached out
during cyclone AILA. It caused inundation of large areas of land for up to two years
until embankments were repaired. During this period, tide-induced sedimentation rate
was higher than normal rate in newly connected channels. According to Auerbach et al
(2015), newly shaped landscape received tens of centimeters of tidally deposited
sediment, equivalent to decades’ worth of normal sedimentation. Cyclone AILA caused
damage to an already weakened embankment system (due to lack of proper
maintenance and time being) and washed away 1,742km of embankments, creating
regular flooding during high tide for over a year after the cyclone. Due to these reasons
evaluation of morphological change of channel in Bangladesh coast for cyclone induced
hydrodynamic shock is a vital task.
1.2 Rationale of the Study
Development of river morphology in an estuarine environment is a complex
phenomenon. Mechanism of sediment transport is more complex during the
hydrodynamic shock event like cyclone generated storm surge. The Ganges-
Brahmaputra-Meghna (GBM) delta is one of the most dynamic tide-dominated deltas in
the world. The coastal regions of Bangladesh are subject to severe cyclones almost
every year. During cyclone AILA in 2009, a significant change was observed in river
planform in the area of polder 32. Currently many modeling suites like Mike21, Delft-
3D, SOBEK, SMS etc. are available for the simulation of river morphodynamics and
hydrodynamics. These are all numerical models that need expert knowledge and special
setup of computing base to simulate natural phenomena related to river morphology.
As a result, it is not possible to quick assessment of morphological change with limited
information due to hydrodynamic shock generated from cyclone generated storm surge.
A quick assessment of morphological changes due to storm surge generated
hydrodynamic shock is necessary to identify morphologically vulnerable river location
after or before a storm surge. This information will add value to emergency planning
during disaster management. This study is expected to contribute to fill the gap in
scientific knowledge in this particular area.
4
1.3 Objectives of the Study
The main objectives of this research are:
1. To identify the driving parameters of physical processes causing changes in
channel morphology.
2. To develop a semi-analytical model to assess morphological changes in
channels due to cyclone generated hydrodynamic shock.
3. To apply the model for a cyclonic event in Bangladesh coast.
Outcome of the study is the semi-analytical model that can be used to compute
morphological changes of rivers/ estuaries due to any cyclonic event (for example storm
surge) with minimum information and relatively simple calculations
1.4 Outline of the thesis
The first chapter of the study gives a brief presentation on the channel
morphodynamics and different hydrodynamic shock events. It emphasizes the exposure
of cyclone generated storm surge as hydrodynamic shock on Bangladesh coast and need
of proper understanding the channel response before and after the hydrodynamic shock.
It also includes the objectives of the current study.
The second chapter consists of the available information and studies that has
been used for this research to achieve the objectives. Literature reviews have been
summarized in this chapter on: (i) River classification and Morphology; (ii) Factors
governing the morphological changes of channel; (iii) Cyclone in Bangladesh and (iv)
Modeling of river morphology and application.
The third chapter describes the semi-analytical model development to assess
morphological changes in channels due to cyclone generated hydrodynamic shock (e.g.
storm surge).
5
Chapter four presents the methodology to compute the coefficients of the semi-
analytical model. This chapter also describes the application of the numerical which is
required to generate the required data to compute the model coefficients.
Chapter five describes the application of the semi-analytical model for a wide
range of scenarios. Finally chapter six describes the conclusions of the study. This
chapter also describes the limitation of the semi-analytical model which is developed in
this study. Recommendations are suggested for further refinement of the model.
6
CHAPTER TWO
LITERATURE REVIEW
2.0 Introduction
River sedimentation and morphological process are a complex phenomenon in
nature. Morphological changes of the river due to hydrological regime and manmade
intervention is a regular process. The changing of morphological characteristics
depends on the type of the river. Different types of rivers act differently to the changing
boundary conditions and driving parameters like discharge, sediment load etc.
2.1 River Classification and Morphology
Channels can be classified based on several characteristics like cross-sectional
dimension, slope, degree of entrenchment, width to depth ratio, sinuosity and trend and
types of morphological change. Channels are classified in different ways by different
researchers to achieve specific goals. Most developed river classifications based on
channel pattern (i.e., planform geometry) are broadly divided into two approaches: (a)
quantitative relationships (which may be either empirical or theoretical) and (b)
conceptual frameworks (Buffington and Montgomery, 2013). Based on observed data
and experiment result, Leopold and Wolman (1957) classified channels into straight,
meander and braided channel by applying quantitative approach. Leopold and Wolman
(1957) developed a threshold between meandering and braided rivers (specified in
Figure 2.1). The changes in channel pattern were presented as a function of discharge
and channel slope. Several studies (Lane, 1957 as cited in Buffington and Montgomery,
2013; Leopold and Wolman, 1957) also specified that grain size, sediment load,
riparian vegetation, channel roughness, width and depth affect the channel morphology.
Different patterns of rivers are specified by applying conceptual framework in Schumm
(1985). Based on the nature of the materials through which a river flows, stream
channels are classified into bedrock, semi-controlled and alluvial. Behaviors of different
types of channels vary due to exposed environmental condition. Alluvial channels have
the most varying patterns and shifting nature due to alluvium characteristics. Alluvium
is eroded, transported and deposited with the change of sediment load and water
discharge.
7
Figure 2.1: Channel pattern (meandering, straight, braided) as a function of channel
slope and bankfull discharge (Buffington and Montgomery, 2013).
Brice et al. (1978 as cited in Schumm, 1985) recognized three basic types of alluvial
channels that are characterized by degrees of sinuosity, braiding and anabranching.
There were 11 sub-classes in quantitative classifications and 16 sub-classes in
qualitative classifications. Anatomising channels are unique from the anabranched
channels as the individual branches of the channel can be meandering, straight or
braided. From the laboratory study, Schumm and Khan (1972) represented straight,
meandering and braided in terms of sinuosity versus valley slope. Schumm (1977)
proposed a more useful approach to classify channel type from straight, through
meandering to braided channels with no abrupt breaks in between. While classifying the
channel type, there is a range of planform patterns together with the use of an
examination of the geomorphological features displayed by the channel. A generalized
relationship between sediment load, channel stability and channel form was presented
in Schumm (1977) .
According to Leopold and Wolman (1957), channel cross-section and pattern
are controlled by the discharge and sediment load provided by the drainage basin. They
8
showed the interaction of the several variables in channel shape from the observed
characteristics of the channel. From their observation, channel width is a function of
bankfull discharge, in combination with the characteristic resistance of bed and bank to
scour. Bed shear varies with the change of channel width. Larger width of channel
increases the shear on the bed at the expense of that on the bank which is reverse for
narrow width. During the high flow of discharge, width adjustment can take place
rapidly and with the erosion or deposition of relatively small volumes of debris, a
relative stable width at high flow is a primary adjustment. The inter-adjustments
between channel depth, flow velocity, slope and roughness tend to accommodate the
further stability of channel. Channel roughness is determined by the particle size which
is an independent factor related to drainage basin. Roughness is also a function of
characteristics of bed configuration in channels carrying fine materials. By Leopold and
Wolman (1957), a particular slope is associated with the roughness where roughness is
independently determined as well as discharge and sediment load. At the width
determined by the discharge, flow velocity and channel depth must be adjusted to
satisfy a quasi-equilibrium state in accord with the particular slope (Langbein and
Leopold, 1964). When roughness also is variable, depending on the changing bed
configuration, then a number of combinations of velocity, depth and slope will satisfy
equilibrium. These adjustments of several variables tending toward the establishment of
quasi-equilibrium in river channels drive to the different channel patterns observed in
nature. Leopold and Wolman (1957) separated meandering and braided rivers
depending on bankfull discharge and slope of the channel (Figure 2.1). It can be seen
that for a certain bankfull discharge river planform changes from meandering to braided
with increase in longitudinal slope of the channel. Against the complex range of driving
variables and boundary conditions for controlling the channel form, only two
parameters (slope and bankfull discharge) were considered to quantify the geomorphic
threshold between meandering and braided. But there is not an individual threshold that
actually exists.
From the above discussion, it is evident that discharge affects morphological
features heavily. For the current study, by applying quantitative approach, channels are
classified based on the conveyance capacity of the channel derived from the Manning's
equation (George and Schneider, 1989). Detail steps applied for channel classifications
are described in Chapter Four.
9
2.2 Factors Governing the Morphological Changes of Channel
There are many hydrological and morphological factors which can initiate the
morphological change of a channel. Channel accommodates these changes by
adjustment of its own properties like grain-size, width-depth, gradient etc ( Langbein
and Leopold, 1964). These adjustments may take short time (for example few minutes )
or long time (for example a century) depending on the initiating event (Buffington,
2012). Hydrological characteristics are important factor to control the river
morphological changes. During hydrodynamic shock like a cyclone, the change of river
planform depends on several hydrological factors like drainage network, sediment load,
upstream discharge etc.
2.2.1 Response of channel
Rivers are opened to changing environmental conditions over multiple spatial
and temporal scales. Response of the river is regulated to varying degrees by the
imposed environmental conditions and human activity. River morphology is controlled
by topography (valley slope and channel confinement), discharge (magnitude,
frequency and duration of runoff events), sediment supply (volume, capacity and
frequency of sediment delivery) , vegetation (bank strength, roughness) and in-channel
wood debris. According to Hogan and Luzi (2010), factors controlling channel
morphology are divided into independent variables (that are imposed on the watershed)
and dependent variables ( that adjust to the imposed conditions). Dependent variables
like sediment supply, discharge and vegetation are depending on the independent
variables like the geologic, climatic and human activities (Montgomery and Buffington,
1993; Buffington et al., 2003, as cited in Hogan and Luzi, 2010). The relation between
independent and dependent variables is specified in Figure 2.2. Channel morphology is
the result of the combined influence of dependent and independent landscape variables.
The channel responds to changes in these variables by adjustments in one or many of
the dependent channel variables (Figure 2.2).
10
Figure 2.2: Independent and dependent variables impact on channel
morphology. (Hogan and Luzi, 2010)
River stability and response to changing environmental conditions are highly
dependent on local context (e.g. channel type, the nature of imposed sediment,
hydrologic and vegetation regimes. imposed anthropogenic constraints, and the legacy
of past natural and anthropogenic disturbances). Rivers can show a broad range of
responses to changing inputs of water, sediment, and vegetation over human time
scales. Channel response may range from small- scale adjustment of channel
characteristics (grain size, width, depth) to large-scale alteration of reach morphology
and planform pattern.
Geology Climate Human
Frequency, volume and caliber of
sediment
Frequency, volume and duration of
streamflow
Riparian vegetation: bank stability and local
flow hydraulics, in channel large woody
debris (orientation and position)
Sediment supply Stream discharge Vegetation
Width, depth, bed slope, grain size, bedforms, sinuosity, scour depth
Independent Landscape Variables
Dependent Landscape Variables
Dependent Channel
Variables
11
Figure 2.3: Spatial and temporal scales of channel response variables in alluvial rivers
(Partly after Knighton, 1998, as cited in Buffington, 2012).
From Figure 2.3, it is evident that the width and depth is the common part of
grain size, bedforms and stream gradient. Changes in any part of channel response
variables have impact on the width and depth of the channel. According to Buffington
(2012), successive, overlapping, spatial and temporal scales of morphological response
in alluvial channels include:
(1) Grain-scale adjustment, comprising:
a) Local changes in grain size, packing, protrusion, and friction angle;
b) Development of micro-grain forms (e.g., particle clusters, stone cells)
c) Formation of textural patches (i.e., grain-size facies).
(2) Bedforms adjustment: Changes in the type, size, and frequency of bed topography,
ranging from micro-bed forms (e.g., ripples, bedload sheets) to macro-bed forms or
channel units (individual bar, pool, step, and riffle topography).
(3) Reformed channel geometry (e.g., changes in local cross-sectional width, depth, and
downstream variation of those features).
(4) Reformed stream gradient due to reach-scale aggradation/incision and changes in
channel sinuosity. Here, stream gradient is distinguished from valley gradient, which is
not adjustable in the short term.
12
Figure 2.4: Work and scale of channel change accomplished for flood events of
different magnitude and duration (Costa and O’Connor, 1995 as cited in Buffington,
2012). Total work (shaded area) is identical to the events, but the scale of channel
change is greater for the high-magnitude event.
2.2.2 Change of channel in large scale
Larger scales of channel response reflect the cumulative action of smaller-scale
processes, particularly sediment transport of bed and bank materials. Hence, a
progression of successive scales of response can be projected, with the grain-size
adjustment being the first order response (Figure 2.3). Furthermore, because alluvial
rivers show mutually adjusting channel characteristics, changes in any one parameter
can influence all of the others.
The magnitude of channel change that occurs for a disturbance depends on the
amount of work accomplished by the event [flood magnitude times duration (Wolman
and Miller, 1960)] and the time needed for a given scale of response to occur. Wolman
and Miller’s (1960) classic magnitude–frequency argument emphasizes the
effectiveness of frequent, moderate-sized events in accomplishing geomorphic work
over the long term, but large-scale changes in morphology require large events (Figure
2.4). Similarly, different temporal scales of disturbances (seasonal to centennial) will
exhibit characteristic scales of response. Seasonal changes are frequent, typically small-
magnitude events that will lead to similarly small degrees of channel change [e.g., bed
loosening and changes in grain-size structure and texture as seasonal floods begin;
(Milhous, 1973, as cited in Buffington, 2012)], while annual peak floods are typically
moderate-sized events that lead to moderate scales of channel change (e.g., altered
cross-sectional widths and depths). Over decadal scales, rare infrequent events (e.g.,
13
50–100-year floods or debris flows) may cause significant channel change followed by
a period of relaxation from the disturbance (recovery or attainment of some new
equilibrium state; Bull, 1991, Simon and Rinaldi, 2006 as cited in Buffington, 2012).
From the above, it appears that the scale of channel change depends on the event
size and frequency. Total change of channel is dependent on the capacity of work
accomplishment by the event. In this study, time lengths of the cyclonic events are
considered from the historical cyclone events of Bangladesh coast.
2.2.3 Thresholds for morphological change
Geomorphic systems are generally nonlinear and depending largely on their
threshold dominated nature (Phillips, 2006). Geomorphic thresholds can be defined as
the condition at which there is a significant landform change. These thresholds are
classified into intrinsic (Changes will take place without changing the external variable)
and extrinsic (Response of a system to external influences occurs at) thresholds
(Schumm, 1979). Rivers are subjected to thresholds that define significant changes in
processes and morphology and distinguish typical riverine landscapes and habitats.
These thresholds are set by the conditions that govern river channel process and form,
amongst which the most important factors are the flow regime, the quantity and caliber
of sediment delivered to the channel, and the topographic setting (which determines the
gradient of the channel) (Church, 2002). These factors control the sediment transport
regime and the character of alluvial deposits along the channel.
Changes of channel occur systematically along the drainage system as flow,
gradient and sediment character change. So a characteristic sequence of morphological
and habitat types are found in channel. The sequence is closely related to stream
capability to move sediment and with bank stability. The riverine landscape is affected
seasonally by flow thresholds and human actions frequently dictate the character of the
riverine landscape. Human activity can accelerate thresholds crossing which changes
these features significantly. Streamflow determines the regime of a river channel and
since runoff is highly correlated with drainage basin area, river channel regime varies
systematically throughout a drainage basin (Leopold, 1994 as cited in Church, 2002).
Stream flow varies continuously in time according to recent weather and seasonal flow
history, whereas channel morphology remains relatively fixed in the short-term.
Varying flows moving through fixed channel geometry create additional thresholds of
temporal significance.
14
Figure 2.5: Channel change depends on the overlap between frequency
distributions of driving and resisting forces for different scales of morphologic response
(Buffington, 2012).
Morphological changes depend not only on exposed conditions but also on previous
historical events. Most geomorphic processes display thresholds for occurrence;
channel response will depend on the probability of a given disturbance exceeding the
response threshold, the magnitude of which also varies with channel type. Hence, the
likelihood of a given channel change will depend on the degree to which the probability
distributions of driving environmental conditions (changes in discharge, sediment
supply, vegetation) overlap with the probability distributions of process thresholds that
lead to changes in channel characteristics (i.e., overlap of driving versus resisting forces
as shown in Figure 2.5).
From the above discussion, it is evident that every morphological parameter of
alluvial river plays an important role in morphological change. However, it always
depends on the flow regime, magnitude, frequency of the event, threshold values and
bed mobility especially bed shear stress of the particular channel.
2.2.4 Change of morphological characteristics during cyclonic condition
Tropical cyclones generate extreme floods and storm surges in parts of the
tropics and subtropics between about 10° and 30° of latitude. Accounts of the resulting
high shear stress and unit stream power, enhanced stream conveyance, sediment
transport and storage and channel forms are available for a limited number of streams in
15
Australia, South Asia and the Caribbean (Gupta, 2000). These events tend to occur at an
interval of decades, rather than as 50 year or 100 year floods. An important feature of
hydrological behavior during tropical cyclones – when rivers begin to rise soon after
heavy rainfall starts, the water becomes more turbid at the same time. This is because
during rising stage of a storm hydrograph, the sediment concentration carried by a river
generally increases. In addition, powerful cyclone-induced floods are among the few
occasions when coarse bedload sediments are set in motion and transported
downstream. Exceptional amounts of suspended sediment and bedload sediment
transport reflect the ability of cyclone floods to erode and reshape various features in
the fluvial landscape (Terry, 2007). Measurement of sediment transport during a
tropical cyclone is limited due to logistical difficulties in collecting sample overbank
flows and within-channel sediment movement during extreme flood conditions.
The concentration of suspended particles varies considerably within a river’s
cross-section, depending on the erodibility of the riverbed and floodplain, the shear
stress of the flow and the height above the channel bed. At any individual study site, the
ratio of the supply of fine sediment to the total suspended load of the river will fluctuate
over the duration of a flood event, depending on bank failure, erosion patterns in the
watershed and many other factors. This can give relationships between suspended
sediments and river discharge that are hysteretic in form. It means that the peak in
suspended sediments may either precede or lag the peak in water discharge. An
investigation in the Rewa River in Fiji have been able to provide some information on
suspended sediment transport during TC Joni and TC Kina as part of an environmental
assessment of the Rewa basin by a private consulting company (NSR Environmental
Consultants, 1994 as cited in Kostaschuk et al. 2003). A total of 12 rising-stage water
samples were collected from TC Joni and 13 from TC Kina. Laboratory analysis of the
particle grain sizes showed that the samples comprised both fine materials, sometimes
called the wash load and suspended riverbed sands. The sand fractions within the
samples reflect the large shear stresses generated in the river flow during the floods
(Kostaschuk et al. 2003).The highest sediment concentration of 950 g/L was observed
early in the flood event on 1 January 1993. Later on during the main period of flood, the
samples had much lower sediment concentration. This probably reflects exhaustion of
sediment supply from channel and catchment sources. The suspended sediment
concentrations measured in the Rewa River during these tropical cyclones are some of
the highest ever recorded during river floods, in both tropical and temperate regimes
16
(Kostaschuk et al. 2003). During TC Kina, the highest clear-water boundary shear stress
of 61 N/m2 was observed for 8809 m3/s instantaneous discharge. Several studies (e.g.
Xu, 1999 as cited in Kostaschuk et al. 2003) have shown that hyper-concentrated flows
can cause extensive erosion of river beds because of enhanced boundary shear stress in
the sediment-water mixture. Sediment transport only starts when the sufficient force
(lift and drag) is provided by bed shear stress to exceed the weight force (gravity) and
any cohesive effects acting to stabilize the bed sediment ( Allen, 1985; Nielsen, 1992,
as cited in Hughes et al., 2010). This is usually expressed in terms of exceeding the
critical bed shear stress required for sediment entrainment (e.g. Sleath, Soulsby, as
cited in Hughes et al., 2010). The basic work on sediment entrainment and critical bed
shear stress can be found in Shields (1936 as cited in Buffington, 2012).
From above, it can be concluded that, during cyclonic condition, a large amount
of sediment is transported by flow in different phases of flood and storm surge event
which caused morphological changes of the channel. Bed shear stress and flow velocity
increase due to exposed cyclonic conditions which play a vital role in sediment
transport. In this study, bed shear stress is counted as the driving parameter for the
morphological changes of the channel.
2.3 Cyclone in Bangladesh
As the cyclone (Ghurnijhor-local name) is a common name in Bangladesh for its
destructive evidence in past. In this study, only cyclone generated hydrodynamic
shocks are considered to compute the changes in channel morphology. From the
historical properties of the past cyclones, duration of the hydrodynamic shock events
and their strengths are assumed in several classes which is described in detail in
chapter four.
2.3.1 General
Bangladesh is recognized as a global hotspot for tropical cyclones by UNDP
(UNDP 2004, as cited in Adams et al., 2011). Nearly every year, cyclones hit the
country's coastal regions in the early summer (April-May) or late rainy season
(October- November). Between 1877 and 1995, Bangladesh was hit by 154 cyclones,
including 43 severe cyclonic storms, 43 cyclonic storms, and 68 tropical depressions
(IMD 2010, as cited in Adams et al., 2011). On average, a severe cyclone strikes the
country every three years (GOB 2009, as cited in Adams et al., 2011). A tropical
cyclone forming in the Bay of Bengal has a lifetime of one week or longer. The height
17
of the surges is limited as high as 10 meters in the coast. When propagating into the
shallower inland coastal areas, the heights of these waves are further increased due to
shallow water effect. The frequency of a wave (surge plus tide) with a height of about
10 m is approximately once per 20 years. A storm surge of approximately once in 5
years has a height of about 7 m [surge plus tide (Khan, 1995)].
Historical cyclone tracks across Bangladesh are specified in Figure 2.6.
Since 1970, Bangladesh has experienced 36 cyclonic storms resulting in over 450,000
deaths and immeasurable economic losses. Although the polderization of the whole area
has been done since 1960’s but the frequency and high intensity cyclonic storm surge is
making this embankment vulnerable day by day. Detail description of economic loss
and devastation history can be found in World Bank report (Adams et al., 2011).
RP10 -0.9200 -0.2532 50.86 RP13 -0.6169 -0.3153 RP14 (n/a) -0.3492 Overall Model reliability over the estuarine systems 47.30
(b) (a)
43
Figure 4.1: Domain of the large model (a) and locations of the model validation (b).
4.1.6 Numerical model result of cyclone SIDR
As measured morphological data during cyclone is not available in Bangladesh,
Delft-3D model is used to simulate the condition of cyclone SIDR in Bangladesh
coast. Details of the model results can be found in Elahi et. al. (2015). This model
result is used to compare computed morphological changes during the application of the
semi-analytical model. Model generated bed shear stress, flow velocity and resultant
erosion/sedimentation at the time of landfall during cyclone SIDR in Bangladesh coast
are shown in Figure 4.2, 4.3 and 4.4 respectively.
Figure 4.2: Magnitude of bed shear stress at the time of landfall for cyclone SIDR.
Landfall location
44
Model result shows visible morphological changes at the mouth of Baleswar estuary
and Bishkhali estuary. In these locations, bed shear stresses are also found high. The
Baleswar and the Bishkhali estuaries are selected as the study site to apply the semi-
analytical model. Special phenomenon like embankment breaching is not considered
during this application.
Figure 4.3: Magnitude of flow velocity at the time of landfall for cyclone SIDR
Landfall location
45
Figure 4.4: Resultant erosion/sedimentation at the time of landfall for cyclone SIDR
4.2 Channel Classification
The relation between geometric shape factor (G) and width-depth ratio (WDR)
is described by Equation 3.20. This equation is used to generate a wide range of values
of G and WDR. The results are shown in Figure 4.5.
Affected estuarie
46
Figure 4.5: Relation between geometric shape factor G and WDR for different
dimensions of channels
A close observation of Figure 4.5 shows that values of G are clustered into four distinct
ranges. As G and channel conveyance K is related (Equation 3.6), the clusters of G can
be used to classify the channels based on channel conveyance. Accordingly, following
channel classification is proposed based on different ranges of values of G.
Table 4.3: Channel classification based on conveyance
Channel Conveyance Values of G Low 0 – 25000 Medium 25001-50000 High 50000-100000 Very High >100000
4.3 Generating the Scenarios
A number of scenarios are generated using the Delft 3D model which are used
to compute the values of the coefficients i, j, a and b. A specific scenario is constructed
by combining a specific channel property and cyclone intensity. Channels with
different conveyance capacities are used for this purpose. Channel properties with
different conveyances are shown in Table 4.4. Using these channel properties, a total of
140 scenarios are generated.
0
25000
50000
75000
100000
125000
150000
175000
200000
225000
250000
275000
0 500 1000 1500 2000 2500 3000
G
WDR
Low
Medium
Very High
High
47
Table 4.4: Channel properties that are used to generate the scenarios of model runs
Width (m)
Depth (m) Remarks 1 3 5 10 15
Conveyance of Channel Channel slope is 0.00005.
500 Low Low Low Low Low 1000 Low Low Low Medium High 3000 Low Low Medium Very High Very High 5000 Low Medium High Very High Very High
10000 Low High Very High Very High Very High 15000 Low High Very High Very High Very High 30000 Medium Very High Very High Very High Very High
From the last 40 historical cyclones, those made landfalls in Bangladesh coast
and near the coast, observed maximum wind velocity, pressure drop and cyclonic days
are specified in Table 4.5. The longest cyclonic duration was 11 days in 1993 cyclone
and the shortest duration was 2 days during the cyclones of 1995, 1996, 2007 and 2011.
Maximum wind speed of 260 km/hour and pressure drop of 9800 Pa was observed
during the cyclone of 1999. In this study, a storm surge generated by cyclonic event is
considered as a hydrodynamic shock. From the combination of observed maximum
wind speed, pressure drop and cyclone-day, hydrodynamic shocks of cyclones are
classified as Very High Strength Cyclone (VHSC), High Strength Cyclone (HSC),
Medium Strength Cyclone (MSC) and Low Strength Cyclone (LSC). These four classes
of cyclones are incorporated in Delft-3D Dashboard-Flow-Morphology model for all
the channel sections which are mentioned in the Table 4.4.
Table 4.5: Classification of different intensities of cyclones
Duration of cyclone (Days)
11 6 (SIDR) 5 2
Max wind velocity (Km/hr.)
260 213 170 83
Max Pressure drop (Pa) 400 Low 5100 Medium 6600 High 9800 Very High
4.4 Incorporating the Cyclone in Delft 3D
48
Tropical Cyclone Tool available in Delft Dashboard (DDB) developed by
Deltares (Delft Dashboard Team 2013, as cited in Laknath et al., 2014) is used for the
generation of atmospheric conditions during the cyclonic event. It generates the
surface wind and pressure fields on a moving circular spider web grid for the given
track information data, based on the Wind Enhancement Scheme (WES) following
Holland (Holland, 1980, as cited in Laknath et al., 2014). This generated spider web
grid file is used in Delft-3D Flow-Mor coupled model to assess the impact of cyclone
generated storm surge. On the basis of historical intensities of cyclones, different
strengths of cyclones are produced as mentioned in the Table 4.5. The cyclone tracks
generated by using DDB are specified in the Figure 4.6.These cyclone tracks are
assumed to make landfall in 90 degree angle to the channel cross section, as the
longitudinal landfall of cyclone produces the maximum inundation in floodplains
(Sakib et al., in preparation).
LSC MSC
HSC VHSC
49
Figure 4.6: Hypothetical tracks of different strengths of cyclones
4.5 Computing the Variables for the Semi-Analytical Model
The required variables of the semi-analytical model to compute the model
coefficients are computed from numerical model results. The variables computed
include - the bed level elevations, magnitude of bed shear stress, depth average velocity
and cross sectional area. The cross-sectional area is calculated from the bed level
elevations and the bed shear force is calculated from the bed shear stress. From the
computed time series of cross-sectional area, change of WDR is calculated.
4.5.1 Calculating the cross sectional area and the wetted perimeter
The cross-sectional area of the channel is calculated from bed level elevations of
the channel by applying the trapezoidal formula
퐴 = ℎ 푏 + (ℎ +ℎ )푏 + ⋯+ ℎ 푏 ……….
(4.2)
Where, h is the bed level elevation and b is the distance between each point. The
schematic figure of a typical cross-section is shown in Figure 4.7. The shaded area of
Figure 4.7 is the cross-sectional area of channel. The wetted perimeter of the channel
for the bankfull condition is computed by the Equation(4.3). Detail process of the cross-
sectional area and wetted perimeter calculation can be found in Rosca et al.(2015).
푃 = 푏 + ℎ + 푏 + (ℎ − ℎ ) + ⋯+ 푏 + ℎ ……….
(4.3)
50
Figure 4.7: A typical channel cross-section.
4.5.2 Calculation of bed shear force
In present study, the Van Rijn’ 84 equation for bed load transport is utilized.
This equation is also used in Delft 3D morphological computation (Delft 3D-Mor). The
formula is commonly used for situations without waves and calculates the bed load
transport rate according to the non-dimensional particle size. According to the Van
Rijn’ 84 formulations (Deltares 2011), bed shear stress is calculated as:
휏 = 휌 푓 푢 ……….
(4.4)
Where, 휌 = density of water, u = depth average velocity,푓 = friction factor which is
expressed as (Deltares 2009):
푓 = . ……….
(4.5)
51
Where, H is water depth and D90=1.5D50. From Equation 4.4 and 4.5, bed shear stress is
related with depth average velocity and water depth. The magnitude of bed shear stress
is extracted from the model results. Multiplying the bed shear stress of the channel with
the total wetted area per unit length, the bed shear force is obtained. The wetted
perimeter is calculated by using the Equation 4.3.
Figure 4.8: Variation of bed shear force for cyclones with variable strengths
From the Delft-3D morphology model, variation of bed shear force for cyclones with
variables strengths are shown in Figure 4.8. The figure shows that increased strength of
cyclones usually generates high bed shear force.
4.5.3 Generating relation between geometric shape factor and bed shear force
To calculate the required variables for the semi-analytical model, relation
between geometrical shape factor and bed shear force is established based on model
simulation results of 140 scenarios. Following the channel classification as described in
section 4.2, these relations are clustered into four different classes and are shown in
Figures 4.9 to 4.12.
8
10
12
14
0 50 100 150 200 250 300
F/Fo
Time (Hour)
Bed Shear Force
Low strength
Medium strength
High strength
Very high strength
52
Figure 4.9: Relation between geometrical shape factor and bed shear force for channel
with low conveyance
Figure 4.10: Relation between geometrical shape factor and bed shear force for channel
with medium conveyance
0.99500
1.00500
1.01500
1.02500
1.03500
1.04500
0.00 5.00 10.00 15.00 20.00 25.00
G/G
o
F/Fo
Channel type: Low conveyance channel
0.99500
1.00500
1.01500
1.02500
1.03500
1.04500
0 5 10 15 20 25
G/G
o
F/Fo
Channel type: Medium Conveyance Channel
53
Figure 4.11: Relation between geometrical shape factor and bed shear force for channel
with high conveyance
Figure 4.12: Relation between geometrical shape factor and bed shear force for channel
with very high conveyance
0.99500
1.00500
1.01500
1.02500
1.03500
1.04500
0 5 10 15 20 25
G/G
o
F/Fo
Channel type: High Conveyance Channel
0.99500
1.00500
1.01500
1.02500
1.03500
1.04500
7 8 9 10 11 12 13 14 15 16 17
G/G
o
F/Fo
Channel type: Very High Conveyance Channel
54
4.5.4 Computation of coefficients
Using the numerical model simulation results for 140 scenarios as described
above, coefficients of the semi-analytical model is computed by using Equations (3.14),
(3.15), (3.22) and (3.23). The required variables to compute the coefficients are
geometrical shape factor, shear force and WDR. As the main objective of study is to
assess the morphological changes of channel due to cyclone, the values of i and j are
determined for the cyclonic conditions. The values of i and j are computed by applying
Eq. (3.14) and Eq. (3.15) respectively for cyclonic events. The values of a and b are
computed by applying Eq. (3.22) and Eq. (3.23) respectively for a wide range of WDR
and geometrical shape factor as specified in Figure 4.5. To determine the values of
coefficients i and j, a database comprising channel conveyance and bed shear force is
generated by applying numerical model (Delft-3D) for different intensities of cyclones
on different channels. Computed values of coefficients are specified in Table 4.6. Based
on these values, coefficients are re-arranged in different ranges depending on channel
conveyance. Different ranges of coefficients are shown in Table 4.7. Schematic
representation of these ranges are shown in Figures 4.13 to 4.16.
55
Table 4.6. Computed coefficients values for different channel types
Channel conveyance type i j a b Low -1.6 x 10-5 0.999916 8.0883 5910 Medium -4.5 x 10-5 1.000132 7.3698 28127 High 246.8 x 10-5 0.980801 11.001 65290 Very high -3.5 x 10-5 0.999814 43.018 112917
Table 4.7. Ranges of coefficients for different channel conveyance types
Type i j a b Low -3.5 x 10-5< i < 246.8 x 10-5 0.999814 < j < 1.000132 7.3698 < a < 11.001 0 < b < 25000
Medium i < -3.5 x 10-5 0.999916 < j a < 8.0883 25000 < b < 50000 High 0 < i and starts from 246.8 x
10-5 j < 0.999814 8.0883 < a < 43.018 50000 < b < 100000
Very high
-4.5 x 10-5< i < -1.6 x 10-5 0.980801 < j < 0.999916 11.001 < a and starts from 43.018
b >100000
Figure 4.13: Schematic representation of ranges of coefficient i
Figure 4.14: Schematic representation of ranges of coefficient j
Figure 4.15: Schematic representation of ranges of coefficient a
56
Figure 4.16: Schematic representation of ranges of coefficient b
To get the best model result, optimum combination of coefficients values need to be
identified.
57
CHAPTER FIVE
APPLICATION OF SEMI-ANALYTICAL MODEL
5.0 Introduction
This chapter describes application of the semi-analytical model that has been
developed in chapters 3 and 4. In subsequent sections of this chapter, model
applications in a wide range of scenarios are presented.
5.1 Application of Semi-Analytical Model for Different Scenarios
The semi-analytical model is applied for different hydrodynamic scenarios.
These scenarios are generated by Delft 3D model along with Delft Dashboard. For
application of the semi-analytical model, required variable like bed shear force for the
respective hydrodynamic scenario is calculated by applying Equations (4.2), (4.3),
(4.4) and (4.5). Reference bed shear force and reference geometrical shape factor are
computed by applying Equations (3.10) and Eq. (3.21). Using the reference bed shear
force (F0) and reference geometrical shape factor (G0), and knowing the resulting bed
shear force during a cyclone (F), the change in channel morphology represented by
WDR is computed by using Eq. 3.25. Required coefficients values are used for the
respective channel class. From the prescribed ranges of coefficient values as shown in
Table 4.7, the coefficient values that are computed for different channel classes for
different scenarios are given in Table 5.1. The model result is compared with the Delft
3D model result. The simulation result of Delft 3D model is termed as ‘measured
values’ when comparing the results of the semi-analytical model with the Delft 3D
model results. Results of the semi-analytical model for each of the channel classes are
presented in Figures 5.1 to 5.4.
Table 5.1. Coefficient values of the semi-analytical model.
Channel conveyance type i j a b Low -1.6 x 10-5 0.999916 8.0883 5910 Medium -4.5 x 10-5 1.000132 7.3698 28127 High 246.8 x 10-5 0.980801 11.001 65290 Very high -3.5 x 10-5 0.999814 43.018 112917
58
Figure 5.1: Variation of WDR with time for low conveyance channel when cyclone
condition is HSC (see Section 4.3). The model shows 0.5% error.
Figure 5.2: Variation of WDR with time for medium conveyance channel when cyclone
condition is VHSC (see Section 4.3). The model shows 1% error.
A semi-analytical model is developed to compute changes of channel
morphology due to changes in channel bed shear force. The channel bed shear force is
taken as a measure of hydrodynamic forcing. In this study, hydrodynamic scenario
generated from cyclone generated storm surge is considered as a particular event which
is termed as ‘hydrodynamic shock’. Change of channel morphology is represented by
width depth ratio of the channel cross section. The semi-analytical model is
developed using the basic laws of hydraulics. The geometrical properties of the channel
representing the morphological changes are mathematically related with the
hydrodynamic forcing causing the morphological changes. During the process of model
development, four coefficients appear in the semi-analytical model which is evaluated
numerically. A set of values of the coefficients are suggested that can be used for a
wide range of channel and cyclone categories. Using the suggested model coefficient
values, the semi-analytical model is applied to compute changes in channel morphology
due to several cyclone generated hydrodynamic shocks and the results are compared
with a calibrated & validated numerical model results (Delft 3D morphology and
dashboard model). The performances of the model are found to be within ±0.5% to
±5% of the error margin.
6.2 Limitations of the Model
1. The model does not consider effects of sediment load, change of flow direction
due to flood and ebb tides, exchange of momentum between the main channel
and the floodplain.
2. Bankfull condition of the channel is assumed to compute the reference
geometrical shape factor. However, the bankfull condition alone can initiate
change in channel morphology during flood flow.
3. Information on the ‘history of shock’ is not explicitely included in the model.
77
4. Hydrodynamic shock due to cyclonic event is considered to occur as an
independent event. However, during a cyclone, embankment breach can impose
additional forcing which is not considered in the model.
5. Due to lack of field data during a cyclonic event, an extensive compariosn of the
model results with field measurements were not possible. Comparison with the
numerical model results implicitely contain numerical errors which might
artificially reduce model error.
6.3 Recommendations for Future Study
1. In the present form of the model, coefficient values are determined numerically.
A close observation of the expressions of the coefficients (Equation 3.14, 3.15,
3.22 and 3.23) shows that it is possible to derive analytical expressions of the
coefficients. This will make the model truly analytical in nature and will
improve the performance of the model.
2. As measurement of channel cross section just before and after a hydrodynamic
shock is difficult to take, it is recommended to conduct flume study in a
controlled condition. This will provide realistic data of the event which can be
used to validate the model. This will also provide further insight into the process
and will make it possible to include more drivers in the model that are
responsible to cause morphological changes during a hydrodynamic shock.
78
REFERENCES
Adams, N., Dasgupta, S., and Sarraf, M. (2011). Bangladesh Development Series: The Cost of Adapting to Extreme Weather Events in a Changing Climate. The World Bank, Paper No. 28 , 1–64.
Ahamed, S., Rahman, M. M., and Faisal, M. A. (2012). Reducing Cyclone Impacts in the Coastal Areas of Bangladesh : A Case Study of Kalapara Upazila, Journal of Bangladesh Institute of Planners, 5, 185–197.
Akter, J., Sarker, M. H., Popescu, I., and Roelvink, D. (2015). Evolution of the Bengal Delta and Its Prevailing Processes. Journal of Coastal Research. DOI: 10.2112/JCOASTRES-D-14-00232.1
Alam, S. and Matin, M. A. (2013). Application of 2D Morphological Model to Assess the Response of Karnafuli River due to Capital Dredging. Journal of Water Resources and Ocean Science, 2(3), 40. http://doi.org/10.11648/j.wros.20130203.13
Ali, A. (1999). Climate change impacts and adaptation assessment in Bangladesh. Climate Research, 12(2-3 SPEC. ISS. 6), 109–116. http://doi.org/10.3354/cr012109
Auerbach, L. W., Goodbred Jr, S. L., Mondal, D. R., Wilson, C. A., Ahmed, K. R., Roy, K., … Ackerly, B. A. (2015). Flood risk of natural and embanked landscapes on the Ganges–Brahmaputra tidal delta plain. Nature Climate Change, 5(2), 153–157. http://doi.org/10.1038/nclimate2472
Bergeron, N. E., and Abraham, A. D. (1992). Estimating shear velocity and roughness length from velocity profiles. Water Resources Research, 28(8), 2155–2158.
Biedenharn, D. S., Copeland, R. R., Thorne, C. R., Soar, P. J., and Hey, R. D. (2000). Effective Discharge Calculation : A Practical Guide. Coastal and Hydraulics Laboratory, U.S. Army Engineer Research and Development Center.
Brammer, H. (2012). The physical Geography of Bangladesh. The University Press Ltd, Dhaka.
Brouke, Mary C.; Pickup, G. (1999). Fluvial Form Variability in Arid Central Australia. (ed. by Miller, A. J., and Gupta, A.), 249-271, John Wiley & Sons Ltd.
Buffington, J. M. (2012). Changes in Channel Morphology Over Human Time Scales, in Gravel-Bed Rivers: Processes, Tools, Environments (eds M. Church, P. M. Biron and A. G. Roy), John Wiley & Sons, Ltd, Chichester, UK. doi:10.1002/9781119952497, chap. 32, pp 433-463, 2012.
Buffington, J. M., and Montgomery, D. R. (2013). Geomorphic Classification of Rivers. Treatise on Geomorphology, Vol. 9. http://doi.org/10.1016/B978-0-12-374739-6.00263-3, pp. 730-767.
Burston, J. M., Nose, T. and Tomlinson, R. (2013). Real-time numerical simulation of storm surge inundation using high-performance computing for disaster management, Queensland. MODSIM, December 2013, Adelaide.
BWDB (2008). Guidelines for River Bank Protection, Jamuna-Meghna River Erosion Mitigation Project (JMREMP).
Cañizares, R., and Irish, J. L. (2008). Simulation of storm-induced barrier island morphodynamics and flooding. Coastal Engineering, 55(12), 1089–1101. http://doi.org/10.1016/j.coastaleng.2008.04.006
Choudhury, A. M. (2007). Natural Disasters, Adaptation and the People of Bangladesh. In Biswas, S. K. (eds), Bangladesh Quarterly, 27(4), Srijoni Printers, Dhaka, Bangladesh.
Church, M. (2002). Geomorphic thresholds in riverine landscapes. Freshwater Biology, 47(4), 541–557. http://doi.org/10.1046/j.1365-2427.2002.00919.x
Condon, Andrew J., and Veeramony, J. (2013). The US Navy Coastal Surge and Inundation Prediction System (CSIPS): Making Forecasts Easier. Florida Shore and Beach Preservation Association National Conference on Beach Preservation Technology – Jacksonville, FL February 14, 2013. Retrieved from fsbpa.com/2013TechPresentations/Condon_Veeramony.pdf
Dasgupta, S., Huq, M., Khan, Z. H., Ahmed, M. M. Z., Mukherjee, N., Khan, M. F., & Pandey, K. (2014). Cyclones in a changing climate: the case of Bangladesh. Climate and Development, 6(2), 96–110. http://doi.org/10.1080/17565529.2013.868335
Deltares. (2011). Simulation of multi-dimensional hydrodynamic flows and transport phenomena, including sediments: User Manual. Version: 3.15, Revision: 14499, Deltares, The Netherlands
Ding, Y., and Wang, S. S. Y. (2008). Development and Application of a Coastal and Estuarine Morphological Process Modeling System Development and Application of a Coastal and Estuarine Morphological Process Modeling System. Journal of Coastal Research, Special issue 52, 127–140. http://dx.doi.org/10.2112/1551-5036-52.sp1.127
Ding, Y., Hsieh, T., and Yeh, K. (2015). Modeling Morphological Changes Due to Multiple Typhoons in the Danshui River Estuary. World Environmental and Water Resources Congress 2015, 1522-1531. doi: 10.1061/9780784479162.149
Elahi, M. W., Dustegir, M. M., Karim, R., Haque, A., and Rahman, M. M. (2016). Scenarios of Planform Changes of Estuarine Systems in the Ganges-Brahmaputra-Meghna Delta. Unpublished, in writing.
Elahi, M. W., Haque, A., Rahman, M. M. and Husna, N. (2015). Impacts of Coastal Floodplain Sedimentation on Net Subsidence in the Ganges-Brahmaputra-Meghna
Delta. Conference Proceedings of International Conference on Recent Innovation in Civil Engineering for Sustainable Development (IICSD-2015), 1032-1038.
Elahi, M.W.E., Sakib, M., Nihal, F., Karim, R., Haque, A., and Rahman, M. M. (2015). Changes of morphological characteristics of Bishkhali estuary due to cyclone SIDR. Presented in International Conference on Climate Change and Water Security, Military Institute of Science & Technology.
George, A. J. J., and Schneider, V. R. (1989). Guide for selecting Manning’s roughness coefficients for natural channels and flood plains. U.S. Geological Survey Water-Supply Paper 2339.
George, A. J. J., and Schneider, V. R. (1989). Guide for selecting Manning’s roughness coefficients for natural channels and flood plains. U.S. Geological Survey Water-Supply Paper 2339.
Gupta, A. (1998). Large floods as geomorphic events in the humid tropics. In: Flood Geomorphology (ed. by Baker, V. R., Kochel, R. C., and Patton, P. C.), 301-315, John Wiley & Sons, New York, USA.
Gupta, A. (2000). Hurricane floods as extreme geomorphic events. in The Hydrology-Geomorphology Interface: Rainfall, Floods, Sedimentation, Land Use (M.A. Hassan, O. Slaymaker and S. Berkowicz, Eds.), International Association of Hydrological Sciences Publication 261, Wallingford, 215-228.
Haque, A., Sumaiya, and Rahman, M. M. (2016), Flow Distribution and Sediment Transport Mechanism in the Estuarine Systems of Ganges-Brahmaputra-Meghna Delta. International Journal of Environmental Science and Development, 7(1), 22-30.
Haque, A., Sumaiya, Salehin, M., and Rahman, M. M. (2014). Reliability analysis of dynamic models. Unpublished, in writing.
Hassan, A., Martin, T. C., and Mosselman, E. (1999). Island topography mapping for the Brahmaputra-Jamuna River using remote sensing and GIS. Geological Society, London, Special Publications, 163(1), 153–161. http://doi.org/10.1144/GSL.SP.1999.163.01.13
Hickin, E. J. (1995). Chapter 4: Sediment Transport. River Geomorphology, 44(2). Wiley.
Hogan, D., and Luzi, D. S. (2010). Channel geomorphology: Fluvial forms, processes, and forest management effects. Compendium of Forest Hydrology and Geomorphology in British Columbia, 331–371.
Hossain, S. M. N. (2008). Addressing the Problems of Flood Management at Sirajganj District: A Planning Perspective on Regional – Local Context. LAP Lambert Academic Publishing.
Hu, K., Ding, P., Wang, Z., and Yang, S. (2009). A 2D/3D hydrodynamic and sediment transport model for the Yangtze Estuary, China. Journal of Marine Systems, 77(1-2), 114–136. http://doi.org/10.1016/j.jmarsys.2008.11.014
Hughes, M.G., Harris, P.T and Brooke, B.P. (2010) Seabed exposure and ecological disturbance on Australia’s continental shelf: Potential surrogates for marine biodiversity. Geoscience Australia Record 2010/43, 78pp.
Jacobson, R. B., Miller, A. J., and Smith, J. a. (1989). The role of catastrophic geomorphic events in central Appalachian landscape evolution. Geomorphology, 2, 257-284. http://doi.org/10.1016/0169-555X(89)90015-9
Kabir, T. (2014). Change of Channel Morphology at Intervened Condition after Aila. Unpublished project work under SAWA fellow program, Institute of Water and Flood Management, Dhaka, Bangladesh.
Kay, S., Caesar, J., Wolf, J., Bricheno, L., Nicholls, R. J., Saiful Islam, A. K. M., Haque, A., Pardaens, A., and Lowe, J.A. (2015). Modelling the increased frequency of extreme sea levels in the Ganges–Brahmaputra–Meghna delta due to sea level rise and other effects of climate change. Environ Sci Process Impacts, 17(7), 1311-22. doi: 10.1039/c4em00683f.
Khan, S.R. (1995). Geomorphic Characterization of cyclone hazards along the coast of Bangladesh. MSc Thesis, International Institute Geo-Information Science and Earth Observation (ITC), Enschede, The Netherlands.
Kostaschuk, R., Terry, J., and Raj, R. (2003). Suspended sediment transport during tropical-cyclone floods in Fiji. Hydrological Processes, 17(6), 1149–1164. http://doi.org/10.1002/hyp.1186
Kuiry, S. N., Ding, Y., and Wang, S. S. Y. (2014). Numerical simulations of morphological changes in barrier islands induced by storm surges and waves using a supercritical flow model. Frontiers of Structural and Civil Engineering, 8(1), 57–68. http://doi.org/10.1007/s11709-014-0235-0
Laknath, D. P. C., Ito, K., Honda, T., and Takabatake, T. (2014). Storm Surge Simulation in Nagasaki during the Passage of 2012 typhoon Sanba. Coastal Engineering Proceedings, 34, 1–10. http://doi.org/http://dx.doi.org/10.9753/icce.v34.management.4
Langbein, W. B., and Leopold, L. B. (1964). Quasi- equilibrium states in channel morphology. American Journal of Science, 262, 782–794.
Leopold, B. L., and Wolman, M. G. (1957). River Channel Patterns: Braided, Meandering, and Straight. GEOLOGICAL SURVEY PROFESSIONAL PAPER, U.S. Geological Survey, 282-B. Retrieved from https://pubs.er.usgs.gov/publication/pp282B
Lesser, G. R. (2009). An Approach to Medium-term Coastal Morphological Modelling. PhD thesis, Delft University of Technology, The Netherlands.
MES II (Meghna Estuary Study), 2001. Hydro-Morphological Dynamics of the Meghna Estuary. Dhaka, Bangladesh: MES II, 80p.
MoEF (2005). National Adaptation Programme of Action Multi‐purpose Cyclone Shelter Project (MCSP) 1993. Summary Report, Ministry of Environment and Forest, Government of the People’s Republic of Bangladesh.
Mohanty, U. C., Mohapatra, M., Sing, O. P., Bandyopadhyay, B. K., and Rathore, L. S. (2014). Monitoring and Prediction of Tropical Cyclones in the Indian Ocean and Climate Change. New York. Springer Science & Business Media, Science.
Montgomery, D.R. (1999). Process domains and the river continuum. Journal of the American Water Resources Association, 35, 397–410.
Montgomery, D.R. and Buffington, J.M. (1997). Channel-reach morphology in mountain drainage basins. Geological Society of America Bulletin, 109, 596–611.
Montgomery, D.R. and Buffington, J.M. (1997). Channel-reach morphology in mountain drainage basins. Geological Society of America Bulletin, 109, 596–611
Murty, T. S., and El-Sabh, M. I. (1992). Mitigating the effects of storm surges generated by tropical cyclones: A proposal. Natural Hazards, 6(3), 251–273. http://doi.org/10.1007/BF00129511
Nihal, F., Sakib, M., Elahi, M.W.E., Haque, A., Rahman, M., and Rimi, R. A. (2015). Sidr-like cyclones in Bangladesh coast. in Second International Conference on Environment Technology & Energy 2015, 26-31. Retrieved from http://uniqueca.com/archieves/proceedingspage/
Pattanayak, S., Mohanty, U. C., and Gopalakrishnan, S. G., (2014). Improvement in Track and Intensity Prediction of Indian Seas Tropical Cyclones with Vortex Assimilation. Monitoring and Prediction of tropical Cyclones in the Indian Ocean and Climate Change, 219-229. Springer.
Phillips, J. D. (2006). Evolutionary geomorphology: thresholds and nonlinearity in landform response to environmental change. Hydrology and Earth System Sciences Discussions, 3(2), 365–394. http://doi.org/10.5194/hessd-3-365-2006
Proudman, J. (1955). The Propagation of Tide and Surge in an Estuary. Proceedings of the Royal Society of London A: Mathematical, Physical and Engineering Sciences, 231(1184), 8–24. http://doi.org/10.1098/rspa.1955.0153
Rahman, M. M., Haque, A., Nicholls, R. J., Jisan, M. A., Nihal, F., Ahmed, I., and Lázár, A. N. (2015). Storm Surge Flooding in the Ganges-Brahmaputra-Meghna Delta : Present and Future Scenarios. E-proceedings of the 36th IAHR World Congress, 28 June- 3 July 2015, The Hague, The Netherlands.
Rahman, M.M. (1998). Studies on Deformation Process of Meandering Channels and Local Scouring around Spur-Dike-Like Structures. PhD thesis, Faculty of Engineering, Kyoto Univ., Kyoto, Japan.
Rosca, S., Bilasco, S., Petrea, D., Fodorean, I., and Vescan, I. (2015). Bankfull discharge and stream power influence on the Niraj river morphology. Carpathian Journal of Earth and Environmental Sciences, 10(1), 133–146.
Sakib, M., Nihal, F., Elahi, M.W.E., Haque, A., and Rahman, M. M. (2015). Impact of landfall location of cyclone on floodplain inundation. Unpublished, in writing.
Sánchez-Arcilla, A., García-León, M., and Gracia, V. (2014). Hydro-morphodynamic modelling in Mediterranean storms - Errors and uncertainties under sharp gradients. Natural Hazards and Earth System Sciences, 14(11), 2993–3004. http://doi.org/10.5194/nhess-14-2993-2014
Schumm, S. A. (1979). Geomorphic Thresholds: The Concept and Its Applications. Transactions of the Institute of British Geographers, Vol. 4, No. 4, 485-515. Retrieved from https://www.jstor.org/stable/622211
Schumm, S. A. (1985). Patterns of Allvial Rivers. Annual Review of Earth and Planetary Science, 13, 5–27.
Schumm, S. A., and Khan, H. R. (1972). Geological Society of America Bulletin Experimental Study of Channel Patterns. America, 83(6), 1755. http://doi.org/10.1130/0016-7606(1972)83
Stockdon, H. F., Sallenger, A. H., Holman, R. A. and Howd, P. A. (2007). A simple model for the spatially-variable coastal response to hurricanes. Marine Geology, 238(14), 1–20. http://doi.org/10.1016/j.margeo.2006.11.004
Terry, J. P. (2007). Tropical Cyclones: Climatology and Impacts in the South Pacific. Springer New York. Retrieved from https://books.google.com.bd/books?id=CUh3FXQFiDMC
Terry, J. P., Kostaschuk, R. A., and Wotling, G. (2008). Features of tropical cyclone-induced flood peaks on Grande Terre, New Caledonia. Water and Environment Journal, 22(3), 177–183. http://doi.org/10.1111/j.1747-6593.2007.00098.x
Trouw, K. J. M., Zimmermann, N., Mathys, M., Delgado, R., and Roelvink, D. (2012). Numerical Modelling of Hydrodynamics and Sediment Transport in the Surf Zone : a Sensitivity Study with Different Types of Numerical Models. Coastal Engineering Proceedings, 1(33), 23. http://doi.org/10.9753/icce.v33.sediment.23
Wolman, M. G., and Miller, J. P. (1960). Magnitude and frequency of forces in geomorphic processes. The Journal of Geology, 68(1):54-74. Retrieved from https://www.researchgate.net/publication/239062331